Opportunities and challenges for T cell-based influenza vaccines.

Vaccination remains our main defence against influenza, which causes substantial annual mortality and poses a serious pandemic threat. Influenza virus evades immunity by rapidly changing its surface antigens but, even when the vaccine is well matched to the current circulating virus strains, influenza vaccines are not as effective as many other vaccines. Influenza vaccine development has traditionally focused on the induction of protective antibodies, but there is mounting evidence that T cell responses are also protective against influenza. Thus, future vaccines designed to promote both broad T cell effector functions and antibodies may provide enhanced protection. As we discuss, such vaccines present several challenges that require new strategic and economic considerations. Vaccine-induced T cells relevant to protection may reside in the lungs or lymphoid tissues, requiring more invasive assays to assess the immunogenicity of vaccine candidates. T cell functions may contain and resolve infection rather than completely prevent infection and early illness, requiring vaccine effectiveness to be assessed based on the prevention of severe disease and death rather than symptomatic infection. It can be complex and costly to measure T cell responses and infrequent clinical outcomes, and thus innovations in clinical trial design are needed for economic reasons. Nevertheless, the goal of more effective influenza vaccines justifies renewed and intensive efforts.

Antigenic drift and subtype interference shape A(H3N2) epidemic dynamics in the United States.

Influenza viruses continually evolve new antigenic variants, through mutations in epitopes of their major surface proteins, hemagglutinin (HA) and neuraminidase (NA). Antigenic drift potentiates the reinfection of previously infected individuals, but the contribution of this process to variability in annual epidemics is not well understood. Here we link influenza A(H3N2) virus evolution to regional epidemic dynamics in the United States during 1997-2019. We integrate phenotypic measures of HA antigenic drift and sequence-based measures of HA and NA fitness to infer antigenic and genetic distances between viruses circulating in successive seasons. We estimate the magnitude, severity, timing, transmission rate, age-specific patterns, and subtype dominance of each regional outbreak and find that genetic distance based on broad sets of epitope sites is the strongest evolutionary predictor of A(H3N2) virus epidemiology. Increased HA and NA epitope distance between seasons correlates with larger, more intense epidemics, higher transmission, greater A(H3N2) subtype dominance, and a greater proportion of cases in adults relative to children, consistent with increased population susceptibility. Based on random forest models, A(H1N1) incidence impacts A(H3N2) epidemics to a greater extent than viral evolution, suggesting that subtype interference is a major driver of influenza A virus infection dynamics, presumably via heterosubtypic cross-immunity.

A Bayesian approach to incorporate structural data into the mapping of genotype to antigenic phenotype of influenza A(H3N2) viruses.

Surface antigens of pathogens are commonly targeted by vaccine-elicited antibodies but antigenic variability, notably in RNA viruses such as influenza, HIV and SARS-CoV-2, pose challenges for control by vaccination. For example, influenza A(H3N2) entered the human population in 1968 causing a pandemic and has since been monitored, along with other seasonal influenza viruses, for the emergence of antigenic drift variants through intensive global surveillance and laboratory characterisation. Statistical models of the relationship between genetic differences among viruses and their antigenic similarity provide useful information to inform vaccine development, though accurate identification of causative mutations is complicated by highly correlated genetic signals that arise due to the evolutionary process. Here, using a sparse hierarchical Bayesian analogue of an experimentally validated model for integrating genetic and antigenic data, we identify the genetic changes in influenza A(H3N2) virus that underpin antigenic drift. We show that incorporating protein structural data into variable selection helps resolve ambiguities arising due to correlated signals, with the proportion of variables representing haemagglutinin positions decisively included, or excluded, increased from 59.8% to 72.4%. The accuracy of variable selection judged by proximity to experimentally determined antigenic sites was improved simultaneously. Structure-guided variable selection thus improves confidence in the identification of genetic explanations of antigenic variation and we also show that prioritising the identification of causative mutations is not detrimental to the predictive capability of the analysis. Indeed, incorporating structural information into variable selection resulted in a model that could more accurately predict antigenic assay titres for phenotypically-uncharacterised virus from genetic sequence. Combined, these analyses have the potential to inform choices of reference viruses, the targeting of laboratory assays, and predictions of the evolutionary success of different genotypes, and can therefore be used to inform vaccine selection processes.

