Cardiomyocyte‐Derived USP20 Attenuates Diabetic Cardiomyopathy by Facilitating the Degradation of STING and Mitigating STING‐Mediated Inflammation

ABSTRACT

Although extensive clinical and basic research has been conducted on diabetic cardiomyopathy (DbCM), the therapeutic efficacy for this condition remains significantly limited. Ubiquitin‐specific peptidase 20 (USP20), a deubiquitinating enzyme, plays an essential role in regulating protein ubiquitination and modulating various cellular processes. In this study, we aimed to investigate the effect of USP20 on the pathogenesis of DbCM, which may provide a novel therapeutic target for its treatment. The cardiomyocyte‐specific USP20 conditional knockout (USP20CKO) mice were employed in this study. The type 2 diabetes mouse model was established using db/db leptin receptor–deficient mice and high‐fat diet/streptozotocin–induced mice. USP20 expression was downregulated in the myocardium of diabetic mice. Cardiomyocyte‐specific USP20 deficiency aggravated cardiac remodeling and myocardial dysfunction in diabetic mice. LC–MS/MS analysis, along with Co‐IP results, demonstrated the interaction between stimulator of interferon genes (STING) and USP20. In mechanism, USP20 directly binds to STING and promotes its degradation through the autophagy pathway by deubiquitinating p62 via its active site C154, thereby alleviating the myocardial inflammation and improving ventricular remodeling and heart failure induced by diabetes.

USP20 directly binds to STING and promotes the degradation of STING through the autophagy pathway by deubiquitinating p62 via its active site C154, thereby alleviating myocardial inflammation and improving ventricular remodeling and heart failure induced by diabetes.

Abbreviations

ANP

atrial natriuretic peptide

DbCM

diabetic cardiomyopathy

DUBs

deubiquitinating enzymes

EF

ejection fraction

FS

fractional shortening

IL‐1β

interleukin‐1β

STING

stimulator of interferon genes

UB

ubiquitin

UPS

ubiquitin‐proteasome system

USP20

ubiquitin‐specific peptidase 20

1. Introduction

Diabetic cardiomyopathy (DbCM), a progressive cardiac disorder independent of coronary artery disease or hypertension, caused by metabolic disorders in diabetes, leading to abnormal structure and function of the heart [1]. DbCM is mainly characterized by myocardial apoptosis, myocardial hypertrophy and fibrosis, as well as cardiac systolic and diastolic dysfunction, which seriously affects the prognosis of patients [2]. Despite extensive clinical and basic research conducted on this issue, several commercial drugs are available for the management of DbCM, the therapeutic outcomes for DbCM remain considerably limited. Therefore, it is imperative to elucidate the pathogenesis of DbCM and to develop efficacious interventions to improve the symptoms of DbCM.

Protein ubiquitination represents one of the most significant post‐translational modifications, playing a crucial role in regulating protein degradation and activity while contributing to the maintenance of normal physiological processes [3, 4]. Deubiquitylation is the reverse process of ubiquitination, which is catalyzed by deubiquitinating enzymes (DUBs) to remove ubiquitin from substrate proteins [5, 6]. Recently, a growing body of research has demonstrated that DUBs are involved in the pathogenesis of cardiovascular diseases. OTUD1 has been reported to deubiquitinate STAT3, thereby promoting cardiac remodeling and contributing to the progression of heart failure [7]. Ubiquitin‐specific protease 28 (USP28) prevents mitochondrial morphological and functional defects and ameliorates diabetic myocardial dysfunction by deubiquitinating PPARα to regulate its stability [8]. Moreover, USP25 alleviates cardiac hypertrophy by deubiquitinating and stabilizing SERCA2a [9]. In the present study, we observed a reduced expression of ubiquitin‐specific protease 20 (USP20) in heart tissue of diabetic mice, suggesting that USP20 may take part in the pathogenesis of DbCM.

USP20, a crucial member of the ubiquitin‐specific protease family, plays a significant role in regulating protein ubiquitination and thus modulates various cellular processes such as cell survival, proliferation, and inflammation [10]. The physiological functions of USP20 have been identified in several conditions, including cancer, metabolic disorders, and neurological diseases [11, 12, 13, 14]. It has been reported that USP20 induces cancer cell proliferation, invasion, and migration by deubiquitinating β‐catenin to regulate its stability [15]. USP20 contributes to maintaining ER homeostasis by deubiquitinating and stabilizing reticulophagy regulator 1 (RETREG1) [16]. Building upon evidence that USP20 regulates inflammatory pathways, including its reported mitigation of neuroinflammation and neuronal death during ischemic stroke via PTEN signaling [12] and its inhibition of tumor necrosis factor‐induced vascular smooth muscle cell inflammation in atherosclerosis [17], we noted that chronic inflammation and metabolic stress are central to diabetic cardiomyopathy. This condition is characterized by persistent low‐grade inflammation and myocardial remodeling. Importantly, several ubiquitin‐specific proteases have been identified as key regulators of inflammation in diabetic cardiomyopathy, primarily by modulating core inflammatory components. For example, USP13 and USP14 stabilize the NLRP3 inflammasome, while USP24 exacerbates injury through NF‐κB pathway activation [18, 19, 20]. This established role of USPs in fine‐tuning inflammatory signaling in the diabetic heart prompted us to investigate whether USP20 also plays a significant part in diabetic cardiomyopathy.

Our findings reveal a dramatic downregulation of USP20 in diabetic myocardium, and cardiomyocyte‐specific USP20 deficiency aggravates cardiac remodeling and myocardial dysfunction in diabetic mice. Mechanistically, USP20 promotes the degradation of stimulator of interferon genes STING through the autophagy pathway by deubiquitinating p62 via its active site C154, thereby alleviating myocardial inflammation and improving ventricular remodeling and heart failure induced by diabetes. This mechanism is distinct and novel as USP20 does not directly regulate NLRP3 or NF‐κB but instead governs the upstream innate immune sensor STING. By deubiquitinating and activating the selective autophagy adapter p62, USP20 drives the autophagic‐lysosomal degradation of STING. This USP20‐p62‐STING axis represents a previously unrecognized layer of regulation, positioning USP20 as a critical checkpoint that suppresses inflammation by controlling the turnover of a key immunostimulatory molecule through the autophagy pathway. These findings advance our understanding of DbCM pathogenesis and propose a novel therapeutic strategy for managing DbCM, thereby offering a new therapeutic target for this condition.

2. Methods

2.1. Human Studies and Ethics Statement

In this study, human left ventricular (LV) tissue samples were obtained from explanted hearts of six diabetic patients, including three with preserved cardiac function and three with heart failure. Briefly, upon heart excision during transplantation, LV tissue was immediately dissected and snap‐frozen for subsequent analysis. All procedures involving human heart specimens were reviewed and formally approved by the Ethics Committee of the First Affiliated Hospital of Wenzhou Medical University (Approval No. KY2022‐156).

