Osimertinib-induced cardiotoxicity is driven by HDAC-dependent epigenetic repression and rescued by vorinostat

Abstract

Osimertinib, a third-generation EGFR tyrosine kinase inhibitor (TKI), has improved outcomes in non-small cell lung cancer (NSCLC) patients harboring the T790M mutation; however, emerging clinical evidence indicates a risk of cardiotoxicity. Here, we establish the first in vivo preclinical model of osimertinib-induced cardiotoxicity using transverse aortic constriction (TAC) in mice. Osimertinib treatment resulted in cardiac dysfunction, impaired hypertrophic remodeling, and increased markers of heart failure and fibrosis. Unbiased transcriptomic profiling revealed a myocardial stress response characterized by activation of p53-associated cell death pathways, mitochondrial dysfunction, and negative enrichment of histone acetyltransferase (HAT) complexes, indicating epigenetic repression. Mechanistically, osimertinib-treated hearts exhibited increased expression of multiple histone deacetylase (HDAC) isoforms, reduced acetylation of histones, and enhanced cardiomyocyte apoptosis via Bax/caspase-mediated pathways. There was a minimal, transient effect on inflammation, supporting a type I, cell-autonomous cardiotoxic mechanism. Consistent with this, in vitro and in vivo analyses demonstrated suppression of prosurvival ERK/AKT signaling, mitochondrial dysfunction, and activation of intrinsic apoptotic pathways. Given the central role of HDAC activation, we tested whether pharmacologic HDAC inhibition could mitigate osimertinib-induced cardiotoxicity. Treatment with the FDA-approved HDAC inhibitor vorinostat (SAHA) restored histone acetylation, attenuated p53 activation, reduced cardiomyocyte death, and rescued cardiac function in osimertinib-treated mice. Translational studies in human NSCLC-derived PC9 cells further demonstrated that SAHA enhances osimertinib antitumor efficacy while alleviating cardiotoxicity. Collectively, these findings define HDAC-dependent epigenetic repression as a key mechanism underlying osimertinib-induced cardiotoxicity and identify HDAC inhibition as a therapeutically actionable strategy to improve both cardiac safety and cancer treatment efficacy.

Introduction

Cardiovascular (CV) toxicity is one of the major side effects of all kinds of anticancer therapies, including chemotherapy, radiation, and targeted therapy.^1–3 Adverse cardiac effects of non-targeted therapies are well-recognized and extensively studied. Tyrosine kinase inhibitors (TKIs) have emerged as a leading targeted cancer therapy, revolutionizing the treatment of various malignancies. These small-molecule drugs inhibit tyrosine kinases that activate proteins involved in cell signaling pathways regulating cell growth, differentiation, and survival.^4,5 Recent improvements in TKIs have enhanced their efficacy and led to more favorable clinical outcomes.^6,7 TKIs are generally associated with fewer adverse effects than non-targeted cancer therapies.^8–10 However, these newer targeted therapies are not the magic bullets, and there are multiple reports of adverse effects, including cardiotoxicity.^11–13 For instance, these TKIs are related to multiple incidences of cardiotoxicity, such as hypertension, cardiomyopathy, QTc prolongation, and myocardial ischemia.¹⁴ However, the mechanism and frequency of targeted therapies-induced cardiotoxicity are largely unknown and likely underestimated.

Osimertinib is a TKI approved to treat non-small cell lung cancer (NSCLC) in adults whose tumors have specific epidermal growth factor receptors (EGFR) gene mutations.^15,16 Specifically, osimertinib is the most effective treatment option for EGFR-NSCLC patients with EGFR T790M resistance mutation to all other approved EGFR TKIs.^15,16 Indeed, compared to previous generations of EGFR-TKI, osimertinib has exhibited improved outcomes for NSCLC patients.¹⁶ However, it is associated with a higher risk of cardiotoxicity.¹⁷ Of note, the cardiotoxicity of osimertinib is well-recognized in real-world clinical scenarios of human patients.¹⁸ Osimertinib clinical trials reveal that 2.6% of 1,142 patients on osimertinib develop cardiomyopathies, and 3.9% of 908 osimertinib-treated patients developed cardiac dysfunction as marked by a > 10% decrease in ejection fraction (EF).^18,19 Thus, the key adverse cardiac events associated with osimertinib are decreased left ventricular ejection fraction (LVEF) and the development of heart failure.^18,20,21 Additionally, among patients with EGFR-mutant NSCLC, osimertinib is associated with a higher incidence of cancer therapy-related cardiac events (CTRCEs) compared with other EGFR TKIs.²⁰ Despite the well-established cardiotoxicity of osimertinib in clinical settings, there is no reported in vivo preclinical modeling to investigate the underlying mechanism and potential strategies to mitigate osimertinib-induced cardiotoxicity.

Herein, we present the first optimized in vivo preclinical mouse model that recapitulates the cardiac dysfunction observed with osimertinib treatment. Using integrated in vivo, in vitro, and transcriptomic approaches, we identify maladaptive stress signaling, mitochondrial dysfunction, and epigenetic remodeling as central features of osimertinib-induced cardiotoxicity. Mechanistically, osimertinib disrupts mitochondrial integrity, suppresses ERK/AKT prosurvival signaling, and activates cardiomyocyte apoptotic pathways. Based on our unbiased gene expression analysis from osimertinib-treated heart samples, validation with western blot analysis, and emerging evidence implicating histone deacetylase (HDAC) activation in cardiac stress responses, we evaluated the FDA-approved HDAC inhibitor vorinostat (SAHA). We demonstrate that HDAC inhibition preserves cardiomyocyte survival and cardiac function during osimertinib treatment, supporting epigenetic regulation as a therapeutically actionable strategy to mitigate osimertinib-associated cardiotoxicity.

Results

Osimertinib-treated animals developed cardiac dysfunction and adverse remodeling

Osimertinib is cardiotoxic to humans in the real-world patient scenario.^{17,18,22–25} However, there is no in vivo preclinical model to investigate osimertinib-induced cardiotoxicity. Given clinical evidence demonstrating that pre-existing cardiovascular comorbidities increase susceptibility to osimertinib-associated heart failure,^{17,18,22–25} to model a clinically relevant comorbid setting and unmask latent cardiotoxicity, 8-week-old C57BL/6J mice were subjected to transverse aortic constriction (TAC) surgery to establish a cardiac-specific pressure-overload model, followed by osimertinib treatment. One week post-TAC, mice were given osimertinib (25 mg/kg/day, 4 weeks, oral gavage). As optimized by the osimertinib developer pharmaceutical company, AstraZeneca, once-daily administration of 25 mg/kg of osimertinib to mice approximates the clinically approved 80-mg dose to patients.²⁶ Echocardiography was performed to assess the heart function at baseline, 2 and 4 weeks post-osimertinib treatment (Fig. 1a). There was no significant difference in body weight among the different groups (Fig. 1b). Echocardiography assessment revealed a significant decline in ejection fraction (EF) and fractional shortening (FS), indicating cardiac dysfunction in osimertinib-treated TAC hearts compared to placebo-treated TAC hearts (Fig. 1c, d). Consistently, left ventricle end-systolic interior dimension (LVIDs) was significantly increased, indicating the development of excessive dilative cardiac remodeling in the osimertinib-treated TAC hearts compared to respective controls (Fig. 1e, f). Importantly, Osimertinib did not lead to cardiac dysfunction in the Sham groups, indicating the critical role of comorbidity in osimertinib-induced cardiotoxicity (Fig. 1c–f). Given reported sex-specific differences in cardiac responses to stress and injury, we evaluated left ventricular function separately in male and female mice following treatment. EF and FS were analyzed separately in male and female mice. At 2 weeks post-treatment, male mice did not exhibit significant alterations in cardiac function; however, by 4 weeks, both male and female mice showed a reduction in EF and FS, indicating development of cardiac dysfunction in both sexes (Supplementary Fig. 1a–d). To assess gene expression of cardiac damage markers and histological analysis, hearts were collected after 4 weeks of osimertinib treatment. The expression of signature genes related to pathological cardiac hypertrophy, ANP, and BNP was significantly increased in the osimertinib-treated TAC hearts compared to controls (Fig. 1g, h). Consistently, Masson’s Trichrome staining revealed excessive fibrotic remodeling of osimertinib-treated hearts at 4 weeks post-treatment but not at 2 weeks (Fig. 1i, j). Additionally, the expression of signature genes related to fibrosis (COL1A1 and COL3A1) was significantly increased in the osimertinib-treated TAC hearts compared to controls (Fig. 1k, l). Taken together, these data indicate that osimertinib treatment leads to adverse cardiac remodeling and cardiac dysfunction.

