Uterine Peroxidase As a Marker for Estrogen Action (Steroid Action/Antiestrogen/Enzyme Induction) C

Home , Nafoxidine

Proc. Nati. Acad. Sci. USA Vol. 74, No. 8, pp. 3162-3166, August 1977 Biochemistry Uterine peroxidase as a marker for estrogen action (steroid action/antiestrogen/enzyme induction) C. RICHARD LYTTLE AND EUGENE R. DESOMBRE Ben May Laboratory for Cancer Research, The University of Chicago, Chicago, Illinois 60637 Communicated by Elwood V. Jensen, May 9,1977

ABSTRACT Administration of a single dose of estradiol to might be consistent with a role of "induced protein" as an early immature rats gives rise to the appearance of substantial key intermediate protein (15) in estrogen action, it would ob- amounts of peroxidase (donor:hydrogen-peroxide oxidoreductase, EC 1.11.1.7) enzyme activity in the uterus. This enzyme viously put expression of this protein prior to the observed in- induction, which is inhibited by administration of actinomycin creases in macromolecular synthesis involved in the actual D and cycloheximide, can be detected at 4 hr after adminis- growth response of the uterus. tration of estradiol, reaches a maximum level by 20 hr, and Increases in uterine peroxidase (donor:hydrogen-peroxide thereafter declines."The amount of uterine peroxidase seen at oxidoreductase, EC 1.11.1.7) content after estrogen adminis- 20 hr after a single dose increases with dose from 0.1 to 100 ;g tration were reported over 20 years ago by Stotz and coworkers of estradiol. Estrone and estriol also show dose-dependent induction of peroxidase, and the quantitative peroxidase responses (16) but attributed to an influx of eosinophils into the tissue (17). to these steroids follow their uterotropic capacities. The an- More recent evidence, both biochemical (18, 19) and histo- tiestrogen CI28, capable of low levels of enzyme induction by chemical (20-22), clearly indicates a uterine origin for the en- itself, can inhibit the induction due to estrogen. zyme. Although the capability of uterine peroxidase and H202 Solubilization of the uterine enzyme with divalent cations, to participate in the metabolism of ['4C]estradiol has been the especially calcium, results in a substantially increased yield of basis for one of the frequently used assays of uterine peroxidase peroxidase. This extraction method provides an enzyme of about 50,000 molecular weight-in distinction to the large aggregated (23), it has not been established that the enzyme actually par- form extraction with sodium chloride. ticipates in the uterine metabolism of estradiol in vivo. Histo- obtained by the usual chemical studies suggested that the estrogen-induced peroxidase The significant progress of the last 15 years toward a better accumulation occurred in only those tissues, such as uterus, understanding of the role of steroid hormones in controlling vagina, and carcinogen-induced mammry tumor (24), whose target cell functions has been possible to a large extent by the growth is effected by estrogen and not other estrogen target elucidation of the steroid-receptor interaction pathway (1-3), tissues, like the pituitary and hypothalamus, where the response coupled with investigations of biochemical parameters whose is primarily functional. We suggested therefore that peroxidase expression is regulated by these steroids (3-5). Studies on the induction may be a useful marker for estrogen-effected growth mechanism of estrogen action have used two main model sys- responses (22). tems. The atrophic rodent uterus, demonstrating dramatic This paper reports the development of a modified extraction growth response to estrogen, has been widely studied regarding procedure using calcium chloride which more effectively and the estrogen-receptor interaction pathway as correlated with completely solubilizes uterine peroxidase and shows that the tissue growth and control of RNA synthesis (4, 6, 7). The chick sensitivity of the guaiacol assay of the enzyme enables quanti- oviduct, after primary steroid stimulation to produce a differ- tation of the peroxidase activity of individual uteri of immature entiated, functional organ, responds to estrogen by producing rats. The quantitative data presented, both regarding the time large quantities of the specific protein, ovalbumin (3, 5). There course of estradiol induction of uterine peroxidase and the has been considerable increase in our knowledge of the nature sensitivity of uterine peroxidase to modulation by various es- of the hormone regulation of specific protein synthesis resulting trogens and antiestrogen, confirm that this enzyme is a useful from studies with the oviduct model in recent years (8-10). gene product for assessing effects of hormones on uterine However, in the uterine model no analogous specific gene growth. product, characteristically a marker for estrogen action, has been available to allow comparable, detailed studies in this more MATERIALS AND METHODS complex, mammalian steroid target organ. There have been Rats and Injection. Female Spragie-Dawley rats were numerous reports of uterine proteins, in particular enzymes, obtained from ARS/Sprague-Dawley, Madison WI at 20 days whose concentration are increased after estrogen administra- of age, allowed free access to food and water and, as immature tion. However, even with those showing significant increases rats, used by age 23 days. Castrate rats were of the same strain, in proportion in the uterus after estrogen stimulation, the extent subjected to bilateral ovariectomy under ether anesthesia when of the steroid induction