Contribution of Melanocortin Receptor Exoloops to Agouti-related Protein Binding
1999; Elsevier BV; Volume: 274; Issue: 20 Linguagem: Inglês
10.1074/jbc.274.20.14100
ISSN1083-351X
AutoresYing-kui Yang, Chris J. Dickinson, Qun Zeng, Ji Yao Li, Darren A. Thompson, Ira Gantz,
Tópico(s)melanin and skin pigmentation
ResumoAgouti-related protein (AGRP) is an endogenous antagonist of melanocortin action that functions in the hypothalamic control of feeding behavior. Although previous studies have shown that AGRP binds three of the five known subtypes of melanocortin receptor, the receptor domains participating in binding and the molecular interactions involved are presently unknown. The present studies were designed to examine the contribution of extracytoplasmic domains of the melanocortin-4 receptor (MC4R) to AGRP binding by making chimerical receptor constructs of the human melanocortin-1 receptor (MC1R; a receptor that is not inhibited by AGRP) and the human MC4R (a receptor that is potently inhibited by AGRP). Substitutions of the extracytoplasmic NH2 terminus and the first extracytoplasmic loop (exoloop) of the MC4R with homologous domains of the MC1R had no effect on AGRP (87–132) binding affinity or inhibitory activity (the ability to inhibit melanocortin-stimulated cAMP generation). In contrast, cassette substitutions of exoloops 2 and 3 of the MC4R with the homologous exoloops of the MC1R resulted in a substantial loss of AGRP binding affinity and inhibitory activity. Conversely, the exchange of exoloops 2 and 3 of the MC1R with the homologous exoloops of the MC4R was found to confer AGRP binding and inhibitory activity to the basic structure of the MC1R. Importantly, these substitutions did not affect the ability of the α-melanocyte stimulating hormone analogue [Nle4,d-Phe7] melanocyte stimulating hormone to bind or activate the chimeric receptors. These data indicate that exoloops 2 and 3 of the melanocortin receptors are important for AGRP binding. Agouti-related protein (AGRP) is an endogenous antagonist of melanocortin action that functions in the hypothalamic control of feeding behavior. Although previous studies have shown that AGRP binds three of the five known subtypes of melanocortin receptor, the receptor domains participating in binding and the molecular interactions involved are presently unknown. The present studies were designed to examine the contribution of extracytoplasmic domains of the melanocortin-4 receptor (MC4R) to AGRP binding by making chimerical receptor constructs of the human melanocortin-1 receptor (MC1R; a receptor that is not inhibited by AGRP) and the human MC4R (a receptor that is potently inhibited by AGRP). Substitutions of the extracytoplasmic NH2 terminus and the first extracytoplasmic loop (exoloop) of the MC4R with homologous domains of the MC1R had no effect on AGRP (87–132) binding affinity or inhibitory activity (the ability to inhibit melanocortin-stimulated cAMP generation). In contrast, cassette substitutions of exoloops 2 and 3 of the MC4R with the homologous exoloops of the MC1R resulted in a substantial loss of AGRP binding affinity and inhibitory activity. Conversely, the exchange of exoloops 2 and 3 of the MC1R with the homologous exoloops of the MC4R was found to confer AGRP binding and inhibitory activity to the basic structure of the MC1R. Importantly, these substitutions did not affect the ability of the α-melanocyte stimulating hormone analogue [Nle4,d-Phe7] melanocyte stimulating hormone to bind or activate the chimeric receptors. These data indicate that exoloops 2 and 3 of the melanocortin receptors are important for AGRP binding. The melanocortin peptides, α-, β-, and γ-melanocyte stimulating hormone (MSH) 1The abbreviations used are: MSH, melanocyte stimulating hormone; AGRP, Agouti-related protein; MC4R, melanocortin-4 receptor; MC1R, melanocortin-1 receptor; MCR, melanocortin receptor; PCR, polymerase chain reaction; NDP-MSH, [Nle4,d-Phe7] MSH 1The abbreviations used are: MSH, melanocyte stimulating hormone; AGRP, Agouti-related protein; MC4R, melanocortin-4 receptor; MC1R, melanocortin-1 receptor; MCR, melanocortin receptor; PCR, polymerase chain reaction; NDP-MSH, [Nle4,d-Phe7] MSHand adrenocorticotropic hormone, are a group of peptides derived from the pro-opiomelanocortin prohormone that share a common message sequence (His-d-Phe-Arg-Trp). These peptides have been implicated in diverse physiological processes but are most widely recognized for their roles in melanogenesis, steroidogenesis, and, more recently, feeding behavior (1Tatro J.B. Neuroimmunomodulation. 