A strategy for eliciting antibodies against cryptic, conserved, conformationally dependent epitopes of HIV envelope glycoprotein

Abstract

Background: Novel strategies are needed for the elicitation of broadly neutralizing antibodies to the HIV envelope glycoprotein, gp120. Experimental evidence suggests that combinations of antibodies that are broadly neutralizing in vitro may protect against challenge with HIV in nonhuman primates, and a small number of these antibodies have been selected by repertoire sampling of B cells and by the fractionation of antiserum from some patients with prolonged disease. Yet no additional strategies for identifying conserved epitopes, eliciting antibodies to these epitopes, and determining whether these epitopes are accessible to antibodies have been successful to date. The defining of additional conserved, accessible epitopes against which one can elicit antibodies will increase the probability that some may be the targets of broadly neutralizing antibodies.

Methodology/principal findings: We postulate that additional cryptic epitopes of gp120 are present, against which neutralizing antibodies might be elicited even though these antibodies are not elicited by gp120, and that many of these epitopes may be accessible to antibodies should they be formed. We demonstrate a strategy for eliciting antibodies in mice against selected cryptic, conformationally dependent conserved epitopes of gp120 by immunizing with multiple identical copies of covalently linked peptides (MCPs). This has been achieved with MCPs representing 3 different domains of gp120. We show that some cryptic epitopes on gp120 are accessible to the elicited antibodies, and some epitopes in the CD4 binding region are not accessible. The antibodies bind to gp120 with relatively high affinity, and bind to oligomeric gp120 on the surface of infected cells.

Conclusions/significance: Immunization with MCPs comprised of selected peptides of HIV gp120 is able to elicit antibodies against conserved, conformationally dependent epitopes of gp120 that are not immunogenic when presented as gp120. Some of these cryptic epitopes are accessible to the elicited antibodies.

Figure 1. Binding of anti MCPs sera to gp120 is blocked by preincubation with the specific MCP immunogen.

(A) Antiserum to 426–441 MCP, (B) Antiserum to 363–384 MCP. Sera derived from 426–441 MCP immunized C57Bl/10 mice (diluted 7250 fold) or from 363–384 MCP immunized C3H/HeJ mice (diluted 140 fold) were incubated with MCPs. Binding of antibodies to gp120 was determined by ELISA.

Figure 2. MAbs 11A8 and 2D3 compete for binding to gp120.

Binding to gp120 of biotinylated 2D3 in the presence of unconjugated MAbs 2D3 or 11A8 was evaluated in ELISA. Concentration of biotinylated 2D3 in the assay was 1.2 nM, concentrations of competing, unconjugated MAbs 2D3 and 11A8 were as indicated.

Figure 3. Antibodies elicited by MCPs fail to recognize denatured gp120.

Recombinant gp120_MN or RCM gp120_MN were either maintained at 4° (control) or were heated for 5 minutes at 95° with 1 mM DTT, 0.1% SDS (denatured). Gp120 was then immobilized in ELISA wells using Ab D7324 and binding of antibodies was determined. Antigens captured on the plate were as follows: (A) gp120 (control); (B) gp120 (denatured). MAbs: 11A8 and 2D3, antisera to MCP 419 and to MCP 426 were all derived from C57Bl/10 immunized mice and antisera to gp120 was derived from C3H/HeJ mice. All sera and MAb preparations were diluted as indicated in the Figure. Protein concentration of the stock of undiluted MAb preparations used was, respectively: 1.96 mg/ml (MAb 11A8), and 0.5 mg/ml (MAb 2D3).

Figure 4. MCPs elicit antibodies to cryptic epitopes of gp120.