is only 5- to 10-fold (11). Considerable 23-35 days old and used 1-3 weeks after surgery. Estradiol-173, interest has been shown in IP or "induced-protein," an acidic estrone, and estriol (Sigma Chemical Co.) and the antiestrogen uterine protein first described by Notides and Gorski (12). C1628, 1-$2-(pja(p-methoxyphenyl)-(-nitrostyryliphenoxy)- Recent studies have shown that this protein, inducible in vitro ethyll pyrrolidine, monocitrate, obtained by the courtesy of J. (13), is an early'response, detectible by 15 min, and declining Reel, Parke-Davis Co., were administered subcutaneously in in amount after 2 hr (14). While such a time course of induction saline, ethanol solution, or in sesame oil. Actinomycin D and cycloheximide were administered by intraperitoneal injec- The costs of publication of this article were defrayed in part by the tion. payment of page charges from funds made available to support the research which is the subject of the article. This article must therefore Extraction of Peroxidase Activity. At the indicated times be hereby marked "advertisement" in accordance with 18 U. S. C. the rats (four to eight per group) were sacrificed by decapita- §1734 solely to indicate this fact. tion. The entire uterus from oviduct to cervix was removed, 3162 Downloaded by guest on September 28, 2021 Biochemistry: Lyttle and DeSombre Proc. Natl. Acad. Sci. USA 74 (1977) 3163 Table- 1 Comparison of solubilization of uterine peroxidase activity with NaCi and CaCl2 Peroxidase Extractant activity* 0.6 M NaCl 1.4 1.2 M NaCl 1.1 0.5 M CaCl2 5.3 2.5 M CaCl2 5.9 * Peroxidase activity is given as enzyme units per ml of extract of 02 aliquots ofa minced pool of uteri of rats 20 hr after administration x of 5 1Ag of estradiol. Extracts of the 39,000 X g sediment corre- a.0 0.1 sponded to 50 mg of uterus per ml. (27)]. Investigating alternative solubilization con- 10 20 30 40 50 60 70 80 90 100 peroxidase Fraction Number ditions, we discovered that calcium chloride was a definitely FIG. 1. Gel chromatography of uterine peroxidase in 0.6 M NaCl superior extractant (Table 1). The total amount of enzyme extract. Uteri of mature rats, 20 hr after administration of 5 Mg of obtained by calcium chloride extraction of an aliquot of pooled estradiol, were homogenized and extracted with 0.6 M NaCl as de- rat uteri was considerably larger than thatextracted with so- scribed in Materials and Methods. The extract was applied to a dium chloride. Moreover, further extraction of the 1.2 M NaCI 180-ml Sephadex G-200 column packed in and eluted with 0.6 M NaCl pellets with either 0.5 M CaCl2 or 1.2 M NaCl did not provide in glycine buffer, pH 9, (1.5 ml per fraction). Vo indicates start of any significant additional amounts of peroxidase. It therefore elution of blue dextran marker in a separate elution. Peroxidase ac- a more tivity (0) is enzyme units per ml by the guaiacol assay; absorbance appeared that extraction with calcium chloride provided (0) is at 280 nm. active enzyme preparation. Gel filtration of the CaCl2 extract on Sephadex G-200 (Fig. 2) clearly indicated a different mo- quickly dissected free of fat and connective tissue, nicked, lecular size for the peroxidase activity. Now the enzyme was blotted on hard filter paper to excise luminal fluid, and separated from the majority of the 280 nm absorbing materials weighed. Each uterus was minced with a scissors into a 12-ml and eluted as a somewhat broad peak of activity in the 50,000 polycarbonate tube and homogenized, 10 sec with ice cooling molecular weight region. Peroxidase solubilization with calcium in 10 mM Tris-HCI, pH 7.2, buffer (25 mg of tissue per ml un- chloride shows a distinct concentration dependence (Fig. 3). less otherwise stated), with a polytron PT lOST homogenizer Little enzyme activity is solubilized with 0.2 M CaCl2 while at setting 6. After removal of aliquots for DNA analysis (25), essentially all the peroxidase appears in the 0.3 M extract. Such the homogenates were centrifuged for 45 min at 39,000 X g at a sharp extraction profile may suggest that a specific solubili- 20. The hypotonic supernatant fraction contains no peroxi- zation is occurring. dase activity. The peroxidase activity was solubilized by reho- Peroxidase Induction with Estradiol. The uterus of the mogenizng the 39,000 X g sediment with 10 mM Tris.HCI, pH immature or castrate rat has little detectable peroxidase activity. 7.2, buffer containing 0.6 or 1.2 M NaCl, or more effectively After a single dose of 10 gg of estradiol was given (Fig. 4), with 0.5 M CaC12. The extract was collected by centrifugation peroxidase was first detected at 4 hr, continued to increase in for 45 minat 39,000 X g at 20. amount until about 20 hr, and declined thereafter. A similar Peroxidase Assay. Assays were performed by modification time course is seen in both the castrate and immature rat and of the method of Himmelhoch et al. (26); the rate of oxidation after 1 ,Ag of estradiol is given, as well as after the 10-lg dose of guaiacol was measured at 25°. The assay mixture, 3.0 ml total shown in Fig. 4. The appreciable standard deviation seen near volume, contained 13 mM guaiacol and 0.3 mM H202 in the the peak induction times are believed to result from animal extraction buffer. The reaction was started by the addition of variability in the time at which the peak of activity occurs, since enzyme (extract), and initial rates of guaiacol oxidation were determined from the increase in absorbance at 470 nm for the 1.2- first minute. An enzyme unit was defined as the amount of 0

