Portal de Periódicos da CAPES

AXOR12, a Novel Human G Protein-coupled Receptor, Activated by the Peptide KiSS-1

2001; Elsevier BV; Volume: 276; Issue: 31 Linguagem: Inglês

10.1074/jbc.m102743200

1083-351X

Alison I. Muir, Larissa Chamberlain, Nabil A. Elshourbagy, David Michalovich, Darren J. Moore, Amy Calamari, Philip G. Szekeres, Henry M. Sarau, Jon Chambers, Paul R. Murdock, Klaudia Steplewski, Usman Shabon, Jane Miller, Susan Middleton, John G. Darker, Christopher Larminie, Shelagh Wilson, Derk J. Bergsma, Piers C. Emson, Richard L. M. Faull, Karen L. Philpott, David C. Harrison,

Neuropeptides and Animal Physiology

A novel human G protein-coupled receptor named AXOR12, exhibiting 81% homology to the rat orphan receptor GPR54, was cloned from a human brain cDNA library. Heterologous expression of AXOR12 in mammalian cells permitted the identification of three surrogate agonist peptides, all with a common C-terminal amidated motif. High potency agonism, indicative of a cognate ligand, was evident from peptides derived from the gene KiSS-1, the expression of which prevents metastasis in melanoma cells. Quantitative reverse transcriptase-polymerase chain reaction was used to study the expression of AXOR12 and KiSS-1 in a variety of tissues. The highest levels of expression of AXOR12 mRNA were observed in brain, pituitary gland, and placenta. The highest levels ofKiSS-1 gene expression were observed in placenta and brain. A polyclonal antibody raised to the C terminus of AXOR12 was generated and used to show localization of the receptor to neurons in the cerebellum, cerebral cortex, and brainstem. The biological significance of these expression patterns and the nature of the putative cognate ligand for AXOR12 are discussed. A novel human G protein-coupled receptor named AXOR12, exhibiting 81% homology to the rat orphan receptor GPR54, was cloned from a human brain cDNA library. Heterologous expression of AXOR12 in mammalian cells permitted the identification of three surrogate agonist peptides, all with a common C-terminal amidated motif. High potency agonism, indicative of a cognate ligand, was evident from peptides derived from the gene KiSS-1, the expression of which prevents metastasis in melanoma cells. Quantitative reverse transcriptase-polymerase chain reaction was used to study the expression of AXOR12 and KiSS-1 in a variety of tissues. The highest levels of expression of AXOR12 mRNA were observed in brain, pituitary gland, and placenta. The highest levels ofKiSS-1 gene expression were observed in placenta and brain. A polyclonal antibody raised to the C terminus of AXOR12 was generated and used to show localization of the receptor to neurons in the cerebellum, cerebral cortex, and brainstem. The biological significance of these expression patterns and the nature of the putative cognate ligand for AXOR12 are discussed. G protein-coupled receptor reverse transcriptase polymerase chain reaction Chinese hamster ovary neuropeptide FF high pressure liquid chromatography The G protein-coupled receptors (GPCRs)1 form a large family of membrane bound proteins that share a unique structural feature comprising seven transmembrane α-helices. These molecules act as receptors for a diverse range of extracellular signaling molecules including small molecules (amino acids and biogenic amines), lipids, small bioactive peptides, and large polypeptides (1Wilson S. Bergsma D.J. Chambers J.K. Muir A.I. Fantom K.G.M. Ellis C. Murdock P.R. Herrity N.C. Stadel J.M. Br. J. Pharmacol. 1998; 125: 1387-1392Crossref PubMed Scopus (114) Google Scholar). They have been used successfully as drug targets by the pharmaceutical industry for a number of years. Attention has focused on a number of proteins that are known to be GPCRs through structural homology but for which no ligand has been identified: so-called orphan receptors. At the same time as the recent discovery of new GPCRs, there has been a renewed focus on discovering potential novel peptides that may act as endogenous ligands for these receptors. Here, we describe the cloning of a novel human orphan receptor, a class I GPCR with sequence similarity to receptors for the neuropeptide galanin. This receptor was given the name AXOR12 in accordance with its position in a series of receptors identified in our organization. AXOR12 has a high degree of homology to the rat orphan receptor GPR54 (2Lee D.K. Nguyen T. O'Neill G.P. Cheng R. Liu Y. Howard A.D. Coulombe N. Tan C.P. Tang-Nguyen A.T. George S.R. O'Dowd B.F. FEBS Lett. 1999; 446: 103-107Crossref PubMed Scopus (380) Google Scholar) (81% amino acid identity), which suggests that these two receptors may be