Fig. 1.

Osimertinib treatment induces cardiac dysfunction after TAC surgery. a Experimental design schematic created with BioRender.com. C57BL/6J mice were subjected to TAC or SHAM surgery, and after one week, they were treated with a placebo or osimertinib (25 mg/kg/day) for four weeks. Echocardiography was performed at baseline, two, and four weeks post-treatment. n = 10–15 mice per group. b Weight. c Ejection fraction (EF). d Fractional shortening (FS). e LV end-diastolic interior dimension (LVID;d). f LV end-systolic interior dimension (LVID;s). Four weeks after treatment, heart tissue was harvested for histological analysis, and RNA was extracted from the LV for gene expression analysis via qPCR. g ANP gene expression. h BNP gene expression. Assessment of cardiac fibrosis by Masson’s Trichrome staining; i Representative Trichrome-stained LV regions and j Quantification of LV fibrosis. Scale bar = 50 µm. k COL1A1 gene expression. l COL3A1 gene expression. Each data point represents an individual mouse. Data were analyzed using Two-way ANOVA followed by Tukey’s post hoc analysis and represented as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

Osimertinib-impaired TAC-induced adaptive hypertrophy

The TAC-induced elevated hemodynamic stress initially triggers a compensatory hypertrophic response characterized by increased cardiomyocyte size and myocardial mass. However, sustained pressure overload ultimately drives a transition to maladaptive remodeling, including dilation and dysfunction.²⁷ To examine the osimertinib-induced cardiac hypertrophy, we stained heart sections with Masson’s Trichrome staining and quantified the CM’s cross-sectional area (CSA). Surprisingly, osimertinib-treated hearts demonstrated impaired TAC-induced cardiomyocyte cellular hypertrophic growth. Indeed, as expected, the placebo-treated TAC heart showed hypertrophic growth (Fig. 2a, b). A complementary method to assess cardiac hypertrophy is measuring the heart weight to tibia length ratio (HW/TL). Consistent with cellular hypertrophy data, the HW/TL ratio demonstrated impaired hypertrophic growth in the osimertinib-treated TAC group compared to controls (Fig. 2c).

Fig. 2.

TAC-induced hypertrophy growth is impaired by osimertinib. Morphometric studies were performed four weeks after TAC surgery and osimertinib treatment. a Representative image of Masson’s Trichrome-stained heart sections. b Quantification of cardiomyocytes' cross-sectional area (CSA). Scale bar = 50 µm. n = 5–7 mice per group. Assessment of cardiac hypertrophy; c Heart weight (HW) to tibia length (TL) ratio. n = 5–19 mice per group. Scale bar = 50 µm. Multiple sections were analyzed per heart and averaged; n represents the number of independent hearts. Data were analyzed using Two-way ANOVA followed by Tukey’s post hoc analysis and represented as mean ± SD. *p < 0.05, **p < 0.01, ****p < 0.0001

Osimertinib induces maladaptive transcriptional reprogramming in pressure-overloaded hearts

Since osimertinib treatment resulted in cardiac dysfunction and blunted hypertrophic remodeling, we next performed unbiased transcriptomic profiling to define the molecular mechanisms underlying this maladaptive response. RNA-sequencing was performed on heart tissues from TAC-placebo and TAC-osimertinib groups. Principal component analysis (PCA) demonstrated a clear and robust segregation between the two groups, indicating a distinct transcriptional reprogramming induced by osimertinib treatment (Fig. 3a).

Fig. 3.

Pro-apoptotic pathway activation occurs with HDAC suppression and increased histone acetylation. We performed unbiased transcriptomic profiling to define the molecular mechanism underlying osimertinib-induced cardiotoxicity. RNA-sequencing was performed on heart tissues from TAC-Placebo and TAC-Osimertinib groups. a Principal component analysis (PCA). b Differential gene expression analysis-volcano plot. c Functional pathway enrichment analysis of differentially expressed genes in TAC-Osimertinib group, Canonical pathways (Reactome). d Gene Ontology (biological process). e Gene set enrichment analysis. f–p Western blot analysis was performed to assess HDACs expression and histone acetylation in osimertinib-treated NRVMs. NRVMs were treated with vehicle (DMSO) and 2.5 µM of osimertinib for 2 h. f Representative western blot showing protein levels of HDACs 1–7, H3K9-Ac, and Total Histone H3 with β-Tubulin as loading control. g–n Densitometric analysis of blots and quantification of the blot in (f). g HDAC7 expression quantification. h HDAC6 expression quantification. i HDAC5 expression quantification. j HDAC4 expression quantification. k HDAC3 expression quantification. l HDAC2 expression quantification. m HDAC1 expression quantification. n H3K9-Ac/total-H3 ratio expression quantification. o Representative western blot showing protein levels of H2AK5-Ac and Total Histone H2A with β-Tubulin as loading control. The quantification plot shows the H2AK5-Ac/total-H2A ratio post-densitometric analysis of blots. p Representative western blot showing protein levels of H4K8-Ac and Total Histone H4 with β-Tubulin as loading control. The quantification plot shows the H4K8-Ac/total-H4 ratio post-densitometric analysis of blots. RNA-sequencing was performed on heart tissue from 5 mice; Protein lysates were prepared from 6 independent DMSO or osimertinib-treated NRVM lysates; each data point represents one biological replicate (individual lysate). Data were analyzed using an unpaired t-test and represented as mean ± SD (g–n, o, and p). *p < 0.05, **p < 0.01, ****p < 0.0001

Differential gene expression analysis, visualized by a volcano plot, revealed significant upregulation of genes associated with cell death, mitochondrial dysfunction, and stress responses, whereas multiple genes involved in cell survival and adaptive remodeling were downregulated in the TAC-osimertinib group compared with TAC-placebo controls (Fig. 3b). To further define the biological pathways affected, we performed gene set enrichment analysis (GSEA) which revealed significant enrichment of several stress- and cell death–related pathways in osimertinib-treated hearts (Fig. 3c, d), including regulation of oxidative stress-mediated programmed cell death, activation of p53 signaling pathways (stabilization of p53, p53-dependent G1/S DNA damage response, and p53-dependent G1/S checkpoint control), and pathways related to mitochondrial stress and quality control. Consistent with this, GSEA also revealed enrichment of multiple catabolic and atrophy-associated pathways, including mitophagy, macroautophagy, and cellular responses to hypoxia. The activation of these pathways suggests that osimertinib promotes a shift toward metabolic stress adaptation and organelle turnover rather than anabolic growth, which likely contributes to the impaired hypertrophic response observed in vivo. In parallel, pathway enrichment analysis showed significant downregulation of gene programs involved in chromosome organization and DNA metabolic processes, further supporting the presence of defective adaptive transcriptional and proliferative remodeling. Collectively, these transcriptomic changes indicate that osimertinib reprograms the myocardial response to pressure overload away from adaptive hypertrophy and toward a state characterized by cellular stress, autophagy, mitochondrial dysfunction, and activation of cell death pathways, thereby providing mechanistic insight into its deleterious effects beyond classical hypertrophic signaling.

Furthermore, GSEA also revealed significant negative enrichment of the histone acetyltransferase (HAT) complex in osimertinib-treated hearts (Fig. 3e), suggesting a shift toward a more transcriptionally repressive chromatin landscape and implicating epigenetic remodeling as an additional regulatory mechanism contributing to the observed maladaptive gene expression profile. Consistent with this finding, osimertinib treatment in NRVMs markedly increased the expression of multiple HDAC isoforms, including HDAC3, HDAC5, HDAC6, and HDAC7 (Fig. 3f–m). This was accompanied by a significant reduction in acetylation of key histones associated with active transcription, including H3K9-Ac, H2AK5-Ac, and H4K8-Ac (Fig. 3f, n–p), indicating enhanced chromatin deacetylation.

