TRIM21enhances IL-17A signaling and drives autoimmune myocarditis by promoting TRAF3 lysosomaldegradation in cardiac fibroblasts

Significance

Myocarditis is a potentially fatal disease that can progress to inflammatory dilated cardiomyopathy (DCMi) and heart failure, yet current therapies do not target the transition to chronic cardiac injury. Understanding the mechanisms driving this progression is critical for developing new interventions. In this study, we identify TRIM21 as a key regulator in cardiac fibroblasts (CFs), amplifying IL-17A signaling through TRAF3 degradation, forming a self-reinforcing positive feedback loop. This mechanism underscores how CFs contribute to sustaining inflammation and fibrosis. Our findings suggest TRIM21 as a potential therapeutic target to prevent the progression from myocarditis to DCMi and heart failure.

Abstract

Myocarditis is a potentially fatal disease that can progress to inflammatory dilated cardiomyopathy (DCMi), a condition that is a leading cause of noncongenital heart failure in young adults. However, the mechanisms driving the transition from acute myocarditis to DCMi remain poorly understood. Although IL-17A-mediated inflammation is a well-known driver of this progression, the intracellular regulators that sustain and amplify this response remain unclear. Here, we identify the E3 ubiquitin ligase TRIM21, which is upregulated in the hearts of DCMi patients and experimental autoimmune myocarditis (EAM) mice, as a key amplifier of IL-17A-driven inflammation in cardiac fibroblasts (CFs), promoting disease progression and fibrosis. TRIM21 deficiency significantly attenuates EAM progression, reducing inflammatory cell infiltration and fibrotic remodeling. Mechanistically, TRIM21 is induced by IL-17A in CFs and promotes K27-linked polyubiquitination and lysosomal degradation of TRAF3, a negative regulator of IL-17A signaling, thereby establishing a self-reinforcing IL-17A-TRIM21–TRAF3 positive feedback loop. Finally, specific knockdown of TRIM21 in CFs via the AAV9-Postn strategy effectively alleviates EAM severity. These findings identify TRIM21 as an intracellular mediator of sustained IL-17A signaling in myocarditis and highlight its potential as a therapeutic target for preventing progression to DCMi.

Inflammatory dilated cardiomyopathy (DCMi) is a major cause of noncongenital heart failure in young individuals, often progressing to end-stage disease requiring cardiac transplantation. However, the molecular and cellular mechanisms that drive the transition from immune-mediated myocardial injury to irreversible remodeling remain poorly understood (1, 2). Autoimmunity is frequently implicated in this process, and the experimental autoimmune myocarditis (EAM) model recapitulates key features of disease progression. In this model, BALB/c mice immunized with cardiac myosin peptide exhibit myocardial inflammation by day 14, peaking at day 21, followed by fibrotic remodeling and ventricular dilation between days 35 and 60 (3), providing a valuable platform to dissect the immune–fibrotic transition.

Recent studies have established a critical role for Th17 cells and their signature cytokine interleukin-17A (IL-17A) in promoting myocardial inflammation and fibrotic progression in human and murine myocarditis (4–6). In many inflammatory-fibrotic conditions, IL-17A drives robust early chemokine/cytokine induction and neutrophil recruitment, and later contributes to fibrosis via fibroblast activation, proliferation, and matrix deposition (7–10). In the heart, cardiac fibroblasts (CFs) are major IL-17A-responsive stromal cells that produce inflammatory chemokines, sustain leukocyte recruitment, and thereby couple acute inflammation to later remodeling (11). Although IL-17A may be dispensable for disease initiation (12), multiple studies support its contribution to fibrotic progression, consistent with an early inflammation and late fibrosis paradigm. Despite this framework, the intracellular pathways that enable CFs to sustain or amplify IL-17A-driven responses remain poorly defined, and clarifying these mechanisms is important for explaining disease progression and for developing fibroblast-targeted therapies (10, 11).

TRIM21, originally identified as the autoantigen Ro52/SS-A, is a cytosolic E3 ubiquitin ligase widely expressed across tissues and implicated in the regulation of autoimmunity (13). Its immunosuppressive roles have been primarily studied in hematopoietic cells. Trim21-deficient mice develop signs of systemic lupus, with enhanced proinflammatory cytokine production in Trim21-deficient splenocytes and macrophages through the ubiquitination of interferon regulatory factors (IRFs) 3, 5, and 8 (14). In human inflammatory bowel disease, TRIM21 is enriched in CD4⁺ T cells and restrains their differentiation into Th1 and Th17 cells (15). Similarly, TRIM21 deficiency promotes atherosclerosis by enhancing Th17 responses (16). Despite its established role in immune cells, the contribution of TRIM21 to inflammatory and fibrotic processes in nonhematopoietic cells remains poorly defined.

Here, we identify TRIM21 as a key amplifier of IL-17A signaling in CFs during autoimmune myocarditis. TRIM21 is upregulated in the hearts of DCMi patients and EAM mice, with its expression induced by IL-17A, forming a positive feedback loop that sustains inflammation. TRIM21 deficiency reduces cardiac inflammation, immune cell infiltration, and fibrosis. Mechanistically, TRIM21 promotes K27-linked polyubiquitination and lysosomal degradation of TRAF3, a negative regulator of IL-17 signaling. These findings suggest TRIM21 as a potential therapeutic target for preventing the transition from myocarditis to heart failure.

Results

TRIM21 Is Significantly Upregulated in the Hearts of DCM Patients and EAM Mice.

Our previous research demonstrated that Trim21 knockout (KO) mice were more susceptible to viral myocarditis (VMC) due to suppression of CVB3 replication (17). To explore the role of TRIM21 in later stages, including autoimmune myocarditis and cardiac fibrosis, we first examined its expression in heart tissues from patients with DCM and other related cardiovascular diseases using publicly available datasets. In the Gene Expression Omnibus Series, two DCM (GSE141910 and GSE120895) and one chronic chagasic cardiomyopathy (CCC) (GSE191081) studies showed elevated TRIM21 expression in patients compared to healthy subjects (Fig. 1A), indicating a potential role in myocarditis and pathological remodeling.

