This study was reviewed and approved by the Institutional Review Board of the University of South Florida (IRB: Pro00033948) with participants’ written informed consent. We recruited seven subjects who were members of families affected by DCM associated with LMNA mutations. Detailed clinical information is provided in Table 1. Three control induced pluripotent stem cell (iPSC) lines were purchased from ATCC (ATCC-1026, ATCC-1028, ATCC-1029) without LMNA mutation.

Peripheral blood mononuclear cells (PBMCs) were reprogrammed as previously described (Angulo et al., 2021; Argenziano et al., 2021; Yang et al., 2021a, b). Briefly, PBMCs were isolated from whole blood and transduced by Integration-free CytoTune^®-iPSC Sendai Reprogramming Kit (Thermo Fisher) following the manufacturer’s protocol. Colonies were purified by handpicking. iPSCs were cultured on Matrigel (Corning) coated 6- or 12-well plates in mTeSR plus medium (STEMCELL Technologies) and were dissociated with DPBS containing 0.5 mm EDTA solution (Invitrogen). Cells were passaged every 3–5 days with ROCK inhibitor Y27632 (10 μm; Sigma-Aldrich). iPSCs used for differentiation ranged from Passage 20 to Passage 40. Cells were maintained at 37°C, 95% air, 5% CO₂. Mycoplasma tests were routinely performed.

Patient-derived and control iPSC lines were differentiated into iCMs using STEMdiff^™ Cardiomyocyte Differentiation Kit (STEMCELL Technologies), as per the manufacturer’s protocol. Briefly, iPSCs were seeded on matrigel-coated 12-well plates and reached >95% confluency before starting differentiation. Cells were then treated with STEMdiff^™ Cardiomyocyte Differentiation Medium A containing matrigel (1:100) for 2 days, STEMdiff^™ Cardiomyocyte Differentiation Medium B for 2 days, STEMdiff^™ Cardiomyocyte Differentiation Medium C for 4 days, and STEMdiff^™ Cardiomyocyte Maintenance Medium. After 15 days of differentiation, cardiomyocytes were maintained in RPMI 1640 medium (Thermo Fisher) supplemented with B27 (Thermo Fisher) until day 30. Purification of cardiomyocytes was achieved by a metabolic-selection method as previously described (Sharma et al., 2015) with glucose starvation (RPMI-glucose + B27) for 5 days before functionality analysis. Fresh media was renewed every other day. Cardiomyocyte aggregates were dissociated using cardiomyocyte dissociation media (STEMCELL Technologies) for downstream experiments.

Cardiac fibroblasts were generated from iPSCs using the protocol described by the Wu group (Zhang et al., 2019). Briefly, iPSCs were seeded on Matrigel-coated 6-well plates and were cultured for 3–5 days until they reached 50–70% confluence when differentiation started (day 0). iPSCs were treated with 6 μm of CHIR99021 for 2 days, recovered in RPMI + B27-insulin for 24 h, and then treated with 5 μm of IWR1 (I0161, Sigma) for 2 days. On day 5, human induced pluripotent stem cell-derived cardiac progenitor cells (iPSC-CPCs) were switched to advanced DMEM medium (12,634,028, Gibco^®, Life Technologies) supplemented with 5 μm of CHIR99021 and 2 μm of retinoic acid (R2625, Sigma-Aldrich) for 3 days, and recovered in advanced DMEM for 4 days. Human iPSC-EPCs were treated with 10 μm of FGF2 (100-18B, PeproTech) and 10 μm of SB431542 (S1067, Selleck chemicals) in a fibroblast growth medium (116–500, Cell applications) for another 6 days. Cells can be maintained in the fibroblast growth medium and passed 4–5 times for subsequent studies.

