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: Doxorubicin, Cardiotoxicity, CISD2, Hesperetin, iPSC derived cardiomyocytes 1. Introduction Doxorubicin (DOX) has served as a mainstay of life-saving chemotherapeutic treatment of various cancers, such as lymphoma, sarcoma, breast cancer, and pediatric leukemia, for more than five decades [[51]1,[52]2]. Nevertheless, unresolved side effects, particularly DOX-induced cardiotoxicity, persist as a significant unmet need and challenge. Considering the increasing number of cancer survivors who suffer from DOX-induced cardiotoxicity, understanding the pathophysiology and the formation of appropriate treatment guidelines is greatly needed. Currently, dexrazoxane stands as the sole FDA-approved medication for mitigating DOX-induced cardiotoxicity [[53]3]. However, there is a concern that dexrazoxane may increase the risk of developing secondary malignancies by reducing the antineoplastic effect of DOX [[54]4]. DOX-induced cardiotoxicity arises via a wide range of mechanisms, including oxidative stress, mitochondrial dysfunction [[55]5], and the ternary complex (DNA/DOX/Top-II) formation [[56]6]. Mitochondria, which are pivotal for energy metabolism and the primary generators of reactive oxygen species (ROS), play a central role in DOX-mediated cardiomyocyte injury [[57]7]. Increased ROS production, particularly from mitochondrial complex I during DOX treatment, disrupts cellular homeostasis and impairs cardiac function [[58]5]. Mitochondrial DNA can be damaged by ROS, leading to a down-regulation of the respiratory complexes, a reduction in respiratory coupling and an enhancement of electron leakage from the respiratory chain [[59]7]. This contributes to cardiomyocyte injury. CDGSH Iron Sulfur Domain 2 (CISD2) is a pro-longevity gene that plays a critical role in maintaining cellular homeostasis in multiple organs of mice, including muscle and heart, the organs most affected during chemotherapy. CISD2 protein localizes to the endoplasmic reticulum (ER), outer mitochondrial membrane (OMM) and mitochondria-associated membrane (MAM). Previously we have shown that CISD2 mediates mitochondrial integrity and lifespan in mice [[60][8], [61][9], [62][10], [63][11]]. Our studies have highlighted that CISD2 plays a pivotal role in regulating cytosolic Ca^2+ homeostasis, ER integrity, and mitochondrial function [[64][11], [65][12], [66][13]]. In the heart, CISD2 levels decrease with age, whereas a sustained elevation of CISD2 effectively delays cardiac aging [[67]14]. CISD2 deficiency has been shown to disrupt the integrity of the intercalated discs and to induce cardiac electromechanical dysfunction. Furthermore, increasing CISD2 expression in cardiomyocytes via genetic overexpression or pharmaceutical augmentation significantly reduces ROS production and preserves mitochondrial functioning [[68]14,[69]15]. Hesperetin, which is a potent CISD2 activator, is able to restore CISD2 expression levels and attenuate age-associated functional decline in the hearts of aged mice [[70]15]. Hesperetin can be converted from hesperidin, a naturally occurring flavanone-glycoside enriched in the peels of citrus fruits, by probiotics in the gastrointestinal tract [[71]16], and has been developed into a functional food. Given the crucial role of CISD2 in cardiac functioning via the maintenance of ROS homeostasis and mitochondrial functioning, it is important to investigate whether CISD2 function is impaired after doxorubicin treatment and to explore whether restoring CISD2 function is able to reduce DOX-induced cardiotoxicity. This study demonstrates the involvement of CISD2 in DOX-mediated cardiotoxicity and provides proof-of-concept for the hypothesis that a CISD2 activator, such as hesperetin, is able to ameliorate DOX-induced cardiotoxicity via enhancing CISD2 levels in the heart, which protects the organ against chemotherapy-induced heart failure. 2. Materials and METHODS 2.1. Mice Cisd2 transgenic (Cisd2TG) mice and Cisd2 floxed allele (Cisd2^f/f) mice have been generated previously [[72]13]. Inducible cardiac-specific Cisd2 knockout (Cisd2cKO) mice were generated by crossing Cisd2^f/f mice with the aMHC-MerCreMer mice (JAX stock number 005657). The mice were administrated with tamoxifen (40 mg/kg; Sigma-Aldrich) on three consecutive days by i.p. injection at 8 week-old. All mice used in this study are males and have a pure or congenic C57BL/6 background. They were bred and housed in a specific pathogen-free animal facility. After each specific treatment, the mice were sacrificed by CO[2] inhalation, which is a humane method of euthanasia. Mice were treated according to the criteria outlined in the NIH Guide for the Care and Use of Laboratory Animals. The animal protocols were approved by the Institutional Animal Care and Use Committee of Chang Gung Memorial Hospital (No. 2017103002 and 2017030901) and National Yang Ming Chiao Tung University (No. 1040104r). The animal protocols were designed to respect the associated guidelines and the 3R principles according to the “Animal Protection Act” of Taiwan. 2.2. Animal protocols for acute DOX injury and the tumor-bearing mouse model For acute DOX injury, the mice at 3-month-old were received a single intraperitoneal (i.p.) injection of DOX (25 mg/kg), as previously described [[73]6]. Treatment of hesperetin (10 mg/kg) was performed two days before the DOX injection by i.p. injection daily. The vehicle control of DOX was 10 % dimethyl sulfoxide (DMSO) in PBS, for hesperetin was 10 % propylene glycol (PG) in PBS. The cardio function analysis was carried out at day 4 (Echo) and day 5(ECG), and the mice were sacrificed at day 5. For the tumor-bearing mice, LLC1 cells (10^6 cells) were subcutaneously implanted. Upon reaching a tumor volume of 100 mm^3 on day 9, the mice were randomly grouped into control, hesperetin, DOX and DOX combined with hesperetin. mice were randomly allocated into control, hesperetin, DOX, and DOX combined with hesperetin groups. Mice received DOX (5 mg/kg) via i.p. injection every other day, while the hesperetin (10 mg/kg) was administrated using i.p. injection daily. The cardio function analysis was carried out at day 16 (Echo) and day 17 (ECG), and the mice were sacrificed at day 17. The tumor volume was calculated as (a^2 x b)/2, where ‘a' represents width and ‘b' represents length. All the mice used in this study were male. 2.3. Western blotting Cardiac muscle tissue samples were homogenized using a MagNA Lyser (Roche) in RIPA buffer and then denatured in 2 % SDS sample. Total protein lysate was separated by SDS-polyacrylamide gel electrophoresis (Bio-Rad), which was followed by electro-transfer to a polyvinylidene fluoride transfer membrane (NEF1002001PK, PerkinElmer). Next, these membranes were blocked with 5 % (w/v) non-fat dried milk in TBST for 60 min at room temperature, and then incubated with primary antibody for 14–16 h at 4 °C. This was followed by washing three times with TBST, probing with the secondary antibody for 60 min at room temperature. Finally, detection was carried by ECL (34580, Thermo). The following antibodies were used for the Western blotting: Cisd2, and Gapdh (MAB374, Millipore). 