The viruses which cause influenza are highly similar in structure but diverse in host range, virulence, and antigenic properties. Understanding the strategies by which these viruses continue to produce pandemic disease is largely dependent on our ability to determine how this functional diversity contributes to the ability of the virus to survive in the human population as well as in other mammalian and avian hosts. Two aspects of virus structure are involved. The virus has a segmented RNA genome which is synthesized by a virus-encoded polymerase that lacks the proof-reading capabilities needed for high fidelity genome replication. Uncloned populations of this virus therefore consist of particles with sequence diversity at both the nucleotide and protein level. Secondly, the virus contains surface glycoproteins whose structures are determined by the amino acid sequence encoded in the viral genome and by the oligosaccharide processing pathway of the host cell in which the virus is grown. Cell-specific glycosylation of the viral hemagglutinin (HA) can, therefore, enhance the biological diversity already present as a consequence of low fidelity genome replication. We are interested in determining the role of HA sequence diversity in determining the ability of these viruses to grow in cells of avian and mammalian origin. Our approaches include (1) determining the structural characteristics of Has on virions recovered directly from human and avian sources, and (2) characterizing a collection of HA variants each of which differs from the others by a single amino acid. In the latter approach we use both naturally occurring mutants and mutants made-to-order by site specific mutagenesis. The receptor binding properties of these HA variants, determined using well characterized neoglycolipids and/or neoglycoproteins, are then related to their ability to grow in cultured cells from different species. Using these approaches we can identify the structural properties of the HA which determine its conformational stability, its ability to interact with specific sialylated glycans, and its ability to infect cells from different animal species. Since the ability of antibodies to neutralize this virus is determined in part by the affinity of the HA for host cell receptors,we expect this information to help us design better diagnostic tests and vaccines. We also expect it to provide clues forpreventing the transfer of influenza viruses to and from natural reservoirs inlower animals.