Proc. Nail. Acad. Sci. USA Vol. 83, pp. 3253-3256, May 1986 Biochemistry Specific binding of human corticosteroid-binding globulin to cell membranes (plasma steroid hormone-binding proteins/membrane receptors/steroid hormone action/prostate) DANIEL J. HRYB*, M. SAEED KHANt, NICHOLAS A. ROMAS*, AND WILLIAM ROSNERtt Departments of *Urology and tMedicine, St. Luke's-Roosevelt Hospital Center, and Columbia University, College of Physicians and Surgeons, 428 West 59th Street, New York, NY 10019 Communicated by Seymour Lieberman, January 21, 1986

ABSTRACT Specific binding sites for corticosteroid-bind- All of the foregoing observations have appeared in a ing globulin were detected on membranes prepared from fragmented way, and no cohesive focus has forced their human prostates. The binding sites are typical of membrane assembly into a meaningful physiological model. We felt that receptors: they are saturable and specific and have high ifthey all were both correct and ofphysiological importance, affinity. There was little specific binding at 40C and 230C. then a membrane receptor for CBG should exist. The Maximal specific binding was obtained at 370C. Scatchard existence of such a receptor would provide both a specific analysis revealed the presence of a single set of binding sites mechanism whereby CBG could be internalized into cells via with an apparent dissociation constant of 8.7 x 10-7 M and a receptor-mediated endocytosis, and a center around which to binding capacity of 22 pmol/mg of membrane protein. The bring together the aforementioned disparate observations. sites were specific for corticosteroid-binding globulin; binding We therefore set out to determine whether there are specific was not inhibited by human testosterone/estradiol-binding cell membrane binding sites for CBG and found that CBG globulin, by albumin, or by transferrin. The density of specific binds saturably and specifically to human prostatic mem- binding sites in membranes obtained from several organs from branes, as well as to membranes from other tissues. A the rhesus monkey is consistent with the hypothesis that preliminary report of some aspects of this study has been corticosteroid-binding globulin is involved in the transport of presented.§ steroid hormones into target tissues. MATERIALS AND METHODS The steroid hormones cortisol and progesterone are secreted Prostatic tissue was obtained from men undergoing transure- into the circulation by the adrenal cortex and ovary, respec- thral prostatectomy (prostate < 65 g) or open prostatectomy tively. They are not in simple solution in the plasma but rather (prostate > 65 g) for benign prostatic hyperplasia and, on two are largely bound to the same trace protein, corticosteroid- occasions, from open prostatectomy on men undergoing binding globulin (CBG) (1-3). The current model of steroid surgery for bladder cancer. Plasma membranes (100,000 x g hormone action envisions that the free (i.e., non-protein- pellet) were prepared by differential centrifugation (21). The bound) steroids diffuse passively across the capillary and cell final 100,000 x g pellet (prostatic membranes) was resus- membranes to the cell's interior where they initiate hormone pended in 1 volume of 25 mM Tris Cl, pH 7.4/10 mM action (4). CaCl2/0.2% bovine serum albumin (buffer A) and either used Over the years, evidence has appeared that indicates it for binding experiments immediately or stored at -700C. might be necessary to modify this model. None of it, Results were the same whether membranes were prepared however, has been either sufficiently convincing or error-free from fresh tissue or from tissue frozen for 2-4 weeks. to upset the free-hormone model. For instance, Rosner and Membranes (100,000 x g fraction) were also prepared