Removal of a single N-linked glycan in human immunodeficiency virus type 1 gp120 results in an enhanced ability to induce neutralizing antibody responses

Abstract

Glycans on human immunodeficiency virus (HIV) envelope protein play an important role in infection and evasion from host immune responses. To examine the role of specific glycans, we introduced single or multiple mutations into potential N-linked glycosylation sites in hypervariable regions (V1 to V3) of the env gene of HIV type 1 (HIV-1) 89.6. Three mutants tested showed enhanced sensitivity to soluble CD4. Mutant N7 (N197Q) in the carboxy-terminal stem of the V2 loop showed the most pronounced increase in sensitivity to broadly neutralizing antibodies (NtAbs), including those targeting the CD4-binding site (IgG1b12) and the V3 loop (447-52D). This mutant is also sensitive to CD4-induced NtAb 17b in the absence of CD4. Unlike the wild-type (WT) Env, mutant N7 mediates CD4-independent infection in U87-CXCR4 cells. To study the immunogenicity of mutant Env, we immunized pig-tailed macaques with recombinant vaccinia viruses, one expressing SIVmac239 Gag-Pol and the other expressing HIV-1 89.6 Env gp160 in WT or mutant forms. Animals were boosted 14 to 16 months later with simian immunodeficiency virus gag DNA and the cognate gp140 protein before intrarectal challenge with SHIV89.6P-MN. Day-of-challenge sera from animals immunized with mutant N7 Env had significantly higher and broader neutralizing activities than sera from WT Env-immunized animals. Neutralizing activity was observed against SHIV89.6, SHIV89.6P-MN, HIV-1 SF162, and a panel of subtype B primary isolates. Compared to control animals, immunized animals showed significant reduction of plasma viral load and increased survival after challenge, which correlated with prechallenge NtAb titers. These results indicate the potential advantages for glycan modification in vaccine design, although the role of specific glycans requires further examination.

FIG. 1.

N-linked glycan mutants in HIV-1 Env. Potential N-linked glycosylation sites in HIV-1 89.6 surface antigen gp120 are indicated by “⋆” symbols (complex type) and “×” symbols (high-mannose type). The positions of mutants N1-N7 and NV3 relative to variable loops V1 to V5 are indicated at the top. Amino acid sequence of V2 and V3 loops and the specific sites changed in mutants N6, N7, and NV3 are indicated at the bottom.

FIG. 2.

Neutralization sensitivity of WT HIV-1 89.6 and N-linked glycan mutants. Neutralization of pseudotyped virus generated with WT or mutant HIV-1 89.6 Env was performed as described in Materials and Methods. Neutralization was quantified as the percent inhibition of viral infectivity measured by RLU expressed in TZM-bl indicator cells. The data shown are representative results obtained from three individual experiments.

FIG. 3.

(A) Infectivity of replication competent WT and mutant SHIV 89.6 in TZM-bl cells. The amount of input virus was normalized by the p27 level in the inoculum and was determined to be in the linear range of infectivity as measured by luciferase gene expression. Values on the y axis represent RLU (10⁴). (B) CD4-independent infection by glycan mutants. The infectivity of HIV-1 pseudotyped with WT or mutant Env was measured in U87 cells expressing CXCR4. The inoculum used for each mutant virus was normalized by the infectivity in U87-CXCR4-CD4 cells measured at 200,000 RLU. Values on the y axis in this panel represent RLU (10³). The results in both panels represent the average and standard deviation obtained from three individual experiments.

FIG. 4.

Immunization and challenge scheme. See Materials and Methods for details.

FIG. 5.

HIV-specific antibody responses after recombinant vaccinia and protein immunizations. HIV-1 SF162 gp120 (>95% pure, produced in stably transformed Chinese hamster ovary cells by Chiron Corp. and obtained from the NIH ARRRP, Division of AIDS, National Institute of Allergy and Infectious Disease, NIH) was used as the capturing antigen. Geometric mean endpoint titers of HIV-1 gp120-specific antibody response in immunized animals (n = 6/group) were measured before immunization (“Preimmune,” week 0), 2 weeks after the second recombinant vaccinia virus immunization (“After vaccinia,” week 10), on the day the animals received the last immunization (“Before last boost,: week 77 for groups 1 and 2 and week 66 for groups 3 and controls [see Materials and Methods]), and 2 weeks after the last immunization (“2 wk after last boost,” week 79 for groups 1 and 2 and week 68 for groups 3 and controls). Vertical bars indicate the standard deviations within each group.

FIG. 6.

NtAb response on the day of challenge. (A) Neutralization titers were measured as the highest serum dilutions that resulted in 50% reduction of infectivity in the TZM-bl cell assay against replication-competent viruses: homologous WT virus HIV-1 89.6, challenge virus SHIV89.6P-MN, or a heterologous virus HIV-1 SF162. The boxplots indicate the median, as well as the range of responses, including the maximum, minimum, and the 75 and 25% quartiles as indicated. The P values for the differences in response between group 1 or group 2 and the control group were ≤0.004 and ≤0.002, respectively (Mann-Whitney U test). The P value for the differences between groups 1 and 2 was ≤0.002 by the same test. (B) Neutralization of viruses pseudotyped with Env from 12 primary HIV-1 subtype B isolates by day-of-challenge serum from two animals immunized with the N7 mutant Env (group 2). Neutralization was expressed as the percent reduction of RLU at a 1:15 serum dilution with the values from preimmune sera subtracted. The data shown represent results obtained from two (B) or three (A) independent experiments.

FIG. 7.

Plasma viral load (left panels) and peripheral blood CD4⁺ T cells (right panels) in animals after SHIV89.6P-MN challenge. (A to H) Data for individual animals in each immunization or control group as indicated on the left margin. (I and J) Means and standard deviations of plasma viral load and peripheral blood CD4⁺ T cells, respectively, in control and experimental groups. The large “X” denotes the approximate date of euthanasia due to AIDS.

FIG. 8.

Correlate of protection and survival curve of challenged animals. The log₁₀-transformed 50% neutralization titer of each animal on the day of challenge against the challenge virus SHIV89.6P-MN was plotted against peak plasma viral load at week 2 (A), setpoint viral load at week 16 (B), or peripheral blood CD4⁺ T cells at week 28 (C). (D) Kaplan-Meier analysis of AIDS-free survival after challenge infection. Correlation coefficients (Spearman's r value) and P values (two-tailed) were calculated by using Prism 4 (GraphPad) software.