Discoidin domain receptor signalling: from crystal structures to mechanisms

We want to find out how specialised sensor proteins on the cell surface, called DDRs, transmit a signal into the cell. This research is important because faulty DDR signalling in humans can cause disease, for example arthritis and cancer. In order to develop drugs that target the DDRs, it is necessary to know the structure and understand the signalling mechanism of the DDRs. Our research will contribute towards this aim. The DDRs are composed of three parts: one part sticks out of the cell, a second part is embedded in the cell membrane, and a third part faces the interior of the cell. Previous studies have established that the DDRs are sensing the presence of collagen, which is a major constituent of all connective tissues. This allows cells to react to their environment. When collagen binds to the exterior part of DDRs, the interior part becomes active and informs the cell that collagen has been recognised. We do not understand what happens during this process, but we believe that the DDRs change their shape when they react with collagen. We will use a range of experimental techniques to study how DDRs bind collagen and how they respond to collagen binding. Collagen is difficult to study because it is a very large molecule. In our experiments we will use collagen peptides, which are small synthetic fragments of collagen. We have found that some collagen peptides can stimulate DDR activity, whereas others block DDR activity. Our previous experiments have revealed how DDR recognises a stimulatory collagen peptide. We now want to see how a blocking peptide is recognised. By comparing the DDR structures obtained with stimulatory and blocking peptides, we hope to learn more about the shape changes that occur during signal transmission. Our previous experiments have also revealed that the whole exterior part of DDRs pairs with itself to form what is called a dimer. We will make small, precise changes in the DDR dimer and study their effects on signalling. By combining many such observations, we can construct a detailed map of the regions within the DDR molecule that are involved in transmitting a signal across the cell membrane. Finally, we have created a very useful tool for studying DDR function: a set of antibodies directed against DDR1 (DDR1 is one of the two DDRs in humans). Antibodies are normally made by animals as a defence against pathogens, but by injecting mice with DDR1 protein, we have been able to obtain antibodies that bind to DDR1. Some of these antibodies block DDR1 signalling, whereas others stimulate this activity even in the absence of collagen. In order to understand how these antibodies influence DDR1 activity, we want to determine where on the DDR1 molecule they bind. This information will complement our map of functionally important regions within the DDR1 molecule. We also plan to create antibodies against the second human DDR, DDR2. Our antibodies will be extremely useful research tools that can be shared with other researchers. They could also be used for diagnostic or therapeutic purposes, but the development of such reagents would have to be carried out in a pharmaceutical company.

The human discoidin domain receptors, DDR1 and DDR2, are unusual receptor tyrosine kinases that are activated by a major constituent of the extracellular matrix, collagen. The DDRs play fundamental roles in cell adhesion, migration and proliferation, and regulate matrix remodelling. With regards to human health, they are implicated in the progression of fibrotic diseases, arthritis and several types of cancer. Understanding DDR function at the molecular and cellular level is of great importance, but many fundamental questions remain unanswered. We previously showed that the DDRs are constitutively dimeric in the absence of ligand, unlike conventional receptor tyrosine kinases, and that specific residues within the DDR transmembrane domain are essential for receptor activation. We also determined crystal structures of the entire DDR1 ectodomain and of the DDR2 discoidin domain bound to an agonistic collagen peptide. The aim of this proposal is to define the mechanism of transmembrane signalling by constitutive DDR dimers. Using protein crystallography and our own function-blocking anti-DDR1 monoclonal antibodies we will reveal the binding modes of antagonistic collagen sequences and the conformation of the DDR1 dimer in its inactive state. Using extensive site-directed mutagenesis we will identify the DDR1 regions and dimer contacts that are associated with receptor activation. Finally, we want to generate functional anti-DDR2 monoclonal antibodies to enable further structure-function studies. The proposed research will provide detailed mechanistic insight into how the dimeric DDRs transmit a signal across the cell membrane. In addition, our research will generate unique tools that may prove valuable in the development of therapies that target unwanted DDR signalling in human diseases.

The proposed research will significantly advance the knowledge base of academic research into receptor tyrosine kinases and collagens. Researchers in many fields will profit directly from the fundamental insights and reagents that will be generated by our research (for details, see the section on academic beneficiaries). Although the proposed research is basic, it is conceivable that commercially exploitable results will be obtained. Any intellectual property arising from this research grant will be exploited through Imperial College Innovations Ltd, the College's technology transfer company. As the DDRs are new drug targets for a number of human diseases, pharmaceutical companies aiming to develop anti-DDR compounds will be interested in the results of our studies. In this regard, BL has already consulted for Novartis Institute of Biomedical Research, Horsham, UK and Boehringer Ingelheim Pharmaceuticals USA, who are establishing a drug development programme against unwanted DDR signalling. A detailed biochemical understanding of the DDR activation mechanism is likely to be required for any such programme. Other receptor tyrosine kinases (RTKs) are the targets of several drugs, but our recent research has identified fundamental mechanistic differences between DDRs and canonical RTKs. The proposed research will provide the necessary framework for future diagnostic and therapeutic applications. The generation of anti-DDR2 monoclonal antibodies (mAbs) in rabbits may benefit patients in the long term, if these mAbs can be developed into drugs. We already obtained mouse anti-DDR1 mAbs that inhibit DDR1 signalling, and aim to obtain rabbit mAbs against DDR2 with similar characteristics. Like mouse mAbs, rabbit mAbs will need to be humanised before they can be used as drugs in humans, but this process is firmly established. Rabbit mAbs have a potential advantage over mouse mAbs in that they are often cross-reactive with the mouse antigens. Thus rabbit mAbs can be used in mouse models of human diseases, which is extremely useful for pre-clinical evaluations.