T H E DISTRIBUTION AND SOURCE O F OESTRIN I N THE PREGNANT MARE H. R. C A T C H P O L E A N D H. H. COLE College o f Agriculture, University of California, Davis, California Oestrin is present in mare serum between the one hundred and fiftieth day of pregnancy and term in low concentration (Cole and Hart, '30 and '30 a ) but extremely large amounts are excreted in the urine (Zondek, '30); frequently the hormonal content of the urine is ten times greater than that of the pregnant woman. Glud et al. ( ' 3 3 ) have shown that considerable amounts are likewise excreted in the feces. A systematic study has been made by Cole and Saunders (unpublished) of the hormonal outputs of a number of individual mares which have been available throughout pregnancy. They have found that oestrin appears in the urine between the fiftieth and one hundredth day of pregnancy. The range in time of appearance of the hormone is undoubtedly less than was indicated by Cole et al. ( '33). I n arriving at this earlier estimate of a range extending from the sixtieth to the one hundred and seventy-fifth day of pregnancy one case was included in which, as was later disclosed, there was fetal resorption and oestrin never appeared in the urine. We have been interested in this hormone in relation to other phenomena occurring in the mare in pregnancy. These include : 1) the output of gonad-stimulating hormone during the first half of pregnancy; 2) the hypertrophy of the interstitial cells of the gonads of the fetus and the finding in them of fairly considerable amounts of female sex hormone; 3) the behavior of the maternal ovaries which after regression both of the primary corpus luteum of pregnancy, and also of the secondary crop of corpora induced presumably by the gonad 335 336 H. R. C A T C H P O L E A N D H. H. COLE stimulating hormone of the serum, become apparently quiescent during the last third of pregnancy. A brief review of these findings has been reported by Cole et al. ( '33). The fetal gonads reach their maximum size a t a fetal C. R. length of 45 to 65 em., a time when the maternal ovaries appear quiescent, and when the oestrin output in urine is high. This tempted a suggestion that the fetal gonads might be one source of the large amounts of hormone that are appearing in the urine of the mare throughout this period. Although aware from a parallel study of gonad-stimulating hormone of the limitations of this particular method in dealing with questions of source of hormone, we were nevertheless sufficiently interested in the phenomenon of hormonal distribution in this species to make a number of comparative assays of various tissues for their oestrin content. These data are of greater significance when correlated with data which we have accumulated on the disappearance of oestrin from mare and foal urine following parturition. MATERIALS Our source of materials was the same as described in a preceding paper (Cole et al., '33). I n view of the labor involved in preparing extracts for the assay of oestrin, a very limited series of tissues in which the fetal length varied from 28 em. to term were obtained. I t was, unfortunately, not possible in all the cases to obtain a sample of urine from the slaughtered animals. From the mares we took the ovaries and samples of liver, kidney, endometrium and blood. From the fetal side we took gonads, liver, kidneys, placenta and blood. I n addition to the slaughter-house material we were interested in studying the hormonal content of mare urine preceding and following parturition. We also determined the oestrin content of newborn foal urine. The placentas of these mares, dropped at term, were also investigated f o r oestrin content. OESTRIN I N T H E PREGNANT MARE 337 METHODS a. Preparation. o f extracts. The preparation of oily extracts containing the female sex hormone, oestrin, is possible by a variety of methods relying essentially on the same principle, namely, destruction of the tissue by rather drastic chemical treatment, followed by extraction of the product successively with various lipoid solvents. The method used by Fevold, Hisaw and Leonard ('32) for the preparation of progestin embodies these same principles. Allen and Meyer ('33) remark that the preparation of crude oestrin and crude progestin extracts is