Structural Insights into the Loss-of-Function R288H Mutant of Human PPARγ

Abstract

Peroxisome proliferator-activated receptor gamma (PPARγ) is a nuclear receptor and the molecular target of thiazolidinedione-class antidiabetic drugs. It has been reported that the loss of function R288H mutation in the human PPARγ ligand-binding domain (LBD) may be associated with the onset of colon cancer. A previous in vitro study showed that this mutation dampens 15-deoxy-Δ^12,14-prostaglandin J2 (15d-PGJ2, a natural PPARγ agonist)-dependent transcriptional activation; however, it is poorly understood why the function of the R288H mutant is impaired and what role this arginine (Arg) residue plays. In this study, we found that the apo-form of R288H PPARγ mutant displays several altered conformational arrangements of the amino acid side chains in LBD: 1) the loss of a salt bridge between Arg288 and Glu295 leads to increased helix 3 movement; 2) closer proximity of Gln286 and His449 via a hydrogen bond, and closer proximity of Cys285 and Phe363 via hydrophobic interaction, stabilize the helix 3–helix 11 interaction; and 3) there is steric hindrance between Cys285/Gln286/Ser289/His449 and the flexible ligands 15d-PGJ2, 6-oxotetracosahexaenoic acid (6-oxoTHA), and 17-oxodocosahexaenoic acid (17-oxoDHA). These results suggest why Arg288 plays an important role in ligand binding and why the R288H mutation is disadvantageous for flexible ligand binding.

INTRODUCTION

Peroxisome proliferator-activated receptor gamma (PPARγ) is a target molecule for insulin-sensitizing agents such as pioglitazone, and is reported to be associated with various diseases and pathological conditions such as inflammation and cancer.^1,2) PPARγ is expressed in a variety of both tumor and normal cells. PPARγ-expressing cancer cells can be induced to inhibit cell proliferation³⁾ or to undergo apoptosis⁴⁾ by PPARγ agonists such as thiazolidinedione (TZD)⁵⁾ and 15-deoxy-Δ^12,14-prostaglandin J2 (15d-PGJ2) (Fig. 1), and mutagenesis of PPARγ is associated with colon cancer. PPARγ therefore may play a role as a tumor suppressor.⁶⁾ In the PPARγ mutant R288H,⁷⁾ an arginine (Arg) residue at position 288 is mutated to histidine (His) during carcinogenesis and the mutant is found in colorectal cancer tumors. This mutant reduces the activity of 15d-PGJ2 and decreases ligand binding. Helix 3 of PPARγ forms the largest area of contact with the ligand because agonists wrap around it. Arg288 is located in the center of helix 3 and its role in ligand binding is not well understood. We report here structural biology studies of the R288H mutant and elucidate the mechanism of its change in transcriptional activity and the role of R288H in the protein.

Fig. 1. Transcriptional Activity of PPARγ WT and R288H Stimulated by Various Ligands (Pioglitazone, LT175, 15d-PGJ2, 6-OxoTHA and 17-OxoDHA)

Cos7 cells were transfected with a GAL4-PPAR chimera expression plasmid (pSG5-GAL-hPPARγ), a reporter plasmid (MH100 × 4-TK-Luc), and an internal control plasmid containing sea pansy luciferase expression constructs (pRL-CMV).

RESULTS

We evaluated how the R288H mutation affects transcriptional activity by using various ligands and evaluating gene transcriptional activity by the luciferase assay. Pioglitazone and (S)-LT175 (LT175)⁸⁾ were used as sp² carbon-rich (rigid) agonists and 15d-PGJ2,⁹⁾ 6-oxotetracosahexaenoic acid (6-oxoTHA)¹⁰⁾ and 17-oxodocosahexaenoic acid (17-oxoDHA)¹¹⁾ were used as sp³ carbon-rich (flexible) agonists. As previously reported, the activity of 15d-PGJ2 decreased in the presence of R288H,⁷⁾ as did the activities of the flexible agonists 6-oxoTHA and 17-oxoDHA. However, the rigid agonists pioglitazone and LT175 did not show any decrease in activity, indicating different behavior of the mutant depending on the ligand properties. This interesting result cannot be explained by the decrease in activity of 15d-PGJ2 being due to destabilization of the ligand binding domain (LBD) by the mutation. We therefore compared the structural stability of the apo-forms of wild type (WT) and R288H.

