Contents
Contributors
Search

EVIDENCE FOR INHERITED FACTORS DETERMINING FAT MASS

Heritability estimates the proportion of the total phenotypic variance attributable to genetic variation under a polygenic model, by comparing the similarity of a trait within monozygotic twins with the similarity within dizygotic twins. Heritability is both a function of the number of genes influencing a phenotype and the proportion of phenotypic variation accounted for by each of these genes.

Twin studies

Traditionally the most favoured model for separation of the genetic component of variance is based on studies of twins, as monozygotic co-twins share 100 percent of their genes and dizygotes 50% on average. Twin studies suggest a heritability of fat mass of between 40 and 70% with a concordance of 0.7-0.9 between monozygotic twins compared to 0.35-0.45 between dizygotic twins (1). However, in traditional nuclear families, family members generally share both genes and environments to some degree, so it is difficult to assess the contribution of either of the individual components.

Correlation of monozygotic twins reared apart is virtually a direct estimate of the heritability, although monozygotic twins do share the intrauterine environment, which may contribute to lasting differences in body mass in later life. Estimates vary from 40% to 70%, depending on age of separation of twins and the length of follow-up (reviewed in (2)). Studies of Swedish twins (3) have suggested a heritability of 0.70 for men and 0.66 for women, whilst a heritability of 0.61 was observed in a cohort of UK twins (4). In a meta-analysis of results derived from Finnish, Japanese and American archival twins, Allison observed similar correlations (5). In addition, Price and colleagues have shown that these correlations did not differ significantly between twins reared apart and twins reared together, and between twins reared apart in more similar (ie: with relatives) versus less similar environments (4).

Adoption studies

Complete adoption studies are useful in separating the common environmental effects since adoptive parents and their adoptive offspring share only environmental sources of variance, whilst the adoptees and their biological parents share only genetic sources of variance. One of the largest series, based on over 5000 subjects from the Danish adoption register which contains complete and detailed information on the biological parents, showed a strong relationship between the BMI of adoptees and biological parents across the whole range of body fatness but none when compared with the adoptive parents (6). The Danish group have also shown a close correlation between BMI of adoptees and their biological full siblings who were reared separately by the biological parents of the adoptees, and a similar, but weaker relationship with half-siblings (7).

PLEIOTROPIC OBESITY SYNDROMES

It is well established that obesity runs in families, although the vast majority of cases do not segregate with a clear Mendelian pattern of inheritance (8, 9). There are about 30 Mendelian disorders with obesity as a clinical feature but often associated with mental retardation, dysmorphic features and organ-specific developmental abnormalities (ie: pleiotropic syndromes). A number of families with these rare pleiotropic obesity syndromes have been studied by linkage analysis and the known chromosomal loci for obesity syndromes are summarised in Table 1. For a comprehensive list of syndromes in which obesity is a recognised part of the phenotype, see Online Mendelian Inheritance in Man (OMIM). World Wide Web URL: http://www.ncbi.nlm.nih.gov/omim/

Prader-Willi Syndrome

The Prader-Willi syndrome is the most common syndromal cause of human obesity with an estimated prevalence of about 1 in 25,000 (10). It is an autosomal dominant disorder and is caused by deletion or disruption of a paternally imprinted gene or genes on the proximal long arm of chromosome 15.

The Prader-Willi syndrome (PWS) is characterized by diminished foetal activity, obesity, hypotonia, mental retardation, short stature, hypogonadotropic hypogonadism, and small hands and feet (11). The diagnostic criteria arrived at by a consensus group (12) were based on a point system, 1 point each was allowed for each of 5 major criteria, such as feeding problems in infancy and failure to thrive, and one-half point each for 7 minor criteria, such as hypopigmentation. A minimum of 8.5 points was considered necessary for the clinical diagnosis of PWS.

Obesity phenotype

There is mild prenatal growth retardation with a mean birth weight of about 6 lbs (2.8 kg) at term, hyporeflexia and poor feeding in neonatal life due to diminished swallowing and sucking reflexes; infants often require assisted feeding for about 3 to 4 months. Feeding difficulties generally improve by the age of 6 months. From 12 to 18 months onward, uncontrollable hyperphagia results in severe obesity invariably associated with abdominal striae. Diabetes mellitus is not a diagnostic criterion for Prader-Willi syndrome (PWS), but it is often found in older PWS patients (13).

Eilholzer presented data on body composition and leptin levels in 13 young, still underweight children and 10 older overweight children with PWS (14). These authors and others have concluded that leptin production appears to be intact and that body composition in PWS is already disturbed in infancy, before the development of obesity (15).

Lindgren studied the eating behaviour of patients with PWS (mean age 10 years) and obese and normal weight controls of the same age. Subjects with PWS had a longer duration of eating and 56 % of the eating curves were non-decelerating, compared with 10% of the normal weight group and 30% of the obese group (16). Whilst hyperphagia is a dominant feature in PWS subjects, the eating behaviour in PWS might be due to decreased satiation as well as increased hunger. One suggested mediator of the obesity phenotype in PWS patients is the novel enteric hormone Ghrelin, which is implicated in the regulation of meal-time hunger in rodents and humans and is also a potent stimulator of growth hormone secretion (17-19). Cummings et al. have found that fasting plasma ghrelin levels are 4.5-fold higher in PWS subjects than equally obese controls and thus may be implicated in the pathogenesis of hyperphagia in these patients (20).

