Analysis of cardiac single-cell RNA-sequencing data can be improved by the use of artificial-intelligence-based tools

Single-Cell RNA Sequencing: Technological Progress and Biomedical Application in Cancer Research

Single-cell RNA-seq (scRNA-seq) is a revolutionary technology that allows for the genomic investigation of individual cells in a population, allowing for the discovery of unusual cells associated with cancer and metastasis. ScRNA-seq has been used to discover different types of cancers with poor prognosis and medication resistance such as lung cancer, breast cancer, ovarian cancer, and gastric cancer. Besides, scRNA-seq is a promising method that helps us comprehend the biological features and dynamics of cell development, as well as other disorders. This review gives a concise summary of current scRNA-seq technology. We also explain the main technological steps involved in implementing the technology. We highlight the present applications of scRNA-seq in cancer research, including tumor heterogeneity analysis in lung cancer, breast cancer, and ovarian cancer. In addition, this review elucidates potential applications of scRNA-seq in lineage tracing, personalized medicine, illness prediction, and disease diagnosis, which reveals that scRNA-seq facilitates these events by producing genetic variations on the single-cell level.

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

BACKGROUND

Our previous analyses of cardiomyocyte single-nucleus RNA sequencing (snRNAseq) data from the hearts of fetal pigs and pigs that underwent apical resection surgery on postnatal day (P) 1 (ARP1), myocardial infarction (MI) surgery on P28 (MIP28), both ARP1 and MIP28 (ARP1MIP28), or controls (no surgical procedure or CTL) identified 10 cardiomyocyte subpopulations (clusters), one of which appeared to be primed to proliferate in response to MI. However, the clusters composed of primarily proliferating cardiomyocytes still contained noncycling cells, and we were unable to distinguish between cardiomyocytes in different phases of the cell cycle. Here, we improved the precision of our assessments by conducting similar analyses with snRNAseq data for only the 1646 genes included under the Gene Ontology term "cell cycle."

METHODS

Two cardiac snRNAseq datasets, one from mice (GEO dataset number GSE130699) and one from pigs (GEO dataset number GSE185289), were evaluated via our cell-cycle-specific analytical pipeline. Cycling cells were identified via the co-expression of 5 proliferation markers (AURKB, MKI67, INCENP, CDCA8, and BIRC5).

RESULTS

The cell-cycle-specific autoencoder (CSA) algorithm identified 7 cardiomyocyte clusters in mouse hearts (mCM1 and mCM3-mCM8), including one prominent cluster of cycling cardiomyocytes in animals that underwent MI or Sham surgery on P1. Five cardiomyocyte clusters (pCM1, pCM3-pCM6) were identified in pig hearts, 2 of which (pCM1 and pCM4) displayed evidence of cell cycle activity; pCM4 was found primarily in hearts from fetal pigs, while pCM1 comprised a small proportion of cardiomyocytes in both fetal hearts and hearts from ARP1MIP28 pigs during the 2 weeks after MI induction, but was nearly undetectable in all other experimental groups and at all other time points. Furthermore, pseudotime trajectory analysis of snRNAseq data from fetal pig cardiomyocytes identified a pathway that began at pCM3, passed through pCM2, and ended at pCM1, whereas pCM3 was enriched for the expression of a cell cycle activator that regulates the G1/S phase transition (cyclin D2), pCM2 was enriched for an S-phase regulator (CCNE2), and pCM1 was enriched for the expression of a gene that regulates the G2M phase transition and mitosis (cyclin B2). We also identified 4 transcription factors (E2F8, FOXM1, GLI3, and RAD51) that were more abundantly expressed in cardiomyocytes from regenerative mouse hearts than from nonregenerative mouse hearts, from the hearts of fetal pigs than from CTL pig hearts, and from ARP1MIP28 pig hearts than from MIP28 pig hearts during the 2 weeks after MI induction.

CONCLUSIONS

The CSA algorithm improved the precision of our assessments of cell cycle activity in cardiomyocyte subpopulations and enabled us to identify a trajectory across 3 clusters that appeared to track the onset and progression of cell cycle activity in cardiomyocytes from fetal pigs.

Promoting cardiomyocyte proliferation for myocardial regeneration in large mammals

From molecular and cellular perspectives, heart failure is caused by the loss of cardiomyocytes-the fundamental contractile units of the heart. Because mammalian cardiomyocytes exit the cell cycle shortly after birth, the cardiomyocyte damage induced by myocardial infarction (MI) typically leads to dilatation of the left ventricle (LV) and often progresses to heart failure. However, recent findings indicate that the hearts of neonatal pigs completely regenerated the cardiomyocytes that were lost to MI when the injury occurred on postnatal day 1 (P1). This recovery was accompanied by increases in the expression of markers for cell-cycle activity in cardiomyocytes. These results suggest that the repair process was driven by cardiomyocyte proliferation. This review summarizes findings from recent studies that found evidence of cardiomyocyte proliferation in 1) the uninjured hearts of newborn pigs on P1, 2) neonatal pig hearts after myocardial injury on P1, and 3) the hearts of pigs that underwent apical resection surgery (AR) on P1 followed by MI on postnatal day 28 (P28). Analyses of cardiomyocyte single-nucleus RNA sequencing data collected from the hearts of animals in these three experimental groups, their corresponding control groups, and fetal pigs suggested that although the check-point regulators and other molecules that direct cardiomyocyte cell-cycle progression and proliferation in fetal, newborn, and postnatal pigs were identical, the mechanisms that activated cardiomyocyte proliferation in response to injury may differ from those that regulate cardiomyocyte proliferation during development.

