Overcoming Stress, Hunger, and Pain: Cocaine- and Amphetamine-Regulated Transcript Peptide’s Promise

Abstract

Cocaine- and amphetamine-regulated transcript encodes an eponymous peptide, CARTp, which exerts diverse pharmacologic actions in the central and peripheral nervous systems, as well as in several endocrine organs, including pancreas. Here we review those diverse actions, the physiological relevance of which had remained unestablished until recently. With the identification of a CARTp receptor, GPR160, the physiologic importance and therapeutic potential of CARTp or analogs are being revealed. Not only is the CARTp-GPR160 interaction essential for the circadian regulation of appetite and thirst but also for the transmission of nerve injury-induced pain. Molecular approaches now are uncovering additional physiologically relevant actions and the development of acute tissue-specific gene compromise approaches may reveal even more physiologically relevant actions of this pluripotent ligand/receptor pair.

The initial excitement that follows the identification of a previously unrecognized, endogenous hormone continues to grow when genetic mutations are created in experimental animals and the consequences of loss of function mutations are reported in human populations. All too often, however, that initial enthusiasm for the hormone’s potential as a therapeutic agent wanes because the development of clinically relevant ligands is stalled by the failure of efforts to identify a cognate receptor. Until such a receptor is identified, the field progresses along with a host of suggestive evidences, descriptions of localization, potential sites of action, and observations of changes in gene transcription or hormone levels under a variety of experimental conditions or human disease states. Enthusiasm is suspended by many in the field, but never completely abandoned by some investigators, particularly those who value the promise of the development of novel ligands for the treatment of common diseases that have not been controlled by available pharmacologic tools. One such example of this barrier to progress recently may have been overcome, and the promise of treatment of a wide variety of human disease states has been renewed.

Discovery and Early Descriptions of Cocaine- and Amphetamine-Regulated Transcript Peptide Actions

The transcript identified to be upregulated by psychostimulants in rodent brain, cocaine- and amphetamine-regulated transcript (CART) (1), was found to encode a large peptide, CARTp, which exerts multiple pharmacologic effects in rodents including in pancreas, as well as central and peripheral nervous systems (Fig. 1). But the physiological relevance of those actions remained uncertain until the recent identification of a CARTp receptor (2). Over 20 years ago, CARTp encoding messenger RNA was observed in rat islet tumors and normal islets of Langerhans (3), and in situ hybridization combined with immunostaining revealed the presence of CARTp in delta cells of the rat islet. CARTp-immunoreactivity was subsequently observed in pancreatic nerves (4,5). Although originally hypothesized to act within the islet to enhance proliferation during development, prior studies failed to identify proliferative effects in cultured INS-1 cells (4). Glucose homeostasis was altered in embryonic CART knockout (KO) mice (6); animals displayed impaired insulin secretion to glucose challenge at 40 weeks of age in vivo and impaired insulin secretion from harvested islets in vitro. CARTp was demonstrated to potentiate glucose-stimulated insulin secretion in islets treated with glucagon-like peptide-1 (GLP-1) or forskolin, but the peptide was, if anything, inhibitory to glucose-stimulated insulin secretion in the absence of agents that elevated protein kinase A activity in those tissues (7).

Figure 1.

Summary of diverse pharmacologic actions of CART peptide, described in experimental animals, discussed in the text. Asterisks indicate actions demonstrated to be physiologically relevant.

Emerging Evidence of Potential Physiological Relevance

More recently, Wierup and colleagues (8) provided evidence suggesting a physiologically relevant role for endogenous CARTp in beta cell function. Using a small interfering RNA approach they demonstrated CART silencing to alter the expression of transcription factors involved in insulin production and secretion. Thus, not only was insulin secretion compromised in CART KO animals, a potential role for endogenous CARTp in islet cell survival had been established. More work from the same group (9) revealed the expression of CART in alpha as well as beta cells and the upregulation of CART expression in islets from diabetic humans and mice. Importantly, that work also identified an action of CARTp to inhibit glucagon secretion from human and mouse islets leading these authors to suggest that CART or analogs of the peptide might be developed as antidiabetic agents. But one major problem remained that seemed to block the development of therapeutic agents: no unique CARTp receptor had yet to be identified.

