Published online 5 December 2022 Nucleic Acids Research, 2022, Vol. 50, No. 22 13063–13082 https://doi.org/10.1093/nar/gkac1119 The multivalency of the glucocorticoid receptor ligand-binding domain explains its manifold physiological activities Downloaded from https://academic.oup.com/nar/article/50/22/13063/6858817 by Fundacio Biblioteca Josep Laporte user on 09 October 2023 Alba Jiménez-Panizo1,2,3,† , Andrea Alegre-Martı́1,2,† , Theophilus T. Tettey3,† , Gregory Fettweis3 , Montserrat Abella1,2 , Rosa Antón4 , Thomas A. Johnson3 , Sohyoung Kim3 , R. Louis Schiltz3 , Israel Núñez-Barrios5 , Joan Font-Dı́az2,6 , Carme Caelles2,7 , Annabel F. Valledor2,6 , Paloma Pérez8 , Ana M. Rojas 5 , Juan Fernández-Recio9 , Diego M. Presman10 , Gordon L. Hager 3,* , Pablo Fuentes-Prior 4,* and Eva Estébanez-Perpiñá 1,2,* 1 Department of Biochemistry and Molecular Biomedicine, Faculty of Biology, University of Barcelona (UB), 08028 Barcelona, Spain, 2 Institute of Biomedicine of the University of Barcelona (IBUB), University of Barcelona (UB), 08028 Barcelona, Spain, 3 National Cancer Institute, National Institutes of Health, Bethesda, MD 20892-5055, USA, 4 Biomedical Research Institute Sant Pau (IIB Sant Pau), 08041 Barcelona, Spain, 5 Andalusian Center for Developmental Biology (CABD-CSIC). Campus Universitario Pablo de Olavide, 41013 Sevilla, Spain, 6 Department of Cell Biology, Physiology and Immunology, Faculty of Biology, University of Barcelona, 08028 Barcelona, Spain, 7 Department of Biochemistry and Physiology, Faculty of Pharmacy and Food Sciences, University of Barcelona, Barcelona 08028, Spain, 8 Instituto de Biomedicina de Valencia (IBV)-CSIC, 46010, Valencia, Spain, 9 Instituto de Ciencias de la Vid y del Vino (ICVV), CSIC - Universidad de La Rioja - Gobierno de La Rioja, 26007 Logroño, Spain and 10 IFIBYNE, UBA-CONICET, Universidad de Buenos Aires, Facultad de Ciencias Exactas y Naturales, Buenos Aires C1428EGA, Argentina. Received May 03, 2022; Revised October 28, 2022; Editorial Decision October 28, 2022; Accepted November 08, 2022 ABSTRACT Our results highlight the relevance of non-canonical dimerization modes for GR, especially of contacts The glucocorticoid receptor (GR) is a ubiqui- made by loop L1–3 residues such as Tyr545. Our tously expressed transcription factor that controls work illustrates the unique flexibility of GR’s LBD metabolic and homeostatic processes essential for and suggests different dimeric conformations within life. Although numerous crystal structures of the cells. In addition, we unveil pathophysiologically rel- GR ligand-binding domain (GR-LBD) have been re- evant quaternary assemblies of the receptor with im- ported, the functional oligomeric state of the full- portant implications for glucocorticoid action and length receptor, which is essential for its transcrip- drug design. tional activity, remains disputed. Here we present five new crystal structures of agonist-bound GR-LBD, along with a thorough analysis of previous structural INTRODUCTION work. We identify four distinct homodimerization in- Nuclear receptors (NRs) are transcription factors that con- terfaces on the GR-LBD surface, which can associate trol central physiological processes ranging from reproduc- into 20 topologically different homodimers. Biologi- tion and development to metabolism, homeostasis, and ul- cally relevant homodimers were identified by study- tradian rhythms (1,2). Steroid receptors are a NR subclass ing a battery of GR point mutants including crosslink- that includes the glucocorticoid receptor (GR/NR3C1) ing assays in solution, quantitative fluorescence mi- (Figure 1A), the androgen receptor (AR/NR3C4), the pro- croscopy in living cells, and transcriptomic analyses. gesterone receptor (PR/NR3C3), the mineralocorticoid re- ceptor (MR/NR3C2), as well as estrogen receptors ␣ and * To whom correspondence should be addressed. Tel: +34 93 403 11 19; Email: evaestebanez@ub.edu Correspondence may also be addressed to Pablo Fuentes-Prior. Email: fuentespriorpablo@gmail.com Correspondence may also be addressed to Gordon L. Hager. Tel: +1 240 760 6618; Email: hagerg@dce41.nci.nih.gov † The authors wish it to be known that, in their opinion, the first three authors should be regarded as Joint First Authors. Published by Oxford University Press on behalf of Nucleic Acids Research 2022. This work is written by (a) US Government employee(s) and is in the public domain in the US. 13064 Nucleic Acids Research, 2022, Vol. 50, No. 22 Downloaded from https://academic.oup.com/nar/article/50/22/13063/6858817 by Fundacio Biblioteca Josep Laporte user on 09 October 2023 Figure 1. The ligand-binding domain of GR self-associates in solution. (A) Schematic representation of GR domain organization. GR␣ and GR␤ are identical up to residue 727 (H10/11), but the last 50 (in GR␣) and 15 residues (in GR␤, green box) are unrelated. Other GR isoforms are shown (right). (B) Overall structure of the GR-LBD monomer. The module is shown in standard orientation in the middle of the panel (i.e. with H1 and H3 displayed facing the viewer and the AF-2 pocket on the left). Four additional views are shown to present other LBD areas. Models are depicted as cartoons with helices (blue), loops (pink) and beta-sheets (purple). DEX (salmon spheres) and SHP peptide (yellow cartoon) are shown. The BF-3 pocket is also labeled. (C) Sequence alignment of LBDs between WT human GR, two engineered variants used in several structures (PDBs 3CLD and 4CSJ), and the ancGR1 and 2 forms. Strictly conserved residues are white with black shading; other conservatively replaced residues are shaded gray. Residues mutated in the current study are marked with asterisks. (D, E) SPR analysis of GR-LBD self-association according to (D) 1:1 or (E) multisite models. The results of experiments conducted in duplicate are shown along with the calculated affinity constants (kD ). Data were fitted to the 1:1 and multisite models with ␹ 2 values of 4.52 and 1.25, respectively. Nucleic Acids Research, 2022, Vol. 50, No. 22 13065 ␤ (ER␣/NR3A1 and ER␤/NR3A2) (3–5). NRs display ent GR-LBD homodimers. We have identified the most fa- a modular architecture with an unstructured N-terminal vored GR homodimers and suggest how they can combine domain (NTD) followed by a ‘core’ formed by a highly into pathophysiologically relevant oligomeric assemblies in conserved DNA-binding domain (DBD), an interdomain cells. linker or hinge, and a C-terminal ligand-binding domain (LBD) (Figures 1A and Supplementary Figure S1) (6–8). MATERIALS AND METHODS GR binds glucocorticoids (GCs; either endogenous such as Peptides and proteins Downloaded from https://academic.oup.com/nar/article/50/22/13063/6858817 by Fundacio Biblioteca Josep Laporte user on 09 October 2023 cortisol, or synthetic, e.g. dexamethasone (DEX)) in an in- ternal cavity of the LBD. This ligand-binding pocket (LBP) A peptide corresponding to residues Gln18-Lys27 of the is allosterically coupled to activation function 2 (AF-2), a small heterodimer partner (SHP/NR0B2, box 1 motif; pocket that binds coregulators (9,10). A nearby cavity, topo- NH2-Q18 GAASRPAILYALLSSSLK27 -OH) was custom- logically equivalent to AR binding function-3 (BF-3) inter- synthesized at Pepmic. Recombinant ancGR2-LBD (cor- acts with cochaperones (11,12). Finally, the LBDs of GR responding to residues 529–777 of the human receptor) and the related oxosteroid receptors (AR, PR and MR) fea- cloned into a pMALCH10T vector was expressed as fusion ture a unique C-terminal extension termed F-domain (5,13) protein with an N-terminal maltose-binding protein (MBP) (Figure 1B). and a hexahistidine tag and purified to homogeneity using NR3C1 is constitutively expressed in nearly all vertebrate standard chromatographic procedures (38). cells. Upon ligand binding, GR translocates to the nucleus (14) where it binds to chromatin (15) and integrates sig- Crystallization and structure determination nals ranging from available ligands to chromatin remodel- ing complexes (16) to control cognate target genes (up to Purified, concentrated DEX-bound ancGR2-LBD was 17% of the human transcriptome (17)) to regulate inflam- combined with a 3-fold molar excess of SHP peptide and mation, cell proliferation, and differentiation in a tissue- incubated for one hour at RT. Drops of the ancGR2-LBD- specific manner (18–20). GR also antagonizes the activity of SHP mixture were equilibrated against 0.1 M Tris–HCl, pH other transcription factors such as activator protein 1 (AP- 8.0, 0.2 M sodium chloride, 2.0 M ammonium sulfate (P31 , 1) and nuclear factor kappaB (NF-␬B) (21). Even though I41 22 and I41 32 crystals); 0.1 M PIPES, pH 7.0, 0.1 M am- monomeric GR is believed to play a DNA-independent role monium acetate, 2.5 M sodium formate (P61 crystals); or 85 in modulating inflammation (22), more recent work sug- mM sodium cacodylate trihydrate, pH 6.5, 0.17 M sodium gests the relevance also of both direct DNA binding of GR acetate trihydrate, 25.5% (w/v) PEG8000, 15% (v/v) glyc- dimers/tetramers (23–28). In line with these manifold ac- erol (C2 crystals) using the sitting drop vapor-diffusion tivities, alterations in GR signaling pathways due to muta- method. Diffraction data were collected at 100 K at the tions in NR3C1 lead to impaired sensitivity to GCs, which ALBA-CELLS synchrotron and processed using MOS- may manifest as either GC resistance (Chrousos syndrome) FLM (http://www.mrc-lmb.cam.ac.uk/harry/mosflm/) and or hypersensitivity (29). GR is thus an important pharma- CCP4 (http://www.ccp4.ac.uk/). The crystal structures were cological target to treat inflammatory diseases, but chronic solved and refined using MOLREP, REFMAC5 and COOT GC use results in drug resistance and side effects (16). from the CCP4 package. Crystal packing was analyzed us- Structures of the DBD dimer, both free and DNA-bound ing PISA (http://www.ebi.ac.uk/), model quality was as- have been solved (30–33), and the LBD has been extensively sessed with MolProbity (http://molprobity.biochem.duke. studied in complex with agonists or antagonists (3,15,34– edu/), and figures were prepared with PyMOL (http://www. 37). To date, however, no structure of full-length (FL) GR pymol.org). has been solved. Thus, several key issues regarding GR ter- tiary and quaternary structure remain unresolved: what is Surface plasmon resonance (SPR) analyses the conformation adopted by dimeric GR and DNA-bound SPR analyses were performed at 25◦ C in a BIAcore T200 tetramers in vivo, and how do LBDs associate in these mul- instrument (GE Healthcare). Highly purified, DEX-bound timers? Are topologically distinct GR conformations pos- recombinant WT ancGR2-LBD and its Y545C and Y545A sible, and are they linked to specific biological functions mutants were diluted in 10 mM sodium acetate, pH 5.0, (e.g. activation vs. repression of transcription)? Knowledge and directly immobilized on CM5 chips (GE Healthcare) of the oligomeric conformations of oxosteroid receptors by amine coupling at densities between 300 and 400 reso- would aid in the design of novel selective, potent GR mod- nance units (RU). As a reference, one of the channels was ulators that may lessen side effects of current drugs. also amine-activated and blocked in the absence of protein. Here we present a systematic structure-and-function in- The running buffer was 50 mM HEPES, pH 7.2, 50 mM vestigation of GR multimerization using X-ray crystallog- Li2 SO4 , 5% glycerol, 5 ␮M DEX, supplemented with var- raphy, state-of-the-art bioinformatics tools, surface plas- ious concentrations of dithiothreitol (DTT). Sensorgrams mon resonance (SPR) and crosslinking experiments in solu- were analyzed with the BIAcore T200 Evaluation software tion, quantitative fluorescence microscopy in live cells, and 3.0 and fitted according to the Langmuir 1:1 and multisite RNA-seq analysis of relevant GR variants. We report five models. new structures of DEX-bound GR-LBD (GR-LBD·DEX) and integrate this information