Integrative Single-Cell RNA-Seq and ATAC-Seq Analysis of Mesenchymal Stem/Stromal Cells Derived from Human Placenta

Abstract

Mesenchymal stem/stromal cells derived from placenta (PMSCs) are an attractive source for regenerative medicine because of their multidifferentiation potential and immunomodulatory capabilities. However, the cellular and molecular heterogeneity of PMSCs has not been fully characterized. Here, we applied single-cell RNA sequencing (scRNA-seq) and assay for transposase-accessible chromatin sequencing (scATAC-seq) techniques to cultured PMSCs from human full-term placenta. Based on the inferred characteristics of cell clusters, we identify several distinct subsets of PMSCs with specific characteristics, including immunomodulatory-potential and highly proliferative cell states. Furthermore, integrative analysis of gene expression and chromatin accessibility showed a clearer chromatin accessibility signature than those at the transcriptional level on immunomodulatory-related genes. Cell cycle gene-related heterogeneity can be more easily distinguished at the transcriptional than the chromatin accessibility level in PMSCs. We further reveal putative subset-specific cis-regulatory elements regulating the expression of immunomodulatory- and proliferation-related genes in the immunomodulatory-potential and proliferative subpopulations, respectively. Moreover, we infer a novel transcription factor PRDM1, which might play a crucial role in maintaining immunomodulatory capability by activating PRDM1-regulon loop. Collectively, our study first provides a comprehensive and integrative view of the transcriptomic and epigenomic features of PMSCs, which paves the way for a deeper understanding of cellular heterogeneity and offers fundamental biological insight of PMSC subset-based cell therapy.

A Biophysical Model for ATAC-seq Data Analysis

“…In this sense, the possible states of a locus form a partially ordered set. For instance, 010100 ≺ 010101, yet 010100 ⊀ 011001, (20) where the partial order, 𝑎 ≺ 𝑏, indicates that sequence 𝑎 can be obtained, starting from 𝑏, via technical noise. In this example, the first two strings are connected by a transition encoded in N , but the second two are not: the technical noise process cannot create false positives.…”

“…Such methods have been used to compare gene regulation across tissues between healthy and pathological conditions (10)(11)(12)(13)(14)(15), and across tissue types (16), as well as to identify new gene-regulatory pathways (17,18). Combined use of scRNA-seq and scATAC-seq data can also help to distinguish cell-types (19,20) and scATAC data can be used to impute scRNA data (21). An exciting recent development has been the ability to perform registered scATAC-seq and scRNA-seq, identifying DNA fragments and mRNA transcripts from the same individual cells (22,23).…”

“…While both registered and unregistered data have been useful for a variety of applications, analysis methods have not incorporated biophysical modeling, and have instead relied on heuristics and phenomenological inferences (20). While progress has been made in building biophysically motivated and interpretable models for RNA-seq data (26)(27)(28), and Gorin et al have outlined a general chemical master equation (CME) formulation for describing various models of transcription with intrinsic and extrinsic noise (29), these tools have not been applied to ATAC-seq, or joint analysis of ATAC-seq and RNA-seq.…”

Genome-wide identification of transcriptional enhancers during human placental development and association with function, differentiation, and disease

“…In this regard, a number of studies using microarrays [13][14][15][16], bulk RNA-seq [17][18][19], and single cell RNA-seq (scRNA-seq) [20][21][22][23][24][25][26][27] have examined aspects of how gene expression in the placenta changes over time in normal or pathologic pregnancies [10,12,28]. A number of studies have focused on differences in gene expression in placenta tissue in response to disease [12,22,[27][28][29][30][31][32][33], while others have used genomic profiling to define distinct cell types in the placenta [10,11,21,[33][34][35][36][37]. Fewer studies, however, have examined how diverse genomic features [e.g., transcription, histone modification, DNA methylation, transcription factor (TF) binding, chromatin accessibility, and chromatin interactions] change over the course of normal development.…”

Single cell RNA sequencing reveals mesenchymal heterogeneity and critical functions of Cd271 in tooth development

“…scRNA-Seq also provides insights into specific changes in cell lineages, trajectory inference, and the identification of biomarkers[ 21 , 22 ]. scRNA-seq has been used to study MSCs derived from adipose tissue, bone marrow, endometrium, placenta and dental pulp[ 23 - 27 ]. However, application of scRNA-seq in maxillary-process-derived MSCs is still absent.…”