An additional role for CARTp in endocrine function initially was proposed when an extensive distribution of CARTp-positive cells was identified in hypothalamus—in particular, in arcuate, paraventricular, supraoptic, and anterior periventricular nuclei (10). Koylu et al also identified CARTp-positive cells in the anterior pituitary gland, although the staining there was not as intense as that observed in nerve fibers in the adjacent neural lobe (10). Those observations, paired with the detection of CARTp-positive cells in the adrenal medulla, led the authors to hypothesize a role for CARTp in the hypothalamo-pituitary-adrenal (HPA) axis (10). An even more extensive characterization of CARTp staining subsequently characterized the distribution in both rat and human brains (11). That study revealed colocalization of CARTp with several other neuropeptides known to be involved in energy homeostasis and thus support for a role of CARTp-producing neurons in metabolism was introduced. A literature has developed indicating CARTp to exert significant effects on autonomic function (12-16), thermoregulation (17), energy homeostasis (18), and appetite (see following discussion). In terms of the potential actions of CARTp within the HPA axis, 2 groups independently identified actions of CARTp on neurons producing corticotropin-releasing hormone (CRH). Vrang and colleagues observed increased c-Fos staining in CRH and oxytocin neurons following central administration of CARTp and a concomitant increase in plasma corticosterone and oxytocin levels (19). Plasma glucose levels also increased following CARTp administration into the lateral cerebroventricle, which may have reflected activation of oxytocinergic projections to sympathetic centers in brainstem (19). Stanley and coworkers similarly administered CARTp centrally and observed, at a dose of the peptide that inhibited food intake, a significant elevation in plasma adrenocorticotropin and corticosterone levels (20). Plasma levels of prolactin increased transiently as did plasma growth hormone concentrations at a single time point. When a lower dose of CARTp was administered directly into the paraventricular nucleus, a transient but significant increase in plasma adrenocorticotropin, but not prolactin or growth hormone, levels was observed. These results were matched by hypothalamic explant studies where CARTp stimulated the release of CRH, neuropeptide Y, thyrotropin-releasing hormone, and alpha-melanocyte–stimulating hormone. In summary, anatomic and pharmacologic studies strongly suggested that CARTp was an important contributor to the control of the HPA axis, but once again, in the absence of potent selective antagonists, the physiological relevance could not be firmly established. A receptor would have to be identified.

Identification of a Potential CARTp Receptor

Our group had developed a strategy with which we identified potential receptors for neuronostatin (21) and phoenixin (22), 2 novel peptides we discovered using a bioinformatics approach (23,24). We also had been able with our deductive reasoning strategy to identify potential receptors for proinsulin-connecting peptide (25) and adropin (26). We were impressed by the broad pharmacologic portfolio of actions of CARTp and curious why no cognate receptor had been identified. There existed in the literature strong evidence that CARTp signaled via a G protein-coupled receptor (GPCR) (27-29), and thus we utilized our receptor identification approach to determine if one of the list of orphan GPCRs cataloged by International Union of Basic and Clinical Pharmacology (30) might be a receptor for CARTp. We employed both in vivo and in vitro approaches in our search. The existing literature suggested a role for CARTp in the transmission of neuropathic pain (31-34), including peptide expression in dorsal root ganglion of the lumbar region and pain responses elicited by intrathecal administration of the CARTp. In an established model [chronic constriction of the sciatic nerve (CCI)] of pain, we searched for transcripts that might be elevated in dorsal horn laminae 1 and 2 of the spinal cord on the ipsilateral side of the injury, compared to the contralateral side, and identified several possible candidates. From that list of candidates, we then examined databases for nonorphan GPCRs, known to be elevated in pain states. Those receptor sequences were then used to search for homologous orphan GPCRs that were known to be expressed in tissues where CARTp had been demonstrated to exert pharmacologic effects. We quickly focused on the orphan GPCR, GPR160, as our top candidate. Indeed we characterized the increased expression of GPR160 on the ipsilateral side of the nerve injury by immunostaining, reverse transcription polymerase chain reaction, and RNA-Seq analysis (2).

