SUMMARY
Immune imprinting is a phenomenon in which prior antigenic experiences influence responses to subsequent infection or vaccination1,2. The effects of immune imprinting on serum antibody responses after boosting with variant-matched SARS-CoV-2 vaccines remain uncertain. Here, we characterized the serum antibody responses after mRNA vaccine boosting of mice and human clinical trial participants. In mice, a single dose of a preclinical version of mRNA-1273 vaccine encoding Wuhan-1 spike minimally imprinted serum responses elicited by Omicron boosters, enabling generation of type-specific antibodies. However, imprinting was observed in mice receiving an Omicron booster after two priming doses of mRNA-1273, an effect that was mitigated by a second booster dose of Omicron vaccine. In both SARS-CoV-2 infected or uninfected humans who received two Omicron-matched boosters after two or more doses of the prototype mRNA-1273 vaccine, spike-binding and neutralizing serum antibodies cross-reacted with Omicron variants as well as more distantly related sarbecoviruses. Because serum neutralizing responses against Omicron strains and other sarbecoviruses were abrogated after pre-clearing with Wuhan-1 spike protein, antibodies induced by XBB.1.5 boosting in humans focus on conserved epitopes targeted by the antecedent mRNA-1273 primary series. Thus, the antibody response to Omicron-based boosters in humans is imprinted by immunizations with historical mRNA-1273 vaccines, but this outcome may be beneficial as it drives expansion of cross-neutralizing antibodies that inhibit infection of emerging SARS-CoV-2 variants and distantly related sarbecoviruses.
Multiple vaccines targeting the spike protein of the Wuhan-1 isolate of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) were rapidly developed to control the coronavirus disease 2019 (COVID-19) pandemic. Despite the initial success of these vaccines in limiting symptomatic SARS-CoV-2 infections, the evolution of variant strains with increasing numbers of escape mutations in the spike protein has limited their efficacy3-6 and enabled breakthrough infections in vaccinated individuals7-9. In response to this loss of protection, vaccine boosters encoding variant spike proteins were designed and deployed, and neutralizing antibody responses against the corresponding SARS-CoV-2 variant were increased4,10-13.
Although the goal of variant-matched boosters is to enhance protection against the variant strains, this outcome could be limited by effects of prior SARS-CoV-2 infection or vaccination with spike proteins corresponding to historical strains. Immune imprinting describes the phenomenon in which immunity against a second antigen is shaped by antecedent exposures to a prior, related antigen, and is thought to be mediated by the recall of cross-reactive memory B cells (MBCs) at the expense of stimulating de novo B cell responses. The phenomenon, first termed “original antigenic sin”, described the observation that individuals had higher antibody titers against influenza viruses that they had been exposed to during childhood instead of more contemporary strains14.
Multiple studies have suggested imprinting effects by prior SARS-CoV-2 antigen exposures on subsequent infections with variant strains or immunizations with variant-matched vaccines15-18. For example, in one cohort of triply-vaccinated (BioNTech BNT162b2) healthcare workers, BA.1 (Omicron) infection increased BA.1-reactive antibody responses in the serum of infection-naïve individuals but not in those who had also experienced Wuhan-1 infections15. Studies in animals and humans have shown relatively small differences in variant spike-binding antibody titers between subjects boosted with Wuhan-1-encoding vaccines and variant-matched vaccines4,11-13,19. Imprinting effects on SARS-CoV-2 immunity generally have been studied by analyzing the spike-binding reactivity of B cells via flow cytometry or through B cell repertoire and monoclonal antibody (mAb) analysis11,13,16,18. Rhesus macaques immunized with a primary series of the Wuhan-1 spike-encoding mRNA-1273 vaccine and boosted with B.1.351 (Beta)- or BA.1-matched vaccines showed poor variant-specific MBC responses11,13. The vast majority of mAbs derived from MBCs isolated from human mRNA vaccine recipients primed with the Wuhan-1 spike-encoding vaccine and boosted with the BA.1-matched vaccine were cross-reactive to the Wuhan-1 strain, with few BA.1-specific mAbs (< 1%) detected18. Although these results indicate a strong imprinting effect in the MBC compartment, the serum antibody profile may differ since it reflects contributions from the antibody-secreting cells, including long-lived plasma cells (LLPCs) in the bone marrow, which may not share the same antigen specificities as MBCs20. A key question remains as to whether heterologous booster vaccines induce serum antibody responses that are type-specific to the variant strain or cross-reactive to the original priming strain, and what fraction each of these contributes to neutralization of the variant strain.
Here, we characterized virus-specific and cross-reactive antibodies in the serum of heterologously vaccinated mice and human clinical trial participants using antibody depletion, multiplexed spike-binding detection, and neutralization assays. We evaluated the imprinting effect of antecedent mRNA-1273 vaccines encoding Wuhan-1 spike on mouse and human serum antibody responses after boosting with one or two doses of mRNA vaccines encoding BA.1, BA.5, or XBB.1.5 spike proteins.
RESULTS
Quantification of spike-reactive antibodies.
To measure the reactivity of serum antibodies against multiple spike protein variants of SARS-CoV-2, we developed a flow cytometry-based multiplexed assay. Biotinylated spike proteins of different SARS-CoV-2 strains were bound to streptavidin-coated microparticles with varying intensities of SPHERO™ yellow fluorophore, incubated with serum samples, and the level of antibody binding to each spike protein was determined (Extended Data Fig. 1a-b). To normalize antibody titers acquired across experiments, a mAb, CV3-25, that recognizes a highly conserved epitope in the S2 stem region21 was run in parallel to create a standard curve (Extended Data Fig. 1b-c).
To validate the assay, a panel of mAbs with SARS-CoV-2 spike protein reactivity across the receptor-binding domain (RBD), N-terminal domain (NTD), and S2 stem region were tested (Extended Data Fig. 1d-e). 2B04 bound to Wuhan-1 spike but not the other variant spike proteins tested, consistent with its loss-of-binding to variants with mutations at residue 48422,23. SARS2-71 bound to Wuhan-1 and B.1.351 but not to BA.1 and BA.5 spike proteins, consistent with this mAb losing binding to spike proteins with S477N mutations24. MAb SARS2-38 bound to Wuhan-1, B.1.351, and BA.5 but not BA.1 spike, which harbors a mutation at residue 446, a key contact site24. NTD-binding mAb SARS2-57 bound to Wuhan-1 and BA.5 but not B.1.351 and BA.1 spike proteins, presumably due to the deletions at residues 242 to 244 and 144, respectively25. In comparison, the S2-binding mAbs S2P6 and CC40.826,27 bound to all spike proteins tested. Together, the binding characteristics of the tested mAbs support the utility of the multiplexed spike-binding assay in distinguishing the spike-binding reactivity of antibodies. As an additional validation step, we showed that the anti-spike titers of the mAbs obtained from the multiplexed spike-binding assay correlated with endpoint titers achieved by ELISA (Extended Data Fig. 1f).
Measurement of virus type-specific antibodies in serum.
Although the multiplexed assay can measure antibody binding titers to multiple spike proteins, it cannot discern the fraction of virus-specific and cross-reactive antibodies. We designed a depletion assay to distinguish variant type-specific and Wuhan-1-cross-reactive antibody responses using streptavidin-coated magnetic beads loaded with a biotinylated spike protein of interest (Extended Data Fig. 2a). Binding analysis of the pre-cleared flow through sera (Extended Data Fig. 2b-e) provides quantitative information on the antibodies that are non-reactive to the spike protein on the depletion beads but reactive to the spike proteins on the detection beads. By coupling the assays, we can measure serum antibodies that bind to the spike protein of one SARS-CoV-2 strain but not another.
To validate our assays, we first tested SARS-CoV-2 Wuhan-1 strain-specific (2B04) or Wuhan-1 and B.1.351 cross-reactive (SARS2-38) mAbs. As expected, the binding of SARS2-38 to Wuhan-1 spike-charged beads was abolished when this cross-reactive mAb was pre-cleared with either Wuhan-1 or B.1.351 spike proteins (Extended Data Fig. 2b-c). In comparison, the binding to Wuhan-1 spike-coated beads of mAb 2B04, which does not react with B.1.351 spike protein, was minimally impacted by pre-clearing with B.1.351 spike-coated beads. To verify the ability of the assay to distinguish samples containing mixtures of antibodies of more than one specificity, SARS2-38 and 2B04 were mixed at 99:1 and 90:10 ratios. Depletion with B.1.351 spike protein decreased the Wuhan-1 spike binding titer by 99.2% and 91.2%, respectively, consistent with the expectation that 99% or 90% of the antibody mixture cross-reacts with the B.1.351 spike protein. Thus, our assays can detect the fraction of type-specific antibodies within a mixture with reasonable dynamic range.
To evaluate virus type-specific antibodies in serum using the depletion and multiplexed binding assays, samples from preclinical versions of mRNA-1273 (encodes Wuhan-1 spike) or mRNA-1273.351 (encodes B.1.351 spike) vaccine primed-boosted mice (1273/1273 and 351/351, respectively) were depleted with Wuhan-1 or B.1.351 spike proteins and then measured for binding reactivity. First, we confirmed that sera from mice and humans without a history of SARS-CoV-2 vaccination or infection did not bind to the spike proteins evaluated (Extended Data Fig. 1g). Complete depleting activity by Wuhan-1 and B.1.351 spike proteins was confirmed by loss of serum antibody binding to Wuhan-1 and B.1.351 spike-coated detection beads, respectively (Extended Data Fig. 2d, light grey circles). To determine whether the mRNA-1273 vaccine induced Wuhan-1 spike-specific antibodies, 1273/1273 sera were incubated first with B.1.351 spike-coated depletion beads, and then the pre-cleared sera were evaluated for residual binding to Wuhan-1 spike detection beads. Depletion of 1273/1273 sera with B.1.351 spike proteins led to reduced Wuhan-1 spike-binding reactivity, with only a fraction (approximately 11%) of the Wuhan-1 spike-reactive response not cross-reacting with B.1.351 spike (Extended Data Fig. 2d-e, left panels, 1273/1273 group). As expected, nearly all Wuhan-1-reactive antibodies in 351/351-immunized mice cross-reacted with B.1.351 spike (Extended Data Fig. 2d-e, left panels, 351/351 group). Similarly, B.1.351 spike-specific antibodies were detected in 351/351- but not in 1273/1273-immunized mice. Thus, our coupled depletion and binding assays distinguish type-specific from cross-reactive antibodies in serum samples.
Variant-specific antibodies after heterologous boosters in mice.
To begin to address whether boosting with variant-matched vaccines elicits variant-specific serum antibodies, we examined serum collected from mice that received heterologous mRNA vaccine booster injections. We immunized three cohorts of mice with either one prime/one boost, two primes/one boost, or two primes/two boosts of mRNA vaccines. Groups of 6-week-old female C57BL/6J mice were intramuscularly immunized with preclinical versions of mRNA-1273 vaccine or mRNA-1273.529 vaccine, the latter of which encodes the BA.1 spike. In the one prime/one boost cohort, mice were vaccinated one month apart with homologous (1273/1273 and 529/529) or heterologous (1273/529 and 529/1273) mRNA vaccines. Serum was collected 3 weeks after the second vaccine dose, and IgG responses to Wuhan-1 and BA.1 spike proteins were determined (Fig. 1a). To measure BA.1 spike-specific antibody in 1273/529-immunized mice, sera were depleted with Wuhan-1 spike and assayed for binding to BA.1 spike-coated detection beads. Despite some variation, a fraction (0 to 56%; median 10%) of BA.1-reactive IgG from 1273/529-immunized mice was virus-type-specific and did not cross-react with Wuhan-1 spike (Fig. 1b-c, right panels). Reciprocally, in the 529/1273-vaccinated group, Wuhan-1 spike-specific IgG responses were observed (5 to 17%; median 9%; Fig. 1b-c, left panels). The response to heterologous boosters was minimally imprinted by the priming dose, as statistical differences were not detected in the fraction of Wuhan-1 spike-specific IgG between 1273/1273 (range 13 to 91%; median 36%) and 529/1273 vaccinated mice (p = 0.31), nor in the fraction of BA.1 spike-specific IgG between 529/529 (range 0 to 55%; median 16%) and 1273/529 immunized mice (p > 0.99). Thus, in the context of a one prime/one boost immunization series in mice, boosting with a heterologous mRNA vaccine enabled generation of type-specific antibodies in serum, although most of the post-boost response consisted of cross-reactive antibodies. The fraction of BA.1 type-specific IgG antibodies in 1273/529-vaccinated mice trended lower than that seen with a single 529 immunization (Extended Data Fig. 3a-b), which suggests that prior immunity induced by a single vaccine dose may suppress subsequent type-specific variant responses.
