Browse the corpus

Rodent models in which monogenic alterations cause obesity in the absence of environmental changes have confirmed earlier inferences regarding the biologic/genetic control of energy balance in mammals. Spontaneous models of obesity in rodents have identified fundamental molecular/cellular systems underlying the control of feeding and energy homeostasis, including the melanocortin system and CNS circuits that respond to leptin. Engineered mutations in rodents have revealed additional genes and pathways participating in the control of body weight. Most genes that impact body weight and adiposity affect the brain systems that control “regulatory” and/or hedonic aspects of feeding behavior. Effects on energy expenditure are frequently present, but of smaller magnitude. Differences in adipocyte physiology between obese and lean individuals appear to be largely secondary phenomena . Most of the rodent monogenic obesities have human orthologs that result in comparably severe obesity-related phenotypes. While these mutations are rare, they confirm that the molecular predicates for the control of body weight in humans are fundamentally the same as those in the rodents. Engineered mutants in rodents have also permitted the analysis of putative genetic contributors to obesity in humans, as well as providing an initial blueprint of the molecular components, pathways and physical interconnections of these systems.

Historically, obesity has been considered a disorder of voluntary behaviors, exacerbated by the ready availability of food and reduced need for energy expenditure afforded by modern societies. Rodent models in which monogenic alterations provoke obesity in the absence of environmental changes have, however, conclusively demonstrated the biologic control of energy balance in mammals. Indeed, orthologs of many of these genes cause or contribute to obesity in humans. Spontaneous mono and polygenic models of obesity in rodents (along with obesity phenotypes of engineered mutations) have identified fundamental molecular/cellular systems underlying the control of feeding and energy homeostasis. Importantly, most genes that impact body weight and adiposity affect the brain systems that control feeding. Genetic studies in rodents have provided an initial blueprint of the molecular constituents and interconnection of these systems.

The first law of thermodynamics (the conservation of energy) dictates that body energy stores reflect the difference between energy taken in and energy expended. More intake with relatively less expenditure leads to energy storage (generally, in adipose tissue); conversely, when expenditure exceeds intake, energy/fat stores decline. In this context, energy is generally taken in by eating (or drinking calorie-containing beverages). Behaviorally, two related systems govern eating (1): The circuits that control the incentive and reward values (wanting and liking) of food, and the satiety system, which promotes meal termination associated with the sensation of “fullness.” Each of these systems is subject to long and short-term regulation. They are physiologically integrated, but for simplicity frequently studied and described as distinct entities.

Energy loss/expenditure includes energy consumed but not absorbed (and which therefore passes out of the body in the stool; this is ~5% under most normal circumstances); energy metabolized in the process of breaking down and storing nutrients (diet-induced thermogenesis); energy metabolized to maintain baseline cellular functions at rest (resting metabolic rate, BMR); and energy consumed in physical activity (non-resting energy expenditure; NREE) (2). RMR accounts for about 70% of total energy expenditure (TEE) in sedentary adults, and is determined by body composition, age, wakefulness, and genetic factors. NREE is the next largest contributor, averaging approximately 20% of TEE. In sedentary individuals, low-level physical activities (fidgeting, short bouts of ambulation, etc.) make up most of NREE. Smaller amounts of energy (7%) are accounted for by diet-induced thermogenesis. Recently, there has been growing interest in the contributions of the gut microbiome to systemic energy homeostasis by possible effects on the efficiency of nutrient utilization in the gut and the consequences of such bacterial metabolism for release of metabolites that could affect energy intake or expenditure (3).