into the wealth of previous Crosslinking experiments structural data to generate the most thorough catalog to date of possible GR homodimers. Four distinct interfaces Purified recombinant ancGR2-LBD (33 ␮M) was in- have been observed to participate in 20 topologically differ- cubated with 100-fold molar excess of crosslinkers 1- 13066 Nucleic Acids Research, 2022, Vol. 50, No. 22 ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochlo- reaction. Melting curves were acquired from eight repli- ride (EDC, Pierce) or disuccinimidyl dibutyric acid (DSBU) cates to determine the average melting temperature (Tm). for 1–2 h at 37◦ C following the manufacturer’s instructions. GR-LBD samples (0.5 mg/mL) were prepared in 20 mM In some experiments the Y545C variant was incubated after HEPES, pH 8.0, 200 mM NaCl, 10% glycerol, 50 mM im- affinity purification for 30 min at room temperature without idazole, 1 mM DTT, 50 ␮M DEX, and centrifuged 5 min. further treatment. Samples of the reaction mixtures were at 14,000 rpm immediately before measurements. SYPRO boiled in the presence of Laemmli sample buffer, either Orange dye (Sigma-Aldrich) was firstly prepared at 80× in Downloaded from https://academic.oup.com/nar/article/50/22/13063/6858817 by Fundacio Biblioteca Josep Laporte user on 09 October 2023 reducing (EDC- and DSBU-crosslinked proteins) or non- the same buffer, starting from a 5,000× commercial dilu- reducing (in the case of the Y545C variant) and resolved by tion. The final concentration of the dye in each well was 5×. SDS-PAGE. The plates were sealed with optical-quality sealing film (Mi- croseal® B Seals, BIO-RAD) and centrifuged at 2,000 × g for 30 s. Samples were equilibrated for 60 s and analyzed Nano-LC–MS/MS mass spectrometry using a linear gradient from 16 to 95◦ C with increments CBB-stained bands corresponding to monomeric, dimeric of 1◦ C/min, recording the SYPRO orange fluorescence and tetrameric GR-LBD after crosslinking with EDC or throughout the gradient using the iQ5 Optical System Soft- DSBU were excised from the gels and subjected to in- ware 2.0. Values were fitted using the online tool JTSA with gel digestion following standard protocols. Briefly, excised the four-parameter logistic equation, and the calculated flu- bands were reduced (10 mM DTT) in 50 mM bicarbon- orescence midpoints were compared with an unpaired t-test ate buffer, pH 8.0, for 45 min at 56◦ C, alkylated (50 mM for equal variances using GraphPad Prism 8. iodoacetamide in 50 mM ammonium bicarbonate buffer for 30 min at 25◦ C) and digested with either trypsin alone Cell line generation and culture or followed by GluC treatment, or with chymotrypsin overnight at 37◦ C in 100 mM ammonium acetate buffer, pH Mammary adenocarcinoma 3617-derived GRKO cells for 8.0 (sequencing-grade endoproteases were from Promega). number and brightness assays (N&B) were transiently In the case of the Y545C mutant, proteins in the excised transfected with the various forms of mouse GR (WT, bands were directly alkylated without previous DTT treat- W718A, Y551A, Y551A-Dim, P643A, D647V, P643A- ment to prevent reduction of the Cys545–Cys545’ disulfide Tetra, D647V-Tetra) using jetOPTIMUS™ reagent (Poly- bridge. Plus) according to the manufacturer’s instructions. The cell Tryptic peptides were diluted in 1% formic acid and lines used for total RNA-sequencing (RNA-Seq) experi- loaded onto a 180 ␮m × 20 mm C18 Symmetry trap column ments were developed in the same GRKO cell line (25,39). (Waters) at a flow rate of 15 ␮l/min using a nanoAcquity Briefly, GFP-tagged versions of mouse GR (WT, Y551A, Ultra Performance LCTM chromatographic system (Wa- Y551A-Dim, D647V, D647V-Tetra) were stably integrated ters). Peptides were separated using a C18 analytical col- into the GT-Rosa26 locus via Crispr Cas9/homology di- umn (BEH130 C18, 75 mm × 25 cm, 1.7 ␮m; Waters) with rected repair with puromycin selection and FACS sorting. a 120-min run, comprising three consecutive linear gradi- Cells for the RNA-Seq and N&B assays were grown in ents: from 1 to 35% B in 100 min, from 35 to 50% B in 10 Dulbecco’s modified Eagle’s medium (DMEM, Invitrogen) min, and from 50 to 85% B in 10 min (A = 0.1% formic acid supplemented with 5 ␮g/ml tetracycline (Sigma), 10% fe- in water, B = 0.1% formic acid in CH3 CN). The column tal bovine serum (Sigma), sodium pyruvate, non-essential outlet was directly connected to an Advion TriVersa Nano- amino acids, and 2 mM L-glutamine maintained in a hu- Mate (Advion) fitted on an LTQ-FT Ultra mass spectrom- midifier at 37◦ C. The FBS-supplemented media was re- eter (Thermo), which was operated in positive mode us- placed with media supplemented with charcoal/dextran- ing the data-dependent acquisition mode. Survey MS scans treated FBS (CSS) to remove glucocorticoids from cells for were acquired in the FT-ICR cell with the resolution (de- 24 h prior to hormone treatment (100 nM DEX). All vari- fined at 400 m/z) set to 100,000. Up to six of the most in- ants of GFP-GR for RNA-Seq and N&B were generated tense ions per scan were fragmented and detected in the lin- with the QuikChange II XL Site-Directed Mutagenesis Kit ear ion trap. The ion count target value was 1,000,000 for (Stratagene). the survey scan and 50,000 for the MS/MS scan. Target ions already selected for MS/MS were dynamically excluded for Number and brightness (N&B) analysis 30 s. Spray voltage in the NanoMate source was set to 1.70 kV. Capillary voltage and tube lens on the LTQ-FT were Images were taken at the CCR, LRBGE Optical Mi- tuned to 40 and 120 V, respectively. The minimum signal re- croscopy Core facility in a LSM 780 laser scanning mi- quired to trigger MS to MS/MS switch was set to 1,000 and croscope (Carl Zeiss, Inc.) equipped with an environmental activation Q value was set at 0.25. Singly charged precursors chamber. Cells were imaged from 20 min up to a maximum were rejected for fragmentation. of 2 h after DEX addition. We used a 63× oil immersion objective (NA = 1.4). The excitation source was a multi-line Ar laser tuned at 488 nm. Fluorescence was detected with a Differential scanning fluorimetry GaAsP detector in photon-counting mode. Thermofluor experiments were performed in an iQ5 Mul- N&B measurements were done as previously described ticolor Real Time PCR Detection System (BIO-RAD) us- (24). For each studied cell, a single-plane stack of 150 im- ing 96-well plates (Hard-Shell® High-Profile Semi-Skirted ages (256 × 256 pixels) were taken with a pixel size of 80 PCR Plate, BIO-RAD) and a 25 ␮l total volume for each nm and the pixel dwell time of 6.3 ␮s. In all stacks, we Nucleic Acids Research, 2022, Vol. 50, No. 22 13067 discarded the first 10 images to reduce the effect of image (http://foldxsuite.crg.eu/), which has an estimated error of bleaching on the results. The frame time under these con- ∼0.7 kcal/mol. Ten iterations were conducted for each mu- ditions is 0.97 s, which guarantees independent sampling of tation, and later averaged. Free energy differences between molecules according to previously reported FCS measure- mutant and WT proteins (G) < 1 kcal/mol were consid- ments (40). Each stack was further analyzed using the N&B ered not significant, those between 1 and 2, 2 and 4, and >4 routine of the SIMFCS 2.0 software (Global Dynamics). In kcal/mol as slightly, mildly, and strongly destabilizing, re- this routine, the average fluorescence intensity () and its spectively. variance (␴ 2 ) at each pixel of an image are determined from Downloaded from https://academic.oup.com/nar/article/50/22/13063/6858817 by Fundacio Biblioteca Josep Laporte user on 09 October 2023 the intensity values obtained at the given pixel along the im- Sequence conservation analyses age stack. The apparent brightness (B) is calculated as the ratio of ␴ 2 to while the apparent number of moving The sequence of human GR-LBD was used as query against particles (N) corresponds to the ratio of to B (41). We the whole NR database, from which we selected ∼880 se- used SIMFCS 2.0 software to classify pixels as occurring in quences and included the ancestral GR (38). We also down- the nucleus or the MMTV array (42) according to their in- loaded the sequences corresponding to the LBD regions tensity values. Selection of cells for analysis followed these from a representative fraction of proteomes at PFAM rp55 criteria. (i) In the case of stimulated cells, an accumulation (PF00104 rp55). (Note that the sequences included in the of signal at the MMTV array must be visible. (ii) The aver- PFAM alignment are truncated, as they lack for instance age apparent number of molecules (N) in the nuclear com- the non-conserved F-domain). We followed three different partment must have a range of 3–30 units in all cases, (iii) approaches to ensure sequence and alignment diversity and no saturation of the detector at any pixel (N < 60), and (iv) thus stability of the analyses. First, we aligned the ∼880 bleaching cannot exceed 5–10%. In a previous work it has sequences to a structure-based profile from entries 5UFS been demonstrated that B is equal to the real brightness ε (GR-LBD) and 5JJM (AR-LBD). The resulting alignment, of the particles plus one (41). Therefore, ε at every pixel of 880 aln, was used to run pySCA in addition to the original images can be easily extracted from B measurements. Im- SCA5 method. Secondly, we aligned our ∼880 sequences to portantly, this analysis only provides information regarding a profile generated from the PF00104 rp55 removing frag- the moving or fluctuating fluorescent molecules since fixed ments to generate the 840 aln. Finally, we used the PFAM molecules (relative to our frame time) will give B values alignment as retrieved from the PFAM database, which con- equal to 1. The experiments were independently repeated tains ∼13,000 sequences (PF00104 rp55). two times for each treatment/condition. Multiple correspondence analysis (MCA): identifying specificity-determining positions. We have performed both Total RNA collection and sequencing supervised and unsupervised runs of the S3DET method (http://csbg.cnb.csic.es/JDet/) on 840 aln using default RNA isolations were performed using the PureLink RNA parameters to maintain a large sequence diversity. kit (Thermo Fisher Scientific 12183018A) as per the man- ufacturer’s instructions. We collected three biological repli- Statistical coupling analysis (SCA). We have used the cates of each condition and sample quality was assessed us- SCA5 8/2011 version with the three different alignment ver- ing the Agilent Bioanalyzer. Strand-specific sequencing li- sions given above, and the updated version pySCA with braries were generated from rRNA-depleted (Illumina RS- 880 aln. Each alignment produced unique sets of residues 122-2301) total-RNA samples, using Illumina stranded to- termed Sca5 880, Sca5 840 and Sca5 onPFAMrp55, re- tal RNA (Illumina20020596) according to the manufac- spectively. Next, we labeled residues in the three different turer’s instructions. Raw reads were demultiplexed into sectors that emerged as outputs of the program with letters Fastq format allowing up to one mismatch using Illumina A, B and C, following their order of appearance. The sta- Bcl2fastq v2.17. Reads of the samples were trimmed for bility of the identified sectors was assessed with a statistical adapters and low-quality bases using Cutadapt1.18. RNA- test based on hypergeometric calculations of the groups of Seq alignment to the mouse mm10 genome was performed residues belonging to given sectors between pairs of align- by STAR (43). DESEQ2 (44) was used to normalize the ments. P-values were adjusted using FDR. Next, specific data by read depth, identify differentially expressed genes residues from the significant sectors were extracted and se- for each form of GR (DEX/vehicle) and calculate log2 lected according to their rank. For instance, if a particular fold changes and false discovery rates (FDR) for each residue appears only in one sector on a