Domestic Animal Endocrinology 59 (2017) 116–133

Contents lists available at ScienceDirect

Domestic Animal Endocrinology

journal homepage: www.domesticanimalendo.com

Systemic and intrafollicular components of follicle selection

in mares
O.J. Ginther a, b, *
a
Eutheria Foundation, Cross Plains, WI 53528, USA
b
Department of Pathobiological Sciences, School of Veterinary Medicine, University of Wisconsin-Madison, Madison, WI 53706, USA

a r t i c l e i n f o a b s t r a c t

Article history: Mares are superb models for study of follicle selection owing to similarities between mares
Received 20 October 2016 and women in relative follicle diameters at specific events during the follicular wave and
Received in revised form 6 December 2016 follicle accessibility for experimental sampling and manipulation. Usually, only 1 major
Accepted 10 December 2016
follicular wave with a dominant follicle (DF) greater than 30 mm develops during the 22 to
24 d of the equine estrous cycle and is termed the primary or ovulatory wave. A major
Keywords:
secondary wave occasionally (25%) develops early in the cycle. Follicles of the primary
Dominant follicle
wave emerge at 6 mm on day 10 or 11 (day 0 ¼ ovulation). The 2 largest follicles begin to
Follicle deviation
Follicle selection deviate in diameter on day 16 when the future DF and largest subordinate follicle (SF) are
IGF 23 mm and 20 mm, respectively. The deviation process begins the day before diameter
Mare deviation as indicated in the future DF but not in the future SF by (1) increase in promi-
Subordinate follicle nence of an anechoic layer and vascular perfusion of the wall and (2) increase in follicular-
fluid concentrations of IGF1, vascular endothelial growth factor, estradiol, and inhibin-A.
A systemic component of the deviation process is represented by suppression of circu-
lating FSH from secretion of inhibin and estradiol from the developing DF. Production of
inhibin is stimulated by IGF1 and LH, and estradiol is stimulated by LH and not by IGF1 in
mares. A local intrafollicular component involves the production of IGF1, which apparently
increases the responsiveness of the future DF to FSH. The roles of the IGF system have been
well studied in mares, but the effect of IGF1 on increasing the sensitivity of the follicle cells
to FSH is based primarily on studies in other species. The greater response of the future DF
than the SF to the low concentrations of FSH is the essence of selection. During the
common growth phase that precedes deviation, diameter of the 2 largest follicles increases
in parallel on average when normalized to emergence or retrospectively to deviation.
Study of individual waves indicates that (1) the 2 follicles change ranks (relative diameters)
during the common growth phase in about 30% of primary waves and (2) after ablation of
1, 2, or 3 of the largest follicles at the expected beginning of deviation, the next largest
retained follicle becomes the DF indicating that several follicles have the capacity for
dominance; therefore, it is proposed that the deviation process represents the entire
mechanism of follicle selection in mares.
Ó 2016 Elsevier Inc. All rights reserved.

1. Introduction length of the interovulatory interval (IOI) is about 22 d

In mares, the ovulatory season in the northern
temperate zone extends from about April to October, the
* Corresponding author. Tel.: þ1 608 798 3777; fax: þ1 608 798 3722.
E-mail address: oj.ginther@wisc.edu.
(horses) and 24 d (ponies), and ovulation occurs from a
large follicle (eg, 40 mm) near the end of a long estrus (eg,
6 d) [1]. One or 2 follicular waves occur during an IOI, and
each consists of many follicles (eg, 10) that reach greater
than 6 mm in diameter. The ovulatory wave develops
13-mm follicles by day 10 or 11 (day 0 ¼ ovulation) at the

0739-7240/$ – see front matter Ó 2016 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.domaniend.2016.12.005
 O.J. Ginther / Domestic Animal Endocrinology 59 (2017) 116–133
peak of the wave-stimulating FSH surge. The 2 largest
follicles grow approximately in parallel (common growth
phase) until day 15 or 16 when the 2 largest follicles are a
mean of 23- and 20-mm, and the FSH surge is declining [2].
At the end of the common growth phase, follicle deviation
begins and is characterized by continued growth of the
largest follicle to become the dominant follicle (DF) and the
beginning of slower growth rate of the second largest
follicle to become the largest subordinate follicle (SF).
During deviation, the follicle destined to ovulate must
escape a degenerative process (atresia), which affects the
contemporaries of the wave. The DF of the last wave of an
IOI becomes the ovulatory follicle.
In heifers, the process of deviation begins about 12 h
before the manifestation of diameter deviation and
involves an increase in granulosa LH receptors and estra-
diol-17b (estradiol) and maintenance of intrafollicular free
IGF1 in the future DF (review [3]). Thereby, the future DF is
developmentally prepared to use the declining FSH in the
wave-stimulating FSH surge. A follicle that emerges first
may maintain the largest diameter ranking and become the
DF by being first to reach a critical developmental stage.
However, the early size advantage is not a requisite
component of the deviation process in heifers as indicated
by (1) occasional switching in follicle diameter rankings
during the common growth phase owing to intraovarian
factors that affect growth of individual follicles, (2) a follicle
that reaches 5 mm regardless of diameter ranking will
become a DF unless it is selected against during deviation,
and (3) a follicle is not developmentally prepared to
become a DF until it is larger than 7.0 mm. It has been
proposed that the mechanisms that underlie deviation and
begin 12 h before the diameter manifestation of deviation
fully represent the theory of follicle selection in cattle [3].
The biological selection of the DF or ovulatory follicle
from a wave of many follicles in monovulatory species
(cattle, horses, and humans) has been the subject of hun-
dreds of reports during more than 5 decades (reviews that
include mares [1,4–10]). Comparisons and contrasts of
follicle selection or the deviation mechanism between
mares and women are also available [11–13]. The present
review is a history of published research results on follicle
selection in mares before and after the introduction of
transrectal ultrasonic imaging for the study of follicle dy-
namics [14]. A purpose of the review is to examine whether
the deviation process may represent the entire follicle se-
lection mechanism in mares as proposed for heifers [3].
2. Chronology of research studies on follicle selection
in mares
A 90-yr history (1926–2016) on observations and
research findings that contributed to our knowledge on the
mechanisms that underlie the follicle selection process in
mares is listed (Table 1). After an original observation in
1926, the number of reports increased gradually during
1972 to 2001 followed by many reports in 2002 to 2004 and
a gradual decrease in 2005 to 2016. The progress was
disjointed as expected owing at least partly to diverse
interests and in the availability of instrumentation and
expertize. Perhaps, the most salient milestones in research
117
progress on follicle selection in mares were the introduc-
tion of an FSH assay and transrectal ultrasonic imaging and
development of the concept of follicle deviation. Deviation
or the dissociation in diameters of the 2 largest follicles was
discovered in 1997, during the study of diameter graphs of
each of the 2 largest follicles in individual waves. The
growth rate of the future DF and the future SF changed
relatively abruptly and served as a manifestation of selec-
tion as represented by the point of departure in changing
diameters between the future DF and SF. The observation of
diameter deviation allowed focusing on a specific event as a
reference point or end point for studies on follicle selection.
The influx of reports in 2002 to 2004 was related to (1) the
development of transvaginal ultrasound techniques for
sampling or injecting into a specific follicle including later
sampling or determining whether the follicle became a DF
or SF, (2) transrectal color Doppler ultrasound for assessing
ovarian and follicular blood flow, and (3) the use of
experimental deviation to compare time lines among sys-
temic and intrafollicular factors.
3. Reports on selection before availability of
ultrasonic imaging
Before the availability of ultrasonic imaging, sequential
monitoring of follicle diameters throughout the IOI in
mares was done by transrectal digital palpation. On the
basis of palpation, it was reported in 1926 that several
follicles of 20 to 30 mm are present before estrus, but
during estrus, only 1 or 2 chosen follicles increase in
diameter to an average of 50 mm [15]. The observation was
subsequently confirmed [61]. Although the term follicle
selection was not used, the phenomenon has been known
in mares for at least 90 yr. A palpation study in 1972 noted
that the follicle that was the largest on the first day of estrus
(mean length, 7.7 d) remained largest and became the
ovulatory follicle in 15 of 24 (62%) mares [16]. Daily
monitoring by palpation throughout the IOI in 14 pony
mares in 1979 indicated that the number of follicles greater
than 20 mm did not increase significantly until after Day 10
[24]. The number of 20-mm follicles began to decrease on
Day 7 or on the first day that the preovulatory follicle was
the only follicle to increase in diameter (Fig. 1) [24]. In
another palpation study, the preovulatory follicle became
the largest follicle on the first or second day of estrus in 64%
and 20%, respectively, of 56 estrous cycles or on a mean of
5.7 d before ovulation [24]. Although these findings were
described before the availability of ultrasonic imaging, they
provided the first indications that selection of the preovu-
latory follicle in mares is manifested by diameter dissoci-
ation between the largest follicle and other follicles
beginning about 7 d before ovulation.
4. Transrectal and transvaginal ultrasonic imaging
The use of transrectal ultrasonic imaging for research in
reproduction in large animals has been reviewed [62]. The
first study on transrectal ultrasonic monitoring of repro-
ductive structures in any species was done in mares and
was reported in 1980 [25]. The ultrasound instrument at
that time had relatively low transducer frequency
118 O.J. Ginther / Domestic Animal Endocrinology 59 (2017) 116–133

Table 1
A 90-yr chronology leading to the theory of follicle selection in mares.

