International Journal of Obesity (2003) 27, 70–74 ß 2003 Nature Publishing Group All rights reserved 0307–0565/03 $25.00 www.nature.com/ijo PAPER -3-receptor variants in morbid obesity

C Schalin-Ja¨ntti1,2*, K Valli-Jaakola1, L Oksanen1, E Martelin3, K Laitinen1, T Krusius4, P Mustajoki5, M Heikinheimo3,6 and K Kontula1

1Department of Medicine, University of Helsinki, Helsinki, Finland; 2Department of Endocrinology, Helsinki University Hospital, Helsinki, Finland; 3Department of Pediatrics, University of Helsinki, Helsinki, Finland; 4Finnish Red Cross Blood Transfusion Service, Helsinki, Finland; 5Peijas Hospital, Vantaa, Finland; and 6Department of Pediatrics, Washington University in St. Louis, MO, USA

BACKGROUND: Linkage and knock-out mice studies suggest that the melanocortin-3-receptor (MC3R) is a candidate gene for obesity. OBJECTIVE: To evaluate whether MC3R mutations underlie morbid obesity. SUBJECTS AND METHODS: MC3R coding and 50-flanking regions were sequenced in 48 subjects and the detected variants genotyped in 252 morbidly obese (BMI  40 kg/m2) Finns. Gel shifts were used to examine whether a mutation in the putative promoter alters GATA-factor binding. RESULTS: Three common MC3R variants were found: a 17C > A variant, changing Thr6?Lys in 16%, a 241G > A variant changing Val81?Ile in 15%, and a 7239A > G substitution in the GATA binding site in 21% of the subjects. Four other variants were detected in the 50 flanking region. Frequencies of the three common variants did not differ between obese and contol subjects. Among the obese, the 17C > A and 241G > A variants were coinherited and associated with increased – glucose ratios (P < 0.05) and leptin levels (P < 0.05). GATA-4 bound efficiently to wild type oligonucleotide, but only weakly to the oligonucleotide with the 7239A > G mutation. CONCLUSIONS: MC3R gene variants are common and do not explain human morbid obesity. These variants associated with subtle changes in onset of weight gain, hyperleptinemia and insulin – glucose ratios. The 7239A > G mutation abolishes binding of GATA-4 to the MC3R promoter region. International Journal of Obesity (2003) 27, 70 – 74. doi:10.1038=si.ijo.0802184

Keywords: melanocortin-3-receptor; MC3R gene variants; GATA-4; EMSA; ; insulin resistance; insulin – glucose ratio; obesity; energy balance; brain

Introduction hyperphagia and obesity in mice.4 MC4R mutations underlie The proportion of body fat is reflected by circulating leptin up to 4% of severe early-onset or adult obesity.5,6 Observed and insulin levels, and long-term homeostasis of body linkage between obesity and the chromosomal region 20q13 weight is accomplished by integration of these hormonal also makes the MC3R gene a plausible candidate gene for signals by hypothalamic centers.1–3 derived human obesity.7 Furthermore, disruption of MC3R in mice from the ACTH precursor pro-opiomelanocortin (POMC) results in increased fat mass, hyperleptinemia and insulin and melanocortin receptors (MCRs) play a key role in this resistance.8,9 We examined whether mutations in the MC3R complex control of appetite and body weight.1,2 MC3R and gene underlie morbid obesity and/or are associated with MC4R are both highly expressed in brain areas involved in changes in metabolic parameters. regulation of energy balance.2 The MC4R plays an important role in energy homeostasis, as its targeted disruption causes

