Abstract Background Dilated cardiomyopathy (DCM) constitutes a major cause of heart failure, characterized by high mortality rates and a limited availability of effective therapeutic options. A substantial body of evidence indicates that mutations in the Nexilin (NEXN) gene are significant pathogenic contributors to DCM, but the pathogenic mechanism for dilated cardiomyopathy is unclear. Methods A human NEXN homozygous knockout cardiomyocyte model was established by combining CRISPR/Cas9 gene editing technology and human induced pluripotent stem cells (hiPSCs)-directed differentiation technology. Cell model phenotypic assays were done to characterize the pathological features of the resulting NEXN-deficient cardiomyocytes. Results NEXN gene knockout did not affect the pluripotency and differentiation efficiency of hiPSCs. NEXN-deficient cardiomyocytes showed disordered junctional membrane complexes, abnormal excitation-contraction coupling, increased oxidative stress and decreased energy metabolism level. Moreover, levo-carnitine and sarcoplasmic reticulum calcium ATPase (SERCA2a) Activator 1 were identified as promising therapeutic agents for the treatment of DCM. Conclusion We demonstrated that NEXN was one of the important components in maintaining the structure and function of cardiomyocyte junctional membrane complexes (JMCs), excitation-contraction coupling and energy metabolism of cardiomyocytes, while the loss of its function would lead to DCM. This model represents an important tool to gain insight into the mechanism of DCM, elucidate the gene-phenotype relationship of NEXN deficiency and facilitate drug screening. Supplementary Information The online version contains supplementary material available at 10.1186/s13287-025-04484-2. Keywords: Human cardiomyocyte, Dilated cardiomyopathy, Junctional membrane complexes, Oxidative stress, Drug discovery Background Dilated cardiomyopathy (DCM) is one of the common causes of heart failure and is the most prevalent cardiomyopathy in children [[40]1]. The incidence of DCM has been reported to be 5–7 patients per 100,000 populations per year [[41]2]; it is characterized by dilatation of the left ventricle followed by systolic dysfunction. Embryonic and childhood development of DCM will lead to severe adverse prognoses. There is no treatment for the disease except heart transplants, which have very high mortality rates in children [[42]1–[43]4]. The etiology of DCM is varied, and genetic mutations can be detected in about 35% of the patients [[44]3]. There is no clear pathogenic mechanism for dilated cardiomyopathy, and elucidating its molecular mechanism will provide new ideas and a strong basis for diagnosis and treatment. Nexilin (NEXN) is an actin filament-binding protein coded by the NEXN gene located at chromosome 1p31.1. It is primarily expressed in cardiomyocytes and comprises 675 amino acids [[45]5]. NEXN is recognized as an indispensable protein to preserve the integrity of the Z-discs in muscle fibers. It interacts with proteins residing in the Z-discs of cardiac muscle, playing a pivotal role in securing the attachment of filaments from adjacent sarcomeres [[46]6]. Evidence-based researches show that mutations in human. NEXN are intimately associated with DCM. Heterozygous deletion of the NEXN gene can lead to dilated cardiomyopathy in infants or adults [[47]7, [48]8]. In mouse models, NEXN deletion is demonstrated to cause T-tubule disorganization, contributing to lethal cardiomyopathy [[49]9, [50]10]. These findings show that the NEXN loss of function plays a significant role in DCM, however, the underlying mechanisms are not well understood. Although genetically edited mouse models for NEXN deficiency have been established, there are notable different phenotypes between species. A heterozygous mutation of the NEXN gene causes progressive DCM in humans, whereas shows no signs of cardiac disease in mice [[51]8, [52]9]. This suggests a necessity for the development of more precise human cardiac models to better understand the pathogenic mechanisms associated with NEXN deficiency in humans. Compared to the challenge in acquisition and in vitro cultivation of primary human cardiomyocytes, human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs) have the advantages of sufficient sources and can be cultured in vitro for a long time [[53]11]. They offer a unique platform that captures the genetic diversity and phenotypic manifestations, which can model cardiomyopathies with a high degree of fidelity, thereby advancing the frontiers of personalized medicine [[54]12]. Here, we established a NEXN-deficient human cardiac myocyte model to elucidate the mechanism of DCM caused by NEXN deletion. We first showed that NEXN was crucial for maintaining the structure and function of cardiomyocytes JMCs, and played a key regulatory role in maintaining the normal excitation-contraction coupling of cardiomyocytes. NEXN loss contributed to abnormal energy metabolism and ultimately caused cardiomyocyte death. Moreover, we identified that Levo-carnitine (LC) and SERCA2a Activator 1 (SA1) were two therapeutic agents that demonstrably enhance energy metabolism. Thus, our study reveals mechanism of NEXN deficiency in leading to DCM and provides new clues for diagnosis and treatment of DCM. Methods HiPSC culture and cardiac differentiation The hiPSCs-U2 cell line obtained from Cellapy, China, was cultured in 6-well plates coated with Matrigel from Corning, USA. The culture medium was refreshed daily until reaching 80–90% confluence, at which point it was switched to a myocardial differentiation medium from Cellapy, China. The hiPSCs were differentiated into hiPSC-CMs using a small molecule-based method as previously described. The chemically defined medium consists of three components: the basal medium RPMI 1640, L-ascorbic acid 2-phosphate, and rice derived recombinant human albumin. Small molecule CHIR99021 is used to activate Wnt signaling, and Wnt-C59 is employed to