1926. Transrectal palpation indicated that only 1 or 2 follicles increase in diameter during estrus and are chosen to become the ovulatory follicles [15].
1972. The follicle that was the largest on the first or second day of estrus (mean length, 7.7 d) remained the largest and became the ovulatory follicle in
most mares [16].
1972. Discovery in Thoroughbreds of a 25% incidence of a preovulatory-sized DF with ovulation during diestrus indicating the occasional presence of
another follicular wave [17].
1973. Demonstration of essentiality of gonadotropins for follicle growth by treatment with an antiserum to a pituitary extract [18].
1973. Characterization of concentrations of circulating LH during the equine estrous cycle [19].
1974. First demonstration of hormonal induction of ovulation and multiple ovulations in mares [20].
1975. First report on assay of circulating FSH. Two FSH surges were found during an estrous cycle [21].
1978. Report that circulating LH is considerably elevated during the ovulatory season, and ovarian steroids rhythmically modify the seasonal effect
during estrous cycles [22].
1979. Demonstration of an FSH-inhibitory effect presumably from inhibin by treatment with a steroid-free equine follicular fluid [23].
1979. Number of 20-mm follicles begins to increase on day 10 and decrease 7 d before ovulation along with differential growth of the preovulatory
follicle [24].
1980. Introduction of transrectal ultrasonic imaging of large follicles (8 mm or greater) [25].
1981. A steroid-free fraction of follicular fluid (inhibin) and estradiol has a synergistic effect in suppressing FSH [26].
1984. First characterizations by ultrasound of changing follicle diameters beginning with 2-mm follicles throughout the IOI with differential growth of
the ovulatory follicle beginning 7 d before ovulation [14].
1987. Greater content of LH receptors in the presumptive ovulatory follicle 1 or 2 d before the ovulatory follicle is identifiable by largest diameter [27].
1989. Introduction of ultrasonic tracking of individual follicles that were 15 mm or greater [28].
1989. Demonstration using an inhibin antiserum of the negative effect of inhibin on FSH and follicles [29].
1991. First report on circulating concentrations of immunoreactive inhibin throughout the estrous cycle; the relationship between inhibin and FSH was
inverse [30].
1992. First documentation that emergence of each major waves (with a DF) and each minor wave (without a DF) is temporally associated with an FSH
surge [31,32].
1997. First characterization of the common growth phase and use of the term deviation. The FSH surge reached a peak when follicles were about 13.0
mm and was declining at deviation (22.5 mm future DF) [2].
1999. Echotextural changes in the follicle wall distinguish the future DF from the future SF a day before diameter deviation [33].
1999. Estradiol begins to increase in follicular fluid and circulation the day before diameter deviation [34].
1999. The increasing estradiol and decreasing FSH at the expected beginning of deviation are functions solely of the future DF [35].
1999. An LH increase is not needed for the deviation process but is needed for full growth of the DF after deviation [36].
2001. Decline in the FSH surge is a function of multiple follicles before deviation and the future DF alone at and after deviation [37].
2001. Immunoreactive inhibin (apparently b-subunits) increases during the decline in the FSH surge [37,38].
2001. Both inhibin and estradiol contribute to the continuing FSH decline after deviation [38].
2001. Concentration of LH increases and then decreases encompassing deviation. The transient LH increase does not affect the intrafollicular aspects of
the deviation process but stimulates FSH suppressors (estradiol, inhibin) [38].
2002. Free IGF1, estradiol, inhibin-A, and activin-A increase in follicular fluid of the future DF during the day before the beginning of diameter deviation [39].
2002. IGF binding proteins increase in the future SF at the beginning of deviation [39].
2002. Addition of FSH to cultured equine granulosa cells stimulates IGF binding proteins [40].
2002. Functional relationships between IGF1 and binding proteins are mediated by specific proteases in the follicular fluid [41].
2002. The absence of an estradiol increase in the future SF is associated with reduced utilization of an androgen substrate [39].
2002. After ablation of the future DF at expected diameter deviation, IGF1 begins to increase in SF before the induced deviation and estradiol, inhibin-A,
and activin-A begin to increase after the induced deviation [42].
2003. Estradiol suppresses LH and the combination of inhibin and estradiol suppresses FSH near the beginning of deviation [43].
2003. Inhibin and season (photoperiod) interact to regulate FSH concentrations [44].
2004. First characterization of the frequency of changes in diameter ranking among follicles during the common growth phase, and the capacity of each
follicle to become the future DF at the end of the common growth phase [45].
2004. Diameters at deviation are similar between major waves of the ovulatory and anovulatory season, but the underlying systemic and intrafollicular
hormone mechanisms are quite different [46].
2004. Description of a predeviation follicle (incidence; 37%) that attains the diameter of the future DF before deviation but does not become a DF [12].
2004. Color-Doppler study reveals greater blood flow in the wall of the future DF than in the SF beginning a day before diameter deviation [47].
2004. The increased concentrations of inhibin-A, activin-A, and VEGF but not estradiol in the future DF at the beginning of diameter deviation are
functions of endogenous IGF1 [48].
2004. Injection of an IGF-binding protein into the future DF at deviation decreases the concentrations of free IGF1, estradiol, inhibin-A, and activin-A and
stops the growth of the DF [49].
2005. Injection of the protease (PAPP-A) into the future SF at the beginning of deviation increases the concentrations of free IGF1, inhibin-A, follistatin,
and VEGF [50].
2005. Surges of FSH and LH are shaped by the negative effect of ovarian hormones on the positive effect of photoperiod [51].
2007. Estradiol has a negative effect on LH throughout the LH ovulatory surge presumably by dampening the positive effect of photoperiod [52].
2008. Induction of a wave by follicle ablations and PGF2a treatment on day 10 resulted in a more prominent FSH surge and more double ovulations [53].
2009. Length of interval from ovulation to deviation (mean, 15.6 d) was significantly correlated between 2 successive IOIs in individuals indicating
repeatability [54].
2009. Two separate deviations about 2 d apart result in 2 DFs (incidence 34%) [55].
2009. Demonstrations that FSH is essential for follicle growth before and after the peak of the FSH surge [56].
2010. A GnRH antagonist and replacement of LH indicated that LH is not needed for growth of the DF during the 48 h after deviation [57].
2010. The increase in free IGF1 in association with deviation is gonadotropin dependent [58].
2015. Estradiol increases and anti-Müllerian hormone decreases concomitantly in growing follicles before deviation [59].
2016. Growth of preantral follicles is greater for the CL ovary than that for the non-CL ovary [60].
2016. Proposal that the deviation process represents the entire theory of follicle selection in mares (present review).

Abbreviations: CL, corpus luteum; DF, dominant follicle; IGF, insulin-like growth factor; IOI, interovulatory interval; PGF2a, prostaglandin F2a; PPAP-A,
pregnancy-associated plasma protein-A; SF, largest subordinate follicle; VEGF, vascular endothelial growth factor.
 O.J. Ginther / Domestic Animal Endocrinology 59 (2017) 116–133
4.0 1979
3.5
3.0
Follicles ≥ 20 mm
Number
2.5
2.0
Largest follicle
1.5
1.0
Diestrus Estrus
1 4 7 10 13 16 –7 –4 –1
Number of days from ovulation
Fig. 1. Mean  standard error of the mean number of follicles 20 mm or
greater, and diameter of the largest follicle in 27 mares as determined by
transrectal palpation. These data provided the first indication of the days of
development of the primary wave including selection of the ovulatory fol-
licle. Adapted from [24].
(3.0 MHz); and therefore, follicles smaller than 8 mm were
not identifiable. Although not significant, the increasing
mean diameter of the future ovulatory follicle appeared to
dissociate from diameter of the second largest follicle about
6 d before ovulation consistent with the report [24] that
was based on digital palpation.
In 1984, the introduction of 5.0-MHz transrectal ultra-
sound transducers for study of equine [14] and bovine [63]
ovaries provided greater resolution and allowed the first
characterizations of follicle diameters including 2-mm
follicles throughout the IOI in each species. The accuracy
of the instrumentation in each species was demonstrated
by a high correlation (greater than 0.90) between in situ
measurements by ultrasonic imaging and grossly after
removing and slicing the same ovaries. The study in mares
[14] confirmed the increase in number of 20-mm follicles
beginning at mid-diestrus and demonstrated the changing
follicle diameters of the 2 largest follicles during the
ovulatory wave (Fig. 2). The identity of follicles from day to
day was unknown and likely accounts for the depicted
greater diameter of the SF than when identity is main-
tained. Beginning on day 8, the 2 largest follicles grew in
parallel for 7 d (common growth phase) and began to differ
in growth rate 6 d before ovulation (equivalent to day 16).
Growth rate of the largest follicle continued unabated from
the point of dissociation, whereas the second largest follicle
decreased in growth rate for a day, reached a static phase,
and then began to regress. These relationships in pony
mares were subsequently confirmed including dissociation
in diameter between the 2 largest follicles 6 or 7 d before
ovulation in Quarter Horses and Appaloosas [64,65]. An
increase and decrease in number of 2- to 5-mm follicles
occurred between Days 1 and 8 (minor wave), and dif-
ferences in diameter changes among months during the
estrous cycles of the ovulatory season were described. The
changes in the follicle population during the estrous cycle
and season were compatible with earlier reports on
changes in FSH concentrations [21,66,67].
119
50 40 1984 Ov
45 Ov Primary
35 wave
40
Diameter (mm)
30
Diameter (mm)
LSD
35
25
Largest follicle
30
20
25
15
20 2nd largest follicle
10
–3 0 3 6 9 –9 –6 –3 0 3
Days from ovulation (Ov)
Fig. 2. Mean diameter of the largest follicles as determined by the trans-
rectal ultrasonic imaging. The vertical LSD bar represents magnitude of the
least significant difference. These data first demonstrated the time lines for
origin, parallel growth of the 2 largest follicles, and the beginning of
dissociation in diameters. LSD, least significant difference; OV, ovulation.
Adapted from [14].
Tracking or maintaining the identity of individual folli-
cles 15 mm or greater from examination to examination
was reported for mares in 1989 [28]. The tracking technique
has had a profound effect on characterizing follicular waves
and follicle diameters associated with deviation (review
[68]). In addition to providing for direct noninvasive
sequential monitoring of the changing diameter of indi-
vidual follicles, tracking facilitated studying the changing
echotexture of the follicle walls [33] and estimating the
changes in follicle vascular perfusion by color Doppler
ultrasound [47]. These additional end points have indicated
which follicle is destined to become the DF before the fol-
licle is identifiable by diameter differences (Section 8).
In addition to transrectally determining follicle di-
ameters and characteristics, ultrasonic imaging has greatly
impacted knowledge on the deviation process in mares by
serving as the basis for other experimental approaches.
Transvaginal ultrasound guiding for accessing follicles in
association with transrectal manipulation of an ovary has
been an effective tool (review [68]). Sampling the fluid of a
specific follicle can be done transvaginally [34], or follicle
cells can be recovered as demonstrated in cattle [69].
Sampling of a small amount of fluid (20 mL) for estradiol
assay from the largest follicle when it attained 15 mm or
20 mm did not alter the day of deviation or diameter of the
2 largest follicles at deviation [34]. Injecting an intra-
follicular factor into a specific follicle with subsequent
sampling of the follicular fluid has been used to determine
the effect of a factor on other factors or on follicle diameter
[50,70]. In one study [70], the future SF was sampled (50 or
90 mL) and treated at the expected beginning of deviation
and then sampled twice during the next 24 h. The follicles
were initially reduced in diameter but returned to the
original growth rate and follicle outcome (dominant, sub-
ordinate, and ovulatory).
The ablation of follicles by ultrasound-guided trans-
vaginal aspiration of follicle contents [2] is useful for initi-
ating a new wave uncomplicated by the intermingling of
120 O.J. Ginther / Domestic Animal Endocrinology 59 (2017) 116–133
follicles from a previous wave [2,36,58] and studying the
endocrinologic effect of specific follicles [37]. A useful
procedure is the ablation of all follicles at a specified
diameter (eg, greater than 5 mm) on day 10. The follicles of
the new wave can be tracked beginning at a smaller
diameter (eg, 6 mm) than when a new wave is not induced
(eg, 15 mm). Ablation of the largest follicle at the beginning
of expected deviation when it reaches 22.5 mm is being
used to study the timing of events involved in the conver-
sion of SF to a new DF (experimental or induced deviation)
[42]. The transvaginal route can be used also for biopsy of
the CL [71], a section of an antral follicle, or a section of
ovarian tissue that may include preantral follicles [60].
5. Follicular waves in mares
The topic of follicular waves is germane to this review in
that it raises the question whether selection of a DF occurs
in other waves in addition to the ovulatory wave. A major
wave develops a DF of greater than 28 or 30 mm, [72] with
or without ovulation. The largest follicle in a minor follic-
ular wave does not attain more than 28 mm; and therefore,
there is apparent absence of deviation or incomplete de-
viation. The terms primary and secondary major follicular
waves have been used for the equine estrous cycle [1].
A primary wave is a major wave that originates during mid-
diestrus and gives origin to a DF that terminates in the
primary ovulation (ovulation at end of IOI). A secondary
major wave emerges during late estrus or early diestrus in
about 25% of Standardbreds and Thoroughbreds [28],
Quarter horses and Appaloosas [72], and Brazilian Breton
mares (small draft horses) [12]. A diestrous ovulation dur-
ing a secondary wave is common (eg, 25% of cycles) in some
breeds of horse mares such as Thoroughbreds [17] and
Standardbreds [28], unusual in some other horse breeds
such as Quarter horses and Appaloosas (eg, 4% of cycles)
[73], and rare in ponies [16]. Mares and breeds without a
major secondary wave often have a minor wave at the time
of a secondary wave in other mares or breeds. In pony
mares [18] and miniature mares [74], a secondary wave is
rare, but a minor wave may occur in early diestrus. A study
of excised ovaries in pony mares in 1973 [18] found that a
minor follicular wave occurred on average in pony mares
from stimulation by a gonadotrophin during the first week
of the IOI. Major anovulatory and ovulatory waves and
minor waves also occur during the first 40 d of pregnancy
[31,32,75].
Critical study of deviation in secondary waves has not
been reported and is needed. Inspection of published [28]
diameter profiles in 11 mares indicated that follicle diam-
eter deviation (a manifestation of selection) occurred dur-
ing 2 of 3 major secondary waves when the future DF was
about 24 mm. The effect of mare age on the deviation
mechanism also has not been reported. Comparative study
of the characteristics and robustness of deviation under
different hormonal milieu during various reproductive
states may be useful for studying the mechanisms of de-
viation (primary vs secondary waves, ovulatory secondary
vs anovulatory secondary waves, anovulatory waves during
the ovulatory vs anovulatory season, and young vs old
mares).
During the transitional period before the first ovulation
of the ovulatory season, 1 to several major anovulatory
waves with a DF of 30 mm or more may emerge about
every 10 d preceding the first ovulation of the ovulatory
season [1,65,76,77]. Based on diameter alone, deviation can
be judged to occur during major waves of the anovulatory
season [78]. During the decline in the FSH surge after the
peak, FSH concentration is greater for a 30-mm anovula-
tory follicle than for a 30-mm ovulatory follicle indicating a
deficiency in the FSH-suppressing abilities of the follicle.
The DF during anovulatory waves develops under hor-
monal deficiencies [46,78], including low concentrations of
LH and estradiol [46,77–79]. However, the peak concen-
tration of a wave-stimulating FSH surge is similar between
anovulatory and ovulatory waves [79]. Blood flow area in
an anovulatory DF after 25 mm remains considerably lower
than that in an ovulatory DF [68] despite similar increases
in diameter of the anovulatory and ovulatory follicles.
Follicular-fluid concentrations of estradiol, IGF1, inhibin-A,
and vascular endothelial growth factor (VEGF) in 30-mm
anovulatory follicles are lower than those in 30-mm
ovulatory follicles. Based on the temporal increase in FSH,
functional impairment of the future dominant anovulatory
follicle begins when the follicle is approximately 20 mm
indicating that the apparent deviation in anovulatory
waves is a different phenomenon from deviation in
ovulatory waves. These findings indicate similarities in
follicle diameter at deviation between major waves of the
anovulatory and ovulatory seasons, but the underlying
hormonal events during apparent deviation are profoundly
different between seasons. In addition, the largest follicle
reaches different maximum mean diameters between the
anovulatory and ovulatory waves, indicating that the
mechanisms for continued growth of the DF after the
initiation of deviation are also different between the 2
types of major waves. Both major and minor follicular
waves also occur during the anovulatory months immedi-
ately after the end of the ovulatory season [80].
6. Use of the terms divergence, deviation, and
common growth phase
Divergence is a term that orginated in cattle (review [3])
and mares [72] in the early 1990s to describe the gradual
average dissociation in diameter between the future DF and
largest SF beginning at a reference point early in a wave (eg,
ovulation, emergence at a specified diameter). Plotting of
the diameter of the 2 largest follicles with ovulation as the
reference point in mares results in gradual departure in
diameters (divergence) for 1 or 2 d before the future DF is at
the diameter characteristic of deviation (Fig. 3) [2,34,81].
However, a study of individual diameter graphs of each of
the 2 largest follicles in each wave rather than the averages
in cattle in 1996 [82] and in ponies in 1997 [2] indicated
that the dissociation in diameters usually occurred rela-
tively abruptly. The relatively abrupt dissociation was
termed deviation to distinguish it from the term divergence
for gradual dissociation. The impression of gradual disso-
ciation (divergence) was attributed to the abrupt dissocia-
tion (deviation) occurring at different times in individual
waves relative to a previous fixed reference point as shown
 O.J. Ginther / Domestic Animal Endocrinology 59 (2017) 116–133 121

