Abstract

   Cardiomyopathy encompasses a range of diseases that severely affect the
   complex functions of the heart, involving structural and functional
   abnormalities, and is associated with high mortality. Recent studies
   have highlighted the critical role of ferroptosis in regulating
   oxidative stress and inflammation in cardiomyopathy. In this study, we
   established that the C6 S100A4+ SMCs subpopulation is critical by
   performing an integrated single-cell analysis of the known publicly
   available data [36]GSE145154. We validated the role of S100A4 in SMCs
   through in vitro experiments, providing evidence for its potential as a
   therapeutic target. Furthermore, these cells interact with endothelial
   cells through the PTN-NCL pathway, influencing disease progression. Key
   transcription factors, including KLF2, FOS, FOSB, and JUNB, were
   identified. This key subpopulation, along with its associated signaling
   pathways, marker genes, stemness genes, and transcription factors, may
   offer new insights for preventing the onset and progression of
   cardiomyopathy, particularly ischemic cardiomyopathy.

   Keywords: Single-cell RNA sequencing, Cardiomyopathy, Smooth muscle
   cells, Ferroptosis, Exosomes, Inflammation

   Subject terms: Cytokines, Immunochemistry, Ion channels, Metabolomics,
   Proteins, Structural biology, Bioinformatics, Cytological techniques,
   Immunological techniques, Metabolomics, Proteomic analysis, Sequencing,
   Cell adhesion, Cell death, Cell division, Cell growth, Cell migration,
   Cell signalling, Mechanisms of disease, Senescence, Cardiovascular
   biology, Biological techniques, Cell biology, Immunology, Physiology,
   Stem cells, Anatomy, Biomarkers, Cardiology, Health care, Medical
   research, Pathogenesis, Risk factors

Introduction

   Cardiomyopathy refers to a group of diseases characterized by diverse
   structural and functional changes in the heart, which often result in
   poor prognosis and a high risk of mortality^[37]1,[38]2. According to
   the classification system proposed by the American Heart Association
   (AHA) in the early twenty-first century, cardiomyopathies are divided
   into primary and secondary types. Primary categories include dilated,
   hypertrophic, and restrictive cardiomyopathies, while secondary causes
   encompass various etiologies such as ischemic, metabolic, infectious,
   toxic, autoimmune, and neuromuscular factors^[39]1. Dilated
   cardiomyopathy (DCM) is the most common form of non-ischemic
   cardiomyopathy and the most frequently encountered subtype
   overall^[40]3,[41]4. Restrictive cardiomyopathy (RM) comprises a
   heterogeneous group of disorders characterized by persistent
   restrictive pathophysiology, diastolic dysfunction, non-dilated
   ventricles, and atrial enlargement. These features persist regardless
   of ventricular wall thickness or systolic function, highlighting the
   diverse nature of RM^[42]5. Ischemic cardiomyopathy (ICM) is the most
   prevalent type of cardiomyopathy, primarily resulting from coronary
   atherosclerosis. Obstruction of blood flow due to atherosclerosis leads
   to myocardial hypoxia and subsequent cell death, triggering cardiac
   remodeling, including fibrosis and scar formation. Over time, the
   cumulative effects of hypoxia and fibrosis impair myocardial
   contractility, leading to cardiac hypertrophy, dilation, and,
   eventually, heart failure^[43]6,[44]7. Treatment of ICM primarily
   targets the underlying cause, with anatomical revascularization being
   the preferred method to address coronary atherosclerosis.
   Pharmacological therapies focus on managing risk factors for
   atherosclerosis, while surgical approaches such as coronary artery
   bypass grafting (CABG), endarterectomy, and percutaneous coronary
   intervention (PCI) are commonly employed to restore coronary blood flow
   and slow disease progression^[45]8–[46]11. Recent studies have
   illuminated the roles of ferroptosis and oxidative stress in the
   pathogenesis of atherosclerosis, which may contribute to the
   development of ICM. These processes also play a significant role in
   conditions like diabetes, which can further predispose individuals to
   DCM^[47]12. Furthermore, ferroptosis has been increasingly recognized
   as a key factor in the development of cardiomyopathy, atherosclerosis,
   and heart failure. Ferroptosis contributes to oxidative stress and
   inflammation, processes which are central to cardiovascular
   diseases^[48]13. Ferroptosis plays a critical role in promoting
   inflammation and oxidative stress, thereby contributing significantly
   to the formation of atherosclerotic plaques^[49]14.

   Additionally, emerging evidence suggests that the Bach1/HO-1 pathway
   regulates oxidative stress and promotes ferroptosis in
   cardiomyopathies^[50]15. Thus, oxidative stress likely contributes to
   the development of cardiomyopathies, arising from an imbalance between
   the generation of reactive oxygen species (ROS) and the capacity of the
   oxidative clearance system^[51]16. Currently, oxidative stress has been
   recognized as a prominent hallmark in the pathogenesis of heart failure
   resulting from cardiomyopathies^[52]17,[53]18. Oxidative stress is
   strongly linked to ICM.

   Oxidative stress is strongly linked to ICM. Vascular smooth muscle
   cells (VSMCs), in conjunction with fibroblasts and endothelial cells,
   play a crucial role in the pathogenesis of cardiomyopathies. VSMCs are
   integral to atherosclerotic plaque formation and contribute to
   pathological processes such as apoptosis and autophagy^[54]19. VSMCs
   also mediate vascular remodeling, a critical aspect of both vascular
   and myocardial changes^[55]20. Notably, a study by Ye et al.^[56]21
   demonstrated that VSMCs undergo ferroptosis, which promotes vascular
   calcification, contributing to the narrowing of coronary arteries and
   the progression of ICM. Additionally, the proliferation and migration
   of VSMCs are essential for vascular remodeling and are closely linked
   to the pathogenesis of atherosclerosis and other vascular
   diseases^[57]22. Therefore, smooth muscle cells (SMCs) may contribute
   to myocardial damage through various mechanisms, including ferroptosis,
   oxidative stress, proliferation, and migration, leading to the
   development of different cardiomyopathies. However, research
   specifically targeting SMCs in cardiomyopathies is limited. The
   relationship between SMCs and cardiomyopathies, particularly aspects
   such as cell stemness, transcription factor activity, and intercellular
   interactions, remains underexplored and warrants further investigation.

   This study aims to utilize single-cell RNA sequencing (scRNA-seq) to
   investigate the spectrum of SMCs involved in cardiomyopathies,
   particularly ICM, with the goal of providing a comprehensive
   understanding of their cellular heterogeneity. By identifying specific
   subpopulations closely associated with ICM onset and progression, this
   research seeks to integrate single-cell data related to cell cycle,
   differentiation, stemness, communication, and transcriptional
   regulation. The identification of potential therapeutic targets within
   these subpopulations could lead to novel strategies to intervene in the
   progression of cardiomyopathies and prevent heart failure.

Materials and methods

Data collection and processing

   The scRNA-seq data for smooth muscle cells in cardiomyopathy were
   obtained from the GEO database ([58]https://www.ncbi.nlm.nih.gov/geo/)
   under accession number [59]GSE145154. The numbers of the sample we
   selected are [60]GSM4307519, [61]GSM4307515, [62]GSM4307516,
   [63]GSM4307517, [64]GSM4307518, [65]GSM430752, [66]GSM4307520,
   [67]GSM4307521, [68]GSM4307522, [69]GSM4307523, [70]GSM4307529,
   [71]GSM4307525, [72]GSM4307526, [73]GSM4307527, [74]GSM4307528,
   [75]GSM4307530, [76]GSM4307531, [77]GSM4307532, [78]GSM4307533,
   [79]GSM4307539, [80]GSM4307536, [81]GSM4307538, [82]GSM4307544,
   [83]GSM4307540, [84]GSM4307541, [85]GSM4307543, [86]GSM4307534,
   [87]GSM4307545, [88]GSM4307546, [89]GSM4307549, [90]GSM4307550 and
   [91]GSM4307551. These samples correspond to scRNA-seq data from
   N-1-Bld, N-1-LVP, N-1-LVN, N-1-RVP, N-1-RVN, DCM-2-Bld, DCM-2-LVP,
   DCM-2-LVN, DCM-2-RVP, DCM-2-RVN, DCM-3-Bld, DCM-3-LVP, DCM-3-LVN,
   DCM-3-RVP, DCM-3-RVN, ICM-1-Bld, ICM-1-MIP, ICM-1-MIN, ICM-1-NMIP,
   ICM-1-NMIN, ICM-2-Bld, ICM-2-LVN, ICM-2-RVN, ICM-3-Bld, ICM-3-LVP,
   ICM-3-LVN, ICM-3-RVN, RM-36, RM-37, RM-40, RM-42 and RM-44. The 10X
   Genomics data for each sample were imported into R software (v4.3.3)
   using the Seurat package (v4.1.1)^[92]23,[93]24. To maintain data
   integrity, potential doublet cells were identified and removed using
   the DoubletFinder program (v2.0.3)^[94]25–[95]27. Furthermore, we
   applied a rigorous filtering process to exclude low-quality cells.
   Specifically, cells were retained only if they met the following
   quality thresholds: 300 < nFeature < 7500 and 500 < nCount < 100,000.
   Additionally, cells with mitochondrial gene expression exceeding 20% of
   the total expression were excluded. Consistent with our study’s
   criteria, cells with fewer than 500 or more than 6,000 identified genes
   were excluded as low-quality.

