0013-7227/90/1275-2430$02.00/0 Endocrinology Copyright © 1990 by The Endocrine Society

Vol. 127, No. 5 Printed in U.S.A.

Developmental Changes in Steroidogenesis by Equine Preovulatory Follicles: Effects of Equine LH, FSH, and CG* J. SIROIS, T. L. KIMMICH, AND J. E. FORTUNE Section of Physiology in the Division of Biological Sciences and Department of Physiology in the College of Veterinary Medicine, Cornell University, Ithaca, New York

ABSTRACT. Ovulation in mares is preceded by a long and variable estrous period. The differentiation of equine preovulatory follicles with respect to steroidogenic capacity and responsiveness to equine gonadotropins was studied by culturing pieces of follicle wall (FW = theca + attached granulosa cells) from preovulatory follicles isolated during late diestrus (day 14 of cycle, n = 5 mares), early estrus (lst-2nd day of estrus, n = 6) or late estrus (4th or 5th day of estrus, n = 6). FW was cultured with or without equine LH, FSH, LH+FSH, or CG (10 or 100 ng/ml) and medium was collected and replaced at 3, 6, 12, 24, 48, and 72 h of culture. Follicular fluid from presumptive ovulatory follicles and medium from FW cultures were assayed for progesterone, androstenedione, and estradiol-17/3. The cumulative secretion of all three steroids after 72 h of culture was significantly lower in FW isolated during late diestrus (P < 0.05) as compared with early or late estrus. Maximal progesterone secretion was observed with FW from late estrus whereas maximal androstenedione and estradiol secretion in vitro occurred with FW from early estrus. In contrast to results obtained in vitro, concentrations of progesterone in follicular fluid were not different among stages of follicular development, and concentrations of androstenedione and estradiol in follicular fluid were

maximal in late estrous follicles. Equine gonadotropins had their greatest stimulatory effect on steroidogenesis with FW obtained during late diestrus. As compared with controls, the addition of LH, FSH, or LH+FSH (100 ng/ml) increased progesterone secretion by FW from late diestrus (48x, 64x, and 58X, respectively, P < 0.01), early estrus (24x, 32x, and 36x, P < 0.01) and late estrus (9X, 9x, and 9x, P < 0.01). Equine LH and FSH also increased androstenedione secretion by follicles obtained during diestrus and estrus. In contrast, estradiol secretion showed a more rapid loss in responsiveness to gonadotropin stimulation, with both LH and FSH stimulating estradiol secretion by FW from late diestrous follicles (P < 0.01), but neither stimulating FW from early or late estrous follicles. Overall, eCG was a less potent stimulator of steroidogenesis in vitro than LH and FSH. In conclusion, this study indicates that, in the mare, concentrations of steroids in follicular fluid do not precisely reflect changes in steroidogenic capabilities of presumptive ovulatory follicles during their final stages of development, and that the developmental transition from late diestrus to late estrus is characterized by a marked decrease in follicular responsiveness to gonadotropins. {Endocrinology 127: 2423-2430,1990)

T

HE ESTROUS cycle of mares is, in several respects, unique as compared with cycles of other large animal species. The follicular phase is characterized by a long period of estrus (4-6 days) and the diameter reached by the developing ovulatory follicle (40-45 mm) is considerably larger than in other species (1-3). Ovulation in the mare is limited only to a specific region of the ovary, the ovulation fossa, and is not triggered by a typical LH surge (4). In contrast to other mammalian species, LH concentrations increase progressively during estrus and peak on average 1 day after ovulation (4-6). The devel-

Received June 6, 1990. Address all correspondence and reprint requests to: Dr. J. E. Fortune, 823 Veterinary Research Tower, Cornell University, Ithaca, NY 14853. * This study was presented at the 23rd annual meeting of the Society for the Study of Reproduction (Knoxville, TN, 1990). This study was supported by a grant from the Harry M. Zweig Memorial Fund for Equine Research and by a fellowship from the Medical Research Council of Canada (J. Sirois).

opment and application of real-time B-mode ultrasonography has greatly contributed to the understanding of ovarian follicular development during the equine estrous cycle (7-10). The use of ultrasonography to follow the patterns of growth and regression of individual follicles > 15 mm has shown that follicular development occurs in waves, with the most frequent pattern being one wave per cycle (10). The average interval of time from follicular recruitment to ovulation is 14-15 days (10, 11). In contrast to other farm animal species, horses have been relatively resistant to attempts to efficiently manipulate their reproductive cycles, including synchronization of estrus/ovulation and induction of superovulation (1,12-17). Repeated efforts to improve the management of equine reproduction have emphasized the limited, and rather controversial, nature of our understanding of the hormonal and cellular dynamics associated with the development of the ovulatory follicle in mares. For example, the relative roles of theca interna and granulosa cells

