Comparitive Computational Models of TRPM8 Help Explain Modulation Mechanisms

Background. Transient Receptor Potential (TRP) channels are a diverse class of ion channels noted for polymodal gating mechanisms and roles in sensation, including vision, thermosensation, and mechanosensation. TRPM8 is a TRP channel implicated in cold sensation, nociception, and a variety of human diseases, including obesity and cancer. Despite sustained interest in TRPM8 since its discovery in 2001, many of the molecular mechanisms that underlie function are not yet clear. Knowledge of these properties could have implications for medicine and physiological understanding of sensation and signaling. Structures of TRP channels have proven challenging to solve, but recent cryo-EM structures of TRPV1 provide a basis for homology-based modeling of TRP channel structures and interactions. Here, I present a computational homology model of TRPM8 based on TRPV1 and in silico binding experiments with known TRPM8 ligands, including PIP2 and menthol. Methods. Bioinformatics tools were used to generate a multiple sequence alignment of human TRPM8 to templates- including TRPV1- based on structural similarity. The Rosetta 3.5 loop rebuilding utility was used in conjunction with this alignment to construct a set of homology models. These models were ranked using Rosetta's scoring functions, which include membrane-specific terms for lipophilicity and secondary structure. RosettaMembrane and RosettaSymmetry were used, which should improve the predictive power of the models. The top scoring models were analyzed for consistency and presence of relevant binding pockets, then docked to menthol and PIP2. Results. Initial models suggest plausible binding pockets for menthol near Y745, an essential residue for menthol modulation according to electrophysiology studies. There is also evidence suggestive of a PIP2 binding site near the TRP box, a well-conserved region in the TRP superfamily that has been implicated in PIP2 modulation by functional studies. This binding site is formed in part through interaction between S0, an N-terminal amphipathic helix and the TRP domain. Conclusion. The models presented provide a platform to investigate the structural basis of TRPM8 ligand modulation complementary to existing functional and structural information. The in silico binding experiments provide testable hypotheses that can guide in vitro experiments, including solution NMR binding studies. For example, the model predicts that the S0 and the TRP box are close in space. This prediction can be validated with a variety of experimental biophysical and structural techniques.