1997 Oxford University Press Nucleic Acids Research, 1997, Vol. 25, No. 21 4307–4313
Allosteric interaction of the 1
α
,25-dihydroxyvitamin D
3
receptor and the retinoid X receptor on DNA
Jean-Pierre Kahlen
1and Carsten Carlberg
1,2,*
1Clinique de Dermatologie, Hôpital Cantonal Universitaire, CH-1211 Genève 14, Switzerlandand 2Institut für
Physiologische Chemie I, Heinrich-Heine-Universität Düsseldorf, D-40001 Düsseldorf, Germany
Received July 7, 1997; Revised and Accepted September 22, 1997
ABSTRACT
Genomic actions of the hormone 1α ,25-dihydroxy-vitamin D3 (VD) are mediated by the transcription
factor VDR, which is a member of the nuclear receptor superfamily. VDR acts in most cases as a heterodimeric complex with the retinoid X receptor (RXR) from specific DNA sequences in the promoter of VD target genes called VD response elements (VDREs). This study describes a mutation (K45A) of the VDR DNA binding domain that enhances the affinity and ligand responsiveness of VDR–RXR heterodimers on some VDREs. In analogy to a homologous mutation in the glucocorticoid receptor (K461A), this lysine residue appears to function as an allosteric ‘lock’. Interestingly, overexpression of RXR was found to reduce the responsiveness and sensitivity of wild type VDR to VD, but enhance the response of VDRK45A. Moreover, the transactivation domains of both VDR and RXR were shown to be essential for obtaining responsiveness of the heterodimers to VD and 9-cis retinoic acid (the RXR ligand). This indicates that RXR is an active rather than silent partner of the VDR on the VDREs tested. Taken together, transactivation by VDR–RXR heterodimers can be triggered individually by all components of the protein–DNA complex, but full potency appears to be reached through allosteric interaction.
INTRODUCTION
The specific binding of transcription factors to response elements, which are found in the promoter region of target genes, is a key step in the regulation of transcription. In addition to these specific protein–DNA interactions, are protein–protein interactions of the transcription factor with other regulatory proteins of central importance. In the case of nuclear receptors the interaction with a specific ligand inducing a conformational change plays an additional central regulatory role.
The nuclear receptor for 1α,25-dihydroxyvitamin D3 (VD), VDR, is a member of the nuclear receptor superfamily (1). The VDR mediates the pleiotropic genomic effects of VD on calcium homeostasis, cellular growth, differentiation and apoptosis (2). Nuclear receptors are characterized by a highly conserved DNA binding domain (DBD) containing two zinc finger structures (3).
The VDR DBD interacts with hexameric core binding motifs of the consensus sequence RGKWSA (R = A or G, K = G or T, W = A or T and S = C or G) generally found in VD response elements (VDREs) (4). In addition, the DBD provides a dimerization interface for cooperative dimerization with the DBDs of VDR partner receptors. VDR has been shown to form functional homodimers and heterodimers with nuclear receptors for retinoic acid (RA, RAR) and thyroid hormone (T3, T3R), but its main partner appears to be the retinoid X receptor (RXR) (5). However, like other nuclear receptors, the VDR interacts with its partner receptors most effectively through a dimerization interface within the ligand binding domain (LBD). VDR–RXR heterodimers bind with high affinity to response elements formed by two core binding motifs in a directly repeated orientation, which are spaced by three nucleotides (DR3-type VDREs) (6–8). Potent VDREs are also formed by motifs with four spacing nucleotides (DR4-type VDREs) (9) or by motifs in an inverted palindromic orientation with a distance of nine intervening nucleotides (IP9-type VDREs) (10). The selectivity of VDR–RXR heterodimers for these VDRE types is directed mainly by the use of different dimerization interfaces within the VDR DBD [(11), M.Quack and C.Carlberg, unpublished results].
