Temporal evolution of single-cell transcriptomes of Drosophila olfactory projection neurons

Qijing Xie, Maria Brbić, Felix Horns, Sai Saroja Kolluru, Robert C. Jones, Jun Li, Anay R. Reddy, Anthony Xie, Sayeh Kohani, Zhuoran Li, Colleen N. McLaughlin, Tongchao Li, Chuanyun Xu, David Vacek, David J. Luginbuhl, Jure Leskovec, Stephen R. Quake, Liqun Luo, Hongjie LieLife 2021 · DOI

TL;DR

This study maps the dynamic transcriptomic landscape of Drosophila olfactory projection neurons (PNs) across four developmental stages, revealing how gene expression changes drive distinct biological processes like neurite pruning, circuit assembly, and adult function. By developing novel methods to track PN types over time, the authors provide a valuable resource for understanding neural development and cell-type diversity.

What Problem Does This Paper Address?

The precise wiring and function of the nervous system rely on the incredible diversity of neurons and their specific connections. While the Drosophila olfactory projection neurons (PNs) are a well-studied model system, the underlying transcriptomic basis for their development, diversification, and functional specialization across different life stages remained largely unknown. Understanding these transcriptomic changes is crucial for deciphering the molecular mechanisms that govern neural circuit formation and function.

Methods

The researchers employed single-cell RNA sequencing (scRNA-seq) to profile PNs at four key developmental stages: 0-6 hr, 24-30 hr, 48-54 hr after puparium formation (APF), and adult. They utilized a combination of GAL4 driver lines (GH146-GAL4, VT033006-GAL4, and several split-GAL4 lines) to label specific PN populations.

Key methodological contributions include:

  • Decoding PN identities: They developed strategies to match transcriptomic clusters to specific PN types, overcoming the lack of universal type-specific markers by identifying combinatorial gene expression patterns.
  • Tracking cell types across development: Novel computational approaches, including similarity calculations (Jaccard index) and marker-based matching, were used to reliably track the same PN types across different developmental stages, despite significant transcriptomic shifts.
  • Dimensionality reduction and clustering: Techniques like tSNE and HDBSCAN were used to visualize and group cells based on their gene expression profiles.
  • MARS (Meta-learned Representations for Single-cell data): This deep learning approach was employed to annotate cell types across developmental stages, leveraging the highly informative 24 hr APF stage as a reference.

Datasets

The study generated a new dataset of approximately 4900 high-quality single-cell transcriptomes from Drosophila PNs across four developmental stages. This dataset has been deposited in the Gene Expression Omnibus (GEO) under accession code GSE161228.

Previously published datasets were also utilized for comparison and validation.

Key Insights and Ideas

  • Developmental stages have distinct transcriptomic signatures: PNs undergo dramatic transcriptomic shifts throughout development, reflecting stage-specific biological processes.
  • Peaked diversity at circuit assembly: Transcriptomic diversity among PN types is highest around 24 hr APF, coinciding with the critical period of olfactory circuit assembly and wiring decisions.
  • Birth order influences early transcriptomes: At early developmental stages (0 and 24 hr APF), PNs with adjacent birth orders exhibit more similar transcriptomes, suggesting a carry-over of progenitor-related gene expression patterns. This similarity diminishes as PNs mature.
  • Adult PNs specialize in signaling: While transcriptomic differences between PN types decrease in adults, they still exhibit distinct gene expression profiles, particularly related to neuronal signaling and ion transport, reflecting functional specialization.
  • A valuable resource for future research: The comprehensive dataset and refined methods provide a foundation for identifying novel genes involved in PN development, wiring, and function.

Key Results

  • 21 transcriptomic clusters identified: Corresponding to 20 distinct PN types at the 24 hr APF stage.
  • Successful matching of PN types across stages: The authors were able to match many PN types across multiple developmental stages, enabling the study of temporal gene expression dynamics.
  • Identification of stage-specific gene expression programs:
    • 0 hr APF: Characterized by neurite pruning, with enrichment of genes involved in morphogenesis and cytoskeleton organization.
  • 24 hr APF: High transcriptomic diversity, reflecting the complexity of olfactory circuit assembly and wiring.
  • Adult: Enriched for genes involved in neuronal signaling and ion transport.
  • Discovery of ~500 differentially expressed genes: Across different PN types, with a significant proportion being transcription factors and cell surface molecules, many of which are implicated in wiring.

Limitations and Future Work

  • Batch effects: While addressed, some residual batch effects were observed between cells sequenced using different platforms, necessitating correction methods.
  • Incomplete decoding: Not all PN types could be definitively decoded, particularly for some vPNs and IPNs, highlighting the need for new driver lines.
  • Post-transcriptional regulation: The study acknowledges that transcriptomic analysis may miss regulatory mechanisms operating at the post-transcriptional level.

Future work suggested by the authors includes:

  • Leveraging the dataset for candidate gene discovery: Identifying novel molecules involved in neurite pruning, neural development, wiring specificity, and adult neuronal signaling.
  • Developing driver lines for specific PN types: Based on genes with consistent expression patterns across development.
  • Investigating post-transcriptional regulation: To gain a more complete understanding of gene function.
  • Applying the developed methods to other complex tissues: Where unique markers for cell types are scarce.

Summary updated

Abstract

Neurons undergo substantial morphological and functional changes during development to form precise synaptic connections and acquire specific physiological properties. What are the underlying transcriptomic bases? Here, we obtained the single-cell transcriptomes of <i>Drosophila</i> olfactory projection neurons (PNs) at four developmental stages. We decoded the identity of 21 transcriptomic clusters corresponding to 20 PN types and developed methods to match transcriptomic clusters representing the same PN type across development. We discovered that PN transcriptomes reflect unique biological processes unfolding at each stage-neurite growth and pruning during metamorphosis at an early pupal stage; peaked transcriptomic diversity during olfactory circuit assembly at mid-pupal stages; and neuronal signaling in adults. At early developmental stages, PN types with adjacent birth order share similar transcriptomes. Together, our work reveals principles of cellular diversity during brain development and provides a resource for future studies of neural development in PNs and other neuronal types.
BibTeX
@inproceedings{xie2021temporal,
  author = {Qijing Xie and Maria Brbić and Felix Horns and Sai Saroja Kolluru and Robert C. Jones and Jun Li and Anay R. Reddy and Anthony Xie and Sayeh Kohani and Zhuoran Li and Colleen N. McLaughlin and Tongchao Li and Chuanyun Xu and David Vacek and David J. Luginbuhl and Jure Leskovec and Stephen R. Quake and Liqun Luo and Hongjie Li},
  title = {{Temporal evolution of single-cell transcriptomes of Drosophila olfactory projection neurons}},
  booktitle = {eLife},
  year = {2021},
  doi = {10.7554/elife.63450},
}
Figure