Systematic in vivo candidate evaluation uncovers therapeutic targets for LMNA dilated cardiomyopathy and risk of Lamin A toxicity

Abstract

Background

Dilated cardiomyopathy (DCM) is a severe, non-ischemic heart disease which ultimately results in heart failure (HF). Decades of research on DCM have revealed diverse aetiologies. Among them, familial DCM is the major form of DCM, with pathogenic variants in LMNA being the second most common form of autosomal dominant DCM. LMNA DCM is a multifactorial and complex disease with no specific treatment thus far. Many studies have demonstrated that perturbing candidates related to various dysregulated pathways ameliorate LMNA DCM. However, it is unknown whether these candidates could serve as potential therapeutic targets especially in long term efficacy.

Methods

We evaluated 14 potential candidates including Lmna gene products (Lamin A and Lamin C), key signaling pathways (Tgfβ/Smad, mTor and Fgf/Mapk), calcium handling, proliferation regulators and modifiers of LINC complex function in a cardiac specific Lmna DCM model. Positive candidates for improved cardiac function were further assessed by survival analysis. Suppressive roles and mechanisms of these candidates in ameliorating Lmna DCM were dissected by comparing marker gene expression, Tgfβ signaling pathway activation, fibrosis, inflammation, proliferation and DNA damage. Furthermore, transcriptome profiling compared the differences between Lamin A and Lamin C treatment.

Results

Cardiac function was restored by several positive candidates (Smad3, Yy1, Bmp7, Ctgf, aYAP1, Sun1, Lamin A, and Lamin C), which significantly correlated with suppression of HF/fibrosis marker expression and cardiac fibrosis in Lmna DCM. Lamin C or Sun1 shRNA administration achieved consistent, prolonged survival which highly correlated with reduced heart inflammation and DNA damage. Importantly, Lamin A treatment improved but could not reproduce long term survival, and Lamin A administration to healthy hearts itself induced DCM. Mechanistically, we identified this lapse as caused by a dose-dependent toxicity of Lamin A, which was independent from its maturation.

Conclusions

In vivo candidate evaluation revealed that supplementation of Lamin C or knockdown of Sun1 significantly suppressed Lmna DCM and achieve prolonged survival. Conversely, Lamin A supplementation did not rescue long term survival and may impart detrimental cardiotoxicity risk. This study highlights a potential of advancing Lamin C and Sun1 as therapeutic targets for the treatment of LMNA DCM.

Graphical Abstract

Highlights

• After evaluation of 14 potential candidates in a cardiac-specific Lmna DCM model, we demonstrated that Smad3 shRNA, Yy1, combination of Bmp7 and Ctgf (Bmp7-Ctgf shRNA), Yap1, Sun1 shRNA, Lamin A, and Lamin C improved cardiac function. Sun1 shRNA and Lamin C particularly prolonged a long-term survival.

• We uncovered that inflammation and DNA damage markers were among the top list of highly correlated markers in addition to traditional HF and fibrosis markers, suggesting these additional markers are important to evaluate the candidates for the treatment efficacy of LMNA DCM.

• The study revealed that treating Lmna DCM with Lamin A did not work as expected and had toxic effects. The toxicity was found to be dose-dependent and not caused by prelamin A processing.

Background

Dilated cardiomyopathy (DCM) is often caused by genetic pathogenesis, with an estimated prevalence up to 1: 250 [1]. The DCM gene panel containing a list of over 100 genes provides a comprehensive genetic evaluation for patients with a personal or family history of hereditary DCM. Among them, the LMNA gene encoding Lamin A and Lamin C is the second most frequently mutated DCM gene [2, 3]. Lamin A is initially expressed as prelamin A and undergoes a series of post-translational modifications to form mature Lamin A [4]. Lamin C arises from alternative splicing. Lamin A and C are identical through their N- termini but diverge at the C-terminus with Lamin C being the shorter isoform, terminating in 6 unique amino acids. Furthermore, as Lamin C does not have the CAAX motif located at the C-terminal end of prelamin A, it lacks a post-translational modification unique to Lamin A.

It is believed that defective function of LMNA leads to DCM wherein mechanistic traits and phenotypes are shared in global and especially cardiac specific knockout mouse models [5,6,7]. Molecular mechanisms underlying LMNA DCM have been intensively studied. Currently, there are three main hypotheses for molecular mechanisms underlying LMNA related DCM, including dysregulated signaling pathways, lamina-chromatin interactions and lamina-cytoskeleton links [8,9,10]. Accordingly, many potential candidates have been shown to suppress LMNA DCM. Among them, suppression of mTor, Mapk, Pdgf signaling pathways, inhibition of BET bromodomain, or disruption of lamina-cytoskeleton links have been proposed to suppress or ameliorate LMNA DCM by using various models including global/cardiac specific knockouts, LMNA mutants as well as patient specific iPS derived cardiomyocytes [6, 7, 11,12,13]. However, it is unknown whether these candidates can effect comparable long term survival, nor have efforts been made to evaluate and compare their efficacy in suppression of LMNA DCM. Recently, we produced a cardiac specific knockdown model for LMNA DCM (designated as Lmna DCM) which is comparable to cardiac specific Lmna knockouts [14]. Importantly, this new model does not require intensive crossing to generate compound genetically modified mice.

Autosomal recessive diseases have been successfully treated with recombinant adeno-associated virus (rAAV) mediated gene replacement, exemplified by approval of rAAV9-SMN gene therapy for spinal muscular atrophy (SMA), the leading genetic cause of infant mortality [15]. However, a gene therapy approach for heart failure (HF) was not successful at phase 2 clinical trial due to lack of efficacy, which highlights the importance of prior candidate evaluation [16, 17]. Recently, AAV vector toxicity also emerged in some clinical trials especially at high dosage of virus administration needed for sufficient gene expression and virus distribution [18]. Besides toxicity of AAV vectors, the gene product itself should also be evaluated as, for example, ectopic expression of the EGFP reporter induced DCM in a sensitive mouse strain [19]. Apart from recessive diseases, dominant cardiovascular diseases such as cardiomyopathy were also assessed by gene therapy. Proof of principle studies have been attempted by targeting the causal genes of cardiomyopathy including MHC, MYPBC3 and MYL2 [20,21,22]. Here, to identify potential targets for the treatment of LMNA DCM, we evaluated the efficacy of 14 candidates by modulating their gene expression using rAAV9 in Lmna DCM. Lamin A and Lamin C produced by the Lmna gene were also included in the assessment since cardiac Lamin A partially rescued global Lmna knockouts [23].