The stabilities of WT apo and R288H apo were compared using guanidine denaturation experiments.¹²⁾ Guanidine hydrochloride (GdmCl) was added at concentrations of 0, 0.5, 1.0, 2.0, 3.0, and 6.0 M to the expressed and purified PPARγ-LBD WT and mutant R288H apo LBDs, and the circular dichroism (CD) spectra were measured (Figs. 2a, b). Monitoring at 222 nm, in 0 M GdmCl WT and R288H gave minima of −69 and −75 mdeg, respectively, and in 6 M GdmCl the minima were −21 and −56 mdeg, respectively. Figure 2c shows the degree of denaturation at different GdmCl concentrations, where the difference between the minimum value and zero (mdeg, 222 nm) was converted to % and compared. GdmCl (6M) resulted in 69% denaturation of the WT and 26% of the R288H mutant, indicating that the R288H mutation structurally stabilizes the apo-form of PPARγ-LBD.

Fig. 2. Biophysical Characterization and Evaluation of (a) 20 µM PPARγ WT LBD and (b) R288H LBD in the Apo-Form

Circular dichroism (CD) spectra from 200 to 250 nm, and the results of denaturation using GdmCl (0, 0.5, 1.0, 2.0, 3.0, 6.0 M). Each step from 0 to 6 M is shown in black and increasingly paler shades of gray. (c) The degree of denaturation of WT and R288H is shown, where the difference between the minimum value and zero at 222 nm in (a) or (b) was converted to % and compared.

We performed X-ray crystal structure analysis of PPARγ apo LBD R288H (204–477) to investigate the mechanism underlying its structural stabilization. The cell dimension (Table 1) was almost identical to those reported for LBD WT apo crystal-(protein data bank (PDB) accession code: 2ZK0)

Table 1. Summary of Data Collection Statistics and Refinement of Crystal Structures

X-Ray source	KEK-PFAR NW-12A
Wavelength (Å)	1.00000
Detector	ADSC Quantum 270
Space group	C 1 2 1
Unit cell dimensions (Å)	a = 91.04, b = 60.92, c = 116.73
(°)	α = 90.00, β = 102.24, γ = 90.00
Resolution range (Å)	57.04–3.20 (3.42–3.20)a)
Total number of reflections	36992
No. of unique reflections	10209
Rmerge	0.167 (0.261)a)
I/σI	6.3 (4.2)a)
% completeness	97.5 (99.9)a)
Redundancy	3.6 (3.8)a)
Refinement statistics
Resolution range (Å)	114.08–3.20
R factor (Rfree/Rwork)	0.3133/0.2338
No. atoms	3935
Protein	3910
Water	25
RMSDs
Bond lengths (Å)	0.0135
Bond angles (°)	1.6819
Ramachandran plot
Favored regions (%)	94
Outliers (%)	0

a) Values in parentheses are for the highest-resolution shell.

The obtained X-ray crystal structure of R288H apo (PDB ID: 7E2O) and WT apo (PDB ID: 2ZK0)⁹⁾ were compared (Figs. 3a–d, f) and showed that Arg288 and Glu295 formed a long-range salt bridge^13–21) in WT (Fig. 3b). This salt bridge was absent in the R288H mutant and thus the mobility of helix 3 was increased. In WT, Gln286 resides on the opposite side of helix 3 from Arg288 and formed hydrogen bonds with Ser464 and Gln283 (Fig. 3b). In R288H mutant, Gln286 flipped toward the ligand binding cavity and formed a hydrogen bond with His449 in helix11. The Cα distance between these two amino acids was 12.8 Å in WT and 10.7 Å in R288H (Figs. 3c, d), indicating that they were much closer in the mutant and that the ligand binding cavity decreased in width by approximately the size of a water molecule. Furthermore, Cys285, located in the center of the ligand binding cavity and adjacent to Gln286 in WT, was pushed to the inside of the ligand binding cavity in R288H, resulting in the side chain forming tight hydrophobic interactions with Phe363 in helix 7. This suggests that the hydrogen bonding and hydrophobic interactions increased the stability of the R288H mutant in the apo-form (Fig. 3d). In addition, Ser289 approached His449, as did Gln286 and Cys285. Superimposition of 6-oxoTHA bound to PPARγ (PDB ID: 3X1I) onto R288H revealed that the ligand was located within steric clash distance of Cys285, Gln286, Ser289 and His449 (Fig. 3e), indicating that these amino acid residues can inhibit ligand binding.

