FAT10 protects against ischemia-induced ventricular arrhythmia by decreasing Nedd4-2/Nav1.5 complex formation

The human leukocyte antigen F-associated transcript 10 (FAT10) is a member of the small ubiquitin-like protein family that binds to its target proteins and subjects them to degradation by the ubiquitin–proteasome system (UPS). In the heart, FAT10 plays a cardioprotective role and affects predisposition to cardiac arrhythmias after myocardial ischemia (MI). However, whether and how FAT10 influences cardiac arrhythmias is unknown. We investigated the role of FAT10 in regulating the sodium channel Nav1.5, a major regulator of cardiac arrhythmias. Fat10 was conditionally deleted in cardiac myocytes using Myh6-Cre and Fat10F/F mice (cFat10−/−). Compared with their wild-type littermates, cFat10−/− mice showed prolonged RR, PR, and corrected QT (QTc) intervals, were more likely to develop ventricular arrhythmia, and had increased mortality after MI. Patch-clamp studies showed that the peak Na+ current was reduced, and the late Na+ current was significantly augmented, resulting in a decreased action potential amplitude and delayed depolarization. Immunoblot and immunofluorescence analyses showed that the expression of the membrane protein Nav1.5 was decreased. Coimmunoprecipitation experiments demonstrated that FAT10 stabilized Nav1.5 expression by antagonizing Nav1.5 ubiquitination and degradation. Specifically, FAT10 bound to the lysine residues in the C-terminal fragments of Nav1.5 and decreased the binding of Nav1.5 to the Nedd4-2 protein, a ubiquitin E3 ligase, preventing degradation of the Nav1.5 protein. Collectively, our findings showed that deletion of the Fat10 in cardiac myocytes led to increased cardiac arrhythmias and increased mortality after MI. Thus, FAT10 protects against ischemia-induced ventricular arrhythmia by binding to Nav1.5 and preventing its Neddylation and degradation by the UPS after MI.


Introduction
Electrical remodeling of cardiac ion channels after myocardial ischemia (MI) plays a vital role in the development of ischemic arrhythmia, which is a major cause of sudden cardiac death 1 . Several cardiac ion channels, such as K + , Ca 2+ , and Na + channels, participate in susceptibility to cardiac arrhythmias after MI. The main isoform of the voltage-gated Na + channel Nav1.5 (a.k.a. Scn5a) encoded by the Scn5a gene is responsible for the generation of the Na + current (I Na ) and is the critical channel for maintaining normal cardiac rhythm 2 . Nav1.5 remodeling provides a critical substrate for the generation of reentrant ventricular arrhythmias in border zones of the infarcted heart 3,4 . Post-translational modification of Nav1.5 is critical for regulating cardiac Na + channels 5,6 . Nav1.5 is known to undergo Neddylation by Nedd4-2 (also known as Nedd4L), a ubiquitin E3 ligase, and subsequent degradation by the ubiquitin-proteasome system (UPS). The UPS plays a key role in regulating cardiac function in the ischemic heart 7,8 .
Ubiquitin-like proteins (UBLs) are a family of small proteins with structural and functional similarity to ubiquitin 9 and are known to target ion channels. For example, a small ubiquitin-like modifier (SUMO) targets Kv4.2, thus affecting its biophysical properties and surface expression 10 . It also targets the current density of Cav1.2, which has been linked to sudden death and cardiomyopathy 11,12 . The human leukocyte antigen F-associated transcript 10 (FAT10) is a UBL that binds to proteins and tags them for proteasomal degradation. In addition to protein degradation, FAT10 is also involved in cell cycle regulation, immune response, and cell apoptosis 13 . We have shown that FAT10 antagonizes the ubiquitination of specific substrates and prevents their degradation 14 .
According to our findings, FAT10 protects cardiomyocytes against apoptosis in response to ischemic injury 15,16 . In the present study, we investigated the role of FAT10 in protecting against ischemic ventricular arrhythmia by binding to Nav1.5 and preventing its Neddylation and degradation by the UPS.

