Loss of Elmsan1 in Cardiomyocytes Leads to Age-Dependent Cardiac Dysfunction and Reduced Lifespan

Abstract

Histone deacetylase (HDAC) complexes regulate pathological gene programs during heart disease progression. The recently identified mitotic deacetylase complex (MiDAC), which includes DNTTIP1, ELMSAN1, and HDAC1/2, remains the least characterized among these complexes. ELMSAN1 has been implicated in left ventricular remodeling, and its global deletion in mice leads to heart malformation. To investigate its role in mouse heart, we generated cardiomyocyte-specific Elmsan1 knockout (ELM cKO) using αMHC-driven Cre recombinase. We analyzed both male and female animals across three experimental groups: αMHC-Cre (Cre control), ELM fl/fl (floxed control) and ELM cKO. In male ELM cKO mice, ejection fraction (EF) was significantly reduced by 12 weeks (45.64 ± 3.12%), compared to αMHC-Cre (55.91 ± 1.29%) and ELM fl/fl (59.16 ± 3.70%) controls. By 24 weeks, EF declined further to 20.79% ± 4.52, representing a reduction of 46.4% (P < 0.01) and 62.1% (P < 0.0001) compared to αMHC-Cre and ELM fl/fl mice respectively. The heart failure phenotype in ELM cKO mice was supported by cardiomyocyte hypertrophy morphology, ventricular dilation, and shortened lifespan. Female ELM cKO mice exhibited similar defects with delayed onset. To investigate early molecular changes, we performed RNA sequencing on presymptomatic hearts from 8-week-old mice. A total of 1055 genes were differentially expressed in ELM cKO hearts, with 460 upregulated and 595 downregulated. Gene enrichment analysis revealed suppression of tricarboxylic acid cycle and key cardiac genes. These transcriptional changes were accompanied by decreased mitochondrial respiratory chain complex proteins, ultrastructural mitochondrial abnormalities, and impaired calcium handling. Our study demonstrates that Elmsan1 is pivotal for maintaining the heart function and hemostasis with advanced age.

New & Noteworthy

Our study demonstrates that Elmsan1, a unique component of the mitotic deacetylase complex (MiDAC), is essential for maintaining cardiac function. Loss of Elmsan1 in cardiomyocytes leads to age-related cardiac dysfunction and mitochondrial abnormalities in mice. Using a cardiomyocyte-specific Elmsan1 knockout model, we show that Elmsan1 preserves adult heart function by regulating genes involved in calcium handling and energy metabolism, underscoring the specific role of MiDAC in maintaining heart hemostasis.

1. Introduction

Heart diseases are the leading cause of mortality in the United States [1]. Although improvements in surgical outcomes reduced heart disease mortality, heart disease has steadily risen over the past decade [1]. Heart diseases fundamentally arise from cardiomyocyte dysfunction and/or loss, secondary to genetic mutations and/or pathogenic stresses, such as atherosclerosis, obesity, diabetes, and aging [2–6]. Increasing evidence indicates that histone modifications, which mainly change chromatin structure and accessibility, contribute to the development of cardiovascular diseases, such as heart failure [7–9]. Elucidating the impact of histone modifiers on the pathogenesis of heart failure has the potential to facilitate the development of future therapeutic interventions.

Histone acetyltransferases (HATs) and histone deacetylases (HDACs) mediate protein (e.g., histone) posttranslational modification through acetylation and deacetylation, respectively. Class II and Class III HDACs have been reported to be cardioprotective [10, 11], whereas Class I HDACs are associated with cardiac arrhythmia, hypertrophy, and dysfunction [10, 12]. Consequently, HDACs have emerged as potential therapeutic targets for heart diseases [13]. With their catalytic core, HDACs function as the primary enzymatic component within large multiprotein complexes, known as HDAC complexes [14]. Well-characterized HDAC complexes include NuRD-Sin3-CoREST-HDAC1/2, as well as SMRT/NCoR-HDAC3 [15]. These complexes are involved in a diverse array of cellular processes, including transcriptional repression, cell division, and differentiation [14, 15].

Despite extensive studies on the NuRD-Sin3-CoREST-HDAC1/2 and SMRT/NCoR-HDAC3 complexes, far less is known about the mitotic deacetylase complex (MiDAC). Comprising HDAC1/2, ELMSAN1 (MiDEAS), and DNTTIP1 (TDIF1), MiDAC was initially identified through chemoproteomic approaches [16]. ELMSAN1 contains an ELM2-SAN domain that recruits HDAC1/2 [17]. The N-terminal domain of DNTTIP1 interacts with both HDAC1/2 and ELMSAN1, while the C-terminal domain of DNTTIP1 enables binding directly to DNA and nucleosomes [17]. The involvement of ELMSAN1 and DNTTIP1 as scaffold proteins not only provides structural stability but also ensures the precise targeting of HDACs to specific genomic loci for transcriptional regulation [18–20]. MiDAC components are evolutionarily conserved across diverse species, underscoring the potential importance of this complex [16, 21, 22]. Germline deletion of either ELMSAN1 or DNTTIP1 results in embryonic lethality associated with malformed smaller hearts [20]. The pathological relevance of ELMSAN1 remains poorly characterized; to date, only a single study has identified ELMSAN1 among the top 10 genes associated with left ventricle remodeling, diastole function, prevalent HFpEF, and incident HFpEF in humans [23]. Collectively, these studies suggest that ELMSAN1 may play important roles in cardiac development and/or contractile function; however, direct evidence, particularly from in vivo genetic models, is lacking.

