Cyclooxygenase 2 messenger RNA levels in canine follicular cells: interrelationship with GDF-9, BMP-15, and progesterone.

Cyclooxygenase 2 (COX-2) encoded by the Cox-2 gene within the periovulatory follicles is a critical mediator of oocyte development. Growth differentiation factor 9 (GDF-9) and bone morphogenetic protein 15 (BMP-15) participate in the modulation of certain target genes in the ovary, possibly influencing the Cox-2 gene expression. However, this relationship has not been characterized in canines. This study aimed to examine the possible relationships among BMP-15, GDF-9, progesterone, and Cox-2 gene expression in granulosa-cumulus cells in dogs. Granulosa cells from antral follicles and their corresponding cumulus-oocyte complexes and follicular fluid (FF) were separately obtained from 56 ovaries collected from adult bitches at estrus (n = 15) and proestrus (n = 13) after ovariohysterectomy. Total RNA extraction was performed in follicular cells, and Cox-2 gene expression was assessed by quantitative PCR analysis. Progesterone, BMP-15, and GDF-9 were determined in the FF samples using ELISA assays. Cumulus-oocyte complexes were subjected to in vitro maturation (IVM) with or without (control) recombinant GDF-9 and BMP-15. After 72 h of culture, Cox-2 transcript analyses were performed in cumulus cells via quantitative PCR. Data were evaluated by ANOVA. An increase (P < 0.05) in Cox-2 messenger RNA levels was observed in follicular cells from follicles at estrus with respect to those at proestrus. However, the levels of BMP-15 and GDF-9 in FF decreased (P < 0.05), whereas progesterone increased (P < 0.05) from the proestrus phase to the estrus phase. The expression of Cox-2 gene in cumulus cells was 4-fold greater (P < 0.01) than that in the control when both growth factors were added to the IVM culture. In conclusion, although BMP-15 together with GDF-9 appears to upregulate the levels of Cox-2 transcripts during IVM, the inverse relationship of these paracrine factors with Cox-2 gene expression and the positive correlation of progesterone with Cox-2 transcripts suggest that the high progesterone levels could be more relevant in the local mechanisms regulating the Cox-2 gene expression.

The impact of prenatal environment on postnatal life and performance: Future perspectives for prevention and treatment.

The present review aims to offer a non-comprehensive outline of the current state-of-the-art and future perspectives on management and therapeutic tools for intrauterine growth restriction (IUGR) and associated prenatal programming in both human and animal species. Animals are used as models for the study of phenomena related to IUGR, but also for research on prenatal therapies with the main objective of designing and developing preventive and therapeutic strategies. The research is currently paying attention on maternal-focused pharmacological treatments and nutritional strategies but also on fetal-focused treatments. Fetal-focused treatments, administered either directly at the fetus or by using infusion of umbilical cord, amniotic sac or placenta, which avoids the administration of substances at high doses to the mother for allowing their availability at the fetoplacental level. The results obtained in this area of research using large animals (rabbits, pigs and ruminants) have a dual interest, for translational biomedicine and for veterinary medicine and animal production.

Ram semen deterioration by short-term exposure to high altitude is prevented by improvement of antioxidant status.

Ovine reproduction efficiency in herds at high altitude (ha) is lower than that at low altitude (la). In ewes, ha effects are due to hypoxia and oxidative stress. Our aim was to establish the effect of antioxidant vitamin supplementation on semen traits and antioxidant status of rams exposed to short or long time ha. A total of 32 rams native to la (~500 m) were used, 16 were kept at la and the other 16 were brought to ha (~3600 m), where they were placed in the same flock as the ha native rams (n=16). Half of the animals in each group were supplemented daily with vitamins C 600 mg and E 450 IU per os, during the entire experimental period, starting the 4th day after animal's arrival at ha (day 0). At days 0, 30 and 60 of treatment, blood and semen samples were collected for evaluation of antioxidant status and semen standard characteristics. Data were compared within each experimental time by analysis of variance using a general linear model. Elevated concentrations of oxidative stress biomarkers were present in blood from animals maintained at ha. Ejaculates from ha exposed rams showed decreased sperm concentration, progressive motility and viability, in addition to decreased antioxidant status in seminal fluid. A total of 30 days of oral supplementation with vitamins C and E prevented some ha negative effects on semen characteristics, mainly in recently ha exposed rams. It is concluded that exposure of rams to ha negatively affects semen quality, where oxidative stress plays a predominant role. These effects are mainly prevented by oral supplementation of vitamins C and E, which constitutes a simple and cheap alternative to improve semen quality of rams when they are moved to ha.

