1086 Diabetes Volume 67, June 2018

A Pharmacogenetic Approach to the Treatment of Patients With PPARG Mutations

Maura Agostini,1,2 Erik Schoenmakers,1 Junaid Beig,3 Louise Fairall,4 Istvan Szatmari,5 Odelia Rajanayagam,1,2 Frederick W. Muskett,4 Claire Adams,1,2 A. David Marais,6 Stephen O’Rahilly,1,2 Robert K. Semple,1,2 Laszlo Nagy,5 Amit R. Majithia,7 John W.R. Schwabe,4 Dirk J. Blom,8 Rinki Murphy,3,9 Krishna Chatterjee,1,2 and David B. Savage1,2

Diabetes 2018;67:1086–1092 | https://doi.org/10.2337/db17-1236

Loss-of-function mutations in PPARG cause familial partial indicate that detailed structural and functional classifica- lipodystrophy type 3 (FPLD3) and severe metabolic disease tion can be used to inform therapeutic decisions in patients in many patients. Missense mutations in PPARG are present with PPARG mutations. in ∼1 in 500 people. Although mutations are often binarily classified as benign or deleterious, prospective functional classification of all missense PPARG variants suggests Peroxisome proliferator–activated receptor-g (PPARg)is that their impact is graded. Furthermore, in testing novel a nuclear receptor originally identified in adipocytes (1). mutations with both prototypic endogenous (e.g., prosta- Although widely expressed, cell-based loss-of-function stud- glandin J2 [PGJ2]) and synthetic ligands (thiazolidinediones, ies attest to its primary role in regulating adipogenesis and tyrosine agonists), we observed that synthetic agonists adipocyte function, with rodent knockout studies robustly selectively rescue function of some peroxisome proliferator– corroborating these data (2,3). Heterozygous, dominant- PPARG METABOLISM activated receptor-g (PPARg) mutants. We report on negative, loss-of-function mutations in human were patients with FPLD3 who harbor two such PPARg mutat- first described in 1999 (4), with subsequent identification ions (R308P and A261E). Both PPARg mutants exhibit of many more receptor defects (5–9). Clinical findings in negligible constitutive or PGJ2-induced transcriptional ac- such patients have refined the phenotype (now known as tivity but respond readily to synthetic agonists in vitro, with familial partial lipodystrophy type 3 [FPLD3]) characterized structural modeling providing a basis for such differential by a paucity of limb fat; preserved abdominal fat; insulin ligand-dependent responsiveness. Concordant with this resistant diabetes; dyslipidemia with particularly labile, finding, dramatic clinical improvement was seen after diet-sensitive hypertriglyceridemia; polycystic ovarian syn- pioglitazone treatment of a patient with R308P mutant PPARg. drome; and hypertension. A patient with A261E mutant PPARg also responded ben- Like many nuclear receptors, PPARg has an amino- eficially to rosiglitazone, although cardiomyopathy pre- terminal activation domain (AF1), a central DNA-binding cluded prolonged thiazolidinedione use. These observations domain, and a carboxy-terminal ligand-binding domain.

1The University of Cambridge Metabolic Research Laboratories, Wellcome Trust- Corresponding author: David B. Savage, [email protected]; Krishna MRC Institute of Metabolic Science, Cambridge, U.K. Chatterjee, [email protected]; Rinki Murphy, r.murphy@auckland 2The National Institute for Health Research Cambridge Biomedical Research Centre, .ac.nz; Dirk J. Blom, [email protected]; or John W.R. Schwabe, js336@leicester Cambridge, U.K. .ac.uk. 3 Greenlane Diabetes Centre, Auckland Hospital, Auckland, New Zealand Received 12 October 2017 and accepted 27 March 2018. 4Leicester Institute of Structural and Chemical Biology, Department of Molecular This article contains Supplementary Data online at http://diabetes and Cell Biology, University of Leicester, Leicester, U.K. .diabetesjournals.org/lookup/suppl/doi:10.2337/db17-1236/-/DC1. 5Department of Biochemistry and Molecular Biology, Faculty of Medicine, Uni- versity of Debrecen, Debrecen, Hungary M.A. and E.S. contributed equally to this work. 6Division of Chemical Pathology, Department of Pathology, University of Cape J.W.R.S., D.J.B., R.M., K.C., and D.B.S. are joint senior authors. Town and National Health Laboratory Service, Cape Town, South Africa © 2018 by the American Diabetes Association. Readers may use this article as 7Program in Medical and Population Genetics, Broad Institute, Cambridge, MA long as the work is properly cited, the use is educational and not for profit, and the 8Department of Medicine, University of Cape Town Health Science Faculty, Cape work is not altered. More information is available at http://www.diabetesjournals Town, South Africa .org/content/license. 9Department of Medicine, Faculty of Medical and Health Sciences, University of Auckland, Auckland, New Zealand diabetes.diabetesjournals.org Agostini and Associates 1087