Figure 1. Serum antibody responses after BA.1-matched vaccine boosters in mice.
a, e, i, Schemes of the one prime/one boost (a), two primes/one boost (e), and two primes/two boosts (i) immunization regimens and blood collections. a, Female C57BL/6J mice were immunized twice over a 4-week interval with the indicated mRNA vaccines in b-d (n = 10, two experiments). e, Mice were immunized one month apart and boosted with a third dose of mRNA vaccine three months later (n = 10, two experiments; data in f-h). i, Mice were vaccinated one month apart, boosted with a third dose three months later, and received a fourth dose ten months later (1273/1273/1273/1273 group, n = 10; 529/529/529/529 group, n = 14; 1273/1273/529/529 group, n = 12; two experiments; data in j-l). Sera were collected 3 weeks after the last dose. b, f, j, Antibody binding of sera pre-cleared with empty beads (no Dp), BA.1 spike-loaded beads (BA.1 Dp), or Wuhan-1 spike-loaded beads (Wuhan-1 Dp) to Wuhan-1 (left) and BA.1 (right) spike proteins (connecting lines represent sera from the same mouse; dotted lines show the limit of detection, LOD; values at the LOD are plotted slightly below). c, g, k, Percentages of Wuhan-1 (left) and BA.1 (right) spike-specific antibodies (bars and numbers at the top denote median values). Pie charts illustrate the fraction of type-specific and cross-reactive IgG in each group. d, h, l, Neutralizing activity of pre-cleared sera against VSV pseudoviruses displaying Wuhan-1 (left) or BA.1 (right) spike proteins (numbers above data are geometric mean titers (GMT); fractions indicate the numbers of mice with detectable neutralization by type-specific antibodies). Pie charts illustrate the fraction of neutralization mediated by type-specific and cross-reactive antibodies. Serum samples with an NT50 below the LOD after pre-clearing with empty beads (no DP) were not used in pie chart analyses. c, g, k, Kruskal-Wallis ANOVA with Dunn’s post-test.
To address how virus-type-specific or cross-reactive antibodies contribute to neutralization, we tested pre-cleared sera for their ability to inhibit infection of vesicular stomatitis virus (VSV)-based chimeric SARS-CoV-2 pseudoviruses28 (Fig. 1d). Naïve mouse sera and pre-pandemic human sera showed minimal to no neutralizing activity against the chimeric VSV pseudoviruses used in this study (Extended Data Fig. 1g). In mice immunized with 1273/529, BA.1-specific antibodies neutralized BA.1 pseudovirus and accounted for 20% of the neutralizing activity (Fig. 1d, right panel, NT50 of 102/498). In 529/1273-immunized mice, Wuhan-1-specific antibodies accounted for 44% of the neutralization of Wuhan-1 pseudovirus (Fig. 1d, left panel, NT50 of 99/225). Thus, in the context of a one prime/one boost vaccination scheme in mice, virus-specific antibodies elicited by heterologous booster vaccines can mediate virus neutralization. When we compared BA.1-specific serum neutralization titers from mice immunized with a single 529 vaccine dose to those obtained after a 529 booster (1273/529), in contrast to that seen for type-specific IgG titers (Extended Data Fig. 3b), no fractional difference was observed (529 only: NT50 of 86/417, 21%; Extended Data Fig. 3c), suggesting that both type-specific and cross-reactive serum antibodies contribute to variant neutralization.
In the two primes/one boost vaccination regimen, mice were immunized with a primary two-dose series of the preclinical version of mRNA-1273 one month apart and boosted three months later with a third dose of mRNA-1273 (1273/1273/1273) or mRNA-1273.529 (1273/1273/529) (Fig. 1e). Mice receiving homologous mRNA-1273.529 vaccines (529/529/529) were included for comparison. At 3 weeks after the third vaccination, BA.1-specific antibodies were detected in 78% of 1273/1273/529-vaccinated mice but at low levels (Fig. 1f, right panel), and the fraction of BA.1-specific response was lower in 1273/1273/529-vaccinated mice than in 529/529/529 mice (6% versus 19%, p = 0.034; Fig. 1g, right panel), reflecting imprinting by the primary immunization series. Although neutralization of BA.1 pseudoviruses was observed in all groups, BA.1-specific IgG in 1273/1273/529 mice did not contribute to this inhibition (Fig. 1h, right panel). At 15 weeks after the third vaccine dose, we observed waning of both type-specific and cross-reactive antibodies (Extended Data Fig. 4). The low fractions of BA.1-specific antibodies in heterologously boosted mice at this memory phase suggest that imprinting effects persist over time after booster vaccinations. As compared to the one prime/one boost regimen, serum antibody response in mice receiving an extra priming dose encoding the Wuhan-1 spike showed greater effects of imprinting, suggesting that the number of antecedent antigen exposures shapes the serum antibody response.
As repetitive SARS-CoV-2 spike antigen exposures that occur during breakthrough infections can overcome imprinting and enhance variant-specific responses29, we tested whether this also happens in the context of mRNA vaccination. We immunized the two primes/one boost mice with a fourth vaccine dose of BA.1-matched vaccine (1273/1273/529/529) ten months after the third dose and collected sera 3 weeks later (Fig. 1i). Mice immunized with homologous mRNA-1273.529 (529/529/529/529) vaccines were included for comparison. BA.1-specific IgG titers in 1273/1273/529/529- and 529/529/529/529-immunized mice were comparable (Fig. 1j, right panel), although there was a trend towards a lower fraction of BA.1 type-specific antibodies in the heterologously boosted mice (5% versus 20%, p = 0.10; Fig. 1k, right panel) as was seen in 1273/1273/529- and 529/529/529 immunized mice (Fig. 1g). Whereas all heterologously boosted mice lacked type-specific neutralizing antibodies against BA.1 after the first BA.1-matched vaccine dose (Fig. 1h, right panel), serum from 8 of 12 mice neutralized BA.1 infections with BA.1-specific antibodies after the second matched vaccine dose (Fig. 1l, right panel). Thus, in mice boosted with two doses of BA.1-matched vaccines, imprinting by antecedent vaccines is mitigated such that variant-specific antibodies in serum contribute more to the neutralizing response.
Variant-specific antibodies after a third or fourth vaccine dose.
We next analyzed sera from individuals participating in clinical trials of B.1.351, BA.1, and BA.5 variant-matched vaccines (NCT04405076 and NCT04927065) that had no prior SARS-CoV-2 infection history as determined by negative anti-nucleocapsid (anti-N) antibody tests. In one study, participants were immunized with a primary two-dose series of 100 μg of mRNA-1273 and boosted at least six months later with 50 μg of mRNA-1273 (1273/1273/1273), mRNA-1273.351 (1273/1273/351), or a bivalent mRNA-1273.211 vaccine composed of a 1:1 ratio of mRNA-1273 and mRNA-1273.351 (1273/1273/211) (Extended Data Fig. 5a). Although Wuhan-1 spike-specific antibodies were detected in all cohorts, none of the booster vaccines induced appreciable B.1.351 spike-specific antibody responses (Extended Data Fig. 5b-c). Correspondingly, neutralizing activity by Wuhan-1-specific antibodies against Wuhan-1 pseudovirus was detected in all groups, but neutralization of B.1.351 pseudovirus was not observed by B.1.351 spike-specific antibodies (Extended Data Fig. 5d). A greater Wuhan-1-specific response was observed in 1273/1273/211 than in 1273/1273/351 participants, likely due to the inclusion of the Wuhan-1 spike-encoding mRNA in the mRNA-1273.211 booster. Despite the limited sample size, these results suggest that the binding and neutralization of B.1.351 seen across all three vaccination groups are conferred principally by cross-reactive antibodies. Moreover, the lack of B.1.351-specific antibodies in heterologously boosted participants suggests that responses were shaped by the primary vaccination series.
In a second study, individuals received two 100 μg and one 50 μg doses of mRNA-1273 and were boosted at least six months later with a fourth 50 μg dose of mRNA-1273.214, a bivalent vaccine composed of mRNA-1273 and mRNA-1273.529 (1273/1273/1273/214) (Fig. 2a). Sera were collected immediately before and one month after the fourth dose. Although binding reactivities against Wuhan-1 and BA.1 spike proteins were boosted by the bivalent vaccine by 2.8- and 3.1-fold, respectively (p = 0.0020 and p = 0.0039; Fig. 2b), we detected Wuhan-1 spike-specific and Wuhan-1/BA.1 cross-reactive but not BA.1 spike-specific antibodies (Fig. 2c). The fraction of Wuhan-1 spike-specific antibodies decreased slightly after boosting with the bivalent vaccine (24 to 19%, p = 0.023), likely due to expansion of cross-reactive antibodies (Fig. 2d). Functional analysis of pre- and post-boost sera revealed increases in neutralizing activity against Wuhan-1 and BA.1 pseudoviruses by 2.8- and 1.8-fold, respectively (p = 0.0020; Fig. 2e). Whereas inhibition of Wuhan-1 pseudovirus infection was conferred by both Wuhan-1 spike-specific and cross-reactive antibodies, neutralization of BA.1 pseudovirus was conferred exclusively by cross-reactive antibodies (Fig. 2f). Thus, in humans, a Wuhan-1/BA.1 bivalent mRNA booster vaccine following three doses of mRNA-1273 induced cross-reactive but not BA.1 spike-specific antibodies. Consistent with these data, when we compared spike-binding properties of MBCs from the blood of the same individuals before and after (approximately 1 or 3 months) the bivalent mRNA-1273.214 vaccine boost, we observed no increase in the fraction of BA.1 spike-specific cells (Extended Data Fig. 6).
Figure 2. Low levels of Omicron-specific antibodies in humans boosted with a fourth dose of BA.1- or BA.5-matched vaccines.
a, g, Schemes of immunizations and blood draws. Clinical trial participants received a primary two-dose series of 100 μg of mRNA-1273, a third dose of 50 μg of mRNA-1273, and then were boosted at least six months later with 50 μg of Wuhan-1/BA.1 bivalent vaccine (mRNA-1273.214) (a, n = 10; data in b-f) or Wuhan-1/BA.5 bivalent vaccine (mRNA-1273.222) (g, n = 15; data in h-l). Sera were collected immediately before and one month after the fourth dose. Subjects with no prior SARS-CoV-2 infection history (anti-N negative) were selected for analysis. b, h, Paired analysis of pre- and post-boost serum antibody reactivity against Wuhan-1 and BA.1 (b) or BA.5 (h) spike proteins. c, i, Antibody binding to Wuhan-1, BA.1 (c), or BA.5 (i) spike proteins of pre- and post-boost sera pre-cleared with empty beads (no Dp), BA.1 spike-loaded beads (BA.1 Dp), BA.5 spike-loaded beads (BA.5 Dp), or Wuhan-1 spike-loaded beads (Wuhan-1 Dp) (connecting lines represent sera from the same individual; dotted lines show the LOD; values at the LOD are plotted slightly below). d, j, Percentages of Wuhan-1 (top), BA.1 (d, bottom), and BA.5 (j, bottom) spike-specific antibodies before and after the fourth immunization (bars and numbers at the top denote median values). Pie charts illustrate the fraction of type-specific and cross-reactive antibodies. e, k, Paired analysis of pre- and post-boost serum neutralizing activity against Wuhan-1, BA.1 (e), and BA.5 (k) pseudoviruses. f, l, Neutralizing activity of pre-cleared sera against Wuhan-1 (top), BA.1 (f, bottom), and BA.5 (l, bottom) pseudoviruses (numbers above data are GMTs; fractions indicate the numbers of individuals with detectable neutralization by type-specific antibodies). Pie charts illustrate the fraction of neutralization by type-specific and cross-reactive antibodies. Serum samples with an NT50 below the LOD after pre-clearing with empty beads (no DP) were not used in pie chart analyses. Statistical analysis was performed using a two-sided Wilcoxon signed-rank test.