2.2. Animal Experiment

The establishment of STZ/HFD models for diabetic cardiomyopathy involved administering a high‐fat diet (comprising 45% fat, 35% carbohydrate, and 20% protein) to eight‐week‐old mice over an eight‐week period, with body weights recorded weekly. Following the completion of the feeding period, intraperitoneal injections of low‐dose streptozotocin (40 mg/kg) were administered to the mice for three consecutive days in order to induce a type 2 diabetes mellitus model. Control mice received an equal volume of citrate buffer intraperitoneally at the same time. Blood glucose levels in the mice were measured 1 week after the injection. Mice exhibiting fasting plasma glucose levels exceeding 11.1 mmol/L or random blood glucose levels greater than 16.7 mmol/L for two consecutive days were selected for subsequent studies and continued on an additional 8 weeks of high‐fat diet (45% fat, 35% carbohydrate, and 20% protein). Oral glucose tolerance tests and insulin tolerance tests were conducted 1 week prior to euthanasia. All experimental procedures were conducted in accordance with the ethical regulations and guidelines established by the US National Institutes of Health. The animal welfare conditions received approval from the Laboratory Animal Ethics Committee at the First Affiliated Hospital of Wenzhou Medical University (approval No. WYYY‐IACUC‐AEC‐2025‐018). The animals had unlimited access to water and food, and they were housed in the Animal Center of the First Affiliated Hospital of Wenzhou Medical University, where temperature and humidity levels were maintained within appropriate ranges.

2.2.1. Experiment 1st

In order to elucidate the role of USP20 in diabetic cardiomyopathy, we obtained eight‐week‐old cardiac‐specific USP20 knockout (USP20CKO) mice on a C57BL/6 background and C57BL/6 wild‐type (WT) mice from GemPharmatech Co. Ltd. (Nanjing, China). Usp20^flox/flox(Usp20^fl/fl) and homozygous cardiomyocytes‐conditional knockout Usp20^{flox/flox,Myh6‐iCre}(USP20CKO) mice were generated by CRISPR/Cas9‐mediated genome editing technique. To identify the mouse genotypes, cDNA was extracted from mouse tail and primers for identification of the USP20^fl/fl type and USP20CKO type. The genotyping primers were as follows: USP20 (F: ATTCAGGAGCGGGAGGAG; R: CCAGAATGAGCCCAGCAA); Myh6‐iCre (F: CCTGCTGTCCATTCCTTATTCCATA; R: ATATCCCCTTGTTCCCTTTCTGC). The mice were randomly assigned to four groups: CON+USP20^fl/fl group (n = 10), CON+USP20CKO group (n = 10), HFD/STZ + USP20^fl/fl group (n = 10), and HFD/STZ + USP20CKO group (n = 10). Mice of both sexes were included, with equal numbers in each group.

2.2.2. Experiment 2nd

To further investigate the impact of USP20 on STING, mice were infected with adeno‐associated virus serotype 9 (AAV9‐cTnT), which included a cardiac‐specific promoter cTNT (cTNTpMCS‐3Flag‐T2A‐EGFP, GV571; Genechem Co. Ltd.), along with either an empty vector (AAV9‐cTnT‐EV) or USP20 (AAV9‐cTnT‐USP20^oe). The STING inhibitor C‐176 was procured from MCE (#HY‐112906). Seven‐week‐old db/db mice from Jackson Laboratory were divided into four groups: AAV9‐cTnT‐EV + vehicle group (n = 10), AAV9‐cTnT‐USP20^oe + vehicle group (n = 10), AAV9‐cTnT‐EV + C‐176 group (n = 10), and AAV9‐cTnT‐USP20^oe + C‐176 group (n = 10). The mice received injections of AAV9 via the tail vein at a dose of 2 × 10¹¹ v.g./mouse/month for 2 months. Additionally, the mice treated with the STING inhibitor received daily intraperitoneal injections of C‐176 in corn oil (1.34 g/kg) for 8 weeks, while the control group was administered an equivalent volume of corn oil. Mice of both sexes were included, with equal numbers in each group.

2.3. Echocardiography

The evaluation of cardiac function in murine subjects was conducted using transthoracic echocardiography, performed with the Vevo LAB 3.1.1 High‐Resolution Preclinical Imaging System (FUJIFILM Visual Sonics). This device features the capability for automated calculation of all cardiac functional parameters, thereby enabling a comprehensive analysis of the resulting data.

2.4. Histological Ananlysis

Mouse myocardial tissues were fixed in 4% paraformaldehyde (PFA, Solarbio, #P1110), then dehydrated and embedded in paraffin. Following the removal of paraffin with xylene, the samples were rehydrated through a series of alcohol gradients. Tissue sections (6 μm) were stained using Hematoxylin and Eosin (H&E, Solarbio, #G1120), Masson's Trichrome (Solarbio, #G1340), and Sirius Red (Solarbio, #S8060) to assess the extent of cardiac hypertrophy and fibrosis. The remaining heart tissues were fixed in OCT (Sakura Finetek, #4583). The sections underwent staining with Wheat Germ Agglutinin (WGA, Solarbio, #I3300) to evaluate the cross‐sectional area of cardiomyocytes. Images were acquired using either a fluorescence microscope or a light microscope and subsequently analyzed using ImageJ software.

2.5. Cell Culture and Transfection

The human embryonic kidney 293T (HEK293T) cells (# GNM 6) and HL‐1 cardiomyocyte cell line (HL‐1, #SCSP‐502) were obtained from the Shanghai Academy of Chinese Sciences. Neonatal rat primary cardiomyocytes (NRPCs) were isolated from the hearts of Sprague–Dawley rats, following the methodology outlined in a previously published study by our research team [21]. Cells were cultured in Dulbecco's modified Eagle's medium (DMEM, Thermo Fisher, #C1199), with 10% foetal bovine serum (FBS, Opcel, #BS‐1105), 1% penicillin–streptomycin (100 U/mL penicillin and 100 mg/mL streptomycin, Gibco, #15140122), and incubated in a 5% CO₂ incubator at 37°C.

To establish a diabetic cardiomyopathy cell model, the cells were treated with high glucose (HG, 35 mM D‐glucose) in conjunction with palmitate (PA, 500 μM) for a duration of 24 h, while the control group was administered 5.5 mM D‐glucose (LG).

Small interfering RNA (siRNA) and negative control (NC) siRNA were synthesized by RiboBio (Guangzhou, China). LipofectAMINETM 2000 (Thermo Fisher, #No. 3035197) was used for USP20 gene silencing according to the manufacturer's instructions. Flag‐USP20 (mouse), His‐STING (mouse), His‐P62 (mouse), HA‐UB (mouse), and HA‐UB‐K63 (mouse) were synthesized by GeneChem (Shanghai, China). Expression of these plasmids was achieved by LipofectAMINETM 3000 (Thermo Fisher, #No. 3039420) according to the manufacturer's instructions.

2.6. The Construction of USP20‐Deficient NIH/3 T3 Cells

We generated USP20‐knockout NIH/3 T3 cells (gUSP20‐NIH/3 T3) using lentiviral CRISPR‐Cas9. The lentivirus, carrying a USP20‐targeting gRNA, was produced in 293 T cells. NIH/3 T3 cells were infected with the collected viral supernatant and selected with blasticidin. After picking monoclonal colonies, the knockout efficiency was confirmed by Western blot.

