THE ROLE OF THE PROTEIN-LIPID BOUNDARY IN THE GATING OF THE MECHANOSENSITIVE CHANNEL MSCS, AND THE THERMODYNAMICS OF ARGININE-PHOSPHATE INTERACTIONS

Abstract

Bacteria are exceptionally adaptive to a wide range of conditions. The E. coli mechanosensitive channel MscS is a low-threshold osmolyte release valve that provides environmental stability by regulating turgor pressure to prevent cell swelling and lysis in response to hypoosmotic shock. MscS is an adaptive multi-state channel that gates directly in response to tension in the surrounding lipid bilayer. Although MscS has three functional states (closed, open, inactivated), there are only two classes of structures: (1) nonconductive, characterized by splayed lipid-facing helices, kinked pore-lining helices, and lipid perturbations at the cytoplasmic interface and (2) semi-open conductive, characterized by an expanded pore that does not fully satisfy the experimental conductance. Currently, there is no consensus on how to relate these structural states to functional states. By default, the nonconductive structure is regularly assumed closed in the literature. In this thesis, I contribute to the body of existing experimental evidence that strongly suggests that the nonconductive structure corresponds to the inactivated state, rather than the closed state. Specifically, I focus on the channel as a membrane-embedded physical object and look to examine how lipids mediate tension-driven conformational dynamics. I use mutagenesis and patch-clamp electrophysiology to determine how MscS mutants with different protein-lipid interactions alter functional state distributions and transition rates. I then leverage these data to inform structure interpretation. Correctly identifying the structures of MscS that correspond to each functional state and the physical factors that stabilize them is critical towards understanding the underlying mechanism for MscS mechanosensitivity and its adaptive functional cycle. Chapter 2 explores how mutations of conserved anchor residues R46 and R74, interacting with lipid phosphates, affect gating transitions. We find that mutations at these positions predominantly alter the kinetics and voltage dependence of slow inactivation transitions, suggesting that extensive lipid rearrangement around these residues is a structural feature of inactivation. We also identify membrane potential as a factor regulating MscS state distribution. Chapter 3 investigates the role of protein-lipid interactions at both cytoplasmic and periplasmic interfaces in MscS functional behavior. Results indicate that MscS requires TM helix mobility at the periplasmic interface, but helix stability at the cytoplasmic interface for proper state transitions. We also find an interesting mutant, R46L/R74L, that is highly predisposed toward the inactivated state in giant spheroplasts, but apparently distributed normally into the closed state in actively metabolizing bacteria, providing evidence that the MscS population is under metabolic control. Finally, Chapter 4 aims to improve methods for examining these interactions in silico. The thermodynamics of arginine small peptide interactions with POPC, POPA, and POPG phospholipids is determined using ITC, and the affinity is found to depend on the accessibility of the lipid phosphate group. I also identify the ensemble of peptide-membrane bound states by constructing Markov state models from clustered trajectory data, revealing discrepancies between experimental and simulation results. These data are the first steps toward improving FF descriptions of arginine-phosphate interactions within membranes.