Entrectinib binds to HMGB1 and activates cardiomyocyte autophagy by inhibiting OTUD5-MTORC1 signaling to induce cardiotoxicity

ABSTRACT

Entrectinib stands unparalleled as the sole neurotrophic tyrosine receptor kinase (NTRK) inhibitor that has demonstrated clinical efficacy in treating brain metastases across various cancer types. However, its potential to induce severe cardiotoxicity, compounded by the current lack of effective intervention strategies, poses a substantial risk of treatment failure, underscoring the critical need for in-depth research on the molecular mechanism. Here, we utilized proteomics analysis and a murine model with cardiomyocyte-specific atg7 deletion to reveal that entrectinib activated autophagy in cardiomyocytes, subsequently triggering apoptosis and leading to cardiac dysfunction. Mechanistically, entrectinib directly bound to the HMGB1 protein at the 103rd phenylalanine residue, enhancing its nuclear localization. In the nucleus, HMGB1 suppressed the transcription of the deubiquitinating enzyme OTUD5, a vital regulator of the MTORC1 pathway, which subsequently inhibited the MTORC1 pathway, culminating in the activation of macroautophagy/autophagy. Furthermore, our research demonstrated that HMGB1 inhibition could prevent the cardiotoxicity induced by entrectinib in both in vivo and in vitro models. Specifically, we found that tanshinone IIA could mitigate the cardiotoxic effects of entrectinib by reducing HMGB1 protein levels. Taken together, our findings elucidated the mechanism underlying entrectinib-induced cardiotoxicity, offering a theoretical foundation for the safer clinical application of this targeted therapy.

Abbreviations and Acronyms: ATG7: autophagy related 7; CHX: cycloheximide; CKMB: creatine kinase myocardial band; CQ: chloroquine; c-PARP: cleaved poly (ADP-ribose) polymerase family; DAPI: 4‘6-diamidino-2-phenylindole; EF: ejection fraction; FS: fractional shortening; GSEA: gene set enrichment analysis; H&E: hematoxylin and eosin; HW:TL: ratio of heart weight to tibia length; KEGG: Kyoto Encyclopedia of Genes and Genomes; MYH6: myosin heavy polypeptide 6, cardiac muscle, alpha; MYH7: myosin, heavy polypeptide 7, cardiac muscle, beta; NPPA: natriuretic peptide type A; NPPB: natriuretic peptide type B; PI: propidium iodide; qPCR: quantitative real-time PCR; SD: standard deviation; SRB: sulforhodamine B; WGA: wheat germ agglutinin.

Introduction

Adverse cardiovascular events are among the greatest challenges in the management of cancer patients and have promoted the emergence of cardio-oncology, an important cardiovascular subspecialty. In August 2019, the FDA granted approval for entrectinib, a landmark therapy for adults and pediatric patients battling advanced recurrent solid tumors characterized by NTRK gene fusions [1]. Entrectinib is a first-in-class, CNS-active, potent inhibitor of NTRK1/TRKA (neurotrophic receptor tyrosine kinase 1), NTRK2/TRKB and NTRK3/TRKC, ROS1, and ALK, approved for the treatment of NTRK fusion-positive solid tumors and ROS1-positive non-small cell lung cancer [2]. Its ability to cross the blood-brain barrier makes it particularly valuable for treating patients with brain metastases, a population with limited therapeutic options [3]. This drug has demonstrated an impressive objective response rate of 57.4% across a diverse array of 10 tumor types, showcasing its potent brain penetration and achieving a full remission rate of 100% in patients with intracranial lesions [4,5]. Despite these successes, 3.4% of patients may experience congestive heart failure as a side effect of entrectinib treatment, with 2.3% of patients being classified as having severe (grade III – IV) heart failure [6,7]. Cardiotoxicity is among the most important adverse effects of entrectinib; owing to its unclear mechanism and the lack of effective strategies, once cardiotoxicity occurs, entrectinib must be interrupted or even permanently discontinued for these patients [3], which is an issue that must be addressed in the field of cardio-oncology. Given the increasing use of entrectinib in clinical practice, there is an urgent clinical need to study the mechanism of cardiotoxicity induced by entrectinib.

Autophagy is a highly conserved cellular process that degrades and recycles damaged organelles and proteins to maintain cellular homeostasis [8]. Dysregulated autophagy has been implicated in numerous human diseases, including neurodegenerative disorders such as Alzheimer and Parkinson diseases, where impaired autophagy contributes to the accumulation of toxic protein aggregates [9]. In metabolic diseases like diabetes and obesity, autophagy modulates insulin sensitivity and lipid metabolism [10]. Furthermore, in the context of cancer, autophagy can exert both tumor-suppressive and tumor-promoting effects depending on the stage and type of malignancy [11]. Autophagy plays a vital role in cardiac health, shielding the heart from stress and preserving its function during ischemic episodes by clearing misfolded proteins and combating oxidative stress [12]. Dysfunctional autophagy is associated with cardiac issues associated with obesity, aging, and doxorubicin-induced cardiomyopathy [13–15]. However, it can paradoxically harm the heart in scenarios such as reperfusion injury and contribute to drug-induced cardiotoxicity by disrupting the protein balance within cardiomyocytes [16,17]. Understanding the dual nature of autophagy is essential for optimizing heart health and the effects of medications. Our group has conducted an exhaustive investigation into the intricate role of autophagy in various types of drug-induced cardiotoxicity [18,19], revealing that autophagy is a pervasive factor in the manifestation of these adverse effects and is likely involved in the cardiotoxicity induced by entrectinib. The precise nature of autophagy activation and its subsequent impact as either a protective or deleterious force in the context of entrectinib-induced cardiotoxicity remain enigmas that warrant further exploration.

The MTOR (mechanistic target of rapamycin kinase) complex 1 (MTORC1) is a pivotal negative regulator of autophagy. High levels of nutrients promote the inhibition of autophagy by phosphorylating ULK1 and ATG13 via MTORC1, thus curbing autophagy activation. However, in the face of nutrient depletion or cellular stress, MTORC1 releases its grip on ULK1, initiating autophagy [20]. Additionally, MTORC1 orchestrates autophagy by modulating the expression of lysosomal genes, with transcription factors such as TFEB and TFE3 playing pivotal roles in this regulatory network [21]. The activity of MTORC1 is governed by a plethora of stimuli, including growth factors, nutrient availability, energy levels, and stress signals, as well as fundamental signaling pathways such as phosphoinositide 3-kinase/PI3K, MAPK, and AMP-activated protein kinase (AMPK) [22]. Despite these known regulatory inputs, a comprehensive understanding of the modulatory mechanisms of the MTORC1 pathway remains elusive. In particular, the potential for nuclear factors to regulate MTORC1 activity within the nucleus is an area that has not yet been fully elucidated and presents an intriguing avenue for future research endeavors.

In this study, we revealed that entrectinib triggered apoptosis in cardiomyocytes through the activation of autophagy, culminating in the impairment of cardiac function. Mechanistically, entrectinib directly interacted with the HMGB1 protein, facilitating its nuclear translocation. Once within the nucleus, HMGB1 impeded MTORC1 signaling, thereby activating cardiomyocyte autophagy. This activation occurred through the downregulation of OTUD5, a key deubiquitinating enzyme. Building upon these insights, we ultimately confirmed the therapeutic efficacy of HMGB1 inhibitors in mitigating the cardiotoxic effects of entrectinib.

Results

Entrectinib causes systolic dysfunction and cardiomyocyte damage in mice

To establish an in vivo model that faithfully represents the cardiotoxic potential of entrectinib, a study in which C57BL/6J mice were orally administered entrectinib at a daily dosage of 200 mg/kg (double the human equivalent dose) for six weeks, after which echocardiographic assessments were performed to evaluate cardiac function (Figure 1(A)). The echocardiographic data revealed a notable decrease in both the left ventricular ejection fraction (LVEF) and fractional shortening (FS) in the mice that received entrectinib compared with those in the placebo (vehicle)-treated group (Figure 1(B) through 1E). This reduction was observed while a consistent heart rate was maintained (Figure 1E and Table S1), indicating a detrimental effect of entrectinib on cardiac systolic function. These findings were consistent with the clinical presentation of congestive heart failure associated with the use of entrectinib, validating our in vivo model as a reliable tool for studying the cardiotoxic effects of this drug.

Figure 1.

Entrectinib induces myocardial injury and left ventricular dysfunction. C57BL/6J mice were treated with vehicle or 200 mg/kg entrectinib for 6 weeks. n = 9. (A) The diagram of the experiment. (B–E) The cardiac function of mice was measured by echocardiography. (B) Representative M-mode echocardiographic images from each group in mice. (C) The statistics of ejection fraction. (D) The statistics of fractional shortening. (E) The statistics of heart rate. (F) Representative whole heart images. (G) Heart weight to tibia length ratio (HW:TL). (H) Representative images of cardiac sections stained by WGA. (I and J) Serum from the mice was analyzed for CK and CK-MB levels. (K) Representative images of cardiac sections stained by hematoxylin-eosin (H&E). (L–M) Representative images of cardiac longitudinal sections stained by Masson or sirius red. (N) the mRNA levels of cardiac remodeling gene. (O) CCC-HEH-2 cells were treated with entrectinib and the survival fraction was detected via SRB assay. (P) Representative images of cell morphology. (Q) Apoptosis rates were detected by PI and ANXA5 co-staining and flow cytometry. (R) CCC-HEH-2 cells were treated with entrectinib and Z-VAD-FMK. The protein levels of c-PARP were detected by western blot. GAPDH was used as the loading control. (S) The mitochondrial membrane potential was detected by JC-1 staining and flow cytometry. (T) Representative images of TUNEL staining of mouse heart tissue. (U) Representative images of transmission electron microscopy of mouse heart tissue. Results are presented as mean ± SD. All in vitro experiments were performed with three biological replicates. Unpaired t test was performed to detect the significance of difference between two groups. One-way ANOVA with Sidak’s multiple comparisons test was performed to detect the significance of difference among multiple groups. *p < 0.05, ***p < 0.001, ****p < 0.0001 (vs. The first group). ^####p < 0.0001(vs. the second group).