Methods

Animal protocols

Animal procedures were performed in accordance with the Singapore National Advisory Committee for Laboratory Animal Research guidelines. All mice were housed in animal care facilities and studied using protocols approved by the Institutional Animal Care and Use Committee (IACUC) of National University of Singapore or of the Biological Resource Centre, Agency for Science, Technology and Research. The methods used in the study conformed to the Guidelines on the Care and Use of Animals for Scientific Purposes (NACLAR, Singapore, 2004) as well as the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication, 8th Edition, 2011). Male C57BL/6JINV (Jax) mice were used in this study. Lmna cardiac-specific knockout mice were previously described [7]. A power calculation was performed to estimate sample size required to demonstrate significant improvement (error alpha 0.05, power 0.8). Details of virus injection is performed as previously described [14]. For heart harvesting, each mouse was euthanized by cervical dislocation under deep anaesthesia with 5% isoflurane and the heart was exposed by opening of the chest. 15% KCl was then injected into the inferior vena cava to achieve asystole at diastole, and the heart was rapidly isolated. Half of the apex was isolated and immersed it in RNALater (Qiagen, 76104) for RNA extraction. The other half of the apex was snap frozen in liquid nitrogen for protein extraction. For histology experiments, the remaining section of the heart was fixed in 4% paraformaldehyde for 24 h and subsequently embedded in paraffin. A dose of 2.0E13 vg/kg was injection for Lmna DCM and 1.0E13 vg/kg was injected for candidates.

Echocardiogram (Echo)

To measure the cardiac dimensions and function of mice after virus transduction, echocardiography was conducted by using a 40 MHz-550S probe on the VisualSonics Vevo 2100 machine or on a Prospect T1 ultrasound systems from S-Sharp Inc. During the echo and analysis, the researchers were blinded to the animal group allocation and identities. To obtain accurate measurements, LV tracings were averaged from at least three consecutive heartbeats of M-mode, and LVDD (LV diastolic dimensions), LVWT (LV posterior wall thickness), EF (ejection fraction), and FS (fractional shortening) were obtained from short axis images.

Cell culture and transfection

Transfection of shRNA constructs and other plasmids were done in HEK293T cells (ATCC CRL-1573™, RRID: CVCL_0063). Two transfection methods, PEI (Polysciences. Inc, 24765-2) and Lipofectamine 3000 (Invitrogen L3000015), were used to transfect the cells based on the instructions provided by the manufacturers.

Recombinant adeno-associated virus (rAAV) production, purification and titration

To produce rAAVs, HEK293T cells were transiently triple transfected with rAAV viral vector with gene, helper plasmid pAdΔF6, and plasmid pAAV2/9 (Penn Vector Core). The viruses were harvested three days post-transfection, and purified using Optiprep density gradient medium (Sigma, D-1556) before being stored at − 80 °C. To determine virus titration, a forward primer (gataaaagcagtctgggctttcaca) and a reverse primer (gagcccatataagcccaagctattg) were used to target the rAAV cTnT promoter region, and quantified by qPCR to determine physical titers.

Vector construction

For shRNA candidates, primers designed to target 21 base-pair gene-specific regions (Invitrogen) were inserted at the miR-155 backbone of the AAV-cTnT-EGFP vector. The sequences of shRNAs are as follows:

Lmna shRNA	agtctcgaatccgcattgaca
LacZ shRNA	aaatcgctgatttgtgtagtc
Raptor shRNA	attacagcaagaatgaaggct
Tgfb1 shRNA	ttcctaaagtcaatgtacagc
Yap1 shRNA	tagttccgatccctttcttaa
Mapk14 shRNA	tccactgtctggttatagtgc
Smad2 shRNA	aagagcagcaaattcttggtt
Smad3 shRNA	aatgccagcagggaagttagt
Ctgf shRNA	cctgtcaagtttgagctttct
Sun1 shRNA	tttggaatgtgttccatggtg

For vectors used in gain of function studies, EGFP in AAV-cTnT-EGFP vector was replaced by Serca, Fgf16, Bmp7, Yy1, Lamin A or C (RNAi resistant forms with three silent mutations at the target region). DNSUN1 was obtained by gene synthesis and comprises a human serum albumin signal peptide, the last 453 amino acids of human Sun1, and the KDEL ER-Golgi retrieval signal. DNSUN1 was cloned into a WPRE-containing AAV vector from Addgene (Plasmid #105921, RRID: Addgene_105921) where the CBh promoter was replaced by a cTnT promoter and intron from AAV-cTnT-EGFP. AAV-cTnT-3Flag-hYAP1 S127A (aYAP1) was from Addgene (Plasmid #86558, RRID: Addgene_86558). The cloning primers are as follows:

Serca-F	ggggctagcatggagaacgctcacacaaagaccg
Serca-R	gggggtaccttactccagtattgcgggttgttcc
Fgf16-F	atagctagcatggcggaggtcgggggc
Fgf16-R	tatgcggccgcttacctatagcggaagagatctctg
Bmp7-F	ccggctagcatgcacgtgcgctcgctgcgcg
Bmp7-R	ccgggtaccctagtggcagccacaggcccggac
Yy1-F	gggcaattgatggcctcgggcgacaccctctacat
Yy1-R	gggggtacctcactggttgtttttggctttagcg
KASH-F	atagctagcatgcgagccttcctgttccggatcctc
KASH-R	tatgcggccgctcagagtggaggaggaccgttggta
Lamin A-F	cccgctagcatggagaccccgtcacagcg
Lamin A-R	cccggtaccttacatgatgctgcagttctgggagc
Lamin C-F	cccgctagcatggagaccccgtcacagcg
Lamin C-R	cccggtacctcagcggcggctgccactca

RNAseq library preparation and next generation sequencing

To establish RNAseq libraries, total RNA was extracted from apex of male mice (n = 3 per group). RNA library was prepared using NEBNext® Ultra™ II Directional RNA library prep kit for Illumina (NEB #E7760) according to NEB’s protocol. Libraries were sequenced using the Illumina NovaSeq 6000 sequencing system and paired-end 150 bp reads were generated for analysis. Raw sequencing reads were filtered to remove reads with adapter contamination, those with uncertain and low-quality nucleotides to obtain clean reads. RNAseq data were aligned to GRCm38/mm10 reference genome by using HISAT2 [24]. Following that, a reference-based approached was used to assemble the mapped reads of each sample using StringTie [25]. To count the reads numbers mapped to each gene, FeatureCounts [26] was used. The FPKM (Fragments Per Kilobase of transcript per Million mapped reads) method was used to calculate the expression of each gene. Differentially expressed genes were selected on the basis of adjusted p value (adjP). RNA-seq data was deposit (GSE240745). Potential genetic candidates (adj P < 0.001) were identified from Lmna DCM control group compared to Control, Lamin C or Lamin A treated group. TBtools was used togenerate the volcano plot [27]. The genes that were expressed differently were submitted to Morpheus, where they were subject to Hierarchical clustering and displayed in a heat-map that was color-coded. The Hallmark gene sets were evaluated using Gene Set Enrichment Analysis (GSEA, Broad Institute).