For Alkaline Phosphatase staining, the iPSC colonies were fixed in 4% paraformaldehyde (w/v; Sigma-Aldrich) for 10 min and stained using Stemgent^® Alkaline Phosphatase Staining Kit II (REPROCELL). 50,000 cells/well were plated onto a 4-well chamber slide (NEST Scientific) and fixed with 4% paraformaldehyde for 10 min, permeabilized with 0.2% Triton X-100 (Sigma-Aldrich) for 10 min, followed by blocking in 1% BSA (Sigma-Aldrich), 22.52 mg/ml glycine (Fisher Scientific) in PBST (PBS + 0.1% Tween 20) for 30 min. Cells were stained with the primary antibody in 1% BSA overnight at 4°C, incubated with secondary antibody (Thermo Fisher) for 1 h at room temperature in the dark and mounted with VECTASHIELD HardSet Antifade Mounting Medium with DAPI (VECTOR Laboratories). Both primary and secondary antibodies were listed in Supplementary Table S1. Images were acquired with BZ-X800 Microscope (Keyence) or FluoView FV1000 Confocal Microscope (Olympus). Nuclear circularity was quantified with ImageJ software using the equation circularity = 4π*area/perimeter^2. Sarcomere organization was analyzed using TT power analyses with TTorg plugin in ImageJ (Pasqualin et al., 2015).

Cells were lysed using RIPA buffer (Thermo Fisher) complemented with protease and phosphatase inhibitors (Thermo Fisher). Lysates were sonicated in short bursts and centrifuged at 15,000 rpm for 15 min. Protein concentration was measured using Bradford assay (Thermo Fisher). 20–50 ug of total protein was loaded per sample and separated on a 4–15% SurePAGE gradient gel (GenScript). The proteins were transferred onto nitrocellulose membrane using Trans-Blot Turbo Transfer System (Bio-Rad). REVERT stain (Li-COR) was used to normalize for total protein loaded. Membranes were blocked with PBS Odyssey Blocking Buffer (Li-COR) for 1 h at room temperature and incubated overnight at 4°C with primary antibodies, followed by secondary antibodies for 1 h at room temperature. Both primary and secondary antibodies were listed in Supplementary Table S1. Images were obtained using an Odyssey FC or CLX imaging system.

The cell lysate was homogenized using the QiaShredder (Qiagen). RNA was isolated using RNeasy Mini Kit (Qiagen). The on-column DNase treatment was performed to remove the genomic DNA. cDNA was converted using the SuperScript VILO cDNA synthesis kit (Thermo Fisher). qPCR reaction was performed using Applied Biosystems TaqMan Fast Advanced Master Mix (Thermo Fisher) and run on the Step One Plus system (Applied Biosystems). Target gene expression levels were normalized to GAPDH mRNA levels and presented as relative to control cell lines in triplicate. GraphPad Prism was used for statistical analysis. The following TaqMan Assay probes (Thermo Fisher) were used: GAPDH (Hs02786624_g1), GJA1 (Hs00748445_s1), SCN5A (Hs00165693_m1), KCNH2 (Hs04234270_g1), KCNQ1 (Hs00165003_m1), ATP2A2 (Hs00544877_m1), CASQ2 (Hs00154286_m1). For Sanger sequencing, RNA of iPSCs was extracted and converted to cDNA, amplified with DreamTaq polymerase (Thermo Fisher) and purified by Monarch PCR & DNA cleanup kit (New England Biolabs). Samples were sent for Sanger sequencing (Genewiz).

Micro Electrode Array (MEA) 24-well plates (Axion Biosystems) were coated with 5 μl droplets of fibronectin (F1141, Sigma-Aldrich) at 50 μg/ml onto each electrode region and incubated for 1 h at 37°C, 5% CO₂. iCMs after metabolic selection at day 35 were seeded at 100,000 viable cells per well. For coculture, 90,000 iCMs at day 35 and 10,000 iCFs at passage 2 were mixed and plated on MEA per well. The cells were then incubated at 37°C/5% CO₂ for 2 weeks with media renewed every 2 days. Field potential and propagation were recorded for 10 min from spontaneously beating monolayers using a Maestro MEA system (Axion Biosystems) at 37°C. The field potential duration was corrected (FPDc) for beating rate according to Fridericia’s formula: FPDc = FPD/Beat period^1/3 (Asakura et al., 2015). Data from each cell line were collected from 3 wells and repeated three times. Three biological controls from a total of 27 wells were combined. Statistical analysis was performed using GraphPad Prism.

