Functional characterization of the Arabidopsis transcription factor bZIP29 reveals its role in leaf and root development

Abstract

Plant bZIP group I transcription factors have been reported mainly for their role during vascular development and osmosensory responses. Interestingly, bZIP29 has been identified in a cell cycle interactome, indicating additional functions of bZIP29 in plant development. Here, bZIP29 was functionally characterized to study its role during plant development. It is not present in vascular tissue but is specifically expressed in proliferative tissues. Genome-wide mapping of bZIP29 target genes confirmed its role in stress and osmosensory responses, but also identified specific binding to several core cell cycle genes and to genes involved in cell wall organization. bZIP29 protein complex analyses validated interaction with other bZIP group I members and provided insight into regulatory mechanisms acting on bZIP dimers. In agreement with bZIP29 expression in proliferative tissues and with its binding to promoters of cell cycle regulators, dominant-negative repression of bZIP29 altered the cell number in leaves and in the root meristem. A transcriptome analysis on the root meristem, however, indicated that bZIP29 might regulate cell number through control of cell wall organization. Finally, ectopic dominant-negative repression of bZIP29 and redundant factors led to a seedling-lethal phenotype, pointing to essential roles for bZIP group I factors early in plant development.

Keywords: Arabidopsis thaliana; bZIP group I transcription factors; cell proliferation; cell wall; chromatin immunoprecipitation; leaf cell number; plant development; root cell number; tandem affinity purification..

Fig. 1.

bZIP29 expression analysis. (A–I) Histochemical GUS staining of a Promoter_bZIP29:GFP-GUS reporter line. In roots, expression was detected in the primary root tip (A) with maximum expression in the QC and the columella cells (B), in lateral root primordia (C) with maximum expression at the boundaries (D and E), and later during lateral root development in the lateral root tip and at the transition from the proliferation to expansion zone (F). In leaves, strong expression was detected in fully developed guard cells (G), and in the progenitor meristemoid (G, inset). This stomatal expression pattern was confirmed in the stomata present in the hypocotyl (H) and in the anthers (I). (This figure is available in color at JXB online.)

Fig. 2.

Genome-wide identification of genes bound by bZIP29 through TChAP-seq analysis. (A) Genome-wide distribution of bZIP29 TChAP-seq peaks present in both replicates, in relation to the gene structure. (B) Distance of the summit of the peaks from both replicates, relative to the translation start site (TSS) of the nearest gene. (C) De novo motif analysis of the TChAP-seq data, identifying significant over-representation of similar motifs in both replicates. Through database comparison, similarity to motifs bound by bZIP30 and bZIP18 was found. (D) Sequence position location (bp) of the identified motifs in relation to the peak summit. (E) Logo representing motif 3 identified in the first replicate. (F) Integrative Genomics Viewer (IGV) screenshot (Thorvaldsdóttir et al., 2013) of sequenced reads from bZIP29, mock, ERF115 and PPD2 TChAP-seq experiments, mapped to the mitotic cell cycle regulator cyclin B1;2, the cell cycle inhibitor SMR4, and the cell wall organizing genes XTH9, EXP-like A1 and EXP-like A3. The gene regions are represented by bars. (This figure is available in color at JXB online.)

Fig. 3.

Dominant-negative repression of bZIP29 and phenotypical analyses of mutant lines. (A) Transactivation activity in tobacco protoplasts transfected with a UAS:fLUC reporter construct, or GAL4DBD effector constructs fused to bZIP29 or to bZIP29-SRDX. Mean values are compared with the GAL4 control (n=8; ***P<0.001, Student’s t-test). Fusion to the SRDX motif abolishes the transcriptional capacity of the bZIP29 TF. (B) Seedling-lethal phenotype of the P35S:bZIP29-SRDX dominant-negative repressor line at 7 DAS, compared with the wild-type phenotype. (C) Phenotype in soil (28 DAS) and leaf series of Promoter_bZIP29:bZIP29-SRDX line 1 and the out-segregated wild-type line 1, at 21 DAS, grown in vitro. (D) Area of individual leaves of the two independent Promoter_bZIP29:bZIP29-SRDX lines, compared with the corresponding out-segregated wild-type lines, and of the overexpressor line (P35S:bZIP29), grown in vitro, at 21 DAS (n=3). Images show phenotype at 21 DAS, grown in vitro. (This figure is available in color at JXB online.)

Fig. 4.

Leaf cellular analysis of bZIP29 repressor and overexpressor lines. (A) Leaf area, cell area, and cell number of leaf 1 and 2 obtained from two independent Promoter_bZIP29:bZIP29-SRDX repressor lines or from the P35S:bZIP29 overexpressor line, grown in vitro, 21 DAS (n=5). Normalization was done relative to out-segregated wild-type plants. *P<0.05; **P<0.01; ***P<0.001 (Student’s t-test). (B–D) Scanning electron microscopy pictures of the abaxial epidermis of leaf 1 or 2 (21 DAS) from wild-type (B), Promoter_bZIP29:bZIP29-SRDX line 1 (C), or Promoter_bZIP29:bZIP29-SRDX line 2 (D). Scale bars: 100 μm. The insets show stomata of irregular (C) or normal (B and D) shape. (This figure is available in color at JXB online.)