Results
The frequency and mortality of ventricular arrhythmia are increased in ischemic cFat10 −/− mice The FAT10 gene was deleted specifically in cardiac myocytes (cFat10 −/− ) upon insertion of a floxed allele encompassing exon 2 of the Fat10 gene by homologous recombination and subsequent deletion of the floxed exon using myosin heavy-chain 6-Cre (Myh6-Cre) mice (Fig. 1A). Successful removal of the Fat10 gene was confirmed by western blot (WB) and immunofluorescence (IF) analyses ( Fig. 1B and Supplementary Fig. S1A). The cFat10 −/− mice displayed normal cardiac size and function, as assessed by echocardiography and histopathologic analysis, compared to those of their age-and sex-matched wild-type littermates (Fat10 fl/fl ) (Fig. 1C, Supplementary  Fig. S1B, and Supplementary Table S1). Likewise, myocardial fibrosis was not increased in the cFat10 −/− mice, as assessed by Masson's trichrome and Sirius red staining ( Supplementary Fig. S1C-E).
Cardiac rhythm was assessed in conscious mice by telemetric cardiac rhythm monitoring, and no spontaneous arrhythmias were observed in the cFat10 −/− or Fat10 fl/fl mice. However, the PR (Fig. 1D), corrected QT (QTc) (Fig. 1E), RR (Fig. 1F), and QT intervals ( Supplementary  Fig. S1F) were significantly prolonged and the heart rates were significantly decreased ( Supplementary Fig. S1G) in the cFat10 −/− mice compared to Fat10 fl/fl mice.
Nav1.5 expression in both the membrane and cytoplasm is reduced in ventricular myocytes isolated from cFat10 −/− mice Analysis of the proteomics data from three cFat10 −/− and control mouse hearts showed that membrane Nav1.5 expression was significantly reduced in the cFat10 −/− heart ( Fig. 2A), which was confirmed by WB (Fig. 2B, C). However, there was no significant difference in the expression levels of the Nav1.5 trafficking proteins SAP97 and MOG1 ( Supplementary Fig. S2A) or in the Scn5a transcript levels between cFat10 −/− and Fat10 fl/fl mice ( Supplementary Fig. S2B), while the Nav1.5 protein levels were reduced in both the cytoplasmic subfractions and the whole-heart protein extracts ( Fig. 2D-G). Unlike Nav1.5, the expression of Connexin-43, Cav1.2, and Na + /K + ATPase was not changed in the hearts of cFat10 −/− mice, as detected by immunohistochemistry and WB ( Fig. 2D and Supplementary Fig. S2C, D). Consistent with these findings, IF staining of isolated cardiac myocytes demonstrated reduced expression and membrane localization of Nav1.5 as determined by Pearson's correlation coefficient in cardiac myocytes isolated from the cFat10 −/− mice (Fig. 2H, I). These results suggested that decreased Nav1.5 expression is a putative mechanism for the increased incidence of ventricular arrhythmia after MI in cFat10 −/− mice.
The electrophysiological function of Nav1.5 is reduced in cFat10 −/− cardiomyocytes To determine the functional effect of reduced Nav1.5 levels, the whole-cell patch-clamp technique was used to record the I Na currents. The peak I Na density was significantly reduced in the cFat10 −/− ventricular myocytes (Fig. 3A, B) (−38.7 ± 2.9 vs. −29.4 ± 2.0 pA/pF, p = 0.01), while changes in steady-state activation and inactivation were not observed ( Supplementary Fig. S3A, B). However, the late I Na (I Na,L ) current was significantly increased in the cFat10 −/− myocytes (p = 0.02) (Fig. 3C, D).
To further corroborate these findings, a heterologous expression system was used to generate Fat10 knockout HEK293 (Fat10 −/− ) cells by transfection with CRISPR-Cas9 plasmids 14 . Compared to that in wild-type (Fat10 +/+ ) cells, the peak I Na current was moderately but significantly decreased in Fat10 −/− cells (−125.1 ± 12.5 vs. −81.6 ± 6.9 pA/pF; p = 0.003) (Fig. 3E, F). However, I Na,L was significantly increased in the  Supplementary Fig.  S4). These AP changes in the cFat10 −/− cardiomyocytes were consistent with the ECG) phenotype (prolonged QTc and RR intervals), and these findings collectively suggested that Fat10 deletion prolongs APD because of decreased I Na and Na + channel activity.

