BIOLOGY OF REPRODUCTION 61, 31–39 (1999)
Follicle Deviation and Intrafollicular and Systemic Estradiol Concentrations
in Mares 1
E.L. Gastal,3 M.O. Gastal,5 M.C. Wiltbank,4 and O.J. Ginther2,3
Departments of Animal Health and Biomedical Sciences,3 and Dairy Science,4 University of Wisconsin,
Madison, Wisconsin 53706
Department of Animal Science,5 Federal University of Viçosa, Viçosa, MG 31570-000, Brazil
ABSTRACT
the largest follicle is about 23 mm [1]. Growth rates then
begin to differ between one follicle and the other follicles,
and this process is called deviation. One follicle, usually
the largest, becomes dominant and continues to grow to a
preovulatory diameter of $ 30 mm, whereas the other follicles of the wave (subordinate follicles) regress.
The diameter relationships of the two largest follicles
have been studied with a two-follicle model, wherein other
follicles were ablated to allow precise tracking of the two
largest follicles by transrectal ultrasonography [1]. The future dominant follicle emerged at 6 mm a mean of 1 day
earlier than the future subordinate follicle, the growth rates
for the two follicles between emergence and deviation (6
days later) did not differ, and the future dominant follicle
was a mean of 3 mm larger than the future subordinate
follicle at the beginning of deviation (23 vs. 20 mm). These
results supported the hypothesis that the follicle destined to
become dominant has a size advantage and is the first to
reach a critical stage at approximately 23 mm. At that time,
the selected follicle is involved in a deviation mechanism
that inhibits the other follicles before the second-largest follicle reaches a similar diameter an average of 1 day later.
Similar results have been reported in cattle, except for the
differences between species in follicle diameters [2, 3]. The
specific mechanism for continued growth of one follicle
and inhibition of the others is not known in either species,
although estradiol has been suggested as a potential facilitator of deviation in cattle [2]. On the basis of in vivo
sampling of follicular fluid in heifers, estradiol concentrations were not higher in the dominant follicle than in the
largest subordinate follicle until the day after the two follicles began to deviate in growth rates [4]. These observations indicate that the future dominant follicle cannot be
identified in cattle reliably by estradiol production before
the deviation in growth rates between the two largest follicles.
The experiments reported here were designed to evaluate
the interrelationships among changing diameters of the
larger and smaller follicles, day of deviation, and intrafollicular and circulating estradiol concentrations using the
two-follicle model in mares. In order to perform this research, an in vivo method was developed and validated for
sampling follicular fluid before deviation in growth rates
between the two follicles.
By definition, follicle deviation begins on the day the two
largest follicles of a wave begin to differ in growth rates. The
relationships between follicle deviation and intrafollicular and
systemic estradiol concentrations were studied in ponies, using
a two-follicle model in which all but the two largest follicles
were ablated. A 20-ml sample of follicular fluid was obtained
from each of the two follicles by transvaginal ultrasonography.
In experiment 1, the two follicles were sampled when the larger
follicle reached 15 mm. No differences ( p . 0.05) in post-sampling follicle characteristics were found between control (n 5
6) and sampled (n 5 8) groups except that the growth rate was
slower ( p , 0.01) in the larger follicle between the day of sampling and the next day (0.7 6 0.7 mm per day) than in the
controls (3.3 6 0.3 mm per day). The growth rates between 2
and 5 days after sampling were not different between groups.
Follicular fluid estradiol-17b concentrations were higher ( p ,
0.007) in the larger follicle (460 6 67 ng/ml; diameter, 16.4 6
0.4 mm) than in the smaller follicle (322 6 50 ng/ml; diameter,
14.6 6 0.6 mm). In experiment 2, the pair of follicles was sampled when the larger follicle reached 15 mm, 20 mm, or 25 mm
(n 5 5 per group). There were no significant differences among
the three groups for day of deviation and diameters of larger
and smaller follicles at deviation. The difference in diameter between the larger and smaller follicles was similar for the 15-mm
(2.2 6 0.9 mm) and 20-mm (3.1 6 1.0 mm) groups, but the
difference between follicles for the 25-mm group (7.9 6 1.2
mm) was greater ( p , 0.004) than for the other two groups. In
contrast, the differences in estradiol concentrations between the
larger and smaller follicles increased ( p , 0.0001) progressively
for the 15-mm (13.0 6 86.8 ng/ml), 20-mm (722.0 6 173.8 ng/
ml), and 25-mm (1873.5 6 310.3 ng/ml) groups. The first significant ( p , 0.007) increase in systemic estradiol occurred between the day before and the day of the beginning of deviation.
Detection of an increased difference in estradiol concentrations
between the two follicles before the detection of a change in
differences in diameter suggests, on a temporal basis, that estradiol is a candidate for involvement in the mechanism that
leads to follicle-diameter deviation in mares.
INTRODUCTION
Pony mares usually have one major follicular wave during an estrous cycle of 24 days. The wave emerges as 6mm follicles at mid-diestrus (approximately Day 12; Day
0 5 ovulation), and the follicles grow at similar rates until
MATERIALS AND METHODS
Accepted February 2, 1999.
Received December 1, 1998.
1
This study was supported by Equiservices Publishing and by Equiculture, Inc., Cross Plains, WI. E.L.G. and M.O.G. were supported by the
Federal University of Viçosa and by a CAPES scholarship, Brazil. This
work was presented in part at the 25th Annual Conference of the International Embryo Transfer Society, Québec, Canada, 1999.
