Rare mutations in the proopiomelanocortin (POMC) gene cause severe early-onset childhood obesity. However, it is unknown whether common variants in POMC are responsible for variation in body weight or fat distribution within the commonly observed range in the population. We tested for association between three polymorphisms spanning the POMC gene and obesity phenotypes in 1,428 members of 248 families. There was significant association between genotypes at the C8246T (P < 0.0001) and C1032G (P = 0.003) polymorphisms and waist-to-hip ratio (WHR) corrected for age, sex, smoking, exercise, and alcohol consumption. Each T allele at C8246T (or G allele at C1032G) was associated with a 0.2-SD–higher WHR in a codominant fashion. When WHR was additionally corrected for BMI, thus providing a measure of body fat distribution throughout the range of BMI, there remained significant evidence for association with both markers that was of similar magnitude and statistical significance. There was no association between genotype at any polymorphism and BMI or plasma leptin level. These data show that genetic variants at the POMC locus influence body fat distribution within the normal range, suggesting a novel role for POMC in metabolic regulation.

Abdominal obesity is a major risk factor for type 2 diabetes, coronary heart disease (CHD), colon cancer, and several other common and serious conditions. Although the definition of obesity typically relies on BMI, several studies suggest that, with respect to risk stratification, measurement of the waist-to-hip ratio (WHR) may be a superior predictor both of type 2 diabetes and CHD, possibly because of specific metabolic abnormalities accompanying central, rather than generalized, obesity (1,2). Both WHR and BMI are heritable traits, and various studies suggest that genes account for 25–70% of the observed variability (3,4).

Melanocortin signaling in the hypothalamus plays a central role in the control of energy homeostasis. The proopiomelanocortin (POMC) gene is expressed in response to leptin signaling by neurons of the hypothalamic arcuate nucleus. Intracellular posttranslational processing of the POMC propeptide by prohormone convertase 2 leads to the production in these neurons of α-, β-, and γ-melanocyte–stimulating hormones. These peptides signal to downstream target neurons in the lateral hypothalamus that express the melanocortin receptors MC3R and MC4R with resultant decrease in food intake and increase in energy expenditure (5). Rare mutations in the POMC gene cause monogenic, severe, early-onset obesity in humans; however, the influence of common polymorphisms in POMC on obesity phenotypes in less extreme individuals is unclear (6). POMC is also a precursor of adrenocorticotrophic hormone (ACTH); pathological ACTH excess results in severe central obesity. We have investigated the role of common polymorphisms of the POMC gene on abdominal obesity in a large family study.

Subject collection and phenotyping.

The collection strategy of this family study has been previously described (7). Briefly, families were ascertained through a proband with essential hypertension. To be suitable for the study, families were required to consist of at least three siblings clinically assessable for blood pressure if at least one parent of the sibship was available to give blood for DNA analysis and to consist of at least four assessable siblings if no parent was available. Where members of the sibship were found to be hypertensive, families were extended and the spouses and offspring of hypertensive sibs collected. Thus, the majority of the individuals in the family collection have blood pressure values within the normal range, and the family collection includes some extended families, though most are nuclear families. A full medical and lifestyle history was taken. Anthropometric measurements including height, weight, and waist and hip circumferences were made (waist measured at the natural waist, and hip measured at the level of the greater trochanters). A total of 1,428 individuals from 248 families participated in the study.

Laboratory methods.

Plasma leptin was measured using a commercially available enzyme-linked immunosorbent assay kit (Linco Research, St. Charles, MO) with sensitivity 0.5 ng/ml, intra-assay precision 2.6–4.6%, and interassay precision 2.6–6.2%. Plasma cortisol and 24-h urinary excretion rates of cortisol metabolites were measured as previously described (8). We used the sum of urinary tetrahydrocortisol, allo(tetrahydrodeoxycortisol), and tetrahydrodeoxycortisol as a measure of total cortisol production.