1996; 3: 259-284Crossref PubMed Scopus (127) Google Scholar, 2Eberle A.N. The Melanotropins: Chemistry, Physiology and Mechanisms of Action. Karger, Basel, Switzerland1988: 210-319Google Scholar, 3Huszar D. Lynch C.A. Fairchild-Huntress V. Dunmore J.H. Fang Q. Berkemeier L.R. Gu W. Kesterson R.A. Boston B.A. Cone R.D. Smith F.J. Campfield L.A. Burn P. Lee F. Cell. 1997; 88: 131-141Abstract Full Text Full Text PDF PubMed Scopus (2538) Google Scholar, 4Fan W. Boston B.A. Kesterson R.A. Hruby V.J. Cone R.D. Nature. 1997; 385: 165-168Crossref PubMed Scopus (1660) Google Scholar). There are five known seven transmembrane G-protein-coupled melanocortin receptors that couple to the stimulatory G-protein, Gs (Refs. 5Mountjoy K.G. Robbins L.S. Mortrud M.T. Cone R.D. Science. 1992; 257: 1248-1251Crossref PubMed Scopus (1451) Google Scholar, 6Chhajlani V. Wikberg J.E.S. FEBS Lett. 1992; 309: 417-420Crossref PubMed Scopus (577) Google Scholar, 7Gantz I. Konda Y. Tashiro T. Shimoto Y. Miwa H. Munzert G. Watson S.J. DelValle J. Yamada T. J. Biol. Chem. 1993; 268: 8246-8250Abstract Full Text PDF PubMed Google Scholar, 8Gantz I. Miwa H. Konda Y. Shimoto Y. Tashiro T. Watson S.J. DelValle J. Yamada T. J. Biol. Chem. 1993; 268: 15174-15179Abstract Full Text PDF PubMed Google Scholar, 9Gantz I. Shimoto Y. Konda Y. Miwa H. Dickinson C.J. Yamada T. Biochem. Biophys. Res. Commun. 1994; 200: 1214-1220Crossref PubMed Scopus (288) Google Scholar). In addition, two endogenous melanocortin antagonists, Agouti and AGRP, have been discovered that function as modifiers of melanocortin action. The most well-documented function of Agouti is its action as a paracrine signaling molecule that modifies melanocortin action at the MC1R (the melanocyte MCR) (10Bultman S.J. Michaud E.J. Woychik R.P. Cell. 1992; 71: 1195-1204Abstract Full Text PDF PubMed Scopus (694) Google Scholar, 11Lu D. Willard D. Patel I.R. Kadwell S. Overton L. Kost T. Luther M. Chen W. Woychik R.P. Wilkison W.O. Cone R.D. Nature. 1994; 371: 799-802Crossref PubMed Scopus (932) Google Scholar). Agouti is temporally produced by cells adjacent to the hair follicle melanocyte and is responsible for the Agouti phenotype (dark hair with a sub-apical yellow band).In vitro studies have demonstrated that recombinant Agouti is a potent antagonist of melanocortin action at MCR subtypes 1, 2, and 4 (11Lu D. Willard D. Patel I.R. Kadwell S. Overton L. Kost T. Luther M. Chen W. Woychik R.P. Wilkison W.O. Cone R.D. Nature. 1994; 371: 799-802Crossref PubMed Scopus (932) Google Scholar, 12Blanchard S.G. Harris C.O. Ittoop O.R.R. Nichols J.S. Parks D.J. Truesdale A.T. Wilkison W.O. Biochemistry. 1995; 34: 10406-10411Crossref PubMed Scopus (62) Google Scholar, 13Yang Y.-K. Ollmann M.M. Barsh G.S. Yamada T. Gantz I. Mol. Endocrinol. 1997; 11: 274-280Crossref PubMed Scopus (121) Google Scholar). AGRP was originally identified from its sequence similarity to Agouti (14Shutter J.R. Graham M. Kinsey A.C. Scully S. Lthy R. Stark K.L. Genes Dev. 1997; 11: 593-602Crossref PubMed Scopus (554) Google Scholar, 15Fong T.M. Mao C. MacNeil T. Kalyani R. Smith T. Weinberg D. Tota M.R. Van der Ploeg L.H.T. Biochem. Biophys. Res. Commun. 1997; 237: 629-631Crossref PubMed Scopus (198) Google Scholar, 16Ollmann M.M. Wilson B.D. Yang Y.-K. Kerns J.A. Chen Y. Gantz I. Barsh G.S. Science. 1997; 278: 135-138Crossref PubMed Scopus (1538) Google Scholar). Both AGRP and Agouti have a COOH-terminal cysteine-rich motif. In contrast to Agouti, AGRP is a potent antagonist of the melanocortin-3 receptor and the MC4R and has also been shown to have a lesser degree of inhibitory action at the melanocortin-5 receptor. Although a full understanding of the biological spectrum of AGRP action remains to be determined, its expression in the hypothalamus, its ability to inhibit the action of α-MSH at the MC4R, and its ability as a transgene (under the control of a β-actin promoter) to cause obesity in mice indicate that it is involved in the regulation of feeding behavior. In the hypothalamus, α-MSH is believed to act a satiety-inducing factor that mediates its action through the MC4R, whereas AGRP is one of several opposing orexigenic agents. AGRP is transcribed as 132 amino acids in man (131 amino acids in mouse), and although it is not presently known whether it is post-translationally processed in mammals, we have recently shown that a 46-amino acid COOH-terminal AGRP variant has the same ability to selectively bind MCR subtypes and functionally inhibit melanocortins as the full-length molecule (17Yang Y.