CGRP receptor heterogeneity: a role for receptor component protein?

Abstract

We read with interest the recent review ‘Calcitonin generelated peptide in pregnancy and its emerging receptor heterogeneity’ by Yallampalli et al. [1]. We welcome the interest in this peptide (CGRP). However, whilst the authors are correct to point out that there appears to be heterogeneity in the receptors activated by CGRP, we feel that the evidence in the literature does not support their explanation. The concept of CGRP receptor heterogeneity was introduced over a decade ago [2] based on the differential antagonist affinity of CGRP fragments, such as CGRP8–37. Two classes of CGRP receptor have been identified pharmacologically: CGRP1 receptors are potently antagonized by CGRP8–37, whereas CGRP2 receptors require higher concentrations of this antagonist to achieve the same effect [2]. We refer readers to the recent account of CGRP pharmacology produced by the International Union of Pharmacology (IUPHAR) CGRP nomenclature subcommittee for a full discussion of this matter [3]. It is now clear that the CGRP1 receptor defined in IUPHAR guidelines is a heteromeric complex including the seven-transmembrane calcitonin receptor-like receptor (CRLR or CL) and receptor activity modifying protein 1 (RAMP1) [3,4]. This corresponds to the ‘CGRP-A’ receptor discussed by Yallampalli et al. A molecular correlate for the CGRP2 receptor has yet to be identified. Whilst the nomenclature issue might be minor, we have more serious concerns about the ‘CGRP-B’ receptor described by Yallampalli et al. The authors do not explain how this relates to the CGRP2 receptor that has been proposed on pharmacological grounds, and suggest that their ‘CGRP-B’ receptor could be a complex between receptor component protein (RCP) and a ‘RDC-1 or G10d-like’ receptor. G10d and RDC1 were once claimed to be receptors for adrenomedullin and CGRP [5,6], but this work could not be repeated [4,7] and these receptors are no longer considered to be candidate CGRP/adrenomedullin receptors [3]. Why such receptors should be considered to have any affinity for CGRP is unclear. Furthermore, RCP is an intracellular peripheral membrane protein that was first identified as conferring CGRP signaling to oocytes after injection of in vitro transcribed RNA [8]. Unfortunately, the literature on this protein has been misunderstood and the account in [1] contains errors, particularly with regard to the receptor with which RCP interacts. There are four important misrepresentations: (1) ‘The RCP, a cytosolic protein, has been proposed to modulate CGRP binding via an as yet unidentified 7TM receptor’. RCP has never been implicated in ligand binding but is instead important for signal transduction [9,10]. Furthermore, co-immunoprecipitation studies clearly suggest that RCP works in conjunction with the previously identified receptor, CRLR [9,10]. Although it is possible that RCP functions with additional receptors, it is misleading to present an ‘unidentified 7-TM receptor’ as the target of RCP actions. (2) ‘CRLR and CTR, with RAMPs as accessory proteins, constitute one system, and RDC1and G10d-like receptors, with RCP as an accessory protein, could form another’. All the experimental data to date suggest that CRLR, RAMP1 and RCP are present in the same receptor system [9,10]. (3) ‘These discrepant findings could either be explained by diffused low-level expression of CRLR mRNA in the cerebellum in spite of a high density of CGRP-binding sites, or it might implicate the existence of another putative CGRP receptor type (Ref. 37 in [1]). This concept is supported by the report that RCP is not associated with CRLR (Ref. 50 in [1])’. CRLR mRNA expression might not always correlate with protein expression [12]. Furthermore, the reference cited for a lack of interaction between RCP and CRLR, Aiyar et al. [11] (1992), pre-dates the publication of RCP by four years (1996). (4) ‘The well-characterized monoclonal antibody that was raised against ligand affinity purified CGRP receptor protein from porcine cerebellum (Refs 6,49,51 in [1]) does not react with CRLR’. The sequence of the protein to which this antibody was raised has never been published, and is notably uncharacterized.