Dimensionality reduction, clustering, and cell type identification via
scRNA-seq

   Gene expression in each cell was quantified using the log(x + 1)
   transformation method, which scales the gene counts to a fraction of
   10,000, followed by a natural logarithm transformation^[96]28–[97]30.
   To ensure proper normalization, the expression matrix was normalized,
   and subsequently, the top 2000 highly variable genes (HVG) were
   identified^[98]31–[99]33. Principal Component Analysis
   (PCA)^[100]34,[101]35 was employed to analyze these genes, and to
   mitigate potential batch effects across samples, the Harmony method was
   applied^[102]36–[103]38. For subsequent dimensionality reduction and
   clustering, we selected the top 30 principal components (PCs). The
   results obtained were then visualized on a 2D plot using the UMAP
   algorithm^[104]39, facilitating cell type
   identification^[105]30,[106]40.

   To annotate the resulting cell clusters and identify distinct cell
   types, we utilized known cell markers from previous literature and the
   CellMarker database ([107]http://xteam.xbio.top/CellMarker/)^[108]41.
   These marker genes were employed to assign cell types to the clusters.
   Furthermore, for the SMCs identified through this process, additional
   clustering was performed to classify each SMC subpopulation based on
   specific marker genes.

Enrichment analysis and AUCell analysis

   Differentially expressed genes (DEGs) for each cell type were
   identified using the“FindMarkers”function^[109]40,[110]42 with default
   parameters, employing the Wilcoxon rank-sum test. DEGs were
   specifically selected from clusters showing a logFC value greater than
   0.25 and expressed in more than 25% of cells within the cluster. To
   gain further insights into the functional roles of these DEGs within
   each cell type, enrichment analysis was conducted using the
   clusterProfiler (v4.6.2) and SCP (v0.4.8) packages. Pathway enrichment
   analysis utilized Gene Ontology (GO)^[111]43,[112]44 Biological Process
   (BP) terms, revealing significantly enriched biological processes and
   molecular functions associated with the DEGs in each cell type. This
   analysis provided valuable information about the functional
   implications of these genes within their respective cellular
   contexts^[113]45–[114]49.

   Additionally, we integrated a novel method called AUCell to identify
   active gene sets within our scRNA-seq data^[115]50. AUCell offers a
   unique approach to assess the“activity”of specific gene sets within
   individual cells. Utilizing this method, we obtained outputs that
   provided valuable information about the activity levels of gene sets
   across our dataset.

Slingshot and CytoTRACE analysis

   To explore developmental and differentiation variations among distinct
   subpopulations of SMCs in cardiomyopathy, we utilized
   the“Slingshot”package (v2.6.0)^[116]51–[117]53 to delineate the
   developmental trajectories of each SMC subpopulation. Subsequently, we
   employed CytoTRACE analysis to rank the differentiation states of all
   SMC subpopulations^[118]54,[119]55. The
   functions“getlineage”and“getCurves”were employed to infer the
   differentiation trajectories of each SMC subpopulation and evaluate
   temporal changes in expression levels.

Construction of pseudotime trajectories for SMC subpopulations

   We utilized Monocle (v2.24.0) to construct pseudotime trajectories of
   SMC subpopulations. Through the analysis of pseudotemporal ordering in
   scRNA-seq data, Monocle aims to reveal the cellular changes occurring
   during the redifferentiation process of SMCs.

Cell–cell communication

   To visualize interactions among all cell types, including interactions
   between SMC subpopulations and endothelial cells, we utilized the
   CellChat package (v1.6.1) based on scRNA-seq data^[120]56,[121]57. The
   CellChatDB.human reference database for ligand-receptor interactions
   was employed. Intercellular interactions across different cell types
   were predicted, with a significance threshold set at p-value
   0.05^[122]58.

Utilization of SCENIC for gene regulatory network reconstruction

   To reconstruct gene regulatory networks and identify stable cell states
   from scRNA-seq data, we utilized the pySCENIC package (v0.12.1) in
   Python (v3.9.19) with default parameters^[123]59–[124]61. In this
   study, we generated AUCell matrices to assess the enrichment of
   transcription factors (TFs) and the activity of regulatory
   factors^[125]57.

Cell culture

   The HA-VSMC cell line was sourced from Procell in Wuhan, China. The
   cells were cultured in Endothelial Cell Medium (ECM, SCIENCELL)
   supplemented with 10% fetal bovine serum, penicillin (100 U/mL), and
   streptomycin (0.1 mg/mL). Cultures were maintained in a standard
   incubator at 37 °C, 5% CO2, and saturated humidity. Cells in the
   logarithmic growth phase were harvested for subsequent experiments.

SiRNA transfection

   For the experimental procedure, cells were seeded onto 6-well plates at
   a density of 2 × 10^5 cells per well. After 24 h, the cells were
   transfected with siRNAs obtained from GenePharma (Shanghai, China) at a
   final concentration of 20 µM. Transfection was performed using RNAiMax
   (Life Technologies, ThermoFisher Distributor; Brendale QLD, Australia)
   following the manufacturer’s instructions. Cells were collected 24 h
   post-transfection for further analysis. The sequences of the siRNAs
   used in the experiment were as follows: siRNA1: UCAACAAGUCAGAACUAAA;
   siRNA2: UCCAAGAGUACUGUGUCUU; NC siRNA: UUCUCCGAACGUGUCACGU.

RNA extraction and quantitative real-time PCR

   To isolate total RNA, TRIzol reagent from Thermo Fisher Scientific
   (Waltham, MA, USA) was used following the manufacturer’s instructions.
   Subsequently, 500 ng of RNA was used for cDNA synthesis using the
   PrimeScript RT Reagent Kit from TaKaRa (Tokyo, Japan). For quantitative
   real-time PCR, the SYBR® Premix Ex Taq™ from TaKaRa was utilized. PCR
   was conducted on the ABI V7 instrument from ABI (Indianapolis, IN,
   USA). The specific primers used for amplification were designed for the
   S100A4 gene as follows: Forward primer: 5’-CGAGGTGGACTTCCAAGAGT-3’ and
   Reverse primer: 5’-TCATTTCTTCCTGGGCTGCT-3’. In addition, the
   housekeeping gene we used was GAPDH, and its primer sequence was as
   follows: GAPDH-homo-qF: GGACTCATGACCACAGTCCA; GAPDH-homo-qR:
   TCAGCTCAGGGATGACCTTG.

Analysis of western blotting

   Lysis buffer (1 M Tris–HCL, pH 6.8, 10% SDS, and 80% glycerin) was used
   to remove all of the cellular protein. Following the manufacturer’s
   instructions, we utilized a bicinchoninic acid (BCA) test to find out
   how much protein was in the sample. In short, a 10% SDS-PAGE gel was
   used to separate the 30 μg of total protein, and then the proteins were
   transferred to polyvinylidene fluoride (PVDF) membranes (Millipore;
   Burlington, MA, USA) using electrophoresis. Next, the membranes were
   washed three times with Tris buffered saline plus Tween (TBST) and then
   completely blocked with 5% skim milk. We incubated the blocked PVDF
   membranes with primary antibodies overnight at 4 °C. Secondary
   antibodies were used for two hours after the original antibodies had
   been incubated. We used an enhanced chemiluminescence kit (ECL;
   ThermoFisher Scientific, Waltham, MA, USA) to check if we could see the
   protein bands.

Cell viability assay

   Cell viability was evaluated using the Cell Counting Kit-8 (CCK-8) from
   DOJINDO (Kumamoto, Japan). Cells were seeded at 1 × 10^3 cells per well
   in 96-well plates and cultured overnight. A 100 µL detection reagent
   was added to each well and incubated for 1 h. Absorbance at 450 nm was
   measured daily over 4 days. Growth curves were plotted by correlating
   OD450 values with time to assess cell viability.

Clone formation assay

   Logarithmically growing cell suspensions were diluted and plated into
   6-well plates at a density of 1 × 10^3 cells per well. The cells were
   cultured for 10 days with regular monitoring. Upon visible clone
   formation, the culture medium was removed, and cells were washed twice
   with ice-cold PBS. Subsequently, cells were fixed in 4%
   paraformaldehyde for 20 min, followed by staining with 0.1% crystal
   violet for 10 min. Colonies were then counted using the Gel imaging
   analysis system (Syngene, GBOX-F3EE; Bangalore, India).

Edu analysis

   Logarithmically growing cell suspensions were prepared, diluted, and
   plated into 6-well plates at a density of 1 × 10^3 cells per well. An
   EdU assay was conducted according to the manufacturer’s instructions
   (RiboBio, China). Treated cells were visualized using fluorescence
   microscopy, and quantification was performed by counting at least 6
   random fields.

Wound healing assay

   Cells were seeded at a density of 2 × 10^5 cells per well in 6-well
   plates and cultured overnight. A scratch was generated in the cell
   monolayer using a 10 µL pipette tip held perpendicular to the culture
   plate. The cells were then washed three times with PBS to remove any
   detached cells, and fresh serum-free medium was added to the plates.
   Subsequently, the cells were incubated at 37 °C in a 5% CO2 chamber.
   Bright-field microscopy was employed to capture images of the scratch
   area at 0 h and 48 h.

Transwell migration and invasion assay

   A Transwell assay was performed to evaluate cell migration and
   invasion. Cells were seeded in 24-well Transwell insert chambers (BD
   Biosciences, USA) with serum-free medium. For the invasion assay, the
   insert chambers were pre-coated with 2% Matrigel. The lower chamber
   medium was supplemented with 20% FBS, acting as a chemo-attractant.
   After a 2-day incubation period, non-migrated cells on the upper side
   of the membrane were gently removed. Cells that had migrated or invaded
   through the membrane were fixed and stained with crystal violet.
   Finally, the migrated or invaded cells were observed and photographed
   under a microscope.