Fertility of a high-altitude sheep model is compromised by deficiencies in both preovulatory follicle development and plasma LH availability.

At high altitude, hypoxia and/or oxidative stress may compromise fertility. This study tested the relative effect of short- or long-term exposure to high-altitude hypobaric hypoxia and oxidative stress in sheep on preovulatory follicle dynamics and gonadotrophin secretion. Thus, growth dynamics, stereidogenic function and competence to ovulate of preovulatory follicles, as well as FSH and LH availability throughout the entire oestrous cycle, were compared among sheep native from low and high altitude, and sheep newcomers to high altitude. The results indicates that short-term exposure in sheep newcomers to high altitude has a deleterious effect on both the ovarian function (affecting preovulatory follicular development) and the pituitary function (diminishing plasma LH availability). On the other hand, there were no detected differences in the preovulatory follicular development in sheep adapted to high altitude for generations and, conversely, LH secretion was increased, which suggests an adaptive mechanism. The treatment with antioxidant agents during a relative short period for the time of folliculogenesis (approximately 1 month and a half) changed substantially the development of preovulatory follicles in short-term exposed sheep to similar patterns than in sheep native and living to both high and low altitude. These results highlight the role of oxidative stress in the detriment of the reproductive function in individuals recently exposed to high-altitude hypoxic environment.

Sperm nuclear decondensation induction capacity of in vitro and in vivo matured canine oocytes.

The objectives of this study were to evaluate the sperm nuclear decondensation capacity of ovulated and in vitro-matured (IVM) canine oocytes during different culture times and correlate this decondensation ability with the state of oocyte nuclear maturation in vitro and in vivo. Fresh ejaculates from three dogs were used for in vitro fertilization (IVF). Dog spermatozoa were cocultured with ovulated or IVM oocytes after each culture period (0, 48, 72 and 96 h) for 24 h. The nuclear stage of the oocytes and the appearance of the sperm nucleus were determined, and data were analysed with a chi-square test. The rates of decondensation and meiotic development in IVM oocytes increased up to 72 h of culture. In contrast, almost all in vivo-matured oocytes showed MII nuclear stage and sperm chromatin decondensation. The percentages of oocytes at MII stage were much lower (p < 0.05) in all IVM groups compared with ovulated oocytes; the rate of sperm chromatin decondensation was higher in ovulated oocytes than in those matured in vitro. Thus, IVM canine oocytes are able to decondense the sperm chromatin during IVF, and this ability increases with time. Nevertheless, sperm chromatin decondensation is less efficient than in ovulated oocytes and may not be completely synchronized with nuclear development as it occurs in vivo.

In vitro fertilization and development of cumulus oocytes complexes collected by ultrasound-guided follicle aspiration in superstimulated llamas.

The objective was to evaluate the developmental competence of cumulus-oocyte complexes (COC) collected by follicular aspiration in llamas treated with FSH or eCG. Llamas were assigned randomly to two groups (n = 16 per group) and treated, at the time of ovarian follicular wave emergence, with either: 1) 25 mg of FSH im, twice daily for 4 d; or 2) 1000 IU of eCG as a single i.m. dose. The start of gonadotropin treatment was considered Day 0. Both groups were given 5 mg of Armour Standard LH im on Day 6, and COC were collected by follicle aspiration on Day 7. Expanded COC collected from FSH- (n = 157) and eCG-treated llamas (n = 151) were fertilized in vitro using epididymal sperm, and presumptive zygotes were in vitro cultured in SOF medium for 8 d. The FSH and eCG treatment groups did not differ with respect to: the number of follicles ≥7 mm (16.0 ± 2.7 vs 14.0 ± 1.9, respectively; P = 0.5); the number of COC collected (11.5 ± 1.9 vs 9.7 ± 1.2; P = 0.4); the number of expanded COC (9.8 ± 1.4 vs 9.4 ± 1.2; P = 0.8); or the percentage of presumptive zygotes which developed into 2 to 8 cell stage embryos (65.3 vs 63.1), morulas (46.2 vs 42.5), or blastocysts (23.1 vs 20.5; P > 0.05). In conclusion, FSH and eCG treatments were equally effective for recovery of a high number of expanded COC which were used directly for in vitro fertilization. Furthermore, rate of embryo development was not significantly affected by the gonadotropin treatment used.