*Correspondence: C Schalin-Ja¨ntti, Departments of Medicine and Subjects and methods Endocrinology, University of Helsinki, BOX 340, Haartmaninkatu 4, A cohort of 252 morbidly obese (BMI  40 kg/m2, 182 FIN-00290 Helsinki, Finland. females/70 males, age 21 – 67 y, prevalence of diabetes and E-mail: camilla.schalin-jantti@hus.fi Received 6 February 2002; revised 10 June 2002; hypertension 24 and 45%, respectively) patients was col- accepted 22 July 2002 lected between 1989 – 1995.10 History of weight development MC3R gene in morbid obesity C Schalin-Ja¨ntti et al 71 was assessed by questionnaire. For mutation detection, Results 1400 bp of the 50-flanking region and the whole coding MC3R gene variants region of the MC3R gene were sequenced in 48 subjects and Two common missense mutations were found in the MC3R the variants detected genotyped in the rest of the cohort. coding region: nucleotide 17C > A changing amino acid Blood samples from 312 healthy blood donors (153 females 6 from threonine to lysine and 241G > A changing and 159 males) from the Finnish Red Cross Blood Transfusion amino acid 81 from valine to isoleucine. Five additional Service served as controls for estimation of allele frequencies variants, 7939G > C, 7911G > A, 7803T > C, 7373G > T in the background population. Leptin concentrations were and 7239A > G were detected in the putative promoter determined by radioimmunoassay (Linco Research Inc., region. The 7239A > G variant was localized in a consensus St Charles, MO, USA) with intra- and interassay CV’s of less GATA transcription factor binding site13 (aaacaaGA- than 5%. Glucose, insulin and lipid levels were measured as TAaaaact), comprising nucleotides 7244 to 7228 of the previously described.10 MC3R gene (common allelic form in bold and underlined). The genotype frequencies for the 17C > A, 241G > A and 7239A > G variants did not differ between morbidly obese DNA sequencing and genotyping and the control population (Table 1A). After PCR, the MC3R gene was sequenced using primers pairs: 50-TGACCAGAGCAGACTACTTTCA-30 and 50-TGAGCAAA- 0 0 0 7 > GACAACAGCCACT-3 ;5-TCTTCTTCCACCCCAGACTC-3 The 239A G variant and GATA binding and 50-GCAAGACAGGGGATGTGTTA-30;50-CCCTTGGCCAA- In EMSAs, nuclear extracts abundantly containing GATA-4/ TATGAAAAA-30 and 50-CCAGATACGTCTTTTGGATGC-30; GATA-6 yielded a strong retarded band A with the wild type 50-TCTCTACCCTCCCCATCCTT-30 and 50-GGGCATTGGACA- probe (Figure 1, lane 1), whereas assays using the MC3Rmut CACTTACC-30;50-CATCGTCAGTCTGCTGGAAA-30 and oligonucleotide revealed only a weak – DNA com- 50-GAGCATCATGGCGAAGAAC-30; and 50-CTCGGAGAG- plex (Figure 1, lanes 9 and 10). In an excess of unlabelled CAAAATGGTCA-30 and 50-TCACGTGGATGGAAAGTCAA-30. Sequencing was carried out using the ABI Prism 377 DNA Table 1A MC3R genotypes in morbidly obese and control subjects sequencer. Genotyping of the 7239A > G, 17C > A and 241G > A variants were performed using restriction Variant Morbidly obese n (%) Control subjects n (%) enzymes AlwI, HpyCH4IV and BseDI, respectively, and gel 7239A > G electrophoresis. AA 198 (79) 238 (76) AG 51 (20) 67 (21) GG 3 (1) 7 (2) Total 252 (100) 312 (100)* Electrophoretic mobility shift assay (EMSA) Nuclear were prepared11 from a mouse Sertoli tumor 17C > A (Thr6?Lys) CC 204 (84) 217 (80) cell MSC-1, Jurkat T-cell lymphoma and NIH 3T3 fibroblast CA 37 (15) 46 (17) lines, expressing GATA-4 and GATA-6 (MSC-1), GATA-3 AA 3 (1) 8 (3) (Jurkat) and no known GATA proteins (NIH-3T3). Probes Total 244 (100) 271 (100) were prepared by annealing oligonucleotides 50-TTTCTATG- 241G > A (Val81?Ile) TAAACAAGATAAAAACTGCTCCTCCT-30 (corresponding to GG 206 (85) 186 (83) 0 nucleotides 7253 to 7220) and 5 -AGGAGGAGCAGTTTT- GA 35 (14) 35 (16) TATCTTGTTTACATAGAAA-30 (MC3R), 50-TTTCTATGTAAA- AA 3 (1) 4 (2) 0 0 CAGGATAAAAACTGCTCCTCCT-3 and 5 -AGGAGGAGCA- Total 244 (100) 225 (100)* GTTTTTATCCTGTTTACATAGAAA-30 (MCR3mut, 7239A > G mutation underlined), and 50-CCCATAAAGATAGGGA-30 and 50-TCCCTATCTTTATGGG-30 (corresponding to nucleo- tides 7182 to 7163 of the steroidogenic factor-1 (SF-1) Table 1B Relationship between coding region MC3R gene variants and gene). One of the oligonucleotides of each probe was metabolic parameters in morbidly obese subjects 50-end-labelled with [g-32P]ATP before annealing. Binding 17C 241G reactions were carried out as described.13 The antibodies were from Santa Cruz Biotechnology Inc. (Santa Cruz, CA). Variant CC AC/AA GG AG/AA

