Human influenza A viruses escape antibody-mediated immunity through changes in the hemagglutinin (HA) and neuraminidase (NA) glycoproteins. HA antigenic evolution has been studied extensively, with more recent interest in NA due to its importance in influenza vaccine efficacy. Here, the antigenic properties of the NA of more than 300 A(H3N2) and A(H2N2) viruses isolated since 1957 were quantified with a NA inhibition enzyme-linked lectin assay and visualized using antigenic cartography, with follow-up molecular studies using recombinant viruses. The antigenic evolution of N2 NA was more gradual than that described for H3 HA, and antigenic changes in NA and HA were discordant. Multiple substitutions around the NA active site and tetramer lateral side that alter the charge, volume, or hydropathy of amino acids collectively determined antigenic properties. These data facilitate sequence-based genomic surveillance and inference of antigenic phenotypes from genotypes and offer opportunities to improve influenza vaccine effectiveness through increased focus on NA.

Highly pathogenic avian influenza viruses (HPAIVs) derive from H5 and H7 low pathogenic avian influenza viruses (LPAIVs). Although insertion of a furin-cleavable multibasic cleavage site (MBCS) in the hemagglutinin gene was identified decades ago as the genetic basis for the LPAIV-to-HPAIV transition, the mechanisms underlying the occurrence of insertion are unknown. Here, we show that transient H5 RNA structures, predicted to trap the influenza virus polymerase on purine-rich sequences, drive nucleotide insertions, providing empirical evidence of RNA structure involvement in MBCS acquisition. Introduction of H5-like sequences and structures into an H6 hemagglutinin resulted in MBCS-yielding insertions. Our results show that nucleotide insertions that underlie H5 HPAIV emergence result from an RNA structure-driven diversity-generating mechanism, which could also occur in other RNA viruses.

Germinal centers (GCs) are specialized sites within secondary lymphoid organs where B cells expand, are selected, and mature to produce high-quality antibodies. Their structural complexity makes them difficult to model in vitro. Here, we developed a human 3D lymphoid culture system combining lymphoid and stromal cells to better mimic GC environments than conventional 2D cultures. Tonsil cells were cultured with or without fibroblastic reticular cells (FRCs) in 2D or 3D hydrogels and stimulated with viral antigens or vaccines. FRC-supported 3D cultures significantly improved B and T cell survival and promoted reaggregation into follicle-like structures with. 3D FRC-supported co-cultures higher levels of antigen-specific antibodies, increased frequencies of S- or HA-specific B cells, and enhanced differentiation into antibody-secreting cells. Importantly, these cultures also showed reduced cell death and lower bystander activation and CXCR4 and CXCR5 expression on CD27+CD38+ B cells indicated GC-like polarization. Autologous and allogeneic FRCs performed comparably, supporting the scalability of the model for high-throughput applications. This 3D platform offers a more physiologically relevant system for studying human GC-associated immune responses and may facilitate mechanistic research and screening of vaccine immunogens and adjuvants in a controlled laboratory setting.

SUMMARY Influenza viruses evade vaccine and infection mediated immunity by accumulating mutations in their hemagglutinin (HA) protein. Predicting this evolution might be possible via selective mutational scanning (SMS) – the generation of many specific mutants of interest from currently circulating viruses and characterizing their escape potential and fitness with high accuracy (Mögling 2016). However, this task is challenging, even when focusing on a reduced set of key HA positions (Koel et al. 2013). Here we describe a high-throughput SMS method to address this challenge. Our approach consists of a three-stage pipeline: (1) a parallel optimized virus rescue process that generates balanced target mutant virus libraries (2) an assay to assess replicative fitness and neutralisation of these variants as a mixture, and (3) a bespoke statistical model to quantify statistically significant differences between these observables. We tested the pipeline on libraries of up to 134 variants finding excellent correlation to classical hemagglutination inhibition (HI) and plaque growth assays used to assess antigenic phenotype and replicative fitness respectively, as well as remarkable repeatability overall. Notably, the method reduces the timeline required to carry out such assessments with classical methods from about a year to several weeks. By enabling rapid and efficient characterization of influenza virus variants, this approach has the potential to greatly enhance surveillance efforts, transforming reactive monitoring into proactive forecasting.