carried out in exactly the same way, and we had previously (Cole et al., '33) employed the rapid progestin method of Fevold for the preparation of oestrin from fetal gonads. It was found convenient t o use this method in the present study. An important feature of the method is the initial boiling of the tissue with a fairly strongly acid solution of 95 per cent ethyl alcohol. We have applied this same process in the preparation of oestrin from mare urine after initial concentration of the urine by evaporation, with no indication that much hormone was being lost in the process, and consider it entirely suitable for the rather rough assays in the present instance. The method employed consisted in the working up of usually 100 gm. tissue by the 'rapid' method (Fevold et al., '32) a s described for progestin carrying the process to the methyl alcohol stage for removal of most of the cholesterol. The products were in every case thick brown oils. The urine samples cited in tables 2 and 3 were not treated chemically. b. Assay of hormowe. The products were injected in three doses at 8-hour intervals into 4-month-old spayed virgin female rats. Vaginal smears were taken at 48,72 and 96 hours after the first injection. The animals were used over again a t intervals of 2 or 3 weeks, with the usual precaution of initial smearing before injection. The doses corresponded to a certain original tissue weight. After a preliminary experiment to determine the approximate concentration, the rat unit was determined by halving the dose until the smallest 338 H. R. CATCHPOLE AND H. H. COLE amount was found which would give a smear of nucleated epithelial cells and cornified cells, or the latter alone, in at least three out of four rats within 48 hours. I n the testing of some extracts only two rats were used at each level and in these cases the rat unit was designated as the minimal dose at which both rats reacted.l I n the results, the weight of original tissue corresponding to a rat unit will be stated. The earlier paper (Cole et al., '33) deals with the oestrin content of fetal ovaries, testes, livers and blood. Referred t o original tissue weights, it was found that with the fetal gonads, from 8 to 32 gm. were necessary to give a positive vaginal reaction by the method of subcutaneous injection, although by smearing the oils into the vaginae of the test animals as little as 0.15 gm. was effective. This was explained by the poor absorption of the injected preparations. It was found that extracts made by the above procedure, whether intended for oestrin or progestin were invariably strongly acid in reaction. This acidity seems to be responsible for the skin lesions so often reported after the use of these preparations. If the extracts be properly neutralized, such lesions are never produced, and absorption is good. This is reflected in a decrease in the amounts of tissue required to produce a reaction. I n the case of fetal gonad extract, the minimal dose fell to 1.5 to 3.0 gm. A discrepancy between the methods of injection and vaginal smearing still exists, although to a less degree. It must be attributed t o a localization of action in the latter method whereby the hormone passes directly to its site of action, without entering the general circulation, whence it is probably rapidly excreted. I n the series reported in this paper the extracts were injected subcutaneously in all instances. Conce&ration of oestrin in horse tissues The activity of several maternal and fetal tissues from four cases is shown in table 1. It may be seen that in very late 'Judging from the values given by Zondek ('31) f o r oestrin in mare urine, our rat unit is equivalent to ten t o twenty of his mouse units. 339 OESTRIN I N T H E PREGNANT MARE stages of pregnancy (case H12 with a fetus having a crownrump length of 90 em.) there is less oestrin present in all tissues than in the cases taken a t earlier stages of pregnancy. Oestrin is evidently distributed in both maternal and fetal tissues. I t is clear that mare kidney, endometrium and fetal placenta contain larger amounts than other tissues listed. The high activity of the mare kidney accords with the high oestrin content of the urine and must be due partly to included urine and partly t o the presence of oestrin in the tubular or TABLE 1 A s s a y of oestrin an mare and f e t a l tissues a t different stages of pregnancy - ~~ . ~ ~~ ~~ .. .. - ~ . .