LincRNA-EPS impairs host antiviral immunity by antagonizing viral RNA-PKR interaction

Abstract

LincRNA-EPS is an important regulator in inflammation. However, the role of lincRNA-EPS in the host response against viral infection is unexplored. Here, we show that lincRNA-EPS is downregulated in macrophages infected with different viruses including VSV, SeV, and HSV-1. Overexpression of lincRNA-EPS facilitates viral infection, while deficiency of lincRNA-EPS protects the host against viral infection in vitro and in vivo. LincRNA-EPS^-/- macrophages show elevated expression of antiviral interferon-stimulated genes (ISGs) such as Mx1, Oas2, and Ifit2 at both basal and inducible levels. However, IFN-β, the key upstream inducer of these ISGs, is downregulated in lincRNA-EPS^-/- macrophages compared with control cells. RNA pulldown and mass spectrometry results indicate that lincRNA-EPS binds to PKR and antagonizes the viral RNA-PKR interaction. PKR activates STAT1 and induces antiviral ISGs independent of IFN-I induction. LincRNA-EPS inhibits PKR-STAT1-ISGs signaling and thus facilitates viral infection. Our study outlines an alternative antiviral pathway, with downregulation of lincRNA-EPS promoting the induction of PKR-STAT1-dependent ISGs, and reveals a potential therapeutic target for viral infectious diseases.

Keywords: PKR; antiviral immunity; lincRNA-EPS; lncRNA; type I interferon.

Figure 1. Downregulation of lincRNA‐EPS by host antiviral immunity

1. A

   RT–qPCR analysis of lincRNA‐EPS transcripts in the BMDMs infected with VSV (MOI 1), SeV (MOI 1), or HSV‐1 (MOI 5) for indicated time points.

2. B

   RT–qPCR analysis of lincRNA‐EPS transcripts in the BMDMs transfected with 1 μg/ml low molecular weight (LMW), high molecular weight (HMW) polyI:C, and polydA:dT for 6 h.

3. C, D

   RT–qPCR analysis of lincRNA‐EPS transcripts in the BMDMs stimulated with 500 U/ml IFN‐α for different time points (C) or different concentrations for 2 h (D).

4. E, F

   RT–qPCR analysis of lincRNA‐EPS transcripts in the BMDMs stimulated with 500 U/ml IFN‐β for different time points (E) or different concentrations for 2 h (F).

5. G, H

   RT–qPCR analysis of lincRNA‐EPS transcripts in the WT and Ifnar1 ^−/− BMDMs transfected with 1 μg/ml polyI:C (LMW) for 6 h (G) and infected with SeV (MOI 1) for indicated time points (H).

6. I

   BMDMs were pretreated with NF‐κB inhibitor BAY11‐7082 (1 μM) for 1 h and infected with VSV (MOI 1) for 6 h. The percentage of lincRNA‐EPS transcripts downregulation in the VSV‐infected group compared with the Mock group was calculated.

7. J

   Cell nucleus and cytoplasm were separated from untreated (Mock) and VSV‐infected iBMMs, RNA was extracted for RT–qPCR analysis and compared with Mock group.

Data information: Data of (A‐J) are shown as the mean ± s.d. from three independent experiments. *P < 0.05 and **P < 0.01 by unpaired Student’s t‐test.

Figure EV1. The signaling pathways modulating lincRNA‐EPS expression

1. RT–qPCR analysis of lincRNA‐EPS transcripts in the WT and Ifnar1 ^−/− BMDMs infected with WSN (MOI 1) and VSV (MOI 1) for 6 h.

2. BMDMs were pretreated with NF‐κB inhibitor BAY11‐7082 (1 μM), p38 MAPK inhibitor SB203580 (5 μM), ERK inhibitor PD98059 (50 μM) and JNK inhibitor SP600125 (20 μM) for 1 h prior and infected with VSV (MOI 1) for 6 h. The phosphorylation and total proteins were detected by Western blot. GAPDH and α‐tubulin were shown as loading control.

3. The percentage of lincRNA‐EPS transcripts downregulated in the VSV‐infected group compared with Mock group were calculated.

