Pregnancy diagnosis in cats using a rapid, bench-top kit to detect relaxin in urine.

Relaxin is a pregnancy-specific hormone in the queen and is produced by the placenta. Both serum and urinary relaxin levels can be used to diagnose and monitor pregnancy in the cat; however, only serum levels are commonly measured in practice. The present study aimed to assess whether urine could be used for the rapid diagnosis of pregnancy at an early stage in domestic cats using a bench-top kit to detect relaxin. Paired serum and urine samples were collected during the first month of gestation in six cats. The samples were tested by applying neat serum, urine or urine diluted in non-pregnant cat serum to the Witness Relaxin kit. Relaxin concentrations in the paired samples were also measured by radioimmunoassay. All undiluted urine samples from pregnant cats tested negative using the bench-top kit; however, the kit was able to detect relaxin in urine after dilution with non-pregnant cat serum. Using this as the test sample, the kit was accurate at diagnosing pregnancy from 28 days after mating and some samples tested positive at 21 days after mating. This preliminary work could lead to the development of a home pregnancy test for cats.

Canine preprorelaxin: nucleic acid sequence and localization within the canine placenta.

Employing uteroplacental tissue at Day 35 of gestation, we determined the nucleic acid sequence of canine preprorelaxin using reverse transcription- and rapid amplification of cDNA ends-polymerase chain reaction. Canine preprorelaxin cDNA consisted of 534 base pairs encoding a protein of 177 amino acids with a signal peptide of 25 amino acids (aa), a B domain of 35 aa, a C domain of 93 aa, and an A domain of 24 aa. The putative receptor binding region in the N'-terminal part of the canine relaxin B domain GRDYVR contained two substitutions from the classical motif (E-->D and L-->Y). Canine preprorelaxin shared highest homology with porcine and equine preprorelaxin. Northern analysis revealed a 1-kilobase transcript present in total RNA of canine uteroplacental tissue but not of kidney tissue. Uteroplacental tissue from two bitches each at Days 30 and 35 of gestation were studied by in situ hybridization to localize relaxin mRNA. Immunohistochemistry for relaxin, cytokeratin, vimentin, and von Willebrand factor was performed on uteroplacental tissue at Day 30 of gestation. The basal cell layer at the core of the chorionic villi was devoid of relaxin mRNA and immunoreactive relaxin or vimentin but was immunopositive for cytokeratin and identified as cytotrophoblast cells. The cell layer surrounding the chorionic villi displayed specific hybridization signals for relaxin mRNA and immunoreactivity for relaxin and cytokeratin but not for vimentin, and was identified as syncytiotrophoblast. Those areas of the chorioallantoic tissue with most intense relaxin immunoreactivity were highly vascularized as demonstrated by immunoreactive von Willebrand factor expressed on vascular endothelium. The uterine glands and nonplacental uterine areas of the canine zonary girdle placenta were devoid of relaxin mRNA and relaxin. We conclude that the syncytiotrophoblast is the source of relaxin in the canine placenta.

Pyrogenicity of etiocholanolone and interleukin-1 in New and Old World Monkeys.

Etiocholanolone (5beta-androstan-3alpha-ol-17-one; designated E) is one of the major products of metabolism of testosterone and androstenedione (androst-4-ene-3,17-dione) in many mammalian species, including humans. E and several other 5beta-reduced steroids have been found to induce fever in humans. The pyrogenic effect of these steroids has been shown to be due to the release of interleukin-1 (IL-1) from the leukocytes that are mobilized in response to the steroid injections. Old World Monkeys such as Rhesus monkeys (Macaca mu/atta), metabolize androgens similarly to humans, and E is a normal metabolite. However, New World Monkeys such as Squirrel monkeys (Saimiri sciureus), lack hepatic 5alpha- and 5beta-steroid reductases and excrete androgens primarily in an unaltered state; E is not produced. Therefore, we postulate that Squirrel monkeys likewise may have lost the ability to respond to 17-ketosteroids such as E. To test this hypothesis, adult male Rhesus and Squirrel monkeys were treated with E, and their rectal temperatures were recorded over a 24-hr period. Rhesus monkeys exhibited a rise of up to 3 degrees F following E injection. Squirrel monkeys, on the other hand, did not exhibit any increase in rectal temperature over the 24-hr period, even when doses up to 250 times the effective human dose were used. However, both species responded to injected IL-1alpha with a robust increase in rectal temperature. The data show that E is pyrogenic in Rhesus, but not Squirrel monkeys. The findings support the notion that injected E may induce release of IL-1 in Rhesus monkeys, but not in Squirrel monkeys.