The POMC gene occupies a short genomic segment of ∼7.5 kb on chromosome 2p24. No common polymorphisms exist within the coding sequence of the gene in Caucasians. We therefore studied three common noncoding polymorphisms, which spanned the gene (Fig. 1A), by PCR, restriction enzyme digestion, and agarose gel electrophoresis. The RsaI polymorphism 1798 bp 5′ to the beginning of exon 1 (representing a C/T substitution, dbSNP ID rs3754860) was genotyped using primers 5′GTAACTTCAGGAGGCTGCTGG3′ and 5′CCAGTGATTCTAAATGCAGTTC3′, annealing temperature 52°C and 0.5 mmol/l MgCl2, and restriction digestion with RsaI. A C/G polymorphism within the first intron (referred to here as C1032G, dbSNP ID rs1009388) was typed using primers 5′CTCGACAAACTTTCTGCGC3′ and 5′CGGTGAGCAGAGATCACC3′, annealing temperature 69°C and 1.5 mmol/l MgCl2, and restriction enzyme AvaI. The C8246T polymorphism in the 3′ untranslated region (referred to in some previous studies as C7566T, dbSNP ID rs1042571) was genotyped using primers 5′AGCCCCGCAGCGATGGT3′ and 5′CCTTTCACGCTCACTAAGTCCTGTG3′, annealing temperature 63°C and 1.5 mmol/l MgCl2, and restriction enzyme MboII. Reference individuals of known genotype were included in each run.

Statistical analysis.

Mendelian inheritance of all genotypes was checked using PedCheck (9). Possible genotyping errors not producing Mendelian inconsistencies were screened for by confirming the correspondence of genotype frequencies to Hardy-Weinberg proportions in the genotyped founders (n = 269–275 for the different polymorphisms) and by examining the families for recombination in this short genomic segment using MERLIN (10). The estimated genotyping error rate was <1%. Disequilibrium coefficients and haplotype frequencies were estimated in the founders using FUGUE (11). Phenotypes of interest were examined for normality and log transformed in the case of BMI and plasma leptin. Significant covariates of these phenotypes were then determined by linear regression using MINITAB software; the adjusted values from these regressions were standardized (so each phenotype has a mean of 0 and an SD of 1) and used in the genetic analyses. Association between genotypes and the adjusted phenotypes was assessed in the families by calculating identity-by-descent vectors for each individual using MERLIN followed by variance components analysis using the QTDT (quantitative transmission disequilibrium test) program (12). Empirical significance levels were determined by Monte Carlo permutation of 5,000 replicates (13). A three-locus measured genotype analysis was also carried out using the Pedigree Analysis Package (PAP version 5.0, available at http://hasstedt.genetics.utah.edu/pap5) in which multilocus additive effects (i.e., no gene interactions) were fitted. Linkage disequilibrium was allowed in the exact likelihood calculations as well as residual intrafamilial correlations in a class D regressive model. A hierarchical analysis was performed to evaluate the fit of nested models.

Characteristics of the study subjects are shown in Table 1. Sixty percent of families comprised between four and six genotyped members. Median values for BMI, WHR, and plasma leptin lie within the normal range for an unselected U.K. Caucasian population. Correction for age, sex, alcohol consumption, smoking, and habitual exercise accounted for ∼15% of the total variability in log BMI and 46–49% of the total variability in WHR and log plasma leptin level. The heritability of the phenotypes after correction was 35% for log BMI, 28% for WHR, and 36% for log plasma leptin (all P < 0.0001).

Genotyping was successful for >95% of all subjects for all polymorphisms. Genotype frequencies at all polymorphisms satisfied Hardy-Weinberg equilibrium (Table 2). Linkage disequilibrium was, as expected, strong in this small genomic region; D′ values were 0.868 (between C8246T and C1032G), 0.964 (between C8246T and RsaI), and 0.933 (between C1032G and RsaI), respectively. There were four common haplotypes of the three polymorphisms (online appendix Table 1 [available at http://diabetes.diabetesjournals.org])

