Title: Progressive loss of conserved spike protein neutralizing antibody sites in Omicron sublineages is balanced by preserved T-cell recognition epitopes

Abstract


Main Text
The SARS-CoV-2 Omicron variant of concern (VOC), that emerged in November 2021, contains over 30 amino acid alterations in its spike (S) glycoprotein as compared to the original Wuhan-Hu-1 (wild-type) strain that mediate partial escape from previously established immunity (1-3).  mAb) cilgavimab (9). Cilgavimab in combination with tixagevimab (Evusheld TM ) is used for BQ.1.1 (11), which additionally has a K444T alteration. The S protein of XBB, a recombinant strain (12), is altered at seven amino acid positions (four in NTD, three in RBD) relative to the 70 main Omicron sublineages BA.1, BA.2, and BA.4/5 and also bears the R346T alteration. It is to be expected that virus evolution will continue. Real-time understanding of the transmissibility, pathogenicity, and immune evasion properties of new variants in conjunction with immunity patterns in the human population that are being shaped through repeated infections with different VOCs, vaccination, and booster cycles with different vaccines, will continue to be critical to 75 assessing the level of risk to public health going forward.
In the face of this highly dynamic situation, we have set up a COVID-19 pandemic preparedness and rapid response strategy. Our approach includes the screening of emerging and circulating variants through an artificial intelligence/machine learning-based early warning system (13), testing prototypical mRNA vaccines adapted to these variants in mouse studies (16), and 80 mapping SARS-CoV-2 T-cell epitopes recognized by the human T-cell repertoire (19,20).
Further, studying the neutralizing antibody activity of individuals following breakthrough infections with the latest circulating variants allows us to detect potential immune escape patterns early and informs on the need for rapid vaccine adaptation strategies (14)(15)(16)(17)(18).
In continuation of this work, the current study assessed Omicron sublineages BA. 4  (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made We next sought to determine whether distinct cross-neutralization of individual SARS-CoV-2 variants is reflected by the degree of B-cell epitope conservation with respect to the wild-type strain. For this, we analyzed a total of 506 neutralizing B-cell epitopes within the S glycoprotein NTD and RBD, consisting of experimentally confirmed epitopes from the Immune Epitope 145 Database (IEDB) and epitopes computationally deduced from structures of SARS-CoV-2 neutralizing antibodies from the Coronavirus Antibody Database (CoV-AbDab). Of these 506 epitopes, 462 (91%) included a position that was altered in at least one of the analyzed variants.
We found that B-cell epitopes were partially conserved in the earlier variants Alpha, Beta, and Delta (≥43%) (Fig 2, Fig S3, and Table S2) This renders T-cell immunity in the population harder to evade through accumulation of mutations. Cytotoxic CD8 + T cells that recognize HLA class I presented epitopes play a critical 160 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted December 15, 2022. ; https://doi.org/10.1101/2022.12.15.520569 doi: bioRxiv preprint role in susceptibility to severe COVID-19 disease, hospitalization and death (18,19). HLA class II dependent CD4 + helper T cells typically orchestrate the immune response by their pleiotropic functions and are also considered to have an important role in the prevention of severe disease (20, 21).
To estimate the impact of SARS-CoV-2 evolution on T-cell immunity, we assessed the degree of 165 conservation of T-cell epitopes localized in the S glycoproteins of various VOCs. To this aim we filtered experimentally confirmed SARS-CoV-2 epitopes reported for HLA class I and II alleles from the IEDB database. In total, 594 and 559 unique epitopes were collected for HLA class I and class II, respectively. For stringency, in this analysis we only retained experimentally confirmed minimal epitopes of either HLA class I or HLA class II in a human host and removed 170 any inferred or predicted entries. Epitopes were also filtered by length (8-14 for HLA class I and 12-20 for HLA class II). Following filtering, 260 unique HLA class I epitope sequences remained, of which 244 were found in the wild-type strain S glycoprotein. Of these, 71 epitopes Approximately 90% of CD8 + and CD4 + T-cell epitopes of the wild-type S glycoprotein were fully conserved in the Alpha, Beta, and Delta variants (Fig 2, fig S3, Tables S3 and S4) and over 80% of CD8 + and ~70% CD4 + T-cell epitopes were fully conserved in Omicron sublineages 180 including BA.2.75.2, BQ.1.1, and XBB ( Fig. 2 and fig. S3), suggesting that T-cell responses against Omicron sublineages may remain largely intact in individuals immunized with wild-type strain-based vaccines.
. CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made In summary, our findings provide insights into two mechanisms: firstly, the neutralizing activity of human sera in the context of currently circulating VOCs, and second, the potential of the 185 SARS-CoV-2 S glycoprotein directed CD8 + and CD4 + T-cell repertoire.
With regard to humoral immunity, we show that exposure of vaccinated individuals to Omicron (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted December 15, 2022. ; https://doi.org/10.1101/2022.12.15.520569 doi: bioRxiv preprint that disrupt critical B-cell neutralization epitopes, and that further boosters or variant adaptations may be required in future.
With regard to cell-mediated immunity, we show that HLA class I and class II presented T-cell epitopes remained mostly unaltered across the evolution of SARS-CoV-2 including Omicron 210 sublineages, suggesting that CD8 + and CD4 + T-cell immunity against Omicron BQ.1.1, BA.2.75.2, and XBB may be largely intact, despite profound neutralizing antibody evasion. This fits in with previous reports, e.g. describing the existence of degenerate T-cell epitopes located within conserved S protein regions (25) and showing that wildtype-strain-vaccinated individuals retain T-cell immunity against Omicron BA.1 (26)(27)(28). Indeed, a fundamental difference of T-215 cell versus B-cell mediated immunity is that owing to the highly polymorphic nature of HLA molecules (29)(30)(31), the T-cell mediated layer of immunity is more robust against population-level breaches by VOCs. Our observations indicate that T-cell immunity may mitigate the lack of neutralizing antibody activity in preventing or limiting severe COVID-19 and further encourages development of vaccine formats that induce functional and broad SARS-CoV-2-directed CD8 + 220 T-cell immunity concurrently to boosting antibody responses.
. CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted December 15, 2022. ; (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted December 15, 2022. ; https://doi.org/10.1101/2022.12.15.520569 doi: bioRxiv preprint VOC pVN 50 GMTs normalized against the wild-type strain pVN 50 GMT (ratio VOC to wildtype). Group geometric mean ratios with 95% confidence intervals are shown. The non-235 parametric Kruskal-Wallis test with Dunn's multiple comparisons correction was used to compare the VOC GMT ratios between cohorts. ****, P<0.0001; ***, p<0.001; **, P<0.01; *, P<0.05. Serum was tested in duplicate.
. CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made  (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted December 15, 2022. ; https://doi.org/10.1101/2022.12.15.520569 doi: bioRxiv preprint the one used in our early warning system (13). T-cell epitopes were retrieved from the Immune Epitope Database (IEDB) on November 11, 2022.
. CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted December 15, 2022.

Materials and Methods
Figs. S1-S4 290 Tables S1-S9 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted December 15, 2022. ; https://doi.org/10.1101/2022.12.15.520569 doi: bioRxiv preprint