TITAN-RNA: A hybrid-capture sequencing panel detects known and unknown Flaviviridae for diagnostics and vector surveillance

Abstract

Clinical testing and public health surveillance can be significantly improved by incorporating sequencing-based molecular detection and subtyping for real-time monitoring of virus evolution. With phylogenetic analysis used for speciation and variant subtyping, target analyte specificity can be relaxed well beyond typical parameters acceptable in PCR-based diagnostics. Hybrid capture is a promising way to enrich large numbers of sequences with maximal flexibility, using standard molecular biology laboratory equipment and small benchtop sequencers. Here, we report the development and bench validation of a hybrid capture based next-generation sequencing diagnostic panel for all known viral tick-borne pathogens, TITAN-RNA. Based on systematic testing with simulated novel viruses and field samples, we determined a 10% tolerance for evenly distributed mutations or 27% tolerance for naturally occurring viral divergence. The TITAN-RNA extrapolated limit of detection in blood is 19.1 genome copies by complementary log-log analysis, and linearity performance (R² ≥ 0.99) is amenable for its use as a quantitative assay. As proof of principle for public health surveillance and evolutionary studies, we report two putatively novel segmented Flavi-like viruses in New York State, USA, identified from the invasive Haemaphysalis longicornis tick.

Figure 1.. Divergent Sequences.

Heatmap showing the pairwise-distance relationship between experimental and naturally occurring sequences. The bluest shade represents lowest dissimilarity, and the reddest shade represents the highest dissimilarity.

Figure 2.. Linearity results.

Linearity plotted as log₁₀ Mean Contig R (number of reads aligned per taxonomically matched contig) versus log₁₀ Concentration (copies/μl) (n = 3). A.) POWV, B.) HRTV, C.) DBV. The X-axis represents the log₁₀ of total copies and the Y-axis represents the log₁₀ of mean contigs detected.

Figure 3.. Limit of Detection Results.

Limit of detection plotted as fraction-positive samples versus total copies (n ≥ 3) A.) POWV synthetic RNA in nuclease-free water via qRT-PCR, B.) POWV synthetic RNA in nuclease-free water via TITAN-RNA (TBD_Virus), C.) POWV synthetic RNA in ear notches via TITAN-RNA (TBD_Virus), D.) POWV synthetic RNA in blood samples via TITAN RNA (TBD_Virus).

Figure 4.. Mean Coverage Depth.

Log₁₀ of mean coverage depth of NS5 using the TBD_Virus panel on random POWV sequence variants. Detection was challenged with reference sequences of two NS5 genes not included in this panel (TBEV and ZIKV) to assess the ability of POWV baits to pull down related known species.

Figure 5.. Phylogenetic analysis of tick-borne Flaviviridae RdRP (NS5 or NS5-like).

Maximum likelihood tree of segmented and unsegmented Flaviviridae RNA-dependent RNA Polymerase (NS5) homologs. Scale represents number of nucleotide substitutions. Bootstraps = 1000.

Figure 6.. Phylogenetic analysis of tick-borne Flaviviridae glycoprotein (E or VP1/1ab).

Maximum likelihood tree of segmented and unsegmented Flaviviridae glycoprotein homologs. Scale represents number of nucleotide substitutions. Bootstraps = 1000.