2
Correspondence: O.J. Ginther, Department of Animal Health and Biomedical Sciences, University of Wisconsin-Madison, 1656 Linden Drive,
Madison, WI 53706. FAX: 608 262 7420; e-mail: ojg@ahabs.wisc.edu
Animals and Ultrasonography
Nonlactating cyclic pony mares (n 5 37), 4–13 yr and
218–431 kg, were used from October to November of Year
1 (experiment 1) and from May to June of Year 2 (experiment 2). The mares were kept in partially sheltered outdoor
paddocks under artificial light (15 h per day) from December to March preceding the studies. They were maintained
31
32
GASTAL ET AL.
on alfalfa/grass hay and had access to water and trace-mineralized salt. The ultrasound scanner was equipped with a
5-mHz linear-array transrectal transducer (Tokyo Keiki
LS300; Products Group International, Lyons, CO) for examinations of the ovaries and uterus. An ultrasound scanner
(Aloka SSD-500V; Aloka, Wallingford, CT) equipped with
a 5-mHz convex-array transvaginal transducer (Aloka
UST974V–5) was used for follicle ablations and follicular
fluid sampling. All mares were determined to be in the
ovulatory season by ultrasonic examinations for detection
of ovulation and the formation of a corpus luteum [5]. Mares
with indications of ovarian or uterine abnormalities were
not used. To minimize the potential for obscuring the results, criteria were established for removing mares from the
experiments. The criteria are given in the Results section
for each experiment.
Before the start of the study, mares were synchronized
with prostaglandin-F2a (Lutalyse; Pharmacia & Upjohn Co.,
Kalamazoo, MI). Follicles $ 30 mm in diameter were monitored daily until ovulation. The experiments were started
on Days 10 and 9 in experiments 1 and 2, respectively.
Initially, before the establishment of the two-follicle model,
the average of height and width of the antrum at the maximal area from a single image was used for follicle diameter.
Both experiments used the two-follicle model, prepared
as described [1]. Briefly, on Days 9 or 10, follicles were
ablated by transvaginal aspiration of follicular contents.
Follicles $ 5 mm were ablated in experiment 1, but to
simplify the procedure, follicles $ 8 mm were ablated in
experiment 2. Growing follicles ($ 4 mm) of the new or
post-ablation wave were tracked daily [5]. The relative location of follicles, corpus luteum, and follicle ablation sites
(echoic areas) were used as references for identifying and
tracking. Subsequent ablation sessions were done for ablating follicles of the new wave when the largest follicle
reached 15 mm; all follicles $ 5 mm (experiment 1) or $
8 mm (experiment 2) were ablated except the two largest.
Ablation sessions were repeated whenever a new follicle
reached 15 mm or an ablated follicle refilled, and this continued until the larger of the two retained follicles reached
25 mm. Subsequent to the Day 9/10 ablation session, an
average of 1.2 6 0.1 ablation sessions for establishing the
model were needed per mare. The two retained follicles
were monitored daily until ovulation or Day 30. Height and
width of the two follicles were taken from three images
during a scanning session, and the average of the six measurements per follicle was used as the follicle diameter for
each day. The diameter of the corpus luteum was evaluated
daily as described [5].
Emergence of a follicle was defined as occurring on the
day before the follicle first exceeded 6 mm. The dominant
follicle (one that grew to a preovulatory size of $ 30 mm)
and subordinate follicle (one that grew to a moderate size
and regressed) were chosen retrospectively according to the
maximum attained diameters. The beginning of deviation
between the future dominant and subordinate follicles was
defined as the day the two follicles began to differ in
growth rates [1]. Thus, the beginning of deviation refers to
the examination preceding the first change in differences in
diameters between the two follicles.
Follicle Ablation and Sampling
Mares were held in a padded squeeze stock to prevent
excessive movement. Sedation and analgesia were induced
with detomidine hydrochloride (Dormosedan, 0.02–0.04
mg/kg i.v.; Pfizer Animal Health, West Chester, PA) and
butorphanol tartrate (Torbugesic, 0.05 mg/kg i.v.; Fort
Dodge Laboratories, Fort Dodge, IA). Rectal relaxation was
induced with hyoscine N-butyl bromide (Buscopan, 0.2 mg/
kg i.v.; Sigma Chemical Co., St. Louis, MO). A tail bandage was applied, and the perineal area was aseptically prepared.
Follicle ablation was done transvaginally by ultrasoundguided follicle entry similar to the procedure described in
cattle [6]. The face of the convex transducer was applied
to the wall of the vaginal fornix. The ovary containing the
targeted follicle was positioned transrectally against the
vaginal wall over the transducer face, so that the follicle
was transected by the built-in line on the ultrasound monitor representing the projected needle path. The needle path
to the follicle was positioned so that it transected the ovarian stroma but not other detected follicles or luteal tissue.
For ablation, a 17-gauge needle (outer diameter, 1.5 mm;
inner diameter, 1 mm; length, 55 cm) was used to puncture
individual follicles. Follicular contents were removed using
a vacuum pump (250–300 mm Hg). Follicle ablation was
defined as collapse of the antral follicle after evacuation of
follicular contents.
The procedure for ultrasound-guided transvaginal sampling of follicular fluid was similar to that used for cattle
[4]. The larger and smaller follicles were sampled once per
mare. A double-channel needle system (RAM IVF Supply,
Madison, WI) was used for the sampling and consisted of
a 20-gauge outer needle (outer diameter, 0.91 mm; inner
diameter, 0.55 mm; length, 53 cm) and a 25-gauge inner
needle (outer diameter, 0.52 mm; inner diameter, 0.24 mm;
length, 57 cm). The inner needle was filled with physiological saline, and the syringe was left attached. The inner
needle was inserted into the outer needle until the tip was
approximately 1 cm from the tip of the outer needle, and
the needle set was then inserted into the needle-guide handle of the transvaginal probe. When the ovary and targeted
follicle were in position, the needles were advanced by a
second operator until the image of the tip of the outer needle became visible on the scanner screen, indicating that
the vaginal wall and peritoneum were penetrated. The inner
needle was then advanced until the image of the inner needle tip was centered within the targeted follicle. Sampling
of follicular fluid was done by the second operator with a
100-ml syringe (Hamilton Co., Reno, NV) preset to the desired sample volume (20 ml). Successful sampling required
continuous observation of the needle tip. The needles and
probe were withdrawn immediately after sampling to avoid
exerting continued pressure on the newly sampled follicle.