Fig. 1.

TRIM21 is significantly upregulated in hearts of human DCM and EAM mice. (A) TRIM21 mRNA levels in human cardiac tissue were analyzed in three datasets, including nonfailing donors (NFD; n = 166) and DCM patients (n = 166) from GSE141910, donors with normal left ventricular ejection fraction (N-LVEF; n = 8) and DCM patients (n = 47) from GSE120895, and healthy controls (HC; n = 6) and chronic Chagasic cardiomyopathy (CCC; n = 8) from GSE191081. (B) Trim21 expression in cardiac tissues from EAM mice at the indicated time points (Day 0: n = 8; Day 14: n = 6; Day 21: n = 7; Day 35: n = 8) was detected by qPCR. (C) TRIM21 protein levels in cardiac tissues from EAM mice at the indicated time points detected by western blot (n = 3). (D) Representative images of H&E staining with inflammation scoring (Top), Sirius Red staining with fibrosis score (Middle), and TRIM21 IHC with positive area score (Bottom) in EAM hearts at the indicated time points (n = 6 per time point). Error bars represent group means ± SD (A) or group median (B and D Right). Statistical analysis: Data in (A) were analyzed by unpaired two-tailed Student’s t test. Data in (B and D Right) were analyzed by one-way ANOVA followed by Tukey’s post hoc test. Significance: *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; ns, not significant.

To investigate whether TRIM21 is similarly induced during myocarditis and cardiac fibrosis in the mouse model, we next assessed TRIM21 expression throughout EAM progression. Both mRNA and protein levels of cardiac TRIM21 were significantly elevated by day 14 and increased further by day 35 (Fig. 1 B and C and SI Appendix, Fig. S1A). Histopathological analysis indicated a biphasic course, in which inflammatory infiltration (H&E) begins at day 14 and peaks at day 21, whereas collagen deposition (Sirius Red-positive) begins at day 21 and intensifies by day 35, consistent with transition toward a remodeling fibrotic phase. Immunohistochemistry (IHC) confirmed a progressive increase in TRIM21 throughout disease progression (Fig. 1D and SI Appendix, Fig. S1B). Taken together, these results demonstrate that cardiac TRIM21 increases during EAM progression, suggesting its potential role in EAM pathogenesis.

TRIM21 Deficiency Protects Mice from EAM.

To investigate the role of TRIM21 in EAM progression, we induced EAM in Trim21 deficient (Trim21 KO) and wild-type (WT) mice immunized with MyHCα_614-629 peptide. By day 21, WT-EAM hearts were enlarged and discolored with an increased heart weight-to-body weight (HW/BW) ratio versus WT-Ctrl. In contrast, Trim21 KO-EAM hearts were smaller, displayed fewer white lesions, and showed a lower HW/BW ratio (SI Appendix, Fig. S2 A and B). Consistently, echocardiography at day 21 showed slightly reduced ejection fraction (EF) and fractional shortening (FS) in WT-EAM mice, indicating incipient ventricular dilation, whereas Trim21 KO-EAM mice preserved systolic function (Fig. 2A). Meanwhile, serum CK-MB, a biomarker of myocardial injury (12), was significantly lower in Trim21 KO-EAM mice, indicating less cardiac damage (Fig. 2B). Histologically, H&E demonstrated greater inflammatory infiltration, and Sirius Red revealed a larger collagen-positive fibrotic area in WT-EAM hearts, whereas Trim21 KO-EAM hearts showed significantly less inflammation and minimal fibrosis (Fig. 2C). Consistently, α-SMA⁺ activated fibroblasts were unchanged at baseline (WT-Ctrl vs. Trim21 KO-Ctrl), increased during EAM, and were significantly decreased in Trim21 KO-EAM compared to WT-EAM (SI Appendix, Fig. S2C). Collectively, Trim21 deficiency reduces inflammation and fibrosis during EAM.

Fig. 2.

TRIM21 deficiency protects against EAM in mice. (A–D) Six-week-old WT and Trim21 KO mice were immunized with MyHCα_614-629 peptide emulsified in CFA to induce EAM and were evaluated by day 21 (n = 6 per group). (A) Representative echocardiography images with EF and FS. (B) Serum CK-MB activity. (C) Representative gross cardiac morphology, H&E staining with inflammation scores, and Sirius Red staining with fibrosis scores. (D) qPCR of Tnfα, Il6, Cxcl1, Cxcl5, and Cxcl10 in heart tissues and Gapdh normalization. (E) Heart-infiltrating immune cell populations (n = 6 per group). Estimated number and percentages of CD45⁺ cells in live single cells. Percentages of CD45⁺CD11b⁺Ly6G⁺ neutrophils and CD45⁺CD11b⁺Ly6C⁺ Mo/Mφ in CD45⁺ cells. Percentage of CD45⁺CD11b⁺Ly6C^high Mo/Mφ in CD45⁺ CD11b⁺Ly6C⁺ Mo/Mφ cells. Error bars represent group means ± SD, and each data point represents an individual mouse. Statistical analysis: Data in (A Right, B and C Right and D) were analyzed by two-way ANOVA with Sidak multiple comparisons test. Data in (E) were analyzed using unpaired two-tailed Student’s t test. Significance: *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; ns, not significant.