2423

2424

OVARIAN FOLLICULAR STEROIDOGENESIS IN MARES

in the production of estradiol-17jS are still unclear (1826). An important limitation to some earlier studies is that they were performed on materials collected from animals of unknown reproductive history (ovaries collected at slaughterhouse). More recently, Fay and Douglas (27) showed that by day 14 of cycle (day 0 = day of ovulation) the presumptive ovulatory follicle has already been selected, as judged by increases in the number of LH receptors in the granulosa cell layer and in estradiol concentrations in follicular fluid. However, the effects of equine pituitary gonadotropins on follicular steroidogenesis in vitro have not been characterized. Also, the effects of equine chorionic gonadotropin (eCG), formerly called PMSG, on follicular steroidogenesis in mares have yet to be documented. This latter gonadotropin, detectable in the circulation between days 40 and 120 of pregnancy, is thought to play a role in the production of accessory corpora lutea by inducing ovulation and/or luteinization of recruited follicles (28, 29). Therefore, the first objective of the present study was to characterize in vitro the steroidogenic capabilities of presumptive ovulatory follicles at three stages of their development, including late diestrus, and early and late estrus. The second objective was to test the effects of equine LH, FSH, and CG on follicular steroidogenesis, and to determine if responsiveness to these gonadotropins changes as the ovulatory follicle matures toward ovulation.

Endo • 1990 Vol 127-No 5

by Vaughan (30). On one occasion (early estrous follicle), the presumptive ovulatory follicle ruptured during the procedure whereas in 16 of 17 mares intact follicles were collected. Once recovered, the ovary was immediately immersed in ice-cold Eagle's Modified Essential Medium (MEM, Gibco, Grand Island, NY) supplemented with penicillin (50 U/ml)—streptomycin (50 fig/m\, Gibco), L-glutamine (2.0 mM, Gibco) and nonessential amino acids (0.1 mM, Gibco). Dissection of the preouulatory follicle The presumptive preovulatory follicle was carefully dissected from the surrounding ovarian tissue with a scalpel. Then the follicular fluid was aspirated and stored at -20 C until assayed for progesterone, androstenedione, and estradiol-17/3. The follicle was cut in half, and both halves were transferred into a new Petri dish containing fresh medium. The theca interna of the selected follicle was characterized by a high degree of vascularization, as described previously for equine presumptive ovulatory follicles (27). In addition to its typical pattern of active follicular development before the surgery and its morphological appearance in vitro, elevated concentrations of estradiol- 17/8 in the follicular fluid were also used in retrospect to confirm that the selected follicle was healthy (27). Only one half of the follicle was used in the present study. Under a dissecting microscope, the theca externa and other surrounding tissues were dissected away from the theca interna using fine forceps. The theca interna and attached granulosa cells were subsequently referred to as a follicle wall (FW) preparation. The FW was cut into small pieces of equal size using iridectomy scissors.

Materials and Methods Animals and reproductive management

Incubation conditions and gonadotropin treatments

Seventeen standard bred mares, 2-10 years old and weighing approximately 375-450 kg, were used. They were kept outdoors and fed daily with hay and concentrates. Mares were teased with a vigorous pony stallion for detection of estrus. Ultrasound examinations of ovarian structures were made daily from day 8 of the cycle (day 0 = day of ovulation) using real-time Bmode ultrasonography, and the patterns of growth and regression of all follicles > 15 mm were determined as previously described (10).

Pieces of FW were incubated in multiwell plastic culture dishes (3 pieces/well; 24 wells per dish, Costar, Cambridge, MA) in 500-^1 defined media with or without gonadotropins. The amount of tissue present in each well represented l/64th (1.56%) of the whole follicle. The culture medium consisted of modified MEM (as described above) supplemented with cortisol (40 ng/ml; Sigma, St. Louis, MO), insulin (1 Mg/ml; Eli Lilly and Company, Indianapolis, IN), and transferrin (5 Mg/ml; Collaborative Research Inc., Waltham, MA). Equine gonadotropins were generously supplied by Dr. Harold Papkoff and included highly purified (> 99% pure) eLH (E98A, E263B; 31) and eCG (PM230GB; 32). The equine FSH preparation (E276B) used in this study contained 5-7% LH (Papkoff H, personal communication). Gonadotropin treatments included: 1) control (no hormone added), 2) LH (10 and 100 ng/ml), 3) FSH (10 and 100 ng/ml), 4) LH + FSH (100 ng/ml), and 5) CG (10 and 100 ng/ml). The gonadotropin concentrations used in culture were selected to mimic physiological conditions. Each treatment was applied in duplicate and the cultures were incubated at 37 C in a humidified incubator gassed with 95% air and 5% CO2. Media were collected and replaced at 3, 6, 12, 24, 48, and 72 h of culture and frozen at —20 C until assayed for steroid content.

Ovariectomies Mares were ovariectomized during three stages of the estrous cycle, i.e. either during late diestrus (day 14 of cycle, n = 5 mares), early estrus (1st or 2nd day of estrus, n = 6), or late estrus (4th or 5th day of estrus, n = 6). Identification of the ovary bearing the presumptive ovulatory follicle was based on the pattern of follicular development characterized by ultrasonography prior to the surgery (10). Neurolepanalgesia was induced during the surgery with a combination of xylazine (Rompun, Haver, Bayvet Division, TX; 0.66 mg/kg iv) and morphine (Morphine Sulfate Injection, Eli Lilly and Company, Indianapolis, IN; 0.66 mg/kg iv). Ovariectomies were performed via colpotomy using a chain ecraseur and a technique described