The 15-ml portion of the 20-ml sample closest to the needle
tip was inserted into tapered microtubes and stored at
2208C.
Estradiol Assays
Follicular fluid estradiol-17b concentrations were evaluated with specific ELISA as previously described [7] and
modified for direct use with follicular fluid [8]. Briefly, follicular fluid samples were either diluted 1:100 or 1:200 in
assay buffer and analyzed directly with the ELISA. The
standard curve was made in assay buffer containing a 1:
100 or 1:200 dilution of charcoal-treated equine follicular
fluid with estradiol concentrations from 31 to 2000 pg/ml.
The intraassay and interassay coefficients of variation were
13.0% and 14.0%, respectively. The sensitivity (3 standard
FOLLICLE DEVIATION AND ESTRADIOL CONCENTRATIONS
33
FIG. 1. Examples of follicular profiles of dominant (closed circles) and largest subordinate (open circles) follicles in individual mares in experiment 1
(a: control; b, c: sampled when larger follicle reached 15 mm) and experiment 2 (d–f: sampled when larger follicle was 15, 20, or 25 mm, respectively).
The smaller follicle before deviation becoming the dominant follicle after deviation is illustrated in c. The transient effect of sampling on growth rate
of the larger follicle is illustrated in b, d–f. Day of deviation (solid arrow) and day of sampling (dashed arrow) are indicated for each mare. OV,
Ovulation.
deviations from maximum bound) was equivalent to a concentration of 0.96 pg/ml.
Systemic estradiol-17b concentrations were measured by
a sensitive RIA procedure that we have modified for use
with equine plasma. The standard curves were made in
charcoal-treated equine plasma and contained estradiol concentrations from 1.25 to 40 pg/ml. All standard curve, quality control, and unknown samples had estradiol extracted
from plasma using diethyl ether. Briefly, 500 ml of plasma
was combined with 2.5 ml of diethyl ether and vortexed
for 2 min. Samples were then frozen in dry ice/methanol,
and the ether fraction was poured into a glass assay tube
(12 3 75 mm). The extraction procedure was repeated, and
ether was evaporated overnight in a ventilated hood. Dried
samples were resuspended in 100 ml of assay buffer and
were directly evaluated with a commercially available RIA
(Ultra-sensitive estradiol DSL-4800; Diagnostic Systems
Laboratory, Webster, TX). For the assay, we used 30 ml of
antiserum, 50 ml of 125I-estradiol, and 1 ml of precipitating
reagent. According to the manufacturer, the assay has low
cross-reactivity with estrone (2.4%), estriol (0.65%), estrone-3-sulfate (0.01%), equilin (0.34%), and estradiol-3sulfate (0.17%). All samples were extracted and evaluated
in duplicate in a single assay. The standard curve was extracted in an identical manner to those of unknown samples
in order to account for extraction efficiency in the assay.
The assay had a sensitivity of 0.31 pg/ml (3 standard deviations from maximum bound), an 80% point of 1.41 pg/
ml, a 50% point of 7.39 pg/ml, an intraassay coefficient of
variation of 6.6% (calculated from quality controls), and a
mean intraassay coefficient of variation for samples of
14.3% (calculated from duplicate extractions of samples).
Experiment 1. Follicular Fluid Estradiol Concentrations
before Deviation
Mares were randomized into a control group (unsampled, n 5 9) and a sampled group (20 ml of follicular fluid
removed from each of the 2 follicles, n 5 10). Sampling
of the two follicles was done once when the larger follicle
first exceeded 14.9 mm (defined as 15 mm; actual diameters
ranged from 15.2 to 17.8 mm).
Experiment 2. Follicular Fluid and Systemic Estradiol
Concentrations Associated with Deviation
Mares (n 5 6 per group) were randomized into three
groups for sampling of 20 ml of follicular fluid from each
of the two follicles once when the larger follicle first exceeded 14.9 mm (15-mm group; before deviation), 19.9 mm
(20-mm group; approximately at deviation), or 24.9 mm
(25-mm group; after deviation). Actual diameters for the
three groups were 15.0–18.0 mm, 20.8–23.5 mm, and
25.0–26.8 mm, respectively. Blood samples were collected
daily, beginning on Day 10 and ending on the day of the
next ovulation. Samples were collected by jugular veni-
34
GASTAL ET AL.