2.4. Transthoracic echocardiography To evaluate cardiac function, we employed the VisualSonics VeVo 2100 Imaging System (VisualSonics, Toronto, Ontario, Canada). Male mice were subjected to anesthesia with 1 % isoflurane in 95 % O2. To maintain a stable body temperature, we used a heated pad (TC-1000, CWE Inc. USA) and monitored it within the range of 36 °C–37 °C. Continuous monitoring of electrocardiograms (ECGs) was conducted throughout the procedure. Cardiac function assessment was performed using a high-frequency probe operating at 30–50 MHz. Subsequent data analysis was executed with the aid of VisualSonics software (VisualSonics). The individuals responsible for data acquisition were kept unaware of the animal group assignments, ensuring unbiased data collection. 2.5. Electrocardiography (ECG) Mice underwent cardiac ECG functional testing as previously described [[74]14]. All procedures were conducted during the light phase. Anesthesia was initiated by exposing the mice to 3 % isoflurane in a chamber for 3–5 min, and it was maintained with 1.5 % isoflurane via a nose cone while the mice were positioned on a heated pad for ECG recording. Continuous 5-min ECGs were recorded using subcutaneous electrodes and a PowerLab data acquisition system (model ML866, ADInstruments, Colorado Springs, CO) along with an Animal Bio Amp (model ML136, ADInstruments). ECG analysis was carried out impartially using LabChart 7 Pro version 7.3.1 (ADInstruments, Inc). QTc intervals, QRS intervals, and Tpeak-Tend intervals were assessed based on Mouse ECG parameters. Statistical comparisons were made using the Mann-Whitney U test, and significance was defined as a p-value of <0.05. 2.6. Immunofluorescence (IF) and Cx43/pan-cadherin colocalization coefficient assessment The IF staining was carried out in accordance with the previously established protocols [[75]14]. In brief, cryosections embedded in optimal cutting temperature (16 μm) were subjected to staining with the following antibodies: pan-cadherin (C3678, Sigma), Desmoplakin (CBL173, Millipore), Cx43 (C8093, Sigma), Collagen I (ab270993, Abcam), and α-actinin (A7811, Sigma). Additionally, wheat germ agglutinin (WGA) or α-actinin (Rhodamine phalloidin) was co-stained in combination with nuclear staining dyes DAPI or Hoechst. Gap junction remodeling was quantified by assessing the degree of colocalization between Cx43 and pan-cadherin at the intercalated discs, employing the Pearson–Spearman correlation colocalization (PSC) plugin in ImageJ. 2.7. Transmission electron microscopy (TEM) The TEM was performed as described previously [[76]14]. In brief, mouse cardiac muscles were fixed in a TEM fix buffer (1.5 % glutaraldehyde and 1.5 % paraformaldehyde in 0.1 M cacodylate buffer at pH 7.3), post-fixed in 1 % OsO4 and 1.5 % potassium hexanoferrate and then tissues were washed in cacodylate and 0.2 M sodium maleate buffers (pH 6.0) followed by block-stained with 1 % uranyl acetate. Following dehydration, the cardiac muscles were embedded in Epon (EMS, Hatfield, PA, USA, #14120) and sectioned for TEM analysis. Quantification of mitochondrial morphology was performed using ImageJ/Fiji software. The structure of the mitochondrial cristae was analyzed by applying skeletonization to TEM images, and the total cristae length was calculated by summing the lengths of all skeletonized branches. A total of 50 mitochondria were randomly selected and analyzed from 5 to 7 representative fields per experimental group. 2.8. Exploring potential transcription factors for CISD2 To identify potential transcription factors regulating CISD2, we analyzed multiple ChIP-seq datasets obtained from the ENCODE project. Candidate transcription factors were selected based on their binding regions being located within 5 kb upstream of the transcription start site and being supported by more than half of the ChIP-seq experiments. A total of 30 transcription factors met these criteria. Then these candidate regulators were knock-down by shRNAs (National RNAi Core Facility, Academia Sinica, Taipei, Taiwan) and the expression of CISD2 was observed by QPCR in HEK293 cells. Total RNA from knockdown cells was extracted via TRIzol reagent (Thermo Fisher Scientific, Waltham, MA, US) according to the manufacturer's protocol. The mRNA expression levels of CISD2 were quantified by normalizing the quantification cycle (Ct) values of CISD2 to ACTB values using qPCR analysis. The primers and Taqman probes (CISD2: Hs00381903_m1, ACTB: Hs01060665_g1) for qPCR analysis were obtained from Thermo Fisher (Thermo Fisher Scientific, Waltham, MA, US). 2.9. Chromatin immunoprecipitation - droplet digital PCR (ChIP-ddPCR) assays The genomic DNA was cross-linked with DNA-binding proteins and then sonicated to desired fragment size. Antibodies against TAF1 (sc-735, Santa Cruz biotech, Santa Cruz, CA, US and ab28450, Abcam, Cambridge, UK), TCF12 (sc-28364, Santa Cruz biotech, Santa Cruz, CA, US. and 14419-1-AP, Proteintech, Rosemont, IL, US), mouse IgG (PP54, Merk Millipore, Burlington, MA, US), and rabbit IgG (ab172730, Abcam, Cambridge, UK) were coupled to magnetic beads (Dynabeads protein G,1004D, Thermo Fisher Scientific, Waltham, MA, US) to capture protein-DNA complexes. DNA isolation is performed according to the manufacturer's protocol. The DNA fragments were then analyzed using the QX200 ddPCR system (Bio-Rad, Hercules, CA, USA). For the ddPCR analysis, the sample, primers, supermix, and mineral oil were loaded into a droplet generator to form thousands of droplets. PCR was performed with primer pairs designed for the binding sites and control ([77]Supplemental Table S1). Droplets were then aspirated and read by the Droplet Reader, and the data were analyzed using QuantaSoft analysis software (Bio-Rad, Hercules, CA, US). For the ChIP-reChIP assay, the complexes of primary ChIPs were eluted from the magnetic beads in elution buffer containing 10 mM DTT for 30 min at 37 °C, and then diluted 10 times using dilution buffer. This was followed by a second round of ChIP using IgG, TAF1 or TCF12 antibody. 2.10. CRISPR/Cas9-mediated deletion of TCF12 binding site To disrupt the consensus binding motif of TCF12 in AC16 cardiomyocytes, the genome region containing putative TCF12-binding consensus sequences was deleted using the CRISPR/Cas9 system. The SgRNA/Cas9 expression vector (Sg12-1: GGGAAGAGGCGGGGCGAGAG, Sg12-2: GATGCTGGAGACCAGCAGAA) was obtained from Dr. Tsai-Yu Tzeng, Cancer and Immunology Research Center-Genome Editing Core Facility, NYCU, Taiwan. For SgRNA/Cas9, AC16 cells were transiently transfected with the SgRNA-Cas9 plasmid and then sorted using FACSAria (BD). The deleted sequence in the target region was cloned using a TA Cloning™ Kit (Promega), followed by sequencing. 2.11. RNA isolation from tissue, RNA sequencing and pathway analysis Total RNA was isolated from the cardiac muscle using TRI Reagent (T9424, Sigma) and phenol/chloroform extraction. The quality of the total RNA was assessed using an Agilent 2100 Bioanalyzer (Agilent Technologies); samples with a RNA Integrity Number (RIN) higher than 8 were subjected to RNA sequencing (RNA-seq). The RNA-seq was conducted by the Genome Research Center at National Yang Ming Chiao Tung University. The analysis was generated to a depth of at least 20 million reads for each sample by single-end sequencing. Differentially expressed genes (DEGs) and normalized counts were identified using DESeq2 in a pairwise manner across the different conditions, and the Wald test statistic was used to identify significance. For the pathway enrichment analysis, pairwise Gene Set Enrichment Analysis (GSEA) was conducted using total transcriptome data. To determine the functional analysis of DEGs, KEGG enrichment analysis was performed using the Database for Annotation, Visualization, and Integrated Discovery (DAVID) Bioinformatics Resources 6.8 ([78]http://david.abcc.ncifcrf.gov/). Additionally, Ingenuity Pathway Analysis (IPA) from Ingenuity® Systems ([79]http://www.ingenuity.com) was employed to identify the pathways associated with the selected DEGs. Bar charts and bubble plots were then generated using GraphPad Prism v9.0. 