Diabetes Mellitus: A Fundamental and Clinical Text
3rd Edition
© 2004 Lippincott Williams & Wilkins
←↑→
55
Genetics of Human Obesity
Oluf Pedersen
Søren M. Echwald
Obesity is a common disorder, affecting about 25% of individuals in many postindustrial societies. The prevalence of obesity is increasing worldwide, and during the last 30 years the percentage of obese and overweight people in the United States and Europe has increased dramatically among adults and similarly among children despite a growing awareness of the health risks associated with obesity (1,2).
Obesity-related diseases like type 2 diabetes, arteriosclerosis, hypertension, and osteoarthritis are a cause of excess mortality and morbidity, and decreased quality of life, especially among middle-aged and elderly people (3). The severity of complications is proportional to the degree of obesity. Considering also the significant contribution to expenses on health care, obesity warrants a strong focus on research in pathogenesis, treatment, and prevention.
The pandemic of obesity has arisen with the ready availability of a large variety of inexpensive and highly palatable foods and diminishing daily physical activity, which are essential features of a contemporary lifestyle in many parts of the world. However, the factors causing this imbalance between energy intake and energy expenditure are more complex than they may seem from this simple observation and are largely unknown. It has been demonstrated that obese subjects actually eat more food, especially fat, rather than just having a reduced metabolic turnover, suggesting that altered behavioral aspects may be part of the pathogenesis rather than a result of the obese state (4,5). Also, intervention such as food restriction against obesity is hampered by a high recurrence rate (6). This might be caused in part by an inherited or acquired set point for body weight regulation, deviations from which may be effectively counterregulated.
As a phenotype, obesity is heterogeneous, and there are at least two distinct but frequently overlapping subtypes: general obesity, which results in increased body fat mass, particularly in subcutaneous tissues, and visceral/android obesity, with fat accumulation especially in the abdominal cavity. These subtypes have different physiologic, clinical, and prognostic implications. The phenotypes seem to have some of the same genetic and environmental influences in common (7,8).
It will be obvious to most clinicians dealing with obesity that this is a disorder that runs in families. Obese children tend to have obese parents, and often obesity is a family characteristic. However, in order to establish to what degree the familiality of obesity is due to shared environment or shared genes, careful and systematic investigations of the nature of the inheritance of obesity has been necessary.
This chapter discusses evidence for an inherited component in the pathogenesis of obesity along with the approaches to identify obesity genes and current knowledge of the genetic elements of syndromic, monogenic, and apparently polygenic forms of human obesity. Likewise, the potential of genetic studies to give insights into novel pathways of appetite and body weight regulation as well as to provide clues for new pharmaceutical treatments of obesity is discussed.
Heritability of Obesity
There is no doubt that obesity is strongly influenced by environmental risk factors such as nutrient intake and sedentary lifestyle (7). The prevalence of obesity increases so rapidly in many populations that the changes cannot be attributed to changes in inheritance. There are differences in prevalence of obesity between populations and between various groups within populations, and these differences are in many cases closely associated with environmental factors, especially social and behavioral factors (2). The degree of obesity in one individual can be modified by interventions that alter that person’s energy intake or expenditure. However, it is beyond question that obesity is also influenced by genetics.
Many family studies have been conducted to estimate the heritability (or familiality) of common obesity features such as variations in fat mass or body mass index (BMI) (9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25). In general, all studies identify a considerable percentage of the phenotypic variation being attributable to genetic risk factors. It is important to note, however, that what are referred to as heritability studies do not refer to actual molecular biology, but rather to statistical analysis of variances and covariances of phenotypes, and their distributions among relatives. Consequently, these family studies do not provide actual evidence of specific modes of inheritance but rather present statistical answers to whether a given phenotypic distribution lends support to a theoretical model of inheritance. Commingling analyses (i.e., the splitting of phenotypic measures into subcomponent distributions) show uni-, bi-, or trimodal phenotypic distributions that can be explained
P.828

by the segregation of one or more major genes but also by effects of major environmental influence or by threshold traits. The relevance of these analyses must await the identification and analysis of causative gene variants and their segregation with obesity, for which novel data now are surfacing.
An advantage of using families for studies of heritability is that the combination of several types of relatives provides the possibility of testing more complex segregational models. The use of several types of relatives may also reduce the bias of shared environment. The consensus findings from family studies are that general heritability of BMI may account for 30% or more of the phenotypic variation (or h2 = 0.3) and that the segregation of the phenotype most probably is described by an oligo- or polygenic model of inheritance (9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25).
Studies of twins show the highest degrees of heritability of the obese phenotype compared with other studies. Overall, results from twin studies give estimates of heritability from 40% up to 80%, with variations on the effect of age and gender from different studies (26,27,28,29,30,31,32). Also based on twin studies, it has been claimed that the heritability of BMI is most marked during late childhood and adolescence (33). Interestingly, the gene determinants for weight regulation may not be the same throughout life. Thus, it has been shown that only about 40% of the genetic risk factors that control BMI at age 20 years are still operative 28 years later (34).
It should be noted, however, that twin studies derive their heritability estimates from comparison of the differences in phenotypic variations between monozygotic and dizygotic twins, who share 100% and 50% of their genetic material, respectively. If it is assumed that a fraction of obesity susceptibility genes interacts with each other with regard to the phenotypic expression, obese monozygotic twins are likely to be relatively more identical than individuals sharing less identity due to the additional variation arising from gene-to-gene interactions. This might explain the lower estimates of heritability when more distantly related family members sharing 25% or less genetic identity are included in the segregational models. One could hypothesize that although relatives may share 50% or 25% of genes, they may not share a similar amount of the gene–gene interactions. Although this is a tentative interpretation of the data on heritability, it may imply that the concepts of heritability versus genetic influence should be kept apart. Therefore, as illustrated by studies of twins, genetic influence may, in fact, explain up to 80% of the variation in obese phenotypes such as BMI, although heritability in general based on family studies is on the order of 30%.
Intervention studies (35), in which twin pairs were put on a high-calorie diet, showed a high degree of heritability in the capacity to gain weight, given the same food challenge. Overall, it seems that not only the cross-sectional weight measures show heritability, but so does the propensity for weight change. However, the limitation in twin studies is that it is difficult to estimate how much of the similarity in a given trait is due to shared environment.
Adoption studies and studies of twins who are not reared in the same environment are approaches for investigating the impact of shared environment compared with genetic determinacy. Several of these studies concern the same twin samples, but analyzed differently (36,37,38,39,40,41,42,43,44). A recurrent finding in studies of adoptees and twins reared apart is that not only shared rearing environment but also inherited factors have significant impact on the phenotypic variation in BMI. Correlations between variations in BMI in biologic parents versus foster parents and biologic siblings versus adopted siblings show that a large part of the variation in BMI is explained by the biologic relatives (i.e., shared genes) and not by the adopted/foster relatives. The remaining environmental variation is thus explained by the effects of adult environment. Estimates of heritability from these studies range from 20% to 60%. However, applying complex segregational models using information from many types of relatives (43), adoption studies show estimates of heritability on the order of 30%, in line with results from family studies.
To sum up, the widespread forms of obesity show a level of heritability between 30% and 80%. There is evidence of the major inherited component being either mono- or polygenic in nature and the various obese phenotypes are hypothesized to be modified by an individual repertoire of variation within a pool of polymorphic genes—each with a small effect—that interact with the major susceptibility genes. Eventually, various environmental pressures are thought to elicit the phenotypic expression. Still, the accurate evidence for these genotype–phenotype interactions in common forms of obesity is weak due to our modest knowledge about specific obesity risk gene variants. However, cases of rare single gene obesity have recently been discovered and will be discussed. Characteristically, for monogenic obesity the correspondence between genotype and phenotype is close to 1:1.