We also evaluated a separate human clinical trial group that was boosted with a fourth 50 μg dose of mRNA-1273.222, a Wuhan-1/BA.5 bivalent vaccine (1273/1273/1273/222) (Fig. 2g). Sera were collected immediately before and one month after the fourth dose. This bivalent vaccine increased binding titers against Wuhan-1 and BA.5 spike proteins by 4.3- and 5.1-fold, respectively (p < 0.0001; Fig. 2h). Whereas Wuhan-1 spike-specific and cross-reactive antibodies were observed before and after boosting, BA.5 spike-specific antibodies were detected in only 2 of 15 individuals and at low levels after boosting (Fig. 2i, bottom panel). The fraction of Wuhan-1 spike-specific antibodies in serum decreased (25 to 18%, p = 0.0010), suggesting an expansion of cross-reactive antibodies (Fig. 2j). Neutralizing titers against Wuhan-1 and BA.5 pseudoviruses were increased by 5.6- and 4.4-fold, respectively, after boosting (p < 0.0001; Fig. 2k). Whereas Wuhan-1 pseudovirus was neutralized by both Wuhan-1 spike-specific and cross-reactive antibodies, BA.5 pseudovirus was inhibited entirely by cross-reactive antibodies (Fig. 2l). Overall, administration of one dose of the Wuhan-1/BA.5 bivalent mRNA vaccine following three doses of mRNA-1273 expanded cross-reactive antibodies and induced minimal levels of BA.5-specific antibodies.
Virus-specific antibodies in humans receiving a fifth vaccine dose.
Omicron subvariant XBB.1.5 arose in December of 2022 and soon became prevalent in most parts of the world30. A clinical trial (NCT04927065) was designed to examine the immunogenicity of two XBB.1.5-matched vaccines, a bivalent BA.5/XBB.1.5 vaccine (mRNA-1273.231) and a monovalent XBB.1.5 vaccine (mRNA-1273.815). We analyzed serum samples of individuals immunized with a primary series of two 100 μg doses of mRNA-1273, boosted with a 50 μg dose of mRNA-1273, a 50 μg dose of mRNA-1273.222 (Wuhan-1/BA.5 bivalent vaccine), and then finally a 50 μg dose of BA.5/XBB.1.5 bivalent vaccine (1273/1273/1273/222/231, Fig. 3a) or XBB.1.5 monovalent vaccine (1273/1273/1273/222/815, Fig. 4a). The primary vaccine series was given one month apart, and all boosters were given at least six months after the prior immunization. Subjects with or without a history of natural SARS-CoV-2 infection (anti-N positive or negative, respectively) were analyzed separately. Sera were collected immediately before and one month after the fifth dose.
Figure 3. Humans administered a BA.5/XBB.1.5 bivalent vaccine after a previous BA.5-matched booster develop cross-neutralizing antibodies.
a, Scheme of immunizations and blood draws. Clinical trial participants received a primary two-dose mRNA-1273 series, a third dose of mRNA-1273, a fourth dose of mRNA-1273.222 (Wuhan-1/BA.5 bivalent vaccine), and a fifth dose of mRNA-1273.231 (BA.5/XBB.1.5 bivalent vaccine) (n = 10). Sera were collected immediately before and one month after the fifth dose. Subjects with no prior SARS-CoV-2 infection history were selected for analysis. b, Paired analysis of pre- and post-boost serum antibody reactivity against Wuhan-1 (top), BA.5 (middle), and XBB.1.5 (bottom) spike proteins. c, Antibody binding to Wuhan-1, BA.5, and XBB.1.5 spike proteins of pre-cleared pre- and post-boost sera. No Dp, depletion with empty beads; BA.5 Dp, depletion with BA.5 spike-loaded beads; BA.5 & XBB.1.5 Dp, depletion with a 1:1 mixture of BA.5-loaded and XBB.1.5-loaded beads; Wuhan-1 Dp, depletion with Wuhan-1 spike-loaded beads (connecting lines represent sera from the same individual; dotted lines show the LOD; values at the LOD are plotted slightly below). d, Percentages of Wuhan-1-specific (BA.5- and XBB.1.5-non-reactive), BA.5-specific (Wuhan-1-non-reactive), and XBB.1.5-specific (Wuhan-1-non-reactive) antibodies before and after the fifth immunization (bars and numbers at the top denote median values). Pie charts illustrate the fraction of type-specific and cross-reactive antibodies. e, Paired analysis of pre- and post-boost serum neutralizing activity against Wuhan-1, BA.5, and XBB.1.5 pseudoviruses. f, Neutralizing activity of pre-cleared sera against Wuhan-1, BA.5, and XBB.1.5 pseudoviruses (numbers above data are GMTs; fractions indicate the numbers of individuals with detectable neutralization by type-specific antibodies). Pie charts illustrate the fraction of neutralization by type-specific and cross-reactive antibodies. Serum samples with an NT50 below the LOD after pre-clearing with empty beads (no DP) were not used in pie chart analyses. Statistical analysis was performed using a two-sided Wilcoxon signed-rank test.
Figure 4. Humans administered an XBB.1.5 monovalent vaccine after a previous BA.5-matched booster develop cross-neutralizing antibodies.
a, Scheme of immunizations and blood draws. Clinical trial participants received a primary two-dose mRNA-1273 series, a third dose of mRNA-1273, a fourth dose of mRNA-1273.222 (Wuhan-1/BA.5 bivalent vaccine), and a fifth dose of mRNA-1273.815 (XBB.1.5 monovalent vaccine) (n = 14). Sera were collected immediately before and one month after the fifth dose. Subjects with no prior SARS-CoV-2 infection history (anti-N negative) were selected for analysis. b, Paired analysis of pre- and post-boost serum antibody reactivity against Wuhan-1 and XBB.1.5 spike proteins. c, Antibody binding to Wuhan-1 and XBB.1.5 spike proteins of pre- and post-boost sera pre-cleared with empty beads (no Dp), XBB.1.5 spike-loaded beads (XBB.1.5 Dp), or Wuhan-1 spike-loaded beads (Wuhan-1 Dp) (connecting lines represent sera from the same individual; dotted lines show the LOD; values at the LOD are plotted slightly below the LOD). d, Percentages of Wuhan-1 and XBB.1.5 spike-specific antibodies before and after the fifth immunization (bars and numbers at the top denote median values). Pie charts illustrate the fraction of type-specific and cross-reactive antibodies. e, Paired analysis of pre- and post-boost serum neutralizing activity against Wuhan-1 and XBB.1.5 pseudoviruses. f, Neutralizing activity of pre-cleared sera against Wuhan-1 and XBB.1.5 pseudoviruses (numbers above data are GMTs; fractions indicate the numbers of individuals with detectable neutralization by type-specific antibodies). Pie charts illustrate the fraction of neutralization by type-specific and cross-reactive antibodies. Serum samples with an NT50 below the LOD after pre-clearing with empty beads (no DP) were not used in pie chart analyses. Statistical analysis was performed using a two-sided Wilcoxon signed-rank test.
Binding reactivity against Wuhan-1, BA.5, and XBB.1.5 spike proteins in infection-naïve participants was boosted by the bivalent BA.5/XBB.1.5 vaccine (11.6-, 5.7-, and 6.6-fold, respectively, p = 0.0020, Fig. 3b). Over ~90% of Wuhan-1 spike-reactive IgG before and after the fifth dose cross-reacted with the BA.5 spike protein, as pre-clearing with BA.5 spike-coated beads decreased Wuhan-1 spike reactivity by approximately 10-fold (Fig. 3c). To determine whether the residual Wuhan-1 spike-reactive antibodies were Wuhan-1-specific or cross-reactive to XBB.1.5 spike, we co-depleted sera with BA.5 and XBB.1.5 spike proteins (Fig. 3c-d). The remaining Wuhan-1 spike-reactive antibodies were type-specific and did not cross-react with BA.5 or XBB.1.5. Despite the increased serum antibody reactivity to BA.5 and XBB.1.5 spike proteins after the BA.5/XBB.1.5 bivalent booster, BA.5 or XBB.1.5 spike-specific antibodies were not appreciably induced. The BA.5/XBB.1.5 bivalent vaccine also enhanced neutralizing responses against Wuhan-1, BA.5, and XBB.1.5 pseudoviruses (2.6-, 6.1-, and 4.6-fold, respectively; p = 0.0039, 0.0020, and 0.0020, Fig. 3e). Depletion analysis revealed that over 70% of serum neutralizing activity against Wuhan-1 was mediated by cross-reactive antibodies (Fig. 3f), whereas the inhibition of BA.5 and XBB.1.5 infections was conferred entirely by antibodies that cross-reacted with Wuhan-1 spike. These data suggest an imprinting effect of the antecedent mRNA vaccines, in which two exposures to Omicron subvariant spike proteins after three exposures to the Wuhan-1 spike resulted in expansion of cross-reactive, cross-neutralizing antibodies that target the XBB.1.5 variant in participants not previously infected with SARS-CoV-2. In humans, this regimen was insufficient to elicit appreciable levels of Omicron-specific antibodies.
We next evaluated serum from individuals that received a fifth dose of mRNA-1273.815, the currently administered monovalent mRNA vaccine booster targeting the XBB.1.5 spike. In subjects without evidence of a prior SARS-CoV-2 infection, antibody titers against Wuhan-1 and XBB.1.5 spike proteins were increased by 4.5- and 5.7-fold, respectively (p = 0.00012, Fig. 4b). The fraction of Wuhan-1 spike-specific antibodies decreased after the booster dose (27 to 12%, p = 0.00012), again consistent with expansion of cross-reactive antibodies (Fig. 4c-d). XBB.1.5 spike-specific antibodies were absent before mRNA-1273.815 boosting and detected in only 3 of 14 individuals at low levels after boosting, indicating that the increased reactivity to XBB.1.5 spike was principally mediated by cross-reactive antibodies. The XBB.1.5 monovalent booster also augmented neutralizing activity against Wuhan-1 and XBB.1.5 pseudoviruses, by 3.0- and 6.6-fold, respectively (p = 0.00037 and p = 0.00012, Fig. 4e). Whereas a fraction of Wuhan-1 neutralization before and after boosting was conferred by Wuhan-1-specific antibodies, neutralizing activity against XBB.1.5 was mediated entirely by antibodies that cross-reacted with Wuhan-1 spike (Fig. 4f). Together, these data suggest that monovalent XBB.1.5-matched mRNA vaccine given as a fifth dose boosted cross-reactive neutralizing antibodies that were shaped by the original mRNA-1273 vaccine in infection-naïve individuals.
As vaccination-induced antibody responses might be affected by antecedent SARS-CoV-2 infections15,29, we also examined the sera of subjects with a history of SARS-CoV-2 infection (anti-N positive) prior to receiving the XBB.1.5 monovalent vaccine. Antibody titers against Wuhan-1 and XBB.1.5 spike proteins were increased by 3.1- and 5.1-fold, respectively (p = 0.00006, Extended Data Fig. 7b). The fraction of Wuhan-1 spike-specific antibodies decreased after the booster dose (40 to 15%, p = 0.00006), consistent with expansion of cross-reactive antibodies (Extended Data Fig. 7c-d). XBB.1.5 spike-specific antibodies were absent before mRNA-1273.815 boosting and detected in only 6 of 28 individuals and at low levels after boosting, suggesting that the increased reactivity to XBB.1.5 spike was principally mediated by cross-reactive antibodies. Neutralizing activity against Wuhan-1 and XBB.1.5 pseudoviruses was increased by 2.9- and 10.0-fold, respectively (p = 0.00006, Extended Data Fig. 7e). Whereas Wuhan-1 neutralization before and after boosting was conferred by both Wuhan-1-specific and cross-reactive antibodies, neutralization against XBB.1.5 was mediated almost entirely by antibodies that cross-reacted with Wuhan-1 spike (Extended Data Fig. 7f). Overall, comparable results were observed in subjects with or without (Fig. 4) a SARS-CoV-2 infection history, suggesting that prior infections did not alter the imprinting effect of the original mRNA-1273 vaccines in human clinical cohorts.
Breadth of cross-reactive neutralizing antibody responses.