FAT10 stabilizes Nav1.5 expression in cardiomyocytes by inhibiting its ubiquitination
To determine the mechanisms by which FAT10 deficiency affects Nav1.5 function, endogenous physical interactions of Nav1.5 and FAT10 were detected by coimmunoprecipitation (Co-IP) in neonatal rat cardiomyocytes (NRCMs) and adult cardiac tissue extracts ( Fig. 4A and Supplementary Fig. S5A). The interaction was confirmed in HEK293 cells transfected with Flagtagged Ssc5a and HA-tagged Fat10 plasmid DNAs (Fig.  4B). Confocal microscopy showed colocalization of FAT10 and Nav1.5 in adult mouse myocytes and NRCMs (Fig. 4C). In NRCMs under normoxic and hypoxic conditions, overexpression of FAT10 was associated with increased levels of Nav1.5 protein expression (Fig. 4D, E), whereas downregulation of FAT10 had the opposite effect ( Fig. 4F, G), and no changes in the mRNA levels were observed ( Supplementary Fig. S5C). Interestingly, FAT10 protein expression was increased by hypoxia (Supplementary Fig. S5D, E). Moreover, Nav1.5 expression was also upregulated by endogenous FAT10 overexpression (induced by interferon-γ/tumor necrosis factor-α (IFN-γ/ TNF-α) 18 ) in NRCMs ( Supplementary Fig. S5F). FAT10 affects the expression of its substrates by regulating the level of their ubiquitination 14 . Given that Nav1.5 is known to be degraded by ubiquitination, we analyzed the regulation of Nav1.5 by FAT10 through ubiquitination. Therefore, the Nav1.5 protein levels were assessed after treatment with cycloheximide (CHX) in the absence or presence of the MG132 proteasome inhibitor. As shown in Fig. 5A (upper), treatment with CHX significantly decreased the endogenous Nav1.5 protein expression in NRCMs in a time-dependent manner, which was nulled by the presence of MG132 (Fig. 5A, lower). Thus, these results indicate that the Nav1.5 protein undergoes degradation via the UPS pathway.
In a complementary set of studies, the effects of FAT10 on the expression of Nav1.5 in NRCMs were shown to be negated by inhibition of the UPS with MG132 (Supplementary Fig. 5B, C). Regardless of MG132 treatment, knockdown or overexpression of FAT10 in NRCMs was associated with an increased or reduced degradation rate of the Nav1.5 protein, respectively (Fig. 5D, E). In vitro experiments showed that the Nav1.5 ubiquitination levels were significantly increased in cFat10 −/− cardiomyocytes (Fig. 5F). We then performed an in vitro ubiquitination assay in HEK293 cells upon overexpression of Nav1.5 and FAT10, revealing that FAT10 overexpression decreased the levels of ubiquitinated Nav1.5 (Fig. 5G). Furthermore, treatment with IFN-γ/TNF-α was associated with reduced binding of Nav1.5 to ubiquitin in Fat10 +/+ cells, but not in Fat10 −/− cells (Fig. 5H). In addition, in the Fat10 −/− cells, the binding of Nav1.5 to ubiquitin was rescued upon expression of HA-Fat10 using plasmids, but not upon treatment with IFN-γ/TNF-α (Fig. 5I). Overall, these results show that FAT10 stabilizes Nav1.5 expression in cardiomyocytes by inhibiting ubiquitination. To identify the Nav1.5 domain that binds to the FAT10 protein, nine fragments of the Scn5a (Nav1.5) gene were generated, as depicted in Fig. 6A, expressed in HEK293 cells and analyzed for their ability to bind FAT10. Co-IP studies showed that the C-terminal fragment (Scn5a-C), comprising residues 1978-2016, contained the FAT10 interaction domain (Fig. 6B). Therefore, a glutathione Stransferase (GST) pull-down assay performed using a GST-Nav1.5-C fusion protein (GST-Nav1.5) showed a direct interaction with FAT10 (Fig. S6A). Because FAT10 modifies its substrates by binding to the lysine sites on substrates 14 , a plasmid construct expressing a mutant Scn5a-C-terminal (GST-Scn5a-Mut) was constructed by mutating nine lysine residues to arginine. Interestingly, the introduction of the mutations abolished the binding of FAT10 to Nav1.5 in the Co-IP and GST pull-down assays (Fig. 6C, D). These findings suggested that FAT10 affects Nav1.5 expression by directly interacting with the lysine of Scn5a-C.