Mitochondrial distribution and meiotic progression in canine oocytes during in vivo and in vitro maturation.

The objective was to evaluate mitochondrial distribution, and its relationship to meiotic development, in canine oocytes during in vitro maturation (IVM) at 48, 72, and 96 h, compared to those that were non-matured or in vivo matured (ovulated). The distribution of active mitochondria during canine oocyte maturation (both in vitro and in vivo) was assessed with fluorescence and confocal microscopy using MitoTracker Red (MT-Red), whereas chromatin configuration was concurrently evaluated with fluorescence microscopy and DAPI staining. During IVM, oocytes exhibited changes in mitochondrial organization, ranging from a fine uniform distribution (pattern A), to increasing clustering spread throughout the cytoplasm (pattern B), and to a more perinuclear and cortical distribution (pattern C). Pattern A was mainly observed in germinal vesicle (GV) oocytes (96.4%), primarily in the non-matured group (P < 0.05). Pattern B was seen in all ovulated oocytes which were fully in second metaphase (MII), whereas in IVM oocytes, ∼64% were pattern B, irrespective of duration of culture or stage of nuclear development (P > 0.05). Pattern C was detected in a minor percentage (P < 0.05) of oocytes (mainly those in first metaphase, MI) cultured for 72 or 96 h. In vitro matured oocytes had a minor percentage of pattern B (P < 0.05) and smaller mitochondrial clusters in IVM oocytes than ovulated oocytes, reaching only 4, 11, and 17% of MII at 48, 72, and 96 h, respectively. Thus, although IVM canine oocytes rearranged mitochondria, which could be related to nuclear maturation, they did not consistently proceed to MII, perhaps due to incomplete IVM, confirming that oocytes matured in vitro were less likely to be competent than those matured in vivo.

hCG-induced ovulation in thoroughbred mares does not affect corpus luteum development and function during early pregnancy.

UNLABELLED: Our aim was to compare Corpus luteum (CL) development and blood plasma concentration of progesterone ([P4]) in thoroughbred mares after spontaneous ( CONTROL: C) or human chorionic gonadotrophin (hCG)-induced ovulation. Lactating mares (C=12; hCG=21) were daily teased and mated during second oestrus post-partum. Treated mares received 2500 IU hCG i.v. at first day of behavioural oestrus when dominant follicular size was >35, or=45 mm. Dominant follicle before ovulation, CL and gestational sac were measured by ultrasound and [P4] by radioimmunoassay (RIA). Blood sampling and ultrasound CL exams were done at days 1, 2, 3, 4, 8, 12, 16, 20, 25, 30, 35, 40, 45, 60 and 90 after ovulation and gestational sac from day 12 after ovulation in pregnant (P) mares; non-pregnant (NP) were followed until oestrus returned. Data analyses considered four subgroups: hCG-P, hCG-NP, C-P and C-NP. Preovulatory follicular size was smaller in hCG mares than in C: 39.2+/-2.7 mm vs 51.0+/-1.8 mm (p<0.0001). All hCG mares ovulated 24-48 h after treatment and presented similar oestrus duration as controls. C. luteum size in P mares showed the same pattern of development through days 4-35, presenting erratic differences during initial establishment. Thus, on days 1 and 3, CL was smaller in hCG-P (p<0.05); while in hCG-NP, CL size was greater than in C-NP on day three (p=0.03). Corpus luteum size remained stable until day 90 in hCG-P mares, while in C-P a transient and apparently not functional increase was detected on days 40 and 45 (p<0.05) and the decrease from day 60 onwards, made this difference to disappear. No differences were observed in [P4] pattern between P, or between NP subgroups, respectively. So, hCG-induced ovulation does not affect CL development, neither [P4] during early pregnancy. One cycle pregnancy rate tended to be lower in hCG mares while season pregnancy rates were similar to controls.

Cytologic, hormonal, and ultrasonographic correlates of the menstrual cycle of the New World monkey Cebus apella.