Osimertinib induces robust cardiomyocyte (CM) death

Because RNA-sequencing and pathway analysis identified prominent enrichment of apoptotic and stress-response programs in osimertinib-treated hearts, we next investigated whether these transcriptional changes were accompanied by increased cardiomyocyte death in the osimertinib-treated TAC hearts. Indeed, CM loss plays a crucial role in the pathogenesis of heart failure. Apoptosis can contribute to the loss of CM’s, which can ultimately lead to heart failure.²⁸ Therefore, we decided to perform a TUNEL assay to determine CM cell death in vivo. Only cells that were both TUNEL positive and α-sarcomeric actin positive (cardiomyocyte marker), with TUNEL colocalizing with 4,6-diamidino-2-phenylindole (DAPI), were counted as apoptotic cardiomyocytes. TUNEL assay of osimertinib-treated hearts identified a significant increase in CM apoptosis as early as 2 weeks post-osimertinib treatment (Fig. 4a–c). Guided by our RNA-sequencing findings and observed CM’s apoptosis, we examined key components of the apoptotic pathway. Western Blot analysis revealed a significant increase in pro-apoptotic BAX expression (Fig. 4d, e), along with elevated levels of cleaved caspase-3/caspase-3 and cleaved caspase-8/caspase-8 ratios in osimertinib-treated hearts (Fig. 4d, f, g). These findings suggest that osimertinib induces excessive cardiomyocyte death by recruiting Bax/caspase-mediated classical apoptotic pathways.

Fig. 4.

Osimertinib enhances cardiomyocyte cell death and activates cell death pathways. a–c Assessment of apoptotic cell death by TUNEL staining was performed at two weeks after TAC surgery and osimertinib treatment. a Representative full-field images (40×) of TUNEL-stained heart sections with clearly indicated regions of interest. Scale bar = 50 µm. b Magnified views of the selected areas. Scale bar = 50 µm. c Quantification of TUNEL-positive nuclei in heart sections. n = 5–6 hearts per group with 10 random fields analyzed per heart. Data were analyzed using Two-way ANOVA followed by Tukey’s post hoc analysis and represented as mean ± SD. d–g Western blot analysis was performed to detect the activity of apoptotic signaling pathways in osimertinib-treated hearts. Lysates were collected from the LV of TAC placebo and TAC osimertinib-treated mice, followed by western blot analysis. d Representative Western blot showing protein levels of BAX, Total Caspase-3 & 8, and Cleaved caspase-3 & 8 with GAPDH as loading control. e BAX/GAPDH expression quantification. f Casp-3/Cleaved Casp-3 expression quantification. g Casp-8/Cleaved Casp-8 expression quantification. h–k NRVMs were treated with vehicle (DMSO), ponatinib, imatinib, and various concentrations of osimertinib as indicated, and cell viability was measured using the CellTiter-Glo luminescence assay. h Concentration-dependent cell death after 48 h of treatment. i NRVM’s time kinetics. After 48 h of vehicle (DMSO) or osimertinib treatment, cells were collected and digested to get a single-cell suspension, and flow cytometry analysis was carried out. j Representative figure of flow cytometry showing a gating strategy. k NRVM’s percentage of pre-apoptotic cells (Annexin V+7AAD−). l, m HiPSC-CMs were treated with vehicle (DMSO) or osimertinib (100 nm or 500 nm) for 48 h, cells were collected and digested to get a single-cell suspension, and flow cytometry analysis was carried out. l Representative figure of flow cytometry showing a gating strategy. m HiPSC-CMs percentage of apoptotic cells (Annexin V+7AAD+). Protein lysates were prepared from the hearts of independent mice; each data point represents one biological replicate (individual lysate). Each data point represents an individual mouse. Data were analyzed using unpaired t-test, One-way, and Two-way ANOVA followed by Tukey’s post hoc analysis and represented as mean ± SD. *p < 0.05, **p < 0.01, ****p < 0.0001

Osimertinib exhibits type 1 cardiotoxicity (direct) by suppressing central cardiac prosurvival ERK/AKT signaling

Unbiased transcriptomic profiling and functional data revealed activation of cell death and stress-response pathways without sustained inflammatory enrichment. We therefore evaluated the contribution of inflammation to osimertinib-induced cardiotoxicity. Our prior work in a cardio-oncology model of the TKI ponatinib identified excessive inflammation as a central driver of cardiac injury.²⁹ To test if similar pathways are also operational in osimertinib-induced cardiotoxicity, we examined markers of inflammation using qPCR and flow cytometry. Consistent with preserved cardiac structure and function at 1-week post-treatment (Supplementary Fig. 2), expression of pro- and anti-inflammatory markers was comparable in the osimertinib and control groups (Supplementary Fig. 3a–i). A modest but significant elevation in a few inflammation markers was observed transiently at the 2-week time point (Supplementary Fig. 3j–r). However, by four weeks, expression of all-inflammatory markers returned to baseline levels (Supplementary Fig. 4a–i), suggesting the absence of sustained inflammation. Consistently, flow cytometry data indicated transient activation of certain immune cells but not sustained inflammation (Supplementary Fig. 5a–f). Notably, the inflammation-related changes in the osimertinib-treated heart were very modest compared with those observed in our recent report on the CML TKI ponatinib.²⁹ Together, these data suggest that inflammation is not a primary driver of osimertinib cardiotoxicity, leading us to examine type I, cell-autonomous cardiomyocyte toxicity using isolated cardiomyocyte models. The 1st cell-culture model we utilized is NRVM, a high-quality preclinical research model. In these experiments, the known cardiotoxic TKI ponatinib was used as a positive control,³⁰ whereas imatinib was used as a negative control. Along with these controls, various concentrations of osimertinib were tested to establish the dose-dependent effect of osimertinib on NRVM’s viability.²⁹ After 48 h of treatment, NRVMs were collected, and the CellTiter-Glo (Promega, #G7570) assay was employed to determine viability. NRVM’s viability was significantly decreased in response to various concentrations of osimertinib (Fig. 4h). We performed a time kinetics assay to examine the osimertinib cardiotoxicity over time. A decrease in the NRVM’s viability was observed as early as 24 h post-osimertinib treatment, which was more significant at later time points (Fig. 4i). Based on our dose-response and time kinetics results and previous reports, we decided to use 2.5 µM concentration for further experiments. As an alternative approach, flow cytometry was performed after 48 h of treatment with osimertinib (2.5 µM), and an increase in the percentage of pre-apoptotic cells (Annexin V+7AAD+) was observed (Fig. 4j, k). Next, we employed the HiPSC-CMs culture model to test the human relevance of findings from NRVMs. HiPSC-CMs were treated with vehicle and osimertinib (100 nM or 500 nM) for 48 h, followed by flow cytometry. Consistent with the findings of the NRVM model, a significant increase in the percentage of apoptotic cells was observed in response to osimertinib treatment (500 nM) (Fig. 4l, m). These data indicate that osimertinib has a direct cytotoxic effect on cardiomyocytes (type 1 toxicity), leading to cell death. These findings align with the recognized paradigm in TKI cardio-oncology that isolated cardiomyocytes are highly sensitive to kinase inhibitor-induced cytotoxicity. While in vitro models define direct, cell-autonomous toxicity, in vivo models capture complex multicellular and systemic contributors to cardiotoxicity. Together, these complementary approaches provide a comprehensive framework for defining TKI-associated cardiac injury.

