The Ovarian Artery as the Final Component of the Local Luteotropic Pathway Between a Gravid Uterine Horn and Ovary in Ewes

BIOLOGY OF REPRODUCTION 15, 414-421 (1976) The Ovarian Artery as the Final Component of the Local Luteotropic Pathway Between a Gravid Uterine Horn and Ovary in Ewes R. J. MAPLETOFT, D. R. LAPIN and 0. J. GINTHER Department of Veterinary Science, University of Wisconsin-Madison, Madison, Wisconsin 53706 ABSTRACT Fifty-five mature, cross-bred ewes which had been bred at estrus were used in 3 experiments. In Exp. 1, the role of the ovarian artery in the unilateral luteotropic effect of a gravid uterine horn was studied. On Days 5-7 after breeding the uterine horns were separated surgically in 16 bilaterally ovulating ewes. One uterine horn was left open to the abdominal cavity to produce a nongravid horn. Ewes were randomized into 4 groups: 1) unilaterally pregnant controls, 2) surgical anastomosis of ovarian branch of ovarian artery from a point near the ovary on the gravid side to the ovarian artery distant to the ovary (about 3 cm before bifurcation of artery) on the nongravid side, 3) nonpregnant controls in which the endometrium (and ability to produce luteolysin) on the gravid side was destroyed with phenol cautery, and 4) nonpregnant controls with phenol treatment (as in Group 3) and surgical anastomosis of ovarian artery (as in Group 2) from cauterized to nongravid side. At necropsy on Day 20, mean CL weights were less (P<0.01) on the nongravid side of Group 1(117mg) and on the noncauterized side of Groups 3 (171 mg) and4 (122 mg) than on the gravid side of Group 1 (542 mg), cauterized side of Group 3(361 mg), and the nongravid side of Group 2 (722 mg). Mean CL weights were not significantly different among the nongravid side of Group 1 and the noncauterized side of Groups 3 and 4 The luteotropic (prevention of luteolysis) effect of a gravid uterine horn was studied in unilaterally ovulating ewes (Exp. 2 and 3). In Exp. 2, 28 bred ewes were randomized into a 2 X 2 factorial (sham-operation or hysterectomy X 0 or 200 ig PG F2 ). Surgery was done on Days 5-7 and on Day 13 PCF20 or vehicle was infused into the ovarian artery on the CL side. On Day 15, mean CL weight was less (P<0.05) in hysterectomized ewes which were given 200 g(339 mg, CL regressed in 6/7 ewes) than in hysterectomized ewes which were given 0 ig (725 rng, CL regressed in 0/7 ewes) or in intact ewes which were given either 200 (540 mg, CL regressed in 2/7 ewes) or 0 zg (602 mg CL regressed in 1/7 ewes). There was no significant difference between doses in intact ewes. In Exp. 3, uterine horns were separated surgically on Days 5-7 and ewes were randomized into 2 groups: 1) surgical anastomosis of ovarian branch of ovarian artery from a point near the ovary on the nongravid side to a point near the ovary on the gravid side (6 ewes), 2) surgical anastomosis of ovarian branch of ovarian artery from a point near the ovary on the nongravid side to ovarian artery distant to ovary (as in Exp. 1) on the gravid side (5 ewes). Ewes were treated with progesterone in oil (25 mg daily) from Day 11 and necropsied on Day 18. Mean CL weight was less (P<0.05) in Group 1 ewes (410 mg, 2/6 ewes with maintained CL) than in Group 2 ewes (597 mg, 5/5 ewes with maintained CL). Results indicated that the gravid uterine horn exerts an antiluteolytic effect at the level of the adjacent ovary and that the effect is exerted through a local utcroovarian venoarterial pathway. INTRODUCTION In the nonpregnant ewe the uterus causes regression of the corpus luteum (CL) through a direct or local pathway between the uterine horn and the adjacent ovary. There is considerable evidence that the local pathway between a uterine horn and adjacent ovary in sheep is venoarterial in nature (reviewed by Ginther, 1974) involving the main uterine vein (uterine branch of ovarian vein) as the uterine compon- Accepted June 15, 1976. Received May 4, 1976. ent and the ovarian branch of the ovarian artery as the ovarian component (Fig. 1). There is also considerable evidence that the uterine luteolytic substance in sheep is prostaglandin (PG) Faa or a closely related substance (Goding, 1974), and that secretion of PGF2a by the sheep uterus increases just prior to and during luteal regression (Wilson et al., 1972a; Barcikowski et al., 1974). Paradoxically, PGF20 has been found to be present in high levels in the uterine venous effluent, not only during the time of luteal regression in cycling ewes, but also during the corresponding days of pregnancy when the CL is maintained (Wilson et al., 1972b; 414