To further investigate pathways affected by Trim21 deficiency, we performed bulk RNA-seq on WT-Ctrl, WT-EAM, and KO-EAM hearts (day 21). Compared to WT-Ctrl, 541 genes were altered in WT-EAM mice and 669 in Trim21 KO-EAM mice, with 175 shared (SI Appendix, Fig. S3 A and B). Functional enrichment of DEGs between WT-EAM and Trim21 KO-EAM highlighted cytokine-mediated signaling and regulation of inflammatory responses (SI Appendix, Fig. S3 C–E), indicating that TRIM21 promotes EAM progression primarily by modulating immune signaling. qPCR analysis validated this, showing reduced Tnfα, Il6, Cxcl1, Cxcl5, and Cxcl10 in Trim21 KO hearts versus WT (Fig. 2D). In addition, we reanalyzed independent RNA-seq data from WT-Ctrl versus Trim21 KO-Ctrl hearts and identified 95 DEGs. Apart from Lcn2, few inflammation-related transcripts were detected, supporting that the EAM phenotypes reflect disease-context amplification by TRIM21 rather than preexisting differences (SI Appendix, Fig. S3F). Thus, Trim21 deficiency does not alter a baseline inflammatory and cardiomyopathy-like transcriptome. Moreover, by day 35, when EAM has advanced to the chronic inflammation and fibrosis stage, Trim21 KO mice retained better cardiac function and exhibited reduced inflammation and fibrosis compared with WT mice, consistent with the differences observed at day 21 (SI Appendix, Fig. S4 A–E).

Finally, to analyze immune cell infiltration, we analyzed infiltrating leukocytes within the heart at day 21 (gating strategy in SI Appendix, Fig. S2D). Trim21 KO hearts contained fewer CD45⁺ cells overall, including both frequencies and estimated counts. Notably, within the myeloid compartment, neutrophils (Nφ, CD11b⁺Ly6G⁺) and monocytes/macrophages (Mo/Mφ, CD11b⁺Ly6G⁻) were decreased in KO-EAM and the Ly6C^high inflammatory Mo/Mφ subset was markedly reduced (Fig. 2E). In sum, Trim21 deficiency confers protection in EAM primarily by limiting inflammatory cell infiltration, thereby alleviating myocarditis and fibrotic remodeling.

TRIM21 Expression in Nonhematopoietic Cardiac Cells Is Required for EAM Progression.

To determine whether the alleviated EAM severity in Trim21 KO mice arises from hematopoietic (immune) or nonhematopoietic (cardiac-resident) cells, we generated four bone marrow chimeras by transplanting bone marrow from Trim21 KO or WT donors into sublethally irradiated WT or Trim21 KO recipients: WT → WT, KO → WT, WT → KO, and KO → KO. After an 8-week reconstitution to achieve stable chimeras, mice were immunized with MyHCα_614-629 (Fig. 3A). Histologically, KO → WT animals were indistinguishable from WT → WT evaluated by H&E inflammation score and Sirius Red fibrosis (Fig. 3 B and C), indicating that TRIM21 in hematopoietic cells is dispensable. In contrast, both WT → KO and KO → KO mice exhibited significantly reduced inflammatory infiltration and collagen deposition compared to WT → WT and KO → WT (Fig. 3 B and C), implicating TRIM21 expression in nonhematopoietic cells is indispensable for EAM. Consistently, mRNA levels of Tnfα, Il6, and Cxcl1 were significantly lower in Trim21 KO recipients than in WT recipients, regardless of donor genotype (Fig. 3D). Furthermore, TRIM21 protein in CD45⁺ immune cells within the myocardium remained unchanged throughout EAM progression (SI Appendix, Fig. S5A). Together, these data indicate that TRIM21’s contribution to EAM pathogenesis is mediated by radio-resistant, nonhematopoietic cardiac-resident cells rather than the hematopoietic immune cells.

Fig. 3.

TRIM21 expression in nonhematopoietic cells is required for EAM progression. (A–D) Irradiated WT or Trim21 KO mice were transplanted with WT or Trim21 KO bone marrow to generate the following bone marrow chimeras. After 8 wk, the chimeric mice were immunized to induce EAM (n = 5 per group). Data are representative of at least two independent experiments. (A) A schematic representation of the generation of bone marrow chimeras is shown. (B–D) By day 21 after immunization, mice were killed to evaluate EAM. (B) Representative images of H&E staining (Top) and Sirius Red staining (Bottom). (C) Pathological scores for each group, including inflammation score and fibrosis score. (D) qPCR analysis of Tnfα, Il6, and Cxcl1 in hearts and Gapdh normalization. (E) Representative IF images showing TRIM21 (red) and α-SMA (green) in EAM hearts at days 0, 14, 21, and 35 postimmunization. Fluorescence intensity calculated by ImageJ (n = 5 per time point). Error bars represent group median, and each data point represents an individual mouse. Data in C–E (Right) were analyzed by one-way ANOVA with Tukey’s post hoc test. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; ns, not significant.

Given the central abundance of cardiomyocytes (CMs) in the heart, we first assessed TRIM21 expression in TNI-positive CM regions and observed baseline expression in CMs, but no significant upregulation during EAM progression (SI Appendix, Fig. S5B). Furthermore, α-SMA-positive areas expanded from day 14 to day 35, indicating accumulation of activated CFs and advancing fibrosis. TRIM21 signal increased over time and substantially colocalized with α-SMA, indicating that elevated TRIM21 in EAM hearts is predominantly localized in activated CFs (Fig. 3E and SI Appendix, Fig. S5C).

IL-17A-Induced TRIM21 Reinforces IL-17A Downstream Signaling in Cardiac Fibroblasts.