from Hochberg (5) provided biochemical evidence years ago that the kidneys, liver, spleen, and uterus of a single adult female CBG-bound cortisol was as active, or possibly more active, rhesus monkey (Macaca mulatta) which, as part of a differ- than free cortisol in the induction of tyrosine aminotrans- ent protocol, had been treated with estradiol valerate (2 ferase activity in rat liver. Others provided biochemical mg/day for 10 days). The rhesus monkey was anesthetized evidence for the presence of CBG or a CBG-like binder in with sodium barbital, after which the desired organs were various glucocorticoid and progesterone target organs: pitu- removed. Membranes were prepared as above. Membrane itary (6, 7), kidney (8), uterus (9-11), muscle (12), lung (13), protein was determined by washing an aliquot of the mem- breast cancer (14), and lymphocytes (15). A CBG-like mol- branes with 0.3 M sucrose and then treating it with 0.1 M ecule has also been identified in plasma-membrane fractions NaOH at 70'C for 30 min. The membrane solution was then from liver (16) and pituitary cells (17). Some of the biochem- centrifuged at 16,000 x g and the protein content of the ical demonstrations have been supported by the immuno- supernatant was ascertained (22). cytochemical visualization of intracellular CBG in liver (18), CBG, containing 1 mol of cortisol per mol of CBG, was uterus (2), kidney (2), and lymphocytes (15); but some of purified from pregnancy plasma by affinity and hydroxylap- these reports have been flawed by the omission of critical atite chromatography (23, 24). Its purity was verified by controls and/or a lack ofevidence that the observations were polyacrylamide gel electrophoresis, both under nondenatur- not due to contamination with plasma (19). More recently, ing conditions and in the presence of NaDodSO4, and by however, in a well-controlled study, guinea pig CBG has been immunoelectrophoresis against antiserum to whole human specifically localized immunohistochemically inside those pituitary cells (corticotrophs) which secrete corticotropin (20). Abbreviations: CBG, corticosteroid-binding globulin; 1251I-CBG, 125I_ labeled CBG; TEBG, testosterone/estradiol-binding globulin. tTo whom reprint requests should be addressed. The publication costs of this article were defrayed in part by page charge §Hryb, D. J., Khan, M. S., Romas, N. & Rosner, W., the 67th payment. This article must therefore be hereby marked "advertisement" Annual Meeting of. the Endocrine Society, June 19-21, 1985, in accordance with 18 U.S.C. §1734 solely to indicate this fact. Baltimore, MD, abstr. 244. 3253 Downloaded by guest on September 30, 2021 3254 Biochemistry: Hryb et al. Proc. Natl. Acad. Sci. USA 83 (1986) serum. It was iodinated, using chloramine-T, to a specific indicates that there was minimal binding at 4°C or at 23°C. activity of0.1-0.3 mol of'251 per mol of CBG (24). To be sure Binding increased with time at 37°C and began to plateau at6 hr, that the iodinated CBG was native, over 90% was shown to which was chosen as the incubation time for subsequent be specifically bound by a monospecific anti-CBG antiserum experiments. These data are shown again in the main figure, before using it in receptor binding experiments (data not where total and nonspecific binding are also indicated. shown). Human testosterone/estradiol-binding globulin Fig. 2 is a representative Scatchard analysis (28) of the (TEBG) and albumin were isolated as described (25-27). specific binding of CBG to prostatic membranes at 37°C. The Human transferrin was purchased from Sigma. data are fit best by a straight line and indicate that CBG is Binding took place in 1.5-ml Eppendorf polypropylene bound to a single set of sites. The inset is a saturation plot centrifuge tubes that were pretreated with 1.0 ml of buffer A using the same data as was used for the Scatchard analysis. for 1 hr at 370C to decrease test tube blanks. 