Numerous studies have established that myocardial hypertrophy is a fundamental pathological process underlying heart failure [23]. Upon the administration of entrectinib, we observed a significant reduction in heart volume and a decrease in the heart weight to tibia length ratio (HW:TL) in mice (Figure 1(F,G)), suggesting a potential impact on cardiac structure. However, WGA revealed no significant difference in the cross-sectional area of cardiomyocytes between entrectinib- and vehicle-treated mice (Figure 1(H)), suggesting that myocardial hypertrophy was not apparently induced by entrectinib within this short study period. In parallel to sunitinib, which is known to cause cardiomyocyte damage leading to heart volume reduction and heart failure [18], we conducted histological evaluations using H&E staining of cardiac tissues. The findings indicated cytoplasmic dispersion, vacuolation, and a decrease in cardiomyocyte density in the hearts of entrectinib-treated mice (Figure 1(K)), indicating the presence of myocardial injury. Biochemical analysis of serum samples from entrectinib-treated mice revealed significant increases in the levels of CK (creatine kinase) and its isoenzyme CK-MB (creatine kinase MB), further confirming the presence of myocardial damage (Figure 1(I,J)). Histological staining using Masson’s trichrome and sirius red revealed a significant increase in the area of collagen fiber positivity in the entrectinib group compared with the control group (Figure 1(L–M) and S1E). We conducted additional analyses on the mRNA levels of genes linked to cardiac remodeling and reported that entrectinib notably increased the expression of Nppa, Nppb, and Myh7 but downregulated Myh6 expression (Figure 1(N)). These changes suggest the onset of myocardial fibrosis and pathological cardiac remodeling in entrectinib-treated mice.

To delve deeper into the myocardial damage induced by entrectinib, we conducted an in vitro study using human embryonic myocardial tissue-derived immortal cardiomyocytes (CCC-HEH-2). After treatment with entrectinib, the cells were analyzed, yielding valuable insights into the nature of cellular damage. SRB staining revealed a marked decrease in the survival rate of CCC-HEH-2 cells following entrectinib treatment (Figure (O)), confirming the presence of myocardial injury. Light microscopy revealed that the cardiomyocytes underwent morphological changes that were characteristic of apoptosis, such as rounding, detachment from neighboring cells, and reduced cell size (Figure 1(P)). To further investigate the apoptotic response, the cells were subjected to dual staining with ANXA5/annexin V and PI and subsequently analyzed by flow cytometry. The results revealed a significant increase in the number of apoptotic cardiomyocytes after treatment with entrectinib (Figure 1(Q) and S1F), underscoring the potential of this drug to trigger programmed cell death. A western blot analysis confirmed the apoptotic effects caused by entrectinib, as evidenced by the significant upregulation of the protein levels of cleaved PARP (poly (ADP-ribose) polymerase family; c-PARP) and cleaved CASP3 (caspase 3; c-CASP3), which are well-established markers of apoptosis (Figure S1G), and the reversal of the apoptotic effects by the use of Z-VAD-FMK, a broad-spectrum caspase inhibitor (Figure 1(R)). Cardiomyocyte apoptosis, a common side effect of numerous antineoplastic agents, is frequently associated with mitochondrial dysfunction [18,19]. Therefore, we employed JC-1 staining to assess the mitochondrial membrane potential in cardiomyocytes as an indicator of mitochondrial health. Our findings revealed a significant increase in the proportion of JC-1 green-positive cardiomyocytes following entrectinib treatment (Figure 1(S) and S1H), suggesting a substantial decrease in mitochondrial membrane potential and indicating the occurrence of mitochondrial damage. To substantiate these in vitro observations in a biological context, we conducted TUNEL staining and transmission electron microscopy on heart tissue samples from mice. TUNEL staining revealed a marked increase in the positive signal in the cardiac tissue of mice treated with entrectinib (Figure 1(T) and S1I), indicating heightened DNA fragmentation, a hallmark of apoptosis. Moreover, transmission electron microscopy provided visual evidence of severe disruption of mitochondrial crista structure in the cardiomyocytes of mice treated with entrectinib (Figure 1(U) and S1J), further highlighting the detrimental effect of entrectinib on mitochondrial integrity.

Entrectinib activates autophagy in cardiomyocytes

Apoptosis is widely recognized as the terminal event in cellular stress responses, marking a critical point in the life cycle of cells under adverse conditions. To elucidate the pivotal initial pathways that are modulated in cardiomyocytes following exposure to entrectinib, we conducted a proteomic analysis on CCC-HEH-2 cells. Our findings revealed a significant shift in the protein expression profile, with 232 proteins upregulated and 348 proteins downregulated in response to entrectinib treatment (Figure 2(A) and S2A). Further analysis using KEGG pathway enrichment indicated that the NTRK receptor signaling, TOR signaling, and autophagy pathways were among the pathways most significantly altered after treatment with entrectinib (Figure 2(B)).

Figure 2.

Entrectinib-activated autophagy in cardiomyocytes mediates the occurrence of cardiotoxicity. (A-C) CCC-HEH-2 cells treated with entrectinib or vehicle were subjected to proteomic analysis. (A) Volcano plot of proteomics. (B) KEGG pathway terms of differentially expressed genes between entrectinib-treated and the control cells. (C) GSEA on autophagy pathway. (D and E) H9c2 and CCC-HEH-2 cells were infected with mCherry-GFP-LC3 virus and then treated with entrectinib or CQ after infection. Autophagic flux assays were performed with the confocal microscope. (F) Quantification of the number of LC3 puncta per cell for panel D. (G) Quantification of the number of LC3 puncta per cell for panel E. CCC-HEH-2 cells were transfected with siRNA targeting ATG7, and then treated with or without entrectinib. (H) Representative images of cell morphology. (I) Apoptosis rates were detected by PI and ANXA5 co-staining and flow cytometry. (J) The protein levels of c-PARP and LC3 were detected by western blot. GAPDH was used as the loading control. (K) Construction and toxicity modeling diagram of mice with cardiac-specific heterozygous knockout of Atg7. The number of mice in each of the four groups is 8, 7, 7, and 8 respectively. The cardiac function of mice was measured by echocardiography. (L) Representative M-mode echocardiographic images from each group in mice. (M) the statistics of ejection fraction. (N) the statistics of fractional shortening. (O) Representative images of cardiac sections stained by hematoxylin-eosin (H&E). (P) Representative images of TUNEL staining of mouse heart tissue. (Q) Representative images of transmission electron microscopy of mouse heart tissue. (R) Representative images of immunohistochemical image targeting LC3 protein of mouse heart tissue. Results are presented as mean ± SD. All in vitro experiments were performed with three biological replicates. Unpaired t test was performed to detect the significance of difference between two groups. One-way ANOVA with Sidak’s multiple comparisons test was performed to detect the significance of difference among multiple groups. ****p < 0.0001 (vs. The first group). ^###p < 0.001, ^####p < 0.0001(vs. the second group). p < 0.0001(vs. The fourth group).

A growing body of research has implicated dysregulated autophagy as a critical factor in disrupting the cellular homeostasis and organ toxicity associated with various drugs [17,19,24]. Given this consideration, our attention was directed toward the autophagy pathway. GSEA revealed significant enrichment of autophagy-related genes in the entrectinib-treated group (Figure 2(C)), suggesting that autophagy might be a central player in the cardiotoxic effects of this drug. To further investigate this phenomenon, we first assessed the levels of autophagy in cardiomyocytes following entrectinib administration. The expression of MAP1LC3/LC3, a protein localized to phagophore and autophagosomal membranes and a well-established autophagy marker, was evaluated using western blot and immunofluorescence assays (Figure S2B through S2E). The results demonstrated a notable increase in LC3 protein levels and the formation of distinct punctate aggregates in entrectinib-treated cardiomyocytes. Given that LC3 protein accumulation can also occur because of the inhibition of autophagosome-lysosome fusion or lysosomal dysfunction, we introduced a GFP-mCherry-LC3 dual-fluorescence reporter system. Our findings revealed a substantial increase in red fluorescence signals in cardiomyocytes following entrectinib treatment, and cotreatment with the lysosome inhibitor chloroquine (CQ) resulted in a significant increase in yellow fluorescence signals (Figure 2(D) through 2 G). These observations suggest that the lysosomal function in cardiomyocytes remained intact and that autophagy was activated in response to entrectinib. To corroborate these in vitro findings, we performed an immunohistochemical analysis of LC3 expression in mouse heart tissue. The results confirmed the significant upregulation of LC3 in the hearts of mice treated with entrectinib (Figure S2F), further substantiating the activation of autophagy in cardiomyocytes by entrectinib. Transmission electron microscopy analysis of mouse heart tissues revealed that entrectinib significantly increased the number of autophagosomes in cardiomyocytes (Figure S2G), clearly demonstrating its role in activating autophagy in these cells. We further investigated whether entrectinib affected the level of mitophagy in cardiomyocytes through a mito-QC reporter plasmid system, and the results revealed no significant change in the level of mitophagy in cardiomyocytes after entrectinib administration (Figure S2H).

