Introduction

Secondary cell wall is a structure found in many plant cells, located between the primary cell wall and the plasma membrane. The cell starts producing thick secondary cell wall after the primary cell wall biosynthesis is complete and the cell has stopped expanding (Lerouxel et al. 2006; Somerville 2006; Zhong and Ye 2007). Secondary cell wall consists mainly of cellulose, hemicellulose and lignin. Secondary cell wall is a defining feature of the cells in xylem fibers and vessels that provide mechanical support for their growing body and serve as a conduit for long-distance transport of water/solutes, respectively (Fukuda 1997; Ye 2002). In addition, secondary wall from plant biomass is a major source of lignocellulosic material and one of the most promising sources of sugars for liquid biofuel production (Han et al. 2007).

Formation of secondary wall requires a coordinated transcriptional activation of the genes involved in the biosynthesis of secondary wall components such as cellulose, hemicellulose and lignin. However, our understanding of the transcriptional regulatory network for secondary wall biosynthesis is limited. Recent studies on NAC and MYB transcription factors have provided an insight into the complex network of transcriptional regulation of secondary wall biosynthesis (Zhong and Ye 2007). ANAC012/SND1 (At1g32770) functions as a potential regulator of secondary wall thickening in xylary fibers (Zhong et al. 2006; Ko et al. 2007). Several MYB transcription factors were also identified as important regulators of secondary wall biosynthesis in Arabidopsis. MYB46, a direct target of ANAC012/SND1, is capable of inducing secondary wall formation (Zhong et al. 2007; Ko et al. 2009). Overexpression of MYB46 results in ectopic deposition of secondary walls in the cells that are normally parenchymatous, while suppression of its function reduces secondary wall thickening (Zhong et al. 2007; Ko et al. 2009). However, the mechanism underlying MYB46-mediated transcriptional regulation is still largely unknown, mainly due to the lack of information on downstream targets of MYB46. Information on cis-acting regulatory elements (i.e., transcription factor binding motifs) that are recognized by MYB46 will facilitate the identification of the target genes of MYB46, which are involved in transcriptional regulation of secondary wall biosynthesis.

The goal of the current study is to identify cis-regulatory motifs in MYB46-mediated transcriptional regulation. Recently, we have identified AtC3H14 (At1g66810) as a direct target of MYB46 (Ko et al. 2009). AtC3H14 is a plant specific CCCH-type zinc finger protein in Arabidopsis (Wang et al. 2008) and one of the 68 CCCH-type zinc finger proteins in Arabidopsis having two identical C-x8-C-x5-C-x3-H motifs (TZF; Tandem CCCH Zinc Finger) (Wang et al. 2008). However, the exact function of AtC3H14 is not characterized yet. We then carried out a series of electrophoretic mobility shift assays (EMSA) using MYB46 protein and AtC3H14 promoter sequences with various deletions. The analysis narrowed down the MYB46 protein binding region to a 25-nucleotide fragment and identified a core MYB46-responsive cis-regulatory element (M46RE). Using in vivo transient activation analysis, we confirmed that the M46RE is necessary and sufficient for MYB46-responsive transcription. Genome-wide analysis of promoter sequences in Arabidopsis revealed that the most of the MYB46-induced secondary wall biosynthetic genes have one or more M46REs in their promoter region. We discuss the significance of this motif in the transcriptional regulation of secondary wall biosynthesis.

Materials and methods

Plant materials and growth conditions

Arabidopsis thaliana, ecotype Columbia (Col-0), was grown on soil in a growth chamber (16 h light/8 h dark) at 23°C. Rosette leaves of three-week-old Arabidopsis were used in transcriptional activation analysis (described below).

Protein expression and purification

MYB46 was fused in frame with GST and expressed in Escherichia coli strain Rosetta gami (Novagen). The induction of the recombinant MYB46-GST proteins was performed by adding 0.3 mM IPTG (isopropyl β-d-thiogalactopyranoside) for 16 h of incubation at 16°C. The recombinant proteins were purified using Glutathione immobilized particles (MagneGST; Promega) and used for EMSA.

