N-linked glycan modifications in gp120 of human immunodeficiency virus type 1 subtype C render partial sensitivity to 2G12 antibody neutralization

Abstract

The monoclonal antibody (MAb) 2G12 recognizes a cluster of high-mannose oligosaccharides on the human immunodeficiency virus type 1 (HIV-1) envelope glycoprotein gp120 and is one of a select group of MAbs with broad neutralizing activity. However, subtype C viruses are generally resistant to 2G12 neutralization. This has been attributed to the absence of a glycosylation site at position 295 in most subtype C gp120s, which instead is typically occupied by a Val residue. Here we show that N-linked glycans in addition to the one at position 295 are important in the formation of the 2G12 epitope in subtype C gp120. Introduction of the glycosylation site at position 295 into three subtype C molecular clones, Du151.2, COT9.6, and COT6.15, did increase 2G12 binding to all three mutagenized gp120s, but at various levels. The COT9-V295N mutant showed the strongest 2G12 binding and was the only mutant to become sensitive to 2G12 neutralization, although very high antibody concentrations were required. Introduction of a glycosylation site at position 448 into mutant COT6-V295N, which occurs naturally in COT9, resulted in a virus that was partially sensitive to 2G12. Interestingly, a glycosylation site at position 442, which is common among subtype C viruses, also contributed to the 2G12 epitope. The addition of this glycan increased virus neutralization sensitivity to 2G12, whereas its deletion conferred resistance. Collectively, our results indicate that the 2G12 binding site cannot readily be reconstituted on the envelopes of subtype C viruses, suggesting structural differences from other HIV subtypes in which the 2G12 epitope is naturally expressed.

FIG. 1.

Amino acid sequence alignment of C2-to-C5 region of gp120s of the three HIV-1 subtype C clones used in this study. Consensus subtype B and C sequences were aligned with the three subtype C envelope sequences and the HXB2 reference sequence. Potential N-glycan attachment sites, determined using N-Glycosite (http://hiv-web.lanl.gov/content/hiv-db/GLYCOSITE/glycosite.html), are highlighted in gray. The sites indicated with arrows, except for position 442, have been implicated in the formation of the 2G12 epitope. Those marked with asterisks were analyzed in this study. The N-linked glycosylation site at position 392 in Du151.2 is shifted by one amino acid. The env gene nucleotide sequences can be obtained from GenBank under accession numbers DQ447272 (Du151.2), DQ447266 (COT9.6), and DQ411851 (COT6.15).

FIG. 2.

Introduction of an N-glycan attachment site at position 295 in viruses Du151.2, COT9.6, and COT6.15 leads to an increase in molecular mass. The electrophoretic mobilities of virion-associated gp120s from the wild type (WT) and V295N mutants of Du151.2, COT9.6, and COT6.15 are shown on a 5% sodium dodecyl sulfate-polyacrylamide gel. Western blots were visualized using the antibody D7324, an anti-sheep Ab conjugate, and enhanced chemiluminescence reagents.

FIG. 3.

V295N mutation increases 2G12 antibody binding to subtype C gp120. The binding of MAbs 2G12 (10 μg/ml), IgG1b12 (10 μg/ml), and A32 (10 μg/ml) to the gp120s of Du151.2, COT9.6, and COT6.15, with and without a glycan at position 295, was measured by ELISA. The graphs represent the means for three separate experiments. OD, optical density.

FIG. 4.

Neutralization of wild-type and V295N mutant subtype C envelope-pseudotyped viruses by 2G12. Results are shown as reductions of virus infectivity relative to that of the virus control (without MAbs), with 50% inhibition indicated by a horizontal dotted line. MAb IgG1b12 and virus QH0692 (subtype B) were used as positive controls.

FIG. 5.

Impact of V295N and S448N mutations on 2G12 binding to monomeric and oligomeric gp120 from COT6.15. Antibody binding was evaluated by gp120 ELISA (A) and flow cytometric analysis (B) of the envelope protein expressed on the surfaces of transfected 293T cells. ELISA results are the averages for four independent experiments, and fluorescence-activated cell sorting results are from one experiment representative of three. OD, optical density.

FIG. 6.

N-glycosylation at position 442 affects 2G12 and IgG1b12 binding to monomeric gp120. The ability of MAbs 2G12 (10 μg/ml) and IgG1b12 (10 μg/ml) to bind gp120s from the glycosylation mutants of COT9.6 (A) and COT6.15 (B) was assessed by ELISA. The graphs represent the means for three separate experiments. A32 and IBU21 were used to standardize the amount of gp120 bound to the plate (data not shown). The subtype B virus QH0692 was used as a positive control for 2G12 and IgG1b12 binding. OD, optical density.

FIG. 7.

Location of the Asn at position 442 relative to the location of the V3 loop region and other N-glycans involved in 2G12 binding. The gp120 molecule is viewed with the outer domain facing forward. The structure is rendered as a ribbon diagram, with the V3 loop shown in yellow and the rest of gp120 shown in green. The Asn residues bearing the N-linked glycans that constitute the core of the 2G12 epitope on subtype B gp120 are highlighted in blue as space-filling models, while those supposedly involved in limiting glycosidase trimming are shown in red. The Asn at position 442, identified here as potentially important for the formation of the 2G12 epitope on subtype C gp120, is highlighted in orange. Coordinates were taken from the structure of the gp120_JRFL core with V3 ligated with CD4 and X5 (Protein Data Bank accession no. 2B4C). The figure was generated with PyMOL (DeLano Scientific LLC, South San Francisco, CA [http://www.pymol.org]).