To begin to define the epitopes recognized by cross-neutralizing antibodies from mRNA-1273.815-boosted individuals (Fig. 4a), we developed a bead-based mAb competition assay (Extended Data Fig. 8a). XBB.1.5 spike-coated beads were incubated first with serum samples and then, without washing, fluorescently tagged neutralizing mAbs that bind to mapped epitopes were added for detection by flow cytometry. We examined binding of 8 published and unpublished XBB.1.5 cross-reactive neutralizing mAbs that recognize previously defined RBD antigenic sites31 (class 1: rCOV2-3731; class 3: S309, rCOV2-3872, rCOV2-3967, and rCOV2-3889; class 4: rCOV2-4030) and the S2 stem helix region (CC40.8 and CV3-25)21,27,32. Competition was evaluated by measuring reductions in the mean fluorescence intensity associated with mAb binding. Naïve sera from individuals with no prior SARS-CoV-2 infection or vaccination history were included as negative controls. Serum cross-reactive antibodies from most 1273/1273/1273/222/815-vaccinated individuals bound at or near all tested neutralizing epitopes in XBB.1.5 RBD and S2 and were boosted to similar extents (Extended Data Fig. 8b-c). This analysis suggests that the XBB.1.5 monovalent vaccine booster induced antibodies that recognize multiple cross-reactive and neutralizing epitopes on the spike protein.
Given that some neutralizing epitopes on the SARS-CoV-2 spike protein are evolutionarily conserved33,34, we hypothesized that antibodies from the serum of XBB.1.5-boosted individuals might neutralize other circulating SARS-CoV-2 variants and more distantly related coronaviruses. Accordingly, we tested these sera for inhibitory activity against pseudoviruses encoding the spike protein of Omicron lineage variants EG.5.1 (95.7, 89.7, 99.0, and 96.5% sequence identity in NTD, RBD, S2, and spike compared to Wuhan-1) and BA.2.86 (92.2, 88.3, 99.3, and 95% sequence identity in NTD, RBD, S2, and spike compared to Wuhan-1), as well as pangolin coronavirus Guangdong (Pang/GD, 67.9, 96.4, 97.6, and 89.8% sequence identity in NTD, RBD, S2, and spike compared to SARS-CoV-2 Wuhan-1), severe acute respiratory syndrome coronavirus 1 (SARS-CoV-1, 47.4, 73.1, 88.4, and 76.0% sequence identity in NTD, RBD, S2, and spike compared to SARS-CoV-2 Wuhan-1), and Middle East respiratory syndrome coronavirus (MERS-CoV, 16.3, 18.1, 38.0, and 28.6% sequence identity in NTD, RBD, S2, and spike compared to SARS-CoV-2 Wuhan-1) (Extended Data Fig. 9). We also modified our depletion assay to determine the contribution to neutralization of cross-reactive antibodies that engaged the NTD, RBD, or S2 protein. We first compared the effect of depletion of Wuhan-1 NTD, RBD, or S2 targeting antibodies on neutralization of Wuhan-1 or XBB.1.5 pseudoviruses (Fig. 5a-b and Extended Data Fig. 10a). In individuals with no SARS-CoV-2 infection history, neutralization of Wuhan-1 pseudovirus was detected in serum before and after XBB.1.5-boosting, and this activity was mediated principally by anti-RBD antibodies (12.6-fold reduction in GMT after RBD depletion of post-boost sera) compared to anti-NTD (1.8-fold reduction) or anti-S2 (no reduction) antibodies (Fig. 5a, third left panel). In comparison, pre-boost serum had poor inhibitory activity against XBB.1.5. However, after boosting, serum neutralized XBB.1.5 pseudovirus infection, and this activity was abolished in sera pre-cleared with Wuhan-1 RBD, suggesting that inhibitory activity against XBB.1.5 was mediated exclusively by anti-RBD antibodies (Fig. 5b, third left panel). Comparable results were observed in individuals with a history of SARS-CoV-2 infection, as individuals who received an XBB.1.5 vaccine booster had neutralizing antibodies against Wuhan-1 and Omicron lineage variants. While the inhibitory activity was completely pre-cleared by Wuhan-1 spike (Fig. 5a-c, right panels, and Extended Data Fig. 10b and 11a), for XBB.1.5, EG.5.1, and BA.2.86, neutralizing antibodies were reduced but not abrogated by pre-clearing with RBD in vaccinated, infected individuals.
Figure 5. Neutralizing activity of vaccine-derived serum against distantly related sarbecoviruses.
1273/1273/1273 post-boost sera (first left panels, described in Extended Data Fig. 5), 1273/1273/1273/222 post-boost sera (second left panels, described in Fig. 2), or pre- and post-boost sera relative to immunization with mRNA-1273.815 (XBB.1.5 monovalent vaccine) of SARS-CoV-2 non-infected (third left panels, described in Fig. 4) and infected (right panels, described in Extended Data Fig. 7) individuals were pre-cleared with Wuhan-1 NTD, RBD, S2, RBD together with S2 (RBD & S2), or spike protein. Pre-cleared sera were tested for neutralizing activity against pseudoviruses expressing Wuhan-1 (a), XBB.1.5 (b), EG.5.1 (c), Pang/GD (d), or SARS-CoV-1 (e) spike proteins (bars and numbers at the top are GMTs; dotted lines show the LOD; values at the LOD are plotted slightly below the LOD; individuals that became infected (anti-N antibody negative before the fourth vaccine dose and positive before the fifth vaccine dose) are indicated with red borders in the right panels).
Although serum obtained before XBB.1.5 boosting of infection-naïve individuals had inhibitory activity against Pang/GD, neutralizing titers were increased 3.2-fold (p = 0.037) at one month after boosting with the XBB.1.5 mRNA vaccine (Fig. 5d, third left panel). In comparison, individuals who received multiple doses of only mRNA-1273 showed little to no inhibitory activity against Pang/GD one month after the last dose (Fig. 5d, first left panel), suggesting that heterologous variant boosting expands the pool of broadly cross-neutralizing antibodies. Like results with XBB.1.5, EG.5.1, and BA.2.86, neutralization of Pang/GD was mediated almost entirely by Wuhan-1 RBD-targeting antibodies in most infection-naïve individuals. SARS-CoV-1, a more distantly related clade 1 sarbecovirus, was poorly neutralized by serum obtained before XBB.1.5 boosting or one month after multiple mRNA-1273 immunizations (Fig. 5e, first and third left panels). One month after XBB.1.5 mRNA vaccine boosting, neutralization of SARS-CoV-1 was observed in serum from non-infected or infected subjects, and this activity was mediated by spike-reactive antibodies imprinted by the mRNA-1273 vaccines (Fig. 5e, third left panel and right panel). Pre-clearing of sera with Wuhan-1 RBD only partially reduced SARS-CoV-1 neutralization, suggesting a less dominant role of RBD-targeting antibodies in the protection against this more distantly related coronavirus. However, for the non-infected population, depletion of serum antibodies with both Wuhan-1 RBD and S2 proteins reduced SARS-CoV-1 neutralization almost to baseline, indicating significant contributions of both anti-RBD and anti-S2 antibodies to neutralization (Extended Data Fig. 10c). Although boosting with XBB.1.5 enhanced neutralization against two other sarbecoviruses, inhibitory activity was not detected against the even more distantly related merbecovirus, MERS-CoV (Extended Data Fig. 11b). Similar trends were observed in individuals with a history of SARS-CoV-2 infection, as inhibitory activity was observed against Pang/GD and SARS-CoV-1 but not MERS-CoV (Fig. 5d-e and Extended Data Fig. 11b, right panels). Neutralization activity against these sarbecoviruses was completely pre-cleared by Wuhan-1 spike. However, pre-clearing with Wuhan-1 RBD or S2 had less of an effect, as higher residual neutralizing activity was apparent in vaccinated, infected individuals. Collectively, these results suggest that the XBB.1.5 monovalent vaccine booster promoted neutralizing activity against multiple sarbecoviruses in vaccine recipients with or without a SARS-CoV-2 infection history, and this inhibitory activity is imprinted by antibodies targeting the Wuhan-1 RBD and S2 regions.
DISCUSSION
We profiled the type-specificity and cross-reactivity of vaccine-induced serum antibodies, which have been shown to correlate with protection against infection and disease35. In mice, the induction of BA.1-specific neutralizing antibodies by BA.1-matched boosters was limited by imprinting, but this effect varied depending on the number of antigen exposures with the original and the boosting antigen. In humans with or without a history of SARS-CoV-2 infection, we detected little to no variant-specific serum antibody response in clinical trial participants primed with two or three doses of mRNA-1273 and boosted with B.1.351-, BA.1-, BA.5-, or XBB.1.5-matched monovalent or bivalent vaccines, suggesting strong imprinting effects of antecedent mRNA vaccinations.
Imprinting after SARS-CoV-2 infection and vaccination has been suggested by studies that analyze the differential reactivity of serum and MBCs to different spike proteins4,11-13,15-19. Our data showing that the spike binding and neutralizing response against Omicron strains is almost entirely cross-reactive with the Wuhan-1 spike suggest that human serum antibody responses are imprinted by antecedent mRNA vaccinations. A prior study that performed RBD-based antibody depletion reported similar findings in the serum of individuals boosted with Wuhan-1/BA.5 bivalent mRNA vaccines36. Our results are relevant to a recent study, which suggested that repeated Omicron infections could override ancestral SARS-CoV-2 immune imprinting29. In that publication, individuals received two doses of the inactivated whole virus CoronaVac vaccine and then experienced two Omicron breakthrough infections. Differences in the immune stimulus of the initial vaccine or subsequent Omicron exposures could dictate whether cross-reactive memory B cells or de novo type-specific naïve B cells are activated to produce antibodies. Consistent with our findings, two recent studies also described substantial imprinting of human serum antibody responses after boosting with XBB-based vaccines37,38.
Our experiments with human samples showed a lack of preferential expansion of BA.1 spike-specific MBCs after Wuhan-1/BA.1 bivalent mRNA vaccine boosting. B cell repertoire analysis in humans also showed few (< 1%) BA.1-specific MBCs in blood of individuals boosted with the BA.1-matched monovalent mRNA vaccine after priming with multiple doses of mRNA vaccines encoding Wuhan-1 spike protein18. While the MBC repertoire does not always reflect the serum antibody composition due to possible discrepancies in the antigen specificity between MBCs and LLPCs20, in the context of SARS-CoV-2 mRNA vaccine boosting, substantial imprinting effects likely occur in both memory cell compartments. Although secondary germinal centers are formed after boosting and include newly recruited naïve B cells39, for most individuals who had multiple mRNA vaccine exposures, antibodies derived from these de novo activated B cells contributed to a minority of the serum response against Omicron variants. This result might differ if the initial antigen exposure is immunologically weaker, or the subsequent exposure is substantially stronger.
A fate mapping study in mice primed once with Wuhan-1 spike mRNA vaccine and heterologously boosted twice with BA.1 spike mRNA vaccine identified de novo serum antibodies derived from newly recruited B cells that had not participated in the original germinal center reaction40. Our experiments in mice also show that while repeated administration of the mRNA-1273 vaccine promotes imprinting, administration of multiple boosters encoding variant spike proteins can mitigate imprinting and yield type-specific neutralizing antibodies in serum. It remains to be determined how many doses of variant-matched vaccines (or breakthrough infections) are required to overcome mRNA vaccine-induced imprinting to elicit type-specific antibody responses in humans that contribute to the serum neutralizing activity. The difference in imprinting response between mice and humans may reflect multiple variables including the priming and boosting intervals, the number of doses of mRNA-1273 vaccine administered, and immunodominance hierarchies. Notably, increasing the dose of antigen in booster doses enhanced antibody titers against specific influenza virus antigens41, although it remains unclear if this was due to greater induction of variant type-specific or cross-reactive antibodies.
The cross-reactive antibodies expanded by boosting exhibited broad spike-reactivity and mediated neutralization of the priming strain, the boosting strain, subsequent circulating Omicron variants, and even distantly related sarbecoviruses. In this context, immune imprinting by SARS-CoV-2 mRNA vaccines may not constitute original antigenic “sin”. Although the sequence identity of spike protein between Wuhan-1 and XBB.1.5 strains (96.7%) versus Wuhan-1 and Pang/GD (89.8%) or SARS-CoV-1 (76.0%) is greater, some individuals had detectable yet low levels of neutralizing antibody against Pang/GD and SARS-CoV-1 but little or none against XBB.1.5, EG.5.1, and BA.2.86, prior to XBB.1.5 mRNA vaccine boosting; this result is consistent with studies42 showing that post-COVID-19 populations have a high prevalence of cross-reactive antibodies to spike proteins from all Orthocoronavirinae genera. Thus, despite the greater spike sequence identity between Wuhan-1 and Omicron lineage variants than Pang/GD and SARS-CoV-1, the antigenic distance is greater. This result is likely because Omicron strains have evolved key substitutions that evade neutralizing antibody responses induced by Wuhan-1 spike directed vaccines.