Quantitative real-time PCR (qPCR)

The levels of transcription were measured using qPCR. Following RNA extraction using Trizol, cDNA was synthesized using Pure-NA™ First Strand cDNA Synthesis Kit (Research Instrument, KR01-100). All qPCR was carried out by KAPA SYBR Fast qPCR Master Mix kit (KAPA Biosystems, KR0389) and relative expression levels were quantified using ΔCT. Transcription data were normalized to Ctcf expression. All qPCR primers are listed as follows:

Primer	Forward	Reverse
Nppa	tttcaagaacctgctagaccacctg	gcttttcaagagggcagatctatcg
Nppb	agtcctagccagtctccagagcaat	cgaaggactctttttgggtgttctt
Myh7	agcattctcctgctgtttcctt	tgagccttggattctcaaacg
Col1a1	gagcctgagtcagcagattgagaac	cctgtctccatgttgcagtagacct
Col1a2	acccttctcactcctgaaggctcta	tatgagttcttcgctggggtgttta
Tgfb1	gaaggacctgggttggaagtggatc	tgtgttggttgtagagggcaaggac
Smad 2	ctctccggctgaactgtctcctact	tccgagtttgatgggtctgtgaagc
Smad 3	tccgatgtccccagcacacaataac	ttccggttgacattggacagtaggc
Atp2a2	tggtgctgaaaatctccttgcctgt	cataatgagcagcacaaacggccag
Fgf16	aactggtacaacacctatgcctcca	catggagggcaacttagaaggatct
Mapk14	agctgtgaacgaagactgtgagctc	atgatgcagcccacggaccaaatat
Sun1	tatccagaaggagctggaagaaacc	tctctaatagccactcgagggaacc
Bmp7	agaatcgctccaagacgccaaagaa	ctctccctcacagtagtaggcagca
Ctgf	acacctaaaatcgccaagcctgtca	aatggcaggcacaggtcttgatgaa
Yy1	cggggaataagaagtgggagcagaa	caggagggagtttcttgcctgtcat
Yap1	gaccctcgttttgccatga	attgttctcaattcctgagac
Raptor	tgagtgtcaatggagatgtgcgctt	ccgttgtagatggctgtgaactggt
KASH	gccttgtacccatgtcagagaaaga	gttacgtctcgagcatacagccttc
Ctcf	atgtcacaccttacctttgcctgaa	ccttcctgctgttcttcctcaaaat
Lmna	gaggctcttctcaactccaaggaag	ctgtagcctgttctcagcatccact

Histological and immunostaining analysis

Heart sample collection and staining were as described previously [14]. Quantification of fibrosis was calculated as the red-stained areas relative to total ventricular area, using NIS-Elements software (Nikon). For antigen retrieval of αSMA, Ki67, Iba1, Yap1, Yy1, cTnI and pH2AX, samples were boiled in citrate buffer (pH 6.0). As for CD3, Sun1 and Nesprin1, samples were boiled in EDTA (pH 9.0) for antigen retrieval. The following primary antibodies were used: αSMA (Sigma, A5228, RRID:AB_262054), Ki67 (Abcam, ab15580, RRID:AB_443209), CD3 (Dako A045201, RRID:AB_2335677), Phospho-Histone H2A.X (Ser139) (Cell Signaling Technology, #9718, RRID:AB_2118009), Iba1 (Wako Laboratory Chemicals (019-19741, RRID:AB_839504), Sun1 (gift from Dr. Brian Burke), YAP1 (Cell Signaling Technology, #14074, RRID:AB_2650491), YY1 (Abcam, ab38422, RRID:AB_778962), Lamin A/C (Cell Signaling Technology, 2032, RRID:AB_10694918), Nesprin1 (Abcam, ab192234, RRID_AB_2917992), PCM1 (Sigma, HPA023374), RRID:AB_1855073 and cTnI (Abcam, ab8295, RRID:AB_306445). DAPI was used for nuclear staining (ThermoFisher, D1306, RRID: AB_2629482). To check for signal specificity, species-specific secondary antibody only controls were used. For each heart sample, investigators unaware of group identities counted the total number of positive signals from three cross-sections, and this data was then standardized to the total number of nuclei or total ventricular area.

Western blots

Heart tissue collection and western blot were performed as previously described [14]. The following primary antibodies were used: phospho-Smad2 (Ser465/Ser467) (Cell Signaling Technology, 18338, RRID:AB_2798798), phospho-Smad3 (Abcam, ab52903, RRID:AB_882596), Smad 2/3 (Cell Signaling Technology, 5678, RRID:AB_10693547), phospho-p38 (Cell Signaling Technology,9211, RRID:AB_331641), p38 (Cell Signaling Technology, 9212, RRID:AB_330713), phospho-p70S6K (Cell Signaling Technology, 9234, RRID:AB_2269803), p70S6K, (Cell Signaling Technology, 9209, RRID:AB_2269804), phospho-mTOR (Cell Signaling Technology, 5536, RRID:AB_10691552), mTOR (Cell Signaling Technology, 2972, RRID:AB_330978), Lamin A/C (Cell Signaling Technology, 2032, RRID:AB_10694918), Ctgf (Santa Cruz Biotechnology, Inc, sc-365970, RRID:AB_10917259), Yy1 (Thermofsher, PA5-29171, RRID:AB_778962), Bmp7 (Proteintech group, Inc, 12221-1-AP), YAP1 (Cell Signaling Technology, 14074, RRID:AB_2650491), Sun1 (Abcam, ab103021, RRID:AB_2890137), Nesprin1 (Abcam, ab192234, RRID_AB_2917992) and Lamin B1 (Abcam, ab16048, RRID:AB_443298). The secondary antibody used were Donkey anti-Rabbit IgG (H + L) Highly Cross-Adsorbed Secondary Antibody, HRP (Thermofisher, A16035, RRID: AB_2534709) and m-IgGκ BP-HRP (Santa Cruz Biotechnology, sc-516102, RRID: AB_2687626). Protein levels on the blots were detected using the enhanced chemiluminescence system (GE Healthcare, RPN2106) according to the manufacturer’s instructions. Protein band intensity was quantified using Image J (NIH, 1.52e RRID:SCR_003070) and protein levels were normalized to Lamin B1 for mouse hearts.

