Scanning electron microscopy of the equine oviduct and observations on ciliary currents in vitro at day 2 after ovulation

There are considerable differences between mammalian species in the distribution and activity of ciliated cells within the oviduct, and limited information is available concerning either the distribution or activity of cilia within the equine oviduct. Patterns of ciliary activity were characterized in the ampulla and isthmus of oviducts recovered at 2 d after ovulation from 10 mares, and scanning electron microscopy was used to examine regional differences in the distribution of cilia in oviducts from 3 of these mares. Based upon the motility of 15 microm latex microspheres, ciliary activity was significantly (P < 0.001) greater in the ampullar oviduct compared with that of the isthmic oviduct. The direction of ciliary beat was consistently toward the uterus in all regions of the oviduct. Scanning electron microscopy revealed ciliated and secretory cells in both regions of the oviduct at 2 d after ovulation, with no apparent differences in the proportion of ciliated versus secretory cells.     

Attachment and release of spermatozoa from the caudal isthmus of the hamster oviduct

Female hamsters were mated shortly after the onset of oestrus. At 3 or 6 h after mating, the right oviduct was flushed in situ with 30, 90 or 180 microliters medium to remove spermatozoa from the lumen, leaving only those firmly attached to the isthmic mucosa of the oviduct. When eggs were recovered from oviducts at 20 h after flushing the majority were fertilized, indicating that the spermatozoa that were firmly attached to the mucosa were capable of detaching and ascending to the ampulla to fertilize eggs. Neither the time of flushing nor the volume of flushing medium had a significant effect on the percentage of spermatozoa that remained in the isthmus after flushing. These results suggest that there is no change in the surface of the oviduct mucosa that causes the release of spermatozoa from the caudal isthmus near the time of ovulation. When incapacitated spermatozoa were introduced into the oviduct, many of them attached to oviductal mucosa, while capacitated spermatozoa did not. This indicates that it is a change in the sperm surface, rather than the mucosal surface, that causes the release of spermatozoa, i.e. spermatozoa remain attached to the isthmic mucosa until they become capacitated and then detach and migrate to the ampulla to fertilize the eggs.     

Variations in rabbit oviduct microvascular architecture after ovulation induced by hCG

Rabbits were induced to ovulate by injection with hCG and vascular corrosion casts of the oviducts were examined by scanning electron microscopy after 24 and 48 h, when the ova would be expected to be at the ampullary-isthmic junction, and traversing the isthmus respectively. At 24 h there was dilatation of the isthmic subserosal venous plexus. It is suggested that venous distension in the isthmic subserosal venous plexus, due to raised venous pressure or to reduced venous wall tone, may occlude the isthmic lumen to ova, and thus explain the known pre-isthmic delay in ovum transport. By 48 h after hCG, distension was no longer evident, consistent with the possibility of ovum transport.     

Effects of transferring in vitro-cultured rabbit embryos to recipient oviducts on mucin coat deposition, implantation and development

A mucin coat is deposited on rabbit embryos during passage through the oviduct; rabbit blastocysts cultured from the 1-cell stage in vitro have no mucin coat. When cultured blastocysts are transferred to recipients, the lack of mucin coat might account in part for subsequent failure of pregnancy. We have investigated the possibility that mucin coat deposition is induced following transfer of in vitro 72 h-cultured blastocysts to oviducts of asynchronous or synchronous recipients. One-cell embryos were collected by flushing oviducts 19-20 h post-coitus and were cultured in vitro for 72 h until they reached the blastocyst stage. The blastocysts were transferred to the oviducts of recipients that were synchronized either with the donors (synchronous) or 1 day later than the donors (asynchronous). They were recovered after 24-48 h and the mucin coat thickness and embryo degeneration rate were measured. The degeneration rate of blastocysts recovered from uteri of synchronous recipients was higher than that from asynchronous recipients (72.2% vs 40.0%). The mucin coats around embryos recovered from oviducts of asynchronous recipients after 48 h were thicker than those from synchronous recipients. More asynchronous recipients were pregnant and gave birth to more pups than synchronous recipients. These results indicate that the oviducts of asynchronous recipients secreted more mucin around the transferred embryos, causing higher rates of implantation of the in vitro-cultured blastocysts.     

