Isolation of a novel insect-specific flavivirus with immunomodulatory effects in vertebrate systems

Abstract

We describe the isolation and characterization of a novel insect-specific flavivirus (ISFV), tentatively named Aripo virus (ARPV), that was isolated from Psorophora albipes mosquitoes collected in Trinidad. The ARPV genome was determined and phylogenetic analyses showed that it is a dual host associated ISFV, and clusters with the main mosquito-borne flaviviruses. ARPV antigen was significantly cross-reactive with Japanese encephalitis virus serogroup antisera, with significant cross-reactivity to Ilheus and West Nile virus (WNV). Results suggest that ARPV replication is limited to mosquitoes, as it did not replicate in the sandfly, culicoides or vertebrate cell lines tested. We also demonstrated that ARPV is endocytosed into vertebrate cells and is highly immunomodulatory, producing a robust innate immune response despite its inability to replicate in vertebrate systems. We show that prior infection or coinfection with ARPV limits WNV-induced disease in mouse models, likely the result of a robust ARPV-induced type I interferon response.

Keywords: Aripo virus; Flavivirus infection; Flavivirus pathogenesis; Insect-specific flavivirus; Super-infection exclusion; West nile virus.

Figure 1:

A maximum likelihood (ML) tree was constructed using NS5 gene sequences of representative members of the flavivirus genus (a). The ML phylogeny shows that ARPV clusters together with dual host ISFVs which show a close evolutionary relationship with vertebrate-infectious flaviviruses within the mosquito-borne flavivirus clade. Values at nodes indicate percent bootstrap values. The arrow indicates ARPV and the vertical line indicates vertebrate-infectious flaviviruses, and the red shaded box shows representative dual host associated ISFVs. Taxon labels include virus species name, virus species abbreviation and accession number. The scale bar represents percent nucleotide divergence. Transmission electron micrographs of ARPV in infected C6/36 cells showing (b) a cluster of virus particles inside an expanded vacuole; (c) virus particles are indicated with arrows and smooth membrane structures in the granular endoplasmic reticulum cisterns. Scale bars = 100 nm.

Figure 2:

Phase-contrast micrographs showing the cytopathic effects of ARPV in representative cell lines; (a) negative control for C7/10, (b) C7/10 cells infected with ARPV, (c) negative control for Cx. tarsalis, (d) Cx. tarsalis cells infected with ARPV; (e) negative control for T. ambionensis, and (f) T. ambionensis cells infected with ARPV. Images were taken three days post-infection. Scale bars = 200 μm.

Figure 3:

Replication kinetics of ARPV in representative mosquito and vertebrate cell lines post-infection (A) and post-electroporation (B). Monolayers were inoculated with an MOI of ~0.1. Data points represent the mean number of genome copies/ml for duplicate infections titrated in triplicate using qPCR tests in (3A) and triplicate infections titrated in triplicate in (3B). Error bars indicate the standard deviation of the mean. Transmission electron micrographs of ARPV-infected Vero cells show that ARPV enters by clathrin-mediated endocytosis by 15 sec (c), 1 min (d), and 5 mins (e) post-infection. An ARPV virion is localized at the surface of the cellular plasma membrane (c). A clathrin-coated pit is formed, and endocytosis is initialized after viral protein-host cell receptor binding (d) prior to ARPV entry via endocytosis (e). Scale bar (100 nm) is shown in the bottom left of each micrograph.

Figure 4:

ARPV infection of macrophages results in robust expression of interferon-associated gene responses. Bone-marrow-derived macrophages from naïve C57BL/6 mice were inoculated with ARPV, UV-inactivated ARPV, ZIKV, or culture media as a negative control. RNA was harvested six hours post-infection for DEG analysis. Gene expression data from each of the experimental groups were analyzed using edgeR to generate (a) volcano plots of differential gene expression and using Ingenuity Pathway Analysis (IPA) to generate (b) a heatmap schematic to illustrate signaling pathways identified as significantly up- or down-regulated following ARPV infection relative to controls. Color intensity represents the respective z-score range provided by IPA. Green squares represent up-regulated, and red squares represent down-regulated pathways. Significant pathways identified were grouped and displayed as either upstream regulators, canonical pathways activated or predicted biological functions impacted by ARPV infection.

Figure 5:

ARPV infection of macrophages results in robust cytokine production. Bone-marrow-derived macrophages from naïve C57BL/6 mice were inoculated with ARPV or culture media as a negative control (mock). Cytokine levels in cell culture supernatant were determined by bead-based Bioplex immunoassay. Infected macrophages produced significant levels of cytokines (a) TNF-α, (b) IL-6 and (c) IL-12, (d) IFN-γ, (e) IL-10, and (f) IL-17 at both 6- and 24-hours post-infection. Data points represent mean values, and error bars indicate standard deviation of the mean for triplicate infections. Significance was determined by two-way ANOVA with ad hoc Tukey’s test. Statistically significant differences are denoted by * p<0.05.

Figure 6:

ARPV limits WNV disease in mice due to a robust innate immune response. Six-week-old CD-1 mice were divided into nine (9) groups of n=6 and subcutaneously inoculated seven days before or one day before the WNV challenge with ARPV, YFV 17D, PBS, or C6/36 culture media. One group was also simultaneously co-inoculated with WNV and ARPV. Mice were then challenged with 10³ PFU of WNV and were monitored daily for weight loss (a) and survival (b) for 21 days post-challenge. Six-week-old Ifnar1^−/− mice were also divided into groups of n=6 and subcutaneously inoculated one day before the WNV challenge with ARPV or PBS. Mice were then challenged with 10³ PFU of WNV and monitored daily for weight loss (c) and survival (d) for five days post-challenge. Data points indicate mean values, and error bars indicate standard deviation of the mean. Significance was determined by two-way ANOVA (a,c) and by log-rank (Mantel-Cox) test (b,d). Statistically significant values are denoted by * p<0.05.