Adipose tissue represents the major repository for ingested energy that exceeds immediate needs (2). The energy density of adipose tissue is nearly 10-fold greater than muscle (protein) or liver (glycogen). The ability to store such energy protects against environmental vicissitudes that might result in starvation, fetal wastage, and inability to provide sufficient breast milk to the young. Therefore, it is likely that evolution has promoted genes/alleles that favor energy storage and conservation. The existence of environments in which excess calories are readily available with minimum or no effort is a very recent occurrence in human evolution. Human genetic “makeup” is presumably designed for the opposite circumstance. Contrary to earlier prevailing views, adipose tissue is not a passive energy depot, but participates in homeostatic processes that regulate food intake (e.g. production of leptin), and storage (e.g., insulin) and release (e.g., catecholamines) of the acylglycerides stored within them (4).

In 1902, French geneticist L. Cuenot described the obese Yellow ( A y /a ) mouse, which had been bred and maintained by European mouse fanciers since the 1800s (5). This was the first report of a spontaneously obese mouse, which prompted investigation of additional spontaneous obese mouse models, including by investigators at the Jackson Laboratories (Jax) in Bar Harbor, Maine. The autosomal recessive obese (ob) mutation was discovered at Jax in 1949-50, after spontaneously arising in a non-inbred strain (6). Sixteen years later, a phenotypically similar mouse was identified (7). The diabetic state of these latter animals (studied on the diabetes-prone coisogenic KsJ background) distinguished them from ob (studied on the B6 background) and hence the mutation was designated diabetes (db). In 1990, Coleman and colleagues described additional, milder recessive obese mutations in mice: tubby (tub), and fat (fat) (8,9) . Fewer obesity-related spontaneous mutations have been detected in rats, due in part to the absence of explicit screening of large numbers of progeny for such phenotypes, and the greater cost of rat husbandry. In addition to identifying mutations in genes identical to some of the murine genes above (e.g. leptin receptor mutations in db mice and the Zucker and Koletsky rats(10,11)), however, the OLETF obese rat has also been described (12).

Each of these mutant animals is hyperphagic compared to controls. Furthermore, classic genetic studies revealed that each of these obese phenotypes had predictable, monogenetic heritability, demonstrating the genetic underpinnings of feeding, as well as overall energy balance. The subsequent finding that some of these rodent obesity genes control body weight in humans confirms that biologic/genetic factors control feeding and the predisposition to obesity in humans, as well as in rodents (13).

In addition to spontaneous models of obesity, genetic engineering (generally in mice) has provided many examples in which genetic alterations modulate body weight and adiposity. In the early 1980s, genetic manipulation techniques became available in rodents, enabling the analysis of systemic and organ-specific effects on physiology of one or more genes selected by the investigator (14). Several genes important for energy balance that have been examined by such approaches are discussed below. Historically, gene manipulation in mammals has been accomplished by one of two distinct means: standard transgenesis and gene targeting .

In standard transgenesis , artificial genes (often comprising promoter sequences designed to produce desired patterns of expression plus the coding sequences for the molecule to be expressed) are directly introduced into a fertilized oocyte, which is then implanted in a female surrogate to permit the development of the transgenic animal (15). In this method, the site of the transgene insertion in the genome is random; hence, the insertion may inadvertently disrupt endogenous genes, and the expression pattern may be influenced by the site of insertion as well as by the promoter sequences used. The use of very large genomic regions (such as those derived from bacterial artificial chromosomes (BACs)) to drive the gene of interest can mitigate some, but not all, of these expression issues (16). Multiple independent transgenic lines must therefore be screened to identify correctly-expressing progeny.

Gene targeting is accomplished by using homologous recombination to introduce genetic sequences designed to modify specific genes, while leaving the rest of the genome intact (17). Generally, manipulated DNA sequences are introduced into undifferentiated murine embryonic stem (ES) cells to recombine with native DNA, producing a modified ES cell line in which a specific gene is inactivated (a “knockout” or KO) or altered by editing or by the introduction of new genetic material (a “knockin” or KI) . The modified ES cells are then injected into blastocysts, which are implanted into surrogate mothers. The resultant pups generally contain cells derived from the ES cells along with cells donated by the recipient blastocyst (hence, these animals are termed “chimeras”). The chimeras are then bred in an effort to obtain germline transmission of the ES cell-derived genes. Thus, while gene targeting is quite specific in terms of the types and locations of manipulations, the use of ES cells requires a greater up-front investment of time and resources (generating the homologous targeting construct, screening ES cell clones for correct targeting, breeding for germline transmission, etc.) than does standard transgenesis.

