The prevalence of obesity has increased dramatically in recent decades. Increased body habitus correlates with enhanced availability of low-cost, high-calorie foods. A mismatch in caloric intake and energy expenditure leads to excess body fat accumulation. There are, however, interindividual differences in susceptibility to obesity with exposure to the same obesogenic environment. It is likely that this variability in metabolism, energy storage, or neural hormonal components affecting behavior has genetic, epigenetic, or genomic components.

Gregor Johann Mendel discovered the laws of heredity that bear his name¹ by combining diligent observation, prudent choice of experimental organism, and mental discipline. These allowed him to discern and then explain patterns consistent with unitary particulate determinative factors governing discrete phenotypic traits (genes). This insight is an example of Pascal’s dictum that “chance favors the prepared mind.”² Unitary genetic traits such as sickle cell disease are the delight of educators, who employ them to convey foundational concepts in biology with elegant simplicity: one gene, one protein, one phenotype. This, of course, allows for oversimplification. By contrast, complex traits such as intelligence involve the interaction of multiple genes with various environmental influences. Readers of this text will likely have encountered Mendelian genetics. However, anyone whose genetics learning occurred more than a decade ago may reference and update their vocabulary of new concepts in genomics by referencing various genetics textbooks.^3,4

The first single-gene mutation associated with obesity was identified in 1994 by positional cloning of the ob/ob mouse strain. The mutation was mapped to the leptin (lep) gene, which encodes a 16-kDa secreted hormone produced primarily by white adipose tissue.⁵ Subsequently, a mutation in the leptin receptor gene (lepr) was identified in another obese strain of mice. The leptin receptor is primarily expressed in the hypothalamus, and this hormonal signaling pathway plays an important role in regulation of food intake in both strains of mice. As it eventually became clear, the same is true in humans. Rare homozygous mutations and polymorphisms in both the LEP and LEPR genes have been identified in patients with severe forms of early-onset obesity and hyperphagia⁶ (see Sidebar 1-1). Leptin deficiency also manifests with reduced insulinlike growth factor 1 (IGF-1) levels, elevated insulin levels, hypothalamic hypothyroidism, impaired growth hormone secretion, and hypogonadotropic hypogonadism. Replacement of leptin by administration of the exogenous peptide has shown remarkable reversal of obesity in patients with homozygous mutations in LEP.⁷

Further research into the leptin signaling pathway elucidated other genes involved in energy homeostasis. Proopiomelanocortin (POMC) has its expression positively regulated by leptin. Loss-of-function (LOF) mutations in POMC are notable for early-onset hyperphagia and obesity, adrenal insufficiency, pale skin, and red hair.⁸ It is produced by the arcuate nucleus of the hypothalamus⁹ and is cleaved to produce alpha-melanocyte-stimulating hormone (α-MSH), a ligand that acts on the melanocortin 4 receptor (MC4R), a seven-transmembrane G-protein-linked receptor that is also expressed in the hypothalamus. MC4R likely transmits the signals from the leptin pathway to downstream effectors through the paraventricular nucleus of the hypothalamus.¹⁰ Both autosomal recessive and dominant LOF mutations in MC4R have been found,¹¹ and they are strongly associated with obesity (Sidebar 1-2).¹² Population studies show that pathogenic MC4R mutations account for approximately 2% of childhood and adult obesity cases.^13,14

PCSK1 codes for proprotein convertase subtilisin/kexin type 1,¹⁵ an integral component of the post-translational proteolytic processing of POMC.^16,17 Mutations in the PCSK1 (also known as PC1) gene have been identified in patients with severe childhood obesity, abnormal glucose homeostasis, reduced plasma insulin with elevated proinsulin levels, hypogonadotropic hypogonadism, and hypocortisolemia. Even carriers of mutations causing partial PCSK1 deficiency have a significantly higher risk of obesity.¹⁸

SIM1 is one of the effectors of MC4R in controlling energy homeostasis and has been shown to have an essential role in embryonic neurogenesis within the ventral neurogenic region in Drosophila.¹⁹ LOF mutations in SIM1 have been characterized in mouse models and identified in humans, resulting in severe hyperphagia and obesity, developmental delays, short extremities, and hypotonia.^20,21,22 Functional studies of hyperphagic obesity in Sim1-deficient mice suggest a role in the leptin-melanocortin-oxytocin pathway and that postnatal central nervous system (CNS) deficiency of Sim1 is sufficient to cause hyperphagic obesity.²³

Brain-derived neurotrophic factor (BDNF) is another effector that regulates eating behavior and energy balance downstream of MC4R.²⁴ Knockout studies in mice demonstrated obesity and hyperactivity. LOF mutation in one copy of the gene was found in a girl with hyperphagia, early-onset obesity, cognitive impairment, and hyperactivity. A missense variant in the BDNF receptor encoded by NTRK2 has also been reported in a boy with severe obesity and memory impairment.²⁵

Adiponectin (encoded by the ADIPOQ gene) is a hormone secreted by adipose tissue that modulates metabolic processes, including glucose regulation and fatty acid breakdown.²⁶ It is associated with both obesity and diabetes mellitus type 2 (DMT2), where it has notably reduced expression levels.²⁷ Similarly, proline-to-alanine substitution at position 12 in the peroxisome proliferator-activated receptor gamma (PPAR) gene is associated with both obesity and DMT2.^28,29