LOCAL LUTEOTROPIC PATHWAY IN EWES 415 Pexton et al., 1975). In the pregnant ewe, the embryo must be present in the uterine horn on approximately Day 12 after estrus to prevent luteal regression (Moor and Rowson, 1966a). Infusion of homogenates from frozen sheep embryonic tissue cotlected on Day 14 and 15 of pregnancy into the uteri of nonpregnant ewes (beginning on Day 12) also prevented luteal regression (Rowson and Moor, 1967). In addition, experiments involving transfer of fertilized ova to surgically separated uterine horns in the ewe resulted in luteal regression when ova were transferred to the horn contralateral to the CL and luteal maintenance when ova were transferred to the horn ipsilateral to the CL (Moor and Rowson, 1966b). Similarly, in ewes with CL in both ovaries, surgical separation of the uterine horns and transfer of fertilized ova to one horn resulted in luteal maintenance on the gravid side and luteal regression on the opposite nongravid side. Therefore, luteal maintenance in pregnancy also is mediated through a local pathway. However, the nature of the local pathway and the mechanism by which the embryo overcomes the effect of the uterine luteolysin are unresolved. The terms luteotropin and antiluteolysin will be used interchangeably herein, to describe the substance from a gravid uterine horn that prevents regression of the CL. Likewise, luteotropic effect and antiluteolytic effect will be used synonymously. Experiments involving mtrafollicular injection of PGF2a in a CL-bearing ovary were interpreted as indicating that the conceptus produced an antiluteolytic factor which inhibited the negative effect of PGF2a on luteal function (lnskeep et al, 1975). Direct evidence that a gravid uterine horn or its contents secretes a blood-borne luteotropin which unilaterally inhibits the effect of the uterine luteolysin, was provided by an experiment involving surgical anastomosis of the main uterine veins in bilaterally ovulating, unilaterally pregnant ewes (Mapletoft et al., 1975). The CL regressed when the ipsilateral uterine vein contained blood from only the nongravid horn, indicating the presence of uterine luteolysin, whereas CL were maintained when the ipsilateral uterine vein contained venous blood from a gravid horn, whether or not it also contained blood from a nongravid horn. Results indicated that the main uterine vein was involved in the unilateral luteotropic effect of a gravid horn and raised the question whether the luteotropin, like the uterine luteolysin, passes from gravid horn to adjacent ovary through a local venoarterial pathway. The present experiments were designed to test the hypothesis that the unilateral luteotropic effect of a gravid uterine horn in ewes is exerted through a local venoarterial pathway (Fig. 1). The specific objectives were to demonstrate that the ovarian artery serves as the final or ovarian component of the unilateral luteotropic pathway and that the gravid uterine horn exerts its antiluteolytic effect at the level of the adjacent ovary. MATERIALS AND METHODS Fifty-five mature, cross-bred ewes were used in 3 experiments. Ewes were observed twice daily for estrus using vasectomized rams. When a ewe stood for mounting, estros was recorded and the day was designated as Day 0 of the estrous cyde. All ewes were mated at least twice using two fertile rams, beginning at the onset of estrus, and twice every 12 h thereafter as long as ewes stood for mounting. At surgery, on Days 5, 6 or 7 of the estrous cycle, ewes were laparotomized mid-ventrally under halothane anaesthesia (induced with thiamylal sodium), and CL were marked with India ink. Ewes with CL in both ovaries were assigned to Exp. 1 and those with CL in only 1 ovary were assigned to Exp. 2 and 3. Experiment 1, Involvement of Ovarian Artery In 28 bilaterally ovulating ewes the uterine horns were separated by severing the intercornual ligament to the point of internal bifurcation of the uterus. A silk ligature was placed around one horn ax the bifurcation (side selected at random) and the horn was transected on the ovarian side of the ligature to produce an isolated uterine horn. Hemostasis was maintained through minimal use of electrocautery. The transected end of the isolated horn was left open to the abdominal cavity to prevent the establishment of pregnancy in the transected (isolated) horn. In this manner, a gravid and a nongravid horn were established in bilaterally ovulating ewes. Ewes were randomized into 4 groups (Fig. 2). Group 1 ewes served as controls. In Group 2 ewes, the ovarian branch of the ovarian artery (donor artery) was surgically anastomosed from a point near the ovary (within 1 cm) on the intact (gravid) side to the ovarian artery (recipient artery) at a point distant to the ovary on the isolated (nongravid) side. The point of the recipient ovarian artery described as distant from the ovary was about 3 cm before its bifurcation into ovarian, uterine, and tubal branches. Group 3 ewes served as controls in which 20 ml of 10 percent phenol solution was injected into the lumen of the uterine horn on the intact side (phenol cauterized side) in onier to destroy the embryo and the endometrium on that side. In Group 4 ewes, the uterine horn on the intact side was treated with 10 percent phenol solution and the ovarian branch of the ovarian artery on the intact (phenol cauterized) side was