Intracellular calcium handling was studied using single-wavelength fluorescent calcium dye Fluo-4 AM (F14217, Thermo Fisher) diluted in anhydrous DMSO (Thermo Fisher) and imaged with BZ-X800 Microscope (Keyence). iCMs at day 35 were dissociated into single CMs onto Matrigel-coated 12-well plates. Four to 7 days post-dissociation, iCMs were loaded with 5 μm Fluo-4 AM in calcium-free Tyrode’s solution (NaCl 134 mm, KCl 5.4 mm, MgCl₂ 2 mm, HEPES 10 mm, Glucose 10 mm) for 15–20 min and washed with Tyrode’s solution for 15 min at 37° C before imaging in Tyrode’s solution containing 2 mm Ca²⁺. Calcium imaging was recorded for 1–2 min at each field of interest. The mean intensity of the calcium transients was analyzed by drawing a region of interest over the whole cells using ImageJ software. The background noise was subtracted before further processing. The Ca²⁺ levels are presented as values of ΔF/F0. Bradycardia is defined as a beating rate 2 standard deviations lower than the average of the controls.

The migration of human iCFs was determined using in vitro scratch assay (Liang et al., 2007). Three biological iCFs were used as controls. Briefly, after confluent monolayers of iCFs were formed, cells were scratched using a pipette tip (20 μl) to create a consistent gap width. Cells were then washed in PBS to remove residual cells in the wounded area. Images were captured using a transmitted-light microscope in identical areas at 0 h, 5 h, 10 h and 24 h post-scratch. The cell-free areas at each time point were measured by ImageJ software using the MRI wound healing tool (Suarez-Arnedo et al., 2020).

All experiments were repeated at least two times and data were expressed as mean ± standard errors of the means (S.E.M.). Statistical comparisons were determined using unpaired two-tailed Student’s t-test between two groups. Multiple comparison correction analysis was performed using one-way ANOVA followed by Tukey’s post-hoc test. Statistical significance was defined as a value of p < 0.05.

Blood samples obtained from seven subjects with DCM who were heterozygous for mutation in LMNA were used to generate iPSC clones. These missense point mutations in LMNA are M1I, R216C/R399H, R216C-male (R216C.m), R216C-female (R216C.f), R335Q, R377H and R541C. Table 1 illustrates demographic, and clinical information of the subjects. It is worthwhile noting that R216C.m and R216C.f are 2 unrelated subjects carrying the same mutation derived from different families. Genotyping by Sanger sequencing confirmed the presence of the specific mutation in LMNA at the expected site (Figure 1A). Three biological iPSC lines were used as controls and their absence of LMNA mutation was validated (data not shown). The verification of the pluripotent identity of R541C-iPSC has been previously published (Yang et al., 2021a). Likewise, all the cell lines showed positive staining of pluripotency markers including SOX2, TRA-1-60, SSEA4, OCT4 and alkaline phosphatase (Figure 1B). The characterization of one representative control line was shown. A normal karyotype was confirmed in all the mutant lines (Supplementary Figure S1A). PCR validated the presence of pluripotency markers including Oct4, Sox2, Klf4, Nanog and c-Myc (Supplementary Figure S1B). In vitro three germ-layer differentiation validated the iPSCs’ ability to differentiate into all three embryonic germ layers (Supplementary Figure S1C).

Patient-derived and control iPSC lines were differentiated towards iCMs using STEMdiff^™ Cardiomyocyte Differentiation Kit, maintained in RPMI 1640 + B27 media until day 30 and selected by glucose starvation (Figure 1C). iCMs had spontaneous contractions and expressed the cardiac-specific marker troponin T by immunostaining (Figure 1D).

To establish cardiac fibroblasts, all iPSC lines were subjected to a small molecule-directed differentiation protocol (Zhang et al., 2019; Figure 1E). On day 18 of differentiation, the cells showed a typical fibroblast morphology. The presence of fibroblast-specific marker vimentin was confirmed by immunolabeling in all lines (Figure 1F).

After we established iPSC-derived cardiomyocytes (iCMs) and cardiac fibroblasts carrying different LMNA mutations, we sought to compare their protein expression of lamin A/C. Three biological wildtype control cell lines were combined and averaged for analysis. Immunoblot indicated a significantly lower quantity of lamin A/C protein in all LMNA mutant iCMs except variant R216C.m compared to the wildtype iCMs (Figure 2A). Among all the variants of cardiac fibroblasts, lamin A/C expression was substantially declined in M1I but was significantly elevated in R377H and R541C relative to controls (Figure 2B).