inhibit Wnt signaling [[55]13]. On the 10th day of differentiation, spontaneously beating cells were observed. Genome editing The sgRNA targeting NEXN (5′- AGAAGGCGUUUGCUGAAGCA − 3′) was designed using an online tool ([56]https://design.synthego.com) and introduced into hiPSCs using electroporation with the epiCRISPR vector (Addgene, 135,960), which expressed SpCas9 and was a gift from Dr. Yongming Wang. Electroporation was performed using the 4D Nuclear Receptor System (Lonza, Germany) and the CA137 program. Following transfection, cells were re-seeded into 6-well plates and subjected to drug screening (puromycin) once reaching 40% confluence to select for homozygous transfected strains. Immunofluorescent staining Cells were cultured on Matrigel-coated 20 mm coverslips, fixed with 4% paraformaldehyde for 15 min, permeabilized with 0.2% Triton X-100 (Sigma, USA), and blocked with 3% bovine serum albumin (Sigma, USA) for 30 min. Primary antibody incubation occurred at 4 °C overnight, followed by secondary antibody incubation (Invitrogen, USA) at 37 °C for 1 h in the absence of light. The cells were then washed with phosphate-buffered saline and counterstained with DAPI (4,6-diamino-2-phenylindole) in fluorescence-blocking tablets. Imaging was performed using a confocal microscope (Leica DMI 4000B, Germany). Antibodies and their working dilutions are listed in Additional file 2: Table [57]S1. Western blotting Cells were subjected to trypsin digestion, centrifugation, and lysis using a protein extraction reagent (Thermo Fisher, USA) supplemented with a combination of phosphatase and protease inhibitors (Thermo Fisher, USA). The entire cleavage procedure was conducted on ice, with samples agitated every 10 min using an oscillator, followed by centrifugation at 12,000 rpm for 15 min to collect the supernatant as the protein sample. Protein concentration was determined using the BCA method. Equal amounts of protein were loaded onto a polyacrylamide gel for gel electrophoresis, electrophoresed, and subsequently transferred to a polyvinylidene fluoride (PVDF) membrane. The membranes were obstructed with a 5% solution of skim milk powder for 1 h at a temperature of 37℃, followed by incubation with the primary and secondary antibodies as detailed in Additional file 2: Table [58]S1. Flow cytometry The cardiomyocytes that were identified underwent treatment with CardioEasy Human Cardiomyocyte Digestive Fluid (Cellapy, China) to create a single-cell suspension. Following a PBS wash, the cells were fixed (excluding those used for viable cell assays), stained with a variety of antibodies, re-filtered through a 300-mesh filter, and promptly analyzed using FACS (Beckman, USA). Each group contained no fewer than 1 million cells for examination. The outcomes were assessed using the FlowJo program. RNA extraction and quantitative real‑time PCR Total RNA was isolated from cells utilizing TRIZOL reagent (Invitrogen, USA) and subsequently reverse-transcribed into cDNA employing the Prime Script Reverse Transcription system (Takara, Japan). Quantitative RT-PCR was performed utilizing SYBR Green II (Takara, Japan) and the iQ5 instrument (Bio-Rad, Hercules, CA). A comparative CT method was employed to assess the relative alterations in gene expression, with the outcomes normalized to the expression of β-actin (internal control). Primer sequences are listed in Additional file 2: Table [59]S2. RNA sequencing (RNA‑Seq) assay The mRNA library was constructed using purified cardiomyocytes on the 45th day of differentiation. Total RNA was isolated from the cardiomyocytes and treated with RNase H and DNase I for purification. Subsequently, cDNA synthesis was performed, followed by purification of the double-stranded DNA using magnetic beads and amplification via PCR. The resulting PCR products were denatured and converted into single-stranded circular DNA (ssCir DNA), which was further amplified to generate the final library. The qualified libraries were then subjected to analysis using the BGISEQ500 platform, utilizing single-ended 50-base reads (SE50). High-quality reads were obtained by removing reads with low quality, contamination, or high levels of unknown bases. These clean reads were subsequently aligned to reference genome sequences using Bowtie2 and HISAT2. Differentially expressed genes (DEGs) meeting the criteria of|log2 Fold change|≥1 and adjusted P value (FDR, Q-value) < 0.001 were identified as significant. Gene Ontology and KEGG pathway enrichment analysis of these DEGs were conducted using the Phyper under the Hypergeometric test ([60]https://en.wikipedia.org/wiki/Hypergeometric_distribution). The Q-value of terms or pathways less than 0.05 was significantly enriched. Myocardial contractility The cells were cultured in 6-well plates (Corning, USA) coated with Matrigel (Corning, USA). The video recordings were obtained using a Leica DMI 4000B microscope. Videos capturing cardiomyocyte contractions were recorded for 3–5 s, saved in the original.czi format, and subsequently converted to an uncompressed.avi format at a frame rate of 70 frames per second (fps). The software plugin MUSCLEMOTION was integrated into ImageJ to analyze the video data [[61]14]. Contraction amplitude is defined as the maximum developed impedance generated by cardiomyocyte during contraction, Peak time refers to the time interval measured from the onset of cardiomyocyte contraction to the attainment of maximum contraction intensity, and relaxation time is defined as the time interval measured from the waveform peak to the restoration of baseline. Ca^2+ imaging Cardiomyocytes were cultured and treated with Fluo-4 (Beyotime, China) before being subjected to confocal microscopy (Leica, USA) to capture spontaneous calcium transient signals at a rapid frame rate. Data analysis was performed using ImageJ and IGOR Pro software. Microelectrode array (MEA) HiPSC-CMs were enzymatically dissociated using CardioEasy Human Cardiomyocyte Digestive Fluid (Cellapy, China) and