Vaginal Sensitivity to Estrogen as Related to Mammary Tumor Incidence in Mice J.J. TRENTIN,PH.D.* (From the Department of Anatomy, Yale Unirersity School of Medicine, New Hauen, Connecticut) The demonstration of the importance of ovarian to estrogen bears no relation to the maternal ex- secretion or exogenous estrogen to a high mam trachromosomal factor. mary tumor incidence in mice was followed by Mühlbock(2,3), from the same laboratory, con numerous studies of the estrous cycles of mice of firmed the higher estrogen requirement (5-7 times different strains, in an attempt to correlate high as much) of the high tumor dba strain as compared mammary tumor incidence with some peculiarity to the low tumor C57 and 020 strains for a com of the cycle. In general, no significant and consist parable degree of vaginal stimulation by either in- ent correlation could be demonstrated in this travaginal or subcutaneous administration of regard. Korteweg and co-workers, however, fo estrogen. cused attention on the possibility that, whereas Sliimkin and Andervont (4) also reported on the outward manifestations of the cycle might be rela vaginal estrogen-sensitivity of three strains of mice tively constant, fundamental differences may exist of known mammary tumor incidence—the C57, in the sensitivity of the genital tissues to estrogen, C, and C3H strains. Again the high tumor strain and that such differences may be related to the was more resistant, the C3H strain requiring twice mammary tumor incidence. Van Gulik and Korte as much estrogen as the low tumor C57 and C, weg (5) reported that of one high tumor and two strains for a positive response in 50 per cent of the low tumor inbred strains, compared with regard to mice. It was also reported that, whereas foster vaginal estrogen-sensitivity, the high tumor dba nursing of the high tumor strain by a low tumor strain was the most resistant to exogenous estrogen strain, or vice versa, altered the mammary tumor stimulus, requiring roughly 3 times the amount of incidence, it did not alter vaginal sensitivity to estrogen needed by the low tumor €¿57strainfor estrogen. a comparable degree of vaginal stimulation. They The present investigation was undertaken to pointed out also that Bonser (1) observed a some determine whether the reported inverse relation what higher percentage of positive vaginal smears ship of vaginal estrogen-sensitivity and mammary in a low mammary tumor strain than in a high tumor incidence is of universal occurrence or mammary tumor strain when both were injected restricted to the few inbred strains of mice studied with equal but varying doses of estrogen. On the thus far. In view of the apparent lack of effect of basis of such findings, they theorized that, since a milk factor on vaginal estrogen-sensitivity, it normal estrous cycle is essential for continued ex would seem improbable that such an inverse rela istence, those strains of low vaginal estrogen-sen tionship could apply to all strains of mice, since sitivity probably produce larger amounts of estro some strains may have a low mammary tumor in gen and that this may be causally related to a high cidence only because of the absence of the milk mammary tumor incidence. They reported also factor. Nevertheless, such an inverse relationship that reciprocal hybrids between the dba and C.57 might conceivably apply when comparison is made strains showed an almost equal vaginal sensitivity only between strains known to possess the milk to estrogen, intermediate between that of the pa factor. rental strains, indicating that vaginal susceptibility MATERIALS AND METHODS «Fellowof The Jane Coffin Childs Memorial Fund for Included in the present study are mice of three Medical Research. This investigation has been aided by a grant strains of high mammary tumor incidence, the from The Jane Coffin Childs Memorial Fund for Medical Re C3H, CBA, and A strains, and three of low tumor search. incidence, the C57, JK, and \ strains. In addition, Received for publication June 9. 1950. mice of two high tumor strains free of milk factor 580 Downloaded from cancerres.aacrjournals.org on September 28, 2021. © 1950 American Association for Cancer Research. TRENTIN—Vaginal Estrogen Sensitivity and Mammary Tumor Incidence 581 as a result of foster nursing are included. These are with the reservation that it may possess a post- the C3H and A strains, designated, respectively, castration extra-ovarian source of estrogen, in the BL and AL. Also studied were three groups of Fl absence of which responsiveness to exogenous hybrids, the CCI (C57 9 X CBA <f), the CC2 estrogen might have been lower. (CBA 9 X C57 (?), and the AB2 (CBA 9 X A single injection of estradici benzoate2 in 0.0.5 Ad1).1 cc. of sesame oil was used. Smears were taken by The mammary tumor incidence of breeding lavage, and only a fully cornified smear