Challenges and Prospects in Developing Future SARS-CoV-2 Vaccines: Overcoming Original Antigenic Sin and Inducing Broadly Neutralizing Antibodies

Understanding the interactions between SARS-CoV-2 and the humoral immune response is critical to developing strategies for inducing broad and durable protection from COVID-19. The SARS-CoV-2 spike (S) protein mediates virus entry into host cells through recognition of the virus receptor angiotensin converting enzyme 2 (ACE2) and serves as a key target of the host humoral immune response (1) and is the main target of neutralizing Abs (NAbs) for vaccine development.

The NAb response is a key determinant of immune protection from viral infection and severe disease (2). It has been noted that pre-existing NAbs are key correlates for protection from secondary SARS-CoV-2 outbreaks as observed in well-monitored populations (3). Similarly, in vivo studies on rhesus macaques have demonstrated that humoral immune protection, especially NAbs, is critical for blocking SARS-CoV-2 infection (4–6).

In this Brief Review, we discuss recent findings that inform the design of the next generation of COVID-19 vaccines. In the context of prior SARS-CoV-2 evolution, we discuss strategies under investigation for the induction of broadly neutralizing Abs (bNAbs) to restrict SARS-CoV-2. Additionally, we examine recent evidence for the importance of mucosal immunity in restricting SARS-CoV-2 transmission. We further consider the challenges associated with these next-generation vaccine strategies, including overcoming “original antigenic sin” (OAS), while highlighting recent evidence elucidating the mechanism of OAS. The induction of broad, long-lasting protection against SARS-CoV-2 remains a critical goal to control the COVID-19 pandemic and address future viral pandemics.

Shortly after the emergence of SARS-CoV-2, a single mutation, D614G, in the S protein represented a key adaptation of SARS-CoV-2 for transmission in humans, quickly becoming dominant in circulating SARS-CoV-2 (Fig. 1) (7). The presence of this single mutation in the C-terminal end of the S1 subunit has been shown to be associated with enhanced virus infectivity, increased transmission, and higher virus titers in the nasopharynx, while increasing sensitivity to neutralizing Abs (7–10).

Subsequently, several SARS-CoV-2 variants emerged with mutations in the receptor binding domain (RBD) that generated concern over possible NAb escape. These included the Alpha variant (Pango lineage B.1.1.7), the Beta variant (B.1.351), and the Gamma variant (P.1). However, although the Beta and Gamma variants exhibited more substantial NAb resistance than did the Alpha variant, it was the Alpha variant that became the next dominant SARS-CoV-2 strain (Fig. 1). The Alpha variant was succeeded by the emergence of the Delta variant (B.1.617.2) (Fig. 1), characterized by the RBD mutations L452R and T478K that conferred enhanced neutralization resistance and transmissibility over the Alpha variant, although this neutralization resistance was not as substantial as the Beta variant (11, 12). This is emblematic of the initial stage of viral adaptation to a novel host species wherein virus evolution is driven by selection for enhanced transmission and utilization of host factors while immunological pressures are minimal in the immune-naive population. However, mounting immunologic pressures would continue to drive virus evolution.

The Omicron variant (B.1.1.529) marked a key turning point in the COVID-19 pandemic. The emergence of the Omicron variant caused significant concern as its S protein contained an alarming >30 mutations including 16 in the RBD. Subsequent characterization of the Omicron variant demonstrated its significant NAb escape (13, 14). Additionally, the Omicron variant exhibited substantially enhanced transmissibility, resulting in an unprecedented spike in COVID-19 cases (Fig. 1) (15, 16). However, the disease severity of the Omicron variant was substantially reduced (17, 18). Notably, in the Syrian hamster model of SARS-CoV-2 infection, Omicron exhibited comparable transmission to Delta; however, the Omicron variant was outcompeted by the Delta variant in the absence of NAbs (19). Thus, the Omicron variant possesses a key immune escape advantage over the Delta variant. The Omicron variant has since diversified into numerous subvariants, including the currently dominant XBB clade, which appear to be exhibiting increasing NAb escape (20–25).

