Development of a novel amplicon based whole-genome sequencing framework for improved surveillance of Toscana virus

Abstract

Toscana virus (TOSV), a Phlebovirus transmitted by sandflies, is a leading cause of aseptic meningitis in the Mediterranean region. Despite its clinical significance, underreporting and limited availability of complete genomic data hinder a thorough understanding of its genetic diversity and evolution. This study presents a novel amplicon-based whole-genome sequencing (WGS) method using Illumina library preparation kits and proprietary software to optimize workflows and enhance bioinformatic analyses. Primers targeting TOSV lineage A genomes were designed with PrimalScheme to generate 400 bp amplicons, incorporating degenerate bases to improve coverage. Library preparation utilized Illumina Microbial Amplicon Prep (iMAP) kits, followed by de novo assembly using BaseSpace DRAGEN Targeted Microbial software. The method's sensitivity was tested on viral propagates at various RNA concentrations (10⁴ to 10 copies/μL), demonstrating robust performance at concentrations above 10² copies/μL. Validation on high-titre viral propagates (n = 7), low-titre clinical samples (n = 15), and phlebotomine pools (n = 5) confirmed its reproducibility and ability to comprehensively cover coding regions. Cerebrospinal fluid samples yielded the most consistent results compared to urine and sandfly pools. This innovative WGS approach represents a significant advancement in TOSV genomic surveillance, enabling large-scale studies of its genetic diversity and evolutionary dynamics, which are critical for improving diagnostics and public health strategies.

Fig. 1. Maximum Likelihood phylogenetic tree of segment S, M and L of Toscana virus.

This figure illustrates the phylogenetic relationships among the sequences originally used for primer design and degeneration (marked in red) and other TOSV lineage A annotated as complete or near-complete segment sequences in GenBank. Phylogenetic trees were rooted using TOSV lineage B sequences: KC_776214.1 for segment S, KC_776215.1 for segment M, and KC_776216.1 for segment L, with the outgroup marked in bold. Bootstrap values ranging from 50 to 100 are shown. Where available, isolation location and isolation date are reported in the tree. The tree was inferred using IQ-TREE v.2.1.4, with automated model selection (TVMe+I+R2 for segment S (A), TVM+F+R2 for segment M (B), and GTR+F+G4 for segment L (C)) and 10,000 ultrafast bootstrap replicates.

Fig. 2. Graphical representation of mean % coverage and segment % coverage across dilutions.

A Comparison of mean percentage genome coverage at varying RNA input concentrations (copies/µL) for two TOSV isolates, IZSLER_181135/14 (blue) and RN_050724 (red), with error bars representing standard deviation. Percentage genome coverage of TOSV segments (L, M, S) across varying RNA input concentrations (copies/µL) for IZSLER_181135/14 (B) and RN_050724 (C), highlighting segment-specific performance with error bars indicating standard deviation. Images are representative of two independent experiments. Error bars represent standard deviations.

Fig. 3. Correlation between viral genome copies/µL and mean percentage genome coverage across different sample types (Vero E6 propagates, CSF, urine, and phlebotomine pools).

The size of the circles represents sequencing depth, highlighting the relationship between viral load (indicated by copies/µL), sequencing performance (mean genome coverage), and consensus sequence reliability (depth). Higher viral loads are associated with higher coverage and sequencing depth, with Vero E6 samples consistently achieving near-complete coverage. Variability is observed in clinical (CSF and urine) and vector (phlebotomine pool) samples, reflecting differences in matrix complexity and RNA quality, leading to different sequencing outputs. Vero E6 propagates are represented in green, CSF samples are represented in purple, urine samples are represented in pink, and phlebotomine are represented in orange.