OVARIAN FOLLICULAR STEROIDOGENESIS IN MARES

2425

RIAs

3200

Nonextracted aliquots of culture media were assayed for progesterone, androstenedione, and estradiol-17jS using assays previously described (33). The progesterone antiserum used cross-reacts 13.6% with 5/?-dihydroprogesterone, < 6% with other progestins tested, and < 0.01% with the androgens and estrogens tested. The sensitivity of the assay is 12.5 pg per assay tube. The androstenedione antibody used was provided by D. T. Amstrong and cross-reacts < 1% with all steroids tested except for testosterone (2%) and 5a-androstane 3, 17dione (5%). The sensitivity of the assay is 12.5 pg per assay tube. The estradiol-17/3 antiserum used was provided by G. D. Niswender and cross-reacts 2.6% with estrone, 5.2% with estriol, and < 7% with other steroids tested (34). The sensitivity of the assay is 5 pg per assay tube. Intra-assay coefficients of variation for progesterone, androstenedione, and estradiol were 6.0%, 4.3%, and 8.8%, respectively, and interassay coefficients of variation were, in the same order, 14.9%, 13.0%, and 12.5%. Aliquots of follicular fluid samples were extracted twice with 2-ml diethyl ether and then assayed for steroid content as described above. Recovery rates for progesterone, androstenedione, and estradiol were 83%, 97%, and 89%, respectively.

2400

l

1600

800

PROGESTERONE ANDROSTENEDIONE

ESTRADIOL

FIG. 1. Concentrations of steroids (ng/ml, mean ± SEM) in follicular fluid from follicles isolated on day 14 of cycle (D14, late diestrus, n = 5 follicles), on the 1st or 2nd day of estrus (El-2, early estrus, n = 5 follicles), or on the 4th or 5th day of estrus (E4-5, late estrus, n = 6 follicles).

matured from late diestrus to early or late estrus (1681 ± 190, 2451 ± 280, and 3064 ±176 ng/ml, respectively, P < 0.05). Effect of stage of cycle on basal steroid secretion in vitro

Statistical analyses Steroid accumulation in the culture medium was first calculated per well, and then expressed per follicle. For each interval of culture (0-3, 3-6, 6-12, 12-24, 24-48, 48-72 h), data were expressed as rate of steroid accumulation (/^g/follicle/h) and were analyzed using a three-way, mixed-model analysis of variance (ANOVA) with experiment, time of culture, and treatment as factors. Two-way ANOVAs (stage of cycle and time of culture) were used to test the effect of stage of follicular development (i.e late diestrus, early and late estrus) on basal rate of steroid accumulation in vitro. For each stage of follicular development, two-way ANOVAs (experiment and treatment) were also performed to test the effect of gonadotropins on cumulative steroid secretion over 72 h of culture. One-way ANOVAs were used to test the effect of stage of cycle on steroid concentrations in follicular fluid. When ANOVAs indicated significant differences, Duncan's multiple range tests were used to compare individual means. When heterogeneity of variance was observed, data were transformed to logarithms before analysis.

Results Effect of stage of cycle on steroid concentrations in follicular fluid

Progesterone concentrations in follicular fluid were low at all three stages of follicular development and did not significantly differ among follicles isolated during late diestrus, early estrus, or late estrus (Fig. 1). Similarly, no significant differences were observed in follicular fluid androstenedione concentrations among stages of the cycle, although levels tended to increase with follicular development (Fig. 1). In contrast, estradiol-17,8 concentrations in follicular fluid increased as follicles

In general, the capacity of FW preparations to secrete progesterone in the absence of gonadotropins increased with follicular differentiation (Fig. 2, upper panel). After the first 3 h of culture, FW from late estrous follicles secreted more progesterone in vitro as compared to FW obtained during late diestrus (P < 0.05). After 12 h of culture, FW from late estrus secreted more progesterone

than FW from both earlier stages of follicular development (Fig. 2, upper panel, P < 0.01). However, a significant difference in progesterone secretion between FW isolated during early estrus vs. late diestrus was observed only after the first 24 h of culture (P < 0.05). Cumulative secretion in vitro of androstenedione and estradiol-17/3 followed similar patterns (Fig. 2, middle and lower panels). Accumulation of both steroids was the lowest with FW isolated during late diestrus; but, in contrast to progesterone secretion in vitro, maximal androstenedione and estradiol secretion occurred with early, and not late, estrous follicles. FW from early and late estrous follicles secreted more androstenedione than FW from late diestrus after the first 6 h (P < 0.05) and 24 h (P < 0.01) of culture, respectively. However, differences in androstenedione secretion between FW from early vs. late estrous follicles were not significant. In contrast, estradiol secretion by FW from early estrus was greater than by FW from late estrus during the first 12 h of culture (P < 0.05). Also, estradiol secretion by FW from early and late estrous follicles was significantly increased as compared to FW obtained during late diestrus during most peroids of culture (P < 0.05).

