Surges of serum antibodies after immunization and infection are highly specific for the offending antigen, and recent studies demonstrate that vaccines induce transient increases in circulating antibody-secreting cells (ASCs). These ASCs are highly enriched but not universally specific for the immunizing antigen, suggesting that a fraction of these ASCs could arise from polyclonal bystander stimulation of pre-existing memory cells to unrelated antigens. This model is proposed to explain maintenance of long-lived serological memory in the absence of antigen exposure. To test this model, we measure the ability of RSV and influenza virus infection or immunizations to influenza virus, tetanus toxoid, hepatitis B, and human papilloma virus to stimulate bystander memory cells specific for other major environmental antigens that represent a large fraction of the pre-existing memory B compartment. Bystander or non-specific ASC responses to RSV and tetanus could not be detected above the background levels in healthy adults, despite the presence of circulating memory B cells specific for the corresponding antigens. Non-specific ASC responses in the healthy subjects were similar to frequencies in cord blood samples. In contrast, both vaccination and infection induce massive expansion of circulating antigen-specific ASCs without significant increases in the frequencies of ASCs against unrelated antigens. Hence, non-specific stimulation of memory B cells is unlikely to contribute to the mechanisms of long-term serological memory against major human pathogens. Additionally, high specificity of circulating ASC after antigenic challenge highlight the diagnostic value of interrogating ASCs as an ideal one-time-point diagnostic immune surrogate for serology during acute infection.

Human CD4 T cell recall responses to influenza virus are strongly biased towards Type 1 cytokines, producing IFNγ, IL-2 and TNFα. We have now examined the effector phenotypes of CD4 T cells in more detail, particularly focusing on differences between recent versus long-term, multiply-boosted responses. Peptides spanning the proteome of temporally distinct influenza viruses were distributed into pools enriched for cross-reactivity to different influenza strains, and used to stimulate antigen-specific CD4 T cells representing recent or long-term memory. In the general population, peptides unique to the long-circulating influenza A/New Caledonia/20/99 (H1N1) induced Th1-like responses biased toward the expression of IFNγ+TNFα+ CD4 T cells. In contrast, peptide pools enriched for non-cross-reactive peptides of the pandemic influenza A/California/04/09 (H1N1) induced more IFNγ−IL-2+TNFα+ T cells, similar to the IFNγ−IL-2+ non-polarized, primed precursor T cells (Thpp) that are a predominant response to protein vaccination. These results were confirmed in a second study that compared samples taken before the 2009 pandemic to samples taken one month after PCR-confirmed A/California/04/09 infection. There were striking increases in influenza-specific TNFα+, IFNγ+, and IL-2+ cells in the post-infection samples. Importantly, peptides enriched for non-cross-reactive A/California/04/09 specificities induced a higher proportion of Thpp-like IFNγ−IL-2+TNFα+ CD4 T cells than peptide pools cross-reactive with previous influenza strains, which induced more Th1 (IFNγ+TNFα+) responses. These IFNγ−IL-2+TNFα+ CD4 T cells may be an important target population for vaccination regimens, as these cells are induced upon infection, may have high proliferative potential, and may play a role in providing future effector cells during subsequent infections.

No abstract