Statistical analyses

Prism 10.0.2 (GraphPad Software, La Jolla, California, RRID:SCR_002798) was used to conduct statistical analysis. Normality of sample distribution was assessed by the Shapiro–Wilk normality test. For data with two groups, a two-tailed, unpaired T-test with Welch correction was performed for data that followed a normal Gaussian distribution whereas Mann–Whitney test was performed for data that depart from normality. Brown-Forsythe and Welch ANOVA test with Dunnett’s T3 multiple comparisons were performed for multiple groups. Survival curve was generated according to the Kaplan–Meier method and P value was calculated using Log- rank (Mantel- Cox) test. Quantitative data were shown as mean ± SD. ns, non-significant.

Results

Lmna gene products suppress Lmna DCM

To assess whether Lmna gene products supress Lmna DCM, we administered Lamin A via rAAV9 upon the development of Lmna DCM mice. HF markers Nppa and Nppb, fibrosis markers Col1a1 and Col1a2, fibrosis as well as cardiac function were significantly improved by Lamin A compared to EGFP control (Fig. 1a, b, Table 1). To further assess whether improvement of cardiac function can be translated into long term protection in Lmna DCM, we performed survival analysis for Lamin A treatment. Unexpectedly, the protective effect of Lamin A on Lmna DCM lapsed starting at 2 months (Fig. 1c). Supplementation with Lamin C, the other product produced by the Lmna gene, preserved heart function, supressed HF/fibrosis marker genes, and imparted long-term survival of Lmna DCM mice (Fig. 1d–f, Table 1). The difference between Lamin A and Lamin C treatment was not due to excessive overexpression, as Lamin A and Lamin C dose and detection in cardiac tissue were comparable (Additional file 1: Fig. S1).

Fig. 1

Effect of Lamin A and Lamin C on Lmna DCM in mice. (a and d) Experimental timeline showing timepoints of virus injection and echocardiogram. Sirius Red (SR) and H&E staining of paraffin heart sections and quantifications of Lmna DCM mice supplemented with control, Lamin A or Lamin C. Quantification of myocardial fibrosis of SR sections, n = 5, Brown-Forsythe and Welch ANOVA test with Dunnett’s T3 correction. For complete heart images: magnification = 4 × , scale bar = 1000 µm; for enlarged images: magnification = 20 × , scale bar = 100 µm. (b and e) Quantitative real-time PCR analyses of Nppa, Nppb, Col1a1 and Col1a2 in Lmna DCM mice supplemented with control, Lamin A or Lamin C, n = 5, Brown-Forsythe and Welch ANOVA test with Dunnett’s T3 correction. (c and f) Survival curve of Lmna DCM mice supplemented with Lamin A (Red) or Lamin C (Red) compared to Lmna DCM mice (Black) and Ctrl (Blue), n = 10, Log- rank (Mantel- Cox) test

Table 1 Effect of potential candidates on Lmna DCM in mice

Difference between Lamin A and Lamin C for the treatment of Lmna DCM

To dissect the difference between Lamin A and Lamin C treatment, we compared the transcriptional profiles of various groups including control, Lmna DCM, Lamin A treated, and Lamin C treated by RNAseq. Hierarchical clustering compared differentially expressed genes (DEGs, P < 0.001) among the different groups. The Lamin C treated group was closely clustered with the control group, in which up to 80% of DEGs in Lmna DCM were reversed by Lamin C (Fig. 2a). In contrast, less than 40% of DEGs in Lmna DCM were reversed by Lamin A (Fig. 2b). We further analysed the list of upregulated DEGs that cannot be reversed by Lamin A using GSEA (Fig. 2c, d). Significantly upregulated gene sets identified “Epithelial Mesenchymal transition” as a top enriched set, which is associated with fibrosis. Consistent with this, cardiac fibrosis in Lmna DCM was not reduced to a similar level by Lamin A as compared to Lamin C (Fig. 1a, d). Upregulated gene sets were also enriched in “Inflammatory response”, “Allograft rejection” and “TNFa Signaling Via NFkB”, suggesting cardiac inflammation remains activated in Lamin A treated hearts.

Fig. 2

RNAseq analysis of Lmna DCM with Lamin A or Lamin C upregulation. (a and b) Heat map representing color-coded expression level of genes that are significantly changed in Lmna DCM mice supplemented with (a) Lamin C or (b) Lamin A. Venn diagram of overlap between upregulated (red) and downregulated (green) genes in Lmna DCM with (a) Lamin C or (b) Lamin A upregulation compared to Lmna DCM mice. Mice were harvested four weeks after transduction, n = 3. (c) Venn diagram of overlap between Lmna DCM upregulated genes in Lamin A or Lamin C treatment. (d) Hallmark signature depicting top 10 dysregulated pathways of upregulated differentially expressed genes (DEGs) which cannot be reversed by Lamin A as designated by GSEA, arranged by NES

