Modeling of lamin A/C mutation premature cardiac aging using patient-specific induced pluripotent stem cells

Altered nuclear architecture and enhanced electric-stimulation induced apoptosis in LMNA^R225X/WT dermal fibroblasts

We have established primary cultures of dermal fibroblasts from the proband (II-7) harboring the heterozygous R225X LMNA mutation and from a control individual (Figure 1). Compared with control dermal fibroblasts, LMNA^R225X/WT dermal fibroblasts had reduced expression of the lamin A/C proteins (Figure 2A). Ultra-structural analysis using electronic microscopy revealed dramatic morphological alterations suggestive of nuclear senescence in LMNA^R225X/WT dermal fibroblasts including focal loss of nuclear membrane, clustering of nuclear pores, large aggregates of highly condensed heterochromatin, bleb and micronucleus formation, and accumulation of mitochondria around the nuclear envelope (Figure 2B-I). Although these dysmorphic features of nuclear senescence, commonly seen in aging cells [26], could also be observed in the control fibroblasts with immunofluorescence analysis, they occurred in lower frequency (3.6 ± 3.6%, vs. 18.0 ± 1.8%, P<0.05) with less-severe morphological abnormalities in comparison with LMNA^R225X/WT dermal fibroblasts (Figure 3A-E).

Figure 1. Clinical phenotype and LMNA genotype
(A) The pedigree of the proband (Patient II-7) with cardiac laminopathy featuring atrial fibrillation, atrioventricular block, dilated cardiomyopathy, ventricular tachyarrhythmia and sudden cardiac death. Four-generation family had 9 affected members showing autosomal-dominant cardiac disease. (See also table 1)(Square = male; circle = female; slash = deceased). A nonsense LMNA mutation (R225X) segregates with atrial fibrillation, atrioventricular block, dilated cardiomyopathy and ventricular tachyarrhythmia (“+” = LMNA^R225X/WT, “-” = LMNA^WT/WT). (B) The sequence chromatogram of the mutant allele of LMNA. The mutation altering amino acid sequence (in red) predicts a premature stop codon. (C) The schematic of lamin A/C proteins. Alternative splicing results at 572 and 664 amino acid lamin C and lamin A proteins. The R225X mutation results in a premature stop codon in the α-helical rod domain, thus a truncated lamin protein.

Figure 2. LMNA^R225X/WT dermal fibroblasts showing nuclear defects and accelerated apoptosis upon electrical stimulation. (A) The expressions of lamin A/C proteins in control and LMNA^R225X/WT dermal fibroblasts with western blot analysis using anti-LMNA antibody targeting both N-terminus. (B-I) Electronic micrographs of nuclei of cultured dermal fibroblasts from controls and LMNA^R225X/WT: (B) Normal nuclear envelope in control dermal fibroblast;(C) Focal loss of the nuclear membrane in LMNA^R225X/WT dermal fibroblasts (arrow);(D) Clustering of nuclear pores;(E, F & G) Bleb and micronucleus formation; (H) Accumulation of mitochondria around the nuclear envelope; (H & I) Irregular shape nucleus, and condensed chromatin.

Figure 3. Effect of electrical stimulation on LMNA^R225X/WT dermal fibroblasts. (A-D) Immunofluorescence showing typical nuclear morphologies of control and LMNA^R225X/WT dermal fibroblasts before and after electrical stimulation. (E) Effects of electrical stimulation on the percentage of control and LMNA^R225X/WT dermal fibroblasts with nuclear blebs. (F) Effects of electrical stimulation and U0126 (ERK1/2 pathway blocker) on the apoptosis of control and LMNA^R225X/WT dermal fibroblasts. The application of U0126, a highly selective blocker of the MEK1 pathway, nearly completely abolished the apoptotic effects of electrical stimulation on LMNA^R225X/WT dermal fibroblasts. (G-H) Electrical stimulation-activates mitogen-activated protein kinases (MAPK) pathway in iPSC-CM: (G) MEK1 and (H) extracellular signal-regulated protein kinases 1 and 2 (ERK1/2) in LMNA^R225X/WT dermal fibroblasts, but not in control fibroblasts. *p- value <0.05.

