Upregulation of utrophin improves the phenotype of Duchenne muscular dystrophy hiPSC-derived CMs

Abstract

Duchenne muscular dystrophy (DMD) is a genetic neuromuscular disease. Although it leads to muscle weakness, affected individuals predominantly die from cardiomyopathy, which remains uncurable. Accumulating evidence suggests that an overexpression of utrophin may counteract some of the pathophysiological outcomes of DMD. The aim of this study was to investigate the role of utrophin in dystrophin-deficient human cardiomyocytes (CMs) and to test whether an overexpression of utrophin, implemented via the CRISPR-deadCas9-VP64 system, can improve their phenotype. We used human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs) lacking either dystrophin (DMD) or both dystrophin and utrophin (DMD KO/UTRN^(+/−)). We carried out proteome analysis, which revealed considerable differences in the proteins related to muscle contraction, cell-cell adhesion, and extracellular matrix organization. Furthermore, we evaluated the role of utrophin in maintaining the physiological properties of DMD hiPSC-CMs using atomic force microscopy, patch-clamp, and Ca²⁺ oscillation analysis. Our results showed higher values of afterhyperpolarization and altered patterns of cytosolic Ca²⁺ oscillations in DMD; the latter was further disturbed in DMD KO/UTRN^(+/−) hiPSC-CMs. Utrophin upregulation improved both parameters. Our findings demonstrate for the first time that utrophin maintains the physiological functions of DMD hiPSC-CMs, and that its upregulation can compensate for the loss of dystrophin.

Graphical abstract

Andrysiak and colleagues demonstrated that utrophin can compensate for the loss of dystrophin in human dystrophin-deficient CMs. They used CRISPR-deadCas9-VP64 coupled with sgRNA targeting UTRN promoter to upregulate the utrophin, which resulted in improved action potential and calcium oscillations characteristics of these cells.

Introduction

Duchenne muscular dystrophy (DMD) is an X-linked genetic disorder caused by mutations in the DMD gene encoding dystrophin, a protein that plays an important role in maintaining the integrity of the cell membrane. Thus, the loss of this protein leads to severe muscle damage, which clinically manifests as a gradual development of muscle weakness.¹ In its later stages, the disease is also associated with respiratory failure and progressive cardiomyopathy.² While the latter is the main cause of death in patients with DMD, its underlying mechanisms remain elusive.

Although a limited number of therapies have been approved to slow the progression of DMD and/or improve the quality of life of patients, no effective cure currently exists. One of the FDA-approved therapies relies on the exon skipping strategy using antisense oligonucleotides (ASOs),³ which allow for the omission of mutated exons during translation and the restoration of the biosynthesis of a truncated but still functional protein. However, these drugs are intended only for patients with mutations in specific exons. Moreover, ASOs are poorly delivered to the heart and are therefore unable to prevent the development of heart failure.^4,5 Research is already underway to solve this problem, mainly by designing ASOs with additional modifications, such as incorporating the peptide-conjugated morpholino oligomer backbone and conjugation to cell-penetrating peptides or by developing of new classes of ASOs, such as tricyclo-DNAs.^6,7 Recently, the FDA also conditionally approved a gene therapy that relies on delivery of the adeno-associated virus (AAV) encoding microdystrophin.⁸ However, the therapeutic outcome of restoring dystrophin entails the problem of the immune response against dystrophin, which is recognized as foreign by the immune system. To this end, the upregulation of non-immunogenic utrophin could be beneficial independently of the type of mutation.

Utrophin is a 395-kDa protein encoded by the UTRN gene located on chromosome 6 in humans (or chromosome 10 in mice).^9,10 Utrophin is homologous to dystrophin in terms of its structure and function: the two proteins share a similar spatial arrangement of functional domains (Figure 1A) and form analogical complexes with glycoproteins.¹¹ In humans, utrophin is predominantly expressed during fetal development in the muscle sarcolemma, where it plays a stabilizing role; postnatally, its function is replaced by dystrophin.¹² Nevertheless, utrophin is still present in mature muscle fibers, though its expression is restricted to neuromuscular junctions (NMJs) and myotendinous junctions (MTJs).¹³ There are two full-length utrophin isoforms, utrophin A and utrophin B, which differ in their N-terminal acting binding site and their site of expression: utrophin A expression is restricted to muscle cells, while utrophin B is usually present in endothelial cells and blood vessels.¹⁴ Accordingly, for the treatment of muscular dystrophies, utrophin A is considered the therapeutic target. Importantly, the most commonly used animal model of DMD, mdx mice, shows a relatively mild phenotype of the disease, which has been associated with the lifetime presence of utrophin expression.¹⁵ In accordance with the above, mice deficient in both dystrophin and utrophin (mdx/utrn^−/−) demonstrate a strikingly severe phenotype with a fairly rapid development of cardiomyopathy, unseen in the mdx strain.¹⁶ Therefore, it was hypothesized that utrophin can compensate for loss of functional dystrophin, and its overexpression may alleviate the symptoms of DMD.¹⁷ This hypothesis could be further supported by previous observations that the upregulation of utrophin in DMD patients correlates with a decreased severity of the disease.^18,19,20 However, there are also literature reports that contradict this correlation.^21,22

Figure 1.

Generation of the utrophin-deficient DMD hiPSC line

(A) Schematic comparison of the full-length utrophin vs. the full-length dystrophin with approximate percentage similarity of the individual domains marked. NT, N-terminal domain; H1–H4, hinges 1–4; CR, cysteine-rich domain; CT, C-terminal domain. Created with BioRender.com. (B) Surveyor nuclease assay verifying the efficiency of the designed sgRNA sequences. (C) Surveyor nuclease assay verifying editing occurrence in individual hiPSC clones. (D) confirmation of the lack of utrophin in hiPSC-CMs differentiated from the generated DMD KO/UTRN^(+/−) hiPSC line by western blotting.

Although the role of utrophin has been extensively investigated in the context of dystrophic skeletal muscles, its role in the human heart has not been thoroughly explored to date. Therefore, the aim of this work was to test whether the presence of utrophin in human CMs may serve as a compensatory mechanism and improve the physiological properties of dystrophin-deficient cells. To this end, we used human induced pluripotent stem cell (hiPSC)-derived cardiomyocytes (hiPSC-CMs). In our previous studies, we generated the hiPSC line with a deletion of exon 50 in the DMD gene using clustered regularly interspaced short palindromic repeats (CRISPR)-CRISPR-associated protein 9 (Cas9) (CRISPR-Cas9) gene editing technology.²³ Here, applying the same method, we introduced a frameshift mutation in the UTRN gene in these cells and generated a double-knockout hiPSC line deficient for both dystrophin and utrophin. In order to overexpress utrophin in DMD hiPSC-CMs, we used the single AAV vector-mediated delivery of the CRISPR-deadCas9-VP64 (CRISPR-dCas9-VP64) system, coupled with sgRNA targeting the UTRN promoter. Our results demonstrate that utrophin maintains the physiological functions of DMD hiPSC-CMs, and that its upregulation can compensate for the loss of dystrophin.

