Regular article
High-resolution solution structure of the retinoid X receptor DNA-binding domaina

https://doi.org/10.1006/jmbi.1998.1908Get rights and content

Abstract

The retinoid X receptor (RXR) is a member of the nuclear hormone receptor superfamily of transcriptional regulators and plays a central role in the retinoid and, through its ability to heterodimerize with other nuclear hormone receptors, non-steroid signaling pathways. The DNA-binding and recognition functions of RXR are located in a conserved 83 amino acid residue domain that recognizes the consensus sequence AGGTCA. In order to provide a detailed picture of its structure, we have calculated a high-resolution solution structure of the C195A RXRα DNA-binding domain. Structures were calculated using 1131 distance and dihedral angle constraints derived from 1H, 13C and 15N NMR spectra. The structures reveal a perpendicularly packed, “loop-helix” fold similar to other nuclear hormone receptor DNA-binding domains and confirm the existence of the C-terminal helix, which was first observed in the low-resolution NMR structure. The C-terminal helix is well formed and is stabilized by packing interactions with residues in the hydrophobic core. The solution structure of RXR is very similar to that determined by X-ray crystallographic studies of the RXR-TR heterodimer complex with DNA, except that in the latter case no electron density was observed for residues corresponding to the C-terminal helix. Other differences between the X-ray and NMR structures occur in the second zinc-binding loop, which is disordered in solution. Heteronuclear 15N NOE measurements suggest that this loop has enhanced flexibility in the free protein.

Introduction

The retinoid X receptor, RXRα, is a central member of the nuclear hormone receptor superfamily of ligand-controlled transcription factors (Mangelsdorf & Evans, 1995). These receptors activate and repress gene expression by binding cooperatively as dimers to hormone response elements composed of two copies of a six nucleotide motif. RXRα recognizes the consensus half-site sequence AGGTCA Mangelsdorf et al 1990, Naar et al 1991. While a subset of the receptors, including the estrogen and glucocorticoid receptors (ER and GR), bind to half-site sequences arranged as inverted repeats, forming symmetric protein-protein dimers, the RXRα half-sites are arranged as direct repeats such that the protein must dimerize in a head-to-tail fashion (Mangelsdorf et al., 1995). Like other members of the superfamily, RXRα transactivates as a homodimer in response to binding of its activating ligand, 9-cis- retinoic acid Zhang et al 1992b, Mangelsdorf et al 1991. However, RXRα also regulates transcription by forming heterodimers with a number of other nuclear hormone receptors, including the thyroid receptor (TR), retinoic acid receptor (RAR), vitamin D receptor (VDR), and peroxisome proliferator-activated receptor (PPAR;Leid et al 1992, Yu et al 1991, Kliewer et al 1992a, Kliewer et al 1992b, Zhang et al 1992a, Bugge et al 1992). Selective recognition of the hormone response elements by the various heterodimers is encoded, in part, by variable spacing between the half-site repeats. RXRα heterodimers with PPAR, VDR, TR and RAR transactivate through response elements with one, three, four or five base-pairs between half-sites, respectively (Umesono et al., 1991).

The DNA-binding function of the receptor is located in an approximately 70 amino acid residue domain, which has a high degree of homology within the hormone receptor superfamily and includes nine conserved cysteine residues (Mangelsdorf et al., 1990). The structure of this DNA-binding domain (DBD) has been determined for ER Schwabe et al 1990, Schwabe et al 1993, GR Luisi et al 1991, Hard et al 1990, Van Tilborg et al 1995, Baumann et al 1993, RAR (Knegtel et al., 1993), TR (Rastinejad et al., 1995) and RXRα Lee et al 1993, Rastinejad et al 1995. The general characteristics of the DBD structures are that eight cysteine residues tetrahedrally coordinate two zinc ions forming two “loop-helix” zinc finger domains. The two helices of the zinc fingers pack perpendicularly to each other, forming a globular domain held together by hydrophobic side-chains that pack into the hydrophobic core.

A low-resolution structure of the RXRα DBD was initially determined in solution by NMR spectroscopy (Lee et al., 1993). Subsequently, a crystal structure of the RXRα DBD-TR DBD heterodimeric complex bound to a hormone response element with a four base-pair spacing (DR4) was determined (Rastinejad et al., 1995). The solution structure revealed an overall fold for the zinc finger domains similar to that of other nuclear hormone receptor and steroid receptor DBDs. At the C terminus, however, an additional helix (residues 202 to 209, seeFigure 1) was observed that has not been found in other DBD structures. This novel helix follows the conserved Gly-Met sequence and is formed by residues that are part of the “T-box”, a region that has been shown to play a key role in homodimeric binding to a hormone response element with a one base-pair spacing between half-sites (DR1;Wilson et al 1992, Lee et al 1993). Interestingly, this additional helix was not seen in the crystal structure of the heterodimer-DNA complex; no electron density was observed for this region of the protein (Rastinejad et al., 1995).

Here, we report the calculation of a high-resolution NMR structure of the RXRα DBD as a monomer in solution. The results unequivocally confirm the existence of the C-terminal helix in free RXRα. The improved structural definition allows a more detailed comparison of the free and DNA-bound states of the RXRα DBD, giving a better understanding of the structural changes that might occur as a consequence of dimerization and DNA-binding. Important questions remain about the nature of the cooperative binding of RXR to DNA, such as whether or not it involves local folding of the RXRα on the DNA and how RXRα accommodates such a variety of partners and response elements. We compare our structure to the RXRα-TR crystal structure to obtain some insight into these questions for RXRα on DR4. We also model the RXRα homodimer on DR1 to attempt to understand the role of the third helix in homodimer formation.