~ ~ ~~~ ~- I E2 70 C A S E DESIGNATION P X T A L C. R. LENGTH IN CENTIMETERS , I H12 90 IMILLIGRAMS OF TISSUE NECESSARY TO EVOKE AN OESTROUS REACTION I N T H E SPAYED VIRGIN F E M A L E R A T ~ _. hiarc urine Mare kidney 1M are liver Mare serum Mare ovaries ~- ~ 400 I >3200 Endoinetriuin Fetal placenta Fetal gonads Fetal liver Fetal kidney Fetal blood 800 3000 I ~ 3200 3200 150 800 2000 >3200 1600 >3200 I 4000 >3200 500 1400 3400 1100 1700 >3200 1600 >3200 3200 >3200 >3200 - . -. - - glomerular epithelium. It appears certain that the kidney is removing oestrin selectively from the blood stream, since no tissue yet investigated contains a concentration of oestrin comparable with urine. The activity of the endometrium is comparable to that of the mare kidney while that of the fetal placenta is somewhat less. The high concentration of oestrin in the endometrium and fetal placenta may be accounted for in a number of ways: It is possible that oestrin is being formed in one or both of these tissues. Evidence to be submitted later in this paper will have a bearing upon this assumption. O r it may be that oestrin is formed in the fetus 340 H. R. C A T C H P O L E A N D H. H. C O L E and merely becomes concentrated in these tissues because they are concerned with its excretion or transfer from the fetal t o the maternal blood stream. I n another paper (Cole et al., '33) we suggested that possibly the fetal gonads were concerned with the production of oestrin. Although there is an interesting correlation between the development of the fetal gonads and the concentration of oestrin in the urine (the maximum concentration of oestrin in the urine of the mare occurs at the time when the fetal gonads reach their greatest development, and the subsequent fall in the oestrin content of the urine a t the time when the fetal gonads are regressing), the evidence in this paper does not support such a theory. All fetal tissues studied, gonads, livers and kidneys, contain oestrin in about the same concentration and thus the presence of oestrin in the fetal gonads loses the special significance which we had previously attached t o it. Further, data t o be given later on the disappearance of oestrin from foal urine after parturition makes the assumption of oestrin production by the fetal gonads still less tenable. Finally, in explaining the high concentration of oestrin in endometrium and fetal placenta, it might be assumed that oestrin is produced by the maternal gonads and selectively stored or used by the placental membranes. This is improbable for it has been shown (Cole, Howell and Hart, '31) that the maternal ovaries are in an apparently quiescent state during that phase of pregnaney when the oestrin content of the urine is greatest. Further, we have shown here that the maternal ovaries contain relatively small amounts of oestrin. An experiment now in progress involving the removal of the maternal gonads during pregnancy should completely settle the question as to their involvement in the oestrin secretion of the mare during this period. I n addition to the tissues listed in table 1we have extracted and assayed three term placentas. In one an equivalent of 1600 mg. of fresh tissue and in another 3200 mg. represented a rat unit; in the third 3200 mg. was negative. 341 OESTRIN IN T H E PREGNANT MARE Oestrin in the urine We have studied t,he oestrin content of the urine of four mares for 2 months antepartum and shortly after this event in the hope that it might throw additional light on the source of the oestrin of pregnancy (table 2). Further, we have examined the urine of newborn foals. It may be seen that during the last 2 months of pregnancy there is a steady fall in TABLE 2 Oestrin in mare urine preceding and following parturition RAT UNITS PER LITER O F URINE I N THE FOLLOWING MARES 1 DAYS PRECEDINQ PARTURITION c7 c9 c2 I ~ 40-55 25-39 15-24 5-14 1-2 16000 8000 4000 4000 2000 8000 4000 4000 4000 1000 <zoo Day of foaling -I_____ 16000 8000 ~ I I. ~ c3 16000 8000 ~ 1 4000 i l--l 1 <loo 4000 '_ _ ~ _ _ TABLE 3 Oestrin in t h e urine of newborn foals 1 Day of foaling 1day after foaling 2 days after foaling RAT UNITS O F