. 416 MAPLETOFT ET AL. ua.3) Pr tbov tboa FIG. 1. Dorsal view of the right uteroovarian vascular pedide from an ewe. Veins were injected with blue latex (dark) and arteries were injected with red latex (light). The uterine horn is drained primarily by the uterine branch of ovarian vein (main uterine vein) and the ovary is nourished by the ovarian branch of the ovarian artery. The local uteroovarian pathway for uterine-induced luteolysis is venoarterial in nature involving the main uterine vein as an initial or uterine component and the ovarian artery as the distal or ovarian component. The venoarterial transfer apparently occurs in areas of close contact between veins containing uterine venous blood and the adjacent ovarian artery. In the present studies, the hypothesis was tested that the luteotropic effect (prevention of luteolysis) of a gravid uterine horn in sheep is also exerted through a local uteroovarian venoarterial pathway. 0 = ovary; uh = uterine horn; ut uterine tube; oa = ovarian artery; ov = ovarian vein (uteroovarian vein); uboa uterine branch of ovarian artery; ubov = uterine branch of ovarian vein (main uterine vein); oboa = ovarian branch of ovarian artery; obov ovarian branch of ovarian vein; tboa = tubal branch of ovarian artery; tbov = tubal branch of ovarian vein. surgically anastomosed to the ovarian artery on the isolated (nongravid) side as in Group 2. The donor artery was occluded with a noncrushing cardiovascular clamp for 20-45 mm while an end-to-end surgical anastomosis was being done. The techniques for freeing the arteries and performing the anastomosis have been described (Ginther et al., 1973). The proximal segment of the recipient artery was ligated twice between the aorta and the point of transection to minimise development of collateral vessels so that the ovary on the recipient side would be supplied only through the anastomosis between the donor and recipient arteries. Sodium heparin solution was given iv (10,000 units) to ewes in Groups 2 and 4 immediately before vessels were transected. Penicillin and streptomycin were given once daily for 3 days. Ewes were observed daily for estrus and were necropsied on Day 20. The marked CL were removed, weighed and classified as maintained, partially regressed or regressed as described (Mapletoft and Ginther, 1975). The uterine horns were examined for the absence of fluid accumulation or embryos on the isolated side (all groups) and for the presence of an intact embryo on the uterine intact side of Groups 1 and 2 and for endometrial destruction (pitting of caruncles) on the uterine intact side of Groups 3 and 4. The surgical anastomosed vessels (Groups 2 and 4) were examined for patency by gently flushing saline solution through the donor ovarian artery. Patency was indicated by oozing of saline from the site of the recovered CL Weights of CL were analyzed statistically by analysis of variance with a hierarchical classification and mesn CL weights were compared by the protected lad test for multiple comparisons (significant F value in the analysis of variance). Experiment 2, Effect of Exogenous PGF20 in Pregnant Ewes At surgery on Days 5-7 after estrus, 28 bred, unilaterally ovulating ewes were randomized into a 2 X 2 factorial experiment (uterine intact or hysterectomized X 0 or 200 PGF20; Table 1). Fourteen of the ewes were hysterectomized completely (removal of the entire uterus including the most cranial cervical ring and all of the oviduct) and 14 ewes were sham operated. When multiple CL were present in 1 ovary, all but 1 were enudeated to minimize the possibilities of confounding effects between number of CL and dose of PGF2. On Day 13 after breeding, ewes were relaporatomized midventrally and PGFSa was injected. The ovarian branch of the ovarian artery of the CL-bearing ovary was freed using a vasodilator solution (0.1 percent phenoxybenzamine hydrochloride in 1 percent lidocaine hydrochloride) and blunt dissection. Injections of PGF20 (0 or 200.ug) were given into the ovarian branch of the ovarian artery approximately 2 cm from the ovary, over a period of 1 min while the ewe was under halothane anaesthesia and in dorsal recumbancy. PGF20 or vehicle (distilled water) was administered in a volume of 1 ml using a 1 ml syringe and 27 gauge needle. PGF20 stock solution (THAM salt in distilled water) and vehicle solution were prepared at the beginning and were kept refrigerated during the course of the experiment (approximately 2 mo). The dose of