Endocrine disorder

Children with Prader-Willi syndrome (PWS) display diminished growth, reduced muscle mass (lean body mass), and increased fat mass - body composition abnormalities resembling those seen in growth hormone (GH) deficiency (21). Diminished GH responses to various provocative agents, low insulin-like growth factor-I levels, and the presence of additional evidence of hypothalamic dysfunction support the presence of true GH deficiency (GHD) in many children with PWS (22). GH treatment in these children decreases body fat, and increases linear growth, muscle mass, fat oxidation and energy expenditure (23). Physical strength and agility are also improved (24). These improvements are most dramatic during the first year of GH therapy, although prolonged treatment does not completely normalize these parameters.

Hypogonadism is thought to be due to hypogonadotrophic hypogonadism in a majority of PWS patients and treatment with clomiphene citrate has been shown to raise plasma luteinizing hormone, testosterone, and urinary gonadotropin levels to normal and result in normal spermatogenesis and physical signs of puberty (25).

Establishing a genetic diagnosis

It is clear that chromosomal mechanisms are principally responsible for PWS and that the syndrome is caused by lack of the paternal segment 15q11.2-q12 (26). There are 2 mechanisms by which such a loss can occur: either through deletion of the paternal 'critical' segment or through loss of the entire paternal chromosome 15 with presence of 2 maternal homologues (uniparental maternal disomy) (27). The opposite, i.e., maternal deletion or paternal uniparental disomy, causes another characteristic phenotype, the Angelman syndrome (AS). This indicates that both parental chromosomes are differentially imprinted, and that both are necessary for normal embryonic development.

In the vast majority of PWS, chromosomal abnormalities occur sporadically. These instances include virtually all interstitial deletions, the large majority of de novo unbalanced translocations, all instances of maternal uniparental disomy with normal karyotype or with a de novo rearrangement involving chromosome 15, and almost all cases of maternal uniparental disomy with a familial rearrangement involving chromosome 15.

Laboratory diagnostic tests

Deletions account for 70 to 80% of cases; the majority are interstitial deletions, many of which can be visualized by prometaphase banding examination (28). A minority consist of unbalanced translocations, mostly de novo, which are easily detected by routine chromosome examination (29). The remainder of cases are the result of maternal uniparental disomy. In most of these latter cases, cytogenetic examinations yield normal results.

Using restriction digests with the methyl-sensitive enzymes HpaII and HhaI and probing Southern blots with several genomic and cDNA probes, Driscoll systematically scanned segments of 15q11-q13 for DNA methylation differences between patients with PWS and those with AS (30). They found that the sequences identified by the cDNA DN34, which is highly conserved in evolution, demonstrate distinct differences in DNA methylation of the parental alleles at the D15S9 locus. Clayton-Smith used DN34 to perform methylation analysis and showed that the methylation pattern varied according to the parent of origin, providing further evidence for the association of methylation with genomic imprinting (31). Thus, DNA methylation can be used as a reliable postnatal diagnostic tool (30, 32).

Candidate genes for PWS

Within the 4.5Mb PWS region in 15q11-q13, where there is a lack of expression of paternally imprinted genes, several candidate genes have been studied and their expression shown to be absent in the brains of PWS patients (33). These include necdin (34, 35) and small nuclear ribonucleoprotein polypeptide N (SNRPN) (36-38). More recent reports have suggested that other non-expressed genes may include the Ring Zinc finger 127 polypeptide gene (39, 40), the MAGE-like 2 gene (41, 42) and the Prader-Willi critical region 1 gene (43). .  Wevrick and colleagues have showed that Necdin and Magel2 bind to and prevent proteasomal degradation of FEZ1, a protein implicated in axonal outgrowth and kinesin mediated transport and also bind to the Bardet Beidl syndrome (BBS) protein BBS4 invitro (43.1).  Interactions amongst these molecules occur at or near centrosomes and centrosomal dysfunction has been implicated in BBS and may be relevant to the common phenotypic features of learning disabilities, hypogonadism and obesity.  However, the precise role of these genes and the mechanisms by which they lead to a pleiotropic obesity syndrome remain elusive.

Albright hereditary osteodystrophy

Gs is the ubiquitously expressed heterotrimeric G protein that couples receptors to the effector enzyme adenylyl cyclase and is required for receptor-stimulated intracellular cAMP generation. Inactivating and activating mutations in the gene encoding G alpha s (GNAS1) are known to be the basis for 2 well-described contrasting clinical disorders, Albright hereditary osteodystrophy (AHO) and McCune-Albright syndrome (MAS). AHO is an autosomal dominant disorder due to germline mutations in GNAS1 that decrease expression or function of G alpha s protein. Heterozygous loss-of-function mutations lead to AHO, a disease characterized by short stature, obesity, skeletal defects, and impaired olfaction. Maternal transmission of GNAS1 mutations leads to AHO plus resistance to several hormones (e.g., parathyroid hormone) that activate Gs in their target tissues (pseudohypoparathyroidism type IA), while paternal transmission leads only to the AHO phenotype (pseudopseudohypoparathyroidism). Studies in both mice and humans demonstrate that GNAS1 is imprinted in a tissue-specific manner, being expressed primarily from the maternal allele in some tissues and biallelically expressed in most other tissues, thus multi-hormone resistance occurs only when Gs (alpha) mutations are inherited maternally (44, 45).

