An Evolutionary Indicator for Monitoring Morbidity of Avian Influenza A Virus Subtypes

Background: Avian influenza A has been a major threat to worldwide public health. The evolutionary patterns of the pathogenic viruses between humans and birds are essential to effective monitoring of disease propagation. In this study, we focused on finding an appropriate evolutionary measure for avian influenza A to reflect its adaptation to human hosts and outbreaks in human population. We estimated the synonymous and nonsynonymous substitution rates (Ks and Ka) for two hemagglutinin domains (HA1 and HA2) and neuraminidase protein in five virus subtypes (H1N1, H2N2, H3N2, H5N1, and H9N2) from human and avian hosts. We analyzed the relation between Ka and Ks and detected meaningful evolutionary patterns. Results: Our results reveal that the slopes of Ka-Ks linear regression lines in HA1 are strongly indicative of virus subtypes. We also show that dramatically increased Ka-Ks regression slopes are associated with outbreaks of H5N1(year 1997 and 2003) related human infections, suggesting that the slope might be a good indicator for monitoring influenza propagation. Analysis of Ka/Ks ratios further suggests that HA1 in H5N1 is undergoing accelerated evolution. Furthermore, the distribution pattern of Ka vs. Ks in HA1 of H1N1 viruses indicates that the evolution of avian influenza virus antigenicity between humans and birds may have a “ceiling” of Ka = ~ 0.2 (Ka = ~ 0.1 in HA). In comparison, the majority of Ka values of HA1 in H5N1 and H9N2 are much smaller than 0.1, implying large room for changes at the amino acid level. Conclusions: In this study, we proposed a straightforward indicator, the Ka-Ks regression slopes between human and avian hosts, for monitoring influenza propagation to predict outbreaks. The indicator was highly correlated to avian influenza transmission and might provide a different direction for studying.