2.7. Western Blotting

The ice‐cold buffer (Boster Biological Technology, #AR0103/AR0101), containing a protein phosphatase inhibitor (Solarbio, #P1260), was used to lyse heart tissues and cells. After quantifying the protein concentration using the BCA Protein Assay Kit (Solarbio, #PC0020), the samples were mixed with 5X dual‐color protein loading buffer (Fude Biological Technology, #FD006) and boiled for 10 min. The total proteins were then separated by SDS‐PAGE and subsequently transferred to mixed cellulose ester transfer membranes (NC, Merck, #HATF00010). Before incubating the membranes with primary antibodies for at least 8 h at 4°C, they were blocked with either 5% skim milk (Saiguo, #1172GR500) or 5% bovine serum albumin (BSA, Solarbio, #G8590) for 2 h at room temperature. Following this step, membranes were probed with secondary antibodies (Beyotime, #P0948 and P0946) for 2 h at 37°C. Finally, blots were developed using Meilunbio fg supersensitive ECL luminescence reagent (Meilunbio, #MA0186), and results were quantified using ImageJ software. Primary antibodies against GAPDH (1:1000, HUABIO, ET1601‐4), USP20 (1:1000, proteintech, 17491‐1‐AP), MyHC (1:1000, HUABIO, ET1702‐88), ANP (1:200, SANTA, sc‐515701), STING (1:1000, HUABIO, HA722832), P62 (1:1000, HUABIO, HA721171), p‐STING (1:1000, HUABIO, HA723137), TBK1 (1:1000, proteintech, 28397‐1‐AP), p‐TBK1 (1:1000, proteintech, 82383‐1‐RR), IRF3 (1:1000, proteintech, 11312‐1‐AP), p‐IRF3 (1:1000, proteintech, ET1705‐81), HA (1:1000, proteintech, 51064‐2‐AP).

2.8. Coimmunoprecipitation (Co‐IP)

The myocardial tissues or HL‐1 cells were lysed in lysis buffer (Boster Biological Technology, #AR0103/AR0101). A small portion of the supernatant was retained as input. The resulting supernatants were incubated with primary antibodies overnight at 4°C. Subsequently, 20 μL of protein A/G‐coupled agarose beads (Beyotime, #P2055) were added, followed by incubation at 4°C with gentle rotation for 2 h. The agarose beads were then eluted using PBS (GIBCO, #C20012500BT), and co‐precipitated proteins were analyzed by Western blotting as previously described.

2.9. Immunohistochemical Staining

Using a microwave‐based antigen retrieval technique on paraffin‐embedded sections, the samples were first incubated with 3% H₂O₂ to block endogenous peroxidase activity and subsequently treated with 5% BSA (Solarbio, #G8590) to reduce nonspecific binding. The sections were then incubated overnight at 4°C with the primary antibody, followed by incubation with HRP‐conjugated secondary antibodies for 1 h. Finally, the sections were labeled using the DAB Staining Kit (ZSJQ‐BIO, #ZLI‐9018).

2.10. Immunofluorescence Staining

Frozen sections or cells were fixed in 4% paraformaldehyde (PFA) (Solarbio, #P1110) for 15 min. Subsequently, they were permeabilized using 0.5% Triton X‐100 (Solarbio, #9002‐93‐1) in phosphate‐buffered saline (PBS, Solarbio, #P1010) for 20 min. The samples were then incubated with 5% bovine serum albumin (BSA, Solarbio, #G8590) at room temperature for 30 min. Afterward, the samples were treated with primary antibodies and incubated overnight at 4°C. The following day, fluorescent secondary antibodies (HUABIO, #HA1121 and HA1125) were applied at room temperature for 1 h. Finally, the samples were stained with DAPI (Yeasen, #36308ES11) to visualize nuclei.

2.11. ELISA

The levels of IL‐1β (westang, #F10770) and ANP (Elabscience, #E‐EL‐M0166) in mouse heart tissues were measured by ELISA kits according to the manufacturer's instructions. Absorbance at 450 nm was measured in each well by visualization of color intensity development.

2.12. LC–MS/MS Analysis

Plasmids encoding Usp20 cDNA or a control empty vector were transfected into HL‐1 cells. After 24 h, the cells were treated with high glucose and palmitic acid for an additional 24 h. Subsequently, cell lysates were prepared for immunoprecipitation. The resulting samples underwent liquid chromatography–tandem mass spectrometry (LC–MS/MS) analysis, which was performed by Shanghai Aimmass (Shanghai, China).

2.13. Real‐Time Quantiative PCR

The total RNA extraction from heart tissues and HL‐1 cells was conducted utilizing TRIzol (Thermo Fisher, #340312). cDNA synthesis was performed using the RevertAid First Strand cDNA Synthesis Kit (Fermentas, #K1622). RT‐qPCR analyzes were carried out to evaluate target genes with TB Green R Ex Taq TM II (TaKaRa, #RR820A) on a TANON 5000 chemiluminescence imaging system. A list of the primer pairs utilized in this study is available in the Table S1.

2.14. Transcriptome Sequencing

Transcriptome sequencing was conducted on heart tissue samples from mice. Total RNA was extracted using TRIzol. Library preparation and transcriptome sequencing were executed by Personalbio (Shanghai) Co. Ltd. Differentially expressed gene analysis was performed with a fold change threshold of ≥ 1.2 and an adjusted p value of < 0.05.

2.15. Statistical Analysis

Data are presented as mean ± standard error of the mean (SEM). The sample sizes (n) denote biological replicates. Statistical analyzes were conducted using Student's t‐test for comparisons between two groups, and ANOVA followed by post hoc Tukey's test was employed for comparisons among multiple groups. A p‐value of less than 0.05 was considered statistically significant. All statistical analyzes related to the population studies were performed utilizing GraphPad Prism version 10.0 software.

3. Results

3.1. Identification of USP20 as a Critical Regulator in Diabetic Cardiomyopathy

To characterize the cell type‐specific expression profile of USP20 in diabetic hearts, we performed single‐cell RNA sequencing (scRNA‐seq) on myocardial tissue from genetically diabetic‐obese (db/db) mice. Uniform Manifold Approximation and Projection (UMAP) clustering identified six distinct cardiac cell populations: cardiomyocytes, fibroblasts, macrophages, endothelial cells, neutrophil granulocytes, and T cells. Notably, USP20 was predominantly expressed in cardiomyocytes (Figure 1A). Subsequently, cardiomyocytes, cardiac fibroblasts, and macrophages were exposed to combined high glucose and palmitic acid (HG/PA) stress. Quantitative immunoblotting showed that USP20 expression was significantly reduced in cardiomyocytes under HG and PA stimulation. In contrast, no significant change in USP20 expression was observed in macrophages or fibroblasts under the same metabolic stress conditions (Figure 1B). Furthermore, immunofluorescence co‐staining confirmed that USP20 expression was markedly decreased in cardiomyocytes but remained unchanged in fibroblasts and macrophages within the heart tissue of diabetic mice (Figure 1C). We also found that USP20 mRNA levels were substantially lower in patients with diabetic heart failure (Figure 1D and Table S4). This result was further supported by data from diabetic mouse heart tissue and HG/PA‐induced HL‐1 cells, with consistent findings both in vivo and in vitro. Specifically, USP20 protein and mRNA levels were significantly reduced in diabetic mouse heart tissue (Figure 1E) and in HL‐1 cells exposed to HG and PA (Figure 1F). Additionally, we achieved overexpression of USP20 in cardiomyocytes via transfection with a USP20 plasmid (Figure S1A). HG and PA exposure led to a pronounced increase in ANP expression in these cardiomyocytes; however, USP20 overexpression effectively attenuated this elevated ANP expression under the same conditions (Figure 1G). Together, these findings suggest that USP20 plays a critical regulatory role in diabetic cardiomyopathy.