Ligand binding mediates a structural change of the LBD that results in the exposure of a small, conserved region to the surface of the receptor (12). These few amino acids are located close to the receptor C-terminus and are referred to as the AF-2 domain (13,14). The AF-2 domain provides an interface for an interaction of the nuclear receptor with a co-activator that mediates signal ‘activation’ to the basal transcriptional machinery (15,16). RXR is known to be the nuclear receptor of 9-cis RA (17,18) and this raises the question, whether the RXR is transcriptionally active in VDR–RXR heterodimers, i.e., whether the RXR AF-2 domain is necessary for functionality of the VDR–RXR heterodimer. It has been suggested that within RAR–RXR and T3R–RXR heterodimers the RXR is transcriptionally silent (19), whereas for other RXR-partnered heterodimers (e.g., LXR–RXR, PPAR–RXR and NGFI-B–RXR) responsiveness to 9-cis RA has been shown (20,21). VD-stimulated VDR–RXR heterodimers can be superactivated by 9-cis RA (8), if limiting RXR amounts do not cause a 9-cis RA-dependent squelching effect (10,22).
Transcriptional activation through VDR–RXR heterodimers is therefore the result of a complex interplay between the DBD and the LBD of the VDR, the partner receptor RXR and the VDRE. *To whom correspondence should be addressed at: Institut für Physiologische Chemie I, Heinrich-Heine-Universität Düsseldorf, Postfach 10 10 07,
Nucleic Acids Research, 1997, Vol. 25, No. 21
4308
It can be hypothesized that all these components of the protein–DNA complex interact allosterically. The lysine residue (K461) of the P-box of the glucocorticoid receptor (GR) has been recently highlighted for its potential role as an allosteric ‘lock’ by restricting the receptor to less active conformations (23). The P-box is known to confer target gene specificity (24) and the lysine residue therein is highly conserved across members of the nuclear receptor superfamily. This implies that the suggested principle applies also to other members of the superfamily.
In this study the P-box lysine residue (K45) of the VDR was point mutated and functionally characterized. Moreover, the responsiveness of VDR–RXR heterodimers to VD and 9-cis RA was tested using VDR and RXR proteins that were deleted or point mutated in their AF-2 domain. The results suggest that the different domains of the heterodimeric complex functionally interact. MATERIALS AND METHODS
DNA constructs
The cDNA for human VDR and human RXRα were subcloned into the expression vector pSG5 (Stratagene) (8). They were used as templates for linear PCR reactions using Pfu DNA polymerase (Stratagene) with a profile of 94C for 30 s, 55C for 1 min and 68C for 11 min for 16 cycles. Mutated receptors VDRK45A, VDRK413STOP, VDRV418A, VDRL419A and RXRD444STOP (K = lysine, A = alanine, V = valine, L = leucine and D = aspartic acid) were generated using the following primer pairs:
K45A+ TATGACCTGTGAAGGCTGCGCAGGCTTCTTCA and K45A– TGAAGAAGCCTGCGCAGCCTTCACAGGTCATA, K413STOP+ AGTGCAGCATGTAGCTAACGC and
K413STOP– GCGTTAGCTACATGCTGCACT,
V418A+ GCTAACGCCCCTTGCGCTCGAAGTGTTTG and V418A– CAAACACTTCGAGCGCAAGGGGCGTTAGC, L419A+ AACGCCCCTTGTGGCCGAAGTGTTTGGCA and L419A– TGCCAAACACTTCGGCCACAAGGGGCGTT, D444STOP+ AGCTCATCGGGTAGACACCCATTGACA and D444STOP– TGTCAATGGGTGTCTACCCGATGAGCT.
