Академический Документы
Профессиональный Документы
Культура Документы
APOMIXIS:
From Mechanisms to
Genetic Engineering
,. CIMMYT.
•
EUrapelln Union
Institut de recherche
pour le developpement
Institut de Recherche pour le Developpernent (IRD) is a French public research institution under the
auspices of the ministers in charge of research and cooperation. For the last 50 years, it has conducted
important research in tropical and subtropical areas. With an annual budget of US$160 million, IRD
employs approximately 750 scientists (of a total of 2,300 employees), with more than 250 of them on
long-term assignments in 26 different countries.
The European Union (EU) is the result of a process of cooperation and integration that began in 1951
between six countries and today has 15 Member States and is preparing for its fifth enlargement. The
EU's mission is to organize relations both among the Member States and their 374 million citizens in a
coherent manner and on the basis of solidarity. The main objective of the European Union's Research,
Technology, and Development (RTD) program FAIR (Agriculture and Fisheries) is the promotion and
harmonization of research in the major European food and non-food production sectors of agriculture,
horticulture, forestry, fisheries, and aquaculture. The program seeks to promote links between research
and the input and processing industries, with rural economic activities, end-users, and consumers.
The International Maize and Wheat Improvement Center (CIMMYT)® (www.cimmyt.org) is an
internationally funded, nonprofit, scientific research and training organization. Headquartered in
Mexico, ClMMYT works with agricultural research institutions worldwide to improve the productivity,
profitability, and sustainability of maize and wheat systems for poor farmers in developing countries.
It is one of 16 food and environmental organizations known as the Future Harvest Centers.
© ClMMYT, IRD, EC 2001. All rights reserved. The opinions expressed in this publication are the sole
responsibility of the authors and do not necessarily reflect the views or positions of ClMMYT, IRD, or
the Commission of the EU. The designations employed in the presentation of materials in this publication
do not imply the expressions of any opinion whatsoever on the part of ClMMYT, IRD, or the EU, or
their contributory organizations concerning the legal status of any country, territory, city, or area, or of
its authorities, or concerning the delimitation of its frontiers or boundaries. The publishers encourage
fair use of this material. Proper citation is requested.
Correct citation: Savidan, Y, J.G. Carman, and T. Dresselhaus, (eds.). 2001. The Flowering of Apomixis:
From Mechanisms to Genetic Engineering. Mexico, D.E: ClMMYT, IRD, European Commission OC VI
(FAIR)
Abstract: Apomixis, the asexual reproduction of plants through seeds, has received increasing attention
as technological advances have led to a rapid increase in knowledge about cellular biology, molecular
genetics, and the mechanisms and pathways behind plant reproduction. The fourteen chapters of this
book address a wide range of theoretical and technical issues related to apomixis, as well as its potential
impact on agriculture in both the developing and developed world. The technical chapters address
aspects of two complementary research paths in the ultimate quest to produce apomictic food crop
plants. One path essentially seeks to either transfer the apomictic trait from a wild apomictic relative
into a crop plant or mutagenize sexual genes into apomictic genes in the crop plant itself. This research
is currently being conducted in important food crops such as maize, wheat, and millet, as well as
forages used for livestock, and model plant species such as Arabidopsis. The other path is rigorously
exploring apomictic and sexual mechanisms and pathways in order to provide a more complete
understanding of the overall apomixis process. This could ultimately allow scientists to target and
induce the interrelated processes of apomixis through natural or artificial means.
AGROVOC Descriptors: Apomixis; Genetic engineering; Asexual reproduction; Sexual reproduction;
Biotechnology; Chromosome translocation; Molecular genetics; Genetic variation; Genetic control;
Genomes; Tripsacum; Zea mays; progeny forms; Plant breeding; Breeding methods
Additional Keywords: ClMMYT, IRD, European Commission
AGRIS Category Codes: F30 Plant Genetics and Breeding
Dewey Decimal Classification: 631.523
ISBN: 970-648-074-9
Contents
Hi Contents
viii Tables
ix Figures
x Acknowledgments
xi Foreword
CHAPTER 1. FEEDING THE WORLD IN THE 21ST CENTURY: PLANT BREEDING, BIOTECHNOLOGY,
AND THE POTENTIAL ROLE OF ApOMIXIS
(GARY H. TOENNIESSEN)
1 Population Projections
2 Plant Breeding
3 Biotechnology
6 Potential Role of Apomixis
7 References
CHAPTER 2. ApOMIXIS AND THE MANAGEMENT OF GENETIC DIVERSITY
UULIEN BERTHAUD)
8 Introduction
9 Progeny of Apomictic Plants
11 Diversity in Wild Apomictic Populations
12 Ploidy Cycles and Organization Of Agamic Complexes
12 TaraxaCllm and Parihenium Agamic Complexes (Asteraceae)
13 Capillipedillm-Diclranthium-Bothriochloa Agamic Complex (poaceae)
13 Panicum llraximllm Agamic Complex (Poaceae)
14 Paspalum Agamic Complex (Poaceae)
14 Tripsacum Agamic Complex (Poaceae)
16 Cycles and Sexuality
16 Management of Apomictic Varieties
17 Transfer of Apomixis Gene(s) and Evolution of Landraces
20 2n + n Progeny
20 Relationship between Wild Relatives and Apomictic Varieties
21 Promoting Genetic Diversity and Release of Apomictic Varieties
22 References
CHAPTER 3. CLASSIFICATION OF ApOMICTIC MECHANISMS
(CHARLES F. CRANE)
24 Introduction
24 Types of Gametophytic Apomixis
25 Nine Types of Embryo-Sac Development
25 1) The Allium odorum-type
25 2) The Taraxacum-type
26 3) The Ixeris-type
26 4) The Blumea-type
26 5) The Elymus rectisetus-type
26 6) The Antennaria-type
26 7) The Hieracium-type
26 8) The Eragrosris-type
26 9) The Panicum-type
27 Subsequent Steps of Development
27 1) Embryos
28 2) Endospenns
28 Alternative Classifications
29 Developmental Interpretation
29 Meiotic Development of Megagametophytes
30 Ameiotic Developments of Megagametophytes
31 Subsequent Steps of Development
iv
33 Outlook
33 References
35 Appendix: Methods to Clear Angiospenn Ovules
CHAPTER 4. ULTRASTRUCTURAL ANALYSIS OF ApOMICTIC DEVELOPMENT
(TAMARA N. NAUMOVA AND JEAN-PHlLll'PE VIELLE-CAJLA.DA)
44 Introduction
45 Nucellar and Integumentary Embryony
46 Diplospory
47 Apospory
47 Differentiation of ApOSpOTOUS Initials
48 ApOSpOTOUS Megagametogenesis
48 The Cellularized Aposporous Megagametophyte
57 Parthenogenesis and Fertilization
58 Apogamety
59 Discussion
61 Future Trends
62 References
CHAPTER 5. GENFTIC ANALYSIS OF ApOMIXIS
(ROBERT T. SHERWOOD)
b4 Introduction
64 Methods
65 Chromosome Number
65 Progeny Testing
65 Embryo-Sac Cytology
66 Sectioning or Clearing Pistils to Classify Reproductive Type
66 Markers
67 Biological Tests for Parthenogenesis
67 Combined Cytological, Progeny, Biological, and Marker Testing
68 Controlled Pollination
69 Reciprocal Crossing
69 Creating Tetraploid Parents
70 Identification of Genomes and Chromosomes with Apomixis Genes
70 Testing Inheritance
70 Starting Point
70 Crossing Schemes
71 Classification and Grouping
71 Testing Genetic Models
71 Inheritance of Apomixis
71 Monopolar Apospory (Gramineae-Panicoideae)
Bipolar Apospory
75 Mitotic Diplospory
Restitutional Diplospory
76 Multicellular Archesporia
7f, Towards a Comprehensive Model of Inheritance
70 Regulation of Monopolar Apospory
77 Regulation of Diplospory
77 Regulation of Facultative Expression
78 The Lethal Gene as the Basis for Heterozygosity
79 Summary
79 References
CHAPTER 6. ApPLICATIONS OF MOLECULAR GENFTICS IN ApOMIXIS RESEARCH
(DANIEL GRIMANELLI, JOE TOHME, AND DIEGO GONzALEZ-DE-LE6N)
83 Introduction
84 Some Biological Aspects of Apomixis Worth Studying Using Molecular Genetics
84 Nonreduction followed by Parthenogenesis
85 Expression of Apomixis and Ploidy Levels
86 Endosperm Development
86 The Single-Gene Model Revisited
\i
189 Genetic Screens For Mutants Displaying Apomictic Traits In Sexual Model Systems
189 Arabidopsis Mutants with Autonomous Seed Development
191 Screen for Pseudogamous Apomixis in Cereals
192 Enhancer Detection as a Powerful Tool to Study Sexual Reproduction in Arabidopsis
192 Enhancer Detection and Gene Trap Systems
193 Generation of Transposants and Ongoing Screens
195 Identification of Developmentally Regulated Genes and Their Promoters
196 Introduction of Apomixis into Sexual Species
196 Introgression and Genetic Synthesis
199 De novo Engineering through Biotechnology
200 Field-Level Regulation of Apomictic Traits
201 Conclusions and Prospects
202 Acknowledgments
202 References
CHAPTER 13. INDUCTION OF ApOMIXIS IN SEXUAL PLANTS BY MUTAGENESIS
(UTA PRAEKELT AND ROD Scon)
212 Introduction
213 Considerations
213 Components of Apomixis
213 1. Avoidance of meiosis
213 2. Formation of aposporous embryo sacs
213 3. Parthenogenesis
214 4. Endosperm development
214 Genetic Control of Apomixis
215 How Important is Polyploidy?
215 The Problem of the Endosperm
216 Which Mutagen?
217 Some Early Work with Mutants
217 Induction of Sexuality in Apomicts
218 Mutants of Sexual Plants with Apomictic Characteristics
218 1. Meiotic mutants
219 2. Parthenogenetic mutants
219 3. Aposporous mutants
220 4. Conclusions
220 Current Approaches to the Isolation of Apomictic Mutants in Model Sexual Plants
221 Screning for Elongated siliques in the Absence of Pollination
222 Screening for Dominant Mutations in the M 1 after Pollination
225 Transposon Mutagenesis for the Isolation of Apomictic Mutants of Arabidopsis and Petunia
225 Branching Out in the Brassicas
226 Conclusions and Perspectives
227 References
CHAPTER 14. GENETIC ENGINEERING OF ApOMIXIS IN SEXUAL CROPS: A CRITICAL ASSESSMENT
OF THE ApOMIXIS TECHNOLOGY
(THOMAS DRESSELHAUS, JOHN G. CARMAN, AND YVES SAVlDAN)
229 Introduction
230 Transfer of the Apomixis Trait to Sexual Crops
230 Breeding and Introgression from Wild Relatives
231 Mutagenesis Approaches
232 Known Gene Approaches
236 Transformation and Inducible Promoter Systems
237 Main Limitations
238 Intellectual Property Rights (lPR)
239 Risk Assessment Studies
240 Summary
241 References
VIII
Tables
Figures
Acknowledgments
Foreword
One of the most urgent applications for the technology will be feeding
and raising the standard of living for the burgeoning populations of
the developing world. It is fitting that in the book's opening chapter,
Gary Toenniessen, Director of Food Security for the Rockefeller
Foundation, succinctly sets forth the magnitude and gravity of the
situation we face, and the tremendous potential apomixis holds for
helping to meet those challenges. By producing crops that produce
asexually through seeds, we can greatly hasten the development of new
higher-yielding hybrid varieties, a keystone of past productivity gains
and one that will be required to boost productivity in coming years.
With costs of development coming down, seed prices to farmers may
also decrease. Of particular import to small-holder farmers, apomixis
will allow scientists to efficiently breed varieties specifically tailored to
a multitude of niche environments, many of them situated in the most
marginal agricultural areas. Finally, because apomictic seed is self-
replicating, developing world farmers should be able to recycle seed
without losing valuable hybrid characteristics. Furthermore, the
technology could be used in such a way that farmers may be able to
better fix the traits they deem desirable within their own indigenous
varieties and land races. Needless to say, however, there is work yet to
be done.
food crops such as maize, wheat, and millet, as well as forages used for
livestock, and model plant species such as Arabidapsis. The other path is
rigorously exploring apomictic and sexual mechanisms and pathways
in order to provide a more complete understanding of the overall apomixis
process. This should ultimately allow scientists to target and induce the
interrelated processes of apomixis through natural or artificial means.
The knowledge gained through research following both approaches has
significantly accelerated advances in the field as a whole.
\~~
Professor TlftIOthy Reeyes
Director General,
The International Maize and Wheat Improvement Center (CIMMYf)
El Batan, Mexico.
March,2001
Chapter 1
Feeding the World in the 21st Century:
Plant Breeding, Biotechnology, and the
Potential Role of Apomixis
GARY H. TOENNIESSEN
surplus of food in many of the world's about 10 percent to more than 50 percent
A wealthier countries has led to a certain
complacency there about future supplies and
during the last three decades; and it is
estimated that there are at least an additional
availability. But for the vast majority of the 100 million women who wish to regulate their
world's people, who live in poorer developing fertility, but who are not now using
countries faced with growing populations and contraceptives. If effectivefamily planning and
increasing demand for food, concern rather reproductive health services were provided to
than complacency is the order of the day. For all those wishing to use them, demographers
the nations of the South, the task of feeding now predict that replacement level fertility
their future generations presents a critical and could be reached as early as 2020 and that the
formidable challenge for agriculture over the world's population would stabilize at 8-11
next half century or longer. billion people near the middle of the 21st
century (Bongaarts 1994;Lutz et al. 1997).
Population Projections
Although the task of curbing population
Fortunately, there are reasons to be optimistic
growth will be arduous, generally speaking
that an end to population growth is finally in
the agencies and institutions that provide
sight, albeit at some distance (Lutz et al. 1997).
family planning services have the technical
The rate of world population growth peaked
know-how required to achieve this goal; now
around 1970 and has been steadily declining
they are working on mobilizing the necessary
since then. As societies have moved from
financial resources and political commitments.
dependence on subsistence agriculture to
To complement this effort, the agricultural
more intensive agriculture and more modem
sector must provide the basic nutrition and
economies-in the process providing
economic growth needed to fuel the desire for
improved nutrition and health care and
smaller families and the requisite family
expanded educational opportunities to their
planning services, until the time that
girls and boys-desired family size has
replacement level fertility is reached.
dropped. A family planning revolution in the
developing world, under way now for more These encouraging population trends will,
than two decades, has lowered the average over the long term, be good for agriculture, as
number of children in a family from six to they imply that sometime during the next
three, which is reflected in a respective decline century the ever-increasing demand for
in annual population growth from 2.5 to greater food production should finally
around 1.8 percent (United Nations 1997). stabilize. The downside is that even given
Contraceptive use by women of child-bearing these positive trends, the developing world
age in developing countries has risen from will need to produce two to three times as
2 Gary H.Tonoitt. .
broadly applicable across such diverse developed for many different types of
agronomic and socio-economic conditions, agronomic and socioeconomic niches. Such
and plant breeders are just beginning to niche breeding has been successful in a few
provide improved varieties tailored to a few locations and has potential for expansion, but
of the multitude of niche environments found it is a slow process when based on
in the region. conventional breeding technology. Notably,
while there is no such thing as low input/high
Of the major developing regions, improved
output agriculture, average yields in Africa are
varieties have had the least impact in sub-
so low (often less than 1t/ha) that a doubling
Saharan Africa; food production there has
or tripling of production should be possible
lagged behind rapid population growth. In
with locally well-adapted varieties using just
Africa as a whole, more than 168 million
minimal inputs. Undoubtedly, better
people are chronically undernourished, and,
management practices would help boost
alarmingly, nearly a fourfold increase in
yields (the use of nutrient and soil-enhancing
population, from 740 million in 1996 to 2.8
crop rotations and associations looks
billion by the end of the 21st century, is now
especially promising), but over the long term,
projected (FAO 1992, 1998; Bongaarts 1994).
greater use of inputs, particularly fertilizer,
The defining characteristics of African
will be necessary.
agriculture are its complexity and
heterogeneity. Most farmers have small
Biotechnology
holdings on which they grow a variety of
Modern plant breeding, which revolutionized
crops, often intercropped wi th one another. In
agriculture in the 20th century, is now on the
each of the continent's countries, soils and
verge of significantly extending its
climate are highly diverse and variable.
technological potential. New genetic
Economic realities limit the development of
monitoring and manipulation tools, in
irrigation and other forms of yield enhancing
aggregate commonly referred to as
and risk averting infrastructure. As in much
biotechnology, are becoming available as a
of Latin America, no elite breeding lines are
result of advances in molecular and cellular
broadly applicable and improved varieties
biology. As indicated in Figure 1.1, these new
wi th specific characteristics need to be
tools are contributing to both phases of plant
Foreign genes and madmed genesl I Related wild species I I Existing crop germplasm
Trod~1 crossing
Breeding lines
Anther cuhure Genotype selection
IBS I I BAC
ISW J
Na~onal Crop Improvement Programs
'-+ ond Extension Agencies- NGOs I
I
I
_ _ _ _ .J
Farmers
even more double and triple annual cropping Under this scheme, variability would be
cycles and more extensive use of yield- generated through traditional hybridization
enhancing technologies such as hybrid seed, or any other technique noted in the
by the overwhelming majority of farmers, evolutionary phase (see Figure 1.1). The
including those with limited purchasing resulting population of plants would be
power. Hybrid seed's potential to increase grown and evaluated under local conditions,
production has been demonstrated by hybrid and the plants that performed best could be
rice in China. From 1980 to 1990, China selected and quickly developed into
increased its rice production by roughly 32.5 genetically stable superior cultivars by
million tons, or 22 percent, while decreasing incorporating the gene(s) for apomixis. For
the area planted to rice by roughly 2.2 million crops that are normally reproduced from
hectares, or six percent (FAO 1990).Yuan Long- tubers or vegeta tive cuttings, apomixis would
ping, the "father of hybrid rice" in China, enable the multiplication and dissemination
speculates that full exploitation of the heterosis of improved varieties as true seed.
available in rice could provide another 30-50
In short, apomixis has the potential to make
percent increase in yield (Yuan 1993). New
a significant contribution toward meeting
hybrid lines that are suitable for other regions
food production demand throughout the
of Asia are slowly becoming available.
developing world in the 21st century. Because
Biotechnological tools (such as genetically
of its limited profit potential, this technology
engineered male sterility systems for elite
will probably not be fully developed in the
breeding lines) and the use of molecular
private sector. Therefore, if the full potential
markers to select parental lines that combine
of apomixis as a breeding tool to help the poor
high levels of heterosis with other desirable
is to be realized, the necessary research and
characteristics can accelerate this process and
development must be undertaken by the
make the use of hybrid rice technology more
public sector international agricultural
broadly applicable. And, as reported later in
research system-and the results must
this book, progress is being made on using
remain freely available to public sector crop
apomixis as the ultimate tool for fixing
breeding programs.
heterosis in cereals, thereby making the
benefits of hybrid seed available to farmers at References
minimal cost. Bongaorts,1. 1994. Population policy options in the developing world.
5liefl(e263: 771-76.
African and Latin American farmers could also FAO. 1990. 5eIeded Indicators ofFood and Agrirolture Development in
benefit from hybrid seed that self-replicates 'he Asian·Pacific Region. Rome: RAPA Publications.
FAO 1992. The Slate of Food and Agriculture. Rome: FAO.
through apomixis, although the application of
FAOSTAT. 1998. Population Statistics. Rome: FAO.
apomixis to niche breeding could yield even wh, W., W. Sanderwn, and S. Xherbav. 1997. DouDIing of world
more consequential results. If apomixis can be population unlikely. Nature 387: 803-05.
Un~ed Nations. 1997. World Population Prospeds: The 1996 Revisioo.
introduced into staple food crops, cultivars
New York: Un~ed Nations Population Division.
that perform well under local conditions could Yuan, LP. 1993. Progress of fwo.line system in hybrid rice breeding. In
be genetically fixed early in the selection cycle. K. Murolidhoran and EA Siddig (eek.!, New frontiers inRice
Researdr. ttyderobad, Indio: Directorote of Rice Research. pp. M-
93.
Chapter 2
Apomixis and the Management of
Genetic Diversity
JULlEN BERTHAlID
Progeny of Apomictic Plants (RAPDs) have been used (Huff and Bara 1993:
In most apomicts, the apomictic mode of Barcaccia et al. 1994). It is difficult to find
reproduction is linked to pseudogamy, detailed results of progeny analyses in the
whereby the endosperm develops only after literature. Frequently, only morphological
fertilization while the embryo develops distinctions between true (maternal) and off-
parthenogenetically (Nogler 1984;see Crane, type progeny are reported.
Chap. 3). Apomixis can be split into two logical Data on Panicum maximum (Combes 1975)are
stages, which does not necessarily imply two presented in Table 2.2. In the F2 generation of
different genetic controls (see Sherwood, a P infestum x P maximum cross (T19progeny),
Chap. 5): (i) development of an embryo sac frequencies of plants produced through
without reduction and (ii) parthenogenetic sexuality and of haploid plant production
development of the embryo without were high. In one case (progeny from T19-36-
fertilization. This results in an embryo with 5),40% of a 177-plant progeny were off-types,
2n + 0 chromosomes that is genetically including seven haploids. F2 progeny from
identical to the maternal plant. However, in other crosses involving accession T19 were less
some cases, the embryo sac is reduced while variable; .only four haploids (n + 0) were found
in others fertilization occurs. It is therefore out of 1,500 observed. In P maximum, the
possible to distinguish four categories in the proportion of off-types, including 2n + nand
progeny of apomictic plants, with respective n + n was 3%,based on a total of 2,100progeny
frequencies dependent on success rates of the observed. Wecan therefore conclude apomixis
different stages (Table 2.1). in P maximum is facultative.
The four categories can be identified through For the Parthenium (Asteraceae) species,
the use of chromosome counts (or flow Guayule and Mariola, frequencies of the four
cytometry) and isozymes or molecular categories of progeny (Table 2.3) were
markers. The IRD-CIMMYf Apomixis team extracted from Powers and Rollins (1945).
used isozymes to score Tripsacum progeny Haploid plants were produced at a low rate.
(Berthaud et al. 1993). In Poa, isozymes and Most plants were produced from unfertilized
random amplified polymorphic DNA unreduced female gametes, however, the
Table 2.1 Genetk constitution of progeny from Table 2.2 Size of four categories defined in Table
apomktk plants 2.1 for two Panicum max;"",m clones (from
Combes 1975)
FetIIGIe ateiosis yes no aGIle Progtly size ...0 It'" 21t+8
Table 2.3Size of four categories defined Ih Table 2.1 for three types of progeny involving two
Partlteai"", species. Adapted from Powers and Rotms (1945)
2It+f1(%) 211+0 (%)
P. argentatum x P. argentatum 342 0 14 (4) 5 (1.5) 323 (94.51
P. argen""'" x P. incanum 888 2 48 (5.4) 78 (8.81 760 (85.61
P. incanum x P. argentmwrl 567 0 76 (13.4) 66 (11.71 425 (75.01
10 J.... IenMooI
categories 11 + nand 211 + n appeared at signifi- presented in Table 2.5.According to the table,
cant rates. Stebbins and Kodani (1944) showed it appears that within one species, but in
a frequency of occurrence of 211+11 progeny of various populations, the rate of 211 + 11
5.6%, ranging from 0.14% to 49%. Thus, progeny is variable and significant, being
apomixis in Partheniuni is largely facultative. quite high in the case of population #39 "La
Torna." Experiments are in progress to
In Tripsacuni we found an average 2.7% (11 +
analyze the effect of the environment
11) progeny, 8.1% (211 + 11), and 89.2% (211 + 0)
(flowering and pollination) on the stability of
progeny (Table 2.4). From seeds collected in
these parameters.
wild populations, we analyzed the occurrence
of 211 + 11 progeny (it is difficult to test for 11 + 11 For Dichanthiuni and Boihriochloa (Poaceae),
progeny in this situation because clones are Harlan et al. (1964) reported rates of 211+11
distributed in small niches of land and progeny from crosses between tetraploid
interpollination occurs from identical species. These combinations, however, are
genotypes, making detection of new isozyme interspecific and therefore are difficult to
patterns difficult). The frequencies for three compare with the former examples. Bashaw
wild populations of Tripsacuni we observed are et al. (1992) showed that in crosses between
Table 2.4 Estimation ofapomixis rateand categories of progeny from chromosome counts and isozyme
analyses of Tripsacum populations (Berthaud et al., unpublished data)
Species 2ft = %
Pop ID PI...t ID tested Size 72 90 108 If+ft 2lf+ft If+ft 2lf+ft 21f+0
24 143 DHBV 10 10 0 0 0 0 0 0 100
2B 163,164 MZ 20 lB 0 2 0 2 0 10 90
29 lB3 MZ 12 10 0 2 0 2 0 16.7 B3.3
37 282, 283, 772 OM 46 43 0 3 1 3 2.1 6.5 91.4
43 35B,361 OM 14 14 0 0 0 0 0 0 100
47 414 BV 12 12 0 0 3 0 25 0 75
4B 421,423 OM 39 38 0 1 0 0 0 5 95
52 497 OM 10 5 0 5 0 5 0 50 50
53 545 IT 19 17 0 2 2 2 10.5 10.5 79
54 5BB OH 15 13 0 2 0 2 0 13.3 B6.7
55 608 OH 18 12 2 4 0 6 0 33.3 66.7
59 641 OM 7 7 0 0 0 0 0 0 100
60 654, 655 OM 23 21 0 2 0 2 0 8.7 91.3
62 675 OH 14 12 0 2 0 2 0 14.3 85.7
63 689 OH 12 11 0 1 0 1 0 B.3 91.7
67 734 OH 11 11 0 0 0 0 0 0 100
71 B53 BV 14 13 0 1 0 1 0 7.1 92.8
72 B79 OM 19 18 0 1 0 1 0 5.3 94.7
74 B98 OM 21 20 0 1 5 1 23.B 4.B 71.4
83 960 OM B 8 0 0 0 0 0 0 100
B7 990 OM 5 5 0 0 0 0 0 0 100
96 1076 JL 35 31 0 4 • 1 4 2.85 11.4 B5.7
9B 1093 OHIT 23 23 0 0 0 0 0 0 100
100 11,201,121 IT 40 3B 0 2 0 2 0 5 95
Total 446 12 36 2.7 8.1 89.2
Abbreviations used: TBY=T. bravum Gray, IDH=T.dactyloides var. hispidum (Hitchc.) De Wel etHarlan, TDM=T.dartyloides var. mexicolHJm De Wet et
Harlan, TIT=T. inlermedium De Wet er Harlan, TJl=T. ialapense De Wel er Brink, ne=T. lancealalum Ruprechr ex Faurnier, TlT=T.lalifolium Hitmc.,
TM2=T. maizar Hernandel etRandalph.
Apelllil;' aod rH .......,., Gnolk DMnity 11
Table 2.6 Size ofcategories defined in Table 2.1 for two Pennisefum flaccitJum x P. mezionum crosses.
From Bashaw et al. (1992)
Progeny type Progeny size IH-O IH-ft 2IH-ft & IH-n" 2IH-ft 211+0 ('It)
Pl3l5868xPI214061 2,505 51 20 77 2428 (96.9%1
P1220606xPI214061 3,040 58 72 148 2892 (95.1 %)
• This hybrid category OOS been recognized on morphological traits. Not 011 the hybrids were analyzed cytologically.
two tetraploid species were distributed in In summary, studies of wild populations
clones of variable size (Table 2.7). The genetic demonstrate that apomixis does not produce
diversity in this population was distributed the uniformity that is often simplistically
among 54 different diploid plants, six triploid suggested. Diversity is maintained in these
clones (11 plants), and 18 tetraploid clones (83 populations. Mechanisms generating and
plants). We conclude that there are almost no maintaining this diversity may involve
"widespread" genotypes in these populations. genetic exchanges between different
Moving from one population to another, new Tripsacum types and genetic recombination as
genotypes of the same species are found. In previously described.
Mexico, populations #38 and #39 ar e about 10
km apart and both contain T. brauum and T. Ploidy Cycles and
dactyloides mexicanum. Nevertheless, their Organization of Agamic
genotypes are distinct. As a rule of thumb, the Complexes
probability of finding distinct genotypes In agamic complexes, two pools exist: one is
within a distance of 50 to 100 m is quite high. sexual diploid and the other is apomictic
In population #38, we analyzed 94 asexually polyploid (very often triploids and
reproducing triploid and tetraploid plants, tetraploids). Plants considered to be apomictic
distributed in 24 clones, i.e., four plants per present a certain amount of sexuality, at a rate
genotype on average. Ellstrand and Roose we will call "k." Authors of reviews on
(1987) observed 5.9 plants per clone in a apomixis (Nogler 1984; Asker and Jerling
literature survey of studies involving asexually 1992)conclude that facultative apomixis is the
reproducing plants. Wild populations of most common. Obligate apomixis, when
dandelion (TaraxaCllm sp, Asteraceae) and found, occurs when k = 0, and is under the
Ante/llwria sp.(Asteraceae) are comparable same genetic control as facultative apomixis.
(Lyman and Ellstrand 1984;Ford and Richards In many cases, apomixis and pseudogamy
1985;Bayer 1990). (endosperm produced after fertilization by
pollen) are found together. Pseudogamy is the
rule for apomictic Poaceae, Rosaceae, and
Ranunculaceae. In Taraxacum (Asteraceae),
Table 2.7 Distribution ofdones in TripSlIC1RfI wild fertilization is not needed for endosperm
population "La Toma" development (Ford and Richards 1985),while
ChromD- ChrOlllD- in Parihenium, which belongs to the same
Type· SOllle 110. Size Type· some no. Size family, seeds are produced only after
BV] 72 33 DM12 54 I pollination, demonstrating that fertilization is
BV2 72 3 DM13 54 2 needed for endosperm development (Powers
DMl 72 4 DM14 72 1 and Rollins 1945).
DM2 72 2 DM15 72 1
DM3 72 2 DM16 72 I Taraxacum and Parthenium Agamic
DM4 72 27 DM17 54 I Complexes (Asteraceae)
DM5 54 5 DM18 54 1 TaraxaCllm sp. is present on five continents and
DM6 90 1 DM19 54 1 about 2,000species have been described. The
DM7 108 1 DM20 72 1 base chromosome number is eight, and
DM8 72 1 DM21 72 1
DM9 72 1 DM22 72 1 diploid and tetraploid forms exist. Diploid
DMIO 72 1 DM23 72 1 forms are sexual and, depending on the
• BY = T. brarurrr, DM = Ldaetyloides nJeJl;canum species, self-incompatible or self-compatible.
ApoooiIis _ ...........' ,f Gootlf< Diwnity 13
Table 2.8Distribution ofclones according toploidy cases, the two pools are at the diploid and
level from the P. maximum colledion estabhshed in tetraploid level (group 3 of Table 2.9). In some
Cote d'lvoire (Combes 1915)
species, however, the sexually self-
Tolol 2x 3x" 4x Si" 6x'" incompatible plants are tetraploid and the self-
551 19 0 506 12 13 compatible apomictic plants are hexaploid or
• 3~ ~oiclylevel nol found in wild populOlions octoploid (group 6 of Table 2.9).
•• 5~ ond 6~ overrepresenled in this collection (from oveHollecting in
these populolionsl Some species that are sexual and self-
compatible at the tetraploid or hexaploid level
Table 2.8 shows the distribution of clones (groups 5 and 7 of Table 2.9) are not apomictic
according to their ploidy level in a collection at higher ploidy levels. In other species,
established in Cote d'Ivoire. triploids are often apomictic and are found in
Triploid plants have been experimentally species with sexual diploids and apomictic
obtained from hexaploids (11 + 0 progeny) as tetraploids.
well as from diploid x tetraploid crosses. As with previously cited agamic complexes,
(Poly-) haploids also have been sexual forms are found at the lowest ploidy
experimentally obtained from tetraploids, level and apomictic forms at the other levels.
and the resultant plants have been either However, in this example, the relationship
sexual or sterile (potentially apomictic as includes the incompatibility system.
shown by embryo sac analyses). These Apomictic plants are self-eompatible and the
findings led Savidan and Pernes (1982) to corresponding sexual plants are self-
propose an evolutionary scheme based on incompatible. Experiments should be
ploidy cycles involving di-tetra-haploid levels conducted to determine whether this also
as in the Dichanthium complex. The change occurs in other agamic complexes.
from diploid to tetraploid is realized through
Tripsacum Agamic Complex (Poaceae)
211 + 11 hybridization with pollen from
The Tripsacllm genus is restricted to the New
tetraploid plants. In this system, sexuality is
World, from 42<'N to 24"5. Its center of diversity
maintained at the diploid level. Contact
(or origin) is located in Mexico and Guatemala,
between diploid and tetraploid plants allows
and 11of the 16 species described for the genus
genetic exchange between these pools
are found in this region. These 11species show
(compartments) and creation of sexual
different ploidy levels both within and among
tetraploid plants, allowing the release of new
themselves. The collection the team assembled
genetic diversity at the tetraploid level.
from Mexico displayed the following
Paspalum Agamic Complex (Poaceae) distribution (unit = one ploidy level of one
The center of diversity for the genus Paspalum species in one population): diploids. 16.4%;
is in South America. Studies conducted by triploids, 7.9%; tetraploids, 72%; penta- and
Quarin (1992), Norrman et al. (1989) and hexaploids, 3.7%.
collaborators at the Instituto de Botanica del
When compared to other agamic complexes,
Nordeste. Corrientes, Argentina (IBONE)
a high frequency of triploid plants in the
show that many species in this genus have
Tripsacum complex was observed. These wild
genetic pools at two or more ploidy levels
triploid plants are apomictic, produce fertile
(Table 2.9). In the pool with the lowest ploidy
pollen, and set good seed. All of the natural
level, plants are sexual and self-incompatible,
polyploids we observed were apomictic
while in pools with higher ploidy levels, they
(Leblanc et al. 1995; and unpublished data).
are apomictic and self-compatible. In many
""xis . . tk ........,., GeMIic Dhenity 15
Diploids are sexual, and progeny with 2/1 + /1 residual sexuality exists in apomicts, which
chromosomes from apomictic plants occur at permits production of n + 11 progeny. This
a significant frequency (Tables 2.4 and 2.5). sexuality favors creation of new diversity at
Through this mechanism, many hexaploids the tetraploid level by allowing crosses
were produced experimentally or detected in between apomictic plants.
seeds collected from a wild population.
Our model (Figure 2.2) suggests that in the
Natural hexaploid plants in wild populations
Tripsacum agamic complex, sexuality fosters
were observed at a lower frequency than in
two stages: (i) a change from tetraploidy to
the seed progeny we analyzed.
Management of
Apomictic Varieties
Two types of apomictic varieties can be
distinguished: forage varieties, which are
already released as apomictic varieties, and
Figure 2.2 Evolution ofploidy levels in Tripsacum apomictic varieties of crops such as maize and
from fertilization offemale gamete (n or 2n) by a pearl millet, which may be released in the
male gamete (n) from 21,4x or 6Jt plants or near future.
parthenogenetic development ofegg eeA (n+O).
In the breeding of apomictic forage grasses, would involve production of triploid plants,
sexuality is involved at different steps and which are usually male sterile; so
permits genetic recombination (Valleand Miles dissemination through triploids should be
1992;see Valle and Miles, Chap. 10). Released negligible. However, in agamic complexes,
varieties are apomictic and have been apomixis seldom occurs at the diploid level.
distributed mainly outside their centers of Some mechanism may suppress the expression
diversity. In this instance, breeding activity is of apomixis or impeach transmission to the
generating new genetic diversity. diploid level. In the pearl millet program, there
is no clear evidence that apomixis can be
Because projects are now underway to transfer
expressed at the diploid level. In contrast, a
apomixis to pearl millet, maize, wheat, and
few BC2 diploid-like hybrids in the maize-
rice, we must consider the consequences of
Tripsacum program were found to express
apomixis on the diversity management of
apomixis (Leblanc et al. 1996).These plants are
landraces and that apomixis drastically
2n = 28 with x = 10 from maize and x = 18 from
reduces the recombination rate. It is important
Tripsacum. Furthermore, triploid Tripsacum are
to remember that these landraces and their
male and female fertile. Thus, tetraploid
wild ancestors represent our current reservoir
apomictic varieties of maize will probably not
of genetic diversity. Thought should also be
restrict diffusion of apomixis gene(s) to other
given to conserving the diversity of wild
maize lines or its wild ancestor, teosinte.
ancestors that grow near fields planted with
Therefore, the models of diffusion of apomixis
apomictic varieties, which could be recipients
discussed below are based on diploidy.
of apomixis genes through naturally occurring
gene flow. Apomixis fixes heterosis, thereby presenting
two options for its use: (i) to produce apomictic
Projects to transfer apomixis to pearl millet and
F, hybrids through breeding programs and
maize have reached an intermediate stage:
release them to farmers as end products; and
advanced generations of interspecific hybrids
(ii) to release to farmers apomictic varieties that
between apomictic forms and cultivated
would be used to transfer (diffuse) gene(s) to
species have been produced that retain the
landraces, which would eventually become
apomictic trait. In the case of rice, possible
apomictic. In the latter case, breeding for
sources of apomixis are yet to be identified. For
apomixis would be a local activi ty. In fact, these
wheat, F l and BCI hybrids between Triticum
two options are complementary and related
and Elymus have been produced (Peel et al.
as they pertain to the diffusion of apomixis
1997;Savidan et al., Chap. 11).Pearl millet and
gene(s). Fl apomictic hybrids could be released
maize are allogamous crops and so methods
in an area where landraces and wild relatives
must be developed to maintain genetically
still exist. The transfer of the gene to these
adaptative processes once this new mode of
landraces and wild relatives will depend on
reproduction is introduced. In its current
the parameters cited above in option 2.
design, the Penniseium project considers the
creation of tetraploid apomictic varieties of Transfer of Apomixis Gene(s) and
pearl millet (Dujardin and Hanna 1989).Upon Evolution of Landraces
release, the distinct ploidy levels of currently We deduce from Sherwood (see Chap. 5), that
cultivated millet and the tetraploid apomictic apomixis is probably initiated by one
new varieties will act as a genetic barrier dominant gene (see also Valle and Savidan
between them. Dissemination of apomixis 1996). The active A allele of this "apomixis
gene(s) from the tetraploid to the diploid level gene" would be found mostly in the
heterozygous condition (Aa). The Equilibrium is reached for R = 1, the
homozygous stage (AA) has been considered population being entirely sexual, or for R = 0,
lethal in some cases (Nogler 1984). the population being completely apomictic.
Nevertheless, in discussing apomixis transfer, This model is identical to the model proposed
we will consider three models: (i) apomixis is by Fisher (1941) for autogamy. In fact,
active as a dominant trait, either heterozygous apomictic plants self-reproduce, however they
or homozygous (Aa or AA) with the recessive simultaneously release pollen with the
homozygote (aa) being sexual; (ii) apomixis is dominant allele to the sexual plant forms;
active only as a heterozygote (Aa), with the consequently, a portion of the progeny of
recessive homozygote (aa) being sexual; and sexual forms becomes apomictic.
(iii) apomixis is only expressed as a recessive
If we take into account a rate of residual
homozygote (ss), while sS and SS are sexual.
sexuality, k, the variation in frequency for A
We will also consider a residual rate of
allele becomes Pn + 1 + Qn + 1 = (Pn + Qn)(l +1/
sexuality, k, in apomictic plants, withO < k < 1.
2(1-k)R n) (Pernes 1971).
Simple models of population genetics predict,
The change in frequency of allele A from
in the absence of selection, the diffusion of the
generation n to generation n + 1 is a function
apomixis gene (Pernes 1971; Marshall and
of Rn, the frequency of the recessive allele, and
Brown 1981). According to the models, it is
a function of k. A zero value for k (obligate
possible for the apomixis gene to transfer to
apomixis) maximizes the frequency of A, while
land races, such as maize or pearl millet, and
higher values of k reduce the frequency of A.
to inadvertently move to wild relatives (Pernes
This variation would be zero if k = 1, i.e., when
1971; van Dijk and van Damme 2000).
all plants are sexual with either the A or a allele.
In model 1, there is one dominant allele for
In this model, we assume random mating of
apomixis and three categories of genotypes at
gametes. Transfer would be favored if an
generation n: AA (apomictic) at a frequency of
apomictic variety, homozygous for A, were
Pn. Aa (apomictic) at a frequency of 2Qn, and
interplanted with the variety (land race) to be
aa (sexual) at a frequency of Rn' Gametes for
modified. In the case of maize, by detasselling
generation n+1 are distributed according to the
and harvesting only the land race, only
following frequencies: male gametes A have a
heterozygous progeny would be produced,
frequency of Pn + Qn and gametes a have a
These new plants would be apomictic and
frequency of Q n + Rn; female gametes A have
genetically fixed. Their ability to evolve would
a frequency of 0, gametes a, a frequency of Rn'
rely on the rate of residual sexuality, k. A
gametes AA, a frequency of P n. and gametes
proportion k of the apomictic forms can be
Aa a frequency of 2Qn'
fertilized by pollen from other sources.
Three genotypes will appear at generation n + Moreover, pollen from the first generation of
1 with the following frequencies (random apomictic forms can be used to pollinate the
mating of gametes): AA at a frequency of landrace. After several cycles of s uch
Pn_ 1 = Pn' Aa at a frequency of 2Qn + 1 = 2Q n + backcrossing, the new variety will be identical
Rn(Pn + Qn)' and aa at a frequency of Rn + 1 = to the land race except that it carries the
Rn(R n + Qn)· apomixis gene. Evolution in these "new"
land races will depend on the rate of residual
With Pn + 2Qn + Rn = 1, we obtain Q n = 1/2(1-
sexuality that is retained at the end of the
Pn- Rn) and the recurrence relation:
transfer process.
Ape.,;.is .... IN. . . . .' 0/ Gnoli< Ilhonity 19
markers to retain the aallele, (il) the production filled kernels (with 2n + 11 embryo) at
of near isogenic lines through backcrossing harvest time. There is a potential loss of
with the landrace, and (iii) the selfing of production due to the presence of these
isogenic, heterozygous (Aa) lines to produce 2n + 11 embryos, but these kernels would
aa apomicts. not be selected as seed for the next
generation.
2n + n Progeny
2. Endosperm development is not affected
In Tripsacum, we saw an average of 10% of
by a ratio of embryo ploidy to endosperm
progeny come from 2n + n hybridization; in ploidy different from 2:3 (or 2:5). In this
some samples, this rate rises to 35%. Crosses case, kernels with 2n + n embryos would
between apomictic species of Penniseium also go undetected and could be used as seed
produced this type of progeny (Bashaw et al. for the next generation. Apomictic plants
1992). These forms are less frequent in other obtained from such embryos are triploid:
species, such as Panicum maximllm. If this trait they may produce normal seeds but the
is inherited during the transfer of apomixis, pollen could be sterile, which could limit
what behavior can be expected from cultivated field prod uction. If the pollen is still
apomictic forms? fertile, as noted with triploid Tripsacum,
no loss in production should be detected.
The transfer projects now underway consider
However, ploidy buildup will occur, and
a type of apomixis linked to pseudogamy.
many different ploidy levels will be
Once apomictic varieties are produced, most
stored in the same variety. This ploidy
probably they will be also pseudogamous. In
buildup could raise chromosome
this case, we are concerned with the ratio numbers to levels far above the optimum
between embryo ploidy and endosperm for productivity, potentially resulting in
ploidy, as it has been often reported that a ratio lower production.
different from 2:3 (or 2:5) would introduce In nature, 211 + n progeny production is a
some developmental incompatibility at the strategy that takes advantage of genetic
seed level and a loss in productivity recombination, as these plants would give rise,
(endosperm development also depends on after meiosis, to some haploid progeny by
matemal:paternal genome ratio; see Chap. 6, parthenogenetic development of reduced
11,12, and 13). However, for the Tripsacum, we embryo sacs. In the case of an apomictic crop,
observed that triploid plants produce seeds it is a trait that should be reduced or
even when their pollen environment comes eliminated.
mostly from tetraploid plants. In this case, the
ploidy ratio between embryo and endosperm Relationship between Wild Relatives and
Apomictic Varieties
is 3:8. The 2/1 + /I progeny we detected were
For the purpose of discussing the relationship
from normal seeds with normally developed
between wild relatives and apomictic varieties,
endosperm. In Tripsacum, the 2:3 ratio (or 2:5)
we will use the maize-teosinte model,
between embryo ploidy and endosperm
however, it is our belief that it can be
ploidy does not appear to be necessary for seed
extrapolated to pearl millet in instances where
filling. In general terms, we have two
wild relatives are still in contact with cultivated
hypotheses to consider:
plants. Teosinte is only found in Mexico and
1. Endosperm development is deficient Guatemala. Relationships between wild
when the ratio of embryo ploidy to relatives and maize are not identical over the
endosperm ploidy is different from 2:3 distribution area of teosinte. The variety
(or 2:5). In this case, ears display poorly
........ IN. . . . . .' 01 ..... DhonlIy 21
paroiglumis may be found in southwest Mexico hybrids and the wild relatives. As a lack of
and is considered to be a very wild form, with synchronization between the two types of
almost no link to modem maize. In the states plants is anticipated, the gene flow between
of Michoacan and Mexico, teosinte should be them should be minimal. These observations
considered a weed. An incompatibility system deviate considerably from the assumptions
exhibited by these weedy teosintes, which posited in the model in which apomictic plants
efficiently controls gene flow from maize to are expected to engage if! pollination in
teosinte, has been detected and analyzed proportion to their frequency in the
(Kermicle and Alien 1990). Moreover, as population. Moreover, in the long run, the
described by Wilkes (1967), teosintes generally apomictic intermediate forms should have a
have a flowering period that is distinct from lower fitness than the sexual forms, because
maize. These mechanisms limit gene flow the latter can take advantage of more new
between this wild relative and maize. recombinations and adapt faster to
environmental changes. As noted earlier, a
If we use model 1 to explain the transfer of
stable polymorphism between sexual and
apomixis from apomictic plants to landraces,
apomictic forms is possible when fitness values
we can envisage the following process. The
of the two forms reach a certain ratio. Wehave
first generation hybrid between teosinte
also observed that the speed of apomixis
(sexual, aa) and apomictic maize (AA) would
diffusion is a function of the rate of residual
be apomictic (Aa), and BC!plants with teosinte
sexuality-a high level of residual sexuality
as female would produce Aa (apomictic) and
will slow apomixis diffusion.
aa (sexual) progeny. At each generation, the
apomictic forms are fixed but they still Promoting Genetic Diversity and the
participate in the next generation from sexual Release of Apomictic Varieties
plants through their pollen, which can transfer We base our models for apomixis diffusion on
the apomixis allele to sexual plants. Therefore, the hypothesis that this mode of reproduction
a portion of each generation's progeny is under a simple genetic control. Current
becomes apomictic. We can then deduce that knowledge about the mechanisms underlying
the apomictic allele will diffuse into the wild apomixis, however, is very incomplete,
population. However, the assumptions made especially regarding the expression of an
to simplify the model may not prove accurate apomixis gene in a new genetic background,
when applied to the relationship between as would be the case with a Tripsacum apomixis
cultivated plants and wild relatives. gene transferred into a maize background. If
genetic control of apomixis in landraces and
Cultivated maize and its wild teosinte relatives
new varieties involves several genes or a major
are, morphologically, widely distinct.
gene and modifiers, the dynamics of diffusion
Apomictic maize x teosinte F! hybrids will be
will be more difficult to describe and
apomictic and will breed true. Sexual maize x
transformation of current varieties to apomictic
teosinte F!s are known to have a low fitness
varieties would have to be carried out by
due to their intermediate morphology and
professional breeders. In this instance,
adaptation, and they are easily recognized
apomixis could be used as a genetic fixation
morphologically. When they grow in a field,
tool and new varieties with a complex genetic
they are not harvested. However, if the hybrid
structure could be created and released. Such
is apomictic, its pollen will transmit the A allele
varieties would contribute to the maintenance
at a rate of 50%.Pollination efficiency depends
of diversity at the farmer's field level.
on synchronization between flowering of these
22 J.... IertMoII
- - . 1997. Asynchrol1OUl expression of duplicote genes in Lymon, J.e., and N.e. Ellstrond. 1984. Clonal diversily in Taraxacum
ongiosperms may COUle oponixis, bispory, letrllSpOly, and affldnale (Asteroceoe), on oponid. Heredity 53: 1-10.
polyembryony. BioJogicol Journal afthe Unneon Society 61: 51-94. Morsholl, D.R., and A.H.D. Brown. 1981. The evolution of oponixis.
Combes, D. 1975. PoIymorphiwe etmodes de reproduction dol1llo Heredity47: 1-15.
section des Maximoe du genre Panicvm (Groninees) en Alrique. Mogie, M., and H. Ford. 1988. SexuoJ and asexual TaraxOMl species.
Memoires ORSTOM (Paris) #77. Biological Journal afthe Unneos Society 35: 165-68.
Combes, D., and J.Pemes 1970. Yorio!iol1l dol1lles nambres Nokojima, K., T. Kornotsu, N. Mochizuki, and 5. Suzuki. J979.lsoIotion of
chromosoniques de I'r1nicum maxilllJllJ en relation ove< lemode de diploid and tetraploid sexual ~onts in Guinea gross (I'r1nicvm
reproduction. (R. Acad. Sei. Paris 270: 782-85. maximum). Japan Jaurnal afBreeding. 29: 228-38.
De Wet, J.MJ. 1968. Diploid-tetroploid-hoploid cycles and lhe origin of Nogler, GA. 1984. Gomelophytic apomixis. In 8.M. John (ed.),
voriabitlly in Dichonthium. Evolution 22: 394·397. Embryalogy afAngiosperms. New York: Springer-Yerlog. Pp. 475-
De Wet, J.MJ., and J.R. Horlon. 1970. Apomixis, poly~oidy, and 518.
speOotion in Dichonthium. Evolution 24: 270-77. Norrmann, GA, (L Quarin, and B.L 8ursan. 1989. Cytogenetics and
Dujordin, M., and W.W. Honno 1989. Developing opomiclic pearl nillel- reprodudive behovior of different chromosome roces in six I'ospolum
choroderizotion of 0 8~ plont. Journol afGenetics and Breeding 43: species. Journal afHfi(edity 80: 24-28.
145-50. Peel, M.D., J.G. Carman, l.W. [ju, and R.R.e. Wong 1997. Meiotic
8ktrond, N.e., and M. Roose. 1987. Ponems of genolypK diversity in OnomotleS in hybrids between wheat and opomiclic f1ymus rectise/IJS
clonal ~tlpet:ies. American JOIIIlDI afBoIDny 74: 123-31. (Nees in Lehm) A. Liive and Comar. Crap Science 37: 717-23.
FISher, RA. 1941. Average excess and overoge effet:t of 0 gene Pernes, J.1971. Elude du mode de reproduction: oponUie facultative du
substitution. Ann. Eugen. 11: 53--63. poinl de vue de 10 genelique des populolions. Trovuux et documen~
Ford, H., and AJ. Richords. 1985. Isozyme vuriolion within and between de I'ORSTOM (Paris) #9.
Taraxacum ogomospedes in 0 single locality. Heredity 55: 289-91. - - . 1975. Organisation evolutive d'un groupe agonique: 10 sedion
Grimaneltl, D., O. Leblonc, E. Espinolo, E. Perolli, D. Gonzlilez de LeOn, des Maximae du genre Panicum. Memoires omON (Paris) #75.
and Y. Sovidon. 1998. Non-Mendelian Ironsmission of opomixis in Powe~, L1945. Fertilizotion without redution in guoyule (I'r1rthenium
maize- Trips«um hybrids coUled by 0 transmission rotio dislortion. argentafum Gray) and 0 hypothesis os 10!he evolution of opomixis
Heredity 80: 40-47. and polyploidy. Genetics 30: 323-%.
Honna, w., M. Dujdn,POzios·Akins, E. Lubbe~, ond LAl1hur. 1993. Powe~, L, and R.( RoIlil1l. 1945. Reproduction and poIinotion studies on
Reproduction, cytology, and fertility of pearl millet x Pennisetum guoyule, Parthenium argentotum Gray and P. inrooom H.B.K. JoumoI
squamuJatum 8C4 plants. Journol afHeredity. 84: 21l--16. afthe Ameriron Society afAgronomy: 96-112.
Horlon, J.R., M.H. Brooks, D.S. Borgoonlcar, and J.MJ. de Wet. 1964. Quorin, (L 1992. The nature of apomixis and its origin in Ponicoid
Noture and inherilonce of oponixis in Bothriochloa and Dichonthium. grosses. Apomixis Newslener 5: 7-15.
Batanicol Gazene 125: 41.-.6. Rutishouser, A. 1948. f'5eudogomie und polymorphie inder Gottung
Horlon, H., ond J.M.J. De Wet. 1975.lln 0. WIIlge and 0 prayer: !he Paten/ilia. Arch. Julilll KJaIll-Stih VererbungsfrxJdl23: 267.-.24.
or~ins of poIyloidy. Bot. Rev. 41: 361-69. Somon, Y., and J.Pernes. 1982. Di~oid-tetroploid-dihoploid cycles and
Huff, D.R., and J.M. Boro. 1993. Deternining genetic ~in of aberrant the evolution of I'r1nicum maxitnum lecq, Evolution 36: 596-600.
progeny fram fowhotive opomidic Kentudcy b1uegross using 0 Slebbins, G.L, and M. Kodoni. 1944. Chromosomal vuriotion in GuoyuIe
combinotion of flow cytometry and silver-stained RAPD marke~. and Moriolo. Journal afHeredity. pp. 16l--72.
Theoretical and Ap,Jied Genetics 87: 201-208. Yolle, e.8. do, and J.W. Miles. 1992. 8reeding of oponictic spe<ies.
Kermicle, J1,and J.O. Alen. 1990. Cross~ncompo1ibility between maize Apomixis NewsJener 5: 37.-.7.
and leosinle. Maydiro 35: 399-408. Yolle, (B. do, and Y. Sovidon 1996. Genetics, cytogenetics, and
Leblonc, 0.,M.D. Peel, J.G. Corman, ond y. Somon. 1995. reproductive biology of Brodriaria. In J.W. Miles, B.L MaUlS, ond e.B.
Megasporogenesis and megagometogenesis in several TripstJcum do Yolle leds.l, Brachiaria: Biology, Agronomy, and /flfJlavement.
species (POlKeoe). Ameriron Journol afBotany 82: 57-63. Coli, Colombia: C1AT-EMBRAPA. pp.147-63.
Leblanc, 0., D. GtimaneBi, N.I. Faridi, J.8erthoud, and Y. Somon. 1996. van Dijk, P, and J.vun Domme. 2000. Aponixis led1no1ogy and the
Reproduc!ive behovior on maize- TripstJcum poIyhapIoid plonts: parodox of sex. Trends in /'/om Science 5(2): 81-84.
imp!icoliol1llor the trol1lfer of apomixis into maize. Journal af Wilkes, H.G. 1967. Teosinte: the cIosesl relotive of maize. Ph.D.
Heredity87: 108-11. dissertation. 8ussey InsliMe, Hurvord University, CoImridge,
MossachUleIlS.
Chapter 3
Classification of Apomictic Mechanisms
CHARLES F. CRANE
-',~ ',~;
? ? ?
gives rise to the mature
8-nucleate embryo sac Elymus 3 - _..
.... ,.), /~,
after three rounds of
teJ T
I;
\.
:,t.::::: _\!J ~~J
? ? ?
mitotic division. Some- :.--\
times a second round of Antennoria r.: I
~~.'
endomitosis precedes
'.1
~~
r_
3) The Ixeris-type. The Ixeris-type follows the sacs has not been thoroughly investigated, but
Taraxacum-type through the formation of a the chalazal member of dyads typically
meiotic restitution nucleus. The next nuclear undergoes three rounds of mitosis and forms
division is not followed by formation of a cell the mature, 8-nucleate embryo sac. The
wall, nor is the division after it. Cell walls form micropylar nucleus ofhemidyads sometimes
only as the embryo sac matures after the third degenerates as if it had been completely cut
round of divisions. All daughter nuclei from off.
the restitution nucleus contribute to the 6) The Antennaria-type. In the Antennaria-
mature, 8-nucleate embryo sac. type, there are no meiotic stages. The
4) The Blumea-type. In the Blumea-type, a megasporocyte undergoes three rounds of
more or less mitotic division of the mitosis, typically after an initial period of
megasporocyte yields a dyad of 2/1 enlargement and vacuolation. The mature
megaspores. The chalazal megaspore gives rise embryo sac has eight nuclei, which are
to the mature, 8-nucleate embryo sac. The arranged as in the conventional Polygonum-
development is identical to the Taraxacum- type.
type after the restitution nucleus has formed. 7) The Hieracium-type. In the Hieracium-
Therefore, existence of the Blumea-type must type, one to several nucellar cells enlarge,
be demonstrated by the absence of the become vacuolate, and go through three
restitutional stages of the Taraxacum-type; this rounds of mitosis, resulting ideally in a
proof requires careful, thorough sampling and conventionally organized 8-nucleate embryo
ordering of all the meiotic stages. sac. The megasporocyte undergoes meiosis in
5) The Elymus rectisetus-type. In Elumus some species and degenerates in others. In
rectisetus, three interrelated developments some cases, reduced and unreduced embryo
occur that superficially resemble the Blumea- sacs can coexist in the same nucellus. The
type. The megasporocyte nucleus enlarges and ability to form multiple embryo sacs in the
is deformed in all three types. The nucleus same ovule increases the frequency of
appears to be under tension, as evidenced by polyembryony.
parallel alignment of its chromatin, and by its 8) The Eragrostis-type. In the Eragrostis-type,
chalazally pointed or occasionally wasp- there are no meiotic stages. The
waisted shape. Triads and dyads with unequal megasporocyte undergoes only two rounds
nuclei indicate amitotic division in a very low of mitotic division, leading to a 4-nucleate
percentage of megasporocytes. The first visible embryo sac with an egg, two synergids, and
prophase follows the nuclear deformation one polar nucleus, or alternatively, to an egg,
episode and leads to what could be considered, one synergid, and two polar nuclei. The
more or less, a mitotic division. Here the three original description of the type also included
subtypes diverge: this division can result in a bipolar sacs with more than four nuclei, which
dyad of completely separated cells, a at tha t time were interpreted as modifica tions
hemidyad of incompletely separated cells, or of the Antennaria-type (Streetman 1%3). Such
a binucleate embryo sac. All three behaviors sacs may instead represent facultative
can occur on the same individual. The occurrence of the Polygonum-type.
chalazal-end nucleus in all three cases often 9) The Panicum-type. In the Panicum-type,
undergoes a milder form of the deformation one to 25 nucellar cells enlarge, increase in
seen in the megasporocyte. The fate of vacuolation, and ideally divide twice
hemidyads and directly binucleate embryo mitotically, although some divide only once
and others not at all. The mature embryo sac Asker and ]erling (1992), but the case is
has an egg, two synergids, and a polar nucleus, scarcely closed. The evolution of automixis is
or else, an egg, one synergid, and two polar implausible in angiosperms, because
nuclei. The megasporocyte can undergo autogamy evolves easily in most groups and
meiosis or degenerate, depending on the has the same genetic result over the course of
species. The potential for polyembryony is generations.
high in this type.
Subsequent Steps of Development
Most of these developments were discussed
1) Embryos. Apomictic embryo development
by Battaglia (1963), who also discussed various
from egg cells has been summarized in earlier
aberrations, and later by Nogler (1984) and
research, e.g., Asker and Jerling (1fi}92) and
Asker and Jerling (1992). Descriptions were
Battaglia (1963). There are three major types:
first published for the Antennaria-type in 1898
pseudogamy, semigamy (= hemigamy), and
OueI1898), the Taraxacum-type in 1904 (luel
autonomous parthenogenesis. There is no
1904),the Hieracium-type in 1907(Rosenberg
tradition for naming these last three
1907),the Ixeris-type in 1919(Holmgren 1919;
developments for the type genera in which
Okabe 1932),the Allium odorum-type in 1951
they occur. The developments differ in the
(Hakansson and Levan 1951, 1957), the
order in which egg and primary
Panicum-type in 1954 (Warmke 1954), the
endospermatic nuclei divide, the ability of the
Eragrostis-type in 1963(Streetman 1963;Voigt
central-cell nuclei to divide without
and Bashaw 1972),and the Elymus-syndrome
fertilization, the number of polar nuclei that
in 1956 (Hair 1956), with significant
contribute to the endosperrn, and the ability
amendments in 1987 (Crane and Carman
of the egg cell to undergo plasmogamy.
1987). The Blumea-type, described for Erigeron
Pseudogamy is traditionally defined as
ramoslIs by Holmgren (1919), was later
asexual seed set that requires pollination and
generically refuted (Fagerlind 1947)and finally
does not involve adventitious embryos.
rehabilitated for Blumea eriantha
However, a more restricted definition seems
(Chennaveeraiah and PatiI1971).
preferable: pseudogamy is seed set through
A number of previous classifications, e.g., fertilization of the central cell, but not the egg,
Asker and Jerling (1992), have recognized in the absence of adventitious embryos.
automixis, which is the fusion of nuclei in the
5emigamy is seed set with full fertilization of
same gametophyte to produce a 2n,
the central cell and only plasmogamy in the
homozygous egg cell. Automixis has been
egg, where the sperm nucleus can be walled
purported in several species oi Axonopus
off,or divide one to several times, or contribute
(Gledhill1967), but the evidence can easily be
equally to chimeric or twinned embryos. The
reinterpreted as early degeneration of one of
last of these behaviors is not found routinely
the synergids in the Polygonurn-type before
in recurrent apomicts, but is typical of the Se
fertilization. Like gametophytic selfing in
mutant in cotton (Turcotte and Feaster 1969),
ferns, the mechanism would produce
in which the sectors can be easily visualized
individuals that are completely homozygous
with leaf-color mutations. In Se cotton,
within genomes. There is no genetic evidence
complex chimeras with maternal haploid,
for or against apomixis in the affected
paternal haploid, and hybrid sectors are
Axonoplls species. Automixis has also been
frequent; they probably arise from coalescence
claimed for Rubus caesills, as discussed by
of one pair of adjacent spindle poles upon
28 ~f.c...
differentiating the egg and micropylar-end undertake meiosis II, although this induction
polar nucleus. The egg cannot form if the third would not necessarily result in a restitution
mitotic division does not occur; its precursor nucleus. Direct induction of meiotic
defaults to being a polar nucleus. Thus meiosis interphase would likely accelerate
II and the first embryo-sac mitosis can be megasporogenesis relative to sexual ovules,
skipped without affecting female fertility or whereas entering a "waiting state" would
offspring ploidy, as in the Adoxa type of sexual cause a significant delay.
embryo sac, however the last divisions are
Induction of meiotic interphase also explains
critical for female fertility. The details of the
the Ixeris-type, but in this case the induction
numerical abnormali ties observed in Cooperia,
is superimposedon a bisporic or tetrasporic
and their implications for the evolution of
type of reduced embryo sac, e.g., the Fritillaria-
sexual embryo sacs, are reported elsewhere.
type in Rudbeckia (Battaglia 1946, 1963).
Following are the author's interpretations of
Assuming a checkpoint for chromosomal
the nine apomictic types of embryo sacs.
double-strandedness, bisporic and tetrasporic
Ameiotic Developments of types lack the second meiotic division, and
Mgagametophytes thus have their own induction of megaspore
Endomitosis in the Allium odorum-type germination after the first meiotic division
would result from relaxation of the (bisporic types) or during the first meiotic
chromosomal double-strandedness check- division (tetrasporic types). Accordingly, the
point postulated above for DNA synthesis. restitution nucleus awaits the first embryo-sac
Endomitosis can occur in a wide variety of mitosis rather than meiosis II in the Ixeris-type,
plant cells, including, most relevantly, the and there is no cross-wall to separate 211
anther tapetum (Crane et al. 1993), which is "rnegaspores."
homologous to cells in the nucellus.
In the Blumea-type, meiotic interphase is
Megasporocyte endomitosis does not affect the
induced in the premeiotic megasporocyte. Its
morphological course of meiosis; it can
genetic basis could be the same as in the
precede any of the sexual types of embryo-sac
Taraxacum-type, and the two types could
development, and it is probably
coexist in the same species or individual if the
underreported. One consideration is how the
timing of induction varied. Proving the
chromosomes behave after chromatid
existence of the Blumea-type is intrinsically
separation and secondary replication in the
difficult because it must be shown that the
premeiotic G 2 nucleus. If they scarcely move,
restitutional stage of the Taraxacum-type is
the former sister chromatids would lie in close
absent. Therefore, measuring the frequencies
proximity as pairing begins in leptotene, and
of these two types in a polymorphic individual
--nmltivalents rarely, if ever, would form.
might not be feasible with microscopic
The Taraxacum-type results from induction of techniques that kill the ovule before it is
the meiotic interphase during the prophase or examined. The outcome of superimposing the
metaphase of meiosis I. The contracted Blumea-type on a bisporic or tetrasporic
chromosomes are thus ind uced to decondense sexual type remains unknown. It might
and await the second meiotic division. superficially conform to the Antennaria-type.
Another possible explanation for the
The Elymus-type appears to result from
Taraxacum-type is induction of an Elymus-
incomplete differentiation of the
style "waiting state" (see under Elymlts below)
megasporocyte from the nucellus. Not only is
during prophase I, with recovery in time to
C1...jfj<Olioo ofApoooidk ......... 31
its callosic wall deficient, but the pattern of tetrasporic sexual types also involve induction
vacuolation and the occasional presence of an of megaspore germination during meiosis I,
oxalate crystal in the megasporocyte conform and that having tetrasporic sacs does not seem
to typical behaviors of the adjacent nucellar to predispose individual taxa to evolve the
cells. The extent to which this similarity causes Antennaria-type or vice versa. Thus different,
the wall to be deficient is not clear, nor is the separately inducible aspects of megaspore
degree to which the deficiency prevents a germination might impact meiosis in different
necessary physiological isolation for full ways, or the event that calls up megaspore
differentiation of the megasporocyte. The germination might fundamentally differ
cytoskeleton of the megasporocyte is between the Antennaria- and tetrasporic
disturbed, as evidenced by frequently oblique sexual types.
division planes and the deformation of the
The Hieracium-type results from induction of
nucleus. Perhaps the meiotic spindle
megaspore germination in the affected
apparatus is not completely suppressed, and
nucellar cells. The ability of an induced
so interacts with the preapomeiotic nucleus.
embryo sac to mature is related to its position,
But this does not explain the second
the time of its induction, and the number of
deformation in the chalazal 211 megaspore.
competing sacs. In theory, Hieraci urn-type sacs
Another feature of the Elymus-type is the need
can coexist with the Polygonum-type,
for parallel control of development. The
resulting in polyembryony and facultative
megasporocyte enters a "waiting" state and
sexuality. One would expect apomicts with the
variably emerges in time to undertake meiosis
Hieracium-type to accumulate mutations that
11, the first embryo-sac mitosis, or an
debilitate the Polugonum development and
in termed ia te condi tion, respective Iy
permit a single, unreduced sac to mature.
accounting for dyads, directly binucleate sacs,
or hemidyads. During the "wait," a master The Eragrostis- and Panicum-types share
"clock" continues to run, albeit abnormally induction of the second embryo-sac mitosis (or
slowly in comparison to the Polygonurn-type its preceding interphase), respectively, in the
in related sexual genotypes; this clock allows megasporocyte or in one to many nucellar
recovery from the wait. cells. The single nucleus is induced to behave
like the micropylar-end nucleus at the
The Antennaria-type can be explained in two
binucleate stage in the Polygon urn-type. These
ways: by induction of megaspore germination
types are known only from related subfamilies
in the megasporocyte or as a much belated
of grasses and possibly share the same genetic
recovery from an Elymus-style waiting
basis, modified only for site of induction. From
c?ndition. The frequently abnormal
a developmental viewpoint, they are related
enlargement and vacuolation of the
to the sexual Oenothera-type, in which a
megasporocyte are consistent with the latter
megaspore behaves as if it were a binucleate
explanation, but intraspecific polymorphism
embryo sac.
for the Antennaria- and Hieracium-types in
Potentillaand POll is consistent with the former Subsequent Steps of Development
explanation. Details of gene expression in In a broad sense, apomictic egg cells are
germinating Polygonum-type megaspores induced to act like zygotes, with the possible
should provide clues as to which examples of exception of semigamy. The question about
the Antennaria-type result from megaspore semigamy is whether it functions exclusively
germination. It should be noted that the in the embryo sac; anecdotal examples of
32 c.Io< F. er-
haploids from normal cotton plants pollinated imply that the genes for each could be
with semigamous pollen have been reported separated and that genotypes lacking the
(Turcotte and Feaster 1963).In the absence of autonomous endosperm induction would be
such rare haploids, one is tempted to presume fully pseudogamous. This is possible in genera
that semigamy results from induction of a where instances of pseudogamy and
subset of zygotic behavior that prevents autonomous parthenogenesis are known, e.g.,
karyogamy but not plasmogamy. If semigamy Poa and Crataegus, but it is inconsistent with
were found to function in both eggs and unreduced autonomous parthenogenesis in
sperms, one would instead suspect a general the absence of pseudogamy in apomictic
defect in fertilization capacity. Cichoreae (Taraxacum, Chondrilla, Crepis, and
Ixeris).
The variation from autonomous to pollination-
dependent embryo induction in pseudogamy Genomic imprinting of gametic nuclei is
suggests that embryos can be induced in more another consideration in the developmental
than one way. Perhaps one way triggers full interpretation of apomictic embryogenesis. In
embryonic development, while another the Polygonum-type, the endosperm has a 2:1
merely primes the egg (or sexual zygote) to maternal:paternal genome ratio; deviations
respond to a stimulus from the dividing from this ratio frequently cause endosperm
endosperm. In the latter case, the apomictic abortion in interp loidy crosses. In semigamous
egg of an unfertilized ovule would degenerate Cooperia, the reduced sperm nucleus fuses
without ever dividing, whereas in the former with both unreduced polar nuclei (Crane
case, the apomictic egg would eventually 1978), resulting in a 4:1maternal:patemal ratio
divide, even if it does not regularly do so in the endosperm, which functions adequately
before the primary endosperm nucleus for seed germination in this genus. In
divides or before un fertilized embryo sacs pseudogamous Ranunculus, both sperm nuclei
degenerate. Savidan (1989)has proposed that can fertilize the central cell, resulting in a 4:2
because pseudogamous Panicum maximum maternal:paternal ratio and normal
undergoes such early ind uction and endosperm function (Nogler 1972). In
maturation of the embryo sac, the egg cell pseudogamous Crataegus, the polar nuclei do
completes its cell wall before pollination. not fuse with each other, and a sperm fuses
While the formation of the egg wall is with only one polar nucleus, resulting in the
progressive in grasses (Cass et al. 1986), it is normal 2:1 maternal:patemal ratio (Campbell
not clear that old eggs inevitably would et al. 1987). The situation in autonomous
complete their wall. In vitro fusion of egg and parthenogenesis is not clear, because the
sperm protoplasts of maize leads to rapid cell number and biochemical effect of imprinted
wall formation (Breton et al. 1995, and loci is not obvious. Imprinting of a tubulin
references cited therein), suggesting that locus (Lund et al. 1995) and the dzrl locus
induction of zygotic behavior is responsible (Chaudhuri and Messin 1994) has been
for the physical barrier to fertilization of documented in developing maize endosperrn,
pseudogamous eggs. and it is possible that the number and sex-
specificity of imprinted loci vary greatly
Autonomous parthenogenesis would seem to
among plant species. What matters here is the
require separate inductions of endosperm
number of loci tha t orchestrate the imprinting
division and either the "triggered"
pattern. Since the evolution of apomixis is
(Wikstroemia) or merely the "primed" (Crepis)
easier if fewer loci control it, one might expect
condition. Yet separate inductions would
one or a few key loci to affect this pattern. imprinting relationships might well become
Ehlenfeld t and Hanneman (1988) analyzed the the biggest obstacle to the utilization of
crossing behavior of diploid hybrids between apomixis in sexual crop species.
two diploid, sexual Solanum species that differ
At the level of cell and molecular biology,
in endosperm balance number, and they
critical questions remain regarding the
concluded that three unlinked loci controlled
sequential and parallel nature of clocks that
it. Obviously, comparable evidence is needed
govern the relative and absolute duration of
for the immediate relatives of autonomously
events in the Polygonum-type, and thus the
parthenogenetic apomicts.
number of control points and interconnections
that suffice to push the whole program along.
Outlook
Diagnostic mRNA or protein suites have not
The speculative nature of the preceding
yet been recognized for the steps in the
discussion emphasizes how little is
Polygonum-type and applied to recognize
experimentally known about meiosis,
parallel steps in apomictic developments. The
megasporogenesis, gametogenesis, and
promoters, reading frame sequences, and
fertilization in sexual plants, and how little of
action of genes at control points remain a
the diversity of apomictic behaviors has been
mystery. Thus the basis of one of the main
studied with modem methods of genetics, cell
assumptions in the preceding developmental
biology,and molecular biology. Most apomicts
interpretations, namely that the induction of
are polyploid, perennial herbs with long
late steps cancels intervening steps within the
generation times, and the effort to maintain
Polygonum-type, has yet to be justified.
suitably large, uncontamina ted populations of
advanced generation hybrids has repeatedly Natural apomicts are difficult experimental
defeated even rudimentary attempts to subjects, but they provide embryological
understand the inheritance of apomixis. The behaviors that might be even more difficult to
pattern of genomic affinity is not known for screen from mutagenized populations of
most pseudogamous or semigamous Arabidopsis thaliana or annual crop species,
apomicts; recombination between the especially if the appropria te mu tations require
centromere and the causative loci can further gains in function. Apomictic studies might
change the expected segregation. The general contribute significantly to our future
problem is that more than one equally best- understanding of sexual plant reproduction
fitting model can be found for the same and to the successful utilization of apomixis
segregation data, especially when multiple in currently sexual crops.
allelism and complex dosage requirements are
considered and advanced generation selfs and References
backcrosses are not available. Since the Asker, S. E., and LJerIilg. 1992. Apomixis inI'/onll. , Rom Ratan,
Florida: (I( Press.
number of developmental inductions is Bottoglia, E. 1946. Ri<erche mriologiche edermnologiche sui genet"e
constrained by the number and interaction of tudWiII (Astera<eae). I-Y.U gometalito femminile e nmdliIe c5 t.
causative loci, the difficulty in completely bicoll)( Nun., t. him L, t. him L YOI Meine Freunde Hart., t.
o~xi(ouIisYahl et. purpureoL (= frhinocea pwpII"eDMoerKh).
understanding the genetics of apomixis has Nuovo GiIIT. BoI. hot 53: 1-69.
hamstrung the more direct developmental - - . 1963. Apomixis..In P. Maheshwuri (ed.), tecenl AdvOflCes in
approaches. This is particularly true for the tire Embryology of~osperms. New Delhi: Indian Sociery of PIont
Morphologists. pp. 221-64
relationship among imprintable loci, Brelan, t, J. E. Faure, and L Dumas. 1995. FrlJ(n in vitro lertiization 10
endosperm function, and embryo induction early embryogenesis in lIIIize. f'rolop/mmo 1B7: 3-12.
in autonomous apomicts. Endosperm
34 a.los F. er..
Compbell, C. S., C. W. Greene, and xon E. Ehlenfeldt, M. l, and R. E. Hanneman. 1983. NoumaYll, 1. 1993. Apomixis inAngiosperms:
Bergquisl. 1987. Apanixis and sexuality in Genelic control of endosperm balance Nucellar and Integumentary Embryony.
three species of AmeIonchier, shadbush number (EaN): three odd~ive loci in a 8am Raton, Florida: CRC Presl..
(Rosa<eae, Maloideae). Amer. 1. Bol. 74: threshold-like system. Theor. Appl. Genel. Nogler, G. A. 1972. Gene~k der Apasparie bei
321-28. 75: 825-32. Rammcu/Ul aurical1lJl. 11.
ColS, D. D., D. J.Peleya, and 8.L Rabemon. Fogerfind, F. 1940. Die Terninalogie der Endaspermzytalogie. 8er. Schweiz. Bol.
1986. Megagametophyte development in Apomixis-Prozesse. Heredilas26: 1-22. Ges. 82: 54-63.
Hordeum nJlgore. 2. Later ItOges of wall - - . 1944. 11 my terminology afthe - - . 1984. Gametaphytic opomixis..In 8.
deyelopment and morphologicol 0Ipe(lI af apanictic phenomeno of 1940 incorrect M. Johri (ed.), Embryology of
megagametophyte cell differenlialion. Con. and inappropriate? Heredilrrs 30: 590-96. Angiosperms., 8erlin: Springer-Verlag..Pp.
J. BoI. M: 2327-36. - - . 1947. Mocrogametophyte formation 475-518.
Chaudhuri, S., and J.Messing 1994. Allele- in I'tIIIl agafllOlpermous &igeron lpecies. Okabe, 5. 1932. Parthenogenesis in Ixeris
specific parentol ifl1lrinling of duI, a Acla Hart. Berg. 14: 223-47. dentola. Bol. Mug. Tokyo 46: 518-23.
posttral1\Cliptional regulator of zein Gledh~t D. 1967. Embryo IOclOImOtion in RCMnberg, O. 1907. Cytologicolltudies on the
acCUI1lJIation. Proc. Notl. Mod. 5O.IUSAl African AxonopUlIpe<ies. apagamy of Hieradum. 5vemk. Bol. rKJSKr.
91: 4867-71. Phylomorphology 17: 214-23. 28: 143-70.
(honnoYeeraiah, Mo S., and R. M. Pati!o 1971. GUltafsson, A. 1946. Apomixis in higher ~anb. Soman, Y. 1989. Another'working hypothesis'
Apomixis in Blumeo. I'frytomorphology 21 : Part I.The mechanism of apanixil. Lundl for the control of parthenogenesis in
71-76. Univ. Amkrift N. F. 42: 1-67. Panicum maximum: the egg cell wall
Coe, G. E. 1953. Cytology of reproduction in Hair, J.8. 1956. Subsexual reproduclion in completion hypathesis. Apomixis
CooperiD peduncu/alo. Amer. 1. Bol. 40: Agropyron. Heredilyl0: 129-60. NewsJeller 1: 47-51.
