Yuelong Shu

Background: Influenza remains a significant global public health threat, causing substantial morbidity and mortality worldwide. While vaccination serves as the best preventive strategy, considerable interindividual variability in vaccine-induced immune responses persists. The ZBTB46 rs2281929 polymorphism (c.A31G; p.T11A; ACG>GCG), which corresponds to the evolutionarily conserved mouse mutation Zbtb46^T11A (c.A31G; ACT>GCT), has been associated with enhanced antibody responses to influenza vaccination in humans, though its functional mechanisms remain unknown. This study aimed to investigate how this mutation affects influenza vaccine immunogenicity using a knock-in mouse model. Methods: A Zbtb46^T11A knock-in mouse model was generated using CRISPR/Cas9 technology. Homozygous (HO) and wild-type (WT) mice were immunized with a quadrivalent influenza vaccine in a prime-boost regimen. Humoral immune responses were assessed by Enzyme-linked immunosorbent assay, hemagglutination inhibition (HI), and microneutralization (MN) assays. Antibody-secreting cells (ASCs) were quantified by Enzyme-linked immunospot assays. Germinal center B cells, plasma cells, plasmablast cells, conventional dendritic cell (cDC) subsets, and T helper (Th) cells were analyzed by flow cytometry. Statistical comparisons were performed using a two-sample t-test. Results: The Zbtb46^T11A mutation did not alter Zbtb46 protein expression or its abundance in cDCs. Following vaccination, HO mice exhibited significantly enhanced humoral responses, including higher HA-specific IgG titers, HI and MN antibody levels, and increased numbers of ASCs. Flow cytometry revealed elevated proportions of germinal center B cells and plasma cells in HO mice. Furthermore, HO mice showed a selective expansion of type 1 cDCs (cDC1s) and a concomitant increase in Th1 cell frequencies and IFN-γ-secreting cells, while cDC2 proportions and Th2 responses remained unchanged. Conclusions: The Zbtb46^T11A mutation is associated with enhanced influenza vaccine immunogenicity, concomitant with increased cDC1 proportions, Th1 polarization, and germinal center-dependent humoral immunity. These observed associations suggest a candidate mechanism whereby Zbtb46 modulation may shape adaptive immunity, though further functional studies are required to establish causality. These findings provide insights into host genetic variation in vaccine responsiveness and may inform personalized vaccination strategies.

Genome-wide association studies (GWAS) have revealed extensive polygenic signals and overlapping genetic architectures across human traits, creating a need for resources that connect trait-level genetic relationships with gene-level functional evidence. Here, we developed the Multi-Omics Causal Resource Database (MOCR-DB), an interactive platform that integrates large-scale GWAS summary statistics from UK Biobank, FinnGen, and the COVID-19 Host Genetics Initiative with molecular quantitative trait locus (QTL) datasets. In total, 613 traits with significant heritability were retained and harmonized using the Unified Medical Language System. MOCR-DB integrates phenotype-to-phenotype analyses, including genetic correlation and Mendelian randomization, with phenotype-to-gene analyses based on QTL-informed summary-data-based Mendelian randomization analysis (SMR) within a single searchable and interactive framework. The platform supports exploration of cross-trait genetic correlations, putative causal relationships, and candidate functional gene associations. An AI-assisted module provides concise plain-language summaries to help contextualize statistical findings. As a case study, we examined obesity and COVID-19 severity, where genetically predicted obesity showed a stronger association with critical COVID-19 and lung eQTL-based SMR analyses revealed distinct immune- and neuronal-related molecular patterns across severity groups. MOCR-DB thus provides a unified and accessible resource for investigating shared genetic architectures and prioritized functional gene candidates across complex traits, supporting the generation of reproducible and biologically interpretable hypotheses. The database is publicly available at https://chenhongwei.net/public/MOCRdb/. Graphical Abstract Data resources, analytical framework, and interpretation in MOCR-DB The Multi-Omics Causal Resource Database (MOCR-DB) integrates large-scale GWAS summary statistics and molecular QTL datasets to provide a unified framework for genetic correlation, causal inference, and functional mediation. Data resources include GWAS summary statistics from UK Biobank, FinnGen, and the COVID-19 Host Genetics Initiative, together with 53 xQTL datasets across 49 tissues (eQTL, mQTL, sQTL, and caQTL). The analytical framework combines linkage disequilibrium score regression (LDSC) for estimating heritability and cross-trait genetic correlation, Mendelian randomization (MR) to infer potential causal relationships between traits, and summary-data-based Mendelian randomization (SMR) to identify tissue-specific functional genes. Results are presented through interactive genetic network searches that link diseases, biomarkers, lifestyle factors, and molecular traits via correlation, causality, and functional annotation. An AI-assisted module further facilitates causal and functional interpretation by summarizing complex results from LDSC, MR, and SMR analyses into accessible biological insights. Together, MOCR-DB provides systematic exploration of shared genetic architectures and functional mediators across complex human traits.

Seasonal influenza viruses accumulate antigenic changes, eroding population immunity and necessitating recurrent vaccine updates. Hemagglutination inhibition (HI) assays are the standard for measuring antigenic relationships between circulating and vaccine strains; however, their limited throughput constrains the scale and timeliness of surveillance. Here, we present fluProfiler, a foundation-model-based framework that learns a stable mapping from viral sequences to antigenic space and uses this representation to support influenza antigenic prediction, vaccine strain evaluation, and diversity-driven sampling. fluAgPredictor aligns hemagglutinin (HA) and neuraminidase (NA) sequence representations with HI-derived antigenicity, enabling accurate and consistent inference of pairwise antigenic distances across surveillance-aligned evaluation settings. Without prior annotation of antigenic sites, it identifies key residues in immunodominant epitopes and reveals the cooperative contributions of HA and NA to antigenic variation. Building on this antigenic-space representation, fluVacSelector provides antigenic coverage scores that are concordant with World Health Organization (WHO) vaccine recommendations while also flagging potential candidates that may offer broader coverage ahead of formal consultations. fluAgEnhancer further leverages the same representation to prioritize antigenically informative and diverse strains for experimental characterization, achieving comparable predictive accuracy with approximately 25% fewer HI measurements than random sampling. Together, these modules provide a high-throughput and interpretable complement to HI testing, converting routine genomic surveillance into a more proactive, data-driven support system for antigenic monitoring and vaccine strain selection.