enzyme required to produce an increase of 1 absorbance unit/ .D 1.0 - 0 min under the assay conditions. I .0 4 .8- RESULTS Extraction of Uterine Peroxidase. Since rat uterine peroxidase activity is associated with the endoplasmic reticulum (21) and vesicles (19) and is not found in particulate-free cytosols of hypotonic homogenates, some extraction or solubilization method is necessary. The usual method, based on the work of Klebanoff (17), effected solubilization by extraction with NaCl. The resultant enzyme shows a very large molecular size, as indicated in Fig. 1, being excluded from Sephadex G-200 along 0 10 20 30 40 50 60 70 80 90 100 with the major part of the proteins of the NaCl extract. Fur- Fraction Number thermore, since the enzyme is also excluded from Sepharose 4B, FIG. 2. Gel chromatography of uterine peroxidase in 0.5 M CaCl2 extract. Conditions were as described in the legend of Fig. 1 except the approximate size of several million daltons would suggest that the 39,000 X g uterine sediment was extracted with 0.5 M CaCl2 an appreciably larger enzyme than has been reported for other in 10 mM Tris-HCl, pH 7.2, buffer which was also used as column el- mammalian peroxidase enzymes [for example, 150,000 mo- uant (2 ml per fraction). o, Enzyme units per ml; 0, absorbance at lecular weight for myeloperoxidase (26) and 300,000 for thyroid 280 nm. Downloaded by guest on September 28, 2021 3164 Biochemistry: Lyttle and DeSombre Proc. Nati. Acad. Sci. USA 74 (1977)

Table 2. Actinomycin D and cycloheximide inhibition of estradiol induction of uterine peroxidase activity Treatment, hr* Peroxidase 14.0- Act. D Cyclo. activityt 1.08 0.42 12.0 - 2 0.05 + 0.01 o 0.10 0.05 2 0.09 0.04 13 10.0 - 4 0.08 0.03 0 (0E._ 6 0.20 I 0.07 a-%0 0. .R8.0 - * Time at which inhibitors [150 ,gg of actinomycin (Act. D) or 250 jig of cycloheximide (Cyclo.), intraperitoneally] were given is hours after administration of 1 Isg of estradiol; uteri were assayed 20 hr 6.0 - after estradiol injection. t Peroxidase activity of calcium chloride extract (see Materials and Methods) is given as the mean enzyme units per mg of DNA + 4.0- SEM.