2.12. Metabolomic analysis The tissue samples were homogenized with 300 μL 50 % methanol. The supernatant was mixed with 120 μL methanol and 900 μL methyl-tert-butyl ether, and then the mixture was incubated for 1 h at room temperature. The 66 μL ddH2O was added and mixed. After centrifugation, the lower level was collected and mixed with 1 mL acetonitrile (ACN). The mixture was centrifuged at 12,000 rpm, 4 °C for 30 min. The supernatant was dried and dissolved with LCMS grade H[2]O for LCMS analysis. The extraction samples were subjected to TQ-XS (Waters Corp., Manchester, UK). The analysis system was Waters ACQUITY UPLC, and ACQUITY BEH C18 column (1.7 μm, 2.1 × 100 mm, Waters Corp., Milford, MA, USA) was used in this analysis. Column temperature and flow rate were maintained at 45 °C and 0.3 mL/min. H[2]O with 10 mM Tributylamine (TBA) and 15 mM acetic acid was used as mobile phase A, and 50 % ACN with 10 mM TBA and 15 mM acetic acid was used for mobile phase B. The ionization mode was ESI in the negative mode, and the capillary was set to 1 kV. The desolvation temperature, desolvation gas flow and cone gas flow were set at 500 °C, 1000 L/h, and 150 L/h. The data were processed and analyzed by MassLynx (version 4.2, Waters Corp.). 2.13. Human induced pluripotent stem cell culture and differentiation The human induced pluripotent stem cells (iPSCs) were provided by Dr. Joseph C Wu at Stanford Cardiovascular Institute [[80]17] and cultured and differentiated as described previously [[81]18]. In brief, the SCVI 51 and SCVI 79 (two DOX-sensitive iPSC lines; patients with cardiotoxicity after treatment of chemotherapy) were routinely maintained in E8 on 1/200 growth factor-reduced Matrigel. When they came to 80 % confluence, 8 μM CHIR99021 was added in RPMI/B27 insulin-free medium to induce mesoderm differentiation at day 0–2. After replacing in RPMI/B27 insulin-free medium for 24 h, the cells were incubated with 5 μM IWR-1 (Sigma) for 48 h (day 4–5). At day 5–7, cells were changed to RPMI/B27 insulin-free medium, and then changed to RPMI/B27 medium. The RPMI/B27 medium was changed every other day until the cells started beating. For purification of cardiomyocytes, cells were cultured in glucose-free RPMI/B27 for 4 days. The beating cells were seeded at 2-3x10^6 per 6-well and the experiments were performed at day 35–40 when they came matured. For treatment, the cells were pre-treated with 10 μM hesperetin for 24 h, and then co-treated with 1 μM DOX for another 24 h for Western blot and functional analysis. For the cell survival, cells were treated with a total of 72 h of DOX. 2.14. Intracellular calcium wave To assess cytosolic Ca2+ levels, iPSC-CMs were subjected to staining with 2 μM fura-2/AM (Invitrogen, San Diego, CA) in RPMI/B27 culture medium at 37 °C for 30 min. Excitation of fura-2/AM occurred alternately between 340 nm and 380 nm utilizing the Polychrome IV monochromator (Till Photonics, Grafelfing, Germany). Imaging was conducted using an Olympus IX71 inverted microscope equipped with a xenon illumination system and an IMAGO CCD camera (Till Photonics). Fluorescence intensity at 510 nm was continuously monitored, digitally stored, and subsequently analyzed employing TILLvisION 4.0 software (Till Photonics, Grafelfing, Germany). 2.15. Statistical analysis All the data are presented as mean ± SD. Comparisons among groups of more than two were carried out using the Kruskal-Wallis test followed by the Dunn's multiple comparisons test, and a p value of <0.05 was considered significant. Statistical analysis was carried out using the software package Graphpad Prism 9.0. 3. RESULTS 3.1. DOX downregulates CISD2 and induces cardiotoxicity, while an enhancement of CISD2 level in transgenic mice ameliorates DOX-induced heart failure To examine whether chemotherapeutic drugs affect the expression of CISD2, we utilize the HEK293-CISD2 reporter cell line, which monitors the expression of human CISD2 gene at the transcriptional level [[82]15]. We treated these cells with various agents, including Cisplatin, DOX, Taxol and Imatinib. Interestingly, a significant decrease in CISD2 reporter was observed following DOX treatment; however, the other three drugs were found to have no overt effects on CISD2 reporter ([83]Fig. 1A). Given the well-established association between DOX and cardiomyopathy, we next investigated the biological significance of CISD2 suppression in relation to DOX-induced cardiotoxicity. Notably, both the levels of Cisd2 mRNA and Cisd2 protein were downregulated in the HL-1 mouse cardiomyocyte cell line and in mouse cardiac muscle following DOX treatment ([84]Fig. 1B–D). To assess the causal role that Cisd2 may play in DOX-induced cardiotoxicity at the cellular level, we treated a Cisd2-overexpressing (OE) HL-1 stable line with DOX. Remarkably, Cisd2-OE shows attenuated DOX-induced mitochondrial depolarization ([85]Fig. 1E and F). These results underscore the importance of Cisd2 downregulation to DOX-induced cardiotoxicity and suggest that restoring Cisd2 expression may serve as a promising strategy for cardiac chemoprotection. Fig. 1. [86]Fig. 1 [87]Open in a new tab 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. The Cisd2TG mice, which is a long-lived mouse model carrying four-copies of the Cisd2 gene [[88]13], were subjected to an acute DOX injury protocol ([89]Fig. 1G). Functional analysis of echocardiography (ECG) revealed the presence of left ventricular contractile dysfunction in DOX-treated wild-type (WT) mice, namely a significant reduction in the left ventricular (LV) ejection fraction. Notably, Cisd2TG mice displayed an improved trajectory for the LV ejection fraction reduction ([90]Fig. 1H and I). Furthermore, our ECG results showed a decrease in heart rate, an elevation of the corrected QT interval (QTc) and Tpeak-Tend in DOX-treated WT mice. Importantly, in DOX-treated Cisd2TG mice, these parameters were significantly improved ([91]Fig. 1J–N). Thus, enhanced Cisd2 expression not only protects against the detrimental effects of DOX in relation to electric conductance, but also preserves cardiac contractile capability, which otherwise is decreased after DOX treatment. 