LOCAL LUTEOTROPIC PATHWAY IN EWES 417 PGF20 administered was on a free-acid basis and was based on results of a preliminary experiment in which this dose was found to consistently cause luteal regression in nonpregnant ewes during middiestrus. On Day 15 (48 h after injection) ewes were necropsied and CL were removed, classified, and weighed. An attempt was made to confirm pregnancy by recovery of embryos but because degenerating embryos were difficult to distinguish from endometrial fragments, this criterion was dropped. Weights of CL were analyzed statistically by analysis of variance for a 2 X 2 factorial design (uterine intact or hysterectomized X 0 or 200 ig PGF20) and mean CL weights were compared by the protected lad test for multiple comparisons. Experiment 3, Anti-Luteolytic Effect of a Gravid Horn At surgery (Days 5-7) 18 bred, unilaterally ovulating ewes were randomized into 2 groups (Fig. 3). Uterine horns were separated surgically (as in Exp. 1) and the horn contralateral to the CL was isolated. In Group 1 ewes, the ovarian branch of the ovarian artery was surgically anastomosed from a point near the ovary on the isolated side to the corresponding vessel at a point near the ovary on the intact (gravid) side. In Group 2 ewes, the ovarian branch of the ovarian artery was surgically anastomosed from a point near the ovary on the isolated side to the ovarian artery at a point distant to the ovary (as in Exp. 1) on the intact (gravid) side. The techniques for freeing the arteries, performing the anastomosis and ligating the proximal segment of the recipient artery were also as in Exp. 1. All ewes were treated with sodium heparin solution iv (10,000 units) immediately before vessels were transected and penicillin and streptomycin were given once daily for 3 days. Progesterone in corn oil (25 mg) was given daily beginning on Day 11 to maintain pregnancy in ewes in which CL regressed. Ewes were necropsied on Day 18 and marked CL were removed, dassified, and weighed. The uterine horns were examined for absence of fluid accumulation on the isolated side and the presence of a viable embryo on the intact side. In addition, the arterial anastomosis was examined for patency by gently flushing saline solution through the donor ovarian artery. Weights of CL were analyzed statistically by one-way analysis of variance and numbers of ewes demonstrating evidence of luteal maintenance were compared by chi-square analysis. RESULTS In Exp. 1 (Fig. 2), 16 ewes (4 per group) had an isolated horn free of embryos and fluid accumulation (all groups), an embryo in the intact horn (Groups 1 and 2) and a patent surgical anastomosis (Groups 2 and 4) at necropsy on Day 20. Based on the analysis of variance of CL weights, the interaction of group and side was highly significant. Mean weight of CL on the isolated (nongravid) side (Fig. 2) was less (P<0.01) in Group 1 control ewes (117 mg;4 of 4 CL regressed) Group 3 phenol-treated control ewes (171 mg; 3 of 4 CL regressed) and Group 4 phenol-treated ewes with arterial anastomosis (122 mg;4 of 4 CL regressed) than on the intact (gravid) side of Group 1 control CONTROL ANASTOMOSIS CONTROL ANASTOMOSIS Isolated Gravid Isolated Gravid Isolated Cauterized Isolated Cauterized Mean (mgi Mean (mgi Mean (mgi Mean (mg) Mean (mgi Mean (mgi Mean (mgi Mean (mgi ± SEN ± SEN ± SEN 2 SEN ± SEN ± SEN 2 SEN 2 SEN 117 542c 722d 109a 171a 1b 12? 17? ±21 ±51 ±54 ±32 ±69 ±93 ±24 ±63 FIG. 2. Mean corpus luteum weights in bilaterally ovulating ewes with surgically separated uterine horns as effected by surgical anastomosis (A) of the ovarian artery. Surgery was done on Days 5-7 after breeding and necropsies were done on Day 20. In all ewes, the uterine horns were surgically separated and one was ligated and transected at the internal bifurcation to produce an isolated (nongravid) horn and an intact (gravid) horn. The gravid horn in Groups 3 and 4 was treated with a solution of 10 percent phenol to produce a cauterized horn. Groups 1 and 3 served as controls. In Groups 2 and 4 the ovarian branch of the ovarian artery was surgically anastomosed (A) from a point near the ovary and on the intact side to the ovarian artery at a point distant to the ovary on the isolated side. The arrow indicates the direction of arterial blood flow through the anastomosis. Mean CL weights with different superscripts are different (P<0.01 ; 4 ewes/group).