BACKGROUND:Highly pathogenic avian influenza A(H5) viruses pose a pandemic threat, with a history of mammalian adaptation and zoonotic spillovers into humans. We aimed to determine whether pre-existing cross-reactive immune responses to A(H5) clade 2.3.4.4b influenza viruses detected between 2020 and 2024 are present in the general population. METHODS:We conducted an observational, cross-sectional study within the prospective Surveillance of Respiratory Viruses in Healthcare and Animal Workers in the Netherlands (SENTINEL) cohort, in which we analysed a subset of health-care workers aged 18 years or older who provided blood samples at a periodic study visit in August or September, 2024. Blood samples were analysed for influenza A(H5)-specific antibody binding, haemagglutination inhibition, Fc-effector functions, neuraminidase (NA) inhibition, and T-cell responses. FINDINGS:We included 107 health-care workers. Participants' median age was 50·0 years (IQR 40·0-58·0); 77 (72%) health-care workers were female, 29 (27%) were male, and one (1%) did not report their biological sex. Virus-specific antibodies were measured in 106 serum samples. Low-level binding antibodies directed against the A(H5) haemagglutinin (HA) head were detected in up to 28 individuals (depending on the antigen), but without haemagglutination inhibition activity. Nevertheless, we detected A(H5)-reactive antibodies with Fc-effector functions in all participants. Additionally, we observed high levels of antibodies with NA inhibition activity (geometric mean titre 208 [95% CI 153-284]) in up to 97% of the health-care workers against avian N1, and T-cell responses against HA and NA from A(H5) influenza viruses in 43-69% (46-74 of 107) health-care workers. A(H5)-specific responses correlated with immune responses targeting A(H1N1). INTERPRETATION:Together, our findings suggest that partial cross-reactive immunity to A(H5) influenza viruses exists in humans, likely induced by previous exposures to seasonal influenza viruses. This partial cross-reactive immunity might play an important role during future outbreaks, potentially by blunting disease severity. Characterising pre-existing baseline immunity is crucial for accurate pandemic risk assessment and preparedness planning. FUNDING:The Netherlands Organization for Health Research and Development, the EU's EU4Health programme DURABLE, the Dutch Ministry of Agriculture, Fisheries, Food Security and Nature, the Dutch Ministry of Health, Welfare and Sport, and the National Institute of Health-National Institute of Allergies and Infectious Diseases.

Highly pathogenic H5Ny influenza A viruses are causing unprecedented, season-independent outbreaks across avian and mammalian species, including dairy cattle, a novel reservoir. The sialoside-binding properties of influenza A hemagglutinin (HA) are strongly related to its ability to infect and transmit between hosts. Mucin-like O-glycans, omnipresent in respiratory tracts, have been understudied as viral receptors due to their complexity. To address this, we synthesized 25 O-linked glycans with diverse sialosides, including modifications by fucosides and sulfates. Our findings reveal that H5Ny 2.3.4.4b viruses bind core 3 sialyl-Lewisx and Sia-Gal-β3GalNAc, O-linked glycans not recognized by classical H5 or other avian viruses. By determining crystal structures, we resolved the structural features of four glycans in an H5 hemagglutinin (HA) from a 2016 2.3.4.4b virus. While these viruses do not bind human-type receptors, their broad receptor specificity enhances binding to human tracheal tissues, suggesting that O-glycan recognition could contribute to the continues spillover of this clade.

H5Nx viruses continue to wreak havoc in avian and mammalian species worldwide. The virus distinguishes itself by the ability to replicate to high titers and transmit efficiently in a wide variety of hosts in diverse climatic environments. Fortunately, transmission to and between humans is scarce. Yet, if such an event were to occur, it could spark a pandemic as humans are immunologically naïve to H5 viruses. A significant determinant of transmission to and between humans is the ability of the influenza A virus hemagglutinin (HA) protein to shift from an avian-type to a human-type receptor specificity. Here, we demonstrate that a 2016 2.3.4.4e virus HA can convert to human-type receptor binding via a single Q226L mutation, in contrast to a cleavage-modified 2016 2.3.4.4b virus HA. Using glycan arrays, X-ray structural analyses, tissue- and direct glycan binding, we show that L133a Δ and 227Q are vital for this phenotype. Thus, whereas the 2.3.4.4e virus HA only needs a single amino acid mutation, the modified 2016 2.3.4.4b HA was not easily converted to human-type receptor specificity.