Methylated template DNA was then digested selectively with
DpnI and supercompetent Escherichia coli XL-1 (Stratagene)
was transformed with non-digested, PCR-generated plasmid DNA. The respective point mutations were confirmed by sequencing. All VDREs were subcloned as a single copy into the
XbaI site of pBLCAT2 in front of the thymidine kinase (tk)
promoter driving the expression of the chloramphenicol acetyl transferase (CAT) gene, as described (25,26). The core sequences of the six VDREs used in this study are given in Figure 1. In vitro translation of nuclear receptors and DNA binding assays
Linearized cDNAs for VDRwt, VDRK45A and RXRwt were used for in vitro transcription as recommended by the supplier (Promega). Five µg of each RNA was mixed with 175 µl rabbit reticulocyte lysate, 100 U RNasin and 20 µM complete amino acid mixture (all from Promega) in a total volume of 250 µl and incubated at 30C for 180 min. Gel-purified HindIII/BamHI fragments of the VDRE-carrying pBLCAT2 derivatives were labelled by a fill-in reaction using [α-32P]dCTP and Klenow polymerase (Promega). In each binding reaction, 2.5 µl of in vitro translated VDRwt or VDRK45A was preincubated with 2.5 µl RXRwt or unprogrammed lysate for 10 min at room temperature in a total volume of 20 µl binding buffer [10 mM HEPES (pH 7.9),
Figure 1. Core sequences of the VDREs used in this study. One copy of each
element was subcloned into the XbaI site in front of the tk promoter of pBLCAT2. Arrows indicate the position and orientation of hexameric core binding motifs; bold numbers between the arrows show the number of intervening nucleotides. Small numbers indicate the nucleotide position of the response element within the promoter of the respective gene.
80 mM KCl, 1 mM DTT, 0.2 µg/µl poly(dI/dC) and 5% glycerol). Approximately 1 ng of radiolabelled probe (50 000 c.p.m.) was added and the incubation was continued for a further 20 min. Protein–DNA complexes were resolved on a 5% non-denaturing polyacrylamide gel (at room temperature) in 0.5× TBE [45 mM Tris, 45 mM boric acid, 1 mM EDTA (pH 8.3)].
Transfection and CAT assays
MCF-7 human breast cancer cells or Cos-7 SV40-transformed African Green monkey kidney cells were seeded into 6-well plates (1–1.5 × 105 cells per well) and grown overnight in phenol red-free DMEM supplemented with 5% charcoal-treated fetal calf serum (FCS). Liposomes were formed by incubating 1 µg of the reporter plasmid, indicated amounts of pSG5-based receptor expression vectors and 1 µg of the reference plasmid pCH110 (Pharmacia) with 15 µg N-[1-(2,3-dioleoyloxy)propyl]-N,N,N-trimethylammo-nium methylsulfate (DOTAP, Boehringer Mannheim) for 15 min at room temperature in a total volume of 100 µl. After dilution with 0.9 ml phenol red-free DMEM, the liposomes were added to the cells. Phenol red-free DMEM (500 µl) supplemented with 15% charcoal-treated FCS was added 4–8 h after transfection. At this time ligand (100 nM VD, 1 µM 9-cis RA, alone or in combination) was also added. The cells were harvested 40 h after treatment and CAT-assays were performed as described (27). The CAT activities were normalized to β-galactosidase activity and induction factors were calculated as the ratio of CAT activity of ligand-stimulated cells to that of mock-induced controls.
RESULTS
VDRK45A, a DBD mutant of the VDR that corresponds to the K461A mutation in the GR (23), was compared to VDRwt for its ability to bind as a homodimer or a heterodimer with RXRwt to the IP9-type VDREs of the mouse c-fos (28) and calbindin D9k (10) gene, the DR4-type VDRE of the rat pit-1 (9) gene and the
4309
Nucleic Acids Research, 1994, Vol. 22, No. 1Nucleic Acids Research, 1997, Vol. 25, No. 21 4309
Figure 2. Differential DNA binding properties of VDRwt–RXRwt and VDRK45A–RXRwt heterodimers. Gel shift experiments were performed using in vitro translated
VDRwt, VDRK45A and RXRwt alone or in combination (as indicated) with the corresponding 32P-labelled IP9-type, DR4-type and DR3-type VDRE probes (for core
sequences see Fig. 1). VDR–RXR heterodimer complexes were separated from non-specific (NS) complexes. A representative experiment is shown. The relative amount of VDR–RXR heterodimer formation was determined by scintillation counting of at least three independent experiments. The ratios between VDRK45A–RXRwt and VDRwt–RXRwt heterodimers were calculated for each VDRE and standard deviation is given in parentheses.
