Figure 1 Estimated number of COVID-19 infections based on US wastewater surveillance data since the pandemic (from Twitter @JPWeiland)

    The receptor-binding domain (RBD) on the Spike glycoprotein of SARS-CoV-2 is an important domain that plays a crucial role during cell entry and is the main target of neutralizing antibodies (NAbs) induced by vaccination and infection. As a saltation variant, the RBD of BA.2.86 has 14 amino acid mutations compared to BA.2 and 12 compared to the previously prevalent XBB.1.5. JN.1 and other subvariants have even more mutations, indicating that the BA.2.86/JN.1 lineage is likely to exhibit distinct immunogenicity and antigenicity from previous strains. Meanwhile, during the circulation, the JN.1 lineage has also continuously accumulated more escape mutations in the RBD region, leading to subvariants like KP.2 and KP.3. These mutants have also quickly exhibited advantages, posing a severe challenge to the effectiveness of existing vaccine booster shots mainly based on the XBB.1.5. Therefore, analyzing the antigenicity and immunogenicity of the JN.1 lineage at the molecular level, determining the protective efficacy of existing vaccines and infections, and clarifying the similarities and differences between the antibodies elicited by XBB.1.5 and JN.1 are crucial for understanding the viral evolution and determining vaccine development strategies in the near future.

Figure 2 Global prevalence of SARS-CoV-2 variants since October 2023

    On November 7, 2024, the research group led by Prof. Yunlong Cao in the Biomedical Pioneering Innovation Center (BIOPIC) of Peking University, Peking University-Tsinghua University Joint Center for Life Sciences, and Changping Laboratory published a research article titled "Evolving antibody response to SARS-CoV-2 antigenic shift from XBB to JN.1" in Nature. Through a systematic analysis of the antibodies elicited by XBB and JN.1 infections, this study described the differences in immunogenicity between the XBB and JN.1 lineages, emphasized the strong immune escape capabilities of JN.1 lineage mutants such as KP.2, KP.3, especially KP.3.1.1, and detailed the distinct neutralization activities of ancestral cross-reactive and Omicron-specific antibodies. Class 1 broad-spectrum antibodies derived from IGHV3-53/3-66, which play a major neutralizing role among cross-binding antibodies, can compete with all Omicron-specific neutralizing epitopes, potentially masking the activation of specific naive B cells during Omicron breakthrough infections (BTIs), especially in the population vaccinated with highly immunogenic mRNA vaccines. This model can explain the reason why mRNA vaccination leads to exceptionally strong immune imprinting only in humans but not in mice, suggesting a potential "East-West immunity difference" of the global population due to different vaccination history and immune imprinting. The results indicate that in order to efficiently enrich truly broad-spectrum Class 1 (or epitope group A1) NAbs and eliminate the immune imprinting, it is essential to develop vaccine booster shots based on the new JN.1 subvariants, such as KP.2 and KP.3, which contain multiple A1 -specific escape mutations. After the preprint of this article was released on bioRxiv on April 19, 2024, it received extensive attention from the WHO and the international academic community, providing an important reference and guidance for the update of COVID-19 vaccines based on JN.1/KP.2 in 2024 autumn and winter.

Figure 3 Antigenic map of major SARS-CoV-2 strains in mice

    First, the authors immunized mice with mRNA vaccines encoding the Spike and used pseudovirus neutralization assays to evaluate the immunogenicity and antigenicity of major variants in mice. The analysis found that BA.2.86 and its subvariants JN.1, KP.2, and KP.3 exhibit completely different immunogenicity and antigenicity from the previously prevalent BA.2/BA.5/XBB lineages. SPR experiments demonstrated that the combined effect of the F456L and Q493E mutations in KP.3 significantly increased the ACE2 affinity.