PPARg heterodimerizes with retinoid X receptor a,and Assessing Transcriptional Activity of PPARg Mutants transcriptional activation is triggered by ligand binding, Characterization of transcriptional activity of PPARg var- resulting in the release of a corepressor complex and re- iants was undertaken as described previously (16). In brief, cruitment of a coactivator complex. Fatty acids and eicosa- HEK293 Epstein-Barr nuclear antigen cells cultured in noids can activate PPARg, with prostaglandin J2 (PGJ2) DMEM and 10% FCS were transfected with Lipofectamine considered prototypic of such putative endogenous PPARg 2000 in 96-well plates and assayed for luciferase and b- ligands (10,11). Structural studies have suggested that the galactosidase activity after a 36-h incubation with or without ligand-binding pocket of PPARg is promiscuous and can ac- ligand. Results represented the mean 6 SEM of at least three commodate several different fatty acids (3). Thiazolidinediones independent experiments in triplicate. (TZDs), a class of synthetic PPARg agonists, promote adipo- genesis and improve insulin sensitivity, underpinning their Structural Modeling of PPARg Mutants therapeutic use as insulin sensitizers in patients with type 2 Crystallographic modeling of PPARg mutants was under- diabetes (12). taken by using PPARg structures (1PRG, 2PRG, 1FM9, 2ZK1, That as many as 1 in 500 people may have missense 3DZY, 2XKW) with different ligands. Results were illus- mutations in PPARG prompted Majithia et al. (7) to generate trated by using the molecular graphics system MacPyMOL and functionally characterize all possible missense PPARG (Schrödinger, New York, NY). Additional methodological mutations to expedite clinical interpretation of the growing details are available in the Supplementary Data. number of missense variants identified in patients. This resource should aid prompt functional classification of RESULTS PPARG novel variants. For individuals with established loss- Identification of PPARG Mutations of-function mutations and a disease phenotype, therapeutic Two different heterozygous missense mutations in the possibilities are limited. Current options include strict dietary ligand-binding pocket of PPARg were identified in patients fatandcalorierestriction,metformin,insulin,andglucagon- who presented with typical features of FPLD3. An R308P like peptide 1 agonists. Leptin has been tried in patients with mutation was detected in a New Zealand woman, and an very low leptin levels (13). Isolated reports of TZD use also A261E mutation was identified in two unrelated women exist (9,14,15), but responses were variable (summarized in from South Africa (see Supplementary Data for clinical Supplementary Table 1). details and Table 1 for biochemical results). In characterizing the properties of all possible PPARG missense mutations (7), we were struck by two observations. Functional Studies of PPARg Mutants First, the spectrum of functional scores exhibited by the Both R308 and A261 in PPARg were highly conserved (Fig. range of all missense PPARG variants suggested that even 1A). In transfection assays that used reporter constructs con- mutations associated with a monogenic disease are likely to taining either synthetic [(PPARE)3TKLUC] or natural (human perturb protein function to a variable degree, predisposing FABP4-LUC) enhancer/promoter elements, both R308P and to a similarly variable phenotype rather than fitting an arbi- A261E mutants exhibited negligible basal transcriptional trary designation as disease-causing or benign. Such gradation activity and minimal responsiveness with PGJ2 (Fig. 1B and C). of PPARg dysfunction is also likely to translate into differ- However, moderate (100 nmol/L farglitazar) or higher con- ential, graded responses to metabolic stress and to molec- centrations (1 mmol/L rosiglitazone; 10 mmol/L pioglitazone) ularly targeted therapeutic interventions. Second, we noted of synthetic agonists restored transcriptional activity com- that a few variants, like R308P, manifested a clearly abnor- parable to wild-type receptor (Fig. 1B and C). Of note, the mal transcriptional response to prototypic natural ligand (e.g., R308P variant returned an intermediate, nondiagnostic PGJ2), whereas their function,whentestedwithasynthetic functional score when tested in a high-throughput cellular agonist, was near normal (7). These in vitro observations assay (7) (Table 1), reflecting a similar discordance between suggest that patients harboring such receptor mutants might failure to respond to PGJ2 and activation with rosiglitazone. respond to treatment with synthetic PPARg agonists. Thus, These results suggest that the R308P mutant is transcrip- we report the dramatic clinical response of a patient who tionally resistant to both natural ligands present endogenously harbors the Arg308Pro (R308P) PPARg variant after treat- within transfected cells and PGJ2, with such loss-of-function ment with rosiglitazone. We also describe a novel Ala261Glu likely contributing to the patient’s lipodystrophic phenotype. (A261E) PPARg mutation, present in two apparently unre- lated families, with similarly discordant responses to PGJ2 Structural Modeling versus synthetic PPARg agonists. In the crystal structure of the PPARg ligand-binding domain, A261 and R308 are situated on different sides of the ligand- binding pocket (Fig. 2A). R308, located close to the amino RESEARCH DESIGN AND METHODS terminus of helix 3 (Fig. 2A and B), participates in an extensive Participants provided informed written consent, and inves- hydrogen-bond network (Fig. 2D–F) involving E287 in helix tigations were approved by local (Cambridge, U.K., and Cape 2 and residues in the loop between helix 2 and 3. In the Town, Africa) research ethics committees and conducted in unliganded receptor, R308 also makes hydrogen bonds within accordance with the Declaration of Helsinki. helix 3 (Fig. 2D). Upon ligand binding, the loop between helix 1088 Pharmacogenetically Informed Therapy in FPLD3 Diabetes Volume 67, June 2018