Results

Generation of the hiPSC line with utrophin deficiency

To verify how the loss of utrophin affects dystrophin-deficient CMs, the DMD hiPSC line additionally lacking utrophin protein was generated. First, the DMD hiPSC were nucleofected with plasmids encoding one of the three designed sgRNAs that target exon 11 of the UTRN gene. Then, the Surveyor nuclease assay was carried out to assess the efficiency of the sequences, and sgRNA1 was selected as the most prominent (Figure 1B). Therefore, nucleofected cells were seeded for single-cell-derived clones, and the presence of mutations in the targeted UTRN region in collected cells was verified by the Surveyor nuclease assay (Figure 1C), followed by Sanger sequencing. The mutation, specifically 108-bp deletion encompassing the splice acceptor site on the 5′ end and the fragment of UTRN exon 11, was detected in clone 9, although in one allele only. Importantly, western blot analysis confirmed the presence of utrophin in the control and dystrophin-deficient hiPSC-CMs, and the lack of this protein in hiPSC clone 9-derived hiPSC-CMs (Figure 1D), which are hereinafter referred to as DMD KO/UTRN^(+/−). Further studies are underway to elucidate the consequences of the knockout of one or both UTRN alleles on the phenotype of human DMD CMs. Of note, the CRISPR-Cas9 approach did not compromise their genomic integrity, as evidenced by the intact karyotype (Figure S1A) and lack of alterations in putative sites of sgRNA1 off-target activity (Figure S1B).

Proteomic analysis revealed considerable differences between DMD and DMD KO/UTRN^(+/−) hiPSC-CMs

Despite their wide application in in vitro studies, a well-known limitation of hiPSC-CMs obtained from hiPSC by a widely established cardiac differentiation protocol is their immature, fetal-like phenotype, which may not reflect all the characteristics of adult CMs. However, in our study, this limitation proved to be advantageous. Since utrophin is still expressed in the stage of fetal development, hiPSC-CMs could be a perfect model to verify whether the presence of this protein affects the phenotype of dystrophic CMs. Therefore, we carried out proteomic analysis of DMD KO/UTRN^(+/−) hiPSC-CMs and compared the results to those of DMD and control hiPSC-CMs. Principal component analysis (PCA) demonstrated that the distribution of the tested samples (Figure 2A, top panel) was affected by the differentiation batch, indicating the presence of technical factors in each differentiation that were independent of dystrophin and dystrophin/utrophin deficiency in hiPSC-CMs, which additionally contributed to the observed variability between the samples. Thus, we applied the batch effect correction procedure, which improved the sample distribution (Figure 2A, bottom panel). In the next step, we analyzed the proteomic profiles of all tested hiPSC-CM groups and identified 118 differentially abundant proteins (DAPs) in DMD hiPSC-CMs compared to control hiPSC-CMs, 329 DAPs in DMD KO/UTRN^(+/−) hiPSC-CMs compared to control hiPSC-CMs and 205 DAPs in DMD KO/UTRN^(+/−) hiPSC-CMs compared to DMD hiPSC-CMs (Figure 2B). The most striking result that emerged from the data is that coexisting utrophin deficiency in dystrophic hiPSC-CMs has a considerable effect on the number of proteins with changed abundance—even greater than dystrophin deficiency alone compared to the control. To understand the biological and functional significance of these proteins, we performed enrichment analysis. Our data support the notion that utrophin is crucial for maintaining cytoskeleton and extracellular matrix organization, as well as cell-cell and cell-surface adhesion, as well as cell adhesion mediated by integrins (Figures 3A and 3B). Moreover, the protein was associated with biological processes such as muscle contraction and muscle system processes (Figures 3A and S2A), and the DAPs were assigned to the following cellular compartments: myofibrils, contractile fiber, sarcomere, Z disc and I band (Figure 3B). Interestingly, the Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis indicated dilated, hypertrophic, and arrhythmogenic right ventricular cardiomyopathy as enriched pathways (Figure 3C), as well as estrogen signaling, focal adhesion, and PI3K-Akt signaling pathways (Figures 3C and S2B)—well-documented mediators of heart failure.^24,25,26 Finally, enrichment analysis illustrated the involvement of DAPs in suppressing molecular functions related to alpha-actinin and extracellular matrix binding (Figure 3D), which are crucial for cytoskeleton organization.

Figure 2.

Proteomic analysis of CTRL vs. DMD vs. DMD KO/UTRN^(+/−) hiPSC-CMs

(A) Principal component analysis (PCA) demonstrating the distribution of the samples before (upper panel) and after (lower panel) batch correction and (B) volcano plots demonstrating differentially abundant proteins (DAPs) between the groups based on p value <0.05 + log2 fold change > |−0.5|.

Figure 3.

Enrichment analysis of DAPs in DMD KO/UTRN^(+/−) hiPSC-CMs compared to DMD hiPSC-CMs

(A) Altered biological processes, (B) cellular compartments, (C) KEGG pathway, and (D) molecular functions.

Changes in the stiffness of dystrophin-deficient hiPSC-CMs

The crucial function of dystrophin in muscle cells is to connect the extracellular matrix with the intracellular cytoskeleton and thereby stabilize the sarcolemma during continuous cycles of contraction and relaxation.²⁷ Indeed, our proteomic analysis illustrated changes in the cytoskeleton and extracellular matrix organization in DMD KO/UTRN^(+/−) hiPSC-CMs compared to DMD hiPSC-CMs. Furthermore, staining for dystroglycan (DAG1), a central component of the dystrophin-associated glycoprotein complex (DGC), clearly demonstrated that dystrophin deficiency is associated with disrupted DGC formation, and that this effect is even more pronounced with the additional loss of utrophin (Figure S3A). Consequently, the loss of both these proteins may affect the mechanical properties of the cell membrane and the cortex underneath. Since cell membrane stiffness is dependent on the cytoskeleton arrangement, atomic force microscopy (AFM) measurements were carried out to examine this physical parameter in DMD vs. DMD KO/UTRN^(+/−) hiPSC-CMs. Specifically, the cell elasticity, known as Young’s modulus, was investigated by measuring cellular deformability in response to compression with the probe. Our results show that Young’s modulus was significantly higher in DMD hiPSC-CMs, which corresponds to greater stiffness of these cells (Figure S4A). Interestingly, the values for DMD KO/UTRN^(+/−) hiPSC-CMs were lower, suggesting that the incorporation of utrophin into DGC may contribute to the increased stiffness of DMD hiPSC-CMs. Additionally, the measurements revealed higher values of contraction amplitude, i.e., differences in cell deformability during contraction and relaxation in DMD hiPSC-CMs compared to their control counterparts (Figure S4B). Although these values were also higher for DMD KO/UTRN^(+/−) cells compared to the control group, the effect was not as evident as it was for DMD hiPSC-CMs (Figure S4B), and no statistical significance was found between DMD and DMD KO/UTRN^(+/−) cells.

Action potential characteristics of DMD and DMD KO/UTRN^(+/−) hiPSC-CMs

As stated above, one of the objectives of the study was to test the hypothesis that there is a degree of functional redundancy between dystrophin and utrophin, and thus the simultaneous loss of both these proteins will result in a greater deterioration of hiPSC-CMs functions compared to dystrophin knockout alone. Given that the proteome analysis identified the utrophin function in cellular compartments associated with cell contraction and related KEGG pathways, we decided to measure the electrophysiological properties of a single hiPSC-CMs using a whole-cell patch-clamp method. Our results revealed an increased amplitude of afterhyperpolarization (AHP) (Figure 4A), which corresponds to the greater difference between the resting membrane potential and the maximal diastolic potential reached in the final stage of the action potential in DMD hiPSC-CMs compared to their control counterparts. The additional loss of utrophin in dystrophin-deficient cells did not significantly affect AHP values, and no marked differences were present in other measured parameters: resting membrane potential (Figure 4B), upstroke velocity (Figure 4C), and action potential durations (APDs) at 20%, 50%, and 90% of repolarization between the study groups (Figures 4D–4F).

Figure 4.

