Journal of Integrative Agriculture 2016, 15(0): 60345-7

Available online at www.sciencedirect.com

ScienceDirect

RESEARCH ARTICLE

Characterization and function of Tomato yellow leaf curl virus-

derived small RNAs generated in tolerant and susceptible tomato
varieties

BAI Miao1*, YANG Guo-shun1*, CHEN Wen-ting1, LIN Run-mao2, LING Jian2, MAO Zhen-chuan2,
XIE Bing-yan2

1
College of Horticulture and Landscape, Hunan Agricultural University, Changsha 410128, P.R.China
2
Institute of Vegetables and Flowers, Chinese Academy of Agricultural Sciences, Beijing 100081, P.R.China

Abstract
Virus-tolerant plant, which allows the accumulation of virus and then generates virus-derived small RNAs (vsRNAs), valu-
able materials to reveal the antiviral efficiency of vsRNAs. Here, a comparison of vsRNAs in Tomato yellow leaf curl virus
tolerant and in susceptible tomato varieties showed the consistent trend of vsRNAs’ distribution on virus genome, which is
presented as an obvious characteristic. However, the expression level of vsRNA in tolerant variety is less than that in sus-
ceptible variety. Slicing targets of vsRNA-mediated viral transcripts were investigated using parallel analysis of RNA ends,
and geminivirus DNA methylation was determined by bisulfite sequencing, which uncovered that not all vsRNAs participated
in viral mRNA degradation and DNA methylation. Additionally, by comparing with the expression pattern of vsRNAs, viral
DNA and mRNA, we proposed the quantity of vsRNAs is corresponding to the expression level of viral mRNA, while the
virus-suppression of vsRNAs is not high-efficient.

Keywords: Tomato yellow leaf curl virus, virus-induced RNA silencing, virus-derived small RNA, degradome

2010). Indeed, virus-derived small RNAs (vsRNAs), which

1. Introduction
In the wild, plants may be infected by a variety of viruses.
RNA silencing has been evolved as an immune strategy
against virus infection, termed virus-induced RNA silencing
or RNA-based antiviral immunity (Li and Ding 2006; Ding
Received 29 October, 2015 Accepted 3 February, 2016
BAI Miao, Mobile: +86-18374879885, E-mail: baimiao1984@
qq.com; Correspondence XIE Bing-yan, Tel/ Fax: +86-10-
82105979, E-mail: xiebingyan@caas.cn.
*
These authors contributed equally to this study.
© 2016, CAAS. All rights reserved. Published by Elsevier Ltd.
doi: 10.1016/S2095-3119(15)61315-6
trigger RNA silencing, have been detected in many plants
infected with viruses (Ding 2010; Llave 2010). Similar to the
situation in host plants, the canonical pathway of virus-in-
duced RNA silencing requires host proteins participating in
three stages. First, Dicer-like (DCL) ribonucleases recognize
viral precursors and slice them into 21–24 nt primary vsRNAs
(Blevins et al. 2006; Bouche et al. 2006; Deleris et al. 2006).
vsRNAs are then incorporated into the RNA-induced silencing
complex, which contains a distinct active component, ARGO-
NAUTE (AGO), and guides the target viral RNA degradation
and/or translational inhibition in a sequence-specific manner
(Morel et al. 2002; Jones et al. 2006). By contrast, interaction
with target viral DNA molecules causes transcriptional repres-
sion through the methylation (Raja et al. 2008). Amplification
of vsRNAs involves the activity of host RNA-dependent RNA
 *** et al. Journal of Integrative Agriculture 2016, 15(0): 60345-7
polymerases (RDRs) that synthesize double-stranded small
RNA (dsRNA) with single-stranded RNA (ssRNA) as template
(Garcia-Ruiz et al. 2010; Wang et al. 2010), which serves as
a substrate for the DCL-dependent formation of secondary
vsRNAs. Secondary vsRNAs support the systemic silenc-
ing that spreads throughout the plant (Molnar et al. 2010).
To counteract this host antiviral strategy, viruses have also
evolved genes that encode viral RNA silencing suppressors
(Wu et al. 2010).
Although virus-induced RNA silencing has been proven
to exist widely in host plants, it obviously does not work
efficiently in virus-susceptible plant varieties. A part from
the RNA silencing suppressors encoded by the virus, this
may be explained by insufficiency of vsRNAs. Using a
transgenic strategy, via strong promoter-inducing abundant
virus-homologous small interfering RNA (siRNA), suscepti-
ble plant varieties can acquire viral resistance (Waterhouse
et al. 1998; Zhu et al. 2009). However, the mechanism of
host resistance genes competing against virus molecules is
seemingly different from that of virus-induced RNA silencing.
Tomato yellow leaf curl virus (TYLCV, genus Begomovirus,
family Geminiviridae), with an ssDNA circular genome, in-
fects tomato plants with obvious symptoms (typical yellowing
and curling of the leaves) and causes significant losses in
tomato production (Czosnek 2007; Yadava et al. 2010).
Five genes (loci), named Ty-1 to Ty-5, with different levels
of resistance to TYLCV, have been reported so far. Ty-1,
the first documented locus, is a partially dominant gene that
originated from Solanum chilense accession LA1969 and
was mapped to the distal end of chromosome 6 (Michelson
et al. 1994; Zamir et al. 1994). Recently, Bai Yuling’s group
found that Ty-1 and Ty-3 are allelic and represent a new
class of resistance gene that encodes an RDR belonging to
the RDRγ type (Verlaan et al. 2013). Nevertheless, whether
the amplification of virus-induced RNA silencing by RDR ac-
tivity results in tomato TYLCV resistance remains unknown.
Extremely resistant material can suppress virus repli-
cation in the initial infection period, and identify vsRNAs in
infected single cell is difficult; therefore, most virus-induced
RNA silencing studies have been based on susceptible host
varieties, not on resistant varieties. Upon infection with
TYLCV, tomato plants containing the extreme resistance
gene Ty-2 do not express any vsRNAs which assessed using
deep sequencing. The reaction of the Ty-1 locus-carrying
lines to TYLCV isolates has been described as ‘tolerance’
because the plants became infected (with detectable levels
of viral DNA) but displayed attenuated symptoms (Barbieri
et al. 2010). The study of virus-induced RNA silencing in this
tolerant material and comparison with susceptible varieties
will help us to understand of the relationship between the
two anti-virus strategies (resistance genes and vsRNA), and
will permit the evaluation of vsRNA efficiency.
3
2. Materials and methods
2.1. Plant materials and viral inoculations
Tomato (Solanum lycopersicum) cv. Moneymaker (TYLCV
susceptible strain) and FL505 (containing Ty-1 gene, Asian
Vegetable Research and Development Center, Taiwan of
China) at six to eight leaf stages (five weeks old) were agroi-
noculated with infectious clones of pBinPLUS-SH2-1.4A
that contained TYLCV-[CN:SH2] (AM282874), as described
previously (Zhang et al. 2009). The two samples were
termed ‘MMS’ and ‘TY1S’, respectively. Mock inoculations
was performed by inoculating plants with the Agrobacterium
tumefaciens strain GV3101 containing pBinPLUS, these
