Proc. Natl. Acad. Sci. USA Vol. 91, pp. 340-344, January 1994 Cell Biology In vivo functional protein-protein interaction: Nuclear targeted hsp9O shifts cytoplasmic steroid receptor mutants into the nucleus KWANG IL KANG*, JOCELYNE DEVIN*, FRANSOISE CADEPOND*, NICOLE JIBARD*, ANNE GUIOCHON-MANTELt, ETIENNE-EMILE BAULIEU*, AND MARIA-GRAZIA CATELLI*t *Institut National de la Sant6 et de la Recherche Medicale U33, Lab Hormones, 80 rue du Gen6ral Leclerc, 94276 Le Kremlin-Bicetre Cedex, France; and tlnstitut National de la Sant6 et de la Recherche M6dicale U135, Hormones et Reproduction, 78, rue du G6n6ral Leclerc, 94275 Le Kremlin-Bicetre Cedex, France Contributed by Etienne-Emile Baulieu, September 27, 1993

ABSTRACT In target tissue extracts, heat shock protein complexes are relevant only in the case of cytoplasmic hsp9O has been found associated to aDl unganded steroid steroid receptors. Here we demonstrate, in the absence of receptors. Modulation ofimportant functions ofthese receptors, ligand, a functional in vivo intracellular hsp90-steroid recep- including prevention of DNA binding and optimiztion of tran- tor interaction using steroid receptors made exclusively scriptional activity, has been attributed to hsp9O. However no cytoplasmic by suppression of their hormone-independent unequivocal in vivo demonstration ofinteraction between recep- NLS and a nuclear targeted hsp90. Furthermore, with wild- tors and hsp9O has been presented. We targeted chicken hsp9O, type proteins, we show the presence in the nucleus of a mainly cytoplasmic protein, with the nucleoplamin nuclear progesterone receptor and some hsp90. lozation signal (90NLS). After transfection into COS-7 ceDls, 9ONLS was found in the nucleus with specific immunofluores- cence and confocal microscopy techniques. A human glucocor- MATERIALS AND METHODS ticosteroid receptor mutant devoid of NLS sequence was also Expression Vectors and Transfection of COS-7 Cells. Syn- expressed in COS-7 cells and found exclusively cytoplasmic. thetic DNA encoding the nucleoplasmin NLS sequence (11) Coexpression of 9ONLS and of the cytoplasmic human gluco- was inserted in the unique PflMl site ofchicken hsp90 cDNA corticosteroid receptor mutant led to complete nuclear localiza- (12) in the pSP72 vector. After sequencing, the wild-type tion ofthe receptor, indicating its piggyback transport by 9ONLS (90WT) and the modified cDNAs (90NLS) were subcloned in and thus physical and functional interaction between the two the pSVK3 vector (Pharmacia). The C-terminal proline in the proteins in the absence of hormone. The same nuclear localiza- NLS sequence resulted from subcloning. Transfection of tion was obtained after cotransfection of 9ONLS and a cytoplas- COS-7 cells, human glucocorticosteroid receptor (hGR) and mic rabbit progesterone receptor mutant. Finally, coexpression rabbit progesterone receptor (rPR) cDNAs inserted in COS of wild-type rabbit progesterone receptor (nuclear) and wild- cell expression vectors and their deletion mutants have been tpe hsp9O (cytoplasmic) into COS-7 cells provoked partial described (2, 13). relocalization of hsp9O into the nucleus. These experiments lay Antibodies. The antibodies utilized were monoclonal BF4 the groundwork on which to study hsp9O as a chaperone, and D7a anti-avian hsp90 (4, 14), which did not recognize regulating activities ofsteroid receptors and possibly participat- endogenous COS-7 hsp90, and Al antipeptide (GluOK-Tyr280) ing in their nuclear-cytoplasmic shuttling. hsp90 antiserum, which, while not recognizing 8S steroid receptor complexes (15), cross-reacted among species. A Steroid hormone receptors are hormone-dependent tran- hGR-specific antipeptide (Ala149-Gln1") antiserum (16) and scriptional activators bearing hormone-independent nuclear anti-rPR Let 126 monoclonal antibody were used (2). localization signals (NLSs) whose efficiency is the primary Preparation of