Invertebrate Iridescent Virus 6, a DNA Virus, Stimulates a Mammalian Innate Immune Response through RIG-I-Like Receptors

Abstract

Insects are not only major vectors of mammalian viruses, but are also host to insect-restricted viruses that can potentially be transmitted to mammals. While mammalian innate immune responses to arboviruses are well studied, less is known about how mammalian cells respond to viruses that are restricted to infect only invertebrates. Here we demonstrate that IIV-6, a DNA virus of the family Iridoviridae, is able to induce a type I interferon-dependent antiviral immune response in mammalian cells. Although IIV-6 is a DNA virus, we demonstrate that the immune response activated during IIV-6 infection is mediated by the RIG-I-like receptor (RLR) pathway, and not the canonical DNA sensing pathway via cGAS/STING. We further show that RNA polymerase III is required for maximal IFN-β secretion, suggesting that viral DNA is transcribed by this enzyme into an RNA species capable of activating the RLR pathway. Finally, we demonstrate that the RLR-driven mammalian innate immune response to IIV-6 is functionally capable of protecting cells from subsequent infection with the arboviruses Vesicular Stomatitis virus and Kunjin virus. These results represent a novel example of an invertebrate DNA virus activating a canonically RNA sensing pathway in the mammalian innate immune response, which reduces viral load of ensuing arboviral infection.

Fig 1. IIV-6 elicits a type I IFN response in mammalian cells.

(A-B) MEFs, (C) human A549 cells, or (D) mouse RAW 264.7 macrophages were infected with IIV-6 or DCV at an MOI of 1 TCID₅₀/cell. 24 hours post-infection supernatant and total RNA were collected to measure (A) IFN-β and TNF-α induction by qRT-PCR, (B, D) IFN-β secretion by ELISA, or (C) ISRE-luciferase activity in HEK 293T cells. Poly(dA:dT) was transfected into A549 cells with lipofectamine as a positive control for ISRE activity. (D) Mock-infected cells did not secrete significantly higher levels of IFN-β at 24 hours as compared to either mock- or IIV-6-infected cells at 6 hours. ELISA and qRT-PCR assays were performed in biological duplicate, and ISRE-luciferase assay for IFN-β secretion from A549 cells was completed in biological triplicate (*, P < 0.05).

Fig 2. IRF3 and STAT1 are activated during IIV-6 infection in mammalian cells.

(A) A549 cells or (B-C) MEFs were infected with IIV-6 at an MOI of 1 TCID₅₀/cell. Cellular protein lysates were analyzed by Western blot for phosphorylated IRF3 and STAT1, total IRF3 and STAT1, and actin. 1 μg/mL poly(dA:dT) was transfected into cells using lipofectamine as a positive control, and blank lipofectamine was used as a negative control.

Fig 3. NFκB is activated during IIV-6 infection.

(A, C) RAW 264.7 macrophages or (B, D) MEFs were infected with IIV-6 at an MOI of 10 TCID₅₀/cell. (A-B) 24 hours post-infection cells were fixed onto coverslips, and stained for NFκB and DAPI. (C-D) Protein samples were analyzed using Western blot for phosphorylated IκB, total IκB, and actin.

Fig 4. IIV-6 enters mouse embryonic fibroblasts.

S2 cells and MEFs were infected with IIV-6 at an MOI of 1 TCID₅₀/cell. S2 cells were fixed at 24 hours post-infection and MEFs were fixed at 72 hours post-infection. The top row indicates S2 controls for the presence of IIV-6, and the zoom of infected S2 cells illustrates the presence of a viral factory in which capsids are developing. Arrows indicate representative virions, and arrowheads indicate empty capsids. The bottom row depicts MEFs, and the zoom indicates the presence of viral particles.

Fig 5. IFN-β secretion following IIV-6 infection is reduced in the absence of RLRs and Dicer.

