Public endowment of B cells

Etoc

Using a transgenic mouse strain expressing the human V_H1-69 germline gene used by many broadly neutralizing antibodies to influenza A virus (IAV), Sangesland et al. show that the V_H1-69 gene segment provides the essentials for mounting antibody responses against the broadly neutralizing hemagglutinin (HA) stem epitope.

The humoral response to the IAV HA is dominated by antibodies to variable epitopes located around the receptor binding site. Due to the variability of their epitopes, these antibodies are highly strain-specific and only neutralize the eliciting or closely related IAV strains. Heterosubtypic neutralizing antibodies (hnAbs), i.e. antibodies that can bind and neutralize multiple strains and subtypes (serotypes) of IAV, are rare but can be found in most individuals at low levels (Corti et al., 2010; Kohler et al., 2014; Sui et al., 2011). hnAbs specific for the conserved epitope located in the stem of IAV HA surface protein are special in that they are preferentially encoded by the V_H1-69 germline gene, display a preference for phylogenetic group 1 viruses, and almost exclusively use their heavy chain for binding to HA (Avnir et al., 2014; Corti et al., 2010; Ekiert et al., 2009; Pappas et al., 2014; Wheatley et al., 2015). V_H1-69 encoded hnAbs constitute a shared or ‘public’ antibody solution as very similar B cell clonotypes specific for this epitope can be found in various individuals. While it is quite possible that independent but convergent B cell evolution leads to public solutions, it has also been speculated that mammalian B cell repertoires have evolved to encode for public solutions to improve responses to commonly encountered pathogens. In this issue, Sangesland et al. present a study that uses transgenic mice expressing a single human V_H gene in the context of all human D and J_H elements to show that the public V_H1-69 HA stem solution is indeed inherent to this V region, and that the presence of V_H1-69 is sufficient to enable antibody responses to this desirable public but subdominant IAV HA epitope.

Antibodies capable of slowing viral dissemination during the first days of infection can be crucial for the survival of the infected host and help to bridge the time until the adaptive immune system has fully kicked in (Hangartner et al., 2006). Part of this first line of defense are natural antibodies, i.e. germline encoded immunoglobulin M (IgM) antibodies that help to prime the immune system by recruiting antigen to secondary lymphoid organs and by keeping early pathogen dissemination at bay (Ochsenbein and Zinkernagel, 2000). The activity of germline-encoded, bona fide low affinity antibodies is augmented by their secretion as pentameric or hexameric IgM antibodies whose higher valency of binding can compensate for poor affinity until higher affinity antibodies have been selected. IgM is also a potent activator of the complement system, which helps both restricting pathogen dissemination and recruitment of antigen to secondary lymphoid organs.

While hardwiring of public specificities into the B cell repertoire is beneficial for surviving common pathogens, it may be at the expense of skewing the antibody repertoire: providing public specificities with a head start will put non-public specificities at a disadvantage, and possibly can result in immunodominance. As a matter of fact, immunodominance has proven to be a major issue for the development of universal vaccines against variable viruses such as HIV or influenza virus. While the variable and hence strain-specific epitopes of the surface proteins of these viruses are highly immunogenic, efforts to immunize against their conserved epitopes have proven to be difficult. In the case of influenza viruses, experimental vaccines can elicit antibodies to the conserved stem epitope but those typically fail to neutralize the virus.

The specificity of antibodies typically comes from the heavy and light chain complementarity determining regions (CDRs). While two of the three CDRs for each chain are encoded in the V region gene (CDR1 and 2), the third (CDR3) arises as the result of the V(D)J recombination during B cell ontogeny. Due to higher variability compared to CDR1 and 2, CDR3, especially that of the heavy chain (HCDR3), often provides the essential antigen contacts with the antigen. V_H1-69-encoded antibodies are special in that they can evolve from the V_H1-69 germline gene with as little as little as one or two V-region substitutions when combined with a properly positioned HCDR3 Tyrosine: e.g. replacement of HCDR2 Ile 52a with a smaller small amino acid, such as Serine, Alanine, Glycine, allows the germline encoded but polymorphic HCDR2 Phenylalanine 54 and the HCDR3 Tyrosine, to be inserted into adjacent pockets to establish the key contacts (Avnir et al., 2014; Pappas et al., 2014) between antibody and HA stem. Since removal of the HCDR3 Tyrosine but not of the HCDR2 Phenylalanine can be tolerated in many V_H1-69 hnAbs, Sangesland et al. speculated that V-segment encoded residues form ‘primordial’ germline contacts that enable selection of V_H1-69-expressing hnAb precursor B cells.