Fragile X syndrome

Fragile X syndrome is characterized by moderate to severe mental retardation, macro-orchidism, large ears, prominent jaw, and high-pitched jocular speech associated with mutations in the FMR1 gene (46). Expression is variable, with mental retardation being the most common feature. A Prader-Willi-like subphenotype of the Fragile X syndrome has been described (47). The features were extreme obesity with a full, round face, small, broad hands and feet, and regional skin hyperpigmentation. Unlike the Prader-Willi syndrome, the patients lacked the neonatal hypotonia and feeding problems during infancy followed by hyperphagia during toddlerhood (48). In a group of 26 patients with suspected Prader-Willi syndrome but without detectable molecular abnormalities of chromosome 15, one fragile X patient was found. Schrander-Stumpfel and colleagues found the FMR1 mutation in a 3-year-old boy with unexplained extreme obesity and delayed motor and speech development (49). They compared the clinical features with those in 9 reported patients with the fragile X syndrome and extreme obesity. Behavioural characteristics such as hyperkinesis, autistic-like behaviour, and apparent speech and language deficits may help point toward the diagnosis of the fragile X syndrome (50, 51). It has been suggested that a reasonable estimate of frequency is 0.5 per 1000 males (52).There are new guidelines which describe criteria for genetic counselling and testing of individuals with Fragile X syndrome as well as carriers and potential carriers for Fragile X mutations (52.1).

Bardet Biedl Syndrome

Bardet-Biedl syndrome (BBS) is a rare (prevalence <1/100,000), autosomal recessive disease characterized by obesity, mental retardation, dysphormic extremities (syndactyly, brachydactyly or polydactyly), retinal dystrophy or pigmentary retinopathy, hypogonadism or hypogenitalism (limited to male patients) and structural abnormalities of the kidney or functional renal impairment. Although these features were present in the patients of Biedl and Bardet, the patients of Laurence and Moon had a distinct disorder with paraplegia and without polydactyly and obesity and residual heterogeneity exists even after the Laurence-Moon syndrome is separated (53). The differential diagnosis includes Biemond syndrome II (iris coloboma, hypogenitalism, obesity, polydactyly, and mental retardation) and Alstrom syndrome (retinitis pigmentosa, obesity, diabetes mellitus and deafness) (the nosology of these and related syndromes is reviewed in (54)).

Bardet-Biedl syndrome is a genetically heterogeneous disorder that is now known to map to at least eight loci (reviewed in (55)), a number of which have now been identified at the molecular level. Although BBS is usually transmitted as a recessive disorder, some families have exhibited so called “tri-allelic” inheritance where the clinical manifestation of the syndrome requires two mutations in one BBS gene plus an additional mutation in a second, unlinked BBS gene (56, 57).
The recent discovery of the novel BBS3,5,7 and 8 genes by phylogenetic/genomic approaches has led to rapid progress in the understanding of Bardet Beidl syndrome (58). These proteins are all involved in basal body and centrosomal function and impact on ciliary development and transport (59). BBS1, 2, BBS6 and BBS4 are all involved in intracellular trafficking (60) and mice lacking the Bbs4 protein recapitulate the major components of the human phenotype, including obesity and retinal degeneration (61). It remains to be seen what role these proteins and organelles play in energy balance, cognitive impairment and renal development.

Identification of causative genes

The BBS1 gene has recently been identified using a positional cloning approach (57). However, the protein does not show any significant similarity with any known protein or protein family and its function is unknown. Families with BBS mapping to BBS6 have been found to harbour mutations in MKKS (58) which has sequence homology to the alpha subunit of a prokaryotic chaperonin in the thermosome Thermoplasma acidophilum. Mutations in this gene also cause McKusick-Kaufman syndrome (hydrometrocolpos, post-axial polydactyly and congenital heart defects). In addition, the genes underlying BBS2 (59) and BBS4 (60) have recently been identified by positional cloning. However, they have no significant homology to chaperonins and the functions of these proteins remain unknown.

Genotype-phenotype correlations

The clinical manifestations of BBS have been compared in 3 unrelated, extended Arab-Bedouin kindreds in which linkage had been demonstrated to chromosomes 3 (BBS3), 15 (BBS4), and 16 (BBS2) (61). Observed differences included the limb distribution of the postaxial polydactyly and the extent and age-association of obesity. It appeared that the chromosome 3 locus is associated with polydactyly of all 4 limbs, while polydactyly of the chromosome 15 type (BBS4) is mostly confined to the hands. The chromosome 15 type is associated with early-onset morbid obesity, while the chromosome 16 type appears to present the 'leanest' BBS phenotype.