FIGURE 1.

USP20 expression is cardiomyocyte‐specifically downregulated in diabetic myocardium. (A) Single‐cell sequencing analysis of Usp20 expression in heart tissue from genetically diabetic‐obese (db/db) mice. For each group, single‐cell suspensions from 3 hearts were pooled as 1 sample. The tSNE dimensional reduction shows 6 main cell types of heart, including cardiomyocytes, endothelial cells, fibroblasts, neutrophil granulocytes, macrophages, and T cells. The red dots represent Usp20. (B) The expression level of USP20 in neonatal rat cardiomyocytes (NRCMs), fibroblasts, and macrophages induced by combined high glucose and palmitic acid (HG + PA) assessed by western blotting and the statistical results. n = 5. (C) Representative immunofluorescence images showing the colocalization of USP20 and α‐Actin, USP20 and Vimentin, USP20 and F4/80 in heart tissues from control and diabetic mice. (D) The mRNA expression levels of Usp20 in heart tissue of diabetic patients were assessed by PCR. n = 3. (E) The expression level of USP20 in heart tissue of diabetic mice tested by western blotting. (F) The statistical results of (E) data have been normalized against GAPDH and the mRNA expression levels of Usp20 in heart tissue of diabetic mice assessed by PCR. n = 6. (G) The expression level of USP20 in cardiomyocytes was assessed by western blotting. (H) The statistical results of (G) data have been normalized against GAPDH and the mRNA expression levels of Usp20 in cardiomyocytes assessed by PCR. n = 5. (I) Expression of atrial natriuretic peptide (ANP) in cardiomyocytes overexpressing USP20 induced by HG + PA tested by western blotting. (J) The statistical results of (I). n = 5. Data are expressed as the mean ± standard error of the mean (SEM). ***p < 0.001; **p < 0.01; *p < 0.05, p > 0.05, ns: No differences. n: The number of independent biological replicates.

3.2. Cardiomyocyte‐Specific USP20 Deficiency Aggravates Diabetes‐Induced Cardiac Remodeling and Myocardial Dysfunction

Cardiomyocyte‐specific USP20 deficiency (USP20CKO) mice were generated for subsequent experiments (Figure S2A,B). To explore the role of USP20 in DbCM, we successfully established a mouse model of type 2 diabetes by combining a high‐fat diet with STZ injection (Figure 2A). T2MD mice showed increased body weight relative to Control mice. Meanwhile, USP20CKO did not lead to further weight gain compared to USP20^fl/fl mice subjected to the same high‐fat diet and STZ treatment (Figure S3A). Additionally, we measured fasting blood glucose and insulin levels in the mice. Interestingly, the absence of USP20 did not alter HFD‐induced hyperglycemia or insulin resistance (Figure S3B–D). Then, we assessed the impact of USP20 on myocardial function in diabetic mice. Both ejection fraction (EF), fractional shortening (FS) levels and E/A ratio were significantly diminished in diabetic mice. Moreover, the deficiency of USP20 in cardiac tissue further aggravated myocardial dysfunction both in systolic and diastolic function, leading to a further decline in EF, FS values and E/A ratio (Figure 2B–E and Table S2). Serum levels of ANP were elevated in USP20CKO mice compared to their USP20^fl/fl counterparts (Figure 2F). Histopathological analysis revealed that the myocardium of diabetic mice exhibited cardiac dilation, cardiomyocyte hypertrophy and fibrosis. Notably, the absence of USP20 intensified these pathological alterations (Figure 2F–N). Furthermore, expression levels of β‐MyHC and ANP were significantly increased in heart tissue from diabetic mice. This effect was even more pronounced in diabetic USP20CKO mice (Figure 2O,P). Importantly, no baseline phenotypic differences were observed across all parameters between non‐diabetic USP20^fl/fl and USP20CKO mice.

FIGURE 2.

Cardiomyocyte‐specific USP20 deficiency aggravates diabetes‐induced cardiac remodeling and myocardial dysfunction. The mice were stratified into four experimental groups: (1) USP20^fl/fl non‐diabetic controls, (2) USP20 conditional knockout (CKO) non‐diabetic controls, (3) USP20^fl/fl diabetic mice, and (4) USP20 CKO diabetic mice. (A) Schematic diagram depicting the experimental strategy for DbCM models. (B) Representative echocardiography images across experimental groups. (C–E) Quantification of left ventricular ejection fraction (EF) (C) and fractional shortening (FS) (D) measured by M‐mode echocardiography. Ratio between mitral E wave and A wave (E/A) by pulse‐wave Doppler (E). n = 10. (F) Serum levels of ANP in mice from each group. n = 10. (G) Representative macroscopic morphology of heart tissue from each group. (H) Representative H&E staining images of heart tissue from each group (scale bar: 500 μm). (I, J) Representative wheat germ agglutinin (WGA) staining images of heart tissue from each group (scale bar: 100 μm) (I) and statistical results (J). (K, L) Representative Masson staining images of heart tissue from each group (scale bar: 100 μm) (K) and statistical results (L). (M, N) Representative Sirius red staining images of heart tissue from each group (scale bar: 100 μm) (M) and statistical results (N). (O, P) Expression of MyHC and ANP in heart tissue from each group tested by western blotting (O) and the statistical results, data has been normalized against GAPDH (P). n = 10. Data are expressed as the mean ± SEM. ***p < 0.001; **p < 0.01; *p < 0.05, p > 0.05, ns: No differences. n: The number of independent biological replicates.