subsequently seeded at a density of 2 × 10^4 cells on a microelectrode array (MEA) coated with 5% Matrigel (Corning, USA). The following day, 300 µl of medium was added to each well. Upon the resumption of spontaneous beating by the hiPSC-CMs, experimental data were acquired using a Maestro EDGE system (Axion Biosystems, Inc., Atlanta, USA) by the MEA manual. Subsequent data analysis was performed using the Cardiac Analysis Tool, AxIS Navigator, AxIS data export tool, and Origin software. MEA was used to detect beat period; field potential duration (FPD), the time from the depolarization spike to the peak of the T-wave; action potential duration at 90% repolarization (APD 90), the time interval from the onset of the action potential upstroke (depolarization phase) to the point where the membrane potential has repolarized to 90% of its peak amplitude relative to the resting membrane potential; excitation-contraction coupling delay (ECCD), the time between the depolarization spike and the beginning of the contraction; standard deviation of ECCD (ECCD SD). Transmission electron microscopy The cardiomyocytes that were identified underwent treatment with CardioEasy Human Cardiomyocyte Digestive Fluid (Cellapy, China) before being centrifuged and fixed with a solution containing 0.10% glutaraldehyde and 4% paraformaldehyde. Following dehydration, the specimens were processed and subsequently analyzed using transmission electron microscopy (Thermo, USA) for image capture. Co-Immunoprecipitation The myocardial cells were digested and collected, followed by the addition of an appropriate amount of cell lysis solution containing protease inhibitors for cracking. PMSF was added in proportion to the mixture, which was then centrifuged to collect the supernatant. The collected supernatant was incubated with HA agarose gel beads, boiled, and combined with SDS-PAGE for Western blot detection. Seahorse assay The cells were seeded into a cell culture plate and the Seahorse XFe24 Flux Assay Kit was used. Then, the Agilent Seahorse XFe24 system was used to detect the respiratory and reserve functions of the cells [[62]15]. Detection of cellular ATP content A volume of 200 µl of lysate was dispensed into individual wells of a 6-well plate, followed by the collection of supernatant after cell lysis. The cells were then examined using an ATP detection kit (Beyotime, China), and the relative light unit values were measured with a luminometer or liquid scintillation meter to measure the ATP content in the cells. The ATP concentration within the sample was determined through the utilization of standard curves and the enumeration of cell quantities. Data analysis and statistics The findings were presented as the mean ± standard deviation, and statistical analysis was conducted using GraphPad Prism 8.0.1 for Windows. A two-sided unpaired Student’s t-test was utilized for comparing two groups with a normal distribution, while a one-way analysis of variance (ANOVA) was employed for comparing three or more groups. Before conducting the t-test and ANOVA, all assumptions of normality and homogeneity of variance were met. Statistical significance was defined as a p-value less than 0.05 (*P < 0.05, **P < 0.01, ***P < 0.001; ns, not significant). Results Modeling NEXN-deficient hiPSC-CMs We generated a NEXN-deficient hiPSC model using the CRISPR/Cas9 system. Specific sgRNA targeting the NEXN gene was designed and incorporated into plasmids containing Cas9 elements. These constructs were introduced into hiPSCs (WT hiPSCs) via electroporation, followed by selection with puromycin. Sanger sequencing confirmed the genotypes of surviving clones, demonstrating successful homozygous knockout of NEXN (NEXN^−/− hiPSCs, KO hiPSCs) (Additional file 1: Figure [63]S1A). NEXN^−/− hiPSCs exhibited normal clonogenic growth and maintained expression of stem cell markers OCT4 and SSEA4 (Fig. [64]1A). Karyotype analysis confirmed a stable chromosome number in NEXN^−/− hiPSCs (Additional file 1: Figure [65]S1B), and teratoma assays validated their trilineage differentiation potential (Additional file 1: Figure [66]S1C). Western blotting confirmed the absence of NEXN protein expression in NEXN^−/− hiPSCs (Fig. [67]1B, full-length blots of the WB result are presented in Additional file 1: Figure [68]S2). To exclude mutations at important sites of genes, STR detection was performed on WT hiPSCs and KO hiPSCs. The detection results indicated that KO-hiPSCs and WT-hiPSCs were derived from the same individual at 24 key sites and did not have mutations (Additional file 1: Figure [69]S1D). Fig. 1. [70]Fig. 1 [71]Open in a new tab Characterization of NEXN^−/− hiPSC-CMs. (A) Immunofluorescence of SSEA4 (Alexa Fluor 488, green), OCT4 (Alexa Fluor 568, red) and DAPI (blue) in WT and NEXN^−/− hiPSCs. Scale bar, 75 μm. (B) Western blot image of NEXN in WT and NEXN^−/− hiPSCs. And quantification of NEXN protein expression normalized to β-actin (n = 3). Full-length blots are presented in Supplementary Figure [72]S2. (C) Flow cytometry analysis and bar graph summarizing the percentage of TNNT2-positive cells in WT as well as NEXN^−/− hiPSC-CMs (n = 3). (D) Immunofluorescence of phalloidin (Alexa Fluor 488, green) and DAPI (blue) in WT and NEXN^−/− hiPSC-CMs. Scale bar, 25 μm. (E) Quantitative analysis of green fluorescence area in (D) (n = 6). (F) Cluster analysis of differentially expressed genes between WT and NEXN^−/− hiPSC-CMs (n = 3). (G) Bubble chart of KEGG enrichment terms associated with the differentially expressed genes between WT and NEXN^−/− hiPSC-CMs. Data is represented as means ± SD. *P < 0.05, ***P < 0.001, ns, not significant. DAPI, 4′,6-diamidino-2-phenylindole WT hiPSCs and NEXN^−/− hiPSCs were successfully differentiated into hiPSC-CMs (WT) and NEXN^−/− hiPSC-CMs (KO) and this differentiation protocol could efficiently generate TNNT2^+ cardiomyocytes. Flow cytometry was used to detect the positive rate of TNNT2 on two types of cardiomyocytes, revealing