3.2. The CISD2 activator, hesperetin, ameliorates DOX-induced cardiotoxicity in a CISD2-dependent manner without affecting the anti-cancer activity of DOX We next evaluated whether pharmaceutical activation of CISD2 is able to protect against the DOX-induced cardiotoxicity by using hesperetin to enhance CISD2 expression. In DOX-treated HL-1 cardiomyocytes, hesperetin preserved the expression level of CISD2. Additionally, DOX-induced damage to mitochondrial function, as monitored by JC-1 staining as part of mitochondrial membrane potential assays, was mitigated by hesperetin treatment ([92]Supplemental Fig. S1A–B). To study if the beneficial effects of hesperetin functions in a CISD2-dependent manner, we generated Cisd2cKO mice (aMHC-MerCreMer x Cisd2^f/f; [93]Supplemental Fig. S1C). Both the control (WT or Cisd2^f/f) and Cisd2cKO mice were subjected to an acute DOX injury protocol ([94]Fig. 2A). In DOX-treated WT mice, the Cisd2 level was significantly downregulated in the heart. Notably, hesperetin treatment was able to maintain a high level of Cisd2 expression ([95]Fig. 2B and C). In [96]Fig. 2B, the band labeled Cisd1 is a cross-reaction signal of the polyclonal Cisd2 antibody; the two proteins, namely Cisd1 and Cisd2, have a high amino acid similarity [[97]9] and therefore the antibody shows cross-reaction; notwithstanding the above, DOX treatment can be seen to have no overt effect on Cisd1 expression. Notably, the enhanced Cisd2 expression appears to ameliorate DOX-induced mechanical dysfunction ([98]Fig. 2D and E) and minimize the electrical disturbance ([99]Fig. 2F and G; [100]Supplemental Fig. S1D). Furthermore, pathological analysis revealed that hesperetin effectively decreased cytosolic vacuole and myofibril degeneration ([101]Supplemental Fig. S1E) that are induced by DOX. Moreover, the serum levels of creatine kinase-MB form (CKMB), a biomarker for cardiac damage, were significantly reduced by hesperetin when mice underwent DOX treatment ([102]Fig. 2H). Intriguingly, in Cisd2cKO mice, in the absence of Cisd2, hesperetin lost its beneficial effect with respect to improvements related to mechanical dysfunction ([103]Fig. 2D and E), electrical disturbance ([104]Fig. 2F and G; [105]Supplemental Fig. S1D), improvements related to histopathological damage ([106]Supplemental Fig. S1E), and the damage biomarker serum CKMB, as well as Troponin I (TnI) ([107]Fig. 2H and I). Together, these results indicate that the action of hesperetin needs the presence of Cisd2 in order to bring about its beneficial effects and that the functioning of hesperetin in this context is mainly dependent on Cisd2. Fig. 2. [108]Fig. 2 [109]Open in a new tab 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. Importantly, hesperetin treatment does not compromise the anti-cancer efficacy of DOX. To closely mimic DOX treatment in cancer patients, we employed a tumor-bearing C57BL/6 mouse model carrying isogenic Lewis lung carcinoma LLC1 cells ([110]Fig. 2J). Notably, DOX treatment effectively reduced the tumor burden, while, interestingly, co-treatment with DOX and hesperetin did not compromise the anti-cancer efficacy of DOX ([111]Fig. 2K and L). Furthermore, DOX significantly decreased cardiac contractility, as evidenced by echocardiographic analysis; while hesperetin significantly improved cardiac functioning ([112]Supplemental Fig. S1F–G). Additionally, co-treatment with hesperetin rescued DOX-induced pathological damage, namely myofibrillar loss and cytoplasmic vacuolation, in the mouse heart ([113]Supplemental Fig. S1H), as well as decreasing serum CKMB and TnI levels ([114]Fig. 2M and N). Together, these results reveal that hesperetin is able to protect against DOX-induced cardiotoxicity and that hesperetin has no overt effect on the anti-cancer activity of DOX in the tumor-bearing mouse model. 3.3. Hesperetin ameliorates damage caused by DOX at the cellular and ultrastructural levels To further elucidate the impact of DOX and hesperetin on the components at the ultrastructural level that form intercalated discs (ICDs), namely gap junctions, desmosomes and fascia adherens, we conducted a comprehensive analysis that employed both immunofluorescence (IF) staining and transmission electron microscopy (TEM). In the DOX-treated mice, we observed a notable amplification and lateralization of the gap junctions that were marked by staining for Cx43, a gap junction protein. Moreover, the colocalization coefficient of Cx43 and pan-cadherin, an ICD protein, showed a significant reduction in DOX-treated mice, which indicated that a substantial proportion of lateralized gap junctions did not colocalize with ICDs. Remarkably, these aberrations in gap junctions were significantly reversed upon co-treatment with hesperetin ([115]Fig. 3A and B). Furthermore, severe ultrastructural alterations were found in DOX-treated hearts, including disorganization of the fascia adherens, as well as fragmentation and loss of gap junctions. Notably, degenerated and swollen mitochondria, together with disorganized and degenerated myofibrils, were readily observed in DOX-treated mice ([116]Fig. 3C). To quantify mitochondrial morphology, we measured the aspect ratio of the mitochondria, namely the ratio of the heights of the mitochondria to the widths of the mitochondria. In addition, we also measured the cristae length. These quantitative results revealed that DOX treatment significantly reduced both the aspect ratio and the length of the cristae. Importantly, hesperetin effectively mitigated these detrimental effects on mitochondrial ultrastructure caused by DOX treatment ([117]Fig. 3C and D). Fig. 3. [118]Fig. 3 [119]Open in a new tab 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. 3.4. DOX downregulates CISD2 expression via decreasing its transcription regulators TCF12 and TAF1 The regulatory mechanisms governing the transcription of CISD2 remain to be elucidated. therefore, we sought to identify potential transcription regulators of CISD2 by leveraging the ENCODE project database, which comprises over 2000 CHIP-seq datasets from various transcription factors derived from cell lines and tissues. Thirty candidates having the potential to bind to the CISD2 promoter were selected; the screening strategy and validation of the candidate regulators are shown in [120]Supplemental Fig. S2A–B. Interestingly, TAF1 and TCF12 consistently emerged as CISD2 activators based on the fact that their knockdown by appropriate shRNAs in HEK293 cells significantly reduced the expression of CISD2 ([121]Supplemental Fig. S2C). Notably, a correlation in the expression patterns of Cisd2, TAF1, and TCF12 was observed in the hearts of DOX-treated mice and HL-1 mouse cardiomyocytes ([122]Fig. 4A; [123]Supplemental Fig. S2D and S2E). Fig. 4. [124]Fig. 4 [125]Open in a new tab 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. To validate the interaction between these two transcription regulators and the CISD2 promoter, we performed chromatin immunoprecipitation (ChIP) coupled with droplet digital PCR (ddPCR). Clustering CHIP-seq data related to TAF1 and TCF12 from public domain databases was able to provide their potential binding sites on the CISD2 promoter; primers for ddPCR were then designed accordingly to determine whether these sites are specifically enriched by ChIP. Our results revealed a significant binding of TCF12, but not TAF1, onto the CISD2 promoter in AC16 human cardiomyocytes, in which DOX treatment downregulated CISD2 expression ([126]Supplemental Fig. S2F). Intriguingly, the TCF12 binding signal was significantly reduced after DOX treatment ([127]Fig. 4B and C; [128]Supplemental Fig. S2G). Although it seems that TAF1 does not directly bind to CISD2 promoter, it is possible that TAF1 interacts with TCF12 and this resulting TAF1/TCF12 complex, which then binds to CISD2 promoter. To explore this possibility, we carried out a ChIP-reChIP experiment using AC16 human cardiomyocytes. Indeed, TAF1 appeared to interact with TCF12 and the paired proteins then activate the CISD2 promoter via binding to the TCF12 consensus sequence as a protein complex ([129]Fig. 4D and E). To study the cis-element of TCF12-binding