samples were called MMC and TY1C, respectively. Inoculat-
ed plants and controls were kept in an insect-free chamber at
25–27°C with 16 h of light per day. PCR was used to check
for the presence of TYLCV in MMS and TY1S. Systemically
infected plant leaves were harvested at 21 and 30 days post
inoculation (dpi), at the same stage. Healthy leaves from
mock-inoculated plants were used as a control. Symptoms
and phenotype before and after virus inoculation are shown
in Appendix A. Each leaf sample was divided into several
parts, each part was used, separately, for small RNA (sRNA)
and degradome library construction by deep sequencing,
RNA extractions to identify viral RNA by quantitative RT-
PCR (qRT-PCR), for total DNA extractions by quantitative
PCR (qPCR) and for bisulfite sequencing PCR (BSP) for
the detection of viral DNA methylation (Appendix B). Three
biological replicates for each treatment were harvested for
RNA isolation (for deep sequencing, all the replicates were
combined into one sample).
2.2. Total DNA and RNA extractions
Total DNA from infected and mock-inoculated tomato leaves
was isolated. RNA was removed with RNase I (Invitrogen,
Carlsbad, CA, USA) treatment and then used for viral DNA
qPCR analyses. Total RNA was isolated using TRIzol re-
agent (Invitrogen), according to the manufacturer’s protocol
and DNA was removed with DNase I (TaKaRa, Japan) treat-
ment. First-strand cDNA was synthesized from 1 μg total
RNA using an oligo(dT) primer and SuperScript™ III cDNA
Synthesis Kit (Invitrogen), according to the manufacturer’s
instructions.
2.3. Quantitative analyses of viral RNA and DNA
Quantitative PCR was performed in 20-µL reactions includ-
ing 20 ng of cDNA synthesized from viral RNA (or viral DNA),
0.2 mmol L–1 primer (primers used are listed in Appendix
C, and 10 µL of SYBR Premix ExTaq (TaKaRa). PCR was
4 *** et al. Journal of Integrative Agriculture 2016, 15(0): 60345-7
performed using Bio-Rad CFX96 Real-Time system (Bio-
Rad, USA) under the following conditions: 95°C for 2 min;
40 cycles of 95°C for 10 s, 62°C for 10 s, 72°C for 10 s, and
95°C for 15 s to obtain melt curves. Each gene was analyzed
in triplicate, after which the average threshold cycle (CT)
was calculated per sample. The tomato β-actin gene was
used as an internal control for normalization. The relative
expression levels were calculated using the 2–ΔΔCt method.
Bio-Rad CFX Manager software (ver. 1.5) was used to obtain
the relative expression levels of each sample.
2.4. Small RNA and degradome library construction
After total RNA isolation, low-molecular-weight RNAs were
isolated as described previously (Hafner et al. 2008). After
ligation of 5´ and 3´ RNA adapters and low-cycle PCR,
sRNAs were collected by polyacrylamide gel purification
and subjected to Solexa high-throughput sequencing using
sequencing by synthesis.
The tomato degradome library was constructed as previ-
ously described (German et al. 2009). In this study, we used
identical samples to those used for sRNA deep sequencing
to identify the degradome by PARE via high-throughput
sequencing. The degradome data of MMC and TY1C were
named as MMCD and TY1CD, respectively; in addition, we
balanced the mix of DNA of 21 and 30 dpi samples of MMS
or TY1S, and named MMSD and TY1SD, respectively.
2.4. Bisulfite sequencing
The bisulfite sequencing procedure was carried out as
previously described (Raja et al. 2008). DNA was isolat-
ed from infected tissues (21 and 30 dpi samples of MMS
and TY1S were collected as one sample, respectively).
Proteinase K digestion was carried out overnight, followed
by bisulfite conversion is using the CT conversion reagent
(EZ-DNA Methylation Gold; Zymo Research). Eight primer
pairs designed against converted template and covering
the whole sense virion (AV, +) strand of viral genome (Ap-
pendix D) were used for overlapping PCR amplification; two
primer pairs, MIRCF and MIRR (Appendix D), were used to
amplify a complementary (AC, –) strand of the intergenic
region (TYLCV coordinates 2 616–2 781 and 1–147). The
PCR products were purified using Promega TA, cloned,
and individual clones were sequenced. For conversion
control, plasmid pBinPLUS-SH2-1.4A was added to a vast
excess of healthy plant DNA extract and treated with the
bisulfite reagent.
2.5. Bioinformatic analyses
The sequence tags from Solexa sequencing were subjected
to data cleaning, which included deleting the low-quality
tags and several kinds of contaminants from the tags. The
3´ adapter sequences were trimmed from raw reads and
sRNAs of 16–27 nt in length were extracted. Afterwards,
standard bioinformatics analyses were used to annotate
the clean tags into different categories. The entire analysis
process is shown in Appendix B. Computational analyses
of sRNA sequences were performed using a set of Perl
scripts. In the alignment and annotation, some sRNA tags
may be mapped to more than one category. To make every
unique sRNA map to only one annotation, we follow the
following priority rule: rRNA/tRNAs/snRNAs/snoRNAs (in
which GenBank>Rfam)>known micro RNA (miRNA)>repeat
associated RNA.
Ultimately, sRNAs were mapped to the TYLCV-[CN:SH2]
and tomato genome using short oligonucleotide analysis
package (SOAP) software. Only sRNA reads that were
100% identical or complementary to TYLCV-[CN:SH2] ge-
nomic sequences were recognized as vsRNAs and those
were extracted for further analysis. siRNA is a 21–24 nt
long double-strand RNA, each strand of which is 2 nt longer
than the other on the 3´ end. According to this structural
feature, we distinguished viral-derived siRNAs (vsiRNAs)
from total vsRNAs.
VMir software (http://www.hpi-hamburg.de/fileadmin/
downloads/VMir.zip) (Grundhoff and Sullivan 2011) was
used to predict viral miRNA precursors (pre-miRNA-like).
According to the degradome sequencing principle,
raw reads were processed to remove 5´ and 3´ adaptor
sequences; tags with sizes of 20 or 21 nt (the sizes ex-
pected from MmeI cleavage) were retained. Tags that did
not correspond to structural RNAs (rRNA, tRNA, snRNA,
and snoRNA) were then mapped to the TYLCV-[CN:SH2]
genome; tags that matched more than one transcript were
repeatedly normalized.
We employed Excel and a set of Perl scripts to statis-
tically analyze the vsRNA data and used Matlab to draw
peak figures. The sRNA library and degradome library
sequencing data are available under NCBI-GEO accession
no. GSE50085.
3. Results
3.1. Distribution and characteristics of Small RNAs
After inoculating TYLCV-[CN:SH2], total sRNA from six
samples (including mock inoculated) was identified by
high-throughput sequencing and bioinformatics analysis.
In each of the six libraries, more than 10 million redundant
tags and more than 3 million non-redundant tags were
harvested (Table 1). After removing 3´ adaptor and other
contaminants, over 98% clean sequences were obtained
 *** et al. Journal of Integrative Agriculture 2016, 15(0): 60345-7 5