-K. Thompson D. Dickinson C.J. Wilken J. Barsh G.S. Kent S.B.H. Gantz I. Mol. Endocrinol. 1999; 13: 148-155Crossref PubMed Scopus (162) Google Scholar). Pharmacological studies using this chemically synthesized truncated AGRP variant, AGRP (87–132), indicate that AGRP is a competitive antagonist of α-MSH at MCR subtypes 3, 4, and 5. AGRP has very little amino acid sequence similarity to melanocortins. Even in its artificially truncated form, AGRP is 46 amino acids and contains 10 cysteine residues capable of forming five disulfide bonds. In contrast, the predominant melanocortin in the hypothalamus, α-MSH, is only 13 amino acids in length and has no cysteine residues. These observations suggest that the members of this agonist-antagonist pair have significantly different tertiary structures, despite their use of the same receptors. It is also important to note that with respect to their competitive interaction, α-MSH is capable of activating MCR subtypes 1, 3, 4, and 5, whereas AGRP can only inhibit the action of α-MSH action at MCR subtypes 3, 4, and 5. In view of their apparent structural dissimilarity and subtype specificity, it is likely that α-MSH and AGRP have receptor binding determinants that are not identical. To date, little is known about the molecular forces and receptor domains involved in AGRP-MCR binding. Based on assumptions about its structure and presently held concepts about the way larger peptides bind seven transmembrane G-protein coupled receptors, we hypothesized that one potential binding determinant for AGRP might be the extracytoplasmic domains of the MCRs. The amino acid sequences of the MC1R and MC4R were examined by a hydrophobicity plot (Genetics Computer Group, Inc., Madison, WI) and examined manually by comparing their sequences to a previously published alignment of seven transmembrane G-protein-coupled receptor α-helices (18Baldwin J.M. EMBO J. 1993; 12: 1693-1703Crossref PubMed Scopus (883) Google Scholar). The chimeras utilized in these studies are schematically diagrammed in Fig.1. The chimeras were constructed by polymerase chain reaction (PCR) using Pfu polymerase (Stratagene, La Jolla, CA). The human MC1R and MC4R served as a template. During an initial round of PCR, partial-length receptor fragments were generated. The sequence of one of the PCR primer oligonucleotides consisted of an extracytoplasmic domain of interest coupled to a portion of the transmembrane domain required to form a chimeric receptor. The second oligonucleotide primer consisted of either the 5′ or 3′ end of the MC1R or MC4R. Receptor fragments were separated by agarose gel electrophoresis and used in a second round of PCR in which full-length chimeric receptor constructs were assembled by cycling the appropriate fragments together for 10 cycles before adding both 5′ and 3′ receptor primers. The chimeric receptors were subcloned into the M13 vector for single-stranded dideoxynucleotide sequencing to check that the desired sequences were present and that no sequence errors had been introduced by PCR. The constructs were then subcloned into the eukaryotic expression vector pcDNA 3.1 (Invitrogen, Carlsbad, CA). A comparison of the exchanged amino acid sequences of the NH2-terminal, first, second, and third exoloops of the MC1R and MC4R is shown in Fig. 2. The sequence of the wild-type MC4R used in these studies can be found in GenBankTM under accession number L08603 (8Gantz I. Miwa H. Konda Y. Shimoto Y. Tashiro T. Watson S.J. DelValle J. Yamada T. J. Biol. Chem. 1993; 268: 15174-15179Abstract Full Text PDF PubMed Google Scholar). The sequence of the wild-type MC1R used in these studies can be found in GenBankTM under accession number X65634 (5Mountjoy K.G. Robbins L.S. Mortrud M.T. Cone R.D. Science. 