A single-cell and spatial wheat root atlas with cross-species annotations delineates conserved tissue-specific marker genes and regulators

Abstract

Despite the broad use of single-cell/nucleus RNA sequencing in plant research, accurate cluster annotation in less-studied plant species remains a major challenge due to the lack of validated marker genes. Here, we generated a single-cell RNA sequencing atlas of soil-grown wheat roots and annotated cluster identities by transferring annotations from publicly available datasets in wheat, rice, maize, and Arabidopsis. The predictions from our orthology-based annotation approach were next validated using untargeted spatial transcriptomics. These results allowed us to predict evolutionarily conserved tissue-specific markers and generate cell type-specific gene regulatory networks for root tissues of wheat and the other species used in our analysis. In summary, we generated a single-cell and spatial transcriptomics resource for wheat root apical meristems, including numerous known and uncharacterized cell type-specific marker genes and developmental regulators. These data and analyses will facilitate future cell type annotation in non-model plant species.

Keywords: CP: Plants; cluster annotation; gene regulatory networks; root meristem atlas; single-cell transcriptomics; untargeted spatial transcriptomics; wheat.

Graphical abstract

Figure 1

Single-cell RNA-seq and cluster annotation of wheat root tips (A) UMAP visualization of the three replicates in our scRNA-seq experiment and corresponding atlas metrics. (B) Expression of cell type markers across each cluster. Dot diameter, proportion of cluster cells in a cluster expressing a given gene; color, mean expression across cells in that cluster. (C) Sankey plot showing annotations transferred from Arabidopsis (Ath), rice (Osa), maize (Zma), and single-nuclei wheat (snTae) to our wheat atlas (Tae) and corresponding q value. (D and E) Annotated UMAPs with cell type (D) and cell state (E) annotations. Please note that cluster 6 was manually annotated as pericycle based on evidence from STOmics Stereo-seq data and known pericycle marker genes and was therefore marked with an asterisk.

Figure 2

scRNA-seq-derived marker gene expression patterns in STOmics Stereo-seq root sections (A) A cross-section of wheat root apical meristem with major cell types annotated. (B–F) UMAP feature plot and STOmics Stereo-seq visualization of marker genes from epidermis (B), cortex (C), phloem (D), xylem (E), and root cap (F).

Figure 3

Reprocessed scRNA-seq datasets (A) Reprocessed and transferred annotation of the Arabidopsis scRNA-seq dataset. (B) The wheat scRNA-seq dataset generated in this study. (C) Reprocessed and transferred annotation of the maize scRNA-seq dataset. (D) Reprocessed and transferred annotation of the rice scRNA-seq dataset.

Figure 4

Tissue-specific markers conserved across Arabidopsis, wheat, rice, and maize or unique to the monocot clade (A) UpSet plot showing the intersections of orthologous groups of xylem markers across Arabidopsis, wheat, rice, and maize. (B–E) Feature plots of a xylem-specific marker across species. (F and G) Spatial expression in STOmics Stereo-seq data (F) and ternary plot showing genome asymmetry information (G) of the same xylem-specific marker in the wheat root meristem. (H) UpSet plot showing the intersections of orthologous groups of cortex markers across Arabidopsis, wheat, rice, and maize. (I–L) Feature plots of a cortex-specific marker unique to monocots. (M and N) Spatial expression in STOmics Stereo-seq data (M) and ternary plot showing genome asymmetry information (N) of the same cortex-specific marker in the wheat root meristem.

Figure 5

Cross-species overlap of predicted tissue-specific regulators (A) UpSet plot showing the intersections of predicted regulators across Arabidopsis, wheat, rice, and maize. (B) Cytoscape visualization of the GRN around wheat VND4. Dot size is proportional to the number of wheat genes in each orthologous group.