SignificanceAntigenic variation requires frequent revision of annual influenza vaccines. Next-generation vaccine design strategies aim to elicit a broader immunity by directing the human immune response toward conserved sites on the principal viral surface protein, the hemagglutinin (HA). We describe a group of antibodies that recognize a hitherto unappreciated, conserved site on the HA of H1 subtype influenza viruses. Mutations in that site, which required a change in the H1 component of the 2017 vaccine, had not previously “taken over” among circulating H1 viruses. Our results encourage vaccine design strategies that resurface a protein to focus the immune response on a specific region.

For broad protection against infection by viruses such as influenza or HIV, vaccines should elicit antibodies that bind conserved viral epitopes, such as the receptor-binding site (RBS). RBS-directed antibodies have been described for both HIV1–3 and influenza virus4–8, and the design of immunogens to elicit them is a goal of vaccine research in both fields. Residues in the RBS of influenza virus hemagglutinin (HA) determine a preference for the avian or human receptor, α -2,3-linked sialic acid and α -2,6-linked sialic acid, respectively9,10. Transmission of an avian-origin virus between humans generally requires one or more mutations in the sequences encoding the influenza virus RBS to change the preferred receptor from avian to human9,11,12, but passage of a human-derived vaccine candidate in chicken eggs can select for reversion to avian receptor preference13–15. For example, the X-181 strain of the 2009 new pandemic H1N1 influenza virus, derived from the A/California/07/2009 isolate and used in essentially all vaccines since 2009, has arginine at position 226, a residue known to confer preference for an α -2,3 linkage in H1 subtype viruses13,14; the wild-type A/California/07/2009 isolate, like most circulating human H1N1 viruses, has glutamine at position 226. We describe, from three different individuals, RBS-directed antibodies that recognize the avian-adapted H1 strain in current influenza vaccines but not the circulating new pandemic 2009 virus; Arg226 in the vaccine-strain RBS accounts for the restriction. The polyclonal sera of the three donors also reflect this preference. Therefore, when vaccines produced from strains that are never passaged in avian cells become widely available, they may prove more capable of eliciting RBS-directed, broadly neutralizing antibodies than those produced from egg-adapted viruses, extending the established benefits of current seasonal influenza immunizations.

Antibody rejection is often accompanied by non-donor HLA specific antibodies (NDSA) and self-reactive antibodies that develop alongside donor-specific antibodies (DSA). To determine the source of these antibodies, we immortalized 107 B cell clones from a kidney transplant recipient with humoral rejection. Two of these clones reacted to HLA class I or MICA. Both clones were also reactive to self antigens and a lysate of a kidney cell line, hence revealing a pattern of polyreactivity. Monoclonality was verified by the identification of a single rearranged immunoglobulin heavy chain variable region (VH) sequence for each clone. By tracking their unique CDR3 sequence, we found that one such polyreactive clone was highly expanded in the patient blood, representing ~0.2% of circulating B cells. The VH sequence of this clone showed evidence of somatic mutations that were consistent with its memory phenotype and its expansion. Lastly, the reactivity of the expanded polyreactive B cell clone was found in the patient serum at time of rejection. In conclusion, we provide here proof of principle at the clonal level that human antibodies can cross-react to HLA and self. Our findings strongly suggest that polyreactive antibodies contribute to DSA, NDSA as well as autoantibodies, in transplant recipients.

Background B cell infiltrates are common in rejected kidney allografts, yet their composition is still unclear. The aim of our study was to characterize the clonal composition of B cell infiltrates of rejected human kidney grafts. Methods We used a molecular approach to characterize the partial B cell repertoires of 5 failed human kidney grafts with detectable B cell infiltrates. A comparison between the intragraft and blood repertoire was also conducted for one case. Results Redundant sequences were observed in both blood and graft, although the level of clonal amplification was significantly higher for the graft. Somatic hypermutations (SHM) were also more frequent in sequences found in the graft compared to the blood. The rate of non-silent mutations was significantly higher in complementarity determining regions (CDR) compared to framework regions in blood sequences as well as in graft sequences found at low frequency. In contrast, this preferential distribution was lost in sequences found at high frequency in the graft, suggesting a lack of affinity maturation in situ. Lastly, follicular dendritic cells were undetectable in CD20+ infiltrates in all samples examined. Conclusions We provide here evidence that B cell clones expand and undergo SHM in situ. However, the even distribution of non-silent SHM in high frequency graft sequences together with the absence of FDC do not support the view that infiltrating B cells are part of functional germinal centers.