The 20th century witnessed three influenza pandemics: the Spanish flu of 1918 (H1N1), the Asian flu of 1957-58 (H2N2) and the Hong Kong flu of 1967-68 (H3N2). Among these subtypes the H1N1 and H3N2 continue to circulate in the human population leading to epidemic outbreaks annually and the H1N1 subtype was responsible for the 2009 ‘swine flu’ pandemic (2009 H1N1). The H2N2 subtype had stopped circulating in humans by 1968, however H2 subtype viruses are occasionally isolated from swine and avian species 
. The circulation of avian H2 strains in domestic birds and pigs increase the risk of human exposure to these viruses and reintroduction of the viruses to the human population. Such a reintroduction will pose a significant global health threat given the lack of pre-existing immunity in a huge subset of the human population born after 1968.
One of the main steps in the evolution of a pandemic influenza virus is the acquisition of genetic changes that enable it to adapt to the human host in order to replicate efficiently and transmit rapidly resulting in widespread and sustained disease in humans 
. The critical first step in the host infection by the virus is the binding of the viral surface glycoprotein hemagglutinin (HA) to sialylated glycan receptors, complex glycans terminated by N
-acetylneuraminic acid (Neu5Ac) expressed on the host cell surface 
. Glycans terminating in Neu5Ac that is α2→6-linked to the penultimate sugar are predominantly expressed in human upper respiratory epithelia 
and serve as receptors for human-adapted influenza A viruses (henceforth referred to as human receptors
). On the other hand, glycans terminating in Neu5Ac that is α2→3 -linked to the penultimate sugar residue, serve as receptors for the avian-adapted influenza viruses (henceforth referred to as avian receptors
The molecular interactions of HA with avian and human receptors have been captured using a topology-based definition of glycan receptors 
. Glycan array platforms comprised of representative avian and human receptors have been widely employed to study the glycan receptor binding of HAs and whole viruses 
. The relative binding affinities of recombinantly expressed HAs from avian- (such as H1N1 and H5N1) and human-adapted (such as H1N1 and H3N2) viruses to avian and human receptors have been quantified by analyzing these HAs (or whole viruses) in a dose-dependent manner on glycan array platforms 
. Furthermore, the glycan array binding properties of the HAs have been shown to correlate with their binding to physiological glycan-receptors in human respiratory tissues 
. Importantly, it has been shown that the human receptor-binding affinity of H1N1 HAs correlated with the efficiency of airborne viral transmission in the ferret animal model 
, which is an established model to evaluate viral transmissibility in humans 
. Such a relationship has yet to be shown for the H2N2 subtype.
Previous structural and biochemical studies have provided insights into interactions of the receptor binding site (RBS) of HA with avian and human receptors for both wild type (WT
) and mutant forms of HA derived from the 1957-58 H2N2 pandemic strains 
. However, it has been recently demonstrated that changes in the interactions between amino acids within and proximal to the RBS, arising from substitutions due to antigenic drift or reassortment, have profound effects on HA-glycan interactions which in turn influences the glycan binding affinity of HA 
. This observation is particularly relevant to HA from recent avian-H2 strains that have diverged considerably in sequence compared to the HA sequence of the pandemic H2N2 strains 
. Therefore in order to monitor changes in the recent avian H2-subtype viruses that would possibly lead to their human-adaptation, it is important to understand the mutations in their HA that would confer human receptor-binding affinity that is quantitatively in the same range as that of HA from the 1957-58 human-adapted H2N2 pandemic viruses.
In this study, we have systematically analyzed the effects of mutations in the glycan RBS of pandemic and recent avian H2N2 HAs on their respective glycan-binding specificities. The HA from a representative 1957-58 pandemic H2N2 strain, A/Albany/6/58 (Alb58
), was chosen as a reference human-adapted HA. The HA from a representative avian H2N2 virus, A/Chicken/Pennsylvania/2004 (CkPA04)
, which is among the most recent strains isolated from birds was also evaluated in this study 
. We first characterized the glycan receptor-binding affinity and human respiratory tissue binding properties of these avian- and human-adapted H2N2 HAs. The glycan receptor-binding affinity of HA is quantitatively defined using an apparent binding constant Kd'
that takes into account the cooperativity and avidity in the multivalent HA-glycan interactions as described previously 
. Next, using homology-based structural models of Alb58
HA-human receptor and CkPA04
HA-avian receptor complexes we analyzed the RBS of these HAs and designed and evaluated mutations in CkPA04
HA that would make its human receptor binding affinity in the same range as that of Alb58