Flow cytometry for apoptosis analysis

   To evaluate cellular apoptosis, the Annexin V-FITC/PI apoptosis
   detection kit from Sigma-Aldrich, Germany, was employed. Cells were
   incubated and washed with cold PBS, followed by suspension in a binding
   buffer containing Annexin V-FITC and PI stains. After incubation in the
   dark, the stained cells were analyzed using flow cytometry. Flow
   cytometry data was processed using FlowJo (v10.0.7) software to
   quantify the extent of apoptosis within the cell population.

Statistical analysis

   Statistical analysis was performed using R software (v4.3.3) and Python
   software (v3.9.19). Wilcoxon’s test and Pearson correlation coefficient
   were employed to assess the significance of differences between
   different groups. Significance levels were indicated as follows:
   *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. The term“ns”was
   used to denote non-significant differences between groups. These
   statistical tests and significance indicators were utilized to evaluate
   the statistical significance of the findings and provide confidence in
   the results.

Results

Overview of the study

   Cardiomyopathy is strongly associated with ferroptosis. The scRNA-seq
   of 4,184 SMCs found that C6 S100A4 + SMCs were critical, exhibiting
   high expression of ferroptosis, oxidative stress, and inflammation,
   which we demonstrated through in vitro experiments. In addition,
   pathways such as PTN-NCL may provide new treatments (Fig. [126]1).

Fig. 1.

   [127]Fig. 1
   [128]Open in a new tab

   Principles of research.

Identification and characterization of distinct smooth muscle cell
subpopulations in cardiomyopathy

   Following batch correction, The sample sources of high-quality cells
   are shown in Figure [129]S1A. we categorized 179,927 high-quality cells
   into various cell types, including T cells, NK cells, endothelial cells
   (ECs), myeloid cells, fibroblasts, pericytes, SMCs, B cells, plasma
   cells, cardiomyocytes (CMs), proliferating cells, EndoCs, and
   endothelial progenitor cells (EPCs). These samples were stratified into
   four groups based on cardiomyopathy types: DCM, ICM, RM, and normal
   (N). Utilizing UMAP plots, we observed that SMCs were predominantly
   associated with ICM and DCM cases (Fig. [130]2A). All high-quality
   cells were in different cell cycle stages(Figure [131]S1B). Further
   batch correction enabled the identification of 4,184 SMCs from 24
   samples, which were segmented into nine distinct subpopulations
   characterized by specific marker genes: C0 SOCS3+ SMCs, C1 CD36+ SMCs,
   C2 SORBS2+ SMCs, C3 RGS5+ SMCs, C4 ACTG2+ SMCs, C5 IGF2+ SMCs, C6
   S100A4+ SMCs, C7 C12orf75+ SMCs, and C8 EGR1+ SMCs (Figs. [132]2B and
   [133]C). All subpopulations exhibited differences in cell cycle
   distribution, with the majority of cells being in the G1 phase (Figure
   [134]S1C). Each subpopulation exhibited unique highly expressed genes
   alongside their defining markers. For example, C6 S100A4+ SMCs
   demonstrated elevated expression of ID4, ID2, SERPINF1, and ACTG2,
   while C4 ACTG2+ SMCs displayed heightened levels of PLN, SORBS2, CKB,
   MFAP4, CLU, and GADD45G (Fig. [135]2D). S100A4 was expressed at the
   highest level in the C6 subpopulation, and was therefore selected as
   the marker gene (Figures [136]S1D-E). Analysis of cell cycle phases
   within these SMC subpopulations revealed a predominant presence in the
   G1 phase, with C6 S100A4+ SMCs showing the highest proportion in G1 and
   the lowest in G2M phases. Notably, C6 S100A4+ SMCs were predominantly
   derived from the ICM group, with other subpopulations also exhibiting
   significant contributions from ICM. However, C0 SOCS3+ SMCs showed
   nearly equal proportions from DCM, ICM, and N groups (Fig. [137]2E).
   Additionally, C6 S100A4+ SMCs exhibited the highest nFeature_RNA and
   nCount_RNA values, followed by C0 SOCS3+ SMCs, whereas C8 EGR1+ SMCs
   displayed the lowest values (Fig. [138]2F). Volcano plots were utilized
   to visualize top upregulated and downregulated genes in each SMC
   subpopulation. For instance, in C6 S100A4+ SMCs, top upregulated genes
   included S100A4, SERPINF1, INHBA, SFTA1P, and ID4, while top
   downregulated genes encompassed FABP4, IGF2, AQP1, TESC, and TIMP3.
   Similarly, other subpopulations exhibited distinct gene expression
   profiles (Fig. [139]2G). To gain insights into biological processes
   associated with each subpopulation in cardiomyopathy progression, we
   conducted GO-BP enrichment analysis. Genes with high Z-score
   differentials in C6 S100A4 + SMCs, such as ID1, ID4, ID2, FN1, COL14A1,
   COL18A1, S100A4, SERPINF1, and APOE, were significantly enriched in
   processes such as regulation of neurogenesis, regulation of nervous
   system development, positive regulation of neurogenesis, circadian
   regulation of gene expression, and positive regulation of nervous
   system development (Fig. [140]2H).

Fig. 2.

   [141]Fig. 2
   [142]Open in a new tab

   Visualization of smooth muscle cell subpopulations in cardiomyopathy
   (A) The spatial distribution of all cells in the selected
   cardiomyopathy samples (left) and their distribution across different
   groups (right) were depicted in a 3D UMAP plot. (B) The UMAP plot
   illustrated the specific distribution of 24 samples selected from
   [143]GSE145154. (C) Smooth muscle cells were categorized into nine
   subpopulations based on differential expression of marker genes. The
   UMAP plot displayed their distributions and provided information about
   the proportions of different groups and cell cycles within each
   subpopulation. (D) Average expression levels of differentially
   expressed genes in the nine smooth muscle cell subpopulations were
   presented in a bubble plot. (E) Pie charts directly showed the
   proportions of cell cycles and different groups within each
   subpopulation on the UMAP plot. (F) Box plots illustrated the
   nFeature_RNA and nCount_RNA of the nine smooth muscle cell
   subpopulations. (G) Volcano plots depicted the top 5 upregulated and
   top 5 downregulated genes in the nine smooth muscle cell
   subpopulations. (H) Highly expressed differential genes in the nine
   smooth muscle cell subpopulations were subjected to GO-BP enrichment
   analysis.

Differences in stemness and oxidative phosphorylation among SMC
subpopulations, with a focus on C6 S100A4 + SMCs

   To explore the proliferative characteristics of SMC subpopulations, we
   conducted a stemness scoring analysis. Initially, we visualized the
   distribution of stemness AUC values across the nine SMC subpopulations
   using UMAP and facet plots (Fig. [144]3A). Box plots were employed for
   a comprehensive comparison, revealing that C6 S100A4+ SMCs and C0
   SOCS3+ SMCs displayed the highest stemness AUC values (Fig. [145]3B).

Fig. 3.

   [146]Fig. 3
   [147]Open in a new tab

   Stemness and metabolic activity of smooth muscle cell subpopulations
   (A) The distribution of stemness AUC values across the nine smooth
   muscle cell subpopulations was visualized using UMAP plots and faceted
   plots. (B) Box plots provided a clear comparison of stemness AUC values
   among the nine smooth muscle cell subpopulations. (C) Violin plots
   illustrated the differential expression of stemness genes across the
   nine smooth muscle cell subpopulations. (D) UMAP plots displayed the
   expression and density distribution of five highly expressed stemness
   genes (EPAS1, CTNNB1, MYC, HIF1A, NES) across all smooth muscle cell
   subpopulations. (E) Heatmaps depicted the AUC values of the top 20
   metabolic pathways in C6 S100A4+ SMCs. (F) Heatmaps presented the AUC
   values of the top 20 metabolic pathways across the four groups. (G)
   UMAP plots showed the density distribution of the oxidative
   phosphorylation pathway. (H) UMAP plots displayed the expression
   intensity distribution of the oxidative phosphorylation pathway. (I)
   Box plots visually represented differences in the expression levels of
   the oxidative phosphorylation metabolic pathway among the nine smooth
   muscle cell subpopulations.

   Furthermore, we identified several genes with elevated stemness
   expression in these SMC subpopulations, including EPAS1, CTNNB1, MYC,
   HIF1A, and NES (Fig. [148]3C). UMAP plots were then used to illustrate
   the distribution and density of AUC values for these five stemness
   genes within the subpopulations (Fig. [149]3D). Notably, strong
   cellular stemness often correlates with heightened metabolic activity,
   indicative of high proliferative capacity, active cell growth, and
   division. Therefore, we performed AUCell scoring on the metabolic
   pathways of C6 S100A4 + SMCs and represented the top 20 pathways in a
   heatmap. Our analysis underscored oxidative phosphorylation as the most
   active metabolic pathway in C6 S100A4+ SMCs (Fig. [150]3E).

   To further explore the role of oxidative phosphorylation across
   different groups, we conducted AUCell scoring of this pathway. The
   results indicated that the N group exhibited the highest scores,
   followed by the DCM and ICM groups (Fig. [151]3F). UMAP plots were
   utilized to visualize the density and expression levels of oxidative
   phosphorylation within the SMC subpopulations (Figs. [152]3G and
   [153]H). Finally, comparing the rankings of oxidative phosphorylation
   scores among all SMC subpopulations revealed that C2 SORBS2+ SMCs had
   the highest scores, followed by C6 S100A4+ SMCs, while C8 EGR1+ SMCs
   exhibited the lowest scores (F[154]ig. [155]3I).

Differences in ferroptosis and oxidative stress among SMC subpopulations

   To investigate the potential association between SMC subpopulations and
   the development of cardiomyopathy, particularly ICM, we examined the
   roles of ferroptosis and oxidative stress, both of which are implicated
   in coronary atherosclerosis and subsequent cardiomyopathy. We conducted
   scoring for ferroptosis and oxidative stress-related pathways across
   all SMC subpopulations.