Action potential characteristics of DMD and DMD KO/UTRN^(+/−) hiPSC-CMs

(A) AHP amplitude, (B) resting membrane potential values, (C) AP upstroke velocity, (D) APD at 20% repolarization, (E) APD at 50% repolarization, and (F) APD at 90% repolarization. Data are presented as mean ± SD (∗p < 0.05, one-way ANOVA with Tukey’s multiple comparison, test. n = 20–30 CMs/group).

Alterations of Ca²⁺ oscillations in DMD KO/UTRN^(+/−) hiPSC-CMs

In the next step, we decided to study how a simultaneous loss of dystrophin and utrophin affects the physiological properties, particularly Ca²⁺ homeostasis, of hiPSC-CMs seeded in small syncytia-like clusters. For this purpose, real-time measurements of intracellular Ca²⁺ oscillations were carried out in these cells. The results revealed that the physiological oscillatory pattern was frequently altered in DMD hiPSC-CMs and even more so in DMD KO/UTRN^(+/−) hiPSC-CMs, as shown by representative traces (Figure 5A) and the total oscillatory pathology score (Figure S5A). Both in DMD and in DMD KO/UTRN^(+/−) hiPSC-CMs, approximately one-third of all oscillations showed the presence of multiple Ca²⁺ spikes (Figure 5B, left panel). However, the DMD KO/UTRN^(+/−) cells had a higher fraction of oscillations on top of the prolonged Ca²⁺ plateau (Figure 5B, left panel), which primarily contributed to their statistically higher incidence of abnormal oscillations compared to both control and DMD hiPSC-CMs (Figure 5B, right panel). Moreover, both DMD and DMD KO/UTRN^(+/−) hiPSC-CMs had lower oscillation frequency on average compared to the control (Figure 5C, right panel), while in the DMD KO/UTRN^(+/−) group, “high-frequency oscillations” accounted for only about one-fifth of all oscillations (Figure 5C, left panel). In addition, the CMs deficient in dystrophin showed random increases in intervals (breaks) between oscillations (Figure 5D, left panel), and this effect was essentially the same in DMD and DMD KO/UTRN^(+/−) hiPSC-CMs (Figure 5D, right panel).

Figure 5.

Alterations of Ca²⁺ oscillations in DMD and DMD KO/UTRN^(+/−) hiPSC-CMs

(A) Representative Ca²⁺ traces recorded in control, DMD, and DMD KO/UTRN^(+/−) hiPSC-CMs.

(B–D) The analysis of Ca²⁺ oscillation characteristics is depicted in pie charts (fraction of scored parameters, left) and bar charts (average score, right); the scoring was performed in the following categories: (B) oscillatory pattern, (C) frequency of oscillations frequency, and (D) continuity of oscillations. The value of each parameter was calculated based on the scoring system shown in Table S6. Data are presented as mean ± SEM (∗∗p < 0.01, ∗∗∗p < 0.005, ∗∗∗∗p < 0.0001, Kruskal-Wallis test with Dunn’s multiple comparison assay, n = 4 independent experiments with 246–260 analyzed cells per group).

Upregulation of utrophin in hiPSC-CMs

In order to activate the utrophin in DMD hiPSC-CMs, the CRISPR-dCas9 strategy was applied using single-AAV-vector transduction. Based on previous reports from our group,²⁸ we decided to use the AAV6 serotype, as it provides markedly improved hiPSC-CMs transduction efficiency compared to the AAV9 serotype. In the next step, the hiPSC-CMs were transduced with AAV vectors encoding CRISPR-dCas9-VP64 and one of four variants of the sgRNA sequence (P1–P4) or a control AAV without any sgRNA (AAV-empty). The AAV vector encoding the sgRNA P1 sequence was the most effective in utrophin activation, resulting in the highest levels of utrophin transcript (Figure 6A) and protein (Figure 6B). Utrophin activation was tested again after transduction of hiPSC-CMs with additionally purified AAV vectors, and an almost 4-fold increase in the level of UTRN mRNA was found (Figure 6C).

Figure 6.

Activation of utrophin in DMD hiPSC-CMs

(A and B) mRNA (A) and protein level (B) of utrophin after transduction with AAV6 vectors encoding different sgRNA sequences (P1–P4).

(C) Utrophin mRNA level in DMD hiPSC-CMs transduced with purified AAV6-CRISPR-dCas9-VP64-sgRNA P1 vectors. Data presented as mean ± SD.

Characteristics of the action potential of DMD hiPSC-CMs after UTRN activation

Once the dystrophic hiPSC-CMs with activated utrophin expression were generated, we subjected them to analogous patch-clamp analysis to investigate whether the greater abundance of the protein in DMD hiPSC-CMs can affect their action potential. For this approach, DMD hiPSC-CMs were transduced with AAV vector encoding sgRNA P1 to activate utrophin (DMD AAV-VP64-UTRN), while both control and DMD hiPSC-CMs were transduced with an AAV-empty vector (CTRL AAV-empty and DMD AAV-empty, respectively) as a control. Importantly, the upregulation of utrophin restored AHP, the parameter elevated in DMD hiPSC-CMs, to the values seen in control cells (Figure 7A). No additional significant differences were found between the DMD AAV-empty and DMD AAV-VP64-UTRN hiPSC-CMs groups in regard to other action potential parameters: resting membrane potential (Figure 7B), upstroke velocity (Figure 7C), and APDs at 20%, 50%, and 90% of repolarization (Figures 7D–7F).

Figure 7.

Action potential characteristics of CTRL AAV-empty, DMD AAV-empty, and DMD AAV-VP64-UTRN hiPSC-CMs

(A) AHP amplitude, (B) resting membrane potential values, (C) AP upstroke velocity, (D) APD20, (E) APD50, and (F) APD90. Data are presented as mean ± SD (∗p < 0.05, one-way ANOVA with Tukey’s multiple comparison test, n = 20–30 CMs/group).

Utrophin upregulation improves altered intracellular Ca²⁺ oscillations caused by the loss of dystrophin

Following the patch-clamp method, we carried out real-time measurements of intracellular Ca²⁺ oscillations to test whether utrophin upregulation could rescue the phenotype of DMD hiPSC-CMs. As illustrated by the representative Ca²⁺ traces (Figure 8A) and the total oscillatory pathology score (Figure S5B), the pattern, frequency, and continuity were indeed improved in DMD AAV-VP64-UTRN hiPSC-CMs compared to DMD hiPSC-CMs transduced with empty vectors (DMD AAV-empty hiPSC-CMs). The oscillations resembled those of the control cells. Our quantitative analysis showed that the upregulation of utrophin resulted in a reduced fraction of abnormal oscillations compared to the DMD AAV-empty hiPSC-CMs, in which the majority of oscillations took the form of multiple spikes or a prolonged plateau (Figure 8B). Furthermore, although the oscillation frequency was lower on average in DMD AAV-empty hiPSC-CMs, it increased back to the control levels (CTRL AAV-empty) after upregulation of utrophin (Figure 8C). Finally, the prevalence of both short and long breaks in oscillations was significantly lower in DMD AAV-VP64-UTRN hiPSC-CMs compared to DMD AAV-empty hiPSC-CMs, in which only one-third of the oscillations showed normal continuity (Figure 8D). Taken together, our results demonstrate that utrophin upregulation not only restores DGC formation, as shown by the increased levels of DAG1 in DMD AAV-VP64-UTRN hiPSC-CMs (Figure S3B), but it also improves the dystrophic phenotype of hiPSC-CMs in terms of their physiological activity, strongly suggesting that utrophin can compensate for the loss of dystrophin in DMD cells.