n 204 40 206 38 Insulin – glucose ratio 3.1Æ 0.2 4.1Æ 0.4* 3.1Æ 0.2 4.0Æ 0.5** Statistical analyses Leptin (ng/ml) 38.5Æ 1.5 46.4Æ 3.6* 38.5Æ 1.5 47.0Æ 3.7** BMI (kg/m2) 42.7Æ 0.5 43.3Æ 1.1 42.7Æ 0.5 43.2Æ 0.5 Mann – Whitney rank-sum test was used to analyse differ- 2 BMI at 20 y (kg/m ) 26.9Æ 0.4 29.3Æ 1.4 26.9Æ 0.4 29.0Æ 1.4 ences between group means and Fisher’s exact test for frequency distributions. *P < 0.05 vs CC; **P < 0.05 vs GG. Values are meansÆ s.e.m.

International Journal of Obesity MC3R gene in morbid obesity C Schalin-Ja¨ntti et al 72

Figure 1 GATA-4 binds to wild type MC3R oligonucleotide (MC3R). Nuclear extracts (NE) from MSC-1, NIH-3T3, and Jurkat cells were incubated with 32P-labelled oligonucleotide probes. When MSC-1 extract (10 mg) and labelled MC3R were used, two major bands A and B were formed (lane 1). The binding of proteins was competed with unlabelled MC3R in 15- and 100-fold molar excess (lanes 2 and 3), or mutated MC3R (MC3Rmut, lane 4) or SF-1 (lane 5) oligonucleotides in 100-fold molar excess. In supershift experiments, antibodies against GATA-4 (lane 6), GATA-3 (lane 7), and GATA-6 (lane8) were used. Labelled MC3Rmut was used as a probe with either 10 mg (lane 9) or 20 mg (lane 10) MSC-1 nuclear extract. NIH-3T3 extract (10 mg, lane 11), without any known GATA proteins, and Jurkat cell extract (10 mg, lane 12 and 13), containing abundant GATA-3 protein, were incubated with labelled MC3R. Supershift experiment with GATA-3 antibodies is shown on lane 13.

MC3R or SF-1 oligonucleotide, band A was clearly attenu- MC3Rmut did not change the formation of band A ated (Figure 1, lanes 2, 3 and 5) indicating specific binding (Figure 1, lane 4). In supershift experiments, GATA-4 anti- and suggesting that, of the various GATA proteins, GATA-4 body abolished band A indicating that GATA-4 binds to the is responsible for formation of band A. Unlabelled MC3R probe, whereas neither GATA-3 nor GATA-6 caused