Cellular Extracts, Sedimentation Analysis, determinant oftheir subcellular localization in the absence of and Western Blot. After transfection, COS-7 cell cytosols hormone; thus the glucocorticosteroid receptor is predomi- were prepared and glycerol gradient centrifugations were nantly located in the cytoplasm, whereas estrogen and pro- performed as described (13). hGR was labeled with [3H]tri- gesterone receptors are essentially nuclear (1-3). Despite amcinolone acetonide (40 nM). Before centrifugation, 10 01 of this, however, a puzzling observation is that all unliganded labeled cytosol were incubated 2 hr at 0°C with 5 Al ofascitic steroid receptors form "8S" complexes with 90-kDa heat fluid (BF4 or D7a) or 10 Al of Al antiserum. shock protein (hsp90) in the cytosol of target cell homoge- Aliquots of gradient fractions, after glucose oxidase (7.9S) nates (4-7), hsp90 being an abundant and ubiquitous protein and peroxidase (3.6S) assay, were counted and/or submitted described as essentially cytoplasmic. Two functions have to 7.5% PAGE and transferred to a nitrocellulose membrane been ascribed to receptor-bound hsp90: masking of the (Schleicher & Schuell). Proteins 90WT and 90NLS were receptor DNA binding domain and maintenance ofthe ligand detected with BF4 antibody and avidin-biotin Vectastain binding domain in a functional hormone-binding conforma- reagent kit (Vector Laboratories). tion, at least for gluco- and mineralocorticosteroid receptors Indirect Immunofluorescence Staining and Confocal Micros- (reviewed in refs. 8 and 9). The in vivo requirement of copy. COS-7 cells grown on coverslips were transfected with appropriate levels of hsp90 for efficient hormonal activation 2 ,g of the plasmids indicated in the legend of figures using of steroid receptors in yeast has also been reported (10). The the DEAE-dextran technique (13). Cells were fixed in 4% lack ofdirect evidence for the in vivo interaction between the paraformaldehyde/0.1% Triton X-100 in phosphate-buffered two proteins has, however, led to the suspicion that hsp90- saline (20 min at room temperature). 9OWT and 9ONLS were steroid receptor complexes represent an artifactual associa- detected with avian-specific hsp90 BF4 rat monoclonal an- tion occurring during cell homogenization or that these Abbreviations: hsp90, heat shock protein of 90 kDa; NLS, nuclear localization signal; 9OWT, wild-type chicken hsp90; 9ONLS, nuclear The publication costs ofthis article were defrayed in part by page charge targeted chicken hsp90; hGR, human glucocorticosteroid receptor; payment. This article must therefore be hereby marked "advertisement" rPR, rabbit progesterone receptor. in accordance with 18 U.S.C. §1734 solely to indicate this fact. tTo whom reprint requests should be addressed. 340 Downloaded by guest on October 1, 2021 Ceff Biology: Kang et aL Proc. Natl. Acad. Sci. USA 91 (1994) 341 tibody (ascitic fluid 1:5000, 30 min at room temperature) and confocal microscopy of immunofluorescent cells that were fluorescein-conjugated anti-rat antibody (1:200, 30 min at detected by the avian-specific anti-hsp90 BF4 antibody (4). room temperature; Southern Biotechnology Associates). Even though the complete nuclear exclusion of 9OWT is The hGR-specific antipeptide antiserum (1:4000) (16) and a difficult to demonstrate because it is abundant in the cyto- rhodamine-conjugated anti-rabbit antibody (1:200) (Southern plasm, Fig. 1 shows that transfected 9OWT was essentially Biotechnology Associates) were used. The specificity of the cytoplasmic (Fig. 1A), whereas 9ONLS was always present in secondary antibodies allowed the simultaneous processing of nuclei (Fig. 1 B and C). Forty-five percent ofcells expressing hsp90 and hGR detection. For rPR the Let 126 monoclonal 9ONLS showed an intense nuclear and cytoplasmic fluores- antibody was used as indicated (2). To avoid cross-reactivity cence, 50% showed a predominant, intense