Repeated exposure to prolactin is required to induce luteal regression in the hypophysectomized rat.

We investigated whether prolactin (PRL) treatments resembling the intermittent PRL surges of estrous cycles could induce luteal regression in hypophysectomized rats. Immature female rats were stimulated to ovulate and form corpora lutea with exogenous gonadotropins, and were hypophysectomized following ovulation. A single s.c. injection of either vehicle (VEH) or PRL was administered to each rat on post-hypophysectomy Day 8 and again on Day 11. The four resulting treatment groups consisted of rats that received two injections of VEH, VEH followed by PRL, PRL followed by VEH, or two injections of PRL. Rats were killed 24 or 72 h following the second injection. Plasma 20alpha-dihydroprogesterone, luteal weight, and total luteal protein were determined. One ovary was sectioned for immunohistochemistry for monocytes/macrophages, apoptotic nuclei, and major histocompatibility class II (MHC II) molecules. No effect of time (following injection) was observed on any endpoint, indicating that PRL does not have an ongoing regressive action. Time groups from within each treatment group were therefore pooled for analysis. Significant declines (P: < 0.05) in plasma concentrations of 20alpha-dihydroprogesterone, luteal weight, and protein per corpus luteum occurred only after two injections of PRL. Numbers of luteal monocytes/macrophages, apoptotic nuclei, and MHC II-positive cells were low in all groups; numbers of luteal monocytes/macrophages increased following two injections of PRL (P: < 0.05). We conclude that PRL has a cumulative regressive effect on the corpus luteum of the hypophysectomized rat. Drawing a parallel with the estrous cycle, we suggest that continued exposure to PRL, over several cycles, is necessary to induce full luteal regression.

Downregulation of long-form prolactin receptor mRNA during prolactin-induced luteal regression.

OBJECTIVE: Prolactin is capable of both trophic and lytic actions in rat corpora lutea. In corpora lutea responding to a trophic prolactin signal, the long form of the prolactin receptor is the dominant form and is upregulated by prolactin. We investigated whether mRNA for the short form of the prolactin receptor was dominant in corpora lutea responding to a lytic prolactin signal, and whether the relative concentrations of the mRNAs for both forms of the prolactin receptor were changed during this response. DESIGN AND METHODS: Immature rats were ovulated by injection of 5 IU equine chorionic gonadotrophin and 5 IU human chorionic gonadotrophin, and were hypophysectomized shortly after ovulation. Nine days after hypophysectomy, rats were injected with prolactin (500 microg/day) or vehicle for 24 (n=6, n=6) or 72 h (n=13, n=5). Total RNA was isolated from corpora lutea and mRNA for both types of prolactin receptor were analyzed by semiquantitative RT-PCR using the ribosomal protein S16 as the internal control. RESULTS: The intensities of the long- and short-form prolactin receptor signals were normalized to the S16 internal control and expressed as relative densitometric units. The normalized values at 24h for prolactin-treated vs vehicle-treated rats were 0.23 +/- 0.05 vs 0.49 +/- 0.15 (P>0.05) for the short form and 4.04 +/- 0.8 vs 4.23 +/- 0. 6 (P>0.05) for the long form. The values for 72 h were 0.30 +/- 0.05 vs 0.24 +/- 0.05 (P>0.05) for the short form and 2.76 +/- 0.4 vs 5. 53 +/- 0.3 (P<0.01) for the long form respectively. CONCLUSION: The long form of the prolactin receptor is the dominant form at both time-points; however, the concentration of mRNA for this receptor isoform was specifically downregulated by prolactin treatment. Our results suggest that the short form of the prolactin receptor alone is unlikely to mediate the luteolytic action of prolactin, but that luteolytic events may be influenced via a change in the ratio of the two receptor isoforms.

Effects of a 3beta-hydroxysteroid dehydrogenase inhibitor on monocyte-macrophage infiltration into rat corpus luteum and on apoptosis: relationship to the luteolytic action of prolactin.