Autophagy inhibition rescues the cardiotoxicity of entrectinib

We next sought to determine the role of autophagy in entrectinib-induced cardiomyocyte apoptosis and the associated impairment of cardiac systolic function. Western blot results demonstrated that entrectinib-induced activation of autophagy in cardiomyocytes occurred significantly earlier than apoptosis (Figure S2I). While the apoptosis inhibitor Z-VAD-FMK reversed cardiomyocyte apoptosis without affecting autophagy levels (Figure S2J), the lysosomal inhibitor chloroquine, which blocks autophagic flux, markedly attenuated entrectinib-induced apoptosis (Figure S2K). These findings collectively suggest that autophagy activation may be a key contributor to apoptosis in cardiomyocytes. Given the pivotal role of the autophagy-related genes ATG5 and ATG7 in the biogenesis of autophagosomes, facilitating the lipidation of LC3 and its recruitment to the phagophore membrane, we used small interfering RNA (siRNA) to selectively silence these genes in cardiomyocytes. Our findings indicated that the silencing of either ATG5 or ATG7 led to a significant increase in cardiomyocyte survival and a reversal of the apoptotic signaling cascade activated by entrectinib (Figure 2(H) through 2J and S2L through S2N). These results collectively suggest that autophagy was a critical mediator of the apoptotic response and cardiotoxicity observed in cardiomyocytes following entrectinib treatment.

To explore the role of autophagy in the cardiotoxic effects of entrectinib in in vivo models, we generated a mouse model with a cardiomyocyte-specific heterozygous atg7 knockout (Atg7^f/+; Myh6-MerCreMer mice, namely, Atg7^−/+) by crossing Atg7^flox/flox (Atg7^f/f) mice with Myh6-MerCreMer mice (Figure 2(K)). Echocardiographic assessments revealed that the LVEF and FS in Atg7^f/+ mice treated with entrectinib significantly decreased. In contrast, Atg7^−/+ mice demonstrated resistance to the decline in LVEF and FS induced by entrectinib (Figure 2L through 2N and S2O and Table S2, suggesting that the suppression of autophagy could mitigate the detrimental impact on cardiac systolic function. Further analysis indicated that following entrectinib administration, the HW:TL ratio of wild-type mice markedly decreased (Figure S2P through S2R), whereas the serum level of CK-MB significantly increased (Figure S2S). Notably, these parameters were ameliorated in entrectinib-treated Atg7^−/+ mice , indicating that inhibiting autophagy could ameliorate the cardiotoxic phenotype induced by entrectinib. A histological evaluation via H&E staining demonstrated that heterozygous atg7 knockout significantly attenuated the pathological cardiac injury caused by entrectinib (Figure 2O). Additionally, sirius red and Masson’s trichrome staining revealed that the collagen-positive areas in the hearts of Atg7^−/+ mice treated with entrectinib were considerably reduced (Figure S2T), suggesting that the heterozygous knockout of Atg7 could ameliorate the myocardial fibrosis associated with entrectinib treatment. We also examined the transcriptional levels of genes associated with cardiac remodeling and reported that heterozygous knockout of Atg7 significantly reversed the alterations induced by entrectinib (Figure S2U). These findings collectively suggest that the inhibition of autophagy reversed not only myocardial injury but also the pathological remodeling of the heart induced by entrectinib.

TUNEL staining revealed a pronounced increase in the number of TUNEL-positive cells in Atg7^f/+ mice following entrectinib treatment, indicating increased apoptosis. In contrast, the number of TUNEL-positive cells was significantly reduced in Atg7^−/+ mice after treatment, suggesting that autophagy inhibition protected against drug-induced cell death (Figure 2(P) and S2V). Transmission electron microscopy further confirmed that Atg7^−/+ mice exhibited a substantial amelioration of mitochondrial structural damage compared with Atg7^f/+ mice, underscoring the role of autophagy in mediating entrectinib’s impact on mitochondrial integrity (Figure 2(Q) and S2W). Immunohistochemical analysis of LC3 expression in cardiac tissue revealed that entrectinib significantly elevated LC3 levels in Atg7^f/+ mice, whereas heterozygous atg7 knockout effectively reversed this increase, indicating marked suppression of autophagy signaling in Atg7^−/+ mice (Figure 2(R)). Collectively, these findings suggest that the inhibition of autophagy could effectively reverse the apoptosis and mitochondrial damage induced by entrectinib in cardiomyocytes.

Our in vivo and in vitro studies collectively demonstrated that the suppression of autophagy could mitigate the apoptotic and functional cardiac effects triggered by entrectinib, indicating that autophagy activation was a central mechanism in entrectinib-induced cardiotoxicity. To further elucidate the pathway through which autophagy activation led to myocardial damage, the aforementioned proteomic data were further analyzed, revealing that entrectinib significantly upregulated oxidative stress pathways in cardiomyocytes (Figure S3A and S3B). These findings were substantiated by the detection of increased levels of reactive oxygen species (ROS) using the fluorescence probe DCFH-DA following entrectinib treatment (Figure S3C). Given the established role of autophagy in modulating cellular ROS levels by clearing oxidized cellular components [25–27], we hypothesized that entrectinib might activate autophagy in cardiomyocytes, which could lead to the abnormal degradation of intracellular antioxidant components, resulting in dysregulated ROS levels. The resulting oxidative stress could then precipitate DNA damage and mitochondrial dysfunction, culminating in apoptosis.

Entrectinib activates cardiomyocyte autophagy by inhibiting MTORC1 signaling

To identify kinases that may regulate the activation of autophagy in cardiomyocytes and contribute to the cardiotoxicity of entrectinib, we conducted a phosphoproteomic analysis on CCC-HEH-2 cells, a powerful technique that involves identifying phosphorylated proteins. We identified their phosphorylation sites, quantified their expression levels, and provided an in-depth analysis of the phosphorylated proteome within cells, including cellular kinases that are key regulators of autophagy pathways. The phosphoproteomic results revealed significant changes in phosphorylation levels, with 577 protein phosphorylation sites upregulated and 776 sites downregulated following entrectinib treatment (Figure 3(A)). The KEGG pathway enrichment analysis revealed notable enrichment of the autophagy pathway, further substantiating the role of entrectinib in activating autophagy in cardiomyocytes. Furthermore, significant enrichment was observed in the NTRK receptor signaling pathway, ATP metabolism regulation pathway, and MTOR signaling pathway (Figure 3(B)). MTOR, a serine/threonine protein kinase, exists in two distinct complexes, MTORC1 and MTORC2, which play crucial roles in cardiac development and heart adaptation to stress [22,28,29]. Among these, MTORC1 is widely acknowledged as the predominant negative regulatory kinase of autophagy [30]. Consequently, our focus turned to the MTORC1 signaling pathway. Analysis of the MTOR pathway heatmap indicated that entrectinib treatment led to a significant downregulation of the phosphorylation levels of proteins such as RPS6, EIF4B, and EIF4EBP1, which are positively regulated by MTORC1. Conversely, the phosphorylation levels of proteins such as ULK1 and ULK2, which are under negative regulation by MTORC1, notably increased. These findings indicated that entrectinib inhibited MTORC1 signaling (Figure 3(C)). Further GSEA corroborated these findings, revealing that entrectinib treatment inhibited MTORC1 signaling (Figure S4A).

Figure 3.

Entrectinib activates autophagy through MTORC1 signaling. (A-C) CCC-HEH-2 cells treated with entrectinib or vehicle were subjected to phosphoproteomic analysis. (A) Volcano plot of phosphoproteomics. (B) KEGG pathway terms of differentially expressed genes between entrectinib-treated and the control cells. (C) Heatmap of the MTOR signaling pathway of phosphoproteomics. (D) Representative images of immunofluorescence image targeting p-RPS6 protein of CCC-HEH-2 cells treated with entrectinib. (E) CCC-HEH-2 cells were treated with entrectinib and the protein levels of c-PARP were detected by western blot. GAPDH was used as the loading control. (F) Representative images of immunohistochemical image targeting p-RPS6 protein of mouse heart tissue. (G) CCC-HEH-2 cells were infected with mCherry-GFP-LC3 virus and then transfected with siRNA targeting ULK1. Autophagic flux assays were performed with the confocal microscope. (H) Quantification of the number of LC3 puncta per cell for panel G. (I) CCC-HEH-2 cells were transfected with siRNA targeting ULK1, and then treated with or without entrectinib. Apoptosis rates were detected by PI and ANXA5 co-staining and flow cytometry. (J-K) The protein levels of c-PARP and p-RPS6 and LC3 were detected by western blot. ACTB was used as the loading control. (L) The protein levels of OTUD5 in proteomics. (M-O) CCC-HEH-2 cells were transfected with siRNA targeting OTUD5. (M) Representative images of immunofluorescence image targeting p-RPS6 protein of CCC-HEH-2 cells. (N-O) the protein levels of p-RPS6 and LC3 and c-PARP were detected by western blot. ACTB was used as the loading control. Results are presented as mean ± SD. All in vitro experiments were performed with three biological replicates. Unpaired t test was performed to detect the significance of difference between two groups. One-way ANOVA with Sidak’s multiple comparisons test was performed to detect the significance of difference among multiple groups. **p < 0.01 (vs. The first group). ***p < 0.001 (vs. The first group). ****p < 0.0001 (vs. The first group). ^#p < 0.05(vs. the second group). ^####p < 0.0001(vs. the second group). p < 0.0001(vs. The second group). ^&&&&p < 0.0001(vs. the fourth group).

