Identification of Key Genes as Early Warning Signals of Acute Myocardial Infarction Based on Weighted Gene Correlation Network Analysis and Dynamic Network Biomarker Algorithm

Chenxi Song, 正 石橋, Luonan Chen, Jing Ge, Rui Zhang, Sheng Yuan, Xiaohui Bian, Chunyue Wang, Qianqian Liu, Lei Jia, Rui Fu, Kefei DouFrontiers in Immunology 2022 · DOI

This study, published in Frontiers in Immunology, presents a systems-biology approach to identifying the "tipping point" of heart disease. By analyzing the RNA of peripheral blood cells, researchers identified a 13-gene signature that acts as an early warning system for the transition from stable coronary artery disease to acute myocardial infarction (heart attack).

TL;DR

Researchers combined traditional transcriptomics with a "Dynamic Network Biomarker" (DNB) algorithm to identify 13 key genes that signal the critical transition from stable heart disease to a heart attack. This approach pinpoints the "intermediate lesion" stage as the vital window for early warning, potentially allowing for medical intervention before irreversible damage occurs.

What Problem Does This Paper Address?

Coronary Artery Disease (CAD) is a leading cause of death globally. While we have excellent biomarkers for diagnosing a heart attack after it happens (like cardiac troponin), we lack robust "early warning" signals that can predict when a stable patient is about to cross the threshold into an Acute Myocardial Infarction (AMI).

Traditional bioinformatics usually compares "healthy" vs. "diseased" states. However, disease progression is non-linear; there is often a "pre-disease" state or a tipping point where the system becomes highly unstable before shifting into a full-blown pathological state. This paper seeks to identify the molecular signatures of that specific tipping point.

Methods

The researchers used a multi-pronged computational pipeline to analyze the transcriptomes of peripheral blood monocytes (a type of immune cell that plays a major role in arterial plaque formation).

  1. RNA-Seq: They sequenced the RNA of 24 patients categorized into three stages: Normal (NCA), Intermediate Lesion (IML), and Heart Attack (AMI).
  2. WGCNA (Weighted Gene Correlation Network Analysis): This standard technique clusters genes into "modules" based on how their expression levels correlate across samples, helping identify groups of genes linked to clinical traits.
  3. DNB (Dynamic Network Biomarker) Algorithm: This is the study's novel core. Unlike static methods, DNB looks for a group of molecules that exhibit three specific behaviors near a tipping point:
  4. A drastic increase in the standard deviation (fluctuation) of the DNB genes.
  5. A drastic increase in the correlation within the DNB gene group.
  6. A drastic decrease in the correlation between DNB genes and other genes in the system.
  7. Integration: The final "key genes" were identified by finding the overlap between Differentially Expressed Genes (DEGs), the most relevant WGCNA modules, and the DNB results.

Datasets

  • Primary Dataset: Original transcriptomic (RNA-seq) data from peripheral blood monocytes of 24 age- and sex-matched patients at Fuwai Hospital, China.
  • Groups: 8 patients with normal coronary arteries (NCA), 8 with intermediate lesions (IML, 50%–70% stenosis), and 8 with AMI.
  • Reference: Human reference genome (GRCh38.91).

Key Insights and Ideas

  • The IML Stage is the Tipping Point: The DNB score (a composite index of system instability) peaked at the Intermediate Lesion (IML) stage. This suggests that patients with moderate arterial narrowing are in the "pre-disease" state where the molecular network is most volatile and prone to shifting toward a heart attack.
  • Monocytes as Sentinels: By looking at monocytes in the blood rather than the heart tissue itself, the study identifies a non-invasive way to monitor the "inflammatory weather" of the coronary arteries.
  • Biological Drivers: The key gene modules were heavily enriched for pathways involving T-cell activation, cytokine production, and apoptosis. This reinforces the idea that a heart attack isn't just a plumbing problem (clogged pipes) but an immunological crisis (inflamed, unstable plaques).

Key Results

  • The 13-Gene Signature: The study identified 13 genes as critical early warning signals: SGPP2, DAZAP2, INSIG1, CD82, OLR1, ARL6IP1, LIMS1, CCL5, CDK7, HBP1, PLAU, SELENOS, and DNAJB6.
  • WGCNA Findings: Identified 9 gene modules. The "Pink" module ($r^2 = 0.64, p = 8e-04$) and "Turquoise" module ($r^2 = 0.55, p = 0.005$) showed the strongest positive association with disease progression.
  • DNB Identification: The DNB algorithm specifically identified 103 genes that characterize the critical transition point.
  • Validation: Violin plots of the 13 genes showed significant, progressive changes in expression levels from the normal stage through the intermediate stage to the AMI stage.

Limitations and Future Work

  • Sample Size: The study is limited by a small cohort (n=24). While the DNB algorithm is designed to work with small samples, larger cohorts are needed for clinical validation.
  • Retrospective Nature: This was a cross-sectional study. A prospective study is required to prove that these 13 genes can actually predict a future heart attack in a healthy-appearing patient.
  • Lack of qPCR Validation: The results are based on RNA-seq data; the authors acknowledge that quantitative PCR (qPCR) validation of these gene expression levels is a necessary next step.
  • Functional Gaps: While some genes like CCL5 (inflammation) and OLR1 (lipid uptake) have known roles in heart disease, the specific roles of others like DAZAP2 and HBP1 in atherosclerosis remain a mystery.

Summary updated

Abstract

The current study identified a total of 13 genes that may play key roles in the progression of atherosclerotic plaque and provides new insights for early warning biomarkers and underlying mechanisms underlying the progression of CAD.
BibTeX
@inproceedings{song2022identification,
  author = {Chenxi Song and 正 石橋 and Luonan Chen and Jing Ge and Rui Zhang and Sheng Yuan and Xiaohui Bian and Chunyue Wang and Qianqian Liu and Lei Jia and Rui Fu and Kefei Dou},
  title = {{Identification of Key Genes as Early Warning Signals of Acute Myocardial Infarction Based on Weighted Gene Correlation Network Analysis and Dynamic Network Biomarker Algorithm}},
  booktitle = {Frontiers in Immunology},
  year = {2022},
  doi = {10.3389/fimmu.2022.879657},
}
Figure