Discussion
FAT10 is a relatively newly discovered member of the UBL family. We previously reported that FAT10 protects cardiomyocytes against apoptosis, which is required to maintain cardiac function post-7 day MI 15,16 . In the present study, we identified an important role of FAT10 in ischemia-induced ventricular arrhythmia. We showed that FAT10 bound to the C-terminal fragments of Nav1.5, stabilizing Nav1.5 expression upon prevention of its Neddylation and degradation, ultimately increasing the I Na current and AP in cardiomyocytes.
Recently, several ion channels, including kv7.1 and Nav1.2, have been shown to be modified by UBLs via processes such as SUMOylation 20 . The function of Nav1.5 is critical for cardiac excitability and electrical conduction 21 , and ubiquitination of Nav1.5 is an important mediator of Nav1.5 trafficking, internalization, and degradation 19,22 . However, little is known about whether Nav1.5 is modulated by UBLs. The findings of the present study provide the first evidence of Nav1.5 modification by FAT10 in cardiomyocytes. Ubiquitination and degradation of Nav1.5 were first reported in 2004 19 , and several ubiquitin E3 ligases have been implicated in the regulation of Nav1.5 ubiquitination 19,23 . For example, Nedd4-2 has been shown to regulate Nav1.5 ubiquitination by the PY motif in the C-terminal region, suggesting an important role of the C-terminal region in Nav1.5 ubiquitination.
Consistent with previous studies, we showed that FAT10 regulated ubiquitination by interacting with the C-terminal domain of Nav1.5, exploiting lysine residues, as changing these residues abolished the interaction of Nav1.5 with FAT10. The findings illustrate that FAT10 prevents Neddylation and UPS-mediated degradation of Nav1.5.
Interestingly, we found that FAT10 expression was increased, while that of Nav1.5 was decreased under quantitative assessment of the protein (right) levels of Nav1.5 in NRCMs treated with cycloheximide (CHX, 20 μM) or MG132 (15 μM) for the indicated times. *p < 0.05 and **p < 0.01 vs. treatment in the 0 h group. B, C Representative immunoblots or quantitative data of Nav1.5 expression in NRCMs after adenoviral knockdown (B) or overexpression (C) of Fat10. Cells were treated with or without MG132 (*p < 0.05). D Western blotting was performed to detect the expression of FAT10 and Nav1.5 in NRCMs with or without FAT10 knockdown (*p < 0.05) treated with CHX. E Western blotting was performed to detect the expression of FAT10 and Nav1.5 in NRCMs with or without FAT10 overexpression treated with CHX (*p < 0.05). F Representative immunoblots (left) and quantitative data (right) of Nav1.5 ubiquitination levels after immunoprecipitation of Nav1.5 from cultured cardiomyocytes isolated after MG132 treatment for 8 h (*p < 0.05). G Representative immunoblots of the Nav1.5 ubiquitination levels in HEK293 cells transfected with the indicated plasmid DNA. Cells in each group were treated with MG132. H Representative immunoblots of the Nav1.5 ubiquitination levels in Fat10 +/+ and Fat10 −/− HEK293 cells treated with or without TNF-α/IFN-γ. I Representative immunoblots of the Nav1.5 ubiquitination levels in the indicated groups of Fat10 −/− HEK293 cells. All results are expressed as the mean ± SEM of independent experiments. hypoxic conditions, and several factors may underlie this result. First, the degradation of Nav1.5 is regulated by complex molecular mechanisms, such as ubiquitination 5 and autophagy 24 . In the present study, FAT10 overexpression stabilized the Nav1.5 protein by regulating its ubiquitination. The FAT10 protein is not unique in regards to the regulation of Nav1.5 ubiquitination, as several proteins have been demonstrated to participate in this process, such as UBC9 25 , UBR3, and UBR6 26 . Furthermore, under ischemia or hypoxia, the UBS and autophagic activities were increased 7,27 , which was potentially responsible for the decreased Nav1.5 protein expression 5,24 However, the degradation of Nav1.5 could partially but not totally be rescued by the increased FAT10 expression under the conduction of ischemia by antagonizing Nav1.5 ubiquitination, demonstrating a cardioprotective effect of FAT10. These results might explain the increased FAT10 expression and decreased Nav1.5 expression under ischemic/hypoxia conditions. Reduced expression of Nav1.5 decreases peak I Na , which results in progressive cardiac conduction defects with the widening of the QT interval 21 . In the present study, Fat10 knockout was associated with reduced sodium channel activity because of decreased Nav1.5 levels, while steadystate activation and inactivation were not affected. The C terminus of Nav1.5 is lysine-rich and is one of the most conserved regions of the protein. Mutations or deletions involving lysine in the C-terminal domain may lead to abnormal Nav1.5 function and severe human arrhythmias. Relevant to these findings, a deletion mutation affecting Fig. 6 FAT10 antagonizes Nav1.5 ubiquitination by competing with Nedd4-2. A Schematic diagram of Nav1.5 deletions and mutants. B HEK293 cells were transfected with the indicated constructs and then lysed for the Co-IP assay using anti-Flag beads to detect HA-FAT10 binding. C Representative immunoblots of FAT10 binding with Nav1.5 in HEK293 cells co-transfected with HA-Fat10 and Flag-Scn5a wild-type or the Flag-Scn5a mutation plasmid as detected by Co-IP. D GST pull-down experiments were performed to assess the association between FAT10 and Nav1.5. E Competitive binding of Nav1.5 was analyzed in a GST pull-down experiment. F, G Fat10 +/+ HEK293 cells were transfected with Si-Nedd4-2 or Si-scramble and treated with IFN-γ/TNF-α for the indicated times. Co-IP was performed to detect Nedd4-2-Nav1.5 and FAT10-Nav1.5 complex formation. H, I Fat10 −/− HEK293 cells were transfected with Si-Nedd4-2/Si-scramble and the Flag-Fat10 plasmid. Co-IP was performed to detect Nedd4-2-Nav1.5 and FAT10-Nav1.5 complex formation. HEK293 cells were treated with IFN-γ/TNF-α to induce FAT10 expression. All results are expressed as the mean ± SEM of independent experiments. lysine-1872 (located in the C terminus) has been reported in a large family with Brugada syndrome, a disease manifesting as decreased peak I Na