Table 1—Biochemical findings in probands with FPLD3 PPARg mutation Characteristic A308P proband 1 A261E proband 2 A261E proband 3 Normal range Sex Female Female Female Age at time of assessment (years) 16 22 39 Age at first presentation (years) 16 20 30 Height (m) 1.46 1.53 1.45 Weight (kg) 48.0 61.0 55.0 BMI (kg/m2)232626 Total body fat (%) 20 NA NA Predicted body fat (%)** 27 NA NA Truncal fat (%) 22 NA NA Leg fat (%) 18 NA NA Hypertension No No Yes T2DM or IGT¶ Yes Yes Yes PCOS§ Yes Yes Yes NAFLD# Yes NA NA Triglyceride (mmol/L) 13.0 16.6 11.3 ,1.7 HDL cholesterol (mmol/L) 0.4 0.5 0.5 .1.0 Total cholesterol (mmol/L) 4.7 8.2 5.1 ,5.1 Insulin (pmol/L) 405 1253 NA ,60 Glucose (mU/L) 22.4 6.4 12.3 ,6.1

HbA1c (mmol/mol) 61 NA 78 20–40 ALT (units/L) NA 20 9 ,30 GGT (units/L) NA 32 18 ,35 Familial cosegregation Unaffected mother and Affected male sibling with Affected male sibling with sibling are mutation negative the mutation the mutation Functional score*** 20.932 23.798 23.798 Fat mass and distribution was assessed with DXA performed with GE Lunar iDXA version 15. ALT, alanine aminotransferase; GGT, g-glutamyl transferase; IGT, impaired glucose tolerance; NA, not available; NAFLD, nonalcoholic fatty liver disease; PCOS, polycystic ovary syndrome; T2DM, type 2 diabetes mellitus. **Predicted body fat = (1.48 3 BMI) 2 7. ¶Yes or no indicates the presence or absence of either condition. §Yes or no indicates the presence or absence of this syndrome. #Yes indicates NAFLD as confirmed by ultrasound and magnetic resonance spectroscopy. ***Functional score as derived from http://miter.broadinstitute.org.