TABLE 1. Mean (6 SEM) intervals, follicle diameters, and growth rates of larger and smaller follicles in control mares and in mares with in vivo
sampling of 20 ml of follicular fluid.a
Diameter at sampling
Experiment 1
End point
Intervals (days) from:
Emergencec to deviation
Larger follicle
Smaller follicle
Sampling to deviationd
Sampling to ovulation
Day 10e to ovulation
25 mm to ovulation
Follicle diameters (mm):
At sampling
Larger follicle
Smaller follicle
At the beginning of deviation
Larger follicle
Smaller follicle
On the day before ovulation
Preovulatory follicle
Growth rates (mm/day) from:
Day before sampling to day of sampling
Larger follicle
Smaller follicle
Day of sampling to 1 day after sampling
Larger follicle
Smaller follicle
1 Day after to 2 days after sampling
Larger follicle
Smaller follicle
Controls
(n 5 6)
6.2 6
5.3 6
1.7 6
10.3 6
16.8 6
7.2 6
0.3
0.3
0.2
0.4
0.8
0.6f
Experiment 2
mmb
15
(n 5 8)
6.0
5.0
1.4
8.8
14.8
4.9
6
6
6
6
6
6
0.3
0.3
0.4
0.9
0.9
0.7g
mmb
15
(n 5 5)
5.4
4.6
1.6
8.6
13.8
4.8
6
6
6
6
6
6
0.5
0.6
0.7f
0.8f
0.6
0.7
20 mmb
(n 5 5)
5.6
5.0
20.2
7.8
14.4
5.6
6
6
6
6
6
6
0.5
0.5
0.2g
0.6f
0.5
0.5
25 mmb
(n 5 5)
6.2
5.2
21.4
4.6
12.6
4.6
6
6
6
6
6
6
0.5
0.6
0.2g
0.5g
0.5
0.5
16.8 6 0.5
15.1 6 0.9
16.4 6 0.4
14.6 6 0.6
16.3 6 0.7f
14.1 6 1.3f
22.2 6 0.6g
19.1 6 1.0g
26.0 6 0.4h
18.1 6 1.5g
22.2 6 1.0
18.9 6 1.1
19.4 6 1.5
17.6 6 1.0
20.4 6 1.9
17.1 6 0.8
21.3 6 1.0
18.8 6 1.2
21.7 6 1.1
19.0 6 1.5
38.3 6 2.0
35.2 6 1.7
37.1 6 1.7fg
38.7 6 2.2f
32.2 6 1.0g
3.4 6 0.3
3.2 6 0.5
2.7 6 0.3y
2.8 6 0.2
3.6 6 0.8y
3.0 6 0.6f
4.1 6 0.5y
3.0 6 0.6f
3.3 6 0.4y
20.8 6 0.6g
3.3 6 0.3f
2.6 6 0.4
0.7 6 0.7gz
1.4 6 0.9
0.2 6 1.4z
1.9 6 0.6f
20.5 6 0.9z
20.2 6 1.2f
0.4 6 1.0z
24.5 6 2.0g
2.9 6 0.4
1.2 6 0.4
3.3 6 0.6y
2.2 6 0.4
2.5 6 1.0yz
2.7 6 0.4f
2.5 6 1.0y
2.1 6 1.2f
2.7 6 0.5y
21.2 6 0.9g
All follicles $ 5 mm (experiment 1) or $ 8 mm (experiment 2) were ablated on day 10 post-ovulation. When the largest follicle of the post-ablation
wave reached 15 mm, a two-follicle model was established by periodically ablating all but the two largest follicles.
b Larger and smaller follicles sampled when the larger follicle was 15 mm (experiment 1) and 15 mm, 20 mm, or 25 mm (experiment 2).
c Emergence, day before the follicle exceeded 6 mm.
d Deviation, day the two follicles began to differ in growth rates.
e Day 10, 10 days after ovulation.
fgh Means within rows in each experiment with no common superscript are different (p , 0.03 to p , 0.0001).
yz Means for growth rate of the larger follicle within a column with no common superscripts are different (p , 0.003 to p , 0.03) or tend to be different
(p , 0.06 to p , 0.08).
a
puncture into heparinized tubes and held for 1–2 h at 48C
until sedimentation. Plasma was decanted and placed in vials for cold storage (-208C) until estradiol assay.
Combined Data
The association between follicle diameters and estradiol
concentrations was determined using all sampled follicles
of the two experiments. To confirm previously reported [1]
relationships between future dominant and subordinate follicles, the control group of experiment 1 (no sampling) and
the 25-mm group of experiment 2 (deviation before sampling) were combined (n 5 11 follicular waves) and studied
from emergence of the dominant follicle until ovulation.
Statistical Analyses
Data were normalized to day of follicular fluid sampling,
day of emergence of the dominant follicle, and day of the
beginning of deviation. Growth profiles of the larger and
smaller follicles at the time of sampling were analyzed by
a group-by-day factorial ANOVA for sequential data. Oneway ANOVA was used to compare groups within the same
end points. Paired and unpaired t-tests were used to compare various characteristics between days within a follicle,
between follicles within end points, or between days. A
two-sample t-test applied to the ranks [9] was used to com-
pare the diameters of the dominant and subordinate follicles
on the day of emergence of the dominant follicle; the ranking test was used because actual diameter of the subordinate
follicle was not available for the day of emergence of the
dominant follicle in 5 of 11 mares. Pearson correlation analyses were used to compare follicle diameters just before
sampling and intrafollicular estradiol concentrations. Significance was defined as p , 0.05.
RESULTS
Sampling of follicular fluid was done in 28 pairs of follicles (total follicles, 56) over the two experiments. Detected effects of follicular fluid sampling on the day after
sampling were a small echoic spot on the follicular wall
(29% of follicles), tiny echoic spots floating in the follicular
fluid (5%), and an apparent blood clot comprising 50–80%
of the antrum (4%). Samples were clear (94%), tinged with
blood (2%), or extensively contaminated with blood (4%).
The follicle sampling success rate, excluding samples with
extensive blood contamination, was 96% (54 of 56). The
sample tinged with blood was included in the data analysis
because the concentration of estradiol did not appear to be
different from those in the other five clear samples in the
same group (15-mm). Data for three pairs of follicles were
not used because one sample of a pair was contaminated
FOLLICLE DEVIATION AND ESTRADIOL CONCENTRATIONS
35
FIG. 2. Profiles (mean 6 SEM) of larger
and smaller follicles (on the day of sampling) normalized to the time of sampling.
In sampled groups, follicular fluid (20 ml)
was removed from each follicle on the day
the larger follicle reached the indicated diameter. a) Main effect of day ( p , 0.0001)
and an interaction of day by group ( p ,
0.03). b) Main effects of day and group ( p
, 0.0001). c) Main effect of day ( p ,
0.0001). d) Main effect of day ( p ,
0.0001) and an interaction of day by
group ( p , 0.0001).
with blood or a follicle developed a blood clot post-sampling. In summary, both satisfactory samples and data from
post-sampling follicle monitoring were obtained from both
follicles for 25 of 28 (89%) mares. Examples of follicular
profiles of the dominant follicle and largest subordinate follicle before and after deviation and follicular fluid sampling
in individual mares of experiments 1 and 2 are shown (Fig.