Since our RNA-seq analysis revealed enrichment of stress and apoptotic signaling pathways, we next examined whether osimertinib disrupts key prosurvival signaling cascades essential for cardiomyocyte homeostasis. The prosurvival AKT and ERK pathways are key growth-signaling pathways critical to cardiomyocyte survival and cardiac homeostasis.³¹ Furthermore, osimertinib primarily targets the prosurvival EGFR pathway, which is upstream of the ERK and AKT cascades.¹⁵ In osimertinib-treated NRVMs, we observed a significant decrease in AKT and ERK phosphorylation (Fig. 5a, b). In addition, osimertinib treatment significantly increased STAT3 phosphorylation (Fig. 5c, d), while phosphorylation of p38 was reduced (Fig. 5c, e); no significant changes were observed in JNK activation (Fig. 5f). Phosphorylation of AKT, ERK, and p38 significantly decreased in the osimertinib-treated hearts (Fig. 5g–I). These data suggest that osimertinib mediated its cytotoxic effect primarily by targeting prosurvival ERK/AKT and p38 signaling pathways.

Fig. 5.

Osimertinib treatment modulates the phosphorylation of survival and stress pathway activity. Western blot analysis was performed to detect survival and stress pathways activity in osimertinib-treated NRVMs. Cells were treated with vehicle (DMSO) or osimertinib (2.5 µM) for the time indicated, followed by lysate collection and western blot. Representative blots are shown alongside the quantification of band intensities. a Representative western blot showing protein levels and corresponding P-AKT/AKT ratio densitometric analysis of blots. b Representative western blot showing protein levels and corresponding P-ERK/ERK ratio densitometric analysis of blots. c Representative western blot showing protein levels of P-STAT3, total-STAT3, P-p38, and total-p38 with β-Tubulin as loading control. d P-STAT3/total-STAT3 ratio quantification post-densitometric analysis of blots. e P-p38/total-p38 ratio expression quantification. f Representative western blot showing protein levels and corresponding P-JNK/total-JNK ratio densitometric analysis of blots. Data were analyzed using Two-way ANOVA followed by Tukey’s post hoc analysis and represented as mean ± SD. g–i LV lysates from TAC placebo and TAC osimertinib-treated mice were analyzed by Western blot to assess the expression of prosurvival signaling pathways. Representative blots are shown alongside the quantification of band intensities. g P-AKT/total-AKT ratio densitometric analysis of blots. h P-ERK/total-ERK ratio densitometric analysis of blots. i P-p38/total-p38 ratio densitometric analysis of blots. n = 5–8 mice per group. Protein lysates were prepared from the hearts of independent mice; each data point represents one biological replicate (individual lysate). Each data point represents an individual mouse. Data was analyzed using an unpaired t-test and represented as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

Osimertinib activates FOXO1 signaling and enhances p53-mediated stress responses

Consistent with suppression of AKT-mediated prosurvival signaling, we next examined the downstream target FOXO1. FOXO transcription factors are key downstream effectors of the PI3K/AKT pathway, regulating genes involved in apoptosis, oxidative stress responses, and metabolic homeostasis.³² Osimertinib treatment significantly reduced phosphorylation of FOXO1 at Ser256 (Fig. 6a–c), indicating relief from AKT-dependent inhibitory signaling, and was accompanied by increased total FOXO1 abundance in the nuclear fraction (Fig. 6a). These findings were further supported by enhanced nuclear localization of FOXO1 in osimertinib-treated NRVMs, as assessed by immunofluorescence (Fig. 6e).

Fig. 6.

Osimertinib modulates FOXO1/FOXO3a nuclear activity and p53 stress-response signaling. a–d Western blot analysis was performed to assess key stress-responsive and pro-apoptotic regulators. a NRVMs were treated with vehicle (DMSO) and 2.5 µM of osimertinib for 2 hours, followed by subcellular fractionation. Cytoplasmic and nuclear fractions were isolated and analyzed by Western blot. b Representative Western blots showing protein levels of P-FOXO1 (S256), total FOXO1, P-p53(S15), and total-p53 with β-Tubulin as loading control. c, d Densitometric analysis of blots in (b). c P-FOXO1/total FOXO1 ratio expression quantification. d P-p53/total-p53 ratio expression quantification. e NRVMs were treated with vehicle (DMSO) and 2.5 µM of osimertinib for 2 hours and then stained for FOXO1. e Representative immunofluorescence image of FOXO1-stained NRVMs. Red arrows indicate the nuclear localization of FOXO1. Scale bar = 20 µm. Each data point represents one biological replicate (individual lysate). Data was analyzed using an unpaired t-test and represented as mean ± SD. **p < 0.01, ***p < 0.001

In parallel, phosphorylation of class IIa HDACs (HDAC4/5/7) was reduced in the cytoplasmic fraction following osimertinib treatment compared with DMSO controls (Fig. 6a), consistent with altered chromatin regulatory signaling and transcriptional repression. Moreover, in agreement with RNA-seq analysis (Fig. 3d) and increased expression of the pro-apoptotic marker BAX (Fig. 4d–e), osimertinib treatment markedly increased phosphorylation of p53 at Ser15 (Fig. 6b, d), a canonical indicator of stress-induced p53 activation.³³ Collectively, these data indicate that osimertinib suppresses AKT-dependent prosurvival signaling while activating stress-responsive transcriptional programs, culminating in p53 activation and BAX-associated apoptotic signaling.

Osimertinib impairs mitochondrial homeostasis

Guided by RNA-seq analysis indicating enrichment of mitochondrial stress and apoptotic pathways, and prior evidence that TKIs induce mitochondrial injury, we investigated whether osimertinib directly impairs mitochondrial integrity in cardiomyocytes.³⁴ Our data indicate that osimertinib treatment increases apoptosis, along with elevated BAX expression and caspase cleavage within the intrinsic apoptosis pathway. Given these findings, we next examined the potential effect of osimertinib on mitochondrial integrity and function. NRVMs and HiPSC-CMs were treated with osimertinib for 48 h, followed by flow cytometry. Analysis revealed a significant decrease in mitochondrial mass (Fig. 7a, f, Supplementary Fig. 6), membrane potential (Fig. 7b, g, Supplementary Fig. 6), reactive oxygen species (ROS) generation (Fig. 7c, d, h, i, Supplementary Fig. 6), and nitric oxide level (Fig. 7e, j, Supplementary Fig. 6) in response to osimertinib. Together, these data indicate that osimertinib disrupts mitochondrial homeostasis and compromises outer mitochondrial membrane integrity, consistent with activation of intrinsic apoptotic signaling.

Fig. 7.

Osimertinib-induced mitochondrial dysfunction in CM. a–e NRVMs were treated with vehicle (DMSO) or osimertinib (2.5 µM) for 48 h, followed by flow cytometry analysis to assess mitochondria content and function. a Mean fluorescent intensity (MFI) of Mitotracker. b MFI of TMRE. c MFI of CellROX. d MFI of MitoSOX. e MFI of DAF-FM. Unpaired t-test calculated significance. f–j HiPSC-CMs were treated with vehicle (DMSO) or osimertinib (100 nM or 500 nM) for 48 h, cells were collected and digested to get a single-cell suspension, and flow cytometry analysis was carried out. f MFI of Mitotracker. g MFI of TMRE. h MFI of CellROX. i MFI of MitoSOX. j MFI of DAF-FM. Each data point represents NRVMs independently isolated from different pups and cultured separately. HiPSC-CMs were cultured in independent plates, and each data point represents a biological replicate. Data were analyzed using One-way ANOVA followed by Tukey’s post hoc analysis and represented as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

FDA-approved HDAC inhibitor Vorinostat (SAHA) rescued the osimertinib-induced cardiotoxicity and strengthened the cancer efficacy

