Single cell sequencing reveals gene expression signatures associated with bone marrow stromal cell subpopulations and time in culture

Shutong Liu, David F. Stroncek, Yingdong Zhao, Victoria Chen, Rongye Shi, Jinguo Chen, Jiaqiang Ren, Hui Liu, Hee Joon Bae, Steven L. Highfill, Ping JinJournal of Translational Medicine 2019 · DOI

TL;DR

This study leverages single-cell RNA sequencing (scRNA-seq) to dissect the heterogeneity of bone marrow stromal cells (BMSCs) and how their gene expression profiles change with time in culture. The researchers identified distinct subpopulations within BMSCs, with a particular focus on a subpopulation expressing high levels of FGFR2 in early passages, which may represent skeletal stem cells.

What Problem Does This Paper Address?

Bone marrow stromal cells (BMSCs) are a promising cell therapy for various conditions due to their regenerative and immunomodulatory properties. However, BMSCs are a heterogeneous population, and their characteristics can change significantly with time in culture (passage number). This heterogeneity and plasticity make it challenging to standardize BMSC products for clinical use and to fully understand their functional potential. Previous studies using bulk analysis masked this cellular diversity. This paper aims to address this by using single-cell resolution to identify and characterize distinct BMSC subpopulations and how they evolve over time in culture.

Methods

The study employed a multi-pronged approach:

  • Bulk RNA Sequencing (RNA-Seq) and Microarray Analysis: Initially, unseparated BMSCs from different passages (2, 4, 6, 8, 9, 10) were analyzed using both microarray and RNA-Seq to establish a baseline understanding of gene expression changes over time. This allowed for comparison and validation of the two technologies.
  • Single-Cell RNA Sequencing (scRNA-seq): This was the core technique used to achieve single-cell resolution. BMSCs from selected passages (3, 4, 6, 8, 9) were isolated using microfluidics (Fluidigm C1 system) and their transcriptomes were analyzed by scRNA-seq. This allowed for the identification of distinct cell populations based on their unique gene expression profiles.
  • 3D Diffusion Map Visualization: This technique was used to visualize the relationships between individual cells and to identify clusters or subpopulations based on their transcriptomic similarity.
  • Hierarchical Clustering and Similarity Matrix Analysis: These methods were used to group cells based on their gene expression patterns, revealing how cells from different passages relate to each other.
  • Ingenuity Pathway Analysis (IPA): Used to identify biological pathways enriched in differentially expressed genes between early and late passage BMSCs.

Datasets

The study utilized BMSC samples from a single healthy subject cultured through passage 10. The gene expression data from both bulk and single-cell analyses were generated and analyzed. The raw data is available in the Gene Expression Omnibus (GEO) under accession number GSE124862.

Key Insights and Ideas

  • BMSC Heterogeneity is Significant: Even within cells derived from a single donor, there is substantial transcriptomic diversity that becomes apparent only at the single-cell level.
  • Passage Number Drives Subpopulation Shifts: As BMSCs are cultured over time, the proportions of different subpopulations change, and their gene expression profiles evolve.
  • FGFR2 as an Early Passage/Skeletal Stem Cell Marker: The study identified high expression of FGFR2 in early passage BMSCs, suggesting it could be a marker for skeletal stem cells, which are thought to be more prevalent in early passages and potentially lost with extensive culturing.
  • PLAT as a Late Passage/Senescence Marker: Conversely, PLAT expression was found to increase with passage number, indicating it might be a marker for late-passage BMSCs or senescent cells.
  • Distinct Subpopulations Identified in Passage 3: By analyzing passage 3 BMSCs, the researchers were able to delineate three distinct subpopulations (A, B, and C) with varying expression patterns of key markers like FGFR2, RUNX2, FGF5, VCAM1, and PLAT, suggesting different functional potentials.

Key Results

  • Bulk vs. Single-Cell Analysis: While bulk RNA-Seq and microarray analyses showed similar trends in gene expression changes across passages, single-cell analysis revealed more nuanced differences and distinct subpopulations that were masked in bulk analysis.
  • Clustering of Cells: 3D Diffusion map visualization of single-cell data showed that cells from very early (passage 3) and very late (passage 9) passages were distinct, while middle passages (4, 6, 8) showed considerable overlap.
  • Marker Identification: FGFR2 was identified as a marker for early passage BMSCs, while PLAT was identified as a marker for late passage BMSCs.
  • Subpopulation Characterization: In passage 3 BMSCs, three subpopulations were identified: one with high FGFR2 and low PLAT (potentially skeletal stem cells), one with lower FGFR2 and higher FGF5, and one with high PLAT and VCAM1 (potentially involved in angiogenesis).

Limitations and Future Work

  • Single Subject Study: A significant limitation acknowledged by the authors is that the study was conducted on BMSCs from only one healthy subject. This limits the generalizability of the findings.
  • Need for Larger Studies: Future work should include BMSCs from a greater number of subjects to confirm these findings and to better characterize BMSC subpopulations across diverse individuals.
  • Technological Advancements: The authors suggest that newer, more robust single-cell RNA sequencing platforms will enable even better characterization of BMSC subpopulations.
  • Functional Validation: While the study identifies potential markers and subpopulations, further functional studies are needed to confirm the specific roles and potentials of these identified cell types.

Summary updated

Abstract

Single BMSCs can be separated by microfluidics and their transcriptome analyzed by next generation sequencing. Single cell analysis of early passage BMSCs identified a subpopulation of cells expressing high levels of FGFR2 that might include skeletal stem cells.
BibTeX
@article{liu2019single,
  author = {Shutong Liu and David F. Stroncek and Yingdong Zhao and Victoria Chen and Rongye Shi and Jinguo Chen and Jiaqiang Ren and Hui Liu and Hee Joon Bae and Steven L. Highfill and Ping Jin},
  title = {{Single cell sequencing reveals gene expression signatures associated with bone marrow stromal cell subpopulations and time in culture}},
  journal = {Journal of Translational Medicine},
  year = {2019},
  doi = {10.1186/s12967-018-1766-2},
}
Figure