Fig. 3. Comparison of WT (Gray) and R288H (Cyan) Structures of PPARγ-LBD

a) Overlay of the whole structure. b) Enlarged view near helix 3. c), d) Gln286 flips and approaches His449. e) The superimposed structure of 6-oxoTHA (3X1I.pdb) and PPARγ R288H. 6-OxoTHA causes steric clash with Cys285, Gln286, Ser289 and His449. f) RMSD (Å) at the Cα’s of LBD between WT and R288H. (Color figure can be accessed in the online version.)

We calculated the root-mean-square deviation (RMSD) at the Cα’s of LBD (Fig. 3f) and found a large difference around helix 2. The true change was probably larger because it occurred in the outer section of the disordered region. Differences in the loop between helix 11 and helix 12 may be due to the absence of a hydrogen bond at Ser464 (Fig. 3b), thus increasing the degree of movement of helix 12 (Fig. 3a). It should be noted that Han and co-workers recently published the structure of R288H co-crystallized with the co-activator SRC-1.²²⁾ Although helix 2 in the complex is also disordered, differences around helix 11 and 12 are not significant, in contrast to our crystal structure. We speculate that the interaction of SRC-1 with R288H may have led to the canonical active conformation of R288H.

DISCUSSION

The structural stability of the LBD of WT PPARγ and R288H was evaluated using CD spectra. The secondary structure was maintained even in 6 M GdmCl. Assuming that the molten globule state is in equilibrium with the canonical ternary structure, the mutation did not lead to the unfolded state so much, in contrast to WT, because of strong interactions of the amino acid residues rearranged by the mutation. The crystal structure suggests that the R288H mutation results in the formation of new hydrogen bonds in the ligand binding cavity resulting in a more stable LBD structure. The results of the dual-luciferase assay showed that the effect of the mutation was different depending on the properties of the ligand.

In the R288H mutant, flexible ligands with reduced gene transcriptional activity made it difficult for the ligand to bind, possibly due to the increased stability of PPARγ in the apo-form, which reduces the benefit of ligand-induced structural stabilization. All the fatty acids tested in this study are covalent modifiers.

It is possible that covalent modification by nucleophilic addition of the thiol group of cysteine is less likely to proceed because Cys285 is in close proximity to the surrounding amino acid residues, which reduces the probability of the enone moiety in flexible ligands, 15d-PGJ2, 6-oxoTHA and 17-oxoDHA, approaching the thiol group.²¹⁾ R288H is similar to WT in the efficacy of the transcriptional activity of 6-oxoTHA (Fig. 1). Thus, once the ligand enters the LBD, the LBD adopts the canonical active conformation, suggesting that the R288H mutation may be stronger effect on ligand binding than on LBD activation. On the other hand, for rigid ligands, which have a strong entropic advantage, structural stabilization by complex formation is favorable rather than by helix-helix interactions in apo-form. This study showed the specific role of Arg288. Our findings may be useful for the design of new PPARγ ligands.

MATERIALS AND METHODS

Materials

6-OxoTHA and 17-oxoDHA were prepared in our laboratory. All other substrates and reagents were purchased from commercial sources and were used without further purification.

Plasmid Construction and Point Mutagenesis Study

hPPARγ-LBD R288H mutant were generated using Pfu Turbo™ DNA polymerase (Agilent Technologies, Santa Clara, CA, U.S.A.). The R288H mutant plasmid was transformed into E. coli DH5α cells and plasmids were purified using a QIAprep spin miniprep kit (Qiagen, Valencia, CA, U.S.A.). The sequence of the R288H mutagenic primer (forward direction only) is 5′-GTT TCA CTC CGT GGA GGC TGT G-3′. The plasmid was confirmed by di-deoxy sequencing.

