Download - Mechanism of gene expression by the glucocorticoid receptor: Role of protein-protein interactions

lain J. McEwan, Anthony P. H. Wright and Jan-Ake Gustafsson

Summary

The glucocorticoid receptor belongs to an important class of transcription factors that alter the expression of target genes in response to a specific hormone signal. The glucocorticoid receptor can function at least at three levels: (1) recruitment of the general transcription machinery; (2) modulation of transcription factor action, independent of DNA binding, through direct protein- protein interactions; and (3) modulation of chromatin structure to allow the assembly of other gene regulatory proteins and/or the general transcription machinery on the DNA. This review will focus on the multifaceted nature of protein-protein interactions involving the glucocorticoid receptor and basal transcription factors, coactivators and other transcription factors, occurring at Accepted these different levels of regulation.

Introduction Glucocorticoid hormones have diverse biological functions, including effects on metabolism, the immune system and the brain, and are widely used clinically as immunosuppressive and anti-inflammatory agents. The biological effects of glucocorticoids are generally thought to be mediated by an intracellular receptor protein, the glucocorticoid receptor (GR), which transduces the hormone signal to the nucleus and participates directly in gene regulation. In the absence of hormone, the receptor appears to be predominantly in the cytoplasm complexed with other proteins, most notably hsp 90 (see ref. 1 and references therein). Upon binding steroid, this complex dissociates and the receptor can enter the nucleus, dimerise and bind to specific DNA sequences, glucocorticoid response elements (GREs), which are associated with target genes (see ref. 2 and references therein). This review will focus on the role of protein-protein interactions involved in gene regulation by the activated receptor protein. This will include discussion of (1) the domain organ- isation of the GR, with emphasis on the structure of the N- terminal transactivation domain; (2) reconstitution of receptor transactivation activity under cell-free conditions; and (3) the interaction of the GR with components of the general transcription machinery, coactivators and other transcription factors. The concluding section contains a review of proteins, identified in physical or functional screens, that bind to other members of the steroid-thyroid hormone receptor

11 November 1996

superfamily, and the implications of these interactions for receptor function are discussed.

Domain structure of the human GR As alluded to above, the GR belongs to a superfamily of nuclear receptors which share a common domain organisa- tion (Fig. l ) , while mediating the effects of structurally diverse signals.

The ligand and DNA binding domains The ligand binding (LBD) domain is localised within the C- terminal portion of the protein and is characterised by lim- ited sequence identity between different receptors. The LBD is proposed to have a common ligand binding pocket con- sisting of conserved amino acid residues and receptor- specific amino acids (see refs 3-5 and references therein). A region critical for hsp 90 binding has been mapped to the LBD of the mouse and rat GRs and corresponds to amino acids 568-653 of the human GR (see ref. 1 and references therein).

The DNA binding domain (DBD), located N-terminally of the LBD, shares the highest homology between the different members of the steroid-thyroid hormone receptor superfamily; for example, the GR~~~419-500 shares 46% and 44% amino acid identity with the DBDs of the oestrogen (ER) and retinoid X (RXR) receptors, respectively. Nuclear magnetic resonance (NMR) spectroscopy revealed the presence of two

a-helices folded perpendicular to each other, while subsequent refinement of the solution structure and crystallo- graphic analysis of the GRDBD-DNA complex identified a third a-helix (see ref. 2 and references therein). Fig. 1 shows schematically the location of the three helices within the DBD.

Receptor transactivation domains Transactivation activities have been localised in the human, rat and mouse GRd6-*). In the human GR the major transactivation domain, termed t l , was mapped to the N-terminal region of the protein between amino acids 77 and 262(’s9) and can activate reporter genes when fused to the GR- or a heterologous DBD (see below). Deletion analysis revealed that sequences near the C terminus of TI were important for transactivation (Fig. I ) , such that a 58-amino-acid core domain (residues 187 to 244) was almost as active as the intact ~l domain and squelched basal transcription to com- parable levels(lO). Squelching describes the inhibitory effect on transcription seen by increasing the concentration of the activator protein. The inhibition is thought to result from protein-protein interactions involving the activator and a limiting pool of target factors, so that the latter are unavailable for normal function (reviewed in ref. 11).

Additional regions involved in transactivation have been mapped to the LBD (Fig. l ) , and include ~ 2 ( ~ , ~ ) and sequences at the C terminus of the protein (amino acids 727-763, TC or AF-2)(I2). The latter region is highly conserved within the family of steroid-thyroid hormone nuclear receptors and forms part of an amphipathic a-helix that is thought to undergo a dramatic conformational rearrange- ment when the receptor binds ligand(3-5).

Analysis of ~1 and the 58-amino-acid polypeptide containing the t l core, using circular dichroism spectroscopy (CD), revealed that both polypeptides were relatively unstructured in aqueous solution(i3). The 51 core acquired significant a-helical content in the presence of the hydrophobic solvent trifluoroethanol (TFE), and three protein segments with a-helical character within the 58-amino- acid core sequence have been identified(I3) (Fig. 1). The purified transactivation domains of VPI 6, Ga14, Gcn4, NF- KB (p65) and c-myc are also largely unstructured in aqueous solution at neutral pH. Under acidic conditions the Gal4 and Gcn4 transactivation domains adopted a P-sheet conformation, while those of VP16, NF-KB (p65) and c-myc adopted an a-helical conformation in the presence of TFE (see ref. 14 and references therein). For the GR-Tl core transactivation domain, mutational analysis has shown that the ability to activate a reporter gene in vivo correlates with the ability to form an a-helical conformation in vitro(15). Thus, transactivation domains have a propensity for structure formation under specific conditions and some evidence suggests that such conditions may be fulfilled upon interaction with target p ro te in~( ’~~~6) .