Global Pandemic Preparedness: Optimizing Our Capabilities and the Influenza Experience.

The coronavirus disease 2019 (COVID-19) pandemic has prompted rapid investigation and deployment of vaccine platforms never before used to combat human disease. The severe impact on the health system and the high economic cost of non-pharmaceutical interventions, such as lockdowns and international border closures employed to mitigate the spread of COVID-19 prior to the arrival of effective vaccines, have led to calls for development and deployment of novel vaccine technologies as part of a "100-day response ambition" for the next pandemic. Prior to COVID-19, all of the pandemics (excluding HIV) in the past century have been due to influenza viruses, and influenza remains one of the most likely future pandemic threats along with new coronaviruses. New and emerging vaccine platforms are likely to play an important role in combatting the next pandemic. However, the existing well-established, proven platforms for seasonal and pandemic influenza manufacturing will also continue to be utilized to rapidly address the next influenza threat. The field of influenza vaccine manufacturing has a long history of successes, including approval of vaccines within approximately 100 days after WHO declaration of the A(H1N1) 2009 influenza pandemic. Moreover, many advances in vaccine science and manufacturing capabilities have been made in the past decade to optimize a rapid and timely response should a new influenza pandemic threat emerge.

Temporal and Gene Reassortment Analysis of Influenza C Virus Outbreaks in Hong Kong, SAR, China.

From 2014 to week 07/2020 the Centre for Health Protection in Hong Kong conducted screening for influenza C virus (ICV). A retrospective analysis of ICV detections to week 26/2019 revealed persistent low-level circulation with outbreaks occurring biennially in the winters of 2015 to 2016 and 2017 to 2018 (R. S. Daniels et al., J Virol 94:e01051-20, 2020, https://doi.org/10.1128/JVI.01051-20). Here, we report on an outbreak occurring in 2019 to 2020, reinforcing the observation of biennial seasonality in Hong Kong. All three outbreaks occurred in similar time frames, were subsequently dwarfed by seasonal epidemics of influenza types A and B, and were caused by similar proportions of C/Kanagawa/1/76 (K)-lineage and C/São Paulo/378/82 S1- and S2-sublineage viruses. Ongoing genetic drift was observed in all genes, with some evidence of amino acid substitution in the hemagglutinin-esterase-fusion (HEF) glycoprotein possibly associated with antigenic drift. A total of 61 ICV genomes covering the three outbreaks were analyzed for reassortment, and 9 different reassortant constellations were identified, 1 K-lineage, 4 S1-sublineage, and 4 S2-sublineage, with 6 of these being identified first in the 2019-1920 outbreak (2 S2-lineage and 4 S1-lineage). The roles that virus interference/enhancement, ICV persistent infection, genome evolution, and reassortment might play in the observed seasonality of ICV in Hong Kong are discussed. IMPORTANCE Influenza C virus (ICV) infection of humans is common, with the great majority of people being infected during childhood, though reinfection can occur throughout life. While infection normally results in "cold-like" symptoms, severe disease cases have been reported in recent years. However, knowledge of ICV is limited due to poor systematic surveillance and an inability to propagate the virus in large amounts in the laboratory. Following recent systematic surveillance in Hong Kong SAR, China, and direct ICV gene sequencing from clinical specimens, a 2-year cycle of disease outbreaks (epidemics) has been identified, with gene mixing playing a significant role in ICV evolution. Studies like those reported here are key to developing an understanding of the impact of influenza C virus infection in humans, notably where comorbidities exist and severe respiratory disease can develop.