nuclear fluores- between anti-mouse secondary antibody and rat BF4, COS-7 cence, and 5% showed a moderate nuclear and cytoplasmic transfected cells were first treated for rPR detection and then staining. The level of 9OWT or 9ONLS expression, deter- treated for hsp90 detection. mined by semiquantitative Western blot analysis (not Stained cells were viewed with an Optiphot II Nikon shown), was about 2% of total soluble proteins. Since the microscope (objective, 60x) and scannings were performed number of transfected cells expressing the construct was using a Bio-Rad MRC 600. For fluorescein, an argon ion laser 10%, the intracellular level of avian hsp90 would actually set to A = 488 nm was used. The fluorescence emission was represent 10-20%o of the soluble proteins, 10 times the separated by a dichroic mirror (DR565) and a long-pass filter endogenous level. Importantly, the 9ONLS was still recog- (540DF30) was placed in front of the photomultiplier; the nized by anti-chicken hsp90 BF4 and D7a antibodies and was signal was digitalized by photon counting. Each section was not preferentially retained in the nuclear fraction of cell scanned 30 times. The pinhole of the confocal system was homogenate but was recovered in the cytosol. In glycerol adjusted to allow a field depth of about 0.5 ,Am. For confocal density gradients, the 9ONLS sedimented at 6-7 S, exactly as scanning of rhodamine staining, a helium/neon ion laser set did 9OWT (Fig. 1E) or hsp90 native dimer (19). to A = 543 nm was used. The emission long-pass filter was Complexes Between Receptors and Wild-Type or Nuclear EF600LP. Double fluorescence images were acquired in two Targeted hsp9O. To test the ability of9OWT or 9ONLS to form passes-fluorescein first, rhodamine second-to avoid 8S complexes with a steroid receptor, the two constructs bleeding from one channel into the other. Photographs were were each cotransfected with -the hGR. In the absence of taken on Kodak T-Max film using a camera mounted on a VM hormone, hGR is preferentially localized in the cytoplasm of 1710 Lucius and Baer film recorder. COS-7 transfected cells, its hormone-independent NLS being almost completely repressed, and it forms complexes with the endogenous monkey hsp90 (13, 20, 21). The 8S steroid RESULTS AND DISCUSSION binding complexes were found after cotransfection of hGR SubceHular Localization ofWild-Type and Nuclear Targeted with 9OWT and 9ONLS (Fig. 2 A-D). Interestingly, the avian hsp9O. The bipartite NLS of Xenopus laevis nucleoplasmin hsp90 was predominant in both of the 8S forms, despite the (11), which bears many similarities to the NLS of steroid presence of significant levels of endogenous hsp90. This was receptors (17), was inserted into the N-terminal region of demonstrated by the displacement of both 8S complexes, on chicken hsp90 between Pro-li and Met-12 (Fig. 1D). The glycerol gradients, by the two anti-chicken hsp9o antibodies presence of the NLS in avian hsp90 does not significantly (Fig. 2A-C). The 6-7S dimeric receptor-free avian hsp90 was alter the predicted secondary structure nor the hydropathy also displaced in the gradient by BF4 antibody (Fig. 2E, a, a', plot ofthe protein according to the Garnier-Robson program, b, b') and by D7a antibody (not shown). In addition, the Al the NLS sequence being located in a flexible, exposed antipeptide antiserum, directed against the highly charged protein segment (18). The intracellular location of the wild- region A of chicken hsp90, which is a candidate for interac- type (90WT) or NLS-bearing (90NLS) chicken hsp9o, ex- tion with steroid receptor (12, 15), reacted with all receptor- pressed transiently in COS-7 cells, was assessed by laser free hsp90 (not shown), both avian and simian, but not with

D E GO PO