418 MAPLETOFT ET AL. TABLE 1. Effect of intraovarian arterial injection of 200 g PGF20 on mean corpus luteum (CL) weight in pregnant and hysterectomized ewes. Dose of PGF20b Pregnant Hysterectomizeda Corpus luteum 0 g 200 g Og 200 g Mean weight (mg) #{243}o2de 5.40d 725e 339C SEM ±62 ±70 ± 32 ± 74 No. ewes with regressed CL 1/7 2/7 0/7 6/7 ahysterectomizes were done on Day 7, PGF20 was injected on Day 13 and necropsies were done on Day 15. bpgf0 (0 and 200 Mg) fl 1 ml distilled water injected into ovarian artery on CL side over 1 mm. c,d,emean CL weights with no superscript letter in common differ significantly (P<0.05). ewes (542 mg; 4 of 4 CL maintained), the intact (cauterized with phenol solution) side of Group 3 control ewes (361 mg; 3 of 4 CL maintained) and on the isolated (nongravid) side of Group 2 ewes with arterial anastomosis (722 mg; 4 of 4 CL maintained). Mean weights of CL were not significantly different among the isolated (nongravid) side in Group 1 and the isolated (noncauterized) side in Groups 3 and 4. In Exp. 2 (Table 1), mean CL weight was less (P<0.05) in hysterectomized ewes which were given 200 pg PGF2a (339 mg;6 of 7 ewes with regressed CL) than in hysterectomized ewes receiving vehicle only (725 mg; 0 of 7 ewes with regressed CL) or intact ewes which were given 200 pg PGF2a (540 mg; 2 of 7 ewes with regressed CL) and intact ewes which were given vehicle only (602 mg; 1 of 7 ewes with regressed CL). There was no significant difference in mean CL weights between doses in intact ewes. In Exp. 3 (Fig. 3), 11 ewes had a patent surgical anastomosis and an intact embryo at necropsy on Day 18 (6 in Group 1 and 5 in Group 2). Mean weight of CL for ewes pregnant at necropsy was less (P<O.05) in Group 1 in which the arterial anastomosis was from a point near the ovary on the isolated (nongravid) side to a point near the ovary on the intact (gravid) side (410 mg; 2 of 6 ewes with maintained CL) than in Group 2 in which the arterial anastomosis was from a point near the ovary on the isolated (nongravid) side to a point distant to the ovary on the intact (gravid) side (597 mg; 5 of 5 ewes with maintained CL). At necropsy, 6 of 12 ewes ultimately assigned to Group 1 and 5 of 6 ewes ultimately assigned to Group 2 were pregnant. DISCUSSION In Exp. 1 (Fig. 2), regression of CL on the isolated (nongravid) side (Groups 1 and 3) and maintenance on the intact (gravid) side (Group 1) confirmed that the nongravid horn produced a Iuteolysin and the gravid horn either produced a luteotropin or prevented production or release of the uterine luteolysin. It is likely that unilateral regression of CL on the isolated (non- CorDon SEN NO. Ews$ P,.gn.nt M.int.in.d ut.. CL Isoisted Grand Isct.t.d Gravid 410 59? ±59 ±49 9/12 5/6 2/6 515 FIG. 3. Effect on corpus luteum of surgical anastomosis (A) of the ovarian artery from a nongravid side to a gravid side in unilaterally pregnant ewes. Ewes were surgically operated on Days 5-7, treated with progesterone daily from Day 11 and necropsied on Day 18. Uterine horns in unilaterally ovulating ewes were surgically separated to produce an isolated (nongravid) and an intact (gravid) uterine horn. The ovarian branch of the ovarian artery was surgically anastomosed (A) from a point near the ovary on the isolated side to a point on the gravid side near the ovary (Group 1) and distant to the ovary (Group 2). The arrow indicates the direction of arterial blood flow through the anastomosis. CL weights were analyzed only for ewes pregnant at necropsy (6 in Group 1 and in Group 2). Mean CL weights and numbers of ewes with maintained CL were significantly different (P<0.05) between the 2 groups.

