An integrated resource for ischemic heart disease defines hallmarks and heterogeneity across time and space

Tianhao Wang, Yining Hu, Wenbo Guo, Haoran Li, Menglei Wang, Bowen Chen, H. -r. Bao, Meng Gao, Xiang Li, Qian Chen, Mingjun Shen, Xin Shao, Jie Liao, Xiaohui FanActa 2025 · DOI

TL;DR

This paper introduces ssIHDB, a comprehensive web resource that integrates single-cell and spatial transcriptomics data to map the complex cellular landscape of ischemic heart disease (IHD) across different time points and spatial locations. By resolving cellular heterogeneity at unprecedented resolution, ssIHDB reveals a novel fibrotic region in cardiomyocytes, suggesting a new therapeutic target for IHD.

What Problem Does This Paper Address?

Ischemic heart disease (IHD) is a major global health challenge, yet its underlying pathology, particularly at the cellular and tissue level, remains incompletely understood. Traditional methods struggle to capture the intricate microstructures and dynamic cellular changes occurring during IHD progression. While single-cell RNA sequencing (scRNA-seq) and spatial transcriptomics (ST) have opened new avenues, limitations in technical resolution, sample size, and inconsistencies in experimental protocols across different studies have hindered a unified understanding. This paper aims to bridge this gap by creating an integrated, high-resolution resource that resolves the spatiotemporal heterogeneity of IHD, enabling more robust cross-validation and deeper biological insights.

Methods

The authors employed a multi-pronged approach:

  • scRNA-seq Data Integration: They integrated multiple wild-type mouse scRNA-seq datasets using the scVI tool, which proved effective in removing batch effects while preserving developmental relationships. This allowed for a hierarchical clustering of cells into broad categories, minor categories, and basic cell types, and further into functional subtypes.
  • Meta-program (MP) Analysis: To overcome limitations in data volume and batch effects for NMF analysis, they developed a modified meta-program analysis. This strategy was applied to identify programs (gene sets) significantly associated with IHD, tissue remodeling, and immune responses.
  • Spatial Transcriptomics (ST) Analysis: They utilized Bayesian Analytics for Spatial Segmentation (BASS) to define spatial domains within the infarcted myocardium across different time points post-injury. Differential expression analysis was performed on these domains to characterize their molecular signatures.
  • Integrative Analysis: The scRNA-seq and ST data were integrated to map the spatiotemporal distribution of identified meta-programs and cell types. This involved scoring ST spots for MP activity and deconvoluting cell types from ST data.
  • Knowledge Graph Construction: A comprehensive knowledge graph (IHDKG) was built by manually curating genes, drugs, and comorbidities related to IHD, along with molecular interactions. This graph serves as a foundation for network analysis tools within the ssIHDB portal.
  • Web Portal Development: The ssIHDB web portal was developed to provide a user-friendly platform for visualizing, reusing, and performing online analyses on the integrated single-cell and spatial data.

Datasets

The study utilized publicly available scRNA-seq data from wild-type mouse models of IHD. Spatial transcriptomics data from mouse models at various time points post-myocardial infarction (MI) and ischemia-reperfusion (IR) injury were also used. The authors also curated a dataset of 6035 genes, 3157 small-molecule or biotech drugs, and 79 comorbidities relevant to IHD.

Key Insights and Ideas

  • Unprecedented Cellular Resolution: The integration of scRNA-seq data allowed for the identification of a highly granular hierarchy of cell types, including rare cell populations that are often missed in single-dataset analyses. This provides a much clearer picture of the cellular complexity in IHD.
  • Discovery of a Novel Fibrotic Cardiomyocyte Subtype: A key finding is the identification of a specific fibrotic region characterized by cardiomyocytes highly expressing Ltbp2. This region, termed "fibrosis zone" (FZ), was consistently identified through integrative spatial analysis and suggests that cardiomyocytes themselves can actively contribute to fibrosis, a paradigm shift from viewing them solely as passive bystanders.
  • Spatiotemporal Dynamics of Disease Hallmarks: The meta-program analysis revealed distinct temporal and spatial patterns of gene programs associated with tissue remodeling and immune responses, highlighting how these processes evolve during IHD progression.
  • CMs as Potential Therapeutic Targets: The identification of Ltbp2-hi CMs in the fibrotic zone points towards these cells as a potential new target for anti-fibrotic therapies in IHD.
  • Integrated Resource for IHD Research: The ssIHDB portal consolidates diverse data types and analytical tools, providing a valuable, accessible resource for the IHD research community to validate findings, explore new hypotheses, and accelerate drug discovery.

Key Results

  • The scRNA-seq atlas resolved IHD into 6 major cell categories, 24 minor categories, and 40 basic cell types, with further division into functional subtypes.
  • Meta-program analysis identified 16 programs significantly associated with IHD, with MP32 being particularly enriched in MI and IR injury.
  • Spatial analysis delineated 12 distinct spatial domains within the infarcted myocardium.
  • The MP32 score was found to be highest in a fibrotic cardiomyocyte subtype (Ltbp2-hi CMs), which was localized to a specific spatial domain (IZ-C), later named the "fibrosis zone" (FZ).
  • The ssIHDB knowledge graph contains 12,810 nodes and 89,596 edges, integrating genes, drugs, and comorbidities.

Limitations and Future Work

  • Mouse Model Focus: The current atlas is based on mouse models. While valuable, translating these findings directly to human IHD requires further validation in human datasets.
  • Technical Limitations: The authors acknowledge that while scVI effectively removed batch effects, some technical limitations and inconsistencies in experimental operations between laboratories can still affect cross-validation.
  • Future Directions: The authors suggest that the ssIHDB portal can be continuously expanded with new datasets and analytical tools. The identification of the Ltbp2-hi CMs as a potential therapeutic target opens avenues for developing targeted anti-fibrotic therapies. Further investigation into the precise mechanisms by which these CMs contribute to fibrosis is warranted.

Summary updated

BibTeX
@inproceedings{wang2025integrated,
  author = {Tianhao Wang and Yining Hu and Wenbo Guo and Haoran Li and Menglei Wang and Bowen Chen and H. -R. Bao and Meng Gao and Xiang Li and Qian Chen and Mingjun Shen and Xin Shao and Jie Liao and Xiaohui Fan},
  title = {{An integrated resource for ischemic heart disease defines hallmarks and heterogeneity across time and space}},
  booktitle = {Acta},
  year = {2025},
  doi = {10.1016/j.apsb.2025.11.020},
}
Figure