2.0 hr gives rise to some subsequent enzymatically assayable peroxidase. The amount of uterine peroxidase present 20 hr after estra- 0.4 0.8 1.2 1.6 2.0 2.4 2.8 diol administration is clearly dose dependent (Fig. 5). Although M Ca Cl2 the control animals had detectible activity, 0.1 sg of estradiol effected a significant increase in uterine peroxidase activity. FIG. 3. Extraction of uterine peroxidase with various concen- Interestingly, continued increases in peroxidase were seen with trations of calcium chloride. Aliquots of a pool of minced uteri from rats given 5 Mg of estradiol 20 hr prior to sacrifice were homogenized 1, 10, and 100 ,ug of estradiol. However, it would appear that and the 39,000 X g sediments (see Materials and Methods) were ex- this wide range dose response may be due to the fact that only tracted at 50 mg/ml with the various concentrations of calcium a single dose of estradiol was used. Indeed, on extended daily chloride in 10 mM Tris-HCl buffer pH 7.2. Peroxidase and protein injections, saturation of the response of uterine peroxidase oc- content (34) were assayed on the extracts. 0, Enzyme units per ml; curs at lower doses. 0, enzyme units per mg of protein. Peroxidase and Other Estrogens and Antiestrogen. Dose- dependent responses of uterine peroxidase are seen after in- assay replication errors are small compared to the differences jection of estrone and estriol also (Fig. 6). Consistent with the seen between individual animals. comparable effects of estrone and estradiol on uterine growth The estradiol induction of uterine peroxidase appears to re- and their relative affinities for the cytosol estrogen receptor, quire both RNA and protein synthesis (Table 2). Cycloheximide about 10 times as much estrone is required to obtain a compa- given 2 hr after 1 Mg of estradiol prevents the induction of en- rable peroxidase induction seen with estradiol. Estriol is zyme otherwise seen 20 hr after estrogen is given. Actinomycin D given the same time as or either 2 or 4 hr after estradiol also 100- prevented the appearance of peroxidase at 20 hr. When this inhibitor of RNA synthesis was given 6 hr after estradiol, the increase in peroxidase activity seen in the uterus 20 hr after 0 to estradiol would suggest that RNA produced during the first 6 u_ 80-

4- I- 60-

-,2- \ a-

20 0!

rw T Control 0.1 Ug 1.0 Jg 10 100 g 4 8 12 16 20 24 E2 E2 E2 E2 Hours After 10 pLg Estradiol FIG. 5. Dose dependence of estradiol induction of peroxidase in FIG. 4. Time course of uterine peroxidase induction. Calcium immature rat uteri. Animals injected subcutaneously with the indi- chloride extracts of individual uteri of immature rats (eight per time cated amount of estradiol (E2) in 0.2 ml of saline/ethanol or vehicle point) at various times after subcutaneous administrations of 10 Mg alone were sacrificed at 20 hr and the uteri were assayed for peroxidase of estradiol were assayed for peroxidase activity. activity. SEM are shown. Downloaded by guest on September 28, 2021 Biochemistry: Lyttle and DeSombre Proc. Natl. Acad. Sci. USA 74 (1977) 3165 Table 3. Antiestrogen blockade of estrogen induction of uterine peroxidase activity 4.0 - Treatment Peroxidase yg estradiol Mug CI628 activity* z 0.0 0.1 2.38 1.04 0 :0'1 1.0 29.0 9.6 E~3.0- 10.0 42.1 + 9.2

c._ 50 0.38 0.24 500 0.58 0.29 1.0 500 6.03 I 3.33 >, 2.0- 10.0 500 7.61 + 3.84 S._ * Peroxidase activity (mean enzyme units per gof tissue 1 SEM) of five rat uteri per group 20 hr after subcutaneous administration of estradiol and/or.CI628 in 0.2 ml of sesame oil, or of vehicle alone. 1.0-