New technologies have emerged over the past several years that promise to facilitate gene targeting. These generally involve the use of site-specific nucleases (zinc finger nucleases, TALENs, CRISPRs, etc.) to create breaks in the genomic DNA of fertilized oocytes or ES cells (18,19). These site-specific breaks can be employed not only to produce KO animals, but also to increase the efficiency of site-specific recombination within the oocyte, so that the co-injection of homologous templates can be used to generate KI animals without the use of ES cells.

In addition to the production of standard KO animals, and animals with edited coding or regulatory sequences in specific genes, gene targeting is also commonly used to produce conditional null or specific gene-expressing animals, often by employing the Cre recombinase/LoxP system (19). This bacteriophage-derived system is composed of two components- Cre (a site-specific DNA recombinase) and LoxP (the short DNA sequences recognized by Cre). In most versions of the system, Cre removes the DNA sequences that are flanked by LoxP sites (“floxed”). By combining tissue-specific Cre expression with floxed genes of interest, the floxed genes may be disrupted in a tissue- and time-specific manner. Cre-expressing animals may be generated by standard transgenesis (with the caveats, above, regarding site of integration), or can be delivered to specific sites in the genome by homologous targeting. Animals carrying floxed alleles are produced by delivering LoxP sites to the desired locations in the genome by homologous targeting.

While many homologous targeting events are designed to be benign, they may have unintended consequences for the expression of the targeted gene (e.g., alterations in expression) or surrounding genes; these must be controlled for carefully. Another important consideration in mouse models of obesity generated by gene targeting is genetic background effects. Targeting has been most consistently successful in the 129 mouse strain because the ES cells of 129 mice are relatively easy to culture and manipulate. However, this strain exhibits increased levels of anxiety in response to environmental stressors that could potentially distort food intake and metabolic phenotypes (20). ES cells from C57BL/6 require carefully controlled in vitro conditions, and even then, often fail to transmit the introduced mutation to the germ line. Serial backcrossing of a progenitor 129 or other strains, e.g., C57BL/6, can be used to transfer the mutation to a more suitable “background”. For example, mice overexpressing melanin-concentrating hormone (MCH) have an obese phenotype only when backcrossed for 7 generations onto the C57BL/6J background. On an FVB background, they appear to have a normal phenotype with regard to body composition (21).

The obese ( ob ; now Lep ob ) mouse was identified as an autosomal recessive mutation in a noninbred strain (Stock V) at Jackson Laboratory in 1949 (6). Mice segregating for the obese gene were backcrossed for many generations to generate a congenic line on the C57BL/6J background strain. Lep ob mice on C57BL/6J, despite their early-onset obesity and transient glucose intolerance, are not diabetic; however, the Lep ob mutation coisogenic on the diabetes-prone C57BL/KsJ line results in severe, early-onset type 2 diabetes (22). Diabetes (db ; now Lepr db ) is a spontaneous recessive mutation that was first noted in a C57BL/KsJ mouse colony at the Jackson Laboratory. The KsJ Lepr db mutant is hyperphagic and obese, but also develops severe type 2 diabetes (7). Backcrossing Lepr db onto the C57BL/6J background attenuates the diabetic phenotype; C57BL/6J- Lepr db is virtually identical to C57BL/6J -Lep ob .

