Hierarchical Transcription Factor and Chromatin Binding Network for Wood Formation in Black Cottonwood (Populus trichocarpa)

Abstract

Wood remains the world's most abundant and renewable resource for timber and pulp and is an alternative to fossil fuels. Understanding the molecular regulation of wood formation can advance the engineering of wood for more efficient material and energy productions. We integrated a black cottonwood (Populus trichocarpa) wood-forming cell system with quantitative transcriptomics and chromatin binding assays to construct a transcriptional regulatory network (TRN) directed by a key transcription factor (TF), PtrSND1-B1 (secondary wall-associated NAC-domain protein). The network consists of four layers of TF-target gene interactions with quantitative regulatory effects, describing the specificity of how the regulation is transduced through these interactions to activate cell wall genes (effector genes) for wood formation. PtrSND1-B1 directs 57 TF-DNA interactions through 17 TFs transregulating 27 effector genes. Of the 57 interactions, 55 are novel. We tested 42 of these 57 interactions in 30 genotypes of transgenic P. trichocarpa and verified that ∼90% of the tested interactions function in vivo. The TRN reveals common transregulatory targets for distinct TFs, leading to the discovery of nine TF protein complexes (dimers and trimers) implicated in regulating the biosynthesis of specific types of lignin. Our work suggests that wood formation may involve regulatory homeostasis determined by combinations of TF-DNA and TF-TF (protein-protein) regulations.

Figure 1.

Xylem Fiber and Vessel Cell–Specific Expression, Subcellular Localization, and Transcriptional Activity of PtrMYB021 and PtrMYB074. (A) to (D) Xylem cross section before (A) and after (B) LCM isolation of vessel cells. Xylem cross section before (C) and after (D) LCM isolation of fiber cells. Red circles in (A) and (B) indicate the dissected vessel cells, and green rectangles in (C) and (D) indicate the dissected fiber cells. Scale bar = 50 µm in (A) to (D). (E) to (G) Quantitative measurement of transcript abundances of PtrSND1-B1 (E), PtrMYB021 (F), and PtrMYB074 (G) in SDX cells isolated by LCM. The label 3 Cell types denote fiber, vessel, and ray cells, the three major cell types in SDX. Relative transcript abundance and SE are calculated from three biological replicates. (H) to (M) Subcellular location of GFP fused PtrMYB021 (H-J) and PtrMYB074 proteins (in [K] to [M]). The MYB021-GFP signal (H), the MYB074-GFP signal (K), and the H2A-mCherry signals (I) and (L), and merged images (J) and (M) are shown. (N) to (P) Transcript abundance (normalized count per million [CPM] reads) of PtrMYB021 (N), PtrMYB074 (O), and PtrSND1-B1 (P) in five P. trichocarpa tissues: xylem, phloem, leaf, shoot, and root. Error bars represent 1 se of three biological replicates.

Figure 2.

Methodology Flowchart. Flowchart depicts a pipeline of how the PtrSND1-B1 four-layered network was constructed and validated using experimental approaches. The six-step experimental procedures were used for generating and validating the regulatory relationships between TFs and their target genes. These relationships were used for generating the TRN in wood formation.

Figure 3.

Identification of Cell Wall Biosynthetic Genes Directly Targeted by PtrMYB021 and PtrMYB074 by ChIP Assays. (A) Approximate location of the promoter fragments (I to IV) amplified by PCR following ChIP assays. Rectangle shows the gene, and the line represents a gene promoter. The arrowheads show the approximate location of the promoter region that was assigned for designing primers for PCR amplification. The precise locations of the promoter sequences tested in ChIP-PCR are shown in Supplemental Figure 4. (B) and (C) ChIP-PCR assays of cell wall biosynthetic genes regulated by PtrMYB021 (B) and PtrMYB074 (C). Four promoter fragments of each target gene were amplified by PCR following ChIP assays. Input, mock, and anti-GFP represent PCR reactions using the chromatin preparations before immunoprecipitation, immunoprecipitated with preimmune serum and immunoprecipitated with anti-GFP antibody, respectively. Four independent biological replicates for each ChIP assay were performed, and the results of one biological replicate are presented. A TF–DNA interaction is considered true when at least three of the four biological replicates are positive. The direct targets of PtrMYB021 and PtrMYB074 are framed in (B) and (C), and the indirect targets are not. The promoter fragments that can be bound by PtrMYB021 or PtrMYB074 proteins are marked with an asterisk (*). The promoter fragments of PtrACTIN were assayed as a negative control. (D) Diagram depicting the regulation of PtrMYB021 and PtrMYB074 for the cell wall (CW) biosynthetic genes. Arrows indicate the protein–DNA regulatory interaction with activation ability.

Figure 4.