low-ranking pair gene. RNA-Seq analyses was performed in two batches of alignments (e.g. rank 19, with a borderline P-value) this that were sequenced and analyzed separately: (i) GFP-GR residue will not be selected as part of a sector. On the con- WT (batch1), GFP-GR-Y551A, GFP-GRdim-Y551A; (ii) trary, if a residue appears in high-ranking pairs, it will be re- GFP-GR WT (batch2), GFP-GR-D647V, GFP-GRtetra- tained. Residues termed as ‘A’ and ‘B’ appeared to be equiv- D647V. The two sets of GR WT RNA-Seq data are biolog- alent in different pairs of alignments, so they were assigned ical replicates collected and sequenced separately with each to the class sector 2, while residues belonging to the ‘C’ batch. group were stable, and thus assigned to class sector 1. Impact of GR point mutations on protein folding/stability Clustering of interaction surfaces in GR-LBD dimers We estimated the impact of the generated mutations on We grouped the interaction surfaces observed in the 20 the overall protein stability with the FoldX empirical force topologically different GR-LBD homodimers by a hier- 13068 Nucleic Acids Research, 2022, Vol. 50, No. 22 archical clustering analysis, using ad-hoc R scripts. For RESULTS each dimer, the interaction surface was defined as the set GR-LBD self-associates in solution of solvent-exposed residues in the monomers (i.e. residues with >25% relative accessible surface area, ASA) that be- To characterize GR-LBD behavior in solution, we per- came buried (<25% relative ASA) in the corresponding formed SPR experiments with the ancient variant of the hu- dimer. Relative ASA were calculated using ICM (Molsoft man GR (ancGR2; Figure 1C). This construct has been re- LLC). Then, we calculated the center of coordinates of the peatedly used in GR structure-function studies because of Downloaded from https://academic.oup.com/nar/article/50/22/13063/6858817 by Fundacio Biblioteca Josep Laporte user on 09 October 2023 residues forming each interaction interface. To compare two its higher solubility and stability in vitro (7). For simplic- pairs of homodimers, we computed the Euclidean distances ity, we will refer to all variants of the receptor as GR-LBD, between the centers of their interaction surfaces after su- unless otherwise specified. Briefly, purified GR-LBD·DEX perimposition on a common monomer. The 40 × 40 dis- was expressed and purified, immobilized on CM5 chips us- tance matrix representing the distances between all pairs of ing amine coupling and increasing concentrations of the interaction surfaces was used to perform hierarchical clus- same agonist-bound protein (between 0.2 and 25 ␮M) were tering with Ward’s method. Finally, the dendrogram gen- run over as analyte. Although GR-LBD immobilization to erated from this analysis was sorted in order of increasing the CM5 chip might occlude some interaction surfaces, SPR distance. experiments clearly demonstrate interactions between solu- ble and immobilized GR. Several affinity models were used Docking experiments and analysis to interpret SPR data (Figure 1D, E and Supplementary Figure S2A, B). GR-LBD self-association behavior could Homodimeric models of GR-LBD were built using pyDock be fitted to a non-covalent, 1:1 Langmuir model with an docking and scoring method (45). First, protein models affinity constant (kD ) of 15.3 ± 0.9 ␮M (Figure 1D). In- were prepared by removing all cofactors and heteroatoms, terestingly, we obtained a statistically significant (F-test, P- and missing side chains were modeled with SCWRL 3.0 value = 0.001) better fit of the data by using a model with (46). Then, the Fast Fourier Transform (FFT)-based dock- non-covalent multisite interactions, with two independent ing programs FTDock (47) (with electrostatics and 0.7-Å binding sites (kD1 = 2.2 ± 0.4 ␮M and kD2 = 27.9 ± 1.9 grid resolution) and ZDOCK 2.1 (48) were used to gener- ␮M, respectively; Figure 1E). These results are consistent ate 10,000 and 2,000 rigid-body docking poses, respectively. with GR-LBD tetramer formation, as previously reported These were merged in a single pool for subsequent pyDock for FL-GR in living cells (24). scoring, based on energy terms previously optimized for rigid-body docking. The pyDock binding energy is basically composed of ASA-based desolvation, Coulombic electro- Novel crystal structures of agonist-bound GR-LBD highlight statics, and Van der Waals (VdW) energy terms. Electro- its versatility for self-association statics and VdW contributions were limited to –1.0/+1.0 Next, we performed crystallization trials with GR- and 1.0 kcal/mol for each inter-atomic energy value, respec- LBD·DEX in the presence of the AF-2 targeting peptide tively, to avoid excessive penalization from possible clashes Gln12-Lys30 from SHP. Coregulator peptides are usually derived from the rigid-body approach. added to induce the active conformation of H12 forming Predicted dimer interfaces. Optimal docking areas (ODA) the AF-2 pocket (38). We conducted solubility screens us- per surface-exposed protein residues were obtained by com- ing all commercially available kits (over 4,800 conditions) puting surface patches with optimal desolvation energy to identify several unreported crystallization conditions. based on the selection of low-energy docking regions gen- Diffraction data from crystals that belong to five different erated from each surface residue (49). ODA hot spots space groups (C2, P31 , P61 , I41 22 and I41 32, from lower to (residues with low ODA values, usually less than –10.0 higher symmetry) were collected using synchrotron radi- kcal/mol) indicate regions with favorable desolvation en- ation. The major features of the inter-monomer contacts ergy upon interaction with a partner protein. observed in these new GR-LBD structures are shown in From the resulting docking poses, normalized interface Figure 2 and Supplementary Table S1. propensity (NIP) values were obtained for each residue with C2 crystals contain a single molecule of GR-LBD·DEX the built-in patch module of pyDock, implementing the py- complexed with the SHP peptide, which is well defined DockNIP algorithm (50). A NIP value of 1 indicates that by electron density. Two different inter-monomer contacts the corresponding residue is involved in all predicted in- were identified: the larger, symmetric interaction surface terfaces of the 100 lowest energy docking solutions, while is centered on the L1–3 loops of both monomers and is a value of 0 means that it appears as expected by ran- stabilized by aromatic ␲-stacking interactions between op- dom chance. Finally, a negative NIP value implies that the posite Tyr545/Tyr545’ residues (Figure 2A; residues from residue appears at the low-energy docking interfaces less of- the second monomer are primed). Further stability is pro- ten than expected by random chance. Usually, residues with vided by a network of hydrogen bonds (H-bonds), involv- NIP ≥ 0.2 are considered as hot-spot residues when using ing most notably Asp549 (L1–3), Arg569 (H3), and Asp626 FTDock. (␤-strand S1). A significantly smaller, asymmetric inter- face features Glu688 (H9), whose carboxylate engages in Energetic characterization of GR dimers. The binding en- strong H-bonds with the main chain N atom and the hy- ergy of the different crystal dimers was computed with the droxyl of Ser556’ (L1–3). Additional interactions involve pyDock bindEy module, using the same scoring function as H1 (Leu532) and H9 residues (Lys695, Lys699) from one in docking. LBD molecule facing H6 (Glu632’) from the neighbor. Nucleic Acids Research, 2022, Vol. 50, No. 22 13069 Downloaded from https://academic.oup.com/nar/article/50/22/13063/6858817 by Fundacio Biblioteca Josep Laporte user on 09 October 2023 Figure 2. New crystal structures of GR-LBD·DEX reveal a variety of quaternary assemblies. For all structures, the overall crystal packing is shown in the central panels. Monomers are depicted as cartoons with helices colored blue and loops colored salmon. DEX molecules are represented as salmon spheres and SHP peptides as yellow ribbons. Details of intermonomer interfaces are given in the lateral panels, in which the side chains of interacting residues are shown as sticks. (A) C2 crystals. Note that major contacts are centered on L1–3 with stacked Tyr545 phenol rings from two neighboring molecules. A monoclinic structure of ancGR2-LBD bound to another synthetic GC, triamcinolone acetonide, and complexed to a shorter SHP peptide had been previously reported (5UFS; (38)). Interestingly, 5UFS features a Tyr545-centered parallel dimer almost identical to the topologically equivalent arrangement in our current C2 crystals. (B) Related P31 and P61 crystals. Tyr545 engages in heterologous contacts with a neighboring molecule in these crystals (see the position of the Trp712’ side chain). Structures of ancGR2-LBD bound to either DEX or a different GC (mometasone furoate) and complexed to a TIF-2 peptide had been previously reported in a similar hexagonal crystal form (3GN8 and 4E2J; a and b axes are ∼5% longer in our crystals, while the c axis is ∼4% shorter). These relatively small differences in the cell constants compared to the current P61 structure result in a markedly different small intermonomer interface, however, which is asymmetric in 3GN8/4E2J. (C, D) Common packing of I41 22 and I41 32 crystals. Note that the phenol rings of two Tyr545 residues stack as in the C2 crystals, although the two interacting monomers are fully differently oriented relative to each other. Note also that the largest interface in these crystals features abutting Asp641 side chains from three monomers organized around a local (I41 22) or exact 3-fold axis (I41 32; right side panel in D). 13070 Nucleic Acids Research, 2022, Vol. 50, No. 22 Residue Pro637’ (L5–6) is part of this interface, making structures. This analysis revealed 20 topologically distinct strong VdW interactions with Glu688. homodimers (see Supplementary Table S3 for interacting Two additional, related structures were solved in the residues in all monomer pairs) numbered #1–20 through- enantiomorphic trigonal and hexagonal space groups, P31 out the manuscript. These homodimeric arrangements ap- and P61 (Figure 2B and Supplementary Table S1). In the pear to cover the whole GR-LBD self-association land- GR-LBD·DEX homodimer with the larger interface, the scape. Along with 11 symmetric (isologous) dimeric ar- cleft between H9 and H10–11 is filled with side chains from rangements (i.e. between the same secondary structure el- ements / residues, such as in the Tyr545-mediated dimers Downloaded from https://academic.oup.com/nar/article/50/22/13063/6858817 by Fundacio Biblioteca Josep Laporte user on 09 October 2023 neighboring L1–3’ and L5–6’ loops. In particular, the aro- matic side chains of Trp712 and Phe715 dock into a shal- described above), asymmetric or heterotypic homodimers low groove formed by residues at the C-terminal end of (i.e. where the contacting GR-LBDs engage in interactions H5 and the following loop. This arrangement is thus topo- using different elements) are also common (nine arrange- logically unrelated to the canonical dimerization mode, in ments). which H10–11 helices from two monomers run parallel to We further explored the homodimerization potential of each other, resulting in much larger interaction areas of GR-LBD with a state-of-the-art protein-protein docking ∼1,000 Å2 . This interface is strengthened by salt bridges procedure and a total of 12,000 generated docking dimers. between residues Arg690 and Asp549’ and by several H- This analysis revealed the existence of at least one dock- bonds (e.g. between the carbonyl oxygen of Phe774 and the ing orientation close to 16 of the 20 representative ‘crystal hydroxyl of Ser550’). Noteworthy, the side chain of Tyr545’ homodimers’ (with a root-mean-square deviation (RMSD) engages also in important contacts at this interface, docking ≤ 10 Å) (Supplementary Figure S3). Interestingly, the 