Previous studies have shown that IL-17A-driven inflammation in CFs promotes EAM progression and fibrosis (10, 11). In our dataset, RNA-seq DEGs indicated attenuation of IL-17 downstream signaling in Trim21 KO-EAM hearts, and the reduced expression of IL-17A-responsive cytokines and chemokines (e.g., Il6, Cxcl1, Cxcl5, Cxcl10) likely contributes to the decrease in neutrophils and inflammatory monocytes/macrophages observed in vivo (SI Appendix, Fig. S3G). Building on this, we next assessed IL-17 receptor expression to identify the predominant IL-17-responsive compartment; Il17ra was enriched in CFs (Fig. 4A). To further confirm whether IL-17A signaling is required for TRIM21 induction, we treated primary CFs from WT and Il17ra-deficient mice with supernatants from homogenized EAM or control hearts (designated as EAM-sup and Ctrl-sup, respectively). EAM-sup significantly upregulated TRIM21 expression in WT-CFs compared to Ctrl-sup, but had no effect in Il17ra KO-CFs (Fig. 4B and SI Appendix, Fig. S6A), indicating that IL-17A-IL-17RA signaling is necessary for TRIM21’s induction. To directly test this, we stimulated CFs with recombinant IL-17A and found markedly increased TRIM21 expression at both mRNA and protein levels, peaking at 36 h and declining by 48 h (Fig. 4 C and D and SI Appendix, Fig. S6B). IF analysis further confirmed TRIM21 upregulation at 24 h with predominant cytoplasmic localization (Fig. 4E and SI Appendix, Fig. S6 C and D). Together, these data indicate that IL-17A directly induces TRIM21 expression in CFs.

Fig. 4.

TRIM21 is induced by IL-17A and amplifies IL-17A signaling in cardiac fibroblasts (CFs). (A) Primary CMs and CFs were isolated to detect Il17ra, Il17rb, and Il17rc mRNA expression levels with expression normalized to the level of Il17rc (n = 6). (B) TRIM21 expression in CFs from WT (Left) and il17ra KO (Right) mice treated with heart tissue homogenates from EAM or control mice, analyzed by western blot. (C and D) Primary CFs isolated from WT mice were cultured and stimulated with murine IL-17A (100 ng/mL) for 0, 12, 24, 36, and 48 h. qPCR analysis of Trim21 mRNA levels (C, n = 6) and western blot analysis of protein level (D, n = 3). (E) Representative IF images showing TRIM21 (red), α-SMA (green), and DAPI (blue) in IL-17A-treated and control L929 cells (n = 6). (F) CFs from WT and Trim21 KO mice were treated with murine IL-17A (100 ng/mL) for 24 h. qPCR analysis of inflammatory and fibrotic genes (n = 6) and Gapdh normalization. (G) Western blot of IL-17A-stimulated signaling in CFs from WT and Trim21 KO mice for the indicated times (100 ng/mL, n = 3) including phosphorylated JNK, ERK, p38, and p65 with corresponding total proteins. (H–K) WT and Trim21 KO mice were intraperitoneally injected with anti-IL-17A or isotype antibodies (mAb) every other day from Day-1 to Day 21 (n = 5 per group) and were evaluated by day 21 postimmunization. H&E staining with inflammation scores (H and I), Sirius Red staining (H and I), serum CK-MB activity (J), and qPCR of Tnf-α, Il6, and Cxcl1 using gapdh for normalization (K). Error bars represent group means ± SD (A, C, and F) or group median (I–K). Data in (A and F) were analyzed by two-way ANOVA with Tukey’s post hoc test. Data in (C and I–K) were analyzed by one-way ANOVA with Tukey’s post hoc test. Significance: *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; ns, not significant.

Having established that IL-17A induces TRIM21 in CFs, we next investigated whether TRIM21 modulates IL-17A downstream signaling. Primary CFs from WT and Trim21 KO mice were isolated, stimulated with IL-17A, and knockout efficiency was confirmed at mRNA and protein levels (SI Appendix, Fig. S7 A and B). Upon IL-17A stimulation, Trim21 KO CFs showed attenuated induction of cytokines (Il6, Tnfα, Csf2, and Lcn2) and chemokines (Cxcl1, Ccl2, Cxcl5, Ccl7, and Cxcl10) relative to WT, while Col3a1 expression remained unchanged (Fig. 4F and SI Appendix, Fig. S7C). Because IL-17A induces proinflammatory cytokines and chemokines through NF-κB and MAPK pathways, we next assessed these downstream signals. IL-17A-induced phosphorylation of JNK, ERK1/2, p38, and p65 was markedly attenuated in Trim21 KO CFs, suggesting that TRIM21 is required for full activation of these signaling cascades (Fig. 4G and SI Appendix, Fig. S7D). In vivo, serial-section IHC confirmed reduced phosphorylated p38 accompanied by lower TRIM21 in KO hearts versus WT (SI Appendix, Fig. S7E). Conversely, AAV-mediated TRIM21 overexpression enhanced IL-17A-induced inflammatory gene expression (SI Appendix, Fig. S7 F and G) and promoted activation of the same signaling pathways (SI Appendix, Fig. S7H). Collectively, these findings demonstrate that TRIM21, induced by IL-17A, serves as a positive regulator that amplifies IL-17A-driven inflammatory signaling in CFs.

To further determine whether TRIM21 aggravates EAM through IL-17A signaling, we treated WT and Trim21 KO mice with an anti-IL-17A antibody (anti-IL-17-mAb) or isotype antibody (iso-mAb) every other day until day 21. In isotype-treated cohorts, Trim21 KO showed the expected protection with attenuated inflammatory infiltration and Sirius Red⁺ collagen deposition (Fig. 4 H and I and SI Appendix, Fig. S8), lower CK-MB (Fig. 4J), and decreased cardiac Il6, Tnfα and Cxcl1 (Fig. 4K). Strikingly, IL-17A blockade eliminated these genotype differences. Anti-IL-17A reduced inflammation and fibrosis scores and suppressed cytokine expression to a similar extent in both WT and Trim21 KO mice, resulting in WT+anti-IL-17A-mAb and Trim21 KO+anti-IL-17A-mAb being indistinguishable across measures (Fig. 4 H–K). These findings indicate that the protective effects of Trim21 KO are mediated by suppression of IL-17A-driven inflammatory signaling.

TRIM21 Enhances IL-17A-Mediated Signaling by Interacting With TRAF3.