35 1997 Predeviation follicle

35
Dominant
30 follicle
Dominant
30 follicle
25 (n = 27)

Diameter (mm)
20 Divergence
25
Deviation
Subordinate
15 follicle
20 Predeviation
follicle
10 (n = 10)
Number deviating Subordinate
on indicated days 15 follicles
5 (1) (2) (6) (2) (2) (1)
Diameter (mm)

0 1 2 3 4 5 6 7 8 9 10
Days after future dominant follicle was 6 mm –5 –4 –3 –2 –1 0 1 2 3 4 5
Days from observed beginning of deviation
35
Dominant Fig. 4. Mean  standard error of the mean diameters of 3 follicles types
30 follicle during primary waves in 27 mares showing the relationship of a predeviation
follicle to other follicles. The number of waves with various follicle types is
shown. Dominant and subordinate follicles are centralized to the beginning of
25 e diameter deviation as represented by the horizontal broken line. The prede-
as
ph
viation follicle is centralized to mean day at the beginning of its regression
h
20 wt
which occurred on the day before diameter deviation. Adapted from [12].
gro
o n- Subordinate
15 m follicle
m identifiable in a published diameter profile of individual
Co
follicles in a mare in each of 2 reports involving different
10 Diameter breeds and laboratories [4,28]. Unlike a minor wave that
deviation may impinge on the major wave, the predeviation follicle
5 emerges along with the follicles of the major wave but
emerges before the DF when centralized to the beginning
–6 –5 –4 –3 –2 –1 0 1 2 3 4 of regression 1 d before diameter deviation of DF. The
Days from beginning of diameter deviation mechanism associated with a predeviation follicle is un-
known, but it appears to reach a diameter characteristic of
Fig. 3. Mean  standard error of the mean diameter of the dominant and deviation of a future DF before a mechanism is in place to
subordinate follicles showing gradual dissociation (divergence) when begin the deviation process in the future DF (Section 7).
normalized to emergence (upper panel) and a more abrupt dissociation
(deviation) when centralized to observed deviation (lower panel). The future
subordinate follicle does not reach a diameter compatible with the initiation 7. The common growth phase and deviation
of the deviation process before the larger follicle reaches diameter deviation.
This study first used the term deviation in mares and characterized the
The first specific study of the common growth phase
growth profiles of the tracked follicles that became the dominant and largest
subordinate follicles. Adapted from [2]. and follicle deviation during the primary wave in mares
was done in ponies in 1997 [2]. The operator was over-
whelmed by the large number of follicles greater than
(Fig. 3). The abrupt deviation observed in individual waves 5 mm from intermingling of follicles from a previous wave.
in mares is preceded by a common growth phase that ex- Regressing follicles of the previous wave masked the day-
tends from follicle emergence (eg, at 6 mm) to the begin- to-day retrospective identity of the future DF and SF dur-
ning of diameter deviation. The end of the common growth ing the common growth phase that ends in deviation for
phase and the beginning of deviation are synonymous. the primary wave. To remove the regressing follicles and to
Studies of follicle selection in mares in this laboratory for reduce the number of follicles in the primary wave, a
the past 20 yr beginning with the 1997 report [2] used 2-follicle model was developed, wherein all follicles that
diameter deviation as a focused reference point to repre- were 5 mm or greater were ablated on day 10. When the
sent the first diameter manifestation of follicle selection. largest follicle of the postablation or primary wave reached
Another consideration in mares is the predeviation 15 mm, all follicles except the 2 largest were ablated. Re-
follicle that occurs in about one-third of primary waves in sults indicated that (1) the future DF emerged at 6 mm, an
mares (Fig. 4) and women [12]. The predeviation follicle average of 1 d earlier than the future SF, (2) the 2 follicles
attains a maximum diameter similar to the diameter of the grew approximately in parallel during the common growth
future DF at deviation but with regression beginning an phase when normalized to deviation and examined retro-
average of 1 d before the beginning of diameter deviation of spectively, (3) deviation occurred 6 d after emergence of a
the future DF. In retrospect, a predeviation follicle is 6-mm future DF, and (4) mean diameters of the future DF
122 O.J. Ginther / Domestic Animal Endocrinology 59 (2017) 116–133
and SF on the day of deviation were 23 and 20 mm,
respectively, equivalent to a 1-d difference in diameter.
The finding of a mean parallel diameter increase in
the 2 largest follicles when all follicles of the primary
wave were retained and normalized to deviation has
been substantiated in subsequent studies in mares of
various types and breeds [12,33,78]. In addition to the
planned studies of deviation, the study by hindsight of
previously published graphs from the initial report of
tracking the diameters of each of the 2 largest follicles in
mares (Thoroughbreds and Standardbreds) [28] indi-
cated that deviation occurred when the future DF was
about 22 to 28 mm in 7 of 11 graphs of individual pri-
mary waves. In the remaining waves, only the DF was
shown, or the second largest follicle was too small for
judging if it reflected inadequate development or a
response to a deviation mechanism. The largest follicle
at the beginning of deviation in the primary wave usu-
ally becomes the ovulatory follicle [83]. For example, the
first follicle to reach 20 mm became the ovulatory fol-
licle in 90% of 51 waves [8].
Although the mean diameter of the 2 largest follicles in
the initial deviation study [2] maintained the largest and
second largest positions, switching in diameter ranking
was reported to occur occasionally in individuals. The
specific diameter interrelationships among follicles during
the common growth phase and the capacity of individual
follicles for dominance have been described [45]. The 4
largest follicles of an induced primary wave (Section 5)
were ranked (largest ¼ rank 1, and so forth) according to
day of emergence at 6.0 to 6.9 mm. Rankings were also
made independently according to the diameter at the ex-
pected beginning of deviation (largest follicle first reached
greater than 20 mm; actual diameter, 22 mm). The follicles
grew in parallel on average until the end of the common
growth phase. However, the frequency of a follicle
becoming the largest at deviation when its rank at emer-
gence was 1, 2, 3, or 4 was 61%, 25%, 9%, or 5%, respectively.
That is, 39% of the first follicle that emerged at 6 mm did
not become the DF. When ranked at expected deviation and
studied retrospectively, the 2 largest follicles grew in par-
allel according to the means but did not maintain their
ranking or remain separated in diameter for 4 d before
deviation in 19 of 59 (33%) individual waves. That is, the
future DF emerged first and maintained its advantage
throughout the common growth phase in about two-thirds
of the waves, regardless whether ranking was in reference
to emergence or deviation. In the remaining one-third, the
diameter ranking among follicles was not constant; first
emergence did not establish that a follicle was destined to
become the DF. Intraovarian factors such as location of the
CL may account for some of the switching in ranks as has
been reported for cattle [3]. In this regard, a recent report in
mares indicated that the CL had a positive intraovarian
effect on preantral follicles [60].
When 1, 2, or 3 of the largest follicles were ablated at the
expected beginning of deviation, the largest remaining
follicle became the DF in 76% of primary waves [45]. In
many of the remaining waves, the SFs seemed undersized
and may have regressed from underdevelopment rather
than from the deviation process. These results supported
the hypothesis that the capacity for dominance at the ex-
pected beginning of deviation is shared by many follicles of
the wave, and the potential is greatest for the largest
(highest ranking) follicle and decreases progressively ac-
cording to decreasing rank. Furthermore, the second largest
follicle retains its dominance capability in most waves for 1
or 2 d after the expected beginning of deviation. The
capability of SF to become a DF if ablation of DF is done 1 or
2 d after deviation implies that the SF remains viable or
dormant for 1 or 2 d after being subjected to the deviation
process. Based on switching in diameter rankings during
the common -growth phase in about one-third of waves
and the capacity of several follicles for dominance, it is
proposed that the deviation process represents the entire
mechanism of follicle selection in mares as recently pro-
posed for heifers [3].
Deviation is manifested by differential changes in
growth rate (diameter) of the future DF and SF. However,
the deviation process begins about a day before diameter
deviation based on (1) the beginning of structural changes
in the follicle wall and increased blood flow in the future DF
1 d before diameter deviation (Section 8), (2) circulatory
concentrations of estradiol and LH (Sections 10 and 11) and
intrafollicular concentrations of IGF1, estradiol, inhibin-A,
and VEGF begin to increase in the future DF during the
day before diameter deviation (Sections 12–15), and (3)
higher thecal content of LH receptors apparently identifies
the presumptive future ovulatory vs nonovulatory follicle
on day 14 or 1 or 2 d before expected estrus or before
identification of the preovulatory follicle based on com-
parison of diameters [27].
The predeviation and deviation events occur during the
declining portion of the FSH surge (Section 9). Apparently,
the future SF has not reached a similar stage of develop-
ment and has inadequate intrafollicular sensitizers for an
enhanced response to low FSH. This differential response
between the future DF and SF to the low levels of FSH is
considered to be the essence of follicle deviation in mares
as reported for heifers [3]. The coupling between changes
in FSH concentration and changes in follicle dimension may
be too close in time to allow the SF to develop the condi-
tions needed for continued growth rate. Detectable
coupling involves only 2 or 3 h in heifers [3], but this has
not been determined in mares. An alternate but unlikely
possibility is that the receptors of the developing DF
remove much of the available FSH so that inadequate
amount of FSH remains for the SF.
Another consideration that has been mostly neglected is
the choice of a characteristic to represent follicle dimen-
sion. Diameter is used traditionally and is easily compre-
hended. However, the follicles are approximately spherical,
and the layers of granulosa and theca cells are at the
periphery of the follicle. That is, surface area is a better
representation of the functional portion of a follicle. Simi-
larly, volume is a better representation of a solid gland (eg,
CL). For example, surface area of a follicle increases expo-
nentially or curvilinearly and is associated with a curvi-
linear increase in circulating estradiol [84]. Two follicles of
different diameters may grow in parallel when based on
diameter. However, the curvilinear increase of the larger
follicle over time is increasingly greater than that for the
 O.J. Ginther / Domestic Animal Endocrinology 59 (2017) 116–133
smaller follicle. That is, the larger follicle may gain a
dimensional and functional advantage over the smaller
follicle as deviation approaches [84].
8. Changes in wall of the future DF before diameter
deviation
The temporal relationships between echotextural and
diameter changes in the 2 largest follicles in association
with deviation in the primary wave have been studied
using thickness of granulosa, echogenicity of granulosa,
and prominence of an anechoic band beneath the gran-
ulosa [33]. An anechoic area indicates the relative absence
or scarcity of reflecting tissue interfaces such as in areas
with a high proportion of fluids as in the lumen of blood
vessels and the antrum of follicles [85]. The first day of an
overall echotextural difference between the 2 largest folli-
cles (echotextural deviation) was about 1 d earlier than the
beginning of diameter deviation. Increasing echotextural
values for the anechoic layer were more attributable to
differences in status between a future DF and SF, whereas
the increasing values for granulosa (thickness and echo-
genicity) were more attributable to follicle growth inde-
pendent of future status. That is, the extent of development
of an anechoic band beneath the granulosa distinguishes
the future DF from the future SF before the beginning of
diameter deviation.
In a color Doppler study, the blood-flow area (cm2)
within the follicle wall began to differentially increase in
the future DF, compared to the future SF, in an equivalent of
1 d (average of 3.0-mm difference in diameter) before the
beginning of diameter deviation [47] similar to the increase
in score for the anechoic layer [33]. These studies have
indicated that deviation in structural and vascular perfu-
sion of the 2 follicles precedes diameter deviation by 1 d.
That is, the deviation process begins a day before the
manifestation of diameter deviation during follicle selec-
tion in mares. The differential changes in echotexture and
vascularity of the follicle wall can be attributed to
increasing exposure of the follicle wall and vasculature of
the future DF to estrogens (Section 13) or VEGF (Section 15).
An increase in the difference between the 2 follicles in
follicular-fluid estradiol concentrations begins about 1 d
before the beginning of diameter deviation [34] or on about
the day of the echotextural and vascular changes [33,47].
9. Relationships between an FSH surge and follicle
events
Circulating FSH is required for the stimulation of follicles
in mammals [86] including horses [87]. During the decade of
the 1970s, the technology and experimentation on circu-
lating concentrations and roles of FSH and LH in follicle
dynamics were extended to mares. Two decades later, the
acquired technology and knowledge were applied to
unraveling the mechanism for the systemic component of
the deviation process. The detrimental effects of reducing
the concentrations of gonadotropins on equine follicles and
CL were first shown in 1972 by treatments with an anti-
serum against an equine pituitary extract [18,88,89] and
more specifically in 1979 by reducing circulating FSH with a
123
steroid-free equine follicular fluid [23,90]. That follicle
growth can be stimulated and that ovulation and multiple
ovulations can be induced with gonadotropin treatments
(equine pituitary extract) in mares were first reported in
1974 [20] and has been confirmed [91]. The concomitant
suppression of both FSH and follicle activity by a steroid-free
equine follicular fluid was also reported for both nonpreg-
nant [92] and pregnant [90] mares. More recently, a GnRH
antagonist that competitively inhibits GnRH action at
pituitary receptor sites [56] and an equine recombinant FSH
[57] have been used for more detailed study on the positive
effect of FSH on follicles and the negative systemic effect of
factors from the follicles on FSH before, during, and after
deviation. A radioimmunoassay for circulating concentra-
tions of FSH in mares was first reported in 1975 [21]; 2 surges
of FSH during an IOI were described. The apparent surges
occurred in early and late diestrus corresponding to the days
when an antiserum against an equine pituitary extract was
reported previously to suppress follicle development [18].
Also, subsequent findings showed that a minor or secondary
wave occurred during early diestrus followed by beginning
of a primary wave during late diestrus (Section 5).
The temporal association between emergence of a
major or minor follicular wave and an FSH surge in mares
was first reported in 1992 [31,32] and was subsequently
confirmed [93]. A significant increase and decrease in FSH
encompassed emergence of each wave and reached a peak
when the follicles were 12 to 15 mm, similar to results in
cattle for wave emergence when follicles were 4 or 5 mm
[94]. In mares as well as in cattle, the declining portion of
the FSH surge was accompanied by divergence in diameter
between the future DF and SF [32]. In another study [95], a
significant increase in diameter of the 6 largest follicles to a
mean of 12 mm was followed by a significant decrease
representing emergence of the primary wave. Emergence
was temporally associated with the greatest mean FSH
concentration of the estrous cycle. The divergence in these
studies was taken as a manifestation of follicle selection.
However, the deviation phenomenon was not yet
described, and it was not known if the decline in FSH was a
component of the selection process. In this regard, FSH-rich
equine pituitary extracts can override divergence in mares
indicating that the decline in FSH plays a role in the se-
lection process [66,96,97].
The extension of the concept of deviation from heifers to
mares in 1997 [2] became an improved reference and
normalization point for study of the temporal relationships
between FSH and diameter dissociation between the 2
largest follicles. The FSH surge temporally associated with
an induced primary wave begins to increase after the first
examination when the future DF is about 6.0 mm, peaks 3 d
later when the future DF is about 13.0 mm, declines for 3 d
from the FSH peak to deviation when the follicle is
22.5 mm, and continues to decline to minimal levels until a
few days after deviation (Section 9; Fig. 5). The FSH decline
occurs while progesterone is decreasing [2,12]. The tem-
poral relationship between FSH and the follicles when
normalized to deviation in mares has been confirmed [38].
Reducing circulating FSH by treatment with a GnRH
antagonist indicated that follicle growth before 11 mm was
independent of FSH but was essential for maintaining
124 O.J. Ginther / Domestic Animal Endocrinology 59 (2017) 116–133