Influenza’s pandemic threat is driven by antigenic drift, which limits the efficacy of conventional vaccines. To address this challenge, we established a clinical serum-anchored computational design pipeline for a broad-spectrum multi-epitope mRNA vaccine (MEMV), bridging the gap between pure in silico design and clinical applicability. Using 36 longitudinal sera (d0/d28/d365) from 12 well-characterized human cohorts (6 vaccine recipients and 6 influenza patients) and high-density antibody-peptide microarrays, we empirically identified 12 immunodominant B-cell linear epitopes from the nucleoprotein (NP) of influenza A (H1N1/H3N2) and B viruses. These experimentally validated epitopes were combined with in silico-predicted conserved helper T-lymphocyte (HTL)/cytotoxic T-lymphocyte (CTL) epitopes (from NP/HA/NA) to construct MEMVs candidates, ensuring high antigenicity, non-toxicity, and 95.63% global HLA coverage. Molecular docking and 100 ns molecular dynamics (MD) simulations confirmed favorable conformational compatibility between MEMVs and Toll-like receptor 3 (TLR3) in silico immunization via C-ImmSim predicted robust B/T-cell responses and protective cytokine (IFN-γ/IL-10) production. Collectively, this pipeline shortens the preliminary design cycle for influenza vaccines, provides a standard epitope-combination strategy, and offers direct targets for follow-up in vitro/in vivo experiments.

Currently, mpox virus (MPXV) continues to pose a global public health challenge, with immunocompromised individuals often exhibiting more severe clinical symptoms. This study screens three male mouse models (ICR, IFNAR1^-/-, SCID) and identifies SCID mice as the optimal model for modeling severe patient symptoms, including pneumonia, rash, and localized inflammation, which is applied to evaluate the antiviral efficacy of tecovirimat and cidofovir. Both drugs prevent systemic MPXV spread when administered within two days post virus exposure. Local antiviral efficacy differs between the two drugs, particularly after intradermal infection. Prolonged treatment up to 28 days post infection results in 100% survival of SCID mice but fails to effectively suppress localized rash and inflammatory swelling, suggesting that the drugs have limited impact at later stages of the disease. These findings indicate that the therapeutic efficacy of tecovirimat and cidofovir depends on the timing of intervention initiation and the stage of disease progression.

Influenza, a highly pathogenic infectious disease, causes nearly half a million deaths annually worldwide. Thus, effective vaccine‐based prevention and control are crucial. Although live attenuated influenza vaccines (LAIVs) can induce mucosal immunity, existing vaccines effectiveness remains relatively low, posing a significant threat to public health. Thus, we developed a novel mosaic H1N1 LAIV candidate by integrating mosaic antigen design with established LAIV technology. This vaccine incorporates most potential T‐cell epitopes of hemagglutinin and neuraminidase antigens into an attenuated master donor strain, ensuring safety and broad immunity. We compared it with commercial monovalent attenuated and inactivated vaccines in mice. The mosaic H1N1 LAIV induced robust cross‐reactive humoral and mucosal immune responses, enhanced antigen‐specific cellular immunity, and established tissue‐resident memory T and B cells in the respiratory tract. Challenge experiments confirmed its protective efficacy against homologous and heterologous strains. It provided complete protection against homologous strains with low epitope similarity and partial protection against the ancestral H3N2 virus. Our study highlights the mosaic H1N1 LAIV as an excellent universal vaccine candidate capable of inducing broad cross‐reactive immune responses and providing robust protection against distinct influenza A viruses, demonstrating a promising strategy to address the limitations of current commercial vaccines.

The interaction between influenza virus and other respiratory pathogens is increasingly recognized, yet the extent to which influenza vaccination alters the dynamics remains to be fully elucidated. Here we evaluate the associations between influenza vaccination and subsequent infection risk of 11 respiratory pathogens using a test-negative case-control study. Additionally, a dynamic “SVIRS×SIRS” model considering both direct and indirect pathways was constructed to explore the primary mechanisms underlying the observed associations. Among the 52,725 paediatric patients included, 11,895 (22.6%) had been vaccinated against influenza within 2 years of the test date. We identified robust associations between influenza vaccination and subsequent infection with influenza A virus (OR 0.52, 95% CI 0.47–0.58), influenza B virus (OR 0.22, 95% CI 0.19–0.26), and Mycoplasma pneumoniae (MP) (OR 1.24, 95% CI 1.20–1.27). The potential increase in human adenovirus and human metapneumovirus infection risk following influenza vaccination, in contrast to the decreased risk of human coronavirus and human respiratory syncytial virus, necessitates further research. No statistically significant associations were detected for the other four pathogens: chlamydia, human rhinovirus, human parainfluenza, and human bocavirus. Simulations using the dynamic model indicated that associations on MP was predominantly mediated by indirect pathway, i.e., the suppression of the influenza virus leaves a void for MP, rather than the direct side effect of the vaccine. These findings underscore that distinguishing between the direct and indirect impacts of vaccination is extremely important considering the currently fragile global confidence in vaccines. Furthermore, understanding these vaccine-pathogen interactions helps strengthen integrated prevention systems for multiple respiratory infectious diseases.