(A) Schematic representing the homologous DExD/H-box helicase domains shared by Drosophila Dicer-2, and human or mouse Dicer, RIG-I, and MDA5. Orange numbers indicate the amino acid residues where the helicase domain begins and ends and the total length of the protein. Percentages on the right are the percent identity of the amino acid sequence of Drosophila Dicer-2 to the mammalian protein in the helicase domain determined by Clustal Omega alignment. (B-F) STING^-/-, Dicer^-/-, MAVS^-/-, RIG-I^-/-, or MDA5^-/- MEFs and their corresponding wild-type MEFs were infected with IIV-6 at an MOI of 1 TCID₅₀/cell in biological duplicate. Supernatant from cells was collected at 6 and 24 hours post-infection and analyzed for secreted IFN-β by ELISA. (G-H) Ddx58 and Ifih1, the genes encoding RIG-I and MDA5, respectively, were knocked down in MEFs using siRNAs, as determined by qRT-PCR. (I) 72 hours after siRNA transfection, MEFs were infected with IIV-6 at an MOI of 1 TCID₅₀/cell. Total RNA was collected for qRT-PCR 24 hours post-infection to measure IFN-β gene induction in biological triplicate. (B-F) Knockout cell lines were compared to the wild-type parental line at each time point, and (I) IFN-β expression for siDDX58 or siIFIH1 was compared to siControl (*, P < 0.05).

Fig 6. Host RNA polymerase III activity contributes to IFN-β activation following IIV-6 infection.

(A-B) Genomic DNA was extracted from purified IIV-6 particles, treated with RNase A, and transfected into MEFs at 1 μg/mL in biological triplicate. (A) Protein was collected at 6 hours post-infection, triplicate protein samples were combined, and samples were analyzed by Western blot for IRF3 activation. Poly(dA:dT) was transfected as a positive control for IRF3 activation. (B) Supernatant was collected at 24 hours post-infection and analyzed by ELISA for secreted IFN-β. (C-D) MEFs were transfected with control siRNA (siControl) or siRNA targeting RNA Pol III (siPol III) for 72 hours and (C) total RNA was collected to measure RNA Pol III levels by qRT-PCR. (D) After RNA Pol III knockdown, cells were infected with IIV-6 at an MOI of 1 TCID₅₀/cell. IFN-β secretion was measured by ELISA 24 hours post-infection. (E) RNA Pol III in MEFs was inhibited by treating cells with 100 μM ML-60218 for 12 hours prior to infection with IIV-6 or transfection with poly(dA:dT). Samples for ELISA and qRT-PCR were measured in biological triplicate. IFN-β induction was compared to (B) blank lipofectamine or (D) siControl (*, P < 0.05).

Fig 7. The mammalian immune response to IIV-6 restricts arbovirus infection.

(A) Schematic of priming experiment: MEFs were infected with IIV-6 at an MOI of 1 TCID₅₀/cell. 24 hours post-infection cell culture supernatant was removed, and viral particles in the cell culture supernatant were removed by centrifugation. Supernatant was used to prime new MEFs for 24 hours. Following priming, MEFs were infected with VSV or KUNV at an MOI of 1 PFU/cell, and cell culture supernatant was collected. (B-C) At the indicated time points post-infection, supernatant was collected for plaque assay on BHK21 cells to determine titers of (B) VSV or (C) KUNV. (D) Schematic of co-infection: MEFs were infected with IIV-6 at an MOI of 1 TCID₅₀/cell for 1 hour, followed by infection with VSV or KUNV at an MOI of 1 PFU/cell, and cell culture supernatant was collected following infection for plaque assay. (E-F) At various times, supernatant was collected for plaque assay on BHK21 cells to measure titer of (E) VSV or (F) KUNV. Values above black bars (B-C, E-F) represent the fold-change decrease as compared to time-matched mock-primed MEFs. Assays with VSV were completed in biological duplicate, and assays with KUNV were completed in biological triplicate. Mock infection was compared to IIV-6 infection or priming for each time point (* P < 0.05; *** P < 0.001).