Virus evolution for the evasion of a NAb response has been a longstanding problem in developing effective vaccines against highly mutable RNA viruses. NAbs have been shown to drive evolution of several RNA viruses including IAV, HIV, HCV, PRRSV, and MERS-CoV (26–31). For example, persistent antigenic drift requires the development of annual influenza vaccines. This strategy has yielded vaccines of variable efficacy and limited uptake by the general population. Given the continual evolution of SARS-CoV-2 for NAb escape, several strategies have been proposed for the induction of bNAbs targeting all SARS-CoV-2 variants and for the development of pan-coronavirus vaccines.

Several studies have isolated and characterized coronavirus bNAbs that efficiently target SARS-CoV-2 (32). Notably, isolation of mAbs from SARS-CoV-2 “elite neutralizers” demonstrated the potential for the induction of Abs capable of neutralizing all human coronaviruses (33). Epitope mapping of bNAbs identified a class of bNAbs targeting the stem-helix bundle of S2 in prefusion S trimers that are predicted to disrupt S-mediated membrane fusion (34). Critically, these bNAbs cross-neutralize the human-infecting Betacoronaviruses, including SARS-CoV-2, SARS-CoV, Middle East respiratory syndrome (MERS)-CoV, HKU1, OC43, as well as other Betacoronaviruses including mouse hepatitis virus (34). Notably, these stem helix–targeting bNAbs failed to neutralize the Alphacoronaviruses NL63 and 229E (34). Other studies have identified bNAbs capable of neutralizing both the Alphacoronaviruses and Betacoronaviruses infecting humans by targeting the fusion peptide (35, 36). Taken together, these results indicate that induction of a bNAb response capable of inducing pan-coronavirus or pan-Betacoronavirus immunity may be possible.

Following the emergence of SARS-CoV-2 Omicron subvariants with escape from NAbs, the development of reformulated mRNA vaccines has been frequently proposed (37, 38). This ultimately resulted in the development of the bivalent mRNA booster vaccines by Moderna and Pfizer/BioNTech (39, 40). This bivalent mRNA vaccine included mRNA for both the prototype S sequence and the BA.5 variant S sequence (or the BA.1 variant S as in the United Kingdom) and was introduced during the BA.5 wave of infections. It was suspected that the presence of BA.5 in the vaccine would induce a naive Ab response against neutralizable epitopes unique to the Omicron clade and boost the induction of NAbs against conserved epitopes. Furthermore, the presence of the prototype S would maintain high NAb titers against earlier SARS-CoV-2 variants to provide protection if a variant unrelated to the Omicron clade began to dominate.

The bivalent mRNA boosters appear to induce a strong NAb response against the BA.5 variant. Early reports described minimal difference in NAb titers between recipients of a monovalent fourth dose and recipients of a bivalent fourth dose (41–43); however, studies supported by Pfizer/BioNTech demonstrated that the bivalent fourth dose enhanced neutralization of Omicron subvariants by 3.6- to 8.9-fold compared with a monovalent fourth dose (44). This enhancement of Omicron neutralization by the bivalent vaccine seems dependent on prior SARS-CoV-2 exposure, with patients having prior SARS-CoV-2 infection exhibiting a more modest difference in NAb breadth between the monovalent and bivalent booster groups (44). The discrepancy in the potential advantage of the bivalent booster over the monovalent booster requires further investigation, as the results from the independent studies used smaller sample sizes compared with the studies supported by Moderna and Pfizer/BioNTech, and their cohorts may include individuals with unreported COVID-19 infections due to minimal symptoms. Importantly, in vivo studies have demonstrated reduced virulence of Omicron infection in mice vaccinated with two doses of the monovalent vaccine and one dose of the bivalent vaccine compared with mice vaccinated with three monovalent doses (45). Finally, examinations of vaccine efficacy have indicated that recipients of three monovalent doses and one bivalent dose exhibit 33.9% higher vaccine efficacy against hospitalization compared with recipients of four monovalent doses (46). Thus, it is possible that the bivalent booster regimen exhibits modestly enhanced protection against Omicron subvariants.