Transfer of cultured rabbit embryos

Embryo transfer experiments were carried out to study the developmental capacity of cultured rabbit embryos when transferred to recipients of variable postovulatory maturity. Rabbit embryos were flushed from the oviduct at 26 hours postcoitum (pc) and cultured in a modified Ham's F-10 medium supplemented with bovine serum albumin (BSA) for a period of 70 hours. At 96 hours pc the cultured embryos, which ranged from the early morula to the expanding blastocyst stage, were transferred to pseudopregnant recipients mated to vasectomized males 36 to 96 hours prior to the transfer procedure. Greatest embryo survival occurred when transfers were made to either the oviducts or uterine horns of recipients at 48 hours pc. Intermediate results for both implantation rates and number of young born were obtained with recipients at 36, 60, 72, and 84 hours pc. Transferred embryos consistently failed to survive the uterine environment of recipients 96 hours pc at transfer although this group was synchronous with embryonic chronological age. Oviductal transfers were generally more successful than uterine transfers. Markedly higher rates of embryo survival resulted from embryos that were collected 60 and 72 hours pc and transferred directly to synchronous recipients without an interim period of culture. Dissimilarity of development for in vivo grown rabbit embryos and those cultured in synthetic medium is demonstrated.     

Distribution and number of spermatozoa in the oviduct of the golden hamster after natural mating and artificial insemination

A group of female hamsters was mated with males of proven fertility either several hours before or during ovulation. Another group of females was artificially inseminated several hours before ovulation. Females were killed at various times after the onset of mating or artificial insemination, oviducts were fixed and sectioned serially, and spermatozoa were counted individually as to their location in the oviduct. Regardless of the type or time of insemination, the vast majority of spermatozoa that entered the oviduct remained in the lower segments of the isthmus (the intramural and caudal isthmus) without ascending to the ampulla. The lower segments of the oviduct, particularly the caudal isthmus, appeared to be acting as a "sieve" and/or "sperm reservoir." In females mated or artificially inseminated prior to ovulation, virtually no spermatozoa reached the cephalic isthmus or ampulla until the commencement of ovulation. Although a few spermatozoa reached the ampulla by 1 h after the onset of mating, they were the exception rather than the rule. When females were mated during ovulation, spermatozoa spent a minimum of about 3 h in the caudal isthmus before ascending to the ampulla. The number of spermatozoa that entered the oviduct after artificial insemination was considerably lower than in naturally mated animals, but this low number was apparently large enough to ensure complete fertilization.     

The movement characteristics of rabbit spermatozoa before and after activation

Spermatozoa were flushed with mineral oil from the lower isthmus of the rabbit oviduct at four hours postcoitus (pc) and 11 hours pc. Videotapes were made of sperm behavior in the native isthmic fluid and after dilution of the fluid with culture medium. The tapes showed that, initially, spermatozoa in the native isthmic fluid were virtually immotile, but immediately resumed movement on contact with the culture medium. Isthmic sperm motility then evolved over a five- to 10-minute interval into the characteristic biphasic pattern of activated movement. Cine films of isthmic spermatozoa taken with a high-speed camera were analyzed to determine flagellar beat frequency, maximum flagellar curvature, and swimming velocity. Progressiveness ratios and hydrodynamic power outputs were then calculated for individual spermatozoa. Two phases of activated sperm movement, a whiplash phase and a progressive phase, were identified and characterized. The power output of activated spermatozoa increased twentyfold in comparison with the preactivated state. The power output of activated spermatozoa did not differ between the two phases of activated movement.     

Effect of in vitro and in vivo culture on embryo development from prepubertal goat IVM-IVF oocytes