in each library, which implied good sequencing quality 60 MMC

sRNAs within the same sample (%)

Relative content of different length
(Appendix E). The data showed that tomato sRNA length MMS-21dpi
50
distribution was mainly 21–24 nt, in which 24-nt sequences MMS-30dpi
TY1C
accounted for up to 30% (Fig. 1). 40
TY1S-21dpi
30 TY1S-30dpi
3.2. Characteristics of vsRNA
20

To investigate the composition of the vsRNA populations, we 10

searched for sRNAs that perfectly matched the sequence of
the TYLCV genome, and vsiRNAs were identified according
to the characteristic of siRNA. The characteristics are shown
in Table 1 and Fig. 2. In previous studies, perfectly matched
vsRNAs varied from 3 to 64% of the total sequenced sRNAs
in virus-infected plants (Donaire et al. 2009; Qi et al. 2009;
Wang et al. 2010). However, in our data, only 0.4–5.49%
were vsRNAs among the total sRNAs in each sample (Ta-
ble 1). There were significantly fewer reads in the tolerant
plant, TY1S, than in the susceptible cultivar, MMS. The
vsRNA proportion of the two samples (21 and 30 dpi) in
MMS were 4.39 and 5.49% by total sRNAs, respectively,
while in TY1S, vsRNAs were 0.4% at 21 dpi and 2.01% at 30
dpi. In addition, no vsRNAs were found in MMC and TY1C,
implying that there is no TYLCV homologous endo-siRNAs
in the two investigated tomato varieties. In MMS and TY1S,
a small proportion of sRNAs (less than 0.03%) mapped to
both the viral and tomato genome. Follow-up analysis found
that these sRNAs from MMS and TY1S did not exist in the
MMC or TY1C sRNA libraries. Therefore, these sRNAs were
considered to come from TYLCV, but not from the host plant.
The size distribution of vsRNAs revealed prominent
peaks for 21- and 22-nt species (more than 80% of total
vsRNAs of each sample), and less than 8% of 24-nt species
in all samples (Fig. 2-A). Our data were somewhat different
from a previous study using TYLCCNV-infected S. lycop-
ersicum and Nicotiana benthamiana plants, in which 22-nt
vsRNAs accumulated to higher levels than the 21-nt species,
but were the same as the data obtained for TYLCCNV and
TYLCCNB when co-infected in plants (Yang et al. 2011).
0
<21 nt 21 nt 22 nt 23 nt 24 nt >24 nt
Fig. 1 Size distribution of total small RNAs (sRNAs).
Histograms showing the size distribution of sRNAs in different
samples. The frequency of redundant total sRNAs is expressed
as a percentage of the total number of sequences for each
sample. MMS and TY1S represent Moneymaker or FL505
plants infected with Tomato yellow leaf curl virus (TYLCV),
respectively, and 21 or 30 days post inoculation (dpi) are
marked. Plants inoculated with pBinPLUS as mock inoculations
are termed MMC and TY1C, respectively.
This indicated that the size distribution of vsRNA may be
diverse within the species of virus and host plants. Given the
differences in length of sRNAs processed by various DCLs,
the above results suggested that certain members of the
tomato DCL family were involved in the pathway of vsRNA
production. It has been reported that sorting of sRNA into
specific AGO complexes is largely conditioned by their 5´
terminal nucleotide (Mi et al. 2008; Takeda et al. 2008; Mach
2010). We investigated the relative abundance of the four
different nucleotides of the vsRNAs from MMS and TY1S.
The result showed the same relative abundance in the four
samples (Fig. 2-B). Subsequent analysis of the relative
abundance in TYLCV-[CN:SH2] genome sequences showed
that it was the same as that of the vsRNAs (Fig. 2-B). Sub-
sequent data analysis has revealed that vsRNAs provide an
almost continuous coverage of every genomic position of
TYLCV-[CN:SH2] (Appendixes F–I). This implied that host
DCL proteins may target any nucleotide position along the

