Editorial: Hypothalamic Melanocortin Signaling in Cachexia

Understanding of the regulation of appetite and satiety has been revolutionized by the discovery of leptin (1), a white adipose tissue-derived cytokine that is secreted into the bloodstream. Overall, a picture of normal energy homeostasis has emerged wherein leptin acts on the central nervous system to help orchestrate the regulation of food intake and caloric disposition, and conversely, the neuroendocrine and metabolic adaptation to inadequate nutrient availability (2). To summarize this concept, during periods of nutrient abundance, leptin secretion is increased, leading to decreased appetite and increased caloric disposal, whereas caloric deficit leads to decreased leptin secretion, increased appetite, and a shift to a neuroendocrine profile that facilitates metabolic adaptation (3). Nevertheless, recent studies of metabolic dysregulation syndromes have revealed important contributions of hypothalamic systems functioning independently of leptin, as illustrated by two recent studies of tumor-induced anorexia and cachexia, including the accompanying paper by Wisse et al. (4, 5). Cachexia is a common pathological syndrome associated with cancer and other chronic illnesses, that encompasses both the loss of appetite (anorexia) and the inability to conserve energy. Ultimately, there is loss of fat and lean body mass, which is the hallmark of the disorder, contributing to morbidity, mortality, and reduced quality of life in such patients (6). While the pathophysiology of cachexia is undoubtedly multifactorial (6, 7), elucidation of the mechanisms whereby appetite, satiety, and energy conservation are normally regulated have permitted new insights into the understanding of cachexia. These findings may have significant clinical and pharmacotherapeutic implications. At the core of the central metabolic regulatory system is the neuroanatomic organization of specific hypothalamic neuron groups that produce an expanding repertoire of classical and newly discovered neuropeptides. The primary site of leptin’s action is the central nervous system, where it appears to act via specific receptors (Ob-Rb) in the hypothalamic arcuate nucleus to influence the activities of specific neuropeptide-producing neurons (8, 9). At least two separate groups of neurons in the hypothalamic arcuate nucleus with opposing functions are responsible for the actions of leptin on the brain, POMC-producing neurons that also co-express cocaine and amphetamine-regulated transcript (CART), and agouti-related peptide (AGRP)-producing neurons that co-express NPY (9). These neurons send monosynaptic projections to identical target regions within the hypothalamic paraventricular nucleus and lateral hypothalamus, where the signals are integrated and then relayed by independent pathways to regions of the brain governing feeding behavior, energy expenditure, and neuroendocrine (hypophysiotropic) function (3, 8–10). When circulating leptin levels are suppressed, such as during fasting, expression of the genes encoding the anorexigenic peptides, POMC and CART, are reduced simultaneously with a marked increase in the genes encoding the orexigenic peptides, AGRP and NPY (3, 8–10). The close interaction between the opposing components of this regulatory system is underscored by the discovery that AGRP exerts its biological effects by the novel mechanism of acting on melanocortin receptors, both as a competitive antagonist and inverse agonist (11–14). Therefore, during fasting, an increase in AGRP cooperates in the down-regulation of melanocortin signaling by antagonizing the action of α-MSH, concurrently with an inhibition of POMC gene expression (15–17). The gain and loss of body weight, however, are only partly explained by net caloric intake resulting from the interactions of α-MSH and AGRP/NPY. Although most attention to date has focused on the regulation of feeding and satiety, equally important in determining energy balance are the mechanisms by which caloric disposition is controlled. The mechanisms involved in regulation of energy expenditure and catabolic processes are not yet well understood but include the coordinated effects of these very same peptides on central pathways governing the sympathetic nervous system and the thyroid axis (18). In addition to the peptides described above, a rapidly expanding list of other neuropeptides including melanin concentrating hormone (MCH), orexins (hypocretins), and ghrelin have been described, which may also contribute to the regulation of appetite and satiety (9). However, compelling evidence points to the melanocortin system as one of the principal regulators of body weight. As opposed to the NPY knockout mouse, which has little or no phenotypic or metabolic abnormalities (19), genetic alterations affecting POMC gene expression or melanocortin receptors (20, 21) result in profound physical, behavioral, and metabolic changes. Targeted