GATA Transcription Factors Coordinate Developmental Progression in the Arabidopsis Root

For proper organ development, cell proliferation and cellular differentiation must balance in a tightly regulated spatiotemporal manner. Here, we show that a phased gene expression profile from meristematic cells at the Arabidopsis root niche, to differentiating cortex- and endodermis cells, is defined by the action of distinct transcription factor (TF) families. We show that while chromatin accessibility serves as a strong predictor of the activation of developmental genes, a key expression 'switch' from dividing- to elongating/differentiating cells is underpinned by a considerable level of chromatin rearrangement. Within this structure, a GATA TF subfamily is proposed to sit at the heart of a mutually antagonistic regulatory module, acting to promote differentiation and feedback on cell proliferation through the regulation of both lineage specific differentiation processes and general maturation programs. Key to this is integration of the auxin- and cytokinin responsive PLT- and ARR-B TF families. This model provides a level of coordination in the regulation of meristem size and the timing of differentiation, at both the tissue- and organ levels. Overall design: Roots of six day old seedlings or A. thaliana Col-0 were cut 0.5 - 0.7 cm from their tip and a total of 180 mg of root tissue was collected. Single cell ATAC-seq libraries were produced from freshly isolated nuclei in accordance with the 10X Genomics Chromium Single Cell ATAC-seq kit protocol (v 1.1 protocol, Rev D). Bulk RNA-seq samples are added. Two previously published (Denyer et al., 2019) datasets (Series GSE123818; samples GSM3511858 and GSM3511859) were reanalyzed to improve atlas quality. Reads from two replicates were processed with Cell Ranger (v7.1.0; 10X Genomics) using the TAIR10 reference genome and TAIR10.55 genome annotation. Downstream processing was done with the Seurat R package (v5.1.0; Hao et al., 2024). Cells with fewer than 500-, or more than 100,000 UMIs, were removed from subsequent analysis, as were genes detected in fewer than five cells. Further, cells containing more than 5 % reads mapped to mitochondrial genes were removed. Count matrices were normalized with the ?SCTransform? function, and replicates integrated following Seurat?s integration pipeline using the canonical correlation analysis (CCA) method and 50 principal components. Genes previously identified as induced by the protoplasting procedure (Denyer et al., 2019) were excluded from the integration procedure. UMAP visualizations were generated using integrated dimensionality reduction and cell clusters identified with the resolution of 0.5. Clusters were annotated using tissue specific marker genes as previously described (Denyer et al., 2019). *************************************************************** The table below lists GEO accessions reused/reanalyzed for this study. ***************************************************************