orthologs. To identify a ligand for AXOR12, we expressed this receptor in mammalian cells and screened the transfected cells in a functional assay against a library rich in known and putative peptide transmitters. Although there was no activity in response to galanin, we identified three peptides that acted as low potency agonists of AXOR12. These peptides were all derived from invertebrates and shared a C-terminal LRF- or LRW-amide motif. During the preparation of this article, a search of patent literature revealed the existence of additional high potency agonists with sequence similarities to the surrogate agonists identified from the screen. These peptides were derived from a precursor known as KiSS-1. The KiSS-1 gene was identified originally as being up-regulated in melanoma cells that have lost the potential to metastasize after microcell-mediated transfer of human chromosome 6 (3Lee J.-H. Miele M.E. Hicks D.J. Phillips K.K. Trent J.M. Weissman B.E. Welch D.R. J. Natl. Cancer Inst. 1996; 88: 1731-1737Crossref PubMed Scopus (809) Google Scholar). Subsequent studies have shown that the exogenous expression ofKiSS-1 in other tumor cell lines also prevents metastasis (4Lee J.-H. Welch D.R. Cancer Res. 1997; 57: 2384-2387PubMed Google Scholar). The mechanism by which this occurs remains largely unknown; however, KiSS-1 has structural features that suggest that it may be the precursor of a secreted peptide with an LRF-amide motif at the C terminus. We synthesized the putative processed products of KiSS-1 and observed that they acted as high potency agonists of AXOR12. To gain insight into the physiological role of this receptor, we used quantitative RT-PCR to localize the mRNA expression of AXOR12 and KiSS-1 in a range of human tissues. We observed high levels of AXOR12 expression in the brain. Further RT-PCR analysis of brain expression revealed a distinctive pattern of mRNA localization that was further explored by immunohistochemistry using an antibody raised to the extreme C-terminal tail of the receptor. The human AXOR12 gene was identified initially within a genomic clone (GenBankTMaccession number AC005379) as five coding exons interrupted by four introns. The full-length cDNA was obtained by a modification of a previously described cDNA capture method (5Shepard A.R. Rae J.L. Nucleic Acids Res. 1997; 25: 3183-3185Crossref PubMed Scopus (30) Google Scholar). Briefly, 5 µg of plasmid DNA from a human brain cDNA library was screened with a biotinylated oligonucleotide (5′-biotin with an 18-atom spacer) corresponding to the 5′-coding region (5′-GATGCGGACCGTGACCAACTTCTAC-3′). Two additional 40-mers (5′-GGAACTCGCTGGTCATCTACGTCATCTGCCGCCACAAGCC-3′ and 5′-ATCGCCAACCTGGCGGCCACGGACGTGACCTTCCTCCTGTG-3′), corresponding to the immediate 5′ and 3′ regions of the biotinylated probe, were also used as blocking oligos. Bacterial colonies from the second round of selection were screened by PCR using AXOR12-specific primers. Five positive clones were identified, and the entire inserts were sequenced on both strands using an ABI sequencer. Two of the sequenced clones were identical to each other and to the full-length AXOR12 cDNA predicted from the genomic DNA sequence. The AXOR12 cDNA was subcloned into the mammalian expression vector pCDN (6Trill J.J. Shatzman A.R. Ganguly S. Curr. Opin. Biotech. 1995; 6: 553-560Crossref PubMed Scopus (94) Google Scholar) as described previously (7Aiyar N. Disa J. Stadel J.M. Lysko P.G. Mol. Cell. Biochem. 1999; 197: 179-185Crossref PubMed Scopus (55) Google Scholar). Transient transfections of HEK293 cells with AXOR12 alone or AXOR12 co-transfected with Gqi5 (a chimeric G protein α-subunit consisting of Gαq with the C-terminal five amino acids substituted with the corresponding amino acids from Gαi2, which facilitates GPCR signaling through phospholipase C) were prepared for functional studies. A Ca2+ mobilization assay using the microtiter plate-based Ca2+ mobilization fluorometric imaging plate reader was performed as described previously (8Chambers J. Ames R.S. Bergsma D. Muir A. Fitzgerald L.R. Hervieu G. Dytko G.M. Foley J.J. Martin J. Liu W.S. Park J. Ellis C. Ganguly S. Konchar S. Cluderay J. Leslie R. Wilson S. Sarau H.M. Nature. 