3.3. USP20 Is Directly Bound to STING

In line with transcriptomic data, comparative Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis and Gene Set Enrichment Analysis (GSEA) revealed that differentially expressed genes were significantly enriched in the oxidative phosphorylation pathway (Figure 3A). Subsequently, we employed co‐immunoprecipitation (Co‐IP) combined with liquid chromatography–tandem mass spectrometry (LC–MS/MS) to identify substrate proteins regulated by USP20 in the context of diabetic cardiomyopathy (DbCM). Through bioinformatics analysis, we selected three protein datasets: those associated with the oxidative phosphorylation pathway, cardiovascular conditions, and diabetes‐related proteomics. Notably, STING emerged as the only overlapping candidate substrate across all three groups (Figure 3B), prompting us to investigate whether an interaction exists between USP20 and STING. Co‐IP assays performed in HL‐1 cells co‐transfected with STING (Figure S4A) and USP20 plasmids demonstrated a direct physical interaction between the two proteins (Figure 3C). This interaction was consistently observed in both cardiomyocytes and myocardial tissue samples (Figure 3D,E). We then examined the specific domains through which USP20 interacts with STING. To this end, we constructed USP20 mutant plasmids targeting three distinct domains: the catalytic USP domain, dual‐specificity phosphatase 1 (DUSP1), and dual‐specificity phosphatase 2 (DUSP2) (Figure 3F). Domain‐focused Co‐IP analyzes revealed that USP20 binds exclusively to STING via its catalytic USP domain, a conclusion supported by the complete loss of interaction in the catalytic USP domain mutant (Figure 3G).

FIGURE 3.

Identification of STING interaction with USP20. (A) KEGG and GSEA of differentially expressed genes in USP20^fl/fl diabetic mice and USP20 CKO diabetic mice. (B) The workflow for USP20 substrate screening. The candidate substrates of USP20 screened by the liquid chromatography tandem mass spectrometry (LC–MS/MS). (C) Co‐immunoprecipitation (Co‐IP) of USP20 and STING in HL‐1 cells co‐transfected with plasmids encoding Flag‐USP20 and His‐STING. Exogenous USP20 was immunoprecipitated by anti‐Flag antibody. (D) Co‐IP of USP20 and STING in cardiomyocytes induced by HG + PA. Endogenous USP20 was immunoprecipitated by anti‐USP20. (E) Co‐IP of USP20 and STING in heart tissue of diabetic mice. Endogenous USP20 was immunoprecipitated by anti‐USP20. (F) Schematic illustration of the USP20 domain. (G) Co‐IP of Flag‐WT‐USP20, Flag‐mut‐USP20, and His‐STING in NIH/3 T3 cells co‐transfected with plasmids of Flag‐WT‐USP20, Flag‐mut‐USP20, and His‐STING. Exogenous WT or mutated USP20 was immunoprecipitated by anti‐Flag antibody. WT, wild type.

3.4. USP20 Facilitates the Degradation of STING Through the Autophagy Pathway

As a catalytically active deubiquitinase, USP20 plays a crucial role in the dynamic regulation of protein stability and functional activity. The subsequent experiments aim to elucidate the impact of USP20 on STING. Initially, we quantified STING protein levels in USP20CKO mice under diabetic conditions. Immunoblot analyzes demonstrated a significant accumulation of STING in the myocardial tissues lacking USP20 compared to the USP20^fl/fl group (Figure 4A,B). Given that higher levels of the STING protein are expected to correlate with increased phosphorylated STING, it might be hypothesized that USP20 naturally enhances STING phosphorylation in cardiomyocytes. To test this hypothesis, we examined the effect of USP20 on STING phosphorylation in cardiomyocytes. The results demonstrated that overexpression of USP20 decreased both the total STING level and the levels of phosphorylated STING (p‐STING) in HL‐1 cells (Figure S5A). Conversely, knockdown of USP20 using si‐USP20 in HL‐1 cells led to an increase in both total STING and p‐STING levels (Figure S5B,C). These findings suggest that USP20 modulates the overall abundance of STING but does not directly influence its phosphorylation activity. We then engineered HL‐1 cells to overexpress USP20 through plasmid transfection and conducted cycloheximide (CHX) chase assays. The results indicated a notable increase in the degradation of the STING protein as CHX treatment duration was extended, particularly evident in cells with elevated levels of USP20 expression (Figure 4C,D) and USP20 does not significantly affect the ubiquitination status of STING (Figure S5D). Following this, we introduced autophagy inhibitors (3‐Methyladenine, 3‐MA and Chloroquine, CQ) and proteasome inhibitors (MG132) to delineate the degradation pathways associated with STING. Findings revealed that STING underwent significant degradation in cardiomyocytes with increased expression of USP20. Furthermore, this specific degradation mediated by USP20 was effectively inhibited by both 3‐MA and CQ without being influenced by MG132 (Figure 4E,F). In subsequent experiments, HL‐1 cells were transfected with either plasmids expressing USP20 or siRNA targeting USP20 to further validate our hypothesis. Immunofluorescence co‐localization analysis revealed that overexpression of USP20 enhanced the co‐localization of STING with the autophagosome marker LC3, whereas siRNA‐mediated knockdown of USP20 reduced their co‐localization (Figure 4G,H). We then employed the autophagy inducer Earle's Balanced Salt Solution (EBSS) to assess functional outcomes. Consistently, enhancement of autophagic flux via EBSS synergistically promoted STING clearance when combined with overexpression of USP20 (Figure 4I,J), while silencing USP20 negated EBSS‐induced degradation of STING (Figure 4K,L). Collectively, these data establish that USP20 orchestrates autophagic‐lysosomal degradation of STING through mechanisms dependent on its deubiquitination activity.

FIGURE 4.

USP20 facilitates the degradation of STING through the autophagy pathway. (A, B) Expression of STING protein in heart tissue of diabetic USP20^fl/fl and diabetic USP20 CKO mice tested by western blotting and statistical results, data has been normalized against GAPDH (B). n = 6. (C, D) Expression of STING protein in HL‐1 cells transfected with Flag‐USP20 plasmids or Flag‐EV plasmids and performed cycloheximide (CHX) for different times and statistical results, data has been normalized against GAPDH (D). n = 3. (E, F) Protein level of STING in HL‐1 subjected to proteasome inhibitor (MG132, 10 mM) or autophagy inhibitors (3‐MA, 10 mM and CQ, 50 mM) and statistical results, data has been normalized against GAPDH (F). (G, H) Double immunofluorescence staining for STING (red), LC3 (green) in the HL‐1 cells treated with Flag‐USP20(G) or si‐USP20(H). Merged images (orange) showing colocalization; scale bar, 20 μm. (I, J) Expression of STING protein in cardiomyocytes transfected with Flag‐USP20 plasmids or Flag‐EV plasmids and treated with EBSS (G) and statistical results, data have been normalized against GAPDH (H). n = 3. (K, L) Expression of STING protein in cardiomyocytes transfected with si‐USP20 or si‐NC and treated with EBSS (I) and statistical results, data has been normalized against GAPDH (J). n = 3. Data are expressed as the mean ± SEM. ***p < 0.001; **p < 0.01; *p < 0.05, p > 0.05, NS: No differences. n: The number of independent biological replicates.

