Deep Sequencing of H7N9 Influenza A Viruses from 16 Infected Patients from 2013 to 2015 in Shanghai Reveals Genetic Diversity and Antigenic Drift

Abstract

Influenza A virus (IAV) infections are a major public health concern, including annual epidemics, epizootic outbreaks, and pandemics. A significant IAV epizootic outbreak was the H7N9 avian influenza A outbreak in China, which was first detected in 2013 and which has spread over 5 waves from 2013 to 2017, causing human infections in many different Chinese provinces. Here, RNA from primary clinical throat swab samples from 20 H7N9-infected local patients with different clinical outcomes, who were admitted and treated at one hospital in Shanghai, China, from April 2013 to April 2015, was analyzed. Whole-transcriptome amplification, with positive enrichment of IAV RNA, was performed, all 20 samples were subjected to deep sequencing, and data from 16 samples were analyzed in detail. Many single-nucleotide polymorphisms, including ones not previously reported, and many nonsynonymous changes that could affect hemagglutinin head and stalk antibody binding epitopes were observed. Minor populations representing viral quasispecies, including nonsynonymous hemagglutinin changes shared by antigenically variant H7N9 clades identified in the most recent wave of H7N9 infections in 2016 to 2017, were also identified.IMPORTANCE H7N9 subtype avian influenza viruses caused infections in over 1,400 humans from 2013 to 2017 and resulted in almost 600 deaths. It is important to understand how avian influenza viruses infect and cause disease in humans and to assess their potential for efficient person-to-person transmission. In this study, we used deep sequencing of primary clinical material to assess the evolution and potential for human adaptation of H7N9 influenza viruses.

Keywords: DNA sequencing; H7N9; avian viruses.

FIG 1

Mapped reads and called SNP numbers. (A) Numbers of called SNPs for all samples. (B) Correlation of the number of mapped reads and the number of called SNPs. (C) Correlation of the average coverage and the number of called SNPs.

FIG 2

Phylogenetic trees of unique HA segments of human H7N9 in China from 2013 to 2017 and consensus HA sequences (SNP frequency, >50%) of our samples. Approximately maximum-likelihood midpoint-rooted phylogenetic trees with Shimodaira-Hasegawa (SH) values are presented. Purple, reference genome; green, our samples in “sample number_year_outcome” format; blue, viruses in the Pearl River Delta HA lineage; red, viruses in the Yangtze River Delta HA lineage. See Fig. S1 for an expanded form of this tree, in which individual viral taxa are labeled by strain name.

FIG 3

The structure of HA with mutations observed on head regions of sample 17 shown on an H7 HA monomer (A/Anhui/1/2013; see Materials and Methods).

FIG 4

The structure of the HA A396E mutation at the HA stalk related to stalk neutralizing antibody CT149. (A) HA396E. (B) HA396A shown on an H7 HA monomer (A/Anhui/1/2013; see Materials and Methods).

FIG 5

Venn diagram of protein sequences identified in three H7N9 waves between 2013 to 2015 compared to the avian consensus protein sequence. Set analysis was performed on translated nonsynonymous SNPs identified in waves 1, 2, and 3. Novel mutations identified in this study are shown in green; SNPs previously identified in <10% or ≥10% of sequences in the Influenza Research Database are indicated by black text or blue text, respectively. Changes from the avian to the human consensus sequences are shown in red. Percentages of protein sequence changes were rounded to nearest whole-number value. SNPs identified in only a single patient are marked with an asterisk (*). The Venn diagram was created using Venny 2.1 (http://bioinfogp.cnb.csic.es/tools/venny/).