Upon IL-17A binding, IL-17RA recruits ACT1, which engages TRAF6 to activate NF-κB and MAPK pathways, while phosphorylated ACT1 recruits TRAF2 and TRAF5 to stabilize IL-17 target mRNAs (18, 19). Recent studies show that TRAF3 binds IL-17RA and blocks ACT1 recruitment, thus suppressing IL-17A signaling and acting as a negative regulator (20). To investigate how TRIM21 modulates this pathway, we performed coimmunoprecipitation (Co-IP) in HEK293T cells cotransfected with TRIM21 and various TRAF proteins. TRIM21 was found to interact with TRAF3 and TRAF6, but not with TRAF2 or TRAF5 (Fig. 5A). In primary CFs, TRIM21 overexpression enhanced IL-17A-induced cytokine expression, and this effect was not altered by the TRAF6 inhibitor (Fig. 5B). In contrast, knockdown of TRAF3 by siRNA abolished the difference in cytokine expression between WT and Trim21 KO CFs, indicating that TRAF3 is a functional target of TRIM21 in this pathway (Fig. 5C). Co-IP experiments in HEK293T cells confirmed the interaction between TRIM21 and TRAF3 (Fig. 5D). This interaction was further validated by endogenous Co-IP in primary CFs and L929 cells (Fig. 5E).

Fig. 5.

TRIM21 enhances IL-17A-mediated signaling by interacting with TRAF3. (A) Myc-TRIM21 expression plasmids were transfected into HEK293T cells alongside Flag-tagged TRAF2, TRAF3, TRAF5, and TRAF6 to examine potential interactors of TRIM21. (B) CFs were transfected with EGFP-TRIM21 or EGFP-NC AAV (MOI = 150) and then treated with TRAF6 inhibitor, followed by stimulation with IL-17A (100 ng/mL). 24 h later, cells were subjected to qPCR (n = 6). (C) WT and Trim21 KO CFs were transfected with si-Traf3 and then treated with IL-17A (100 ng/mL). 24 h later, cells were subjected to qPCR (n = 6). (D) Cells expressing TRIM21 and Flag-TRAF3 were lysed and subjected to immunoprecipitation using anti-Myc beads (Top) or anti-Flag beads (Bottom), followed by western blot detection with the indicated antibodies. (E) Co-IP analysis of endogenous TRIM21 with endogenous TRAF3 was performed on WT CFs (Left) or L929 cells (Right) stimulated with IL-17A (100 ng/mL) for 24 h. (F) Molecular docking visualized the TRIM21–TRAF3 interaction site. (G) HEK293T cells were transfected with TRIM21 or its truncation mutants alongside Flag-TRAF3 for 36 h, followed by IP with anti-Flag beads and western blot analysis. (H) Co-IP assays were conducted to determine the interaction between Myc-TRIM21 and TRAF3 or its truncation mutants in HEK293T cells. Error bars represent group means ± SD. Statistical analysis: Data in (B and C) were analyzed by two-way ANOVA with Sidak multiple comparisons test. Significance: *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; ns, not significant.

To further elucidate the interaction between TRIM21 and TRAF3, we used protein–protein docking techniques and domain deletion analysis. TRIM21 protein comprises an N-terminal RING domain, a B-box domain, a coiled-coil dimerization domain, and a C-terminal PRY-SPRY domain. TRAF3 protein contains a RING domain followed by a zinc finger domain, a coiled-coil dimerization domain, and a C-terminal TRAF domain. Docking analysis revealed that the PRY-SPRY domain of TRIM21 and the C-terminal TRAF domain of TRAF3 are crucial for their interaction (Fig. 5F). Domain deletion analysis further clarified the domain requirements for the TRIM21–TRAF3 interaction. On the TRIM21 side, deletion of either the RING or B-box domains did not disrupt binding to TRAF3, whereas deletion of the PRY-SPRY domain completely abolished the interaction (Fig. 5G). On the TRAF3 side, deletion of the RING or zinc finger domains had no effect, but removal of the TRAF domain eliminated binding to TRIM21 (Fig. 5H). Notably, both the full C-terminal region (CT, encompassing the coiled-coil and TRAF domains) and the TRAF domain alone were sufficient to mediate interaction with TRIM21 (Fig. 5H). These data suggest that the PRY-SPRY domain of TRIM21 and the TRAF domain of TRAF3 are critical for their interaction, which is essential for TRIM21’s ability to enhance IL-17A signaling.

TRIM21 Promotes TRAF3 Lysosomal Degradation Via Lys-27-Linked Polyubiquitination.

As an E3 ubiquitin ligase, we hypothesized that TRIM21 regulates TRAF3 ubiquitination. In HEK293T cells transfected with Flag-tagged TRAF3 alongside increasing doses of TRIM21, we observed that TRIM21 overexpression significantly reduced TRAF3 protein levels (Fig. 6A). In L929 cells and primary CFs, TRIM21 overexpression decreased endogenous TRAF3 protein without affecting mRNA levels (Fig. 6 B and C), indicating regulation at the translational or posttranslational level. Then, we performed cycloheximide (CHX) pulse–chase experiments to assess TRAF3 protein stability and found that TRIM21 overexpression significantly shortened TRAF3’s half-life (Fig. 6D), indicating that TRIM21 promotes TRAF3 degradation.

Fig. 6.