Having defined the molecular mechanisms underlying osimertinib-induced cardiotoxicity, we next explored therapeutic strategies to mitigate cardiac injury without compromising anticancer efficacy. Transcriptomic and mechanistic analyses identified maladaptive stress signaling, mitochondrial dysfunction, and HDAC-dependent epigenetic repression as central drivers of cardiomyocyte death. Because direct activation of prosurvival pathways such as ERK/AKT is incompatible with cancer therapy, we focused on approaches with dual oncologic and cardioprotective potential.³⁵ An FDA-approved HDAC inhibitor, Vorinostat (SAHA), is proposed to treat NSCLC and has also been shown to be cardioprotective by improving CM’s survival and preserving mitochondrial function in multiple settings.^36,37 Indeed, preclinical studies have demonstrated the enhancement of the antitumor effects of osimertinib by combining HDAC inhibitor treatments.³⁸ However, to our knowledge, this combination has not been assessed in the context of its cardioprotective effects to mitigate osimertinib cardiac adverse events in cancer patients. Considering its dual potential to alleviate cancer growth and cardiac injury, alongside shared underlying mechanisms, we investigated whether SAHA could mitigate osimertinib-induced cardiotoxicity. As shown in Fig. 8a, we treated TAC mice with vorinostat (25 mg/kg, oral gavage) two times a week, in parallel with osimertinib (25 mg/kg/day, oral gavage) treatment for four weeks. There was no significant difference in body weight among the different groups (Fig. 8b). Notably, intervention with vorinostat largely rescued the detrimental cardiac effect of osimertinib (Fig. 8c–e). Consistently, vorinostat treatment reduced the osimertinib-induced elevation in markers of cardiac damage (ANP and BNP) (Fig. 8f, g) and fibrotic remodeling (COL1A1 and COL3A1) (Fig. 8h, i). Consistently, the excessive fibrotic remodeling of osimertinib-treated hearts was significantly reduced by vorinostat treatment (Fig. 8j–k). Additionally, the impaired hypertrophic growth in the osimertinib-treated TAC group was prevented by vorinostat (Fig. 8l). Moreover, osimertinib-induced CM’s death effect was abolished by vorinostat treatment, both in vivo and in vitro (Fig. 8m, n). Consistently, the osimertinib’s direct cytotoxic effect on isolated cardiomyocytes was abolished by vorinostat (Fig. 8o).

Fig. 8.

SAHA protects against osimertinib-induced cardiotoxicity. a Experimental design schematic created with BioRender.com. C57BL/6J mice were subjected to TAC surgery, and after one week, they were treated with a placebo, osimertinib (25 mg/kg/day), SAHA (25 mg/kg/biweekly), or a combination (osimertinib + SAHA) for four weeks. Echocardiography was performed at baseline, two- and four weeks post-treatment. n = 5–10 mice per group. b Weight. c EF. d FS. e LVIDs. f–i Four weeks after treatment, heart tissue was harvested for histological analysis, and RNA was extracted from the LV for gene expression analysis via qPCR. f ANP gene expression. g BNP gene expression. h COL1A1 gene expression. i COL3A1 gene expression. Data was analyzed using Two-way ANOVA followed by Tukey’s post hoc analysis and represented as mean ± SD. j, k Assessment of cardiac fibrosis by Masson’s Trichrome staining; j Representative Trichrome-stained LV regions and k Quantification of LV fibrosis. Scale bar = 50 µm. n = 5–7 mice per group. Assessment of cardiac hypertrophy: l HW/TL. n = 7–8 mice per group. m–n Assessment of apoptotic cell death by TUNEL staining was performed at four weeks after TAC surgery and osimertinib treatment. m Representative full-field images (20×) of TUNEL-stained heart sections with clearly indicated regions of interest. Scale bar = 100 µm. n Magnified views of the selected areas and quantification of TUNEL-positive nuclei in heart sections. n = 5–8 hearts per group with 10 random fields analyzed per heart. Scale bar = 100 µm. o NRVMs were treated with vehicle (DMSO), osimertinib (2.5 µM), SAHA (0.06 µM), or combination (osimertinib + SAHA) for 48 h, and cell viability was measured using the CellTiter-Glo luminescence assay. o Percentage of cell viability. p–s Western blot analysis was performed to evaluate FOXO1 and p53 phosphorylation status and histone H3K9 acetylation in NRVMs that were treated with vehicle (DMSO), osimertinib (2.5 µM), SAHA (0.06 µM), or combination (osimertinib + SAHA) for 2 h. p Representative western blots showing protein levels of the indicated proteins. q–s Densitometric analysis of blots. q P-FOXO1/FOXO1 ratio expression quantification. r P-p53/p53 ratio expression quantification. s H3K9-Ac/H3 ratio expression quantification. t Human EGFR-mutant NSCLC PC9 cells harboring an EGFR exon 19 deletion (delE746-A750) were treated with osimertinib, vorinostat (SAHA), or the combination for 48 h, followed by cell viability assessment. t Percentage of cell viability. Each data point represents an individual mouse. Data were analyzed using One-way ANOVA followed by Tukey’s post hoc analysis and represented as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

To determine whether HDAC activation is functionally central to osimertinib-induced cardiotoxicity, we pharmacologically inhibited HDAC activity using the pan-HDAC inhibitor SAHA. SAHA treatment, both alone and in combination with osimertinib, significantly reversed the osimertinib-induced reduction in FOXO1 phosphorylation at Ser256, suppressed the upregulation of p53 phosphorylation at Ser15, and restored H3K9 acetylation levels (Fig. 8p–s), demonstrating that HDAC-dependent chromatin remodeling is a key mediator of the adverse cardiac phenotype.

In addition, we employed the NSCLC cell line PC9 derived from human lung adenocarcinoma to test our hypothesis that vorinostat provides dual therapeutic benefits. PC9 cells harbor an EGFR exon 19 deletion mutation (delE746-A750), which makes them highly sensitive to EGFR TKIs, such as osimertinib. PC9 cells are commonly used in preclinical studies to investigate mechanisms of EGFR signaling and targeted therapies in EGFR-mutant NSCLC. PC9 cells were treated with osimertinib, vorinostat, and a combination for 48 h. As expected, osimertinib induced a potent cytotoxic (anticancer) effect on PC9 cells, as reflected by a significantly reduced viability (Fig. 8t). To our complete surprise, vorinostat also induced a significant cytotoxic effect but to a lesser extent than osimertinib. Most importantly, combining osimertinib and vorinostat showed a synergistic effect, as reflected by significantly reduced PC9 cell viability (Fig. 8t).

To assess the long-term effects of HDAC inhibition, TAC mice were treated with osimertinib ± vorinostat for eight weeks (Supplementary Fig. 7a). Body weight was unchanged across groups (Supplementary Fig. 7b). Vorinostat largely prevented osimertinib-induced cardiac dysfunction (Supplementary Fig. 7c–f), reducing elevations in heart failure and fibrotic markers (Supplementary Fig. 7g–j), and preserving hypertrophic growth (Supplementary Fig. 7k). Although aging is an important factor influencing cardiac remodeling and therapeutic responses, the present study was conducted in young adult mice; therefore, future studies incorporating aged models will be important to further define the translational relevance of these findings.

To further assess the systemic safety of SAHA, serum clinical chemistry parameters were evaluated in TAC-treated mice (Supplementary Fig. 8). Markers of liver injury revealed increased ALP levels in osimertinib-treated mice, whereas SAHA alone or in combination normalized ALP to placebo levels; AST levels were unchanged across groups, and ALT elevations observed in TAC controls were reduced with SAHA treatment (Supplementary Fig. 8a–c). Assessment of hepatic synthetic function showed modest changes in albumin and total protein, with normalization of total protein levels in SAHA-treated and combination groups (Supplementary Fig. 8d–e). Renal function, assessed by BUN, was comparable among all groups (Supplementary Fig. 8f). Metabolic parameters demonstrated elevated cholesterol in osimertinib-treated mice that was reduced with SAHA treatment, while glucose levels remained unchanged (Supplementary Fig. 8g, h). Collectively, these data indicate that SAHA does not exacerbate systemic toxicity and may alleviate select hepatic and metabolic alterations associated with osimertinib.

In summary, herein, we have established the 1st in vivo mouse model successfully recapitulating the observations in human patients, identified key molecular drivers of osimertinib’s cardiotoxicity, and demonstrated that aberrant HDAC activation underlies maladaptive cardiac remodeling. Importantly, these findings establish epigenetic regulation as a therapeutically actionable pathway and highlight HDAC inhibition as a promising strategy to preserve cardiac function while maintaining the therapeutic efficacy of osimertinib.