that NEXN deficiency did not affect cardiomyocyte differentiation (Fig. [73]1C). NEXN^-/- hiPSC-CMs demonstrated a hypertrophic phenotype In dilated cardiomyopathy, cardiomyocytes experience cardiac hypertrophy and myocardial remodeling, which serve as the primary contributors to the development of heart failure [[74]16]. To investigate the effects of NEXN loss on cardiomyocytes, we first measured the cell surface area of both groups of cardiomyocytes. We cultured the two groups of cardiomyocytes for 40 days respectively, and then stained the cytoskeleton of the two groups of cardiomyocytes with phalloidin, and then observed through confocal microscopy, and found that the surface area of cardiomyocytes was significantly expanded after NEXN knockout, showing dilated cardiomyopathy-like hypertrophic myocardium (Fig. [75]1D, E). Q-PCR analysis was performed on indicators associated with cardiac hypertrophy and fibrosis in two groups of cardiomyocytes. The results indicated a significant increase in mRNA levels of ANP and BNP (Additional file 1: Figure [76]S3A, B). Additionally, the ratio of MYH7/MYH6 was found to be significantly elevated (Additional file 1: Figure [77]S3C), suggesting a progression towards myocardial hypertrophy in NEXN^−/− hiPSC-CMs. Proper myofilament sarcomere organization is essential for maintaining the normal contractile function of cardiomyocytes. A significant proportion of the etiological factors contributing to dilated cardiomyopathy involve aberrant myofilament myotome structure in cardiomyocytes [[78]17]. Given the findings from a previous study indicating a substantial increase in cardiomyocyte volume following NEXN knockout, resulting in a morphology resembling dilated cardiomyopathy [[79]10], further investigation is warranted to ascertain whether these morphological alterations are attributable to abnormal myofilament myotome structure in cardiomyocytes. Immunofluorescence staining was employed to simultaneously label myocytes from two distinct groups with TNNT2 and α-actinin. The findings revealed that the myofilaments of myocytes in both groups exhibited a well-organized and evenly distributed pattern (Additional file 1: Figure [80]S4A), suggesting that the deletion of NEXN did not induce alterations in the myofilament structure of myocytes. Transmission electron microscopy was utilized to examine the myofilament myotome structure in two groups of cardiomyocytes at the subcellular level. The findings indicated that the myofibrillar fibers in both groups were organized in a parallel manner, devoid of any disarray. Additionally, there was no notable disparity in the length of intercalated discs (Additional file 1: Figure [81]S4B, C), aligning with the outcomes of immunofluorescence analysis. NEXN-deficient induced gene expression changes in cardiomyocytes To explore the molecular mechanism of the phenotype of dilated cardiomyopathy in cardiomyocytes after NEXN knockout, we detect and analyze the enrichment changes in gene expression profiles of the two groups of cardiomyocytes through RNA sequencing (RNA-seq). Analysis of the differential genes of the two groups of cardiomyocytes showed that compared with the WT group, the expression of 2656 genes in KO cardiomyocytes was significantly different, of which 546 genes were significantly up-regulated and 2110 genes were significantly down-regulated (Fig. [82]1F). To delve deeper into the distinctions in gene expression between WT and KO, KEGG analysis was performed to elucidate the variations in pathway expression (Fig. [83]1G). The findings indicated that KO exhibited significant enrichment of cardiomyopathy-related pathways, including Diabetic cardiomyopathy, Hypertrophic cardiomyopathy, and Dilated cardiomyopathy, compared to WT. Besides, pathways associated with cardiac remodeling, such as ECM-receptor interaction and Focal adhesion, were also significantly enriched in the KO group. The findings were in line with the gene expression patterns associated with myocardial cell remodeling identified in prior studies. The results indicated a significant enrichment of pathways related to cardiac muscle contraction in KO cardiomyocytes. Furthermore, our analysis revealed significant upregulation of reactive oxygen species (ROS) level, reduced oxidative phosphorylation, and dysregulated Citrate cycle (TCA cycle) in KO cardiomyocytes. These results indicated that the knockout of NEXN may impact the energy metabolism of cardiomyocytes. The contractile function of NEXN^-/- cardiomyocytes was abnormal The contractile function of cardiomyocytes serves as a crucial indicator of their normal function [[84]18]. In this study, high-resolution digital imaging analysis technology was employed to assess the contractile force levels of cardiomyocytes in two groups at 40 days (Fig. [85]2A, B, Additional file 1: Figure [86]S4D). This technique is capable of simultaneously assessing the contractile function of a population of cells, providing a more accurate representation of the contractile force variations in cardiomyocytes within the population. This method has undergone extensive validation and implementation in various studies. The results showed that compared with WT hiPSC-CMs, the contraction intensity of KO hiPSC-CMs decreased significantly, and the peak time and relaxation time were significantly extended (Fig. [87]2C, D, E). In addition, we also found that the contraction frequency of KO cardiomyocytes decreased significantly, suggesting that the heart rate of KO cardiomyocytes decreased. All these results indicate that NEXN deficiency can cause the dysfunction of myocardial contractile function. Fig. 2. [88]Fig. 2 [89]Open in a new tab Knockout of NEXN impacted the contractile function of cardiomyocytes. (A, B) The representative contraction waveform of WT and NEXN^-/- hiPSC-CMs. (C-E) Respective quantification of contraction amplitude, peak time and relaxation time (n = 10). (F-J) Respective quantification of beat period, field potential