At the beginning of 2020, the COVID-19 caused a short chaos in China, but the epidemic was quickly brought under control, due to the strict and effective epidemic prevention and control policies. Along with the increasing vaccination rate, the control policies were gradually loosened. At the beginning of 2022, over 87% of people had been fully vaccinated, and most of the them were administrated with three doses of inactivated vaccine, which was an inactivated vaccine based on the original SARS-CoV-2 strain (www.gov.cn). The herd immunity built by inactivated vaccines had played important roles against breakthrough infection in China. However, over 80% of Chinese were infected by Omicron BA.5.2 or BF.7 variants from December 2022 to February 2023, leading to a peak of infection ( Figure 1A ). Fortunately, the case fatality rate of Omicron variants was significantly lower than the former strains, which could be explained by the vaccination, better medical conditions, and the decrease in the pathogenicity of the virus itself (Yuan and Ye, 2022; Bálint et al., 2022; Fan et al., 2022). After this infection peak, only mini-waves emerged, with the rapid changes in the circulating Omicron sublineages in China. It was worth noting that the Omicron XBB.1.5 caused a small infection peak in the summer and autumn of 2023 ( Figure 1A ).

To monitor the immune response in vaccinees, we tracked a cohort consisting 30 participants in Wuhan, who were infected by Omicron BA.5.2 subvariant in December 2022 and were infected by Omicron XBB.1.5 in April–May 2023. As shown in Figure 1B , the sera samples of this cohort were collected at multiple time points: 1 month and 9 months after the last dose inactivated vaccine, 1 month and 4 months after each infection.

It was reported that D614G can significantly improve infectivity of SARS-CoV-2 (Zhou et al., 2021; Ma et al., 2022). The Omicron subvariants carried more than 30 mutations in addition to D614G in Spike ( Figure 2A ). In order to investigate the potential effect of mutations on the infectivity, we detected the hACE2 affinity and infection features of SARS-CoV-2 wild type (WT), Delta variant, and several Omicron subvariants. Through biolayer interferometry (BLI) assay, the hACE2 binding capacity of RBDs from SARS-CoV-2 WT and other seven previous/currently circulating variants were measured. The kinetic binding affinity (K_D ) values between RBDs and hACE2 were from 22.34  nmol/L to 65.21 nmol/L, showing an increasing trend of hACE2 affinity from WT to XBB.1.5 and JN.1 but still in same order magnitude ( Figures 2B, C ).

We then tested the infection features of these strains through pseudovirus infection assay. Pseudovirus bearing WT-Spike protein with D614G mutation (WT-D614G) was used as a reference virus as the D614G is critical for higher viral infectivity. Interestingly, no significant difference was observed in all tested strains including WT-D614G, Delta, Omicron BA.5.2, BF.7, BQ.1.1, XBB.1.5, EG.5, and JN.1 at the same median tissue culture infectious dose (TCID₅₀) in BHK-21-hACE2 cells ( Figure 2D ).

The emergence of immune evasion SARS-CoV-2 variants repeatedly, especially Omicron subvariants (BA.5.2, BF.7, XBB.1.5, EG.5, etc.), is still a global threat (Wang et al., 2022; Kurhade et al., 2023; Miller et al., 2023). To monitor the immune response against these variants in vaccinees, we first conducted the pseudovirus neutralization assay with sera samples collected at 1 month and 9 months after the third dose of inactivated vaccine administration. One month after the third dose, the serum mean neutralization efficiency against WT-D614G was 60.6%. However, the efficiency reduced to 17.6%, 15.5%, 3.4%, 1.4%, 1.1%, and 0.6% for BA.5.2, BF.7, BQ.1.1, XBB.1.5, EG.5, and JN.1, respectively, when we tested the sera samples at 25-fold dilution, indicating the significant immune evasion of these Omicron subvariants. Nine months after the third dose of inactivated vaccine, the mean neutralization efficiency of the sera against WT-D614G drop to 33.2%, about a half than before. Moreover, the efficiency against Omicron subvariants dropped to near background when tested at 25-fold dilution ( Figure 3A ). By November 2022, it has been more than 9 months since the last vaccination for most people in China. Considering the immune evasion of BA.5.2 and BF.7, the major prevalent subvariants (https://www.chinacdc.cn), as well as the dropped herd immunity established by vaccines, the infection peak of Omicron BA.5.2/BF.7 infection during this period could be explained.

