Background:  The established role of the gp41 tryptophan (W)-rich membrane proximal external region (MPER) (K₆₆₅WASWNWFN₆₇₄ITNW₆₇₈LWYIK) in HIV-1 fusion and the presence of broadly neutralizing epitopes (including 2F5 and 4E10) make it an important domain for study. The MPER is known to be membrane-disruptive, and the aromatic residues form a “collar” along the helix potentially enabling “velcro-like” interactions with the viral membrane. In this study, the importance of MPER collar arrangement and bilayer-destabilizing properties in fusion was examined.

Methods:  “Phase” mutants (D674, insA674, ins674-675) were created that disrupt the collar, and several membrane-disruptive chimera (MDC) were designed that replaced the MPER with part of or the entire W-rich antimicrobial indolicidin peptide (ILPWKWPWWPWRR) and its alanine-substituted variant CP10A, yielding KWILPWKWPWWPWRRAYIK (Ind1), and KWILAWKWAWWAWRRAYIK (CP1). We designed 2 MDC with a spacer (ITN) before W₆₇₈, KWILP/AWKWP/AWITNWP/AWYIK (Ind2 and CP2), to mimic MPER W-spacing. Finally, the 2 arginines in Ind1 were substituted with alanines (KWILPWKWPWWPWAAAYIK (Ind3). Cell-cell fusion and viral entry were quantitated using a luciferase assay and JC53BL indicator cells.

Results:  The phase mutants retained fusion efficiencies up to 58% of WT, arguing that a strict a-helical organization of this region is not critical for fusion. This conclusion is supported by the surprising observation that substitution of the proline-rich indolicidin peptide yielded low levels of fusion (5%) that were increased to 33% that of WT if the charged arginines were replaced by alanines. The Ind2 mutant, that lacks charges and has a W-spacing similar to MPER, retained fusion up to 46% of WT even though it is highly proline-rich. CP1 (~1%) and CP2 (~7%) showed fusion efficiencies lower than Ind1 and Ind2, respectively, implying that the capacity to form an a-helix is not crucial for fusion. Viral entry of all the mutants was dramatically reduced, with the best (insA674) exhibiting only ~1% that of WT. 

Conclusions: The replacement of HIV-1 gp41 MPER with a proline-rich membrane-disruptive peptide does not prevent fusion, strongly indicating that the aromatic collar and a-helical potential are not essential for this function. In contrast, the same substitutions effectively block viral entry, suggesting that other structural constraints are imposed on this region, possibly at the level of glycoprotein incorporation.