Approaches Taken to Dissect the Genetics of Obesity
The genetic variants that need to be identified in widespread forms of obesity are likely to be common, to be present in both obese and nonobese subjects, to be associated with a variable degree of increase in individual risk, and in most cases possibly to interact in a complex and nonlinear way with other genetic and environmental susceptibility factors. To cope with these challenges in the dissection of obesity genetics, a variety of complementary approaches is being applied, including the biologic candidate gene approach and the random genome mapping and linkage analyses (genome screens) of dichotomous or quantitative traits applied on extended pedigrees or sibling pairs and most recently whole-genome single nucleotide polymorphism (SNP) association studies (reviewed in reference 45). Genome screens are a top-down strategy, linking complex phenotypes with genomic markers, whereas the conventional candidate gene approach is a bottom-up strategy, which takes known genes as a starting point and then works up to the phenotype. In all cases, large well-characterized patient populations involving family collections and obesity-related phenotypes are required.
Biologic Candidate Gene Approach
With the biologic candidate gene approach, a number of obese subjects are investigated for the presence of mutations or polymorphisms
P.829

in regulatory, untranslated, and coding regions of genes that, when mutated, are likely to display the phenotype of interest based on available information. Obesity candidate genes include genes encoding proteins involved in adipocyte differentiation, eating behavior, and energy expenditure. Once a candidate gene has been investigated for mutations, the impact of identified mutations is investigated in a large number of individuals to determine whether the genetic variant is likely to confer the obese phenotype. The term association studies refers to studies of case control design, where the prevalence of the variant of interest in subjects carrying the phenotype (i.e., are obese) is compared with its prevalence among age-, gender-, and population-matched control subjects. When applying the candidate gene method, a thorough knowledge of the pathophysiology of obesity is necessary in order to choose relevant genes for investigation, and only genes already discovered are eligible for detection of genetic variation. Genotype–phenotype interaction studies are comparisons of the obesity-related phenotype of interest (e.g., waist circumference) among subjects characterized by having different genotypes of a genetic variant. By haplotype association studies, a number of polymorphisms can be compared together, because the linkage disequilibrium (LD) present between closely spaced polymorphisms will cause these to be inherited as patterns. Creating haplotypes is a computationally demanding analysis, because the most likely haplotype constellations usually will have to be computed from genotype data. The advantage of haplotype association analysis is that it includes combinations of variants in patterns that may interact to determine functional variations that are involved in phenotypic variation. The traditional case control design is sensitive to unintended differences in population stratification, a problem that can be circumvented by using family-based association tests, which at the same time provide estimates of association and linkage.
Genome-Wide Screens
Another way for identifying obesity susceptibility genes is the application of random genome mapping. This approach requires no a priori hypothesis with regard to which kind of gene can confer phenotypic variation in opposition to the candidate gene approach. The basic method is a comparison of affected and genetically related subjects in order to identify regions of the genome in which there is excess allele sharing (identity by descent) compared with the random expectation. The investigated subjects are pedigrees or affected sibling pairs, but other designs can be used equally well. Polymorphic microsatellite markers are applied to estimate the regions of the genome that are shared between subjects, and for a genome-wide scan markers are often spaced every 10 centimorgans (cM), giving 350 to 400 markers to genotype in each subject under investigation. Subsequent computational linkage analysis allows identification of those chromosomal regions that show statistically significant cosegregation with disease and are thus likely to harbor susceptibility genes. However, this locus identification only represents the first step in the gene hunt. The locus arising from a genome scan typically covers about 20 cM, which is likely to contain about 20 million nucleotides and 200 genes or more. From here at least two different paths can be taken. One is the so-called LD mapping to narrow down the broad chromosomal peak that shows linkage to obesity to a few centimorgans and from there to perform positional cloning, which still is a huge enterprise. The second manner of attack concentrates on biologically relevant candidates in the locus taking advantage of bioinformatics, including results of tissue expression profiling. Obviously, the two complementary strategies are often used simultaneously.
Besides being used in many studies of common human obesity, this random genomic-based approach has been successfully applied to unravel the specific genetic defects in several forms of obesity in mice that segregate as mendelian traits.
Quantitative trait loci (QTL) analysis is a linkage approach that compares a variable phenotype (e.g., percentage of body fat) with the allele sharing between related subjects. This approach, which complements analysis of dichotomous traits, has been used extensively to map susceptibility loci in rodent models. In rodents, a QTL study is usually performed on a backcross of the F2 generation of two inbred strains divergent for the trait of interest, and this increases the possibility that the strains have different alleles in genes controlling the trait. The inbred background of the strains minimizes genetic variance. Rodent QTLs linked to obesity can be extended to the syntenic regions of human chromosomes, where already mapped genes or expressed sequence tags can be examined for mutations.
The availability of the drafts of human and several animal genomes has opened for the possibility of identifying obesity-causing genes in silico by the positional information available in databases. This is commonly referred to as the positional candidate gene strategy.
A promising new tool to unravel genetic susceptibility to obesity is the application of whole-genome association studies based on high throughput genotyping of, for instance,100,000 SNPs in or close to every known gene in pooled or individual DNA samples from obese and carefully matched nonobese subjects.
Obesity Syndromes with a Mendelian Mode of Inheritance
There exist a number of human pleiotropic syndromes with obesity as one of the clinical features, such as Prader-Willi syndrome, Bardet-Biedl syndrome, Cohen syndrome, Borjeson-Forssman-Lehman syndrome, and Wilson-Turner syndrome, among others. Applying cytogenetics and random genome mapping approaches to several of these disorders, segregating chromosomal locations have been determined and can be accessed through the Online Mendelian Inheritance of Man (OMIM) database (http://www3.ncbi.nlm.nih.gov\omim\).
The syndromes segregate as either autosomal-dominant or -recessive traits, some of which are X linked. Syndromic obesity is obviously distinct from widespread cases of obesity, but less severe genetic variation at these loci might increase the risk for more common obesity. When testing a total of 17 chromosomal markers (46) in a sibling pair analysis of 207 siblings from 44 families segregating for extreme obesity, the obesity phenotype
P.830

did not segregate with any of the loci for a number of the known human obesity-related syndromes (Prader-Willi, Bardet-Biedl, Cohen, Borjeson-Forssman-Lehmann, or Wilson-Turner).
Recently, causative genes have been mapped in three distinct subtypes of the Bardet-Biedl syndrome (BBS) (reviewed in reference 47). BBS6 is caused by mutations in the MKKS gene, which presents homologies with a bacterial chaperonin. BBS2 and BBS4 genes encode for proteins with yet unknown functions. Thus, in no case of syndrome-associated obesity has a clear mechanistic link between the product of the mutant gene and disordered energy balance been identified so far.
One lesson from these syndromes, however, is the recurrent finding of mental retardation and obesity in the same affected person. These findings point to the role of the central nervous system in controlling human adiposity, and important information of the regulation of satiety and energy expenditure may follow the cloning and characterization of more genes underlying syndromic obesity.
Insights into Appetite Regulation Derived from Studies of the Genetics of Obesity in Rodents
The genes for most of the known inbred models of obesity in rodents have been identified by genome mapping and positional cloning or positional candidate gene strategies (Table 55.1). In addition, biologic candidate genes for obesity have been investigated by generation of single gene knockout or transgenic mice. The first obesity-linked locus found in the mouse contained the agouti gene, in which dominant mutations confer an obese phenotype with increased linear growth and yellow fur (48,49,50). Other mouse mutations are recessive and lead to complex phenotypes associating obesity with hormonal and metabolic abnormalities, including the obesity (ob) (51), diabetes (db) (52,53), fat (fat) (54) mouse or sensory dysfunctions like in the tubby mouse (55). The melanocortin-4 receptor gene (mc4r) and proopiomelanocortin gene (pomc) knockouts (56,57) produced obesity with a pattern similar to mutations in the agouti gene.
Table 55.1. Summary of eight known rodent strains with monogenic forms of obesity, with indication of dominant/ recessive inheritance, and the gene defect shown to underlie the obese phenotype
Strain   Inheritance Encoded Protein Defect
Obese (ob) Mouse Recessive Leptin Stop codon/promoter defect in leptin
Diabetes (db) Mouse Recessive Leptin receptor Defect lepR
Agouti yellow (aγ) Mouse Dominant Agouti Ectopic expression of melanocortin receptor antagonist
Tubby (tub) Mouse Recessive Phosphodiesterase Apoptosis in the brain?