duration, 90% repolarization time, excitation-contraction coupling delay time and excitation-contraction coupling delay time dispersion degree, measured by MEA technology (n = 10). Data is represented as means ± SD. **P < 0.01, ***P < 0.001 NEXN-deficient resulted in increased ECCD The proper contractile function of the heart is primarily reliant on the coordinated contraction and relaxation of the myocardium. Excitation-contraction coupling (ECC) serves as the mechanism that links the depolarization of cardiomyocytes, marked by alterations in membrane potential, with the contraction process facilitated by the sliding of myofilaments [[90]19]. This process is essential for enabling cardiomyocytes to effectively carry out their contractile function. The period from the initiation of cardiomyocyte excitation signals to the onset of cardiomyocyte contraction is referred to as the excitation-contraction coupling delay (ECCD). In healthy cardiomyocytes, the ECCD is expected to remain consistent at a specific value; however, in cases of excitation-contraction decoupling, the ECCD may become prolonged and variable [[91]20]. MEA technology was utilized to examine the Beat Period (BP), field Potential duration (FPD), and 90% repolarization time (APD 90) of WT and KO hiPSC-CMs. Moreover, excitation-contraction coupling delay time (ECCD mean) and excitation-contraction coupling delay time dispersion degree (ECCD SD) were measured (Additional file 1: Figure [92]S4E). Quantitative statistical analysis revealed that KO hiPSC-CMs exhibited a significantly prolonged beating cycle and irregular heart rate compared to WT hiPSC-CMs (Fig. [93]2F). The FPD and APD 90 of KO cardiomyocytes exhibited a more than two-fold increase compared to WT, accompanied by an evident dispersion distribution (Fig. [94]2G, H). Furthermore, the ECCD of WT remained stable at approximately 190 ms. In contrast, the ECCD of KO was prolonged to over 400 ms (Fig. [95]2I), leading to a significant increase in ECCD dispersion among KO cardiomyocytes (Fig. [96]2J). These results suggested that after NEXN knockout, the excitation-contraction decoupling appeared, which led to myocardial systolic dysfunction. The JMC structure was disturbed after the NEXN knockout The maturity of cardiomyocytes formed by hiPSCs-induced differentiation was not sufficient, the proximity between the myofilament and sarcoplasmic reticulum of these cardiomyocytes was consistently maintained at approximately 12 nm [[97]21]. Utilizing transmission electron microscopy, we compared WT and KO cardiomyocytes (Fig. [98]3A). Quantitative analysis revealed a significant increase in the distance between myofilament-sarcoplasmic reticulum connections in the KO group. Fig. 3. [99]Fig. 3 [100]Open in a new tab Knockout of NEXN disturbed JMC structure. (A) Transmission electron microscopy images of WT and NEXN^-/- hiPSC-CMs. And the proximity between the myofilament and sarcoplasmic reticulum of these cardiomyocytes was quantificated. Scale bar, 100 nm. (n = 3) (B) Co-immunoprecipitation assay was performed to show the interaction between NEXN and several JMCS-related structural proteins. Full-length blots are presented in Supplementary Figure [101]S5. (C) Immunofluorescence of JPH2 (Alexa Fluor 488, green), NEXN (Alexa Fluor 568, red) and DAPI (blue) in WT and NEXN^-/- hiPSC-CMs. Scale bar, 10 μm. (D) Representative Ca transient signals in WT and NEXN^-/- hiPSC-CMs measured using Fluo-4 AM. (E) Waveform diagram of calcium transient. (F, G) Quantification of peak value of calcium transient and calcium decay time in (D) (n = 4 in F and n = 6 in G). (H) Ca^2+ imaging of WT and NEXN^-/- hiPSC-CMs treated with caffeine. (I, J) Quantification of peak value of calcium transient and recovery time of 90% calcium in (H) (n = 4 in I and n = 3 in J). Data is represented as means ± SD. ***P < 0.001, ns, not significant The interaction between NEXN and JMCS-related structural proteins was detected by co-immunoprecipitation and immunofluorescence staining. The results suggested that NEXN interacts with JPH2 and Actin, but does not interact with CACNA2D1 (Fig. [102]3B, full-length blots of the WB result are presented in Additional file 1: Figure [103]S5). In normal cardiomyocytes, NEXN and JPH2 were colocalized under the cell membrane. After NEXN knockout, the expression of JPH2 in cardiomyocytes was significantly reduced, and the distribution of JPH2 in the cells was dysregulated and decreased (Fig. [104]3C). Western blotting detected the expression of JMC structure-related proteins in the KO group. In NEXN^-/- hiPSC-CMs, JPH2 decreased, CACNA1C expression increased, and SERCA2a expression decreased significantly in myocytes (Additional file 1: Figure [105]S4F, G). NEXN^-/- hiPSC-CMs showed abnormal Ca^2+ transport After incubation with Fluo-4 AM, the two groups were monitored and recorded by laser confocal microscopy (Fig. [106]3D, E). In comparison to normal cardiomyocytes, NEXN^-/- hiPSC-CMs exhibited a significant reduction in the peak amplitude of calcium transient (Fig. [107]3F). Furthermore, the calcium decay time was nearly three times longer in the KO group (Fig. [108]3G). The caffeine assay measured the total amount of calcium in the sarcoplasmic reticulum of the two groups (Fig. [109]3H). The results showed that the peak calcium release in KO cardiomyocytes did not change significantly compared with WT, but the recovery time of 90% calcium was greatly extended (Fig. [110]3I, J). NEXN-deficient changed ROS levels in cardiomyocytes The RNA sequencing analysis revealed that KO cardiomyocytes exhibited increased expression of genes and pathways associated with oxidative stress and fatty acid metabolism in comparison to WT cardiomyocytes, indicating that the loss of NEXN may impact the mitochondrial function of cardiomyocytes. The ROS levels of WT and KO cardiomyocytes were assessed with DHE, revealing a significant elevation in ROS levels in KO cardiomyocytes (Fig. [111]4A, B). The majority of ROS generated