Figure 8.

The effect of utrophin upregulation in DMD hiPSC-CMs on Ca²⁺ oscillations

(A) Representative Ca²⁺ traces recorded in CTRL AAV-empty, DMD AAV-empty, and DMD AAV-VP64-UTRN hiPSC-CMs.

(B–D) The analysis of Ca²⁺ oscillation characteristics is depicted on pie charts (fraction of the parameters scored, left) and bar charts (average score, right); the scoring was done in the following categories: (B) oscillatory pattern, (C) frequency of oscillations, and (D) continuity of oscillations. The value of each parameter was calculated based on the scoring system shown in Table S6. Data are presented as mean ± SEM (∗∗∗∗p < 0.0001, Kruskal-Wallis test with Dunn’s multiple comparison assay, n = 4 independent experiments with 246–268 analyzed cells per group).

Discussion

A growing body of evidence in the literature suggests that utrophin overexpression could be a promising therapeutic approach to alleviate the effects of dystrophin deficiency in DMD patients.^14,29,30,31 The rationale behind this approach is based on these facts: (1) utrophin is a paralog of dystrophin; (2) utrophin is capable of replacing dystrophin in anchoring the sarcolemma to the cortex cytoskeleton; (3) as expressed prenatally, utrophin is non-immunogenic when overexpressed, as opposed to dystrophin, which often becomes recognized as a foreign protein by the immune system in patients with DMD³²; and (4) approach does not depend on the type of DMD mutation found in affected individuals. Indeed, several preclinical studies demonstrate the beneficial effects of utrophin upregulation in skeletal muscles.^17,32,33,34 However, the therapeutic potential of this protein in the heart has not yet been thoroughly investigated. In particular, we lack data on the role of utrophin in human dystrophic CMs. To fill this gap, we applied hiPSC-CMs, which have become a commonly used model of human heart cells in vitro for both investigating the pathophysiology of cardiac diseases and drug testing.³⁵

In this study, we first demonstrated that, in contrast to adult human CMs, our hiPSC-CMs still express utrophin. To test the hypothesis that the presence of utrophin may affect the phenotype of DMD hiPSC-CMs and partially compensate for the loss of dystrophin, we generated the hiPSC-CM line lacking both proteins (DMD KO/UTRN^(+/−)). This was followed by extensive proteomic analysis of DMD KO/UTRN^(+/−), dystrophin-deficient cells (DMD), and control hiPSC-CMs. Both PCA analysis and volcano plots representing changes in protein abundance between the groups revealed considerable differences between DMD and DMD KO/UTRN^(+/−) hiPSC-CMs. Enrichment analysis indicated that the functional and biological significance of most of these proteins is related to the mechanisms of contraction, cardiomyopathy development, the cytoskeleton, and related cellular compartments (sarcomeres, myofibrils, contractile fibers, etc.). As a result, we focused our studies on unraveling how the loss of utrophin in dystrophic hiPSC-CMs may affect the mechanical and physiological properties of these cells.

Interestingly, we found that DMD hiPSC-CMs have increased membrane stiffness compared to their control counterparts, and this difference is reversed by the coexisting loss of utrophin. This implies that the latter protein, while compensating for the absence of dystrophin, may alter the mechanical properties of the sarcolemma and the subcellular cytoskeletal cortex. Indeed, previous studies have indicated different modes of actin binding by dystrophin and utrophin.³⁶ Additionally, Ramirez et al. recently revealed that membrane stiffness may depend on the posttranscriptional modification of utrophin, primarily phosphorylation.³⁷

Our studies demonstrated that double-knockout hiPSC-CMs have more severely altered spontaneous cytosolic Ca²⁺ oscillations, with a higher incidence of discontinuous oscillations, a prolonged plateau, and a lower frequency of oscillations compared to DMD hiPSC-CMs. On the other hand, we did not find significant differences between DMD KO/UTRN^(+/−) and DMD hiPSC-CMs in terms of the electrophysiological properties analyzed by whole-cell patch-clamp technique. This discrepancy may be explained by the location of utrophin in CMs, which, in addition to its presence in the sarcolemma, was also found in intercalated discs, as previously shown in the murine heart.^38,39 Intercalated discs are structures that connect individual CMs and are involved in maintaining signal conductivity.⁴⁰ This is achieved by a number of protein complexes located in the three main regions of intercalated discs: fascia adherens junctions, desmosomes, and gap junctions, which are responsible for electrical, mechanical, and chemical signal transmission.⁴¹ Alterations in the normal functioning of intercalated discs have been reported in several cardiac diseases: arrhythmogenic cardiomyopathy, dilated cardiomyopathy, Friedreich cardiomyopathy, and Brugada syndrome.^{42,43,44,45,46} Utrophin has been found in two types of intercellular connections: the aforementioned NMJs and MTJs. Previous studies in utrophin-deficient mice highlighted the role of this protein in stabilizing the neuromuscular synapses,^47,48,49 in which the loss of utrophin resulted in much fewer acetylcholine receptors and postsynaptic infoldings. In our experimental setup, the physiological properties of hiPSC-CMs were analyzed in both single cells and small syncytia-like clusters of cells, revealing that the loss of utrophin particularly affects the characteristics of the latter. This may indicate a crucial role of utrophin in signal transduction between CMs, considering that its absence does not entail any changes in the action potential properties of single cells, but it significantly worsens spontaneous cytosolic Ca²⁺ oscillations.

These observations concur well with our proteomic data, which pointed to the involvement of utrophin in such biological processes as cell-cell and cell-surface adhesion, focal adhesion, and cell adhesion mediated by integrins. It is certainly related to impaired DGC formation and dysfunction in ECM binding and cytoskeleton organization caused by a loss of utrophin in glycoprotein complex. Furthermore, dystroglycan, which is an important mediator of cell adhesion signaling,⁵⁰ becomes downregulated in DMD KO/UTRN^(+/−) hiPSC-CMs. Consequently, impaired intracellular communication involves alterations in electrical coupling and signal transduction. Cardiomyopathies resulting from mutations in genes that encode protein components of cell adhesion are well documented.⁵¹ Importantly, our data demonstrate that CRISPR-dCas9-VP64-mediated utrophin upregulation restored normal expression of DAG1 and intracellular Ca²⁺ handling in dystrophic hiPSC-CMs, which provides additional strong evidence to support the postulated mechanism.

Another possible explanation for the utrophin-dependent changes in both Ca²⁺ oscillations and AHP parameters is its involvement in regulating the opening of mechanosensitive (MS) ion channels, as reported previously.⁵² The mode of action of MS channels strictly depends on the mechanical properties of the sarcolemma and ECM, and it is well known that they can modulate Ca²⁺ ion entry and can interact with calcium regulatory proteins.⁵³ Earlier studies have shown prolonged opening of the MS ion channels in DMD⁵⁴; the gating of these channels has also been shown to depend on utrophin.⁵² Consequently, this phenomenon might lead to an excessive influx of extracellular Ca²⁺ ions and potential disturbances of excitation-contraction coupling. To the best of our knowledge, a connection between the MS ion channels and prolonged voltage-gated potassium channel opening (which are responsible for higher values of AHP) has not been described yet, but it would certainly be worth exploring.