Dogfish shark (Squalus acanthias) testes contain a relaxin.

Relaxin is a 6-kd polypeptide that exerts important hormonal effects in many female mammals. Relaxin is produced by the ovary, placenta, or uterus in many mammalian species. The functions of relaxin in the male mammal are not yet firmly established, but there is some evidence suggesting an exocrine effect on sperm motility and fertilizability. In the male mammals that have been studied, relaxin is produced by the prostate gland (human) or seminal vesicles (boar). However, in the bird, the testis is the likely source of relaxin. Among the elasmobranchs, ovaries obtained from dogfish sharks have been shown to contain a polypeptide hormone that is structurally, biologically, and immunologically similar to mammalian relaxins, but the male reproductive tract of this species has not previously been investigated as a potential source of relaxin. Extracts of testes obtained from mature dogfish sharks have now been tested by a specific relaxin bioassay and by a homologous porcine radioimmunoassay for the presence of relaxin. Both crude and partially purified testicular extracts contained unmistakable guinea pig pubic symphysis-"relaxing" activity and relaxin-like immunoactivity. Following immunoaffinity purification, the shark testis polypeptide had an apparent specific activity of 88 microg porcine relaxin equivalents per milligram in the radioimmunoassay, which is similar to the immunoactivity of pure shark ovarian hormones. These data, therefore, strongly support the view that in dogfish sharks, the male as well as the female gonad produces relaxin. Furthermore, as the dogfish shark has existed as a species for about 200 million years, the data suggest that testicular relaxin appeared early in vertebrate evolution.

Pattern and source of secretion of relaxin in the reproductive cycle of the spotted hyena (Crocuta crocuta).

Female spotted hyenas are highly masculinized at birth and have no external vagina. Copulation with males and birth of young are accomplished through the central urogenital canal of the clitoris. This unusual adaptation requires remarkable changes in the elasticity of the connective tissues of the clitoris, without which neither copulation nor birth would be possible. We hypothesized that relaxin, a hormone that increases the extensibility of the connective tissues of the uterus and cervix of many other mammalian species, plays a role in the clitoral changes observed in hyenas. Serum relaxin was determined by specific RIA. Relaxin was not detected in serum of males, pubertal or nonpregnant adult females, or ovariectomized females. Immunoactive relaxin was detected in serum of juveniles at the time of initial growth of the urogenital meatus. High concentrations of immunoactive relaxin appeared in the serum of pregnant hyenas in the 2 wk preceding parturition. Immunoassays of extracts of hyena tissues and serum obtained from uterine and ovarian veins indicated that the placenta was the predominant source of relaxin, with possible ovarian contributions. Circulating relaxin decreased promptly following cesarean section near term. We conclude that relaxin secretion coincides with changes in extensibility of clitoral connective tissues 1) during growth of the clitoris in juveniles and 2) near the time of parturition in adults.

Chemical synthesis of a Zwitterhormon, insulaxin, and of a relaxin-like bombyxin derivative.

The structural motif of insulin and relaxin is frequently seen in molecules of divergent functions and origins. The insect developmental factor bombyxin, the relaxin-like factor from Leydig cells, and the insulin-like factor 4 (INSL4) all are made of two disulfide-linked chains and have one disulfide bond within the A-chain. The polyclonal antibody R6, which was raised against porcine relaxin, reacts with a wide variety of naturally occurring relaxins from primates, marine and terrestrial mammals, and elasmobranchs but does not recognize insulin. The antibody binds mainly to the arginines that occur in the N, N+4 positions in the B-chains of all relaxins which also constitute the receptor-binding site. The receptor-binding haptens were incorporated by total synthesis into human despentapeptide insulin and bombyxin II, a developmental factor from the silk moth Bombyx mori. In the process the insect factor became a perfect antigen for the anti-relaxin antibody, whereas the human insulin was transformed into a bona fide relaxin. The conversion was affected by changing four critical residues so that the insulin activity was retained to the extent of 10% of the original level. This, to the best of our knowledge, is the first designer protein to incorporate two unrelated biological functions in one primary sequence, and we are therefore proposing that, analogous to zwitterion, the generic name "Zwitterhormon" (German spelling) be used for this type of construct.

