Molecular genetic characteristics of influenza A virus clinically isolated during 2011‐2016 influenza seasons in Korea

Background The influenza virus is reportedly associated with 3‐5 million cases of severe illness and 250 000‐500 000 deaths annually worldwide. Objectives We investigated the variation of influenza A virus in Korea and examined the association with death. Methods A total of 13 620 cases were enrolled in the Hospital‐based Influenza Morbidity & Mortality surveillance system in Korea during 2011‐2016. Among these cases, a total of 4725 were diagnosed with influenza using RT‐PCR (influenza A; n = 3696, influenza B; n = 928, co‐infection; n = 101). We used 254 viral sequences from the 3696 influenza A cases for phylogenetic analysis using the BioEdit and MEGA 6.06 programs. Results We found that the sequences of A/H3N2 in the 2011‐2012 season belong to subgroup 3C.1, whereas the sequences in the 2012‐2013 season pertain to subgroup 3C.2. The sequences in the 2013‐2014 and 2014‐2015 seasons involve subgroups 3C.3a and 3C.2a. The A/H1N1pdm09 subtype belongs to subgroup 6 and contains two clusters. In addition, sequence analysis confirmed the several substitutions of internal genes and gene substitutions associated with drug resistance (I222V in NA and S31N in M2) in the fatal cases. While statistical analysis found no significant associations between genetic differences in the viruses and mortality, mortality was associated with certain host factors, such as chronic lung disease. Conclusions In conclusion, influenza A virus clade changes occurred in Korea during the 2011‐2016 seasons. These data, along with antigenic analysis, can aid in selecting effective vaccine strains. We confirmed that fatality in influenza A cases was related to underlying patient diseases, such as chronic lung disease, and further studies are needed to confirm associations between mortality and viral genetic substitutions.

A viruses are enhanced by the absence of RNA proofreading enzymes; in particular, substitutions in the HA protein alter the viral antigenic epitopes sufficiently to avoid the host immune response (antigenic drift). These antigenic changes can trigger the generation of new strains and subgroups, leading to pandemics or new epidemics worldwide. [1][2][3] Influenza A virus subtypes H3N2 (A/H3N2) and H1N1 (A/ H1N1), and influenza B virus have been circulating in the human population every winter, and account for 3 to 5 million cases of severe illness and 250 000 to 500 000 deaths annually, mostly caused by secondary bacterial pneumonia in young children and the elderly. 4 have provided comprehensive understanding of evolutionary models as well as epidemiologic insight based on analysis of antigenic determinants, drug resistance, and a variety of sequence-based bioinformatics methods. 2,8,9 Here, we analyzed the HA and NA genes of influenza A viruses prevalent in Korea during the 2011-2016 seasons. These viruses were identified through the Hospital-based Influenza Morbidity & Mortality (HIMM) surveillance system. 10 In addition, we investigated the effect of mutations in the 8 segmented genes of influenza A viruses on risk of mortality for infected patients.

| Sequence analysis
For the phylogenetic study, the sequences were compared with the NCBI-registered full-length nucleotide sequences of influenza A/

| Hemagglutination inhibition (HAI) assay
HAI titers were determined using standard procedures. In brief, antisera were pre-treated overnight with receptor-destroying enzyme

| Neuraminidase inhibitor (NAI) assay
The NA-Fluor TM assay kit (Thermo Fisher Scientific, Waltham, MA, USA) was used according to the manufacturer′s protocol.
Neuraminidase inhibitor was prepared in 10-

| Genetic analysis of 8 segmented genes in the fatal influenza A virus isolates
Viral virulence can be increased by mutating non-structural proteins (Table S1). Therefore, internal gene sequences from influenza A viruses from fatal cases were used in the genetic analysis. Genetic  (Table 3).
In the A/H3N2 analysis, we confirmed variations for the cluster differences (3C.2a and 3C.3a) in the HA and NA genes. In addition, the NA sequence of the 2011-2012 season (n = 2) carried an I222V substitution, an NA inhibitor resistance mutation. [13][14][15][16] The NA se-  were consistent with clade 3C.2a in the HA phylogenetic tree. As shown in Table 4, we confirmed that the substitutions of NA, NS1, and PA coincided with HA-based genetic clade. In addition, the PB1 and PB2 genes revealed multiple non-synonymous substitutions (Table S2).
Statistical analysis was performed to study the association between amino acid substitution and death in fatal cases (Table 6).
Although some substitutions (I39M of M2 and M481I of NP) showed a P value of .054, they were not significant in the correlation analysis of genetic substitution and mortality. As shown in Table 7, our correlation analysis revealed that chronic lung disease was more frequently associated with fatal than non-fatal cases.
In conclusion, statistical analysis showed no significant association between the viral genetic differences and mortality; however, the mortality was increased by host factors such as chronic lung disease.  and H274Y substitutions. 15,16,19,[22][23][24][25][26] The S31N substitution in the M2 protein was frequently detected in the more recent viral sequences and reference sequences. 27