Interaction of the GR-tl transactivation domain with limiting factors required for transcriptional activation Initiation of transcription in eukaryotic cells Accurate initiation of transcription of mRNA coding genes requires a number of basal or general transcription factors (TFIIA, B, D, E, F and H), which together with RNA polymerase II assemble at the start site of transcription (for a

Fig. 1. Domain structure of the human GR. Within the DBD, cysteine residues co-ordinating Znz+ ions are identified by a closed circle underneath; amino acids involved in DNA binding specificity (P-box) and dimerisation (D-box) are highlighted in bold; amino acids that, when mutated, are deficient in AP-1 trans- repression are marked with an asterisk@*). The position of a-helical regions are indicated by cylinders above or below the amino acid sequence.

review see ref. 17). While the basal factors or sub-sets of these factors, together with the RNA polymerase II, are sufficient to initiate transcription under cell-free conditions, regulated levels of transcription require the action of additional factors, termed coactivators or mediators (see ref. 17 and references therein). Activator proteins have been shown to bind to basal transcription factors, notably the TATA-binding protein (TBP), a component of TFIID, and TFIIB, as well as coactivators, and it is therefore likely that multiple interactions between activators and the general transcription machinery are involved in the control of gene transcription (reviewed in ref. 17). With the recent cloning of a number of the general transcription factors, together with coactivator proteins, from yeast to man, it will soon be possible to study the role of different protein-protein interactions in the transcription process using highly defined reconstituted systems.

Reconstitution of G R-t 1 activity under cell-free conditions High levels of transcriptional activators have been reported to inhibit (squelch) transcription by sequestering a limiting cellular factor(s) important for transcription (see ref. 11 and references therein). Previously, we have described in vivo and in vitro squelching experiments showing that the isolated GR-rl domain competed for limiting factors required for basal transcription (see ref. 18 and references therein). Although it was possible to reverse r l -dependent squelching using a chromatography fraction derived from a yeast whole-cell extract, attempts with known recombinant basal transcription factors have so far proved unsuccessful. Thus, one possibility is that r1 may interact with a target that is in a complex containing general transcription factors (reviewed in ref. 18). An attractive candidate for the latter would be the recently described holo-RNA polymerase, which may contain all or a subset of the general transcription factors as well

as coactivators and the RNA polymerase subunits (see ref. 17 and references therein). While the basal transcription factors TBP and TFllB bind to other nuclear receptors, the functional significance of such interactions has not always been dem~nstrated('~-~*) (Table 1). Interestingly, in the case of the thyroid hormone receptor (TR), the interaction with TBP is disrupted by ligand and appears to be important for TR-dependent repression of gene expression in the absence of thyroid hormone(27).

Three monoclonal antibodies (mAbs) that specifically inhibited r l -dependent transactivation have been identified and the epitopes for these mAbs mapped to the rl core domain. These antibodies were unable to reverse the squelching activity of 71, however, and this suggests that interactions with the general transcription machinery described by the squelching experiments, although important, are not sufficient to explain the mechanism of transcriptional activation by r l (reviewed in ref. 18). The kinetics of squelching are compatible with rl having a role at an early step in preinitiation complex (PIC) formation, as once assembled, the PlCs were more refractory to the inhibitory effect of the isolated r l -domain. This is also consistent with studies on the full-length human GR and a derivative of the rat GR in HeLa and Drosophila embryo nuclear extracts, respectively (see ref. 18 and references therein). The inhibitory effect of the mAbs persist longer, however, suggesting that .rl is also important for events after PIC formation. Based on these studies, our current model suggests a role for the transactivation domain, r l , at more than one step in the initiation reaction, and involves multiple interactions with the general transcription machinery.

Although both the full-length and discrete domains of the receptor have been shown to activate reporter genes under cell-free conditions, these studies to date have involved crude nuclear extracts (see ref. 18 and references therein

Table 1. Summary of known interactions between cloned target factors and members of the steroid-thyroid hormone receptor superfamily Target Source Function

TFllB TBP',' TAFii30 TAFiilIOt ARA7Ot CBPt RAP46

RIP140' SRC-1 familyt (SRC-la-e, GRlPl, TIF-2) T IF - l t Trip-li SUGl t Swi3p

RIP-1 lot

Human Human Human Human Human

Human Mouse Human Human (Mouse) Mouse HumaniMouse Yeast

Basal factor Basal factor TBP associated factor/coactivator TBP associated factor/coactivator Coactivator CREB binding proteinicoactivator Coactivator ? Coactivator ? Coactivator ? Coactivator

Coactivator ? Coactivator ? Chromatin remodeling

Receptors References

ER, PR, TR, VD3R, COUP, ARP-1. HNF-4 ~ ~- ~

19-24, 28 ER, RXR, TR ER PR, RXR AR ER, GR, RAR, RXR, TR GR, AR, ER, TR RAR, RXR, TR ER, RAR, RXR, TR PR, ER, GR, AR, RAR, RXR, TR ER, RAR, RXR, VD3R, PR TR, ER, RAR. VD3R GR

25-27 55 26,69 60 41 39 68 22,62 40-43

56 53,54 49

'Ligand disrupts interaction of TR. tinteraction with RXR is ligand-dependent. 'Interaction is ligand-dependent.