Our results suggest that boosting with XBB.1.5 mRNA vaccine expands cross-reactive antibodies that bind multiple neutralizing epitopes. More precise mapping of cross-neutralizing epitopes could be informed by landscape mutagenesis43,44 or polyclonal antibody epitope mapping and cryo-electron microscopy45,46. Our depletion studies with Wuhan-1 NTD, RBD, and S2 suggest that the class of cross-neutralizing antibody may differ depending on the sarbecovirus. Whereas neutralizing responses against Omicron variants and Pang/GD in uninfected individuals were largely depleted by RBD, those against SARS-CoV-1 required pre-clearing with both RBD and S2 proteins to reduce inhibitory responses. A history of SARS-CoV-2 infection altered this pattern: whereas neutralizing responses against Omicron variants, Pang/GD, and SARS-CoV-1 were depleted by Wuhan-1 spike, less of an effect was observed with RBD and S2. This discrepancy might be because natural infection can also select for epitopes exposed on virions compared to vaccine-induced cell-associated spike proteins.
We note several limitations in our study. (a) We characterized the serum antibody responses following immunization with mRNA but no other platform vaccines; (b) We examined the boost response at a 1-month time point in most cohorts. Variant-specific responses could emerge at later time points after sustained germinal center reactions47; and (c) Our human studies lacked a control group that was immunized only with Omicron lineage vaccines. Inherent differences in immunodominance to Omicron spike could affect induction of type-specific neutralizing antibodies, as described for earlier variants48. Future studies with samples from immunized young children might allow assessment of type-specific responses from this lineage.
In summary, our study shows that in humans, variant-matched mRNA boosters administered after mRNA-1273 vaccinations primarily expand Wuhan-1-cross-reactive antibodies. Despite this, heterologous mRNA vaccine boosters induced cross-reactive antibodies with broadly neutralizing activity that likely contribute to their continued efficacy against severe SARS-CoV-2 disease. Intriguingly, administration of future boosters that continue to drift from the original Wuhan-1 spike might expand broadly neutralizing antibodies against multiple and even more distantly related sarbecoviruses.
METHODS
Cells and viruses.
Vero-TMPRSS2 cells49 were cultured at 37°C in Dulbecco’s Modified Eagle medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 10 mM HEPES pH 7.3, 1 mM sodium pyruvate, 1X nonessential amino acids, 100 U/mL of penicillin–streptomycin, and 5 μg/mL of blasticidin.
Chimeric VSV pseudoviruses expressing spike proteins corresponding to the ancestral SARS-CoV-2 Wuhan-1 virus, variant strains B.1.351, BA.1, BA.5, XBB.1.5, EG.5.1 and BA.2.86, and other coronaviruses including Pang/GD, SARS-CoV-1, and MERS-CoV were produced according to published methods28. Briefly, the spike genes of different coronaviruses including SARS-CoV-2 (Wuhan-1, GenBank MN908947.3; B.1.351, GenBank OX008568.1; BA.1, GenBank OR321077.1; BA.5, GenBank QY979334.1; XBB.1.5, GenBank OP790748.1; EG.5.1, GenBank OX463749.1; and BA.2.86, GenBank OR461132), Pang/GD (GenBank MT799524.1), SARS-CoV-1 (GenBank JN854286.1), and MERS-CoV (GenBank MK357909.1) were synthesized (Integrated DNA Technologies) and inserted into an infectious molecular clone of VSV50,51 as described previously28. Modifications to the cytoplasmic tail (21 amino acid deletion) of the spike genes were generated to promote secretion. Other VSV N, P, L and G expression plasmids have been described51.
Recovery of recombinant VSV was performed as described51. BSRT7/5 cells52 were inoculated with vaccinia virus vTF7-3 and subsequently transfected with T7-expression plasmids encoding VSV N, P, L, and G, and an antigenomic copy of the viral genome. Cell culture supernatants were collected at 56-72 h, clarified by centrifugation at 1,000 x g for 5 min, and filtered through a 0.22 μm filter. Virus was plaque-purified on Vero CCL81 cells in the presence of 25 μg/mL of cytosine arabinoside (Sigma-Aldrich), and plaques in agarose plugs were amplified on Vero CCL81 cells. Viral stocks were amplified on MA104 cells (ATCC, CRL-2378.1) at a multiplicity of infection (MOI) of 0.01 in Medium 199 containing 2% FBS and 20 mM HEPES pH 7.7 at 34°C. Viral supernatants were harvested upon extensive cytopathic effect and clarified of cell debris by centrifugation at 1,000 x g for 5 min. Aliquots were stored at −80°C.
Animal experiments.
Animal studies were carried out in accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. Protocols were approved by the Institutional Animal Care and Use Committee at the Washington University School of Medicine (assurance number A3381–01). Six-week-old female C57BL/6J mice purchased from The Jackson Laboratory (Cat. #000664) were used for all experiments. Animals were housed in groups of five and fed standard chow diets. Photoperiod was 12 h on:12 h off dark/light cycle. The ambient animal room temperature was 70°F, controlled within ± 2°F. The room humidity was 50%, controlled within ± 5%. Mice described in Extended Data Fig. 2d-e were primed via intraperitoneal injection with 0.25 μg and boosted with 1 μg of mRNA vaccines one month apart, and serum was collected 3 weeks later. Mice described in Fig. 1 and Extended Data Fig. 4 were immunized according to the depicted schemes. The primary vaccines were given 4 weeks apart, the third dose vaccine was given 3 months later, and the fourth dose vaccine was given 10 months after the third dose. All vaccines given to mice are preclinical materials and were administered via intraperitoneal injection at 0.25 μg mRNA vaccine per mouse. Serum samples were collected 3 (Fig. 1) or 15 (Extended Data Fig. 4) weeks after the final immunization. In vivo studies were not blinded, and mice were randomly assigned to vaccination groups. No sample-size calculations were performed to power each animal study. Instead, sample sizes were determined based on prior in vivo immunization experiments.
Clinical study design and human participants.
Human samples were obtained from open-label phase 2/3 studies (NCT04405076 and NCT04927065) that evaluate the immunogenicity, safety, and reactogenicity of mRNA booster vaccines in adults (18 years and older) who had already received 2-dose primary series (100 μg) of mRNA-1273. Participants, with informed consent, were enrolled in a sequential, non-randomized manner and received boosters of 50 μg of mRNA-1273, mRNA-1273.351, or mRNA-1273.211 (cohort 1, Extended Data Fig. 5), 50 μg of mRNA-1273 followed by 50 μg of mRNA-1273.214 or mRNA-1273.222 (cohort 2, Fig. 2), or 50 μg of mRNA-1273 followed by 50 μg of mRNA-1273.222 and then 50 μg of mRNA-1273.231 (cohort 3, Fig. 3), or 50 μg of mRNA-1273 followed by 50 μg of mRNA-1273.222 and then 50 μg of mRNA-1273.815 (cohort 4, Fig. 4).
For participants in trial NCT04927065 boosted with a fifth dose of mRNA-231 or mRNA-1273.815, median (interquartile ranges) intervals were 8.2 (7.8-9.8) and 9.0 (7.7-11.5) months between the 2nd and 3rd doses (both mRNA-1273), 9.8 (8.3-10.3) and 9.2 (8.2-10.3) months between the 3rd and 4th doses (mRNA-1273 and mRNA-1273.222, respectively), and 8.3 (8.1-8.4) and 8.2 (8.1-8.3) months between the 4th and 5th doses, respectively. Interim 29-day analysis results are reported.
The trials are being conducted across 23 sites in the United States of America, in accordance with the International Council for Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use, Good Clinical Practice guidelines. The central Institutional Review Board/Ethics Committee (Advarra, Inc., 6100 Merriweather Drive, Columbia, MD 21044) approved the protocol and consent forms. All participants provided written informed consent. Eligible participants included healthy male and female adults >18 years of age.
Viral antigens.
Recombinant soluble spike proteins (spike, RBD, and NTD) from Wuhan-1, B.1.351, BA.1, BA.5, and XBB.1.5 SARS-CoV-2 strains were expressed as described53,54. Recombinant proteins were produced in Expi293F cells (ThermoFisher, Cat. #A14635) by transfection of DNA using the ExpiFectamine 293 Transfection Kit (ThermoFisher, Cat. #A14524). Supernatants were harvested 4-5 days post-transfection, and recombinant proteins were purified using Strep-Tactin XT resin (IBA Lifesciences, Cat. #2-5010), then buffer exchanged into phosphate-buffered saline (PBS) and concentrated using Amicon Ultracel centrifugal filters (EMD Millipore, Cat. #UFC905096). SARS-CoV-2 (Wuhan-1 strain) S2 protein (Hexapro-SS-2W) included residues 697–1208 of the spike with an artificial signal peptide, proline substitutions at positions 817, 892, 899, 942, 986, and 987, cysteine substitutions at positions 704 and 790, a glutamate substitution at 957, tryptophan substitutions at 991 and 998, and a C-terminal foldon trimerization motif, HRV 3C cleavage site, 8X His tag and Twin-Strep-tag55. Soluble Hexapro-SS-2W protein was expressed and purified as described55 in FreeStyle 293F cells (ThermoFisher, Cat. #R79007).
For multiplexed bead-binding assays and antibody depletion assays, recombinant spike proteins were biotinylated with EZ-Link™ NHS-PEG4-Biotin (ThermoFisher, Cat. # A39259) for 2 h at 4°C and processed through Zeba spin desalting columns (ThermoFisher) to remove excess unbound biotin.
ELISA.
Purified recombinant Wuhan-1, B.1.351, BA.1, or BA.5 (2 μg/mL) spike proteins were coated onto 96-well Maxisorp plates (ThermoFisher, Cat. #439454) in 50 μL of 50 mM sodium carbonate pH 9.6 (coating buffer) overnight at 4°C. Coating buffer was removed, and wells were washed 3 times with PBS + 0.05% Tween-20 (PBST) and blocked with PBST + 2% BSA (blocking buffer) for 1 h at room temperature. Sera were serially diluted in blocking buffer and added to the plates. Plates were incubated for 1 h at 37°C and then washed 3 times with PBST, followed by addition of 50 μL of 1:750 anti-mouse IgG-HRP (Sigma-Aldrich, Cat. #A5278) in blocking buffer. Following a 1 h incubation at 37°C, plates were washed 3 times with PBST and added with 50 μL of 1-Step Ultra TMB-ELISA (ThermoFisher, Cat. #34028). Reactions were stopped by the addition of 50 μL of 2 M sulfuric acid. Optical density measurements were taken at 450 nm using a microplate reader (BioTek). The endpoint serum dilution was calculated with curve fit analysis of optical density (OD) values for serially diluted sera with a cut-off value set to mean plus three times the standard deviation of the background signal.
Multiplexed spike-binding assay.
Recombinant biotinylated spike proteins (spike, RBD, NTD, S2) from Wuhan-1, B.1.351, BA.1, BA.5, and XBB.1.5 strains of SARS-CoV-2 were incubated with different fluorescence intensity peaks of the SPHERO streptavidin coated fluorescent yellow particles (Spherotech, Cat. #SVF3-2552) at 20 ng per μg beads for 30 min at room temperature on an end-over-end mixer. Free biotin (Avidity, Cat. #BIO200) was added to the beads at 5 μM and incubated for 15 min at room temperature. Beads of different intensity and loaded with different spike proteins were then pooled and washed with PBS supplemented with 2% FBS, 2 mM ethylenediaminetetraacetic acid (EDTA), and 0.1% sodium azide (FACS buffer), mixed with serially diluted pre-cleared serum samples or mAbs, and incubated for 30 min at room temperature. Beads were washed twice with FACS buffer and stained with a mixture of anti-mouse IgG Alexa Fluor™ 647 (ThermoFisher, Cat. #A-21235, 1:3000) and anti-human IgG Alexa Fluor™ 647 (ThermoFisher, Cat. #A-21445, 1:3000) for 10 min at room temperature in dark. After two washes with FACS buffer, beads were resuspended in FACS buffer and acquired on flow cytometer iQue®3 (Sartorius). Population gating and analysis of fluorescence intensity were performed with iQue Forecyt 9.0 software. The average of geometric mean fluorescence intensity of no-antibody control wells was defined as the background signal. Background subtracted fluorescence intensity of samples was divided by background values to obtain the fold change and interpolated with CV3-25 (spike and S2), SARS2-38 (RBD), or SARS2-57 (NTD) standard curves performed in duplicates. Interpolated values were multiplied by the sample dilution factor, and the final binding titers were shown as ng/mL equivalents of CV3-25, SARS-38, or SARS2-57.