Correction: Breadth and function of antibody response to acute SARS-CoV-2 infection in humans.

[This corrects the article DOI: 10.1371/journal.ppat.1009352.].

Recruitment of dendritic cell progenitors to foci of influenza A virus infection sustains immunity.

Protection from infection with respiratory viruses such as influenza A virus (IAV) requires T cell­mediated immune responses initiated by conventional dendritic cells (cDCs) that reside in the respiratory tract. Here, we show that effective induction of T cell responses against IAV in mice requires reinforcement of the resident lung cDC network by cDC progenitors. We found that CCR2-binding chemokines produced during IAV infection recruit pre-cDCs from blood and direct them to foci of infection, increasing the number of progeny cDCs next to sites of viral replication. Ablation of CCR2 in the cDC lineage prevented this increase and resulted in a deficit in IAV-specific T cell responses and diminished resistance to reinfection. These data suggest that the homeostatic network of cDCs in tissues is insufficient for immunity and reveal a chemokine-driven mechanism of expansion of lung cDC numbers that amplifies T cell responses against respiratory viruses.

Reduced sialidase activity of influenza A(H3N2) neuraminidase associated with positively charged amino acid substitutions.

Neuraminidase (NA) inhibitors (NAI), oseltamivir and zanamivir, are the main antiviral medications for influenza and monitoring of susceptibility to these antivirals is routinely done by determining 50 % inhibitory concentrations (IC50) with MUNANA substrate. During 2010-2019, levels of A(H3N2) viruses presenting reduced NAI inhibition (RI) were low (~0.75 %) but varied year-on-year. The highest proportions of viruses showing RI were observed during the 2013-2014, 2016-2017 and 2017-2018 Northern Hemisphere seasons. The majority of RI viruses were found to contain positively charged NA amino acid substitutions of N329K, K/S329R, S331R or S334R, being notably higher during the 2016-2017 season. Sialidase activity kinetics were determined for viruses of RI phenotype and contemporary wild-type (WT) viruses showing close genetic relatedness and displaying normal inhibition (NI). RI phenotypes resulted from reduced sialidase activity compared to relevant WT viruses. Those containing S329R or N329K or S331R showed markedly higher Km for the substrate and Ki values for NAIs, while those with S334R showed smaller effects. Substitutions at N329 and S331 disrupt a glycosylation sequon (NDS), confirmed to be utilised by mass spectrometry. However, gain of positive charge at all three positions was the major factor influencing the kinetic effects, not loss of glycosylation. Because of the altered enzyme characteristics NAs carrying these substitutions cannot be assessed reliably for susceptibility to NAIs using standard MUNANA-based assays due to reductions in the affinity of the enzyme for its substrate and the concentration of the substrate usually used.

Correction: Protective porcine influenza virus-specific monoclonal antibodies recognize similar haemagglutinin epitopes as humans.

[This corrects the article DOI: 10.1371/journal.ppat.1009330.].

A Sanger sequencing protocol for SARS-CoV-2 S-gene.

We describe a Sanger sequencing protocol for SARS-CoV-2 S-gene the Spike (S)-glycoprotein product of which, composed of receptor-binding (S1) and membrane fusion (S2) segments, is the target of vaccines used to combat COVID-19. The protocol can be used in laboratories with basic Sanger sequencing capabilities and allows rapid "at source" screening for SARS-CoV-2 variants, notably those of concern. The protocol has been applied for surveillance, with clinical specimens collected in either nucleic acid preservation lysis-mix or virus transport medium, and research involving cultured viruses, and can yield data of public health importance in a timely manner.

Reduced antibody cross-reactivity following infection with B.1.1.7 than with parental SARS-CoV-2 strains.

