Deubiquitinase USP13 alleviates doxorubicin-induced cardiotoxicity through promoting the autophagy-mediated degradation of STING

Abstract

Doxorubicin (Dox) is an anthracycline drug widely applied in various malignancies. However, the fatal cardiotoxicity induced by Dox limits its clinical application. Post-transcriptional protein modification via ubiquitination/deubiquitination in cardiomyocytes mediates the pathophysiological process in Dox-induced cardiotoxicity (DIC). In this study, we aimed to clarify the regulatory role and mechanism of a deubiquitinating enzyme, ubiquitin-specific peptidase 13 (USP13), in DIC. RNA-seq analysis and experimental examinations identified that cardiomyocyte-derived USP13 positively correlated with DIC. Mice with cardiac-specific deletion of USP13 were subjected to Dox modeling. Adeno-associated virus serotype 9 (AAV9) carrying cTNT promoter was constructed to overexpress USP13 in mouse heart tissues. Cardiomyocyte-specific knockout of USP13 exacerbated DIC, while its overexpression mitigated DIC in mice. Mechanistically, USP13 deubiquitinates the stimulator of interferon genes (STING) and promotes the autolysosome-related degradation of STING, subsequently alleviating cardiomyocyte inflammation and death. Our study suggests that USP13 serves a cardioprotective role in DIC and indicates USP13 as a potential therapeutic target for DIC treatment.

Keywords: Autophagy; Cardiomyocyte; Cardiotoxicity; Deubiquitinating enzyme; Doxorubicin; Inflammation; STING; USP13.

Graphical abstract

Figure 1

Cardiomyocyte USP13 expression is increased in Dox-induced cardiomyopathy. (A, B) Heatmaps illustrate the DUBs' expression levels in datasets GSE23598 (Dox-induced mouse heart tissues) and GSE120895 (cardiac tissues from patients with dilated cardiomyopathy). (C) The level of lactate dehydrogenase (LDH) in cell supernatants. HL-1 cells were transfected with 8 different DUBs before the addition of Dox (1 μmol/L, 24 h) (n = 6). (D) Comprehensive combined GEO database and cell viability analysis for identification of USP13 as a regulator in DIC. Red denotes upregulated DUBs in GSE23598, blue represents upregulated DUBs in GSE120895, and green indicates results from functional screening. (E, F) The mRNA or protein level of USP13 in Dox-induced mice heart tissue (n = 6). (G) The single-cell sequencing data reveals the distribution of USP13 in mice heart. (Single-cell suspensions from 3 to 4 hearts were pooled as 1 sample). (H) The protein levels of USP13 in different cells. (I) The protein level of USP13 in HL-1 cells with Dox (1 μmol/L) for 0–24 h. (J) Immunofluorescence staining was performed on heart sections from Dox mice to detect USP13 (green) along with α-actin or vimentin (both red). Merged images indicate co-localization (yellow). Data are presented as mean ± SEM; ∗∗P < 0.01.

Figure 2

Cardiomyocyte specific knockout of USP13 exacerbates Dox-induced cardiac dysfunction and injury in vivo. USP13^fl/fl and USP13CKO mice were intraperitoneally injected with either PBS or Dox (15 mg/kg, 3 times/week for 2 weeks). DIC was assessed at 4 weeks after Dox injection. (A) Representative M-mode images of indicated groups. (B) Ejection fraction (EF) and fractional shortening (FS) in different groups (n = 6). (C–E) Serum atrial natriuretic peptide (ANP), LDH, and creatine kinase-MB (CK-MB) in different groups (n = 6). (F) Whole heart images of indicated groups. (G) HE staining of cardiac sections. (H, I) Representative wheat germ agglutinin (WGA) fluorescence staining images and quantification showing relative cardiomyocyte size (n = 6). (J, K) Masson staining of cardiac sections and quantification of collagen (n = 6). (L, M) TUNEL staining of cardiac sections and quantification of TUNEL-positive cells (n = 6). Data are presented as mean ± SEM; ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001.

Figure 3

Cardiomyocyte-specific overexpression of USP13 alleviates cardiac dysfunction and injury induced by Dox. (A) Schematic diagram outlining animal experiments: mice were injected with AAV9 viruses (2 × 10¹¹ v.g., 100 μL) after 2 weeks of injection of Dox (15 mg/kg, 3 times/week). The mice were harvested at 6 weeks after Dox injection. (B) Protein level of Flag–USP13 in heart tissues (n = 6). (C) Representative M-mode images of indicated groups. (D) EF and FS in different groups (n = 6). (E, F) Biochemical analysis of serum LDH, and CK-MB (n = 6). (G) Whole heart images of indicated groups. (H) HE staining of cardiac sections. (I, J) Representative WGA fluorescence staining images and quantification showing relative cardiomyocyte size (n = 6). (K, L) Masson staining of cardiac sections in heart sections and quantification of collagen (n = 6). (M, N) TUNEL staining of cardiac sections and quantification of TUNEL-positive cells (n = 6). Data are presented as mean ± SEM; ns = no significance, ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001.

