Evolution of Human Respiratory Syncytial Virus (RSV) over Multiple Seasons in New South Wales, Australia

Abstract

There is an ongoing global pandemic of human respiratory syncytial virus (RSV) infection that results in substantial annual morbidity and mortality. In Australia, RSV is a major cause of acute lower respiratory tract infections (ALRI). Nevertheless, little is known about the extent and origins of the genetic diversity of RSV in Australia, nor the factors that shape this diversity. We have conducted a genome-scale analysis of RSV infections in New South Wales (NSW). RSV genomes were successfully sequenced for 144 specimens collected between 2010⁻2016. Of these, 64 belonged to the RSVA and 80 to the RSVB subtype. Phylogenetic analysis revealed a wide diversity of RSV lineages within NSW and that both subtypes evolved rapidly in a strongly clock-like manner, with mean rates of approximately 6⁻8 × 10^-4 nucleotide substitutions per site per year. There was only weak evidence for geographic clustering of sequences, indicative of fluid patterns of transmission within the infected population and no evidence of any clustering by patient age such that viruses in the same lineages circulate through the entire host population. Importantly, we show that both subtypes circulated concurrently in NSW with multiple introductions into the Australian population in each year and only limited evidence for multi-year persistence.

Keywords: evolution; multi-year persistence; phylogenetics; respiratory syncytial virus.

Figure 1

Incidence of respiratory syncytial virus (RSV) in New South Wales (NSW). (A) Number of RSV genome sequences per four-week period: RSVA = blue, RSVB = red. (B) Number of positive samples per 1000 specimens reported across NSW per four-week period: RSV = orange, Influenza = green, Other, including parainfluenza virus, adenovirus, and human metapneumovirus = grey.

Figure 2

Geographic distribution of respiratory syncytial virus in NSW. Number of sequences per postcode in NSW (left) and greater Westmead area (right). The location for Westmead Hospital is indicated.

Figure 3

Global phylogeny of respiratory syncytial virus subtype A (RSVA) and local clustering within NSW. The maximum likelihood tree shown was estimated using complete RSV genome sequences. Known genotypes and the sequence time range is indicated. The tree was rooted using an RSVB outgroup. Tree tips are colored according to the geographic region and sequences from this study are shown in white. Local clusters comprising NSW sequences are marked within the global tree and the four clusters with potential multi-season transmission events are colored grey and enlarged on the right side. Node supports are indicated and branch lengths are scaled according to the number of substitutions per site.

Figure 4

Global phylogeny of respiratory syncytial virus subtype B (RSVB) and local clustering within NSW. The maximum likelihood tree shown was estimated using complete RSV genome sequences. The tree was rooted using RSVA as an outgroup and the BA and BA10 genotypes are marked. Tip colors represent the sampling location and sequences from this study are shown in white. Local clusters comprising NSW sequences are marked within the global tree and the three clusters with potential multi-season transmission events are colored grey and enlarged on the right side. Node supports are indicated and branch lengths are scaled according to the number of substitutions per site.

Figure 5

Evolutionary rates in respiratory syncytial virus (RSV). (A) Linear regressions of root-to-tip genetic distances against sampling date based on maximum likelihood trees. The R² value for each regression is indicated and corresponding p values <0.001 are indicated with an asterisk. Sequences from NSW in this study are shown on the left and global sequences on the right. (B) Estimates of the nucleotide substitution rate per site per year are shown for sequences from NSW (BEAST and least-square dating algorithm (LSD) estimates) and globally (LSD estimates only). Rates are shown as mean values (circles) and the 95% highest posterior density (HPD) and confidence intervals for BEAST and LSD, respectively (error bar). (RSVA: blue; RSVB: red).