Excessive autophagic degradation of MYLK3 causes sunitinib-induced cardiotoxicity

Abstract

Sunitinib is a receptor tyrosine kinase inhibitor used for the treatment of renal cell carcinoma and imatinib-resistant gastrointestinal stromal tumors. Clinical data have shown that patients receiving sunitinib develop reduced cardiac function, arrhythmia and heart failure, thereby largely limiting its clinical use. However, the molecular mechanisms underlying sunitinib-induced arrhythmogenesis remain unclear. Here, utilizing the human induced pluripotent stem cell-derived cardiomyocyte (iPSC-CM) model, we found that sunitinib caused a variety of deleterious phenotypes, including cardiomyocyte death, sarcomeric disorganization, irregular Ca²⁺ transients, impaired ATP2A2a/SERCA2a (ATPase sarcoplasmic/endoplasmic reticulum Ca²⁺ transporting 2a) activity, arrhythmia, and excessive macroautophagy/autophagy. Mechanistically, SQSTM1/p62 (sequestosome 1) interacts with MYLK3 (myosin light chain kinase 3) and drives excessive autophagic degradation of MYLK3 in sunitinib-treated iPSC-CMs. Downregulation of MYLK3 suppresses the phosphorylation of CAMK2/CAMKII (calcium/calmodulin dependent protein kinase II), thereby reducing the phosphorylation level of its downstream substrate PLN (phospholamban), leading to impaired ATP2A2a/SERCA2a activity and subsequent Ca²⁺ dyshomeostasis and arrhythmia. Moreover, pharmacological intervention of the cardiac myosin activator omecamtiv mecarbil (OM) or overexpression of MYLK3 significantly restored the expression of MYLK3 and reversed pathogenic phenotypes in sunitinib-treated iPSC-CMs. Nanoparticle delivery of OM effectively prevented sunitinib-induced cardiac dysfunction in mice. Our findings suggest that sunitinib-induced MYLK3 degradation causes the inhibition of the CAMK2-PLN-ATP2A2a signaling pathway and leads to sunitinib-induced arrhythmogenesis, and that MYLK3 can act as a novel cardioprotective target for sunitinib-induced cardiotoxicity.Abbreviation: ACTN:actinin alpha;APD:action potential duration; ATG:autophagy related;ATP2A2a/SERCA2a:ATPase sarcoplasmic/endoplasmicreticulum Ca2+ transporting 2a;BafA1:bafilomycin A₁;Caff: caffine; CAMK2/CAMKII:calcium/calmodulin dependent protein kinase II;CASP3:caspase 3;CQ, chloroquine;DADs:delayed afterdepolarizations; EAD:early afterdepolarization; ECG: electrocardiogram; EF: ejectionfraction; FS: fractional shortening; iPSC:inducedpluripotent stem cell;iPSC-CM: inducedpluripotent stem-cell-derived cardiomyocyte;ISO: isoprenaline; LVIDs: left ventricular end systolic diameter;LVIDd: left ventricular end diastolic diameter;MAP1LC3/LC3:microtubuleassociatedprotein 1 light chain 3;MYL2v/MLC2v:myosin light chain 2 v;MYLK3:myosin light chain kinase 3;OE: overexpression; OM:omecamtiv mecarbil; PLN: phospholamban;SIC:sunitinib-induced cardiotoxicity; SR:sarcoplasmic reticulum; TUNEL:TdT-mediated dUTP nick end labeling.

Keywords: Arrhythmia; MYLK3; autophagy; iPSC-derived cardiomyocytes; sunitinib-induced cardiotoxicity.

Figure 1.