28
. The findings demonstrated that the C terminus might play an important role in regulating I Na through the UPS pathway. Previous studies support the involvement of the C-terminal (Ser-Ile-Val [SIV]) domain of Nav1.5 in UPS-mediated degradation 29 . Another notable finding was increased I Na,L in the cFat10 −/− cardiomyocytes, which may be attributed to reduced expression of the caveolin-3 (Cav3) protein, as reported previously 15 . Decreased Cav3 expression has been shown to increase the S-nitrosylation of Nav1.5, resulting in increased I Na,L in cardiomyocytes 30 .
The findings of prolonged PR, RR, and QT intervals on ECG suggested slow conduction in the cFat10 −/− hearts, which may stem from altered tissue architecture, cell-tocell coupling, and/or cardiomyocyte excitability [31][32][33] . In the present study, Nav1.5 expression was reduced, as were I Na , AP upstroke velocity (dV/dtmax, a measure of sodium channel availability), and AP amplitude, which might explain the slow conduction velocity 34,35 . Moreover, the QTc interval was also prolonged in the cFat10 −/− hearts, suggesting abnormal repolarization, which may be partly attributed to the increased late current in the model. I NaL has been linked to the manifestation of inherited and acquired cardiac diseases, including long QT variant 3 syndrome and heart failure 36,37 .
The cFat10 −/mice exhibited abnormal I Na and ECG parameters, but did not show ventricular arrhythmias at baseline and developed ventricular arrhythmia after MI. Changes in I Na observed at baseline in the cFat10 −/− hearts were modest and likely insufficient to provoke spontaneous ventricular arrhythmias. Thus, additional electrical and/or structural perturbations, such as ischemic events, were needed to unmask ventricular arrhythmias. The secondary changes included alterations in peak I Na and I NaL as well as reduced conduction velocity, prolonged ventricular AP, and promotion of early afterdepolarizations, which could lead to a predisposition to the reentrant circuit and trigger activity 33 .
Our findings on the regulatory effects of FAT10 on sodium channels might shed light on the role of UBLs in ischemia-induced arrhythmia, providing additional insight into the complex mechanisms underlying arrhythmogenesis during MI. In addition, FAT10 affects the expression of Cav3, a critical scaffolding protein involved in the genesis of t-tubules and the t-tubular Ca 2+ current. We speculate that FAT10 may have similar modulatory effects on cardiac electrical activity in other pathophysiological conditions, such as heart failure and hypertrophic cardiomyopathy, which needs to be further investigated.
Several limitations should be considered in this study. First, because our experiments were confined to young adult mice, the findings might not pertain to older mice, which may exhibit arrhythmias secondary to age-associated changes, such as fibrosis 38 . Second, other Ca 2+ or K + channel proteins were not detected in the tandem mass tag-mass spectrometry proteomics experiments, which might be due to the limitation of the methods. Therefore, the potential role of Ca 2+ and K + currents, which are also crucial in AP repolarization, was not investigated. However, WB showed that Cav1.2 expression was not significantly changed in cFat10 −/− cells, which might exclude the effect of Ca 2+ channels. Third, the protective role of FAT10 in ischemic arrhythmia was not tested in a cardiac FAT10 overexpression mouse model. Finally, the study was mainly focused on the susceptibility of cFat10 −/− mice to arrhythmias under ischemia. However, modulatory effects on cardiac electrical activity in other pathophysiological conditions (e.g., aging and hypertrophic cardiomyopathy) remain unexplored.
In conclusion, these findings provide the first evidence of a protective role of FAT10 in ischemia-induced ventricular arrhythmia. Deletion of the Fat10 gene was associated with abnormal APD because of a dysregulated sodium current as a mechanism for ischemia-induced ventricular arrhythmia. In addition, the removal of Fat10 was related to the reduced Nav1.5/FAT10 binding and increased Neddylation of Nav1.5, resulting in its subsequent degradation reduced expression levels and thereby explaining the molecular mechanisms underlying the abnormal sodium current. Our findings highlight the significance of FAT10 in regulating ventricular arrhythmia during MI.