Fat (fat) Mouse Recessive Carboxypeptidase E Carboxypeptidase E activity abolished
Zucker/fatty (fa) Rat Recessive Leptin receptor Defect lepR
Koletsky (kol) Rat Recessive Leptin receptor Defect lepR
Corpulent (cp) Rat (SHR/NIH-cp or LA/N-cp) Recessive Leptin receptor Defect lepR
These pioneering studies of obesity genetics in rodents have given exciting insights into key molecular mechanisms operating predominantly in hypothalamus and adipose tissue to regulate energy balance. It has turned out that all the proteins encoded by the above-mentioned genes are part of the same pathway regulating eating behavior. For instance, cloning of ob and db genes has provided the molecular basis for an understanding of how leptin, a cytokine-like protein secreted by adipose cells, serves as the afferent mediator that controls energy balance through behavioral and metabolic effectors (58,59,60,61). Leptin appears to act as an endocrine and a paracrine factor and perhaps also as an autocrine factor, and has a series of roles: as a growth factor in a range of cell types, as a permissive factor for puberty, and as a signal of metabolic status and modulation between the fetus and the maternal metabolism. In all of these interactions, its role is to interact with other hormonal mediators and regulators of energy status and metabolism such as insulin, glucagon, the insulin-like growth factors, growth hormone, and glucocorticoids.
Another important piece of knowledge has come from the discovery of the genetic defect in the agouti mouse. Subsequent transgenic studies revealed that obesity in these Ay mutant mice reflects the ability of the agouti protein, which normally is expressed in melanocytes but which in these mice is abnormally expressed in the brain, to mimic the neuropeptide agouti-related protein (Agrp). The latter protein is normally expressed in the hypothalamus and signals through the central nervous system specific melanocortin-4 receptor (MC4R) (62,63,64). Agouti protein and Agrp are antagonistic ligands for melanocortin receptors.
P.831

The endogeneous agonist ligands for these receptors are the anorexigenic peptide α-melanocyte–stimulating hormone (α-MSH) and the adrenocorticotropic hormone (ACTH), which are produced by cleavages of the precursor proopiomelanocortin (POMC) by the proteolytic enzymes PC1 and carboxypeptidase E. Thus, α-MSH influences energy expenditure and food intake by activation of MC4R in the hypothalamus. The genetic predictions of these relationships have been borne out by transgenic experiments: gain-of-function Agrp mutations produce an obesity phenotype similar to that displayed by loss-of-function mutations in pomc or mc4r. Taken together, eating behavior is now viewed as controlled by homeostatic mechanisms whereby hormonal signals from adipose tissue send feedback to hypothalamic circuitry to regulate appetite (65,66,67,68,69,70,71,72,73,74) (Fig. 55.1).
Figure 55.1. Model of homeostatic circuit regulating energy balance through the melanocortin 4 receptor (MC4R). Increased adiposity leads to increased leptin production in fat tissue. Leptin stimulates neurons in the arcuate nucleus of the hypothalamus that coexpress the anorexigenic hormones α melanocyte-stimulating hormone [α-MSH, a cleavage product of proopiomelanocortin (POMC)] and cocaine- and amphetamine-regulated transcript. Leptin also inhibits neurons in the arcuate nucleus that coexpress the orexigenic hormones agouti-related protein and neuropeptide Y. The neurons in the arcuate nucleus project to other regions of the hypothalamus (including the paraventricular nucleus and the lateral hypothalamic area–parafornical area), where α-MSH binds to its receptor, MC4R, resulting in an upregulation of anorexigenic effectors such as corticotropin-releasing hormone (CRH) and thyrotropin-releasing hormone (TRH) and a downregulation of orexigenic effectors such as melanin-concentrating hormone (MCH) and orexin. Agouti-related protein acts as an antagonist of MC4R.
Monogenic Obesity
The landmark discoveries in the genetics of body weight regulation in rodents have inspired intensive gene hunting in subjects with an extreme obese phenotype applying a biologic candidate gene approach, and mutations in genes with homology to the mouse obesity genes or in genes acting in the same metabolic pathways have been identified (Table 55.2). Phenotypic features shared between mice and humans with homologous obesity
P.832

mutations exhibit a fundamental conservation of the underlying biologic pathways. Yet, special characteristics in some of the human cases of obesity caused by single gene mutations show aspects of energy homeostasis that are unique to human biology. In humans, seven genes causing monogenic obesity have been reported to date. These genes fall into two categories. The first group involves the genes coding for leptin, leptin receptor, POMC, and prohormone convertase-1 (PC1) (75,76,77,78,79). Mutations in these genes confer extremely rare cases of recessive obesity associated with multiple endocrine abnormalities having been identified in just a few handfuls of families throughout the world. Obesity causing mutations in MC4R (80,81) and the small heterodimer partner gene (SHP) (82) belong to the second category. MC4R-linked obesity exhibits a codominant mode of inheritance and may account for up to 5% of morbid obesity, whereas obesity due to mutations in SHP is inherited in a dominant fashion and has so far only been reported among Japanese, where it may account for about 6% of early-onset mild obesity. Also, a rare mutation in the peroxisome proliferator–activated receptor-γ2 gene has been reported to cause massive obesity (83). Apart from SHP and PPAR-γ2, the proteins encoded by these genes are functionally related in the central signaling system of energy homeostasis and feeding behavior.
Table 55.2. Comparison of major phenotypic features of monogenic forms of Human obesity
Gene Obesity Birth weight Endocrine abnormalities Hyperphagia Inheritance Chromosome
LEP Severe Normal Low leptin
Hypogonadism
High thyroid-stimulating hormone
High insulin		 + Recessive 7q31.3
LEPR Severe ? High leptin
Pituitary dysfunction
Hypogonadotrophic hypogonadism
Hypothalamic hypothyroidism
Sympathetic dysfunction
High insulin		 + Recessive 1p31
POMC Severe Normal Red hair pigmentation
ACTH deficiency, hypocortisolism
Low α-MSH		 + Recessive 2p23.3
PC1 Severe ? Hypogonadotrophic hypogonadism
Hypocortisolism
High proinsulin, low insulin
Postprandial hypoglycemia
High POMC		 ? Recessive 5q1.5-2.1
MC4-R Severe Normal Not observed + Dominant 18q22
NROB2 Mild High Mild hyperinsulinemia – Dominant 1p36.1
LEP, leptin; LEPR, leptin receptor; POMC, pro-opiomelanocortin; PC1, prohormone convertase 1; MC4-R, melanocortin-4 receptor; ACTH, adrenocorticotropic hormone; α-MSH, α-melanocyte-stimulating hormone.
Obesity Due to Mutations in Leptin and Leptin Receptor Genes
The breakthrough that led to the cloning of the ob and db genes and the characterization of their products, leptin and the leptin receptor, respectively, was based on the classical parabiosis experiments in the obese ob/ob and db/db mouse strains, involving the surgical cross-anastomosis between the circulatory systems of the obese and wild-type animals (84,85). The results of these studies led to the prediction of the existence of a sensory pathway that would report the status of the adipose tissue via a circulating factor (product of the ob gene) and its receptor (product of the db gene).
After being secreted primarily from adipose cells, leptin circulates in the bloodstream and is transported into the central nervous system where it acts on the hypothalamus. The leptin receptor, which is a member of the cytokine receptor superfamily, mediates leptin signaling through activation of the Janus kinase (JAK) and signal transducers and activators of transcription (STAT) members of transcription factors. The receptor is expressed in various isoforms in multiple tissues, but predominantly in the hypothalamic neurons. Rodent strains carrying homozygous ob and db mutations exhibit similar phenotypes, characterized by early onset of obesity, hyperphagia, low core temperature, insulin resistance, and susceptibility to diabetes.