1999; 400: 261-265Crossref PubMed Scopus (460) Google Scholar). HEK293 cells co-transfected with AXOR12 and Gqi5 were screened against a large library of over 1500 known and putative GPCR agonists including all available mammalian neuropeptides as described previously (9Szekeres P.G. Muir A.I. Spinage L.D. Miller J.E. Butler S.I. Smith A. Rennie G.I. Murdock P.R. Fitzgerald L.R. Wu H. McMillan L.J. Guerrera S. Vawter L. Elshourbagy N.A. Mooney J.L. Bergsma D.J. Wilson S. Chambers J.K. J. Biol. Chem. 2000; 275: 20247-20250Abstract Full Text Full Text PDF PubMed Scopus (151) Google Scholar). Peptides in this library were tested at a final concentration of >100 nm and other potential small molecule agonists at >1 µm. To obtain mammalian cells stably expressing AXOR12, the cDNA was subcloned into the vector pCD (a derivative of pCDN lacking the gene for neomycin resistance) and co-transfected with pCDN Gqi5 (10Conklin B.R. Farfel Z. Lustig K.D. Julius D. Bourne H.R. Nature. 1993; 363: 274-276Crossref PubMed Scopus (599) Google Scholar) into Chinese hamster ovary cells using LipofectAMINE PLUS (Life Technologies, Inc.) at a DNA ratio of 10:1 (CHO/AXOR12:Gqi5 cells). 48 h later the cells were seeded into 96-well plates at 103cells/well and selected with G418 (Life Technologies, Inc.) (400 µg/ml) and in the absence of nucleosides. Doubly selected cells were screened by Northern blotting to confirm AXOR12 expression, and positive clones were screened functionally in the fluorometric imaging plate reader calcium assay. The clone that responded most sensitively to surrogate agonists was used in all future experiments. The peptides KiSS-1-(107–121), KiSS-1-(112–121), KiSS-1-(114–121), NPFF, neuropeptide AF, RF-amide-like peptide-1 and -3, and galanin-like peptide were synthesized by conventional solid-phase techniques using Fmoc (N-(9-fluorenyl)methoxycarbonyl) chemistry (11Chan W.C. White P.D. Fmoc Solid Phase Peptide Synthesis: A Practical Approach. Oxford University Press, Oxford2000Google Scholar), and purification was conducted by preparative reverse phase HPLC. All final products showed a purity of >95% by analytical reverse phase HPLC, and peptide identities were confirmed by electrospray mass spectrometry. The peptides KiSS-1-(58–65), KiSS-1-(68–91), KiSS-1-(68–80), KiSS-1-(68–121), KiSS-1-(96–121), and KiSS-1-(125–144) were prepared by California Peptide Research Inc., CA. Antho-RW-amides I and II, Peptide F1, and galanin were purchased from Bachem. RNA purification and TaqMan RT-PCR analysis of human tissue were performed as described previously (12Sarau H.M. Ames R.S. Chambers J. Ellis C. Elshourbagy N. Foley J.J. Schmidt D.B. Muccitelli R.M. Jenkins O. Murdock P.R. Herrity N.C. Halsey W. Sathe G. Muir A.I. Nuthulaganti P. Dytko G.M. Buckley P.T. Wilson S. Bergsma D.J. Hay D.W.P. Mol. Pharmacol. 1999; 56: 657-663Crossref PubMed Scopus (301) Google Scholar). TaqMan primer and probe sets for AXOR12, KiSS-1, and the housekeeping gene glyceraldehyde-3-phosphate dehydrogenase were designed using Primer Express V1.0 software (Applied Biosystems). Primer and probe sequences (forward primer, reverse primer, TaqMan probe) were; AXOR12, 5′-TGGCACCCACGCAGCTA-3′, 5′-AGTTGCTGTAGGACATGCAGTGA-3′, 5′-CCGCCTACGCGCTTAAGACCTGG-3′; KiSS-1, 5′-ACTCACTGGTTTCTTGGCAGCT-3′, 5′-CAGAGGCCACCTTTTCTAATGG-3′, 5′-CTTTTCCTCTGTGCCACCCACTTTGG-3′; and glyceraldehyde-3-phosphate dehydrogenase, 5′-CAAGGTCATCCATGACAACTTTG-3′, 5′-GGGCCATCCACAGTCTTCTG-3′, 5′-ACCACAGTCCATGCCATCACTGCCA-3′. The levels of mRNA measured were calculated as copies of mRNA detected per ng of reverse transcribed poly(A)+ RNA. A unique synthetic peptide (CVLGEDNAPL) located at the extreme C terminus of the human AXOR12 receptor sequence, corresponding to amino acids 389–398, was synthesized (Bio Synthesis Inc.). Polyclonal antibodies were produced as described in detail elsewhere (13Moore D. Chambers J. Waldvogel H. Faull R. Emson P. J. Comp. Neurol. 2000; 421: 374-384Crossref PubMed Scopus (124) Google Scholar). In brief, New Zealand White rabbits were injected with a peptide-purified protein derivative of tuberculin conjugate and boosted at regular intervals. Crude bleeds were tested for antibody titer using a standard enzyme-linked immunosorbent assay protocol. AXOR12 antiserum was purified from crude rabbit serum by immunoaffinity chromatography on peptide-coupled Sulfolink™ gel (Pierce). Western blotting was carried out essentially as described elsewhere (13Moore D. Chambers J. Waldvogel H. Faull R. Emson P. J. Comp. Neurol. 2000; 421: 374-384Crossref PubMed Scopus (124) Google Scholar). In brief, membranes were prepared from selected tissue regions of human brain (frontal cortex, hippocampus, and basal ganglia) and CHO AXOR12:Gqi5 cells or nontransfected CHO cells. Protein concentrations were determined using the BCA protein assay kit (Pierce) according to manufacturer instructions. For SDS-polyacrylamide gel electrophoresis, 10 µg of membrane protein was denatured in Laemmli sample buffer (14Laemmli U.K. Nature. 