with no females of the inbred strains is presented in Table more than occasional leukocytes was considered 1. These data arc compiled largely from Dr. positive. The mice were primed with 0.1 /¿gof TABLE 1 estradici benzoate, and assays were performed at .1-2-week intervals thereafter, depending on the INCIDENCEOFMAMMARYADENOCARCINOMAINBREEDING amount of the previous injection and the length of FEMALEMICE LIVING150 DAYSOR MORE time required for a return to a negative smear. A centwithmammarytumors8387580 Per cent with graded series of estrogen dosages were adminis ofStrain No. No. of mammary miceC3H Strain mire tumors tered, and different dose levels repeated as often 369CBA JK 37 0 as necessary to obtain a fairly smooth sigmoid 154A Al 50 0 type dose-response curve for each strain. 156C57 BL Õ5K 62* 138Per The dams of the mice used for vaginal assay * In addition to the above-mentioned adenocareinonjas in tile BLstrain, were kept to determine mammary tumor inci seventeen mice, or 6.7 percent, had at autopsy one or more small nodular dence. Two of the BL dams developed mammary foci of .squamous metaplasia of the mammary glands. Four more animals, or 1.6 per cent, had squamous tumors of 4-15 mm. diameter. The two smaller tumors—one an adenocarcinoma and the other a tumors were adenomas containing squamous cysts. The largest was a hard, white, granular, rapidly growing squamous-cell carcinoma with scattered squamous-cell carcinoma with small scattered smallarcas of adcnocarcinoma. The fourth was a 1-cm. hiird, white, granular tumor mass, ou which sections are not available. areas of adenocarcinoma. Progeny of both of these mice were eliminated from the experimental series. Gardner's colony over the period of approximately The incidence of mammary adenocarcinomas the past ten years and from the relatively fewer in the BL strain, since it was obtained from Dr. animals of the author's colony during the past two Bittner in 1941, has been 6.2 per cent. These tu years. Both colonies were maintained under com mors have appeared relatively more frequently in parable conditions. Breeding mice received Purina recent years, raising some question as to the com laboratory chow, and nonbreeding experimental plete absence of milk factor in this strain at the animals received Purina Fox Chow. Adequate present time. The data for the BL strain are never numbers of animals are not available to report theless presented, since the tumor incidence is still mammary tumor incidence for the N strain under much below that of the C3H strain, and progeny comparable conditions. However, mammary tu of tumor-bearing animals have been eliminated. mors are rare in mice of this strain in Dr. Strong's colony, and none have been observed in the few RESULTS AND DISCUSSION breeding females used for the present experiment. The percentage of positive vaginal smears ob Groups of approximately twenty virgin female tained from mice of the different strains at the mice of each of the stocks mentioned previously various levels of estrogen are presented in Table 2. were ovariectomized at approximately 1 year of The dose-response curves for the different strains age (8-14 months range), allowed to rest for .5-7 are presented graphically in Figures 1, 2, 3, and 4. weeks, then smeared 4 times on alternate days. In Table 3 lists the different strains in decreasing the JK strain many of these smears contained, in order of vaginal sensitivity, as measured by the addition to leukocytes, numerous small nucleated minimum dose of estrogen required for a positive epithelial cells, indicating a possible extra-ovarian response in approximately .50per cent of the mice. source of estrogen. After estrogen administration, The dose-response curves of the three high tu the smears became fully cornified, followed by a mor and the three low tumor inbred strains do not completely diestrous type of smear. It is probable, affirm the reported inverse relationship of vaginal therefore, that during the course of the assay, in estrogen-sensitivity and mammary tumor inci volving repeated estrogen injections, the amount dence. The range of sensitivity represented by the of endogenous estrogen was negligible. Neverthe two groups of curves overlaps considerably. It is less, the JK strain is included in the present report apparent why an inverse relationship might be 1The BL and AL strains were obtained from Dr. W. I'. indicated in the comparison of a smaller number Gardner who, in turn, received them from Dr. J. J. Bittner in of strains, one of which is the low tumor C57 1941 and 1944, respectively. The remaining inbred strains were 2The estradici benzoate was generously provided by the obtained from Dr. (¡ardnerand Dr. L. C. Strong. Schering Corporation. Downloaded from cancerres.aacrjournals.org on September 28, 2021. © 1950 American Association for Cancer Research.