Two human kindreds with defects in leptin have been reported (75,76). Both loss of function and missense mutations in the leptin gene cosegregate with obesity. Homozygous mutations result in low levels of serum leptin, hyperphagia, and onset of obesity in first weeks of life. Much like in the corresponding mouse model, leptin-deficient humans fail to undergo normal puberty due to hypogonadotropic hypogonadism. Heterozygous missense mutations in the leptin gene are likewise associated with obesity and low levels of serum leptin, similar to what is seen in the ob/+ mouse.
One family has been identified with a leptin receptor mutation (77). In the affected homozygous mutation carriers, a truncation of the receptor before the transmembrane domain abolishes leptin signaling, causing a phenotype characterized by
P.833

high levels of circulating leptin, hyperphagia, early onset obesity, no pubertal development, hypergonadotropic hypogonadism, hypothalamic hypothyroidism, growth delay, and subnormal growth hormone response to hypoglycemia. The high circulating level of leptin in both homozygous and heterozygous carriers of the mutation is bound to the truncated leptin receptor, giving rise to a prolonged plasma leptin half-life. Patients with leptin receptor mutations differ from the hormonal-metabolic phenotype of the db mice since they do have normal fasting and postprandial plasma glucose, insulin, and lipid levels as well as a normal hypothalamic–pituitary axis.
Thus, in human obesity cases, the evident difference between those with mutations in leptin compared with those having mutations in its receptor is the presence of significant growth retardation and central hypothyroidism in the carriers of leptin receptor mutation. By contrast, mice deficient for leptin or its receptor display remarkably similar endocrine abnormalities; both exhibit reduced levels of growth hormone and retarded linear growth. One possible explanation for these disparities would be if the human but not the mouse leptin receptor were capable of stimulating a low level of certain hypothalamic-releasing hormones in the absence of leptin, such that loss of function in the receptor leads to a more severe defect than loss of function in the ligand.
Finally, comparison of leptin-signaling defects in mice and humans show additional differences in the hypothalamus–pituitary–adrenal axis; db/db and ob/ob mice exhibit strikingly increased glucocorticoid production that is not seen in any of the human cases. Patients with leptin receptor mutations also differ from the hormonal-metabolic phenotype of the db mice since they do have normal fasting and postprandial circulating levels of glucose, insulin, and lipids.
Obesity Due to Mutations in Proopiomelanocortin
Being the precursor molecule of peptides involving adrenocorticotropin (ACTH) and melanocyte-stimulating hormones (MSH), POMC is another key protein in the leptin–melanocortin pathway. pomc knockout mice show obesity, altered pigmentation, and defective adrenal development (74). These features are similar to the striking phenotypes of the two reported patients who harbor either a homozygous or a compound heterozygous loss-of-function mutation in the coding region of POMC (78). These patients are characterized by the deficiency of pituitary peptides derived from POMC for signaling through several melanocortin receptors. The deprivation of α-MSH leading to the lack of the melanocortin ligand for the melanocortin-4 receptor causes obesity while the absence of the ligand for the melanocortin-1 receptor is responsible for the altered pigmentation and red hair color. Lack of ACTH to stimulate the melanocortin-2 receptor leads to adrenal deficiency.
Obesity Caused by a Mutation in Prohormone Convertase-1
The fat mouse exhibits late-onset moderate obesity (progressively developing between 8 and 12 weeks after birth), hyperproinsulinemia, transient hyperglycemia, and infertility (54). The mutation causing the phenotype is a single missense mutation in the carboxypeptidase E gene (cpe), which eradicates all enzymatic activity of the carboxypeptidase. Carboxypeptidase E is expressed in various tissues, including the brain, pancreas, and adrenal glands, and is involved in the processing of prohormones and neuropeptides, including proinsulin and POMC. It is therefore likely that the obese phenotype observed in the fat mouse is due to insufficient processing of POMC to yield α-MSH.
No sequence variation in the cpe homologue in humans has been related to obesity. Instead, mutations in another processor of POMC, PC1, have been reported to be associated with obesity in a single subject (79). The individual is compound heterozygous for two loss-of-function mutations in PC1, and her phenotype involves moderate adult obesity but extreme childhood obesity, markedly elevated circulating levels of POMC, hyperproinsulinemia, lack of true insulin in serum, impaired adrenal function, and hypogonadism with infertility. Thus, even though carboxypeptidase and PC1 code for nonhomologous genes, their loss of function causes similar metabolic abnormalities, which in both cases involve obesity.
Obesity Caused by Mutations in the Melanocortin-4 Receptor
MC4R is a seven-transmembrane G protein–coupled receptor that activates the cyclic adenosine monophosphate second messenger system. With the discovery of the obese phenotype of the agouti mouse and the antagonism of the MC4R stimulated adenylate cyclase activity by the agouti-related protein, it became obvious that the MC4R is implicated in the regulation of food intake. Even stronger indication of the involvement of the MC4R in the regulation of energy metabolism accumulated when the mc4r knockout mouse was developed (56,57). The mc4r knockout mice show progressive weight gain, increased linear growth, hyperphagia, hyperinsulinemia, and hyperglycemia. The neuroendocrine apparatus comprising growth, thyroid, and reproduction is normal. The mice heterozygous for the null mutation have a phenotype intermediary to that of null mice and wild-type mice, suggesting a dose-dependent response to changes in the MC4R. Thus, the phenotype of mc4r knockout mice strongly resembles the phenotype of agouti mice.
Human monogenic obesity caused by MC4R mutations is so far the most common form of causally explained severe childhood-onset obesity (80,81,86,87,88,89,90,91). Some studies point to a gender difference similar to the results in MC4R mice, where female carriers are more obese than males. However, the prevalence of pathogenic mutations varies considerably among the different studies, with estimates ranging from about 0.5% to about 6%. Mutations in MC4R result in a type of obesity syndrome that is inherited in a codominant manner. However, although all identified homozygous carriers are morbidly obese, not all heterozygous carriers in the families are obese, findings that cannot be explained by functional studies of the specific mutations. These differences in clinical expression may point to the impact of modifying environmental and genetic risk factors in some pedigrees. Thus, such nonobese carriers might have inherited
P.834

alleles of genes protecting them from developing obesity. Mutations leading to complete loss of function are associated with a more severe phenotype, whereas subjects with mutations retaining residual signaling capacity have a less severe phenotype. The correlation between the signaling properties of these mutant receptors and energy intake emphasizes the key role of this receptor in the control of eating behavior in humans. Human obesity caused by MC4R deficiency is characterized by an increase in lean body mass and bone mineral density, increased linear growth, and severe hyperinsulinemia disproportionate to the degree of obesity. The latter may suggest that impaired MC4R signaling could be involved in eliciting hyperinsulinemia through impaired negative neuronal control of insulin secretion. Studies of children with pathogenic MC4R mutations show that their ad libitum energy intake is greatly increased as compared with their unaffected siblings. This finding is consistent with their reported food-seeking behavior in the free-living situation. Interestingly, binge eating (a condition that involves recurrent episodes, i.e., two to four per week, of eating an abnormally large amount of food and feeling of a lack of control over this behavior) appears to be a feature of subjects with mutations in MC4R. However, all subjects with MC4R deficiency, including those who are homozygous for a deletion of MC4R, are reported to have a lower ad libitum food intake than those with leptin deficiency, suggesting that some of the inhibitory effects of leptin on food intake may be mediated by other neuropeptides.
Obesity Caused by Mutations in the Small Heterodimer Partner
Subjects with SHP mutations exhibit no endocrine dysfunction and show relatively modest obesity (82). Thus, these SHP mutations reveal a pathway to increased body weight that is apparently independent of the hypothalamic pathway. This pathway could be related to effects of SHP on one or more of a number of nuclear receptors associated with metabolic regulation. These include the steroid, retinoid, and thyroid hormone receptors, the orphan hepatocyte nuclear factor 4α (HNF-4α), and the PPARs, which play a key role in adipogenesis.
Interestingly, SHP mutations are also associated with increased birth weight (82). Because SHP inhibits HNF-4α, the loss of SHP function should lead to increased activity of HNF-4α, resulting in increased insulin secretion. Because insulin is one of the key hormones in adipogenesis, increased insulin secretion occurring in the fetus might increase intrauterine growth. Increased fetal growth could also underlie the postnatal obesity in the affected subjects. The prevalence of obesity-related SHP mutations, if any, in Western populations is at present not known.