125I-labeled CBG Using the computer program LIGAND (29), we analyzed (125I-CBG, 250,000-300,000 cpm per tube, about 2 nM), plus four experiments simultaneously. The analysis yielded an or minus 14 ,M radioinert CBG, was incubated with 100-200 association constant (Ka) of 1.16 ± 0.26 (SEM) x 106 M-1 and ,ug of prostatic membrane protein in 25 mM Tris Cl, pH a binding capacity of 22 pmol/mg of membrane protein. 7.4/25 mM CaCl2/0.2% bovine serum albumin, in a final To determine the ligand specificity of the receptor, the volume of 0.25 ml. All assays were done in triplicate, and ability ofradioinert CBG to compete with the specific binding membrane-bound 125I-CBG was separated from free by of '25I-CBG to prostatic membranes was compared with that centrifugation for 10 min at 40C. After the supernatants were of TEBG and two other human plasma proteins, transferrin aspirated, the tip ofeach tube was cut 0.5 cm from the bottom and albumin (Fig. 3). None of these proteins significantly and the 1251I therein was measured in a gamma counter at an inhibited specific CBG binding to the membranes, attesting to efficiency of 72%. Test tubes without membranes served as the specificity of the CBG-membrane interaction. blanks. Specific binding was calculated as the difference Human and rhesus monkey (Macaca mulatta) CBGs are between total binding and nonspecific binding, which was immunologically very similar (30). Therefore, in order to that binding seen in the presence of 14 AM radioinert CBG. study the tissue distribution of specific CBG binding, and in Generally, about 2% of the added 125I-CBG was specifically the absence of the ready availability of fresh human tissues, bound. There were no statistically significant differences in we examined CBG binding in membranes freshly prepared binding between membranes prepared from prostates ob- from the kidney, spleen, liver, and uterus ofa rhesus monkey tained by either transurethral or open prostatectomy. (Fig. 4). All four tissues examined bound substantially more CBG than did human prostate. RESULTS DISCUSSION The dependence of the binding of 125I-CBG to human prostatic membranes as a function oftime and temperature is The studies reported here demonstrate that human prostatic shown in Fig. 1. The inset, which shows only specific binding, membranes contain CBG binding sites with characteristics that are typical of membrane receptors. The sites are satu- 5] 37°C rable and specific and have high affinity. The studies of specific CBG binding sites in membranes from various organs ._ _ of the rhesus monkey indicate that the kidney, spleen, liver, and uterus bind substantially more CBG than does the human C E 1 23°C prostate. Kidney, spleen, and liver are important target 4°C

4

x E E 1.5 0. u 0 0

.0 -6.0 S~~~~~~~~~O\ X \O0 CBG added, nmol 41) m~~oI\~ 0.50 4 Time, hr 0

FIG. 1. Time-course of binding of I251-CBG to prostatic mem- branes. 1251-CBG (300,000 cpm, about 2 nM) was added to prostatic membranes (0.4 mg of protein per ml), with or without radioinert 0 CBG (14 ,uM), and incubated at 37°C, 23°C, or 4°C. The main figure 5 10 15 shows total (o) and nonspecific (A) as well as specific (o) binding at Bound, nM 37°C. Generally, about 2% of the added 125I-CBG bound specifically to the prostatic membranes during 6 hr ofincubation at 37°C. Specific FIG. 2. Scatchard analysis. 1251I-CBG (300,000 cpm, 2 nM) was binding is the difference between total binding (2-3% of added incubated with prostatic membranes (0.4 mg/ml) in the presence of 125I-CBG at 370C) and nonspecific binding [binding in the presence of various amounts ofradioinert CBG (0-7 ,uM) at 37°C for 6 hr. (Inset) a 7000-fold excess of radioinert CBG (about 25% of total binding)]. A saturation plot of specific CBG binding. Both plots are corrected (Inset) Specific binding only. for nonspecific binding in the presence of 14 ,uM radioinert CBG. Downloaded by guest on September 30, 2021 Biochemistry: Hryb et A Proc. Natl. Acad. Sci. USA 83 (1986) 3255

