Mitochondrial Amount Determines Doxorubicin-Induced Cardiotoxicity in Cardiomyocytes

Abstract

Doxorubicin, an anthracycline commonly used for treating cancer patients, is known for its cardiotoxic side-effects. Although dose-dependent, but susceptibility remains variable among patients, and childhood-exposure-adult-onset remains challenging. Besides topoisomerase toxicity, Doxorubicin is also toxic to the mitochondria yet the underlying late onset mechanism remains elusive. Here, it is observed that the mitochondrial copy number in PBMCs of patients treated with anthracycline chemotherapy is negatively correlated with the change in plasma BNP levels after treatment. Isogenic hiPSC-CMs are generated with high, norm, and low mitochondrial copy numbers using mitochondrial transplantation and the YFP-Parkin system. Remarkably, lower mitochondria copy number translates to lower IC₅₀, suggesting increased susceptibility. Mitochondria supplementation by intramyocardial injection prevents doxorubicin induced heart failure. Mechanistically, doxorubicin treatment leads to mPTP opening and mitochondrial DNA (mtDNA) leakage. This mtDNA leakage event activates the cGAS-STING pathway and drives inflammation and myocardial senescence. Cardiomyocyte-specific knockout of Sting (Myh6-Cre/Sting^flox/flox; Sting^CKO) and over expression of mitochondrial tagged DNase1 in mice partially rescue doxorubicin-induced cardiac dysfunction. In conclusion, the work establishes a negative correlation between cardiomyocyte mitochondrial copy number and doxorubicin toxicity. Molecularly, it is demonstrated that mtDNA leakage activates cGAS-STING pathway and accelerates myocardial dysfunction. These insights offer new co-administration strategies for cancer patients.

Keywords: cGAS‐STING; doxorubicin induced cardiotoxicity; mitochondrial amount; mitochondrial transplantation; senescence.

Figure 1

Patients with a higher mitochondrial copy number may have a better prognosis in anthracycline‐based chemotherapy. a) Correlation between doxorubicin dosage and plasma BNP levels in patients (n = 11). b) Correlation between doxorubicin dosage and plasma cTnT levels in patients (n = 11). c) Correlation between doxorubicin dosage and plasma cTnI levels in patients (n = 11). d) Correlation between mitochondrial DNA copy number and plasma BNP levels (n = 11). e) Correlation between mitochondrial DNA copy number and plasma TnT levels (n = 11). f) Correlation between mitochondrial DNA copy number and plasma TnI levels (n = 11).

Figure 2

HiPSC‐CMs with varying mitochondrial DNA copy numbers exhibit different levels of doxorubicin tolerance. a) Representation of cells with varying mitochondrial DNA copy numbers. b) Confocal microscopy shows hiPSC‐CMs with three distinct mitochondrial DNA copy numbers, TOM20 (yellow), cTnT (purple), YFP (green), nuclear (blue), scale bar 10 µm. c) qPCR showing differences in mitochondrial DNA copy numbers among three distinct hiPSC‐CMs types, n = 3, quantification values are expressed as mean ± SEM. d) Seahorse analysis showing differences in basal oxygen consumption among hiPSC‐CMs with three distinct mitochondrial DNA copy numbers, n = 6, quantification values are expressed as mean ± SEM. e) hiPSC‐CMs with three distinct mitochondrial DNA copy numbers show significant differences in maximal oxygen consumption and basal respiratory oxygen consumption, n = 6, quantification values are expressed as mean ± SEM. f,g) hiPSC‐CMs with higher mitochondrial DNA copy numbers exhibit stronger beating capacity, n = 30, quantification values are expressed as mean ± SEM. h) Apoptosis curves show that hiPSC‐CMs with fewer mitochondrial DNA copy numbers have lower doxorubicin tolerance, n = 3, quantification values are expressed as mean ± SEM, TUNEL assay after doxorubicin treated 24 h. i) TMRM staining shows that after doxorubicin treatment, hiPSC‐CMs with more mitochondria maintain mitochondrial membrane potential (orange), whereas hiPSC‐CMs with fewer mitochondria exhibit decreased membrane potential, doxorubicin 1 µM, 24 h, scale bar 100 µm, n = 30, quantification values are expressed as mean ± SEM. j) mito‐SOX staining shows that after doxorubicin treatment, hiPSC‐CMs with more mitochondria produce less mitochondrial reactive oxygen species (superoxide, red), while hiPSC‐CMs with fewer mitochondria produce more mitochondrial superoxide, doxorubicin 1 µM, 24 h, scale bar 100 µm, n = 30, quantification values are expressed as mean ± SEM. k) RT‐qPCR analysis revealed that following doxorubicin treatment 1 µM, 24 h, mito^high cells exhibited reduced cytosolic mtDNA leakage, n = 2, quantification values are expressed as mean ± SEM.