The administration of prolactin to hypophysectomized rats results in regression of the corpora lutea, accompanied by immune-inflammatory events such as infiltration of monocytes and macrophages. Recent reports indicate an autocrine role for progesterone during the lifespan of the corpus luteum. In the present study, an inhibitor of 3beta-hydroxysteroid dehydrogenase, Trilostane, was used to investigate the hypothesis that a decrease in luteal tissue steroids precipitates the cascade of immune-inflammatory events leading to luteal regression in prolactin-treated hypophysectomized rats. Immature rats were induced to ovulate by administering eCG-hCG, and hypophysectomized on the day after ovulation (at 32 days of age). Rats were injected s.c. 9-11 days after hypophysectomy with (a) Trilostane (80 mg kg(-1) day(-1)), (b) ovine prolactin (500 mg day(-1)), (c) Trilostane plus prolactin, or (d) vehicle. Plasma and luteal tissue progesterone and 20alpha-dihydroprogesterone ('progestin') were quantified; luteal tissue monocytes-macrophages and apoptotic nuclei were counted, and luteal wet mass was determined. Rats treated with prolactin alone showed the expected markers of luteal regression: decreased plasma progestin, increased numbers of monocytes-macrophages and apoptotic nuclei in luteal tissue, and decreased luteal wet mass; however, progestin concentration in luteal tissue was unchanged. Treatment with Trilostane reduced plasma and luteal tissue progestin, but did not result in an infiltration of monocytes-macrophages or increased numbers of apoptotic nuclei in the corpora lutea, or any change in luteal wet mass. Trilostane in combination with prolactin reduced plasma and luteal tissue progestin and produced the expected markers of regression, with the exception of luteal tissue mass, which remained unchanged. In conclusion, inhibition of steroidogenesis does not initiate luteal regression or augment prolactin-induced luteal regression in hypophysectomized rats. Prolactin-induced infiltration of monocytes-macrophages is not accompanied by a decrease in luteal tissue progestin, at least in the early stages of luteal regression.

The proestrous prolactin surge is not the sole initiator of regressive changes in corpora lutea of normally cycling rats.

During the estrous cycle, secretion of prolactin is largely restricted to a surge on proestrus. We investigated whether this proestrous prolactin surge initiates regression of the corpora lutea of the preceding cycle. Adult rats were killed prior to the prolactin surge (Proestrus group), following the prolactin surge (Estrus group), after chemical blockade of the prolactin surge with bromocryptine (Estrus+BRC group), and after blockade of the prolactin surge and administration of prolactin (Estrus+BRC+PRL group). Corpora lutea of the current (proestrus) or preceding (estrus) cycle were dissected out, weighed, and sectioned for immunohistochemistry or cultured for examination of in vitro progestin production. Numbers of luteal monocytes/macrophages, differentiated macrophages, and apoptotic nuclei per high-power field were greater for Estrus and Estrus+BRC+PRL than for Estrus+BRC, which in turn had greater numbers than Proestrus (P< 0.05). In contrast, BRC completely reversed the decline in luteal weight observed between Proestrus and Estrus (P<0.05). Number of major histocompatibility complex II-positive cells was not different between groups (P>0.05). Finally, progestin production by corpora lutea in vitro was lower for Proestrus than for the other groups (P<0.05). The results indicate that the prolactin surge alone is not responsible for initiation of apoptosis or immune cell infiltration in regressing corpora lutea of the estrous cycle, although prolactin increases these markers of regression. Prolactin does cause a decline in luteal weight; however, the corpora lutea retain the capacity for steroidogenesis. We conclude that although prolactin has a role in luteal regression, it is not solely responsible for the initiation of this process.

Glucocorticoids stimulate the accumulation of lipids in the rat corpus luteum.