CANDIDATE GENE MUTATIONS IN MONOGENIC HUMAN OBESITY

The recent identification and characterisation of single gene defects in rodent obesities inherited in either a dominant (yellow, Ay/a) or recessive (ob/ob, db/db, fa/fa, tb/tb) manner has given substantial insights into the physiological disturbances that can lead to obesity, the metabolic and endocrine abnormalities associated with the obese phenotype, and the more detailed anatomical and neurochemical pathways that regulate energy intake and energy expenditure (see Chapters 3 and 4). These studies provide the basic framework upon which the understanding of the more complex mechanisms in humans can be built.

In the past five years several human disorders of energy balance that arise from genetic defects have been described. All of these are in molecules identical or similar to those known to cause obesity in genetic and experimental syndromes of obesity in rodents. These mutations all result in morbid obesity in childhood without the developmental pleiotropic features characteristic of the recognised syndromes of childhood obesity.

Congenital leptin deficiency

The first monogenic human obesity syndrome to be reported was congenital leptin deficiency. Two severely obese cousins in a highly consanguineous family of Pakistani origin were found to have undetectable levels of serum leptin (62). They were homozygous for a frameshift mutation in the ob gene (DG133), which resulted in a truncated protein that was not secreted (63). These children were severely hyperphagic, constantly demanding food, with an intense drive to eat, which was never satisfied. They developed severe disabling obesity (an 8yr old girl weighing 86kg and a 2yr old boy weighing 29kg) characterised by an excess deposition of fat but without any of the clinical features suggestive of the recognised childhood obesity syndromes. No disturbance of 24 hour energy expenditure measured by indirect calorimetry or free-living energy expenditure measured by isotope dilution was identified. Mild hyperinsulinaemia, advanced skeletal maturation, impaired T cell mediated immunity and hypogonadotropic hypogonadism are other features also seen in three leptin deficient subjects from a Turkish family homozygous for a missense mutation in the ob gene (64). Studies of the heterozygote members of the DG133 families indicate that their leptin levels are lower than ethnically matched control subjects and this partial leptin deficiency results in increased fat mass (mean 23% more than predicted) (65). Thus the phenotype of human leptin deficiency due to mutations in the ob gene is remarkably similar to that seen in leptin-deficient rodents.

The administration of leptin to leptin-deficient ob/ob mice results in a decrease in food intake, weight loss and restoration of fertility and T cell mediated immune function (66-69). Three leptin-deficient children have been treated with daily subcutaneous injections of recombinant human leptin for up to four years with sustained, beneficial effects on appetite, fat mass, hyperinsulinaemia and hyperlipidaemia (70). The major impact of leptin on human energy balance was mediated via its suppressive effects on food intake with a marked reduction in caloric consumption during a test meal and this was associated with parental reports of a near-normalisation of eating behaviour in the domestic setting (Figure 1 depicts clinical response). In contrast, no effect of leptin on basal metabolic rate or free-living energy expenditure was seen. Leptin administration permits the full progression of appropriately timed puberty but does not appear to cause precocious activation of the pubertal process in younger children (Figure 2). Although this syndrome is rare, these studies have established that in humans, leptin has profound effects on appetite and may act as a metabolic gate for the onset of puberty at an appropriate developmental age (70).

Figure 1. Response to leptin therapy inCongenital leptin deficiency

Figure 2. Role of leptin in human physiology - lessons from congenital leptin deficiency

Leptin receptor deficiency

A mutation in the leptin receptor has been reported in one consanguineous family (71). Affected individuals were homozygous for a mutation that truncates the receptor before the transmembrane domain and the mutated receptor circulates bound to leptin. There are a number of phenotypic similarities with the leptin-deficient subjects. Leptin receptor deficient subjects were also born of normal birthweight, exhibited rapid weight gain in the first few months of life, with severe hyperphagia and aggressive behaviour when denied food. Basal temperature and resting metabolic rate were normal, cortisol levels were in the normal range and all subjects were normoglycaemic with mildly elevated plasma insulins as seen in leptin-deficient subjects. In contrast, some neuroendocrine features were unique to leptin receptor deficiency. The presence of mild growth retardation in early childhood with impaired basal and stimulated growth hormone secretion and decreased IGF-1 and IGF-BP3 levels and evidence of hypothalamic hypothyroidism in these subjects, suggest that loss of the leptin receptor results in a more diverse phenotype than loss of its ligand leptin.

POMC deficiency

The behavioural and neuroendocrine effects of leptin are thought to be mediated through its actions at hypothalamic leptin receptors (72, 73). Pro-opiomelanocortin (POMC) is produced by hypothalamic neurones of the arcuate nucleus, is regulated positively by leptin and is sequentially cleaved by prohormone convertases to yield peptides including aMSH that suppress feeding in rodents (reviewed in (74)).