N-terminal of chicken hsp9O t M PE A V TDPMEE T F A F... 9OWT 9ONLS KRPAATKKAGQAKKKK P NLS of nucleoplasmin 4 5 6 7 8 9 1011 12 13 14 15 1617 18 19 FIG. 1. (A-C) Subceliular localization of chicken 9OWT (A) and 9ONLS (B and C) in COS-7 transfected cells detected by indirect immunofluorescence staining and laser confocal microscopy. (Bar = 10 pm.) (D) Insertion of nucleoplasmin NLS in the N-terminal region of chicken hsp9o. (E) Sedimentation analysis of 9OWT and 9ONLS on glycerol gradients, detected by Western blot with BF4 antibody. Internal markers are indicated. GO, glucose oxidase; PO, peroxidase. Downloaded by guest on October 1, 2021 342 Cefl Biology: Kang et al. Proc. Natl. Acad. Sci. USA 91 (1994) DPM Several important deductions stem from these results. (i) 800 GO P0 The hGR seems to interact well with the cotransfected A GO P0 B chicken hsp90, possibly due to their concomitant translation (9). The cotransfection system, combined with the use of appropriate specific antibodies, is therefore suitable for 600 studying the interactions ofhsp90, steroid receptor, and other proteins forming heterooligomeric complexes and the do- mains eventually involved. (ii) During the homogenization procedure, no major exchange takes place between receptor- 400- bound avian hsp90 and the endogenous one, suggesting that hsp90-receptor interaction occurs in vivo. (iii) Nuclear tar- geting of the hsp90 does not impede 8S complex formation with the hGR, a predominantly cytoplasmic receptor. Nuclear Targeting of hsp9O and Subcellular Distribution of 200 "Cytoplasmic" Steroid Receptor Mutants. The finding of the interaction between 9ONLS and hGR prompted us to inves- tigate the possibility that 9ONLS was able to change the hGR subcellular distribution. To assess the colocalization of9OWT 0 or 9ONLS with the hGR directly, we used laser confocal 800 C GO PO GO PO microscopy and specific double immunofluorescence detec- tion. However, in the absence of hormone, as already ob- served by immunohistochemistry (20), we found that some hGR transfected alone was nuclear. Consequently, for a 600- clear-cut experimental design, we chose the deletion mutant A491-515 hGR that, when transfected alone, showed an exclusive cytoplasmic localization as well as the ability to form complexes with endogenous hsp90 (13, 20). When 400 cotransfected with 9OWT, A491-515 hGR remained exclu- sively cytoplasmic (Fig. 3C) as did 9OWT (Fig. 3A). In contrast, when cotransfected with 9ONLS, the hGR mutant was nuclear in all cotransfected cells showing an exclusive 9ONLS nuclear location (Fig. 3 B and D). Note that in Fig. 3B 200- both cells express 9ONLS, whereas only one (see Fig. 3D) expresses the mutant hGR shifted to the nucleus. In the cotransfected cells showing intense nuclear and cytoplasmic 9ONLS fluorescence, the GR mutant was also nuclear or 0 nuclear and cytoplasmic. It is worthy of note that 9ONLS was 0 10 20 0 10 20 unable to shift cotransfected cytoplasmic 13-galactosidase FRACTION NUMBER into the nucleus (not shown). Thus, by swapping the hor- mone-independent nuclear localization function from a ste- roid receptor to hsp90, we have provided direct evidence for a an in vivo interaction between the two proteins, thus allowing the nuclear transfer of the cytoplasmic hGR mutant. When a complex is made between two NLS engineered proteins, the a nuclear translocation-deficient subunit is carried into the nucleus by the nuclear translocation-proficient subunit. We also were able to obtain exactly the same result with 9ONLS and the deletion mutant A593-640 of the rPR (2). Whereas the __- ~~b wild-type form of the PR is nuclear, the A593-640 mutant displays a cytoplasmic colocalization with 9OWT in the absence of hormone (ref. 2 and Fig. 3 Eand G). It was shifted into the nucleus (Fig. 3H) by 9ONLS (Fig. 3F). In glycerol FIG. 2. (A-D) Sedimentation analysis on glycerol gradients of 8S density gradient analysis, wild-type rPR and A593-640 mutant hGR complexes, 9OWT and 9ONLS, after cotransfection ofhGR with 8S with the or cotransfected 9OWT or 