4. Copy‐number analysis of lincRNA‐EPS transcripts in several cell types by RT–qPCR. Standard curve was generated using in vitro transcribed RNA molecule of lincRNA‐EPS as template.

Data information: Data of (A, C) are shown as the mean ± s.d. from three independent experiments, **P < 0.01 and ns, not significant by unpaired Student’s t‐test. Data of (B, D) are representative results from three independent experiments.

Figure 2. LincRNA‐EPS facilitates viral infection in macrophages

1. A

   RT–qPCR analysis of lincRNA‐EPS transcripts in the lincRNA‐EPS‐stably‐overexpressed iBMMs (lincRNA‐EPS) and corresponding control cells (EV).

2. B

   LincRNA‐EPS iBMMs were infected with VSV (MOI 1) for 8 h, VSV titers were measured by plaque assay.

3. C, D

   The EV and lincRNA‐EPS iBMMs were infected with SeV (MOI 1) (C) or HSV‐1 (MOI 5) (D) for 12 h. The viral RNA was measured by RT–qPCR.

4. E

   The transcripts level of lincRNA‐EPS in control (sgCtrl) and lincRNA‐EPS knockdown (sglincRNA‐EPS) RAW264.7 cells were measured by RT–qPCR.

5. F, G

   sgCtrl and sglincRNA‐EPS RAW264.7 cells were infected with VSV‐GFP (MOI 0.1) for 6 h or infected with VSV (MOI 0.1) for 8 h. The fluorescence of GFP was checked by microscope (F) and the viral titer of VSV from the cell supernatant was measured by plaque assay (G).

6. H, I

   sgCtrl and sglincRNA‐EPS RAW264.7 cells were infected with SeV (MOI 1) for 8 h (H) and HSV‐1 (MOI 5) for 12 h (I). The viral RNA was measured by RT–qPCR.

Data information: Data of (A–E) and (G–I) are shown as the mean ± s.d. from three independent experiments. *P < 0.05 and **P < 0.01 by unpaired Student’s t‐test. Data of (F) are representative images from three independent experiments, scale bar, 100 μm.

Figure 3. Knockout of lincRNA‐EPS enhances host antiviral ability

1. A, B

   WT and lincRNA‐EPS ^−/− iBMMs were infected with VSV‐GFP (MOI 0.1) for 6 h, and the lincRNA‐EPS transcripts were measured by RT–qPCR (A), and the fluorescence of GFP were checked by microscope (B).

2. C

   Cell supernatant from VSV‐infected WT and lincRNA‐EPS ^−/− iBMMs (MOI 0.1, 8 h) were collected and the viral titer were measured by plaque assay.

3. D, E

   WT and lincRNA‐EPS ^−/− iBMMs were infected with SeV (MOI 1) for 8 h (D) or HSV‐1 (MOI 5) for 12 h (E). The viral RNA was measured by RT–qPCR.

4. F

   Peritoneal macrophages isolated from WT and lincRNA‐EPS ^−/− mice were infected with VSV (MOI 1) for 10 h, and the viral titer were measured by TCID₅₀ assay.

5. G

   Eight weeks female lincRNA‐EPS ^−/− mice (n = 12) and WT littermates (n = 12) were injected (i.v.) with VSV (lethal dose, 1 × 10⁸ pfu/g), and the survival situation was monitored for 120 h.

6. H–J

   Eight weeks female lincRNA‐EPS ^−/− mice (n = 5) and WT littermates (n = 5) were injected (i.v.) with VSV (sub‐lethal dose, 6 × 10⁷ pfu/g) for 12 h, and negative control groups were injected with PBS (n = 3). Pathological section of liver and lung were harvested by H&E staining (H). Serum, liver, and lung from the VSV‐infected mice were collected. The viral load of serum and tissue homogenate were measured by TCID₅₀ assay (I), and the serum IFN‐β protein level were checked by ELISA (J).

Data information: Data of (A, C–F) are shown as the mean ± s.d. from three independent experiments, data of (I, J) are shown as the mean ± s.d. of a typical representative result from three independent experiments, and one dot represents a mouse, *P < 0.05 and **P < 0.01 by unpaired Student’s t‐test. Data of (G) are calculated with Log‐rank (Mantel‐Cox) test, **P < 0.01. Data of (B, H) are representative images from at least three independent experiments, scale bar,100 μm.