To gain insight into the potential roles of ELMSAN1 in the heart, we generated cardiomyocyte-specific Elmsan1 knockout (ELM cKO) mice using αMHC-Cre-mediated recombination. We hypothesized that depletion of Elmsan1 in adult cardiomyocytes would impair cardiomyocyte function and disrupt cardiac hemostasis. Serial echocardiography revealed that ELM cKO mice develop heart failure by 24 weeks of age, associated with cardiomyocyte hypertrophy, cardiac fibrosis, induction of molecular stress markers, and ultrastructural abnormalities. Furthermore, we took advantage of the opportunity to comprehensively understand Elmsan1’s function in the heart by examining the effects of its loss at histological, cellular and molecular levels. Our findings provide direct evidence that Elmsan1 is essential for maintaining cardiac structure and function, and further suggest a potential contribution of MiDAC in the pathogenesis of dilated cardiomyopathy-associated heart failure.

2. Methods

2.1. Mice

Male and female cardiomyocyte (CM)-specific Elmsan1 knockout (ELM cKO) mice were generated by crossing Elmsan1 floxed mice with α myosin heavy chain (αMHC)-Cre mice. Elmsan1 floxed mice were generated by Biocytogen (Co., Beijing, China), through use of CRISPR/Cas9 technology; more specifically, exon 4 of the Elmsan1 gene was flanked with LoxP sites (Figure 1A). αMHC-Cre mice were a generous gift from Dr. Gangjian Qin (UAB). The Institutional Animal Care and Use Committee of the University of Alabama at Birmingham approved all animal procedures and experiments used in this study (protocol # IACUC-22306). All mice were housed at a constant temperature and were kept on a 12-hour light-dark cycle with free access to water and food.

Figure 1. Generation of cardiomyocyte-specific Elmsan1 KO mice.

A, The targeting strategy to conditionally knock out mouse Elmsan1 exon 4 by Cre-Loxp system. B, Genotyping of Elmsan1 floxed mice by using 5’ Loxp (up) and 3’ Loxp (down) primers. C, Breeding strategies to obtain cardiomyocyte-specific Elmsan1 KO (ELM cKO) mice and control littermate Elmsan1 fl/fl (ELM fl/fl) mice. D, Cre-mediated exon 4 deletion was confirmed in heart tissues by genomic DNA PCR. n=2 per group. E, Illustration and Sanger sequence results showing deletion of exon 4 in Elmsan1 transcript and frameshift of amino acid codes. F, RT-qPCR analysis of Elmsan1 in the ELM fl/fl and ELM cKO mouse hearts and isolated cardiomyocytes (Iso. CM). The relative expression was normalized as indicated. The value in ELM fl/fl was set as 1. n=3 per group. Data are presented as the mean ± SEM and analyzed an unpaired two-tailed Student′s t test. *P < 0.05, **P < 0.01.

Mouse genotyping was performed by PCR of tail DNA. To assess the presence or absence of the Elmsan1 floxed allele, the following primers were utilized: 5’-GTAGTTTACAGTCGGACATGAGGGAAG-3’ (forward) and 5’-GCGTTTGTCAGTTTGACATCAACC-3’ (reverse) for 5’ Loxp; 5’-GACCAGCAGGCAGTGAATGAGGC-3’ (forward) and 5’-GGCCTCACCTTTCTCCAAGTTAATC-3’ (reverse) for 3’ Loxp. PCR products for the 5’ Loxp were 317bp (wild-type allele) and 481bp (floxed allele), while for the 3’ Loxp, the products were 367bp (wild-type allele) and 453bp (floxed allele, Figure 1A). Genotyping of αMHC-Cre mice was performed using previously published PCR strategies [24]. To assess cardiac-specific deletion of exon 4 in the Elmsan1 gene of ELM cKO mice, genomic DNA was isolated using TRIzol^™ Reagent (#15596026, Invitrogen) based on manufacturer’s instruction from heart, liver, kidney, spleen, lung, and skeletal muscle, followed by PCR analysis using the following primers: 5’-GAGAGTGGGATGGTACCCCT-3’ (forward) and 5’-CTCCCCCATCACAGAGAGGA-3’ (reverse). A PCR product of 360 bp was specific only for the gene with exon 4 deletion (Figure 1D).

2.2. Adult Mouse Cardiomyocyte (AMVM) Isolation

Cardiomyocytes were isolated through use of previously described Methods [25]. Briefly, the heart from each adult male mouse was harvested and cannulated through the aorta and then perfused with perfusion buffer and digested with digestion buffer. The digested heart tissue was minced and gently pipetted to small fragments, followed by filtering through 100μm strainer. Cells were repeatedly centrifuged and resuspended by gradually increased Ca²⁺ solution. Cells were finally seeded on laminin-coated tissue culture dishes and cultured in plating medium (DMEM, 10% FBS, 1% Penicillin/Streptomycin, 1% ATP, 0.1% Blebbistatin) for 1–3hrs, followed by overnight culture in culture medium (MEM, 1% Penicillin/Streptomycin, 0.1% Blebbistatin, 1% BSA).

2.3. Echocardiographic Assessment

Echocardiography was performed with MS400 transducer on Vevo2100 or F2 (VisualSonics) to capture images in the parasternal short-axis view by M-mode. In brief, male and female mice were anesthetized by inhalation of isoflurane (1.5–2%) and then secured to a 37 °C heating pad. Hair was removed from the chest using Nair. Heart rates were monitored and maintained at 400–500 beats per minute. M-mode images were used to measure left ventricular (LV) interventricular septal thickness (IVS), LV internal dimensions (LVID) and LV posterior wall thicknesses (LVPW) at diastole (D) and systole (S). For each mouse, 3–5 measurements were taken, and the average value was calculated to ensure accuracy and consistency. LV ejection fraction (EF), LV fractional shortening (FS) and LV mass were also calculated using Vevo Lab software.