OVARIAN FOLLICULAR STEROIDOGENESIS IN MARES

2426 30-

t > E4-5

Progesterone

24-

1.2

Endo • 1990 Vol 127 «No 5

Late Diestrus

0.9

/

F100

18-

0.6 12-

LI 00

0.3

6-

o

2

0

24

48

• El-2 i D14 72

0.0 1.2

Androstenedione

CG100

0-3

3-6

6-12 12-24 24-48 48-72

Early Estrus

E1-2 0.9-I E4-5

0.6

'o V f

D14

e

0.3 0.0

E1-2

1.2

E4-5

0.9

0-3

3-6

Late Estrus

6-12 12-24 24-48 48-72 F100 L1OO

0.6

0

24

48

Hours of Culture FIG. 2. Effect of stage of cycle on cumulative secretion of progesterone, androstenedione and estradiol-17/3 (/ug/follicle, mean ± SEM) by follicle wall over 72 h of culture. Presumptive ovulatory follicles were isolated on day 14 of cycle (D14, late diestrus, n = 5 follicles), on the 1st or 2nd day of estrus (El-2, early estrus, n = 6 follicles), and on the 4th or 5th day of estrus (E4-5, late estrus, n = 6 follicles). SEM is not shown when smaller than symbol.

Effect of stage of cycle on responsiveness to gonadotropins in vitro Significant increases in steroid accumulation in culture were observed when gonadotropins were used at a concentration of 100 ng/ml (Figs. 3-5), but 10 ng/ml was rarely effective (data not shown). Overall, the stimulation obtained with a combination of LH and FSH (LH+FSH, 100 ng/ml) was not significantly different from that observed when LH or FSH was used alone (data not shown), and equine CG was less potent than eLH and eFSH in stimulating steroidogenesis in vitro (Figs. 3-5). The greatest stimulation of steroid secretion by gonadotropins over control cultures was observed when FW from the earliest stage of follicular development were used (late diestrous (day 14) follicles, Figs. 3-5, upper panels). At that stage of follicular development (day 14 of cycle), all three equine gonadotropins significantly

0-3

3-6

6-12 12-24 24-48 48-72

Hours of Culture FIG. 3. Rate of progesterone accumulation (^g/follicle/h, mean ± SEM) during six intervals of culture (0-3, 3-6, 6-12,12-24, 24-48, and 48-72 h) of follicle wall preparations. Presumptive ovulatory follicles were isolated during late diestrus (day 14 of cycle, n = 5 follicles, upper panel), early estrus (1st or 2nd day of estrus, n = 6 follicles, middle panel), and late estrus (4th or 5th day of estrus, n = 6 follicles, lower panel). Follicle wall preparations were incubated in medium alone (control, C), or in medium supplemented with 100 ng/ml of equine FSH (F100), LH (L100), or CG (CG100). SEM is not shown when smaller than symbol.

stimulated progesterone, androstenedione, and estradiol secretion over 72 h of culture. The maximal stimulation of progesterone secretion in vitro was obtained with eFSH, whereas androstenedione and estradiol accumulation were maximally stimulated by eLH. Equine CG consistently had the weakest stimulatory effect on steroid secretion by late diestrous follicles. The developmental transition from late diestrus to early estrus was characterized by marked decreases in responsiveness of FW to all three gonadotropins (Figs. 3-5, middle panels). Progesterone secretion was the least affected because eLH, eFSH, and eCG maintained a stimulatory effect on its secretion over 72 h of culture. Interestingly, during the transition from late diestrus to early estrus, LH became as potent as FSH in stimulating progesterone secretion in vitro (Fig. 3, comparison be-

OVARIAN FOLLICULAR STEROIDOGENESIS IN MARES

2427

1.6

1.6

Late Diestrus

Late Diestrus

F100

0.8

1.6

0.8

s: 0.4

0.4 0.0

1.2

LI 00

1.2

.2? o

0-3

3-6

6-12 12-24 24-48 48-72

=

0.0

Early Estrus

0-3

3-6

6-12 12-24 24-48 48-72

Early Estrus c .2 1.2

1.2 F100 LI 00

0.8

0.8

r-T

0.4

o a:

CG100

0.0 0-3

3-6

6-12 12-24 24-48 48-72

Q

1.6

Late Estrus

.1 CG1O0 - L100

0.4

CO LJ

o

0.0

0-3

3-6

6-12

12-24 24-48 48-72

1.6 Late Estrus

a> •*->

a or

1.2

1.2

0.8

0.8

0.4

0.4

0.0

0-3

3-6

6-12 12-24 24-48 48-72 Hours of Culture FIG. 4. Rate of androstenedione accumulation (/ig/follicle/h, mean ± SEM) during six intervals of culture (0-3, 3-6, 6-12, 12-24, 24-48, and 48-72 h) of follicle wall preparations. Presumptive ovulatory follicles were isolated during late diestrus (day 14 of cycle, n = 5 follicles, upper panel), early estrus (1st or 2nd day of estrus, n = 6 follicles, middle panel), and late estrus (4th or 5th day of estrus, n = 6 follicles, lower panel). Follicle wall preparations were incubated in medium alone (control, C), or in medium supplemented with 100 ng/ml of equine FSH (F100), LH (L100), or CG (CG100). SEM is not shown when smaller than symbol.

tween upper and middle panels). Equine LH and FSH, but not CG, had a stimulatory effect (P < 0.01) on androstenedione secretion by FW from early estrous follicles. Estradiol-170 secretion appeared to be the most affected by this developmental transition inasmuch as its cumulative secretion over 72 h of culture could no longer be stimulated by LH and FSH. However, eCG had a stimulatory effect on cumulative estradiol secretion by FW from early estrous follicle (P < 0.05). The subsequent transition from early estrus to late estrus was characterized by a further decrease in responsiveness to gonadotropins (Figs. 3-5, lower panels). As compared with follicles obtained during early estrus, the decreased stimulatory effect of all gonadotropins on progesterone secretion by late estrous follicles was accompanied by a significant increase in basal progesterone