STEROID BINDING

DIMERISATION TRANSACTIVATION :INNDING TRANSACTIVATION

I 71 262 41 5 500 777

Fig. 2. Summary of GR-protein interactions. See text for details.

and refs 29,70). It will therefore be important to show receptor function in a reconstituted assay together with known or suspected coactivator proteins, in order to further elucidate the mechanisms of GR-dependent transcriptional activation biochemically.

Proteins interacting with the GR Interactions with the transcription factors AP- 1 and NF-KB Glucocorticoids are used clinically to inhibit inflammation and cell proliferation in the treatment of diseases such as rheuma- toid arthritis. An important target gene for glucocorticoid repression is the metalloprotease collagenase and studies from a number of laboratories demonstrated that the AP-1 response element in the collagenase promoter was critical for the glucocorticoid response (reviewed in refs 30 and 31). Fur- thermore, the repression by glucocorticoids was not dependent upon protein synthesis, strongly suggesting that the receptor might interfere with AP-1 activity by direct protein- protein interaction^(^^,^'). There is now evidence of the GR binding to the c-fos and c-jun components of AP-1 (see ref. 31 and references therein). The integrity of the GRDBD appears to be critical for the repression activity(32), although additional sequences outside the DBD may also play a role (see refs 30,31 and references therein) (Fig. 2). AP-1 can also antago- nise the activity of the GR, and the basic-zipper regions of c- fos and c-jun are necessary for this interaction, although the N terminus of c-fos, containing the transactivation domain, has also been implicated (reviewed in ref. 30).

A more complex picture has emerged for GR/AP-l interactions on the proliferin gene, where the target sequence for the receptor is an overlapping GRE and AP-I response element (see ref. 33 and references therein). The GR has a positive effect in the presence of c-jun homodimers and a repressive function with c-fos/c-jun heterodimers. A further difference from the collagenase model is that sequences in

HSPW

NF-KB (p6S)

Basal Transcription Factors Co-Activators( s)

AP- I ( C-FOS)

GRIP1 (TIF-2. SRC-I)

Swi3

the N terminus of the receptor were reported to be important for GR-dependent repression in this

Glucocorticoids are also known to inhibit the transcription of key genes involved in the immune response such as interleukin (IL)-6 and -8 (reviewed in ref. 31), although there is no evidence of regulatory sequences binding the GR. On the contrary, the region shown to be critical for glucocorticoid-mediated down-regulation of the IL-6 promoter was mapped to an NF-KB responsive element. NF-KB is a heterodimer of 50 kDa (p50) and 65 kDa (p65, re1 A) subunits and is an important regulator of genes involved in the immune and acute phase responses (see ref. 31). The GR can inhibit DNA binding by NF-KB and thereby down-regu- late the transcription of genes dependent on N F - K - B ( ~ ~ - ~ ~ ) . By analogy with the interactions seen with AP-1, this action of the GR involves direct binding to the p65 subunit and requires the DBD and possibly regions in the LBD of the r e ~ e p t o r ( ~ ~ - ~ ~ ) (Fig. 2). Thus, it will be interesting to determine if the mutations in the DBD identified by Heck et

(Fig. 1) affecting GR-AP-1 interactions also disrupt interactions between GR and p65.

In contrast to the collagenase/AP-1 model, it has recently been reported that some of the negative effects of glucocorticoids are dependent upon protein synthesis and that glucocorticoids induce the synthesis of the inhibitor protein I K B ( ~ ~ , ~ * ) . In unstimulated cells, NF-KB is retained in the cytoplasm bound to IKB. Upon stimulation the IKB component is phosphorylated, resulting in its breakdown and release of NF-KB, which can then enter the nucleus and activate target genes. Thus, an increase in the synthesis of IKB would result in less NF-KB in the nucleus. The relative importance of these two mechanisms to the immunosuppressive actions of glucocorticoids remains to be established.

Interactions with putative coactivator proteins GR-.rl -affinity chromatography has revealed a number of

specifically interacting proteins from yeast whole cell extracts, although the identity of the proteins remains to be determined(’*). Recently, Zeiner and Gehring(3g) reported the isolation of cDNA for a 46 kDa protein (RAP46), of unknown function, which bound to a number of nuclear receptors, including the GR, in a ligand-dependent manner (Table 1). A novel approach, in which the GR crosslinked to a 32P-labelled GRE DNA sequence was used as a probe, identified a protein of 170 kDa (GR interacting protein, GRIP170), which co-purified with an activity that gave a modest but reproducible stimulation of GR-induced tran- s~ription(*~). In addition the GR has been reported to interact with a protein termed GRIP1 (40) and the CREB-binding protein, CBP(41), in a ligand-dependent manner (Table 1 and Fig. 3). It is now known that GRIP1 belongs to a family of related proteins that have been identified in different laboratories as binding to the progesterone receptor (PR)(42) and ER(41,43), and termed SRC-1 (steroid receptor coactivator 1) and TIF-2 (transcriptional intermediary protein-2) (see Table 1). GRIP1 is actually the mouse homologue of TIF-2(43). Significantly, CBP, SRC-1 and the TR have been shown to form a ternary complex, at least in an in vitro protein-protein interaction assay(41). With respect to the GR mechanism of action, it will be important to map which domains of the receptor interact with CBP and SRC proteins in order to determine if there is any overlap of the binding sites and to test if the receptor transactivation domains participate directly in the binding.