To substantiate the impact of entrectinib on MTORC1 signaling, we evaluated the phosphorylation status of RPS6, a key downstream effector of MTORC1, in cardiomyocytes following treatment with entrectinib. We observed a significant reduction in phosphorylated RPS6 (p- levels in cardiomyocytes treated with entrectinib (Figure 3(D,E)), suggesting that entrectinib had the capacity to suppress MTORC1 signaling. Consistent with our in vitro observations, the levels of p-RPS6 were markedly decreased in the hearts of mice exposed to the drug (Figure 3(F)). These concordant results from both cellular and tissue-level analyses provided compelling evidence that entrectinib exerted an inhibitory effect on MTORC1 signaling in cardiomyocytes.

To address whether inhibition of the MTORC1 pathway mediated the autophagy and apoptosis of cardiomyocytes triggered by entrectinib, we focused on ULK1, a pivotal autophagy initiation kinase and a key downstream effector in MTORC1-mediated regulation of autophagy [31]. We employed siRNA to silence the ULK1 gene in cardiomyocytes, thereby disrupting the regulatory control of MTORC1 over autophagy. A GFP-mCherry-LC3 fluorescence assay revealed a significant increase in red fluorescent LC3 in cardiomyocytes following entrectinib treatment, and silencing of ULK1 reduced the number and proportion of red fluorescent LC3, indicating that ULK1 silencing could inhibit autophagy activated by entrectinib (Figure 3(G,H)). Further investigation using light microscopy revealed that silencing ULK1 and cotreatment with entrectinib led to a significant increase in the survival of cardiomyocytes (Figure S4B). ANXA5-PI staining and western blot analysis demonstrated that ULK1 silencing inhibited the apoptosis induced by entrectinib (Figure 3(I) and S4C and S4D). These findings suggest that inhibiting ULK1 could reverse the autophagy and apoptosis in cardiomyocytes activated by entrectinib, suggesting that MTORC1 inhibition might be a critical driver of entrectinib-induced cardiotoxicity. AKT1S1/PRAS40, a negative regulatory component of MTORC1, is capable of sensing and binding to substrates, thereby inhibiting MTORC1 activity [32]. We silenced AKT1S1 in cardiomyocytes to increase MTORC1 activity and conducted a series of assays. The results indicated that AKT1S1 gene silencing significantly suppressed autophagy and reversed the apoptosis triggered by entrectinib (Figure S4E and S4F). Additionally, we introduced MHY1485, a pharmacological agonist of MTORC1. Compared with entrectinib alone, the combination of MHY1485 and entrectinib significantly increased the number of surviving cardiomyocytes (Figure S4G). A western blot analysis revealed significant inhibition of apoptotic signaling in cardiomyocytes in the combined treatment group (Figures 3J, K), demonstrating that an MTORC1 agonist could effectively counteract the proapoptotic effects of entrectinib on cardiomyocytes.

OTUD5 is a regulator of MTORC1 signaling and cell survival in cardiomyocytes

Our above findings suggest that entrectinib stimulated autophagy in cardiomyocytes by suppressing MTORC1 signaling. To further elucidate the mechanism by which entrectinib inhibited the MTORC1 pathway, we conducted an in-depth analysis of the proteomic data. The omics data analysis revealed that OTUD5, a positive regulator of MTORC1 signaling, was among the top 10 most significantly downregulated molecules in response to entrectinib. This investigation revealed that the expression of OTUD5 markedly decreased following entrectinib administration (Figure 3(L)). As a deubiquitinating enzyme, OTUD5 plays a crucial role in positively regulating MTORC1 signaling [33,34]. To confirm the impact of entrectinib on OTUD5 expression, we performed a western blot analysis, which revealed a significant decrease in OTUD5 levels in cardiomyocytes treated with entrectinib (Figure S4H). This phenomenon was further corroborated by the results of immunohistochemical staining (Figure S4I), which revealed substantial downregulation of OTUD5 protein levels in the cardiac tissue of mice exposed to entrectinib.

Building on our findings, we delved into the implications of reduced OTUD5 expression on the inhibition of MTORC1 signaling by entrectinib. Immunofluorescence assays revealed a significant decrease in p-RPS6 expression following OTUD5 silencing (Figure 3M), suggesting that OTUD5 played a regulatory role in MTORC1 activity in cardiomyocytes. Further assessment of autophagy levels in cardiomyocytes revealed that OTUD5 silencing robustly activated autophagy (Figure S4J), suggesting that OTUD5 might suppress autophagy in cardiomyocytes through the promotion of MTORC1 activity. We subsequently investigated the potential involvement of OTUD5 in the regulation of apoptosis in cardiomyocytes induced by entrectinib. Previous studies have shown that OTUD5 safeguards against myocardial ischemia‒reperfusion injury by mitigating iron-induced cell death [35], highlighting its critical role in preserving cardiomyocyte viability. Our experimental data revealed a significant reduction in p-RPS6 protein levels and a concurrent increase in c-PARP protein levels upon OTUD5 silencing (Figures 3(N, O)), suggesting that decreased OTUD5 expression could precipitate cardiomyocyte apoptosis. Moreover, the overexpression of OTUD5 reversed the apoptotic effects induced by entrectinib (Figure S4K). These findings underscored OTUD5 as a pivotal factor in modulating MTORC1 signaling and maintaining cell survival in cardiomyocytes.

Entrectinib inhibits transcription of the OTUD5 gene through nuclear HMGB1

To elucidate the molecular mechanism underlying the entrectinib-induced reduction in OTUD5 expression, we detected the stability of the OTUD5 protein and found that etrectinib did not alter OTUD5 protein stability (Figures 4(A, B)). Real-time quantitative PCR revealed a significant decrease in OTUD5 mRNA levels upon entrectinib treatment (Figure 4(C) and S5A), suggesting that entrectinib might lower OTUD5 protein levels by suppressing OTUD5 gene transcription. To further explore the regulation of OTUD5 gene expression, we utilized a luciferase reporter gene system. The promoter region of the OTUD5 gene was fused with a plasmid expressing the luciferase gene, resulting in the construction of a luciferase expression plasmid. The results demonstrated a significant inhibition of luciferase activity in response to entrectinib (Figure 4(D)), suggesting that entrectinib suppressed the transcriptional activity of the OTUD5 gene.

Figure 4.

Entrectinib inhibits OTUD5-MTORC1 by directly binding to HMGB1 and promoting its nuclear localization. (A and B) CCC-HEH-2 cells were treated with CHX (20 μg/mL) with or without entrectinib. (A) the protein levels of OTUD5 were detected by western blot. ACTB was used as the loading control. (B) relative quantification. (C) CCC-HEH-2 cells were treated with entrectinib and the mRNA levels of OTUD5 gene were detected by qPCR. (D) Dual-luciferase reporter assay for OTUD5 gene. Relative luciferase activity in CCC-HEH-2 cells treated with entrectinib. (E) Representative images of immunohistochemical image targeting HMGB1 protein of mouse heart tissue. (F) Representative images of immunofluorescence image targeting HMGB1 protein of CCC-HEH-2 cells. (G) CCC-HEH-2 cells were treated with 3 μM entrectinib for 24 h. Nuclear and cytoplasmic fractions were separated to detect the levels of HMGB1 by western blot. (H) relative luciferase activity in CCC-HEH-2 cells transfected with empty vector or HMGB1 plasmid. (I) CCC-HEH-2 cells were transfected with siRNA targeting HMGB1 and the mRNA levels of OTUD5 gene (left) and HMGB1 gene (right) were detected by qPCR. (J) Schematic showing the structure of HMGB1 protein. (K and L) CCC-HEH-2 cells were transfected with empty vector or HMGB1-Flag plasmid or HMGB1-Δ-Flag plasmid or HMGB1-mutant-Flag plasmid. (K) Representative images of immunofluorescence image targeting Flag. (L) Relative luciferase activity. (M) Molecular dynamics simulation of the interaction between HMGB1 protein and entrectinib. (N) Microscale thermophoresis (MST), illustrating the interaction between HMGB1 protein and entrectinib. (O) CCC-HEH-2 cells were transfected with plasmids as indicated. Representative images of immunofluorescence targeting Flag. Results are presented as mean ± SD. All in vitro experiments were performed with three biological replicates. Unpaired t test was performed to detect the significance of difference between two groups. One-way ANOVA with Sidak’s multiple comparisons test was performed to detect the significance of difference among multiple groups. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 (vs. The first group). ^###p < 0.001(vs. the second group).