The aim of this study was to analyze different culture systems on embryo development of prepubertal goat oocytes. We compare (i) the effect of the age of donor (goat) of oocytes on in vitro maturation, fertilization and subsequent embryo development, (ii) the effect of the origin of oviduct cells from coculture of prepubertal goat embryo development, and (iii) the effect of in vivo culture in rabbit oviducts for 1, 2 and 3 days on the development of prepubertal goat embryos produced in vitro. In Experiment 1, at 24 h post-insemination (hpi), oocytes from adult goats were allocated in TCM199 with oviduct cells from adult goats, and oocytes from prepubertal goats were randomly placed in drops with oviduct epithelial cells from adult (aOEC) or prepubertal (pOEC) goats. Cleavage rate and embryo development were evaluated at 48 hpi and after 7 days coculture, respectively. In Experiment 2, at 24 hpi, prepubertal oocytes were allocated in TCM 199 with pOEC. At 40-42 hpi, a group of embryos remained in the coculture (control group), and the rest were transferred to rabbit oviducts (three rabbits for replicate) for culturing in vivo for 24, 48 and 72 h. After these in vivo cultures, embryos were recovered, evaluated and placed in TCM199 with pOEC until Day 8 post-insemination. The maturation, fertilization and blastocyst rates did not differ significantly between oocytes obtained from adult and prepubertal goats. The percentage of blastocysts obtained from prepubertal goat embryos cocultured with aOEC or pOEC was also similar (12.1% versus 12.2%). The transfer of prepubertal goat embryos to rabbit oviducts for 1, 2 and 3 days did not improve the blastocyst rate compared to the control group (9.7, 10.9, 4.1 and 11.5%, respectively). In conclusion, in our conditions, there were no significant differences in embryo development between oocytes obtained from prepubertal and adult goats, and the embryo development from prepubertal goat oocytes were similar in the different culture systems compared.     

Capacitation status of hamster spermatozoa in the oviduct at various times after mating

Female hamsters were mated shortly after the onset of oestrus or immediately after ovulation. At various times after mating, spermatozoa were flushed from the isthmus of the oviduct using a modified Tyrode's medium supplemented with 20% hamster serum. Cumulus oophorus-free eggs were introduced into the suspensions of isthmic spermatozoa. Some eggs were removed every 30 min and examined for evidence of fertilization. For females mated shortly after the onset of oestrus, spermatozoa recovered from the oviducts 8 h after mating (about 1.5 h after ovulation) could penetrate eggs within 30 min and were considered fully capacitated. When spermatozoa were recovered at earlier times (1, 2, 4 and 6 h after mating) they required additional time (2, 1.5, 1 and 1 h respectively) in vitro before penetrating eggs. Therefore, when mating occurs shortly after the onset of oestrus, spermatozoa in the oviduct do not appear to become fully capacitated until about the time of ovulation. For females mated immediately after ovulation, spermatozoa recovered from the oviducts at 4 h after mating could penetrate eggs within 30 min. Spermatozoa recovered at 1 and 3 h after mating required 2 and 1 h respectively in vitro before penetrating eggs. These results suggest that sperm capacitation proceeds at a faster rate when mating occurs after ovulation.     

Motility characteristics of sperm from the uterus and oviducts of female mice after mating to congenic males differing in sperm transport and fertility

Motile sperm were videotaped after removal from the uterus and isthmus of the oviducts of female mice 1 h after mating with congenic males carrying none, one, or two t complexes. Males carrying one t complex (tw32/+) are fertile, and sperm carrying the t complex have an advantage in fertilization; males carrying two complexes (tw32/t0) are sterile. For each sperm, 2 sec of movement of the head-midpiece junction were traced from the videotape. For each tracing, five motility parameters were used: curvilinear velocity (Vc), and index of the sperm's mean swimming speed; coefficient of variation of move length (CVML), an index of speed constancy; progressiveness ratio (PR), an index of all deviation of the sperm's movement from a straight line; linear index (LI), an index of the straightness of the sperm's trajectory; and curvilinear progressiveness ratio (PRc), an index of the degree of lateral oscillation about that trajectory. Uterine sperm from fertile males were progressive, with straight trajectories and little lateral oscillation. There were no consistent differences in any motility parameter between uterine sperm from tw32/+ and congenic +/+ males. Uterine sperm from sterile tw32/t0 males were extremely slow and showed very little progressive movement, which could explain their lack of transport to the oviduct. For all fertile males, isthmic oviductal sperm differed significantly from uterine sperm in every motility parameter except Vc: isthmic sperm were less consistent in swimming speed, and less progressive with less straight trajectories and more lateral movement. One or more of these motility characteristics may be related to hyperactivation. A large proportion of isthmic sperm from tw32/+ males had nonlinear trajectories (LI less than .50); these nonlinear sperm were faster than nonlinear isthmic sperm from congenic +/+ males. These motility characteristics of isthmic sperm from tw32/+ males may be related to hyperactivation, or to their previously observed abnormal transport within the oviduct.     

Prostaglandin E2-specific binding to the equine oviduct.