1992; 257: 1248-1251Crossref PubMed Scopus (1451) Google Scholar), except that position 163 is Arg, and position 164 is Gln. This genetic polymorphism in transmembrane 4 does not cause any apparent change in either basal or stimulated cAMP or 125I-NDP-MSH binding compared with the published MC1R sequence.Figure 2Displacement of the radioligand 125 I-NDP-MSH from MC4R chimeras by (A) NDP-MSH and (B) AGRP (87–132). C depicts the displacement of125I-AGRP (87–132) by AGRP (87–132) from the chimeric MC4R. This set of chimeric receptors consists of the basic structure of the MC4R with substitutions of various extracellular portions of the MC1R. MC4RWT, wild-type MC4R; MC1RWT, wild-type MC1R; MC4R/NH2MC1R, MC4R containing the NH2terminus of the MC1R; MC4R/1eMC1R, MC4R containing the first exoloop of the human MC1R; MC4R/2eMC1R, MC4R containing the second exoloop of the MC1R; MC4R/3eMC1R, MC4R containing the third exoloop of the MC1R; MC4R/2e,3eMC1R, MC4R containing second and third exoloops of the MC1R.View Large Image Figure ViewerDownload Hi-res image Download (PPT) cAMP assays were performed on transiently transfected HEK-293 cells as described previously (13Yang Y.-K. Ollmann M.M. Barsh G.S. Yamada T. Gantz I. Mol. Endocrinol. 1997; 11: 274-280Crossref PubMed Scopus (121) Google Scholar) using a competitive binding assay (Amersham Pharmacia Biotech cAMP assay kit TRK 432). NDP-MSH was obtained from Peninsula Laboratories (Belmont, CA), and human AGRP (87–132) was provided by Gryphon Sciences (South San Francisco, CA). Data was analyzed using Graphpad Prism (Graphpad Software, San Diego, CA). All experiments represent n≥ 3 ± S.E. [125I](Iodotyrosyl2)-NDP-MSH and125I-AGRP (87–132) were prepared by simple oxidative methods using chloramine T and Na125I (Amersham Pharmacia Biotech) as described previously (13Yang Y.-K. Ollmann M.M. Barsh G.S. Yamada T. Gantz I. Mol. Endocrinol. 1997; 11: 274-280Crossref PubMed Scopus (121) Google Scholar). NDP-MSH was purchased from Peninsula Laboratories, and AGRP (87–132) was provided by Gryphon Sciences. Binding experiments were performed on transiently transfected HEK-293 cells using conditions described previously (13Yang Y.-K. Ollmann M.M. Barsh G.S. Yamada T. Gantz I. Mol. Endocrinol. 1997; 11: 274-280Crossref PubMed Scopus (121) Google Scholar), with some modification. Briefly, 12 h before the experiments, 0.2 million cells were plated on 24-well plates. Before initiating the binding experiments, cells were washed twice with minimum Eagle's medium. Cells were then incubated with different concentrations of unlabeled ligand containing 0.2% bovine serum albumin and either 1 × 105 cpm of 125I-NDP-MSH or 1 × 105 cpm of 125I-AGRP. After a 1-h incubation, the cells were again washed twice with minimum Eagle's medium, and the experiment was terminated by lysing the cells with 0.1 nNaOH and 1% Triton X-100. Radioactivity present in the lysate was quantified using an analytical gamma counter. Nonspecific binding was determined by measuring the amount of 125I label remaining bound in the presence of 10−5m unlabeled ligand, and specific binding was obtained by subtracting the nonspecific bound radioactivity from the total bound radioactivity. The binding displacement curves were drawn using Graphpad Prism. The agonist-antagonist complex dissociation constant was calculated using the equation K b = [AGRP]/(EC50ratio − 1). All experiments represent n ≥ 3 ± S.E. Fig. 2 A demonstrates that cassette substitutions of the NH2 terminus, first, second, and third exoloop of the MC4R (alone or in combination) with homologous regions of the MC1R did not alter 125I-NDP-MSH binding. Data in Figs. 2 and 4 are expressed as total counts per minute (cpm) to emphasize the point that the various HEK-293 cell lines expressed roughly the same numbers of receptors. Fig. 2 B summarizes the ability of AGRP (87–132) to displace125I-NDP-MSH from the MC4R/MC1R chimeras. IC50values are listed in Table I. As shown, AGRP (87–132) had only a minute ability to displace125I-NDP-MSH binding from the wild-type MC1R. In contrast, AGRP (87–132) dose-dependently displaced all125I-NDP-MSH from the MC4R. These data are consistent with the known MCR subtype specificity of AGRP. Substitution of the NH2 terminus or the first exoloop of the MC4R with homologous loops of the MC1R had no effect on the ability of AGRP (87–132) to displace 125I-NDP-MSH. However, substitution of the second or third exoloop of the MC4R with those of the MC1R resulted in a significantly reduced IC50 of AGRP (87–132) in inhibiting 125I-NDP-MSH binding to chimeric receptor