Figure 3

Doxorubicin induces mitochondrial DNA release into the cytoplasm through BAX/BAK and VDAC channels, activating the cGAS‐STING pathway. a,b) Immunoelectron microscopy shows mitochondrial leakage into the cytoplasm in hiPSC‐CMs treated with doxorubicin. Gold particles indicated by arrows represent mtDNA, doxorubicin 1 µM, 24 h, scale bar = 200 nm, n = 12, quantification values are expressed as mean ± SEM. c) Immunofluorescence shows localization of dsDNA (red) and mitochondria (TOM20, green) in heart section slide, scale bar = 20µm. d) Immunofluorescence shows expression of STING (red) and cTnT (green) in heart section slide, scale bar = 50 µm. e) Immunofluorescence shows expression of cGAS (red) and cTnT (green) in heart section slide, scale bar = 50 µm. f) qPCR shows activation of the cGAS‐STING pathway and increased expression of downstream inflammatory cytokines in hiPSC‐CMs after doxorubicin treatment, doxorubicin 1 µM, 24 h, n = 2, quantification values are expressed as mean ± SEM. g,h) Immunoblotting demonstrates that knock down of BAX/BAK and VDAC prevents doxorubicin‐induced activation of the cGAS‐STING pathway in hiPSC‐CMs, doxorubicin 1 µM, 24 h, n = 3, quantification values are expressed as mean ± SEM. i) Immunofluorescence shows localization of dsDNA (red) and mitochondria (TOM 20, yellow) in hiPSC‐CMs, scale bar = 20 µm.

Figure 4

MtDNA leakage‐activated cGAS‐STING pathway promotes senescence in cardiomyocytes. a) Bulk RNA‐seq reveals increased expression of inflammatory cytokines in myocardial cells after doxorubicin treatment mice, n = 3. b) A mitochondria‐targeted DNase1 has been designed to eliminate extruded mtDNA post‐doxorubicin treatment. c) Confocal microscopy shows localization of DNase1 (red) on mitochondria (TOM20, green) in infected hiPSC‐CMs, scale bar = 10 µm. d) qPCR reveals a significant decrease in inflammatory cytokine levels in hiPSC‐CMs expressing mitochondria‐targeted DNase1 following doxorubicin treatment, n = 3, quantification values are expressed as mean ± SEM. e) Immunoblotting demonstrate a significant decrease in p21 levels in hiPSC‐CMs expressing mitochondria‐targeted DNase1 following doxorubicin treatment, doxorubicin 1 µM, 24 h, n = 3, quantification values are expressed as mean ± SEM. f) SA‐β‐gal staining shows a significant reduction in the proportion of senescence cells in hiPSC‐CMs expressing mitochondria‐targeted DNase1 following doxorubicin treatment, scale bar = 40 µm. g) p21 immunofluorescence staining shows a significant reduction in the proportion of p21 positive cells in neonatal mouse ventricular myocytes expressing mitochondria‐targeted DNase1 following doxorubicin treatment, scale bar = 100 µm.