Prostaglandin E2 (PGE2) bound specifically (P less than 0.001) to ampullary and isthmic tissue on Day 2 and Day 5 after ovulation. No significant differences (P greater than 0.8) were detected between Day 2 and Day 5 in the specific binding of ampullary or isthmic tissue. Significantly more (P less than 0.05) PGE2 bound specifically to ampullary versus isthmic tissue on both days. Detection of PGE2-specific binding in the oviductal isthmus on Day 2 and Day 5 indicates that the oviduct is responsive to PGE2 when it is capable of transporting equine embryos.     

Components of oviduct physiology in eutherian mammals

Recalling the evolutionary sequence of development first of gonad and subsequently of oviducts, ovarian endocrine regulation of all known components of oviduct physiology is reviewed. Ovaries not only influence oviducts via the systemic blood circulation, but also locally by counter‐current transfer of relatively high concentrations of steroid hormones and prostaglandins between the ovarian vein and oviduct branch of the ovarian artery. The efficiency and impact of such counter‐current transfer is greatest around the time of ovulation, the transfer process receiving further inputs from hormones present in peritoneal fluid. Classical oviduct physiology is summarised, and the potential molecular consequences of temperature gradients within the duct lumen examined. At ovulation, an oocyte‐cumulus complex is displaced in minutes from the follicular surface to the site of fertilisation at the ampullary‐isthmic junction of the oviduct. This rapid initial phase is contrasted with the subsequent slow progression of embryos to the uterus in days, still encompassed within a zona pellucida. Regarding transport of spermatozoa, the formation of a pre‐ovulatory reservoir in the caudal portion of the oviduct isthmus is noted, with suppression of motility and sperm‐head binding to epithelial organelles acting to maintain fertilising ability. Completion of capacitation is prompted shortly before ovulation, predominantly by Ca²⁺ influx into bound spermatozoa. A controlled release of spermatozoa coupled with their hyperactivation results in initial sperm:egg ratios at the site of fertilisation close to unity, thereby avoiding the pathological condition of polyspermy. Both the oviduct milieu and embryonic development are influenced by paracrine activity of follicular granulosa cells released at ovulation and remaining in suspension in the vicinity of the oocyte or embryo. These cells may amplify early pregnancy signals from a zygote to the endosalpinx. Beneficial effects of the oviduct on domestic animal embryos are contrasted with anomalies arising as a consequence of in vitro culture. Primate embryos do not require exposure to an oviduct for normal development, perhaps due to overlapping compositions of endosalpingeal and endometrial secretions. Additionally, primate endometrial secretions may be modified by viable gametes or an embryo in the presence of a cumulus cell suspension.     

Sperm transport and motility in the mouse oviduct: observations in situ

Sperm transport and motility were studied through the transparent walls of the mouse oviduct by direct microscopic observation and videomicrography. Observations were made on excised female tracts 1-2 h post-coitus (pc) and 1-2 h before and after the approximate time of ovulation. Motile sperm were seen at the uterine entrance to the uterotubal junction (UTJ) in all females at 1-2 h pc, but in fewer females at later times. The intramural UTJ was usually constricted and held few sperm. The extramural UTJ and adjacent lower isthmus contained many motile sperm at 1-2 h pc. Apparently, the column of sperm moved upwards because in some females, sperm were found in the upper isthmus and not in the UTJ at the later time points. Few sperm were seen in the ampulla in the periovulatory period, and none at 1-2 h pc. There appeared to be two mechanisms retaining sperm in the lower oviduct: immobilization and adherence to the epithelium. Columns of immotile sperm were seen in the lower isthmus of some females. Motile sperm usually appeared to adhere by their heads to the oviductal epithelium, only occasionally breaking free to move vigorously about the lumen.     

Hyperactivated sperm progress in the mouse oviduct.