MC4R/2eMC1R or MC4R/3eMC1R. Simultaneous substitution of the second and third exoloops of the MC4R with those of the MC1R (chimera MC4R/2e, 3eMC1R) led to further inhibition of AGRP binding.Table IEffect of NDP-MSH and AGRP on 125I NDP-MSH binding on HEK cells transfected with chimeras of human MC4RChimerasBinding IC50(nm)NDP-MSHAGRPMC4RWT3.48 ± 0.269.1 ± 1.2MC4R/NH2MC1R2.78 ± 0.149 ± 0.5MC4R/1eMC1R3.25 ± 0.149.4 ± 2.8MC4R/2eMC1R4.1 ± 0.2263 ± 5.9MC4R/3eMC1R3.1 ± 0.16126 ± 5.2MC4R/2e,3eMC1R3.27 ± 0.65562 ± 26MC1RWT0.42 ± 0.02>1000 Open table in a new tab To further examine AGRP binding to the chimerical receptors, binding studies were performed with 125I-AGRP (87–132) (Fig.2 C). As expected, the wild-type MC1R demonstrated little if any specific 125I-AGRP (87–132) binding. In contrast, significant 125I-AGRP (87–132) binding was observed at the MC4R. Notably, there was a progressive loss in specific125I-AGRP (87–132) binding at chimeric MC4R with substitutions of the second and third exoloop with homologous exoloops of the MC1R (chimeras MC4R/2eMC1R, MC4R/3eMC1R, and MC4R/2e,3eMC1R). Substitution of the third exoloop had a greater effect than substitution of the second loop, and simultaneous substitution of exoloops 2 and 3 had an additive effect. These data are entirely consistent with the ability of AGRP to displace125I-NDP-MSH binding presented in Fig. 2 B. The chimerical MC4R containing both exoloops 2 and 3 of the MC1R (MC4R/2e, 3eMC1R) retained only approximately 38% of the specific125I-AGRP (87–132) binding of the wild-type MC4R. To study the functional effects of exoloop substitution in more detail, the ability of AGRP (87–132) to inhibit melanocortin action at the MC4R/MC1R chimeras was examined in cAMP assays (Fig.3). Consistent with the known effects of AGRP, 3 × 10−7m AGRP (87–132) potently inhibited NDP-MSH action at the MC4R (Fig. 3 A) but had no effect on NDP-MSH-stimulated cAMP generation at the MC1R (Fig.3 E). Fig. 3, B and C, reveals that there was a loss in the ability of AGRP to inhibit NDP-MSH-stimulated cAMP generation at the chimeric MC4R/MC1R containing the second or third exoloops of the MC1R. When compared with the wild-type MC4R, simultaneous substitution of exoloops 2 and 3 of the MC4R with those of the MC1R resulted in a greater loss in the ability of AGRP (87–132) to inhibit NDP-MSH-stimulated cAMP than substitution of either individual exoloop (Fig. 3 D). EC50 andK b values are reported in TableII.Table IIEffect of AGRP on cAMP formation stimulated by NDP-MSH on HEK cells transfected with chimeras of the human MC4RNDP-MSHNDP-MSH + AGRP (3 × 10−7m)NDP-MSH + AGRPAGRP K bNDP-MSHEC50 nmEC50nmnmMC4RWT0.91 ± 0.1228.8 ± 34247.81.2 ± 0.15MC4R/2eMC1R3.5 ± 0.05210 ± 1560.05.1 ± 0.35MC4R/3eMC1R4.8 ± 0.14196 ± 1240.87.5 ± 1.1MC4R/2e,3eMC1R4.65 ± 0.0757.3 ± 2012.3226.5 ± 4.1MC1RWT0.31 ± 0.040.36 ± 0.61.32937.5 ± 54 Open table in a new tab Because replacement of exoloops 2 and 3 of the MC4R with the homologous exoloops of the MC1R reduced AGRP binding affinity and inhibitory activity, we sought to ascertain the effects of the reciprocal substitutions. We hypothesized that if exoloops 2 and 3 were important to AGRP binding, then cassette substitution of those MC1R exoloops with those homologous exoloops of the MC4R would result in a chimeric MC1R that, unlike the wild-type MC1R, would be able to interact with AGRP. Fig. 4 Ademonstrates that MC1R chimeras with substitutions of their second and third exoloops of the MC4R retained a 125I-NDP-MSH binding affinity comparable to that of the wild-type MC1R. Fig. 4 B shows that AGRP (87–132) had only a minimal ability to displace 125I-NDP-MSH binding at the wild-type MC1R. A small displacement of 125I-NDP-MSH binding was observed only at very high concentrations of AGRP (87–132). In contrast, AGRP (87–132) dose-dependently displaced all bound125I-NDP-MSH from the MC4R. Remarkably, substitution of the second or third exoloop of the MC4R into the sequence of the MC1R caused a dramatic increase in the ability of AGRP (87–132) to displace125I-NDP-MSH from the basic structure of the MC1R, whose nonmutated form lacks responsiveness to AGRP. A stepwise increase in the ability of AGRP ability to displace 125I-NDP-MSH binding from chimeras MC1R/2eMC4R, MC1R/3eMC4R, and MC1R/2e,3eMC4R was observed. Simultaneous substitution of both the second and third exoloops