Inhibition of viral pathogenesis and promotion of the septic shock response to bacterial infection by IRF-3 are regulated by the acetylation and phosphorylation of its coactivators

Abstract

Interferon (IFN) is required for protecting mice from viral pathogenesis; reciprocally, it mediates the deleterious septic shock response to bacterial infection. The critical transcription factor for IFN induction, in both cases, is IRF-3, which is activated by TLR3 or RIG-I signaling in response to virus infection and TLR4 signaling in response to bacterial infection. Here, we report that IRF-3's transcriptional activity required its coactivators, β-catenin and CBP, to be modified by HDAC6-mediated deacetylation and protein kinase C isozyme β (PKC-β)-mediated phosphorylation, respectively, so that activated nuclear IRF-3 could form a stable transcription initiation complex at the target gene promoters. β-Catenin bridges IRF-3 and CBP, and the modifications were required specifically for the interaction between β-catenin and CBP but not β-catenin and IRF-3. Consequently, like IRF-3(-/-) mice, HDAC6(-/-) mice were resistant to bacterial lipopolysaccharide-induced septic shock. Conversely, they were highly susceptible to pathogenesis caused by Sendai virus infection. Thus, HDAC6 is an essential component of the innate immune response to microbial infection.

FIG 1

PKC-β and HDAC6 are required for TLR-induced IRF-3 transcriptional activity. (A) P56/IFIT1 induction was analyzed in TLR3-stimulated HT1080 cells in the presence of a PKC-β inhibitor by Western blotting. (B) P56/IFIT1 induction was analyzed in TLR3 [poly(I⋅C), 100 µg/ml]-stimulated shCon and shHDAC6 cells by Western blotting. (C) P56/IFIT1 induction was analyzed in IFN-β (1,000 U/ml)-stimulated shCon or shHDAC6 cells by Western blotting. (D) P56/IFIT1 induction was analyzed in RLH [transfected-poly(I⋅C)]-stimulated shCon or shHDAC6 cells by Western blotting. (E) P54/Ifit2 induction was analyzed in RLH [transfected-poly(I⋅C)]-stimulated wt or HDAC6^−/− MEFs by Western blotting. KO, knockout. (F) Primary bone marrow-derived macrophages (BMDMs) from wt or Hdac6^−/− mice were treated with TLR4 (LPS, 1 µg/ml) or TLR3 [poly(I⋅C) (pI:C), 25 µg/ml] agonists, Ifit1 induction was analyzed by RT-PCR, and 18S rRNA was used as a loading control. (G) Splenocytes isolated from wt or Hdac6^−/− mice were treated with LPS (at the indicated concentrations) or poly(I⋅C) (50 µg/ml), and P54/Ifit2 induction was analyzed by Western blotting. (H) BMDMs were isolated from wt or Hdac6^−/− mice, and IL-6 (Il6) and TNF-α induction was analyzed by RT-PCR upon LPS treatment.

FIG 2

PKC-β and HDAC6 activities are required for promoter occupancy of IRF-3 but not its phosphorylation and nuclear translocation. (A) IRF-3 phosphorylation (P-Ser396) and P56/IFIT1 induction in TLR3-stimulated HT1080 cells in the absence or the presence of TSA. (B) Nuclear translocation of IRF-3 was analyzed in TLR3-stimulated shCon or shHDAC6 cells by Western blotting. (C) Nuclear translocation of IRF-3 was analyzed in PKC-β inhibitor- or Gӧ6976-treated, TLR3-stimulated HT1080 cells by Western blotting. (D, E) Chromatin immunoprecipitation (ChIP) of IRF-3 and RNA Pol II on the IFIT1 promoter was analyzed in TLR3-stimulated, Gӧ6976-treated HT1080 (D) and shHDAC6 (E) cells. (F) ChIP of STAT2 and RNA Pol II on the IFIT1 promoter was analyzed in IFN-β-stimulated shCon and shHDAC6 cells. n.s., nonspecific.