We investigated the physiological basis for the trophic effect of glucocorticoids in rat corpora lutea in the absence of pituitary gonadotropins. Immature (Day 29) Sprague-Dawley rats were given eCG and hCG to induce the development of corpora lutea and were hypophysectomized on Day 32. Beginning on Day 40, rats received twice-daily s.c. injections of either dexamethasone (dex; 200 microg/rat/day) or vehicle (controls) and then were killed on Day 44. Plasma 20alpha-dihydroprogesterone, a major steroid produced by the corpora lutea, was higher (p 2-fold of plasma 20alpha-dihydroprogesterone concentration compared to controls. Glucocorticoid receptor protein (about 92 kDa) was detected in both luteal and nonluteal ovarian tissues in this animal model. These effects of glucocorticoids and the presence of the glucocorticoid receptor raise the possibility of a physiological role for glucocorticoids in the rat corpus luteum.

Estrogen withdrawal induces macrophage invasion in the rabbit corpus luteum.

Macrophages within the corpus luteum are associated with spontaneous luteal regression in a number of species. However, an understanding of the consequences of macrophage recruitment on the functional capacity and responsiveness of the luteal tissue has remained elusive. Here we investigate the temporal appearance of macrophages and their potential impact in corpora lutea of rabbits, in which a rapid fall in progesterone synthesis and premature regression of the corpus luteum are initiated by withdrawal of the luteotropic hormone estradiol-17beta. Removal of estradiol implants, placed subcutaneously, induced a significant increase in the average number of macrophages per high-power field (hpf) in corpora lutea (p < 0.05) within 72 h. Replacement of the estradiol implants 48 h after their removal resulted in a marginal rebound of plasma progesterone and a variable number of luteal macrophages (range: 6-160 macrophages/hpf) among the 11 rabbits. A third experiment revealed that the relative numbers of macrophages within the corpora lutea have no apparent relationship to rates of progesterone synthesis in vitro: progesterone production (ng/mg tissue) did not differ (p > 0.05) between corpora lutea of estradiol-maintained rabbits and those of estradiol-replaced rabbits despite obvious differences in numbers of luteal macrophages (2 +/- 1 vs. 42 +/- 10 macrophages/hpf, respectively; p < 0.05). We conclude that the entry/recruitment of macrophages into the rabbit corpus luteum is sensitive to the luteotropic hormone estradiol-17beta and that the presence of macrophages does not preclude the continuation of progesterone production in surviving luteal tissue revitalized after estradiol removal/replacement.

Expression of the steroidogenic acute regulatory protein in the corpus luteum of the rabbit: dependence upon the luteotropic hormone, estradiol-17 beta.

The recent characterization of the mitochondrial protein, Steroidogenic Acute Regulatory (StAR) protein, as a rate-limiting protein in steroidogenesis prompted us to investigate whether StAR is expressed in the rabbit corpus luteum and whether the expression of StAR is responsive to estradiol-17 beta, the luteotropic hormone in the rabbit. In rabbits treated continuously with exogenous estradiol through Day 13 of pseudopregnancy (n = 9), immunoblot analysis revealed that luteal expression of StAR was stable, ranging from 8.5 to 9.7 U of corrected integrated optical density. Plasma progesterone concentration (mean +/- SEM) remained elevated in these rabbits (14.3 +/- 2.1 ng/ml). In contrast, expression of StAR decreased in corpora lutea of rabbits deprived of estradiol for the last 48 and 72 h of the experiment (4.9 +/- 2.2 and 0.3 +/- 0.2 U, respectively, n = 3 per group), and was associated with a decline in plasma progesterone (0.8 +/- 0.1 and 0.5 +/- 0.3 ng/ml, respectively). Replacement of estradiol after 48 h of estradiol deprivation (n = 3) stimulated the reappearance of StAR (10.3 +/- 2.6 U) and the restoration of plasma progesterone (10.4 +/- 4.9 ng/ml). [35S]Methionine labeling of proteins in rabbit corpora lutea revealed that several isoforms of StAR protein were specifically synthesized in response to estradiol treatment. Collectively, these observations are consistent with a proposed role for StAR in the mediation of the luteotropic effect of estrogen to promote the synthesis of progesterone in the rabbit.