Both Lamin A and its mature form induce DCM in wildtype mice

The impaired protection of Lamin A in Lmna DCM and detection of retained fibrosis and inflammatory gene expression by RNAseq analysis, hinted at a detrimental impact of Lamin A following its administration. The specific detrimental effect of Lamin A supplementation was tested by application of an equivalent Lamin A dose to wildtype mice. Although EF was not significantly altered, HF and fibrosis markers, as well as fibrosis were modestly upregulated by Lamin A (Additional file 2: Table S1, Additional file 1: Fig. S2). Since AAV mediated gene delivery is among the most common tools in current gene therapy, it is crucial to dissect the mechanism of this rAAV-Lamin A induced toxicity. An increased dose of Lamin A resulted in significantly impaired cardiac function, enlarged heart chamber of the left ventricle (LV) and reduced LV wall thickness coupled with interstitial fibrosis (Fig. 3a, Additional file 2: Table S2). HF markers Nppa and Nppb as well as fibrosis markers Col1a1 and Col1a2 were significantly upregulated by Lamin A compared to EGFP control (Fig. 3b, c). Consistent with cardiac fibrosis, Lamin A transduced hearts showed a significant upregulation of p-Smad2 and the myofibroblast marker αSMA (Fig. 3d, e). Additionally, Lamin A induced cardiac inflammation is indicated by increased presence of infiltrating Iba-1 + macrophage and CD3 + T cells (Fig. 3f, g). Taken together, these results demonstrate that supplementation of Lamin A by rAAV is detrimental and induces DCM in a dose dependent manner (designated as Lamin A DCM). We noted that accumulation of prelamin A was observed in Lmna transduced mice, suggesting that prelamin A is not properly processed upon upregulation of Lamin A by rAAV (Fig. 3d). We hypothesised that this accumulation of prelamin A in CMs could drive detrimental heart function.

Fig. 3

Increased dose of Lamin A in wildtype mice. (a) Experimental timeline showing timepoints of virus injection and echocardiogram. SR and H&E staining of paraffin heart Section 4 weeks after transduction of EGFP control or Lamin A. Quantification of myocardial fibrosis of SR sections, virus dose, 2.0E + 13 vg/kg, n = 5, two-tailed, unpaired T-test with Welch correction. For complete heart images: magnification = 4 × , scale bar = 1000 µm; for enlarged images: magnification = 20 × , scale bar = 100 µm. (b, c) Quantitative real-time PCR analyses of Nppa, Nppb, Col1a1 and Col1a2 in mice transduced with EGFP control or Lamin A, n = 5, two-tailed, unpaired T-test with Welch correction and Mann–Whitney test. (d) Western blot and quantitative analysis of phosphorylated Smad2 (p-Smad2) protein levels in mouse heart tissue of mice transduced with EGFP control or Lamin A, n = 5, two-tailed, unpaired T-test with Welch correction. (e–g) Paraffin heart sections (left) and quantifications (right) of (e) αSMA (red), (f) Iba-1 (red), (g) CD3 (red), cTnI (green) and DAPI (blue) positive cells in mice after transduction of EGFP control or Lamin A, n = 5, two-tailed, unpaired T-test with Welch correction and Mann–Whitney test, scale bar = 50 µm

Lamin A, initially expressed as prelamin A, undergoes farnesylation and cleavage by Zmpste24 to generate mature Lamin A [28, 29]. The complete processing of farnesylated prelamin A to mature Lamin A is crucial as the accumulation of farnesyl-prelamin A is toxic. Importantly, mice expressing prelamin A only in the absence of Lamin C developed DCM, suggesting that prelamin A accumulation and farneysylation contribute to cardiomyopathy development [30]. To test this possibility, we constructed an AAV vector expressing a mature form of Lamin A, bypassing prelamin A synthesis and farnesylation steps. Transduction of mature Lamin A also led to impaired cardiac contraction, enlarged LV chamber size, reduced LV wall thickness and interstitial fibrosis (Additional file 1: Fig. S3a, Additional file 2: Table S3). Furthermore, we observed a significant increase in HF/fibrosis markers in the mature Lamin A DCM mice compared to controls (Additional file 1: Fig. S3b–e). Consistently, these mice showed significant upregulation of fibrotic mediators as well as cardiac inflammation markers (Additional file 1: Fig. S3f–g). Therefore, upregulation of either pre- or mature Lamin A induced DCM pathogenesis, suggesting that these detrimental effect is caused by the Lamin A product alone and not by impaired prelamin A processing.

Lamin A/C preserves the integrity of the nucleus and DNA. DNA damage is commonly observed in LMNA DCM and Lmna animal models [31,32,33,34]. To further dissect the molecular mechanisms of Lamin A upregulation, we examined DNA damage marker γH2AX in the hearts of Lamin A DCM. Interestingly, γH2AX was significantly increased in the CMs of Lamin A DCM, indicating that Lamin A upregulation induces DNA breaks in mouse hearts and contributes to the detrimental effects of Lamin A on heart function (Additional file 1: Fig. S4).

Selection of potential candidates for the treatment of Lmna DCM

LMNA gene replacement is able to treat LMNA DCM caused by haploinsufficiency through compensation for the loss of functional LMNA. However, LMNA DCM caused by dominant negative mutations are not amenable to gene replacement. To evaluate other potential candidates, we analysed our generated RNAseq data for Lmna DCM. Significant dysregulated genes were detected by Gene Set Enrichment Analysis (GSEA, Broad Institute) (Fig. 4a, b). Dysregulated gene sets identified hallmark signature associated signaling pathways including KRas, Tgfb and mTorc1, many of which have been shown to be deregulated in various LMNA related models. To validate these pathways in Lmna DCM, we examined the key factors in these pathways in heart tissues by Western blot. Consistent with previous studies, we observed that Tgfβ and mTor signaling pathways were dysregulated in Lmna DCM [11, 35] (Fig. 4c). However, the p-p38/Mapk pathway did not change significantly. Based on our pathway analysis/validation as well as previous studies, we selected to evaluate candidates including Tgfb1, Smad2, Smad3 for Tgfβ signaling pathway, Raptor for mTor signaling pathway, Fgf16 and Mapk14 for Mapk signaling pathway, Serca2 for calcium handling and Sun1 for the linker of nucleoskeleton and cytoskeleton (LINC). Our recently studied genes such as Bmp7, Ctgf, Yy1, Yap1 were also included in the comparison [14, 36] (Fig. 4d). We further examined their gene expression in heart tissue of Lmna DCM. The expression of Tgfb1, Fgf16, Bmp7, Ctgf, Serca2, Smad2, Raptor and Yap1 were dysregulated in Lmna DCM. Other signalling downstream molecules or transcriptional regulators such as Smad3, Mapk14, Yy1 and Sun1, were not transcriptionally regulated in Lmna DCM (Additional file 1: Fig. S5).