Table 1 Statistics of small RNA sequences from tomato samples1)

Category Total clean Structural RNAs vsRNA vsiRNA
Sample NR R NR R NR R NR R
MMC 5232249 16215708 282351 1532461 – – – –
MMS-21dpi 5202028 16330550 281128 1499385 11786 716029/4.38%2) 8591 481650/67.27%3)
MMS-30dpi 4115022 12390094 339094 2047136 20134 679812/5.49% 9719 489787/72.05%
TY1C 5795418 16152166 344319 1936069 – – – –
TY1S-21dpi 5881226 16978643 341653 2040114 7879 67091/0.40% 2030 23809/35.49%
TY1S-30dpi 3261414 10258175 303100 1994479 13003 206592/2.01% 4824 110594/53.53%
1)
Structural RNAs, including tRNA, rRNA, scRNA, snRNA, snoRNA, repeat sequence; vsRNA, virus-derived small RNA; vsiRNA, viral-
derived siRNA; NR, non-redundant sequences; R, redundant sequences.
2)
/%, percentage means reads of vsRNA compared with total sRNA reads.
5)
/%, percentage means reads of vsiRNA compared with total vsRNA reads.
6 *** et al. Journal of Integrative Agriculture 2016, 15(0): 60345-7

A 30 000 7 000
MMS-21dpi TY1S-21dpi
Number of reads (RPM)
25 000 MMS-30dpi 6 000 TY1S-30dpi
5 000
20 000
4 000
15 000
3 000
10 000
2 000
5 000 1 000
0 0
<21 nt 21 nt 22 nt 23 nt 24 nt >24 nt <21 nt 21 nt 22 nt 23 nt 24 nt >24 nt
B
TYLCV genome
MMS-21dpi TY1S-21dpi MMS-30dpi TY1S-30dpi

U A A A U A U A
U U
31.68% 27.26% 30.82% 26.00% 30.23% 25.26% 30.64% 25.90% 30.35% 25.65%
C C C C
C G G G G
G 20.03% 20.53% 20.00% 20.29%
19.06% 23.15% 23.97% 23.46% 23.72%
22.01%

MMS-21dpi TY1S-21dpi MMS-30dpi TY1S-30dpi

C
100
90
U
Percentage (%)

80
70 G
60
50 C
40
30 A
20
10
0
<2 2 2 2 2 >2 <2 2 2 2 2 >2 <2 2 2 2 2 >2 <2 2 2 2 2 >2
1 1n 2n 3n 4n 4 1 1n 2n 3n 4n 4 1 1n 2n 3n 4n 4 1 1n 2n 3n 4n 4
nt t t t t nt nt t t t t nt nt t t t t nt nt t t t t nt

MMS-21dpi TY1S-21dpi MMS-30dpi TY1S-30dpi

D
100
90
80 +
Percentage (%)

70
-
60
50
40
30
20
10
0
<2 21 22 23 24 >2 <2 21 22 23 24 >2 <2 21 22 23 24 >2 <2 21 22 23 24 >2
1 n n n n 4 1 n n n n 4 1 n n n n 4 1 n n n n 4
nt t t t t nt nt t t t t nt nt t t t t nt nt t t t t nt

Fig. 2 Characteristics of virus-derived small RNAs. A, comparison of reads between different conditions of MMS and TY1S-vsRNAs
in 21 or 30 dpi samples. RPM=reads per million of total sRNA. B, the relative abundance of the four different nucleotides of the
TYLCV-[CN:SH2] genome (TYLCV) and total vsRNAs from MMS and TY1S samples. U=uridine, A=adenosine, C=cytosine, and
G=guanine. C, the relative abundance of the four different nucleotides of vsRNAs 5´ terminus from MMS and TY1S sequences
of different lengths. D, histograms showing the ratio of virion-sense (+) and complementary-sense (−) vsRNAs obtained by deep
sequencing of four samples of different lengths.

virus genome and produced vsRNAs with set at one-nu- of AGOs could be concealed by the activity of various host
cleotide intervals. These results also suggested that when DCLs during subsequent slicing. However, after checking
the data quantity of vsRNA is considerable, the 5´-terminal the base content of vsRNAs of different lengths, we found
nucleotide preference of vsRNA caused by the selectivity that the base-content difference correlated with the diversity
 *** et al. Journal of Integrative Agriculture 2016, 15(0): 60345-7
of vsRNA lengths to some extent (Fig. 2-C). For example,
relative to the viral genome base density, U is more frequent-
ly found in the 5´ terminal nucleotide of 21-, 22- and 23-nt
vsRNAs. Moreover, approximately equal ratios of vsRNAs
from the virion-sense (+) and complementary-sense (–)
strands were detected in all of the samples. Analysis of
vsRNA polarity using reads from each individual size class
or using unique sequences showed an essentially similar
ratio of (+)- and (–)- vsRNAs (Fig. 2-D) and the amount of
vsRNA was consistent in both chains of the gene open read-
ing frame (ORF) region (Appendixes F–I). These findings
suggested that vsRNAs accumulated not only from virus
transcriptions incised by DCL, but also from long dsRNA
synthesized by host plant RDR proteins.
Summarily, the trends of sRNAs from MMS and TY1S
were coherent in their length distribution, base content,
and ratio of the (+)- and (–)- strands (Fig. 2). However,
subsequent analyses found the raw reads of vsRNAs in
MMS were expressed three to ten times higher than those
in TY1S. These phenomena prompted us to perform a more
detailed analysis.
3.3. vsRNA generation and function
In this study, we conducted a detailed analysis and classifi-
cation of vsRNAs, and used a Parallel analysis of RNA ends
(PARE) strategy to detect the targets of the vsRNAs. The
comprehensive analysis of the characteristics of the induced
vsRNAs was designed to increase our understanding of
vsRNA generation, metabolism, and function. Indeed, a
number of interesting phenomena emerged (Fig. 3).
Hotspots As mentioned earlier, vsRNAs were distributed
over the whole viral genome (Appendixes F–I). We con-
structed a distribution density map of vsRNAs (Fig. 3-A and
B) and found that they showed a significant non-random
distribution (i.e., ‘hotspots’), which was similar to previous
studies (Navarro et al. 2009; Qi et al. 2009; Yang et al. 2011;
Miozzi et al. 2013). The hotspots had three main features:
i) The vsRNAs were mainly located in the coding region,
particularly in the overlapping part of the coding region
compared with the non-overlapping zone (Fig. 3-B). For
instance, many vsRNAs were observed from the region of
overlap between V1 and V2 near the 420-nt site, at 1 226–1
633 nt in the overlapping region of C3, C2, and C1, and at
2 171–2 464 nt in the overlap of C1 with the C4. ii) vsRNA
hotspots were also found in severe non-gene overlapping
regions. For example, hotpots were detected both in the
(+)- and (–)- strands of V1 (715 and 884 nt) and C1 (1 704,
1 771, and 2 000 nt). iii) The positions of vsRNAs hotspots
were the same in the susceptible and resistant samples, but
fewer reads were obtained in the TY1S library compared
with the MMS library (Appendixes F–I).
7
vsRNAs and vsiRNAs In our data, some vsRNAs (21–24
nt) multiple sense/antisense sRNA duplexes with two
3´-protruding nucleotides were found at consecutive po-
sitions along the viral genome, which were designated as
viral-derived siRNAs (vsiRNAs) to distinguish them from the
other vsRNAs (we called non-vsiRNA-vsRNAs) (Fig. 3-B,
with gray color). In both of MMS and TY1S, the vsiRNAs
and non-vsiRNA-vsRNAs were mostly 21 and 22 nt, and
some were 24 nt. In these samples, the proportion of
vsiRNA from vsRNA are in the range from 35.47 to 72.05%
(Table 1), which differed from a previous study reported that
most vsiRNAs were vsRNAs (Yang et al. 2011).
In addition, the distribution trends on the TYLCV genome
of vsiRNAs and vsRNAs were consistent (Fig. 3-B). How-
ever, the figure showed that, in some hotspot of vsRNA, the
quantity of vsiRNA was not proportional to vsRNA’s (e.g.,
the site of 715, 886, 1 787 and 2 399 nt, et al.). This result
implied that some vsRNA could not derive from dsRNA,
but from other type of sRNA. Recent work has shown that
from the positive strand of RNA, vsRNAs accumulates
predominantly, suggesting that at least some sRNAs are
produced as miRNA-like duplexes (Akbergenov et al.
2006). A class of virus-encoded miRNA was discovered
in mammalian cells infected by various viruses (Pfeffer
et al. 2004; Grundhoff 2011; Grundhoff and Sullivan 2011).
However, unlike in mammalian systems, virus-encoded
miRNA-like molecules have not yet been detected in host
plants. We used the VMir software to predict viral miRNA
precursors (pre-miRNA-like) from sense transcripts: the
vast majority of bona fide pre-miRNA-like sequences reach
values >115. Nine viral pre-miRNA-like sequences were
predicted (Fig. 3-C). The pre-miRNA-like sequences were
located mostly in vsRNA hotspots of non-gene-overlapping
areas. We also found some pre-miRNA-like sequences
located in hotspots comprising significantly higher yields of
non-vsiRNA-vsRNA (Fig. 3-B, masked with gray). miRNAs
are characterized by an asymmetrical miRNA/miRNA* ratio;
therefore, we suggested that the vsRNAs located in hotspots
of non-gene-overlapping areas may be caused by their viral
miRNA-like composition.
Slicing targets of vsRNAs vsRNAs could recruit AGOs
and mediate viral mRNA degradation through the principle
of base pairing. As vsRNAs were distributed over the entire
viral genome, it seemed unreasonable to assume that the
target sites covered the whole viral genome, simply using
bioinformatics methods. In this study, a high-throughput
and direct identification of the target method (German et al.
2008; German et al. 2009), PARE, was used for degradome
sequencing in the identical copy of samples that were used
for sRNA deep sequencing (Appendix B). More than 16
million redundant tags were harvested from MMCD, TY1CD,
MMSD, and TY1SD, respectively. In each of the samples,
8 *** et al. Journal of Integrative Agriculture 2016, 15(0): 60345-7