2 and 3 adopts varying conformations, depending on the denaturation temperature compared with the wild-type nature of the ligand (Supplementary Fig. 1). Mutation R308P receptor (Supplementary Table 1). Because PGJ2 docks in would completely disrupt both the intra- and interhelical this part of the ligand-binding cavity, its binding to receptor hydrogen-bond networks (Fig. 2G). Although PGJ2 does is expected to be impaired (Fig. 2C). In contrast, receptor not alter the structural architecture of this region (Fig. 2H), occupancy by rosiglitazone, farglitazar, and pioglitazone is binding of farglitazar, rosiglitazone, and pioglitazone can not structurally dependent on this region, correlating with potentially alter the conformation of the loop between helix preservation of transcriptional activation of the A261E mu- 2 and 3, thereby providing a mechanism that counteracts the tant (Fig. 2B and Supplementary Fig. 1). destabilizing effect of the R308P mutation and preserving transcriptional responsiveness to these synthetic ligands Responses to TZD Therapy (Fig. 2F and I and Supplementary Fig. 1B–D). In keeping with The R308P proband had previously been treated with dietary this prediction, proton nuclear magnetic resonance spectral advice and metformin, but her metabolic control remained analysis confirmed that pioglitazone can bind effectively to suboptimal, so pioglitazone 30 mg/day was commenced, the R308P mutant (Supplementary Fig. 2). resulting in dramatic improvements in glycemic control A261 is located in helix 2a (Fig. 2A–C), and the size and and dyslipidemia (Table 2). Her hirsutism, hyperandrogen- charge difference of the A261E mutation will cause displace- ism, and acanthosis nigricans also improved. These changes ment of helix 2a and the loop to helix 2b, thereby desta- were largely sustained over a 3-year period without a sub- bilizing the ligand-binding pocket. This was confirmed by stantial change in BMI (23.7–22.0 kg/m2 at 12 months and using circular dichroism studies showing a lower thermal 23.0 kg/m2 at 24 months). diabetes.diabetesjournals.org Agostini and Associates 1089

Figure 1—A: Schematic representation of the three major domains of PPARg, showing the locations of the two mutations and the conservation of the mutated residues between species (A261, R308—PPARg2 nomenclature). B: Transcriptional responses of empty vector (pcDNA), R308P, or A261E mutant PPARg2 to PGJ2 and rosiglitazone, pioglitazone, and farglitazar (doses in nmol/L on x-axis) when tested with a (PPARE)3TKLUC reporter construct and Bos-b-gal internal control plasmid. Results are expressed as a percentage of the maximum activation achieved with wild- type (WT) PPARg2 and represent the mean 6 SEM of at least three independent experiments in triplicate. C: Transcriptional responses of empty vector, R308P, or A261E mutant PPARg2 to PGJ2 and rosiglitazone, pioglitazone, and farglitazar (doses in nmol/L on x-axis) when tested with a human FABP4-TKLUC (hFABP4-TKLUC) promoter construct and Bos-b-gal internal control plasmid. Results are expressed as a percentage of the maximum activation achieved with WT PPARg2 and represent the mean 6 SEM of at least three independent experiments in triplicate.