Figure 4 Comparison of plasma neutralization titers of different lineages of COVID-19 mutant strains in populations with different immune backgrounds

    Therefore, this study collected peripheral blood samples from seven cohorts with different immune histories and tested the neutralizing activities of their plasma against variants. In the unvaccinated population, there was almost no cross-neutralizing antibody between the infection of the XBB lineage and that of the JN.1 lineage, which was consistent with the results of mouse immunization. However, for those who had been vaccinated with inactivated vaccines or infected with BA.5/BF.7 before being infected with XBB or JN.1, XBB and JN.1 lineages could elicit cross-neutralization, indicating the existence of potentially broadly neutralizing antibodies. Meanwhile, the study found that the recent KP.3.1.1 (KP.3 + S31del) consistently showed the strongest immune escape ability, explaining its high growth advantage. Overall, JN.1 BTI showed significantly higher neutralizing antibody titers against JN.1 subvariants than that of the XBB lineage, including HK.3, supporting the importance of developing vaccine boosters based on the JN.1 lineage.

Figure 5 Specificity and SHM of isolated mAbs from different cohorts

    To further understand the impact of immune histories on the humoral responses to the XBB/JN.1 lineages, the authors further isolated RBD-specific memory B cells by FACS, determined ~2,000 BCR sequences by single-cell V(D)J sequencing, and expressed them as monoclonal antibodies (mAbs). The memory B cells in all cohorts encoded antibodies that could cross-bind with the ancestral strain, as well as antibodies specific to the Omicron lineage. Due to immune imprinting, the proportion of cross-reactive antibodies in the XBB single-breakthrough was the highest, and the imprinting was weakened by repeated infections. In the vaccinated population, the somatic hypermutation (SHM) rate of the cross-reactive antibodies was significantly higher than that of the Omicron-specific antibodies; but not in the unvaccinated population, which was also in line with expectations.

Figure 6 Epitope mapping of XBB.1.5 and JN.1 RBD-specific antibodies

Figure 7 Neutralizing activities of RBD antibodies targeting different epitopes

    Next, the authors applied the previously developed high-throughput deep mutation scanning (DMS) to determine the escape mutation profiles of the RBD-specific mAbs. Among the 12 epitope groups identified, A1, D2, E1/E2.1, E2.2, E3, and F1.1 were mainly ancestral cross-reactive antibodies, while the remaining mainly contained Omicron-specific antibodies. A1 is the only epitope that can cross-bind with the ancestral strain and has strong neutralizing activity against new variants. Notably, the proportion of A1 antibodies was positively correlated with the number of immunizations experienced. Among the Omicron-specific antibodies, the F3 epitope group exhibited strongest neutralization and was observed in multiple cohorts. However, previous studies have shown that even after repeated exposures to Omicron, mRNA vaccine recipients could not effectively elicit Omicron-specific antibodies. To further study the effective neutralizing antibody components of different groups, especially those who had been reinfected with different strains, the authors next focused on the contributions of ancestral-cross-binding antibodies and Omicron-specific antibodies to the neutralization against latest JN.1 subvariants.

Figure 8 Neutralizing activities of the ancestral cross-binding antibodies against new subvariants

    Consistent with the previous analysis, among the neutralizing antibodies that cross-bind with the ancestral strain, A1 antibodies play a major neutralizing role against JN.1, KP.2, and KP.3. A1 antibodies are mainly the classic Class 1 NAbs derived from IGHV3-53/3-66, and their binding sites highly overlap with that of the receptor ACE2. These findings are consistent with the high growth advantage of A1-specific escape mutations in the real world. Although the mutations such as F456L and Q493E in KP.2 and KP.3 are located on the epitopes of such antibodies, many A1 antibodies still maintain high neutralizing efficiency. Moreover, the antibodies from HK.3/JN.1 reinfection exhibit better neutralization breadth, reflecting that the mutations in HK.3/JN.1 can effectively enrich broad-spectrum A1 NAbs. These results indicate that developing vaccines based on JN.1, or even further including KP.2/KP.3 with extensive A1 epitope mutations, is helpful for enriching A1 broadly neutralizing antibodies.