9ONLS. (A) hGR plus 9OWT (o) and hGR plus 9OWT with formed complexes endogenous BF4 antibody (o). (B) hGR plus 9ONLS (o) and hGR plus 9ONLS with avian hsp90 (9OWT) (not shown). BF4 antibody (o). (C) hGR plus 9ONLS (o) and hGR plus 9ONLS with Cytoplasmic mutants of hGR and rPR have been deleted, D7a antibody (o). (D) hGR plus 9OWT with nonimmune serum (o) to different extents, in their hormone-independent NLS. and hGR plus 9OWT with anti-hsp90 Al serum (o). Internal markers They both lack the positively charged motif conserved are as in Fig. 1. GO, glucose oxidase; PO, peroxidase. (E) Western throughout the steroid receptor subfamily (3), corresponding blot analysis of fractions 4-17 of gradients A [a (o), a' (e)] and B [b to C-terminal basic regions of the bipartite NLS (17). In (o), b' (o)] with BF4 antibody. addition, rPR lacks the NLS located in the zinc finger region (2, 3). Our results therefore confirm that GR interacts with the 8S complexes (Fig. 2D), thus confirming that Al epitopes hsp90 before cell lysis, as indicated by metabolic labeling are masked in the 8S complexes (15). The hGR, on the basis studies (22), and that the integrity of glucocorticosteroid of a binding assay, was expressed at 2 pmol/mg of protein, receptor NLS is not necessary for complex formation with a concentration 100-fold lower than that found for the co- hsp90 (13). They also demonstrate that deletion of all of the transfected hsp90. Thus the ratio of the two overexpressed hormone-independent NLSs ofrPR does not compromise the proteins in cotransfected COS-7 cells, glucocorticosteroid interaction with hsp90. receptor and hsp90, was maintained in the physiological Nuclear Localization of Cotransfected rPR and Chicken range of about 1:100. hsp9O. Our experimental evidence for a "cytoplasmic" ste- Downloaded by guest on October 1, 2021 Cefl Biology: Kang et aL Proc. Natl. Acad. Sci. USA 91 (1994) 343

FIG. 4. Subcellular localization of rPR and 90WT detected by simple or double indirect immunofluorescence labeling and confocal microscopy of COS-7 cells. (A) Cell expressing rPR (Let 126 anti-PR and rhodamine staining). (B) Cell expressing 9OWT (BF4 anti-hsp90 and fluorescein staining). (C and D) Cell coexpressing rPR (rhoda- mine staining, C) and 90WT (fluorescein staining, D). (Bar = 10 Am.) steroid receptor complexes are thus localized in the cyto- plasm or in the nucleus (e.g., glucocorticosteroid and pro- gesterone receptor 8S complexes, respectively) according to the potency of the hormone-independent NLS. In the ab- sence of hormone, a fraction of hsp90 may be thus trans- ported "piggyback" into the nucleus by nuclear steroid receptors. A similar mechanism has been reported for other protein complexes (reviewed in ref. 25) and for progesterone receptor dimers (3). It is possible that complexes of hsp90 with predominantly cytoplasmic or nuclear steroid receptors really shuttle, at different rates, between the two compart- ments. In fact, receptor recycling has been described for FIG. 3. Subcellular localization of A491-515 hGR and A593-640 progesterone and glucocorticosteroid receptors (26, 27). In rPR cotransfected with 9OWT or 9ONLS detected by double indirect shuttling of hsp90, possibly car- immunofluorescence labeling and confocal microscopy of COS-7 turn, nuclear-cytoplasmic cells. (A and C) Cell coexpressing 90WT (A) and AhGR (C). (B and ried by other proteins, may result in a nuclear level of hsp90 D) Two cells expressing 9ONLS (B), one ofwhich coexpresses AhGR able to influence the activation of the receptor by the hor- (D). (E and G) Cell coexpressing 9OWT (E) and ArPR (G). (F and H) mone, a step that, in vitro, requires hsp90 dissociation (8, 9). Cell coexpressing 9ONLS (F) and ArPR (H). (A, B, E, and F) hsp90, Functions of Steroid Receptor-Bound hsp9O. Additional fluorescein staining. (C, D, G, and H) Steroid receptors, rhodamine proteins from the large chaperone