A TYLCV-[CN:SH2] genome:
148 498 1 084 1 226
1 V2 V1 1 633 2 171 C4 2 464 2 781
IR C2 IR
308
1 081 C3 1 485 1 542 C1 2 615

B vsRNA & vsiRNA :

2 400 424
1 800 MMS-21dpi
886 2 211
1 200 1 704
564 +
600 186
0
52 106
600
Number of reads (RPM)

1 200 1 517 1 787 2 007

_
1 800 2 399
400 1 252
2 400 429 715 1 380
3 000
240
120 TY1S-21dpi +
0
60
180 _
300
2 400
1 800
MMS-30dpi
1 200
+
600
0
600
1 200
1 800 _
Number of reads (RPM)

2 400
3 000
600
300 TY1S-30dpi +
0
300 _
600
900

C Viral pre-miRNA-like:
MD1:166.2 MD2:140.8 MD3:133.9 MD4:123
MR9:124.5 MR8:153.7 MR6:150.9 MR5:162.6
MR7:174.1

D Targets from degradome:

6.25
5.00 MMS
3.75
2.50 +
1.25
0
Number of reads (RPM)

1.25
2.50
_
3.75
5.00
6.25
2.5
2.0 TY1S
1.5 +
1.0
0
1.0
_
1.5
2.0
2.5

E Methylated sites:
5
4
3 MMS
2 +
1
0
1
Methylated cytosines

2
3 _
4
5
5
4
3 TY1S +
2
1
0
1
2
3 _
4
5

Fig. 3 Mapping of vsRNAs, degradome and methylated sites. A, genome organization of TYLCV-[CN:SH2] is indicated in linear
form (the 5´-3´ orientation is from right to left), each open reading frame (ORF) is marked and viral transcripts are denoted by arrows
indicating the 5´-3´ orientation. Genes of virion-sense (AV, +) or complementary-sense (AC, −) strand polarity are in orange or blue,
respectively. IR: intergenic region. AV1, AV2, AC1, AC2, AC3, and AC4 encode the capsid protein (CP), replication initiator protein
(Rep), a transcriptional activator protein (TrAP), a replication enhancer protein (REn), and a symptom and movement determinant,
respectively. B, expression of vsRNAs and vsiRNAs in four samples. Perfectly matched reads of vsRNAs are presented in both
strands of the TYLCV genome. Frequency of vsRNAs at each locus was calculated and peak figures were constructed by Matlab
software. vsRNAs were divided into two groups: vsiRNAs (+, orange; −, blue) and non-vsiRNAs-vsRNAs, which are presented in
gray. Significant hotspots sites are also marked. C, viral pre-miRNA-like sequences were predicted by VMir software and marked
in red on the TYLCV genome. MD1–4 and MR5–9 are predicted from the transcripts of virion-sense (AV) and complementary-
sense (AC) strand, respectively, and values were calculated by VMir software after tagging. D, reads and sites of perfectly matched
degradome sequences are statistics plotted to the (+) (orange) and (−) (blue) strands of the TYLCV genome. Two time samples
(21 and 30 dpi) were collected as one sample, marked as MMS and TY1S, respectively. E, the frequency of methylation of each
cytosine in the whole virion-sense strand (+) genome and the two strands of the IR region, as analyzed by the bisulfite sequencing
PCR (BSP). Samples surveyed as the same as described in D. Each bar represents the position of a cytosine (guanine in
complementary-sense strand) in the TYLCV genome. Ten independent clones were sequenced and analyzed for each sample.
 *** et al. Journal of Integrative Agriculture 2016, 15(0): 60345-7
21–22 nt tags accounted for more than 98% of the total
reads, which meant that our date was reliable (Appendix J).
When matched to the TYLCV genome, 709 (with 13 931 raw
reads) and 613 (with 2 896 raw reads) targets were identified
in MMSD and TY1SD, respectively (Appendixes K and L).
No target was found in MMCD and TY1CD. These targets
were allocated to the TYLCV genome map (Fig. 3-D). Most
targets corresponded to the direction of the ORF-coding
strand; for instance, those located in the regions of the V1
and V2 genes were in the virion-sense orientation and those
located in C1-4 were in the complementary orientation. This
implied that vsRNAs induced slicing of sense viral gene
transcripts. In addition, the targets were distributed mainly
in the mRNA 3´ end. This may reflect the efficiency of the
reverse-transcription step before sequencing (German et al.
2009). It could also have been caused by the highly compe-
tent vsRNA slicing on the 3´ end compared with the 5´ end.
Interestingly, at the vsRNA hotspot of 2 007 bp (Fig. 3-B),
a significant degradome peak existed on the V chain. The
basically consistent distribution of targets in MMS and TY1S
suggested that sRNAs that participated in viral-mRNA slicing
are stable. Nevertheless, from a quantitative point of view
(row reads of targets divided by total reads, see Appendixes
K and L), we suspected that only some of the vsRNAs are
involved in viral transcript slicing.
Most of the transcript data on begomoviruses have come
from analyses using Tomato golden mosaic virus (TGMV),
African cassava mosaic virus (ACMV), Abutilon mosaic
virus (AbMV), or Tomato leaf curl virus (ToLCV); however,
there are no transcript data on TYLCV (Gronenborn 2007).
This study was the first to use the degradome to provide
unequivocal evidence of vsRNA-induced viral mRNA slicing.
Our results may provide clues for the study of viral mRNA
transcripts. As shown in Fig. 3-D, over the whole genome,
including the non-gene (even in the IR region), degradation
molecules existed on both the (+)- and (–)- strands. Only
the transcript with poly(A) tails could be detected by de-
gradome sequencing (German et al. 2009), therefore, this
result suggested that the transcripts of TYLCV covered the
entire viral genome.
3.4. Viral DNA methylation
In the intergenic region, higher numbers of vsRNAs came
from the (–)- strand (80%) than the (+)- strand (Appendix
M). The intergenic region (IR) also contained more 24-nt
vsRNAs (Appendix M). Does this means that DCL3’s par-
ticipation in the formation of the 24-nt vsRNAs targeting viral
DNA methylation represents an additional antiviral pathway
(Seemanpillai et al. 2003)? To answer this question, we
performed bisulfite sequencing PCR (BSP) to determine
methylated cytosines. First, we identified methylation of
9
the whole virion-sense chain of TYLCV, and found that the
degree of methylation was extremely low; only 14 sites on
MMS and 12 sites on TY1S were methylated (Fig. 3-E), this
seemingly showed a different result to many other works
(Sharma et al. 2013; Sahu et al. 2014). The 24-nt vsRNA
covered almost the whole TYLCV genome (Appendixes F–I);
therefore, this result implied that in our samples, the degree
of 24 nt vsRNA-induced DNA methylation in the virion-sense
chain is limited. We also investigated the methylation of the IR
region in the complementary strand. Interestingly, we found
that this region was highly methylated: all cytosines of MMS
were methylated (Fig. 3-D, Appendix N). There are more 24-
nt vsRNAs in the IR region than in the other regions and they
are present mostly on the complementary strand (Appendix
O); therefore, we hypothesized that the efficiency of 24-nt
vsRNA-induced methylation was dependent on the site and
traceable chain of vsRNAs. However, why vsRNA-induced
methylation has regional selectivity awaits further research.
3.5. Generation of vsRNAs depends on viral mRNA
Interestingly, regardless of the distribution of vsRNA, the
cutting targets or methylation patterns are surprisingly
similar; however, the numbers are higher for MMS than
TY1S. We attempted to find out the reason underlying this
observation. After investigating vsRNAs, we detected the
amount of viral mRNA and DNA, and these showed some
interesting results:
i) Viral DNA production was investigated using IR prim-
ers by qPCR (Fig. 4-A). Relative to β-actin, the viral DNA
content in both MMS and TY1S were increased from 21
to 30 dpi, with the fold of 1.67 (506.08/302.88) and 2.02
(453.85/224.19), respectively. Compared with the virus
loadings of two samples at the 30 dpi, we found that the
quantity of virus DNA of TY1S was attached to MMS’s
(MMS:TY1S=506.08/453.85=1.12). This implied that viral
DNA replication has not been effectively suppressed by
Ty-1, after the virus have enough time to infected plants,
systemically.
II) Specific primers were designed (Appendix D) and
qRT-PCR was performed to investigate the expression of
each viral gene (Fig. 4-B). We found that the relative ex-
pression tendency of viral mRNA from each ORF region was
consistent within the samples. From the view of average
expression, MMS or TY1S showed no significant differ-
ences between 30 and 21 dpi samples (relative to β-actin
expression, MMS-30dpi/MMS-21dpi=0.89 and TY1S-30dpi/
TY1S-21dpi=1.04, respectively). The viral mRNA expres-
sion difference in the two samples (21 and 30 dpi) of MMS
was larger than in TY1S (MMS-21dpi/TY1S-21dpi=17.18
times and MMS-30dpi/TY1S-30dpi=14.16 times on aver-
age, respectively). This result implied that Ty-1 effectively
10 *** et al. Journal of Integrative Agriculture 2016, 15(0): 60345-7

A Viral DNA B Viral mRNA

10 7

Relative Fold Expression

MMS-21dpi MMS-30dpi
Relative Fold Expression

600 MMS 9
TY1S TY1S-21dpi 6 TY1S-30dpi
500 8
7 5
400 6 4
300 5
4 3
200 3 2
100 2
1 1
0 0 0
21dpi-sample 30dpi-sample ActinV2 V1 C3 C2 C4 C1 ActinV2 V1 C3 C2 C4 C1

n
C vsRNA D

tio
Host cell

a
lic
Reads per base site (RPM)

MMS-21dpi MMS-30dpi

p
25 35 Viral DNA

Re
TY1S-21dpi TY1S-30dpi
30
20

ion
25

Tr
a
lat
15 20

ns
thy

cr
15 Viral protein

ipt
10

Me

ion
10 Ty-1 Tr
5 an
5 RDR sla
tio
0 0 vsRNA DCL n
IR V2 V1 C3 C2 C4 C1 IR V2 V1 C3 C2 C4 C1 Viral mRNA
Slicing/Supression

Fig. 4 Relative quantification of vsRNAs, viral transcripts, and DNAs. A, viral DNA expressions related to tomato β-actin identified
in MMS and TY1S. B, transcription levels of viral mRNA in TYLCV-infected plants. The viral mRNA expression level of each gene
was investigated relative to tomato β-actin. Each bar represents the mean standard error of triplicate readings and the values are
calculated using the comparative 2–ΔΔCt method. C, vsRNA abundance of four samples assigned to each viral gene. The reads of
vsRNAs in each gene region were divided by the length of the gene to obtain the probability of a vsRNA emerging from one base
site in each gene region (RPM=reads per million of total vsRNA). This value was then used to evaluate the relationship with each
viral gene, compared by their trends in expression levels. D, overview of the TYLCV virus molecule circulating in the host plant cell.

