Activation of CISD2 as a protective strategy against doxorubicin-induced cardiotoxicity

Abstract

Background: Cardiotoxicity of doxorubicin, a chemotherapy medication, remains the most dangerous side effect. CISD2 plays a critical role during cardiac aging.

Objectives: We use a potent CISD2 activator, hesperetin, to ameliorate doxorubicin-induced cardiotoxicity by upregulating CISD2 in mice.

Methods: Two animal models, an acute and a tumor-bearing doxorubicin-induced cardiotoxicity model, were used in this study. Both genetic and pharmacological approaches were employed. Transgenic mice and a potent CISD2 activator, hesperetin, were utilized to ameliorate doxorubicin-induced cardiotoxicity by upregulating CISD2 expression in mice. Additionally, a human-derived iPSC system was used to provide human-relevant evidence. Comprehensive biological, histological, transcriptomic, and metabolomic analyses were conducted.

Results: Five findings are pinpointed. Firstly, doxorubicin suppresses Cisd2 expression resulting in cardiac electromechanical dysfunction. Intriguingly, transgenic overexpression of Cisd2 mitigates doxorubicin-induced cardiotoxicity. Secondly, hesperetin effectively sustains a high level of Cisd2 and improves cardiac function in a Cisd2-dependent manner after doxorubicin treatment. Importantly, hesperetin doesn't influence the anti-cancer efficacy of doxorubicin. Thirdly, doxorubicin downregulates the transcription of CISD2 by decreasing the expression of two transcription regulators, TAF1 and TCF12. Fourthly, analysis of transcriptomic and metabolomic datasets reveals that hesperetin protects the heart via a network connecting glucose, fatty acids and amino acids metabolism, thereby ensuring a sufficient energy supply. Additionally, hesperetin improves antioxidation capacity via reinstating the pentose phosphate and glutathione pathways. Finally, in human iPSC-derived cardiomyocytes, hesperetin significantly upregulates CISD2 and protects the cells from doxorubicin-induced toxicity and functional damage.

Conclusions: Our results highlight the potential utility of Cisd2 and its activator hesperetin in chemotherapy involving doxorubicin.

Keywords: CISD2; Cardiotoxicity; Doxorubicin; Hesperetin; iPSC derived cardiomyocytes.

Fig. 1

Doxorubicin (DOX) downregulates the expression of CISD2, while transgenic overexpression of CISD2 protects mice from DOX-induced cardiotoxicity. (A) DOX decreased the luciferase activity in the HEK293-CISD2 reporter cell line. However, Cisplatin, Taxol and Imatinib have no overt effects on the CISD2 reporter (n = 3). (B–C) The mRNA (B) and the protein (C) levels of Cisd2 were decreased in a dose-dependent manner in HL-1 mouse cardiomyocytes after DOX treatment for 24 h (n = 3). (D) The mRNA level of Cisd2 was decreased at 16 h in the cardiac muscle of DOX-treated mice after injection with a single dose of DOX (25 mg/kg) (n = 4 for control group, and n = 5 for DOX-treated group). (E) The level of Cisd2 in the Cisd2 overexpression (Cisd2-OE) stable line of HL-1 mouse cardiomyocytes was analyzed by Western blot. (F) Cisd2-OE improved DOX-induced mitochondrial dysfunction as monitored by JC-1 staining of mitochondrial membrane potential (n = 3). (G) Animal protocol for acute DOX-injury model (H–N). WT and Cisd2 transgenic (Cisd2TG) mice were injected with a single dose of DOX (25 mg/kg, i.p. injection). Cardiac function was monitored at day 4 (Echo) and day 5 (ECG) after DOX treatment. (H–I) Echocardiography (Echo) is used to measure the left ventricular ejection fraction and end-diastolic diameter. Representative echocardiographic (ECG) results (H) and the quantification data (I) from the WT and Cisd2TG mice are shown (n = 7 or 8 for each group). (J) Representative waterfall plots of ECG results of WT and Cisd2TG mice that received a single dose of DOX. (K–N) Cardiac electrical dysfunction was detected by ST elevation (K), heart rate (L), QTc (M) and T-peak-T-end (N) (n = 7 or 8 for each group). Data represent mean ± SD from at least three independent biological replicates per group, and the numbers are indicated as n. All data are analyzed by the Kruskal-Wallis test with the Dunn's test. ∗p < 0.05, ∗∗p < 0.005.