family, such as hsp70 and staining. (Bars = 10 ,um.) p59-HBI (28), have been found associated with the hetero- roid receptor shifted into the nucleus by 90NLS implies that oligomeric complex of steroid receptor and hsp90 that was in the described initially (29, 30). Their participation in steroid the natural situation would be the reverse-that is, receptor functions remains to be determined. Only hsp90 has absence of hormone, some cytoplasmic hsp90 is relocalized been proven, by various immunological and biochemical into the nucleus by steroid receptors possessing a constitu- approaches, to interact directly with receptors (reviewed in tively functional NLS. In fact, a hsp90-containing 8S pro- refs. 8, 9, 30). gesterone receptor complex has been obtained from purified Therefore, it would seem that, starting from steroid recep- nuclei ofrabbit uterus in the absence ofthe cognate hormone tor translation, hsp90 displays its chaperone function by (23). On the other hand, the reported conclusion that the maintaining an appropriate receptor conformation that facil- nuclear progesterone receptor is mainly hsp90-free (24) itates hormone binding and the efficiency ofhormonal effects seems to be due to the inability to detect the small fraction of (8-10). In addition to this positive role, we and others have the abundant cytoplasmic hsp9O that would be colocalized in already proposed that the unliganded hormone binding do- nuclei with PR. Here we were able to demonstrate that some main plus the interacting hsp90 be considered as the in vivo 9OWT, when cotransfected with rPR, is localized in the repression unit of the ligand-induced activities of steroid nucleus. Fig. 4D clearly shows that a low but detectable level receptors, such as DNA binding and transcriptional activa- of 9OWT was found in nuclei after cotransfection with rPR, tion (4, 6, 8, 9). Our results also point to a possible novel role when compared to the lack of detectable nuclear 9OWT for hsp90: the in vivo functional interaction, in the absence of transfected alone under the same experimental conditions hormone, between hsp90 and steroid receptor may imply that (Figs. 4B and 1A). hsp90 would participate in their nuclear localization or nu- This last result suggests that possibly all steroid receptors, clear-cytoplasmic shuttling. hsp90 may help in this step by and not only those found in the cytoplasm, interact with assisting steroid receptor folding-unfolding, assembly or hsp90 during translation (9) and, most importantly, that this disassembly with other proteins such as NLS receptors or interaction is strong enough in vivo to persist during the component of the nuclear pore, and/or by furnishing, via its hormone-independent nuclear localization step. hsp9O- ATPase activity (31), the energy required in the NLS- Downloaded by guest on October 1, 2021 344 Cell Biology: Kang et al. Proc. Natl. Acad. Sci. USA 91 (1994) dependent translocation (25). This hypothesis remains to be Bigogne, C., Levin, J. M., Garnier, J., Baulieu, E. E. & Catelli, tested in further studies concerning the hsp9o-steroid hor- M. G. (1989) Biochem. Biophys. Res. Commun. 159, 140-147. 13. Cadepond, F., Schweizer-Groyer, G., Segard-Maurel, I., mone receptor mechanism and the more general biological Jibard, N., Hollenberg, S. M., Giguere, V., Evans, R. M. & function of signal transduction to the cell nucleus. Baulieu, E. E. (1991) J. Biol. Chem. 266, 5834-5841. 