inhibited TYLCV mRNA production. 4. Discussion

III) The relative abundance of the vsRNAs was split into
different gene regions (reads in gene-overlapping areas 4.1. Role of host RNA-silencing core proteins
were equally distributed to each gene) (Fig. 4-C). The
generated frequency showed that the distribution trend of As core components of plant RNA silencing, DCLs, AGOs,
vsRNAs and viral mRNA in each of the four samples was and RDRs participate in various functions in a hierarchical
basically the same. Quantitatively, MMS or TY1S showed manner. Antiviral RNA silencing also depends on some
differences between the 21 and 30 dpi samples (on average, of these core factors participating in the biogenesis and
the fold change were MMS-30dpi/MMS-21dpi=1.25 times activity of vsRNAs.
and TY1S-30dpi/TY1S-21dpi=6.32 times, respectively, com- Four DCL proteins have been identified in Arabidopsis
pared with 0.89 and 1.04 times for viral mRNA); the vsRNA thaliana (Margis et al. 2006). Genetic studies have revealed
expression difference of two samples (21 and 30 dpi) of MMS the hierarchical access of DCL4, 3, and 2 to viral dsRNA in
was larger than for TY1S (MMS-21dpi/TY1S-21dpi=11.66 the biogenesis of distinct vsiRNA size classes (Bouche et al.
times and MMS-30dpi/TY1S-30dpi=2.31 times, on average, 2006; Moissiard and Voinnet 2006; Du et al. 2007; Donaire
respectively, compared with 17.18 and 14.16 times for viral et al. 2008; Qu et al. 2008; Garcia-Ruiz et al. 2010). DCL4
mRNA). Although not completely accurate, these figures is the primary sensor of viral dsRNAs and produces 21-nt
could imply that the amount of vsRNA positively correlated vsRNAs. In the absence of DCL4, 22- and 24-nt vsRNAs
with viral mRNA. are produced by DCL2 and 3, respectively (Blevins et al.
We note that the total number of vsRNAs in MMS was 2006; Deleris et al. 2006; Fusaro et al. 2006; Diaz-Pendon
12.34 times higher than that in TY1S, while the viral mRNA et al. 2007). DCL1, is probably a minor contributor to vsR-
expression ratio of MMS/TY1S was 26.51, and the viral NA formation in plants infected with viruses (Bouche et al.
DNA production ratio of MMS/TY1S was 1.115, on average. 2006; Moissiard and Voinnet 2006; Donaire et al. 2008;
These results indicate that the Ty-1 gene did not significantly Garcia-Ruiz et al. 2010). Earlier, we identified four DCLs
inhibit viral DNA accumulation, but it may directly or indirectly of the DCL2 family and many of these showed up-regulated
affect the production of viral transcripts. expression under TYLCV infection (Bai et al. 2012). This
 *** et al. Journal of Integrative Agriculture 2016, 15(0): 60345-7
implied that DCL2 could play a crucial role in antiviral immu-
nity. In the present study, 21–24 nt vsRNA were identified
at various concentrations, suggesting that four classes of
DCLs participate in the generation of vsRNAs. The iden-
tification of a high proportion of 24-nt vsRNA and the DNA
hypermethylation in the IR region implied that the function
of the 24-nt vsRNA, mediated by DCL3, was also critical in
the fight against DNA viruses. Ten AGO proteins have been
identified in Arabidopsis, a proportion has been functionally
associated with different classes of sRNAs (Vaucheret
2008). AGO1, AGO2, and AGO5 selectively bind vsiRNAs
with a 5´ terminal nucleotide preference (Zhang et al. 2006;
Takeda et al. 2008). However, high-throughput sequencing
results showed that the preference is not reflected in our
samples. However, the consistency of the base ratio of
vsRNAs compared with the TYLCV genome suggested
that AGOs showed the bias in selecting vsRNAs could be
masked by DCLs’ subsequent slicing activity.
In plants, RDR1 or RDR6 produces secondary vsiRNAs
following viral RNA replication-triggered biogenesis of prima-
ry siRNAs (Curaba and Chen 2008; Vaistij and Jones 2009;
Wang et al. 2010). RDR2 has a crucial role in the biogenesis
of hc-siRNAs that induce a DNA methylation pathway, with
the participation of AGO4 and DCL3 (Xie et al. 2004; Lu et al.
2006; Kasschau et al. 2007; Raja et al. 2008). The ratios
of (+)- and (–)- strands vsRNAs in TYLCV-infected tomato
cells implied that vsRNA mainly from the dsRNA may be
processed by host RDRs. Several vsRNA hotspots were
observed in the gene-overlapping region and we believe this
reflects the involvement of RDRs continued participation at
the degradative PCR (Lipardi et al. 2001). Compared with
non-gene-overlapping areas, in the gene-overlapping areas,
there are more viral transcripts that are recognized by DCLs
and processed to vsiRNAs. Subsequently, these vsiRNAs
serve as primary siRNAs to pair with nascent viral transcripts
and triggering dsRNA synthesis by some host RDRs. This
would lead to an induction of vsRNAs exponential growth in
those gene-overlapping regions. That hypothesis still cannot
explain the generation of hotspots in non-gene-overlapping
regions. We proposed that vsRNA hotspots in these regions
are formed as viral pre-miRNA-like sequences. It should
be noted that, in the non-gene-overlapping areas, there are
several hotspots not covered by pre-miRNA-like sequenc-
es. This may be the result of imprecision of the prediction
of software, or other factors that affect vsRNA production.
Thus, both the stem-loop structural features of viral RNA
and the intermolecular base pairing of viral RNA strands
synthesized by RDRs contribute to the origin of hotspots
in the infected cells. However, it still cannot explain why
hotspots emerged at the same position. Further studies
should be performed to clarify the dynamic interactions of
vsRNAs in the host.
11
4.2. vsRNA function and the interplay between plants
and viruses
Genome-scale approaches, along with the deep sequencing
of vsRNAs and their targets, could shed light on the origins
and compositions of vsRNAs and their potential function
for controlling viral gene expression in the virus-plant in-
teraction.
In virus-infected host plants, vsRNAs were identified
that mostly cover the whole viral genome. Nevertheless,
direct evidence for viral mRNA targets was lacking. RACE
is a method for identified sRNA slicing targets but it is a not
high-throughput method. Now, a novel approach called
PARE for high-throughput identification of slicing targets of
sRNA from the RNA degradome has been developed (Ger-
man et al. 2009). Relying on the PARE strategy, we showed
that not all vsRNAs participated in slicing viral mRNA. Inter-
estingly, the number of slicing targets is proportional to the
number of vsRNAs, rather than the inverse ratio, which did
not support the hypothesis that the potency of virus-induced
RNA silencing is positively correlated with the abundance of
vsRNAs (Vanitharani et al. 2003; Aliyari et al. 2008).
In subsequent analyses, by comparing the relations
among viral DNA, mRNA, and vsRNA and its functional
characteristics, we put forward a model to explain the
relationship between Ty-1 gene and virus-induced RNA
silencing (Fig. 4-D).
I) In the susceptible variety, the normal growth cycle of
viral DNA in the host proceeded, including replication, tran-
scription, and translation. Using viral mRNA transcripts, the
host RNA silencing system-related factors produced vsRNA
that negatively regulated viral expression. However, the
frequency of vsRNAs is directly correlated with the structure,
content, and distribution of viral transcripts, rather than with
the inverse ratio of its viral mRNA slicing. Combining the
above observations with the limited slicing of viral transcripts
from degradome, we hypothesized that the amount of viral
mRNA determines the number of vsRNAs, but the negative
controlling function of vsRNAs is limited. This because, first,
a ‘slicer’ activity has only been demonstrated for some of
the AGO family members (Hutvagner and Simard 2008),
and redundant but incapable vsRNAs could compete with
the AGOs or other RNA silencing system elements and
lead to reduced RNA silencing. Thus, we have to redefine
the hypothesis concerning the function of vsRNA. Perhaps
below a certain threshold vsRNA function is proportional
to its content. Therefore, vsRNA below the threshold can
achieve effective inhibition of the virus, but this is also
dependent on the strength of viral replication. Beyond this
threshold, the accumulation of vsRNA is a reflection of viral