Electrophoretic mobility shift assay (EMSA)

DNA fragments for EMSA were obtained by PCR-amplification or synthesized and labeled with [γ-32P]ATP using T4 polynucleotide kinase (NEB, http://www.neb.com/). The end-labeled probes were purified with Microspin S-200 HR column (GE Healthcare, http://www.gelifesciences.com/). The labeled DNA fragments were incubated for 25 min with 50 ng of MYB46-GST in a binding buffer [10 mM Tris, pH 7.5, 50 mM KCl, 1 mM DTT, 2.5% glycerol, 5 mM MgCl2, 100 μg/mL BSA, and 50 ng/μL poly(dI-dC)]. Five percent polyacrylamide gel electrophoresis (PAGE) was used to separate the recombinant protein-bound DNA fragments from the unbound ones. The gel was dried, and placed in a film cassette and exposed to X-ray film (Kodak) for over night. Radioactive fragments were visualized by autoradiography.

Transcriptional activation analysis

Preparation of Arabidopsis leaf protoplasts and transient transfection of reporter and effector constructs were carried out as described previously (Ko et al. 2009). Briefly, transfected protoplasts were lysed, and the soluble extracts were used for analysis of GUS, after 16 h of incubation. For the time-course experiment, each transfected protoplasts were harvested at the indicated time for GUS activity measurement. In each experiment, the expression level of the GUS reporter gene in the protoplasts transfected with the reporter construct alone was used as control. The data were the average of three biological replications with S.E. pTrGUS vector was used to produce both effector and reporter constructs. pTrGUS is a home-made vector designed to reduce the vector size around 4 kb to improve transfection efficiency and was provided by Dr. Soo-Un Kim (Seoul National University, Korea). For effector constructs, the full-length cDNAs of transcription factors were ligated between CaMV 35S promoter and NOS terminator after removing GUS of the pTrGUS vector. The reporter constructs were created by placing promoter fragments in front of the GUS reporter gene after removing 35S promoter region of the pTrGUS. High quality plasmids were prepared by PureHeix™ Fast-n-Pure plasmid kit (Nanohelix, South Korea). Transfection and transient expression by using this vector system was successfully demonstrated (Ko et al. 2009). Primers used for PCR amplification of full-length genes and promoter fragments were summarized in Table S1.