Sunitinib-induced morphological changes and apoptosis in iPSC-CMs. (A) The sketch map of SIC. (B) Bar graph to compare the cell viability among vehicle-, 3 μM sunitinib-, and 10 μM sunitinib-treated iPSC-CMs by one-way ANOVA (Tukey method). **p < 0.01 and ****p < 0.0001. n = 4 biologically independent experiments. (C) Immunofluorescent staining of vehicle-, 3 μM sunitinib-, and 10 μM sunitinib-treated iPSC-CMs using cardiac-specific markers TNNT2 (green) and ACTN (red). DAPI indicates nuclear staining (blue). Scale bar: 100 μm. (D) Sarcomeric arrange analysis in vehicle-, 3 μM sunitinib-, and 10 μM sunitinib-treated iPSC-CMs using MATLAB (MathWorks). Sarcomeres were marked using TNNT2 (green) and ACTN (red). Scale bar: 10 μm. (E) Representative confocal images showing TUNEL staining in vehicle-, 3 μM sunitinib-, and 10 μM sunitinib-treated iPSC-CMs. Scale bar: 100 μm. (F) Bar graph to compare the ratio of TUNEL:DAPI between different groups in E by one-way ANOVA (Tukey method). **p < 0.01 and ****p < 0.0001. n = 10 fields of vision. (G) Western blot analysis of protein expression of cleaved CASP3 and CASP3 in vehicle-, 3 μM sunitinib-, and 10 μM sunitinib-treated iPSC-CMs. (H) Bar graphs to compare the cleaved CASP3 expression (cleaved:total) among three different groups in G by one-way ANOVA (Tukey method). ****p < 0.0001. n = 4 biologically independent experiments.

Figure 2.

Sunitinib-induced arrhythmia and irregular Ca²⁺ handling in iPSC-CMs. (A-B) Representative action potential waveforms recorded from vehicle- and 3 μM sunitinib-treated iPSC-CMs. Red arrows indicate arrhythmic events with DADs. (C) Bar graph to compare arrhythmic rate in vehicle- and 3 μM sunitinib-treated iPSC-CMs by Fisher’s test. *p < 0.05. n = 28–35 cells. (D) Representative single action potential waveforms of vehicle- and 3 μM sunitinib-treated iPSC-CMs. (E-F) Scatter graphs to compare APD₅₀ and APD₉₀ between vehicle- and 3 μM sunitinib-treated iPSC-CMs by unpaired two-tailed Student’s t-test. *p < 0.05 and **p < 0.01. n = 15–17 cells. (G) Bubble chart of KEGG enrichment clarifying the differential pathway between vehicle- and 3 μM sunitinib‐treated iPSC-CMs. (H) Western blot analysis of the protein expression of phosphorylated and total CAMK2, phosphorylated and total PLN, and total ATP2A2a in vehicle-, 3 μM sunitinib- and 10 μM sunitinib-treated iPSC-CMs. (I-K) Bar graphs to compare the protein expression of phosphorylated and total CAMK2, phosphorylated and total PLN, and total ATP2A2a among three different groups in H by one-way ANOVA (Tukey method). **p < 0.01, ***p < 0.001 and ****p < 0.0001. n = 4 biologically independent experiments. (L-M) Representative Ca²⁺ transient waveforms recorded from vehicle- and 3 μM sunitinib-treated iPSC-CMs. Red arrows indicate irregular Ca²⁺ transient events. Ca²⁺ imaging was conducted using Fluo-4 AM. (N) Bar graph to compare irregular Ca²⁺ transient rate between vehicle- and 3 μM sunitinib-treated iPSC-CMs by Fisher’s test. *p < 0.05. n = 58–70 cells. (O-P) Scatter graphs to compare transient duration 90 and decay 90 between vehicle- and 3 μM sunitinib-treated iPSC-CMs by unpaired two-tailed Student’s t-test. ****p < 0.0001. n = 42–52 cells. (Q) Representative Ca²⁺ transient tracings following stimulation with 10 mM caffeine in iPSC-CMs treated with vehicle or 3 μM sunitinib. (R) Scatter graph to compare the SR Ca²⁺ load between different groups by unpaired two-tailed Student’s t-test. ****p < 0.0001. n = 120–153 cells.

Figure 3.

