Initial antibodies binding to HIV-1 gp41 in acutely infected subjects are polyreactive and highly mutated

Abstract

The initial antibody response to HIV-1 is targeted to envelope (Env) gp41, and is nonneutralizing and ineffective in controlling viremia. To understand the origins and characteristics of gp41-binding antibodies produced shortly after HIV-1 transmission, we isolated and studied gp41-reactive plasma cells from subjects acutely infected with HIV-1. The frequencies of somatic mutations were relatively high in these gp41-reactive antibodies. Reverted unmutated ancestors of gp41-reactive antibodies derived from subjects acutely infected with HIV-1 frequently did not react with autologous HIV-1 Env; however, these antibodies were polyreactive and frequently bound to host or bacterial antigens. In one large clonal lineage of gp41-reactive antibodies, reactivity to HIV-1 Env was acquired only after somatic mutations. Polyreactive gp41-binding antibodies were also isolated from uninfected individuals. These data suggest that the majority of gp41-binding antibodies produced after acute HIV-1 infection are cross-reactive responses generated by stimulating memory B cells that have previously been activated by non-HIV-1 antigens.

Figure 1.

Plasma cell response in AHI. (A) Sorting of plasma cells from AHI subjects (681–7, 684–6, 001–4, 0689 and 065–0). Dot plots were gated on CD3⁻CD14⁻CD16⁻CD235a⁻CD19⁺ B cells (cells were also gated as CD20^lo/− during sorting). Ovals designate the populations that were sorted as plasma cells (CD27^hi/+ and CD38^hi/+). Numbers indicate percentage of cells in the sort gate. LKP, leukapheresis cells. (B) Specificity of antibodies isolated from plasma cells from each subject, or from all five AHI subjects pooled together. Total numbers of antibodies were indicated at the center of each pie chart. Percentage of antibodies binding to gp41, gp120, AT-2 inactivated HIV-1 clade B virion, and non–HIV-1/nondefined antigens are indicated in different colors. (C and D) Frequency of somatic mutations in V_H gene segments of antibodies isolated from all five AHI subjects or from individuals 14 d after the second immunization (42 d after first dose) or 14 d after the fourth immunization (182 d after first dose) with gp120/NefTat/AS01B (D).

Figure 2.

Phylogenetic trees of antibody clonal lineages derived from AHI. Clonal lineages 558 (A), 572 (B), 712 (C), 728 (D), and 2495 (E) were obtained from AHI subject 684–8. Clonal lineage 2495 was obtained from AHI subject 001–4. Antibodies highlighted in red for each clonal lineage were gp41 reactive and antibodies labeled in black font were gp41-negative in the initial screening. The length of lines in scale indicates extent of nucleotide differences between individual antibodies. Numbers underneath the scale lines indicate frequency of mutation; 0.01 = 1% sites mutated or 0.1 = 10% sites mutated. Inferred unmutated ancestors (RUAs) are indicated by red triangles in each clonal lineage. For clonal lineage 558 (A), numbers under or beside lines denote numbers of nucleotides of V_H and V_K chain genes of individual antibodies that differ from each other, and numbers in parentheses indicate the number of antibodies that had identical sequences. Inferred reverted intermediate antibodies are indicated by red square boxes for clonal lineage 558.

Figure 3.

Binding of RUA antibodies and observed antibodies to 684–6 transmitted/founder gp140 and to HIV-1 MN rgp41 was analyzed by SPR. RUAs and observed antibodies were derived from HIV-1 gp41-reactive clonal lineages 558, 572, 712, and 728 as indicated at the top of each graph. Approximately 1,000–3,000 Response Units of 684–6 gp140 (A) or HIV-1 MN rgp41 (B) were immobilized on adjacent flow cells of the same sensor chip. Each antibody was injected in a dose range starting at 100 µg/ml for 3 min at 30 µl/min on a BIAcore 3000 instrument (GE Healthcare) using PBS, pH 7.4, as running buffer. For binding K_d measurements, each antibody that bound to 684–6 gp140 (A) or to HIV-1 MN rgp41 (B) was injected at concentrations ranging from 80 to 1µg/ml and at a flow rate of 50 µl/min for 3min. Data are representative of three separate experiments.

Figure 4.

Characteristics of anti–HIV-1 gp41 Env clonal lineages. (A) Analysis of the reactivity of inferred RUA, intermediates, and observed antibodies in clonal lineage 558 with HIV-1 rgp41, autologous Env gp140, and CON-S Env gp140. The RUA (triangle) and reverted intermediates 1, 6, and 9 (square boxes), as well as observed antibody H828-K616, were assayed for reactivity with rgp41, autologous 684–6 gp140, and the group M consensus gp140 CON-S by Luminex assays. (B) V_H sequences of clonal lineages identified by 454 deep sequencing. Shown are the numbers of clonal sequences (on the y axis) with the percentage of nucleotide change from germline sequences (on the x axis) for each of HIV-1 Env-reactive clonal lineage m (indicated by clonal lineage number with number of sequences found in parentheses) identified by 454 deep sequencing from genomic DNA of B cells of acute HIV-1 subjects. Sequences with identical V, D, and J segment usage and identical V-D and D-J junctions were considered to be members of an individual B cell clone. Data are representative of three separate experiments.

Figure 5.

Binding of the RUA, inferred reverted intermediate antibodies, and observed antibodies in clonal lineage 558 to HEp-2 epithelial cells in the indirect immunofluorescence staining assays. Numbers in square boxes are inferred intermediate antibodies. The observed antibody members are highlighted in red font. Positive or negative immunofluorescence staining of antibodies on HEp-2 epithelial cells is indicated above the pictures for the indicated antibodies. Data are representative of three separate experiments.

Figure 6.

Reactivity of antibodies isolated from uninfected subjects. Three IgG antibodies (Ab1144, Ab929, and Ab931) and one IgM antibody (C14) in threefold dilutions ranging from 50 to 0.2 µg/ml (x axis) were evaluated for reactivity with HIV-1 rgp41 by Luminex assays (A) and for reactivity with a panel of autoantigens by ANA assays (B). The dotted lines indicate the cutoff values ≥120 luminance units used to denote positivity. Binding affinity of Ab1144, Ab929, and Ab931 in dose ranges (starting concentration at 40 µg/ml) to HIV-1 MN gp41 was measure by SPR (C). Data are representative of at least two separate experiments.