Just demonstrating expression changes in the pain model wasn’t enough to convince us that GPR160 was necessary for the transmission of neuropathic pain. Instead, we compromised GPR160 expression using intrathecal injections of a small interfering RNA construct designed to selectively target the Gpr160 transcript and found that this approach not only prevented the onset of both spinal injury–induced CCI and spared-nerve injury–induced (35) mechano-allodynia but also reversed established pain in the CCI model (2). Passive immunoneutralization of endogenous GPR160 also resulted in a reversal of CCI-induced mechano-allodynia. The effects were observed in male and female animals. Our task then became the final link, that CARTp bound to GPR160. In human KATO III cells, knockdown of Gpr160 prevented the ability of CARTp to increase cfos expression and CARTp co-immunoprecipitated with GPR160 from KATO III cells membrane extracts. In addition, Gpr160 knockdown abrogated CARTp-induced extracellular signal-regulated kinase phosphorylation in cultured PC-12 cells. Physical association of CARTp with GPR160 was demonstrated using a proximity ligation assay. We believe we had identified at least 1 CARTp receptor (2).

Is GPR160 Necessary for Other Actions of CARTp?

Using the passive immunoneutralization approach, we presented evidence that CARTp requires the presence of GPR160 to exert its pharmacologic and physiologic anorexigenic actions (36). CARTp had long been known to exert anorexigenic action when administered into the ventricular system or rodents (17,37-41), and both hypothalamic and brainstem sites of action had been hypothesized. In fact, as early as 1998, Kristensen and colleagues had demonstrated that passive immunoneutralization of endogenous CARTp levels resulted in exaggerated normal, light-entrained food intake in rats (42), and more recently, Lee and colleagues demonstrated a similar effect of anti-CARTp antibody when injected into the nucleus solitarius of the rodent brainstem (43), a site of abundant Gpr160 expression (36). Those authors also established that the anorexigenic action of CART in brainstem was due to delivery of the peptide via vagal afferents to appetite-related areas adjacent to the area postrema (43) and that the anorexigenic effect of cholecystokinin required release of CARTp into brainstem feeding centers. Cholecystokinin isn’t the only gut hormone known to control the activity of vagal afferents; in fact, GLP-1 receptors are present in nodose ganglion and peripheral administration of the GLP-1 inhibits food intake via an effect on those afferents and possibly GLP-1 receptors in brainstem (44). CARTp interacting with GPR160 appears also to be important in the control of thirst (36,45), an effect that at least temporally differs from that on feeding (36,46).

Translational Potential for the Development of CARTp Analogs or Mimetics

Embryonic KOs of the CART gene in mice resulted in varying phenotypes. Wierup and colleagues (6) reported that CART KO mice, provided standard lab chow, displayed impaired insulin responses to glucose even before any significant changes in body weight were observed, compared to wild-type (WT) controls. In addition, pancreatic islets isolated from KO mice released less insulin in response to glucose than islets from WT mice. Another group (35) compared CART homozygous KO mice to heterozygous CART KO and WT mice on a high-fat diet. Homozygous male KO mice consumed more food and gained more bodyweight and fat mass than heterozygous male KO mice and WT mice. In females, both homozygous and heterozygous mice ate more food and gained more body weight and fat mass than WT females (35). These were global gene compromise models, and therefore the reason for the observed phenotypes can only be hypothesized. Was it due to the loss of the protropic effect on the beta cell or perhaps the loss of the inhibitory effect on glucagon secretion? Could it have been due to the loss of CARTp’s actions in the central nervous system? Even site-specific gene compromise of CARTp production might not address the issue of site of deficit since CARTp circulates in blood and those levels, as well as tissue concentrations, are affected by multiple factors including leptin (15) and adrenal steroids (47). However, these KO animals still have important value. It would be interesting to know if reintroduction of CART expression in these animals or simply daily or continuous infusion of exogenous peptide would reverse some of the phenotypes observed. If so, then there might be promise for treatment of individuals described next.

