Cross-species single-cell transcriptomic integration for plant cell type identification

Overview

This research develops computational methods to integrate single-cell transcriptomic datasets across multiple plant species, enabling direct biological comparison and the identification of homologous cell types and cell states. By creating a common embedding space using orthogroup-based approaches, the work addresses a key challenge in plant biology: transferring knowledge from well-characterized model species like Arabidopsis to commercially important crops.

The research focuses on flowers, which determine fruit and seed productivity, and includes three commercial crops. This integration capability will reveal how evolutionary conservation and divergence shape plant physiology, and will permit the transference of biological insights across species boundaries.

Technical specifications

Key features:

• Integration of single-cell floral transcriptomic atlases from four species with different evolutionary histories: two eudicots (Arabidopsis and Brassica rapa) and two monocots (asparagus and banana)
• Use of orthogroups (sets of homologous genes) inferred through multiple algorithms to provide a shared gene space, circumventing reliance on single-copy orthologs in divergent species
• Application of the 10x Genomics platform, where single protoplasts or nuclei are combined with barcoded gel beads on a microfluidic device before library preparation and sequencing
• Mapping of transcriptomes to reference genomes to identify cell populations sharing transcriptional features
• Marker gene identification characterizing each cell population, validated through reporter constructs in Arabidopsis
• Label transfer methodology extending cell identity from model Arabidopsis to crop species

Technology readiness level

The project builds upon already-generated single-cell floral atlases for all four species, with thousands of floral cells including those with key roles in plant reproduction already sequenced and mapped. Cell populations have been identified and marker genes characterized. Future validation will perform integration between closely related species (Arabidopsis and Brassica) to explore parameter space, then extend to more divergent species (asparagus and banana) and finally all four species together. The study is estimated to take one year to complete, with plans to extend the approach to other crops of interest.