DR3-type VDREs of the mouse osteopontin (29), rat atrial natriuretic factor (ANF) (25) and human parathyroid hormone (PTH) (30) gene, respectively. Weak binding of VDRwt homodimers were demonstrated to the VDREs of the pit-1 and the osteopontin gene, however, no formation of VDRK45A homodimers did result (Fig. 2). VDRwt–RXRwt heterodimer formation was observed on all six VDREs ranging from high affinity binding on the pit-1 VDRE to low affinity binding on the PTH VDRE. VDRK45A– RXRwt heterodimers clearly differed from VDRwt–RXRwt hetero-dimers in their VDRE preference, showing reduced affinity to the VDREs of c-fos (38%), calbindin D9k (40%), pit-1 (50%) and osteopontin (65%), but increased affinity to the VDREs of ANF (185%) and PTH (286%) (Fig. 2). Comparable results were also obtained with nuclear extracts from VDRwt-, VDRK45A- and RXRwt-overexpressing Cos-7 cells (data not shown).
The pit-1 DR4-type and the ANF DR3-type VDRE were selected for a functional analysis in MCF-7 breast cancer cells, as both IP9-type VDREs showed low affinity for VDRK45A–RXRwt heterodimers and a single copy of the PTH VDRE fused to the tk promoter is not functional in reporter gene assays (M.Schräder and C.Carlberg, unpublished results). This cell line is known to endogenously express VDR and RXR (8). In the presence of relatively low amounts of endogenous RXR, a moderate overexpression of VDRK45A (1 µg of expression plasmid) in comparison to VDRwt resulted in a dominant negative effect on the VD responsiveness of both VDREs (Fig. 3). Interestingly, overexpression of RXRwt resulted in reduced VD inducibility of VDRwt–RXRwt heterodimers on both VDREs, but in an increased response of VDRK45A–RXRwt heterodimers. In accordance with
Figure 3. The functional effect of the DBD mutant VDRK45A. MCF-7 cells
were transfected with CAT reporter constructs containing either the DR3-type ANF VDRE or the pit-1 DR4-type VDRE (for core sequences see Fig. 1) and the indicated amounts of expression vectors for VDRwt, VDRK45A and RXRwt.
The cells were treated for 40 h with 100 nM VD, then β -galactosidase-normal-ized CAT activities were determined and fold induction was calculated in comparison to solvent-induced controls. Columns represent means from triplicates and bars indicate standard deviations.
Nucleic Acids Research, 1997, Vol. 25, No. 21
4310
Figure 4. Ligand sensitivity of VDR–RXR heterodimers. MCF-7 cells were transfected with the ANF DR3-type VDRE-tk CAT reporter construct and expression
vectors for VDRwt or VDRK45A (1 µg) either with or without RXRwt-expression plasmid (3 µg). Cells were treated with graded concentrations of VD for 40 h, then
β-galactosidase-normalized CAT activities were determined and fold induction was calculated in comparison to solvent-induced controls. Each data point represents the mean of triplicates.
the DNA binding studies on the ANF DR3-type VDRE (Fig. 2) at high RXRwt levels VDRK45A–RXRwt heterodimers were shown to be more potent than VDRwt–RXRwt heterodimers.
VDRwt–RXRwt and VDRK45A–RXRwt heterodimers were compared at low and high RXR expression levels for their ligand sensitivity from the ANF DR3-type VDRE (Fig. 4). The dose responses to VD indicated that high RXRwt-levels reduced not only the maximal response of VDRwt–RXRwt heterodimers, but augmented also the half-maximal activation value (an increase by a factor of 10 or more was estimated). In contrast, the dose response of VDRK45A–RXRwt heterodimers stayed about the same at both RXR concentrations.
