Multiscale simulations and Markov state models of peripheral membrane proteins

Peripheral membrane proteins (PMPs) interact with the cell membrane. They are important for many cell processes and are important therapeutic targets. This thesis seeks to investigate the membrane recognition and binding processes that are integral to the function of two PMP systems. Phosphatidylinositol phosphates (PIPs) are lipid signaling molecules that play key roles in many cellular processes. The PIP5K1A kinase catalyses phosphorylation of PI4P to form PIP2, which in turn interacts with membrane and membrane-associated proteins. We present a multiscale simulation study of the PIP5K1A kinase and its interaction with PIPs in the membrane. The study showed that the PIP5K1A kinase interacts with PIPs via its activation loop, leading to reorientation and binding. It also showed that the dimeric PIP5K1A binds with PIPs via only one catalytic site at a time, which has important implications for the role of membrane curvature and composition in PMP binding. This is addressed by large-scale simulations of PIP5K1A signalling. Furthermore, Markov state models (MSMs) were applied to the activation loop. This work showed that the loop remains disordered in the membrane but adopts specific conformations that may help explain its specificity. These methods were extended to another PMP system of interest, α- synuclein. The results showed that interaction with the membrane is led by the interhelical region, and offers a structural perspective of a break in the helical structure, which is supported by experimental studies. This may inform our understanding of α- synuclein aggregation, and more generally aid in our understanding of PMP signalling.