Despite the growing interest in the application of utrophin upregulation in the treatment of DMD-associated cardiomyopathy, no studies on its role in human CMs have been conducted. To our knowledge, this is the first report demonstrating that upregulation of utrophin markedly improves the physiological functions of dystrophic CMs of human origin. Utrophin overexpression in hiPSC-CMs was achieved through the CRISPR-dCas9-VP64 system delivered by AAV vectors. In the past, other approaches have also been proposed to upregulate utrophin for therapeutic purposes. Ezutromid (SMT C1100), a small molecule compound developed by Summit Therapeutics, is one of the most documented, as it has been validated in clinical trials. Initial studies carried out in vitro and in vivo (in mdx mice) were very promising: ezutromid caused an increase in utrophin mRNA and protein levels and improved muscle function.³¹ Since the drug was well tolerated in the phase 1 clinical trials, it progressed to phase 2 to test its effectiveness in the DMD patient cohort. While the first reports of the 24-week analysis were encouraging and showed that ezutromid increased utrophin levels and alleviated muscle damage, the 48-week observations did not meet the expected endpoints, and the study was discontinued (NCT02858362). Although not entirely successful, these studies do not devalue utrophin as a promising therapeutic target but rather indicate the need to develop effective methods for its upregulation.

Although the therapeutic potential of upregulating utrophin using the dCas9-VP64 system in DMD hiPSC-CMs requires further investigation, this approach is an innovative and exciting alternative to existing treatment strategies. A similar approach was previously tested by Wojtal et al., who used the CRISPR-dCas9-VP16 system to activate utrophin in myoblasts.⁵⁵ They compared the upregulation efficiency of utrophin A vs. B and reached the conclusion that utrophin B is an easier and more attainable target for overexpression (3.8- to 6.9-fold increase) compared to the more challenging utrophin A (1.7- to 2.7-fold increase).⁵⁵ The observed difference likely results from substantial methylation of the utrophin A promoter and the associated difficulty of finding accessible open chromatin sites, as demonstrated by an increased DNAse I footprint density in the promoter of utrophin A compared to utrophin B.⁵⁵ The CRISPR-dCas9-VP64 system, used in our study, provides a reliable tool for upregulating of utrophin A, which can be of clinical benefit. Additionally, the system relies on non-immunogenic and non-integrating AAV vectors, providing long-lasting transgene expression. These vectors can be successfully used in clinical application, as AAV-based therapies have already been approved by the FDA for the treatment of inherited retinal disease (Luxturna, AAV2), spinal muscular atrophy (Zolgensma, AAV9), hemophilia B (Hemgenix, AAV5) and hemophilia A (Roctavian, AAV5). Just recently, AAV therapy was also conditionally accepted for DMD (Elevidys, AAVrh74).⁸ Of note, we developed a single-vector system that concomitantly delivers both a specific sgRNA and dCas9-VP64, limiting the potential toxicity of AAV application. Nevertheless, more research is required to fully validate the therapeutic potential of this approach in preclinical settings.

The last few years have witnessed considerable progress in the development of new treatment strategies for patients with DMD. It was reported that the aforementioned gene therapy, ELEVIDYS, which involves delivering a shortened version of dystrophin (microdystrophin), provides functional improvement and long-term safety in treated individuals. Interestingly, despite previous concerns about immunity to dystrophin, most of the patients involved in clinical trials of this drug did not experience these types of adverse reactions. However, based on the experience of this study—in which one patient suffered from myositis⁵⁶—and previous reports,⁵⁷ it was hypothesized that immune reactions to dystrophin occur only in patients with mutations in a specific region, namely deletions of exons 9–13.⁵⁶ This can be explained by the fact that this region encodes Hinge 1, which is not identical to the corresponding sequence present in utrophin; therefore, when provided by microdystrophin, it can induce the unwanted production of anti-dystrophin antibodies.⁵⁸ This observation can hopefully resolve the problem of such adverse side effects related to dystrophin immunity by redesigning the microdystrophin sequence in the future. Otherwise, potential utrophin-based therapy could be considered for patients excluded from microdystrophin-based treatment due to mutation in this specific region.

Future clinical applications of gene editing will also have to consider the risks of additional immune response against the nucleases. In studies in which CRISPR-Cas gene editing has been applied to restore dystrophin expression, both cellular and humoral immune responses against Cas9 have been reported; however, it was not found when treating neonatal mice.⁵⁹ Nevertheless, further studies demonstrated a strong immune response in DMD dogs.⁶⁰ Moreover, DMD gene editing appeared to be more persistent in the heart than skeletal muscles of dystrophic mice, which could be related to the cycles of necrosis and regeneration typical of the latter tissue.⁶¹ Therefore, extensive in vivo animal studies are necessary to address these potential problems, including for utrophin upregulation therapy.

Our study also provides new insights into hiPSC-based models of DMD, which likely can be extended to other diseases as well. Almost all types of hiPSC-derived cells generated according to current protocols suffer from an immature phenotype, which can limit their application in research on the pathophysiological causes of relevant diseases. This is particularly the case for diseases that do not manifest early in life or result from age-related changes in tissue (e.g., deficiencies or deposit accumulation). hiPSC-derived cells may share the same physiological or metabolic characteristics and preventive mechanisms as those present in young/immature cells, making it more difficult to reveal in hiPSC-derived cells all the disease characteristics described in adults. This was exactly the complication we encountered in our hiPSC-derived CMs, which, unlike adult cells, still showed utrophin expression. In the future, it would be worth considering using hiPSC-derived cells for DMD studies (skeletal muscle cells, CMs, etc.) that are more mature or have additional utrophin knockout, as its presence may obscure the effects of dystrophin deficiency.

Taken together, our findings support the idea that utrophin can compensate for the loss of dystrophin in DMD hearts. This may have a number of implications for research and clinical practice, particularly in the application of utrophin upregulation as a potential therapy for DMD-associated cardiomyopathy, as well as for future in vitro models of DMD using hiPSC.

Study limitations

Although we evaluated the role of utrophin in human CMs in an in vitro model, our study does not provide the validation of the CRISPR-dCas-VP64 system in vivo. This would require the evaluation of new sgRNA sequences (specific to a given species) and other AAV vector serotypes, and thus, it would not be a good representation of the system’s performance and function in human CMs. Moreover, as hiPSC-CMs represent an immature, fetal-like phenotype, the role of utrophin in these cells may differ from that of adult CMs. Therefore, it would be important to examine the effect of utrophin upregulation in mature cells, although the application of the CRISPR-dCas9 system in this in vitro setup could be challenging. Despite these aspects, we believe that our work could be a starting point for understanding the role of utrophin in the human heart, as well as for further investigation of its therapeutic potential in DMD-associated cardiomyopathy.

Clinical perspectives

Our study demonstrates that the upregulation of utrophin in human dystrophin-deficient CMs improves their physiological functions, particularly those related to the transmission of electrical signals, which can reduce the risk of arrhythmias and can be clinically beneficial for patients suffering from DMD-associated cardiomyopathy. The CRISPR-dCas9-VP64 system utilized here efficiently upregulated utrophin expression in hiPSC-CMs, which potentially opens a new avenue of upregulating utrophin in humans. Importantly, the use of effective methods to deliver the system to specific tissue types (e.g., using the appropriate AAV serotype) may allow it to be applied in other DMD-relevant tissues such as skeletal muscle, in addition to CMs. Therefore, further studies should validate the therapeutic efficiency and safety of this approach in vivo, which could potentially benefit DMD patients regardless of the type of mutation.