Circulatory hormones equine follicular fluid suppressed FSH in ovariectomized

14 mares [23]. Circulating concentration of immunoreactive
(36 mm)
(ir) inhibin in mares throughout the IOI was first reported
FSH in 1991 [30]. An inverse relationship in peripheral con-
12
centrations of FSH and ir-inhibin was found. Concentra-
Concentration FSH, ir-inhibin, LH (ng/mL); estradiol (pg/mL)

Largest tions of ir-inhibin [30,37,38] and apparently inhibin-A [98]

10 follicle begin to increase before the declining portion of the wave-
stimulating FSH surge. Concentration of circulating ir-
8 Estradiol inhibin increased when FSH was increasing, and follicles
(22 mm) were growing from 8 to 12 mm. This result was attributed
6 to reports that some forms of inhibin (a-subunits) that
would have been measured by the assay do not have FSH-
4 (12 mm) suppressing activity [99]. The ir-inhibin continues to in-
crease during the decline in the FSH surge apparently from
(7 mm)
2 forms that do suppress FSH (Fig. 5) [37,38].
A functional relationship between FSH and inhibin was
shown by the finding that an inhibin antiserum in mares
40 during diestrus increases the concentrations of FSH [29].
ir-Inhibin The role of follicular inhibin in the decline in FSH between
the peak of the surge (largest follicle, 13 mm) and diameter
35 deviation (largest follicle 22 or 23 mm) was shown by
ablation of follicles when they were 15 mm so that all fol-
30 licles, 3 largest, or only the largest were retained [37]. Re-
sults indicated that the secretion of inhibin and the decline
25 LH in FSH between the peak of the surge and follicle deviation
were a function of multiple follicles that are greater than
20 13 mm. One follicle was less effective than 3 in secreting
Diameter inhibin and suppressing FSH, and 3 follicles were as effective
deviation as all follicles. The continued FSH decline after the beginning
15
Emergence of deviation was a function of the developing DF. Ablation of
all follicles of 6 mm or greater on day 12.5 or before expected
deviation induced an immediate increase in FSH but not
–7 –6 –5 –4 –3 –2 –1 0 1 2 3 4 5 6 when ablation was done on day 0.5 or 6.5 [100].
Days from diameter deviation In ovariectomized mares, each of estradiol and a steroid-
free fraction of follicular fluid suppress FSH concentrations,
Fig. 5. Mean  standard error of the mean concentrations of circulating but a combination of estradiol and the follicular-fluid
hormones and diameter of the largest follicle in 15 mares. Data are
fraction have a greater or synergistic effect on FSH sup-
normalized to expected beginning of diameter deviation (largest follicle,
22 mm). Concentration of FSH continued to decrease after the beginning of
pression than either alone [23,26]. In intact mares, an
the deviation, and estradiol began to increase before deviation. Inhibin estradiol increase was not detected during the first 2 d of
began to increase on the day before FSH began to decrease, and LH increased the FSH decline (Fig. 5) [35,38]. However, circulating
and formed a mean plateau encompassing deviation. The plateau represents estradiol [34,38] and estradiol in the follicular fluid of the
a decrease in LH on different days in individuals. The increase in LH occurs
largest follicle [34,39] increase between the day before and
during a decrease in progesterone and is depicted elsewhere [2]. Adapted
from [38]. the day of diameter deviation. These temporal relationships
between the FSH peak and deviation indicate that inhibin is
the sole FSH suppressor for the first 2 d of the decline in the
follicle viability including during the FSH decline [56]. The surge, and estradiol begins to contribute to the decline
suggestion that FSH was not needed early in the wave is during the third day or about 1 d before diameter deviation.
suspected in that the antagonist reduced FSH concentra- Both inhibin and estradiol continue to be secreted by the
tions only by 50%. The reduction was greater, however, than developing DF for a few days after deviation, and both
when steroid-free follicular fluid (inhibin) or estradiol was contribute to the continuing FSH decline [38].
used. On a temporal basis, there is an association between
an increase in number of 2- and 3-mm follicles and an in- 11. LH and the deviation mechanism
crease in FSH concentrations [95]. The diameter of follicles
in the primary wave at the time of the first increase in the Radioimmunoassay of circulating LH concentrations
associated FSH surge needs clarification. during the equine estrous cycle was reported in 1973 [19].
The ovulatory LH surge begins on about day 14 or 15
10. Regulation and role of the FSH surge (before diameter deviation) and reaches maximum the day
after ovulation. There was considerable delay in publication
A role for an FSH-suppressing form of inhibin of follicle of the initial report on the protracted LH surge in mares
origin in the decline in FSH was first suspected from the owing to reluctance of reviewers to accept that the LH
report that a steroid-free or proteinaceous fraction of surge in this species departed drastically from the rapid
 O.J. Ginther / Domestic Animal Endocrinology 59 (2017) 116–133
preovulatory surge in other species [1]. The LH increase
before deviation has been well documented by studies that
used the beginning of deviation as the reference point
[2,38]. On average, LH reaches a plateau before deviation
that extends for a few days after deviation and then in-
creases rapidly until the day after ovulation (Fig. 5) [1,38].
However, in most individuals (eg, 80%), the LH that en-
compasses deviation is transient as represented by an in-
crease before deviation and a decrease after deviation
followed by the major portion of the ovulatory LH surge
[38]. The mean LH plateau is misleading and results from
the postdeviation decrease occurring on different days in
individuals.
The presence of LH receptors in the future DF and SF in
association with deviation has not been specifically studied
in mares. The LH receptor protein content in granulosa cells
is greater for follicles that are 15 to 19 mm than that for
smaller follicles [101]. An earlier report indicated that the
thecal content of LH receptors identified the presumptive
ovulatory follicle on day 14 [27]. Day 14 was 1 or 2 d before
expected estrus or before the future ovulatory follicle was
identifiable by greater diameter. Specific study on the dif-
ferential presence of LH receptors between the future DF
and SF in association with deviation is needed.
The role of LH in the deviation mechanism has been
studied by reducing LH concentration with various doses of
progesterone [36]. A dose that decreased LH 2 d before
deviation in controls did not delay the onset of diameter
deviation based on no difference from the controls in
diameter of the largest follicle or in day of deviation.
However, maximum diameter of the DF was reduced
beginning 1 d after deviation. Follicles grew at the same
rate as in controls until the day of deviation. These results
indicated that the LH increase that encompasses deviation
was not needed for the deviation process in this species but
was needed for full growth of the DF after deviation. A
project with shortening of the luteal phase with prosta-
glandin F2a or lengthening the phase with exogenous
progesterone led to the same conclusion [38]. In a more
recent study, manipulation of LH and FSH concentrations
with a GnRH antagonist and replacement with an equine
recombinant LH or equine recombinant FSH indicated that
FSH but not LH was needed for growth of the DF until 48 h
after the beginning of deviation [57]. However, an experi-
ment with exogenous progesterone indicated that the LH
elevation played a role in the production of estradiol and
inhibin [38]. That is, the only role of LH in the deviation
process that has been demonstrated in mares is the pro-
duction of estradiol and inhibin, which in turn are needed
for the predeviation and continuing postdeviation sup-
pression in FSH (Section 10). However, this role of LH may
not be essential; experimental depression of LH did not
delay deviation [36]. The systemic component of the devi-
ation process is represented in the developing DF by sup-
pression of circulating FSH from stimulation of inhibin by
LH and IGF1 (Section 12) and stimulation of estradiol by LH.
Estradiol has a negative effect on LH [43] throughout the
ascending portion of the ovulatory LH surge including on
the day of the beginning of expected deviation (Fig. 5)
based on treatments with physiologic doses of estradiol
[52]. The endogenous estradiol is from the future and
125
established DF as indicated by ablation of follicles
throughout the ascending portion of the surge including
before deviation [51]. The paradox of both a negative and
positive effect of the increasing estradiol on the increasing
LH including during deviation is attributable to a damp-
ening (ie, negative effect) on the positive seasonal effect
from increased photoperiod [1,22]. Thereby, LH is reduced
to a level that is compatible with a positive effect on the
follicles. A similar balance and counterbalance between
ovarian control and photoperiod control also is involved in
regulating an effective concentration of FSH [22,44].
12. The IGF system
The IGF system has received considerable attention as
an initiator of deviation. The ability of IGF1 to increase the
responsiveness of follicle cells to FSH has been shown by
tissue-culture studies in nonequine species [102,103]. The
complexities of the IGF system including forms of IGF and
IGF-binding proteins have been reviewed [104,105]. Several
types of IGF1-binding proteins are produced by equine
granulosa cells in tissue culture when stimulated by FSH
[40]. Concentrations of IGF binding protein-2 in mares
begin to increase at the beginning of deviation in the future
SF but not in the future DF [39]. Other studies have shown
that several types of IGF-binding proteins are higher in
concentration in SF than in DF [41,106].
The functional relationships between IGF1 and binding
proteins during the deviation process are mediated by
specific proteases found in follicular fluid. The proteases
degrade the binding proteins and thus increase the
bioavailability of free IGF1 in a specific follicle. Based on
studies in other species, the protease action for releasing
free IGF1 from its binding protein may occur at least partly
in the follicular fluid, and the free IGF1 interacts with re-
ceptor sites at the cell membranes (review [107]). Proteo-
lytic activity for several binding proteins has been reported
in the DF during the follicular phase in mares [41,106].
Pregnancy-associated plasma protein-A (PAPP-A) is a pro-
tease involved in proteolysis in follicular fluid by the release
of IGF1 from its binding protein [108,109]. In mares, PAPP-A
was injected into the future SF at the beginning of devia-
tion, and follicular fluid was sampled (200 mL) the next day
[50]. The PAPP-A increased the concentration of IGF1, and
the effects on other intrafollicular factors were attributed to
the release of the free IGF1. The induction of proteolytic
activity in the IGF system is the earliest reported identifi-
able intrafollicular event in selection of a DF in horses.
Concentration of IGF1 in the intrafollicular fluid of the
future DF increases before the beginning of diameter de-
viation (Fig. 6) [39]. Diameter deviation began when the
largest follicle was in a range of 20.0 to 23.9 mm. A differ-
ential IGF1 increase began earlier at 16.0 to 19.9 mm or
before diameter deviation. After ablation of the future DF at
expected deviation, IGF1 began to increase in SF 12 h before
diameter deviation between SF and the next largest SF
(Fig. 7) [42]. On a temporal basis, these 2 studies have
indicated that a role of the IGF system in mares is the
initiation or activation of the deviation process.
The roles of intrafollicular IGF1 have been shown on a
functional basis. Intrafollicular treatment of SF with a
126 O.J. Ginther / Domestic Animal Endocrinology 59 (2017) 116–133