of MC4R had an additive effect on this parameter. Notably, substitution of MC4R exoloop 3 had a greater effect on the ability of AGRP (87–132) to displace 125I-NDP-MSH binding than did MC4R exoloop 2. The IC50 values are reported in TableIII.Table IIIEffect of NDP-MSH and AGRP on 125I NDP-MSH binding on cells transfected with chimeras of the human MC1RChimerasBinding IC50NDP-MSHAGRPnMMC1RWT0.38 ± 0.05>1000MC1R/2eMC4R0.5 ± 0.12316 ± 26.5MC1R/3eMC4R0.76 ± 0.21100 ± 21MC1R/2e,3eMC4R0.71 ± 0.1432 ± 7.6MC4RWT3.81 ± 0.329 ± 1.2 Open table in a new tab Fig. 4 C shows that individual or simultaneous substitutions of the second and third exoloops of the MC4R into the sequence of the MC1R (chimeras MC1R/2eMC4R, MC1R/3eMC4R, and MC1R/2e,3eMC4R) conferred upon the MC1R a dramatically increased ability to bind125I-AGRP (87–132). These data are consistent with the ability of AGRP to displace 125I-NDP-MSH binding presented in Fig. 4 B. As shown, substitution of the third exoloop had a greater effect than substitution of the second loop, and simultaneous substitution of MC4R exoloops 2 and 3 had an additive effect. The chimerical MC1R containing both exoloops 2 and 3 of the MC4R (MC1R/2e,3eMC4R) demonstrated approximately 76% of the specific125I-AGRP (87–132) binding of the wild-type MC4R. As in the case of the MC4R/MC1R chimeras, we sought to more fully examine the functional effects of exoloop substitutions by examining the ability of AGRP (87–132) to inhibit melanocortin-stimulated cAMP generation. Fig.5 A demonstrates that 3 × 10−7m AGRP (87–132) had no ability to inhibit melanocortin-stimulated cAMP generation at the wild-type MC1R. In contrast, AGRP (87–132) is a potent antagonist of NDP-MSH action at the wild-type MC4R (Fig. 5 E). As shown in Fig. 5,B and C, placement of MC4R exoloop 2 or exoloop 3 into the sequence of the MC1R led to a chimerical receptor that, unlike the wild-type MC1R, was inhibited by AGRP. Substitution of the third exoloop of the MC1R with the third exoloop of the MC4R had a greater effect on establishing functional AGRP inhibition than did substitution of the second exoloop of the MC4R. Simultaneous substitution of exoloops 2 and 3 of the MC1R with the homologous exoloops of the MC4R led to a chimerical MC1R that could be significantly inhibited by AGRP (Fig. 5 D). The cAMP data of the MC1R/MC4R chimeras are completely consistent with radioligand binding studies from the same chimeric receptors (Fig. 4). EC50 and K bvalues are reported in Table IV.Table IVEffect of AGRP on cAMP formation stimulated by NDP-MSH on HEK cells transfected with chimeras of the human MC1RNDP-MSHNDP-MSH + AGRP (10−7m)NDP-MSH + AGRPAGRPK bNDP-MSHEC50nmEC50 nmnmMC1RWT0.31 ± 0.040.36 ± 0.61.16276.8 ± 32MC1R/2eMC4R0.18 ± 0.030.59 ± 0.113.3130 ± 13MC1R/3eMC4R0.14 ± 0.051.2 ± 0.218.639.47 ± 6.8MC1R/2e,3eMC4R0.35 ± 0.0638.4 ± 6.2109.72.75 ± 0.87MC4RWT0.92 ± 0.12228.8 ± 34247.8.11.2 ± 0.05 Open table in a new tab Recent insights into the hypothalamic control of feeding behavior indicate that α-MSH, which is produced by neurons in the arcuate nucleus, purveys a satiety signal that is mediated through the MC4R (the role of the melanocortin-3 receptor within the arcuate nucleus has not yet been clearly defined). Produced by adjacent neuropeptide Y-containing neurons, AGRP is believed to act as an opposing orexigenic agent. In its role as an important mediator of satiety, the MC4R has become a prime target for anti-obesity drug development. An understanding of the molecular basis underlying the ability of α-MSH and AGRP to bind the MC4R is of potential importance to those pharmaceutical discovery efforts. Previous three-dimensional modeling of the MC1R, extensive point mutagenesis of the human MCR subtypes 1, 3, and 4, and structure-activity studies using substituted and truncated adrenocorticotropic hormone 1–13 fragments have led us to formulate a structural model of MCR in which agonists bind to the MCRs in a relatively shallow pocket formed by the transmembrane α-helices of those receptors (19Haskell-Luevano C. Sawyer T.K. Trumpp-Kallmeyer S. Bikker J. Humblet C. Gantz I. Hruby V.J. Drug Des. Discov. 