A6 a A Z '

0 m- 0 A

0

Competitor, M FIG. 3. Competition by human CBG, TEBG, transferrin, and albumin for 125I-CBG binding sites on prostatic membranes. 125I-CBG was incubated with prostatic membranes in the presence of various amounts of radioinert human CBG (A), human TEBG (A), human transferrin (0), and human albumin (o). Bo, specific binding of '25I-CBG in the absence of competitor; B, specific binding of 251I-CBG in the presence of competitor. tissues for cortisol, whereas the uterus is a major target tissue does cortisol (1)]. Although the prostate is not considered to for progesterone [progesterone binds as tightly to CBG as be a major target organ for cortisol or progesterone, its stroma, which constitutes about 50% of the gland, contains 3or elements such as connective tissue, blood vessels, and Kidney smooth muscle which are well known to be responsive to . . . glucocorticoids. Hence, these data are compatible with the * * o Total * * o hypothesis that CBG binding to membranes is involved with . . . Sploleen cortisol and progesterone action. **..o ,... C *o ** There is a previous report of a putative CBG receptor on Nonspecific ** **@ *@ *-e human liver membranes (31). The data in that report showed 0. *@* ** of to ** *** specific binding 125I-labeled asialo-CBG human hepatic 0 Specific ** *** Liver membranes. However, native CBG was almost four orders of o ** * 20 *v- *- II* * o magnitude less effective than asialo-CBG itself in the dis- *** ** *o S. *** ** *-@ placement of 125I-labeled asialo-CBG (31). Thus, the ob- *** * *-@ served can to 0 ** ** **@ binding activity be attributed most reasonably * *@ the o *@* * ...***-Jterus interaction of125-labeled asialo-CBG with the well-known * * ... .., liver membrane asialoglycoprotein receptor (32). x * * 44... .., Displace- E ** * *** *e ment of the 125I-labeled asialo-CBG could easily have been 0. * *- ... ., * * * * ** due to a trace (<0.1%) of asialo-CBG contaminating the ** ** ... ., 0 * *** * * * * * purified CBG used as competitor. * ** ... .., * *@ *eX * We recently reported (21) that membranes from human D 10 I-_ * *@ *v * prostate contain specific binding sites for another plasma ** *v I *** * I - 1 * *v ** * steroid-binding protein, TEBG. It is important to be aware of 0 * *v I ** * * ** I *- * the fact that the apparent dissociation constants (Kd) of the ** ** **- ** ** *X ** ** CBG and TEBG receptors [8.7 x 10-7 M and 1.5 x 10-8 M * *- ** * * *v *** * (21), respectively] are close to the average concentration of * * *** ** * * ** * CBG and TEBG in male plasma. This relationship dictates .. ...1 ** that the density of CBG and TEBG on the surface of target _* cells (i.e., that which is receptor-bound), and hence their _ _ biological effects, would be altered by changes in either their Human Rhesus monkey plasma concentration or the concentration oftheir respective receptors. The system is clearly designed to be responsive to FIG. 4. Distribution of '251-CBG binding to membranes prepared both steroid-binding protein and to receptor. from several organs of the rhesus monkey. Membranes were pre- We have documented the existence of saturable and pared from human prostate and the kidney, spleen, liver, and uterus specific binding sites for CBG in membranes derived from the of a single female rhesus monkey. Membranes (0.2-0.4 mg/ml) were human prostate and from target organs for cortisol