Use of synthetic canine relaxin to develop a rapid homologous radioimmunoassay.

Canine relaxin (cRlx) was synthesized by a combination of solid-phase methods and sequential site-directed disulfide bond formation. Proof that the intended molecule had been synthesized was obtained by analytical HPLC of the intact and reduced molecule, by amino acid and sequence analysis, and by receptor binding and in vivo mouse interpubic ligament bioassays. Antisera to synthetic cRlx were raised in six male rabbits; these cross-reacted with relaxins of other species, but not with insulin, LH, FSH, hCG, or prolactin (PRL). Three of the antisera neutralized relaxin-induced interpubic ligament formation in estrogen-primed mice in vivo. A new homologous cRlx RIA was developed through the use of rabbit antiserum 79888, synthetic cRlx for standards and 125l-labeled trace, and a goat anti-rabbit lgG-polyethylene glycol precipitant. The new RIA can be completed in 26 h and has a sensitivity of 0.195-0.39 ng cRlx/tube. Intra- and interassay coefficients of variation were 3% and 12.5%. During pregnancy in bitches, serum cRlx rose to about 10 micrograms/ml. Immunoactive cRlx was also detected in serum, colostrum, and milk of lactating bitches, but not in large volumes (100-300 microliters) of serum of pseudopregnant or estrous bitches. Immunoreactive cRlx was also found in seminal plasma, but not in serum, of male dogs. The new homologous cRlx RIA is simple, rapid, sensitive, and specific, and will be used in future studies of canine relaxin physiology.

Lipoprotein profiles and glucose tolerance in lean and obese chimpanzees.

We compared serum lipid profiles and glucose tolerance of obese and lean chimpanzees maintained on a 10.9% fat diet. Seven of 14 obese and 6 of 17 lean chimpanzees were hypercholesterolemic (low density lipoprotein cholesterol > 160 mg/dl), three obese and three lean animals had total cholesterol/high density lipoprotein cholesterol ratios of 5.9-10.7, and two obese and one lean chimpanzee had abnormal glucose tolerance. Useful numbers of captive chimpanzees thus exhibit metabolic abnormalities without recourse to high fat diets and could serve as surrogates in studies of human metabolic diseases.

Patterns of relaxin and steroids in the reproductive cycle of the common marmoset (Callithrix jacchus): effects of prostaglandin F2 alpha on relaxin and progesterone secretion during pregnancy.

We measured the concentrations of relaxin (Rlx), progesterone, and estradiol-17 beta in serum samples obtained twice or three times weekly from marmosets during the estrous cycle and pregnancy. The cyclic patterns and concentrations of progesterone and estradiol-17 beta were similar to those reported by previous investigators. Rlx was not detected in individual serum samples ( < 0.62-1.25 ng/ml) obtained from nonpregnant marmosets. However, pooling of luteal serum from all animals permitted assay of much larger volumes of serum (0.4 ml vs. 0.1 ml), and a concentration of about 1 ng/ml was detected. Rlx was first detected in serum in the second or third week of the 21-wk marmoset pregnancy, rose to a peak during Weeks 10-14, and then declined slowly as the time of parturition approached. The pattern of Rlx was unlike that observed during pregnancy in Old World monkeys, chimpanzees, or women, and resembled, instead, that seen in rodents, carnivores, and equids. Progesterone and estradiol-17 beta likewise increased throughout pregnancy, and their patterns were similar to those previously described for marmosets by other investigators. The concentrations of the steroids and Rlx in serum of pregnant marmosets was 10-fold or more higher than those found in Old World monkeys, baboons, chimpanzees, or women. Spontaneous abortions in two of the marmosets were accompanied by precipitous falls in serum levels of progesterone, estradiol-17 beta, and Rlx. Following s.c. injection of the luteolytic agent prostaglandin F2 alpha (PGF2 alpha) into two marmosets at midpregnancy, serum progesterone and Rlx fell to low levels. These animals received a progestin, 17 alpha-ethyl-19-nortesterone, to preclude abortion. Serum progesterone rose again, but serum Rlx remained low for the duration of pregnancy.(ABSTRACT TRUNCATED AT 250 WORDS)

Relaxin and its role in pregnancy.

