Possible mechanisms for reduction of circulating concentrations of progesterone by interferon-\g=a\ in cows: effects on hyperthermia, luteal cells, metabolism of progesterone and secretion of LH C. M. Barros, J. G.
Betts, W. W. Thatcher and P. J. Hansen
Departamento de Farmacologia, Universidade Estadual Paulista, Botucatu, Brazil *Dairy Science Department, University of Florida, Gainesville, Florida 32611-0701, U.S.A. (Requests for offprints should be addressed to P. J. Hansen) received 6 September 1991 ABSTRACT
Experiments were performed to determine the mechby which recombinant bovine interferon-\g=a\I1 (rbIFN-\g=a\)causes an acute reduction in plasma concentrations of progesterone. In experiment 1, administration of a prostaglandin synthesis inhibitor blocked rbIFN-\g=a\-induced hyperthermia but did not prevent the decline in plasma concentrations of progesterone. The decline in progesterone concentrations caused by
did not affect basal and LH-induced release of progesterone from cultured luteal slices, indicating that rbIFN-\g=a\is unlikely to affect luteal function directly. Injection of rbIFN-\g=a\did, however, cause a decrease in plasma concentrations of LH in ovariectomized cows (experiment 4) that coincided temporally with the decrease in progesterone concentrations seen in cows having a functional corpus luteum. The present results strongly suggest that rbIFN-\g=a\acts to reduce secretion of progesterone by interfering with pituitary support for luteal synthesis of progesterone. The finding that rbIFN-\g=a\ can inhibit LH secretion implies that interferon-\g=a\molecules should be considered among the cytokines that can regulate hypothalamic or pituitary function. Journal ofEndocrinology (1992) 133, 175\p=n-\182
INTRODUCTION
feron called trophoblast protein-1 (oTP-1 and bTP1 for sheep and cow respectively) (Roberts et al. 1990; Bazer, Thatcher, Hansen et al. 1991). Both oTP-1 and bTP-1 extend luteal function by reducing endometrial secretion of prostaglandin (PG) F2a (Godkin, Bazer, Thatcher & Roberts, 19846; Vallet, Bazer, Fliss & Thatcher, 1988; Helmer, Gross, Hansen & Thatcher, 1989a; Helmer, Hansen, Thatcher et al. 19896; Salamonsen, Manikhot, Healy & Findlay, 1989; Barros, Plante, Thatcher &
anism
rbIFN-\g=a\was, therefore, not a direct consequence of the associated hyperthermia or of pathways mediated through prostaglandin synthesis. It is also unlikely that rbIFN-\g=a\ acts to increase the clearance of progesterone since injection of rbIFN-\g=a\ did not decrease plasma concentrations of progesterone in ovariectomized cows given an intravaginal implant of progesterone (experiment 2). In experiment 3, rbIFN-\g=a\
In the last several years, it has become clear that various regulatory molecules usually associated with the immune system play important roles as endo¬ crine or paracrine regulatory molecules in other tis¬ sues, especially in the reproductive tract and central nervous system (McCann, Rettori, Milenkovic et al.
1989; Hunt, 1989; Shibata, 1990; Roberts, Cross,
Farin et al. 1990; Tartakovsky & Ben-Yair, 1991). One of the best-studied examples is provided by the regulation of the luteal lifespan by the ruminant em¬ bryo. In sheep, cattle and goats, the embryo pre¬ vents destruction of the corpus luteum during early pregnancy by secreting a type I trophoblast inter-
Hansen, 1991).