and incubated with about 300,000 cpm of 125I-CBG in the presence progesterone from the rhesus monkey. We previously dem- (nonspecific binding) or absence (total binding) of 14 JIM radioinert onstrated receptors for TEBG in human prostatic membranes CBG. Data shown for prostate represent means ± SD from nine separate experiments using membranes prepared from a single (21). Taken together, our results on the CBG and TEBG human prostate and three different batches of 1251-CBG. (The height receptors, along with the recently published report of a of the bars, for the rhesus monkey, represents total binding.) TEBG receptor in human decidual endometrium membrane Downloaded by guest on September 30, 2021 3256 Biochemistry: Hryb et al. Proc. Natl. Acad. Sci. USA 83 (1986) (33), raise the possibility that CBG and TEBG may serve to 10. Rosenthal, H. E., Ann-Paul, M. & Sandberg, A. A. (1974) J. transport steroid hormones into target tissues via specific Steroid Biochem. 5, 219-225. membrane receptors. Since it is generally accepted that the 11. Al-Khouri, H. & Greenstein, B. D. (1980) Nature (London) specificity of receptor-mediated endocytosis resides exclu- 287, 58-60. 12. Mayer, M., Kaiser, N., Milholland, R. J. & Rosen, F. (1975) J. sively in the initial binding step between a ligand and its Biol. Chem. 250, 1207-1211. receptor (34), this hypothesis both supports and is supported 13. Giannopoulos, G. (1976) J. Steroid Biochem. 7, 553-558. by previous reports indicating the intracellular presence of 14. Amaral, L. & Werthamer, S. (1976) Nature (London) 262, CBG in various cortisol and progresterone target cells (2, 589-590. 6-15, 18, 20); the presence of CBG-like molecules on mem- 15. Werthamer, S., Samuels, A. J. & Amaral, L. (1973) J. Biol. branes from rat pituitary (17) and rat liver cells (16); the Chem. 248, 6398-6407. intracellular presence of TEBG in various androgen and 16. Suyemitsu, T. & Terayama, H. (1975) Endocrinology 96, estrogen target cells (35-39); and the observation that cortisol 1499-1508. bound to CBG is physiologically active (5, 40). However, 17. Koch, B., Lutz-Bucher, B., Briaud, B. & Mialhe, C. (1978) J. Endocrinol. 79, 215-222. direct evidence for the above hypothesis is not yet at hand, 18. Perrot-Applanat, M., David-Ferreira, J. F. & David-Ferreira, and it is possible that the binding of CBG to its receptor may K. L. (1981) Endocrinology 109, 1625-1633. subserve an as yet unknown function that is unrelated to the 19. Ballard, P. L. (1979) in Glucocorticoid Hormone Action, eds. transport of steroids into target cells. Baxter, J. D. & Rousseau, G. G. (Springer, New York), pp. Finally, the classic, two-step model of steroid hormone 25-48. action has been questioned. The classic model posits a 20. Perrot-Applanat, M., Racadot, 0. & Milgrom, E. (1984) En- cytosolic receptor for the steroid, whose association constant docrinology 115, 559-569. ensures that there is essentially no unbound steroid in the cell 21. Hryb, D. J., Khan, M. S. & Rosner, W. (1985) Biochem. cytoplasm. Thus, in addition to transporting the steroid into Biophys. Res. Commun. 128, 432-440. 22. Waddell, W. J. (1956) J. Lab. Clin. Med. 47, 311-315. the nucleus, the receptor creates a steep concentration 23. Rosner, W. & Bradlow, H. L. (1975) Methods Enzymol. 36, gradient down which the plasma free hormone can flow into 104-109. the cytoplasm by simple diffusion (4). Recent evidence 24. Rosner, W., Polimeni, S. & Khan, M. S. (1983) Clin. Chem. indicates that, at least for progesterone (41, 42) and estrogens 29, 1389-1391. (43-46), the cytosolic receptor may be an artifact of cell 25. Rosner, W. & Smith, R. N. (1975) Biochemistry 14, disruption and that all of the receptor is nuclear. Therefore, 4813-4820. we might be left without a mechanism to create a large 26. Khan, M. S., Ewen, E. & Rosner, W. (1982) J. Clin. positive steroid concentration gradient between the cell Endocrinol. Metab. 