Background Highly pathogenic avian influenza H5N1 viruses of the A/Goose/Guangdong/1/1996 lineage pose a global threat to wildlife, domestic animals, and humans. Cross-species transmission events to mammals, including humans, in the past 4 years highlight this threat. For influenza A viruses, crucial determinants of cross-species and intraspecies transmission to and among mammals include attachment to and replication in respiratory airway epithelial cells. Although these determinants have been studied for H5N1 viruses in the past, limited studies for clade 2.3.4.4b viruses exist. Therefore, the aim of this study was to determine the ability of recent clade 2.3.4.4b H5N1 viruses to attach to human respiratory tissues, to replicate in human airway epithelial cells and the associated immune response. Methods In this in-vitro study, we investigated three H5N1 clade 2.3.4.4b viruses (H5N1Gull2022, H5N1Polecat2022, and H5N1Bovine2024) in comparison with previously studied 2.1.3.2 H5N1 (H5N12005) and a seasonal H3N2 virus. First, we compared virus attachment patterns by virus histochemistry. Second, we investigated the infection and replication efficiency, and innate immune responses in infected human respiratory epithelium in vitro. Third, we measured polymerase complex activity using a minigenome assay. Findings Clade 2.3.4.4b viruses and H5N12005 virus differed by five amino acids located near the receptor binding site of the haemagglutinin. All clade 2.3.4.4b viruses attached more efficiently to cells of the human upper and lower respiratory tract compared with H5N12005 virus. All clade 2.3.4.4b viruses replicated in human nasal and tracheobronchial respiratory epithelium cultures. In the tracheobronchial respiratory epithelium cultures, H5N1Gull2022 virus replicated more efficiently than H5N12005 virus (p=0·0050) and reached titres similar to H3N22003 virus. Polymerase complex activity of H5N1Gull2022 virus was not significantly different from that of H5N12005 and was significantly lower compared with H3N22003 virus (p≤0·0001). Infection with H5N1Gull2022 virus induced a broader antiviral immune response than H5N12005 virus. Interpretation Clade 2.3.4.4b H5N1 viruses have phenotypic characteristics that are different from a clade 2.1.3.2 H5N12005 virus. The ability of clade 2.3.4.4b viruses to attach to and replicate in respiratory epithelium likely contributes to an increased risk for both human infection and virus adaptation to humans. Funding The EU, the Dutch Research Council, the Netherlands Organization for Health Research and Development, and the Dutch Ministries of Agriculture, Fisheries, Food Security and Nature, and Health, Welfare and Sport.

Highly pathogenic avian influenza A(H5) viruses globally impact wild and domestic birds, and have caused severe infections in mammals, including humans, underscoring their pandemic potential1-5. The antigenic evolution of the A(H5) haemagglutinin (HA) poses challenges for pandemic preparedness and vaccine design6. Here the global antigenic evolution of the A(H5) HA was captured in a high-resolution antigenic map. The map was used to design immunogenic and antigenically central vaccine HA antigens, eliciting antibody responses that broadly cover the A(H5) antigenic space. In ferrets, a central antigen protected as well as homologous vaccines against heterologous infection with two antigenically distinct viruses. This work showcases the rational design of subtype-wide influenza A(H5) pre-pandemic vaccines and demonstrates the value of antigenic maps for the evaluation of vaccine-induced immune responses through antibody profiles.

Finland faced an outbreak of highly pathogenic clade 2.3.4.4b A(H5N1) avian influenza in 2023, which spread from wild birds to fur farms. Vaccinations of at-risk individuals began in June 2024 using the MF59-adjuvanted inactivated A(H5N8) vaccine (Seqirus; A/Astrakhan/3212/2020, clade 2.3.4.4b). Here, in an observational study, we assessed vaccine-induced immune responses in occupational at-risk individuals participating in the phase IV trial, including virus-specific antibody (n = 39 individuals) and T-cell (n = 18 individuals) responses. Vaccination elicited functional antibodies against the vaccine virus and two heterologous clade 2.3.4.4b strains associated with outbreaks on Finnish fur farms and dairy cattle in the United States. Among previously unvaccinated individuals, seroprotection rates against the vaccine virus were 83% (95% CI 70-97%) by microneutralization assay (titre ≥20) and 97% (90-100%) by haemagglutination inhibition assay (titre ≥40). In those previously vaccinated against avian influenza, a single dose induced seroprotection. A(H5N8)-specific memory CD4+ T-cell responses were detectable, with ~5-fold increase in IFNγ secretion after two doses. These results demonstrate that the vaccine probably provides cross-protection against circulating H5 clade 2.3.4.4b viruses. EU Clinical Trial Number 2023-509178-44-00.