role of the trophoblast proteinregulate uterine prostaglandin secretion, these molecules do possess biological activity characteristic of type 1 interferons, including While the
major
1 molecules is to
antiviral and Torres, Vallet
immunosuppressive activity (Pontzer,
al. 1988; Newton, Vallet, Hansen & Bazer, 1989; Plante, Hansen, Thatcher et al. 1990; Skopets, Li, Thatcher et al. 1992). Interestingly, administration of non-embryonic interferon-a mol¬ ecules can mimic the antiluteolytic effect of bTP-1 and oTP-1 when given to cyclic animals (Plante, Hansen & Thatcher, 1988; Plante, Hansen, Martinod et al. 1989; Stewart, Flint, Lamming et al. 1989; Newton, Martinod, Hansen et al. 1990) and can reduce secretion of PGF2a from endometrial tissue in vitro (Salamonsen et al. 1989; Barros et al. 1991) and in vivo (Piante, Thatcher & Hansen, 1991). One of the paradoxical effects of systemic administration of recombinant bovine interferon-a, 1 (rbIFN-a) to cattle is that it extends the lifespan of the corpus luteum but also causes an acute decline in serum concentrations of progesterone which reaches a nadir at 4-6 h after injection and returns to preinjection values by 8 h after injection (Newton et al. 1990; Plante et al. 1991). The transitory decline in circulating concentrations of progesterone and the hyperthermia-induced rbIFN-a are of greatest magnitude following an initial injec¬ tion. These effects of rbIFN-a decline and then dis¬ appear during a regimen of repeated injections (Newton et al. 1990; Plante et al. 1991). Interferon- reduces circulating concentrations of steroid hormones in other systems as well. Administration of interferon-a reduced circulating concentrations of oestradiol and progesterone in women (Kaupilla, Cantell, Janne et al. 1982) and testosterone in men (Orava, Cantell & Vikho, 1986). Since the timing and refractoriness of the effect of interferon-a on body temperature parallels those for serum concentrations of progesterone, it is possible that effects on progesterone are caused indirectly by changes in hyperthermia. Perhaps temperature alters the function of the hypothalamo-pituitary-ovarian axis (Flowers & Day, 1990) or clearance of progester¬ one. Interferon- has been reported to increase drug clearance in a body-temperature-related manner (Ehrsson, Eksborg, Wallin et al. 1990). It is also poss¬ ible that interferon-a alters progesterone secretion through mechanisms unrelated to hyperthermia. One possibility is that rbIFN-a acts directly on luteal cells to inhibit progesterone secretion. Pretreatment of cultured porcine Leydig cells with interferon-a reduced human chorionic gonadotrophin (hCG)stimulated testosterone secretion (Orava, 1989). Another possibility is that rbIFN-a acts directly on the hypothalamic-pituitary axis to reduce the secretion of luteinizing hormone (LH) and cause a subsequent decline in progesterone secretion. These possible mechanisms for the effects of rbIFN-a on the plasma concentrations of progesterone are examined in this paper.
MATERIALS AND METHODS
et
Materials
Bovine interferon-a, 1 and placebo (buffered mannitol solution) were donated by Ciba-Geigy (Basel, Switzerland). The intravaginal progesterone implants used in the study were CIDR@-B devices obtained from Carter Holt Harvey Agricultural Division, Hamilton, New Zealand. Each implant contained 1-9 g progesterone. Flunixin meglumine (FM) was from Schering-Plough (Kenilworth, NJ, U.S.A.). Lutalyse was from Upjohn (Kalamazoo, MI, U.S.A.).
Dulbecco's modified minimal essential medium (DMEM) and Ham's nutrient mixture F-12 medium, (F12) were obtained from Sigma (St Louis, MO, U.S.A). LH (USDA-bLH-B-5) and antiserum to bov¬ ine LH (USDA-309-684P) were a gift from Dr D. J. Bolt (USDA, Beltsville Agricultural Research Center,
Beltsville, MD, U.S.A.).
Experiment 1: interrelationship between hyperthermia and plasma concentrations of progesterone Seven non-lactating Holstein cows (466-597 kg) were given two injections of 25 mg PGF2a (Lutalyse) 11 days apart to synchronize oestrus. Each cow received the following series of treatments at 3-day intervals starting on the morning of day 7 of the oestrous cycle: placebo (4 ml), rbIFN-a (20mg/cow, i.m., 4 ml); rbIFN-a (20 mg, i.m., 4 ml) and FM (a prostaglandin synthesis inhibitor, 22 mg/kg, i.m., two injections at 0 and 5 h relative to injection of rbIFN-a); and FM alone. Treatments were given according to the schedule dictated by a 2x4 Latin square design, where one square was incomplete. At 0,2,4,6,8,10 and 24 h after injection of rbIFN-a, rectal temperatures were measured and a 10 ml sample of jugular blood was collected into heparinized tubes via venipuncture for subsequent determination of progesterone concen¬ trations in plasma. Experiment 2: clearance of progesterone In this experiment, it was reasoned that possible changes in plasma progesterone concentrations induced by rbIFN-a in ovariectomized cows given
progesterone implants would be indicative that rbIFN-a affects clearance of progesterone from the
blood. Non-lactating Holstein cows (fl 8) were ovariectomized through a vaginal approach at least 2 weeks prior to receiving an intravaginal progesterone implant. The experimental design was a single cross¬ over experiment. On day 3 after placement of the implant, cows were randomly assigned to receive rbIFN-a (20 mg, i.m., sequence group 1) or placebo (4 ml; sequence group 2). Rectal temperatures and 10ml jugular blood samples for determination of plasma =
concentrations of progesterone were obtained at 0, 2, 4, 6, 8, 10 and 24 h after injection. The procedure was
on day 6 after implant insertion except that sequence group 1 received placebo and sequence group 2 received rbIFN-a.