3.5. USP20 Regulates the Degradation of STING by Deubiquitinating p62 via Its Active Site C154

Next, we sought to elucidate how USP20 regulates STING protein stability. Co‐IP assays were conducted, confirming that USP20 interacts with p62 but not with LC3 (Figure 5A). Furthermore, we demonstrated that USP20, p62, and STING constitute a trimeric complex (Figure S6A–C). Notably, knockdown of p62 abolished USP20‐mediated degradation of STING in HL‐1 cardiomyocytes under HG/PA stress, thereby establishing p62 as a crucial adaptor within this regulatory pathway (Figures 5B,C and S6D). Considering the canonical deubiquitinase activity associated with USP20, we examined the dynamics of ubiquitination on p62. The results indicated that overexpression of USP20 resulted in a significant reduction in the amount of ubiquitin (UB) conjugated to p62 (Figure 5D). Previous studies have shown that the UBA domain of p62 is capable of binding both K63‐linked and K48‐linked ubiquitin chains, exhibiting a higher affinity for K63 linkages [22, 23]. Consequently, our study detected K63‐linked ubiquitin chains. Our findings demonstrated that overexpression of USP20 was associated with a marked decrease in ubiquitin molecules linked to p62 in cells transfected with a plasmid designed to retain only K63 activity (UB‐K63) (Figure 5E), implying that USP20 specifically targets and removes K63‐linked ubiquitination from the p62 protein. Subsequently, we investigated specific sites where USP20 deubiquitinates p62. The cysteine residue at position 154 (C154) and histidine residue at position 645 (H645) were proposed as catalytic residues critical for the deubiquitinating activity of USP20. Accordingly, mutant variants of USP20 plasmids, C154A (mutation converting cysteine to alanine at C154) or H645A (mutation substituting histidine with alanine at H645), were generated. While both mutants retained their capacity to bind STING (Figure 5F), only the USP20C154A mutation completely abrogated USP20's ability to enhance EBSS‐induced clearance of STING (Figure 5G,H). A consistent trend was observed in USP20‐NIH/3 T3 cells transferred with Flag‐USP20 and Flag‐USP20C154A (Figures 5I,J and S6E). Additionally, HL‐1 cells were co‐transfected with Flag‐USP20WT, Flag‐USP20C154A or an empty vector along with HA‐UB‐K63 plasmids. Co‐IP analysis demonstrated that the USP20C154A mutation substantially increased the levels of HA‐UB‐K63 on p62 compared to USP20WT (Figure 5K), indicating a loss of deubiquitinating activity for the p62 protein associated with the USP20C154A mutation. The C154 residue in USP20 is critical for its deubiquitination function towards p62. These findings suggest that USP20 enzymatically removes K63‐linked ubiquitin chains from p62 through its active site at C154, thereby facilitating the autophagic sorting of STING for lysosomal degradation mediated by p62 (Figure 5L).

FIGURE 5.

USP20 regulates the degradation of STING by deubiquitinating p62 via its active site C154. (A) Co‐IP of USP20, p62, and LC3 in HL‐1 cells transfected with plasmids encoding Flag‐USP20. Exogenous USP20 was immunoprecipitated by anti‐Flag antibody. (B, C) The expression of STING in HL‐1 cells co‐transfected with plasmids encoding Flag‐USP20 and si‐p62 (B) and statistical results, data has been normalized against GAPDH (C). n = 5. (D) Immunoprecipitation of p62 was performed in HL‐1 cells co‐transfected with HA‐ubiquitin (UB), His‐p62, and Flag‐USP20. The ubiquitinated p62 was detected by immunoblotting using a His‐specific antibody to elucidate the ubiquitination patterns regulated by USP20. (E) Immunoprecipitation of p62 was performed in HL‐1 cells co‐transfected with HA‐K63, His‐p62, and Flag‐USP20. The ubiquitinated p62 was detected by immunoblotting using a His‐specific antibody to elucidate the ubiquitination patterns regulated by USP20. (F) Immunoprecipitation of STING in HL‐1 cells that were co‐transfected with plasmids encoding His‐STING, Flag‐USP20‐C154A, and Flag‐USP20‐H645A plasmids. The ubiquitinated form of STING was detected through immunoblotting utilizing a His‐specific antibody. (G, H) Expressions of STING protein in HL‐1 cells overexpressing USP20‐WT, USP20‐C154A, and USP20‐H645A plasmids and treated with EBSS (G) and statistical results, data have been normalized against GAPDH (H). n = 3. (I, J) Expressions of STING protein in gUSP20‐NIH/3 T3 cells overexpressing USP20‐WT and USP20‐C154A plasmids and treated with EBSS (I) and statistical results, data have been normalized against GAPDH (J). n = 3. (K) Immunoprecipitation of p62 in HL‐1 cells that co‐transfected with plasmids encoding HA‐K63, His‐p62, Flag‐USP20‐WT, and Flag‐USP20‐C154A. Ubiquitinated p62 was detected by immunoblotting using an His‐specific antibody. (L) Schematic illustration of USP20 regulating the STING degradation. Data are expressed as the mean ± SEM. ***p < 0.001; **p < 0.01; *p < 0.05. n: The number of independent biological replicates.

3.6. USP20 Attenuates Inflammatory Response via Modulation of the STING Signaling Pathway

The STING pathway is known to mediate pyroptosis and inflammatory responses, contributing significantly to the pathogenesis of diabetic cardiomyopathy (DbCM) [24]. To investigate the regulatory effects of USP20 within this context, we systematically analyzed STING pathway activation and levels of inflammatory factors under varying states of USP20 expression. Overexpression of USP20 substantially attenuated the HG/PA‐induced upregulation of STING protein and inhibited phosphorylation of the downstream effectors TBK1 and IRF3 (Figure 6A,B). Consistent with this, USP20 overexpression markedly reduced HG/PA‐induced mRNA levels of multiple inflammatory cytokines in cardiomyocytes, including Il‐1β, Il‐6, Il‐18, Tnf‐α, Cxcl1, Ccl2, Icam1, and Vcam1 (Figure 6C). Conversely, knockdown of USP20 exacerbated STING pathway activation, enhancing both STING expression and TBK1/IRF3 phosphorylation (Figure 6D,E). This enhancing effect was also reflected in the transcriptional activity of inflammatory factors, where USP20 deficiency amplified HG/PA‐induced gene expression (Figure 6F). Importantly, in vivo validation using USP20CKO mice corroborated these findings by demonstrating significantly elevated mRNA levels for inflammatory factors within the heart tissue from diabetic USP20 conditional knockout mice (Figure 6G).

FIGURE 6.