1996; 14: 197-211PubMed Google Scholar, 20Yang Y. Dickinson C. Haskell-Luevano C. Gantz I. J. Biol. Chem. 1997; 272: 23000-23010Abstract Full Text Full Text PDF PubMed Scopus (143) Google Scholar). In general, this model of melanocortin binding to the MCRs is consistent with current concepts regarding the molecular interactions of small peptide hormones and seven transmembrane G-protein-coupled receptors (21Ji T.H. Grossmann M. Ji I. J. Biol. Chem. 1998; 273: 17299-17302Abstract Full Text Full Text PDF PubMed Scopus (547) Google Scholar). A model for AGRP binding to the MCRs remains to be developed. Our previous characterization of AGRP (87–132) described a molecule that consisted only of the COOH-terminal cysteine motif of AGRP (17Yang Y.-K. Thompson D. Dickinson C.J. Wilken J. Barsh G.S. Kent S.B.H. Gantz I. Mol. Endocrinol. 1999; 13: 148-155Crossref PubMed Scopus (162) Google Scholar). Although this molecule contained no NH2-terminal amino acids proximal to the first cysteine residue present in the AGRP sequence (Cys-87), it retained full biological activity. Whereas it is plausible to assume that the deletion of cysteine residues would disrupt the AGRP tertiary structure by removing disulfide bonds and would result in a molecule with markedly diminished functional capabilities, those experiments did not test this point, and it cannot definitely be said that AGRP (87–132) is a truly “minimized” AGRP molecule. Based on the probable size of AGRP, we hypothesized that AGRP was likely to bind to receptor domains other than or in addition to transmembrane domains. Extracytoplasmic receptor domains constitute a logical site that could be involved in AGRP binding. Seven transmembrane receptors have four potential extracytoplasmic domains including the NH2 terminus, first, second, and third extracytoplasmic loops (exoloops) at which binding might occur. Importantly, the NH2 terminus of the MCRs is not long, unlike the glycoprotein, metabotropic glutamate, and ion-sensing receptors in which lengthy NH2 termini have been found to play a crucial role in ligand binding. In the present experiments, we demonstrate that exoloops 2 and 3 of the MC4R are crucial determinants of AGRP binding affinity and inhibitory activity. Substitution of MC4R exoloops 2 and 3 with those homologous exoloops of the MC1R resulted in a loss of AGRP binding affinity and inhibitory activity, whereas placement of those MC4R exoloops into the sequence of the MC1R led to the establishment of significant AGRP activity at the MC1R. A comparison of the relative degree of change in AGRP binding affinity and inhibitory activity displayed by the MCR chimeras indicates that substitution of exoloop 3 induced more profound effects than substitutions of exoloop 2. In this regard, it is interesting to note that the amino acid sequence of exoloop 3 of the MC1R and MC4R is longer than that of exoloop 2 and that exoloop 3 has a greater degree of dissimilarity in amino acid side chain charge and hydrophobicity than exoloop 2 (Fig. 1). Importantly, the observation that similar receptor numbers were expressed suggests that receptor expression is not the basis of the observed changes in cAMP assays. Whereas the present studies have clearly identified receptor domains (exoloops 2 and 3) that are involved in antagonist binding, they do not conclusively reveal the mechanism underlying those changes. Inherent to most mutagenesis studies is the inability to unequivocally discern whether pharmacological effects result from a direct disruption of atomic interactions involving the ligand and receptor or whether they result from conformational changes imposed on the receptor that only indirectly affect ligand binding. In other words, it is difficult to state with absolute certainty that the changes in AGRP binding affinity and inhibitory activity observed in these studies truly resulted from the addition or removal of specific amino acids (present in the exoloops) that have direct atomic interactions with the amino acids of AGRP. Instead, it is possible that the observed pharmacological changes simply resulted from an alteration in chimeric receptor structure unrelated to the pharmacophore binding pocket that permitted or discouraged AGRP binding, despite being uncharacteristic of the wild-type receptor on which they were based. Several observations favor an interpretation that specific sites of interaction were identified in these studies. First, all chimeras retained a 125I-NDP-MSH binding affinity similar to that observed at the wild-type MCRs, and all chimeras were activated by nanomolar concentrations