repeated
Experiment 3: basal and LH-induced secretion of
progesterone from luteal slices
Corpora lutea obtained from cows (n 5) on day 12 of the oestrous cycle were collected immediately after slaughter and transported to the culture room within
conditioned culture medium to the assay was parallel to the standard curve. The RIA for LH was per¬ formed as described by Lucy, Staples, Michel et al. (1991). The coefficients of determination for intraand interassay coefficients of variation for the LH assay were 7-4 and 9-6% respectively. Values for LH concentrations are expressed relative to the USDA-bLH-B5 standard.
=
5 min. To confirm that cows had a functional corpus luteum at slaughter, corpus luteum weight and plasma concentrations of progesterone were measured for each cow. Each corpus luteum was cut into 250 mg slices utilizing a Stadie-Riggs microtome. Individual slices were placed in 6 ml of a 1:1 (v/v) mixture of DMEM and F12 medium in a 60 15 mm Petri dish. Tissue slices were preincubated (38 °C, 5% C02) for 30 min. Medium was then replaced and rbIFN-a (0, 10, 50 and 100 pg/1) and LH (0 or 100 pg/1) added in duplicate cultures in a 4x2 factorial arrangement. Placebo was given in appropriate volumes so that all doses of rbIFN-a contained the same buffer compo¬ sition. After 2 h at 38 °C and 5% C02, culture medium 20 °C) until analysis by was harvested and stored (
radioimmunoassay (RIA) for progesterone. —
Experiment 4: concentrations of LH Non-lactating Holstein cows ( 8) that had been ovariectomized for more than 30 days were used in a single cross-over design. On the first day of the exper¬ iment, a catheter was placed in the jugular vein of each cow as described previously (Helmer et al. 1989 ). Blood samples (8 ml) for obtaining plasma to measure LH concentrations were collected into heparinized
Statistical
analyses Data were analysed by least-squares analysis of vari¬ ance using the General Linear Models procedure of the Statistical Analysis System (SAS, 1990). For experiment 1, the model contained effects of period, cow, treatment (placebo, rbIFN-a, FM, FM-frbIFNa), residual 1 (to test aforementioned effects), time, time treatment and residual 2 (to test time and time treatment). The model used for experiments 2 and 4 included effects of group, cow (sequence group), period, sequence group period (i.e. treatment), cow (sequence group) period, time, time sequence group, time cow (sequence group), time period, sequence group period time (i.e. treatment time) and residual. For experiment 3, the model included effects of cow, LH ( + or ), concentration of rbIFN-a (0, 10, 50 and 100 ng/ml), cow LH, cow rbIFN-a, —
rbIFN-a and the residual. In all models, cow considered as a random effect and other main effects were considered fixed. LH
was
=
tubes at 15-min intervals for 2 h before and 8 h after i.m. injection of rbIFN-a (20 mg; sequence group 1) or placebo (4 ml; sequence group 2), at which time catheters were removed. Rectal temperature was measured at 0, 2, 4, 6, 8 and 10 h after treatment. After 7 days, catheters were reinserted and the same procedure was repeated except that treatments were reversed as in a single cross-over design (i.e. sequence group 1 recieved placebo and sequence group 2 received rbIFN-a).
