Breaching peripheral tolerance promotes the production of HIV-1-neutralizing antibodies

Abstract

A subset of characterized HIV-1 broadly neutralizing antibodies (bnAbs) are polyreactive with additional specificities for self-antigens and it has been proposed immunological tolerance may present a barrier to their participation in protective humoral immunity. We address this hypothesis by immunizing autoimmune-prone mice with HIV-1 Envelope (Env) and characterizing the primary antibody response for HIV-1 neutralization. We find autoimmune mice generate neutralizing antibody responses to tier 2 HIV-1 strains with alum treatment alone in the absence of Env. Importantly, experimentally breaching immunological tolerance in wild-type mice also leads to the production of tier 2 HIV-1-neutralizing antibodies, which increase in breadth and potency following Env immunization. In both genetically prone and experimentally induced mouse models of autoimmunity, increased serum levels of IgM anti-histone H2A autoantibodies significantly correlated with tier 2 HIV-1 neutralization, and anti-H2A antibody clones were found to neutralize HIV-1. These data demonstrate that breaching peripheral tolerance permits a cross-reactive HIV-1 autoantibody response able to neutralize HIV-1.

Figure 1.

B6.Sle123 and B6 mice mount a similar Env-specific IgG antibody response. B6.Sle123 (n = 23 Env; n = 11 Alum) and B6 (n = 24 Env; n = 13 Alum) mice were immunized with alum alone (gray symbols) or with Env + alum (black symbols) and Env-specific Ig titers measured by ELISA. (A) gp120-reactive IgM is shown for B6 (squares) and B6.Sle123 (circles) mice. (B) Total IgM serum concentrations are shown for B6 (squares) and B6.Sle123 (circles) naive (open) mice and 14 d after alum alone (gray) or Env + alum (black). (C) gp120-specific IgG is shown for B6 (squares) and B6.Sle123 (circles) mice 14 d after immunization with alum alone (gray symbols) or with Env + alum (black symbols). (D) Total IgG serum concentrations are shown for B6 (squares) and B6.Sle123 (circles) naive (open) mice and 14 d after alum alone (gray) or Env + alum (black). All P-values were calculated using Student’s t test assuming unequal variance. Each symbol represents one mouse and all data are plotted as the arithmetic mean ± SEM. *, P < 0.05; **, P < 0.001; ***, P < 0.0001; ns, not significant. Data shown is from five independent experiments.

Figure 2.

B6.Sle123 mice neutralize tier 1 and 2 HIV-1 strains after alum treatment alone. HIV-1 neutralization by (A) B6.Sle123 or (B) B6 mice 14 d after immunization with Env + alum or by B6.Sle123 mice after (C) alum treatment alone or (D) no treatment (naive). These are same mice as in Fig. 1 and data are expressed as ID₅₀, calculated as the reciprocal dilution of sera required for 50% neutralization. The number of alum and Env + alum–treated B6.Sle123 mice able to neutralize ≥1 virus (neutralizers) is statistically higher compared with similarly treated B6 mice (Env, P = 0.0007; alum, P = 0.023). Naive B6.Sle123 mice able to neutralize ≥1 virus was statistically lower than treated mice (Env, P = 0.017; alum, P = 0.028). Statistics calculated for A–D using Fisher’s exact test. (E) Mean ID₅₀ for neutralization of indicated tier 1 and 2 virus by B6.Sle123 treated with alum alone (open) or Env + alum (filled). Data are shown as the geometric mean ± SEM. (F) Percent of all B6.Sle123 mice treated with alum (left) or Env + alum (right) able to neutralize 0, 1, 2, 3, or 4 HIV-1 strains. Data are from five independent experiments.

Figure 3.

Neutralization of HIV-1 is observed in MRL/lpr mice. MRL/lpr mice (8–16 wk old) were immunized with alum (n = 5) or Env + alum (n = 11), sacrificed day 14 pi, and analyzed for neutralization. The ID₅₀ is calculated as in Fig. 2. (A) The mean ID₅₀ was calculated for the responders per virus for MRL/lpr Env immunized (filled bars) and alum injected (open bars) mice. Data are plotted as the geometric mean ± SEM. (B) The percentage of responding MRL/lpr alum (left, 60%) and Env (right, 82%) immunized mice was similar (P = 0.55). Statistics were calculated using Fisher’s exact test. Data are shown from two independent experiments.

Figure 4.

HIV-1 neutralization correlates with total IgM levels. B6.Sle123 HIV-1 neutralizers (black circles) and nonneutralizers (gray circles) from Fig. 2 were evaluated separately for (A) age at time of sacrifice and serum levels of (B) total IgM (P = 0.025) and (C) total IgG. Serum autoantibodies were also measured in B6.Sle123 neutralizers and nonneutralizers for IgM (top) and IgG (bottom) reactive with (D) chromatin, (E) Smith antigen, and (F) cardiolipin. Neutralizer is defined as mice neutralizing ≥1 strain of HIV-1. Each symbol represents one mouse and data are shown as the arithmetic mean ± SEM. All P-values were calculated using Student’s t test assuming unequal variance. *, P < 0.05; **, P < 0.001; ***, P < 0.0001. Only total IgM levels are significantly different between B6.Sle123 neutralizers and nonneutralizers (P = 0.025).