Background: The degree of heterotypic immunity induced by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) strains is a major determinant of the spread of emerging variants and the success of vaccination campaigns, but remains incompletely understood. Methods: We examined the immunogenicity of SARS-CoV-2 variant B.1.1.7 (Alpha) that arose in the United Kingdom and spread globally. We determined titres of spike glycoprotein-binding antibodies and authentic virus neutralising antibodies induced by B.1.1.7 infection to infer homotypic and heterotypic immunity. Results: Antibodies elicited by B.1.1.7 infection exhibited significantly reduced recognition and neutralisation of parental strains or of the South Africa variant B.1.351 (Beta) than of the infecting variant. The drop in cross-reactivity was significantly more pronounced following B.1.1.7 than parental strain infection. Conclusions: The results indicate that heterotypic immunity induced by SARS-CoV-2 variants is asymmetric. Funding: This work was supported by the Francis Crick Institute and the Max Planck Institute for Dynamics of Complex Technical Systems, Magdeburg.

Diversity in the Circulation of Influenza A(H3N2) Viruses in the Northern Hemisphere in the 2018-19 Season.

While vaccination is considered the most effective means to prevent influenza infection, its seasonal effectiveness varies, depending on the circulating influenza strains. Here, we characterized the circulation of influenza strains in October-2018 and March-2019 around the world. For this, we used nasopharyngeal samples collected from outpatient and hospitalized patients in Israel and data reported in ECDC, CDC, and WHO databases. Influenza A(H3N2) was dominant in Israel, while in Europe, Asia, and USA, A(H1N1)pdm09 virus circulated first, and then the A(H3N2) virus also appeared. Phylogenetic analysis indicated that A(H3N2) viruses circulating in Israel belonged to clade-3C.3a, while in Europe, Asia, and USA, A(H3N2) viruses belonged to subclade-3C.2a1, but were later replaced by clade-3C.3a viruses in USA. The vaccine A(H3N2) components of that year, A/Singapore/INFIMH-16-0019/2016-(H3N2)-like-viruses, belonged to clade-3C.2a1. The circulation of different influenza subtypes and clades of A(H3N2) viruses in a single season highlights the need for universal influenza vaccines.

Protective porcine influenza virus-specific monoclonal antibodies recognize similar haemagglutinin epitopes as humans.

Pigs are natural hosts for the same subtypes of influenza A viruses as humans and integrally involved in virus evolution with frequent interspecies transmissions in both directions. The emergence of the 2009 pandemic H1N1 virus illustrates the importance of pigs in evolution of zoonotic strains. Here we generated pig influenza-specific monoclonal antibodies (mAbs) from H1N1pdm09 infected pigs. The mAbs recognized the same two major immunodominant haemagglutinin (HA) epitopes targeted by humans, one of which is not recognized by post-infection ferret antisera that are commonly used to monitor virus evolution. Neutralizing activity of the pig mAbs was comparable to that of potent human anti-HA mAbs. Further, prophylactic administration of a selected porcine mAb to pigs abolished lung viral load and greatly reduced lung pathology but did not eliminate nasal shedding of virus after H1N1pdm09 challenge. Hence mAbs from pigs, which target HA can significantly reduce disease severity. These results, together with the comparable sizes of pigs and humans, indicate that the pig is a valuable model for understanding how best to apply mAbs as therapy in humans and for monitoring antigenic drift of influenza viruses in humans, thereby providing information highly relevant to making influenza vaccine recommendations.

Breadth and function of antibody response to acute SARS-CoV-2 infection in humans.