of NDP-MSH. If the tertiary structure of the chimeric receptors had been drastically altered, one would expect that the ability of NDP-MSH to bind and activate the chimeric receptors should have been altered. Second, exchange of exoloops between the MC1R and MC4R led not only to a loss of the ability of AGRP to bind chimeric MCRs and inhibit NDP-MSH action (MC4R/MC1R chimeras) but also to a gain in the functional ability of AGRP to bind chimeric MCRs and inhibit NDP-MSH action (MC1R/MC4R chimeras). Despite these salient points, the possibility that small, localized conformational changes in the chimeric receptors are responsible for the observed pharmacological effects lingers. However, regardless of the distinction between direct and indirect causation, it is still possible to conclude that the present studies have identified an important intrinsic property of the MCRs residing in exoloops 2 and 3 that facilitates AGRP binding. The present studies should not be interpreted to mean that the only receptor domains involved in AGRP binding are exoloops 2 and 3. In fact, the observation that exoloops 2 and 3 of the MC4R were not sufficient to confer full AGRP binding affinity and inhibitory activity to the MC1R (chimera MC1R/2e,3eMC4R) could be viewed as evidence that other MCR domains also participate in AGRP binding. Similarly, removal of exoloops 2 and 3 from the MC4R (MC4R/2e,3eMC1R) was not sufficient to lead to a complete loss of AGRP (87–132) binding affinity and inhibitory activity. The studies reported herein primarily address the binding of AGRP to the MCRs. Nonetheless, some additional comment and a specific word of caution are warranted regarding the implications of these studies to NDP-MSH binding. Existing literature is conflicting regarding the contribution of exoloops to the binding of melanocortin agonists to the MCRs. Chhajlani et al. (22Chhajlani V. Xu X. Blauw J. Sudarshi S. Biochem. Biophys. Res. Commun. 1996; 219: 521-525Crossref PubMed Scopus (35) Google Scholar) reported that mutating several hydrophilic residues in the NH2 terminus and exoloops 1, 2, and 3 of the MC1R resulted in significant shifts in125I-NDP-MSH binding. In contrast, Schiöth et al. (23Schiöth H.B. Muceniece R. Szardening M. Prussi P. Wikberg J.E.S. Biochem. Biophys. Res. Commun. 1996; 229: 687-692Crossref PubMed Scopus (16) Google Scholar) performed mutagenesis studies on exoloop 2 of the human melanocortin-3 receptor and concluded that this exoloop was not involved in NDP-MSH binding. At the time of publication, neither of these groups had the advantage of using radioligand125I-AGRP (87–132), which is of a distinctly different nature from 125I-NDP-MSH, to check the structural integrity of their mutated receptors. Unfortunately, the present studies are not capable of resolving the controversy over whether exoloops are significantly involved in NDP-MSH binding. The fact that changing exoloops did not alter NDP-MSH binding affinity or potency could support the notion that exoloops are not critical to its binding. On the other hand, because both the MC1R and MC4R bind NDP-MSH, any binding determinants in exoloops 2 and 3 could simply have been switched in the construction of the chimeras. These data also make it tempting to hypothesize that a component of the subtype specificity of the MCRs for Agouti resides in the second and third receptor exoloops. It is not unreasonable to speculate that the relative affinity of Agouti for selected MCR subtypes could be increased or decreased by altering exoloops 2 and 3. Cassette mutagenesis experiments could be used to address this question (e.g. by exchanging the second and third exoloops between the melanocortin-3 receptor or the melanocortin-5 receptor and the MC1R and observing the increases or decreases in the affinity or inhibitory activity of Agouti). In summary, the present data have identified an important MCR structural property involved in AGRP binding that resides in exoloops 2 and 3. Whereas these exoloops appear to have a preeminent effect on AGRP binding, it will be important in the future to determine whether other MCR domains are also involved in AGRP binding and to define the relative importance of those domains compared with these exoloops. We thank Dr. L. H. T. Van der Ploeg and Dr. T. M. Fong for helpful discussions and critical evaluation of this work.
Referência(s)