ESI‘BUS, ovummn AND CHANGES IN SOME 30mm: DURING THE Esme-us CYCLE 0F MAKES AND AFTER PROSTAGLANDIN 112mgA A Disserioflon {or ”10 Degree of Db. D. MICEIGAN STATE UNIVERSITY Patricia Ann Noden 1975 MICHIGA 11111111111111 \Bllllllllll F g5 ;-. mu3 |1293 106145 ’3 . . M ........ - . This is to certify that the thesis entitled Estrus, Ovulation and Changes in some Hormones during the Estrous Cycle of Mares and after Prostaglandin an presented by Patricia Ann Noden has been accepted towards fulfillment of the requirements for Ph.D. Physiology 6 degree in , , Dalry Sc1ence Major professor Date April 30, 1975 0-7639 l L‘s \ o E rps‘r ABSTRACT r“ ) ‘ 31‘ E ESTRUS, OVULATION AND CHANGES IN SOME HORMONES ; all DURING THE ESTROUS CYCLE OF MARES AND AFTER PROSTAGLANDIN F20 By Patricia Ann Noden The principal purpose of this research was to determine whether PGFZa was luteolytic in horses, as had been reported in other species. Plasma progesterone was measured after intrauterine injection or subcutaneous injection of PGan as an indication of luteal function. In the first experiment, I measured duration of estrus, time of ovulation during estrus, and plasma progesterone, estrone, estradiol, androstenedione and LH during four consecutive estrous cycles in six or eight mares. The first and third cycles served as controls for the second which was initiated by 10 mg intrauterine (iu) PGFZa' The fourth cycle was initiated by 15 mg subcutaneous (sc) PGFZa‘ During a control estrous cycle the interestrual interval was 14.5 days, estrus persisted 5.2 days and ovulation was detected l.5 days before estrus ended. At 7 to 9 days after ovulation dur- ing a control cycle, 10 mg PGFZa-Tham salt was deposited in the uterus of six mares. After iu PGan, estrus began 2.2 days later; ovulation occurred 1.6 days before the end of estrus which per- sisted 7.5 days. These criteria measured during the subsequent Patricia Ann Noden control estrous cycle (cycle III) were similar to those measured during the control cycle which preceded PGFZa treatment; estrus persisted 5.6 days, ovulation occurred 1.8 days before the end of estrus. ' In all three cycles, during the 3 days before estrus began, progesterone decreased from values typical of diestrus (13 to l7 ng/ml) to values below l ng/ml when estrus began (P < .01); and then increased for 7 to 9 days after ovulation to diestrual values. In all three cycles, estradiol increased from low values (2 to 4 pg/ml) during diestrus to a peak (l0 to l7 pg/ml) l or 2 days before ovulation, and decreased to diestrual concentrations by the time ovulation occurred. Androstenedione also increased to a peak before the time of ovulation, but estrone did not change sig- nificantly throughout the three cycles. Plasma LH began to increase l day before estrus, had increased lO-fold (P < .Ol) by the time ovulation occurred 6 to 8 days later, and remained high until progesterone increased near the end of estrus. Then LH decreased gradually for 7 to 8 days after the end of estrus. When sc PGFZa (l5 mg) was given to eight mares 7 to 9 days after ovulation, estrus began 2.8 days later, persisted 7.8 days, and ovulation was detected 2.l days before the end of estrus. Hormone changes resembled those in the first three cycles. In summary, PGFZa was luteolytic in mares, whether given iu or sc, and ovulation occurred with relatively precise synchrony--at 8.0 i 0.6 and 8.0 t 1.0 days after PGFZa' Patricia Ann Noden In a second experiment designed to determine the minimal effective dose of PGFZa-Tham salt, mares were given, 2, 3, 5 or 10 mg PGan (sc) 7 to 9 days after ovulation. As determined by changes in blood progesterone and interval to onset of estrus, 2 or 3 mg PGFZa each caused luteolysis in‘3 of 4 mares, whereas 5 or 10 mg PGFZa caused luteolysis in all mares treated on day 7 to 9 after ovulation. Among the mares which responded to PGFZa? estrus began within 3.2 i 0.3 days and estrus persisted 6.9 i 0.4 days. Ovula- tion occurred 1.8 i 0.2 days before the end of estrus, about 8.3 i 0.4 days after PGFZa' Therefore, 5 mg PGFZQ appeared to be the minimal effective dose to cause luteolysis; 3 mg PGFZa caused luteolysis in some mares. A third experiment was designed to determine the time after ovulation when the corpus luteum (CL) becomes susceptible to lute- olysis by PGFZa' On day 1, 3, 5 or 7 after ovulation, mares were given 10 mg PGan-Tham salt (sc). As determined by changes in blood progesterone and interval to onset of estrus, PGFZa was luteolytic in none of five mares treated on day l, two of five on day 3, and in all mares treated on day 5 and day 7. Among the mares which responded to PGFZa’ estrus began 2.6 i 0.4 days after PGFZa and persisted 7.6 i 0.6 days. Ovulation occurred 1.9 i 0.3 days before the end of estrus--8.3 i 0.8 days after treatment with PGan. Therefore, while PGan induced luteolysis in some mares 3 days after ovulation, it appeared to be luteolytic in all mares by 5 days after ovulation. Patricia Ann Noden The luteolytic property of PGFZa was also tested in non- cycling mares with abnormally extended luteal function. In three mares which had exhibited estrus for an average of 35 days, PGFZa treatment caused 1uteolysis--estrus and ovulation followed.f Estrus began within 2.6 i 0.4 days after PGFZa and persisted 5.0 i 1.1 days; ovulation occurred 0.7 i 0.6 days before the end of estrus, 6.9 i 0.5 days after PGF2a_treatment. To test fertility, eight mares were mated during the estrus after PGan treatment. Six of the eight mares conceived. one aborted in early gestation; while in the remainder, the dura- tion of gestation was normal (332 days) and the foals were healthy. I conclude that PGan is luteolytic in mares, in sc doses 2.5 mg (Tham salt), when given at least 4 or 5 days after ovulation. The estrus following PGan does not differ from control cycles. PGan appears to have practical potential for control of estrus and synchronization of ovulation in mares. Possibly mares could be bred once about 7 or 8 days after PGan without regard to estrous behavior. ESTRUS. OVULATION AND CHANGES IN SOME HORMONES DURING THE ESTROUS CYCLE 0F MARES AND AFTER PROSTAGLANDIN an By Patricia Ann Noden A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of ’ DOCTOR OF PHILOSOPHY Departments of Physiology and Dairy Science 1975 To the mind-- it can think if you use it! ii BIOGRAPHICAL SKETCH of Patricia Ann Noden I was born in Los Angeles, California, on June 12, 1942, to Winifred and William Fail. My parents had migrated to California 5 years earlier to escape the results of dust bowl and depression in the Kansas farming community in which they had resided. However, in 1948, my parents purchased a farm in Wilson County, Kansas, and I began my grade school education in a one-roomed rural school. After completion of my eighth year, I entered Midway Consolidated High School near Buffalo, Kansas. I received an A.A. degree from Chanute Junior College in 1962 and a 8.3. and M.S. in Zoology from Kansas State University in 1964 and 1966. During the next year I worked with Dr. Webb Haymaker, a neuropathologist at the National Aeronautics and Space Administration Base at Moffett Field, Cali- fornia. Then during the next 4 years I taught biology, general science and general mathematics at Sanger High School near Fresno, California. In September of 1970 I enrolled at Michigan State Univer- sity in the Physiology Department. During this year I was accepted by Dr. Harold Hafs and the Department of Dairy Science to continue my studies for a Ph.D. degree. The past 4 1/2 years at Michigan State University have been rewarding. I have assisted in teaching classes in three departments and participated in research projects in four laboratories. These experiences and the association with Dr. Harold Hafs, his associates and fellow graduate students in the Animal Reproduction Laboratory have contributed immensely to not only my professional but also my personal maturation. The comple- tion of this thesis and the Ph.D. degree in June 1975 brings anticipation of what lies ahead: an opportunity to continue to improve as a teacher, a researcher and as a person. iv ACKNOWLEDGMENTS I would like to take this opportunity to thank the Depart- ments of Physiology and Dairy Science and their chairmen, Dr. Frances Haddy and Dr. Charles Lassiter, for providing financial support and facilities for my graduate studies. Also to be thanked for finan- cial assistance are the Michigan Agriculture Experiment Station and the Michigan State University General Funds.' I am deeply indebted to my major professor, Dr. Harold Hafs, for continued encouragement, support and direction during my gradu- ate studies. I especially appreciate his continued efforts to impart to me his optimism, enthusiasm and intellectual concern for teaching and reSearch and their application to practical situations. I also appreciate his help with technical procedures, statistical analysis, and seminar and publication preparation. I am also grate- ful for the advice and support of my committee members, Drs. S. 0. Aust, J. L. Gill, G. D. Riegle and H. A. Tucker. This thesis is a summation of the continued effort of several people in addition to myself. Dr. Wayne 0. Oxender pro- cured animals, directed and assisted with the rectal palpations, estrus behavior determinations and blood sampling. Additionally, the unselfish help and encouragement of Mary Vomachka is gratefully acknowledged. Rex Payne, Alan Swanson, Steve Wilson during 1972, and David Bolenbaugh and Cherolee Ricky during 1973 helped with collection of samples, organization of data and afforded encourage- ment. Providing additional assistance with RIA methods were Drs. Wayne Oxender, Lee Edgerton and Edward Convey. For technical assistance with RIA's, column chromatography and radioiodination of LH I wish to thank Candice Rankin, Janice Mapes, Patty Kaneshiro, David Jolly and Jim Byrne. For advice with statistical analysis, I acknowledge Dr. Roger Neitzel. Finally, I wish to thank my husband, Wayne Oxender, for continued encouragement and inspiration and in addition, for proof- reading, criticizing and commenting on the preliminary draft of this thesis; and it was he, who above all, understood. To my typist, Belinda Oxender, I extend deepest.appreciation for an excellent job of typing, and helping me "get it all together," and to Thomas Oxender my thanks for understanding there were times "when she has to study." To all of these people and to numerous others I say, "Thank you for your help and encouragement." Without them this thesis would not have been written. vi Table 10. 11. LIST OF TABLES Recovery of androstenedione added to 0.5 m1 mare plasma as determined by direct extraction or after chromatography with benzene: methanol (85: 15) on LH- 20 Sephadex . . Recovery of estrone or estradiol added to 1 ml mare plasma as determined by direct extraction or after chromatography with benzenezmethanol (85:15) on LH-20 Sephadex mini-columns Recovery of estrone added to 1 m1 of mare plasma with or without competition of added andros- tenedione and/or estradiol . . Relative activity of selected steroids in the radioimmunoassay for estradiol and estrone . Frequency of bleeding, teasing and rectal palpa- tion of reproductive organs in mares given PGFZG . . . . . . . Estrual behavior rating for mares during four con- secutive estrous cycles . . . . . Estrus and ovulation during two control cycles and after PGan (10 mg, iu) in six mares . . Plasma hormone concentrations during a control estrous cycle in six mares (cycle I) . . Plasma hormone concentrations during the estrous cycle following treatment with prostaglandin an (10 m , intrauterine) in six mares (cycle II? Plasma hormone concentration during a control estrous cycle after a cycle induced by PGFZQ in seven mares (cycle III) . . . . . Estrus and ovulation during a control cycle and after PGan (15 mg/sc) in eight mares ix Page 30 32 32 34 38 44 45 46 58 62 65 Table 12. 13. 14. 15. 16. 17. 18. Plasma hormone concentrations during the estrous cycle following treatment with PGan (15 mg, sc) in eight mares (cycle IV) . Estrus and ovulation in mares after varying doses of PGFZQ (sc) 7 to 9 days after ovulation Mean serum progesterone in mares after varying doses of PGFZa (sc) 7 to 9 days after ovulation . . . Estrus and ovulation in mares given 10 mg PGFZa (so) on various days after ovulation . Serum progesterone in mares after 10 mg PGFZG (sc) on various days after ovulation . Serum progesterone for mares in which PGFZQ did not result in estrus within 8 days of treatment . . . . . . Estrus and ovulation after PGF a in three mares with abnormally prolonged luEeal function Page 67 73 74 76 77 78 84 LIST OF FIGURES Radioimmunoassay of standard equine LH (LER 1138-1) and relative activities of equine FSH, bovine TSH and equine blood plasma . . . . . Plasma LH and thyroxine after thyrotropin releasing hormone (TRH) in two mares . Plasma pro esterone during the control estrous cycle (I) in six mares . . Plasma estradiol during the control estrous cycle (I) in six mares . . . . . . . . Plasma LH during the control estrous cycle (I) in six mares . . . . . . . . . . . Plasma androstenedione during the control estrous cycle (I) in six mares . . . Relative changes of plasma progesterone, LH and estradiol during the control estrous cycle (I) in six mares . . . . . . Plasma progesterone after PGan (10 mg, iu) in six mares . . . . . . Relative changes of plasma progesterone, LH and estradiol in mares during four consecutive estrous cycles . . . . xi Page 36 37 47 49 51 54 56 59 7O TABLE OF CONTENTS LIST OF TABLES . LIST OF FIGURES INTRODUCTION REVIEW OF LITERATURE . Reproductive Patterns of the Mare (Equus caballas) . Seasonal Changes The Estrous Cycle Pregnancy . . Anestrus . Attempts to Control the Estrous Cycle of Mares Gonadotropic Hormones . . Steroid Hormones Artificial Light . . Uterine Infusion With Saline . . Induction of Luteolysis With Prostaglandin an (PGFZQ) . . . . . . Utero- Ovarian Relationships . Luteolytic Agents PGFZa METHODS AND MATERIALS General Methods and Data Collection . Animals . Blood Collection and Sample Storage Estrual Behavior Rating . . Ovulation Detection Preparation of Treatments Radioimmunoassays (RIA) Experimental Design Luteolysis, Estrus, Ovulation and Plasma Hormones. After PGF a in Mares Estrus, Ovu ation and Serum Progesterone After 2, . 3, 5 or 10 mg of PGF a . . Estrus, Ovulation and erum Progesterone After PGF a on 1, 3, 5 or 7 Days After Ovulation Breed1ng, Pregnancy and Foaling After PGan Vii Page ix xi RESULTS AND DISCUSSION Luteolysis, Estrus, Ovulation and Plasma Hormones After PGFZQ in Mares . . Control Estrous Cycle (I) . Estrous Cycle After Intrauterine PGan (II) Control Estrous Cycle (111).. Estrous Cycle After Subcutaneous PGan (IV) Estrus, Ovulation and Serum Progesterone After 2, 3, 5 or 10 mg of PGan Estrus and Ovulation Progesterone . . Estrus, Ovulation and Serum Progesterone After 10 mg PGF a Administered on 1, 3, or 7 Days After Ovu ation . . . . . . . . . . Estrus and Ovulation Progesterone . . Breeding, Pregnancy and Foaling After PGFZG Side Effects of PGF . . . Abnormally Prolonged0L Luteal Function GENERAL DISCUSSION Hormonal Interactions During the Estrous Cycles in Mares . . . . Mechanism of Action of PGFZQ Practical Applications CONCLUSIONS . Luteolytic Effects of PGFZQ in Mares LITERATURE CITED viii Page 86 94 98 101 101 104 INTRODUCTION "The notorious infertility of mares has raised the question of the relationship of the time of service to ovulation" (Asdell, 1946). It is now generally agreed that an insemination or mating which occurs no more than 24 hours prior to ovulation is more fertile than those earlier or later in the estrual period. To ensure this requires either repetitive breeding (or insemination) throughout estrus, or accurate prediction of ovulation time. How- ever, the time of ovulation is difficult to predict in the mare, since it occurs sometime during a 4- to 10-day estrus. Unfortu- nately, the interval from onset of estrus to ovulation is highly variable while the interval from ovulation to the end of estrus is relatively constant. In addition, estrus in the mare is difficult to detect except in the presence of a stallion. I The prolonged estrus and variable time of ovulation are not the only problems associated with reproductive efficiency and conception in the mare. In addition, the mare is a seasonal breeder, with an anovulatory period during the fall and winter. This limits opportunity for conception. Furthermore, fertility may be decreased by physiological factors common in other species as well as in mares, factors such as ovarian dysfunction or uterine infection. Finally, owing to the birthdate requirements of horse breed associations, the most desirable breeding period is as near after the first of the year as possible, a time when most mares have just begun to cycle and breedings result in a low fertility. Because of the low fertility in mares, many researchers have attempted to control the time of ovulation. Gonadotropic hormones which stimulate follicular growth and ovulation, steroid hormones which delay or hasten the onset of estrus and changes in the photoperiod have been used with moderate success, but precise synchronization of ovulation had not been achieved. Prostaglandin F2“ (PGFZQ), a ZO-carbon fatty acid, caused luteolysis (i.e., the death or demise of the corpus luteum) in rats (Pharris and Wyngarden, 1969), sheep (McCracken et al., 1972) and heifers (Louis et al., 1972). Ovulation occurred with reasonable synchrony (99 i 12 hours) after PGFZa treatment in cows (Louis et al., 1972). If PGan induced luteolysis, estrus and synchronous ovulation in horses, it could contribute materially to efficient breeding management of mares. The major objective of this study was to determine if PGan was luteolytic in mares. I compared the luteolysis after deposit (If PGFZQ in the uterus (iu) with that of various doses of PGF injected subcutaneously (sc) during diestrus. Also, I 201 determined how soon following ovulation PGan would cause 1ute- olysis. Furthermore. to determine if endocrine events associated with the estrous cycle differed from normal after PGan-induced luteolysis, I determined blood progesterone, estradiol, estrone, androstenedione and LH, and I estimated the duration of estrus and the time of ovulation in mares given PGFZa and in control mares. REVIEW OF LITERATURE Reproductive Patterns of the Mare (Egyus caballas) Seasonal Changes Walter Heape was one of the first (1900) to state that the mare is a "polyoestrous animal with a tendency toward monoestrum." Eckstein and Zuckerman (1956b) reviewed the evidence in wild species of the genus Eggg§_and concluded that "The breeding season lasts from April to September, ovulation apparently being limited to the period May-August." Observations by John Hammond in 1936 indicated that the "breeding season" was a function of degrees latitude, being less marked nearer the equator than in Canada and Britain (Eckstein and Zuckerman, 1956b). Marshall (1936) concluded that "The results as a whole undoubtedly suggest a correlation between the sexual season and the incidence of daylight," because seasonal breeders (ewes, mares) transferred from the southern hemisphere to the northern hemisphere exhibited two "seasonal sea- sons" during one calendar year. For example, a pony mare after having exhibited normal estrous cycles in the southern hemisphere during November to February was then transported to Scotland (56°N). This mare adjusted to the increasing ambient photoperiod and exhibited cyclic estrual activity--as did her new herdmates--in the spring. This same phenomena was observed when ewes, ferrets and red deer were transferred from the southern to the northern hemisphere (Marshall, 1936; Eckstein and Zuckerman, 1956b). 