Prolactin-induced regression of the rat corpus luteum: expression of monocyte chemoattractant protein-1 and invasion of macrophages.

Monocyte chemoattractant protein-1 (MCP-1) is a potential mediator of the recruitment of monocytes/macrophages into the regressing corpus luteum (CL). We investigated whether the luteolytic effect of prolactin in the rat is associated with the expression of MCP-1 and an invasion of monocytes/macrophages. Ovulation was induced in immature female rats by injection of eCG (5 IU, s.c.) at 30 days of age. All rats were hypophysectomized 3 days later. Rats received injections of ovine prolactin (250 micrograms, s.c.) at 12-h intervals on Day 9, 10, and 11 posthypophysectomy; controls received injection of vehicle. Rats were killed by decapitation 24, 48, or 72 h after the first injection of prolactin or vehicle. In rats treated with prolactin, immunoreactive MCP-1 was detected in the CL at 24 h after the first injection, and a consistent level of staining was reached by 72 h with immunodetectable MCP-1 diffused throughout individual CL. The number of monocytes/macrophages in the CL (mean +/- SEM) increased significantly after prolactin treatment, from 3.1 +/- 1.8 at 24 h to 49.3 +/- 8.2 at 72 h (p < 0.05), and the number of monocytes/macrophages was different from that in control, vehicle-treated rats at 72 h (10.3 +/- 4.1; p < 24 and 72 h in prolactin-treated rats (p < 0.05). It is concluded that a potentially important component of the luteolytic effect of prolactin in the rat is the expression of MCP-1 and invasion of monocytes/macrophages into the CL.

Expression of monocyte chemoattractant protein-1 in the corpus luteum of the rat.

The known accumulation of macrophages in corpora lutea (CL) at the time of luteal regression prompted us to investigate whether the chemoattractant protein monocyte chemoattractant protein-1 (MCP-1) is expressed in the rat CL. On the day of confirmed mating (Day 0 of pregnancy), regressing CL from the previous (nonfertile) estrous cycle contained immunodetectable MCP-1 and numerous monocytes/macrophages, whereas the newly formed CL of pregnancy, within the same ovary, contained little MCP-1 and few monocytes/macrophages. MCP-1 diminished in the regressing CL on Days 3 and 9 of pregnancy, although numerous monocytes/macrophages remained. The CL of pregnancy on Days 3 and 9 of pregnancy contained minimal MCP-1 and relatively few monocytes/macrophages. By Days 17 and 21 of pregnancy, however, prior to parturition and prior to an accumulation of monocytes/macrophages, expression of MCP-1 increased in the CL of pregnancy. Northern blots revealed a resurgence of luteal MCP-1 mRNA on Day 21 of pregnancy: 3805 +/- 1077 on Day 21 vs. 1059 +/- 177 on Day 9 (p < 0.05; expressed as densitometric units relative to beta-actin). In conclusion, the expression of MCP-1 in the rat CL in association with, or preceding, the appearance of monocytes/macrophages at the time of luteal regression is consistent with the known role of MCP-1 as a potent chemoattractant for monocytes/macrophages. This suggests that MCP-1 might have a prominent role in the immunological process of luteal regression.

Estrogen uncouples steroidogenesis from 3',5'-cyclic adenosine monophosphate regulation in the rabbit corpus luteum.

The hypothesis was investigated that estradiol, the luteotrophic hormone in the rabbit, uncouples luteal progesterone production from regulation by LH/cyclic AMP. Progesterone production by corpus luteum (CL) incubated with vehicle, 3-isobutyl-1-methyl-xanthine (IBMX), hCG, or hCG+IBMX was compared in pseudopregnant rabbits treated continuously with estradiol (estradiol-maintained), withdrawn from estradiol for 24-48 h (estradiol-withdrawn), or withdrawn and then replaced with estradiol for 6 or 24 h (estradiol-replaced). Progesterone production in estradiol-maintained rabbits was not altered by hCG and/or IBMX, but was stimulated significantly in estradiol-withdrawn rabbits. This response was reversed (i.e., abolished) in estradiol-replaced (24 h) rabbits. The loss of responsiveness to hCG was not attributable to impaired accumulation of cyclic AMP: basal and hCG-stimulated cyclic AMP concentrations were similar in luteal tissues of estradiol-maintained and estradiol-withdrawn rabbits. The loss of responsiveness to hCG was also not a consequence of maximal progesterone production: CL of estradiol-replaced (6 h) rabbits were also insensitive to hCG, and this occurred before progesterone production attained a maximal rate. We conclude that a striking feature of the luteotrophic action of estrogen is to uncouple the regulation of progesterone production from cyclic AMP.

