Identification and characterization of novel lineage 1 Powassan virus strains in New York State

Abstract

Powassan virus (POWV, family Flaviviridae) is a reemerging tick-borne virus endemic in North America and Russia. In 1997, a POWV-like agent was isolated from Ixodes scapularis in New England and determined to be genetically distinct from the original POWV isolate. This revealed the existence of two lineages: lineage 1, prototype Powassan virus (POWV-1) and lineage 2, deer tick virus (DTV). POWV-1 is thought to be primarily maintained in a cycle between I. cookei and woodchucks and I. marxi and squirrels, while DTV is primarily maintained in a cycle between I. scapularis and small mammal hosts. Recent tick, mammalian, and human isolates from New York State (NYS) have been identified as DTV, but for the first time in 45 years, we detected four POWV-1 isolates, including the first reported isolation of POWV-1 from I. scapularis. We aimed to investigate genotypic and phenotypic characteristics of recent NYS isolates through sequence analysis and evaluation of replication kinetics in vitro and in vivo. Our sequencing revealed genetic divergence between NYS POWV-1 isolates, with two distinct foci. We found that POWV-1 isolates displayed variable replication kinetics in nymphal ticks but not in cell culture. POWV-1 isolated from I. scapularis displayed increased fitness in experimentally infected I. scapularis as compared to historic and recent POWV-1 isolates from I. cookei. These data suggest the emergence of divergent POWV-1 strains in alternate tick hosts and maintenance of genetically and phenotypically discrete POWV-1 foci.

Keywords: Ixodes scapularis; New York State; Powassan virus; deer tick virus; lineage 1.

Figure 1.

Powassan virus (POWV-1) and deer tick virus (DTV) surveillance across New York State (NYS) between 2007-2019. All counties are sampled yearly for questing adult Ixodes scapularis and tested for POWV except for 4 counties outside of New York City. Counties shaded in gray represent those in which at least 1 tick has been positive for POWV and those in white represent no viral detection. Counties with stars indicate regions which recent NYS POWV-1 isolates were collected. Stars do not represent location within the county of the collection site and were placed concentrically.

Figure 2.

Maximum likelihood phylogeny of Powassan virus (POWV-1) and deer tick virus (DTV) based on full genome nucleotide sequences with tickborne encephalitis virus as an outgroup. Bootstrap values are displayed at each node (range: 38.9-100). Recent New York State (NYS) isolates (arrows); POWV-1 46-001, POWV-1 63-002, and POWV-1 19065–059 cluster with historic POWV-1 isolates.

Figure 3.

Growth kinetics of Powassan virus (POWV-1) and deer tick virus (DTV) in baby hamster kidney (BHK-21) cells. Replication in BHK-21 is similar among all isolates. Data points represent mean +/- SEM (n = 3 per strain). Two-way ANOVA and Tukey’s multiple comparisons test (p = 0.8261).

Figure 4.

Overall infection rates of New York State (NYS) Powassan virus (POWV-1) and deer tick virus (DTV) isolates in Ixodes scapularis nymphs following immersion. Black shading represents the percent of infected ticks with POWV-1 or DTV and gray shading represent the percent of uninfected ticks. Infected ticks across all timepoints (7-,14-,21-, and 28-days post infection) were combined to calculate overall percentage of POWV-infected ticks (n = 80/strain for POWV-1 64–7062 and POWV-1 46-001, n = 90/strain for POWV-1 63-002, POWV-1 19065-059, and DTV 18071-054). POWV-1 64-7062, POWV-1 46-001, and POWV-1 63–002 displayed significantly lower infection rates compared to POWV-1 19065–059 and DTV 18071–054 (chi-squared test, **p < 0.01, ***p < 0.001).

Figure 5.

Growth kinetics of New York State (NYS) Powassan virus (POWV-1) and deer tick virus (DTV) isolates in Ixodes scapularis nymphs following immersion. Individual data include two separate, statistically equivalent experiments (t-test, p > 0.05, n = 4-19). Time (DPI) significantly influenced viral load (two-way ANOVA, p < 0.0001). Pairwise comparisons indicated statistical equivalence of viral load among strains, with the exception of POWV-1 63-002, which was significantly lower than all other strains at 21 DPI (two-way ANOVA, Tukey’s multiple comparisons, p < 0.017).