54, 705-710. and the such a is 27. Smithies, 0. (1955) Biochem. J. 61, 629-634. cytoplasm plasma. Obviously, gradient not 28. Scatchard, G. (1949) Ann. N.Y. Acad. Sci. 51, 660-672. a necessary condition for steroid hormone action. The effects 29. Munson, P. T. & Rodbard, D. (1980) Anal. Biochem. 107, of steroid hormones could still proceed by the passive 220-239. diffusion of steroids through the plasma and nuclear mem- 30. Rosner, W., Pugeat, M. M., Chrousos, G. P. & Khan, M. S. branes and into the nucleus. However, concurrent with a (1986) Endocrinology 118, 513-517. small, steady diffusion of free steroid hormones into cells, 31. Strel'chyonok, 0. A. & Avvakumov, G. V. (1983) Biochim. steroids could be augmented in target cells, as needed, by the Biophys. Acta 755, 514-517. internalization of bound steroids via membrane receptors for 32. Ashwell, G. & Harford, J. (1982) Annu. Rev. Biochem. 51, CBG and TEBG. Thus, the active transport of steroid 531-554. 33. Strel'chyonok, 0. A., Avvakumov, G. V. & Survilo, L. I. hormones into target cells by CBG and TEBG by way of (1984) Biochim. Biophys. Acta 802, 459-466. specific receptors would lend an additional degree of speci- 34. Silverstein, S. C., Steinman, R. M. & Cohn, Z. A. (1977) ficity and control to steroid hormone action. Annu. Rev. Biochem. 46, 669-722. 35. Bordin, S. & Petra, P. H. (1980) Proc. Natl. Acad. Sci. USA We acknowledge the technical assistance of Mr. Frank Elias and 77, 5678-5682. Ms. Susan Polimeni. This work was supported in part by Public 36. Egloff, M., Vendrely, E., Tardivel-Lacombe, J., Dadoune, Health Service Grant AM28562. J. P., & Degrelle, H. (1982) C.R. Seances Acad. Sci. Ser. C 295, 107-112. 1. Rosner, W. (1976) in Trace Components ofPlasma: Isolation 37. Tardivel-Lacombe, J., Egloff, M., Mazabraud, A. & Degrelle, and Clinical Significance, eds. Jamieson, G. A. & Greenwalt, H. (1984) Biochem. Biophys. Res. Commun. 118, 488-494. T. J. (Liss, New York), pp. 377-395. 38. Rosen, V., Jung, I., Baulieu, E. E. & Robel, P. (1975) J. Clin. 2. Siiteri, P. K., Murai, J. T., Hammond, G. L., Nisker, J. A., Endocrinol. Metab. 41, 761-770. Raymoure, W. J. & Kuhn, R. W. (1982) Recent Prog. Horm. 39. Steins, P., Kreig, M., Hollmann, H. J. & Voigt, K. D. (1974) Res. 38, 457-571. Acta Endocrinol 75, 773-784. 3. Westphal, U. (1983) J. Steroid Biochem. 19, 1-15. 40. Faict, D., De Moor, P., Van Baelen, H., Heyns, W., Stevens, 4. Lan, N. C., Karin, M., Nguyen, T., Weisz, A., Birnbaum, E. & Ceuppens, J. (1983) J. Steroid Biochem. 19, 1397-1401. M. J., Eberhardt, N. L. & Baxter, J. D. (1984) J. Steroid 41. Perrot-Applanat, M., Logeat, F., Groyer-Picard, M. T. & Biochem. 20, 77-88. Milgrom, E. (1985) Endocrinology 116, 1473-1484. 5. Rosner, W. & Hochberg, R. (1972) Endocrinology 91, 42. Ylikomi, T., Gasc, J. M., Isola, J., Baulieu, E. E. & 626-632. Tuohimaa, P. (1985) Endocrinology 117, 155-160. 6. De Kloet, E. R. & McEwen, B. S. (1976) Biochim. Biophys. 43. King, W. J. & Greene, G. L. (1984) Nature (London) 307, Acta 421, 115-123. 747-749. 7. Koch, B., Lutz, B., Briaud, B. & Mialhe, C. (1976) Biochim. 44. Welshons, W. V., Lieberman, M. E. & Gorski, J. (1984) Biophys. Acta 444, 497-507. Nature (London) 307, 747-749. 8. Feldman, D., Funder, J. W. & Edelman, I. S. (1973) Endocri- 45. Gorski, J., Welshons, W. V. & Sakai, D. (1984) Mol. Cell. nology 92, 1429-1441. Endocrinol. 36, 11-15. 9. Milgrom, E., Atger, M. & Baulieu, E. E. (1970) Nature 46. McClellan, M. C., West, N. B., Tacha, D. E., Greene, G. L. (London) 228, 1205-1206. & Brenner, R. M. (1984) Endocrinology 114, 2002-2007. Downloaded by guest on September 30, 2021