Section snippets

Design of expression construct

The construct used for the low-resolution structure (Lee et al., 1993) consisted of RXRα residues 130 to 229 and contained all nine of the conserved cysteine residues. This protein had a tendency to precipitate and, as a result, the maximum sample concentration for NMR studies was approximately 0.6 mM. It was suspected that cysteine 195, which neither coordinates zinc nor participates in an intramolecular disulfide linkage, might be causing precipitation through oxidative dimerization. The

Conclusions

The high-resolution solution structure of RXRα DBD has allowed a detailed comparison of the free and DNA-bound states of the protein. Two important structural features are revealed, both involved in formation of the dimer interface. The first is the presence of a C-terminal helix, which is well ordered at its N terminus as a consequence of packing interactions with the hydrophobic core but is frayed over the last four or five residues. Modeling of the RXRα homodimer on DNA suggests that the

Construction of C195A

In order to prevent oxidative dimerization of the RXRα DBD, the non-coordinating Cys195 was substituted with an alanine following an established PCR-based mutagenesis procedure (Higuchi et al., 1988) and using the fusion GST-RXRα(130-223) expression plasmid Lee et al 1993, Lee et al 1994 as a template. The following primers were designed to incorporate the mutation (underlined): 5′-GCCGCTACCAGAAGGCTCTGGCCATGGG-3′ and 3′-GCGATGGTCTTCCGAGACCGGTACCCG-5′.

The mutated gene was amplified using two

Acknowledgements

We thank David Case, Ishwar Radhakrishnan, John Love, Jarrod Smith, Tom Macke, Debbie Wuttke and Shohei Koide for helpful discussions; Xiang Li for making available the XPKASGN program; John Chung for assistance with NMR experiments; and Linda Tennant and Martine Reymond for technical assistance. This work was supported by grant GM36643 from the National Institutes of Health. S.M.A.H. and M.P.F. are recipients of American Cancer Society Fellowships.

References (60)

  • A.M Näär et al.

    The orientation and spacing of core DNA-binding motifs dictate selective transcriptional responses to three nuclear hormone receptors

    Cell

    (1991)
  • M Nilges

    Calculation of protein structures with ambiguous distance restraints. Automated assignment of ambiguous NOE crosspeaks and disulphide connectivities

    J. Mol. Biol.

    (1995)
  • J.W.R Schwabe et al.

    The crystal structure of the estrogen receptor DNA-binding domain bound to DNA: how receptors discriminate between their response elements

    Cell

    (1993)
  • D.S Sem et al.

    NMR spectroscopic studies of the DNA-binding domain of the monomer-binding nuclear orphan receptor, human estrogen related receptor. 2. The carboxyl-terminal extension to the zinc-finger region is unstructured in the free form of the protein

    J. Biol. Chem.

    (1997)
  • K Umesono et al.

    Direct repeats as selective response elements for the thyroid hormone, retinoic acid, and vitamin D3receptors

    Cell

    (1991)
  • M.A.A Van Tilborg et al.

    Structure refinement of the glucocorticoid receptor-DNA-binding domain from NMR data by relaxation matrix calculations

    J. Mol. Biol.

    (1995)
  • V.C Yu et al.

    RXRβa coregulator that enhances binding of retinoic acid, thyroid hormone, and vitamin D receptors to their cognate response elements

    Cell

    (1991)
  • H Baumann et al.

    Refined solution structure of the glucocorticoid receptor DNA-binding domain

    Biochemistry

    (1993)
  • H Berglund et al.

    Backbone dynamics of the glucocorticoid receptor DNA-binding domain

    Biochemistry

    (1992)
  • T.H Bugge et al.

    RXRα, a promiscuous partner of retinoic acid and thyroid hormone receptors

    EMBO J.

    (1992)
  • N.A Farrow et al.

    Backbone dynamics of a free and a phosphopeptide-complexed Src homology 2 domain studied by 15N NMR relaxation

    Biochemistry

    (1994)
  • S Grzesiek et al.

    Correlating backbone amide and side-chain resonances in larger proteins by multiple relayed triple resonance NMR

    J. Am. Chem. Soc.

    (1992)
  • S Grzesiek et al.

    Amino acid type determination in the sequential assignment procedure of uniformly 13C/15N-enriched proteins

    J. Biomol. NMR

    (1993)
  • S Grzesiek et al.

    A simple and sensitive experiment for measurement of JCCcouplings between backbone carbonyl and methyl carbons in isotopically enriched proteins

    J. Biomol. NMR

    (1993)
  • S Grzesiek et al.

    Multiple-quantum line narrowing for measurement of Hα-HβJ coupling in isotopically enriched proteins

    J. Am. Chem. Soc.

    (1995)
  • P Güntert et al.

    Improved efficiency of protein structure calculations from NMR data using the program DIANA with redundant dihedral angle constraints

    J. Biomol. NMR

    (1991)
  • T Härd et al.

    Solution structure of the glucocorticoid receptor DNA-binding domain

    Science

    (1990)
  • R Higuchi et al.

    A general method of in vitro preparation and specific mutagenesis of DNA fragments: study of protein and DNA interactions

    Nucl. Acids Res.

    (1988)
  • E.G Hutchinson et al.

    HERA - a program to draw schematic diagrams of protein secondary structures

    Proteins: Struct. Funct. Genet.

    (1990)
  • E.G Hutchinson et al.

    PROMOTIF-a program to identify and analyze structural motifs in proteins

    Protein Sci.

    (1996)
  • Cited by (0)

    a

    Edited by F. Cohen

    View full text