Fig. 2

The CISD2 activator hesperetin ameliorates DOX-induced cardiac dysfunctions without affecting the anti-cancer efficacy of DOX and hesperetin functions in a Cisd2-dependent manner. (A) Animal protocol for pharmaceutical approach using an acute DOX-injury model (B–I). Cisd2^f/f and Cisd2cKO mice were injected with a single dose of DOX (25 mg/kg, i.p. injection) with or without hesperetin treatment (10 mg/kg/day, i.p. injection). (B–C) The protein level of Cisd2 was detected by Western blot (n = 3–4). The mice used in this experiment are WT C57BL/6 males. (D) Echocardiography was used to measure the left ventricular ejection fraction and end-diastolic diameter. Representative echocardiographic images are shown. (E) Quantification results of the ejection fraction (n = 7 or 8). (F) Representative waterfall plots of the electrocardiographic results. (G) The cardiac electrical dysfunction was monitored by ST elevation (n = 7 or 8). (H) Serum levels of CKMB (n = 5–7). (I) Serum levels of Troponin I (n = 6). (J) Animal protocol for tumor-bearing mice (K–N). WT mice were implanted with LLC1 cells by subcutaneous (s.c.) injection and received six doses of DOX injection (5 mg/kg, i.p. injection) with or without hesperetin treatment. (K–L) Tumor volume (K) and tumor weight (L) were measured (n = 8 for each group). (M–N) Serum levels of CKMB (M) and serum troponin I (N) in different groups of tumor-bearing mice were examined (n = 8). V or Veh: vehicle control; Hes: hesperetin treatment. Data are presented as mean ± SD and are analyzed by the Kruskal-Wallis test with the Dunn's test. ∗p < 0.05; ∗∗p < 0.005.

Fig. 3

The CISD2 activator hesperetin improves the ultrastructural abnormalities present in the cardiac muscle of DOX-treated mice. (A) Lateralization of gap junctions in the cardiac muscle of DOX-treated mice at 3 months old. Representative IF images of heart sections stained with antibodies against Cx43 (green) to localize the gap junctions, with antibodies against pan-cadherin (red) to localize the intercalated discs, and with wheat germ agglutinin (WGA; purple) to stain cell membranes by binding to membrane glycoproteins. The sections were also stained with Hoechst (blue) to identify the nuclei. White arrows indicate lateralization of gap junctions. (B) Colocalization coefficient of gap junctions (Cx43) and intercalated discs (pan-cadherin) was analyzed by Pearson's correlation. The quantification data is presented as the Cx43/pan-cadherin colocalization coefficient. Data were collected from 6 to 10 randomly selected fields for each heart sample (n = 5 mice for each group). Data are presented as mean ± SD and are analyzed by one way ANOVA with Tukey correction. ∗p < 0.05. (C) Co-treatment with hesperetin ameliorates the DOX-induced ultrastructural defects, including disorganization of the fascia adherens, fragmentation and loss of gap junctions, and degenerated and swollen mitochondria (indicated by yellow stars), as well as disorganized and degenerated myofibrils. (D) Quantitative analyses of mitochondrial aspect ratio (height/width), average cristae length, total cristae length, and total cristae length normalized to mitochondrial area. ∗p < 0.05, ∗∗p < 0.005.

Fig. 4

DOX downregulates CISD2 gene expression via decreasing the transcription regulators, TCF12 and TAF1, in AC16 human cardiomyocytes. (A) A positive correlation between the expression patterns of Cisd2, TAF1 and TCF12 in the heart of mice treated with a single dose of DOX (25 mg/kg) for 16 h (n = 3 for Veh-treated group, and n = 4 for DOX-treated group). (B–C) ChIP-ddPCR assays to detect the transcription regulators binding to the CISD2 promoter with or without DOX treatment. ChIP assays for the CISD2 promoter were performed using TAF1 and/or TCF12 antibodies. The captured DNA fragments were quantified by ddPCR (n = 3). (D) Schematic illustration of TAF1 and TCF12 as putative transcription regulators of CISD2. (E) ChIP-reChIP assays to detect the binding of TAF1 and TCF12 on the CISD2 promoter (n = 3). (F) Disrupting the binding site of TCF12 by CRISPR/Cas9-mediated DNA deletion downregulates CISD2 expression (n = 3). The CISD2 levels in clone #13 and clone #28 were measured using Western blotting. Data represent mean ± SD from at least three independent biological replicates per group, and the numbers are indicated as n. All data were analyzed by one way ANOVA with the Tukey correction. ∗p < 0.05, ∗∗p < 0.005.