Sperm from naturally mated mice were observed and videotaped moving within mouse oviducts. The typical pattern of sperm progress involved intermittently breaking free and swimming a short distance, then reattaching to the epithelium. The proportion of sperm that swam freely (were not attached to the epithelium) was calculated and analyzed for effects of oviductal region, ovulation status, and sperm location relative to the lumen. A significantly higher proportion of sperm were free in the ampulla than in the isthmus (26.3% +/- 0.8% vs. 11.8% +/- 1.0%; p less than 0.0001) and in post-ovulatory than pre-ovulatory (16.2% +/- 2.0% vs. 10.6% +/- 1.6%; p less than 0.05) oviducts. Flagellar curvature ratio values showed that free sperm (0.716 +/- 0.024) had more sharply curved tails than stuck sperm (0.782 +/- 0.013). While this difference is significant (p = 0.01), the effect of attachment status interacted significantly (p less than 0.05) with the oviductal region such that there was a greater difference in the isthmus than in the ampulla. Only sperm using the more curved tail beats of hyperactivation were seen to break free from the epithelium and to progress along the oviduct. These results indicate that hyperactivation plays a role in moving sperm out of the isthmic reservoir and to the site of fertilization.     

Detection of early pregnancy factor (EPF) using the rabbit ovary and oviduct perfused in vitro

Two peaks of rabbit serum EPF activity were seen over the course of pregnancy. Rabbit ovaries with or without attached oviducts were perfused in vitro for 5 h beginning 12, 16, 24, 48, 72 and 120 h after mating. Perfused isolated ovaries did not produce EPF in vitro, but significant EPF activity was detected in the perfusate of the ovary together with oviduct. Pseudopregnant animals and those rabbits that did not ovulate exhibited no perfusate EPF activity. Perfusate EPF activity was highest at the time embryos were at the pronuclear stage and continued through the morula stage. Although the location of embryos at 72 h after mating varied between oviduct and uterus, EPF activity was maintained over the perfusion period. The results suggest that EPF release occurs within 3 h of fertilization and that the presence of the preblastocyst embryo is crucial for EPF release.     

Effect of nitric oxide synthase inhibitors on ovum transport and oviductal smooth muscle activity in the rat oviduct

The effect of the inhibition of nitric oxide synthase (NOS) on ovum transport and oviductal motility in rats was investigated. Three different NOS inhibitors were injected into the ovarian bursa at oestrus or day 3 of pregnancy. Oviducts and uteri were flushed 24 h later and the presence of ova was recorded. In oestrous and pregnant rats, treatment resulted in accelerated egg transport, as shown by a decrease in the number of ova present in the oviducts. In cyclic rats, intrabursal injection of 1 mg kg-1 of either N-monomethyl-L-arginine (L-NMMA) or N omega nitro-L-arginine methyl ester (L-NAME) elicited a 30% reduction in the number of ova present in the oviducts, whereas in pregnant animals, the same dose of L-NMMA produced a reduction of 40%. Simultaneous administration of the NO donor spermine NONOate (5 mg kg-1) completely reversed the effect of L-NMMA. Tubal motility was assessed by microsphere displacement analysis within the oviduct. Surrogate ova were transferred to the oviductal lumen at oestrus and 24 h later the effect of intraoviductal injection of 1 microgram L-NMMA or vehicle was assessed. The microspheres in the isthmus showed an oscillating motion, and periods in which movement was not detectable. However, L-NMMA treatment produced a 3.6-fold increase in the maximum instant velocities and a significant reduction in the resting periods of the microspheres compared with the control group (P < 0.001). These results provide evidence that NO inhibition increases tubal motility that results in accelerated ovum transport, and indicate that NO could act as a paracrine signal between different layers of the oviductal wall, providing a role for endogenous NO in regulation of tubal function.     

The polyol pathway in the bovine oviduct

The oviducts likely provide optimized micro‐environments for the final maturation of gametes, fertilization, and early embryo development. Hexoses, including glucose, fructose, and sorbitol, are involved in these critical reproductive events. Monosaccharide production is controlled, in part, by the polyol pathway and requires two enzymes: an aldose reductase (AR) that reduces glucose into sorbitol, followed by its oxidation into fructose by sorbitol dehydrogenase (SDH). We analyzed the expression of AR and SDH in the isthmus and ampulla of the bovine oviduct at the proliferative, mid‐luteal, and late‐luteal phases of the estrous cycle by quantitative PCR and immunoblots. Immunochemistry and an assay of SDH activity were also performed. The quantity of hexoses in whole sections of isthmus and ampulla were determined by liquid chromatography coupled to mass spectrometry. In sum, AR expression was restricted to the isthmus, while SDH was mostly expressed in the isthmic–ampullary junction and the ampulla, specifically concentrated in the luminal epithelium of the oviduct. The estrous cycle had no impact on protein expression of AR and SDH. Instead, the levels of AR and SDH expression were associated with higher ratios of sorbitol to fructose in the isthmus (1.6) than in the ampulla (4.1; P = 0.005). These results are discussed in light of physiological events occurring in the oviduct. Mol. Reprod. Dev. 79: 603–612, 2012. © 2012 Wiley Periodicals, Inc.     