These data suggested that RXR influenced VDRwt-mediated transactivation, possibly through its AF-2 domain. RXRD444STOP (Fig. 5F–I and O–R), a RXR mutant lacking the complete AF-2 domain, was therefore analyzed in comparison to RXRwt (Fig. 5B–E and K–N). The respective heterodimeric complexes with either VDRwt (Fig. 5A–I) or VDRK45A (Fig. 5J–R) mediated VD- and 9-cis RA-stimulated transactivation from the ANF DR3-type VDRE in Cos-7 cells. These cells express very low endogenous amounts of VDR, therefore overexpression of either VDRwt or VDRK45A was required to obtain measurable VD-stimulated transactivation. The endogenous RXR levels in Cos-7 cells were apparently also limiting for potent ligand stimulation (Fig. 5A and J), therefore overexpression of RXRwt up to an approximate equimolar ratio to VDR (Fig. 5B and C), increased stimulation by VD and 9-cis RA alone or in combination. If the response to a combination of VD and 9-cis RA is found to be higher than the response to VD alone, it is here referred to as superactivation. Comparable to the MCF-7 system, higher RXR levels resulted in a decrease of ligand responsiveness of VDRwt–RXRwt heterodimers (Fig. 5D and E). With VDRK45A (Fig. 5K–N) comparable results were obtained, but the overall responsiveness to 9-cis RA was reduced, whereas increasing RXR amounts resulted in increased stimulation with VD (Fig. 5M and N). With VDRwt at all RXRwt levels (Fig. 5B–E) a superactivation could be observed, whereas with VDRK45A at high RXRwt levels (Fig. 5M and N) superactivation was lost.
Superactivation as well as 9-cis RA responsiveness was also completely lost when using the RXRD444STOP mutant (Fig. 5G–I and O–R). In contrast to VDRK45A–RXRD444STOP heterodimers (Fig. 5P–R), the VD responsiveness of VDRwt–RXRD444STOP heterodimers was clearly reduced (Fig. 5F–H) and at high RXRD444STOP levels nearly abolished (Fig. 5I).
In order to test whether the VDR AF-2 domain influences the responsiveness of VDR–RXR heterodimers to 9-cis RA, the effects of AF-2 domain point mutations VDRL419A (Fig. 6B and E) and VDRL418A (Fig. 6C and F) as well as the AF-2 domain deletion VDRK413STOP were analyzed (Fig. 6D and G). RXRwt (Fig. 6A–D) and RXRD444STOP (Fig. 6E–G) were co-transfected into Cos-7 cells at approximate equimolar VDR:RXR ratios and ligand responsiveness of the respective VDR–RXR heterodimers were tested again on the ANF DR3-type VDRE. Compared to VDRwt (Fig. 6A), VDRL419A (Fig. 6B) and VDRV418A (Fig. 6C) clearly reduced the VD responsiveness of the respective hetero-dimers and VDRK413STOP (Fig. 6D) completely abolished it. However, the AF-2 domain mutations did not influence 9-cis RA responsiveness and superactivation of the respective VDR–RXR heterodimers (Fig. 6B–D). As shown in Figure 5G for VDRwt, the heterodimers of VDRL419A (Fig. 6E), VDRV418A (Fig. 6F) and VDRK413STOP (Fig. 6G), respectively, also did not display either 9-cis RA responsiveness nor superactivation with RXRD444STOP. DISCUSSION
The precise transcriptional activity of the VDR is cell type specific and is determined largely by the structure of the response element. Different VD target tissues can be characterized by expression levels of VDR, its main partner receptor RXR and their respective co-factors. The findings presented in this study provide evidence that the DBD of the VDR and the AF-2 domains of both the VDR and the RXR allosterically interact. This interaction appears to determine the ligand responsiveness of VDR–RXR heterodimers in contact with different VDREs.