ChIP Identification of TF Genes Directly Targeted by PtrMYB021 and PtrMYB074. (A) Approximate locations of the promoter sequences amplified by PCR following the ChIP assays, as described in Figure 3. The precise locations of the promoter sequences tested in ChIP-PCR are shown in Supplemental Figure 5. (B) and (C) ChIP-PCR assays of TF genes from PtrMYB021-regulated genes (B) and PtrMYB074-regulated genes (C). Four promoter fragments of each target gene were amplified by PCR following ChIP assays. Input, mock, and anti-GFP represent PCR reactions using the chromatin preparations before immunoprecipitation, immunoprecipitated with preimmune serum, and immunoprecipitated with anti-GFP antibody, respectively. Four independent biological replicates for each ChIP assay were performed, and the results of one biological replicate are presented. A TF–DNA interaction is considered true when at least three of the four biological replicates are positive. The direct targets of PtrMYB021 and PtrMYB074 are framed in (B) and (C), and the indirect targets are not. The promoter fragments that can be bound by PtrMYB021 or PtrMYB074 proteins are marked with an asterisk (*). For these experiments, the promoter fragments of PtrACTIN were assayed as a negative control. (D) Diagram showing that PtrMYB021 and PtrMYB074 directly regulate the TF genes. The arrows indicate the protein–DNA regulatory interaction with activation ability.

Figure 5.

Expression Profiles of 36 Cell Wall Biosynthetic Genes in SDX Protoplasts Overexpressing Each of 15 TFs Directly Regulated by PtrMYB021 and PtrMYB074. Heatmap represents the relative expression (fold change) of the genes in SDX protoplasts overexpressing each TF compared with the control. The average expression of three biological replicates is presented. Fold change and se for RT-qPCR are shown in Supplemental Data Set 8. The overexpressed TFs are shown on left side, and the cell wall biosynthetic genes are shown in upper panel. Twenty-five (highlighted in green) of the 36 cell wall genes were regulated by nine TFs (highlighted in yellow). Color bar at the bottom depicts the relative expression levels, where blue, white, and red represent downregulation, no change, and upregulation, respectively. Asterisks represent genes regulated by PtrMYB021 (*), PtrMYB074 (**), or both PtrMYB021 and PtrMYB074 (***).

Figure 6.

Characterization of TF–DNA Regulatory Interactions between the Third-Layered TFs and Their Regulated Cell Wall Biosynthetic Genes. (A) Locations of the promoter sequences amplified by PCR following the ChIP assays, as described in Figure 3. The accurate locations of the promoter sequences tested in ChIP-PCR are shown in Supplemental Figure 6. (B) to (G) ChIP assays showed that the direct regulation of PtrMYB090 (B), PtrMYB161 (C), PtrMYB174 (D), PtrWBLH1 (E), PtrWBLH2 (F), and PtrNAC123 (G) regulate 15 cell wall genes based on TF–promoter interactions. Four promoter fragments of each target gene were amplified by PCR following ChIP assays. Input, mock, and anti-GFP are PCR reactions using the chromatin preparations before immunoprecipitation, immunoprecipitated with preimmune serum and immunoprecipitated with anti-GFP antibody, respectively. Four independent biological replicates for each ChIP assay were performed, and the results of one biological replicate are presented. A TF–DNA interaction is considered true when at least three of the four biological replicates are positive. The negative results and control for the ChIP-PCR assays of PtrMYB090, PtrMYB161, PtrMYB174, PtrWBLH1, PtrWBLH2, and PtrNAC123 are shown in Supplemental Figure 7. (H) The network depicts how six TFs directly regulate 15 cell wall biosynthetic genes. The genes are categorized into four functional groups shown by different colors: red for TFs, ivory for secondary cell wall (SCW) synthases, gray for primary cell wall (PCW) synthases, and yellow for monolignol biosynthesis enzymes. Arrows indicate the positive protein–DNA regulatory interactions, and the blunt line indicates a negative protein–DNA regulatory interaction.

Figure 7.

Four-Layered PtrSND1-B1–Mediated TRN Depicting Transcriptional Regulation of Cell Wall Biosynthesis in P. trichocarpa SDX. The 18 TF and 27 cell wall (CW) biosynthetic genes (gene IDs and their annotations are shown in Supplemental Data Set 9) are connected by direct TF–DNA interactions (black lines) in this TRN. Ptr-SND1-B1 is at the top (first layer) of this TRN. The second layer of the TRN consists of PtrMYB021 and PtrMYB074 based on Lin et al. (2013). Twenty-seven direct targets of PtrMYB021 and PtrMYB074 are presented in the third layer, according to ChIP assays coupled with RNA-seq. The 27 direct targets include 15 cell wall biosynthetic genes and 12 TF genes. PtrMYB021 and PtrMYB074 share eight common direct targets, including three cell wall biosynthetic genes (PtrLAC21, PtrLAC26, and PtrPec9-1) and five TF genes (PtrMYB059, PtrMYB090, PtrMYB161, PtrNAC123, and PtrWBLH2). In the forth layer, 15 cell wall biosynthetic genes are directly regulated by six TFs (PtrMYB090, PtrMYB161, PtrMYB174, PtrWBLH1, PtrWBLH2, and PtrNAC123). Black lines indicate regulatory protein–DNA interactions. The number on each black line represents fold change of the downstream targets induced by the upstream TF overexpression. Arrows indicate positive regulation (activation), and the blunted line indicates negative regulation (repression). Green lines and curved lines represent TF-TF (protein–protein) interactions. PCW, primary cell wall; SCW, secondary cell wall.