Discussion

To our knowledge, this is the first study to establish a preclinical mouse model that effectively recapitulates the osimertinib-induced cardiotoxicity observed in cancer patients. Our results demonstrate that osimertinib impairs key survival signaling pathways, disrupts mitochondrial homeostasis, induces cardiomyocyte apoptosis, and dysregulates histone acetylation through altered HDAC activity, ultimately leading to cardiac dysfunction and adverse remodeling. Importantly, we identified HDAC inhibition as a promising cardioprotective strategy, as vorinostat effectively mitigated osimertinib-induced cardiotoxic effects. Furthermore, findings with human NSCLC-derived PC9 cells indicate that SAHA provides dual therapeutic benefits by enhancing osimertinib’s cancer efficacy and alleviating cardiotoxicity.

Osimertinib is a third-generation EGFR-TKI with well-documented clinical efficacy; however, its cardiotoxic effects remain poorly understood due to the lack of a suitable preclinical model.³⁹ Recent clinical observations and meta-analyses have highlighted that patients with cardiovascular comorbidities are especially vulnerable to TKI therapies associated with cardiotoxicity.^40,41 To address this, we employed a transverse aortic constriction (TAC) model of pressure overload to mimic the cardiovascular comorbidities in cancer patients. Indeed, a clinically relevant dose of 25 mg/kg/day, which approximates the human-equivalent FDA-approved dose of 80 mg/day, was sufficient for the robust cardiac phenotype. These findings are consistent with previous reports demonstrating cardiac stress enhances vulnerability to drug-induced cardiotoxicity.^40,41 This well-optimized clinically relevant in vivo model could be further utilized for mechanistic insight and testing therapeutic interventions of osimertinib cardiotoxicity.

One of the key findings in our investigation is that osimertinib-treated hearts failed to develop a typical adaptive hypertrophic response to pressure overload. The significantly reduced cardiomyocyte cross-sectional area confirmed that impaired cellular adaptation primarily drove the stunted organ-level hypertrophy. Indeed, we establish that osimertinib elicits type 1 cardiotoxicity, characterized by impaired growth signaling and excessive cardiomyocyte death. Cardiomyocyte survival relies on the integrity of prosurvival signaling pathways that promote adaptation to stress and prevent cell death.⁴² Disruption of these prosurvival mechanisms can lead to cardiomyocyte apoptosis and contribute to the progression of cardiac dysfunction.³⁰ Osimertinib suppressed the key cardiac prosurvival AKT/ERK1/2 cascades, while relieving inhibitory control of FOXO1 and activating stress-responsive p53–BAX signaling, thereby compromising cardiomyocyte viability and mitochondrial homeostasis. These observations align with previous reports that therapeutic TKIs can impair survival pathways and promote mitochondrial injury, supporting the notion that these effects may be a shared mechanism of TKI-induced cardiotoxicity.^30,43 However, this is starkly distinct from our recent report with CML TKI ponatinib, which primarily induces cardiotoxicity by excessive activation of immune cells and inflammation.²⁹ Of note, osimertinib did not exhibit sustained activation of inflammatory pathways, suggesting that non-inflammatory mechanisms are primarily responsible for cardiotoxicity. These mechanistic differences highlight the importance of not generalizing the effects of all TKIs and emphasize the need to understand the specific actions of each one to guide cardioprotective strategies and therapeutic development.

Mitochondria are critical for maintaining cardiomyocyte function and survival, and their dysregulation is a hallmark of various types of cardiac injury.^44,45 RNA-sequencing revealed activation of mitochondrial stress and oxidative pathways, and this was supported by functional evidence of reduction in mitochondrial membrane potential, mitochondrial mass, and intracellular NO levels, along with elevated ROS production, collectively highlighting loss of mitochondrial integrity as a central pathological feature in osimertinib-mediated cardiotoxicity. These results suggest that osimertinib disrupts mitochondrial homeostasis, thereby compromising cardiomyocyte viability and contributing to cardiac dysfunction. These findings are consistent with a previous study with a HiPSC-CMs-based cell-culture model of osimertinib cardiotoxicity.⁴⁶ Previous studies have implicated mitochondrial damage as a mechanism of cardiotoxicity of other TKIs, supporting the notion that this may be a class-wide effect of TKIs.³⁴ Together, these findings suggest that mitochondrial dysregulation may play a central role in osimertinib-associated cardiac pathologies; therefore, therapeutic strategies to preserve mitochondrial integrity may be a viable target to combat osimertinib cardiotoxicity.

We identified changes in HDAC expression and histone acetylation as key features of osimertinib-treated hearts, highlighting epigenetic remodeling as a potential driver of cardiotoxicity. Pharmacological inhibition of specific HDACs has been shown to restore histone acetylation, reduce cardiomyocyte death, and improve cardiac function in models of cardiac stress.⁴⁷ Consistent with these established roles of HDAC isoforms in regulating transcriptional programs governing cardiac development, hypertrophy, metabolism, and remodeling, our RNA-seq data identify HDAC dysregulation and altered histone acetylation as key features of osimertinib-treated hearts, supporting epigenetic targeting as a potential strategy to mitigate cardiotoxicity. Thus, we prioritized identifying rescue strategies that preserve both cardiovascular function and anticancer efficacy. The choice of SAHA for rescue studies was strategically guided by its known cardioprotective ability and favorable impact on mitochondrial biogenesis. Moreover, previous oncology-focused studies had suggested a synergistic anticancer effect when HDAC inhibitors were combined with EGFR TKIs, a concept we confirmed in our NSCLC PC9 cell experiments. The HDAC inhibitor vorinostat (SAHA) effectively mitigated the structural and functional cardiac deterioration caused by osimertinib. Beyond blunting cardiac apoptosis and fibrosis, SAHA restored mitochondrial health, improved cellular survival signaling, and potentially enhanced osimertinib’s antitumor efficacy in EGFR-mutant NSCLC cells. This dual benefit strongly supports its candidacy for combinatorial regimens in NSCLC patients at risk of cardiotoxicity. However, it’s essential to recognize that future preclinical studies employing the NSCLC cancer-bearing mouse model are warranted to test the true translational potential of this pathway to manage osimertinib-mediated cardiotoxicity.

In summary, we report the first in vivo preclinical mouse model of osimertinib-induced cardiotoxicity and identify mitochondrial disruption and impaired survival signaling leading to excessive cardiomyocyte death as core mechanisms. From a translational perspective, we demonstrate that HDAC inhibition offers a promising strategy to preserve cardiac function while sustaining antitumor efficacy. As osimertinib becomes increasingly entrenched as first-line therapy for EGFR-mutated NSCLC patients, our work advocates a paradigm shift that integrates cardiovascular risk assessment into oncologic treatment planning.

Materials and methods

Chemicals

Tyrosine Kinase Inhibitors Osimertinib, Ponatinib, and Imatinib were purchased from LC Laboratories and suspended in DMSO to prepare 10 mM stock for cell-culture experiments. For mice treatment, osimertinib was dissolved in 5% DMSO + 40% PEG300 + 5% Tween-80 + 50% ddH₂O. Vorinostat was purchased from MedChemExpress and suspended in DMSO to prepare a 10 mM stock for cell-culture experiments. For mice treatment, vorinostat was dissolved in 5% DMSO + 40% PEG300 + 5% Tween-80 + 50% Saline.

Animals

Wild-type (WT) C57BL/6J mice were purchased from Jackson Laboratories (Bar Harbor, ME). Mice of both genders at 8 weeks of age were utilized for both in vitro and in vivo experiments. For in vivo dosing, osimertinib (25 mg/kg) was administered daily by oral gavage for 4 weeks. A fresh stock solution was prepared every 7 days and stored at 4 °C. This dose of osimertinib approximates the clinically approved 80-mg dose for patients.²⁶ All mice were maintained in a pathogen-free environment at the University of Alabama (UAB) animal facilities and at the Louisiana State University Health Shreveport (LSUHS). All animals were included in the analysis. All procedures were performed after approval from the Institutional Animal Care and Use Committees (IACUC), in agreement with the NIH Guide for the Care and Use of Laboratory Animals.