Relaxin is a 6000-d polypeptide, structurally related to insulin and the insulin-like growth factors. Unlike insulin, the structure of which is remarkably well conserved among the vertebrates, relaxin sequences can vary by more than 50% between different species. Despite these large sequence variations, relaxins (with few exceptions) have very similar biologic activities in animal test systems. The reason for this has recently come to light: the receptor binding region of the B chain, in contrast to the rest of the molecule, is highly conserved between species. Relaxin is measured by bioassays employing interpubic ligament formation in mice and guinea pigs, and by inhibition of uterine motility. A more sensitive and efficient bioassay is urgently needed. In women, the target organs for relaxin are the uterine cervix, myometrium, endometrium, and decidua. Other presumptive but unproven targets are the pubic symphysis and sacroiliac joints, mammary glands, and pituitary gland. Circulating relaxin is secreted by the corpus luteum. The placenta, decidua, or both also produce relaxin, which does not enter the circulation but may act in an autocrine or paracrine fashion. hCG is a stimulus to luteal relaxin secretion. Other regulatory factors are poorly defined. Aluteal women are hyporelaxinemic, and yet are capable of normal vaginal delivery of their infants. Local effects of placental or decidual relaxin cannot be discounted in such subjects. Hyperrelaxinemia may occur in women with multiple gestations and ovarian stimulation, and may be associated with increased premature births. Serum relaxin also is elevated in pregnant diabetics, but its role in this condition has not been defined. Clearly, further investigations are needed to delineate the precise role of relaxin in human pregnancy.

Inhibition of relaxin-induced pubic symphyseal "relaxation" in guinea pigs by glycosaminoglycan polysulfates and pentosan polysulfate.

There are similarities between the actions of estrogen and relaxin on the connective tissues of the pubic symphysis and those of neutral proteases on cartilage in osteoarthritis, including cartilage hydration, proteoglycan loss, and dissolution of collagen fibers. We hypothesized that compounds known to inhibit cartilage breakdown in animal models of osteoarthritis, such as polysulfated GAGs, would also antagonize the actions of estrogen and relaxin that increase the laxity and mobility of the pubic symphyses of guinea pigs. Estrogen-primed guinea pigs were injected with relaxin or with relaxin and the test compound. The pubic symphyses were manually palpated 6 h later and the degree of mobility scored. Glycosaminoglycan polysulfates and pentosan polysulfate inhibited relaxin-induced pubic symphyseal relaxation, whereas other types of agents were without effect. The guinea pig pubic symphysis assay for relaxin may thus provide a novel rapid screening test for compounds with potential chondroprotective activity.

Transmission of relaxin from lactating bitches to their offspring via suckling.

The 6-kDa polypeptide hormone relaxin (Rlx) has been identified in human and bovine milk, and we recently reported its presence in canine milk. We postulated that Rlx might be transferred via suckling to the newborn pups, where, by virtue of its known effects to increase the distensibility of the pelvic connective tissues, it could play a role in causing the excessive laxity of the capsule and ligaments of the coxofemoral joint that precedes the development of hip dysplasia in genetically predisposed animals. Rlx was found in the serum of dysplastic (HD+) bitches for up to 6 wk of lactation, whereas it was detected in the serum of nondysplastic (HD-) bitches for only 1-2 wk of lactation. Rlx concentrations in milk were up to 60-fold greater than in serum. Milk Rlx levels varied markedly, but were highest during the first week of lactation and decreased thereafter. There were no significant differences in milk Rlx concentrations between HD+ and HD- bitches. Although the source of Rlx in milk is unknown, it cannot be the ovary or uterus, since hystero-ovariectomy performed at the time of cesarean section did not eliminate Rlx from milk during subsequent lactation. In serum samples taken from newborn pups before suckling, there were significant quantities of Rlx, demonstrating that the hormone enters the fetus in utero. However, Rlx rapidly disappears from serum of pups prevented from suckling for five hours.(ABSTRACT TRUNCATED AT 250 WORDS)

Maternal relaxin concentrations in diabetic pregnancy.