Radioimmunoassays The progesterone RIA, performed without extraction, was described elsewhere (Knickerbocker, Thatcher, Bazer et al. 1986). Intra- and interassay coefficients of
variation were 8-9 and 7-6% respectively. Samples of conditioned medium were diluted 100-fold with dis¬ tilled water before use in the assay. The inhibition curve caused by adding various amounts of diluted
RESULTS
Experiment 1: interrelationship between hyperthermia and plasma concentrations of progesterone Rectal temperature was affected by treatment (P
Betts, W. W. Thatcher and P. J. Hansen
Departamento de Farmacologia, Universidade Estadual Paulista, Botucatu, Brazil *Dairy Science Department, University of Florida, Gainesville, Florida 32611-0701, U.S.A. (Requests for offprints should be addressed to P. J. Hansen) received 6 September 1991 ABSTRACT
Experiments were performed to determine the mechby which recombinant bovine interferon-\g=a\I1 (rbIFN-\g=a\)causes an acute reduction in plasma concentrations of progesterone. In experiment 1, administration of a prostaglandin synthesis inhibitor blocked rbIFN-\g=a\-induced hyperthermia but did not prevent the decline in plasma concentrations of progesterone. The decline in progesterone concentrations caused by
did not affect basal and LH-induced release of progesterone from cultured luteal slices, indicating that rbIFN-\g=a\is unlikely to affect luteal function directly. Injection of rbIFN-\g=a\did, however, cause a decrease in plasma concentrations of LH in ovariectomized cows (experiment 4) that coincided temporally with the decrease in progesterone concentrations seen in cows having a functional corpus luteum. The present results strongly suggest that rbIFN-\g=a\acts to reduce secretion of progesterone by interfering with pituitary support for luteal synthesis of progesterone. The finding that rbIFN-\g=a\ can inhibit LH secretion implies that interferon-\g=a\molecules should be considered among the cytokines that can regulate hypothalamic or pituitary function. Journal ofEndocrinology (1992) 133, 175\p=n-\182
INTRODUCTION
feron called trophoblast protein-1 (oTP-1 and bTP1 for sheep and cow respectively) (Roberts et al. 1990; Bazer, Thatcher, Hansen et al. 1991). Both oTP-1 and bTP-1 extend luteal function by reducing endometrial secretion of prostaglandin (PG) F2a (Godkin, Bazer, Thatcher & Roberts, 19846; Vallet, Bazer, Fliss & Thatcher, 1988; Helmer, Gross, Hansen & Thatcher, 1989a; Helmer, Hansen, Thatcher et al. 19896; Salamonsen, Manikhot, Healy & Findlay, 1989; Barros, Plante, Thatcher &
anism
rbIFN-\g=a\was, therefore, not a direct consequence of the associated hyperthermia or of pathways mediated through prostaglandin synthesis. It is also unlikely that rbIFN-\g=a\ acts to increase the clearance of progesterone since injection of rbIFN-\g=a\ did not decrease plasma concentrations of progesterone in ovariectomized cows given an intravaginal implant of progesterone (experiment 2). In experiment 3, rbIFN-\g=a\
In the last several years, it has become clear that various regulatory molecules usually associated with the immune system play important roles as endo¬ crine or paracrine regulatory molecules in other tis¬ sues, especially in the reproductive tract and central nervous system (McCann, Rettori, Milenkovic et al.
1989; Hunt, 1989; Shibata, 1990; Roberts, Cross,
Farin et al. 1990; Tartakovsky & Ben-Yair, 1991). One of the best-studied examples is provided by the regulation of the luteal lifespan by the ruminant em¬ bryo. In sheep, cattle and goats, the embryo pre¬ vents destruction of the corpus luteum during early pregnancy by secreting a type I trophoblast inter-
Hansen, 1991).