GPR120 (also known as O3FAR1) encodes a G-protein-coupled receptor for unsaturated long-chain free fatty acids that plays a critical role in energy homeostasis by obesity, glucose intolerance, fatty liver with decreased adipocyte differentiation, and lipogenesis. Furthermore, a deleterious arginine-to-histidine mutation at position 270 inhibits GPR120 signaling in obese subjects.³⁰ As a lipid sensor, GRP120 provides a link between physiological response and the environment, in particular providing insights to the mechanism by which an obesogenic, high-calorie, fatty diet may contribute to obesity and associated complications like fatty liver. Such environment-sensitive genes provide particular interest in translational research as potential drug targets for prevention or treatment of obesity.³¹

Regions of genetic material may be lost or duplicated in one individual compared to others, with no apparent phenotypic consequence. Such changes (microdeletions or microduplications) are called copy number variants (CNVs). They may involve many tens of thousands of nucleotides, such that the loss or gain would seem intuitively to have important phenotypic consequence, but does not. Knowledge of the human genome has evolved rapidly over the last decade. We now have a large catalog of CNVs that are known to have no phenotypic importance, but we also have another list of deletions and duplications that are well established to have causative relationship to more or less well-defined adverse phenotypes. There is also a third list of changes that appear to imply pathology but have not as yet been documented as such in the literature.

That third group is cause for consternation in the clinical setting, where theoretical importance does not suffice to explain an adverse phenotype or to enlighten the family in terms of optimum management or long-term prognosis. The role of CNVs in causing adverse phenotypes is coming into focus gradually; there is evidence that apparently unimportant CNVs might become important if an individual has more than one of them.³² Testing for genomic changes (comparative genomic hybridization, or chromosome microarray) is now a routine clinical approach to the evaluation of individuals with intellectual disability, autism spectrum, or multiple anatomic anomalies.^33,34 Studying the genome in greater depth, with whole-exome analysis (next-generation DNA sequencing), is used only in the most complex cases, but as the cost of this approach falls, it is likely to replace microarray in the future.³⁵ Such testing sometimes allows for predictive or preventive measures.³⁶

A portion of a chromosome may be missing or extra, with more or less serious consequence. Some are well known, such as cri-du-chat syndrome, which is associated with missing part of the short arm of number 5; or Wolf-Hirshhorn syndrome, which is associated with missing a portion of the short arm of number 4. Prader-Willi syndrome usually involves missing a portion of the long arm of number 15 (see Sidebar 1-3), while Williams syndrome is found with a missing part of the long arm of number 7. The 22q11 deletion syndrome involves loss of that segment and goes by several names (DiGeorge syndrome, Shprintzen syndrome, velocardiofacial syndrome).

The CNVs are considered either common variants with frequencies greater than 5% or rare variants with frequencies less than 1% in the general population. Several structural variants have also been found to have associations with obesity. One such association is the common CNV on chromosome 10q11.22 that encompasses the pancreatic polypeptide receptor 1 (PPYR1) gene, with low copy numbers and deletions observed in individuals with increased body mass index (BMI).³⁷ Another common CNV on 11q11, encompassing olfactory receptor genes OR4P4, OR4S2, and OR4C6, has a deletion allele that is associated with extreme early-onset obesity.³⁸ Large meta-analysis studies have demonstrated deletions associated with increased BMI near the hypothalamic-expressed neuronal growth regulator 1 (NEGR1) gene, while a 21-kb CNV was also mapped upstream of GPRC5B with a nonrisk deletion allele.³⁹ Further studies of GPRC5B in deficient mice found protection from diet-induced obesity and highlight its role as a major node in adipose signaling systems that link diet-induced obesity to DMT2.⁴⁰

Large deletions in 16p11.2 have been shown to result in a 43-fold increase in the risk of morbid obesity.⁴¹ Interestingly, reciprocal duplications at this locus have been found to demonstrate a mirror effect on phenotype, with an 8-fold increase in the risk of being underweight.⁴² SH2B1 is the common gene to all 16p11.2 deletions, and genome-wide association studies (GWAS) have revealed single-nucleotide polymorphisms (SNPs) within SH2B1 that are associated with BMI.⁴³ Functional studies in knockout mice show disrupted leptin and insulin signaling that supports the role of this gene in energy homeostasis regulation.⁴⁴ In addition, LOF mutations in SH2B1 have been associated with aggressive and maladaptive behaviors.⁴⁵

Many genes have been implicated in monofactorial forms of obesity, and a large number of polymorphisms with smaller effect size associated with adiposity have been discovered through GWAS. Although monogenic forms of obesity are rare, they are often severe with a large effect size, having clinical importance in screening and early intervention. The complex contribution of polygenic variants and epigenetics may not currently have significant clinical utility, but researchers are finding applications for risk and outcome prediction to enhance clinical decision-making that eventually will be integrated into clinical practice. Research in this field has enhanced our understanding of the pathophysiology of obesity and is rapidly evolving with advances in sequencing technologies toward the promise of new therapeutics and personalized medicine.

Polygenic (or common) obesity is the result of the combined contribution of multiple genetic variants in the setting of environmental risk factors. The candidate gene study of the genetics of obesity in mice and humans has provided great insights into the mechanism of energy homeostasis. Yet, this approach has been limited to known pathways with strong monogenic components because we see what we know and expect to see in complex situations. The advent of GWAS in part addresses this issue due to its global capability and power in screening for novel loci of association without requiring a priori knowledge of the biological mechanism.

Many new loci and pathways in obesity have been discovered via GWAS. One of the most notable genes discovered this way is the fat mass and obesity-associated FTO gene, which harbors one of the first common variants found to be strongly correlated with obesity in humans.⁴⁶ The cluster of SNPs in FTO was discovered in 2007 during GWAS for DMT2 by the Wellcome Trust Case Control Consortium.⁴⁷ Once the data were corrected for BMI, it was realized that the associated phenotype was obesity rather than DMT2.⁴⁸ Multiple studies in both children and adults of different ancestries have validated these findings.^49,50,51