Intrafollicular factors Experimental deviation

4000 Estradiol Follicles

Diameter (mm)
3000 25 New DF
Future
dominant Ablated DF
2000 follicle
SF1 New SF1
20
1000 SF2
Largest
subordinate
0 15

Free IGF1 15 Free IGF-1

Follicular-fluid concentration (ng/mL)

25
20
10
15
10
5
5
0 0

Follicular-fluid concentration (ng/mL)

200 Inhibin-A
Estradiol
150 1500
100
1000
50

500
Activin-A
300
Inhibin-A
200 200

100 150
Diameter deviation

12.0-15.9 16.0-19.9 20.0-23.9 24.0-27.9 100

Diameter (mm) range for largest follicle
Fig. 6. Mean  standard error of the mean concentrations of follicular-fluid
factors from the beginning of diameter deviation, which occurred in the
20.0- to 23.9-mm range of the future dominant follicle. Differential changes
between follicles in concentrations of estradiol, IGF1, inhibin-A, and activin-
A began to change significantly in 1 diameter range before the beginning of
diameter deviation. Adapted from [39].
pharmacologic dose of IGF1 stimulated SF to become a DF
and to ovulate in 24 of 31 (77%) mares compared to 24%
saline-injected controls combined for 2 reports [48,83]. The
injection of IGF1 into 8 of 12 SFs when they reached 20 mm
stimulated dominance in the follicle even when 20 mm was
reached 1 to 3 d after the beginning of deviation [83]. This
result is consistent with the maintenance of dominant ca-
pacity by SF for a few days after the beginning of deviation
(Section 7). The long-term effects of a single pharmacologic
dose did not occur with a single physiologic dose and is
attributable to prolonged presence vs transitory presence
of IGF1 for the high vs low doses. A physiologic dose of IGF1
was predetermined by an amount that resulted in an
Induced
50 deviation
0 12 24 48 60
Hours after ablation of DF at the
expected beginning of deviation
Fig. 7. Mean  standard error of the mean follicle diameters and follicular-
fluid concentrations of free IGF1, estradiol, and inhibin-A during experi-
mental deviation in 48 mares. The dominant follicle (DF) was ablated at
expected diameter deviation (hour 0). The largest subordinate follicle (SF1)
converted to a new DF with a new diameter deviation at hour 24. Concen-
tration of IGF1 began to deviate before diameter deviation (hour 12), whereas
estradiol and inhibin-A began to deviate after diameter deviation (hour 48).
The arrows indicate the beginning of significance between SF1 and SF2 in the
differences in values between sequential examinations. Adapted from [42].
immediate exogenous concentration in the treated future
SF that was similar to the endogenous concentration in the
future DF at the beginning of deviation [48]. At the first
sampling 24 h after injection of a physiologic dose into the
future SF, there was an increase in inhibin-A, activin-A, and
VEGF but not an increase in estradiol. In addition, an
 O.J. Ginther / Domestic Animal Endocrinology 59 (2017) 116–133
increase in IGF-binding protein and androstenedione was
prevented. These results demonstrated that endogenous
IGF1 is involved in the regulation of the production of
inhibin-A and activin-A but not estradiol in the future DF. In
contrast to the noneffect of IGF1 on estradiol in mares,
intrafollicular [70] or intrastromal [104] treatment with
IGF1 in cattle immediately increases the concentration of
estradiol in follicular fluid.
The role of the IGF system in follicle selection has also
been studied in mares by injecting an IGF-binding protein
into the largest follicle at the expected beginning of
deviation and sampling follicular fluid 24 h later [49].
In many of the waves, injection of the binding protein into
the DF stopped the follicle from growing followed by
regression, and the second largest follicle became the DF.
In addition, injection of an IGF-binding protein into the
future DF decreased the follicular-fluid concentrations of
free IGF1, estradiol, inhibin-A, and activin-A and increased
androstenedione concentrations at the first sampling
in 24 h.
The large follicles of mares have favored extensive
studies of the IGF system by intrafollicular treatment with
IGF, PAPP-A, or an IGF-binding protein followed by later
sampling of the treated follicles (Section 4). As a result, the
role of the IGF system in follicle deviation has been more
thoroughly studied in mares by challenge experimentation
than that in any other species. It is concluded that the IGF1
system plays an initiating and essential intrafollicular role
in selection and development of a DF in mares [6,49].
13. Role of estradiol
Estradiol in mares begins to increase in the plasma
(Fig. 5) [34] and in the antral fluid of the future DF (Fig. 6)
[39] at the beginning of the deviation process when the
future DF reaches about 20 mm or about 1 d before the
beginning of diameter deviation. This is about the time of a
negative systemic effect of estradiol on FSH concentration
(Section 10). The estradiol increase in mares is primarily a
function of the aromatase system from a positive effect on
the transient increase in LH that encompasses deviation
(Section 11). In heifers, intrafollicular estradiol is involved
in the deviation process as an FSH sensitizer; systemic
treatment of heifers with an antiserum against estradiol
delays the beginning of deviation despite an increase in
FSH [110]. A similar study has not been done in mares.
An injection of a pharmacologic [70] or physiologic [50]
dose of IGF1 into SF at the beginning of expected deviation
did not increase the concentration of estradiol during 24 h
post-treatment. The lack of an intrafollicular role for
estradiol in deviation in mares is also consistent with the
results of experimental conversion of SF to a DF (Fig. 7) [42].
After ablation of the future DF at expected deviation, the
experimental diameter deviation during conversion of SF to
a new DF began in 24 h but estradiol in the converted DF
did not begin to increase until 48 h. That is, the induced
deviation in diameter began well before deviation in
follicular-fluid concentration of estradiol. Despite the role
of estradiol in the systemic component of the deviation
process in mares and in both the systemic and intra-
follicular components in heifers [3], available reports
127
indicate that estradiol is not involved as an FSH sensitizer in
the intrafollicular component in mares.
Although estradiol has a systemic role by contributing to
suppression of FSH but not a local direct role in the devi-
ation process in mares, it likely has a local role in growth of
follicles. That is, the apparent growth effect is similar be-
tween a future DF and SF of similar diameter independent
of a difference in the day that each follicle attains a given
diameter. A growth function in mares is consistent with the
reported increase in granulosa cell content of aromatase
[111]. However, the nature of the effect of estradiol on the
cells of the follicle wall must be assumed on a comparative
basis from studies in nonequine species. In other farm and
laboratory species, estradiol promotes development of
preantral follicles and stimulates steroidogenesis in gran-
ulosa and theca cells in vitro [112]. In pigs and sheep,
estradiol promotes the synthesis of IGF1 [104]. In rats,
estradiol has the capacity in vitro to augment its own
synthesis by upregulating the thecal androgens [113].
Androgens and progestins are substrates for estradiol
synthesis. Androgens are produced in the thecal layer and
then aromatized in the granulosa layer in mares, as in
other species [114]. The synthesis of androgen in the mare,
but not in cattle, involves progesterone as an intermediary
steroid. The absence of an estradiol increase in the future
SF as deviation approaches is consistent with reduced
utilization of the androgen substrate as indicated by the
increasing androstenedione concentrations in SFs [39].
Progesterone concentration increases in the DF at the
beginning of deviation [39] although deviation usually
occurs during luteolysis [2,38]. The comparative (horses
vs cattle) relationships of androgens and progesterone
to the beginning of deviation and the high androgen
concentration in the SF in horses may be a productive
research area.
14. Roles of inhibin-A and activin-A
Inhibin-A and activin-A as well as free IGF1 and estra-
diol begin to increase in the future DF but not in the future
SF before the beginning of deviation (Fig. 6) [39]. Intra-
follicular injection of a pharmacologic or physiologic dose
of IGF1 stimulates an increase in inhibin-A and activin-A at
the first post-treatment examination at 24 h [48]. More-
over, an intrafollicular injection of PAPP-A into the future SF
at the beginning of deviation simultaneously increases the
concentrations of free IGF1, inhibin-A, and follistatin [50].
Inhibin-A and activin-A concentrations decrease in the
follicle that was expected to be the DF 24 h after an injec-
tion of an IGF-binding protein [49]. With the physiologic
dose of IGF1, concentrations of free IGF1, inhibin-A, and
activin-A in the SF 24 h after treatment were similar to the
concentrations of these factors in the future DF at the time
of treatment of SF. These results indicate that the concen-
trations of these factors in the future DF at expected devi-
ation are functions of endogenous IGF1 beginning before
diameter deviation. The increase in inhibin-A as well as
estradiol before spontaneous deviation is also a conse-
quence of the transient increase in LH and is consistent
with the apparent development of LH receptors before
deviation (Section 11).
128 O.J. Ginther / Domestic Animal Endocrinology 59 (2017) 116–133

