Optimization and qualification of a memory B-cell ELISpot for the detection of vaccine-induced memory responses in HIV vaccine trials

Abstract

Various aspects of the human immune system can be analyzed to determine the efficacy of a vaccine. We have developed a B-cell ELISpot to measure HIV-specific antibody-secreting B cells in the peripheral blood as a result of vaccination or natural infection. Our method includes stimulating peripheral blood mononuclear cells with interleukin-2 and a polyclonal activator, R848, to induce memory B cells to differentiate into antibody-secreting cells. Total immunoglobulin-secreting as well as antigen-specific B cells are then quantified. We have tested several HIV Env gp120 and gp140 proteins from different HIV subtypes, as well as a sensitive consensus group M Env gp140. Our findings indicate that the B-cell ELISpot provides a sensitive and specific tool to detect antigen-specific memory B-cell responses, and it is equally suited to detect antibody-secreting plasmablasts present in the circulation shortly after infection or vaccination.

Keywords: ASC; Ag; Antibody-secreting cells; B-cell ELISpot; HBSS; HIV; Hanks balanced salt solution; Ig; KLH; Keyhole Limpet Hemocyanin; Memory B-cell responses; Mucosal responses; PBMC; PVDF; SFC; TMB; Vaccine; antibody-secreting cell; antigen; immunoglobulin; peripheral blood mononuclear cells; polyvinylidene fluoride; spot-forming cells; tetramethylbenzidine.

Figure 1. Titration of R848 concentration, R848/IL-2 and cell number per well in an HIV-infected subject

A) R848 was titrated in duplicate in twofold dilutions from 2μg/ml to 0.015 μg/ml to find the optimal concentration for B-cell stimulation using 5,000 cells/well and total IgG as a readout. B) R848 and IL-2 were tested in duplicate at 1μg/ml / 10ng/ml or 0.5μg/ml / 5ng/ml for total IgG, ConS and BaL gp140 using 5,000 cells/well for total IgG and 500,000 cells/well for ConS and BaL. C) The number of PBMC added per well for total IgG was titrated in duplicate in twofold dilutions from 10,000 to 312 cells/well to determine the optimal number that would give a readable result on the AID plate reader while still being able to detect low level responders.

Figure 2. R848/IL-2 provides superior stimulation than CpG-C in five HIV-infected subjects

A) The total number of CD19⁺ B cells prior to stimulation (open bars) as well as after 5-day stimulation with CpG-C (light grey bars) or R848 (dark grey bars) was assessed on a Guava Counter. B) R848/IL-2 leads to higher magnitude of HIV-specific memory B-cell responses (dark grey bars) than CpG-C (light grey bars) without increasing background (open bars) using 200,000 cells/well for each stimulation. The dotted grey line represents the positivity threshold of 20 SFC/million. C) Phenotyping of PBMC prior to stimulation (top row) and after 5 days of stimulation with CpG-C/IL-2 (middle row) or R848/IL-2 (bottom row). Not show is the gating tree for singlets, live cells (AViD-negative) and lymphocytes (FSC/SSC). R848 stimulation leads to a higher percentage of total B cells and higher proportion of CD20^lo/− CD27^hi CD38^hi plasmablasts that are class-switched (IgD⁻) than CpG-C.

Figure 3. Optimal readout depends on PBMC and antigen concentration used in the B-cell ELISpot

A) PBMCs from an HIV-infected subject were titrated by two-fold dilutions in duplicate using wells coated with 10μg/ml (top rows) or 5μg/ml (bottom rows) SF162 gp140. Input cell numbers ranged from 500,000 to 31,250 cells/well. B) The concentration of AVI-tagged ConS was titrated by two-fold dilutions in duplicate from 4μg/ml to 0.5μg/ml using 250,000 cells/well.

Figure 4. B-cell ELISpot qualification

A) Assay specificity was determined by testing triplicate samples from 20 HIV-negative and 4 HIV subtype B-infected subjects enrolled in the Seattle Area Control (SAC) cohort with the negative control (KLH, open circles) and two HIV antigens (gp140 ConS, black squares; and gp140 BaL, grey diamonds). The dotted line at 20 SFC/million PBMC represents the cut-off under which responses are considered negative. PBMC were plated at 500,000 cell/well. B and C) The variability of the assay was determined by calculating the coefficient of variation (%CV) for the intra-assay, inter-technician and inter-day variables for gp140 ConS (B) and gp140 BaL (C). Only data from positive responses were included. The cut-off was set at 20% (dotted line). None of the samples reached this level.

Figure 5. Vaccine-induced memory B-cell responses

Vaccine recipients from HVTN 204 were tested at baseline (left columns) and four weeks after the Ad5 boost (right columns) with ConS (diamonds) and BaL (circles) gp140 proteins. The dotted line at 20 SFC/million PBMC represents the cut-off under which responses are considered negative. Open symbols represent negative responses, closed symbols represent positive responses.

Figure 6. IgG/IgA FluoroSpot for PBMC

PBMC from two vaccine recipients after vaccination (Participant 1 and 2) and one at baseline (Participant 3) were added to a FluoroSpot plate to determine the reactivity towards different HIV proteins and the negative control (KLH), and to measure total IgG and total IgA. IgA was detected using FITC-labeled secondary antibodies and develops as green spots. IgG was detected using Streptavidin-Red-labeled secondary antibodies and develops as yellow spots. 5×10⁵ stimulated PBMC were added to wells coated with KLH, gp140 ConS and gp140 BaL. 2.5×10⁵ stimulated PBMC were added to wells coated with gp140 MN and gp120 A244. 1 × 10⁵ stimulated PBMC were added to wells coated with gp140 ConS. 5×10³