There was evidence for significant association between genotypes at the C8246T polymorphism and WHR (0.197 ± 0.047 [means ± SE] higher-adjusted standardized WHR per T allele, P < 0.0001) and between genotypes at the C1032G polymorphism and WHR (0.133 ± 0.042 higher-adjusted standardized WHR per G allele, P = 0.002) (Table 3). There was no significant association between genotypes at the RsaI polymorphism and WHR (P = 0.145) (Table 3). There was also no significant association between any polymorphism and BMI, plasma leptin level, plasma cortisol, or urinary cortisol metabolite excretion (data available on request). To investigate more specifically the effect of POMC polymorphisms on fat distribution, we additionally adjusted the WHR phenotype for log BMI. Log BMI was responsible for an additional 12.5% of the variability in adjusted WHR (P < 0.0001). Following adjustment for log BMI, there remained significant association between genotypes at the C8246T polymorphism and WHR (0.184 ± 0.044 [means ± SE] higher-adjusted standardized WHR per T allele, P < 0.0001) and between genotypes at the C1032G polymorphism and WHR (0.119 ± 0.040 higher-adjusted standardized WHR per G allele, P = 0.003). There was no association between genotypes at the RsaI polymorphism and WHR (P = 0.292) (Table 3). These associations remained significant at the P < 0.01 level after Monte Carlo simulation of 5,000 replicates, and significance (at P < 0.05) persisted after application of a Bonferroni correction for 12 analyses. There was no evidence of heterogeneity in the strength of the association between men and women (F1,1,315 = 2.043, P = 0.153). Measured genotype analysis using the Pedigree Analysis Package showed no support for additional genetic effects over the C8246T marker on the WHR phenotype (P > 0.45) and no evidence for dominance effects encoded by C8246T (P = 0.44). Models robust to stratification fitted in the QTDT program (that is, incorporating transmission from heterozygous parents only) confirmed support for the association at C8246T (P = 0.026) and C1032G (P = 0.0127). The proportion of the total variance in the WHR phenotype associated with the C8246T polymorphism (narrow-sense heritability) was estimated to be 1.1% (95% CI 0.2–2.7) (Fig. 1B). Post hoc power calculations indicated that the cohort had >90% power to detect a 0.2 SD difference between genotypes at either the C8246T or C1032G polymorphisms.

Patients with a complete deficiency of POMC due to homozygous or compound heterozygous loss-of-function mutations exhibit a characteristic syndrome marked by childhood-onset severe obesity, red hair, and hypocortisolism (14). With respect to nonsyndromic obesity, four mutation screening studies involving 601 nonsyndromic obese subjects have shown missense or nonsense mutations in just 7 individuals, implying that coding sequence variation in POMC is a very uncommon cause of nonsyndromic obesity (1518). Two genome-screening studies, in French obese sibling pairs and in Mexican-American extended families, suggested strong evidence for linkage of plasma leptin levels to the region of chromosome 2 harboring the POMC gene (19,20). Association between haplotypes of the RsaI and C8246T polymorphisms and plasma leptin levels (though not BMI or WHR) was subsequently shown in the Mexican-American cohort, with higher plasma leptin observed in the haplotype comprising a present RsaI restriction site (+ in this study) and a T allele at C8246T (21). Although we found no association between any genotype and plasma leptin, the same T allele at C8246T is associated with higher WHR in the present study. No association between POMC polymorphisms and either plasma leptin or BMI was found in the French cohort (22). No previous study has reported either positive or negative findings with respect to common POMC polymorphisms and WHR.

The finding in the present study that POMC polymorphisms affect WHR even after adjustment for BMI suggests that the principal effect of these common polymorphisms is on susceptibility to central rather than generalized obesity. Since dysregulation of appetite and energy expenditure (known to be mediated via hypothalamic actions of POMC) might be expected to predispose to generalized obesity, our observations may reflect actions of POMC-derived peptides whose effects on obesity are extrahypothalamic. Posttranslational processing of POMC differs in different tissues and yields a large number of such peptides. For example, β-endorphin is produced in lymphocytes, macrophages, and the small intestine in addition to the central nervous system (23). Plasma concentrations of β-endorphin are higher in obese individuals even after successful weight loss when compared with nonobese control subjects, and gene-targeted mice that have normal melanocortin signaling but produce no β-endorphin are obese and hyperphagic (24). β-Lipotropin, another POMC-derived peptide, is an important promoter of lipolysis in adipocytes and is present in higher concentrations in the plasma of centrally obese subjects (25). Further research will be necessary to investigate whether the polymorphisms we have typed are correlated with plasma and tissue concentrations of these and other POMC-derived peptides. Although differences in ACTH production caused by POMC polymorphisms would be a strong candidate mechanism for an effect on central obesity, we found no association between POMC variants and steroid phenotypes in this study.