ERK pathway hyperactivation has been implicated in the pathophysiological perturbation of cardiac laminopathies (Siu et al., 2012; Chatzifrangkeskou et al., 2018). We therefore evaluated ERK signaling cascade by calculating the ratio of phosphorylated ERK_1/2 (pERK_1/2) over total ERK_1/2 through immunoblotting in iCMs and iCFs for all the subjects. In agreement with previous findings, the LMNA-iCMs had significant upregulation of pERK/ERK relative to controls in patients R216C.f, R335Q, R377H and R541C (Figure 2C). Markedly, the presence of ERK pathway hyperactivation was more prevalent in LMNA-mutant iCFs, with only R335Q showing no difference compared to controls (Figure 2D).

Increased apoptosis has been observed as a possible mechanism of LMNA-associated pathophysiology (Siu et al., 2012; Lee et al., 2017). Immunoblotting was performed to quantify the expression of cleaved caspase-3, a critical executioner of apoptosis, in iCMs and iCFs. In iCMs, patient R216C/R399H, R216C.m, R216C.f and R377H exhibited a higher degree of apoptotic marker compared to other LMNA variants or controls (Figure 2E). In iCFs, by contrast, there was no elevation in cleaved caspase-3 compared to control lines and there was no significant difference among the iCF cell lines (Figure 2F).

These results suggest that cardiac fibroblasts might be more significantly involved in the pathobiology of LMNA-related DCM concerning the abnormally activated ERK pathway than previously recognized. Increased levels of apoptosis markers in LMNA-iCMs are consistent with previous evidence. Further, this data supports that different LMNA mutation locations may not affect the same molecular pathways equally.

Previous studies on mature adult CMs and iPSC-derived CMs from patients affected with other LMNA mutations have shown aberrant nuclear morphology (Siu et al., 2012; Lee et al., 2017). We did not observe significantly misshapen nuclei in our LMNA-iCMs matured for 35–40 days as measured by fluorescence microscopy (Figures 3A,C). The LMNA-iCFs however, demonstrated irregular LMNA distribution and nuclear envelope shape in immunostaining for Lamin A/C antibody. Vimentin was co-stained to confirm the fibroblasts’ identity (Figure 3B). We observed an obvious wide range of nuclear dysmorphisms associated with different LMNA mutations in iCFs, such as nuclear blebbing, nuclear invagination, and atypical lamin A/C distribution. The nuclear deformity and lamin A/C aggregation at the peripheral nucleus were especially dramatic in variants R335Q and R377H. Objective assessment of reduced nuclear circularity supported the nuclear shape irregularities in variant R216C/R399H, R216C.m, R216C.f, R335Q and R377H (Figure 3D). Additionally, quantification of the diverse abnormal nucleus further confirmed an increased percentage of lamina structure impairment in all the LMNA-iCFs (Figure 3E).

Sarcomere architecture in iCMs was assessed by immunostaining using the α-actinin antibody (Figure 4A). Various LMNA mutations resulted in altered myofilament structures, suggesting irregular sarcomere organization (Figure 4B). The number of cells with severe sarcomeric abnormalities was significantly increased in all DCM iPSC-CMs compared to control cell lines (Figure 4C).

To assess functional abnormalities in iCFs, a wound-healing assay was conducted to measure cell migration rate (Figure 5A). Quantitative data showed a significant reduction in the migration rate over time in LMNA-iCFs (Figure 5B). This reduction was most prominently observed at the 10-h timepoint after the scratch (Figure 5C).

We investigated the intracellular Ca²⁺-handling properties of the single dissociated iCMs by Fluo-4 calcium imaging. Besides regular Ca²⁺ transients (Figure 6A), LMNA-iCMs exhibited a wide variety of impaired calcium dynamics including alternans, slow beat rate, as well as more perturbed dysrhythmias such as early afterdepolarization, oscillations, and fibrillations (Figure 6A). Analysis of calcium transients demonstrated that LMNA mutants displayed a prominently higher frequency of these potentially proarrhythmic events (Figure 6B). Collectively, these findings indicate that dysregulation of intracellular calcium signalling exists at the cellular level and may provide an arrhythmogenic substrate for clinical disturbances in LMNA-DCM.