site, we conducted a CRISPR interference (CRISPRi) study ([130]Supplemental Fig. S2H), which is a genetic manipulation technique derived from the CRISPR/Cas9 system. We designed 11 sgRNAs that are targeted at the CISD2 promoter region ([131]Supplemental Table S2) and transfected these sgRNAs into HEK293-CISD2 reporter cells. Our result revealed that sgRNA5 and sgRNA9, which have target sites close to the TCF12-binding site, significantly suppressed the transcription of the CISD2 reporter ([132]Supplemental Fig. S2I). To further investigate the relationship between the TCF12-binding site and CISD2 expression, we employed the CRISPR/Cas9 system to disrupt the DNA sequence of the TCF12-binding site in AC16 human cardiomyocytes. Two independent clones, namely clone #13 and clone #28, were isolated. Both clones are heterozygous lines that carry one WT allele and one mutant allele, the latter with a DNA deletion covering the TCF12-binding site located in the upstream region of the CISD2 promoter. Additionally, we have confirmed that there is no detectable off-target effect in the mutant clones ([133]Supplemental Fig. S2J–M). Strikingly, the expression levels of CISD2 drop down to ∼50 % in these two heterozygous mutant clones ([134]Fig. 4F), indicating that the cis-element of the TCF12-binding site is essential for the normal expression of CISD2. Together, these results suggest that TAF1 and TCF12 play an essential role in the activation of CISD2 transcription and that DOX downregulates the expression of CISD2 via decreasing the expression in cardiomyocytes of these two transcription activators, namely TAF1 and TCF12. 3.5. Decipher the molecular mechanism underlying the benefits that hesperetin confers in relation to DOX-induced cardiotoxicity through trans-omics analysis To elucidate the molecular mechanisms by which the CISD2 activator hesperetin confers cardio-protection against DOX-induced toxicity, we performed a time-course transcriptomic analysis of mouse hearts collected at 16 h (an early phase) and 72 h (a later phase) after a single dose injection of DOX (25 mg/kg i.p.). These timepoints were selected to capture gene expression changes preceding (16 h) and during (72 h) the development of an overt phenotype. It should be noted that CISD2 expression is significantly downregulated at both time points. At 16 h, differential expression analysis revealed 3587 genes altered (1943 downregulated; 1644 upregulated) by DOX compared to vehicle. Co-treatment of DOX with hesperetin (DOX + Hes) resulted in 376 differentially expressed genes (DEGs) (69 downregulated; 307 upregulated) compared to DOX alone. At 72 h, 4104 genes were differentially expressed in DOX-treated hearts (2054 downregulated; 2050 upregulated), while co-treatment of DOX with hesperetin (DOX + Hes) altered the expression of 2007 genes (916 downregulated; 1091 upregulated) ([135]Fig. 5A). Principal component analysis (PCA) and sparse partial least squares-discriminant analysis (sPLS-DA) revealed distinct clustering of the DOX + Hes group away from the DOX group and toward the vehicle control at both timepoints ([136]Fig. 5B, [137]Supplemental Fig. S3A). An unsupervised hierarchical heatmap of the top 100 DEGs further supported this observation, with the DOX + Hes transcriptome showing closer similarity to the vehicle controls. ([138]Fig. 5C). Notably, hesperetin reversed a subset of DOX-altered genes. At 16 h, 363 DOX-affected genes (303 upregulated; 60 downregulated) had their changes reversed by hesperetin. At 72 h, 1578 DOX-affected genes (858 upregulated; 720 downregulated) had their changes reversed by hesperetin ([139]Fig. 5D). Gene Set Enrichment Analysis (GSEA) indicated that DOX treatment led to upregulation of the gene sets related to apoptosis and inflammation, while hesperetin was found to promote oxidative phosphorylation (OXPHOS) and myogenesis-related gene sets ([140]Supplemental Fig. S3B). To further analyze the biological pathways affected by hesperetin, Ingenuity Pathway Analysis (IPA) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis were employed. Notably, hesperetin-reverted genes at both timepoints are associated with metabolic signaling (HIF-1), mitochondrial function and survival (calcium signaling, ROS metabolism, and necroptosis), and cardiac contractility pathways ([141]Fig. 5E and [142]Supplemental Fig. S3C–D). At 72 h, additional enrichment was observed in lipid and amino acid metabolism, PPAR and mTOR signaling, and autophagy-related survival pathways ([143]Fig. 5F). Interestingly, the temporal pathway comparison of the DOX and the DOX + Hes groups, namely 72 h vs. 16 h, revealed that dynamic transcriptional rewiring was taking place. Co-treatment with hesperetin of the DOX-injured heart appears to maintain the pathways related to metabolism, mitochondrial integrity, and cardiac function, suggesting a protective effect of hesperetin over time. In contrast, DOX alone led to progressive dysregulation of lipid metabolism, cellular senescence, and cardiac dysfunctions ([144]Supplemental Fig. S3E–G). In essence, the administration of DOX disrupts genes that are part of various critical pathways, including those governing heart contractions (TNC, MYOSIN, and ACTIN), calcium regulation (DGPR, PLN, CASQ2, ANT, and VDAC), mitochondrial functionality (ANT and VDAC), antioxidant mechanisms (MnSOD), apoptosis (CYTC), and cellular senescence (TGFb, and ING1), as well as inflammation and fibrosis (TGFb and NFkB) at both timepoints ([145]Supplemental Fig. S3H). As summarized in [146]Supplemental Fig. S3I, these DOX-induced abnormalities are likely to lead ultimately to dysfunction of the heart. Strikingly, hesperetin treatment appears to restore the expression levels of these dysregulated genes, thereby improving the functioning of the heart when mice are treated with DOX. Fig. 5. [147]Fig. 5 [148]Open in a new tab 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. Given that the DEGs reverted by hesperetin in DOX-treated hearts are primarily associated with metabolism ([149]Fig. 5E and F, [150]Supplemental Fig. S3C–D), we therefore applied metabolomic analysis to the cardiac tissues collected at 16 and 72 h after DOX treatment ([151]Fig. 6A). PCA of the metabolomic profiles at 16 h revealed no overt separation among the four groups, which suggests minimal metabolic alterations at this early timepoint ([152]Supplemental Fig. S4A). Remarkably, at 72 h, the DOX group exhibited distinct metabolomic profiling. Intriguingly, the metabolomic profiling of the DOX + Hes group is closer to the vehicle control, suggesting that hesperetin restored, at least in part, the metabolic dysregulation present. These observations were further supported by the PCA score plots, and hierarchical clustering of the metabolomic profiles ([153]Fig. 6B and C, [154]Supplemental Table S3). Fig. 6. [155]Fig. 6 [156]Open in a new tab 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. Since transcriptomic changes are expected to precede metabolic alterations, we integrated the 16-h transcriptomic datasets with the 72-h metabolomic datasets to perform a trans-omics analysis. To do this, we used MetaboAnalyst ([157]https://www.metaboanalyst.ca), a web-based platform for the comprehensive analysis and integration of multi-omics data. Interestingly, the trans-omics integration revealed that the DEGs and DEMs (differentially expressed metabolites) are involved in a wide range of