Antibody depletion assay.
Streptavidin magnetic beads (BioLabs, Cat. #S1420S) were washed twice with PBST and incubated with recombinant biotinylated spike proteins (spike, RBD, NTD, S2) from Wuhan-1, B.1.351, BA.1, BA.5, or XBB.1.5 strains of SARS-CoV-2 in PBS at 30 μg per mg beads for 30 min at room temperature on an end-over-end mixer. Beads were washed twice with PBS and incubated with diluted sera samples for 30 min at room temperature with agitation. Depleted sera were collected on the KingFisher Flex extraction robot (ThermoFisher) and stored at 4°C.
Pseudovirus neutralization assay.
Vero-TMPRSS2 cells were seeded into 96 well plates in DMEM supplemented with 10% FBS without blasticidin. The next day, depleted serum samples or PBS were serially diluted with DMEM supplemented with 2% FBS and mixed with VSV pseudoviruses and incubated for 1 h at 37°C. Antibody-virus complexes were added to Vero-TMPRSS2 cell monolayers and incubated for 1 h at 37°C. Subsequently, cells were overlaid with 1% (w/v) methylcellulose in minimum essential medium (MEM). Plates were harvested 12 h (MERS), 20 h (Wuhan-1, B.1.351, BA.5, XBB.1.5, EG.5.1, and BA.2.86), 26 h (BA.1 and SARS-CoV-1), or 30h (Pang/GD) later by removing overlays and fixed with 4% paraformaldehyde in PBS for 20 min at room temperature. Plates were washed and sequentially incubated with anti-VSV nucleoprotein antibody (1:5,000 dilution, Sigma-Aldrich, Cat. #MABF2348) or anti-VSV matrix protein antibody (1:5,000 dilution, Sigma-Aldrich, Cat. #MABF2347) and subsequently with HRP-conjugated goat anti-mouse IgG (1:750 dilution, Sigma-Aldrich, Cat. #A5278) in PBS supplemented with 0.1% saponin and 0.1% bovine serum albumin. Pseudovirus-infected cell foci were visualized using TrueBlue peroxidase substrate (KPL, Cat. #5510-0050) and quantitated on an ImmunoSpot analyzer (Cellular Technology).
B cell probe binding.
Cryopreserved peripheral blood mononuclear cells (PBMCs) were thawed, washed briefly with RPMI supplemented with 4% heat inactivated newborn calf serum, and incubated with aqua live/dead viability dye (ThermoFisher) for 20 min at room temperature. Cells were stained with the following antibodies (monoclonal unless indicated) for 20 min at room temperature: IgD FITC (1:160, goat polyclonal, Southern Biotech), IgM PerCP-Cy5.5 (1:20, clone G20-127, BD Biosciences), IgA Dylight 405 (1:80, goat polyclonal, Jackson Immunoresearch Inc), CD20 BV570 (1:40, clone 2H7, Biolegend), CD27 BV650 (1:20, clone O323, Biolegend), CD14 BV785 (1:80, clone M5E2, Biolegend), CD16 BUV496 (1:80, clone 3G8, BD Biosciences), CD4 BUV737 (1:320, clone SK3, BD Biosciences), CD19 APC (1:80, clone J3-119, Beckman), IgG Alexa 700 (1:20, clone G18-145, BD Biosciences), CD3 APC-Cy7 (1:40, clone SP34-2, BD Biosciences), CD38 PE (1:160, clone OKT10, Caprico Biotechnologies), CD21 PE-Cy5 (1:40, clone B-ly4, BD Biosciences) and CXCR5 PE-Cy7 (1:40, clone MU5UBEE, ThermoFisher). Stained cells were then incubated with streptavidin-BV605 (BD Biosciences) labeled BA.1 S-2P (1:6.25), streptavidin-BUV661 (BD Biosciences) labeled S-2P WA-1 (1:25), and either streptavidin-BUV395 labeled S-2P BA.5 (1:25) or streptavidin-BUV395 labeled S-2P BQ.1.1 (1:25) for 30 min at 4°C (protected from light). Cells were washed and fixed in 0.5% formaldehyde (Tousimis Research Corp) prior to data acquisition. All antibodies were titrated on human PBMC to determine the optimal concentration. Samples were acquired on an BD FACSymphony cytometer and analyzed for B cell phenotype using FlowJo version 10.9.0.
Production of mAbs.
To produce mAbs recombinantly, ~15 to 30 mL ExpiCHO cell cultures were transfected using the Gibco ExpiCHO Expression System (ThermoFisher) as described by the vendor’s protocol. MAbs were purified from clarified cell culture supernatants using HiTrap MabSelect SuRe columns (Cytiva, Cat. #GE11-0034-95) on a 24-column parallel protein chromatography system (Protein BioSolutions). Purified mAbs were buffer exchanged into 1X Dulbecco's phosphate-buffered saline (DPBS), concentrated using Amicon Ultra 50 KDa-cutoff centrifugal filter units (Millipore Sigma), and stored at 4°C until use.
MAb competition assay.
Recombinant biotinylated XBB.1.5 spike protein was incubated with SPHERO streptavidin coated fluorescent yellow particles (Spherotech, Cat. #SVF3-2552) at 20 ng per μg beads for 30 min at room temperature on an end-over-end mixer. Free biotin (Avidity, Cat. #BIO200) was added to the beads at 5 μM and incubated for 15 min at room temperature. Beads were then washed with FACS buffer, mixed with serially diluted serum samples, and incubated for 30 min at room temperature followed by 5 min at 4°C. Without washing, mAbs (final concentrations of 0.002 μg/mL of rCOV2-3731, rCOV2-3872, rCOV2-3967, rCOV2-3889, or rCOV2-4030, 1 μg/mL of S309, or 0.1 μg/mL of CC40.8 or CV3-25) that were pre-labeled with Alexa Fluor 647 (using Zip Alexa Fluor™ Rapid Antibody Labeling Kit (ThermoFisher, Cat. #Z11235)) were added to the beads and incubated for 30 min at room temperature in dark. After two washes with FACS buffer, beads were resuspended in FACS buffer and acquired on flow cytometer iQue®3 (Sartorius). Population gating and analysis of fluorescence intensity were performed with iQue Forecyt 9.0 software. The average geometric mean fluorescence intensity (GMT) of no mAb control wells was defined as the background signal. Background-subtracted fluorescence intensity of samples was divided by the average GMT of no serum control wells to obtain % mAb binding.
Statistical analysis.
P values were determined using GraphPad Prism version 10. Tests (Mann-Whitney test with Bonferroni correction; Kruskal-Wallis ANOVA with Dunn’s post-test; Wilcoxon signed-rank test), number of animals, numbers of humans, median or mean values, and statistical comparison groups are indicated in the Figure legends.
Extended Data
Extended Data Figure 1. Multiplexed spike-binding detection assay.
a, Gating strategy for the multiplexed spike-binding assay. After gating on singlets, beads pre-bound with the spike protein of different SARS-CoV-2 strains were separated based on their intensity of SPHERO™ yellow fluorophore. b, Representative flow cytometry plots of mAb CV3-25 binding to Wuhan-1, B.1.351, BA.1, and BA.5 spike-charged detection beads at the indicated antibody concentrations. c, A CV3-25 standard curve of binding to Wuhan-1, B.1.351, BA.1, and BA.5 spike-charged detection beads (pooled from two experiments, error bars denote standard deviations [SD]). d, Literature-based SARS-CoV-2 spike binding characteristics and escape mutations of the indicated mAbs to Wuhan-1, B.1.351, BA.1, and BA.5 spike proteins. CHK-265 (anti-CHIKV E2) was included as a negative control. +, positive binding; -, negative binding; amino acid substitutions indicate the mutations in the corresponding SARS-CoV-2 strain that results in loss of mAb binding. e, Representative flow cytometry plots of the indicated mAbs to Wuhan-1, B.1.351, BA.1, and BA.5 spike-charged detection beads (three experiments). f, Pearson’s correlation between the spike-binding and ELISA endpoint titers of the mAbs evaluated in d-e. R2 and p values are indicated. g, Antibody binding (top panels) and neutralizing activity (bottom panels) of naïve mouse sera and pre-pandemic human sera to Wuhan-1, BA.1, BA.5, and XBB.1.5 spike proteins and chimeric VSV pseudoviruses (n = 4; one experiment). Positive controls of serum collected from a mouse immunized with two doses of mRNA-1273 followed by two doses of mRNA-1273.529 (encodes BA.1 spike protein) and serum collected from a SARS-CoV-2 non-infected clinical trial participant immunized with three doses of mRNA-1273 and boosted with mRNA-1273.222 (Wuhan-1/BA.5 bivalent vaccine) are included (bars indicate mean values; dotted lines show the LOD; values at the LOD are plotted slightly below the LOD for visualization). −, negative SARS-CoV-2 exposure history; +, positive SARS-CoV-2 exposure history.
Extended Data Figure 2. Antibody depletion assay.
a, Schematic of the anti-spike antibody depletion assay. Dp, depletion. b-c, Wuhan-1 spike-binding and antibody titers obtained after interpolating with CV3-25 standard curves of mAbs SARS2-38 and 2B04 mixed at the indicated ratios and pre-cleared with empty depletion beads (no Dp), B.1.351 spike-loaded beads (B.1.351 Dp), or Wuhan-1 spike-loaded beads (Wuhan-1 Dp) (numbers at the top in b indicate the derived fraction of Wuhan-1 spike-specific antibody; three experiments; mean values ± SDs are shown, and dotted lines show the LOD). d, Antibody binding of 3-week post-boost sera harvested from mice primed with 0.25 μg and boosted one month later with 1 μg of mRNA-1273 (1273/1273) or mRNA-1273.351 (351/351) (n = 10, two experiments) to Wuhan-1 and B.1.351 spike proteins. Sera were pre-cleared with empty beads, B.1.351 spike-loaded beads, or Wuhan-1 spike-loaded beads (connecting lines represent sera from the same mouse; dotted lines show the LOD; values at the LOD are plotted slightly below the LOD for visualization). e, Percentages of Wuhan-1 (left) and B.1.351 (right) spike-specific antibodies, derived from d. The fraction of Wuhan-1 spike-specific IgG is calculated by dividing the Wuhan-1 spike-binding titer of B.1.351 spike depleted sera (blue circles in d left panel) by that without depletion (dark grey circles in d left panel) from the same individual (B.1.351 Dp/no Dp) (type-specific titers at the LOD in d are designated 0% type-specific response in e; bars and numbers at the top indicate median values). e, Two-sided Mann-Whitney test with Bonferroni correction (p = 0.000011).
Extended Data Figure 3. Effect of heterologous vaccine boosting on type-specific serum antibody binding and neutralizing responses.
Sera were collected 26 days post-vaccination from mice immunized with a single dose of mRNA-1273 (1273) or mRNA-1273.529 (529) (n = 10, two experiments), or 21 days after the last vaccine dose from mice vaccinated with the one prime/one boost (1273/529, n = 10), two primes/one boost (1273/1273/529, n = 9), or two primes/two boosts (1273/1273/529/529, n = 12; also described in Fig. 1) regimens. All vaccines were given at 0.25 μg mRNA per dose. 1273, mRNA-1273; 529, mRNA-1273.529. a, Antibody binding of sera pre-cleared with empty beads (no Dp), BA.1 spike-loaded beads (BA.1 Dp), or Wuhan-1 spike-loaded beads (Wuhan-1 Dp) to Wuhan-1 (left) and BA.1 (right) spike proteins (connecting lines represent sera from the same mouse; LOD for the single dose and the one prime/one boost cohorts are 1,600 ng/mL equivalents (eq, derived from a mAb standard curve); LOD (dotted lines) for the two primes/one boost and two primes/two boosts cohorts are 3,200 ng/mL eq; values at the LOD are plotted slightly below the LOD for visualization). b, Percentages of Wuhan-1 (left) and BA.1 (right) spike-specific antibodies (bars and numbers at the top denote median values). Pie charts illustrate the fraction of type-specific and cross-reactive IgG in each group. c, Neutralizing activity of pre-cleared sera against VSV pseudoviruses displaying Wuhan-1 (left) or BA.1 (right) spike proteins (numbers above data points indicate the GMT; fractions in parentheses at the top indicate the numbers of mice with detectable neutralization by type-specific antibodies; LOD for the single dose and the one prime/one boost cohorts are 1/25 serum dilution; LOD (dotted lines) for the two primes/one boost and two primes/two boosts cohorts are 1/50 serum dilution). Pie charts illustrate the fraction of neutralization mediated by type-specific and cross-reactive antibodies in each group. Serum samples with an NT50 below the LOD after pre-clearing with empty beads (no DP) were not used in pie chart analyses. b, Kruskal-Wallis ANOVA with Dunn’s post-test.