LOCAL LUTEOTROPIC PATHWAY IN EWES 419 gravid) side in Groups 1 and 3 was exerted through a venoarterial pathway as has been demonstrated in nonpregnant sheep (Ginther et al., 1973; Mapletoft and Ginther, 1975). Luteal maintenance on the phenol-treated side in Group 3 was attributed to chemical cauterization of the endometrium and, therefore, an inability of the horn to produce adequate amounts of uterine luteolysin (Foote et al., 1974). Mean CL weight was significantly larger on the phenol-treated side of Group 3 than on the isolated (nongravid) side in either Group 3 or 4, indicating that the ovarian artery on that side contained inadequate luteolysin. A preliminary experiment involving surgical separation of the uterine horns and phenol treatment on the CL side resulted in luteal maintenance in 4 of 4 control ewes, whereas, surgical anastomosis of the ovarian branch of the ovarian artery from a point near the ovary on a phenol-treated side to a point distant to the ovary on the isolated (CL) side (as in Group 4) resulted in luteal regression on the recipient side in 4 of 4 ewes. These preliminary results therefore also indicated that phenol treatment resulted in luteal maintenance and that surgical anastomosis of the ovarian artery as in Groups 2 and 4 should have provided sufficient venoarterial contact area on the recipient side to permit transfer of the luteolysin. Venoarterial transfer of luteolysin into the ovarian artery on the recipient side in Group 4 resulted in luteal regression on that side in spite of the CL being supplied by arterial blood from a phenol-cauterized side. A similar result would have been expected in the Group 2 ewes if the embryo had prevented the production or release of luteolysin on the donor (intact) side. However, luteal maintenance occurred on the isolated (recipient) side in Group 2. This indicates that a luteotropin from the intact (gravid) side passed through the arterial anastomosis and protected the CL on the isolated side from the luteolysin. Both luteolysin and luteotropin apparently passed from veins containing uterine venous blood into the ovarian artery by venoarterial transfer. Results of this experiment, therefore, provide support for those reported previously which demonstrated the involvement of the main uterine vein in the unilateral luteotropic effect of a gravid horn (Mapletoft et al., 1975), and indicates that the ovarian artery serves as the final component of the same pathway. Results collectively support the hypothesis that the unilateral luteotropic effect of a gravid uterine horn on the CL in the adjacent ovary is exerted through a local venoarterial pathway. Luteal regression and ovarian inactivity (small and nearly devoid of follicles) occurred on the donor sides in Exp. 1 (Fig. 2, Groups 2 and 4) in all ewes except one in Group 4 which showed evidence of necrosis and partial luteal regression. These findings were considered due to impared ovarian arterial blood supply on that side. Therefore, luteal regression induced by restricted blood flow in these groups was considered to be nonphysiologic. Administration of PGF2a by way of the ovarian branch of the ovarian artery has been shown to be an effective method of inducing luteolysis with a dose of PGF2 that was ineffective by systemic routes in the pony mare (Douglas et al., 1976). In preliminary experiments, 200 pg PGF2a (free acid) in sterile water was luteolytic when administered int o the ovarian artery of intact middiestrus ewes. This dose falls far below the considered minimum effective systemic luteolytic dose of PGF2a in ewes (Douglas and Ginther, 1973). As was expected, hysterectomized ewes treated with vehicle in Exp. 2 showed no evidence of luteal regression and mean CL weight was not significantly different from pregnant ewes treated with vehicle. Administration of 200 pg PGF2a into the ovarian artery supplying the CL-bearing ovary resulted in complete luteal regression within 48 h in 6 of 7 hysterectomized ewes, whereas 200 pg PGF2a was ineffective in 5 of 7 pregnant ewes (Table 