Figure 5

Heart mitochondrial transplantation alleviates doxorubicin‐induced cardiotoxicity. a) Schematic diagram of the chronic doxorubicin‐induced heart failure model with in situ mitochondrial transplantation. b) Mouse weight tracking shows that in situ transplantation of cardiac mitochondria protects against weight loss following doxorubicin treatment, n = 5, quantification values are expressed as mean ± SEM. c) Echocardiography shows that in situ transplantation of cardiac mitochondria partially alleviates doxorubicin‐induced cardiac dysfunction, n = 5, quantification values are expressed as mean ± SEM. d) RT‐qPCR demonstrate a significant decrease in cGAS and STING levels and downstream inflammation factors in AMCM cells of in situ mitochondria transplant mouse following doxorubicin treatment, n = 3, quantification values are expressed as mean ± SEM. e) Masson staining shows that in situ transplantation of cardiac mitochondria partially alleviates doxorubicin‐induced cardiac fibrosis, with blue indicating fibrotic areas in mouse heart section, scale bar = 200 µm. f) Ionoptix detection of isolated adult mouse cardiomyocytes under simulated physiological beating conditions shows that in situ transplantation of cardiac mitochondria alleviates doxorubicin‐induced cardiotoxicity, n = 100, quantification values are expressed as mean ± SEM. g) Transmission electron microscopy shows that in situ transplantation of cardiac mitochondria alleviates doxorubicin‐induced mitochondrial rupture, membrane swelling, and cristae fragmentation, scale bar = 500 nm. h) MitoTracker Green staining of isolated adult mouse cardiomyocytes shows that in situ transplantation of cardiac mitochondria increases mitochondrial copy numbers reduced by doxorubicin, n = 100, quantification values are expressed as mean ± SEM. i) TMRM staining of isolated adult mouse cardiomyocytes shows that in situ transplantation of cardiac mitochondria partially alleviates doxorubicin‐induced mitochondrial membrane potential decline, n = 100, quantification values are expressed as mean ± SEM.

Figure 6

Myocardial specific STING knockout and overexpression of mitochondria‐targeted DNase1 both confer protection against doxorubicin‐induced cardiotoxicity. a,h) Schematic representation of the chronic heart failure model induced by doxorubicin in myocardial specific STING knockout and mitochondria‐targeted DNase1 OE mice. b,i) Echocardiography shows that myocardial specific STING knockout and mitochondria‐targeted DNase1 OE partially alleviates doxorubicin‐induced cardiac dysfunction, n = 5, quantification values are expressed as mean ± SEM. c,j) Masson staining shows that myocardial specific STING knockout and mitochondria‐targeted DNase1 OE partially alleviates doxorubicin‐induced cardiac fibrosis, blue indicating fibrotic areas, scale bar = 200 µm. d,k) Transmission electron microscopy shows that myocardial specific STING knockout and mitochondria‐targeted DNase1 OE partially alleviates doxorubicin‐induced mitochondrial rupture, membrane swelling, and cristae fragmentation, scale bar = 500 nm. e,l) TMRM staining of isolated adult mouse cardiomyocytes shows that STING knockout and mitochondria‐targeted DNase1 OE partially alleviates doxorubicin‐induced mitochondrial membrane potential decline, n = 100, quantification values are expressed as mean ± SEM. f,m) RT‐qPCR demonstrate a significant decrease in cGAS and STING levels and downstream inflammation factors in AMCMs of STING knockout and mitochondria‐targeted DNase1 OE mice following doxorubicin treatment, n = 3, quantification values are expressed as mean ± SEM. g,n) Immunoblotting shows that myocardial specific STING knockout and mitochondria‐targeted DNase1 OE partially reduces doxorubicin‐induced p21 expression, n = 3.

Figure 7

Schematic presentation showing the signaling mechanism and function of the mtDNA‐cGAS‐STING pathway in doxorubicin injury and rescue strategies.