1).
Experiment 1. Follicular Fluid Estradiol Concentrations
before Deviation
Mares were removed from the experiment because of an
anovulatory wave with no follicle reaching . 27 mm (3
control and 2 sampled mares); subsequent monitoring indicated that these mares had entered the anovulatory season. Thus, a total of 6 control and 8 sampled mares were
available for statistical analyses.
No significant differences between the control and sampled groups were found for any of the following end points:
1) length of intervals between emergence and deviation,
sampling and deviation, sampling and ovulation, and Day
10 to ovulation (Table 1); 2) follicle diameters at sampling
and at deviation (Table 1); 3) day of follicle deviation (controls, 18.2 6 0.5; sampled, 17.2 6 0.4); 4) difference in
diameter between larger and smaller follicles at sampling
(controls, 1.7 6 0.4 mm; sampled, 1.8 6 0.6 mm); and 5)
number of larger follicles at 15 mm that became dominant
follicles (controls, 6 of 6; sampled, 5 of 8). The growth
rates from the day before sampling to the day of sampling
and from 1 day after to 2 days after sampling were not
different between groups within follicles (Table 1). However, the growth rate between the day of sampling and the
next day was slower (p , 0.01) in the sampled group than
in the control group. The subsequent four daily growth rates
were not different between groups. For the diameter of the
larger follicle, the interaction of group and day was significant (p , 0.03; Fig. 2a). For the smaller follicle, there was
no significant group effect nor an interaction (Fig. 2c). Intrafollicular estradiol concentrations were higher (p ,
0.007) in the larger (460 6 67 ng/ml) than in the smaller
(322 6 50 ng/ml) follicles.
Experiment 2. Follicular Fluid and Systemic Estradiol
Concentrations Associated with Deviation
Mares were removed from the experiment for the following reasons: loss of identity of the smaller follicle after
sampling (one mare in the 15-mm group); larger follicle
regressed after reaching 27.5 mm in the presence of a maintained corpus luteum (anovulatory wave, one mare in the
20-mm group); and double ovulations (codominant follicles, one mare in the 25-mm group). Thus, five mares per
group were available for statistical analyses.
There were no significant differences among the three
groups (15-mm, 20-mm, and 25-mm) for day at the begin-
36
GASTAL ET AL.
Growth profiles and statistical data for the two follicles
when normalized to day of sampling are shown (Fig. 2, b
and d). As for experiment 1, a transient decrease in growth
rates of the larger follicle occurred immediately after sampling in all groups.
Mean follicle diameters and follicular fluid and systemic
estradiol concentrations of the larger and smaller follicles
on the day of sampling for the three groups are shown (Fig.
3, a–c). The diameter of the larger follicle increased (p ,
0.0001) progressively over the three groups, whereas the
diameter of the smaller follicle increased (p , 0.04) only
between the 15-mm and 20-mm groups. The difference in
diameters between the larger and smaller follicles was similar (p . 0.05) for the 15-mm and 20-mm groups, but the
difference was greater (p , 0.004) for the 25-mm group
than for either of the other 2 groups. Estradiol concentrations were not different between the larger and smaller follicles in the 15-mm group (493 6 91 vs. 480 6 106 ng/
ml) but were for the 20-mm (1353 6 91 vs. 631 6 168 ng/
ml; p , 0.01) and 25-mm (2220 6 328 vs. 346 6 176 ng/
ml; p , 0.009) groups. The differences in estradiol
concentrations between the larger and smaller follicles increased (p , 0.0001) progressively over the 15-mm, 20mm, and 25-mm groups (Fig. 3b). The differences were
significant between the 15-mm and 20-mm groups and between the 20-mm and 25-mm groups. For systemic estradiol, the difference among the three groups on the day of
sampling was not significant (p , 0.09; Fig. 3c). When the
systemic daily estradiol concentrations were normalized to
the beginning of deviation, the day effect was significant
(p , 0.0001), but the group effect and interaction were not
(Fig. 4a). The first significant (p , 0.008) increase in daily
circulating estradiol concentrations occurred between Day
21 (day before the beginning of deviation) and Day 0 (Fig.
4b).
Combined Data
FIG. 3. Results of experiment 2, showing means 6 SEM for a) diameters
of larger (black bars) and smaller (white bars) follicles, b) follicular fluid
estradiol concentrations of larger (black bars) and smaller (white bars)
follicles, and c) systemic estradiol concentrations when the larger follicle
reached 15 mm (n 5 5), 20 mm (n 5 5), or 25 mm (n 5 4). The larger
and smaller follicles and systemic blood were sampled once per mare.
*Difference ( p , 0.02 to p , 0.0001) between groups (diameters at sampling) within a follicle. Different letters (x, y, z) indicate differences ( p ,
0.01) between groups in the difference between the larger and smaller
follicles. For systemic estradiol (c), the difference among groups was not
significant ( p , 0.09).
ning of deviation (Day 16.8 6 0.4, 16.4 6 0.5 and 16.8 6
0.8, respectively), interval from Day 10 to ovulation (Table
1), and diameters of larger and smaller follicles at the beginning of deviation (Table 1). Diameter of the preovulatory follicle on day before ovulation was smaller (p , 0.05)
in the 25-mm group than in the 20-mm group. The growth
rates of the larger follicle for the 3 days encompassing deviation did not differ among days or between groups (Table
1), but a difference (p , 0.02) in growth rates during the
three intervals was observed between the smaller follicle
of the 25-mm group and the other two groups (Table 1).