To elucidate the mechanism by which entrectinib suppressed OTUD5 transcription, we reexamined the proteomic data, focusing on molecules known to have transcriptional regulatory functions. Our analysis revealed significant upregulation of HMGB1 (high mobility group box 1) expression in response to entrectinib treatment (Figure S5B). HMGB1, a nonhistone chromosomal binding protein predominantly found in the eukaryotic nucleus, is capable of modulating gene transcription by influencing chromatin structure [36]. Furthermore, HMGB1 acts as a modulator of autophagy, promoting its initiation through various mechanisms [37]. We confirmed the increased expression levels of HMGB1 in entrectinib-treated cardiomyocytes and heart tissues using western blot, immunofluorescence, and immunohistochemical analyses (Figure 4E and S5C). Notably, immunofluorescence analysis revealed a marked increase in the nuclear localization of HMGB1 in cardiomyocytes following entrectinib treatment (Figure 4F), suggesting that entrectinib might modulate the subcellular distribution of HMGB1. Further investigation was conducted using a cytoplasmic and nuclear fractionation assay on cardiomyocytes, and a noteworthy increase in nuclear HMGB1 levels was observed (Figure 4G). These findings suggest that entrectinib increased not only the levels of HMGB1 protein but also its nuclear accumulation. The overexpression of HMGB1 significantly inhibited the transcription of the OTUD5 gene (Figure 4H). In contrast, silencing HMGB1 expression significantly increased the mRNA levels of the OTUD5 gene (Figure 4I and S5D). These results suggest that HMGB1 was a key negative regulator of OTUD5 gene transcription in cardiomyocytes. To further confirm the regulatory role of nuclear HMGB1 on OTUD5, we constructed nuclear localization sequence (NLS) mutants of HMGB1 (Figure 4J). HMGB1 has two NLSs with amino acid positions of 27–43 and 178–186 [36], and we deleted both the NLS1 and NLS2 sequences to construct a double-deletion mutant of HMGB1 (Mutant 1, HMGB1-Δ-Flag). In addition, mutation of serine (Ser, S) to glutamate (Glu, E) in two NLS segments of HMGB1 can block its entry into the nucleus [38]. Therefore, we constructed the HMGB1^{S35,39,42,46,181E} plasmid (Mutant 2, HMGB1-Mutant-Flag). The immunofluorescence results revealed that the wild-type HMGB1 plasmid localized to both the nucleus and the cytoplasm, with the fluorescence signal predominantly observed in the nucleus (Figure 4K). In contrast, the two mutant plasmids were expressed primarily in the cytoplasm , indicating that mutations in both nuclear localization sequences could block HMGB1 from entering the nucleus. The results of the luciferase experiments revealed a significant decrease in the transcription level of the OTUD5 gene following the overexpression of the HMGB1 wild-type plasmid, while the two mutant plasmids did not (Figure 4L). These findings suggest that the regulatory function of HMGB1 in negatively impacting OTUD5 gene transcription was dependent on its nuclear localization. On the basis of the above experimental results, we concluded that OTUD5 was a downstream gene that was regulated by HMGB1 in the nuclei of cardiomyocytes. To further confirm the role of HMGB1, we studied the regulatory effect of HMGB1 on the MTORC1 pathway. Immunofluorescence staining revealed that overexpression of HMGB1 significantly decreased the expression level of p-RPS6 in cardiomyocytes (Figure S5E), indicating that HMGB1 inhibited MTORC1. Secreted HMGB1 can bind to the membrane surface receptor RAGE and subsequently regulate the phosphoinositide 3-kinase-AKT-MTOR signaling pathway [39–41]. Therefore, we examined the proteins secreted by cardiomyocytes and observed an absence of HMGB1 secretion regardless of the presence of entrectinib (Figure S5F). We concluded that entrectinib might inhibit OTUD5 and MTORC1 signaling by promoting the nuclear localization of HMGB1.

Entrectinib directly binds to HMGB1 to promote its nuclear localization

We further investigated the mechanism by which entrectinib promoted the nuclear localization of HMGB1. Proteins with nuclear localization sequences can translocate into the nucleus under the action of ligands. On this basis, we used molecular docking to simulate the binding mode of entrectinib as a ligand that directly binds to HMGB1 and reported that entrectinib demonstrated a high binding potential with the HMGB1 protein (docking score of −8.787 kcal/mol) (Figure 4(M)). Microscale thermophoresis can accurately detect interactions between biomolecules. The results showed that entrectinib could interact with the HMGB1 protein at a reasonable concentration (Kd = 65.46 µM) (Figure 4(N)), suggesting that entrectinib might directly bind to HMGB1 as a ligand and promote its nuclear entry. On the basis of the potential binding sites of entrectinib with HMGB1 obtained from the molecular docking analysis, we constructed a series of HMGB1 site-specific amino acid deletion plasmids and used immunofluorescence to detect the subcellular localization of the plasmids. We found that deletion of the 103rd amino acid of phenylalanine could prevent the nuclear entry of HMGB1 caused by entrectinib (Figure 4(O)), indicating that entrectinib might bind to HMGB1 through the 103rd amino acid. The above results indicated that entrectinib promoted the nuclear entry of HMGB1 by directly binding to it and that their interaction depended on the 103rd amino acid of HMGB1.

Inhibition of HMGB1 reverses the autophagy and apoptosis of cardiomyocytes activated by entrectinib

Our research indicated that the activation of HMGB1 might represent an early trigger in the cascade of events leading to cardiotoxicity, specifically through the inhibition of MTORC1-activated autophagy by entrectinib. This assertion is corroborated by a body of literature suggesting that HMGB1 is a promising therapeutic target for various cardiac conditions, including heart failure, myocardial ischemia, and myocardial hypertrophy [42]. Motivated by these insights, we explored the potential of HMGB1 as an interventional target to mitigate the cardiotoxic effects of entrectinib. Utilizing siRNA to silence HMGB1 in cardiomyocytes prior to entrectinib administration, we observed significant suppression of autophagy, as indicated by the results of the GFP-mCherry-LC3 fluorescence assay (Figure 5(A,B)). These findings suggest that HMGB1 silencing could effectively counteract the autophagy induced by entrectinib. Furthermore, light microscopy revealed a substantial increase in the viability of cardiomyocytes following HMGB1 knockdown and drug treatment (Figure S6A). Additional analyses, including western blot and ANXA5-PI staining, demonstrated that HMGB1 silencing reversed the apoptotic signaling initiated by entrectinib and decreased the percentage of apoptotic cells (Figure 5(C) and S6B). These results collectively suggest that targeting HMGB1 might offer a viable strategy to ameliorate the cardiotoxicity associated with entrectinib treatment.

Figure 5.

Inhibition of HMGB1 is protective against entrectinib-induced cardiotoxicity. (A–C) CCC-HEH-2 cells were transfected with siRNA targeting HMGB1, and then treated with or without entrectinib. (A) Autophagic flux assays were performed with the confocal microscope. (B) Quantification of the number of LC3 puncta per cell for panel A. (B) Apoptosis rates were detected by PI and ANXA5 co-staining and flow cytometry. (D–F) CCC-HEH-2 cells were treated with tanshinone IIA with or without entrectinib. (D) Autophagic flux assays were performed with the confocal microscope. (E) Quantification of the number of LC3 puncta per cell for panel D. (F) Apoptosis rates were detected by PI and ANXA5 co-staining and flow cytometry. (G) construction and toxicity modeling diagram of mice with cardiac-specific knockout of Hmgb1. The number of mice in each of the four groups is 8, 8, 6, and 6 respectively. (H) The statistics of ejection fraction. (I) the statistics of fractional shortening. (J) Representative images of cardiac sections stained by hematoxylin-eosin (H&E). (K–N) the mRNA levels of cardiac remodeling gene. (O) Representative images of TUNEL staining of mouse heart tissue. (P) Representative images of transmission electron microscopy of mouse heart tissue. Results are presented as mean ± SD. All in vitro experiments were performed with three biological replicates. Unpaired t test was performed to detect the significance of difference between two groups. One-way ANOVA with Sidak’s multiple comparisons test was performed to detect the significance of difference among multiple groups. ***p < 0.001, ****p < 0.0001 (vs. The first group). ^##p < 0.01, ^###p < 0.001, ^####p < 0.0001 (vs. The second group). p < 0.0001(vs. The second group). ^&&&&p < 0.0001(vs. the fourth group).

Tanshinone IIA, a principal bioactive component extracted from the root of Salvia miltiorrhiza, is known as an HMGB1 inhibitor because of its ability to suppress the expression and secretion of HMGB1 [43–45]. In our study, we examined the effects of combining tanshinone IIA with entrectinib and discovered that this combination could effectively curb the autophagy triggered in cardiomyocytes by entrectinib (Figure 5(D,E)). Moreover, tanshinone IIA significantly reversed the apoptotic effects of entrectinib on cardiomyocytes (Figure 5(F) and S6C through S6E), suggesting that the inhibition of HMGB1 represented a potent strategy to counteract autophagy and apoptosis in cardiomyocytes activated by entrectinib.

Cardiomyocyte-specific knockout of Hmgb1 prevents the cardiotoxicity of entrectinib

To further assess the therapeutic potential of targeting HMGB1 in vivo to protect against the cardiotoxic effects of entrectinib, we generated a mouse model with cardiomyocyte-specific hmgb1 knockout by crossing Hmgb1^f/f mice with Hmgb1^f/f; Myh6-MerCreMer mice. Entrectinib was then administered to these mice via gastric gavage for six weeks to evaluate the effect of Hmgb1 deficiency on entrectinib-induced cardiotoxicity (Figure 5(G) and Table S3). Following entrectinib treatment, echocardiographic assessments revealed a notable decrease in the LVEF and FS in the control group mice, confirming the establishment of a reliable cardiotoxicity model. In contrast, the LVEF and FS of hmgb1^−/− mice were nearly normalized after treatment, suggesting that the absence of Hmgb1 in cardiomyocytes could effectively counteract the cardiotoxicity of entrectinib (Figure 5(H) through 5I and S6F through S6I).

Histological evaluation via H&E staining demonstrated that the knockout of Hmgb1 significantly ameliorated the pathological cardiac injury induced by entrectinib (Figure 5(J)). Serum biochemical analysis further revealed that hmgb1 knockout reversed the increase in the serum level of CK-MB (Figure S6J). Additionally, sirius red and Masson’s trichrome staining demonstrated a significant reduction in collagen-positive areas in the hearts of hmgb1^−/− mice (Figure S6K), suggesting the attenuation of myocardial fibrosis. Transcriptional analysis of cardiac remodeling markers revealed that the expression levels of Nppa, Nppb, Myh7, and Myh6 were significantly restored in hmgb1^−/− mice following entrectinib treatment compared with those in the model group (Figures 5(K) through 5N). These observations indicated a reversal of myocardial injury and pathological remodeling. TUNEL staining of cardiac tissue further confirmed that hmgb1 knockout markedly reduced the number of apoptotic cells, which was increased by entrectinib (Figure 5(O) and S6P). Transmission electron microscopy provided evidence that mitochondrial structure in cardiomyocytes was preserved in hmgb1^−/− mice after entrectinib treatment (Figure 5(P) and S6Q), indicating that Hmgb1 deficiency could prevent the mitochondrial damage associated with this drug.