Few reports on the reproductive physiology of Cebus apella have been published. In this study we characterized menstrual cycle events by means of vaginal cytology, ultrasonography (US), and hormonal measurements in serum during three consecutive cycles in 10 females, and assessed the probability that ovulation would occur in the same ovary in consecutive cycles in 18 females. The lengths and phases of the cycles were determined according to vaginal cytology. Taking the first day of endometrial bleeding as the first day of the cycle, the mean cycle length +/- SEM was 19.5+/-0.4 days, with follicular and luteal phases lasting 8.2+/-0.2 and 11.3+/-0.4 days, respectively. The follicular phase included menstruation and the periovulatory period, which was characterized by the presence of a large number of superficial eosinophilic cells in the vaginal smear. The myometrium, endometrium, and ovaries were clearly distinguished on US examination. During each menstrual cycle a single follicle was recruited at random from either ovary. The follicle grew from 3 mm to a maximum diameter of 8-9 mm over the course of 8 days, in association with increasing estradiol (E(2)) serum levels (from 489+/-41 to 1600+/-92 pmol/L). At ovulation, the mean diameter of the dominant follicle usually decreased by >20%, 1 day after the maximum E(2) level was reached. Ovulation was associated with an abrupt fall in E(2), a decreased number of eosinophilic cells, the presence of leukocytes and intermediate cells in the vaginal smear, and a progressive increase in progesterone (P) levels that reached a maximum of 892+/-65 nmol/L on days 3-6 of the luteal phase. The menstrual cycle of Cebus apella differs in several temporal and quantitative aspects from that in humans and Old World primates, but it exhibits the same correlations between ovarian endocrine and morphologic parameters.

Post-coital administration of levonorgestrel does not interfere with post-fertilization events in the new-world monkey Cebus apella.

BACKGROUND: Experimental evidence to disprove the belief that emergency contraception with levonorgestrel (LNG) prevents pregnancy by interfering with post-fertilization events is lacking. Here we determined the effect of post-coital and pre-ovulatory administration of LNG on fertility and ovulation, respectively, in the Cebus monkey. METHODS: To determine the effect on fertility, LNG 0.75 mg or vehicle were administered orally or s.c. once or twice within the first 24 h after mating occurring very close to the time of ovulation. Females that became pregnant were aborted with mifepristone and re-entered the study after a resting cycle until each of 12 females had contributed, in a randomized order, two LNG and two vehicle-treated cycles. To determine the effect on ovulation, LNG 0.75 mg or vehicle were injected twice coinciding with follicles smaller or larger than 5 mm in diameter. Six females contributed five treated cycles each. RESULTS: The pregnancy rate was identical in vehicle- and LNG-treated cycles. LNG inhibited or delayed ovulation only when treatment coincided with a follicle <5 mm diameter. CONCLUSION: In Cebus monkeys, LNG can inhibit or delay ovulation but, once fertilization has taken place, it cannot prevent the establishment of pregnancy. These findings do not support the hypothesis that emergency contraception with LNG prevents pregnancy by interfering with post-fertilization events.

Prediction of gestational age by ultrasonic fetometry in llamas (Lama glama) and alpacas (Lama pacos).

Fetal biparietal diameter (BPD) and thorax height (TH) were measured by ultrasound during intrauterine growth in pregnant llamas (Lama glama) and alpacas (Lama pacos). The goal was to establish representative curves that allows estimation of gestational age (GA) from real-time ultrasonic measurements of these fetal structures at any stage of gestation. Llamas and alpacas were mated under controlled conditions. Ultrasound exams were conducted to determine pregnancy status 1 month later. Measurements of fetal BPD and TH were conducted from the second month of pregnancy until term. Observation and assessment of fetal TH was difficult during the last 3 months of pregnancy, specially in llamas. Regression curves were calculated from the data as a function of GA, with the best fit represented by the following equations: llama GA=(BPD-0.002399)43.02293,r=0.98,P<0.001; llama GA=(TH-0.07137)46.94485, r=0.95,P<0.001; alpaca GA=(BPD-0.11376)47.23287, r=0.98,P<0.001; alpaca GA=(TH-0.36436)52.87663, r=0.96,P<0.001, where GA was measured in days and BPD and TH in centimeters. Results indicate that ultrasonic measurement of these fetal biometric variables constitute a valuable tool to estimate GA at any stage of pregnancy in these domestic South American camelids.

Plasma progesterone in alpaca (Lama pacos) during pregnancy, parturition and early postpartum.

Plasma progesterone concentration during pregnancy and the early postpartum period was measured by radioimmunoassay in five alpacas (Lama pacos), of the Huacaya breed, whose pregnancy length had a mean of 344.8 +/- 4.4 days. Concentration of progesterone increased (P < 0.001) from low premating values (0.11 +/- 0.06 nmol/l) to greater values at 30 days of pregnancy (8.05 +/- 1.13 nmol/l) and remained high up to 2 months of pregnancy. A slight transitory decline was observed between 3 to 7 months. Concentration of plasma progesterone dropped markedly during the 72 h before parturition especially, at the day of parturition.