OESTRIN PER LITTER O F URINE I N THE FOLLOWING FOALS c2 foal -1 c3 foal ___2000 1 <loo <~OO the oestrin content of mare urine. I n two instances in which samples were obtained 1 to 2 days before foaling it took Q to 1 cc. to produce a response in spayed rats. I n each of the four cases cited in the table doses of 5 cc. of urine collected after parturition were negative. This rapid disappearance of oestrin from the urine of the mare is duplicated in the foal (table 3 ) . I n two foals (c7 and c9) we were fortunate in obtaining the first urine voided after birth. These samples 342 H. R. CATCHPOLE A N D H. H. COLE contained oestrin whereas urine on subsequent days was negative in the amounts administered. I n foal c2 the sample of urine collected on the first day after parturition contained oestrin. This urine was thick, dark brown, contained considerable mucus and had every appearance of the first voided urine. It is quite likely that some foals do not urinate for 24 hours after birth for we have waited 12 to 14 hours t o obtain the first voided urine in other instances. I t is significant that oestrin disappears from maternal and foal urine in the course of a few hours following parturition. If either the maternal or fetal gonads are concerned with the production of part or all of the oestrin excreted in the urine, their activity is precipitously terminated by the expulsion of the fetus and placenta. No corresponding morphological change is apparent in the gonads at this time to account f o r such a sudden change in physiological activity. DISCUSSlON O F RESULTS a. The placenta as an endocrine organ. Halban ( '05) suggested that the placenta takes over hormonal functions during pregnancy, and a considerable number of investigators since, relying frequently on merely finding hormone in this location, have supported his theory. The most recent work on this point has dealt with hormonal output following extirpation of the ovaries during pregnancy. It has been known for many years that pregnancy in woman will take a normal course after bilateral ovariectomy. Waldstein ( '29) performed this operation on a woman in a case of bilateral dermoid cyst a t the estimated thirty-fourth day of pregnancy. The pregnancy was maintained and the birth was normal. The oestrin content of the blood in the eighth month was normal ; that of the urine was lower than the usual very large amounts recorded for normal pregnancies. This implies a considerable ovarian contribution in normal pregnancy. I n the case of woman, it may be remarked, there occurs no fibrous regression of the ovary as in the mare. The hormonal content of the placenta in Waldstein's case was normal, but even so contained only OESTRIN IN T H E PREGNANT MARE 343 a fraction of the total daily excretion of hormone. I n this case it is impossible to regard the hormone as coming out of a placental reservoir, since ovariectomy was performed at such an early age. Following birth, atrophy of the uterus and, up to 7 months following birth, absence of menstruation bore witness to the completeness of the operation. Waldstein concludes that the placenta is to be regarded as a principal source of the oestrous hormone-a conclusion concurred with by Novak, J. Halban and 0. 0. Fellner. That the ovary, if present, may continue to secrete hormone, and that it is necessary for furthering implantation during the first month of gestation, is not denied by these investigators. Szarka ('30) reports a similar case with bilateral ovariectomy at the third month. He found just previously to birth 1000 mouse units of oestrin in the urine as compared with a normal of 5000 to 8000 M.U. This hormone might possibly be proceeding from the fetus, which the observation of de Snoo ('28) does not entirely rule out. The latter worker finds excessive amounts of oestrin being excreted in a case of chorioepithelioma of the tube, in which trophoblastic elements alone are present. But the ovaries were also present, and the extensive ovarian changes shown by Novak and Koff ( '30) to occur in these conditions render this finding unreliable, not only f o r the exclusion of fetal participation, but even for the assumption of a chorionic role. Zondek ( '31) regards Waldstein's experiment as conclusive, and in the German literature, rather