Extended Data Figure 4. Persistent effects of immune imprinting in the serum of mice receiving heterologous mRNA vaccines.
a, Scheme of immunization regimen and blood draw. Mice were immunized one month apart and boosted with a third dose of mRNA vaccine three months later (n = 10, two experiments, also in Fig. 1e). Sera were collected 15 weeks after the last dose. 1273, mRNA-1273; 529, mRNA-1273.529. b, Antibody binding of sera pre-cleared with empty beads (no Dp), BA.1 spike-loaded beads (BA.1 Dp), or Wuhan-1 spike-loaded beads (Wuhan-1 Dp) to Wuhan-1 (left) and BA.1 (right) spike proteins (connecting lines represent sera from the same mouse; dotted lines show the LOD; values at the LOD are plotted slightly below the LOD). c, Percentages of Wuhan-1 (left) and BA.1 (right) spike-specific antibodies (bars and numbers at the top denote median values). Pie charts illustrate the fraction of type-specific and cross-reactive IgG in each group. d, e, Paired analysis of 3- and 15-week post-boost undepleted sera (d, 15-week data also shown in b "no Dp") and type-specific (e, 15w data also shown in b "BA.1 Dp” or “Wuhan-1 Dp”) antibody reactivity against Wuhan-1 (left) and BA.1 (right) spike proteins (numbers on the top indicate GMT fold reduction). c, Kruskal-Wallis ANOVA with Dunn’s post-test.
Extended Data Figure 5. Absence of B.1.351 spike-specific antibodies in humans receiving B.1.351-matched booster as a third-dose vaccine.
a, Scheme of immunizations and blood draw. Participants received two doses of 100 μg of mRNA-1273 one month apart and were boosted with 50 μg of the indicated mRNA vaccine at least six months later. Sera were collected one month after the third dose (n = 5). 1273, mRNA-1273; 351, mRNA-1273.351; 211, mRNA-1273.211 (Wuhan-1/B.1.351 bivalent vaccine). b, Antibody binding to Wuhan-1 (top) and B.1.351 (bottom) spike proteins of sera pre-cleared with empty beads (no Dp), B.1.351 spike-loaded beads (B.1.351 Dp), or Wuhan-1 spike-loaded beads (Wuhan-1 Dp) (connecting lines represent sera from the same individual; dotted lines show the LOD; values at the LOD are plotted slightly below the LOD). c, Percentages of Wuhan-1 (top) and B.1.351 (bottom) spike-specific antibodies (bars and numbers at the top denote median values). Pie charts illustrate the fraction of type-specific and cross-reactive IgG in each group. d, Neutralizing activity of pre-cleared sera against Wuhan-1 (top) and B.1.351 (bottom) pseudoviruses (numbers above data points indicate the GMT; fractions in parentheses at the top indicate the numbers of individuals with detectable neutralization by type-specific antibodies). Pie charts illustrate the fraction of neutralization mediated by type-specific and cross-reactive antibodies. Serum samples with an NT50 below the LOD after pre-clearing with empty beads (no DP) were not used in pie chart analyses. c, Kruskal-Wallis ANOVA with Dunn’s post-test.
Extended Data Figure 6. Imprinting effects on MBCs after boosting with bivalent mRNA-1273.214 vaccine.
a, Scheme of immunizations and blood draws. b, Gating scheme for identification of spike-positive MBCs in blood. c-d, Study participants (n = 20) who previously received three doses of mRNA-1273 were boosted with mRNA-1273.214 (Wuhan-1/BA.1 bivalent vaccine), and peripheral blood mononuclear cells (PBMCs) were obtained on days 0 (pre-boost), 28 (post-boost), and 90 (post-boost). c, Representative flow cytometry dot plots showing MBC specificities for pairs of fluorescently labeled WA-1/2020 and BA.1 spike proteins. Quadrant frequencies represent percentage of total MBCs. d, Fractions of spike-binding MBCs with indicated specificities (WA-1+BA.1+, left; WA-1+BA.1−, middle; or WA-1−BA.1+, right) at days 0, 28, and 90 (numbers at the top denote median values). d, Kruskal-Wallis ANOVA with Dunn’s post-test.
Extended Data Figure 7. Humans with a SARS-CoV-2 infection history administered an XBB.1.5 monovalent vaccine after a previous BA.5-matched booster develop cross-neutralizing antibodies in serum.
a, Scheme of immunizations and blood draws, also described in Fig. 4. Clinical trial participants received a primary two-dose mRNA-1273 series, a third dose of mRNA-1273, a fourth dose of mRNA-1273.222 (Wuhan-1/BA.5 bivalent vaccine), and a fifth dose of mRNA-1273.815 (XBB.1.5 monovalent vaccine) (b-d post-boost, n = 28; b-d pre-boost and e-f, n = 15). Sera were collected immediately before and one month after the fifth dose. Subjects with a history of SARS-CoV-2 infection (anti-N antibody positive before the mRNA-1273.815 vaccination) were selected for analysis. 1273, mRNA-1273; 222, mRNA-1273.222; 815, mRNA-1273.815. b, Paired analysis of pre- and post-boost serum antibody reactivity against Wuhan-1 and XBB.1.5 spike proteins. c, Antibody binding to Wuhan-1 and XBB.1.5 spike proteins of pre- and post-boost sera pre-cleared with empty beads (no Dp), XBB.1.5 spike-loaded beads (XBB.1.5 Dp), or Wuhan-1 spike-loaded beads (Wuhan-1 Dp) (connecting lines represent sera from the same individual; dotted lines show the LOD; values at the LOD are plotted slightly below the LOD). d, Percentages of Wuhan-1 and XBB.1.5 spike-specific antibodies before and after the fifth immunization (bars and numbers at the top denote median values). Pie charts illustrate the fraction of type-specific and cross-reactive antibodies. e, Paired analysis of pre- and post-boost serum neutralizing activity against Wuhan-1 and XBB.1.5 pseudoviruses. f, Neutralizing activity of pre-cleared sera against Wuhan-1 and XBB.1.5 pseudoviruses (numbers above data points indicate the GMT; fractions in parentheses at the top indicate the numbers of individuals with detectable neutralization by type-specific antibodies). Pie charts illustrate the fraction of neutralization mediated by type-specific and cross-reactive antibodies. Serum samples with an NT50 below the LOD after pre-clearing with empty beads (no DP) were not used in pie chart analyses. All statistical analysis was performed using a two-sided Wilcoxon signed-rank test.
Extended Data Figure 8. XBB.1.5 monovalent booster increases serum binding to multiple neutralizing epitopes on the XBB.1.5 spike protein.
a, Schematic of the mAb competition assay (competing serum antibody, red; non-competing serum antibody, salmon; fluorescence-labeled mAb, blue). b, Percent binding of the indicated mAbs competed by serially diluted pre- and post-boost serum of XBB.1.5 monovalent vaccine boosted individuals described in Fig. 4 to XBB.1.5 spike protein. Two naïve sera (black curves, no SARS-CoV-2 infection or vaccination history) were included as control. c, Competition titer of pre- and post-boost sera that inhibits 50% binding of the indicated mAbs (numbers on the top indicate GMT fold increases), derived from b.
Extended Data Figure 9. Spike protein sequence alignment.
Multiple sequence alignment of SARS-CoV-1 (P59594, UNIPROT), Pang/GD (EPI_ISL_410721, GISAID), XBB.1.5 (EPI_ISL_16713058, GISAID), and Wuhan-1 (EPI_ISL_402124, GISAID) spike proteins. The Wuhan-1 spike is shown in the last row with relative variant sequence changes indicated. The color ribbons beneath the sequence correspond to the specific regions of the spike protein (NTD, green; RBD, blue; S2, pink).
Extended Data Figure 10. Depletion of Wuhan-1 NTD, RBD, S2, and spike-reactive antibodies from serum.
Antibody binding to Wuhan-1 NTD, RBD, S2, or spike proteins of pre- and post-boost sera of 1273/1273/1273/222/815-vaccinated individuals without (a) or with (b) a history of SARS-CoV-2 infection after pre-clearing with empty beads (no Dp), Wuhan-1 N-terminal domain protein-loaded beads (NTD Dp), RBD-loaded beads (RBD Dp), S2 protein-loaded beads (S2 Dp), a 1:1 mixture of RBD-loaded and S2-loaded beads (RBD & S2 Dp), or spike-loaded beads (spike Dp) (dotted lines show the LOD; values at the LOD are plotted slightly below the LOD; individuals that became infected (anti-N antibody negative before the fourth dose and positive before the fifth dose) are indicated with red borders; all other individuals in b are anti-N positive before the fourth dose). Binding to Wuhan-1 NTD, RBD, and S2 was interpolated from SARS2-57, SARS2-38, and CV3-25 standard curves, respectively. c, Serum neutralizing activity of infection-naïve post-boost sera pre-cleared with empty beads (no Dp), Wuhan-1 RBD-loaded beads (RBD Dp), or a mixture of Wuhan-1 RBD-loaded and S2-loaded beads (RBD & S2 Dp) against SARS-CoV-1 pseudovirus (data also shown in Fig. 5e, third panel) (connecting lines represent sera from the same individual). Statistical analysis: two-sided Wilcoxon signed-rank test.
Extended Data Figure 11. Neutralizing activity of vaccine-derived serum against Omicron BA.2.86 and MERS-CoV.
As described in Fig. 5, 1273/1273/1273 post-boost sera (first left panels, described in Extended Data Fig. 5), 1273/1273/1273/222 post-boost sera (second left panels, described in Fig. 2), or pre- and post-boost sera relative to immunization with mRNA-1273.815 (XBB.1.5 monovalent vaccine) of SARS-CoV-2 non-infected (third left panels, described in Fig. 4) and infected (right panels, described in Extended Data Fig. 7) individuals were pre-cleared with Wuhan-1 NTD, RBD, S2, RBD together with S2 (RBD & S2), or spike protein. Pre-cleared sera were tested for neutralizing activity against pseudoviruses expressing BA.2.86 (a) or MERS-CoV (b) spike proteins (bars and numbers at the top are GMTs; dotted lines show the LOD; values at the LOD are plotted slightly below; individuals that became infected (anti-N antibody negative before the fourth vaccine dose and positive before the fifth vaccine dose) are indicated with red borders in the right panels).
Supplementary Material
Acknowledgements.
This study was supported by the National Institutes of Health (R01 AI157155, NIAID Centers of Excellence for Influenza Research and Response (CEIRR) contracts 75N93021C00014 and 75N93019C00051, to M.S.D.), the Intramural Research Program of the Vaccine Research Center, NIAID, and a sponsored Research Agreement with Moderna. We thank S. Mackin and B. Ying for assistance with some of the animal studies, L. Lu for help with production of some of the chimeric VSV pseudoviruses, and I-T. Teng for the S-2P probes used with the MBC assay. Some of the Figure graphics were created using BioRender.com.
Footnotes
Competing interests. M.S.D. is a consultant or advisor for Inbios, Vir Biotechnology, IntegerBio, Moderna, Merck, and GlaxoSmithKline. The Diamond laboratory has received additional unrelated funding support in sponsored research agreements from Vir Biotechnology, Emergent BioSolutions, and IntegerBio. G.A.A., S.M.E, and D.K.E. are employees of and shareholders in Moderna, Inc. J.E.C. has served as a consultant for Luna Biologics, is a member of the Scientific Advisory Boards of CompuVax and Meissa Vaccines and is Founder of IDBiologics. The Crowe laboratory at Vanderbilt University Medical Center has received sponsored research agreements from Takeda Vaccines, IDBiologics, and AstraZeneca. Vanderbilt University has applied for a patent concerning antibodies that are related to this work (U.S. Provisional Patent Application No. 63/513,255). All other authors declare no competing interests.
Data and reagent availability. All data supporting the findings of this study are available within the paper or its Extended Data files. Any additional information related to the study also is available from the corresponding author upon reasonable request. All reagents are available through a Material Transfer Agreement.