14. Schuh, S., Yonemoto, W., Brugge, J., Bauer, V. J., Riehl, K.I.K. and J.D. contributed equally to this work and should be R. M., Sullivan, W. P. & Toft, D. 0. (1985) J. Biol. Chem. 260, considered equally as first authors. We thank G. Geraud (Service de 14292-14296. Microscopie Confocale, Institut Jacques Monod, University of Paris 15. Catelli, M. G., Binart, N., Vourc'h, C., Devin, J. & Baulieu, VII) for assistance in confocal laser microscopy; the laboratories of E. E. (1990) in Activation of Hormone and Growth Factor R. Evans and E. Milgrom for steroid receptor plasmids; R. Fiddes, Receptors, eds. Alexis, M. N. & Sekeris, C. E. (Kluwer, K. Rajkowski, M. Renoir, and F. Delahaye for helpful suggestions Boston), pp. 239-256. and discussion; and C. Legris, J. C. Lambert, and M. Balhoul for 16. Cadepond, F., Binart, N., Chambraud, B., Jibard, N., Schwei- manuscript and figure preparation. This work was supported by zer-Groyer, G., Segard-Maurel, I. & Baulieu, E. E. (1993) Institut National de la Sante et de la Recherche Medicale and Proc. Natl. Acad. Sci. USA 90, 10434-10438. Association pour la Recherche sur le Cancer (Grant ARC MGC 17. Robbins, J., Dilworth, S. M., Laskey, R. A. & Dingwall, C. was of an Association pour la Recherche (1991) Cell 64, 615-623. 6329). K.I.K. the recipient J. M., Robson, B. & Gamier, J. sur le Cancer fellowship. 18. Biou, V., Gibrat, J. F., Levin, (1988) Protein Eng. 2, 185-191. 19. Radanyi, C., Renoir, J. M., Sabbah, M. & Baulieu, E. E. (1989) 1. Picard, D. & Yamamoto, K. R. (1987) EMBO J. 6, 3333-3340. J. Biol. Chem. 264, 2568-2573. 2. Guiochon-Mantel, A., Loosfelt, H., Lescop, P., Sar, S., Atger, 20. Cadepond, F., Gasc, J. M., Delahaye, F., Jibard, N., Schwei- M., Perrot-Applanat, M. & Milgrom, E. (1989) Cell 57, 1147- zer-Groyer, G., Segard-Maurel, I., Evans, R. & Baulieu, E. E. 1154. (1992) Exp. Cell Res. 201, 99-108. 3. Ylikomi, T., Bocquel, M. T., Berry, M., Gronemeyer, H. & 21. Pratt, W. B., Jolly, D. J., Pratt, D. V., Hollenberg, S. M., Chambon, P. (1992) EMBO J. 11, 3681-3694. Giguere, V., Cadepond, F., Schweizer-Groyer, G., Catelli, 4. Joab, I., Radanyi, C., Renoir, J. M., Buchou, T., Catelli, M. G., Evans, R. M. & Baulieu, E. E. (1988) J. Biol. Chem. M. G., Binart, N., Mester, J. & Baulieu, E. E. (1984) Nature 263, 267-273. (London) 308, 850-853. 22. Howard, K. J. & Distelhorst, C. W. (1988) J. Biol. Chem. 263, 5. Catelli, M. G., Binart, N., Feramisco, J. R. & Helfman, D. 3474-3481. (1985) Nucleic Acids Res. 13, 6035-6047. 23. Renoir, J. M., Radanyi, C., Jung-Testas, I., Faber, L. E. & 6. Catelli, M. G., Binart, N., Jung-Testas, I., Renoir, J. M., Baulieu, E. E. (1990) J. Biol. Chem. 265, 14402-14406. Baulieu, E. E., Feramisco, J. R. & Welch, W. J. (1985) EMBO 24. Tuohimaa, P., Pekki, A., Blauer, M., Joensuu, T., Vilja, P. & J. 4, 3131-3135. Ylikomi, T. (1993) Proc. Natl. Acad. Sci. USA 90, 5848-5852. 7. Sanchez, E. R., Toft, D. O., Schlesinger, M. J. & Pratt, W. B. 25. Garcia-Bustos, J., Heitman, J. & Hall, M. N. (1991) Biochim. (1985) J. Biol. Chem. 260, 12398-12401. Biophys. Acta 1071, 83-101. 8. Baulieu, E. E., Binart, N., Cadepond, F., Catelli, M. G., 26. Guiochon-Mantel, A., Lescop, P., Christin-Maitre, S., Loos- Chambraud, B., Garnier, J. M., Groyer-Schweizer, G., Oblin, felt, H., Perrot-Applanat, M. & Milgrom, E. (1991) EMBO J. M. E., Radanyi, C., Redeuilh, G., Renoir, J. M. & Sabbah, M. 10, 3851-3859. (1989) in The Steroid/Thyroid Hormone Receptor Family and 27. Madan, A. P. & DeFranco, D. B. (1993) Proc. Natl. Acad. Sci. Gene Regulation, eds. Gustafsson, J. A., Eriksson, H. & USA 90, 3588-3592. Carlstedt-Duke, J. (Birkhauser, Basel), pp. 301-318. 28. Gething, M.-J. & Sambrook, J; (1992) Nature (London) 355, 9. Pratt, W. B. (1992) BioEssays 14, 841-848. 33-45. 10. Picard, D., Khursheed, B., Garabedian, M. J., Fortin, M. G., 29. Callebaut, I., Renoir, J. M., Lebeau, M. C., Massol, N., Lindquist, S. & Yamamoto, K. R. (1990) Nature (London) 348, Burny, A., Baulieu, E. E. & Mornon, J. P. (1992) Proc. Natl. 166-168. Acad. Sci. USA 89, 6270-6274. 11. Dingwall, C., Dilworth, S. M., Black, S. J., Kearsey, S. E., 30. Smith, D. F. & Toft, D. 0. (1993) Mol. Endocrinol. 7, 4-11. Cox, L. S. & Laskey, R. A. (1987) EMBO J. 6, 69-74. 31. Nadeau, K., Das, A. & Walsh, C. T. (1993) J. Biol. Chem. 268, 12. Binart, N., Chambraud, B., Dumas, B., Rowlands, D. A., 1479-1487. Downloaded by guest on October 1, 2021