One of the patients with A261E was treated twice with are benign or mild in their impact, others are pathogenic but rosiglitazone (4 mg twice daily) when her glycemic and triglyc- likely in a graded rather than a binary categorical fashion. eride control deteriorated significantly.Oneachoccasion, Because synthetic PPARg ligands are licensed treatments this intervention was accompanied by substantial falls in her and given the severity of the metabolic complications seen in HbA1c as well as improvements in fasting triglyceride levels, patients with FPLD3, use of TZDs is the obvious therapeutic although these remained labile (Supplementary Fig. 3). How- option. Theoretically, patients with FPLD3 could be 1)re- ever, therapy was discontinued because it exacerbated the sistant to TZDs because of the extreme deleterious nature of patient’s severe congestive heart failure, which ultimately the underlying PPARg defect; 2) responsive to therapy with caused her death at age 26 years. mutations that are unresponsive to low-affinity, endogenous ligands yet activated by higher-affinity synthetic agonists; or DISCUSSION 3) potentially hyperresponsive to specific designer ligands The remarkable increase in access to and use of next- that can overcome the molecular defect particular to a spe- generation sequencing has accelerated the discovery of novel cific receptor mutation. Mendelian disorders and detection of mutations in genes We report on two FPLD3-associated PPARg mutations known to cause monogenic disorders like FPLD3, where ;1 (A261E, R308P) whose properties fall into the latter cate- in 500 people harbor missense mutations (7). Although most gories (2 and 3 above). Despite its transcriptional efficacy 1090 Pharmacogenetically Informed Therapy in FPLD3 Diabetes Volume 67, June 2018

Figure 2—Crystallographic modeling on the basis of structures of unliganded PPARg (1PRG) or bound to PGJ2 (2ZK1) or pioglitazone (2XKW). One mutated residue (A261) is in proximity to PGJ2 (A), whereas the other amino acid (R308) is in the vicinity of pioglitazone (B). Substitution of glutamic acid for alanine at residue 261 (A261E) can interfere with PGJ2 binding through steric hindrance (C). The side chain of arginine 308 (R308) participates in a network of intrahelical (H3) and interhelical (e.g., E287 in H2) hydrogen bonds in unliganded (D) and liganded (E and F) PPARg. Mutation of this residue to proline likely disrupts this hydrogen-bond network (G). PGJ2, which binds elsewhere in the ligand-binding cavity, is unable to prevent loss of such interactions (H), whereas pioglitazone, which binds in the vicinity, forms hydrogen bonds with E287 and could preserve receptor conformation (I). H, helix; PIO, pioglitazone. with wild-type PPARg, PGJ2, a ligand that is prototypic mediated by A261E or R308P mutants (Fig. 1B), whereas ex- of the various endogenous fatty acid and eicosanoid posure to high-affinity synthetic ligands like rosiglitazone, PPARg activators, was unable to fully activate transcription pioglitazone, and farglitazar achieved full transcriptional diabetes.diabetesjournals.org Agostini and Associates 1091

Table 2—Comparison of investigations before and after pioglitazone treatment Investigation Before treatment At 12-month treatment At 24-month treatment Reference range Weight (kg) diabetes profile 49.0 48.6 50.5