Interestingly, CBP appears to be a general coactivator required for the transactivation function of a number of transcription factors, including AP-1, in addition to nuclear receptors. Thus GR antagonism of the AP-1 signalling path- way has been suggested to be the result of competition for a limiting pool of CBP, and indeed ectopically expressed CBP relieves GR inhibition of AP-1 activation(41). DNA binding and transactivation activities of the GR, however, have been distinguished from the trans-repression function by mutational analysis and the use of steroid analogues(32). In light of this it will be of importance to map the interaction domains on the GR and CBP, and to compare the effect of receptor trans-repression mutations and steroid analogues on GR/AP-1 and GR/CBP binding, in order to unravel the complex nature of GR/AP-1 interactions.

Role of chromatin structure on gene activation by the GR Many of the experiments that have contributed to our under- standing of gene activation by the GR have used non-chromatin DNA templates in transient transfection or cell-free transcription experiments. Endogenous target genes for the GR are generally packaged as chromatin, imposing additional constraints on gene activity that are encountered by regulatory transcription factors during gene activation. For example, the GR is able to interact with enhancer sequences in the tyrosine aminotransferase gene and initiate gene activation in liver cells but not in fibroblasts, even

though fibroblasts can mediate GR-dependent activation from a transiently transfected template(44). Even in poten- tially active chromatin domains, chromatin structure appears to play a repressing role in the absence of regulatory stimuli such that access of transcription factors to their binding sites is often occluded. This appears to be the case for the tyrosine aminotransferase gene in which the GR plays a key role by disrupting two positioned nucleosomes in an enhancer region, in a hormone-dependent manner, thus allowing access to a range of other transcription factors(44). Removal of hormone or addition of a glucocorticoid antago- nist causes equally rapid reversion to the repressed state. Positioned nucleosomes are also found on the Long Termi- nal Repeat (LTR) of the Mouse Mammary Tumour Virus (MMTV), which is a target for induction by both the GR and the PR(45). In this case the GR can interact with its binding sites in vitro without disrupting the nucleosome that occu- pies the same sequences(46). Early studies suggested that the nucleosome in question was severely disrupted in vivo on addition of hormone. More recently, however, it has been reported that the nucleosome remains in place during activation and that receptor binding causes local conformational changes at the dyad axis of symmetry of the nucleosome that allow access to NF-I and possibly OTF-1 (see ref. 47 and references therein). Thus on different promoters there appear to be at least two distinct mechanisms of chromatin derepression used by the GR, involving either nucleosome disruption or more localised changes allowing the binding of additional transcription factors.

The mechanism of chromatin derepression by the GR and other transcriptional activator proteins is still unclear.

RIP140

SRC-1 +/-

+I- 0 TIF-2

RING mow

Fig. 3. Schematic representations of receptor interacting proteins A R A ~ O ( ~ ~ ) , RIP140(22~62), SRC-1 (42), TIF-1(56) and TIF-2(43). The solid bar above each protein represents the region involved in receptor binding, while S, T, 0 represent serine-, threonine-, proline- and glutamine-rich domains and +/- are clusters of charged residues. The hatched boxes indicate the RING and BROMO domains of TIF-l(56).

Fig. 4. Model for transcriptional activation by the GR. The AF- l ls l and AF-2 transactivation domains of the GR are shown interacting directly or indirectly (via SRC and CBP) with the transcriptional machinery. The coactivators SRC and CBP are represented as part of a coactivator complex, which may contain additional, as yet unidentified, proteins. ? represents unidentified target proteins for the GR. See text for full discussion.

One possibility is that receptor transactivation domains interact directly or indirectly with the Swi/Snf complex of proteins that has been implicated in chromatin derepression in yeast (reviewed in ref. 48). Homologous proteins are found in Drosophila and humans and they have been shown to affect chromatin structure in vitro in an ATP-dependent manner (see ref. 48 and references therein). Interestingly, the gene activation potency of the GR in yeast is severely reduced in swi/snf mutants(4g). Recently, additional polypeptides have been described as components of the swi/snf complex, and one of these, swp73p, was identified as being important for GR function (see ref. 50 and references therein). So far no direct interaction between the GR transactivation domains and the Swi/Snf complex has been reported, although the DNA-binding domain of the GR has been reported to interact with the Swi3p subunit of the com- p l e ~ ( ~ ~ ) (Fig. 2). Several other protein complexes, such as TFIID, the holo-RNA polymerase II and the ADA adapter complex, have been shown to be important for later steps in transcriptional transactivation in cell-free systems using non-chromatin templates. It remains a possibility, however, that some or all of these complexes also function in chromatin derepression prior to their roles in pre-initiation complex formation and subsequent events. Consistent with this, it has recently been reported that Swi/Snf proteins can be components of holo-RNA polymerase and that the Gcn5 subunit of the ADA complex contains a histone transacetylase activity(S2).