metabolic processes, namely energy metabolism, amino acid metabolism, fatty acid metabolism, and carbohydrate metabolism. Furthermore, most of the DEM- and DEG-associated pathways seem to be linked with key molecules and these in turn are able to be modulated by hesperetin ([158]Fig. 6D). Furthermore, DOX severely damages pathways responsible for the energy supply within the heart. To illustrate the complex network of connections revolving around these DEGs and DEMs, we have summarized the relevant pathways associated with DOX treatment and with co-treatment with DOX and hesperetin ([159]Fig. 6E). Briefly, DOX impairs a number of antioxidant and energy metabolic pathways in the heart. Specifically, these are as follows. (1) In the cytosol, the antioxidation branch of pentose phosphate pathway (PPP) and the glutathione metabolism are dysregulated. The levels of glucose and NADP^+ are thus significantly decreased. However, the level of glutathione reductase (GSR), which catalyzes the reduction of oxidized glutathione (GSSG) to reduced glutathione (GSH) using NADPH as a cofactor, is significantly increased; this is accompanied by an increase in GSH. This suggests that DOX creates a consistent presence of ROS along with an aberrant regulation of antioxidation processes. (2) In the peroxisome, the β-oxidation of very long-chain fatty acids is disrupted by DOX. The levels of acyl-CoA and octanonyl-CoA are significantly decreased, while the level of octanonyl-carnitine is significantly increased. This suggests that there are abnormalities present that affect fatty acid metabolism. (3) Glycolysis in the cytosol, as well as the tricarboxylic acid (TCA) cycle and the electron transport chain in the mitochondria, are all dysregulated. Remarkably, many enzymes involved in glycolysis and TCA cycle are downregulated by DOX ([160]Supplemental Table S4–S6). Furthermore, the levels of many associated metabolites are disturbed, some of which are downregulated (e.g. 2 PG and PEP), while others are up-regulated (e.g. cis-aconitate and malate). Additionally, several metabolic regulation systems are disrupted by DOX. For example, a decrease in acetyl-CoA and its regulatory enzyme pyruvate dehydrogenase (PDH), a pivotal part of the metabolic hub, occurs. Furthermore, a decrease in pyruvate kinase M (PKM) along with an increase of pyruvate and lactate are observed, indicating that there is a dysregulation of pyruvate metabolism. Moreover, the enzyme responsible for converting fatty acids to acetyl-CoA, namely acetyl-CoA acetyltransferase (ACAT), is decreased. In addition, genes encoding components of complexes I and III of the electron transport chain (ETC) are found to be down-regulated by DOX. Intriguingly, hesperetin protects the heart against DOX-induced damage related to energy metabolism via the preservation of the operation of the pathways connecting the metabolism of glucose, fatty acids and amino acids, thereby ensuring an adequate supply of energy. In addition, hesperetin improves the antioxidation capacity of cells via reinstatement of the functioning of the PPP and glutathione antioxidant pathway. Moreover, we found that DOX treatment causes a decrease in the level of arginine, a known cardioprotective agent [[161]19], and that co-treatment with hesperetin is able to rescue the level of arginine. Together, these effects brought about by hesperetin with respect to maintaining the normal operation of the biological pathways within the heart provide an explanation for the cardioprotective effects of hesperetin ([162]Fig. 6E). To further evaluate the effect of transcriptomic changes on metabolic rewiring at a later stage, we performed trans-omics analysis by integrating the 72-h transcriptomic datasets with the 72-h metabolomic datasets. While most DEGs at 72 h overlapped with those at 16 h, several amino acid metabolic pathways, including those for cysteine (GOT1/2, MDH1), aspartate (GOT1/2, PCX), alanine (GPT), and branched-chain amino acids (BCKDH), were found to be further suppressed at 72 h. In addition, components of complexes II and IV of the ETC were also downregulated at the later timepoint. Interestingly, genes involved in lipid uptake were increased at 72 h in the DOX-treated group; this likely reflects a compensatory response to energy crisis brought about by DOX. Importantly, hesperetin appears to reverse these alterations ([163]Supplemental Fig. S4B, [164]Supplemental Table S7–S10). In summary, our data demonstrate that transcriptomic alteration precedes a metabolic shift. At an early stage, namely 16 h post-DOX treatment, there are already significant changes in transcriptomic profiling; however, the metabolomic profiling has undergone only minimal alteration. This supports the concept that early gene regulation events drive downstream metabolic remodeling. Intriguingly, hesperetin confers cardio-protection against DOX-induced injuries by preserving cardiac metabolic function and the redox balance. Moreover, trans-omics analysis of transcriptomic and metabolomic datasets further revealed that hesperetin preserves mitochondrial function, enhances antioxidant defenses, and maintains bioenergetic homeostasis, thereby mitigating DOX-induced metabolic dysfunction in the heart. 3.6. Hesperetin ameliorates DOX-induced cellular damage and improves the functioning of human iPSC-CMs Induced pluripotent stem cell-derived cardiomyocytes (iPSC-CMs) have been previously demonstrated to exhibit structural and functional similarities to in vivo cardiomyocytes. Recent studies have further established that iPSC-CM derived from patients are able to replicate the phenotypes of DOX-induced cardiotoxicity observed in animal models [[165]17]. Accordingly, we employed iPSC-CMs as our cellular model in order to investigate whether the beneficial effects of hesperetin can be reproduced using human cardiomyocytes. The iPSC-CMs used in this study were generated by following previously described protocols, and the drug treatment procedure is as outlined in [166]Fig. 7A. Consistent with our findings in mice, DOX also downregulated CISD2 expression, while hesperetin restored the level of CISD2, when human iPSC-CMs were treated with DOX ([167]Fig. 7B). In addition, hesperetin significantly rescued DOX-induced mitochondrial depolarization as revealed by JC-1 staining of mitochondrial membrane potential ([168]Fig. 7C). Furthermore, hesperetin significantly decreased the cytosolic and mitochondrial ROS levels induced by DOX ([169]Fig. 7D and E). To assess whether hesperetin is able to sustain Ca^2+ homeostasis when this is disrupted by DOX, we carried out Ca^2+ wave analysis using Fura-2/AM staining of the beating iPSC-CMs ([170]Fig. 7F). Quantification of the data revealed that the DOX-mediated dysregulation of Ca^2+ waves was significantly improved in iPSC-CMs treated with hesperetin ([171]Fig. 7G). Additionally, DOX treatment disrupted sarcomeric organization, which is accompanied by the complete disappearance of the actin cable and Troponin I expression. Notably, hesperetin treatment appears to partially restore sarcomeric organization ([172]Fig. 7H). Moreover, DOX-induced cell death was attenuated by hesperetin as revealed by a significant increase in the cell survival rate ([173]Fig. 7I). Fig. 7. [174]Fig. 7 [175]Open in a new tab 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 Ca^2+ waves of beating iPSC-CM were measured by confocal microscopy using fura-2/AM staining. Total peak area and maximal Ca^2+ 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. In summary, hesperetin acts as a potent CISD2 activator and this enhances the expression level of CISD2 in iPSC-CMs, which is otherwise downregulated by DOX. Consequently, as demonstrated in mice, enhanced CISD2 expression appears to protect human iPSC-CMs, a cell model that mimics cardiomyocytes in human patients, from DOX-induced cellular toxicity and functional damage. 