FIG 3

HDAC6-deficient cells fail to recruit complex containing IRF-3, CBP, and β-catenin. (A) Co-IP of IRF-3 with CBP in TLR3-stimulated HT1080 cells in the absence or the presence of TSA. The input represents the levels of IRF-3 in cell lysates. IB, immunoblotting. (B) Co-IP of IRF-3 with CBP in TLR3-stimulated shCon and shHDAC6 cells. The input represents the levels of IRF-3 in cell lysates. (C) HT1080 cells were transfected with nontargeting (NT) or β-catenin-specific siRNA, and induction of P56/IFIT1 was analyzed upon TLR3 stimulation by Western blotting. Knockdown of β-catenin was confirmed by Western blotting (middle panel). (D) β-Catenin acetylation (acetylated K49 [Ac-K⁴⁹]-specific antibody) was analyzed in shCon and shHDAC6 cells by Western blotting. (E) Co-IP of IRF-3 with CBP in nontargeting (NT) or β-catenin (β-cat)-specific, siRNA-transfected HT1080 cells upon TLR3 stimulation. The input represents levels of β-catenin in cell lysates. (F, G) Western analyses of β-catenin levels (F) and P56/IFIT1 induction (G) in TLR3-stimulated [poly(I⋅C)] HT1080 cells in the absence or the presence of Wnt signaling activators (as indicated). Actin was used as a loading control. SB, SB216763.

FIG 4

HDAC6-mediated deacetylated β-catenin is required for interaction with CBP but not IRF-3. (A) Flag-tagged β-catenin was expressed in shCon and shHDAC6 cells. IP of β-catenin with CBP was analyzed upon TLR3 stimulation. The input represents the levels of β-catenin in the cell lysates. (B) Flag-tagged β-catenin and V5-tagged IRF-3 were coexpressed in shCon and shHDAC6 cells. Co-IP of β-catenin with IRF-3 was analyzed upon TLR3 stimulation. The input represents the levels of β-catenin in the cell lysates. (C) Flag-tagged β-catenin was expressed in HT1080 cells. Co-IP of β-catenin with CBP was analyzed upon TLR3 stimulation in the absence or the presence of the PKC-β inhibitor (PKCβ-i). The input represents the levels of β-catenin in the cell lysates. (D) Flag-tagged β-catenin and V5-tagged IRF-3 were coexpressed in HT1080 cells. Co-IP of β-catenin with IRF-3 was analyzed upon TLR3 stimulation in the absence or the presence of the PKC-β inhibitor. The input represents the levels of β-catenin in the cell lysates. (E) Flag-tagged β-catenin (wt or the K49R mutant [KR]) was expressed in shHDAC6 cells. Co-IP of β-catenin with CBP was analyzed upon TLR3 stimulation. The input represents the levels of β-catenin in the cell lysates. (F) The wt or K49R mutant of β-catenin was expressed in shHDAC6 cells, and P56/IFIT1 induction was analyzed upon TLR3 stimulation [poly(I⋅C)]. (G) The Flag-tagged β-catenin mutant (K49R) was expressed in HT1080 cells. Co-IP of β-catenin with CBP was analyzed upon TLR3 stimulation in the absence or the presence of the PKC-β inhibitor. The input represents the levels of β-catenin in the cell lysates.

FIG 5

HDAC6-mediated transcriptional activation of IRF-3 contributes to LPS-induced septic shock and protection against Sendai virus infection in mice. Shown are percentages of survival of IRF3^−/−, TRIF^−/−, and wt mice (C57BL/6) after intraperitoneal injection with LPS (1 mg/mouse) (A), of IRF3^−/− and wt mice (C57BL/6) after intranasal infection with SeV (strain 52, 1,000 PFU/mouse) (B), of HDAC6^−/− and wt mice (129s) after intraperitoneal injection with LPS (1 mg/mouse) (C), and of HDAC6^−/− and wt mice (129s) after intranasal infection with SeV (strain 52, 1,000 PFU/mouse). The statistical significance of survival differences between the knockout mice and their wt controls are indicated as P values. pi, postinfection.