Transverse aortic constriction (TAC) surgery in mice

TAC surgery was performed as previously described.^48,49 Briefly, mice were sedated with isoflurane (induction, 2%; maintenance, 1.5%) and anesthetized to a surgical plane with intraperitoneal ketamine (100 mg/kg) and xylazine (10 mg/kg). Anesthetized mice were intubated, and a midline cervical incision was made to expose the trachea and carotid arteries. A blunt 20-gauge needle was inserted into the trachea and connected to a volume-cycled rodent ventilator on supplemental oxygen at a rate of 1 l/min, with a respiratory rate of 140 breaths/min. Aortic constriction was performed by tying a 7-0 nylon suture ligature around a 27-gauge needle. The needle was then promptly removed to yield a constriction of approximately 0.4 mm in diameter. To confirm the efficiency of TAC surgery, the pressure gradient across the aortic constriction was measured by Doppler echocardiography. A sham operation was similarly conducted without suture ligation.

Cell isolation and culture

Isolation of neonatal rat ventricular cardiomyocytes (NRVMs)

Neonatal rat ventricular cardiomyocytes (NRVMs) were isolated following established procedures from 1–3-day-old Sprague-Dawley rats.^50,51 The isolated cells were plated onto Primaria cell-culture dishes (BD) and cultivated in Ham’s F-10 medium supplemented with 5% horse serum, 5% fetal bovine serum, and 1% antibiotics. After applying the intended treatments, the cells were lysed for subsequent experiments.

Human induced pluripotent stem cell-cardiomyocytes (HiPSC-CMs)

Human iPS cells (ethics approval number B26-001) were cultured within 6-well plates coated for 1 h at room temperature with hESC-qualified Matrigel supplied from Corning (354277) at the manufacturer-recommended concentration in order to maintain pluripotency. The cells are cultured using MTeSR1 (Stemcell Technologies 85850) and receive daily media changes. At 75–85% confluency, cells are disassociated using Accutase (Stemcell Technologies 07920) and reseeded in MTeSR1 supplemented with 10 μM Y-27632, ROCK1 inhibitor (Selleckchem S1049) at a split ratio of 1:6. After 12 h, hiPSCs cells are switched to MTeSR1 without ROCK1 inhibitor. After 3 days, hiPSCs are at a confluence of 75–85% and are switched to RPMI medium (Gibco 11875093) supplemented with B27 without insulin (Gibco A18956-01) and 10 μM WNT promoter CHIR99021 (Stemcell Technologies 72052); this is considered Day 1 of differentiation. After exactly 24 h, cells are washed and switched to media without CHIR99021. Cells are maintained with daily media changes until Day 4, when the media is then supplemented with 5 μM WNT inhibitor IWP-2 (Stemcell Technologies 72124). Cells are cultured with IWP-2 until Day 6. On Day 8, cells are switched to RPMI medium supplemented with B27 with insulin (Gibco 17504044). At this point, the cells are referred to as hiPSC-derived cardiomyocytes and have begun spontaneously beating. Medium is changed daily until Day 10, when it is refreshed thereafter every two days. Only plates that contain greater than 60% beating hiPSC-CMs are used for further study.

PC9 (human adenocarcinoma, differentiated)

PC9 (Human adenocarcinoma, differentiated) (ABC-TC0907) was purchased from AcceGen Biotech (Fairfield, NJ). Cells were cultivated in RPMI 1640 medium supplemented with 2 mM Glutamine and 10% fetal bovine serum. After applying the intended treatments, the cells were lysed for subsequent experiments.

Subcellular fraction

Subcellular compartment extracts were prepared using the Thermo Scientific Subcellular Protein Fractionation Kit for Cultured Cells (Cat. No. 78840) according to the manufacturer’s instructions. Briefly, adherent cells were harvested using trypsin-EDTA and centrifuged at 500 × g for 5 min. Suspension cells were collected by centrifugation under the same conditions. Pellets were washed once with ice-cold PBS and transferred to 1.5 mL microcentrifuge tubes. Approximately 1–10 × 10⁶ cells were used per fractionation. Cells were pelleted again at 500 × g for 2–3 min, and the supernatant was carefully removed. Cell pellets were resuspended in ice-cold Cytoplasmic Extraction Buffer (CEB) supplemented with protease inhibitors. For all subsequent steps, buffer volumes were scaled proportionally to pellet size using the manufacturer’s recommended ratios. Samples were incubated with CEB for 10 min at 4 °C with gentle mixing and centrifuged at 500 × g for 5 min. The resulting supernatant (cytoplasmic extract) was transferred to a pre-chilled tube. The remaining pellet was resuspended in Membrane Extraction Buffer (MEB) containing protease inhibitors, vortexed for 5 s at maximum speed, and incubated for 10 min at 4 °C with gentle mixing. Samples were centrifuged at 3000 × g for 5 min, and the supernatant (membrane extract) was collected. Pellets were then resuspended in Nuclear Extraction Buffer (NEB) containing protease inhibitors, vortexed for 15 s at maximum speed, and incubated for 30 min at 4 °C with gentle mixing. Samples were centrifuged at 5000 × g for 5 min, and the supernatant containing the soluble nuclear extract was transferred to a clean pre-chilled tube. All fractions were kept on ice throughout the procedure and stored at −80 °C until use.

Flow cytometry

Heart tissue, NRVMs, and HiPSC-CMs were collected in media to create a single-cell suspension. Cell counts were performed, and 0.5–1 × 10⁶ cells were used for surface staining after Fc receptor blocking (1 μg/ml) in 3% FBS for 30 min on ice. For intracellular staining, 0.5–1 × 10⁶ cells were cultured in 96-well plates (Nunc, USA) and incubated for 5 h in DMEM media containing Phorbol 12-myristate 13-acetate, or PMA (0.1 mg/ml, Sigma), ionomycin (1 mg/ml, Sigma), and a 1:1000 dilution of Golgistop/Golgiplug (BD Biosciences). Following stimulation, cells were first surface-stained for 30 min on ice. After washing with PBS, cells were fixed and permeabilized using the BD Cytofix/Cytoperm kit (BD Biosciences) for 30 min at 4 °C, followed by washing with the BD Perm Wash kit (BD Biosciences) and then stained with intracellular fluorescently labeled antibodies. The fluorescence intensity of fluorochrome-labeled cells was measured using flow cytometry (BD LSR-II), and sorting, as needed, was performed with the BD FACS Aria III cytometer at the UAB Flow Cytometry Core Facility. FACS Diva was utilized for cell acquisition, and the final data analysis was conducted using FlowJo (Tree Star, USA). Details of antibodies used in this study are provided in Supplementary Table 1.

Transthoracic echocardiography

Echocardiography was performed using the Vevo 3100 high-resolution ultrasound imaging system equipped with an MX400 transducer (Visual Sonics Inc., Toronto, Canada). Mice were anesthetized with 1.5–2% isoflurane and positioned on a temperature-controlled platform in a supine orientation. After removing the hair, a layer of warmed ultrasound transmission gel was applied to the chest area. Transthoracic M-mode echocardiography was performed in the parasternal short-axis view, with a 12-MHz probe (VisualSonics). LV end-systolic interior dimension (LVID;s), LV end-diastolic interior dimension (LVID;d), ejection fraction (EF), and fractional shortening (FS) values were obtained by analyzing data using the Vevo 3100 program.

Tissue processing and histological analysis

Animals were euthanized under anesthesia. Heart tissues were fixed in 10% neutral buffered formalin, embedded in paraffin, and sectioned at 5 μm thickness. Sections were deparaffinized with xylene and rehydrated by incubating with decreasing concentrations of ethanol. For the detection of fibrosis, sections were stained with Masson Trichrome (Sigma-Aldrich, #HT15) as per the manufacturer’s instructions. The images of the LV region were captured using a Nikon Eclipse E200 microscope with NIS Element software version 5.20.02. The quantification of LV fibrosis was determined with ImageJ version 1.52a software (NIH). For fibrosis measurement, 8–10 images of the LV region were taken, and LV fibrosis was quantified as a percentage of the total LV area scanned. At least 100,150 cardiomyocytes per heart were taken for CSA measurement. The representative images were chosen as per the mean/average.