   Initially, we visualized the distribution of ferroptosis AUC values
   across the nine SMC subpopulations using UMAP and facet plots,
   segmented by different cell cycle phases and groups (Fig. [156]4A).
   Overlaying these values onto a unified UMAP plot with different shapes
   representing distinct groups, we observed that ferroptosis AUC values
   were notably elevated in C4 ACTG2+ SMCs, C6 S100A4+ SMCs, and C0
   SOCS3+ SMCs (Fig. [157]4B). Differences in ferroptosis scores across
   cell cycle phases were minimal, with slightly higher scores observed
   during G1 and S phases compared to the G2M phase (Fig. [158]4C).
   Similarly, variations in scores across the four groups (N, DCM, ICM,
   RM) were minor, with slightly higher scores in the N and RM groups
   (Fig. [159]4D).

Fig. 4.

   [160]Fig. 4
   [161]Open in a new tab

   Differences in ferroptosis and oxidative stress scores across all
   smooth muscle cell subpopulations (A) UMAP plots were used to visualize
   the distribution of ferroptosis AUC values. Faceted plots provided
   detailed insights into the distribution differences of ferroptosis AUC
   values within each subpopulation, across different groups, and during
   different cell cycles. (B) Detailed variations in ferroptosis scores
   were illustrated, highlighting specific groups within each
   subpopulation. Comparisons of ferroptosis scores among the nine smooth
   muscle cell subpopulations were also presented. (C) Variations in
   ferroptosis scores across different cell cycles were displayed. (D)
   Variations in ferroptosis scores across different groups were
   illustrated. (E) UMAP plots depicted distribution differences of four
   oxidative stress-related scores. (F) Box plots visually compared the
   differences in the four oxidative stress-related scores across the
   different smooth muscle cell subpopulations.

   Next, we assessed four gene sets related to oxidative stress pathways
   across all SMC subpopulations. These pathways included response to
   oxidative stress, intrinsic apoptotic signaling pathway in response to
   oxidative stress, positive regulation of oxidative stress-induced
   intrinsic apoptotic signaling pathway, and negative regulation of
   response to oxidative stress. Using UMAP plots, we depicted the
   distribution and strength of pathway scores across different
   subpopulations and groups, where shapes represented different groups
   (Fig. [162]4E). Box plots revealed that C6 S100A4+ SMCs exhibited the
   highest scores in response to oxidative stress, intrinsic apoptotic
   signaling pathway in response to oxidative stress, and positive
   regulation of oxidative stress-induced intrinsic apoptotic signaling
   pathway (Fig. [163]4F). Interestingly, C6 S100A4 + SMCs, along with C4
   ACTG2+ SMCs and C7 C12orf75+ SMCs, showed the lowest scores in negative
   regulation of response to oxidative stress.

Slingshot analysis reveals developmental and differentiation characteristics
of SMC subpopulations

   To gain insights into the developmental stages and differentiation
   levels of SMC subpopulations, we employed Slingshot analysis to model
   developmental trajectories, identifying three distinct SMC lineage
   trajectories: Lineage 1, Lineage 2, and Lineage 3. Lineage 1: C1
   CD36+ SMCs → C0 SOCS3+ SMCs → C6 S100A4+ SMCs → C4 ACTG2+ SMCs → C2
   SORBS2+ SMCs → C8 EGR1+ SMCs → C5 IGF2+ SMCs → C1 CD36+ SMCs; Lineage
   2: C1 CD36+ SMCs → C0 SOCS3+ SMCs → C6 S100A4+ SMCs → C4
   ACTG2+ SMCs → C2 SORBS2+ SMCs → C3 RGS5+ SMCs; Lineage 3: C4
   ACTG2+ SMCs → C6 S100A4+ SMCs → C0 SOCS3+ SMCs → C1 CD36+ SMCs → C5
   IGF2+ SMCs → C3 RGS5+ SMCs → C7 C12orf75+ SMCs. In our analysis, we
   utilized UMAP plots to visualize the sequential expression of SMC
   lineage trajectories across all subpopulations, aligning them with a
   temporal order observed in another UMAP plot (Fig. [164]5A). This
   allowed us to observe how different subpopulations progressed along
   distinct differentiation paths.

Fig. 5.

   [165]Fig. 5
   [166]Open in a new tab

   Slingshot analysis of smooth muscle cell subpopulations (A) Three
   fitted differentiation trajectories of smooth muscle cell
   subpopulations were depicted. The temporal order and distribution of
   these trajectories were further validated using UMAP plots. The
   trajectories were as follows: Lineage 1:
   C1 → C0 → C6 → C4 → C2 → C8 → C5 → C1; Lineage 2:
   C1 → C0 → C6 → C4 → C2 → C3; Lineage 3:
   C4 → C6 → C0 → C1 → C5 → C3 → C7. (B) Heatmaps illustrated the
   differential expression of genes along the three trajectories,
   accompanied by their GO-BP enrichment analysis. (C) The differential
   expression of marker genes in the nine smooth muscle cell
   subpopulations was shown as they progressed through the three
   trajectories. (D) The differential expression of the five
   aforementioned stemness genes (EPAS1, CTNNB1, MYC, HIF1A, NES) was
   presented as they evolved along the three trajectories.

   Next, we investigated differential gene expression along the three
   identified differentiation trajectories, focusing particularly on the
   characteristics of C6 S100A4+ SMCs within these trajectories. Our
   exploration revealed several genes closely associated with C6
   S100A4+ SMCs, including IER2, C11orf96, ZFP36, PPP1R15A, BTG2, SERTAD1,
   SOCS3, and EGR1.

   Furthermore, we conducted GO-BP enrichment analysis on the
   differentially expressed genes related to SMC subpopulations
   (Fig. [167]5B). This analysis highlighted enrichment in key biological
   processes such as radiation response, mRNA processing, nuclear
   transcription, proliferation, stress response, transcription
   regulation, muscle development, cell death, and mRNA destabilization,
   underscoring their relevance in SMC differentiation and function.
   Additionally, we explored the expression timing of nine selected genes
   across the three lineages. Notably, S100A4 exhibited early expression
   in Lineage 3 and mid-stage expression in Lineages 1 and 2
   (Fig. [168]5C).

   Here are the expression patterns observed: EPAS1: Highly expressed
   early in Lineage 1 and in the mid-stage of Lineage 3. CTNNB1: Shows a
   similar expression pattern across the three lineages, with slightly
   higher mid-stage expression. MYC: Primarily expressed early in all
   three lineages, with significantly higher expression compared to other
   stages. HIF1A: Predominantly expressed early in Lineage 3, with a
   noticeable downregulation later; slightly upregulated mid-stage in
   Lineages 1 and 2 before downregulation. NES: Upregulated mid-stage in
   Lineage 3 before significant downregulation; early and late
   upregulation in Lineages 1 and 2 with mid-stage downregulation
   (Fig. [169]5D).

Pseudotime analysis of SMC subpopulations

   To validate the findings from Slingshot analysis, we utilized CytoTRACE
   to assess differentiation levels among SMC subpopulations. CytoTRACE
   results confirmed that C6 S100A4 + SMCs exhibited the lowest
   differentiation level, while C3 RGS5 + SMCs showed the highest
   (Figs. [170]6A and [171]B).

Fig. 6.

   [172]Fig. 6
   [173]Open in a new tab

   CytoTRACE analysis and pseudotime trajectory validation of smooth
   muscle cell subpopulations (A) UMAP plots displayed the results of
   CytoTRACE analysis conducted on smooth muscle cell subpopulations. (B)
   Based on the CytoTRACE results, the differentiation levels of the nine
   smooth muscle cell subpopulations were visualized through box plots.
   (C) UMAP plots showcased the distribution of pseudotime. (D) A
   two-dimensional plot further elucidated the pseudotime trajectory,
   progressing from left to right with two branches in the middle phase
   and one branch in the late phase. (E) Due to branching in the
   pseudotime trajectory, the entire pseudotime was divided into seven
   stages (State 1–7), and their distribution across all smooth muscle
   cell subpopulations was displayed using UMAP plots. (F) A dedicated
   two-dimensional plot specifically depicted the distribution of State
   1–7 along the established pseudotime trajectory. (G) Ridge plots
   displayed the distribution of the nine smooth muscle cell
   subpopulations on the pseudotime trajectory. (H) Box plots provided a
   more intuitive quantification of the expression levels of the nine
   smooth muscle cell subpopulations on the pseudotime trajectory. (I) A
   two-dimensional plot visualized the specific distribution of the nine
   smooth muscle cell subpopulations on the established pseudotime
   trajectory. (J) The distribution characteristics of each smooth muscle
   cell subpopulation on the trajectory were individually displayed in
   sequence. (K) Visualization of the expression patterns of marker genes
   for each smooth muscle cell subpopulation over time on the pseudotime
   trajectory. (L) Pie charts on UMAP plots revealed the proportions of
   different states within each smooth muscle cell subpopulation. (M) Bar
   graphs depicted the proportions of different smooth muscle cell
   subpopulations within each state. (N) Bar graphs showed the percentage
   of each state occupied by each smooth muscle cell subpopulation.

   Subsequently, we constructed a pseudotime trajectory using Monocle and
   visualized it via a UMAP plot, illustrating the temporal progression of
   each SMC subpopulation (Fig. [174]6C). The detailed trajectory
   (Fig. [175]6D) revealed three branches as time unfolds. Based on these
   branches, we segmented the trajectory into seven states, shown in
   Fig. [176]6F, and mapped these states onto the UMAP plot depicting SMC
   subpopulations (Fig. [177]6E).