Previous study in a lamin A-haploinsufficiency mouse model of DCM had suggested that reduction of lamin A/C proteins render premature apoptosis, particularly in electrically active cardiomyocytes [18.] Therefore, we applied electrical stimulation to cultured dermal fibroblasts to simulate an in-vivo cardiac electric field, and scored the fractions of abnormal nuclei present in control and LMNA^R225X/WT dermal fibroblasts (Figure 3A-D). After electrical stimulation (at 6.6 V/cm, 1 Hz for 3 days), the percentage of cell with nuclear senescence (82 ± 9.7% vs. 47.4 ± 7.1%, P<0.05) and apoptosis (27.5 ± 4.1% vs. 6.5 ± 3.4%, P<0.05) were significantly higher in LMNA^R225X/WT dermal fibroblasts as compared with control dermal fibroblasts (Figure 3E-F).

Mechanisms of electrical stimulation-induced apoptosis

We further explored the possible stress-related signaling pathway involved in electrical stimulation-induced apoptosis in LMNA^R225X/WT dermal fibroblasts. Upon electrical stimulation, MEK1 and its downstream candidate, extracellular signal-regulated protein kinases 1 and 2 (ERK1/2), were activated only in LMNA^R225X/WT dermal fibroblasts, but not in control fibroblasts (Figure 3G-H). The application of U0126, a highly selective blocker of the MEK1-ERK1/2 pathway, nearly completely abolished the apoptotic effects of electrical stimulation on LMNA^R225X/WT dermal fibroblasts (Figure 3F). These data suggest that increased nuclear senescence and apoptosis induced by electrical stimulation in LMNA^R225X/WT dermal fibroblasts was mediated via the MEK1 pathway.

Generation of LMNA^R225X/WT iPSC and cardiac differentiation

Despite the near-ubiquitous LMNA expression in most somatic cells, DCM is the most prominent phenotype of the R225X mutation. Therefore, we generated iPSCs from dermal fibroblasts isolated from the proband (II-7) to derive cardiomyocytes for recapitulating the cardiac phenotype of LMNA-related familial DCM in-vitro.

Altogether 4 clones of iPSCs were generated from the proband; however, chromosomal translocation was detected in 1 of these clones. As a result, the subsequent experiments were from the remaining 3 clones, which we did not detect any significant clone-to-clone variation. The genomic LMNA was sequenced in all iPSCs, and the expected mutation, R225X in LMNA was detected from the iPSCs generated from the dermal fibroblasts of the proband but not from those of the control. Besides, as we previously generated another iPSC line derived from DCM patient bearing frame-shift mutated LMNA [27], it was used as another LMNA halploinsuffiency model in the present study.

After cardiac differentiation, spontaneously beating outgrowths appeared approximately 14 to 21 days in embryoid bodies from both LMNA^R225X/WT, LMNA^{Frameshift/WT} and control iPSC lines (Figure 4A-C and Supplementary video I-III). The beating outgrowths were micro-surgically dissected out from the EBs and were dissociated into single cell clusters (Figure 4D-F, and Supplementary video IV-VI). Immunofluorescence analysis revealed the typical pattern of a cardiac-specific protein, α-actinin in beating cells from both LMNA^R225X/WT, LMNA^{Frameshift/WT} and control iPSC lines, confirming their cardiac identity (Figure 4G-I). Morphologically, there were no observable difference between the cardiomyocytes derived from the three iPSC lines, except the more prominent nuclear alterations of nuclei in LMNA^R225X/WT andLMNA^{Frameshift/WT} iPSC-CMs (Figure 4J-L).

Figure 4. Cardiomyocytes derived from LMNA^R225X/WT, LMNA^{Frameshift/WT} and LMNA^WT/WT iPSCs. (A, B & C) Beating embryoid bodies derived from LMNA^R225X/WT, LMNA^{Frameshift/WT} and LMNA^WT/WT iPSCs;(D, E & F) After micro-surgical dissection, spontaneously beating cell clusters were plated onto glass coverslips. Videos of these beating embryoid bodies and clusters were available in supplemental materials. (G, H & I) Immunofluorescence co-staining showing the expression of cardiac specific marker, α-actinin and lamin A/C in these cells, (J, K & L) Electronic microscopy of nucleus of cardiomyocytes derived from LMNA^R225X/WT, LMNA^{Frameshift/WT} and LMNA^WT/WT iPSC line. The black arrows indicate the loss of nuclear envelope in the LMNA^R225X/WT iPSC-derived cardiomyocytes.

Furthermore, we characterized the electrophysiological properties of LMNA^R225X/WT iPSC-CMs using whole-cell patch clamp experiments. Figure 4I depicts representative single action potentials recorded from cardiomyocytes derived from LMNA^R225X/WT and control iPSCs. According to action potential recording, cells were then classified into atrial- and ventricular-like cardiomyocytes. There were no significant differences in the proportion of atrial-like and ventricular-like cardiomyocytes between LMNA^R225X/WT iPSCs and control. Furthermore, there were no significant differences in action potential duration at 50% and 90% repolarization (APD90) between the two groups (Data not shown).