In a previous report that used a Drosophila cell line model (10), a low ratio of RXR to VDR overexpression had been shown to
4311
Nucleic Acids Research, 1994, Vol. 22, No. 1Nucleic Acids Research, 1997, Vol. 25, No. 21 4311
Figure 5. Allosteric interaction of VDR and RXR. Cos-7 cells were transfected with the ANF VDRE-tk CAT reporter construct and indicated amounts of VDRwt-,
VDRK45A-, RXRwt- and RXRD444STOP-expression plasmids. Cells were treated with 100 nM VD and 1 µM 9-cis RA either alone or in combination for 40 h, then
β-galactosidase-normalized CAT activities were determined and fold induction was calculated in comparison to solvent-induced controls. Columns represent means from triplicates and bars indicate standard deviations.
mimic the inhibiting effects of 9-cis RA on VD signalling caused by limiting RXR amounts as reported from a rat osteosarcoma cell line (22). However, in MCF-7 and Cos-7 cells the respective endogenous RXR expression is apparently high enough to prevent such squelching effects. Interestingly, high amounts of RXR in these cells caused a clear decrease of VD responsiveness and sensitivity of VDRwt–RXRwt heterodimers. This observation is in contrast to the common assumption that high amounts of a partner receptor would increase ligand responsiveness. In fact, increasing RXR levels have a more prominent effect on the augmentation of the basal promoter activity than on ligand stimulated activity (data not shown).
High RXR levels stimulated the VD responsiveness of VDRK45A from DR3-type and DR4-type VDREs. In the case of the homodimer forming GR the lysine residue that is homologous to K45 (K461) has been defined as an allosteric ‘lock’ (23).
However, the VDR generally acts as a heterodimer with RXR. Therefore its functional response in part relies on its partner receptor. K461 has been shown to be involved in direct contact of the GR DBD to guanidine bases within the hexameric core binding motif (31) and, even though a crystal structure of the VDR DBD has not yet been resolved, a similar role can be extrapolated for the K45 residue in the VDR. At approximate equimolar VDR:RXR ratios, VDRK45A–RXRwt heterodimers compared to VDRwt–RXRwt heterodimers, showed a decreased affinity to IP9-type and DR4-type VDREs, but an increased affinity to DR3-type VDREs. This effect apparently depends on the spacing and the relative orientation of the two core binding motifs, but also seems to be directed by the sequence of the core binding motif. In case of a thymidine in the third position of the core binding motif that is contacted by the VDR (Fig. 1), the mutant VDRK45A appears to lead in general to a decrease of DNA
Nucleic Acids Research, 1997, Vol. 25, No. 21
4312
Figure 6. VDR–RXR heterodimers have two functional AF-2 domains. Cos-7 cells were transfected with the ANF VDRE-tk CAT reporter construct and with each
0.3 µg of the indicated VDR and RXR expression vectors. Cells were treated with 100 nM VD and 1 µM 9-cis RA either alone or in combination for 40 h, then
β-galactosidase-normalized CAT activities were determined and fold induction was calculated in comparison to solvent-induced controls. Columns represent means from triplicates and bars indicate standard deviations.
binding affinity, whereas a guanidine in this position has the opposite effect. This explains the effects observed with the DR3-type VDRE from the mouse osteopontin gene. Interestingly, both DR3-type VDREs that demonstrated effective function from the mutation K45A (from the rat ANF and the human PTH gene) have negative function in their respective natural promoter context. This supports a model (23), to which a class of receptor point mutations can be characterized by an inappropriate presentation of activation surfaces in repressing contexts, i.e., in a complex with co-repressors. In fact, mutated receptors such as VDRK45A would help in classifying VDREs by their differential interaction with VDR–RXR heterodimers.
Interestingly, a naturally occurring mutant of the lysine residue at position 45 into glutamic acid (K45E) has been reported (32). K45E has been shown to cause hereditary VD resistant rickets (32) and it can be postulated that K45A would show a similar functional profile.
