PKR Binds Enterovirus IRESs, Displaces Host Translation Factors, and Impairs Viral Translation to Enable Innate Antiviral Signaling

Figures

• 

  FIG 1. Time course of PVSRIPO infections in primary human monocyte-derived dendritic cells (mdDCs), A375 melanoma, Hs683 glioma, and M059J glioma cells. (A and B) mdDCs from donors 1 (A) and 2 (B) were infected with PVSRIPO and lysates were analyzed by immunoblot for eIF4G1 cleavage, viral (2C) translation, STAT1 activation [p-STAT1(Y701)/STAT1], and ISG induction (IFIT1). (C to E, left) A375 (C), Hs683 (D), and M059J cells (E) were infected with PVSRIPO and lysates were studied by immunoblot for eIF4G1 cleavage (*, eIF4G1 cleavage fragment), viral translation (2C/2BC), IRF3 [p-IRF3(S396)], and STAT1 activation. A minimum of 3 independent series were performed; representative results are shown. (C to E, right) Viral (2C) translation and type I interferon (IFN) signaling [p-STAT1(Y701)] were quantified over a 0–32 hpi period. The values represent the average abundance of signal in 3 independent series (%max. detected, normalized to GAPDH). Standard error of the mean (SEM) is indicated by error bars. MOI, multiplicity of infection; hpi, hours postinfection.

• 

  FIG 2. Immediate early innate interference with PVSRIPO involves host double-stranded (ds)RNA sensors interacting with the viral 5′ untranslated region (UTR). (A) A375 cells were pretreated with recombinant human IFNα2 or vehicle (16 h) prior to mock (0 h), CAV21 or PVSRIPO infection. Lysates were collected and analyzed by immunoblot as described for Fig. 1; representative results observed in 3 independent assays. (B) List of proteins of interest identified in our first series of liquid chromatography-mass spectrometry/mass spectrometry (LC-MS/MS) analyses of samples pulled down with PVSRIPO 5′-UTR bait RNA (see Table S2). The list includes all hits that were enriched in both bait+ samples with low/absent binding to bait- bead samples (see main text for detail). “Total spectrum count” represents a measure of (bait RNA) binding activity of the protein in question.

• 

  FIG 3. IFN-inducible host innate dsRNA sensors identified to interact with the PVSRIPO 5′ UTR. (A) The top 21 RNA-binding proteins that specifically interacted with the biotinylated RNA bait. Cytosolic fractions of “plain” A375 lysates, or of lysates from A375 cells treated with 100 U/mL recombinant human IFNα2 (16 h) (A375+IFNα), or of lysates from A375 cells treated with PVSRIPO (MOI 10, 48 h) (A375+PVSRIPO) were incubated with biotinylated HRV2 internal ribosomal entry site (IRES) bait (30 min) and pulled down with Streptavidin beads. The pulled down samples were analyzed by LC-MS/MS. Total Spectrum Count represents maximal binding value. Proteins with non IFN-responsive (red) and responsive (yellow) binding, subjected to in-depth analyses, are highlighted. (B) Confirmatory immunoblots of proteins identified by LC-MS/MS (A) in cytosolic fractions of lysates from plain, from IFNα-pretreated, or from PVSRIPO-infected A375; representative results from 3 independent experimental series.

• 

  FIG 4. Depletion of double-stranded RNA-activated protein kinase (PKR) enhances early PVSRIPO translation. A375 cells with dox-inducible depletion of DHX9 (A), LGP2 (B), ADAR1 (C), or PKR (D) were infected with PVSRIPO and viral translation was assessed by immunoblot of viral 2C protein. Dox-inducible cell lines were generated using 2 (DHX9, ADAR1, or PKR) or 3 (LGP2) different shRNAs for each depletion target (Table S1). The results shown are from the respective cell lines with more efficient dox-inducible, target-specific depletion. Two series were conducted with each cell line; representative results are shown. Levels of viral 2C in dox-induced (depleted) relative to mock-induced (nondepleted) samples were quantified for each interval in PVSRIPO-infected samples. Graphs represent means ± SE. *, P < 0.05, **, P < 0.005, respectively.