3 Thus, depending on environmental conditions, there is a period of ovarian inactivity sometimes termed the “winter anestrum" in the majority of mares. In Michigan, this period begins in September and continues until April in most mares. In Kentucky, 75 to 85% of mares showed signs of estrus and ovulated about every 19 to 22 days from April through October, but only 20 to 25% of the mares showed evidence of ovarian cyclicity as early as January and February (Hughes et al., 1972) Nishakawa (1959) concluded that on the average, mares in Japan (similar in latitude to Kentucky) exhibited regular estrous cycles from April to October. Among mares in the southern hemisphere, 75% or more ovulated cyclically between mid-November to mid-April, and the lowest incidence of ovulation was from mid-June to mid-September (Osborne, 1966; Van Niekerk, 1967). The Estrous Cycle During the time of the year when the majority of mares were exhibiting regular estrous cycles, the estrous cycles ranged from 19 to 23 days and the duration of estrus ranged between 4 and 9 days (Asdell, 1946). Andrews and McKenzie (1941) reported an average estrous cycle of 20.7 days and 5.3 days duration of estrus in 45 draft and 38 light mares over the breeding seasons of 2 years. Data on duration of the estrous cycle and duration of estrus were similar for horses and ponies in Japan (Nishakawa, 1959). Attempts to characterize vaginal, cervical and external genitalial changes during estrus in the mare indicated most were not reliable predictors of estrus. Color changes in the labia and vaginal mucosa occurred, but they were extremely variable. The vaginal and cervical mucosa were highly vascular near the time of ovulation, and secretions from these tissues were most abundant during estrus. Changes in vaginal cell types did not correlate consistently with stages of the estrous cycle; however, some cornification of vaginal surface epithelium occurred during estrus presumably due to increased estrogens (Andrews and McKenzie, 1941). Nishakawa (1959) demonstrated changes in viscosity of cervical mucus throughout the cycle, but like Andrews and McKenzie (1941) found no consistent changes in vaginal cytology. A more reliable predictor of estrus is the muscular tone and position of the cervix in the vaginal canal. During estrus the cervix became flaccid and endematous, and the cervical lumen was easily penetrated. During diestrus the cervix acquired more muscu- lar tone tightly sealing the cervical lumen, and extended as a noticeable protrusion into the vagina. In addition, the vascu- larity of the cervix decreased during diestrus, resulting in detectable color differences (Andrews and McKenzie, 1941). The cervical changes in color and muscular tone during the estrous cycle could be used to detect the phase of estrus (or diestrus) in mares but they were not specific enough to predict when ovulation would occur. Even if these criteria accurately reflected stage of estrous cycle, collection of cervical mucus, vaginal cytology and observa- tion of cervical color or tone would require frequent use of a vaginal speculum. The technical difficulty of the above procedures coupled with the increased hazard of uterine infection following use of the vaginal speculum limit their usefulness as monitors of stage of estrous cycle. In addition, these changes were not suf- ficiently specific to predict ovulation. However, estrus behavior can be detected by use of a stallion. Estrus is the only portion of the cycle during which mares respond positively to the attentions of a stallion. Posi- tive responses to the "teaser" stallion included elevation of the tail, spreading the rear legs and squatting. This characteristic stance was accompanied by "winking" (eversion of the clitoris through the labia) and frequent urination (Andrews and McKenzie, 1941). Nishakawa (1959) used a five-point scale to rate intensity of estrual behavior during estrus, while Andrews and McKenzie (1941) rated sexual behavior during the entire cycle on a six- point scale. Both authors reported an intensification of sexual receptivity before ovulation. Anatomically, the ovary of the mare differs from that of other species in two important ways. Firstly, the ovary is cov- ered by by a fibrous capsule except for the ovulation fossa, a groove in the ventral surface of the ovary. Secondly, it is through this groove that follicular rupture occurs (Asdell, 1946). The small follicles distributed throughout the ovarian stroma migrated to the ovulation fossa as they grew (Eckstein and Zucker- man, l956a). After ovulation, the ruptured follicle collapsed, filled with blood to form a corpus hemorrhagicum and the corpus luteum began to form from granulosa cells. The corpus luteum (CL), was mushroom-shaped with the stem turned toward the ovulation fossa. Since the CL was only about 50 to 75% the size of the follicle and buried in the stroma of the ovary, it was difficult . to detect via rectal palpation (Day, 1940; Andrews and McKenzie, 1941; Asdell, 1946). Follicles are detectable (l to 2 cm) via rectal palpation around the onset of estrus and increase to a maximum size (5 to 6 cm diameter) just before ovulation. Andrews and McKenzie (1941) palpated follicles ranging from 1.0 cm to 7.5 cm during estrus; those of draft mares were larger than those in light mares. Fol- licles on the left ovary were larger than those on the right, probably because about 60% of the ovulations occurred from the left ovary. 'Double ovulations were infrequent (3.8%). Although somewhat smaller than follicles of mares, the follicles of ponies also increased in size as estrus progressed (Nishakawa, 1959). Nishakawa (1959) reported that 25% of the 107 ovulations recorded occurred 2 days before the end of estrus and 75% occurred 1 day before the end of estrus. In agreement, estrus ended 24 to 48 hours after ovulation in mares studied by Andrews and McKenzie (1941). Follicular fluid (Andrews and McKenzie, 1941; Short, 1961) and urine of mares (Loy, 1970; Dohn and Cole, 1971) contain estrogens. Urinary estrone increased from 20 ug/l to 60 pg/l dur- ing the week before ovulation, and urinary estradiol increased in smaller amounts from 9 ug/l to 18 pg/l (Loy, 1970). Mayer et a1. (1940) reported two urinary estrogen peaks; one during estrus and the other near the time of luteolysis 10 to 15 days after estrus. Asdell (1946) suggested the peak of estrogen which occurred during the luteal phase of the estrous cycle may cause luteolysis, since at about the same time, blood progesterone decreased from 10 to 20 ng/ml to concentrations below 1 ng/ml during estrus (Short, 1959; Smith et al., 1970; Stabenfeldt et al., 1972). Our- ' ing estrus, progesterone remained below 1 ng/ml until after ovula- tion; it increased significantly by 24 hours after ovulation and reached maximal values (10 ng/ml) by 6 days postovulation (Staben- feldt et al., 1972). Blood progesterone decreased beginning 3 days before the next estrus, reflecting corpus luteum regression. Follicular fluid of mares contained progesterone as well as estradiol and estrone (Short, 1960). Presumably, the granulosa and thecal cells secreted these steroids into the follicular fluid. Short (1960) also found other steroids in follicular fluid: cortisol, 17-a-hydroxy-progesterone, androstenedione, epitestos- terone and two unidentified compounds (an estrogen and a l7—keto- steroid). The concentration of estrogenic steroids in follicles did not change throughout the cycle, but the total estrogen increased as the follicles increased in size (Mayer et al., 1940). On the other hand, Knudsen and Velle (1961) measured 6- tolO-fold greater con- centrations rat pituitary or castrate rat serum > ovine, bovine, or porcine pituitary. Hartree et a1. (1968) and Reichart and Wilhelmi (1965) each isolated an equine LH fraction with potency of 0.73 units/ mg and 0.96 units/mg, respectively, when assayed by the ovarian ascorbic acid depletion method (OAAD) against NIH-LH-Sl (National Institutes of Health standard preparation of sheep LH). Equine follicle stimulating hormone (FSH) was isolated and purified by Saxena et a1. (1962) and Hartree et a1. (1968): relative potency measured 7 units/mg of NIH-FSH-Sl. That is, the FSH isolated was seven times as active as the NIH-FSH-Sl standard. Subse- quently, Braselton and McShan (1970) isolated a purer preparation of each hormone than those isolated by Hartree et a1. (1968) or Reichart and Wilhelmi (1965). Their LH preparation measured 5.3 units/mg NIH-LH-Sl and the FSH was equal to 90 units/mg NIH-FSH-Sl. They suggested that equine LH and FSH had higher molecular weights than LH and FSH molecules of other species. Equine LH also had a 5- to lO-fold higher sialic acid content than gonadotropins of other species (Landefield et al., 1972), as well as a longer biological half-life in rats (Parlow, 1963). The biological potency of glycoproteins as well as other plasma proteins partially depends on their sialic acid content. Desialylated human FSH and human chorionic gonadotropin (HCG) 11 had low biological potencies in vivo and were rapidly removed from the circulation and destroyed by the liver (Morrell et al., 1971). Therefore the high sialic acid content of equine LH may account for its higher potency than LH of other species in biological assays. To my knowledge, peripheral plasma concentrations of equine gonadotropins had not been reported prior to initiation of this project. Ovulatory surges of LH occurred coincidentally with onset of estrus in the cow (Swanson et al., 1972), during early estrus in the ewe and sow (Hansel and Echternkamp, 1972) and on the second day of estrus in the bitch (Boynes et al., 1972). The ovulatory surge of LH persisted for a period of less than 24 hours and was followed by ovulation in about 30 hours in the cow, sow and ewe (Hansel and Echternkamp, 1972). Pregnancy The duration of gestation in the mare varies among breeds from 329 to 345 days (Asdell, 1946). The first estrus period occurred about 1 week following parturition; sometimes called the "foal heat," it persisted an average of 11.4 days (Andrews and McKenzie, 1941). In addition to progesterone produced by the primary corpus luteum (CL), secondary ovulations and corpra lutea were formed during the first 70 days of pregnancy. These corpra lutea presumably also secrete twogesterone (Cole et al., 1931). Short (1959), using paper chromatography, found progesterone in the blood of pregnant mares during the first 90 days of gestation; 12 after this time it was nondetectable (< 4 ng/ml). However, the placenta of the mare produces ZO-a-hydroxypregn-4-en-3-one (Short, 1957). GonadotrOpic activity increased in pregnant mare serum (PMSG) beginning about 40 days postconception, reached a maximum between 50 and 80 days of gestation and was not detectable after 180 days (Eckstein and Zuckerman, 1956b). The endometrial cups of the mare produce PMSG (reviewed by Rowlands, 1963), but it is not of maternal origin. Allantochorion girdle cells of the fetal trophoblast stimulate endometrial cup formation and PMSG produc- tion (Allen and Moor, 1972). PMSG, a glycoprotein (58,000 M.W.) composed of two subunits (Gospodarowicz, 1972), had both LH and FSH activity throughout purification procedures (Raacke et al., 1957), and a high sialic acid content compared to other glyco- proteins (13.5%). PMSG probably stimulates accessory CL forma- tion (Cole et al., 1931) but this is supposition. Although the secondary corpra lutea formation and PMSG production both occur around 40 days of gestation, massive doses of PMSG did not stimu- late follicular development in mares (Nishakawa, 1959). In addition to PMSG and progestogens, the placenta of the pregnant mare also synthesized large quantities of equilin and equilienin (Girard et al., 1932). These are estrogens with an unsaturated 8 ring (Girard et al., 1932) as well as the saturated A ring found in ovarian estrogens. 13 Anestrus The term anestrus is a source of confusion in animals which are seasonal breeders. The seasonal anestrus for mares occurs during the short daylight period of the year (Hughes et al., 1972). Anestrus can also refer to conditions during the breeding season which cause failure of the estrous cycle. Irregularities of the cycle included (1) long periods of diestrus (termed anestrus by some), (2) irregular periods of estrus and (3) silent estrus (no estrus period detectable although ovulation occurred). Some causes of anestrus other than seasonal anestrous were (1) diseases such as ovarian hypoplasia, pituitary tumors or uterine infections, (2) pregnancy, (3) failure to exhibit estrual behavior ("silent" estrus) and (4) lactation (Roberts, 1971). Attempts to Control the Estrous Cycle of Mares The horse is increasingly popular for recreation in many countries; other countries require large numbers of horses for work. Therefore, recently, increased research efforts have been aimed to estrous cycle and/or ovulation control in mares. Unfor- tunately, until recently, attempts to control the estrous cycle with exogenous hormones, photoperiod control and uterine infusions of saline were only partially successful in the mare. Gonadotropic Hormones Induction of follicular growth and/or ovulation would be one means of estrous cycle control. Sequential injection of extracts containing FSH activity followed by LH caused ovulation 14 in rats (Casida, 1934). This experimental result was followed by attempts to induce ovulation in cows (Casida et al., 1943), in sheep and sows (Nishakawa, 1959), and in mares (Day, 1940, Nisha- kawa, 1959). PMS or FSH failed to induce follicular growth in the mare (Day, 1940; Bain, 1957; Nishakawa, 1959). However, ovulation in the mare can be hastened if follicles are already present. In mares treated with 1000 to 2000 R.U. (rat units) of human pregnancy urine at various times during estrus, ovulation occurred 20 to 40 hours later providing a follicle was present during treatment (Day, 1940). Davidson (1947) attempted to moni- tor maturity of the follicle; when the follicles were large and thin-walled he injected 500 IU of pregnancy urine to induce ovu- lation. Nineteen of the 23 barren mares (i.e., those that had failed to conceive and foal during 1 or more previous years) ovu- lated within 24 to 48 hours. Estrus ended 24 hours later. Although there were no control animals mentioned, Davidson (1947) apparently compared the duration of estrus in these mares to stable mates or to their prior records since he claimed the treatment shortened the duration of estrus. Not only did pregnancy urine induce ovulation during estrus but LH was also partially successful. Bain (1957) administered LH to 134 mares that had a history of ovulation failure and 65 to 80% ovulated. Similarly, the proportion of mares pregnant to breeding after induced ovulation (65%) was similar to the proportion of con— trol or treated mares pregnant after HCG treatment (Loy and Hughes, 1966). 15 In controlled experiments, administration of Prolan (a commercial preparation of urinary LH; Nishakawa, 1959) or human chorionic gonadotropin (HCG; Loy and Hughes, 1966; Ginther, 1971; Sullivan, 1972) to mares in early estrus hastened ovulation and reduced the duration of estrus. However, Sullivan (1972) indi- cated successive HCG treatment may have caused antibody formation since mares treated with HCG during three successive estrous periods did not ovulate as early following the third treatment as did control mares; because ovulation occurred near the end of estrus the treated mares then had increased duration of estrus com- pared to controls. Repetitive use of gonadotropic hormones to stimulate follicular growth and ovulation in mares will require research to obviate antihormone formation. Steroid Hormones Exogenous steroids have been used to attempt control of the estrous cycle in mares, since steroids may suppress or induce release of pituitary gonadotropins in mares as they do in other animals. Progesterone or progestogens administration may block estrus, follicular growth and/or ovulation; estrogens may cause estrual behavior and/or cause pituitary gonadotropin release resulting in follicular growth. Progesterone (100 or 400 mg in oil) injected every other day during the luteal phase of the cycle for 15 days blocked estrus and ovulation in 12 mares (Loy and Swan, 1966). Return to estrus following progesterone withdrawal was dependent on the 16 dose of progesterone and the season of the year. Small doses (50 mg/day) blocked estrus but not ovulation. Stabenfeldt et a1. (1972) detected-occasional ovulation during the luteal phase of the cycle in untreated mares, and suggested that progesterone had less influence in suppression of ovulation in the mare than in other animals. Moreover, oral progestogens in doses much higher than the effective dose in cattle or sheep did not block estrus or ovu- lation in mares (Loy and Swan, 1966; Britt and Ulberg, 1971). Use of progestogens to synchronize estrus in mares awaits the develop- ment of more efficacious compounds. In addition, attempts to induce estrus and ovulation with estrogens have failed. Although therapeutic use of estrogens to treat noncycling mares (Burkhart, 1947b; Bain, 1957) induced a receptive period, results were not clear since there were no observations on "normal" or control mares. In addition, the effect of estrogen treatment on follicle growth was confused with con- current administration of pregnancy urine to induce ovulation. Mares exhibited a receptive period after stilbestrol treatment during winter anestrus, but without follicular development and ovulation (Nishakawa, 1959). Artificial Light The "sexual season" of the equine female is correlated with increasing duration of daylight. Burkhart (1947a) placed a 1000 watt Sollux lamp above four mares to increase the photoperiod by 2 to 6 hours over the normal daylight period in Cambridge, England. 17 These anestrus mares began estrous cycles 38 to 64 days after the first addition of light on January 1. Control mares did not begin estrous cycles until 21 or more days later than the last of the treated mares. Similar observations were reported by Nishakawa (1959) who, in addition, suggested that the breeding season can be extended until January by addition of 5 hours of artificial light during August through December. Hormonal changes occurring at or before the first estrus in these mares are unknown. Uterine Infusion With Saline Irritation of the uterine endometrium modified luteal function in the cow and ewe (reviewed by Anderson et al., 1969) and irrigation of the uterus with saline has been used to treat noncycling mares in an attempt to reinitiate estrous cycles or increase "intensity of display" (Bain, 1957). For example, after intrauterine infusion of saline into 148 noncycling mares, estrus began within 2 to 12 days (Bain, 1957). Ginther (1971) confirmed Bain's observation: saline infused on the sixth or seventh day of diestrus was followed by onset of estrus in 4 days, signifi- cantly earlier than control mares. Induction of Luteolysis With Prostaglandin FZGTPGFZO) Utero-Ovarian Relationships Loeb (1923) demonstrated that the uterus was involved in the regulation of luteal function; hysterectomized guinea pigs had greatly prolonged estrous cycles. (Loeb, 1927). Removal of the 18 entire uterus during the luteal phase of the cycle caused main- taince of the [CL for a period longer than (ewe and sow) or nearly equal to pregnancy (cow; Anderson et al., 1969). If only a por- tion of the uterus remained, the cow continued to cycle about every 21 days; the sow required about one-quarter of one uterine horn and the guinea pig needed nearly one uterine horn to maintain nor- mal estrous cycles (Anderson et al., 1969). The influence of the uterus on luteal function in mares was substantiated by Ginther and First (1971). When mares were killed 30 days after hysterec- tomy, the CL marked with India ink at the time of hysterectomy was stillfunctional;meanwhile, the marked CL on the ovaries of con- trol mares regressed. In some species such as the guinea pig, cow, sow and ewe, the uterus exhibited a local control over the CL (Anderson et al., 1969). That is, the luteal function was controlled by the horn of the uterus nearest the ovary containing the CL. Removal of the uterine horn adjacent to the CL resulted in maintaince of the CL, 'whereas removal of the contralateral lunv1 had no effect in these species (Anderson et al., 1969). If one-half of the uterus (ewe and guinea pig) was removed and at the same time the adjacent ovary removed, cycling continued. However, if one uterine horn and the contralateral ovary were removed, luteal function was pro- longed (Anderson et al., 1969). In contrast, 2 of 5 mares with a unilateral hysterectbmy (on the side adjacent to the corpus luteum) continued to undergo normal estrous cycles; 3 of 5 mares with a corpus luteum on the ovary contralateral to the unilateral 19 hysterectomy continued to cycle (Ginther and First, 1971). This suggested that the corpus luteum of the mare is not controlled by local transfer of a substance from the uterus as in some species. Integrity of the uterine vein is essential to luteal regression: when the uterine vein was ligated on the side of the uterus adjacent to the CL, the estrous cycle was prolonged in ewes (Anderson et al., 1969). In contrast, ligation of the artery alone or of the artery and vein on the contralateral side of the uterus did not effect luteolysis. However, no direct link between the uterine vein and the blood supply to the corpus luteum has been found (Anderson et al., 1969). A counter-current transfer mechanism has been proposed (McCracken et al., 1972). In the ewe the ovarian artery is closely associated with the utero-ovarian vein, and sub- stances might be passed directly between them in some manner, since separation of the two vessels prevented luteolysis at the expected time (McCracken et al., 1972). Luteolytic Agents Insertion of foreign bodies and infusion of various toxic materials into the uterus as well as administration of oxytocin modified luteal function (Anderson et al., 1969). Insertion of an inert foreign body into the uterus may reduce or prolong the cycle depending on the stage of the cycle at the time of insertion. A glass or plastic bead, or plastic coil inserted into the uterus of a cow, ewe or guinea pig during the early part of the estrous cycle caused premature luteal regression (Anderson et al., 1969). 