When the future DF is ablated at the expected beginning Double deviation

of diameter deviation the following changes occur during Follicles
conversion of the SF into a new DF: (1) free IGF1 in the SF
35
begins to increase in 12 h (Section 12), (2) the experimental 1 deviation
or induced diameter deviation begins in 24 h, and (3) 1 dominant follicle
inhibin-A and estradiol begin to increase in 24 h and 48 h, 30 (n = 21)
respectively, after the induced deviation (Fig. 7) [42]. The F1
concentrations of circulating FSH were not different be-
tween the expected beginning of spontaneous deviation 25
and the beginning of experimental deviation. These results * F2
indicated that deviation in intrafollicular concentrations of

Diameter (mm)
F3
inhibin-A and activin-A as well as estradiol did not occur 20
before the experimental deviation and therefore appar-
ently were not part of the intrafollicular component of the
deviation process. As a caveat, some reservation is war-
Follicles
ranted in comparison of the mechanisms of experimental
35
vs spontaneous deviation. It is concluded that intra-
2 deviations
follicular inhibin-A and activin-A increase from IGF1 2 dominant follicles
stimulation but apparently (more study needed) do not
have intrafollicular roles in the initiation of deviation. An
30 (n = 5) F1
*
earlier review [4] interpreted on a temporal basis that 25 F2
estradiol, inhibin-A, and activin-A contribute to deviation
by sensitizing the follicle cells to increase their respon-
* F3
siveness to FSH. However, a more recent review [6] and the 20
present review have concluded that IGF1 is the only one of
the 4 intrafollicular factors that have been shown to play a 15
direct role in the deviation process in mares apparently by
18 FSH
increasing the sensitivity of the follicle cells to FSH during
the low FSH. 16
Concentration (ng/mL)

The likelihood that inhibin-A and activin-A are used for

follicle growth in mares is based on the results of experi- 14
ments in other species [113,115]. The potential for local
12
* 1 dominant
regulatory functions of inhibins, activins, and follistatin in 2 dominant follicle
the ovary have been reviewed for various species [116], but 10 follicles
specific studies apparently are not available for mares.
Intrafollicular inhibin enhances LH-induced androgen 8
production in the thecal cells of cattle [113]. In vitro studies
6
in rats and cattle have indicated that activin-A induces the
proliferation and expression of FSH receptors, increases 4
granulosa cell steroidogenesis and proliferation, and in-
–4 –3 –2 –1 0 1 2 3 4
creases aromatase activity and estradiol production (re-
view [116]). Follistatin is also present in follicular fluid of Days from beginning of first observed deviation
mares [50]. In cattle, follistatin apparently acts as a binding
Fig. 8. Mean  standard error of the mean diameters of F1 (largest follicle),
protein for activin [113], but a follistatin role in mares F2, and F3 for ovulatory waves with 1 deviation and 1 dominant follicle
apparently has not been studied. (upper panel), 2 deviations and 2 dominant follicles (middle panel), and
Excised ovaries of mares have been used to identify associated concentrations of FSH (lower panel). The asterisk indicates the
genes coexpressed with anti-Müllerian hormone (AMH) in day of the first greater difference (P < 0.05) between means. For the follicles,
an asterisk indicates a greater difference between means than for the day
granulosa cells of growing follicles (largest follicle, 15– before, whereas for FSH, the difference is greater than for 2 d before.
20 mm) before expected follicle deviation [59]. The Adapted from [55].
expression of AMH and genes coexpressed with AMH in the
cells and concentrations of AMH in the follicular fluid
decreased in the growing follicles during predeviation. 15. Role of VEGF
Intrafollicular estradiol increased concomitantly. These
changes did not occur in an established DF of 35 mm. The The intraovarian factor VEGF is angiogenic and is higher
reported pattern of circulating FSH during follicle emer- in the DF than that in the SF 1 d after the beginning of
gence and deviation (Section 9) is similar to the observed deviation in mares [48], but earlier temporal relationships
expression of AMH and FSH receptors in the granulosa cells. apparently have not been reported. In mares, VEGF in-
This temporal association tentatively indicates that FSH has creases in the future SF when first examined 24 h after
a stimulatory effect on the expression of AMH accounting injection of IGF1 at the beginning of expected deviation
for the decrease in AMH when FSH decreases. A role for a [48]. When an IGF-binding protein is injected into the
decrease in AMH in follicle deviation has not been shown. future DF, concentration of VEGF decreases [49]. The
 O.J. Ginther / Domestic Animal Endocrinology 59 (2017) 116–133 129

Follicle selection in mares

Positive effect Gonadotropins
Negative effect Follicle diameter Systemic component
Intrafollicular component of selection (via vascular system)
LH
Pituitary
FSH

sion
up pres
Hs
FS wth
Gro
FS
mm H
R
31 Ih Ih Ih Ih Activin-A
Inhibin-A
E2 E2 Estradiol
28
DF

Se
25 IGF

nsi
system

tize
IGF R LH

r
22 Ih
E2 IGFBP Proteolysed

Incre
IGF1 IGFBP
19
IGFBP PAPP-A

ase
(protease)

d
16 VEGF

blo
e SF

od
as
ph

f
13

low
th
r ow Beginning of
IGF1
g
10 on- deviation process
m
m
Co Beginning of Intrafollicular component
7 diameter deviation (IGF1 sensitizes cells to FSH)

–4 –3 –2 –1 0 1 2 3 E2 Estradiol
Days from diameter deviation IGF1 Insulin-like growth factor 1
IGFBP IGF binding protein
Ih Inhibin
12 13 14 15 16 17 18 19 R Receptor for FSH, LH
Days from ovulation VEGF Vascular endothelial growth factor

Fig. 9. The postulated interactions of the systemic and intrafollicular components of the deviation mechanism between the dominant follicle (DF) and largest
subordinate follicle (SF) in mares. The depicted follicular events occur primarily in the cells of the theca and granulosa of the future DF during the deviation
process that begins during the day before diameter deviation; however, protease activity to free IGF1 from its binding protein may occur at least partly within the
follicular fluid. The solid and broken arrows represent positive and negative effects, respectively, as demonstrated in mares. The dotted arrows are positive effects
on follicle growth based on studies in other species. The follicle wall is shown as a whole without regard to cellular layers. Inhibin-A and estradiol are systemic
suppressors of FSH and account for the low FSH during the deviation process. The IGF system is shown as the initiator of deviation, wherein IGF1 is freed from its
binding protein by a protease and increases the responsiveness of the future DF to the declining portion of the wave-stimulating FSH surge. Activin-A, inhibin-A,
and estradiol have roles in follicle growth (dotted arrows) based on studies in other species. That is, deviation is associated with an increase in free IGF1 in DF and
an increase in bound IGF1 (IGFBP) in SF. A more detailed description of the depicted mechanisms is given in the conclusions (Section 17), and even more detailed
literature sources are given in Sections 10–15.