Continued evolution of SARS-CoV-2 Omicron subvariants has resulted in neutralization escape from bivalent mRNA vaccine-induced sera. Recently, the emerging Omicron subvariants including BQ.1.1 and XBB have been shown to exhibit resistance to bivalent booster-induced sera (25, 47, 48). This may require recurrent reformulation of mRNA booster vaccines to induce a NAb response to circulating SARS-CoV-2. However, increasing the valency of the mRNA vaccines may diminish the NAb response against individual variants owing to increased Ag competition and increased epitope suppression (49).

Ag-presented nanoparticle vaccines are a new type of protein subunit vaccine in which multiple Ags are presented on one self-assembling particle. These nanoparticle vaccines induce Ab responses more potently than simple subunit vaccines, including for SARS-CoV-2 nanoparticle vaccines in preclinical and clinical trials (50–56). The multimeric nature of these nanoparticle vaccines enhances a process known as BCR cross-linking. BCR cross-linking is a process by which Ag-bound BCRs colocalize to enhance induction of B cell activation (57–60). Multimeric Ag-presenting nanoparticles can facilitate BCR cross-linking to greatly enhance B cell activation, leading to a robust B cell response.

Attempts have been made to use such nanoparticle vaccines to generate pan-coronavirus or pan-sarbecovirus vaccines. These include nanoparticle vaccines containing S or RBD from multiple coronaviruses (50, 61, 62). Critically, this mosaic nanoparticle preferentially facilitates BCR cross-linking for B cells targeting epitopes conserved among the displayed Ags. This will enhance the induction of Abs targeting conserved epitopes over those targeting variable epitopes. The “Mosaic-8b” nanoparticle vaccine candidate bears RBDs of diverse sarbecoviruses including SARS-CoV-2 and the bat coronaviruses RaTG13, SHC014, Rs4081, RmYN02, Rf1, and WIV1 as well as the pangolin coronavirus Pang17 (61, 62). Vaccination of mice and rhesus macaques with Mosaic-8b induced a NAb response against diverse sarbecoviruses and protected mice and macaques from challenge with both SARS-CoV-2 and SARS-CoV (61, 62). These promising results may indicate potential success for nanoparticle vaccines in the production of pan-coronavirus or pan-sarbecovirus vaccines.

It is unclear how the successful implementation of a pan-coronavirus vaccine may impact SARS-CoV-2 evolution. Notably, bNAbs capable of pan-coronavirus or pan-Betacoronavirus neutralization occur relatively rarely, with variable epitope–targeting NAbs predominating (35, 61). As such, these conserved epitopes targeted by bNAbs are not under substantial immunologic pressure. However, if the immune landscape is altered by widespread implementation of bNAb-inducing vaccines, immunologic pressure for alterations in these conserved epitopes will increase. It is therefore possible that development of a bNAb-inducing vaccine will ultimately not prevent the emergence of escape mutants. This of course depends on the functional importance of the conserved epitopes being targeted. The pan-sarbecovirus Mosaic-8b vaccine induces the production of Abs targeting more conserved regions of the RBD that interact with other elements of the S protein, and it has therefore been suggested that these epitopes may not tolerate substantial mutagenic change (61). Similarly, the stem-helix bundle and fusion peptide are critical elements for S protein–mediated fusion, and minimal escape mutation to evade pan-coronavirus bNAbs may be tolerated in these regions. However, given that NAbs targeting these regions are rare components of the NAb responses, it is unclear whether these regions remain conserved due to their functional importance permitting minimal change or if a lack of immunologic pressure has not driven antigenic drift in these regions. Given that the high mutation rate of RNA viruses permits the sampling of mutagenic alterations, it is not unrealistic to expect that permissible alterations in functionally important regions may emerge under immunologic pressure. However, if induction of a broad, potent, and durable NAb response can be achieved, persistent sterilizing immunity may prevent SARS-CoV-2 from evolving under this immunologic pressure, as occurs with measles virus-induced immunity (63, 64).