The electrophysiological phenotype in LMNA-iCMs was investigated by the generation of field potentials at the multi-cellular level with MEA. A representative control tracing of extracellular field potential duration (FPD) was shown (Figure 6C). Based on the baseline electrophysiological characteristics, the LMNA-iCM aggregates displayed increased beat period (diminished beating rates) and prolonged FPD compared to the aggregates from controls (Figures 6D,E). As for conduction properties, iCMs from M1I, R216C/R399H, R335Q, R377H and R541C displayed reduced conduction velocity, as well as a prolonged excitation-contraction delay (Figures 6F,G), compared to healthy control iPSC-CMs.

Given the contributions of ion channel expression to electrophysiology characteristics, we examined the mRNA levels of several ion channels genes. SCN5A, the coding gene for fast voltage-dependent sodium channels (Nav1.5), exhibited variably reduced transcripts among different LMNA subject iCMs (Figure 7A). This finding is consistent with altered beating rates and conduction velocities. Meanwhile, the extension of corrected FPD in M1I and R541C may be attributed to depressed mRNA levels of KCNH2 and KCNQ1 (Figures 7B,C) two important potassium channels responsible for the repolarization phase of the cardiac action potential.

To study the genes involved in the regulation of the contraction/relaxation cycle, we evaluated the genes involved in calcium signalling including ATP2A2, PLN (phospholamban) and CASQ2 (calsequestrin 2). ATP2A2 encodes an enzyme called sarcoplasmic/endoplasmic reticulum calcium-ATPase 2 that mediates translocation of calcium from the cytosol into the sarcoplasmic reticulum lumen. Phospholamban is a key regulator of cardiac contractility and modulates sarcoplasmic Ca²⁺ sequestration. Calsequestrin is a calcium-binding protein that acts as a calcium buffer within the sarcoplasmic reticulum. Interestingly, LMNA-M1I and -R377H iCMs showed a significant elevation of mRNA levels of ATP2A2, PLN and CASQ2 by qPCR (Figures 7D–F). These findings may imply the effect of calcium handling perturbations on electrophysiological properties.

Connexin 43 (Cx43), an important component of cardiac gap junctions, has been implicated in the pathogenesis of LMNA mutation (Sun et al., 2010; Chen et al., 2013; Macquart et al., 2019). GJA1, the gene encoding Cx43, exhibits reduced expression in iCMs from R335Q and R377H (Figure 7G). All mutant LMNA cell lines exhibit reduced GJA1 expression in iCFs relative to controls (Figure 7J). With regards to the expression of Cx43 protein, there were no significant differences observed in the mutant LMNA-iCMs in comparison to the control iCMs (Figures 7H,I). However, the protein abundance of Cx43 was markedly decreased in all the mutant LMNA-iCFs relative to the control iCFs (Figures 7K,L), suggesting that cell-cell communication in CFs may play an important role in LMNA-DCM pathophysiology.

To examine the potential crosstalk between iCMs and iCFs, we performed coculture experiments by combining 90% iCMs and 10% iCFs from the identical parent iPSC lines. As described previously, immunoblotting/qPCR showed perturbed Cx43 expressions in LMNA-iCFs more prominently than in LMNA-iCMs. We, therefore, hypothesized that iCFs coupling to iCMs via Cx43-mediated gap junctions could contribute to the electrical disturbances caused by LMNA mutations. We focused on three mutations (M1I, R335Q and R541C) and three biological controls. Immunofluorescent co-staining illustrated cardiac-specific marker α-actinin and Cx43 was performed on the cocultured cells. The fluorescence signal of Cx43 was markedly weaker in the LMNA mutant patients than the wild-type controls (Figure 8A). Of note, although Cx43 was mainly localized at the cell-cell contacts in control coculture, it was more confined to the cytoplasm in LMNA mutant cells. Next, we further examined the electrophysiology characteristics of the cocultured cells on MEA. Interestingly, compared to the controls, M1I-, R335Q- and R541C-LMNA mixture monolayer showed a trend towards exaggerated electrical perturbations over iCMs alone, including prolonged beat period and FPDc, slower conduction velocity accompanied by increased excitation-contraction delay (Figures 8B–E). These results suggest that the in vitro electrical response of LMNA mutant cardiomyocytes was enhanced by LMNA-iCFs, possibly due to the lower abundance of Cx43.