33>-43. Hakansson, A., and A. Leyan. 1951. Streelman, L 1. 1963. Reproduction of ,he
Crane, C. F. 1978. Apanixil and crossing Parthenogenesis in Allium. Bal. Noliser IoYegrOlIel, the genUl &ograstis. I. f.
inco.ibililies in lOme Zephyranlheae. 1951: 143-79. chloromelas Steud., f. cufVIIla IXhrod.)
Ph.D. dissertation, the University of Texal al - - . 1957. Enda-duplicatianol meiosil in NeeI, f. lehmanniono NeeI and E. superbo
Austin, AUltin, lexes, Allium odorum. Heredilos43: 179-200. Peyr. Wrighlio 3: 41-51.
Crane, C. F., and J.G. Corman. 1987. HoImgren, I. 1919. Zytologische Studien (uber Turcone, E. L,and C. V. Feaster. 1963. Haploidl:
Med1anilml of apomixis in f/ymus die Fortpflanzung bet der Gattungen High-frequen<y production from ling Ie-
rectiselUlfrom eastern AUltrana and New &igeron und Eupatorium. K. 5Yen. embryo seeds in a line of Pima conon.
Zealand. Amer. 1. Bol. 74: 477-96. VelenlicapsoJcod. Hondl. 59: 1-118. Science 140: 1407~8.
Crane, C. F., H. J.Price, D. M. Stel~, D. G. Jahri, 8. M., l 8.Ambegaokar, and P. S. - - . 1969. Semigametic production of
Czeschin, and 1. D. McKnight. 1993. SriYOItaYll. 1992. Compora/iye haplaids in Pima canon. Crop 50. 9: 653-
Identification of a homealogoul Embryology ofAngiOlperms. Bernn: 55.
mrol11OlOme pair by in I~U DNA Springer-Verlog.. - - . 1973. The origin of 2n and nsectorl
ITybridization to ribasarnal RNA loci in JueI, O. 1898. Parthenogenesil bet AntennariD of chimerol Pima conan ~anb. Crop 50.
meiotic chrOl11OlOmel of cotton (Gossypium alpino LBer. Bal. Zbl. 74: 369-72. 13: 111-12.
hifSulumL). Genome 36: 1015-22. - - . 1904. Die Tetrodenteilungen inder - - . 1974. MelllOds of producing
Edman, G. 1931. Apameialis und Apomixis bet Somenonlage yon Taraxacum. ATk. 801. 2: haplaids: Semigamelic production of conon
Atrophaxis frvtescem C. Koch. Aclo Hart. 1-9. haploids. In KJ. Kosho. (ed.), Hop/aids in
Berg. 11: 11-66. Lund, G., J.Mesling, and A. V"lOtli. 1995. Higher I'fan~: Advances and Palentiol
Endasperm-lpecific dernethylation and Ontario, Canada: UniY. of Guelph. pp. 53-
activatian of spe<ific alleles afalpha- M.
tubulin genes of lea mars L Mol. Gen. Voigl, P. W., and E. C. Bashaw. 1972. Apomixis
Genel. 246: 716-22. and sexuolity in &ograstis cufVIIla. Crop
Sci. 12: 843-47.
Warmke, A. E. 1954. Apomixis in Panicum
maximum. Amer. J.Bal. 41: 5-11.
Classililalioo of..,.lio: ........ 35
media. There may also be a statistical Both of Herr's 4-1/2 media contain chloral
correlation of optimum refractive index to hydrate, a federally controlled substance that
DNA content per chromosome, with large sometimes presents legal problems for institutions
genomes tending to require a higher no. that possess it. They also contain eugenol, which
For example, Zephyranthes can be cleared turns yellow upon exposure to light, and phenol,
in methyl salicylate (no = 1.537), but which likewise discolors and is a chemical contact
Nothoscordum (which has bigger hazard. On the other hand, excellent photographs
chromosomes) requires a higher nO' as have been obtained with Herr's media in
provided by 2:1 benzyl benzoate: dibutyl appropriate species, and their refractive index
phthalate (no = 1.542). Possibly this probably can be adjusted upward without
correlates more directly to G-C content,
pitch of the double helix, or degree of
cytosine-methylation in heterochromatin, Table 3.1 Refractive index (nJ of common and
than to total DNA content; the smaller potential clearing media
genomes simply carry less Subst.ce DD
heterochromatin. 20% sucrose 1.363
40% sucrose 1.399
Table 3.1 presents refractive indices of
60% sucrose 1. 441
various published and candidate clearing 70% sucrose (saturated) 1.465
media as measured with an Abbe saturated sucrose in 50% g~cerol 1.468
refractometer at ca. 20°C. Most ovule- 80% fructose (saturated) 1.483
clearing studies have used one of two 80% fructose in 10M Kacetate, 10 mM KOH 1.494
80% fructose in saturated (ca. 8.6 M) KI, 10 mM KOH 1.517
classes of media: modified lactophenol
95% g~cerol saturated with KI, 1mM KOH 1.506
(Herr's4-1/2and BB4-1/2, Herr 1971and saturated CaCI 2 in H20 1.469
1974) and aromatic esters (Crane 1978; saturated CaCI 2 in g~cerol (glassy solid atroom temperature) 1.534
Younget al. 1979; Crane and Carman 1987). saturated 2nCI 2 in H20 1.550
The latter have been combined successfully saturated 2na 2 in 80% (v:v) g~cerol 1.532
1:1:4 (v:v:w) water, g~cerol, 2nCI 2 1.518
with staining in hemalum (Stelly et al.
Fea3 6H 20 liquefied with aIrace of g~cerol 1.560
1984) or azure dyes (G. L. Hodnett, 50% pa/yvinylpyrrolidone (MW 10000) 1.429
personal cornrn.) for brightfield 75% polyvinylpyrrolidone (MW 10000) 1.487
observation. The Herr media do not fully saturated naphthalene in g~cerol I.472
dehydrate the ovule because of the 15% Hell's 4 1/2 medium 1.503
Herr's 8B 41/2medium 1.510
wa ter in commercial lactic acid, and
methyl salicylate 1.537
organelles and cell walls therein seem to dibutyl phthalate (OBP) 1.492
clear optimally around no 1.51, whereas benzyl benzoate (BB) 1.569
the optimum in aromatic esters is around BB:OBP mixes, v:v:
no 1.53 or 1.54. A third class, nearly 50:50 1.530
55:45 1.533
saturated sugar solutions, has been used 60:40 1.537
to document callose deposition (Peel et al. 65:35 1.541
1997), with lesser success in displaying 70:30 1.545
nuclear and cellular locations. A fourth 75:25 1.549
class, salts in glycerol or sugar solutions, 80:20 1.552
saturated KI + 2.5 mg/ml Iris base in 50% OMSO 1.475
remains to be tested, but appropriate saturated KI + 2.5 mg/ml Iris base in 70% OMSO 1.496
refractive indices have been obtained for saturated KI + 2.5 mg/ml Iris base in 80% DMSO 1.508
high-quality clearing (Table 3.1). saturated KI + 2.5 mg/ml Iris base in 90% DMSO 1.519
saturated KI + 2.5 mg/ml Iris base in DMSO 1.525
Gnsifi<.....fApoalidi< ......... 37
affecting their general properties by partially not clear the tissue well. This is attributed to
substituting isoeugenol (n D = 1.57-1; Windholz iodination of the macromolecules in the cell,
et al. 1983) for eugenol (n D = 1.541;Windholz which raises its refractive index. The problem
et al. 1983) in the recipes given below. with CaClz is the high viscosity of its solutions
in polyols. Its hygroscopicity also causes the
Aromatic oils require complete or nearly
refractive index of the preparation to vary with
complete dehydration of the ovule, which
the relative humidity in the room. Ferric and
typically lengthens the clearing procedure and
zinc chlorides are extremely corrosive to
leads to shrinkage and possibly distortion. The
metals and human flesh, and they destroy
ovules normally become hard and brittle,
nuclei relatively rapidly. Potassium
which is good on the microscope slide but not
thiocyanate gives a high refractive index in
on the dissecting stage. While the oils that have
solution, but is chaotropic (destroys nuclei
been used are only mildly toxic, this is not
again) and hazardous to the mental health of
uniformly true, and some oils (polynuclear
the slide user (Handbook of Chemistry and
aromatic hydrocarbons) must be rejected as
Physics). Finally, the problem with DMSO for
components of clearing media because of their
embryological clearings is its denaturant
carcinogenicity.
action on DNA and proteins, resulting in poor
Both the lactophenol and aromatic-oil clearing quality of nuclei or no nuclei at all. This action
agents offer limited opportuni ties for is the basis of a very effective recipe (provided
cytochemistry. Most of the established below) for destructively clearing leaves for
cytochemical procedures are based on vascular and epidermal study.
reactions in water and may behave abnormally
The following protocols are mostly based on
in less polar solvents. Opaque reaction
fixation in organic solvents that kill the cell
products can obscure unstained regions
quickly and preserve the chromosomes well.
behind them, and accessibility to reactants is
The choices include FAA (37% aqueous
always more difficult in intact ovules than in
formaldehyde, acetic acid, and 50% or 70%
sections thereof. Reaction products can be
ethanol, 1:1:18 by volume); FPA (the same
mobilized or lost upon dehydration and
solution with propionic acid substituted for
infiltration with clearing agents. Nevertheless,
acetic acid), 3:1 ethanol:acetic acid;, Carnoy's
there is interest in distinguishing various types
solution (6:3:1 ethanol:chloroform:acetic acid,
of cell walls (especially callose) and cellular
by volume); and 2:1 acetone:acetic acid. Craf
inclusions (e.g., starch grains) in cleared
and Allen-Bouin-type fixatives can also be
ovules, and a medium that can be used to this
used, but require longer dehydrations.
end is introduced here for further
Formaldehyde and glutaraldehyde, alone or
investigation.
in combination, also require long dehydrations
Several inorganic salts, including those listed and typically leave the ovules yellow. They
in Table 3.1, dissolve to a considerable degree also contribute to autofluorescence in the
in DMSO, glycerol, concentrated sugar detection of callose with aniline blue. The less
solutions, or polyethylene glycol. None of polar fixatives generally cause severe
them has proven completely satisfactory as an plasmolysis of embryo sacs with two or more
alternative to the aromatic oils or 4-1/2 media, nuclei, but they also extract chlorophyll and
but they might permit additional carotenoid pigments more completely from
cytochemistry to be performed on whole ovary walls. Acetone/acetic acid is
mounts. Some media with KI equal the 4-1/2 particularly effective at this and often leaves
media in refractive index (Table 3.1\, but do tissues snow-white.
38 c.In F. er-
7. The nuclei can be stained in [JAPI (3 g/ for attempting the aniline blue reaction
ml in McIlvaine's buffer), dehydrated, for callose. However, this reaction is
and cleared in aromatic
,.., oils. known to fail in DMSO, beca use the
Epifluorescence microscopy of such callose is denatured and partially
preparations with ultraviolet excitation dissolves. Likewise, the iodine test for
can yield stunning photographs, but the starch fails for media with more than
necessary control of destaining and the 50% DMSO; the helical structure of the
requisite photographic skill are much starch is presumably disrupted upon
higher than for azure dyes. solvation with DMSO.
Clearing in salty glycerol (Crane, Rapid clearing of leaf tissue (Crane and Y.
unpublished data): Ma, unpublished data)
1. Fix specimens in FAA-50 or 2:1 (v:v) 1. Cut the specimen into ca. 1 cm squares.
acetone:acetic acid (AnA), as in the IN A WORKING FUME HOOD,
general aromatic-oils protocol. immerse a few of these in 50 ml of a 2:1
2. Bring the specimens to 70% ethanol. (v:v) mixture of DMSO and chloroform
3. Infiltrate the specimens with 3:1:4 (v:v:w) and heat on a hot plate until it begins to
glycerol: polyethylene glycol (MW ca. boil. Move the beaker on and off the hot
200): CaCl 2 _ 2H 20 . A series of 6-10 half- plate to keep the solution just below
hour gradations, e.g., 7:1, 6:2, 5:3, 4:4, 3:5, boiling (ca. 700C) until the solution has
2:6, 1:7, of 70% ethanol to glycerol become green. Change to fresh 2:1
solution is reasonable. The necessary DMSO: chloroform and continue. Repeat
number of gradations is determined by until leaf pigments are no longer
specimen size and permeability to gases. extracted into the liquid. Use tongs to
If air is exsolved, use more steps in the handle the beaker, and wear
infiltration series. Although it is tempting impermeable gloves. THIS FIXATIVE IS
to heat the clearing agent at the latter CARCINOGENIC!!!
stages of the infiltration in order to 2. Infiltrate the squares with 3:1 or 7:3 (v:v)
reduce its viscosity, doing so risks benzyl benzoate:dibutyl phthalate. The
denaturing callose and obliterating its optimum mixture depends on the
aniline-blue response. thickness of the specimen and its
secondary cell walls. Use at least 2 ml of
4. Mount the specimen on a slide, apply a fluid per square. One possible schedule is
.cover slip, and seal it with rubber cement 30 min each in 3:1, 2:2, 1:3 mixtures of the
or viscous silicone oil. The latter is messy fixative with the clearing agent. Finally
but will probably provide a more change to the pure clearing agent, wait
permanent seal. Unsealed slides 30 min. and change again to the pure
eventually would take up water from the clearing solution. This will reduce the
air. amount of chloroform liberated around
5. Alternative clearing agents for this recipe the user when the slide is mounted.
can be found in Table 3.1. Dimethyl 3. Mount the cleared square in the clearing
sulfoxide dissolves many salts, but it medium on a slide, apply a cover slip,
denatures proteins over time and cannot and examine it with phase-contrast or
reliably give sharp nuclei. One can add differential interference contrast optics.
tris-hydroxyaminomethane (tris base; 2.5 Keep track of the abaxial and adaxial
mg/ml) to these media to raise the pH surfaces of the specimen.
4. The method works also for meristerns, &delska, D., H. H. Heunert, 1. Hard, J. Koeding, and H. W"dlmam. 1979.
floral primordia, and other general Befrudrtung und FriiIre Entwidclung 1'00 Embryo IIId Endosperm beim
SchneegliicJcclren. Inslilul fUr den W"1SSeIlSdlaflIic Film, lID. 1465.
subjects where one wants to see only the Fredrikson, M. 1990. Embryological study of Heminiwn monorchis
cell walls. Oxalate crystals and opaline (Orchidoceoe) usilg conforal smming loser microKopy. Amerimn
silica survive the procedure intact. If the JOIJfnaJofBotony. 77: 123--27.
Hen, J. M. 1971. Anew dealing-squash technique for the slucly of lMJle
epidermis separates from the mesophyll, development in angiosperlllS. Amerimn JOIXIIOI ofBotDny. 58: 76(l.
use less heat next time during fixation. 785.
- - . 1974. Aclearing-squash technique for rhe study of ovule and
Acknowledgments megagametophyte development in ongiosperms.1n A. Elodford, W.
The author would like to thank Drs. David M. C. Dickinson, J. R. Mossey, and C. I. Bell (eds.l, Vascular I'font
Systemotia. New York: Harper and Rowe. pp. 230-35
Stelly and H. J. Price for their help during the Hen, J. M., Jr. 1982. An allD!y5is of methods for permlI1IIlIty mcuding
writing of this chapter. This research was a~ cleared in four-ilnd-il-hoIf type de«ing I'uids. SIoil
supported by TexasAdvanced Technology and Technology 57: 161-69.
Hodnen, G.L, c.F. Crone, and D.M. S.eIy. 1997. Arapid 5lDin-de«ing
Research Program Grant 999902090. method for video based cytological arOysis of colloll
megagome.ophytes. Biotedrnic and Histochemistry 7t. 16--21.
References Peel, M.D., J.c. Cannon, Z.W.liu, and u.-e. Wang. 1997. MeioIK
Berlyn, G. P., and J. P. Mhhe. 1976. Botanical MiCTotechnique and anomolies in hybrids be_n wheat and oponicIic: E1ymus reetiwIus.
(ytochemistry. Ames, Iowa: The Iowa SIote Univenily Press. (Nees in Lehm.) A.LDve and Connoc. Crop Sri. 37: 717-23.
Crone, C. f. 1978. Aporrixis and crosWg ilKompotibitllies insome Soss, J. E. 1958. Botonical Microtechnique. Ames,Iowa: Iowa 5tDle CoIege
Zephyrantheae. Ph.D.I5ssertDlion, the UniverYly of Texas a'Austin, Press.
Aus~n, Texas.
5.e1~, D. M., S. J. Peloquin, R. G. PoImer, and C. F. Crone. 1984. Moyer's
Crone, C. F., and J. G. Call11Cl1. 1987. MedIOnis1r6 of apomixis in E1ymus hemolum-methyt sotKYiote: AstoilKleorilg techrique for
rectisehJs from easlern Australia and New ZeoIond. Amerimn. JOIJfnol observations within whole 0¥Ules. Stoil Tedrn%gy 59: 155-61.
ofBolr1ny. 74: 477-96. W"lIldholz, M., 5. 8udovori, R. f.81umelli, and E. 5. Otterbein. 1983. The
&delska, D., H. K. Golle, H. H. Heunert, and K. Phiipp 1971. Merck Index. Railway, NJ.: Merck and Ca., IlK.
Samenonloge und frUhe EndospermentwicHmg 1'00 Jasione montDno Young, B. A., R. 1. Sherwood, and E. C. Boshow. 1979. Oemed pislil and
((omponuIrKeoeJ. InsliluI fUr den ~nsdlafIfidlen Film. Ihick -sectioning .ed1niques for delecti1g aposporDUS oponixis it
grosses. (onodion.Journo1 ofBotany. 57: 166Pr-72.
Chapter 4
Ultrastructural Analysis of Apomictic
Development
TAMARA N. NAUMOVA AND JEAN-PHILlPPE VIELLE-CALZADA
(Figure 4.2a,b,c). Their cytoplasm contains ciliare, as in most reported aposporous species,
numerous plastids and mitochondria, and female gametophytes usually contain two
plasmodemata are scarce or appear to be filled synergids, an egg cell, and a single polar
with a cellulose-like matrix. The degeneration nucleus in the central cell; however, a variable
of meiotically derived megaspores and early number of female gametophytes (up to 20%
stages of aposporous initiation has recently in certain genotypes) may be composed of a
been investigated in Brachiaria brizantha. In this single synergid, an egg cell, and two polar
grass, aposporous initials appear to contain nuclei. On rare occasions, embryo sacs
dedifferentiated organelles reminiscent of containing three polar nuclei and no synergids
organellar populations found in pre-meiotic have been observed. This variable organization
MMCs (AC. Guerra de Araujo and V.Carneiro, appears to be associated with the localization
personal communication). of nuclei prior to cellularization, and suggests
that positional information plays a role in
Aposporous Megagametogenesis
gametophytic cell specification.
Several aposporous embryo sacs can develop
in a single ovule (Figure 4.2d). Aposporous In all aposporous grasses ultrastructurally
female gametophytes show a different examined to date, the egg apparatus
orientation with respect to the micropylar- differentiates with morphological
chalazal axis, and usually differentiate characteristics similar to those found in
heterochronically with respect to each other. sexually functional synergids and egg cells.
The transition from aposporous initial to one- These cells are attached at the micropylar apex,
nucleated embryo sac is characterized by an but the triangular organization of the egg
increase in cell size (Figure 4.3a).The first signs apparatus is not necessarily conserved, as the
of cellular polarity are the consequence of unreduced egg cell may appear in a more
vesicular fusion; two large vacuoles are formed lateral position with respect to one of the
at opposite sides of a centrally located nucleus. synergids. Except for the presence of the
Cell wall thickness also increases, and no filiform apparatus, few differences are found
plasmodesmata can be discerned (Figure in the ultrastructural constitu tion of the
4.2d,e). In Penniseium ciliare, the nucleus synergids and the egg cell in unpollinated
migrates to the periphery of the cell before pistils of Panicum maximum. The three cells
dividing (Figure 4.3b).The first mitotic division appear vacuolated, with a centrally located
is perpendicular to the long axis of the cell and nucleus, and organelles preferentially located
usually close to the micropylar region of the in the micropylar pole (Figure 4.2f,g,h).
one-nucleated embryo sac (Figure 4.3c). The Naumova and Willemse (1995)characterized
second mitotic division is synchronous in both the ultrastructure of aposporous embryo sacs,
sister nuclei, giving rise to a four-nucleated but only observed events prior to pollination.
type of embryo sac that lacks antipodals at the In unpollinated pistils, the egg cell and
chalaza I pole (Figure 4.3c,d). synergids were similar to the egg apparatus
of an embryo sac of the Polygonum-type
The Cellularized Aposporous
typically found in the grasses. The egg cell wall
Megagametophyte
The second mitotic division is followed by the showed variable thickness, and the chalazal
cellularization of individual nuclei. Little is plasma membrane surface of the egg, central
known about the processes that regulate cell, and synergids were in direct contact
cellularization and differentiation of (Figure 4.2f,g,h). Chapman and Busri (1994)
aposporous megagametophytes. In Penniseium described the ultrastructure of mature
Ultra.,.. <1.raI Aoalysi••f Apoooktk D....,...., 49
Figure 4.4 Organization of the mature aposporous egg apparatus in Pennisetum ciliare.
(e) The egg cell three hours after pollination.The chalazaIend is campletely covered by a cell wall (CW)" vacuole (V), eggcell (EC);
(b) Detail of thecell wall (CWl separating the chalazal regionof the egg cell (EC) from the central cell ((C) cytoplasm; M,
mitochondria; (cl Micropylar region of theaposparous egg apparatus, filiform apparatus (FA), egg cell (EO; (d) Numerous Golgi
(arrowheads) are present in the apical packet, a regian located between the central cell ((C) wall and theegg apparatus, egg cell (EO,
plastid (P), vacuole (V), and cell wall (CWl.
Ultraslnnl,rlll A.lIIysis of Apomidic Denlopmtll 55
aposporous embryo sacs in the progeny of a of both cells contains small vacuoles, and few
Pentusetuni glaucunt (sexual) x PClll1iscllll1l organelles can be identified. In general, the
souantulatuni (obligate apomict) interspecific cells are similar in appearance to the sexual
cross. In facultative apomictic genotypes, they synergids during the first two hours following
compared sexual and aposporous embryo sacs pollination, but before pollen tube arrival.
in the same ovule. Particular attention was
During the first two hours following
given to the distribution of internal cell wall
pollination, the cytoplasm of one of the
ingrowths (transfer walls) in the central cell.
synergids becomes electron-dense, the plasma
Many wall projections were observed in the
membrane appears disrupted, organelles
micropylar region, but few ingrowths were
cannot be identified, and the nucleus is
found in the chalaza I region of both sexual anti
irregularly shaped and pressed to the plasma
apomictic megagametophytes. Plasmodes-
membrane. Increased amounts of vesicular
mata were present in the cell wall that separates
traffic are observed. Later, the cytoplasm of the
the central cell from the antipodals, but
second synergid also shows signs of
appeared to be absent at the chalazal pole of
degeneration. Two to three hours after
aposporous embryo sacs.
pollination, the degenerated synergid has
Parthenogenesis and Fertilization entirely collapsed and its remnants appear as
In Penniseiumciliate. most genotypes reprod uce an electron-dense fringe closely associated
by obligate apospory. Breeding efforts are with the egg cell. The time of initiation of the
based on the identification of rare genotypes degenerative process is variable in both sexual
that have completely lost the ability to and aposporous embryo sacs, and in some
differentiate aposporous megagametophytes cases, highly degenerated synergids are
and only form sexually functional, reduced present in unpollinated ovules during or just
(haploid) embryo sacs of the Polygonum-type. after embryo-sac cellularization.
The comparison of the egg apparatus in sexual
Significant differences are also observed
and aposporous megagametophytes of P ciliare
between the aposporous and the sexual egg
offers an opportunity to analyze the cellular
cell. After cellularization but before pollination,
dynamics of fertilization in apomictic plants.
the aposporous egg cell is characterized by a
Such a comparison is particularly valuable if
conspicuous centrally located nucleus that
conducted where independent genotypes of
contains a single nucleolus (Figure 4.3e). The
apomictic and sexual germplasm are available
nucleus of the sexually derived egg cell is
within a population segregating for method
generally centrally located, but has a smaller
of reproduction.
nucleolus. The chalazal vacuole present in the
In P. ciliare, fertilization occurs 3-4 hours after sexually derived egg cell is replaced by several
pollination. The examination of the egg smaller vacuoles that restrict the cytoplasm to
apparatus of buffelgrass at several time a region located around the nucleus. In contrast
intervals after pollination has provided some to the sexual egg cell, the cytoplasm contains
information on the structural and functional abundant mitochondria and polysomes, but
features that distinguish sexual and apomictic few Golgi bodies and endoplasmic reticulum
development (Vielie et al. 1995). Compared to (ER) can be observed. At the micropylar region,
the synergids of the sexually derived egg both sexual and aposporous eggs share
apparatus, the degenerative process in one or common cell walls with both synergids, but
both aposporous synergids appears to be these walls disappear in the middle portions
accelerated. Prior to pollination, the cytoplasm
of the cell. The plasma membranes of the egg is endemic to the extreme eastern region of the
and synergids are in direct contact at the former USSR. Synergid embryos develop in
chalazal pole. more than 80% of the embryo sacs investigated
in this species (Naumova 1978, 1990). Light
Three to four hours after pollination, striking
microscopy studies have shown that the egg
changes are detected in the ultrastructure of
cell and the synergids undergo limited
the aposporous egg cell. A cell wall without a
differentiation. They are characterized by a
midd le lamella has covered the chalazal region
centrally located nucleus and no central
of the egg plasma membrane, separating the
vacuole. The three cells appear similar in size
cell from the degenerated cytoplasm of a
and shape, but the presence of a conspicuous
collapsed synergid (Figure 4.4a,b). No cell wall
filiform apparatus is characteristic of the
covers the chalaza I end of the sexual egg cell,
synergids. Differences between their
even after pollen tube discharge. The
cytoplasmic constitution can only be discerned
aposporous egg cell cytoplasm appears
at the ultrastructural level. In contrast to the
vacuolated and contains numerous
egg cell, synergids are rich in endoplasmic
undifferentiated plastids preferentially
reticulum and Golgi cisterneae. Plastids.
organized in clusters at the periphery of the
mitochondria, and polysomes are abundant
nuclear membrane. A thin layer of central cell
in the egg cell. Before pollen tube arrival into
cytoplasm is associated with the external
one of the synergids, the three cells have an
surface of this de /lOVO formed cell wall (Figure
incomplete cell wall at their chalazal pole
4.4c,d). The central cell cytoplasm contains a
(Figure -l.Sa.b),
large number of Golgi cisternae and
mitochondria, particularly in the so-called Following fertilization of the egg and central
apical pocket, a region of the central cell formed cell, drastic ultrastructural changes occur in
by the proximity of the egg apparatus to the the persistent synergid. An increase in the
central cell wall (Figure 4.4d). The unreduced nuclear and nucleolar size is followed by a
polar nucleus usually contains more than one complete reorganization of the organellar
nucleolus. In some rare occasions, a population, which becomes similar to the one
multicellular embryo can be present before present in the egg cell before sperm cell
pollen tube arrival into the micropyle. delivery. Mitochondria with numerous cristae
and plastids dense in stroma are abundant.
Apogamety Whereas the sexually derived embryo is
Apogamety designates the formation of already surrounded by a thick cell wall devoid
embryos from a cell of the megagametophyte of plasmodesmata, the cell wall of the 2-
other than the egg. Even if this phenomenon cellular embryo derived from autonomous
has rarely been reported in sexual and synergid activation thickens (Figure 4.5c,d)
apomictic species (Asker and Jerling 1992), it and becomes progressively isolated through
implies the autonomous activation of the loss of all plasmodesmata. In summary, the
reproductive cells and can be considered a ultrastructural transformations taking place in
nonrecurrent form of apomixis. embryogenic synergids of T. canischatcense are
The fine structure of synergids undergoing comparable to the changes occurring in the egg
cell during the gametophytic to sporophytic
autonomous activation has been described in
transition of sexual flowering plants.
Trillium camschatcense Ker. Gawl., a species that
0.....,..., 59
Ultra.'..".... AoaIy.i••t ......1it
Dickinson, H.G., ond U. Po"er. 1978. Cyto~osmic ligrOl1e, R., J.G. Ducke", and K.S. Renzoglia. Pennell, R.I. 1992. Cell surfoce orobinogolacton
chonges occomponying the female meiosis 1993. The gametophytic·sparophytic proteins, orabinogoloctons and ~ont
in Ulium longif/orum Thunb. 1. (ell Sei. 29: junctiOl1 in land plants. Advanced 80tanical development. In J.A. Callow ond JR. Green
147~9 Research 19: 231-317. (eds.!, Perspectives in PIonl (ell Recognition.
Faure, J·E, C. D~OI1net, ond e. Dumas. 1994. An Mortinez, EJ., F. Espinoza, and (L Ouorin. Society of Exp. Bioi. Series 48. Cormridge
in vitro system for odhesiOl1 ond fusion of 1994. Bill progeny (211tn) from apamictic U.K.: Carmridge Univ. Press.
moize gometes. Seience 263: 1598--1600. Paspolum nolalum obtoined through eor~ Pennell, H, LJanniche, PKjelbam, G.N. Scafield,
Gerossimavo-Novoshino, H. 1957. On some pallinotion. Journal ofHeredity 85: 295- J.M. Peort, ond K.A. Roberts. 1991.
cytologicol principles underlying 97. Developmental regulotion of a plosma
fertilizotion Phytomorphology 7: 1582-98. Mogensen, H.L 19B8. Exclusion of mole membrane arobinogalocton prOlein epitope in
Golubovskoyo, I.N. 1979. Genetic cOl1trol of mitochondrio ond plostids during syngamy oibeed rope Rowers. Plonl (ellJ: 1317-26.
meiosis. 1nl. Rev. (ytal. 58: 247-90. in barley as 0 bosis fOl molernol Pennell, R.I., and K.A. Roberts. 1990. Sexuol
GoIubavskoya, I.N, N.A. Avalkin, and W.F. inheritonce. Prac. Natl. Acad. Sa. (USA) 85: development in the peo is presaged by
Sheridan. 1992. Effect of several meio~c 2594--97. oltered expression of orobinogalacton
mUlonls OI1femole meiosis in moize. Oev. Noumova, IN. 1978. Peculiority of protein. Hature 344: 547-49.
Genel. ,3:411-24. mocrogometogenesis and posl·fertilization PhiliplOl1, M.N. 1978. Apomixis in (artoderio
GustaksOl1,A. 1947. Apomixis in h~her plants development of lrillium calTl!(hatcense iubala (Gromineoe). Hew Zeal. J 80t. 16:
Il.lunds Univ. Arsskr NF1I43: 71-179. Ker. -Gowt. Soc.8al. Fr. Actualitfis 45-59
Huang, B.O, and S.D. RusseR 1992. The femole Solaniques 1-2: 183--86. Remy, W., PG. Gensel. and H. Hoss 1993. The
germ unit. In S.D. Russell and e.Dumos - - . 1990. Fami~ Trillioceoe. In M.S. gametophyte generaliOl1 of some eor~
(eds.!, Sexual Reproduction in Flowering Vakalev ond lB. Botygina (eds!, DevOl1ion lond ~onts. 1nl. 1. Planl Sei. 154:
Plants. 1nl. Rev. (ytoJ. 140: 233--93. (amporative Embryology ofFlowering 35-58.
Higashiyomo 1, H. Kuroiwo, S. Kawono, ond 1 Plants (Butornoceoe·lemnoceoe). Russell, S.D. 19B5. Preferentiolfertilizalion in
Kuroiwa. 2000. Explosive dischorge of Leningrod: Nouko, (in Russion) Pp. 151- Plumbago; Ultrastructurol evidence fOl
pallen tube contents in Torenia foumieri. 59. gomete-Ievel recognitiOl1 in on angiosperm.
Planl Phys.122: 11-14. - - . 1993. Apomixis in Angiaspernrs: Prac. Norl. Acad. sa. (USA) 82: 6129-32.
Jensen, W.A. 1965. The ultrastructure and Huce/lar and Inlegumenlary Embryony. - - . 1992. Double fertilizotiOl1. In S.D.
histochemistry of the synergids of ca"OI1. Boco ROtOl1, Florido: CRC Press. Russell and e. Dumos (eds.l, Sexual
Amer. 1. 80t. 52: 238--56. Noumovo, IN., A.P.M. den Ni~, and M.T.M. Repoduction in Aowering PIonts. 1nl. Rev.
Jensen, W.A_, and D.B. fISher. 1968. Canon Winernse. 1992. Ouantitative ono~is of (ytal. 140: 3S7-88.
embryogenesis: the entrance ond discharge oposporous parthenogenesis in Poa Sovidon, V. 1982. Nature ethellidite de I'apomixie
of the pollen tube in the embryo soc. Plonta pralensis genotypes. Acta 801. Heerlondico chez Panicum maximum Jocq. Travel Ooc
78: 158--83 43: 1-14. ORSTOM 153:1-159.
Knox, J.P 1992. Molecular probes for the ~onl Noumovo, IN., ond F. Matzk. 1998. Differences Schulz, S.R., and W.A. Jensen 1968. (apse/Ia
cell surface PrOloplasmo 167: 1-9. in the initiotiOl1 of the zygotic ond embryogenesis: lhe egg, zygole ond young
Knox, R.B., and J. He~op-Harrison. 1990. Direct parthenogenetic pathway in the SolmOl1 embryo. Amer. 1. 801. SS: 807-19.
demonstrntion of the low permeability of lines of wheot: ultrostructurol studies. Sex V'ielle, J-p' B.L BursOl1, H. Boshow, ond M.A.
the angiosperm meiotic tetrod using 0 Planl Reproduction 11: 121-30. Hussey. 1995. Eorly fertilizotion events in the
Auarogenic filter. l. Pf/onzenphys;al. 62: Naumovo, IN_, J.V. Osodtchiy, V.K. Shormo, P sexual and oposporaus egg apparatus of
451-59 Dijkhuis, and K.S. Ramulu. 1999. Apamixis Pennilelum ciliore tl.) link. Planl1. 8: 309-
Knox, JP,PJ. linsteod, J.King, e.Cooper, and K. in ~onts: structural ond functional ospe<ts 16.
Roberts. 1990. Pectin esteriocation is of of di~osply in Pao nemorolis and P. V'ielle·Calzoda, l-P., CF. Crone, and D.M. Stel~.
spaliolly regulated bath within cell walb and polllSlris. Praloplasma 208: 186-95. 19960. Apomixis: the osexuol revolution.
between developing tissues of root opices. Noumovo, IN., ond M.lM. Willemse. 19B2. Science 274: 1322-23.
Planla 181:512-21 lIucellar pa~ermryony in Sorcococca Vielle-Calzada, J-P, M.L Nuccio, M.A. 8udimon, IL
Kohunow, A.M., K. SoI~, N. Nobumoso, ond S. humilis: ultrastructurol ospects Thornos, B.L Burson, M.A. Hussey, and RA
Mcdure. 1995. Anther, ovule, and nucellar PhytornorphoJogy 32: 94--1 D8. Wing. 1996b. Comparotive gene expression
embryo development in GrrllS sinensis cv --.1995. Ultrostructural in sexuol and opomicti< ovaries of
Valencia. (an. 1. 8ot. 73: 1567-82. choracterization of apospory in Panicum Penniselum ciliore. Planl Mol. 8iol. 32:
Kranz, E., J. 80uIOl, and H.liirz. 1991 In vitro maximum. Sex. PIont Repr. 8: 197-204. 1085-92.
fertilizotion of single, isolated gametes of Nogler, G.A. 1984. Gametophytic opomixis. In Voigt, P.W., and E.e. 80show. 1972. Apamixis and
moize medioted by electrofusion. Sex. Plant B.M. Johri (ed.), Embryology of sexuolity in fJogroslis curvvla. (rap 5<i. 12:
Reprod. 4: 12-16. Angiospernrs. Berlin: Springer·Verlog. Pp. B43--47.
leblanc, 0., M.D. Peel. J.G. Carman, and V. 475-518. Wakona, A., and S. Uemoto. 1987. Adventive
Sovidon. 1995. Megasporogenesis and Peel. M.D, J.G. Carman, and O.Leblanc. 1997. embryogenesis in GrrllS (Rutoceoe) 11. The
megogometogenesis in several Tripsacum Megosporacyte collose in apomictic occurrelKe of adventive embryos without
species (Pooceoe) Amer. 1. 8at. 82 57-63. buffelgross, Kenlucky bluegross, pollination 01 fertilizatiOl1 Amer. 1. 801. 74:
Penniselum squamulalum Fresen, 517-30.
Tripsacum L and weeping lavegrass. (rap Wilms, H.J., J.L von Went, M. Cresti, and F.
5ci. 37 717-23. Gompalini. 1983. Adventive embryogenesis
in CiIIUS. (aryologio 36: 65-78.
Chapter 5
Genetic Analysis of Apomixis
ROBERT T. SHERWOOD
Chromosome pairing and disjunction usually The distinction between meiotic (= sexual) and
can be determined by examining apomeiotic (= apomictic) events becomes
microsporogenesis (Hignight et al. 1991; cytologically discernable after differentiation
Burson 1992). Pollen viability is determined of the megaspore mother cell (MMC) in all
using I2KI stain, fluorescein diacetate stain, or ovules (see Leblanc and Mazzucato, Chap. 9).
germination tests (Dujardin and Hanna 1989; In the normal monosporic Polygonum-type
Hill et al. 1989). Female fertility is estimated meiosis, walls of megasporocytes and
by determining percentage of seed set per 100 megaspores (tetrad cells) become invested
florets in open or self-pollinated inflorescences with callose. It is a simple matter to visualize
(Hignight et al. 1991). this cage-like indicator of meiotic activity
using fluorescence microscopy of intact,
Progeny Testing
aniline blue treated pistils (see Leblanc and
Progeny testing originally was practiced as
Mazzucato, Chap. 9). The fully differentiated
whole plant morphological comparison of
Polygonum-type sac is based on an 8-nucleate
progeny with the maternal parent. Broadly
speaking, comparative analysis of any trait scheme, with an egg apparatus, polar nuclei,
and antipodal cells.
66 loberl T. SWwtod
genes been located. No conventional length probes (RFLP) from a maize- Tripsacunt
morphological, agronomic, or physiological dactvtoide« F 1 population that cosegregated
traits are specifically associated with apomixis. with diplospory. They also were linked on the
The lack of linkage information is hardly long arm of chromosome 6 of maize.
unexpected given the obstacles to traditional
Biological Tests for Parthenogenesis
mapping in species that have the barrier to
Matzk (1991a) devised an auxin test for
crossing imposed by apomixis and for the most
detecting parthenogenetic capacity.
part are alloploids with indistinguishable
Unpollinated plants are treated with DIC;
chromosomes, irregular chromosome
2,-1-D; 2,-l,5-T; or CPAA. Parthenogenetic
duplication, and secondary economic status.
individuals form grains with a mature embryo
Conventional, unlinked, monogenically but no endosperm. Results are positive for
inherited traits have been used as markers to parthenogenetic mutants of nonapomictic
distinguish maternal from hybrid progeny. If species (barley, wheat) and for apomictic plants
the maternal parent is homozygous for a of apomictic species. The test can be used to
recessive trait, and the pollen parent is screen for parthenogenetic plants in sexual
homozygous dominant, uniformi ty of progeny species and to detect sexual plants in apomictic
for the maternal marker suggests maternal populations. It has proven useful in
inheritance. Homozygous or heterozygous characterizing POil pratellsis lines that vary in
dominant markers in the pollen parent have degree of facultative apomixis (Matzk 1991b;
been employed to reveal hybrid F1s(Hanna et Mazzucato et al. 1996).
al. 1970;Hanna and PO\,vell1973; Dujardin and
Naumova et al. (1993) described a cytological
Hanna 1989; Hignight et al. 1991).
test for quantitative analysis of
Isozyme polymorphism has been used to parthenogenesis in POI1 pratensie. Embryo sacs
characterize variability in apomictic parents were isolated mechanically and examined for
and progeny (Marshall and Downes 1977; spontaneous embryogenesis.
Hacker 1988; Cruz et al. 1989; Roy and
An ovule culture medium facilitated
Rieseberg 1989; Bayer et al. 1990; Kojima et al.
identification of apomixis in diplosporous
1991; Poverene and Voigt 1995; Gustine et al.
Allium t uberosum (Kojirna and Kawaguchi
1996; Berthaud, Chap. 2; Leblanc and
1989). Up to 80% of apomictic embryos, but
Mazzucato, Chap. 9).
no sexual embryos, showed development on
Several molecular markers that apparently are the medium.
linked with apomixis genes have been found
Combined Cytological, Progeny,
(Leblanc and Mazzucato, Chap. 9; Grimanelli Biological, and Marker Testing
et al.. Chap. 6) Ozias-Akins et al. (1993, 1998) When used alone, none of the progeny testing
and Lubbers et al. (1994) described a random methods discussed above can unequivocally
amplified polymorphic DNA (RAPD) marker establish the reprod uctive sta tus of every plant.
and a sequence-tagged site (STS) marker Two or more approaches applied together are
tightly linked with apomixis in Pennisetuni more informative (Naurnova et al. 1993;
species. Gustine et al. (1997) described two Mazzucato et al. 1996).
additional linked RAPD markers in P ciliate
and derived a preliminary linkage map of three Whole plant progeny testing views the end
markers with the apospory locus. Leblanc et product of seed formation and is the ultimate
al. (1995b) prepared three restriction fragment test of whether apomixis is functional.
Cytological examination of ovules during
681....,T.~
Brachiaria did not totally prevent selfing. diploids. If fully sexual tetraploids are not
Techniques have been published for available, they may be produced by various
emasculating small flowered grasses (Burson strategies. Colchicine treatment of sexual
1985,1992; Richardson 1958). diploids has created sexual tetraploids used
in hybridizations (Button and Forbes 1960;
4) Male gametocide. Young inflorescences of
Richards 1970; Savidan 1981; Miles and Valle
Penniseium ciliare were sprayed with a male
1991; Valle et at. 1991). Leblanc et al. (1995a)
gametocide (Bashaw and Hignigh 1990;
treated embryogenic calli of sexual diploid
Hignight et al. 1991).
Tripsacum with colchicine to induce
5) Self incompatibility. Self incompatibility of chromosome doubling; the regenerated
sexual female lines has been used to advantage tetraploid plants reproduced sexually.
in Tara;'(awm (Richards 1970),Paspalum noiatum However, when a sexual line of Paspalum
(Burton and Forbes 1960), and Hieracium hexastachium was doubled, the tetraploid was
(Gadella 1987). Dujardin and Hanna (1989) facultatively aposporous (Quarin and Hanna
used male sterile pearl millet (Pennieetum 1980). Asker (1967) started with an apomictic
glaucllm) as a female parent. Self- diploid (possibly a dihaploid?) biotype of
incompatibility is incomplete in guineagrass Potentilla argeniea and obtained a partially
iPanicum maximum); the degree of cross sexual tetraploid. Thus colchicine doubling
fertilization depends upon the procedure may reveal latent capacities for apomixis or
practiced for isolation (Savidan et al. 1989; sexuality.
Savidan 199Ob). Because genotypes can vary
Tetraploids can be created by hybridization.
in degree of spontaneous selfing, it may be
Harlan et al. (1964) selected a completely
necessary to use control tests to establish
sexual tetraploid of Bothriochloa grahamii from
reliability of each female parent (Matzk 1989;
a cross between two facultatively apomictic
Valle et al. 1991).
tetraploids. Savidan's (1981) crosses of
facultative tetraploids of genotype Aaaa
Reciprocal Crossing yielded sexual:aposporous (S:A) progeny in
Reciprocal crossing detects nuclear and the ratio of 1:3, indicating that the sexual
cytoplasmic maternal effects (including progeny were genotype aaaa. The genetic
matrocliny). The apomixis gene(s) has such a
analyses of Harlan et al. (1964) also used a
powerful maternal effect that geneticists have sexual tetraploid Dichanthium annulaium
been discouraged from using this technique,
accession that origina ted from a Bill
but apomixis need not be a deterrent to its use.
hybridization of an emasculated diploid sexual
It is feasible to use one or two facultative
plant with a tetraploid male. Burton and
parents in the crossing scheme (Savidan 1981).
Hanna (1986) grew diploid sexual Pensacola
Nogler (l984b) conducted reciprocal
bahiagrass iPaspolum notatum) in isolation with
backcrosses with a sexual diploid genotype as
an apomictic tetraploid to produce a triploid
the male or female in pairings with facultative
Bm hybrid. Open pollination of the triploid
lines. [assem (1990) used reciprocals in beets.
with Pensacola bahiagrass yielded ~II hybrids
Creating Tetraploid Parents at the tetraploid level. These facultatively
It is best if both parents are at the same ploidy apomictic lines could be used as maternal
level. The tetraploid level seems to be the parents in crosses and selfs to create sexual
natural milieu for expression of apomixis, tetraploids. Quarin (1992) successfully used a
whereas apomixis is rarely confirmed in similar scheme.
70 .oMrt T. SIotrw....
generation Sj plants of sexual parents were (1992). Reports issued since 1950 are
analyzed. Information from Fjs and SlS has summarized here, with a reinterpretation of
only limi ted power for testing alternative some of the results. Symbol "A" is used to
genetic models. Interpretations can be denote the dominant allele of the putative
strengthened using advanced generations, as apomixis gene regardless of the symbols used
in studies by Burton and Forbes (1960);Nogler in the original reports.
(1984b); Savidan (1981); and Valle and Miles
Monopolar Apospory (Gramineae-
(Chap. 10). Panicoideae)
Classification and Grouping Paspalum notatum (bahiagrass). Burton and
How should facultative progeny displaying Forbes (1960) crossed co\chicine-ind uced
various degrees of apomixis be grouped when sexual autotetraploid lines of Pensacola
testing segregation ratios? Current practice bahiagrass with an apomictic tetraploid line.
places all plants showing any apomixis into Progeny were classified only by whole plant
one group (deemed apomictic) and all plants progeny testing. The crosses produced
devoid of apomixis into a second group sexual:apomictic (S:A) segregation ratios near
(deemed sexual) (Savidan 1981; Voigt and 3:1.Selfing the sexual progeny gave an F2 with
Burson 1983; Miles and Valle 1991; Sherwood a ratio near 35:1.Selfing the apomictic progeny
et al. 1994).Most facultative plants produce far gave only apomictic F2 progeny. Burton (1992)
more apomictic sacs and progeny than sexual postulated an A gene dominant for apomixis
sacs or nonmatemal progeny, and classification and an independent 5 gene dominant for
is relatively straightforward. With only two sexuality. He assigned genotype Aaaassss to
groups being recognized, it is inevitable that MH, and genotype aaaaSSSS to the sexual
the genetic models are for simple gene action parents. Some progeny from selfing sexuals
with Mendelian interpretations. Unimodal appeared apomictic based on their uniformity
distributions of reproductive types indicative but actually may have been sexual; this could
of quantitative or polygenic inheritance have give rise to the 35:1 ratios. In the crosses giving
not been tested. 3:1 ratios, some facultatively apomictic hybrid
F j progeny may have been inadvertently
Testing Genetic Models
classified sexual, skewing the ratios in favor
All reasonable genetic models and
of high numbers of sexual progeny. More
permutations of genotypes should be tested.
recent data support the view that apomixis is
Predicted ratios should consider the effects of
coded for by dominant genes (Burton and
number of loci, dominance, epistasis, and
Hanna 1992).The bahiagrass system deserves
lethality. If the degree of alloploidy versus
reexamination using modern classification
autoploidy of tetraploids is unknown, both
methods.
disornic and tetrasomic inheritance and both
partial and complete random assortment of Dichan th ill m-Bothriochloa (b l uestem).
chromosomes or chroma tids shou Id be Harlan et al. (1964) crossed sexual tetraploids
considered (Sherwood et al. 1994). of D. annulatum and D. grall1llllii with apomictic
tetraploid pollen parents. Crosses of sexual x
Inheritance of Apomixis sexual provided only sexual Fjs. Crosses of
Litera ture on inheritance of apomixis has been sexuals x apomicts gave a S:A of 1:4.1. They
reviewed by Stebbins (1941), Gustafsson (1946- postulated random assortment of two disornic
19·m, Asker (1980), Nogler (1984a), Bashaw genes and assigned genotype A j a1A 2a2 to
and Hanna (1990), and Asker and Jerling tetraploid apomicts, and a ja ja 2a 2 to sexuals. If
72 •..., T. sw-d
we postulate that the tetraploids were nor any recessive gene models. Data were
tetrasomic with genotype AAaa, the observed compatible with a one-tetrasomic-gene model
ratios fit ratios expected for the assumption with apospory regulated by dominant allele A,
of random chromatid assortment (1:3.67) or under either of two assumptions:(i) random
the assumption of a recessive lethal effect of assortment of chromatids, or (ii) A acting as a
the A allele (1:4) (Sherwood et al. 1994). recessive lethal in gametes. Sexual plants were
assigned genotype aaaa; apomicts were Aaaa
Pennisetum ciliare (buffelgrass). A naturally
and AAaa. Data on linkage of apospory in
occurring tetraploid sexual plant, designated
Penniseium with molecular markers (Gustine et
B-1S, was selfed and crossed with two
al. 1997;see Grimanelli et al., Chap. 6) provides
aposporous biotypes by Taliaferro and
additional evidence that a single major locus
Bashaw (1966). The S:A ratios (near 13:3 for
regulates apospory in Pennisetum.
SlS and 5:3 for Fls) suggest two disomic
independent genes with the dominant allele Panicum maximum (guineagrass). Hanna et al.
of gene A being required for apospory, and (1973) reported that four naturally sexual
the dominant allele of gene B being epistatic tetraploid accessions produced SI progenies
to A and conferring sexuality. Sexual parent segregating in a combined ratio of 116S:54A.
8-1S was assigned genotype AaBa. Further Crosses of sexuals x apomicts gave 21S:28A.
evaluation of Fls, F2s, and a BCI from 8-1S, They proposed a digenic, disomic additive
identified true breeding sexual progeny of model using the assignment of AaBb for the
apparent genotype aabb and apomictic plants sexual plants and Aabb, aaBb, or aabb for
of putative genotypes Aabb and AAbb (Read apomicts (two dominant doses required for
and Bashaw 1969;Bashaw et al. 1970). sexuali ty).
Crane (1992) proposed a tetraploid single gene Savidan (1981) crossed a colchicine-induced
model to explain the Taliaferro and Bashaw autotetraploid sexual plant and a natural
segregations. Three alleles were postulated: a apomictic tetraploid. Ten kinds of crosses were
(wild type sexual), A (aposporous), and A+ tested (Table 5.1). All the data fit perfectly with
(super sexual). Only genotypes AAAA, AAAa,
and AAaa would be apomictic. Chromosomal Table 5.1 Segregations for mode of reproduction in
segregation patterns were postulated to be 10 crosses of Ptmicum IIHIXimum (Savidan 1981;
preferential and to differ during Savidan et aL 1989)
microsporogenesis and megasporogenesis. sexual x opomidic crosses sum apo sex
Fl hybrids: Six Al 133 71 62
Sherwood et al. (1994) studied inheritance of 3-woy hybrids: (S1xAIlsex xA2 279 135 144
embryo sac type of sexual tetraploid plant B- Backcross: (S1xAIlsex xAl 26 14 12
25. From open pollinated B-25, five sibling Backcross: sexuol 3-woy hybrid xA2 170 73 97
sexual 3-woy xapomictic 3-woy 60 26 34
sexual lines and five sibling aposporous lines
Backcross: Six (S1xAllopo 23 13 10
were selected as parents. Segregations were 10101 sex xapo crosses 691 332 359
determined for crosses of sexual x aposporous
sexual x sexual crosses (or selfed)
lines and sexual x sexual lines, and selfs of sexual Fl hybrids selfed 126 0 126
sexual lines. Selfs and crosses of sexual plants sexual 3-woy hybrids seKed 57 0 57
gave only sexual progeny. Fls from sexual x sexual 3-woy xsexual 3-woy 82 0 82
aposporous combinations segregated for S:A 10101 sex xsex crosses 265 0 265
at ratios near 15:13 for four aposporous lines apomictic x apoaUrtK crosses
and 1:2.8 for the other line. Segregations did apomictic 3-woy xapomictic 3-woy' 71 53 18
not fit any one- or two-disomic gene models, •• ana~is mode of off-types (IIlOternol types nol counted)
the hypothesis of one tetraploid gene Bipolar Apospory
dominant for apomixis, with all sexual parents Ranunculus (buttercup, Ranunculaceae).
assigned genotype aaaa and all apomicts Nogler (1984b) crossed diploid sexual R.
assigned genotype Aaaa. cassubicijolius with tetraploid apomictic R.
megacarpus (Figure 5.1). Four fertile triploid
Brachiaria (Gramineae). Valleand coworkers
facultatively aposporous progeny were
(1991, 1992, 1993, and Chap. 10) conducted
obtained and used to initiate three generations
extensive studies of Brachiaria along the lines
of reciprocal backcrossing to the sexual diploid.
of Savidan's (1981) guineagrass program. The
Nogler deduced that apospory is caused by a
results pointed to a single dominant gene
dominant factor A (designated A- in Nogler
determining apospory with genotypes of aaaa
1984b), the wild allele of which (a) (designated
for the colchicine-induced tetraploid sexual
A + in N ogler 1984b) does not enhance
parents and Aaaa for the aposporous
apospory. Dominance of A is incomplete and
tetraploid parents.
additive. Furthermore.A, when homozygous
p
rn levelof
polyploidy 2x
4x
fl
d' AAaa
I plant."
I la I a] obtained
Aaa
D
B n hybnd: n- n
dihaploid: n .0
Bul hybrid: ~n n
maternal: 2" 0
aa (A I a ] •• a Aaa I a. aI a AaI (mal.) • trisomic hybrid: ~n= 17
9·~9~E?Q~ f.
.,.~
BC,
42
E2~r "
BC~
EJ
aa
13
17
Aa
rn·~·~:.·.·.·
0
(mat.)
5
72
o[] ~,~/ ~ tJ ~ ~ n
, IS
38
16/
OIHAPLDIDS
2.1 + 5 veg.
:,a
aaa
66
A aa (mal.) A a a a a
29 U
~~~ I a A a I a A a {mat.} A. a a
EJ DDD~'::":D
, 61 0 / 16 0 31 20
6.1
Figure 5.1 Genealogical tree ofthe aoss RanlRlcullls CGSSllbicifoliu S =c,. 2% =16, meiotic (sexual) x R.
megacarpllS =M, 4% =32, partially aposporous C-totaUy" apomictic) and the different backaosses with
the sexual parent Co The number of plants obtained, the level ofpolyploidy, the approximate degree of
apospory, and the genotype are incficated for each offspring.
Reproduced with permission 01 the publishel, Birkhouser-Yerlog AG, B~el, Switzerland. From Nogler, G.A. 19B4. Genelia 01 apospory in aponiclic
Ronuncu/us OuriCOlllUS. Y. Conclusion. Boronico HelYeli<o 94: 411-422. Updated by G.A. Nogler (personol comm.) ond NoglerIl99S). See Nogler
(1984b) for deroils.
74 Ro~.rt T. Sllorwood
in the g arnet op hy te. is lethal to the Poa pratensis (Kentucky bluegrass, Poaceae),
gametophyte: there are no functional A, AA, or There Me several inconclusive and
AAA gametes because of recessive lethalitv. contr adict or y reports on inher i t ance of
Cametes must carry the wild type a allele to be apomixis in Kentucky bluegrass. Bluegrass
viable. Several lines of evidence point to populations have complex arrays of
lethalitv, including failure to find ilPOSPOrous polyploidy and aneuploidy, with chromosome
diploid hybrid progeny. A highly ilposporous numbers ranging continuously from 2-1 to 12-1.
211 + 1 aneuploid. line 1, was believed to have In addition, somatic cells within plants milY
originated from an a egg and an Aa sperm and vary by as milny as 30 chromosomes (Huff
have genotype Aaa. Line T transmitted a and 1992). Plants vilry Widely in the degree of
A'l gilmetes, but not A gametes. The data. for apomixis. Almgard (1966) concluded that the
the most p art. did not permit testing of presence of aposporous embryo sacs showed
segregation ratios, but cross aa x Aaaa and its dominant inheritance, but retention of the
reciprocal did yield 1:1 ratios as expected. In maternal phenotype (functional apomixis)
trisomic apomictic lines assigned genotype Aaa, was recessive.
there was declining strength of apomixis with
Matzk (1991 b) explored regula tion of the
advance from BCl to BC,. Nogler believed this
parthenogenetic capacity of P. pmtellsis using
might be due to modifying genes for sexuality
the auxin test. He concluded that the apomictic
that increased with each generation of
parents were heterozygous for one or more
backcrossing, Within generations there was a
dominant alle les for parthenogenesis. The
d osag e effect; for example genotype Aaa
results were consistent with expectations for
showed greater apospory than Aaaa. Some
a single major gene.
plants were identified as d ihaploids with
genotype Aa and were highly aposporous. Study of the genetic regulation of facultative
Dihaploids are formed by parthenogenetic expression (Noglers modifying genes for
development of reduced Aa eggs of sexuality) milYbe feasible using molecular and
facultiltively apomictic AAall or Aaall biological probes developed for P. prateusi«
tetraploids: in other species, dihaploids usually (Huff and Bara 1993; Naumova et ,11. 1993;
are sexual (Nogler 1984,1). Mazzucato et ill. 1995; Mazzucato et ill. 1996).
There is also an interesting report of
Line T expressed low parthenogenicity. Nogler
Hieraciurn-type embryo-sac formation in
(198-1b) postulated that parthenogenicity was
diploid specimens of three Tribolium species
coded by il separate gene closely linked to A.
(Poaceae) (Visser and Spies 199-1).
Later tests revealed that apomictic line T
transmitted genes for par thenogenicitv. Hieracium (hawkweed, Asteraceae). From a
Therefore, it was not necessary to insist that cross of diploid sexual H. auricuta x tetraploid
apomixis and parthenogenesis were coded by apomictic H. IlllrtJlltillClIIIl, Christoff (19-12)
separate genes (Nogler 1989, 1995). Noglers found 32 sexual and 27 apomictic progeny, a
(19S-1b) study encountered all the problems ra tio tha t fits the expected 1:1 if the ma le has a
anticipated from interspecific crossing at single dominant allele for apomixis. Cadella
different ploidy levels-poor fertility, poor seed (1987) believed results from il cross of
set, facultative expression, aneuploidy, and tetraploid sexual H. pilose/lll with a pentaploid
ambiguous segregation ratios-but succeeded apomict could be explained by monogenic
because of exceptional effort and inSight. dominant inheritilnce with the pentaploid
Gooeti< """'lis .1 a...... is 75
Chap. 8); Koltunow et al. (1995) believe progeny (21 plants) were sexual. In the
Hicracuun can be a model system for studying reciprocal. most progeny, as expected, were
molecular genetics of apomixis. maternal or polyploid apornicts, but four
diploid sexual F1s were formed (Gerstel et al.
Mitotic Diplospory
1953) They concluded that apomixis genes
Most reports on diplospory. summarized
acted recess ivelv but additively and
below, were handicapped by lack of suitable
postulated that polyploids with two apomixis
sexual parents and by unavailability of
genomes and one sexual genome were
convenient classification techniques.
apomictic.
Eragrostis curuula (weeping lovegrass,
There is an al ter nate interpretation. The
Poaceae). Crosses of naturally occurring
st,lfting materials and results with Parthcniunt
tetraploid sexual plants with tetraploid and
resemble those of Nog\er (1984b) for
hexaploid apornicts gave F j progeny test
aneuploid plant T of Ranunculus (Figure 5.1).
segregations indicating that apomixis IS
The Partlicnium results may indicate the same
monogenic and dominant (Voigt and Burson
control as in Ranunculus, i.c.. a single factor
1983). Results of a cross with an aneuploid
dominant for apomixis, that acts as a recessive
indicated possible dosage effects, but m,ly
lethal, with the polyhaploid parent being
have been confounded by Bill hybridization
genotype Aaa and the diploids being aa.
or chromosome elimination.
Restitutional Diplospory
Tripsacum dactyloides (Eastern gamagrass,
TaraxaclIIII (dandelion, Asteracae).
Poaceae), Sherman et al. (1991) crossed a
Eutriploids (21/ = 24) and many hypotriploids
sexual diploid female parent with a triploid
(211 = 23) have facu Itative diplospory. Certain
apomict. Forty-six hyperdiploid progeny were
211 = 23 aberrants are primarily scx u.il
identified as hybrids. All but two of these
(Sorensen 1958). Mogie (1988) offered the
showed cytological indications of apomixis;
following interpretation of earlier studies.
the degree of diplospory ranged from
Expression of apomictic phenotype in
predominantly sexual to highly apomictic.
TtlrtJxaclIllI depends on one or more genes
The authors believed this indicated that
located on one chromosome and on dos,lge.
apomixis is incompletely dominant, or that
At least two copies of the mutant apomixis
minor additive genes on various
,11/1'11' ilrl' required to obtain apomixis; the
chromosomes affect penetrance of apomixis.
al lc h- prevents meiosis in diplosporou-.
Recent production of sexual tetraploids by
a porn icts. The dominance rc la t iorish i p
colchicine doubling should facilitate future
between the wild type and mutant allele is
study of inheritance. The tetraploid T
dch-rrniru-r] by balance and envrronrnent.
aactvtoutc« parent genotype is simplex for the
76 •...,T.sw-.I
Environment plays a role in expression The Lethal Gene as the Basis for
(Nogler 198-la; den Nijs and van Dijk 1993). Heterozygosity
A short photoperiod increases the frequency Nogler (198-lb) concluded that functional
of aposporous vs. sexual embryo sacs in gametes of RallllllCIIlIIs contain a copy of the
Dicuantliiu m arisi atu m (Knox 1967) and wild type a allele. Noirot (1993) reviewed
Paspalum cllrol1lyorrhizoll (Quarin 1986). Salt evidence that the A allele may act adversely
stress affects facultative expression in P. ciliare in Panicuni I1JaXIIIlllnl; his report focused on
(Gounaris et al. 1991). male and female sterility. Male and female
sterility is encountered in spontaneous
Harlan et al. (196-l) accurately assessed the
dihaploids and trihaploids of Pall/CIIIIl of
relation between sexual and aposporous
putative genotypes Aa and Aaa, respectively
reproduction. "Apomixis (read apospory)
(Combes 1975).Mogie (1988)proposed that the
and sexual reproduction are not alternative
wild type a allele has a function that is essential
modes of reproduction, either genetically or
to normal plant processes. If gametophytes
operationally, but are simultaneous and
and gametes bearing dominant allele A must
independent phenomena. The genes
also bear the wild type a allele to remain
controlling normal sexual reproduction are
functional, this would account for the
not allelic to those controlling apomixis in the
observation (Harlan et al. 196-l) that
con ve n tional sense." This accoun ts for
aposporous apomicts invariably are
f acul tat ive expression of apospory. In
heterozygous at the apomixis locus. No
aposporous lines, meiotic reduction of the
apomict has ever transmitted an exclusive
archesporia I nucleus and apomeiotic
capacity for apomixis to the offspring; a
induction of apospory in nucellar cells go
capacity for sexual reproduction is always also
forward at about the same time. Aposporous
transmitted, although it may not surface until
initials and unreduced embryo sacs normally
later generations. Heterozygous Aa gametes
crowd out the reduced megaspores and sacs.
can be formed by dihaploids, triploids, or
Facultatively displosporous plants also tetraploids. In the case of the dihaploid. the
possess all of the genetic information required Aa gamete is from an unreduced (apomictic)
for completion of both meiotic and sac, and the progeny are either maternal (no
apomeiotic embryo sacs. However, in contrast fertilization) or Bill hybrids (fertilization), as
to facultative apospory, the two events cannot shown for the dihaploids of RallllllClIllIs
proceed simultaneously in a facultatively (Figure 5.1). In the case of triploids and
dip losporous ovule, for they compete for the tetraploids, the Aa gamete occurs in sexually
same si te in the ovule. Events beginning in reduced (meiotic) sacs, and the egg is usually
the megaspore mother cell can proceed only fertilized (BII hybridization), but occasionallv
towards normal meiosis or apomeiosis, but may parthenogenetically form an Aa
not both. Variability within facultative dihaploid as part of a diploid-
diplosporous or aposporous types indicates tetraploid-dihaploid cycle (de Wet and Harlan
that the entire apomictic developmental 1970; Savidan and Pernes 1982). Insofar as
process cannot be explained on the basis of a sexual transmission of A is concerned, the
single gene (Grimanelli et al. 1995; Savidan parent must be polyploid and heterozygous;
et al. 1995). for asexual transmission, the parent must be
heterozygous.
Ge..1ic AoaIysis .f Apoooi.is 79
- - . 1980. Gametaphyli< apomixis: BosOOw, E.C., and K.W. Hignight. 1990. Gene
References elements and geneli< regulotion. Hereditas transfer in apomictic buffelgrossthraugh
Almgard, G. 1966. Experiments with Poa. Ill. 93: 277-93. fertilization of an unreduced egg. Clop 56.
Further studies of Poo Iongifolio Trin. with Asker, S.E., and LJerling. 1992. Apomixis in 30: 571-75.
special reference to its crOlS with Poa Plants. 8aca Ratan, Florida: CRC Press. BasOOw, E.C., and E.c. Halt. 1956.
protensis L umtbrukshogskolon Annales 8asOOw, H. 1962. Apomixis and sexualily in MegO\flOlagenesis, embryo IQ(
~2: 3-64. buHelgrOls. (rap Sei. 2: 412-15. development and embryogenesis in
Asker, S.E. 1967. Induced sexuolily aher - - . 1980. Apomixis and its app~cDlian daltllQrass, Pospolum dilatotum, Pair.
chrOl1lOlOme doubling in an apomictic in crop improvement.. In W.R. Fehr and Agran. 1. 50: 753-56.
Paten/ilia argenleo-biolype. Hereditas 57: H.H. Hodley (eds.), Hybridizotion of(rap 8asOOw, E.c., A.W. Havin, and E.C. Halt. 1970.
339--42. Plonts. Madisan, Wisconsin: ASA and CSSA. Apomixis, its evolutionary significance and
- - . 19700. Apomixis and sexuality in the Pp 4>-63 utilization in plant breeding. PrO(. 11th 1n/.
Palen/illo argenlea complex. I. Crosses with BasOOw, H., and w.w. Hanna. 1990. Apomictic Gross/. (angr. Pp. 24S--48.
other species. Hereditas 66: 127--44. reproduction..In G.P. Chapman (ed.], Bashaw, H., M.A. HllIsey, and K.W. Hignight.
- - . 1970b. Apomixis and sexuality in the Reproductive Versatility in the Grosses. 1992. Hybridization (N+N and 2N+Nl of
Polen/illa argenlea complex. 11. Crosses CaniHidge U.K.: Cambridge Universily facu~ative apomictic species in the
within the complex. Hereditas 66: 189- Press. Pp. 100-30 Penniselum agamic com~ex. 1nl. 1.PIonl
204. Sri. 153: 466-70.
Banaglia, E. 1963. Aponnis. In P. Maheshwlri ChrisIolI, M. 1942. Die genefische grundloge der Grimonel5, D., D. Leblonc, D. Ganzalez deleon,
(ed.], Recent Advances in the Embryology aporrik1isdlen Fortpflonzung bei Hierocium ond Y. Sovidan. 1995. Mopping apomixis
ofAngiospemrs. DeIJi, India: 1nl. Sot of ourontiocum. Zeitschrih fur Inductive in tetroploid Tripsocum, preliminary resuh5.
PIont ~, Univ. of Deft1i. pp. Abstuntnllngs·und Vererbungslehre 80: Apomixis Newsiener 8:37-39.
221-64. 103. GUSlakson, A. 1946--1947. Apomixis in higher
Bayer, H, K. Ritbnd, ond B.G. Purely. 1990. Combe5, D. 1975. Polymorphisme et modes de planll. I.The mechanism of apomixis.
Evidence of portial opomixis in AntennoriJ reproduction dons la section des Moximoe Lands Univ., Amlcr. N.F. AIlv. 242(31:1-
media IAsleroceae:lnuleael dete<led by the du genre Ponicum (Granineesl en Alrique. 66. 11. The cousal ospecI of opomixis. ADv.
segrego1ion of genetic morkenl. Amer. J. Mfimoires OmOM 77: 1-99. 243(21:69-182.11I. Biotype and species
Bot. 77: 107~3. Crone, U 1978. Aponixil and crOlsing formotion. AIlv. 243(12): 183-370.
BickneH, R. 1994. Iflerociurrr. Amodel system inCOf1'4lO1ibit11ies in some Zephyranthaceae. Gusline, D.l, R.T. Sherwood, and D.R. Hull.
for sludying the molecular genetics of Ph.D. thesis, Univ. of TexOl, Austin, Iexes, 1997. Apospory·~nked molecular morken
aporrixis. Apomixis NewsJener 7: 8-10. - - . 1992. Arationale for the in bullelgrOls. (ropSci37: 947-51.
Bicknen, RA, and N.K. Borst. 1996. Isolation of investigation of certain wild aporricll. In Gusfine, D.l, R.T. Sherwood, Y. Gounaris, and
reduced genatypes ofHitHocium pilosello Bgin, LH, and lP.Miklche (eds.) Proc. D.R. Hull. 1996. Isozyme, protein, and
using anathet' culture. l'Iont (el, flSsue and Apomixis Workshop, February 11-2, 1992, WOmorkers within 0 half-sib fani~ of
Organ (uitlIre 45: 37-41. Atlanta, Georgia. USDA, AR'). ARS-104. pp. bullelgross segregating for opospory. (rop
Brown, W.V., and W.H.P. &nery. 1958. Aponixis 58-61. Xi. 36: 723-27.
in the Gramineoe: Panicoideoe. Amer. 1. Crone, U, and lG.Carman. 1987. Mechanisms Hacker, lB. 1988. Sexuality and hybridization
BoI. 45: 253--63. of apomixis in Bymris rectisetus from in signal gross, Brochiorio decuniJens.
BurlOn, B.lI985. Cytology of Pospolum eos1ern Australia ond New Zealand. Amer. TropiCDI Grosslonth 22: 139-44.
chocoence and P. durifoNum and their Hot. 74: 477-96. Hair, lB. 1956. 5ubsexual reproduclion in
relationship to P. diIototum. BoI. Goz. 146: Cruz, I.,lW. Miles, W. Roca, and H. Romirez. Agropyron. Heredity 10: 129-60.
124-29. 19B9. Apomixis and sexua~ty in Brochiorio. Honna, W.W., ond 18.Powel!. 1973. Stubby
- - . 1992. Cytology and reproductive I. Biochemicallllxlies. (uban J.Agric Xi. head, on induced focuhative aponicl in
behovior of hybrids between Prtspolum 23: 317-21. pearl I11tllel. (rop Xi. 13: 726--28.
urvillei and two hexaploid P. dilototum den Nijs, A.P.M. 1990. Experimenting with Hanna, W.W., J.B. Powell, le. Millal, and G.W.
biatypes. IienooJe 35: I 002~6. aporrixis and sexuality in POD protensis. Burton 1973. Cytology of obt,ate sexual
Burlon, G.W. 1992. Monipuloting apomixis in Apomixis NewsJeffer2: 52-54. planll in Ponicum maximum Jocq. and
Prtspolum. In lH.Bgin and lP.Miksche den Nijs, A.P.M. and G.E. vun Dijk 1991 their use in cantrolled hydrids. (rop Sri 13:
leds.), Proc. Aponixis Workshop, February Apomixis.. In M.D. Hayward, N.D. 695-97.
11·12, 1992, Adanta, GA. USDA, AIS. Bosemark, and I. RomogOlO (eels.), 1'I0nt Honna, W.W., KJ5chertz, and E.e. Bashaw.
ARS-104. pp.16--19. Breeding: Principles and Prospects. Landon: 1970. Apospary in Sorghum bicolor ll.)
Burlon, G.W., and I. Forbes. 1960. The geoelics Chopmon and HoII. pp. 229-45 Moendl. Science 170: 338--39.
and manipulation of obIigote apomixis in de Wet, lMJ., and J.R. Horlan. 1970. Aponnis, Harlon, JR., M.H. Broalcl, D.S. Borgaanlcar, and
convnon Bahio gross IPospolum IIOtulum polyploidy, and lpeCiotian in Oichonthium. lMJ.de Wet. 1964. Nature and
Auggel. Proc. BIb 1nl. Grossl. (ongr. Evolution 24: 270-77. inheritance of apomixis in Bothriochloo and
pp.66--71. Oujardin, M., and W.W. HOMO. 1989. Oicbonlhium. Bol. Goz. 125: 41-46.
Burtan, G.W., and W.W. Honno. 1986. Bahia Developing aporri<lic pearl millel- Herr, lM. 1971. Anew dearing-lqumh
gross (PospoIum IIOIDhlml letroploids choroderizalion of 0 B~ plant. 1. Genel. ledmique for the ltuely ofovule
produced by making laporriclic tetraploid Breed. 43: 145-51. developrnent in angiOlpelI11l. Amer. J. Bot.
xdiploid) xdiplaid hybrids. Crop Xi. 26: Fisher, W.D., E.e. Bashaw, and E.e. Holt. 1954. 58: 785-90.
1254-56. Evidence for apomixis in Penniselum 61iore - - . 1982. An analysis of methods for
- - . 1992. Using apomiclic letra~oids 10 and Cenchrus setigerus. Agron. J. 46: 401- permonenl~ mounting ovules in four-ond-
make 0 self-in~ible ciploid Pemo<oIa 04. a-ha~ type clearing fluids. Slain Techn. 57:
Bahiogross done set seed. J. Hered. 83: GadeIIo, T.WJ. 1987. Sexual tetraploid ond 161-67.
30~6. aponidic penlo~oid popuIotions of Hignight, K.W., E.e. Boshow, and M.A. Hussey.
Carman, 1 1997. Asynchronous expression of HitHocium piIocello (Compositoel.l'I. SysI. 1991. Cytological and morphological
duplicate genes in angiosperms may cause Evol 157: 219-46. diversity of native apomiclic buIIeIgrO>l,
aporrixis, bispory, tetraspory, and Gerstel, D.U., B.l Hommond, ond e.Kidd. 1953. Penniselurn ciliore Il.) link. BoI. Goz. 152:
polyembryony. Bio/. J. ofthe oon. Sac. 61: An additional nole an the inheritance of 214-18.
51-94. aponixis in guayule. BoI. Goz. 115: 89- H~I, u, M.l Dahmer, MA Hussey, and 0.0.
Carman, lG., U Crone, and D. Riero-lizorazu. 93. Dean. 1989. Evaluolian of lechniquel for
1991. Compcrolive histology of eel wuls Gerslel, D.U., ond W. Mishonec 1950. On the assessing pollen viability in (enchrus cilioris
during meoIic ond opomeiolic inherilonce of aporrixis in Porthenium l Agronomy Abstracts 1989: 85.
megasporogenesis in two hexoploid. orgentatum. BoI. Gaz. 112: 96--1 06. Hull, D. 1992. Apomixis in POD•.In lH.Bgin
Aus1Ta~an E1ymus species. (rop Sci. 31 : Gounoris, E.K., R.T. Sherwood, I. Gaunaris, R.H. and lP.Milclche (eels.), Proc. Aponixis
1527-32. Honihan, and D.lGusline. 1991. WOfhhop, February 11-12,1992,
Inorganic lCIhs modify embrya IOC AIlanta, Georgia. USDA, AR'). ARS-l 04. pp.
development in sexual and aposporous 20-25.
(enchrus cioris. Sex.l'lont Reproo. 4:
188-92.
Gooolit ,"""lis 01 A"'Ii. 81
Huff, R.D., and J.M. Bara. 1993. Delermining --.19910. AnOYeI oppraodJ la Ozios-Akins, P., E.L Lubbers, W.W.lIonna, md
genetic arigils of abenonl progeny from dilferentialed embryos in the absence of J.W. McNoy. 1993. Transmission of the
hKuhatiYe ~ Kentucky b1uegrms endosperm. Sex. l'Ion1 Reprod. 4:88-94. opomiclic made of reproduction in
using acombinatian of flow cy1ametry and - - . 1991 b. Hew effoolla overcame Penniselurrr. c~inheritonce of the tro~ and
silver-stained WO markers. Theor. AppI. apomixis in Pea prulemis L Euplrytico 55: maIeruIor rrakers. Theor. AppI. Genel. 85:
Gene!. 87: 201~8. 6>-72. 632-38.
Jassem, 8.1990. Apomixis inthe genus BelD. Mazzucoto, A. 1996. Which gene(s) are we Peacock, WJ. 1993. Genetic engineering and
Apomixis 1Iew5Ietter 2: 7-23. laoKing far? Apomixis Newsle"er 9: 7. mutagenesis far apomixis inrice. In KJ.
Jeffersan, RA 1993. Strategic development of Mozzucoto, A., 6.Barocoo, M. Pezzani, and M. Wi1son (ed.), Proe. Internotiorrd Workshop
apomixis as a general taollar agricuhure.. Fakineli. 1995. 8iod1emicol and malecular on Apomixis in Rice, Jan. 13-15, 1992,
In KJ. Wilson (ed.), PrO!. IntematioooI markers far investigating the made of Changsha, China. Canberra, Australia:
Workshop on Aponixis in Rice, Jan. 13- reproduction inthe focut1olive apomict POD CAM81A. pp. 11-21.
IS, 1992, Changsha, China. Canberra, prutensis L Sex. Plant Reprod. 8: 133-38. Paverene, M.M., and P.W. Vaigt. 1995.
Australia: CAM81A. pp. 206-17. Mozzucoto, A., A.P.M. den Hill, and M. Falcineln Identilicocion de hibridos opomic1icos a
Knox, R.8. 1967. Apomixis: seasonoI and 1996. Estimolion of porthenagensis sexuoles de Erogrostis ClKVIJIa (Sdvod.l
populatian dilferences inagrass. Science frequency in Kentucky bluegrass wilh Hees medionte 0n05sis isoenzimolico.
157:32>-26. auxin-induced porthenoearpic seeds. Crop Mencle/iono 11: 29-36.
KalIunaw, A.M., RA 8icknell, and A.M. Sci. 36: 9-16. Ouorin, CL 1986. 5easonoI changes in the
Chaudhury. 1995. Apomixis: maIeruIor Mozzucoto, A., M. Wogenvoort, and A.P.M. den incidence of apomixis of diplaid, triploid,
strategies far the generation of genetical~ Hijs 1994. Row cy1amelrk onaiy5es la and letraploid plants of PMpolum
identical seeds wiIIlout fertilizalian. l'Iont estimate the made of reproduction in POD cromyorrlrizon. Evplrytico 35: 51 >-22.