1970; 227: 680-685Crossref PubMed Scopus (205531) Google Scholar). After electrophoresis, the proteins were blotted onto 0.45-µm nitrocellulose membranes, blocked in 5% milk solution in Tris-buffered saline/0.1% Tween 20, and incubated with affinity-purified AXOR12 antiserum (1:5000). Immunoreactivity was detected and visualized by incubating the membrane with a horseradish peroxidase-conjugated goat anti-rabbit secondary antibody (1:5000, Promega, Madison, WI) followed by chemiluminescent detection (ECL, Amersham Pharmacia Biotech). Controls included pre-absorption of AXOR12 antibody with immunogenic peptide (10 µm) prior to incubation with the blot and omission of the primary antibody. Immunocytochemistry was performed on stably transfected CHO AXOR12:Gqi5 cells. The cells were fixed in 4% paraformaldehyde/PBS for 20 min at 4 °C, permeabilized for 5 min in 0.5% Triton X-100/PBS, and blocked for 30 min in 50% fetal calf serum. The fixed cells were incubated in a 1:250 dilution of affinity-purified AXOR12 antiserum overnight at 4 °C, washed three times in PBS, and incubated for 30 min in anti-rabbit secondary antibody conjugated to fluorescein isothiocyanate. Coverslips were mounted with vectashield containing 4′,6-diamidino-2-phenylindole (DAPI, Vector Laboratories). Immunohistochemistry was carried out on perfusion-fixed post-mortem human brain tissue obtained from the New Zealand Neurological Foundation Human Brain Bank (University of Auckland) with consent from the University of Auckland Human Subjects Ethics Committee. All tissues used in this study were from cases with no previous history of neurological disorders or abnormalities after thorough pathological examinations. The post-mortem delay from death until tissue fixation ranged from 10 to 18 h. Coronal 50-µm-thick brain sections were cut on a freezing sledge microtome and collected in 0.1 mPBS, pH 7.4. Endogenous peroxidase activity was quenched for 30 min. Sections were pre-incubated with normal goat serum (1%) in buffer A (0.1 m PBS/0.3% (v/v) Triton X-100) for 1 h followed by incubation with affinity-purified AXOR12 antiserum (1:2000) overnight at 4 °C. Sections were then incubated with biotinylated goat anti-rabbit secondary antibody (1:200, Vector Laboratories) for 2 h followed by incubation with ABC reagent (1:200, Vector Laboratories) for 45 min. Sections were visualized using 0.5 mg/ml 3′3-diaminobenzidine as a substrate and 0.03% H2O2. Stained tissue sections were mounted onto microscope slides, air-dried, and coverslipped with Depex (BDH Laboratory Supplies). Controls included pre-absorption of primary antibody with 50 µm peptide antigen overnight at 4 °C prior to incubation in addition to the omission of the primary antibody or the use of pre-immune serum. As part of an ongoing effort to identify ligands for novel human orphan GPCRs, we cloned a novel GPCR, originally identified within a genomic clone, using a cDNA capture method. As shown in Fig.1 this gene, AXOR12, encodes a 398-amino acid protein. TMHMM, a program that uses a hidden Markov model for predicting transmembrane helices (15Sonnhammer E.L.L. von Heijne G. Krogh A. Proceedings of the Sixth International Conference on Intelligent Systems for Molecular Biology. American Association for Artificial Intelligence Press, Menlo Park, CA1998: 175-182Google Scholar), predicted that the protein contained seven hydrophobic putative transmembrane domains. AXOR12 has sequence homology with class I members of the GPCR superfamily, exhibiting the highest sequence homology (81% amino acid identity) to GPR54, a rat receptor previously characterized as a galanin receptor-like orphan (2Lee D.K. Nguyen T. O'Neill G.P. Cheng R. Liu Y. Howard A.D. Coulombe N. Tan C.P. Tang-Nguyen A.T. George S.R. O'Dowd B.F. FEBS Lett. 1999; 446: 103-107Crossref PubMed Scopus (380) Google Scholar). The human GPCRs to which AXOR12 has closest homology are members of the galanin receptor family, sharing 28, 30, and 30% amino acid identity with human GalR1, GalR2, and GalR3, respectively. We adopted a “reverse pharmacological” strategy (16Stadel J.M. Wilson S. Bergsma D.J. Trends Pharmacol. Sci. 1997; 18: 430-437Abstract Full Text PDF PubMed Scopus (141) Google Scholar) to identify agonists for AXOR12. Thus, HEK293 cells co-transfected with AXOR12 and Gqi5 were challenged with a large library of more than 1500 putative ligands, and responses were measured in a microtiter plate-based (fluorometric imaging plate reader) calcium mobilization assay. We observed that specific responses were elicited by two neuropeptides originally isolated from the sea anemone, Anthopleura elegantissima (antho-RW-amide I (17Graff D. Grimmelikhuijzen C.J. Brain Res. 1988; 442: 354-358Crossref PubMed Scopus (29) Google Scholar) and antho-RW-amide II (18Graff D. Grimmelikhuijzen C.J. FEBS Lett. 1988; 239: 137-140Crossref PubMed Scopus (21) Google Scholar)), and a neuropeptide from the lobster, peptide F1 (19Trimmer B.A. Kobierski L.A. Kravitz E.A. J. Comp. Neurol. 