Higher levels of B-cell mutation in the early germinal centres of an inefficient secondary antibody response to a variant influenza haemagglutinin

Abstract

Designing improved vaccines against mutable viruses such as dengue and influenza would be helped by a better understanding of how the B-cell memory compartment responds to variant antigens. Towards this we have recently shown, after secondary immunization of mice with a widely variant dengue virus envelope protein with only 63% amino acid identity, that IgM⁺ memory B cells with few mutations supported an efficient secondary germinal centre (GC) and serum response, superior to a primary response to the same protein. Here, further investigation of memory responses to variant proteins, using more closely related influenza virus haemagglutinins (HA) that were 82% identical, produced a variant-induced boost response in the GC dominated by highly mutated B cells that failed, not efficiently improving serum avidity even in the presence of extra adjuvant, and that was worse than a primary response. This supports a hypothesis that over a certain level of antigenic differences, cross-reactive memory B-cell populations have reduced competency for affinity maturation. Combined with our previous observations, these findings also provide new parameters of success and failure in antibody memory responses.

Keywords: B-cell memory; affinity maturation; antibodies; cross-reactive; influenza virus.

Figure 1

Primary and secondary response to homotypic PR8 haemagglutinin (HA). (a) Anti‐PR8 HA serum IgG response. Bar shows mean. Px, x days since priming; Bx, x days since boosting; BO, adjuvant‐only primed, PR8 HA boosted, analysed day 6; His, B6 serum sample reactivity with irrelevant C‐terminal His‐tagged protein (Bacillus clausii pdxR). (b) FACs gating strategy. (c) GC B‐cell response to PR8 HA. Bar shows mean. NI, not immunized; other labels as for panel (a). (d) VH mutations in single sorted germinal centre (GC) B cells. From n = 3 mice (Px) and n = 6 mice (Bx). Bar shows median value. Labels as for panel (a). (e) Relative serum IgG avidity for PR8 HA. Bar shows mean. Labels as for panel (a).

Figure 2

Secondary responses to variant haemagglutinin (HA) boosting. (a) Anti‐Brisbane 59/07 (Bris) HA serum IgG titres after Bris HA boosting. Bars indicate mean. x‐axis labels: P44, PR8 HA1 prime only, day 44. All other labels: Px, x days since priming with antigen indicated below; Bx, x days since boosting with antigen indicated below. Bris, Bris HA, Bris + A, Bris HA including adjuvant. Note: The Bris* prime was performed with Bris HA, not HA1, to allow appropriate comparison with Bris HA‐boosted groups. (b) Germinal centre (GC) B‐cell response to Bris HA boost. Bars show mean. First three data sets reproduced from Fig. 1(c) for comparison. x‐axis labels as for panel (a). (c) VH mutations in single sorted GC B cells. From n = mice in each group indicated. Bar shows median. x‐axis labels as for panel (a). First two data sets reproduced from Fig. 1(d) for comparison. (d) Proportion of IgM⁺ GC B‐cells in day 6 GC. Bar shows mean. x‐axis labels as for panel (a). (e) Relative serum IgG avidity for Bris HA. Bars show mean. x‐axis labels as for panel (a). (f) VH mutations in IgM⁺ and IgG⁺ GC B cells. Number of mice in each group indicated as n = . Bars indicate median. x‐axis labels as for panel (a).