role of the trophoblast proteinregulate uterine prostaglandin secretion, these molecules do possess biological activity characteristic of type 1 interferons, including While the
major
1 molecules is to
antiviral and Torres, Vallet
immunosuppressive activity (Pontzer,
al. 1988; Newton, Vallet, Hansen & Bazer, 1989; Plante, Hansen, Thatcher et al. 1990; Skopets, Li, Thatcher et al. 1992). Interestingly, administration of non-embryonic interferon-a mol¬ ecules can mimic the antiluteolytic effect of bTP-1 and oTP-1 when given to cyclic animals (Plante, Hansen & Thatcher, 1988; Plante, Hansen, Martinod et al. 1989; Stewart, Flint, Lamming et al. 1989; Newton, Martinod, Hansen et al. 1990) and can reduce secretion of PGF2a from endometrial tissue in vitro (Salamonsen et al. 1989; Barros et al. 1991) and in vivo (Piante, Thatcher & Hansen, 1991). One of the paradoxical effects of systemic administration of recombinant bovine interferon-a, 1 (rbIFN-a) to cattle is that it extends the lifespan of the corpus luteum but also causes an acute decline in serum concentrations of progesterone which reaches a nadir at 4-6 h after injection and returns to preinjection values by 8 h after injection (Newton et al. 1990; Plante et al. 1991). The transitory decline in circulating concentrations of progesterone and the hyperthermia-induced rbIFN-a are of greatest magnitude following an initial injec¬ tion. These effects of rbIFN-a decline and then dis¬ appear during a regimen of repeated injections (Newton et al. 1990; Plante et al. 1991). Interferon- reduces circulating concentrations of steroid hormones in other systems as well. Administration of interferon-a reduced circulating concentrations of oestradiol and progesterone in women (Kaupilla, Cantell, Janne et al. 1982) and testosterone in men (Orava, Cantell & Vikho, 1986). Since the timing and refractoriness of the effect of interferon-a on body temperature parallels those for serum concentrations of progesterone, it is possible that effects on progesterone are caused indirectly by changes in hyperthermia. Perhaps temperature alters the function of the hypothalamo-pituitary-ovarian axis (Flowers & Day, 1990) or clearance of progester¬ one. Interferon- has been reported to increase drug clearance in a body-temperature-related manner (Ehrsson, Eksborg, Wallin et al. 1990). It is also poss¬ ible that interferon-a alters progesterone secretion through mechanisms unrelated to hyperthermia. One possibility is that rbIFN-a acts directly on luteal cells to inhibit progesterone secretion. Pretreatment of cultured porcine Leydig cells with interferon-a reduced human chorionic gonadotrophin (hCG)stimulated testosterone secretion (Orava, 1989). Another possibility is that rbIFN-a acts directly on the hypothalamic-pituitary axis to reduce the secretion of luteinizing hormone (LH) and cause a subsequent decline in progesterone secretion. These possible mechanisms for the effects of rbIFN-a on the plasma concentrations of progesterone are examined in this paper.
MATERIALS AND METHODS
et
Materials
Bovine interferon-a, 1 and placebo (buffered mannitol solution) were donated by Ciba-Geigy (Basel, Switzerland). The intravaginal progesterone implants used in the study were CIDR@-B devices obtained from Carter Holt Harvey Agricultural Division, Hamilton, New Zealand. Each implant contained 1-9 g progesterone. Flunixin meglumine (FM) was from Schering-Plough (Kenilworth, NJ, U.S.A.). Lutalyse was from Upjohn (Kalamazoo, MI, U.S.A.).
Dulbecco's modified minimal essential medium (DMEM) and Ham's nutrient mixture F-12 medium, (F12) were obtained from Sigma (St Louis, MO, U.S.A). LH (USDA-bLH-B-5) and antiserum to bov¬ ine LH (USDA-309-684P) were a gift from Dr D. J. Bolt (USDA, Beltsville Agricultural Research Center,
Beltsville, MD, U.S.A.).
Experiment 1: interrelationship between hyperthermia and plasma concentrations of progesterone Seven non-lactating Holstein cows (466-597 kg) were given two injections of 25 mg PGF2a (Lutalyse) 11 days apart to synchronize oestrus. Each cow received the following series of treatments at 3-day intervals starting on the morning of day 7 of the oestrous cycle: placebo (4 ml), rbIFN-a (20mg/cow, i.m., 4 ml); rbIFN-a (20 mg, i.m., 4 ml) and FM (a prostaglandin synthesis inhibitor, 22 mg/kg, i.m., two injections at 0 and 5 h relative to injection of rbIFN-a); and FM alone. Treatments were given according to the schedule dictated by a 2x4 Latin square design, where one square was incomplete. At 0,2,4,6,8,10 and 24 h after injection of rbIFN-a, rectal temperatures were measured and a 10 ml sample of jugular blood was collected into heparinized tubes via venipuncture for subsequent determination of progesterone concen¬ trations in plasma. Experiment 2: clearance of progesterone In this experiment, it was reasoned that possible changes in plasma progesterone concentrations induced by rbIFN-a in ovariectomized cows given
progesterone implants would be indicative that rbIFN-a affects clearance of progesterone from the
blood. Non-lactating Holstein cows (fl 8) were ovariectomized through a vaginal approach at least 2 weeks prior to receiving an intravaginal progesterone implant. The experimental design was a single cross¬ over experiment. On day 3 after placement of the implant, cows were randomly assigned to receive rbIFN-a (20 mg, i.m., sequence group 1) or placebo (4 ml; sequence group 2). Rectal temperatures and 10ml jugular blood samples for determination of plasma =
concentrations of progesterone were obtained at 0, 2, 4, 6, 8, 10 and 24 h after injection. The procedure was
on day 6 after implant insertion except that sequence group 1 received placebo and sequence group 2 received rbIFN-a.
