Cell-Cycle-Specific Autoencoding Improves Cluster Analysis of Cycling Cardiomyocytes

Area of Science:

• Cardiology
• Molecular Biology
• Genomics

Background:

• Previous single-nucleus RNA sequencing (snRNAseq) identified cardiomyocyte subpopulations with potential for proliferation post-myocardial infarction (MI).
• Distinguishing cell cycle phases within cardiomyocyte clusters was previously challenging.
• This study focused on cell cycle-related genes for enhanced precision.

Purpose of the Study:

• To precisely assess cardiomyocyte cell cycle activity using a refined snRNAseq analysis.
• To identify specific cardiomyocyte subpopulations and their cell cycle dynamics.
• To uncover genes and pathways regulating cardiomyocyte proliferation.

Main Methods:

• Applied a cell-cycle-specific analytical pipeline to mouse and pig cardiac snRNAseq data.
• Identified cycling cardiomyocytes by co-expression of five key proliferation markers.
• Utilized the cell-cycle-specific autoencoder (CSA) algorithm for cluster analysis.

Main Results:

• The CSA algorithm identified 7 clusters in mouse hearts and 5 in pig hearts, with distinct cell cycle activities.
• In pigs, cardiomyocyte clusters pCM1 and pCM4 showed cell cycle activity, with pCM4 prominent in fetal hearts and pCM1 in injured hearts.
• Pseudotime analysis revealed a cell cycle progression pathway (pCM3 → pCM2 → pCM1) in fetal pig cardiomyocytes, linked to specific cell cycle regulators.
• Identified four transcription factors (E2F8, FOXM1, GLI3, RAD51) upregulated in regenerative contexts.

Conclusions:

• The CSA algorithm significantly improved the precision of cardiomyocyte cell cycle assessment.
• A clear trajectory of cardiomyocyte cell cycle onset and progression was identified in fetal pigs.
• This research provides insights into the molecular mechanisms governing cardiomyocyte regeneration.