Obesity Caused by Mutations in Peroxisome Proliferator Activated Receptor-γ2
The PPARs are a group of ligand-activated transcription factors that influence numerous biologic processes, including energy metabolism, cell proliferation, and inflammation (reviewed in reference 92). Four different PPAR-γ isotypes are known. The PPAR -γ2 isotype is predominantly expressed in adipose tissue, where it stimulates adipogenesis. It is the target of a group of antidiabetic drugs called thiazolidinediones. In a study of 121 obese, unrelated Germans, four markedly obese subjects were heterozygous carriers of a Pro115Gln mutation in PPAR-γ2 (83). Three of the four obese subjects also had type 2 diabetes. In vitro analyses showed a permanent activation of PPAR- γ2 that led to an accelerated rate of adipocyte differentiation and increased fat accumulation in a tissue-culture model of adipogenesis. Unfortunately, the same mutation has failed to emerge in large-scale follow-up screening studies of obese subjects.
Studies of Common Forms of Obesity
Genes for common obesity are expected to be less deleterious compared with genes causing monogenic obesity and thus are not expected to be selected out of a population. Because the ability of fat accumulation is a crucial factor for survival and fitness in most species, including humans, it is likely that a large part of the genetic variation predisposing for common human obesity has been founded early in the evolution and is widespread in the contemporary population.
The candidate gene approach has proven to be successful for unraveling causes of rare monogenic obesity in humans. In contrast, application of the two mainstream approaches, the candidate gene strategy and the random genomic strategy, to the widespread and assumed oligo-/polygenic form of obesity has so far been less rewarding.
Numerous biologic candidate genes of relevance for the pathogenesis of obesity, including the genes involved in monogenic forms of obesity, have been analyzed for variations that modify the risk for obesity (47). There have been a large number of negative findings, but exciting results also have been reported on associations between gene variants and obesity, such as reports on a frequent, functional promoter variant in the gene encoding uncoupling protein 2 (93) and a rare, functional missense mutation in pomc (94,95), which might influence the risk of developing obesity. Both of these findings need to be replicated in independent and statistically powered studies. Collectively, most of the gene variants identified with the candidate gene approach and claimed to associate with estimates of obesity (e.g., previous or current BMI, waist circumference, weight gain during follow-up, maximal body weight, or age of obesity onset) have been refuted in subsequent investigations in various populations (47).
There may be several possible explanations for these limitations in hunting for obesity genes. A major problem is our lack of insights into the pathogenesis of obesity and lack of knowledge of which obesity-related phenotype should be related to which biologic candidate gene at which time point during the course of obesity development. Obesity markers like BMI and waist circumference are crude and late clinical end points of the obesity disease processes and arise after many years in which a range of more relevant obesity-predisposing phenotypes have
P.835

been evident and available for genetic studies (e.g., altered eating behavior, including taste, olfactory functions, food preferences, mood, stress tolerance, anxiety, or level of physical activity, all of which are genetically influenced). Also, most studies have only examined parts of the genes (often the coding regions), and many pathogenic variants may obviously reside in regulatory regions and in exon/intron borders. Another set of problems in previous studies originates in small study samples, which means that many investigations have been statistically underpowered. In the best-case scenario, we may expect to find obesity-associated alleles conferring relative risks of 1.5 or less (i.e., the study samples should be on the order of several thousands of well-characterized individuals who ideally are recruited from longitudinal and population-based or family-based settings). Most published studies in this research field were also destined to fail because no efforts were made to control for how potential obesity susceptibility genes interact with each other, their genetic background, behavior, and environment to elicit the distal phenotype. In a simplistic view, such interactions might be additive when two risk factors (genes or environment) acting in the same direction to express a more distinct phenotype than either of them alone, or may be multiplicative where one gene variant positively enhances the function of another gene or an environmental risk factor. Thus, a series of related challenges in methodologies and research paradigms in relation to the candidate gene strategy need to be carefully considered if the pertinent obesity genes are to be found.
When it comes to the genomic strategy, more than 20 genome-wide mapping studies involving obesity and obesity-quantitative traits have been reported in various ethnic groups. In subsequent linkage analyses, specific chromosomal regions have been identified and replicated in independent studies with strongly linked loci on several chromosomal regions, including 2p, 10p, and 20q (96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124). Yet, the identity of the potential obesity-causing gene variants within these regions still awaits discovery. Will they turn out to be susceptibility gene haplotypes with minor or modest impact on risk for obesity? Or will some of them have a major and unforeseen biologic role in the pathogenesis of subsets of common obesity? The tremendous pace with which the field of molecular genetics is progressing makes it likely that, in the not too distant future, we will identify some of the common pieces of genes that fit into the extremely complex and challenging picture of obesity-related gene-to-gene and gene-to-behavior/environment interactions.
Implications of Recent Progress in Genetic Studies of Obesity
Since the discovery in the early 1990s of the first genes causing obesity in mice, numerous genes have been mutation scanned as candidates for causing obesity. Also, the entire genome has been examined for loci linked to obesity. A few genes have been found to cause monogenic forms of obesity in humans, and up to 6% of massive obesity may now be explained by genetic variation in these genes, predominantly mc4r. Unfortunately, no breakthrough has been made in the elucidation of the molecular genetic basis of common obesity. However, the quantum leap that has been taken with monogenic obesity has provided insights into novel energy regulatory pathways. A first picture is emerging of a complexity in which there appears to be multiple overlap as well as redundancy between various neural networks with a wide range of feedback mechanisms. Much more will be added to this picture, but pieces of the already achieved advances have been transferred to genomic pharmacotherapy. For instance, the new knowledge gained from the discovery of the leptin signaling pathway has been tested in obese children with congenital leptin deficiency, in whom subcutaneous injection of recombinant human leptin in some of the cases has led to a definite improvement of eating behavior with a profound reduction in energy intake and a dramatic and sustained loss of fat mass, normalization of metabolic disturbances and gonadal function, and marked improvements in quality of life (125,126). Unfortunately, the fact that leptin is antigenic for these children has triggered the production of neutralizing antibody in some of them (126).
A crucial question is whether molecules that are therapeutically effective in monogenic forms of obesity might be relevant for the treatment of common (polygenic) forms of obesity. Concerning the potential therapeutic role of recombinant leptin in common forms of obesity, the situation is even less clear (127). Clinical trials involving large obese populations have been undertaken, and although some degree of dose-response–related loss in body weight occurred, the overall response to leptin treatment was modest and showed no advantages compared with those of other drug interventions in obesity (127,128). The disappointing results may be related to the fact that most obese subjects appear to be leptin resistant (128). Yet, it remains to be examined whether subsets of obese subjects with relatively low levels of circulating leptin may exhibit a greater weight loss response following leptin therapy. Pharmaceutical efforts are likewise under way to discover molecules that activate the postreceptor leptin signaling pathway. One example is the ciliary neurotrophic factor (CNTF) that activates JAK/STAT-mediated signaling in leptin receptor–expressing cells. It has been shown that injection of CNTF in obese animals lacking the leptin receptor was associated with decreased food intake and weight loss, so the effect of CNTF agonists on body weight control should be tested (129).
Probably the most compelling evidence stems from the discovery of the highly obesigenic impact of the mc4r mutations in humans. Rather than hypometabolism, obesity in MC4R mutation carriers is caused by hyperphagia and is associated with an increased occurrence of binge eating; thus, this subset of obesity relates more to a behavioral type defect than a metabolic defect. These findings have pointed to the MC4R as an obvious target for therapeutic intervention, and synthetic agonists of MC4Rs might be strong candidate molecules to regulate energy balance in subsets of common obesity. This idea is supported by a recent study of nonobese subjects who showed loss of body fat after intranasal treatment with an MSH/ACTH fragment molecule (130).
What may be an important outcome of these recent genetic studies is the possibility of identifying subphenotypes among obese individuals, depending on the nature of the genetic defects they carry. Knowing whether a subject is predisposed to a
P.836

type of obesity, related primarily to the rate of metabolism, or whether a central effect on appetite regulation is at play may in the future prove an important clue in designing personalized treatment and prevention regimens.