Highly pathogenic H5Nx influenza A viruses are causing unprecedented, season-independent outbreaks across avian and mammalian species, including dairy cattle, a novel reservoir. The sialoside-binding properties of influenza A hemagglutinin (HA) are strongly related to its ability to infect and transmit between hosts. Mucin-like O-glycans, omnipresent in respiratory tracts, have been understudied as viral receptors due to their complexity. To address this, we synthesized 25 O-linked glycans with diverse sialosides, including modifications by fucosides and sulfates. Our findings reveal that H5Nx 2.3.4.4b viruses uniquely bind core 3 sialyl-Lewisx and Sia-Gal-β3GalNAc, glycans not recognized by classical H5 or other avian viruses. By determining its crystal structure, we resolved the structural features of both structures in an H5 hemagglutinin (HA) from a 2016 2.3.4.4b virus. While these viruses do not bind human-type receptors, their promiscuous receptor specificity enhances binding to human tracheal tissues, suggesting that O-glycan recognition contributes to their zoonotic potential. ### Competing Interest Statement The authors have declared no competing interest. Austrian Science Fund, , FWF, J-4260-B21 ICRAD, , °862605 (Flu-Switch) NWO, TOP-PUNT, , 718.015.003 National Institutes of Health NIAID Centers of Excellence for Influenza Research and Response, , 75N93021C00015 / PENN CEIRR

There is an increasing need for reproducible long-term in vitro primary cell culture systems that are representative of the avian respiratory tract to study pathogens like highly pathogenic avian influenza viruses (HPAIVs), which threaten poultry, wildlife, and human health. Self-renewing organoid cultures allow for long-term culture, due to the presence of tissue-resident stem cells, and can approximate the in vivo cellular diversity and organization of tissues. Efforts to establish avian organoid cultures have been limited to the intestinal tract. Here, we describe the isolation and long-term culture of chicken tracheal organoids (CTOs). The CTO cultures were passaged for three to four months and cryopreserved at different stages. Mucociliary differentiation of CTOs was promoted by culture at air–liquid-interface, after which the pseudostratified epithelial cell layer of the avian trachea was recapitulated, including ciliated, goblet, and basal cells. Inoculation of CTO-derived 2D cultures with low pathogenic avian influenza viruses (LPAIVs) and HPAIVs showed that the appropriate receptors, as confirmed by virus histochemistry, and proteases to sustain multi-cycle replication of LPAIVs were expressed and that HPAIVs preferentially disseminated to the endothelium of epithelial/endothelial co-culture systems. Taken together, CTOs represent a useful tool for research on the avian respiratory tract, and their application will generate new insights into host–pathogen interactions, including HPAIV tropism.

The structural complexity of secondary lymphoid organs (SLOs) and their role in shaping antigen-specific B cell responses, pose significant challenges in modeling human germinal center (GC) response in vitro . A human 3D lymphoid model incorporating lymphoid and stromal cell types recapitulates key immune and structural features, enabling the study of antigen-specific B and T cell interactions beyond current 2D culture limitations. In this study, human tonsil cells were cultured with and without tonsil-derived fibroblastic reticular cells (FRCs) either in 2D or within a 3D PEG-4MAL hydrogel culture. Antigen-specific B cell responses in co-cultures were studied by comparing unstimulated cultures to stimulation with antigen (SARS-CoV-2 spike (S) or Influenza hemagglutinin (HA), both with or without adjuvant R848), S-nanoparticles and influenza vaccines. Combination of FRCs with the 3D matrix significantly improved B and T cell survival and facilitated reaggregation into follicle-like structures. Antigen-specific responses were most pronounced in 3D FRC-supported co-cultures, with increasing S- or HA-specific B cell frequencies, antibody-secreting cell differentiation, and secretion of antigen-specific antibodies. Importantly, cell death and unspecific bystander activation was lowest in 3D FRC-supported cultures. Additionally, GC-associated chemokine receptors CXCR4 and CXCR5 showed distinct expression patterns on CD27⁺CD38⁺ B cells, reflecting GC-like dark and light zone organization typically observed in SLOs in vivo . Autologous and allogeneic FRC-supported cultures yielded comparable results, demonstrating the platform’s potential for high-throughput applications. The 3D FRC-supported lymphoid cultures offer a physiologically relevant platform for studying human GC responses in vitro , supporting mechanistic research into adaptive immunity and enabling the screening of vaccine immunogens and adjuvants in a controlled setting. ### Competing Interest Statement The authors have declared no competing interest.

