MrgX2 Is a High Potency Cortistatin Receptor Expressed in Dorsal Root Ganglion*

MrgX2 is a recently identified orphan G-protein-coupled receptor whose ligand and physiological function were unknown. Here we describe cortistatin, a neuropeptide for which no specific receptor has been identified previously, as a high potency ligand at MrgX2. Cortistatin has several biological functions including roles in sleep regulation, locomotor activity, and cortical function. Using a “reverse pharmacology” approach, we have identified a number of additional cyclic peptide agonists for MrgX2, determined their rank order of potency, and demonstrated that this receptor has a pharmacological profile distinct from the other characterized members of the Mrg (Mas-related genes) family. In MrgX2-expressing cells, cortistatin-stimulated increases in intracellular Ca2+ but had no effect on basal or forskolin-stimulated cAMP levels, suggesting that this receptor is Gq-coupled. Immunohistochemical and quantitative PCR studies show MrgX2 to have a limited expression profile, both peripheral and within the central nervous system, with highest levels in dorsal root ganglion.

G-protein-coupled receptors (GPCRs) 1 make up the largest and most diverse family of transmembrane proteins, which respond to a wide variety of stimuli including biogenic amines, peptides, bioactive lipids, hormones, and light. Agonist binding to these receptors activates intracellular signaling events, mediated by G-proteins, such as modulation of adenylate cyclase or Ca 2ϩ mobilization (1,2). Completion of the human genome sequencing project has identified ϳ140 "orphan" GPCRs for which the ligand and function are unknown (3)(4)(5). We have used bioinformatic and tissue distribution analysis to prioritize those orphans with potential therapeutic relevance followed by a "reverse pharmacology" approach to identify cognate and surrogate ligands (6,7). The orphan receptor MrgX2 is a member of a family of ϳ5 human, 31 mouse, and 2 rat coding sequences that are related to the Mas1 oncogene (8). This group of genes has been termed Mrg (Mas-related genes) or SNSR (sensory neuron-specific G-protein-coupled receptors). Tissue distribution studies have shown these receptors to be expressed mainly in the small sensory neurons of the dorsal root ganglia (DRG), suggesting a role in nociception. To date, cognate ligands have been identified for five of these receptors: proenkephalin A gene products, especially BAM22 (EC 50 ϳ13 nM), have been shown to potently activate human MrgX1 (SNSR3) (9). The neuropeptide RF amides are agonists at murine MrgA1 (NPFF, EC 50 ϳ200 nM), MrgC11 (NPFF, EC 50 ϳ54 nM), and MrgA4 (NPAF, EC 50 ϳ60 nM) (10). Adenine has been identified as a high affinity ligand (K i ϳ18 nM) for a rat member of the Mrg family (11). However, none of these ligands are high potency agonists at MrgX2.
In an attempt to identify the cognate ligand for MrgX2, we screened a library of ϳ1200 endogenous GPCR ligands including small drug-like molecules, peptides, and bioactive lipids.
Here we describe cortistatin-14 as a high affinity ligand for MrgX2 and examine the effects of other cyclic peptides at this receptor. In addition, immunohistochemistry and quantitative PCR were used to describe the tissue distribution of MrgX2 and determine co-expression with cortistatin.