Two unrelated obese German children with homozygous or compound heterozygous mutations in POMC have been reported (75). These children were hyperphagic, developing early-onset obesity presumably as a result of impaired melanocortin signalling in the hypothalamus. They presented in neonatal life with adrenal crisis due to isolated ACTH deficiency (POMC is a precursor of ACTH in the pituitary) and had pale skin and red hair due to the lack of MSH function at melanocortin 1 receptors in the skin (Figure 3). Three further patients with homozygous or compound heterozygous complete loss of function mutations of the POMC gene have been described (76). Recently, a number of groups have identified a heterozygous missense mutation (Arg236Gly) in POMC that disrupts the dibasic amino acid processing site between b-MSH and b-endorphin (77-79). This mutation results in an aberrant b-MSH/b-endorphin fusion peptide which binds to MC4R with an affinity identical to that of a- and b-MSH but has a markedly reduced ability to activate the receptor. Thus this cleavage site mutation in POMC may confer susceptibility to obesity through a novel molecular mechanism.

Figure 3. POMC derived peptides and the syndrome of human POMC deficiency

Prohormone Convertase 1 deficiency

Further evidence for the role of the melanocortin system in the regulation of body weight in humans comes from the description of a 47 year old woman with severe childhood obesity, abnormal glucose homeostasis, very low plasma insulin but elevated levels of proinsulin, hypogonadotropic hypogonadism and hypocortisolaemia associated with elevated levels of POMC (80). This subject was found to be a compound heterozygote for mutations in prohormone convertase 1, which cleaves prohormones at pairs of basic amino acids, leaving C-terminal basic residues which are then excised by carboxypeptidase E (CPE) (81). We have recently identified a child with severe, early-onset obesity who was a compound heterozygote for complete loss of function mutations in PC1 (personal observations). Although failure to cleave POMC is a likely mechanism for the obesity in these patients, PC1 cleaves a number of other neuropeptides in the hypothalamus, such as Glucagon-like-peptide 1, which may influence feeding behaviour. The phenotype of these subjects, is very similar to that seen in the CPE deficient fat/fat mouse implicating this part of the pathway in the control of body weight in humans. To date, however, no humans with CPE defects have been described.

Melanocortin 4 Receptor deficiency

Of the five known melanocortin receptors, the melanocortin 4 receptor has been most closely linked to the control of energy balance in rodents. Mice homozygous for a deleted MC4 receptor become severely obese; heterozygotes have a body weight intermediate between wild type and homozygote null animals (82). In 1998, two groups reported families in which heterozygous mutations in the MC4 receptor were associated with dominantly inherited obesity (83, 84). Since then, heterozygous mutations in MC4R have been reported in obese humans from different ethnic groups (85, 86). In a study of 500 severely obese probands approximately 5% were found to harbour pathogenic mutations in the MC4R gene, thus mutations in the MC4R appear to be the commonest monogenic cause of obesity thus far described in humans (87). Recent studies in the Danish population provide an important indication of the likely population prevalence of this disorder. Pedersen and colleagues showed that in Danish men recruited to the draft board, 2.5 percent of those with a BMI > 30 had a pathogenic mutation in MC4R (87.1) confirming that MC4R deficiency is the commonest obesity syndrome described to date and is one of the commonest monogenic  diseases.The fact that obesity is expressed in heterozygotes and the lack of any apparent effect of the mutations on reproductive function are factors which probably contribute to the maintenance of a reasonably high disease frequency. More recently, a small number of homozygotes for MC4R mutations have been described, however, the heterozygotes in these families do have an intermediate phenotype consistent with a co-dominant mode of inheritance.”

The fact that obesity is expressed in heterozygotes and the lack of any apparent effect of the mutations on reproductive function are factors which probably contribute to the maintenance of a reasonably high disease frequency. More recently, a small number of homozygotes for MC4R mutations have been described, however, the heterozygotes in these families do have an intermediate phenotype consistent with a co-dominant mode of inheritance.

Detailed phenotypic studies of patients with melanocortin 4 receptor mutations reveal that this syndrome is characterised by an increase in lean body mass and bone mineral density, increased linear growth throughout childhood, hyperphagia and severe hyperinsulinaemia (87). These features are similar to those seen in MC4R knockout mice (Figure 4), suggesting the preservation of the relevant melanocortin pathways between rodents and humans.While at present there is no specific therapy for MC4R deficiency it is possible that subjects heterozygous for MC4R mutations might respond to pharmacological therapy with small molecule MC4R agonists that would increase signalling from the intact copy of the MC4R. 

Recent studies in the Danish population provide an important indication of the likely population prevalence of this disorder. Pedersen and colleagues showed that in Danish men recruited to the draft board, 2.5 percent of those with a BMI > 30 had a pathogenic mutation in MC4R (87.1) confirming that MC4R deficiency is the commonest obesity syndrome described to date and is one of the commonest monogenic  diseases.The fact that obesity is expressed in heterozygotes and the lack of any apparent effect of the mutations on reproductive function are factors which probably contribute to the maintenance of a reasonably high disease frequency. More recently, a small number of homozygotes for MC4R mutations have been described, however, the heterozygotes in these families do have an intermediate phenotype consistent with a co-dominant mode of inheritance.”

Figure 4. Effects of MC4R deficiency in rodents and humans

TrkB deficiency
A single patient with a denovo mutation in the neurotrophin receptor TrkB has been reported in association with severe hyperphagia and obesity, delayed speech and language development, impaired short term memory and loss of nociception (87.2).