0 Previous histochemical studies detected uterine peroxidase 12 I hr after, 200 ug of estradiol was given (20), but using more physiologic dose levels no peroxidase stain was evident 6 hr after Fi-1 1 1 1 estradiol; weak activity was seen 24 hr (22) after a single dose Control lg 1Og 10Pg I g 100 10 P9s E2 El El El E3 E3 E3 of estrogens. This would suggest that the histochemical stain FIG. 6. Comparison of estradiol (E2), estrone (El), and estriol for peroxidase is probably less sensitive than the biochemical (E3) induction of uterine peroxidase activity in the immature rat. assay by guaiacol oxidation. Probably a more serious defect Procedure is as described for Fig. 5, 20 hr after administration of associated with histochemical detection of peroxidase is its es- steroid or vehicle alone, but with results presented based on DNA sentially qualitative nature. Hence, previous studies detected assay of uterine homogenate (25). SEM as shown. C1628 induction of uterine peroxidase but could not differen- tiate peroxidase levels among animals given estrogen alone, somewhat less dose-effective, especially at the higher dose C1628 alone, or the two in competition. Our present results levels. It is noteworthy that appreciable uterine peroxidase is clearly indicate that although C1628 effects low levels of uterine induced by each of these estrogens. peroxidase induction such levels are significantly less than those Antiestrogens, like the Parke-Davis compound C1628 used seen after estrogen, and the combination of estrogen and C1628 here, frequently act as weak estrogens themselves. As shown gives intermediate amounts. On the other hand, the capability in Table 3, both 50 and 500 Aig of C1628 gave rise to small of the histochemical stain to identify individual cell types amounts of uterine peroxidase, each less than that seen after containing peroxidase provides a valuable kind of Information administration of 0.1 jug of estradiol. However, the simultaneous not readily available from guaiacol assay of uterine extracts. administration of C1628 with either or 1 or 10,gg of estradiol The assay measuring the conversion of [14Cjestradiol to water effected an appreciable reduction in the peroxidase levels soluble products, like the guaiacol assay, is sufficiently sensitive compared to animals given the respective doses of estradiol to detect increased peroxidase activity prior to 12 hr after es- alone. tradiol administration (18, 23). The higher baseline seen with the assay for water-soluble products and the maximum con- DISCUSSION version of about 60% of the estradiol tend to minimize the ap- A number of desirable qualities can be described for an optimal, parent extent of the stimulation of uterine peroxidase. The or at least useful, biologic marker for estrogen action in the evidence that actinomycin D and cycloheximide prevent the uterus. In the first place, sensitive methods should be available steroid-induced appearance of peroxidase confirms evidence for its detection and quantitation so that meaningful changes from previous histochemical (20) and biochemical (18) reports in marker levels in an individual uterus or even in individual that macromolecular synthesis is required for uterine expression uterine cells should be possible. Second, the marker should of the marker. More detailed studies will be needed to establish clearly be a uterine product that is absent, or alternatively the timing of required RNA and protein synthesis. present only in insignificant amounts, in the atrophic tissue. All the available evidence concerning uterine peroxidase Administration of estrogen should effect induction of the levels under various hormonal conditions is consistent with the marker in a dose-dependent manner. The half-life of the proposed marker function. In addition to the reported absence marker should be sufficient so that its degradation does not in the uterus of the immature or ovariectomized host and its complicate measurement of the induction. Third, modulation induction with estradiol, the variation with the estrus cycle (18, of marker expression should follow closely the regulation of 20, 22) in rats as well as in human beings (28) has been docu- uterine growth by various estrogens, steroids, and antagonists. mented. In this report we have shown that the relationship of It would furthermore be advantageous if the marker were a well efficacies of estradiol, estrone, and estriol for induction of characterized product, or if not, that the marker were present uterine peroxidase relates to their previously documented in sufficient amounts in the target organ to allow purification uterotropic activities (29) and with each a dose dependency is of the marker and isolation and quantitation of its mRNA. observed. Furthermore the quantitative peroxidase results using To an appreciable extent peroxidase activity would appear the antiestrogen, CI628, indicate, as previously suggested, that, to be an appropriate marker for estrogen action in the uterus. while weakly estrogenic itself (30), C1628 is inhibitory to es- As presented in this paper, using the guaiacol assay the enzyme trogen action when given with estradiol (31, 32). Another an- is readily quantitated in individual uteri with a sensitivity that tiestrogen, nafoxidine or U 11100A, has been recently reported allows its detection as early as 4 hr after estrogen administration. by McNabb and Jmlinck (33) to show a similar effect on uterine Downloaded by guest on September 28, 2021 3166 Biochemistry: Lyttle and DeSombre Proc. Nati. Acad. Sci. USA 74 (1977)