Diurnal changes in light intensity inside the pregnant uterus in sheep.

Penetration of light into the pregnant sheep uterus was studied in 9 ewes, gestational ages 40 to 142 days (term 147 days). Light sensors were placed inside the pregnant horn and over the flank skin overlying the position of the uterine horn. To perform the experiments, the ewes were placed in a study cage outdoors and light sensors were connected to a luxometer. Simultaneous measurements were obtained from the intrauterine and the external sensors in the shade at noon. The amount of light detected inside the uterus increased with gestational age from two lux at 40 days to 51.1 +/- 16.5 (n = 5) lux at 142 days (0.2 and 5.4% of the amount of light detected at the maternal flank). Measurements through the 24 h were done in four pregnant ewes at 142 days gestation under natural photoperiod (13.5 light:10.5 dark). In these experiments, the intensity of intrauterine light changed through the 24 h, reflecting the changes in the intensity of the sunlight. Maximal intrauterine light values were observed at noon, corresponding to 4.7% of incident light. Small but detectable values were observed at 0900 and 1800 h. Our data show that, at mid gestation, light reaches the pregnant uterus and that, at late gestation, changes in intrauterine lighting throughout the 24 h are present reflecting the changes in external daylight. Therefore the sheep fetus is exposed to light-dark transitions at dawn and dusk, and to a peak of light at midday.

Study of prenatal growth in the capuchin monkey (Cebus apella) by ultrasound.

Capuchin monkey (Cebus apella) is a new world primate that in recent years has become important in biomedical research. The purpose of this study was to establish and correlate normal fetal growth parameters with gestational age in capuchin monkeys. In seven pregnant animals serial ultrasonic assessment of gestational sac (GS), embryo/fetal greatest length (GL), biparietal diameter (BPD), thorax height (TH), and femur length (FL) were performed. Identification of the GS was possible on day 23+/-2.8 (X +/- SE). The embryo and its heartbeat was detected on day 32.7+/-2.8, the GL being measurable thereafter. By day 45.4+/-1.4 BPD and TH were measurable. FL could only be measured from day 70.6+/-2.1. Predictive regression equations of gestational age (GA) were modeled with data obtained. In addition, preliminary data of fetal heart rate showed a decrease in frequency with advancing gestation.

Early pregnancy diagnosis in alpaca (Lama pacos) and llama (Lama glama) by ultrasound.

An ultrasonography study of early pregnancy diagnosis was carried out in 19 alpacas and 12 llamas, after controlled matings. The aim was to determine the earliest gestational age at which pregnancy diagnosis by transrectal ultrasonography could be achieved, and to generate an empirical formula for gestational sac diameter (GSD) growth as a function of gestational age (GA), allowing an estimate of GA during the first month of pregnancy. We found that pregnancy diagnosis may be carried out as early as 9 days after mating in alpacas and 7 days in llamas. This diagnosis was found to be accurate at 23 days in alpacas and 34 days in llamas. The empirical relations that best describe the relationship between GSD and GA were GA = logGSD + 1.2339/0.0585 r = 0.85; P < 0.001 in alpacas, and GA = logGSD + 1.2649/0.0546 r = 0.77, P < 0.001 in llamas, where GA is measured in days and GSD in centimeters. Our results also indicate that ultrasonography is a reliable technique for early pregnancy diagnosis. Furthermore, the empirical formulae reliably make it possible to estimate GA from GSD during the first month of pregnancy and their use might improve the efficiency of camelid breeders.

Effect of constant light on fetal and maternal prolactin rhythms in sheep.

A 24-h rhythm of plasma PRL is present in fetal sheep. This rhythm is synchronized to an environmental clue (zeitgeber). We determined whether the light-dark cycle (L:D) is a zeitgeber for the fetal PRL rhythm and, if so, whether the mother might convey this zeitgeber to the fetus. We kept nine ewes (twin pregnancies) in constant light (L:L) and five ewes (singleton) in 14:10 L:D from 110 days gestation. Fetuses and mothers were catheterized at 119 days gestation. Blood samples were taken hourly for 24 h after 16 days under L:L or L:D. A mean 24-h rhythm of PRL was found (by RIA) in fetuses under L:D, but not in those under L:L. However, fetuses under L:L showed individual 24-h PRL rhythms (cosinor analysis) whose acrophases were distributed around the clock. Nonsynchronized rhythms persisted after 23 and 30 days of L:L. Acrophases of PRL rhythms within a set of twins were closer than those between sets, suggesting that twins were responding to a common signal. These findings indicate that the L:D cycle is a zeitgeber for the PRL rhythm in fetal sheep and suggest that the mother might convey the zeitgeber.