in the mitochondria. Consequently, MitoSOX Red was utilized to measure ROS levels within the mitochondria of both WT and KO cardiomyocytes. The results revealed a significant increase in mitochondrial ROS levels in KO (Fig. [112]4C, D), suggesting a higher level of mitochondrial oxidative stress in KO cardiomyocytes. The membrane potential levels of the two groups were assessed using JC-1 fluorescent probe. Analysis revealed a significant decrease in red fluorescence and a corresponding increase in green fluorescence in KO cardiomyocytes (Fig. [113]4E, F), suggesting a significant reduction in mitochondrial membrane potential following NEXN knockout. Fig. 4. [114]Fig. 4 [115]Open in a new tab Knockout of NEXN increased ROS levels in cardiomyocytes. (A) Fluorescence images of total ROS in WT and NEXN^-/- hiPSC-CMs stained with DHE. Scale bar, 50 μm. (B) Quantification of mean fluorescence intensity in (A) (n = 4). (C) Fluorescence images of mitochondrial ROS in WT and NEXN^-/- hiPSC-CMs stained with MitoSOX Red. Scale bar, 50 μm. (D) Quantification of mean fluorescence intensity in (C) (n = 3). (E) Fluorescence images of mitochondrial membrane potential in WT and NEXN^-/- hiPSC-CMs stained with JC-1. Scale bar, 50 μm. (F) Quantification of the mean fluorescence intensity ratio between Texas Red and FITC (n = 3). (G, H) The OCR and ECAR curves of WT and NEXN^-/- hiPSC-CMs were detected by Seahorse assay (n = 5). (I-K) Quantification of basal, maximal respiration levels and ATP production rate in WT and NEXN^-/- hiPSC-CMs (n = 15). Data is represented as means ± SD. *P < 0.05, **P < 0.01 and ***P < 0.001 NEXN-deficient caused mitochondrial dysfunction In normal cardiomyocytes, ATP is primarily generated through oxidative phosphorylation in mitochondria [[116]22]. Seahorse XF analysis was employed to assess the oxidative phosphorylation levels in WT and KO cardiomyocytes. After detection, we could see that the oxygen consumption rate of KO cardiomyocytes was significantly reduced (Fig. [117]4G), suggesting that the level of oxidative phosphorylation was decreased; At the same time, the extracellular acidification rate of KO cardiomyocytes was also significantly lower than that of WT cardiomyocytes, suggesting that the glycolytic capacity of KO cardiomyocytes was significantly reduced (Fig. [118]4H). The curve of the final detection results showed that basal respiration, maximal respiration, and ATP production of KO cardiomyocytes were significantly lower than those of WT cardiomyocytes (Fig. [119]4I, J, K). The analysis of RNA-seq results suggested that the enrichment of fatty acid metabolism related pathways and oxidative phosphorylation related pathways in NEXN^-/- hiPSC-CMs was changed, combined with the previous experiment confirming that NEXN knockout would lead to mitochondrial function damage in cardiomyocytes. Therefore, we next tested the changes in energy metabolism in NEXN knockout cardiomyocytes. Q-PCR was used to detect the fatty acid metabolism in WT and KO cardiomyocytes, the results hint that the gene expression of fatty acid metabolism proteins decreased in NEXN^-/- hiPSC-CMs (Fig. [120]5A). Fig. 5. [121]Fig. 5 [122]Open in a new tab Knockout of NEXN induced mitochondrial dysfunction. (A) The mRNA levels of fatty acid metabolism related genes in WT and NEXN^-/- hiPSC-CMs were detected by Q-PCR assay (n = 3). (B) Transmission electron microscopy images of mitochondria in WT and NEXN^-/- hiPSC-CMs. Scale bar, 200 nm. (C) The expression levels of mtDNA related genes ND1 and ND2 in WT and NEXN^-/- hiPSC-CMs were detected by Q-PCR assay (n = 3). (D) Fluorescence images of mitochondrial calcium content in WT and NEXN^-/- hiPSC-CMs stained with Rhod-2. Scale bar, 50 μm. (E) Quantification of mean fluorescence intensity in (D) (n = 3). (F) Representative Western blot image of proteins involved in the mitochondrial apoptosis pathway. Full-length blots are presented in Supplementary Figure [123]S6. (G) Quantification of grayscale value in (F) (n = 3). Data is represented as means ± SD. *P < 0.05, ***P < 0.001 and ns, not significant Transmission electron microscopy was utilized to visually analyze the number and morphology of mitochondria in cardiomyocytes from two groups. Following the observation, it was determined that the volume of mitochondria was typically enlarged and the mitochondrial crista was relatively reduced in KO cardiomyocytes (Fig. [124]5B). The two groups of cells were tested for mitochondrial DNA content (ND1 and ND2), Q-PCR detection showed the amount of mtDNA was equal in both groups of cells (Fig. [125]5C). Then, we used Rhod-2 to detect the mitochondrial calcium content of the two groups, and the NEXN^-/- hiPSC-CMs showed obvious mitochondrial calcium overload (Fig. [126]5D, E). The mitochondria emerged as the primary pro-apoptotic target of elevated levels of ROS [[127]23], with the findings from RNA sequencing indicating enrichment of gene expression related to the PI3K-AKT pathway in NEXN^-/- hiPSC-CMs. The expression levels of proteins involved in the mitochondrial apoptosis pathway in WT and KO cardiomyocytes were detected by Western blotting (Fig. [128]5F, G, full-length blots of the WB result are presented in Additional file 1: Figure [129]S6). After NEXN knockout, the overall level of apoptosis in cardiomyocytes is increased, and it is caused by the mitochondrial apoptosis pathway. The NEXN^-/- hiPSC-CMs model was used to screen the therapeutic drugs for DCM As evidenced by the aforementioned findings, NEXN knockout led to hiPSC-CMs energy metabolism disorder and abnormal Ca^2+ transport. Therefore, we treated NEXN^-/- hiPSC-CMs with metabolism-enhancing agent levo-carnitine (LC) and SERCA2a activator 1 (SA1) respectively [[130]24, [131]25], evaluating the efficacy of LC and SA1 in NEXN deficiency-induced DCM. Following initial experiments, we investigated the drug concentration gradient of two therapeutic drugs and ultimately established the optimal drug concentrations