Figure 5.

B6.Sle123 HIV-1 neutralizers harbor elevated levels of IgM anti-histone H2A. (A) Sera from B6 (open), B6.Sle123 nonneutralizers (gray), and neutralizers (black) were interrogated with an autoantigen array and results for the IgM reactive with the indicated anti-DNA antigens (top left), RNA-binding proteins (bottom left), and anti-histone antigens (top right), as well as IgG anti-histone antigens (bottom right) are shown. Relative serum titers of (B) IgM anti-H2A (left) and anti-H2B (right) measured by ELISA in B6.Sle123 neutralizers and nonneutralizers. Each symbol represents measurements for one mouse, and all data are plotted as the arithmetic mean ± SEM. P-values were calculated using the Mann-Whitney nonparametric test; *, P < 0.05; **, P < 0.001. Histone H2A IgM titers are significantly higher in B6.Sle123 neutralizers versus nonneutralizers (P = 0.005).

Figure 6.

Wild-type mice neutralize tier 2 HIV-1 after an experimental breach in immunological tolerance. (A) Schematic of experimental protocol. Serum neutralization of tier 1 and 2 HIV-1 strains by B6 mice treated with (B) pristane alone for 30 d or after (C) 2 (2X) or (D) 3 (3X) subsequent immunizations with alum alone or Env + alum as indicated. Data are from the same mice serially bled. NS, insufficient sera available for analysis. (E) Percentage of mice able to neutralize tier 1 and 2 HIV-1 viruses after pristane only (open bars) or after 2X Env + alum immunizations (black bars). (F) Mean ID₅₀ for the mice that neutralized tier 1 and 2 HIV-1 viruses after pristane only (open bars) or after 2X Env + alum immunizations (black bars). Data are plotted as the geometric mean ± SEM. (G) Mean ID₅₀ are shown for B6 mice able to neutralize tier 2 HIV-1 strains after pristane alone (open bars), after two immunizations with alum alone (gray bars) or after 2 immunizations with Env + alum (black bars). (H) Percentage of mice able to neutralize tier 1 and 2 HIV-1 viruses after 3 Env + alum immunizations. (I) Mean ID₅₀ for the mice that neutralized tier 1 and 2 HIV-1 viruses after 3X Env + alum immunizations. (J) gp140 YU2-specific ELISAs for IgM and IgG between 2X alum (open circles) and 2X Env + alum (black circles) immunized mice. Data are representative of two independent experiments. P-values were calculated by Student’s t test assuming nonequal variance; *, P < 0.05; **, P < 0.001.

Figure 7.

Elevated IgM anti-histone H2A titers correlate with tier 2 HIV-1 neutralization by pristane treated wild-type C57BL/6 mice. Total serum concentrations of (A) IgM, (B) IgG, and (C) relative titers of serum IgM anti-H2A are shown for naive (open circles) B6 mice or B6 mice treated 30 d with pristane only (gray circles) and subsequently immunized 2X or 3X (black circles) with alum alone or Env + alum. Serum from individual B6 mice (regardless of treatment with pristane alone, alum alone, or Env + alum) were separated based on neutralization of ≥1 tier 2 HIV-1 strains and measured for (D) IgM anti-H2A or (E) IgM anti-H2B relative titers. Mice neutralizing only tier 1 strains were included in the tier 2 nonneutralizer group (mostly 3X). Each symbol represents measurements for one mouse, and all data are plotted as the arithmetic mean ± SEM. All P-values were calculated using Student’s t test assuming unequal variances. *, P < 0.05; **, P < 0.001; ***, P < 0.0001; ns, not significant. Data are representative of two independent experiments.

Figure 8.

Histone H2A-reactive IgM monoclonal antibodies, isolated from B6.Sle123 mice, neutralize tier 2 strains of HIV-1. Hybridomas were generated from splenocytes isolated from B6.Sle123 mice that displayed serum tier 2 HIV-1 neutralization (n = 2). (A) Purified monoclonal IgM and IgG antibodies (n = 8) were tested for HIV-1 neutralization against 4 strains of HIV-1 and reported as IC₅₀ (the concentration of antibody required for 50% neutralization). (B) Specificities of the monoclonal antibodies were tested using ELISA against the histone H2A, histone H2B, and chromatin nuclear antigens in addition to HIV-1 gp140 (YU2) Env and the CD4bs (RSC3) epitope, as well as a CD4bs-negative control (ΔRSC3). The two neutralizing IgM mAbs (P4E4 and O4C5) are shown with red symbols, and lines and non-neutralizing mAbs shown with solid black symbols and lines for IgM and gray symbols and dashed lines for IgG. IgM and IgG control antibodies were used to determine the background of the assay and for which ODs above this line were considered positive. Data are representative from three independent experiments.