Materials and methods

hiPSC culture

The generation of a control (under approval of the Institutional Review Board and Bioethical Committee, no. 122.6120.303.2016 and with informed consent, in accordance with the Declaration of Helsinki) and isogenic dystrophin-deficient hiPSC cell line (obtained by the CRISPR-Cas9-mediated deletion of DMD exon 50) has been described elsewhere.^23,62 hiPSCs were cultured in Geltrex-coated 12-well plate wells in Essential 8 (E8) medium (Thermo Fisher Scientific), which was refreshed daily. The cells were passaged using 0.5 mM EDTA, and 10 μM Rho-associated protein kinase (ROCK) inhibitor Y-27632 (Abcam) was added to the medium for the first 24 h after each passage. Cell culture was performed under standard conditions (37°C, 5% CO₂, and 20% O₂) and was regularly monitored for Mycoplasma infection.

Construction of a plasmid to introduce the mutation in the gene encoding utrophin

To introduce the mutation, three different sgRNAs targeting exon 11 of the UTRN gene (Table S1) were designed using the software program CHOPCHOP and were cloned into the pSpCas9(BB)-2A-Puro (PX459) V2.0 plasmid⁶³ (a gift from Feng Zhang, Addgene plasmid #62988; http://n2t.net/addgene:62988; RRID: Addgene_62988). Briefly, the plasmid was digested using the BpiI restriction enzyme (Thermo Fisher Scientific) and dephosphorylated at the 5′ ends with FastAP alkaline phosphatase (Thermo Fisher Scientific). Subsequently, it was separated on 1% agarose gel, and the appropriate band was excised from the gel and purified using the Zymoclean Gel DNA Recovery Kit (Zymo Research) according to the manufacturer’s instructions. Each pair of DNA oligos encoding the specific fragment of sgRNA was mixed, phosphorylated using T4 PNK (New England Biolabs), and annealed to obtain oligoduplexes. The digested plasmid and the annealed oligoduplexes were subjected to ligation with the Quick Ligation Kit (New England Biolabs) according to the manufacturer’s instructions. They were then used for the One Shot Stbl3 Chemically Competent E. coli bacteria (Thermo Fisher Scientific) transformation, followed by plating on LB agar plates containing ampicillin (Sigma-Aldrich) and further plasmid amplification and isolation using the Plasmid MIDI AX kit (A&A Biotechnology) according to the manufacturer’s instructions.

Nucleofection

The nucleofection of dystrophin-deficient hiPSC was performed using the Human Stem Cell Nucleofector Kit 1 (Lonza) according to the manufacturer’s protocol. Briefly, 500,000 cells were suspended in 100 μL of the Human Stem Cell Nucleofector Solution 1 and Supplement 1 mixture (at a ratio of 4.5:1). Subsequently, 5 μg of plasmid DNA was added, and nucleofection was performed on the Amaxa Nucleofector (Lonza) using the A-023 program. The cells were then seeded in a Geltrex-coated well in E8 medium supplemented with ROCK inhibitor. The next day, cells were selected by stimulation with 0.3 μg/mL of puromycin (Sigma-Aldrich) for 24 h; resistant cells were used in further procedures.

DNA isolation

DNA isolation was performed using the Genomic Mini kit (A&A Biotechnology) according to the manufacturer’s instructions. The DNA concentration and purity were determined using the NanoDrop 1000 spectrophotometer.

Surveyor nuclease assay

In the first stage of the assay, the PCR reaction was prepared, consisting of 20 ng of DNA isolated from hiPSCs after nucleofection (a mixture of cells or single clones) or control unedited hiPSCs, filled up to 10 μL with water, 1.25 μL of each primer (forward: 5′-GGTGTGAAGACAGGACTATGG-3′, reverse: 5′-CTATCAATTCCACCCTGTGAGCT-3′), and 12.5 μL KAPA2G Fast Genotyping Mix (Sigma-Aldrich) according to the reaction conditions described in the manufacturer’s instructions. Subsequently, 4.5 μL of the PCR product from the control sample and 4.5 μL of the product from the tested sample were mixed with 1.5 μL of celery juice extract (CJE) buffer (prepared according to the protocol described by Till et al.⁶⁴) and subjected to a heteroduplex formation reaction performed under the following conditions: 5 min at 95°C, then a temperature drop by 2°C/s to 85°C, followed by a temperature drop of 0.1°C/s to 25°C. The remaining volume of the PCR product was run on 2% agarose gel to confirm that the target sequence has been amplified (data not shown). Next, 1 μL of CJE containing mismatch-specific DNA endonuclease (prepared according to the protocol described by Till et al.⁶⁴) and 3.5 μL of water were added to the product of the heteroduplex formation reaction. The samples were incubated for 45 min at 45°C, and the reaction product was then analyzed on 2% agarose gel.

Cardiac differentiation

hiPSCs were differentiated into CMs according to the protocol described by Lian et al.⁶⁵ Briefly, 3 × 10⁴ hiPSCs were seeded in wells of a 24-well plate and cultured in E8 medium until they reached 90% confluency. At that time (day 0), the cells were stimulated for 24 h with the 12 μM small molecule GSK3 inhibitor CHIR99021 (Sigma-Aldrich) in RPMI-1640 medium (Biowest) complemented with B-27 Supplement Minus Insulin (Thermo Fisher Scientific). On day 3, the cells were stimulated with the 5 μM Wnt pathway inhibitor IWR-1 (Sigma-Aldrich) for 2 days and then refreshed with RPMI-1640 medium with B-27 Supplement Minus Insulin. From day 7, differentiating cells were cultured in RPMI-1640 medium supplemented with B-27 (Thermo Fisher Scientific). To increase the purity of the CM population and the differentiation yield, metabolic selection was used. This method takes advantage of the fact that CMs can use lactate as an energy source, thus eliminating other cells whose metabolism is dependent on glucose.^66,67 For this purpose, differentiating cells were cultured from day 10 to day 16 in RPMI glucose-depleted medium (Thermo Fisher Scientific) supplemented with 4 mM sodium DL-lactate (Sigma-Aldrich). After that, the cells were detached using TrypLE Select Enzyme (Thermo Fisher Scientific), harvested in RPMI-1640 medium supplemented with 20% fetal bovine serum (FBS) (Biowest), centrifuged at 200 × g for 5 min, and reseeded in Geltrex-coated wells in RPMI medium supplemented with B-27.

Construction of plasmid for utrophin activation

For the purpose of utrophin activation in cells, the CRISPR-dCas9-mediated transcriptional activation system was used based on the catalytically inactive Cas9 and VP64 activation domain. In this study, the pJEP304-pAAV-EFS-dSaCas9-VP64-pA plasmid (Addgene #113679⁶⁸) was used, as it enables the delivery of the editing system to CMs using AAV vectors. In the first step, the U6 promoter and guide RNA (gRNA) scaffold for the Staphylococcus aureus CRISPR-Cas9 system was cloned from the pX601-AAV-CMV::NLS-SaCas9-NLS-3xHA-bGHpA; U6::BsaI-sgRNA plasmid (Addgene #61591⁶³) backbone to the pJEP304-pAAV-EFS-dSaCas9-VP64-pA plasmid backbone. For this approach, the fragment of the sequence encoding the U6 promoter and gRNA was amplified by PCR using forward and reverse primers introducing the MluI restriction enzyme cleavage site at the ends of the targeted sequence. Subsequently, both the amplified fragment and the pJEP304-pAAV-EFS-dSaCas9-VP64-pA plasmid were digested using MluI restriction enzyme (New England Biolabs) and separated by electrophoresis on 2% and 1% agarose gel, respectively. Proper bands were excised from the gel, purified, and then ligated, and the resulting plasmid was used for bacteria transformation as described above.