To test their hypothesis, the authors used two transgenic mouse strains expressing either the human V_H1-69 or V_H1-2 gene segments in conjunction with all human D and J_H elements (Figure 1). They then immunized these humanized mouse strains with infectious virions, and virus like particles (VLPs) displaying either full-length HA or a specially crafted ‘headless’ HA lacking the immunodominant head epitopes. The authors found that V_H1-69 but not V_H1-2 humanized mice were able to mount stem-specific antibodies, enrich for B cells expressing B cell receptors (BCRs) with the aforementioned heterosubtypic antibody signatures, and were protected against a heterosubtypic virus challenge. Thereby they could prove that the V_H1-69 region indeed provides all the necessary features for eliciting HA stem specific hnAbs.

Figure 1.

Antibodies to the conserved stem epitope (yellow circle) of influenza virus hemagglutinin (HA, blue and gray) are difficult to elicit, and immunization of wild-type mice primarily elicits antibodies to the antigenically variable head of HA. Closely related stem-specific and broadly reactive V_H1-69-encoded antibodies can be found in many humans, indicating that they constitute a public specificity. V_H1-69-encoded stem antibodies are special in that they partially rely on germline encoded residues for strain-independent binding and neutralization. Using a mouse strain that has been genetically engineered to encode for a single human germline gene in combination with human D and J elements, Sangesland et al could demonstrate, that the presence of the V_H1-69, but not the human V_H1-2 gene segment enabled mice to mount antibodies to the conserved stem-epitope and thereby provide evidence that the public specificity for this type of stem-specific heterosubtypic antibodies is segregating with the V_H1-69 gene segment.

In contrast to HIV, where potential vaccine recipients are immunologically naïve, most IAV vaccine recipients can be expected to be have been previously exposed to the virus. Sangesgard et al. could show that also in IAV pre-exposed V_H1-69 mice, headless-HA VLP immunization could enrich for B cells expressing hnAb signatures and refocus the virus-induced antibody response on the HA-stem hnAb epitope. To demonstrate that their observations are not an artifact of the drastically increased precursor B cell frequencies in the transgenic mice, Sangesgard et al., then used an elegant H1 and H5 HA dual staining method to determine heterosubtypic B cell precursor frequencies in humans, and matched these frequencies by adoptively transferring transgenic B cells into wt mice. Importantly, they could show that their headless HA VLP could also elicit V_H1-69-encoded stem-specific antibodies at human physiological precursor B cell frequency.

The study demonstrates that V-regions can incorporate the essentials for public antibody solutions, which under the right conditions can enable the immune system to mount antibody responses against difficult but desirable subdominant epitopes. However, the findings in this study also indicate that no matter how well an immunogen has been engineered for optimal presentation of a particular epitope, it will fail if there is not a sufficient number of precursor B cells available. Conversely, also in this model, availability of increased precursor frequencies of desired B cells was not sufficient to prevent the immunodominant, strain-specific epitopes from taking over eventually, indicating that the V_H1-69-encoded public hnAb endowment is not very competitive compared to other epitopes. Nonetheless, the results presented by Sangesgard et al. back current reverse vaccinology approaches to elicit antibodies against the subdominant broadly neutralizing epitopes of HIV and IAV. Since many of these approaches aim to increase precursor B cell frequencies, e.g. with germline-targeting immunogens, and to seek to avoid immune responses to dominant variable epitopes, e.g. by masking or switching the dominant epitopes, these results are encouraging for the further development of such universally protective vaccines.