4. Discussion This study demonstrates that sustained upregulation of Cisd2 expression, which can be achieved either by transgenic overexpression or by pharmacological activation, is able to significantly improve cardiac function under DOX treatment. As a proof-of-concept for translational medicine, we have used a potent Cisd2 activator, hesperetin, to reveal that enhancing Cisd2 is a promising strategy for improving the outcome when there is potential DOX-induced heart failure. Five findings can be pinpointed in this study. Firstly, doxorubicin suppresses Cisd2 expression and causes a variety of cellular and ultrastructural damage; these result in cardiac electromechanical dysfunction. Intriguingly, transgenic overexpression of Cisd2 in mice mitigates this DOX-induced cardiotoxicity. Secondly, hesperetin effectively sustains a high level of Cisd2 and this improves cardiac structure and function under DOX treatment. Importantly, hesperetin functions in a Cisd2-dependent manner, as hesperetin loses its beneficial effects in Cisd2cKO mice. Additionally, hesperetin treatment does not influence the anti-cancer efficacy of DOX in tumor-bearing mouse model. Thirdly, DOX downregulates CISD2 expression via a decrease in two transcription regulators, TAF1 and TCF12, which play an essential role in the transcription of human CISD2. Fourthly, trans-omics analysis of cardiac transcriptomic and metabolomic datasets revealed that hesperetin protects the heart against DOX-induced disturbance of a network of pathways that connect glucose metabolism, fatty acid metabolism and amino acid metabolism, thereby ensuring a sufficient energy supply to the heart. Additionally, hesperetin improves the antioxidation capacity of the heart via reinstating the pentose phosphate pathway and glutathione antioxidant pathway. Finally, in human iPSC-CMs, hesperetin significantly upregulates CISD2 expression, thus protecting the cells from DOX-induced cellular toxicity and functional damage. All together, these findings underscore the crucial role of CISD2 in mitigating DOX-induced cardiotoxicity and highlight the potential utility of the CISD2 activator hesperetin in a clinical setting involving DOX chemotherapy. 4.1. CISD2 protects mitochondria from DOX-induced injury in cardiomyocytes Anthracyclines, such as DOX, are recognized for their efficacy when they are used as chemotherapeutic agents; they primarily act through the inhibition of topoisomerase 2 alpha (Top2a), which leads to the induction of double-stranded DNA breaks and cessation of cell division [[176]6]. However, their clinical use is limited by the presence of cumulative, dose-dependent cardiotoxicity that is now recognized to affect myocardial mitochondrial functioning [[177]20]. These effects include the suppression of the ETC, increased production of ROS, the creation of Ca^2+ overload, the disruption of iron uptake and storage, and the induction of DNA damage via complex formation with topoisomerase [[178][21], [179][22], [180][23], [181][24]]. Transfer of mesenchymal stem cell-derived vesicle-bound viable mitochondria restores DOX-damaged mitochondrial function in cardiomyocytes [[182]25]. DOX also initiates an autophagic response that aims to remove damaged mitochondria [[183]26]. This process involves altering mitochondrial dynamics, activating autophagy [[184]27,[185]28], and inhibiting lysosomal function [[186]29,[187]30], which leads to an accumulation of autolysosomes [[188]31] and an impaired autophagic flux. Additional layers of regulation involve non-coding RNAs (e.g., microRNAs, lncRNAs) that modulate oxidative stress and apoptosis [[189]32,[190]33]. Canonical apoptotic signals driven by p53 activation, ROS generation, and Bax translocation further exacerbate cardiomyocyte loss, whereas interventions targeting Bax, infiltrating macrophages, or APE1/Ref-1 have been shown to attenuate DOX injury [[191][34], [192][35], [193][36]]. Targeting the pathways with small compounds or natural compounds is a promising strategy to relieve cardiotoxicity [[194]36,[195]37], and this will then increase the usability of DOX in clinical cancer patients. Our previous studies have revealed that a consistently high level of Cisd2 is able to preserve mitochondrial function, maintain intracellular Ca^2+ homeostasis and provide ROS buffering capacity in cardiomyocytes during aging [[196]14,[197]38,[198]39]. Furthermore, we have extended these genetic findings to pharmaceutical intervention and have demonstrated that dietary supplementation with the Cisd2 activator, hesperetin, mitigates age-related organ dysfunction [[199]15,[200]40]. Here we have demonstrated that pharmaceutical activation of Cisd2 expression by hesperetin in a mouse model and in human iPSC-CMs is capable of counteracting the reduction of Cisd2 brought about by DOX treatment and this then alleviates the adverse effects brought about by DOX. Although Cisd2 has been reported to be a double-edged sword in cancer [[201]41,[202]42], our data demonstrate that Cisd2 activation in tumor-bearing mice mitigates DOX-induced cardiotoxicity without compromising its anticancer efficacy. Thus, Cisd2 activators, such as hesperetin or curcumin [[203]43], offer a promising avenue for mitigating DOX-induced cardiotoxicity. 4.2. The role of CISD2 in Ca^2+ homeostasis under DOX treatment The presence of DOX-induced electrical abnormalities, such as a prolonged QTc, QRS interval, and Tpeak-Tend interval, suggests that the injury caused by DOX is primarily a consequence of reduced Ca^2+ transients in cardiomyocytes [[204]44]. DOX-induced intracellular elevation of the concentration of Ca^2+, which results in an increase in ROS level and is followed by apoptosis, is both a cause and a consequence of DOX-induced cardiotoxicity [[205]45]; this can be treated with a Ca^2+ chelator [[206]46]. This Ca^2+ disequilibrium is associated with decreased Serca2a function; as a consequence, this further leads to an enhancement of Ca^2+ leakage from the sarcoplasmic reticulum of the heart [[207]47]. In addition, DOX treatment elevates cytosolic Ca^2+ levels in cardiomyocytes, which subsequently activates protein kinase C alpha (PKCα). This activation is crucial for the phosphorylation regulation of desmoplakin, an essential component involved in desmosome dynamics and internalization, which is necessary for maintaining the structural integrity and signal transduction within the heart [[208]48]. Furthermore, an increase in cytosolic Ca^2+ level affects the expression and distribution of Cx43, a vital gap junction protein that facilitates cell-to-cell communication in the heart. An alteration in Cx43 expression and localization has been linked to various cardiac diseases, including hypertrophic cardiomyopathy [[209]49], heart failure [[210]50], and ischemic cardiomyopathy [[211]51]. These observations underscore the significant impact of DOX-induced cytosolic Ca^2+ changes on both Cx43 behavior and the broader pathophysiology of heart diseases. One of our previous studies has shown that a sustained high expression of Cisd2 prevents Serca2a oxidative modification and this maintains its function during cardiac aging [[212]14]. Here we provide evidence that the DOX-induced cellular and ultrastructural damage, includes (a) lateralization of gap junctions as revealed by Cx43 IF staining, (b) structural defects of intercalated discs as revealed by the co-localization coefficient of Cx43 and pan-cadherin, (c) disorganization of the fascia adherens and fragmentation of gap junction as revealed by TEM, and (d) the presence of degenerated and swollen mitochondria, along with degenerated myofibrils, as revealed by TEM. All these structural abnormalities caused by DOX are effectively mitigated by hesperetin; consequently, cardiac mechanical contraction, which is triggered by electrical activation, namely electromechanical coupling, is largely preserved when there is co-treatment with DOX and hesperetin. 