Maternal serum concentrations of relaxin, an insulin homologue produced both by the corpus luteum of pregnancy and by the fetoplacental unit, are highest in the first trimester and fall to their lowest level in the third trimester. Relaxin is thought to influence carbohydrate metabolism in the uterus, and it has been suggested that serum concentrations of relaxin in diabetic women are higher than those of non-diabetic women. We show that maternal serum relaxin concentrations are significantly higher at each stage of pregnancy in insulin-dependent diabetic mothers than in non-diabetic mothers. This elevation in relaxin concentrations is not related to other indices of diabetic control. The physiological importance of the higher concentrations of relaxin in the serum of diabetic women--in particular, whether they contribute to the higher incidence of major anomalies in the fetuses of diabetic mothers--is yet to be determined.

Determination of the time of ovulation in chimpanzees by measurement of LH, estrone sulfate, and pregnanediol 3 alpha-glucuronide in urine: comparison with serum hormone patterns.

The concentrations of LH, total estrogens, and pregnanediol 3 alpha-glucuronide (PdG) were determined by specific radioimmunoassays on daily overnight urine samples obtained in 13 menstrual cycles of six adult female chimpanzees during the periods of increasing, maximal, and decreasing tumescence of the perineal sex skin. The peaks of estrogens and LH and the rise in PdG in urine accurately reflected the peaks of estradiol-17 beta and LH and the subsequent rise in progesterone in the serum of the same animals during the same menstrual cycles, and can be used to predict and verify the occurrence of ovulation, thus avoiding the repeated tranquilizations necessary to obtain daily blood samples.

Serum concentrations of relaxin, chorionic gonadotropin, estradiol-17 beta, and progesterone during the reproductive cycle of the chimpanzee (Pan troglodytes).

Levels of serum relaxin were measured by a specific RIA and correlated with serum patterns of estradiol-17 beta, progesterone, LH, or cCG during a single menstrual cycle in each of 10 female chimpanzees, and throughout 24 pregnancies in 21 chimpanzees. Significant concentrations of relaxin, higher than those reported for the human being, were detected in serum of nonpregnant chimpanzees during the late luteal phase of the menstrual cycle. During pregnancy in the chimpanzee, serum relaxin concentrations, exceeding levels found during the luteal phase, were highest during the first third of gestation, and declined thereafter. Although the absolute concentrations were higher, the patterns of relaxin secretion throughout the reproductive cycle in chimpanzees was qualitatively very similar to that observed in other primates, including the human being. The chimpanzee should thus provide a useful model for examining the role of relaxin in human reproduction.

The role of estrogen and relaxin in the reproductive abnormalities of mice with Hertwig's anemia.

Mice homozygous for a mutant allele (an/an) causing a lifelong macrocytic anemia (Hertwig's anemia) also demonstrate an inability to deliver their offspring, despite normal ovulation, conception, implantation, and fetal development. We investigated the roles of estrogen and relaxin in the etiology of the reproductive defect in the Hertwig's anemia mice. Immunoreactive relaxin levels were undetectable in the nonpregnant controls, whereas levels in both timed-pregnant controls and timed-pregnant affected mice were significantly higher than in nonpregnant controls, but not significantly different from each other. Mean interpubic ligament length in the pregnant Hertwig's anemia mice was significantly greater than that in nonpregnant controls, but significantly less than that in the pregnant controls on Day 18 of pregnancy. Porcine relaxin was administered to nonpregnant affected and unaffected littermates and to nonpregnant controls. Whereas controls showed a significant response to porcine relaxin, neither the Hertwig's anemia mice nor their unaffected littermates responded to the porcine relaxin. Additional study was performed to determine estradiol effects in the affected and control animals utilizing detailed computerized morphometric analysis of uterine horns and cervices from immature, estradiol-injected controls and Hertwig's anemia mice. Results demonstrated a statistically significant trophic effect of estradiol upon uterine horn and cervical enlargement, as assessed by weight and volume, in controls. Only a slight, non-significant effect was seen in Hertwig's anemia mice. Additional histological effects of estradiol, including endometrial enfolding observed in controls, were not present in Hertwig's anemia mice. Lack of response to both estrogen and relaxin is responsible for the parturitional defect in Hertwig's anemia mice.

Diurnal variation of serum progesterone, but not relaxin, prolactin, or estradiol-17 beta in the pregnant bitch.