Genetic defects of as yet uncertain significance

Mutations in a number of other genes have been found in association with severe obesity in a small number of individuals, however, often the significance of these findings remains unclear as segregation studies are either not performed or are inconclusive and data on in vitro function or function of aberrant molecules in rodents is lacking.

Two groups have found missense mutations in the cocaine- and amphetamine-regulated transcript (CART), a neuropeptide implicated in the control of feeding behaviour in rodents (88). A Ser66Thr mutation was found in heterozygous form in 2 unrelated UK probands but did not co-segregate with obesity in family studies (89). In an Italian study, the Leu34Phe CART mutation was identified in the heterozygous state in a 10-year-old obese boy and a number of obese family members, but no functional data was provided (90).

Holder and colleagues studied a girl with early-onset obesity and a balanced translocation between 1p22.1 and 6q16.2 (91). The child displayed an aggressive, voracious appetite, and the obesity was thought to be due to increased energy intake, as measured energy expenditure was normal. The translocation did not appear to affect any transcription unit on 1p, but it disrupted the SIM1 gene on 6q. The Drosophila single-minded (sim) gene is a regulator of fruit fly neurogenesis and in the mouse Sim1 is expressed in the developing kidney and central nervous system and is essential for formation of the supraoptic and paraventricular (PVN) nuclei which express the melanocortin-4 receptor (92). It could thus be hypothesized that haploinsufficiency of SIM1, possibly acting upstream or downstream of MC4R in the PVN, was responsible for severe obesity in this patient.

It is of note that in all the genetic syndromes thus far described it appears that the major physiological perturbance is in appetite and energy intake (93). To date, there have been no compelling descriptions of obese humans with primary disorders of metabolic rate or nutrient partitioning. One possible exception is the description of three German subjects with mutations in the N terminus of the nuclear hormone receptor PPAR gamma, an important determinant of adipogenesis (94). These subjects all have a mutation, which interferes with a negative regulatory site in the molecule. Receptors containing this mutation are more powerful inducers of adipogenesis when transfected into cultured cells. Unfortunately, however, there is no information available regarding the co-segregation of this mutation with obesity in pedigrees and as this mutation, to date at least, appears to be unique to the original population it has not been possible to test this issue in independent families.

GENETIC STUDIES IN POLYGENIC OBESITY

Genetic influences are not confined to the extremes of obesity, but exert their effect across the whole range of body weight and at a population level are consistent with the polygenic inheritance of fat mass. The genetic determinants of inter-individual variation in body fat mass are likely to be multiple and interacting, with each single variant producing only a moderate effect. Because of this complexity, the search for genes predisposing to common obesity has been a challenging undertaking (95).

The strategies that have been, and are currently being, undertaken in the search for genetic variants that underlie both common obesity and the normal population variance in fat mass vary in a number of key elements.

1. The outcome variable measured. Many studies rely on the relatively crude, anthropometric measure of BMI while others attempt to measure fat mass and distribution more directly through techniques such as skin fold thickness, bioelectrical impedance, DEXA, isotope dilution and MRI scanning. Some studies either focus entirely on, or incorporate " intermediate phenotypes" in their analysis (reviewed in (96)). Such traits have the theoretical advantage, at least true for certain candidate gene studies, that they may be more proximally related to the function of the gene under study. Thus, a gene which influenced beta adrenergic receptor function might be easier to identify if one studied resting metabolic rate as the outcome variable. However, even though such a gene might influence fat mass in a population, in a study focusing only on BMI, a much greater number of subjects might be needed to see an effect as it would be diluted by multiple other influences on the latter phenotype. Intermediate phenotypes frequently used include resting metabolic rate, respiratory quotient, insulin sensitivity, and plasma leptin levels.
2. Use of obesity/non obesity as a dichotomous variable or adiposity as a continuous trait. To some extent, the utility of these two approaches depend on what one is interested in finding, and whether the genetic determinants of fat mass around the 'normal" range are likely to be the same as those that determine the extremes. To a large extent these types of approach can be viewed as complementary
3. Association vs linkage. The relative merits and difficulties of these approaches are discussed in more detail below.
4. The use of inbred vs outbred populations. Human populations that are geographically or culturally isolated and descended from a relatively small founder population exhibit greater genetic homogeneity and linkage disequilibrium, both of which can increase statistical power. Populations currently under study for obesity-related traits include Pima Indians, Nauruan and Kosrae islanders, Icelanders, Finns, and the Old Order Amish. Disadvantages of this approach include the possibilities that findings will not be generalisable and the risk that, in certain populations, "obesity predisposing' variants will be so highly prevalent as to make their detection problematic.

Association studies

Genetic association studies assess correlations between genetic variants at a polymorphic site and a phenotype/trait of interest. Such variants can either be directly involved in disease predisposition or indirectly involved through linkage disequilibrium with pathogenic variants in close proximity. Association studies may either be population- or family-based. The latter strategies overcome the problem of population stratification (see below) by using transmission vs non-transmission of parental alleles as the comparator e.g. the Transmission Disequilibrium Test (TDT) (methodology reviewed in (97)). However, methods such as TDT are reliant upon the availability of parental genotypes, which are often difficult to obtain in families with late onset diseases such as obesity. Recently, a number of modifications to the classical TDT approach have been developed for multi-allelic markers, multiple siblings and quantitative traits (reviewed in detail (97-99)).

