Molecular characterization of the HA gene of influenza type B viruses

Nucleotide sequences of the HA1 subunit of influenza B viruses isolated in Portugal between 1994 and 2003 influenza winter seasons were analyzed by the Neighbor-Joining algorithm and rates of HA1 evolution estimated by linear regression. From 1994 to 2002, all influenza B viruses studied were of the Yamagata lineage. Strains isolated from 1994 to 1996, 1996 to 1999, and 1999 to 2002 revealed a high similarity with B/Beijing/184/93, B/Yamanashi/166/98, and B/Sichuan/379/99, respectively, and strains isolated during 1994-1995, 1996-1997, and 1998-1999 clustered in more than one branch of the phylogenetic tree. Victoria-related strains reappeared during 2002/2003 and formed only one branch in the phylogenetic tree revealing a closer relationship to B/Shandong/7/97. Evolutionary rates for strains from the Yamagata lineage were estimated as 3.82x10(-3) nucleotides/site/year and 2.62x10(-3) nucleotides/site/year for Victoria-related strains. In order to identify putative influenza B HA1 codons under selective pressure, a codon-substitution model for heterogeneous selective pressure at amino acid sites was used. A percentage of 97.3% of codons under negative selective pressure and 2.7% of codons under positive selective pressure (omega=dN/dS=2.65) were estimated, with posterior probability higher than 0.90. Amino acid sites 75, 197, and 199 were found more likely to be under positive selective pressure.