20 This premature regression was probably caused by a local luteolytic influence from the stimulated portion of the uterus. Irritation of the uterine endometrium by infusion of ethanol, water, iodine, phenol, silver nitrate or saline also modified luteal function (Anderson et al., 1969). Infusion of these compounds during the early luteal phase of the cycle resulted in an early return to estrus in cows, ewes and mares (Bain, 1957; Anderson et al., 1969). In contrast, infusion of these or other compounds later in diestrus caused prolonged luteal function. In the cow, oxytocin caused premature luteal regression. The duration of the estrous cycle was reduced following daily injection of large amounts of oxytocin during the first week after estrus (Armstrong and Hansel, 1959). Oxytocin treatment on day 15 to 22 or in hysterectomized cows did not alter the estrous cycle. Epinephrine and atropine blocked the luteolytic effects of oxytocin described above (Black and Duby, 1965). In contrast to these results in cows, oxytocin did not alter the estrous cycle in mares (Ginther, 1971). That the uterus produces a luteolytic factor which normally causes the demise or death of the CL is widely accepted. However, to my knowledge, the uterine 1uteolysin has not been isolated or identified. Active luteolytic materials isolated from uterine flushings of sows were thought to be high molecular weight protein; whereas active material in the ewe was thought to have a low molecular weight of approximately 1500 (Anderson et al., 1969). Anderson et a1. (1969) suggested that either the uterine 1uteolysin 21 had a short half-life, or was utilized by the target organ because of its local action in some species. PGan Discovered in the early 19305, in the semen of rams by Goldblatt and von Euler (reviewed by Oesterling et al., 1972), prostaglandins were purified and identified during the 19505 and 19605 in Sweden by Bergstrom, Samuelsson and van Dorp (Oesterling et al., 1972). They are unsaturated 20-carbon fatty acids, with a molecular weight of about 350. It was apparent very early that seminal extracts had powerful effects on the uterine and vascular smooth muscle. The relatively recent discovery, that PGan was luteolytic in pseudopregnant rats (Pharriss and Wyngarden, 1969), various laboratory animals (Pharriss et al., 1972) and sheep (McCracken, 1970) has prompted a renewed effort to find the uterine 1uteolysin. Pharriss et a1. (1972) reviewed the action of PGFZa on luteal function in various species and concluded that ”PGan has the qualities and effectiveness to be the long sought uterine 1uteo- lytic factor and fits the current hypothesis for the mechanism of uterine directed luteolysis." Several lines of evidence supported his conclusion. (1) PGFZa was luteolytic in species in which the uterus controls luteal function, such as guinea pigs, rabbits, rats and sheep. (2) Prostaglandin caused luteolysis in intact animals when administered systemically; vaginal, intravenous, subcutaneous and intracardiac as well as intrauterine treatments 22 were effective. I assume if PGFZa is the uterine 1uteolysin, it would also be luteolytic in hysterectomized animals. Local infu- sion of PGFZa into the ovarian artery and uterine vein of sheep caused luteolysis in doses lower than systemic treatment (McCracken et al., 1972). This may be because PGFZa was cleared rapidly from systemic circulation (Piper et al., 1970). (3) PGFZa was present in or released by the endometrium at a time when the uterus is known to induce luteolysis. PGFZa was released after distention of the uterus in guinea pigs and after estradiol administration. In addition to PGFZa release, both uterine distention and estra- diol treatment were luteolytic in guinea pigs; estradiol worked via the uterus since it was ineffective in hysterectomized guinea pigs (Blatchley et al., 1971). Bland et a1. (1971) reported increased PGFZa activity in the endometrium on day 14 in the ewe. An IUD inserted into the uterus increased PGan in the endometrium of the ewe; and caused luteolysis in sheep (Wilson et al., 1972). (4) Pregnancy should counteract the presence or action of PGan since pregnancy results in prolonged luteal function. Although the luteotropic factor of pregnancy is unknown, various gonado- tropins or luteotropic complexes can overcome PGan-induced luteolysis (Pharriss et al., 1972). (5) PGan caused biochemical changes (a shift from progesterone to 20a-hydroxy-pregn-4-en-3- one), as well as morphological changes (irregular shape, pycnotic nuclei, hypoplasia and atrophy of luteal cells) in corpora lutea of rats and hamsters (Pharriss et al., 1972). 23 All five of the above observations supported PGFZQ as a candidate for the uterine 1uteolysin. McCracken et a1. (1972) presented evidence that PGFZQ was the uterine 1uteolysin in the ewe; PGFZa caused luteolysis with a concomitant fall in proges- terone secretion in ewes. In addition, PGFZa was released from the uterus of the ewe in high concentrations during day 15 and 16 of the estrous cycle (McCracken et al., 1972) and furthermore, 3H-PGF2a was transferred from the uterine vein to the ovarian artery in support of the proposed counter-current transfer mechanism for the uterine 1uteolysin in ewes (McCracken et al., 1972). However, Lukaszewaska and Hansel (1970) and Caldwell et al. (1968) reported that the uterine 1uteolysin in bovine and porcine endometrial preparations was of a high molecular weight, probably a protein. This would eliminate PGFZa as the uterine 1uteolysin unless, as Shaw suggested to Pharriss et a1. (1972), prostaglandins bind to proteins. It is possible that the active fraction of the bovine endometrium was a protein bound PGan. These reports of the luteolytic action of PGan led to the discovery in our laboratory that intrauterine PGan treatment was followed by estrus at 72 hours and ovulation at 99 hours in cattle (Louis et al., 1972). If PGan would induce luteolysis, estrus and ovulation as precisely in mares as it did in cattle, it could obviate the most important breeding management problem in horses. Mares could be bred at a predetermined time after PGFZa . Conse- quently, the overall goal of this thesis was to test whether PGFZa was luteolytic and to characterize the luteolysis induced by PGFZQ 24 in mares. Since uterine control of luteal function operates systemically in mares rather than locally as in cows and ewes, part of this research compared luteolysis after systemic or local (uterine) administration of PGFZa' Secondly, because the response of the CL to various treatments depended on the age of the corpus luteum, part of the research involved treatment of mares with PGan at l, 3, 5 and 7 days after ovulation. Finally, I administered 2, 3, 5 or 10 mg of PGan at 7 days after ovulation to determine minimal effective dose. METHODS AND MATERIALS General Methods and Data Collection Animals We conducted these experiments during June, July and August of 1972 and 1973. The mixed-breed mares weighted 350 to 500 kg and ranged from 3 to 20 years of age. Except during periods of data collection, all mares grazed on 40 acres of pasture adjoin- ing the Michigan State University Veterinary Research Barn No. 3. A pony stallion, that served as "teaser,” was housed in a box stall in the barn. Blood Collection and Sample Storage Blood (40 ml) was collected by jugular puncture (16 gauge needle). During the 1972 project, blood was mixed with powdered oxalic acid (41 mg/tube) to prevent coagulation and placed on ice until transferred to the laboratory centrifuge l to 6 hours later. After centrifugation (3000 rpm for 30 min.), the plasma was decanted into a plastic vial and stored at -20°C. Blood was taken four times daily during estrus, and daily during diestrus. During 1973, the blood was allowed to coagulate, and after centrifugation the decanted serum stored at -20°C. Estrual Behavior Rating, The mares were tested once daily for estrus by exposure to a pony stallion (teasing). For each encounter, both the stallion 25 26 and the mare were rated on a 6-point scale similar to that used by Andrews and McKenzie (1941). Six was recorded for a mare if she stood unrestrained and allowed the stallion to mount (estrus). Winking, urination, tail raise and stance which accompanied estrus were recorded, but if these behavioral displays were not accompanied by standing to be mounted, the mare was rated 4 or 5 depending on the number of signs displayed. Kicking or striking, squealing and/ or biting and caudad orientation of the ears were considered nega— tive reactions and the mare was rated 1 or12 depending on the degree of these negative responses. A passive reaction toward the stallion (rated 3) was the absence of the above signs except occa- sional caudad orientation of the ears. Because the mares were teased once daily, the onset of estrus was considered to be 12 hours previous to first rating of 6, and the end of estrus was estimated to be 12 hours previous to the first rating less than 6. Ovulation Detection Follicular changes were monitored by palpation of the ovaries, daily during estrus or after PGan treatment, and every third day during diestrus. The diameter of follicles greater than 1.0 cm was estimated to the nearest 0.5 cm. Ovulation was (estimated to have occurred 12 hours previous to first detection of a collapsed follicle or corpus hemorrhagicum. Preparation of Treatments PGan-Tham salt (475.6 M.W.) was supplied in a powdered form by Dr. J. W. Lauderdale of the Upjohn Company. It was diluted 27 to the desired concentration in 0.85% saline and stored at -20°C until it was used. For intrauterine (iu) administration, 10 mg of PGFZa was thawed, aspirated into a sterile disposable 5 m1 syringe and injected through a sterile plastic insemination catheter which had been inserted asceptically through the cervix. Then 2 m1 of air was injected through the insemination catheter to force the remainder of the PGFZa into the uterus. In other experiments PGFZa was given subcutaneously (sc) in the neck. Because cows responded to 5 mg of PGan (Louis et al., 1972) deposited into the uterus, I decided to use 10 mg PGan for iu treatment in mares because of their larger size and the lack of local control of CL function by the uterus in mares. For 5c treat- ment, 15 mg of PGFZa was selected as the initial dose because 20 mg PGFZQ caused luteolysis in a 550 kg test mare. Radioimmunoassays (RIA) The antisera for my research, supplied by Dr. Gordon D. Niswender, Colorado State University, Fort Collins, Colorado, were tested for specificity, sensitivity and recovery of added mass. The steroid antisera were prepared in rabbits or sheep against a 68-succiny1 or 6-oxime steroid conjugated to bovine serum albumin. The assay for progesterone was described by Louis et a1. (1973) and the specificity of the progesterone antiserum (#869) was reported by Niswender (1973). To further validate the pro- gesterone assay for equine blood, 22 samples representing all stages of the estrous cycle were extracted with benzenezhexane 28 (1:2) and chromatographed on LH-20 Sephadex with chrloroform: ethanol (96:4) to isolate progestogens from estrogens and gluco- corticoids (Swanson et al., 1972). The progestogens were rechromatographed on another LH-20 Sephadex column with heptane: chloroform:ethanolzwater (200:200:1:saturated) to isolate pro- gesterone from other progestogens (Gwazdauskas et al., 1972). Progesterone eluted (fraction 8) well in advance of 17d-hydroxy- progesterone (fraction 12) and 20a-hydroxy-pregn-4-en-3-one (fraction 20) in the second solvent system. The RIA values for the chromatographically isolated progesterone were highly corre- lated (r = 0.98; P < 0.01) with values for the same samples deter- mined by RIA directly on benzenezhexane (1:2) extracts without chromatography (avg. 4.5 vs 4.6 ng/ml, respectively). In addition, extraction of tritiated progesterone from serum with benzene: hexane (1:2) was high (80 i 1%; Louis et al., 1973) compared with other tritiated steroids; estrone (56%), estradiol (27%), cortisol (1.4%) and corticosterone (8%) with the same solvent (n = 6 each). Therefore, progesterone was quantified in the blood of nonpregnant mares by RIA directly on solvent extracts without chromatographic isolation from other steroids. The androstenedione assay was described by Mongkonpunya et a1. (1975). Various steroids were tested for their ability to displace antibody (#866) bound tritiated androstenedione in this RIA; estradiol had 11% activity, estrone 1% and progesterone, cortisol and corticosterone less than 0.3%. Validation of the assay for mare blood paralleled that described above for 29 progesterone. Androgens were isolated in the 10th fraction from the first column (chloroformzethanol) and androstenedione eluted well before (fractions 5 to 7) dihydrotestosterone (fractions 14 to 16) and testosterone (fractions 17 to 19) on the second column (heptanezchloroform:ethanol:water). Values for 10 samples assayed after direct extraction with benzene:hexane (1:2) or after double column isolation were not highly correlated (r = -0.1) and the averages differed (P < .01, 160 i 17 vs 68 i pg/ml, respectively). Consequently, androstenedione was assayed in another 12 samples after direct extraction, after double column chromatography or after one chromatography on mini-columns. The mini-column (Thatcher, 1974) was a 2.5-ml disposable glass syringe containing LH-20 Sephadex with benzene:methanol (85:15; Carr et al., 1971) solvent. Androstenedione (0.2 to 1.2 m1) eluted prior to estrone (1.2 to 1.5 ml) and estradiol (3.0 to 5.0 ml). The average value for androstenedione in the 12 samples was 137 i 14, 94 i 15 and 126 .t 14 pg/ml, respectively, when measured after isolation by the three methods. Androstenedione values after mini-column isolation were highly correlated (r = .8, P < .01) with those after direct extraction, but values from the double column were not highly cor- related (r = .11, r = .04) with either of the other procedures. Therefore, recovery of various amounts of androstenedione added to mare blood plasma was determined after isolation from the mini- column and after direct extraction (Table 1). Androstenedione was highly recoverable by both methods (90.7% vs. 104.5%). Therefore, androstenedione was quanitified by RIA without chromatographic 30 TABLE l.--Recovery of androstenedione added to 0.5 m1 mare plasma as determined by direct extraction or after chromatog- raphy with benzenezmethanol (85:15) on LH-20 Sephadex. Isolation System Androstenedione Direct Mini- Extraction Column (pg) ----------- (pg/ml) ---------- 0 55 103 50 106 --- 100 149 173 200 --- 310 500 650 516 1000 1108 --- Overall % recoverya 104.5:4.6% 90.7:5.7% apg androstenedione observed pg androstenedione expected lated from the value when no androstenedione was added, plus the amount added to the mare serum. x 100%; expected was calcu- isolation in blood from the experimental mares. The cross reactions of estradiol (10.7%) or estrone (.8%) in the RIA should not have influenced the results, since quantities of estradiol and estrone are estimated to be only about 2 to 10% of androstenedione in mare blood. In 13 assays (n = 639 samples), an average of 84.4% of the androstenedione was extracted with benzene:hexane (1:2); RIA values were corrected for procedural loss of hormone. The estradiol (antiserum #825) assay was similar to that reported by Wettemann et a1. (1972) and Britt et a1. (1974). To validate the estradiol assay, estradiol in 18 samples of mare blood 31 plasma (representing all stages of the estrous cycle) were quan- tified by RIA after extraction with benzene (direct extraction) or after benzene extraction and chromatographic isolation on the mini-columns described for the androstenedione RIA. Estradiol aver- aged 44.7 t 0.8 pg/ml after direct extraction and 3.7 i 3.8% pg/ml after mini-column chromatography and results from the two methods were highly correlated (r = .91, P < .01). In addition, 87% of the estradiol added to 1 ml of mare serum was recovered by direct extraction whereas only 64% was recovered after mini-column chroma- tography (Table 2). Ten plasma samples were processed similarly to validate the estrone assay (antiserum #84; Britt et al., 1974) for mare plasma. Estrone averaged 9.5 pg/ml after direct extraction and only 4.9 pg/ml after mini-column chromatography, and the two meth- ods were not highly correlated (r = .28). Higher values after direct extraction may have been caused by failure to isolate estrone from other cross—reacting steroids; because RIA after direct extraction recovered more estrone (121 t 7.6%) added to 1 m1 of mare serum prior to quantification than RIA after chroma- tographic isolation (66.5 i 6.0%; Table 2). Substantial estrone was lost during mini-column chromatography. To test for interference of androstenedione and/or estra- diol on the estrone assay, estrone (0, 10 or 20 pg) was added to 1 ml of mare plasma (Table 3) to which estradiol (0 or 50 pg) and/or androstenedione (0 or 500 pg) had been added. Estradiol (50 pg) increased the estrone values by 2.1 to 3.6 pg. whereas 32 TABLE 2.--Rec0very of estrone or estradiol added to 1 ml mare plasma as determined by direct extraction or after chromatography with benzenezmethanol (85:15) on LH-20 Sephadex mini-columns. Estradiol Estrone Added Estrogen Direct Mini- Direct Mini- Extraction Column Extraction Column (p9) --------------------- (pg/m1) ---------------------- O 1.7 2.1 2.0 1.7 5 6.3 4.6 9.7 5.5 10 9.3 8.6 15.6 ‘ 7.6 20 19.7 12.9 24.6 14.5 50 --- --- 55.1 27.9 Overall % o o o recoverya 87.0i4.2% 64.0i3.8% 121i7.6% 66.5:6.0% aobserved x 100%. expected TABLE 3.--Recovery of estrone added to 1 ml of mare plasma with or without competition of added androstenedione and/0r estradiol. Picograms Added Estradiol (E2) or Androstenedione (A) Estrone Added 0 1:2 50 £2 0 £2 50 52 +0 A +0 A +500 A +500 A (99) --------------------- (pg/m1) ----------------------- 0 1.23 3.32 3.08 4.35 10 11.34 14.91 11.03 14.51 20 19.52 21.92 23.21 23.08 33 androstenedione inflated estrone values by 0 to 2.7 pg. However, the inflation of estrone values by estradiol and androstenedione would be maximal when the hormones are at maximum concentrations, since androstenedione and estradiol increased simultaneously in women (Moghissi et al., 1972; Abraham, 1974). Therfore, if these three hormones increase simultaneously in mares, peak values of estrone would be most inflated and estrone changes would therefore not be masked, but enhanced, by estradiol and/or androstenedione. Specificity of the estrone and estradiol RIA were accessed by determining competition of several different steroids with tritiated (3H) estradiol or estrone for their respective antisera (Table 4). Relative activity (RA), used to determine specificity, was calculated by comparing the amount of a compound which dis- placed a constant amount of the estradiol or estrone from their respective antisera. The estrone antisera was relatively specific for estrone; none of the hormones tested had more than 1.4% the relative activity (RA) of estrone (estradiol = 1.38%). However, the estradiol antisera bound testosterone (7.67% RA) and estrone (9.17% RA) in relatively large amounts. Androstenedione which is present in larger quantities than either estrogen in mare blood had 5.1.5% RA (Table 4) on either RIA. I decided to quantify estrone and estradiol by their respective RIA following direct extraction of the unknowns with benzene (without chromatography), because (1) estrogen recovery was reduced by column chromatogra- phy, (2) cross reaction of either RIA with other steroids extracted by benzene was negligible (benzene extracted 65% of added 34 TABLE 4.--Relative activitya of selected steroids in the radio- immunoassay for estradiol and estrone. Relative Activity in RIA for Crossreacting Steroid or Sterol Estradiol Estrone Estrone .0917 1.0000 Estradiol .0000 .0138b Cholesterol .0033c .0008c Corticosterone .0033C .0008C Cortisol .0033c .0008c progesterone .0033c .0008c 20-0-hydroxyprogesterone .0033c .0008C Androstenedione .0150 .0011b Testosterone .0767 .0008b aRelative activity = pg of estrogen at 50% displacement of pg of steroid at 50% displacement of 3H-estrogen bInterpolation from slope since 50% depression was not achieved with highest amount tested. 3H-estrogen c10,000 ng was the maximum amount of steroid tested; therefore 10,000 was used for the denominator in the above frac- tion although 50% depression was not achieved. 3H-androstenedione, 82% 3H-progesterone, 71% 3H-corticosterone) and (3) I expected relatively low plasma concentrations of unidentified estrogens or androgens in plasma from nonpregnant mares. Plasma LH was determined using a double antibody RIA similar to that described (Niswender et al., 1968). Purified ovine LH (LER-1056 CZ, supplied by Dr. L. E. Reichert, Emory University, Atlanta, Georgia) was iodinated (1251) and diluted to 20,000 cpm per 100 pl in 0.01 M phosphate buffered (pH 7.0) saline (PBS) with 35 0.1% Knox gelatin. Rabbit anti-ovine LH (GDN #15, supplied by Dr. Gordon Niswender, Colorado State University, Fort Collins, Colorado) was diluted 1:32,000 in 0.01 M PBS (pH 7.0) with 1:400 normal rabbit serum. Equine LH (LER 1138-1, 0.27 U NIH LH-Sl/mg, supplied by Dr. L. E. Reichert) was used as a standard, and PBS (pH 7.0) with 0.1% gelatin was used to dilute unknowns. Other- wise, details of this RIA for equine LH resembled those for the rat LH RIA (Niswender et al., 1968). The RIA values for LH in four dilutions (10 to 300 pl) of plasma from each of three mares were parallel to the standard curve (Figure 1). Ten to 100 ng of equine FSH (supplied by Mr. L. Nuti, University of Wisconsin, Madison, Wisconsin) and 25 to 300 ng of highly purified bovine TSH (supplied by Dr. J. Pierce, UCLA, Los Angeles, California) cross reacted 8% and 9%, respectively in the LH assay (Figure 1). To further test for TSH cross reactivity, two mares at about 190 days of pregnancy were given intravenously 100 pg thyrotrophin releasing hormone (TRH, supplied by Dr. R. Rippel,Abb0ttLabs., Chicago). Jugular blood was taken by venipuncture hourly for 6 hours, and plasma thyroxine was determined by the Tetrasorb method (Hernandez et al., 1972) as a reflection of serum TSH. After injection of TRH, serum thyroxine more than doubled at 60 min. (P < .01) and remained elevated for 6 hours (Figure 2, on page 37). Although serum thyroxine (and presumably TSH) was increased markedly after TRH, LH measured by my RIA was unchanged during this period. This observation indicates that the 9% cross reactivity of bovine TSH 36 Lu HDOr- (3 ch 0 EQUINE FSH :3 - “ .5 o 1— \ o m 80» ‘ fl _ . eova 1511 .5 \‘ 3 6° " \LASMA I D o ’ ‘\‘ m A» .A H 40— D 2! ‘\\\‘\\\ '5 F m '_ 20 _ EQUINE LH PLASMA 2 z w PLASMA 3 o .. a: E O l l l I I l l I 4 IO 20 4o 60 100 300 ng of HORMONE or pl BLOOD PLASMA Figure l.--Radi0immun0assay of standard equine LH (LER 1138-1) and relative activities of equine FSH, bovine TSH and equine blood plasma. in the LH RIA (Figure 1) may have been associated with LH contamina- tion in the bovine TSH preparation. Alternatively, the antiovine LH may cross react with bovine TSH and not equine TSH. Experimental Design Luteolysis, Estrus, Ovulation and Plasma Hormones after PGFZG in Mares In June 1972, six mares with normal ovaries and uterus were selected for the initial experiment to quantify luteolysis induced 37 05°F ;2 'Tlifl (ICND,zg iV) sE“* 1 A: L. Thyroxine -I Egao~ l a: _ ‘ ’ C: 0: 01.1120” .E—l x 1- 22' .: F-';J: - A -' 0 M— . a. . ’ 0.5 o 1 2 3 4 5 Hours from TRH Figure 2.--P1asma LH and thyroxine after thyrotropin releasing hormone (TRH) in two mares. by PGF Seven days after the ovulation which occurred during 20' estrus of a control cycle (cycle I), 10 mg of PGFZa-Tham salt in 1 ml of saline was deposited in the uterus. Estrus and ovula- tion followed PGan (cycle 11) and the mares were allowed a second control estrous cycle (cycle III). Then, 7 days after ovulation during the second control cycle, 15 mg of PGan was injected 5c in eight mares; blood sampling, palpation, and estrual behavioral observations were continued (Table 5) through the estrus which 38 TABLE 5.