Serological and plasmablast responses and plasmablast-derived IgG monoclonal antibodies (MAbs) have been analysed in three COVID-19 patients with different clinical severities. Potent humoral responses were detected within 3 weeks of onset of illness in all patients and the serological titre was elicited soon after or concomitantly with peripheral plasmablast response. An average of 13.7% and 3.5% of plasmablast-derived MAbs were reactive with virus spike glycoprotein or nucleocapsid, respectively. A subset of anti-spike (10 of 32) antibodies cross-reacted with other betacoronaviruses tested and harboured extensive somatic mutations, indicative of an expansion of memory B cells upon SARS-CoV-2 infection. Fourteen of 32 anti-spike MAbs, including five anti-receptor-binding domain (RBD), three anti-non-RBD S1 and six anti-S2, neutralised wild-type SARS-CoV-2 in independent assays. Anti-RBD MAbs were further grouped into four cross-inhibiting clusters, of which six antibodies from three separate clusters blocked the binding of RBD to ACE2 and five were neutralising. All ACE2-blocking anti-RBD antibodies were isolated from two recovered patients with prolonged fever, which is compatible with substantial ACE2-blocking response in their sera. Finally, the identification of non-competing pairs of neutralising antibodies would offer potential templates for the development of prophylactic and therapeutic agents against SARS-CoV-2.

A COVID-19 vaccine candidate using SpyCatcher multimerization of the SARS-CoV-2 spike protein receptor-binding domain induces potent neutralising antibody responses.

There is need for effective and affordable vaccines against SARS-CoV-2 to tackle the ongoing pandemic. In this study, we describe a protein nanoparticle vaccine against SARS-CoV-2. The vaccine is based on the display of coronavirus spike glycoprotein receptor-binding domain (RBD) on a synthetic virus-like particle (VLP) platform, SpyCatcher003-mi3, using SpyTag/SpyCatcher technology. Low doses of RBD-SpyVLP in a prime-boost regimen induce a strong neutralising antibody response in mice and pigs that is superior to convalescent human sera. We evaluate antibody quality using ACE2 blocking and neutralisation of cell infection by pseudovirus or wild-type SARS-CoV-2. Using competition assays with a monoclonal antibody panel, we show that RBD-SpyVLP induces a polyclonal antibody response that recognises key epitopes on the RBD, reducing the likelihood of selecting neutralisation-escape mutants. Moreover, RBD-SpyVLP is thermostable and can be lyophilised without losing immunogenicity, to facilitate global distribution and reduce cold-chain dependence. The data suggests that RBD-SpyVLP provides strong potential to address clinical and logistic challenges of the COVID-19 pandemic.

Integrating genotypes and phenotypes improves long-term forecasts of seasonal influenza A/H3N2 evolution.

Seasonal influenza virus A/H3N2 is a major cause of death globally. Vaccination remains the most effective preventative. Rapid mutation of hemagglutinin allows viruses to escape adaptive immunity. This antigenic drift necessitates regular vaccine updates. Effective vaccine strains need to represent H3N2 populations circulating one year after strain selection. Experts select strains based on experimental measurements of antigenic drift and predictions made by models from hemagglutinin sequences. We developed a novel influenza forecasting framework that integrates phenotypic measures of antigenic drift and functional constraint with previously published sequence-only fitness estimates. Forecasts informed by phenotypic measures of antigenic drift consistently outperformed previous sequence-only estimates, while sequence-only estimates of functional constraint surpassed more comprehensive experimentally-informed estimates. Importantly, the best models integrated estimates of both functional constraint and either antigenic drift phenotypes or recent population growth.

Vaccination is the best protection against seasonal flu. It teaches the immune system what the flu virus looks like, preparing it to fight off an infection. But the flu virus changes its molecular appearance every year, escaping the immune defences learnt the year before. So, every year, the vaccine needs updating. Since it takes almost a year to design and make a new flu vaccine, researchers need to be able to predict what flu viruses will look like in the future. Currently, this prediction relies on experiments that assess the molecular appearance of flu viruses, a complex and slow approach. One alternative is to examine the virus's genetic code. Mathematical models try to predict which genetic changes might alter the appearance of a flu virus, saving the cost of performing specialised experiments. Recent research has shown that these models can make good predictions, but including experimental measures of the virus' appearance could improve them even further. This could help the model to work out which genetic changes are likely to be beneficial to the virus, and which are not. To find out whether experimental data improves model predictions, Huddleston et al. designed a new forecasting tool which used 25 years of historical data from past flu seasons. Each forecast predicted what the virus population might look like the next year using the previous year's genetic code, experimental data, or both. Huddleston et al. then compared the predictions with the historical data to find the most useful data types. This showed that the best predictions combined changes from the virus's genetic code with experimental measures of its appearance. This new forecasting tool is open source, allowing teams across the world to start using it to improve their predictions straight away. Seasonal flu infects between 5 and 15% of the world's population every year, causing between quarter of a million and half a million deaths. Better predictions could lead to better flu vaccines and fewer illnesses and deaths.