Fig. 4

Candidate selection for the treatment of Lmna DCM. (a) Volcano plot showing differentially expressed genes in heart samples of Lmna DCM mice. Downregulated genes are reflected in green and upregulated genes in red. (b) Hallmark signature depicting top 10 dysregulated pathways in Lmna DCM mice as designated by GSEA, arranged by Normalised Enrichment Score (NES). (c) Western blot and quantitative analysis of phospho-Smad2, phospho-Smad3, phospho-p38, phospho-p70S6K and phospho-mTOR protein levels in mouse heart tissues of control or Lmna DCM mice, n = 5, two-tailed, unpaired T-test with Welch correction. (d) Schematic diagram of selected candidates in the cytoplasm and nucleus of cardiomyocyte as well as extracellular matrix (ECM)

Positive candidates extend survival of Lmna DCM

To evaluate the functions of these selected candidates in Lmna DCM, we modulated their expression by rAAV9. The knockdown efficacy in vitro was assessed by a 2-color system developed previously [37] (Additional file 1: Fig. S6). At least 75 percent knockdown was achieved for loss-of-function candidates. All candidates were assessed in Lmna DCM. The effects of each candidate on cardiac function were compared using echocardiography (Fig. 5a, Additional file 2: Table S4). These candidates alone did not affect cardiac function and survival of wildtype mice in the observed time window (Additional file 2: Table S5, Additional file 1: Fig. S7). Among them, Smad3 shRNA, Yy1, combination of Bmp7 and Ctgf shRNA (Bmp7-Ctgf shRNA), aYAP1, Sun1 shRNA, Lamin A, and Lamin C, can significantly improve the ejection fraction (EF) in Lmna DCM. Their upregulation or knockdown efficacy in vivo was validated by Western Blot (Additional file 1: Fig. S8). Selected candidates including Smad2, Tgfb1, Mapk14, Smad3, Yy1, Yap1 and Sun1 were included to assess the effect of dosage on gene expression. Echocardiography results showed that increased dosage of the selected candidates did not change their effect on cardiac function (Additional file 2: Table S6). In addition, the knockdown and overexpression efficacy of Sun1, Yap1 and Yy1 at different dosages was examined by immunostaining (Additional file 1: Fig. S9 and S10). To further assess whether improvement of cardiac function can be translated into long term protection in Lmna DCM, we performed survival analysis for all positive candidates (Fig. 5b). Two negative candidate shRNAs Tgfb1 and Mapk14 were included for comparison. Consistently, knockdown of these two negative candidates did not prolong survival for Lmna DCM. Although cardiac EF in Lmna DCM was significantly improved to ~ 35% upon Smad3 shRNA treatment, the median survival was not significantly extended, suggesting improved cardiac function does not always extend to prolonged animal survival. Yy1, aYAP1 or Bmp7-Ctgf shRNA extended the median survival of Lmna DCM by a modest 15–30%. Strikingly, Sun1 shRNA extended survival of Lmna DCM by at least tenfold. Importantly, cardiac function in Lmna DCM was preserved by Sun1 shRNA, similar to Lamin C (Additional file 1: Fig. S11). As a key component of the Linker of Nucleoskeleton and Cytoskeleton (LINC) complex, Sun1 plays a pivotal role in linking the nuclear lamina to the cytoskeleton by bridging nuclear lamins with cytoskeleton-interacting KASH domain proteins [9].

Fig. 5

Positive candidates extend the survival of Lmna DCM. (a) Comparison of EF (%) assessed by echocardiography in control and Lmna DCM mice supplemented with potential candidates compared to control (P value in green) or Lmna DCM (P value in black). Echocardiography was performed 3 weeks after transduction, n = 5, Brown-Forsythe and Welch ANOVA test with Dunnett’s T3 correction. (b) Survival curve of Lmna DCM mice supplemented with selected candidates (Red) compared to Lmna DCM mice (Black), n = 10, Log- rank (Mantel- Cox) test

To dissect the roles of Sun1 in Lmna DCM, we first assessed Sun1 expression and protein levels. No significant change of Sun1 was detected in Lmna DCM compared to that in control mice (Additional file 1: Fig. S5, S9a and S12a). Similar to Lamin A/C, Sun1 signal was detected around the nuclei in control cardiac cells (Additional file 1: Fig. S12b). Interestingly, we observed the number of cardiomyocyte (CM) nuclei with abnormal Sun1 distribution was significantly increased in Lmna DCM compared to those in the controls (Additional file 1: Fig. S12b), suggesting that Lmna knockdown perturbs Sun1 distribution. Analysis of the nuclear shape of CM with abnormally distributed Sun1 revealed a significant increase of nuclear protrusion when compared to CM with normal Sun1 distribution (Additional file 1: Fig. S12c). Importantly, the number of nuclei with abnormal nuclear protrusion in Lmna DCM mice was significantly reduced by Sun1 shRNA (Additional file 1: Fig. S12c), indicating that knockdown of Sun1 preserved nuclear integrity of the CMs in Lmna DCM.

Nesprins interact with Sun1 via their conserved KASH domain [38]. To further elucidate the role of LINC complex in Lmna DCM, we investigated Nesprins and found no significant changes in Syne1 (coding for Nesprin-1) expression and protein levels (Additional file 1: Fig. S13a, b). Sun1 and Nesprin-1 were co-localized in the nucleus (Additional file 1: Fig. S13c). Interestingly, we observed a significant increase of abnormal Nesprin-1 distribution in CM nuclei in Lmna DCM compared to controls (Additional file 1: Fig. S13d), indicating that Lamin A/C knockdown also induces abnormal distribution of Nesprin-1. Furthermore, the proportion of nuclear protrusion with abnormally distributed Nesprin-1 was significantly increased compared to those with normal Nesprin-1 distribution in Lmna DCM CMs (Additional file 1: Fig. S13d).

Our earlier work suggested that mutating the KASH domain of Nesprin-1 could ameliorate Lmna DCM [39]. To perturb the interaction between Sun1 and Nesprins in Lmna DCM mice, we introduced the KASH domain of Nesprin-1 to compete for binding with Sun1. Interestingly, upregulation of the KASH domain suppressed Lmna DCM. The EF of Lmna DCM mice was significantly improved by KASH domain expression (Additional file 2: Table S7). Furthermore, cardiac fibrosis was significantly reduced in Lmna DCM mice with KASH domain upregulation (Additional file 1: Fig. S14a, b). Consistently, we observed a significant decrease in HF markers Nppa and Nppb as well as fibrosis markers Col1a1 and Col1a2 with KASH treatment (Additional file 1: Fig. S14c). Taken together, these results indicate that the LINC complex is a key target for treatment of Lmna DCM.

