Single-cell network analysis reveals gene expression programs for Arabidopsis root development and metabolism

Abstract

Single-cell RNA-seq (scRNA-seq) datasets of Arabidopsis roots have been generated, but related comprehensive gene co-expression network analyses are lacking. We conducted a single-cell gene co-expression network analysis with publicly available scRNA-seq datasets of Arabidopsis roots using a SingleCellGGM algorithm. The analysis identified 149 gene co-expression modules, which we considered gene expression programs (GEPs). By checking their spatiotemporal expression, we identified GEPs specifically expressed in major root cell types along their developmental trajectories. These GEPs defined gene programs regulating root cell development at different stages and are enriched with relevant developmental regulators. As examples, a GEP specific for quiescent center (QC) contains 20 genes regulating QC and stem cell niche homeostasis, and five GEPs are expressed in sieve elements (SEs) from early to late developmental stages, with the early-stage GEP containing 17 known SE developmental regulators. We also identified GEPs for metabolic pathways with cell type-specific expression, suggesting the existence of cell type-specific metabolism in roots. Using the GEPs, we discovered and verified a columella-specific gene, NRL27, as a regulator of auxin-related root gravitropism response. Our analysis thus systematically revealed GEPs regulating Arabidopsis root development and metabolism and provided candidate genes for root biology studies.