Although pan-coronavirus vaccination could lead to the emergence of SARS-CoV-2 variants with further immune evasion, the development of a pan-coronavirus vaccine is still critical. Given that conserved coronavirus epitopes are generally in regions of the S protein with diminished immunogenicity, such as the fusion peptide, which is often buried within the prefusion S trimer, a potent NAb response against these epitopes is rarely generated. In animal reservoirs of coronaviruses, including bats, it is likely that there is minimal induction of Abs targeting conserved epitopes. Thus, if a human pan-coronavirus vaccine is to be implemented to mitigate the COVID-19 pandemic, the bNAb response induced would also reduce the risk of zoonotic transmission of novel coronaviruses into humans.

Although NAbs are key correlates for protection from virus infection, the contribution of Abs beyond neutralization should not be overlooked. The nonneutralizing functions of Abs are largely mediated through binding of the Ab Fc domain to the Fc receptor on immune cells (65). This Fc domain allows for immune complex formation, neutrophil and macrophage activation, complement activation, chemoattractant production, and NK cell–mediated cytotoxicity among other functions (65). The role of these various effector functions in protection from severe COVID-19 remains to be elucidated.

The Fc-dependent functions of NAbs and nonneutralizing Abs have been shown to diminish disease severity during SARS-CoV-2 infection. In mice, mutations in the Fc Ab domain diminish the protective effect of therapeutically and prophylactically administered NAbs (66–68). Corroborating these findings, NAbs with Fc domains optimized for binding to activating Fc receptors exhibited enhanced protection in Fc receptor humanized mice (69). Thus, the protection conferred by NAbs is not solely determined by their epitope binding but also by their capacity to induce effector functions. This may indicate a similar mechanism of function for nonneutralizing Abs. In fact, it has been demonstrated that Fc optimization of a nonneutralizing Ab allowed for diminished virus replication in mice compared with the nonoptimized version (70). However, full protection was only achieved in the presence of an administered NAb (70). Further studies distinguishing the importance of Fc-mediated effector functions from neutralizable epitope binding are needed to elucidate the importance of nonneutralizing Abs for protection from severe COVID-19.

The nature of the adaptive immune response is dependent on the location of the Ag exposure. Ag exposure within the respiratory tract leads to dendritic cell uptake of Ag and migration to MALT including the nasopharynx-associated lymphoid tissue (71). Induction of the adaptive immune response within nasopharynx-associated lymphoid tissue leads to the production of IgA-secreting plasma cells. IgA is the Ig predominantly responsible for mucosal immunity, exhibiting high concentrations in the respiratory and gastrointestinal mucosa (71). This mucosal IgA can neutralize SARS-CoV-2 virions at the site of initial infection with SARS-CoV-2, such as the respiratory mucosa, to better restrict the initiation of virus infection and spread.

In addition to strategies for the generation of durable bNAbs to prevent SARS-CoV-2 immune evasion, many studies have sought to induce a mucosal immune response to diminish SARS-CoV-2 transmission. This strategy may help to prevent the emergence of neutralization escape variants of SARS-CoV-2, as establishment of virus infection may be more efficiently blocked. Notably, such mucosal immunity is not strongly induced by mRNA vaccination but can be induced by natural SARS-CoV-2 infection (72, 73). However, mRNA vaccination in previously infected patients has been demonstrated to boost saliva anti-S IgA (73). Thus, systemic immunization in the context of prior mucosal immunity can induce a mucosal immune response. Furthermore, in healthcare workers (HCWs) having received three mRNA vaccine doses, Omicron infection elicited a mucosal IgA response (74). However, this IgA response was more rapid and achieved higher titers in HCWs with prior SARS-CoV-2 infection (74). Hence, priming with systemic immunizations still permits the induction of a mucosal immune response.

