The molecular definition of the differentiation process of cardiovascular lineages is of intense interest for developmental biologists. Recent studies have demonstrated that multipotent cardiac progenitors are capable of giving rise to multiple cardiovascular lineages such as endothelium, endocardium, epicardium and myocardium. Moreover, Nkx2-5 and Etv2 are the key genes that regulate the specification of mesodermal lineages. Understanding the global gene expression pattern and ontogeny of Nkx2-5 and Etv2 positive cells are critical steps for the decoding of the signaling cascade and the molecular mechanism that govern mesodermal lineage differentiation.
Using Nkx2-5- and Etv2-EYFP transgenic embryos, from which progenitors of cardiovascular/circulatory lineages can be isolated with high purity, these researchers performed single cell transcriptome analysis. Single cell RNA sequencing (scRNA-seq) is a new exciting technique that allows profiling of the expression profiles at the single cell level. However, due to the heterogeneity of the scRNA-seq profiles and high technical noise (especially the dropout noise), accurately discovering the hidden cell population and reconstructing the complex lineage hierarchies from the scRNA-seq data became a challenging task in this field. To address the dropout noise, extract features from the high dimensional scRNA-seq data, and reconstruct the differentiation pathways, the researchers developed a series of novel concepts and computational methods including: dpath (recovering the lineage trajectories from the scRNA-seq data); TCM (topographic cell map - visualizing the temporal scRNA-seq data); DrImpute (imputing the dropout events in the scRNA-seq data); scNCA (neighborhood component analysis for single cell RNA-seq - correcting the batch effects of temporal scRNA-seq data); DCLEAR (distance-based cell lineage reconstruction; SeATAC (V-plot analysis of ATAC-seq by deep learning); and EnsembleMerge (Integration of single cell RNA-seq datasets using an ensemble approach). These scRNA-seq analysis methods have proved superior to conventional methods. The results demonstrated the power of combining machine learning algorithms and single cell RNA-seq for discovering hidden cell types, novel lineage markers and reconstructing lineage hierarchies.
In 2024, the lab plans to generate the single cell expression profiles of other important mesodermal genes such as T (Brachyury), Kdr, Gata1, Isl1, etc. They will also study the role of Etv2 as a pioneer factor during reprogramming using the single cell technique. They will collaborate with Jay Zhang's group at the University of Birmingham, Alabama, to use scRNA-seq and snRNA-seq to study pig injury models and to study cardiovascular disease in large anmial models. They aim to improve the current single cell RNA-seq pipeline, and develop new computational methods for discovering various differentiation pathways, predicting key genes controlling for lineage switching, and integrating multiple dimensional genomic data at the single cell level. It is hoped that this entire project will generate a full blueprint of lineage hierarchies during early embryonic development across multiple species and provides insights into related biological mechanisms, disease process, and designing therapeutic strategies.