The AF-2 domain mutations of VDR and RXR showed that both transactivation domains contribute to the activity of the heterodimeric complex. This supports the view that RXR is an active rather than a silent partner of the VDR. The approximate 2-fold stimulation of VDR–RXR heterodimers by 9-cis RA (in the absence of VD) on the rat ANF VDRE is low, but significant and is clearly dependent on a functional RXR AF-2 domain. Interestingly, a functional RXR AF-2 domain is necessary for a full responsiveness of VDR–RXR heterodimers to VD, but the AF-2 domain of the VDR does not seem to be necessary for a 9-cis RA responsiveness. AF-2 domain deletions for both VDR and RXR, or overexpression of AF-2 domain deleted RXR protein completely abolishes ligand responsiveness of VDR–RXR heterodimers. Moreover, the RXR AF-2 domain was found to be necessary to obtain superactivation of VD responsiveness by 9-cis RA with VDR–RXR heterodimers. This indicates that both receptors in the VDR–RXR heterodimer interact simultaneously with their specific co-factors. However, since VDR–RXR heterodimers are apparently a part of a multi-protein complex that includes several co-factors, CBP,
TFIIB and other so far unknown proteins (33–35), a more complex explanation for the allosteric effects within VDR–RXR heterodimers would postulate that these partner proteins mediate the communication between the AF-2 domains of VDR and RXR.
Taken together, the ligand responsiveness of VDR–RXR heterodimers can be triggered individually by the different factors, the structure of the VDRE, the relative amount of RXR, the conformations of the VDR DBD and the AF-2 domains of VDR and RXR. Although locally distinct, all the above components of the protein–DNA complex can interact allosterically, resulting either in enhanced or reduced activity of the VDR–RXR heterodimer.
ACKNOWLEDGEMENTS
We would like to thank P.Polly for critical reading of the manuscript and the LEO Research Foundation for support. REFERENCES
1 Mangelsdorf, D. J., Thummel, C., Beato, M., Herrlich, P., Schütz, G., Umesono, K., Blumberg, B., Kastner, P., Mark, M., Chambon, P. and Evans, R. M. (1995) Cell, 83, 835–839.
2 Walters, M. R. (1992) Endocrine Rev., 13, 719–764. 3 Freedman, L. P. (1992) Endocrine Rev., 13, 129–145. 4 Carlberg, C. (1995) Eur. J. Biochem., 231, 517–527. 5 Carlberg, C. (1996) Endocrine, 4, 91–105.
6 Umesono, K., Murakami, K. K., Thompson, C. C. and Evans, R. M. (1991) Cell, 65, 1255–1266.
7 Yu, V. C., Delsert, C., Andersen, B., Holloway, J. M., Devary, O. V., Näär, A. M., Kim, S. Y., Boutin, J.-M., Glass, C. K. and Rosenfeld, M. G. (1991) Cell, 67, 1251–1266.
8 Carlberg, C., Bendik, I., Wyss, A., Meier, E., Sturzenbecker, L. J., Grippo, J. F. and Hunziker, W. (1993) Nature, 361, 657–660. 9 Rhodes, S. J., Chen, R., DiMattia, G. E., Scully, K. M., Kalla, K. A.,
Lin, S.-C., Yu, V. C. and Rosenfeld, M. G. (1993) Genes Dev., 7, 913–932. 10 Schräder, M., Nayeri, S., Kahlen, J.-P., Müller, K. M. and Carlberg, C.
(1995) Mol. Cell. Biol., 15, 1154–1161.
11 Towers, T. L., Luisi, B. F., Asianov, A. and Freedman, L. P. (1993)
4313
Nucleic Acids Research, 1994, Vol. 22, No. 1Nucleic Acids Research, 1997, Vol. 25, No. 21 4313 12 Bourguet, W., Ruff, M., Chambon, P., Gronemeyer, H. and Moras, D.
(1995) Nature, 375, 377–382.