more than in the English, the vicarious production of oestrin, at least as far a s the placenta is concerned, seems to be fully accepted. b. Studies im the puerperiurn. Philipps' studies ( ' 3 0 ) show oestrin to be present in the newborn up to 72 hours after birth. Thereafter it diminishes and none is present after 6 days. He speaks in favor of a placental origin, and considers that since the conditions are the same in girls and boys a fetal origin is excluded. I n the urine of post-pregnant women hormone disappears in the course of 8 days (Zondek, '31). Briihl finds female sex hormone in the urines of newborn children without distinction of sex up to the fourth day. T H E ANATOMICAL RECORD, VOL. 5 9 , NO. 3 344 H. R. CATCHPOLE A N D H. H. COLE The theory that the placenta is contributing to the large amounts of hormone that are being produced in the pregnant woman receives strong support from some of the foregoing evidence. The actual amounts that are coming from the active organs must wait on future pregnancy castration studies in this and other forms. c. Deductions from the present studies. Oestrin disappears from the body of the mare even faster than it does in women, and hormone has yet to be demonstrated in the postpartum blood and urine of the mare. The actual daily production of oestrin in the mare at 1 or 2 days antepartum-assuming a daily production of 10 liters of urine and a rat unit in 0.5 to 1.00 cc.-is from 10,000 to 20,000 R.U. The day following birth this figure has fallen t o one-fifth or less of its value, and this abrupt failure in hormone production must be explained. I n both woman and the mare the uniform concentration of hormone in the blood is a striking feature. I n woman this figure is not appreciably altered by castration, although the excretion is reduced. I n both these forms there is an overproduction of hormone reflected in the excretion of the excess. This over-production is specially marked in the mare, in which the blood concentration of hormone is only a little higher than in woman, whereas the excretion is ten times a s great. I n the foal, all the hormone that is present in the body at birth appears to pass completely into the urine during the first day, and is excreted. Whatever its function in the mare it is improbable that its presence in the fetus is more than incidental. The findings with foal urine would appear to rule out the fetal gonads as significant sources of the hormone at this time. The morphological evolution of the fetal gonads is not interrupted by the incident of birth, and were these responsible for only a fraction of the amounts of hormone found immediately antepartum it should be possible to recognize them in a state of hormonic activity beyond the point at which we actually find hormone in the foal-namely at the first urination, and at this point only. The fetal gonads in later stages are much degenerated, and the interstitial cells are vastly different from the healthy interstitial cell masses OESTRIN IN T H E PREGNANT MARE 345 which are present at the time of maximal development; at this latter time the possibility of their hormonic activity cannot be denied. But the present demonstration of the approximate equivalence of fetal body tissues in respect of hormonal content leave histological considerations as the sole support for the theory of fetal gonadal participation in the active production of oestrin in the mare. The question of male sex hormone in the fetal gonads of the male was touched upon in the earlier paper, but has not been further investigated. The ovaries of the mare have been described as undergoing regression in the last third of pregnancy. During this time they are found to be reduced in size, being tough and fibrous, with no follicles grossly recognizable. Highly pigmented (orange or red) flecks are often present in the stroma, representing carotinoid remains of corpora. We would suggest that their inactivity is absolute, as compared with the relative activity of the human ovaries. This would explain the total failure of oestrin in the mare following parturition. There remains to be considered the placenta. The results of the German investigators have shown that it possesses a definite capacity for production of oestrin in the pregnant castrated