Susceptibility of iPSC-CMs to electrical stimulation

Because electrical stimulation increased the fraction of cells with nuclear senescence as well as apoptosis in LMNA^R225X/WT dermal fibroblasts, we further tested whether LMNA^R225X/WT iPSC-CMs exhibited similar susceptibility to electrical stimulation. At baseline, there was no significant difference in the percentage of cardiomyocytes with nuclear senescence between the control and LMNA^R225X/WT iPSC-CMs (3.1 ± 0.7 % vs. 6.8 ± 2.2%) (Figure 5A-B). Upon electrical stimulation, the percentage of nuclear senescence in LMNA^R225X/WT and LMNA^{Frameshift/WT} iPSC-CMs markedly increased to 48.2 ±4.0% (P<0.005 vs. baseline) and 52.3 ± 7.9% (P<0.05 vs. baseline) respectively and was significantly higher than those of the control iPSC-CMs (8.6±2.5%, P<0.01 in both mutated LMNA group; Figure 5B). More importantly, electrical stimulation resulted in a remarkably higher percentage of apoptotic cells in both LMNA^R225X/WT and LMNA^Framshift/WT iPSC-CMs compared with control (13.8±2.9% and 11.2±0.8% vs. 6.3±1.2%; n=3 P<0.05) as determined by APO-BrdU TUNEL assay quantitatively (Figure 6F and G). Co-staining of cardiac marker with TUNEL reaction further qualitatively verify the apoptotic events happened in LMNA^R225X/WT and LMNA^Framshift/WT iPSC-CMs upon electrical field stimulation.

Figure 5. Electrical stimulation inducing nuclear abnormality in cardiomyocytes derived from LMNA^R225X/WT &LMNA^{Frameshift/WT} iPSCs. (A) Representative immunofluorescence staining of cardiomyocytes derived from LMNA^WT/WT iPSCs, LMNA^R225X/WT iPSCs, LMNA^{Frameshift/WT} iPSCs, LMNA^WT/WT iPSCs treated with mock shRNA, and LMNA^WT/WT iPSCs treated with shLMNA at baseline and after electrical stimulation. (B) The percentage of cardiomyocytes showed nuclear blebs.

Figure 6. Electrical stimulation inducing apoptosis in cardiomyocytes derived from LMNA^R225X/WT &LMNA^{Frameshift/WT} iPSCs. Representative TUNEL assay and co-immunofluorescence staining of alpha-actinin in cardiomyocytes derived from (A)LMNA^WT/WT iPSCs. (B)LMNA^R225X/WT iPSCs. (C)LMNA^{Frameshift/WT} iPSCs. (D)LMNA^WT/WT iPSCs treated with mock shRNA. (E)LMNA^WT/WT iPSCs treated with shLMNA at baseline and after electrical stimulation. (F, G) Quantification of apoptotic cardiac differentiated iPSCs by APO-BrdU TUNEL assay. The percentage of cardiomyocytes with apoptosis was determined by FACS analysis by FL-1 positive gating. Unpaired t-test was performed between treatment and baseline n=3-5, *p-value<0.05.

To further verify whether lamin insufficiency contributes to the enhanced susceptibility of cardio-myocytes to electrical stimulation, we altered the expression of LMNA in control hiPSC-CMs by shRNA-mediated down-regulation (~60% as revealed by western blotting by probing of N-terminal LMNA A/C antibodies, data not shown). shRNA-treated control iPSC-CMs exhibited significantly higher percentage of nuclear senescence at baseline and after electrical stimulation (p<0.05) (Figure 5B). Likewise, shRNA mediated down-regulation of LMNA resulted in a significantly higher percentage of apoptotic cells than in control iPSC-CMs after electric stimulation (Figure 6E).

Intriguingly, the application of U0126 and selumetinib, AZD6244 (an anti-cancer drug in phase I-II clinical trial) to block MEK1 pathway in LMNA^R225X/WT and LMNA^Framshift/WT iPSC-CMs and shRNA treated control iPSC-CMs showed a trend that attenuated or even completely abolished the apoptotic effects of field electric stimulation on these lamin-deficient cardiomyocytes as seen in the dermal fibroblast (Figure 3 and 6) whereas 10 μM of U0126 significantly rescued the rate of apoptosis mediated by electrical stress (11.2 ± 0.8% vs. 4.4 ± 1.4%; n=3, P<0.05). On the other hand, although it has recently demonstrated that rapamycin reverses progeric phenotype in laminopathy, the addition of rapamycin (0.68 mM) did not alleviate the electrical stimulation mediated apoptosis of mutant LMNA iPSC-CMs (Supplementary Figure 3).