Unlocking cardiac motion: assessing software and machine learning for single-cell and cardioid kinematic insights

The heart coordinates its functional parameters for optimal beat-to-beat mechanical activity. Reliable detection and quantification of these parameters still represent a hot topic in cardiovascular research. Nowadays, computer vision allows the development of open-source algorithms to measure cellular kinematics. However, the analysis software can vary based on analyzed specimens. In this study, we compared different software performances in in-silico model, in-vitro mouse adult ventricular cardiomyocytes and cardioids. We acquired in-vitro high-resolution videos during suprathreshold stimulation at 0.5-1-2 Hz, adapting the protocol for the cardioids. Moreover, we exposed the samples to inotropic and depolarizing substances. We analyzed in-silico and in-vitro videos by (i) MUSCLEMOTION, the gold standard among open-source software; (ii) CONTRACTIONWAVE, a recently developed tracking software; and (iii) ViKiE, an in-house customized video kinematic evaluation software. We enriched the study with three machine-learning algorithms to test the robustness of the motion-tracking approaches. Our results revealed that all software produced comparable estimations of cardiac mechanical parameters. For instance, in cardioids, beat duration measurements at 0.5 Hz were 1053.58 ms (MUSCLEMOTION), 1043.59 ms (CONTRACTIONWAVE), and 937.11 ms (ViKiE). ViKiE exhibited higher sensitivity in exposed samples due to its localized kinematic analysis, while MUSCLEMOTION and CONTRACTIONWAVE offered temporal correlation, combining global assessment with time-efficient analysis. Finally, machine learning reveals greater accuracy when trained with MUSCLEMOTION dataset in comparison with the other software (accuracy > 83%). In conclusion, our findings provide valuable insights for the accurate selection and integration of software tools into the kinematic analysis pipeline, tailored to the experimental protocol.

Single-cell RNA sequencing analysis identifies one subpopulation of endothelial cells that proliferates and another that undergoes the endothelial-mesenchymal transition in regenerating pig hearts

Background: In our previous work, we demonstrated that when newborn pigs undergo apical resection (AR) on postnatal day 1 (P1), the animals' hearts were completely recover from a myocardial infarction (MI) that occurs on postnatal day 28 (P28); single-nucleus RNA sequencing (snRNAseq) data suggested that this recovery was achieved by regeneration of pig cardiomyocyte subpopulations in response to MI. However, coronary vasculature also has a key role in promoting cardiac repair. Method: Thus, in this report, we used autoencoder algorithms to analyze snRNAseq data from endothelial cells (ECs) in the hearts of the same animals. Main results: Our results identified five EC clusters, three composed of vascular ECs (VEC1-3) and two containing lymphatic ECs (LEC1-2). Cells from VEC1 expressed elevated levels of each of five cell-cyclespecific markers (Aurora Kinase B [AURKB], Marker of Proliferation Ki-67 [MKI67], Inner Centromere Protein [INCENP], Survivin [BIRC5], and Borealin [CDCA8]), as well as a number of transcription factors that promote EC proliferation, while (VEC3 was enriched for genes that regulate intercellular junctions, participate in transforming growth factor β (TGFβ), bone morphogenic protein (BMP) signaling, and promote the endothelial mesenchymal transition (EndMT). The remaining VEC2 did not appear to participate directly in the angiogenic response to MI, but trajectory analyses indicated that it may serve as a reservoir for the generation of VEC1 and VEC3 ECs in response to MI. Notably, only the VEC3 cluster was more populous in regenerating (i.e., ARP1MIP28) than non-regenerating (i.e., MIP28) hearts during the 1-week period after MI induction, which suggests that further investigation of the VEC3 cluster could identify new targets for improving myocardial recovery after MI. Histological analysis of KI67 and EndMT marker PDGFRA demonstrated that while the expression of proliferation of endothelial cells was not significantly different, expression of EndMT markers was significantly higher among endothelial cells of ARP1MIP28 hearts compared to MIP28 hearts, which were consistent with snRNAseq analysis of clusters VEC1 and VEC3. Furthermore, upregulated secrete genes by VEC3 may promote cardiomyocyte proliferation via the Pi3k-Akt and ERBB signaling pathways, which directly contribute to cardiac muscle regeneration. Conclusion: In regenerative heart, endothelial cells may express EndMT markers, and this process could contribute to regeneration via a endothelial-cardiomyocyte crosstalk that supports cardiomyocyte proliferation.