13 Danielian, P. S., White, R., Lees, J. A. and Parker, M. G. (1992) EMBO J.,
11, 1025–1033.
14 Durand, B., Saunders, M., Gaudon, C., Roy, B., Losson, R. and Chambon, P. (1994) EMBO J., 13, 5370–5382.
15 Horwitz, K. B., Jackson, T. A., Bain, D. L., Richer, J. K., Takimoto, G. S. and Tung, L. (1996) Mol. Endocrinol., 10, 1167–1177.
16 Torchia, J., Rose, D. W., Inostroza, J., Kamei, Y., Westin, S., Glass, C. K. and Rosenfeld, M. G. (1997) Nature, 387, 677–684.
17 Levin, A. A., Sturzenbecker, L. J., Kazmer, S., Bosakowski, T., Huselton, C., Allenby, G., Speck, J., Kratzeisen, C., Rosenberger, M., Lovey, A. and Grippo, J. F. (1992) Nature, 355, 359–361.
18 Heyman, R. A., Mangelsdorf, D. J., Dyck, J. A., Stein, R. B., Eichele, G., Evans, R. M. and Thaller, C. (1992) Cell, 68, 397–406.
19 Forman, B. M., Umesono, K., Chen, J. and Evans, R. M. (1995) Cell, 81, 541–550.
20 Perlmann, T. and Jansson, L. (1995) Genes Dev., 9, 769–782. 21 Willy, P. J., Umesono, K., Ong, E. S., Evans, R. M., Heyman, R. A. and
Mangelsdorf, D. J. (1995) Genes Dev., 9, 1033–1045. 22 Macdonald, P. N., Dowd, D. R., Nakajima, S., Galligan, M. A.,
Reeder, M. C., Haussler, C. A., Ozato, K. and Haussler, M. R. (1993)
Mol. Cell. Biol., 13, 5907–5917.
23 Starr, D. B., Matsui, W., Thomas, J. R. and Yamamoto, K. R. (1996)
Genes Dev., 10, 1271–1283.
24 Umesono, K. and Evans, R. M. (1989) Cell, 57, 1139–1146.
25 Kahlen, J.-P. and Carlberg, C. (1996) Biochem. Biophys. Res. Commun.,
218, 882–886.
26 Luckow, B. and Schütz, G. (1987) Nucleic Acids Res., 15, 5490. 27 Pothier, F., Ouellet, M., Julien, J.-P. and Guérin, S. L. (1992) DNA
Cell Biol., 11, 83–90.
28 Schräder, M., Kahlen, J.-P. and Carlberg, C. (1997) Biochem. Biophys. Res.
Commun., 230, 646–651.
29 Noda, M., Vogel, R. L., Craig, A. M., Prahl, J., DeLuca, H. F. and Denhardt, D. T. (1990) Proc. Natl. Acad. Sci. USA, 87, 9995–9999. 30 Demay, M. B., Kieran, M. S., DeLuca, H. F. and Kronenberg, H. M.
(1992) Proc. Natl. Acad. Sci. USA, 89, 8097–8101.
31 Luisi, B. F., Xu, W. X., Otwinowski, Z., Freedman, L. P., Yamamoto, K. R. and Sigler, P. B. (1991) Nature, 352, 497–505.
32 Rut, A. R., Hewison, M., Kristjansson, K., Luisi, B., Hughes, M. R. and O’Riordan, J. L. H. (1994) Clin. Endocrinol., 41, 581–590.
33 MacDonald, P. N., Sherman, D. R., Dowd, D. R., Jefcoat, S. C. and DeLisle, R. K. (1995) J. Biol. Chem., 270, 4748–4752.
34 Kamei, Y., Xu, L., Heinzel, T., Torchia, J., Kurokawa, R., Gloss, B., Lin, S.-C., Heyman, R. A., Rose, D. W., Class, C. K. and Rosenfeld, M. G. (1996) Cell, 85, 403–414.