The autonomy of the rabbit corpus luteum.

The rabbit corpus luteum possesses LH receptors that are coupled to adenylyl cyclase, but paradoxically it does not require LH as a luteotrophic factor for the maintenance of progesterone secretion. This suggests that rabbit luteal cells may not respond physiologically to LH. Therefore, the present study was undertaken to investigate the responsiveness of the rabbit corpus luteum of pseudopregnancy to human chorionic gonadotrophin (hCG) which acts on the same receptor as LH. Pseudopregnancy was induced by injection of 40 IU pregnant mare serum gonadotrophin followed 50 h later by an injection of 40 IU hCG (day 0). On days 7 and 11 of pseudopregnancy, corpora lutea were obtained and incubated for 2 or 5 h in the presence of either 0.1 or 1 microgram/ml hCG or 1 mM monobutyryl cyclic AMP (bcAMP). Neither hCG nor bcAMP stimulated progesterone production by the isolated corpus luteum, despite a sustained high rate of progesterone production by the tissue throughout the incubation period. By contrast, Graafian follicles removed from the same ovaries and incubated under the same conditions responded both to hCG and bcAMP with large increases in progesterone production. To determine whether the cyclic AMP content of the corpus luteum was altered by in vitro exposure to hCG, day 7 and day 11 corpora lutea were incubated for 5 or 15 min with various concentrations of hCG, and cyclic AMP in the tissue was then measured. Even at the highest concentration of hCG tested (10 micrograms/ml), the cyclic AMP content of the corpus luteum was unaltered.(ABSTRACT TRUNCATED AT 250 WORDS)

Effect of treatment with methylprednisolone on duration of pseudopregnancy and on macrophages and T lymphocytes in rabbit corpora lutea.

The potential role of macrophages and T lymphocytes in the destruction of the corpus luteum at the end of the luteal phase was investigated by treating pseudopregnant rabbits with the immunosuppressant glucocorticoid methylprednisolone. Eleven specific pathogen-free New Zealand White rabbits were injected with pregnant mares' serum gonadotrophin (40 iu, i.m.), followed 2 days later by human chorionic gonadotrophin (40 iu, i.v.) to stimulate ovulation. The following day (day 1 of pseudo-pregnancy) all animals had an oestradiol-filled Silastic capsule implanted s.c., to ensure that oestradiol, the luteotrophic hormone in this species, would not be limiting. From day 10 of pseudopregnancy, three animals were injected with a low dose of methylprednisolone (2 mg kg-1 per day) until day 20. Three other animals were injected with a higher dose of methylprednisolone (20 mg kg-1 per day) from day 13 of pseudopregnancy until day 19. Five animals served as control, vehicle-injected animals. Blood samples were taken at intervals and assayed for progesterone. Immunofluorescence was used to stain luteal tissue for macrophages, T lymphocytes and class II antigens, and positive cells were counted under high-power magnification. Methylprednisolone treatment reduced (by about 70%), but did not eliminate, the macrophages in the regressing corpora lutea. In contrast, the high dose of methylprednisolone essentially eliminated T lymphocytes, and reduced (by about 90%) the number of cells expressing class II antigen in the luteal tissue. Despite the effects of methylprednisolone on these cells, serum progesterone profiles were not altered by treatment with methylprednisolone, and pseudopregnancy was of normal duration.(ABSTRACT TRUNCATED AT 250 WORDS)

Insulin-like growth factor-I stimulates steroidogenesis in rabbit luteal cells.