In conclusion, ample evidence exists that obesity is in part a genetically determined disease, and the discovery of the specific genetic defects underlying rare cases of monogenic forms of obesity has revolutionized our understanding of the intricate molecular mechanisms underlying the control of body weight. It is likely that these insights will be translated into new therapeutics for obesity and perhaps efficient tools in obesity prevention.
References
1. World Health Organization. Obesity: preventing and managing the global epidemic. Report of a WHO consultation on obesity. Geneva: WHO, 1998:1–276.
2. Friedman JM. Obesity in the new millennium. Nature 2000;404: 632–634.
3. Lean MEJ, Hans TS, Seidell JC. Impairment of health and quality of life in people with large waist circumference. Lancet 1998;351: 853–856.
4. Lichtman SW, Pisarska K, Berman ER, et al. Discrepancy between self-reported and actual caloric intake and exercise in obese subjects. N Engl J Med 1992;327:1893–1898.
5. Bandini LG, Schoeller DA, Cyr HN, et al. Validity of reported energy intake in obese and nonobese adolescents. Am J Clin Nutr 1990;52:421–425.
6. Drenick EJ, Johnson D. Weight reduction by fasting and semistarvation in morbid obesity. Int J Obes 1978;2:123–132.
7. Comuzzie AG, et al. Genetic and environmental correlations among hormone levels and measures of body fat accumulation and topography. J Clin Endocrinol Metab 1996;81:597–600.
8. Maes HH, Neale MC, Eaves LJ. Genetic and environmental factors in relative body weight and human adiposity. Behav Genet 1997;27: 325–351.
9. Karlin S, Williams PT, Jensen S, et al. Genetic analysis of the Stanford LRC family study data. I. Structured exploratory data analysis of height and weight measurements. Am J Epidemiol 1981;113:307–324.
10. Borecki IB, Bonney GE, Rice T, et al. Influence of genotype-dependent effects of covariates on the outcome of segregation analysis of the body mass index. Am J Hum Genet 1993;53:676–687.
11. Lecomte E, Herbeth B, Nicaud V, et al. Segregation analysis of fat mass and fat-free mass with age-dependent and sex-dependent effects—the Stanislaz family study. Gen Epidemiol 1997;14:51–62.
12. Tiret L, Andre JL, Ducimetiere P, et al. Segregation analysis of height-adjusted weight with generation- and age-dependent effects. Gen Epidemiol 1992;9:389–403.
13. Borecki IB, Rice T, Perusse L, et al. Major gene influence on the propensity to store fat in trunk versus extremity depots—evidence from the Quebec family study. Obes Res 1995;3:1–8.
14. Zonta LA, Jayakar SD, Bosisio M, et al. Genetic analysis of human obesity in an Italian sample. Hum Heredity 1987;37:129–139.
15. Byard PJ, Siervogel RM, Roche AF. Sibling correlations for weight/ stature and calf circumference. Hum Biol 1983;55:677–685.
16. Nirmala A, Mitchell LE, Rice T, et al. Assessment of adiposity in an Indian population: familial correlations. Gen Epidemiol 1993;10: 133–143.
17. Heller R, Garrison RJ, Havlik RJ, et al. Family resemblances in height and relative weight in the Framingham Heart Study. Int J Obes 1984;8:399–405.
18. Friedlander Y, Kark JD, Kaufmann NA, et al. Familial aggregation of body mass index in ethnically diverse families in Jerusalem. The Jerusalem Lipid Research Clinic. Int J Obes 1988;12:237–247.
19. Bouchard C, Rice T, Lemieux S, et al. Major gene for abdominal visceral fat area in the Quebec Family Study. Int J Obes 1996;20: 420–427.
20. Rice T, Borecki IB, Bouchard C, et al. Segregation analysis of fat mass and other body-composition measures derived from underwater weighing. Am J Hum Gen 1993;52:967–973.
21. Hasstedt SJ, Ramirez ME, Kuida H, et al. Recessive inheritance of a relative fat pattern. Am J Hum Genet 1989;45:917–925.
22. Comuzzie AG, Blangero J, Mahaney MC, et al. Major gene with sex-specific effects influences fat mass in Mexican Americans. Gen Epidemiol 1995;12:475–488.
23. Price RA, Ness R, Laskarzewski P. Common major gene inheritance of extreme overweight. Hum Biol 1990;62:747–765.
24. Moll PP, Burns TL, Lauer RM. The genetic and environmental sources of body mass index variability. Am J Hum Genet 1991;49: 1234–1255.
25. Sørensen T, Sonne-Holm S. Risk in childhood of development of severe adult obesity: retrospective population-based case-cohort study. Am J Epidemiol 1988;127:104–113.
26. Maes HH, Neale MC, Eaves LJ. Genetic and environmental factors in relative body weight and human adiposity. Behav Genet 1997;27: 325–351.
27. Carey DG, Nguyen TV, Campbell LV, et al. Genetic influences on central abdominal fat: a twin study. Int J Obes 1996;20:722–726.
28. Bouchard C, Perusse L, Leblanc C, et al. Inheritance of the amount and distribution of human body fat. Int J Obes 1988;12:212–215.
29. Korkeila M, Kaprio J, Rissanen A, et al. Effects of gender and age on the heritability of body mass index. Int J Obes 1991;15:647–654.
30. Korkeila M, Kaprio J, Rissanen A, et al. Consistency and change of body mass index and weight: a study on 5967 adult Finnish twin pairs. Int J Obes 1995;19:310–317.
31. Herskind AM, McGue M, Sørensen T, et al. Sex and age specific assessment of genetic and environmental influences on body mass index in twins. Int J Obes 1996;20:106–113.
32. Allison DB, Heshka S, Neale MC, et al. A genetic analysis of relative weight amongst 4,020 twin pairs with an emphasis on sex effects. Health Psychol 1994;13:362–365.
33. Pietilainen KH, Kaprio J, Rissanen A, et al. Distribution and heritability of BMI in Finish adolescents aged 16y and 17y: a study of 4884 twins and 2509 singletons. Int J Obes Rel Metab Disord 1999; 23:107–115.
34. Fabsitz RR, Carmelli D, Hewitt JK. Evidence for independent genetic influences on obesity in middle age. Int J Obes Rel Metab Disord 1992;16:657–666.
35. Bouchard C, Tremblay A, Despres JP, et al. The response to long-term overfeeding in identical twins. N Engl J Med 1990;322:1477–1482.
36. Bouchard C, Savard R, Despres JP, et al. Body composition in adopted and biological siblings. Hum Biol 1985;57:61–75.
37. Allison DB, Kaprio J, Korkeila M, et al. The heritability of body-mass index amongst an international sample of monozygotic twins reared apart. Int J Obes 1996;20:501–506.
38. Stunkard AJ, Sørensen T, Hanis C, et al. An adoption study of human obesity. N Engl J Med 1986;314:193–198.
39. Price RA, Cadoret RJ, Stunkard AJ, et al. Genetic contributions to human fatness. Am J Psychiatry 1987;144:1003–1008.
40. Sørensen T, Price RA, Stunkard AJ, et al. Genetics of obesity in adult adoptees and their biological siblings. BMJ 1989;298:87–90.
41. Sørensen T, Stunkard AJ. Does obesity run in families because of genes? An adoption study using silhouettes as a measure of obesity. Acta Psychiatr Scand Suppl 1996;370:67–72.
42. Sørensen T, Holst C, Stunkard AJ. Childhood body mass index—genetic and familial environmental influences assessed in a longitudinal adoption study. Int J Obes 1992;16:705–714.
43. Vogler GP, Sørensen T, Stunkard AJ, et al. Influences of genes and shared family environment on adult body mass index assessed in an adoption study by a comprehensive path model. Int J Obes 1995;19: 40–45.