I CG100

"A

L100 F100 C

0.0 0-3

3-6

6-12 12-24 24-48 48-72 Hours of Culture

FIG. 5. Rate of estradiol accumulation (jig/follicle/h, mean ± SEM) during six intervals of culture (0-3, 3-6, 6-12, 12-24, 24-48, and 48-72 h) of follicle wall preparations. Presumptive ovulatory follicles were isolated during late diestrus (day 14 of cycle, n = 5 follicles, upper panel), early estrus (1st or 2nd day of estrus, n = 6 follicles, middle panel), and late estrus (4th or 5th day of estrus, n = 6 follicles, lower panel). Follicle wall preparations were incubated in medium alone (control, C), or in medium supplemented with 100 ng/ml of equine FSH (F100), LH (L100) or CG (CG100). SEM is not shown when smaller than symbol.

secretion (P < 0.01, Fig. 3, lower panel). In contrast, androstenedione and estradiol secretion by late estrous follicles could no longer be stimulated by any gonadotropin in vitro (Figs. 4 and 5, lower panels). Table 1 summarizes the effects of all three gonadotropins on cumulative progesterone, androstenedione, and estradiol-17/3 secretion over 72 h of culture by follicles obtained during the three stages of follicular development, and their changes in responsiveness to gonadotropins during the transition from late diestrus to late estrus.

Discussion To our knowledge this study describes for the first time the developmental changes in basal steroidogenic capacities in vitro of equine preovulatory follicles, and

OVARIAN FOLLICULAR STEROIDOGENESIS IN MARES

2428

TABLE 1. Effects of equine LH, FSH, and CG on cumulative secretion of progesterone, androstenedione, and estradiol-17/3 by follicle wall preparations from preovulatory follicles isolated during late diestrus, and early and late estrus. Time of isolation of presumptive ()vulatory follicles" Progesterone* LH (100 ng/ml) FSH (100 ng/ml) LH + FSH (100 ng/ml) CG (100 ng/ml) A ndrostenedioneb LH (100 ng/ml) FSH (100 ng/ml) LH + FSH (100 ng/ml) CG (100 ng/ml)

Day 14

Estrus 1-2

48.4 Xc 64.1 Xc 58.4 xc 8.9 x c

24.0 Xc 31.9 x c 36.0 x c 8.3 xc

37.6 Xc 24.3 x c 36.3 Xc 5.9 x c

3.8 Xc 4.0 x c 2.9 Xd 2.3 x

1.4 X 2.1 x 1.9 X 1.5 x

1.3 X 1.2 X 1.2 X

1.2 1.1 1.0 1.3

Estrus 4-5 8.7 9.3 8.7 3.5

xc xc xc xc

Estradiol-17"Pb

LH (100 ng/ml) FSH (100 ng/ml) LH + FSH (100 ng/ml) CG (100 ng/ml)

11.5 8.1 6.9 6.0

xc xc xc xc

1.7 xc

X X x x

0

Follicle wall preparations were obtained from follicles isolated on day 14 of cycle (late diestrus, n = 5 follicles), on the 1st or 2nd day of estrus (Estrus 1-2, early estrus, n = 6 follicles) or on the 4th or 5th day of estrus (Estrus 4-5, late estrus, n = 6 follicles). 6 For each steroid, gonadotropin treatment, and stage of cycle, data are presented as ratios of the total steroid secretion over 72 h of culture in cultures supplemented with gonadotropins vs. respective control cultures (i.e. as fold increase (x) over control cultures). c Indicates a significant increase over control cultures (P < 0.01). d Indicates a significant increase over control cultures (P < 0.05).

changes in their responsiveness to equine gonadotropins as they mature toward ovulation. The concentrations of progesterone, androstenedione, and estradiol observed in follicular fluid are comparable to those previously reported for presumptive ovulatory follicles (27, 35-37). However, results from our experiments in vitro indicate that steroid levels in follicular fluid in mares may not accurately reflect the steroidogenic capabilities of developing ovulatory follicles. For example, although no differences were observed in follicular fluid progesterone levels between preovulatory follicles isolated during late diestrus vs. late estrus (Fig. 1), as also reported by Fay and Douglas (27), progesterone accumulation over 72 h of culture clearly showed that late estrous follicles have an increased ability to produce progesterone in vitro (Fig. 2). This finding is likely explained by the exposure of follicles isolated during the later stage to elevated LH levels in vivo (4-6). Androstenedione and estradiol concentrations in follicular fluid gradually increased with stage of follicular development to reach maximal levels in late estrous follicles. In contrast, secretion of both steroids in vitro indicated that follicles with the greatest capacities to produce estradiol and androstenedione were those collected on the 1st or 2nd day of estrus (Fig. 2).