Materials and methods
Cardiac Fat10 knockout mouse model and generation of CRISPR-Cas9 knockout HEK293 cell lines Cardiac conditional knockout of the Fat10 gene in C57B/6L mice was based on homologous recombination technology with help from the Model Animal Research Center of Nanjing University. Briefly, when the Fat10 gene is located between two loxP sites, the Cre enzyme is expressed, loxp can be deleted, and a frameshift mutation occurs. Subsequent knockdown of Fat10 was further confirmed by WB and DNA-sequencing. Fat10 deletion was induced by tamoxifen 6 weeks after the mice were born. The generation of CRISPR-Cas9 knockout HEK293 cell lines was previously reported 14 .
In brief, HEK293 cells were transfected with CRISPR-Cas9 plasmids for 2 days and then treated with 1 µg/ml puromycin to enrich for transfected cells. Then, the viable clones were picked up, and Fat10 knockout HEK293 cells were confirmed by both WB and DNAsequencing.
Animal and ischemic animal model Newborn (6-10 g, 2 days) male Sprague-Dawley (SD) rats and C57/B6L mice (20-25 g, 12-16 weeks) were used in this study. All animal experiments and procedures were approved by the Animal Ethics Committee of Nanchang University and performed in accordance with the Guide for the Care and Use of Laboratory Animals.
As previously described 15 , adult C57/B6L mice were randomly used to create an MI model by ligating the left anterior descending coronary artery of the heart.

Cell lines and cell culture
NRCMs were isolated from the ventricular myocardia of newborn SD rats through an enzymatic digestion method according to previous studies 15,39 . Cultured cardiomyocytes were subjected to hypoxic injury by incubation in a CO 2 /N 2 /O 2 (94:5:1 ratio) environment at 37°C for 8 h.
HEK293 cells were purchased from American Type Culture Collection (Rockville, USA) and cultured in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum. Cells were cultured at 37°C and 5% CO 2 incubator in a humidified atmosphere.

Cellular electrophysiology
The APs were recorded by whole-cell patch-clamp experiments in the current-clamp mode as previously described 39 . For recording, the isolated ventricular myocytes were kept at room temperature and perfused for 5 min with Tyrode's solution 39 . The APs were stimulated with 5 ms, 900 mV square-wave pulses of 2 Hz pacing frequencies. All patch-clamp data were recorded using an EPC-10 patch-clamp amplifier (HEKA Electronik, Lambrecht, Germany).
The conventional whole-cell patch-clamp technique was used to record the Na + current from HEK293 cells and rod-shaped, quiescent singly isolated mouse ventricular myocytes. The sodium current of HEK293 cells was recorded using a previously described method and standard solutions 40 .

Liquid chromatography with tandem mass spectrometry (LC-MS/MS) analyses
LC-MS/MS analyses were performed as previously described 41 . Briefly, proteins were extracted from three cFat10 −/− and control mouse hearts and analyzed by labelfree tandem mass tag and LC-MS/MS. The obtained samples were then analyzed by LC-MS/MS with the help of Shanghai Applied Protein Technology Co., Ltd., who provided technological assistance. The MS data were analyzed using MaxQuant software version 1.3.05 (Max Planck Institute of Biochemistry in Martinsried, Germany).

Statistical analysis
Experiments were independently repeated a minimum of three times. Continuous data are presented as the mean ± standard error of the mean. Unpaired Student's t tests (equal variance) or unpaired Student's t tests with Welch's correction (unequal variance) were used for comparisons between two independent groups, and one-way analysis of variance with Tukey's post hoc test was used for multiple comparisons. Categorical data are expressed as percentages and were compared between groups using Fisher's exact test. All analyses were performed using the GraphPad Prism 5.0 software. Statistical significance was defined as p < 0.05.