Future directions Within the last 2-3 years a plethora of novel proteins interacting with members of the nuclear receptor superfamily have been identified by two-hybrid screening, functional complementation studies, far western blotting and expression cloning (Table 1). Using a yeast two-hybrid screen, two groups have recently described the isolation of

a protein (Trip-I or mSUG1) that interacts with the TR and retinoic acid receptor (RAR) in a ligand-dependent manner(53,54). Trip-I is the mammalian homologue of the yeast protein Sugl, which is a component of the holo-RNA polymerase II complex (see ref. 53 and references therein). Chambon and co-workers have described a number of distinct proteins binding to the C-terminal transactivation domain of several nuclear receptors, including a novel 30 kDa TBP-associated factor(55), which binds directly to the C- terminal transactivation function of the ER and mediates ER-dependent transcriptional activation, and a 1 12 kDa protein (termed TIF-1)(56), which enhanced the activity of AF- 2/zc domain of the RAR and RXR when co-expressed in yeast. Proteins of 140 kDa and 160 kDa have also been shown to bind to the AF-2/rc region of the ER(57,58) and RAR(59), one of which, the 140 kDa protein referred to as RIP140, has been cloned(22). The 160 kDa protein, p160, may be related to a family of proteins termed SRC-1, SRC- la-e and GRIPI/TIF-2, which bind to and potentiate the lig- and-dependent transactivation of several members of the nuclear receptor s ~ p e r f a m i l y ( ~ ~ - ~ ~ ) . In contrast to the above receptor-interacting proteins, which may be general targets for nuclear receptors, one novel protein has been described that bound to the androgen receptor (AR) and selectively enhanced AR-dependent transactivation in a cell culture system(60). Two proteins, termed MBFl (18 kDa) and MBF2 (22 kDa), have been shown to be necessary for the Drosophila orphan receptor BmFTZ-F1 transactivation function in vitro, possibly by functioning as a bridge between the receptor and TBP(61).

Fig. 3 shows schematically several of the cloned receptor-interacting proteins, summarising features of their pri- mary amino acid sequence and the region(s) involved in receptor binding. Recently, RIP140 has been shown to contain two receptor binding sites localised in the N- and C-terminal halves of the protein(62). Furthermore, although SRC- 1 and GRIPITTIF-2 are clearly members of the same protein family, the regions identified as binding to nuclear receptors do not overlap (Fig. 3) (see ref. 43). Thus these proteins may have more than one receptor binding site, which would allow multiple interactions with one receptor or, more intriguing, simultaneous binding to two or more receptors. Significantly, reconstitution of ligand-dependent transcriptional activation with the individual N- and C-terminal portions of the ER is enhanced in the presence of SRC-1 (63). Although the underlying mechanism remains unknown, it is tempting to specu- late that SRC-1 may physically facilitate the interaction of the N- and C-terminal domains of the receptor. Alternatively, and as discussed by Katzenellenbogen and c o - ~ o r k e r s ( ~ ~ ) , the effect of SRC-1 could be indirect and involve other proteins. It is therefore of note that nuclear receptors and SRC- l /p l60 both interact with another coactivator protein, CBP, to form a ternary complex(41). Furthermore, a complex of up to nine polypeptides, associated with the TR, has been shown to mediate receptor-dependent transactivation in in

vitro transcription assays(64). The identity of the polypeptides remains unknown, but the complex did not contain CBP, RIP140 orTIF-1(64).

With the identification and cloning of putative coactivators for the GR and other members of the nuclear receptor superfamily we may soon have the reagents necessary to determine the underlying mechanisms for receptor-dependent transactivation (Fig. 4), as well as cell- and promoter- specific receptor action (see refs 55,65) and synergism between receptors and other transcription factors (see refs 66,67). In the meantime, questions that remain to be answered include the identification of (1) downstream targets for receptor-interacting proteins identified so far and (2) target proteins for receptor AF-1 or N-terminal transactivation functions, such as GR-rl. Most of the proteins identified to date appear to be ubiquitously expressed and to interact with multiple members of the nuclear receptor superfamily, although perhaps with differing affinities and specifici- ties(62,54). It is possible that all these proteins function as coactivators, mediating a link between the receptor and the general transcription machinery, but it cannot be excluded that they could act at some other step in the receptor path- way, for example in nuclear transport, DNA binding or receptor stability and turnover. The challenge now is to under- stand the role(s) these different interacting proteins play in receptor-dependent transactivation.

Acknowledgements This work was supported by a grants from the Swedish Medical Research Council (1 3X-2819), Swedish Natural Science Council (K-KV9756-301) and the Swedish cancer Society (3466-Bg4-02XBB). We thank members of the nuclear receptor unit for helpful discussions.