Clinical history and genetic phenotype

A 56 year-old Chinese man (II-7) presented to us with atrial fibrillation and cardioembolic stroke. He had history of progressive conduction system disease, evolving from 1^st degree to 3^rd degree atrioventricular block, and required permanent cardiac pacemaker implantation at age of 49 years (Table 1, and Figure 1A). Subsequently, he developed progressive heart failure and transthoracic echocardiogram revealed dilated right and left ventricles with impaired left ventricular ejection fraction of 35%. Then he underwent upgrading of his pacemaker to cardiac resynchronization therapy with defibrillator, and had received multiple device therapy for ventricular tachyarrhythmias during follow-up. Sequencing of the LMNA gene revealed a heterozygous single base exchange (672G→T) in exon 4, resulting in an R225X nonsense mutation (truncation in the coil 1B domain of both lamin A and C proteins), previously known to be associated with familial DCM (Figure 1B and IC)[43-45]. Indeed, he had a strong family history of DCM, conduction system disease and sudden death spanning 3 generations (Figure 1A). His father (I-1) suffered sudden death at age of 59, and one of his sisters (II-2) suffered from atrial fibrillation, heart failure and 3^rd degree atrioventricular block with pacemaker implanted, who died suddenly and had documented ventricular fibrillation from her pacemaker memory. In addition, his two other sisters (II-5 and II-9) both had atrial fibrillation and 3rd degree atrioventricular block with permanent pacemaker implantation at 51 years and 48 years respectively. Subsequent screening of members of his family revealed asymptomatic 1^st to 2^nd degree atrioventricular block in 3 younger, apparently healthy family members (III-2, III-3, and III-4), in whom 24-hour ECG monitoring detected non-sustained ventricular tachyarrhythmia in two of them. Cardiac magnetic resonance imaging revealed patchy fibrosis over right ventricular myocardium (III-3 and III-4). Genetic testing showed that these family members had the same heterozygous LMNA mutation, confirming autosomal dominant inheritance in this family.

Mitogen-Activated Protein Kinase (MAPK) phosphorylation analysis

Total cell lysate obtained from mutant and control primary fibroblast underwent electric stimulation from five individual experiment was subjected to MAPK pathway phosphorylation analysis using the Milliplex MAP 10plex MAPK/SAPK signaling kit (Millipore, St. Charles, Missouri, USA) according to manufacturer instructions.[46] Bio-Plex Manager software version 4.1.1 was used for data acquisition and analysis. Each lysate was measured in duplicated wells and the results were present as mean with standard error.

Generation of disease-specific induced pluripotent stem cells

For the generation of iPSCs, we recruited the proband (II-7) (LMNA^R225X/WT) and another patient with frame-shift-mutated LMNA (contained a GCCA insertion at base 50 in lamin A/C gene, LMNA ^{Frameshift/WT})along with one healthy age- and sex-matched control subject (LMNA^WT/WT). The male patient (45 years old) bearing framshift-mutated LMNA was chosen also suffered from DCM and conduction system defectas previously described [27]. The study protocol of procurement of human tissue for the generation of iPSCs was approved by the local Institutional Review Board and was registered at the Clinical Trial Center, the University of Hong Kong (HKCTR-725, http://www.hkclinicaltrials.com).

After obtaining Informed consents from all participants, skin biopsies were taken, minced and then plated in 6-well culture dishes with 10% FBS medium as previously described[47]. Dermal fibroblasts growing out from the skin tissues were expanded and transduced with lentiviruses encoding OCT4, SOX2, KLF4, and c-MYC. Putative iPSC clusters were typically observed 14-21 days after lentiviral transduction, which were manually dissected and expanded in new Matrigel™-coated 6-well plates (Supplementary Figure 1). The authenticity of the human iPSCs was confirmed with the expression of a panel of pluripotent markers, transgene silencing, OCT4 promoter demethylation, and teratoma formation after inoculation into severe combined immunodeficiency mice (Supplementary Figure 1 and Result section in Supplementary Appendix). All the stem cell characterization of LMNA^{Frameshift/WT} iPSC has been reported previously [27].