It is critical to determine whether the induction of mucosal immunity provides enhanced protection from SARS-CoV-2. In HCWs who had received three monovalent mRNA vaccine doses, pre-existing mucosal IgA titers were predictive of protection from Omicron infection, with HCWs with the highest IgA titers (>75th percentile) exhibiting 65% lower risk of becoming RT-PCR positive for Omicron during the study window (74). However, mucosal IgG, which correlates with serum IgG, did not exhibit a similar protective effect (74). Similarly, in Syrian hamsters, intranasal (i.n.) administration of the AstraZeneca ChAdOx adenovirus-based SARS-CoV-2 vaccine provided comparable protection from SARS-CoV-2 challenge as i.m. vaccination, and i.n. immunization provided enhanced protection from SARS-CoV-2 transmission and subsequent virus shedding in cohoused hamsters (75). Taken together, these findings suggest that induction of a mucosal immune response can provide superior protection from SARS-CoV-2 infection.

Several preclinical studies have examined the efficacy of various vaccine platforms for the induction of mucosal immunity. Virus vector–based vaccines can often be administered i.n., allowing for the induction of a mucosal immune response. In a murine model, an i.m. prime with an mRNA SARS-CoV-2 S vaccine followed by an i.n. boost with an adenovirus vector-based SARS-CoV-2 S vaccine induced both IgA production and a mucosal T cell response (76). Critically, the i.m. mRNA prime with i.n. adenovirus vector boost induced higher IgG, IgA, and NAb titers in bronchoalveolar lavage samples than either a homologous mRNA i.m. prime and boost or an adenovirus i.m. prime and boost series (76). However, these viral vector–based systems can induce antivector immunity, resulting in reduced efficacy, particularly if annual boosters are required.

Protein subunit mucosal vaccines may help to address this potential concern of viral vector immunity. In mice, primary i.m. immunization with a SARS-CoV-2 S mRNA vaccine followed by i.n. immunization with recombinant SARS-CoV-2 S induced higher IgA and IgG in bronchoalveolar lavage and nasal wash and provided superior protection from SARS-CoV-2 challenge compared with prime alone (77). Notably, i.n. boosting with a poly(amine–co-ester)-coated mRNA nanoparticles also provided enhanced protection from SARS-CoV-2 challenge (77). Finally, both i.m. prime with i.n. boost strategies induced a systemic immune response comparable to i.m. prime and boosting (77). Similarly, in a macaque model, i.m. immunization with recombinant SARS-CoV-2 S1 followed by i.n. boosting with nanoparticle-enclosed recombinant SARS-CoV-2 S1 provided comparable protection to SARS-CoV-2 challenge as i.m. prime and boost (78). However, this i.m. prime and i.n. boost series induced lower mucosal IgG and IgA titers compared with the i.m. prime and boost series (78). This may indicate weak immunogenicity in i.n. administered, nanoparticle-enclosed S1.

The development of live attenuated SARS-CoV-2 vaccines provides another potential avenue for the induction of mucosal immunity. One such live attenuated vaccine, sCPD9, is the result of rational design and subsequent screening to introduce synonymous codon deoptimization mutations in nsp15 (endonuclease) and nsp16 (2′-O-methyltransferase) (79). This deoptimization is expected to mildly disrupt protein expression. Critically, sCPD9 exhibits attenuated pathogenesis while remaining highly immunogenic (79, 80). When compared with i.m. immunization with Pfizer BNT162b2 or an adenovirus-vectored SARS-CoV-2 S vaccine, sCPD9 i.n. immunization, either as a two-dose regimen or as a booster after BNT162b2 priming, exhibits enhanced protection in hamsters from SARS-CoV-2 challenge and induces higher mucosal IgA and NAb titers than two homologous doses of either BNT162b2 or the adenovirus-vectored S vaccine (81). A similar attenuated vaccine, COVI-VAC, contains synonymous codon deoptimization in the S gene and a deletion in the S furin cleavage site (82). In hamsters, COVI-VAC exhibited substantial attenuation including diminished replication in the olfactory bulb and minimal replication in the lungs. COVI-VAC exhibited comparable protection from wild-type SARS-CoV-2 infection as prior wild-type SARS-CoV-2 infection (82). However, the efficacy of COVI-VAC relative to other approved vaccines remains to be assessed (82). Other live attenuated vaccine candidates include cold-adapted SARS-CoV-2, SARS-CoV-2 containing deletions of open reading frames 3, 6, 7, and 8, as well as SARS-CoV-2 containing a point mutation inactivating nsp16 all exhibit attenuated pathogenicity and induce varying degrees of protection in mouse and hamster models (83–85). These live attenuated SARS-CoV-2 vaccines are promising candidates for the induction of mucosal immunity but require thorough safety testing and must include sufficient attenuating mutations to prevent reversion to a more pathogenic phenotype.