deletion of the type 4 melanocortin receptor (MC4-R) in mice produces an obesity syndrome characterized by hyperphagia, hyperinsulinemia, and reduced energy expenditure (21, 22). Increased adiposity due to decreased energy expenditure is also caused by targeted deletion of the MC3-R (23, 24), another major melanocortin receptor subtype expressed in the brain (25). Similarly, in humans, mutations that interfere with the functions of the MC4-R, the POMC gene, or the processing enzymes necessary to generate a fully matureα -MSH, result in severe obesity (26–29). Maintaining adequate tone in the melanocortin signaling system, therefore, has an important role in the maintenance of normal body weight. This is highlighted by the recent finding that the rapid weight loss in animals that occurs after involuntary overfeeding is prevented by pretreatment with a melanocortin receptor antagonist, and in fact, the animals continue to gain weight (30). Loss of tone in the melanocortin signaling system as a result of senescence of the arcuate nucleus POMC neurons (31) has also been proposed as a mechanism to explain the tendency for weight gain with aging (32). The paper by Wisse et al. (4) brings attention to a seemingly paradoxic aspect of melanocortin signaling, not observed under usual circumstances. In this study, animals bearing prostate carcinoma cells that developed tumor-associated cachexia, significantly increased their food intake and gained weight after the intracerebroventricular administration of the MC3/4-R antagonist, SHU-9119. Similar findings were reported by Marks et al. (5) using AGRP (83–132), a synthetic fragment of the endogenous MC3/MC4 receptor antagonist, to prevent cachexia in mice induced by sarcoma tumors. Furthermore, MC4-R-deficient mice were relatively resistant to sarcoma-induced anorexia and weight loss, even with continued progression of the tumor (5). These observations indicate that despite marked loss of body weight, which would normally be expected to down-regulate the melanocortin signaling system as a way to conserve energy stores, during cancer-induced cachexia, the melanocortin signaling system remains active. Because leptin levels were appropriately suppressed (4), the findings cannot be attributed to increased circulating leptin levels as described in certain inflammatory states (33, 34). Maladaption of the melanocortin signaling system is not unique to cancer cachexia and has also been described in other anorectic states. Like tumor-induced anorexia, the anorexia induced by systemic treatment of rats with lipopolysaccharide (LPS) was reversed by central administration of the MC3/MC4 antagonist SHU-9119 (35), a finding that was confirmed and extended in mice (5). Similarly, anorexia developing in rats as a result of repeated immobilization stress (36), can be at least partially reversed by the administration of the MC4-R antagonist, HS014. The catabolic and clinical features of cachexia in chronic illness are quite distinct from other anorectic states (6), so that one should avoid a rush to overgeneralize. Nevertheless, based on these converging lines of evidence, it is tempting to speculate that aberrant melanocortin signaling may be a common contributing factor in anorexia and cachexia in a variety of chronic illnesses. The mechanisms contributing to persistent anorexigenic/cachexic activity of the central melanocortin system during illness are unknown. Proinflammatory cytokines are widely considered to be candidate common mediators of LPS- and tumor-induced anorexia (6, 7), prompting Wisse et al. (4) to speculate that elevated circulating cytokine levels in tumor-bearing animals may activate central melanocortin release. IL-1, IL-6, TNFα, CNTF, and leukemia inhibitory factor are all capable of producing anorexia (6, 7, 37), and as cytokines generally act in a cascade fashion, multiple candidates on this list, acting in concert, may contribute to the anorectic response. Indeed, there is abundant evidence that elevations of proinflammatory cytokines stimulate pituitary POMC synthesis and ACTH secretion (38), but available evidence for cytokine-mediated activation of hypothalamic melanocortin-producing neurons is presently limited. No increase in hypothalamic POMC mRNA levels was observed (39) in the tumor-induced cachexia model used by Wisse et al. (4), but increased hypothalamic POMC levels did occur following systemic LPS treatment (40), and an earlier study reported increased central release of α-MSH in febrile rabbits (41). In contrast, LPS treatment of rats at a dose exceeding that required to cause anorexia (35) suppressed thyroid axis activity and TRH gene expression in hypophysiotropic paraventricular neurons (42). Because centrally administered α-MSH increased neuronal TRH mRNA levels (43), these findings are not consistent with increased hypothalamic melancortin secretion during LPS-induced anorexia. It is conceivable, therefore, that cytokines selectively