USP20 attenuates inflammatory response via modulation of the STING signaling pathway. (A, B) Expression of STING, TBK1, p‐TBK1, IRF3, and p‐IRF3 in cardiomyocytes overexpressing USP20 induced by HG + PA tested by western blotting (A) and the statistical results, data has been normalized against GAPDH (B). n = 5. (C) The mRNA expression levels of Il‐1β, Il‐6, Il‐18, Tnf‐α, Cxcl1, Ccl2, Icam1, and Vcam1 in cardiomyocytes overexpressing USP20 induced by HG + PA. n = 5. (D, E) Expression of STING, TBK1, p‐TBK1, IRF3 and p‐IRF3 in cardiomyocytes transfected with si‐USP20 or si‐NC and induced by HG + PA tested by western blotting (D) and the statistical results, data have been normalized against GAPDH (E). n = 5. (F) The mRNA expression levels of Il‐1β, Il‐6, Il‐18, Tnf‐α, Cxcl1, Ccl2, Icam1 and Vcam1 in cardiomyocytes transfected with si‐USP20 or si‐NC and induced by HG + PA. n = 5. (G) The mRNA expression levels of Il‐1β, Il‐6, Il‐18, Tnf‐α, Cxcl1, Ccl2, Icam1, and Vcam1 in heart tissues of USP20^fl/fl non‐diabetic controls, USP20 conditional knockout (CKO) non‐diabetic controls, USP20^fl/fl diabetic mice, and USP20 CKO diabetic mice. Data are expressed as the mean ± SEM. ***p < 0.001; **p < 0.01; *p < 0.05, p > 0.05, ns: No differences. n: The number of independent biological replicates.

3.7. USP20 Ameliorates DbCM Through Negative Regulation of STING

In the subsequent experiments, we administered AAV9 vectors that included a cardiac‐specific promoter cTNT and encoded USP20 (AAV9‐cTnT‐USP20^oe) to achieve cardiomyocyte‐specific overexpression in db/db mice (Figure S7A). C‐176, an inhibitor of STING, was utilized in this study to further elucidate the role of USP20 in DbCM (Figure 7A). Our findings indicated that compared to control db/db mice, those with USP20 overexpression in cardiomyocytes or treated with C‐176 demonstrated a significant improvement in cardiac systolic and diastolic function, as evidenced by elevated levels of EF and FS and reduced E/A ratio (Figure 7B–E and Table S3). We measured body weight, fasting blood glucose and insulin levels in the mice. Interestingly, the inhibition of STING or the overexpression of USP20 did not alter HFD‐induced hyperglycemia or insulin resistance (Figure S7B–E). Moreover, the increase in serum levels of IL‐1β and atrial natriuretic peptide (ANP) induced by diabetes was significantly reduced following either USP20 overexpression or C‐176 treatment. However, it is noteworthy that the overexpression of USP20 within the cardiomyocytes of C‐176‐treated diabetic mice did not confer any additional cardioprotective effect compared to diabetic mice receiving only C‐176 treatment (Figure 7F,G). Additionally, gross cardiac morphology and histological analyzes using H&E staining revealed that both USP20 overexpression and C‐176 treatment ameliorated cardiac enlargement and myocyte swelling associated with diabetes (Figure 7H). Furthermore, WGA staining along with Masson's trichrome and Sirius red staining indicated a notable reduction in interstitial collagen deposition and myocardial fibrosis triggered by diabetes. This protective effect was observed following either USP20 overexpression or C‐176 administration. Nevertheless, no further benefits were noted in diabetic mice exhibiting both USP20 overexpression and treated with C‐176 when assessed against those receiving only C‐176 therapy (Figure 7I–N). Moreover, both approaches, USP20 overexpression and treatment with C‐176, effectively diminished expression levels of MyHC and ANP within heart tissue from diabetic mice (Figure 7O). These findings provide mechanistic evidence that USP20 predominantly mediates its cardioprotective effects in DbCM through the negative regulation of the STING signaling cascade.

FIGURE 7.

USP20 ameliorates DbCM through negative regulation of STING. The mice were stratified into four experimental groups: (1) db/db + Veh + AAV9‐cTnT‐EV, (2) db/db + Veh + AAV9‐cTnT‐USP20^oe, (3) db/db + C‐176 + AAV9‐cTnT‐EV, and (4) db/db + C‐176 + AAV9‐cTnT‐USP20^oe. (A) Schematic diagram depicting the experimental strategy for STING inhibition and USP20 activation. (B) Representative echocardiography across experimental groups. (C–E) Quantification of cardiac EF (C) and FS (D) of mice from each group measured by M‐mode echocardiography. Ratio between mitral E wave and A wave (E/A) by pulse‐wave Doppler (E). n = 10. (F, G) Serum levels of IL‐1β (F) and ANP (G) in mice from each group. n = 10. (H) Representative macroscopic morphology of heart tissue from each group. (I) Representative HE staining images of heart tissue from each group (scale bar: 500 μm). (J, K) Representative WGA staining images of heart tissue from each group (scale bar: 100 μm) (J) and statistical results (K). (L, M) Representative Masson staining images of heart tissue from each group (scale bar: 100 μm) (L) and statistical results (M). (N, O) Representative Sirius red staining images of heart tissue from each group (scale bar: 100 μm) (N) and statistical results (O). (P) The mRNA expression levels of MyHC and ANP in heart tissue from each group n = 10. Data are expressed as the mean ± SEM. ***p < 0.001; **p < 0.01; *p < 0.05. n: The number of independent biological replicates.

4. Discussion

The present study elucidates a previously unrecognized cardioprotective mechanism orchestrated by USP20 in DbCM. In our study, HFD/STZ induced myocardial USP20‐deficient diabetic mice and USP20‐cardiomyocyte‐specific overexpression in db/db mice were used to identify the role of USP20 in DbCM, suggesting that activated USP20 exerts protective effects on the hearts of DbCM mice. USP20 promotes the degradation of STING through the autophagy pathway by deubiquitinating p62 via its active site C154, thereby alleviating the myocardial inflammation and improving ventricular remodeling and heart failure induced by diabetes. These discoveries fundamentally reshape our understanding of ubiquitination code reprogramming in metabolic heart disease and provide actionable therapeutic targets for clinical translation.

Ubiquitination and deubiquitination are critical for maintaining cellular homeostasis, with deubiquitinating enzymes (DUBs) fine‐tuning protein stability and function [25]. USP20, a deubiquitinating enzyme governing substrate specificity in protein degradation, participates in modulating various cellular processes [10]. USP20 exerts its effect by deubiquitinating substrate proteins and regulating protein function. Emerging evidence has demonstrated that DUBs are involved in the pathogenesis of cardiovascular diseases [26, 27, 28]. Our study identifies USP20 as a novel molecular determinant in DbCM, through its regulation of the STING. The cGAS‐STING axis, traditionally recognized as a cytosolic DNA sensor mediating antiviral immunity [29], exhibits context‐dependent activation in metabolic disorders. Mechanistically, STING recruits TANK‐binding kinase 1 (TBK1) upon binding cyclic GMP‐AMP (cGAMP), triggering IRF3 phosphorylation and NF‐κB nuclear translocation, and leading to inflammatory response [30]. In addition to its key role in antiviral immune response, the cGAS‐STING signaling pathway has also been shown to be involved in several other important biological processes, including cellular senescence [31, 32], neurological disorders [33], and myocardial infarct‐associated inflammation [34]. Moreover, the cGAS‐STING signaling pathway was activated by increased mtDNA in the cytosol of DbCM mouse hearts [35], which mediated pyroptosis and the inflammatory response, leading to myocardial injury [24]. Inhibition of the cGAS‐STING signaling pathway ameliorated DbCM‐induced myocardial remodeling and heart failure [36, 37]. The present study demonstrated that USP20 interacted with STING and facilitated its degradation, consequently alleviating the myocardial inflammation and improving ventricular remodeling and heart failure induced by diabetes.