Four pairs of sgRNA sequences targeting the utrophin A promoter region (sgRNA P1-P4) were designed (Table S2) and cloned into the pJEP304-pAAV-EFS-dSaCas9-VP64-pA plasmid with U6 promoter and gRNA. First, the plasmid was digested using the BsaI restriction enzyme (New England Biolabs) and dephosphorylated at the 5′ end with FastAP alkaline phosphatase. Subsequently, the plasmid was separated on 1% agarose gel, and the appropriate band was excised from the gel and purified using the Zymoclean Gel DNA Recovery Kit according to the manufacturer’s instructions. Each pair of DNA oligos encoding sgRNA was mixed, phosphorylated using T4 PNK, and annealed to obtain oligoduplexes. The digested plasmids and annealed oligoduplexes were subjected to a ligation reaction with the use of the Quick Ligation Kit according to the manufacturer’s instructions.

AAV production

AAV vectors were produced for the purpose of activating utrophin in hiPSC-CMs. AAV6 vectors were used, as they demonstrate high transduction efficiency in hiPSC-CMs.²⁸ In the study, a helper-free system consisting of two plasmids and human embryonic kidney 293-derived AAV293 packaging cells was used.

AAV293 cells were cultured in DMEM medium (Biowest) supplemented with 10% FBS and 1% penicillin/streptomycin (Thermo Fisher Scientific, 10,000 U/ml for both antibiotics). Cell passage was performed when the confluency reached ∼70% using trypsin-EDTA solution (Thermo Fisher Scientific) diluted five times. Cell culture was performed under standard conditions (37°C, 5% CO₂, and 20% O₂).

First, AAV293 cells were seeded in 15-cm² culture plates and grown until they reached ∼60% confluency. They were then transfected with 40 μg of AAV6 helper plasmid (pDGM6) (Addgene #110660) and 20 μg of pJEP304-pAAV-EFS-dSaCas9-VP64-pA plasmid (either encoding sgRNA or with no sgRNA sequence inserted) using 2.58 mg/mL (1:1 w/v) of polyethylenimine PEI MAX (Polysciences). After 72 h, the cells were scratched from the surface, collected, and centrifuged at 350 × g for 10 min at 4°C. The cell pellet was washed with PBS and reconstituted in PBS with Ca and Mg ions (Lonza) and then lysed by repeated freezing and thawing cycles in liquid nitrogen and a 37°C water bath. Subsequently, the lysates were digested with 50 U/ml of HS nuclease (MoBiTec) for 1 h at 37°C and centrifuged at 4,000 × g for 30 min at 4°C.

The AAV vectors were subjected to purification by ultracentrifugation at 300,000 × g (2 h, 18°C) on an iodixanol density gradient. Iodixanol dilutions (OptiPrep Density Gradient Medium, Sigma-Aldrich) were prepared in PBS with 0.5 mM MgCl₂ and 2.5 mM KCl in the following ratio: 20 mL of processed crude lysate, 6 mL of 15% iodixanol, 4.5 mL of 25% iodixanol, 4.5 mL of 40% iodixanol, and 4.5 mL of 54% iodixanol. After centrifugation, the 40% iodixanol fraction contained purified AAV particles. Therefore, it was collected and concentrated using Amicon Ultra-15 Centrifugal Filters (Merck Millipore).

For the purpose of titrating AAV viral genome copies, a small volume of vectors (5 μL) was taken for DNA isolation using phenol-chloroform extraction. Subsequently, quantitative PCR (qPCR) with the dCas9-binding TaqMan probe was performed. 10-fold serial dilutions of linearized plasmid DNA served as standards for generating a standard curve. The reaction contained 2 μL of template (extracted DNA/standard), 10 μL of TaqMan Gene Expression Master Mix (Thermo Fisher Scientific), 1 μL of probe (6-FAM AGGCGGCATAGAATCCAGAGAGTGA BHQ-1) at a final concentration of 20 nM, 1 μL of each primer (forward: 5′-AAGAGGCCAACGTGGAAA-3’; reverse: 5′- AGCAGGTTGTAGTCGAACAG-3′) at a final concentration of 100 nM, and 5 μL of nuclease-free water. The reaction was performed in a StepOne Plus Real-Time PCR thermocycler using the conditions listed in Table S3.

Transduction of hiPSC-CMs with AAV vectors

hiPSC-CMs were transduced with the AAV6 vectors with an MOI of 10,000 viral genomes per cell. The medium was changed 24 h after transduction, and all experiments were performed at least 6 days after transduction.

RNA isolation, reverse transcription, and qPCR

RNA was isolated according to Chomczynski and Sacchi’s method⁶⁹ using Fenozol (A&A Biotechnology). RNA concentration and purity were determined using the NanoDrop 1000 spectrophotometer (Thermo Fisher Scientific). Reverse transcription was performed using RevertAid Reverse Transcriptase polymerase (Thermo Fisher Scientific) according to the manufacturer’s protocol for 500 ng of RNA. The mixture for qPCR contained 7.5 μL of SYBR Green JumpStart Taq ReadyMix (Sigma-Aldrich), 0.75 μL of forward primers, 0.75 μL of reverse primers (working concentration of 0.5 μM), 3 μL of water, and 3 μL of complementary DNA (cDNA) (diluted 5 times after reverse transcription). The reaction was performed in the StepOne Plus Real-Time PCR (Applied Biosystems), as presented in Table S4. The sequences of the primers used in this study are provided in Table S5. The troponin T (TNNT) gene served as a housekeeping gene in the hiPSC-CMs qPCR analysis.

Protein isolation

The cells were washed twice with ion-free PBS and then lysed for 20 min in 1% Triton X-100 (BioShop Canada) (in the case of samples collected for western blot analysis) or 10% SDS (BioShop Canada) (in the case of samples collected for proteome analysis) in PBS supplemented with cOmplete Protease Inhibitor Cocktail (Sigma-Aldrich) on ice. Subsequently, cells were centrifuged at 14,000 × g for 20 min, (at 4°C when 1% Triton X-100 was used or room temperature [RT] when 10% SDS was used), and the supernatants were collected in separate tubes. Protein concentration was determined using a bicinchoninic acid assay (Sigma-Aldrich) according to the manufacturer’s instructions.

Proteomic analysis

Before the material was collected, the high cardiac differentiation efficiency (more than 80%) was confirmed by flow cytometric analysis of cardiac troponin T-positive cells for each sample. The samples were lysed in 6 M guanidine-hydrochloride (GuHCl), heated for 10 min at 95°C, and sonicated for 30 s. After dilution to 0.5 M GuHCl, the proteins were digested with LysC (1:200 enzyme:substrate ratio) and trypsin (1:100 enzyme:substrate ratio) overnight at 27°C. After clean-up with in-house SDB-RPS stage tips, the peptides were reconstituted in 2% acetonitrile (ACN) and 5% formic acid (FA). To separate the peptides, an in-house 30-cm fused silica emitter (75 μm diameter) was packed with 5 μm C18 Poroshell resin (Agilent) and applied to an Easy nLC1200 (Thermo Fisher Scientific) with the column temperature being maintained at 50°C through the integrated column oven. For solvent A, 0.1% FA was used, and for solvent B, 80% ACN with 0.1% FA was used. A 90-min segmented gradient of 4%–32% solvent B over 72 min, 32%–55% solvent B over 13 min, and 55%–95% over 2 min at a flow rate of 250 nL/min was applied to elute the peptides. They were measured with the Orbitrap Eclipse Tribrid (Thermo Fisher Scientific) equipped with the FAIMS.Pro interface (Thermo Fisher Scientific). Spectra were acquired through a data-independent acquisition (DIA) method with staggered windows with compensation voltage (CV) of −50, employing a duty cycle of 2 sets of DIA acquisitions shifted by a ½ isolation window from 400 to 900 m/z.