Chromatin immunoprecipitation analysis

The full-length cDNA of MYB46 was fused in frame with GFP cDNA and inserted behind the GAL4 upstream activation sequence of the pTA7002 binary vector (Aoyama and Chua 1997). The construct was introduced into wild-type Arabidopsis thaliana (Col-0) by Agrobacterium-mediated transformation and the resulting transgenic plants were used for chromatin immunoprecipitation analysis as described in Zhong et al. (2007) with slight modification. Briefly, the MYB46-GFP/pTA7002 transgenic plants were grown on soil for 3 weeks and treated 10 μM DEX with 0.02% silwet surfactant (Lehle Seeds) for 8 hours. The seedlings were then harvested and cross-linked in 1% formaldehyde for 10 min under vacuum and quenched by adding 0.125 M of glycine for 5 min. The seedlings were washed twice with deionized water and then ground in liquid nitrogen. To extract chromatin, 2 g of ground powder was resuspended in 30 mL of extraction buffer 1 (10 mM Tris–HCl, pH 8.0, 0.4 M sucrose, 5 mM 2-mercaptoethanol, 1 mM PMSF, 1X protease inhibitor cocktail, and 4 μg/mL pepstain A) and filtered through two layers of miracloth before centrifugation at 2,500 g for 20 min at 4°C. The pellet was resuspended in 1 mL of extraction buffer 2 (10 mM Tris–HCl, pH 8.0, 0.25 M sucrose, 10 mM MgCl2, 1% Triton X-100, 5 mM 2-mercaptoethanol, 1 mM PMSF, 1X protease inhibitor cocktail, and 4 μg/mL pepstain A) and centrifuged at 14,000 g for 10 min at 4°C. The pellet was resuspended in 300 μL of extraction buffer 3 (10 mM Tris–HCl, pH 8.0, 1.7 M sucrose, 0.15% Triton X-100, 2 mM MgCl2, 5 mM 2-mercaptoethanol, 1 mM PMSF, 1X protease inhibitor cocktail, and 4 μg/mL pepstain A) and layered on top of 300 μL of extraction buffer before centrifugation at 14,000 g for 1 h at 4°C. The chromatin pellet was resuspended in 500 μL of ice-cold nuclei lysis buffer (50 mM Tris–HCl, pH 8.0, 10 mM EDTA, 1% SDS, 1 mM PMSF, 1X protease inhibitor cocktail, and 4 μg/mL pepstain A) and sonicated to small fragments with an average size of 600–800 bp. Sonication was done by Omni-Ruptor 250 (OMNI International Inc., USA) with four cycles of 12 s pulse (40% duty) and 1 min pause. The sonicated chromatin was diluted 1:10 in ChIP dilution buffer (16.7 mM Tris–HCl, pH 8.0, 1.1% Triton X-100, 1.2 mM EDTA, 167 mM NaCl, 1 mM PMSF, 1X protease inhibitor cocktail, and 4 μg/mL pepstain A) and pre-cleared by incubation with Protein A agarose beads (Roch) for 1 h. The pre-cleared chromatin was then incubated with 2 μg of GFP antibody (Abcam) overnight at 4°C. The immune complexes were collected by incubation with Protein A agarose beads for 1 h and subsequent washing for 5 min with 1 mL of each in low-salt wash buffer (20 mM Tris–HCl, pH 8.0, 150 mM NaCl, 0.2% SDS, 0.5% Triton X-100 and 2 mM EDTA), high-salt wash buffer (20 mM Tris–HCl, pH 8.0, 500 mM NaCl, 0.2% SDS, 0.5% Triton X-100 and 2 mM EDTA), LiCl wash buffer (10 mM Tris–HCl, pH 8.0, 0.25 M LiCl, 0.5% NP-40, 0.5% sodium deoxycholate and 1 mM EDTA) and TE buffer. The immune complexes were then eluted with 500 μL elution buffer (1% SDS and 0.1 M sodium bicarbonate) at 65°C for 15 min with agitation. To reverse protein-DNA cross-linking, the eluted chromatin was incubated in 0.2 M NaCl at 65°C without agitation for overnight. Chromatin DNA was further treated with proteinase K (0.2 mg/mL) for 1 h to remove proteins before used for quantitative PCR analysis. The enrichment of the AtC3H14 promoter sequence was examined by real-time quantitative PCR analysis of chromatin samples before (input) and after (bound) immunoprecipitation. The promoter sequence of the MYB46 gene, which has been shown not to be a downstream target of MYB46 (Ko et al. 2009), was used as an internal control for nonspecific binding of genomic DNA. The values of bound over input for AtC3H14 promoter fragment was normalized against that of the control DNA (MYB46 promoter, −206 to −440 bp). A promoter sequence (−264 to −529 bp) of MYB54 (At1g73410) was used as a negative control. The experiments were repeated three times and obtained similar results.

Results

Identification of MYB46 binding site in the promoter region of AtC3H14

In order to identify a cis-acting regulatory motif recognized by MYB46, we characterized the promoter sequence of AtC3H14 gene, a known direct target of MYB46 (Ko et al. 2009; Fig. S1). MYB46 activates AtC3H14 transcription in vivo (Fig. S1a and b) as shown previously (Ko et al. 2009). We then confirmed that the recombinant MYB46 protein binds to the AtC3H14 promoter fragment evidenced by the mobility shift in the EMSA (Fig. S1c). This interaction was sequence-specific because the addition of a cold (i.e., unlabeled) AtC3H14 promoter fragment (Pro_AtC3H14-S2; −250 to −500 bp upstream of the start codon) effectively competed with the binding in a highly dose-dependent manner, and the mobility shift was not observed when the AtC3H14 promoter fragment was incubated with GST alone (Fig. S1c).

To narrow down MYB46 binding region in the AtC3H14 promoter, pro_AtC3H14-S2 was divided into three fragments (P1, P2, and P3) with 25-bp overlap (Fig. 1a) and the fragments were used as competitors against pro_AtC3H14-S2 probe in the EMSA experiments. Cold pro_AtC3H14-S2 was used as a positive control in the experiments (Fig. 1b). The result showed that one region, pro_AtC3H14-P3, exhibited very strong binding by MYB46 (Fig. 1b). We next divided the pro_AtC3H14-P3 fragment into seven 25-bp oligonucleotides (P3-1 to P3-7) with 13-bp overlap to further narrow down the sequence that was recognized by MYB46 (Fig. 1c, d). Following EMSA analysis identified the pro_AtC3H14-P3-2 as the binding region recognized by MYB46 (Fig. 1d).