2.4. Histological Examination

Heart tissue paraffin embedding, sectioning and H&E staining were performed by the UAB Pathology Core Research Laboratory. Sections were stained with Masson Trichrome (HT15, Sigma-Aldrich) as per the manufacturer’s instructions. Images of LV region were captured using Olympus 800. The average value of four different areas in LV region per male mouse heart was measured (at least 3 hearts per group). The quantification of LV fibrosis was determined with Image J software.

2.5. Wheat Germ Agglutinin (WGA) Immunostaining

Paraffin-embedded sections were stained using wheat germ agglutinin (WGA) staining kit (Invitrogen^™, W32466) as per manufacturer’s instruction. Images were captured using Olympus 800. WGA-stained cross-sectional area of cardiomyocytes was determined with Image J software. Three different areas in LV region and 15 cardiomyocytes in each area (at least 3 hearts per group) were assessed, and the average value was used for each male mouse heart.

2.6. Optical Recordings of Calcium Transients

Isolated adult mouse cardiomyocytes were stained with a low-affinity Ca-sensitive dye Calbryte-520 AM at a concentration of 5 μM for 1 h [26], transferred to a perfusion chamber mounted on an inverted microscope, and perfused with DMEM medium supplemented with 20 μM of Blebbistatin. Cells were stimulated with 5-ms rectangular pulses delivered at 500-ms intervals via a glass pipette filled with Hank’s solution; the return electrode was located at the chamber periphery. Fluorescence was excited using a 200-W Hg/Xe arc lamp and optical signals were recorded with a 16×16 photodiode array (Hamamatsu) at a spatial resolution of 110 μm per diode as previously described [27]. Calbryte-520 AM fluorescence was excited at 480/40 nm and measured at 535/50 nm. Optical signals were digitally filtered to increase the signal-to-noise ratio. The onset of Ca transients was measured at 50% of the maximum transient amplitude. Calcium transient durations were measured at 50% and 80% of signal recovery (CaD₅₀ and CaD₈₀, respectively). In each glass coverslip, Ca measurements were performed at 4–6 different locations.

2.7. Transmission Electron Microscopy (TEM)

Mouse heart samples were prepared by the UAB High Resolution Imaging Core as previously described [28], with modifications. Briefly, mouse left ventricle heart tissue was cut into a cube less than 2mm square and incubated in a fixative buffer (3% glutaraldehyde in 0.15M NaCaCo buffer; pH 7.4). After fixation, specimens were rinsed several times with NaCaCo buffer followed by post-fixation with 1% osmium tetroxide for one hour. Following further rinsing, specimens were dehydrated using ethyl alcohols. The tissue was subsequently embedded and sectioned. Sections were placed on either copper or nickel mesh grids, and stained with heavy metals (uranyl acetate and lead citrate) for contrast. Five different areas (images) from each grid with longitudinal sarcomere section were randomly picked to image. Imaging was performed using a JEOL 1400 FLASH 120kv TEM (JEOL USA Inc, Peabody, MA). Digital images were taken using an AMT NanoSprint43 Mark II camera (AMT Imaging, Woburn, MA). The average area of 100 mitochondria per male mouse heart was counted and used to analyze (3 hearts per group).

2.8. Citrate synthase assay

Frozen mouse heart (5 male mice at age of 10 weeks for each group) was pulverized, homogenized and protein quantified (5 μg protein used for citrate synthase assays). Citrate synthase was measured using the coupled reaction with oxaloacetate, acetyl-CoA, and 5,5-dithiobis-(2,4-nitrobenzoic acid) [29–31]. Citrate synthase was used as a surrogate index of mitochondrial volume [30, 32–35].

2.9. mRNA Expression Analysis

Total RNA was isolated from male mice hearts using TRIzol^™ Reagent (#15596026, Invitrogen) according to manufacturer’s instructions. cDNA synthesis was performed using High-Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific, Cat#: 4368813). To validate the deletion of exon 4 in Elmsan1 mRNA, we performed PCR amplification of the indicated regions using forward (5’-GAGAGTGGGATGGTACCCCT-3’) and reverse (5’-CTCCCCCATCACAGAGAGGA-3’) primers and analyzed Sanger sequencing results. Quantitative real-time PCR (RT-qPCR) was performed using SYBR Green Master mix (#B21202, Bimake) with specific primers on the ABI Applied Biosystem 7900HT Fast Real-Time PCR instrument. Relative gene expression was normalized to housekeeping gene and represented as fold change. Forward and reverse primer sequences are listed in Table 1.

Table 1.

Primer sets for RT-qPCR analysis.

	Forward	Reverse
Gapdh	AGGTCGGTGTGAACGGATTTG	TGTAGACCATGTAGTTGAGGTCA
Cre	GCTAACCAGCGTTTTCGTTC	GCATTTCTGGGGATTGCTTA
Elmsan1	TTCATCGCCCCTCCTGTCTA	GGGTTCGATGGGCACTCTTT
β-Actin	GGCTGTATTCCCCTCCATCG	CCAGTTGGTAACAATGCCATGT

2.10. RNA Sequencing (RNA-seq) Analysis

RNA-Seq was performed by Novogene (Co., China). In brief, RNA-Seq library were prepared using the NEBNext Ultra RNA library Prep Kit (New England Biolabs). Quality of the library was assessed using Bioanalyzer High Sensitivity DNA Chip. GRCm38 (mm10) was used to align and map the paired-end reads. RNA-Seq data were analyzed using DESeq2, which calculated fold change, P- and adjusted P- (False Discovery Rate; FDR) values for genotype effects. Gene Ontology (G0) and KEGG analyses were performed on differentially expressed genes (FC ≥ 2, adjusted p-value < 0.05). Gene enrichment analysis (GSEA) was performed using the molecular signatures database (https://www.gsea-msigdb.org).