EXPERIMENTAL PROCEDURES
Fluorometric Imaging Plate Reader Assay (FLIPR®)-HEK293-G␣15 cells, obtained from Aurora Biosciences Inc. (San Diego, CA), and native HEK293 cells were grown in Dulbecco's minimum essential medium supplemented with 10% fetal calf serum and L-glutamine (2 mM) at 37°C in 5% CO 2 , 95% air. These cells were transiently transfected with the expression plasmid encoding MrgX2 (coding sequence as GenBank TM AY042214) using LipofectAMINE Plus (TM) reagent (Invitrogen), according to the manufacturer's instructions. Twenty-four h after transfection, the HEK293 and HEK293-G␣15 cells were separately seeded into black clear-bottomed, 96-well, poly-D-lysine-coated plates at a density of 5 ϫ 10 4 cells/well. The cells were cultured for a further 24 h. The medium was then removed from the cells and replaced with 100 l of warm (37°C) dye loading solution (50 g of Fluo4 (Molecular Probes) in 20 l of Me 2 SO ϩ 20 l of 20% pluronic acid, added to 11 ml of serum-free Dulbecco's modified Eagle's medium containing 2.5 mM probenecid). The plates were then incubated for 1 h at 37°C. The cells were subsequently washed three times with 150 l of wash buffer/well (Hank's balanced salt solution, 2.5 mM probenicid, 0.1% bovine serum albumin, pH 7.4). The plates were incubated at room temperature for 30 min prior to processing within the FLIPR®. Small molecule ligands and bioactive lipids were screened at a concentration of 10 M, and peptides were screened at a concentration of 1 M. Peptides were diluted to the appropriate concentration in Hank's balanced salt solution, 0.1% bovine serum albumin. For dose-response studies, all data points were carried out in triplicate. Mock-transfected cells (transfected with empty plasmid) were screened in parallel to MrgX2-transfected cells.
Cyclic AMP-responsive Element (CRE)-Luciferase Assay-HEK293 cells were cultured to ϳ80% confluence, at which time they were cotransfected with plasmids pCRE-Luc (CRE 4 -luciferase, Stratagene) and a construct containing MrgX2 using the lipid transfection reagent LipofectAMINE Plus (TM) (Invitrogen) as recommended by the manufacturer. Cells were cultured for 24 h after transfection before being washed with phosphate-buffered saline (pH 7.4), recovered in Dulbecco's modified Eagle's medium supplemented with 10% (v/v) fetal calf serum, and seeded as 100-l aliquots (5 ϫ 10 4 cells) into white 96-well tissue culture plates. Following a further 24 h incubation, 3-isobutyl-* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1-methylxanthine was added to a final concentration of 100 M, and the cells were incubated at 37°C for at least 30 min. Drug treatments were added in a 20-l volume to triplet wells. For cAMP inhibition studies, this was followed by the addition of forskolin to a final concentration of 5 M. The plates were left to incubate for a minimum of 5 h. The cells were then equilibrated to room temperature. 100 l of SteadyGlo luciferase detection reagent (Promega) was added, and the cells were incubated at room temperature for at least 5 min. The level of luminescence was quantified on a Tecan Ultra (TM) Reader (100 ms/well).
Taqman Tissue Distribution Analysis-300 nM of the forward and reverse primer and 150 nM of probe were used with 2ϫ TaqMan Universal PCR Master Mix (Applied Biosystems). The MrgX2 primer/probe set is as follows: MrgX2-Forward, 5Ј-tgtggcaaggagaccctga-3Ј; MrgX2-Reverse, 5Ј-ttcctaccagcccgaccag-3Ј; MrgX2-Probe, 5Ј-6-carboxyfluorescein-cccggtcttcctgatccttttcattgc-6-carboxy-N,N,N,N-tetramethylrhodamine-3Ј. The cortistatin primer/probe 96-well set is as follows: cortistatin (CST)-Forward, ggaaaagcagcctcctgacttt; CST-Reverse, cccggcactggcctg; CST-Probe, ctcgcttggtggtttgagtggacctc. One ng of cDNA from Multiple Tissue cDNA (MTC™) panels, Human I and Human II (Clontech), was used as the template. In addition, total RNA was isolated from human thoracic, lumbar, and cervical DRG, and cDNA was generated using the Superscript double-stranded cDNA synthesis kit (Invitrogen). The PCR reaction was run on the TaqMan 7700 sequence detection system as a single reporter, real time quantitation with 25-l reaction volume and every point in triplicate. PCR conditions were as follows: 50°C for 2 min, 95°C for 10 min, and 95°C for 15 s, 60°C for 1 min, ϫ 40 cycles. The data were analyzed by the sequence detection software with the baseline set at 3-15 cycles. The results file was exported to Excel (Microsoft), and the data are expressed as copies per nanogram of template and normalized to ␤-actin.
