A multiomic atlas of human early skeletal development

Ken To, Lijiang Fei, J. Patrick Pett, Kenny Roberts, Krzysztof Polański, Tong Li, Nadav Yayon, Peng He, Chuan Xu, James Cranley, Ruoyan Li, Kazumasa Kanemaru, Ni Huang, Stathis Megas, Laura Richardson, Rakesh Kapuge, Shani Perera, Elizabeth Tuck, Anna Wilbrey-Clark, Ilaria Mulas, Fani Memi, Batuhan Çakır, Alexander V. Predeus, Dave Horsfall, Simon Murray, Martin Prete, Pavel Mazin, Xiaoling He, Kerstin B. Meyer, Muzlifah Haniffa, Roger A. Barker, Omer Ali Bayraktar, Christopher D. Buckley, Sarah A. TeichmannbioRxiv 2024 · DOI

TL;DR

This paper presents the first comprehensive multiomic atlas of early human skeletal development, mapping cell types, gene regulation, and spatial organization from 5 to 11 post-conception weeks. This resource reveals novel cell states and regulatory networks, advancing our understanding of skeletal formation and its links to diseases like osteoarthritis and craniosynostosis.

What Problem Does This Paper Address?

The intricate processes of bone and joint formation during early human development, including the differentiation of progenitor cells into specialized bone and cartilage cells, are not well understood. Specifically, the precise cell states, epigenetic mechanisms, and regulatory factors that guide these lineage commitments during ossification and joint formation have remained largely unexplored at a high resolution. This knowledge gap hinders our ability to understand developmental disorders and regenerative medicine approaches.

Methods

The study employed a multi-modal approach combining:

  • Paired Single-Nuclei Transcriptional and Epigenetic Profiling (snRNA-seq and snATAC-seq): This allowed for simultaneous measurement of gene expression and chromatin accessibility in individual nuclei, providing a detailed view of cellular states and regulatory potential.
  • Spatial Transcriptomics (In Situ Sequencing - ISS, Visium CytAssist): These techniques enabled the mapping of cell types and gene expression within the anatomical context of developing tissues, revealing spatial organization and zonation.
  • Novel Computational Tools:
    • ISS-Patcher: A tool developed to impute cell labels from low-resolution droplet-based data onto high-resolution ISS data, enhancing spatial resolution.
  • OrganAxis: A tool for annotating spatial transcriptomics data to define developmental axes and trajectories.
  • SNP2Cell: A tool that integrates GWAS data with cell-type specific regulatory networks to identify potential causal links to complex traits like osteoarthritis.
  • In Silico Perturbations: Gene knockouts and overexpression simulations were used to predict the functional impact of transcription factors on developmental trajectories.
  • Trajectory Analysis (scFates, Palantir, Monocle3): These methods were used to reconstruct developmental pathways and infer cell fate decisions.
  • Gene Regulatory Network Analysis (SCENIC+): Used to predict transcription factor binding sites and gene regulatory networks.

Datasets

The study utilized human fetal skeletal tissue samples from 5 to 11 post-conception weeks (PCW), covering various anatomical regions including the shoulder, hip, knee, and cranium. The data generated includes:

  • 336,162 high-quality transcriptome and chromatin-accessibility droplets from paired snRNA-seq and snATAC-seq.
  • Spatial transcriptomics data from multiplexed in situ sequencing (ISS) and Visium CytAssist.

The processed data is available for visualization and download at https://developmental.cellatlas.io/skeleton-development.

Key Insights and Ideas

  • Unprecedented Cellular Diversity: The atlas revealed previously undescribed cellular diversity within osteogenic and chondrogenic lineages, including novel chondrocyte cell states.
  • Spatial Zonation of Joint Development: Spatial mapping identified distinct cellular mechanisms of zonation during bone and joint formation, particularly in the synovial joint.
  • Differential Ossification Mechanisms: The study characterized regionally distinct osteoprogenitor populations and regulatory networks governing both intramembranous and endochondral ossification.
  • Linking Development to Disease: The atlas provides a framework to link cell-type specific regulatory networks to polygenic traits like osteoarthritis and to investigate mechanisms of monogenic disorders like craniosynostosis.
  • Potential Cross-Lineage Contributions: Evidence suggests potential contributions of muscle and Schwann cell lineages to chondrogenesis, challenging traditional views.