2nd on H9 from a neighboring monomer. A second, symmet- best-scoring docking orientation was close (7.3 Å RMSD) ric homodimer is centered on the aromatic Tyr638 (L6–7) to one of the dimers (#20). In two further cases (#8, #10), and Phe735/Tyr738 (C-terminal end of H11) facing each there were docking orientations within 5 Å from the crystal other. However, since positions 638 and 738 are occupied structures, although with no optimal docking scoring. by smaller polar residues in WT GR (Cys and Gln, respec- For an unbiased estimate of the similarity between dif- tively; Figure 1C), this arrangement is unlikely to be signif- ferent homodimers, we first considered the overlapping of icant in vivo. shared contact residues. To this end, we mapped these sets Finally, two related, medium-resolution structures of of residues into a multiple sequence alignment and calcu- GR-LBD·DEX·SHP were obtained in the tetragonal and lated distances between the resulting vectors using differ- cubic space groups (I41 22 and I41 32, respectively; Figure ent metrics (Figure 4A). This Jaccard analysis confirmed 2C, D). Also in this case, we observe a symmetric ho- e.g. the topological similarity between two front-to-front modimer in which the Tyr545/Tyr545’ aromatic rings are homodimers: the first described non-canonical conforma- stacked, although the overall arrangement of LBD modules tion (#1, 1M2Z (3)) and #2, an arrangement observed differs strongly from the Tyr545-directed dimer found in C2 in 4P6W (51) (Supplementary Tables S2 and S3). On the crystals. Additional H1-H3’ contacts result in a more com- other hand, front-to-front homodimer pairs #6 and #11, pact conformation, which is stabilized by H-bonds between although sharing interface residues, differ strongly in that both main- and side-chain atoms of the two monomers, in- the two monomers are arranged parallel and antiparallel to cluding a Glu542–Arg569’ salt bridge. The largest interac- each other, respectively. Similarly, homodimers #9 and #10 tion interface in these forms features a trimeric arrangement share the important residue Ile628 at the center of their in- in which loops L1–3’/H3’, S2-L6’ and L11–12’ dock per- termonomer interfaces, but the two modules are quite dif- pendicularly onto H10–11. The large, buried surface area ferently oriented relative to each other. The way the four in this trimer appears to compensate the electrostatic re- distinct areas on the GR-LBD surface assemble to generate pulsion of abutting Asp641 carboxylates from the three the 20 topologically different homodimers is schematically monomers around a pseudo (in the tetragonal form) or ex- presented in Figure 4B. act 3-fold axis (in the cubic cell). Alternatively, GR-LBD homodimers were superimposed on a common origin and classified by mapping the cen- ters of coordinates of their interaction surfaces (Supple- A thorough catalog of homodimeric arrangements illustrates mentary Figure S3A). Supplementary Figure S3B shows the the multivalent potential of the GR-LBD orientations of these surfaces, represented by vectors be- The fact that even minor changes in protein complexes and tween the common center of coordinates and each inter- crystallization conditions result in different GR-LBD ar- action surface. A hierarchical clustering analysis based on rangements, as demonstrated by the variety of crystal con- the Euclidean distance between these vectors grouped all tacts described above (Figure 2A–D), prompted us to sys- homodimers into six clusters (Supplementary Figure S3C). tematically analyze protein-protein contacts in all crystal From the spherical coordinates of the vectors representing structures of the domain previously deposited in the PDB the orientations of the interaction surfaces (Supplementary (Figure 3A, B; Supplementary Tables S2 and S3). GR-LBD Figure S3E), we found that these clusters can be associ- residues involved in homodimer formation cluster in four ated with combinations of the previously defined surfaces: areas on the surface, the ‘front’, ‘back’, ‘top’ and ‘base’ of cluster 1 (top-front), cluster 2 (top-back), cluster 3 (base- the domain (Figure 3B–D), in the standard view of NRs back), cluster 4 (base-front), cluster 5 (front), and cluster (Figure 1B). 6 (base). While each cluster may contain surfaces with dif- Next, we analyzed which combinations of these four ho- ferent binding energy values (Supplementary Figure S3D), modimerization interfaces have been encountered in crystal arrangements corresponding to top and back interaction Nucleic Acids Research, 2022, Vol. 50, No. 22 13071 Downloaded from https://academic.oup.com/nar/article/50/22/13063/6858817 by Fundacio Biblioteca Josep Laporte user on 09 October 2023 Figure 3. Experimental structures and bioinformatics analyses unveil four major homodimerization surfaces on GR-LBD. (A) Relative frequencies of residue involvement in GR-LBD homodimer formation. Bar height indicates how often a given residue engages in crystal contacts in all available struc- tures of GR-LBD, normalized to the residue most frequently found in homodimer interfaces, Leu741. Secondary structure elements given below the plot correspond to the crystal structure of human GR-LBD resolved at the highest resolution, 6NWL. (B) Residues involved in GR-LBD homodimerization cluster in continuous patches on its front (colored purple), base (coral), back (blue) and top (pink) faces. The association of these four faces yields the cat- alog of GR-LBD dimers represented in Figure 4. Models are shown in the same orientation and at the same magnification in panels (C)–(E). (C) Predicted protein-protein interaction ODA. ODA ‘hotspots’ (residues with favorable docking energy; ODA < –10.0 kcal/mol) are colored red, residues with ODA > 0 kcal/mol are shown in blue, and intermediate values are scaled accordingly. ODA hotspots form continuous surface patches that essentially overlap with the four protein-protein interaction interfaces shown in panel (B). (D) Hotspot interface residues predicted from docking experiments. Surface residues are colored according to their NIP. Residues with NIP > 0.4 and < 0 are colored red and blue, respectively; intermediate values are scaled accordingly. (E) SCA identifies two sectors of clustered, physically connected residues in GR-LBD. The front and back orientations of GR-LBD are depicted, and in both cases the module is represented as a solid surface and as a cartoon, with helices shown as rods and labeled. Residues belonging to sectors I and II are shown with their side chains atoms as spheres, colored cyan and dark blue, respectively. Residues belonging to both sectors are colored pink. SHP peptide is colored yellow, and DEX is shown as salmon spheres. 13072 Nucleic Acids Research, 2022, Vol. 50, No. 22 Downloaded from https://academic.oup.com/nar/article/50/22/13063/6858817 by Fundacio Biblioteca Josep Laporte user on 09 October 2023 Figure 4. An integrated catalog of GR-LBD homodimers. The four distinct GR-LBD protein-protein interfaces associate to generate 20 topologically different homodimers. (A) A dendrogram based on a hierarchical analysis of protein-protein contacts using Jaccard’s index groups the 20 unique GR- LBD assemblies into six different clusters. (B) Relationships between the different GR-LBD homodimers. For orientation, monomers highlighting the four interacting surfaces are placed at the cardinal points in this panel (top, nord; front, east; base, south; and back, west), colored-coded as in Figure 3. Monomers in 10 representative homodimers are depicted as cartoons; each monomer is colored according to the face used to associate with its partner. Dimers are placed closest to the generating monomers. An equivalent schematic representation of GR-LBD homodimerization potential is shown at the upper left corner, with the position of major interacting residues indicated. Nucleic Acids Research, 2022, Vol. 50, No. 22 13073 surfaces have in general more favorable binding energy. The 5A). To identify crosslinked Asp/Glu-Lys pairs, gel bands distribution of interfaces in the top 1,000 docking models corresponding to oligomeric GR-LBD were excised, en- also shows significant clustering around regions with favor- zymatically digested with either trypsin or chymotrypsin, able binding energy (Supplementary Figure S3F). However, and analyzed by MS (Supplementary Figure S4A, Supple- docking solutions also clustered in other regions with less mentary Tables S4 and S5). Most notably, we found that optimal energy, such as front and base surfaces. elements essential for top-to-top (#20) and front-to-front (#1, #2, #6, #9, #10 and #11), non-canonical dimeriza- Downloaded from https://academic.oup.com/nar/article/50/22/13063/6858817 by Fundacio Biblioteca Josep Laporte user on 09 October 2023 tion are overrepresented among EDC-linked peptides, with Two sectors define the internal circuits linking key GR-LBD the most common contacts involving (i) residues of H1 and interaction sites L1–3, on the one side, and from H3, on the other, which Positions differentially conserved within protein subfami- would correspond to front-to-front interactions, as well as lies (termed ‘specificity determining positions’, SDPs) are (ii) H9 and the L9-10 linker from two monomers, which is related to functional specificity (e.g. binding of different co- compatible with homodimer #20 (Figure 5B, C and Sup- factors). Recent bioinformatics methods based on MCA plementary Figure S4B, Supplementary Table S4). Simi- have allowed identification of the subtle patterns of conser- lar results were obtained with the MS-cleavable, urea-based vation of SDPs within large protein families (52). Our initial crosslinker, DSBU (Supplementary Table S6). MCA run did not replicate the previously reported family classification (38), and only identified three residues that di- Mutant Y545C exhibits GR-LBD non-canonical homodimer- vide the NR superfamily into two clusters. This prompted ization in solution us to use more sophisticated statistical tools to search for evolutionary conserved units in GR-LBD. To analyze if Inspection of homodimer interfaces in GR-LBD revealed self-association surfaces may be allosterically coupled to several symmetric arrangements in which the side chain of other functional regions, we performed a SCA, which en- a solvent-exposed residue from one monomer is located tirely relies on correlated amino acid variations across the within VdW distance of the same residue from a crystal domain without considering its 3D structure (53,54). In- neighbor (Figure 2A and Figure 5D). To verify whether deed, this analysis identified 40 residues that decompose some of these conformations are present in solution, we the GR-LBD sequence into two quasi-independent groups have generated several Cys mutants of the GR-LBD. All of correlated residues or ‘sectors’ (Figure 3E). Sector 1 studied mutants were properly folded, as indicated by only comprises 17 residues in and around H1 and H10, most minor decreases in melting temperatures in DSF analysis, in notably LBP residues Met601 and Arg611 along with the line with the results of a systematic bioinformatics analysis nearby Phe606, whereas sector 2 features 20 residues mostly of mutant stability performed using Fold-X (not shown). from H3 and H5 (e.g. Met604 in the LBP, Lys579, Phe584, Incubation of purified Y545C in low-reducing conditions Gln597 of AF-2 and Trp577 in an internal path connect- resulted in the rapid formation of covalent dimers (Figure ing AF-2 to the LBP). Finally, Gly583 (BF-3), Leu596 (AF- 5E). By contrast, neither the WT protein nor other Cys 2) and the internal Tyr663 belong to both sectors. Interest- mutants tested (e.g. D641C, S744C) dimerized under the ingly, all residues cluster in the upper half of the LBD, where same conditions (Figure 5E). These findings strongly sug- both sectors are physically interconnected (Figure 3E). Sec- gest that the side chains of Tyr545 from two monomers are tor 1 