2.11. Cell Lysate Preparation and Western Blotting

Total protein was extracted from male mouse heart tissue using a lysis buffer containing a protease inhibitor cocktail (#05056489001, Roche). Proteins were separated by SDS–PAGE (8% −15% gel) and transferred to polyvinylidene fluoride (PVDF) membranes (#45–004-021, Cytiva Amersham). After blocking with 5% non-fat milk (#1706404, Bio-rad) for 1 h, the membranes were incubated with the following primary antibodies overnight at 4 °C: anti-ANP (1:2000, PA5–29559, Thermo Fisher), OXPHOS (1:1000, #ab110411, Abcam), anti-Hdac1 (1:200, sc-81598, Santa Cruz), anti-Hdac2 (1:10000, 12922–3-AP, Proteintech), and anti-Dnttip1 (1:100, sc-166296, Santa Cruz). Membranes were next incubated with horseradish peroxidase-conjugated secondary antibodies (#7076, Santa Cruz) at room temperature for 1 h. Blots were visualized with an enhanced chemiluminescence kit (WBKLS0500, Millipore) and analyzed using Amersham Imager 600. All densitometry data were normalized to Amido Black (#SLCP7209, Sigma).

2.12. Quantification and Statistical Analysis

Statistical analyses were performed using Prism10.0.2 (GraphPad Software). Normality of data was assessed through the use of the Shapiro–Wilks test; all experimental data fit a normal distribution. Differences between data groups were evaluated for significance with the use of unpaired parametric two-tailed Student t-tests for comparison between 2 experimental groups. ANOVA was performed when the number of experimental groups was greater than 2. To determine the effect of 3 genotypes (αMHC-Cre, ELM fl/fl, ELM cKO) and ages (i.e. weeks old), a two-way ANOVA with two-stage linear step-up procedure of Benjamini, Krieger, and Yekutieli (BKY) test was used for multiple comparisons. All data are presented as mean ± SEM, unless noted otherwise. Survival data were analyzed by using a Kaplan–Meier survival analysis with a log rank (Mantel-Cox) test. For all analyses (except RNA-seq analysis), a p-value <0.05 was considered statistically significant.

3. Results

3.1. Generation and validation of cardiomyocyte-specific Elmsan1 knockout mice

To investigate the potential roles of Elmsan1 in cardiac biology, we generated cardiomyocyte (CM)-specific Elmsan1 knockout mice (ELM cKO). Elmsan1-floxed mice (with exon 4 flanked by Loxp sites) were generated using the CRISPR/Cas9-based genome editing system (Figure 1A). Both 5’ and 3’ Loxp insertion was confirmed by genomic PCR (Figure 1B). To deplete Elmsan1 specifically in heart, Elmsan1-floxed mice were crossed with αMHC-Cre mice, which express Cre recombinase under the control of the α-myosin heavy chain (αMHC) promoter [24, 36] (Figure 1C). The excision of exon 4 was validated by PCR analysis of the Loxp loci in heart genome using the forward primer for 5’Loxp and the reverse primer for 3’Loxp (Figure 1D). Sanger sequencing for cDNA amplicon further confirmed deletion of exon 4 in ELM cKO hearts, resulting in a frameshift mutation and premature termination of the coding sequence (Figure 1E). The knockout efficiency was evaluated by qPCR analysis, showing a 52% reduction in Elmsan1 mRNA levels in homozygous ELM cKO (Elmsan1 fl/fl::αMHC-Cre/+) hearts compared with littermate flox-control ELM fl/fl (Elmsan1 fl/fl) mice (Figure 1F). To exclude the contamination in non-myocytes, we isolated cardiomyocytes from ELM cKO hearts and revealed 99.9% reduction in Elmsan1 expression (Figure 1F), suggesting sufficient deletion. Consistent with the cardiac-specific nature of the ELM cKO model, neither Elmsan1 loci (assessed by genomic PCR) nor Elmsan1 mRNA (assessed by RT-PCR) were altered in liver, kidney, spleen, lung, and skeletal muscle tissues (Supplemental Fig. S1). Therefore, this ELM cKO mouse model allows us to study the specific role of Elmsan1 in cardiac muscles.

3.2. Cardiomyocyte-specific Elmsan1 deletion does not cause the cardiac dysfunction in young mice

Although αMHC-Cre is expressed from embryonic day 10.5 (E10.5) [37], the cardiac ablation of Elmsan1 did not result in an embryonic lethal phenotype or abnormality in embryos or pups (Supplemental Fig. S2A). Mice with the mutant genotype were observed born at expected Mendelian ratios (Supplemental Fig. S2B). No defect or behavior abnormality was observed in cKO mice up to the age of sexual maturity (8 weeks old). To further evaluate the physiological impact of Elmsan1 deficiency in the heart, we performed echocardiographic analysis in male and female mice at either 4 or 8 weeks of age (Figure 2A–2F and Supplemental Fig. S3A-S3C). No significant differences were determined in echocardiographic parameters between ELM cKO and littermate flox-control male mice. However, female αMHC-Cre mice showed a relative lower, though still close to normal range, ejection fraction (EF, 48.72±3.64%) and fractional shortening (FS, 23.83±2.11%) at 4 weeks (Figure 2A–2B). At the same time point, male αMHC-Cre mice showed relatively high left ventricular (LV) MASS (Figure 2C) and interventricular septal at end-diastole (IVSD) (Supplemental Fig. S3B) compared with other strains, likely due to genetic background-related variability during early development [38]. To assess potential histological alterations in ELM cKO hearts, hematoxylin and eosin (H&E) and Masson Trichrome staining were performed. These analyses indicated no significant structural abnormalities or cardiac remodeling (e.g., fibrosis) in male ELM cKO mice at 8 weeks of age (Figure 2G and 2H). Collectively, these data demonstrate that cardiomyocyte-specific Elmsan1 deletion in mice does not adversely affect cardiac function or structure up to 8 weeks of age (i.e., sexual maturity).