Transfection and Transactivation Assay

COS-7 cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 5% fetal bovine serum (FBS). Cells were seeded on 24-well plates at a density of 2 × 10⁴ per well. After 24 h, a mixture containing 0.18 µg of a reporter plasmid (MH100 × 4-TK-Luc),⁸⁾ 0.05 µg of a pSG5-GAL-hPPARγ chimera expression plasmid, 0.02 µg of the internal control plasmid containing sea pansy luciferase expression constructs (pRL-CMV), and 0.75 µL of the Trans IT-LT1 reagent (Mirus, Madison, WI, U.S.A.) were added to each well.¹⁰⁾ The MH100 × 4-TK-Luc reporter plasmid contains four copies of the MH100 GAL4 binding site.⁹⁾ After 8 h incubation, the cells were treated with either the ligand or ethanol vehicle and were cultured for 16 h. Cells in each well were harvested with a cell lysis buffer, and their luciferase activity was measured using a luciferase assay kit (Promega, WI, U.S.A.). Transactivation that was measured as luciferase activity was normalized with the internal control. All experiments were performed in triplicate.

Protein Expression and Purification

The human PPARγ-ligand binding domain (LBD) (aa 204–477) was expressed using a modified pET30a vector with an N-terminal 6 × histidine (His) tag cleavable by TEV protease. Escherichia coli (E. coli) Rosetta (DE3) was freshly transformed with the plasmid and grown in four flasks containing 1 L of 2 × TY medium with kanamycin 34 µg mL⁻¹ and chloramphenicol 50 µg mL⁻¹ at 37 °C to an optical density (OD) at 600 nm of 1.0. Protein synthesis was then induced with 0.5 mM isopropyl-β-_D-thiogalactopyranoside and the cultures were further incubated at 20 °C for 18 h. Cells were harvested and resuspended in 50 mL lysis buffer (20 mM Tris–HCl pH 8.0, 100 mM NaCl, 1 mM TCEP, 0.5 mM ethylenediaminetetraacetic acid (EDTA), × 1 Protease inhibitor cocktail). Cells were lysed by sonication, and the soluble fraction was isolated by centrifugation (18000 × g for 20 min). The supernatant was applied to cOmplete His-Tag Purification Resin and the resin was thoroughly washed in wash buffer (20 mM Tris–HCl pH 8.0, 100 mM NaCl, 1 mM TCEP, 10 mM imidazole). The human PPARγ-LBD was eluted with elution buffer (20 mM Tris–HCl pH 8.0, 100 mM NaCl, 1 mM TCEP, 250 mM imidazole). TEV protease was added to the eluate and the mixture was dialyzed overnight at RT, dialysis with 500 mL of buffer (20 mM Tris–HCl pH 8.0, 1 mM TCEP, 0.5 mM EDTA). The cleaved protein was passed through cOmplete His-Tag Purification Resin. The flow-through was loaded onto a Resource Q (6 mL) column (GE Healthcare, Chicago, IL, U.S.A.) equilibrated with buffer (20 mM Tris–HCl pH 8.0, 1 mM TCEP, 0.5 mM EDTA). The column was eluted with a NaCl gradient from 0 to 0.5 M in the starting buffer. The eluted fractions were concentrated and loaded onto a Superdex 75 (24 mL) gel-filtration column equilibrated with buffer (20 mM Tris–HCl pH 8.0, 1 mM TCEP, 0.5 mM EDTA)

GdmCl Denaturation Experiment

GdmCl (≥99%) was purchased from Nacalai Tesque (Kyoto, Japan). In each GdmCl denaturation experiment, samples of PPARγ were titrated with GdmCl from 0 to 6.0 M at a protein concentration of 20 µM. Unfolding was initiated by dilution of a concentrated protein stock into the appropriate GdmCl buffer. All samples were incubated at 20 °C for 2 h before measurement using a JASCO J-1500 CD spectrometer.¹⁰⁾

Protein Crystallization

Crystallization was performed by vapor diffusion at room temperature using a hanging drop that was made by mixing protein solution and reservoir solution.