Collectively, these mucosal vaccine strategies may provide more robust sterilizing immunity to block SARS-CoV-2 transmission and diminish the capacity for SARS-CoV-2 immune evasion mutants to emerge. However, given the ability of i.m. mRNA vaccination to boost pre-existing mucosal immunity and the high prevalence of SARS-CoV-2 breakthrough infection, mucosal vaccines may exhibit minimal advantage over current mRNA vaccine platforms in the context of a population with high rates of prior SARS-CoV-2 infection. More clinical data on mucosal vaccine efficacy is needed, including comparisons to approved mRNA vaccines in previously infected subjects and subjects with no prior infection.

A chief concern in the development of next-generation COVID-19 vaccines is the impact that pre-existing immunity may have on the response to vaccination with Ags from more recent SARS-CoV-2 variants (86, 87). Of particular concern is OAS, or the propensity for exposure to a given Ag to preferentially induce an immune response against epitopes conserved with a previous related Ag. OAS was first characterized for influenza A virus when it was noted that exposure to one influenza A virus strain related to a prior infecting strain preferentially induced a B and T cell response that cross-reacted with the earlier strain but provided weak immunity to the new strain (88–90). For novel SARS-CoV-2 vaccine design, overcoming any OAS will be critical given that the general population is no longer immune-naive to SARS-CoV-2. As such, any novel vaccines should be assessed for their ability to overcome OAS. Mechanistically, OAS has been used to refer to both the preferential induction of a memory response over a de novo response (antigenic seniority) or the active suppression of a de novo B cell response against novel epitopes (primary addiction) (Fig. 2).

OAS resulting from exposure to mildly pathogenic human coronaviruses may influence the humoral immune response in COVID-19 patients. It has been demonstrated that in severe COVID-19 patients, SARS-CoV-2 infection led to the induction of Abs targeting the Betacoronaviruses HKU1 and OC43, but not the Alphacoronavirus 229E (91). Notably, this induction of cross-reactive Abs against HKU1 and OC43 was associated with a diminished de novo B cell response against SARS-CoV-2–specific epitopes, particularly affecting neutralizable epitopes (91). Similarly, mice immunized with HKU1 S followed by SARS-CoV-2 S mounted a diminished NAb response against SARS-CoV-2 compared with mice immunized first with SARS-CoV-2 S followed by HKU1 S (92). Taken together, these results suggest that SARS-CoV-2 infection can induce an OAS-like response that preferentially induces Abs that cross-react with seasonal coronaviruses to the detriment of the de novo Ab response against SARS-CoV-2–specific (and often neutralizable) epitopes. Critically in humans, SARS-CoV-2 infection has been shown to induce Abs that cross-react with seasonal coronaviruses more potently than SARS-CoV-2 vaccination (93). This may be related to the disruption of germinal centers during SARS-CoV-2 infection, but not during vaccination (94, 95).