A potential role of insulin-like growth factor I (IGF-I) in the regulation of steroidogenesis in the rabbit corpus luteum was investigated using a primary culture of luteal cells obtained 3 days after ovulation. Dissociated cells were cultured for 1 day in the presence of medium 199 and 10% fetal bovine serum; thereafter, the cells were cultured in medium 199 containing 0.1% BSA, gentamicin (50 micrograms/ml), and hormones or growth factors, and without serum. IGF-I (human recombinant, 100 ng/ml) was as effective as LH (ovine, 10 ng/ml) in maintaining progesterone accumulation through 4 days of culture. Estradiol (10(-8) M), either alone or in combination with LH or IGF-I failed to stimulate progesterone accumulation, which was not surprising since these cells did not possess estrogen receptors. The stimulation of progesterone by IGF-I was not detectable until 24-36 h after introduction of the growth factor to the cultures, whereas stimulation by LH was observed within 2 h. The steroidogenic effect of IGF-I was not attributable to increased cell number, as DNA values or [3H]thymidine incorporation were unchanged by IGF-I. IGF-I increased functional enzymatic activity, observed as increased progesterone accumulation in the presence of 25-hydroxycholesterol used as exogenous substrate. These data indicate that luteal cells have the capacity to respond to IGF-I, raising the possibility that IGF-I has a role in the regulation of steroidogenesis in the rabbit corpus luteum.

Luteal enzymes of the luteinizing hormone and beta-adrenergic signal transduction pathways in hypophysectomized rabbits do not require pituitary hormone support.

Experiments were conducted to determine whether continuous pituitary hormone support is required for expression of the LH and beta-adrenergic cAMP signal transduction pathways in rabbit CL during pseudopregnancy. Parameters of the LH and catecholamine cAMP signal transduction pathways in CL of estrogen-treated hypophysectomized rabbits were compared to those of pituitary-intact rabbits. Results showed that each of the parameters of the LH and beta-adrenergic cAMP signal transduction pathways was retained in CL taken from estrogen-treated pseudopregnant rabbits that had been hypophysectomized for as long as 13 days at levels not significantly different from those of estrogen-treated pituitary-intact rabbits. These included luteal basal, and LH-, epinephrine-, and fluoride-stimulated adenylyl cyclase activities; total luteal cAMP levels; the number and affinity of cAMP-dependent protein kinase regulatory subunit cAMP binding sites; binding activity of the type I and type II regulatory subunits; and the amount of catalytic subunit protein of cAMP-dependent protein kinase. We conclude that expression of the proteins of the cAMP signal pathway for LH and beta-adrenergic hormones in CL of estrogen-treated rabbits does not require pituitary hormone support.

The biosynthesis of cholesterol side-chain cleavage cytochrome P-450 in the rabbit corpus luteum depends upon estrogen.

To gain a better understanding of the luteotropic action of estrogen, we have investigated the effect of estrogen on the synthesis of the enzyme, cholesterol side-chain cleavage cytochrome P-450 (P-450scc) in the rabbit corpus luteum. Using an established protocol, rabbits were treated with estradiol, and the estradiol was then withdrawn on day 9 of pseudopregnancy, which caused an 88% fall in serum progesterone within 48 h. In other rabbits, estradiol was replaced at 48 h which stimulated a 6.6-fold increase in serum progesterone concentration within the next 24 h. Luteal tissues were incubated with [35S]methionine and homogenized, and a mitochondrial fraction lysate was obtained. Equal trichloroacetic acid-precipitable radioactivity was taken for immunoprecipitation using a well-characterized polyclonal antiserum against bovine adrenal P-450scc. The immunoisolated proteins were resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, and radioactivity was visualized by autofluorography. The results indicate that the rate of synthesis of P-450scc in 48 h-estradiol withdrawn animals was markedly reduced, and by 72 h of withdrawal was barely detectable. When estradiol was reintroduced, the synthesis of P-450scc was increased. Despite the prominent changes in P-450scc synthesis, immunoblotting revealed only a minimal (approximately 30%) decrease in relative P-450scc content by 72 h after estradiol withdrawal. Analyses of DNA and protein contents of luteal tissues revealed an increase in DNA per mg luteal tissue, a decline in total tissue protein/DNA ratio, but no change in mitochondrial fraction protein/DNA ratio after estrogen withdrawal. The results indicate that de novo synthesis of P-450scc in the corpus luteum is sensitive to estrogen; however, the estrogen-sensitive rate-limiting step(s) for steroidogenesis are at other sites in the steroid biosynthetic pathway.

