The National Animal Health Laboratory Network (NAHLN) is a collaborative network of veterinary diagnostic laboratories that provide testing capacity for ongoing surveillance programs and respond quickly and effectively to disease outbreaks as they occur. In 2019, Colorado State University, a member of the NAHLN, was activated by the National Veterinary Services Laboratories (NVSL) to respond to the VSV outbreak in Colorado. The samples sequenced herein were tested as part of these efforts. Swabs of vesicular lesions were collected by veterinarians in Colorado and submitted for VSV diagnostic testing at the Colorado State University Diagnostic Laboratories between July and October 2019.

RNA was extracted from samples using an automated bead-based platform and tested by real-time reverse transcriptase polymerase chain reaction (rRT-PCR) in accordance with the NAHLN standard operating procedure. Extracted RNA was stored at −80°C after initial testing. In some instances, original samples that had been stored at −80°C were re-processed in an attempt to improve sequencing quality. RNA from these samples was extracted using the same NAHLN SOP as previously used during initial diagnostic testing.

Samples with resulting cycle threshold (C_t) values of ≤26.99 upon initial rRT-PCR testing were selected for next generation sequencing. This cut-off was chosen based on results of a pilot study sequencing a subset of samples with Cts ranging from 16.09 to 33.29. RNA from the original extraction was sequenced in duplicate as previously described (14, 15). Briefly, DNA was depleted from extracted samples using the DNA-free DNase kit (Ambion, Austin, TX) following manufacturer's instructions. First-strand synthesis was performed using a Superscript™ II Reverse Transcriptase kit (Invitrogen, Carlsbad, CA) with random hexamer primers and one conserved VSV intergenic region specific primer (5). A NEBNext^® Ultra Non-Directional RNA Second Strand Synthesis Module was used to produce the second strand of cDNA. cDNA was purified using AmPure XP beads (Beckman-Coulter, Brea, CA), followed by the creation of cDNA libraries using a Nextera XT DNA Library Preparation kit. Next generation sequencing (NGS) was performed using an Illumina NextSeq550 instrument, with paired-end reads. Reads were quality trimmed and de novo assembled. Trimmed reads were also mapped to a VSIV reference genome [IN98COE, GenBank accession #AF473864; (16)]. Consensus sequences were extracted from each mapping and assembly for each sample then aligned, and a final consensus sequence was extracted from the alignment for each sample. All analyses were performed in CLC Workbench v21 using default parameters.

A total of 86 near-full-length sequences were generated (Supplementary Table S1), representing 27 counties in Colorado. The 10,821–11,185 nucleotide (nt) genomes encode for 5 structural and non-structural proteins: nucleocapsid protein (1,269 nt), phosphoprotein (798 nt), matrix protein (690 nt), glycoprotein (1,536 nt), and L-polymerase (6,330 nt), with intergenic spacer regions separating each protein coding region. These sequences contain a 14–22 nt insertion in the G-L intergenic region compared to IN98COE. Interestingly, insertions in the G-L intergenic region have been noted in other VSIV lineages originating in Central America (16), suggesting this lineage may have also originated in that region. However, further research is needed to determine the origin of the causative lineage of the 2019–2020 VSIV outbreak in the US.

The 86 sequences generated in the current study were aligned with all publicly available sequences from the 2019–2020 VSIV outbreak in the US (n = 20), and IN98COE was included as an outgroup to investigate genetic variation within the outbreak. Sequences were aligned using MUSCLE (17) implemented in Geneious Prime v2022 (www.geneious.com). The Tamura 3-parameter model with uniform rates was identified as the most appropriate model based on Bayesian information criterion (BIC), and a phylogenetic tree was constructed using maximum likelihood implemented in Mega X (18). The final consensus tree was visualized using FigTree v1.4.4 (19).

Phylogenetic analysis revealed three distinct groups within the 2019 sequences, and the 2020 sequences formed an additional distinct group (Figure 1). The 2019 sequences grouped by geographic area rather than time. Isolates collected at timepoints throughout the outbreak are represented in each of the 2019 groups. Isolates in group 1 were collected in counties in the northeast part of Colorado, while group 2 were collected in counties in the southwest, and group 3 were collected in the central part of the state. Within-group similarity was 99.93–99.96%, and between-group similarity was 99.80–99.90% (Table 1). Similar to patterns in previous outbreaks, the sequences from the expansion year of the outbreak (2020) form a monophyletic group within the sequences from the incursion year (2019) (14). Overall, there was low variation among isolates from the 2019–2020 outbreak, which is consistent with previous findings of low genetic variability in VSIV, especially in comparison with VSNJV (14, 20). Ongoing analyses of these sequences are investigating additional evolutionary, epidemiological, and ecological aspects of the outbreak. Similar to previous work with VSNJV (21), we are also developing an infectious clone of this epidemic VSIV strain for use in experimental investigations of the pathogenesis of epidemic VSV in livestock.