Endo•1990 Vol 127 • No 5

Therefore, it can be hypothesized that the high levels of androstenedione and estradiol in follicular fluid observed in late estrous follicles may result from the accumulation of both steroids in the fluid during follicular development. A similar and progressive accumulation of androgens and estrogens in follicular fluid has been previously associated with the final maturation of preovulatory follicles during normal estrous cycles (27) and in hCGtreated mares (37). This is in contrast to observations on several other species including rats, humans, cattle, and sheep in which follicular fluid concentrations of androgens and/or estrogens have been shown to decrease (38-43), and progesterone to increase (39-42), as the ovulatory follicle approaches the time of ovulation, and most notably following the LH surge. Overall, all three equine gonadotropins had a stimulatory effect on follicular steroidogenesis in vitro during at least one stage of follicular development. Equine FSH (100 ng/ml) was the most potent stimulator of progesterone secretion by FW from late diestrous follicles, whereas LH (100 ng/ml) became as potent as FSH with preparations obtained from early estrous follicles (Fig. 3). This transition from late diestrus to early estrus has been associated in the mare with a tendency for an increase (27), and in other species with a significant increase (44, 45) in the number of LH receptors on granulosa cells. Because granulosa cells are thought to be the primary site of progesterone production in mare preovulatory follicles (24), an increase in LH receptors would likely be responsible for the enhanced responsiveness to LH of FW from early estrous follicles. The stimulation of FW androstenedione secretion by eFSH was unexpected because androgens are thought to derive from the theca interna, and thecal cells have LH, but not FSH, receptors (27). This stimulation of androgen secretion could have resulted from a 5-7% contamination of the eFSH preparation by eLH (i.e. 5-7 ng/ml of LH in culture medium containing 100 ng/ml of FSH; Papkoff H, personal communication). However, cultures treated with 10 ng/ml of eLH showed no increase in androstenedione secretion, as mentioned above. Alternatively, the stimulatory effect of FSH on progestin secretion by granulosa cells could have played a role by providing progestin precursors for thecal androgen synthesis, an interaction previously shown in bovine preovulatory follicles (46). Future studies on the effects of equine gonadotropins on separated cell types should clarify this point. Equine chorionic gonadotropin has been used in several species as a potent stimulator of ovarian follicular development, with the notable exception of the mare in which it has only very little, if any, stimulatory effect (2, 13, 14). Stewart and Allen (28) provided an explanation for this phenomenon when they showed that eCG does not appear to bind to the equine FSH receptor, and binds

OVARIAN FOLLICULAR STEROIDOGENESIS IN MARES

to equine LH receptors with about one tenth the affinity with which it binds to the LH receptors of rats, pigs, and cows. Results obtained in the present study are consistent with these previous observations. Overall, eCG was the weakest stimulator of follicular steroidogenesis in vitro, as compared with LH and FSH. Amino acid sequence analyses have recently revealed that the equine CG-/3 subunit is 100% identical to equine LH-/3 subunit (47, 48). This identity in the primary amino acid sequence of both hormones contrasts with the major differences that we and others (49) have observed in their respective effects in vitro. Interestingly, in a receptor binding study, Stewart and Allen (29) have shown that eCG had only 2.7% the affinity of eLH for equine follicular tissue. Known differences between eCG and eLH in their carbohydrate content (45 vs. 24%, respectively; 50, 51) could be responsible for differences in biological activities between the two glycoproteins. Changes in responsiveness of FW to equine gonadotropins were observed at two levels. First, a "vertical effect" was noted, within the first two stages of follicular development, along the steroidogenic pathway, consisting of decreasing stimulatory effects of gonadotropins on progesterone, androstenedione, and estradiol secretion (Table 1). However, this reduced effectiveness may be attributed in part to differences in respective basal steroid secretion, because an inverse increasing relationship was observed in basal steroid accumulation in vitro from progesterone to estradiol with late diestrous and early . estrous follicles. Second, a "horizontal effect" was noted for each steroid along stages of follicular development, with gonadotropins becoming decreasingly stimulatory as follicles were isolated from late diestrus to late estrus (Table 1). These results, along with the reported tendency for a decrease in LH receptors in presumptive ovulatory follicles isolated during late estrus (27), further support the concept of down-regulation, which likely results in the mare from the prolonged exposure of late estrous follicles to elevated levels of LH in vivo. When taken together, the marked increase in basal progesterone secretion and the decrease in basal androstenedione and estradiol secretion by FW isolated from late vs. early estrous follicles clearly suggest that follicular luteinization has been initiated on the 4th or 5th day of estrus (i.e. before the time when LH reaches maximal levels). These developmental changes in steroid synthesis are qualitatively similar to those induced by the LH surge in rats, humans, cattle, and sheep (38-43). Richards et al. (52, 53) showed that, in the rat preovulatory follicle, these changes can be explained by the effects of high levels of gonadotropins/cAMP in selectively turning on or off the expression of specific genes involved in rate limiting steps of ovarian follicular steroidogenesis. However, the mare differs from other species, in that lutein-

2429

ization and ovulation are not triggered by a typical LH surge, but rather by a progressive increase in LH levels during estrus (4-6). In summary, although concentrations of steroids in follicular fluid are good indicators of follicular health (27), this study shows that in the mare they do not mirror changes in steroidogenic capabilities of presumptive ovulatory follicles during their final stages of development. Overall, equine LH and FSH are more potent stimulators of follicular steroidogenesis than eCG, and the developmental transition from late diestrus to late estrus is characterized by a marked decrease in follicular responsiveness to gonadotropins. Considering the unique aspects of the mare's reproductive cycle, including a long follicular phase, the relative ease with which the presumptive ovulatory follicle can be identified up to 6-7 days before ovulation, and the large size of the follicle, this study suggests that the mare can offer unique advantages for studying the regulation of ovarian follicular steroidogenesis.