• 

  FIG 5. PKR binds to stem-loop domains (SLDs) 5–6 in the HRV2 IRES as a dimer, resulting in autophosphorylation of the kinase activation loop (see Fig. S2 and S3 for extended data). (A) Schematic representation of the HRV2 IRES SLD structure with an approximate footprint for eukaryotic initiation factor (eIF) 4G:eIF4A (3). (B) Binding of PKR, ADAR1, PCBP2, and DHX9 to different SLD fragments as indicated. Cytosolic fractions of A375 cells, “plain” or infected with PVSRIPO (MOI 10, 48 h), were incubated with biotinylated RNA fragments (30 min) prior to pull down with Streptavidin beads. Bound proteins were analyzed by immunoblot (see Fig. S2 for corresponding input blots). (C) PKR and DHX9 binding with biotinylated wild-type (WT) SLD 5 and three distinct SLD 5 variants carrying mutations (Mut 1–3) disrupting base-pairing structure as shown. (D) C16 prevented PKR activation, evident as autophosphorylation of T446 in the kinase activation loop, upon binding to biotinylated SLD 1–3, SLD 5–6, or poly(I·C). Methylcrotonyl-coa carboxylase subunit 1 (MCCC1) nonspecifically associates with streptavidin beads (in the absence of biotinylated RNA bait) and was used as a control to ensure equal…

• 

  FIG 6. PKR catalytic activity suppresses immediate early PVSRIPO translation (see Fig. S4 and S5 for extended data). (A) Detailed long-range kinetics of the innate antiviral response to PVSRIPO in A375 cells. The dynamics of key innate antiviral signaling cascades were examined by immunoblot (left) and quantitated over time (right). The results of a similar assay in Hs683 cells is shown in Fig. S4. (B) Effect of PKR inhibition with C16 on early viral translation in A375 cells. Cells were infected as shown in Fig. 1 and C16 was added to medium at the time of infection (0-h treatment delay) or at the indicated intervals thereafter. The effect on viral translation was determined by immunoblot of viral 2C at 4 and 6 hpi. The experiment was repeated 3 times; average 2C levels at 6 hpi (relative to mock-treated cells) are graphed at bottom; error bars represent SEM. See Fig. S5 for dose-optimization of small molecule inhibitors used in this study. (C) Schematic representation of inhibitors of PKR downstream signaling: PKR catalytic activity proper (C16), IKKα:β binding/NF-κB activation (IKK16), and the integrated stress response (ISRIB). (D) The effect of C16, IKK16, or ISRIB on early (6 hpi)…

• 

  FIG 7. A biphasic innate response to PVSRIPO: early sensing by PKR and a late, polar TBK1-IRF3 inflammatory response orchestrated by MDA5 (see Fig. S6 and S7 for extended data). (A) The effect of TBK1 (Bx795) or PKR (C16) inhibition on early (6 hpi) and late (24, 48 hpi) viral translation in M059J cells, determined by 2C immunoblot. Protein levels at each time point were normalized to corresponding mock-treated samples and plotted in the graph at right. See Fig. S6 for a similar assay in A375 cells. (B and C) Comparing the effect of PKR (C16; B) and JAK/STAT (Ruxolitinib [Rux]; C) inhibition on early (6 hpi) or late (24-48 hpi) viral translation and STAT1 activation in PVSRIPO-infected A375 cells. A375 cells were treated with C16 (B) or Rux (C) at the time of PVSRIPO addition (0 h); viral translation (2C) and STAT1 activation/MDA5 induction were assessed by immunoblot. See Fig. S7 for testing the effect of Rux on PVSRIPO infection of Hs683 cells. All assays were repeated 3 times; representative immunoblots are shown. Relative expression levels of viral 2C were calculated for each time point and are plotted at the bottom.