Day-night changes in c-fos expression in the fetal sheep suprachiasmatic nucleus at late gestation.

The suprachiasmatic nucleus (SCN) is a circadian oscillator in mammals and shows day-night changes in metabolic activity. To investigate whether the fetal sheep SCN behaves as a circadian oscillator, day-night changes in c-fos expression, a marker of neuronal activity, were measured. Eight fetal sheep were sacrificed at 135 days gestation--four at day-time (1200 hours) and four at night-time (2400 hours). Fetal brains were fixed, removed and cut in 40-microns serial coronal sections. Alternate sections were incubated with anti-Fos antibody (1:500) and Fos expression was revealed with extra-avidin-peroxidase and 3,3'-diaminobenzidine or stained with cresyl violet. The number of Fos-immunoreactive (Fos-ir) neurons per mm2 in the rostral, intermediate and caudal regions of the fetal sheep SCN was counted. Fetuses sacrificed in the day-time showed a higher number of Fos-ir neurons per mm2 (mean +/- s.e.; 516.7 +/- 60.1) in the three regions of the SCN than fetuses sacrificed at night-time (140.5 +/- 21.8). In addition, at night-time Fos-ir neurons were mainly localized to the ventrolateral area of the SCN. These findings demonstrate day-night changes in molecular activity consistent with the presence of a circadian oscillator in the fetal sheep SCN.

Prenatal development of the retinohypothalamic pathway and the suprachiasmatic nucleus in the sheep.

Circadian rhythms are present during fetal life in several mammalian species. To characterize the ontogeny of the neural mechanisms that account for circadian rhythmicity in a precocious species, we studied the prenatal development of the retinohypothalamic pathway in lambs (gestation period of 147 days), using horseradish peroxidase and wheat germ agglutinin as anterograde tracers. The suprachiasmatic nucleus was present as early as embryonic day 52 (E52). After E58, the suprachiasmatic nucleus reached its full number of neurons, estimated by the disector method in about 160,000 cells per nucleus at E62. The retinohypothalamic axons invaded the suprachiasmatic nucleus from E58, while neuroblasts were still migrating to the nucleus. At E62, there was a strong retinal projection that evolved until E121, when the retinal afferents established their definitive pattern of distribution in the ventral and central regions of the suprachiasmatic nucleus, and adjacent hypothalamic structures. The development of the retinohypothalamic pathway was delayed by about a week relative to the innervation of other subcortical visual centers. The present findings demonstrated an early prenatal development of the visual pathways in lambs, including the retinohypothalamic pathway, suggesting that the mechanisms for the visual entrainment of circadian rhythms in lambs may be functioning several weeks before birth.

The retino-hypothalamic tract is involved in prolactin regulation in fetal sheep.

To investigate the role of the retino-hypothalamic tract on fetal prolactin regulation, we examined the effect of ocular enucleation on fetal plasma prolactin. Eleven fetuses of Suffolk ewes were chronically catheterized during fall, and six of them were subjected to bilateral ocular enucleation. All ewes were kept at 12h:12h light:dark cycle (lights on at 0800 and off at 2000). The experiments were performed 5-9 days after surgery (GA control fetuses 125 +/- 1.5, optical enucleation 121.3 +/- 1.5 days). Blood samples were taken from fetuses hourly around the clock, and plasma prolactin and cortisol were measured by radioimmunoassay (RIA). Luteinizing hormone (LH) and Growth hormone (GH) were measured in pooled plasma samples from control and enucleated fetuses by RIA. Average plasma prolactin was 5-fold lower in enucleated than in control fetuses (9.6 +/- 0.5 and 54.2 +/- 3.3 ng/ml, SEM; P < 0.005). Both control and enucleated fetuses presented circadian rhythm of prolactin with acrophase between 1400 and 1830 h. An enucleated fetus was tested for response of prolactin to TRH. Prolactin increased as described in the literature. There was no change in plasma concentration of cortisol, LH or GH after ocular enucleation. Our data indicate that the optical pathway participates in prolactin regulation in the fetal sheep.