Figure EV2. Rescued expression of lincRNA‐EPS inhibits host antiviral ability

1. A

   The lincRNA‐EPS transcripts were measured by RT–qPCR in the control (EV) and lincRNA‐EPS‐rescued (lincRNA‐EPS) iBMMs.

2. B, C

   LincRNA‐EPS‐rescued iBMMs and control group were infected with VSV (MOI 1) and SeV (MOI 1) for 12 h, VSV titers were measured by plaque assay (B) and SeV RNA level were checked by RT–qPCR (C).

Data information: All data are shown as the mean ± s.d. from three independent experiments, *P < 0.05 and **P < 0.01 by unpaired Student’s t‐test.

Figure 4. Greater induction of antiviral ISGs in the lincRNA‐EPS ^−/− macrophages

1. Overview of GSEA analysis using the whole transcriptome of RNA‐seq from biological duplicates of VSV (MOI 0.1)‐infected WT and lincRNA‐EPS ^−/− iBMMs for 6 h.

2. The mean RPKM values of the biological duplicates of Mock or VSV‐infected WT and lincRNA‐EPS ^−/− iBMMs were calculated, and the upregulated antiviral genes of lincRNA‐EPS ^−/− iBMMs comparing to WT iBMMs were listed by heatmap.

3. RT–qPCR analysis of Mx1, Oas2, Ifit2, Irf7 mRNA level in the WT and lincRNA‐EPS ^−/− iBMMs transfected with 1 μg/ml polyI:C, polydA:dT or infected with VSV (MOI 0.1) and SeV (MOI 1) for 6 h, or HSV (MOI 5) for 12 h.

4. RT–qPCR analysis of Mx1, Oas2, Ifit2, Irf7, Ifnb1 mRNA level and lincRNA‐EPS transcripts in the WT and lincRNA‐EPS ^−/− PMs infected with VSV (MOI 1) for 6 h and 10 h.

Data information: Data of (C, D) are shown as the mean ± s.d. from three independent experiments, *P < 0.05 and **P < 0.01 by unpaired Student’s t‐test.

Figure EV3. Greater induction of antiviral ISGs in the lincRNA‐EPS ^−/− macrophages

1. Heatmap analysis of differential genes expression in top three pathways of GSEA.

2. Enrichment plot of Interferon alpha response was extracted from GSEA.

3. Control (sgCtrl) and lincRNA‐EPS knockdown (sglincRNA‐EPS) RAW264.7 cells were transfected with 1 μg/ml polyI:C and polydA:dT for 10 h, the transcripts of lincRNA‐EPS and Mx1, Oas2, Ifit2, and Irf7 were measured by RT–qPCR.

Data information: Data of (C) are shown as the mean ± s.d. from three independent experiments, *P < 0.05 and **P < 0.01 by unpaired Student’s t‐test.

Figure EV4. Deficiency of lincRNA‐EPS enhanced ISGs expression in primary macrophages and infected mice model

1. A

   RT–qPCR analysis of lincRNA‐EPS transcripts, viral RNA and Ifnb1, Mx1, Oas2, Ifit2, Irf7 mRNA level in the WT and lincRNA‐EPS ^−/− BMDMs infected with VSV (MOI 5) for 10 h.

2. B, C

   Eight weeks female lincRNA‐EPS ^−/− mice (n = 5) and WT littermates (n = 5) were injected (i.v.) with VSV (sub‐lethal dose, 6 × 10⁷ pfu/g) for 12 h, and negative control groups were injected with PBS (n = 3). RT–qPCR analysis of lincRNA‐EPS transcripts and Mx1, Oas2, Ifit2, Ifnb1 mRNA level from liver (B) and lung (C) tissue homogenate.

Data information: Data of (A) are shown as the mean ± s.d. from three independent experiments, data of (B, C) are shown as the mean ± s.d. of a typical representative result from three independent experiments, and one dot represents a mouse. *P < 0.05 and **P < 0.01 by unpaired Student’s t‐test.