The raw MS data were analyzed using DIA-NN (1.7.2) and processed with an in-house R script using the libraries data.table, tidyr, magrittr, stringr, and dplyr. Statistical analysis was performed within Perseus (1.5.5.3). Two-sided t tests were performed to identify differentially expressed proteins between experimental conditions. The mass spectrometry proteomics data have been deposited in the ProteomeXchange Consortium via the PRIDE⁷⁰ partner repository with the dataset identifier PRIDE: PXD044407.

Gene set enrichment analysis (GSEA) was performed in R (v.4.1.2) using clusterProfiler (v.4.2.2) and the BH p-adjustment method. Batch effects estimation and correction were performed using the ComBat function from the sva (v.3.42.0) R library. The results were visualized using the ggplot2 (v.3.3.5), pathview (v.1.34.0), and enrichplot (v.1.14.2) libraries. The GSEA was prepared on a set of analyzed proteins sorted and filtered by |log2 fold change| >0.1.

Western blot

Protein in the amount of 20 μg in 20 μL of lysis buffer and 5 μL of loading buffer (8% SDS, 0.4% bromophenol blue, 40% glycerol, 200 mM Tris-HCl [pH 6.8]) was separated on 6% polyacrylamide gel by electrophoresis. Subsequently, the proteins were transferred onto nitrocellulose membranes (Bio-Rad) through overnight wet transfer carried out at 30 V. The membranes were then blocked in 5% non-fat milk in Tris-buffered saline with 0.1% Tween 20 (BioShop Canada) (TBST) buffer (blocking buffer) for 1 h at RT and incubated overnight at 4°C with monoclonal anti-utrophin (Santa Cruz Biotechnology, 8A4) or anti-α-tubulin antibody (Sigma-Aldrich) primary antibodies diluted in blocking buffer (1:500 and 1:1,000, respectively). The next day, the membranes were washed four times for 5 min in TBST buffer and incubated for 1 h at RT with horseradish peroxidase (HRP)-conjugated goat anti-mouse secondary antibodies (BD Pharmingen) diluted in blocking buffer (1:5,000). Afterward, the membranes were washed again four times for 5 min in TBST buffer, and chemiluminescence was detected using Immobilon Western Chemiluminescent HRP Substrate (Sigma-Aldrich) and the ChemiDoc Imaging System (Bio-Rad).

AFM measurements

For AFM measurements, 250,000 hiPSC-CMs were seeded onto 28-mm-diameter Geltrex-coated round cover slips in RPMI medium supplemented with B-27. 2 days later, the cells were mounted on a mechanical stage platform in HBSS buffer at 37°C; the nanomechanical parameters, elastic modulus, and contraction amplitude were measured using the NanoWizard 3 NanoScience AFM system (JPK Instruments) with the spherical probe (4.5 μm in diameter) (Novascan). The analysis of elastic modulus was performed using JPK Data Processing software, while the contraction amplitude was analyzed by MATLAB software.

Patch-clamp analysis

For patch-clamp analysis, 10,000 hiPSC-CMs were seeded onto 11-mm-diameter Geltrex-coated round cover slips in RPMI medium supplemented with B-27. 2 days later, the cells on the cover slips were placed under the Axioskop 2 FS microscope fitted with infrared differential interference contrast, in a perfusing system containing an extracellular Tyrode solution (140 mM NaCl, 5.4 mM KCl, 1.8 mM CaCl₂, 1 mM MgCl₂, 5.5 mM glucose, and 5 mM HEPES [pH = 7.4]) at 37°C. Whole-cell patch-clamp recordings of membrane potential were performed using borosilicate electrodes (7–9 MΩ, Sutter Instruments) filled with electrolyte solution (125 mM K-gluconate, 20 mM KCl, 5 mM NaCl, and 10 mM HEPES) and mounted onto an Ag/AgCl electrode. The signal was amplified using the SC 05LX amplifier (NPI), low-pass filtered at 3 kHz and digitized at 20 kHz.

All recordings were made in current clamp mode, using Signal software (Cambridge Electronic Design). The resting membrane potential was recorded at a holding current of 0 nA, whereas for the action potential measurements, each cell was manually brought down to −65 mV by negative current injection, from which 10 successive action potentials were evoked with a pacing of 1 Hz by short (10-ms) pulses of positive current (60–200 pA). These 10 traces were then averaged and analyzed for the AHP magnitude, the maximum upstroke velocity, and action potential duration at 20%, 50%, and 90% repolarization. A liquid junction potential of −15 mV was added to the recorded values.

Measurement of intracellular Ca²⁺ oscillations

For intracellular Ca²⁺ measurements, 250,000 hiPSC-CMs were seeded onto 32-mm-diameter Geltrex-coated round cover slips in RPMI medium supplemented with B-27. After 48 h, the cells were washed with PBS and loaded with the 1 μM fluorescent Ca²⁺ indicator Fluo-4 AM (Thermo Fisher Scientific) for 30 min at 37°C. Both loading and subsequent measurements of cytosolic calcium oscillations were performed in NaHEPES buffer, consisting of 140 mM NaCl, 4.7 mM KCl, 10 mM HEPES, 1 mM MgCl₂, and 10 mM glucose (pH 7.2), supplemented with 1 mM CaCl₂. Real-time spontaneous Ca²⁺ oscillations were recorded at 37°C in Fluo-4-loaded hiPSC-CMs in a flow chamber filled with NaHEPES-based extracellular solution using the Leica DMi8 fluorescence microscope. The signal was captured at a resolution of 384 × 288 (binning 5 × 5) with 0.4-s intervals between consecutive frames. Excitation was set at approximately 488 nm (LED light source), and green emission was collected at approximately 510 nm. Each group contained 246–268 cells and was analyzed on four independent cover slips in eight different fields of view (on average).

The data were analyzed by scoring the individual parameters in three categories, as presented in Table S6, and are shown as an average score for the individual parameter. The total score presented as the oscillations pathology is the average sum of all three parameters.

Immunofluorescent staining

The cells were fixed with 4% paraformaldehyde (PFA) (Santa Cruz Biotechnology) for 15 min at RT and were washed with PBS. Afterward, 0.1% Triton X-100 (BioShop Canada) was added for permeabilization (15 min at RT), and the cells were washed again with PBS. In order to block non-specific binding sites, cells were incubated for 1 h at RT with 3% bovine serum albumin in PBS (blocking buffer). Primary rabbit anti-DAG1 antibodies (ABclonal) diluted in blocking buffer to 1:100 were added for overnight incubation at 4°C. The next day, the cells were washed three times with PBS, and AF488 Donkey Anti-Rabbit IgG secondary antibodies (Thermo Fisher Scientific), diluted in PBS to 1:400, were added for 1 h at RT. Next, the cells were washed again three times with PBS, and the nuclei were counterstained with 0.2 μg/mL of 4′,6-diamidino-2-phenylindole (Sigma-Aldrich). Images were captured using the Carl Zeiss LSM-510 meta laser scanning confocal microscope.

Statistical analysis

All statistical analyses were performed using GraphPad Prism version 9.3.1 software (GraphPad Software). Statistical significance was evaluated using one-way ANOVA, followed by Tukey’s multiple comparison test for the comparison of several datasets unless otherwise stated. Grubb’s test was used to identify outliers. All data are presented as mean ± SD unless otherwise specified. Data with a p value <0.05 were considered statistically significant.

Data and code availability

The mass spectrometry proteomics data has been deposited in the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PRIDE: PXD044407.