Development of Picomolar Affinity Antagonists and Long‐acting Agonists for the Adrenomedullin and CGRP Receptors Using Combinatorial Peptide Library and Structure‐guided Design Approaches

Abstract

Adrenomedullin (AM) and calcitonin gene‐related peptide (CGRP) have diverse functions in the cardiovascular, lymphatic, and nervous systems including regulation of vascular tone, cardioprotection, and pain transmission. These peptides signal through three heterodimeric receptors comprising the class B GPCR, the calcitonin receptor‐like receptor (CLR), and one of three receptor activity‐modifying proteins (RAMP1‐3) that determine ligand selectivity. CLR‐RAMP complexes are proven or promising drug targets for several diseases, however, the moderate binding affinities and short plasma half‐lives of AM and CGRP limit their utility as therapeutics. Here, we used synthetic peptide combinatorial library and rational structure‐guided design approaches to develop AM and CGRP variant peptides with dramatically enhanced receptor affinities. The binding of the variants to purified RAMP‐CLR extracellular domain (ECD) complexes was characterized and a high‐resolution crystal structure of an AM variant bound to RAMP2‐CLR ECD explained its enhanced affinity. Incorporation of the variants into different length AM and CGRP scaffolds yielded picomolar affinity antagonists or agonists that exhibited substantially prolonged cAMP signaling after ligand washout. These activities were demonstrated in a model cell line transiently expressing the receptors and in primary human cells or immortalized cell lines that naturally express the receptors. This work provides a suite of AM and CGRP variants that will be valuable pharmacological tools and may have promise as peptide therapeutics with long receptor residence times. In addition, the properties of the variants support a mechanism for RAMP function in which direct RAMP‐peptide contacts and allosteric modulation of CLR cooperate to determine ligand selectivity.Support or Funding InformationThis work was supported by grants NIH R01GM104251 and OCAST HR16‐005 (AAP) and AHA predoctoral fellowship 18PRE33990152 (JMB).