1987; 266: 16-26Crossref PubMed Scopus (125) Google Scholar) (sequences shown in Table I). Further experiments demonstrated that these responses were not dependent upon co-transfection of the recombinant chimeric G protein and were concentration-dependent with EC50 values in the low micromolar range (Fig. 2). For further studies, CHO AXOR12:Gqi5 cells were used. The CHO parental cells did not respond to the peptides studied here, but the transfected cell line responded to the three surrogate agonists (Fig. 4 and TableI). Galanin and galanin-like peptide, when tested at concentrations up to 1 µm, did not activate CHO AXOR12:Gqi5 cells (data not shown). The common feature of all activating peptides is the presence of an amidated LRF or LRW motif at the C terminus. This suggests that the cognate ligand for this receptor is likely to have a similar structure at the C terminus.Table IpEC50 of peptides tested on CHO AXOR12:Gqi5 cells as determined from dose-dependent changes in intracellular Ca2+PeptideSequenceSpeciesMean pEC50S.E.M.Antho-RW-amide IIpESLRW-amideSea anemone5.730.28Antho-RW-amide IpEGLRW-amideSea anemone6.420.28Peptide F1TNRNFLRF-amideLobster5.630.08KiSS-1-(114–121)WNSFGLRF-amideHuman8.030.51KiSS-1-(112–121)YNWNSFGLRF-amideHuman9.300.39KiSS-1-(107–121)KDLPNYNWNSFGLRF-amideHuman8.920.37KiSS-1-(94–121)IPAPQGAVLVQREKDLPNYNWNSFGLRF-amideHuman8.760.43KiSS-1-(68–121)GTSLSPPPESSGSRQQPGLSAPHSRQ—Human8.000.16—IPAPQGAVLVQREKDLPNYNWNSFGLRF-amideKiSS-1-(68–91)GTSLSPPPESSGSRQQPGLSAPHSHuman<4N/AKiSS-1-(68–80)GTSLSPPPESSGSHuman<5N/AKiSS-1-(58–65)PAATLRSHuman<4N/AKiSS-1-(125–144)EAAPGNHGRSAGRGWGAGAGQHuman<4N/AThe peptide sequences and source species of the peptides are shown along with the mean pEC50 and S.E. derived from three experiments. N/A, not applicable. Open table in a new tab Figure 4CHO AXOR12:Gqi5 cells responded to nonmammalian peptides containing the LRF- or LRW-amide motif, but peptides deduced from KiSS-1, containing LRF-amide, were more potent activators as measured by mobilization of intracellular Ca2+. The data are expressed as the change in fluorescent intensity units (FIU) over background and are from a single experiment, representative of three such experiments. Each point was determined in triplicate and is given as a mean. ○, KiSS-1-(112–121); ■, KiSS-1-(94–121); ●, KiSS-1-(107–121); ▪, KiSS-1-(114–121); ▾, Antho-RW-amide I; ▵, Antho-RW-amide II; ▴, peptide F1.View Large Image Figure ViewerDownload (PPT) The peptide sequences and source species of the peptides are shown along with the mean pEC50 and S.E. derived from three experiments. N/A, not applicable. From a search of the patent literature, KiSS-1, a potential ligand for AXOR12, was identified (patent number WO200024890, Takeda Chemical Industries, Ltd). An analysis of the peptide sequence of KiSS-1 showed it to have features typical of secreted neuropeptides including a signal sequence, as predicted by signalP (20Nielsen H. Engelbrecht J. Brunak S. von Heijne G. Protein Eng. 1997; 10: 1-6Crossref PubMed Scopus (4911) Google Scholar), several potential dibasic cleavage sites, and a cleavage/amidation site (Fig.3). This would result in a putative 54-amino acid-secreted peptide product corresponding to residues 68–121 of the full-length KiSS-1. Most interestingly, this contains a C-terminal LRF-amide sequence, as predicted from our studies with nonmammalian peptides. A range of peptides of different lengths were synthesized, corresponding to the C-terminal end of the putative secreted segment of KiSS-1, and were tested in CHO AXOR12:Gqi5 cells using the fluorometric imaging plate reader assay as described above. A comparison of the peptides derived from the putative secreted segment of KiSS-1 (68–121, 94–121, 107–121, 112–121, and 114–121) (Table I) revealed that these peptides were substantially more potent than the nonmammalian LRF- or LRW-amide peptides tested (Fig.4, Table I). Of the four truncated KiSS-1 sequences, the decapeptide 112–121 possessed the highest potency. Further N-terminal deletion of this peptide resulted in nearly a 20-fold drop in functional potency. Chain elongation to the pentadecapeptide 107–121 or the longer fragment 96–121 afforded smaller reductions in potency (Fig. 4). Likewise, elongation to the 54-amino acid peptide corresponding to the entire putative secreted segment of KiSS-1, tested in separate experiments, resulted in a further reduction in potency (Table