Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 9 (1), 3520

Multidomain Architecture of Estrogen Receptor Reveals Interfacial Cross-Talk Between Its DNA-binding and Ligand-Binding Domains

Affiliations

Multidomain Architecture of Estrogen Receptor Reveals Interfacial Cross-Talk Between Its DNA-binding and Ligand-Binding Domains

Wei Huang et al. Nat Commun.

Abstract

Human estrogen receptor alpha (hERα) is a hormone-responsive nuclear receptor (NR) involved in cell growth and survival that contains both a DNA-binding domain (DBD) and a ligand-binding domain (LBD). Functionally relevant inter-domain interactions between the DBD and LBD have been observed in several other NRs, but for hERα, the detailed structural architecture of the complex is unknown. By utilizing integrated complementary techniques of small-angle X-ray scattering, hydroxyl radical protein footprinting and computational modeling, here we report an asymmetric L-shaped "boot" structure of the multidomain hERα and identify the specific sites on each domain at the domain interface involved in DBD-LBD interactions. We demonstrate the functional role of the proposed DBD-LBD domain interface through site-specific mutagenesis altering the hERα interfacial structure and allosteric signaling. The L-shaped structure of hERα is a distinctive DBD-LBD organization of NR complexes and more importantly, reveals a signaling mechanism mediated by inter-domain crosstalk that regulates this receptor's allosteric function.

Conflict of interest statement

The authors declare no competing interests.

Figures

Fig. 1
Fig. 1
Contact residues between the DBD and LBD identified by footprinting. a Structural domains of hERα. Human ERα contains a DNA-binding domain (DBD; blue), a ligand-binding domain (LBD; green), and functions as a homodimer. b, c The crystal structures of DBD dimer (b light/dark blue) in complex with ERE–DNA (gray) (1HCQ.pdb), and of LBD dimer (c light/dark green) in complex with estradiol and a coactivator TIF2 peptide (1GWR.pdb). The C-terminal helix H12 of the LBD is highlighted (red). d Hydroxyl radical footprinting of hERα. High logPF values of six residues (red asterisks) indicate their involvement in domain contacts. Duplicates were performed and standard deviations were indicated. e Solvent accessibility surface area (SA) values of residue side chains calculated from the crystal structure of individual domains. f Correlation between logPF and SA values. Differentiation of the six contact residues (red dots) is shown from the rest of 14 residues (black dots). The latter have a Pearson’s correlation coefficient −0.77 (p-value = 0.001). g, h Structural mapping of contact residues. Contact residues (red) are Y191/Y195/W200 on the surface of the DBD (blue blobs) and I326/W393/L409 on the LBD (green blobs)
Fig. 2
Fig. 2
Overall architecture of the hERα homodimer revealed by data integration. a Fitting against experimental data. The fit of computationally generated conformations (dot) is simultaneously assessed against hydroxyl radical protein footprinting (φ2) and small-angle X-ray scattering (χ2). Lower χ2 and φ2 values are better in fitting. The best-fit ensemble structures lie at the bottom corner of the fit plot, below the red dashed line. b Ensemble of best-fit hERα structures. It contains both LBD monomers (light/dark green) and DBD monomers (light/dark blue). The C-terminal helix H12 of the LBD is in red, and ERE–DNA is in gray. The LBD–DBD connecting loops are shown as light green ribbons. The structure models (within the red circle) are within 3 Å Cα-RMSD of the best-fit structure. c A rotated view of the best-fit hERα structures. d Goodness of fit to measured SAXS data. Theoretical SAXS data were the ensemble average of the set of hERα structures above. The scattering intensity, log10I(q), is plotted as a function of the scattering angle (q). The goodness of fit χ2 = 1.2. Inserted is the Guinier plot with a linear fit, yielding the radius of gyration Rg = 38.0 ± 0.3 Å. The bottom graph shows residuals from subtraction between calculated and experimental profiles. A total of six scattering images were used and standard deviations were indicated. e Goodness of fit to footprinting data. Measured footprinting protection factors (logPF) are plotted against average accessible surface areas (SA) derived from the ensemble structures. Linear correlation coefficient is ρ = −0.95. A total of seven structures were used for ensemble calculations and standard deviations were indicated
Fig. 3
Fig. 3
The DBD–LBD interface and its functional relevance. a Close-up view of the DBD–LBD interface. Highlighted are contact residues with mutation sites (red) and the fluorescence site at W200 (blue). b Cartoon of interfacial residues. Dashed lines indicate a probability >75% of making a residue contact within the structure ensemble. c Effect of interfacial mutations on ER transcription activity. The Y191H mutation increases the transcription luciferase activity of the receptor, while N407A reduces the activity. Triplicates were carried out and standard deviations were indicated. d Tryptophan fluorescence site at W200. Shown are interactions between hydrophobic residues N407 (red circle), Y195 (gray circle), and W200 (blue circle) at the interface. e A schematic representation of tryptophan surroundings upon mutation. Illustrated are possible structural changes near W200 before and after mutation. f Quenching of tryptophan fluorescence. Emission fluorescence intensity is reduced in mutant N407A. A protein concentration of 0.1 mg/ml was used before and after mutation. Excitation at 295 nm
Fig. 4
Fig. 4
Transcriptional regulation of mutations at interfacial LBD residues. a Sites of LBD mutation (I326, Y328, P406, and L409), highlighted in black circles. bc Transient transfection reporter activity of the hERα and the Gal4-DBD/hERα-LBD fusion protein, using an ERE-TK-Luc (b) or a Gal4-TK-Luc reporter construct (c), respectively. Triplicates were carried out and standard deviations were indicated
Fig. 5
Fig. 5
Multidomain architecture and cross-talk at the domain interface of the hERα. a The hERα homodimeric complex contains both LBDs (green) and DBDs (blue). A hormone ligand (yellow) is capped underneath the LBD’s C-terminal helix H12 (ribbon). b The LBD–DBD interface consists of the LBD’s two β-strands, distant from the ligand-binding pocket and coactivator-binding sites. Disruption of this interfacial cross-talk, which serves as an allosteric channel to transmit the signaling of ligand binding from the LBD to a distant DBD, suppresses hormone-induced transcription. c Alteration of the domain cross-talk at the structural level is monitored by intrinsic tryptophan fluorescence (i.e., Trp200 in the middle of the interface as a probe in red blobs), using our genetically engineered hERα construct

Similar articles

See all similar articles

Cited by 6 articles

See all "Cited by" articles

References

    1. Kumar V, et al. Functional domains of the human estrogen receptor. Cell. 1987;51:941–951. doi: 10.1016/0092-8674(87)90581-2. - DOI - PubMed
    1. Deroo BJ, Korach KS. Estrogen receptors and human disease. J. Clin. Invest. 2006;116:561–570. doi: 10.1172/JCI27987. - DOI - PMC - PubMed
    1. Kumar V, Chambon P. The estrogen-receptor binds tightly to its responsive element as a ligand-induced homodimer. Cell. 1988;55:145–156. doi: 10.1016/0092-8674(88)90017-7. - DOI - PubMed
    1. Dahlman-Wright K, et al. International union of pharmacology. LXIV. Estrogen receptors. Pharmacol. Rev. 2006;58:773–781. doi: 10.1124/pr.58.4.8. - DOI - PubMed
    1. Chandra V, et al. Structure of the intact PPAR-gamma-RXR- nuclear receptor complex on DNA. Nature. 2008;456:350–356. doi: 10.1038/nature07413. - DOI - PMC - PubMed

Publication types

Substances

Feedback