Since their first detection in 1996, highly pathogenic avian influenza viruses with H5 haemagglutinin of the A/Goose/Guangdong/1/1996 (Gs/Gd) lineage have caused outbreaks in domestic and wild animals associated with mass morbidity and mortality, and economic losses as well as sporadic human infections. These viruses have spread to hosts across the European, Asian, African, and North and South American continents, and most recently Antarctica, representing a major threat to wildlife, domestic animals and humans. Owing to continuous circulation in poultry, Gs/Gd lineage viruses have diversified into numerous distinct genetic and antigenic (sub)clades, and genetic diversity has further increased by extensive reassortment with low pathogenic avian influenza viruses of wild birds. In this Review, we discuss the historical emergence of Gs/Gd lineage viruses and their evolution and geographical spread. An overview of the major determinants of host range and cross-species transmission is provided to summarize phenotypic changes that may signal increased zoonotic or pandemic risks. The recent unusual outbreaks in wild carnivorous mammals and dairy cows is discussed, as well as the changing risk to humans. Countermeasures and mitigation strategies are described from the One Health perspective for future (pre-)pandemic preparedness. In this Review, Fouchier and colleagues explore the evolution, spread and zoonotic risks of highly pathogenic H5 avian influenza viruses. They highlight recent unusual outbreaks, discuss future preparedness strategies from a One Health perspective, and describe countermeasures to mitigate the risks.

In 2023, Finland faced an outbreak of highly pathogenic avian influenza caused by clade 2.3.4.4b A(H5N1) viruses, which spread from wild birds to fur farms. Vaccinations of individuals at-risk, such as fur and poultry farm workers, veterinarians, and laboratory workers, began in June 2024 using the MF59-adjuvanted inactivated (H5N8) vaccine manufactured by Seqirus (based on clade 2.3.4.4b A/Astrakhan/3212/2020). We investigated antibody responses following a two-dose vaccination regimen in 39 subjects. Vaccination induced comparable levels of functional antibodies both against the vaccine virus and two clade 2.3.4.4b viruses, either associated with outbreaks in fur animals in Finland or cattle in the United States. Upon two doses of the vaccine for previously unvaccinated people, the seroprotection rate against the vaccine virus was 83 % (95 % CI 70-97 %, titer ≥20) and 97 % (95 % CI 90-100 %, titer ≥40) using microneutralization or hemagglutinin inhibition assays, respectively. In a subset of previously H5-vaccinated individuals, the first dose already led to seroprotective titers, indicative of immunological recall. These data show that the vaccine is expected to confer cross-protection against currently circulating H5 clade 2.3.4.4b viruses. ### Competing Interest Statement Hanna Nohynek is member of the National Immunization Technical Advisory Group, THL, Finland and chair of the WHO Strategic Advisory Group of Experts. Ritva Syrjanen has acted or acts as sub-investigator in a COVID-19 study sponsored by Pfizer, pneumococcal carriage study sponsored by Merck Sharp & Dohme and influenza, pertussis and meningitis studies sponsored by Sanofi Pasteur, not related to this work. Her current affiliation FVR, Finnish Vaccine Research conducts clinical trials and studies sponsored by almost all vaccine providers, not related to this work. Merit Melin is member of the National Immunization Technical Advisory Group, THL, Finland. ### Clinical Protocols ### Funding Statement The Finnish Institute for Health and Welfare (THL) funded the clinical vaccine study. For the immunological studies THL and University of Turku received funding from the Academy of Finland, Avian and seasonal influenza vaccine-induced humoral and cell-mediated immune responses, Decision number 362192 and 362193. ### Author Declarations I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained. Yes The details of the IRB/oversight body that provided approval or exemption for the research described are given below: The study was conducted according to the guidelines of the Declaration of Helsinki and was submitted for evaluation through the EU Clinical Trial Information System. The study has been approved by the National Committee on Medical Research Ethics (decision number TUKIJA/7/2024) and received authorization from the Finnish Medicines Agency Fimea (EU CT number 2023-509178-44-00). I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals. Yes I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance). Yes I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable. Yes Data subject to third party restrictions. The data that support the findings of this study are available from Finnish Social and Health Data Permit Authority Findata. Restrictions apply to the availability of these data, which were used under license for this study under informed consent form. Anonymized data are available after permission by Findata.