Figure 2. Evaluation of heart function and physiology in ELM cKO mice at 4 and 8 weeks.

A through F, Heart function of male and female αMHC-Cre, ELM fl/fl and ELM cKO mice was measured by transthoracic echocardiogram at 4 weeks and 8 weeks of age. A, Ejection fraction (EF). B, Fractional shortening (FS). C, Left ventricle mass (LV mass). D, Left ventricle internal dimension at end-systole (LVIDS). E, Interventricular septal end-systole (IVSS). F, Left ventricle posterior wall thickness at end-systole (LVPWS). Male mice, n=6–8 per group, Female mice, n=3–8 per group. Data are presented as the mean ± SEM and analyzed using two-way ANOVA. * P < 0.05, ** P < 0.01, **** P < 0.0001. G, Representative images of H&E staining for heart tissues at 8 weeks of age. H, Representative Mason Trichrome-stained LV regions and the quantification of LV fibrosis in male ELM fl/fl (n=4) and ELM cKO (n=7). Data are presented as the mean ± SEM and analyzed using an unpaired two-tailed Student′s t test. n.s., not significant.

3.3. Age-onset cardiomyopathy in ELM cKO mice

Continued phenotypic characterization of aging mice revealed the onset and progression of cardiac dysfunction. Serial echocardiography was performed at 12 and 24 weeks in male and female mice to monitor changes in heart function (Figure 3A-F and Supplemental Fig. S4A-S4C). ELM cKO mice showed a progressive decline in EF and FS over time. By 24 weeks of age, they developed significant heart failure, with markedly reduced EF and FS compared to ELM fl/fl or αMHC-Cre controls (Figure 3A and 3B). Similar patterns were observed in female ELM cKO mice, although the onset of EF and FS reduction occurred later than that in males (Figure 3). Despite the decline in systolic function, LV mass and heart/tibia ratio remained unchanged between male ELM cKO and control mice throughout the study period (Figure 3C and 3G). Instead, increase in left ventricular internal dimension (LVIDS and LVIDD, Figure 3D and Supplemental Fig. S4A) and left ventricular posterior wall thickness at end-systole (LVPWS, Figure 3F) at 12 and 24 weeks of age indicated dilated remodeling in ELM cKO hearts. Consistently, the histology evaluation showed reduced wall thickness and chamber dilation (Figure 3H). Additionally, the cardiac stress marker atrial natriuretic peptide (ANP) was significantly increased in ELM cKO hearts at 12 weeks (Figure 3I). Taken together, these findings indicate that Elmsan1 deletion in cardiomyocytes leads to age-onset cardiac dysfunction and structural abnormalities.

Figure 3. Age-dependent cardiac dysfunction in ELM cKO mice.

A through F, Heart function of male and female αMHC-Cre, ELM fl/fl and ELM cKO mice was measured by echocardiogram at 12 weeks, 24 weeks of age. A, Ejection fraction (EF). B, Fractional shortening (FS). C, Left ventricle mass (LV Mass). D, Left ventricle internal dimension at end-systole (LVIDS). E, Interventricular septal end-systole (IVSS). F. Left ventricle posterior wall thickness at end-systole (LVPWS). Male mice, n=4–9 per group, Female mice, n=4–8 per group. Data are presented as the mean ± SEM, and analyzed using two-way ANOVA. * P < 0.05, ** P < 0.01, *** P < 0.001,**** P < 0.0001. G. Heart weight (HW) to tibia length ratio and left ventricle weight (LVW) to tibia length ratio in 24 weeks old αMHC-Cre, ELM fl/fl and ELM cKO male mice. Data are presented as the mean ± SEM, and analyzed an unpaired two-tailed Student’s t test. n.s. not significant. H, Representative images of H&E staining for heart tissues at 24 weeks of age. I, Immunoblot and quantification results showing significant increase of ANP in 12 week-old male ELM cKO mice, compared with ELM1 fl/fl mice. Amido black staining results were served as loading controls. n=3 per group. Data are presented as the mean ± SEM, and analyzed an unpaired two-tailed Student′s t test. * P < 0.05.