TRIM21 promotes lysosomal degradation of TRAF3. (A) TRAF3 level in HEK293T cells transfected with Flag-TRAF3 and increasing doses of Myc-TRIM21, detected by western blot. (B) TRAF3 level in L929 cells transfected with Myc-TRIM21 at increasing doses and stimulated with IL-17A (100 ng/mL), detected by western blot. (C) Relative Traf3 mRNA levels in CFs or L929 cells infected with EGFP-Trim21 or EGFP-NC AAV and stimulated with IL-17A, detected by qPCR (n = 6). (D) HEK293T cells transfected with Flag-TRAF3 and Myc-TRIM21 or control vectors for 18 h were treated with the protein synthesis inhibitor CHX (10 μM) for the indicated times. Flag-TRAF3 levels were detected by western blot (n = 3). (E) TRAF3 levels in HEK293T cells treated with MG132 (10 μM), CQ (10 μM), or NH₄Cl (10 mM) posttransfection with Flag-TRAF3 and Myc-TRIM21. Flag-TRAF3 levels were detected by western blot (n = 3). (F) TRIM21 or its truncation mutants were transfected into HEK293T cells alongside Flag-TRAF3 for 24 h and were subjected to immunoblotting with the indicated antibodies. (G and H) Endogenous TRAF3 ubiquitination assessed in TRIM21 overexpressing L929 cells or Trim21 KO primary CFs. (I and J) HEK293T cells transfected with Flag-TRAF3 and the indicated ubiquitin mutant (K6O/R, K11O/R, K27O/R, etc.) together with Myc-TRIM21 or not, and then treated with chloroquine (10 μM). Cell lysates were subjected to IP and IB with the indicated antibodies. Error bars represent group means ± SD. Statistical analysis: Data in C were analyzed by unpaired two-tailed Student’s t test. Data in E (Bottom) were analyzed by one-way ANOVA with Tukey’s post hoc test. Significance: ***P < 0.001; ns, not significant.

To determine whether TRIM21-mediated degradation of TRAF3 is proteasome-dependent or lysosome-dependent, we treated TRIM21-overexpressing cells with the proteasome inhibitor MG132 and lysosomal inhibitors bafilomycin A1 (BAF), chloroquine (CQ), and NH4Cl. Inhibition of lysosomal degradation with BAF, CQ, or NH4Cl effectively blocked TRIM21-induced TRAF3 degradation, while MG132 had no effect (Fig. 6E). Deletion of TRIM21’s RING domain, which is essential for its E3 ligase activity, or the PRY-SPRY domain, responsible for its interaction with TRAF3, both impaired TRAF3 degradation, highlighting the critical role of these domains in the process (Fig. 6F).

To investigate whether TRIM21 could ubiquitinate TRAF3, we cotransfected HEK293T cells with vectors expressing Flag-TRAF3, Myc-TRIM21, and HA-tagged ubiquitin. TRIM21 overexpression significantly enhanced polyubiquitination of TRAF3 (SI Appendix, Fig. S9A). Endogenous ubiquitination assays in IL-17A-treated L929 cells and CFs showed similar results, with TRIM21 overexpression enhancing TRAF3 ubiquitination, and TRIM21 knockout reducing it (Fig. 6 G and H). Moreover, a TRIM21 mutant lacking the RING domain (ΔRING), which loses E3 ligase activity, failed to ubiquitinate TRAF3 (SI Appendix, Fig. S9B). To determine the nature of TRIM21-mediated TRAF3 ubiquitination, we performed ubiquitination assays in HEK293T cells transfected with TRIM21- and TRAF3-expressing vectors, along with wild-type ubiquitin (HA-Ub) or its mutants. TRIM21 primarily promoted Lys-27(K27O)-linked polyubiquitination of TRAF3 (Fig. 6I), and mutation of Lys-27(K27R) completely abolished this effect (Fig. 6J). These findings suggest that TRIM21 primarily mediates nonclassic K27-linked polyubiquitination to facilitate TRAF3 degradation via the lysosomal pathway.

Specific Knockdown of TRIM21 in CFs Alleviates Myocarditis and Cardiac Fibrosis.

To validate the functional role of TRIM21 in CFs during EAM progression and evaluate its potential as a therapeutic target at the late stage, we employed a previously established approach using adeno-associated virus serotype 9 (AAV9) driven by the fibroblast-specific periostin (Postn) promoter to deliver EGFP-tagged Trim21-targeting RNA interference (AAV9-Postn-Trim21-RNAi) (21). AAV9-Postn-Trim21-RNAi or a control virus was administered via tail vein injection to mice at day 21 post–EAM induction (Fig. 7A). We first validated the targeting specificity and knockdown efficiency of AAV9-Postn in vivo. IF analysis confirmed efficient delivery to cardiac tissue (SI Appendix, Fig. S10A). Next, α-SMA⁺ CFs were isolated from heart tissues (SI Appendix, Fig. S10B), and western blotting showed a reduction of TRIM21 expression in CFs transduced with AAV9-Postn-Trim21-RNAi compared to controls (Fig. 7B).

Fig. 7.

TRIM21 knockdown in CFs attenuates myocarditis and fibrosis in mice. (A) Schematic representation of the CF-specific Trim21 knockdown strategy. (B) Mice received AAV9-Postn-NC-RNAi or AAV9-Postn-Trim21-RNAi at day 21 post-EAM and were killed at day 45 (n = 3 per group). TRIM21 protein levels in α-SMA-positive CFs isolated from these hearts were detected by western blot. (C–F) Mice received AAV9-Postn-NC-RNAi or AAV9-Postn-Trim21-RNAi at day 21 post-EAM and were evaluated at day 45 (n = 6 per group). (C) Serum CK-MB activity. (D) Echocardiography (EF, FS). (E) Gross heart morphology; H&E with inflammation score; Sirius Red with fibrosis score. (F) qPCR of Tnfα, Cxcl1, Il6, Csf2, Cxcl5, and Cxcl10 in whole hearts and Gapdh normalization. Error bars represent group means ± SD. Each data point represents an individual mouse. Statistical analysis: Data were analyzed by two-way ANOVA with Sidak multiple comparisons test. Significance: *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; ns, not significant.

CF-specific TRIM21 knockdown reduced cardiac enlargement, decreased white lesions, lowered HW/BW ratios, and diminished myocyte injury (lower CK-MB) compared with controls (Fig. 7C and SI Appendix, Fig. S10C). Echocardiography revealed that AAV9-Postn-Trim21-RNAi administration significantly improved left ventricular systolic function, with increased EF and FS (Fig. 7D). Histopathological analysis demonstrated substantial attenuation of myocardial inflammation and fibrosis in AAV9-Postn-Trim21-RNAi-treated mice (Fig. 7E). Furthermore, qPCR demonstrated suppressed IL-17-driven inflammatory responses in mice with CFs knockdown of TRIM21, marked by downregulation of Il6, Tnfα, Cxcl1, Cxcl5, Cxcl10, and Csf2 in cardiac tissues (Fig. 7F). These data provide in vivo evidence that knockdown of TRIM21 in CFs alleviates EAM progression.