Fig. 1
figure 1

Determination of MYB46 binding region in AtC3H14 promoter. a Diagram of the AtC3H14 promoter region used in the EMSA experiment in b. b MYB46 binds to the 100-bp fragment (P3) of AtC3H14 promoter, which is located between −400 and −500 relative to the start codon. GST was used as control. For competition analysis, competitors (unlabeled DNA fragments) were added to the reaction mixture with 30-fold molar excess relative to the labeled probes. S2 indicates Pro_AtC3H14-S2 in a. c Diagram of overlapping 25-bp DNA oligonucleotides in the AtC3H14 promoter fragment (pro_AtC3H14-P3) used in the EMSA experiment in d. d Competition analysis of MYB46 binding to Pro_AtC3H14-P3 by the overlapping 25-bp DNA oligonucleotides shown in c. Competitors (25-bp unlabeled DNA oligonucleotides) in 50-fold molar excess were used in the EMSA. A 25-bp oligonucleotide, P3-2, competed with Pro_AtC3H14-P3 for binding to MYB46. GST was used as control. The free DNA probes are indicated by arrow

Full size image

Defining MYB46-responsive cis-regulatory element (M46RE)

To identify the specific MYB46 binding sequence, we created a series of point mutations in the pro_AtC3H14-P3-2 sequence (Fig. 2a). Further competition analysis using these mutated oligonucleotides as competitor revealed that mutations in m13, m15, m16, m17, m18, m19 or m20 resulted in loss of binding ability to MYB46 (Fig. 2b), identifying the 8-bp AGTAGGTG as a putative core sequence.

Fig. 2
figure 2

Determination of MYB46-responsive cis-regulatory element in the AtC3H14 promoter. a Wild-type (WT) and mutated Pro_AtC3H14-P3-2 oligonucleotides (m1–m22). Each double-stranded probe contains a single base substitution as indicated. Dashes indicate no change. b Competition analysis of MYB46 binding to Pro_AtC3H14-P3 by mutated Pro_AtC3H14-P3-2 oligonucleotides shown in a. Competitors (unlabeled WT and mutated Pro_AtC3H14-P3-2 oligonucleotides) in 50-fold molar excess were used in the EMSA. Note that m13, m15, m16, m17, m18, m19 and m20 could not compete with Pro_AtC3H14-P3 for binding to MYB46. The free DNA probes are indicated by arrow

Full size image

The identified 8-bp core sequence is present in the overlapping region of P3-1 and P3-2 fragments (Fig. S2a). Interestingly, the P3-1 sequence, which starts with the core sequence, could not compete with pro_AtC3H14-P3 probe for MYB46 binding (Fig. 1c, d). It is known that, for protein binding, extra nucleotides on both sides of a short motif are commonly needed in order to form a helical structure. To confirm this possibility, we performed additional competitive EMSAs using mutated versions of P3-1, which includes additional three nucleotides lead sequence in front of the core sequence (Fig. S2a). The mutated versions of P3-1, both P3-1-m1 and P3-1-m2, successfully compete with the pro_AtC3H14-P3 probe (Fig. S2b), indicating that at least three additional non-specific nucleotides are necessary for MYB46 protein binding on the core sequence.

To identify a consensus MYB46 binding motif, the 8-bp core sequence was further mutated by substituting one nucleotide at a time with all other three nucleotides (Fig. 3a). These mutated oligonucleotides were then used as competitor in the EMSA assay (Fig. 3c). From the EMSA results we identified RKTWGGTR (where R stands A or G; K for G or T; W for A or T) as a MYB46-responsive cis-regulatory element (named M46RE) (Fig. 3c). Interestingly, M46RE is almost identical to MBSIIG (GKTWGGTR) that was identified by ‘binding site selection’ with four representative R2R3-MYB transcription factors, AtMYB15, AtMYB77, AtMYB84, and AtMYBGI1 (Romero et al. 1998).