3.4. Cardiac transcriptome alterations in ELM cKO mice

To investigate gene programs associated with Elmsan1 loss and its role in cardiac dysfunction, we performed RNA sequencing (RNA-seq) analysis on hearts collected from 8-week-old ELM fl/fl (n=2), αMHC-Cre (n=2), and ELM cKO (n=4) mice. To exclude potential effects of gene-edited flox sites and/or Cre recombinase, we conducted comparisons between: 1) ELM cKO versus ELM fl/fl; and 2) ELM cKO versus αMHC-Cre samples. A total of 1,055 differentially expressed genes (DEGs) were identified as commonly altered (fold change ≥ 2 and adjP < 0.05) in ELM cKO heart, regardless of control group used. Among these genes, 460 (43.6%) were significantly upregulated and 595 (56.4%) were downregulated upon Elmsan1 deletion (Figure 4A). Upregulated genes enriched for gene ontologies (GO) related to cytoplasmic translation and collagen-containing extracellular matrix (Figure 4B). It is noteworthy that many key cardiac genes were significantly downregulated in ELM cKO hearts, including contractility genes Pln (encoding phospholamban), Atp2a2 (encoding SERCA) and calcium handling genes Ryr2 (encoding ryanodine receptor 2), Jph1 (encoding junctophilin 1), and Slc8a1 (encoding NCX1, Figure 4B). This coordinated suppression of calcium transport genes suggests a global disruption in calcium homeostasis machinery. Consistent with this, calcium transient measurements in adult cardiomyocytes isolated from 12-week-old hearts showed prolonged decay times, with increased CaD50 and CaD80 value (time to 50% and 80% recovery of calcium transient duration) in Elmsan1-null cells (Supplemental Fig. S5A, indicating impaired calcium reuptake. The GO analysis also showed that downregulated genes encode proteins that largely localize to the mitochondrial matrix and membrane and sarcoplasmic reticulum (Figure 4C). Additionally, the gene set enrichment analysis (GSEA) indicated that T-tubule related genes and tricarboxylic acid (TCA) cycle genes were under-representative in ELM cKO samples (Figure 4D), suggesting potential alterations in contractility and metabolism. In summary, these global changes in gene expression suggest that cardiac dysfunction following Elmsan1 deletion might be associated with alterations in calcium handling, cardiac contractility and mitochondrial function, implicating Elmsan1 as a key transcriptional regulator of cardiac homeostasis.

Figure 4. Transcriptome changes in ELM cKO hearts at 8 weeks.

A, Venn diagram showing differentially expressed genes (DEGs, fold change ≥ 2, adjP < 0.05) identified by comparisons between ELM cKO vs ELM fl/fl and ELM cKO vs αMHC-Cre, at 8-week-old age. B, Heatmap showing DEG patterns, enriched Gene Ontology (GO) and representative genes. C, GO terms enriched in downregulated genes in ELM cKO. Cellular Component (CC). D, Gene set enrichment analysis (GSEA) showing T-tubule and tricarboxylic acid cycle (TCA) related genes underrepresented in ELM cKO compared to ELM fl/fl control. BP, Biological Process; NES, normalized enrichment score.

3.5. Elmsan1 involves in the regulation of energy metabolism

Mitochondrial metabolism is closely linked to cardiac disease and comprises two major processes: the TCA and the electron transport chain (ETC) (Figure 5A), both of which were enriched among DEGs in ELM cKO heart. Thus, to further investigate the impact of Elmsan1 depletion on ETC, we assessed the levels of mitochondrial respiratory chain complex subunits. Western blot analysis revealed significant reduction in mitochondrial complex I and complex II proteins in ELM cKO hearts at 8 weeks (with trends for decreases in other complexes; Figure 5B). Since complex I and II in the ETC replenish NAD + and FAD, respectively, allowing the oxidative TCA cycle to function, reduction in these proteins could impair TCA cycling [39]. Consistent with this, our transcriptome data showed that TCA cycling-related genes, including Sdha, Sdhd, Fh1, Aco1, and Aco2, were significantly downregulated in ELM cKO hearts (Figure 5C). Additionally, two master regulators of cardiac energetics, Ppara (encoding PPARα) and Ppargc1a (encoding PGC1α) were also decreased [40] (Figure 5D), suggesting disrupted energy metabolism in ELM cKO hearts. Although the quantification of citrate synthase (CS) activity showed no significant differences between 10-week-old ELM cKO and ELM fl/fl hearts (Supplemental Fig. S6A), transmission electron microscopy (TEM) revealed enlarged mitochondria and abnormal cristae in ELM cKO hearts at 12 weeks (Figure 5E). Together with the downregulation of key regulators of mitochondrial fission (Dnm1l, Mff) and mitophagy (Pink1, Supplemental Fig. S6B), these findings suggest that mitochondrial dynamics and quality control pathways are compromised in ELM cKO hearts. All these findings support the conclusion that Elmsan1 loss-induced cardiac dysfunction is associated with altered regulation of mitochondrial metabolism.

Figure 5. Mitochondrial defects in ELM cKO mice hearts.

A, Schematic of TCA cycle and electron transport chain (ETC) complexes in mitochondria. B. Western blot and quantification of ETC complexes from 8-week-old mice heart tissue. Amido black staining results were served as loading controls. The protein level in ELM fl/fl sample was set as 1. C I-C V, ETC complexes as indicated in A, n=3 per group. Data are presented as the mean ± SEM, and analyzed using an unpaired two-tailed Student′s t test. * P < 0.05, ** P < 0.01. C, Heatmap showing expression changes of TCA cycle genes in heart tissues from 8-week-old ELM fl/fl mice, αMHC-Cre mice and ELM cKO mice. D. Normalized RNA-seq read counts of mitochondria metabolism genes, Ppara and Ppargc1a, in heart tissues with or without Elmsan1 depletion. The statistical results were indicated by adjusted p values generated by DESeq2 (see Methods). E, Representative images and quantification results of TEM in 12-week-old male mice. Data are presented as the mean ± SEM, and analyzed using an unpaired two-tailed Student′s t test. ** P < 0.01. Star indicates mitochondria with abnormal cristae.