Immunohistochemical analysis of key signaling molecules involved in entrectinib-induced cardiotoxicity revealed that although entrectinib significantly increased nuclear HMGB1 levels in cardiomyocytes, hmgb1^−/− mice maintained low levels of HMGB1 and were resistant to drug-induced upregulation (Figure S6L). The expression of OTUD5, which was suppressed by entrectinib, was restored in hmgb1^−/− mice (Figure S6M). Similarly, the downregulation of p-RPS6, a marker of MTORC1 signaling, induced by entrectinib was reversed in the knockout mice (Figure S6N). Finally, the increased expression of LC3, a marker of autophagy, in response to entrectinib was mitigated by hmgb1 deletion (Figure S6O). We further investigated the role of the OTUD5-MTORC1 signaling pathway in the protective effects of the HMGB1 inhibitor against cardiotoxicity. Results showed that either silencing OTUD5 or treating with rapamycin, the MTORC1 inhibitor, abolished the protective effect of tanshinone IIA against entrectinib-induced cardiomyocyte apoptosis (Figure S6R through 6T), suggesting that restoration of OTUD5-MTORC1 signaling is a critical mechanism through which tanshinone IIA exerts its cardioprotective function. In conclusion, our study provided compelling evidence that targeting HMGB1, either through pharmacological inhibition or genetic knockout, could effectively ameliorate the cardiotoxic signaling pathways activated by entrectinib.

Tanshinone IIA can intervene in the cardiotoxicity of entrectinib

Next, we examined the protective effects of the HMGB1 inhibitor tanshinone IIA when it was coadministered with entrectinib in vivo. Echocardiographic assessments were conducted to evaluate the effect of this combination on entrectinib-induced cardiotoxicity (Figure 6(A) and Table S4. Compared with the control treatment, entrectinib treatment led to a notable decrease in LVEF and FS in mice. Promisingly, LVEF and FS in mice treated with both tanshinone IIA and entrectinib significantly improved, demonstrating that tanshinone IIA could mitigate the cardiotoxic effects of entrectinib (Figure 6(B) through 6C).

Figure 6.

Tanshinone IIA can rescue entrectinib-induced cardiotoxicity. C57BL/6J mice were treated with vehicle, entrectinib (200 mg/kg), tanshinone IIA (20 mg/kg) or tanshinone IIA plus entrectinib for 6 weeks by means of intragastric administration. The number of mice in each of the four groups is 9, 9, 8, and 8 respectively. (A) Representative M-mode echocardiographic images from each group in mice. (B) The statistics of ejection fraction. (C) the statistics of fractional shortening. (D) Representative whole heart images. (E) Heart weight to tibia length ratio (HW:TL). (F) serum from the mice was analyzed for CK-MB levels. (G) Representative images of cardiac sections stained by hematoxylin-eosin (H&E). (H-I) Representative images of cardiac longitudinal sections stained by Masson or sirius red. (J) Representative images of TUNEL staining of mouse heart tissue. (K) Representative images of transmission electron microscopy of mouse heart tissue. (L) Representative images of immunohistochemical image of mouse heart tissue. Results are presented as mean ± SD. Unpaired t test was performed to detect the significance of difference between two groups. One-way ANOVA with Sidak’s multiple comparisons test was performed to detect the significance of difference among multiple groups. ****p < 0.0001 (vs. The first group). ^#p < 0.05, ^##p < 0.01, ^###p < 0.001, ^####p < 0.0001 (vs. The second group).

Our findings revealed that tanshinone IIA effectively mitigated the decrease in heart size and in the HW:TL ratio induced by entrectinib (Figure 6(D) through 6E). Histological assessment via H&E staining demonstrated that concurrent administration of tanshinone IIA with entrectinib markedly ameliorated the pathological myocardial damage typically caused by entrectinib alone (Figure 6(G)). Furthermore, serum biochemical analyses indicated that the combination therapy significantly decreased the serum levels of CK-MB in mice treated with entrectinib (Figure 6F). Collagen deposition, as evidenced by sirius red and Masson’s trichrome staining, was substantially reduced in the hearts of mice treated with the tanshinone IIA and entrectinib combination than in those treated with entrectinib alone (Figure 6(H,I) and S6U). These observations supported the capacity of tanshinone IIA to reverse the myocardial injury and the pathological remodeling processes triggered by entrectinib. TUNEL staining of cardiac tissue revealed that tanshinone IIA significantly decreased the increase in the number of apoptotic cells induced by entrectinib treatment (Figure 6(J)). Transmission electron microscopy further confirmed that entrectinib disrupted the mitochondrial crista structure in cardiomyocytes and that the mitochondria in mice concurrently treated with tanshinone IIA remained intact (Figure 6(K)). Collectively, these results suggest that tanshinone IIA not only countered the pro-apoptotic and mitochondrial-damaging effects of entrectinib but also promoted the preservation of mitochondrial integrity, thereby potentially safeguarding against drug-induced cardiotoxicity.

We conducted an immunohistochemical analysis to assess the expression levels of key signaling molecules involved in the cardiotoxicity associated with entrectinib (Figure 6(L)). Our findings indicated a notable increase in nuclear HMGB1 levels in cardiomyocytes following entrectinib treatment, with tanshinone IIA effectively reversing this upregulation. The expression of OTUD5, which was significantly diminished by entrectinib, was substantially restored upon the coadministration of tanshinone IIA. This restoration was also observed for the expression of p-RPS6, an essential MTORC1 substrate, which exhibited a significant downregulation under entrectinib treatment and was normalized with tanshinone IIA. Furthermore, the elevated levels of LC3, an autophagy marker induced by entrectinib, were attenuated by tanshinone IIA. These collective results underscored the therapeutic potential of HMGB1 inhibition in mitigating the cardiotoxic effects of entrectinib, confirmed that HMGB1 was a viable interventional target for countering the cardiotoxicity of entrectinib and validated the efficacy of the HMGB1 inhibitor tanshinone IIA in both in vivo and in vitro models.

Discussion

In this study, we utilized a combination of in vitro and in vivo approaches to elucidate the mechanisms of cardiotoxicity induced by entrectinib. Our results demonstrated that entrectinib initiated apoptosis by triggering autophagy in cardiomyocytes, ultimately culminating in myocardial damage and cardiotoxicity. Mechanistically, entrectinib operates through direct molecular interactions with the HMGB1 protein, facilitating its nuclear translocation. In the nuclear milieu, HMGB1 disrupts MTORC1 signaling, consequently activating autophagy in cardiomyocytes through the suppression of OTUD5 gene transcription. Furthermore, we substantiated the therapeutic potential of tanshinone IIA, an HMGB1 inhibitor, in countering the cardiotoxic effects of entrectinib, highlighting its promise as a mitigating agent in the clinical landscape of cancer therapeutics (Figure 7).

Figure 7.

Illustration of the proposed mechanism by which entrectinib induces cardiotoxicity. Entrectinib binds to HMGB1 and activates cardiomyocyte autophagy by inhibiting OTUD5-MTORC1 signaling to induce cardiotoxicity. Tanshinone IIA can rescue entrectinib-induced cardiotoxicity by inhibiting HMGB1.

Entrectinib is a unique pan-tumor targeted therapy with proven efficacy against intracranial neoplasms. However, its adverse impact on the cardiovascular system poses a formidable challenge in the comprehensive care of cancer patients, potentially undermining the success of optimal tumor treatment [6]. In a comprehensive analysis of clinical trials involving 355 patients treated with entrectinib, congestive heart failure was identified in 3.4% of the patient population, with grade 3 severity observed in 2.3% of cases [7]. The onset of congestive heart failure necessitates the interruption, and potentially the permanent cessation, of entrectinib therapy for these individuals, thereby risking the failure of their tumor treatment regimen. Consequently, addressing the clinical toxicity of congestive heart failure is a critical challenge in oncology. The integrity of the cardiomyocyte count is fundamental to the maintenance of cardiac contractility, making myocardial injury an important contributor to heart failure [46,47]. In our study, we identified cardiomyocyte apoptosis as the precipitating factor in entrectinib-induced heart failure. Additionally, among these 355 patients, 3.1% experienced a prolongation of the QTcF interval by more than 60 ms following the initiation of entrectinib, with 0.6% exhibiting a QTcF interval exceeding 500 ms [7]. Prolongation of the QTcF interval is a known cardiac risk, and studies have correlated myocardial injury with this electrocardiographic alteration [48]. The involvement of cardiomyocyte apoptosis in QTcF interval prolongation associated with entrectinib treatment represents an important research question that merits further investigation. Furthermore, two cases of myocarditis in entrectinib-treated patients have been reported [49,50]. Considering that myocarditis might also cause the death of cardiomyocytes, the secondary toxic effect of entrectinib-induced myocarditis on cardiomyocytes might also partially contribute to cardiotoxicity.

In recent years, the pivotal role of autophagy in drug-induced cardiotoxicity has attracted considerable interest [51,52]. Research has demonstrated that the modulation of autophagy, either through its activation or inhibition, canresult in the dysregulation of critical proteins within cardiomyocytes, precipitating cellular damage or dysfunction [18,19]. Our investigation revealed that entrectinib triggered autophagy in cardiomyocytes and that the suppression of this process could ameliorate the myocardial injury and deterioration of cardiac systolic function associated with entrectinib treatment. These findings suggest that autophagy activation is a primary contributor to the cardiotoxic effects of entrectinib. On the basis of our proteomic analysis, we hypothesize that entrectinib may stimulate autophagy in cardiomyocytes, leading to the degradation of essential intracellular antioxidants. This degradation can disrupt the intracellular reactive oxygen species (ROS) balance, culminating in mitochondrial damage and the initiation of apoptotic pathways. However, the key factors involved in the activation of autophagy and oxidative stress damage and their regulatory mechanisms require further investigation.