HBA1c (mmol/mol) 61 42 31 20–40 Glucose (mmol/L) 12.0 4.0 3.7 ,6.1 Insulin (pmol/L) 405 ND ND 10–60 Liver enzymes ALT (IU/L) 64 35 33 ,30 GGT (IU/L) 34 16 15 ,35 Hormonal profile Free testosterone (pmol/L) 198 84 ND ,50 SHBG (nmol/L) 16 14 ND 20–90 Free androgen index 450 214 ND ,80 FSH (IU/L) 8.8 4.5 7.1 3–25 LH (IU/L) 14.8 4.3 5.1 2.0–25 Lipid profile TG (mmol/L) 13.2 2.4 1.6 ,1.7 HDL (mmol/L) 0.4 0.6 0.7 .1.0 Total cholesterol (mmol/L) 4.7 3 3.8 ,5.1 LDL (mmol/L) 2 1.3 2.4 ,3.4 ALT, alanine aminotransferase; FSH, follicle-stimulating hormone; GGT, g-glutamyl transferase; LH, luteinizing hormone; ND, not done; SHBG, sex hormone–binding globulin; TG, triglyceride. activity. Crystal structures of PPARg bound to either preference to other standard glucose-lowering therapies, farglitazar, rosiglitazone, pioglitazone, or PGJ2 show differ- resulting in substantial clinical improvements in all met- ences between these ligands in the nature of their occupancy abolic abnormalities paralleled by specific redistribution of the binding cavity (Supplementary Fig. 1), and structural of body fat away from visceral and with expansion of modeling provides a plausible basis for differential mutant subcutaneous depots. Moreover, structural modeling and PPARg responsestoprototypicendogenousversussynthetic verification with studies of mutant receptor function ligands. Although synthetic agonists do not occupy the region in vitro provide a plausible explanation for in vivo ob- of the pocket where A261 is situated, this residue is in prox- servations. Specifically, impaired receptor activation by imity to PGJ2 and other fatty acid ligands (16). Modeling of endogenous ligands presumably mediates diminished the A261E mutation suggests that the alanine to glutam- adipogenesis and the FPLD3 phenotype; the subsequent ic acid change is likely to perturb PGJ2 binding directly profound therapeutic response to pioglitazone likely reflects through steric hindrance, whereas receptor interaction the ability of this synthetic agonist to bypass or overcome with rosiglitazone, pioglitazone, and farglitazar would be the molecular consequences of this mutation. In patients preserved, correlating with the observed transcriptional harboring A261E mutant PPARg,wehavedocumented responses (Fig. 1B and C). similar discordant transcriptional responses to prototypic The R308P mutation involves a different part of the endogenous ligand versus synthetic agonists. We have ligand-binding cavity, and this residue does not make direct shown that this translates into a beneficial therapeutic contact with ligands. Structural modeling suggests that the response to rosiglitazone in one patient with this receptor arginine to proline change would disrupt local hydrogen-bond defect. networks, with deleterious conformational consequences af- Although our observations are based on prismatic case fecting transcriptional function of the receptor. Rosiglitazone, studies, they appear to be supported by structural analyses of pioglitazone, and farglitazar (but not PGJ2), bind in proximity the isolated reports of TZD use in patients with FPLD3 to R308, possibly stabilizing receptor structure. In particular, (Supplementary Table 1). Such concordance among struc- pioglitazone, which we have shown to bind effectively to the tural modeling of PPARg mutations, transcriptional responses R308P mutant receptor, makes a hydrogen bond with E287 in of mutant PPARg to ligands in vitro, and clinical responses helix 2, as does R308, counteracting the effect of the R308P to treatment with synthetic agonists in vivo highlight the mutation, which is predicted to disrupt this interaction (Fig. potential for this approach to inform individualized ther- 2F–I). apeutic choices. In vitro studies with R308P mutant PPARg mirrored the patient’s dramatic and sustained response to pioglitazone therapy. Thus, her case highlights the importance of recog- Acknowledgments. The authors are grateful to the participating patients. nizing and then establishing the genetic basis for severe, Funding. J.W.R.S., K.C., and D.B.S. are supported by the Wellcome Trust (grants early-onset, metabolic disease. Identification of a PPARG WT100237, WT095564, and WT107064, respectively), the MRC Metabolic Disease mutation enabled early treatment with pioglitazone in Unit, the National Institute for Health Research (NIHR) Cambridge Biomedical Research 1092 Pharmacogenetically Informed Therapy in FPLD3 Diabetes Volume 67, June 2018