Totaled over all experiments, intrafollicular estradiol
concentrations were positively and significantly correlated
to diameter of the larger follicles (r 5 0.92; p , 0.0001)
and to diameter of the smaller follicles (r 5 0.59; p ,
0.004). Mean day-to-day changes in diameter of the future
dominant and subordinate follicles of the control group of
experiment 1 and the 25-mm group of experiment 2 are
shown, using as reference points the day of emergence of
the future dominant follicle at 6 mm (Fig. 5a) and the day
at the beginning of deviation (Fig. 5b). Data were truncated
to the days when observations were available for all mares.
The mean day of follicle deviation was 6.2 6 0.3 days after
emergence of the dominant follicle or on Day 17.4 6 0.4
after ovulation. The future dominant follicle emerged earlier than the future subordinate follicle and was larger than
the future subordinate follicle on the day of emergence of
the dominant follicle and on the day at the beginning of
deviation (Table 2). The growth rates of the future dominant
and subordinate follicles between emergence and deviation
(6 days later) did not differ between follicles.
DISCUSSION
The number of mares (8 of 37, 22%) removed from the
data analyses after successful sampling was large but was
done according to the protocol. Removal of mares was most
important for experiment 1, to minimize the effects of the
approaching anovulatory season. It is not known whether
other seasonal effects influenced the results, but a main goal
FOLLICLE DEVIATION AND ESTRADIOL CONCENTRATIONS
37
FIG. 4. Results of experiment 2, showing
daily systemic estradiol concentrations
(mean 6 SEM) normalized to the beginning of deviation for groups in which the
follicles were sampled when the larger follicle reached the indicated diameter. a)
The day effect was significant ( p ,
0.0001), but the group effect and interaction were not. b) *Differences ( p , 0.05)
between days.
of experiment 1 was development of the sampling technique. Removal of mares when the wave did not produce
an ovulatory follicle was prudent because of the possibility
that the largest follicle was too small to express dominance
(no deviation mechanism). In mares removed for this reason, five (14%) had an anovulatory wave [1, 10], and four
of these entered the anovulatory season [5]. One mare (3%)
was also removed because the larger follicle regressed in
the presence of a maintained corpus luteum, resulting in a
prolonged interovulatory interval (48 days compared to a
normal mean of 24 days) [1, 10]. Prolonged maintenance
of the corpus luteum occurs occasionally during the estrous
cycle in mares [11], especially at the end of the ovulatory
season [12]. The identity of the smaller follicle was lost in
one mare (3%) because of follicular fluid leakage after sampling, and the mare was removed from the study. The double-ovulation rate in the present study (3%) was similar to
a reported rate of 2% in ponies [10] and resulted in removal
of one mare.
The diameter advantage of the future dominant follicle
over the future subordinate follicle, beginning on the day
of emergence at 6 mm, is consistent with a previous study
[1]. The future dominant follicle emerged an average of 1
day earlier than the future subordinate follicle and was larger on the day of emergence and on the day of deviation.
The two follicles, on average, grew in parallel as indicated
by the absence of a significant difference in growth rates
between the days of emergence and deviation. Several studies in cattle have indicated that removal of the largest follicle before or soon after the beginning of deviation allows
the second-largest follicle to become dominant [2]. Apparently, selection of the dominant follicle is not complete until deviation occurs. Results of the present study and previous reports in ponies [1] and cattle [2, 3] indicate the need
for focusing on the deviation mechanism. In a previous
study in cattle [4], intrafollicular but not systemic estradiol
concentrations were examined in relationship to time of deviation. In the present studies in mares, the relationships
between follicle deviation and intrafollicular and systemic
estradiol concentrations were considered.
Ultrasound-guided follicular entry for injection of substances has been done in mares for eCG [13], in heifers for
hCG [14], and in cattle for sampling follicular fluid 1–4
days after follicular-wave emergence [4]. In the present
studies, sampling of follicular fluid by the ultrasound-guided technique was effective for 10- to 27-mm follicles in
mares that were accustomed to the padded squeeze stock
and transrectal and transvaginal procedures. The volume of
follicular fluid sampled (20 ml) was approximately 1.1%,
0.5%, and 0.2% of the fluid volume of a spherical follicle
with an antral diameter of 15, 20, and 25 mm, respectively.
Usable samples and follicle data were obtained from 25 of
28 (89%) pairs of targeted follicles. The undesirable effects
of follicular fluid sampling were similar to those reported
FIG. 5. Mean (6 SEM) diameter of follicles combined for the control group (experiment 1; n 5 6) and 25-mm group (experiment 2; n 5 5). a) Data normalized to
the day of emergence of the future dominant follicle at 6 mm. The number of mares with follicle deviation on each day is
indicated in parentheses, and the mean (6
SEM) day of follicle deviation was 6.2 6
0.3. b) Data normalized to deviation.
Growth rates of the 2 follicles were not
different ( p . 0.05) for 1–5 days after
emergence (a) and 23 to 0 days before
deviation (b).
38
GASTAL ET AL.
TABLE 2. Intervals, follicle diameters, and growth rates combined for controls (experiment 1) and the 25-mm
group (experiment 2).
End point
Intervals (days) from:
Day 10a to emergenceb of follicle
Dominant follicle
Subordinate follicle
Emergence of follicle to deviationc
Dominant follicle
Subordinate follicle
Follicle diameters (mm):
On day of emergence of dominant follicle
Dominant follicle
Subordinate follicle
At the beginning of deviation
Dominant follicle
Subordinate follicle
Growth rates (mm/day) from:
Emergence of follicle to deviation
Dominant follicle
Subordinate follicle
Mean 6 SEM
(n 5 11)
p value
1.3 6 0.4
2.2 6 0.4
p , 0.002
6.2 6 0.3
5.3 6 0.3
p , 0.002
5.3 6 0.2
,4.8d
p , 0.008
21.9 6 0.7
18.9 6 0.9
p , 0.0001
2.7 6 0.1
2.5 6 0.1
NSe
a
Day 10, 10 days after ovulation.