(A) Binding of MAb 11A8 to gp120 is not inhibited by antisera to gp120. Competition for gp120_MN binding of biotinylated MAb 11A8 and murine sera was evaluated in ELISA as described in Materials and Methods. All competing sera were obtained from C57Bl/10 mice prior to immunization (preimmune), from mice immunized with 426–441 MCP (130, 000) or with rgp120MN (550, 000). Titers of sera antibodies recognizing gp120 are listed in parenthesis. Results are expressed as % inhibition of biotinylated 11A8 binding to gp120 by competing antibodies. (B) HIVIG does not inhibit the binding to gp120 of murine antibodies elicited by MCPs 419–439, 363–384 or 105–117. Competition between HIVIG and anti-MCP sera for binding to gp120 was assayed by ELISA as described in Materials and Methods. Gp120 was incubated with HIVIG at the indicated concentrations and the binding of anti-MCP sera (colored lines) from C3H/HeJ mice (anti-MCPs 105–117 and 363–384), C57Bl/10 mice (419–439 MCP) or of HIVIG (shown as black line) was determined. Results are expressed as absorbance at 405 nm.

Figure 5. Determination of antibodies in 426–441 MCP immune sera that can compete with 11A8 for gp120 binding.

Competition of antibodies and biotinylated 11A8 (0.02 µg/ml) for binding to gp120_MN was determined in ELISA as described in Materials and Methods. Source of competing antibodies: (A) Purified, unconjugated 11A8 - used as a standard; (B) Sera harvested from mice immunized with 426–441 MCP. MAb 11A8 and 426–441 MCP sera were derived from C57Bl/10 mice. Results are expressed as percent inhibition of biotinylated 11A8 binding by competing antibodies.

Figure 6. A MAb elicited by 426 MCP (11A8) recognizes HIV-1 on the surface of infected cells.

(A) Uninfected CEM-4 cells, (B) CEM-4 cells infected with HIV-1 clade B. Cells were incubated, as indicated, with 5 µg of 11A8 or of anti-TNP (isotypic control), or with no primary antibody (reagent control). Antibody binding was detected using PE-conjugated secondary antibody and measured by flow cytometry. Flow histogram plots are displayed.

Figure 7. 11A8 binding to HIV on infected cells is blocked by gp120 and by homologous MCP.

11A8 was incubated for 30 min at 37° with 426–441 MCP, 363–384 MCP or with gp120 in 2% FCS RPMI and subsequently the binding of 11A8 to HIV-1 clade B infected CEM-4 cells was assayed. Final concentration of 11A8 in the binding assay was 0.5 µM, and concentrations of competing antigens were as indicated. MAb 11A8 binding to cells was determined by flow cytometry and % inhibition of MAb binding by antigens was calculated.

Figure 8. 11A8 binds to cells infected by a clade D HIV isolate.

CEM-4 cells infected with (A) HIV-1 clade B or (B) with clade D isolate, were incubated with MAbs 11A8, b12 or with HIVIG at 10 µg/ml. Antibody binding to cells was detected using the appropriate secondary antibodies. The flow histogram plots of the fluorescence intensity of antibody binding to cells are displayed. (C) Antibody binding to clade D infected cells: concentration dependence. Clade D infected cells were incubated with MAbs 11A8, MOPC (murine isotypic control), b12, F105 or with HIVIG at the indicated concentrations. Percentage of cells binding antibody to the cell surface was determined by flow cytometry.

Figure 9. Antibodies elicited by MCPs 419 and 426 partially compete for CD4 binding to gp120.

(A) CD4 inhibition of antibodies binding to gp120. (B) MAb 11A8 weakly inhibits CD4 binding to gp120. Reciprocal competition of CD4 and antibodies for binding to gp120 was evaluated by ELISA as described in Materials and Methods with gp120 captured on the antibody to the C-terminal sequence of gp120.

Figure 10. MAb 11A8 neutralizes SF162 pseudovirus.

(A) Concentration curve of 11A8 and the murine isotypic control anti-TNP for SF162 pseudovirus neutralization. (B) Neutralization of SF162 by 11A8 is abolished by preincubation with the homologous MCP. MAb 11A8 was incubated before neutralization assay for 1 h at 37° with 426–441 MCP or 105–117 MCP or with medium (control). The final concentration of 11A8 in neutralization assay was 0.1 µM and of the MCPs was 1 µM.