Figure 5. Stronger activation of IFNAR downstream signaling independent of IFN‐β induction in the lincRNA‐EPS ^−/− macrophages

1. A, B

   RT–qPCR analysis of Ifnb1 mRNA level in the WT and lincRNA‐EPS ^−/− iBMMs infected with VSV (MOI 0.1) for 6 h (A), ELISA analysis of supernatant IFN‐β protein after infecting for 8 h (B).

2. C, D

   Mx1 (C) and Oas2 (D) mRNA level were measured in the control (EV) and lincRNA‐EPS‐rescued (lincRNA‐EPS) iBMMs after infecting with VSV with different titer for 6 h.

3. E

   LincRNA‐EPS‐stably overexpressed iBMMs (lincRNA‐EPS) and corresponding control group (EV) were infected with VSV for different titer (MOI 0.1 and MOI 0.5), Mx1 and Oas2 mRNA level were measured by RT–qPCR after 6 h.

4. F

   The basal level of ISGs and proinflammatory genes in the WT and lincRNA‐EPS ^−/− BMDMs (Datasets from the ArrayExpress database under the accession number E‐MTAB‐4088) were listed by heatmap.

5. G

   ChIP assays using anti‐H3K4me3 antibody was carried out in WT and lincRNA‐EPS ^−/− iBMMs with or without polyI:C (1 μg/ml) transfection for 4 h, Mx1 and Oas2 enrichment relative to input were measured by qPCR.

6. H, I

   WT and lincRNA‐EPS ^−/− iBMMs were infected with VSV (MOI 0.1) for 6 or 10 h (H) or transfected with polyI:C (1 μg/ml) for 4 or 8 h (I), phosphorylation and total protein expression were analyzed by Western blot with α‐tubulin and β‐Actin as loading controls.

Data information: data of (A–E) and (G) are shown as the mean ± s.d. from three independent experiments, *P < 0.05 and **P < 0.01 by unpaired Student’s t‐test. Data of (H, I) are representative results from three independent experiments.

Figure EV5. LincRNA‐EPS inhibits PKR‐STAT1‐dependent induction of ISGs

1. A

   RT–qPCR analysis of Ifnb1 mRNA level of polyI:C (1 μg/ml) transfected iBMMs for 6 h.

2. B

   20 nM siRNA for negative control (NC), PKR and STAT1 were transfected into iBMMs for 48 h. STAT1 and PKR protein level relative to naïve cells (Ctrl) were detected by Western blot and α‐tubulin was shown as a loading control.

3. C

   Empty vector (pMSCV‐EV) and pMSCV‐PIG‐lincRNA‐EPS (pMSCV‐EPS) were transfected into PKR‐overexpressed MEF cell lines (pBABE‐PKR) and control cells (pBABE‐EV), phosphorylation and total levels of STAT1 and PKR were detected by Western blot and β‐Actin was shown as a loading control.

4. D, E

   20 nM siRNA for NC (siNC) and PKR (siPKR#2) were transfected into WT and lincRNA‐EPS ^−/− iBMMs, Mx1, Ifit2, Isg15 and Pkr mRNA level were measured by RT–qPCR (D), and relative basal expression change rates of the antiviral ISGs in lincRNA‐EPS ^−/− iBMMs were calculated compared with WT group (E).

5. F

   20 nM siRNA for NC and PKR were transfected into WT and lincRNA‐EPS ^−/− iBMMs following transfected with 1 μg/ml polyI:C for 6 h, Mx1, Ifit2, Isg15 mRNA level were measured by RT–qPCR. Relative expression change rates of ISGs in lincRNA‐EPS ^−/− iBMMs were calculated compared with WT group.

6. G

   20 nM siRNA for NC and STAT1 were transfected into WT and lincRNA‐EPS ^−/− iBMMs, Mx1, Oas2 mRNA level were measured by RT–qPCR. Relative expression change rates of ISGs in lincRNA‐EPS ^−/− iBMMs were calculated compared with WT group.

Data information: Data of (A, and D–G) are shown as the mean ± s.d. from three independent experiments, *P < 0.05 and **P < 0.01 by unpaired Student’s t‐test. Data of (B–C) are representative images from three independent experiments.