I). Peptides corresponding to the N terminus of the putative secreted segment and also to putatively nonsecreted segments of KiSS-1 were inactive (pEC50 < 5) (Table I). NPFF, neuropeptide AF, and the RF-amide like peptides active at the type I neuropeptide FF receptor, NPFFR1 (21Hinuma S. Shintani Y. Fukusumi S. Iijima N. Matsumoto Y. Hosoya M. Fujii R. Watanabe T. Kikuchi K. Terao Y. Yano T. Yamamoto T. Kawamata Y. Habata Y. Asada M. Kitada C. Kurokawa T. Onda H. Nishimura O. Tanaka M. Ibata Y. Fujino M. Nat. Cell Biol. 2000; 2: 703-708Crossref PubMed Scopus (498) Google Scholar), were also tested, and all failed to generate a response (pEC50 < 5, data not shown). The calcium mobilization response seen after the activation of AXOR12 when transiently transfected without additional G protein α-subunits into HEK293 cells (Fig. 2) suggests that this receptor is coupled to G proteins of the Gq/11 subfamily. In agreement with this hypothesis, KiSS-1-(112–121) caused identical calcium mobilization in both control and pertussis toxin-treated HEK293 cells transiently expressing AXOR12 (data not shown), suggesting that activation of G proteins from the Gi/o family and subsequent Gβγ-mediated activation of phospholipase Cβ does not contribute to the functional response observed. In addition, neither basal nor forskolin-elevated levels of intracellular cAMP were modulated by KiSS-1-(112–121) in HEK293 cells transiently expressing AXOR12 (data not shown), suggesting that this receptor does not couple strongly to G proteins of the Gs and/or Gi/o subfamilies. To characterize the expression pattern of both the receptor AXOR12 and its putative ligand KiSS-1 we carried out quantitative RT-PCR analysis on a broad range of human tissues. AXOR12 was widely expressed in human tissues including the placenta, brain, and pituitary at high levels (Fig. 5), with lower levels detected in lymphocytes, pancreas, and adipose tissue. Within the human central nervous system, AXOR12 mRNA was widespread in its expression including the amygdala, nucleus accumbens, hippocampus, and cingulate gyrus (Fig. 6). KiSS-1 mRNA was detected predominantly in the placenta (Fig. 5) but was also widespread throughout the central nervous system, although at much lower levels (Fig. 6).Figure 6Localization by quantitative RT-PCR of AXOR12 and KiSS-1 in the human central nervous system together with expression of the housekeeping gene glyceraldehyde-3-phosphate dehydrogenase. Bars indicate mean number of copies detected per ng of reverse transcribed poly(A)+ RNA from two male and two female individuals. Error barsindicate standard error.View Large Image Figure ViewerDownload (PPT) To gain a greater understanding of the specific cellular localization and level of protein expression of AXOR12, we generated an antiserum directed against a C-terminal peptide from AXOR12. Western blot analysis of membranes from CHO AXOR12:Gqi5 cells revealed a broad immunoreactive band of ∼75 kDa, which was absent in membranes from untransfected CHO cells (Fig.7 A). Similarly, analysis of human brain membrane proteins revealed a specific band of ∼75 kDa (as well as a larger band at ∼125 kDa) in hippocampus, basal ganglia, and frontal cortex (Fig. 7 B). These immunoreactive bands were competed out by the immunizing peptide (data not shown). In both cases, the 75-kDa protein detected agrees with the predicted size of the native AXOR12 receptor together with a probable degree of glycosylation at extracellular consensus sites (see Fig. 1). The AXOR12 receptor-specific antiserum was used to localize receptor immunoreactivity in CHO AXOR12:Gqi5 cells. Expression of the receptor was detected both at the membrane and within the cells (Fig.7 C), although only a subset of the cells was immunoreactive for AXOR12 (Fig. 7, C and D). Immunohistochemical analysis of human brain sections revealed prominent neuronal expression in the regions sampled including the cerebral cortex, thalamus, pons-medulla, and cerebellum. In the cerebral cortex, AXOR12-specific staining was present on a large number of pyramidal neurones of layers III, V, and VI (Fig.8 A). However, in the basal ganglia (caudate nucleus, putamen, globus pallidus, and substantia nigra), staining was primarily localized to AXOR12-immunoreactive fibers and processes. In the cerebellum AXOR12 was strikingly localized to the surface of the Purkinje cells and their apical dendrites (Fig.8 B) and to a lesser extent the cells of the deep cerebellar nuclei. In the pons medulla AXOR12 immunoreactivity was widespread in a number of nuclei including the raphe nuclei, inferior olive, and hypoglossal nuclei (Fig. 8 C). AXOR12 staining was abolished when the antibody was pre-absorbed with 10 µm immunogenic peptide (Fig. 8, D and E) or when the primary antibody was omitted (data not shown). In this article we describe the cloning of a potential human ortholog of the rat G protein-coupled receptor GPR54, which we term AXOR12. A number of peptides possessing an LRW-amide or an LRF-amide motif at the C terminus are identified as surrogate ligands of AXOR12. The first RF-amide peptide to be described was FMRFamide, a cardioexcitatory neuropeptide isolated from the bivalve molluskMacrocallista nimbosa (22Price D. Greenberg M. Science. 