--Frequency of bleeding, teasing and rectal palpation of reproductive organs in mares given PGFZa' Cycle n Bleeding Teasing Palpation -------- (daily)------- (weekly) I Contro1 6 1-4X 1X 3-7X II ??52%g’ iu) 5 2-4x 1x 3-7x III Control 8 1-4X 1X 3-7X IV ??E2gg’ SC) 8 2-4x 1x 3-7x followed the sc administration of PGFZa (cycle IV). After each PGan administration, mares were observed for side effects. Statistical analysis of differences in intervals between cycles (e.g., duration of estrus, interval from ovulation to end of estrus, interval from onset of estrus to ovulation) and folli- cle size were analyzed using analysis of variance (ANOVA). For analysis of hormone changes various segments of the cycle were selected for analysis. For example, the decrease in blood pro— gesterone for 3 days after PGan in treated mares was compared with that during luteolysis in untreated mares by ANOVA. In addi- tion, values for all five hormones on 6 selected days of the cycle were compared by least squares ANOVA within and among cycles. The values used for analysis were from samples taken at (l) 5 days before estrus (cycle I and III) or on the day PGan was given (cycle II and IV), (2) onset of estrus, (3) 1 day before ovulation, 39 (4) day of ovulation, (5) end of estrus and (6) 7 days after ovula- tion (cycle I, II, III or IV) which was the day of PGFZa adminis- tration that began cycle II or IV. Several other analyses for specific changes in specific hormones will be described in the Results and Discussion section. Statistical analysis among all four cycles was not possi- ble, because although 6 mares were included in all four cycles, two additional mares were included in cycle III and IV that had not been treated with iu PGan. Thus some mares were nested within cycles III and IV, others were factorially related to all four cycles. Therefore, for purposes of statistical analysis the two parts of the experiment were considered separately. That is, for the first three cycles data for 6 mares were compared by ANOVA on selected days. Then a separate set of ANOVA was used to com- pare cycle III and IV for 8 mares each. Thus in each analysis mares were considered to be random effects; treatments and selected days were fixed effects. The lines in the ANOVA table were cycles (C), mares (M), selected days (0; where applicable) and the various two-way interactions (CM, CO, MD) and the three-way interaction (CMD). To test for treatment differences, the F ratio of the mean squares for C and CM was compared (C/CM) with critical values of the F distribution (Rohlf and Sokal, 1969). The difference between various days was tested by D/DM and the pattern of the hormone changes was tested by the F ratio of the mean squares CD/CMD. In an experiment such as this in which blood samples were taken sequentially the error may be correlated and not independently 4O distributed. This violation of the assumptions of the ANOVA (that the error is randomly and independently distributed) would result in overestimation of significance. Estrus, Ovulation and Serum Proges- teronefiAfter 2, 3, 5 or 10 mg of PGFZa A second experiment, started in June 1973, was an attempt to determine the minimal luteolytic dose of PGFZa' Mares were given (sc) 2, 3, 5 or 10 mg of PGan-Tham salt in 1 ml of saline at 7 to 9 days after ovulation. On the first day (after initi- ation of the project) they exhibited estrual behavior, the 12 nonpregnant mares available for this experiment were assigned to be given one of the various doses of PGan until each treatment was given to a minimum of four mares. For example, the first mare exhibited estrual behavior on June 6th; she was assigned to be given 5 mg PGan on the seventh day after ovulation was detec- ted. Two more mares exhibited estrus for the first time on June 15; they were assigned to be given 3 mg and 2 mg, respectively. The mares were not allowed control cycles between treatments for two reasons. There were only a small number of mares available for the experiment and it was necessary to confine the experiment to the months of June, July and August. One mare was given 3 treatments, five mares were given 2 treatments and six mares were given 1 treatment. Although the possibility exists of residual effects from one treatment to the next, it was not considered to be a major problem since there was no indication of residual 41 effects for the criteria measured h1the previous experiments (1972). The mares were teased daily, palpated daily during estrus and serum samples were taken daily beginning 1 day prior to PGFZa treatment and continuing until 3 days after treatment for RIA of serum pro- gesterone. Estrus, Ovulation and Serum Pro- gesterone After PGFZG on 1, 3, 5 or 7 Days After Ovulation A third experiment was designed to test the interaction between stage of CL development and luteolysis induced by PGan. Ten mg of PGan-Tham salt was administered to each ofW4 to 6 mares on day 1, 3, 5 or 7 after ovulation. On the first day the estrual behavior of a mare was rated a 6, she was assigned to a treatment group. The mares were given 10 mg PGFZa; the first on 1 day after ovulation, the second mare on day 3, the third on day 5 after ovulation et cetera until a minimum of four mares was treated on each of the selected days after ovulation. A total of 12 mares was treated; one mare was treated 3 times (three different treat- ments), seven mares were treated twice and four mares were treated once. No mare was included in the same treatment group twice. The mares were checked for estrual behavior once daily, palpated daily during estrus and 10 m1 of blood was drawn once daily from the day before PGan was administered until 3 days after treatment. In both the experiment to determine minimal luteolytic dose of PGFZa and in the experiment to test luteolysis induced by PGan at various days after ovulation, I used one-way ANOVA to 42 detect differences in intervals from (1) PGFZa to onset of estrus, (2) PGFZa to ovulation, (3) onset of estrus to ovulation, (4) ovulation to end of estrus and (5) onset of estrus to end of estrus. A two-way ANOVA designed to test repeat measurements (Gill and Hafs, 1972) was used to test for differences among treatments and the interaction of treatment and time. Breeding, Pregnancy and Foaling After PGFZQ To determine fertility of mares during estrus following sc PGFZa’ beginning on the second day of estrus eight mares were mated every other day to a fertile stallion until ovulation occurred. Pregnancy was detected by palpation at 30 to 60 days after breeding and by foaling. RESULTS AND DISCUSSION Luteolysis, Estrus, Ovulation and Plasma Hormones After PGFZa in Mares Control Estrous Cycle (I) Behavioral response of the mare to a teaser stallion during the first control cycle averaged 2.7 i 0.3 (on an arbitrary 6—point scale)!5days before onset of estrus (Table 6) indicating negative and/or passive responses. The numerical rating of respone increased to 6.0 i 0.0 on the day estrus began when the mares allowed the stallion to mount. During estrus all mares were given a rating of 6 (by definition), then on the day of ovu- lation one mare did not exhibit estrual behavior and the estrual behavior rating for six mares averaged 5.7 i 0.2. Behavior of the mares during mid-diestrus was characterized by aggressive hostility or passive reaction toward the teaser stallion. The average rating was 2.4 i 0.2 (Table 6). During the first control estrous cycle, estrus persisted 5.2 i 0.5 days; ovulation occurred 3.8 days after the onset of estrus or 1.5 days before the end of estrus (Table 7 on page 45). For two mares in which an earlier estrus had been observed the interestrual interval was 14.5 days (Table 7). Similar data for duration of estrus and time of ovulation during estrus were reported earlier (Andrews and McKenzie, 1940; Stabenfeldt et al., 1970). 43 44 TABLE 6.--Estrual behavior ratinga for mares during four consecu- tive estrous cycles. Cycle Day I II III IV Control iu PGF2 Control sc PGF2 Five days before estrus 2.7i0.3 --- 2.4i0.2 --- or on day of pGan --- 2.44 2 --- 3.0i0.5 Two days before estrus 3.6i0.4 --- 3.1i0.1 --- or one day after PGFZQ --- 2.4i0.2 --- 3.1iO.3 One day before estrus 4.7i0.4 --- 3.0i0.4 --- or two days after PGFZQ --- 4.5i0.7 --- 2.9i0.5 Onset of estrus 6.0i0.0 6.0i0.0 6.0i0.0 5.9i0.1 One day before ovulation 6.0i0.0 6.0i0.0 6.0i0.0 6.0i0.0 Ovulation detected 5.7i0.2 6.0i0.0 5.4i0.6 6.0i0.0 End of estrus 3.8i0.6 2.6i0.4 2.9:0.5 3.0:0.4 Mid-diestrus 2.4i0.2 2.5:0.2 3.0:0.5 2.7:0.3 aThe behavior of a mare exposed to a stallion was rated on a scale from 1 to 6. Negative responses were rated 1 if accompa- nied by kicking, squealing, and switching of the tail; passive responses were rated 3; and positive responses were rated 6 if the mare allowed the stallion to mount. bArbitrary units of estrual behavior. Mean :_SEM. (See page 25 for details.) 45 TABLE 7.--Estrus and ovulation during two control cycles and after PGFZQ (10 mg, iu) in six mares.a ~ Control PGFZQ Control Interval Cycle (I) Cycle (II) Cycle (111)b ---------------- (days)------------------- From PGFZQ to estrus --- 2.2:.3 --- From PGFZa to ovu- ___ ___ lation 8'0i'6 From onset estrus b to ovulation 3.8:.5 5.81.7 3.8:.4 From ovulation to end estrus 1.5i.5 1.6:.2 1.8:.6 Duration of estrus 5.2:.5b 7.5:.9 5.6:.8 Interestrual interval 14.5:2.5c 8.8:.9 14.2:.4 aEstrus and ovulation were estimated at 12 hours previous to time first detected. Values are means i S.E.M. bn = 5. cn = 2. Plasma progesterone averaged 17.1 i 2.3 ng/ml during diestrus (Table 8; Figure 3); it decreased from 9.5 i 2.7 ng/ml 3 days before onset of estrus to 0.7 i 0.2 ng/ml the first day of estrus (P < .01), and remained below 1 ng/ml until ovulation occurred. With formation of the corpus luteum, progesterone increased nearly linearly from 5.0 i 1.2 ng/ml the first day the mares did not exhibit estrus (defined as end of estrus in the following discussion) to 13.6 i 2.2 ng/ml about 7 to 9 days after ovulation (mid-diestrus). Similar changes of progesterone during the estrous cycle were reported in mares (Smith et al., 1970; 46 TABLE 8.--Plasma hormone concentrations during a control estrous cycle in six mares (cycle I). Plasma Hormone Day . Andros- Progesterone LH Estrad1ol Estrone tenedione ------- (ng/m1)------- -------------(pg/ml)------------ Five days a before 17.1i2.3 71:20 2.7il.l 11.012.7 190i40 estrus Two days before 5.3i1.9 102111 5.6i0.4 9.2iO.9 230:76 estrus One day before 1.4i0.2 236i75 5.9i0.9 10.9:2.1 147:20 estrus Onset of estrus 0.7i0.2 323178 7.1i1.7 9.7i0.9 182:30 One day 1 2:0 8 before (0'4,0'])d 661i113 11.512.7 12.5i2.3 380:73 ovulation ' ' ' Ovulation 1.9i1.2 End of + estrusb 5.0i1.2 806i124 4.8—0.9 ll.9i1.7 188:18 Mld' 13 5+2 2 123+35 4 4+0 8 10 0+1 1 207+35 diestrus ' ‘ ’ ‘ ° ‘ ' ° ‘ ° - aValues are means i S.E.M. b . . F1rst day mares were not 1n estrus. C7 to 9 days after ovulation; day of PGFZa administration for cycle II. dThese values omit data for one mare with abnormal proges- terone pattern during estrus (see page 83). 47 End estrus \ Progesterone (ng lml) .1 1 (g D-5 -3 -l «estrus-b l 3 5 Days from estrus Figure 3.--Plasma progesterone during the control estrous cycle (I) in six mares. Values (means i S.E.M.) are cen- tered on the onset of estrus and the end of estrus. Plotka et al., 1972; Stabenfeldt et al., 1972) and in pony mares (Squares et al., 1970; Sharp, 1972; Allen and Hadley, 1973). Sta- benfeldt et a1. (1972) suggested that the progesterone decline which began 3 days before onset of estrus (Figure 3) reflected the end of full luteal activity, and reported a significant increase in progesterone the day after ovulation. In the first control cycle in my mares, progesterone increased from 0.4 i 0.1 ng/ml approximately 12 hours before ovulation, to 2.7 i 1.3 ng/ml 48 36 hours after ovulation suggesting rapid formation of luteal tissue following ovulation. As blood progesterone declined during the 3 days before estrus, plasma estradiol increased 2-fold from 2.7 i 1.0 pg/ml 5 days before estrus to 5.6 i 0.4 pg/ml 2 days before estrus and to a maximum of 17.7 i 3.7 pg/ml (P < .01)2 days prior to ovula- tion (Table 8; Figure 4). Follicular growth accompanied increased estradiol production; follicles averaged 1.7 i 0.4 cm in diameter prior to estrus and increased to 4.2 i 0.2 cm the day ovulation was detected. Estradiol decreased to 4.8 i 0.9 pg/ml by the end of estrus and then ranged from 3 to 6 pg/ml until mid-diestrus. The peak in plasma estradiol 2 days before ovulation (Figure 4) is in agreement with the peak of urinary estrogen near ovulation in mares (Loy, 1970). Nett (1974) reported increased blood total estrogens during the foal estrus; Pattison et a1. (1974) obtained lO-fold higher values for blood estradiol, but the relative changes during the cycle resembed those in Figure 4. However, in marked constrast to estradiol, estrone did not change significantly throughout the cycle (Table 8). The lack of parallel change in estrone and estradiol con- centrations is not clear to me, particularly in light of the significant changes in urinary estrone concentrations in mares (Loy, 1970). Urinary estrone increased near ovulation in women (Moghissi et al., 1972); plasma estrone and estradiol both increased near estrus in cows (Hansel and Echterncamp, 1972; 49 OVULNHON lDETECTED Estradiol (pg/ml) 4,, G—ESTRUS—N» l l l l l I'Lil l I l 4L 1 l l J l -2 o —2 o 2 4 6 Days Figure 4.--P1asma estradiol during the control estrous cycle (I) in six mares. Values (means i S.E.M.) are centered on the onset of estrus (0) and the day ovulation was first detected (0). 50 Wettemann et al., 1972). Hansel and Echterncamp (1972) also reported peaks of estrone during diestrus, but these were highly variable. In contrast, others report that estrone did not change significantly during the estrous cycle of cows (Glencross et al., 1973). Adrenal production of estrone may have contributed to the plasma estrone pattern observed in my mares. The changes in urinary estrone throughout the estrus cycle may reflect metabolism of plasma estradiol, because the quantity of urinary estrone exceeded the estradiol in mares (Loy, 1970) and in humans (Moghissi et al., 1972); urinary estriol was even more abundant than estrone or estradiol in women (Moghissi et al., 1972). In contrast, Younglai (1970) and Knudsen and Velle (1961) found greater quantities of estradiol than estrone in follicular fluid of mares. Thus the ratio of estradiol to estrone apparently changes between the time estrogen is produced by the follicle and when it is excreted into the urine; estradiol predominates in the follicle and estrone predominates in urine. During mid-diestrus, LH ranged from 53 to 70 ng/ml (base- line); it increased (P < .01) by the day before estrus (236 i 75 ng/ml), and continued to increase to a broad peak of 800 to 1000 ng/ml which persisted until 2 days after ovulation (Figure 5). Then it decreased gradually during the next 7 or 8 days to 123 i 35 ng/ml when PGan was deposited into the uterus to begin cycle 11. The broad elevation of LH during estrus (Figure 5; Table 8) agrees with comparable data reported by Whitmore et a1. (1973) and Pattison et a1. (1974). In all three studies, blood 51 35C) I IZCX) ml:- (1‘ L 8 49' “1000 I. ...£ a 'J'. I ’II III: . A -'l: O ///’ 800F ‘-- Q 600* I; LH (ng LER ll38-l/ml ' "iii?” [/7 «égflhfizé; wag; sf”. // / (71 0 Pi 19 42 ..,’ / 1 .b O O 1 V3 57 / l Ovulation ' \‘ o 1 200 _ e ectled -1 I—Estrus --~>l ’_ 1 '15C) 0 l l l 1 ML I:#L 1 1 ML 1 1 1 l L l 1 -4 -2 O -2 O 2 4 6 8 Days from estrus Days from ovulation Figure 5.--Plasma LH during the control estrous cycle (I) in six mares. Values are centered an onset of estrus (0) and the day ovulation was first detected (1). The shaded area represents the S.E.M. around each plotted point. LH began to increase near the onset of estrus to maximal levels near ovulation and remained elevated until after ovulation; the subsequent decrease appeared nearly symmetrical to the increase. However, plasma LH values in this study are higher than the others. A major portion of this difference may be due to RIA differences; Whitemore et a1. (1973) used an anti-PMS first anti- body and an equine LH standard of approximately lO-fold higher potency than that used in our study. Pattison et a1. (1974) used the same anti-ovine LH used in the present study, but they J. o o 9 52 used ovine LH (LER-1056-C2) for a standard which is of higher potency than my equine LH standard. Consequently, the LH standards may account for most of the quantitative differences among these three reports. Sustained elevations of plasma LH throughout estrus with maximum concentrations after ovulation in mares is in contrast to the brief 8-hour surge of LH prior to ovulation in most other species. The LH surge occurs coincidentally with onset of estrus in the cow (Swanson et al., 1972; Hansel and Echternkamp, 1972), during early estrus in the ewe and saw (Hansel and Echternkamp, 1972), and on the second day of estrus in the bitch (Boynes et al., 1972; Smith and McDonald, 1974). The significance of the prolonged elevation of LH during estrus in mares is not apparent to me. That this may be an artifact of cross reaction with other pituitary hormones seems unlikely. The data in Figure 2 dis- counted any significant interference of TSH in the equine LH RIA, because while plasma thyroxine doubled within 1 hour after TRH injection, LH did not change. Furthermore, in addition to TSH release, TRH injections also released prolactin and growth hormone (GH) in several other species (Convey, 1973). If TRH released prolactin and/or GH in my mares, these hormones did not interfere with my RIA of LH because LH did not change after TRH injection. As outlined in the Review of Literature, the duration of estrus in mares is 2 to 5 times longer than in most other species. During estrus, the ovulatory follicle migrates through the ovarian stroma to the ovulation fossa where rupture occurs (Asdell, 1946). 53 It is possible that enzymes necessary for follicular migration and/or thinning of the follicular wall prior to ovulation may require prolonged LH stimulation. Rondell (1970) suggested that hydrolytic enzymes in the follicle wall require LH to trigger their ovulatory function. When the follicles of mares were manually ruptured on day 2 of estrus, luteinization did not occur whereas luteinization followed follicle rupture on day 5 (Asdell, 1946). This suggested that LH or a product produced in response to LH was not present in sufficient quantities to support luteinization before day 5 of estrus. Equine granulosa cells from large follicles luteinized when grown in vitro (Channing, 1969) whereas granulosa cells from small follicles did not, indicating changing metabolic capability of the granulosa cells as the follicle grew. Greater amounts of LH may bind to each equine granulosa cell in large follicles than to these cells in small follicles. Large granulosa cells of sows bind more HCG/cell than do granulosa cells of small follicles (Channing, 1974). Possibly LH directly or indirectly stimulates the number of LH receptors, thus affecting the ability of the granulosa cells to undergo luteinization. Equine LH has a slower disappearance rate than LH of other species when disappearance of exogenous LH is monitored in a horse (Ginther et al., 1974) or in a rat (Parlow, 1963). As discussed in the Review of Literature, the long half-life of equine LH is probably due to a high sialic acid content. The slow disappearance 54 rate of equine LH may in part contribute to the slow decrease in endogenous LH after ovulation in the mare. Plasma androstenedione (Figure 6; Table 8) increased (P < .05) from 182 i 30 pg/ml at onset of estrus to a peak of 380 i 73 pg/ml the day before ovulation, and decreased to 188 i 18 PQ/ml the first day of diestrus. Androstenedione was present in 3M3C1F- 340 F- 300 - 260 - OVULATION DETECTED ./ 220 - ANDROSTENEDIONE (pg/m1) BEGIN 180 ¢/ESTRUS / END ) esraus 01111,,L11fi,#1 111 <——55'raus m 1 3 DAYS FROM ESTRUS Figure 6.--Plasma androstenedione during the control estrus cycle (I) in six mares. Values (means i S.E.M.) are centered on onset of estrus, ovulation and end of estrus. 55 follicular fluid of mares (Short, 1960; 1961). Younglai (1971) reported highest concentrations of androstenedione in vascular follicles present during estrus. Similar to the peak of andros- tenedione before ovulation in mares (Figure 6), peripheral blood androstenedione peaked prior to ovulation in rats (Dupon and Kim, 1973) and in women (Abraham, 1974) where the ovary was the origin of 45 to 70% of the androstenedione. The function of androstene- dione in the female is unknown; perhaps it affects estrual behavior, follicular maturation or ovulation, or it may reflect estrogen synthesis. The reason for the changes in androstenedione before ovulation and the lack of changes in estrone concentrations are not apparent. The composite of the changes in progesterone, estradiol, and LH for control mares in Figure 7 indicates that progesterone decreased to near basal values before the LH surge began near the onset of estrus. Furthermore. the LH surge persisted until pro- gesterone began to increase at 1 or 2 days after ovulation. These observationssuggestincreasing progesterone may cause a decrease in LH release in mares. 0n the other hand, the LH decrease may be a delayed response to cessation of estradiol production several days earlier, since estradiol may stimulate LH production in mares. Estradiol increases pweceded or accompanied the increase in LH (Figure 7). However, neither LH nor estradiol increased until progesterone had decreased by 2.50%. These facts suggest that estradiol and progesterone both influence LH release in the mare. 56 ‘—__'”' ESTRUS w ,. - I200 gv'ula‘tiojn . e ec e 2° ’ 1 41000 ' Estradiol R ' ‘ 1 16 f ‘ ' \ LH . 800 i- ) V’A‘I I Progesterone 600 400 LH (ng LER ll38-l/rnI) 200 Progesterone (ng/ml) or Estradiol (pg/ml) days to estrus days from ovulation Figure 7.