OBJECTIVETo establish a rapid, specific and effective technique for identifying subtyping A(1), A(3) and B of influenza virus isolates and clinical specimens as well as to analyze the sequences of nucleotides and deduced amino acids of HA1 regions from isolates of influenza virus A(3) isolated from 1996 to 2002.

METHODSSix inner and outer sets of oligonucleotide primers were designed to detect, type and subtype human influenza A and B. The first two corresponding sets differentiate type A and B of matrix (M) gene while, the second two corresponding sets identify the H(1) and H(3) subtypes of type A virus HA gene. To type and subtype influenza viruses in clinical isolates, a mixture of inner primer sets specific for H(1), H(3) and B were used in a single PCR reaction tube. To detect influenza viruses in clinical specimens, a mixture of the outer primer sets were used in a single primary PCR tube, and the inner ones in a single second PCR reaction tube. Amplified products were visualized in 1.2% agrose gel containing ethidium bromide. HA1 regions of hemagglutinin of 5 field strains (H3N2) isolated from 1996 to 2002 in Beijing were amplified by RT-PCR and sequenced directly.

RESULTSThere was 100% correlation between multiplex RT-PCR and culture to type and subtype influenza viruses from clinical isolates. For typing and subtyping, 76.9%, 57.1% and 86.5% were positive for A(1), A(3) and B by multiplex nested-PCR compared within virus isolation on culture, respectively. The sequence data of HA1 of A(3) strains showed that there was a high homology of nucleotide and amino acid, and the closer the date of isolating was, the higher homology showed.

CONCLUSIONSMultiplex RT-PCR and nested-PCR for influenza viruses could provide a useful alternative to existing methods of influenza detected and identified from clinical isolate and specimens. There were certain, continuous mutations and increasing glycosylated sites which might cause the antigen drift in the A(3) strains during 1996-2002 in Beijing area.