An Arabidopsis gene regulatory network for secondary cell wall synthesis

Abstract

The plant cell wall is an important factor for determining cell shape, function and response to the environment. Secondary cell walls, such as those found in xylem, are composed of cellulose, hemicelluloses and lignin and account for the bulk of plant biomass. The coordination between transcriptional regulation of synthesis for each polymer is complex and vital to cell function. A regulatory hierarchy of developmental switches has been proposed, although the full complement of regulators remains unknown. Here we present a protein-DNA network between Arabidopsis thaliana transcription factors and secondary cell wall metabolic genes with gene expression regulated by a series of feed-forward loops. This model allowed us to develop and validate new hypotheses about secondary wall gene regulation under abiotic stress. Distinct stresses are able to perturb targeted genes to potentially promote functional adaptation. These interactions will serve as a foundation for understanding the regulation of a complex, integral plant component.

Extended Data Figure 1. Number of novel and previously described protein-DNA interactions and transcription factors involved in secondary cell wall biosynthesis and xylem development

Venn diagrams of overlap between previously reported (A) interactions or (B) transcription factors and those of the xylem-specific gene regulatory network. *=includes genes that were not included in yeast one hybrid screen.

Extended Data Figure 2. Activation or repression of VND7 by E2Fc is dynamic and dose-dependent

(A) Intensity of LUC bioluminescence quantified using Andor Solis image analysis software. Data are means ± s.d. (n=20). Asterisks denote significance at p<0.05 determined by Student’s t-test. (B) Quantitative real-time PCR of E2Fc and VND7 transcripts in ΔN-E2Fc (E2Fc overexpressor line lacking the N-terminal domain) expressing plants versus Col-0 control. Red dashed line marks the point at which VND7 is unchanged compared to control. Each data point is an individual biological replicate with 3 technical replicates. (C) 3-week old tobacco leaves were infiltrated with the p19 silencing inhibitor and either the reporter VND7::GUS or VND7::GUS and either 35S::E2Fc:MYC or 35S::RBR:GFP, or both. Extracted protein was then used in a quantitative MUG fluorescent assay, where relative fluorescence was measured 60 min after incubation with substrate. Data are means ± s.d., n=3.

Extended Data Figure 3. Binding of NST2 and SND1 to fragments of CESA7, CESA8, and KOR promoters

Electrophoretic mobility shift assays showing NST2 (A–D) and SND1 (E–F) protein specifically binds the promoters of cellulose-associated genes. Probe was incubated in the absence or presence of GST or GST:SND1 protein extracts. The arrowheads indicate the specific protein-DNA complexes, while arrows indicate free probe.

Extended Data Figure 4. Sub-networks of network genes differentially expressed in response to iron deprivation of high salinity

Sub-network of genes with q-values of ≤0.01 and whose fold change between mean expression values was ≥1.5 in either direction in iron deprivation (A) or high NaCl (B) stress microarray dataset. Nodes are colored according in in-degree as shown on scale bars below sub-networks. Transcription factors with the highest in-degree are labeled and indicated with a black circle.

Extended Data Figure 5. The reconstructed gene regulatory consensus network based on analysis of the iron-deprivation expression dataset by different network inference methods

(A) Unsupervised, (B) supervised in the first pass, (C) Supervised after the validated two connections have been added in the training set. Edge transparency denote p ≤ 0.06 for the Pearson Correlation Coefficient (PCC); edge width is proportional to PCC; edge value correspond to the total edge score; a greater value corresponds to more significant score. Yellow and red nodes correspond to transcription factor and target gene nodes, respectively; black and blue edges denote Y1H-derived and inferred interactions, respectively.