   To examine the positional differences of all SMC subpopulations along
   the pseudotime trajectory, we generated a ridge plot (Fig. [178]6G).
   This analysis indicated that C0 SOCS3+ SMCs, C4 ACTG2+ SMCs, C6
   S100A4+ SMCs, and C8 EGR1+ SMCs primarily occupied early pseudotime
   stages, while C1 CD36+ SMCs, C2 SORBS2+ SMCs, and C3 RGS5+ SMCs were
   predominantly in the middle to late stages, and C7 C12orf75+ SMCs were
   mainly in the latest stage. These findings were further validated by
   box plots (Fig. [179]6H) and illustrated the distribution across
   pseudotime (F[180]ig. [181]6I), as well as individually showcasing each
   SMC subpopulation (Fig. [182]6J).

   We further examined the expression patterns of nine specified genes
   across pseudotime in SMC subpopulations (Fig. [183]6K). These genes
   exhibited diverse expression dynamics, with some upregulated at
   pseudotime extremes and downregulated in the middle. Additionally, we
   mapped the seven defined states onto the UMAP plot, using pie charts to
   display the distribution of each SMC subpopulation within these states
   (Fig. [184]6L), further substantiated by proportions shown in a bar
   graph (Fig. [185]6M). In addition, we used a bar chart to show the
   proportion of state1-7 in each subpopulation (The sum of the
   proportions of each state in each subpopulation is 1) (Fig. [186]6N).

Interaction between C6 S100A4 + SMCs and ECs via the PTN-NCL signaling
pathway

   To explore the potential interaction between C6 S100A4+ SMCs and ECs,
   we employed CellChat for intercellular interaction analysis.
   Figure [187]7A presents a comprehensive overview of interaction
   quantity and strength among all cell types in myocardial disease
   samples. Our focus was on C6 S100A4+ SMCs as signal senders interacting
   with other cells. Notably, C6 S100A4+ SMCs showed robust interactions
   with T cells, NK cells, myeloid cells, and proliferating cells
   (Fig. [188]7B).

Fig. 7.

   [189]Fig. 7
   [190]Open in a new tab

   Cell communication landscape in cardiomyopathy (A) The left plot
   depicted the abundance of interactions among all cells in
   cardiomyopathy, represented as a circle diagram. The right plot
   visualized the intensity of these interactions among all cells, also
   shown as a circle diagram. (B) In the left plot, the quantity of
   interactions was shown when C6 S100A4+ SMCs acted as signal emitters to
   all other cells. The right plot showcased the strength of these
   interactions when C6 S100A4+ SMCs acted as signal emitters to all other
   cells. (C) The upper plot displayed the quantity of interactions when
   ECs acted as signal receivers from all other cells. The lower plot
   presented the strength of interactions when ECs acted as signal
   receivers from all other cells. (D) Bubble charts depicted the
   expression levels of distinct proteins when all cells acted as signal
   receivers (top) and when all cells functioned as signal emitters
   (bottom). (E) Chord diagrams illustrated the patterns of interaction
   among all cells within the PTN signaling pathway network. (F) Chord
   diagrams demonstrated the differential expression of receptor proteins
   when ECs acted as signal receivers from all smooth muscle cell
   subpopulations. (G) The interaction patterns among all cells within the
   PTN-NCL receptor-based cell interaction pathway. (H) Bubble charts
   depicted the stable expression levels of PTN protein in C6 S100A4+ SMCs
   and NCL protein in ECs, respectively. (I) Bubble charts presented the
   expression levels of various receptor proteins when C6 S100A4+ SMCs
   interacted with other cells (excluding other smooth muscle cell
   subpopulations).

   Subsequently, we examined ECs as signal receivers. ECs received signals
   from various cell types except CMs and C7 C12orf75+ SMCs. Importantly,
   ECs received prominent and intense signals from C6 S100A4+ SMCs, C0
   SOCS3+ SMCs, and fibroblasts (Fig. [191]7C), underscoring a significant
   relationship between C6 S100A4+ SMCs and ECs. To delve deeper into the
   interaction between C6 S100A4+ SMCs and ECs, we investigated protein
   pathways involved. A bubble plot (Fig. [192]7D) visualized protein
   expression differences when cells acted as signal senders and
   receivers. Notable proteins highly expressed in ECs included CD99, APP,
   MIK, VEGF, ITGB2, PECAM1, ANGPTL, SELL, CDH5, and JAM, whereas C6
   S100A4+ SMCs exhibited high expression of COLLAGEN, MIF, APP, PTN,
   ESAM, THBS, GAS, NOTCH, PDGF, and VEGF.

   Further analysis focused on the PTN signaling pathway network
   (Figs. [193]7E and [194]F), revealing a specific interaction between
   PTN from C6 S100A4+ SMCs and NCL protein on ECs, confirmed through
   visualizations. A chord diagram (Fig. [195]7G) depicted the PTN-NCL
   signaling pathway, illustrating how C6 S100A4+ SMCs interact with ECs.
   Additionally, bubble plots (Fig. [196]7H and [197]I) verified PTN
   expression in C6 S100A4 + SMCs and NCL presence in ECs, highlighting
   the PTN-NCL pathway among various signaling pathways through which
   these cells interact.

Gene regulatory network analysis of SMC subpopulations

   To identify core TFs associated with myocardial disease in SMC
   subpopulations, we utilized PySCENIC for analysis. Initially, we
   determined the top5 TFs for each SMC subpopulation (Fig. [198]8A). For
   C6 S100A4+ SMCs, these TFs were identified as KLF2, FOS, JUN, FOSB, and
   ATF3. We then ranked regulatory factors based on their regulon
   specificity score (RSS) and highlighted specific subpopulations with
   red dots on the UMAP plot. Additionally, we illustrated the
   distribution of binary regulon activity scores (RAS) of the top
   regulons on the UMAP plot with green dots, consistent with the heatmap
   (Fig. [199]8B).

Fig. 8.

   [200]Fig. 8
   [201]Open in a new tab

   Gene regulatory network analysis of smooth muscle cell subpopulations
   in cardiomyopathy (A) Heatmaps displayed the differential expression of
   the top 5 TFs in the nine smooth muscle cell subpopulations. (B)
   Regulators in smooth muscle cell subpopulations associated with
   cardiomyopathy were ranked based on the Regulator Specificity Score
   (RSS) (left). Each smooth muscle cell subpopulation was highlighted
   with a red dot in the UMAP plots (middle). The distribution of the
   highest regulatory subunits, mapped on UMAP plots based on the binary
   RAS (normalized using Z-score across all samples, with a threshold of
   2.5 set to convert to 0 or 1), was highlighted with green dots. (C)
   UMAP plots illustrated the expression levels of EGR1, NR2F2, HEY2,
   IRX1, and KLF2 in the smooth muscle cell subpopulations. (D) UMAP plots
   depicted the distribution of AUC values for EGR1, NR2F2, HEY2, IRX1,
   and KLF2 in the smooth muscle cell subpopulations. (E) Box plots
   quantified the differential expression levels of the top 5 highly
   active regulatory subunits in the nine smooth muscle cell
   subpopulations. (F) Box plots visually compared the AUC values of the
   aforementioned 5 regulators in the nine smooth muscle cell
   subpopulations. We used R software (v.4.3.3)
   ([202]https://cran.r-project.org/) to visualize the results calculated
   by the pySCENIC module (v0.12.1) based on Python (v3.9.19)
   ([203]https://www.python.org/). Finally, we integrated all the images
   in Adobe Illustrator (2024)
   ([204]https://www.adobe.com/products/illustrator.html).

   In summary, the top 5 active TFs identified were EGR1, NR2F2, HEY2,
   IRX1, and KLF2. Notably, KLF2 emerged as the most active TF in both C4
   ACTG2+ SMCs and C6 S100A4+ SMCs. NR2F2 showed prominence as the top TF
   in both C1 CD36+ SMCs and C5 IGF2 + SMCs. IRX1 was notably active in C3
   RGS5+ SMCs, C7 C12orf75+ SMCs, and C8 EGR1+ SMCs.

   We displayed the expression levels and density distribution of these 5
   TFs sequentially on the UMAP plots (Fig. [205]8C) alongside their
   distribution of AUC values (Fig. [206]8D). To compare expression
   intensity differences among these TFs in each subpopulation, we used
   boxplots. EGR1 exhibited higher expression in C0 SOCS3+ SMCs, C8
   EGR1+ SMCs, C4 ACTG2+ SMCs, and C6 S100A4+ SMCs. NR2F2 showed higher
   expression in C1 CD36+ SMCs, C5 IGF2+ SMCs, C0 SOCS3+ SMCs, and C6
   S100A4+ SMCs. HEY2 was predominantly expressed in C4 ACTG2+ SMCs and C6
   S100A4+ SMCs, with relatively lower levels in other subpopulations.
   IRX1 exhibited minimal expression across all SMC subpopulations. KLF2
   displayed higher expression in C4 ACTG2+ SMCs and was also detected in
   C0 SOCS3+ SMCs and C6 S100A4+ SMCs (Fig. [207]8E).

   Regarding AUC values (Fig. [208]8F), EGR1 showed the highest in C0
   SOCS3+ SMCs, followed by C4 ACTG2+ SMCs and C6 S100A4+ SMCs. NR2F2 had
   the highest AUC in C1 CD36+ SMCs, followed by C5 IGF2+ SMCs. HEY2
   exhibited relatively high AUC values in C0 SOCS3+ SMCs, C4 ACTG2+ SMCs,
   C6 S100A4+ SMCs, and C2 SORBS2+ SMCs. IRX1 had the highest AUC in C8
   EGR1+ SMCs, followed by C3 RGS5+ SMCs and C7 C12orf75+ SMCs. KLF2
   showed relatively high AUC values in C4 ACTG2+ SMCs, C6 S100A4+ SMCs,
   and C0 SOCS3+ SMCs.