--Relative changes of plasma progesterone, LH and estra- diol during the control estrous cycle (I) in six mares. Values (means i S.E.M.) are centered on the onset of estrus (0) and the day ovulation was first detected (1). Estrous Cycle After Intra- uterine PGFZQ (II)— Seven to nine days after the ovulation which occurred dur- ing the first control estrus, 10 mg PGan-Tham salt was deposited into the uterus of each of six mares. An average of 2.2 days later, the mares exhibited signs of behavioral estrus (Table 6); and ovulation occurred 5.8 days after the onset of estrus (Table 7). Therefore after PGan, ovulation occurred 2 days later 57 (P < .05) after onset of estrus than in the preceding control estrus (Table 6). However, the interval from ovulation to the end of estrus (1.6 days) resembled that in the control estrus, and the duration of estrus after PGFZa was not significantly longer. The interestrual interval (8.8 days) was reduced 5 days (P < .01) by comparison with the previous interestrual interval (Table 7) owing to luteolysis precipitated prematurely by the PGFZa' Estrual behavior rating during this estrous cycle resembled those in cycle I (Table 6). Plasma progesterone averaged 13.6 ng/ml when PGan was infused into the uterus (Table 9; Figure 8,cn1page 59). Proges- terone decreased to 5.8 ng/ml within 12 hours (P < .01), and the decrease continued to 2.6 ng/ml and 0.9 ng/ml by 24 and 48 hours (Figure 8). The precipitous decrease in progesterone after PGan resembled that after PGFZa in heifers (Louis et al., 1973) and although it appeared to be more rapid, the decrease in proges- terone for 3 days after PGan did not differ significantly from that during the 3 days prior to estrus during normal luteolysis in mares (Figures 3 and 8). Progesterone in blood plasma increased rapidly after ovulation averaging 4.1, 11.5 and 13.7 ng/ml on days 1, 4 and 6 after estrus as in the control cycle (Figure 3). Similar to changes after normal luteolysis, plasma estra- diol (P < .06) and androstenedione (P < .01) increased during the estrus after PGan—induced luteolysis to maximum values before ovulation and had begun to decrease by the time ovulation was 58 TABLE 9.--Plasma hormone concentrations during the estrous cycle following treatment with prostaglandin F2“ (10 mg, intrauterine) in six mares (cycle II). Plasma Hormone Day . Andros- Progesterone LH Estrad1ol Estrone tenedione ------- (ng/ml)------- -------------(pg/ml)------------- On day of a PGFZa l3.6:2.2 123:35 4.4i0.8 10.0i1.1 207i36 treatment One day after 2.6:0.5 188:49 4.4i0.4 9.6il.5 163i10 PGF2 0. Two days after 0.9i0.2 284261 5.7il.O 7.9i0.8 130:15 PGF2 (1 Onset of estrus O.7:O.2 317i64 6.9i0.8 8.0i0.8 143i11 One day before 0.3:0.1 870:221 9.7:2.3 9.4i0.9 238i28 ovulation Ovulation detected 0.6i0.1 952:158 6.0i1.4 9.0i0.8 232i36 End of estrusb 4 1:0 8 10254186 4.9:1.1 8.8i1.1 162i20 fllggtrusc 13.5:2.3 208460 4.0:0.9 8.1i0.6 132417 aValues are means i S.E.M. bFirst day mares were not in estrus. C7 to 9 days after ovulation. 59 1' I6 - ::: ,. “r,’IDGBFé“ 5 12 - o C _- e 8- 2 End at F estru5\ 3 4 - (3' _ ovulition CL. 0 l I l 1 4w -2 o 2 4 '_" Days from PGFZ“ Figure 8.--Plasma progesterone after PGFZQ,(H)mg, iu) in six mares. Values are means i S.E.M. and are centered on PGF a treatment (0), the day ovulation was detected (1) and the end of estrus. detected (Table 9). Estradiol and androstenedione did not achieve peak values after PGan in cycle 11 equal to those observed in the first control cycle (Figure 6; Tablesfiand 9). Follicles increased from a diameter of 1.6 i 0.3 cm at onset of estrus to 3.4 i 0.3 cm the day before ovulation. As in the control cycle, blood estrone did not change significantly during the PGan-induced cycle (Table 9). 60 After the control estrus, LH gradually returned toward baseline; it averaged 123 i 35 ng/ml when PGFZa (10 mg) was deposited in the uterus at 7 to 9 days after ovulation. On the first day of the estrus after PGan, LH averaged 317 i 64 ng/ml (P < .01); it increased to a broad peak of about 900 to 1100 ng/ml during the 3 days near ovulation (Table 9). Then LH declined nearly continuously for 8 days, roughly symmetrically with the previous increase, and averaged 69 i 8 ng/ml for 5 days during mid-diestrus; these changes in LH after PGan in cycle II gen- erally resembled those during the first control cycle. The slightly longer duration of estrus after PGFZa treat- ment may have been due to additional time necessary for immature follicles to mature; the diestrum was terminated prematurely by PGan and average follicular size was smaller (P < .05) at onset of estrus during cycle II than during cycle I (1.6 i 0.3 vs 2.8 i 0.2 cm). However, the differences in follicle size between these two cycles may be attributable to relative experience of the individuals doing rectal palpation on due to chance alone, because the size of follicles at onset of estrus and at ovulation for cycles III and IV were similar to those for cycle II. ANOVA of the size of follicles for six mares during the first three cycles indicated a difference (P < .01) in size of follicles at onset of estrus, but not at ovulation among cycles. In addition, follicle size is probably not a very objective measure of follicular maturity, because the largest follicle is not necessarily the 61 ovulatory follicle and ovulation does not occur when a given size is achieved. Control Estrous Cycle (III) Although eight mares were originally assigned for observa- tion during this estrous cycle, only the six mares also assigned to cycles I and II will be considered for comparisons to the first control cycle and the estrous following iu PGFZa' However, one of these six mares did not exhibit estrual behavior during this cycle although an ovulation was detected. Therefore, the data from this mare could not be included in the data for cycle III. None of the intervals measured during this cycle dif- fered from those during cycle I (Table 7); estrus persisted 5.6 days and ovulation occurred 3.8 days from the onset of estrus. During this control cycle, diestrus lasted 14.2 i 0.4 days (Table 7) similar to the diestrum during cycle I. Like the estrual intervals, the changes in hormones during the second control cycle (III) resembled those during the first control cycle (Table 8); and because these hormone changes were similar to those during cycle I they are not separated from the other twomares included in cycle III (Table 10) as a control for cycle IV. However as previously stated, for strictly legitimate statistical analysis, the hormone changes during the first three cycles for six mares were done separately from changes during cycle III and IV for eight mares. With respect to progesterone changes during cycle III: (1) the significant decrease in progesterone for the 3 days before 62 TABLE lO.--Plasma hormone concentrations during a control estrous cycle after a cycle induced by PGFZa'hiseven mares (cycle III). Plasma Hormone Day . Andros- Progesterone LH Estrad1ol Estrone tenedione ------- (ng/m1)------ ------------(pg/m1)------------ Five days a before 10.0i0.9 70:14 3.4:l.3 11.2:0.5 132:17 estrus Two days before 3.8:l.3 62:12 5.1il.l 9.6:1.0 155127 estrus One day before 1.0:0.2 148547 6.1:1.2 ll.4il.2 158:33 estrus Onset of estrus 0-410 1 316:49 5.7:1.4 1o.7:1.0 179:29 One day before O.3:0.0 736i118 ll.5:3.l 11.4:l.8 226i36 ovulation Ovulation detected 0.8i0.3 742:154 15.817.2 11.6:2.6 260:40 End of estrus 3 0i0 9 971:156 4.2:0.8 9.1:0.8 227:40 fllggtrusc 10.7:o.4 161:66 4.1:o.4 12.0:1.1 181:20 aValues are means i S.E.M., n = 7 since one mare did not return to estrus, and two new mares were added. bFirst day the mares were not in estrus. c7 to 9 days after ovulation; day of PGan administration for cycle IV, n = 8. 63 estrus (P < .01) was not significantly different from the proges- terone decrease that occurred during luteolysis before the first control estrus; (2) the levels of progesterone during estrus (the day before estrus, 6 days of estrus and the day after estrus) during the two control cycles did not differ; (3) the increase in progesterone during corpus luteum growth (the last day of estrus and the three days following) did not differ between the two con- trol cycles, and finally, (4) the significant (P < .05) proges- terone changes on 6 days throughout the estrous cycle [(a) 5 days before estrus or day of PGan treatment, (b) onset of estrus, (c) 1 day before ovulation, (d) ovulation, (e) end of estrus and (f) mid-diestrus] were not different amont the three estrous cycles. Estradiol and androstenedione during the second control cycle--as in the first two cycles--increased to peak levels near ovulation. Estradiol averaged 7.3 i 1.9 pg/ml at onset of estrus, increased to 10.1 i 3.8 pg/ml the day before ovulation was detec- ted, and averaged 3.9 i 1.0 pg/ml by the end of estrus (similar to data in Table 10). Androstenedione peaked at 220 pg/ml on the day of ovulation (P < .05). Both estradiol and androstenedione increased significantly during estrus before ovulation (P < .05) and the production of these steroids was accompanied by increases in follicular size; from 1.6 i 0.4 cm at onset of estrus to 3.5 i 0.3 cm before ovulation. Although estradiol concentrations were highest on the day of ovulation, somewhat later than in the two previous cycles, 64 estradiol concentrations appeared to be more variable (see SEM's, Tables 8, 9 and 10) near the day of ovulation. Average estrone fluctuated nonsignificantly between 9 and 12 pg/ml throughout the third cycle, as during the first and second cycles. The pattern of changes in LH around the second control estrus did not differ from that during the first control estrus; LH increased from a baseline of 56 i 11 ng/ml on the 2 days before estrus to 1237 t 345 ng/ml 2 days after ovulation and then declined to 200 ng/ml 7 to 9 days after ovulation when PGan was given subcutaneously (similar to Table 10). Analysis of variance of LH data from mares on 5 days (the day before estrus, the first day of estrus, the day of ovulation, the first day after estrus and day 7 after ovulation) indicated there was no difference in LH values between the two control cycles and the cycle which began after intrauterine PGan. In addition, analysis of LH con- centrations for six selected days throughout the estrous cycle indicated that (1) during cycle III, LH changed significantly (P < .05) and (2) there was no difference in LH patterns (cycle by day interaction) among the three cycles. The lack of significant differences in all criteria mea- sured between the two control cycles was interpreted to mean that there were no residual effects of PGan on the subsequent control cycle. In other words, the ovulation which occurred during the estrus after PGFZa (10 mg, iu) was followed by CL function, luteal regression, estrus and ovulation statistically indistinguishable from those during control cycles and because the second control 65 cycle did not differ from the first, it was used as a control for changes after sc PGFZa (cycle IV). Two new mares were added to the project during cycles IIIand IV. When the hormone values for the seven mares during cycle III are averaged, changes in progesterone, LH, estradiol, estrone and androstenedione were similar to those already detailed for five mares during this cycle (Table 10). They were in estrus 6.0 i 0.4 days after a 14.5 i 0.4 day diestrus (Table 11), similar to the length of estrus in the first two cycles. Behavioral responses to the teaser stallion resembled those during cycles I and II (Table 6). TABLE ll.--Estrus and ovulation during a control cycle and after PGFZa (l5 mg/sc) in eight mares.a __. Control PGF 2 I"terva‘ Cycle (III) Cycle TIV) -------------- (days)------------- From PGFZa to estrus --- 2.8i0.3 From PGan to ovulation --- 8.0:1.ob From onset of estrus b b to ovulation 4.1:0.4 5.1:0.8 From ovulation to b b end of estrus 1.9i0.5 2.1:0.3 Duration of estrus 6.0i0.6b 7.8:O.8 Interestrual interval l4.5:0.4C 9.4:0.4 aEstrus and ovulation were estimated at 12 hours previous to time first detected by palpation. Values are means i S.E.M. n = 7. n = 6. 66 Estrous Cycle After Sub- cutaneousPGFZa (IV)' When 15 mg PGFZa-Tham salt was injected sc 7 to 9 days after the ovulation during the second control estrus (cycle III), the eight mares began estrus 2.6 days later, estrus persisted 7.8 days and ovulation occurred 2.1 days prior to the end of estrus (Table 11). None of these intervals differed significantly from comparable intervals during the control cycle (III) or after iu PGFZa in cycle II (Table 7). Changes in estrual behavior resembled the changes during cycles I, II and III (Table 6). A On the day of PGan treatment, progesterone averaged 10.7 i 0.4 ng/ml (Table 12) and had decreased to 4.6 i 0.3 ng/ml by 12 hours later. As in the cycle following iu PGan, proges- terone continued to decrease and was below 1 ng/ml when the mares first exhibited estrual behavior. Progesterone changes on the 6 selected days throughout the estrous cycle after so PGan were not statistically different than progesterone changes on the 6 days during the control cycle (III). Therefore the decline in progesterone during the 3 days before estrus (cycles I and III) or the 3 days after PGan (cycles 11 and IV) for the six mares represented in all four cycles were compared. The decline in progesterone following 5c PGan did not differ significantly from that after iu PGFZa (Tablesiland 12). On the average, the decline in progesterone for 3 days after PGan treatment (cycles II and IV) was more rapid (P 2 .06) than that during the 3 days before estrus during the control cycles (Tables 9 and 12). 67 TABLE 12.--Plasma hormone concentrations during the estrous cycle following treatment with PGan (15 mg, 5c) in eight mares (cycle IV). Plasma Hormone Day . Andros- Progesterone LH Estrad1ol Estrone tenedione ------- (ng/ml)------- -----------(pg/m1)------------- On day of a PGFZQ 10.7:0.4 161:66 4.1:0.4 12.0:l.1 181:19 treatment One day after ‘ 2.0:0.l 372:90 5.2:1.3 10.65120 160:10 PGF2 0. Two days , after 0.6i0.1 298364 5.6:1.6 8.6:0.7 146116 PGF 20 Onset of estrus 0.4:0.1 463:54 5.4:1.1 8.1:0.4 178118 One day before 0.4:0.2 585170 9.9:2.1 9.3:0.5 220:34 ovulation Ovulation detected 0.4:O.l 663:71 6.3:1.5 9.8:0,9 195:33 End Ofb' estrus 2.7iO.5 936:105 3.8i0.5 8.5:0.3 218i46 Mid- diestrusc 7.8:1.5 264:84 3.4:0.4 10.7:0.7 175:19 aValues are means i S.E.M. bFirst day mares were not in estrus. C7 to 9 days after ovulation. 68 When 5c PGan was administered LH averaged 161 i 66 ng/ml (Table 12); by 24 hours later LH averaged 372 i 90 ng/ml--a 2-fold increase. LH remained between 300 and 500 ng/ml the next 2 days and then increased to peak values of 936 i 105 ng/ml near the end of estrus (Table 12). This pattern of LH changes during the cycle after PGFZa differed (P < .05) from those during the control cycle (III). Firstly, LH concentrations appear to be higher before estrus and higher when estrus began (Table 12) than at the comparable times during cycle III (Table 10); and secondly, LH concentrations appear to be lower the day before ovulation than during cycle.III. However, ovulation occurred normally relative to the end of estrus. The LH changes after PGan appeared to be accompanied by similar changes in estradiol (Table 12); estradiol remained near 5.0 pg/ml for 3 days near the onset of estrus. As in previous cycles, estradiol peaked (P < .05) the day before ovulation at 9.9 i 2.1 p9/m1. The point is this: even though PGan may have altered the normal pattern of LH and/or estradiol during estrus, the duration of estrual behavior and time of ovulation were apparently normal (Table 11). In addition to the alterations of LH changes following PGan, androstenedione did not change as markedly as during the control cycle, and these changes only approached significance (P = .16). However, the highest concentrations (220 i 34 pg/ml) occurred the day before ovulation, similar to the time of peak 69 androstenedione concentrations during the first three estrus periods (Tables 8, 9, 10 and 12), Furthermore, in contrast to the control cycle, estrone decreased significantly before estrus (P < .01) during the 2 to 3 days after sc PGFZa (Table 12). Inspection of data from cycle II (Table 9) reveals a similar decrease in estrone during the 3 days after iu PGan. However the decrease in estrone after iu PGan was not statistically significant. The 2-fold LH increase 1 day after sc PGFZa’ or the sig- nificant decrease in estrone during the 3 days following sc PGFZa’ or lack of significant changes in androstenedione, may have been a result of altered steroid metabolism after PGFZa' However, the mare is a seasonal breeder, and by some manner the ovary becomes quiescent during the winter anestrus, which begins as early as September in some mares in Michigan. The hormonal changes that take place to bring about the onset of seasonal anestrus are unknown. I am suggesting that significant changes observed during cycle IV may have been a result of seasonal changes. There cer- tainly were changes in hormone pattern that appear to occur through the season. For example (Figure 9) the maximal mid-diestrus con- centrations of progesterone decreased over the four cycles. Seasonal decrease in luteal progesterone was suggested by Staben- feldt (1972). Secondly, LH appears to reach maximal concentrations later during estrus as the season progressed; and thirdly, peak androstenedione values were lower as the season progressed (Tables 8, 9, 10 and 12). Whether the differences in LH, 70 It 7‘ 29.23005 OOOOOOOOOOOOOOOOOO OOCOCOCOCOOCCOOOCC DDDDDDDDDDDDDDDD . IIIIIIIIIIIII 0000000000000 O. I. .0. O O I .0. O 0000000000000000 0.000.000 0000000 0000000000 O 00.... 0.0.0.0... ..... 00...... 0000000000 00 I)", O 0.... 000000000000000 0000000000000000 OOOOOOOOOOOOOOOO 00000 00000000000 .0623 65 Days The patterned bars indicate the period of sexual receptivity; the vertical lines within the The second and fourth cycles were initiated patterned bars indicate the day of ovulation. estradiol in mares during four consecutive estrous \II IY ( a .2 SF 66 10! C yy Cb Figure 9.--Re1ative changes of plasma progesterone, LH and 71 androstenedione and estrone after sc PGFZa are due to PGan altered steroid metabolism, or due to seasonal changes in the pituitary-ovarian relationships or due in part to both is unknown. Nevertheless, during cycle IV, seemingly normal estrual behavior occurred, ovulation took place, and so it follows that the overall changes in hormones associated with these physi- ological events were also normal. It was evident from these experiments that PGFZa caused dramatic luteolysis in mares. Each treated mare began estrus within 4.5 days after PGan treatment, in agreement with obser- vations for ponies (Douglas and Ginther, 1972). Similarly, an analog of PGFZa (ICI-79939) was luteolytic in Welsh ponies (Allen and Rowson, 1973). Estrus, Ovulation and Serum Progesterone After 2, 3, 5 or 10 mg of PGan Since 15 mg of sc PGan resulted in rapid luteolysis when given to mares in the 1972 experiments, 10 mg PGan (sc) was tested in one 550 kg mare and proved to be effective to cause luteolysis. In addition, Douglas and Ginther (1972) caused luteolysis in pony mares with the free acid of PGan equivalent to about 9 ug/kg PGFZa-Tham salt in pony mares. Thus, about 3 to 5 mg PGan-Tham salt should be near the minimal effective dose in 500 kg mares. Consequently, in this experiment 2, 3, 5 or 10 mg of sc PGan was given 7 to 9 days after ovulation to determine if in fact 5.3 mg PGFZa would be luteolytic in mares. 72 Most mares during the 1972 experiment returned to estrus 2 to 3 days after PGan and all were in estrus by the end of the fifth day. Furthermore, if PGFZa treatment failed to cause luteolysis, normal demise of the CL would occur about 8 days later since the average duration of diestrus in control mares was 13 to 15 days and PGan was given 5 to 7 days after the end of estrus. Therefore in this experiment if a mare exhibited estrual behavior within 4 days after treatment, PGFZa was considered to have induced luteolysis. The decrease in blood progesterone after PGFZa was taken as the main indication of luteolysis if the mare began estrus between 4 and 7 days. Luteolysis was not considered induced by PGFZa in mares beginning estrus 3.7 days after PGan' \ Estrus and Ovulation After administration of 5 or 10 mg of PGFZa sc, all of the treated mares began estrus within 4 days (Table 13); estrus persisted 6 to 8 days. However, 2 or 3 mg of PGan sc failed to uniformly cause luteolysis; one of the four mares in each group did not begin estrus within 7 days after PGFZQ and, consequently, they were eliminated from the data in Table 12. However, most mares given 2 or 3 mg PGan began estrual behavior by about 4 days after PGan treatment (Table 13) as in the previous experiments when mares were given PGan iu or sc. Progesterone Serum progesterone values, excluding the two mares which did not respond with luteolysis after PGFZQ, are listed TABLE l3.--Estrus and ovulation in mares after var PGFZa (sc) 7 to 9 days after ovulation. ying doses of Dose of Prostaglandin an (mg) Interval 2 3 5 10 ------------------ (days)-----------—------- From PGFZQ to estrus 4.5il.O 2.8:O.4 3.3:O.6 2.5:0.4 From PGF to ovulatioga 8.8:O.3 6.8il.2 8.3tO.6 9.0:],1 From onset of estrus to ovulation 4'3i1'2 4°0i0-4 5-0i0-5 6.3:0.8 From ovulation to end of estrus 2.7:O.3 2.0:O.6 1.0:0.o l.8i0.3 Duration of estrus 7.0:].2 6.0:].0 5,010.5 8,110.7 aPGan was given subcutaneously in 1 ml of saline; values are means i S.E.M. Only three of four mares began estrus in response to 2 or 3 mg PGan, while the data for 5 and 10 mg PGan are from five and six mares, respectively. in Table 14. Similar to progesterone changes after the intra- uterine and subcutaneous treatments (Figure 8; Tables 9 and 12), progesterone decreased markedly within 24 hours after 2, 3, 5 or 10 mg PGan (sc) and continued to decrease during the 72-hour sampling period. Statistical analysis of the data for treatment effects and differences over time indicated significant differences in the rate of decline of progesterone (P < .05) among treatment groups. Serum progesterone decreased more rapidly after 5 or 10 mg of PGan; the concentration of progesterone fell under 1 ng/ml by 48 hours. However, after 2 or 3 mg PGFZa , progesterone had not 74 TABLE l4.