I'/rysiol. 108: 134>-52. prulensisL Apomixis NewsJe"er7: 22-24. - - . 1992. The nature of apomixis and its
Kajima, A., and 1 Kawagudi. 1989. Apomictic Miles, lW., and C8. do Volle. 1991. Assessment arigin in Ponicaid grosses. Apomixis
nature of Chinese dive (AtUII nJJerosum of reproductive beiJovior of interspecific NewsJe"er 5: 8-15.
Rattl.) deteded by unpoIinated ovule 8rochiorio hybrids. Apomixis Newsle"er 3: Ouarin, CL, and W.W.lIonn. 1980. Effect of
ruhure. lapJ.Breeing 39: 449-56. 9-10. three ploidy levels on meiosis and made of
Kajima, A., Y. Nogata, and llinata. 1991. Mogie, M. 1988. Amadel far lhe evolution and reproduction inPMpolum hexoslDdryum.
Degree of apomixis in 0Wlese chive cantral of generative apomixis. Bio/. l. Crop Sci. 20: 69-75.
(AJlium nJJerosuml estimoted by esterase Unn. Soc. 35: 127-53. Reod,J.C, and E.t8oshcrw. 1969.
isozyme analysis. lap. 1. Breeting 41: 73- Hokojimo, l, and H. Mochizuki. 1983. Degrees (ytatoxononic relationship and the role of
84. of I8xUOIi!y insexuol plants of guinea apomixis in SjMl(io1ion inbuIIelgrms and
leblOlK, 0., M. Dueiios, M. Hemcindez, 5.Bello, grass by the si~~1ied embryo soc birdwoodgross. Crop Sci. 9: 805416.
V. Garcia, J. 8ertllaud, and Y. XMdon. analysis. lap. 1. Breeding 33: 4>-54. Richords, A.J. 1970. Eutriploid focut1otive
I99Sa. Ovamosome doubling in Houmava, 1,A.P.M. den HIli, nnd M.T.M. ogomaspermy in Taraxacum. New
Tripscwrrr. the production of artifioal, WiBernse. 1993. OuontilOlivt analysis of Phytologist6t. 761-74.
sexuoltetraploid plants. Plant Breeding aposporous porthenogenesis in POD - - . 1996. Why is gometophytic apomixis
114: 226-30. prolensis genotypes. Ado Bol. Neerlond. almost restricted to poIypIaids? The
leblanc, 0., D. 6rimanelli, D. Ganzlilez de Leiln, 42: 299-312. gometaphytHxpressed lethal model.
and Y. Savidon. 1995b. Deledian of the Nogler, 61. 19840. Gametaphytic apomixis..In Apomixis Newsletter 9: 3.
apomiclic made of reproduction in maize- 8.M. Jalvi (ed.!. Embryology of Richordsan, W.L 1958. Ated1nique of
Tripscwm hybrids using maize RFlP Angiosperms. 8erlin: Springer-Verlog. emasculating smal grass flare!\. Indian 1.
markers. Theor. AppI. Genel. 90: 1198- pp.47>-518. Genel. l'Iont Breed 18: 69-73.
1203. - - . 1984b. Genetics of apospory in Ray, B1., and LH. Rieseberg. 1989. Evidence
leblanc, 0., M.D. Peel, J.6. Carman, and Y. apomictic Ranunculus ouricol1XK. V. far apomixis in Arobis. l. Hered. 80: 506-
Saman. 1995<. Megasporogenesis and (andusian. Bot. Helvelico 94: 4' 1-22. 08.
megagometogenesis in several Tripscwm - - . 1989.(ytagenetics of Ru!ishouser, A. 1969. Embryologie und
species (Paoceoe). Amet. 1.BoI. 82: 57- porthenogensis - first resuhs an forlpflonzungsbiologie der CIIgiospermen.
63. RolHRICulus ouricomut Apomixis Wllln: Springer.Verlag.
tittefars, A. 1955. (ytalagical studies in Sorbus. Newsletter 1: 44-47. Saman, Y.H. 1981. Genetics and utizolian of
Acto Ham Bf!f9ioni 17: 46--113. - - . 1990 Simptllied methods far apomixis for the improvement of
Lubbers, E.L, LArthur, W.w. Honna, and P. embryalogkal studies. Apomixis Newsle"er 6uineogrms (I'r1nicum moxilOOrrJ Joeq.).
Ozios-Akins. 1994. Molecular markers 2: 56--58. Proe. XIV Int. Grossl. (angr., lexington,
shored by diverse opomidic Penniselum - - . 1995. Genelics of apomixis in Kentucky. pp. 182-i4.
species. Theor. AppI. Genet. 89: 636--42. Ranunculus OurKOlOOt VI. Epilogue. Bol. - - . 19820. Embryologkol analysis of
Morsholl, D.R., and R.W. Downes. 1977. Atesl He/velico 105: 111-15. facuhalive apomixis inPonicum moxilOOrrJ
far abligole apomixis in groin sorghum Hairol, M. 1993. AI1etK ratias and sterility in the Jocq. Crop Sci. 22: 467 ~B.
R473. EvplJytico 26: 661--M. agamic complex of lhe Moximoe - - . 1982b. Nature elllered~e de
Motzk, F. 1989. Genetic studies an (Pankaideoel: evalutionary role of the raponixie chez Ponicum maximum locq.
parthenogenesis in POD prulemis L residual sexuality. 1. &0/. Bioi. 6: 9>-101. Tro'IUux e! Docurnents. ORSTOM 153: 1-
Apomixis Newsle"er 1: 32-34. 159.
82 R"'rt T. Silo<wood
- - . 1989. Apomixis in plant breeding: Sherman, R.A., PW. Vaigt, B18urSO/l, and Ll, Valle, tB.de, G. Leguizamon, and N.R. Guedes.
transfer vs. synthesis. Apomixis Newslelter Dewold. 1991. Apomixis in diploid xtriploid 1991.Interspecilic hybrids of Bromiorio
1:22-24. Tripsorum dOdyloides hybrids. Genome 34: (Gramineoel. Apomixis Newslelter 3: lO-
- - . 19900. The genetic cantral of 528-32. ll.
apomixis. Apomixis Newslelter 2: 24--27. Sherwood, R.T. 1995. Nuclear DNA amount Valle, tB. do, ond lW. Miles. 1992. Breeding 01
- - . 1990b. Commentary an the during SIlOIogenesis and gametogenesis in apomictic spedes. Apomixis Newslelter 5:
preteding contribution. Apomixis sexual and apaspolaus buffelgrass. Sex. 37-47.
Ne"lnJelter 2: 51. Plont Reprod. 8: 8S-90. V"lSser, N.t, and JJ. Spies. 1994. Cylogeneti<
--.19910. La cr~ique est esee maisl'art Sherwood, R.T., tt Berg, and 8.A. Young. 1994. studies in the genus Tribolium (Pooceoe:
est dillicile. Apomixis Newslelter 3: 21r-27. Inheritance af apospory in buffelgrass. Crop Donthanieoel.lI. Areport an embryo soc
- - . 1991 bDo we need aclassificatian of xi. 34: 1490-94. development, with special reference to the
apomixis? Apomixis Newslelter3: 27-29. Sherwood, R.T., BA Yaung, and E.tBoshaw. accurrelKe of apomixis in di~oid
- - . 1992. Genetic cootrol and screening 1980. Focuhative apomixis in buffe~ross. specimens. South Alric. 1. Bot. 60: 22-26.
took far apomixis. Apomixis Newslelter 5: Crop xi. 20: 37S-79 Voigt, PW., and H. Bashow. 1972. Apomixis
l1r-19. Snyder, LA., A.R. Hernandez, and H.E. Warmke. and sexual~ in uogrostis rurvulo. Crop
Savidan, Y.H., D. Grimanelli, and O. Leblonc. 1955. The methanism afapomixis in Sri. 12: 843-47
1995. Apomixis expression in maize- Pennisetum riliore. Bot. Gal. 116: 209-21. Voigt, PW., and 8.L Bursan. 1983. Breeding of
Tripsocum hybrid derivatives and the Sarensen, 1. 1958. Sexual chromosome- apomictic uogrostis rurvulo. Proc. 14th Int.
impli<ations regarding ~ control and aberrants in triploid apomictic Taraxaca. Gross/. (ongr. Pp.16Q--63.
potential for manipulation. Apomixis Botonisk TidssJcrih 54: 1-22. - - . 1992. Apomixis in uogrostis.. In J.H.
Ne"lnJelter 8: 3S-37. Stebbins, G.L 1941. Apomixis in the Bgin and lP. Miksche (eds.), Proe.
Savidan, Y.H, LJank, and CB. do Valle. 1989. angiosperms. Bot. Rev. 7: 507-42. Apomixis Workshop, February 11-12,
Breeding Panicum maximum in Brazil. 1. - - . 1950. Variation ond Evolution in 1992, Atlanta, Georgia. USDA, ARS. AR5-
Genetic resources, modes af repraductioo Plants. New York: Columbia Univ. Press. 104. pp. 8-11.
and breeding procedures. Eup!rytico 41: Talioferro, tM., and H. 8oshow. 1966 Young, B.A., R.T. Sherwood, and E.tBashow
107-12. Inheritance and control of obligate 1979. Geared-pistil and thick-sectioning
Savidan, Y.H, and J.Pernes. 1982. apomixis in breeding buffe~rass, te<hniques far detecting apasporous
Diplaid-Ietraploid·dihoplaid cycles and the Pennisetum eiliore. Crop Sei. 6: 473-76. apomixis in grasses. (on. 1. Bot. 57:
evolutioo 01 Ponirum moximum Jacq. VaJIe, tB. do, and t Glienke. 1993. Towards 1668-72.
Evolution 36: 591r-600. defining the inheritance of opomixis in Yudin, B.F. 1994. Towards progeny-tests as
Broehiorio. Apomixis Newslelter 6: 24--25. recognifian tool of opomicts. Apomixis
Ne"lnlelter7: 10-12.
Chapter 6
and they are not necessarily informative when endosperm, dosage requirements would have
it comes to manipulating apomixis genes acted as a barrier against the emergence of
beyond their respective species. Indeed, it apomixis by preventing endosperm
could well be that a single mutation in those formation. Hence, only families in which the
species gave rise to an apomictic genotype. regulation of endosperrn development had
This does not rule out the possibility that somehow been modified would have been
several other genetic factors may be required prone to the emergence of apomixis.
to ensure the expression of apomixis. Such
By the same token, it is conceivable that
factors would not necessarily be detected
different families would have been inclined
through classical genetic analysis, simply
to different types of apomixis. Strong
because of a lack of polymorphism for those
supporting evidence that different species are
characteristics, but the factors would be
compatible with different forms of apomixis
revealed by manipulating apomixis beyond
can be found in the phylogenetic pattern of
the limits of specific species or genera. Those
distribution of the various forms of apomixis
characteristics, necessary but not sufficient,
(Richards 1986;Asker and [erling 1992;Mogie
probably would have accumulated during the
1992; Carman 1997). Most apomictic taxa
evolution of those species prior to their switch
(75%) belong to only three families:
from sexual to apomictic modes of
Asteraceae, Poaceae, and Rosaceae, which
reproduction.
together comprise no more than 10% of
Several observations support this hypothesis. angiosperm species. Diplospory is common
During various attempts to transfer apomixis among the Asteraceae, but less so among the
from wild relatives to cultivated crops, the Rosaceae and the Poaceae, while apospory is
observed transmission of apomixis through common among the Rosaceae and the
generations of backcrossing did not conform Poaceae, but less so among the Asteraceae.
to a simple genetic model. In the case of the Autonomous apomixis appears to be
maize- Tripsacum system, genetic data show restricted mostly to the Asteraceae and is
that the expression of functional apomictic found only infrequently in the Poaceae and
reproduction depends on a complex mode of Rosaceae. Clearly, the occurrence of apomixis
inheritance (Leb lanc, personal comm.; and the distribution of its various forms are
Savidan, Chap. 11). The conditions of not random (Carman 1997).This might reflect
endosperm development in pseudogamous (i) that not all taxa are compatible with the
apomictic grasses is another strong illustration emergence of apomixis, and (ii) that different
of this hypothesis. Angiosperm apomicts taxa are not compatible with the same types
evolved from sexual ancestors that may have of apomixis.
been subject to dosage effects in the
Hence, introducing apomixis into otherwise
endosperrn, as apparently many angiosperms
sexually reproducing crops may depend on
have to a variable degree (see Birchler 1993).
more than the few genes responsible for
This suggests that some adjustments in the
polymorphism in modes of reproduction
mechanisms governing endosperm
wi thin agamospecies. Other factors ma y need
development might have accompanied the
to be considered, such as the endosperrn, that
evolution of apomixis; because the switch from
represent necessary conditions for the success-
sexual to apomictic reproduction simul-
ful expression of the apomictic genes per se.
taneously changes the genomic ratio of the
Applications of Molecular considered two alternatives. The first
Genetics to Apomixis Research alternative is to use existing apomictic species
What Material? that fulfill, as much as possible, the criteria
It could be speculated that the diffusion of described earlier. Bicknell proposed Hieracium
apomixis in crops could be achieved through as a model system and has been developing a
the isolation and manipulation of genes from transposon tagging approach for aposporous
a well-chosen model system. It is worth apomixis in that species (Chap. 8). On the other
considering, then, whether this model could hand, the wealth of genetic information
be defined for apomixis research. But is there available in the grass families, including the
solely one "universal apomixis," despite the remarkable level of genomic synteny found in
amazing diversity of apomictic processes? In the Poaceae (Bennetzen and Freeling 1993;Ahn
other words, should we consider the different and Tanksley 1993; Moore et al. 1995), make
types that have been described in the literature apomictic grasses an attractive model because
as different expressions of the same genetic the various forms of apomixis can be
components, or should different sources of compared.
apomixis be studied as distinct and unrelated The second alternative relies on generating or
processes? According to Sherwood (Chap. 5), transferring the components of apomixis into
a single gene might be responsible for the a well-characterized. easily handled organism,
induction of both diplospory and apospory. such as Arabidopsis or maize.Three approaches
Still, apomixis has occurred in a seemingly for this alternative have been proposed: (i) the
independent fashion in various taxa during transfer of apomixis from a wild species to a
their evolution through different processes, related and genetically well-studied crop
which might also be viewed as evolutionary through sexual hybridizations, (ii) the de novo
convergence. Although answering these generation of apomixis in normally sexual
questions is undoubtedly an important long- organisms by mutagenesis and manipulation
term goal, given our current knowledge, the of gene expression, and (iii) the de novo
choice of a model system for apomixis research generation of apomixis through wide
is more a matter of technical considerations. hybridization after selection of the appropriate
Some of those considerations are proposed by parental reproductive phenotypes (based on
Bicknell (Chap. 8) in this volume, and include Carman's hypothesis [Carman 1997,Chap. 7]).
the ability for both in vivo and in vitro culture, A review and discussion of the first two
a short generation time, easy hybridization, the approaches follow later in this chapter. A
availability of related sexual and apomictic description of the third approach may be
biotypes, good characterization at the genome found in Chapter 7.
level, and ability for transgeny. Another
Most current work in apomixis research
important consideration is that using diploid
essentially focuses on the very first event in
apomicts greatly simplifies genetic analyses.
the apomixis mechanism, i.e., the failure or
Furthermore, access to efficient mutagenesis
absence of meiosis. This is partly a
procedures, including transposon mutagen-
consequence of the prevailing hypothesis that
esis, would provide attractive tools for
apomixis processes in their entirety, or at the
functional analyses.
very least, apomeiosis, might depend on a
While no known taxon fulfills all of the above single-gene regulation. As opinions evolve
criteria, researchers working on the genetics regarding this regulation, more effort will be
and molecular biology of apomixis have
AppiadIoos ef. . . Geotlics io ApooWs ...... 89
directed toward identifying the components DNA markers linked with both apospory and
required for the expression of functional diplospory have been reported for apomictic
apomixis and dissecting their genetic basis. Pennisetum, Brachiaria, Taraxacum, Tripsacum,
Most of the works presented herein deal with and Erigeron species, among others (Ozias-
apomeiosis. The stra tegies described, however, Akins et al. 1998; Pessino 1997; Leblanc et al.
apply to most aspects of apomictic 1995; Grimanelli et al. 1998b; Noyes and
development. Rieseberg 2000). Interestingly, these diverse
reports reach common conclusions about
Molecular Mapping of Apomixis
several aspects of apomixis. Taken together,
The first molecular work on apomixis
they demonstrate that apomeiosis is likely
essentially focused on the development of
controlled by one or several genes located on
molecular maps and the localization of the
a single chromosome segment. Furthermore,
DNA regions that control apomixis in various
reports on Tripsacum (Grimanelli et al. 1998a)
organisms. Part of the interest in developing
Penniseium (Ozias Akins et al. 1998) and
genetic maps lays in the nature of molecular
possibly Erigeron (Noyes and Riesberg 2000)
markers; their Mendelian inheritance is
indicate that this segment might be
independent of either environmental
characterized by a very strong restriction to
conditions, or our ability to actually observe a
recombination. In Tripsacum, where the
given phenotype. Therefore, by studying their
mapping data could be compared between
cosegregation with any trait of interest, one
apomictic and sexual accessions, this
can identify and characterize chromosomal
restriction to recombination appears to be
regions that play a role in the expression of
apomict-specific; while in the sexual forms the
that trait. Once mapped, any trait can
mapped alleles underwent a significant rate
theoretically be studied or followed-
of recombination, complete linkage was
regardless of its expression and with a known
observed in the apomict for the alleles detected
confidence level-by detecting and analyzing
by the same probes. Clearly, recombination is
the segregation of linked molecular markers.
restricted at the tetraploid (apomictic) level as
Chapter 10 is devoted to the genetic opposed to the diploid (sexual) level in both
improvement of apomictic cultivars. Most Tripsacum and maize, as seen in their RFLP
applications of molecular maps in plant maps. In Pennisetum, the segment itself seems
improvement are also relevant to apornicts, to be apomixis specific, as revealed by
and comprehensive reviews on such Southern analysis.
applications are readily available. Note,
Because the specific chromosome segment
however, that molecular markers are
shows a restricted level of recombination, the
particularly valuable for studying characters
classical model of monogenic inheritance for
that are expressed late in plant development,
apomixis probably warrants a careful review,
such as apomixis or other reproductive traits.
because regardless of the number of genes
By using DNA markers, reproductive
involved, they behave as a Single locus in
behavior can be rapidly predicted at the
seedling stage, with confidence levels that segregating populations. This number of genes
might be particularly important within the
depend mainly on the linkage between the
framework of a gene isolation program.
marker and the mapped gene(s). Moreover, as
opposed to cytoembryological tests, molecular
marker analysis is not destructive.
90 o.iel cn-a-. lot to.... aodDMgo Goo,iIn......leOo
Cloning the Apomixis Gene(s) Using because most, if not all, of the candidate
Molecular Genetics Tools species for a map-based cloning project are
A major difficulty encountered by those highly heterozygous tetraploids. for which
interested in cloning "apomixis genes" is little genomic characterization exists.
simply defining what they are. Introducing
Furthermore, when attempting positional
apomixis into crops implies that specific genes
cloning, the first step is to identify a
are transferred or altered and expressed in the
chromosomal region, defined by two or more
target crops. Most likely. not all of the genes
molecular markers, that flanks the gene under
involved in the apomictic process should be
study. The precision of the estimated position
targeted: most, if not all, of them should
of the gene is therefore limited by the smallest
already be present and playing a role in
measurable recombination unit, meaning one
sexually reproducing plants. The issue then is
recombinant in a given mapping population.
which alleles of pertinent genes must be
Hence, the recombination level around the
transmitted or manipulated for the induction
apomixis gene(s) presents another significant
and successful development of apomictic
challenge: positional cloning will prove
embryos and seeds. To date, all efforts to tag
efficient only insofar as recombination can be
apomixis genes, including those presented in
observed near the locus of interest. As
this paper, have focused on the mechanism of
mentioned earlier, recombination near the
nonreduction, mainly because it is an excellent
apomictic alleles is very likely restricted, at
indicator of apomictic development and it is
least in Pennisetum and Tripsacum.
probably the easiest one to score. Nevertheless,
Consequently, the smallest recombination unit
it should be remembered that apomixis is
defined by two markers that encompasses the
probably more complex than the simple
apomixis locus might well be a relatively large
process of nonreduction. The importance of
amount of DNA.
this constraint will likely emerge when
attempts are made to synthesize de /lOVO Transposon tagging of apomixis genes. Some
apomicts in sexual organisms model plants, such as maize, rice, tomato,
Arabidopsis, and Petunia have undergone
"Map-based" cloning in apomictic species.
extensive genome characterization. Specific
Once a gene has been located on a genetic map.
approaches are available for gene tagging
subsequent efforts to specify its position can
these plants that might be considered for
ultimately lead to its isolation (for the first
tagging apomixis gene(s), provided that
successful efforts in plants, see Giraudat et al.
components of apomixis occur in one of these
1992; Martin et al. 1994). The recent
organisms.
development of powerful new approaches for
physically mapping chromosome segments A very promising approach is that of
combined with the ability to clone large DNA transposon tagging. Transposable elements
fragments (Burke et al. 1987; Shizuya et al. are short DNA sequences that have the
1992). and progress in genome sequencing property to transpose to more or less random
techniques have created new and higher locations in the genome (see Walbot 1992, for
standards for positional cloning in plants. It is a review). They were discovered in maize, but
still a laborious and risky task outside of a few have since been identified or introduced in
well-characterized model genomes, but the very diverse organisms. They have been used
number of genes cloned in this manner are in a wide range of genetic studies, and have
rapidly increasing. However, positional been found to be highly effective for gene
cloning for apomixis is not \"ery promising tagging and cloning.
"'lea.....f lIoIt<oD Goeolia it Apoooillis....... 91
Transposon tagging in apomicts presents some respective alleles, might represent prospective
constraints, including access to transposable "candidates" for the apomixis gene(s), i.e., the
elements and the genetic control of the trait. gene(s) that would code for identical
To the best of our knowledge, transposon functions as their apomicitic counterparts.
activity has not been demonstrated in The best, though not the only, candidates are
apomictic species. This might be overcome by the yeast genes responsible for the induction
introducing functional transposable elements of meiosis and the meiotic mutants identified
into apomicts, either through transformation in higher plants.
(as in Hieracium, Bicknell, Chap. 8) or through
Major biochemical pathways involved in the
hybridization with a close relative (as with
regulation of the cell cycle and meiosis appear
maize and Tripsacum, Grimanelli 1997).In both
to be relatively well conserved between
cases, maize transposable elements were
distant organisms such as yeast and higher
successfully introduced into an apomictic
plants, and the advance of whole-genome
background, and transposable activity was
sequencing puts provides complete catalogs
demonstrated.
of putative candidate genes. This progress
In our view, the main issue concerning offers great promise, but it is tempered by the
transposon tagging of apomixis is genetic fact that it is usually difficult to verify
control of the trait. While this approach is whether a yeast gene of known function plays
efficient for phenotypes controlled by single a similar role in plants. One powerful way to
genes, it might yield no, or disappointing, corroborate such gene functions is the so
results ifapomixis is genetically more complex. called "reverse genetics" strategy, using either
But taken further, it would at least provide an insertional mutagenesis or homologous
elegant method to determine whether recombination. When based on transposon or
apomixis is controlled by one or several genes: T-DNA insertions, reverse genetics (or site-
if a Single allele controls the trait, then a Single specific transposon mutagenesis) implies that
mutation should allow complete reversion to transposon tagging is performed to identify
sexuality; if a more complex system is individuals carrying a transposon insertion
involved, then individual mutations should in a gene of known sequence. The expected
lead to abnormal or only partial expression of function of that given gene can then be
the trait. corroborated by confirming that its disruption
leads to the loss or alteration of the expected
Candidate gene approaches. Although
function. Powerful reverse-genetic systems
apomixis is unknown in major crop plants or
are available in various plant species,
other genetically well-characterized
including maize, Arabidopsis, and tomato.
organisms, useful information can be derived
from detailed analyses of the reproduction A specific candidate gene strategy based on
processes of select sexual organisms. For comparative mapping can also be undertaken
example, genes involved in the control of ovule within the grass family. The identification of
development, the initiation of meiosis, orthologous genes between species (i.e.,genes
embryogenesis, and endosperm development that diverged from a common gene at the time
have been described in various organisms, and that the species harboring them diverged)
a close look at these genes might provide could be used to understand the relationships
useful information about the regulation of between the genes responsible for various
apomixis. Such genes, but not necessarily their components of apomixis in apomictic plants,
92 D.itl ........ Jot. . . . . . Diop Gooziln.....
Gicaudot, J., B.M. Houge,l Yalon, J.Smolle, F. l!b!anc, D., D. Grimaneli, D. Gonmlez de L20n,
References Parry, and H.M. Goodman. 1997. Isolation and Y. Savidon. 1995. Deteclian afthe
Ahn, S.N., and S.D. Tanks/ey. 1993. afthe Arobidopsis AB, 3 gene by posilianal apani<lic made afreproduclian in maize-
(omparative linkage maps afthe rice and
maize genames. Proc NaIl Atod sa (USA)
cloning. Planta« 1751--'1. Tripsacum hybrids using maize RFlP
Grimaneli, D. 1997. I:opomixie dons le genre markers. Theor. AppI. Genet. 90: 119S-
90:79~.
Tripsocum: cootr6Ie genetique et 1703.
Asker, S., and LJeriing. 1997. Apomixis in controinles ossodt!es iJ son IrDIISFel1 rhez l!b!anc, 0., D. Grimaneli, N.I. Faridi, J.
Plants. Baco Ratan, Florida: (RC Press. Ies plonles cuhivf!es. Ph.D. dissertation, Berthaud, and Y. Savidan. 1996. Can
Bennetzen, J.L, and M. Freeling. 1993. Grasses Inslilut Nalianal Agronomique Paris· diplaid cereals reproduce through
as asingle genetic system: genome Grignan. apanixis? Evidence fram maize- Tripsocum
compasilian, calinearily and campatibilily. Grimaneli, D., M. llemandez, E. Peralli, and Y. hybrids. JoumaJ of HereGty B7: 10i-15.
Trends in Genetics 9:759-61. Savidon. 1997. Dosage effects in the Lua, M., P. Bilodeau, E.S. Dennis, Ill. Peocadt,
BidcneU, R.A., N.K. Barslk, and A.M. KaIhmw. endosperm afdiplolparaus apomidic andA. Chaudhury. 7000. Expression and
7000. Monogenic inheritllKe afapanixis Tripsocum (Paauae). 5exlJll1 Plant parent·af-origin effects far F1S7, MEA, and
in two Hieraoum species with l&stincl Reproduclioo1D: 779·7B7 RE inthe endosperm and embryo af
developmenlal mechanisms. Heredity. B4 Grimanelli D., D. l!bIarK, E. Peralli, D. develaping Arabidopsis seeds. Proc NaIl
:778-37. Ganzcilez de Lecin, and Y. Savidan..19980. Acad sa (USA) 97(19):10637-47.
Chaudhury, A.M., L Ming, L Miler, S. (raig, E.S. Nan-Medeb transnission af apomixis in Ma, H. 1999. Seed Develapment, withar
Dennis, and WJ. Peocadt. 1997. maize· Tripsocum hybrids caused by a without sex? u.renl Biology 9: 636-39.
Fertilization-independent seed transmiisian ratio distortion, Heredity BD: Martin, G.B., S.H. Brammanschenkel, l.
development in Arabidopsis /hoiona. hoc 4(1...44. Chunwongse, A. Frary, M.W. Ganal, L
NaIl Atod sa (USA) 94: 47~7B. - - . 199Bb. Diplasporaus apanixis in Spivey, T. Wu, E.D. Earte, and S.D. Tanksley.
Birchler, JA 1993. Dosage analysis afmaize tetraploid Tripsocum: one gene ar several 1994. Map-bosed daring afapratein
endasperm development. Annual Review of genes? HemJityBO: 33-39. kinase gene canferring disease resistance
Genetics 77: IBI-704. Grassniklous U, J.P. Y"Calzado, MA in tamala. Science 67:1437-36.
Burke, DJ, G.F. Oarke, and M.Y. DIson. 19B7. Haeppner, and W.B. Gogliana. 1998. Magie, M. 1988 Amodel far the e'IOOtian and
Oaning aflarge segments afhexagenaus Maternal cantrol afembIyogenesis by cantrol afgenerative apanWs. Bioi. 1.
DNA inta yeast by mean of artificial MmEA, a polycarnb group gene in unn. Soc. 35: 177-53.
chramosome vectors. Science 736: 806- Ara!Jidopsis. Science. 780(5367): 446-50. - - . 1997. The erolution ofasexool
11. Grassnik!aus, U., A. Kahooow, and M. van reproduction inplon/:l. l.andon: Chapman
Calzoda, J.P·Y., IF. (rane, and D.M. Stelly. Loakeren Ca~. 1999 Abright MUle and HaD.
1996. Apomixis: an asexual revolution. lar apomixis. Trends PIont sa. 3: 415-16. Moare, G., K.M. Devas, Z. Wang,. and M.D.
Science 774: 1377-73. Jefferwl, RA, and R. BickneI. 1996. The Gale. 1995. Grasses: line-~ and form a
Carman, J.G. 1997. Asynchronous expression af potential impocI af apanixis: a maIecuIar cirde. Cflrenl Biology 5: 738--47.
duptKated genes in angiasperms may genetic appraadt. In B.W.S Sabral led.!, Neuffer, G.M., E.H. Cae, and S.R. WessIer. 1997.
cause apanixis, bispory, letraspory ond ThelllfKJd ofPIonl MoIeoJor Genetics. Mutants afmaize. Cold Spring Horbar, New
poIyembryany. Bioi. 1. linneon Soc. 61, Baslan: 8irkhaiiser. Jersey (USA): (aid Spring Harbor
51·94 Kajima, A., T. Kazana, Y. Nagola, and K. Hinata. laboratory Press.
Chaudhury, A.M., L Ming, L Miller, S. (roig, E.S. 1994. Non-parthenogenetic ~ants detected Nagler, GA 1982. How taobtain diploid
Deoois, and WJ. Peocadt. 1997. in Chinese Chive, afacuhative apomict. apanictic Ranunculus ouricamus plants naI
Fert~izatian-independant seed Breeding Science 44:143-49. found in the wild slate. Bot.Helvelico 97:
development in Arabidopsis tha~ana. Proc. 13-27.
Notl. Acod. Sci. IUSAI 94:4773-7B.
94 o.iel Gno-e.., lot TolNooe, aod Diego G"",ciIe,"'lHI
- - . 1984a. Gamelaphytic apomixis. In Ozias-AXioI, P. 1998. Tighl clustering and Shizuya, H., B. Birren, U.· J. Kim, Y. Mancina, 1.
8.M. lOOri (ed], Embryology of hemizygasity af apomixis-linked malecular Slepok, Y. Tachiiri, and M. Siman. 1992.
Angiosperms. Berlin: Springer-Yerlog. pp. markers in Pennisetum squolTXJlotum Cloning and liable mainlenonce of 300·
475-SlB. implies genefic conlrol of apaspary by a kilobase-poir frogments 01 human DNA in
- - . 1984b. Genelia of apospory in divergent locus that may have no allelic Esd1erichia cali using a f-faclar-bosed
apomictic Ronunculus ouricomus Y farm in sexual genotypes. Proc. Norf. kod. veclar. Pro. Natl. Acod. Sri. (USA) 89:
Conclusions. Botonico Helvetic094: 411- Sei.IUSAJ 95: 5127-32. 8794--97.
22. P~no, S.c. 1997.ldentifica1ion of amaize von Dijk, PJ. 1999. Crosses between sexual and
Noirol, M. 19lJ3. Allelic ratios and slermty in the ~nkoge graup releted 10 apomixis in apomictic dondelions IToroxocum!. 11. The
agamic com~ex af the Maximeae 8racharia. Theor. Appl. Genet. 94: 439- breakdown af apomixis. Heredity 83:
(Panicaideae): evolutionory role of the 44. 715-21.
re~dualsexualily. J. Evol. BioI. 6:95-101. RKhords, AJ. 1986. Plont breeding s~tems. V"lelle-Calzado H, J.Thomas, C. Spillane, A.
Nares, R.D., and R.H. Rieseberg. 2000. Two London: Gearge Alien and Unwin. CoIu«io, MA Haeppner, and U.
independenllocj control agamaspermy Sovidan, Y. 1982. Nature etheredife de Grossniklaus. 1999. Mainlenonce of
(apamix~) in the Iri~aid Aawering ~anl r opolTixie chez Ponicum maxilTXJm Jocq. genomic imprinting a1the Arobidopsis
Erigeron onnuus. Genetics 155: 379-90. PftD. dissertotion Universite af Paris XI. medea locus requires zygatic DDM1
Ohad, N., LMargassian, H. Hsu, C. Williams,P. - - . 2000. Apomixis: Genelics and activity. Genes Oev. 13(22): 2971 ~2.
Repeni, and R.L FISher. 1996. Amutation 8reeding. Plont Breeding Reviews 18: 13- Walbol, Y. 1992. Slralegies far mutagenesis and
ihot allows endasperm developmenl 86. gerle cloning using Iransposan lagging and
development without lert~iza1ion. Prac. Sheridan, W.F., AA Nadezhda, 1.1. Shanyov, lB. T-DNA mutogenes~. Ann Rev. Plont P!rysiol.
Natl. Acad. Sei. (USA) 93: 5319-24. Batygino, and I.N. GoIunovskaya. 1996 Plont Mol. BioI. 43: 49~2.
The mac gene: cantralling the commitment
tathe meiotic path'MJY in maize. Genetia
142: 1009-20.
Chapter 7
The Gene Effect:
Genome Collisions and Apomixis
JOHN G. CARMAN
mechanisms (Figure 7.1). For example, are found in Allium. The recognition of these
Antennaria-type diplospory is identical to developmental similarities and phylogenetic
tetraspory (sexual), except that the nuclear associations led Carman (1997) to analyze
divisions leading to a tetranucleate embryo sac taxonomic data for all known species that
are meiotic in tetraspory but mitotic in express these anomalies. It was found that
Antennaria-type diplospory. Both Antennaria- apomictic, polysporic, and polyembryonic
type diplospory and tetraspory occur in species are polyphyletic and tend to be
Antenllaria, Erigeron, Limonium, and Rudbeckia phylogenetically related. Many highly
(Carman 1997).In both anomalies, MMC lack significant associations were discovered.
callose (Peel et al. 1997a). Ixeris-type
The variation represented in Figure 7.1 is
diplospory is even more similar to tetraspory
somewhat continuous. For example,
because, as in tetraspory, a meiotic prophase
Antennaria-type diplospory is similar to both
occurs. However, in the diplosporous
tetraspory and apospory. In some families,
mechanism, a first division restitution ensues.
both species and genera span this continuum.
Ixeris-type diplospory is identical to bispory
For example, apomixis and polyspory occur
(sexual) except that meiosis I failsin the former
together in 13 of 127 apomixis-containing
(Figure 7.1). Both Ixeris-type diplospory and
genera (Allium, Antennaria, BlIrnumnia, Cordia,
bispory (and apospory and tetraspory) occur
Cynoglossum, Erigeron, Eurvoiopsi«, Leontodon,
in Erigeron. Allium odorum-type diplospory
Limonium, Rubus, Rudbeckia, Sambllclls, and
and bispory differ only in that a chromosomal
Tridax) and 18 of 33 apomixis-containing
endoreduplication occurs in the former. Both
Abundant
families. Differences between observed and evidence that certain genera are particularly
expected values for these associations, vulnerable to evolutionary processes that
assuming independent distribution, were asynehronize ovule development. Several
highly significant (Carman 1997). Such possible reasons for this vulnerability have
developmental similarities and phylogenetic surfaced from our examinations of the
associations suggest that the gene effect genomic makeup of reproductively-
mechanisms responsible for apomixis are in anomalous species.
some way mechanistically and evolutionarily
related to those responsible for polyspory and Genomes of Reproductively-
polyembryony. Hence, gene effect hypotheses Anomalous Species
that explain apomixis should also explain the High chromosome base numbers (x ~ 10)
related anomalies. suggest paleopolyploidy, which means
polyploidy followed by diploidization with or
Two other female anomalies appear
without ascending or descending aneuploidy.
mechanistically and phylogenetically related
Multiple base numbers per genus reflect
to apomixis and polyspory. Both repeat a
diploidization following ascending or
segment of female development in an
descending aneuploidy and is further
asynchronous manner. The first, preleptotene
evidence of paleopolyploidy (Goldblatt 1980;
chromosome condensation, inserts an
Lewis 1980). Carman (1997) found
addi tional mi totic prophase between
chromosome base numbers for 80% of all
premeiotic Sand leptonema. The chromo-
genera identified as containing apomictic,
somes condense to a mitotic metaphase state,
polysporic, or polyembryonic species.
decondense, and resume meiotic prophase
Statistical analyses indicated that polysporic
(Bennett and Stem 1975). This phenomenon
and polyembryonic species are generally
is similar to Allium odorum-type diplospory
paleopolyploid (x = 15.7 and 13.2,
(see Crane, Chap. 3) in that a segment of the
respectively), while many apomicts, which are
cellcycle (prophase, not S)is duplicated before
generally polyploid (Asker and Jerling 1992),
meiosis. Preleptotene chromosome
contain primary genomes (x = 9.6).
condensations occur in certain species of
Furthermore, genera with polysporic but not
Liliitm, Tradescaniia, Trillium, Fritillaria,
apomictic species have more x values per
Nicoiiana, Vieia, and Psilotum, with the first five
genus (2.7 ± 0.4SE) than genera with apomictic
also containing polysporic species.
but not polysporic species (1.7 ± 0.1). This
In the second anomaly, observed in Rosaceae, means apomicts tend to have balanced sets of
an apparently normal MMC forms, duplicate genes (primary genomes) and
degenerates, and is subsequently replaced by polysporic and polyembryonic species tend to
one or more secondary and fully functional have imbalanced sets of genes (paleopolyploid
MMCs. This anomaly occurs in Alehemilla, genomes, i.e., partially duplicated or
Cotoneasier, Sorbus, Rubus, Aphanes (Davis triplicated due to aneuploid series formation
1966) and Waldsteinia (Czapik 1985), all of followed by diploidization). Thus, a distinct
which also contain apomictic or polysporic divergence in genome composition occurs
species (Czapik 1985; Carman 1997). The fact between apomixis on the one hand and
that preleptotene chromosomal condensations tetraspory and polyembryony on the other,
and repetitive MMC formations occur in even though these anomalies are
genera that in almost all cases contain species phylogenetically and mechanistically related.
expressing apomixis or polyspory is strong Hence, gene effect hypotheses attempting to
98 JeIo.G.c....
explain the existence of apomixis must also Several questions relevant to the evolution of
address these highly significant peculiarities female developmental anomalies can be
in genome com posi tion, ph ylogenetic form ula ted from this informa tion. For
related ness, and developmental (or example, how is the synchrony of female
mechanistic) affinities. development affected when some of the
duplicated genes responsible for megasporo-
Other genome-related factors may abnormally
genesis, embryo-sac development, and
affect female development in polyploids or
embryony from one of two genomes are
paleopolyploids. For example, meiotic rates are
silenced or lost during diploidization
linearly correlated with DNA content, but the
(paleopolyploid formation)? What happens
regression line is much steeper in polyploids.
when there are duplicate doses of genes for
And, meiosis in tetraploids usually requires the
certain stages of meiosis or embryo sac
same period of time as in related diploids
development and not other stages, as is
containing half the DNA. This occurs because
anticipated in highly aneuploid paleo-
genes for meiosis in polyploids are duplicated
polyploid polysporic species? Could such
(Bennett 1977). That the meiotic rate to DNA
imbalances cause some of the anomalies of
content regression slopes in paleopolyploids
embryo sac development observed in
reflect either a diploid or a polyploid condition
polysporic species, such as a precocious
may be critical to the evolution of apomixis and
gametophytization of the MMC or the
related anomalies, Examples include Scilla
formation of 4 to 32 nucleate embryo sacs?
nonscriptus, 21/ = 2y = 16, which belongs to an
aneuploid series with x = 6 to 9, 15, and 17 as Total quantities of DNA also influence the
stabilized base numbers, and Comiallaria types of life cycles angiosperms assume. For
niaialis, 211 = 2" = 38, with a sole base number example, species of Fritillaria (many of which
of x = 19. Both are paleopolyploid diploids with are polysporic paleopolyploids) have large
large quantities of DNA, but their meioses amounts of nuclear DNA and their meioses
occur in only 50% of the time predicted for non- may require 3-4 weeks to complete. In
paJeopolyploid dipJoids with similar amounts contrast, annuals have little DNA and very
of DNA, i.e., they behave as polyploids. In short meioses (Bennett 1977), and apomixis
contrast, Ornithogalum oirens (2/1 = 6), which is and related anomalies are rare among them
at the bottom of a descending aneuploid series (Asker and Jerling 1992). Hence, a minimum
in which x = 3 to 5 and 7, Allium cepa (2/1 = 16), threshold in duration of meiosis may be a
which is at the middle of an aneuploid series prerequisite for the evolution of certain
with x = 7 to 9, and Fritillaria meleagris (211 = reproductive anomalies.
2-1), which is at the top of an ascending
Reproductive anomalies in angiosperms might
aneuploid series with x = 7, 9, and 12, are
also be influenced by differences in meiotic
paleopolyploid diploids with slow meioses
durations between genders. In cereals, female
that is indicative of diploids with considerable
and male meioses are generally synchronous
DNA. Duplicate genes for meiosis in these
and similar in duration. However, in species
species have either been lost through
in which female meiosis occurs later than male
aneuploidy or genetically silenced. Hexaploid
meiosis, differences in duration may be as
nulli 5B tetra 50 wheat is another example.
great as 50 % (Bennett 1977). Such differences
Meiotic rates in this line reflect a tetraploid, not
might encourage anomalous development in
a hexaploid, probably because of irnbalanced
one gender but not the other.
sets of meiotic genes (Bennett 1977).
l\e Go.. Elf.d' Go_ UIisioo........iI 99
Finally, ecotype divergences resulting from photoperiod responses, i.e., only 3.8% of
divergent selection pressures associated with genera are known to express reproductive
high versus low latitudes (flowering responses anomalies. Thus, if rep rod uctive anomalies
to specific photoperiods), high versus low occur independently of distinct adaptations to
elevations, highly shaded versus full sun photoperiod, then only 3.8% (not 33%) of the
conditions, or moist versus dry conditions genera identified as containing species with
may contribute to the evolution of apomixis distinct photoperiod adaptations should have
and related anomalies. Many plants have been also expressed reproductive anomalies. This
categorized according to their response to 33% breaks down as follows: 12% contain
photoperiod. For example, long-day plants are gametophytic apomicts (compared with 1% of
adapted to higher latitudes and flower in the all angiospermous genera, i.e., a 12-fold
spring and early summer when days are long. increase); 13% contain polysporic species
Short-day plants are often found in lower (compared with 1.6% of all angiospermous
latitudes (tropics) and often flower during the genera, i.e., an eightfold increase); and 7.5 %
tropical "winter" when days are short. Dual- contain polyembryonic species (compared
day-length plants require either short or long with 1.7% of all angiospermous genera, i.e., a
days to induce reproductive bud formation, 4.4-fold increase). These associations between
followed by long or short days, respectively, genera containing species that express
to cause the formed buds to mature into reproductive anomalies and genera containing
flowers. Intermediate-day plants will not species that express distinct adaptations to
flower if days are too long or too short. Day- photoperiod (different latitudes, etc.) should
neutral plants show little adaptation to day be studied in more detail, as they suggest that
length and flower induction occurs under a photoperiod responses ei ther contribu te to the
broad range of day lengths. In addition, evolution of reproductive anomalies or are at
several other specialized categories exist. least correlated with unknown causal factor(s)
(see Carman 2000).
Salisbury and Ross (1992) selected 78 species
from among approximately 300 species of These data indicate that apomictic polyploids
plants studied for flowering responses to could contain interracial or interspecific
different photoperiods. These 78 species were genomes polygenically "coadapted" (Wallace
chosen independently of any reproductive 1991) to divergent environmental conditions.
anomalies that might be expressed either If true, the sexual progenitors of apomicts
within themselves or within other species in could have expressed cytologically-detectable
the genera they represent. In contrast, they temporal divergences in gross ovule
were chosen to distinctly represent specific development relative to the time at which
photoperiod response categories. Sixty-seven female meiosis occurs. Such differences have
different genera are represented by these 78 now been observed among the putative sexual
species. It is interesting to note that 33% of diploid progenitors of apomictic Antennaria
these genera (22 of 67) contain species with rosea and Tripsacum dactyloides (Carman and
female reproductive anomalies (gametophytic Kowallis, unpublished), and these divergences
apomixis, polyspory, or polyembryony; probably represent coadaptions to different
compare Salisbury and Ross 1992, Table 23-1, photoperiods (latitudes or elevation), shading
with the appendix in Carman 1997). This is a regimes (full sun or highly shaded
ninefold increase in the number of genera environments), elevations (d u ra tion of
expected if reproductive anomalies occur growing season), or other environmental
independently of adaptations to distinct
100 JoMG.e-
Table 1.1 Phylogenetic, genomic, and developmental peculiarities that hypotheses for thegenetic regulation
ofapomixis and related reproductive anomatles must explain
PIIy10genetic
Apomixis, polyspory, and polyembryony are rare yet po~phyletic
Apomixis, polyspory, and polyembryony tend to occur in the some species/genera/families
MMC degeneration/replacement occurs in apomictic or po~sporic genera
Preleptotene chromosomal condensations lend to occur in po~sporic genera The majority of apomicts evolved during the post
three million years, i.e., during the lost 2% of the evolutionary existence of angiosperms (see Carman, 2000).
GenOlllic/developmental
Chromosome base numbers are low in apomicts and high in polysporic/po~embryonic species
Apomicts tend to be genome-balanced, polysporic/poIyembryonic species tend not to be
Apomixis, polyspory, and po~embryony are general~ absent in annuak (Iow amounts of DNA/rapid meioses)
Paleapo~ploids may behave as di~oids or polyploids with respect to meiotic duration
Polygonum-type reproduction is facu~ative in apomictic, polysporic, and po~embryonic species
Tendencies to apomixis are observed in some wide hybrids including Raphanabrossica (consistent aposporous embryo sac
formation)
Male and female meioses in some species stan atdifferent times and have different durations
Apomixis, polyspory, and po~embrony tend to occur in genera capable of strong adaptations to photoperiod
Apomixis, polyspory, and polyembryony ore characterized by osynchronously-expressed substitutions, replacements, or
duplications of discrete reproductive phases Most apomicts display relaxed endosperm balance number requirements (see
Grimanelli etel, Chap 6; Grossniklaus, (hap 12)
Since these first reports, the absence of callose apomixis. The precocious induction hypothesis
around diplosporous MM Cs has been is attractive because it explains precocious
documented in Poa nemoralis (Naurnova et al. initiation of embryo-sac formation in apospory,
1993), various Tripsacum species (Leblanc et diplospory, tetraspory, and bispory. However,
al. 1995; Peel et al. 1997a), Eragrostis cunntla the single-mutation-based induction of Esi fails
(Peel 1997a), and Antennaria rosea (Carman, to explain other features of female reproductive
unpublished). Although usually present, anomalies not temporally or spatially
callose envelopment of MMCs is often associated with embryo-sac induction
abnormal in aposporous apomicts (Naumova (Koltunow 1993).
et al. 1993;Naumova and Willemse 1995;Peel
In many reproductive anomalies, two or more
1997a), and these abnormalities may be related
developmental processes occur simultaneously
to the time at which aposporous embryo sacs
and asynchronously (Figure 7.1). During
are initiated (Peel et aI1997a).
apomixis, meiosis and embryo-sac formation
The callose hypothesis states that a genetic often occur simultaneously, and partheno-
lesion(s) frequently prevents or reduces genesis is often initiated before fertilization of
callose deposi tion in the walls of MMCs. This the central cell. Preleptotene chromosomal
causes apomixis by allowing developmental replica tions and condensations, MMC
Signals to move in a less restricted manner abortions and subsequent replacements, and
through the ovule, thus confusing elimination or duplication of nuclear divisions
development. If this underlying concept is resulting in from 4 to 32-nucleate tetrasporic
correct, i.e., if a role of callose is to contain embryo sacs would require additional
developmental signals, we would expect to mutations. The mutation-based nature of this
see this mechanism associated with other hypothesis also fails to explain why
reproductive anomalies. However, MMC reproductively-anomalous taxa are phyloge-
callose is generally not lacking in apospory netically related and why apomicts are
or bispory, and no apparent connection exists generally polyploid with balanced genomes
between absence of MMC callose and and low chromosome base numbers, while
parthenogenesis or the proliferation of polysporic and polyembryonic species are
embryo-sac nuclei, in various forms of typically paleopolyploid with unbalanced
tetraspory (Johri et al. 1992). The callose genomes, high chromosome base numbers, and
hypothesis also fails to address (i) genome multiple base numbers per genus (see Table 7.1
differences between apomicts and polysporic for other phylogenetic, genomic, and
or polyembryonic species, and (ii) the developmental correlations and phenomena
occurrence of "tendencies for apomixis" that not readily explained by Simple mutation).
occur in many wide hybrids (Asker and Thus, while the developmental displacement
Jerling 1992;Carman 1997). component of this hypothesis is intriguing, its
mutation-based explanation is questionable.
The Precocious Induction Hypothesis
That apospory is caused by the precocious and The Hybridization-Derived Floral
ectopic expression of a normal master gene Asynchrony Theory
for embryo sac formation (embryo sac The hybridization-derived floral asynchrony
induction gene, ESI) was proposed by Peacock (HFA) theory expands on the developmental
(1993), who suggested that the precocious displacement component of the precocious
expression of Esi may be caused by a Single induction hypothesis but suggests a non-
mutation. If this is correct, inducing mutations mutation origin. The foundation of this theory
in maize, rice, or Arabidopsis could produce
102 J". G.Cannao
Figure 7.2Model of how asynchronously-expressed dup6cate genes cause diplospory and apospory in
polyploids containing two genomes divergent in the temporal expression offemale developmental schedules
(floral induction, megaspore formation. gametophyte development, and embryony). Bolded developmental
phases are skipped as described below.
Developmentally-uitical stages·
Genome 2 3 4
Genome I Archespore Meiosis Embryo sac Double fertilization/
(unmodified) early embryolly
Genome I Embryo soc Double fertibation/ Fertiliza~on of
(modified) eor~ embryony central cell on~
Genome 11 Meiosis Embryo soc Double fertilization/ Fertiliza~on of
eor~ embryony central cell on~
Ovule development ~ initiated by developmentollignols from genome I,which po~geni(olly encode odoptolions for lote meioss relotive togross ovule development,
ond genome 11, which p~genkol~ encode odoptotians for eor~ meiosis 'elotive to gross ovule development
I. Atlhe beginning ofstoge 1,genome 11 produces signok for meiosis, whilh foil bemuse the MMC ~ not reody for meiosis, i.e., itdevelops oton intermediote role
dictoted by ,he intermediate phenotype
2. AI the beginning of stoge 2,embryo 5llC development signok from genome 11 may synchronize genome I with genome 11 in0 monner sim~or tothat observed in
osynchronous yeast heterokoryons lreviewed herein). 11 meio~s ~ succe'isful~ preempled, one ofseverol forms of diplospory IFig I)OlfUIl, i.e., on embryo soc forms
precOlious~ from the megasporocyte {Antennario-type diplosporyl or young femole meiOcy1e (Toroxocum or lxer~-types 01 diplospory). If meiosis ~ unsuccessful~
preempted, either locukorive sexuolity oroposflOry [Fig 11 Olcurs. In the loner «se, one ormore embryo sers form from odjacent nucellor celk.Th~ OlCUIl Pfimari~
inspecies contoining mukiple or ill-defined orchegonial celk. In both opospory ond diplospory, 0 geneticol~ unreduced embryo 5llC develops. Development of the
nongometophytic t~sues of the ovule ond ovory continues 10Olcur 010 normol role. In conlrost, embryo soc development continues toO(cur precOlious/y befouse of
Ihe precO(ious~-expressed embryo 5llf developmenl genes ofgenomes Iond 11.
III 1 P,ecOlious signok from the two synchrOnized genomes induce egg formotion ond porthenogenes~, bolh 01 which occur pre(Olious~ relotive 10the development 01
nongometophytic ovule ond ovory '~sues.
4. Pollinotion ocrurs ocrording 10the intermediate phenotype schedule, but tbe egg is no longer receptive ond in mony reses hos olreody divided. The (ent,ol cell, if
not outonomous, is fertilized, ond the endospe,m ond porlhenogenetil embryo develop
104 JoMG.e-
effects), which could be caused by differences The HFA theory also predicts ambiguous
in genetic background, or (ii) environmental outcomes regarding the sexuality of progeny
factors that reduce the degree of asynchrony when an apomict is crossed with a sexual or
by accelerating or decelerating gene with another apomict, regardless of the
expression from one genome relative to that closeness or wideness of the cross. The mode
of another (photoperiod or temperature of reproduction expressed in the progeny will
responses, e.g., as occurs in Dicanthium, depend on how the added or removed
Themeda), thus allowing sexual development genome(s) affect asynchrony, and this cannot
to occur facultatively (Carman 2000). be predicted without some a priori knowledge
of the female developmental schedules
According to the HFA theory, polyspory and
encoded by the involved genomes (Carman
polyembryony result from the competitive
1997). That these many inconsistencies in the
expression of grossly imbalanced genomes
apomixis literature are explained by the HFA
(incompletely duplicated sets of reproductive
theory is strong evidence for its validity.
genes) in which some checkpoint systems are
missing. In contrast, competitive expression
Testing the Gene Effect
among genomes is terminated by checkpoint
genes in apomicts, which generally contain
Hypotheses
If apomixis is the result of one or a few
balanced sets of reproductive genes (Carman
mu rations, similar artificially ind uced
1997), thus allowing a somewhat smooth
transition to apomixis (Figure 7.2). mutations might produce apomicts from
sexual species. Research programs currently
Apomixis is much more prevalent among exploring this possibility are reviewed by
outcrossing species than inbreeding species Bicknell (Chap. 8), Grossniklaus (Chap. 12),
(Asker and }erling 1992). This is consistent and Praekelt and Scott (Chap. 13). Likewise,
with the HFA theory because outcrossing simple inheritance should permit transfer of
species are much more prone to form apomixis gene(s) to sexual species. To date,
interecotypic or interspecific polyploids when introgression projects have failed to confer
secondary contacts occur, e.g., during the apomixis upon sexual species by adding
numerous major climatic shifts associated with anything less than at least one complete alien
the Pleistocene glaciations (Frakes et al. 1992; chromosome (Savidan, Chap 11). Kindiger et
Carman 2000). Likewise, more apomicts are al. (1996)reported a condition that might lead
allopolyploid than autopolyploid because to an exception. They isolated, from their
polyploidization by Bm hybrid formation is maize- TripsaclIm backcross program, a line (30
expected to occur more frequently in nature Mz + 9 Tr chromosomes) that appears to
in interspecific hybrids than interracial contain a maize Tripsacum translocation
hybrids. Similarly, the chances of Bm hybrid possessing gene(s) for apomixis. However,
formation occurring in interecotypic or Blakey et al. (1997, reviewed below)
interspecific Fj hybrids that are sterile and determined that the genes required for
annual are low compared to their formation apomixis occur in five distinct Tripsacum
in sterile perennials, which may flower linkage groups that are syntenic to regions on
annually for many years. This factor limits the three maize chromosomes. These data cast
chances of annuals becoming apomictic and doubt on Kindiger's simple inheritance model.
explains their low frequency in nature.
If apomixis is caused by hybridization, the A plus a probe unique to early embryo-sac
parental phenotypes (divergent floral development from genome B detect their
induction stimuli and meiotic start times in respective genome-specific mRNAs in ovules
ovules, etc.) may be identified by a fixed and sectioned during meiotic prophase,
combination of phenological and cytological and (ii) the probe for meiotic prophase from
studies. Because many developmental genome B does not detect its respective
pathways occur asynchronously in apomicts genome-specific mRNA product (or vice
(Figure 7.1),it may be possible to use molecular versa). This would confirm that one genome
probes to determine if the asynchronous codes for meiosis (but not embryo-sac
signals originate from the same genomes or development) at the same time the other
different genomes. For example, do both genome is coding for embryo-sac develop-
genomes in a tetraploid apomict produce both ment. This test would not be valid if the mRNA
meiotic and embryo-sac development signals synthesis identified by the respective probes
at the same time, or does one genome prod uce is produced in response to transacting
meiotic signals (and not embryo-sac regulatory genes.
development signals) at the same time the
Implications of the HFA Theory
other genome is producing embryo-sac
If the HFA theory is correct, much of the
development signals (see Figure 7.2)?
apomixis literature will need to be
Many criteria should be considered in testing reinterpreted. This includes interpretations of
for such asynchrony. First, it would be how apomixis evolved, the role of apomixis
desirable for the apomict to (i) be allotetraploid in evolution, the genetic control of apomixis,
with known sexual diploid progenitors, (ii) and how apomixis might be transferred to (or
contain well-mapped genomes, (iii) be induced to occur in) sexual species.
amenable to transposon tagging, (iv) be easily Evolution of Apomixis and
grown with a short vegetative phase, and (v) Related Anomalies
have ovules readily accessible to cytological According to HFA theory, many secondary
analyses. At present, no apomict meets all of contacts must have occurred between ecotypes
these criteria. Second, molecular probes that (or closely related species) that had been
recognize mRNAs specific to different isolated from each other for many years along
developmental stages would need to be latitudinal or other ecological gradients. Major
produced, and the genes from which they are climatic shifts could account for such
developed would need to contain adequate secondary contacts (Carman 2000). The
intergenomic sequence divergence such that distribution patterns of most apomicts indicate
probes unique to each genome could be a Pleistocene origin (Stebbins 1971;Asker and
produced. Such probes may currently be under Jerling 1992),i.e., the geographic distributions
development. and centers of diversity of many apomicts are
Portions of meiotic prophase and early centered near the margins of the Pleistocene
embryo-sac development occur simulta- glaciations, but their ranges often encompass
neously in most aposporous apomicts and all the ecological ranges of the putative sexual
Taraxacum- and Ixeris-type diplosporous progenitors, which lie north and south of the
apomicts (Carman 1997;Peel et al. 1997a).The glacial margins. Eight major glaciations, which
HFA theory would be confirmed if the covered as much as 20% of the earth's surface,
following two conditions are observed: (i) a occurred during the Pleistocene. These were
probe unique to meiotic prophase of genome separated by warm interglacial periods lasting
106 JolI.G.e..-
for several thousand to a hundred thousand Once asynchrony is initiated, further shifts in
years each. Likewise, most of the major glacial facultativeness might occur in response to
events consisted of glacial advances growing conditions. G. Ledyard Stebbins
interrupted by minor interglacial periods that (personal communication, 1997) suggests that
lasted for a few thousand years (Frakes et al. conditions favoring rapid growth (high
1992). Hence, during the Pleistocene, the high temperatures, moisture, and light) might
latitudes of both hemispheres were repeatedly enforce competition between asynchronous
deglaciated and revegetated by cosmopolitan genomes causing an increased frequency of
taxa capable of adapting to cool climates, long apomixis. This prediction was observed in
days, and short growing seasons. clones of F} hybrids obtained between wheat
and diplosporous Elvmus reciisetus. Clones
A long-day flowering response and a
grown under favorable conditions (high
precocious meiosis and embryo-sac
temperatures and light intensities) produced
development in young ovules of sexual
more aporneiotic-like MMCs (Peel et al. 1997b).
Antennaria, and probably many other taxa, are
Additional research should be conducted to
adaptations to short summers in high latitudes
elucidate such effects on facultativeness (Asker
or altitudes (Carman, unpublished). Glacial
and ]erling 1992). Such research could provide
advances, which followed the numerous
clues concerning the nature of the divergent
interglacial periods, cooled the mid latitudes,
parental phenotypes that may have produced
permitting higher latitude flora to invade
apomicts upon hybridization and poly-
mid latitude flora. This provided opportunities
ploidization during the Pleistocene.
for ecotypes with a putatively-precocious
female development (from higher latitudes or The HFA theory also suggests that the role of
elevations, i.e., temperate to arctic climates) to apomixis in evolution is more prominent than
form polyploids with ecotypes (or related previously supposed. What happens if some
species) with a putatively-delayed female of the many checkpoint genes permitting a
development (from lower latitudes, i.e,. tropic reasonably smooth replacement of
to full-sun temperate climates). Such developmental segments (resulting in
polyploids may have given rise to modern apomixis) are mutated or lost during
apomicts (Carman 1997, 20(0). diploidization? Phylogenetic and genomic
evidence suggests that such confusions of
This interpretation is consistent with the
development may manifest themselves as
observed effects of climatic factors on
polyspory or polyembryony. Thus, apomixis,
facultativeness in certain apomicts. For
instead of being an evolutionary dead end,
example, exposing Dicanthium annulatum, D.
may occasionally serve as an evolutionary
intermedium, and Themeda triandra to short days
springboard in the evolution of normal or
and low temperatures increases the frequency
developmentally-novel paleopolyploid sexual
of apospory. The opposite conditions increase
species and genera (Carman 1997).
the frequency of sexuality in a partially-sexual
Dicantliium hybrid (reviewed by Asker and Mendelian Analyses of Apomixis
]erling 1992; Carman 2000). Such shifts in Essentially all Mendelian analyses of apomixis
facultativeness are expected if signal face reinterpretation if the HFA theory is
transduction pathways vary among genomes correct. Data from a variety of apomicts have
in sensitivity to morphogens (hormones, etc.) on one or more occasions (or when grown in
that accumulate in response to changing certain environments) tended to fit a
seasons and photoperiods. tetrasomic inheritance model with apomixis
1\eGnt (HKt Gooo... C.lisioo..... ApoooUis 107
conferred by a single dominant gene "A" (or primarily among weakly apomictic F IS and
linkat). The putative linkat encodes sufficient was defined by a very high level of ovule
information to , - t embryo-sac formation abortion (48-68%). The apomictic P.
and function (Sherwood, Chap. 5). However, squamulatum parent is an autoallohexaploid
such "segregation ratios" are in some cases and contains four copies of the S genome and
influenced by the environment in which the two copies of an S' genome (Patil et al. 1961).
putative segregants are grown, much like the The sexual inbred autotetraploid pearl millet
Dicanthium hybrid (reviewed above), which is parent contained four copies of the A genome.
sexual when grown in one environment and From this information, several reasonable
facultatively apomictic when grown in scenarios involving "intergenomic hetero-
another. zygosity" could be proposed for the
expression of apomixis in the Fjs.
Dujardin and Hanna (1983) concluded that
PeIlllisetl/nl squamuuuum is heterozygous for One scenario is that the A and S genomes are
apomixis beca use some Fjs in a cross between similar with regard to reproductive timing, but
aposporous P. squamulatum and sexual both differ from the S' genome. Thus, the Fj
tetraploid inbred pearl millet (PellIlisetllnl genome combination AASSS' would be
ameriCallllnl) were mostly sexual. This apomictic. However, because the S genomes
conclusion may not be warranted. Six of 20 in P. squamuiatum are heterozygous, shifts in
Fjs obtained in the study were highly facultativeness are expected because of
apomictic, but 13 of the 14 remaining Fjs were recombination among S genome chromo-
facultative apomicts with, on average, 3.2% of somes prior to fertilization. This may explain
nonaborted ovules containing aposporous the three reasonably distinct reproductive
embryo sacs. No aposporous embryo sacs phenotypes observed: strongly apomictic (6 of
were observed in one of the 20 Fjs, but only 64 20 Fjs, both genomes strongly expressed);
nonabortive ovules from this Fj were studied. weakly apomictic/highly abortive (5 Fjs, one
It is possible, given the small sample size and genome expressed more strongly than the
the possible influence of environment on other); and weakly apomictic/weakly abortive
facultativeness, that this F 1 also was (9 Fjs, one genome expressed much more
facultative. strongly than the other). In this regard, it is
interesting that most of the apomictic progeny
Many published studies may prove useful in
are late-maturing relative to pearl millet
interpreting how genetic background and
(Hanna et al. 1992), which suggests they are
environment shift degrees of facultativeness.
expressing intermediate phenotypes with
According to HFA theory, if genetic
regard to reproductive development.
background favors the expression of one
genome over another, sexual development will Replicated Mendelian analyses with consistent
be favored. A genetic background favoring segregation ratios have been conducted in
both genomes would favor apomixis, and a Panicum (Savidan 1982), Tripsacum (Leblanc et
genetic background that completely silences al. 1995b),and Brachiaria (Valleand Miles 2000,
reproductive signals from one genome would Chap 10),and each study suggests apomeiosis
enforce sexuality. (detected cytologically) is controlled by a
single dominant allele. However, other recent
In the Dujardin and Hanna (1983)study, a third
studies challenge this conclusion, e.g., the
category consisting of about five of the 20 Fjs
apomeiosis "allele" in the TripsaCHnl accession
was observed. This category occurred
studied by Leblanc et al. (1995b) is part of a
108 ~G.e-
encoded by the recipient plant is Ba!ItJ9ia, E. 1989. The evoIutian of !he female gametophyte of
angiosperms: an inIerpretatiYe key. AnnoIIi BataticD 47: 7-144.
appropriately asynchronous with that Bemett, M.D. 1977. The lineand dwation of meiosis. 1'f11osop/iaJ/
encoded by the alien segment. Hence, HFA Trarrslatiam ofthe loyal Society oflDntJon, Seiies 8. 277: 201-26.
Bemett, M.D., and H. Stern. 1975. The ~me and dll"atian of preleptolene
theory predicts that programs designed to
chrOlllOlOlllll candensatian stage in Liwmhybrid cv. black beauty.
introgress apomixis into crops will experience Proceedings ofthe loym Society oflDntJon, Series B. 188: 477-93.
difficulties producing apomictic lines that Blakey, U, CL DewOOl, and S.L Galdman. 1997. (o.segregalian of DNA
menenwilh Tripsoamfertity. ~42: 36~9.
possess anything less than one to a few alien
BrurrmiI1, R.K. 1992. VoscvGr Ibrt Fomfes tmd Genera. Whitstable, Kent,
chromosomes. U.K.: Royal Botanic Gardens, Kew, Whitstable litho, IJd.
Carman, J.G. 1986. Genetic engineering: potentials for undenIanding and
It should be possible to produce apomicts in using opannis. Presentation a1!he worhhop "The Polenliollke of
at least some crops by (i) selection for Apomixis inCrop Improyement,' AfriI21-26, 1986, Rockefeller
Foundation, Bellagio Study and Conference (enter; BeIagio, 1IuIy.
appropriate parental phenotypes (divergence
11 0 Jo" G.Canon
- - . 1997. Asynchronous expr~ion of Heslop-Horrison, 1., end A. Mackenzie. 1967. Patil, B.O., M.W. Hardas, and A.B. Joshi. 1961.
duplicate genes in ongiosperms moy cause Autoradiography of soluble (2. 140- Auto-olloploid nature of Pennisetum
apomixis, bispory, tetrDlpory, and thymidine derivatives during meiosis and squamulatum Fresen. Nature 189: 419-20.
po~embryony. Biological Journal ofthe micrDlporogenesis in Lilium anthers. Peocock, 1. 1993. Genetic engineering and
Linnean Sodety61: 51-94. Journal ofCell Science 2: 387-400. mutagenesis for apomixis in rice. In KJ.
- - . 2000. The evolulion of gometophytic Johri, B.M., K.B. Ambegookar, and PS. SrivDltro. Wilson led), Proceeding ofthe
apomixis. In 1. Botygina Ied.l, Embryology 1992. Comparative Embryology of International Workshop on Aporrixis in Ilice,
ofFlowering PIonts, Terminology and Angiosperms. Volume 1and 2. New York: Chongsho, China. New York: RockefeHer
Concepts, Vol. 3, The Systems of Springer-Verlog. Foundotion. Pp. 11-22.
Ileproduction., SI. Petmburg: Russian Kindiger, B.K., 8. Dopeng, and V. Sokolov. 1996. Peel, M.D., 1.G. Cormon, and D. Leblont 19970.
Academy of Sciences, World and Fomi~ Cytological and molecular studies of Megosporocyte callose in apomictic
Press. (in press). apomictic moize- Tripsacum hybrids. buHelgrOlS, KentlKky bluegrOls,
Carman, J.G., and U. Crone. 19B6. Agronomy Abstracts. Pp. 132. Pennisetum squomulotum Fresen, Tripsacum
Comparolive cytochemical analyses of Kohunow, A.M. 1993. Apomixis: embryo sacs Land weeping lovegroll. Crap Science 37:
megasporogenesis ~etween apomictic and and embryos formed wilhout meiosis or 724-32.
sexual Australasian 8ymusspecies. lertilization in ovules. The Plant Cell 5: Peel, M.D., J.G. Carman, ZW. Liu, and R.R.C
Agronomy Abstrocts. Pp. 60. 1425-37. Wong. 1997b. Meiotic anomalies in hybrids
Cormon, J.G., CF. Crone, and D. Riero·Lizorozu. leblonc, D., D. Grimanelli, D. Gonzolez-de-lelin, between wheol and apomictic 8ymus
1991. Comporotive histology of cell wolls and Y. Somon. 1995b. Detection of the rectiserus (Nees in lehm.l A. love and
during meiotic and oporneiotic apomictic mode of reproduction in maize- Connor. Crop Science 37: 717-23.
megasporogenesis in two australosian Tripsocum hybrids using mo;ze RFlP Pessino, S.C, 1.PA Drt;z, M.D. Hayward, and CL
Bymus Lspecies. Crop Science 31: 1527- markers. Theoretical and Applied Genetics Quarin. 1999. The moIe<ular genetics of
32. 9D: 1198-1203. gomelophytic apomixis. Hereditas 130: I-
Crone, CF., and 1.G. Carman. 19B7. Me<honisrm leblanc, D., M.D. Peel, J.G. Carman, and Y. ll.
of apomixis in Bymus rectisetuslrom Somon. 19950. Megasporogenesis and Roche, D., PCang, Z. (hen, WW. Honna, D.L
eastern Australia and New Zealand. megogometogenesis in several Tripsacum Gustine, R.T. Sherwood, and PDzias-Akins.
Americon Journal ofBatany 74: 477-96. species (Pooceael. American Journal of J999. An opospory-specifi< genomic region
(zopik, R. 19B5. Apomictic embryo IOCS in Botany 82: 57-63. is conserved between buHelgross (Cenchrus
diploid Wakkteinio geoides Willd. le~ch, H, W. Mosgiiller, 1. Schworzacher, M.D. cilioris L) and Pennisetum squoroolotum
(Rosoceoe). Acta Biologica Cracovlensio 27: 8ennen, and 1.S. Heslop-Horrison. 1990. Fresn. The Plont Journal J9: 203--08.
29-37. Genomic in situ hybridization to sectioned Rodkiewia, B. 1970. Callose in cell wolls during
Oujordin, M., and W.W. Honna. 19B3. Apomictic nuclei shaWl chromosome domoins in gross megasporogenesis in ongiosperms. Planta
and sexual pearl millet XPennisetum hybrids. Journal afCell Science 95: 335- 93: 39-47.
lquamulatum hybrids. Jaurnal afHeredity 41. Solisbury, F.B., and CW. Ross. 1992. Plant
74: 277-79. Lewin, B. 1994. Genes V. New York: Dxford Physiology. Belmont, California: Wadsworth,
Oovis, GL 1966. Systematic Embryalogy ofthe University Press Inc.
Angiosperms. New York: Wiley and Sons. lewis, W.H. 1980. Polyploidy in angiosperms: Somon, Y. 1982. Nolure ethered~e de
Frokes, LA., 1.E. Froncis, and J.I. SyktUl. 1992. dicotyledons..In W.H. lewis [ed.l, I'apomixie chez Panicum moximum Jacq.
Oimote Modes ofthe PhonerazoiL Palyplaidy. New York: Plenum Press. Pp. Ph.D. thesis, Universite Paris XI, France.
Cambridge: Cambridge University Press. 241-68. Shermon, RA, PW. Voigt, B.L Burson, and CL
Goldblan, P19BO. Po~ploidy in ongiosperms: Mogie, M. 1992. The Evolution ofAsexual Dewold. 1991. Apomixis in diploid Htriploid
rTlOIlO<oty\edons.ln. W.H. Lewis (ed.1. Ileproduction in Plants. London: Chopmon Tripsacum dactylaides hybrids. Genome 34:
Polyploidy. New York: Plenum Pr~. Pp and Hall. 528-32.
219-39 Noumova,1.N. 1993. Apomixis in Angiosperms: Stebbins, G.L 1971. Chromosomal Evolution in
Grimonelli, 0., D. Leblonc, E. EspinalO, E. Peroni, Nuce/Ior and Integumenrary Embryany. Higher PIonl5. London: AddilOn-Wesley.
D. Gonzolez-de-Le6n, and Y. Soman. Boco Ralon, Florida: CRC Press. Theissen, G., A. Be<ker, A.D. Roso, A. Kanna, J.T.
1998. Mopping diplosporous apomixis in Noumovo, 1.N., A.PM. den Ni~, and M.1.M. Kim,1. Munster, K.U. Winter, and H. Soedler.
tetraploid Tripsocurrr. one gene or severol Willemse. 1993. Quanlitolive analysis of 2000. Ashort history of MADS-box genes in
genes? Hllfedity 80: 33--39 oposporous porthenogenesis in Poo plonts. Plant Mo/e<ular Biology 42: I' 5-
Grinonelli, D., POzias-Akins, 1. Tohme, ond D. pratensis genotypes. Acta Botanica 49.
Gonzolez-de-lecin. 2000. Molecular Neerlondica 42: 299-312. Isseret, B., E.B. Eson, and 1. Muroshige. 1979.
generics inapomixis research. In Y. Sovidon Naumovo, 1.N., M.T.M. Willemse. 1995. Somotic embryogenesis in angiosperms.
ond 1.G. Cormon (eds.), Advancel in Ultrastructural characterization of opospory Horticuhurallleviewsl: 1-77.
Apomixis Ilesearch. Rome: FAD. in Panicum moximum. Sexual Plant Volle, CB. Do,ond J.W. Miles. 2000. Breeding of
Honna, W.W., M. Dujardin, PDzios-Akins, and L lleproduction8: 197-204. apomictic species. In Y. Sovidon and 1.G
Authur. 1992. Tronsler 01 apomixis in Nogler, GA 19B4. Gometophytic apomixis. In Cormon (eds), Advances in Apomixis
Pemisetum.. In .l.H. 8gin and LP Miksd1e B.M. Johri led), Embryology of Ilesearch. Rome: FAD.
(eds.], Proceedings afthe apomixis AngioSPllfms. New York: Springer·Verlag. Wolloce, 8.1991. COOdOplotion revis~ed. Journol
workshop, February 11-12, 1992, Allonlo, pp. 475-518. ofHeredity. 82: 89-95.
Geargia.lISDA-ARS, ARS-l04. Pp 3D-33.
Chapter 8
1993). Before the trait can be successfully use a species in which apomixis can be easily
commercialized, more information is assessed, preferably in a format that can be
required about the impact of environmental quantified, to facilitate the study of allelic
variables on the expression of apomixis and differences and epistasis.
their interaction with sexuality, particularly
The inheritance of apomixis is typically
with regard to fertility and resource
assessed by crossing sexual and apomictic
allocation. Ultrastructural studies are also
biotypes, often using the sexual plant as the
required to more clearly elucidate the
maternal parent. Although it is usually also
cytological and histological events involved
possible, and sometimes necessary, to perform
in apomixis, such as the role of differential
the reciprocal cross, it requires the separation
callose deposition (Carman et al. 1991;
of recombinant (Bu and Bm hybrids) from non-
Leblanc et al. 1995a; Peel et al. 1997).
recombinant (maternal and polyhaploid)
Attributes of A Model System progeny. Inheritance studies, therefore, require
that cross-eompatible sexual and apomictic
The innate advantages and Iimita tions of any
biotypes are available, and if a sexual recipient
system will dictate the available research
is used, tha t microsporogenesis and
opportunities and impact on the rate of
microgametogenesis are functional in the
possible progress. Therefore, before
apomictic biotypes used. Sexual biotypes are
discussing candidates, it is helpful to consider
also useful in molecular studies for assessing
the features that would facilitate their use,
the developmental importance of reintro-
specifically in a study of the developmental
duced, putative control sequences.
biology and molecular genetics of apomixis.
Types of Apomixis
Biological Attributes
Two principal mechanisms of apomixis have
To facilitate experimental progress, there are
been reported: (i) adventitious embryony, in
a number of cultural characteristics to
which the maternal embryo arises directly
consider in choosing a model system. For
from a somatic cell, and (ii) gametophytic
simplicity, the plant should be easy to
apomixis, in which the maternal embryo is
cultivate, both in vivo and in vitro. Ideally it
derived from an egg cell within an unreduced
should also be a perennial that can be easily
embryo sac. Gametophytic apomixis is further
propagated vegetatively to permit the
divided into diplosporous and apospoi'OUS
maintenance of sterile or self-incompatible
mechanisms, depending on whether the
sexual biotypes. A small, compact plant
unreduced embryo sac arises from a
stature, which does not require training, will
modification of the meiotic apparatus or from
facilitate the manipulation of large
a separate cell, respectively. Intermediates
populations, such as those used during
between the two forms of gametophytic
mutant screens and inheritance studies. For
apomixis have been reported (Gustafsson
the rapid turnover of experimental
1946), indicating that they possibly represent
populations it is convenient to use a species
modifications of a Single developmental
with a short generation time, abundant seed
mechanism. For more detailed descriptions of
set, and adequate seed fertility. This is
mechanisms, see Nogler (1984a) and
particularly important for studying apomixis
Koltunow (1993). Most current research efforts
because plants are assessed at the time of
on native apomixis focus on gametophytic
flowering or seed formation. For the
mechanisms, including studies of both
evaluation of apomixis, it is advantageous to
aposporous and diplosporous species.
For an experimental system using a native breeders alike. Fortunately, most gametophytic
apomict, it is convenient to use a facultative apomicts of all types appear to be facultative,
species in which both maternal and although they differ in the relative importance
nonmaternal ("aberrant") progeny can be of the meiotic and apomeiotic pathways of
harvested from the same plant. Rutishauser seed formation.