Figure 9 JN.1 can efficiently elicit Omicron-specific F3 neutralizing antibodies

    Among the Omicron-specific antibodies, the F3 contributed mainly to the neutralization against JN.1 lineage. This epitope is spatially overlapped with the broad-spectrum neutralizing antibody SA55 previously reported by the authors, but these specific F3 cannot bind to the ancestral RBD. Interestingly, except for G504, which has been proven to possibly cause conformational changes and is unfavorable, the main escape sites of such antibodies are all sites that have been mutated in Omicron. Further mutations of these sites may lead to the recovery of the escaped antibodies, which may also be unfavorable. This indicates that such antibodies may be good Omicron broad-spectrum NAbs that are difficult to be escaped by future mutants. Compared to XBB/HK.3, F3 from JN.1 infection exhibit significantly better neutralization against the JN.1 lineage. F3 NAbs are mainly derived from IGHV2-5 and IGHV5-51 genes. JN.1 reinfection can better enrich the IGHV5-51 F3 antibody which exhibit higher neutralization breadth. These findings also support the vaccine boosters development based on the JN.1 lineage.

Figure 10 The broad-spectrum A1 public antibodies compete with all Omicron-specific neutralizing epitopes

    Previous studies have shown that the immune imprinting induced by mRNA vaccination in humans cannot be eliminated by even repeated exposures to Omicron. However, this phenomenon does not exist in mice. Even when immunized with the ancestral mRNA vaccine, mice that are repeatedly immunized with Omicron can efficiently generate Omicron-specific antibodies. The main difference in the humoral immunity between humans and mice lies in the composition of V(D)J genes. Based on the epitope analysis of the Omicron-specific antibodies above, the authors found that the "public antibodies" using IGHV3-53/3-66 of the A1 epitope can compete with all Omicron-specific neutralizing epitopes, and this was verified by SPR competition experiments. Therefore, the authors proposed a hypothesis that the strong immune imprinting caused by mRNA vaccination is related to the response of the IGHV3-53/3-66 public antibodies.

Figure 11 Proposed model to explain the differences in the results of immune imprinting caused by mRNA vaccination in humans and mice

    Due to the strong immunogenicity of mRNA vaccines, strong IGHV3-53/3-66 antibody responses could be induced and matured during the initial immunization, and since the escape mutations on their epitopes are limited due to the high receptor mimicking ability of these NAbs, they are not completely escaped by the Omicron. These remaining memory B cells are recalled upon the first Omicron immunization, and due to their competition with Omicron-specific antibodies, the epitope masking makes it difficult for the Omicron-specific naive B cells to be activated and matured. As a result, even when continuously exposed to Omicron, there is a tendency to repeatedly activate these A1 antibodies rather than produce novel specific antibodies.

    On the other hand, in the population mainly vaccinated with inactivated vaccines in China, because the immunogenicity of inactivated vaccines is weaker than that of mRNA vaccines, and due to the “zero-COVID” policy, the long-term immune attenuation process reduces the role of such antibody memories. Therefore, when the population vaccinated with inactivated vaccines analyzed in this study is infected with Omicron, the remaining A1 antibodies or memory B cells are not sufficient to form epitope masking, allowing specific antibodies to be produced and matured, and then reactivated during a second immunization. On the other hand, there are no genes in mice that can cause similar IGHV3-53/3-66 immune responses, nor can they effectively prevent the production of Omicron-specific antibodies. The above analysis requires further experimental research verification in the future.

    Link: https://www.nature.com/articles/s41586-024-08315-x

    Yunlong Cao, a principal investigator at the Biomedical Pioneering Innovation Center of Peking University, Peking-Tsinghua Center for Life Sciences and the Changping Laboratory, is the corresponding author of this article. Fanchong Jian, Jing Wang, Dr. Yisimayi Ayijiang, Weiliang Song, and Yanli Xu are the co-first authors. This research was financially supported by the Ministry of Science and Technology of China, Changping Laboratory, and the National Natural Science Foundation of China.

    Cao Lab is dedicated to applying single-cell sequencing and high-throughput screening technologies, combined with advanced computational methods, to deeply explore frontier issues in virology and immunology, and develop broad-spectrum vaccines and antibody therapeutics. We sincerely invite talented investigators in engaging in relevant basic research and translational directions to join us. If you are interested in the positions, please refer to the website https://yunlongcaolab.com or contact Prof. Yunlong Cao at yunlongcao@pku.edu.cn.