residues cluster around the N-terminus of the LBD close enough in solution, at least in a subset of GR-LBD and are thus likely candidates to interact with the hinge and molecules. DBD. Perhaps more relevantly, sector 2 comprising residues To directly confirm that residues Cys545/Cys545’ are profusely innervate the LBP and AF-2 regions while Arg611 responsible for disulfide bridge-mediated dimerization in from sector 1 is essential to position the hormone. The three solution, bands corresponding to the dimer were excised residues that belong to both sectors are strategically located from the gel, treated with iodoacetamide to block free to link the LBP with AF-2 and BF-3. Taken together, our re- Cys residues, and subjected to enzymatic digestion with sults suggest that both sectors couple functionally relevant trypsin and GluC. MS analysis of these digests allowed in- regions (e.g. ligand and coregulator binding or chaperone deed the identification of peaks corresponding to peptide docking/release). V543 LCSGYD549 crosslinked to either V543 ’LCSGYD549’ or V543’ LCSGYDSTLPDTSTR558 ’, thus confirming Cys545- mediated covalent bond formation (Supplementary Figure In vitro crosslinking experiments corroborate non-canonical S4C and Supplementary Table S7). GR-LBD dimerization To further assess the contribution of Tyr545 to GR ho- The results presented above suggest that many surface- modimerization in solution, we performed SPR assays with exposed residues of GR-LBD engage in a variety of crys- its Y545C and Y545A mutants. Indeed, experiments con- tal contacts. To clarify their role in homodimer formation ducted with the Cys mutant revealed significant increases in solution, we took advantage of intermolecular H-bonds in affinity. The increase was the highest when the mutant between Glu/Asp and Lys at some of the crystal interfaces, was used as both ligand (i.e. chip-immobilized) and analyte, which can be ‘frozen’ upon incubation with the zero-length as indicated by dramatic increases in affinity (kD = 0.85 vs. crosslinker, EDC. 13.5 ␮M for the WT-WT association at 100 ␮M DTT; Sup- Indeed, we observed rapid formation of a covalent dimer plementary Figure S2C, D) as well as in complex lifetime, upon incubation of GR-LBD with EDC, as well as a fainter with a 4-fold increase at 10 ␮M DTT (t1/2 = 1,130 vs. 270 band corresponding to a tetrameric arrangement(s) (Figure s). By contrast, truncation to Ala at position 545 led to a 13074 Nucleic Acids Research, 2022, Vol. 50, No. 22 Downloaded from https://academic.oup.com/nar/article/50/22/13063/6858817 by Fundacio Biblioteca Josep Laporte user on 09 October 2023 Figure 5. Several GR-LBD homodimers are populated in solution. (A) SDS-PAGE analysis of GR-LBD after incubation with EDC. Notice bands with relative molecular masses corresponding to GR-LBD dimers (D) and tetramers (T) in all the lanes except control lane 1 (no EDC added). Lanes 2 and 3, protein incubated at about 0.37 mg/ml; lanes 4 and 5, protein incubated at about 1.5 mg/ml. Samples in lanes 2 and 4 were treated at RT; those in lanes 3 and 5 at 30◦ C. Bands corresponding to co-purified DnaK and GroEL E. coli chaperones are marked with (*) and (**), respectively. (B) Schematic representation of GR-LBD structure, with side side chains of residues identified using EDC shown as spheres (Asp in green, Lys in lilac, and Glu in yellow). (C) Crosslink map of EDC-treated GR-LBD showing all intermonomer crosslinks identified by MS analysis. Regions corresponding to the top, front, base, and back surfaces are colored as in Figures 3 and 4. A secondary structure plot is shown above the map. (D) Closeup of the major homodimerization interface in C2 crystals (front-to-front homodimer #11), dominated by stacked Tyr545/Tyr545’ residues. (E) Non-reducing SDS-PAGE analysis of purified GR-LBD(Y545C) (lane 3) shows spontaneous dimerization in solution. Note that the WT protein does not form dimers when incubated at the same concentration (lane 2). Nucleic Acids Research, 2022, Vol. 50, No. 22 13075 slightly less tight association (Supplementary Figure S2E). rowheads). Of note, reduced genome-wide chromatin bind- We also crystallized and solved the structure of the Y545A ing for GRmon has recently been shown (39), and only a variant, thus confirming proper folding of the generated very small percentage of GRmon cells form visible arrays point mutants. Noteworthy, Y545A crystallized in the P61 (23,24). Interestingly, we did not detect any arrays in cells space group, which does not involve symmetric contacts be- with GRmon/Y545A (Figure 6B), suggesting an even more tween the side chains of residues at position 545 (Figure 2B). drastic phenotype for this triple mutant. Altogether, our results confirm that residue Tyr545 plays an While W712A oligomerizes as the WT receptor both in Downloaded from https://academic.oup.com/nar/article/50/22/13063/6858817 by Fundacio Biblioteca Josep Laporte user on 09 October 2023 important role in GR-LBD homodimerization in solution. the nucleoplasm and at the array (Figure 6C, D), Tyr545Ala substitution appears to slightly decrease dimerization in the nucleoplasm (ε = 1.83). By contrast, Pro637Ala produces Residues Tyr545 and Asp641 modulate multimerization of a slight increase in oligomerization (ε = 2.46), even though FL-GR neither difference achieves statistical significance. Since FL- Quantitative fluorescence microscopy in living cells (the GR dimerizes at least through both DBD and LBD (Figure number and brightness method, N&B) allows to estimate 6A), we tested the effect of the Tyr545Ala mutation in the the average oligomeric state of a fluorescent protein from its GRdim background, which has impaired DBD-DBD con- molecular brightness (ε; (41)). For N&B experiments we use tacts yet mostly dimerizes in the nucleoplasm (Figure 6C, mouse adenocarcinoma cells, which possess a tandem array (23,24)). Indeed, the GRdim/Y545A double mutant shows sig- of DNA binding sites for GR, the MMTV array (42). We nificant tendency to remain monomeric in the nucleus (ε = have knocked out expressed GR protein in these cells, thus 1.16), like the previously characterized GRmon (Figure 6C, designated GRKO (25). These cells are transiently trans- (23,24,56)). Taken together, these results confirm the impor- fected with GFP-tagged full length mouse GR (GFP-mGR) tant role of Tyr545 in the dimerization of GR in live cells. or mutants thereof, and the oligomeric state of fluorescently On the other hand, the Asp641Val mutation, linked to tagged GR molecules is quantified by comparing to a con- a GC resistance phenotype known as Chrousos syndrome stitutively monomeric GR variant (N525*) (24). The mu- (57), and Pro637Ala promoted higher-order oligomeriza- tated residues of the mouse GR variants correspond to the tion at the array (Figure 6D), possibly hexamers or a mix- human versions discussed above and are noted in the fig- ture of tetra- and octamers. These findings prompted us to ure legends. Further, presence of the MMTV array in the analyze the impact of these two mutants when combined GRKO cells allows us to differentially assess GR oligomer- with a DBD mutation that enforces GR tetramerization, ization not only in the entire nucleoplasm but also in a re- Pro474Arg. Unexpectedly, instead of a synergistic effect, the gion highly enriched in GR binding sites that is visible under combined mutations reversed to GRtetra oligomerization in the microscope as a bright spot (Figure 6B, arrowheads). the nucleus (Figure 6C). By contrast, the GRtetra mutation To test the relevance for the full-length receptor (FL-GR) abrogated the ability of P637A and D641V variants to form of key surface-exposed residues identified in GR-LBD ho- higher-order oligomers at the array level (Figure 6D). These modimer interfaces in vitro, we generated Ala mutants of observations highlight a complex relationship between the mouse GR at positions topologically equivalent to human different structural domains of the NR. residues Tyr545, Pro637, Asp641 and Trp712, all of which are conserved in ancGR2 (Figure 1C). Further, all residues Role of residues Tyr545 and Asp641 on GR transcriptional but Asp641 are strictly conserved from fish to humans, and activity Asp641 is conservatively replaced by a glutamate in non- mammals (Supplementary Figure S1). To study the impact Having demonstrated that residues Tyr545 and Asp641 play of the Tyr545→Ala exchange in the background of two essential roles in FL-GR oligomerization, both in solu- other variants previously shown to be important for GR ho- tion and in live cells, we wondered if they would modu- modimerization (23), we also generated the double mutant late the activity of the receptor. To this end, we generated (Ala458Thr, Tyr545Ala) (in following termed GRdim/Y545A ) stable cell lines expressing GFP-tagged mouse versions of and the triple mutant (Ala458Thr, Tyr545Ala, Ile628Ala), the GRY545A and GRdim/Y545A variants (see methods), and or GRmon/Y545A . Finally, we also generated double mutants compared their transcriptional programs to that of mouse in which a GR variant that tetramerizes both in the nu- GR WT by RNA-seq, after 2h of 100 nM Dex stimulation cleus and at the array (Pro474Arg, (24)) was combined (Supplementary Table S8 and Supplementary Figure S5A– with either P637A or D641V (These variants, (Pro474Arg, F). As summarized in Figure 7A, GRY545A upregulates only Pro637Ala) and (Pro474Arg, Asp641Val), are in following 56% of the genes compared to WT GR, and only three out termed GRtetra/P637A and GRtetra/D641V , respectively; Figure of the 21 genes downregulated by the WT receptor. In line 6A). with the more significant impact of the GRdim/Y545A double Next, we transiently transfected GFP-mGR and the gen- mutant on receptor homodimerization, its transcriptional erated mutants into GRKO cells and performed N&B ex- activity was more compromised than GRY545A single mu- periments as previously described (23,24,55). All mutants tant, with only 28% of the number of genes being modulated translocate to the nucleus upon hormone stimulation (Fig- respect to WT GR (Figure 7A). Of the 17 common genes be- ure 6B), indicating the mutants can bind ligand at the tween the three GR types, all but 2 have only modestly lower tested hormone concentration of 100 nM DEX. Further, fold change with Dex between GR WT and GRY545A , while all variants except those carrying the GRmon double mutant most of the same common genes exhibit a more severe im- (Ala458Thr, Ile628Ala) were visible at the MMTV array, pact upon hormone treatment in the GRdim/Y545A cells (Fig- suggesting that DNA binding was possible (Figure 6B, ar- ure 7B). 