Emerging severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) Omicron subvariants are considered antigenically distinct, defined by escape from neutralising antibodies. To assess the antigenic distance between variants and predict immunity gaps in the population, neutralisation data are used to construct antigenic maps and antibody landscapes, respectively.1Wilks SH Mühlemann B Shen X et al.Mapping SARS-CoV-2 antigenic relationships and serological responses.Science. 2023; 382eadj0070Crossref PubMed Scopus (10) Google Scholar However, these tools do not consider T-cell immunity. Spike (S)-specific T cells are cross-reactive between emerging SARS-CoV-2 variants, and virus evolution has thus far not led to antigenic variation that facilitates complete T-cell escape.2Tarke A Coelho CH Zhang Z et al.SARS-CoV-2 vaccination induces immunological T cell memory able to cross-recognize variants from Alpha to Omicron.Cell. 2022; 185: 847-859Summary Full Text Full Text PDF PubMed Scopus (515) Google Scholar Additionally, contrasting with neutralising antibodies, SARS-CoV-2-specific T cells can target multiple (internal) viral proteins besides S. Since these internal proteins are less prone to variation, T cells are more resilient to emerging variants. SARS-CoV-2-specific T cells were previously associated with protection from symptomatic coronavirus disease-2019 (COVID-19) in children who were vaccinated, suggesting that T-cell immunity is clinically relevant.3Zhong Y Kang AYH Tay CJX et al.Correlates of protection against symptomatic SARS-CoV-2 in vaccinated children.Nat Med. 2024; 30: 1373-1383Crossref PubMed Scopus (0) Google Scholar Consequently, profiling and landscaping virus-specific T-cell responses would be a useful addition to antibody-based mapping and landscaping, to more accurately identify immunity gaps on a population level. To generate extensive T-cell profiles before and after monovalent XBB.1.5 vaccination, we used the IFNγ release assay (IGRA). Small volumes of whole blood were stimulated with commercially available overlapping peptide pools covering the S protein from 12 different coronaviruses: nine SARS-CoV-2 variants (D614G, Delta, BA.1, BA.2, BA.5, BQ.1.1, XBB.1.5, EG.5.1, and BA.2·86), two common cold coronaviruses (CCCs; NL63 and OC43), and Middle East respiratory syndrome-related coronavirus (MERS-CoV; appendix 1 p 2). Because EG.5.1 was not circulating at the start of this study, EG.5.1 responses have only been assessed in a subset of donors. Moreover, since the BA.2.86 overlapping peptide pool was not commercially available at the time of the study, we used an in-house designed peptide pool that was used in a separate analysis. By comparison with T-cell assays that use peripheral blood mononuclear cells, the IGRA allows for easy multiplexing of numerous antigens. We generated T-cell profiles from a well characterised cohort of health-care workers4Tan NH Geers D Sablerolles RSG et al.Immunogenicity of bivalent omicron (BA.1) booster vaccination after different priming regimens in health-care workers in the Netherlands (SWITCH ON): results from the direct boost group of an open-label, multicentre, randomised controlled trial.Lancet Infect Dis. 2023; 23: 901-913Summary Full Text Full Text PDF PubMed Scopus (0) Google Scholar before (n=164) and 28 days after (n=57) monovalent XBB.1.5 vaccination (appendix 1 pp 3, 9 and appendix 2). In general, S-specific T cells cross-recognised all SARS-CoV-2 variants assessed, both before and after vaccination (figure A and B; appendix p 4). Pre-vaccination, cross-reactivity with several Omicron subvariants was reduced (at approximately 80% when expressed relative to ancestral reactivity; figure C), indicative of polyclonal T-cell escape. This finding corresponds with previous observations reporting that 85% of T-cell reactivity with Omicron BA.1 was maintained.2Tarke A Coelho CH Zhang Z et al.SARS-CoV-2 vaccination induces immunological T cell memory able to cross-recognize variants from Alpha to Omicron.Cell. 2022; 185: 847-859Summary Full Text Full Text PDF PubMed Scopus (515) Google Scholar Reactivity with the Omicron BA.5 subvariant was the most reduced pre-vaccination: 75% relative to ancestral in the total cohort, but only 71% relative to ancestral in health-care workers who had received a previous bivalent Omicron BA.1 vaccine. In addition, reactivity with the Omicron BA.5 subvariant was 83% relative to ancestral in health-care workers who received a bivalent Omicron BA.5 vaccine (appendix p 5). 28 days post-vaccination, T-cell responses to all SARS-CoV-2 variants were boosted, including to newer variants (figure A, B, and appendix pp 4, 6). Polyclonal T-cell escape was less apparent post-vaccination (Omicron subvariant reactivity at 87–92% of ancestral), showing that monovalent XBB.1.5 vaccination led to an overall broadening of T-cell responses (figure C). The increase in S-specific T-cell responses was maintained when exclusively analysing health-care workers from whom paired pre-vaccination and post-vaccination samples were available (appendix pp 6, 7). Variant-specific T-cell responses correlated well with ancestral-specific T-cell responses, and de novo induction of variant-specific T-cells by monovalent vaccination was not detected (appendix p 8). In a subset of individuals, ancestral-specific T-cell responses correlated less with variant-specific T-cell responses, possibly indicating low-level T-cell escape that could be HLA-dependent. Reduced presentation of mutated T-cell epitopes could explain some of the individual differences observed. T-cell responses to the Omicron BA.2.86 subvariant were higher than ancestral-specific responses pre-vaccination, but not post-vaccination (appendix pp 4, 6). We cannot exclude that these divergent results are a consequence of different peptide pools (in-house produced vs commercially produced), although a separate ancestral control was included to correct for this. Alternatively, it is possible that mutations in the BA.2.86 S protein led to the exposure of immunogenic T-cell epitopes, as previously speculated.5Sette A Sidney J Grifoni A Pre-existing SARS-2-specific T cells are predicted to cross-recognize BA.2.86.Cell Host Microbe. 2024; 32: 19-24Summary Full Text Full Text PDF PubMed Google Scholar S-specific T-cell responses targeting the CCCs NL63 and OC43 were found in 116 (70%) of 166 donors and 55 (33%) of 166 donors pre-vaccination, and in 46 (78%) of 59 donors and 29 (49%) of 59 donors post-vaccination, respectively. Compared with stimulation with SARS-CoV-2 S overlapping peptide pool, the absolute concentrations of IFNγ after stimulation with CCC S overlapping peptide pools were low, probably because of the high number of total exposures to SARS-CoV-2 in the population. Both geometric mean NL63-specific and OC43-specific T-cell responses were boosted by monovalent XBB.1.5 vaccination in paired donors (appendix p 7), potentially because of boosting cross-reactive T cells.6Mateus J Grifoni A Tarke A et al.Selective and cross-reactive SARS-CoV-2 T cell epitopes in unexposed humans.Science. 2020; 370: 89-94Crossref PubMed Scopus (784) Google Scholar Although we did not expect to detect MERS-CoV-specific T cells, low concentrations of IFNγ were observed in some health-care workers pre-vaccination (12 [7%] of 166) and post-vaccination (10 [17%] of 59) that were not boosted by vaccination (figure A, appendix p 7). Because MERS-CoV does not cause infections in Europe, this finding suggests that previous coronavirus infections or COVID-19 vaccinations induced MERS-CoV cross-reactive T-cell responses in these individuals. The T-cell profiling described here shows that emerging Omicron subvariants that escape antibody-mediated neutralisation are still recognised by cross-reactive T-cells, and should not be considered antigenically distinct by definition. Furthermore, vaccination with a monovalent XBB.1.5 vaccine boosted T-cell responses, including T cells that were cross-reactive with variants that emerged after XBB.1.5. By combining T-cell profiles with antibody landscapes, it will become possible to better identify COVID-19 immunity gaps in the population and potentially guide the selection of vaccine antigens. AS is a consultant for AstraZeneca, Calyptus Pharmaceuticals, Darwin Health, EmerVax, EUROIMMUN, F Hoffman-La Roche, Fortress Biotech, Gilead Sciences, Granite Bio, Gritstone Oncology, Guggenheim Securities, Moderna, Pfizer, RiverVest Venture Partners, and Turnstone Biologics. AG is a consultant for Pfizer. La Jolla Institute for Immunology has filed for patent protection for various aspects of T-cell epitope and vaccine design work. All other authors declare no competing interests. Download .pdf (1.0 MB) Help with pdf files Supplementary appendix 1 Download .xlsx (.07 MB) Help with xlsx files Supplementary appendix 2