Previous studies that have examined the POMC gene and obesity phenotypes have in general enrolled participants with significantly higher BMIs and plasma leptin levels than the present study. Thus, one limitation of our study is that, since the majority of our study population has BMIs in the high normal rather than the obese range, it is not possible to rule out absolutely effects of common POMC variants on BMIs >30 kg/m2 or on the highest end of the distribution of plasma leptin level. However, it seems unlikely that strong effects on these phenotypes are present within the range of BMI and plasma leptin present in our sample (which would encompass ∼80% of the U.K. population). Furthermore, no previously reported study has found association between POMC polymorphisms and BMI, even in highly selected samples. Strengths of our study include its large size, the genotyping of three informative polymorphisms spanning POMC (thus likely capturing all common variation in this small gene), its family-based design, which circumvents concerns about population stratification, and its detailed ascertainment, which permits statistical correction of obesity phenotypes for important lifestyle confounders.

Central obesity may be a more powerful predictor of both CHD and type 2 diabetes risk than generalized obesity. These POMC polymorphisms would be suitable candidates for testing in large-scale case-control studies of genetic susceptibility to these diseases. Although their effect on abdominal obesity is relatively modest, they are common and thus could potentially contribute significantly to the population attributable risk.

Additional information for this article can be found in an online appendix at http://diabetes.diabetesjournals.org.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Han TS, Feskens EJ, Lean ME, Seidell JC: Associations of body composition with type 2 diabetes mellitus.
Diabet Med
15
:
129
–135,
1998
2.
Yusuf S, Hawken S, Ounpuu S, Dans T, Avezum A, Lanas F, McQueen M, Budaj A, Pais P, Varigos J, Lisheng L: Effect of potentially modifiable risk factors associated with myocardial infarction in 52 countries (the INTERHEART study): case-control study.
Lancet
364
:
937
–952,
2004
3.
Stunkard AJ, Foch TT, Hrubec Z: A twin study of human obesity.
JAMA
256
:
51
–54,
1986
4.
Borecki IB, Higgins M, Schreiner PJ, Arnett DK, Mayer-Davis E, Hunt SC, Province MA: Evidence for multiple determinants of the body mass index: the National Heart, Lung, and Blood Institute Family Heart Study.
Obes Res
6
:
107
–114,
1998
5.
Yeo GS, Farooqi IS, Challis BG, Jackson RS, O’Rahilly S: The role of melanocortin signalling in the control of body weight: evidence from human and murine genetic models.
QJM
93
:
7
–14,
2000
6.
Krude H, Biebermann H, Gruters A: Mutations in the human proopiomelanocortin gene.
Ann N Y Acad Sci
994
:
233
–239,
2003
7.
Keavney B, McKenzie CA, Connell JM, Julier C, Ratcliffe PJ, Sobel E, Lathrop M, Farrall M: Measured haplotype analysis of the angiotensin-I converting enzyme gene.
Hum Mol Genet
7
:
1745
–1751,
1998
8.
Keavney B, Mayosi B, Gaukrodger N, Imrie H, Baker M, Fraser R, Ingram M, Watkins H, Farrall M, Davies E, Connell J: Genetic variation at the locus encompassing 11-beta hydroxylase and aldosterone synthase accounts for heritability in cortisol precursor (11-deoxycortisol) urinary metabolite excretion.
J Clin Endocrinol Metab
90
:
1072
–1077,
2005
9.
O’Connell JR, Weeks DE: PedCheck: a program for identification of genotype incompatibilities in linkage analysis.
Am J Hum Genet
63
:
259
–266,
1998
10.
Abecasis GR, Cherny SS, Cookson WO, Cardon LR: Merlin: rapid analysis of dense genetic maps using sparse gene flow trees.
Nat Genet
30
:
97
–101,
2002
11.
