Pan-neutralizing, germline-encoded antibodies against SARS-CoV-2: Addressing the long-term problem of escape variants

Abstract

Despite the initially reported high efficacy of vaccines directed against ancestral SARS-CoV-2, repeated infections in both unvaccinated and vaccinated populations remain a major global health challenge. Because of mutation-mediated immune escape by variants-of-concern (VOC), approved neutralizing antibodies (neutAbs) effective against the original strains have been rendered non-protective. Identification and characterization of mutation-independent pan-neutralizing antibody responses are therefore essential for controlling the pandemic. Here, we characterize and discuss the origins of SARS-CoV-2 neutAbs, arising from either natural infection or following vaccination. In our study, neutAbs in COVID-19 patients were detected using the combination of two lateral flow immunoassay (LFIA) tests, corroborated by plaque reduction neutralization testing (PRNT). A point-of-care neutAb LFIA, NeutraXpress™, was validated using serum samples from historical pre-COVID-19 negative controls, patients infected with other respiratory pathogens, and PCR-confirmed COVID-19 patients. Surprisingly, potent neutAb activity was mainly noted in patients generating both IgM and IgG against the Spike receptor-binding domain (RBD), in contrast to samples possessing anti-RBD IgG alone. We propose that low-affinity, high-avidity, germline-encoded natural IgM and subsequent generation of class-switched IgG may have an underappreciated role in cross-protection, potentially offsetting immune escape by SARS-CoV-2 variants. We suggest Reverse Vaccinology 3.0 to further exploit this innate-like defense mechanism. Our proposition has potential implications for immunogen design, and provides strategies to elicit pan-neutAbs from natural B1-like cells. Refinements in future immunization protocols might further boost long-term cross-protection, even at the mucosal level, against clinical manifestations of COVID-19.

Keywords: B cell memory; B-1 B cells; IgM; Reverse Vaccinology; SARS-CoV-2; neutralizing antibodies; somatic hypermutation; vaccines.

Figure 1

NeutAbs against SARS-CoV-2 detected with NeutraXpress™ in the early phase of natural infection are enriched in COVID-19 patients possessing the IgM+ signature. (A) Illustration of NeutraXpress™ design (23). T1 is striped with recombinant His-tagged human ACE2 protein. T2 is striped with anti-human IgM + IgG Abs. The conjugate pad is impregnated with colloidal gold nanoparticles (GNP)-labeled recombinant RBD from Spike protein of SARS-CoV-2, as well as GNP-labeled chicken IgY used as a tracer to indicate the completion of the lateral flow when it is captured by goat anti-chicken antibody at the C line. If there is no neutAb or binding antibody in the specimen, GNP-RBD is captured by ACE2 at T1 line and the T2 line should not appear. If the specimen contains neutAbs, the interaction between GNP-RBD with ACE2 at the T1 line is blocked and T1 disappears or shows reduced intensity, in comparison with T1 from the control well with added diluent only. The appearance of the T2 line indicates the presence of IgM and/or IgG Abs specific for RBD, i.e., T2 shows the totality of both neutralizing and non-neutralizing RBD-binding IgM and IgG Abs. T2 intensity correlates with higher titers for RBD-binding IgM + IgG Abs, but T2 does not provide information on neutAbs. (B) Examples of NeutraXpress™ showing blood samples from healthy subjects at 21 days (left) and 161 days (right) after receiving the 2^nd dose of mRNA vaccine. Note that the left sample wells were added with diluent only, whereas the right sample wells were added with 1 drop of whole blood. (C) PCR-confirmed COVID-19 patients were sub grouped based on their serum IgM/IgG profiles of anti-RBD reactivity, using an LFIA rapid test DISCOVID™ and graphed according to the number of days following symptom onset or PCR positivity (in asymptomatic patients) that the sample was taken. Each square symbol represents one patient. Each data point was further differentiated to show the presence of neutAbs in the serum samples as determined by a second LFIA rapid test NeutraXpress™. Green coloration indicates the presence of neutAbs and red coloration indicates the absence of neutAbs. Samples positive for SARS-CoV-2 neutAbs were concentrated in patients having an IgM+IgG+ anti-RBD profile, compared with patients that were anti-RBD IgG+ only (*P <0.05, two-sided two proportion z test).

Figure 2

Workflow of Reverse Vaccinology 3.0. (A) In Reverse Vaccinology 1.0, complete genome sequences of pathogens are analyzed and genes coding for surface exposed proteins are identified (a). Potential surface-exposed proteins are expressed (b), and used to immunize the mice (c). Immunogens that can most efficiently elicit protective responses are screened (d), and further optimized as clinical trial leads for human vaccines (e). This process obviates the need to directly culture the pathogen. (B) In Reverse Vaccinology 2.0, structural data are utilized to guide immunogen design. First, MBCs or plasmablasts from PBMCs of subjects seropositive through infection are enriched and sorted (a). Then sorted single B cells are cultured and stimulated to screen for antigen-specific B cell clones in neutralization assays and their paired VH and VL genes are PCR amplified and sequenced (b). Recombinant neutAbs are expressed in mammalian cells, e.g., HEK293 cells or CHO cells (c), to obtain sufficient materials for function confirmation in animal models (not shown), and for structural characterization of such neutAbs bound to their target antigen (Ag). Co-crystallization analysis of neutAb (usually Fab) and antigen provides detailed structural information of the protective epitope or the conformation of the antigen in general, which is different from the one in the metastable form of the antigen that often can fail to induce neutAbs (d). Protein engineering is guided by 3D modeling to stabilize the monomer immunogen and present it in a multimeric nanoparticle format as a potential candidate for human vaccine (e). (C) In Reverse Vaccinology 3.0, the goal is to design a germline-antibody-binding immunogen and gradually evolve germline neutAbs into ones with sufficient breadth to neutralize the current pathogen and its future variants. This process starts by enriching of B cells from the first-line of defense, i.e., fluids in the body cavity and mucosa [e.g., from sputum, ref (71)], where natural antibody-secreting B-1 B cells are predominantly present (a’). Antigen-binding single B cells are functionally screened and their Ig sequences are scrutinized with software [e.g., IgSCUEAL, ref (72)] to identify the germline sequences (b’). Next generation sequencing of enriched B-1 B cells from a cohort of infected patients at early stages of the disease (e.g., stratified by our method in Figure 1 ), may help identify convergent germline sequences induced by infection (38). Modeling with software such as Rosetta allows for the calculation of the interacting area between the Fab of germline neutAbs and the antigen (c’). Since the native antigen usually does not bind the germline antibody sequences (73, 74), it is necessary to generate a yeast surface displayed random mutational library of the antigen to select for variants that bind germline neutAbs as the initial immunogen (d’). Such immunogens would be presented in multimeric form on self-assembled nanoparticles during primary vaccination (e’), followed by sequential boosting with homologous and/or heterologous immunogens to facilitate somatic hypermutation and nurture broad neutAbs that also protect against future variants (f). This strategy relies heavily on computational bioinformatics, and has a species restriction on Ig repertoire, hence humanized mice with knock in human Ig locus would be the preferred preclinical animal model (75).