Background: Highly pathogenic avian influenza (HPAI) H5 viruses of the A/Goose/Guangdong/1/1996 (GsGd) lineage pose significant global risks to wildlife, domestic animals, and humans. Recent cross-species transmission events to mammals, including humans, highlight this risk. Critical determinants for cross-species and intra-species transmission include the ability to attach to and replicate in respiratory epithelial cells. Although these factors have been studied for HPAI H5N1 viruses in the past, limited studies are available for currently circulating strains. Methods: We compared level of adaptation to human respiratory tract of a HPAI H5N1 clade 2.3.4.4b (H5N1.2022) virus with those of well characterized HPAI H5N1 clade 2.1.3.2 (H5N1.2005) and seasonal H3N2.2003 viruses by three methods. First, we compared pattern of virus attachment by virus histochemistry. Second, we compared efficiency of infection and replication, as well as innate immune responses in human respiratory epithelium in vitro. Lastly, we compared polymerase complex activity in a minigenome assay. Findings: The H5N1.2022 virus attached more abundantly to and replicated more efficiently in cells of the human respiratory tract compared to H5N1.2005 and H3N2.2003 viruses. This increased replication was not associated with an increased polymerase activity of H5N1.2022 virus compared to H3N2.2003 virus. The efficient replication of H5N1.2022 virus infection induced a robust innate immune response almost comparable to H3N2.2003. Interpretation: The pattern of virus attachment and replication efficiency of a HPAI H5N1.2022 virus resembled that of H3N2.2003 virus more closely than a HPAI H5N12005. This could contribute to an increased risk for both human infection and virus adaptations to humans. ### Competing Interest Statement The authors have declared no competing interest.