Consecutive AAV delivery into the same mouse is not possible due to the development of neutralizing antibodies to AAV capsids following initial AAV exposure. To assess whether potential candidates can halt or reverse the DCM phenotype, alternative Lmna DCM models other than AAV9-Lmna shRNA are required. Previously, we established an inducible Lmna DCM mouse model by CM-specific Lmna deletion in CMs in vivo. Disruption of the LINC complex by AAV9-mediated expression of a dominant negative SUN1 (DNSUN1) construct suppressed Lmna DCM and resulted in an increased lifespan in this DCM model [7]. A follow-up study showed that AAV9-DNSUN1 can suppress inducible Lmna DCM 2.5 weeks after induction of DCM, when EF has begun to decline (Additional file 2: Table S8). Importantly, the lifespan of inducible Lmna DCM mice was significantly prolonged, demonstrating that LINC complex disruption is a potential treatment, and not just a prophylaxis, for Lmna DCM (Additional file 1: Fig. S15a). However, whether expression of DNSUN1 or the KASH domain, or silencing Sun1 to disrupt the LINC complex, or overexpression of Lamin C, can suppress fully developed and late stage Lmna DCM requires further optimization due to the quick development of the phenotype and rapid mortality of our currently used mouse models.

Additional markers are required to evaluate candidates for the treatment of Lmna DCM

For a comprehensive understanding of whether the candidates and markers were related to the treatment of Lmna DCM, we aligned EF with commonly used markers including Nppa and Nppb for HF markers, and Cola1 and Cola2 for fibrosis related markers (Fig. 6a). Additionally, we incorporated picrosirius red staining for fibrosis level, αSMA staining for myofibroblast activation, p-Smad2 level for the activation of Tgfβ signaling pathway, Iba1 and CD3 staining for the infiltration of macrophages and T cells, γH2AX staining for DNA damage, and Ki67 staining for cardiac cell proliferation. Most markers except Ki67 were negatively correlated with cardiac function (Spearman’s correlation, − 0.640 ~ − 0.921, P < 0.05–0.001). HF and fibrosis markers as well as fibrosis level ranked closely with EF (Fig. 6b). However, these commonly used markers in most HF studies were not sufficient to predict the long-term efficacy for the treatment. The lapses of protective effect of the candidates were associated with an increase of markers including γH2AX, Iba1 and CD3. We further analyzed the correlation of the markers with animal survival for the treatment of Lmna DCM (Fig. 6c). Similar to EF, most markers except Ki67 and p-Smad2 were negatively correlated to animal survival (Spearman’s correlation, − 0.782 ~ − 0.927, P < 0.05–0.001). In addition to HF and fibrosis markers, top ranking markers included immune (CD3) and DNA damage markers (γH2AX) which had a highly negative correlation with animal survival, suggesting that these additional markers are important to evaluate candidates for the treatment of Lmna DCM (Fig. 6d).

Fig. 6

Evaluation of disease markers in Lmna DCM. (a and c) Heat map representing color-coded log₂ fold change of the gene expression of Nppa, Nppb, Myh7, Col1a1 and Col1a2, immunostaining (αSMA, Iba1, CD3, γH2AX, Ki67), Western blot (p-Smad2) and histology (fibrosis). (a) EF or (c) survival of Lmna DCM mice supplemented with selected candidates or control, n = 5. (b and d) Hierarchical clustering and correlation analysis of log₂ fold change of selected markers with (b) EF or (d) survival of Lmna DCM mice supplemented with selected candidates or control by nonparametric Spearman correlation, n = 5

Discussion

We have systematically evaluated candidates related to key pathways in Lmna DCM including Tgfβ/Smad, Fgf/Mapk and mTor. Among them, the Tgfβ/Smad signaling pathway is commonly activated in Lmna DCM as well as many other HF models [40]. Smad3 plays an essential role in heart remodelling in MI or TAC-induced heart failure models, although it plays different roles in CFs and CMs [41,42,43]. Consistent with MI or TAC models, CM specific reduction of Smad3 ameliorated the impaired cardiac function in Lmna DCM mice. However, cardiac fibrosis, markers for HF/fibrosis and activation of myofibroblasts were only modestly reduced. Critically, this protective effect quickly lapsed, resulting in no significant improvement of survival of Lmna DCM mice. p38α MAP kinase and mTor signaling pathways are activated in Lmna^H222P/H222P or Lmna -/- mouse models. Inhibition of these pathways by p38α inhibitor ARRY-371797 or mTOR inhibitor rapamycin prevents cardiac dilation, improves cardiac function and/or modestly extends survival of diseased mice [11, 44, 45]. However, knockdown of Mapk14 (coding for p38α), Raptor (a key component for mTORC1 signaling) or overexpression of Fgf16 (a downstream gene of Gata4) specifically in CMs did not significantly suppress Lmna DCM, suggesting their protective roles might be contributed by non-CMs or are specific to certain disease models [46]. Moreover, modulation of both Bmp7 and Ctgf, the secreted factors regulated by Yy1, improved cardiac function and extended survival by ~ 30%. However, Lmna DCM mice died quickly in a short time window after treatment, suggesting modulation of signaling pathways has a broad but modest therapeutic benefit for Lmna DCM. Similarly, modulation of calcium handling and proliferation related genes has a limited beneficial role in Lmna DCM.