repeated
Experiment 3: basal and LH-induced secretion of
progesterone from luteal slices
Corpora lutea obtained from cows (n 5) on day 12 of the oestrous cycle were collected immediately after slaughter and transported to the culture room within
conditioned culture medium to the assay was parallel to the standard curve. The RIA for LH was per¬ formed as described by Lucy, Staples, Michel et al. (1991). The coefficients of determination for intraand interassay coefficients of variation for the LH assay were 7-4 and 9-6% respectively. Values for LH concentrations are expressed relative to the USDA-bLH-B5 standard.
=
5 min. To confirm that cows had a functional corpus luteum at slaughter, corpus luteum weight and plasma concentrations of progesterone were measured for each cow. Each corpus luteum was cut into 250 mg slices utilizing a Stadie-Riggs microtome. Individual slices were placed in 6 ml of a 1:1 (v/v) mixture of DMEM and F12 medium in a 60 15 mm Petri dish. Tissue slices were preincubated (38 °C, 5% C02) for 30 min. Medium was then replaced and rbIFN-a (0, 10, 50 and 100 pg/1) and LH (0 or 100 pg/1) added in duplicate cultures in a 4x2 factorial arrangement. Placebo was given in appropriate volumes so that all doses of rbIFN-a contained the same buffer compo¬ sition. After 2 h at 38 °C and 5% C02, culture medium 20 °C) until analysis by was harvested and stored (
radioimmunoassay (RIA) for progesterone. —
Experiment 4: concentrations of LH Non-lactating Holstein cows ( 8) that had been ovariectomized for more than 30 days were used in a single cross-over design. On the first day of the exper¬ iment, a catheter was placed in the jugular vein of each cow as described previously (Helmer et al. 1989 ). Blood samples (8 ml) for obtaining plasma to measure LH concentrations were collected into heparinized
Statistical
analyses Data were analysed by least-squares analysis of vari¬ ance using the General Linear Models procedure of the Statistical Analysis System (SAS, 1990). For experiment 1, the model contained effects of period, cow, treatment (placebo, rbIFN-a, FM, FM-frbIFNa), residual 1 (to test aforementioned effects), time, time treatment and residual 2 (to test time and time treatment). The model used for experiments 2 and 4 included effects of group, cow (sequence group), period, sequence group period (i.e. treatment), cow (sequence group) period, time, time sequence group, time cow (sequence group), time period, sequence group period time (i.e. treatment time) and residual. For experiment 3, the model included effects of cow, LH ( + or ), concentration of rbIFN-a (0, 10, 50 and 100 ng/ml), cow LH, cow rbIFN-a, —
rbIFN-a and the residual. In all models, cow considered as a random effect and other main effects were considered fixed. LH
was
=
tubes at 15-min intervals for 2 h before and 8 h after i.m. injection of rbIFN-a (20 mg; sequence group 1) or placebo (4 ml; sequence group 2), at which time catheters were removed. Rectal temperature was measured at 0, 2, 4, 6, 8 and 10 h after treatment. After 7 days, catheters were reinserted and the same procedure was repeated except that treatments were reversed as in a single cross-over design (i.e. sequence group 1 recieved placebo and sequence group 2 received rbIFN-a).
Radioimmunoassays The progesterone RIA, performed without extraction, was described elsewhere (Knickerbocker, Thatcher, Bazer et al. 1986). Intra- and interassay coefficients of
variation were 8-9 and 7-6% respectively. Samples of conditioned medium were diluted 100-fold with dis¬ tilled water before use in the assay. The inhibition curve caused by adding various amounts of diluted
RESULTS
Experiment 1: interrelationship between hyperthermia and plasma concentrations of progesterone Rectal temperature was affected by treatment (P