1977; 197: 670-671Crossref PubMed Scopus (917) Google Scholar). FMRFamide and related RF-amides are widespread among invertebrates including Caenorhabditis elegans, which has at least 22 such peptides. These are expressed in the central and peripheral nervous systems and have a diverse range of functions including control of defecation, feeding, and reproduction (23Li C. Kim K. Nelson L.S. Brain Res. 1999; 848: 26-34Crossref PubMed Scopus (180) Google Scholar). More recently, several RF-amides have been discovered in mammalian species: neuropeptides FF and AF (NPFF and neuropeptide AF (NPAF)) (24Yang H.Y. Fratta W. Majane E.A. Costa E. Proc. Natl. Acad. Sci. U. S. A. 1985; 82: 7757-7761Crossref PubMed Scopus (559) Google Scholar, 25Perry S.J. Huang E.Y.-K. Cronk D. Bagust J. Sharma R. Walker R.J. Wilson S. Burke J.F. FEBS Lett. 1997; 409: 426-430Crossref PubMed Scopus (197) Google Scholar), prolactin releasing peptide (26Hinuma S. Habata Y. Fujii R. Kawamata Y. Hosoya M. Fukusumi S. Kitada C. Masuo Y. Asano T. Matsumoto H. Sekiguchi M. Kurokawa T. Nishimura O. Onda H. Fujino M. Nature. 1998; 393: 272-276Crossref PubMed Scopus (526) Google Scholar), and RF-amide related peptides-1 and -3 (21Hinuma S. Shintani Y. Fukusumi S. Iijima N. Matsumoto Y. Hosoya M. Fujii R. Watanabe T. Kikuchi K. Terao Y. Yano T. Yamamoto T. Kawamata Y. Habata Y. Asada M. Kitada C. Kurokawa T. Onda H. Nishimura O. Tanaka M. Ibata Y. Fujino M. Nat. Cell Biol. 2000; 2: 703-708Crossref PubMed Scopus (498) Google Scholar). Some orphan receptors have been shown to be activated by naturally occurring peptides from lower organisms that have structural similarity to the cognate neuropeptide ligand for these receptors (27Elshourbagy N.A. Ames R.S. Fitzgerald L.R. Foley J.J. Chambers J.K. Szekeres P.G. Evans N.A. Schmidt D.B. Buckley P.T. Dytko G.M. Murdock P.R. Milligan G. Groarke D.A. Tan K.B. Shabon U. Nuthulaganti P. Wang D.Y. Wilson S. Bergsma D.J. Sarau H.M. J. Biol. Chem. 2000; 275: 25965-25971Abstract Full Text Full Text PDF PubMed Scopus (236) Google Scholar). For this reason we screened AXOR12 against a library with a wide diversity of mammalian and nonmammalian peptides. On the basis of our initial screen we reasoned that the natural ligand for AXOR12 was likely to contain a LRF- or LRW-amide motif at its C terminus. KiSS-1, a putative human neuropeptide with an LRF-amide motif identified from the patent literature, was therefore a candidate cognate ligand for this receptor. Structurally, KiSS-1 bears all of the hallmarks of a secreted neuropeptide, with a putative signal sequence and several sites amenable to cleavage including an amidation/cleavage site that would result in a number of amidated peptide fragments of various lengths. We synthesized a number of putative N-terminally truncated products of this peptide and tested them against AXOR12. The results showed that the C-terminal decapeptide from the putative full-length secreted segment of KiSS-1 possessed sub-nanomolar activity at AXOR12. A reduction in chain length to the octapeptide 114–121 resulted in a significant drop in functional activity, and it is postulated that Tyr112 and Asn113 may each play an important role in receptor interaction and activation. Chain elongation resulted in minor decreases in activity, suggesting that the most relevant pharmacophore resides in the C-terminal fragment 112–121. AXOR12 failed to be activated by other RF-amide peptides, NPFF, neuropeptide AF, and RF-amide-like peptide-1 and -3, indicating the selectivity of activity of the KiSS-1 peptides. Expression of KiSS-1 in normal tissues was detected initially only in placenta (3Lee J.