4.3. Activation of CISD2 ameliorates DOX-induced dysregulation of metabolism Previous studies have demonstrated that DOX induces metabolic reprogramming in cardiomyocytes [[213]52]. This reprogramming is accompanied by ROS production, which results in mitochondrial dysfunction, reduces glycolysis, perturbs fatty acid metabolism, and impairs OXPHOS [[214][52], [215][53], [216][54], [217][55]]. Additionally, the interaction between DOX and the topoisomerase IIβ complex is able to inhibit mitochondrial biogenesis and related gene expression, such as Wt1, thereby suppressing OXPHOS and decreasing energy production [[218]53,[219]56]. Indeed, our findings reveal that DOX downregulates a panel of DEGs that are involved in various pathways important to the heart; these include pathways related to metabolism, cell survival, inflammation, ROS and mitochondrial function, as well as pathways associated with cardiac function. 4.4. Limitations of this study This study has several limitations. Firstly, all the mice used in this study are males since our initial pilot experiments revealed that female mice exhibited only minor DOX-induced cardiac injury compared to males. Accordingly, we focused our study on male mice. Thus this design precludes analysis of sex-specific differences in CISD2 expression, in DOX metabolism, and in cardio-protective responses. This may be particularly important given the well-documented impact of sex differences on cardiovascular disease presentation, drug response, and outcomes [[220][57], [221][58], [222][59]]. In females, estrogen is known to enhance mitochondrial function and antioxidant capacity; thus this may potentially reduce their susceptibility to DOX-induced cardiotoxicity. On the other hand, in males, testosterone seems to have adverse effects on cardiac redox homeostasis [[223]60]. Secondly, sex differences in pharmacokinetics and pharmacodynamics can influence both the toxicity and efficacy of DOX and cardio-protective agents [[224]61]. Thirdly, the acute, high-dose DOX model used here does not fully replicate the chronic, cumulative exposure patterns typical in clinical oncology, nor does it account for sex-dependent differences in drug response and the tumor microenvironment [[225]59]. The tumor-bearing mouse model may also not adequately capture the complexity of the human tumor microenvironments and immune interactions, which are increasingly recognized as sex-influenced. Lastly, the current mechanistic study focused on TAF1/TCF12 transcriptional regulation of CISD2 influenced by DOX. It does not address other potential possibilities such as post-translational modifications of CISD2 or alternative regulatory pathways that are activated during prolonged DOX exposure. Future research could tackle sex difference-related effects, employ chronic dosing regimens, and utilize sex-matched human iPSC-derived cardiomyocytes to clarify how sex-specific factors modulate CISD2's cardio-protective effects; this will help with the development of personalized therapy. 5. Conclusion and perspectives This study provides compelling evidence that DOX downregulates CISD2 expression and that upregulation of CISD2 expression is an effective approach to preserving cardiac function in the context of DOX chemotherapy. This highlights a potential strategy for mitigating DOX-induced heart failure. The pivotal role of CISD2 in cardioprotection was demonstrated here via both genetic and pharmacological approaches. Moreover, the identification of TAF1 and TCF12 as upstream regulators of the transcription of CISD2 gene offers a promising avenue for further investigation into their roles in DOX-mediated CISD2 downregulation and cardiotoxicity. Ultimately, this study underscores the importance of maintaining a high level of CISD2 for protecting the heart from DOX-induced toxicity and for preserving cardiac function when using DOX as a chemotherapeutic agent. Our findings provide experimental evidence and lay the groundwork for future clinical applications and interventions using the CISD2 activator hesperetin, a functional food originally derived from the peels of citrus fruits, to combat DOX-induced heart failure. CRediT authorship contribution statement Yi-Ju Chou: Writing – original draft, Methodology, Investigation, Formal analysis, Data curation. Chi-Hsiao Yeh: Writing – original draft, Supervision, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization. Chian-Feng Chen: Validation, Methodology, Investigation, Formal analysis, Data curation. Chi-Jen Lo: Methodology, Investigation, Data curation. Jian-Hsin Yang: Methodology, Data curation. Wen-Tai Chiu: Validation, Supervision, Methodology, Investigation, Data curation. Cheng-Heng Kao: Supervision, Methodology, Formal analysis, Data curation. Tsai-Yu Tzeng: Methodology, Investigation. Zhao-Qing Shen: Methodology, Data curation. Chien-Yi Tung: Methodology, Formal analysis. Chung-Kuang Lu: Resources, Methodology. Mei-Ling Cheng: Methodology, Investigation, Formal analysis, Data curation. Patrick C.H. Hsieh: Supervision, Investigation, Conceptualization. Shu-Ling Fu: Writing – review & editing, Validation, Supervision, Funding acquisition, Formal analysis, Data curation, Conceptualization. Ting-Fen Tsai: Writing – review & editing, Validation, Project administration, Investigation, Funding acquisition, Formal analysis, Conceptualization. Data statement The data supporting the findings of this study are available from the corresponding authors upon reasonable request. Disclosures The authors declare no conflict of interest. Sources of funding This research was funded by grants from National Science and Technology Council (MOST 108-2320-B-182A-003, MOST 109-2320-B-182A-020 and MOST 110-2320-B-182A-021 to CHY; MOST 108-2320-B-010-009, MOST 109-2320-B-010-043 and MOST 110-2320-B-A49A-543 to TFT; MOST 108-2320-B-010-010, MOST 109-2320-B-010-044 and MOST 110-2320-B-A49A-537 to SLF), and from the Ministry of Health and Welfare (NHRI-11A1-CG-CO-07-2225-1, NHRI-12A1-CG-CO-07-2225-1, NHRI-13A1-CG-CO-07-2225-1 and NHRI-14A1-CG-CO-07-2225-1 to TFT), as well as partially supported by the Interdisciplinary Research Center for Healthy Longevity of National Yang Ming Chiao Tung University from The Featured Areas Research Center Program within the framework of the Higher Education Sprout Project by the Ministry of Education (MOE) in Taiwan. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgments