Loss of TRIM29 mitigates viral myocarditis by attenuating PERK-driven ER stress response in male mice

Abstract

Viral myocarditis, an inflammatory disease of the myocardium, is a significant cause of sudden death in children and young adults. The current coronavirus disease 19 pandemic emphasizes the need to understand the pathogenesis mechanisms and potential treatment strategies for viral myocarditis. Here, we found that TRIM29 was highly induced by cardiotropic viruses and promoted protein kinase RNA-like endoplasmic reticulum kinase (PERK)-mediated endoplasmic reticulum (ER) stress, apoptosis, and reactive oxygen species (ROS) responses that promote viral replication in cardiomyocytes in vitro. TRIM29 deficiency protected mice from viral myocarditis by promoting cardiac antiviral functions and reducing PERK-mediated inflammation and immunosuppressive monocytic myeloid-derived suppressor cells (mMDSC) in vivo. Mechanistically, TRIM29 interacted with PERK to promote SUMOylation of PERK to maintain its stability, thereby promoting PERK-mediated signaling pathways. Finally, we demonstrated that the PERK inhibitor GSK2656157 mitigated viral myocarditis by disrupting the TRIM29-PERK connection, thereby bolstering cardiac function, enhancing cardiac antiviral responses, and curbing inflammation and immunosuppressive mMDSC in vivo. Our findings offer insight into how cardiotropic viruses exploit TRIM29-regulated PERK signaling pathways to instigate viral myocarditis, suggesting that targeting the TRIM29-PERK axis could mitigate disease severity.

Fig. 1. TRIM29 knockout inhibits PERK-mediated ER stress and apoptosis induced by cardiotropic viruses in mouse cardiomyocytes.

a Immunohistochemistry (IHC) analysis of TRIM29 and ER marker KDEL expression in mouse heart tissues from wild-type mice without or with CVB3 infection for 2 days. Scale bars represent 100 μm. b Immunoblot analysis of PERK, IRE1α, ATF6 and TRIM29 levels in mouse primary neonatal cardiomyocytes from Trim29^+/+ and Trim29^–/– mice by infection without or with CVB3 and EMCV for 6 h at an MOI of 1. RT-qPCR analysis of Atf4 (c), Chop (d), Bim (e), Noxa (f) and Puma (g) at the mRNA level in mouse primary neonatal cardiomyocytes from Trim29^+/+ and Trim29^–/– mice by infection without or with CVB3 and EMCV for 3 h at an MOI of 1. h Immunoblot analysis of CHOP, cleaved caspase-3, BCL-2, BAX and ANP levels in mouse primary neonatal cardiomyocytes from Trim29^+/+ and Trim29^–/– mice by infection without or with CVB3 and EMCV for 6 h at an MOI of 1. i Cell viability quantification analysis of mouse primary neonatal cardiomyocytes from Trim29^+/+ and Trim29^–/– mice by infection without or with CVB3 and EMCV for 12 h at an MOI of 1 using CellTiter-Glo assay. Mock, mouse neonatal cardiomyocytes without infection. Data are shown as the mean ± SD. Statistical significance was determined by a two-tailed, unpaired Student’s t test. NS, not significant. Data are representative of three independent experiments. Source data are provided as a Source Data file.

Fig. 2. TRIM29 knockout unleashes type I interferon production to restrict cardiotropic viruses by relieving ROS-mediated TBK1 inhibition in mouse cardiomyocytes.

a ROS production analysis of mouse primary neonatal cardiomyocytes from Trim29^+/+ and Trim29^–/– mice by infection without or with CVB3 and EMCV for 6 h at an MOI of 1 using a DCFDA cellular ROS assay kit. The fluorescence intensity of the mock group was defined as 100%. b Immunoblot (IB) analysis of TBK1 and FITC precipitated with anti-TBK1 from whole-cell lysates incubated with 5 μM 5-IAF for 1 h labeling of free thiols in mouse neonatal cardiomyocytes from Trim29^+/+ and Trim29^–/– mice by infection without or with CVB3 and EMCV for 6 h at an MOI of 5. c Immunoblot analysis of IRF3, TBK1 and native TBK1 monomer and dimer in mouse neonatal cardiomyocytes from Trim29^+/+ and Trim29^–/– mice by infection without or with CVB3 and EMCV for 6 h at an MOI of 5. ELISA of IFN-α (d) and IFN-β (e) production by mouse neonatal cardiomyocytes from Trim29^+/+ and Trim29^–/– mice by infection without or with CVB3 and EMCV for 16 h at an MOI of 5. Each circle represents one value of the three biological replicates; small horizontal lines indicate the average of triplicates. f Quantification of the expression of CVB3 and EMCV viral genome copies relative to GAPDH in mouse neonatal cardiomyocytes from Trim29^+/+ and Trim29^–/– mice by infection without or with CVB3 and EMCV for 6 h at an MOI of 1. Mock, mouse neonatal cardiomyocytes without infection. Data are shown as the mean ± SD. Statistical significance was determined by a two-tailed, unpaired Student’s t test. NS, not significant. Data are representative of three independent experiments. Source data are provided as a Source Data file.

Fig. 3. Cardiomyocyte-specific TRIM29 deficiency protects mice from viral myocarditis in vivo.

a Survival of Trim29^fl/fl and cardiomyocyte-specific TRIM29 knockout (Trim29^MyHC-KO) mice after intraperitoneal infection with CVB3 (1 × 10⁷ PFU per mouse) (n = 10 per group). b Hematoxylin and eosin (H&E)-staining of heart sections from Trim29^fl/fl and Trim29^MyHC-KO mice after intraperitoneal infection without (Mock) or with CVB3 (1 × 10⁷ PFU per mouse) for 4 days. Scale bars represent 1000 μm for original images and 400 µm for enlarged images. c Representative M-mode echocardiography images of hearts from Trim29^fl/fl and Trim29^MyHC-KO mice on day 4 after CVB3 infection. Cardiac function analysis of ejection fraction (EF) (d) and fractional shortening (FS) (e) of hearts from mice as in (c) (n = 5 per group). f Assessment of heart weight/baseline body weight in Trim29^fl/fl and Trim29^MyHC-KO mice (n  =  5 per group) on day 0 or day 6 after CVB3 infection. ELISA of creatine kinase production in sera (g) and viral titers in homogenates of heart, pancreas, and spleen (h) from Trim29^fl/fl and Trim29^MyHC-KO mice on day 0 (Mock) and day 2 after CVB3 infection (n = 5 per group). ELISA of IFN-α (i), IFN-β (j) IL-6 (k), TNF-α (l) and IL-1β (m) in hearts from Trim29^fl/fl and Trim29^MyHC-KO mice on day 2 after CVB3 infection (n  =  5 per group). Data are shown as the mean ± SD. Statistical significance was determined by a two-tailed, unpaired Student’s t test and Gehan-Breslow-Wilcoxon test for survival analysis. NS, not significant. Data are representative of three independent experiments. Source data are provided as a Source Data file.

Fig. 4. TRIM29 deficiency reduces PERK-mediated immunosuppressive mMDSC to enhance functions of antiviral CD8 T cells during viral myocarditis in vivo.

Flow cytometry (a, c) and quantification (b, d) analysis of mouse MDSC (CD11b⁺Gr1⁺) cells of heart infiltrated immune cells (a, b) and spleen immune cells (c, d) from Trim29^fl/fl and Trim29^MyHC-KO mice infected without or with CVB3 for 2 days using CD11b-BV785 and Gr1-APC antibodies. Flow cytometry analysis of mouse mMDSC in gated MDSC (CD11b⁺Gr1⁺) from heart (e) and spleen (f) of CVB3 infected Trim29^fl/fl mice as in (a), (c) using Ly6C-PE and Ly6G-FITC antibodies. Flow cytometry (g) and quantification (h) analysis of proliferating CD8 T cells cocultured with mouse MDSC (1:1/4 ratio) isolated from heart of Trim29^fl/fl and Trim29^MHC-KO mice with a 3-day infection of CVB3 infection for 0 (blue line) and 3 days (red line) using carboxyfluroescein succinimidyl ester (CFSE) staining. Flow cytometry (i, k) and quantification (j, l) analysis of mouse IFN-γ producing CD8 T cells (IFN-γ⁺ CD8 T cells) of heart infiltrated lymphocytes (i, j) and spleen lymphocytes (k, l) from Trim29^fl/fl and Trim29^MyHC-KO mice infected without or with CVB3 for 2 days followed by PMA/ionomycin stimulation for 4 h using CD8-PerCP and IFN-γ-APC antibodies. Flow cytometry data were acquired on an LSR-II flow cytometer (Beckton Dickinson) and analyzed using FlowJo v10 software (Tree Star). Data are shown as the mean ± SD. Statistical significance was determined by a two-tailed, unpaired Student’s t test. NS, not significant. Data are representative of three independent experiments. Source data are provided as a Source Data file.

Fig. 5. TRIM29 interacts with PERK to promote PERK-mediated ER stress and apoptosis.

a Immunoblot (IB) analysis of endogenous proteins TRIM29 and PERK precipitated with anti-PERK or immunoglobulin G (IgG) from whole-cell lysates of mouse primary neonatal cardiomyocytes from Trim29^fl/fl and WTMyHC-Cre mice with or without (Mock) CVB3 and EMCV infection for 6 h at an MOI of 5. Schematic diagram showing full-length PERK (Full, b) or TRIM29 (Full, c) and serial truncations of PERK (b) or TRIM29 (c) with deletion (Δ) of various domains (left margin); numbers at ends indicate amino acid positions (top). Luminal, ER luminal domain; TM, transmembrane domain; Kinase, protein kinase domain; BBOX, the B-box zinc-finger domain; CC, the coil-coil domain; OmpH, the outer membrane protein H domain. Immunoblot analysis of purified HA-tagged full-length PERK and serial truncations of PERK with deletion of various domains alone (d) or purified HA-tagged full-length TRIM29 and serial truncations of TRIM29 with deletion of various domains alone (e) with anti-HA antibody (top blot) or after incubation with Myc-tagged TRIM29 (d) or Myc-tagged PERK (e) and immunoprecipitation with anti-Myc antibody (second blot), and immunoblotting analysis of purified Myc-tagged TRIM29 (d) or Myc-tagged PERK (e) with anti-Myc antibody (third blot) or after incubation with Myc-tagged TRIM29 (d) or Myc-tagged PERK (e) and immunoprecipitation with anti-Myc antibody (bottom blot). f Immunoblot analysis of endogenous phosphorylated PERK (pPERK), HA-tagged PERK and Myc-tagged TRIM29 from whole-cell lysates of HEK293T cells co-transfected with HA-PERK and vector or Myc-TRIM29 for 24 h. g Immunoblot analysis of endogenous proteins TRIM29 and phosphorylated PERK (pPERK) precipitated with anti-pPERK or immunoglobulin G (IgG) from whole-cell lysates of mouse primary neonatal cardiomyocytes from Trim29^fl/fl mice followed by CVB3 infection at an MOI of 5 and treatment with the PERK inhibitor GSK2656157 or DMSO for 6 h. The position of protein markers (shown in kDa) is indicated on the right. Data are representative of three independent experiments. Source data are provided as a Source Data file.

Fig. 6. TRIM29 promotes SUMOylation of PERK to maintain its stability.

a Immunoblot (IB) analysis of HA-TRIM29 and Flag-Ubc9 precipitated with anti-HA or immunoglobulin G (IgG) from whole-cell lysates of HEK293T cells transfected with Flag-Ubc9 and HA-TRIM29 Full, its mutants HA-TRIM29 N or C. b Immunoblot analysis of TRIM29-mediated PERK SUMOylation with SUMO1, SUMO2 or SUMO3 from whole-cell lysates of HEK293T cells transfected with the indicated plasmids for 24 h followed by Ni²⁺-NTA agarose affinity pull-down assay using anti-HA, anti-Myc, anti-Flag and anti-His antibodies. c Immunoblot analysis of SUMO1-mediated PERK SUMOylation by TRIM29 Full or its mutant N from whole-cell lysates of HEK293T cells transfected with the indicated plasmids for 24 h followed by Ni²⁺-NTA agarose affinity pull-down assay using anti-HA, anti-Myc, anti-Flag and anti-His antibodies. d Immunoblot analysis of endogenous SUMO1-mediated PERK SUMOylation by TRIM29 precipitated with anti-PERK from whole-cell lysates of mouse primary neonatal cardiomyocytes from Trim29^fl/fl, WTMyHC-Cre and Trim29^MyHC-KO mice without (Mock) or with CVB3 infection at an MOI of 5 using anti-PERK, anti-TRIM29, anti-SUMO1 and anti-SUMO2/3 antibodies. e Immunoblot analysis of TRIM29-mediated SUMOylation of wild-type (WT) PERK or its mutants, including K637R, K641R, K672R, K937R and K641/672R, from whole-cell lysates of HEK293T cells transfected with the indicated plasmids for 24 h followed by Ni²⁺-NTA agarose affinity pull-down assay using anti-HA, anti-Myc, anti-Flag and anti-His antibodies. f Immunoblot analysis of endogenous SUMO1-mediated PERK SUMOylation by TRIM29 precipitated with anti-PERK from whole-cell lysates of mouse primary neonatal cardiomyocytes from Trim29^fl/fl mice without (Mock) or with CVB3 infection at an MOI of 5 and treatment with the PERK inhibitor GSK2656157 or DMSO using anti-PERK, anti-TRIM29 and anti-SUMO1 antibodies. The position of protein markers (shown in kDa) is indicated on the right. Data are representative of three independent experiments. Source data are provided as a Source Data file.

Fig. 7. A PERK inhibitor ameliorates viral myocarditis in vivo.

a Schematic illustration of the animal experiment. b Survival of wild-type (WT) mice after intraperitoneal infection with CVB3 (1 × 10⁷ PFU per mouse) and treatment with the PERK inhibitor GSK2656157 or DMSO (n = 10 per group). c Hematoxylin and eosin (H&E)-staining of heart sections from WT mice after intraperitoneal infection without (Mock) or with CVB3 (1 × 10⁷ PFU per mouse) and treatment with the PERK inhibitor GSK2656157 or DMSO for 4 days. Scale bars represent 1000 μm for original images and 400 µm for enlarged images. d Representative M-mode images of hearts from WT mice after intraperitoneal infection without (Mock) or with CVB3 and treatment with the PERK inhibitor GSK2656157 or DMSO for 4 days by echocardiography analysis. Cardiac function analysis of ejection fraction (EF) (e) and fractional shortening (FS) (f) of hearts from mice as in (d) (n = 5 per group). g Assessment of heart weight/baseline body weight in WT mice (n  =  5 per group) on day 0 or day 6 after CVB3 infection and treatment with the PERK inhibitor GSK2656157 or DMSO. ELISA of creatine kinase production in sera (h) and viral titers in homogenates of hearts (i) from WT mice on day 0 (Mock), day 2 and day 4 after CVB3 infection and treatment with the PERK inhibitor GSK2656157 or DMSO (n = 5 per group). ELISA of IFN-α (j), IFN-β (k) IL-6 (l), TNF-α (m) and IL-1β (n) in hearts from WT mice on day 2 after CVB3 infection and treatment with the PERK inhibitor GSK2656157 or DMSO (n  =  5 per group). Data are shown as the mean ± SD. Statistical significance was determined by a two-tailed, unpaired Student’s t test and Gehan-Breslow-Wilcoxon test for survival analysis. NS, not significant. Data are representative of three independent experiments. Source data are provided as a Source Data file.

Fig. 8. Working model for TRIM29-PERK axis in controlling pathogenesis of viral myocarditis by regulating ER stress and ROS responses.

In WT cardiomyocytes after infection with cardiotropic viruses (CVB3), TRIM29 is strongly induced by cardiotropic viruses and interacts with PERK and induces its SUMO1-mediated SUMOylation for maintaining stability of PERK, thereby promoting ROS-mediated TBK1 oxidization to reduce TBK1-mediated type I IFN production and enhancing PERK ER stress-mediated apoptosis, proinflammatory cytokines, and immunosuppressive mMDSC, which act in combination to cause pathogenesis of viral myocarditis. In contrast, in TRIM29 KO cardiomyocytes or WT cardiomyocytes with GSK2656157 treatment, PERK could not maintain its stability without TRIM29 or with GSK2656157 treatment post infection of cardiotropic viruses, and loss its strong ability to promote ER stress and ROS responses, resulting in enhanced type I IFN and reduced apoptosis, proinflammatory cytokines and immunosuppressive mMDSC, which significantly reduce viral myocarditis.