Excessive autophagic degradation of MYLK3 in sunitinib-treated iPSC-CMs. (A) Representative confocal images of mCherry‐GFP‐LC3 expressed in vehicle- and 3 μM sunitinib-treated iPSC-CMs. Red fluorescence (mCherry⁺ GFP^‐) indicates autolysosomes whereas yellow fluorescence (mCherry⁺ GFP⁺) indicates autophagosomes. Scale bar: 10 μm. (B) Bar graph to compare the percentage of mCherry⁺ GFP^‐ puncta in vehicle- and 3 μM sunitinib-treated iPSC-CMs by unpaired two-tailed Student’s t-test. ****p < 0.0001. n = 9–12 cells in 3 independent experiments. (C) Western blot analysis of the protein expression of LC3, ATG5 and ATG7 in vehicle-, 3 μM sunitinib-, and 10 μM sunitinib-treated iPSC-CMs. (D-F) Bar graphs to compare the LC3-II, ATG5 and ATG7 expression among three different groups in C by one-way ANOVA (Tukey method). *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001. n = 4 biologically independent experiments. (G) Schematic of the analysis of SQSTM1 interactome via MS. (H) Venn diagram of SQSTM1 interactome in vehicle-, 3 μM sunitinib-, and 10 μM sunitinib-treated iPSC-CMs. (I) Co-IP analysis using an antibody against SQSTM1 in vehicle-, 3 μM sunitinib-, and 10 μM sunitinib-treated iPSC-CMs, followed by western blot analysis to detect the protein expression of SQSTM1 and MYLK3. (J) Western blot analysis of the protein expression of SQSTM1, MYLK3 and MYL2v (phosphorylated [phospho-] and nonphosphorylated [nonphospho-]) in vehicle-, 3 μM sunitinib-, and 10 μM sunitinib-treated iPSC-CMs. Red arrow indicates phospho-MYL2v. (K) Bar graph to compare the SQSTM1 expression among three different groups in J by one-way ANOVA (Tukey method). **p < 0.01 and ***p < 0.001. n = 4 biologically independent experiments. (L) Bar graph to compare the MYLK3 expression among three different groups in J by one-way ANOVA (Tukey method). *p < 0.05 and ***p < 0.001. n = 4 biologically independent experiments. (M) Bar graph to compare the phosphorylated MYL2v expression among three different groups in J by one-way ANOVA (Tukey method). ****p < 0.0001. n = 4 biologically independent experiments. (N) Bar graph to compare the mRNA expression of MYLK3 between vehicle- and 3 μM sunitinib‐treated iPSC-CMs. *p < 0.05. n = 3 biologically independent experiments.

Figure 4.

Rescuing SIC phenotypes by inhibition of early-stage autophagy in iPSC-CMs. (A) Representative confocal images of mCherry‐GFP‐LC3 expressed in iPSC-CMs treated with vehicle, 3 μM sunitinib alone, or 3 μM sunitinib with 1 μM MRT68921. Scale bar: 10 μm. (B) Bar graph to compare the percentage of mCherry⁺GFP^‐ puncta among three different groups in a by one-way ANOVA (Tukey method). ****p < 0.0001. n = 9–18 cells in 3 independent experiments. (C) Bar graph to compare the cell viability among iPSC-CMs treated with vehicle, 3 μM sunitinib alone, or 3 μM sunitinib with 1 μM MRT68921 by one-way ANOVA (Tukey method). **p < 0.01 and ****p < 0.0001. n = 5 biologically independent experiments. (D) Sarcomeric arrange analysis in iPSC-CMs treated with vehicle, 3 μM sunitinib alone, or 3 μM sunitinib with 1 μM MRT68921. Sarcomeres were marked using TNNT2 (green) and ACTN (red). Scale bar: 10 μm. (E) Western blot analysis of the protein expression of total MYLK3, total ATP2A2a, phosphorylated and total CAMK2, and phosphorylated and total PLN in iPSC-CMs treated with vehicle, 3 μM sunitinib alone, or 3 μM sunitinib with 1 μM MRT68921. (F-I) Bar graphs to compare the protein expression of total MYLK3, total ATP2A2a, phosphorylated CAMK2 (phosphorylated:total), and phosphorylated PLN (phosphorylated:total) among three different groups in E by one-way ANOVA (Tukey method). *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001. n = 4 biologically independent experiments. (J) Representative action potential waveforms recorded from iPSC-CMs treated with vehicle, 3 μM sunitinib alone, or 3 μM sunitinib with 1 μM MRT68921. Red arrows indicate arrhythmic events with DADs. (K) Bar graph to compare arrhythmic rate among four different groups in J by Fisher’s test. *p < 0.05. n = 15–54. (L) Representative Ca²⁺ transient waveforms recorded from iPSC-CMs treated with vehicle, 3 μM sunitinib alone, or 3 μM sunitinib with 1 μM MRT68921. Red arrows indicate abnormal Ca²⁺ transient events. Ca²⁺ imaging was conducted using Fluo-4 AM. (M) Bar graph to compare irregular Ca²⁺ transient rate among four different groups in L by Fisher’s test. *p < 0.05 and ***p < 0.001. n = 55–108 cells.