International Journal of Obesity MC3R gene in morbid obesity C Schalin-Ja¨ntti et al 73 any supershift (Figure 1, lanes 6, 7 and 8). The mutation Foundation for Cardiovascular Research (LO, KK), the 7239A > G did not affect the binding of protein(s) forming Maud Kuistila Foundation (LO), the Finnish Cultural Foun- band B (Figure 1, lanes 2 – 4). dation (LO), the Helsinki Biomedical Graduate School, Uni- versity of Helsinki (EM), the Sigrid Juselius Foundation (MH, KK) and the Research Funds from the University Central MC3R variants and metabolic parameters Hospital in Helsinki (CS-J, MH, KK) The 17C > A and 241G > A gene variants were coinherited in all but 3 cases, indicating an almost complete linkage dis- equilibrium. Individuals with the combined genotype of 17CC/241GG had lower insulin – glucose ratios and lower References leptin levels and tended to weigh less at 20 y than those with 1 Spiegelman BM, Flier JS. Obesity and the regulation of energy balance. Cell 2001; 104: 531 – 543. the combined variant genotypes (Table 1B). A carrier status 2 Wardlaw SL. Obesity as a neuroendocrine disease: lessons to be for the 7239A > G variant was associated with a tendency learned from and towards lower maximal BMI values when compared to wild mutations in mice and men. J Clin Endocrinol Metab 2001; 86: 1442 – 1446. type homozygotes (45.7Æ 0.7 vs 47.7Æ 0.5 kg/m2; P ¼ 0.06). 3 Considine RV, Sinha MK, Heiman ML, Kriauciunas A, Stephens TW, Nyce MR, Ohannesian JP, Marco CC, McKee LJ, Bauer TL, Caro JF. Serum immunoreactive-leptin concentrations in normal- Discussion weight and obese humans. New Engl J Med 1996; 334: 292 – 295. 4 Huszar D, Lynch CA, Fairchild-Huntress V, Dunmore JH, Fang Q, The GATA-binding proteins are zinc finger transcription Berkemeier LR, Gu W, Kesterson RA, Boston BA, Cone RD, factors regulating gene expression, differentiation and cell Smith FJ, Campfield LA, Burn P, Lee F. Targeted disruption of proliferation.14 We demonstrate that of the different GATA the mouse melanocortin-4 receptor results in obesity in mice. Cell proteins 3, 4 and 6, GATA-4 was responsible for binding to 1997; 88: 131 – 141. 5 Vaisse C, Clement K, Durand E, Hercberg S, Guy-Grand B, the MC3R promoter GATA site. GATA-4 is expressed in Froguel P. Melanocortin-4 receptor mutations are a frequent pituitary and hypothalamic cells, gonadal and adrenal and heterogeneous cause of morbid obesity. J Clin Invest 2000; cells15,16 and MC3R in the brain cortex, thalamus, hippo- 106: 253 – 362. campus, pituitary and hypothalamus.17 – 19 The expression 6 Farooqi IS, Yeo GSH, Keogh JM, Aminian S, Jebb SA, Butler G, Cheetham T, O’Rahilly S. Dominant and recessive inheritance of patterns for these two thus overlap, suggesting that morbid obesity associated with defi- GATA-4 is a regulator of the MC3R in vivo. The 7239A > G ciency. J Clin Invest 2000; 106: 271 – 279. variant was associated with a tendency towards lower max- 7 Lee JH, Reed DR, Li W-D, Xu W, Joo E-J, Kilker RL, Nanthakumar E, imal BMI (45.7Æ 0.7 vs 47.7Æ 0.5 kg/m2; P ¼ 0.06) and was North M, Sakul H, Callum B, Price RA. Genome scan for human obesity and linkage to markers in 20q13. Am J Hum Genet 1999; also an independent predictor of maximal BMI in the obese 64: 196 – 209. (stepwise forward regression analysis; R-squared increment ¼ 8 Chen AS, Marsh DJ, Trumbauer ME, Frazier EG, Guan X-M, Yu H, 0.019; P ¼ 0.03), indicating that it may exert a modulating Rosenblum CI, Vongs A, Feng Y, Linhai C, Metzger JM, Strack AM, effect on development of maximal weight. Camacho RE, Mellin TN, Nunes CN, Min W, Fisher J, Gopal- Truter S, MacIntyre DE, Chen HY, Van der Ploeg LHT. Inactivation Three detected MC3R gene variants were common in the of the mouse melanocortin-3 receptor results in increased fat obese as well as in control subjects and therefore do not mass and reduced lean body mass. Nature Genet 2000; 26: explain morbid obesity. The Thr6?Lys mutation is located 97 – 102. in the extracellular N-terminus and the Val81?Ile in the first 9 Butler AA, Kesterton RA, Khong K, Cullen J, Pelleymounter MA, Dekoning J, Baetscher M, Cone RD. A unique metabolic syn- transmembrane part of MC3R. Within the morbidly obese, drome causes obesity in the melanocortin-3 receptor-deficient the rare amino acid variants were associated with higher mouse. Endocrinology 2000; 9: 3518 – 3521. leptin levels and insulin – glucose ratios compared to subjects 10 Oksanen L, Mustajoki P, Kaprio J, Kainulainen K, Ja¨nne O, homozygous for the common variants. They also reported a Peltonen L, Kontula K. Polymorphism of the b3-adrenergic recep- tor gene in morbid obesity. Int J Obesity 1996; 20: 1055 – 1061. tendency towards higher BMI at 20 y. Hani et al found 11 Andrews NC, Douglas DV. A rapid micropreparation technique increased insulin – glucose ratios in healthy subjects with for extraction of DNA-binding proteins from limiting numbers of the variant MC3R alleles compared to healthy subjects mammalian cells. Nucleic Acids Res; 1991; 19: 2499. with the wild type alleles.20 Collectively, these data indicate 12 Martelin M, Palvimo JJ, Lapatto R, Raivio KO. Nuclear factor Y activates the human xanthine oxidoreductase gene promoter. that the MC3R gene variants may induce subtle changes in FEBS Lett 2000; 480:84– 88. obesity-related parameters and that these changes are not 13 Quandt K, Frech K, Karas H, Wingeder E, Werner T. MatInd and restricted to morbidly obese subjects. MatInspector — New fast and versatile tools for detection of consensus matches in nucleotide sequence data. Nucleic Acids Res 1995; 23: 4878 – 4884. 14 Orkin SH. GATA-binding transcription factors in hematopoietic Acknowledgements cells. Blood 1992; 80: 575 – 581. We thank Ms Tuula Soppela-Loponen, Ms Sirpa Stick and 15 Kiiveri S, Siltanen S, Rahman N, Bielinska M, Lehto V-P, Ms Merja Soininen for excellent technical assistance. This Huhtaniemi IT, Muglia LJ, Wilson DB, Heikinheimo M. Recipro- cal changes in the expression of transcription factors GATA-4 and study was supported by the Yrjo¨ Jahnsson Foundation (CS-J, GATA-6 accompany adrenocortical tumorigenesis in mice and KV-J), the Stockmann Foundation (CS-J), the Finnish humans. Mol Med 1999; 5: 490 – 501.

