The Interferon-Induced Exonuclease ISG20 Exerts Antiviral Activity through Upregulation of Type I Interferon Response Proteins

FIG 1. Inducible expression of murine ISG20 suppresses alphavirus replication. (A) Two clones of tet-off MEFs expressing FLAG-tagged eGFP, murine ISG20, or the exonuclease domain mutant murine ISG20D94G (ExoII) were induced for 72 h and then immunoblotted for the FLAG-tagged protein. DOX, doxycycline. (B to D) Induced tet-off MEFs were infected with (B) WT CHIKV, (C) WT VEEV, or (D) VEEV-G3A (MOI of 0.1). Replicated virus levels were measured in cell supernatants by BHK-21 plaque assay. *, P < 0.5; ***, P < 0.001 (2-way ANOVA with Holm-Sidak post hoc analysis against eGFP control) (n = 6, 2 independent experiments).

FIG 2. ISG20 suppresses translation of non-IRES-containing capped mRNA. (A) tet-off MEFs overexpressing the indicated protein were infected with CHIKV-nsP3-nLuc nonstructural reporter virus at an MOI of 0.1, and lysates were collected at the indicated time points. Reporter signal data are given as ratios of relative luminescence units normalized to cellular protein levels in micrograms (RLU/μg). ***, P < 0.001 (2-way ANOVA with Holm-Sidak post hoc analysis against GFP control) (n = 6, 2 independent experiments). (B) Schematic representation of type 0 capped and polyadenylated nonreplicating translation reporters. The luciferase gene is flanked by 5′ and 3′ nontranslated regions for the indicated mRNA. The CHIKV reporter also encodes the translational initiation sequences of the nsP1 protein (diagonal lines). (C to F) Induced tet-off MEFs overexpressing the indicated protein were electroporated with 7.5 μg of indicated RNA and then sampled for 4 h. Relative luminescence at 3 h is given as a percentage of the initial signal at 30 min. (C) Host reporter. (D) CHIKV reporter. (E) EMCV reporter. (F) CrPV reporter. **, P < 0.01 (2-way ANOVA with Holm-Sidak post hoc analysis against eGFP…

FIG 3. ISG20 expression does not accelerate degradation of CHIKV RNA. (A) MEFs were infected with CHIKV expressing nLuc in frame with the nsP3 nonstructural protein at an MOI of 5 and then treated with 0 or 10 μM cycloheximide (Cyc) to block translation and, ultimately, replication of CHIKV RNA. (B) MEFs were infected with CHIKV-LR (MOI = 5) and then treated with 10 μM cycloheximide. RNA was collected at the indicated time p.i. and analyzed by qRT-PCR for the CHIKV 3′ terminus. (B) Data represent results of 2-way ANOVA with Holm-Sidak post hoc analysis against eGFP control. (n = 6, 2 independent experiments; data were not statistically significant). (C) RNA isolated from ISG20 MEFs is functional. The CHIKV-translation reporter was isolated from tet-off MEFs overexpressing GFP, ISG20, or ExoII mutant, retransfected into MEFs expressing either GFP or ISG20 for 3 h, and assayed for firefly luciferase reporter activity, and luciferase activity data from second transfections are given as ratios of firefly luciferase signals from h 3 versus h 0 samples taken from first transfection. Data represent results of ANOVA with Dunnett’s post hoc analysis against GFP control (n = 6, 2 independent…

FIG 4. ISG20 expression induces antiviral gene transcription. RNA-seq was performed on two sepearte clones of MEFs expressing eGFP, ISG20, or ExoII. (A) Differential gene expression was analyzed, and upregulated canonical pathways are shown. Numbers on top of each column indicate the total number of genes investigated in each pathway. (B) A heat map of the most significant ISG20-induced genes is shown. Temperature ranges are shown as fragments per kilobase of transcript per million mapped reads (FPKM) from green (low) to red (high). P < 0.001 (Cuffdiff 2 gene expression model; false-discovery rate [FDR] = 5%).

FIG 5. ISG20 induces a subset of antiviral genes. tet-off MEFs expressing eGFP, ISG20, or ExoII mutant were induced for 72 h, and total cellular RNA and protein were collected. (A) A selection of genes shown to be induced by ISG20 overexpression through RNA sequencing were confirmed by qRT-PCR (n = 6, 2 independent experiments). NC, no change. (B and C) IFIT1 protein induction was confirmed by Western blotting (B) and quantified (C). (A and C) *, P < 0.05; **, P < 0.01 (Mann-Whitney test). (For panels B and C, n = 2, 2 independent experiments).

FIG 6. ISG20 expression activates transcriptional pathways involving IRF3. (A) MEFs expressing eGFP, ISG20, and ExoII were induced and transfected with the luciferase driven by the IFN-β promoter (pβLUX) or the IRF3 responsive elements (PRDI/III) or the NF-κB element of the IFN-β promoter (PRDII) along with a SV40 promoter-driven Renilla plasmid. Firefly luciferase RLUs were normalized to Renilla RLUs, and induction data are given as percentages of untreated GFP cell results. (B) MEFs expressing GFP, ISG20, and ExoII mutant were induced and treated at 24 h with siIrf3 or siRNA scramble for an additional 48 h. Selected genes were assessed by qRT-PCR, and induction levels are given as fold changes from sham-treated GFP control results. (C) IFN levels were measured in the tet-off MEF cell culture supernatant before and after poly(I·C) lipid transfection. *, P < 0.05; **, P < 0.01; ***, P < 0.001. (A and C) ANOVA with Dunnett’s post hoc analysis (n = 6, 2 independent experiments). (B) Mann-Whitney test (n = 6, 2 independent experiments).

FIG 7. Isg20−/− mouse data support the idea of increased virulence of the IFIT1-sensitive VEEV-G3A mutant. (A to D) B6 and Isg20−/− mice received bilateral subcutaneous inoculations of 103 PFU of VEEV-G3A in rear footpads and were monitored for signs of disease based on (A and B) clinical scoring of behavior, (C) weight loss, and (D) average survival. (E) Mice were sacrificed and perfused at the indicated time point, and tissues were assessed for serum IFN. (F to I) Virus replication was measured in (F) popliteal lymph node, (G) serum, (H) spleen, and (I) brain at the indicated time points. (J to L) IFN-β mRNA was measured in (J) popliteal lymph node, (K) spleen, and (L) brain after VEEV-G3A infection. Data are represented as fold change in IFN-β compared to mock treatment results. *, P < 0.05; **, P < 0.01. (D) log-rank test (n = 6 mice, 2 independent experiments). (E to I) Multiple Student’s t tests (n = 8 mice, 2 independent experiments). (F) Mann-Whitney test (n = 8 mice, 2 independent experiments).

FIG 8. A deficiency of ISG20 leads to increased CHIKV and VEEV-G3A replication in primary cells. Primary cells were isolated from age- and sex-matched B6 and Isg20−/− mice. Cells were treated with 0, 10, or 100 IU/ml IFN-α4/β (1:1 ratio) for 4 h and then infected with (A and B) CHIKV or (C and D) VEEV-G3A (MOI = 1). Virus titers were determined at 24 h.p.i. by plaque assay. (E) ISG induction was measured in uninfected, 6-h IFN-treated B6 or Isg20−/− MEFs by qPCR. *, P < 0.05; **, P < 0.01; ***, P < 0.001 (multiple Student’s t tests). (A to D) n = 6, 2 independent experiments. (E and F) n = 3, 2 independent experiments.