We evaluated Lamin A, Lamin C and the LINC complex component Sun1 as therapeutic candidates for Lmna DCM. These three proteins are co-localized and interact in the nuclear envelope. Lamin A and Lamin C are the products of the Lmna gene via alternative splicing. Lamin A/C forms a matrix to preserve the strength and integrity of CM nuclei. Although Lamin A has a unique C-terminal modification and maturation process compared to Lamin C, Lamin A and Lamin C appear to be functionally equal and interchangeable because Lamin A or Lamin C only knockout mice are normal in general [47, 48]. Prelamin A only transgenic mice instead succumbed to DCM [30]. Consistent with Lamin C only transgenic mice, supplementation of Lamin C suppressed Lmna DCM and achieved long term survival of at least one year. Although Lamin A supplementation also prolonged survival, this protection lapsed, with median survival limited to 110 days. We identified dose-dependent Lamin A toxicity as the novel pathological mediator of this effect. This is inconsistent with previous results that transgenic mice expressing Lamin A alone in the absence of Lamin C had no disease phenotypes [48]. One possible explanation for this discrepancy is that rAAV9 mediated upregulation of transgenes is acute and rapid in comparison to classic transgenic techniques. In addition, we speculate that there is a need to maintain a specific ratio between Lamin A and Lamin C levels in the CMs, given that endogenous Lamin C is the dominant form in CM compared to Lamin A [14]. It is possible that Lamin A upregulation, resulting in an increased Lamin A/C ratio, led to detrimental effects. Hence, the necessity to ensure a careful balance of cellular Lamin A levels has important implications for LMNA DCM gene therapy approaches. We note that this is consistent with a reported higher frequency of misshapen nuclei in mouse embryonic fibroblasts of Lamin A only mice [48]. Lamin A upregulation induced DNA damage, however, the exact mechanisms of how Lamin A upregulation resulted in DCM is yet unknown and warrants further study. Sun1 interacts with Lamin A/C, linking nucleoskeleton to cytoskeleton via KASH domain proteins such as Nesprin-1. Perturbing this interaction via Sun1 shRNA or KASH overexpression significantly suppressed Lmna DCM and achieved long term survival, which is consistent with a protective role of dominant negative Sun1 in a DCM model induced by cardiac specific Lmna knockout [7]. Sun1-/- or mutating the KASH domain of Nesprin-1 also ameliorates Lmna null and progeroid Lmna^Δ9 mutant mice, suggesting that Sun1/Nesprin-1-containing LINC complexes can serve as a therapeutic target for other Lmna related diseases [39, 49, 50]. A comparable beneficial effect between Sun1 knockdown and Lamin C gene replacement for Lmna DCM advances a potential of translational research for LMNA DCM treatment. In addition, this finding implies that perturbing the LINC complex can achieve a similar efficacy as gene replacement for Lmna DCM, which can be potentially extended to the treatment for LMNA DCM caused by different pathological variants. To date, more than 400 LMNA mutations have been described (http://www.umd.be/LMNA/) and among them, 165 unique mutations distributed along the entire LMNA gene have been linked to cardiomyopathy [51]. Missense mutations in LMNA have been proposed to act mainly through a dominant negative pathway while truncating variants likely result in haploinsufficiency [52, 53]. Only a handful of mutations are unique to Lamin A, the majority of which are located on the C terminus of prelamin A, a hotspot for progeria where DCM mutations are rare [51, 52, 54] (Additional file 1: Fig. S16). Hence, Lamin C gene replacement is suitable for the treatment of LMNA DCM with haploinsufficiency mechanism through compensation for the loss of functional Lamin A/C in which the majority of mutations affect both Lamin A and C. On the other hand, the rescue role of Sun1 knockdown, KASH overexpression or DNSUN1 overexpression will extend to LMNA DCM caused by dominant negative mutations.

We compared various pathological markers to heart function and animal survival. Heart function was negatively correlated with the expression of HF/fibrosis markers as well as fibrosis which are utilised commonly in most HF related research. Importantly, tests for B-type natriuretic peptide (BNP, coded by NPPB) and N-terminal pro-B-type natriuretic peptide (NT-proBNP) are extensively utilized as diagnostic and prognosis biomarkers for HF in clinical settings due to their highly sensitivity and specificity. However, these common markers alone cannot predict the long-term efficacy of treatments for Lmna DCM. Activation of resident or infiltration of blood-derived immune cells are involved in pathological inflammatory pathways and tissue reparative processes in HF [55,56,57]. Upregulation of inflammatory cytokines contributes to the pathogenesis of LMNA-cardiomyopathy patients and affects the severity of cardiac phenotype [58]. Reduced immune cell numbers correlated with long term animal survival following treatment by Sun1 shRNA or Lamin C supplementation in Lmna DCM. Whether other types of immune cells, resident or infiltrated are also related to heart function and animal survival requires further investigation. DNA damage was observed in CMs of Lmna DCM as well as LMNA mutant models. Suppression of DNA damage was also highly correlated with animal survival, suggesting that maintenance of DNA/nuclei integrity is important for long term efficacy of Lmna DCM therapy.

In this study, we used rAAV9 as well as a CM specific promoter to modulate gene expression specifically in CMs. Besides cardiac tropism, AAV9 has a broad tropism to other organs including neuron, liver and lung [59]. AAV9’s liver tropism also facilitates CRISPR-based gene therapy for liver targets related to cardiovascular disease including atherosclerotic cardiovascular disease (PCSK9) [60]. However, AAV9 has limited access to cardiac fibroblasts, smooth muscle cells, and endothelial cells which impedes application to non-CMs in hearts. Recently AAV9 capsid modification enhanced potency for skeletal muscle which could extend applications to muscular disease such as Duchenne muscular dystrophy [61]. Since our Lmna DCM model is also cardiac specific, the evaluation of selected candidates specifically in CMs highlights a potential of advancing positive candidates to the treatment of cardiac specific LMNA DCM [62]. Whether these candidates also benefit other LMNA related disease requires further validation.

We found that Lamin C and the LINC complex are promising gene therapy targets for LMNA DCM. However, this study has several limitations. First, we only modulate Lmna and potential candidates in CMs. Hence, we may have overlooked the functions and contribution of these candidates in non-CMs. Second, due to the limitations of rAAV9 delivery, we were unable to obtain satisfactory expression levels via consecutive virus injection to assess whether Lamin C and Sun1 shRNA can reverse established Lmna DCM. To overcome these limitations, alternative animal models such as inducible knockout or knockin mice can be used to further evaluate the potential candidates. Finally, atrioventricular block and arrhythmias frequently occur in patients with LMNA-associated DCM. It is worth evaluating whether modulation of Lamin C and Sun1 are able to suppress these abnormal ECG features.

Here, we evaluated 14 potential therapeutic candidates in Lmna DCM. Heart function, animal survival and pathological cardiac markers were analyzed and compared, revealing Lamin C and Sun1 as viable therapeutic targets, and an unforeseen risk of Lamin A toxicity. This study further provides a simple platform that can be adapted to evaluate and compare other potential candidates for LMNA DCM.

Conclusions

This study demonstrates that gene replacement or LINC complex perturbation has a profound and beneficial impact on LMNA DCM, providing a solid foundation for the therapy development. Currently, researchers are investigating AAV-mediated gene therapy approaches for cardiovascular disease and their potential benefits for HF patients. These findings suggest that gene therapy could be a potential option for the treatment of LMNA DCM.