Douglas Coleman, at Jackson Labs, looking for the molecular predicates of the lipostatic system posited by Kennedy (23) and Hervey (24), performed parabiosis (joined circulation) studies coupling Lep ob mice to either wild-type or Lepr db mice (25). The Lep ob mouse became lean when joined to a wild type, but, when joined to a Lepr db mouse, the Lep ob mouse died of starvation. These findings led Coleman to hypothesize that a blood-borne factor regulating body weight might be deficient in Lep ob , but circulating at high levels in the blood of Lepr db mice. He suggested that obese was the secreted factor and diabetes its receptor (25,26). In 1994, the gene encoding Lep ob was isolated by positional cloning (27) by a group at Rockefeller University, and its product, leptin, was shown to be produced primarily in adipocytes. Leptin is a type 1 cytokine, similar in structure to IL-6. Lep ob mice lack circulating leptin by virtue of an R105X mutation that creates a premature stop codon sequence in the leptin gene, resulting in a truncated protein that is rapidly degraded (27).

The gene ( Lepr ) that encodes the receptor for leptin (LepR) was identified by expression cloning (28); the first genetic mutation in Lepr was identified in the Lepr db mouse (10,29,30), thus confirming the conceptual model proposed by Coleman: the obesity of the Lepr db mouse was due to a mutation that precluded leptin signaling by the receptor that binds the ob gene product. Alternative splicing of the Lepr transcript produces multiple isoforms of the receptor (which is a cytokine receptor similar to members of the IL6 receptor family): LepRa, -b, -c, -d, and so forth ( Figure 1 ). The mutation in the Lepr db mouse results from a splicing defect that causes the 3′ terminal exon (18a) of leptin receptor isoform a ( Lepr-a ) to be inserted into Lepr-b. A stop codon at the end of exon 18a prevents transcription of the Lepr-b terminal exon, so that LepRa is produced in place of LepRb (10,29,30). Because the Lepr db mouse synthesizes all leptin receptor isoforms except LepRb, it is clear that this isoform (which contains JAK box and STAT3 domains) is critical to the control of energy homeostasis (31). Indeed, restoration of LepRb on a background null for all other LepR isoforms restores energy balance (32).

Figure 1 . LepR isoforms and function . LepRa (Ra) represents the mostly highly expressed short form of LepR; LepRb (Rb) is the long form. Exon 17 contains half of a Jak docking site (BOX1) common to Ra, Rb and Rc, while exon 18b contains additional motifs required for full Jak2 binding (BOX2) and STAT3 signaling (31,33). Circulating leptin binding protein consists of extracellular domain that has been cleaved from the cell surface, along with the LepRe splice variant that lacks a transmembrane domain. Humans do not generate the splice variant, so that all LepRe is produced by cell surface cleavage, presumably by membrane associated metalloproteases (33).

LepRa, -c, -d and the other so-called “short” isoforms contain the same first 17 exons as LepRb, but diverge within the intracellular domain. LepRb is the only isoform that mediates classical Jak-STAT signaling, as this isoform alone contains the motifs required to interact with Jak2 and to bind STAT proteins for downstream signaling (Figure 1)(34). While the function of LepRb is clear, the functions of the short isoforms are not, although they have been speculated to function in leptin transport into the brain and/or a source of cleaved, circulating extracellular LepR (which, along with LepRe comprises the major circulating leptin-binding protein) (35). The biological role(s) of soluble LepR isoforms (sLEPR) are unclear. Human obesity and fasting are associated with decreased circulating sLEPR; pregnancy with increased sLEPR. sLEPR can block LEP transport across the BBB(36-40).