Molecular Characterization of Influenza C Viruses from Outbreaks in Hong Kong SAR, China.

In 2014, the Centre for Health Protection in Hong Kong introduced screening for influenza C virus (ICV) as part of its routine surveillance for infectious agents in specimens collected from patients presenting with symptoms of respiratory viral infection, including influenza-like illness (ILI). A retrospective analysis of ICV detections up to week 26 of 2019 revealed persistent low-level circulation, with two outbreaks having occurred in the winters of 2015 to 2016 and 2017 to 2018. These outbreaks occurred at the same time as, and were dwarfed by, seasonal epidemics of influenza types A and B. Gene sequencing studies on stored ICV-positive clinical specimens from the two outbreaks have shown that the hemagglutinin-esterase (HE) genes of the viruses fall into two of the six recognized genetic lineages (represented by C/Kanagawa/1/76 and C/São Paulo/378/82), with there being significant genetic drift compared to earlier circulating viruses within both lineages. The location of a number of encoded amino acid substitutions in hemagglutinin-esterase fusion (HEF) glycoproteins suggests that antigenic drift may also have occurred. Observations of ICV outbreaks in other countries, with some of the infections being associated with severe disease, indicates that ICV infection has the potential to have significant clinical and health care impacts in humans.IMPORTANCE Influenza C virus infection of humans is common, and reinfection can occur throughout life. While symptoms are generally mild, severe disease cases have been reported, but knowledge of the virus is limited, as little systematic surveillance for influenza C virus is conducted and the virus cannot be studied by classical virologic methods because it cannot be readily isolated in laboratories. A combination of systematic surveillance in Hong Kong SAR, China, and new gene sequencing methods has been used in this study to assess influenza C virus evolution and provides evidence for a 2-year cycle of disease outbreaks. The results of studies like that reported here are key to developing an understanding of the impact of influenza C virus infection in humans and how virus evolution might be associated with epidemics.

Pandemic potential of highly pathogenic avian influenza clade 2.3.4.4 A(H5) viruses.

The panzootic caused by A/goose/Guangdong/1/96-lineage highly pathogenic avian influenza (HPAI) A(H5) viruses has occurred in multiple waves since 1996. From 2013 onwards, clade 2.3.4.4 viruses of subtypes A(H5N2), A(H5N6), and A(H5N8) emerged to cause panzootic waves of unprecedented magnitude among avian species accompanied by severe losses to the poultry industry around the world. Clade 2.3.4.4 A(H5) viruses have expanded in distinct geographical and evolutionary pathways likely via long distance migratory bird dispersal onto several continents and by poultry trade among neighboring countries. Coupled with regional circulation, the viruses have evolved further by reassorting with local viruses. As of February 2019, there have been 23 cases of humans infected with clade 2.3.4.4 H5N6 viruses, 16 (70%) of which had fatal outcomes. To date, no HPAI A(H5) virus has caused sustainable human-to-human transmission. However, due to the lack of population immunity in humans and ongoing evolution of the virus, there is a continuing risk that clade 2.3.4.4 A(H5) viruses could cause an influenza pandemic if the ability to transmit efficiently among humans was gained. Therefore, multisectoral collaborations among the animal, environmental, and public health sectors are essential to conduct risk assessments and develop countermeasures to prevent disease and to control spread. In this article, we describe an assessment of the likelihood of clade 2.3.4.4 A(H5) viruses gaining human-to-human transmissibility and impact on human health should such human-to-human transmission occur. This structured analysis assessed properties of the virus, attributes of the human population, and ecology and epidemiology of these viruses in animal hosts.