References 1 Dalman, F. C., Scherrer, L. C., Taylor, L. P., Akil, H. and Pratt, W. B. (1991). Localisation of the 90-kDa heat shock protein-bindig site within the hormone- binding domain of the glucocorticoid receptor by peptide competition J. Biol. Chem. 266,3482-3490. 2 Zilliacus, J., Wright, A. P., Carlstedt-Duke, J. and Gustafsson, J.-A. (1995). Structural determinants of DNA-binding specificity by steroid receptors. Mol Endocrinol. 9, 389-400. 3 Bourguet, W., Ruff, M., Chambon, P., Gronemeyer, H. and Moras, D. (1995). Crystal structure of the ligand-binding domain of the human nuclear receptor RXR- alpha. Nafure 375,377-382. 4 Renaud, J. P. eta/. (1995). Crystal structure of the RAR-gamma Iigand-binding domain bound to all-trans retinoic acid. Nature 378, 681 -689. 5 Wagner, R. L. eta/. (1995). A structural role for hormone in the thyroid hormone receptor. Nafure 378, 690-697. 6 Danielsen, M., Northrop, J. P. and Ringold, G. M. (1986). The mouse glucocorticoid receptor. mapping of functional domains by cloning, sequencing and expression of wild-type and mutant receptor proteins. EMBO J. 5, 251 3-2522. 7 Giguere, V., Hollenberg, S. M., Rosenfeld, M. G. and Evans, R. M. (1986). Functional domains of the human glucocorticoid receptor. Cell46, 645-52. 8 Godowski, P. J., Picard, D. and Yamamoto, K. R. (1988). Signal transduction and transcriptional regulation by glucocorticoid receptor-LexA fusion proteins. Science241,812-6. 9 Hollenberg, S. M. and Evans, R. M. (1988). Multiple and cooperative transactivation domains of the human glucocorticoid receptor. Cell55, 899-906. 10 Dahlman-Wright, K. et a/. (1994). Delineation of a small region within the

major transactivation domain of the human glucocorticoid receptor that mediates transactivation of gene expression Proc. Natl Acad. Sci USA 91, 1619-1623 11 Ptashne, M. and Gann, A. A. F. (1990). Activators and targets Nature 346. 329-331. 12 Danielian, P. S., White, R., Lees, J. A. and Parker, M. G. (1992) Identification of a conserved region required for hormone dependent transcriptional activation by steroid hormone receptors. EMBO J 11, 1025-1 033 13 Dahlman-Wright, K. et a/. (1 995). Structural characterization of a minimal functional transactivation domain from the human glucocorticoid receptor. Proc. Natl Acad. Sci. USA 92, 1699-1 703 14 McEwan, I. J., Dahlman-Wright, K., Ford, J. and Wright, A. P. H. (1996). Functional interaction of the c-Myc transactivation domain with the TATA binding protein evidence for an induced fit model of transactivation domain folding. Biochemistry35,9584-9593. 15 Dahlman-Wright, K. and McEwan, I. J. (1996). Structural studies of mutant glucorticoid receptor transactivation domains establish a link between transactivation activity in vivo and u-helix forming potential in vifro Biochemistry 35, 1323-1327. 16 Shen, F., Triezenberg, S. J., Hensley, P., Porter, D. and Knutson, J. R. (1996). Transcriptional activation domain of the herpesvirus protein VP16 becomes conformationally constrained upon interaction with basal transcription factors. J. Biol. Chem. 271, 4827-37. 17 Stargell, L. A. and Struhl, K. (1996). Mechanisms of transcriptional activation in vivo: two steps forward Trends Genet. 12, 311-315. 18 McEwan, I. J. eta/. (1 995). Mechanisms of transcriptional activation by nuclear receptors. studies on the human glucocorticoid receptor r l transactivation domain Muf. Res. 333, 15-22. 19 Ing, N. H., Beekman, J. M., Tsai, S. Y., Tsai, M. J. and O'Malley, B. W. (1992). Members of the steroid hormone receptor superfamily interact with TFllB (S300- 11). J. Biol. Chem. 267, 17617-17623. 20 Baniahmad, A,, Ha, I., Reinberg, D., Tsai, S., Tsai, M. J. and O'Malley, B. W. (1 993). Interaction of human thyroid hormone receptor beta with transcription factor TFllB may mediate target gene derepression and activation by thyroid hormone. Proc. NatlAcad. Sci. USA 90, 8832-6. 21 Blanco, J. C. et a/. (1995). Transcription factor TFllB and the vitamin D receptor cooperatively activate Iigand-dependent transcription. Proc. Nall Acad So. USA 92, 1535-9. 22 Cavailles, V. et a/. (1995). Nuclear factor RIP140 modulates transcriptional activation by the estrogen receptor EMBO J. 14, 3741-3751 23 Hadzic, E. et a/. (1995). A 10-amino-acid sequence in the N- terminal a/b domain of thyroid hormone receptor alpha is essential for transcriptional activation and interaction with the general transcription factor TFllB Mol Cell. Biol 15,4507- 451 7. 24 Malik, S. and Karathanasis, S. (1995) Transcriptional activation by the orphan nuclear receptor ARP-1. Nucl Acids Res. 23, 1536-1543. 25 Sadovsky, Y. eta/. (1 995). Transcriptional activators differ in their responses to overexpression of TATA-box-binding protein Mo/ Cell. Biol 15. 1554-1 563. 26 Schulman, 1. G., Chakravarti, D., Juguilon, H., Romo, A. and Evans, R. M. (1995). Interactions between the retinoid x receptor and a conserved region of the tata-binding protein mediate hormone-dependent transactivation. Proc Natl Acad Sci. USA 92, 8288-8292. 27 Fondell, J. D., Brunel, F., Hisatake, K. and Roeder, R. G. (1996). Unliganded thyroid hormone receptor alpha can target tata-binding protein for transcriptional repression. Mol. Cell. Biol. 16.281-287. 28 Malik, S. and Karathanasis, S. K. (1 996). TFIIB-directed transcriptional activation by the orphan nuclear receptor hepatocyte nuclear factor 4. Mol. Cell