female. The case for placental production in the mare is rendered strong by the above considerations that appear to rule out the ovary of the mare as a significant secretory organ in the latter part of pregnancy, and the fetal gonad as an unlikely one. The position of the maternal endometrium assumes in this study an importance equal to that in the case of the gonadstimulating hormone ( Catchpole and Lyons, '34). The endometrium is the main absorber of the products of excretion of the fetus and the allanto-chorion, and in the event that the 'placenta be regarded as an important source of oestrin in this species, the activity of the maternal endometrium in the two cases in which it was studied and found to be markedly potent is immediately clarified, without recourse having to be made to the very unlikely hypothesis that it possesses an internal secretory function of its own in respect of oestrin production. 346 H. R . CATCHPOLE A N D H. H. COLE SUMMARY AND CONCLUSIONS 1. A quantitative study has been made of the distribution of oestrin in the tissues of four mares and their respective fetuses. Hormone has been found in mare kidney, endometrium and ovary; in fetal gonads, liver and kidneys; in the allanto-chorion, both during pregnancy and at term. 2. The amounts of oestrin in the urines of mares and foals at parturition have been investigated. 3. These findings are discussed in relation to the histological pictures in the gonads of the fetus and the ovaries of the mare over the period studied, and also in relation to the excretion of oestrin in mare urine during pregnancy. Their bearing on the source of the oestrous hormone of pregnancy is discussed. 4. It is considered that the theory of placental formation of hormone must be adopted in this form. A basis is thereby given for the quantitative results. L I T E R A T U R E CITED ALLEN,W. M., AND R. K. MEYER 1933 The quantitative separation of progestin from oestrin in extracts of the corpus luteurn. Am. J. Physiol., vol. 106, p. 53. BRUHL, R. 1929 Das Vorkommen von weiblichen Sexualhormon und Hypophysenvorderlappenhormon im Blute und Urin von Neugeborenen. Klin. Woch., Jahrg. 8, 8. 1766. CATCHPOLE, H. R., AND W. R. LYONS 1934 The gonad stimulating hormone of pregnant mares. Am. J. Anat. ( I n press.) COLE,11. H., AND G. H. HART 1930 The potency of blood serums of mares i n progressive stages of pregnancy i n effecting the sexual maturity of the immature rat. Am. J. Physiol., vol. 93, p. 57. 1 9 3 0 a Sex hormones i n the blood serum of mares. 11. The sera of mares from the 222nd day of pregnancy t o the first heat period post partum. Am. J. Physiol., vol. 94, p. 597. COLE,11. H., C. E. HOWELL, AND G. H. HART 1931 The changes occurring i n the ovary of the mare during pregnancy. Anat. Rec., vol. 49, p. 199. COLE,H. H., G. H. HART,W. R. LYONS,AND H. R. CATCHPOLE 1933 The development and hormonal content of fetal horse gonads. Anat. Rec., vol. 56, p. 275. FEVOLD, H. H., F. H. HISAW,AND S. L. LEONARD1932 Hormones of the corpus luteum. Separation and purification of three active substances. J. Am. Chem. Soc., vol. 54, p. 234. OESTRIN I N T H E PREGNANT MARE 347 GLUD, P., K. PEDERSEN-RJERGAARD, AND E. PORTMAN 1933 Uber Graviditatsreaktion bei der Stute. Endokrinologie, Jahrg. 13, S. 21. HALBAN, J. 1905 Die innere Secretion von Ovarium und Placenta, und ihre Bedeutung f u r die Funktion der Milchdruse. Arch. f. Gynak., Jahrg. 75, s. 353. NOVAK,E., AND A. K. KOFF 1930 The ovarian and pituitary changes associated with hydatidiform mole and chorionepithelioma. Am. J. Obstet. and Gynec., vol. 20, p. 481. PHILIPP, E. 1930 Hypophysenvorderlappen und Placenta. Zent. f . Gynak., Jahrg. 54, S. 450. DE SNOO,E. 1928 Chorionepitheliom der Tube. Hormonbildung vom isolierten Trophoblasten (Menformon). Zent. f. Gynak., Jahrg. 52, S. 2703. SZARKA, S. 1930 Hormcnale Verhaltnisse wahrend dcr Schwangerschaft bei Fehlen der Eierstocke. Zent. f . Gynak., Jahrg. 54, S. 2211. WALDSTEIN, E. 1929 Friihkastration in der Schwangerschaft. Zur Genese des Ovarialhormons. Zent. f. Gynak., Jahrg. 53, S. 1305. ZONDEK,B. 1931 Die Hormone des Ovariums und des Hypophysenvorderlappens. Berlin. Verlag von Julius Springer. 1930 Hormonal Schwangerschaftsreaktion aus dem Harn bei Mensch und Tier. Klin. Wchnschr., Jahrg. 9, S. 2285.