Dawson E, Abecasis GR, Bumpstead S, Chen Y, Hunt S, Beare DM, Pabial J, Dibling T, Tinsley E, Kirby S, Carter D, Papaspyridonos M, Livingstone S, Ganske R, Lohmussaar E, Zernant J, Tonisson N, Remm M, Magi R, Puurand T, Vilo J, Kurg A, Rice K, Deloukas P, Mott R, Metspalu A, Bentley DR, Cardon LR, Dunham I: A first-generation linkage disequilibrium map of human chromosome 22.
Nature
418
:
544
–548,
2002
12.
Abecasis GR, Cardon LR, Cookson WO: A general test of association for quantitative traits in nuclear families.
Am J Hum Genet
66
:
279
–292,
2000
13.
McIntyre LM, Martin ER, Simonsen KL, Kaplan NL: Circumventing multiple testing: a multilocus Monte Carlo approach to testing for association.
Genet Epidemiol
19
:
18
–29,
2000
14.
Krude H, Biebermann H, Luck W, Horn R, Brabant G, Gruters A: Severe early-onset obesity, adrenal insufficiency and red hair pigmentation caused by POMC mutations in humans.
Nat Genet
19
:
155
–157,
1998
15.
Miraglia del Giudice E, Cirillo G, Santoro N, D’Urso L, Carbone MT, Di Toro R, Perrone L: Molecular screening of the proopiomelanocortin (POMC) gene in Italian obese children: report of three new mutations.
Int J Obes Relat Metab Disord
25
:
61
–67,
2001
16.
Hinney A, Becker I, Heibult O, Nottebom K, Schmidt A, Ziegler A, Mayer H, Siegfried W, Blum WF, Remschmidt H, Hebebrand J: Systematic mutation screening of the pro-opiomelanocortin gene: identification of several genetic variants including three different insertions, one nonsense and two missense point mutations in probands of different weight extremes.
J Clin Endocrinol Metab
83
:
3737
–3741,
1998
17.
Echwald SM, Sorensen TI, Andersen T, Tybjaerg-Hansen A, Clausen JO, Pedersen O: Mutational analysis of the proopiomelanocortin gene in Caucasians with early onset obesity.
Int J Obes Relat Metab Disord
23
:
293
–298,
1999
18.
Challis BG, Pritchard LE, Creemers JW, Delplanque J, Keogh JM, Luan J, Wareham NJ, Yeo GS, Bhattacharyya S, Froguel P, White A, Farooqi IS, O’Rahilly S: A missense mutation disrupting a dibasic prohormone processing site in pro-opiomelanocortin (POMC) increases susceptibility to early-onset obesity through a novel molecular mechanism.
Hum Mol Genet
11
:
1997
–2004,
2002
19.
Comuzzie AG, Hixson JE, Almasy L, Mitchell BD, Mahaney MC, Dyer TD, Stern MP, MacCluer JW, Blangero J: A major quantitative trait locus determining serum leptin levels and fat mass is located on human chromosome 2.
Nat Genet
15
:
273
–276,
1997
20.
Hager J, Dina C, Francke S, Dubois S, Houari M, Vatin V, Vaillant E, Lorentz N, Basdevant A, Clement K, Guy-Grand B, Froguel P: A genome-wide scan for human obesity genes reveals a major susceptibility locus on chromosome 10.
Nat Genet
20
:
304
–308,
1998
21.
Hixson JE, Almasy L, Cole S, Birnbaum S, Mitchell BD, Mahaney MC, Stern MP, MacCluer JW, Blangero J, Comuzzie AG: Normal variation in leptin levels in associated with polymorphisms in the proopiomelanocortin gene, POMC.
J Clin Endocrinol Metab
84
:
3187
–3191,
1999
22.
Delplanque J, Barat-Houari M, Dina C, Gallina P, Clement K, Guy-Grand B, Vasseur F, Boutin P, Froguel P: Linkage and association studies between the proopiomelanocortin (POMC) gene and obesity in caucasian families.
Diabetologia
43
:
1554
–1557,
2000
23.
Coll AP, Farooqi IS, Challis BG, Yeo GS, O’Rahilly S: Proopiomelanocortin and energy balance: insights from human and murine genetics.
J Clin Endocrinol Metab
89
:
2557
–2562,
2004
24.
Appleyard SM, Hayward M, Young JI, Butler AA, Cone RD, Rubinstein M, Low MJ: A role for the endogenous opioid beta-endorphin in energy homeostasis.
Endocrinology
144
:
1753
–1760,
2003
25.
Genazzani AR, Facchinetti F, Petraglia F, Pintor C, Corda R: Hyperendorphinemia in obese children and adolescents.
J Clin Endocrinol Metab
62
:
36
–40,
1986