Recent studies have established a robust correlation between HMGB1 and the activation of autophagy in cardiomyocytes [42,53]; however, the underlying regulatory mechanisms remain enigmatic. Our examination of entrectinib-induced cardiotoxicity revealed a novel nuclear pathway through which HMGB1 modulated autophagy. Specifically, HMGB1 hindered the MTORC1 signaling cascade by suppressing the transcription of the OTUD5 gene, thereby promoting autophagy. In our study, entrectinib facilitated the nuclear translocation of HMGB1 without altering the expression levels of HSPB1, a reported HMGB1 downstream protein that regulates autophagy by promoting vesicle transport. Our research not only identified a novel downstream molecule of HMGB1 but also revealed a previously unknown regulatory mechanism of the MTORC1 signaling pathway in cardiomyocytes, shedding light on the regulatory factors and mechanisms of the MTORC1 signaling pathway within the nucleus.

In our research, the application of HMGB1 inhibitors or genetic silencing of HMGB1 in cardiomyocytes effectively reversed the myocardial injury and cardiac dysfunction induced by entrectinib in both in vivo and in vitro experimental models. These findings underscored the potential of HMGB1 as a therapeutic target for mitigating the cardiotoxic effects of entrectinib. The activation and upregulation of the HMGB1 protein have been implicated in a spectrum of cardiovascular pathologies, including myocarditis, myocardial hypertrophy, myocardial infarction, and atherosclerosis [54–56]. Most notably, in alignment with the findings of numerous other investigative teams, our research revealed that the activation of HMGB1 was pivotal for the myocardial damage and cardiotoxicity associated with doxorubicin, sunitinib, and entrectinib. This convergence of evidence suggests that HMGB1 in cardiomyocytes could serve as a biomarker for drug-induced cardiotoxicity, a hypothesis that merits further rigorous and systematic exploration. In addition, we suggest that in the process of drug development, the deliberate avoidance of molecular structures that interact with the HMGB1 protein may proactively impede drug-induced myocardial damage by modulating HMGB1, thus potentially decreasing the occurrence of cardiotoxicity. This approach signifies a considerable advancement in the design of safer and more efficacious therapeutics.

Tanshinone IIA, a principal bioactive component extracted from the root system of Salvia miltiorrhiza, has garnered research attention for its potential to modulate HMGB1 activity [43]. Studies have indicated that tanshinone IIA has the potential to suppress HMGB1 by impeding its release or decreasing its intracellular expression [44,45]. The potential of this capability is particularly noteworthy, given that tanshinone sodium IIA sulfonate injection is already in clinical use for the adjunctive treatment of cardiovascular ailments such as coronary heart disease, angina pectoris, and myocardial infarction. Clinical trials have demonstrated its high efficacy in managing angina pectoris, with a favorable side effect profile, positioning it as a valuable agent in cardiovascular therapy [43]. Our research has identified tanshinone IIA as a viable intervention strategy for mitigating the cardiotoxicity associated with entrectinib. This discovery not only bolsters the clinical utility of tanshinone IIA but also paves the way for expanding its clinical indications. By leveraging its HMGB1 inhibitory properties, tanshinone IIA may offer a novel therapeutic avenue to counteract drug-induced cardiotoxicity, enhancing patient safety in oncological treatment protocols.

In conclusion, we identified a novel and central role for autophagy in the regulation of adverse remodeling and dysfunction in entrectinib-treated hearts. Entrectinib activates autophagy in cardiomyocytes by directly binding to HMGB1 and promoting its entry into the nucleus, thereby inhibiting downstream OTUD5-MTORC1 signaling. Our findings unequivocally illustrate that entrectinib-induced autophagy is the primary mechanism leading to adverse cardiac remodeling and ventricular dysfunction. Deletion or inhibition of HMGB1, the protein that interacts with entrectinib, restored the level of entrectinib-induced autophagy in cardiomyocytes and reversed the cardiotoxicity of entrectinib in vivo and in vitro. Our findings not only elucidate the mechanism of entrectinib cardiotoxicity but also provide a potential strategy for clinical intervention in this context.

Materials and methods

Study design

The objectives of this study were to examine the role of autophagy on entrectinib-induced cardiac complications and to elucidate the underlying mechanisms. Our research initially unveiled the stimulatory impact of entrectinib on cardiomyocyte autophagy, utilizing a combination of proteomics and biochemical assays. Subsequent in-depth investigations, employing both in vivo and in vitro models of autophagy suppression, confirmed that autophagy plays a pivotal role in the cardiotoxicity associated with entrectinib treatment. Further exploration through phosphoproteomics revealed that the suppression of the MTORC1 signaling pathway is instrumental in the activation of autophagy within cardiomyocytes. We discovered that entrectinib triggers a decrease in OTUD5 gene expression by activating HMGB1, consequently dampening the MTORC1 signal. Our subsequent experiments, including microscale thermophoresis, suggested that entrectinib might directly interact with HMGB1, thereby promoting its activation. In the culmination of our study, we demonstrated the therapeutic potential of tanshinone IIA, the HMGB1 inhibitor, in mitigating the cardiotoxic effects of entrectinib, as evidenced by both in vivo and in vitro model assessments. All animal studies were performed under approved protocols of the Institutional Animal Care and Use Committee and the Institutional Review Board (IACUC-s21-002). The sample size for animal studies was determined by statistical analysis of variance (ANOVA) and on the basis our experience with similar studies. The sample size for each experimental group is indicated in the figure legends and is between six and eight mice per group. For cell studies, a minimum of three or four experimental replicates were performed and the numbers of replicates are presented in the figure legends. No outliers were excluded from any experiments or analyses reported in this manuscript.

Animal study

The C57BL/6J mice were purchased from Zhejiang Vital River Laboratory Animal Technology Co., Ltd. (Jiaxing, China) and housed in barrier facilities with a 12 h light/dark cycle and free food and water. The mice were housed for 1 week to adapt to the new environment before drug treatment. Before tissue or blood collection, the mice were anesthetized with 2% isoflurane and then sacrificed via cervical dislocation.

Entrectinib (TargetMol USA, T3678) and tanshinone IIA sulfonate sodium (TargetMol USA, T2946) were dissolved in cyclodextrin (Aladdin, H108813) to form a stock solution. The mice (7–9 weeks old) were treated with vehicle, 200 mg/kg entrectinib or 20 mg/kg tanshinone IIA sulfonate sodium daily through intragastric administration for 42 days.

Echocardiography

Cardiac function was evaluated in mice with 1% isoflurane (RWD Lifescience, R510-22) using a Vevo3100 Imaging System (Fujifilm Visual Sonics, Vevo3100). VisualSonics software was used to analyze the left ventricular internal diameter in diastole and systole and the end-diastolic volume or end-systolic volume. The left ventricular EF was calculated according to the following formula: EF (%) = [(end-diastolic volume – end-systolic volume)/end-diastolic volume] × 100. The left ventricular Fractional Shortening (FS) was calculated according to the following formula: FS (%) = [(left ventricular internal diameter in diastole – systole)/left ventricular internal diameter in diastole] × 100.

Histopathological and immunohistochemical analysis

Heart samples were fixed in 10% phosphate-buffered formalin (Sigma-Aldrich, F8775), dehydrated, embedded in paraffin, and sectioned at 3-μm intervals. After de-waxing and rehydration, the sections were stained with H&E, sirius red (Abcam, ab150681), Masson Trichrome (Beyotime Biotechnology, C0189S) or specific antibodies for analysis.

Biochemical analysis

Serum from mice or rats was collected after centrifugation of the blood. The serum cardiac enzyme activity of CK and CK-MB was determined by a fully automatic biochemical detection machine (Cobas c 311, Roche Diagnostics GmbH, Germany) using specific detection kits (Roche, 07190808190).

Transmission electron microscopy analysis

Left ventricular tissues were collected from indicated mice. Tissue blocks of the left ventricle were cut out with a surgical instrument (far away from the apes of the heart to ensure consistent alignment of muscle filaments) and fixed in 1 mL fresh 2.5% glutaraldehyde (Scientific Phygene, PH9003) solution at room temperature for 2 h and stored at 4°C overnight. Tissue blocks were fixed with 1% osmic acid for 1 h and were stained with 2% uranium for 0.5 h. After being dehydrated and embedded, the samples were sectioned with an ultrathin slicer. Transmission electron microscope (Thermo Fisher Scientific, TECNAI 10) was used to collect images.

Cell culture and drug treatment

The human embryonic cardiac tissue-derived cell line CCC-HEH-2 (RRID: CVCL_VU29) was obtained from the National Infrastructure of Cell Line Resource of China. The cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM; Gibco, 10,569,010) supplemented with 10% fetal bovine serum (FBS; HyClone, SV30160.03), 100 U/mL penicillin and 100 μg/mL streptomycin (Gibco, 10,378,016) in a humidified atmosphere with 5% CO₂ at 37°C. All cell lines routinely tested negative for mycoplasma contamination.

The cells were treated with 3 μM entrectinib for the indicated time points or 1.5, 3 or 6 μM entrectinib for 24 h for cell survival analysis and western blot. In specific samples, 10 μM CQ (Sigma-Aldrich, C6628), 20 μg/mL CHX (MedChemExpress, HY-12,320) or 5 μM tanshinone IIA were used.