mRNA transcript content, and its role of virus inhibition will
reach a plateau.
12 *** et al. Journal of Integrative Agriculture 2016, 15(0): 60345-7
II) Assaying viral DNA content showed that, in the tolerant
material, the Ty-1 gene does not have a significant influence
on viral DNA replication, but has an obvious inhibitory effect
on viral mRNA transcription. Although it has been validated
that host plants could resist DNA viruses by methylation
(Bian et al. 2006; Raja et al. 2008; Rodríguez-Negrete
et al. 2009; Flower et al. 2011; Yadav and Chattopadhyay
2011) and that the viral suppressors could avoid post tran-
scriptional gene silencing (PTGS) or DNA methylation from
the host plant (Raja et al. 2008; Hohn and Vazquez 2011),
the consistency of the BSP results implied that Ty-1 is not
directly involved in viral DNA methylation of a particular area
and that no suppressors inhibited viral DNA methylation.
III) After inoculation with TYLCV, no vsRNA and viral
DNA was detected in tomato plants containing the com-
plete resistance gene, Ty-2 (data not shown). However, in
the variety containing the tolerance gene, Ty-1, viral DNA
replication was not inhibited effectively. One explanation
is that Ty-1 might rely on the RNA silencing pathway to
withstand the virus. The hypothesis is based on the fol-
lowing: first, vsRNAs are derived from viral transcripts;
therefore, the viral inhibitory action of vsRNA always lags
behind the transcription and replication of the virus. As
the vsRNA couldn’t completely suppress viral replication,
the viral DNA could be accumulated gradually after each
replication cycle. However, compared with the allele ty-1,
the existence of Ty-1 led to more effective viral inhibition in
the initially infected cells. This hypothesis conforms to the
function of Ty-1 as an RDR (Verlaan et al. 2013). Whether
Ty-1 actually carries out typical RDR catalysis of vsRNAs
amplification in virus-induced RNA silencing requires more
experimental evidence.
5. Conclusion
In this paper, after analyzed Tomato yellow leaf curl
virus-derived small RNAs generated in tolerant and sus-
ceptible tomato varieties, and compared the expression
pattern of vsRNAs, viral DNA and mRNA, we found that
not all vsRNAs participated in viral mRNA degradation
and DNA methylation. Additionally, by comparing with the
expression pattern of vsRNAs, viral DNA and mRNA, we
proposed the quantity of vsRNAs is corresponding to the
expression level of viral mRNA, while the virus-suppression
of vsRNAs is not high efficienct. These results indicated
that, in tolerant material, when after virus infection, espe-
cially in late after triggered systemic infection, the content
of vsRNA may not reflect the material resistance ability.
This paper is designed to remind readers should be paid
attention to this point in related studies by the high-through-
put sequencing of vsRNAs.
Acknowledgements
This work was supported by the Specialized Research Fund
for the Doctoral Program of Higher Education of China
(20134320120013) and the Natural Science Foundation of
Hunan Province, China (14JJ3095). The authors gratefully
acknowledge Prof. ZHOU Xue-ping, Chinese Academy of
Agricultural Sciences, for the gift of a infectious clone of
TYLCV-[CN:SH2].
Appendix associated with this paper can be available on
http://www.ChinaAgriSci.com/V2/En/appendix.htm
References
Akbergenov R, Si-Ammour A, Blevins T, Amin I, Kutter C,
Vanderschuren H, Zhang P, Gruissem W, Meins Jr F,
Hohn T, Pooggin M M. 2006. Molecular characterization of
geminivirus-derived small RNAs in different plant species.
Nucleic Acids Research, 34, 462–471.
Aliyari R, Wu Q, Li H W, Wang X H, Li F, Green L D, Han C
S, Li W X, Ding S W. 2008. Mechanism of induction and
suppression of antiviral immunity directed by virus-derived
small RNAs in Drosophila. Cell Host Microbe, 4, 387–397.
Bai M, Yang G S, Chen W T, Mao Z C, Kang H X, Chen G H,
Yang Y H, Xie B Y. 2012. Genome-wide identification of
Dicer-like, Argonaute and RNA-dependent RNA polymerase
gene families and their expression analyses in response to
viral infection and abiotic stresses in Solanum lycopersicum.
Gene, 501, 52–62.
Barbieri M, Acciarri N, Sabatini E, Sardo L, Accotto G P,
Pecchioni N. 2010. Introgression of resistance to two
mediterranean virus species causing tomato yellow leaf curl
into a valuable traditional tomato variety. Journal of Plant
Pathology, 92, 485–493.
Bian X Y, Rasheed M S, Seemanpillai M J, Ali Rezaian M. 2006.
Analysis of silencing escape of tomato leaf curl virus: An
evaluation of the role of DNA methylation. Molecular Plant-
Microbe Interactions, 19, 614–624.
Blevins T, Rajeswaran R, Shivaprasad P V, Beknazariants D,
Si-Ammour A, Park H S, Vazquez F, Robertson D, Meins Jr
F, Hohn T, Pooggin M M. 2006. Four plant Dicers mediate
viral small RNA biogenesis and DNA virus induced silencing.
Nucleic Acids Research, 34, 6233–6246.
Bouche N, Lauressergues D, Gasciolli V, Vaucheret H. 2006. An
antagonistic function for Arabidopsis DCL2 in development
and a new function for DCL4 in generating viral siRNAs.
The EMBO Journal, 25, 3347–3356.
Curaba J, Chen X. 2008. Biochemical activities of Arabidopsis
RNA-dependent RNA polymerase 6. The Journal of
Biological Chemistry, 283, 3059–3066.
Czosnek H. 2007. Tomato Yellow Leaf Curl Virus Disease.
Springer, The Netherlands.
Deleris A, Gallego-Bartolome J, Bao J, Kasschau K D,
 *** et al. Journal of Integrative Agriculture 2016, 15(0): 60345-7
Carrington J C, Voinnet O. 2006. Hierarchical action and
inhibition of plant Dicer-like proteins in antiviral defense.
Science, 313, 68–71.
Diaz-Pendon J A, Li F, Li W X, Ding S W. 2007. Suppression
of antiviral silencing by Cucumber mosaic virus 2b protein
in Arabidopsis is associated with drastically reduced
accumulation of three classes of viral small interfering
RNAs. The Plant Cell, 19, 2053–2063.
Ding S W. 2010. RNA-based antiviral immunity. Nature reviews.
Immunology, 10, 632–644.
Donaire L, Barajas D, Martinez-Garcia B, Martinez-Priego L,
Pagan I, Llave C. 2008. Structural and genetic requirements
for the biogenesis of Tobacco rattle virus-derived small
interfering RNAs. Journal of Virology, 82, 5167–5177.
Donaire L, Wang Y, Gonzalez-Ibeas D, Mayer K F, Aranda
M A, Llave C. 2009. Deep-sequencing of plant viral small
RNAs reveals effective and widespread targeting of viral
genomes. Virology, 392, 203–214.
Du Q S, Duan C G, Zhang Z H, Fang Y Y, Fang R X, Xie Q, Guo
H S. 2007. DCL4 targets Cucumber mosaic virus satellite
RNA at novel secondary structures. Journal of Virology,
81, 9142–9151.
Flower K, Thomas D, Heather J, Ramasubramanyan S, Jones
S, Sinclair A J. 2011. Epigenetic control of viral life-cycle
by a DNA-methylation dependent transcription factor. PLoS
ONE, 6, e25922.
Fusaro A F, Matthew L, Smith N A, Curtin S J, Dedic-Hagan J,
Ellacott G A, Watson J M, Wang M B, Brosnan C, Carroll B J,
Waterhouse P M. 2006. RNA interference-inducing hairpin
RNAs in plants act through the viral defence pathway.
EMBO Reports, 7, 1168–1175.
Garcia-Ruiz H, Takeda A, Chapman E J, Sullivan C M, Fahlgren
N, Brempelis K J, Carrington J C. 2010. Arabidopsis RNA-
dependent RNA polymerases and dicer-like proteins in
antiviral defense and small interfering RNA biogenesis
during Turnip mosaic virus infection. The Plant Cell, 22,
481–496.
German M A, Luo S, Schroth G, Meyers B C, Green P J. 2009.
Construction of parallel analysis of RNA ends (PARE)
libraries for the study of cleaved miRNA targets and the
RNA degradome. Nature Protocols, 4, 356–362.
German M A, Pillay M, Jeong D H, Hetawal A, Luo S,
Janardhanan P, Kannan V, Rymarquis L A, Nobuta K,
German R, De Paoli E, Lu C, Schroth G, Meyers B C,
Green P J. 2008. Global identification of microRNA-target
RNA pairs by parallel analysis of RNA ends. Nature
Biotechnology, 26, 941–946.
Gronenborn B. 2007. The tomato yellow leaf curl virus genome
and function of its proteins. In: Czosnek H, ed., Tomato
Yellow Leaf Curl Virus Disease. Springer, The Netherlands.
pp. 67–84.
Grundhoff A. 2011. Computational prediction of viral miRNAs.
Methods in Molecular Biology, 721, 143–152.
Grundhoff A, Sullivan C S. 2011. Virus-encoded microRNAs.
Virology, 411, 325–343.
Hafner M, Landgraf P, Ludwig J, Rice A, Ojo T, Lin C, Holoch
13
D, Lim C, Tuschl T. 2008. Identification of microRNAs and