Figure 5.

Overexpression of MYLK3 is sufficient to rescue SIC in iPSC-CMs. (A) Validation of the transfection efficiency. Representative confocal images showing the transfection efficiency in iPSC-CMs overexpressed vector-GFP treated with or without 3 μM sunitinib and iPSC-CMs overexpressed MYLK3-GFP treated with 3 μM sunitinib. Scale bar: 100 μm. (B) Bar graph to compare the cell viability among iPSC-CMs overexpressed vector-GFP treated with or without 3 μM sunitinib and iPSC-CMs overexpressed MYLK3-GFP treated with 3 μM sunitinib by one-way ANOVA (Tukey method). **p < 0.01 and ***p < 0.001. n = 3 biologically independent experiments. (C) Representative confocal images showing TUNEL staining in iPSC-CMs overexpressed vector-GFP treated with or without 3 μM sunitinib and iPSC-CMs overexpressed MYLK3-GFP treated with 3 μM sunitinib. Scale bar: 100 μm. (D) Bar graph to compare the ratio of TUNEL:DAPI among different groups in C by one-way ANOVA (Tukey method). ****p < 0.0001. n = 10 fields of vision. (E) Western blot analysis of the protein expression of MYLK3, cleaved CASP3 and CASP3, phosphorylated and total CAMK2, and phosphorylated and total PLN in iPSC-CMs overexpressed vector-GFP treated with or without 3 μM sunitinib and iPSC-CMs overexpressed MYLK3-GFP treated with 3 μM sunitinib. (F-I) Bar graphs to compare the protein expression of MYLK3, cleaved CASP3 (cleaved:total), CAMK2 (phosphorylated:total), and PLN (phosphorylated:total) among three different groups in E by one-way ANOVA (Tukey method). *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001. n = 4 biologically independent experiments. (J) Representative Ca²⁺ transient waveforms recorded from vehicle and 3 μM sunitinib-treated iPSC-CMs overexpressed vector-GFP or MYLK3-GFP, respectively. Ca²⁺ imaging was conducted using Fura-2 AM and field-stimulated at 1 hz. (K-L) Scatter graphs to compare transient duration 90 and decay 90 among four different groups by one-way ANOVA (Tukey method). **p < 0.01, ***p < 0.001 and ****p < 0.0001. n = 74–100 cells. (M) Representative action potential waveforms recorded from vehicle and 3 μM sunitinib-treated iPSC-CMs overexpressed vector-GFP or MYLK3-GFP, respectively. Red arrows indicate arrhythmic events with DADs. (N) Bar graph to compare arrhythmic event rate in four different groups in M by Fisher’s test. **p < 0.01, ***p < 0.001 and ****p < 0.0001. n = 19–35 cells. (O-P) Scatter graphs to compare APD₅₀ and APD₉₀ among four different groups in M by one-way ANOVA (Tukey method). *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001. n = 13–29 cells.

Figure 6.