After the infection peak of BA.5.2/BF.7 in China, new variants such as BQ.1.1 and XBB.1.5 were circulating around the world (Callaway, 2023). We conducted the pseudovirus neutralization assay with sera samples collected at 1 month and 4 months after the BA.5.2 infection. One month after BA.5.2 infection, the humoral immune was significantly improved, the serum mean neutralization efficiency was 98.7%, 97.7%, 96.8%, 84.3%, 58.0%, 54.3%, and 50.9% against WT-D614G, BA.5.2, BF.7, BQ.1.1, XBB.1.5, EG.5, and JN.1, respectively. Four months after the BA.5.2 infection, the efficiency against WT-D614G, BA.5.2, BF.7, and BQ.1.1 maintained well, but the efficiency against XBB.1.5 dropped to 29.8% ( Figure 2B ), consistent with the mini-wave of XBB.1.5 in the summer of 2023 in China ( Figure 1 ). In addition, the humoral immune against Omicron sublineages was significantly improved by the XBB.1.5 infection. The serum neutralization efficiency against most tested Omicron sublineages kept at relatively high level within 4 months after the reinfection ( Figure 3C ). A potentially unsettling factor is that the neutralization efficiency against JN.1 sublineages dropped to 61.2% at 4 months after the reinfection, indicating the potential infection risk of JN.1 in the spring of 2024 in China.

For further quantitative research on the immune response in vaccinees after breakthrough infection, we assessed the geometric mean 50% neutralization titer (NT₅₀) of sera samples collected after infections against the Omicron sublineages. One month after the BA.5.2 infection, the NT₅₀ against the WT-D614G, BA.5.2, BF.7, BQ.1.1, XBB.1.5, EG.5, and JN.1 was 5869.6, 1957.0, 1607.4, 785.0, 112.7, 65.3, and 60.3, respectively. Comparing with WT-D614G, the NT₅₀ to Omicron subvariants are lower by factors of 3.0–97.3. Four months after the BA.5.2 infection, the NT₅₀ decreased by factor of 1.8–5.6, alerting the reinfection risk of XBB.1.5 at that time ( Figures 4A, B ). Notably, we found out that the infection of Omicron BA.5.2 largely boosted the neutralizing antibodies against WT-D614G ( Figures 3A, B ). Meanwhile, the NT₅₀ against the WT-D614G was higher than NT₅₀ against the BA.5.2 after the BA.5.2 infection ( Figures 4A, B ), indicating the vaccine-induced immune imprinting.

In addition, we also connected another 11 participants who had accepted three doses of inactivated vaccine administrations and then been infected by BF.7 in Huzhou, China. Their immune response at 1 month after BF.7 infection was similar with BA.5.2 convalescents ( Supplementary Figure S1 ), which are understandable since there was only one different mutation (R346T) in their Spike ( Figure 2A ). Regretfully, we lost connection with this BF.7 cohort.

Moreover, 1 month after the XBB.1.5 infection, the NT₅₀ against the WT-D614G, BA.5.2, BF.7, BQ.1.1, XBB.1.5, EG.5, and JN.1 was 983.1, 4671.6, 2155.7, 1128.1, 815.2, 552.0, and 302.2, respectively. Four months after the XBB.1.5 infection, the NT₅₀ decreased by factor of 3.9–5.3 ( Figures 4C, D ). After the infection of XBB.1.5, NT₅₀ against the BA.5.2 became the highest rather than NT₅₀ against WT-D614G. These results indicated that two rounds of Omicron infections largely mitigated the immune imprint of three doses of inactivated vaccine.

Blood samples from the long-tracked cohort (30 participants) who had three-dose vaccination of inactivated vaccine were collected after 1 month and 9–10 months. The same cohort was infected by Omicron BA.5.2, and all participants recovered within 2 weeks. At 1 month and 4 months after the infection, blood samples from this cohort were collected. The same cohort was then infected by Omicron XBB.1.5, and all individuals recovered within 2 weeks. At 1 month and 4 months after the infection, blood samples from this cohort were collected. Another 11 blood samples were collected at 1 month after the BF.7 infection. After centrifugation, serum from the upper layer were collected and incubated at 56°C for 30 min for heat inactivation, and then stored at −80°C until use. The infecting strain was determined by viral genome sequencing of the throat swab from the participants. The participants were randomly recruited to the cohort according to the criteria of the easy sampling principle. Convalescents who had mental disorders, dementia, difficulty in moving freely, or refused to enroll or were not contactable were excluded from the cohort. The age range for the participants was 19–65 years; 16 participants were male and 14 participants were female.