Figure 8.

Validation of Direct Regulation of PtrMYB021, PtrMYB074, and PtrMYB090 in Stable Transgenic P. trichocarpa. The direct target genes of PtrMYB021, PtrMYB074, and PtrMYB090 derived from ChIP assays (shown in Figure 7) were verified by RT-qPCR for their downregulated expression in SDX of stable transgenic P. trichocarpa by knocking down PtrMYB021, PtrMYB074, or PtrMYB090. (A) and (B) Transcript abundances of ChIP-PCR–verified cell wall biosynthetic genes (A) and TF genes (B) in PtrMYB021 transgenic lines and the wild type (control). (C) and (D) Transcript abundances of ChIP-PCR verified cell wall biosynthetic genes (C), and ChIP-PCR verified TF genes (D) in wild-type (control) and PtrMYB074 transgenic lines. (E) Transcript abundances of ChIP-PCR–verified cell wall biosynthetic genes in PtrMYB090 transgenics. For each target gene, the average transcript abundance of three biological replicates of the wild-type (control) plants was set as 1. Error bars represent 1 se of three biological replicates. Asterisks represent significant differences from the wild-type trees, determined using Student’s t test (*P < 0.05; **P < 0.01). The expression of PtrMYB059 and PtrMYB090 in PtrMYB021 transgenics, and PtrMYB059 and PtrMYB088 in PtrMYB074 transgenics, showed no significant change in gene expression compared with the wild-type trees.

Figure 9.

Y2H and BiFC Demonstrate That TF Pairs Binding to the Same Promoter Fragments May Interact with Each Other. TF pairs (PtrMYB021 and PtrMYB074, PtrMYB090 and PtrNAC123, PtrMYB090 and PtrMYB161, PtrMYB161 and PtrWBLH1, PtrMYB090 and PtrWBLH1, PtrMYB161 and PtrWBLH2, and PtrMYB090 and PtrWBLH2) bind to the same promoter fragments, as indicated by ChIP-PCR in Figures 3, 4, and 6. (A) to (G) Y2H demonstrated that these TF pairs can physically interact with each other. Each bait and prey pair were coexpressed in yeast cells and selected on the SD/–Leu/–Trp (–LW) medium: (A) AD-PtrMYB021 and BD-PtrMYB074, (B) BD-PtrMYB090 and AD-PtrNAC123, (C) BD-PtrMYB090 and AD-PtrMYB161, (D) BD-PtrMYB161 and AD-PtrWBLH1, (E) BD-PtrMYB090 and AD-PtrWBLH1, (F) BD-PtrMYB161 and AD-PtrWBLH2, and (G) BD-PtrMYB090 and AD-PtrWBLH2). (A1) to (G1) Protein–protein interaction of each pair was validated by the growth of the transformants on SD/–Leu/–Trp/–His/–X-Gal (–LWHX) medium. (A2) to (G2) BiFC demonstrated that ChIP-validated pairs of TFs heterodimerizes with each other. Each pair of TF proteins, PtrMYB021-YFP^N and PtrMYB074-YFP^C (A2), PtrMYB090-YFP^C and PtrNAC123-YFP^N (B2), PtrMYB090-YFP^C and PtrMYB161-YFP^N (C2), PtrMYB161-YFP^C and PtrWBLH1-YFP^N (D2), PtrMYB090-YFP^C and PtrWBLH1-YFP^N (E2), PtrMYB161-YFP^C and PtrWBLH2-YFP^N (F2), and PtrMYB090-YFP^C and PtrWBLH2-YFP^N (G2), coexpressed into SDX protoplasts gave a positive BiFC signal for heterodimerization, which are colocalized with H2A-1-mCherry signal in the nucleus. As negative controls, transformation of single TF-YFP^N with YFP^C alone ([A3] to [G3]), with a noninteracting TF (SND1-B1-YFP^C) ([A4] to [G4]), or TFs in other complexes ([A5] to [H2]), did not give any YFP signals. (H1) and (I1) Negative control was indicated by BD-53/AD-Lam (H1), whereas positive growth was confirmed using the positive control BD-53/AD-T (I1).

Figure 10.

Three-Layered Network Encompassed by AtSND1 and AtMYB46 Regulating Cell Wall Biosynthetic Genes in Arabidopsis Based on studies using inducible assays, electrophoretic mobility shift assay, Y1H, or ChIP assays (Zhong et al., 2010a; Zhong and Ye, 2012), 14 TF and 7 cell wall biosynthetic genes were identified as the direct targets of AtSND1, and 17 TF and 12 cell wall biosynthetic genes are direct targets of AtMYB46. Using these TF–DNA interactions, a three-layered hierarchical TRN directed by AtSND1 and AtMYB46 was constructed for Arabidopsis stem cell wall biosynthesis. Black lines indicate the regulatory TF–target regulation through protein–DNA interactions. Cyan shading indicates TF–DNA interactions also present in the P. trichocarpa SND1-B1 TRN. CW, cell wall; SCW, secondary cell wall.