(1948) identified three aberrant types among
The formation of the endosperm may also be
the progeny of these plants: BII hybrids,
an important consideration in the choice of an
derived from the fusion of a reduced egg cell
experimental system. The endosperm may
and sperm nucleus; Bm hybrids, derived from
either form spontaneously, as in
the fusion of an unreduced egg cell and sperm
"autonomous" apomicts, or require the
nucleus; and polyhaploids, which arise
fertilization of the polar nuclei by a sperm
through parthenogenesis from a reduced egg
nucleus (pseudogamy). As pseudogamy only
cell. Aberrant progeny are often very useful
requires spontaneous embryogenesis and not
for studying the genetics of apomixis. Hybrid
spontaneous endosperm formation,
progeny permit the evaluation of cytoplasmic-
pseudogamous species may be considered
inheritance and the hybridization of pollen-
potentially simpler models to study. This
sterile biotypes such as interspecific hybrids
possible advantage, however, is offset by
(Savidan et al. 1994). Similarly polyhaploids
experimental constraints associated with
have been used to study the expression of
pseudogamy. In these plants, pollination is
apomixis in diploids (Nogler 1982; Bicknell
required for seed formation, so it becomes
1997).
necessary to demonstrate that the applied
Competence to form both meiotic and pollen led only to the fertilization of the polar
apomeiotic seed is also invaluable for mutation nuclei and not of the egg. This difficulty has
screening. Dominant inheritance (see below) been largely overcome, however, in species
suggests that the selective inactivation of either that can be assessed for apomixis using a
developmental pathway by a mutation will correlated cytological character, such as callose
lead to the exclusive expression of the other. deposition in diplosporous species of
This dual competency provides a useful Tripsacum (Leblanc et al. 1995a)and four-eelled
internal control. The continued formation of embryo-sac formation in aposporous species
at least one class of seed indicates that the of Panicum (Savidan 1980). Finally, there may
mutation(s) is specific to an event(s) associated be end uses for which autonomous apomixis
with apomixis and that related requirements is more sui table,such as the production of male
for floral development, megagametogenesis, sterile or self-incompatible lines, the avoidance
and embryogenesis remain intact. Similarly, as of pollinators, or for crops in which the
facultative mechanisms incorporate a balance endosperm is commercially important, such
between the utilization of parallel as grains.
developmental pathways, they can be used to
Genetic Attributes
study factors that influence that balance, such
Apomixis appears to be relatively Simply
as the impact of physiological stress and of
inherited either as a one- or two-gene trait, at
interactions between genetic modifiers.
least in the small number of cases studied.
Finally, as one clear goal of this research is to Examples include apospory in Ranunculus
incorporate apomixis into crop populations, a auricomus (Nogler 1984b); Panicum maximum
facultative mechanism is likely to provide the (Savidan 1980,1982); Penniseium squamulatum
maximum flexibility for farmers and plant (Dujardin and Hanna 1985); Cenchrus ciliaris
114 10" l.Iilbe1
(Sherwood et al. 1994); Brachuiria sp. (Valle and have already been documented, while for
Savidan 1996); Hicracium (Gadella 1991; others they would need to be established.
Bicknell et al. 2000); and diplospory in Tissue culture and transformation techniques
Tripsacum dactuloides (Leblanc et al. 1995b; are particularly valuable. They permit the
Grimanelli et al. 1998); TaraxaCllm (van Dijk et rapid micropropagation of selected types, the
al. 1999); and Erigeroll (Noyes and Rieseberg retention of somatic genotypes through in
20(0). Earlier work by Dopp (1939) indica tes vitro shoot regeneration, and the maintenance
tha tit may even be true for the fern Druopleris, of val uable genotypes through long-term cold
Simple inheritance would be particularly storage. It is also possible to recover
valuable for analyzing the molecular biology interspecific hybrids using embryo rescue,
of the trait. and to manipulate ploidy in culture through
the use of anther or microspore culture to
For the advancement of a molecular research
isolate reduced genotypes and by colchicine
program, it is particularly advantageous if a
application for chromosome doubling.
model system can be genetically transformed.
This permits the introduction of marker genes As the initiation of the embryo sac is the first
to follow segregation and recombination and decisive event of the female reproductive
mutagenic sequences such as T-DNA tags and phase, a thorough embryological analysis is
transposons to assist in the cloning of an indispensable base for an investigation of
sequences associated with the expression of apomixis, while embryology continues to be
apomixis. Furthermore, transformation necessary a t all stages of the research. Reliable
permits expression studies about putative cytological techniques are therefore essential
regulatory elements by their fusion to marker for this work. Fortunately, this is one area of
sequences and the functional testing of apomixis research that is well documented,
putative control genes by their introduction particularly with respect to the use of
into sexual or mutant genotypes. versatile, routine clearing techniques (Leblanc
et al. 1995a) that enable the use of thick
It would also be preferable to use a species
sections to visualize the complex internal
wi th a small genome size, such as Arabidopsi«,
structure of the ovule.
to facilitate the identification of critical loci.
Similarly, an ideal system would already be Quantifying Apomixis
cha r ac te r ized with respect to genomic Most studies of apomixis require a method
sequence and mapped morphological and for determining the presence of apomixis, and
molecular markers (deletions, translocations, when possible, for quantifying its extent,
RFLPs, RAPDs, transposons. ete.) for the either with respect to prevalence in a
localization of loci. Finally, it would be population or level of expression in a single
advantageous. although not essential, to use genotype. Measuring apomixis, however, is
a crop species to assist in the transfer of a difficult task. When considered as a genetic
research findings into practical outcomes. event, apomixis is reproduction through seed
without either allelic segregation or
Experimental Methods recombination. Assessment, therefore, should
The success of any program on apomixis will be strictly based on whether allelic
depend on both the natural attributes of the rearrangement occurs. This is seldom
species used and on the power of the practical experimentally and most systems for
techniques available to manipulate it assessing apomixis use a correlated character.
evperirnentallv For some species, techniques The reliabilitv of these correlations is clearly
I.,....ta1 lioIotY .f ApooUis
Model Sy.'elM le StillytileGeoetic. ood O... 115
induced mutation or by the accumulation of these genes, specifically with their expression
mutant alleles. A second approach has been during gametogenesis and early
to attempt the transfer of the trait by embryogenesis and that this may be associated
introgression from an apomictic relative. with mechanisms of maternal inheritance
(Vielle-Calzada et al. 1999).
Two important advantages of mu tagenesis are
that it can be conducted on a species without When mutations that alter megasporogenesis,
any known close apomictic relative and it has megagametogenesis, and embryogenesis are
the potential to rapidly provide the already known, it may be possible to create
experimental material required in a an apomictic mechanism by the accumulation
characterized genetic background. The of appropriate aIleles within a single genotype.
principal difficulty with this approach is that In potato, the homozygous representation of
methods must be developed for screening the ds-l allele Significantly reduces chiasmata
mutant populations for apomixis without any frequencies on all chromosomes during both
foreknowledge of the developmental megasporogenesis and microsporogenesis,
mechanism(s) that will arise. In particular, it leading to high levels of 2n gametes through
is not known whether a single mutagenic first division restitution (FDR) Oongedijk et al.
event can cause an unambiguous phenotypic 1991). If this could be combined with
change that can be identified as either an intact parthenogenesis, true potato seed could be
apomixis or a recognizable component of the generated through a synthetic diplosporic
trait. The screening technique must therefore mechanism (Hermsen et al. 1985).
be based both on the appearance of
When an apomictic relative is available, it may
developmental anomalies that are
be possible to transfer the trait into a
characteristic of apomixis and, ultimately, on
characterized system by introgression. This
the retention of the maternal genotype. One
typically involves a backcrossing program
good example of this approach is the mutant
using the sexual species as the recurrent
screening of Arabidopsis thaliana for mutations
parent.lntrogression has been used to attempt
leading to the formation of seed without
the introduction of apomixis into several crop
fertilization (fie and fis mutations) (Ohad et al.
species, normally as part of a crop
19%; Chaudhury et al. 1CJ97). The screens were
improvement program (Asker and Jerling
based on the identification of plants that
1992). Examples of this approach include
developed elongated siliques without prior
attempted transfers from Pennisetum
fertilization. The advantages of Arabidopsis for
souamulatum to cultivated pearl millet
this approach are highlighted by the authors'
(Dujardin and Hanna 1985), Elvmus rectisetus
use of male sterile mutants (popl and pistillata)
to Triticum (wheat) (Torabinejad and Mueller
to avoid self-fertilization, and the use of GUS
1993),and Tripsacum dactyloides to Zea (maize)
as a paternal marker in test crosses with the
(Leblanc et al. 1995b). Similarly, it may be
mutants. The description of these mutants is
possible to transfer the trait from apomictic
a particularly exciting outcome, providing
crucifers, such as Arabis holboellii (Roy 1995)
evidence for the involvement of chromatin
or Draba oligosperma (Mulligan and Findlay
remodeling factors in the control of cell
1969),to Arabidopsis thaliana to take advantage
division at the time of fertilization
of the versatility of this model system. Unlike
(Grossniklaus et al. lCJ98; Kiyosue et al. 1999).
synthesis, transfer has the advantage that the
Furthermore, recently reported data indicates
program is based on a known, functional
that methylation is critical to the regulation of
apomictic mechanism. Furthermore, the
inheritance of that trait can be monitored than 400species of flowering plants from more
through each generation, providing than 35 families (Asker and Jerling 1992;
information on its genetic basis in the system Carman 1997).
under study. Finally,as indicated above, when
Which is the best to study? Different species
the recipient species is a crop, the product may
clearly have different advantages. The unique
be of immediate commercial value. The
biology of ferns, for example, presents some
principal disadvantage of transfer is that the
unusual opportunities to study apomixis
availability of apomictic relatives typically
(Sheffield and Bell 1981). The events of
dictates the mechanism used. It has also
megasporogenesis and megagametogenesis
proven difficult to incorporate the trait into
are physically separated in these plants,
obligate sexual crops with a significant level
permitting the study of the individual
of expression. The reason for this difficulty is
component processes of apomixis in isolation.
unclear. It may be associated with the
Unlike the angiosperm embryo sac, the free
inheritance of modifiers or an important
living fern gametophyte is isolated from
association with polyploidy, which is lost
parental influence, and, in some systems,
during the backcrossing program. Alter-
parthenogenetic development of the
natively, the difficulties encountered may be
sporophyte can be induced in vitro (Sheffield
more a problem of experimental scale. Most
and Bell 1987). Furthermore, the sporogenic
crop species do not have a close apomictic
tissue is relatively exposed in ferns,
relative so introgression requires wide
simplifying the study of events associated with
crossing. This often results in poor fertility
the avoidance of meiosis. It is interesting to
among the progeny and little, if any, crossing
note that Manton (1950) reported that
over during meiosis. Large populations need
unreduced spores of Crytomium arose from
to be formed and assessed for ploidy and
tissue immediately adjacent to meiotic tissue
apomixis, but traditional methods of
in the sporangium. a situation closely
chromosome counting and thin sectioning are
analogous to the development of aposporous
too labor-intensive to be practical for most
initials in the angiosperm ovule. Finally,as the
research teams. Recent advances in DNA
fern sporophyte develops wi thout the need for
quantification through flow cytometry and
an endosperm, difficulties associated with the
analytical cytology using clearing techniques
development of that tissue do not arise. There
are overcoming these difficulties. Over the past
are, of course, several limitations in using ferns
two years, researchers in the maize/Tripsacum
as model systems, particularly for a molecular
program have screened more than 200,000
study of development. They are often large
plants for chromosome number using flow
slow-growing plants that can be difficult to
cytometry (Savidan, personal comm.) and the
cultivate, and they present some real
experimental populations have been
challenges when conducting controlled
advanced to the BCs generation.
fertilizations with motile sperm cells. Finally,
Development of a Model System from an very little is known of the molecular biology
Existing Apomict of this group, which would impede progress
Apomixis occurs throughout the plant in any molecular study of apomixis.
kingdom. Species utilizing various forms of
Apomixis occurs throughout the angiosperrns
asexual reproduction that involve
gametophytic structures have been recorded incl uding representati ves of both
monocotyledonous and dicotyledonous
among the algae, pteridophtyes, and in more
genera. Many of the most comprehensive
118 Roll A.lilbeI
studies of apomixis used monocotyledonous backcrosses. The results indicate that apospory
species, principally relatives of the cereals, such is inherited as a simple dominant Mendelian
as Eivmu« (Torabinejad and Mueller 1993), trait in this system, however, the allele
Tripsacum (Leblanc et al. 1995a), Panicuni conferring apomixis could only be transferred
(Savidan 1982), and important forage grasses, in a diploid or polyploid gamete. Nogler noted
such as Penniseium (Dujardin and Hanna 1985; that this mechanism could explain the very
Ozias-Akins et al. 1998)),Brachiaria (Valleet al. close association observed between
1994),and Puspalum (Bonilla and Quarin 1997). polyploidy and apomixis in native systems.
In contrast, recent results with the
The inheritance of apomixis has been
diplosporous genus Taraxacllm (van Dijk et al.
particularly well characterized for Panicum
1999) indicate that a two loci model better fits
maximllm. Savidan (1990, 2000) partially
the data obtained from controlled crosses
ascribed his success with this system to the use
within this taxon. Taraxacllm is an attractive
of apomictic and sexual forms from within the
experimental system because it is a well
same species and at the same ploidy level. This
studied ecological model (Richards 1986) and
important advantage is shared with only a very
it has also been successfully transformed (Song
small number of studied apomictic taxa. In
and Chua 1991).
contrast, most known apomictic species appear
to have either evolved away from their One dicotyledonous taxon that appears to be
immediate sexual progenitor(s) or their well suited for use as a model system is
progenitor(s) no longer exists. As a result, Hieracium subgenus Pilosella (Koltunow et al.
apomicts have often been crossed with a related 1995). The plants are small herbaceous
but distinctly different sexual species. In such perennial daisies that are easily propagated
studies, the subsequent analysis of the progeny and maintained in the greenhouse. Hieracium
must consider the inheritance of the breeding is a long-day plant, flowering in response to
system while also allowing for unrelated effects extended photoperiods (Yeung 1989).
resulting from interspecific hybridisation. Photoperiodicity is a very useful experimental
Despite this caveat, however, mapping tool as it enables the programmed production
strategies have led to the isolation of molecular of flowers at any time during the year using
markers linked to apospory in the grass genus day-length extension lighting. Both H. pilosella
Pennisetum (Ozais-Akins et al. 1993, 1998). In a and H. aurantiacumset seed within 3-4 months
similar approach, colinearity between grass of germination, allowing 3-4 generations per
genomes has also been used to propose a annum (Bicknell1994a).
linkage to apomeiosis in Tripsacum (Leblanc et
Apomictic biotypes of Hieracium subgenus
al. 1995b) and Brachiaria (Pessino et al. 1997;
Pilosella develop seed by facultative apospory
see Grimanelli et al., Chap. 6).
coupled to autonomous endosperm
Some dicotyledonous species have been used development. Pollination is therefore not
previously as model systems, most notably required for the formation of maternal seed,
Poientilla (Rutishauser 1948;Asker 1970, 1971); and apomixis can be scored by quantifying the
Taraxacllm (Richards 1973;van Dijk et al. 1999); seeds that set after the exclusion of pollen. As
Hvpericuni (Noack 1939); Ranunculus (Nogler with Taraxacllm, the simplest method of
1984a); and Hieracium (Bicknellet al. 2000).The excluding pollen is to decapitate the immature
inheritance of apospory in Ranunculus has been bud, removing both the anthers and stigmas
particularly well studied (Nogler 1984b) (Ostenfeld 1906; Richards 1986).
through four generations of crosses and
ModeIS,.I_ I. SI1'" ~. Geo.tk. ad o.""""h11 lioIotY of ApoooUis 11 9
The embryology of Hieracium is now well - - . 1996. Isolation of redoced genotypes of Hieradum pilose/Ia
using anther culture. Plant (eH, flSSue Organ (uhure 45: 37-41.
documented (Koltunow et al. 1998)and a range Bicknell, R.A., N.K. Borll, and A.M. Kallunow. 2000. Monogenic
of tissue culture techniques have been inheritance of apomixis in two Hieradum speOes with distinct
described, including methods for shoot developmental mechoniSl11l. Heredity 84: 228-37.
Banilla, JR., and CL Quarin. 1997. Diplosporaus and aposporaus
regeneration from leaf tissue (Bicknell 1994b), apomixis in 0 penta~oid race of Paspolum minut Plant Science 127:
the recovery of red uced genotypes from an ther 97-104.
culture (Bicknell and Borst 1994), and an Carman, lG.1997. Asyn<hranous expression of duplicate genes in
angiosperms may cause apomixis, bispory, tetrospary, and
efficient genetic transformation system pa~ernbryany. Biological Journal ofthe linnean Society 61: 51 -94.
(Bicknell and Borst 1996). Carman, lG., CF. Crane, and O. Riere-Lizarazu. 1991. Comporative
histology of cell walls during meiotic and apomeiali<
megasporogenesis in two hexaploid AuIlralosian f1ymus speOes. Crop
Summary Science31: 1S27-32.
The choice of experimental system is a Chaudhury, A.M., LMing, CMiller, S. Craig, and ES Dennis. 1997.
fundamental decision in any research program. Fertilizatian-independant seed developmenl in Arabidopsis thaliona.
Proceedings ofthe Notional Academy ofScience IUSA194: 4223-28.
The relative strengths and weaknesses of the Dopp, W. 1939. Cylalogische und genetisdle unteouchungen inllefholb
chosen system determine the opportunities der ga"ung Oryapteris. Planta 29: 481-533.
available to the researcher, and often dictate Dujardin, M., and W.w. Hanna. 1985. Cylology and reproduclian of
reciprocal bockcrDlsel between pearl m~let and sexual and apomictic
end uses for the information gathered. hybrids of pearl millet x Pennisetum sqvamulotum Crop Science 25:
Although different research objectives often 55--62.
lead to the selection of different systems, there Godello, 1WJ. 1991. Variotion, hybridizalian and reproductive biology of
Hieraaum pilDle/1a L Prac. Kan. Ned. Akad. V. Wetensch. 94: m-
is a need to develop either one, or a small 88.
number of, model system(s) for studying the Grimanelli, D., O. leblanc, LEspinoso, D. Gonzalez de Leon, and Y.
Savidan. 1998. mapping di~osporaus apomixis inlelraploid
developmental biology and molecular genetics
Tripsocum: one gene or several genes? Heredity 80: 33-39.
of apomixis. Several systems have been GrDllniklaus, U., l-P. V"ielle-Calzado, MA Hoeppner, and W.B. Gogoona.
proposed for this purpose, including some °
1998. Maternal control of embryogenesis by MfOEA, PoIytomb
group gene in Arabidopsis. Science 280: 44~50. ..
modified sexual systems such as Arabidopsis, GuslafslDn, A. 1946. Apomixis in Higher Plants. wilds Univ. AmJcr. N. F.
and others that are naturally-occurring Avd. 2, 42(3).
apomicts. Irrespective of the plants used, it is Hermsen, lG1,M.S. Ramanna, and LJangedijk. 19B5. Apomictic
approach to inlrodoce uniformity and vigour into progenies from true
important that collaborative links are palata seed (TPS). 26th Planning Conference of the Intemotional
established and maintained between teams Potato Centre. Dec 12-14, 1983. Lima, Peru: Training and
working on apomixis research in order to Communications Department of ClP. pp. 99-114.
Jangedijk, L, CB. HuIlen, lM.A.SA van der WoIk, and S.U. Schuurmans
expedite the advance of our knowledge in this Stekhovefl. 1991. Synoptic mUlants in potala, SoIonum tuberDlUfn L
fascinating field. Ill.Effect of the Ds·1/ds-lIDC111 (desynopsis) on geneli<
recombination in male and female meiosis. Genome 34: 121-30.
KiyDlue, 1, N. Ohod, et01. 1999. Control of fertilization-independent
References endosperm development by the MfOEA poIycorm gene in
Asker, S. 1970. Apomixis and sexuality in the Patentilla argentea complex. Arabidapsit Proceedings ofthe National Academy ofScience (USA)
Hereditas 66: 189-204. 96: 418~91.
- - . 1971. Apomixis and sexuality in the Patel7filla argentea Kahunow, A.M. 1993. Apomixis: Ermrya lOa and einbryDl farmed
complex. Hereditas 67: 111-42. wilhaUl meiosis of fertilisation in ovules. Plant (eR. 5: 1425-37.
Asker, S.L, and LJerling 1992. Apomixis in Plants. Boco Raton, Aorido: Kohunow, A.M., RA Bicknell, and A.M.. Choudhury. 1995. Apomixis:
CRC Press. Molecular stralegies for the generation of genetical~ identical seeds
Bidmell, R. A. 19940. Hieradunr. Amodel system for Iludying the withoUl fertilization. Plant Physiology 108: 1345-52.
molecular genetics of apomixis. Apomixis News/ener 7: 8-10. Kohunaw, A.M., S.D. Jahnson, and RA Bicknell. 1998. Sexual and
- - '. 1994b. Microprapogation of Hierodum ourontiorum. PIont (e/I apomictic development in Hieradum. Sexual PIont Repradudionll:
Tissue Organ (uhUfe 37: 197-99. 213-30.
- - . 1997.lsoIotion of adiploid, apomictic plont of Hieradum leblanc, 0., D. Grimanelli, D. Gonz6lez de lean, and Y. Savidan. 1995b.
aurantiacum. Sexual Plant Reproduction 10: 168-72. Detection of the apomictic made of reprodoclion in maize·Tripsacum
Bicknell, R.A., and N. K. Barst. 1994. Agrohaderium-mediated hybrids using maize RFlP markers. Theoretical Applied. Genetia 90:
transformalion of Hieradum aurantiorum. International Journal of 1198-1203.
Plant Science 155(41: 467-70.
120 1... 1. ......
l.ebicIK. 0.• M.D. Peel. lG.Carman,lIIld Y. Savidan. 1995o. - - . 1986. PIont Breeding SyslerrJ1 London: Gearge Alien lIIld
Megasporogenesis and rneglIfImetogenesis in several TripsDam Unwin.
species (POlK8Qe). Ameriarn 10lmlI af80IrIJy 81: 57-63. loy. BA 1995. The breed"1Il!I systems afsix species afAtobis
McrIIon, I. 1950. Apogomous ferns: The general ~.1'roIJIems (Brassicoceae). Ameriarn JOInIIlI afBolany 82(7): 869·77.
afCytology and Evolu/ion if tile I'reridophyto. Cariridge, U.l: lutishauser, A. 1948. PIeudogamie unci PoIymocphie in der Gattung
CanDidge University Press. Potenlilo. Arch. Julius-kmus·S~hoog f. Vererb.·ForsdJ. 23: 267--424.
lUigon, G.A.. and IN. Findlay. 1969. Sexual reprodUdion unci Savidan, Y. 1980. o..ornosomal and embryoIagical analyses in sexual x
agomospermy in the genus Draba. CDnadion. JOImIII afBolOny48: aponiclic hybrids afPonicum maxillUll Jacq. Theoretical Appied
269-70. GenrIics. 57: 153-56.
Nm. lL 1939. iiber ~Kreuzoogen YI. - - . 1982. Nature et11Illlldhe de rapornixie chez Ponicum
Fortphflanzungsverhiiltnisse 1nl Baslarde WlIl /typ«iam perlorarom maxinllmJcKq. TrrMlux & OoaJmenls ONON 153: 1-159.
LZ. Indukl. Abstamn VererMgsJeln. 76: 56~ 1. - - . 1989. Another "working hypothesis" far the control af
Nogler, G.A. 1982.11aw laoblain diploid apomidic IDIUICIIAJs I1lIricOl7llS parthenogenesis in Ponicum. Aptxnixis Newslener1: 47-51.
pbmnot fOlllli in the wid s1ale. Bot. Helvel. 92: 13-22. - - . 1990. The gene~c cantrol afaporrixis. Aptxnixis Newsletter 2:
- - . 19840. Gametaphylic aporrixis. In B.M. JoIvi (ed.). 24.
Embryology ofAngiosperms. New York: Sjlmger.Yer~. - - . 2000. Aporrim: Geneoo and 8reeding. In 1 Jaoo (ed.!.
pp.475-518. P/ont Breeding lmews. New York: John Wiley and Sans. 18: 1l-86.
- - . 1984b. Geneoo afapospary in apomidic lanoocuIus Som. Y., O. l.eblan<, and J.8erthDl1d. 1994. The lIIIize- TripsDCPII B(2
auncoroos. Y. Candusian. Bolwr. Helve/. 94 (2): 411-22. prodUdion - gelling Iale lawin the bet? Apotrixis Newslener 7: 27-
Noyes, 1.0., and LH. Rieseberg. 2000. Two independenllod control 28.
agomospermy (apomixis) in the ~ IIawering planl frgeron Sheffield. E., and P.L Bel. 1981. Experimenlal sludies afapospary in
annws. Genetics 155(1): 379-90. ferns. Ann. Bot. 47: 187-95.
Ohad, N., LMorgossian, H. Hsu, C. Yffiallll. P. lepelli. and I.LFisdIer. - - . 1987. Cmen! SIue5es of the Pteridaphyle life (yde. TIre
1996. Amutation that alows endosperm deveIapment WIthout Botanical leriew 53: 442-90.
fertiization. Proceedings aftile IIaIioID AiDtJemy afScience (IISA) Sherwood, R.T.. c.c. &erg. unci BA Young. 1994.lnherhance af~
93: 5319-24. in buffe~rass. (rop Science 34: 1490-94.
aslenfeld, C.H. 1906. Experimenlal and cytological sludies in the Hierada. Song. Y.C.N., and N.·H Chua. 1991. TISSUe cuhure and genem
I. CDstrUlion and hybridization experimenls with some species af transformalian afDandelion. AcIl1. 1/011. 289: 261-62.
Hieracio. BoI. rllhsJcr. 27: 22s-48. Tarabinejad, 1,and IJ. Mueler 1993. Genome analy5is of intergeneric
Ozias·Altins, P., E.L Lubbers, W.w. Hanna, and lW. Mdlay. 1993. hybrids afaponiclic and sexual Australian E1ymus species with wheat.
Transmission afIhe aponidic mode afreprodUdion in Pennisetrmr. barley and rye: implicUlion for the transfer of apomixis lacereak.
cG-inherilance of the trah and molecular IIIIrktn. T1reoteticaI AppNd TheoreliarJ App'ed Genelia 86: 288-94.
Genelia 1993 (85): 632-38. Yale. C.8. da. C. Glienke, and G.O.c. Leguizamon. J994.lnherilClKe af
Ozias·Altins. P., D. lache, and W.W. llama. 1998. Tl!lhl duslering and aponim in Brachiario, a Iropicalfarage grass. Aptxnixis Newsletter
henizygosi!y afapomixis-&nked rnoIeoD IIIIrktn in I'erJniseIum 7: 42--43.
squamu/alum inp5es genetic lDI1lIoI of apospary by a cmrgenllocus Yale, C.8. do, and Y. Savidan. 1996. Genetics, cytCl!leneoo lIld
thal may have no alelic farm in sexual genotypes. Proceedirgs aflire reproductive biology afBracJrma. In lW. Miles. B.L Moass and C. 8.
National Academy afScience (IISA) 95: 5127-32. do Yale (eels.!, Brachioria 8io1ogy, Agronomy, and Ill1Jravement.
Peel, M.D., lG.Call1lln. and O. LebIanc. 1997. Megaspore arIose in Can, CaIombia: 0AJ/EM81APA. pp. 147·163.
aponiclic buffelgrass, Kenndy b1uegrass. I'ennisehxn fqlJl1roolatum van Dijk, PJ.,I.C.Q. Tas, M. Fa~ue, uncIT. Bakx·Schotmon. 1999. Classes
Fresen, Tripsocum L. and .epingIavegrass. Crop Science 37: 724- between sexlJl1l and apomidic dandelions (Taraxacum). 11. The
32. breakdown of aponim. Heredity 83 (6): 71 S-21.
Pessina, S.c.. lP. Ortiz, O. Lebm. C.8. do Yale. C. mns. and M.D. Y"lele-Calzada, J.p, 1 Thomas, el01. 1999. Maintenance afgenomic
Hayward. 1997. lden!ifieUlion afa lIIIize linkage group relaled la imprinting Ul the ArabitJopsis medea Locus requires zygoM DDM 1
aponim in Brochiaria. TheoreliarJ Appied Genelia 94: 439-44. acliviIy. Genes ond Oeve/opmenlI3(22): 2971-82.
Richarels, A.J. 1973. The origin afTaraxQMI agomaspe6es. BioIogicrJ Yeung. E.c. (1989). HierCKium. (R( HandbookofFlowering. 8am IUlon,
JOInIIlI aftile I.iJneon Society 66: 189-211. Florida: (RC Press.
Chapter 9
Screening Procedures to Identify and
Quantify Apomixis
OUVIER LEBLANC AND ANDREA MAZZUCATO
::::.':".::;' /~ ~-""
--r.........,. -.
.lI ...... ES- rri lk
IUK.., 11II ....,..
"1IcoIy(9" U...d".d ES - (2HllI.
.............. ..........
(APOSPORY) -----.. 0 2~2~11 Hierad_type
<,
hull., / ~ , . . ......
c~:=... e."2·~-
0•
-- ~ 2. 0
IlOrh·"'"
".."....,..
Pd.• I,lIltIc
..............
~~ 'l
II....
. ,._ M 0~ .lIe URr.;.co; ES - 12HllI.r.........,. -.
_ _ ..........
Figure 9.1 ,.\echanisms of pseudogamous gametophytic apomixis: consequences and comparison with sexual
reproduction.
Su..... P""Mom to Ideotily Old o-tiIy"'is 123
synergid(s) and one egg cell (Polygonum-type (Naumova et al. 1993; Peel 1993; Peel et al.
development). All nuclei from the meiotic ES 1997), but little light was shed on their
are reduced (n chromosomes). Double reproductive behavior. (Peel 1993; Peel et al.
fertilization is required for embryo and 1997). However, recent studies on callose
endosperm development to begin. Each of the deposition patterns and the dynamics of beta-
sperm cells is involved in different fusion 1,3-glucanase (HpGluc) expression in
events: fusion wi th the egg restores the aposporous Hieracium provide new insights
sporophytic chromosome number, and fusion into the role of callose (Tucker et al. 2(00): both
with the central cell produces a nutritive tissue, altered patterns and persistence of callose
the endosperm. during megasporogenesis occur in apomictic
plants when compared to sexual ones. In
In diplospory, meiosis is totally omitted (e.g.,
addition, the HpGluc enzyme might also play
Antennaria-type) or perturbed (e.g.,
a role in promoting the aposporous pathway
Taraxacum-. Ixeris- or Allium-types). In both
over megasporogenic callose dissolution.
cases, ESsform through three or more mitoses:
(i) from MMCs behaving as unreduced Megagametogenesis in apomicts and in their
megaspores (Antennaria-type); (ii) from sexual counterparts is usually similar. One
unreduced megaspores after restitution exception is the aposporous ESstructure in the
nucleus (Taraxacum- and Ixeris-types); or (iii) Panicoideae and Andropogoneae subfamilies
from 2n megaspores after premeiotic (Poaceae), which is a 4-nucleate ES (Panicum-
chromosome doubling (Allium-type). The type, after Warmke 1954), the sexual
characteristic meiotic sequence (MMC, dyad, counterparts producing 8-nucleate ESs
tetrad) is absent and callose deposition does (Polygonum-type).
not occur (Naumova et al. 1993; Leblanc et al.
Embryo and Seed Formation
1995a)or is strongly disturbed in diplosporous
The forma tion of viable seeds usually requires
pathways (Carman et al. 1991;Peel et al. 1997).
endosperm differentiation. This is achieved in
Lack of callose deposition also has been
apomicts through (i) pseudogamy (single
reported in meiotic mutants of Medicago saliva
fertilization of polar nuclei) or (ii) autonomous
tha t produce u nred uced embryo sacs
endosperm development (central cell develops
(Ba rcaccia et al. 1996) and in apomictic
autonomously). Most apomicts are
Al1lel1l1aria hybrids obtained from parents
pseudogamous.
displaying floral asynchrony (Carman 2(00).
In apospory, several ESs generally differentiate Megagametogenesis in both reproductive
from nucellar (soma tic) cells. In contrast to modes appears to take the same amount of
diplospory, which seems to result from genetic time, but time periods for megasporogenesis
lesions directly affecting meiosis, some authors differ. Such differences have been documented
have stated that meiotic and apomeiotic for several aposporous species (Ranunculus
developments are independent in apospory allricomlls, Nogler 1984; PalliCllm maximllm,
(Harlan et a I. 1964; N ogler 1984). Both Savidan 1982a; Paspalum notaium, Martinez et
developments can theoretically occur at the al. 1994) and two diplosporous species
same time within the same ovule, but usually (Tripsawm zopilotense and T dactyloides, Leblanc
the legitimate sexual line is eliminated in and Savidan 1994). The complete maturation
subsequent developmental stages. Abnormal of the apomeiotic ESbefore the meiotic ES may
patterns of callose deposition have been contribute to the failure of unreduced egg cell
observed in various aposporous species fertilization: by the time the pollen tube reaches
124 llIme< laWoo< ... Aoonto .........
the ovule, unreduced egg cells may not be combination of failure or success in meiosis
receptive. This loss of receptivity is not yet well and fertilization (Table 9.1). A fairly strict
understood, but several hypotheses have been genetic control for both the formation of
proposed, including chemical or mechanical unreduced ES(reviewed by Sherwood, Chap.
barriers (e.g., a complete cell wall around the 5) and the degree of apomixis (Asker 1980)has
egg) and a temporal window of receptivity, been reported in most taxa studied.
among others.
Levels of Screening and
Consequences of Apomictic Seed
Formation Related Tools
In sexual reproduction, the two gametes that There are several indicators of apomixis,
fuse are produced through meiosis. Sexual including high frequency of multiple
development allows genetic recombination seedlings, high seed fertility in plants expected
and segregation, thereby enhancing genetic to be sterile (e.g., wide hybrids, triploids,
diversity. Aside from strict autogamy and from autopolypJoids, and aneuploids), homo-
the very specific case of permanent geneous progenies, etc. (Bashaw 1980;Hanna
translocation heterozygosity (Ellstrand and and Bashaw 1987;den Nijs and van Dijk 1993).
Levin 1982), offspring from sexual plants are They are sometimes difficult to use in the case
new genotypes. Apomictic pathways are of wild materials and, in all cases, further
characterized by unreduced egg cell investigation is required to assess apomixis
parthenogenesis, resulting in offspring that are type and level of expression. The relative
exact genotypic replicas of the mother plant. advantages or disadvantages of the screening
However, genetic recombination may occur procedures presented here are discussed
during apomictic reproduction in plants that further in "Choosing Suitable Procedures."
show partially synaptic and restitutional Analyses at the Plant Level
meiosis or somatic DNA rearrangements 1. Molecular markers cosegregating with
(Richards 1997). apomixis. To date, the identification of
Complete (100%) maternal progenies are isozymic or molecular markers strongly linked
recovered when the mother plant reproduces with apomixis is the only procedure for
through obligate apomixis. But generally, detecting apomixis prior to flowering.
apomixis is facultative and progenies comprise Molecular marker-based analyses in
various types, each resulting from a different apomicts were conducted either to
investigate the molecular basis of
Table 9.1 The four theoretical offspring classes in progenies apomixis, to assist in transferring
from facultative pseudogamous apomids. Note that
apomeiotic mechanisms can induce chromosome losses and apomixis into crops, or to ultimately
result inunbalanced ullleduced female gametes. isolate the gene(s) responsible for its
regulation. Segregating progenies or
Egg eeldevelopmellt after
bulk segregant analyses were used
Fusioa with a Parthenogenesis after determining the reproductive
Egg ceH orig. sperm cell (+8) (+0)
behavior on the basis of
Reduced megospore Sexuality (Poly)ha~oidproduction cytoembryological observations or
after meiosis: ngamete n-n offspring n+O offspring
progeny testing. Because of
New genotypes New genotypes
conflicting results, debate continues
Apomeiosis (apospory "Genomi< accumulation" Apomixis over whether apomixis is controlled
or di~ospory): 2n+n offspring 2n+0 offspring
2n gamete New genotypes Maternal genotypes by a single locus or by multiple loci.
Scmoiot P........... Ideotify . . o-tiIy Apooobis 125
However, molecular markers (RFLp, SSR,and megasporogenesis (i.e., MMC beha vior,
AFLP) for apomixis, apomei osis. and occurrence of meiosis products, see Table 9.2)
parthenogenesis have been reported for in the case of diplospory, on nucellar cell
several aposporous genera iPennisetum: Ozias- initiation in the case of apospory, and on
Akins et al. 1993, 1998;Cenchrus: Gustine et al. differences in mature ESs for apospory of the
1997; Roche et al. 1999; Braclriaria: Miles et al. Panicum-type. To standardize time of
1994; Pessino et al. 1997, 1998; POll: Barcaccia sampling, pistils can be classified according to
et al. 1998; Erygon: Noyes and Rieseberg 2000), pollen developmental stage if the flowers are
and in diplosporous Tripsacum dactyloides hermaphrodite, or by size if they are
(Leblanc et al. 1995b; Kindiger et al. 1996; monoecious. In pseudogamous species
Blakey et al. 1997;Grimanelli et al. 1997a). showing early embryo divisions before
anthesis and endosperm formation
2. Cytoembryology. Cytoembryological
(precocious ernbryony). cytoembryological
differences between sexual and apomictic
observations within ESscan also help identify
developments appear at different times.
and quantify parthenogenesis (Kojima and
Observations to determine the origin of ESs
Nagato 1992a; Naumova et al. 1993).
are therefore based on differences in
Table 9.2 Main characteristics ofmegasporogenesis and megagametogenesis during both sexual
reproduction and gametophytk apomixis
Meiosis Megasporogenesis CytClelllbryology Megogametogenesis Referen<es
Callose
5EluAUTT Completed. The chalazal CDIose depositillll in Meioli< sequence Mature 8·nucleate ES [I] Pennel ond
megaspore of the tetrad Angiosperms prodlKing (MM(, dyad, tetrad). forms from chalozal Robem, 1990.
develops into 1IIIES. mono- and bispori< reduced megaspore [2] Rodkiewi<z, 1970.
~hered expression of embryo soo [2]. through 3(Polygonum· ·Hen, 1971.
arabinogolodllll protein type) or more mnoses. ·Russel, 1978.
was shown to be Mature ES ore produced ·Dumos lIIId
ossoc:iated with sexual loner than in opomeioHc Mogensen, 1993.
development in (meiosis delays
Pisum. [I]. megogamelogenesisl
ApOMIXIS ·Nogler, 1984.
·Crone, chop. 3.
~pospory Meiosis is inniated but Yes (meioli< produm). Concomnllllt ES forms from somati< .Miintzig, 1940.
generally foils soon DDisturbed callose development of the cells through mnoses. ·Wormke, 1954.
or lot1er. ponerns may ind"Kate reproductive cen Polyembryony: Several -Burson ond Bennen,
opospory. through meiosis somoli< cells may 1970.
lsexuoli1yl and somali< develop. Reduced ES Cllll ·Young etal., 1979.
cell(s) through mnoses be formed. Poni<um·type: -Sovidon, 1982b.
after enlargement. 4·nudeate ES. -Naumova etal., 1993.
Hierocium-type: ES ore -Tucker etal., 2000
similor to sexuals.
Diplospory ~ntennorio-type: meiosis No callose deposition MMC enlorges ES forms from the -Yoigt and Boshow,
is totally over·passed. in megosporocyle (TripstJcum spp., reproductive ceo. No 1972.
Taroxocum· and Ixeris- cell walls. Erogrostis curvula) or polyembryony. GeneroUy -Crlllle and (ormoo,
types: meiosis foils early elongates (Elymus similor to sexuality. 1987.
producing 0 restitution ree/ise/us). Relation Binucleate ES shope -Carmllll etal., 1991.
nucleus. ~16um-lype: berween enlorgementl con be charoderisti< -Kopmo and Nogoto,
endomnosis Antennoria type ond (Tripsacum) Four 1992b.
elongation/Toroxacum nucleate ES described in -Peel el 01., 1997
type? Erogrostis currula
126 Ohier leW.c ....... - . . -
be useful for progeny tests and should proven flow cytometry to be a rapid and
therefore be considered. Progeny tests are reliable procedure for determining the mode
usually pe rformed on seedlings or fully-grown of reproduction (Mazzucato et a\. 1994 ;
plants, but other tissues from earlier Brautigam and Brautigam 1996; Grimanelli et
developmental stages, such as ovaries, a!. 1997b; Gupta et al. 1998; Naumova et a\.
endosperms or seeds, can also be used . 1999; Matzk et a!. 2000). Another option for
DNA content estimation of the endosperm
1. Analysis of pollinated ovaries or seeds.
nucl ei is to combine staining with 4'-6-
Determining ploidy levels in pollinated ovaries
diamidino-2-phenylindole (DAPI), fluo-
or seeds (albuminated) provides information
resence microscopy, and image analysis
on both reduction (meiosis) and fertilization
(Naumova et al. 1993; Sherwood 1995; Caceres
events. Ratios between endosperm and
et a\. 1999).
embryos and betwe en female and male
contributions to the endosperm in apomicts 2. Ovule regenerated plants. In tetraploid
often differs from those in sexual plants except accessions of Allium luberosum, Kojima and
fo r the Pan icum-type aposporous Kawaguchi (1989) reported a high frequency
development (Figure. 9.1). For many other of tetraploid regenerated plants from
apomictic pathways, these ratios differ. For unpollinated cultured ovules, suggesting
example, endosperms found in tetraploid apomixis expression. This indicator could be
diplosporous apomicts are higher than in their applied in screening because, in similar
sexual tetraploid counterparts for identical culture media, sexual plants would generate
pollen donors (i.e.. 10x versus 6x if the pollen few (poly)haploids, whereas apomeiotic
is 2x ); endosperm /embryo ratio for ovules would grow mostly into plantlets with
autonomous apomixis is 2:1 and 5:2 [(4x + 4x) the same number of chromosomes as the
+ 2x / 4x + OJ for tetraploid pseudogamous mother plant.
apomicts (tetraploid pollen donor) .
3. Analysis of progeny plants. Progeny tests
Fertilization by unreduced pollen (Chao 1980;
mu st clearly identify either hybrid offspring
Huff and Bara 1993)and endopolyplo idization,
(n + 0 type s are generally poorly represented)
which sometimes occu rs during endosperm
or seed production in absence of pollination
development, is also possible and may further
when pseudo gamous apom ixis or
compli cate analyses. However, endosperm
autonomous apomixis, respectively, are
ploidy level (s) may suggest apomicti c
suspected. Hybrids can be identified using (i)
reproduction or allow the quantification of
morphological descriptors, (ii) cytological
facultative apomixis. Nevertheless, it cannot
data, and / or (hi) marker analyses, if the or igin
reveal the prec ise nature of the apomictic
of the progeny is appropriate.
mechanisms involved .
Remarks on progeny size. The use of
Ploidy level in fertilized ovaries or immature
progenies from controlled crosses is
seed s cannot easily be determined using
recommended . Male parents bea ring
classical chromosome counting methods, but
dis criminating traits (dominant traits,
flow cytometry now permits rapid
different chromosome numbers, etc.) should
measurement of DNA content in a variety of
be chosen when available, limit ing possible
plant tissues, including single embryos, young
confusions between selfed and hyb rid
endosperms, or seeds (Galbraith et al. 1983;
progenies . However, open pollinated
Kowles et a!. 1990; Hignight et al. 1991) .
progenies can be used when mother plants
Analyses in numerous apomictic species have
are s u ffi c ien tly heterozygous to detect
5«.... P.......... w..tiIy . . ClooootIy ....1s 129
segregation after selfing and when there is distribution. Curves shown in Figure 9.4
significant diversity in the surrounding field clearly indicate that, up to n = 100, the
collection, as is the case for most apomictic species, information obtained is poorly significant
which are generally polyploid, polymorphic, and regardless of the value of a. Finally, to
highly heterozygous. obtain good estimations of aberrant rates
(i.e., less than 10% confidence limits), it
Identifying or quantifying apomixis does not
appears that a high number of
require the same number of progeny. To detect
individuals is required.
apomixis, a relatively small number of progeny (15-
25)can be analyzed. Aberrant rates typically are a:n Chromosome number determination
ratios with 'n' the progeny size and 'a' the number within progenies. The sexual or asexual
of aberrants observed in the progeny. Statistically, origin of offspring is not detectable from
such samplings are binomial; 'p' (aberrant rate) is crosses made at the same level of ploidy,
the ratio to be estimated for a given value of n but 2n + nand n + 0 off-types are easily
(progeny size) on the basis of an observed value for detected even at the seedling stage.
a (number of aberrants detected within the Interploidy-level crosses could be used
progeny). Confidence limits for p in a binomial to detect all classes of offspring, but
sampling are given in Figure 9.4 for various values information can be biased by
ofn (a =0.025). Note that for n>30, confidence limits disturbances caused by unstable ploidy
can be estimated using formulas for the normal levels or ploidy barriers. Chromosome
counts can be made from root tips,
microspores, or any somatic tissue using
flow cytometry (Hignight et al. 1991).
Markers for hybrid detection. Traits under Isozymes or molecular markers can be used
simple genetic control are ideal for progeny to assess variation in progenies (fingerprinting
testing by crossing recessive maternal analyses; Nybom 1996). Finding good
genotypes with homozygous dominant testers polymorphic isozyme systems, RFLP probes,
(Hanna et al. 1970; Bashaw and Hanna 1990). or primers for peR as candidates for
Models for estimating levels of apomixis by fingerprint experiments is not a major obstacle.
following marker segregation have been Although genetic analysis is still hindered by
developed (Marshall and Brown 1974), polyploidy, any variation in isozyme or DNA
however, recombination can occur without patterns indicates off-type production,
fertilization, and the presence of dominant provided that somatic recombination does not
traits in progeny tells nothing about the origin occur frequently in the material under study.
of the off-types (n + n or 2n + n) in the absence Esterase and peroxydase were the first systems
of cytological data. Moreover, identification of used to isolate sexual plants from Panicum
such "ideal" markers in apomictic species or mnximum (Smith 1972). Apomixis expression
agamic complexes is not necessarily easy, was also confirmed or quantified using
because traits in polyploid apomicts are isozymes in Taraxacum (Ford and Richard
difficult to analyze genetically. 1985), Arabis holboellii (Roy and Rieseberg
1989), Allium tuberosum (Kojima et al. 1991b),
Morphological descriptors are the easiest
Poa pratensis (Wu et al. 1984; Barcaccia et al.
means for conducting progeny tests. If the
1994), Tripsacum spp. (Leblanc 1995),and Mnlus
tester (pollen donor) differs significantly from
sp. (Ur-Rahman et al. 1997).
the progeny-tested plant, hybrids will vary
sufficiently from the maternal type to allow Mazzucato et al. (1995) showed a slightly
detection. In the case of selfing, because higher capacity of RAPD markers in
apornicts are generally highly heterozygous, discriminating off-types in progenies from the
offspring arising through sexuality will vary same species, when compared with three
sufficiently from the mother plant to be scored polymorphic isozyme systems or with analysis
as off-types. In most species, (poly)haploids of traditional morphological traits. Although
are easily detected because of their particular still seldom used, molecular markers have
phenotypes and the low vigor they exhibit been successfully used for progeny
(Asker and [erling 1992). However, when fingerprinting (e.g., Poa pratensis: Huff and
using morphological descriptors, it is often not Bara 1993; Barcaccia et al. 1997; Paspalum
possible to distinguish between sexuality notatllm: Ortiz et al. 1997).
(n-m) and genomic accumulation (2n + n). But
when morphological and cytological Choosing Suitable Procedures
(chromosome number) data are combined, the Analyses at the Plant Level versus
identification of all classes is theoretically Progeny Tests
possible. Analysis of seedlings has the major 1. Nature of the information obtained.
advantages of timeliness and saving space, but Apomixis results from aporneiotic processes
the most informative descriptors for screening (apospory or diplospory) that produce
purposes are usually expressed at maturity. unreduced ESs, and parthenogenetic embryo
There are few reports of successful progeny development from unreduced eggs. Although
testing for morphology on seedlings after nonreduction and parthenogenesis are
interspecific crosses (Williamson 1981). thought to be closely linked in apomicts,
observations and/or analyses of the plant itself
obviously provides insights only about apospore as measured using cytoembryology
apomeiotic or meiotic events, not about the (Clausen et al. 1947;Kojima and Nagato 1992b).
complete process of apomixis. Data on the next This was confirmed by Savidan (1982a)in one
generation (progeny test) must be collected to Panicum maximum accession: sexual potential
study fertilization and parthenogenesis events was estimated using a clearing procedure at
as well as the degree of apomixis. The choice 22.5%,but only 3% of the open pollinated adult
of the level of analysis (apomeiosis / progeny, were off-types, Elimination of hybrid
parthenogenesis / apomixis) depends on the offspring occurred at germination (-7%)or after
objectives of the research, i.e., whether one transferring plants to the field (-12.5%), because
wishes to determine only cytological of their inbred nature (resulting from selfing
processes, study parthenogenesis, or or hydridization with genetically close
investigate apomixis in its entirety. genotypes in the collection). On the other hand,
after self- or sib-pollination, the lack of
2. Comparing results. Limited information is
heterozygous loci in segregation may cause an
available on diplosporous development, but
overestimation of apomixis, with progeny tests
cytological analyses of parent plants compared
showing the presence of "apparent apomixis"
with progeny tests generally show good
(Bayer et al. 1990).
agreement between apomeiosis and apomixis
screenings in Eragrostis curuula (Voigt and Screening Procedures:
Burson 1981), Allium tuberosum (Kojima and Advantages and Constraints
Nagato 1992b), and Tripsacum spp. (Leblanc Until recently, screening tools for mode of
1995).By contrast, the situation in aposporous reproduction were limited to easy-but-late
species appears more complex: morphological progeny tests or skill-
cytoembryological analyses generally demanding and time-consuming cytological
revealed higher sexual potential than did sectioning methods (see Table 9.3). During the
morphology-based progeny tests in PaniCllm past 15 years, new tools in molecular and cell
maximum (Savidan 1982b), Poa pratensis biology have made screening for mode of
(Nygren 1951), and Bothriochloa-Dichanthium reproduction more efficient, rapid, and reliable.
(Harlan et al. 1964). The same tendency was These techniques include ovary progeny
also observed by Mazzucato et al. (1996) in Poa testing, flow cytometry for determining ploid y
praiensis, when auxin tests and field data were level, auxin test, and molecular markers that
compared. However, using progeny tests on cosegregate with reproductive mode. The
more than 100 Brachiaria F1s, Miles and do major disadvantage of the new methods is their
Valle (1991) classified ten plants that were expense. In addition, though the methods seem
highly facultative apomicts as sexual, to agree wi th cytological and/ or field
according to cytoembryological tests. Sexual observations, additional data are needed to
potential in aposporous tropical grasses has confirm their reliability.
generally been scored according to the 1. Apomixis identification and
formation of 8-nucleate ESs that may develop characterization. As mentioned, apomixis may
concomitantly with several apomeiotic (4- be detected in various ways, but
nucleate) ESs. The competition among ESs- cytoembryological observations are ultimately
more favorable to apomeiotics (Savidan needed to confirm the origin of the ES and to
1982a}--and the possible weakness of certain determine the type of apomixis. Clearing
hybrids that are eliminated early, may explain techniques are now quick and easy but require
the overestimation of sexuality in facultative
the use of phase-contrast or differential
interference contrast optics, both entailing However, these tests do not require much
considerable expense. Stain clearing equipment or technical skill, and can thus be
techniques that allow observation of ovule managed everywhere. Their main drawback
details under traditional optics are less is that they produce frequent errors because
expensive. Molecular markers that cosegregate facultative apomixis occurs more often than
with apomixis, which enable analysis at earlier previously thought. Moreover, progeny with
growth stages than cytoembryology, require sexual origin may resemble the mother plant
the development of special plant materials and in morphology, leading to misclassification
protocols, and the cost of associated supplies and to an overestimation of the degree of
is often beyond the means of many research apomixis. The existence of this gray area in
groups. Moreover, they may not be used with progeny plant classification was reported by
materials that differ in origin from the Williamson (1976), after extensive progeny
materials used to identify the markers, testing in Poa sp. This makes morphological
especially in the case of the highly cross- progeny tests unreliable when apomixis is
specific RAPDs (Williams et al. 1993). highly facultative, but more efficient as
Morphological progeny tests are time- and apomixis expression increases. Early progeny
space-consuming because good descriptors tests using isozymic or molecular markers can
are usually expressed in adult plants and a be cond ucted for apomixis detection on 15-25
minimum of 15 to 25 offspring are needed. offspring. Only a few isozyme systems are
Table 9.3Advantages and disadvantages of important procedures for the investigation of modes of
reproduction at the plant and progeny levels. • See Ragot and Hoisington (1993) for RFLP and RAPD costs.
Plant level analyses Progeny tests
Procedures CytoembrYIl- Molerular markers Auxin tem Chr. counting in Aduh Plants Morphology Fingerprinling
Iogy (dearing Cll-segregating ovaries or seeds CIv. counting'
procedures) with apomixis'
Information Apomixis type Depends on the Indica1ion of Indica1ion of Off-types of Apomixis ident~ica1ion and
expened determination na1ure of the apomixis apomixis 2n+n and quanl~ica1ion; off-types nature ~
and sexual marker(s) expression; expression; n+O na1ure combined with chromosome
potential ident~ied (to estima1ion of eslima1ion of detection. counting.
estima1ion. dote linkage the degree of the degree
with apomeiosisl. porthenogenesis of apomixis.
Plant 15to 100 Already deter- 100 flowers. 50 to 100 Apomixis idenl~ication: 151025 offspring.
materials 1I0wers, mined mo1eriols ovaries/seeds. Apomixis quant~ica1ion: olleostl 00 offspring.
required depending on in segregation
the objedives. for marker
iden'~ica1ion.
Advantages Easy and Analyses can be Easy and quick Eosyand Easy ifflow Easy Analyses on
quick to performed 10 perform quick to cytomelry young offsprings
perform after eor~. after flowering perform after (embryo, possible.
1I0wering. pollina1ion. endosperm,
plan~ets).
Constraints Expensive Preliminary work The auxin test Expensive nme consu· nme and spoce consumming.
equipment to determine hes been main~ equipment far ming (classical Morpholagicallests: unreliable if
for maleriols. Use used to dole in now cytometry. methocls) or apomixis is high~ facuha1ive.
micrOlcopy. of the marke~ cool-season expensive ~
ocrOls occessions grosses. now cytornetry
of different isused.
origins? Expensive.
5creeoiot I'roceoIores to WooIiIy . . Clooolily . . . .is 133
required to indicate apomixis and determine additional species should begin with taxa
the nature of the hybrids detected. RFLPs and presenting these traits. The very first screening
RAPDs can also be used in the same way, but can be based on the expression of the already
at greater expense. mentioned "indicators of apomixis," while
more discriminative procedures may be
2. Degree of apomixis expression. Many
applied to promising specimens. For
offspring are needed to obtain a good estimate
germplasm evaluation, a representative sample
of the degree of apomixis. Both auxin tests and
of the collection must be chosen on the basis of
flow cytometric analyses of pollinated ovaries
morphological and cytological data, and traits
or seeds provide good estimates of sexual
of agronomic value such as disease resistance.
potential, though distinguishing 2n + 0 from
Chromosome number, repro-ductive
n + n offspring might be difficult in certain
development, and degree of apomixis are the
cases. In contrast, systematic chromosome
primary [actors for which basic data must be
counting within progenies is useful for
collected to develop strategies for further
detecting 2n + nand n + 0 off-types, but it does
research. Genetic studies also may be
not separate 2n + 0 from n + n offspring, and
attempted to genetically dissect apomictic
without flow cytometry it becomes
mechanisms (number of genes involved and
tremendously time consuming. Progeny tests
their effects). Following this preliminary work,
combining cytology and marker analyses
appropriate tools for larger-scale screening
represent the best option for identifying the
should be developed or chosen according to
different classes of offspring within apomictic
the apomixis characteristics of the collection
progenies. To limit cytology work (when flow
(e.g.,callose patterns for diplospory, ESclearing
cytometry is not available), markers can be
for apospory of the Panicum-type, etc.).
applied first to separate maternal offspring
from (poly)haploids or hybrids. The origin of Sexual parents involved in crosses for apomixis
the latter may be determined according to the inheritance studies must be carefully chosen
patterns they produce (i.e., 2n + n off-types using cytoembryology. Highly facultative
must carry all bands from the mother plant, apomicts are easily misclassified as sexuals
plus extra bands from the pollen), and then using progeny tests. This causes distortions of
cytologically confirmed. segregation ratios for mode of reproduction
among progeny. In the same way, looking for
Choosing a Procedure
differences between sexual and apomictic
There are four main areas of apomixis research,
development at the molecular level requires the
each with distinct constraints and objectives:
(i) the search for apomixis or elements of
analysis of genotypes that are well
characterized for mode of reproduction. This
apomixis in new taxa, coupled with genetic
may allow the development of near isogenic
studies in wild populations, (ii) germplasm
characterization of apomictic species, lines, an important step in identifying the
gene(s) controlling apomixis.
(iil) genetic and biological studies for further
manipulation of apomixis, and (iv) breeding Before apomixis can be transferred into crops
of apomicts and introduction of apomixis into or used in breeding programs, researchers need
sexual crops. procedures to identify apomictic genotypes
Since gametophytic apomixis is formidably (see do Valle and Miles, Chap. 10; Savidan,
limited to perennial, polyploid, and Chap. 11) and to quantify apomixis in
outcrossing species, the search for apomixis in genotypes selected for varietal release. Progeny
134 om.. LoW- .,j . . . . . Man.....
testing in such programs may help identify in most species apomixis and sexuality do not
apomixis, because offspring are necessarily express at the same ploidy level, estimating
produced as part of breeding schemes, but an chromosome number within progenies using
entire plant cycle must pass before data are flow cytometry allows easy identification of
obtained (a serious drawback in the case of apomictic genotypes in early backcross
annual plants). Notwithstanding, in some generations, but becomes less effective when
cases-i-especially when low female fertility is chromosome numbers close to that of the
affecting the plants (e.g., interspecific or recurrent parent are recovered. Appropriate
intergeneric hybrids)-this may be the best cytological procedures or marker-assisted
way to test for mode of reproduction. Because selection may also be used to identify apomixis.
References Blakey, C. A., C. LDewold, and S. LGoldman. Chao, C. Y. 1980. Automanous developmenl of
1997. CD-segregation of DNA marke~ with embryo in Pospalum conjulotum 8erg. BoI.
Asker, S., 1980. Gametophytic apomixis; Tripsocum fertitlty. Moydico 42: 363--69. No/. 133: 2\~22.
elements and genetic regulation. Hereditas Bradley, M.V. 1948. An acetocarmine squash Chaudhury, A.M., LMing, C. Miller, S. Craig,
93: 277-93. tedmique for mature embryosocs. Slain lS. Dennis, and J.W. Peacock. 1997.
Asker, S., and LJerling. 1992. Apomixis in Technology 23: 29-40. Fertilization·independent seed
plants. 8oco Raton, Florida: CRC Press. Brautigam, S., and E. Brautigam. 1996. developmenl in Arobidopsis tho/hJoo. Proc.
8orcocoa, G., A. Mozzucato, M. Penani, and M. Determination of the plaidy level in the Noli. Acad. 5<;. (USA) 94: 4223-28.
Fokinelli. 1994. ComporilOll between genus Hierocium subgenus Pilosella (Hill) C1ausen, J., 0.0. Keck, and W.M. Hiesey. 1947.
isozyme and RAPD ana/yses to \(reen S.F. Gray by flow cyIometric DNA analysis. Partial apomixis: an evolutionary
aberrant plants in Pea pralensis L FoIhJ GeobolDnKO elI'hytotoxonomico 31 : labyrinth. (omegie Inst. Washing/on Year
progenies. Apomixis NewsJener 7: 29-30. 31~21. BooK46 (1946/47): 101-03.
8ol(acoa, G., A. Mozzucata, M. FolOnelli, and F. Brown, W.V., and H.P. Emery. 1958. Apomixis in Crone, U., and J.G. Carman. 1987.
Veronesi. 1996. Callose localization in cell the Gramineoe: Ponicoideae. Amer. 1. BOI. Me<hanisms of apomixis iQ Bymus
walk during meiotic and apomeio1ic 45: 253--63. rectisetus from Eostern Australia and New
megasporogenesis in diploid ollalfa BurlOn, B.L, P.W. Voigt, RA Sherman, and C.L Zealond. Amer. J. Bol. 74: 477-96.
IMedKOgo spJ (otyologhJ 49: 4~ 56. Dewald. 1990. Apomixis and sexuality in Dorlington, C.D., and LF. Lo Cour. 1966. The
80rcacoa, G., A. Mozzucata, A. Belordinelli, M. Eastern GamagrolS. Clop xi. 30: 86-39. Handling ofChromorornes. 4th ed.
Pezzoni, S. Looeni, and M. Fokinelli. 1997. Burton, G. W., J.c. Millat, and W.G. MOIlIOIl. London: Alien and Unwin.
Inherilance of parental genomes in 1973. Breeding procedures for PonKum den Nijs, A.P.M., and G.E. van Dijk. 1993.
progenies 01 Pea prolemis LFrom sexual maximum suggested by ~ant variability Apomixis. In M.D. Hayward, N.O.
and apomictic genatypes asllSleSsed by and mode of reproduction. Clop xi. 13: Bosemark, and I. Romogoso (em.), Plant
RAPD marXers and flow cyIometry. Thear. 717·20. Breeding: Principles and Prospedl.
Appl. Genel. 95: 516-24. Caceres, M.l, F. Pupiln, C.L Quarin, and S. London: Chapman and Hall. pp. 229-45.
Barcacoa, G., A. Mozzucato, E. A1bertini, J. Arcioni. 1999. Feu~en·DNA densitometry Duich, J.M., and H.B. Musser. 1959. The extent
Zethaf, A. Gerats, M. Pezzolti, and M. of embryo IDCS perm~s discrimination of aberrants produced by 'Merion'
Fokinelli. 1998.lnher~ance of between sexual and apomictic plants in Kentudly bluegralS, Pea prolensis L, os
parthenogenesis in Pea prolensis L: auxin Pospalum simplex. Euphytico 11 0: 161- determined by Ii~t and sec:ond generation
lest and AFLP linkage ana/yses suppart 67. progeny test. Agron. J 51: 421-24.
monogenic control. Theor. Appl. Genel. 97: Carman, J.G. 1997. Asynchronous expression of 8ktrand, N.C., and D.A. Levin. 1982. Genotypic
7~2
duptKale genes in angiosperms may cause dive~ity in Oenolhero laciniato
8oshow, l C. 1980. Apomixis and its im~ications apomixis, bispory, tetrospory, and (Onagraceoe), 0 permanent translocation
in crop improvement. In W.R. Fehr, and H.H. polyembryony. Bioi. 1. Linn. Soc. 61: 51- heterozygote. Evolution 36: 63-69.
Hadley lem.), Hybridization ofClop Plants. 94. Ford, H., and AJ. Richard. 1985. Isozyme
Madison, WlIConsin: American Society of - - . 2000. New ~ndings support new variation within and between Taraxacum
Agronomy. pp. 4~3. modek for the origins and stabilization of agomospe<ies in 0 single locality. Heredity
Bashow, l I, and W.W. Hanna. 1990. Apomictic agomic complexes. XYhh Intemotional 55: 289-91.
reproduction. In G.P. Chapmon (ed.], Congress on Sexual Plant Reproduction, Gadella, lWJ. 1983. Some nates on the
Reproductive Versatility in the Grosses. April 1·5, 8onff, Canada. determination of the mode of reproduction
Cambridge, U.K.: Cambridge Univernty Carman, J.G., tF. Crone, and O. Riera·[jzarazu. in higher ~ants. Proc. Koninklijke Nederl.
Press. pp. 101·29. 1991. Comparative histology of cell WIllis AJcod. van WelellScooppen, \eI. C. 86:
8ayer, R. J., K. R~lond, and B.G. Purdy. 1990. during meiotic and apomeiatic \5~6.
Evidence of partial apomixis in Antenoooo megasporogenesis in two hexaploid Galbra~h D.W., K.R. Harkins, J.M. Maddox, HA
medhJ (Asteraceoe, Inuleoel deteded by the Australasian Bymusspe<ies. Clop 56. 31: Ayres, D.P. Shorma, D.P. Firoozaboby, and
segregation of genetic marXers. Amer. J. \527-32. l Firoozaboby. 1983: Arapid flow
s« 77: 1078-83. cyIometric analysis of cell cycle in intac1
planllissues. xience220: 1049-51.
SueeoioI Pro<etIom to .... tiIy... a-tily ApooUis 135
Grirroneltl D., M. Hernondez, l Peroffi, ond Y. Kindiger, B., D. Boi, and V. Sokolov.I996. l.eblonc, 0., and Y. Sovidan. 1994. Toning af
Sovidon. 1997b. DOIOge effem in the Assignment af agerJe(s) conferring megasporogenesis in Tripsocum species
endosperm of di~osporous opomidic aponixis in Tripsocum taachromosome (Pooceoel os related to the control al
Tripsocum(POIKeoe). Sex. Plant Reprod.IO: arm: cy1a1ogieal and malecular evidence. aponixis and sexuality. Pal. Bot. Stud. B:
279-82. Genome 39: 1133--41. 75-81.
Grirronelli D., O. l.eblonc, l Espinoso, E. Peroni, KOIirro, A., l Hinota, and S. Nodo. 1991 b. An Marshall, R.D., and A.H.D. Brown. 1974.
D. Gonzlilez de l.eOn, ond Y. Sovidon. improvement of squash method for the Estirrotian af the level afapomixis in pIont
19970. Mopping di~asporous oponixis in cy1a1ogical study afferrole meiosis in AHium populotions. Heredity 32: 321-33.
tetroploid TripsOCIJI1[ one gene or severol tuberasum , litMlceoe. (hrorrosome Mortinez, EJ., F. Espinoza, and Ll, Quorin. 1994.
genes? Heredity 80: 33--39. Information Service 50: 5-7. Bill plogeny (2n+n) from opornicIic
Grossniklaus, U., JP. V"JeIle.{olzodo, M.A. Kojirro, A., and 1 Kowoguchi.1989. Apomidic Paspalum natotum obtained through eorly
Hoeppner, ond W.B. Gogliono. 1998. noture afChinese chive (AHium tuberasum pollination. 1. Hered. 85: 295-97.
Moternol control of emlllyogenesis by Ronl.) deleoed in unpollinoted ovule Motzk,F. 1991.A novel approach 10 differentiate
MfOEA, 0 po~comb group gerJe in cuhUle. Jap. J.Breed. 39: 449-56. embryos in the absence of endosperm. Sex.
Arobidopsis. Science 280: 446--SO. Kajirro, A., Y. Nogota, and l Hinota. 1991 a. Plant Reprod. 4: 88-94.
Gupta, M.G., S. Gupto, 8.V. Bhot, V. Bhot, ond SI Degree afapomixis in Chinese Chive (Allium Motzk, F., A. Il\eister, and I.SchOOert. 2000. An
Ahmod. 1998. fslirrotion of frequency of tuberasuml estimated by esterase isozyme efficient screen for reproductive pathways
apomixis by auxin induced porthenocorpy anolysis. Jop. J. Breed. 41: 73-83. using mahKe seeds afITIDIlD(I)ts and dieats.
technique in Oichanthium onoolo1vm Kojirro, A., ond Y. Nogalo. 1992a. PlantJ. 21: 97-108.
(Fonsk.J Slopf. Range ltIanogement and Pseudogomous embryogenesis and the Mozzucato, A., G. Borcocda, M. Pezzalli, and M.
Agroforestryl9: 149-53. degree afporthenogenesis in Allium Folcinelli. 1995. 8iodlemical and rnoIecub
Gustokson, A. 1947. Apomixis in higher plo,m. tuberasum. Sexool Plant Reproduction 5: markers far investigoling the mode af
Port Ill.Biotype and IflIlcles forrrolion. 79-85. reproduction in the Iocuhative apornicl POD
lunds Unversitets Arsskrilt Nova Seriel2. - - . 1992b. Dip\osporous embryo-sac protensis LSex. Plant Reprod. 8: 133--38.
43: 183--370. farrrotion ond the degree of dip\ospory in Mozzucata, A., A.P.M. den Nijs, and M. Fokinel5.
Gustine D.L, RI Sherm, and D.R. Huff. AHium tuberasum. SexUl1I Plant 1996. Eslirrotian of parthenogenesis
1997. Apospory·tmked rrorker1 in Reprcduction 5: 72-78. frequency in Kentucky bkJegross using
buffelgross. Crop 56.37: 947-51. KoItunow, A.M. 1993. Apomixis: EJOOryo 100 auxin-induced porthenocorpic seeds. Crop
Hanno, W.W.,and le. Boshow.1987.Apomixis: and emlllyas formed wilhout meiosis of 56.36: 9-16.
in identilicotion and use in plont breeding. fertifization in ovules. The Plant Cell 5: Mozzucato, A., M. Wogenvoort, and HM.den
Crop 56. 27: 1036--39. 1425-37. Nijs. 1994. Row cy10metrie analyses la
Hanno, W.W., K.F. Schertz, ond le. Boshow. KoItunow, A.M., R.A. Bicknell, and A.M. eslimate the mode af reproduction of Poo
1970. Apospory in Sorghum bicolor (Ll Choudhury. 1995b. Apomixis: Male<ulor protensis L Apomixis Newsletter 7: 22-24.
Moench. Science 170: 338-39. strategiel for the generation of genetical~ hlendel, Gregor.1870. 'Ueber eDge aus
Horlon, J.R., M.H. Brooks, DS. Borgoonkor, ond identical seeds wilhaul fertilization. Plant kunstlichen Belruchtung gewomenen
lMJ.de Wet. 1964. NahKe ond I'/rysiol. 108: 1345-52. Hieraoum-Bostorde: Verhordoogen des
inheritance of apomixis inBothriochloo and KoItunow, A.M., l Sohys, N. Nila, and S. noturfaBChenden Vereines, Abhondlungen,
Oic/rorrthium. Bot. Goz. 125: 41-46. McCIure. 19950. Anther, avule, seed, and Brunn. 8: 26-31.
Herr lM. 1971.A new deoring-squash technique nucellor embrya development in Citrus Miles, lW., and e.B. do VaRe. 1991. Assessment
for the study of ovule development in sinemis cv. Valencia. tsm. J. Bat. 73: 1567- afreproductive behovior afinterspecific
angiospenns. Amer. J. Bot. 58: 785-90. 82. Brochiorio hybrids. Apomixis Newslener 3:
Hignight, lW., le. Boshow, and MA. Hussey. KowIes, R.V., F. Srienc, and R.L Phillips. 1990. 9-10.
1991. Cylological and morphological Endoredupficotion afnuclear DNA in the Miles, lW., F. Pedraza, N. Palocios, and 1 Tohme.
diversity of native opomiclic buffelgross, developing rroize endosperm. Oev. Genetics 1994. MoIecuIor marker far the apomixis
Pennisetum dliore (L)link. Bot. Goz. 152: 11: 125-132. gerJe in Brochiario. In Proc. 2nd Int. Conf.
214-18. l.eblonc, O. 1995. Modes de reproduction dons le Plant Genome, Jaooory 24-27, Son Diego,
Hillory, B.B. 1940. Uses of the Feulgen reaction complexe agomique des Tripsocoor, et Californio.
in cy101ogy. 11. New techniques ond special ano~ d'hybrides rrois·Tripsocum en vue Mogie, M. 1988. Amodel fOf the evolution and
applications. Bot.Goz. 102: 225-35. du transfert de ropornixie chez le rrois. PhD control afgenerolive apomixis. Bioi. J. oon.
Huff, D.R., ond lM. Bora. 1993. Determining thesis, Paris: INA·PG, UniveOOy of Paris VI, Sac. 37: 127-53.
gerJeoo origins afaberrant progeny from UniveOOy afPoris XI. Naurrovo, 1.1992. Apomixis inAngiosperms.
facuhative apomictic Kentucky bluegross l.eblanc 0., D. Grirronelfi, D. Gonzlilez de leOi1, NIKe/lor and Integumentrlry fniJryany. Boco
using acombination afflow cy10metry and and Y. Sovidon. 1995b. Delection af rhe Raton, AOfido: (RC P1ess.
silver·stained RAPD rrorker1. Theor. Appl. aponidic mode of reproduction in maize- Naurrova, 1,A.P.M. den Nijs, and M.TJA.
Genet. 87: 201~8. Tripsocum hyb!ids using rroize RFlP WiUemse.I993. Quonlilative anolysis af
Jahonsen, D.A. 1940. PIont Miuotemnique. New rrorkers. Theor. Appl. Genet. 90:1198- apasporous parthenogenesis in Poo protensis
York: McGrow·Hi11 Book Co. 1203. genotypes. Acto Bat. Neerl. 42: 299-313.
Jongedijk, E. 1987. Aquick enzyme squash l.eblonc, 0., M.O. Peel, lG.Corrron, and Y. Naurrovo,1 N., Hayward, M. D., and
technique for detailed studies on ferrole Sovidon. 19950. Megasporogenesis and Wogenvaort, M. 1999. Apomixis and
meiosis in SoIonum. Stain Temn. 62: 135- megagometogenesis in several Tripsocum sexuality in diploid and tetraploid accessions
41. speciel (Pooceoel. Amer. J. Bot. 82: 57-63. of Barochiorio decumbes. Sex. Plant Reprod.
12: 43--52.
136 0Iiri0< ltWooc ... Aodna 1Iau_
Nogler, G. 1984. Gometophytic apomixis. In 8. Pmino 5., lPAOrtiz, O. Leblanc, C. mns, c. Strosburger, E., and W. HalJause. 1900.
III Johri (ed.l, fmJKyoIogy ofAngiosperrm. do Vale, and IllD. Hayward. 1997. Hanrl100k ofPractical BotOllY. New York:
Berlin: Springer-Verlog. pp. m-518 ldentifimlion of 0maize linkage group MacMillon.
Noyes, R.D., and LH. Rieseberg. 2000. Two related to apomixis in Brodiario. Theor. Tavoletti, 5., A. Morioni,lIld EVeronesi. 1991.
independentlaci control oganmpermy AppI. Genet. 94: 439-44. Cytological analysis of mooo- and
(opamixis) in the triploid flowering plant RogoI,lll, and DA Haisinglon. 1993 Malerulor miaasporogenesis of 0 d~ alfmfo done
ErigerOll OflfllM. Genetics 155: 379--90. markers for plant breeding: comparisons of producing male and female 2n gametes.
Nybam, H. 1996. DNA fingerprinting -0 useful RFlP and RAPO genolyping costs. Theor. ClopSci. 31: 1258-63.
tool in the laxonamy of apomiclic plant AppI. Genet. 86: 97>-84. Tucker,IllR., Poech, NA, and A.M. Kahunow.
groups. Folia GeobotOllica et RoRllIu, K.S., POijldUs, A. Pereira, G.c. 2000. DyI1IIllics of bela· I,J.gluconose
P/rytotaxonomKa 31: 29>-304. Angemenl, U vun Lookeren Call1lDgne, eXJlfession c1urilg embryo SO( formalion in
Hygren, A. 1951. Embryology af Poa. CDmegie and JJ.M. Dons. 1997. EMS and transposan aparridic Hierocium. Proceedings ofthe
Institute ofWashing/OIl Year Book 50 rooIagenesis for the isolation of apomictic XVIth Intemotional (OIlgress 011 Sexual
(1950/511: 113-15. rooIants in plants. In S.1ll Join, 0.5. 8rar, Plant leprodudion, April 1-5, Banll,
Ohod, N., LMorgassian, Y.c. Hsu, S.c. WiUioms, and 8.5. Ahloowulia (eds.l, Somadonal Canada.
PRepenl. and R.LFischer. 1996. ArooIDIion I'Driotion and Induced Mutations in Clop Ur·Rohmon, H., DJ. Jomes, A.M. Hadanau, and
that allows endosperm development without Improvement. Dardrecht: Kluwers Academic PD.S. Caligari. 1997. The use of RAPO for
ferti~zation. PrO(. NotI. AlaJ. Sci. (USA) 93: Publishers. pp. 379-400. verifying the apomiclic status of seedlings of
5319-24. Richords, A.J. 1997. Plant Brmng Syrtems. Mo/us species. Theor. Appl. Genet. 95:
Ortiz lPA, S.c. Pessino,O.l.eblanc,IllD. 2nd edition. London: <hapmon and Hall. 1080-33.
Hayward, and C.L Quorin. 1997. Generic Roche, D., C. PeiSheng, C. lhenBang, W.W. vun Di~, P, and 1 vun Domme. 1999. Aponixis
fingerprin~ng for delermining the mode of Honno, D.L Gusline, R.T. Sherwood, and P. technology and the paradox of sex. Trends
reprodUdion in Pospolum nototum, a Ozias-Akins. 1999. An apospory·specilic PI. Sci. 5: 81-84.
subtropical apomiclic forage grass. Theor.
AppI. Genet. 95: 850-56.
Ozias·Akins, P, E.L Lubbers, W.W. Hanno, and
bullelgross (CenmflJS .is
genonlC region is conserved between
L) and
Perrrisetwn squallXllotum Fresen. Plant 1.
YJeUe-(alzado lP, CF. Crone, and D.M. Srelly.
1996. Apomixis: The asexual revolution.
Science 274: 1322-23.
lW. Ml:Nay. 1993. Transmission aflhe 19: 203-08. Voigt, P.W., and E.c. Bashaw. 1972. Apomixis aos
apomictic mode of reprodUdion in Rodkiewia, 8.1970. CaUase in reU wuUs during sexuatJly in Erogr05tis curvu/o. Clop Sci. 12:
Pennisetum: Co-inherilDIICe of the trail and megasporogenesis in angiasperms. Planta 843-47.
rnole<uIor markers. Theor. AppI. Genet. 85: 93: 39-47. Voigt, PW., and 8.L 8urson. 1981. 8reeding of
632-38. Ray, BA, and LH. Rieseberg. 1989. Evidence apomiclic Erogr05tis CUfYlJIo. PrO(. 14th Int.