Centre, and the NIHR Rare Disease Translational Research Collaboration. J.W.R.S. holds 6. Agarwal AK, Garg A. Genetic basis of lipodystrophies and management of a Royal Society Wolfson Research Merit Award. metabolic complications. Annu Rev Med 2006;57:297–311 Duality of Interest. No potential conflicts of interest relevant to this article were 7. Majithia AR, Tsuda B, Agostini M, et al.; UK Monogenic Diabetes Consortium; reported. Myocardial Infarction Genetics Consortium; UK Congenital Lipodystrophy Consortium. Author Contributions. M.A. and E.S. undertook functional studies of the Prospective functional classification of all possible missense variants in PPARG. Nat mutants, analyzed data, and wrote parts of the manuscript. E.S., L.F., F.W.M., and Genet 2016;48:1570–1575 J.W.R.S. undertook structural modeling of mutations, circular dichroism, and nuclear 8. Robbins AL, Savage DB. The genetics of lipid storage and human lip- magnetic resonance studies and reviewed the manuscript. J.B. and R.M. charac- odystrophies. Trends Mol Med 2015;21:433–438 terized and treated proband 1 and wrote parts of the manuscript. I.S., S.O., R.K.S., 9. Savage DB, Tan GD, Acerini CL, et al. Human metabolic syndrome resulting from L.N., and A.R.M. analyzed data and reviewed the manuscript. O.R. performed se- dominant-negative mutations in the nuclear receptor peroxisome proliferator-activated quencing studies, identified the PPARG mutations, and reviewed the manuscript. receptor-g. Diabetes 2003;52:910–917 C.A. coordinated studies and reviewed the manuscript. A.D.M. and D.J.B. 10. Forman BM, Tontonoz P, Chen J, Brun RP, Spiegelman BM, Evans RM. 15- characterized probands 2 and 3, analyzed data, and wrote parts of the manuscript. Deoxy-delta 12, 14-prostaglandin J2 is a ligand for the adipocyte determination factor R.M., K.C., and D.B.S. planned studies, analyzed data, and wrote the manuscript. PPAR g. Cell 1995;83:803–812 D.B.S. is the guarantor of this work and, as such, had full access to all the data in the 11. Kliewer SA, Lenhard JM, Willson TM, Patel I, Morris DC, Lehmann JM. A study and takes responsibility for the integrity of the data and the accuracy of prostaglandin J2 metabolite binds peroxisome proliferator-activated receptor g and the data analysis. promotes adipocyte differentiation. Cell 1995;83:813–819 12. Soccio RE, Chen ER, Lazar MA. Thiazolidinediones and the promise of insulin sensitization in type 2 diabetes. Cell Metab 2014;20:573–591 References 13. Guettier JM, Park JY, Cochran EK, et al. Leptin therapy for partial lipodystrophy 1. Spiegelman BM. PPAR-gamma: adipogenic regulator and thiazolidinedione linked to a PPAR-g mutation. Clin Endocrinol (Oxf) 2008;68:547–554 – receptor. Diabetes 1998;47:507 514 14. Francis GA, Li G, Casey R, et al. Peroxisomal proliferator activated receptor- – 2. Lazar MA. PPAR gamma, 10 years later. Biochimie 2005;87:9 13 gamma deficiency in a Canadian kindred with familial partial lipodystrophy type 3 3. Hallenborg P, Petersen RK, Kouskoumvekaki I, Newman JW, Madsen L, (FPLD3). BMC Med Genet 2006;7:3 Kristiansen K. The elusive endogenous adipogenic PPARg agonists: lining up the 15. Demir T, Onay H, Savage DB, et al. Familial partial lipodystrophy linked to a novel suspects. Prog Lipid Res 2016;61:149–162 peroxisome proliferator activator receptor -g (PPARG) mutation, H449L: a comparison 4. Barroso I, Gurnell M, Crowley VEF, et al. Dominant negative mutations in human of people with this mutation and those with classic codon 482 Lamin A/C (LMNA) PPARgamma associated with severe insulin resistance, diabetes mellitus and hy- mutations. Diabet Med 2016;33:1445–1450 pertension. Nature 1999;402:880–883 16. Agostini M, Schoenmakers E, Mitchell C, et al. Non-DNA binding, dominant- 5. Semple RK, Chatterjee VKK, O’Rahilly S. PPAR gamma and human metabolic negative, human PPARgamma mutations cause lipodystrophic insulin resistance. Cell disease. J Clin Invest 2006;116:581–589 Metab 2006;4:303–311