Addressing the architecture, dynamics and activation mechanism of the CGRP receptor

G protein coupled receptors (GPCRs) are the largest family of proteins in the human genome and also the largest target for therapeutic drugs; thus they are of enormous scientific and practical interest. They are divided into a number of families. Of these, family-A is the best understood, but family-B includes receptors which are likely to be important in many disease states and so it is important to understand how these function, both to further our knowledge of fundamental biology and also for the design of new drugs.

Calcitonin gene-related peptide (CGRP) is found throughout the nervous system and is particularly important in regulating both the cardiovascular system (the heart and blood vessels) and also the immune system and inflammation. The receptor for CGRP is of special scientific interest as it involves a GPCR called CLR and also a second protein called RAMP1. RAMP1 is a member of a protein family that modulates a number of GPCRs of which the best characterised is CLR. CGRP is also likely to be important both in cardiovascular disorders and any disease that involves inflammation. The peptide is a major cause of migraine and drugs which block CGRP receptors have shown great promise in clinical trials; however, so far it has not been possible to use these clinically because of toxicity problems. Thus, there is an urgent need to develop new drugs that could act on CGRP receptors.

The CGRP receptor is made up of two parts. A portion called the transmembrane domain is found in the membranes of cells. This is connected to the extracellular domain, which is on the outside of cells. CGRP interacts with both parts of this structure and causes the transmembrane domain to change shape. This causes the receptor to interact with other proteins, leading to cell activation. We have a crystal structure of the part of the CGRP receptor that is on the outside of cells. Unfortunately, we do not know how CGRP binds to this, nor do we know how it binds to the transmembrane domain. This severely limits our understanding of the receptor and our ability to design drugs that could target it.

We have previously used experimental data from a technique known as site-directed mutagenesis to construct a computer model of the transmembrane domain of the CGRP receptor. This transmembrane domain is very similar to the transmembrane domains of two family-B GPCRs which were crystallised after our computer model was produced. This gives us confidence that our approach of combining experimental and computational methods is valuable. In this project, we intend to extend the approach to study how CGRP binds to both domains of the receptor and how this causes the receptor to become activated. We will use mutagenesis and also methods where we physically cross-link CGRP to the receptor to identify contact points. We will then use these to construct computational models, which we can refine through further experimentation. Using a computer, we can predict how the receptor shape will change when CGRP binds to it, so identifying the mechanism for receptor activation. This knowledge will be benefitial in the design of new drugs which can either block the receptor or promote its activation.

The CGRP receptor is a particularly interesting family B G-protein coupled receptor (GPCR) having an absolute requirement for an auxiliary protein known as Receptor activity modifying protein 1 (RAMP1). Class B GPCRs consist of a large extracellular domain (ECD) and a transmembrane domain (TMD). They frequently associate with accessory proteins belonging to the family of RAMPs. They act as receptors for a number of peptide hormones and neurotransmitters. They are attractive therapeutic targets but it has proved very difficult to obtain drugs that target them. Several crystal structures exist for the ECDs and there are crystal structures for two class B GPCRs (glucagon and CRF), but neither have bound peptides and the orientation between the TMD and ECD for any receptor remains speculative, as does the mechanism whereby agonists activate the receptors.

We have recently used a combination of site-directed mutagenesis and molecular modelling to propose a structure for CGRP bound to the TMD of CLR. This shows excellent agreement with the crystal structures, which were published after our modelled structures were deposited. Thus we propose that our methodology is robust. Furthermore, the presence of the RAMP provides additional constraints on the orientation of the ECD relative to the TMD, making the CGRP receptor especially amenable to modelling by greatly reducing the number of ways in which it could be modelled incorrectly.

We propose a strategy of photoaffinity cross-linking, disulphide trapping and point mutagenesis to provide experimental information on the architecture of the receptor when bound to CGRP and as a test for the modelling. This information will then be used to produce a model of the complex. We will use molecular dynamics and other modelling techniques to plan the experiments, to interpret the results and hence to determine the conformational changes caused by CGRP binding and so establish how the receptor is activated by its native agonist.

The chief beneficiaries will include industrial scientists who are seeking to develop new drugs and the companies which employ them. In turn this would have economic benefits for the UK and also enhance the quality of life from anyone who might receive such drugs (which would further benefit the UK economy).

The most immediate beneficiaries would be those companies with research programmes directed towards the development of CGRP antagonists for migraine, where there is clinical evidence of the effectiveness of these agents. Migraine alone is estimated to cost the UK economy £2.25 billion per annum (Steiner TJ., Lecture to the All Party Parliamentary Group on Primary Headache Disorders., 19 November 2008) and CGRP antagonists have been shown to be effective against migraine in clinical trials. There have been 66 new patent applications filed worldwide for CGRP antagonists since January 2010. Thus the development of new agents to target the CGRP receptor would be of considerable benefit both to the UK pharmaceutical industry and also the health and well-being of the UK population. The mode of binding of CGRP and the way it activates its receptor is likely to be shared by other peptides in this family such as amylin (implicated in the control of eating) and calcitonin (well-established for the treatment of osteoporosis), further adding to the value of the project. The spectrum of disorders covered by the CGRP family of peptides include many which are common amongst elderly populations (e.g. heart failure, osteoporosis) and so this project is relevant to the BBSRC's initiative on lifelong health and well-being. More broadly, the challenges resulting from CLR modelling have serendipitously resulted in the generation of a helix alignment program that can work below the "twilight zone" to align distantly related proteins (Vohra et al., J. Roy. Soc. 2013, Taddese et al., Plant Phys 2014) and we expect other methodologies to result from this challenging problem. Here, the way in which the modelling is closely integrated with experiment be applicable to a wide range of proteins of pharmaceutical or other interest. These include G-protein coupled receptors but extend far beyond those. In this respect, the project also addresses the BBSRC initiative on Technology development for the biosciences.