The concentrations of progesterone (P), relaxin (Rlx), estradiol (E2) and PRL were measured by specific RIAs in serum samples collected twice daily at 8:00 am and 3:00 pm at weekly intervals after mating and until whelping in five Labrador Retriever bitches. From weeks 3 to 6 of pregnancy, P exhibited a conspicuous and statistically significant diurnal pattern (P less than 0.05), its serum concentration being approximately twice as high at 8:00 am as at 3:00 pm. A similar but nonsignificant trend was observed weeks 2, 7, and 8, and the overall ratio of the am/pm P concentrations was 2.4 +/- 0.28 (P less than 0.001). Rlx, E2, and PRL did not show a diurnal pattern at any time during pregnancy. The glandular sources and regulation of secretion of Rlx were further investigated. Rlx bioactivity was detected in canine placentas and ovaries, providing supportive evidence that these organs may be a dual source of the hormone in canine pregnancy. Injection of ovine PRL in three pseudopregnant bitches significantly increased serum P concentration as compared with seven untreated pseudopregnant controls, but Rlx was not detectable in serum before, during or after PRL treatment. The data support the view that Rlx and P are independently regulated in the bitch; PRL may be a regulator of P, but not of Rlx secretion. However, as PRL showed no diurnal changes in concentration, it does not appear to be directly responsible for the diurnal pattern exhibited by P.

Characterization of the circulating form of porcine relaxin: biological activity and terminal amino acids.

Relaxin is structurally related to insulin, and it induces pregnancy-related changes in the reproductive tract of several mammalian species. Relaxin isolated from the ovaries of pregnant sows has been used for primary structure determination, for much of the biological characterization of the hormone, and for the development of RIAs. Immunoreactive (IR) relaxin is found in peripheral blood during pregnancy in pigs and other species, but it has not been established that the substance identified by RIA is structurally or biologically equivalent to the native ovarian hormone. IR relaxin was, therefore, isolated from peripheral plasma of late pregnant gilts (days 112-114) for bioassay and determination of terminal amino acid residues. IR relaxin was monitored by a specific homologous RIA during fractionation of plasma by gel filtration, cation exchange, and hydrophobic binding to octadecysilica. IR relaxin circulates unbound and is equipotent with ovarian relaxin in the mouse pubic ligament bioassay. Amino acids released from IR relaxin by pyroglutamic aminopeptidase and carboxypeptidase-Y were converted to their 4-dimethylamino-azo-4'-sulfonyl derivatives for identification by HPLC. The B-chain of IR relaxin had an amino-terminal pyroglutamic acid. Amino acids sequentially released from the carboxy-terminal indicated a chain length of 28-30 amino acids, suggesting a heterogeneity reminiscent of that of ovarian relaxin isolated by other methods. Arginine was released from the free amino-terminal by dimethylaminozaobenzene-isothiocyanate degradation, indicating an intact A-chain of 22 amino acids. Blood immunoreactive relaxin in pigs is, therefore, a secreted biologically active form of relaxin with an amino acid composition similar to that of the form stored in the corpus luteum.

Serum relaxin and progesterone concentrations in pregnant, pseudopregnant, and ovariectomized, progestin-treated pregnant bitches: detection of relaxin as a marker of pregnancy.

Serum concentrations of relaxin and progesterone were measured by specific radioimmunoassays in pregnant, pseudopregnant, or ovariectomized (between gestation weeks 4 and 5) pregnant Labrador Retriever bitches. Daily administration of 17 alpha-ethyl-19-nortestosterone was performed to maintain gestation in the ovariectomized pregnant bitches. This synthetic gestagen was selected because it did not interfere with the assay for endogenously secreted progesterone concentration in serum. Serum progesterone concentration was high in ovarian-intact pregnant or pseudopregnant bitches, but the mean progesterone concentration in pseudopregnant bitches (evaluated at 4 weeks after mating) was only 56% of the concentration in pregnant bitches. After ovariectomy, serum progesterone concentration decreased to undetectable values. Unlike progesterone, serum relaxin concentration increased during the latter half of pregnancy in the ovarian-intact and in the ovariectomized pregnant bitches, but relaxin was not detectable at any time in the pseudopregnant bitches. The amount of relaxin measured in the ovariectomized pregnant bitches was less (P less than 0.05) than that in ovarian-intact bitches, suggesting that the ovaries may have contributed to the total circulating relaxin concentration in the latter. Placental production of relaxin might have accounted for the serum relaxin concentration after ovariectomy; thus, the ovary and placenta each may secrete relaxin during gestation in bitches. Regardless of its source, measurement of serum relaxin concentration may offer a useful way of distinction between pregnancy and pseudopregnancy in dogs.