1). The presence of a gravid uterine horn adjacent to the ovary bearing the CL therefore appeared to protect the CL from the exogenous luteolysin. Mean weight of CL in pregnant ewes receiving 200 pg PGF2a was not significantly different from that of pregnant ewes receiving vehicle. Results indicate that the presence of a gravid uterine horn protects the CL in the adjacent ovary from exogenous PGF2a by Day 13. Furthermore, the antiluteolytic effect was exerted within the ovary since the PGF2a was injected into the immediate ovarian arterial supply. The results of Exp. 3 also indicate that a gravid uterine horn exerts its effect at the level of the ovary. Ovarian arterial blood from a nongravid (isolated) side with demonstrated luteolyric activity (Fig. 3, Group 1; 4 of 6 ewes with regressed CL) was rendered nonluteolytic when it was diverted through the uteroovarian vascu-

420 MAPLETOFT ET AL. lar pedicle on the gravid (intact) side (Fig. 3, Group 2; 5 of 5 ewes with maintained CL). The luteotropin apparently passed into the ovarian artery on the gravid side in Group 2 and, upon delivery to the ovary, protected the CL from the luteolysin. Results therefore provide additional evidence for the inhibition of the luteolysin by the luteotropin at the level of the ovary rather than by prevention of venoarterial transfer of the luteolysin or inactivation of the luteolysin in the vein or vessel walls. The mechanism by which the luteotropin passes from veins which contain uterine venous blood into the adjacent ovarian artery is not known. It is likely, however, that passage of the luteotropin, like the luteolysin, from vein to artery is favored by the close apposition of the tortuous ovarian artery to the wall of veins which contain uterine venous effluent in the uteroovarian vascular pedicle (Fig. 1). Available evidence exists only for the venoarterial transfer of substances of low molecular weight, most of which were fat soluble, e.g., steroids, PGF2 (reviewed by Ginther, 1974; Free and Tillson, 1975). Furthermore, the venoarterial transfer of testosterone in the rat pampiniform plexus would appear to involve a concentration gradient-mediated passive transfer (Free and Jaffe, 1975). The pig trophoblast has been shown to be capable of producing steroids of estrogenic activity at about Day 11 of pregnancy, coinciding with maternal recognition of pregnancy (Perry et al., 1973). Similarly, the preimplantation blastocysts of rats, mice, hamsters and rabbits have been shown to have steroidogenic activity (reviewed by Dickmann, 1975). The gravid uterus in the ewe is capable of producing PG s (Wilson et al., 1976). Recently, however, a tetrapeptide (MW 499) which prevented ovulation in nonbred test animals, has been isolated from 2-cell hamster embryos (Kent, 1975). It is not known whether the preimplantation sheep embryo produces a similar molecular weight peptide and more importantly, whether such a substance would pass through a venoarterial pathway. Results of the present experiments do, however, indicate that the luteotropin from a gravid uterine horn in sheep must be of such characteristics that it will pass through the local uteroovarian venoarterial pathway. Attempts at isolation of the luteotropin, therefore, may be based on the assumption that the substance is fat soluble and of low molecular weight; e.g., a steroid or a prostaglandin. ACKNOWLEDGMENTS Supported by the College of Agricultural and Life Sciences and the Graduate School, University of Wisconsin-Madison, a grant from Merck Sharp and Dohme and grant No. 630-0505B from the Ford Foundation. R. J. Mapletoft is a post-doctoral trainee of the Endocrinology-Reproductive Physiology Program, University of Wisconsin-Madison and is supported by the Medical Research Council of Canada. PGF20 was provided by the Upjohn Co., Kalamazoo, MI. Fig. 1 is used herein for illustrative purposes and was first published in Am. J. Vet. Res. 34, (June, 1973); 723-728. REFER ENCES Barcikoki, B., Carison, J. C., Wilson, L. and Mc- Cracken, J. A. (1974). The effect of endogenous and exogenous estradiol-17 on the release of prostagiandin F20 from the ovine uterus. Endocrinology 95, 1340-1349. Dickmann, Z. (1975). Steroidogenis in preimplantation embryos. Res. Reprod. 