-H. Miele M.E. Hicks D.J. Phillips K.K. Trent J.M. Weissman B.E. Welch D.R. J. Natl. Cancer Inst. 1996; 88: 1731-1737Crossref PubMed Scopus (809) Google Scholar). Our own observations have shown that although KiSS-1 mRNA is extremely abundant in the placenta, low levels of message are also found in the brain and more specifically in the hypothalamus and basal ganglia. This distribution pattern is consistent with the mRNA localization of AXOR12, which is also in placenta, several central nervous system regions, and pituitary. Our data on AXOR12 localization broadly correspond to the published data for the putative rat ortholog GPR54 (2Lee D.K. Nguyen T. O'Neill G.P. Cheng R. Liu Y. Howard A.D. Coulombe N. Tan C.P. Tang-Nguyen A.T. George S.R. O'Dowd B.F. FEBS Lett. 1999; 446: 103-107Crossref PubMed Scopus (380) Google Scholar). Immunohistochemical data indicate that expression of the receptor occurs specifically on a number of neuronal cell types in the human central nervous system regions that were examined. Indeed, neuronal localization of AXOR12 in many regions of the cerebral cortex, cerebellum, and medulla fits well with the observed mRNA expression pattern in similar regions of the human central nervous system. The prominent and widespread expression of AXOR12 throughout the central nervous system, especially on a number of projection neurones including pyramidal cells in the cerebral cortex and cerebellar Purkinje cells, indicates that ligands acting on AXOR12 would be able to influence a wide range of central nervous system functions ranging from cognition through to movement and balance. The mapping of AXOR12 to human chromosome 19p13.3 (2Lee D.K. Nguyen T. O'Neill G.P. Cheng R. Liu Y. Howard A.D. Coulombe N. Tan C.P. Tang-Nguyen A.T. George S.R. O'Dowd B.F. FEBS Lett. 1999; 446: 103-107Crossref PubMed Scopus (380) Google Scholar) corresponds with a number of inherited neurological diseases including familial febrile convulsions (28Johnson E.W. Dubovsky J. Rich S.S. O'Donovan C.A. Orr H.T. Anderson V.E. Gil-Nagel A. Ahmann P. Dokken C.G. Schneider D.T. Weber J.L. Hum. Mol. Genet. 1998; 7: 63-67Crossref PubMed Scopus (157) Google Scholar), vacuolar neuromyopathy (29Servidei S. Capon F. Spinazzola A. Mirabella M. Semprini S. de Rosa G. Gennarelli M. Sangiuolo F. Ricci E. Mohrenweiser H.W. Dallapiccola B. Tonali P. Novelli G. Neurology. 1999; 53: 830-837Crossref PubMed Google Scholar), and cayman-type cerebellar ataxia (30Nystuen A. Benke P.J. Merren J. Stone E.M. Sheffield V.C. Hum. Mol. Genet. 1996; 5: 525-531Crossref PubMed Scopus (80) Google Scholar). Furthermore, the syntenic region on mouse chromosome 10 is associated with the allelic mutationsjittery and hesitant, which have neuropathic phenotypes characterized by ataxia, dystonia, and seizures (31Kapfhamer D. Sweet H.O. Sufalko D. Warren S. Johnson K.R. Burmeister M. Genomics. 1996; 35: 533-538Crossref PubMed Scopus (22) Google Scholar). The KiSS-1 peptide was identified originally as being differentially up-regulated in C8161 melanoma cells that have been rendered nonmetastatic by microcell-mediated transfer of human chromosome 6 (3Lee J.-H. Miele M.E. Hicks D.J. Phillips K.K. Trent J.M. Weissman B.E. Welch D.R. J. Natl. Cancer Inst. 1996; 88: 1731-1737Crossref PubMed Scopus (809) Google Scholar). Transfection of KiSS-1 into human breast carcinoma cells also prevents these cells from metastasizing (4Lee J.-H. Welch D.R. Cancer Res. 1997; 57: 2384-2387PubMed Google Scholar). The role of AXOR12 in these systems has not been explored, although expressed sequence tags corresponding to AXOR12 have been identified in a number of tumor cDNA libraries (GenBankTM accession numbers AI823800, AI819198, andAA887801). Interestingly, several neuropeptides are known to have functional roles in tumor biology including galanin, vasopressin, cholecystokinin, and neurotensin by autocrine and paracrine mechanisms (32Seufferlein T. Rozengurt E. Cancer Res. 1996; 56: 5758-5764PubMed Google Scholar, 33Ormandy C.J. Lee C.S.L. Ormandy H.F. Fantl V. Shine J. Peters G. Sutherland R.L. Cancer Res. 1998; 58: 1353-1357PubMed Google Scholar). The high levels of expression of both KiSS-1 andAXOR12 in placenta are also noteworthy in this respect. The placenta is also an invasive tissue, and there are similarities in the behavior of invading cancer cells and that of invading placenta cells (34Murray M.J. Lessey B.A. Semin. Reprod. Endocrinol. 1999; 17: 275-290Crossref PubMed Scopus (219) Google Scholar). It is possible that KiSS-1 and AXOR12 may form part of a mechanism that is common to both of these processes. In conclusion, AXOR12 constitutes a new human G protein-coupled receptor that has now been paired with its neuropeptide ligand, KiSS-1. Although there is still much to be discovered about both the receptor and its ligand, the biological evidence thus far suggests that they may have important physiological roles both in the central nervous system and in tumor biology. As such, they represent an intriguing target for novel therapies in the fields of both neurology and oncology.