13076 Nucleic Acids Research, 2022, Vol. 50, No. 22 Downloaded from https://academic.oup.com/nar/article/50/22/13063/6858817 by Fundacio Biblioteca Josep Laporte user on 09 October 2023 Figure 6. Mutation of LBD–LBD interface residues profoundly affects the multimerization behavior of FL-GR. (A) Schematic 3D model of FL-GR. The NTD is highly disordered and is followed by the globular DBD (DNA-bound conformation is shown) and the LBD. Note that the length and flexibility of the hinge (H) allows for the formation of various homodimers. The side chains of all residues mutated to assess the multimerization behavior of GR are shown as spheres. (B) Subcellular localization of WT GFP-mGR and indicated mutants in 3617-GRKO cells, as assessed by fluorescence microscopy. Variant N525* lacks the entire LBD and remains monomeric, irrespective of DEX treatment (23). White arrowheads point to the MMTV arrays. Scale bar: 5 ␮m. Data for WT GR, GRdim , GRmon and GRtetra were taken from (24) and are shown for comparison purposes. All N&B experiments have been performed with mouse GR as in (23). However, residue numbers correspond to the human protein to facilitate comparisons. (C) GR oligomerization in the nucleus, as determined in N&B assays. The fold increase in molecular brightness (ε) relative to the N525* monomeric control is shown (N = 428 total cells with 21 < N < 37 between treatments). (D) Quaternary structure of DNA-bound GR. The results of N&B assays at the MMTV arrays are represented as in panel C (N = 338 total cells with 17 < N < 35 between treatments). Note that simultaneous disruption of intermonomer interactions mediated by the DBD (Ala458Thr) and the LBD (Tyr545Ala) in the GRdim/Y545A double mutant results in a variant that is monomeric in the nucleus, while at the array it formed mostly trimers. In panels (C) and (D), centered lines show the medians and crosses represent sample means (average numbers below each box-plot). Box limits indicate the 25th and 75th percentiles; whiskers extend 1.5-fold the interquartile range from the 25th and 75th percentiles, with outliers represented by dots. Boxes with different superscript letters are significantly different from each other (P < 0.05; one-way ANOVA followed by Tukey’s multiple comparison test). Nucleic Acids Research, 2022, Vol. 50, No. 22 13077 Downloaded from https://academic.oup.com/nar/article/50/22/13063/6858817 by Fundacio Biblioteca Josep Laporte user on 09 October 2023 Figure 7. Residues Tyr545 and Asp641 have major roles in GR quaternary structures and transcriptional activity. Stable cell lines expressing GFP-tagged variants of the indicated mouse GR mutants were treated with 100 nM DEX for 2 h (note than numbers correspond to hGR, as in Figure 6). (A, C) Venn diagrams of differentially expressed genes from RNA-Seq (|log2 Fold change| > 0.5, False-discovery-rate < 0.01). The two WT GR sets (b1, b2) are biological replicates collected and sequenced separately for each batch. (B, D) Scatter plots of shared hormone-regulated genes (17 and 71 common genes, respectively) between GR mutants, compared and plotted against their respective WT GR control. (E) Models of human GR-LBD homodimers based on the observed assemblies #10, #6, #9 and #11. The critical homodimerization residues, Tyr545 and Ile628, are shown as color-coded spheres in all cases. (F) Model of GR-LBD tetramer generated by docking dimer #11 onto itself. The model is compatible with the results obtained with the Y545C mutant and EDC-crosslinking. (G) Putative GR-LBD hexamer favored by the Asp641Val mutation. The central trimer corresponds to an arrangement observed in tetragonal and cubic crystal forms (#14; Figures 2C, D and 4), while the peripheral monomers dock according to conformation #4. Asp641 residues from the central trimer are encircled. This generates a closed hexamer by additional interactions between the N-terminal end of H10 and H12’ at the third interface. In addition to D641V, other GR mutants associated with Chrousos disease might favor multimeric forms that are incompatible with active GR tetramers on DNA. For instance, replacement of Thr556 by an aliphatic Ile would stabilize this arrangement through contacts with e.g. residues Met560 and Pro637 from its ‘own’ monomer and/or His645’/Asn731’ from a neighboring molecule. Similar considerations apply to mutations such as Arg714Gln, Phe737Leu, Ile747Met and Leu773Pro as well as to variant Pro637Ala, which also forms higher-order oligomers at the array (Figure 6D). 13078 Nucleic Acids Research, 2022, Vol. 50, No. 22 RNA-Seq analysis of cells expressing the higher (Supplementary Figure S6A), (5,13,58) could be adopted by oligomeric variant, D641V, revealed that the mutant mod- other NRs. However, none of the 20 GR-LBD homodimers ulated only two thirds of the hormone-responsive genes can be considered as topologically equivalent to the one compared to WT GR (Figure 7C), consistent with the observed in the crystal structure of AR-LBD, illustrating in vivo phenotype described as Chrousos syndrome (GC a more complex multimerization behavior than previously resistance) (57). The double mutant GRtetra/D641V shows anticipated. Replacement of AR interface residues such as a much larger number of hormone responsive genes with Thr656 by positively charged Lys/Arg residues and of the respect to GR WT, similar to the GRtetra single mutation re- Downloaded from https://academic.oup.com/nar/article/50/22/13063/6858817 by Fundacio Biblioteca Josep Laporte user on 09 October 2023 following Asn657 by bulkier Gln/His residues in all other ported elsewhere (25). This suggests that the constitutively members of the subfamily might preclude the formation tetrameric conformation exhibited by this DBD mutant of AR-like homodimers in GR/MR/PR. This highlights can exert a strong effect on gene response even in tandem the difficulty to extrapolate the quaternary structure of one with the D641V variant of the LBD. However, the tetramer NR to other, even closely related NRs, and the need for mutation cannot completely rescue the relative fold change experimental evidence to identify physiologically relevant of many common DEX-responsive genes, especially genes conformations (13). A related arrangement had been previ- with the highest response (Figure 7D). These observations ously observed in some structures of agonist- or antagonist- reinforce the importance of both DBD and LBD modules bound GR-LBD (3), but its biological relevance has been of the GR in transcription regulation, and underscore the repeatedly questioned (59,60). Here we identify and discuss functional relevance of two surface-exposed LBD residues, several unrelated multimeric assemblies of GR-LBD that Tyr545 and Asp641. might illuminate different structural arrangements of the chromatin-bound full-length protein. DISCUSSION GR can adopt several topologically distinct dimeric confor- Although many structures of GR-LBD have been reported mations (Supplementary Table S2), the physiologically relevant con- formation(s) of GR and other oxosteroid NRs remain con- Our data indicates that ‘non-canonical’ GR homodimers troversial. The new structures presented here highlight the centered on the Tyr545 phenol ring are critical for GR ability of different GR-LBD surfaces to engage in var- homodimerization. In live cells, variant Y545A is mostly ious quaternary arrangements depending on the bound monomeric when combined with an exchange that also dis- agonists/antagonists, cofactors, and other biochemical pa- rupts DBD-DBD contacts (Ala458Thr, GRdim). Consis- rameters. These observations prompted us to dissect the tently, the transcriptional program of the Y545A variant is oligomerization capability of GR-LBD. Since the ancient significantly affected, and the double mutant GRdim/Y545A LBD structure may not accurately mimic the multimeriza- is practically inactive (Figure 7A). Further, an important tion behavior of modern, wild-type GR, our structural pre- fraction of GR-LBD Y545C molecules rapidly and sponta- dictions were experimentally verified by characterizing sev- neously forms covalent homodimers in vitro mediated by eral key mutations of exposed residues in the background of the Cys545–Cys545’ disulfide bond (Figures 5C, D). Fi- mouse FL-GR in live cells. The results of these experiments nally, Ala replacement of Try545 led to a reduced monomer- demonstrate that the multimerization landscape of ancGR2 monomer affinity in SPR experiments (Supplementary Fig- reproduces that of modern GR, in the context of the full- ure S2E). length receptor. Careful inspection of intermonomer con- Taken together, we propose four agonist-bound plausible tacts in all available 3D structures, irrespective of the vari- GR dimers (Figure 7E). Formation of the Cys545–Cys545’ ant used for the structural analysis, allowed us to identify disulfide bridge in the Y545C variant is compatible with 20 topologically different homodimeric architectures (Fig- two topologically different, roughly parallel and antiparal- ure 4, Supplementary Table S3). These experimental GR lel symmetric homodimers (#6 and #11; Figure 7E). Incom- conformations could be grouped into six different clusters plete dimerization of GR-LBD(Y545C) suggests that other considering relationships between the interacting residues dimeric states are also populated in solution. Indeed, in a (Figure 4A, Supplementary Figures S3C and S3E), which third, also antiparallel GR-LBD homodimer the Tyr545 correspond to three topologically distinct front-to-front ho- side chains are located at the borders of the protein-protein modimers (#1, #2, #6, #9, #10 and #11), along with base- interface (#9, Figure 7E). This arrangement is centered on to-base (#12), top-to-top (#20), and back-to-back arrange- residues Ile628/Ile628’, which are important for receptor ments (#15, #16 and #17). SPR and XL-MS experiments in dimerization (23). Indeed, the I628A mutant decreased GR solution coupled with quantitative fluorescence microscopy multimerization of FL-GR in living cells to a larger ex- of FL-GR WT and several mutants in live cells along with tent than the Y545A variant, both in the nucleoplasm and RNA-seq analyses of most relevant variants allow us to pos- DNA-bound (24). The most likely explanation for these dif- tulate the more likely quaternary arrangements of GR-LBD ferences is that Tyr545 and Ile628 affect two topologically associated with its different genetic variants (Supplemen- distinct although interconnected dimerization interfaces of tary Table S9). GR-LBD, in line with our structural analysis (Figures 3 It has long been accepted that the ‘canonical’, H10- and 4). In this regard, it is noteworthy that this side chain mediated homodimeric conformation adopted by ER␣- is less exposed than the Tyr545 phenolic ring, engaging LBD (Supplementary Figure S6B) is hindered by the F- in close VdW contacts with Leu620, Cys622 and Pro625. domain in oxosteroid receptors (5,13). Several observations Thus, the larger impact of the I628A mutant might simply suggested that the non-canonical dimer reported for AR reflect disruption of an extended protein-protein interface Nucleic Acids Research, 2022, Vol. 50, No. 22 13079 area centered on the Ile628 side chain. Finally, a second cleus, but its ability to form tetramers on DNA is com- Ile628-centered, parallel conformation is found in crystals promised (24). Thus, impaired tetramerization of RU486- of ancGR1 (homodimer #10; Figure 7E, (56)). Therefore, bound GR might account for its transcriptional deficiency. both parallel and antiparallel arrangements of LBD mod- ules with different involvement of the Tyr545 and Ile628 Towards a model for FL-GR quaternary arrangements and valences are compatible with current data and are equally their role in physiology and disease possible. Some of these quaternary arrangements in solu- Downloaded from https://academic.oup.com/nar/article/50/22/13063/6858817 by Fundacio Biblioteca Josep Laporte user on 09 October 2023 tion could be verified in our XL-MS experiments (Figure Our earlier live-cell imaging studies have revealed that 5A, Supplementary Table S6). GR-LBD features several dimeric GR might be an intermediate state towards tran- strictly conserved Tyr/Trp residues, two of which, Tyr545 scriptionally active tetramers bound to target DNA se- and Trp712, are repeatedly found at monomer-monomer in- quences (24,25,39) also in line with the current SPR results terfaces (Figure 2 and Supplementary Table S3). Against ex- (Figure 1D, E). Although more speculative, it is possible pectations, elimination of the bulky indole ring at position to generate tentative models of DNA-bound GR tetramers 712 had no effect on the oligomerization properties of FL- that satisfy the constraints derived from current structure- GR in live cells, as the W712A variant behaved as WT both function information (Figure 7F). First, we reasoned that in the nucleoplasm and DNA-bound (Figures 6C and D). since tetramers but not higher order multimers are detected These findings strongly suggest that Trp712 is not important on chromatin, none of the known intermonomer ‘anchors’, for GR multimerization in vivo. This would exclude, among Tyr545 and Ile628, would be available for protein-protein others, the symmetric homodimer #20, which had