Discussion

Our previous studies showed that TRIM21 deficiency exacerbates viral myocarditis by impairing type I interferon-mediated antiviral responses, leading to increased viral replication (17). Interestingly, while TRIM21 exerts opposing effects in viral versus autoimmune myocarditis, its fundamental role as an amplifier of immune signaling remains consistent. In viral myocarditis, TRIM21 is mainly expressed in cardiomyocytes where it enhances antiviral immunity. In contrast, in autoimmune myocarditis, TRIM21 is upregulated in CFs and amplifies IL-17A-driven proinflammatory signaling. Consequently, TRIM21 deficiency reduces IL-17A signaling, immune cell infiltration, and cytokine production, thereby attenuating disease progression. This duality reflects distinct pathogenic mechanisms but underscores TRIM21 as a pivotal enhancer of immune responses beneficial in viral defense but detrimental in autoimmune inflammation. Supporting this, targeted knockdown of TRIM21 in CFs via AAV9-Postn-Trim21-RNAi at the fibrotic stage significantly mitigated myocarditis severity. These results highlight TRIM21 in CFs as a critical modulator and promising therapeutic target in autoimmune myocarditis.

TRIM21, initially identified as an autoantigen in patients with autoimmune diseases such as Sjögren’s syndrome and systemic lupus erythematosus (SLE), is a multifunctional E3 ubiquitin ligase involved in various cellular processes, including immune responses, oxidative stress, antiviral defense, and tumor progression (13, 22–25). Recent research has shown that TRIM21 is widely expressed in carcinoma cells, epithelial cells, endothelial cells, and fibroblasts. In EAM, TRIM21 mRNA and protein are robustly increased in whole-heart lysates, with spatial enrichment in fibrotic regions. Bone-marrow chimeras indicate that nonhematopoietic, rather than hematopoietic, TRIM21 drives disease exacerbation. Coimmunofluorescence localizes TRIM21 to α-SMA-positive areas, with no enrichment in CD45⁺ immune cells and minimal change in isolated CD45⁺ fractions. While TRIM21 may also be expressed in other stromal or parenchymal compartments, the requirement for IL-17A signaling, whose receptor is predominantly expressed in CFs, together with the therapeutic benefit of a Postn-targeted AAV9, supports CFs as the principal cell type in which TRIM21 acts. Thus, we propose a CF-centric mechanism in which TRIM21 tunes IL-17A-responsive signaling to promote inflammatory-fibrotic remodeling, while potential contributions from nonfibroblast cells. The regulation of TRIM21 expression by cytokines such as IFN-α, IFN-β, IFN-γ, and TNF-α, as well as by innate immune signaling pathways triggered by TLRs, is well documented (26, 27). IL-17A, in particular, can activate NF-κB signaling, which may lead to the transcriptional upregulation of TRIM21. Notably, our study reports that IL-17A treatment markedly upregulates TRIM21 expression in CFs.

As is well established, IL-17A rapidly amplifies tissue inflammation by activating NF-κB and MAPK programs that induce chemokines and proinflammatory cytokines. It often acts with TNF to drive neutrophil recruitment and reinforce barrier defense (28). Current evidence supports a temporal distinction. IL-17A is proinflammatory in early or acute phases, with strong chemokine and cytokine induction, neutrophil influx, and epithelial repair. In late or chronic phases, it becomes profibrotic, with fibroblast activation and proliferation, extracellular matrix deposition, and crosstalk with TGF-β and ERK (19, 29, 30). In skin repair, early IL-17A signaling in stromal and epithelial cells initiates the inflammatory phase and accelerates re-epithelialization. Over this short window, fibroblasts preferentially upregulate inflammatory mediators rather than collagen or α-SMA (31–33). The cardiac stroma shows a similar early bias toward inflammation. Cihakova’s group reported that acute IL-17A stimulation in CFs preferentially induces chemokines/cytokines, linking early inflammation to later fibrotic remodeling, rather than immediate collagen or α-SMA induction (11). Consistent with this, we found that loss of Trim21 lowers Il6, Tnf-α, Csf2, Cxcl1, and Ccl2 in CFs at this acute time point without altering collagen transcripts. Our study extends this mechanism by identifying TRIM21 as a CF-intrinsic amplifier of the IL-17A-NF-κB/MAPK axis that enhances and sustains the inflammatory response. Functionally, this amplification increases myeloid cell infiltration and promotes cardiac inflammation at day 21, without implying that IL-17A acutely drives collagen/α-SMA expression. Accordingly, we position TRIM21 as a fibroblast-intrinsic amplifier required for the full magnitude of IL-17A-driven chemokine responses and the day-21 inflammatory burden, thereby connecting early inflammatory signaling to later fibrotic remodeling.

Upon IL-17A binding to its receptors (IL-17RA and IL-17RC), the receptor complex recruits the adaptor protein ACT1 via the SEFIR domain, engaging TRAF6, which is critical for downstream activation of signaling pathways, including NF-κB and MAPK (18). Additionally, IL-17RA contains a C-EBP-β activation domain (CBAD), linked to the negative regulation of this signaling pathway (34). TRAF3 and A20, both known signaling inhibitors, are associated with the CBAD (19, 35). Recent findings indicate that FXYD3 enhances IL-17A signaling by competing with IL-17R for TRAF3 binding in keratinocytes (20). Our study identified that TRIM21 directly interacts with TRAF3 and TRAF6. However, further experiments using Traf3 siRNA transfection or a TRAF6 inhibitor demonstrated that TRIM21 positively regulates the IL-17A downstream pathway by targeting TRAF3. Deletion mutant analysis confirmed that TRIM21 interacts with TRAF3 through its PRY-SPRY domain, whereas TRAF3 engages TRIM21 through its coiled-coil and TRAF domains. TRAF3, a relatively less understood member of the TRAF family, is known to play crucial roles in various signaling pathways, including those regulating inflammation, antiviral immunity, and cell survival. Our study adds to the understanding of TRAF3’s role in IL-17A-stimulated CFs in the context of myocarditis. However, the potential roles of TRAF3 in CFs in response to other stimuli or other cardiac diseases remain to be further investigated.