The DNA sequences of human codon-optimized S proteins from SARS-COV-2 variants (WT-D614G, Delta, BA.5.2, BF.7, BQ.1.1, XBB.1.5, EG.5, and JN.1) and S protein mutations were commercially synthesized or generated by overlapping PCR-based mutagenesis using pCAGGS-SARS-CoV-2-S-C9 as the template and cloned into pCAGGS vector with C-terminal 18 amino acid (aa) truncation to improve VSV pseudotyping efficiency.

BHK-21 (ATCC, CCL-10) and Vero E6 cells (ATCC, CRL-1586) were cultured in Dulbecco’s modified Eagle’s medium (DMEM, Gibco) supplemented with 10% fetal bovine serum (FBS, Excell Bio, Suzhou, China), 2.0 mM L-glutamine, 110 mg/L sodium pyruvate, and 4.5 g/L D-glucose. BHK-21-ACE2 (clone 7), which stably expresses human ACE2 was generated based on BHK-21 through lentiviral transduction and maintained with puromycin at 1 μg/mL. I1-Hybridoma (CRL-2700) secreting a monoclonal antibody targeting VSV glycoprotein was cultured in minimum essential medium with Earle’s salts and 2.0 mM L-glutamine (MEM; Gibco) supplemented with 10% FBS. All cells were cultured at 37°C in 5% CO₂ with regular passage every 2–3 days.

HEK293F cells were cultured in SMM 293T-I medium (Sino Biological) at 37°C, with 5% CO₂. Plasmids of Fc-RBD were transfected into HEK293F when the cell density is 1 × 10⁶ cells per mL with EZ Trans (Life-Lilab) transfection reagent. The cell supernatant was collected by centrifugation to remove the cells, and Protein A Resin (GenScript) was used to purify the mAbs. These mAbs were further purified by a Superose 6 10/300 column (GE Life Sciences) and resuspended into phosphate-buffered saline (PBS), and the purity of the recombinant antibodies was analyzed by SDS-PAGE.

The binding affinity of the purified SARS-CoV-2 Fc-RBD protein with His-hACE2 was monitored by BLI using an Octet-Red 96 device (Pall ForteBio). Briefly, 20 μg/ml of Fc-RBD were loaded onto protein A (ProA) biosensors (ForteBio, 18–5010), and then dipped into phosphate-buffered saline with Tween 20 (PBST) buffer to wash out unbound mAb. Then, the biosensors were dipped into PBST buffer containing soluble His-hACE2 protein with concentrations ranging from 0 nM to 300 nM and then dipped into PBST buffer to record dissociation kinetics. PBST buffer was used to define the background. The affinities were calculated by Octet Data Analysis software.

The pseudovirus packaged with spike proteins from the WT-D614G SARS-CoV-2, and the SARS-CoV-2 variants were produced as we described before (Liu Q. et al., 2022). In brief, Vero-E6 cells were transfected with plasmids expressing different S proteins through Lipofectamine 2000 (Biosharp, Hefei, China). After 24h, the transfected cells were inoculated with VSV-dG-fLuc [1 × 10⁶ 50% tissue culture infectious dose (TCID₅₀)/mL] diluted in DMEM for 5h at 37°C and then replenished with growth medium (DMEM with 10% FBS) containing anti-VSV-G monoclonal antibody (I1-hybridoma, cultured supernatant, 1:20). Twenty-four hours later, the SARS-CoV-2 pseudovirus-containing supernatant was harvested and clarified at 3,000 rpm for 10 min, aliquoted, and frozen at −80°C for storage. The TCID₅₀ of the pseudovirus was determined using a serial dilution-based infection assay on BHK-21-hACE2 cells and calculated according to the Reed-Muench method.

The SARS-CoV-2 pseudoviruses (1 × 10⁵) TCID₅₀/well) were incubated with serial-diluted sera at room temperature for 30 min in 96-well white flat-bottom culture plates and then mixed with trypsinized BHK-21-hACE2 cells at a density of 2 × 10⁴/well. After 16h, the medium of the infected cells was removed, and the cells were lysed with 1× Bright-Glo Luciferase Assay reagent (Promega) for chemiluminescence detection using a SpectraMax iD3 multi-well luminometer (molecular devices). The 50% neutralization dilution titer (NT₅₀) was calculated by GraphPad Prism 7 software with nonlinear regression curve fitting (normalized response, variable slope).