Current Challenges

As previously described, mutations in the gene encoding CARTp resulted in obese phenotypes. Could the same result from mutations in the gene encoding GPR160? Single nucleotide polymorphisms have been cataloged in human tumors (https://www.ncbi.nlm.nih.gov/snp/?term=GPR160); however, patient populations expressing mutations in Gpr160 have yet to our knowledge been described, at least in terms of a characteristic phenotype. We would hypothesize that if they did exist, the clinical presentation of those individuals would mirror that of the families described next. It is, however, possible that more than 1 receptor for CARTp exists, although compromise of GPR160 levels in rodents did abrogate both pharmacologic and pathophysiologic actions of CARTp (2). Just the same, at least 1 group has suggested that in addition to signaling via a G_i/o mechanism, CARTp may signal via a G_s mechanism (48), and therefore either GPR160 must signal via alternative G proteins dependent upon cell type and physiologic condition or more than 1 receptor may exist. This will require further validation.

Prospects for the Future

Information from human mutation studies revealed that alteration in CART prehormone production and processing resulted in profound metabolic and behavioral phenotypes, suggesting that therapeutic administration of CARTp or mimetics might provide some benefit. The first report of mutations in the CART gene detailed the result of a Leu34Phe in an adolescent with early onset obesity (49). A cosegregating history of obesity and diabetes was noted on his mother’s side of the family, as was a history of anxiety and depression (50). When this mutation was introduced into AtT20 cells, evidence suggesting altered processing of the CART prohormone was observed (51). Altered posttranslational processing and intracellular trafficking of CARTp subsequently was demonstrated by Yanik and colleagues (52) in AtT20 cells transfected with the mutant gene (Leu34Phe), and importantly, these authors demonstrated that circulating mature (ie, bioactive) CARTp levels were reduced in an obese kindred bearing the mutation. Indeed, work by some of the same investigators (53) revealed several single-nucleotide polymorphisms in the CART gene that correlated with extreme obesity. Another group reported in a separate population of children that the A1475G mutation was significantly associated with overweight and obesity (54). Thus, mutations in the CART gene have been identified, which cosegregate with obesity, and the underlying reason for the development of that phenotype might well be the altered processing and trafficking within the cell of the CART prohormone.

Could the metabolic consequences of low levels of mature bioactive CARTp be overcome by replacement of the missing peptide in these individuals? This alone highlights the importance of identifying a cognate CARTp receptor. Unless the gene deficit itself can be corrected, perhaps the best approach would be to see whether the administration of exogenous WT CARTp or a stable analog might ameliorate some of the devastating consequences of improper processing and trafficking of the CART gene product. The identification of a CARTp receptor now enables the development of selective and/or biased receptor ligands with which to either drive the CARTp/GPR160 tandem to inhibit appetite and address metabolic issues or, in the case of pain sensation, interrupt that signaling. Challenges to be overcome in those efforts will include issues of delivery and site-selective targeting, in addition to the pharmacokinetic properties of peptide analogs. Just the same, the possibility that safe, effective, and tissue-specific ligands could be developed will certainly contribute to our ability to gain additional insight into the cellular basis of several pathologies and, with any luck, enable the development of novel therapeutic agents.

Additional Information

Disclosures: WKS, DS, and GLCY are holders of patent no. 10,85,378 (Methods for Treating Pain Using Anti-GPR160 Antibodies).

Data Availability

Data sharing is not applicable to this article as no data sets were generated or analyzed during the current study.