--Mean serum progesterone in mares after varying doses of PGFZa (sc) 7 to 9 days after ovulation.a Dose of Prostaglandin an (mg) Hours From PGFZG 2 3 5 10 ----------------- (ng/ml)------------------- -24 10.8i1.7 10.3:o.3 11.4i1.2 l4.8:3.7 o 9.1:3.2 15.1:2.8 11.2:1.o 11.4:1.9 24 3.2:].4 3.3:o.5 3.0:o.4 2.4:o.3 48 2.1:].0 1.5:o.4 0.9:o.1 o.9:o.2 72 1.5:].0 1.1:o.3 O.6:0.1 o.5:o.1 aPGF a was given subcutaneously in 1 ml of saline; values are means i g E.M. Of four mares given 2 or 3 mg PGan only three began estrus, while the data for 5 and 10 mg PGFZG are from five and six mares, respectively. decreased to less than 1 ng/ml by 72 hours when sampling ended (Table 14). Since PGan caused luteolysis in an unequal number of mares in the various treatment groups, the results of statistical analysis are only approximate. I conclude that 5 mg PGan-Tham salt (sc) is adequate to produce luteolysis in mares (5.550 kg). Douglas and Ginther (1972) concluded that 1.25 mg of PGan (free acid, M.W. = 354.5) was luteolytic in pony mares (5.195 kg). Five mg of PGan-Tham salt (475.6 M.W.) is equivalent to 3.68 mg PGan free acid and when compared on a body weight basis the minimal effective dose in my mares (approximately 9 ug/kg) was equivalent to that (8.6 ug/kg) in the pony mares (Douglas and Ginther, 1972). 75 Estrus, Ovulation and Serum Progesterone After 10 mg PGFZQ Administered on 1, 3, 5 or 7 Days After Ovulation Since 10 mg of sc PGan had proved to induce luteolysis in mares on day 7 after ovulation, this dose was chosen to test the relationship between the age of the CL and PGan-induced luteolysis. In cattle, PGFZa was effective on day 5 or 11 but not on day 3 after ovulation (Louis et al., 1973). But luteal tissue appears to form more rapidly in mares, so 1 administered 10 mg sc PGFZa' Tham salt to 5 mares each on day 1, 3 or 5 after ovulation and to 6 mares on day 7. Estrus and Ovulation Some estrual characteristics for the mares which responded to PGFZa are listed in Table 15. The data for mares on day 7 are the same as those included in Tables 11 and 12 under 10 mg. Among the five mares given 10 mg PGan 1 day after ovulation none began estrus within 7 days. Only two of five mares treated 3 days after ovulation responded to treatment with a reduced duration of diestrus. All of the mares given PGFZa on 5 or 7 days after ovu- lation and two mares treated on the third day postovulation began estrus, most within 4 days after PGFZa (Table 15). Average dura- tion of estrus ranged from 6 to 8 days among treatment groups. Although the interval from PGan to ovulation and duration of estrus after PGFZa on day 5 or 7 were somewhat longer than those after treatment on day 3, none of the differences in any of the intervals listed in Table 14 was significant for those mares in 76 TABLE 15.-—Estrus and ovulation in mares given 10 mg PGFZa (sc) on various days after ovulation.a w m Days after ovulation Interval 1 3 5 7 ------------------ (days)------------------- From PGFZa to estrus --- 2.0:O.3 2.9tl.O 2.5:O.4 From PGF to ovulatioga --- 5.5:3.o 8.7il.2 9.0:1.1 Fr°m °"59t °f eStrus --- 3.5:2.5 5.8:o.5 6.3:0.8 to ovulation From ovulation to end of estrus --- 2.0:0.0 2.0:O.6 1.8:0.3 Duration of estrus --- 5.5:2.5 7.8:1.0 8.1:O.7 aPGan was given subcutaneously in 1 ml of saline; values are means i S.E.M. None of the five mares began estrus in response to PGFZG 1 day after ovulation, two of five mares began estrus after treatment on day 3, and the data for day 5 and 7 are for five and six mares, respectively. which the CL function was terminated. Intervals to estrus and ovu- lation and duration of estrus in mares given 10 mg PGan (sc) on 3, 5 or 7 days after ovulation were comparable to those for mares which responded to 2, 3 or 5 mg of PGan on 7 days after ovulation in the previous experiment. Allen and Rowson (1973) also reported that PGF was not luteolytic before 3 days after ovulation in 2a mares; and as in this study some animals responded on day 3 while some did not. 77 Progesterone Table 16 lists serum progesterone for the mares which responded to 10 mg PGFh (sc) on 3, 5 or 7 days after ovulation. . Progesterone decreased rapidly (P < .01) after PGan to _.1 ng/ml by 48 hours after treatment. As expected from data from control cycles (Figure 3; Tables 8 and 10), serum progesterone averaged 6.7 ng/ml at the time of treatment on day 3 after ovulation, less than the 11 or 12 ng/ml for mares treated on day 5 or day 7. As was true for all mares responding to PGFZa (Figure 8; Tables 9 and 12) as well as for control mares during luteolysis (Figure 3; Tables 8 and 10), estrus began when progesterone was near 1 ng/ml TABLE 16. --Serum progesterone in mares after 10 mg PGan (sc) on various days after ovulation. a “fi— ___-___— __—_—..- __- _— ”. ‘_-.——— _- —c——_- —._—.-._ .fi—O Days After Ovulation Hours From PGFZa 3 5 7 ----------------- (ng/ml)----------—------ -24 4.9:0.3 ll.2:1.9 l4.8:3.7 O 6.7:0.4 12.8:3.4 ll.4:l.9 24 l.8:0.3 2.9:O.9 2.4:O.3 48 O.7:0.1 l.O:O.2 O.9:O.2 72 1.8 0.7:0.1 O.6:O.1 aPGan was given subcutaneously in 1 m1 of saline; values are means i S.E.H. None of the five mares began estrus in response to PGFZG 1 day after ovulation, while two of five mares began estrus after treatment on day 3. The data for day 5 and 7 are for five and six mares, respectively. 78 (Tables 15 and 16). I suspected that the two mares which responded to PGan on day 3 may have had a more mature CL than the three mares which did not begin estrus after PGFZQ. However, serum progesterone was no higher at treatment in the mares that began estrus (6.7 i 0.4 ng/ml) than in the mares which did not begin H- estrus (7.6 0.3 ng/ml) within 8 days after treatment on day 3 (Tables 16 and 17). In contrast to the fall in progesterone after PGan treat- ment on day 5 or 7, serum progesterone continued to increase in mares given PGan 1 day after ovulation (Table 17). This increase was comparable to progesterone changes during the 3 or 4 days after ovulation in control cycles (Figure 3). TABLE l7.—-Serum progesterone for mares in which PGF a did not result in estrus within 8 days of treatmegt.a ___—~—-——. ~-.——.——-—— w... Dose of PGan (mg) Days After Ovulation Hours From PGan 2 3 l 3 -24 17.6 9.0 l.6:l.2 3.9:0.5 0 10.6 6.2 l.7:0.8 7.6:0.3 24 6.3 2.9 2.2:0.3 3.8:1.5 48 5.9 1.7 4.2:0.8 4.1il.8 72 7.4 2.4 6.3:1.1 4.3:2.5 aPGF a was given subcutaneously in 1 m1 of saline; values are means : §.E.M. Of four mares given 2 or 3 mg PGFZa, one mare did not begin estrus, while all five mares given 10 mg PGFZQ (sc) on day 1 did not begin estrus and three of five mares treated on day 3 after ovulation did not begin estrus within 8 days. 79 The transient decrease in progesterone after 2 or 3 mg of PGFZa was given on day 7 or after 10 mg PGFZa on day 3 (Table 17) apparently reflected incomplete luteolysis. Evidently these PGF2 treatments were not sufficient to complete luteolysis, because although progesterone decreased 50% at 24 hours after PGFZa’ it began to increase at 72 hours. As a result these five mares did not begin estrus until 8 to 16 days after PGFZQ treatment as would have been expected without PGan treatment. In summary, complete luteolysis resulted after 10 mg of PGan-Tham salt (sc) on or after the fifth day of CL development. Some mares responded with luteolysis after PGan on day 3; others had only a transient decline in progesterone and consequently PGan did not reduce the duration of diestrus in these mares. PGan did not alter the estrous cycle when given on day 1 after ovulation. Breeding, Pregnancy and Foaling After PGF2a To determine fertility after PGan, eight mares (four each in 1972 and 1973) were mated during the estrus that followed 10 or 15 mg PGan-Tham salt (sc). Beginning on the second day of estrus the mares were mated every other day to a fertile stallion until ovulation occurred. Six (75%) of the eight mares mated after sc PGan were pregnant (four in 1972, two in 1973) on the basis of palpations around day 40 of gestation. One mare apparently aborted between 50 and 80 days, a period when mares most frequently abort (Allen, 1972) Parturition for five mares (63%) occurred 80 332 i 3.0 days after ovulation; all deliveries were unassisted and the foals appeared healthy and normal. Foal heat occurred 6.5 i 0.4 days later; it persisted 7.0 i 1.2 days and ovulation occurred 2.2 i 1.1 days before the end of estrus. The fertility and dura- tion of pregnancies and the characteristics of the foal heats in these mares resembled those reported for untreated mares (Asdell, 1946). That the foals were healthy dispelled fears of abnormal fertilization, implantation or embryogenesis following PGFZa treat- ment. A Pennsylvania study (Ganjam, 1975) indicated that fertility was normal (65%) in 350 mares mated (or inseminated) following PGan (5 mg, sc). Therefore, PGFZa’ at least in these two reports, did not decrease conception and pregnancy, because Sullivan (1972) reported that the average pregnancy rate was 51% for the Quarter Horse and 43% for the Thoroughbred. Thus, I am optimistic that PGFZa would not be detrimental to fertility after use for induction of 1uteoly- sis. In fact my data suggests the synchrony of ovulation after PGan is such that mares could be bred at 7 days after PGFZQ with- out regard to signs of estrus. Side Effects of PGFZa After infusion of 10 mg PGan-Tham salt into the uterus of the first mare treated in the first experiment, I noted about 15 minutes later that she was sweating profusely. Her hair and body were wet to the touch, and droplets of water moistened the dry ground beneath her. Three more mares were observed closely 81 after PGFZa (10 mg, iu) treatment. Sweating began about 10 to 15 minutes after PGFZa; first the skin was damp behind and between the front legs and spread until the entire body was damp. Within 15 to 20 minutes after PGan droplets of water covered the hairs and dropped from the body. The mares appeared to have experienced a hard run, however all were tied. During the 20 minutes after PGFZa rectal temperature decreased from an average of 100.5°F, by 0.5°F, 1.2°F, and 1.4°F in the three mares. The sweating sub- sided and the mares had begun to dry off by 35 minutes after treatment. Mares were observed to sweat noticeably after 60% of the PGan treatments during the 1972 project. They began to sweat 10 to 15 minutes after PGFZa but were dry by 45 to 60 minutes. They did not appear to be very uncomfortable. For example, often they continued to eat hay. However, modest symptoms associated with colic were noticed in 70% of the mares: dulled eyes, trem- bling upper lip, involuntary muscle contraction of gut reflected by arched back. Frequent defecation and minor diarrhea were observed once each. These symptoms appeared and disappeared within the 45 to 60 minutes after treatment. During the 1973 experiment, four more mares were observed closely. All four began to sweat 8 to 15 minutes after PGFZa; rectal temperatures dropped from lOO.8°F to 98.0°F at 35 minutes when sweating had again decreased and the mares began to dry off. Lauderdale and Miller (1974) rated the sweating in mares on a four-point (0 to 3) scale and found that there was a graded response of sweating and temperature decrease after various doses 82 of PGFZa' The mares receiving smaller doses had less sweating and a smaller change in rectal temperature. In one mare given 200 mg of PGFZa-Tham salt, Lauderdale (1973) described profuse sweating, muscle tremors and a dullness of the eyes, but full recovery followed within 3 to 4 hours. Lauderdale and Miller (1974) indicated that mares given epinepherine (10 mg) began sweating within 8 to 10 minutes and had a transient but insig- nificant (.25°F) decrease in body temperature; shiving in response to epinepherine may have adjusted body temperature back towards normal. Mares given PGan were sweating 15 to 20 minutes after treatment, and decreased body temperature (2°F) occurred but shiv- ering did not. The PGan mares had a depressed body temperature for 3 to 4 hours until other effects of PGan had disappeared. Lauderdale and Miller (1974) also suggested that the mares may not shiver after PGFZa due to depression of the hypothalamic tempera- ture regulating mechanism, because sweating can be caused by stimulus of the anterior hypothalamus; shivering by stimulus of the posterior hypothalamus (Ganong, 1969). It is possible that while PGFZa stimulates sweating, it may inhibit the shivering response. However, this does not appear to be a major problem since even after a mare was given 40-fold higher than the minimal effec- tive dose of PGan, all of these side effects disappeared within 4 hours. Similar to the side effects following PGFZa treatment, sweating, hypermotility of the gastrointestinal tract, diarrhea, increased pulse and respiration and slight abdominal pain followed 83 treatment with a PGan analog (101 79939) in mares (Allen and Rossdale, 1973). They suggested that it should be possible to develop a PGFZa analog with fewer toxic effects in mares. Abnormally Prolonged Luteal Function As previously stated, one mare did not exhibit estrual behavior during cycle III of the first experiment. Inspection of progesterone values for cycle I and cycle III revealed progesterone near 1 ng/ml on 1 day while she was in estrus during cycle I, while her progesterone was never below 5 ng/ml during cycle III. This was in marked contrast to the usual progesterone pattern in mares (Figure 3). This mare did, however, exhibit a normal estrus fol- lowing PGFZa in cycles 11 and IV, and progesterone was below 1 ng/ml during estrus after PGan in both cases. This observation suggested that PGan had potential for induction of luteolysis in mares with abnormally extended CL function. The mare discussed above and two of her herdmates had abnormal estrous cycles. All three of the mares had previously exhibited estrual behavior but only briefly (2 to 3 days) and with less intensity (rated 5+) than normal. Ovulations which occurred in the absence of estrus were detected when progesterone was 5 to 7 ng/ml in all three mares. However, these ovulations were not accompanied by an increase in LH concentrations and may have been atretic follicles or otherwise abnormal follicles. After 35.7 i 0.9 days of extended luteal func- tion when progesterone was 3 to 11 ng/ml, all three mares were given PGan. For a total of seven treatments of PGan in these 84 three mares, estrus began in every case in an average of 2.6 i 0.4 days (Table 18), it persisted 5.4 i 1.1 days and ovulation occurred 0.7 i 0.6 days before the end of estrus. The interestrual interval between two consecutive estrus periods following PGFZa treatments was 9.5 i 1.5 days. One of these mares was mated; she conceived and foaled 330 days later. Kenney et a1. (1975) treated 73 maiden, barren and foaling mares having extended periods of clinical TABLE l8.--Estrus and ovulation after PGF a in three mares with abnormally prolonged luteal fu ction.a Interval Control Cycleb PGan Cyclec --------------- (days)------------- From PGFZa to estrus --- 2.6iO.4 From PGan to ovulation --- 6.9:O.5 From onset of estrus to ovulation 5.0:0.6 4.310.5 From ovulation to d end of estrus -l.3i0.3 0.7:O.6 Duration of estrus 3.7iO.7 5.0:l.l Interestrual interval 35.7:O.9e --- aEstrus and ovulation were estimated at 12 hours previous to time first detected. Doses were 10, 15 or 20 mg (sc) or 10 mg (iu). Values are means i S.E.M. n = 3. n = 7. dOvulation occurred after the end of estrus. eThe number of days of prolonged luteal function between the end of a previous estrus and the first day of estrus after PGF 2a' 85 anestrus (73 days) during the breeding season with PGFZa' Estrus began in 73% of the mares 4.4 days after 5 mg of PGFZa (sc) and 55% conceived. Allen and Rossdale (1973) related that a PGFZa analog (101 79939) was used as a treatment for nonpregnant mares with no estrual behavior and progesterone levels > 1 ng/ml during the breeding season; 34 of 35 mares began estrus as a result of treatment. It probably is essential for a CL to be present in these mares if PGan treatment is to be effective. PGan (sc) treatment of mares with extended luteal func- tion has advantages over infusion of saline or other materials into the uterus to induce luteolysis. Mares had reduced duration of luteal function after infusion of saline (Bain, 1957; Ginther, 1971), but because of the threat of uterine infections from iu treatments, I favor sc PGan treatment. Likewise, noncycling mares exhibited estrus after stilbestrol or estrogen treatment, but ovulation did not occur unless a follicle was present (Burk- hardt, 1947b; Nishakawa, 1959). Similarly, HCG treatment caused ovulation only when follicles were present (Day, 1940). There- fore, since the 1uteolysis which follows PGFZa appears to trigger the sequence of events which follow normal uteolysis, it should be a useful treatment fin~mares with abnormally prolonged luteal function. GENERAL DISCUSSION Hormonal Interactions During the Estrous Cycle in Mares Blood progesterone decreased rapidly in mares during the 3days before estrus in untreated mares as well as after PGan treatment. Similarly, progesterone decreased during the 3 days before estrus (Wettemann et al., 1972) as well as after PGFZQ treatment in cows and heifers (Louis et al., 1973; 1974). Fur- thermore, the decline in progesterone before estrus in ewes and sows (reviewed by Hansel and Echternkamp, 1972) resembles that in mares and cows. Therefore, CL regression and the onset of estrus appear to be temporally related in the mare much like they are in other large domestic animals. Exogenous progesterone suppresses estrus behavior and/or ovulation in most mammals, including mares (Loy and Swan, 1966). However, quantities of progestogens that blocked estrus did not always block ovulation in mares. For example, 100 mg of proges- terone given mares daily blocked estrus and ovulation, but ovula- tion occurred in the absence of estrual behavior while 50 mg/day was administered. It appears, then, that in the mare ovulation can occur in the absence of estrus, but apparently blood progesterone must be minimal (5.1.0 ng/ml) to allow signs of estrual behavior. Comparatively, maximal progesterone values in the mare were greater (15 ng/ml) than in cattle (4.6 ng/ml; Wettemann et al., 86 87 1972) and sheep (3.0 ng/ml; Thorburn et al., 1969). In addition, much higher doses of exogenous progesterone were required to sup- press estrus in the mare than in the cow and ewe (reviewed by Britt and Ulberg, 1971). Similarly, synthetic progestogens in quantities nearly 10- to 20-fold greater than those effective in cattle or sheep were ineffective in mares. It appears that higher levels of exogenous progesterone are necessary to block estrual behavior and/or ovulation in mares than in cows and sheep. Stabenfeldt et al. (1972) reported that mares occasionally ovulated during early or late diestrus when blood progesterone was at about 50% of maximal diestrual concentrations. In my study, although not accompanied by significant LH changes, four ovulations occurred during periods of abnormally prolonged luteal function; three occurred at times when progesterone was less than maximal for a given mare. Three other ovulations occurred when LH was still elevated during early diestrus. It seems possible that the changes on the ovarian surface called "diestrus ovulations" may in fact be large follicles which subside without ovulation. Although progesterone remains near basal values during estrus in cows, sows, ewes and mares, the duration of estrus in the mare is much longer than that in the others. In mares, before and during estrus, progesterone remained below 1 ng/ml for 6 to 8 days until about 24 to 36 hours after ovulation. Cows also have progesterone below 1 ng/ml for 5 days (Wettemann et al., 1972), but duration of estrus was less than 1 day. Similar to the cow, IDrogesterone in the ewe remained low for 5 days and estrus persisted 88 28 hours (Thorburn et al., 1969; Hansel and Echternkamp, 1972). Thus the prolonged period of estrus in the mare does not appear to be due entirely to a prolonged period of low progesterone. Synthesis of significant quantities of progesterone appears to occur sooner after ovulation in mares than in other species, because blood progesterone increased 7- to lO-fold within 36 hours after ovulation in mares. Formation of luteal tissue occurred in ovaries of mares 24 to 48 hours postovulation (Short, 1964). In contrast, progesterone remained at basal values for 48 to 72 hours after ovulation in the cow and ewe (Thorburn et al., 1969; Swanson et al., 1972). As a result, blood progesterone peaks within about 7 days after ovulation in mares (Table 9, cycle II), but not until 10 or 11 days after ovulation in the ewe and cow. In the rabbit, before ovulation and following the initial LH release, 20a-hydroxypregn-4-en-3-one (20a-0HP) secretion from the ovarian interstitial tissue acts to prolong LH secretion (Hilliard et al., 1967). Preovulatory increases of progestogens in the human, the rhesus monkey and the rat were reported (reviewed by Miyake, 1974), but no such increase in progesterone has been reported for the cow, ewe or sow (Hansel and Echternkamp, 1972). In agreement with data in cows, I detected no preovulatory increase of progesterone in mares. However, this could be due to infrequent bleeding and/or the inability of the progesterone antibody to bind ZOa-OHP. Short (1964) reported the presence of ZOo-OHP in the luteal tissue of mares postovulation. 