Immunohistochemistry-Immunohistochemistry was carried out by Lifespan Biosciences Inc. as part of subscription to their GPCR gene expression data base (www.lsbio.com).

RESULTS
Functional Activation of MrgX2-We initially screened HEK293-G␣15 cells transiently transfected with MrgX2 against a library of 1200 known GPCR ligands using the measurement of intracellular calcium concentration to detect activation of the receptor. Dose-response studies on the activating ligands showed cortistatin-14 (CST-14) to be the highest potency agonist with an EC 50 of 25 nM (Fig. 1). A number of other cyclic peptides showed high nanomolar potency at MrgX2 (Ta-ble I). No significant change in efficacy was seen in response to these agonists. No response to these peptides was seen in the mock-transfected cells.
To investigate the downstream signaling of MrgX2, we mon- itored the increase or decrease of intracellular cAMP in response to MrgX2 activation, using a CRE luciferase reporter system. No changes in cAMP levels were observed in MrgX2transfected cells in response to cortistatin, whereas the same transfected population of cells elicited a potent response to CST-14 in the FLIPR® Ca 2ϩ assay. To further define the native signaling mechanism of MrgX2, the receptor was transfected into native HEK293 cells, and the response to CST-14 was determined in the FLIPR® Ca 2ϩ assay. An EC 50 of 25 nM was obtained, which was equivalent to the value generated using the HEK293-G␣15 cells. These results taken together indicate that this receptor is not G s -or G i -coupled and that it signals via a G q -coupled mechanism.
Human Tissue Distribution Studies-Quantitative PCR was used to determine the distribution of MrgX2 and cortistatin in human tissues. Highest levels of MrgX2 were observed in lumbar DRG with high transcript levels also seen in DRG isolated from thoracic and cervical regions (Fig. 2C). Moderate levels were detected in testis, small intestine, spinal cord, pancreas, lung, heart, thymus, and colon ( Fig. 2A). Overlapping distribution of the receptor and the ligand cortistatin was seen in DRG, testis, and pancreas (Fig. 2, B and 2C). In addition, immunohistochem- istry studies were carried out across a wide range of normal and diseased human tissues using MrgX2-specific anti-peptide antibodies. High levels of receptor expression were detected in DRG samples with the small diameter neurons staining strongly and the large neurons showing moderate staining (Fig. 3). Within the hippocampus, neurons in areas CA2, CA3, and CA4 stained faintly, whereas neurons in area CA1 were negative. Staining was also identified in subsets of neurons in the substantia nigra, hypothalamus, and medulla (data not shown). No MrgX2 was detected in the cerebral cortex. Within the small intestine, faint staining was observed in the endothelium and vascular smooth muscle. Testis samples showed faint to moderate staining of spermatocytic precursors (data not shown). DISCUSSION We have identified cortistatin-14 (CST-14) to be a high affinity ligand for the orphan receptor MrgX2. Cortistatin is known to activate the five somatostatin receptors (SSTR1-5) with a high potency comparable with that of somatostatin (IC 50 range, 0.09 -18.2 nM) (12,13), but to date, no cortistatin-specific receptor has been described previously. In humans, CST is a cyclic peptide that is derived from a prepropeptide of 114 residues (14) and has been shown to play a role in induction of slow wave sleep, inhibition of locomotor activity, and inhibition of cell proliferation (15,16). Cortistatin-14 shares 11 of its 14 amino acids with somatostatin-14, but despite this high similarity, they are products of distinct genes. Cortistatin mediates several of its effects via activation of somatostatin receptors (SSTR1-5), but it also has a number of distinct biological activities, raising the possibility of the existence of a cortistatinspecific receptor (17). Unlike somatostatin, cortistatin increases slow wave sleep with no effect on rapid eye movement (REM) sleep. Cortistatin has been shown to directly inhibit hippocampal pyrimidal neurons, and this may be a mechanism by which slow wave sleep is enhanced. Immunohistochemical studies show moderate levels of MrgX2 expression in subsets of neurons in the CA2, CA3, and CA4 regions of the hippocampus, where it may act as a cortistatin-specific receptor and mediate the distinct effects of cortistatin. The low potency MrgX2 agonists HS024 and HS014 (cyclic ␣-melanocyte-stimulating hormone analogues) are thought to have sedative effects at high doses (18), and it is interesting to speculate on the role of MrgX2 in this process, especially in conjunction with the evidence for the sleep-inducing effects of cortistatin via a nonsomatostatin receptor mediated mechanism (17).