Key Results

  • Comprehensive Atlas: The study generated a detailed, multiomic atlas of human early skeletal development, encompassing over 336,000 cells.
  • Identification of Novel Cell States: Defined 122 fine-grained cell clusters, including novel chondrocyte and osteoprogenitor populations.
  • Spatial Mapping of Joint Formation: Visualized the spatial organization and differentiation trajectories within the developing synovial joint.
  • Regulatory Landscape of Ossification: Identified distinct transcriptional and epigenetic regulatory landscapes driving intramembranous and endochondral ossification.
  • Disease Association: Implicated specific cell states and regulatory networks in osteoarthritis and craniosynostosis through fGWAS and in silico perturbation studies.
  • Development of New Tools: Introduced ISS-Patcher and SNP2Cell, valuable resources for spatial transcriptomics analysis and disease gene association studies.

Limitations and Future Work

  • Temporal Resolution: While covering 5-11 PCW, finer temporal resolution could further elucidate dynamic changes.
  • In Vitro Validation: The study primarily relies on computational predictions; future work could involve experimental validation of identified regulatory networks and cell states.
  • Broader Disease Scope: While craniosynostosis and osteoarthritis were investigated, the atlas has potential to inform research into other skeletal disorders.
  • Cell-Cell Interactions: While ligand-receptor interactions were predicted, further experimental validation of these communication pathways would be beneficial.
  • Second Trimester Development: The authors note that further studies of the second trimester are needed for a complete census of joint development.

Summary updated

Abstract

Abstract Bone and joint formation in the developing skeleton rely on co-ordinated differentiation of progenitors in the nascent developing limbs and joints. The cell states, epigenetic processes and key regulatory factors underlying their lineage commitment to osteogenic and other mesenchymal populations during ossification and joint formation remain poorly understood and are largely unexplored in human studies. Here, we apply paired single-nuclei transcriptional and epigenetic profiling of 336,000 droplets, in addition to spatial transcriptomics, to construct a comprehensive atlas of human bone, cartilage and joint development in the shoulder, hip, knee and cranium from 5 to 11 post-conception weeks. Spatial mapping of cell clusters to our highly multiplexed in situ sequencing (ISS) data using our newly developed tool ISS-Patcher revealed new cellular mechanisms of zonation during bone and joint formation. Combined modelling of chromatin accessibility and RNA expression allowed the identification of the transcriptional and epigenetic regulatory landscapes that drive differentiation of mesenchymal lineages including osteogenic and chondrogenic lineages, and novel chondrocyte cell states. In particular, we define regionally distinct limb and cranial osteoprogenitor populations and trajectories across the fetal skeleton and characterise differential regulatory networks that govern intramembranous and endochondral ossification. We also introduce SNP2Cell, a tool to link cell-type specific regulatory networks to numerous polygenic traits such as osteoarthritis. We also conduct in silico perturbations of genes that cause monogenic craniosynostosis and implicate potential pathogenic cell states and disease mechanisms involved. This work forms a detailed and dynamic regulatory atlas of human fetal skeletal maturation and advances our fundamental understanding of cell fate determination in human skeletal development.
BibTeX
@misc{to2024multiomic,
  author = {Ken To and Lijiang Fei and J. Patrick Pett and Kenny Roberts and Krzysztof Polański and Tong Li and Nadav Yayon and Peng He and Chuan Xu and James Cranley and Ruoyan Li and Kazumasa Kanemaru and Ni Huang and Stathis Megas and Laura Richardson and Rakesh Kapuge and Shani Perera and Elizabeth Tuck and Anna Wilbrey-Clark and Ilaria Mulas and Fani Memi and Batuhan Çakır and Alexander V. Predeus and Dave Horsfall and Simon Murray and Martin Prete and Pavel Mazin and Xiaoling He and Kerstin B. Meyer and Muzlifah Haniffa and Roger A. Barker and Omer Ali Bayraktar and Christopher D. Buckley and Sarah A. Teichmann},
  title = {{A multiomic atlas of human early skeletal development}},
  year = {2024},
  doi = {10.1101/2024.07.10.602965},
  howpublished = {bioRxiv},
}
Figure