International Journal of Obesity MC3R gene in morbid obesity C Schalin-Ja¨ntti et al 74 16 Lawson MA, Mellon P. Expression of GATA-4 in migrating gona- 19 Lorsignol A, Vande Vijver VV, Ramaekers D, Vankelecom H, dotropin-releasing neurons in the developing mouse. Mol Cell Denef C. Detection of melanocortin-3 receptor mRNA in Endocrinol 1998; 140: 157 – 161. immature rat pituitary: functional relation to g3-MSH-induced 17 Arceci RJ, King AAJ, Simon MC, Orkin SH, Wilson DB. Mouse changes in intracellular Ca2þ concentration. J Endocrinol 1999; GATA-4: a retinoic acid-inducible GATA-binding transcription 11: 171 – 179. factor expressed in endodermal derivatives and heart. Mol Cell 20 Hani EH, Dupont S, Durand E, Dina C, Gallina S, Gantz I, Biol 1993; 13: 2235 – 2246. Froguel P. Natural occurring mutations in the melanocortin 18 Gantz I, Konda Y, Tashiro T, Shimoto Y, Miwa H, Munzert G, receptor 3 gene are not associated with type 2 diabetes mellitus Watson SJ, DelValle J, Yamada T. Molecular cloning of a novel in French Caucasians. J Clin Endocrinol Metab 2001; 86: melanocortin receptor. J Biol Chem 1993; 268: 8246 – 8250. 2895 – 2898.

International Journal of Obesity