Fig. 2. SARS-CoV-2-specific immune responses before and after bivalent booster vaccination. (A-C) Blood was collected from 18 HCWs 11 (IQR 5-31) days before (T1) and 57 (IQR 38-65) days after (T2) bivalent booster vaccination with 70 (IQR 69-71) days in between the two timepoints. (D-F) Immune responses of 8 HCWs who received no bivalent booster vaccination were also assessed at similar time points with an interval of 70 (IQR 69.5-75.5) days. (A,D) Spike S1-specific T cell responses. (B,E) Anti-RBD (ancestral) IgG concentrations. (C,F) Omicron BA.5 serum neutralization activity. HCWs were excluded if a previous (booster) vaccination was received within three months before T1 or if the HCW tested SARS-CoV-2 RT-qPCR positive in between T1 and T2. Data are represented as median with IQR and were assessed by a Wilcoxon signed-rank test.
Ratio of Omicron BA.5 to ancestral SARS-CoV-2 strain-specific immunological responses after Omicron BA.5 infection and ancestral/Omicron BA.1 bivalent vaccination
We determined the effect of a recent Omicron BA.5 infection, an ancestral/Omicron BA.1 bivalent booster vaccination, and both on Omicron BA.5 and ancestral SARS-CoV-2 specific immunity. For this purpose, we stimulated PBMCs with solely Omicron BA.5 mutation peptides and the corresponding ancestral strain spike peptides. Also, we performed sVNT assays using either Omicron BA.5 or ancestral SARS-CoV-2 pseudovirus particles.
At T1, the T cell responses against these specific peptides were overall low, and the Omicron BA.5/ancestral ratios were comparable between the two groups (Fig. 3A and Supplementary Fig. 2A) . The neutralization activity was not significantly different between the two groups at T1 (Figure 3B) . However, 92% of recently infected HCWs versus 57% of non-recently infected HCWs exhibited higher (i.e., a ratio of >1) neutralization activity against Omicron BA.5 compared to the ancestral strain.
At T2, T cell responses were also overall low and only the HCWs who were both recently infected with Omicron BA.5 and received a bivalent booster vaccination had 2.2-fold higher T cell responses against Omicron BA.5 than against ancestral strain (Figure 3C and Supplementary Fig. 2C) . Strikingly, this HCW group also demonstrated a 2.4-fold higher neutralizing activity against Omicron BA.5 compared to ancestral strain, which was higher (p = 0.0456) than in HCWs who received a bivalent vaccination but without a recent Omicron BA.5 infection (Fig. 3D) . The latter group exhibited a 1.4-fold higher neutralization activity.
For comparison, we also assessed the Omicron BA.5/ancestral neutralization ratios in pre-Omicron BA.5 serum samples which were obtained before and after monovalent booster vaccination in 2021.12 Accordingly, we observed that there was an overall higher neutralizing activity of the ancestral strain than of BA.5 (Supplementary Fig. 2) .