Ozias·Akins, P, D. Roche, and W.W. Honno. for opanixis in Arubis. 1. Hered. 80: 506- Gross/. (angr., l.exinglon, Kentucky. pp.
1998. TlQht dustering and henizygosily of 08. 160--63.
aparrixis-~nked moIeculDl markers in Russel, 5.0. 1978. FiJe structure of Warmke, H.E. 1954. Apomixis inPonicvm
Pennisetum squamulotum imp~ genetic megagamelophyte development in Zea maxmum. Amer. 1. Bot. 41: >-1 1.
control of opaspary by 0divergent locus trrlYi CDn.1. Bot. 57: 1093-11 ID. WiHioms, lG.K., M.K. Honafey, and J.A. Ralolski.
that may have no oIle1ic form insexual Soran, 5., and J.MJ. de Wet. 1966. Arapid 1993. Genetic analysis using random
genotypes. PrO(. NotI. AlaJ. Sci. (USA) 95: squash method for gross embryology. PrO(. amp~fied polymorphic DNA markers.
5127-32. Oklahoma Acad. Sci. 46: 17-19. ItIethods in Enzyma/ogy 218: 704-40.
PeDlson, E.S., and H. O. Hortley. 1958. Sovidan, Y. 19820. NoIIre etheredite de WiUiomson, U. 1976. Problems in the
Biometrika Tables fat Staticicions. Volume I. rapomixie chez Ponicum maximum Jacq. identilicDlion and u1it.sotion afinlersperific
Cambridge U.K.: Cambridge University Travuux etDocuments de rOISTOM hybrids af Poa in0pIont breeding
Press. 153:1-159. progranvne. Ph.D. thesis, University of
Peel, M.D. 1993. Meiocyte callose in aposparic - - . 1982b. EmIlIyalogical anolysis of Edinburgh.
and di~osporic gr~ and inhybrids Iocuhative opanixis in Ponicvm maxillXlm - - . 1981. Variobi~ty inseedling
between bread wheat and E1ymus rectiselvs. JlKq. Clop Sci. 22: 467~9. progenies and the ellect aflight regimes
MS thesis, Utah State University. Sherwood, R.T. 1995. Nudeor DNA amount during seed production on inlersperili<
Peel, 1llD., lG.Carman, and O. l.eblon<. 1997. during sporogenesis and gametogenesis in hybrids of POd New P/ryt%gis187: 78>-
Megasporocyte callose in opamictic sexual and opasparous buffe~rDSS. Sex. 97.
bullelgross, Kentucky bluegross, PIont leprod. 8: 8>-90. WlIlkler, H. J908. Ober Parthenogenesis unci
Pennisetum squamulotum Fresen, TripsDClXII Smith, R.L 1972. Sexual reproduction in Apagomie im PIIonzerveKh. Progr. lei. Bot.
Land weeping Iovegross. Crop Science 37: Ponicum maxinun JlKq. Clop Sci. 12: 2: 293-454.
724-32. 624-27. Wu, L, A.H. Horivandi, and J.A. Davis. 1984.
Pemell. R.I., and K. Roberts. 1990. Sexual Snyder, LA. 1957. Apomixis in Pospolum secans. IdentiIicDlion afKenlucky bluewoss cultivurs
develapment in the pea is presaged by Amer. J. Bot. 44: 318-24. with esterase and phosphoglucomutase
ahered eXJlfessian of arabinogoloctan Stebbins, G.L 1950. Variation and Evolution in isozyme markers. Clop Sci. 24: 763-68.
prolein. Nature 344: 547-49. Plants. New York: Columbia University Young, BA, R.T. Sherwood, and E.c. Bashaw.
Pessina, S.c., c. mns, J.P.A Ortiz,l. Armslead, Press. 1979. Oeared-jlistis and thidt-sectianing
C.8. do Vole, and IllD. Hayward. 1998. A S1e1Iy, D.M., SJ.PeIoquin, R.G. Palmer, and CF. techniques for detecting aposparllUS
genetic map of the apospary-fegian in Crane.1984. Muyer's hemolum-methyt apomixis in grosses. (on) Bot. 57: 1668-
Brochicm hybrids: identilicDlion aftwo solicilDle: Astain-deoring technique for 72.
markers doseIy osscxioled with the troil. observation within whale ovule. Stain
Hereditas I28: 152-58. Techn. 59: 15>-61.
Chapter 10
facultative apomicts are found (Harlan et al. sources of the desired attributes in the existing
1964; Voigt and Bashaw 1972; Bashaw 1980; germplasm. This presupposes intimate
Savidan 1983;Hanna and Bashaw 1987). knowledge of the species of interest, in order
to identify limiting factors not readily
Differencesin ploidy level are common among
overcome by simple selection of superior
sexual and apomictic species of tropical
genotypes from the available germplasm or
grasses (Burton and Forbes 1960; Carnahan
amenable to improved cultural practices. Once
and Hill 1961; Dujardin and Hanna 1983;
a constraint has been identified (e.g., disease
Norrmann et al. 1989).However, in groups in
susceptibility or low forage quality), the
which apomixis is found, diploid accessions
natural germplasm needs to be screened to
are generally obligatory sexual while
identify candidates for hybridization. Ideally,
polyploids display different degrees of
the desired attribute(s) can be found in
apomixis ranging from essentially sexual to
apomictic or cross-compatible sexual
obligate apomicts (de Wet and Harlan 1970;
accessions with a superior agronomic
Quarin and Norrmann 1987; Valle et al. 1989;
background.
Valle 1990;Asker and Jerling 1992). In species
with higher ploidy levels (6x or 7x),such as B. General Structure of a
liumidicola, sexuality may be found at the
Breeding Program
tetraploid level (Valle and Glienke 1991).
A general selection and breeding scheme for
Sexually reproducing genotypes in the tropical apomictic forage species is presented in Figure
forage grasses outcross and are highly 10.1.Note that Brachiaria serves as the example
heterozygous (Bashaw and Funk 1987).Some for the topics under discussion.
degree of self-incompa tibili ty or strong
Brachiaria is native to the tropical savannas of
inbreeding depression is common (Bashaw
Africa (lBPGR 1984), encompassing about 90
and Funk 1987).Rates of self-fertility in sexual
species with wide morphological and
B. ntZiziellsis were not affected by chromosome
phenological differences (Clayton and
doubling and ranged from 7.2 to 8.4%,
Renvoize 1982; Ren voize et al. 1996).
according to Lu tts et al. (1991). When
Apomictic cultivars of some of these species,
hybridization with apomicts has been
possible, resulting progenies are highly
[+
variable owing to segregation in the
heterozygous parents. ~;:~er~~rodumon
I----+~ Genefic markers
Since hybridization and production of fertile Morphological characterization
progeny are more effective when progenitors elc.
are at the same ploidy level, basic studies I
leading to the determination of chromosome +
SEX
+
APO
number should be undertaken early in the
program to enhance the chances of successful
recombination of attributes by conventional
crossing. +
APO
+
SEX 5 APO
I
A third prerequisite for efficient breeding of I
apornicts, as in any plant improvement Release of ..
program, is the establishment of clear, new cultivors
achievable objectives, and the identification of Figure 10.1 Selection and breeetmg scheme for
apomictic forage species.
derived directly from natural germplasm, are al. 1999). Another line of research now being
widely sown because of their excellent pursued is based on identifying molecular
adaptation to infertile acid soils, good forage markers that cosegregate with the resistance
value, and abundant seed production to spittlebugs (CIAT 1998).
(Bogdan 1977; Keller-Grein et a!. 1996).
Screening for nutritive value and consumption
Cultivated pastures cover more than 45
potential is also a laborious, costly, and time-
million hectares of acid savannas in Brazil
consuming endeavor, many times depending
alone (Anuario Estatistico do Brasill995) and
on grazing trials. New procedures are being
cultivars of two Brachiaria species (B.
developed in an attempt to identify promising
decumbens cv. Basilisk and B. brizantha cv.
genotypes early in the breeding program. Such
Marandu) are grown on more than 85%of that
identification entails correlating measures of
area. Genetic uniformity (associated with
physical traits (shearing strength and grinding
inability to generate new genotypes due to
resistance), anatomical characteristics (patterns
apomixis) resulted in a massive failure of cv.
of lignin deposition and cuticle and epidermis
Basilisk in the Brazilian Amazon, stemming
width), and gas production potential with
from its susceptibility to spittlebugs
nutritive value (Hughes et al. 1998, 2000;
(Homoptera: Cercopidae), (Seiffert 1984).
Sabatel et al. 1999).
Objectives
Germplasm Acquisition and Evaluation
In the B. decumbens-B. brizaniha agamic
Plant introduction programs and subsequent
complex, we are seeking apomictic genotypes
agronomic evaluation were required for
adapted to, and persistent on, low-fertility
Brachiaria, as they are for most tropical forage
acid soils, wi th high levels of antibiotic
grasses. Prior to 1984, the small "world
resistance to a range of species and genera of
collection" of Brachiaria accessions numbered
spittlebug (Homoptera: Cercopidae). For
no more than 150 (Keller-Greinet al. 1996),and
continuously grazed pastures, stoloniferous
the desired combination of attributes had not
growth and strong competitive ability are
been identified.
desired. If a new Brachiaria cultivar is used in
a lay farming system in rotation with annual Extensive collection of Brachiaria was
crops, genotypes with less "weedy" undertaken in 1984-85 in East Africa, the center
characteristics may be preferred (Valle, of origin and diversity for the main species of
unpublished). Forage yield and quality are agronomic importance. Nearly 700 new
clearly important attributes of any forage accessions in 24species were collected in a joint
plant cultivar, especially tropical grasses. CIAT/IBPGR (International Board for Plant
Genetic Resources) venture, with the support
Achieving breeding objectives depends on the
of ILCA (International Livestock Center for
ability to reliably measure attributes of interest
Africa) in Ethiopia and Kenya, and national
in large segregating populations. For example,
institutions in East Africa, including
a greenhouse screening methodology for
Zimbabwe, Grassland Research Station
resistance to spittlebug was developed at
Marondera; Burundi, Institute des Sciences
CIATabout a decade ago (Lapointe et al. 1989,
Agronomiques du Burundi (ISABU); Rwanda,
1992), but its utility for breeders was limited
Institute de Sciences Agronomiques du
by its low capacity. More recently, an
Rwanda (ISAR); and Tanzania, Tanzania
improved methodology was implemented to
Livestock Research Organization (TALIRO)
allow faster and more efficient screening for
(CIAT 1986).
spittlebug resistance (CIAT 1998; Cardona et
As accessions were transferred to CIA T- germplasm into groups of morphologically
Colombia and released from quarantine, large similar accessions, regardless of taxonomic
portions of this collection were subsequently classification. This type of study helps
forwarded to Brazil (approximately 400 researchers define closely related accessions
accessions), Costa Rica (approximately 280 within and among groups from which
accessions), and Peru (approximately 260 individual progenitors may be selected for
accessions) for agronomic evaluation (Grof et future crosses. This analysis revealed the
al. 1989a; CIAT 1992). Because large numbers continuous polymorphism that exists among
of accessions were involved, evaluation three species (B. decumbens, B. brizaniha, and
methodology needed to be simple and B. ruziziensisi and clearly separated typical
efficient, such as that proposed by Toledo accessions of B.humidicola, B. dicrfoneura, and
(1982), in order to discard poorly performing B. [ubaia (Figure 10.2). The selection of
materials. Since then, more detailed and accessions for pasture trials was based on an
intensive evaluation has been cond ucted at the association of agronomic traits with
Embrapa Beef Cattle Research Center in morphological characteristics.
Campo Grande, MS, Brazil. Agronomic
Cytology, Reproductive Mode, and
evaluation began with accessions planted in Inheritance of Apomixis
small plots with three replications. A periodic Basicinformation about mode of reproduction
cutting regime was imposed for three years to and cytogenetics of sexually reproducing
estimate overall and seasonal production, accessions was also ascertained from the
regrowth vigor, seed production, and
resistance to spittlebug and diseases (Valle et
Table 10.1 Agronomic evaluation of .rrxlritria
al. 1993a). At one harvest each year, samples
accessions in Brazil
were analyzed for crude protein content and
in vitro digestibility. The range of variation N lDMY (kg!ha) %DSP R
observed within this collection is remarkable B. brizantha
(Table10.1).Nineteen selected accessions were range 96 2040·9420 9·27 1.9·3.8
overage (ollec. 96 4797 19 2.6
then evaluated in regional trials in different overage select. 10 7503 18 3.1
ecosystems and superior genotypes were
B. decumbens
identifi~d (Valleet al.1997). The next evalua tion range 35 1348·5543 10·25 1.5·3.2
step involved studying the effects of livestock overage tollec, 35 23611 16 2.0
on the pasture. Eight apomictic accessions were overage select. 5 4063 18 2.7
compared in paddocks to the commercial B. humidicola
cultivar. Four of these were selected (Euclides range 21 1908·4435 9·18 2.3·3.7
et al. 2(01) to undergo animal performance overage (ollec. 21 3242 13 2.7
overage select. 4 4358 15 2.6
trials, the last step prior to release as a new
cultivar. B. ruIiIiensis
range 20 1563·2685 9-19 1.4 - 2.7
Morphological characterization, applying overage (ollec. 20 2160 13 1.1
numerical taxonomy and using 26 descriptors, overage select. 2 3099 19 2.1
was carried out for all 340 accessions in the B. iubam
Brazilian Brachiaria collection (Valle et al. range 11 1281 ·2320 7·22 1.5 -3.1
overage (o/Iee. 11 1327 16 2.3
1993c). The objectives were to study the overage select. 4 1864 18 2.7
diversity of the accessions, analyze the
H. =number of lJ((essions; LDMY =Leof dry mane! yield; "DSP =
dispersion and genetic distance between Percentoge dry season production; • = regrowth during roiny season
accessions and species, and organize the (0-6 maxI. Volues ore 3-yeor averoges.
142 CdIa ......... V... ..t ... w. ....
experimental plots. Previous reports on some The diversity of the introduced collection
species of this genus established the basic justified a thorough search for sexuality. The
chromosome number as n = 9, and the most mode of reproduction was determined by
common ploidy level among commercial examination of embryo sacs for 427 accessions
cultivars as 211 = 4x = 36 (Schank and of 15 different species in Colombia and Brazil
Sotomayor-Rios 1968;Ferguson and Crowder (Table 10.2). Flowers were fixed in FAA for 24
1974; Valle 1986). B. ruziziensis is the only hours and later transferred to 70% ethyl
commercially cultivated species that is diploid alcohol. Ovaries were then extracted under a
and obligately sexual, with normal stereoscope and cleared using dehydration
chromosome behavior at meiosis. Other and methyl salycilate (Young et al. 1979).
species are polyploid (4x or 6x) and have Structures were mounted on slides and
irregular meiotic configurations. These examined with interference contrast
polyploids are apomictic, with apospory microscopy. Results include discovery of
characterized by a 4-nucleate embryo sac of obligate sexual accessions in species previously
the Panicum-type. One egg-cell and one considered obligate apornicts, such as B.
(occasionally two) conspicuous polar nucleus decumbens, B. dictyonellra, and B. brizantha, and
can be observed in cleared ovaries. The two determination of mode of reproduction for
synergids are rarely seen. Meiotic embryo sacs species never before studied, such as B. serrata,
of the Polygonum-type with an egg-eell, two B. plaiimota, and B. Sllblllifolia (Valle 1990).
large polar nuclei, and multiple antipodal cells
Chromosome counts were taken on
are found in the sexual accessions and also in
microsporocytes of various sexual accessions
the apomicts, in differing proportions.
using traditional acetocarmin squashes. It was
Brachiaria is pseudogamous, therefore, pollen
determined that the one sexual B. brizantha and
production results from normal meiosis and
all sexual B. decumbens accessions were
is abundant both in apomictic and sexual
diploids, whereas the majority of apomictic
plants.
accessions of these two species were tetraploid.
PRIMl
10..------------------.
Table 10.2 Mode ofreproduction of 15 species of
BrfJC#riaria. based on embryo-sac analysis
-/I
/
/ Spedes no. accessions Rcmge sex SEX APO
/
/
/ B. arrecto 3 79-90 3 o
/1 B. bavanei 4 7- 27 o 4
I.
/ B. brizantha 235 0- 94 1 234
~:~~~\;!'"
B. dewmbens 54 0- 100 22 31
·2 B. def/exa 1 91 1 o
·4
- B. dicfyaneura 6 0·96 1 5
B. dura 1 93 1 o
.6.6 B. humidicala 52 0-100 2 50
·4 -2 o 2 4 6
B. jubata 34 0-94 5 29
PRIM2
B. miliiformis 1 6 1 o
Figure 10.2 Distribution of 253 ac:cessions of B. nigrapedata 3 5- 20 o 3
BrfJChiario (B = B. "rizon'~ D= B. decumbens; B. plafynata 3 3- 97 2 1
R=B. ruziziensis; H =B. humidico/a; J=B. iulHda; B. ruziziensis 24 40-100 24 o
T=B. dictyoneuro) in two planes (PRIN1 and B. serrata 2 30-100 2 o
PRIN2) generated by Prindpal Component Analysis B. subulifalia 4 7-38 o 4
using seven morphological descriptors. Tolal 426 65 361
........U,_kmSptcin 143
The sexual B.humidicola accession is tetraploid, Ndikumana (1985) obtained the first
with regular bivalent pairing, whereas the few interspecific Brachiaria hybrids by crossing the
apomict B.humidtcola examined are hexaploid artificially-induced sexual tetraploid B.
(Valle et al. 1989; Valle and Glienke 1991). ruziziensis to natural tetraploid apomictic
These materials are relevant for breeding accessions of B. decumbens or B.brizantha. The
purposes and also for studies of phylogenetics 35 hybrids obtained were screened for mode
and polyploidization within the genus. of reproduction and chromosome behavior. Of
29 hybrids with B. decumbens, 15 were sexual
More recently, a thorough survey of ploidy
and 14apomictic. Six hybrids with B.brizantha
levels of 435 accessions in 13 species of
broke down into four sexual and two
Brachiaria in Campo Grande, Brazil, was
apomictic. The ratio of sexual:apomictic
accomplished by means of flow cytometry
obtained was close to 1:1, which pointed to a
(Penteado et al. 2000). Traditional cytology was
simple model of inheritance of apomixis. The
used to verify chromosome numbers in
relative ease of crossing after ploidy barriers
specific situations. Despite a predominance of
were removed, together with the
tetraploidy, variation in ploidy level within
chromosomal configuration of hybrids,
species was confirmed and levels never before
suggests a probable agamic complex involving
described, such as pentaploids in B. brizantha
these three species. Further support for this
(Letteriello et al. 1999) and heptaploids in B.
comes from work undertaken in Colombia
[ubaia and B. humidicola, were observed.
and Brazil since 1988. Large numbers of
Cytometry is a valuable tool in breeding
crosses have been made under greenhouse
programs involving polyploid species because
and field conditions between sexual tetraploid
a large number of hybrids must be quickly and
B. ruziziensis (R) and different apomictic
accurately screened for use in additional
genotypes of B.decumbens (0) and B.brizantha
crosses.
(B)(Valleet al. 1991;CIAT 1992). The objectives
Early hybridization studies indicated the need of these hybridization programs were to
for ploidy compatibility to effectively produce explore the potential for manipulating
hybrids in Brachiaria. Attempts by Ferguson apomixis in applied Brachiaria breeding
and Crowder (1974) to hybridize a sexual programs, to study the inheritance of mode of
diploid B. ruziziensis with an apomictic reproduction, and to enlarge the tetraploid
tetraploid B. decumbens proved unsuccessful. sexual pool of Brachiaria. To this end, the
More recently, Hacker (1988)crossed a diploid experimental scheme crossed different clones
sexual B.decumbens with a tetraploid apomictic of sexual tetraploid B. ruziziensis (derived from
cytotype and recovered a single, totally sterile the original Belgian material) to B. decumbens
triploid hybrid. In Belgium, natural diploids cv.Basilisk (which has excellent adaptation to
of B. ruziziensis were successfully acid soils and vigorous stoloniferous growth)
polyploidized using colchicine (Sweene et al. and to B. brizantha cv. Marandu (which has
1981;Gobbe et al. 1981). The resulting induced- spittlebug resistance and vigorous tufted
autotetraploids maintained the obligate growth) (Figure 10.3).
sexuality of the original diploid material.
Greenhouse crosses were accomplished
Recently, three diploid accessions of B.
without prior emasculation of potted plants
decumbens were duplicated in vitro as
of B. ruziziensis or the sexually-reproducing
described by Leblanc et al. (1995a), opening
hybrids. Inflorescences were brought from the
the possibility of intraspecific hybridizations
field on the afternoon before pollination and
(Pinheiro et al. 2000).
144 c.iIlo ....... v...... w.....
kept in vases of water. On the day of Studies conducted at ClAT identified an alpha-
pollination, inflorescences were shaken over beta esterase system capable of discriminating
petri dishes to collect pollen, which was used among putative hybrids of carefully selected
on flowers from which stigmas had just progenitors (Cruz et al. 1989a, 1989b;Calderon
extruded. The inflorescence from the sexual and Agudelo 1990).
plant was prepared by removing unopened
Second generation crosses in Brazil included
and old flowers. After brushing the turgid
sexual x apomictic backcrosses, crosses
stigmata with pollen from the apomictic
between half sibs, full sibs, selfing of sexual
parent, the racemes of the sexual plant were
Fls, 3-way hybrids, and facultative apomictic
individually bagged and labeled. Bags were
x apomictic crosses. Results from this
collected when seed shattering started.
experiment (Table 10.3) point to a single
Scarified seeds were individually germinated
dominant gene determining apomixis, as
4-6 months later in styrofoam trays with a
proposed for Panicnm maximum (Savidan
sand:perlite mixture (2:1) or in petri dishes,
1983), for Brachiaria (Ndikumana 1985), and
and then transferred to plastic bags with soil,
Cenchrus (Sherwood et al. 1994). The excess
from which they were later transferred to the
number of sexual plants observed in some
field (Valle et al. 1991).
crosses may be due to crossing procedures;
Mode of reproduction was determined by sexual maternal plants were not emasculated,
embryo-sac analysis on 30-40 ovaries of 376 and no special precau tions were taken to avoid
individual first-generation hybrids from pollen circulation in the greenhouse, except
greenhouse crosses in Brazil. No reliable after pollination when flowers were bagged.
genetic marker yet exists to determine hybrid
nature of the progeny, therefore attempts to
discriminate among individuals were made
using morphological characteristics and/or Table 10.3 Segregation for mode of reproduction in
electrophoresis. Whenever parental materials 'rachiar;a hybrids
display wide differences in morphology or in Type aoss SEX APO STER Abn
band patterns, hybrids showing intermediate F,
characteristics can be readily identified. B.ruziziensis x B.Jecumbens 79 49
B.ruziziensisx B. brizantha 125 123
F2
B. ruziziensis 2n sexual B.ruziziensis x B.Jecumbens 2 0 2 0
1
4n sexual
colchidne
. . ---{B. decumbens
5 4nopomlcti< B. brizanlha
B.ruziziensis x B. brizantha
BC
7 0 0 0
B.ruziziensis x B.Jecumbens 10 9 0
+
4n apomictic
I
.
4n sexual 5 4n ~some or
B.ruziziensisx B. brizantha 9 7 0
0
0
3-W
Fl ~iclicothe~ B.ruziziensjs x B.Jecumbens 24 21 9 1
4n opomicti< 4n sexual ... B.ruziziensis x B. brizantha 31 6 0 0
F2 ~ FS
B.ruziziensis x B.Jecumbens 38 27 4 0
B.ruziziensis x B. brizantha 24 32 3 1
agronomic HS
evaluations B.ruziziensis x B.Jecumbens 51 53 6 8
Figure 10.3 Hybridization scheme for breeding B.ruziziensis x B. brizantha 61 23 6 0
Brachiaria (adapted from Gobbe et at. 1983). BC: bodmoss, 3-W: three-way hybrids, FS: full-sibs, HS: half-sibs_
.......IA,. .", Spodos 145
The mode of reproduction of an independent reproduction of the mother plant was inferred
set of 107 first generation Brachiaria hybrids from the relative uniformity or heterogeneity
was determined by progeny tests and by of the open pollinated progeny. The results
embryo-sac analysis in Colombia (fable 10.4). were later compared to microscopic
Embryo-sac analysis determined that 56 of the examination of embryo-sac structures of the
plants were sexual and 51 were apomicts hybrid mother plants. The two methods
(Miles and Valle 1991), a finding that agrees agreed closely, except for ten of the progenies,
with the proposed hypothesis of simple in which the degree of sexuality (determined
inheritance. Although interspecific crosses by embryo-sac analysis) ranged between 10
may not be ideal for studying inheritance of and 77%. (Table 10.4).The degree of effective
apomixis, work on the agamic complex sexuality as detected by the progeny test was
formed by B. ruziziensis, B. brizantha, and B. not closely associated with the proportion of
decumbens indicates simple genetic control for sexual embryo-sacs observed microscopically.
apomixis. Sexual x apomictic crosses released Whereas facultative apomictic hybrids with 10
a large amount of phenotypic variation in the or 17% sexuality produced heterogeneous
progeny (plant morphology, growth habit, and progenies, other hybrids, in which sexual
flowering time). embryo-sacs were observed in up to 73% of
the progenies, appeared to behave as obligate
Progeny tests of the 107 open pollinated, first-
apomicts in the test. It is unclear what factor{s)
generation interspecific hybrids were also
contributes to determining effective
used to assess reproductive behavior (Miles
reproductive behavior. Elucidation of the
and Valle 1991). Seeds harvested from
mechanism of apomixis might help explain its
individual plants were sown in five-plant
expression under different circumstances.
plots, in one to four replicates. The mode of
At CIAT and the Institute for Grassland and
Table 10.4 Comparison between progeny lesl and Environmental Research (IGER),
embryo-sac lIIICIIysis for delermination ofmode of Aberystwyth, Wales,U.K.,a molecular marker
reproduction for firsl-generation inlerspecific for the apomixis gene(s) is being sought, which
'racltitJritl hybrids could prove potentially useful for determining
Rate of reproductive mode. Pessino et a1. (1997),using
Mode of reprodudioll selllality (%) a bulk segregant analysis and RFLPs and
embryo-sac embryo-sac RAPDs, were able to identify molecular
progely-Iesl .alysis .alysis markers cosegregating with apomixis in a
54 hybrids sexual sexual small (n = 43) Brachiaria F1 population. Two
37 hybrids apomictic apomictic 10- 73 clones (umc147 and umc72) belong to a
4hybrids undassilied apomictic 7-83 duplicated linkage group that maps to the
2hybrids undassilied sexual
distal part of maize chromosome-1 long arm
FlKlIItative "dS 'hl classified as sellual and chromosome-5 short arm. Another,
541-03 sexual apomictic 10
(OPC4), previously reported as a potential
544-04 sexual apomictic 50
549-02 sexual apomictic 17 marker for apospory in Pennisetum, also
554-02 sexual apomictic 63 cosegregated well in Brachiaria. In later work,
554-03 sexual apomictic 77 Pessino et a1. (1998),using RR..Ps and AFLPs,
683-01 sexual apomictic 52 generated a complete map for the region in
687-0I sexual apomictic 70
693-02 sexual apomictic 47 maize chromosome 5, identifying at least two
694-07 sexual apomictic 43 markers closely linked to the apospory region.
702-06 sexual apomictic 30
146 CdIo ........ V..... JolIo W.....
embryo-sac analysis) and (ii) open pollinated contain both sexual and apomictic genotypes.
progenies of selected sexual segregants. The Seed harvested from a sexual segregant will
open pollinated progenies of sexuals will represent a new half sib family, the result of
pollination by both sexual and apomictic
SEX 5 APO
plants in the crossing block. Seed harvested
S + - - ESA, RM idenli~ed from the apomictic segregants will produce a
uniform apomictic progeny. As reproductive
Year 1 mode of selected individuals in the open
Isolated field pollinated progenies is determined, the
crossing block + recombinant open pollinated progenies and
ogrOllOlli< evaluotiOll the apomictic progenies will be defined for the
SSSSSS •.••••.• subsequent year's crossing block.
1 11 + - - Selection In this scheme, both sexual and apomictic
OP (sexual) progeny genotypes are being improved
,.L-----'---'------'----'--J._ _~I seleded (sexuak) simultaneously, and new apomictic
Year 2 S S S S SS .•••...• recombinants-i-candidates for new cultivars-
S S S S S S
S S S S S S will be identified in each cycle. However, the
expense of determining reproductive mode
with every generation is substantial. A critical
assessment of the relative efficiency, genetic
S S S S S S
and economic, among alternative breeding
Figure 10.4 Simplified diagrlllll ofrecurrent mass
schemes is needed.
selection employed in SEX B,acbitril population.
More complex schemes might be envisioned.
Year 1 ? A ? A ? . Since the desired cultivar is essentially a
A ? A ? A F.uI (CARIMAGUA) hybrid, a recurrent selection scheme based on
? A ? A ? crossing block +
ogrOllOlli< evoluotion performance of hybrid genotypes, such as
(A = known apomict) reciprocal recurrent selection (Hallauer and
(? = hybrid progeny, Miranda 1988) or recurrent selection on
RM unknown)
A?A?A? ••••••
specific combining ability (Hull 1945) may be
Sampled I Selection appropriate. In the latter scheme (Miles and
lor ESA ~ A 4------ Escand6n 1997; Miles 1995, 1997), the sexual
RM OP progeny of population would advance in isolation from
idenfified seleded sexuols
the apomictic tester, which over cycles of
Apomidi< progeny Vegetafive propagation
selection would allow the development of
of selected apomids to sublriol, (oquelo
( . ebu resistan(e) heterotic interactions with one or more
apomictic tester clones. With each cycle of
Year 2 ? A ? A ? A ? selection, the scheme would generate a
A ? A ? A ? A number of apomictic genotypes in the test-
? A ? A ? A ?
cross progenies for further testing and possible
elevation to cultivar status.
References Bogdon, A. V. 1977. TropiaJ/ Pasture and Fodder Burlon, G.W., and I. Forbes. 1960. The genetics
PIonts. New York: Longman. and manipulation of obigate apomixis in
Asker, S.l,and J. Jertmg 1992. Apomixis in Burson, B.L 19B3. Phylogenetic investigo1ions of common bohie9oss (PaspaMn IIOIr1hJm
plants. Boco 100on, Florida: (le Preu. aponictic Paspailm dilotrmmr and relaled Auggel. /'roe. VI/1nl. GrossI. CDnpr., leading,
Boshow, Lt 1980. Apomixis and its opplicotion speOes. In /'roe. XIV Int. GrossJJOIIfJr., U.K. 1960. pp.66-71.
incrop implovemenl. In W.I. FeIr and H.H. Lexinglon, Kentucky, 19B 1. Boulder, Cameron, DJ. 19B3. To breed or nollobreed. In
Hodley (eds.), Hybridizlllion 01 Crop I'brIt Colorado: WesIYiew Preu. pp. 170-73. J.G. Mclvor, and RA Bray (eds.l, Genetic
Madison, Wrsronsin: American Sodety of - - . 19B9. Phylogenetics of apomictic tesources 01 Forage I'/onts. EasI Melbourne,
Agronomy Preu. pp.4~3. I'ospolum dilatotlJm. /'roe. XV/lnt. GrllSsJ. Austro50: (SIlO. pp.237-50.
Boshow, It, and tl. Funk. 19B7. Apomidic COIIfJr., Nice, FrOlKe. 19B9. Asso<. Fr. Prod. Calderon, M. de Ios A., and J.Agudelo-Cortes.
grll1SeS. In W.I. Felv (ed.!, PriOOp/es 01 Fourrogere. v.l. PP.731-32. 1990. Hibridodones interespecilims en el
CuItivor Development, vol.2. New York: Burton, G.W. 1992. Monipu\oting apomixis in genera Brochioria. Undergroduote thesis.
MoaniIlon Pub&shing Ca. pp. 40-82. Paspalum. In J.H. Bgin and LP. Mikshe UniYmidod Ncxionol de Colombia, PlWro.
Botislo, LA.I, and L Godoy 2000. (eels.!, /'roe. 01 the Aporrixis WcrlsOOp. B9p.
CarOderiz~iJo p1e1ininor e ~ de
Arlonto, Georgia. 1992. IlSDA-ARS. 104. pp. Cardono, L,J.W. MMs, ood G. Solelo. 1999. An
germoplosmo do genera Paspalum pero 16-19. improved methodology for mcmive lO'eening
prod~Oo de forrogem. ter. Bros. [od«.,
of Brodriorio spp. genotypes for resistance 10
29(1): 23-32.
Aeneolamio rorio (Hornoplero: Cercopidoe). J.
£coo. fnlomoI. 92(21: 496--96.
150 (tdda ....... do Vale . . , . W. Miles
Carnohon, H.L, and H.D. Hill. 1961. Cytalogy Fernondes, A.T.F., lD.Fernondes, V.P.8. Eoclides, Harlan, J.R., M.H. Brooks, D.S. Borgaankar, and
and genetics ollaragt grllSles. 801. lev. and B. Grof 1992. Avali~iio de aceslOl de J.MJ. de Wet. 1964. Nature and
27: 1-162. Paspolum SW. em cansarcio~iia com inheritance afapomixis in8othriochloo
OAT (Centro Internocionol de Agricuhura Arachis pintoi, em areas umidas de baixa and Oirhanthium. Bol. Goz. 125: 41-44.
Tropical). J986. GermoplllSlTlO,.lnlormE ferlilidade. In ledInlemorionol de Hughes, N.R.G, lB.do Valle, M. Herrero. 1998.
Anuol, 19B5, Pastos Tropicales. Cali, Evaluarioo de Pastos Tropiwles!llEPT: 10. Estimativa da resistencia ao cisalhamenlo e
Calombia: OAT. pp. 5-28. Pastures. leuniOn Sobonas. Brosi/io. Dacumenta de a moagem em quatro espeeies de
- - . 1992. Pastures for the Tropiwl 'trabajo Na. 117. Ca6, Colombia: ClAT. Brorhiorio.. In Proc. 351euniDo Anool Soc.
Lowlands: C/AT'! Contribution. Ca6, PP.5 55--110. 8ros. Zooler., Botucatu: SBI, V~OIO. Pp.
Calombia: OAT. Galindo-Llipez, LF. 1997. Transformaciiln 43-45.
--.1993. Biole!hnology lesearch Unit •" 'geneoca de la graminea forrajera Hughes, N.R.G, lB.do Valle, V. 5obatel, J.
Annuallepot1198B·I992. Ca6, Colombia: Brochiorio SW., medianle 10 tli<nica de Boock, N.S. Jessap, M. HllIrero. 2000.
ClAT. bombardeo de particulas. [Genetic ' Shearing strength os an additional
- - . 1998. GralTineos yLegulTinosas translormalion of the foroge grass selection crfferion lor quality in Brorhiorio
Tropicales: OptilTizaciOn de la diversidad Brochiorio SW. by the technique of particle pasture ecotypes. J.Agric. sa., Cambridge,
genetica para IJ\OImuhi~es (Proyedo Ip· bombardment.] Undergraduate thesis. 135: 123-130.
51. 1998.lnformeAnuoI. 1997. Centro UniveMad Nacional de Calombia, Hull, EH. 1945. Recurrent selectian for specific
Inlernocionol de Agricultura Tropical (OAn, Palmira. 130 p. [+ Appendices). combining abmty in corn. 1. Amer. Soc.
Cat!, Calombia. Dorumenta de Trabajo Gobbe, J., B. Lang~, and B·P Lauant. 1983 Agran. 37: 134-45.
no.174. Cali, Calombia: OAT. Apomixie, sexua6te etamelioratian des IBGE (Institulo Brasileiro de Geogralia e
Clayton, W.D., and SARenvoize. 1982. gralTinees Iropicales. Tropicuhuro 1:5-9. Estatisoca). 1995. AnOOrio Estotimco do
Gramineoe (Part 3). In R.M. Polmll (ed.), Gobbe,J., A. Swenne, and B·P Louant. 1981. Brosilv.55. Rio de Janeiro, Braz~: IBGE.
Flora ofTropiwl Eost Afriro. Kew, U.K.: Diploiides noturels etaUloletrapla"ides IBPGR (International Board for Plant Genetic
Royal Botanical Gardens. pp. 575--1100. induin chez Brochiorio ruziziensis Germain Resources). 1984. Tropiwl and subtropical
Combes, D., and J.Pernes. 1970. Variations etEvrard: crfferes d'identjficatian. Agron. forages. Reporl afworking group. Rome:
dansles nombres chromosalTiques du Trop. 36: 339--46. FAO.
Ponirum moxinxnn Jacq. en relation OVe! Grol, 8., R.P. de Andrade, M.S. Franc;o·Dantos, KeIIllI·Grein, G., B.LMooss, and J.Hanson.
le mode de repraduction. ( I.Arod. sa. and MA Sauza. 1989a. Selection of 1996. Natural variatian in Brochiorio and
Paris. 270: 782-35. Broehiorio SW. for the arid·soil lOYaMas of existing germ~osm colledions..In J.W.
Inn, R., J.W. Miles, W. Roca, ond G. de laCruz. the central ~eau region of Braz~. Proc. Miles, B1Mooss, and lB.do Valle (eds.l,
19B9a. Apomixis ysexuolidad en XVI 1nl. Gross/. Cangr., Nice, France. 1989. Brorhiorio: Biology, Agronotrff, and
Broehiorio. 2.Estudios citaembriolOgicos. MIlK. Fr. Prad Fourragere. v.l. pp. 267- Improvemenl. OAT Pubr.catian No.259.
lev. Cubono de Oencio Agrieolo 23(3): 68. Ca~, Calombia: OAT and Compa Grande,
307-12. Grof, B., R.P. de Andrade, MA Sauza, and J.f.M. Brazil: CNPG(/EMBRAPA. pp. 9-15.
(ruz, R., J.W. Miles, W. Roca, ond H. Ramirez. Valls. I 989b. Selection of Pospolum spp. Lopointe,S.~ G. Aranga, and G. 5otela. 1989.
19B9b. Apomixis ysexuot!dad en adapted toseasanally flooded vOlzea lalllk Amethodology for evaluation afhost planl
Brochiorio. 1.Estudios bioquimicos. lev. incentral Brazil. Proc. XVI 1nl. Gross/. resillance in Brochioria to spinlebug
Cubono de Ciencio Agricolo 23(3): 301- Cangr., Nice, France. 19B9. Msoc. Fr. Prad. (Homoptllla:(ercopidae). Proc. XVllnl.
05. Fourrogere. v.1. pp. 291-92. Gross/. Cangr., Nice, France. 1989.Assoc.
Darliogtan, lD.1939. The Evolution ofGenetir Hodter, lB. 19BB. Sexualily and hybridization Fr. Prad. Fourragere. v.l, pp. 731-32.
Systems. Cambridge, U.K.: Cambridge in signalQloss, Brorhiorio decumben~ Trap. Lopointe, S1,M.S. Serrano, G1Arango,
University Press. Gross/. 22: 139--44. G.5otela, and F. Cordoba. 1992. Antibiosis
de Wet, J.MJ., and J.R. Harlan. 1970. Hallauer, H, and ia Miranda. 1988. 10 spinlebugs (Homoptllla: Cercopidael in
Apomixis, polypIaidy and speciatian in QlHJlIiIotive Genetia in Moize Breeding. accessiOlll of Brochiorio spp. J. Econ.
Oiehonthium. Evolution 24: 270-77. 2nd ed.Ames, lawa: lowo State University &rlom%gy85 (4): 1485-90.
Dujardin, M., and W.W. Hanno. 1983. Apomictic Press. pp. 430-34. Leblanc, 0.,M. Dueiias, M. Hemcindez, S. Bella,
and sexual pearl millet x Pennise1um Hanna, W.W., and E.l Boshaw. 1987. Apomixis: V. Garcio, J.Berthoud, and Y. Savidan.
squolllllolvmhybrids. J. Hered. 74: 277- ill idenlifKation and use in plant breeding. 19950. Ovomosame doubling in
79. Clop sa. 27: 113&-39. Tripsacunz the praduction of artificial,
Euclides, V.P.8., lB.Valle, M.lM. Maceda,and Hanna, W.W., M. Dujardin, P. Ozios·Alins, and L sexualtetra~oid plann.I'lOnlBreeding
M.P.06veira. (2001). Evaluotion 01 Arthur. J992. Transflll of apomixis in 114: 226-30.
Brorhiorio brimntfw ecotypes under Pellllise1um. In J.H. Bgin and LP. Mikshe Leblanc 0., D. Grimaneltl, D. Gonzlilez de LeOn,
grazing in III1CIII plols. In Proc. XIX 1nl. (eds.), Proc. oflhe Aponixis Workshop. and Y. 5ovidon. I995b. Detection of the
Gross/. CDngr., PiracKaba, SOo Paulo, USOA·ARS. Atlanta, Georgia. 1992. apamictic mode of repraductian inmaize-
Broz~. FEALQ, 200I.(in press) Atlanta, Georgia: USDA·ARS 104. pp. 30- Tripsarum hybrids using maize RFLP
Ferguson, J.E., and LV. Crowder. 1974. Cytology 33. markers. TIreor. Appl. Genel. 90: 119~
and breeding behavior al Brochiorio Harlan, H. 1983. The scape far callection and 1203.
roziziensis Germain etEvrard. Clop Sci. 14: improvement afforage plann. In J.G.
893-95. Mclvar and RA. Broy (ed!.), Genetic
lesolHres ofForage l'Ionll. East
Melbourne, Australia: (sIRO. pp. 3-14.
...... of Apoooidi< Specin 151
!Bnnis M. 1998. Desenvolvimento de um Ndikumono, 1 1985. Etude de thybridation Quorin, Ll.,and GA Norrmom. 1987.
metodo de lronsformo!:Oo genetim de enlre espt<es oporniIiques etsexuees dons Relociones entre eI numero de
Braclriaria 1flP. pur bomboredeomento de le genre Brochiorio. Ph.D. Disserlolion. cromosomos, su comportonienlo en 10
porticulos. Musters thesis. Universidode de Univ. of Louvoin, louvain-Lo-Neuve, meiosis yeI sislemo reproductivo en el
Brosilio, DJ..8rod. 131 p. 8e~ium. genero Pospolum. IV Congresso
!Bneriello, G., t8. YoKe, D. Christione, ond NorrmoM, GA, (L Quorin, ond B.lBursan. lDtinoomericona de BoItinico. PrlKeedings.
M.I.O. Penteodo. 1999. Gtologio emodo 1989. Cytogenetics ond reproductive Bogotci, CoIornbio. 3:25-34.
de reprod~Oo de lKessm pentoplOides de behovior of different chromosome roces in Renvoize, SA, W.D. CIoyton, ond tH.S.
B. brizontha. In PrlK. 36 RelXliiio AnIJOI do six Pospolum species. 1.Hered. 80: 24-28. Kobuye. 1996. Morphology, toxonomy,
5oc. BIas. looter., 1999. Porto Alegre, Penteodo, M.l.O, A.tM. Sontos, IJ. Rodrigues, ond nofurol distribution of Btochiorio
S8l/Videolor, SUo Poulo, Brod. (J).ROM. tB.Volle, MAtSeixos, ond A. Esteves. [Trin.) Griseb..In J.W.lAiIes, 8.l Mooss,
Forrogiculturo. Ava~o de Forrogeiros. 2000. DeternillOfiio de plait/io e ovolioriio ond tB.do VoKe (eds.l, Brochi«io:
fOR 140. do qlJOrrtidode de DNA tolol em diferentes Biology, Agronomy, ond Improvement.
lulls, S., 1 Nidikumono, 8.-P. Louonl. 1991. esp«ies do genera Brachiorio. &nbropo Co~, Colombio: OAT ond Compo Gronde,
fert~ity of Blodriorio ruziziemis in Godo de Corte, Compo Gronde, 8razil. Brod: (NPG(·EM8RAPA. pp. 1-15.
inl~ crosses with BrochHxio BoIetim de Pesquiso. N.l1. 32p. Rodrigues-Otubo, 8.M., M.I.O. Penteodo, and
decumbem ond Brodriorio brizlJlthu. Pemes, I, R. Rene.<houme, 1 Rene, ond Y. tB.Volle. 2000. Embryo rescue of
meiotic behovior, pollen viobi~ ond seed Sovidon. 1975. Schema d'omeliorotion interspecific hybrids of Brochi«io 1flP.
set. Evphyrico 57: 267-74. genelique des complexes ogomiques du Plont CeH flSsue Orgon CulMe. 61 (3):
Miles, lW. 1995. Application of recwrent type Ponicum. Cohiers OmOM, ser. Bioi. 175-82.
selection for ljJe(i1ic combining obiity to 10: 67-75. Sobotel, V.O, tB.do Volle, lR.lS. Thiogo, It
the improvement of oponiclic Bradriorio. Pessino,S., t nons, lPAOrtiz, I.Armsteod, Bm, M.tMh\ocedo, ond M. llerrero.
In Harnessing Apomixis, An Intemotional t8. do Volle, and M.D. Hayward. 1998. A 1999. Produ~Oo de gcis como pcramenlrO
Conferencer- 25-27 September 1995. genetic I1lOIl of the opospory-region in cornpcrativo de quolidode entre 20 ocessm
(allege Stotion, TX, I/SA. pp. 64. BrCKhiorio hybrids: identification of two de Blodriorio. In PrlK. 36 RIfIIiiio 5oc.
[Abllroct! (allege Stotion, Texas: FAO/ morkers cIose~ ossocioled with the tro~. Bras. looter., SOo Poulo, Brazil. sall
ORSTOM/USDA-ARS/Rockefeller Hereditas, Lund, 28: 153--58. Yideolor. 1999. 1CD-ROM. 066.
Foundolion/Texos Ag.Exp.Sto'/oepI.Soi Pessino,S.; HA OI1iz, O. leblonc, tB.do Volle, Forrogiculturo. QuolidlKle eVolar Nutritiva.
ond Crop so-TAMU/ Pioneer Hi-Bred t nons, ond "'.0. Hayward. 1997. FOR-2.
Inlernotional, Inc. Identificotion of 0 moize linkoge group Sovidon, Y. 1982. Nofure ethered".e de
- - . 1997. Abreeding mme 10 explo~ reloted with oponixis in Bromiorio. toponixie mez Ponicum moUnum JlKq.
helerosis in oponim. (Abstract). In Theoreticol andApplied Genetics.94: 439- TrovauIC etDocuments OmOM, 153.
OMMYT. BooIc ofAbstracts. The Genetics 44. Poris: ORSTOM.
ond Exploitation ofHeterosis in Craps; An Pinheiro, A., M.T. Pozzobon, t8. do Volle, M.I. - - . 1983. Genetics ond ulizotion of
Intemotional Symposium. Mexico, de O. Penteodo, A.tG. Aroujo, ond V.T. oponixis for the improvement of
DJ.:OMMYI, pp. 182-83. Comeiro. 2000. DuptKotion of the guineogross (Ponicum maxinllm Jocq.l.ln
Miles,J.W., ond M.l Es<ondcin. 1997. FtJther chromosome numbel of di~oid BrlKhiorio J.A. Sm~h and V.W. Hoys (eels.), PrlK. XIV
evidence on the inher~once of reprodudive brizontho ~on~ using colchicine_ Plont CeR Intemotional Gross/ond Congress,
mode in Brochiorio. Con. l. Plant SO. 77: Reports 19(3): 274-78 !Bxington, Kentucky, 1981. 8ouIder,
105-07. Pizorro, EA, ond MA Corvolho. 1992. Cerrodo: CoIorodo: Westview Press. pp.182-84.
Miles, lW., S.llopointe, M.l Es<ondOn, and G. InrroducOOn yevoluociOn ogronilmico de - - . 1987.0 meIhoromenlo genetico do
Sotelo. 1995. Inher~once of spill\ebug forrojeros tropicoles. In Red Intemocional genero Pospolum. In Y. Sovidon et01.
resistonce in interljJe(i1ic Bradriorio de &ollJOciOn de Pastas TropicoleslRlffIl (eels.), fncontro Internacionol sabre
hybrids. 1. Eton. fntom. 88: 1477-81. 10. ReuniOn 5obonos. BrosiIio. Documento Melhoromento Genetico de PospoIum, III
Miles, lW,ond tB.do YoRe. 1991. Assessment de Irobojo No. 117. Coli, CoIombio: OAT. &nbropo/FAO/ORSTOM/BID-IlCA, Novo
of reproductive behovior of interljJe(ific pp.l~8. Odesso, Brozil. PrlKeedings. pp. 37-47.
Brochiorio hybrids. Apomixis Hewslener 3: Quorin, tl 1987. Mejoromiento en Pospalum Sovidon, Y., l Jonk, J.c.G. Coslo, and tB.do
9-10. pur medio de hibridociones interespecilicos. Volle. 1989. 8reeding of Ponicum
- - . 1996. Manipulation of oponixis in In Y. Sovidon et01. (eels.!, fncontro maximum in Brozil. 1.Genetic resources,
Brodriorio breeding..In lW. Miles, B.l Intemocionol sabre Melhoromento GemJtico modes of reproduction ond breeding
Mooss, ond tB.do Volle (elk.), Brodriorio: de Pospalum, Novo Odessa, Brozil: III prlKedures. fuphyric041: 107-12.
Biology, Agronomy, ond Improvement. &nbropo/FAO/ORSTOM/BID-IlCA. pp.27- Schonk, S.t,ond A. Sotomoyor-Rios. 1968.
Coli, Colombio: OAT ond Compo Gronde, 29. Cytologicol studies on Bradriorio species.
Brod: (NPG(-EM8RAPA.pp. 164-77. - - . 1992. The nolure of opomixis ond i~ Soil Crop xi.5oc. Florida PrlK. 28: 156-
Nokofl/1lO, t, M. Ochi, ond N. Mo<hizuki. 1978. origin in Ponicoid grosses. In PrlK. 1st 62.
Charocteristics ond voriations of APONfT Workshop, Montpelliet; APONfT. Seiffert, N.F. 1984. Gromineos forrogeiros do
guineogross stroins collected ond Apomixis Hewslener 5: 8--15. genera Brochioria. EMBRAPA-CNPGC,
introduced from Africo. I. Evaluation of Grculor Tecnico, 1.Compo Gronde, Braz~:
charocteristics ond variotions. Bull. Not/. EMBRAPA-(NPGt
Gross/. Res. Inst. lop. 12: 38--53.
Sherwood, RI,tt 8erg, and BA. Young. Vale, tB. do, S. cmooa, and M.t Arnezquila. Vale, tB. do, Y. ScMdIl1, and LJank. 19B9.
1994. Inheritance of apospary in 19930. Af"0IIIllri< ewumon of 8rrKIriaritJ Apanixis and sexuatlly in Brachiorio
buffe9ass. Crop sa. 34: 1490-94. gennplasrn in Brazil. hac. XVl/int. Grms/. decrriens Slapf. hac XVlInt.
Swenne, A., B-P I.oucI1t, and M.llujlfcin. 19BI. Congr., NZGA, TGSA, MlSAP, ASAP-QId, and GnmI.CDrrJr., Mice, France. 19B9. Ass«.
Inductian par la cakhiOne de formes MOO, Palmeman North, Mew ZeoIand. Fr. Prod. Faurragere. v.1. pp. 407~8.
aulol8lraplaides mez BracIriorio ruziziensis 1993. pp. SII-J2. Vale, tB., and Y. ScMdan. 1996. Genetics,
Germain et EvnJrd (Grlllllinees). Agron. Vale, tB.do, and t G1ienke. 1991. Hew sexual cytagenetics and reploductive biology of
Trop. 36: 134-41. lXmsians in Brachiario. Aponixis 8nKJriIriJ.. In lW. Miles, B.L Maass, II1d
Ta5aferro, c.M., and E.tBashaw. 1966. ItewsJetter 3:11-13. tB.do Vale (eels.), Brachiorio: Biology,
lnheritlIlce and contra! of oIJ5gate Vale, tB.do, t G5enke, and G.D.t AgrOllClllf, and Improveme"'. Cali,
aponixis inbreeGng buffelgrass, l.eglizaman 1993b. BreeGng of apomicti< CoIambia and Campa Grande, Brazil: OAT
Pemisetum diore (CendJrus cicel. Clop Bradri«io through inlerespecilic and (HPG(-EMBRAPA. pp.147-163.
SO. 6: 473-76. hybricisatian. hac. XVl/lnt. Grms/. CDrrJr., Vale, tB., J R. Valerio, S. (a~X1o, A. D.
Toledo, J.M. led.). 1982. ManUlll plFa 10 HlGA, TGSA, MlSAP, ASAP-aid, and MZIAS, Bar• . 1997. Characteristics of selected
eya/uaciOn DgfOII6nico: ledInIemDcionoI Pamersran North, Hew Zealand. 1993. pp. genotypes of 8rrKIriaria for BrazitlOn
de EwIvoci6n de 1'r7stos Tropicules. Ca~, 427-2B. pasl\Jes. In hoc. XVll/lnt. Grass/. (ongr.,
Colombia: QAT. Vale, tB.do, G.l.eglizamon, and M.R. Guedes. Canadian Forage (0111101, 1999. W"lIlnipeg,
Vale, tB. do. 1986. (ytology, mode of 1991. Intmpecilic hybrids of Brachiorio Canada. PP.l/Bl-1/B2. CD-ROM.v.l,
reproductian and forage quality of selected (Granineoe). Aponixis Newsletter 3: 1a- section I. ID 13SB.
species of Brochioria Griseb. Ph.D. Il. Voigt, P.W., and E.tBashaw. 1972. Apomixis
dissertation. University of "ms-Urbana. Vale, tB.do, B1Maass, tB.de Almeida, and and sexuality in uagroslis curtula. Clop
Urbana, Ilinois. USA. ltG. CosIa. 1993<. Morphological sa. 12: 843-47.
- - . 1990. (a¥o de gennap/asnrJ de chara<terization of Brachiaria ger~ Vaigt, P.W., and B.L BursOlI. 1992. Apomixis in
espBcies de Bradriaria no OAT: esturJas hac. XVlllnt. GrrmI. (ongr., MZGA, TGSA, uagroslis. In .I.H. Bgin Jr. and lP.Miksdle
mos rismtJo DO melharamenta genelica. NZSAP, ASAP.Qld, and MZIAS, Palmeman (eds.). Proc. ApoTrixis Wortshap, Allanta,
EMBlAPA-QlPG<, lIoomenlos, 46. Campa North, Hew Zealand. 1993. pp. ~. Georgia, USA. February 11-12, 1992.
Grande, Bran: EMBRAPA-<NPGC. lISDAlARS. ARS-l04. pp. 8--11.
Young, BA, RI Sherwood, and E.tBashaw.
1979. aec..ed-pislyt and lhKk-sectioning
techniques for detecting aposporous
aponiJis ingrasses. Can. J. Bot. 57:
1668--72.
Chapter 11
Transfer of Apomixis through
Wide Crosses
YVES SAVIDAN
search for an opltimum apomichc donor for described as obligate. which meiJnS that 100%
pearl millet has~n conducted, Threes~ies of the observed prog('ny appeared to be
studied by Dujardin and HlllUla (1989) that millemal in field tests (Dujardin and Hanna
show crossability with the crop are Palllisellfm 1984a, b).
Ori~nIQJI!, a
tetraploid with 211 "" 36; P. setaceum,
Advdntages of P. squamlllulum as a donor
a tnplold with 2" = 27; and P. squlImulaJtIlTl, a
species for apomixis include good pollen
hexaploid with Z" .. 54. They all reproduce
fertility, 4-nuc1eate embryo sacs. and a unique
apomictically and their apomods was
unreduced polar nuclei. This trait can also be Crossing species with different flower sizes
used for screening modes of reproduction in and shapes may require special tricks, e.g., in
segregating populations by means of flow the case of maize x Tripsacum, more hybrids
cytometry (Grimanelli et al. 1997). Previous are produced if the silks are shortened to about
studies showing that 5-6 backcrosses are 2-3 cm. Most wide crosses require embryo
needed to produce introgressed 20- rescue techniques, using classical media such
chromosome maize plants provided another as MS (Murashige and Skoog 1962) or N 6 (Chu
advantage to using this species. Disadvantages et al. 1975). Small embryos from maize x
include total male sterility, which is seemingly Tripsacum Fls grew better on 50 g/l sucrose as
retained until reaching addition forms with compared with standard embryo culture
very few Tripsacum chromosomes, and the medium containing 30 g/l sucrose. Several
difference in basic chromosome numbers (x = environmental factors can further affect the
18 compared to x = 10 in maize). production and culture of hybrid embryos. As
a result, the production of hybrids may be good
Production of Interspecific or one year, but poor the next.
Intergeneric F1 Hybrids When apomixis is not found in wild relatives,
Several procedures are available to produce
transfer may be attempted from a more distant
hybrids between cultivated and distantly
apomictic species by using protoplast fusion.
related wild species. Special techniques,
Such a transfer was started for sorghum using
including chromosome manipulation,
apomixis from Cenchrus ciliaris (Bharathi et al.
bridging species, hormonal treatment, embryo
1991). However, no reports of plant
rescue, ovary culture, and in vitro pollination,
regeneration have surfaced to date, apomictic
are available for overcoming the cross
or not, from such protoplast fusions. A more
incompatibility and the sterility of the F IS. The
recent approach, developed by Ramulu et al.
presence of apomixis makes the cross more
(1996), explores the production of
difficult because it can only be performed in
microprotoplasts containing only one or two
one direction, with the apomixis progenitor
alien chromosomes and the direct production
being used as pollinator. Therefore, the donor
of monosomic addition lines after fusion with
must exhibit good pollen fertility. Because most
protoplasts from the receptor species.
apomicts require fertilization with reduced
pollen to produce endosperm, pollen quality Sterility of the F,s
is generally not affected by apomixis. An Sterility in interspecific and intergeneric Fls
exception to this rule is Elymus rectisetus, in and subsequent backcross generations is a
which male infertility is a problem with most characteristic of wide crosses. Restoring
accessions G. G. Carman, personal comm.). fertility of the F I hybrids through chromosome
doubling is the most common approach. In
Crossing Techniques both pearl millet and maize transfer attempts,
Most of the crossing techniques are common
however, Fls from some wild species
to intra- and interspecific crosses. A
accessions were totally sterile, while those
prerequisite is good knowledge of the self-
obtained from other accessions showed some
sterility or self-incompatibility systems existing
degree of fertility, making the chromosome
within the crop. For most crops, however, hand
doubling unnecessary.
emasculation is preferred.
The transfer programs in pearl millet and
wheat have produced F] hybrids with some
degree of male fertility.However, as described 1,730 hybrids were produced. A sample of 64
below for Tripsacum, this is not an absolute segregated 31 apomictic (30 classified as
requirement. Nevertheless, it obviously helps, obligate) and 30 sexual, which suggests
because the Fls generally have morphological dominance of apomixis over sexuality.
features close to that of the wild progenitor,
Relative crossabilities in maize x Tripsacum and
e.g., a limited number of fertile flowers to
pearl millet x wild species of Penniseium are
pollinate. In maize, the F1s have less than 20
shown in Tables 11.1 and 11.2, respectively.
flowers per inflorescence, while the recurrent
According to J. G. Carman (personal cornm.),
maize parent, ifitcould be used as female (i.e.,
the crossability between wheat and apomictic
if the FI hybrid had some male fertility), would
ElYllllIS rectisetus as measured by the same Fls/
offer hundreds.
ear ratio was less than 1%. Differences in
Penniseiumsetaceum (211 = 3x = 27) was the first crossability may possibly be due to relative
apomictic species crossed with pearl millet. F1 differences in genetic distance between the
hybrids had 211 =25 chromosomes, were male crop and its wild relatives or to genetic effects.
sterile, but reproduced apomictically (Hanna
1979).This interspecific cross was abandoned Production of Apomictic
because of male sterility. Pennisetum orientale Progenies through
(211 = 4x = 36) was then crossed with pearl Backcrossing
millet. F1 hybrids had 211 = 25 = 18 P orientale Facultativeness becomes especially important
(Or) + 7 pearl millet (Pm) chromosomes when interspecific or intergeneric hybrids are
(Hanna and Dujardin 1982). They were male totally male sterile. Dujardin and Hanna (1989)
sterile, but backcrossing was attempted using considered male sterility as an impediment to
pearl millet as the pollinator. the transfer of apomixis because their
Pennisetum squamulatum (2n = 6x = 54) was progenitors were apparently obligate
successfully used to pollinate tetraploid pearl apomicts. This was certainly reasonable based
millet. Crosses with diploid pearl millet failed on the available techniques and limited
(Dujardin and Hanna 1989).Of 20 F] hybrids, number of plants used for analysis at the
15 were facultative apomicts, based on beginning of their project in the early 1980s. In
embryo-sac analyses. One FI was classified as the progenies of the maize x Tripsacum BC3
an obligate apomict, although 35% of the
ovules were considered aborted. This may
Table 11.2Crossabilities between pearl milet and
possibly be interpreted in another way if the three apomidic wild Pennisetum species
timing of sexual and aposporic pathways of
Cross COlllbinatiOll eors F15 F15/ eor
development is different (see Issue # 1).Pollen
fertility of this hybrid was surprisingly high pearl millet 12n = 14) x
(66%) and therefore it was used to pollinate
P.orientale 12n = 36) 88 20 0.23
pearl millet (2n =28) x
tetraploid pearl millet to produce a BC I P.orientale (2n = 36) 70 2 0.03
progeny. The BCI plants were totally male pearl millet (2n =141 x
sterile. The breakthrough was found in P.setoceum 12n = 27l 7 28 4.00
making a tri-specific hybrid. The pearl millet pearl millet (2n = 28) x
P.squamulatum (2n = 54) 59 337 5.11
x P. squamulatum male fertile FI (classified as
an obligate apomict) was used to pollinate a overage 224 387 1.73
pearl millet x napier (P. purpureums FI , and ears =number of pearl millel inflores<ences pollinated with the
Penniselum wild species; F,s= number of F, hybrids grown 10
moturity.
160 Ton So.....
hybrid deri\'ativc!\, only 0.9% of the plants family (see Nogler 1984 for review;Sherwoocl,
apparently resulted from fertilluUion of a Chap,5). In 1'CJ1f1iselllfll, male sterile apomictic
reduced egg cell, i.e., the ratc of diplospory in hybrids could have bt-en agood stdrting point
B~s was 99.1 %, which would pmbilbly not for the traru;fcr of apomixis if flo\\ cytomctry
be detectable rf only 30 or 40 plants were had been aVilltable fOr screening of rarge
anillyzed in a progeny lest. prog~nies, but the technology only ~cam«
available 10 plant scien tssever;tr y~ars after
The obligate nature of nponmfl5 mllY b....
the project began (Gl'Ilbrallh et 31. 1(83).
overestim:lted bee,lulleof the populatIon size,
e.g., Burton et al. (1973) claSSified The BC I plants from pearl millet x PnmisetunI
approximalely 80% of their PIIII/CW1t maximum oriell(nle hybrids hild 23, 27, or 32 chromo-
accessions as obligate apomicts bJSed on 10- somes, The latter were 211 + 11 off·types with
planl progeny tests. Savidan (l982bl, ho\.....~er. 25 I 7 Pm. as pearl mHlet WJ!> used as
found only 20"{, of such obligate apomicts pollinator. The 2.1-<hromDSome plants W(,I'C
using .. lOO-ovary ernbryologicaI analystS for described as Cacultatlve apomiclS, with a low
each accession. The~fore. the male-sterilt> rate (or expression) of apomixis.
apomictic interspccific Fj hybrid rmyprobably
From the crosses with P. $(.'tllceulIl. a ill '= V
always be u<;ed ,lS fem,de in the backcross,
BC, plant appc01red 10 be totally male sterile,
provided progenies of ~tlfficient sizt! cm be
but could be pollinated by pcilrl m!llet or P.
screened. One can expect lhtll a few off-types
st'lnceum. Pollination with peilfl millet
will be produced from sexu.,1 reproduction (11
produced no seed, while pollination with P.
, 11 combinations) 10 help bypass the sterility
Stl/llCCl.IIlr produced four plants, three matenlat
barrier. Some may reproduce apomictically,
and one 271 + 77. The p, orie7llalc pathway wa$
assuming the apomixis "allele" is dominant
considered unsUitable for apomrxis transfer
and Simplex. as observed in all sexual x
because of the low expression of apomixis or
apomictiC hybflds produced 50 far In the grass
complete male sterility in the ne\ derivatives,
Hybrids that are totally male sterile and the~fore suggested that a limIted range of
obligately apomicllc are indeed dead ends: vanallon could possibly allo...... sdecllon back
pollmating such hybrids with the crop polhm to obligate apomixis, In the inlergeneric
will produce onty maternal offspring, i.e" background of maize x Tripsocum hybrid
perfecl copie~ of the sterile Fl, However, if derivatives, the variation obSen't.'CI (T<lble n.3)
apomixis is slightly facullatl'll:, off-type; can .lppeared Less stable, possibly because the
beproduced,someofwhich may ben + nand apomictic Tripsacunl progenitor was already
still apomicnc, representing progress toward much more facultative than the guineagrass
a return to the chromosome number of the accessions used by Savidan (t982), By selt."Cting
crop, The rate of facultativeness has 10 be low, among TTlpsaCllrn accessions for their ability to
howevllr, if one expects the backcross produce hybrid derivatiVes II'l backcrossing F1s
procedure to eventually produce iln apomictic with maize, the team possibly selected Ol"le of
crop germplasm with a high degree oi the most facultative of the apomictic
apomixis. Analyse~ made on POlllcum tnpYcums
nl/mnlllnl (S;widan 191!2a,b)show th..'\tlhe rah!
Table 11.4 shows the cumulatrve result olthe
of facultaliveness, and more pra:iscly of 11 "'/1
analysis of approlumately 6,000 progenies
off-types, may remain relallvdy conserved
produced from maize )C TripsllCWtl BCI5 with
through generations of hybndiZ<'\til)ll. It was
plant progeny is shown in Table 11.5,in which fertile apomictic BC4 had been confirmed as
the rate of n + n off-types was below 0.2%. combining 20 maize chromosomes with less
Almost 200 hybrid derivatives have been than 16 Tripsacum chromosomes (Table 11.6).
produced and classified as BC4, with Increasing the progeny size did not change the
chromosome numbers ranging from 2n = 20 trend, an observation suggesting that the
to 2n = 36. Modes of reproduction could be original transfer scheme (Figure 11.2) had to
being determined for some of them by RFLP be reconsidered, especially since its 38-
markers linked with apomixis, by progeny- chromosome plant step could not produce the
tests, or by ploidy of the endosperms addition lines that were expected.
evaluated through flow
cytometry (Table 11.6). The Table ".S Maize x Tripsacum 8C3 progenies, in which the 8C4s are
progeny size was recently the n + n category
increased further. Progl!llies maternal oH-types oH-types oH-types
.total no. 2/H-0=38 2/H-D=48 /H-D=20-36 11+0=10, 28
Screening the modes of
reproduction through flow 125916 114602 10778 158 78 300
(%) 91.01 8.56 0.12 0.06 0.24
cytometry is a unique
opportunity offered by reproduction apomictic apomictic segregating sexual,apo apomictic
diplosporous species such • mosI~ 4n !resfitution nuclei)
as Tripsacum dactyloides. In
sexual plants, triploid Table 11.6 Maize x Tripsacum 8C4 with known mode of reproduction
endosperms result from Plut 2" ISW RFLP Endo PGT plGllt 2" ISH RFLP Eado PGT
the fertilization, by a 1496 20 Sex 1457 27 13M+14Tr Sex
reduced pollen, of two 1500 20 Sex 1476 27 Apo
reduced polar nuclei. 1502 20 20M Sex 1460 28 20M+8Tr? Sex Sex
Diplosporous plants form 1503 20 Sex 1484 28 20M+8Tr? Sex Sex
1516 20 Sex 1348 30 Apo
endosperm as a result of
1529 20 Sex 1346 31 Apo
the fertiliza tion of two 1454 21 Sex Sex 1347 31 Apo
unreduced polar nuclei by 1482 21 Sex Sex 1439 31 Apo
a reduced pollen. The 1489 21 Sex 1453 31 Apo
difference is shown in 1492 21 Sex Sex 1479 31 I7M+ 14Tr Apo
1535 21 Sex Sex 1276 32 Apo
Figure 11.1. Diploid sexual
1275 22 Sex 1339 32 Sex
plants have triploid 1338 22 Sex Sex 1426 32 Apo
endosperms (peak 2 in 1345 22 Sex 1306 33 Sex
Figure ILIa), while 1422 22 Sex 1349 33 18M+15Tr Apo Apo Apo
tetraploid apomictics 1499 22 20M+2Tr Sex 1493 33 Apo Apo
1534 22 Sex 1532 33 Apo
produced endosperms
1393 23 20M+3Tr Sex Sex 1313 34 Sex
(peak 2 in Figure l1.lb), 1515 23 Sex 1394 34 16M+18TrSex? Apo
with a DNA content 2.5 1229 24 Sex 1494 34 16M+18Tr Apo Apo Apo
times that of the embryos 1425 24 Sex Sex 1517 34 Apo Apo
(Grimanelli et al. 1997). 1481 24 Sex 1521 34 Sex Apo
1526 24 20M+4Tr Apo Apo Sex? 1522 34 Apo Apo
Preliminary data indicated 1528 24 20M+4Tr Sex Sex 1523 34 Apo
apomixis could be 1471 25 20M+5Tr Sex Sex 1544 35 Apo
1501 25 20M+5Tr Sex 1308 36 20M+ 16Tr Apo Apo
transmitted to the BC4
generation, although no 'ISH: in situ hybridizolion dolo; RFLP: use 01 markers linked 10 apomixis; EOOo: flow cyIometry
ono~is 01 the ploidy 01 the endosperrrn; PGT= progeny-Iesl (duomosome counts).
164 T... smdoo
3
l~x_
unanswered. However, more progress has and manipulation of its components. Another
been achieved toward producing an apomictic promising avenue, approaches based on
grain during the last ten years than ever before, mutagenesis, is discussed in the following two
mostly because of the development and chapters. These approaches will undoubtedly
application of new techniques. As molecular better our understanding of the regulation of
dissecting tools continue to improve, we will reproduction as a whole. In the end, apomixis
see great progress in our understanding of certainly cannot be manipulated without a
how apomixis is controlled and the isolation thorough understanding of how it is controlled
in the wild.
References Chu, «, «. Wang, cs. Sun, c Hsu, It rill, Grimonelli, D., O. leblanc, l Espinaso, E.
and tY. Chu. 1975. Establishment of an Perani, D. Gonztilez de LeOn, and Y.
Asker, S. 1979. Progress in apomixis research. efficient medium far anther ru~ure 01 rice Savidan. 19980. Mopping diplosporous
Hereditas 91: 231--40. through comparative experiments an the aponixis in tetraploid Tripsorum one gene
Asker, 5., and LJerling. 1992. Apomixis in nitrogen sources. Sei. Sinica 18: 659-68. or several genel? Heredity 80: 33--39.
Plants. 8o<a Raton, Flarida: (RC Press. de Wet, J.MJ.de 1979. Tripsocum introgression - - . )998b. Non-MendetlOn tronsmission
Bharafhi, M., U.R. Murty, K.8.R.S. V'lSOrado, and and agronomic fitness in moize Ilea ma~ 01 apomixis in moize· Tripsocum hybrids
A. Annapuma. 1991. Passib~ity of l.l, Proc. Coni. Broadening Genet. Base in caused by a transmission ratio distortion.
transferring obligate apomixis from Craps. Wageningen: Pudoc Publish. pp. Heredity 80: 40--47.
CenchlllS ciliaris L la Sorghum bicalar (Ll 203--09. Hanna, W.W. 1979.lnterspecific hybrids
Moench. Apomixis NewsJener 3: 13--14. Dujordin, M., and W.W. Hanna. )9840. between peari m~let and faunlaingross. 1.
Boshow, It 1975. Prablems and possibilities of Micrasporogenesis, reproductive behovior Hered. 70: 425-27.
aponixis in !he improvement of tropi<a1 and fertility in live Pennisetum species. Hanna, W.W., and M. Dujordin. 1982. Apomictic
forage grasses. Tropical Forages in T1rear.Appl. Genet. 67: 197-201. interspecjfic hybrids between pearl nillet
livestock Praclvclion Systems. 24: 23--30. - - . 1984b. Pseudagamous and !'cxientale Lt Rich. Crap SO. 22:
Modison, Wisconsin: Ameriron Society of parthenogenesis and fert~izatian of a pearl 857-59.
Agronomy. rrnllet x Pennisetum orientale aponiclic Hanna, W.W., M. Dujardin, P. Ozios-Akins, l
Boshow, It, A.W. Hovin, and E.tHolt. 1970. derivative. 1. Hered. 75: 503--04. lubbers, and LArthur. 1993.
Apomixis, ~ evolutionary significance and --.1986. An aponictic poIyhaploid Reproduction, cytology, and fertility of
utilization inplant breeding. Proc. Xllnt. obtained from a pearl nilJef x Pennisetum pearl m~let x Pennisetum JqlJOtrlJlatum
Grass/. Congr. Madison, WISConsin: squomuIatum aponictic interspecific 8C4 plants. 1. Hered. 84: 213--16.
Amerimn Society of Agronamy. pp. 245- hybrid. Theor. Appl. Genet. 72: 33--36. Harlon, J.R., and lM.J. de 'NlltI977. Pathways
48. - - . 1989. Crossobility of pearl millet of genetic transfer from Tripsocum to leo
Bergquisl, R.R. 1981. Transfer from Tripsocum with wild Pemisetum species. Crap SO. 29: mays. I'roc. NatI. Acad. SO. (USA) 74:
dacty/aides to corn of a major gene Iorus 77--30. 3494-97.
conditioning resislllnce to Pucdnio sorghi. Engle, LM., J.MJ. de Wet, and H. Horlan. James, J. 1979 New moize x Tripsocum hybrids
I'/rytapathalagy 7: 51S-20. 1973. Cytology of backaoss offspring for moize improvement. fuphytica 28:
Bemord, 5., and D.tJeweH. 1985. CrMSing derived from a moize- Tripsocum hybrid. 239--47.
moize with ~um, Tripsocum and nilet Crop Sci 13: 690-94. Kindiger, 8., and V. Sokolov 1995. Occurrence of
!he products and !heir level of - - . 1974. Chromosomal variation partial meiotic behoviars inapomKtic
develcpment following pollination. Theor. among offspring of hybrid derivatives with eastern gamograss.. In 8.Sheridan (ed.),
AppI. Genet. 70: 474-83. 20 leoand 36 Tripsorum chromosomes. Annual Maize Genetics Conference, 37th,
Birchler, JA 1993. Dosage analysis of moize CDryalagio 27: 193--209. Pacific Grave, Cahfornia, 16-19 Morch
endasperm development. Ann. lev. Genet. Galbraith, D.W., K.R. Horkins, lM. Maddox, 1995. Grand Forks, North Dakota:
27: 181-204. NA Ayres, D.P. Shormo, and l University of North Dakota. pp.24.
8rawn, W.V., and H.P. Emery. 1958. Apomixis in Firoozababy. 1983. Arapid Row cyIOOIetric Kindiger, 8., V. Sokolov, and tL Dewald. 1996.
!he Granineoe : Panicoideoe. Amer. 1. Bat. analysis of cell cycle in intact plont tissues. Acomparison of apomKtic reproduction in
45: 253--63. Science 220: 1049-51. eastern glR110grass {Tripsoaxn d«ty/aides
8urton, G.W., It Millat, ond W.G. Monson. Ga~nat, W.t 1971. The origin of moize. Ann. (UU and apomictic maize- Tripsocum
1973. 8reeding procedures far Panicum lev. Genet. 5: 447-78. hybrids. Genetica97: 103--10.
maximum suggested by ~ant variabi~ Gatlllat, W.t, P. Chandravodana, and 8.G.S. leblanc, 0., D. Grimonel5, D. Gonztilez de LeOn,
and mocIe of reproduction. Crap SO. 13: Rao. 1970. Cytologicol mop of Tripsocum and Y. Savidon. 1995b. Detection of !he
717-20. cIoctytoides (2n=36). Maize Genet. Coap. apanidic mocIe of reproduction in moize-
Carman, lG.1997. Asynchronous expresYon of News/. 44: 114-16. Trips~C'Jm hybrids using moize RFLP
duplicale genes inangiosperms may C!Me Grimaneli, D., M. Hernlindez, l Pet'alli, and Y. markers. T1rear. AppI. Genet. 90: 119S-
aponixis, bispory, tetraspary, and Sovidan. 1997. Dosage effe<ts in !he 1203.
polyembryony (review). Bioi. J. Unn.Sac. endosperms of diplosporous aponiclic
61: 51-94. Tripsorum. Sex. Plant leprod. 10: 279-32.
T..sfer ., ApoooUis tftotIo vr. Cm... 167
l.eblanc, 0., D. Grimanelli, N.IsIam-Faridi, 1 Ndikumana, 1 19B5. Etude de thybridation Petrov, DJ., N.I. BeloUlOvo, E.S. Folcina, R.M.
Berthoud, and YSoman. 1996. entre especes apom~iques et sexuees dons Ya15enko, LI. loikova, and IP. Soralcillll.
Reproductive behaviar in maize·TripsoclHO le genre Brochiorio. Ph.D. dissertotion. 1984. Tronsfer of same elemeo15 af
poIyhoploid ~an15: Implications for the Univ. af louvain, louvain·lo·Neuve, opallixis Iram Tripsonm lamaize. In DJ.
transfer afapomixis inta maize. 1. Hered. Be~ium. Pelrav led.), Apomixis and /Is Role in
B7: lD8-1l. Nogler, GA. 19B4. Gametaphytic opomixis. In Evolution ond Plont Breeding. Russian
l.eblanc, 0., M.D. Peel, J.G. Carman, and Y. B.M. Johri Ied.), Embryology of Tronslation Series 22. Ronerdom: AA
Soman. 19950. Megasporogenesis and Angiosperms. Berlin: Springer·Verlag. Bolkeno. pp. 9-78.
megagametogenesis inseveral TripsoClHO pp.47>--518. Purseglave, lW. 1972. Tropim/ Crops,
speGes (Pooceae). Amer. 1. Bot. B2: 57- Nairot, M. 1993. AIlerK ratios and ster~ity in lhe Manocoty/edons. New York: larIQmcm.