Fig. 3
figure 3

Identification of a consensus MYB46-responsive cis-regulatory element (M46RE) in AtC3H14 promoter. a Wild-type (WT) Pro_AtC3H14-core sequence was mutated by substituting one nucleotide at a time with all other three nucleotides (A–X). Dashes indicate no change. b MYB46 binds to 14-bp fragment (core) of the AtC3H14 promoter. c Competition analysis of MYB46 binding to Pro_AtC3H14-core by mutated Pro_AtC3H14-core oligonucleotides shown in a. Competitors (unlabeled WT and mutated Pro_AtC3H14-core oligonucleotides) in 50-fold molar excess were used in the EMSA

Full size image

M46RE is necessary and sufficient for MYB46 binding

To confirm the importance of M46RE in MYB46-induced activation of AtC3H14 transcription in Arabidopsis, a protoplast-based transient transcriptional activation analysis (TAA) (Walker et al. 1987; Skriver et al. 1991; Sheen 2001; Yoo et al. 2007; Yi and Liu 2009; Ko et al. 2009) was performed using GUS reporter constructs driven by wild-type, a single or four base mutations (pro_AtC3H14-M1 to M5) in the core motif, or deletion (pro_AtC3H14ΔP3-2) of the entire M46RE (Fig. 4a, b). While the wild-type version of AtC3H14 promoter (pro_AtC3H14-WT) showed high level of GUS activity by MYB46, all of the mutant versions showed very little GUS activity (Fig. 4c). This result shows that the M46RE is necessary for MYB46-dependent AtC3H14 promoter activity in vivo (Fig. 4c).

Fig. 4
figure 4

M46RE is necessary for MYB46 activation of AtC3H14 transcription in vivo. a Diagram of the effector and reporter constructs used in the transcriptional activation assay. The effector construct contained MYB46 gene driven by CaMV 35S promoter. The reporter constructs contained GUS reporter gene driven by the indicated wild type (WT) or mutated AtC3H14 promoters (M1 to M5 and DP3-2). Mutated region (P3-2, red box) is indicated by arrowhead. DP3-2 means the deletion of P3-2 sequence from AtC3H14 promoter. b Base substitutions of M46RE in the mutated region of AtC3H14 promoter in a. Dashes indicate no change. c MYB46 could not activate AtC3H14 transcription driven by the mutated promoters in vivo. MYB46 was co-expressed in Arabidopsis leaf protoplasts with GUS reporter gene driven by wild-type (WT) or mutated AtC3H14 promoters (M1 to M5 and DP3-2) shown in a. Activation of the promoter by MYB46 was measured by assaying GUS activity after 15-h of incubation. The expression level of the GUS reporter gene in the protoplasts transfected with no effector was used as control and was set to 1. Error bar indicates SE of three biological replications

Full size image

Is M6RE sequence alone sufficient for MYB46 binding? To address this question, we carried out TAA experiments with GUS reporter gene driven by either 500-bp region of MYB46 promoter as negative control or the MYB46 promoter with three tandem repeats of M46RE fused to the upstream and 500-bp region of AtC3H14 promoter as positive control (Fig. 5a). We found that AtC3H14 promoter and MYB46 promoter with the M46REs were able to increase the GUS activity while the native MYB46 promoter did not activate the reporter gene (Fig. 5b), indicating that M46RE sequence alone is sufficient for MYB46 binding.

Fig. 5
figure 5

M46RE is sufficient for MYB46-mediated activation of AtC3H14 transcription in vivo. a Diagram of the effector and reporter constructs used in the transcriptional activation assay. The effector construct contained MYB46 gene driven by CaMV 35S promoter. The reporter constructs contained GUS reporter gene driven by the indicated AtC3H14 promoter, MYB46 or MYB46 promoter plus three tandem repeats of M46RE. b MYB46 could activate GUS gene transcription when the three tandem repeats of M46RE were added to the promoter. MYB46 was co-expressed in Arabidopsis leaf protoplasts with GUS reporter gene driven by wild-type (WT) or MYB46 promoter plus three tandem repeats of M46RE as shown in a. Activation of the promoter by MYB46 was measured by assaying GUS activity after 15-h of incubation. The expression level of the GUS reporter gene in the protoplasts transfected with no effector was used as control and was set to 1. Error bar indicates SE of three biological replications