Identification of TF regulatory modules in cardiomyopathy SMCs

   Using SCENIC, we identified regulatory modules and rules in SMC
   subpopulations associated with cardiomyopathy. Based on AUCell scores’
   similarity, we categorized these rules into two main modules, M1 and M2
   (Fig. [209]9A). Mapping the average activity scores of each module onto
   the UMAP plot revealed distinct distributions across various SMC
   subpopulations, as depicted in the facet plot. From the boxplot
   analysis, M1 prominently featured C0 SOCS3+ SMCs, C4 ACTG2+ SMCs, and
   C6 S100A4+ SMCs, while M2 included fewer subpopulations, with slightly
   higher proportions observed in C8 EGR1+ SMCs, C3 RGS5+ SMCs, and C5
   IGF2+ SMCs.

Fig. 9.

   [210]Fig. 9
   [211]Open in a new tab

   Identification of transcription factor regulatory modules in
   cardiomyopathy smooth muscle cells (A) The heatmap displayed the
   identification of regulatory submodules in smooth muscle cell
   subpopulations associated with cardiomyopathy based on SCENIC
   regulatory rules modules and AUCell scores similarity. Eventually, two
   rule-based submodules were identified. (B, C) UMAP plots and faceted
   plots showed the distribution of AUC scores of two rule-based
   submodules (M1, M2) and their distribution across various smooth muscle
   cell subpopulations (left). Box plots intuitively compared the
   expression levels of M1 and M2 in each smooth muscle cell subpopulation
   (middle). Scatter plots provided a better understanding of the
   expression score rankings of M1 and M2 in each smooth muscle cell
   subpopulation (right). (D) Scatter plot displayed the ranking of
   different TFs in M1 based on their variation scores. (E) Visualization
   of the 5 TFs with high variation scores in M1. UMAP plots illustrated
   the expression distribution of JUNB, ATF3, FOS, FOSB, and JUN, while
   box plots compared their expression differences among the nine smooth
   muscle cell subpopulations. (F) Scatter plot depicted the variation
   scores of different TFs in M2. (G) Visualization of the 2 TFs with high
   ranking in M2. UMAP plots respectively showed the expression
   distribution of these TFs, while box plots compared their expression
   differences among the nine smooth muscle cell subpopulations. (H)
   Scatter plots showed the ranking of the main cell cycle phases in M1
   and M2. (I) Scatter plots showed the ranking of the groups to which
   cells belonged in M1 and M2.

   We ranked the SMC subpopulations within M1 and M2 based on their
   transcription factor activity scores. In M1, the ranking was C0
   SOCS3+ SMCs > C6 S100A4+ SMCs > C4 ACTG2+ SMCs > C1 CD36+ SMCs > C8
   EGR1+ SMCs > C5 IGF2+ SMCs > C2 SORBS2+ SMCs > C3 RGS5+ SMCs > C7
   C12orf75+ SMCs. In M2, the ranking was C8 EGR1+ SMCs > C3
   RGS5+ SMCs > C7 C12orf75+ SMCs > C1 CD36+ SMCs > C5 IGF2+ SMCs > C0
   SOCS3+ SMCs > C2 SORBS2+ SMCs > C4 ACTG2+ SMCs > C6 S100A4+ SMCs
   (Fig. [212]9B and [213]C).

   We visualized the TFs in M1 and M2. In M1, TFs with high variance
   scores included JUNB, ATF3, FOS, KLF2, FOSB, JUN, and EGR1
   (Fig. [214]9D). In contrast, TFs in M2 generally exhibited lower
   variance scores, although IRX1, E2F1, and GATA3 showed relatively
   higher levels of activity (Fig. [215]9F).

   We depicted the expression levels and density distribution of JUNB,
   ATF3, FOS, FOSB, and JUN in M1 on the UMAP plot. From the boxplot, we
   observed that JUNB was highly expressed in C0 SOCS3+ SMCs, C4
   ACTG2+ SMCs, C8 EGR1+ SMCs, and C6 S100A4+ SMCs; ATF3 was highly
   expressed in C0 SOCS3+ SMCs, C4 ACTG2+ SMCs, and C6 S100A4+ SMCs; FOS
   showed high expression in C0 SOCS3+ SMCs, C4 ACTG2+ SMCs, C8
   EGR1+ SMCs, and C6 S100A4+ SMCs; FOSB exhibited high expression in C8
   EGR1+ SMCs, C0 SOCS3+ SMCs, C6 S100A4+ SMCs, and C4 ACTG2+ SMCs; JUN
   displayed high expression in C8 EGR1+ SMCs, C0 SOCS3+ SMCs, C4
   ACTG2+ SMCs, and C6 S100A4+ SMCs (Fig. [216]9E).

   Similarly, we mapped the AUC value distribution of E2F1 and GATA3 onto
   the UMAP plot. From the boxplot, E2F1 showed lower AUC values across
   all SMC subpopulations, whereas GATA3 exhibited slightly higher values
   in C1 CD36+ SMCs, C3 RGS5+ SMCs, and C5 IGF2+ SMCs (Fig. [217]9G).

   Furthermore, we ranked cell cycle phases and groups based on
   transcription factor activity scores (Fig. [218]9H and [219]I). In M1,
   the order was G2M phase > G1 phase > S phase, N > DCM > ICM > RM. In
   M2, the order was S phase > G2M phase > G1 phase, RM > ICM > DCM > N.

Lowering S100A4 suppresses HA-VSMC cell proliferation

   To elucidate the role of S100A4 in myocardial disease, particularly its
   impact on smooth muscle cell behavior in ischemic cardiomyopathy (ICM),
   we conducted rigorous in vitro experiments. Initially, we quantified
   S100A4 expression levels using qRT-PCR after 24-h transfection with
   siRNA in the HA-VSMC cell line, confirming effective S100A4 knockdown
   (Fig. [220]10A). Next, we verified the relative protein expression
   after S100A4 knockdown, and it turned out that S100A4 protein
   expression was significantly reduced (Fig. [221]10B). Subsequently,
   CCK8 assays unequivocally demonstrated a significant decrease in
   HA-VSMC viability following S100A4 knockdown (P < 0.001)
   (Fig. [222]10C). Colony formation assays further supported these
   findings, showing a marked reduction in colony counts in both S100A4
   knockdown groups compared to the non-knockdown control group (NC)
   (Fig. [223]10D). The observed slower rate of colony formation in cells
   with S100A4 knockdown strongly suggests the crucial role of S100A4 in
   promoting HA-VSMC proliferation. Furthermore, results from EdU staining
   experiments provided compelling evidence that downregulation of S100A4
   expression hindered the proliferation of HA-VSMC cells relative to the
   NC group (Fig. [224]10E).

Fig. 10.

   [225]Fig. 10
   [226]Open in a new tab

   Knockdown of S100A4 suppresses the proliferation ability of HA-VSMCs
   (A) Relative expression of S100A4 in Si-S100A4 knockdown HA-VSMCs. (B)
   Differential expression of S100A4 protein in HA-VSMCs with Si-S100A4
   knockdown. (C) The CCK-8 assay revealed a significant decrease in cell
   viability following S100A4 knockdown. (D) Colony formation assay showed
   a markedly reduced number of cell colonies in cells with depleted
   S100A4 expression compared to the NC group. (E) Edu staining assay
   demonstrated that downregulation of S100A4 significantly inhibited
   HA-VSMC proliferation relative to the NC group. Statistical
   significance was indicated as *P < 0.01, **P < 0.001, ***P < 0.0001,
   and ****P < 0.00001.

Knocking down S100A4 suppresses cell migration and invasion of HA-VSMC and
induces apoptosis

   Subsequent findings from scratch wound healing assays revealed a
   significant reduction in the rate of wound closure following the
   downregulation of S100A4 expression (Fig. [227]11A). Transwell assays
   further demonstrated a notable decrease in both the migration and
   invasion capabilities of HA-VSMC cells upon S100A4 knockdown
   (Fig. [228]11B). Moreover, apoptosis in HA-VSMC was assessed using
   Annexin V-FITC/PI staining followed by flow cytometry analysis. As
   illustrated in Fig. [229]11C, knocking down S100A4 led to an increased
   proportion of apoptotic HA-VSMC cells. To ensure the accuracy and
   consistency of our results, all assays were conducted using a single
   smooth muscle cell line (HA-VSMC), and the data are presented as the
   mean standard deviation of independent experiments. Statistical
   significance is indicated as *P < 0.05, **P < 0.01, ***P < 0.001.

Fig. 11.

   [230]Fig. 11
   [231]Open in a new tab

   Knockdown of S100A4 suppresses the migration and invasion abilities of
   HA-VSMCs and induces cell apoptosis (A) The scratch wound healing assay
   revealed that decreased expression of S100A4 significantly impaired the
   wound healing rate. (B) The Transwell assay demonstrated that
   downregulation of S100A4 markedly reduced the migration and invasion
   capabilities of HA-VSMCs. (C) Annexin V-FITC/PI double staining flow
   cytometry assay showed a substantial increase in induced cell apoptosis
   following S100A4 knockdown compared to the NC group. Statistical
   significance was indicated as *P < 0.01, **P < 0.001, ***P < 0.0001,
   and ****P < 0.00001.

