Search

Background: The global surveillance of human influenza viruses has resulted in the generation of a uniquely extensive collection of geographically and temporally comprehensive virus sequence data, which has provided an opportunity to explore the drivers behind the global spread of influenza viruses. However, due to the insufficient spatiotemporally representative virus sequence data from tropical and sub-tropical African countries, especially from sub-Saharan Africa, relatively little is known about the possible role these regions play in the global spread of influenza viruses. Using influenza A(H1N1)pdm09 virus and A(H3N2) virus sequence data, this study aimed to understand how seasonal influenza viruses are introduced and spread across geographically defined regions, whether local, national, continental or global, and their patterns of persistence across these regions. Methods: A laboratory method for whole-genome sequencing (WGS) of influenza A(H1N1)pdm09 and A(H3N2) viruses on Illumina next-generation sequencing (NGS) platform was established at Kilifi, coastal Kenya. This was then used to sequence samples collected between 2009 and 2018 from geographically defined regions: local community in Kilifi (n=66); countrywide in Kenya (n=383); and across Africa from 5 countries (n=100). The arising genomes were analyzed using phylogenetic and phylogeographical methods to investigate the patterns of spread, persistence and fade-out of seasonal influenza viruses at a local community in Kilifi, countrywide in Kenya, and across the African continent. Additionally, a global contemporaneous sequence dataset was analyzed in conjunction with the WGS data from this study in a Bayesian framework for inference of the situation of sub-Saharan Africa in the global network of spread of influenza viruses. Results: A total of 549 new influenza type A virus (IAV) whole genomes were generated during this study; 414 A(H1N1)pdm09 virus and 135 A(H3N2) virus genomes. Phylogeographical analyses revealed that local seasonal community epidemics of IAV were initiated by multiple independent introductions into the community, with each introduction commonly spreading to multiple locations within a relatively short period of time. Countrywide, in Kenya, circulation of IAV was predominantly characterized by virus migration from multiple locations to multiple destinations within the country and between locations in proximity; persistence of IAV countrywide might therefore be modulated by frequent virus introductions from outside the country and virus spread between locations in proximity. Continentwide, strains of IAV from Africa fell into strongly supported multinational lineages that suggested possible intra-continental spread of influenza viruses within Africa, which exhibited a significant northern to southern hemisphere migration. Globally, significant migration pathways from multiple geographical regions to multiple geographical destinations that also includes Africa were observed, which suggests that the seeding of epidemics of influenza viruses globally is driven by different geographical regions that also includes Africa. However, East or Southeast (E-SE) Asia acted as the major source of spread of influenza viruses globally, which is consistent with findings from other studies on the global circulation of influenza viruses. A greater global migration was observed for A(H3N2) virus compared to A(H1N1)pdm09 virus, consistent with greater global migration of A(H3N2) virus compared to (H1N1)pdm09 virus. Conclusions: The global migration dynamics of seasonal influenza viruses are well understood, and several models have been proposed to describe these patterns. However, analysis of virus sequence data from understudied regions, as exemplified in this study, suggests that these migration patterns are far more complex than those proposed by current models alone. For example, the findings from this study support the notion that influenza viruses persist as temporally structured migrating metapopulations in which new virus strains can emerge in any geographical region, including in Africa, with the location of the source population changing regularly. The epidemics across geographically defined regions (local community, countrywide, continentwide, and globally) are also interconnected at various scales of observation to different extents. Therefore, a more complete understanding of the global migration dynamics of influenza viruses requires deeper and wider sampling of viruses from understudied tropical and sub-tropical regions, notably, Africa, South and Central Asia, and South America. Understanding the circulation patterns of influenza viruses across geographically defined regions, together with their origins and patterns of persistence, is useful in selecting the most effective vaccine strains for the circulating seasonal influenza viruses. The rapid and widespread global mixing of viruses from all northern and southern hemisphere countries including in countries in Africa, Asia, Europe, North America, South America, and Oceania as reported in this study emphasize that global vaccine recommendations need well distributed, widespread global influenza A(H1N1)pdm09 virus and A(H3N2) virus sampling from as many localities as possible.