Two microliters of hPPARγ-LBD solution (8.0 mg mL⁻¹, in 20 mM Tris–HCl pH 8.0, 1 mM TCEP, 0.5 mM EDTA) in 1 µL of reservoir solution (0.1 M Tris–HCl pH 7.4, 0.7 M sodium citrate). The mixture was stored in the dark and prismatic crystals appeared after a few days. Crystal was flash-cooled in liquid nitrogen after a fast soaking in a cryoprotectant buffer (reservoir solution with glycerol 24% (v/v)).

X-Ray Crystallographic Analysis

Diffraction data sets were collected at the beamline NW-12A of the Photon Factory Advanced Ring (PF-AR) at the high energy accelerator research organization (KEK, Tsukuba, Japan). Reflections were recorded with an oscillation range per image of 1.0°. Data were indexed, integrated, and scaled using the iMOSFLM^23–25) and the CCP4²⁶⁾ suite of programs. The structure was solved using molecular replacement and was rebuilt and refined using COOT,²⁷⁾ and REFMAC.²⁸⁾ The coordinate data for the structure was deposited in the Protein Data Bank under the accession number 7E2O.

Acknowledgments

This work was financially supported by the Platform for Drug Discovery, Informatics, and Structural Life Science from the Japan Agency for Medical Research and Development (AMED) to K.Y. and a Grant-in-Aid for Scientific Research (No. 22790116) from the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan to T.I. Synchrotron radiation experiments were performed at the Photon Factory (Proposal No. 2013G656), and we are grateful for the assistance provided by the beamline scientists at the Photon Factory.

Conflict of Interest

The authors declare no conflict of interest.

