Structural study of the HIV1 gp41 coat protein

National Institute Of Diabetes And Digestive And Kidney Diseases

Investigators

Linked publications & trials

Abstract

Truncation of the gp41 construct just past its transmembrane domain, leaving residues 1-194, results in a protein that is soluble in detergent micelles. Although previously believed to adopt its natural homo-trimeric form, we have found strong evidence for a rapid dynamic equilibrium between monomeric and trimeric forms of the protein, which can be shifted to the mostly trimeric form by choice of suitable detergents and detergent:protein stoichiometry. The NOE data and 15N relaxation rates show a well structured helical conformation for residues 5-14 of the fusion peptide region, followed by a more disordered FPPR segment which connects it to the ecto-domain. Remarkably, resonances from the C-terminal heptad repeat (CHR) and membrane proximal region MPER) are exchange broadened to such an extent that they do not give rise to observable NMR resonances. Together with exchange broadening observed in the highly mobile immuno-dominant loop region, this points to an equilibrium between early and late stage 3- and 6-helical states for the ecto domain. There is no evidence for direct interaction with the transmembrane fusion domain in our studies carried out at pH4, and both light scattering, SAXS, and NMR diffusion data point to a mass ratio of detergent:protein slightly above 1.0. The protein is found to be quite stable, even at temperatures of 40C. The fusion peptide exhibits effective correlation times that are much shorter than for the ecto domain, indicating that the intact fusion domain helix is highly mobile within the detergent micelle/protein aggregate, while retaining a helical conformation. This excludes the previously hypothesized interaction with the gp41 transmembrane domain under our conditions of pH and detergent. When studying a protein construct that consists of just the N- and C-terminal heptad repeat regions of gp41 (NHR and CHR), with the immunogenic loop region replaces by a short linker, our NMR data indicate the 6-helical bundle structure adopted by the protein is indistinguishable from the crystallographic model reported for this structure. However, we find that in the presence of detergent micelles or small unilamellar vesicles, the trimer falls apart while the NHR and CHR retain their alpha-helical structure. Both helices were found to interact tightly with the phospholipids, suggesting an active role for the gp41 ecto domain in mediating membrane fusion. However, at near neutral pH there is evidence of interaction between the fusion peptide proximal region (FPPR) and the MPER region, potentially providing the initial step in the refolding of the membrane associated monomeric structures into the post-fusion state 6-helical bundle trimeric state, previously identified by X-ray crystallography.