Immunofluorescence assay

For in vitro assay, cells were seeded on poly-D-lysine-coated coverslips (CORNING, 354,086). After treatment, cells were fixed with 4% paraformaldehyde in phosphate-buffered saline (PBS; Gibco, 10,010,023) for 20 min at room temperature. The cells were then permeabilized by 1% Triton X-100 (BioFROXX, 1139ML100) in PBS for cells for 10 min at 4°C and blocked with 4% bovine serum albumin (Sigma-Aldrich, B2064) in PBS for 30 min. Cells were incubated with primary antibodies overnight, and stained with secondary antibody for 2 h and DAPI for 5 min and mounted for fluorescence microscope (Leica, TCS SP8).

The following primary antibodies were used: anti-p-RPS6 (Rodent Specific; Cell Signaling Technology, 4858), anti-HMGB1 (HuaBio Technology, ET1601-2) and anti-Flag (yoche-biotech, AYC01-100). The following secondary antibodies were used: Alexa Fluor 488- or Alexa Fluor 568-conjugated secondary antibodies (Thermo Fisher Scientific, A21202, A10042).

Autophagy flux assay

Cells were seeded in Nunc™ Lab-Tek™ II Chamber Slide™ (Thermo Fisher Scientific, 154,534) and infected with an adenovirus encoding the tandem fluorescent probe mCherry-GFP-LC3 and then treated as indicated in figure legends. Cells were fixed with 4% paraformaldehyde (Sigma-Aldrich, P6148) in PBS for 20 min at room temperature. After washing with PBS, the cells were permeabilized with ice-cold 0.1% Triton X-100 in PBS for 10 min and stained with DAPI (Dojindo, D212) for 5 min. Then cells were mounted for analysis by confocal microscope (Leica TCS-SP8 and Leica DMi8, Germany).

Microscale thermophoresis

HMGB1-His protein was purified and dialyzed. When purified HMGB1-His was used as a target, the protein was labeled with RED-Tris-NTA Protein Labeling Kit (NanoTemper Technologies, MO-L018). 200 nM labeled HMGB1 and 512 uM to 15.6 nM entrectinib was mixed in PBS-T (PBS with 0.05% Tween 20). The sample was loaded into the premium capillaries (NanoTemper Technologies, MO-K025), and the MST measurements were performed at 25°C, 100% excitation nano-red and medium MST-power. Kd values were calculated using the mass action equation available in MO.Affinity Analysis software.

Proteomics

In this research, Applied Protein Biotechnology Company conducted proteomic analysis and preliminary data interpretation. To discern variations in protein expression among the comparison groups, we employed 4D-Label-free quantitative proteomics for sequencing. The study encompassed a blank control group and an entrectinib treatment group, where cells were treated with 3 μM entrectinib for 24 h, with triplicate samples for each. Cellular lysis was achieved using a solution containing 4% (w:v) SDS, 100 mM Tris-HCl, and 1 mM DTT, adjusted to a pH of 7.6. The mass spectrometry procedure comprised several key steps: protein extraction, enzymatic digestion of peptides, chromatographic separation, and acquisition of liquid chromatography-tandem mass spectrometry (LC-MS/MS) data, followed by database searching. Bioinformatics analyses were conducted to identify proteins, assess differential expression, and perform functional analyses. Gene Set Enrichment Analysis (GSEA) was executed using version 4.3.2 of the GSEA software.

Cytoplasmic and nuclear fractionation and western blot

Separation of CCC-HEH-2 cell lysate into nuclear and cytoplasmic fractions was performed with Nuclear and Cytoplasmic Protein Extraction Kit (Beyotime Biotechnology, P0027). For western blot, cell lysates were separated on SDS polyacrylamide gels, transferred to PVDF membranes (Millipore Corporation, IPVH00010) and blocked with 5% skim milk. Incubation of primary antibodies, secondary antibodies and the Western Lightning Plus-ECL Enhanced Chemiluminescence Substrate (PerkinElmer, NEL105001EA) were applied to detect the signal.

Primary antibodies directed against GAPDH (sc-25778) and ACTB (sc-58673) were purchased from Santa Cruz Biotechnology. Antibodies directed against LC3 (2775S), OTUD5 (20087), p-RPS6 (4858) and RPS6 (2317) were purchased from Cell Signaling Technology. Antibodies directed against HMGB1 (ET1601-2), c-PARP (ET1608-10), ATG5 (ET1611-38) and LMNB1 (ET1606-27) were purchased from HuaBio. HRP-labeled secondary antibodies (GAR007 and GAM007) were purchased from MultiSciences (Lianke) Biotech.

Quantitative real-time PCR (qPCR)

Total RNA was extracted by TRIzol reagent (Invitrogen, 15,596,026) and reverse transcribed to cDNA by a reverse transcription kit (Transgene Biotech, AT311-03). Quantitative PCR was performed on a 7500 Fast System (Applied Biosystems, Singapore). More than 3 independent experiments were performed for statistical analysis.

The primer sequences were as follows:

Nppa forward: GAGAGAAAGAAACCAGAGTG

Nppa reverse: GTCTAGCAGGTTCTTGAAATC

Nppb forward: AATTCAAGATGCAGAAGCTG

Nppb reverse: GAATTTTGAGGTCTCTGCTG

Myh6 forward: AATCCTAATGCAAACAAGGG

Myh6 reverse: CAGAAGGTAGGTCTCTATGTC

Myh7 forward: TTGGGAAATTCATCCGAATC

Myh7 reverse: CCAGAAGGTAGGTCTCTATG

Actb forward: ACCTTCTACAATGAGCTGCG

Actb reverse: CTGGATGGCTACGTACATGG

OTUD5 forward: CAGGAGCATTGGTTTGAAAAGG

OTUD5 reverse: GCTTTCGCACAACCTCATG

HMGB1 forward: GATATGGCAAAAGCGGACAAG

HMGB1 reverse: GGCGATACTCAGAGCAGAAG

GAPDH forward: ACATCGCTCAGACACCATG

GAPDH reverse: TGTAGTTGAGGTCAATGAAGGG

Flow cytometry analysis

For ANXA5-PI staining assay, cells were harvested after drug treatment or transfection as indicated and washed with PBS. Then cells were placed in tubes and stained with FITC ANXA5 Apoptosis Detection Kit I (BD Pharmingen, 556,547) according to the manufacturer´s protocol.

For the mitochondrial membrane potential assay, JC-1 probe (5 μM; Sigma-Aldrich, T4069) was used to measure mitochondrial depolarization in CCC-HEH-2 cells. After treatment, cells were digested with trypsin and incubated with an equal volume of JC-1 solution (5 μg/mL) at 37°C for 20 min in the dark. After washing, resuspend cells with 1× buffer.

A FACSCalibur cytometer (BD Biosciences, USA) was employed for analysis.

Plasmid construction

Standard PCR was utilized to prepare the luciferase constructs used in this study. A 2000-bp OTUD5 promoter construct (−2001/-1 OTUD5), corresponding to the sequence from −2001 to −1 (relative to the transcriptional start site) of the 5ʹ-flanking region of the human OTUD5 gene, was generated from human genomic DNA using specific forward and reverse primers, respectively. PCR product was cloned into Bgl II and HindIII sites of pGL3-basic vector.

Mutations of HMGB1 were performed by PCR-based site-directed mutagenesis using PrimeSTAR HS polymerase (Takara) according to the manufacturer’s instructions.

Cell transfection

Cells were seeded into 6-well plates at 8 × 10⁴ per well and grown to 50–60% confluence. si-jetPRIME (Polyplus-transfection; Illkirch, 114–15; France) was used according to the manufacturer’s recommendations. siRNA oligonucleotides were transfected at a final concentration of 12 nM. The transfection solution was changed to fresh medium plus 10% FBS for further study 6 h after transfection.

The following oligonucleotides were provided by GenePharma (Shanghai, China) as siRNAs targeting the indicated genes:

si ATG5 sence: 5’-GCCUGUAUGUACUGCUUUAAC-3’

si ATG7 sence: 5’-CAAUGAGAAGGACAGAUAAGA-3’

si HMGB1 sence: 5’-GGCCCGUUAUGAAAGAGAAAU-3’

si OTUD5#1 sence: 5’-GAUGCUAGAAGACAAGAAACG-3’

si OTUD5#2 sence: 5’-AGCAGAUGCUAGAAGACAAGA-3’

si ULK1 sence: 5’-AGUUCGAGUUCUCCCGCAAGG-3’

si AKT1S1 sence: 5’-GCCTTCAATTTACGTTCTTTA-3’

Luciferase activity assay

OTUD5 promoter-driven luciferase plasmid and pRL-SV40 plasmid expressing renilla luciferase were co-transfected into subconfluent monolayer cells using lipofectamine reagent LipofectamineTM2000 (Invitrogen Corporation, 18,324,012). After 5 h of transfection, cells were treated with various concentrations of entrectinib or transfected with plasmids as indicated. Luciferase activity was detected 12 h after treatment using the dual luciferase reporter assay system (Vazyme Biotechnology Corporation,

DL101-01). The relative luciferase activity was normalized with Renilla luciferase activity.

Statistical analysis

Statistical analysis was performed using GraphPad Prism 9. All data were expressed as the mean value ± standard deviation (SD). When comparing the difference between two groups, Student’s t-test (unpaired, two-tailed) was applied. One-way ANOVA was performed to detect the significance among multiple groups.

Funding Statement

This study was supported by National Natural Science Foundation of China [82474018 to Dr. Xu and 82173893 to Dr. Luo], Natural Science Foundation of Zhejiang Province [LY23H310003 to Dr. Xu], the Fundamental Research Funds for the Central Universities [226–2023-00151 to Dr. Xu] and the “Pioneer” and “Leading Goose” R&D Program of Zhejiang province [2024C03144 to Dr. Xu].

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/15548627.2025.2576619