89 \ In most mammals, blood estrogen production is related to follicular develOpment. In mares both follicular size and estra- diol concentrations increased 2- to 3-fold to a maximum 12 to 36 hours before ovulation (Figures 4 and 7; Tables 8, 9, 10 and 12). Similarly in cows, follicular size increased 2-fold during the 4 to 5 days before ovulation and estradiol was highest 0.5 days before estrus, about 40 hours before ovulation (Hansel and Echtern- kamp, 1972; Wettemann et al., 1972). Total estrogens in the ewe and sow peaked near the onset of estrus, near the time of maximum follicular development (Scaramuzzi et al., 1971; Hendricks et al., 1972). Thus, the pattern of follicular estradiol production in mares resembles that of other species, peaking 24 to 48 hours before ovulation. However, estrogen may be elevated longer in mares during lowered progesterone, resulting in a longer period of estrus. It is also possible that the longer period of estrus in mares is "due to the longer time taken by the follicle to come to the surface of the ovulation fossa and break through" (Eckstein and Zuckerman, 1956a). Other than the peak in estradiol just before ovulation, no other significant changes in estradiol were observed during the estrous cycle of mares. However, other estrogens may reflect follicular changes during the remainder of the cycle. Hansel and Echternkamp (1972) reported three elevations of plasma estrone which they suggested coincided with waves of follicular develop- ment during the estrous cycle in cows, although Marion and Gier (1971) reported development of follicles during all stages of the 90 estrous cycle. Urinary estrone was elevated before ovulation in women (Moghissi et al., 1972). However, there were no significant changes in blood estrone during the estrous cycle in my mares (Tables 8, 9, 10 and 12). This is in contrast to reports of increased urinary estrone prior to ovulation in mares (Loy, 1970). Peak urinary estrone, however, may reflect renal metabolism of plasma estradiol. Estrogens have been implicated as the cause of pituitary LH release in several species. For example, in the cow, ewe and sow estradiol begins to increase 2 to 4 days before LH increases and reaches maximal levels one or two days before the ovulatory surge of LH in the cow and ewe and 4 days before the LH peak in sows (Hansel and Echternkamp, 1972). Furthermore, exogenous extradiol evoked release of LH in ovariectomized heifers (Hobson and Hansel, 1972; Beck and Convey, 1974) in intact heifers after enucleation of the CL (Hobson and Hansel, 1972), and in anestrus (Beck and Reeves, 1973) or ovariectomized ewes (Goding et al., 1970). The increase of estradiol concentrations that preceded and accompanied the LH release in my mares (Tables 8, 9, 10 and 12) suggests that estradiol stimulates release of LH in mares as in the other species. The increase in estradiol began 2 days before estrus, and the LH surge began 1 day later. Then estradiol continued to increase to maximal values the day before ovulation, slightly anticipating the parallel changes in LH. LH is the ovulatory hormone in all mammals. In most species, an ovulatory surge of LH occurs briefly about 1 day 91 before ovulation. In cows, for example, a 10- to 20-fold increase in LH persisted 8 hours and preceded ovulation by an average of 32 hours (Swanson et al., 1972). In contrast to the relatively brief ovulatory surge of LH in the cow, LH in mares was elevated for 12 to 14 days. It began to increase near the onset of estrus and continued to a relatively broad plateau persisting for 3 or 4 days near ovulation. To my knowledge, the broad elevation of LH persisting 12 to 14 days is unique to the mare. LH causes release or synthesis of proteolytic enzymes in the follicle of several species, enzymes which may increase the distensibility (decreased tensile strength) of the follicle wall (Espey, 1974). Perhaps ovulation in mares requires prolonged LH stimulus to produce these enzymes because of unique anatomical features of her ovary. As discussed in the literature review, ovulation nearly always occurs at the ovulation fossa in mares. Therefore, developing follicles must migrate through the ovarian stroma to the fossa before ovulation. Perhaps prolonged LH stimu- lation is required to complete this migration. However, I have no evidence to support either of these theories. That continued LH stimulus is a requirement for ovulation in mares is indicated by response of time of ovulation to differ- ent amounts of LH released to gonadotropin releasing hormone (GnRH) stimulus. When 400 pg of GnRH was given to pony mares once on day 2 of estrus, LH increased 2-fold and the blood LH was sustained 3 hours (Ginther and Wentworth, 1974); the duration from onset of estrus to ovulation was similar to that in control mares. However, 92 when 2 mg of GnRH was given to light mares beginning on day 2 of estrus daily until ovulation, the interval from onset of estrus to ovulation was reduced (Downey et al., 1974). This suggests that release of greater quantities of endogenous LH after continued stimulus with exogenous GnRH caused ovulation to occur sooner after the onset of estrus and possibly that release of some minimal quantity of LH may be necessary for ovulation to occur. Presumably, the brief surge of LH before ovulation and/or the low levels following the LH spike in cows and ewes is suffi- cient to cause luteinization of the granulosa cells and initiate progesterone secretion. Possibly the granulosa cells of the rup- tured follicle of mares requires a prolonged LH stimulus for this to occur. This may be related to the fact that equine granulosa cells have 60% fewer receptors than porcine granulosa cells (Chan- ning, 1974). Assuming LH is luteotropic in the mare as in most species (Hansel and Echternkamp, 1972), the sustained concentration of LH before and after ovulation may account for more rapid devel- opment of luteal function after ovulation in mares than in other species. The anterior pituitary of the mare appears to be sensi- tive to removal of progesterone, as is true in other species; after progesterone decreased 2. 50%, estradiol increased, and increased LH followed 1 day later. Similar changes were reported for the ewe, cow and sow (reviewed by Hansel and Echternkamp, 1972). However, LH does not decline until progesterone is increased in the mare suggesting progesterone may be required to 93 stop anterior pituitary release of LH. An alternative, however, is that a longer time is required for metabolism of equine LH (as discussed in the review of literature). Short (1972) suggested androstenedione may be important in regulating estrous behavior in mares. Although there is some evidence to suggest that androgens are related with sexual behavior in primates (Everitt and Herbert, 1969), to my knowledge it has not been proven. Some authors reported an androstenedione peak before ovulation in women (Abraham, 1974; Baird et al., 1974), while others found no change in androgens at ovulation (Horton, 1965). Similarly, androstenedione was elevated during proestrus in rats (Dupon and Kim, 1973) and ewes (Baird et al., 1968). In my mares, blood androstenedione was elevated during the period when mares exhibited estrual behavior, and peaked l or 2 days before ovulation. Since mare follicular fluid and ovarian vein blood contained large quantities of androstenedione (Short, 1964), presumably the blood androstenedione is of follicular origin in mares as reported for women (Abraham, 1974) and may reflect synthesis of estradiol. The ever increasing levels of estradiol (Figure 4) and androstenedione (Figure 6) were accompanied by ever increasing concentrations of LH in the plasma samples collected 5 to 6 days before ovulation. The continual increase of estradiol was termina- ted 2 days before ovulation (Figure 4, cycle 1) whereas androstene- dione levels reached a maximum 1 day before ovulation (Figure 6, 94 cycle I). LH increased to maximal levels the day of ovulation (Figure 7, cycle I). It would be interesting to test the idea that the continual increase of LH in some manner first blocked the enzymatic conver- sion of androstenedione to estradiol 1 to 2 days before ovulation and secondly blocked the formation of androstenedione just before ovulation. According to this scheme, LH would continue to elimi- nate or block enzymes of the steroidogenic pathway in the ovary resulting in an increase in progesterone. Progesterone in turn might exert a negative influence on the pituitary and/or hypo- thalamus causing diminution of LH release about 2 days after ovu- lation (Figure 7). With further increase in progesterone, LH secretion finally reached basal levels 6 or 7 days after ovulation (Figure 7). Mechanism of Action of PGan The uterus of the cow and ewe exert a local control on the ovary (reviewed by Anderson et al., 1969), but uterine control of luteal function acts systemically in the mare (Ginther and First, 1971). A possible explanation may lie in the fact that the utero- ovarian vein and the ovarian artery were closely related anatom- ically; i.e., the tortuous ovarian artery was adhered to the sur- face of the utero-ovarian vein in the ewe and cow (Ginther, 1974). This might facilitate the direct transfer of the uterine 1uteolysin from the uterine vein into the ovarian artery to deliver 1uteoly- sin in high concentration directly to the ovary. In the mare, 95 however, such an intimate relationship does not exist between the two vessels. Therefore, it was not surprising that PGFZa was no more effective as a 1uteolysin when injected into the uterus than when given systemically (sc) in mares. In contrast, the cow (Hafs et al., 1974; Stellflug et al., 1975) and the ewe (Ginther, 1974) required considerably more PGFZa to cause luteolysis when admin- istered systemically than when PGFZa was injected into the uterus. My data on PGFZa-induced luteolysis agree with Ginther's (1974); PGFZa was equally effective whether given systemically or injected into the uterus of mares. Possibly the uterine 1uteolysin, like PGan, is inactivated rapidly in the lungs (Piper et al., 1970) and/or other tissues such as the liver, kidney and spleen (Oester- ling, 1972). If so, a direct transfer of the uterine 1uteolysin or PGan--from the utero-ovarian vein to the ovarian artery--would avoid degradation by the lungs or other tissues before it reached the CL. On the other hand, the tissues of animals without a local transfer mechanism for the transfer of the uterine luteoly- sin such as the mare may not degrade PGan (or the uterine 1uteolysin) as rapidly (if at all) as in cows and ewes. This would explain not only the phenomena of local vs systemic control but also the differential effect of sc vs iu PGan in the two types of animals having different utero-ovarian relationships. Although luteolysis is interrupted when the uterus is removed, exogenous PGan appears to cause luteolysis in the absence of the uterus; sc PGFZa caused luteolysis in hysterectomized pony mares indistinguishable from that in intact mares given PGFZa 96 (Douglas et al., 1974). Similarly, PGFZQ caused luteolysis in hysterectomized cows (Stellflug et al., 1975). Thus, the lute- olytic effect of PGF2 does not require the presence of the uterus; PGan-induced luteolyzis appears to represent a direct action of PGan on the CL. The CL of mares appears to be more sensitive to PGan than the CL of cows and ewes, since the minimal effective sc dose for PGFZa to induce luteolysis was 22-fold greater on a body weight basis for ewes than for mares (Ginther, 1974). Similarly, the minimal effective dose of PGan in cows (Stellflug et al., 1975) was at least 3-fold higher than that for mares. The minimal effective dose of PGan was 5 mg or 9 ug/kg in the mares of this study, comparable to that in pony mares (Douglas and Ginther, 1972). As an alternative to greater sensitivity to PGFZa in the mare, perhaps the cow and ewe metabolize PGFZa more rapidly. The CL appears to be more sensitive to lytic agents when full luteal function has been developed than during the formative stages of the CL. In my study the CL of mares became susceptible to the luteolytic action of PGFZa before day 5 after ovulation; PGan caused luteolysis in no mares treated on day l, in three of four mares treated on day 3 and in all mares treated on day 5 after ovulation. Similar results were reported for pony mares (Allen and Rowson, 1973) and cattle (Rowson et al., 1972). By what mechanism does PGFZa cause the demise of the CL? (h~iginally, Pharriss and Wyngarden (1969) suggested that veno- constriction of the utero-ovarian vein caused by PGan might be 97 responsible for the resultant luteolysis. The venoconstriction would reduce blood flow to the ovary and "starve" the CL. Fur- thermore, distribution of radioactive microspheres before and after PGFZa demonstrated a redistribution of blood supply to the CL (reviewed by Poyser, 1973, and Pharriss et al., 1972). However, this distribution of blood supply may be a consequence of CL regression and not a cause. PGFZa also inhibits cholesterol ester synthetase, an enzyme necessary for formation of the cholesterol esters essential for progesterone production (Poyser, 1973). In contast, PGan caused increased progesterone production by luteal tissue in vitro during a 3-hour incubation: but in agreement, a longer, incuba- tion of PGan with luteal tissue resulted in decreased proges- terone production (Poyser, 1973). Direct infusion of PGan into the ovaries of ewes caused an initial increase in progesterone levels in the ovarian vein. This initial increase in progesterone was followed by a pronounced decrease in progesterone (Poyser, 1973). If PGan caused an increase in progesterone production in cows it was not detected in a sample of jugular blood taken 10 minutes after PGFZa treatment (Hafs et al., l974)--progesterone decreased in cows immediately and continued to decrease rapidly for 1 hour. In conclusion, the mechanism of action of PGan on the CL of the mare (or of other species) is unknown. It is possible PGan acts initially to cause redistribution of blood and sec- ondly to inhibit cholesterol synthesis. PGFZa may bind directly 98 to the luteal cells since prostaglandin receptors have been reported in these cells (Rao, 1974). Practical Applications Replacement of horses by tractors, at least in the United States, is reflected by the decreased numbers of horses and mules from a peak of about 25 million in 1920 to 2.8 million in 1950 (U.S. Department of Commerce, 1960). Recently, however, the numbers of horses are increasing-~principally for recreation and meat production. There was estimated to be 121 million horses, mules and asses in the world (excluding Russia and Red China) and 8 million in the United States (U.S., Food Agric. Orig., 1972a). In 1972 the meat production from slaughtered horses, mules, asses and crosses was estimated to be 386,000 metric tons (U.S., Food Agric. Orig., 1972b), mostly in Europe and South America. The amount of time spent horseback riding in the U.S.A. in 1973 was equivalent to that spent golfing, waterskiing or snowski- ing (U.S. Department of Commerce, 1974a). Horseracing was attended by 73 million people in 1973, comparable to the total number of people attending college and professional football games (U.S. Department of Commerce, 1974b). Breeding efficiency in mares is low in comparison to cows, due at least in part to the variability in the duration of estrus and thus the time of ovulation is difficult to predict. Pregnancy rate was 43% after Thoroughbred mares were mated once during estrus, and 51% for Quarter Horse mares inseminated artificially or mated naturally several times during estrus (Sullivan, 1972). 99 In my study, ovulation in mares occurred relatively syn- chronously at 8 days after PGan. Thus, possibly mares could be mated (or inseminated) with normal fertility 7 or 8 days after PGFZa without regard to estrus. For such a management regime to be useful, fertility following PGFZa must be at least equivalent to that achieved with present management systems. In the small number of mares mated after PGan in this study, 5 of 8 foaled (63%), and Allen et a1. (1975) reported 83% fertility when mares were mated after an analog of PGan was used to induce luteolysis. PGan did not reduce fertility in either study. Therefore, PGan should be useful for breeding management of mares, but larger trials are required. In one such trial, of 350 mares bred once 7 or 8 days after 5 mg sc PGFZQ, 65% conceived (Ganjam, 1975). Synchronization of ovulation in mares would (1) reduce the number of inseminations or matings, thereby reducing the number of stallions and labor required, and desirable stallions could be used more widely; (2) fertility may be improved if ovulation were more predictable than it normallyis;and (3) since conception would require fewer matings the risk of injury to personnel and/or animals would be minimized and reproductive tract infections would be minimized. Hopefully, as a result more effort could be directed toward genetic improvement of horses for characteristics desirable to the race horses, pack animals, riding horses or work horses by selection and artificial insemination as practiced in cattle. Moreover, there were indications from my data that some mares which had irregular cycles during the breeding season would 100 respond to PGan' Mares with prolonged luteal function (36 days) returned to estrus and ovulated after sc or iu PGF2a~ Similarly, when Kenney et a1. (1975) treated barren, maiden or postpartum mares (with an average anestrus period of 73 days during the breeding season) with PGFZa (5 mg Tham salt, sc or im), the mares began estrus in 4.4 days. Of mares checked by palpation, 86% ovulated 8.9 days later whether or not estrual behavior was exhibited. The mares were bred 7 days after PGFZa’ and 55% of these mares were pregnant after PGFZa' Therefore, in addition to potential for synchronization of ovulation, PGan has considerable potential to initiate estrus in mares which fail to have luteal regression, thereby resulting in prolonged luteal function and the absence of estrus. A note of caution is warranted, however. In addition to the luteolytic effect of PGFZQ, it can cause uterine contractions, bronchoconstriction, increased intraocular pressure, and venocon- striction (Oesterling, 1972). Therefore, use of PGFZa in pregnant mares or mares with respiratory or cardiovascular disease would not be warranted. Care in handling of PGFZQ by humans is also advisable. PGFZa is available from Upjohn company on an experi- mental basis only, and is not available for purchase at the present time. CONCLUSIONS Luteolytic Effects of PGFZa in Mares Abrupt luteolysis occurred when mares were treated with 5 to 15 mg PGFZa-Tham salt on day 5 to 9 after ovulation because in all cases blood progesterone decreased significantly within 12 to 24 hours and continued to decrease until the mares exhibited behavioral estrus about 2.5 days later. PGan was equally effec- tive whether administered into the uterus or systemically (sc). After PGFZa’ the duration of estrus and the interval between ovulation and end of estrus were not different from those of control mares. Ovulation occurred 1 to 2 days before the end of estrus in the estrous cycle of control and PGan-treated mares alike; in treated mares ovulation occurred relatively syn- chronously at 7 to 9 days after PGFZQ. Less than 5 mg of PGan-Tham salt on day 7, or 10 mg given on day 3 did not consistently cause complete luteolysis. Although there was a significant decrease in progesterone in all mares 24 hours after treatment, this decrease did not continue in some mares. PGFZa given the day after ovulation did not cause luteolysis, and in these mares progesterone continued to increase as it did in control mares after ovulation, reaching maximal values about 7 days later. During the control cycle (I) in the first experiment, pro- gesterone decreased from 17.1 to 1.4 ng/ml during the 5 days before 101 102 estrus. Two days before estrus when progesterone was 5.3 ng/ml, estradiol had increased 2-fold to 5.6 pg/ml; LH was significantly increased the following day. During estrus while progesterone remained below 1 ng/ml estradiol, androstenedione and LH increased to peak concentrations. Androstenedione and estradiol decreased to diestrual values by the end of estrus; LH remained elevated until 1 to 2 days after ovulation and decreased gradually for 7 to 9 days while progesterone increased. Estrone concentrations did not change significantly during the estrous cycle. Plasma progesterone changes during the estrous cycle after PGan differed only slightly from comparable changes during con- trol cycles. The progesterone decrease after PGan--although more rapid (P < .01) during the first 24 hours-~was not signifi- cantly different from that which occurred during natural luteoly- sis during the 3 days before estrus. When luteolysis was complete at about 48 to 72 hours, basal progesterone was below 1 ng/ml and estrus began. As with progesterone changes, the changes in estradiol and LH were not influenced by iu PGFZa administration. During both cycles estradiol increases preceded or accompanied the LH increase before ovulation, and like estradiol, androstenedione also peaked l or 2 days before ovulation. Again LH reached maximal values near ovulation, remained high for 48 to 60 hours after ovulation and did not decrease until progesterone began to increase. Estrone did not change significantly. These hormone changes during the eStrous cycle after iu PGan were substantially the same as those 103 that occurred during luteolysis, follicular growth, ovulation and CL function during a control cycle. The second control cycle was identical to the first in every measured criterion. I therefore believe that for these variables, PGan-Tham salt had no residual effect on subsequent estrous cycles. Mares were fertile during the estrus which followed PGan; and the foals born to these mares were physically normal. Thus, PGan did not depress fertility, and has potential for estrus control and synchronization of ovulation in brood mares. Treat- ment of the noncycling mare with PGan also has potential; PGan caused luteolysis in mares with prolonged luteal function during the breeding season. LITERATURE CITED 104 LITERATURE CITED Abraham, G. E. 1974. Ovarian and adrenal contribution to periph- eral androgens during the menstrual cycle. J. Clin. Endocrinol. Metabol. 39: 340-346. Allen, E. and E. A. Doisey. 1924. The induction of a sexually mature condition in immature females by injection of the ovarian follicular hormone. Am. J. Physiol. 69: 577-588. Allen, W. E. and J. C. Hadley. 