Broadly Inhibiting Antineuraminidase Monoclonal Antibodies Induced by Trivalent Influenza Vaccine and H7N9 Infection in Humans.

The majority of antibodies induced by influenza neuraminidase (NA), like those against hemagglutinin (HA), are relatively specific to viruses isolated within a limited time window, as seen in serological studies and the analysis of many murine monoclonal antibodies (MAbs). We report three broadly reactive human MAbs targeting N1 NA. Two were isolated from a young adult vaccinated with trivalent influenza vaccine (TIV), which inhibited N1 NA from viruses isolated from humans over a period of a hundred years. The third antibody, isolated from a child with acute mild H7N9 infection, inhibited both group 1 N1 and group 2 N9 NAs. In addition, the antibodies cross-inhibited the N1 NAs of highly pathogenic avian H5N1 influenza viruses. These antibodies are protective in prophylaxis against seasonal H1N1 viruses in mice. This study demonstrates that human antibodies to N1 NA with exceptional cross-reactivity can be recalled by vaccination and highlights the importance of standardizing the NA antigen in seasonal vaccines to offer optimal protection.IMPORTANCE Antibodies to the influenza virus NA can provide protection against influenza disease. Analysis of human antibodies to NA lags behind that of antibodies to HA. We show that human monoclonal antibodies against NA induced by vaccination and infection can be very broadly reactive, with the ability to inhibit a wide spectrum of N1 NAs on viruses isolated between 1918 and 2018. This suggests that antibodies to NA may be a useful therapy and that the efficacy of influenza vaccines could be enhanced by ensuring the appropriate content of NA antigen.

Segment 2 from influenza A(H1N1) 2009 pandemic viruses confers temperature-sensitive haemagglutinin yield on candidate vaccine virus growth in eggs that can be epistatically complemented by PB2 701D.

Candidate vaccine viruses (CVVs) for seasonal influenza A virus are made by reassortment of the antigenic virus with an egg-adapted strain, typically A/Puerto Rico/8/34 (PR8). Many 2009 A(H1N1) pandemic (pdm09) high-growth reassortants (HGRs) selected this way contain pdm09 segment 2 in addition to the antigenic genes. To investigate this, we made CVV mimics by reverse genetics (RG) that were either 6 : 2 or 5 : 3 reassortants between PR8 and two pdm09 strains, A/California/7/2009 (Cal7) and A/England/195/2009, differing in the source of segment 2. The 5 : 3 viruses replicated better in MDCK-SIAT1 cells than the 6 : 2 viruses, but the 6 : 2 CVVs gave higher haemagglutinin (HA) antigen yields from eggs. This unexpected phenomenon reflected temperature sensitivity conferred by pdm09 segment 2, as the egg HA yields of the 5 : 3 viruses improved substantially when viruses were grown at 35 °C compared with 37.5 °C, whereas the 6 : 2 virus yields did not. However, the authentic 5 : 3 pdm09 HGRs, X-179A and X-181, were not markedly temperature sensitive despite their PB1 sequences being identical to that of Cal7, suggesting compensatory mutations elsewhere in the genome. Sequence comparisons of the PR8-derived backbone genes identified polymorphisms in PB2, NP, NS1 and NS2. Of these, PB2 N701D affected the temperature dependence of viral transcription and, furthermore, improved and drastically reduced the temperature sensitivity of the HA yield from the 5 : 3 CVV mimic. We conclude that the HA yield of pdm09 CVVs can be affected by an epistatic interaction between PR8 PB2 and pdm09 PB1, but that this can be minimized by ensuring that the backbones used for vaccine manufacture in eggs contain PB2 701D.