Disruption of Lep function results in hyperphagia and obesity, and leptin administration to Lep ob mice (but not Lepr db animals), reduces food intake and adiposity, sparing lean tissue (41-43). Thus, Lep ob and Lepr db mice demonstrate that fat mass (along with both energy intake and expenditure) can be controlled by a single molecule. Leptin represents a powerful biologic controller of feeding and energy balance, revealing the existence of an endocrine system that controls feeding and energy balance. In humans, leptin deficiency also elicits a severe obesity phenotype: A rare, recessively inherited LEP mutation was discovered in two children who are members of a highly consanguineous Pakistani family (44). As with the Lep ob mutation in mice, this frameshift mutation introduces a premature stop codon that truncates the leptin protein. While rare, additional leptin-deficient individuals (all of whom are severely obese) have been identified. Daily subcutaneous administration of recombinant leptin dramatically and selectively reduces body fat to normal levels in these individuals (45). A few humans homozygous for LEPR mutations have also been identified; these individuals present a severe obese phenotype similar to those lacking leptin, although – as anticipated - they are not responsive to exogenous leptin (46). In these patients, growth hormone deficiency and central hypothyroidism are phenotypes seen more frequently than in leptin deficiency per se . It is important to note that mice (47) and humans (48) heterozygous for null mutations of either LEPR or LEP are more obese than suitable controls. It is thus possible that individuals heterozygous for functionally null mutations of these and other genes encoding molecular components of the various signaling pathways regulating energy homeostasis discussed in this review constitute a significant proportion of the very obese. Additionally, heterozygosity for several of these mutations would be expected to produce even greater levels of obesity. The increasing use of exome sequencing in evaluating instances of severe obesity will lead to the detection of more instances of obesity caused by such oligogenic mechanisms.

While the role for leptin in the control of appetite and adiposity initially dominated the thinking about its biology, it rapidly became clear that leptin has other functions, and that the effects of high leptin are not as dramatic as those of low leptin. Indeed, obese rodents and humans exhibit high circulating concentrations of leptin, commensurate with their high levels of leptin-producing adipose tissue (49,50). Similarly, in contrast to the Lep ob mice, increasing leptin to supraphysiologic levels in normal animals modestly and briefly blunts food intake and body weight [effect may be more striking than this]. Likewise, supraphysiological doses of leptin have only modest effects on body weight in obese and non-obese humans (51). Thus, the absence of leptin appears to convey a more powerful signal than does its excess. Also, Lep ob mice (and their human counterparts) display additional phenotypes, including impaired growth and gonadal axis function, diminished immune function, infertility, and decreased energy expenditure due to low sympathetic nervous system tone and thyroid function- all of which are reversed by leptin treatment (52). The lack of leptin also promotes increased hepatic glucose production, and leptin treatment suppresses hyperglycemia in models of several diabetes, including T1D (53). This constellation of phenotypes resulting from low leptin mirrors the physiologic response to starvation; indeed, leptin treatment attenuates many of these consequences of very low adiposity (54). Thus, normal leptin concentrations signal the repletion of energy (fat) stores to mitigate hunger and enable energy expenditure, while low leptin indicates the dearth of adipose reserves and promotes food-seeking and the conservation of remaining fat by reducing energy expenditure. The concentration of leptin constituting such a signal of adequacy of fat stores may differ among individuals, reflecting genetic, developmental and acquired differences in the CNS molecules and circuits comprising this system (55).

Transgenic animals that lack adipose tissue exhibit a syndrome that mirrors that of lipodystrophic humans (who lack adipose tissue on a congenital or acquired basis): In spite of their leanness, lipodystrophic people and animals exhibit hyperphagia along with a predisposition to insulin resistance, diabetes and other endocrine and metabolic abnormalities that are not corrected even with caloric restriction (56,57). Due to their dearth of adipose tissue, leptin levels are low and leptin treatment improves their hunger and endocrine/metabolic abnormalities. Indeed, leptin was recently approved for the treatment of lipodystrophy syndromes in humans (58).

The similar phenotypes of Lep ob and Lepr db mice (along with the inability of leptin to alter physiology in Lepr db mice) indicates that leptin action on LepRb-expressing cells must mediate its effects. Consistent with its behavioral effects (e.g., on feeding) and its effects on the neuroendocrine and autonomic systems, most LepRb-expressing cells lie in the brain. Indeed, transgenic overexpression of LepRb throughout the central nervous system (CNS) partially corrects the obesity syndrome of Lepr db-3J mice (which lack all LepR isoforms) (32). Similarly, ablation of CNS LepRb using a neuron-specific Cre in combination with a floxed ( Lepr flox ) allele promotes hyperphagia, neuroendocrine failure, and obesity (59). Some tissues outside of the CNS express LepRb, but the physiologic role for leptin action on these non-CNS cells remains unclear.