peroxidase activity as determined by the assay for water-soluble 14. DeAngelo, A. B. & Gorski, J. (1970) Proc. Nati. Acad. Sci. USA product. The nonsteroidal estrogen diethylstilbestrol is highly 66,693-700. potent for uterine peroxidase while substantial 15. Baulieu, E. E., Wira, C. R., Milgrom, E. & Raynaud-Jammet, C. induction, (1972) in Fifth Karolinska Symposium on Research Methods amounts progesterone and testosterone are ineffective of in Reproductive Endocrinology, ed. Diczfalusy, E. (Karolinska (22). Institute, Stockholm), pp. 396-415. More detailed studies regarding the macromolecular nature 16. Lucas, F. V., Neufeld, H. A., Utterback, J. G., Martin, A. P. & of the hormonal regulation of the uterine peroxidase gene would Stotz, E. (1955) J. Biol. Chem. 214, 775-780. be enhanced by isolation of purified uterine peroxidase and its 17. Klebanoff, S. J. (1965) Endocrinology 76,301-311. mRNA. As suggested by the magnitude of the peroxidase in- 18. Jellinck, P. H. & Lyttle, C. R. (1972) Adv. Enzyme Regul. 11, duction, the estrogen-stimulated rat uterus is an appropriate 17-3. source for purification of this enzyme marker. 19. jellinck, P. H., McNabb, T., Cleveland, S. & Lyttle, C. R. (1976) Adv. Enzyme Regul. 14, 447-465. We thank Richard Garay and R. Dizon for excellent technical as- 20. Brokelmann, J., & Fawcett, D. W. (1969) Biol. Reprod. 1, 59- sistance. This work was supported by The Breast Cancer Task Force, 71. U.S. Public Health Service, Contract 1-CB-23873, and the Population 21. Churg, A. & Anderson, W. A. (1974) J. Cell Biol. 62,449-459. Council, Grant M75-18C. 22. Anderson, W. A., Kang, Y-H. & DeSombre, E. R. (1975) J. Cell 1. Jensen, E. V. & DeSombre, E. R. (1973) Science 182, 126-134. Biol. 64, 668-681. 2. Liao, S. (1975) Ipit. Rev. Cytol. 41, 87-172. 23. Lyttle, C. R. & Jellinck, P. H. (1972) Biochem. J. 127, 481- 3. O'Malley, B. W. & Means, A. R. (1974) Science 183,610-620. 487. 4. Jensen, E. V., Mobla, S., Gorell, T. A. & DeSombre, E. R. (1974) 24. DeSombre, E. R., Anderson, W. A. & Kang, Y-H. (1975) Cancer Vitam. Horm. (N.Y.) 32,89-127. Res. 35, 172-179. 5. Palmiter, R. D., Christensen, A. K. & Schimke, R. T. (1970) J. 25. Burton, K. (1956) Biochem. J. 62, 315-323. Biol. Chem. 245,833-845. 26. Himmelhoch, S. R., Evans, W. H., Mage, M. G. & Peterson, E. 6. Hamilton,, T. H., Widnell, C. C. & Tata, J. R. (1968) J. Biol. A. (1969) Biochemistry 8, 914-921. Chem. 243,408-417. 27. Hosoya, T. & Morrison, M. (1967) J. Biol. Chem. 242, 2828- 7. Glasser, S. R., Chytil, F. & Spelsberg, T. C. (1972) Biochem. J. 2836. 130,947-957. 28. Lucas, E. V., Carnes, V. M., Schmidt, H. J., Sipes, D. R. & Hall, 8. McKnight, G. S., Pennequin, P. & Schimke, R. T. (1975) J. Biol. D. B. (1964) Am. J. Obstet. Gynecol. 88,965-967. Chem. 250, 8105-8110. 29. Huggins, C., Jensen, E. V. & Cleveland, A. S. (1954) J. Exp. Med. 9. Tsai, S. Y., Tsai, M. J., Schwartz, R., Kalimi, M., Clark, J. H. & 100,225-240. O'Malley, B. W. (1975) Proc. Nati. Acad. Sci. USA 72, 4228- 30. Clark, J. H., Peck, E. J., Jr. & Anderson, J. N. (1974) Nature 251, 4232. 446-448. 10. Harris, S. E., Rosen, J. M., Means, A. R. & O'Malley, B. W. (1975) 31. Callantine, M. R., Humphrey, R. R., Lee, L. L., Windsor, B. L., Biochemistry 14, 2072-2081. Schottin, N. H. & O'Brien, 0. P. (1966) Endocrinology 79, 11. Baquer, N. Z. & McLean, R (1972) Biochem. Biophys. Res. 153-167. Commun. 48,729-734. 32. Jensen, E. V., Jacobson, H. I., Smith, S., Jungblut, P. W. & De- 12. Notides, A. & Gorski, J. (1966) Proc. Natl. Acad. Sci. USA 56, Sombre, E. R. (1972) Gynecol. Invest. 3,108-123. 230-235. 33. McNabb, T. & Jellinck, P. H. (1976) Steroids 27,681-689. 13. Katzenellenbogen, B. S. & Gorski, J. (1972) J. Biol. Chem. 247, 34. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. 1299-1305. (1951) J. Biol. Chem. 193,265-275. Downloaded by guest on September 28, 2021