Solid-state nuclear magnetic resonance studies of the structures, membrane locations, cholesterol contact, and membrane motions of membrane-associated HIV Fusion Peptide (HFP

Membrane fusion is the key step during HIV viral entry to cells, and the process is catalyzed by HIV membrane fusion protein gp41. HFP is the ~25-residue N-terminal domain of gp41 and is required for membrane fusion with significant decreases in fusion activity with point mutations. Both viral and host cell membrane contain ~30mol % cholesterol (CHOL), and HFP induced fusion is faster in membrane with CHOL. However, how HFP interacts with membrane lipids and CHOL is unknown. In this thesis, we used the newly developed 13C-2H Rotational Echo Double Resonance (REDOR) solid-state NMR method to study the membrane location of HFP in chemically-native membrane environment. HFP is 13CO labeled at specific residue, and the membrane is deuterated at specific regions of the membrane using selective regions deuterated phosphatidylcholine (PC) and CHOL. We study HFP wild type, HFP_V2E and L9R mutants because these two mutants eliminate and decrease fusion respectively. HFP is predominantly β sheet structure in bilayer membrane for both HFP wild type and HFP_V2E mutant, HFP_L9R has a different structure and is likely helical. Both HFP and HFP_V2E mutant have major deeply-inserted membrane location contacting membrane center and minor shallowly-inserted membrane location contacting half way of one membrane leaflet. The HFP_V2E mutant has bigger fraction of molecules with shallower membrane location, which is consistent with the strong correlation between membrane location insertion depth and the peptide fusogenicity. HFP_L9R mutant has majorly deeply inserted into membrane.By comparing the HFP- PC and HFP- CHOL contact, there is preferential contact between HFP and CHOL vs PC at several residues including G5, G10 and G16. The free energy difference for contacting PC vs CHOL is ~ 0.57(5) kcal.mol-1 for T= 300K. HFP- CHOL contact geometry is successfully modeled by Swiss Dock and YASARA energy minimization with two strands antiparallel HFP (1→16/16→1 registry). There are two energetically favorable binding models between HFP and CHOL, from docking, energy minimization and consistency with REDOR results. The contact models reveal tilted and curved-up tail orientation of Chol_d7. Fusion may be catalyzed by matching the curvature of lipids contacting HFPs with the membrane curvature during the fusion intermediates like the stalk. Membrane motion perturbation by HFP is studied by static deuterium NMR from deuterium powder pattern spectrum, order parameter profile and T2 relaxation time. The DMPC-d54 spectrum becomes ~10% narrower in membrane without CHOL with 4% HFP and in membrane with 33% CHOL with 1% HFP. Accordingly, the order parameter of lipid acyl chain becomes ~ 1-10% disordered by HFP. However, the spectrum becomes 20% broader in membrane with 33% CHOL with 4% HFP, and the order parameter of lipid acyl chain becomes ~ 20- 30% ordered by HFP. With HFP at 37 °C, DMPC-d54 T2 decreases ~ 70 %, and the CHOL T2 decreases ~ 30%. T2 reduction is probably associated with increased membrane curvature induced by HFP. With greater membrane curvature, the C-D bond will experience more orientation diversity relative to the external magnetic field. Thus, the quadrupolar field will have greater change, leading to faster relaxation and shorter T2. Gp41_V2E mutant eliminates cell-cell fusion. Our CD spectroscopy studies show that the FPHM_V2E mutant is helical and the melting temperature is above 90 °C in 10mM Tris buffer + 0.2 % SDS at pH 7.4. Protein is trimer and induces no lipid mixing in PC: CHOL= 2:1 vesicles.