Within the brain, the majority of LepRb-expressing neurons are found within the hypothalamus and brainstem, consistent with the known roles for these structures in the control of feeding and endocrine and autonomic function (60-62). While LepRb ablation in the nucleus tractus solitarius (NTS) in the brainstem reduces satiety (consistent with the known role of this brain structure), pan-hypothalamic ablation of LepRb promotes a phenotype very similar in quality and magnitude to that of whole-body null Lepr db animals (63). Furthermore, ablation of LepRb from broadly-distributed hypothalamic vGat - or Nos1 -expressing neurons promotes dramatic hyperphagia and obesity (64,65). Smaller, more circumscribed sets of hypothalamic LepRb neurons have also been implicated, as well. Within the arcuate nucleus (ARC), an important satiety center in the brain, excision of Lepr flox by Pomc cre and Agrp cre modestly increases feeding and adiposity (66,67). Ablation of LepRb in the SF1-expressing ventromedial hypothalamic nucleus (VMH) blunts the increase in energy expenditure that accompanies increased adiposity, and deletion of SF1 in the lateral hypothalamic area (LHA, which is associated with motivation) diminishes motor activity and promotes obesity (68,69). LepRb neurons in the ventral premammillary nucleus (PMv) play roles in reproduction (70). Importantly, many additional groups of LepRb cells in the hypothalamus (especially the ARC and dorsomedial hypothalamic nucleus (DMH)) and brainstem have as yet undetermined functions.

Expression of the agouti gene ( a ) normally occurs intermittently in the hair follicle resulting in the production of alternate yellow and black pigment bands of the resulting hair; this admixture produces the agouti coat color (71). The molecule acts as primarily as an inverse agonist at the melanocortin receptor (MC1R in skin).

The Yellow mutation of the agouti locus ( A y /a ) is also termed ‘lethal yellow’, since homozygotes for the allele are prenatal lethal. Yellow was bred by mouse fanciers in Europe beginning in the 1800s, and was notable for the dominant inheritance of its striking yellow coat color and obesity proportional to the intensity of the yellow coat (5). In 1960, another spontaneous agouti mutation was detected in the Jackson Laboratory colony; viable yellow ( A vy ) (72). The original, lethal, yellow mutation is a deletion of the Raly gene, which causes a fusion of the constitutively active Raly promoter to the agouti gene, resulting in ectopic continuous overexpression of agouti in all somatic (including brain) cells. A vy /a is also the result of ectopic overexpression of agouti ; this mutation results from insertion of a retrovirus-like repetitive intracisternal A particle (IAP) into a noncoding exon of agouti . The resulting splice variant fuses the constitutively active Raly promoter to the agouti gene, allowing constitutive overexpression of agouti in all somatic (including brain) cells.

The increased body weight of A y /a and A vy /a mice results mainly from hyperphagia, and reflects both increased fat mass and lean body mass (with increased body length) (73). By contrast, Lep ob and Lepr db mice have a selective expansion of fat mass due to increased food intake, decreased energy expenditure, preferential storage of excess calories as fat, decreased body length and lean body mass (74). Thus, agouti overexpression affects food intake similarly to leptin, but alters energy expenditure less dramatically. The agouti gene encodes agouti signaling protein (ASP), a peptide with a high affinity for melanocortin receptors. The yellow coat color of the A y /a mouse results from continuous overexpression of agouti in the skin which blocks (mainly by inverse agonist effects) alpha-melanocyte-stimulating hormone (a-MSH) signaling at melanocortin-1 receptors (MC1R) in the hair follicle (71,75). Since a-MSH activates melanocytes to initiate synthesis of eumelanin (black pigment) instead of phaeomelanin (yellow pigment), antagonism of a-MSH by ASP elicits a yellow coat color. ICV administration of a-MSH and a-MSH agonists decreases food intake and body weight; overexpression of agouti in the A y /a brain antagonizes the anorectic action of a-MSH signaling as well as blunting the endogenous activity of the receptor, thus causing hyperphagia.