as LC 200 µmol/L and SA1 10 µmol/L. HiPSC-CMs were obtained for analysis after a 2-hour drug administration (Additional file 1: Figure [132]S7A, B). We constructed a full-length NEXN lentiviral expression vector and overexpressed (OE) NEXN in KO hiPSC-CMs at day 40 post-differentiation, to restore its protein expression level. There was no significant difference in transfection efficiency among the three virus gradients (MOI = 10, 15 or 20) (Additional file 1: Figure [133]S7C, D). In the subsequent experiments, we selected 10 MOI for cell transfection, and the NEXN OE group was set as a positive control. The transmission electron microscope was used to examine the Spacing between the sarcolemma and sarcoplasmic reticulum in various treatment groups. The proportion of normal JMCs significantly increased in the OE group. However, there was no significant improvement in groups KO + LC or KO + SA1 (Fig. [134]6A). Similarly, overexpression of NEXN in KO hiPSC-CMs increased peak value of calcium transient and shortened calcium decay time, while treatment with LC or SA1 showed no significant improvement in calcium transport (Fig. [135]6B, C). Mitochondrial ROS staining assay indicated that overexpression of NEXN, treatment with LC or SA1 all lowered mtROS level (Fig. [136]6D). Subsequently, MEA was used to detect the ECCD of cardiomyocytes in each group. The ECCD time and diffusion of OE group cardiomyocytes were comparable to those of WT cardiomyocytes, while after drug treatment, only the SA1 treatment group had significant recovery of ECCD time and diffusion (Fig. [137]6E, F). To explore the effects of LC and SA1 on the recovery of energy metabolism in KO cardiomyocytes, the Seahorse assay was performed in each group. The result showed that the energy metabolism level of cardiomyocytes was significantly increased in OE, LC + KO and SA1 + KO group (Additional file 1: Figure [138]S7E, Fig. [139]6G-I). These results suggested that both metabolism-enhancing agent levo-carnitine and SERCA2a activator 1 can effectively improve the energy metabolism of NEXN^-/- hiPSC-CMs. Fig. 6. [140]Fig. 6 [141]Open in a new tab NEXN was a promising therapeutic target for dilated cardiomyopathy. (A) Respective transmission electron microscopy images of ultra microstructure in WT, NEXN^-/- (KO), NEXN^-/- hiPSC-CMs overexpressed NEXN (OE), NEXN^-/- hiPSC-CMs treated with LC (KO + LC) and NEXN^-/- hiPSC-CMs treated with SA1 (KO + SA1). Scale bar, 100 nm. (B, C) Respective quantification of peak value of calcium transient and calcium decay time (n = 3). (D) HiPSC-CMs were stained with MitoSOX Red, the mean fluorescence intensity was quantitatively analyzed (n = 3). (E, F) Respective quantification of ECCD time and diffusion in each group of cardiomyocytes with MEA technology (n = 10). (G-I) Quantification of basal, maximal respiration levels and ATP production rate in each group of cardiomyocytes with Seahorse assay (n = 9). Data is represented as means ± SD. *P < 0.05, **P < 0.01, ***P < 0.001, ns, not significant Discussion In this study, we discovered the mechanism of NEXN deficiency in cardiomyocytes leading to DCM. NEXN knockout disrupted the structural stability of JMCs in hiPSC-CMs, leading to downregulation of calcium-handling protein within JMCs. This dysregulation impaired intracellular calcium transport, ultimately resulting in compromised contractile function of hiPSC-CMs. In studies associated with NEXN deficiency-induced dilated cardiomyopathy, all foundational mechanistic research has been confined to animal models [[142]9, [143]26], which failed to fully recapitulate the pathophysiological context in humans. Therefore, we employed hiPSC-CMs to elucidate the molecular mechanisms underlying NEXN deficiency-mediated DCM. Previous study has revealed structural abnormalities in JMCs of cardiomyocytes resulting from NEXN depletion [[144]9], while our study conducted comparative RNA-seq between WT and KO cardiomyocytes. The analysis demonstrated most significant enrichment changes in mitochondrial function-related pathways, oxidative stress response, and tricarboxylic acid cycle-associated genes following NEXN deletion (Fig. [145]1G). These findings suggested that NEXN knockout may induce myocardial energy metabolism dysfunction, representing a novel mechanistic insight first identified in our investigation. In terms of cellular morphology, our observations revealed spontaneous cardiomyocyte hypertrophy in NEXN-deficient cells, which may be attributed to aberrant Ca²⁺ handling (Fig. [146]3D-J) and energy metabolism dysregulation (Figs. [147]4G, H, and [148]5A) in NEXN KO hiPSC-CMs. This finding contrasted with the conventional view that cardiomyocyte hypertrophy is a secondary response to pressure overload. Previous studies reported Z-disc disorganization in NEXN-mutant zebrafish and adult human cardiomyocytes [[149]27]. In contrast, no discernible alterations in sarcomeric or Z-disc architecture were observed in cardiac tissue-specific NEXN knockout murine models [[150]9], and autopsy of neonates with homozygous NEXN mutations demonstrated normal Z-disc development [[151]7]. Consistent with these findings, our study detected no significant myofilament or sarcomeric disarray in NEXN-knockout cardiomyocytes (Figures [152]S4B, C), suggesting that NEXN, despite its localization at the Z-disc of sarcomeres, may have limited involvement in sarcomeric stabilization. Furthermore, NEXN knockout resulted in mitochondrial enlargement and cristae reduction (Fig. [153]5B), morphological anomalies that likely impair mitochondrial function, reinforcing the hypothesis that NEXN deficiency disrupts cardiomyocyte energy metabolism. During cardiomyocyte maturation, the predominant energy metabolic pathway shifts from glycolysis to oxidative phosphorylation primarily fueled by fatty acid metabolism. Studies demonstrated that, compared to adult primary cardiomyocytes, hiPSC-CMs exhibited metabolic immaturity resembling fetal cardiomyocytes, characterized