29 Eggert, M. eta/. (1 995). A fraction enriched in a novel glucocorticoid receptor- interacting protein stimulates receptor-dependent transcription in vitro. J, B!o/ Chem. 270,30755-30759. 30 Pfahl, M. (1993). Nuclear receptoriAP-1 interaction Endocrinol. Rev. 14. 651- 658. 31 Cato, A. C. B. and Wade, E. (1996). Molecular mechanisms of antiinflammatory action of glucocorticoids. BioEssays 18, 371 -378 32 Heck, S. eta/. (1994). A distinct modulating domain in glucocorticoid receptor monomers in the repression of activity of the transcription factor AP-1. EMBO J 13,4087-4095. 33 Pearce, D. and Yarnarnoto, K. R. (1993) Mineralocorticoid and glucocorticoid receptor activities distinguished by nonreceptor factors at a composite response element. Science259. 1161.1165. 34 Ray, A. and Prefontaine, K. E. (1994). Physical association and functional antagonism between the p65 subunit of transcription factor NF-kappa B and the glucocorticoid receptor Proc. Natl Acad. Sci. USA 91,752-756. 35 Caldenhoven, E. et a/. (1995). Negative cross-talk between RelA and the glucocorticoid receptor: a possible mechanism for the antiinflammatory action of glucocorticoids. Mol. Endocrinol. 9,401 -412. 36 Scheinman, R. I. eta/. (1995). Characterization of mechanisms involved in

Bid. 16, 1824-1831.

transrepression of NF-kappa B by activated glucocorticoid receptors. Mol. Cell.

37 Auphan, N., DiDonato, J. A., Rosette, C., Helmberg, A. and Karin, M. (1 995). Immunosuppression by glucocorticoids: inhibition of NF-kappa B activity through induction of I kappa B synthesis. Science270. 286-90. 38 Scheinman, R. I., Cogswell, P. C., Lofquist, A. K. and Baldwin, A., Jr. (1995). Role of transcriptional activation of I kappa B alpha in mediation of immunosuppression by glucocorticoids. Science 270, 283-286 39 Zeiner, M. and Gehring, U. (1 995). A protein that interacts with members of the nuclear hormone receptor family - identification and cDNA cloning Proc Nafl Acad. So. USA92, 11465-11469. 40 Hong, H., Kohli, K., Trivedi, A., Johnson, D. L. and Stallcup, M. R. (1 996). GRIP1, a novel mouse protein that serves as a transcriptional coactivator for the hormone binding domains of steroid receptors Proc. Nafl Acad. Sci. USA 93, 4948-4952. 41 Kamei, Y. et a/. (1996). A CBP integrator complex mediates trasncriptional activation and AP-1 inhibition by nuclear receptors Cell85, 403-41 4. 42 Onate, S. A,, Tsai, S. Y., Tsai, M.-J. and O'Malley, B. W. (1995). Sequence and characterisation of a coactivator for the steroid hormone receptor superfamily. Science270,1354-1357. 43 Voegal, J. J., Heine, M. J. S., Zechal, C., Chambon, P. and Gronemeyer, H. (1996) TIF2, a 160 kDa transcriptional mediator for the Iigand-dependent activation function AF-2 of nuclear receptors. EM50 J. 15, 3667-3675 44 Reik, A,, Schutz, G. and Stewart, A. F. (1991). Glucocorticoids are required for establishment and maintenance of an alteration in chromatin structure: induction leads to a reversible disruption of nucleosomes over an enhancer. EM50 J. 10.2569-2576. 45 Richard-Foy, H. and Hager, G. L. (1987). Sequence-specific positioning of nucleosomes over the steroid-inducible MMTV promoter. EMBO J. 6, 2321-2328. 46 Perlmann, T. and Wrange, 0. (1988). Specific glucocorticoid receptor binding to DNA reconstituted in a nucleosome. EM50 J. 7, 3073-3079. 47 Chavez, S., Candau, R., Truss, M. and Beato, M. (1995). Constitutive repression and nuclear factor 1 -dependent hormone activation of the mouse mammary tumor virus promoter in saccharomyces cerevisiae. Mol. Cell. Biol. 15. 6987-6998. 48 Peterson, C. L. and Tamkun, J. W. (1995). The SWI-SNF complex: a chromatin remodeling machine? Trends Biochem. So. 20, 143-146. 49 Yoshinaga, S. K., Peterson, C. L., Herskowitz, I. and Yamamoto, K. R. (1992). Roles of SWIl, SW12, and SW13 proteins for transcriptional enhancement by steroid receptors Science 258. 1598-1 604. 50 Carins, B. R., Levinson, R. S., Yamamoto, K. R. and Kornberg, R. D. (1996). Essential role of Swp73p in the fuction of yeast Swi/Snf complex. Genes Dev. 10, 2131-2144. 51 Wilson, C. J. etal. (1996). RNA polymerase II holoenzyme contains swi/snf regulators involved in Chromatin remodeling. Cell84, 235-244. 52 Brownell, J. E. et a/. (1996). Tetrahymena histone acetyltransferase A a homolog to yeast Gcn5p linking histone acetylation to gene activation. Cell 84,

