Residues in the membrane-spanning domain core modulate conformation and fusogenicity of the HIV-1 envelope glycoprotein

Abstract

The membrane-spanning domain (MSD) of human immunodeficiency virus type I (HIV-1) envelope glycoprotein (Env) is critical for its biological activity. Initial studies have defined an almost invariant "core" structure in the MSD and demonstrated that it is crucial for anchoring Env in the membrane and virus entry. We show here that amino acid substitutions in the MSD "core" do not influence specific virus-cell attachment, nor CD4 receptor and CXCR4 coreceptor recognition by Env. However, substitutions within the MSD "core" delayed the kinetics and reduced the efficiency of cell-cell fusion mediated by Env. Although we observed no evidence that membrane fusion mediated by the MSD core mutants was arrested at a hemifusion stage, impaired Env fusogenicity was correlated with minor conformational changes in the V2, C1, and C5 regions in gp120 and the immunodominant loop in gp41. These changes could delay initiation of the conformational changes required in the fusion process.

FIG. 1. Kinetics of cell-cell fusion mediated by the Env

(A) A schematic structure of the HIV-1 Env sequence (FP: fusion peptide; HR: heptad repeat; Loop: immunodominant loop; MSD: membrane-spanning domain; LLP: Lentivirus lytic peptide). (B) Amino acid sequences of HIV-1 MSDs. The consensus sequence of the HIV-1 MSD was generated by the alignment of the Env MSD sequences of all M and N group HIV-1 isolates from the Los Alamos HIV Sequences Database. The amino acid residues of the putative MSD region are shown as upper case letters. The flanking sequences of the HIV-1 MSD are shown as lower case letters. The position of the MSD in the sequence of HIV-1 NL4-3 is marked above the positively charged amino acid residues. The mutated portions of the HIV-1 MSD are underlined. (C) Virus-cell attachment assay. Purified viruses were co-incubated with JC53BL cells at specific temperature for 2 h. Cells were then washed three times in PBS and lysed in PBS containing 0.5% Triton X-100. The amount of p24 in cell lysate was measured by p24 ELISA. The error bars represent the standard deviations of three independent experiments. (D) Cell-cell fusion kinetics. 293T cells were transfected with the proviral vector, pTN6-GFP, containing the MSD mutants. At 48 h posttransfection, the transfected 293T cells were cocultured with the CMAC-blue-loaded JC53BL indicator cells for specific times. Fused cells were observed and quantitated by fluorescent microscopy. For each mutant, at least 50 fusion events were counted at each time point.

FIG. 2. Examination of hemifusion intermediates

(A) Cell-cell fusion mediated by Env. HeLa cells, which were transfected with proviral DNAs, were cocultured with HeLa P4 indicator cells for 12 h, and then mixed cells were treated with 0.1mM CPZ for 30 seconds. After extensive wash, cells were incubated in complete DMEM for 10 h and then were subjected to a β-Gal staining. The cell-cell fusogenicity of Env mutants were measured by counting the number of cells involved into syncytia (blue cells) per well (at least two different wells were counted). (B) Flow cytometry analysis of cell-cell fusion. The proviral DNA-transfected HeLa cells were loaded with either membrane dye DiO or cytoplasmic dye Cell Tracker Green, and then were cocultured with DiD-loaded or Cell Tracker Blue-loaded HeLa P4 indicator cells, respectively, for 6 h. The cell mixtures then were subjected to flow cytometry analysis. The double-positive cell populations are shown as percentage to WT. The error bars represent standard deviations of three independent experiments. (“*” – statistically significant) (C) Three color microscopy of cell-cell fusion mediated by the Env. HeLa cells, which were transfected with proviral DNAs and loaded with CMFDA-Green, were cocultured with HeLa P4 indicator cells, which were double stained with DiI (red) and CMAC-blue, for 3 h. Fused cells were observed by fluorescent microscopy.

FIG. 3. Sensitivity of MSD core mutants to the levels of CD4 and CXCR4 on cell surface

The Env mutants were cloned into a proviral vector, pTN7, in which the nef coding sequence was replaced by a renilla luciferase reporter gene. Viruses produced by transfecting 293T cells with pTN7 constructs were used to infect HI- and RC- cell lines (Kabat). Luciferase activities were measured 48 h postinfection. (A) Cell lines HI-J and RC-25 have the same surface level of CXCR4, but the CD4 level of HI-J is 15-fold higher than that of RC-25. (B) The cell surface levels of CD4 are similar between HI-R and RC-49 cells, while the surface level of CXCR4 on RC-49 is 6-fold lower than that of HI-R cells. The standard deviations of three independent experiments are shown in the error bars.

FIG. 4. gp120 virus capture assay

Purified viruses were incubated at 37°C in 96-well plates which have been pre-coated with anti-gp120 monoclonal antibodies. The coating antibodies target to different epitopes of the gp120 glycoprotein: b12 (A) binds to the CD4-binding site, 697-30D (B) to the V2 region, F425 B4a1 and 902 (C) to the V3 loop, ID6 and Chessie6 (D) to the C1 region, Chessie13–39.1 (E) to the C2 region, 670-30D (F) to the C5 region, F425 A1g8 (G), 17b (H) and E51 (I) to the coreceptor-binding region, and A32 (J) to the C1/C4 region. The bound viruses were lysed in PBS containing 0.5% Triton X-100 and then subjected to a p24 ELISA assay. The experiment was carried out in triplicates for each antibody and the standard deviation is shown in the error bar.

FIG. 5. gp41 virus capture assay

Purified viruses were incubated at 37°C in 96-well plates which were pre-coated with anti-g41 monoclonal antibodies. The coating antibodies target to different epitopes of the gp41 glycoprotein: NC-1 (A) to the 6-helix bundles, T32 (B) to the immunodominant loop of gp41 ectodomain, D50 (C) to the HR2 region, and 4E10 (D) to the MPER region. Bound viruses were lysed in PBS containing 0.5% Triton X-100 and then subjected to a p24 ELISA assay. The experiment was carried out in triplicates for each antibody and the standard deviation is shown in the error bar.

FIG. 6. C34 and T20 inhibition of cell-cell fusion mediated by the Env and viral infection

293T cells transfected with the Env expression vector, pCDNA3.1, were cocultured with JC53BL indicator cells. Cell mixtures were incubated for 24 h in the presence of fusion inhibitors, C34 (A) and T20 (B), at specific concentrations, and then were subjected to a luciferase activity assay. JC53BL indicator cells were infected with the p24-normalized viruses in DMEM containing 1% fetal bovine serum, 80µg/ml DEAE-Dextran and fusion inhibitors, C34 (C) and T20 (D), at specific concentrations. Complete DMEM was added following 2 h incubation and cells were analyzed for luciferase activity 48 h post infection. Error bars represent standard deviations of at least three independent experiments.