Extended Data Figure 6. Iron deprivation and NaCl stress influences lignin and phenylpropanoid biosynthesis associated gene expression

(A) No change was observed in the expression of 4CL1::GFP in 4 DAI roots transferred to a control media (left, n=4) or media with 140 mM NaCl for 48 hours (right, n=4). (B) Increased fuchsin staining of xylem cells as well as of cell walls of non-vascular cells in 4 DAI roots transferred to a control media (left) or media with an iron chelator for 72 h (right). (C) No change was observed in the expression of VND7::YFP in 4 DAI roots transferred to a control media (left, n=4) or media with an iron chelator for 72 h (right, n=5).

Extended Data Figure 7. Schematic diagram of dual-luciferase reporter vector development

(A) Three distinct donor vectors harboring either the transcription factor, VP64 activation domain fused to the 35S minimal promoter, or a promoter fragment. (B) The dual reporter vector, pLAH-LARm, is then recombined with the three donor vectors to generate the (C) single reporter vector.

Figure 1. Regulators of xylem development and secondary cell wall biosynthesis

(A) Gene regulatory network for secondary cell wall biosynthesis in Arabidopsis root xylem. Nodes-transcription factors or promoters, edges-protein-DNA interactions. Edges in feed-forward loops are red. (B) A sample feed-forward loop in red. (C) ‘Power edges’ between node sets. (D) The secondary wall network from sub-fragments of cell wall promoters.

Figure 2. E2Fc represses secondary cell wall gene biosynthesis

(A) E2Fc-DNA interactions. Solid edges=Y1H, dashed edges=literature. (B) Bright field (top) and dark-field (bottom) of representative leaves (n=20) expressing VND7::LUC or together with 35S::E2Fc in 1:0.1, 1:1, 1:2, 1:5, and 1:10 ratios respectively. C) VND6 and VND7 expression relative to UBC10 control in an E2Fc RNAi line relative to wild-type. n= 2 biological replicates with 3 technical replicates. (D) Phloroglucinol staining of lignin (n=6xgenotype, representative images shown) and (E) crystalline cellulose in wild-type and E2Fc-knockdown roots (n=3×1000xgenotype). For all panels, *p<0.05 from Student’s t-test and data are means ± s.d.

Figure 3. Tissue-specific VND7 regulation and VND7 targets

(A) REV and PHB expression relative to β-tubulin control following dexamethasone treatment of 35S::VND7:VP16:GR relative to untreated. n=4, a,b,c = p<0.01). (B) PAL4 expression relative to AT5G15710 control in rev-5 relative to wild-type. (C) PAL4 expression relative to UBC21 control following one hour dexamothasone treatment of 35S:REV:GR relative to untreated. *p<0.05 for panels B and C, n= 2 biological replicates with 3 technical replicates. All panels show data as means ± s.d, with p calculated from Student’s t-test.

Figure 4. Multiple transcription factors bind the CESA4 promoter

(A) Activation of CESA4::LUC by transcription factors in tobacco (n=5). *p<0.05 based on Student’s t-test. Data are means ± s.d. (B–C) EMSA with NST2 (B) and SND1 (C) with promoters. Arrowheads indicate protein-DNA complexes, arrows indicate free probe. (D) ChIP of NST2:GFP with CESA4, CESA7, CESA8, and KOR promoters.

Figure 5. The xylem-specific gene regulatory network is responsive to high salinity and iron deprivation

(A) Network genes responsive to high salinity and/or iron deprivation. (B) VND7, HCT, 4CL1, PAL4 expression after iron deprivation. (C) 4CL1::GFP expression after iron deprivation (representative images shown, n=4xline). (D) Lignin gene expression after iron deprivation in rev-5. G-genotype, F-Fe stress; p-values from ANOVA. (E) VND7, HCT, 4CL1, PAL4 expression after NaCl. (B,D,E) Expression relative to UBC10 and PP2AA3 controls. n= 2 biological replicates with 3 technical replicates. *p≤0.01based on Student’s t-test and data are means ± s.d. (F) Representative images of VND7::YFP (n=5) and (G) fuchsin-staining (n=5) after NaCl. Arrows- (F) non-stele cells and (G) extra metaxylem strand. (H) Proposed regulation of secondary wall biosynthesis.