A cellular expression map of epidermal and subepidermal cell layer-enriched transcription factor genes integrated with the regulatory network in Arabidopsis shoot apical meristem

Abstract

Transcriptional control of gene expression is an exquisitely regulated process in both animals and plants. Transcription factors (TFs) and the regulatory networks that drive the expression of TF genes in epidermal and subepidermal cell layers in Arabidopsis are unexplored. Here, we identified 65 TF genes enriched in the epidermal and subepidermal cell layers of the shoot apical meristem (SAM). To determine the cell type specificity in different stages of Arabidopsis development, we made YFP based transcriptional fusion constructs by taking a 3-kb upstream noncoding region above the translation start site. Here, we report that for ~52% (22/42) TF genes, we detected transcription activity. TF genes derived from epidermis show uniform expression in early embryo development; however, in the late globular stage, their transcription activity is suppressed in the inner cell layers. Expression patterns linked to subepidermal cell layer identity were apparent in the postembryonic development. Potential upstream regulators that could modulate the activity of epidermal and subepidermal cell layer-enriched TF genes were identified using enhanced yeast-one-hybrid (eY1H) assay and validated. This study describes the activation of TF genes in epidermal and subepidermal cell layers in embryonic and postembryonic development of Arabidopsis shoot apex.

Keywords: Cell layer; epidermal; gene regulatory network; shoot apical meristem; subepidermal; transcription factors.

FIGURE 1

Spatiotemporal expression pattern revealed for epidermal cell layer enriched TF genes in embryonic and postembryonic development in Arabidopsis. From left to right the picture represents globular, heart stage, 3 DAG seedlings and inflorescence meristems (IMs), respectively. The Arabidopsis gene identifier and gene name are written for each promoter reporter construct on the left. Of the 17 TF promoter‐reporters studied in IM (d, h, l, p, t, x, ab, af, aj, an, ar, av, az, bd, bh, bl, bp ), ten are active in globular stage embryo (a, e, m, q, u, y, ac, ao, as and aw) followed with 14 in heart stage (b, f, n, r, v, z, ad, ah, al, ap, at, ax, bb and bj), 16 in 3 DAG seedlings (c, g, k, o, s, w, aa, ae, ai, am, aq, au, ay, bc, bg and bk. Scale bars = 20 μM

FIGURE 2

Repression of epidermal cell layer enriched TF genes occur in the inner layers in late globular stage. In the four‐cell and eight‐cell stage of embryos,PDF2 expression is not detected (a and b). PDF2 expresses uniformly in early globular stage embryo (c), in globular stage its expression is visible within the protoderm (d). In contrast, ATML1 expression is seen from 4‐cell stage onward, and its expression get restricted to outer cell layer in the late globular stage (e–h). HDG12 and NF‐YA5 expression also show a similar expression pattern (i–l and m–p). HDG12 expression at four‐cell (i), eight‐cell (j), 16‐cell (k) and early globular stage (l). (m‐p) NF‐YA5 expression at four‐cell (m), eight‐cell (n), 16‐cell (o), and early globular stage (p). Scale bars = 10 μM

FIGURE 3

Subepidermal cell layer enriched TF genes display dynamic expression pattern. For each reporter line from left to right images represent embryonic and postembryonic stages of plant development. Mainly globular and heart stages of embryo development were investigated, whereas, for postembryonic development 3 DAG seedling and four‐week‐old IM were chosen (a–t). Interestingly, HDG4 and HDG7 show subepidermal cell layer specific expression in IM (d and h), in 3 DAG seedling expression of HDG4 expression was sporadic and spotted in subepidermal cell (c), however, in HDG7 it was variable and expressed both in epidermal and subepidermal cell layer (g). Scale bars = 20 μM

FIGURE 4

GRN of epidermal and subepidermal cell layer enriched TFs. Hierarchical clustering of 49 TF genes that are differentially expressed in SAM and used for making baits (a). Microarray data is derived from WUS, HDG4, and ATHMGB15 cell population of the shoot to plot the heatmap. Enrichment degree for bait and prey genes selected for eY1H was calculated (b). Experimental outline followed in conducting the eY1H assay is depicted (c), 49 TF promoters were used to make bait in YM4271 while 327 prey plasmids were transformed into Ya1867 (c). Bait and prey were mated, and diploids were selected on minimal media lacking histidine and tryptophan. A network of interactions deduced using eY1H is represented in the cytoscape (d)

FIGURE 5

Validation of protein‐DNA interaction to construct subnetworks. Protein‐DNA interaction network tested for DEWAX in dewax mutant and overexpression lines (OX) (a). Colored bars indicate inferred relationship (e.g., red, activator; green, repressor and grey, not consistent). In (b) athb34, athmgb15, grf123 mutant lines, and 35S::GRF3 OX used to test the target TFs. In (c) arf9, arf12, azf2, gata8, annac082 mutant lines, and WRKY54, AT2G28810 OX lines were used to infer the regulatory relationship with the targets. A summary of tested networks using RT‐qPCR assay (d). Red arrows indicate activating, and green lines indicate repressing relationships among the tested TFs. Dotted grey lines indicate non‐linear relationships

FIGURE 6

Schematic drawings of expression patterns captured for epidermal and subepidermal cell layers enriched TF genes. TFs expressed in globular stage (a), late heart stage (b), 3‐DAG seedlings (c), and 4‐week old IM (d). The Arabidopsis gene identifier and gene name are written below the representative expression pattern drawing