Figure 6. LincRNA‐EPS and viral RNA competitively interact with PKR

1. A

   LincRNA‐EPS and negative control RNA (NC) were transcribed in vitro to label desthiobiotin and pulldown proteins from iBMMs lysate, the differential protein bands from silver‐stained SDS‐PAGE were cut for mass spectrometry analysis. One of the identified proteins from the band in the red box was PKR.

2. B

   Western blot for verification of RNA pulldown‐MS results with β‐Actin as a negative loading control.

3. C

   The interaction between PKR and lincRNA‐EPS were checked by RIP–qPCR using anti‐PKR antibody, and the efficiency of IP was verified by Western blot.

4. D

   20 nM siRNA targeting negative control (NC) and PKR were transfected into iBMMs for 48 h, the mRNA and protein level of PKR were tested by RT–qPCR and Western blot, respectively.

5. E, F

   VSV titer in the WT and lincRNA‐EPS ^−/− iBMMs transfected with siRNA as described in (D) and infected with VSV (MOI 0.1) for 8 h was measured by plaque assay (E). The alteration rates of viral titers were calculated in lincRNA‐EPS ^−/− iBMMs relative to WT group (F).

6. G

   VSV‐N and VSV‐G mRNA were transcribed from virus particles for RNA‐pulldown, the interactions between VSV RNA with PKR protein was checked by Western blot, β‐Actin was shown as a negative control.

7. H

   LincRNA‐EPS was transcribed in vitro and pulled down proteins from iBMMs infected with VSV for 4 h, PKR protein was detected by Western blot and GAPDH was shown as a loading control.

8. I

   RNA‐pulldown using biotinylated lincRNA‐EPS from cell lysates of polyI:C (1 and 2.5 μg/ml) transfected iBMMs for 4 h, PKR protein was detected by Western blot and β‐Actin was shown as a loading control.

9. J

   HEK293T cells transfected with lincRNA‐EPS and Flag‐tagged PKR were infected with VSV to detect the PKR‐binding lincRNA‐EPS and VSV RNA by RIP–qPCR using anti‐Flag antibody.

Data information: Data of (C, up panel), (D, left panel), (E, F) and (J) are shown as the mean ± s.d. from three independent experiments, *P < 0.05 and **P < 0.01 by unpaired student t‐test. Data of (B), (C, down panel), (D, right panel), (G, H, I) are representative images from three independent experiments.

Figure 7. LincRNA‐EPS inhibits PKR‐STAT1‐dependent induction of ISGs

1. A

   20 nM siRNA for negative control (NC) and PKR were transfected into iBMMs following infected with VSV (MOI 0.1) for 4 h or 8 h, phosphorylation and total levels of STAT1, eIF2α and PKR were detected by Western blot and β‐Actin was shown as a loading control.

2. B

   PKR was transfected into A549 cells for 24 h following stimulated with IFN‐β (50 ng/ml) for 1 h to test STAT1 phosphorylation by Western blot and β‐Actin was shown as a loading control.

3. C–H

   20 nM siRNA for NC and PKR were transfected into WT and lincRNA‐EPS ^−/− iBMMs following infection with VSV (MOI 0.1) for 4 h, Mx1, Ifit2, and Isg15 mRNA level were measured by RT–qPCR (C, E, and G), and relative basal expression change rates of the antiviral ISGs in lincRNA‐EPS ^−/− iBMMs were calculated compared with WT group (D, F, and H).

4. I

   Working model of lincRNA‐EPS in resting and viral infected WT or lincRNA‐EPS ^−/− cells.

Data information: Data of (A, B) are representative images from three independent experiments. Data of (C‐H) are shown as the mean ± s.d. from three independent experiments, *P < 0.05 and **P < 0.01 by unpaired student t‐test.

Figure EV6. Downregulated ISGs after interfering PKR expression

20 nM siRNAs for negative control (siNC) and PKR (siPKR) were transfected into iBMMs for 48 h, and the total RNA was isolated to perform RNA sequencing. The mean RPKM values of the biological duplicates were calculated and the downregulated antiviral ISGs after knocking down PKR were listed by heatmap (left panel). The fold change of ISGs level of lincRNA‐EPS ^−/− iBMMs comparing to the WT iBMMs in siNC and siPKR groups were also illustrated by heatmap (right panel).