Structure-activity relationship studies on the cortistatin peptide show that the extracyclic N-terminal proline and the C-terminal lysine amide are essential for its distinct biological properties (19). This suggests that both of these residues are required for binding of cortistatin to its specific receptor. Interestingly, MrgX2 exhibits high affinity for CST-14 and CST-17, which contain both the N-terminal proline and the C-terminal lysine amide. However, the ligand affinity drops dramatically for somatostatin-14, which lacks these two critical amino acids but shares 11 of its remaining 12 residues with cortistatin. The somatostatin analogue octreotide is unable to functionally activate MrgX2 further, indicating that the structural determinants required for MrgX2 activation are located within the extreme N-and C-terminal residues of the peptide (Fig. 4).
Within the brain, the highest levels of cortistatin expression are found in the cerebral cortex. MrgX2 is not detected in the cortex by quantitative PCR or by immunohistochemistry and thus is unlikely to play a role in the modulation of cortical activity. Quantitative PCR studies show cortistatin to have a broad tissue distribution as compared with the limited expression of MrgX2. This is not unexpected as a number of the diverse actions of cortistatin have been shown to be mediated via interactions with the widely distributed somatostatin family of receptors (17). However, as discussed previously, the existence of a cortistatin-specific receptor is suggested by pharmacological evidence that cortistatin exerts some of its effects via somatostatin receptor-independent pathways. Quantitative PCR and immunohistochemistry data show that MrgX2 is expressed at highest levels in the small diameter neurons of the DRG (Fig. 3), which include primary sensory fibers associated with acute and neuropathic pain. We have also detected the presence of cortistatin mRNA in human DRG by quantitative PCR, and it is interesting to speculate that this ligand-receptor pair may have a role in nociception, but this has yet to be established. However, it may prove challenging to use an animal model to define the physiological effects of MrgX2 activation by cortistatin as bioinformatic similarity searches have shown no mouse or rat orthologues of MrgX2. This divergent cross-species receptor evolution is not unique to the Mrg family. The recently ligand-paired human receptor GPR8 has been shown to be absent in rodents (20), although its high affinity ligand, neuropeptide W, is present in these species (21). Despite these challenges, the identification of MrgX2 as the first human cortistatin-preferring receptor, together with the colocalization of receptor and ligand in the DRG, should promote investigation into the role of cortistatin and its receptor in nociception.
MrgX2 is the sixth member of the Mrg family to be ligandpaired. The neuropeptides NPFF, NPAF, and BAM22 and the purine adenine have been shown to be agonists of MrgA1, MrgA4, MrgX1, and a rat Mrg receptor, respectively. Recently, peptides containing the C-terminal consensus sequence ϪRF(Y)G or ϪRF(Y) amide have been identified as agonists at the MrgA1 and MrgC11 receptors. BAM22 and NPFF activate MrgX2 with an EC 50 of ϳ400 nM, whereas NPAF and adenine are inactive at this receptor. All of these receptors share 45-65% amino acid identity with MrgX2 and exhibit overlapping expression within small diameter fibers of the DRG. However, the structure activity relationships and rank order of potency of their respective ligands are distinct. Thus nociceptive sensory neurons show not only molecular diversity but also a surprising degree pharmacological diversity, suggesting the involvement of the Mrg family in subtle regulation and fine-tuning of nociceptive responses.
Our data have contributed additional facets to the already complex pharmacology of the Mrg family. The future development of MrgX2-selective agonists and antagonists will further elucidate the physiological role of this receptor and its involvement in the biological functions of cortistatin.