intrafollicular injection of VEGF into the future SF at the
beginning of deviation increases the concentrations of free
IGF1 and decreases the concentrations of androstenedione
but does not affect the concentrations of other factors [48].
The positive effect of VEGF on IGF1 may be secondary to an
increase in blood flow of the follicle wall but this has not
been shown (Section 8). The concomitant temporal
relationship among increased blood flow, IGF1, and VEGF
beginning about a day before diameter deviation and the
apparent functional relationship between IGF1 and VEGF
are noteworthy, but needs more study.
The angiogenesis, vascular permeability, and stimula-
tory effects on endothelial cells are functions of VEGF based
on extensive studies in other species (reviews [117,118]). In
130 O.J. Ginther / Domestic Animal Endocrinology 59 (2017) 116–133
addition, IGF1 stimulates VEGF production in cultures of
granulosa cells of other species [118,119]. In mares, an
expansion of an anechoic band and an increase in blood
flow occur within the wall of the future DF beginning a day
before deviation (Section 8). Although information on VEGF
concentration at this time apparently is not available in
mares, VEGF is a likely candidate for a role in developing
follicle vascularity differentially in the future DF before the
beginning of diameter deviation (Section 8). A continued
increase in vascularity of the future DF is a likely advantage
for the developing DF in receiving a preferential supply of
growth factors, gonadotropins, steroid precursors, and
nutrients during the deviation process.
16. Double deviation
Two deviations at separate times during the ovulatory
wave result in 2 DF (greater than 28 mm). Double deviation
has been well described in cattle (review [3]) but is the
subject of apparently only 1 report in mares [55]. In the
study in mares, 1 DF developed in 21 of 32 waves (66%) and
2 DF in 11 waves (34%); 8 of 11 waves with 2 DF had 2
distinctly separate deviations (Fig. 8). The incidences of 2
deviations (deviation 1 and deviation 2) were 8 of 32 (25%),
but the incidence of double ovulation was only 3%. Follicles
at deviation 1 were defined as F1 (largest future DF), F2
(second largest future DF), and F3 (largest future subordi-
nate). Deviation 1 was represented by a combination of F1
and F2 vs F3 and deviation 2 by F1 vs F2 (Fig. 8). The mean
interval between deviations 1 and 2 was 2.5 d. The only
detected differences were that waves with 2 deviations vs 1
deviation had greater diameter of F1, tendency for a greater
diameter of F2, and lesser FSH concentration during both
deviations. The continued growth of F2 until deviation 2
suggests partial production of the sensitizer for FSH
responsiveness in F2 at deviation 1 in association with
incomplete development of F2 at the time of deviation of
F1. However, the concentrations of follicular fluid factors
especially IGF1 have not been compared between waves
with 1 vs 2 deviations.
17. Conclusions
Conclusions on the systemic and intrafollicular compo-
nents of follicle selection as represented by the deviation
process in mares can be best assimilated when viewing
Figure 9. The systemic component of diameter deviation
during the primary or ovulatory follicular wave includes a
decline in the wave stimulating FSH surge. Diameter devia-
tion begins on day 15 or 16 when the future DF is 22 or
23 mm and is characterized by the beginning of continued
growth rate of the DF and reduced growth rate of the SF
followed by SF regression. The deviation process begins in
the future DF when it is about 20 mm or during the day
before diameter deviation. Inhibin is produced by multiple
growing follicles of the wave when each follicle attains 12 or
13 mm. Inhibin is a suppressor of FSH throughout the decline
in the FSH surge. Estradiol is stimulated by a transient cir-
culatory LH increase but not by intrafollicular IGF in mares
and systemically contributes to the FSH decline beginning
the day before diameter deviation. Estradiol also has a
negative effect on LH thereby dampening the positive effect
of photoperiod on LH. The critical intrafollicular initiator of
the deviation process in mares is IGF1, which is released from
a binding protein by a protease at the beginning of the
deviation process during the day before the beginning of
diameter deviation. The protease action in releasing free
IGF1 from it binding protein apparently may occur at least
partly in the follicular fluid, and the free IGF1 attaches to its
receptors at the cell membranes. The IGF1 apparently
increases the responsiveness of the future DF to the positive
cellular effects of FSH during the decline in the FSH surge.
The IGF1 remains bound to its protein in the future SFs; and
therefore, the subordinates are not able to use the low FSH.
The IGF1 in the future DF also stimulates the production of
VEGF, which apparently increases the vascular perfusion of
the future DF during the deviation process. There has been
no indication that estradiol, inhibin-A, and activin-A are a
direct part of the intrafollicular component of the deviation
mechanism in mares but based on studies in other species
have specific functions in the cells of the follicle wall related
to growth of follicles. It is proposed that the deviation
process represents the entire mechanism of follicle selection
in mares for the following reasons: (1) although the future
DF emerges before the future SF on average, the 2 largest
follicles often (eg, 33% incidence) change ranks during a
common growth phase that precedes diameter deviation
and (2) the capacity for dominance is shared by several
follicles, and the future subordinates must be selected
against by the deviation mechanism.
Acknowledgments
Preparation of this review was sponsored by the
Eutheria Foundation, Cross Plains, WI, USA. The author
thanks Maria Hoffman for word processing and figure
preparation and Estela Araujo and Sri Vidya Dangudu-
biyyam for word processing.
References
[1] Ginther OJ. Reproductive biology of the mare: basic and applied
aspects. 2nd edition. Cross Plains, WI: Equiservices Publishing;
1992.
[2] Gastal EL, Gastal MO, Bergfelt DR, Ginther OJ. Role of diameter
differences among follicles in selection of a future dominant fol-
licle in mares. Biol Reprod 1997;57:1320–7.
[3] Ginther OJ. The theory of follicle selection in cattle. Domest Anim
Endocrinol 2016;57:85–99.
[4] Ginther OJ, Beg MA, Gastal MO, Gastal EL. Follicle dynamics and
selection in mares. Anim Reprod 2004;1:45–63.
[5] Ginther OJ. Ultrasonic imaging and animal reproduction: book 4,
color-Doppler ultrasonography. Cross Plains, WI: Equiservices
Publishing; 2007.
[6] Beg MA, Ginther OJ. Follicle selection in cattle and horses: role of
intrafollicular factors. Reproduction 2006;132:365–77.
[7] Ginther OJ. Selection of the dominant follicle in cattle and horses.
Anim Reprod Sci 2000;60-61:61–79.
[8] Ginther OJ, Beg MA, Bergfelt DR, Donadeu FX, Kot K. Follicle se-
lection in monovular species. Biol Reprod 2001;65:638–47.
[9] Raz T, Aharonson-Raz K. Ovarian follicular dynamics during the
estrous cycle in the mare. Isr J Vet Med 2012;67:11–8.
[10] Roser JF, Meyers-Brown G. Superovulation in the mare: a work in
progress. J Equine Vet Sci 2012;32:376–86.
[11] Ginther OJ, Beg MA, Gastal EL, Gastal MO, Baerwald AR, Pierson RA.
Systemic concentrations of hormones during the development of
follicular waves in mares and women: a comparative study.
Reproduction 2005;130:379–88.
 O.J. Ginther / Domestic Animal Endocrinology 59 (2017) 116–133
[12] Ginther OJ, Gastal EL, Gastal MO, Bergfelt DR, Baerwald AR,
Pierson RA. Comparative study of the dynamics of follicular waves
in mares and women. Biol Reprod 2004;71:1195–201.
[13] Mihm M, Evans AC. Mechanism of dominant follicle selection in
monovulatory species: a comparison of morphological, endocrine
and intraovarian events in cows, mares and women. Reprod
Domest Anim 2008;43:48–56.
[14] Ginther OJ, Pierson RA. Ultrasonic anatomy of equine ovaries.
Theriogenology 1984;21:471–83.
[15] Aitken WA. The oestrous cycle of the mare. Vet Pract Bull 1926;8:
178–89.
[16] Ginther OJ, Whitmore HL, Squires EL. Characteristics of estrus,
diestrus, and ovulation in mares and effects of season and nursing.
Am J Vet Res 1972;33:1935–9.
[17] Hughes JP, Stabenfeldt GH, Evans JW. Clinical and endocrine as-
pects of the estrous cycle of the mare. Proc Ann Conv Amer Assoc
Equine Pract San Francisco 1972;18:119–51.
[18] Pineda MH, Garcia MC, Ginther OJ. Effect of antiserum against an
equine pituitary fraction on corpus luteum and follicles in mares
during diestrus. Am J Vet Res 1973;34:181–3.
[19] Whitmore HL, Wentworth BC, Ginther OJ. Circulating concen-
trations of luteinizing hormone during estrous cycle of mares
as determined by radioimmunoassay. Am J Vet Res 1973;34:
631–6.
[20] Douglas RH, Nuti L, Ginther OJ. Induction of ovulation and multiple
ovulation in seasonally-anovulatory mares with equine pituitary
fractions. Theriogenology 1974;2:133–42.
[21] Evans MJ, Irvine CHG. Serum concentrations of FSH, LH and pro-
gesterone during the oestrous cycle and early pregnancy in the
mare. J Reprod Fertil Suppl 1975:193–200.
[22] Garcia MC, Ginther OJ. Regulation of plasma LH by estradiol and
progesterone in overiectomized mares. Biol Reprod 1978;19:447–53.
[23] Miller KF, Wesson JA, Ginther OJ. Changes in concentrations of
circulating gonadotropins following administration of equine follic-
ular fluid to ovariectomized mares. Biol Reprod 1979;21:867–72.
[24] Ginther OJ. Reproductive biology of the mare: basic and
applied aspects. 1st edition. Cross Plains, WI: Equiservices Pub-
lishing; 1979.
[25] Palmer E, Driancourt MA. Use of ultrasonic echography in equine
gynecology. Theriogenology 1980;13:203–16.
[26] Miller KF, Wesson JA, Ginther OJ. Interaction of estradiol and a
nonsteroidal follicular fluid substance in the regulation of gonad-
otropin secretion in the mare. Biol Reprod 1981;24:354–8.
[27] Fay JE, Douglas RH. Changes in thecal and granulosa cell LH and FSH
receptor content associated with follicular fluid and peripheral
plasma gonadotrophin and steroid hormone concentrations in pre-
ovulatory follicles of mares. J Reprod Fertil Suppl 1987;35:169–81.
[28] Sirois J, Ball BA, Fortune JE. Patterns of growth and regression of
ovarian follicles during the oestrous cycle and after hemi-
ovariectomy in mares. Equine Vet J 1989;21:43–8.
[29] Nambo Y, Kaneko H, Nagata S, Oikawa M, Yoshihara T,
Nagamine N, Watanabe G, Taya K. Effect of passive immunization
against inhibin on FSH secretion, folliculogenesis and ovulation
rate during the follicular phase of the estrous cycle in mares.
Theriogenology 1998;50:545–57.
[30] Bergfelt DR, Mann BG, Schwartz NB, Ginther OJ. Circulating con-
centrations of immunoreactive inhibin and FSH during estrous
cycle of mares. J Equine Vet Sci 1991;11:319–22.
[31] Ginther OJ, Bergfelt DR. Associations between FSH concentrations
and major and minor follicular waves in pregnant mares. Ther-
iogenology 1992;38:807–21.
[32] Bergfelt DR, Ginther OJ. Relationships between circulating con-
centrations of FSH and follicular waves during early pregnancy in
mares. J Equine Vet Sci 1992;12:274–9.
[33] Gastal EL, Donadeu FX, Gastal MO, Ginther OJ. Echotextural
changes in the follicular wall during follicle deviation in mares.
Theriogenology 1999;52:803–14.
[34] Gastal EL, Gastal MO, Wiltbank MC, Ginther OJ. Follicle deviation
and intrafollicular and systemic estradiol concentrations in mares.
Biol Reprod 1999;61:31–9.
[35] Gastal EL, Gastal MO, Ginther OJ. Experimental assumption of
dominance by a smaller follicle and associated hormonal changes
in mares. Biol Reprod 1999;61:724–30.
[36] Gastal EL, Bergfelt DR, Nogueira GP, Gastal MO, Ginther OJ. Role of
luteinizing hormone in follicle deviation based on manipulating
progesterone concentration in mares. Biol Reprod 1999;61:1492–8.
[37] Donadeu FX, Ginther OJ. Effect of number and diameter of follicles
on secretion of inhibin and suppression of FSH in mares. Repro-
duction 2001;121:897–903.
131
[38] Bergfelt DR, Gastal EL, Ginther OJ. Response of estradiol and
inhibin to experimentally reduced luteinizing hormone during
follicle deviation in mares. Biol Reprod 2001;65:426–32.
[39] Donadeu FX, Ginther OJ. Changes in concentrations of follicular
fluid factors during follicle selection in mares. Biol Reprod 2002;
66:1111–8.
[40] Davidson TR, Chamberlain CS, Bridges TS, Spicer LJ. Effect of follicle
size on in vitro production of steroids and insulin-like growth
factor (IGF)-1, IGF-II, and the IGF-binding proteins by equine
ovarian granulosa cells. Biol Reprod 2002;66:1640–8.
[41] Bridges TS, Davidson TR, Chamberlain CS, Beisert RD, Spicer LJ.
Changes in follicular fluid steroids, insulin-like growth factors
(IGF) and IGF-binding protein concentration, and proteolytic ac-
tivity during equine follicular development. J Anim Sci 2002;80:
179–90.
[42] Ginther OJ, Meira C, Beg MA, Bergfelt DR. Follicle and endocrine
dynamics during experimental follicle deviation in mares. Biol
Reprod 2002;67:862–7.
[43] Donadeu FX, Ginther OJ. Suppression of circulating concentrations
of FSH and LH by inhibin and estradiol during the initiation of
follicle deviation in mares. Theriogenology 2003;60:1423–34.
[44] Donadeu FX, Ginther OJ. Interactions of follicular factors and sea-
son in the regulation of circulating concentrations of gonadotro-
pins in mares. Reproduction 2003;125:743–50.
[45] Gastal EL, Gastal MO, Beg MA, Ginther OJ. Interrelationships
among follicles during the common-growth phase of a follicular
wave and capacity of individual follicles for dominance in mares.
Reproduction 2004;128:417–22.
[46] Acosta TJ, Beg MA, Ginther OJ. Aberrant blood flow area and
plasma gonadotropin concentrations during the development of
dominant-sized transitional anovulatory follicles in mares. Biol
