Characterization of DCM and Control Samples Using Comprehensive Single-Cell Analysis
This study aimed to delineate the cellular heterogeneity in cardiac tissues of patients with DCM and control subjects using a comprehensive single-cell approach. Three patients diagnosed with DCM and two control subjects without a history of cardiac disease were selected. One DCM sample (DCM_1) did not meet the stringent quality control standards established for subsequent analysis. The criteria for poor quality control were based on a combination of low cell yield, poor library complexity, and high doublet scores. Consequently, DCM_1 was excluded to ensure the integrity and reliability of comparative analysis results. Each sample was subjected to simultaneous scRNA-seq and ATAC sequencing using standard optimized protocols to profile gene expression and chromatin accessibility landscapes, respectively. After processing using the Seurat package (Fig. 1A–B), the RNA-seq data were subjected to quality control (Figure S1A). Violin plots revealed the distribution and expression levels of marker genes in the identified cell populations (Fig. 1C). Distinct cardiac cell types, including cardiomyocytes, endothelial cells, and fibroblasts, as well as immune cell subsets and their proportions, were identified (Fig. 1D).
The ATAC-seq data were processed using the ArchR package, which enabled the assessment of chromatin accessibility across the genome at single-cell resolution (Fig. 1E). The integration of single-cell RNA and ATAC-seq data within the ArchR framework enabled a comprehensive multi-omic characterization with optimal quality control (Figure S1B–C). This analysis confirmed the concordance of cell-type annotations between the transcriptomic and chromatin accessibility data. The cell-type-specific patterns of open chromatin (Fig. 1F) were identified, providing a basis for understanding the regulatory changes in DCM. The chromatin accessibility profiles of cells positive for MKI67, a marker of proliferation, were similar to those of fibroblasts and macrophages. This suggests that the cardiac tissues exhibit a potential proliferative capacity, which can contribute to the pathological remodeling in DCM. However, neuronal cells and MKI67 + proliferating cells did not exhibit distinct open chromatin features, indicating the presence of an epigenetic regulatory mechanism in these cell types that warrants further investigation. The results of single-cell analysis revealed the complex cellular landscape in DCM and provided a foundation for understanding the pathophysiology of DCM at single-cell resolution.
WT1 is a Central Regulator of Neuronal Pathway Dysregulation in the Cardiomyocytes of Patients with DCM
Simultaneous scRNA-seq and scATAC-seq analysis revealed a complex transcriptional and epigenetic landscape in DCM cardiomyocytes. Analysis of differential gene expression between DCM and control samples revealed significantly dysregulated genes (Fig. 2A). The top 100 significant differentially expressed genes (DEGs) in the DCM group were subjected to Enricher analysis [6]. These DEGs were enriched in different pathway collections (Fig. 2B). Consistent with previous findings, pathways related to collagen synthesis, trimerization, and extracellular matrix organization were enriched in the DCM group. Additionally, this study, for the first time, demonstrated the enrichment of neuronal pathways, including synapse, ion-channel, and GABA receptor signaling, in the DCM group. These findings provide a new perspective on DCM pathophysiology, suggesting the presence of an uncharacterized neuro-cardiac interface in DCM.
The enriched neuronal pathways can be novel mechanisms underlying DCM pathogenesis. Among the transcription factors identified from scATAC-seq peak analysis, WT1 was significantly enriched in the DCM group (Fig. 2C–D). The direct role of WT1 in DCM has not been previously established. However, Kumari et al reported that WT1 is a key regulator of synaptic plasticity, modulating memory strength and flexibility [7]. This suggests that WT1 promotes neuron-like characteristics or communication with neuronal cells in the heart. The importance of WT1 was also verified with transcription factor motif analysis in Enrichr using the top regulated DEG from scRNA-seq data (Fig. 2E).
To further investigate the potential implications of dysregulated neuronal pathways, immunostaining was performed for the neuronal cell marker RBFOX3 (Fig. 2F). Immunostaining was performed to validate the bioinformatics analysis findings and establish the correlation between WT1 expression and neuronal features in DCM cardiomyocytes. The number of RBFOX3-positive cells in the DCM group was higher than that in the control group. Thus, the dysregulation of neuronal pathways in DCM was further studied.
Neuronal Influence in DCM Cardiomyocytes
Cell-cell signaling analysis, which was performed using CellChat, revealed the significant pathways and interactions involved in the pathophysiology of DCM. This study focused on GABA receptor signaling, which was suggested to be a target pathway in our previous analyses. Neuronal cells were categorized based on the expression of GABBR1 to precisely elucidate neuronal-cardiomyocyte communication. Network analysis revealed the upregulation of both autocrine and paracrine signaling in GABBR1-positive cells of the DCM group. The results of network analysis were visualized using a network diagram (Fig. 3A and Figure S2A). These findings illustrated the enhanced communication pathways.
Analysis of the specific pathways that differentiated the control group from the DCM group revealed the enrichment of SPP1 and NOTCH signaling pathways in the control group. In contrast, the IGF, NRG, and CDH signaling pathways were enriched in the DCM group (Fig. 3B). These differential pathways were identified using a heatmap of the enriched functional pathways in various cell types (Fig. 3C), enabling a comparative view of cellular communication. Compared with those in the control group, adipocytes and lymphatic cells upregulated IGF signaling in the DCM group. Furthermore, analysis of GABBR1-positive and GABBR1-negative neuronal cells revealed that the NRXN and NRG signaling pathways were upregulated in the DCM group (Fig. 3D) in addition to the top DEGs between GABBR1-positive and GABBR1-negative cells (Fig. 3F), indicating their potential as therapeutic and diagnostic targets. This suggests an intricate interplay in which neuronal signaling pathways mediate the cardiomyocyte response in DCM.
The NOTCH signaling pathway, which is involved in cell differentiation and communication, was downregulated in DCM. This is consistent with the findings of previous studies, which reported that NOTCH signaling has a crucial role in various cellular processes and differentiation pathways [8]. The role of the IGF pathway in the pathogenesis of DCM is interesting as it regulates biliary epithelial cell expansion and inhibits hepatocyte differentiation [9]. Thus, the IGF pathway may be involved in cell proliferation and differentiation. These findings suggest that specific signaling pathways traditionally associated with neuronal or developmental processes are dysregulated in DCM, influencing disease progression and providing avenues for developing novel therapeutic strategies.
Cellular Dynamics in DCM Samples
To further delineate the microenvironment of DCM, one control sample and two DCM samples were subjected to matched spatial transcriptomic analysis. The data were analyzed using CellTrek. Different cell clusters in different samples were identified and mapped in the spatial transcriptomic slides (Fig. 4A–B, E–F, and I–J). The scRNA label was transferred to the slides to predict the cell label on the slide (Fig. 4C, G, and K). Based on the CellTrek results, the colocalization patterns of different types of cells were summarized (Fig. 4D, H, and L). The colocalization of cardiomyocytes with fibroblasts in DCM samples was lower than that in control samples. A previous single-cell study reported that fibroblast activation is a major event in DCM [3]. Additionally, the colocalization of cardiomyocytes with neuronal cells was upregulated in DCM samples, indicating enhanced spatial interactions.