Modelling the pH- and Lipid-Dependent Activity of the Antimicrobial Protein Amoebapore A

Abstract

Amoebapore A (APA-1) is an antimicrobial protein produced by the human parasite Entamoeba histolytica. The activity of APA-1 has been hypothesised to be regulated by a pH-dependent dimerisation mechanism that enables the protein to assemble into a pore structure in membranes. This research thesis has characterised the pH-dependent dimerisation and membrane binding activity of APA-1 using a variety of techniques. Production of APA-1 samples was achieved using cell-free protein synthesis. Combinatorial 15N-labelling combined with 1H chemical shift data and three-dimensional NMR experiments were used to assign NMR spectra. NMR chemical shift perturbation analysis identified residues located at the dimer interface and, in conjunction with HADDOCK a new dimer model was derived. Small angle X-scattering (SAXS) experiments confirmed that the NMR-derived dimer model matched the SAXS data statistically better than the previously derived dimer model. The equilibrium dissociation constant (Kd) for the APA-1 dimer was measured using size-exclusion chromatography and NMR spectroscopy and was found to be 247 ± 32 μM at the optimal activity pH of 5.2, and >850 μM when measured at pH 3.0, indicating that dimer formation involves ionic interactions. The pKa values of Asp, Glu and His residues were determined by NMR spectroscopy to gain insight into the functional role of these residues. Intriguingly, the pKa values of Glu16 and Glu62 (5.3), which are structurally proximal, were found to match the pH range where APA-1 activity dramatically decreases. This suggests that the ionisation state of these residues is important in regulating APA-1 membrane interactions. Quartz crystal microbalance with dissipation and neutron reflectometry were used to examine the membrane binding properties of APA-1. Different bilayers mimicking eukaryotic and prokaryotic membranes on surfaces were chosen and changes in the lipid bilayer architecture upon addition of APA-1 were analysed. APA-1 bound favourably to membranes containing negatively charged phosphatidylglycerol, supporting the hypothesis that positive electrostatic regions on APA-1 are essential for the initial recognition of a negatively charged membrane surface. APA-1 bound more efficiently to loosely deposited phospholipids and bilayers with defects, and the membrane interaction was found to be reversible. APA-1 was capable of removing thick lipid deposits that had not formed stable bilayers, suggesting that APA-1 can facilitate shedding of lipid material. No clear evidence of pore formation was observed.