been pro- interactions in active tetramers. In other words, all GR- posed as the most likely conformation of dimeric GR-LBD LBD valences should be satisfied in bona fide GR tetramers. (61). Nevertheless, the fact that Trp712 participates in six Since the Ile628 side chain is exposed in both conforma- topologically different homodimeric arrangements suggests tions with opposing Tyr545 rings (#6 and #11), different an enhanced propensity for protein-protein interactions. In- combinations with homodimers #9 and #10 are conceivable deed, the recently reported cryo-EM structure of GR-LBD (Supplementary Figures S6E-J). These hypothetical multi- bound to the ‘client-maturation complex’ Hsp90-p23 has meric arrangements reconcile the role of residues Tyr545 revealed that Trp712 occupies the central position in a ma- and Ile628 for GR multimerization. The additional interac- jor binding epitope for the p23 co-chaperone (62). tions predicted in these multimeric arrangements, in addi- Interestingly, antagonist binding appears to result in tion to DBD-DBD interactions (23) and condensation pro- completely unrelated GR-LBD assemblies: the symmetric vided by the NTD (27) would overcome the energy loss due back-to-back (#15) and base-to-base homodimers (#12), to Tyr545→Ala or Ile628→Ala exchanges, explaining why which have much larger interface areas and much lower en- variants Y545A and I628A are still tetrameric on chromatin ergies than all other GR-LBD conformations (Supplemen- (Figure 6D). However, the triple mutant (A458T, Y545A, tary Table S3). These dimers are indeed only observed in I628A) did not bind the array, indicating that simultaneous GR-LBD bound to the antagonist, mifepristone/RU-486 elimination of the Tyr545 and Ile628 valences would gener- (1NHZ and 5UC3, respectively, (59)), and share several ate a well folded, but fully inactive variant. important features. First, the bound antagonist displaces On the other hand, our results suggest that formation H12, which partially disrupts the LBP. More importantly, of non-physiological multimers might underlie the deleteri- in both cases the AF-2 of one monomer is partially cov- ous effect of glucocorticoid resistance mutants. Unexpect- ered by either H12 (#15) or H3 (#12) from a neighboring edly, we observed that a GR mutant linked to Chrousos molecule, thus interfering with coregulator binding (Sup- syndrome, D641V (57), preferentially formed higher-order plementary Figures S5G, H). Further, essentially the same oligomers when bound to DNA (Figure 6D), which was ac- arrangements are found in crystals of the dominant nega- companied by a significant reduction in its transcriptional tive GR␤, which is known to bind only antagonists (63). activity (Figure 7C). A three-fold reduced affinity of the Even though tetrameric assemblies in 1NHZ/5UC3 and in Val641 mutant for DEX was previously proposed as the a third structure of mifepristone-bound GR (3H52; (64)) underlying molecular defect in the D641V variant. How- have been reported, these are not compatible with the results ever, since Asp641 is exposed on the GR surface and is of our XL experiments with EDC (Figure 5A and Supple- located 8 Å away from the closest DEX atom, it is un- mentary Figure S4B) or DSBU (Supplementary Tables S4 likely to play any relevant role in hormone recognition. and S6). Inactive RXR␣ also adopts a disc-like tetrameric This non-conservative mutation has a particularly strong conformation, in which H12 from one molecule protrudes stabilizing impact of 3.4 kcal/mol for each dimer pair on away and docks on the AF-2 of an adjacent monomer (Sup- the GR-LBD trimer observed in current tetragonal and plementary Figure S5I; (65)). Although the overall arrange- cubic crystals (#14; see Figure 2D, Supplementary Table ment differs from the ones adopted by agonist-bound GR- S3). Therefore, these structures provide a plausible model LBD, the fact that AF-2 is occluded in both cases points to for a trimeric arrangement in which the three Asp641 side a general feature of self-repressed or inactive conformations chains face each other around a three-fold axis. The pos- in NRs. Altogether, these findings strongly suggest that the itive effect of the mutation results from the relief of strong most stable, back-to-back (#15) and base-to-base confor- electrostatic repulsion between the three abutting negatively mations (#12) of GR-LBD are associated with inactive, self- charged Asp641 residues, coupled with favorable VdW con- repressed states, which are induced or stabilized by antago- tacts made by the aliphatic Val641 side chains with each nist binding. Previous experiments had demonstrated that other and with Cys638 from a neighboring molecule. No- FL-GR bound to mifepristone is fully dimeric in the nu- tably, conformation #4 is fully compatible with this trimer 13080 Nucleic Acids Research, 2022, Vol. 50, No. 22 and generates a closed hexamer (Figure 7G). This confor- SUPPLEMENTARY DATA mation would occlude the AF-2 pockets of the ‘external’ Supplementary Data are available at NAR Online. monomers and is thus incompatible with coregulator bind- ing, impeding GR transcription. Several additional residues linked to Chrousos disease are located at or close to one ACKNOWLEDGEMENTS of the three monomer-monomer interfaces, and their muta- We thank Erick Ortlund (Emory University) for provid- tions might promote local conformations that also stabilize ing the plasmid for ancGR2-LBD, Arnold T. Hagler for his Downloaded from https://academic.oup.com/nar/article/50/22/13063/6858817 by Fundacio Biblioteca Josep Laporte user on 09 October 2023 this hexamer (Figure 7G). This is, to the best of our knowl- comments on the manuscript, and Ildefonso Cases (CABD) edge, the first time in which non-productive multimers of a for his help with code. We thank ALBA-Cells synchrotron NR are postulated as the molecular basis of a human dis- XALOC team for beamline support. ease. Author contributions: P.F.-P. and E.E.-P. designed and In conclusion, our results suggest that GR multimer- supervised the project, obtained financial support, and ization needs to be reconsidered from the perspective of share overall responsibility. A.J.P., A.A.M., M.A.M. and multiple, energetically roughly equivalent homodimers. It R.A. obtained recombinant proteins. M.A.M., A.A.M. and is therefore conceivable that several different GR dimers A.J.P. performed crystallization trials. A.A.M, J.F.D. and coexist within a cell, all of which, in principle, recruit R.A. generated mutants and analyzed proteins. I.N.B. and different cofactors and thus activate different transcrip- A.R.M. performed and interpreted SCA analyses. J.F.R. tion pathways. ‘Selection’ of a given conformation might performed and interpreted docking, NIP, and ODA cal- be determined not only by the availability of GR-specific culations. T.T., R.L.S. and T.A.J. generated FL-GR mu- agonist/antagonist but also by many other parameters, in- tants. G.F. and D.M.P. performed and analyzed N&B exper- cluding cell state. Current structure-function information iments. A.J.P., A.A.M., R.A., P.F.-P. and E.E.-P. performed suggests that at least four different GR-LBD homodimers and interpreted MS SPR. A.J.P., P.F.-P. and E.E.-P. col- have roughly equal probabilities to be formed in vivo, which lected and interpreted X-ray diffraction data, P.F.-P. solved would in turn generate different tetrameric arrangements and refined structures. T.T., R.L.S., T.A.J., A.V.F., C.C. and on DNA. We are tempted to speculate that these individual P.P. contributed tools. T.T., T.A.J. and SK performed and conformations of tetrameric GR are associated with spe- analyzed the genomics data. D.M.P., and G.L.H. supervised cific transcriptional programs. Future investigations should cell experiments. A.J.P., P.F.-P. and E.E.-P. drafted the arti- verify the validity of this hypothesis and establish whether cle. A.R.M, T.A.J., J.F.R., D.M.P., G.L.H., P.F.-P. and E.E.- specific GR conformations result in unique expression pat- P. critically reviewed the manuscript. All authors approved terns. The high plasticity of GR-LBD for self-association the manuscript. and its potential modulation by different agonists and/or antagonists opens new venues for the potential treatment of GR-related diseases. The ability to promote or stabilize FUNDING these specific GR multimeric states might be an essential E.E.-P. thanks the generosity of the Gemma E. Carretero step towards the development of novel GCs with reduced Fund; MINECO [BFU2017-86906-R, SAF2017-71878- side effects. REDT, SAF2015-71878-REDT to E.E.-P., RTI2018- 101500-B-I00 to P.F.-P., RTI2018-096735-B-100 to A.R.M., DATA AVAILABILITY PID2019-110167RB-I00 to J.F.-R., SAF2017-89510-R to A.V.F. and C.C.]; G.L.H thanks the NIH Intramural The atomic coordinates and structure factors have been Research Program; D.M.P was supported by CONICET. deposited in the PDB with accession codes 7YXC (C2), Funding for open access charge: Spanish Ministry of 7YXD (P31 ), 7YXN (P61 ), 7YXO (I41 22), 7YXP (I41 32) Science (MINECO). and 7YXR (P61 , Y545A). The PDB data has been designed Conflict of interest statement. None declared. ‘for immediate release on publication’. MS data have been deposited to the ProteomeXchange Consortium (PRIDE partner repository: dataset identifier PXD028039). Bioin- REFERENCES formatics data and code for the clustering and SCA are 1. Conway-Campbell,B.L., Pooley,J.R., Hager,G.L. and Lightman,S.L. available at https://github.com/ibn90/GRPROJECT2021. (2011) Molecular dynamics of ultradian glucocorticoid receptor The Gene Expression profiles are deposited in the Gene Ex- action. Mol. Cell Endocrinol., 348, 383–393. 2. Busada,J.T. and Cidlowski,J.A. (2017) Mechanisms of glucocorticoid pression Omnibus (GEO) database and the code assigned is action during development. Curr. Top Dev. Biol., 125, 147–170. GSE212973. All relevant data are available upon request. 3. Bledsoe,R.K., Montana,V.G., Stanley,T.B., Delves,C.J., Apolito,C.J., McKee,D.D., Consler,T.G., Parks,D.J., Stewart,E.L., Willson,T.M. et al. (2002) Crystal structure of the glucocorticoid receptor ligand ACCESSION NUMBERS binding domain reveals a novel mode of receptor dimerization and coactivator recognition. Cell, 110, 93–105. The atomic coordinates and structure factors have been 4. Evans,R.M. and Mangelsdorf,D.J. (2014) Nuclear receptors, RXR, deposited in the Protein Data Bank (PDB) with acces- and the big bang. Cell, 157, 255–266. sion codes 7YXC (C2), 7YXD (P31 ), 7YXN (P61 ), 7YXO 5. Jimenez-Panizo,A., Perez,P., Rojas,A., Fuentes-Prior,P. and (I41 22), 7YXP (I41 32) and 7YXR (P61 , Y545A). The Gene Estebanez-Perpina,E. (2019) Non-canonical dimerization of the androgen receptor and other nuclear receptors: implications for Expression profiles are deposited in the Gene Expres- human disease. Endocr. Relat. Cancer, 26, R479–R497. sion Omnibus (GEO) database and the code assigned is 6. Housley,P.R., Sanchez,E.R., Danielsen,M., Ringold,G.M. and GSE212973. Pratt,W.B. (1990) Evidence that the conserved region in the steroid Nucleic Acids Research, 2022, Vol. 50, No. 22 13081 binding domain of the glucocorticoid receptor is required for both glucocorticoid receptor require DNA binding. Nucleic Acids Res., 48, optimal binding of hsp90 and protection from proteolytic cleavage. A 8393–8407. two-site model for hsp90 binding to the steroid binding domain. J. 27. Garcia,D.A., Johnson,T.A., Presman,D.M., Fettweis,G., Wagh,K., Biol. Chem., 265, 12778–12781. Rinaldi,L., Stavreva,D.A., Paakinaho,V., Jensen,R.A.M., 7. Ortlund,E.A., Bridgham,J.T., Redinbo,M.R. and Thornton,J.W. Mandrup,S. et al. (2021) An intrinsically disordered region-mediated (2007) Crystal structure of an ancient protein: evolution by confinement state contributes to the dynamics and function of conformational epistasis. Science, 317, 1544–1548. transcription factors. Mol. Cell, 81, 1484–1498 8. Meijsing,S.H., Pufall,M.A., So,A.Y., Bates,D.L., Chen,L. and 28. Weikum,E.R., de