The C108g epitope in the V2 domain of gp120 functions as a potent neutralization target when introduced into envelope proteins derived from human immunodeficiency virus type 1 primary isolates

Abstract

Monoclonal antibodies (MAbs) directed against epitopes in the V2 domain of human immunodeficiency virus type 1 gp120 often possess neutralizing activity, but these generally are highly type specific, neutralize only laboratory isolates, or have low potency. The most potent of these is C108g, directed against a type-specific epitope in HXB2 and BaL gp120s, which is glycan dependent and, in contrast to previous reports, dependent on intact disulfide bonds. This epitope was introduced into two primary Envs, derived from a neutralization-sensitive (SF162) and a neutralization-resistant (JR-FL) isolate, by substitution of two residues and, for SF162, addition of an N-linked glycosylation site. C108g effectively neutralized both variant Envs with considerably higher potency than standard MAbs against the V3 and CD4-binding domains and the broadly neutralizing MAbs 2G12 and 2F5. These amino acid substitutions also introduced the epitope recognized by a second V2-specific MAb, 10/76b, but this MAb possessed potent neutralizing activity only in the absence of the glycan required for C108g reactivity. In contrast to other gp120-specific neutralizing MAbs, C108g did not block binding of soluble Env proteins to either the CD4 or the CCR5 receptor, but studies with a fusion-arrested Env indicated that C108g neutralized at a step preceding the one blocked by the gp41-specific MAb, 2F5. These results indicate that the V1/V2 domain possesses targets that mediate potent neutralization of primary viral isolates via a novel mechanism and suggest that inclusion of carbohydrate determinants into these epitopes may help overcome the indirect masking effects that limit the neutralizing potency of antibodies commonly produced after infection.

FIG. 1.

Effects of reduction and deglycosylation on stability of epitopes in the V2 and other domains. Recombinant BaL gp120 was captured on wells of a 96-well ELISA plate either untreated (open squares) or after deglycosylation with peptide N-glycosidase F (PNGase F; closed diamonds) or reduction by DTT (closed triangles). The reactivity of the indicated MAbs to control or treated rgp120 was titrated for 1 h at 37°C and bound antibody detected with alkaline phosphatase-conjugated anti-IgG, followed by incubation with alkaline phosphatase substrate. OD405, optical density at 405 nm.

FIG. 2.

Sequences of the V2 regions of the Env proteins used in this study. The region in HXB2 V2 containing glycosylation site 160 and the adjoining C108g-reactive peptide sequence is boxed, and N-linked glycosylation motifs are underlined. The mutated residues in this region for JR-FL and SF162 are in italics. The SF16Z(NI+GKV) variant contained both the NI and the G-V mutations.

FIG. 3.

(A) Effects of mutations in the JR-FL V2 domain on neutralization by V2-specific MAbs C108g and 10/76b and MAbs to the V3 region (4117C) and CD4-binding domain (5145A). Neutralization assays were performed with env deleted, luc-expressing NL4-3 provirus pseudotypes with either the parental JR-FL Env (closed triangles) or the JR(GKV) variant expressing the C108g epitope in the JR-FL background (open triangles). (B) Effects of mutations in the SF162 V2 domain on neutralization by V2-specific MAbs C108g and 10/76b and MAbs to the V3 region and CD4-binding domains. Neutralization was determined as in panel A for luc-expressing viral pseudotypes with parental SF162 Env (closed squares) or with the SF162 variants containing the glycosylation site at position 160 [SF(NI), open diamonds], the changes at positions 167 to 169 [SF(GKV), open triangles], or the combination of these changes [SF(NI+GKV), closed circles] that expressed the C108g epitope. WT, wild type.

FIG. 4.

Analysis of neutralizing activities of MAbs against various domains for the viruses pseudotyped with the two parental Envs SF162 (closed circles) and JR-FL (closed triangles) and with the two C108g-expressing variants SF(NI+GKV) (open circles) and JR(GKV) (open triangles). The MAb used and the domain it recognizes are indicated at the top of each panel.

FIG. 5.

Fluorescence-activated cell sorter assays of the inhibition of binding of either gp120 to U87 cells expressing CD4 (panel A) or of FLSC to cells expressing CCR5 (panel B) by MAbs or sCD4. Env proteins were used at 1.0 μg/ml, MAbs at 20 μg/ml, and sCD4 at 100 μg/ml. Binding of Env proteins to cells was detected by subsequent staining with biotinylated HIVIG, followed by R-phycoerythrin-conjugated streptavidin and quantitated on a Becton-Dickinson FACScalibur flow cytometer. The MAb used and the extent of inhibition compared to the control in the absence of antibody, measured as the decrease in mean fluorescence intensity compared to the control, are indicated in each panel.

FIG. 6.

Pre- and postbinding neutralization assays of SOS-shackled pseudovirions. Virus pseudotyped with the JR(GKV).SOS Env was either preincubated with MAbs and then added to cells (prebinding neutralization conditions) or allowed to bind to U87/CD4/CCR5 target cells for 2 h and then treated with the indicated antibodies for 2 h at 37°C (postbinding conditions). The virus and antibodies were then removed and the bound virus activated for fusion by brief exposure to DTT. After washing, the cells were incubated for an additional 2 days and assayed for luciferase levels. Antibodies were used at the following concentrations: 4117C, 19 μg/ml; C108g, 5 μg/ml; 2F5, 12 μg/ml; b12, 3.5 μg/ml.