Tumor necrosis factor production and accumulation of inflammatory cells in the corpus luteum of pseudopregnancy and pregnancy in rabbits.

The potential involvement of macrophages, T lymphocytes, and the cytokine tumor necrosis factor (TNF) in regression of the corpus luteum was investigated at different stages of pseudopregnancy and pregnancy by use of immunocytochemical methods and a TNF bioassay. Few macrophages (11 +/- 6 per high power field of 8-microns frozen sections of corpus luteum, Day 10 of pseudopregnancy) were observed until the very end of pseudopregnancy, when the number of macrophages increased greatly (176 +/- 42 per high power field, Day 19 of pseudopregnancy). Pregnancy, of 32 days duration, delayed large-scale macrophage accumulation until 3 days after parturition (154 +/- 30 per high power field). Low TNF activity (approximately 1.0 U/mg protein) was detected in incubations of luteal tissue at all stages; in response to lipopolysaccharide, TNF values in medium increased 10- to 30-fold at times of luteal regression and macrophage accumulation (1 day postpartum and Day 19 of pseudopregnancy). Class II-positive T lymphocytes were observed in luteal tissue, but unlike macrophages, the number of lymphocytes did not increase at the time of regression of the corpus luteum. These data are consistent with the hypothesis that involution of the corpus luteum is promoted through the interactions of inflammatory cells and action of TNF, although the action of TNF has not been determined in this luteal tissue. Through unknown mechanisms, pregnancy postpones the accumulation of macrophages in the corpus luteum, in association with the prolongation of luteal function until the time of parturition.

Physiological and immunocytochemical evidence for a new concept of blood flow regulation in the corpus luteum.

To explain the high rate of blood flow in the corpus luteum, we hypothesize that luteal blood vessels offer minimal resistance to flow and are incapable of vasomotion. This hypothesis was tested in rabbits at mid-pseudopregnancy by measuring blood flow in the corpus luteum and ovarian stroma with tracer-labeled microspheres at three levels of arterial blood pressure, which was manipulated by constricting the aorta above the ovarian artery. In addition, the distribution of vascular smooth muscle in the ovary was evaluated with morphological and immunocytochemical techniques. Decreases in arterial pressure were paralleled by reductions in blood flow in the corpus luteum, whereas ovarian stromal blood flow was unchanged. Consistent with our hypothesis, there was no change in the low level of vascular resistance offered by blood vessels in the corpus luteum, supporting the view that they are maximally dilated and incapable of autoregulation. Morphologically, the vessels within the corpus luteum appeared as large sinusoidal capillaries without smooth muscle, providing an anatomical explanation for the lack of vasomotor control demonstrated physiologically. The absence of vascular smooth muscle was confirmed with immunocytochemistry using an antibody against the muscle-specific intermediate filament, desmin. The fluorescein-labeled antibody decorated arteries and arterioles within the ovarian stroma and near the capsule of the corpus luteum, but did not decorate vessels in the corpus luteum of pseudopregnancy, providing additional evidence that the vessels of the corpus luteum lack the smooth muscle investment necessary to change vascular caliber. From these findings, we have proposed a novel scheme to explain intraovarian blood flow regulation. Vascular resistance in the ovarian stroma, as in most tissues, is acutely regulated by dilation or constriction of intratissue arterioles. In contrast, vascular resistance within the corpus luteum is modeled as a relatively invariable parameter, fixed at a low level by the morphological characteristics of the luteal vasculature. Therefore, the corpus luteum operates on a linear (maximally "vasodilated") pressure-flow curve, does not actively regulate intratissue blood flow, and is subject to acute regulation of perfusion only through changes in extra-luteal vessels.