Cardiac myosin activator omecamtiv mecarbil can rescue SIC in iPSC-CMs. (A) Western blot analysis of the protein expression of MYLK3, SQSTM1, LC3, cleaved CASP3 and CASP3, phosphorylated and total CAMK2, and phosphorylated and total PLN in iPSC-CMs treated with vehicle, 200 nM omecamtiv mecarbil (OM) alone, 3 μM sunitinib alone, or 3 μM sunitinib with 200 nM OM. (B) Co-IP analysis using an antibody against SQSTM1 in iPSC-CMs treated with vehicle, 200 nM OM alone, 3 μM sunitinib alone, or 3 μM sunitinib with 200 nM OM, following by Western blot analysis to detect the protein expression of SQSTM1 and MYLK3. (C-H) Bar graphs to compare the protein expression of MYLK3, SQSTM1, LC3-II (LC3-II:GAPDH), cleaved CASP3 (cleaved:total), CAMK2 (phosphorylated:total) and PLN (phosphorylated:total) among four different groups in a by one-way ANOVA (Tukey method). *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001. n = 4 biologically independent experiments. (I) Representative action potential waveforms recorded from iPSC-CMs treated with vehicle, 3 μM sunitinib alone, and 3 μM sunitinib with 200 nM OM. Red arrows indicate arrhythmic events with DADs. (J) Bar graph to compare arrhythmic event rate in three different groups in I by Fisher’s test. *p < 0.05 and **p < 0.01. n = 28–56 cells. (K) Representative single action potential waveforms recorded from iPSC-CMs treated with vehicle, 3 μM sunitinib alone, and 3 μM sunitinib with 200 nM OM. (L-M) Scatter graphs to compare APD₅₀ and APD₉₀ among three different groups in K by one-way ANOVA (Tukey method). *p < 0.05, ***p < 0.001 and ****p < 0.0001. n = 15–39 cells. (N) Representative Ca²⁺ transient waveforms recorded from iPSC-CMs treated with vehicle, 3 μM sunitinib alone, and 3 μM sunitinib with 200 nM OM. Red arrows indicate irregular Ca²⁺ transient events. Ca²⁺ imaging was conducted using Fluo-4 AM. (O) Bar graph to compare irregular Ca²⁺ transient rate among three different groups in N by Fisher’s test. *p < 0.05, ****p < 0.0001. n = 51–163 cells. (P-Q) Scatter graphs to compare transient duration 90 and decay 90 among three different groups by one-way ANOVA (Tukey method). *p < 0.05, ***p < 0.001 and ****p < 0.0001. n = 38–108 cells. (R) Representative Ca²⁺ transient tracings following stimulation with 10 mM caffeine in iPSC-CMs treated with vehicle, 3 μM sunitinib alone, and 3 μM sunitinib with 200 nM OM. S. Scatter graph to compare the SR Ca²⁺ load among three different groups by one-way ANOVA (Tukey method). n = 28–41 cells. *p < 0.05 and ***p < 0.001.

Figure 7.

Nanoparticle delivery of OM prevents SIC in mice. (A) Schematic of the in vivo experiment workflow to test the protective effect of PCM-PEG2000-DSPE nanoparticles carrying OM in sunitinib-treated mice. (B) Representative images of gross heart morphology from 12-week-old mice treated with vehicle, OM (0.5 mg/kg) alone, sunitinib (40 mg/kg) alone, or sunitinib (40 mg/kg) with OM (0.5 mg/kg). (C) Scatter graph to compare heart weight to body weight (HW:BW) ratio among four different groups in B by one-way ANOVA (Tukey method). *p < 0.05. n = 3 mice. (D) Echocardiographic analysis of mice treated with vehicle, OM alone, sunitinib alone, or sunitinib with OM. (E-H) Scatter graphs to compare key echocardiography parameters among four different groups in D by one-way ANOVA (Tukey method), including ejection fraction (EF), fractional shortening (FS), left ventricular end systolic diameter (LVIDs), and left ventricular end diastolic diameter (LVIDd). *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001. n = 4 mice. (I) Representative images of heart sections stained with H&E from mice treated with vehicle, OM alone, sunitinib alone, or sunitinib with OM. Scale bars: 50 µm. (J) Representative Ca²⁺ transient waveforms recorded from cardiomyocytes isolated from mice treated with vehicle, OM alone, sunitinib alone, or sunitinib with OM. Ca²⁺ imaging was conducted using Fura-2 AM and field-stimulated at 2 hz. (K-N) Scatter graphs to compare Ca²⁺ transient amplitude, time to peak, decay 50, and decay 90 among four different groups in J by one-way ANOVA (Tukey method). *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001. n = 33–45 cells in 4 mice. (O) Proposed work model. Sunitinib causes a variety of deleterious phenotypes including reduced cell viability, apoptosis, sarcomeric disorganization, arrhythmia, irregular Ca²⁺ handling and excessive autophagy. The expression of SQSTM1-binding MYLK3 is downregulated by excessive autophagic degradation in sunitinib-treated iPSC-CMs, which suppresses the phosphorylation of CAMK2, thereby leading to reduced phosphorylation of PLN and impaired ATP2A2a activity. Inhibition of the CAMK2-PLN-ATP2A2a signaling pathway causes dysfunction of Ca²⁺ homeostasis, and subsequent DAD arrhythmia. Early-stage autophagic inhibitor or cardiac myosin activator OM can restore MYLK3 expression and reverse sunitinib-induced pathogenic phenotypes.