44. Price RA, Gottesman II. Body fat in identical twins reared apart. Behav Gen 1991;21:1–7.
P.837

45. McCarthy MI. Susceptibility gene discovery for common metabolic and endocrine traits. J Mol Endocrinol 2002;28:1–17.
46. Reed DR, Ding Y, Xu W, et al. Human obesity does not segregate with the chromosomal regions of Prader-Willi, Bardet-Biedl, Cohen, Borjeson or Wilson-Turner syndromes. Int J Obes 1995;19:599–603.
47. Chagnon YC, Rankinen T, Snyder EE, et al. The human obesity gene map: the 2002 update. Obes Res 2003;11:313–367.
48. Bultman SJ, Michaud EJ, Woychik RP. Molecular characterization of the mouse agouti locus. Cell 1992;71:1195–1204.
49. Michaud EJ, Bultman SJ, Klebig ML, et al. A molecular-model for the genetic and phenotypic characteristics of the mouse lethal yellow (a (y)) mutation. Proc Natl Acad Sci USA 1994;91:2562–2566.
50. Siracusa LD. The agouti gene: turned on to yellow. Trends Genet 1994;10:423–428.
51. Zhang Y, et al. Positional cloning of the mouse obese gene and its human homologue. Nature 1994;372:425–432.
52. Tartaglia LA, Dembski M, Weng X, et al. Identification and expression cloning of a leptin receptor ob-R. Cell 1995;83:1263–1271.
53. Chua SC, Chung WK, Wupeng XS, et al. Phenotypes of mouse diabetes and rat fatty due to mutations in the ob (leptin) receptor. Science 1996;271:994–996.
54. Naggert JK, Fricker LD, Varlamov O, et al. Hyperproinsulinaemia in obese fat/fat mice associated with a carboxypeptidase-e mutation which reduces enzyme-activity. Nat Genet 1995;10:135–142.
55. Kleyn PW, Fan W, Kovats SG, et al. Identification and characterization of the mouse obesity gene tubby—a member of a novel gene family. Cell 1996;85:281–290.
56. Seeley RJ, Yagaloff KA, Fisher L, et al. Melanocortin receptors in leptin effects. Nature 1997;390:349.
57. Fan W, Boston BA, Kesterson RA, et al. Role of melanocortinergic neurons in feeding and the agouti obesity syndrome. Nature 1997; 385:165–168.
58. Friedman JM, Halaas JL. Leptin and the regulation of body weight in mammals. Nature 1998;395:763–770.
59. Ahima RS, et al. Role of leptin in the neuroendocrine response to fasting. Nature 1996;382:250–252.
60. Woods SC, Seeley RJ, Porte D, et al. Signals that regulate food intake and energy homeostasis. Science 1998;280:1378–1383.
61. Ahima RS, Kelly J, Elmquist J, et al. Distinct physiologic and neuronal responses to decreased leptin and mild hyperleptinemia. Endocrinology 1999;140:4923–4931.
62. Rossi M, et al. A C-terminal fragment of agouti-related protein increases feeding and antagonizes the effect of alpha-melanocyte stimulating hormone in vivo. Endocrinology 1998;139:4428–4431.
63. Wilson BD, Ollmann M, Barsh GS. The role of agouti-related protein in regulating body weight. Mol Med Today 1999;5:250–256.
64. Shutter JR, et al. Hypothalamic expression of CART, a novel gene related to agouti, is up-regulated in obese and diabetic mutant mice. Genes Dev 1997;11:593–602.
65. Elmquist JK, Elias CF, Saper CB. From lesions to leptin: hypothalamic control of food intake and body weight. Neuron 1999;22: 221–232.
66. Elias CF, et al. Leptin differentially regulates NPY and POMC neurons projecting to the lateral hypothalamic area. Neuron 1999;23: 775–786.
67. Mizuno TM, Mobbs CV. Hypothalamic agouti-related protein messenger ribonucleic acid is inhibited by leptin and stimulated by fasting. Endocrinology 1999;140:814–817.
68. Broberger C, Johansen J, Johansson C, et al. The neuropeptide Y/agouti gene-related protein (AGRP) brain circuitry in normal, anorectic, and monosodium glutamate–treated mice. Proc Natl Acad Sci USA 1998;95:15043–15048.
69. Elmquist JK, Maratos-Flier E, Saper CB, et al. Unraveling the central nervous system pathways underlying responses to leptin. Nat Neurosci 1998;1:445–450.
70. Seeley R, et al. Melanocortin receptors in leptin effects. Nature 1997; 390:349.
71. Cheung CC, Clifton DK, Steiner RA. Proopiomelanocortin neurons are direct targets for leptin in the hypothalamus. Endocrinology 1997;138:4489–4492.
72. Marsh DJ, et al. Response of melanocortin-4 receptor–deficient mice to anorectic and orexigenic peptides. Nat Genet 1999;21: 119–122.
73. Boston B, Blaydon K, Varnerin J, et al. Independent and additive effects of central POMC and leptin pathways on murine obesity. Science 1997;278:1641–1644.
74. Yaswen L, Diehl N, Brennan, MB, et al. Obesity in the mouse model of pro-opiomelanocortin deficiency responds to peripheral melanocortin. Nat Med 1999;5:1066–1070.
75. Montague CT, Farooqi IS, Whitehead JP, et al. Congenital leptin deficiency is associated with severe early-onset obesity in humans. Nature 1997;387:903–908.
76. Strobel A, Issad T, Camoin L, et al. A leptin missense mutation associated with hypogonadism and morbid obesity. Nat Genet 1998;18: 213–215.
77. Clement K, Vaisse C, Lahlou N, et al. A mutation in the human leptin receptor gene causes obesity and pituitary dysfunction. Nature 1998;392:398–401.
78. Krude H, Biebermann H, Luck W, et al. Severe early-onset obesity, adrenal insufficiency and red hair pigmentation caused by POMC mutations in humans. Nat Genet 1998;19:155–157.
79. Jackson RS, Creemers JW, Ohagi S, et al. Obesity and impaired prohormone processing associated with mutations in the human prohormone convertase 1 gene. Nat Genet 1997;16:303–306.
80. Vaisse C, Clement K, Guy-Grand B, et al. A frameshift mutation in human MC4R is associated with a dominant form of obesity. Nat Genet 1998;20:113–114.
81. Hinney A, Schmidt A, Nottebom K, et al. Several mutations in the melanocortin-4 receptor gene including a nonsense and a frameshift mutation associated with dominantly inherited obesity in humans. J Clin Endocrinol Metab 1999;84:1483–1486.
82. Nishigori H, Tomura H, Tonooka N, et al. Mutations in the small heterodimer partner gene are associated with mild obesity in Japanese subjects. Proc Natl Acad Sci USA 2001;98:575–580.
83. Ristow M, Muller-Wieland D, Pfeiffer A, et al. Obesity associated with a mutation in a genetic regulator of adipocyte differentiation. N Engl J Med 1998;339:953–959.
84. Coleman DL. Effects of parabiosis of obese with diabetes and normal mice. Diabetologia 1973;9:294–298.
85. Coleman DL, Hummel KP. Effects of parabiosis of normal with genetically diabetic mice. Am J Physiol 1969;217:1298–1304.
86. Jacobsen P, Ukkola O, Rankinen T, et al. Melanocortin 4 receptor sequence variations are seldom a cause of human obesity: the Swedish Obese Subjects, the HERITAGE Family Study, and a Memphis cohort. J Clin Endocrinol Metab 2002;87:4442–4446.
87. Miraglia Del Giudice E, Cirillo G, Nigro V, et al. Low frequency of melaocortin-4 receptor (MC4R) mutations in a Mediterranean population with early-onset obesity. Int J Obes Rel Metab Disord 2002; 26:647–651.
88. Gu W, Tu Z, Kleyn PW, et al. Identification and functional analysis of novel human melanocortin-4 receptor variants. Diabetes 1999; 48:635–639.
89. Vaisse C, Clement K, Durand E, et al. Melanocortin-4 receptor mutations are a frequent and heterogeneous cause of morbid obesity. J Clin Invest 2000;106:253–262.
90. Farooqi IS, Keogh JM, Yeo GSH, et al. Clinical spectrum of obesity and mutations in the melanocortin 4 receptor gene. N Engl J Med 2003;348:1085–1095.