Emergence, day before the follicle exceeded 6 mm.
c Deviation, day the two follicles began to differ in growth rates.
d Values unknown in five mares because the subordinate follicle was not detectable on day of emergence of
dominant follicle; data were analyzed by a ranking test.
e NS 5 not significant.
b
for cattle [4]. Extensive blood contamination of the samples
and blood clot formation in the antrum of targeted follicles
were considered evidence of unsuccessful sampling attempts. The failures seemed associated with mare movements or technique problems during the follicle puncture.
The frequency of failures seemed to decrease as the experiments progressed. Better maintenance of the needle tip in
the center of the follicle and better coordination between
the two operators may account for some of the improved
success with increased experience. The results of a trial in
cattle [4] indicated that aspiration of a consistent volume
of follicular fluid required that the inner needle (25-gauge)
be filled with fluid (saline). However, some mixing of the
follicular fluid with the saline occurred, as indicated by a
3–11% reduction in estradiol in the sampled fluid closest
to the saline. In the present study, altered concentrations
because of mixing of fluids presumably was minimized by
using the 15-ml portion of the sample closest to the needle
tip of an aspirated volume of 20 ml.
The mean growth rate of the larger sampled follicles (15,
20, and 25 mm) was reduced on the day immediately after
sampling, but thereafter the follicles grew at a rate similar
to the rate in controls or the rate before sampling. The
effects of sampling on diameter of the smaller follicle were
confounded by regression in association with deviation. As
indicated by the comparisons between the sampled and control groups in experiment 1, sampling at 15 mm did not
significantly alter the subsequent follicle characteristics associated with deviation (interval from sampling to the beginning of deviation, diameter of follicles at deviation). The
larger follicle at sampling was on average significantly larger at deviation; however, in 3 of 8 mares in the sampled
group, compared to 0 of 6 in the control group, the larger
follicle at sampling became the subordinate follicle after
deviation. These ratios were not significantly different, but
the number of mares was small. When data were combined
for the control group of experiment 1 and the 25-mm group
of experiment 2 (dominance established before sampling),
the larger follicle at 15 mm became the dominant follicle
in 11 of 11 mares. Combined data for mares with follicles
sampled at 15 mm in the 2 experiments showed that the
follicle that was larger when sampled became the dominant
follicle in 8 of 13 mares (11 of 11 vs. 8 of 13, p , 0.07).
On this basis, an effect of sampling on the future status of
a follicle (dominant vs. subordinate) was considered suggestive or at least equivocal. For this reason, no attempt
was made to reclassify the larger and smaller follicles at
sampling on the basis of their eventual status as dominant
and subordinate follicles.
There were indications that follicle sampling or puncture
hastened the time of ovulation after puncture at 15 mm in
experiment 1. The interval from attainment of a diameter
of 25 mm to ovulation was reduced by a mean of 2.3 days
compared to that in the controls. The corresponding intervals for the 15, 20, and 25-mm sampled groups in experiment 2 were similar to that of the 15-mm sampled group
in experiment 1. In addition, the diameter of the follicle on
the day before ovulation was reduced by a mean of 5 or 6
mm in the 25-mm group compared to the other 2 groups.
The reason for the ovulation effect is not known. Ovulation
is associated with an inflammatory reaction in many species
[15], and speculatively the inflammation or healing associated with puncture of the follicular wall in mares may
have hastened ovulation.
In experiment 2, follicle deviation had not begun on the
day of sampling in the 15-mm group, but it had begun in
the 25-mm group on the basis of changes in the differences
in diameters between the larger and smaller follicles. The
beginning of deviation, defined as the day before a change
in the differences in diameters between the two follicles,
was on the day of sampling for each mare in the 20-mm
group, as indicated by examination of the growth profiles
of the two follicles for each mare. Furthermore, there was
close agreement between mean diameter of the larger follicle on the day of sampling in the 20-mm group with the
mean diameter of the larger follicle on the day at the beginning of deviation in nonsampled mares. The larger follicle at the time of sampling was a mean of 22 mm, which
FOLLICLE DEVIATION AND ESTRADIOL CONCENTRATIONS
was similar to the diameter at the beginning of deviation
in the controls of experiment 1 (22 mm), in the 25-mm
group of experiment 2 (deviated before sampling; 22 mm),
and in unsampled follicles of a previous study (22–23 mm
in various groups) [1].
The low systemic estradiol concentrations of 1–5 pg/ml
found in this study agree with some findings [12, 16], but
were much lower than others [17, 18]. An explanation for
the apparent differences in systemic estradiol concentrations in the literature could be the different types of assays
used and the cross-reactivity of each assay with other estrogens (e.g., estrone). According to the manufacturer, the
assay used in the present study has low cross-reactivity with
other estrogens.
The most novel aspect of this study involved the assessment of changes in both intrafollicular and systemic concentrations of estradiol in relation to changes in follicle
diameters. In experiment 2, the progressive increase (not
significant, p , 0.09) in systemic concentrations of estradiol in groups 15 mm, 20 mm, and 25 mm was based on
blood samples collected on the day of follicular fluid sampling. More detailed information was obtained by normalizing all mares to the day of the beginning of diameter
deviation and using daily blood samples. The absence of a
difference among groups or an interaction of group and day
indicated that the concentrations followed a similar pattern
in the three groups. The significant day effect resulted from
progressive increases from the day before the beginning of
deviation to the last day considered (2 days after the beginning of deviation). Thus, an increase in systemic estradiol was detected the day before the detection of follicle
deviation, suggesting that follicular estradiol production
changed before the diameter manifestation of follicle deviation in the mare.