activate only a subset of POMC neurons within the arcuate nucleus that project exclusively to adipostatic neurons in the hypothalamus. Testing the effects of cytokines or tumor cell supernatants on identified POMC neurons in vitro (44) may shed some light on whether tumor-derived factors and cytokines do indeed stimulate melanocortin neurosecretion. An alternative or complementary mechanism that may contribute to the inappropriate maintenance of anorexigenic/catabolic melanocortinergic tone during inflammation and cachexia is increased hypothalamic sensitivity to these actions of melanocortins. Supporting this hypothesis is the observation that LPS-treated rats were far more sensitive to the anorexic effect of centrally administered α-MSH (i.e. responded to a 10-fold lower dose), and exhibited a more marked suppression of food intake, than did similarly treated controls (35). Moreover, these effects were not dependent on the severity of accompanying febrile responses, because the exogenous α-MSH suppressed LPS-induced fever concurrently with its potentiation of LPS-induced anorexia. Another example of cytokine-associated increases in central melanocortin responsiveness concerns the febrile response itself. α-MSH acts centrally via the MC3- or MC4-R to inhibit fever (46), but the animals are unresponsive to the thermoregulatory actions of antipyretic doses of α-MSH in the absence of fever (45–47). Therefore, induction of melanocortin responsiveness, rather than activation of POMC neurons alone, is involved in the antipyretic effects of both exogenous and endogenous melanocortins. Mechanisms that could account for such sensitization to the anorexigenic and cachetic effects of melanocortins during illness states are speculative, but several possibilities come to mind. Up-regulation of hypothalamic melanocortin receptor expression or coupling efficiency might occur, but there is as yet no empirical basis to support this. Reduction in the antagonism of MC3/4 receptors by hypothalamic AGRP- containing neurons might also occur, as recent evidence supports a potential role of deficient AGRP signaling in a rodent model of anorexia nervosa. Namely, the anorexic mutant (anx/anx) mouse is deficient in AGRP (as well as co-expressed NPY) projections to hypothalamic centers associated with feeding and satiety (48). Finally, certain single nucleotide polymorphisms in the human AGRP gene are found with higher frequency in anorexia nervosa patients as compared with controls (49), raising the possibility that some alleles that encode AGRP isoforms result in decreased antagonist or inverse agonist activities at the MC4-R. Whether some of these isoforms could potentially predispose individuals to the development of cachexia in association with cancer, merits further investigation. Existing therapeutic options for the treatment of cachexia, largely based on the use of megestrol and medroxyprogesterone, have been of limited efficacy. The observations by Wisse et al. (4) and others (5, 36) that the central melanocortin signaling system contributes to animal models of cachexia, therefore, should encourage the development of small, nonpeptide melanocortin receptor antagonists for human investigation. Such therapy may be applicable not only for individuals with cancer, but also individuals with chronic debilitating infections and anorexia nervosa. Since in experimental animals antagonists of the melanocortin receptors have not been associated with tachyphylaxis (17), analogs for human administration may be useful for long-term therapy. Given that melanocortin receptors are widely distributed in the central nervous system and mediate a variety of responses in addition to appetite and satiety (50), however, one must consider the possibility that inhibition of melanocortin signaling, even if selectively targeted to the MC3- and MC4-R, could have other unintended effects. On the one hand, this could have beneficial effects given that melanocortins increase neuropathic pain in animal models (51) and antagonize opiate analgesia (52). Therefore, patients with cancer might have the additional benefit of improved pain control when treated with melanocortin receptor antagonists. On the other hand, because α-MSH stimulates TRH gene expression in hypophysiotropic neurons and activates the thyroid axis (41), MC3/4-R antagonists might exacerbate the hypothyroid state (nonthyroidal illness syndrome) already present in individuals with debilitating disease (53). Antagonism of the antipyretic effects of melanocortins (44) also could potentially magnify febrile responses in patients receiving chemotherapy, or those who develop secondary infections. Nevertheless, the new findings provide cause for optimism that novel therapeutic strategies based on selective targeting of melanocortin receptors could qualitatively improve the debilitating effects of cachexia in chronic illnesses, significantly improving quality of life for these patients.