The brain contains two predominant melanocortin receptor isoforms- melanocortin receptor-3 and -4 (MC3R and MC4R, respectively) (76). Both isoforms are potently activated by a-MSH. MC4R is expressed in the PVN, DMH, VMH and LHA(75), all of which are hypothalamic sites crucial for the control of food intake. Mice homozygous for a targeted deletion of Mc4r display substantial hyperphagia, 3- to 5-fold increased adiposity, and 50-100% increased body weight compared to littermate controls, while maintaining the same absolute lean body mass as +/+ littermates (77). Heterozygosity for the Mc4r null mutation elicits an intermediate phenotype. Mc4r -/- mice also have increased linear growth, as is characteristic of A y /a mice (77). Mc4r -deficient mice maintain core body temperature when exposed to a cold challenge (78), suggesting that sympathetic tone is not reduced to the same extent as it is in Lep ob and Lepr db mice. Oxygen consumption of Mc4r -/- mice is reduced by 20% as compared to weight-matched controls, however, indicating that MC4R mediates some control of energy expenditure, in addition to affecting feeding.

Approximately 4% of morbid human obesity (BMI > 40 kg/m 2 ) is due to mutations in MC4R (79-81). Preserved lean mass and increased stature are also evident in the human MC4R deficiency syndrome, as in rodent models (82). Most obesity associated with MC4R mutations has been attributed to heterozygosity for such mutations (83). Severe childhood obesity results from a null MC4R receptor, generated by missense, frame shift, deletion, and nonsense mutations (82). MC4R mutations are codominantly inherited, and heterozygous family members are overweight, suggesting that these mutations impair the function of the normal gene product, unlike null mutations in Mc4r -/- mice. Genome-wide association studies (GWAS) have revealed common non-coding polymorphisms within MC4R that are associated with increased adiposity (84). These are likely variants affecting the transcription rate of the gene.

An homology search to identify a protein with a normal physiological function comparable to ASP in the brain revealed a candidate with 25% amino acid homology to ASP (Agouti-related peptide; AgRP) (71). Eutopic Agrp expression is restricted to a set of neurons in the ARC that contain the orexigenic neuropeptide Y (NPY), and which are activated by fasting or leptin deficiency, consistent with a role in controlling (promoting) food intake (85). In vitro binding studies showed that AgRP binds the MC1, MC3 and MC4 receptors, and mice globally overexpressing AgRP (like Agouti mice) are hyperphagic and obese compared to nontransgenic littermates (86). AgRP differs from ASP in that it does not block eumelanin synthesis to elicit a yellow coat color when transgenically overexpressed in mice. MC4R exhibits significant constitutive activity in vitro ; ASP and AgRP not only block binding of a-MSH to the MC4R, but also suppress MC4R constitutive signaling, i.e., act as inverse agonists (71). In support of a physiological role for the orexigenic action of AgRP, fasted animals show increased expression of ARC Agrp mRNA, as do Lep ob and Lepr db mice (86). ICV administration of AGRP to rats elicits a long-lasting hyperphagic response (71).

Interestingly, however, mice congenitally null for Agrp or Npy exhibit minimal alterations in energy balance, as do compound Npy -/- ;Agrp -/- mice (87). This lack of phenotype apparently reflects developmental compensation/reprogramming, however, since mice in which AgRP neurons are ablated early in development exhibit normal energy balance, while ablation of these cells in adults results in aphagia and death by starvation (88,89). Thus, AgRP likely plays an important physiologic role in the promotion of feeding, presumably by blocking melanocortin receptor action.