A total of 52 participants were enrolled in this study. The first group (10 females and 2 males, median age: 27.5) received three doses of inactivated vaccines3,19 and had a breakthrough infection during a wave of BF.7/BA.5.2 circulation in China in December 20226, and their blood samples were collected on 13th Jan 2023. The second group (6 females and 6 males, median age: 30.5) received three doses of ZF2001 vaccine4,20 and had a breakthrough infection during a wave of BF.7/BA.5.2 circulation in China in December 2022, and their blood samples were collected on 5th Jan 2023. The third group (9 females and 3 males, median age: 25.5) received three doses of inactivated vaccine and had two breakthrough infections during the waves of BF.7/BA.5.2 in December 2022 and XBB sub-variants in 2023, respectively. Their blood samples were collected on 26th Jan 2024. The fourth group (8 females and 8 males, median age: 48.5) received three doses of the inactivated vaccine, followed by a booster dose of the ZF2202-A vaccine. ZF2202-A is ZF2001 vaccine’s next-generation COVID-19 vaccine with updated bivalent Delta-BA.5 RBD-heterodimer immunogen8,21. The COVID-19 inactivated vaccines BBIBP-CorV and CoronaVac are widely used in China and contain 4 and 3 μg inactivated prototype SARS-CoV-2, respectively. The ZF2001 and ZF2202-A vaccines contain 25 μg immunogen. All these vaccines are adjuvanted by aluminum hydroxide. Serum samples from the first, second and third groups were collected from the participants in real world, which were approved by the Ethics Committee of the Institute of Microbiology, Chinese Academy of Sciences. Serum samples from the fourth group were collected 3 weeks after boosting from a clinical trial (ClinicalTrials.gov: NCT05850507), which were approved by the clinical research ethics board of the First Affiliated Hospital of Wannan Medical College (Yijishan Hospital) ([2023]KY34). All participants signed the written informed consent. Detailed information about the participants is available in Supplementary Table 1.

Specific pathogen-free (SPF) female BALB/c mice were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. (licensed by Charles River). All mice were allowed free access to water and a standard chow diet and provided with a 12-h light and dark cycle (temperature: 20 °C–25 °C, humidity: 40%–70%). All mice used in this study are in good health and are not involved in other experimental procedures. They were housed under SPF conditions in the laboratory animal facilities at IMCAS. The mice experiments conducted in IMCAS were approved by the Committee on the Ethics of Animal Experiments of the IMCAS, and performed in compliance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the IMCAS Ethics Committee.

Antigen proteins were prepared as previously described21. They will be made available upon request. AddaVax adjuvant was purchased from InvivoGen. Groups of 6- to 8-week-old female BALB/c mice (n = 6) were intramuscularly immunized with three doses of 100 μL vaccines, containing 2 μg immunogens and AddaVax adjuvant. The interval between the first and second doses was 21 days. The interval between the second and third doses was 21 days. The serum samples were collected after anesthesia at 7 days post the last immunization. These mice were then euthanized in a box by carbon dioxide.

The methods for preparing pseudotyped viruses and neutralization assays were described previously22,23. Briefly, the spike protein with 18 amino acids deleted from the C-terminus of each variant was cloned into the pCAGGS vector, respectively. Codon optimization was performed to make the plasmid suitable for mammalian cell expression, and 30 μg of each construct was transfected into HEK-293T cells. VSV-ΔG-GFP pseudotyped virus was added 24 h after transfection. After 2 h of infection, the VSV-ΔG-GFP residues were removed by changing the medium to fresh complete DMEM medium containing anti-VSV-G antibody (I1Hybridoma, ATCC CRL-2700). After incubation at 37 °C for 30 h, the supernatants were collected, filtered through a 0.45 μm filter (Millipore, Cat#SLHP033RB), and then aliquotted and stored tor at −80 °C.

The pseudotyped viruses displaying SARS-CoV-2 spike protein express GFP in infected cells. The heat-inactivated (56 °C for 30 min) human and murine sera samples were 2-fold serial diluted starting from 1:20 and 1:40, respectively. Equivalent pseudovirus (1000 transducing units, TU) was incubated with the sera at 37 °C for 1 h, and the mixture was then added onto pre-plated Vero E6 cells (ATCC CRL-1586) in 96 well plates. The TU numbers were calculated on a CQ1 confocal image cytometer (Yokogawa) after a 15 h incubation. The 50% pseudovirus neutralization titer (pVNT50) was determined by fitting non-linear regression curves (log(inhibitor) vs. normalized response - Variable slope) using GraphPad Prism and calculating the reciprocal of the serum dilution required for 50% neutralization of infection. The model of the non-linear regression curve is Y = 100/(1 + 10^((LogpVNT50-X)*HillSlope)). HillSlope describes the steepness of the family of curves. pVNT50 below the limit of detection was determined as half the limit of detection.

p-values were analyzed with two-tailed Wilcoxon signed-rank tests or Dunn’s multiple comparison tests. Significant levels are indicated by asterisks; ****p < 0.0001, ***p < 0.001, **p < 0.01, *p < 0.05. The details are indicated in the figure legends.