Fig. 5.

Phenotypical analysis of roots from Promoter_bZIP29:bZIP29-SRDX repressor lines. (A) Agravitropic root growth phenotype of two independent Promoter_bZIP29:bZIP29-SRDX repressor lines, compared with the wavy root growth phenotype of out-segregated wild-type plants, grown in vitro (7 DAS) at an inclination of approximately 45°. (B) Root gravitropic response of Promoter_bZIP29:bZIP29-SRDX repressor and wild-type lines following gravistimulation (as depicted in the scheme), determined by a root tip bending assay. (C) Confocal microscopy picture of the Promoter_bZIP29:GFP-GUS reporter line showing expression in the QC and columella cells. Roots were counterstained with propidium iodide. Scale bar: 50 µm. (D) Quantification of the root meristem size (µm) of Promoter_bZIP29:bZIP29-SRDX repressor and out-segregated wild-type lines. ***P<0.001 (Student’s t-test). Representative pictures of root tips from repressor and wild-type lines, counterstained with propidium iodide, are shown below the graph. Length of meristem is shown by yellow lines. Scale bars: 50 µm. (This figure is available in color at JXB online.)

Fig. 6.

Transcriptome analysis of root tips from the Promoter_bZIP29:bZIP29-SRDX repressor line 1 and comparison with the TChAP-seq target list. (A) Venn diagram showing a significant overlap (**P=3.87×10^–5) between genes bound by bZIP29 identified by TChAP-seq in cell culture, and genes down-regulated in the root tips of the Promoter_bZIP29:bZIP29-SRDX line 1 as identified by RNA-seq. (B) Summary of 11 direct target genes bound by bZIP29 and down-regulated upon repression of bZIP29 in root tips. (C) bZIP29- and/or VIP1-dependent activation of the XTH9 promoter in a transient expression assay. Different combinations of bZIP29 and VIP1 were tested. Equal total amount of effector plasmids was used. Indicated values represent average relative luciferase activities compared with a P35S:GUS control for eight biological repeats. ***P<0.001; *P<0.05 (Student’s t-test). (D) Relative expression of BALDIBIS (AT3G45260) in seedlings and in root tips from the Promoter_bZIP29:bZIP29-SRDX repressor lines and the out-segregated wild-types, as determined by qRT-PCR. Expression was normalized using Actine2_2 and AteEF1A-4 as reference genes. (This figure is available in color at JXB online.)

Fig. 7.

Rooted maximum likelihood phylogenetic tree encompassing all Arabidopsis bZIP group I and group E members, and their bZIP interactors as identified by TAP. In addition, four unclassified bZIP factors from Arabidopsis are included, forming one subfamily (HOM03D000835) (dashed line) with the two group E members, according to the PLAZA 3.0 Dicots comparative genomics platform. PLAZA also clustered the unclassified AtbZIP72 (AT5G07160) as part of the same subfamily (HOM03D000341) as the group I members. AtbZIP73 (AT2G13130) was excluded from the analysis, since it encodes a pseudogene, while UNE4 (AT2G12940) was added to group I because it contains the conserved lysine residue and is part of the HOM03D000341 subfamily. Proteins identified by TAP with bZIP29 are shown in bold. Group I orthologs from other plant species are shown in italics. The group A member ABF4/AtbZIP38 was used as an out-group. Protein sequences from Arabidopsis were based on the TAIR10 annotation (Lamesch et al., 2012) and orthologous protein sequences were obtained from UNIPROT (Dimmer et al., 2012). The PhyML tree was constructed using the ‘one click’ mode of the Phylogeny.fr software tool (Dereeper et al., 2008). The branch length is proportional to the number of substitutions per site and branches are annotated with their support values. (This figure is available in color at JXB online.)

Fig. 8.

Summarizing model on bZIP29 function and regulation. bZIP29 is mainly expressed in the QC, columella cells, lateral root primordia (LRP), meristemoids, and stomata. Upstream, bZIP29 is transcriptionally regulated by auxin (+) or ABA (–). Post-translational regulation of bZIP29 activity might be obtained through activating or inhibiting phosphorylation by mitotic CDKB–cyclin B kinase complexes or MPK3, and counteracted by PP2A-4-containing phosphatase complexes. Phosphorylation could regulate its subcellular localization through interaction with 14-3-3 GRFs, or its dimerization preference with other bZIP TFs like VIP1. Homo- or heterodimers of bZIP29 regulate expression of osmosensory and stress genes, and genes involved in cell cycle regulation or cell wall organization, through binding to the core motif mCAGCTGk or other cis-regulatory elements. Repression of bZIP29 and redundant factors alters leaf and root development. (This figure is available in color at JXB online.)