Phosphatidylinositol-specific phospholipase C

Abstract

Phosphatidylinositol-specific phospholipase C (PI-PLC) from B. thuringiensis is activated by phosphatidylcholine (PC) surfaces for both phosphatidylinositol (PI) cleavage to inositol 1,2-(cyclic)-phosphate (cIP) and subsequent hydrolysis of cIP to inositol-1-phosphate. These enzyme kinetics strongly suggest that this PI-PLC has two discrete binding sites for phospholipids - the active site binding PI (or substrate competitors) and an activator site specific for PC. However, it is difficult to determine the orientation and conformation of peripheral membrane proteins when docked to target membranes, let alone where sites for these might be on the protein. In this thesis, various biophysical techniques were applied to this bacterial PI-PLC to obtain structural information in the absence and presence of membranes to characterize specific conformational changes that occur when the protein binds to activating membranes. The crystal structures of an interfacially impaired double mutant of PI-PLC, W47A/W242A, was solved and showed the protein as a homodimer. The major interactions came from four clustered surface tyrosine residues from each monomer. This structure suggested the possibility of PI-PLC dimerization on membrane surfaces as part of the mechanism for interfacial activation. Mutations of these tyrosines showed a loss of activity and membrane binding. Crystal structures of these mutant proteins showed no significant change in the proteins, consistent with either disruption of a dimerization interface of a specific PC binding motif. FRET was used to try and monitor oligomerization of PI-PLC, derivatized on a cysteine introduced at residue 280 (W280C) with either a donor or acceptor fluorophore, on vesicle surfaces. The results suggested some specific aggregation could occur on very PC-rich surfaces but not on phospholipid vesicles with at least 50 mol% anionic phospholipids, strongly suggesting that a stable dimer was not forming when the enzyme was bound to vesicles mimicking conditions where enzyme specific activity is high. If dimerization occurs on surfaces, it must be transient. To examine which portions of the PI-PLC are interacting with membrane and to further explore if there is any evidence for PI-PLC dimerization on membrane surface, deuterium exchange coupled by mass spectrometry experiments were carried out with wild type PI-PLC, W47A/W242A and a covalent dimer formed from W242C that is more active than wild type enzyme. Results showed (i) a stable short helix B (containing an exposed tryptophan thought to insert into membranes) in wild type PI-PLC and its complete destabilization in W47A/W242C, (ii) a flexible surface loop (containing another tryptophan thought to partition into the membrane) that became protected when the protein was bound to vesicles, and (iii) reduced deuterium exchange for the peptide containing the tyrosines that either mediate transient dimerization or form a PC binding site.. These observations modify how we envision the protein anchoring to substrate-containing membranes.