3.6. Reduced lifespan following cardiomyocyte-specific Elmsan1 deletion

To establish the long-term consequences of cardiomyocyte-specific Elmsan1 deletion, we monitored mice for up to 1 year. Compared with ELM fl/fl and αMHC-Cre mice, both male and female ELM cKO mice exhibited a reduced lifespan. More specifically, no ELM cKO survives beyond 11 months (Figure 6A). In addition, ELM cKO mice displayed significant body weight loss relative to ELM fl/fl mice, though not significantly different from αMHC-Cre mice (Supplemental Fig. S7). Histologic examination of ELM cKO hearts at 7 months of age revealed enlarged heart chambers, thinner ventricular walls and septa (Figure 6B), indicating a strong association of death and cardiac pathology. Additionally, both cardiomyocyte size (Figure 6C) and interstitial fibrosis (Figure 6D) were significantly increased in ELM cKO hearts at 24 weeks of age. In conclusion, these data demonstrate that Elmsan1 deletion in cardiomyocytes leads to adverse cardiac remodeling and premature death.

Figure 6. Loss of Elmsan1 results in lethal heart failure.

A, Survival curve of ELM fl/fl (male mice n=6, female mice n=6), αMHC-Cre (male mice n=7, female mice n=9,), and ELM cKO mice (male mice n=6, female mice n=6) analyzed by log-rank test. B. Representative images of heart morphology and H&E staining in 7-month-old male mice. C, Representative wheat germ agglutinin (WGA) images and quantification of LV regions of male mice at 6-month-old, at least 3 mice per group. D, Representative images and quantification of heart Masson Trichrome-stained LV regions of male mice at 6-month-old, at least 3 mice per group. Data are presented as the mean ± SEM, and analyzed using an unpaired two-tailed Student′s t test. * P < 0.05, *** P < 0.001,**** P < 0.0001.

4. Discussion

Our study provides new evidence on the critical role of ELMSAN1 in cardiomyocytes and overall cardiac function. By utilizing cardiomyocyte-specific Elmsan1 knockout (ELM cKO) mice, we revealed that Elmsan1 is essential for maintaining postnatal cardiac function. Our findings demonstrated that Elmsan1 deletion leads to significant age-onset cardiomyopathy, characterized by decreased ejection fraction, fractional shortening, and structural changes in the heart. Additionally, we observed notable alterations in the cardiac transcriptome, particularly affecting mitochondrial metabolism and essential calcium handling functions. These insights underscore the importance of ELMSAN1 in preserving mitochondrial energetics and preventing cardiac dysfunction, with its loss ultimately contributing to a reduced lifespan in ELM cKO mice. Importantly, our results provide additional insights into how modulation of the specific HDAC complex MiDAC influences heart physiology and is distinct from global deletion or inhibition of Hdac1 and/or Hdac2.

4.1. αMHC-Cre induced ELM cKO mice model

Elmsan1 global knockout mice die around embryonic day 16.5, exhibiting severe anemia and pronounced heart malformations [20]. Gene expression analysis of mouse embryonic fibroblasts from homozygous Elmsan1 KO embryos revealed disruptions in developmental pathways [20], indicating that Elmsan1 is essential for development. However, the global deletion of Elmsan1 affects all tissues and stages of development, making it difficult to determine the specific functions of Elmsan1 and identify which phenotypes, such as heart malformations, are directly related to its primary function. Given that Elmsan1 has distinct roles across tissues or cell types [18, 20, 41], it is important to develop a tool for studying its role in specific tissues. Therefore, we generated Elmsan1 floxed mice (Figure 1), providing a versatile genetic model to investigate Elmsan1’s role not only in heart, by crossing with the αMHC-Cre line, but also in other organs if using tissue-specific Cre lines.

Although the αMHC promoter-driven Cre can be detected between E10.5 and E11.5, which is prior to the lethality of global Elmsan1 knockout embryo, we did not observe abnormality during prenatal stage (Supplemental Fig. S2A), and the pups were born at Mendelian frequencies based on our breeding strategy (Supplemental Fig. S2B). Normally, the heart begins to develop different structures from E10.5 [42], such as formation of ventricular septation, atrial septation and valves. It suggests that the heart malformation is likely due to an earlier depletion of Elmsan1 in mouse embryos. ELM cKO mice did not lead to noticeable deficiencies in heart morphology at young ages (4 to 8 weeks), nor exhibited significant differences in cardiac function or structure compared to control mice (Figure 2), indicating that Elmsan1 depletion after E11.5 has minimal influence on developing heart at prenatal and postnatal stage. While our published study has illustrated that ELMSAN1 regulates differentiation and maturation of cardiomyocytes derived from human induced pluripotent stem cells [43], the impact of Elmsan1 deletion on early heart development prior to E10.5 remains unclear.

4.2. Elmsan1 depletion caused dilated cardiomyopathy characterized cardiac phenotypes

Dilated cardiomyopathy (DCM) is a deadly heart condition, characterized by biventricular dilation and significant cardiac systolic dysfunction [44–46]. Our data show that ELM cKO mice develop progressive DCM with advancing age, evidenced by echocardiography and histopathology. Key features included reduced ejection fraction and fractional shortening, ventricular wall thinning, chamber dilation, and hypertrophic cardiomyocytes (Figures 3 and 6). Despite these severe changes, LV mass remained unchanged, indicating unique remodeling features. While most DCM-associated mutations are identified in sarcomere related genes, emerging studies suggest that epigenetic factors, such as histone demethylase KDM8 [47], play critical roles in regulating cardiac metabolism and contributing to DCM development [48]. Our findings suggest an additional epigenetic regulator involved in the cardiac pathology of DCM.