63. ogamic complex of the Maximoe Romulu, K.S., P. Dijkuis, E. Rutgers, 1 Blaos, FA
l.eblanc, 0., and Y. Soman. 1994. Timing af (Panicaideae): evolutionory rale of the Krens, JJ.M. Dons, CM. (oIi;1·Haajmans,
megasparogms in Tripsocum spe<ies residual sexuol~. 1. £1'0/. BiD. 6: 9>--1 Dl. ond HA Verhoeven. 1996.
IPooceae) os related ta the control af Ozios·Akins, P., E.L tubboo, W.W. Honna, ond Miaoprolaplasl-mediated transfer of single
apallixis and sexuality. Polish. Bot. SfIJd. lW. McNay. 1993. Tronsmission of the specific chramasames between sexualy
B: 75-81. opallictic made of reproduction in inca~~ble ~an15. Genome 39: 921-33.
Mangelsdar/, P'C, and R.G. Reeves. 1931. Penniseturrr. co-inher~ol1Ce of the tra~ ond Somon, Y. 19820. Nature el heredile de
Hybridization af maize, Tripsocum, and moIerular marke~. TIreor. Appl. Genet. B5: topamixie chez Ponicum maximum Jacq.
Euchlaellll. 1. Hered. 22: 339-43. 632-38. Tmvaux & Documents ORSTOM 153: J-
Marshal!, R.D., and A.H.D. Brown. J9Bl. ~zios·Akins, P., D.Roche, ond W.W. Hanna. 159.
Estimation afthe level afapomixis in~ont 1998. TIQht dustering ond hellizygas~ of - - . 1982b. Embryologicol analysis of
papulations. Heredity 32: 321-33. apallixis·rtnked molecular markers in foruholive apomixis in Ponicum moximum
Martinez, EJ., f. Espinaza, and CL Quarin. PeMisetum squamulatum im~ies generic: Jocq. CropSci. 22: 467~9.
1994. Bill progeny 12ntnllram apallidic control afopospary by 0divergent locus - - . 2000. Apallixis: Genetics and
Paspa/um IIIltatum obtained through early thot may have na ollelic form in sexuol Breeding. Plant Breemng Reviews 18: 13-
pallilllltion. 1.Hered. 85: 29>--97. genotypes. Proc. Nat. Acod. Sci.IUSA) 95: B6.
Magie, M. 1988 Amodel for the evolution and 5127-32. Simane, G.W., and A.L Hooker. 1976.
control of generative apomixis. Bioi. 1. Pernes, J.1972. Organisation evolutive d'un Monogenic resistonce in com ta
linn. Soc. 35: 127-53. graupe ogallique: la section des Maximoe Helllinthasparium turcicum derived Iram
Murashige, 1, and F. Skaag. 1962. Arevised du genre Ponicum IGrallinees). Ph.D. Tripsocum Roridol1lJm. Proc. Am.
medium for rapid growth and biaas~ dissel1olion, University of Paris. I'hytopothol. Soc. 3: 207.
with tobacca ~swe ruhures. PIrysioI. PIont. Petrav, DJ., N.I. Belausava, ond E.S. Falcina.
15,413-97. J979.lnher~0I1Ce of opallixis and m
elemen15 in corn-TripsoclHO hybrids.
Genetlco 15:1827-36.
Chapter 12
From Sexuality to Apomixis:
Molecular and Genetic Approaches
UELI GROSSNIKLAUS
gametes and embryogenesis (Walbot 1996). its apomictic relative TripsaCllm dactyloides (e.g.,
The meiotic products of plants undergo several Mangelsdorf and Reeves 1931; Harlan and de
division cycles to form a multicellular haploid Wet 1977; Petrov et al. 1984; Savidan, Chap.
organism. The gametes differentiate later in 11), and the extensive synteny among the
gametophyte development. In angiosperms, grasses (Bennetzen and Freeling 1993; Moore
double fertilization concludes the et al. 1995; Gale and Devos 1998; Keller and
gametophytic phase and marks the beginning Feuillet 2000) allows for comparative genomic
of the next sporophytic genera tion. In analyses between sexual and apomictic grass
angiosperm apomixis, the plant life cycle is species. The Mlltator transposon system offers
short-circuited and an unreduced cell lineage highly efficient methods for insertional
gives rise to a megagametophyte (gameto- mutagenesis (Chomet 1994) and for site-
phytic apomixis) or directly to an embryo specific transposon mutagenesis by reverse
(sporophytic apomixis). Because of the close genetic approaches (Das and Martienssen
developmental and evolutionary relationship 1995; Bensen et al. 1995; Mena et al. 1996;
between apomictic and sexual reproduction, Rabinowtiz and Grotewold 2000), originally
a better understanding of the fundamental pioneered in the fruit fly Drosophila
biological principles governing female gameto- melanogaster (Ballinger and Benzer 1989).
genesis and seed development will provide
The small plant Arabidops is thaliana, a member
invaluable tools for the manipulation of the
of the Brassicaceae, has been widely adopted
sexual reproductive system towards apomixis.
as a model system for the developmental
Sexual Model Systems biology and genetics of flowering plants
Two well-established sexual model systems, (Meyerowitz 1989).The small size of the plant,
Arabidopsis thaliana and Zea mays (maize), and its rapid life cycle, and the large number of
a rapidly emerging system, Oryza sativa (rice), seeds it produces make it ideal for the isolation
are of particular interest for genetic and and study of mutants that affect biochemical
molecular investigations. All three are well and developmental pathways. The small
characterized at the genetic level and offer a genome size (-125 Mb), high percentage of
vast array of powerful genetic and molecular single copy DNA (Pruitt and Meyerowitz
techniques (Freeling and Walbot 1993; 1986), large number of molecular markers
Meyerowitz and Somerville 1994;Tanksley and (http://www.arabidopsis.com). and the
McCouch 1997;McCouch et al. 1997).Versatile complete genome sequence (Arabidopsis
transposon systems for insertional Genome Initiative 2000) make Arabidopsis a
mutagenesis and gene tagging are available powerful system for molecular studies. Highly
and offer the opportunity for reverse genetic efficient T-DNA-based transformation
approaches (Walbot 1992; Dellaporta and methods (Bech thold et a1. 1993) and
Moreno 1994;Feldmann et al. 1994;Shimamoto heterologous transposon systems for targeted
et al. 1993; Shimamoto 1995; McKinney et al. gene tagging, genome wide insertional
1995; Sundaresan 1996; Parinov et al. 1999; mu tagenesis, and reverse genetics are
Speulman et al. 1999; Tissier et al. 1999; available (Feldmann et al. 1994; McKinney et
Meissner et al. 1999; Krysan et a!. 1999). a!. 1995; Sundaresan 1996; Parinov et a!. 1999;
Speulman et al. 1999; Tissier et a1. 1999;
Maize, as an agriculturally important member
Meissner et al. 1999; Krysan et a!. 1999). The
of the grass family (http://www.agron.
Arabidopsis genome is the first plant genome
missouri.edu), has some advantages for
to be completely sequenced (Arabidopsis
apomixis research. It can be hybridized with
Fro.Soaoality .. """is:....
ad.... Geetil: Appndos 171
Gunning 1990). The megaspore mother cell is (Randolph 1936;Kiesselbach 1949;Misra 1962;
surrounded by extensive callose depositions, Poliakova 1964; Diboll and Larson 1966',
which isolate it from the sporophytic tissues Russe1l1979; Huang and Sheridan 1994;Webb
of the ovule (Rodkiewicz 1970). and Gunning 1994; Schnei tz et al. 1995;
Christensen et al. 1997; Moore et al. 1997;
In Arabidopsis, the two meiotic nuclear
reviewed in Grossniklaus and Schneitz 1998;
divisions occur before cytokinesis, leading to
Drews et al. 1998). After the first division, the
the formation of tetrads with a multiplanar or,
nuclei migrate to opposing poles of the
more rarely, a linear arrangement (Webb and
developing megagametophyte, and a
Gunning 1990;Schneitz et al. 1995).In contrast,
prominent large vacuole forms in its center. A
cytokinesis accompanies meiosis in maize, first
second vacuole at the chalazaI pole is found
producing two dyad cells and finally the four
in Arabidopsis and some genotypes of maize
megaspores, which form a usually linear tetrad
(Vollbrecht and Hake 1995; Christensen et al.
(Figure 12.1) (Weatherwax 1919; Kiesselbach
1997). As the embryo sac enlarges, the
1949; Russell 1979). Little information is
integuments grow to envelop the nucellus.
available on rice megasporogenesis, but its
Asymmetric growth of the integuments gives
development is likely to be similar to that
the Arabidopsis ovule its characteristic
observed in maize. Only the chalazal-rnost
anatropous shape (Misra 1962). The nuclei at
megaspore survives (functional megaspore)
each pole undergo two synchronous divisions
whereas the other three undergo programmed
to form the 4- and 8-nucleated embryo sac. A
cell death and degenerate. Acertain variability
single nucleus at the chalazal pole starts
with respect to the form of the tetrads and
migrating toward the micropylar pole and
cytoplasm allocation to the functional
becomes one of the two polar nuclei in the
megaspore has been observed (Webb and
central cell. Cellularization leads to the
Gunning 1990; Bedinger and Russell 1994).
formation of seven cells: an egg cell and two
Degenerating and surviving megaspores are
synergids at the micropylar pole, three
initially similar at the ultrastructural level
antipodals at the chalazaI pole, and a central
except for an enrichment of organelles in the
cell harboring the two polar nuclei (Figure
functional megaspore. However, only the
12.1). In Arabidopsis, the nucellar tissue is
degenerating megaspores are surrounded by
absorbed as the embryo sac grows and
a callose rich cell wall, whereas the functional
expands. At maturity, remnants of the nucellus
megaspore remains in direct contact with
are only present at the chalazaI pole. The
nucellar tissues (Rodkiewicz 1970; Webb and
endothelial tissues, which are in direct contact
Gunning 1990; Russell and West 1994). It has
with the megagametophyte, are of
been proposed that the pattern of callose
integumental origin. In maize and rice, ovule
deposition during megasporogenesis plays a
morphogenesis is characterized by an initial
crucial role for the differentiation and survival
intensive proliferation of the nucellar tissue ,
of the chalazal megaspore, which eventually
such that at maturity the embryo sac is still
forms the megagametophyte (Haig and
embedded in this tissue.
Westoby 1986).
The two synergids and the egg cell are
2. Megagametogenesis. The functional
arranged in triangular configuration at the
megaspore gives rise to a mature embryo sac
micropylar pole to form the egg apparatus
of the Polygonum-type by three consecutive
(Mansfield et al. 1991; Webb and Gunning
mitotic divisions that occur in a syncytium
1988; Diboll and Larson 1966; Russell 1979;
Vollbrecht and Hake 1995). The synergids are The egg cell is located at a slightly chalazal
highly specialized cells thought by some to position with respect to the synergids. The
provide constraints for pollen tube attraction distribution of the egg cytoplasm is
and the transport of the sperm cells to the egg asymmetrical with a highly vacuolated
and central cell (Huang and RusseIl1992).Their micropylar pole (or a single large vacuole in
cytoplasm is highly polarized with a chalazally this position) and a chalazally located nucleus
located vacuole and, typically, a centrally (RusseIl1993). In rice, the nucleus is in a more
located nucleus. A highly specialized cell wall, central position and the egg appears less
the filiform apparatus, is associated with the polarized than in other species (jones and Rost
micropylar-most region of the synergids 1989). Nevertheless, the micropylar-ehalazal
(jensen 1968; Mogensen 1988; Russell 1993). (proximo-distal) axis first eminent when ovule
One of the synergids typically degenerates primordia emerge is maintained throughout
prior to fertilization, but the moment for megasporogenesis and megagametogenesis at
initiation of the degenerative process varies the level of the megasporocyte, the embryo sac,
(RusseIl1992; Christensen et al. 1997). and its constituent cells. The polarity of the egg
cell may have important implications for early
-..A-..
\.!)YM
g)
e-~
';::J
h)
An
Naumova 1993; Crane, Chap. 3). An should be kept in mind that in sporophytic
ultrastructural characterization of apomictic apomixis only embryo initiation is affected
development is discussed by Naumova and and, thus, it may be easier to tackle sporophytic
Vielle-Calzada (Chap. 4). Here, I will briefly apomixis at the molecular level.
describe the main developmental features of
In gametophytic apomicts, the embryo results
apomixis in order to facilitate a comparison
from the parthenogenetic development of an
with the sexual pathway.
egg cell produced by an unreduced embryo
Two fundamentally different classes of sac. The unreduced gametophyte can originate
apomictic development can be distinguished either directly from nucellar cells (apospory)
(Gustafsson 1947; Nogler 1984a; Koltunow or from a megaspore mother cell that has
1993) (Figure 12.2). In sporophytic apomixis, undergone no or an aberrant meiosis resulting
an embryo forms directly from a nucellar or in the formation of one (mitotic diplospory) or
integumentary cell in the ovule (adventive two unreduced megaspores (meiotic
embryony). Although adventive embryos are diplospory). Aposporous embryo sacs form
not derived from gametophytic cells, their mitotically from nucellar cells that develop
development depends on the presence of a during or after megasporocyte differentiation
megagametophyte, because they usually rely and are similar in appearance to the megaspore
on sexually derived nutritive endosperm mother cell. Often several aposporic embryo
tissue. Sporophytic apomixis will not be sacs are present in a single ovule in addition to
considered further in this chapter because an the sexually derived one. In diplosporous
engineered switch between sexuality and development, a variety of cytologically distinct
gameotphytic apomixis appears more processes lead to a failure in meiosis; the
attractive for breeding purposes. However, it megasporocyte switches from a meiotic to an
[ '=)
~, ~~
-ed-uc-ed~P-oIoi~ cEooCedE!) I Unrednudeiuced.polar Ill"'""u-nr-ed-uc-ed-egg-1
~,.w:b
~T ~=~~[~"i:.l
IApomictic embryo 11 Endasperm 1 1 Endasperm II-Se-xua-'
em-bry---"'o1 1 Endosperm 11ApomKtic embryo 1
Figure 12.2 'rhe main developmental features of apomixis in relationship tothe sexual pathway.
Unreduced cells are in rectangular boxes, reduced celk are in oval boxes and key events are in darlcly shaded boxes. In sexlllll plants, !he
megaspore mother cell undergoes meiosis and one of the reduced megospores forms the embryo sac. EnUyo and endospenn are formed by
double fertilization. In gametophyti< apomixis, reduction is avoided and embryo soc development initiated from an umduced dipIospare or
an aposporic initial cell. The embryo develops parthenogeneticaly from the unreduced egg while the endospenn forms either autonomously
or through fertilization of !he central cell (pseudogamy). In sporophytic apomixis, the embryo forms directly from on umduced nucelor
initial cell. The apomidi< embryo relies 011 sexually produced endospenn.
176 lIoIGm.......
apomictic pathway with the net result of completely covered by a cell wall (Vielle et al.
prod ucing an unred uced functional 1995) whose presence may prevent the fusion
megaspore (see Crane, Chap. 3). of the apomictic egg with a sperm cell (Asker
and Jerling 1992; Savidan 1992). Some
Unlike in sexual development, the megaspore
apomictic species are truly autonomous and
mother cell of diplosporous species is not
do not require fertilization at all; both embryo
surrounded by callose. In aposporous species,
and endosperm develop autonomously. In
callose deposition around the sexual
contrast, seed development in most apomicts
megasporocyte can be abnormal (Crane and
depends on the fertilization of the central cell
Carman 1987; Carman et al. 1991; Leblanc et
to produce the nutritive endosperm, which is
al. 1993; Naumova et al. 1993; Naumova and
required for successful seed production
Willemse 1995; Peel et al. 1997). As is typical
(pseudogamy) (Nygren 1967;Asker 1979,1980;
for the functional megaspore in sexuals,
Nogler 1984a; Richards 1986; Bashaw and
aposporic initials, which will form apomictic
Hanna 1990; Asker and Jerling 1992). Unlike
embryo sacs, are devoid of callose (Naumova
in sexual species, the apomictic egg cell often
and Willemse 1995). The marked difference of
initiates embryogenesis before the first
callose deposi tion between sex ual and
endosperm division occurs.
diplosporous species is intriguing, but most
likely is not causal, instead being the Interrelationship of Sexual and Apomictic
consequence of a more fundamental lesion Reproduction
(Crane and Carman 1987; Carman et al. 1991; Sexual and apomictic reproduction are
Carman, Chap. 7). developmentally and evolutionarily related.
Apomixis can be viewed as a developmental
The unreduced megaspore of diplosporous
variation of the sexual pathway. Apomictic and
apomicts divides mitotically to give rise to a
sexual modes of reproduction are not mutually
mature embryo sac. Usually, only one
exclusive. Whereas obligate apomicts produce
megaspore of the dyad initiates embryo sac
exclusively clonal progeny, both forms of
development and the other degenerates, but
reproduction coexist in facultative apomicts
bisporic development, in which two
(Asker 1980; Richards 1986; Bashaw and
unreduced nuclei are present in the same spore,
Hanna 1990).They form both reduced egg cells
also occurs (Ixeris-type). In some apomicts, the
that are fertilized to produce genetically
developing megagametophyte undergoes only
diverse progeny as well as apomictic embryo
two mitoses to form a 4-nucleated embryo sac
sacs that give rise to clonal offspring.
where no antipodals form (Panicum-type)
Apomictic and sexual embryo sacs occur in the
(Gustafsson 1947; Nogler 1984a, b; Crane,
same plant or even within the same ovule
Chap. 3). For instance, sexual megagame-
(Asker 1980; Nogler 1984a; Vielle et al. 1995).
tophytes in Penniseium ciliare are of the typical
Facultative apomicts, benefiting from the
seven-celled Polygonum- type, whereas
advantages of both modes of reproduction,
aposporic embryo sacs carry only four nuclei may have an evolutionary advantage and are
and typically form one egg cell, two synergids
more common than obligate apomicts (Nogler
and one polar nucleus (or a variation thereof) 1984a; Richards 1986;Asker and Jerling 1992).
but no antipodals (Taliaferro and Bashaw 1966;
The degree of sexuality versus apomixis in
Vielle et al. 1995).The egg cell of an apomictic facultative apomicts is influenced by a variety
embryo sac ini tia tes embryogenesis of environmental factors, the effects of which
autonomously in the absence of fertilization.
on the reproductive system are not well
In Pennisetum, the aposporic egg cell is
understood (e.g., Knox and Heslop-Harrison activation in sexual species (Mogie 1992;
1963;Knox 1967;Frost and Soost 1968;Cox and Peacock 1992;Koltunow 1993;Grossniklaus et
Ford 1987; Hussey et a!' 1991). al. 1998a).
development without entering meiosis or after 1989; Murray 1992) may ensure a strict
premature meiotic abortion and nuclear sequential order of developmental steps
restitution (Crane, Chap. 3). Likewise, during sexual reproduction. In apomicts,
parthenogenetic embryogenesis is usually developmental checkpoints and feedback
initiated prior to an thesis and often before mechanisms may be ignored or altered,
fertilization of the central cell in pseudogamous leading to the initiation of a developmental
species. Thus, it appears as if specific event before the completion of an earlier one
developmental events are initiated prior to (Koltunow 1993).
completion of the previous ones. Heterochronic
Alternatively, rather than misexpression of
development is a hallmark of apomixis, with
regulatory genes in the nucellus, more general
which specific developmental events are
changes in the cellular machinery could cause
replaced asynchronously or coexist and
apomixis. For instance, an increase in the
compete with events occurring in normal
duration of the cell cycle may allow genes to
sequence.
be expressed at an earlier time in development
In addition to a change in the temporal order than usual. Such a situation has been observed
of developmental processes there is also a for genes with large introns in Drosophila. The
relaxed constraint on cell fate decisions. genes kntrps (klli) and knirps-related (kllrl)
Whereas in sexual species a single nucellar cell encode highly similar proteins, but knrl
is usually committed to the meiotic pathway, contains a large 19 kb intron (Nauber et al.
several nucellar cells in apomictic species have 1988; Oro et al. 1988; Rothe et al. 1989). Thus,
the potential to form unreduced gametophytes. knrl is only functional at nuclear division cycle
The regulation of individual developmental 13 during cleavage, when the cell cycle has
events appears to be conserved between become long enough to allow RNA-
apomictic and sexual pathways. Therefore, it polymerase to transcribe the entire knrl
is likely that key regulatory genes playing transcription unit before the initiation of M-
essential roles in sexual development are phase (Rothe et al. 1992). In contrast, kni is
misregula ted in ei ther space and / or time expressed already at nuclear division cycle 9.
leading to the developmental alterations An intron-Iess knrl gene can fully rescue the
observed in apomicts. Precocious initiation of kni embryo lethal phenotype (Rothe et al.
megagametogenesis and the premature 1992). Two mutants have been isolated that
activation of the egg cell could be caused by allow for the functional substitution of kni by
misexpressed regulatory genes that perform knrl. Both act by lengthening the cell cycle and,
the same functions during sexual reproduction. thus, allow knrl to be transcribed at an earlier
Thus, the gene(s) controlling apomixis does not stage of development than usual (Ruden and
necessarily encode altered gene products, but [ackle 1995). A similar situation may be
rather could be under relaxed or aberrant encountered in apomictic species, with
temporal and / or spatial control. duplicated genes being activated
heterochronically. The duplicated genes may
Several models accounting for the precocious
be paralogs present in the same genome or
induction of developmental events and the
orthologs from two genomes brought together
interrelationship with the sexual pathway have
through hybridization.
been proposed (Peacock 1993;Koltunow 1993).
Developmental checkpoints similar to the ones The models discussed above do not take into
proposed to control proper progression account the tight association of apomixis with
through the cell cycle (Hartwell and Weinert polyploidy, although they are certainly
F.... Sell.lily t. Apomixis: Moloaolar ... Geeeti<......... 179
compatible with it. Several models that observed in apomicts and other reproductive
consider the relationship between polyploidy anomalies. An important aspect of this theory
and apomixis have been put forward (Nogler is that apomixis results from the conflicting
1982; Mogie 1992; Noirot 1993; Carman 1997; action of genes that usually play a regulatory
Grimanelli et al., Chap. 6). For instance, an role during sexual development, i.e., the same
alteration of cell cycle length as hypothesized wild-type genes (not mutant forms) control
above may be caused by polyploidization or both sexuality and apomixis. This reinforces
wide hybridization. In many species with the need for a better understanding of the
isolates of several ploidy levels, diploids are molecular and genetic basis of sexual
usually sexual and polyploids apomictic reproduction for the engineering of apomixis
(Asker and JerJing 1992; Leblanc et al. 1995). in sexual crops.
Autoploidization of sexual diploids has
Another consideration that could influence
resulted in apomictic tetraploid plants in some
research strategies was raised by Jefferson and
species (Burton 1992). It is attractive to
Bicknell (1996).At present, there is no evidence
speculate that the cell cycle length is altered in
to indicate that apomixis is controlled by a
response to changes in ploidy. However, no
trails-acting gene product rather than by a eis-
experimental data is available to support this
acting element. One can envision an alteration
hypothesis since essentially nothing is known
of a eis-acting element, for instance a binding
about the regulation of the cell cycle during
site for a trails-acting factor (e.g., for a
reproductive development in plants. The
transcription factor or chromatin component),
association of polyploidy with apomixis may,
with altered affinity or copy number that could
however, be a secondary effect caused by
cause apomixis by changing the concentra tion
deleterious mutations that accumulated in the
of the trails-acting factor in the cell. For
genome of apomicts. This is supported by the
example, the factor could be titrated out by an
recent isolation of diploid apomicts in
increase in the copy number of its binding sites,
Hieracium and Allium (Bicknell 1997; Kojima
which in turn would result in precocious or
and Nagato 1997). These and earlier findings
inappropriate development of the embryo sac.
suggest that polyploidy is not an absolute
Thus, a dominant locus could readily be
requirement for apomixis (Savidan 1980;
explained. The recent observation that large
Nogler 1982;Hashemi et al. 1989).
genomic regions that are associated with the
An attractive hypothesis, which takes both inheritance of apomixis are not present in
developmental and genomic peculiarities into sexual relatives (Ozias-Akins et al. 1998;Roche
account, has recently been proposed by et al. 1999)is consistent with such a mechanism.
Carman (1997; Chap. 7) based on earlier
models put forward by Ernst (1918). In short, Genetic Control of
the duplicate-gene asynchrony hypothesis Reproduction and Candidate
states that duplicate sets of genes regulating Genes for the Engineering of
reproductive development exist in polyploids.
Apomixis
Polyploidy originally arose through
To date, no fully apomictic mutants have been
hybridization, such that the regulatory control
recovered in sexual species, however, several
of development originating from the two
mutants and spontaneously occurring
genomes may not be in synchrony. The
variations of sexual reproduction display
resulting intergenomic regulatory conflict may
individual components of apomixis. These
then lead to the developmental aberrations
include the production of unreduced spores
180 lIoi Gm.......
Arabidopsis, relatively little is known about the Roeder 1995). In yeast, a large body of
genetic control of megasporogenesis. Recently, knowledge on the molecular mechanisms
the isolation of 270 Arabidopsis mutants with controlling meiosis has been amassed. Many
defective spore development (megasporo- yeast mutants that regulate the entry into
genesis-defective, msd) was reported (Schneitz meiosis and differentiate between meiotic and
et al. 1997). Mutants of the msd class do not mitotic division have been isolated. These
produce a megagametophyte, however, mutants share some characteristics with
sporophytic ovule development proceeds apospory or diplospory of the Antennaria type
normally. The developmental defects during (Koltunow 1993), and their plant homologs
megasporogenesis have not been charac- could be instrumental in the engineering of
terized in detail. All of these mutants also affect apomixis.
microsporogenesis and, therefore, are male
Many genes that play roles in yeast meiosis
and female sterile. They may affect meiosis per
have been characterized, generally by
se rather than female specific processes.
identifying mutants with specific meiotic
Usually, only a single archesporiaI cell, and defects and studying the level of transcripts
consequently a single megasporocyte, of the corresponding genes during meiosis
differentiates in an ovule. However, the (Mitchell and Bowdish 1992; Mitchell 1994).
occurrence of multiple megaspore mother cells These meiotic genes act at different stages of
in some species (Eames 1961; WaIters 1985; meiosis. Genes acting early in the pathway and
Sumner and van Caseele 1998) and of two regulating the decision between mitotic and
megasporocytes in about 5% of the wild type meiotic divisions are of particular interest.
Arabidopsis ovules suggest that several Among the products of early meiotic genes,
nucellar cells have the potential to enter the the meiotic activator IMEl is a master control
meiotic pathway. Once a cell is committed, it gene required for the expression of the genes
appears to inhibit neighboring cells from acting in the early phase of meiosis (Kassir et
doing the same (Grossniklaus and Schneitz al. 1988;Smith and Mitche1l1989; Mitchell et
1998). This view is supported by a recently al. 1990; Kawaguchi et al. 1992). To be
identified mutant in maize. Plants functional, IMEl has to become
homozygous for multiple archegonial eel/51 phosphorylated by RIMl1 (Bowdish et al.
(mac1) contain between three and 21 1994). Upon phosphorylation, the early
megasporocytes in a single ovule (Sheridan meiotic genes are activated and a starved
et al. 1996). Thus, mac1 is only likely to be diploid cell undergoes meiosis to produce four
involved in megaspore mother cell haploid spores. In the fission yeast
determination. The phenotype shows certain Schizosaccharomuces pombe, the Mei3 gene is
similarities to apospory, in which multiple induced by nutrient deprivation. Mei3 inhibits
aposporic initials form around the sexual the protein kinase Pail, that then triggers the
megaspore mother cell. However, unlike in entry into meiosis (reviewed by Yamamoto
apomicts where microsporogenesis is usually 1996). The Patl kinase, in turn, represses the
unaffected, mad mutants also show abnormal Mei2 protein, which is an essential positive
male sporogenesis (Sheridan et al. 1999). factor for entry into meiosis. Thus, regulatory
networks involving phosphorylation and de-
The genetic regulation of meiosis has been
phosphorylation events responding to
extensively studied in maize and the yeast
environmental signals play a crucial role in the
Saccharomyces cereoisiae (Golubovskaya 1979;
commitment to the meiotic pathway.
Golubovskaya et al. 1992; Mitchell 1994;
182 UdGr.........
absolutely essential for egg activation. Case in Riley 1%3; Turcotte and Feaster 1963; Chase
point: the introduction of calcium into the egg 1969) and after induction in isolated ovaries
is sufficient to induce parthenogenetic and ovules. Certain genetic backgrounds and
activation of sea urchin and mammalian eggs mutants produce a high percentage of
(Steinhardt and Eppe11974;Uranga et al. 1996). parthenogenetic haploids in their progeny. A
Recently, in vitro injection experiments with better understanding of the genetic basis of
mouse oocytes have shown that a truncated c- parthenogenetic development will prove
ki t receptor tyrosine kinase leads to instrumental for the engineering of apomixis.
parthenogenetic egg activation that requires In maize, parthenogenetic haploids are
both calcium and phospholipase C activity produced spontaneously at a frequency of
(Sette et al. 1997). about 0.1% (Chase 1969),and as high as 3.2%
in certain genetic backgrounds (Coe 1959;
In plants, our understanding of the events
Sarkar and Coe 1966). In plants homozygous
following fertilization is very limited.
for ig, the frequency of maternal haploids is
Experimentation with angiosperm gametes
increased fourfold to 0.6% as compared to
has been difficult because of their
isogenic lines homozygous for the wild type
inaccessibility and the complex milieu within
Ig allele (Kermicle 1969). Interestingly, ig also
the megagametophyte where double
conditions the production of patroclinous
fertilization occurs (Russell 1993; Dumas and
(androgenetic) offspring at a frequency of more
Mogensen 1993). Specific fusion of isolated
than 2% as compared to 0.001% (1/80,000) in
maize gametes has recently been achieved and
wild type stocks (Chase 1963; Kermicle 1%9,
offers great promise for the study of molecular
1994). Thus, ig not only permits extra rounds
and cellular events underlying fertilization and
of mitosis in the embryo sac (Lin 1978, 1981;
egg activation in vitro (Faure et al. 1994;Dumas
Huang and Sheridan 1996)but it also promotes
and Faure 1995;Tirlapur et al. 1995;Kranz and
the formation of embryos in the absence of
Dresselhaus 1996;Rougier et al. 19%). One of
karyogamy. Extremely high frequencies of
the first visible changes after fertilization of an
haploid production are found in barley plants
isolated maize egg cell is the formation of a
homozygous for haploid initiator (hap) and in
cell wall (Kranz et al. 1995).In the brown algae
the Salmon system of wheat. The hap mutation
FUCIIS, cell wall secretion depends on an
conditions the formation of 30% haploid
increase in the cytosolic calcium concentration
embryos when homozygous (Hagberg and
(Roberts et al. 1994; Roberts and Brownlee 1995;
Hagberg 1980;Asker et al. 1983).The sperm is
Belanger and Quatrano 2000). A transient rise
prevented from fertilizing the egg by an
of the cytosolic calcium concentration has
unknown mechanism whereas the central cell
recently been reported for the first time in
gets fertilized normally to produce the
angiosperms after in vitro fertilization of a
nutritive endosperm (Mogensen 1988).
maize egg cell (Digonnet et al. 1997; Antoine
et al. 2000). These observations suggest that Salmon wheat lines can produce up to 90%
plant and animal egg activation may be similar, parthenogenetic haploids (Matzk 1995). Both
however, the role played by the release of cytoplasmic and nuclear determinants are
calcium in the fertilized plant egg and what involved in parthenogenetic activation: the
events it triggers are currently unknown. presence of the wheat-rye translocation
chromosome IBL-IRS in the Aegilops
Parthenogenetic egg activation is a key
cytoplasm of cat/data or kotschyi Salmon leads
component of apomixis and occurs in many
to haploid production, whereas either the
plant species, both spontaneously (Kimber and
translocation or the Aegilops cytoplasm alone
186 UtI Gn.......
the developmental program for endosperm only a few division cycles or does not divide
development is activated in the absence of at all. In many dicotyledonous species,
fertilization. including Arabidopsis, the endosperm forms
but is essentially degraded by the time seed
It is possible that the same genetic control is
maturation is initiated. In contrast, the
responsible for autonomous endosperm and
endosperm of cereals is persistent and of great
egg activation. This would be in agreement
economic value. Therefore, an engineered
with the hypothesis that the endosperm is
apomixis in grain crops will have to allow for
evolutionary and derived from a second
normal development of the endosperm. In an
embryo, as first proposed by Sargant (1900).
ideal situation, endosperm formation in
This hypothesis is supported by morphological
engineered apomictic crops could be induced
analyses of fertilization in nonflowering seed
autonomously. However, successful formation
plants of the genera Ephedra and Gnetllm
of endosperm in cereals depends on the
(Friedman 1990, 1992; Carmichael and
specialized cytoplasm of the central cell and
Friedman 1995). In addition, molecular and
requires contributions from both maternal and
genetic investigations have shown that there
paternal genomes. This may be because some
is a large overlap in gene activity between the
genes are imprinted, that is their activity
endosperm and embryo. For instance, among
depends on their parental origin (Kermicle
the 855 characterized defective kernel mutants
1970; Kermicle and Alleman 1990; Messing
in maize (representing about 285 loci), the vast
and Grossniklaus 1999). Thus, engineered
majority affect both the embryo and
apomictic grain crops are likely to require the
endosperm and very few are potentially
fertilization of the central cell. The vast
endosperm-specific (Neuffer and Sheridan
majority of apomictic Gramineae are
1980).Similar results were obtained in a study
pseudogamous; possible autonomous
of defective kernel mutants in barley (Bosnes et
apomixis has been observed in very few
al. 1987), suggesting tha t a very large
species, including Calamagrostis, Poa neroosa
percentage of seed-speeific genes are expressed
and Nardus stricta (Iohri et al. 1992).
in both embryo and endosperm, despite their
different development and physiology. In 2. Genomic imprinting. Imprinting in plants
contrast to this extensive overlap in gene is usually regarded as specific to the
expression between embryo and endosperm, endosperm (Kermicle and Alleman 1990;
studies on several recently isolated Arabidapsis Walbot 1996; Alleman and Doctor 2000).
mutants that allow fertilization-independent Formation of androgenetic and gynogenetic
endosperrn formation but not embryogenesis haploids in many species (Kimber and Riley
suggest that endosperm activation may be 1963;Sarkar and Coe 1966;Kermicle 1969)and
controlled by different developmental of asexually derived embryos in apomicts
programs (Ohad et al. 1996;Chaudhury et al. suggest that imprinting does not play a crucial
1997; Grossniklaus and Vielle-Calzada 1998; role for embryogenesis in these species,
Luo et al. 1999;Kiyosue et al. 1999;D. Page, R. although it may exist in ".thers. The
Pru itt, S. Lolle, and U. Grossniklaus, development of embryos from somatic tissue
unpublished data). (Zimmerman 1993;Mordhorst et al. 1997) and
through anther culture (Zaki and Dickinson
The importance of the endosperm for seed
1990) is taken as further evidence tha t
development varies among species depending
imprinting is not involved in plant
on its developmental pattern. In some species
embryogenesis. However, the initial
of the Orchidaceae the endosperm undergoes
188 UoIGns......
development in Tripsacuni is normal under a abortion (Grimanelli et al. 1995; Dujardin and
wide range of ratios of maternal to paternal Hanna 1989) that is likely to result from a
genomes. Similarly,apomictic Paspaluni species genomic unbalance in the endosperm
are insensitive to an imbalanced genome ratio (Grossniklaus et aI. 1998a;Morgan et aI. 1998).
in the endosperm while the sexuals maintain Crosses with pollen donors of higher ploidy
this requirement (Quarin 1999). that maintained the endosperm balance
number could possibly restore fertility. At
Altered modes of fertilization that are expected
present, our understanding of imprinting and
to maintain the endopserm balance number
its importance for seed development is very
have been reported in several cases. This can
limited. A sustained effort toward a better
be achieved if either both sperm cells delivered
understanding of the genetic and molecular
by the pollen tube fuse with the central cell, or
basis governing imprinting is required to
if only one of the two polar nuclei and a single
overcome the potential constraints to the
sperm nucleus participate in karyogamy
engineering of apomictic crops.
(Rutishauser 1954; Reddy and d'Cruz 1969;
Nogler 1972,1984a).Alternatively, unreduced
Genetic Screens for Mutants
pollen could serve as the male parent (Chao
1980). As a very successful alternative to
Displaying Apomictic Traits in
relaxed imprinting requirements, many Sexual Model Systems
apomictic grasses show apospory of the In previous sections, I discussed a number of
Panicum-type, where 4-nucleated embryo sacs genes that control sexual development that
are formed, which most often contain only one could serve as powerful tools for the
polar nucleus that fuses with a single sperm engineering of apomixis. As an alternative, a
nucleus (e.g.,Savidan 1980). Sexual individuals screen for mutants that display apomictic traits
of these agamic complexes usually produce 8- in a sexual species could directly lead to the
nucleated Polygonum-type embryo sacs with identification of key regulatory components
two reduced polar nuclei. Thus, fertilization (Peacock 1992).This approach has been taken
of both sexually and apomictically derived in several laboratories using Arabidopsis as a
central cells produces endosperms with model system. Although no apomictic species
balanced parental genomes (Reddy 1977). have been described in this genus, the close
relative Arabisholboelli is apomictic (Asker and
Apomictic species may be evolutionarily
]erling 1992). While direct experimentation
derived from predisposed genera that had
with Arabis is difficult because of its poor
relaxed imprinting requirements or,
genetic characterization and long generation
alternatively, evolved specific adaptations of
time, its close relationship to Arabidopsis may
the fertilization mechanism that maintained
be useful for comparative and wide
the imprinting requirements (see also
hybridization approaches.
Grimanelli et al., Chap. 6). Such predis-
positions and adaptations are not thought to Arabidopsis Mutants with
exist in most sexual species and imprinting Autonomous Seed Development
may pose a serious problem to the introduction Screens for Arabidopsis mutants that allow seed
of apomixis into sexual crop plants development in the absence of fertilization
(Grossniklaus et al. 1998a; SpilIane et al. 2000; have been performed in several laboratories.
Savidan 2000; Grossniklaus et al. 2001).lndeed, These screens take advantage of male sterile
apomictic maize- Tripsacum and pearl milIet- mutants and aim to identify second site
Pennisetum hybrids show a high degree of seed mutations that pseudo-suppress sterility. In
190 UoIGro........
Arabidopsis, unpollinated pistils do not cellular stage in fisl and fis2. The seed coat
elongate, so that pistil elongation is an easily develops properly but no embryos form
scored phenotype correlated with seed (Chaudhury et al. 1997). Structures that
development. Different male sterile mutants resemble an elongated zygote have been
have been used, including pistil/ata (pi), a observed in fisl and fis2 at a low frequency. In
homeotic flower mutant that lacks stamen fie/fis plants, the mutant allele is either very
(Chaudhury and Peacock 1993; Koltunow et poorly transmitted or not transmitted at all
al. 1995; Chaudhury et al. 1997); the wax through the female gametophyte and can only
biosynthetic mutant eceriferumb (cer6) (Dellaert be recovered through the pollen, because
1979;Preuss et al. 1993;Hulskamp et al. 1995), pollinated seeds derived from a mutant
a conditional male sterile that is fertile under megagametophyte abort (Ohad et al. 1996;
high humidity (Ohad et al. 1996); and the Chaudhury et al. 1997). Thus,fie/fis mutants
temperature sensitive mutant TH154, isolated display a gametophytic maternal effecton seed
in R. Pruitt's laboratory, which is male sterile development, which is similar to the
at 25DC but fully fertile at 18DC (D. Page, R. phenotype observed for mea that also shows
Pruitt, S. Lolle, and U. Grossniklaus, fertilization-independen t end os perm
unpublished results). Silique elongation under development (Grossniklaus et al. 1998b;
the restrictive condition indicates an asexual Grossniklaus and Vielle-Calzad a 1998).
mode of reproduction with full or partial seed Indeed,fisl and two other mutants were found
development, or the development of a fruit to be allelic to mea (Kiyosue et al. 1999; Luo et
without concomitant seed production al. 1999),as were fis3 and fie (Chaudhury et al.
(parthenocarpy). 1997;Ohad et al. 1999).
Using this type of screen with more than 15,000 MEA and FIE encode members of the
M1 plants, we identified three classes of Polycomb group, proteins thought to regulate
mutants (D. Page, R. Pruitt, S. Lolle, and U. gene expression by modulating higher order
Crossniklaus, unpublished results): (i) mutants chromatin structure (Grossniklaus et al. 1998b;
suppressing the male sterility defect, which can Ohad et al. 1999). FIS2 encodes a putative DNA
easily be identified because they produce binding protein with a Zn-finger domain (Luo
functional pollen (some of these will be true et al. 1999).As with their animal counterparts,
revertants of TH154); (ii) mutants displaying the MEA and FIE proteins interact directly in
parthenocarpy; and (iii) mutants displaying a protein complex (Luo et al. 2000;Spillane et
apomictic traits with autonomous al. 2000;Yadegari et al. 2000).The function and
development of seed-like structures in the regulation of MEA, FIE, and FIS2 have been
absence of pollen production. The latter are extensively reviewed (Good rich 1998;Ma 1999;
rather rare and, to date, mutants in only three Preuss 1999;Mora-Garcia and Goodrich 2000;
loci have been reported. These have been Russinova and de Vries 2000; Grossniklaus et
characterized in more detail at the genetic and al. 2001) and will be summarized only very
morphological level. In the Jertilization- briefly. MEA was shown to be regulated by
independent endosperm (fie), mutant endosperm genomic imprinting, thus explaining its
develops autonomously to the free nuclear maternal effect on seed development (Vielie-
stage, the seed coat develops normally, but no Calzada et al. 1999; Kinoshita et al. 1999). While
embryo forms (Ohad et al. 1996). Likewise, FIS2 also seems to be regulated by genomic
fertilization-independent seed mutants (fisl,fis2, imprinting (Luo et al. 2(00), FIE appears to be
fis3) form autonomous endosperm, which was expressed biparentally later during seed
shown to be diploid and can progress to the development (Luo et al. 2000; Yadegari et al.
2000). At earlier stages, a FIE::GUS fusion fertilizing reproductive behavior of
protein was not expressed from the paternal Arabidopsis, however, makes such screens labor
allele, as was observed for a large number of intensive because they are based on
genes expressed early during seed outcrossing and scoring the progeny for
development (Vielle-Calzada et al. 2000), but exclusively maternal inheritance. In maize,
paternal FIE::GUSactivity becomes detectable similar screens are greatly facilitated by the
later on (Luo et al. 2000), suggesting that it is natural outcrossing mode of reproduction, the
regulated in a manner different from MEA. availability of embryo-specific markers that
can be scored on whole kernels, and the
Although the mea.fie, andfis2 mutants do not
multitude of genetic tools available to the
produce fertile asexual seeds, they show
geneticist. Over the last few years, I have
characteristics of autonomous reproduction. It
developed maize stocks to perform a genetic
is possible that some of the yet uncharacterized
screen aimed at isolating mutants with
mutants with fertilization-independent
characteristics of pseudogamous apomictic
endosperm and maternal effect seed abortion
reproduction.
will not only allow the autonomous
e iosperm formation, but also full or partial The screen is based on the Mutator transposon
c.nbryogenesis. Alternatively, the mea.fie, and system, which is used as a potent mutagen and
fis2 mutants can be used as the starting material allows for easier cloning of newly isolated
for second site modifier screens that could mutants (Chomet 1994). The genetic screen
eventually lead to the identification of exploi ts the imprinting barrier for endosperm
additional mutants that allow the formation development and an easily scored embryo and
of fertile seeds. endosperm marker. Lines that are
homozygous for a recessive allele of the red
Screen for Pseudogamous Apomixis in
Cereals color or r locus and have Mutator activity serve
Although the engineering of autonomous as female recipients. An allele of the r locus,
apomixis in which both embryo and which pigments both the crown of the kernel
endosperm develop without fertilization (aleurone layer of endosperm) and the embryo
(R-nj, Figure 12.3,p. 197),is usedas a dominant
might be considered the ideal situation, it may
be extremely difficult to achieve in cereals. As pa ternal marker. Screens aimed independently
outlined earlier, autonomous apomixis is rare at nonreduction and parthenogenesis have
in nature and the imprinting requirements for been devised. Both rely on the R-nj paternal
successful endosperm development pose a marker gene in tetraploid configuration. If a
serious barrier to the introduction of mutation causes nonreduction during meiosis,
autonomous apomixis into sexual crops. the resulting central cell carries four maternal
Therefore, the engineering of apomixis in genomes, and only endosperms receiving two
sexual species will have to target a paternal genomes (from the tetraploid pollen
pseudogamous mode of apomictic donor) maintain the endopserm balance of
reproduction. A search for apomictic mutants 2m:1p (here 4m:2p) and develop normally. All
in sexual model systems that allow fertiliza tion seeds derived from a reduced female
of the central cell may be more appropriate gametophyte will abort such that only kernels
than screens for fully autonomous apomicts. derived from a nonreduced embryo sac will
develop (or a seed, where the requirement for
Screens for Arabidopsis mutants displaying a balanced genome ratio in the endosperm has
pseudogamous apomixis have been proposed been abo lished ). Thus, the imprinting
(Chaudhury and Peacock 1993). The self- requirement provides a powerful selection
192 UtI GrM.......
system to identify mutants that lead to the sporogenesis has been identified (Siddiqi et al.
formation of an embryo sac from an unreduced 2000; Motamayor et al. 2(00) and attempts to
cell lineage. In addition, full-sized kernels can isolate genes that regulate the developmental
bescored for absence of the dominant paternal events initiating female gametogenesis and
R-nj marker in the embryo, which indicates embryo development are just beginning. As
parthenogenetic development. an alternative to the isolation of mutations, we
identify genes expressed specifically during
Because obtaining a mutation that causes both
megasporogenesis and megagametogenesis.
nonreduction and parthenogenesis may be
The inaccessibility of the developing embryo
extremely difficult, a second screen aimed at
sac and the small number of cells involved
the isolation of parthenogenetic mutants has
make this a difficult undertaking using
been developed. It makes use of the el1
conventional molecular methods such as
mutation, which, when homozygous, produces
differential screening techniques. Therefore,
a large fraction of unreduced embryo sacs. In
we use a novel technology, enhancer detection,
el1 homozygotes, independent assortment
which allows the identification of
during meiosis I is not affected and the
developmentally regulated genes based on
resulting gametes are not genetically identical
their pattern of expression. Enhancer detection
(Roades and Dempsey 1966). Nevertheless, it
is one of the most powerful tools to identify
provides a reliable source for unreduced female
tissue specifically expressed genes and their
gametes. Lines homozygous for r1 and el1 and
regulatory sequences. Application of this
displaying high Mutator activity have been
approach in angiosperms will lead to the
constructed to serve as female recipients.
identification of many genes that control
Kernels derived from crosses with a tetraploid
gametogenesis and cellular differentiation in
R-nj pollen donor can be scored for rejection of
the female gametophyte. In addition, it will
the R-nj marker, i.e., for the maternal
identify many cell type- and tissue-specific
phenotype, in order to identify embryos that
regulatory regions that will be required to
developed without a paternal contribution.
express candida te genes in a precise temporal
Such embryos will be diploid (through the
and spatial fashion.
action of el1), which greatly facilitates
subsequent genetic characterization. These Enhancer Detection and
genetic screens aimed at the isolation of Gene Trap Systems
mu ta tions tha t lead to nonred uction, Enhancer detection was first developed in the
parthenogenesis, or a combination of both fruit fly Drosophila rnelanogaster and relies on a
aspects will provide useful material to further mobile genetic element carrying a reporter
our understanding of these processes at the gene under the control of a weak constitutive
molecular and genetic level. promoter (O'Kane and Gehring 1987). This
minimal promoter is usually not active but
Enhancer Detection as a ideally it can be activated in all tissues and at
Powerful Tool to Study all developmental stages. If it comes under the
control of genomic eis-regulatory elements
Sexual Reproduction in
such as enhancers, the reporter gene is
Arabidopsis expressed in a specific temporal and spatial
The molecular and genetic bases of pattern (Figure 12.4, p. 197). This pattern
megasporogenesis and megagametogenesis reflects the expression of a nearby gene
are poorly understood. To date, only one controlled by the same regulatory elements
female-specific mutant that affects and, thus, allows the identification of genes
f_ Se..lllly .. ,,_.ds: ......... Geotil ApproodIos 193
based on their pattern of expression rather than Enhancer detection has some important
on a mutant phenotype (Bellen et al. 1989;Bier advantages over classical genetic screens for
et al. 1989;Grossniklaus et al. 1989; Wilson et the identification of genes required in the
al. 1989).Enhancer detector screens have been gametophytic phase of the life cycle: (i) many
extremely successful in Drosophila genes that encode components of the basic
developmental genetics and similar cellular machinery display a gametophyte
approaches were rapidly adapted to other lethal phenotype if disrupted. Essential genes
model systems. Because of the large are expected to show widespread although not
intergenomic regions in mice, gene traps were necessarily ubiquitous expression, whereas
developed that depend on a modification of expression in particular cell types of the
this approach involving the generation of megagametophyte suggests a function in cell
transcriptional fusions to the reporter gene specification and differentiation; (ii) genes
(Gossleret al. 1990;Skames 1990;Friedrich and required for both micro- and megagameto-
Soriano 1991). In Arabidopsis, similar systems genesis can be isolated, because a large
based on T-DNA insertional mutagenesis percentage of enhancer detector insertions do
(Topping et al. 1991; Fobert et al. 1991; not disrupt gene function. Mutants affecting
Kertbundi t et al. 1991) or the Ac/ Os both male and female gametophytes can
transposable element system from maize usually only be recovered as rare partially
(Sundaresan et al. 1995; Springer et al. 1995; penetrant mutations or in genomic regions that
Smith and Fedoroff 1995) have been can be covered by a duplication (Vollbrecht
developed. 1997); (iii) enhancer detection is the only
efficient technique that allows the
Enhancer detection and gene trap systems
identification of genes expressed in very few
offer an added benefit: they allow the
or even single cells. By focusing on the cells
identification of genes that are not readily
and tissues where a gene is expressed, subtle
amenable to classical genetic analyses (Bellen
phenotypes can be identified that may not
et al. 1989; Bier et al. 1989; Grossniklaus et al.
easily be recognized in phenotypic screens;
1989; Wilson et al. 1989). They have been
and (iv) most importantly, enhancer detector
especially useful in studying developmental
and gene trap transposons greatly facilitate the
processes occurring late in development, i.e.,
molecular cloning of genomic sequences
after the effective lethal phase of a
flanking the insertion site. In addition, they
corresponding mutation. For example, a gene
allow a detailed genetic analysis of the detected
that is required for essential steps during both
gene through remobilization and the recovery
embryo and ovule development would be
of additional alleles, regional chromosomal
identified as an embryo lethal mutation and
rearrangements, and revertant sectors (e.g.,
its function during ovule development would
Grossniklaus et al. 1992; Springer et al. 1995;
be masked. Enhancer detection allowed
Tsugeki et al. 1996;Grossniklaus et al. 1998b).
identification of the first embryo lethal genes
in Drosophila that are also required for Generation of Transposants and
oogenesis or eye development (Grossniklaus Ongoing Screens
et al. 1989;Mlodzik et al. 1990).The dissection To identify genes involved in female
of processes characterized by functional gametogenesis, we have generated nearly
redundancy and high complexity is also 4,300 lines carrying randomly inserted
greatly facilitated by enhancer detection enhancer detector or gene trap transposons (U.
(Wilson et al. 1990;Bellen et al. 1990). Crossniklaus, J. Moore, W.Gagliano, J-P. Vielle
Calzada, unpublished data). We are using the
194 UoIGm.......
system developed by Sundaresan et al. (1995), identified with a restricted expression pattern
which is based on the Ac / Os transposon of in young ovule primordia O-P. Vielle Calzada
maize and allows the recovery of unlinked and U. Grossniklaus, unpublished data).
transposi tion events throughou t the Many of the expression patterns reflect the
Arabidopsis genome. In brief, an enhancer highly polar organization of the ovule and
detector or gene trap transposon is mobilized may be involved in establishing or interpreting
by crossing a starter line homozygous for Os positional information. Other patterns are
to a line carrying a stable Ac transposase associated with the meiotic cell lineage and
source. Self-pollination of the F1plants, which are of particular interest to the engineering of
contain both the Os starter locus and the Ac apomixis. The second screen is aimed at the
transposase construct, results in some F2 identification of genes expressed in individual
progeny carrying a transposed Os element cell types of the embryo sac and concentrates
(transposants). By positively selecting for the on late stages of ovule ontogeny, including
presence of Os but negatively against the post-fertilization stages (R. Baskar, J. Moore,
donor Os locus and Ac, unlinked stable W. Gagliano, U. Grossniklaus, unpublished
transposition events can be recovered data). Some of these patterns are specific to
(Sundaresan et al. 1995). Negative selection the individual cell types of the embryo sac
against the donor locus ensures the recovery incl uding the egg cell and will serve as
of unlinked or loosely linked transposition importan t tools to direct expression of
events, a prerequisite for genome-wide candidate genes to the megagametophyte. The
random insertional mutagenesis. Similar results of this screen are discussed in more
strategies are currently being developed for detail in the next section.
rice (Chin et al. 1999; R. [efferson, personal
Independently, our transposant library is
comm.).
being used for classical genetic screens to
Using six independent starter lines, we isolate insertional mutants that affect fertility.
generated approximately 45,000 F1s,of which The first phenotypic screen identifies mutants
more than 35,000 were grown to maturity to that disrupt the development or function of
harvest their F2 seeds. About 23,000 of the F2 the female gametophyte. These mutants are
families have been put through the positive/ identified in a two-step screen for reduced
negative selection process to recover fertility and a non-Mendelian segregation
transposants. Between 20% and 25% of the F2 ratio, both indicating a gametophytic defect
families yielded an unlinked transposition (Moore et al. 1997; Feldmann et al. 1997;
event. The transposant library of about 4,300 Howden et al. 1998; Christensen et al. 1998;
lines that we generated serves as the basis for Bonhomme et al. 1998; Grini et al. 1999). In a
four large-scale screens aimed at identifying second screen we identify families segregating
genes involved in female reproduction. sterile plants that show a sporophytic
requirement. Of 3,200 families that were
Two of the screens we are performing are
screened, nearly 40 sterile mutants were
designed to identify genes expressed during
identified (Vielle-Calzada et al. 1998).
ovule and megagametophyte development.
Reciprocal outcrosses to wild type showed that
The first one targets early ovule development,
most of these mutants are male sterile or affect
which encompasses the key events of
both sexes, but six were found to be female-
megasporogenesis (Vielle-Calzada et al. 1998).
specific. These recessive female sterile mutants
More than 1,000 transposants have been
affect ovule morphogenesis or megasporo-
screened and about 30 lines have been
genesis and are currently being analyzed in can be visualized by histochemical staining
more detail at the molecular and genetic level (lefferson et al. 1987;Kavanagh et al. 1988).To
(J-P. Vielle-Calzada and U. Grossniklaus, identify genes expressed during ovule
unpublished data). development and female gametogenesis, we
analyzed GUS expression in maturing ovules
For the rapid isolation and sequencing of
of about 2,300 transposants (R. Baskar, J.
genomic regions flanking the Os insertion we
Moore, W. Gagliano, U. Grossniklaus,
adapted a PCR-based procedure, TAIL-PCR
unpublished data). Between 9% and 10% of
(Liu et al. 1995),to be used in conjunction with
the enhancer trap lines and 2% to 3% of the
Os elements (see also Tsugeki et al. 1996).
gene trap lines show spatially restricted GUS
Using a set of Os-specific primers
expression in mature ovules. Approximately
(Grossniklaus et al. 1998a)in combination with
half of these enhancer detector lines show
arbitrary primers, we isolated at least one
expression restricted to sporophyte and
flanking fragment for more than ISO lines that
gametophyte, respectively, whereas very few
we identified in various screens. Using two
show regional expression in both
sets of arbitrary primers, the success rate was
gametophytic and sporophytic tissues.
greater than 95% (Grossniklaus et al. 1998a).
We have identified insertions into a multitude Although many Arabidopsis promoters have
of genes tha t encode essential proteins been found to be highly compact (e.g., Dwyer
involved in basic metabolic and cellular et al. 1994;Thoma et al. 1994; Xia et al. 1996),
processes, putative regulatory proteins, and enhancers that drive reporter gene expression
several novel sequences of unknown function. could be at a considerable distance from the
Based on sequence information, the majority site of insertion. This makes the isolation of
of the detected genes appear to be involved in the detected gene and its promoter more
gene regulation and signaling processes. laborious, requiring that the expression
pattern of an isolated gene be confirmed. To
Identification of Developmentally
Regulated Genes and Their Promoters date, we have analyzed the expression of three
Very few genes expressed in the genes expressed in the megagametophyte; in
megagametophyte have been described situ hybridization indicates that they are
(Nadeau et a1.1996; Belostotsky and Meagher indeed expressed as expected (Vielle-Calzada
1996),and genes expressed in individual cells et al. 2000; R. Baskar, J-P. Vielle-Calzada and
of the embryo sac have not previously been U. Grossniklaus, unpublished data). It would
identified and characterized. Recently, cDNA be preferable to identify gene trap insertions
libraries obtained from isolated egg cell and with cell type-specific expression because they
in vitro fertilized zygotes of maize have been have to be inserted within the transcription
generated, leading to the identification of unit in order to function. However, because
genes expressed in the embryo sac and embryo gene traps must integrate within the gene in
(Dresselhaus et al. 1994, 1996, 1999a,b). the correct orientation (Sundaresan et al. 1995)
Although no cell type-specific genes have been and GUS activity is often weak, the frequency
isolated yet, this approach holds great promise at which highly specific expression patterns
for the identification of embryo sac-specific are recovered is very low. The screening
genes. We use enhancer detector and gene trap process for cell-type specific expression in the
transposons carrying the HidA reporter gene ovule and megagametophyte is extremely
encoding b-glucuronidase (GUS) (lefferson et laborious, requiring preparations for high-
aI.1986;Jefferson 1987). The expression ofGUS resolution light microscopy. Therefore, we
196 UelGro........
Ds element stigma-specific
enhancer
Gene Trapping
stigma-specific Ds element
enhancer
regions controlling these traits are physically sexual crops will pose a serious barrier to the
very large. wee different approaches that rely introgression of apomixis and must be
on class ical breeding are being pursued: (i) addressed at the genetic and molecular levels
introgression of apomixis from wild apomictic to make both introgression and biotech-
relatives through wide crosses (Savidan, Chap. nological approaches possible.
11); (ii) generation of apomicts through
An attractive alternative to introgression
hybridization of sexual progenitors (Carman,
approaches is suggested by Carman, based on
Chap. 7); and (iii) construction of apomixis by
the genome collision theory for the evolution
combining reproductive mutants that display
of apomixis. He proposes that the presence of
certain aspects of apomixis (Asker et al. 1982).
duplicate genomes controlling the
Except for the second strategy, these
reproductive pathway in hybrids cause
approaches are only applicable if wild
. reproductive anomalies such as apomixis. The
apomictic relatives or mutants displaying
theory predicts that hybridization of two
apomictic characters are a vailable for a given
reproductively divergent ecotypes could be
crop spec ies. Thus, a broad introduction of
used to generate apomictic hybrids .
apomixis into sexual crops will depend on
Hybridization as the cause for apomixis was
biotechnology and isolation of the genes that
fi rst proposed by Ernst (191 8), bu t the
manipulate the reproductive system.
mutational hypotheses for apomixis soon
Although apomixis is found in many plant became more popular. The occurrence of
families, it has been described in only a small apomictic hybrids in the offspring of certain
number of agriculturally important species. interspecific crosses that involve only sexual
These include several forage grass crops parents has indeed been reported and is
(Bashaw and Hanna 1990; Savidan 2000; do reviewed by Carman (Chap. 7).
Valle and Miles, Chap. 10), horticultural crops
The genetic synthesis of apomi xis by
such as Citrus, apple, mango, and mangosteen,
combining reproductive mutants that display
as well as orchids (Wakana and Uemoto 1987;
nonrecurrent forms of apomixis has been
Naumova 199 3; Koltunow e t al. 1996).
attempted in barley by Asker et a!. (1982). A
Importantly, apomixis has also been described
mutant that produces a large number of
in relatives of a few important grain crops,
unreduced gametes, tri, was combined with
notably maize, pearl millet, and wheat
hap, a mutant that gives rise to parthenogenetic
(Bashaw and Hanna 1990). Several breeding
haploids at a high frequency. Although the tri
programs have focused on the introgression
mutation is not suitable for the fixation of
of apomixis from wild relatives through wide
heterosis (because of restitution at the second
crosses (Dujardin and Hanna 1989; Petrov et
meiotic division) and no apomictic double
a!. 1994;Savidan and Berthaud 1994; Savidan
mutants have been generated, plants showing
2000; Savidan, Chap 11). Although apomictic
aspects of both phenotypes have been
maize- Tripsacum and pearl millet-Pennisetum
recovered. It should be noted, however, that
hybrids have been generated they are
very few reproductive mutants have been
characterized by a high degree of seed
described to date and that it is essential to
abortion, which is likely caused by an
systematically search for additional mutants
endosperm defect stemming from an
that affect megasporogenesis, partheno-
endosperm imbalance (Grossniklaus et a!.
genesis, and endosperm development. To my
1998; Morgan et a!. 1998; Grimanelli et al.,
knowledge, no sy stematic screens directly
Chap. 6). Imprinting-related phenomena in
targeted at nonreduction or parthenogenesis
have been conducted in angiospenns and the apomixis is underway and will provide us with
mutants described so far have been identified novel tools to manipulate sexuality towards
fortuitously. Well-defined sexual model apomixis. These efforts are being
systems are best suited for such screens. Rather complemented by new screens that specifically
than perfonning large-scale genetic screens for target relevant aspects of reproduction in
reproductive mutants in many different crop sexual species. Several laboratories are also
plants, it may be easier to isolate the relevant trying to isolate plant homologs of yeast
genes from Arabidopsis (or maize) and mimic mutants that could play crucial roles in
the mutant phenotype in crops through genetic determining the meiotic lineage and
engineering. nonreduction. It must be emphasized that our
understanding of the molecular mechanisms
De Novo Engineering of
Apomixis through Biotechnology that control plant reproduction are still
As outlined above, the introduction of extremely limited.
apomixis into a wide variety of sexual crops In addition to regulators of the sexual pathway,
will be most efficiently achieved through many other genes may be useful for the
genetic engineering. To maximize its engineering of apomixis. Such genes include
usefulness and versatility in a bioengineered glucanases that degrade callose, the absence
form apomixis will have to be dominant; of which serves as a consistent indicator of cells
otherwise the fixation of hybrid genotypes will initiating megagametogenesis. Genes
be very slow. The engineering of apomixis promoting cell wall formation could also be
through biotechnology will require a concerted useful to prevent fertilization of the egg cell
effort in three main areas: (i) the identification and promote parthenogenesis. Another
and characterization of candidate genes that important class is made up of genes tha tcontrol
can be used to manipulate the reproductive the cell cycle. Heterochronic or heterotopic
system; (ii) the isolation of promoters that allow expression of such regulators could potentially
a precise control of gene expression at the be used to trigger cell division and initiate
spatial and temporal levels; and (iii) the developmental events such as
development of efficient technologies to megagametogenesis and embryogenesis.
introduce and control transgenes and/or to Recent studies in Arabidopsis have shown that
perform targeted mutagenesis of endogenous misexpression of cyclin1At in root cells can
genes. trigger extra rounds of cell division (Doemer
The identification of regulatory genes that can et al. 19%) and that cyclinD controls the growth
be used to control apomixis is being pursued rate in tobacco (Cockcroft et al. 2(00). It will be
by a variety of approaches using both interesting to see whether similar experiments
apomictic and sexual systems. Insertional can induce cell proliferation in the egg cell.
mutagenesis in Hieracium and Tripsacum Other growth regulators, such as plant
(Bicknell, Chap. 8; Grimanelli et al., Chap. 6) hormones, have not been studied in detail
or positional cloning based on mapping during sexual reproduction, but may play
approaches and comparative genomics will crucial roles in initiating developmental
lead to the identification of the components programs relevant to apomixis.
controlling apomixis in natural apomicts. In Targeted misexpression of candidate genes will
sexual model systems, the characterization and require well-defined regulatory sequences that
molecular isolation of existing reproductive can be used to drive transgene expression. To
mutants that show individual components of date, only a few promoters have been
200 UoIGns.......
described that are active in the ovule, and Field-level Regulation ofApomictic Traits
promoters specific to the gametophyte and To maximize benefits, it will be necessary to
its constituent ~ells have just been isolated (R. control the expression of apomixis such that a
Baskar and U. Grossniklaus, unpublished choice can be made between the sexual and
data). Enhancer detection bears great apomictic modes of reproduction at any stage
potential for the identification of genes that of a breeding program. Apomixis as a
play crucial roles during sexual reproduction, constitutive trait could potentially pose a threat
and also because it serves as an entry point to to genetic diversity, which would preclude the
isolate a multitude of highly specific use of apomixis for crop improvement in a
promoters that are restricted to specific cell versatile and creative way. For instance,
types and regions of the ovule and female apomixis would constitute the default
gametophyte. condition, wherein application of an
exogenous condition or compound would
Although the introduction of transgenes is
suppress apomixis to allow for sexual breeding
now readily achieved in many crop species,
to introgress new germplasm and create
present technology only allows for the
segrega ting populations. Alternatively,
introduction of dominant traits. Transgene
sexuality could be the default condition,
activity is based either on overexpression or
addressing the concern that apomixis could
on homology dependent gene silencing
pose a threat to biodiversity (van Dijk and van
phenomena, such as antisense suppression or
Damme 2000), and apomixis would only be
sense cosuppression (e.g.,Matzke and Matzke
induced at specific steps of the breeding
1995; Jorgensen 1995;Jorgensen et al. 1996).
program and for seed production.
By virtue of the epigenetic nature of these
phenomena, it may prove difficult to Although several inducible systems have been
efficiently and stably introduce the traits that described and shown to work efficiently under
control apomixis. Currently, it is not routinely laboratory conditions, none of these systems
possible to disrupt endogenous genes, and the would allow the induction or suppression of
lack of efficient homologous recombination a trait under field conditions. These systems
techniques, which would allow us to mutate allow either repression or induction of a gene
or otherwise modify endogenous genes, upon addition of a compound. All of the
remains a serious obstacle to genetic existing inducible (suppressible) systems,
engineering in plants. However, recent including the tetracycline repressor (Gatz et al.
reports on successful homologous 1992; Weimann et al. 1994), the copper
recombination in Arabidopsis (Mao and Lam inducible system of the yeast metallothionin
1995; Kempin et al. 1997) may soon lead to gene (Mett et al. 1993; Mett et al. 1996), and
the development of more efficient protocoIs the glucocorticoid receptor (Schena et al. 1991;
for targeted gene disruption. In Arabidopsis Aoyama and Chua 1997) have serious
and maize, gene disruption by site-specific drawbacks for field use, although they are
transposon mutagenesis has become an potent systems for use in the laboratory and
efficient way to inactivate specific genes. greenhouse. These systems often have a high
Although this approach has been adapted to background of noninduced expression and use
rice (Izawa et al. 1997), it is unlikely that it expensive and/or environmentally
will be implemented in a wide variety of other unacceptable inducers that often have poor
crop species. mobility in the plant.
f_Sex",*, .. Apoloiais: ........ _GoMIk ......... 201
As a prerequisite for the use of apomixis in crucial for apomictic development. Therefore,
the field, optimal control systems will have genetic and molecular studies tha t concentra te
to be developed that have low background on the dissection of the sexual pathway will
activity but that can get potently induced. As provide inval uable tools for the engineering of
outlined by [efferson and Nugroho (1998), the apomixis through biotechnology. Several long-
inducer should be water soluble, stable, neglected aspects of sexual reprod uctive
readily transloca table, and should not impair development will have to be addressed, such
the fitness of the plant. The inducer should as the control of megasporogenesis, egg
also be safe to use, biodegradable, and have activation, and imprinting requirements for
no adverse affects on the ecosystem. Recently, endosperm development. Studies on the
two inducible systems that may be developed control of reproductive development in
into agriculturally applicable systems have apomictic species such as Hieracium and
been described. Research at CAMBlA focuses Tripsacum will have to be complemented by a
on the development of a system that is based detailed characterization of the developmental
on the glucuronide repressor, gusR, which is events during sexual reproduction and the
a potent repressor in the absence of molecular mechanisms controlling them.
glucuronides but disengages from the DNA
Even if apomictic species have evolved
when a wide variety of glucuronides are
complex and elusive mechanisms that control
added (Jefferson and Nugroho 1998).
apomixis, which have no immediate molecular
Glucuronides meet many of the criteria for
counterpart in sexual relatives, as suggested by
ideal inducers being inexpensive, benign,
recent findings in Pennisetum and Cenchrus
stable, phloem-mobile, soluble, and not
(Ozias-Akins et al. 1998; Roche et al. 1999),
endogenous to plant cells. Another recently
apomixis could be engineered through a
described system that has potential for field
synthetic approach by targeting the key
use is the ethanol inducible system derived
regulatory steps. This view is supported by the
from the ethanol regulon on Aspergillus
recent findings that the genetic control of
nidulans (Caddick et al. 1998).The system is
apomeiosis and parthenogenesis can be
based on the alcA promoter and its
separated in Taraxacum and Erigeron (van Dijk
transcriptional activator AlcR, which was
et al. 1999; Noyes and Rieseberg 2(00). The
shown to lead to about a hundredfold
engineering of individual elements of apomixis
induction upon the addition of ethanol in
will require a sustained effort to identify more
transgenic plants. Whether it will be
reproductive muta~ts that affect the relevant
practicable to spray or water plants with
developmental processes in those sexual model
ethanol in the field is not known, but the fact
systems that are easily amenable to molecular
that AlcR responds to a variety of other
techniques. Classical genetic screens targeted
alcohols and ketones opens many
at nonreduction, parthenogenesis, seed
possibilities for adapting this system for field
development (mea, fie, fis2 mutants), and
use (Gatz 1998).
regulators of imprinting hold great promise for
Conclusions and Prospects identifying the candidate genes required for the
introduction of apomixis into a wide variety
The introduction of apomixis into sexual
of sexual crops. Enhancer detection provides a
crops will require a multifaceted approach
powerful alternative to classical genetic screens
that builds on the use of both sexual and
and allows rapid cloning of genes involved in
apomictic systems. It is very likely that genes
sexual reproduction and their promoters.
involved in the sexual pathway are also
202 UoIGns......
References Arabidopsis Genome Iniliolive 2000. Analysis of BaUinger, D.G., and S. Henzer. 1989. Targeted
the genome sequence of the flowering gene IllJlolions in Drosophila. Prac. Natl
Mokas, H. 1977. Amutant of borley: lri~oid plant ArDbidapsis rhaliona. Nature 408: Acad. sa. (USA) 86: 9402-06.
inducer. Horley Gene'. NmI. 7: 4-0. 796--815. Barcaccia, G., A. Mollucola, E. AIlertini, 1
AIIemon, M.., and, 1 Dooor. 2000. Genomic Asker, S.E. 1979. Progress in oparnixis research. Zethof, A. GeralS, M.. Pellotli, M.. FoIcinelli.
imprinting in pIonlS: observations and Hereditas 91: 231-40. 1998. Inheritance of parthenogenesis in
evolulionory i",lkotions. Plant Mol. Biol - - . 1980. Gomelophytic apomixis: Poa prutemis L:auxin test lIld AfLP
43: 147~1. 8ements and genetic regulotion. Hereditas 5nkoge analyses support monogenic
Angenent, G.c., and L Colombo. 1996. 93: 277-93. control. Theoretical and Applied Genetics
Molecular centrol of ovule development. Asker, S.E., A. Hagberg, and G. Hagberg. 1983. 97: 74--82.
Trends Plant Sei I: 228--32. Aparrixis in barley? Sver. Utiidesfiiren. Bashaw, E.c., and W.W. Hanna. 1990. Apomicti<
Anloine, A.F., J-E. Foure, S. Cordeiro, C. Dumas, rrlskr. 93: 75-76. reproduction.. In G.P. Choprnon (ed.),
M. Rougier, and lA. FeiiO. 2000. Acalcium Asker, S.E., and LJer~ng. J992. Apomixis in lepraductive Versatility inthe Grosses.
influx is triggered and propagoles in the Plants. Baca Raton, Flarido: OC Press. Cambridge, U.K.: Cambridge Univenily
zygote as 0 wovelront during in vitro Baker, S.c., K. Rabinson-Beerl, J.M. Vilonuevo, Predp. 100-30.
fertilization of flowering ~lIlts. Prac. Nad. lC. Goiser, and C.S. Gasser. 1997. Be<htald, N., 1 Elis, and G. Peletier. 1993. In
Acad. sa. USA 97: 10643--48. InterlKlions among genes regulating ovule pIonta Agrabaderium gene transfer by
Aayomo, 1,and H-H. Chuo. 1997. A development in ArDbiIfopfis t/8iona. in6ltrotion of odull ArDbidapsis t/8iona
glucocorticaid-medioted Iranscriplional Genetics 145: 1109-24. plants. Cl. Aiad. sa. Paris. 316: 1194-99.
induction system in transgenic planlS. Plant
J. 11: 605-12.
Ilecinger, P., and S.D. R~. 1994. Bo1rlOish, K.S., H.E. Yuan, and A.P. Mitchel. 1994. Chin, H.G., M.S. Choe, S.H. Lee, S.H. Park, ic
Gametogenesis in maize. In M. Freeling and Analysis of RIM11, ayeast protein kinase Koa, N.Y. Kin, JJ. Lee, B.G. Oh, G.H. Yi,
Y. Walbat (eeIs.I, The Moize HandW New that phospharyIates the meiotic actiYutor S.( Kirn, H.( Choi, M.J. Cho, and (D.HCII.
York: Springer·Yerlag. pp. 48-61. IMEI. Mol. (eI. Bioi. 14: 7909-19. 1999. MaIecular analysis of rice plants
Belanger K.D., and R.S. Qua1rona. 2000. Brink, RA, lIld D.(Cooper. 1947. The harbaring III At!Os transposable element·
Polarity: the role of localized seaelian. (lIfT. endosperm in seed deveIapment. Bot. lev. mediated gene 1Tapping system. Plant J.19:
Gp. Plant Bioi. 3: 67-72. 13: 423-77. 615-23.
BeReft, HJ., U. O'Kane, cWilsan, U. Buckner, B., and S.L Reeves. 1994. V-Klbt1i1y of Charnel, P.S. 1994. TranspolOl1lagging wilh
Grossniklaus, R.K. PllII'lOI1, IIld WJ. female garnelaphytes that possess Mutator. In M. Freetng and Y. WuIloI (eels.).
Gehrilg. 198Uelement·mediated deliciencies far the regian of chromasarne 6 The tie Handbook.. New York: Springef.
enhancer detedian: a venalile method la contoining the Y1 gene. MayriCll. 4: 247- Yerlag. pp. 243-49.
study development in Drosop/riltJ. Genes lIld 54. Christensen, u., E.J. KiIg. J.L Jordan, and G.N.
Dev. 3: 1288-1300. Burton, G.W. 1992. MooipUating apomixis in Drews. 1997. MegogameIogenesis in
Bellen, HJ., ( Wison, and WJ. Gehring. 1990. I'rtspoNm ftoc. Aporrixis ~, USDA· ATabitJopsiswid type ond the 61111l1ant.
Ilissedilg the CDf11IIexily of the nertOUI ARS, Atlanto, USA. pp. 16-19. Sex. PIant'eprod. 10: 49-64.
system by enhancer detection. Bioessays 10: Caddick, M.X., A.J. Greenlond, I.Jepson, K·P. Christensen, u., S. SulJromanian, and G.N.
199-204. Krause, N.IN, K.V. RiddeI, M.G. Soher, W. Drews. I99B. IdentifimIian of gometap/rylic
Belastal5ky, DA, and R.8. Meagher. 1996. A Schuch, U. SonnewaId, and A.8. Tomsen. mutoIions affecting female garnelophyte
pailen-, owIe-, lIld early embryo-spedfic I99B. An e1hano/ inducille gene switch for develapmen1 inATabit/opsis. Dew. BioI.
poly lA) bindilg pralei1 ~am ATabitJopsis p1anl5 used 10 manipulate carbon 202: 136-51.
camplemenl5 essenIiallun<lions in yeas!. metobaism. Natrle Biotech. 16: 177~. Cackaoh, (E., B.G. den Boer, J.M. Hemy, and
Plant (elB: 1261-75. Carmon, lG.1995. Garnetap/rylic angiospefm JA M1wray. 2000. (ydn Dcon\IOI of
Bemetzen, Il, IIld M. Free5ng. 1993. Grasses aponicIs and the ocCllfence of poIyspary growth rate in p1arm.1tGhn 405: 575-79.
as asingle genetic system: genome and polyembryony among their relatives. (08, E.H. 1959. Hne of maize wiIh Iigh haploid
campasilian, cainearily and compatibility. Apomixis Newsletter. 8: 39-53. ~equency. Amer. NatrI. 93: 381-82.
Trends in Genetics 9: 259-61. - - . 1997. Async!Konaus expression of (08, E.H., M.G. 1Ieuffer, and DA llaisington.
Bensen, H, G.S. W, Y.( Clone, J.T. Tossberg, duplicate genes in ongiO!jlerlTl5 may cause 1988. The genetics of cam. In G.E Sprague
P.S. Schnable, R.B. MeeIey, and S.P. Briggs. aponixis, bispory, letraspary, and and lW. DucIey (eels.!. C«n lIldCorn
1995. Ooning and charooerizalian of the polyembryony. BioIJ.linn. Sac. 61: 51-94. Improvement. Madison, ~:
maize AnI gene. Plant (el7: 75-84. Carmon, lG., c.F. Clone, and O. Riera-Lizorazu. American Society far Agronomy. pp. 81-
Berger, F. 1999. Endasperm development. (lIfT. 1991. Comparative histology of eel waIs 258.