Acknowledgments We would like to thank Dr. B. A. Ball for his assistance with the in vivo aspect of this study, Dr. Harold Papkoff for providing purified equine gonadotropins, and Drs. D. T. Armstrong and G. D. Niswender for androstenedione and estradiol-17/3 antisera, respectively.

References 1. Palmer E 1978 Control of the oestrous cycle of the mare. J Reprod Fertil 54:495 2. Ginther OJ 1979 Reproductive biology of the mare: basic and applied aspects. Equiservices, Cross Plains, WI 3. Hughes JP, Stabenfeldt GH, Kennedy PC 1980 The estrous cycle and selected functional and pathological ovarian abnormalities in the mare. Vet Clin North Am (Large Anim Pract) 2:225 4. Whitmore HL, Wentworth BC, Ginther OJ 1973 Circulating concentrations of luteinizing hormone during estrous cycle of mares as determined by radioimmunoassay. Am J Vet Res 34:631 5. Stabenfeldt GH, Hughes JP, Evans JW, Geschwind II1975 Unique aspects of the reproductive cycle of the mare. J Reprod Fertil (Suppl) 23:155 6. Alexander S, Irvine CHG 1982 Radioimmunoassay and in vitro bioassay of serum LH throughout the equine oestrous cycle. J Reprod Fertil (Suppl) 32:253 7. Palmer E, Driancourt MA 1980 Use of ultrasonic echography in equine gynecology. Theriogenology 13:203 8. Pierson RA, Ginther OJ 1985 Ultrasonic evaluation of the preovulatory follicle in the mare. Theriogenology 24:359 9. Palmer E 1987 New results on follicular growth and ovulation in the mare. In: Roche JF and O'Callaghan DO (eds) Follicular growth and ovulation rate in farm animals. Martinus Nijhoff, Dordrecht, p237 10. Sirois J, Ball BA, Fortune JE 1989 Patterns of growth and regression of ovarian follicles during the oestrous cycle and after hemiovariectomy in mares. Equine Vet J (Suppl) 8:43 11. Driancourt MA, Palmer E 1984 Time of ovarian follicular recruitment in cyclic pony mares. Theriogenology 21:591 12. Douglas RH 1979 Review of induction of superovulation and embryo transfer in the equine. Theriogenology 11:33 13. Irvine CHG 1981 Endocrinology of the estrous cycle of the mare: applications to embryo transfer. Theriogenology 15:85 14. Allen WR 1982 Embryo transfer in the horse. In: Adams CE (ed)

2430

OVARIAN FOLLICULAR STEROIDOGENESIS IN MARES

Mammalian egg transfer. CRC Press, Florida, p 135 15. Woods GL, Scraba ST, Ginther OJ 1982 Prospects for induction of multiple ovulations and collection of multiple embryos in the mare. Theriogenology 17:61 16. Palmer E 1985 Recent attempts to improve synchronisation of ovulation and to induce superovulation in the mare. Equine Vet J (Suppl) 3:11 17. Squires EL, Garcia RH, Ginther OJ, Voss JL, Seidel GE 1986 Comparison of equine pituitary extract and follicle stimulating hormone for superovulating mares. Theriogenology 26:661 18. Short RV 1962 Steroids in the follicular fluid and the corpus luteum of the mare. A "two-cell type" theory of ovarian steroid synthesis. J Endocrinol 24:59 19. Short RV 1964 Ovarian steroid synthesis and secretion in vivo. Recent Prog Horm Res 20:303 20. Ryan KJ, Short RV 1965 Formation of estradiol by granulosa and theca cells of the equine ovarian follicle. Endocrinology 76:108 21. Channing C 1969 Studies on tissue culture of equine ovarian cell types: pathways of steroidogenesis. J Endocrinol 43:403 22. YoungLai EV, Short RV 1970 Pathways of steroid biosynthesis in the intact Graafian follicle of mares in oestrus. J Endocrinol 47:321 23. Mahajan DK, Samuels LT1974 The steroidogenic ability of various cell types of the equine ovary. Steroids 24:713 24. Hay MF, Allen WR, Lewis IM 1975 The distribution of A5-3/3hydroxysteroid dehydrogenase in the Graafian follicle of the mare. J Reprod Fertil (Suppl) 23:323 25. YoungLai EV, Jarrell JF 1983 Release of 3H2O from l/3,2j3[3H] androstenedione by equine granulosa cells. Acta Endocrinologica 104:227 26. Silberzahn P, Almahbobi Gh, Dehennin L, Merouane A 1985 Estrogen metabolites in equine ovarian follicles: gas chromatographic-mass spectrometric determinations in relation to follicular ultrastructure and progestin content. J Steroid Biochem 22:501 27. Fay JE, Douglas RH 1987 Changes in the thecal and granulosa cell LH and FSH receptor content associated with follicular fluid and peripheral plasma gonadotrophin and steroid hormone concentrations in preovulatory follicles of mares. J Reprod Fertil (Suppl) 35:169 28. Stewart F, Allen WR 1979 The binding of FSH, LH and PMSG to equine gonadal tissues. J Reprod Fertil (Suppl) 27:431 29. Stewart F, Allen WR 1981 Biological functions and receptor binding activities of equine chorionic gonadotrophins. J Reprod Fertil 62:527 30. Vaughan JT 1988 The female genital system. In: Oehme (ed), Textbook of large animal surgery. Williams and Wilkins, Baltimore, p 581 31. Matteri RL, Papkoff H, NG DA, Swedlow JR, Chang Y-S 1986 Isolation and characterization of three forms of luteinizing hormone from the pituitary gland of the horse. Biol Reprod 34:571 32. Matteri RL, Papkoff H, Murthy HMS, Roser JF, Chang Y-S 1986 Comparison of the properties of highly purified equine chorionic gonadotropin isolated from commercial concentrates of pregnant mare serum and endometrial cups. Domest Anim Endocrinol 3:39 33. Fortune JE, Eppig JJ 1979 Effects of gonadotropins on steroid secretion by infantile and juvenile mouse ovaries in vitro. Endocrinology 105:760 34. Korenman SG, Stevens RH, Carpenter LA, Robb M, Niswender GD, Sherman BM 1974 Estradiol radioimmunoassay without chromatography: procedure, validation, and normal values. J Clin Endocrinol Metab 38:718 35. Short RV 1961 Steroid concentrations in the follicular fluid of mares at various stages of the reproductive cycle. J Endocrinol

