Single cell transcriptomics reveals chondrocyte differentiation dynamics in vivo and in vitro

John E. Lawrence, Steven Paul Woods, Kenny Roberts, Dinithi Sumanaweera, Petra Balogh, Alexander V. Predeus, Peng He, Tong Li, Krzysztof Polański, Elena Prigmore, Elizabeth Tuck, Lira Mamanova, Di Zhou, Simone Webb, Laura Jardine, Xiaoling He, Roger A. Barker, Muzlifah Haniffa, Adrienne M. Flanagan, Matthew D. Young, Sam Behjati, Omer Ali Bayraktar, Susan J. Kimber, Sarah A. TeichmannbioRxiv 2023 · DOI

TL;DR

This study tackles the challenge of reliably producing functional chondrocytes in vitro for regenerative medicine and disease modeling. By creating a detailed single-cell atlas of human embryonic skeletal development, the researchers benchmarked existing in vitro methods, revealing significant variability and off-target differentiation. They then used this atlas to identify a way to improve in vitro chondrogenesis by inhibiting a key transcription factor, FOXO1.

What Problem Does This Paper Address?

The goal of creating lab-grown chondrocytes that perfectly mimic those found in living organisms is crucial for developing treatments for musculoskeletal diseases and for accurately modeling these conditions. However, current methods for generating these cells in vitro often result in a mixed bag of cell types, with some cells differentiating into unintended lineages (off-target differentiation). This heterogeneity makes it difficult to study specific cellular processes or to develop reliable regenerative therapies. A major hurdle has been the lack of a detailed reference map of how chondrocytes develop naturally in humans, making it hard to compare and improve upon in vitro methods.

Methods

The core of this study involved single-cell RNA sequencing (scRNA-seq). The researchers applied this technique to:

  • Human embryonic hindlimb samples: They collected and sequenced cells from first-trimester human embryos at various gestational ages to build a comprehensive atlas of endochondral ossification.
  • In vitro chondrogenesis protocols: They analyzed existing published protocols designed to generate chondrocytes from stem cells and other sources.
  • Single-nuclei RNA sequencing (snRNA-seq): This was used for a more direct comparison between in vitro and in vivo cells.

To analyze the vast amounts of data generated, they employed:

  • Trajectory alignment algorithms (Genes2Genes): This computational approach allowed them to compare the developmental paths of cells in vitro with the natural developmental paths observed in vivo, highlighting similarities and differences.
  • Transcription factor network inference (DoRothEA): This helped identify key regulatory pathways driving chondrogenesis.
  • RNA in situ hybridization: This technique was used to visualize gene expression patterns directly within tissue sections, validating scRNA-seq findings.
  • Targeted gene inhibition: They used a specific inhibitor for the transcription factor FOXO1 to test its role in regulating chondrocyte differentiation.

Datasets

  • Newly generated dataset: Single-cell RNA sequencing data from 18 human embryonic hindlimb samples (7-9 post-conception weeks).
  • Publicly available datasets: Data from previous studies on embryonic limb development and various in vitro chondrogenesis protocols.
  • Data Repository: All newly generated data is available via the European Genome-Phenome archive (accession number EGAD00001011306). Previously published data is available from ArrayExpress. Processed data can also be visualized at the portal: https://chondrocytes.cellgeni.sanger.ac.uk/

Key Insights and Ideas

  • A detailed in vivo roadmap: The creation of a high-resolution single-cell atlas of human embryonic skeletal development provides an invaluable reference for evaluating and improving in vitro systems.
  • In vitro variability is high: Existing in vitro chondrogenesis protocols show substantial differences in the cell states they produce, with many exhibiting off-target differentiation.
  • FOXO1 inhibition improves chondrogenesis: Targeting FOXO1, a transcription factor implicated in bone development, can steer cells towards a more desirable chondrocyte phenotype in vitro. This demonstrates the power of using developmental atlases to guide tissue engineering.
  • Off-target transcripts, not necessarily off-target cells: The study suggests that in vitro chondrocytes might aberrantly express some off-target transcripts rather than being a mix of completely different cell types, which has implications for understanding tissue function.

Key Results

  • The study generated a developmental atlas of 249,151 cells spanning 5-17 post-conception weeks, revealing diverse cell types involved in skeletal development.
  • Analysis of various in vitro protocols showed substantial variability in cell states and evidence of off-target differentiation in all evaluated methods.
  • Trajectory alignment revealed that, on average, 58% of highly variable genes and 56% of transcription factors showed matched expression patterns between in vitro and in vivo trajectories.
  • Inhibition of FOXO1 with AS1842856 at specific concentrations (4 nM and 20 nM) led to an increased expression of chondrocyte genes (e.g., COL2A1, COL9A1, MATN3) and hypertrophic markers in vitro.

Limitations and Future Work

  • Sampling limitations: The study acknowledges that capturing the full spectrum of cell types, especially mature hypertrophic chondrocytes, can be challenging due to tissue dissociation limitations and the sampling window for embryonic tissues.
  • In vitro variability: The authors note that inter-experiment variability in in vitro differentiation is a known challenge, and even within their own protocol, subtle differences were observed between runs.
  • Future directions: The authors suggest that future work should focus on fine-tuning protocols by targeting other molecules implicated in embryonic chondrogenesis to minimize off-target transcription. They also highlight the need for more densely sampled studies to capture rapid differentiation dynamics and the importance of assessing protein-level and mechanical properties to fully evaluate engineered tissues. Furthermore, exploring the role of Schwann cell precursors in chondrocyte lineage warrants further investigation.

Summary updated

Abstract

Summary The consistent production of in vitro chondrocytes that faithfully recapitulate in vivo development would be of great benefit for musculoskeletal disease modelling and regenerative medicine. Current efforts are often limited by off-target differentiation, resulting in a heterogeneous product. Furthermore, the lack of comparison to human embryonic tissue, precludes detailed evaluation of in vitro cells. Here, we perform single-cell RNA sequencing of embryonic long bones dissected from first trimester hind limbs from a range of gestational ages. We combine this with publicly available data to form a detailed atlas of endochondral ossification, which we then use to evaluate a series of published in vitro chondrogenesis protocols, finding substantial variability in cell states produced by each. We apply single-nuclear RNA sequencing to one protocol to enable direct comparison between in vitro and in vivo, and perform trajectory alignment between the two to reveal differentiation dynamics at the single-cell level, shedding new light on off-target differentiation in vitro . Using this information, we inhibit the activity of FOXO1, a transcription factor predicted to be active in embryonic bone development and in chondrogenic cells in vitro , and increase chondrocyte transcripts in vitro. This study therefore presents a new framework for evaluating tissue engineering protocols, using single-cell data from human development to drive improvement and bring the prospect of true engineered cartilage closer to reality.
BibTeX
@misc{lawrence2023single,
  author = {John E. Lawrence and Steven Paul Woods and Kenny Roberts and Dinithi Sumanaweera and Petra Balogh and Alexander V. Predeus and Peng He and Tong Li and Krzysztof Polański and Elena Prigmore and Elizabeth Tuck and Lira Mamanova and Di Zhou and Simone Webb and Laura Jardine and Xiaoling He and Roger A. Barker and Muzlifah Haniffa and Adrienne M. Flanagan and Matthew D. Young and Sam Behjati and Omer Ali Bayraktar and Susan J. Kimber and Sarah A. Teichmann},
  title = {{Single cell transcriptomics reveals chondrocyte differentiation dynamics <i>in vivo</i> and <i>in vitro</i>}},
  year = {2023},
  doi = {10.1101/2023.12.20.572425},
  howpublished = {bioRxiv},
}
Figure