Highly pathogenic avian influenza viruses (HPAIVs) cause severe disease and high fatality in poultry1. They emerge exclusively from H5 and H7 low pathogenic avian influenza viruses (LPAIVs)2. Although insertion of a furin-cleavable multibasic cleavage site (MBCS) in the hemagglutinin gene was identified decades ago as the genetic basis for LPAIV-to-HPAIV transition3,4, the exact mechanisms underlying said insertion have remained unknown. Here we used an innovative combination of bioinformatic models to predict RNA structures forming around the influenza virus RNA polymerase during replication, and circular sequencing5 to reliably detect nucleotide insertions. We show that transient H5 hemagglutinin RNA structures predicted to trap the polymerase on purine-rich sequences drive nucleotide insertions characteristic of MBCSs, providing the first strong empirical evidence of RNA structure involvement in MBCS acquisition. Insertion frequencies at the H5 cleavage site were strongly affected by substitutions in flanking genomic regions altering predicted transient RNA structures. Introduction of H5-like cleavage site sequences and structures into an H6 hemagglutinin resulted in MBCS-yielding insertions never observed before in H6 viruses. Our results demonstrate that nucleotide insertions that underlie H5 HPAIV emergence result from a previously unknown RNA-structure-driven diversity-generating mechanism, which could be shared with other RNA viruses.

Highly pathogenic avian influenza viruses (HPAIVs) emerge from H5 and H7 low pathogenic avian influenza viruses (LPAIVs), most frequently upon insertions of nucleotides coding for basic amino acids at the cleavage site (CS) of the hemagglutinin (HA). The exact molecular mechanism(s) underlying this genetic change and reasons underlying the restriction to H5 and H7 viruses remain unknown. Here, we developed a novel experimental system based on frame repair through insertions or deletions (indels) of HAs with single nucleotide deletions. Indels were readily detected in a consensus H5 LPAIV CS at low frequency, which was increased upon the introduction of only one substitution leading to a longer stretch of adenines at the CS. In contrast, we only detected indels in H6 when multiple nucleotide substitutions were introduced. These data show that nucleotide sequence is a key determinant of insertions in the HA CS, and reveal novel insights about the subtype-specificity of HPAIV emergence.

Highly pathogenic avian influenza A(H5) viruses globally impact wild and domestic birds, and mammals, including humans, underscoring their pandemic potential. The antigenic evolution of the A(H5) hemagglutinin (HA) poses challenges for pandemic preparedness and vaccine design. Here, the global antigenic evolution of the A(H5) HA was captured in a high-resolution antigenic map. The map was used to engineer immunogenic and antigenically central vaccine HA antigens, eliciting antibody responses that broadly cover the A(H5) antigenic space. In ferrets, a central antigen protected as well as homologous vaccines against heterologous infection with two antigenically distinct viruses. This work showcases the rational design of subtype-wide influenza A(H5) pre-pandemic vaccines and demonstrates the value of antigenic maps for the evaluation of vaccine-induced immune responses through antibody profiles.