Fig. 5

DOX-mediated DEGs are reverted by hesperetin thereby improving DOX-induced cardiac dysfunction at 16 h and 72 h after a single dose of DOX treatment (25 mg/kg). (A) Volcano plots of DEGs. (B) Unsupervised PCA score plots of DEGs (mouse number n = 3 for each group). (C) Unsupervised hierarchical heatmap of the top 100 DEGs (mouse number n = 3 for each group). (D) Venn diagram of DOX-mediated DEGs that can be reverted by hesperetin, namely the hesperetin-reverted DEGs. (E) Functional annotations of DEGs based on the KEGG pathways. The common enriched pathways for both 16-h and 72-h stages are shown. (F) Functional annotations of DEGs based on the KEGG pathways. The unique enriched pathways existed only at the 72-h stage are shown.

Fig. 6

Integrated trans-omics analysis reveals the effects of hesperetin on various interconnected pathways for energy production and antioxidation. (A) Animal protocol for the study of cardiac transcriptomic and metabolomic analyses. Mice were injected with a single dose of DOX (25 mg/kg, i.p. injection) with or without treatment of hesperetin (10 mg/kg/day, i.p. injection). For the transcriptomic and metabolomic analyses, mice were sacrificed at 16 h and 72 h after DOX injection. (B) PCA score plot of the differentially expressed metabolites (DEMs) (n = 5). (C) Unsupervised hierarchical heatmap of the top 25 DEMs. (D) Functional annotations of DEGs and DEMs based on KEGG pathways. (E) The DEGs, DEMs and their associated pathways are summarized to illustrate the network connections of the pathways; this trans-omics analysis integrated the 16-h transcriptomic datasets with the 72-h metabolomic datasets. Left panel: Data were obtained from the comparison between DOX and Vehicle (DOX vs Veh). The red symbols indicate up-regulation in the DOX group; the blue symbols indicate down-regulation in the DOX group. Right panel: Data were obtained from the comparison between DOX + Hes and DOX (DOX + Hes vs DOX). The orange symbols indicate up-regulation in the DOX + Hes group; the green symbols indicate down-regulation in the DOX + Hes group.

Fig. 7

Hesperetin ameliorates DOX-induced cellular damage and improves the functioning of human iPSC-CMs. (A) The scheme for differentiation of human iPSCs into iPSC-derived cardiomyocytes (iPSC-CMs). The concentration of DOX used to treat the iPSC-CMs is 1 μM. (B) The protein level of CISD2 was detected using Western blotting. (C) The mitochondrial membrane potential was measured by JC-1 staining followed by flow cytometry (n = 8). (D) The cytosolic ROS was measured by DCFDA staining and the data were recorded by microplate reader. The area under curve (AUC) was calculated for the quantification (n = 10). (E) The mitochondrial oxidative stress was measured using MitoSOX staining and analyzed by flow cytometry (n = 10). (F–G) The spontaneous Ca²⁺ waves of beating iPSC-CM were measured by confocal microscopy using fura-2/AM staining. Total peak area and maximal Ca²⁺ release (maximum-minimum level) were quantified (n = 7). (H) Representative images of the sarcomeric organization in DOX-treated human iPSC-CMs. TnI: Troponin I; ACTN2: Actinin Alpha 2. (I) The cell survival rate of DOX-treated iPSC-CMs with or without hesperetin (10 μM) treatment (n = 5). The iPSC-CMs were pre-treated with 10 μM hesperetin for 24 h, and then co-treated with 1 μM DOX for another 24 h before Western blot and functional analysis. For cell survival, the iPSC-CMs were treated for a total of 72 h with DOX. Data represent mean ± SD from at least three independent biological replicates per group, and the numbers are indicated as n. All data were analyzed by one way ANOVA with the Tukey correction. ∗p < 0.05, ∗∗p < 0.005.