The immune response induced by SARS-CoV-2 variant infection has been shown to be influenced by OAS. Both Alpha and Delta variant infection in unvaccinated individuals have been shown to induce an infecting variant-biased NAb response, with this bias reduced by prior vaccination (96). Notably, Omicron breakthrough infection in vaccinated individuals has been shown to predominantly induce an Ab response that cross-reacts with the ancestral Wuhan-Hu-1 S, whereas Omicron infection in unvaccinated individuals induces an Ab response more specific to the Omicron clade (97–100). Thus, Omicron breakthrough infection induces an OAS-like response with preferential boosting of Abs cross-reactive against prior variants. Critically, the Abs elicited from this “back boosting” exhibit higher levels of somatic hypermutation than Abs specific to the infecting Omicron strain (97, 99). This may indicate that these cross-reactive Abs are the product of induced memory B cells rather than a naive response. Additionally, Omicron breakthrough infection has been shown to preferentially induce nonneutralizing Abs that cross-react with preceding variants (97, 99) likely due to the substantial divergence of the Omicron clade in neutralizable epitopes. However, Omicron breakthrough infection has been established to successfully enhance NAb titers against the Omicron clade (97–100). Thus, the induction of conserved neutralizable epitopes, additional affinity maturation, or the induction of a de novo B cell response despite preferential induction of the memory B cell response are likely occurring.

A recent report from Schiepers et al. (101) examined the impact of OAS on heterologous boosting. Utilizing mice bearing an Igκ L chain tagged with a floxed Flag tag followed by a Strep tag, the study was able to label with a Strep tag those Abs derived from germinal center B cells following primary immunization. However, all other Abs remain Flag tagged, including those derived from germinal center B cells following booster immunization. Thus, during booster immunization, those Abs derived from a memory response induced by the primary immunization (Strep tagged) could be distinguished from those Abs derived from a de novo B cell response (Flag tagged). The authors demonstrate that the de novo B cell response is suppressed by 55-fold during homologous boosting, providing strong evidence for primary addiction. During heterologous boosting, the suppression of the de novo B cell response is dependent on the antigenic distance between the priming and boosting Ag. Specifically, boosting with BA.1 S following a prototype S prime resulted in only a 3.3-fold suppression in the de novo B cell response to heterologous boosting. Critically, this de novo B cell response to BA.1 S boosting was biased toward those epitopes that were divergent between the BA.1 S and the prototype S while the de novo response against conserved epitopes remained suppressed. Notably, these divergent epitopes were largely neutralizable. This epitope specificity in suppression of the de novo B cell response during heterologous boosting is critically important, as it indicates an epitope-specific mechanism for OAS. The authors suggest that “epitope masking” by the initial prototype S prime-induced Abs may diminish the de novo B cell response against conserved epitopes, while permitting a de novo B cell response against novel epitopes. Taken together, these results suggest that heterologous boosting with divergent S proteins can induce a de novo B cell response capable of recovering neutralizing activity against altered epitopes. This may indicate that annual boosters, updated to circulating SARS-CoV-2 variants, can potentially recover neutralizing activity against novel SARS-CoV-2 strains. Notably absent from the study was an investigation of monovalent versus bivalent boosting, which may impact the de novo response against divergent epitopes and requires further study.

The COVID-19 pandemic will likely continue to be a significant public health concern for years to come. Improved and constantly evolving vaccination strategies will be necessary to mitigate the worst impacts of SARS-CoV-2. Induction of a broad and long-lasting NAb response against SARS-CoV-2 is critical to restricting SARS-CoV-2 evolution for immune escape. Furthermore, developing strategies for the induction of pan-coronavirus or pan-sarbecovirus immunity will be important for preventing, or rapidly containing, the next emerging human coronavirus. Key questions remain to be addressed including the capacity for SARS-CoV-2 evolution to evade bNAbs, the ability of novel vaccine candidates to overcome OAS, the durability of novel vaccine-induced immunity, and the relevance of mucosal immunity for long-lived protection from SARS-CoV-2. Finally, although beyond the scope of this Brief Review, the role for T cell immunity, including that elicited by the nucleocapsid protein, in restricting severe disease and the potential breadth of this immunity in novel vaccine candidates remain to be addressed.

S.-L.L. currently serves as a consultant for GlaxoSmithKline. J.P.E. will be employed by GlaxoSmithKline.