Endo • 1990 Vol 127-No 5

22:153 36. Kenney RM, Condon W, Ganjam VK, Channing C 1979 Morphological and biochemical correlates of equine ovarian follicles as a function of their state of viability or atresia. J Reprod Fertil (Suppl) 27:163 37. Watson ED, Hinrichs K 1988 Changes in the concentrations of steroids and prostaglandin F in preovulatory follicles of the mare after administration of hCG. J Reprod Fertil 84:557 38. Fujii T, Hoover DJ, Channing CP 1983 Changes in inhibin activity, and progesterone, oestrogen and androstenedione concentrations in rat follicular fluid throughout the oestrous cycle. J Reprod Fertil 69:307 39. Bomsel-Helmreich O, Gougeon A, Thebault A, Saltarelli D, Milgrom E, Frydman R, Papiernik E 1979 Healthy and atretic human follicles in the preovulatory phase: differences in evolution of follicular morphology and steroid content in follicular fluid. J Clin Endocrinol Metab 48:686 40. Dieleman SJ, Kruip Th AM, Fontijne P, de Jong WHR, van der Weyden GC 1983 Changes in oestradiol, progesterone, and testosterone concentrations in follicular fluid and in the micromorphology of preovulatory bovine follicles relative to the peak of luteinizing hormone. J Endocrinol 97:31 41. Fortune JE, Hansel W 1985 Concentrations of steroids and gonadotropins in follicular fluid from normal heifers and heifers primed for superovulation. Biol Reprod 32:1069 42. Fortune JE, Sirois J, Quirk SM 1988 The growth and differentiation of ovarian follicles during the bovine estrous cycle. Theriogenology 29:95 43. Webb R, England BG 1982 Relationship between LH receptor concentrations in thecal and granulosa cells and in vivo and in vitro steroid secretion by ovine follicles during the preovulatory period. J Reprod Fertil 66:169 44. Richards JS 1979. Hormonal control of ovarian follicular development: a 1978 perspective. Recent Prog Horm Res 35:343 45. Hansel W, Convey EM 1983. Physiology of the estrous cycle. J Anim Sci 57(Suppl 2):404 46. Fortune JE 1986 Bovine theca and granulosa cells interact to promote androgen production. Biol Reprod 35:292 47. Sugino H, Bousfield GR, Moore WT, Ward DN 1987 Structural studies on equine glycoprotein hormones. Amino acid sequence of equine chorionic gonadotropin /3-subunit. J Biol Chem 262:8603 48. Bousfield GR, Liu WK, Sugino H, Ward DN 1987 Structural studies on equine glycoprotein hormones. Amino acid sequence of equine lutropin /3-subunit. J Biol Chem 262:8610 49. Licht P, Gallo AB, Aggarwal BB, Farmer SW, Castelino JB, Papkoff H 1979 Biological and binding activities of equine pituitary gonadotrophins and pregnant mare serum gonadotrophin. J Endocrinol 83:311 50. Moore WJ Jr, Ward DN 1980 Pregnant mare serum gonadotropin. Rapid chromatographic procedures for the purification of intact hormone and isolation of subunits. J Biol Chem 255:6923 51. Landefeld TD, McShan WH 1974 Equine luteinizing hormone and its subunits. Isolation and physicochemical properties. Biochemistry 13:1389 52. Richards JS, Jahnsen T, Hedin L, Lifka J, Ratoosh S, Durica JM, Goldring NB 1987 Ovarian follicular development: from physiology to molecular biology. Recent Prog Horm Res 43:231 53. Richards JS, Hickey GJ, Oonk RB, Kurten RC, Gaddy-Kurten D, Wong W, Krasnow JS 1989 Diverse mechanisms controlling P450 enzymes and other proteins during ovarian follicular development and luteinization. In: Tsafriri A and Dekel N (eds) Follicular development and the ovulatory response. Serono Symposia, Rome, p211