Discussion

   Cardiomyopathies encompass a broad spectrum of disorders that
   significantly impact the intricate functioning of the heart^[232]62.
   The etiology of these conditions is widely recognized to involve
   complex interactions between genetic predisposition and environmental
   factors^[233]63. These ailments are typically classified into distinct
   types, such as DCM, ICM and RM. Scientific investigation has validated
   a close association between the intricate mechanisms of ferroptosis—a
   regulated form of cell death dependent on disruptions in iron
   metabolism—and various cardiomyopathies, particularly ICM and DCM.
   Specifically, under conditions like coronary atherosclerosis-induced
   ischemia, cardiomyocytes endure damage due to excessive iron
   accumulation^[234]64. Moreover, the occurrence of ferroptosis
   exacerbates oxidative stress, leading to detrimental effects such as
   compromised integrity of cardiomyocyte membranes and dysfunction of
   cardiovascular endothelial cells^[235]65. In a murine model of
   myocardial ischemia, G. Drossos et al. observed a significant decrease
   in the expression of FTH, impairing cardiomyocytes’ ability to
   sequester free iron and resulting in increased oxidative stress and, in
   severe cases, cell death^[236]66. Thus, a plausible link emerges
   between ferroptosis, oxidative stress, and their collaborative roles in
   shaping the progression of cardiomyopathies. Furthermore, ferroptosis
   in VSMCs correlates positively with the advancement of
   atherosclerosis^[237]67. Therefore, our primary research focuses on
   VSMCs as a pivotal area of investigation. We aim to explore key smooth
   muscle cell subsets and hope to find and validate possible target
   genes. To achieve this goal, we rigorously analyzed scRNA-seq data to
   identify diverse SMC subpopulations implicated in cardiomyopathies,
   employing advanced visualization techniques to unravel their intricate
   features.

   To dissect the nuanced variations within SMC subpopulations with
   precision, we meticulously categorized 4,148 high-quality SMCs into
   nine distinct clusters. Each cluster was distinguished by a prominent
   marker gene: C0 SOCS3+ SMCs, C1 CD36+ SMCs, C2 SORBS2+ SMCs, C3
   RGS5+ SMCs, C4 ACTG2+ SMCs, C5 IGF2+ SMCs, C6 S100A4+ SMCs, C7
   C12orf75+ SMCs, and C8 EGR1+ SMCs. Notably, S100A4, a crucial regulator
   of vascular remodeling, plays a pivotal role in promoting angiogenesis
   and inducing oxidative stress-related damage in vascular endothelial
   cells^[238]68–[239]70. However, research on the involvement of S100A4
   in SMCs is currently limited, with few studies establishing a direct
   correlation between S100A4 levels and cardiovascular risk^[240]71. In
   addition to affecting ECs, the secretion of inflammatory cytokines and
   angiogenic growth factors influences both ECs and SMCs, impacting the
   development of atherosclerosis. The proliferation of SMCs, crucial in
   atherosclerosis progression, ultimately contributes to
   ICM^[241]72–[242]74. Among the genes highly expressed in and closely
   associated with C6 S100A4+ SMCs are IER2, C11orf96, ZFP36, PPP1R15A,
   BTG2, SERTAD1, SOCS3, and EGR. We were pleased that several of these
   eight genes have been previously identified as pro-inflammatory genes,
   which has a non-negligible link with the occurrence and development of
   cardiomyopathy that we are studying^[243]75,[244]76. Among these,
   research on IER2 in the context of ICM is scarce. However, studies have
   found that IER2 is upregulated in adipose stromal cells after vitamin C
   treatment and may participate in a novel nutrient-responsive gene
   pathway in adipose tissue^[245]77, potentially relevant to early
   atherosclerotic plaque formation. This warrants further investigation
   into its role in the pathogenesis of ICM. Research on C11orf96 is also
   extremely limited, with only a few studies reporting its upregulation
   in lung cancer tumor tissues^[246]78,[247]79. ZFP36, which is expressed
   in atherosclerotic lesions^[248]80, may have a potential connection to
   ICM. What’s more, ZFP36 is a known pro-inflammatory gene involved in
   the regulation of inflammatory responses. It promotes the expression of
   a variety of inflammatory factors by regulating the stability of
   mRNA^[249]81. Sparse studies have linked BTG2 to cardiac
   hypertrophy^[250]82. Ischemia-induced SOCS3 may promote apoptosis
   during ventricular remodeling, exacerbating myocardial necrosis and
   potentially leading to heart failure^[251]83,[252]84. And SOCS3 is an
   important immunomodulatory factor, which regulates the inflammatory
   response by inhibiting the cytokine signaling pathway, but its effect
   has a negative feedback nature^[253]85. The EGR gene family includes
   multiple members, such as EGR1, which are usually activated when cells
   respond to various stimuli, such as inflammatory stimuli. EGR1 in
   particular is thought to play an important role in the pro-inflammatory
   response^[254]86. Furthermore, overexpression of EGR has been shown to
   induce cardiac hypertrophy and arrhythmias^[255]87. In contrast,
   PPP1R15A and SERTAD1 lack research evidence supporting their
   association with cardiomyopathy. Furthermore, our visualizations
   demonstrated the differences in the cell cycle across subpopulations.
   It is well established that an increase in cells in the G1 phase
   indicates cell cycle arrest^[256]88. Moreover, previous studies showed
   that knocking down S100A4 in other cell types led to cell cycle
   arrest^[257]89. However, whether knocking down S100A4 produced the same
   effect in SMCs lacked corresponding research evidence. To address this,
   we conducted in vitro experiments using a human aortic smooth muscle
   cell line. Following S100A4 knockdown, we observed reduced
   proliferation and migration of VSMCs, along with increased apoptosis.
   More importantly, our annotations revealed that S100A4 was the most
   strongly expressed marker gene in the C6 subpopulation, prompting us to
   designate S100A4 as the naming gene for this group. Overall, the cell
   cycle was only one aspect of evidence for cellular proliferative
   capacity. Other features examined in this study, such as stemness,
   metabolism, differentiation potential, and developmental trajectories,
   further supported the proliferative ability of these cells. Based on
   these findings, we identified the C6 subpopulation as having strong
   proliferative capacity and high S100A4 expression. We therefore
   hypothesized that S100A4 has stronger expression in cells without
   growth arrest. Taken together, we proposed that the C6 subpopulation
   exhibited significant proliferative capacity and functional importance.
   This led us to hypothesize a strong correlation between the C6
   S100A4+ SMCs subpopulation and the pathogenesis as well as the adverse
   progression of cardiomyopathy. Therefore, we decided to focus future
   research on this intriguing C6 S100A4+ SMCs subpopulation.

   Next, a comprehensive analysis of GO-BP has revealed that C6
   S100A4+ SMCs are notably enriched in complex processes such as the
   regulation of neurogenesis, nervous system development, and circadian
   gene expression regulation. Importantly, ferroptosis—a critical
   component in neuronal cell demise and neurodegenerative
   diseases—emerges as an intriguing possibility for potential
   interconnections with C6 S100A4+ SMCs. These insights open up exciting
   avenues for further exploration into the role of S100A4 in SMCs and its
   implications in cardiovascular health and disease, suggesting novel
   therapeutic targets and mechanisms underlying cardiomyopathies.

   Building on our earlier analysis, we have delved deeper into
   visualizing the cellular growth activity within C6 S100A4+ SMCs.
   Leveraging stemness scoring, which correlates closely with cell growth
   dynamics and intrinsic cellular characteristics^[258]90, our
   investigation unequivocally confirms that among various subgroups, C6
   S100A4+ SMCs exhibit the highest degree of stemness, thereby
   substantiating our initial hypothesis. To further elucidate these
   findings, we investigated the expression patterns of five prominent
   stemness genes: EPAS1, CTNNB1, MYC, HIF1A, and NES. Notably, CTNNB1
   plays a dual role in cardiac development, persisting in myocardial
   structures even amid cardiac mutations^[259]91,[260]92. MYC is
   implicated in potential critical roles in heart failure^[261]93, while
   the stability of HIF1A is linked to metabolic shifts towards increased
   glucose utilization in cardiomyopathy and aging^[262]94. Thus,
   establishing potential associations between cardiomyopathies and these
   stemness genes is plausible. The heightened level of cellular stemness
   and robust cell growth observed in C6 S100A4+ SMCs suggest an equally
   active state of cellular metabolism. In this context, we assessed the
   prominence of metabolic pathways within this subgroup, identifying
   oxidative phosphorylation as the predominant metabolic process. Genetic
   anomalies affecting oxidative phosphorylation have been implicated in
   precipitating cardiomyopathies. A notable study by Lu et al.^[263]95
   using a mouse model highlighted significant similarities in
   differentially expressed proteins between ICM and DCM hearts, primarily
   mediated through pathways associated with oxidative phosphorylation.
   This underscores the pivotal role of oxidative phosphorylation in heart
   failure pathogenesis.

   In summary, our comprehensive findings suggest that C6 S100A4+ SMCs may
   hold relevance to cardiomyopathies, particularly concerning cellular
   growth, stemness, and metabolic processes. These insights pave the way
   for further exploration into novel therapeutic targets and mechanisms
   underlying cardiomyopathies, potentially offering new avenues for
   intervention and treatment strategies.

   Subsequently, we proceeded to assign scores to all SMC subpopulations
   using a gene set library specifically associated with ferroptosis.
   Notably, C6 S100A4+ SMCs exhibited relatively elevated scores,
   effectively corroborating our initial hypothesis. In the subsequent
   assessment, we evaluated all subpopulations against four distinct
   pathways linked to oxidative stress. In the first three pathways, which
   involve positive regulation or response to oxidative stress, C6
   S100A4+ SMCs consistently attained higher scores compared to other
   subpopulations. Conversely, in the pathway associated with the negative
   regulation of the oxidative stress response, C6 S100A4+ SMCs recorded
   notably low scores. This compellingly reinforces our contention that C6
   S100A4+ SMCs represent a subpopulation highly likely to be involved in
   ferroptosis and susceptible to oxidative stress.