Vera,I.M.S., Nwachukwu,J.C., Hudson,W.H., Yamamoto,K.R. (2009) DNA binding site sequence directs Nettles,K.W., Kojetin,D.J. and Ortlund,E.A. (2017) Tethering not Downloaded from https://academic.oup.com/nar/article/50/22/13063/6858817 by Fundacio Biblioteca Josep Laporte user on 09 October 2023 glucocorticoid receptor structure and activity. Science, 324, 407–410. required: the glucocorticoid receptor binds directly to activator 9. Rogatsky,I., Wang,J.C., Derynck,M.K., Nonaka,D.F., protein-1 recognition motifs to repress inflammatory genes. Nucleic Khodabakhsh,D.B., Haqq,C.M., Darimont,B.D., Garabedian,M.J. Acids Res., 45, 8596–8608. and Yamamoto,K.R. (2003) Target-specific utilization of 29. Chrousos,G.P., Loriaux,D.L., Tomita,M., Brandon,D.D., transcriptional regulatory surfaces by the glucocorticoid receptor. Renquist,D., Albertson,B. and Lipsett,M.B. (1986) The new world Proc. Natl. Acad. Sci. U.S.A., 100, 13845–13850. primates as animal models of glucocorticoid resistance. Adv. Exp. 10. Pfaff,S.J. and Fletterick,R.J. (2010) Hormone binding and Med. Biol., 196, 129–144. co-regulator binding to the glucocorticoid receptor are allosterically 30. Luisi,B.F., Xu,W.X., Otwinowski,Z., Freedman,L.P., coupled. J. Biol. Chem., 285, 15256–15267. Yamamoto,K.R. and Sigler,P.B. (1991) Crystallographic Analysis of 11. Estebanez-Perpina,E., Arnold,L.A., Nguyen,P., Rodrigues,E.D., the Interaction of the Glucocorticoid Receptor with DNA. Nature, Mar,E., Bateman,R., Pallai,P., Shokat,K.M., Baxter,J.D., Guy,R.K. 352, 497–505. et al. (2007) A surface on the androgen receptor that allosterically 31. Hudson,W.H., Youn,C. and Ortlund,E.A. (2013) The structural basis regulates coactivator binding. Proc. Natl. Acad. Sci. U.S.A., 104, of direct glucocorticoid-mediated transrepression. Nat. Struct. Mol. 16074–16079. Biol., 20, 53–58. 12. Jehle,K., Cato,L., Neeb,A., Muhle-Goll,C., Jung,N., Smith,E.W., 32. Frank,F., Okafor,C.D. and Ortlund,E.A. (2018) The first crystal Buzon,V., Carbo,L.R., Estebanez-Perpina,E., Schmitz,K. et al. structure of a DNA-free nuclear receptor DNA binding domain (2014) Coregulator control of androgen receptor action by a novel sheds light on DNA-driven allostery in the glucocorticoid receptor. nuclear receptor-binding motif. J. Biol. Chem., 289, 8839–8851. Sci. Rep., 8, 13497. 13. Fuentes-Prior,P., Rojas,A., Hagler,A.T. and Estebanez-Perpina,E. 33. Hard,T., Kellenbach,E., Boelens,R., Maler,B.A., Dahlman,K., (2019) Diversity of quaternary structures regulates nuclear receptor Freedman,L.P., Carlstedt-Duke,J., Yamamoto,K.R., Gustafsson,J.A. activities. Trends Biochem. Sci., 44, 2–6. and Kaptein,R. (1990) Solution structure of the glucocorticoid 14. Vandevyver,S., Dejager,L. and Libert,C. (2012) On the trail of the receptor DNA-binding domain. Science, 249, 157–160. glucocorticoid receptor: into the nucleus and back. Traffic, 13, 34. Schacke,H., Berger,M., Rehwinkel,H. and Asadullah,K. (2007) 364–374. Selective glucocorticoid receptor agonists (SEGRAs): novel ligands 15. Weikum,E.R., Knuesel,M.T., Ortlund,E.A. and Yamamoto,K.R. with an improved therapeutic index. Mol. Cell Endocrinol., 275, (2017) Glucocorticoid receptor control of transcription: precision and 109–117. plasticity via allostery. Nat. Rev. Mol. Cell. Biol., 18, 159–174. 35. Biggadike,K., Bledsoe,R.K., Coe,D.M., Cooper,T.W., House,D., 16. Clark,A.R. and Belvisi,M.G. (2012) Maps and legends: the quest for Iannone,M.A., Macdonald,S.J., Madauss,K.P., McLay,I.M., dissociated ligands of the glucocorticoid receptor. Pharmacol Ther., Shipley,T.J. et al. (2009) Design and x-ray crystal structures of 134, 54–67. high-potency nonsteroidal glucocorticoid agonists exploiting a novel 17. Franco,L.M., Gadkari,M., Howe,K.N., Sun,J., Kardava,L., binding site on the receptor. Proc. Natl. Acad. Sci. U.S.A., 106, Kumar,P., Kumari,S., Hu,Z., Fraser,I.D.C., Moir,S. et al. (2019) 18114–18119. Immune regulation by glucocorticoids can be linked to cell 36. Hurt,D.E., Suzuki,S., Mayama,T., Charmandari,E. and Kino,T. type-dependent transcriptional responses. J. Exp. Med., 216, 384–406. (2016) Structural analysis on the pathologic mutant glucocorticoid 18. Sevilla,L.M., Bayo,P., Latorre,V., Sanchis,A. and Perez,P. (2010) receptor ligand-binding domains. Mol. Endocrinol., 30, 173–188. Glucocorticoid receptor regulates overlapping and differential gene 37. Liu,X., Wang,Y. and Ortlund,E.A. (2019) First high-resolution subsets in developing and adult skin. Mol. Endocrinol., 24, crystal structures of the glucocorticoid receptor ligand-binding 2166–2178. domain-peroxisome proliferator-activated gamma coactivator 19. Cain,D.W. and Cidlowski,J.A. (2017) Immune regulation by 1-alpha complex with endogenous and synthetic glucocorticoids. glucocorticoids. Nat. Rev. Immunol., 17, 233–247. Mol. Pharmacol., 96, 408–417. 20. Oh,K.S., Patel,H., Gottschalk,R.A., Lee,W.S., Baek,S., Fraser,I.D.C., 38. Weikum,E.R., Okafor,C.D., D’Agostino,E.H., Colucci,J.K. and Hager,G.L. and Sung,M.H. (2017) Anti-inflammatory Ortlund,E.A. (2017) Structural analysis of the glucocorticoid chromatinscape suggests alternative mechanisms of glucocorticoid receptor ligand-binding domain in complex with triamcinolone receptor action. Immunity, 47, 298–309. acetonide and a fragment of the atypical coregulator, small 21. De Bosscher,K., Desmet,S.J., Clarisse,D., Estebanez-Perpina,E. and heterodimer partner. Mol. Pharmacol., 92, 12–21. Brunsveld,L. (2020) Nuclear receptor crosstalk - defining the 39. Johnson,T.A., Paakinaho,V., Kim,S., Hager,G.L. and Presman,D.M. mechanisms for therapeutic innovation. Nat. Rev. Endocrinol., 16, (2021) Genome-wide binding potential and regulatory activity of the 363–377. glucocorticoid receptor’s monomeric and dimeric forms. Nat. 22. Louw,A. (2019) GR dimerization and the impact of GR dimerization Commun., 12, 1987. on GR protein stability and half-life. Front. Immunol., 10, 1693. 40. Mikuni,S., Tamura,M. and Kinjo,M. (2007) Analysis of intranuclear 23. Presman,D.M., Ogara,M.F., Stortz,M., Alvarez,L.D., Pooley,J.R., binding process of glucocorticoid receptor using fluorescence Schiltz,R.L., Grontved,L., Johnson,T.A., Mittelstadt,P.R., correlation spectroscopy. FEBS Lett., 581, 389–393. Ashwell,J.D. et al. (2014) Live cell imaging unveils multiple domain 41. Digman,M.A., Dalal,R., Horwitz,A.F. and Gratton,E. (2008) requirements for in vivo dimerization of the glucocorticoid receptor. Mapping the number of molecules and brightness in the laser PLoS Biol., 12, e1001813. scanning microscope. Biophys. J., 94, 2320–2332. 24. Presman,D.M., Ganguly,S., Schiltz,R.L., Johnson,T.A., Karpova,T.S. 42. McNally,J.G., Mueller,W.G., Walker,D., Wolford,R.G. and and Hager,G.L. (2016) DNA binding triggers tetramerization of the Hager,G.L. (2000) The glucocorticoid receptor: rapid exchange with glucocorticoid receptor in live cells. Proc. Natl. Acad. Sci. U.S.A., regulatory sites in living cells. Science, 287, 1262–1265. 113, 8236–8241. 43. Dobin,A. and Gingeras,T.R. (2015) Mapping RNA-seq reads with 25. Paakinaho,V., Johnson,T.A., Presman,D.M. and Hager,G.L. (2019) STAR. Curr. Protoc. Bioinformatics, 51, 11.14.1–11.14.19. Glucocorticoid receptor quaternary structure drives chromatin 44. Love,M.I., Huber,W. and Anders,S. (2014) Moderated estimation of occupancy and transcriptional outcome. Genome Res., 29, 1223–1234. fold change and dispersion for RNA-seq data with DESeq2. Genome 26. Escoter-Torres,L., Greulich,F., Quagliarini,F., Wierer,M. and Biol., 15, 550. Uhlenhaut,N.H. (2020) Anti-inflammatory functions of the 13082 Nucleic Acids Research, 2022, Vol. 50, No. 22 45. Cheng,T.M., Blundell,T.L. and Fernandez-Recio,J. (2007) pyDock: for duchenne muscular dystrophy treatment. Proc. Natl. Acad. Sci. electrostatics and desolvation for effective scoring of rigid-body U.S.A., 117, 24285–24293. protein-protein docking. Proteins, 68, 503–515. 57. Hurley,D.M., Accili,D., Stratakis,C.A., Karl,M., Vamvakopoulos,N., 46. Canutescu,A.A., Shelenkov,A.A. and Dunbrack,R.L. Jr (2003) A Rorer,E., Constantine,K., Taylor,S.I. and Chrousos,G.P. (1991) Point graph-theory algorithm for rapid protein side-chain prediction. mutation causing a single amino acid substitution in the hormone Protein. Sci., 12, 2001–2014. binding domain of the glucocorticoid receptor in familial 47. Gabb,H.A., Jackson,R.M. and Sternberg,M.J. (1997) Modelling glucocorticoid resistance. J. Clin. Invest., 87, 680–686. protein docking using shape complementarity, electrostatics and 58. Nadal,M., Prekovic,S., Gallastegui,N., Helsen,C., Abella,M., biochemical information. J. Mol. Biol., 272, 106–120. Zielinska,K., Gay,M., Vilaseca,M., Taules,M., Houtsmuller,A.B. Downloaded from https://academic.oup.com/nar/article/50/22/13063/6858817 by Fundacio Biblioteca Josep Laporte user on 09 October 2023 48. Chen,R., Li,L. and Weng,Z. (2003) ZDOCK: an initial-stage et al. (2017) Structure of the homodimeric androgen receptor protein-docking algorithm. Proteins, 52, 80–87. ligand-binding domain. Nat. Commun., 8, 14388. 49. Fernandez-Recio,J., Totrov,M., Skorodumov,C. and Abagyan,R. 59. Kauppi,B., Jakob,C., Farnegardh,M., Yang,J., Ahola,H., (2005) Optimal docking area: a new method for predicting Alarcon,M., Calles,K., Engstrom,O., Harlan,J., Muchmore,S. et al. protein-protein interaction sites. Proteins, 58, 134–143. (2003) The three-dimensional structures of antagonistic and agonistic 50. Grosdidier,S. and Fernandez-Recio,J. (2008) Identification of forms of the glucocorticoid receptor ligand-binding domain: RU-486 hot-spot residues in protein-protein interactions by computational induces a transconformation that leads to active antagonism. J. Biol. docking. BMC Bioinformatics, 9, 447. Chem., 278, 22748–22754. 51. He,Y., Yi,W., Suino-Powell,K., Zhou,X.E., Tolbert,W.D., Tang,X., 60. Helsen,C. and Claessens,F. (2014) Looking at nuclear receptors from Yang,J., Yang,H., Shi,J., Hou,L. et al. (2014) Structures and a new angle. Mol. Cell Endocrinol., 382, 97–106. mechanism for the design of highly potent glucocorticoids. Cell Res., 61. Bianchetti,L., Wassmer,B., Defosset,A., Smertina,A., Tiberti,M.L., 24, 713–726. Stote,R.H. and Dejaegere,A. (2018) Alternative dimerization 52. Chagoyen,M., Garcia-Martin,J.A. and Pazos,F. (2016) Practical interfaces in the glucocorticoid receptor-alpha ligand binding analysis of specificity-determining residues in protein families. Brief domain. Biochim. Biophys. Acta. Gen. Subj., 1862, 1810–1825. Bioinform., 17, 255–261. 62. Noddings,C.M., Wang,R.Y., Johnson,J.L. and Agard,D.A. (2022) 53. Halabi,N., Rivoire,O., Leibler,S. and Ranganathan,R. (2009) Protein Structure of Hsp90-p23-GR reveals the hsp90 client-remodelling sectors: evolutionary units of three-dimensional structure. Cell, 138, mechanism. Nature, 601, 465–469. 774–786. 63. Min,J., Perera,L., Krahn,J.M., Jewell,C.M., Moon,A.F., 54. Lockless,S.W. and Ranganathan,R. (1999) Evolutionarily conserved Cidlowski,J.A. and Pedersen,L.C. (2018) Probing dominant negative pathways of energetic connectivity in protein families. Science, 286, behavior of glucocorticoid receptor beta through a hybrid structural 295–299. and biochemical approach. Mol. Cell Biol., 38 , e00453-17. 55. Presman,D.M., Ball,D.A., Paakinaho,V., Grimm,J.B., Lavis,L.D., 64. Schoch,G.A., D’Arcy,B., Stihle,M., Burger,D., Bar,D., Benz,J., Karpova,T.A. and Hager,G.L. (2017) Quantifying transcription Thoma,R. and Ruf,A. (2010) Molecular switch in the glucocorticoid factor dynamics at the single-molecule level in live cells. Methods, receptor: active and passive antagonist conformations. J. Mol. Biol., 123, 76–88. 395, 568–577. 56. Liu,X., Wang,Y., Gutierrez,J.S., Damsker,J.M., Nagaraju,K., 65. Gampe,R.T. Jr, Montana,V.G., Lambert,M.H., Wisely,G.B., Hoffman,E.P. and Ortlund,E.A. (2020) Disruption of a key Milburn,M.V. and Xu,H.E. (2000) Structural basis for autorepression ligand-H-bond network drives dissociative properties in vamorolone of retinoid x receptor by tetramer formation and the AF-2 helix. Genes Dev., 14, 2229–2241.