Gp. Plant Bioi. 2: 28-32. during meiaIi< and apomeiotic (auIeau, F., F. BeIziIe, ( HarIaw, O. Grand"I'lI1,
Bicknel, RA 1997. IsoIaIian of a diploid, megasporogenesis in two CIIl5lralasian D. Yezon, lIld M.P. Doutriaux. 1999.
apaniclic plant of HieracilJm DlJllIntiocum. E/ynM Lspedes. Clop 56.31: 1527-32. Random chromasarne segregation wilhaur
Sex. Plant 'eprod 10: 168-72. Carmichael, lS.,and W.E. Friedman. 1995. meiotic arrest in bath male and female
Bicknel, RA, N.K. Barst, and A.M. KaIhmw. Double fertiizalian in Gnellm gnernofT. The meiocyIes of a dmc II1IlIant of Ara6it/opsis.
2000. Monogenic inheritance of apomixis in relatiomhips between the eel cyde and Plant (ellI: 1623-34.
two HiellKilm S4*ies wiIh cistiIcI sexual reproduction. Plant CIf 7: I91~B. (01,1, and H. Ford. 1987. The plastic growth
developmental mechanisms. Heroolty 84: Coss, D.D., and WA Jensen. 1970. Fertiization responses of tlvee agamaspecies of
228-37. in barley. Amer J. Bot. 57: 62-70. dandelion latwo levels of ooIrienI. Am.
Bier, E., H. Yaessin, S. Shepherd, K. Lee, K. Chase, S. 1963. Androgenesis· il5 use for Bot. 59: BI-91.
McCall, S. Barbel, LAckermon, R. Carretla, transfer of maize cytaplasm. J. Hered. 54: Clone, 0., and lG.Carman. 1987. Mecharisms
T. Uemura, E. Grel, LY. Jan, and Y.N. Jan. 152-5B. of apomixis in E/ynM recIiJetus fram
1989. Searching far paIIem and mutation in - - . 1969. MonopIaids and monoploid· easlern Australia and New Zealand. Amer.
lhe Drosop/riltJ genome wilh P·1acZ vector. deriVll1iYes of maize (lea mars L). Bot. 1. Bot. 74: 477-96.
Genes lIld Oev. 3: 127~7. lev. 35: 117~7. Dos, L, and RA Marlienssen. 1995. Site-
Birchler, JA 1993. Dosage lIlII!ygs of maize Chaa, (Y. 1980. Aulon()lTl(M develapmen1 of seleded transpolOl1 mutagenesis atthe
endosperm development. Annu. 'ev. Gene!. the embryo in I'rtspoNm conjugollm Berg. 1rc/1061ocus in llllize. f'IonI CIf 7: 287-
27: IBI-204. Bot Not. 133: 215-22. 94.
Banhamme, S., ( HarIaw, D. Yezon, S. de Chaudhury, A.M., L Ming, ( "'11er, S. C1aig, E.S. Davis, G.L 1966. Systemotic fnrIrtoIogy 01 tile
Loinanliere, A. Guyon, M. Ferault, M.
Marchand, N. Bechtold, and G. Pelletier.
1998. T-DIlA meGated disllJplion of
euenlial garnetophytic genes in Arabidopsis
is unexpectedly rare and camaI be inferred
fram segregalion distortion alone. Mol. Gen.
Genet. 260: 444-52.
Deoois, and WJ. Peacock. 1997.
FeI1iIization-independenl seed develapmenl
inATabidopsis thtHJna. Proc. Notf. AtaJ. Sci.
(USA) 94: 4223-28.
Choudhury, A.M., and 1W. Peacock. 1993.
Approaches tawards isolating apomidic
mularm in ATabidopsJ! t1HJIiJna: prospect
and geneti< 0.
Angiosperms. New York: WiIey and Sons.
DeIaef1, LM.W. 1979. Ererilenm IIIlIanlt in
ArobitJopsis thdiona (L) Heym: Phenolypic
DiboH, A.G., and DA Lorson. 1966. An electron [ames, AJ. 1961. Morphology ofAngiosperrns. Gadello, 1WJ. 1987. Sexuoltetrop\oid and
miaogrophic sIudy of !he motIJe New YOlk: McGrow-Hill. aporriclic pentaploid popuIotions of
megagametophyte in Zea rooys. Amer. 1. Ehlenfeldt, M.K., and R. Orliz. 1995. Evidence on Hierocium pilosello (Compositoe). Plant
Bot. 53: 391-402. !he nature and origin of endosperm dosage 5yst. EvoI. 157: 219-45.
Dioonson, H.G. 1992. PIont reprodudion: Past, requirements in Solanum and other Gaise!, J.e., K. RabillSDll-8eers, and e.S. Gasser.
present and future. In Y. Donee, e.Dunm, ongiosperm genera. Sex. Plant Reprod. 8: 1995. The Arobidopsis 5UPfRMAN gene
and A. Gallais (eels.), Reproductive Biology 189-96. mediates DSyrm1etric growth of the outer
and Plant Breeding. Bertin: Springer-Verlog. Bio", R.e., A.S. 8etzner, E. Hu"ner., M.P Dokes, integument of ovules. Plant (e' 7: 333-
Digomet, L, D. AIson, N. LedIK, e.Dumos, and W.QJ. Tucker, D. Gerenles, PPerez, and 45.
M. Rougier. 1997. Firs! evidence of 0 D.R. Smyth. 1996. AIN7fGUMENTA, on Gale. M.D., and K.M. Devos. 1998. Con1Jorolive
calcium transient in lIowering plants 01 APfTAIA2~ike gene of Arabidopsis with genetics in the grosses. hoc. /Iotl. AlotJ. Sci
fertilization. Development 124: 2867-74. pleiotropic roles in ovule development and USA 95: 1971-74.
Doerner, P, J-E. lllgensen, R. You, J.S1eppuhn, IIorol orgon growth. Plant (eH. 8: 155-68. Gasser, CS., J.8roodhvesl, and 8.A. Houser.
and e.Lonil. 1996. (ontrol of root growth &nsI, A. 1918. Die 8os!ardierung a~ Ursache der 1998. Genetic analysis of ovule
and development by cyetin expression. Apogomie irn PIIonzenreiche. lena, development. Ann. Rev. Plant f'trysioI.Plant
Nature 380: 520-23. Germany: Fisdler. Mol. Bioi. 49: 1-24.
Doutriaux, M.P, E(outeau, e.8ergaunioux, Foure, J.E., e.Digannet, and e.Dunm. 1992. An Gotz, e.1998. An intoxicating switch for plant
and e.White. 1998. 1soI00ion and in vitro sysIem for adhesion and fusion of transgene expression. Nature Biotech. 16:
characterisation of the RAD51 and DMCI maize gametes. Seience 263: 1598-1600. 140.
homalogs from Arabidopsis thliiana. Mol. Feldmaoo, U, D. Coury, and M.L Christionson. Gatz. L, e.Frohberg, and R. Wendenburg.
Gen. Gene/. 257: 283-291. 1997. Exceptional segregation of 0 1992. S1ringent repression and
Dresselhous, 1, (anlls, 5., Heuer, 5., Souter M., selectoble marker (Kanl) identifies genes homogeneous de·repression by tetra·
H. LDrz, and E. Kranz. 1999b. Novel inporIant for gamelaphylic growth and cycline of 0 modified (aMV 355 promoler
ribosomal genes from maize ewe development. Genetics 147: 1411-22. inintact transgenic tobcKco plants. Plant J.
diHerential~ expressed inthe zygotic and Feldmaoo, KA, R.L Molmberg, and e.Dean. 2: 397-404.
somatic cell cycles. Mol. Gen. Gene/. 261: 1994. Mutagenesis in Arabidopsis..In E.M. Gotz, L, and I. Lenk. 1998. Promolen that
416-27. Meyerowilz and e.R. Somerville (eels.), respond to chernicol inducen. Trends Plant
Dresselhous, 1, (OIdts, and H. I.Drz 19990. A Arobidopsis. Cold Spring Harbor, USA: (old Sei3: 352-58.
transcript encoding translation initiation Spring Horbor Loborolary Press. pp. 137- Goodrich J.1998. PIont development: Medeo's
fadOl elF-SA is StOled in!he unferlilised 72. moternol inslincl. (urr. Bioi. 8: 480-84.
egg cell of maize. Plant Mol. Bioi. 39: Finch, RA, ond M.D. 8enne". 1979. Action of GoIubovskoyo,I.N. 1979. Genetic control of
1063-71. triploid inducer (tn)on meiosis in borley meiosis.lnt. Rev. (ytol. 58: 247-90.
Dresselhous, 1, e.Hage!., H. LDrz, ond E. Kronz. (Hordeum vu/gore L). HINedity. 43: 87-93. GoIubovskoyo, I.N., NA Avullcin, and WJ
1996. IsoIotion of 0 ful.Jength cDNA Fobert, PR., 8.L Miki, and V.N. /yer. 1991. Sheridon. 1992. EHed of several meiotic
encoding colreliculin from 0 PO-library of Detection of gene regulatory signo~ in mutants on female meiosis inmaize. Dev/.
in vitro zygates of maize. Plant Mol Bioi plants revealed by T-DNA-mediated Iusions. Gene/. 13: 411-24.
31: 23-34. Plant Mol. Biail7: 837-51. - - . 1997. New insights into !he role of
Dresselhous, 1, H. LDrz, ond E. Kronz. 1994. Franke, R. 1975. Uber dos Aufrelen van the maize ameiatic I locus. Genetics 147:
Representative cDNA libraires from few unreduzierten Gameten bel Angiospermen. 1339-50.
plant cel~. Plant J 5: 605-10. Archiv fiir Ziiclrlvngsfrxschung 5: 201-08. GoIubovskoyo, I.N., I.K. Grebennikova, NA
Drews G.N., D. Lee, and LA. Christensen. 1998. Free/ing, M., and V. Walbot. 1994. The Moize Avulkino, and W.E Sheridon. 1993. The
Genetic analysis of female gamelophy!e Handbook. New York: Springer Verlog. role of !he ameiatic I gene inthe inilio1ion
development and function. Plant (eH 10: Friedman, W.E. 1990. Double ferti~zation in of meiosis and in subsequent meiotic
5-18. Ephedra, 0 non- flowering seed plant: its events in maize. Genetics 135: 1151~.
Dwyer, K.G., M.K. KandaSDmy, D.I. Mohosky, J. bearing on !he origin of angiosperms. Gassier, A., A.L loyner, J.Rossont, and W.e.
Acciai, 8.1. Kudish, J.E. Mille, M. Nosral\oh, Seience247: 951-54. Skornes. 1989. Mouse enilryonic stern
and J.8. Nosrallah. 1994. Asuperfani~ of - - . 1992. Evidence of 0 pre-iJngiosperm cel~ ond reporter constructs to deled
S-Iocus related sequences in Arabidopiir. origin of endosperm: im~ication for !he developmentalfy regulated genes. Science
Diverse slrlKlures and expression poIIerns. evolution of flowering plants. Science 255: 244: 463-115.
Plant (ell 6: 1829-43. 336-39. GrimaneRi, D., M. HerniJndez, E. Peralli, and Y.
Dujordin, M., and W.W. Honna. 1989. Friedrich, G., and PSoIiana. 1991. Promoter Sovidon. 1997. Dosage eHects in !he
Developing aponictic pearl nu11et: traps in embryonic stem cells: 0 genetic endosperm of diplosporous aporriclic
characterization of 0 8~ plant. J. Genet. screen 10 iden~fy and mutate Tripsocum(Pooceae). Sex. Plant Reprod
Breed. 43: 145-51. developmental genes in mice. Genes and 10: 279-82.
Dumas, C, and .I-E. Faure 1995. Use of in vitro Dev 5: 1513-23. Grimanelli, D., O. Leblanc, E. Espinoso, E.
fertilization and zygote culture in crop Frost, H.8., and R.K. Soost. 1968. Seed Perotli, D. Gonzolez de Lecin, ond y.
improvement. Isn. Opinion Biatech. 6: reproduction development of gametes and Soman. 1998. Mopping diplosporous
183-88. embryos. In W. Reuther, LD. 8otchelor, and oponixis in tetraploid Tripsoam one gene
Dumos, c.. and H.L Mogensen. 1993. Gametes HJ. Webber (em.), The Citrus Industry, VoI or several genes? Heredity 80: 40-47.
and fertilization: maize os 0 model system 11. Berkeley, (aMornio: University of
for experimental embryogenesis in (alifornia Press. pp. 29~324
flowering plonh. Plant (ellS: 1337-48.
F_ Sel....., . . . . .is:......... GeHtk Appnaos 205
GrillJlIlelli, D., D. Leblonc, D. Gonzciles de Leoo, Hoig, D., and M. Westoby. 1986. KilKonllict Huong, 8-0., lIld W.F. Sheridon. 1994. Female
and Y. Soft!m. 1995. Aporrixis expression wilhin megaspore letrads. In E.G. Wil5oms, gomelophyte development inmaize:
in maize- Tripsoam hybrid derivatives lIld 1.8. Knox, and D. Imne (eels.), PoIlllltion microtulJulor orgonizolion lIld enUyo !/l(
!he implications regdngits rontrol and '86. Melbourne, Australia: School of Botany, poIorily./brt (e' 6: 845-61.
potential for manipulotion. Apotrixis UniveMy of MeIJourne. pp. 211·14. - - . 1996. Embryo !/l( development in!he
NewsJener 8: 35-37. - - . 1991. Genonic: irnp"inling in mlize indeterminllte gometophyte I
Grini, P. E., A. Sdmillger, H. SdMI-z, I. endospenn: its efled on seed deYelopmen1 mutant AbnormoI oodeor behoYior and
Zimmermcm, 8.SdIwab, G. JUrgens, and incrosses between speOes, lIld belween defeclive microtubule orgorizolion. 1bIt
M. Hiilskomp. 1999.1soIotion of ethyl differenl ploidies of !he some species, and (el/8: 1391-1407.
methonesuHonate-ilduced gomelophytic: its impkations for !he evOOlion of Hiilska"" M., S.D. Kopczok, T.F. Harej5i, 8.K.
mutam inAmbitIopsis dHi.1Il1l by 0 aponixis. 1'fIiI. Trom. R. Soc. Lond. Set. 8. Kill, and I.E. PMt. 1995.ldenIiIiartion 01
segregation al5lortion assay using !he 333: 1-13. genes reqwed for poIIen-sIigmo rel:ogrWIion
muhimarker dllOl11OlOl11e 1.Genetics 151: HoMO, W.W. 1995. Use of oponixis in cultiyor inArrJbidopsis thoIiIlna./brt J.8: 703-14.
849-63. developmenr. Adv. Agron. 54: 333....54. Hussey, M.A., E.t Boshow, [W. Higniflt, lIld
Grossniklous, U., HJ. 8e11en, L WiIson, lIld WJ. HOMO, W.W., and E.t Boshow. 1987. Apomixis: M.L Dohmer. 1991.1nltuenc:e of !he
Gehring 1989. P-element·medialed its idenlilic:ation and use in plonl breeding. photoperiod on !he frequentY of sexual
enhon<er detection oppied 10 !he study of Crop Xi. 27: 1136--39. embryo 500 in focuIroliYe apomicli<
oogenesis in Drosophila. Oe'ielopment 107: Harlon, U.,lIld J.MJ. de WeI. 1977. P~ bufle9l1S1. Euplrytico. 54: 141-45.
189-200. of geneli< transfer from Tripsocum to lea Izowa ,1,1 Ohnishi, 1 Nalcano, N. Is/ida, H.
Grossniklous, U., A. KoIrunow, and M. yon moys. Proe. NotI. Arod. Xi. (USA) 74: Enoki, H. Hashimala, [ Itoh, I.Terado, L
Lookeren Co"1lOP. 1998c. Abrighl 3494-97. Wu, L Miyuzoki, 1 Endo, S. lido , and [
future for oponixis. Trends in f'Iont Sciences Hartwel, LH., and 1H. Weinerl. 1989. Shimomata. 1997. Tramposon logging in
3: 415-16. Che<kpoinl5: controls that ensure the order rice. PIont Mol Bill/. 35: 219-29.
Grossniklous, U., 1 Moore, and W. GogIiano. of cel cyde eYerm. Science 246: 629-34. Jaffe, LF. 1991. The po1h of cakium in cyrosolic:
I99Bo. MoIerulor and geneli< opproodles Hoshemi, A., A. Estilli, lIld 1 Wo~. 1989. cakiurn osciIlotions: 0 unifying hypothesis.
10 undemanding lIld engineering apomixis: (ytogeneli<s and reproductive behovior of Proe. NaIl. AaK/. Xi. (USA) 88: 9883-87.
ArrJbidopsis os 0 powerful tool. In 8. Hardy induced and noturaltetraploid guoyule - - . 1996. Egg membrllleS during
(ed.), Proceedings ofthe 3rd Intemotillool (PortlreniumargentolumGrl/'f). Genome fertilizolion..In S.G. SchuItz et01. (eds.),
Hybrid Rice S~m 1996; Manila, 32: 11 QO--04. MoIecuIor Biology ofMembrooe TrollSpOl1
Philippines: IIII Press. Herr, lM. Jr. 1995. The origin of !he ovule. lJiJon/ets. New York: PIeoom Press. pp.
Grossniklous, U., 1.[ Peol5Oll, and WJ. Gehring. Amet: 1.Bot. 82: 547-64. 367-78.
1992. The Orasoplrio sJowy ptiredlorus Hill, lL,lIld E.I. Lord. 1994. Wild type ber Jeffenon, I.A. 1987. Assoying chimeric genes in
encodes two proteins that show homology development gyneoOol inilialion.ln 1 plum: The GlIS gene fusion system. /brt
10 manrnolion transcription Iu<r~. Genes 8owmon (ed.), ArrJbirJopsir: An AlIos of Mol. Bioi. Rep. 5:387-405.
ond Dey. 6: 1030-51. Morp/roIogy and Oe'ielopment. New York: - - . 1993. Strolegic developmenl 01
Grossniklous, U., lIld [ Schneirz. 1998. The Springer Verlog. pp. 158-59. aponixis os 0generoltoollor O9'icullllle.
genelK and moIeaJIcr basis 01 n and Hirayoma, 1,L Is/ida, 1 Kurornori, S. Obota, L In KJ. Wilson (ed.l, Proe. 1nt.1bhhop on
megagametophyte development. SeminotJ Shimodo, M. Yamamola, [ Shinozoki, and Apotrixis in lice. Conberro, Australia:
ee.
in & Oevl. Bioi. 9: 227-38. L Ohta. 1997. fImionol cloning of 0cDNA CAM81A..
Grossniklous, U., L Spilone, 0.1. Page, and t encoding Mei2~ike prOlein from - - . 1994. Aponixis: AsoOoI revokllion
Kiihler. 2001. Genomic: irnp"inling lIld seed ArrJbidopsis th~itI/lIl usiIg fission yeast for ogricultwe? Biotechnology rnJ
deyelopment endosperm Iormotion wilh pheromone rel:eplor delicient mutant. FfBS Oe'ielopment Monitor 19.
and without sex. (lIfT. Op.1'Iont Xi. 4: 21- Left. 413: 16--20. Jeffenon I.A., lIld I. 8idcneI. 1996. The
27. Halmekler, W. 1849. Die Entffehung des ErrDrya potential impocIs of aponixis: 0molecuIor
Grossniklous, U., lIld J.p' YJe8e.(alzodo. 1998. der I'honerogomen. ~zig, Germany. genelia opprooch. In 8.W.S. Sobral (ed.),
... response: parental connic:l and Homo, SI, 1 Corral, and [Swam. 1993. The TIre IIIfIIlCI of1'Iont MoI«uIor Genetia.
infanticide during embryogenesis. Trends role of mkium inmanrnolian oocyte 8osIon: Bi-khouser. pp. 87-101.
/'Iont Xi3: 328. mo1lJrolion lIld egg lX1iYuIion. Hum. Jeffenon, RA, S.M. 8111gess, and D. Hinh.
Grossniklous, U, YIelIe CoIzado, J.p, Hoeppner, M, Reprod. 8: 1274-81. 1986. 6 Glucuronidase from £sdJerichio Cof
and W. GogIiono 1998b. MaIernoI control of Howden, I.,S.K. Pork, lM. Moore, 1 Orrne, U. os 0gene fusion marker. Proc. NotI. AaK/.
embryogenesis by MEDEA, 0 PoIyromb- GrossnikJaus, and D. Twel!. 1998. Selection Xi. (USA) 83: 8447-51.
group gene in ArrJbidopsis. Science of T-DNA·togged male and female Jeffersoo, 1.A.,·lIld S. Nugroho. 1998. MoIeruIor
280:446--50. gornetophytic: mutom by segregation strolegies for hybrid rice: mole striry and
Guslakson, A. 19470. Aponixis in angiosperms. distortion in ArrJbidopsis. Genetics 149: aponixis. In 8.Hardy (ed.l, Proceedings of
11. Lunds UniY. Arnkr. N.F.1I42: 71-179. 621 ....31. the Jrd IrrIernIltionII/ S~ on HyIriI
Guslakson, A. 1947b. Aponixis inangiosperms. Huong, 8·Q., lIld S.D. lusseI. 1992. The female Rice. 1996. MoIiIo, Phi5ppines: IIII Press.
Ill. Lunds UniY. Arnkr. N.F. 1143: 1U-370. germ unit. In S.D. lusseI and t Dumos Jensen, W.A. 1965. The ultrostr1Klllle and
Hagberg. A., and G. Hilgbefg. 1980. H~ (eels.), Sexwl ReprotlllClion in flowering hislodlemistry of !he synergids of canon.
frequency of lflOI1Ianeous hopIoids inthe /'Ionts. Int. Rey. (yto/. 140, New York: Amet: J. Bot. 52: 238-56.
progeny of on induced mutation in barley. Academic Press. pp. 233-93. - - . 1968. Canon embIyogenesis: The
Hereditas 93: 341-43. zygole. PIontu 79: 346--66.
206 Uei Gre".lIoo,
- - . 1974. Reproduction in flowering Kempin, S.A., SJ. liljegren, LM. 8kxk, S.D. K1imyuk, Y.I., and J.D.G. Jooes. 1997. AtDMC1,
plonts. In: A.W. Robords (ed.), Oynomi< Rounsley, MJ Yanakky, and E. Lam. 1997. the Arobidopsis hamalogue of the yeast
Aspects ofPlant Ultrll5tnKture. New York: Targeted disruptioo in Arobidopsis. Nature. DMCl gene: characterization, transposon-
McGrow-Hill. pp. 481-503. 389: 802~3. induced alletJc variotian and meiosis
Jensen, W.A., M.E. Ashtoo, and CA.lIeosley. Kerbundn, S., H. DeGreve, f. Debaeck, M- Vlll1 associated expression. Plant Jour.nol. 11: 1-
1985. Pollen tube-embrya sec interaction in Mootagu, and J.-P. Hernalsteens 1991. In 14.
caNon. In D.L Mukahy and E. Ottaviona viva random b-glucuranidose gene lusioo in Klucher, K.M-, H. Chow, LReiser, and R.L fischer.
(eds.1, Pollen: Biology and Implications for Arobidopsis tha/iono. Prrx. NolI. Acod. 56. 1996. The AIN7EGUMEN7A gene of
Plonl Breeding. New York: Elsevier (USA) 88: 5212-16. Arobidopsis required far ovule and female
8iomedical. pp. 67-72. Kermi<le, J. 1969. Androgenesis cooditioned by gametophyte development is reloted to the
Ja/1I1sI00, SA, T.P.M. den Nill, SJ. Pelaquin, and a mutatioo in maize. Science 116: 1422- floral homeoti< gene APETAIA2. Plant (eH. 8:
R.E. Honneman Jr. 1980. The significance of 24. 137-53.
genic balorKe la endasperm development in - - . 1970. Dependence of the R-mallled Knax, R.8. 1967. Apomixis: seasonol and
interspe<ilic crosses. Theor. Appl. Genel. 57: aleurone phenotype in maize an the made papulotian differences in 0 grass. 56ence
5-9. of sexual transmission. Genelics66: 69- 157: 325-26.
Ja/1I1sI00, SA, and R.E. Honneman. 1982. 85. Knax, R.8., and J.Heslop-Harrisan. 1963.
Manipulotions of endasperm balonce - - . 1971. P1eialropi< eff~ 011 seed Experimental cootral of apasparaU5
number overcame crossing barriers belYreen development of the indelerminate apamixis in a gross of the AndrapogOlleile.
diploid So/onumspecies. xience 247: 441t- gomelophyte gene in maize. Amer. 1. Bol. BOI. Nol. 116: 127-41.
48. 58: 1-7. Koboyashi, 1,Y. HaNa, and S. labata. 1993.
Ja/ni, 8.M., K.8. Ambegaakar, and PS. - - . 1994. Indelerminole gamelophyte lsolatioo and characterization of ayeast
Srinivasfa. 1992. (omporative Embryology (igJ: biology and use. In M. freeting and Y. gene thot is homologous with a meiosis-
ofAngiosperms. Vols. 1and 2. New York: Walbot (eels.), The Maize Handbook.. New specific cDNA from a plont. Mol. Gen. Genet.
Springer·Yerlog. York: Springer-Yerlog. Pp.388-93. 237: 225-32.
Janes, lJ.,and 1LRosl. 1989. Hista<hemislry Kermi<le, J., and M. Alleman 1990. Gametic Koboyoshi, 1, E. Koboyashi, S. Sata, Y. HaNa, N.
and uhrostructure of rice (Oryla 1DIiv0) imprinling in maize in relation la the Miyajima, A. lanoka, and S. Tabata. 1994.
zygoti< embryogenesis. Amer. J. Bol. 76: angiosperm life cycle. Development Suppl Characterization of cDMAs induced in meiotic
504--20. 1:9-14. prophose in li~ microsparocytes. ONA Res.
Jangedijk, E. 1985. The paltern of Kiesselbach, lA 1949. The structure and 1: 15-26.
megasporogenesis and megogamelogenesis reproduction of cam. Univ. NebraSKa (011. Koboyashi, 1,and K. Tsune'Mlki. 1978. Ha~oid
in di~oid Solanum species hybrids: i~ Agr.ic. Exp. Slotion Res. Bull. 161. 50th induction and its geneti< mechanism in
relevance to lhe origin of 21HQQ5 and the anniversiory edition. Cold Spring Harbor allo~asmi< common wheat. 1. Hered. 71: 9-
inductioo of apomixis. Euphytico 34: 599- (USA): Cold Spring Harbor Laboratory 14.
611. Press. Kajima, A., and Y. Nogata. 1997. Discovery of
Jorgensen, RA 1995. Clrsuppression, flower Kitlan, A., DA Kudra, A. Kleinho~, M. Yano, N. high~ apamicti< and highly amphimicti<
color palterns, and metastable gene Kurala, B. SIeffenson, and 1Sasoki. 1995. dihaploids inAHium tuberasum. Sex. Plant
expression states. Science 268: 681t-91. Ri<e-barley synteny and i~ ap~icatian to Reprod. 10: 8-12.
Jorgensen, RA, P.D. 0U51er, .I. English, Q. Que, saturation mapping of the barley Rpg I KoItunow, A.M. 1993. Apomixis: embryo IOCS and
and CA. Napoli. 1996. Chalcone synthase region. Nucleic Adds Res. 23: 2729-33. embryos farmed without meiosis or
casuppression phenatypes in petunia Kimber, G., and R. Riley. 1963. Haploid fert~izalion in ovules. Plant (eH 5: 1425-
flowers: comparison of sense vs. anmense angiasperms. 801. Rev. 29: 48G-53J. 37.
constructs and single-<opy 'IS. complex T- Kindiger, 8.K., 8.Dopeng, and Y. Soicolav. 1996. KoIIunow, A.M., R.A. Bi<knell, and A.M.
DNA sequences. Plonl Mol. Bioi. 31: 957- Cytalogi<al and moIeculor studies of Chaudhury. 1995. Apomixis: Molecular
73. apamicti< maize-Tripsocum hybrids. strategies far the generalioo of generically
Kossir, Y., D. Granat, and G. Simchen. 1988. Agrooomy Abs1rads. pp. 132. identi<al seeds without fertilization. Plant
IME!, apositive regulator gene of meiosis Kinoshna, 1, R. Yadegari, JJ. Harodo, R.8. Plrysiol. 108: 1345-52.
in 5. cerevisioe. (ell 52: 8S3--62. GoIdberg. and R.L Fischer. 1999. KoItunow, A.M_, 1 Hidaka, and S.P. Rabinson.
Kaul, M.LH., and 1G.K. Murthy. 1985. Mutant Imprinting of the MEOEA Polycomb gene in 1996. Polyembryony in dms, Accumulation
genes affecting higher ~ant meiosis. Theor. the Arobidopsis endosperm. Plant CBH 11 : of seed storage proteins inseeds and in
Appl. Genel. 70: 449-66. 1945-52. embryos cultured in vilra. Plant 1'frysioI.
Kavanagh, TA, RA Jefferson, and M.W. Bevan. Kiyosue,1, N. Ohad, R. Yadegari, M. Honnan, J. 110: 599-609.
1988. Targeting afareign protein to Dinneny, D. Wells, A. Katz, LMorgossion, Kranz, E., and 1 DresselhoU5. 1996. In vitro
chloroplos~ using Iusions to the transit JJ. Horada, R.8. GoIdberg, and R.l flSCher. fertitJzation with isolaled higher plont
peptide of achlorophyll a/bprotein. Mol. 1999. Control of fertitlZatian-independent gametes. Trends Planl Sci 1: 82-89.
Gen. Genel. 215: 38-45. endasperm development by the MEDEA Kranz, E_, P. van W"Mlgen, and H. lOrz. 1995.
Kowoguchi, H., M- Yashida, and I. Yamashita. Po/ycomb group gene in ArobirJopsis. Prrx. Early cytologi<al even~ alter induction of cel
1992. Nutritianol regulatioo of meiosis- Natl. Aiod. 56. USA 96: 4185-91. division in egg cells and zygote development
specific gene expression in Socrhoronryces KIophatz, S., and R.E. Esposita. 1980. Isolation of following in vilro fertibation with
cerevisioe. Biolech. Biochem. 56: 289-97. spol2-1 and spol3-1 from anarural angiospelm gametes. Plant J. 8: 9-23.
KeHer, 8., and C. feuillel. 2000. CoIinearity and variant of yeast that undergoes asingle Krysan, PJ., J. C. Young, and M- R. SU5\1l1rn.
gene density in grlllS genornes. Trends Planl meioli< division. Gene1icl96: 567-88. 1999. T-DNA os an insertional mutogen in
xi.5: 241t-51. Arobidopsil. Plonl (elll1: 2283-90.
F... s....lity.. ApaoUis: MoItaIor ... e;-,...,..... 207
Long, J.D., S. Ray, and A. Ray. 1994. rinl, a Mongelsdorf, P.e., and R.G. Reeves. 1931. transcription factor gene Iomily in
mutation affecting female fertility in Hybridization of maize, Tripsorum and Arobidoprir: assessing the potentiol of
Arobidopsis, in'eracts with modI, in Euchloeno. J. Hered. 22: 329-43. reverse genetics to identify insertionol
recessive modifier. Genetics. 137: 1101-1 O. MonsIield, S.G., and LG. 8rialty. 19900. mutations in R2R3 MY8 genes. I'Iont (ell
Leblanc 0., M.D. Peel, J.G. Carman, and Y. Development of the free-nuclear endasperm 11: 1827-40.
Soman. 1993. Megasporogenesis in sexual in Arobidopris tholiooo L Arobidopris Inf. Mena, M., 8AAmbrase, R.8. Meeley, S.P.
and apomic1ic TripllKllm species using Serv. 27: B-M. 8riggs, M.F. Yanakky, and RJ. Schmidt.
interference contrasl and fluorescence. MonsIield, S.G., and LG. 8rialty. 1990b. 1996. Diversification of (·function in maize
Apomixis NewsJener6: 14-17. Endosperm cellularization in Arabidopris flower development. Science 274: 1537-
- - . 1995. Megasporogenesis and tholiono L Arobidopris Inf. Serv. 27: 65- 40.
megagometogenesis in several Tripsorum 72. Messing, J., and U. Grassniklaus. 1999. Genonic
species (Paoceoe). Amer. 1.BOlony. 82: 57- Monslield, S.G., LG. 8riolty, and s. Erni. 1991. imprinting in plants. Results Prabl in (ell
63. Eorly embryogenesis of Arobidopris Differ. In R. Ohlsson (ed.), Genomk
Lean-K1asterziel, [M., e.J.Keijzer, and M. tholiono. I.The mature embryo 5OC. (on. J. Imprinting 25: 23-40.lIeideIberg,
Kaarnneel. 1994. Aseed shope mutant in 801.69: 447-ilO. Germany :Springer Verlag.
Arobidopris that is affected in integument Matzk, F. 1996. The 'Solmon ~tem" of Men, V.L, LP. Lochhead, and P.H. Reynolds.
develapment. I'Iant (eR. 6: 385-92. wheat - a su~able model for aponixis 1993. (apper-eontroloble gene expression
Un, 8.·Y. 1978. Structural modificalians of the research. Hereditas 125: 299-30l. ~tem for whole plants. PrO<. Nud. Acod.
female gamelophyle associated with the Matzk, F., H.M. Meyer, H. Biiumlein C, HJ. 50. (USA) 90: 4567-71.
indelerminate gamelophyle (ig) mutan' in Balzer, and I. Schubert. 1995. Anavel Men, V.L, E. Podivinsky, A.M. Tenoonl, LP.
maize. (an. J. Genel. (ylol. 20: 249-57. appl'm to the aoolysis of the initiation of Lochheod, W.T. Janes, and P.H. Reynolds.
- - . 1981. Megogame'ogenetic aherations embryo development in Gramiooe. Sex. 1996. Asystem for Iissue~ific copper·
associated with the indelerminate I'Iant Repred. 8: 266--72. cootralloble gene expression in Iransgenic
gomelop/ry1e (ig) mulation in maize. Rev. Matzk, F., H.M. Meyer, e.Hootmann, HJ. 8alzer, plants: nodule-specific onlisense of
Bras. BioI. 41: 557-U H. Biiumlein, and I. Schubert. 1997.A asportole aminotransferase-PI. Tronsgenic.
- - . 1982. A!5O<iotion of endasperm speciIic olpha-tubulin is as5O<io'ed with the Res. 5: 105-13.
reduction with porental imprinting in maize. inilio'ion of por1henogenesis in 'Salmon' Meyerawm, E.M. 1989. Arabidopsis, a useful
Genetics 100: 475--86. whea' ~nes. Hereditas. 126: 219-24. weed. Cell 56: 263-ll9.
- - . 1984. Ploidy borrier 10 endasperm Matzke, MA, and AJ.M. Matzke. 1995. How Meyerowilz, EM., and CR. Somerville. 1994.
development in maize. Genelics. 107: 1DJ.- and why do ~anls inactivate homalogous Arabidopsis. (aid Spring Horbor, USA: (old
15. (tronslgenes? I'Ianl Plrysiol. 107: 679-85. Spring Horbor Lobaratary Press.
Uu, Y·G., N. Minukmwo, T. Oosumi, and R.F. Marsden, M.P.F., and D.W. Meinke.1985. Mioo, I.H., and E Lom. 1995. Targeted
Whittier. 1995. Efficient isolatian and Abnormal development ofthe suspensor in disruption of the TGA310cus inArabidopsis
mapping of Arobidopsis tholiooo T·DNA an embryo lethal mutan! of Arobidopris tholiono. I'IantJ.7: 351J-.65.
insert junctions by thermal asymmetric tholiano. Amer. J. s« 72: 1801-12. Minet, M., M.E. Dufour, and F.l.oaoute. 1992.
interlo<ed PeR. I'Ionl1. 8: 457-U McOin'O<k, 8.1978. Development of maize (ornplementolion of Sochorornyces
Lopes, IU, and 8ALorkins. 1993. Endcnperm endosperm as revealed by clones. In S. cerevisioe auxatrophk muloots by
origin, development, and function. Planl (eN Subtelny and I.M. Su1leX (eds), The Oanol Arobidopsis thaliooo cOMAs. I'tant 1. 2:
5: 138J.-99. Basis ofDevelopment. New York: Acadenic 417-22.
luo. M., P. 8iladeau, E.S. Dennis, WJ. PeacO<k, Press. Pp. 217-37. Miles, J.W., F. Pedroza, N. Palocios, and J.
and A. (houdhury. 2000. Expression and McCauch, S.R., X. <hen, O. Ponaud, S. Temnykh, Tahme. 1994. MoIe<ular mcrker for the
porent-llf-origin effects for FlS2, MEA, and Y. Xu, Y.G. Cho, N. Huong, T. Isnii, ond M apomixis gene in Brochi«io. Abstract in
RE inthe endosperm ond embryo of 8loir. 1997. Micrasalell~e marker Pro<eedings of Plant Genome 11,24-27,
developing Arobidopsis seeds. PrO<. Nat!. development, mapping and application in Jan. 1994, Son Diego, Ca5famia, USA. pp.
Aced. 50(USA) 97: 10637-42. rice genetics and breeding. I'Iont Mol. BioI. 51.
lua, M., P. 8iladeau, A. Kohunow, ES. Dennis, 35: 89-~9. Misra, R.e. 1962. Cantribulian to the
WJ. PeacO<k, and A.M. Choudhury. 1999. McKinney, Ee., N. A1i, A. Trout, KA FeIdmonn, embryology ofArobidopsis thaliooo (Goy
Genes con'rolling fertilizalion-independent DA 8e1ostotsky, J.M. McDoweH, and R.8. ond Mann.). Univ. Hes., Agro. 11: 191-
seed development inArabidopsis thaliooo. Meagher. 1995. Sequence-based 99. .
PrO<. NatI. Acod. 50. (USA) 96: 296--301. identification of T-DNA insertion mutations M~chen, A.P. 1994. (ontral of meioli< gene
Ma, H.1999. Seed development: with or without inArobidopsir actin mutants od2·1 and expression in Socrharomyces cerevisioe.
sex? (urr Bioi. 9: 636--39. od4-I.I'tant1. 8: 61J.-22. Microbiol. Rev. 58: 56--70.
Maheswori, P. 1950. An InlredlKlion 10 the Meissner, R.e., H. Jin, E (oninel~, M. MilcheU, A.P., and K.S. Bowdish. 1992. 5eleetian
Embryology ofAngiospenm. New York: Denekomp, A. Fuertes, R. Greco, H. D. for early meiolic mutants il yeast. Genetics.
McGraw-HiD. Kranz, S. Penlield, K. Petroni, A. Urzainqui, 131: 65-72.
Malone, R., s. 8ullar, M. Hermiston, R. Rieger, M. e.Martin, J.Paz·Ares, S. Smeekens, e. Milchen, A.P., S.E Driscall, and H.E Snith. 1990.
(001, and A. Golbrailh. 1991.1so1alian of Tonelli, 8. Weisshoar, E8aumann, V. Positive contral of sparuIotiOll-specific genes
mutants defective inearly steps of meiolic Khmyuk, S. Marillonnet, K. Patel, E by IMEl and IME2 produ<ts in
recombination in the yeast Socrhoromyces Speulman, A. F. Tl5Sier, D. 8auchez, J.J.D. Socrhoromyces cerevisioe. Mol. (eR. Bioi.
cerevisioe. Genetics. 128: 79-88. lanes, A. Pereira, EWl5man, and M. 8evan. 10: 2104-10.
1999. Function search in alarge
208 UoIGr.........
MIodzik, M., Y. Hironi, U. Weber, ts. Goodmon, Naumova, IN., A.P.M. Den Nijs, and M.T.M. Ohod, N., R. Yodegari, LMargassian, M. HaMOn,
ond G.M. Rubin. 1990. The Orosophilo Wiemse. 1993. Quantitative anolysis of D. Michae5, JJ. Haracla, R.8. Goldberg, ond
smn-up gene, a member afthe steroid aposparaus parthenogenesis in POll R.L frsdler. 1999. Mutations in RE, a WO
re<eptor superlamily contrak photoreceptor proIemis genotypes. Ado Bot. Neer/. 42: pa!ycarm group gene, allow endasperm
cell lutes. (e1I60: 1237-49. 299-12. develapment without fe"~ization. Pbrt (el
Modruson L,LReiser, U Feldmann, R.L Naumova, IN., ond M.T.M. Willernse. 1995. 11: 407-16.
Fischer, and G.W. Houghn. 1994.1Iomeoli< Uhrastructural choracterization of apaspary O'Kane, t, and WJ. Gehring. 1987. Detection in
transforma1ion afavvles inta carpel-5ke in I'onicum moxiroom. Sexuol Plonl leprod. situ of genomic regulatory elements in
structures inArahit/opsis. I'Iant (el6: 8: 197-204. Drosophilo. /'roe. NatI. Atod. sa. (USA) 84:
333-49. Nelson, O.E., and G.8. Gory. 1952. Genic: contral 9123-27.
Mogensen, H.L 1988. Exclusion ofmole of semi-s1eritlly in moize. J. Hered. 43: Okon, H, ond 0.0. Coss. 1981. Changes in
mitochondriJ and plastids during syngamy 205-10. megagametophyte structure in I'opaver
in barley as a basis for maternal Neuffer, M.G., and W.F. Sheridan. 1980. nudicoule following in vitro placental
inheritance. /'roe. NatI. Alod. sa. (USA) 85: Defective kernel mutants of moize. I. pollination. Amer. J. Bol. 68: 133Pr41.
2594-97. Genetic: and Iethatlly studies. Genetics 95: OIsen, O-A., t Umestad, ond S.E. Nic:haIs. 1999.
Mogie, M. 1992. The ewlution of asexual 929-44. Developmental biology of the cereal
reproduction in plants. London: Chapmon Nishiyamo, I., ond 1 YoOOno. 1978. Causal endasperm. Trends Planl sa. 4: 253-57.
and HoI. rela1ionships belween the polar nudei in Oro, A.E., E.S. Dng, 1.5. Margalis, lW. Posakany,
Monlamayor, 1.t, D. Vezan, t 8ajon, A. double fertilizolion and interspecilic: cross- M. McKeown, and R.M. [vans. 1988. The
Souvanet, O. Grandjeon, M. Marchand, N. ~libility in Avena. (ytologio (Tokyo) DrOSopJ.13 gene knirps-reloted isa me.
8e<hlold, G. Pel\etier, and t HarIaw. 2000. 43: 453-66. of the steroid-re<eptor gene superlami~.
Switchl (swil), on Arabidopsis mutant Nogler, GA 1972. Genetik der Apasparie bei Nature 336: 493-96.
affected inthe female meialic switch. Sex. 101lVflClJ1us OUricOllXlS. 11. Ozias-Akins, P., E.L Lubbers, W.W. Honna, ond
Plant leprod. 12: 209-18. Endaspermzytologie. Ber. Sdrweiz. 801. lies. lW. McNay. 1993. Transmissiarl of the
Maare, G., K.M. Devas, Z. Wang, and M.D. Gale. 82:54-U apamdK mode af reproduction in
1995. Grasses: line-up and form 0 circle. - - . 1973. Genetik der Apasparie bei Penniselum: CCHlaminoll(e of the tra~ and
(ooem Bioi. 5: 733--42. 101lVflClJ1us OUricOllXlS. Ill.Ff moIerular markers. 1heor. ApJi. Genel. 85:
Moare, 1, J.p' V"Mlle Colzodo, W. Gogliano, and RUdtkreuzungsbastarde. Ber. Sdrweiz. 801. 632-38.
U. GrOlll1iklaus. 1997. Genetic: lies. 83: 295-305. Ozios-Akins, P., D. Rache, ond W.W. Hamo. 1998.
charocterization of hadad. aI1IJtonl --.1975. Genelio afapaspary in TIQht duslering and hen1zygasily of
disrupting femole gametogenesis in 101lVflClJ1us OuricOIllU1 IV. Embryology afF3 apamxis-nnked moIeculor morkers in
Arobidopsis rholiona. Cold Spring Horbor ond F4 backcross offspring. Pennisetum squomulotum implies genetic:
Symp. Quont. Bioi. 62: 35--47. l'I7ytomorph%gy 25: 485-90. control of apaspary by adivergent laM
Mara·Garcia, 5., and 1 Goodrich. 2000. Seeds of - - . 1982. How to obtain di~oid apaniclic: that may hove no allelic: farm in sexual
conAic1. (00 BioI. 10: 71-74. lonunculus OuricOllXlS plants noIfound in genotypes./'roe. NatI. Alod. sa. (USA195:
Mardhorst, A.p., MAJ. Toonen, and S.t de Vries. the wild slate. Bot. HeIv. 92: 13-22. 5127-32.
1997. PIont errUyogenesis. Crit. lev. PI. - - . 19840. Gometaphytic: apomixis. In Pabner, R.G. I 97l.Cytolagic:alltudies of ameioli<
Sci. 16: 535-76. B.M. Johri (ed.), &nbryo/ogy of and normol maize with referell(e to
Margan, R.N., P. Ozias·Akins, and W.W. Honna. AngiMperrns. Berlin: Springer-Verlag. Pp. premeiotic: pairing. Chromosomo. 35: 233-
1998. Seed set inan apamdK 8~ pert 475-518. 46.
milel. Int. J. PIonl Sci. 159: 89-87. - - . I 984b. Genelia of apaspary in Parinav, 5., S. Mayalagu, D. Ye, W-t Yang, M.
Murray, A.W. 1992. Creative blcxXs: Call cycle opamdK 101lVflClJ1us OUricOllXlS. V. Kumaran, and V. Sundaresan. 1999.
checkpoints and feedba<X controls. Nature Conclusion. 801. Helv. 94: 411-22. Analysis of flanking sequences fram
359: 599--604. Noiral, M. 1993. AIIe'c ratiOl and ster~ity inlhe 0iss06ati0n insertion nnes: Adatabase for
Murgia, M., 8Al. Huang, seTucker, and M.E. agamic: complex of the Maximeoe reverse genetics in Arabidopsi1 Plant (eH
Musgrave. 1993. Embryo SO( locking (Panic:oideae): evolutionary role of the 11: 2263-70.
ontipodal celk inArabidopsis thoJiona residual sexuality. J. EvoI. Bioi. 6: 95-101. Paterson, A.H., I·H. Lon, K.P. Reischmonn, t
(8rassicoceoe). Amer. J. Bot. 80: 824-38. Noyes, R.D., and LH. Rieseberg. 2000. Two Chang, Y.-R. Un, S.-t liu, M.D. Burow, S.P.
Nodeau, JA,X.S. lhong, 1 U, and 5.0. O'Nei1I. independent loci conlrol agamosperl1T1 KOMJIski, tS. Kallar, TA DelManle, U
1996. Ovule devela,nent: Identification of (apomixis) in the lriploirlllowering ~onl Feldmonn, K.F. s<hertz, and lE Wendel.
stage spe<ilic: and tissue speciIic cOMAs. Erigeron onnws. Genetics 1SS: 379-90. 1996. Toward a unified geneli< mop of
Plant Cs1I8: 213-39. Nuccitelli, R. 1991.11aw do sperm octivate eggs? higher plants, transcending the monacot-
Nouber, U., M.J. Pankratz, A. Kienlin, E. Seaert, (urT. Topics Oevl. Bioi. 25:1-16. dic:al divergence. Nature Genelics 14: 380-
U. K1emm, and H. Jii<kIe. 1988. AbdaminaI Nygren, A. 1967. Aponixis in the angiasperms. 82.
segmentation of the Orosophilo embryo Hondb. ded'f/onzenphys. 18: 551-96. Potlerson, E.8. 1978. Properties and uses af
requires a hormone receptor~ike pralein Ohod, N., LMargosWn, Y.-t Hsyu, t Williams, duplicate deli<ienl chromosome compliments
encoded by the gap gene knirps. Nature P. Repelli, and R.L Fischer. 1996. A in moize.ln D. Woldan (ed.), Maze
336: 489-92. mutalion that alows endasperm Breedtrg ond Genetics. New York: John
Naumova, IN. 1993. Apomixis in Angiosperms: develapment without fertilization. /'roe. Wiley and Sans. pp. 693-71 O.
Nucellor ond Integumen~ ftriNyotry. NaIL Atod. Sci. (USA) 93: 5319-24. Peacock, lW. 1992. Geneli< engineering and
80ca Raton, Florida: (RC Press. mutagenesis for apamxis in rice. Apomixis
Newsleller4: 3-7.
,_ s.. ~ to ApooDis: ........ G.etir......... 209
- - . 1993. Genetic engineering and Ray A., K. RobiIlSOll-BeeI1, S. Ray, S.c. Baker, Rothe, M., U. Nauber, and H. Jiickle. 1989.1hree
mutagenesis for aponixis in rice. In KJ. lD.Lang, D. PreUlS, 5.8. ~ligon, and C.S. hormone recep.or~ike OrIlSDphiJ genes
Wil50n (ed.!, /'roe. Int. Workshop 01/ Gasser. 1994. Arubidopsis floral oomeo.ic encocIe and identiml DNA·bindilg filger.
Apomixis illice. Oxmgsha, China. New gene BEll (BELl) cOl/'rols ovule £MBO 1. 8:3087-94.
York: Rockefeler F1lmdation. Pp. 11-2L development through nega'ive regulation Rathe, M., M. PehI, H. Taubert, and H. JackIe.
Peel, M.D., lG.Carman, ond O. Leblanl1997. of AGAMOI./S gene (AGl./'roe. NatI. Acari. 1992. Lass of gene Imction .lvough rapid
Megosporocyle mllose it apomicIic SO. (USA) 91: 5761~5. malic cydes il the OrIlSDphiJ embryo.
buffelgr~, Kentucky bIuegr~, Reddy, PS. 1977. Evolution of apomictic Nature 395: 156-59.
PennisetOOl sqllfllllJlrmm Fresen, medxmisll15 in Grominae: AcOl/cep'- Rougier, M., A.F. An.oine, D. AIdon, and C.
Tripsowm L, and -.ping Iavegrass. frap I'fIytDlllDf/lha/agy27: 4~9. DuIllllS. 1996. New lights il early Sleps of
SO. 37:717-23. Reddy, PS., and R. D'Cruz 1969. Me<hanism of in vitro fertilization in plants. Sex. Plant
Pessino, S.c., lP.A. Ortiz, O. Lelllanc, C.B. aponixis in Oirhanthivm annulatum Reprad. 9: 324--29.
daVolle, C. mns, and M.D.llayward. (Forssk) S1apf. Botanical Gazelle 139: 71- Ruden, D.M., and H. JOckIe. 1995.lI'4tolic delay
1997.ldentilicalion of alIIlIize Iinkoge 79. dependent survival idenlifies c~ of
'0
group related aponixis it Brachioria. Redei, G.P. 1964. Crossing experiences wi.h cell cyde control il the OrD5D(JhiJa
Theor. App/. Genet. 94: 439-44. poIyiloids. ArabirJopsis Inf. Serv. 1. boIstoderm. Development 121: 63--73.
PeIrO'l, DJ., N.I. Belousova, E.S. Fakina, LI. --.1965.lloo-mendelian Rutishauser, A. 1954. Die Enlwiddungsenegung
LaikO'la, R.M. Yatsenko, and IP. Sorokina. megagame'agenesis inArabidapsis. des EndosperIl15 bei pseudogomen
19B4. Transfer of some elernenIs of Genetics 51: 857-972. Ranunculus Arten. Min. naturf. Ges.
aponixis from TripsllCOOl'o lIIlIize. In DJ. Reiser, L, and R.L Fisd1er. 1993. The ovule and SchafflwMen 25: 1--45.
PelrO'l (ed.l, Apomixis and its Role in the embryo sac. PlantCeN. 5: 1291-1301. Rutishouser, A. 1969. Die embJyoIagisdlen unci
Evolution and Breedilg. New DeIIi, India: Reiser, L, Z. Modruson, LMargossian, A. cytogenetischen Grundlagen der
Oxonion Press lid. pp. 9-73. Salllllch, N. Ohad, G.W. Houghn, and R.L Somerriornpatibi&tiit. Her. Schweiz. Bot.
Poliakova, IF. 1964. Development of the male Fischer. 1995. The BELL' gene encodes a Ges. 79: S-48.
and female gamelophyIes of Arabidopsis homeodornain pro'ein involved in ponern Russell, 5.0.1979. Fine structure of
tholiano (U Heynh. IssJed Genet. 2: 12~ forllllltion in the ArabirJopsis ovule megagametophyte development in Zea
33. primordium. Cel183: 73~2. mars. Can. J. Bot. 57: 1093--1110.
PreUlS 0.1999. Chromotin silenOng and Rhoodes, M.M. 1956. Genic conlrol of - - . 1985. Preferential fertiizalian in
Arabidopsis development Arole for chrolllllSOllllll behavior. Maize Genet. CDDfJ. Plun1Iago: Ultraslructural evidence for
Polycamb profeins. Plant CB'. 11: 76~. NewsI. 30: 38--42. gornelt-leYel reeognition in an ongiospenn.
Preuss, D., B. Lemieux, G. Yen, and R.W. Dam. Rhoodes, M.M., and E. Dempsey. 1966. /'roe. NatI. Acad. SO. (USA) 82: 6129-32.
1993. Acanadianal ster*, mutation Induction of chromosome doubling 01 - - . 1992. Double fertiization.ln 5.0.
e1iminales surlace companents from meiosis by the elongate gene in lIIlIize. Russell and C. DOOlIIS (eels.!, SexlIfIl
Arabidopsis pollen and disrupts cell Genetics 54: 50~22. Reprar1JJetiDn inFIowrting Plunts.lnl. Rev.
signal~ng during fertilization. Genes Dev. 7: Richords, AJ. 1986. Plant Breeding Systems. Cytol.I40: 357-88. New York: Acadeni<
974-85. London: George Alien and Unwin. Press.
Pruin, R.E., and E.M. Meyerawilz. 19B6. Raberts, 5., and C. 8rownlee. 1995. Calcium - - . 1993. The egg cel: Development and
Charaderizalion of the genome of ilHux, fertitrsation potential and egg role in fertilization and ear~
Arubidopsis thaliDno. J.Mol. BioI. 187: activation in Fucus serratus Zygote. 3: embryogenesis. Plant CBI 5: 1349-59.
169-83. 191-97. Russell, 5.0., and D.P. West. 1994. Techniques for
Quarin, C.L 1999. Effect of pollen source and Raberts, 5., I. Gillo', and C. 8rownlee. 1994. hisloIagy of lIIlIize megospores and errGryo
pollen ploidy on endosperm forllllllion and Cytoplasmic calcium and Fvcus egg socs. In M. Freelilg and V. WalbaI (eels.),
seed selin pseudogalllllUS apornicti< motion. Development 120: 15~3. The /rIaize Honrlboo«. CaId Spring Harbor,
Paspa/um nolrt.um. Sex. I'Iont Reprad. 11: Robinson-Beers, K., R.E. Pruin, and C.S. Gasser. New Jersey: Cold Spring Harbor Laboratory
331-35. 1992. Ovule develapmen. in wild type Press. pp. 13~39.
Rabinowia PD, and E. Grolewold. 2000. Anavel Arubidapsis and two felllllle-slerile mutants. Russinova, E., and S. de Vries. 2000. Parerdal
reverse-genetic approach (SIMFI idenlifies PlantCeH4: 1237--49. cordribulion 10 plant erOOryos. Plant Ce'12:
Mutator insertions in new Myb genes. Roche, D., PCong, Z. Chen, W.W. Hanna, D.L 461~.
Plan/a 211 :887-93. Gusline, R.T. Sherwood, and POzias·Akil15. Sargonl, E. 1900. Recent work on .he resuhs of
Ral1lCJlu, K.S., PDijkhuis, A. Pereira, G.c. 1999. An apospory·specilic genonic region fertilization inongiosperms. Ann. BDI. 14:
Angenent, M.M. van Lookeren Campagne, is conserved between 8uffelgrass (Cenchrus 689-712.
and JJ.M. Dons. 1997. EMS and lransposon dliDris U ond Pertllisetum sqllflfllCllatum Sarkar, K.R., and E.H. Coe. 1966. Agenelic
mutagenesis for the isoIotion of apomictic Fresen. Plant 1. 19:203--08. anaiy5js of the origin of lIIlI'emoI hoplaids
mutants in plants. In S.M. Join, 0.5. 8rar, Radkiewia, 8. 1970. CoUose in cell walls during in lIIlIize. GeneIicI 54: 453--64.
and 8.5. Ahloowalia (eels.!, Somadanal megasporogenesis in angiosperll15. Planta. Sa'a, 5, Y. HoIIa, and S. Tabato. 1995. Structural
Variation and Induced MuIr1IiDns infrap 93: 39--47. anaiy5js of ree·A like gene inthe genome of
Improvement. Dordrechl, The Netherlands: Roeder, S. 1995. Sex and the single cell: meiosis Arabidopsis thaliDno. DNA Res. 2: 89-93.
Kluver Acadenic Pub&shen. Pp. 379-400 in yeasf./'roe. Narl. Acari. SO. IUSA) 92: Savidan, Y. 1980. Chrornasollllll and
Rondolph, LF. 1936. Developrnentalllllllphology 10450-56. embryological analyses it sexual x
of the caryapsis of lIIlIize. 1. Agr Res. 53: aponidic hybrids of Panicum IIIlIxilllJrn
881-916. Jacq. Theor. Ap';. Genet. 57: 153--56.
21 0 UtI GA".lIon
- - . 1982. Nature ethered~e de Sheridan, W.F., and J.K. Oork. 1994. Fertitlzation Spillane, C, l-P V-Ielle-Calzado, and U.
raponixie mez I'rmicum maximum Jocq. and embryogeny in moize. In M. Freetlng GrOllniklaus. 2000. Parent-of-origin effec15
Travaux etDocumenl5 ORSTOM, Paris153: and V. Walbor (eels.!, The Maize Handboolc. and seed development: genetics and
1-159. New York: Springer-Verlag. pp.3-10. epigenetics. In Y.H. HUi, G.G
--.1992. Progress in research on Sheridon, W.F., J.K. Oork, and M.G. Neuffer. Khochatourians, A. McHughen, and W.K.
aponixis and iI5 transfer 10 major groin 1995. Embryo-endmperm interoc1ion in Nip, R. Scarzo (eds.), Handbook of
uClpI. In Y. Donee, CDumos, and A. GaRais early seed development. Proc. Is/Int. Conf. Transgenic Food. New York: Macel Dekker
(eds.!. Reproductive Biology ond PIont on Apomixis: 'Hofflessing Apomixis. ANew Ine. (in press)
Breeding. Ber~n: Springer-Verlag. Frontier in PIont Science: College Station, Springer, PS., W.R. Mc(ambie, V. Sundaresan,
- - . 2000. Aponixis: Genetics and Texos,USA. ond RA. Martienssen. 1995. Gene trap
breeding. Plant Breeding Reviews 18: 13- Sheridan, W.F., E.A. Galubeva, L1. Abrhamova, tagging of PROLlFERA, on essentiol MCM2-
86. and I.N. Golubovskoya. 1999. The moc 1 3- 5-like gene in Arobidopsis. xience 268:
Sovidan, Y., and 1 Berthoud 1994. Maize x mutation alters the developmental fate of 877-80.
Tripsocum hybridization and the potentiol the hypodermal celk and their ceRulor Steinhordt, RA, ond D. Eppel. 1974. Activation
for apomixis transfer for moize progeny in the maize anther. Genetics 153: of sea urchin eggs by 0 cakium ionophore.
improvement. In Y.PS. 8ajoj led,), 933-41. Proe. Noli. Acod. Sd.IUSAI 71: 1915-19.
Biotechnology inAgriculture ood Fore51ry, Sheridan, W.F., and 8.-0. Huang. 1997. Nuclear Sundareson, V. 1996. Horizontolspread of
Vol. 25 Maize. 8erlin: Springer-Verlog. Pp. behoviar is defective in lhe moize (Zea Ironsposon mutagenesis: new uses for old
69-83. moyJ l) lethol ovu/e2 female elements. Trends Plont Sci. 1: 184-90.
Sovidon, Y., and M. Dujordin. 1992. Apomixie: la gamelophyte. Plont 1. 11: 1029. Sundaresan, V., PS. Springer, 1. Volpe, S.
prochoine revokJtion verte? La Recherche Sherwood, n, Cc. 8erg, and BA. Young. 1994. Howard, lD.G. lanes, e.Dean, H. Mo, and
241: 326-34. Inheritance of apospory in buffe~rass. RA Marlienssen. 1995. Ponerns of gene
Schena, M., A.M. Uoyd, and R.W. Dom. 1991. A Clop. xi. 34: 1490-94. artion in plant development reveoled by
steroid·inducible gene expression l'fIIem Shimamofo, K. 1995. The moleculor biology of enhancer trap and gene trap transposable
for ~ant celk. Proc. Notf. Acod. Sci. (USA) rice. xience 270: 1772-73. elements. Genes and Dev. 9: 1797-1810
88: 10421-25. Shimamato, K., Miyazaki, C, Hashimoto, H., Sumner, MJ., and Lvan Caseele. 1988. Ovule
Schieftholer, U., S. 8alosubramanian, PSieber, Izowa, I, Itoh, K., Terodo, R., and S.lida. development in Brossico campes/ris. Alight
D. Chevatler, E. Willl1On, ond K. Schneitz. 1993. Trons-octivotion and stable microscope study. Con. 1. Bot. 66: 2459-
1999. Molecular ono~ of NOZlIl, 0 integrotion of Ihe maize transposable 2S69.
gerle involved in pottern formation ond elemenl Os cotransfected with the Ac Tolioferro, CM, and E.C Boshow. 1966.
early sporogenesis during sex orgon Iransposose gene in transgenic rice plonts. Inheritance and control of obligate
development in Arobidopsis tholiono. Proc. Mol. Gen. Genet. 239: 35~0. apomixis in breeding buffelgross,
NaI/. Acod. Sci. (USA) 96: 11664--69. Siddiqi, I., G. Ganesh, U. Grossniklaus, and V. Pennisetum ci/iore. Crop Sci. 6: 473-76.
Schneitz, K. 1999. The molecular and geneti< Subbiah. 2000. The riyod gene is required Tank~ey, S.D., ond S.R. McCouch. 1997. Seed
control of ovule development. CUff. Op. for progression through female meiosis in banks ond molecular maps: unlocking
Plant. Bioi. 2: 13-17. Arobidopsis. Development127: 197-207. generic potential from Ihe wild. xience
Schneitz, K., M. Hiilskamp, and R.E. Pruitt. 1995. Singleton, W.R., and PC Mangekdorl. 1940. 227: 1063-66.
W,Id·rype ovule development in Arobidopsis Gametic Ierhok on the fourth chromosome Thoma, S., UHecht, A. Kippers, J.Botello, S.C
tholiooo: 0 light mi<ro\(ope study of cleared of maize. Genetics 25: 366-90. de Vries, and e.Somerville. J994. Issue-
whole-mount tissue. Plont 1. 7: 731-49. Skarnes, W.e. 1990. Entrapment vectors: 0 new specific expression of Ihe carrol gene EP2
- - . 1997. Dissection of sexuol organs tool for mommalian generics. lipid transfer protein gene. Plont Cell 3:
ontogenesis: 0 geneti< ono~ of ovule BioTechnology. 8:827-31. 907-21.
development in Arabidopsis tholiono. Sm~h, D.L, ond N.V. Fedoroff. 1995. LRP1, 0 Tirlopur, U.K, E. Kranz, and M. Cresti. 1995.
Development 124: 1367-76. gerle expressed in lateral and advent~ious Charocterization of isolated egg cells, in
Scon, RJ, M. Spielmon, 1 8ailey, and H.G. root primardia of Arobidopsis. Plont Cell 7: vitro fusion products and zygotes of Zea
Dickil1lOl1. 1998. Parenl-of-origin effec15 in 735--45. mays Lusing ihe technique of imoge
seed development in Arobidopsis tholiono. Sm~h, H.E., ond A.P Mitchell. 1989. A analysis and confocallaser scanning
Development 125: 3329-41. tron\(riplionol (OlI:ade governs entry into microscopy. Zygote 3: 57-64.
Sene, C, A. Bevilocquo, A. 8ionchini, F. Mangio, meiosis in Socchoromyces cerevisioe. Mol. r~sier, H, S. Marillonnet, V. Klimyuk, K. Palel,
R. Geremio, and PRossi. 1997. Cell Bioi. 9:2142-52 M. Angel Icrres, G. Murphy, and J. D.G.
Porthenogenetic octivation of mouse eggs Speulmon, E., PU. Metz, G. von Arkel, B. te lanes. 1999. Multiple independent
by mi<roinjeclion of 0 truncated c-k~ Untel Hekkert, WJ. Stiekemo, and A. defective suppressor-mennor transposon
tyrosine kinase present in spermatozoo. Pereira. 1999. A!wo-<omponenl enhancer· insertions in Arobidopsir 0 tool for
Development 124: 2267-74. inhib~OItransposon mUlagenesissystem for functionol qenomio. Plant Cellll: 1841-
Sheridan, W.F., NA Avolkino, I.! Shamrov, 18. functionol onolysis of the Arobidopsis 52
8atygino, ond I.N. GoIubovskaya. 1996. genome. Plant Celll1: 1853-66. Topping, H, W. Wei, and K. Undsey. 1991.
The mac Igerle: controlling the comm~menl Functional tagging of regulatory elem~nts
to the meioti< pathway in moize. Genetics in the planl genome. Development 112:
142: 1009-20. 1009-19.
Fro.. So. . .y to . . . . is: .... 1ar cM GeHti< Aw-H' 211
Tsugeki, R., EL Kochieva, and N.V. Fedoral!. Vijoyraghovan, M.R., and K. Prabhokar. 1984. Weinmonn, P, M. Gassen, W. Hilen, H. Bujord,
1996. Atransposan insertion in the The endosperm.ln B.M. Johri (ed.), and CGotz. 1994. Achimeric:
Aiabidopsis SSK /6gene causes on embryo- Embryology ofAngiosperms. Berlin: transodivatl)( allows telTo-cyulIIe-
defective lethal mutation. Plont J. 10: Springer Verlag. pp. 319-76. responsive gene expression in whole
479-89. Vizir, I.Y., M.L Anderson, Z.A. Wilson, and BJ. plants. Planl1. 5: 55~9.
Tsunewoki, K., and Y. Mukai. 1990. Wheal Mulligon. 1994. Isolation 01 deficiell<ies in Wesl, MAL, and JJ. Haroda. 1993.
haplaids through the Salmon method. In the Aiobidopsis genome by g-irradiation 01 Embryogenesis in higher plants: An
Y.PS. 8ajaj (ed.), Biotechnology in pollen. Genelicr 137: 1111-19. overview. Plan/CeD5: 1361--69.
Agriculture and Forestry, Voll3 Wheat. Vollbremt, E. 1994.lelhol ovule2 causes Wh~acker, M., and K. Swann. 1993. Lighling the
Heidelberg: Springer·Verlag. Pp. 460--78. aberrant embryo soc development. Maize fuse atfertilization. Developmenll 17: 1-
Turcatffe, E.L, and cv. Feaster. 1963. Haploids: NewsJener 68: 2-l 12.
high·frequency production ham single- Vollbremt, E. and S. Hake. 1995. Deliciell<Y Willemse, M.T.M. 198 J. Polarity during
embryo seeds in 0line 01 Pimo canon. onolysis 01 lemale gametogenesis in maize. megalpologene5is and
Science 140: 1407-D8. Dev/. Gene'. 16: 44-63. megagometogenesis. P/rytomarphology
Urango lA., R.A. Pedersen, and J.Arechaga. Wokono, A., ond S. Uemota. 1987. Adventive 31: 124-34.
1996. Parthenogenetic activation 01 mouse embryogenesis in OIrllS. I.The occurrent:e Willemse, M.T.M., ond J.L van Wenl. 1984 The
oocytes using cakium ionophores and 01 odventive embryos without pollinDlion lemale gamerophyte. In B.M. Johri (ed.l,
protein kinDse ( stimulDlan. 1nl. 1. Dev. I)( fertilizotion. Ainer. 1. Bol. 74: 517-30. fmlxyo/ogy ofAngiosperms. New Yl)(k:
BioI. 40: 515-19. Walbot, V. 1992. Strotegies 11)( mutagenesis and Springer-Verlag. pp. 159-96.
van Dijk, PJ., I.G.O. Ies, M. Fo~ue, and 1. Bokx· gene cloning using tronsposon tagging and Wilson, C, HJ.Bellen, and WJ. Gehring. 1990.
Sehalman. 1999. Crosses between sexuol T-DNA mutogenesis. Ann. Kev. Plant Position el!ecIs on eukaryatic gene
and apomictic dandelions ITOfllxocuml.1I Physiol. Plant Mol. Bioi 43: 49-82. expression. Anoo. Kev. Cell. BioI. 6: 679-
The breokdaWl1 01 apomixis. Heredity 83: - - . 1994. Overview 01 the key steps in 714.
71 5-21. aleurl)(le developmenl.. In M. Freeling and WiIson, C, R.K. PeorlOl1, HJ. BeUen, U.
van Dijk, P, and J.van Damme. 2000. Apomixis V. Wolbot (eds.l, The Maize Hondboolc. O'Kone, U. Grossniklaus, and WJ. Gehring.
technology and the porodax 01 sex. Trends New bk: Springer Verlag. Pp. 78--80 1989. P-elemenl-mediDled enholl<er
Plonl xi5: 81-84. - - . 1996. Sources and cansequent:es 01 detection: on efficient method fl)( isolating
Vielle, J.p, 8.l. 8urson, E.C BosOOw, and MA. phenotypic and genotypic plasticity in and characteriZing developmentally
Hussey. 1995. Eor~ fertilization events in flowering plants. Trends Plonl Sci. 1:27- regulated genes in Drosophila. Genes and
the sexual and oposporous egg apparatus 32. Dev.3: 1301-13.
of Penniselum ci/iore ILl Link. Plant J. 8: Wahers, M.S. 1985. Meiosis reodiness in U1ium. Xio, Y., BJ. Nikolau, and PS. Schnoble. 1996.
309-16. Can. 1. Genel. Cytol. 27: 33-38. Cloning and characterization of CER2, on
Vielle-(olzada, J.p, R. Baskar, and U. Weatherwax, P1919. Gometogenesisand Arabidopsis gene that oIfecls cuticuIor wox
GrDllniklous. 2000. DeIoyed activation 01 fecundotion in Zea lIWlys os the basis 01 accumulation. Plan/ CeB8: 1291-1304.
the paternol genome during seed xenia and heredity in the endasperm. Bull. Yadegari, R., 1. Kinoshita, O. lDIan, G. Cohen, A.
development. Nature 404: 91-94. T!)(fey Bol. Oub46: 73-90. Kotz, Y. Choi, A. Kotz, K. Nokashimo, JJ
'{lelle-(olzado, J·P., CF. Crone, and D.M. Stel~. Webb, M.C, and B.E.S. Gunning 1988. The Harodo, R.B. GoIdberg, R.L Fischer, and N.
1996. Apomixis. The asexual revolution. microtubular cytoskeleton during embrya- Ohad. 2000. Mutations inthe RE and MEA
xience274: 1322-23. IOC development in Aiobidopsis Iho/iono. In genes thot ell<ode interacting Po/ycomb
'{lelle-(oizoda, J.p, J.M. Moore, W.B. Gogliono, R.B. Knox, M.B. Singh, and LF. Traioni proteins cause porent-of-origin el!ecIs on
and U. Grallniklaus. 1998. Altering sexual (eds.l, PoRination 'B8. Melbourne, seed development by I5s1inct mechanisms.
development in Aiobidopsit 1. Plant BioI. Austra~a: School of Bolany, Univenity of Plan/ CeD 12: 2367--82.
41:7~1. Melbourne. Yamomoto, M. 1996. Regulotion 01 meiosis in
'{lelle-(alzada, J-P., J.ThOll1Dl, CSjlillane, A. - - . 1990. Embryo soc development in fission yeast. CelI5/rVd. FOOd. 21: 431-
(aiuccia, M.A. Hoeppner, and U. Aiobidopsis thaliano. I. Megalpologene5is, 36.
Grollniklaus. 1999. Mainlenonce of induding the microtubular cytoskeleton. Yang, W-C, and V. Sundoreson. 2000. Genetics
genomic imprinling DI the Aiobidopsis Sex. PlonI Keprod. 3: 244-56. of gamelophyte biogenesis in Arabidopsis.
medeo Io<us requires zygotic 001141 - - . 1991. The microtubular cytoskeleton Curr. Op. PIont 56. 3: S3-57.
activity. Genes Dev. 13: 2971--82. during the develapment of the zygote, Yang, W-C, D. Ye, J.Xu, and V. Sundoreson.
Villanueva, J.M., J. 8roodhvest, B.A. Hauser, U proembrya and hee-nudeor endosperm in 1999. The SPOKOCYTE1I5S gene of
Meisler, K. Schneitz, and CS. Gasser. 1999. Aiabidopsis thaliano Il.) Heynh. Plonto Aiobidopsis is required fl)( initiation 01
INNER NO OlJTfK regulates abaxial! 184: 187-95. sporogenesis and ell<odes 0 novel nudear
adaxiol ponering in Aiobidopsis ovules. - - . 1994. Embryo soc developmenl in protein. Genes Dev. 13: 2108-17.
Genes Dev. 13: 3160--69. Aiobidopsis thaliano. 11. The cytoskeleton Zaki, M.A.M., and H.G. Oi<kinson. 1990.
Vinkenoog, R., M. Spielmo, S. Adorns, R.L during megogametogenesis. Sex. PIonl Strudural changes during the first divisions
Filcher, H.G. Dickinson, and U Sean. Keprod. 7: 153--6l of embryos resulting frl)(n lIlther and free
2000. HypomehtylDlion promotes Weber, 01 1983. Monosomic ano~sis in di~oid microspore rulture in Brassica nDpllS.
outanomous endosperm development and crop plants. In PK. Gupta and U. Sinha PrmopklsnxJ 156: 14~2.
rescues post-Iertilization letholity in fie (eds.), Cytogenetics ofClop Plants. New Zirnmermon, J.L 1993. Somatic embryogenesis:
mutants. Plonl Cell 12: 2271-82. Delhi, India: Mocmillan India ltd. Pp. 352- o model fl)( early development inhigher
78. plants. Plonl CeH 5: 1411-23.
Chapter 13
Induction of Apomixis in
Sexual Plants by Mutagenesis
UTA PRAEKELT AND ROD SCOTT
Crosses with pollen from different species linkage of several genes, a possibility that is
result in various proportions of matromorphs consistent with the lack of recombination
arising from the parthenogenetic development observed between molecular markers
of unreduced egg cells (Eenink 1974b). Thus associated with apomixis (Grimanelli et al.,
in Brassica, two of the most important Chap. 6). Even if a single mutation, perhaps
ingredients of apomixis are revealed by distant resulting in the inhibition of meiosis, was
pollination: (i) the presence of unreduced responsible for the evolution of apomixis, it is
embryo sacs, and (ii) the inherent capacity for likely to have occurred in a background that
parthenogenesis. Haploid parthenogenesis permitted its expression (Mogie 1988) and has
has recently been induced both in Arabidopsis therefore evolved only in a subsection of genera
and in Brassica [uncea by the application of and families. For these reasons, we do not
brassinolide, a steroid hormone first isolated expect that a single mutation in a sexual plant
from Brassica pollen (Kitani 1994). could result in the production of viable seed in
the absence of fertilization. However, given the
4. Endospenn development. In autonomous
variety of apomictic forms that can be
apomicts, endosperm development occurs
distinguished, such as in diplospory and
spontaneously, but in the case of
apospory, apomixis has probably arisen
pseudogamous apornicts. it depends on
independently in different species and perhaps
fertilization of the central cell nucleus by a
involves different genes in each case.
sperm nucleus. In some cases, pollination
Consequently, there may be ample
without fertilization has been suspected of
opportunities for the induction of some
triggering endosperm development.
element of apomixis in a sexual plant by
Endosperm plays a crucial role in the
mutations in a number of different genes.
formation of viable seed, and requires special
consideration in the design of a mutagenesis An important aspect to consider is the apparent
screen. The problem of the endosperm is dominance, in many cases, of apomixis over
discussed later in more detail. sexuality (Nogler 1984; Mogie 1988; Leblanc
et al. 1995). Mutations that completely abolish
Genetic Control of Apomixis
gene function, such as deletions, can be
With perhaps one exception (Carman 1997;
dominant only if expression of that gene is
Carman, Chap. 7), it is now generally accepted
that apomixis has evolved from sexual subject to gene dosage. One of the hypotheses
that have been put forward on the control of
ancestors by mutation rather than being a
apomixis is that the responsible gene(s) encode
consequence of polyploidization and
regulatory functions that initiate or repress
heterozygocity. This is an important
certain developmental programs (Koltunow et
assumption for mutagenic approaches to the
al. 1995).If apomictic development is repressed
study of apomixis. Much discussion focuses
in sexual plants by a negative regulatory
on whether apomixis is regulated by a single
protein whose concentration is crucial, then a
gene and could therefore be induced by a
reduction in the level of this protein could be
single mutation in a sexual plant. Given the
sufficient to induce a developmental pathway
different components of apomixis, it seems
that is suppressed only when two copies of the
more likely that a number of mutations were
gene are present. The fact that many apomicts
needed in the evolution of a viable apomict
are facultative, and that the proportion of
from a sexual ancestor. The apparent single
apomictic progeny can be influenced by
locus inheritance that has been reported in
environmental factors, supports the hypothesis
several cases could be explained by the tight
........ ., Apo.ixiI io5...01 P1oot, ~........ 215
The above mutants of maize and barley prematurely. In this mutant, parthenogenesis
illustrate that mutations at a number of loci can is the only element of apomixis that has been
induce one element of apomixis-the induced. The formation of a perfectly well-
formation of unreduced gametes. Further developed endosperm, which supports the
development is dependent on fertilization, and production of a viable seed, is presumably
normal endosperm development depends on facilitated by normal fertilization events in the
a balanced maternal to paternal ratio. central cell that result in a genomically
balanced endosperm.