REFERENCES

• 1) Schoonjans K, Staels B, Auwerx J. Role of the peroxisome proliferator-activated receptor (PPAR) in mediating the effects of fibrates and fatty acids on gene expression. J. Lipid Res., 37, 907–925 (1996).
• 2) Janani C, Ranjitha Kumari BD. PPAR gamma gene—a review. Diabetes Metab. Syndr., 9, 46–50 (2015).
• 3) Grommes C, Landreth GE, Heneka MT. Antineoplastic effects of peroxisome proliferator-activated receptor gamma agonists. Lancet Oncol., 5, 419–429 (2004).
• 4) Theocharis S, Margeli A, Vielh P, Kouraklis G. Peroxisome proliferator-activated receptor-gamma ligands as cell-cycle modulators. Cancer Treat. Rev., 30, 545–554 (2004).
• 5) Lehmann JM, Moore LB, Smith-Oliver TA, Wilkison WO, Willson TM, Kliewer SA. An antidiabetic thiazolidinedione is a high affinity ligand for peroxisome proliferator-activated receptor γ (PPARγ). J. Biol. Chem., 270, 12953–12956 (1995).
• 6) Tachibana K, Yamasaki D, Ishimoto K, Doi T. The role of PPARs in cancer. PPAR Res., 2008, 102737 (2008).
• 7) Sarraf P, Mueller E, Smith WM, Wright HM, Kum JB, Aaltonen LA, de la Chapelle A, Spiegelman BM, Eng C. Loss-of-function mutations in PPARg associated with human colon cancer. Mol. Cell, 3, 799–804 (1999).
• 8) Montanari R, Saccoccia F, Scotti E, Crestani M, Godio C, Gilardi F, Loiodice F, Fracchiolla G, Laghezza A, Tortorella P, Lavecchia A, Novellino E, Mazza F, Aschi M, Pochetti G. Crystal structure of the peroxisome proliferator-activated receptor gamma (PPARgamma) ligand binding domain complexed with a novel partial agonist: a new region of the hydrophobic pocket could be exploited for drug design. J. Med. Chem., 51, 7768–7776 (2008).
• 9) Waku T, Shiraki T, Oyama T, Fujimoto Y, Maebara K, Kamiya N, Jingami H, Morikawa K. Structural insight into PPARgamma activation through covalent modification with endogenous fatty acids. J. Mol. Biol., 385, 188–199 (2009).
• 10) Egawa D, Itoh T, Yamamoto K. Characterization of covalent bond formation between PPARγ and oxo-fatty acids. Bioconjug. Chem., 26, 690–698 (2015).
• 11) Egawa D, Itoh T, Akiyama Y, Saito T, Yamamoto K. 17-OxoDHA is a PPARα/γ dual covalent modifier and agonist. ACS Chem. Biol., 11, 2447–2455 (2016).
• 12) Montanari R, Capelli D, Yamamoto K, Awaishima H, Nishikata K, Barendregt A, Heck AJR, Loiodice F, Altieri F, Paiardini A, Grottesi A, Pirone L, Pedone E, Peiretti F, Brunel JM, Itoh T, Pochetti G. Insights into PPARγ phosphorylation and its inhibition mechanism. J. Med. Chem., 63, 4811–4823 (2020).
• 13) Barlow DJ, Thornton JM. Ion-pairs in proteins. J. Mol. Biol., 168, 867–885 (1983).
• 14) Bosshard HR, Marti DN, Jelesarov I. Protein stabilization by salt bridges: concepts, experimental approaches and clarification of some misunderstandings. J. Mol. Recognit., 17, 1–16 (2004).
• 15) Kumar S, Nussinov R. Relationship between ion pair geometries and electrostatic strengths in proteins. Biophys. J., 83, 1595–1612 (2002).
• 16) Gong H, Freed KF. Electrostatic solvation energy for two oppositely charged ions in a solvated protein system: salt bridges can stabilize proteins. Biophys. J., 98, 470–477 (2010).
• 17) Salari R, Chong LT. Desolvation costs of salt bridges across protein binding interfaces: similarities and differences between implicit and explicit solvent models. J. Phys. Chem. Lett., 1, 2844–2848 (2010).
• 18) Salari R, Chong LT. Effects of high temperature on desolvation costs of salt bridges across protein binding interfaces: similarities and differences between implicit and explicit solvent models. J. Phys. Chem. B, 116, 2561–2567 (2012).
• 19) Smith MS, Billings WM, Whitby FG, Miller MB, Price JL. Enhancing a long-range salt bridge with intermediate aromatic and nonpolar amino acids. Org. Biomol. Chem., 15, 5882–5886 (2017).
• 20) Zhou HX, Pang X. Electrostatic interactions in protein structure, folding, binding, and condensation. Chem. Rev., 118, 1691–1741 (2018).
• 21) Itoh T, Fairall L, Amin K, Inaba Y, Szanto A, Balint BL, Nagy L, Yamamoto K, Schwabe JWR. Structural basis for the activation of PPARγ by oxidized fatty acids. Nat. Struct. Mol. Biol., 15, 924–931 (2008).
• 22) Jang DM, Jang JY, Kim HJ, Han BW. Differential effects of cancer-associated mutations enriched in helix H3 of PPARγ. Cancers, 12, 3580 (2020).
• 23) Leslie AGW. The Integration of Macromolecular Diffraction Data. Acta Crystallogr. D Biol. Crystallogr., 62, 48–57 (2006).
• 24) Leslie AGW, Powell HR. Processing diffraction data with mosflm. Springer, Dordrecht, pp. 41–51 (2007).
• 25) Battye TGG, Kontogiannis L, Johnson O, Powell HR, Leslie AGW. IMOSFLM: a new graphical interface for diffraction-image processing with MOSFLM. Acta Crystallogr. D Biol. Crystallogr., 67, 271–281 (2011).
• 26) Collaborative Computational Project, Number 4. The CCP4 suite: programs for protein crystallography. Acta Crystallogr. D Biol. Crystallogr., 50, 760–763 (1994).
• 27) Emsley P, Cowtan K. Coot: model-building tools for molecular graphics. Acta Crystallogr. D Biol. Crystallogr., 60, 2126–2132 (2004).
• 28) Murshudov GN, Vagin AA, Dodson EJ. Refinement of macromolecular structures by the maximum-likelihood method. Acta Crystallogr. D Biol. Crystallogr., 53, 240–255 (1997).