   The degree of cell differentiation is a crucial indicator reflecting
   ongoing cellular growth and proliferation activity. To validate the
   characteristics of C6 S100A4+ SMCs, we employed trajectory inference
   using the Slingshot algorithm. Remarkably, our results indicate that C6
   S100A4+ SMCs predominantly occupy early stages across all three
   lineages studied. This compellingly demonstrates that the majority of
   C6 S100A4+ SMCs exist in a state of active growth, marked by heightened
   mRNA and gene expression. Consequently, these findings suggest a
   profound potential for these SMCs to significantly contribute to the
   process of SMC proliferation.

   Furthermore, the GO-BP analysis of genes showing differential
   expression at the stage where C6 S100A4+ SMCs are located revealed
   significant enrichment in terms related to radiation, mRNA, nuclear
   transcription, proliferation, stress, transcription, muscle, death, and
   destabilization. Most of these terms are directly associated with cell
   growth, proliferation, apoptosis, and stress responses. Thus, this
   thorough analysis provides strong evidence that substantiates our
   initial hypothesis.

   Based on these findings, it can be inferred that C6 S100A4+ SMCs are
   closely linked to the exacerbation of cardiomyopathy, potentially
   contributing to its progression. Adhering to rigorous research
   principles, we utilized CytoTRACE to evaluate cellular stemness,
   revealing that C6 S100A4+ SMCs exhibit the highest stemness level and
   the lowest degree of differentiation. Additionally, Monocle analysis
   further supports these observations by showing that the majority of C6
   S100A4+ SMCs are positioned early in the developmental trajectory, with
   only a small fraction in the terminal segment. Consistency with results
   from Slingshot and stemness scoring provides further confirmation that
   C6 S100A4+ SMCs represent a subset of SMCs characterized by limited
   differentiation but heightened activity in growth and proliferation.

   Moreover, extensive research has focused on exploring the intricate
   interactions between SMCs and ECs^[264]96. Endothelial dysfunction, a
   well-established pathological condition affecting ECs, is closely
   associated with various diseases such as hypertension, atherosclerosis,
   and diabetes. These conditions often involve changes in vascular
   reactivity and remodeling of the vascular wall, which particularly
   affect SMCs. Consequently, these phenomena could significantly
   influence the development and progression of disorders like
   ICM^[265]97,[266]98. Therefore, it is hypothesized that the interplay
   between SMCs and ECs may play a pivotal role in the pathogenesis of
   cardiomyopathies, although the precise mechanisms remain incompletely
   understood.

   Through the application of CellChat analysis, a compelling finding
   emerged: C6 S100A4+ SMCs function as signal transmitters, while ECs act
   as signal receivers, forming a ligand-receptor pair identified as
   PTN-NCL. This interaction has been visually confirmed and validated,
   providing concrete evidence of its occurrence. Notably, PTN expression
   on C6 S100A4+ SMCs has been shown to enhance EC migration, potentially
   leading to an increased local EC population^[267]99,[268]100. NCL, a
   multifunctional protein, plays a dynamic role in the proliferation of
   VSMCs and acts as a mediator for PTN’s stimulatory effects.
   Furthermore, blocking or downregulating NCL on EC surfaces effectively
   inhibits EC migration, impairs capillary tube formation, and promotes
   EC apoptosis^[269]101,[270]102. Specifically, PTN enhances the
   proliferation and migration of ECs, while NCL boosts the proliferation
   of C6 S100A4+ SMCs and inhibits EC apoptosis. The synergistic actions
   of PTN and NCL mutually reinforce their effects, collectively exerting
   a significant impact on the initiation and progression of
   cardiomyopathies.

   To explore the key TFs in specific subpopulations of SMCs implicated in
   cardiomyopathies, we conducted a gene regulatory network analysis.
   Initially, our investigation identified KLF2 as the most dynamically
   active TF in C6 S100A4+ SMCs. Activated KLF2 regulates critical
   functions in ECs, including proliferation, migration, and inflammation,
   playing a protective role in maintaining EC integrity and
   health^[271]103,[272]104. Consequently, C6 S100A4+ SMCs may influence
   cardiomyopathies by promoting EC proliferation and migration through
   heightened KLF2 expression.

   Using data from the Comprehensive Sequence-based Investigation (CSI),
   we identified two primary modules, M1 and M2, within SMC subpopulations
   associated with cardiomyopathies. Importantly, C6 S100A4+ SMCs
   predominantly belong to module M1. Subsequently, we examined TFs with
   the highest rankings in M1. JUNB, ATF3, FOS, FOSB, and JUN were found
   to exhibit elevated expression levels in C6 S100A4 + SMCs. ATF3, in
   particular, shows heightened expression specifically in myocardial
   tissue affected by DCM, although its precise role in cardiac physiology
   remains debated^[273]105,[274]106. Emerging evidence suggests that
   inhibiting FOS may potentially mitigate cardiac fibrosis and
   inflammation^[275]107. Additionally, there is a documented association
   between fibrosis and the activation of TFs FOSB and JUNB^[276]108.
   Cardiac fibrosis is a detrimental outcome associated with
   cardiomyopathies, particularly in cases linked to diabetes-triggered
   DCM. Hence, it is plausible that the release of TFs FOS, FOSB, and JUNB
   by C6 S100A4+ SMCs contributes to exacerbating cardiomyopathies,
   potentially through the induction of cardiac fibrosis and inflammation.

   Finally, as mentioned when selecting the C6 subtype, we conducted in
   vitro experiments to determine whether the knockdown of S100A4
   similarly blocks cell growth and promotes apoptosis in SMCs, with the
   aim of exploring its role in cardiomyopathy, particularly in ICM.
   Initially, using siRNA-mediated knockdown, we observed a significant
   decrease in the viability of HA-VSMCs after S100A4 downregulation, as
   confirmed by CCK8 assays. Further studies involving colony formation
   and Edu staining experiments confirmed that S100A4 positively promotes
   the proliferation of HA-VSMCs. Furthermore, using scratch wound healing
   and Transwell assays, we determined that knockdown of S100A4
   significantly inhibited HA-VSMCs migration and invasion. Flow cytometry
   analysis showed that the proportion of apoptotic HA-VSMCs increased
   after S100A4 knockdown, indicating that S100A4 knockdown exhibits
   apoptosis-promoting properties. Based on these findings, we concluded
   that knockdown of S100A4 in C6 S100A4+ SMCs inhibits SMCs
   proliferation, migration and invasion and promotes apoptosis. This may
   serve as indirect evidence that S100A4 may have a necessary link with
   the proliferation of SMCs, leading to fibrosis and pathological
   ventricular remodeling, which impairs cardiac function and ultimately
   accelerates the progression of cardiomyopathy^[277]109,[278]110. Of
   course, further studies are needed to prove this. In addition, we will
   also conduct further experiments to verify the specific regulatory
   effect of S100A4 on the cell cycle process.

Conclusion

   Based on our comprehensive single-cell characterization of SMC subsets
   in cardiomyopathy, as well as analysis of relevant transcriptional
   regulatory networks and in vitro experimental validation, our study
   concludes that C6 S100A4+ SMCs exhibit increased sensitivity to adverse
   events associated with cardiomyopathy. This sensitivity may be
   influenced by processes involved in ferroptosis, oxidative
   phosphorylation, and inflammatory response, as C6 subgroup has higher
   such scores, but whether it is really affected by them needs further
   research to prove.

   In addition, C6 S100A4+ SMCs had higher scores for genes related to
   ferroptosis, oxidative stress and inflammation, which may indicate a
   new potential therapeutic approach for treating cardiomyopathy and
   preventing heart failure. The identification of PTN-NCL protein
   receptor pairs between C6 S100A4+ SMCs and ECs also provides a possible
   potential target for therapeutic intervention in cardiomyopathy. In
   addition, stem-like genes such as CTNNB1, MYC, and HIF1A may also
   attenuate heart failure after cardiomyopathy. KLF2, a regulator
   identified in our study, has been shown by several studies to inhibit
   EC growth, proliferation and migration, so it also shows potential as a
   therapeutic target for the prevention of cardiomyopathy. The
   transcription factors FOS, FOSB, and JUNB may help prevent adverse
   events such as cardiac fibrosis in cardiomyopathy.

   In summary, these findings provide new insights into the future
   prevention, treatment, and management of cardiomyopathies. However,
   further refinement and advancement of our research are necessary to
   fully harness the therapeutic potential of these discoveries.

Study limitations

   We must acknowledge certain unavoidable limitations that may introduce
   some degree of error. Some of these limitations include, for example,
   the possibility that patient-derived samples may be affected by the
   administration of positive inotropic drugs, which may affect signaling
   mechanisms and differential transcription patterns. This may lead to
   small differences in the strength and frequency of communication
   signals between cells, as well as changes in transcriptional patterns
   within subpopulations. In addition, C6 subpopulation is derived from
   two ventricles, and it has been reported that in different
   pathogenesis, cells from the two ventricles respond to various signals
   differently and tend to adopt different functions, and each ventricle
   is unique. Therefore, the heterogeneity between the two ventricles may
   play a role in the characterization of C6 subgroup. Therefore, fully
   considering these potential errors and striving to minimize their
   effects will remain an important aspect of our work in our future
   studies. Further in vitro studies on ferroptosis, oxidative stress,
   PTN-NCL signaling pathway, and related ligand receptor proteins and
   genes will be performed to obtain more important and interesting
   findings.

Supplementary Information

   [279]Supplementary Information 1.^ (6MB, docx)
   [280]Supplementary Information 2.^ (6.2MB, pdf)
   [281]Supplementary Information 3.^ (471.4KB, pdf)

Acknowledgements