The preferential detection of unreduced eggs
by crosses with tetraploids is a useful reminder 3. Aposporous mutants. In pearl millet, two
of the importance of choosing a sui table pollen mutants that produce aposporous embryo sacs
parent for mutagenesis studies involving have been reported. The first, female sterile ifs),
pollination. On one hand, mutants with is a recessive mutation induced by radiation
unbalanced endosperm may be difficult to treatment (Hanna and Powell1974; Arthur et
isolate initially because of low viability; on the al. 1993). Homozygous mutants are female-
other hand, the shrunken endosperm sterile but produce normal viable pollen.
phenotype could serve as a useful criterion in Mutant ovules are small and immature
the selection for meiotic mutants. compared with the wild type, and only about
half of them contain embryo sacs. Of these, the
2.Parthenogenetic mutants. The EMS-induced
majority are multiple embryo sacs that appear
hap mutant of barley was initially isolated in
to be aposporous, although sexual embryo sacs
the form of a chlorophyll deficient mutant
are observed in some ovules. Only a very small
containing at least three linked mutations, tig,
proportion of ovaries display any endosperm
let (pollen lethality), and hap (Nielsen 1974).
or pro-embryo development, and all ovules
Presence of the hap allele results in a low
degenerate five days after pollination. Pollen
frequency of haploid progeny. After separation
tube growth in the mutant is abnormal, and
of hap from the other two mutations, Hagberg
the inhibition of fertilization has been proposed
and Hagberg (1980) showed that hap is
to explain the absence of seed set.
incompletely dominant over the wild type
allele: heterozygous thap / +) plants produce The second, stubby head, was discovered in
3-6% haploid progeny, whereas homozygous progeny of seed treated with both thermal
(!lap/hap) plants produce up -to 40% haploid neutrons and diethyl sulfate (Hanna and
progeny. Perhaps not surprisingly, crosses Powell1973). This recessive mutation causes a
between a homozygous mutant (!lap/hap )and pleiotropic phenotype, including twin ovules,
wild type (+ / +) plants produce different results shortened internodes, flattened stems, and a
in the F" depending on whether the mutant is stubby inflorescence. It produces both normal
the male or the female parent: a hapfhap female sexual embryo sacs in some ovules and
plant pollinated by wild type (+ / +) pollen multiple embryo sacs, which arise from
results in a high frequency of haploid F, nucellar cells, in others. Test crosses confirmed
progeny, whereas no haploids are produced that stubby head is a facultative apomict,
when a wild type (+ / +) female plant is crossed producing maternal progeny at frequencies
by a !lap/hap male. This indicates that the hap ranging between 23% and 77%.
locus acts only through the maternal tissue, Mircrosporogenesis in this mutant is normal,
either to prevent fertilization of the egg cell or however, seed set is low; this has been
to stimulate the egg cell nucleus to divide attributed partly to nonfertilization because of
competition between the multiple embryo sacs.
220 Uta PnoUIr _.04 *"
Stubby head could indeed be a very useful both give rise to shrunken seeds, and since it
mutant, as it produces viable seed by apospory. has been shown that an unbalanced ratio
Although the nature of the mutation is not affects seed size in Arabidopsis too, this
known, the mutation affects several elements phenomenon could be used as a screen for the
of apomixis in a single step. Whilst this could isolation of Arabidopsis mutants that produce
be due to a single gene mutation, the unreduced egg cells.
pleiotropic nature of the mutant phenotype
could well indicate that it is caused by a Current Approaches to the
deletion encompassing a number of genes. Isolation of Apomictic Mutants
4. Conclusions. The above examples of in Model Sexual Plants
mutants with apomictic characteristics The most important precondition for the large-
illustrate a number of points relevant to scale screening of apomictic mutants is the
mutagenic approaches in model plants. availability of male sterile lines that do not set
seed in the absence of pollination. Arabidopsis
The most important conclusion is that all the
is ideally suited for such an undertaking
elements of apomixis can be induced by
because it offers several ways of establishing
mutation in sexual plants. In one case, stubby
male-sterile lines. The most useful of these
head, several elements were induced
causes conditional male sterility, which allows
simultaneously to produce a viable form of
the propaga tion of homozygous seed by
facultative apospory. It would be very
selfing under permissive conditions. In
interesting to analyze these mutants at the
addition, Arabidopsis provides easily scored
DNA level, but the isolation of genes from
dominant and recessive markers for
these crop species would not be a simple task.
subsequent screening of mutagenized
Unlike Arabidopsis, the cloning of genes via
progeny.
their mutant alleles is not routine.
The most obvious screen for apomictic
None of the mutants described herein were
mutants of a male-sterile sexual plant is for
originally isolated in screens for apomictic
seed set in the absence of pollination. However,
characteristics. The majority were isolated as
there have been no reports of mutants (in any
reduced fertility mutants and some as
model plant) that produce viable seed in the
pleiotropic mutants. Therefore, it seems worth
absence of pollination. This may well be
considering which kinds of mutant
because the number of progeny screened in
phenotypes, other than apomixis itself, could
this way was not large enough to detect such
be screened for in Arabidopsis. First, it might
mutants, as these could indeed be very rare.
be worthwhile to screen for reduced fertility
For the reasons already outlined, however, it
mutants, a simple screen that would involve
also seems likely that more than one gene
testing for seed set in the M 2• A large number
mutation is required to induce a viable form
of male-sterile mutants have been isolated
of apomixis.
from Arabidopsis. Some are present in the
collection of T-DNA tagged lines (available Current efforts in several laboratories are
from the Arabidopsis Stock Centers at directed toward the induction of partial
Nottingham, U.K. and in Ohio, USA.); these apomictic development. Basically two main
could all be screened for maternal progeny types of screens are being conducted. The first
after pollination with a dominantly marked identifies mutants that show partial
paternal line. Second, the tri and et mutants development of fruits in the absence of
holIodIoo 01 ApooUis ;, Se.... rIoots ~......... 221
pollination. Normally, the pistils do not Chaudhury used as the parental plant a
elongate in unfertilized flowers of Arabidopsis, pistillata (pi) mutant, which has no petals or
whereas fertilized flowers produce long fruits stamens. Since this plant is male-sterile in all
(siliques). In the expectation that partial seed conditions, it was mutagenized in the
development would lead to visible silique heterozygous F1 population after crossing to a
elongation, the screen is simply for elongated wild type. Again, 50,000plants, in this case the
siliques on male-sterile plants under petalless portion of the M 1 generation, were
nonpermissive conditions. Whether these are screened for elonga ted siliques. and six
due to autonomous embryo or endosperm mutants were obtained. Three of these ifis1 to
development or both must be established by fis3-fertilization independent seed) were
further cytological analysis of the siliques. The characterized in detail.
second screen is for mutants that produce
All three fis mutations are gametophytic,
autonomous embryos but which require
giving rise to seed-like structures from 50% of
fertilization of the polar nuclei for endosperm
ovules in heterozygous fis/FIS plants. These
development, i.e., for pseudogamous
mutant seeds contain diploid endosperm that
apomicts. The screen is simply for maternal
develops up to the stage of cellularization and
progeny among a population of male-sterile
then atrophies. The majority of mutant seeds
mutants that are pollinated by pollen with a
have no embryos, and only a small proportion
dominant marker.
of fis1 and fis2 seeds contain proembryos
Screening for Elongated Siliques in the arrested at the globular stage. However, after
Absence of Pollination pollination of mutant flowers, each of the
Two laboratories have recently reported the mutant seeds contains an embryo at the more
isolation of Arabidopsis mutants that show advanced torpedo stage, including fislfie3,
partial seed development in the absence of which shows no embryo development in the
fertilization (Ohad et al. 1996; Chaudhury et absence of fertilization. By testing for a
al. 1997).In one case, the parental material was dominant pollen marker (GUS), it could be
a conditional male-sterile pop1 mutant, which shown that these torpedo-stage embryos
is fertile at high humidity but does not produce consist of a mixture of zygotic and maternal
functional pollen at low humidity (Ohad et al. embryos. Even after fertilization by a wild type
1996). Homozygous pop1 seeds were sperm nucleus, the resulting embryo is
mutagenized with EMS, and 50,000 M 1 plants aborted, suggesting that the wild type FIS/FIE
were screened for silique elongation under allele is essential for female gametophyte
nonpermissive conditions. A total of 12 lines development. Thus, the fis/fie mutations can
were isolated, each of which had partially only be transmitted via pollen, by fertiliza tion
elongated siliques appearing as sectors on the of a wild type ovule. Ohad had previously
M 1 plants. The characteristics of one of these shown that the endosperm of fie3 too is
mutants, fie (for fertilization independent fertilized in the presence of pollen, suggesting
endosperm), were described in detail. Because that the mutation does not present a block to
this mutant, now called fie3, as well as two fertilization, either of the egg or the central cell.
more fie mutants now appear to be allelic to
The primary effect of the fislfie mutations is the
the mutants subsequently obtained by
premature proliferation of the endosperm in
Chaudhury, they will not be described
separately. the absence of pollination. A secondary effect
is the formation of a normal seed coat from
the integuments of the ovule and the
222 Uta Pratktlt _Iod 5<011
elongation of the silique. As discussed by Ohad male-sterile seed parents are waxless ecerijerum
et al.. these two processes both involve mutants, eerl and eer6-2 (pop'l), which are self-
maternal developmental programs that must fertile at high humidity and male sterile at low
be induced by signals from the developing humidity. M[ plants have been screened for
gametophyte. The discovery of embryoless sectors with elongated siliques, and M 2
seeds with normal seed coat and elongating families derived from selfed individual M j
siliques suggests that the signal originates in plants have been screened for segregation of
the endosperm rather than in the embryo. plants containing elongated siliques.
It is not clear what the relationship is between Screening for Dominant Mutations in the
fislfie and the genes controlling apomixis. The M, after Pollination
fact that the mutations affect a late We are currently conducting a screen for
developmental stage after embryo-sac dominant mutations that result in the
differentiation make it unlikely that they are production of autonomous embryos but that
alleles of the apomixis genes per se. They may require fertilization of the central nucleus
could, however, be the targets of apomictic for the development of endosperm. Although
regulatory genes. It is noteworthy that fislfie most natural pseudogamous apomicts have
are female lethal mutations, as this is consistent fertile pollen, induced mutations that give rise
with the observation that apomixis appears to to autonomous embryos as a result of meiotic
be controlled by a dominant gene, the recessive disturbances could also have defective pollen.
(sexual) allele of which may be required for Such mutants would not be recovered in the
some function in development. It would be M 2 generation after selfing. For this reason, and
interesting to test the effect of these mutations to take advantage of the smaller number of
in a tetraploid background to see if diploid plants that need to be screened to obtain a
embryo sacs heterozygous for lis/fie have mutant, we decided to screen in the M j .
autonomous endosperm development and are In one version of the screen, mutants are
able to transmit the mutant allele.
detected in the form of maternal progeny after
The three mutants described above were pollination by a pollen parent containing a
identified in two independent large-scale dominant marker. This screen depends on the
screens of M j and M 2 progeny, respectively; formation of fertile seed and would not detect
each mutant was isolated in the form of several mutant apomictic seeds that fail to germinate.
allelic mutations, suggesting that the level of Therefore, we devised a second version that
mutagenesis must have been near saturation. allows the detection of autonomous embryos
The fact that all mutants detected in these two at an immature seed stage, and which can be
screens show autonomous endosperm, and employed should the first version fail.
that many of the seed-like structures contained The seed stock used for mutagenesis is a
no embryos, could mean that a mutation conditional male sterile line (DTA Q3) in the
resulting in autonomous embryo develop- Arabidopsis C24 background. This line is
ment, but one that does not activate homozygous for a transgene encoding a
endosperrn, would not be detected in a screen temperature sensitive diptheria toxin under
for elongated siliques. the control of the tapetum-specific promoter
Finally, a screening program for elongated A9. Growth at 18°C results in male sterility due
siliques in Arabidopsis has been described by to the absence of functional pollen. At 26°C,
Ramulu et al. (1997). Again, the conditional the plants are fertile, allowing the propagation
of homozygous seed by selfing.
lodw<tioI of Aponoili. io Sex'" PI.t. ~ Mot.,...a 223
Seeds were mutagenized with EMSfor 6 hours, the pollen parent (Figure 13.1). The WS
washed, and dried for storage on filter paper. ecotype produces a highly characteristic leaf
The seeds were sown in batches of 100-200 at rosette consisting of tightly-spaced round
regular intervals, and seedlings were grown leaves with abundant trichomes that are
to flowering in individual pots. After growth distinguishable from the smooth elongated
at 18°C for at least one week, the plants were leaves of C24 at an early stage of growth.
cut back to two mature flowers on a single Heterozygous seedlings are indistinguishable
inflorescence. These M. flowers were from the homozygous pollen parent WS.
pollinated by a transgenic line in the WS
As the screen is conducted in the Mp mutants
background homozygous for a Basta herbicide
that give rise to maternal progeny are most
resistance gene. Progeny resulting from cross-
likely to be heterozygous for the mutation.
fertilization are resistant to Basta, whereas
Maternal progeny could potentially arise in a
maternal progeny are Basta-sensitive (Figure
number of different ways: (i) the partheno-
13.1). However, since Basta-sensitive plants
genetic development of a reduced egg in
cannot be rescued, this selection presents a
combination with pseudogamous endosperm
problem in cases in which only a small number
resulting in a haploid seedling; (ii) the parthe-
of seeds are available. We are therefore
nogenetic development of an unreduced egg
currently using a simplified selective criterion,
cell with pseudogamous endosperm and
spoecifically, the dominant leaf morphology of
B. (24 OTA x WS
~~ ~
C. (24 OTA x EM2
O. fl/OTA x TZ/fZ
autonomous endosoerm with parthenogenesis
haploid diploid
E. OTA xenda·GUS ~~ c:::$)U,;UQO!!:>
IQ QQ
Figure 13.1 "Uncoup6ng" genetic screens for apomictic mutations in Arabidopsis.
The generatian of Arobidopsis mutants expressing autonamaus apomixis is regarded as unlike~ for reosons ourlined in the text.
We therefore propose a number of screens that remave the requirement for mutations simultaneously conditioning both
parthenogenesis and autonomaus endosperm develapment. These 'uncoupling" screens aim toidentify porthenogenetic mutants in
the presence of a sexual endosperm and visa versa. (Seed parents: (24 OTA = (24 ecatype with elongated leoves, male sterile at
18 x C. tz/ OTA =thiamine auxotrophic, male sterile at18 x C. Pollen porents: 063 =WS ecotype hamozygous for Rasta herbicide
resistance. EM2 = promoter-trap line expressing GUS on~ in embryos. TZ/fZ = any wild type. endo-GUS = transgenic for GUS
reporter under the control of endosperm-specific promoter. For more details and references, see text)
resulting in a diploid seedling; and (iii) selfing the early detection of maternal progeny that
of the conditional male-sterile plant by allows screening at a much higher density and
infrequently fertile pollen. As illustrated in provides results soon after sowing. If this
Figure 13.1,haploid parthenogenesis arising in screen for maternal progeny mutants is not
the M] should produce equal frequencies of successful, the second version is employed
maternal and hybrid progeny. In the case of that allows the identification of autonomous
diploid parthenogenesis resulting from a embryos (Figure 13.1). For this purpose,
dominant mutation, all progeny could be mutants are pollinated by a pollen parent
maternal. However, in both cases the expected transgenic for an embryo-specific GUS
frequency would be less if the mutation results reporter gene (Topping et al. 1994). GUS
in facultative parthenogenesis or if the mutation expression is easily detectable in the
had incomplete penetrance. developing seeds from eight days after
pollination, well in advance of seed ripening.
Considering pseudogamous development of
GUS-negative embryos again could arise from
the endosperm, both haploid parthenogenesis
haploid or diploid parthenogenetic egg cells
and selfing would result in a balanced maternal:
or from selfing of the plant.
paternal ratio, and therefore the size and shapes
of seeds are expected to be normal. However, In principle, the viable and nonviable seed
in the case of diploid parthenogenesis, the screens could be carried out simultaneously
endosperm may have either a balanced or by combining the use of the thiamine-
unbalanced m:p ratio, depending on whether auxotrophic male-sterile seed parent with a
one or both sperm nuclei fertilize the central pollen donor containing the embryo-specific
cell. Therefore the resulting seed could be either GUS reporter. One of the siliques resulting
of normal size or smaller (maternal excess from the cross could be stained for GUS
phenotype). Small seeds might be activity at an immature stage and the other
distinguishable at the time of sowing, and left on the plant until seed maturation.
appropriate steps could be taken should they Putative apomictic candidates can be further
not readily germinate on normal soil. tested to confirm whether endosperm
development is autonomous or
The scale of any screening program involving
pseudogamous (Figure 13.1). For this purpose,
pollination is limited by its labor intensity and
a pollen donor line has been established in the
by the space required for growing the progeny.
ecotype WS that is transgenic for an
For future screens, we have therefore
endosperrn-specific marker gene, the GUS
incorporated a recessive marker into the seed
gene under the control of a high molecular-
parent that is detectable at a very early stage of
weight glutenin wheat gene (Colot et al. 1987).
seedling growth. We crossed the conditional
This is expressed only in the cells of the
male sterile line DTA Q3 with the thiamine
developing endosperm. After crossing with
auxotrophic tz mutant, and selected progeny
this marker line, GUS expression is tested in
homozygous for both conditions (DTAtz).After
siliques at an immature stage, before
pollination with wild type pollen, heterozygous
endosperm absorption. This system could also
progeny are thiamine prototrophic, but
be used to screen for autonomous endosperm
maternal progeny are auxotrophic (Figure 13.1).
mutants independent of embryo formation.
Homozygous tz mutants emerge with green
cotyledons, but the first true leaves are white. If the cited screens for viable and nonviable
These seedlings can be rescued by spraying apomictic seed prove unsuccessful, a screen
with thiamine. The advantage of this system is involving pollination will be conducted in an
M 2• This allows the detection of recessive as of new mutations, which can be screened
well as dominant mutants. Also, it is well under non permissive conditions for apomictic
known that mutations occur in sectors of M) mutants.
plants, and by screening only a single
For transposon mutagenesis in Petunia,
inflorescence per plant, many mutants may be
Ramulu et al. (1997) are using a two-element
lost. This loss can be avoided by screening in
transposon system found in Petunia, which
theM 2
consists of a nonautonomous element, dTph1,
Transposon Mutagenesis for the and an autonomous element carrying the
Isolation of Apomictic Mutants of transposase, ACT1 (Doodeman et al. 1984;
Arabidopsis and Petunia Gerats et aI. 1990). A line containing more than
Transposon mutagenesis, like T-DNA tagging, 200 copies of dTph1, which produces a high
creates mutations by insertion of the frequency of unstable mutations in selfed
transposon into a gene. Its advantage is that progeny, was used to establish a number of
tagged genes can be isolated by using the transposon genotypes, which were each
inserted sequence as a molecular probe. Also, crossed with a conditional male-sterile plant.
a large number of mutants can be produced Male sterility in this line results from the
simply by repeated selfing of the plants. This absence of flavonols, which is caused by a
approach is currently being applied to chalcone synthase antisense gene (Ylstra et al.
Arabidopsis and Petunia, as described in detail 1994).The application of flavonols, which are
by Ramulu et aI. (1997). We shall briefly required for pollen tube growth, restores
summarize the main points. fertility and allows selfing. Plants homozygous
For Arabidopsis, a two-element system, derived for the male sterility phenotype will be selected
from the maize transposable element En- 1, in F2 populations, and screening for apomictic
was used (Aarts et aI. 1995). The maize mutants will be conducted on a large number
transposon has a 13 bp inverted repeat at each of F3 and F4 plants in the absence of flavonoIs.
terminus and encodes a transposase required Branching Outin the Brassicas
for transposition. The two-element system A benefit of mapping data from several
consists of a nonautonomous "wings-clipped" important crop plants and from Arabidopsis has
En-transposase under the control of the CaMV been the discovery that groups of genes within
355 promoter and a nonautonomous mobile large segments of the chromosomes are
I-element with flanking inverted repeats that arranged in the same linear order between
has been inserted into a kanamycin resistance related species regardless of differences in
(nptIl) gene. Both elements are contained genome size (F1avell and Moore 1996). In
within aT-DNA tha t also carries a hygromycin many cases, molecular markers identified for
resistance marker for selection of one species are found to map to corresponding
transformants. Several lines that contained locations in a related species. This high level
about 20 I-elements and the En-transposase of synteny can be exploited for the isolation
were crossed with homozygous cer1 and cer6-2 of genes from species with large genomes. The
mutants, and homozygous male sterile lines homologous gene can first be isolated from a
were selected from the segregating F 2 related species with a small genome, such as
population. Propagation of these lines for Arabidopsis, where fine mapping and
several generations under permissive chromosome walking are feasible; it can then
conditions is expected to create a large number be used as a probe for direct isolation of the
gene from the species of interest.
226 Ut. P....elt....od 5<011
Of particular relevance to our current efforts complete in a ma tter of a few years. Therefore,
is the observed synteny between the genomes it will be possible to use the markers that
of Arabidopsis and Brassica, a genus that cosegregate with apomixis in Arabis for the
contains many important crop species and, identification of the sexual alleles
like Arabidopsie, belongs to the family corresponding to the apomixis locus, and
Brassicaceae (Kowalski et al. 1994). By consequently, the apomixis genes themselves.
extension, we expect that there is synteny
between A rabidopsis and Arabis,another genus Conclusions and Perspectives
of the Brassicaceae containing several The various examples of mutants in non-
apomictic species. We have initiated a model species and the recently discovered
molecular study of Arabis with a view to Arabidopsis "autonomous end osperrn "
exploit the mapping data available from mutants illustrate a number of points about
Arabidopsis to isolate the apomixis gene(s). mutagenesis for the induction of apomixis in
sexual plants. First, it seems that all elements
The so-called Arabis holboellii complex is a
of apomixis can be induced in sexual species
collection of closely related species at several
by mutagenesis. Which element of apomixis
base numbers and ploidy levels from diploid
can be identified is clearly dependent on the
to triploid, tetraploid and hexaploid. Apomicts
kind of screen employed. Many serendipitous
are common, particularly in the triploid
discoveries have been made from pleiotropic
species, but are also found at the diploid (211 =
effects of the mutations, such as shriveled seed,
14) level (Bocher 1951; Roy 1995). However,
small plant size, or male sterility, which
as pointed out by Carman (personal comm.)
suggests that these parameters should be
this diploid species is likely to be
included in screening programs. In Arabidopsis,
paleopolyploid. Apomixis is of the
previous efforts have relied on screens for
diplosporous type, and endosperm formation
elongated siliques and have resulted in a single
is pseudogamous. It is not clear, however,
category of mutants that produce autonomous
whether pollination results in fertilization of
endosperm. Whilst these mutants have
the endosperm nucleus or is only required to
provided important insights into the
trigger endosperm development (Roy 1995).
interrelationship between fruit development
Pollen viability is high in apomictic Arabisspp., and endosperm proliferation, they illustrate
and therefore the diploid apomict would be the necessity for more sophisticated screens
an ideal candidate for the establishing to isolate mutants with autonomous embryos.
mapping populations wi th the sexual species, One of these, involving post-mutagenesis
such as A. drummondii, The collection of pollination, has been described in detail and
molecular markers available in Arabidopsis can is expected to identify new mutants not related
be used to identify markers associated with to fis/fie.
apomixis in Arabis. Alternatively, if none of the
Evidence gathered so far from mutants in the
Arabidopsis markers show heterozygocity
Brassicaceae suggest that (i) the diplosporous
between the Arabis species to be mapped, new
type of apomixis could be induced more easily
markers can be identified in Arabis and
in Arabidopsis than apospory, since diplospory
subsequently mapped to the Arabidopsis
is found in Arabis , and (ii) the Brassicaceae
genome. Much of the Arabidopsis genome is
may be predisposed to the pseudogamous
already available in the form of yeast or
type of apomixis as found in Arabis. That
bacterial artificial chromosomes (YACs or
Arabidopsis might be predisposed to
BACs), and sequencing of the genome will be
hodo<tioto of Apooilisio s..... "'-ts ~....... 227
pseudogamy can be concluded from the should eventually enable us to understand the
induction of haploid parthenogenesis at high regulation of these traits and to manipulate
frequency by the application of brassinolide, them in the best interests of agriculture.
the steroid hormone present in pollen tha t may
be the trigger for endosperm development References
following normal pollination (Kitani 1994). Aarls, M.G.M., P. (OIzoon, WJ. Stiekema, and A. Pereira. 1995. A!wo
element Enhancer· Inhibitor transposon system inAmbidopsis
The fact that to date mutagenesis has not tholiaoo. Mol. Gen. Genef. 247: SSS-M.
AhoKas, H. 1977. Amutant afbarley: Tri~aid inducer. Borley Genetics
resulted in fertile maternal seed confirms the News/eNer 7: ~.
hypothesis that viable apomixis can only be Ar!hur, L, P. OZios·Akins, and W.W. Hanna. 1993. ~ sterile llIltant in
obtained in species that have acquired the peorl millet: evidence far initiatioo afapospory. 1.Hered. 84: 112-
15.
necessary preadaptations. A long-term Asler, S. 1966. Effem afmutagen trealment an some apamiclic Pofentil/o
consideration in our pursuit of apomictic species. Hereditos 55: 249--65.
- - . 1980. Gametaphylic apomixis: elements and genetic
mutants in Arabidopsis may well be to combine
regulation. Hereditas 93: 277-93.
mutations obtained from different screening Asker, S., and LJerling 1992. Apomixis inPlonts. Bom Ratoo. Aorida:
procedures. For example, it is possible that a (I( Press.
Baker, B.S., A.T.C (arpenter, M.S. Esposita, R.E. Esposita, and L Sondler.
mutation that produces unreduced embryo 1976. The genetic cantral afmeiosis. Ann. Rev. Genet. 10: 53--134.
sacs, in combination with a fislfie mutation Bashow, lC, and BJ. Holf. 1962. Elfem afirradiatioo allllJORlicti<
(perhaps in the heterozygous state), would daltlSgrass. Clop 5<i. 2: SOI--0-t
8icknell, RA. 19940. Mi<raprapagation al Hierodum aurantitKum. Plant
result in a viable form of apomixis. Although (ell flSsue Orgon (uh. 37: 197-99.
such experiments could be carried out without - - . 1994b. Hierodum, a model system lor studying the molecular
further knowledge of the genes involved, an genetics alapanixis. Apomixis HewsleNer 7: 8--10.
- - . 1994<. Evidence far the transposition al the AtIranspasoble
important step toward a controllable system elementlrarn maize, in lhe lacu~ati'le apomict, HienKitm
of apomixis will be the isolation and auronfiocum. Proc. Queens/own Molecular Biology Meeting.
characterization of the mutations and their Queenstawn, New Zealand.
BickneR, R.A., and N.K.Boot. 1994. AgrobocteriUllHnediated
wild type alleles. The mutant genes could be Iranslarmalian al Hierodum ourantiocum. Int.1. Plant Xi. 155:
transferred to normal Arabidopsis via T-DNA, 467-70.
somer, lW. 1951. (y101agi<al and embryalagi<al studies in the amplJi.
to determine what effect they have in a
apanicti< Arobis hoIboeNii complex. Dons!e rHiemJcahemes Selskab
background where the wild type allele is also BioIogiske S/crifter 6: 1-58.
present. It would be interesting to determine Bauchez, D., CCanllleri, and M. Caboche. 1993. Abinary vector based an
Basta resistance lar inplonfa translormation afArabidopsis tho/iana.
if the fislfie mutations are transmissible by CR. Acoo. 5<i Poris 316: 1188--93.
female gametes and, if dominant, give rise to (arman, 1G. 1997. AsynchrallOUl expression of dlJllbte genes in
seeds with autonomous endosperm and angiosperms may cause apomixis, bilpary, tetraspory, and
polyembryony. Bioi. J. linn. S«. 61: 61-94.
embryos. Chaudhury, A.M., LMing, CMiller, S. Croig, lS. Dennis, en! WJ.
Peacock. 1997. Ferti&zotion-independent seed development in
Efforts to identify apomixis genes from Arabis, Arobidapsis fha/iooo. Proc. NaIl. Atoo. Xi. (USA) 94: 4223--78.
as well as other apomicts, should eventually (0101, V., LS. Roberl, TA. Kavanogh, M.W. BeYllll, and R.D. Thompson.
come to fruition. An intriguing area of inquiry, 1987. localizatioo 01 sequences in wheat endosperm pralein genes
whi<h conler tissue-specific expressioo in tobacco. EMBO 1. 6: 3559-
when that comes to pass, will be the nature of 64.
the relationship between these natural (urlis, CA, and G.G. Dayle. 1992. Productioo af aneuploid ond diploid
apomixis loci and any mutations conferring eggs by meiotic llIltants of maize. 1. Hered 83: 335--41.
Doodernan, M., A.G.M. Gerats, A.W.Schram, P. de V1aming. and EBionchi.
apomictic characteristics that have been 1984. Geneti< analysis 01 instab~ily in Petunia hybrida 2.Unstable
identified in the related sexual species. The mutatioos 01 dilferenlloci os the resu~ 01 transpO!ilion of the gene/i<
combined strategies of mutagenesis in sexual element inserted a1 the AnI locus. Theoref. AppI. Genet. 67: 357--66.
Eenink, A.H. 19740. Matromarphy in Brassica aleracea L I.Ternlnolagy,
plants to induce apomixis, and in apomictic parthenogenesis in Cruciferae and lhe formatioo and usobiily 01
plants to identify natural apomixis genes, motrarnorphi< plants. fuphytica 23: 429-33.
- - . 1974b. MutrOl1lOfphy in BmssicD
oIeraceo L 1I.lliffererKes in
JUfhenogenetK ability and
por1henogenesis indlKing ability. EupIrytn
23: 43)....15.
FeldmoM, U, R.LMolmberg, and t Dean.
1994. Mutagenesis in Arobidopsis. In E.M.
TheI.
Jeflenon, U, and R. BKkneI. 1996. The
potential in1IOCfS of aponixis: 0 rnoIectU
geneoo approach. In B.W.5. Sobrol (ed),
ofPlam Molecular Genetia.
Boston: Birkhiiuser.
Jongedijk, E., and M.5. RamoMO. 1988.
5ynoptK l1lJlanll in palata, Solanum
Nielsen, OJ. 1974. Moaomoleculor physiology
of pIostids. XII. TIgrino mulants in barley:
genetK, spectrQ\(opic and llructtKoI
choracterimtion. HfJ(editrls 76: 269-304.
Nogler, GA 1984. Gometophytic apomixis. In
B.M. JoIvi led.), Embryology af
Angiosperms. Berlin, Heidelberg, New
MeyerowitI and tR. Somervile (eek.), ll.Gerosum L I. Expression and identity of York: Springer-Yellog. Pp. 475-51 B.
Arobidopsis. New York: Cold Spring Ibboi genes for desynopsis. Genome 30: 664- Ohod, N., LMorgossion, Y. -t Iku, t WilIioIl1S,
Loborolory Press. pp. 137-72. 70. P. Repelli, and R.L filCher. 1996. A
FIoYeI, R.B., and G. Moore. 1996.l'tonl J~k, E., M.5. Ramonno, Z. Sawor. and mutolion that alows endosperm
genome constituents and their lG.l Hennse. J991. Formation of fim development without fertJlizolion. Prac.
organisolion..ln G.D. Foslel and D. TwelI DIVision reslitulion (FOR) 21H11eg05p01"es NotI. Acod. SO. (USA) 93: 5319-24.
(eek.), Plant Gene Isolation. <hichestel, hough pseudohonxJtypic cMion in rh-I Ramulu, l5., P. Dijkhuis, A. Pereira, G.t
New York, Brisbane, Toronto, Singapore: (desynopsis) l1lJlants of diploid potato: Angenenl, M.M. van Lookelen Compogne,
John WiIey and Sons. pp.I-25. rOUline production of tetraploid progeny and JJ.M. Dons. 1997. EMS and
Finch, RA, and M.D.Bennen. 1979. Action of from 21fDR x2.lf1)R crosses. 1he«. App/. transposon mutagenesis for the isoIotion of
triploid inckKer (1nl an meiosis in bor\ey Genet. B2: 645-56. aponictK l1lJlants in plants. In Somodonol
{Hordeum vu/gore U. Heredity 43: B7- Kin, B.t, M.5. Soh, BJ. Kang, M. FlWI/Ya, and Vario1ion and IndrKed Mutations in Crop
93. H.G. Nom. 1996. Two dominonl Impravemem. Dordrecb" the Netherlands:
Gelcm, A.G.M., H. Hum, E. Yn"jlondt, t MorIllO, phoIomorphogenic l1lJlotions of Klvwers Academic Pubw,eII.
E. Souer, and M. Belli. 1990. ~r Arobidopsis thdiano idenlified as Reddy, P.5..and R. d'Cruz 1969. Mechanism of
characterization of 0 non-outOllOOlOUl suppressor mutations of/ry2.1'font J. 9: aponixis in Dichonlhivm amulotum
transposable elemenl (tlTphl) of Petunia. 441-56. (Forskkl Stopf. Bot. Gaz. 130: 71-79.
I'/ontCeH2: 1121-2B. malgel', B., D. Bai, and Y. Sokolov. 1996. Rhoades, M.M., and E. Dempsey. 1966.
Gol15choIk, W., and H.D. KIein. 1976. The Assigmlenl 010 gene(s) canferring Induction of dvomosome doubling at
influen<e of l1lJlated genes on aponixis in Tripsocum 100 dnmalame meiosis by the elongate gene in moize.
sporogenesis. Asuney an the genetK arm: cytological and moIecuIor evidence. Geneoo 54: 502-22.
control of meiosis in l'iwm soIivvm. TIre«. Genome 39: 1133-41. Roy, BA J995. The breeding sysIeII1S of six
App/. Genet. 4B: 23-34. Kilani, Y. 1994. Induction of porthenogenetK species of Arobis (Brossicocael. AJnet. J.
Gostakson, A., and I. Godd. 1965. Mutotionl in haploid p1anll with brossinolide. Jpn. J. Bot. B2: 869-77.
crop i~O'iernenl. IV. I'oD ptlIIerrsis L Genet. 69: 35-39. Rutishouser, A. 1954. Die Enlwiddungserregung
(Graninoel. Hereditrls 53: 90-102. KoItunow, A.M., RA Bicknell, and A.M. des EndosperIl1S bei pseudogomen
Hogberg, A., and G. Hogberg 1980. High Choudluy. 1995. Apomixis: moIecuIcr lalllmllus-orlen. Mitt. NotIJIfoOOr. ges.
frequency of spontaneous haploids in the lIrategies for the generolian of genelicol~ Scbafflrousen 25: 1--45.
progeny of on inckKed mutolion in barley. identKol seeds without fer1ilizolion.1'font 5coll, RJ., M. $pielmon, J. Bailey, and H.G.
Hereditas 93: 341--43. 1'/ryfioI. lOB: 1345-52. Dickinson. 199B. Porent-of-origin effects
HoMO, W.W. 1995. Use of oponixis in Mtiwr KlllloIski, 5.P., LTlen-Hung, U Feldmom, and an seed development in Arabidopsis
development. ADvan. Aglon. 54: 333-50. A.H. PolellOl1. 1994. Comparative tboiono. 0e~125: 3329-41.
HoMO, W.W., M. Dujordin, P. Ozios-Akins, and L mopping of Arobidopsis thoJiallO and Stringom, G.R. 1970. Acytogenetic anolym of
Arthur. 1992. Transfel ofoponixis in Bnmica oIeraceo chromosomes reveals three synoptK mutanll in Brossica
Pennisetum..Prac. Apomixis Worbhop, islands of cOlllefYed organisation. Genetics collfJeSlris L Can. J. Genet. Cy1DI. J2:
AI\onta, Georgia, USDA. Pp. 30--33. 138: 499-510. 743-49.
HOMO W.W., ond 18.PoweIl. 1973. ShJIby LebicIK, 0., D. GrimoneIIi, D. Gonztilez-ile.l.eOn, Topping, u, F. Agyemon, B. Henricot, and l
head, and induced Iocu~otive opomicI in and Y. Sovidon. 1995. Detection of the Lindsey. 1994.1den~1icotion 01rnoIerulor
pearl millet. Crop SO. 13: 72lt-2B. aponictK mode of reproduction in maize- morkelS of embrogenesis in Arobidopsis
- - . 1974. Rodiotion-induced female- Tripsonm hybrids using maize RFLP tboiono by promoler trapping. Plant J. 5:
sterile l1lJlont in pearl malet. J. Hered. 65: morkel1. TIre«. App/. Genet. 90: 1198- 895-903.
247--49. 1203. Warmke, H.E. 1954. Apomixis in Ponicum
Honson, ll, and F.Y. JUlka. 1962. Induced Lindsey, l, W. Wei, M.t Oarke, HJ. McAnle, maUnum. Amer. J.Bot. 41: S.11.
mutalions in Kentucky Bluegross. Crap Sci. LM. RooIte, and J.F. Topping. 1993. Wilinson, la., M.B.Lonohon, D.G. Ootk, A.B.
2: 369-71. Tagging genonic sequences that direct Bleedcel, t Chong, E.M. MeyerowitI, and
Hemer~, A., J. de A1meido Engler, t transgene expression by octivoIion of0 HJ. KIee. 1997. Adominonl mutanl
Bergounioux, M. van Montogue, G. Engler, promoter trap in plants. Tromgenic receptor from Arobidopsis conlel1 ethylene
O.lnze, ond P. Ferreira. 1995. Doninord leseorm 2: 33-47. insensitiv~ inhelerologous p1anll. I/ahn
negative mulonll of the Cdc2 kinase Mogie, M. 19BB. Amodel for the evolution and Biolem. 15: 444--47.
uncouple cel DIVision from ~eloIive plant control of generative opomixis. Bioi. J. Ylstra, B., J. Busscher,l Franken, P.tH.
development. EMBOJ. 14: 3925-36. Um. Soc. 35: 127-53. HolImll1, J.N.M. Mol, and AJ. van TW1eI1.
NeI, P.M. 1974. Crossing O'iel and diploid egg 1994. Aovonols and lertibotion in Petunia
Iormolion in the elongate mutant of moize. /rybnda: Ioaiizolion and mode of odion
Genetics 79: 435-50. during poIen tube grDWIh. Plant J.6:
201-12.
Chapter 14
Genetic Engineering of Apomixis in
Sexual Crops: A Critical Assessment of
the Apomixis Technology
THOMAS DRESSELHAUS, JOHN G. CARMAN, AND YVES SAVIDAN
Table 14.1 Examples ofisolated genes and their promoters that might beuseful as tools for de IHI~O
synthesis oftheapomixis trait in sexual aops
PrCKess to be . . . .tecl
GetJe (expressiOll/factioa) (Origin) Reference
'lpoIBixis genes'
nol isolaled yell?!
Ovule and ...cehs-spedfic target getIt expressioa
FBPl promoter (ovule-specific) (Petunia) Colomba elaI., 1997
DEFH9 promoler (ovule-specific) (Anthirrhinum) Rotino elal., 1997
WM403 promoler (nlKellus-specific) (waler-melon) Shen el01., unpublished
Nucel/in cDNA (nlKellus-spe<ific) (barley) Chen and FooIad, 1997
Preventioa of meiosis/apo.eiosis
diverse cDNAs (ear~ meiosis-specific) lIi~) Kobayoshi elal., 1994
pAWJl3 cDNA (ear~ meiosis-specific) (wheat) Ji and langridge, 199~
DMCI gene (MMC*-specific) (Arabidopsis) Klimyuk and Jones, 1997
SrNI gene (chrom. condensation/pairing) (Arabidopsis) Bai el 01_, 1999
Partheaogenesis (.tOllOllllNfS embryo developlllent)
SEIlK gene (compelence 10 form embryos) (carrol, Arabidopsis) Smmidl elal., 1997
l£(l gene (compelence 10 form embryos) (Arabidopsis) Lolan el 01., 199B
BBM 1gene (compelence 10 form embryos) (Brassica, Arabidopsis) Boutilier elaI., unpub~shed
ZmES/-4 promoler (embryo soc-specific) (maize) Amien and Dresselhaus, unpublished
(AutonomlNfs) endospenn development
MWFIS 1gene (suppressor) (Arabidopsis) Grossniklaus el01., 1998b
Loo el01., )999
FIS2gene (suppressor) (Arabidopsis) loo elal., 1999
FI£!FIS3 gene (suppressor) (Arabidopsis) Ohad elaI., 1999
ZmES 1-4 promoler (embryo sac-specific) (maize) Amien and Dresselhaus, unpublished
Imprintilg
METI a/s (hypomethylation) (Arabidopsis) Adorns elaI., 2000
Vinkenoog el aI., 2000
Inducille/repressable systems
Sleroid-inducible promoter (mammak) Smena el01., J991
Copper-indlKible promoler (yeasl) Mell elal., 1993
TelTacydine-inducible/-inactivatable promoler (bacterium) Weinmann el01., 1994
Ethonol-inducibele promoler (fungus) Caddick el 01., J998
"MMC: Mega- and Micr~pore mother celll.
234 no.- Orn......, Jo" G.C_ _ lIIIII l,n s...w.
Table 14.2 Examples of patents linked with theengineering oftheapomixis traitin sexual crops.
Sources: Intellectual Property Network (http://www.delphion.coml. European Potent Office (http://ep.dips.org/dips), and Bicknell
and Bicknell (1999).
Apomixis tedlnology
Patent number'
(Publication date) TItle (and content) Applicant(s)
Breeding strategies
W089l10810 Asexual induction of heritable male sterility and apomixis in plants Maxell Hybrids INC
(Feb. 9, 1989) (use of male sterility factors).
CN1124564 Hybrid vigor fixing breeding process for rice apomixis Chen J.
(June 19, 1996) (breeding and selection strategy).
US5710367 Apomictic maize (introgression of apomixis USDA
(Jan. 20, 1998) from Tripsacum to maize).
W0971 0704 Apomixis for producing true-breeding plant progenies (introgression USDA
(Sep. 22, 1998) of apomixis from Pennisetum squomulatum 10 cullivars).
W09833374 Methods for producing apomicitic planls University of Utah Stole
(Aug. 6, 1998) (breeding program).
WOl107434 Novel genetic material for transmission into Eubonks M.W.
(Feb. 17,2000) maize (introgression of apomixis from Tripsacum).
Stimulation ofapomictic reproduction
EP0127313 The production ofhaploid seed, of doubled haploids ond of Rohm & Hoos
(Dee 5, 1984) hamazygous plontlines therefrom (causing opomixis by applying
on apomixic agent).
SU1323048 Stimulolor oflloral opomixis Pollov Selskokhaz IG
(Ju~ 15, 1987) (no file available). Nikinkij
US4818693 Methods and materials for enhanced somatic PGS
(April 4, 1989) embryo regeneration in the presence of auxin.
US5840567 Simplified hybrid seed production by latent diploid porthenogenesis and University of California
(Nov. 24, 1998) porthenote cleavage (induced by controlled environmental condnions).
De novo synthesis ofapomixis (genes and promoters)
W09743427 Production of apomictic seed (using 0 SERK gene for Novartis and inventors
(Nov. 11, 1997) embryogenic potential).
W09808961 Endosperm and nucellus specific genes, promolers and Doan, D.N.P., Olsen,
(March 5, 1998) uses thereaf. O.-A. and Linnestad, C.
W09828431 Transcriptional regulation in plonls John Innes Centre
(Ju~ 2, 1998) (using 0 meiosis specific promater). lnnov. UD and inventors
US5792929 Plants with modified Rowers (modifying Rower celk after PGS
(Aug. 11, 1998) tronsformation with foreign DNA).
W09836090 Means for identifying nucleotide sequences IRD and C1MMYT-ABC
(Aug. 20, 1998) involved in opomixis (isolation ond modification of sexual genes
for the expression of apomixis in Gramineae).
W09837184 Leafy cotyledon I genes and their use (using embryo specific genes University of California
(Aug. 27, 1998) ond their promoters).
US5907082 Ovule-spe<ific gene expression University of California
(May 25, 1999) (using ovule-spe<ific genes).
W09935258 Nucleic ocid markers for opospory·specific University of Geargia
(Ju~ 15, 1999) genomic region (from the genus Paspolum). Research Found. IHC
W09953083 Seed specific polycomb group gene and Cold Spring Harbor lob.
IOct. 21, 1999) methods of use for some (using repressOfS of embryo ond
endosperm development).
W0024914 Apomixis conferred by expression of SERK Hovortis
(May 4, 20(0) interacting proteins (see above W097434271.
•wo, US, Ep, CN and SU refer to World patents, US-, European, Chinese and larmer Sawjel Union patents.
GoIOti< blgioHriog .f A,o.ui. io SUI"I er.ps: ACritical As... _1 ., "'" A....... is TedloolotY 235
gametophyte (Crossniklaus, Chap. 12;Cordts 1995). Very little molecular data concerning
and Dresse lhaus, unpublished results). parthenogenesis are available for higher
Through the use of mutant approaches plants. One protein (a-tubulin) was identified
(Vollbrecht and Hake 1995; Drews et al. 1998; whose expression is associated with the
Yang and Sundaresan 2000; Crossniklaus, initiation of parthenogenesis in wheat (Matzk
Chap. 12;Praekelt and Scott, Chap. 13), we can et al. 1997). And auxin (2,4 D) treated sexual
anticipate that many more genes involved in eggs from maize can be triggered to initiate
female gametophyte development will soon be embryo development.,.at a low frequency
isolated. Gene trap screens such as T-DNA (Kranz et al. 1995), however, the molecular
insertional mutagenesis, transposon mechanism is not understood. Three genes
mu tagenesis, and enhancer detection were used to successfully initiate the formation
(Crossniklaus, Chap. 12) are very powerful of embryo-like structures on vegetati ve tissue
molecular tools for isolating the corresponding (leel: leafy cotvledonl, Lotan et al. 1998; and
genes and/or their promoters from sexual bbml: babyboom 1r Boutilier et al., unpublished
model plants like maize and Arabidopsis. results) or to enhance the rate of somatic
Further tissue/cell-specific genes and their embryos in culture (SERK1: somatic
promoters will be isolated by transcript embryogenesis receptor-like kinase 1, Hecht et al.,
profiling methods (e.g.,Liang and Pardee 1992; unpublished results), respectively. It remains
Welford et al. 1998;Matsumura et al. 1999)and to be demonstrated whether these genes are
from tissue/cell-specific cDNA libraries (e.g., also useful for inducing embryo development
Dresselhaus et al. 1994;Dia tchenko et al. 1996). in reproductive cells.
Initial attempts have been made to compare
Parthenogenesis may also arise as a function
gene expression profiles between sexual and
of timing, taking into account that
apomictic lines within the same species. A few
parthenogenetic embryogenesis is usually
genes that are specifically expressed in the
initiated before anthesis. In contrast to sexual
ovules of either sexual or apomictic lines were
eggs, parthenogenetic eggs (e.g., Penniseium
isola ted (Vielle-Calzada et al. 1996b). These
ciliare and wheat) contain ample amounts of
genes may eventually be useful tools for
ribosomes and polysomes and a large number
inducing apomictic development in sexual
of cristae in mitochondria, thus suggesting a
lines or sexual development in apomictic lines.
highly active metabolic status prior to
Parthenogenetic embryogenesis from pollination (Naumova and Vielle-Calzada,
unreduced eggs is the next required step for Chap. 4; Naumova and Matzk 1998). In
successfully engineering the apomixis trait. contrast to sexual eggs, degeneration of
Whether this will occur spontaneously once synergids in aposporous Pennisetum ciliate
the egg is diploid has yet to be shown. Quarin female gametophyte was precocious and
and Hanna (1980) found that doubling a sexual rapid. In addition, a complete cell wall around
diploid Paspalum line generated a tetraploid the eggs was already generated before the
that was facultative aposporous, thus arrival of the pollen tube (Vielie et al. 1995). In
unreduced egg cells developed partheno- maize, zygotic gene activation (ZGA), the
genetically into embryos. Spontaneous swi tch from maternal to embryonic control of
parthenogenetic development was observed at development, occurs soon after fertilization
a low frequency in maize (Chase 1969; Bantin (Sauter et al. 1998; Dresselhaus et al. 1999;
and Dresselhaus, unpublished results). Wheat Bantin and Dresselhaus. unpublished).
lines have been described that produced up to Precocious expression of zygotic genes before
90% parthenogenetic haploids (Matzk et al. pollination/ fertilization could thus eventually
be used as a tool to induce parthenogenetic successful transformation of maize and wheat
development of sexual eggs, and perhaps has also been reported (lshida et al. 1996;
those same genes might be useful for inducing Cheng et al. 1997). Even so, particle
endosperm development. Although the bombardment of wheat and maize immature
existence of repressor molecules that prevent scutellum tissue remains the most widely used
unfertilized eggs from initiating embryo method in most public laboratories. Relatively
development has not been proven, it is efficient transformation systems are now
reasonable to postulate their reality. Once available for all major crops as well as some
isolated, they might be a useful tool for forage grasses (Spangenberg et al. 1998).
engineering parthenogenetic embryo Development of transformation systems for
development as a component of apomixis. apomictic species is in progress, and
transformation protocols for pearl millet will
Induction of endosperm development will
be established once interesting apomixis genes
probably be the biggest obstacle to the utilizing
become available (P. Ozias-Akins, personal
apomixis in sexual crop species (discussed
comm.). Transformation of Brachiaria and
further under "Main Limitations").
Tripsacum are foci of apomixis programs at the
Nevertheless, an in vitro system for
International Center for Tropical Agriculture
endosperm development in maize was
(ClAT)and the International Maize and Wheat
reported recently (Kranz et al. 1998),providing
Improvement Center (CIMMYf), respectively.
impetus to molecular investigations about
gene expression and regulation during the A major problem related to transgene activity
earliest steps of endosperm development. is the instability of expression (Iorgensen 1995;
Matzke and Matzke 1995).Often inactivation
Transformation and
Inducible Promoter Systems of transgene expression is accompanied by an
Tremendous progress has been made in plant increase in DNA methylation (Meyer 1995).In
genetic engineering since the first reports of addition, transgenes may be integrated in
successful plant transformation appeared in hypermethylated chromosomal regions
the early 1980s, and many commercially displaying a spatial and temporal change of
relevant genes have been transferred to crop methylation during plant growth and
plants (Christou 1996). Agrobacterillm- development (position effect).Transgenes with
mediated transformation has been the method homologous sequences to endogenous genes
of choice for introducing exogenous DNA into may be silenced through the cosuppression
dicotyledonous plants. Agrobacterium effect (Iorgensen 1995; Matzke and Matzke
transformation has proven difficult with 1995). All the same, plants stably expressing
cereals, and consequently, alternative methods the transgenes can be selected over
such as particle bombardment have been generations, although this is time-consuming
employed. Nevertheless, because Agrobac- and expensive. Suggestions have been made
terillm-mediated gene delivery offers many as to how vectors used for genetic
advantages (easy protocols, often low- or even transformation can be optimized in order to
single-copy integrations, mostly full-length minimize the cosuppression effect (Meyer
integration of transgenes, short or no tissue 1995). Single-copy integration of transgenes
culture period), considerable effort has been will be enabled by the deployment of
dedicated to establishing this method for Agrobacterillm-mediated gene delivery. This in
cereals (Komari et al. 1998). Agrobacterillm turn will increase the rate of plants that stably
transformation of rice is now routine, while express the transgenes. Gene targeting by
homologous recombination, i.e., the
Geootk bfioe<riog .f ApooUis io Sex.. C1opI: ACriticoI AJ..._ .f IIle ApMoWs WoooIort 237
generation of null mutants, is probably the different organs, especially in embedded cells
ideal way to stably silence genes. The like megaspore mother cells and the cells of
deployment of this approach, however, is still the embryo sac, which are the main target cells
relatively limited for higher plants (Puchta for the genetic engineering of different
1998).An alternative is homology-dependent apomixis components. Seed producers
gene silencing (HOCS; for review, see Kooter anticipate efficiency rates as high as 99% for
et al. 1999), especially through the use of such systems (http://www.apomixis.de; see
double-stranded RNA (RNAi: RNA panel discussion during the Third European
interference technology) as a template for gene Apomixis Workshop). Existing systems,
silencing (Bass 2(00). Gene silencing at rates therefore, must be optimized, or preferably,
up to 100% was reported with transgenic new systems using natural, easily
plants using the latter approach. biodegradable, and harmless chemicals as
inducers must be developed to satisfy seed
Inducible/ repressible systems are necessary to
producer demands and environmental
engineer the apomixis trait, because genetic
necessities.
recombination through sexual crossing will
always be required for the introduction of new Main limitations
traits into crops. In a panel discussion with Perhaps the biggest obstacle to genetically
industrial representatives during the Third engineering apomictic grain crops is that
European Apomixis Workshop (April 21-24, fertilization of the central cell is likely to be
1999,Gargnano, Italy), it became very clear that required because of dosage effects (Birchler
inducible systems for engineering the 1993; Savidan, Chap. 11) and because
apomictic trait are highly desired (http:/ / autonomous endosperm development occurs
www.apomixis.de;seeworkshops).mainly at low frequencies in cereals. A balanced
because they serve as a natural means of maternal:paternal genome ratio (2m:lp) is an
protecting intellectual property rights (see absolute requirement for endosperm
"Intellectual Property Rights," this chapter). development in cereals (Birchler 1993).In most
The question is whether such systems are cases, deviation from this ratio leads to embryo
practically possible, given the problems abortion or seeds with diminished fertility
encountered with the application of (Birchler 1993; Praekelt and Scott, Chap. 13).
gametocides. Various chemical inducible In contrast to cereals, Scott et al. (1998) have
systems have been reported, e.g., the shown that in Arabidopsis, 2m:2p, 4m:lp and
tetracycline inducible/ inactivatable promoter 4m:2p ratios are allowed. Also observed in
system, and steroid-, copper- and ethanol most pseudogamous apomicts are ratios of
inducible promoter systems (for review, see 4m:1p and 4m:2p. In apomictic lines of the
Gatz and Lenk 1998). Whether these systems maize relative Tripsacum, Grimanelli et al.
are applicable and acceptable for use under (1997) identified 2m:2p, 4m:1p, and 8m:1p
field conditions is doubtful; spraying ratios. Imprinting of gametic nuclei is the
antibiotics, steroids, and heavy metals is genetic reason behind this phenomenon: one
environmentally unacceptable. Ethanol set of alleles is silenced on the chromosomes
systems might offer an alternative. Most of contributed by the mother, while another set
these systems, however, are leaky and have is silenced on the paternal chromosomes. Each
some background activity, or they may be too genome thus contributes a different set of
sensitive. In addition, there is the question of active alleles (Vinkenoog et al. 2000; AlIeman
how homogeneously the induction works in and Doctor 2(00). A few imprinted loci have
238 n.-Dr........ ,JoDG.e-- .... T... s....
apomixis for crop improvement. These improvement of human health (Helier and
apomixis patents raised concerns about the use Eisenberg 1998). In regards to apomixis, it is
of apomixis technology. The Rural unlikely that the situation will change in the
Advancement Foundation International near future because it is still possible to file very
(RAFI), a nongovernmental organization, broad apomixis patents.
recently expressed the concern that apomixis
The question of whether farmers in developing
IPR could wind up in the hands of only a few
countries will get access to disclosed apomixis
dominate global agrobusiness players, and that
technology remains unanswered. One can
farmers in both developed and developing
hope that many of the relevant patents will be
countries might become totally dependent on
secured by public organizations such as the
their seed products. Other concerns are that
Consultative Group on International
genetic diversity could significantly decline
Agricultural Research (CGIAR) and other
and that developing countries will not have
public insti tutions (see Hoisington et aI. 1999),
access to this technology because they will be
thus giving interested parties in developing
unable to afford the required rights and
countries the possibility of acquiring free
licenses (RAFI 1998). The latter concern is
access to this powerful technology. Certainly,
shared by leading apomixis researchers and
the public image of the big agrobusiness
was formalized in 1998 in the Bellagio
players would benefit from freely licensing the
Apomixis Declaration (for full text, see http:! /
technology to CGIAR institutions or directly
billie.harvard.edu/apomixis). Signatories to
helping farmers in developing countries use
the declaration were interested in how to
this technology. The bulk of profits, after all,
develop novel approaches for generating the
will be earned in the more developed
enabling technology, and how to patent and
countries. Introducing the apomixis trait into
license it.Currently, patents related to apomixis
local varieties would give farmers in
enabling technology are dispersed among
developing countries access to powerful and
many parties (Table 14.2). Furthermore, it is
productive hybrid technology (Hoisington et
expected that the number of patents will
aI. 1999).Tosome extent, these farmers should
greatly swell as numerous public and private
have the right to save seed for subsequent
research institutions continue investigating
replanting, thus allowing them to significantly
different aspects of apomictic and amphimictic
increase their crop yield and personal income.
reproduction pathways using different species
and approaches (see e.g., Bicknell and Bicknell Risk Assessment Studies
1999).
Risk assessment research and studies relate to
Another negative impact stemming from the use and or release of genetically modified
apomixis patents is that communication of organisms (GMOs) into the environment. Since
research results to the scientific community is the first release of genetically modified plants
either delayed until patents have been filed or (GMPs) some twelve years ago, many short-
they are simply not communicated at all. A term studies have been conducted (de Vries
Widespread phenomenon in today's 1998). Short- and long-term risk assessment
biomedical research is that while IPR is studies are also needed to evaluate the
growing rapidly, scarce resources are poorly environmental implications of novel apomictic
utilized because too many patent owners are crops. One key issue for investigation is
blocking one another. Paradoxically. more IPR whether the apomixis trait can move to the
may lead to fewer useful products for the land races and wild ancestors of food crop
240 no- O..S...... JoIio G.c.-." T.......
plants. and if so. what would be the impact. and farmers in both developing and developed
This issue is especially important in the centers countries (see also IPR) will be required, and
of origin for the crop plants. Furthermore. the the research results should be communicated
issue of how apomixis might affect genetic to all potential users.
diversity. and whether it would increase or
decrease monoculture farming needs to be Summary
explored. Based on field studies on herbicide The extensive introduction of apomixis into
and/or insecticide resistant plants, we can sexual crops will undoubtedly rely on genetic
probably expect engineered apomixis genes to engineering. as we anticipate that more
move through vertical gene transfer (transfer candidate genes (especially regulatory genes
of a gene from plant to plant via sexual and tissue/cell-specific promoters) and
reproduction/pollen) (Lutrnan 1999).The rate enabling techniques will be identified and
of horizontal gene transfer (asexual gene flow developed in the near future. Transformation
between organisms) is relatively low and the technology for all major crops is now available
risk negligible, however, microbiological risk and inducible systems are currently being
assessment studies in this area could be useful developed and optimized, allowing the control
(Syvanen 1994). Given our current knowledge, of transgene expression and activity even
it appears unlikely that microorganisms could under field conditions. Adventious apomixis
gain some advantage over wild relatives after using already described or novel genes under
uptake of apomixis genes. the control of ovule-, nucellus- or archespore-
specific promoters is probably the easiest way
If apomixis is controlled by multiple genes. the
to engineer the apomixis trait. Plant breeders
probability of diffusing this trait to wild
and seed producers would like to generate
relatives is extremely low. The transfer of
inducible obligate mitotic diplospory in
several genes to a wild plant should lower its
combination with autonomous endosperm
fitness to a level unacceptable for survival in
development. The latter is probably the most
the wild (Berthaud, Chap. 2). If apomixis is
difficult aspect of engineering apomixis.
controlled by a single gene, which would result
especially for cereals such as wheat, rice. and
in obligate apomictic wild races, these races
maize, because of dosage and imprinting
would lose their potential to evolve. If
effects.
dominant, an apomixis gene could rapidly
become fixed in an outcrossing sexual Although apomixis is a hot topic in plant
population. Therefore, in theory, apomixis research, our current understanding of both
transgenes could possess advantages that apomictic and amphimictic reproduction
might result in the uncontrollable spread of pathways in higher plants is still extremely
the transgenes (van Dijk and van Damme limited. The economic potential of apomixis
2000). Inducible apomictic systems and male might provide the impetus to bring apomictic
sterility might circumvent these problems. crops to the marketplace, and in the process it
Nevertheless, the described possibilities may well contribute significantly to our future
indicate that risk assessment studies and understanding of the molecular regulation of
research to investigate the ecological the many different sexual and apomictic plant
implications of novel apomictic crops (once reproduction pathways.
available) to the environment are an absolute International and interdisciplinary approaches
necessity. In addition, socioeconomic studies and efforts are now needed to study and
on the positive and negative implications of manipulate seed reproduction. It will be
this technology for breeders, seed companies,
necessary (i) to characterize the genetic reproduction through seeds in apomictic
regulation of apomixis and isolate the systems and sexual crops," coordinated by T.
responsible genes, (ii) to analyze the genetic Dresse lhaus). In 1999, a transatlantic
and molecular bases of sexual reproduction consortium was initiated between two public
and to isolate the corresponding genes, and institutions (CIMMYT and IRD) and three
(iii) to produce the tissue/cell-specific and private companies (Pioneer Hi-Bred, Novartis,
inducible/repressible promoters that will be and Group Limagrain). This is just a beginning
needed to control the expression of the target and more concerted projects are needed in
genes. Concerted international research efforts order to reach the ambitious aim of
have been made in Europe aimed at manipulating the apomixis trait in crops.
understanding apomictic and sexual
Apomixis technology will offer many exciting
reproduction pathways in order to develop
opportunities for the agriculture of the 21't
tools for the manipulation of the apomictic
century, and indeed many patents already
trait (e.g.. an E.U. Research Technology and
have been filed with many more yet to come.
Development (RTD) project entitled "The
It is critically important that these patents be
manipulation of aporruxis for the
held and used for the good of all. Public
improvement of tropical forages," coordinated
institutions in particular must safeguard the
by M. D. Hayward; a RTD project entitled
access of developing countries to these
"A pornixis in agricul ture: a molecular
enabling technologies. In all likelihood,
approach," coordinated by M. van Lookeren
constraints to the broad and generous use of
Campagne; and a Concerted Action Project
apomixis technology will be political and
entitled "Introducing and controlling asexual
economic rather than technical in the future.
References Blakey, CA, Ll, De'MIld, and S.L Goldman. Calarnbo, L, 1 Franken, A.l Van der Krol, P.E.
1997. Ca-segregaticln 01 DNA markers with W"lIIich, HJ. Dons, lIld G.tAngenent.
s.
Adams, R. V"l/1kenaog, M. Spielman, H.G. Tripsacumledty. Moymcu42: 363-69. 1997. Dawrvegulatian 01 ovule-specific
Dickinsan, and RJ. Scatl. 2000. PlI"enl-ol· Caddidt, M.l, U Greenland, I.Jepson, K.P. MADS box genes ham petunia resuhs in
Qligin effem on seed deYeIolX'*lt in Krause, N. Qu, K.V. RiddeD, M.G. Soher, W. maternally cantraled delem inseed
Arabidopsis tho/iana requre DNA Schud1, U. Sonne'MIld, and A.B. Tamsen. development. fbrtCel9: 703-15.
methylalian. Development 127: 2493- 1995. An ethanol indu<ible gene switch lor Diatchenka, L, Y.-F. Chris Lou, A.P. Ca~beI, A.
2502. p1an15 used lamanipulate carbon Chenchik, F. Maqadom, B. Huang, S.
Alfeman, 1.1., and 1 DooQl. 2000. Genanic metabalism. NaI. Biatechnal. 16: 177-aO. Lukyanov, K. Lukyunow, N. Gurskayo, E.D.
imprinting inplan15: observations and Chase, S. 1969 Manaplaids and monoplaid· Sverdlac, and P.D. Siebert. 1996.
eva!utianory i~ns. fbrtMol. Bioi. derivatives 01 maize (Zeo mars l.], Suppression subrraetiw hybridization: A
43: 147-61. Botanical Review 35: 117-67. method lor generatilg differentially
Asker, H., and LJerling. 1992. Apotrixis in Chen, E, and M.R. Faalad. J997. Malecular regulated or lissue-spedfic cDNA probes
PlanlI. Saca Raton, Florida: (RC Press. organization 01 a gene inbarley which and libraries. Prac. NaIl. Acad. Xi. (USA)
Sai, X., S.N. Peirson, EIlang, cXue, and CA encodes aprotein sinilar to aspartic 93: 6025-30.
Mckaraff. 1999. Isolation and pralease and its specific expression in Dresselhaus, 1,S. Cordl5, S. Heuer, M. Soufer, H.
characterization of SYNI, aRAD21·like nucellar cells during degeneration. Plant LOrz, and E. Kranz. 1999. Navel ribosomal
gene essentiollor meiosis in Arabidopsis. Mol. Bioi. 35: B21-31. genes hom maize are differentially
PlantCelll: 417-30. Cheng, 1.1., lE.Fry, S. Pang, H. Zhou, tM. expressed inthe zygatK and lOIlllIlic eeH
Sass, S.L 2000. Double-llranded RNA as a Hiranoka, D.R. Duncan, 1W. Conner, and Y. cycles. Mol. Gen. Genet. 261: 416-27.
template lar gene silencing. C!l11 DJ: Won. 1997. Genetic tronslQlmaticln of Dresselhaus, 1, H. LOll, and E. Kranz. 1994.
235-3S. wheat medioted by Agroboderium Representative cDNA ~braries ham few
Sidmell, RA, and K.S. Sicknell. 1999. Who wm rumefaciem. Plant Physiol. 115: 971-a0. plant cells. fbrt1. 5: 605-10.
benefit ham apomixis? Biotechnology and Chrislau, P. 1996. Translormaticln technology. Drews, G.N., D. Lee, and CA Christensen. 1995.
Development Mani/ex 37: 17-21. Trends PIont Sci. 1:423-31. Genetic anolysis allemale gamelaphyte
Sirchler, JA1993. Dosage analysis of maize de Vries, G.E. 1995. Past, Present And Future development and function. Plant CeH 10:
endosperm develolX'*lt. Moo. Rev. Considerations In Risk Assessment WIIen 5-17.
Genet. 27: 'SI-204. Using GMOs. Bihhaven, The Nerherlonds: Evans, L1 1995. Feeding the Ten Billion: PIonII
Commission Genetic Modification. and Population Grawth. New York:
Cambridge University Press.
242 1\0-. Dr......... Iou G., _ _ .... , ... s.r.w.
6011. t, and I. lenk. 1998. Promoters lhot Kaboyashi. 1. E. Kabayashi, S. Sato, Y. Hona, N. Motzk. F., H.·M. Meyer. H. Biiumlein, H.·J.
respond to ch~mi<al inducers Trends Plant Miyapma. A. Tanaka. and S. Tabota. 1994. Balzer. and I.Schubert. 1995. Anovel
xi. 3:352-58. Characterizotion of cDNAs induced in approach to the analysis ofthe initiation of
Grimanelli, D.• M. Hernandez. E. Perani. and Y. meioli< prophose in Ii~ micraspores. DNA embryo development in Gramineae. Sex.
Savidan. 1997. Dosage eHem in the Res. I: 15-26. Plant Reprod. 8: 261>--72.
andosperm of diplasparaus apomidic Kohunow, A.M.• R.A. 8i<knell, and AM. Motzk. F., H.·M. Meyer, t Horslmann, H.·J.
Tripsacum (Pooceae). Sex. Plant Reprod. CIloudhury. 1995. Apomixis: Molecular 8olzer. H. Biiumlein, and I. Schubert. 1997.
10: 279-92. strategies far the generation of genelical~ Aspecific olpha-tubulin is associated with
Grassniklaus. U., A. Kahunow. and M. van idanlical seeds without fertilization. Plant the initiation of parthenogenesis in
laokeren (ampogne. 19980. Abright Phyliol. 108: 1345-52. 'Salmao' wheat lines. Hereditas 126: 219-
future for apomixis. Trends Plant xi. 3: Kamori, 1. Y. Hieri. Y. Ishida. 1. Kumoshira, and 24.
415-16. 1. Kubo. 1998. Advances in cereal gene McMeniman, S.• and G. lubulwo. 1997. Project
Grassniklaus, U.• J..P. Vielle·(alzado. MA transfer. (/HT. Opin. Plant Bioi. 1: 161--il5. Development Assessment: An Economic
Hooppner, and W.8. 6ogliona. 1998b. Kooler. J.M.• MA Motzke, and P. Meyer. 1999. Evaluation ofthe Potential Benefits of
Maternal control of embryogenesis by tistening lathe silent genes: transgene Integrating Apomixis in Hybrid Rice.
MfDEA, a Palycamb group gene in silencing, gane regulation and pathogen Canberra: Australian (entre for
Arabidoplis. xience 280: 441>--50. control. Trends Plant 56.4: 34ll-47. International Agricultural Research.
Helier, MA. and RoS. Eisenberg. 1998. (an Kranz. E.• P. von WleQen, and H.LOrz. 1995. Men, V.L. LP. lachhead. and P.H. Reynolds.
patenn deler innovalion? The anti<ammons Ear~ cytological evenn aher induction of 1993. (opper-<ontrallable gene expression
in biamedi<al research. xience 280: 698- cell division in egg cel~ and zygote system for whole plann. Proc. Natl. kad.
70l. development following in vitro fertilization xi. (USA) 90: 4567-71.
Hoisington, D., M. Khoirallah, 1 Reeves, J.·M. with angiosperm gametes. Plant 1. 8: 9-23. Meyer. P. 1995. Understanding and controlling
Ribout. 8.Skavmond. S. Tabo. and M. Kranz. E., P. van WleQan, H. Quader. and H. transgene expression. TIBTECH 13: 332-
Warburtan. 1999. PIont geneti< resources: LOrz. 1998. Endasperm development aher 37.
Whot can they contribute toward increased fusion of isolated. single maize sperm and Noumava, IN.• and F. Matzk 1998. DiHerances
crop productivity. Prac. Nat/. Acad. Sci. cantral cel~ in vitro. Plant (ell 10:511-24. in the initiotion of the zygotic and
(USA) 96: 5937-43. tiang, P.. and A.8. Pardee. 1992. DiHerentiol parthenogenetic pathway in the Salmao
Ishida. Y..H. Saita, S. Ohta. Y. Hiei. 1 Kamori, display of eukaryotic messenger RNA by ~nes of wheat: uhrastructural studies. Sex
and 1. Kumashira. 1996. High efficient means of the pa~merase choin reaction. Plant Reprod. 11: 121-30.
transformation of maize (lea tOOys Ll xience 257: 967-71. NeuHer, M.G.• E.H. (00, and S.R. Wessler. 1997.
medioted by Agrobacterium tumefa6ens. latan, 1.M. Ohio, K.M. Yee. MA Wesl, R. la. Mutants ofMaize. (old Spring Harbor. New
Nature Biotechnology 14: 745-50. R.W. bong, K. Yamagishi. R.L flSCher, R.B. York: (old Spring Harbor laboratory Press.
Jemes, t 1997. GlobolStalus ofTtransgenic GoIdberg, and JJ. Harado. 1998. Ohod, N-r R. Yadegari, LMargassian, M.
Clops in 1997.ISAAA Briefs Na.5. hhoca. Arabidoplis L£AfY COTYUDONI is sufficient Hannan, D. Michoeli, JJ. Harada, R.B.
New York: 1SAAA. to induce embryo development in GoIdbeJg, and R.L Fischer. 1999. Mutations
Ji. L·H., and P.langridge. 1994 An ear~ meiosis vegetative cells. (ell 93: 1195-1205. inRE, a WO Palycarnb group gene. allow
cDNA clone from wheat. Mol. Gen. Ganel. lua, M., P. 8ilodeau, A. Kohunow. E.~. Dennis. andosperm development without
243: 17-23. WJ. Peacock. and A.M. Chaudhury 1999. fertitlzation. Plant (ellll: 407-15.
Jandle, RJ. 1999. PIont breeders' righn and Ganes controlling fertilizalion-independanl Pirrana, V. 1998. Palycarnbing the genome: PeG,
plant biotechnology. In A. Ahman. M. f1v. seed development in Arabidoplis thaliona. IrxG. ond dvomotin silencing. (elI93:
and S.lzhor (eds.). Plant Biotechnology and Proc. NotI. kad. Sci. (USA) 96: 291>--301. 333-36.
In Vitro Biology ofthe 21st (entury. lutman. PJ.W. 1999. Gene flow and agricuhure: Puchta, H. 1998. Towards targeted
Dordrecht, The Netherlands: K1UWllf RelevallCe for transgenic crops. British Clop transformation in plann. Trends Plont Sci.
Acodemic Pubtrshers. pp. 747-49 Protection (ounci/Symposium Proceedings 3: 77-78.
Jorgansen. RA 1995. (o-suppression, Rower No. 72. Quarin, U and W.W. Hanna. 1980. Effect of
color patterns, and metastable gene lulZ, W.• W. Sanderson, and S. Scherbov. 1997. three ploidy levels on meiosis and mode of
expression stales. Science 268: 681>--91. Doubling of world papulotion unlike~. reproduction in Pospolum hexastochyum.
Kinoshita. 1. R. Yadegari. JJ. Horoda. R.B. Nature 387: 8O~5. Clop Sci. 20: 69-75.
GoIdberg. and R.l. fl\Cher. 1999 Matsumura. H.• S. Nirasawo. and R. Terauchi. Rabobank International. 1994. The Work/Seed
Imprinting of the MEDEA Pa~carnb gene in 1999. Transcript profiling in rice (Oryza Marker: Developments and Strategy.
the Arabidoplis andosperm. Plont Cilllll : saliva U seedlings using serial an~ of Utrecht, The Netherlands: Rabobonk.
1945-52. gene expression (SAGE). Plont 1. 20: 719- RAR (Rural AdvallCement Foundation
Klimyuk, n. and J.D. lenes. 1997. AtDMCI. the 26. International). 1998. Terninator Trends:
Arabidoplis homalogue of the yeast DMCI Motzke. MA, and UM.MolZke 1995. How The '5iIent Spring' of'Farmefl Rights': Seed
gene: characterization, transposon·induced and why do ~ann inactivate homalogOllS Saving, The Public Sector and Terminator
allelic variation and meiasis-associoted (Iranslgenes? Plont Physiol. 107: 679-85. Transnolionalt Occasional Paper VoI.5.
expression. PlantJ. 11: 1-14. Matzk. F., A. Meister, and I.Schubert. 2000. An No.l. W"lIlnipeg. (anada: RAFI Publications.
effioanl serean far reproductive patl1wlJy5 Rolina. G.L, E. Perri, M. Zanini. H. Sommer. and
using mature seeds of manocan and dican. A. Speoo. 1997. Genetic angineering of
PlontJ. 21: 97-108. parthenocarpic ~ann. Nat. Biotechna/. 15:
1398-1401.
Geoeti< blgi"";og .f Apemili. illuoal Crops: , (riti<oI As••• _I' of !toe ''''x;, Tedl_1ogy 243
Sauter, M., Pvon Wiegen, H. LOrz, ond E. Kronz. Syvanen, M. 1994. Horizontal gene transfer: Vollbre<hl, E., and S. Hake. 1995. Oeocieocy
1998. Cell cycle regulatory genes from Evidell<e and possible consequences. Annu. ano~ of femole gametogenesis in
maize ore differentiol~ conirolled during Rev. Genet. 28: 237-61. maize. Dev. Genetics 16: 44-63.
fertilization and ~rst embryonic cell van Dijk, P, and J.van Domme. 2000. Weinmann, P., M. Gos~, W. H~Ien, H. Sujard,
division. Sex. Plant Reprod. 11: 41--48. Apomixis te<hnalogy and the paradax af and C. Gatt. 1994. AchilT1efic
Savidan, Y. (2000). Apomixis: genetics and sex. Trends Plant sa. 5: 81-84. transoctivalor aHOWI tetracydine-
breeding. Plant Breeding Reviews 18: 13- Vermon, H., and M. Pierce. 2000. responsive gene expression in whole
86. Transcriplionol regulation of meiosis in plants. Plant 1. 5: 559-69.
xhena, M., A.M. Uoyd, and R.W. Dovis. 1991. A yeast. CUff. Opin. Cell Bioi. 12: 334-39. Wellord, S.M., J. Gregg. E. Chten, D. Garrison,
sleraid·indicible gene eXllrellion ~tem V'telle-Calzada, J.·P., R. Baskar, and U. PH. Sorensen, Cl Denny, and SJ. Nelson.
for ~ont celk. Proc. Nali. Acod. SO. (USA) Grollniklaus. 2000. Delayed activation af 1998. Detection of differenooly expelsed
88: 10421--425. the palernol genome during seed genes in primary lumor tissues using
xhmidt, E.D., F. Guzza, M.A. Toanen, S.c. de development. Nature 404: 91-94. repesentalionol differell<es anolysis
Vries. 1997. Aleucine-rich repeal Me, J.·P., B.L Bunon, E.c. Bashow, and M.A. coupled to microarroy hybridization. Noel.
containing re<eplor·like kinase marks Hussey. 1995. Eor~ fertilization events in Acids Res. 26: 3059-65.
somatic ~ant celk competentta form the sexual and aposparaus egg apparatus Yong, W.·c., and V. Sundoreson. 2000. Geneti<S
embryos Oevelopmentl24: 2049-62. of Penniselum ciliare (Ll link. PIont J.8: and gametophyte biogenesis in
xa", RJ., M. Spielmon, J.8ailey, and H.G. 309-16. Arobidops;s. CUff. Opin. Plant Bioi. 3: 53-
Oickinson. 1998. Parent-Qf-origin effects V'telle-Cal-zada, J.·P., U.Crone, and 0.1.1. Stel~. 57.
on seed development in Arobidopsis 19960. Apomixis: The asexual revolution. Ye, X., S. AJ.Bobi~, A. Kloti, J.Zhang, P. Lucca, P
thaliona. Development 125: 3329--41. Science274: 1322-23. Seyer, and I. Potrykus. 2000. Engineering
Spongenberg, G., Z.Y. Wang, and I. Polrykus V'telle-Calzada, J..p, M.L Noecio, M.A. Budimon, the provilomine A(Horotene) biosynthetk
1998. 8iote<hnology in foroge and turf 1LThomos, B.L Burron, MA Hussey, and pathway inlo (carolenoid·lreel rke
glall improvement Mooogroplrs on RA Wing. 1996b. Comparative gene endosperm. Science 287: 303-05.
Theoretical andApplied Genetics. Vol. 23. expression in sexual and apanictic avaries Yuan, LP 1993. Progress oltwo-tine ~tem in
Heidelberg: Springer Verlag. of Pennisetum cilia re (Lllink. Plant Mol hybrid rke breeding. In K. Murar.dhoran
Spillane, C. 1999. Recent Developments in Bioi. 32: 108>...92. and lA. Siddig (eds.l, New Frontiers in
Biotechnology 01 TIley Relate to Plant Vinkenoag, R., M. Spielmon, S. Adarm, R.L Rice Research. Hyderabod, India:
Genetic Resources for Food and Agriculture. FlSCher, H.G. Oickinson, and RJ. xo". Directorate of Rice Research. pp. U--93.
FAO Background Study Paper No.9. Rome: 2000. Hypomethylalion promoles Zetka, ht,and A. Rose. 1995. The genetks of
FAO. autonomous endosperm development and meiosis in Coenorhobdilis elegallS. Trends
rescues postferlilization lethality in fie Genel. 11: 27-31.
mulants. Plant Cel/12: 2271-82.