MAIN TEXT REFERENCES
- 1.Cobey S & Hensley SE Immune history and influenza virus susceptibility. Curr Opin Virol 22, 105–111 (2017). 10.1016/j.coviro.2016.12.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Knight M, Changrob S, Li L & Wilson PC Imprinting, immunodominance, and other impediments to generating broad influenza immunity. Immunol Rev 296, 191–204 (2020). 10.1111/imr.12900 [DOI] [PubMed] [Google Scholar]
- 3.Garcia-Beltran WF et al. Multiple SARS-CoV-2 variants escape neutralization by vaccine-induced humoral immunity. Cell 184, 2372–2383. e2379 (2021). 10.1016/j.cell.2021.03.013 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Choi A. et al. Safety and immunogenicity of SARS-CoV-2 variant mRNA vaccine boosters in healthy adults: an interim analysis. Nature medicine 27, 2025–2031 (2021). 10.1038/s41591-021-01527-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Pajon R. et al. SARS-CoV-2 Omicron Variant Neutralization after mRNA-1273 Booster Vaccination. New England Journal of Medicine 386, 1088–1091 (2022). 10.1056/NEJMc2119912 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Gruell H. et al. mRNA booster immunization elicits potent neutralizing serum activity against the SARS-CoV-2 Omicron variant. Nature medicine 28, 477–480 (2022). 10.1038/s41591-021-01676-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Buchan SA et al. Estimated effectiveness of COVID-19 vaccines against Omicron or Delta symptomatic infection and severe outcomes. JAMA network open 5, e2232760–e2232760 (2022). 10.1001/jamanetworkopen.2022.32760 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Christensen PA et al. Signals of significantly increased vaccine breakthrough, decreased hospitalization rates, and less severe disease in patients with coronavirus disease 2019 caused by the omicron variant of severe acute respiratory syndrome coronavirus 2 in Houston, Texas. The American journal of pathology 192, 642–652 (2022). 10.1016/j.ajpath.2022.01.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Tan CY et al. Protective immunity of SARS-CoV-2 infection and vaccines against medically attended symptomatic omicron BA. 4, BA. 5, and XBB reinfections in Singapore: a national cohort study. The Lancet Infectious Diseases 23, 799–805 (2023). 10.1016/s1473-3099(23)00060-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Wu K. et al. Variant SARS-CoV-2 mRNA vaccines confer broad neutralization as primary or booster series in mice. Vaccine 39, 7394–7400 (2021). 10.1016/j.vaccine.2021.11.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Corbett KS et al. Protection against SARS-CoV-2 beta variant in mRNA-1273 vaccine–boosted nonhuman primates. Science 374, 1343–1353 (2021). https://doi.org:doi: 10.1126/science.abl8912 [DOI] [PubMed] [Google Scholar]
- 12.Ying B. et al. Boosting with variant-matched or historical mRNA vaccines protects against Omicron infection in mice. Cell 185, 1572–1587.e1511 (2022). 10.1016/j.cell.2022.03.037 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Gagne M. et al. mRNA-1273 or mRNA-Omicron boost in vaccinated macaques elicits similar B cell expansion, neutralizing responses, and protection from Omicron. Cell 185, 1556–1571.e1518 (2022). 10.1016/j.cell.2022.03.038 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Francis T. On the doctrine of original antigenic sin. Proceedings of the American Philosophical Society 104, 572–578 (1960). [Google Scholar]
- 15.Reynolds CJ et al. Immune boosting by B.1.1.529 (Omicron) depends on previous SARS-CoV-2 exposure. Science 377, eabq1841 (2022). https://doi.org:doi: 10.1126/science.abq1841 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Park YJ et al. Imprinted antibody responses against SARS-CoV-2 Omicron sublineages. Science 378, 619–627 (2022). 10.1126/science.adc9127 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Röltgen K. et al. Immune imprinting, breadth of variant recognition, and germinal center response in human SARS-CoV-2 infection and vaccination. Cell 185, 1025–1040.e1014 (2022). 10.1016/j.cell.2022.01.018 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Alsoussi WB et al. SARS-CoV-2 Omicron boosting induces de novo B cell response in humans. Nature 617, 592–598 (2023). 10.1038/s41586-023-06025-4 [DOI] [PubMed] [Google Scholar]
- 19.Scheaffer SM et al. Bivalent SARS-CoV-2 mRNA vaccines increase breadth of neutralization and protect against the BA.5 Omicron variant in mice. Nature Medicine (2022). 10.1038/s41591-022-02092-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Purtha WE, Tedder TF, Johnson S, Bhattacharya D & Diamond MS Memory B cells, but not long-lived plasma cells, possess antigen specificities for viral escape mutants. Journal of Experimental Medicine 208, 2599–2606 (2011). 10.1084/jem.20110740 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Li W. et al. Structural basis and mode of action for two broadly neutralizing antibodies against SARS-CoV-2 emerging variants of concern. Cell reports 38, 110210 (2022). 10.1016/j.celrep.2021.110210 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Alsoussi WB et al. A potently neutralizing antibody protects mice against SARS-CoV-2 infection. The Journal of Immunology 205, 915–922 (2020). 10.4049/jimmunol.2000583 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Su W. et al. Neutralizing monoclonal antibodies that target the spike receptor binding domain confer Fc receptor-independent protection against SARS-CoV-2 infection in Syrian hamsters. MBio 12, e0239521 (2021). 10.1128/mBio.02395-21 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.VanBlargan LA et al. A potently neutralizing SARS-CoV-2 antibody inhibits variants of concern by utilizing unique binding residues in a highly conserved epitope. Immunity 54, 2399–2416.e2396 (2021). 10.1016/j.immuni.2021.08.016 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Adams LJ et al. A broadly reactive antibody targeting the N-terminal domain of SARS-CoV-2 spike confers Fc-mediated protection. Cell Reports Medicine 4, 101305 (2023). 10.1016/j.xcrm.2023.101305 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Pinto D. et al. Broad betacoronavirus neutralization by a stem helix-specific human antibody. Science 373, 1109–1116 (2021). 10.1126/science.abj3321 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Zhou P. et al. A human antibody reveals a conserved site on beta-coronavirus spike proteins and confers protection against SARS-CoV-2 infection. Science Translational Medicine 14, eabi9215 (2022). 10.1126/scitranslmed.abi9215 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Case JB et al. Neutralizing antibody and soluble ACE2 inhibition of a replication-competent VSV-SARS-CoV-2 and a clinical isolate of SARS-CoV-2. Cell Host & Microbe 28, 475–485.e475 (2020). 10.1016/j.chom.2020.06.021 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Yisimayi A. et al. Repeated Omicron exposures override ancestral SARS-CoV-2 immune imprinting. Nature 625, 148–156 (2024). 10.1038/s41586-023-06753-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Centers for Disease Control and Prevention (CDC). Variant Proportions, <https://covid.cdc.gov/covid-data-tracker/#variant-proportions> (2023, November 16).
- 31.Barnes CO et al. SARS-CoV-2 neutralizing antibody structures inform therapeutic strategies. Nature 588, 682–687 (2020). 10.1038/s41586-020-2852-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Hurlburt NK et al. Structural definition of a pan-sarbecovirus neutralizing epitope on the spike S2 subunit. Communications Biology 5, 342 (2022). 10.1038/s42003-022-03262-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Pinto D. et al. Cross-neutralization of SARS-CoV-2 by a human monoclonal SARS-CoV antibody. Nature 583, 290–295 (2020). 10.1038/s41586-020-2349-y [DOI] [PubMed] [Google Scholar]
- 34.Cameroni E. et al. Broadly neutralizing antibodies overcome SARS-CoV-2 Omicron antigenic shift. Nature 602, 664–670 (2022). 10.1038/s41586-021-04386-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Khoury DS et al. Neutralizing antibody levels are highly predictive of immune protection from symptomatic SARS-CoV-2 infection. Nature medicine 27, 1205–1211 (2021). 10.1038/s41591-021-01377-8 [DOI] [PubMed] [Google Scholar]
- 36.Carreño JM, Singh G, Simon V & Krammer F Bivalent COVID-19 booster vaccines and the absence of BA.5-specific antibodies. The Lancet Microbe 4, e569 (2023). 10.1016/s2666-5247(23)00118-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Johnston TS et al. Immunological imprinting shapes the specificity of human antibody responses against SARS-CoV-2 variants. Immunity 57, 912–925.e914 (2024). 10.1016/j.immuni.2024.02.017 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Tortorici MA et al. Persistent immune imprinting occurs after vaccination with the COVID-19 XBB.1.5 mRNA booster in humans. Immunity 57, 904–911.e904 (2024). 10.1016/j.immuni.2024.02.016 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Mesin L. et al. Restricted clonality and limited germinal center reentry characterize memory B cell reactivation by boosting. Cell 180, 92–106.e111 (2020). 10.1016/j.cell.2019.11.032 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Schiepers A. et al. Molecular fate-mapping of serum antibody responses to repeat immunization. Nature 615, 482–489 (2023). 10.1038/s41586-023-05715-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Webster RG, Kasel JA, Couch RB & Laver WG Influenza virus subunit vaccines. II. Immunogenicity and original antigenic sin in humans. Journal of Infectious Diseases 134, 48–58 (1976). 10.1093/infdis/134.1.48 [DOI] [PubMed] [Google Scholar]
- 42.Singh G. et al. The post-COVID-19 population has a high prevalence of cross-reactive antibodies to spikes from all Orthocoronavirinae genera. mBio 15, e0225023 (2024). 10.1128/mbio.02250-23 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Dadonaite B. et al. A pseudovirus system enables deep mutational scanning of the full SARS-CoV-2 spike. Cell 186, 1263–1278.e1220 (2023). 10.1016/j.cell.2023.02.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Greaney AJ et al. Complete mapping of mutations to the SARS-CoV-2 spike receptor-binding domain that escape antibody recognition. Cell Host & Microbe 29, 44–57.e49 (2021). 10.1016/j.chom.2020.11.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Bangaru S. et al. Structural mapping of antibody landscapes to human betacoronavirus spike proteins. Science Advances 8, eabn2911 (2022). 10.1126/sciadv.abn2911 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Han J. et al. Polyclonal epitope mapping reveals temporal dynamics and diversity of human antibody responses to H5N1 vaccination. Cell Reports 34, 108682 (2021). 10.1016/j.celrep.2020.108682 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Turner JS et al. SARS-CoV-2 mRNA vaccines induce persistent human germinal centre responses. Nature 596, 109–113 (2021). 10.1038/s41586-021-03738-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Greaney AJ et al. A SARS-CoV-2 variant elicits an antibody response with a shifted immunodominance hierarchy. PLoS Pathog 18, e1010248 (2022). 10.1371/journal.ppat.1010248 [DOI] [PMC free article] [PubMed] [Google Scholar]
METHODS REFERENCES
- 49.Zang R. et al. TMPRSS2 and TMPRSS4 promote SARS-CoV-2 infection of human small intestinal enterocytes. Science immunology 5, eabc3582 (2020). 10.1126/sciimmunol.abc3582 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Chandran K, Sullivan NJ, Felbor U, Whelan SP & Cunningham JM Endosomal proteolysis of the Ebola virus glycoprotein is necessary for infection. Science 308, 1643–1645 (2005). 10.1126/science.1110656 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Whelan SP, Ball LA, Barr JN & Wertz GT Efficient recovery of infectious vesicular stomatitis virus entirely from cDNA clones. Proceedings of the National Academy of Sciences 92, 8388–8392 (1995). 10.1073/pnas.92.18.8388 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52.Buchholz UJ, Finke S & Conzelmann KK Generation of bovine respiratory syncytial virus (BRSV) from cDNA: BRSV NS2 is not essential for virus replication in tissue culture, and the human RSV leader region acts as a functional BRSV genome promoter. Journal of Virology 73, 251–259 (1999). 10.1128/jvi.73.1.251-259.1999 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53.Amanat F et al. SARS-CoV-2 mRNA vaccination induces functionally diverse antibodies to NTD, RBD, and S2. Cell 184, 3936–3948.e3910 (2021). 10.1016/j.cell.2021.06.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54.Stadlbauer D. et al. SARS-CoV-2 Seroconversion in Humans: A Detailed Protocol for a Serological Assay, Antigen Production, and Test Setup. Current protocols in microbiology 57, e100 (2020). 10.1002/cpmc.100 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 55.Nuqui X. et al. Simulation-Driven Design of Stabilized SARS-CoV-2 Spike S2 Immunogens. bioRxiv (2023). 10.1101/2023.10.24.563841 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.