The intracellular degradation of STING is orchestrated through two distinct proteostatic pathways: lysosome‐dependent processing via endosomal trafficking complexes, and autophagosome‐lysosome fusion‐mediated clearance [38]. In the present study, we found that USP20 mediated the degradation of STING through the autophagy pathway. Autophagy and the ubiquitin‐proteasome system are the two primary quality control pathways that maintain cellular homeostasis. Within the cellular quality control network, the ubiquitin‐proteasome system and autophagy‐lysosomal pathway exhibit functional crosstalk mediated by p62/SQSTM1‐ a multidomain scaffold protein containing ubiquitin‐associated (UBA) and LC3‐interacting region (LIR) domains that enable simultaneous recognition of polyubiquitinated substrates and autophagosomal membranes [39, 40]. Notably, p62 serves as the molecular linchpin for STING's autophagic degradation, with its phosphorylation by TBK1 enhancing binding affinity to STING, and promoting the trafficking of STING to autophagosomes [41]. Our research reveals an additional layer of regulation involving the ubiquitination state of p62 itself. USP20 modifies p62's activity by deubiquitinating it at the active site cysteine 154, a critical function that is lost in the C154A mutant. We propose a model wherein USP20, p62, and STING form a functional ternary complex. The deubiquitination of p62 by USP20 may potentially enhance p62's oligomerization or alter its conformational state, increasing its efficacy as an autophagy receptor for STING. This hypothesis, consistent with reports that deubiquitination promotes p62 oligomerization and autophagic flux, offers a more integrated view of the USP20‐p62‐STING regulatory axis. It suggests that USP20 acts as a molecular switch, reprogramming p62 from a static scaffold to an active vehicle for STING disposal.

Our study is the first to demonstrate that USP20 regulates STING stability by modulating the ubiquitination status of p62, thereby establishing a novel deubiquitination‐dependent checkpoint in the autophagy‐mediated control of innate immune signaling in DbCM. This finding not only distinguishes USP20 from other DUBs implicated in cardiovascular disease but also expands the functional scope of DUBs in metabolic stress responses. However, the present study is subject to certain limitations. First, the clinical relevance of USP20 in diabetic cardiomyopathy requires further validation due to the limited size of analyzed human samples. Expanding to a larger, multi‐center cohort with detailed clinical stratification will be necessary to definitively establish the correlation between USP20 expression and disease severity. Second, although we have delineated the USP20‐p62‐STING signaling axis, key aspects of its regulatory network remain unclear. These include whether USP20 modulates upstream activators such as mtDNA‐cGAS signaling [42], its in vivo effect on autophagic flux, and its potential role in promoting p62 oligomerization to activate autophagy. Addressing these questions will require integrated approaches, including live‐cell imaging, detailed protein‐interaction studies, and tissue‐specific genetic models.

From a translational perspective, targeting the USP20‐p62‐STING axis presents both opportunities and challenges. The cardioprotective effect of USP20 overexpression suggests therapeutic potential. Cardiac‐specific delivery via adeno‐associated virus (AAV) vectors could be a viable strategy to maximize cardiac benefit while minimizing systemic effects [43]. However, the multi‐faceted roles of USP20 in other organs and pathways necessitate careful evaluation. Global USP20 activation or inhibition could lead to unintended consequences, underscoring the need for tissue‐specific targeting approaches. Future efforts should also aim to develop small‐molecule activators of USP20's deubiquitinase activity towards p62 as a more precise pharmacological intervention.

In summary, this study delineates a novel cardioprotective mechanism wherein USP20, by deubiquitinating p62, facilitates the autophagic degradation of STING, thereby alleviating diabetic cardiomyopathy. This work not only deepens our mechanistic understanding of ubiquitin signaling in metabolic heart disease but also highlights the intricate crosstalk between deubiquitination and selective autophagy. While further research is needed to address the outlined limitations and hypotheses, this USP20‐p62‐STING axis represents a promising new frontier for developing targeted therapies against DbCM.

5. Conclusions

Our study has confirmed that cardiomyocyte‐specific USP20 promoted the degradation of STING through the autophagy pathway by deubiquitinating p62 via its active site C154, thereby alleviating the myocardial inflammation and improving ventricular remodeling and heart failure induced by diabetes, which provides compelling preclinical rationale for targeting USP20‐mediated deubiquitination as a next‐generation therapeutic strategy.

Author Contributions

Shanshan Dai, Zhenfeng Cheng, and Zhouqing Huang contributed to the literature search and study design. Yixin Zhou, Miaomiao Ying, Baozhen Qi, Yunxuan Chen, Jiahui Lin, Yucong Zhang, Zexin Yang, Ziyi Huang, and Jun Wu performed the experiments and analyzed the data. Weihong Lin, Chunwu Zhang, Xueli Cai, and Weijian Huang provided technical help. Yixin Zhou and Shanshan Dai participated in the drafting of the article. All authors agree to be accountable for all aspects of work ensuring integrity and accuracy.

Funding

This work was supported by MOST | National Natural Science Foundation of China (NSFC) (Grants 82202380, 82271336 and 80225018), MOST | NSFC | NSFC‐Zhejiang Joint Fund | 浙江省科学技术厅 | Natural Science Foundation of Zhejiang Province (ZJNSF) (Grant LQ23H310005), 市科技局 | Science and Technology Plan Project of Wenzhou Municipality (温州市科技计划项目) (Grant 2024ZY0050), 市科技局 | Science and Technology Plan Project of Wenzhou Municipality (温州市科技计划项目) (Grant Y2023104), Special Funding Project for Characteristic Directional S‐Disciplines of The First Affiliated Hospital of Wenzhou Medical University (Grant wyyy‐2025S02) and Zhejiang Provincial Science and Technology Innovation Program (New Young Talent Program) for College Students (Grant 2024R413C089).

Ethics Statement

All experimental procedures were conducted in accordance with the ethical regulations and guidelines established by the US National Institutes of Health. The animal welfare conditions received approval from the Laboratory Animal Ethics Committee at the First Affiliated Hospital of Wenzhou Medical University (approval No. WYYY‐IACUC‐AEC‐2025‐018). All procedures involving human heart specimens were reviewed and formally approved by the Ethics Committee of the First Affiliated Hospital of Wenzhou Medical University (Approval No. KY2022‐156).

Consent

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Data S1: fsb271540‐sup‐0001‐Supinfo.docx.

Zhou Y., Ying M., Qi B., et al., “Cardiomyocyte‐Derived USP20 Attenuates Diabetic Cardiomyopathy by Facilitating the Degradation of STING and Mitigating STING‐Mediated Inflammation,” The FASEB Journal 40, no. 3 (2026): e71540, 10.1096/fj.202503913R.

Data Availability Statement

All data needed to evaluate the conclusions in this study are presented in this manuscript or the Supporting Information. The materials described in this study are either commercially available or available upon reasonable request from the corresponding authors.