In experiment 1, estradiol concentrations were higher in
the larger follicle when the larger and smaller follicles were
a mean of 16.4 mm and 14.6 mm. However in experiment
2, there was not a significant difference between follicles
in the 15-mm group. This apparent contradiction can be
attributed, at least partly, to more mares in experiment 1 (n
5 8) than in experiment 2 (n 5 5) and to differences in
season (late and early in the ovulatory season for experiments 1 and 2, respectively) [12]. In experiment 2, intrafollicular estradiol concentrations were analyzed at times
encompassing the beginning of deviation. The difference in
follicular fluid estradiol concentrations between the larger
and smaller follicles increased progressively for the 15-mm,
20-mm, and 25-mm groups. In mares in which follicles
were sampled at the beginning of deviation (all in the 20mm group), the difference between the two follicles in estradiol concentrations was greater than for the 15-mm
group. Thus, differences in intrafollicular estradiol concentrations between the two follicles occurred before a change
in the differences in diameter. This result and the observed
changes in systemic estradiol concentrations are compatible
with the conclusion that intrafollicular estradiol production
began to increase in the larger follicle before detection of
the beginning of deviation in diameter. This temporal relationship encourages further study focused on the possible
functional relationship between an increase in intrafollicular estradiol in the largest follicle and the mechanism of
39
diameter deviation. Speculation on the potential manner in
which estradiol could be involved has been reviewed [2].
In conclusion, the results indicated that follicle sampling
did not alter the day of deviation or follicle diameters on
the day of deviation. The follicular fluid sampling technique
developed in these studies may be useful in relating other
intrafollicular factors to the mechanism of follicle selection
in mares. The detection of a change in the differences in
intrafollicular estradiol concentrations between the two follicles before the detection of a change in the differences in
diameter suggested, on a temporal basis, that estradiol is a
candidate for involvement in the mechanism that leads to
follicle-diameter deviation in mares.
ACKNOWLEDGMENTS
The authors thank Pharmacia & Upjohn Co., Kalamazoo, MI, for providing Lutalyse and Ms. Josie A. Lewandowski for helping with the estradiol assays.
REFERENCES
1. Gastal EL, Gastal MO, Bergfelt DR, Ginther OJ. Role of diameter
differences among follicles in selection of a future dominant follicle
in mares. Biol Reprod 1997; 57:1320–1327.
2. Ginther OJ, Wiltbank MC, Fricke PM, Gibbons JR, Kot K. Selection
of the dominant follicle in cattle. Biol Reprod 1996; 55:1187–1194.
3. Ginther OJ, Kot K, Kulick LJ, Wiltbank MC. Emergence and deviation of follicles during development of follicular waves in cattle. Theriogenology 1997; 48:75–87.
4. Ginther OJ, Kot K, Kulick LJ, Wiltbank MC. Sampling follicular fluid
without altering follicular status in cattle: oestradiol concentrations
early in a follicular wave. J Reprod Fertil 1997; 109:181–186.
5. Ginther OJ. Ultrasonic Imaging and Animal Reproduction: Horses.
Book 2. Cross Plains, WI: Equiservices Publishing; 1995: 394.
6. Bergfelt DR, Lightfoot KC, Adams GP. Ovarian synchronization following ultrasound-guided transvaginal follicle ablation in heifers. Theriogenology 1994; 42:895–907.
7. Rasmussen FE, Wiltbank MC, Christensen JO, Grumer RR. Effects
of fenprostalene and estradiol-17b benzoate on parturition and retained placenta in dairy cows and heifers. J Dairy Sci 1996; 79:227–
234.
8. Bodensteiner KJ, Wiltbank MC, Bergfelt DR, Ginther OJ. Alterations
in follicular estradiol and gonadotropin receptors during development
of bovine antral follicles. Theriogenology 1996; 45:499–512.
9. SAS. SAS Procedures Guide, version 6, 3rd ed. Cary, NC: Statistical
Analysis System Institute, Inc.; 1990: 705.
10. Ginther OJ. Reproductive Biology of the Mare: Basic and Applied
Aspects, 2nd ed. Cross Plains, WI: Equiservices; 1992: 642.
11. Ginther OJ. Prolonged luteal activity in mares—a semantic quagmire.
Equine Vet J 1990; 22:152–156.
12. Weedman BJ, King SS, Neumann KR, Nequin LG. Comparison of
circulating estradiol-17b and folliculogenesis during the breeding season, autumn transition and anestrus in the mare. J Equine Vet Sci
1993; 13:502–505.
13. Gastal EL, Kot K, Ginther OJ. Ultrasound-guided intrafollicular treatment in mares. Theriogenology 1995; 44:1027–1037.
14. Kot K, Gibbons JR, Ginther OJ. A technique for intrafollicular injection in cattle: effects of hCG. Theriogenology 1995; 44:41–50.
15. Espey LL, Lipner H. Ovulation. In: Knobil E, Neill JD (eds.), The
Physiology of Reproduction, vol. 1, 2nd ed. New York: Raven Press;
1994: 725–780.
16. Nett TM, Holtan DW, Estergreen VL. Plasma estrogens in pregnant
and postpartum mares. J Anim Sci 1973; 37:962–970.
17. Nelson EM, Kiefer BL, Roser JF, Evans JW. Serum estradiol-17b
concentrations during spontaneous silent estrus and after prostaglandin
treatment in the mare. Theriogenology 1985; 23:241–262.
18. Daels PF, Ammon DC, Stabenfeldt GH, Liu IKM, Hughes JP, Lasley
BL. Urinary and plasma estrogen conjugates, estradiol and estrone
concentrations in nonpregnant and early pregnant mares. Theriogenology 1991; 35:1001–1017.