As an E3 ubiquitin ligase, TRIM21 dictates the fate and function of its substrate proteins through the type of ubiquitination. While K48-linked polyubiquitination is well known for mediating proteasomal degradation, our previous work also demonstrated that TRIM21 targets HBx for degradation via K48-linked polyubiquitination, thereby suppressing HBV replication (36). In this study, we found that TRIM21 promotes TRAF3 degradation through K27-linked rather than K48-linked, polyubiquitination. K27-linked ubiquitin chains have been implicated in diverse cellular processes, including immune signaling, inflammation, and DNA damage response (37). The TRIM family of E3 ligases plays a central role in assembling K27-linked chains and mediating protein–protein interactions (38). Another previous study also reported that TRIM21 promotes the K27-linked polyubiquitination of MAVS, thereby enhancing type I interferon signaling but not promoting its degradation (17). In contrast, MARCH8 promotes the K27-linked polyubiquitination of MAVS at K7, serving as a recognition signal for autophagic degradation (39). Consistently, in this study, our data showed that autophagic-lysosomal degradation inhibitors (BAF, CQ, or NH4Cl), but not the proteasome inhibitor MG132, effectively abrogated TRIM21-induced TRAF3 degradation, suggesting that TRIM21 destabilizes TRAF3 protein in a lysosome-dependent manner. Further, ubiquitination experiments verified TRIM21-mediated K27-linked polyubiquitination of TRAF3 to promote its degradation. These results underscore the complexity of ubiquitination, which may involve different lysine residues and/or E3 ligases, warranting further studies to elucidate these mechanisms.

Together, our data establish that IL-17A induces TRIM21 in CFs, which in turn augments IL-17A signaling through TRAF3, sustaining proinflammatory output and promoting remodeling. However, there are several limitations to this study that should be acknowledged. First, while we demonstrate the involvement of TRIM21 in IL-17A signaling in mice, whether these findings are conserved in human myocarditis or DCM patients remains to be explored. Second, the application of single-cell and spatial transcriptomics in future studies will be valuable to resolve cell–cell interactions and CF state dynamics across inflammatory and remodeling phases. Last, although targeting TRIM21 offers promising therapeutic potential, it will be crucial to consider the specific disease context and stage when evaluating its clinical applicability.

Materials and Methods

Animal Experiments.

All mouse experimental procedures performed in this study were conducted in accordance with guidelines of the Institutional Animal Care and Use Committee at Soochow University. The BALB/c mice were purchased from Shanghai SLAC Animal Inc. (Shanghai, China). Trim21-deficient mice (Trim21 KO) on a C57BL/6J background were generated as described previously (17). Trim21 KO mice on the C57BL/6J background were crossed with BALB/c mice for more than eight generations to establish a Trim21 KO line in the BALB/c background (11). EAM was induced as described previously (12). Six-week-old male BALB/c mice or Trim21 KO mice were immunized on days 0 and 7 by subcutaneous injection at two sites, the axilla and the back. At each site, 100 µL containing 100 µg of α-myosin heavy chain peptide MyHCα_614-629 (Ac-SLKLMATLFSTYASAD-OH; purity ≥95%; Synpeptide, Nanjing, China) was emulsified 1:1 with Complete Freund’ s Adjuvant (CFA) (Sigma-Aldrich, F5881). Control mice received PBS emulsified with CFA. Bone marrow chimera mice were generated by lethally irradiating 6-week-old male WT or Trim21 KO mice with two doses of 600 rad, followed by intravenous injection of 5 × 10⁶ bone marrow cells from WT or Trim21 KO donors, and EAM was induced after an 8-week reconstitution period. For Trim21 knockdown in CFs, an AAV9 vector containing EGFP-tagged Trim21-targeting RNAi under the periostin promoter (AAV9-Postn-Trim21-RNAi) was used. Mice were intravenously injected with 3 × 10¹¹ v.g. of AAV9-Postn-Trim21-RNAi at day 21 post–EAM induction, and were euthanized at day 45 for downstream analyses, all experimental procedures involving animals were performed in strict accordance with protocols approved by the Institutional Animal Care and Use Committee at Soochow University.

RNA Sequencing Analysis.

Total RNA was extracted from WT and Trim21 KO mouse heart tissues collected on day 21 of EAM and sequenced using the Illumina NovaSeq 6000 platform (Shanghai Biotechnology Corporation). High-quality reads were obtained using Seqtk, mapped to the mouse genome with Hisat2, and gene fragments were quantified by StringTie followed by TMM normalization. Detailed analysis and methods are provided in the SI Appendix.

Statistical Analysis.

Data are presented as the mean ± SD or median for each group and were analyzed using GraphPad Prism 10.0 (GraphPad Software, Inc., CA). Data are representative of three independent experiments. Statistical significance was determined using an unpaired two-tailed Student’s t test, one-way ANOVA followed by Tukey’s post hoc test, the Wilcoxon test with Holm–Bonferroni correction, or two-way ANOVA with Sidak’s multiple comparisons test, as appropriate. P value < 0.05 was considered statistically significant. Significance is indicated as follows: *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; ns, not significant.

Details for each experimental method are provided in the SI Appendix, Materials and Methods.

Supplementary Material

Appendix 01 (PDF)

Dataset S01 (XLS)

Dataset S02 (XLSX)

Dataset S03 (TXT)

Dataset S04 (CSV)

Data, Materials, and Software Availability

RNA sequencing data have been deposited in Gene Expression Omnibus database (GSE281873) (40). Study data are included in the article and/or supporting information.