1973. Peripheral blood levels of progesterone in pony mares during the oestrous cycle and early pregnancy. Vet. Rec. 93: 77 (abst). Allen, W. R. 1972. Some recent studies upon reproduction in the mare. London, Henderson Group One. Allen, W. R. and R. M. Moor. 1972. The origin of the equine endometrial cups. 1. Production of PMSG by fetal thropho- blast cells. J. Reprod. Fert. 29: 313-316. Allen, W. R. and P. D. Rossdale, 1973. A preliminary study upon the use of prostaglandins for inducing oestrus in non- cycling thoroughbred mares. Equine Vet. J. 5: 137-140. Allen, W. R. and L. E. A. Rowson. 1973. Control of the mare's oestrous cycle by prostaglandins. J. Reprod. Fert. 33: 539-543. Allen, W. R., F. Stewart, M. J. Cooper, R. C. Crowhurst and P. D. Rossdale. 1975. Induction and synchronization of oestrus with synthetic prostaglandin analogues. J. Reprod. Fert., Supp. 23: in press. Anderson, L. L., K. P. Bland and R. M. Melampy. 1969. Compara- tive aspects of uterine-luteal relationships. Rec. Prog. Horm. Res. 25: 57-104. Andrews, F. W. and F. F. McKenzie. 1941. Estrus, ovulation and related phenomena in the mare. Univ. of Mo. Agric. Exp. Stat. Res. Bull. 329: 1-117. Armstrong, 0. T. and R. O. Greep. 1965. Failure of deciduomal response to uterine trauma and effects of LH upon estrogen secretion in rats with ovaries luteinized by exogenous gonadotropins. Endocrinol. 76: 246-254. 105 106 Armstrong, 0. T. and W. Hansel. 1959. Alteration of bovine estrous cycle with oxytocin. J. Dairy Sci. 42: 533-542. Asdell, S. A. 1946. In: Patterns of Mammalian Reproduction. Ithaca, Comstock PuElishing Co., Inc. Bain, A. H. 1957. Estrus and infertility of the Thoroughbred mare hiAustralasia. J. Am. Vet. Med. Assoc. 131: 179- 185. Baird, D. T., P. E. Burger, G. D. Heavon-Jones and R. J. Scara- muzzi. 1974. The site of secretion of androstenedione in nonpregnant women. J. Endocrinol. 63: 201-212. Baird, D. T., J. R. Goding, Y. Ichikawa and J. A. McCracken. 1972. The secretion of steroids from the auto-transplanted ovary in the ewe spontaneously and in response to systemic gonadotrophin. J. Endocrinol. 42: 238-299. Beck, T. W. and E. M. Convey. 1974. LH after estradiol and GnRH in spayed heifers. J. Anim. Sci. 39: 198 (abst). Beck, T. W. and J. J. Reeves. 1973. Serum luteinizing hormone (LH) in ewes treated with various dosages of 17 beta estradiol at three stages of the anestrous season. J. Anim. Sci. 36: 566-570. Black, 0. L. and R. T. Duby. 1965. Effects of oxytocin, epine- phrine and atropine on the estrous cycle of the cow. J. Reprod. Fert. 9: 3-8. Bland, K. P., E. W. Horton and N. L. Poyser. 1971. Levels of prostaglandin an in the uterine venous blood of sheep during the oestrous cycle. Life Sci. 10: 509-517. Blatchley, F. R., B. T. Donavan, N. L. Poyser, E. W. Horton, C. J. Thompson, and M. L05. 1971. Identification of Prosta- glandin an in the utero-ovarian blood of guinea-pig after treatment with oestrogen. Nature 230: 243-244. Boyns, A. R., G. E. Jones, E. T. Bell, 0. W. Christie and M. F. Parkes. 1972. Development of a radioimmunoassay for canine luteinizing hormone. J. Endocrinol. 55: 279-291. Braselton, W. E., Jr. and W. H. McShan. 1970. Purification and properties of follicle-stimulating and luteinizing hor- mones from horse pituitary glands. Arch. Biochem. & Biophysics 139: 45-48. Britt, J. H., R. J. Kittok and D. S. Harrison. 1974. Ovulation, estrus and endocrine response after GnRH in early post- partum cows. J. Anim. Sci. 39: 915-925. 107 Britt, J. H. and L. C. Ulberg. 1971. The therapeutics of pro- gestogens. Vet. Scope 16: 1-9. Burkhardt, J. 1947a. Transition from anoestrus in the mare and the effects of artificial lighting. J. Agric. Sci. 37: 64-68. Burkhardt, J. 1947b. Anestrus in the mare and its treatment with estrogens. Vet. Rec. 59: 341-342. Caldwell, B. V., R. M. Moor and R. A. S. Lawson. 1968. Effects of sheep endometrium grafts and extracts on the length of pseudopregnancy in the hysterectomized hamster. J. Reprod. Fert. 17: 567-569. Carr, 8. R., G. Mikhial and G. L. Flickinger. 1971. Column chromatography of steroids on Sephadex LH-20. J. Clin. Endocr. 33: 358-360. Casida, L. E. 1934. Production of ovulation by gonadotrophic extracts. Endocrinol. 18: 714-720. Casida, L. E., R. K. Meyer, W. H. McShan and W. Wisnicky. Effects of pituitary gonadotrophins on the ovaries and the induc- tion of superfecundity in cattle. Am. J. Vet. Res. 4: 76-94. Chance, M. R. A., I. W. Rowlands and F. G. Young. 1939. Species variations in thyrotrophic, gonadotrophic and prolactin activities of the anterior hypophyseal tissue. J. Endo- crinol. 1: 239-240. Channing, C. P. 1969. Tissue culture of equine ovarian cell types: culture methods and morphology. J. Endocrinol. 43: 381-390. Channing, C. P. 1975. HCG receptors on equine granulosa cells. Personal communication. Channing, C. P. and S. Kansmerman. 1974. Binding of gonadotropins to ovarian cells. Biol. Reprod. 70: 179-198. Cole, H. H., C. E. Howell and G. H. Hart. 1931. The changes occurring in the ovary of the mare during pregnancy. Anat. Rec. 49: 199-209. Convey, E. M. 1973. Neuroendocrine relationships in farm animals: a review. J. Anim. Sci. 37: 745-757. Davidson, W. F. 1947. The control of ovulation in the mare with reference to insemination with stored semen. J. Agric. Sci. 37: 287-290. 108 Day, F. T. 1940. Clinical and experimental observations on reproduction in the mare. J. Agric. Sci. 30: 244-261. Dohn, P. M. and H. H. Cole. 1971. Sex steroids in the urine of mares. J. Anim. Sci. 33: 252 (abst). Douglas, R. H., M. C. Garcia and O. J. Ginther. 1974. Effect of equine pituitary antiserum and PGFZQ in hysterectomized mares. J. Anim. Sci. 39: 989 (abst). Douglas, R. H. and O. J. Ginther. 1972. Effect of prostaglandin an on length of diestrus in mares. Prostaglandins 2: 265-268. Downey, B. R., D. S. Irvine, W. G. Parker and J. J. Sullivan. 1974. Estrus and ovulation time in GnRH treated mares. J. Anim. Sci. 39: 206 (abst). Dupon, C. and M. H. Kim. 1973. Peripherial plasma levels of testosterone, androstenedione and oestradiol during the rat oestrous cycle. J. Endocrinol. 59: 653-654. Eckstein, P. and S. Zuckerman. 1956a. Morphology of the repro- ductive tract. In: Marshall's Physiology of Reproduction, edited by A. S. Parkes. New York, Longsman, Green and Co., Inc., 43-155. Eckstein, P. and S. Zuckerman. 1956b. The oestrous cycle in the mammalia. In: Marshall's Physiology of Reproduction, edited by A. S. Parkes. New York, Longsman, Green and Co., Inc., 226-396. Espey, L. L. 1974. Ovarian proteolytic enzymes and ovulation. Biol. Reprod. 10: 216-235. Everitt, B. J. and J. Herbert. 1969. Adrenal glands and sexual receptivity in female rhesus monkeys. Nature (London) 222: 1065-1066. Fevold, H. L. 1939. Extraction and standardization of pituitary follicle-stimulating and luteinizing hormones. Endocrinol. 24: 435-446. Ganjam, S. 1975. Fertility in mares after Prostaglandin F2a° Personal Comnunication. Ganong, W. F. 1969. Neural centers regulating visceral function. In: Review of Medical Physiology, Los Altos, Lange Medical Publications, 173-192. 109 Gill, J. L. and H. D. Hafs. 1971. Analysis of repeated measure- ments of animals. J. Anim. Sci. 33: 331-336. Ginther, 0. J. 1971. Some factors which alter estrous cycle in mares. J. Anim. Sci. 33: 1158 (Abst). Ginther, 0. J. 1974. Internal regulation of physiological processes through local venoarterial pathways: a review. J. Anim. Sci. 39: 550-564. Ginther, O. J. and N. L. First. 1971. Maintenance of the corpus luteum in hysterectomized mares. Am. J. Vet. Res. 32: 1687-1692. Ginther, O. J., M. H. Pineda, B. C. Wentworth and L. Nuti. 1974. Rate of disappearance of exogenous LH from the blood of mares. J. Anim. Sci. 39: 397-403. Ginther, O. J. and B. C. Wentworth. 1974. Effect of a synthetic gonadotrOpin releasing hormone on plasma concentrations of luteinizing hormone in ponies. Am. J. Vet. Res. 35: 79-81. Girard, A., G. Sandulesco, A. Fridenson and J. J. Rutgers. 1932. Sur une nouvelle hormone sexuelle cristallisée retiree de l'urine des juments gravides. C. R. Acad. Sci. (Paris) 194: 909-911. Glencross, R. G., I. B. Minro, B. E. Senior, and G. S. Pope. 1973. Concentrations of oestradiol-178, oestrone and progesterone in jugular venous plasma of cows during the oestrous cycle and in early pregnancy. Acta. Endocrinol. 73: 374-384. Goding, J. R., M. A. de Blockley, J. M. Brown, K. J. Catt and I. A. Cumming. 1970. The role of oestrogen in the control of the oestrous cycle in the ewe. J. Reprod. Fert. 21: 368-369. Gospodarowicz, D. 1972. Purification and physiochemical proper- ties of the pregnant mare serum gonadotropin (PMSG). Endocrinol. 91: 101-106. Gwazdauskas, F. C., W. W. Thatcher and C. J. Wilc0X. 1972. Adrenocorticotropin alteration of bovine peripheral plasma concentrations of cortisol, corticosterone, and proges- terone. J. Dairy Sci. 55: 1165-1169. Hafs, H. 0., T. M. Louis, P. A. Noden and W. D. Oxender. 1974. Control of the estrous cycle with prostaglandin FZG in cattle and horses. J. Anim. Sci. 38: (Supp. I) 10-21. 110 Hansel, W. and S. E. Echternkamp. 1972. Control of ovarian function in domestic animals. Amer. Zool. 12: 225-243. Hartree, A. S., J. 8. Mills, R. A. S. Welch and M. J. Thomas. 1968. Fractionation of protein hormones from horse pituitary glands. J. Reprod. Fert. 17: 291-303. Heape, W. 1900. The "sexual season" of mammals, and the relation of the "proestrum" to menstruation. Quart. J. Micr. Sci. 44: 1-17. Hendricks, D. M., H. D. Guthrie and D. L. Handlin. 1972. Plasma estrogen, progesterone and luteinizing hormone during the estrous cycle in pigs. Biol. Reprod. 6: 210-218. Hernandez, M. V., K. M. Etta, E. P. Reineke, W. D. Oxender and H. D. Hafs. 1972. Thyroid function in the prenatal and neonatal bovine. J. Anim. Sci. 34: 780-785. Hilliard, J., R. Penardi and C. H. Sawyer. 1967. A functional role for 20a-hydroxypregn-4-en-3-one in the rabbit. Endocrinol.‘ 80: 901-909. Hughes, J. P., G. H. Stabenfeldt and J. W. Evans. 1972. Estrous cycle and ovulation in the mare. J. Am. Vet. Med. Assoc. 161: 1367-1374. Hobson, W. C. and W. Hansel. 1972. Plasma LH levels after ovari- ectomy, corpus luteum removal and estradiol administration in cattle. Endocrinol. 91: 185-190. Horton, R. 1965. Estimation o; androstenedione in human peripherial blood with 5s thiosemicarbazide. J. Clin. Endocrinol. Metabol. 25: 1237-1245. Hutchison, J.ES.M., D. T. Armstrong and R. O. Greep. 1968. A comparison of luteinizing hormone from different species using the ventral prostate assay with and without delay vehicle and the ovarian ascorbic acid depletion assay. J. Endocrinol. 40: 231-235. Kenney, R. M., V. K. Ganjam, J. Lauderdale and W. Cooper. 1975. Control of the mares oestrous cycle with PGan. J. Reprod. Fert., Supp. 23: in press. Knudsen, 0. and W. Velle. 1961. Ovarian estrogen levels in the nonpregnant mare: relationship to histological appear- ance of the uterus and to clinical status. J. Reprod. Fert. 2: 130-137. 111 Landefield, T. 0., H. J. Grimek and W. H. McShan. 1972. End group and carbohydrate analysis of equine LH. Biochem. Biophys. Res. Comm. 46: 463-469. Lauderdale, J. 1973. Effects of 200 mg PGan-Tham salt in one mare. Personal Communication. Lauderdale J. and P. Miller. 1974. Sweating after PGFZa or epinepherine in mares. Personal Communication. Loeb, L. 1927. The effects of hysterectomy on the system of sex organs and the periodicity of the sexual cycle in the guinea pig. Am. J. Physiol. 83: 202-224. Louis, T. M., H. D. Hafs and D. A. Morrow. Estrus and ovulation after PGan in cows. J. Anim. Sci. 35: 1121 (abst.) Louis, T. M., H. D. Hafs and D. A. Morrow. 1974. Intrauterine administration of prostaglandin an in cows: progesterone, estrogen, LH, estrus and ovulation. J. Anim. Sci. 38: 347-353. Louis, T. M., H. D. Hafs and B. E. Seguin. 1973. Progesterone, LH, estrus and ovulation after Prostaglandin an in heifers. Proc. Soc. Exp. Biol. Med. 143: 152-155. Loy, R. G. 1970. Management and other factors affecting effi- ciency in mares. Proc. 3rd Tech. Conf. Art.Insem. & Reprod. Chicago, Nat. Assn. Art. Breed., 60-64. Loy, R. G. and J. P. Hughes. 1966. The effects of human chorionic gonadotropin on ovulation, length of estrus, and fertility in the mare. Cor. Vet. 56: 41-50. Loy, R. G. and S. M. Swan. 1966. Effects of exogenous progesto- gens on reproductive phenomena in mares. J. Anim. Sci. 25: 821-826. Lukaszewska, J. H. and W. Hansel. 1970. Extraction and partial. purification of luteolytic activity from bovine endome- trial tissue. Endocrinol. 86: 261-270. Marion, G. B. and H. T. Gier. 1971. Ovarian and uterine embryo- genesis and morphology of the nonpregnant female mammal. J. Anim. Sci., Supp. I, 32: 24-47. Marshall, F. H. A. 1936. The Croonian lecture: sexual periodicity and the causes which determine it. Phiolos. Trans. 8. 226: 423-433. 112 Mayer, D. T., F. N. Andrew and F. F. McKenzie. 1940. The estrin content of the follicular fluid and urine of the mare and its relation to phenomena of the estrual cycle. Endocrinol. 27: 867-872. McCracken, J. A., J. C. Carlson, M. E. Glew, J. R. Goding, D. T. Baird, K. Green and B. Samuelsson. 1972. Prostaglandin FZQ identified as a luteolytic hormone in sheep. Nature New Biology 238: 129-134. McCracken, J. A., M. E. Glew and R. J. Scaramuzzi. 1970. Corpus luteum regression induced by prostaglandin an. J. Clin. Endocrinol. Metabol. 30: 544-546. McShan, W. H. and R. K. Meyer. 1941. Effectiveness of heme in the augmentation of gonadotropic extracts from different sources. Endocrinol. 28: 694-700. Meyer, R. K. and W. K. McShan. 1941. Effects of increasing the number of daily injections of gonadotrophic prepara- tions on the ovaries of immature rats. Endocrinol. 29: 31-34. Miyake, T. 1974. Blood concentration and interplay of pitui- tary and gonadal hormones governing the reproductive cycle in female mammals. In: Reproductive Physiology, edited by R. O. Greep. Baltimore, University Park Press, 155- 178. Moghissi, K. S., F. N. Syner and T. N. Evans. 1972. A composite picture of the menstrual cycle. Am. J. Ob. Gynecol. 114: 405-418. Mongkonpunya, K., Y. C. Lin, P. A. Noden, W. D. Oxender and H. D. Hafs. 1975. Androgens in the bovine fetus and dam. Proc. Soc. Exp. Biol. Med. 148: 489-493. Morrell, A. G., G. Gregoriadis, I. H. Scheinberg, J. Hickman and G. Ashwell. 1971. The role of sialic acid in deter- mining the survival of glycoproteins in circulation. J. Biol. Chem. 246: 1461-1467. Nett, T. M., D. W. Holtan, and V. L. Estergreen. 1973. Plasma estrogens in pregnant and postpartum mares. J. Anim. Sci. 37: 962-970. Nishikawa, Y. 1959. Studies on reproduction in horses. Tokyo, Japan Racing Assoc. Niswender, G. D. 1973. Influence of the site of conjugation or the specificity of antibodies to progesterone. Steroids 22: 413-424. 113 Niswender, G. D., A. R. Midgley, S. E. Monroe and L. E. Reichart, Jr. 1968. Radioimmunoassay for rat luteinizing hormone with antiovine LH serum and ovine LH1 -I*. Proc. Soc. Exp. Biol. Med. 128: 807-811. Oesterling, T. 0., W. Morozowich and T. J. Roseman. 1972. Prostaglandins. J. Pharm. Sci. 61: 1861-1895. Osborne, V. E. 1966. An analysis of the pattern of ovulation as it occurs in the annual reproductive cycle of the mare in Australia. Austral. Vet. J. 42: 149-154. Oxender, W. 0., P. A. Noden, T. M. Louis and H. D. Hafs. 1974. A review of prostaglandin F2a for ovulation control in cows and mares. Am. J. Vet. Res. 165: 997-1001. Parlow, A. F. 1963. Species differences in luteinizing hormone (LH, ICSH) as revealed by the slope in the prostate assay. Endocrinol. 73: 509-512. Pattison, M. L., C. L. Chen, S. T. Kelley and G. W. Brandt. 1974. Luteinizing hormone and estradiol in peripherial blood of mares during estrous cycle. Bio. Reprod. 11 :245-250. Pharriss, B. B., S. A. Tillson and R. R. Erickson. 1972. Prosta- glandins in luteal function. Rec. Prog. Horm. Res. 28: 51-89. Pharriss, B. B. and L. J. Wyngarden. 1969. The effect of PGFZG on the progesterone content of ovaries from pseudopreg- nant rats. Proc. Soc. Exp. Biol. Med. 130: 92-94. Piper, P. J., J. R. Vane and J. H. Wyllie. 1970. Inactivation of prostaglandins by the lungs. Nature 225: 600-604. Plotka, E. 0., D. M. Witherspoon and C. W. Foley. 1972. Luteal function in the mare as reflected by progesterone concen- trations in peripheral blood plasma. Am. J. Vet. Res. 33: 917-920. Poyser, N. L. 1973. The physiology of prostaglandins. Clin. Endocrinol. Metab. 2: 393-410. Raacke, I. 0., A. J. Lostroh, J. M. Boda and C. H. Li. 1957. Some aspects of the characterization of pregnant mare serum gonadotrophin. Acta Endocrinol. 26: 377-387. Rao, C. V. 1974. Prostaglandin receptors in bovine corpus luteum cell membranes. Prostaglandins 6: 533-536. Reichert, L. E., Jr., and A. E. Wilhelmi. 1966. Preparation of equine luteinizing hormone. Endocrinol. 76: 762-765. 114 Roberts, S. J. 1971. Infertility in the mare. In: Veterinary Obstetrics and Genital Diseases. Ann Arbor, Edwards Bro., Inc., 512-550. Rohlf, F. J. and R. S. Sokal. 1969. Statistical Tables. San Francisco, W. H. Freeman and Co. Rondell, P. 1970. Biophysical aspects of ovulation. Biol. Reprod., Supp. 2: 64-89. ' Rowlands, I. W. 1963. Levels of gonadotrOpins in tissues and fluids, with emphasis on domestic animals. In: Gonado- tropins, edited by H. H. Cole. San Francisco, Freeman and Co., 74-112. Rowson, L. E. A., H. R. Tervit and A. Brand. 1972. 'The use of prostaglandin for synchronization of oestrus in cattle. J. Reprod. Fert. 29: 145 (abst). Saxena, B. B., W. H. McShan and R. K. Meyer. 1962. Purification of follicle-stimulating hromone from horse anterior pitu- itary glands. Biochem. Biophys. Acta. 65: 394—402. Scarmuzzi, R. J., B. V. Caldwell and R. M. Moon. 1970. Radio- immunoassay of LH and estrogen during the estrous cycle of the ewe. Biol. Reprod. 3: llO-ll9. Sharp, 0. C. 1972. Changes in peripheral progesterone throughout the estrous cycle of pony mares. Biol. Reprod. 7: 135 abst). Short, R. V. 1957. Steroids related to progesterone in the human and mare placenta. J. Endocrinol. 15: i (abst.) Short, R. V. 1959. Progesterone in blood: IV. Progesterone in the blood of mares. J. Endocrinol. 19: 207. Short, R. V. 1960. Steroids present in the follicular fluid of the mare. J. Endocrinol. 20: 147-156. Short, R. V. 1961. Steroid concentrations in the follicular ' fluid of mares at various stages of the reproductive cycle. J. Endocrinol. 22: 153-163. Short, R. V. 1964. Ovarian steroid synthesis and secretion jp_ vivo. Rec. Prog. Horm. Res. 20: 303-340. Short, R. V. 1972. Role of hormones in sex cycles. In: Hormones in Reproduction, edited by C. R. Austin and R. V. Short. Cambridge University Press, 42-72. 115 Smith, I. 0., J. M. Bassett and T. Williams. 1970. Progesterone concentrations in the peripheral plasma of the mare during the oestrous cycle. J. Endocrinol. 47: 523-524. Smith, M. S. and L. E. McDonald. 1974. Serum levels of luteiniz- ing hormone and progesterone during the estrous cycle, pseudopregnancy and pregnancy in the dog. Endocrinol. 94: 404-412. Squires, E. L., B. C. Wentworth and O. J. Ginther. 1974. Pro- gesterone concentration in blood of mares during the estrous cycle, pregnancy and after hysterectomy. J. Anim. Sci. 39: 759-767. Stabenfeldt, G. H., J. P. Hughes and J. W. Evans. 1972. Ovarian activity during the estrous cycle of the mare. Endocrinol. 90: 1379-1384. Stellflug, J. N., T. M. Louis, R. C. Gorewit, W. D. Oxender and H. D. Hafs. 1975. Luteolysis after PGan in hysterec- tomized cattle. J. Anim. Sci. 41: in press. Sullivan, J. J. 1972. Some aspects of reproduction in the horse and present status of artificial insemination with frozen stallion semen. In: Technician's Manual for Artificial Insemination (AI) of Mares. Madison, American Breeders Service, Inc. Sullivan, J. J., W. G. Parker and L. L. Larson. 1973. Duration of estrus and ovulation time in nonlactating mares given human chorionic gonadotropin during three successive estrous periods. J. Am. Vet. Med. Assoc. 162: 895-898. Swanson, L. V., H. D. Hafs and D. A. Morrow. 1972. Ovarian characteristics and serum LH, prolactin, progesterone and glucocorticoids from first estrus to breeding size in holstein heifers. J. Anim. Sci. 34: 284-293. Thatcher, W. W. 1974. Use of mini-columns for Sephadex LH-20 separation of estrone, estradiol and estriol. Personal Communication. Thorburn, G. D., J. M. Bassett and I. D. Smith. 1969. Proges- terone concentration in the peripheral plasma of sheep during the oestrous cycle. J. Endocrinol. 45: 459-469. U.S. Department of Commerce, Bureau of the Census. 1960. Historical Statistics of the United States: Colonial Times to 1957. Chapter K: Agriculture, 289-351. 116 U.S. Department of Commerce, Bureau of the Census. 1973. 1969 Census of Agriculture. Chapter 5: Livestock, Poultry, Livestock and youltry Products, 18. U.S. Department of Commerce, Bureau of the Census. 1974a. Statistical Abstract of the United States. Participants in Selected Outdoor Activities, 210. ' U.S. Department of Commerce, Bureau of the Census. 1974b. Statistical Abstract of the United States. Selected Recreational Activities, 211. U.S., Food and Agriculture Organization. 1972a. Trade and Pro- duction Yearbook, Vol. 26. Livestock Numbers and Products, 183-185. U.S., Food and Agriculture Organization. 1972b. Trade and Produc- tion Yearbook, Vol. 26. Meat Production from Slaughtered Animals, 200. Van Niekerk, C. H. 1967. Pattern of the oestrus cycle of mares. I. The breeding season. J. South African Vet. M. A. 38: 295-298. Wettemann, R. P., H. D. Hafs, L. A. Edgerton and L. V. Swanson. 1972. Estradiol and progesterone in blood serum during the bovine estrous cycle. J. Anim. Sci. 34: 1020-1024. Whitmore, H. L., B. C. Wentworth and D. J. Ginther. 1973. Cir- culating concentrations of luteinizing hormone during estrous cycles of mares as determined by radioimmunoassay. Am. J. Vet. Res. 34: 631-636. Wilson, L., Jr., R. J. Cenedella, R. L. Butcher, and E. K. Inskeep. 1972. Levels of prostaglandins in the uterine endometrium during the ovine estrous cycle. J. Anim. Sci. 34: 93-99. Younglai, E. V. 1971. Steroid content of the equine ovary during the reproductive cycle. J. Endocrinol. 50: 589-597. Younglai, E. V. and R. V. Short. 1970. Pathways of steroid biosynthesis in the intact Graafian follicle of mares in oestrus. J. Endocrinol. 47: 321-331. .5. AN STQTE UNIV. "‘111111111111111111111 1111111113 1 14