Six-helix bundle completion in the distal C-terminal heptad repeat region of gp41 is required for efficient human immunodeficiency virus type 1 infection

Abstract

Background: The native pre-fusion structure of gp120/gp41 complex of human immunodeficiency virus type 1 was recently revealed. In the model, the helices of gp41 (α6, α7, α8, and α9) form a four-helix collar underneath trimeric gp120. Gp41 is a class I fusion protein and mediates membrane fusion by forming a post-fusion structure called the six-helix bundle (6HB). The comparison of the pre- and post-fusion structures revealed the large conformational changes in gp41 during the antiparallel packing of the N- and C-terminal heptad repeats (NHRs and CHRs) in membrane fusion. Several mutagenesis studies of gp41 performed in the past were interpreted based on 6HB, the only available structure at that time. To obtain an insight about the current pre-fusion structural model and conformational changes during membrane fusion, alanine insertion mutagenesis of the NHR, CHR and connecting loop regions of HXB2 gp41 was performed. The effects of mutations on biosynthesis and membrane fusion were analyzed by immunoblotting and fusion assays, respectively. The extent of membrane fusion was evaluated by split luciferase-based pore formation and syncytia formation assays, respectively.

Results: Consistent with the current structural model, drastic negative effects of mutations on biosynthesis and membrane fusion were observed for NHR, loop, and proximal regions of CHR (up to amino acid position 643). The insertions in α9 after it leaves the four-helix collar were tolerable for biosynthesis. These CHR mutants showed varying effects on membrane fusion. Insertion at position 644 or 645 resulted in poor pore and syncytia formation. Efficient pore and syncytia formation almost similar to that of the wild type was observed for insertion at position 647, 648 or 649. However, recovery of virus infectivity was only observed for the insertions beyond position 648.

Conclusions: The mutagenesis data for HXB2 gp41 is in agreement with the recent pre-fusion structure model. The virus infection data suggested that fusion pores sufficiently large enough for the release of the virus genome complex are formed after the completion of 6HB beyond position 648.

Keywords: Envelope protein; Fusion pore; Heptad repeat; Human immunodeficiency virus type 1; Membrane fusion; Six-helix bundle; Split green fluorescent protein; Split luciferase.

Fig. 1

Conformational changes in Env during membrane fusion. a Pre-fusion and post-fusion structures of gp41. Left panel: Schematic representation of the prefusion structure of Env based on the structure of BG505 SOSIP.664, as determined by CryoEM (PDB 4TVP). The upper and lower images depict the side and bottom (from the viral membrane) views, respectively. The main chain of the gp120 subunit is shown in gray, and the gp41 subunit is shown in various colors corresponding to different α-helices. The loop between the NHR and CHR is shown in yellow. Note that only a monomer of Env is shown in the left panel, and bound antibodies PGT122 and 35O22 are removed for clarity. The N- and C- terminus of gp120 are wrapped in the center of the complex formed by gp41 [7]. Right panel: the post-fusion structures of the six-helix bundle (6HB) of gp41 (PDB 1AIK). NHR and CHR are shown in the same color assignment as in the left panel. The side and bottom views are shown. The images were generated by the University of California, San Francisco Chimera program. b The alanine insertion mutants used in this study. The insertion sites for alanine are indicated by arrowheads above the sequence of gp41. The sequence and secondary structures of HIV-1 gp41 are indicated. Both pre-fusion and post-fusion secondary structures are colored corresponding to a. Cylinders represent α-helices, and the disordered regions are indicated by “x”. The regions corresponding to N36 and C34 are also shown. Mutants are named by the position of the inserted alanine residue (such as 645+A). The numbering is based on HXB2 Env

Fig. 2

The protein profiles of Env mutants. The expression and processing of Env were analyzed by immunoblotting. Representative results of immunoblotting analysis of Env with NHR and loop mutants (a) and CHR mutants (c) expressed in the transfected 293FT cells probed with goat anti-gp120 or with Chessie 8 anti-gp41 antibodies are shown. The gp160 and gp41 bands were then quantified (b and d) by measuring the intensity of bands on the anti-gp41 immunoblot. The expression levels were normalized to that of the wild-type protein. Statistical analysis was performed using one way ANOVA. ****p < 0.0001, ***p < 0.001, **p < 0.01, *p < 0.05; ns indicates no statistically significant difference when compared with the WT protein

Fig. 3

Fusion assay of alanine insertion mutants of gp41. a and b The DSP assay was performed at 2 h after coculture to examine the fusion activities of mutants. Surface expression levels of mutant Envs were determined by CELISA, and the fusion activity measured by the DSP assay (open bar) was normalized to the surface expression level (dark bar). Error bars represent the standard deviation of the results of triplicate experiments. The results for NHR and loop mutants are shown in a. b The results for CHR mutants. Statistical analysis was performed using t test. **p < 0.01 and *p < 0.05 when comparison of the indicated mutant with wild type was done. ns indicates no significant differences (p > 0.05) when 647+A and 648+A were compared. c Syncytia formation assays were used to evaluate the fusion activities of the mutants. The number of nuclei in the syncytia was divided by the total number of the nuclei in the field. Five randomly chosen fields were used to measure the value. The degree of syncytia formation is shown as a percentage (the wild-type was set as 100%). Statistical analysis was performed using t test. ns indicates no significant differences (p > 0.05) when compared with the value of the wild type

Fig. 4

Pseudotype virus infection assay. a Infectivity of pseudotype virus bearing mutant Env was evaluated by measuring the reporter (n-luc) activity after 48 h of infection. The CHR mutants, 644+A, 645+A, 646+A, 647+A, 648+A, 649+A, were tested together with the wild type. Error bars represent the standard deviation of the results of triplicate experiments. Statistical analysis was performed using t test. **p < 0.01 when compared with 647+A. b The protein profiles of VLPs of CHR mutants were analyzed by immunoblotting. Anti-gp120, gp41 and p24 antibodies were used to detect the respective bands. The Env expression vector was omitted for mock transfection. The ratio of the band intensities of gp41 and p24 was indicated under the immunoblot

Fig. 5

Schematic representation of the virus and cell fusion. a Prefusion structure of gp41. Side (left panel) and bottom (right upper panel) views of gp41 are shown. The four-helix collar (α6–α9) of gp41 encircled gp120 (gp120 is indicated as a dotted circle). Close-up view of gp41 around residue 647 is shown in the right lower panel. The distances between residues 647 and residues 592 (4.786 Å), 595 (3.199 Å), and 596 (4.829 Å) are indicated. b Schematic description of conformational changes in gp41 during membrane fusion based on our data. After the fusion peptide (FP) reached the host-cell membrane, packing of CHR into the NHR grooves was initiated near the connecting loop region between the CHR and NHR. Pairing of residues in the CHR located before residue 647 may be sufficient for generation of initial unstable fusion pores. Residues beyond 647 of the CHR needed to interact with the NHR to achieve pore enlargement. The cytoplasmic (CT) domain, transmembrane (TM) domains, fusion peptide (FP), membrane proximal external region (MPER), NHR, CHR, and loop region are indicated