To date, association studies have largely been restricted to candidate genes whose dysfunction might reasonably be expected to result in obesity by virtue of their having putative effects on energy intake, energy expenditure or nutrient partitioning (Table 2 for overview). This strategy has been extensively utilised in obesity genetics but brings with it a number of serious problems. These include 1) "Population stratification" i.e. the uncertainty as to whether individuals with and without disease or from upper and lower centiles of a distribution differ in genetic background and therefore that any association with a polymorphism might be due to its association with that genetic background and not with the trait under study. 2) Many association studies have been underpowered, some grossly so and 3) There is a publication bias towards the reporting of positive rather than negative associations which tends to exaggerate the true nature or strength of an association.

Many of these problems are exemplified by a common polymorphism in the ß3-adrenergic receptor, where despite over 40 association studies, involving more than 7,000 subjects, the findings have been markedly inconsistent. A comprehensive and updated reference for all association studies in obesity genetics is available in the form of the obesity gene map established by Bouchard, Chagnon, Perusse and colleagues at The Pennington Biomedical Research Centre (link to http://www.obesite.chaire.ulaval.ca/genemap.html)

Table 2. Association studies in obesity - candidate genes for energy intake , energy expenditure , nutrient partitioning and body fat distribution
Gene	Name	Location	N	Phenotype	P value	Reference
LEPR	Leptin receptor	1p31	20	Percentage of fat	0.003	(125)
			308	Fat-free mass (FFM) in subjects with BMI = 27 Ng/m2	0.03 to 0.05	(126)
			130	Extreme obesity in children	0.02 to 0.04	(127)
			502	BMI, fat mass (FM)	0.005 to 0.03	(128)
			267	BMI, abdominal sagittal diameter	0.041 and 0.046	(129)
			62	Total and subcutaneous abdominal fat	0.03 to 0.04	(130)
			118	BMI	0.01 to 0.04	(131)
			179	Weight loss in overweight women	0.006	(132)
			220	BMI, FM, leptin in postmenopausal women	0.0001 to 0.009	(133)
POMC	Pro- opiomelanocortin	2p23.3	337	Leptin in Mexican Americans	0.001	(104)
			75	Leptin in obese children	<0.025	(134)
GHRL	Ghrelin	3p26-p25	192	Obesity in women	<0.05	(135)
NPY5R	Neuropeptide Y5 receptor	4q31-q32	74	Obesity (BMI 40 to 47 kg/m2) in Pima Indians	<0.05	(136)
CART	Cocaine- and amphetamine- regulated transcript	5q	612	WHR in men	0.0021	(89)
MC4R	Melanocortin 4 receptor	18q22	156	FM, percentage of fat, FFM in women	0.002 < p < 0.004	(137)
CCKAR	Cholecystokinin A receptor	4p15.2-p15.1	1296	Percentage of fat, leptin	0.003 to 0.041	(138)
ADRB2	Adrenergic ß-2-, receptor	5q31-q32	140	BMI, FM, fat cell volume	0.001 < p < 0.009	(139)
			508	BMI in Japanese subjects	0.001	(140)
			836	BW, BMI, waist circumference, hip circumference, WHR	<0.002	(141)
ADRB2	Adrenergic ß-2-, receptor		574	BMI in Japanese subjects	0.009 to 0.003	(142)
			277	BMI in Japanese men	0.004	(143)
			826	Obesity, BMI, waist circumference, hip circumference, WHR	0.05 to 0.01	(109)
			224	BMI > 35 kg/m2 in men	0.01	(144)
			284	Leptin	0.033	(145)
PPARG	Peroxisome proliferative activated receptor-g	3p25	820	Leptin in obese subjects	0.001	(146)
			333	BMI	0.03	(147)
			752	DBMI in obese men	0.002 to 0.008	(148)
			141	Body weight (BW), BMI, lean body mass (LBM), FM, waist and hip girths	0.002 to 0.05	(149)
			375	Severe obesity with early onset	<0.05	(150)
			921	BMI, waist girth, leptin in Mexican American	0.015 to 0.028	(151)
			838	BW, height, BMI waist girth	0.002 to 0.04	(152)
			619	BMI	0.035	(153)
			119	10-year weight gain	0.009	(154)
GRL	Glucocorticoid receptor	5q31-q32	51	Abdominal visceral fat in lean subjects	0.003 < p < 0.007	(155)
			262	BMI, WHR, abdominal saggital diameter, leptin	0.001 to 0.039	(156)
			135	WHR in men	0.01	(157)