7, 3. Douglas, R. H. and Ginther, 0. J. (1973). Luteolysis following a single injection of prostaglandin F30 in sheep. J. Anim. Scs. 37, 990-993. Douglas, R. H., Del Campo, M. R. and Ginther, 0. J. (1976). Luteolysis following carotid or ovarian arterial injection of prostaglandin F30 in mazes. Biol Reprod. 14,473-477. Foote, W. C., Alifakiots, T. A. and Call, J. W. (1974). Endometrial and corpus luteum relationships in the ewe. J. Anim. Sci. 39, 207 (abstract). Free, M. J. and Jaffe, R. A. (1975). Dynamics of venous-arterial testosterone transfer in the pampiniform plexus of the rat. Endocrinology 97, 169-177. Free, J. J. and Tillson, S. A. (1975). Local increase in concentration of steroids by venous-arterial transfer in the pampinform plexus. In: Hormonal Regulation of Spermatogenesis. F. S. French, Ed., Plenum Press, New York. Ginther, 0. J., Del Campo, C. H. and Rawlings, C. A. (1973). Vascular anatomy of the uterus and ovaries and the unilateral luteolytic effect of the uterus: A local venoarterial pathway between uterus and ovaries in sheep. Am. J. Vet. Res. 34, 723-728. Ginther, 0. J. (1974). Internal regulation of physiological processes through local venoarterial pathways: A review. J. Anim. Sci. 39, 550-564. Goding, J. R. (1974). The demonstration that PGF20 is the uterine luteolysin in the ewe. J. Reprod. Fert. 38, 261-271. Kent, H. A., Jr. (1975). Contraceptive polypeptide from hamster embryos: Sequence of amino acids in the compound. Biol. Reprod. 12, 504-507. Inskeep, E. K., Smumy, W. J., Butcher, R. L. and Pexton, J. E. (1975). Effects of intrafollicular injections of prostaglandins in non-pregnant and pregnant ewes. J. Anim. Sd. 41, 1098-1104. Mapletoft, R. J. and Ginther, 0. J. (1975). Adequacy of main uterine vein and the ovarian artery in the local venoarterial pathway for uterine-induced luteolysis in ewes. Am. J. Vet. Res. 36, 957-963. Mapletoft, R. J., Del Campo, M. R. and Ginther, 0. J. (1975). Unilateral luteotropic effect of uterine venous effluent of a gravid uterine horn in sheep.

LOCAL LUTEOTROPIC PATHWAY IN EWES 421 Proc. Soc. Exp. Biol. Med. 150, 129-133. Moor, R. M. and Rowson, L. E. A. (1966a). The corpus luteum of the sheep: functional relationships between the embryo and the corpus luteum. J. Endocrinol. 34, 233-239. Moor, R. M. and Rowson, L. E. A. (1966b). Local maintenance of the corpus luteum in sheep with embryos transferred to various isolated portions of the uterus. J. Reprod. Fert. 12, 539-550. Nett, T. M., Staigmiller, R. B., Akbar, A. M., Diekman, M. A., Ellenwood, W. E. and Niswender, G. D. (1976). Secretion of prostaglandin F20 in cycling and pregnant ewes. J. Anim. Sd. 42, 876-880. Perry, J. S., Heap, R. B. and Burton, R. D. (1975). Oestrogen metabolism by trophoblast and endometrium, and its probable relation to luteal maintenance in the pig. J. Endocrin. 64, 40P-41P. Pexton, J. E., Weems, C. W. and Inskeep, E. K. (1975). Prostaglandins F in uterine venous plasma, ovarian arterial and venous plasma and in ovarian and luteal tissue of pregnant and nonpregnant ewes. j. Anim. Sd. 41, 154-1 59. Rowson, L E. A. and Moor, R. M. (1967). The influence of embryonic tissue homogenate infused into the uterus on the lifespan of the corpus luteum in the sheep. J. Reprod. Fert. 13, 511-516. wilson, L, Jr., Cenedella, R. J., Butcher, R. L. and lnskeep, E. K. (1972a). Levels of prostaglandinsin the uterine endometrium during the ovine estrous cycle. J. Arnm. Sci. 34, 9 3-99. Wilson, L, Jr., Butcher, R. L. and Inskeep, E. K. (1972b). Prostaglandin F20 in the uterus of ewes during early pregnancy. Prostaglandins 1,479482. RECOMMENDED REVIEWS Moor, R. M. (1968). Effect of embryo on corpus luteum function. J. Anim. Sci. Suppl. 1. 29, 97-118. Ginther, 0. J. (1974). Internal regulation of physiological processes through local venoarterial pathways: a review. J. Anim. Sci. 39, 5 50-564.