Our RNA-seq data revealed dysregulation of calcium handling genes in ELM cKO mice, suggesting potential defects in calcium homeostasis. These molecular alterations were supported by the observation of prolonged D50/D80 calcium decay kinetics, likely resulting from impaired sarcoplasmic reticulum calcium reuptake due to decreased Atp2a2 (SERCA) expression and/or reduced membrane extrusion associated with downregulation of Slc8a1 (NCX1, Figure 4B). Collectively, these defects point to diastolic dysfunction and an increased energetic burden, features characteristic of the DCM phenotype ELM cKO. Furthermore, impaired calcium handling may exacerbate mitochondrial dysfunction and contribute to transcriptional alterations in metabolic pathways, although it remains to be determined whether these effects result from direct regulation or secondary, indirect mechanisms.

4.3. The MiDAC histone deacetylase complex in cell and mouse models

Unlike HDAC1/2, ELMSAN1 and DNTTIP1 are unique components of MiDAC. Therefore, studying their roles in vitro or in vivo can help elucidate the distinct functions of MiDAC compared with other class I HDAC complexes. Studies by our group and others have shown that MiDAC is dispensable for growth of pluripotent stem cells [18, 43]. However, loss of MiDAC function impedes neuron and cardiomyocyte differentiation and maturation, highlighting its role in cell fate determination [18, 43]. Interestingly, in 8-week-old ELM cKO mice compared with ELM fl/fl mice, we observed increased Hdac1 and Dnttip1 protein levels (Supplemental Fig. S8A), a finding not consistent with the previous observation in iPSCs or iPSC derived cardiomyocytes[18, 43]. This elevation suggests a compensatory regulation of MiDAC components in the adult heart. Moreover, it implies that the downregulated cardiac genes observed in ELM cKO mice may be indirectly regulated by Elmsan1, reflecting a potential secondary regulatory response during the progression of cardiac dysfunction at this stage. To date, the direct genomic targets or functional histone substrates of MiDAC remain largely unspecified and inconsistent across studies [19, 41, 49, 50]. To address this gap, future investigations using acute depletion models and comprehensive histone acetylation profiling in Elmsan1- or Dnttip1-deficient cardiomyocytes will be critical to further elucidate the molecular mechanisms by which MiDAC regulates cardiac gene expression and function.

In vivo studies further indicate that MiDAC complex is crucial for embryogenesis, as homozygous global deletion of each MiDAC component leads to embryonic and neonatal lethality [20, 51]. Mice with CM-specific Hdac1 or Hdac2 KO induced by αMHC-Cre did not exhibit heart abnormalities from embryo to 8 weeks of age. However, αMHC-Cre induced double KO of Hdac1 and Hdac2 resulted in dilated cardiomyopathy, arrhythmias, and postnatal lethality by day 14, accompanied by aberrant upregulation of calcium channel genes and skeletal muscle genes [51]. Our transcriptome analysis similarly showed an increase in skeletal muscle gene expression in ELM cKO hearts (Supplemental Fig. S8B), suggesting that abnormal activation of skeletal muscle genes may not be the primary cause of the rapid heart disorder and postnatal lethality observed in Hdac1/2 double KO mice. In contrast, CM-specific depletion of Elmsan1 preserves the function of HDAC1/2 in other complexes, resulting in a delayed, age-dependent impact of MiDAC on heart function, along with unique transcriptome changes in energy metabolism genes at early stage. Currently, conditional knockout models for Dnttip1 are unavailable, limiting further exploration of MiDAC’s specific role in heart development and function.

4.4. Implications for HDAC-targeted therapies in heart disease

Our findings have important implications for the therapeutic targeting of HDACs in heart disease. Although specific HDAC1 or HDAC2 inhibitors have shown both beneficial and deleterious effects in cardiac models, the currently available small molecule inhibitors have limited specificity and target multiple deacetylases, which impedes the understanding of individual HDAC complexs [17]. Given the unique role of MiDAC in regulating mitochondrial metabolism and calcium handling, our results suggest that non-selective HDAC inhibition may risk disrupting essential homeostatic functions mediated by MiDAC components. An important feature of the ELM cKO phenotype is its age-related onset of cardiomyopathy, with progressive decline in cardiac function manifesting over time. This observation raises critical considerations regarding the long-term use of HDAC inhibitors, particularly in chronic disease settings or in aging populations.

Importantly, the phenotypic and mechanistic findings from our ELM cKO mouse model closely parallel our previous observations in human iPSC-derived cardiomyocytes [43]. In both systems, Elmsan1 loss led to downregulation of key metabolic genes, impaired calcium handling, and disrupted contractile function. This strong cross-species concordance suggests that MiDAC regulates conserved molecular programs essential for cardiomyocyte health across species and developmental stages. It also suggests that the molecular and physiological alterations observed in our animal model may reflect conserved mechanisms relevant to human cardiomyopathy. Of note, the current in vivo model demonstrates that sustained Elmsan1 function is required to maintain cardiac metabolism and contractile function in the adult heart and that its loss leads to progressive, age-related cardiac dysfunction that cannot be fully captured in iPSC models.

Thus, this study elucidates the crucial role of Elmsan1 in cardiac function and disease. While early cardiac development and function appear unaffected by Elmsan1 deletion, its absence leads to significant age-related cardiac dysfunction and premature death. Late-onset cardiac dysfunction and lethal heart failure could be the consequence of an age-dependent requirement for Elmsan1 in maintaining cardiac metabolism and heart function.

Supplementary Material

Supplementary Figure S1-S8 can be found on FigShare: https://doi.org/10.6084/m9.figshare.29293754