Linkage analysis

In linkage analysis, regions of the genome are identified that co-segregate with disease in families. This has been an enormously powerful technique for the identification of gene defects causing monogenic disorders. Its utility in identifying chromosome regions containing susceptibility genes for complex disorders is less certain, given the lack of clear patterns of inheritance and the multiple genetic and environmental influences on complex traits. In general, linkage analysis has the great advantage over association studies that novel and unexpected genes may ultimately be identified through the recognition of a subchromosomal region that co-segregates with disease or with a quantitative trait. This aspiration has been the basis for the numerous genome wide scans that have been applied to obesity or obesity-related related traits. In a genome wide scan, linkage analysis is conducted using a series of anonymous polymorphisms, spaced at relatively constant intervals over the entire genome e.g., ~350 to 370 markers with an average spacing of 10 cM. Genome scans are complicated by the fact that instead of a single test for linkage, one must conduct multiple tests across the entire genome. In light of this, it has been proposed that a LOD score =3.3 can be taken as strong evidence of linkage and a LOD score =1.9 but <3.3 as evidence suggestive of linkage (100).

Results from reported genome-wide linkage studies that have examined obesity and/or related intermediate traits have identified several loci that show positive evidence for linkage with a LOD score > 2.6 (Table 3). In two studies, one of extended, Mexican-American pedigrees (101) and the other of French sibling-pairs (102), significant linkage of serum leptin levels to chromosome 2p21 was found. In the Mexican American study, suggestive evidence of linkage of fat mass to 2p21 was reported while in the French study no hint of linkage to body mass index was identified. The potential importance of this locus is supported by a study of African-Americans which confirmed linkage with serum leptin levels in this population (103). This region of human chromosome 2 includes the pro-opiomelanocortin (POMC) gene, in which loss-of-function mutations have been demonstrated as rare Mendelian causes of obesity in humans (75). Hixson et al have directly examined extended haplotypes of single nucleotide polymorphisms around the POMC gene and have found a highly significant association of these haplotypes with serum leptin in Mexican Americans (104). Other than the locus on 2p there are no other loci which have been highlighted in more than one of the genome wide scans reported to date.

Gene-environment interactions

Implicit to the susceptible-gene hypothesis is the role of environmental factors that unmask latent tendencies to develop obesity. Susceptibility to obesity may thus be determined largely by genetic factors, but phenotypic expression is determined by the environment. For example, in a population where caloric availability is limited, individuals with high genetic susceptibility might have a relatively higher degree of adiposity, but the mean level would still be relatively low. However, in the presence of high-fat or high-calorie diets, the overall distribution of adiposity would be shifted and high genetic susceptibility would lead to morbid obesity.

The ability to identify possible interactions between genes and environmental factors is difficult because there may be a delay in an individual's exposure to an 'obesogenic' environment and uncertainty about the precise timing of the onset of weight gain. Polymorphisms in obesity candidate genes have been studied in a few population based cohorts on whom extensive and detailed information on diet, physical activity and markers of intermediate metabolism have been measured. The relationship between the Pro12Ala variant in the nuclear receptor peroxisome proliferator-activated receptor-gamma (PPARgamma) (105-107) and the ratio of dietary polyunsaturated fat to saturated fat (P:S ratio) has been studied and there is some evidence for a gene-nutrient interaction (108). Evidence for gene-exercise interactions have been found for variants in lipoprotein lipase in the HERITAGE Family Study and for the Gly16Arg variant in the beta (2)-adrenergic receptor in a French and UK population based cohorts (109-111). These studies, although few in number, emphasize the difficulty of examining the effect of common polymorphisms in the absence of data on non-genetic exposures, and may explain in part the heterogeneity of findings in previous studies.

FUTURE PERSPECTIVES AND STRATEGIES FOR OBESITY GENETICS

A major goal of identifying genes responsible for human obesity in the general population is to provide more rational approaches to therapy, either by elucidating the underlying pathophysiology or by stratifying patients into groups in which the effectiveness of different treatments can be determined empirically. The success of this approach has been limited, in part because of the underlying complexity, and in part because there may be a large number of genes with relatively small effects.

Recently, Risch compared the statistical power to detect multi-factorial genetic effects in linkage and association studies (112, 113). His power calculations highlighted the advantages of the association design in terms of sample size and suggested that the power of association analysis to detect genetic contributions to complex disease can be much greater than that of linkage studies. In the next few years, dense SNP maps across the genome, high throughput genotyping technologies and simultaneous comparisons of groups of loci and statistical measures for assessing genome wide significance may overcome some of the original limitations of the classical association approach (114). The availability of large sets of SNPs throughout the genome, such as those identified by the SNP Consortium (link to http://snp.cshl.org/), will permit linkage disequilibrium mapping to be an increasingly important component for regional and whole-genome association strategies (115) and with the recent compilation of the draft human genome sequence (see http://www.ncbi.nlm.nih.gov/Tour/), genome-wide studies of association in human diseases such as obesity may become a realistic possibility (116).

The detailed study of humans with morbid obesity due to specific monogenic defects should continue to shed light on the molecular mechanisms underlying the regulation of human appetite, body weight and growth and in mediating these and other biological effects that are clearly closely linked. If underlying genetic differences converge on common pathophysiological pathways, application of gene or protein expression profiling may provide additional resolution with which to define rational treatment strategies (see Chapter 21).