Immunofluorescence staining

Cells were cultured on 3% gelatin coated-glass coverslips to approximately 60–80% confluency prior to immunostaining. Cells were fixed in 4% paraformaldehyde for 10–20 min at –20 °C, followed by three washes with PBS to remove residual fixative. Cells were permeabilized with 0.1% Triton X-100 in PBS for 10–15 min at room temperature and then washed three times with PBS. To reduce nonspecific antibody binding, samples were incubated with a blocking solution consisting of 5% bovine serum albumin (BSA) in PBS containing 0.1% Triton X-100 for 1 h at room temperature. Primary antibodies were diluted 1:100 in blocking buffer and applied to the cells for 1–2 h at room temperature or overnight at 4 °C, according to the manufacturer’s recommendations. After incubation, cells were washed three times with PBS. Fluorophore-conjugated secondary antibodies were diluted 1:800 in blocking buffer and incubated with the samples for 30–60 min at room temperature in the dark, followed by three PBS washes. Nuclei were counterstained with mounted DAPI following the manufacturer’s instructions. Images were acquired using an Olympus Spinning Disk Confocal microscope with CellSens Dimension software at the Microscopy Imaging Core at the Research Core Facility at LSU Health Shreveport.

Western blot

Left ventricular (LV) tissues and neonatal rat ventricular myocytes (NRVMs) were harvested and lysed as described below. Aliquots of the same protein lysates were processed using either infrared-based or chemiluminescent detection protocols, depending on the target proteins analyzed.

LV tissues were homogenized, and NRVMs were lysed in either cell lysis buffer (Cell Signaling Technology, #9803) or RIPA buffer (Thermo Fisher Scientific, #89900), each freshly supplemented with protease inhibitors (Sigma-Aldrich, #P8340 or Thermo Fisher Scientific, #78440) and phosphatase inhibitors (Sigma-Aldrich, #P0044 or Thermo Fisher Scientific, #78440). Protein concentrations were determined using the Bio-Rad Protein Assay Dye (#5000006). Equal amounts of protein were denatured in SDS–PAGE sample buffer or 2× Laemmli buffer containing DTT (reducing conditions), resolved by SDS–PAGE (Bio-Rad, #4561096), and transferred onto PVDF membranes (Immobilon-P, Millipore, #IPVH00010; 0.45 µm pore size).

For infrared-based detection, membranes were blocked with Odyssey Blocking Buffer (LI-COR, #927-40000) for 1 h at room temperature and incubated overnight at 4 °C with primary antibodies (Supplementary Table 2), followed by IRDye-conjugated secondary antibodies (LI-COR Biosciences). Protein signals were detected using the Odyssey Infrared Imaging System, and band intensities were quantified using Image Studio software (version 5.2).

For chemiluminescent detection, membranes were blocked with 5% nonfat dry milk in TBST or 5% BSA in TBST (for phospho-specific antibodies) and incubated with the indicated primary antibodies (Supplementary Table 2), followed by HRP-conjugated secondary antibodies (Cell Signaling Technology, #7074S or #7076S). Protein bands were visualized using enhanced chemiluminescence reagents (MilliporeSigma, #WBLUF0100, #WBLUC0100, or #WBULS0100) on a LI-COR Odyssey XF Imaging System. Densitometric analyses were performed using ImageJ software (version 1.54 g).

RNA extraction and quantitative PCR analysis

Total RNA was extracted using the RNeasy Mini Kit (Qiagen, #74104) according to the manufacturer’s protocol. cDNA was synthesized using the iScript cDNA synthesis kit (Bio-Rad, #170-8891) following the manufacturer’s instructions. Gene expression was analyzed by quantitative PCR (qPCR) using the TaqMan Gene Expression Master Mix (Applied Biosystems, #4369016) and TaqMan gene expression assays (Applied Biosystems) on a Quant Studio 3 (Applied Biosystems) Real-Time PCR Detection machine. Details of TaqMan gene expression assays used in this study are provided in Supplementary Table 3. Relative gene expression was determined by using the comparative CT method (2-ΔΔCT) and was represented as fold change. Briefly, the first ΔCT is the difference in threshold cycle between the target and reference genes: ΔCT = CT (a target gene X) − CT (18S rRNA), while ΔΔCT is the difference in ΔCT as described in the above formula between the CTRL and experimental group, which is =ΔCT (KO target gene X) − ΔCT (CTRL target gene X). Fold change is calculated using the 2-ΔΔCT equation.

RNA-sequencing and bioinformatic analysis

RNA-sequencing was performed by LC Sciences, LLC (Houston, TX). Total RNA samples underwent quality control assessment, followed by poly(A)-enriched, strand-specific library preparation according to the provider’s standard protocols. Libraries were sequenced on an Illumina platform, generating 150-bp paired-end reads.

Bioinformatic analysis, including quality assessment, reading alignment, quantification, and differential gene expression analysis, was performed by the Bioinformatics and Modeling Core at LSU Health Shreveport (Shreveport, LA). Briefly, raw reads were aligned to the mouse reference genome and transcriptome (GRCm38) using STAR and quantified with RSEM, following the nf-core/rnaseq pipeline. Differential expression analysis was performed using the limma + voom in R, with significance defined as adjusted p-value (FDR) < 0.05. Downstream analyses, including PCA and pathway enrichment (GSEA/g:Profiler), were conducted using standard R-based pipelines.

Clinical chemistry

Clinical chemistry profiling of mouse serum was outsourced to IDEXX BioAnalytics (North Grafton, MA), where analytes were measured using standardized automated clinical chemistry instrumentation.

Cell viability assay

NRVMs were seeded in a 96-well plate at 40,000 cells per well in F-10 Ham (Gibco) supplemented with 10% FBS, 2 mmol/L L-glutamine, and 1% penicillin–streptomycin, allowed to adhere overnight, followed by 24-h incubation in serum-free media. The next day, cells were treated with a dilution series of test compounds for 24, 48, and 72 h. Cell viability was determined using the CellTiter-Glo kit (Promega), and the results were represented as background-subtracted relative luminescence normalized to a dimethyl sulfoxide (DMSO)–treated control.

TUNEL assay

Animals were euthanized under anesthesia. Heart tissues were fixed in 10% neutral buffered formalin, embedded in paraffin, and sectioned at 5 μm thickness. Sections were deparaffinized with xylene and rehydrated by incubating with decreasing concentrations of ethanol. Staining of apoptotic cells by terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL; kit from Roche) was performed according to the manufacturer’s instructions. The images were captured using an Olympus BX43 microscope with CellSens standard software. The number of apoptotic cells was counted with ImageJ software.

Statistical analysis

We conducted our analyses using GraphPad Prism (version 9.3.1). Data were subjected to statistical analysis based on the following criteria: Statistical differences for multiple group comparisons were performed using 1-way analysis of variance (ANOVA) with the Tukey post hoc test. Differences between data groups with two variables were evaluated for significance using 2-way ANOVA followed by the Tukey post hoc test. We considered a significant level of less than 0.05 (P < 0.05) statistically significant.

Author contributions

A.T.C. and H.L. conceived and designed the research; A.T.C., A.S.B., A.S.T., A.J., B.S., P.U., D.Y.L., and R.S. performed experiments; A.T.C., A.S.T., A.J., B.S., D.Y.L., and R.S. analyzed data; A.T.C. and H.L. interpreted the results of the experiments; A.T.C. prepared figures; A.T.C. and H.L. drafted the manuscript; all the authors contributed to editing, revision, and approved the final version of the manuscript.

Data availability

The Bulk RNA-seq data described in this study can be found at the Gene Expression Omnibus, which is available online. Accession number: GSE328472. All other data supporting the findings of this study are available from the corresponding author upon reasonable request.

Competing interests

P.U. is currently employed by AstraZeneca, MD, USA. A.T.C., A.S.B., A.S.T., A.J., B.S., D.Y.L., R.S., Q.Z., S.K.V., S.T. and H.L. declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41392-026-02814-1.