91. Branson R, Potoczna N, Kral JG, et al. Binge eating as a major phenotype of melanocortin 4 receptor gene mutations. N Engl J Med 2003;348:1096–1103.
92. Kersten S. Peroxisome proliferators activated receptors and obesity. Eur J Pharmacol 2002;440:223–234.
93. Esterbauer H, Schneitler C, Oberkofler H, et al. A common polymorphism in the promoter of UCP2 is associated with decreased risk of obesity in middle-aged humans. Nat Genet 2001;28:178–183.
94. Echwald SM, Sørensen TI, Andersen T, et al. Mutational analysis of the proopiomelanacortin gene in Caucasians with early onset obesity. Int J Obes 1999; 23:293–298.
P.838

95. Challis BG, Pritchard LE, Creemers JWM, et al. 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 2002;17 (11):1997–2004.
96. Norman RA, Thompson DB, Foroud T, et al. Genomewide search for genes influencing percent body fat in Pima Indians: suggestive linkage at chromosome 11q21–q22. Am J Hum Genet 1997;60:166–173.
97. Norman RA, Tataranni PA, Pratley R, et al. Autosomal genomic scan for loci linked to obesity and energy metabolism in Pima Indians. Am J Hum Genet 1998;62:659–668.
98. Hanson RL, Ehm MG, Pettitt DJ, et al. An autosomal genomic scan for loci linked to type II diabetes mellitus and body mass index in Pima Indians. Am J Hum Genet 1998;63:1130–1138.
99. Walder K, Hanson RL, Kobes S, et al. An autosomal genomic scan for loci linked to plasma leptin concentration in Pima Indians. Int J Obes 2000;24:559–565.
100. Duggirala R, Blangero J, Almasy L, et al. A major susceptibility locus influencing plasma triglyceride concentrations is located on chromosome 15q in Mexican Americans. Am J Hum Genet 2000;66: 1237–1245.
101. Duggirala R, Blangero J, Almasy L, et al. A major locus for fasting insulin concentrations and insulin resistance on chromosome 6q with strong pleiotropic effects on obesity-related phenotypes in nondiabetic Mexican Americans. Am J Hum Genet 2001;68:1149–1164.
102. Comuzzie AG, Hixson JE, Almasy L, et al. A major quantitative trait locus determining serum leptin levels and fat mass is located on human chromosome 2. Nat Genet 1997;15:273–276.
103. Kissebah AH, Sonnenberg GE, Myklebust J, et al. Quantitative trait loci on chromosomes 3 and 17 influence phenotypes of the metabolic syndrome. Proc Natl Acad Sci USA 2000;97:14478–14483.
104. Lee JH, Reed DR, Li WD, et al. Genome scan for human obesity and linkage to markers in 20q13. Am J Hum Genet 1999;64:196–209.
105. Zhu X, Cooper RS, Luke A, et al. A genome-wide scan for obesity in African-Americans. Diabetes 2002;51:541–544.
106. Feitosa MF, Borecki IB, Rich SS, et al. Quantitative-trait loci influencing body-mass index reside on chromosomes 7 and 13: the National Heart, Lung, and Blood Institute Family Heart Study. Am J Hum Genet 2002;70:72–82.
107. Chagnon YC, Borecki IB, Perusse L, et al. Genome-wide search for genes related to the fat-free body mass in the Quebec Family Study. Metabolism 2000;49:203–207.
108. Perusse L, Rice T, Chagnon YC, et al. A genome-wide scan for abdominal fat assessed by computed tomography in the Quebec Family Study. Diabetes 2001;50:614–621.
109. Hsueh WC, Mitchell BD, Schneider JL, et al. Genome-wide scan of obesity in the Old Order Amish. J Clin Endocrinol Metab 2001;86: 1199–1205.
110. Hager J, Dina C, Francke S, et al. A genome-wide scan for human obesity genes reveals a major susceptibility locus on chromosome 10. Nat Genet 1998;20:304–308.
111. Öhmann M, Oksanen L, Kaprio J, et al. Genome-wide scan of obesity in Finnish sibpairs reveals linkage to chromosome Xq24. J Clin Endocrinol Metab 2000;85:3183–3190.
112. Parker A, Meyer J, Lewitzky S, et al. A gene conferring susceptibility to type 2 diabetes in conjunction with obesity is located on chromosome 18p11. Diabetes 2001;50:675–680.
113. Perola M, Öhmann M, Hiekkalinna T, et al. Quantitative-trait-locus analysis of body-mass index and of stature, by combined analysis of genome scans of five Finnish study groups. Am J Hum Genet 2001; 69:117–123.
114. van der Kallen CJH, Lindgren CM, Daly MJ, et al. Genome scan for adiposity in Dutch dyslipidemic families reveals novel quantitative trait loci for leptin, body mass index and soluble tumor necrosis factor receptor superfamily 1A. Int J Obes 2000;24:1381–1391.
115. Clement K, Garner C, Hager J, et al. Indication for linkage of the human OB gene region with extreme obesity. Diabetes 1996;45: 687–690.
116. Reed DR, Ding Y, Xu W, et al. Extreme obesity may be linked to markers flanking the human OB gene. Diabetes 1996;45:691–694.
117. Wu X, Cooper RS, Borecki I, et al. A combined analysis of genomewide linkage scans for body mass index from the National Heart, Lung, and Blood Institute Family Blood Pressure Program. Am J Hum Genet 2002;70:1247–1256.
118. Duggirala R, Stern MP, Mitchell BD, et al. Quantitative variation in obesity-related traits and insulin precursors linked to the OB gene region on human chromosome 7. Am J Hum Genet 1996;59:694–703.
119. Lapsys NM, Furler SM, Moore KR, et al. Relationship of a novel polymorphic marker near the human obese (OB) gene to fat mass in healthy women. Obes Res 1997;5:430–433.
120. Roth H, Hinney A, Ziegler A, et al. Further support for linkage of extreme obesity to the obese gene in a study group of obese children and adolescents. Exp Clin Endocrinol Diabetes 1997;105:341–344.
121. Hinney A, Ziegler A, Oeffner F, et al. Independent confirmation of a major locus for obesity on chromosome 10. J Clin Endocrinol Metab 2000;85:2962–2965.
122. Price RA, Li WD, Bernstein A, et al. A locus affecting obesity in human chromosome region 10p12. Diabetologia 2001;44:363–366.
123. Hunt SC, Abkevich V, Hensel CH, et al. Linkage of body mass index to chromosome 20 in Utah pedigrees. Hum Genet 2001;109:279–285.
124. Pietilainen KH, Kaprio J, Rissanen A, et al. Distribution and heritability of BMI in Finnish adolescents aged 16 y and 17 y: a study of 4884 twins and 2509 singletons. Int J Obes Rel Metab Disord 1999; 23:107–115.
125. Farooqi S, Jebb S, Langmak G, et al. Effects of recombinant leptin therapy in a child with congenital leptin deficiency. N Engl J Med 1999;341:879–884.
126. O’Rahilly S. Insights into obesity and insulin resistance from the study of extreme human phenotypes. Eur J Endocrinol 2002;147: 435–441.
127. Dagogo-Jack S. Human leptin regulation and promise in pharmacotherapy. Curr Drug Targets 2001;2:181–195.
128. Heymsfield SB, Greenberg AS, Fujioka K, et al. Recombinant leptin for weight loss in obese and lean adults: a randomized, controlled, dose-escalation trial. JAMA 1999;282:1568–1575.
129. Lambert PD, Anderson KD, Sleeman MW, et al. Ciliary neurotrophic factor activates leptin-like pathways and reduces body fat, without cachexia or rebound weight gain, even in leptin-resistant obesity. Proc Natl Acad Sci USA 2001;98:4652–4657.
130. Fehm HL, Smolnik R, Kern W, et al. The melanocortin melanocyte-stimulating hormone/adrenocorticotropin (4–10) decreases body fat in humans. J Clin Endocrinol Metab 2001;86:1144–1148.
←↑→