53 Lee, J. W., Ryan, F., Swaffield, J. C., Johnston, S. A. and Moore, D. D. (1995). Interaction of thyroid-hormone receptor with a conserved transcriptional mediator. Nature 374,91-94. 54 vom Baur, E. etal. (1 996). Differential ligand-dependent interactions between the AF-2 activating domain of nuclear receptors and the putative transcriptional intermediaryfactors mSUGl and TIFl. EMBOJ. 15, 110-124 55 Jacq, X., Brou, C., Lutz, Y., Davidson, I., Chambon, P. and Tora, L. (1994)

BiOl. 15, 943-953.

843-851

Human TAF(11)30 is present in a distinct TFllD complex and is required for transcriptional activation by the estrogen receptor. Cell79, 107-1 17. 56 LeDouarin, B. eta/. (1 995). The N-terminal part of TIF1, a putative mediator of the Iigand-dependent activation function (AF-2) of nuclear receptors, is fused to B- raf in the oncogenic protein T18 EMBO J 14,2020-2033. 57 Cavailles, V., Dauvois, S., Danielian, P. S. and Parker, M. G. (1994) Interaction of proteins with transcriptionally active estrogen receptors Proc Natl AcadSci USA91.10009-10013. 58 Halachmi, S. et a/. (1994). Estrogen receptor-associated proteins: possible mediators of hormone-induced transcription. Science 264, 1455-1 458. 59 Kurokawa, R. eta/. (1 995). Polarity-specific activities of retinoic acid receptors determined by a co-repressor Nature 377,451 -454 60 Yeh, S. and Chang, C. (1996). Cloning and characterisation of a specific coactivator. ARA70, for the androgen receptor in human prostate cells. Proc Nafl Acad. So. USA 93,5517-5521 61 Li, F.-Q., Ueda, H. and Hirose, S. (1994) Mediators of activation of fushi tarazu gene transcription by BmFTZ-Fl. Mol. Cell. Biol 14, 3013-3021 62 CHorset, F., Dauvois, S., Heery, D. M., Cavailles, V. and Parker, M. (1996). RIP-140 interacts with multiple nuclear receptors by means of two distinct sites. Mol. Cell. Biol. 16.6029-6036. 63 Mclnerney, E. M., Tsai, M.-J., O'Malley, B. W. and Katzenellenbogen, B. S. (1996). Analysis of estrogen receptor transcriptional enhancment by a nuclear hormone receptor coactivator. Proc. Nafl Acad. So. USA 93, 10069-1 0073. 64 Fondell, J. D., Ge, H. and Roeder, R. G. (1996). Ligand induction of a transcriptionally active thyroid hormone receptor complex. Proc. Nafl Acad. So. USA 93,8329-8333 65 May, M., Mengus, G., Lavigne, A.-C., Chambon, P. and Davidson, I. (1996). Human TAF1128 promotes transcriptional stimulation by activating function 2 of the retinoid X receptor. EM50 J. 15. 3093-3104 66 Schule, R., Muller, M., Kaltschmidt, C. and Renkawitz, R. (1988). Many transcription factors interact synergistically with steroid receptors. Science 242, 1418-1 420. 67 Strahle, U., Schmid, W. and Schutz, G. (1988). Synergistic action of the glucocorticoid receptor with transcription factors. EMBOJ. 7, 3389-3395. 68 Lee, J. W., Choi, H. S., Gyuris, J., Brent, R. and Moore, D. D. (1995). Two classes of proteins dependent on either the presence or absence of thyroid hormone for interaction with the thyroid hormone receptor. Mol. Endocrinol. 9, 243-254. 69 Schwerk, C., Klotzbucher, M., Sachs, M., Ulber, V. and Klein-Hitpass, L. (1 995). Identification of a transactivation function in the progesterone receptor that interacts with the TAFII110 subunit of the TFllD complex. J Biol Chem. 270, 21331-21338. 70 Schmitt, J. and Stunnenberg, H. G. (1993) The glucocorticoid receptor hormone binding domain mediates transcriptional activation in vitroin the absence of ligand. Nucl Acids Res. 21,2673-2681

lain J. McEwan*, Anthony P. H. Wright and Jan-Ake Gustafssont are at the Department of Biosciences, Karolinska Institute, NOVUM, Huddinge S-141 57, Sweden. *Present address: Department of Molecular and Cell Biology, University of Aberdeen, Institute of Medical Sciences, Foresterhill, Aberdeen AB25 2ZD, UK. tAlso at the Department of Medical Nutrition.