Reprod 2004;71:637–42.
[47] Acosta TJ, Gastal EL, Gastal MO, Beg MA, Ginther OJ. Differential
blood flow changes between the future dominant and subordinate
follicles precede diameter changes during follicle selection in
mares. Biol Reprod 2004;71:502–7.
[48] Ginther OJ, Gastal EL, Gastal MO, Checura CM, Beg MA. Dose-
response study of intrafollicular injection of insulin-like growth
factor-1 on follicular fluid factors and follicle dominance in mares.
Biol Reprod 2004;70:1063–9.
[49] Ginther OJ, Gastal EL, Gastal MO, Beg MA. Critical role of insulin-
like growth factor system in follicle selection and dominance in
mares. Biol Reprod 2004;70:1374–9.
[50] Ginther OJ, Gastal EL, Gastal MO, Beg MA. In vivo effects of
pregnancy-associated plasma protein-A, activin-A and vascular
endothelial growth factor on other follicular-fluid factors during
follicle deviation in mares. Reproduction 2005;129:489–96.
[51] Ginther OJ, Gastal EL, Gastal MO, Beg MA. Regulation of circulating
gonadotropins by the negative effects of ovarian hormones in
mares. Biol Reprod 2005;76:315–23.
[52] Ginther OJ, Utt MD, Beg MA, Gastal EL, Gastal MO. Negative
effect of estradiol on luteinizing hormone throughout the
ovulatory luteinizing hormone surge in mares. Biol Reprod
2007;77:543–50.
[53] Ginther OJ, Jacob JC, Gastal MO, Gastal EL, Beg MA. Follicle and
systemic hormone interrelationships during spontaneous and
ablation-induced ovulatory waves in mares. Anim Reprod Sci
2008;106:181–7.
[54] Jacob JC, Gastal EL, Gastal MO, Carvalho GR, Beg MA, Ginther OJ.
Temporal relationships and repeatability of follicle diameters and
hormone concentrations within individuals in mares. Reprod
Domest Anim 2009;44:92–9.
[55] Jacob JC, Gastal EL, Gastal MO, Carvalho GR, Beg MA, Ginther OJ.
Follicle deviation in ovulatory follicular waves with one or two
dominant follicles in mares. Reprod Domest Anim 2009;44:248–
54.
[56] Checura CM, Beg MA, Gastal EL, Gastal MO, Wiltbank MC, Parrish JJ,
Ginther OJ. Effect of suppression of FSH with a GnRH antagonist
(acyline) before and during follicle deviation in the mare. Reprod
Domest Anim 2009;44:504–11.
[57] Checura CM, Beg MA, Parrish JJ, Ginther OJ. Positive effect of FSH
but not LH on early development of the dominant follicle in mares.
Reprod Fertil Dev 2010;22:1092–9.
[58] Checura CM, Beg MA, Parrish JJ, Ginther OJ. Functional relation-
ships among intrafollicular insulin-like growth factor 1, circulatory
gonadotropins, and development of the dominant follicle in mares.
Anim Reprod Sci 2010;118:270–8.
[59] Claes A, Ball BA, Troedsson MHT, Curry Jr TE, Squires EL,
Scoggin KE. Mollecular changes in the equine follicle in relation to
132 O.J. Ginther / Domestic Animal Endocrinology 59 (2017) 116–133
variations in antral follicle count and anti-Müllerian hormone
concentrations. Equine Vet J 2016;48:741–8.
[60] Alves KA, Alves BG, Gastal GDA, de Tarso SGS, Gastal MO,
Figueiredo JR, Gambarini ML, Gastal EL. The mare model to study
the effects of ovarian dynamics on preantral follicle features. PLoS
One 2016;11:1–18.
[61] Küpfer M. The sexual cycle of female domesticated mammals. Rpts
Director Vet Ed Dept Agr, Union So Afr 1928:1209–70.
[62] Ginther OJ. How ultrasound technologies have expanded and
revolutionized research in reproduction in large animals. Ther-
iogenology 2014;81:112–25.
[63] Pierson RA, Ginther OJ. Ultrasonography of the bovine ovary.
Theriogenology 1984;21:495–504.
[64] Pierson RA, Ginther OJ. Follicular populations during the estrous
cycle of the mare. Anim Reprod Sci 1987;14:219–31.
[65] Ginther OJ. Folliculogenesis during the transitional period and
early ovulatory season in mares. J Reprod Fertil 1990;90:311–20.
[66] Irvine CHG. Endocrinology of the estrous cycle of the mare: ap-
plications to embryo transfer. Theriogenology 1981;15:85–104.
[67] Turner DD, Garcia MC, Ginther OJ. Follicular and gonadotrophic
changes throughout the year in pony mares. Am J Vet Res 1979;40:
1694–700.
[68] Ginther OJ. Ultrasonic imaging and animal reproduction: book 2,
horses. Cross Plains, WI: Equiservices Publishing; 1995.
[69] Vos PLAM, de Loos FAM, Pieterse MC, Bevers MM, Taverne MAM,
Dieleman SJ. Evaluation of transvaginal ultrasound-guided follicle
puncture to collect oocytes and follicular fluids at consecutive
times relative to the preovulatory LH surge in eCG/PG-treated
cows. Theriogenology 1994;41:829–40.
[70] Ginther OJ, Bergfelt DR, Beg MA, Meira C, Kot K. In vivo effects of an
intrafollicular injection of insulin-like growth factor 1 on the
mechanism of follicle deviation in heifers and mares. Biol Reprod
2004;70:99–105.
[71] Beg MA, Gastal EL, Gastal MO, Si J, Wiltbank MC, Ginther OJ.
Changes in steady-state concentrations of messenger ribonucleic
acids in luteal tissue during prostaglandin F2a induced luteolysis
in mares. Anim Reprod Sci 2005;90:273–85.
[72] Ginther OJ. Major and minor waves during the equine estrous
cycle. J Equine Vet Sci 1993;13:18–25.
[73] Ginther OJ, Pierson RA. Regular and irregular characteristics of
ovulation and the interovulatory interval in mares. J Equine Vet Sci
1989;9:4–12.
[74] Gastal EL, Neves AP, Mattos RC, Petrucci BPL, Gastal MO,
Ginther OJ. Miniature ponies: 1. follicular, luteal and endometrial
dynamics during the oestrous cycle. Reprod Fertil Dev 2008;20:
386–90.
[75] Ginther OJ, Bergfelt DR. Ultrasonic characterization of follicular
waves in mares without maintaining identity of follicles. J Equine
Vet Sci 1992;12:349–54.
[76] Ginther OJ, Gastal EL, Gastal MO, Beg MA. Seasonal influence on
equine follicle dynamics. Anim Reprod 2004;1:31–44.
[77] Watson ED, Heald M, Tsigos A, Leask R, Steele M, Groome NP,
Riley SC. Plasma FSH, inhibin A, and inhibin isoforms containing
pro- and dC during winter anestrus, spring transition, and the
breeding season in mares. Reproduction 2002;123:535–42.
[78] Donadeu FX, Ginther OJ. Interrelationships of estradiol, inhibin and
gonadotropins during follicle deviation in pony mares. Ther-
iogenology 2004;61:1395–405.
[79] Donadeu FX, Ginther OJ. Follicular waves and circulating concen-
trations of gonadotropins, inhibin, and oestradiol during the
anovulatory season in mares. Reproduction 2002;124:875–85.
[80] Ginther OJ, Woods BG, Meira C, Beg MA, Bergfelt DR. Hormonal
mechanism of follicle deviation as indicated by major versus minor
follicular waves during the transition into the anovulatory season
in mares. Reproduction 2003;126:653–60.
[81] Gastal EL, Gastal MO, Nogueira GP, Bergfelt DR, Ginther OJ. Tem-
poral interrelationships among luteolysis, FSH and LH concentra-
tions and follicle deviation in mares. Theriogenology 2000;53:
925–40.
[82] Ginther OJ, Wiltbank MC, Fricke PM, Gibbons JR, Kot K. Selection of
the dominant follicle in cattle. Biol Reprod 1996;55:1187–94.
[83] Ginther OJ, Gastal EL, Gastal MO, Beg MA. Intrafollicular effect of
IGF1 on development of follicle dominance in mares. Anim Reprod
Sci 2008;105:417–23.
[84] Ginther OJ. Ultrasonic imaging and animal reproduction: book 3,
cattle. Cross Plains, WI: Equiservices Publishing; 1998.
[85] Ginther OJ. Ultrasonic imaging and animal reproduction: book 1,
fundamentals. Cross Plains, WI: Equiservices Publishing; 1995.
[86] Padmanabhan V, McNeilly AS. Is there an FSH-releasing factor?
Reproduction 2001;121:21–30.
[87] Ginther OJ, Beg MA, Donadeu FX, Bergfelt DR. Mechanism of fol-
licle deviation in monovular farm species. Anim Reprod Sci 2003;
78:239–57.
[88] Pineda MH, Ginther OJ, McShan WH. Regression of the corpus
luteum in mares treated with an antiserum against an equine pi-
tuitary fraction. Am J Vet 1972;33:1767–73.
[89] Pineda MH, Ginther OJ. Inhibition of estrus and ovulation in mares
treated with an antiserum against an equine pituitary fraction. Am
J Vet Res 1972;33:1775–80.
[90] Bergfelt DR, Ginther OJ. Follicular populations following the inhi-
bition of FSH with equine follicular fluid during early pregnancy in
the mare. Theriogenology 1986;26:735–47.
[91] Lapin DR, Ginther OJ. Induction of ovulation and multiple ovula-
tions in anovulatory and ovulatory mares with an equine pituitary
extract. J Anim Sci 1977;44:834–42.
[92] Bergfelt DR, Ginther OJ. Delayed follicular development and
ovulation following inhibition of FSH with equine follicular fluid in
the mare. Theriogenology 1985;24:99–108.
[93] Bergfelt DR, Ginther OJ. Relationships between FSH surges and
follicular waves during the estrous cycle in mares. Theriogenology
1993;39:781–96.
[94] Adams GP, Matteri RP, Kastelic JP, Ko JC, Ginther OJ. Association
between surges of follicle-stimulating hormone and the emergence
of follicular waves in heifers. J Reprod Fertil 1992;94:177–88.
[95] Ginther OJ, Bergfelt DR. Growth of small follicles and concentra-
tions of FSH during the equine estrous cycle. J Reprod Fertil 1993;
99:105–11.
[96] Pierson RA, Ginther OJ. Ovarian follicular response of mares to an
equine pituitary extract after suppression of follicular develop-
ment. Anim Reprod Sci 1990;22:131–44.
[97] Squires EL, Garcia RH, Ginther OJ, Voss JL, Seidel GE. Comparison of
equine pituitary extract and follicle stimulating hormone for
superovulating mares. Theriogenology 1986;26:661–70.
[98] Watson ED, Thomassen R, Steele M, Heald M, Leask R, Groome NP,
Riley SC. Concentrations of inhibin, progesterone and oestradiol in
fluid from dominant and subordinate follicles from mares during
spring transition and breeding season. Anim Reprod Sci 2002;74:
55–67.
[99] Knight PG, Beard AJ, Wrathall JHM, Castillo RJ. Evidence that the
bovine ovary secretes large amounts of monomeric inhibin alpha
subunit and its isolation from bovine follicular fluid. J Mol Endo-
crinol 1989;2:189–200.
[100] Ginther OJ, Gastal EL, Gastal MO, Duarte LF, Beg MA. Follicle sup-
pression of circulating follicle-stimulating hormone and luteiniz-
ing hormone before versus after emergence of the ovulatory wave
in mares. Theriogenology 2009;72:445–52.
[101] Goudet G, Belin F, Bezard J, Gerard N. Intrafollicular content of
luteinizing hormone receptor a-inhibin, and aromatase in relation
to follicular growth, estrous cycle stage, and oocyte competence
for in vitro maturation in mare. Biol Reprod 1999;60:1120–7.
[102] Webb R, Campbell BK, Garverick HA, Gong JG, Gutierrez CG,
Armstrong DG. Molecular mechanisms regulating follicular
recruitment and selection. J Reprod Fertil 1999;54:33–48.
[103] Glister C, Groome NP, Knight PG. Oocyte-mediated suppression of
follicle-stimulating hormone– and insulin-like growth factor–
induced secretion of steroids and inhibin-related proteins by
bovine granulosa cells in vitro: possible role of transforming
growth factor a. Biol Reprod 2003;68:758–65.
[104] Spicer LJ, Chamberlain CS. Production of insulin-like growth factor-
I by granulosa cells but not thecal cells is hormonally responsive in
cattle. J Anim Sci 2000;78:2919–26.
[105] Spicer LJ. Proteolytic degradation of insulin-like growth factor
binding proteins by ovarian follicles: a control mechanism for se-
lection of the dominant follicles. Biol Reprod 2004;70:1223–30.
[106] Monget P, Fabre S, Mulsant P, Lecerf F, Elsen J-M, Mazerbourg S,
Pisselet C, Monniaux D. Regulation of ovarian folliculogenesis by
IGF and BMP system in domestic animals. Domest Anim Endo-
crinol 2002;23:139–54.
[107] Bondy CA, Zhou A, Arratzoa JA. Growth hormone, insulin-like
growth factors, and the ovary. In: Neill JD, editor. Knobil and
Neill’s physiology of reproduction. 3rd edition, vol. 1. London, UK:
Elsevier Inc; 2006. p. 527–46.
[108] Mazerbourg S, Bondy CA, Zhou J, Monget P. The insulin-like growth
factor system: a key determinant role in the growth and selection
of ovarian follicles? A comparative study. Reprod Domest Anim
2003;38:247–58.
 O.J. Ginther / Domestic Animal Endocrinology 59 (2017) 116–133
[109] Rivera GM, Fortune JE. Selection of the dominant follicle and
insulin-like growth factor (IGF)-binding proteins: evidence that
pregnancy-associated plasma protein A contributes to proteolysis
of IGF-binding protein 5 in bovine follicular fluid. Endocrinology
2003;144:437–46.
[110] Beg MA, Beira C, Bergfelt DR, Ginther OJ. Role of oestradiol in
growth of follicles and follicle deviation in heifers. Reproduction
2003;125:847–54.
[111] Belin F, Goudet G, Duchamp G, Gèrard N. Intrafollicular con-
centrations of steroids and steroidogenic enzymes in relation
to follicular development in the mare. Biol Reprod 2000;62:
1335–43.
[112] Rosenfeld CS, Wagner JS, Roberts RM, Lubahn DB. Intraovarian
actions of oestrogen. Reproduction 2001;122:215–26.
[113] Wrathal JHM, Knight PG. Effect of inhibin related peptides and
estradiol on androstenedione and progesterone secretion by
bovine theca cells in vitro. J Endocrinol 1995;145:491–500.
[114] Fortune JE, Quirk SM. Regulation of steroidogenesis in bovine
preovulatory follicles. J Anim Sci 1988;66:1–8.
133
[115] Campbell BK, Baird DT. Inhibin A is a follicle stimulating hormone-
responsive marker of granulosa cell differentiation, which has both
autocrine and paracrine actions in sheep. J Endocrinol 2001;169:
333–45.
[116] Knight PG, Glister C. Potential local regulatory functions of in-
hibins, activins and follistatin in the ovary. Reproduction 2001;
121:503–12.
[117] Redmer DA, Reynolds LP. Angiogenesis in the ovary. Rev Reprod
1996;1:182–92.
[118] Martinez-Chequer JC, Stouffer RL, Hazzard TM, Patton PE,
Molkness TA. Insulin-like growth factors-1 and -2, but not hypoxia,
synergize with gonadotropin hormone to promote vascular
endothelial growth factor-A secretion by monkey granulosa cells
from preovulatory follicles. Biol Reprod 2003;86:1112–8.
[119] Schams D, Kosmann B, Berisha B, Amselgruber VM, Miyamoto A.
Stimulatory and synergistic effects of luteinizing hormone and
insulin like growth factor 1 on the secretion of vascular endothelial
growth factor and progesterone of cultured bovine granulosa cells.
Exp Clin Endocrinol Diabetes 2001;109:155–62.