Regulation of blood flow to the rabbit corpus luteum: effects of estradiol and human chorionic gonadotropin.

We tested the hypothesis that estrogen and hCG can modify blood flow in the rabbit corpus luteum. Radioactively labeled microspheres were used to measure luteal blood flow in pseudopregnant rabbits in which estrogen had been withdrawn to initiate premature luteal regression and in pseudopregnant rabbits injected with hCG. Removal of estradiol-filled Silastic capsules on day 10 of pseudopregnancy caused an 80% decrease in the serum progesterone concentration within 24 h. Despite the decline in progesterone secretion, luteal blood flow remained at very high levels and was not different from that in control rabbits treated continuously with estradiol. Replacement of estradiol-filled capsules for 3 h did not change the high rate of blood flow to the corpus luteum, but blood flow in the uterus, vagina, and ovarian stroma was increased. The injection of hCG (10 IU, iv) on day 10 of pseudopregnancy caused a 3-fold increase in blood flow to the nonluteal portion of the ovary and a 3-fold increase in the serum progesterone concentration, but luteal blood flow did not change. We conclude that the acute actions of estradiol or hCG in the rabbit corpus luteum are not mediated by changes in luteal blood flow. Further, the results suggest that the luteal vasculature is regulated differently from the vasculature of other estrogen-responsive tissues and that blood flow in the nonluteal tissues of the ovary can be regulated independently of blood flow in the corpus luteum.

Relationship between blood flow and steroidogenesis in the rabbit corpus luteum.

Blood flow in the corpus luteum of the pseudopregnant rabbit was measured with tracer-labelled microspheres before and at 1 and 3 h after saline treatment (N = 8) or after inhibition of progesterone synthesis with aminoglutethimide (N = 10). Before treatment luteal blood flow (29.5 +/- 3.9 ml/min.g-1 (mean +/- s.e.m.] was much higher than blood flow to other tissues (ovarian stroma = 2.9 +/- 0.6; uterus = 0.5 +/- 0.1; adrenal gland = 2.6 +/- 0.2 ml/min.g-1). Aminoglutethimide reduced serum progesterone by 60% within 1 h but luteal blood flow was unchanged (26.2 +/- 3.5 ml/min.g-1). At 3 h after aminoglutethimide, serum progesterone remained low and luteal blood flow was slightly reduced to 22.5 +/- 3.4 ml/min.g-1. This reduction was associated with a significant decline in mean arterial blood pressure which resulted in luteal vascular resistance being unaltered by aminoglutethimide treatment. Further analysis of these data indicated that serum progesterone concentration was not significantly correlated with blood flow to the corpora lutea or with blood flow to other tissues. In contrast, mean arterial blood pressure was highly correlated with blood flow to the corpus luteum (r = 0.80; P less than 0.001) but not to the ovarian stroma (r = 0.04), or adrenal gland (r = 0.06). These results indicate that luteal blood flow is not acutely responsive to changes in luteal progesterone production and suggest that luteal blood flow changes passively with changes in arterial blood pressure.

Tumor necrosis factor-a (TNF-a) production and localization of macrophages and T lymphocytes in the rabbit corpus luteum.

Utilizing immunocytochemistry numerous macrophages were localized in regressing corpora lutea. In contrast, few macrophages were observed in young corpora lutea. Regressing corpora lutea readily produced TNF-a in vitro in response to lipopolysaccharide, whereas young corpora lutea produced significantly less TNF-a. T lymphocytes were identified in young corpora lutea preceding the appearance of macrophages. These observations suggest that cells of the immune system and cytokines could be important participants in physiological regression of the corpus luteum.