Figure 4

USP13 directly binds to STING and promotes the autophagy-dependent degradation of STING. (A) Schematic representation of the integration of Co-IP with LC–MS/MS for USP13 substrate identification. (B) The Co-IP analysis of USP13 and STING in Dox-treated HL-1 cells (1 μmol/L, 24 h). (C) Co-IP of USP13 and STING in heart tissues upon Dox treatment. (D) Co-IP of USP13 and STING in NIH/3T3 cells co-transfected with Flag–USP13 and His–STING plasmids. (E) Immunoblot of USP13 and STING in NIH/3T3 cells transfected with Flag–USP13. (F) The mRNA levels of Usp13 and Sting in NIH/3T3 cells transfected with Flag–USP13 (n = 6). (G) Representative immunoblot images of USP13, p-STING, and STING in HL-1 cells with Spautin-1 (10 μmol/L, 2 h) and then Dox. (H) Representative immunoblot images of USP13, p-STING, and STING in HL-1 cells transfected with Flag–EV or Flag–USP13 plasmids and then Dox stimulation. (I) Representative immunoblot images of Flag–USP13 and STING in HL-1 cells transfected with either Flag–EV or Flag–USP13 plasmids, and subjected to treatment with Mock, MG132 (10 μmol/L), and bafilomycin A1 (Baf A1, 0.2 μmol/L). (J, K) Representative immunoblot images of Flag–USP13 and STING in HL-1 cells transfected with Flag–EV or Flag–USP13 plasmids, and exposed to EBSS for 0, 3, 6, and 9 h (down), quantification of STING (up) (n = 6). (L) Co-IP of STING, P62 and Flag–USP13 in NIH/3T3 cells co-transfected with His–STING, HA–P62 and Flag–USP13 plasmids. Data are presented as mean ± SEM; ns, no significance, ∗P < 0.05, ∗∗∗P < 0.001.

Figure 5

USP13 modulates K63-linked deubiquitination of STING via the C343 site. (A) NIH/3T3 cells were co-transfected with His–STING, HA–Ub or HA–K63 along with Flag–USP13, followed by treatment with 0.2 μmol/L Baf A1 for 8 h. Lysates were then subjected to Co-IP with anti-His, followed by immunoblotting with antibodies against HA, STING, and Flag. (B) A schematic illustration depicting the USP13 active site mutation (C343). (C) Co-IP of Flag–USP13 and STING in NIH/3T3 cells co-transfected with Flag–EV, Flag–USP13–WT, Flag–USP13–C343A, and His–STING plasmids. (D) NIH/3T3 cells were transfected with His–STING and HA–Ub along with either Flag–USP13–WT or Flag–USP13–C343A, followed by treatment with 0.2 μmol/L Baf A1 for 8 h. (E, F) Representative immunoblot images of Flag–USP13 and STING in HL-1 cells transfected with either Flag–USP13–WT or Flag–USP13–C343A plasmids, and exposed to EBSS for 0, 3, 6, and 9 h, along with quantification of STING. (G) Illustrative schematic outlines that USP13 regulates K63-linked deubiquitination of STING via its active site C343 and subsequently promotes the degradation of STING via P62-mediated autophagy. Data are presented as mean ± SEM; ∗∗P < 0.01.

Figure 6

USP13 attenuates Dox-induced cardiomyocyte death and inflammation via negatively regulating STING. (A–D) HL-1 cells were transfected with the USP13 plasmid 24 h subsequent to the 1 μmol/L Dox. (A) Representative photographs stained with propidium iodide (PI) and quantitative data for PI positive nuclear (n = 3). (B) The release of LDH and cell viability in cells (n = 6). (C) Representative immunoblot images of p-TBK1 and TBK1, along with quantification of p-TBK1/TBK1 (n = 6). (D) The mRNA levels of Tnf, Il6, and Ifnb1 in each group (n = 6). (E–H) HL-1 cells were treated with the Spautin-1 2 h before exposure to 1 μmol/L Dox. (E) Representative photographs stained with PI and quantitative data for PI positive nuclear (n = 3). (F) The release of LDH and cell viability in cells (n = 6). (G) Representative immunoblot images of p-TBK1 and TBK1, along with quantification of p-TBK1/TBK1 (n = 6). (H) The mRNA levels of Tnf, Il6 and Ifnb1 in each group (n = 6). Data are presented as mean ± SEM; ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001.

Figure 7

USP13 negatively regulates Dox-induced STING activation and inflammation in vivo. (A–C) Mice were injected with AAV9 viruses (2 × 10¹¹ v.g., 100 μL) after 2 weeks of injection of Dox (15 mg/kg, 3 times/week). The mice were harvested at 6 weeks after Dox injection. (A, B) Protein levels of p-STING, STING, p-TBK1, and TBK1 in heart tissue, along with the quantification (n = 6). (C) The mRNA levels of Tnf, Il6 and Ifnb1 (n = 6). (D–F) USP13^fl/fl and USP13CKO mice were intraperitoneally injected with either PBS or Dox (15 mg/kg, 3 times/week for 2 weeks). Cardiotoxicity was assessed at 4 weeks after Dox injection. (D, E) The protein levels of p-STING, STING, p-TBK1 and TBK1 in heart tissue, along with the quantification (n = 6). (F) The mRNA levels of Tnf, Il6 and Ifnb1 (n = 6). (G) The schematic diagram illustrates the principal findings of this study. Data are presented as mean ± SEM; ∗∗P < 0.01, ∗∗∗P < 0.001.