In vitro cardiac differentiation and characterization of cardiomyocytes

To induce cardiac differentiation, undifferentiated iPSCs were co-cultured with endoderm-like cells as described [48-51]. Briefly, undifferentiated iPSCs were dissociated into clumps using dispase (Invitrogen, CA, USA) and cultured in suspension using low attachment plates to form embryoid bodies. Embryoid bodies were transferred onto irradiated endoderm-like cells for co-culture. Beating outgrowths from iPSC embryoid bodies were micro-surgically dissected 20 days after induction of cardiac differentiation, followed by enzymatic dissociation into single isolated cardiomyocytes for further experiments. Rescued effect of electrical field-induced apoptosis of iPSC-CMs was studied by using mitogen-activated protein kinase kinase 1(MEK1)-inhibitor, 10 μM U0126 (dissolved in DMSO from Toricis), or 25 nM selumetinib, AZD6244 (dissolved in DMSO from Selleck Chemicals), an anti-cancer agent at phase I-II clinical trials; or 0.68 mM rapamycin (dissolved in DMSO from Sigma),[40] a specific inhibitor of the mTORC1 pathway. Standard immunofluorescence staining of alpha-actinin and electrophysiology analysis by patch clamping were performed to confirm their cardiac phenotypes [52-54].

shRNA mediated knockdown of LMNA

To verify the functional effect of LMNA in cardiomyocytes, the 18-nucleotide shRNA sequences designed to knockdown LMNA were tested in control iPSC-CM. shRNA against the common region of lamin A and C RNA position 609-627 from the start codon was cloned into pAd-GW/U6 vectors, and pAd-GFP was used as control. Adenovirus for shRNA against LMNA and control GFP vectors were produced using ViraPower Adenoviral Expression Kit (Invitrogen, CA, USA) and purified using Add-N-Pure Viral Purification kit (Applied Biological Materials Inc., Richmond, BC, Canada) according to manufacturer's instruction respectively. Viral titer was determined by AdEasy Viral titer kit (Stratagene, La Jolla, CA). A MOI of 10 was used for virus transduction. A knockdown efficiency of 30-40% of lamin A and C was observed.

Electrical field stimulation

To mimic the microenvironment of cardiomyocytes, iPSC-CMs were subjected to electrical stimulation. Cells were seeded on 13-mm glass cover slips (Nunc A/S, Rockilde, Denmark) and mounted inside a 6-well plate filled with corresponding medium. Electrical stimulation was then delivered to the cultured cell with carbon electrodes using an eight-channel C-Pace cell culture stimulator (Ion-Optics Co., Milton, MA, USA) at 6.6 V/cm, 1 Hz, 2 ms with alternating polarity for 4 hours [31]. Total number of cells was counted using standard cytometry by two blinded individuals before and after electrical stimulation.

Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay for apoptosis

TUNEL assay was performed using In Situ Cell Death Detection Kit, Fluorescein (Roche Applied Sciences, Mannheim, Germany) according to manufacturer's protocol. Cells were grown on glass coverslips, and after assigned treatment, cells were fixed and permeabilized with 0.1% Triton X-100 in 0.1% sodium citrate for 2 minutes on ice. Cells were then incubated in TUNEL reaction mixture at 37°C for 60 minutes in a humidified dark chamber. Coverslips were mounted onto glycerol-based mountant. Images were acquired using Carl Zeiss fluorescence microscope with AxioVision 6.0 software (Zeiss GmbH, Gottingen, Germany).

Detection of apoptotic cells upon electrical stimulation was further quantified by the APO-BrdU TUNEL Assay Kit (Molecular Probes, Inc, Eugene, OR). The dissociated cells were fixed in 1% (w/v) paraformaldehyde on ice for 15 mins and then washed twice with DPBS at 300 × g for 5 mins. The fixed cells were stored in ice-cold 70% (v/v) ethanol at −20 ° C before nicked ends labeling reaction. Before incubation in TUNEL labeling solution for an hour at 37 ° C (10 μl reaction buffer, 0.75 μl TdT enzyme, 8 μl of BrdU and 31.25 μl dH₂O), the cells were washed with wash buffer. After further washing by rinse buffer, the cells were stained with the Alexa Fluor^® 488 dye-labeled anti-BrdU antibodies for half an hour. The propidium iodide/RNase A staining buffer was finally added to the cell prior to flow cytometry analysis. The positively labeled cells (green signal), a population with DNA fragmentation, were counted as % of apoptotic cells.

Statistical Analysis

Data are expressed as mean ± SEM. Statistic analysis was performed with the unpaired student t-test paired sample t-test as appropriate. Calculations were performed using SPSS software (version 12.0). A P-value < 0.05 was considered statistically significant.