by reduced fatty acid utilization and persistent reliance on glycolysis [[154]28]. In our study, NEXN knockout significantly suppressed fatty acid metabolism in hiPSC-CMs (Figs. [155]1G and [156]5A), suggesting that NEXN deficiency impeded metabolic maturation. This impairment may underlie the high mortality observed in fetuses and neonates harboring homozygous NEXN mutations [[157]4, [158]7, [159]29, [160]30]. In the context of calcium handling, previous studies demonstrated that NEXN knockout in murine embryonic or adult stages reduces protein levels of RyR2, Cacna1c, and Serca2 while elevating Cacna2d1 and Casq1 expression, resulting in significantly diminished calcium transient amplitude, prolonged duration, and delayed decay kinetics [[161]9, [162]10]. Our findings corroborated these observations (Figures [163]S4E, F and 3D-J). Elevated intracellular calcium triggered mitochondrial calcium overload, exacerbating oxidative stress and ultimately inducing cardiomyocyte apoptosis (Figs. [164]4A-F and [165]5D-G). Notably, our study identified a four-phase calcium transient profile in NEXN-knockout cardiomyocytes: rapid decay, plateau phase, secondary rapid decay, and slow decay (Fig. [166]3E), likely mediated by coordinated SERCA and NCX activities. The possible reasons are as follows. Rapid decay phase: this phase is likely mediated primarily by SERCA2-dependent calcium reuptake. Western blot analysis in our study demonstrated reduced SERCA2 protein levels in NEXN^−/− hiPSC-CMs (Figure [167]S4F, G), which may account for the slower calcium reuptake rate during this phase compared to wild-type controls. Plateau phase: the activity of SERCA2 may be further suppressed due to altered phosphorylation status of phospholamban (PLB), a key regulator of SERCA. Reduced PLB phosphorylation could enhance its inhibitory interaction with SERCA2, leading to transient stagnation of calcium reuptake and the observed plateau [[168]31, [169]32]. Secondary rapid decay phase: during the plateau phase, cytosolic calcium accumulation may reach a threshold that activates forward-mode NCX. Concurrent restoration of the sodium gradient by Na⁺/K⁺-ATPase activity enables a transient surge in NCX-mediated calcium extrusion, driving the secondary rapid decay. Slow decay phase: Sodium gradient depletion: prolonged NCX activity elevates intracellular Na⁺ levels, overwhelming the Na⁺/K⁺-ATPase’s capacity to restore the gradient, thereby diminishing NCX efficiency. Normal calcium handling in cardiomyocytes requires intact JMCs. Our study demonstrated that NEXN deficiency reduced JPH2 protein level (Figures [170]S4E, F), disrupting JMCs integrity as evidenced by increased sarcoplasmic reticulum-sarcolemmal separation (Figs. [171]3A, B), ultimately impairing excitation-contraction coupling (Figs. [172]2I, J). It was reported that JPH2 expression was regulated at transcriptional and post-translational levels [[173]33–[174]36]. The molecular mechanism underlying NEXN knockout-induced JPH2 downregulation remains to be elucidated. In our study, rescuing NEXN protein expression significantly enhanced calcium handling and energy metabolism while mitigating mitochondrial oxidative stress. Notably, treatment of NEXN-knockout cardiomyocytes with either LC or SA1 markedly improved excitation-contraction coupling and metabolic function (Figs. [175]6E-I), though no discernible effects on mitochondrial ultrastructure or calcium transient dynamics were observed (Figs. [176]6A-D). These findings collectively highlighted LC and SA1 as potential therapeutic agents for DCM. However, whether combinatorial administration of these compounds yields synergistic benefits requires further investigation. In modeling dilated cardiomyopathy, previous study has demonstrated that hiPSC-CMs resemble cardiomyocytes from early human embryonic heart development [[177]37]. Proteomic profiling further reveals that hiPSC-CMs exhibited low expression of fatty acid metabolism-associated proteins, mirroring the proteomic signature of fetal primary cardiomyocytes [[178]38]. In our study, enhancing fatty acid metabolism via LC treatment improved cellular energy metabolism (Figs. [179]6H, I), suggesting LC as a potential strategy to augment hiPSC-CMs maturation. Current approaches to enhance hiPSC-CMs maturity include prolonged culture duration, engraftment into embryonic hearts, electrical stimulation, engineered 3D tissue/organoid systems, and co-culture with endothelial cells to form microtissues [[180]39]. In future investigations, we will employ these methodologies to study DCM in models that more faithfully recapitulate human pathophysiology. In summary, our study revealed that NEXN worked as a critical regulator of cardiomyocyte excitation-contraction coupling and energy metabolism homeostasis. NEXN deficiency caused JMCs uncoupling, characterized by T-tubule-sarcoplasmic reticulum disconnection, reduced expression and disorganized distribution of JPH2/RyR2 complexes, mitochondrial dysfunction and aberrant calcium handling. Ultimately, this maladaptive cascade drived cardiomyocyte decompensation, pathological hypertrophy, and progression to dilated cardiomyopathy. Conclusions We utilized hiPSCs in conjunction with the CRISPR/Cas9 system to establish a myocardial cell model with a NEXN defect, to investigate the pathogenic mechanism of dilated cardiomyopathy caused by NEXN. This model serves as a valuable tool for investigating the etiology of cardiac diseases in humans, identifying associations between genes and phenotypes, and facilitating the discovery of novel therapeutic agents. The cardiomyocyte model developed in this study demonstrates significant potential for clinical translation and application due to its high customization and versatility as a technology platform. Electronic supplementary material Below is the link to the electronic supplementary material. [181]Supplementary Material 1^ (5.1MB, docx) [182]Supplementary Material 2^ (19.7KB, docx) Acknowledgements