other small regulatory RNAs using cDNA library sequencing.
Methods, 44, 3–12.
Hohn T, Vazquez F. 2011. RNA silencing pathways of plants:
silencing and its suppression by plant DNA viruses.
Biochimica et Biophysica acta, 1809, 588–600.
Hutvagner G, Simard M J. 2008. Argonaute proteins: Key
players in RNA silencing. Nature reviews. Molecular Cell
Biology, 9, 22–32.
Jones L, Keining T, Eamens A, Vaistij F E. 2006. Virus-
induced gene silencing of argonaute genes in Nicotiana
benthamiana demonstrates that extensive systemic
silencing requires Argonaute1-like and Argonaute4-like
genes. Plant Physiology, 141, 598–606.
Kasschau K D, Fahlgren N, Chapman E J, Sullivan C M, Cumbie
J S, Givan S A, Carrington J C. 2007. Genome-wide profiling
and analysis of Arabidopsis siRNAs. PLoS Biology, 5, e57.
Li F, Ding S W. 2006. Virus counterdefense: diverse strategies
for evading the RNA-silencing immunity. Annual Review of
Microbiology , 60, 503–531.
Lipardi C, Wei Q, Paterson B M. 2001. RNAi as random
degradative PCR: siRNA primers convert mRNA into
dsRNAs that are degraded to generate new siRNAs. Cell,
107, 297–307.
Llave C. 2010. Virus-derived small interfering RNAs at the
core of plant-virus interactions. Trends in Plant Science,
15, 701–707.
Lu C, Kulkarni K, Souret F F, MuthuValliappan R, Tej S S,
Poethig R S, Henderson I R, Jacobsen S E, Wang W,
Green P J, Meyers B C. 2006. MicroRNAs and other small
RNAs enriched in the Arabidopsis RNA-dependent RNA
polymerase-2 mutant. Genome Research, 16, 1276–1288.
Mach J. 2010. Differentiation among the ARGONAUTES. The
Plant Cell, 22, 294.
Margis R, Fusaro A F, Smith N A, Curtin S J, Watson J M,
Finnegan E J, Waterhouse P M. 2006. The evolution
and diversification of Dicers in plants. FEBS Letters, 580,
2442–2450.
Mi S, Cai T, Hu Y, Chen Y, Hodges E, Ni F, Wu L, Li S, Zhou H,
Long C, Chen S, Hannon G J, Qi Y. 2008. Sorting of small
RNAs into Arabidopsis Argonaute complexes is directed by
the 5´ terminal nucleotide. Cell, 133, 116–127.
Michelson I, Zamir D, Czosnek H. 1994. Accumulation and
translocation of Tomato yellow leaf curl virus (TYLCV) in
a Lycopersicon esculentum breeding line containing the
L. chilense TYLCV tolerance gene Ty-1. Phytopathology,
84, 928–933.
Miozzi L, Pantaleo V, Burgyan J, Accotto G P, Noris E. 2013.
Analysis of small RNAs derived from Tomato yellow leaf curl
Sardinia virus reveals a cross reaction between the major
viral hotspot and the plant host genome. Virus Research,
2, 287–296.
Moissiard G, Voinnet O. 2006. RNA silencing of host transcripts
by cauliflower mosaic virus requires coordinated action of
the four Arabidopsis Dicer-like proteins. Proceedings of
the National Academy of Sciences of the United States of
14 *** et al. Journal of Integrative Agriculture 2016, 15(0): 60345-7
America, 103, 19593–19598.
Molnar A, Melnyk C, Bassett A, Hardcastle T, Dunn R,
Baulcombe D. 2010. Small silencing RNAs in plants are
mobile and direct epigenetic modification in recipient cells.
Science, 328, 872.
Morel J B, Godon C, Mourrain P, Beclin C, Boutet S,
Feuerbach F, Proux F, Vaucheret H. 2002. Fertile
hypomorphic ARGONAUTE (ago1) mutants impaired in
post-transcriptional gene silencing and virus resistance.
The Plant Cell, 14, 629–639.
Navarro B, Pantaleo V, Gisel A, Moxon S, Dalmay T, Bisztray
G, Di Serio F, Burgyan J. 2009. Deep sequencing of viroid-
derived small RNAs from grapevine provides new insights
on the role of RNA silencing in plant-viroid interaction. PLoS
ONE, 4, e7686.
Pfeffer S, Zavolan M, Grasser F A, Chien M, Russo J J, Ju J,
John B, Enright A J, Marks D, Sander C, Tuschl T. 2004.
Identification of virus-encoded microRNAs. Science, 304,
734–736.
Qi X, Bao F S, Xie Z. 2009. Small RNA deep sequencing
reveals role for Arabidopsis thaliana RNA-dependent RNA
polymerases in viral siRNA biogenesis. PloS ONE, 4, e4971.
Qu F, Ye X, Morris T J. 2008. Arabidopsis DRB4, AGO1,
AGO7, and RDR6 participate in a DCL4-initiated antiviral
RNA silencing pathway negatively regulated by DCL1.
Proceedings of the National Academy of Sciences of the
United States of America, 105, 14732–14737.
Raja P, Sanville B C, Buchmann R C, Bisaro D M. 2008. Viral
genome methylation as an epigenetic defense against
geminiviruses. Journal of Virology,, 82, 8997–9007.
Rodríguez-Negrete E A, Carrillo-Tripp J, Rivera-Bustamante R
F. 2009. RNA silencing against geminivirus: complementary
action of posttranscriptional gene silencing and
transcriptional gene silencing in host recovery. Journal of
Virology, 83, 1332–1340.
Sahu P P, Sharma N, Puranik S, Prasad M. 2014. Post-
transcriptional and epigenetic arms of RNA silencing: A
defense machinery of naturally tolerant tomato plant against
Tomato leaf curl New Delhi Virus. Plant Molecular Biology
Reporter, 32, 1015–1029.
Seemanpillai M, Dry I, Randles J, Rezaian A. 2003.
Transcriptional silencing of geminiviral promoter-driven
transgenes following homologous virus infection. Molecular
Plant-Microbe Interactions, 16, 429–438.
Sharma N, Sahu P P, Puranik S, Prasad M. 2013. Recent
advances in plant-virus interaction with emphasis on small
interfering RNAs (siRNAs). Molecular Biotechnology, 55,
63–77.
Takeda A, Iwasaki S, Watanabe T, Utsumi M, Watanabe Y.
2008. The mechanism selecting the guide strand from small
RNA duplexes is different among argonaute proteins. Plant
Cell Physiology, 49, 493–500.
Vaistij F E, Jones L. 2009. Compromised virus-induced gene
silencing in RDR6-deficient plants. Plant Physiology, 149,
1399–1407.
Vanitharani R, Chellappan P, Fauquet C M. 2003. Short
interfering RNA-mediated interference of gene expression
and viral DNA accumulation in cultured plant cells.
Proceedings of the National Academy of Sciences of the
United States of America, 100, 9632–9636.
Vaucheret H. 2008. Plant argonautes. Trends in Plant Science,
13, 350–358.
Verlaan M G, Hutton S F, Ibrahem R M, Kormelink R, Visser
R G, Scott J W, Edwards J D, Bai Y. 2013. The Tomato
yellow leaf curl virus resistance genes Ty-1 and Ty-3 are
allelic and code for DFDGD-class RNA-dependent RNA
polymerases. PLoS Genetics, 9, e1003399.
Wang X B, Wu Q, Ito T, Cillo F, Li W X, Chen X, Yu J L,
Ding S W. 2010. RNAi-mediated viral immunity requires
amplification of virus-derived siRNAs in Arabidopsis
thaliana. Proceedings of the National Academy of Sciences
of the United States of America, 107, 484–489.
Waterhouse P M, Graham H W, Wang M B. 1998. Virus
resistance and gene silencing in plants can be induced
by simultaneous expression of sense and antisense RNA.
Proceedings of the National Academy of Sciences of the
United States of America, 95, 13959–13964.
Wu Q, Wang X, Ding S W. 2010. Viral suppressors of RNA-
based viral immunity: Host targets. Cell Host & Microbe,
8, 12–15.
Xie Z, Johansen L K, Gustafson A M, Kasschau K D, Lellis A D,
Zilberman D, Jacobsen S E, Carrington J C. 2004. Genetic
and functional diversification of small RNA pathways in
plants. PLoS Biology, 2, E104.
Yadav R K, Chattopadhyay D. 2011. Enhanced viral intergenic
region-specific short interfering RNA accumulation and DNA
methylation correlates with resistance against a geminivirus.
Molecular Plant-Microbe Interactions, 24, 1189–1197.
Yadava P, Suyal G, Mukherjee S K. 2010. Begomovirus
DNA replication and pathogenicity. Current Science, 98,
360–368.
Yang X L, Wang Y, Guo W, Xie Y, Xie Q, Fan L J, Zhou X P.
2011. Characterization of small interfering RNAs derived
from the geminivirus/betasatellite complex using deep
sequencing. PLoS ONE, 6, e16928.
Zamir D, Ekstein-Michelson I, Zakay Y, Navot N, Zeidan M,
Sarfatti M, Eshed Y, Harel E, Pleban T, van-Oss H. 1994.
Mapping and introgression of a Tomato yellow leaf curl virus
tolerance gene, TY-1. Theoretical and Applied Genetics,
88, 141–146.
Zhang H, Gong H, Zhou X. 2009. Molecular characterization
and pathogenicity of Tomato yellow leaf curl virus in China.
Virus Genes, 39, 249–255.
Zhang X, Yuan Y R, Pei Y, Lin S S, Tuschl T, Patel D J, Chua
N H. 2006. Cucumber mosaic virus-encoded 2b suppressor
inhibits Arabidopsis Argonaute1 cleavage activity to counter
plant defense. Genes & Development, 20, 3255–3268.
Zhu C X, Song Y Z, Yin G H, Wen F J. 2009. Induction of
RNA-mediated multiple virus resistance to Potato virus Y,
Tobacco mosaic virus and Cucumber mosaic virus. Journal
of Phytopathology, 157, 101–107.
(Managing editor ZHANG Juan)