Arylsulphatases in the rabbit oviduct: postovulatory changes tested by histochemical and biochemical procedures

Summary A histochemical and biochemical study of the activity of arylsulphatases A and B was carried out on the oviduct of female rabbits during the first days after mating. The histochemical results demonstrated that the ampullary and the isthmic epithelial cells have a positive reaction to the sulphatases during the whole of the postovulatory period tested. The enzymatic activity is mainly localized in the basal cellular cytoplasm. The biochemical results confirmed that both arylsulphatase A and B are active. Arylsulphatase A activity is more intense in the ampulla than in the isthmus and it increases during the whole of the postovulatory period; in the isthmus the activity increases up to 72 h, thereafter decreasing again. The arylsulphatase B activity is always lower than arylsulphatase A activity; maximum activity is reached between 66 to 72 h after mating. The arylsulphatase B is relatively higher in the ampulla than in the isthmus. The biological role of these enzymes is discussed in relation to the regulation of the sulphated glycoconjugates.     

Effects of delayed excision of oviducts/ovaries on mouse oocytes and embryos

To achieve the best and reproducible results of experiments, effects of delayed excision of oviducts/ovaries on mouse ovarian/ovulated oocytes and embryos have been studied. Oviducts/ovaries were excised at different times after death of mice and effects of the postmortem interval on ovarian/ovulated oocytes and embryos were analyzed. When oviduct excision was delayed 10 min, many ovulated oocytes lysed or underwent in vitro spontaneous activation, and this postmortem effect aggravated with the extension of postmortem interval and oocyte aging. Oocytes from different mouse strains responded differently to delayed oviduct removal. Delayed oviduct excision did not cause lysis of zygotes or embryos but compromised their developmental potential. When ovaries were excised at 30 min after death, percentages of atretic follicles increased while blastocyst cell number declined significantly after oocyte maturation in vitro. Preservation of oviducts in vitro, in intact or opened abdomen at different temperatures and histological analysis of oviducts from different treatments suggested that toxic substance(s) were secreted from the dying oviducts which induced oocyte lysis and spontaneous activation and both this effect itself and the sensitivity of oocytes to this effect was temperature dependent. It is concluded that a short delay of oviduct/ovary removal had marked detrimental effects on oocytes and embryos. This must be taken into account in experiments using oocytes or embryos from slaughtered animals. The data may also be important for estimation of the time of death in forensic medicine and for rescue of oocytes from deceased valuable or endangered mammals.     

Origin of oestrus-associated glycoproteins in bovine oviductal fluid.

The aim of the present study was to determine whether the synthesis of an oestrus-associated protein found in bovine oviductal fluid varies with oviductal region, stage of cycle or day of pregnancy. Explant culture was performed using oviducts recovered from naturally cycling animals either at oestrus or 12-14 days after oestrus. Three oviductal regions, the preampulla, ampulla and isthmus, were cultured individually in the presence of 20 muCi [35S]methionine in serum-free medium for 6 h at 37 degrees C. Synthesis of oestrus-associated protein was assessed by one-dimensional SDS-PAGE, fluorography and densitometry of radiolabelled bands. Significantly more oestrus-associated protein was synthesized by the ampullar region of the oviduct, although it was detected in explant culture media from both the isthmic and preampullar regions. A polyclonal antibody produced against oestrus-associated protein was used to localize the protein in paraffin-embedded sections of oviductal explant cultures and other bovine tissues. Localization of the protein in oviductal tissue sections varied with stage of cycle (oestrus > luteal > pregnant) and region of oviduct (ampulla > preampulla/isthmus). These findings indicate an effect of oviductal region and hormonal state (cycling versus pregnant) on the synthesis and secretion of the oestrus-associated protein. Lectin affinity studies indicated that galactosyl(beta 1,3)N-acetylgalactosamine and N-acetylglucosamine residues are associated with the oestrus-associated protein.