Full size image

MYB46 bind promoter of AtC3H14 in planta

To further corroborate the results from the EMSA and TAA, we used a chromatin immunoprecipitation (ChIP) assay to examine whether MYB46 binds to the AtC3H14 promoter in planta. For this experiment, we overexpressed GFP-tagged MYB46 protein in Arabidopsis under the control of a dexamethason (DEX)-inducible promoter (Fig. 6a). Inducible overexpression of GFP-tagged MYB46 resulted ectopic secondary wall thickenings in curled leaves (Fig. 6c), as shown in the transgenic Arabidopsis overexpressing MYB46 (Ko et al. 2009), indicating that the MYB46-GFP fusion protein is fully functional in the activation of secondary wall biosynthesis. After the induction of MYB46-GFP by DEX treatment, formaldehyde cross-linked chromatin was isolated, fragmented and immunoprecipitated with GFP antibodies for enrichment of MYB46-GFP bound DNA fragments. Quantitative real-time PCR was performed to detect AtC3H14 promoter sequence using the immunoprecipitated DNA fragments as templates. The results showed that the pro_AtC3H14-P3 fragments, which were strongly bound by MYB46 in the EMSA, were enriched more than three-fold compared to the control DNA (Fig. 6b). For a negative control, we used the promoter sequence of MYB54 (At1g73410), which is highly up-regulated by MYB46 overexpression (Ko et al. 2009) but has no M46REs in its promoter sequence. As expected, no enrichment of MYB54 promoter was detected (Fig. 6b). The results from the in vitro and in vivo binding analyses demonstrate that MYB46 directly binds to the promoter of AtC3H14.

Fig. 6
figure 6

Chromatin immunoprecipitation (ChIP) analysis of MYB46 binding to the AtC3H14 promoter sequence in planta. a Diagram of the vector construct used for inducible expression of MYB46-GFP in the ChIP analysis. b Real-time PCR analysis shows the enrichment of the AtC3H14 promoter sequence after chromatin immunoprecipitation. The values were normalized against that of the control DNA (promoter of MYB46). MYB54 promoter was used as a negative control which dose not have M46RE. Error bars represent SE from three biological replicates. c Leaf-curling phenotype was observed after 24 h of DEX treatment. d Primer sequences used for the ChIP experiments

Full size image

Significance of M46RE as a cis-element in secondary wall biosynthesis

In order to survey genome-wide distribution of M46RE in the promoter regions (within 1 kb upstream sequence from start codon ATG), we used a ‘Patmatch’ program (http://www.arabidopsis.org/cgi-bin/patmatch/nph-patmatch.pl) with RKTWGGTR as a query. This analysis resulted in identification of 9,836 genes, which is 43.5% out of 22,591 genes included in the ATH1 Genome Array (Affymetrix), that have M46REs in their promoters (Table 1). The frequency of the promoters having M46REs increases up to 70% among secondary wall biosynthesis-related genes, compared to the genome-wide average of 43.5% (Table 1). The frequency goes even higher among the 414 genes that are upregulated three-fold or higher within 6-h of MYB46 induction (i.e., induction of ectopic secondary wall biosynthesis) (Ko et al. 2009). For examples, all of the five MYB46-induced cellulose synthase (CesA) and cellulose synthase-like (CSL) genes had M46RE motifs in their promoter regions while only 48.7% (19 of 39) of CesA/CSL genes in the genome had the motif. Furthermore, 91.3% (21 of 23) of the MYB46-upregulated lignin biosynthesis genes had the motifs while only 53% (187 out of 353) such genes in the genome had the motif (Table 1). Taken together, it appears that M46RE is highly enriched in the promoters of the genes that are involved in MYB46-regulated secondary wall biosynthesis.

Table 1 Global analysis of genes having M46RE in the promoter region
Full size table

Utility of M46RE in the identification of MYB46 target genes

In order to assess the utility of M46RE motif, we surveyed the promoter regions of the 52 genes that we identified as not only co-expressed with but also up-regulated by MYB46. Over 71% (37 out of 52) of the genes contain multiple M46REs in their promoter regions (Tables 1 and 2), suggesting that these genes are likely directly regulated by MYB46. Three (AtC3H14, KNAT7 and MYB63) of the five transcription factors on the list have been shown to be activated by MYB46 in vivo (Ko et al. 2009). EMSA results show that MYB46 specifically bind to the promoter of KNAT7 (Kim et al. unpublished data). Interestingly, many of cell wall modification, secondary wall biosynthesis genes (e.g., cellulose, xylan and lignin) or cellular processes related to cell wall biosynthesis, such as cytoskeletal organization and signal transduction, were present on this list. Specifically, the genes on the list include three secondary wall-associated cellulose synthase (IRX1/AtCesA08, IRX3/AtCesA07 and IRX5/AtCesA04) (Taylor et al. 2003), xylan biosynthesis genes (IRX8, IRX9, IRX14, and IRX15-L) (Peña et al. 2007; Wu et al. 2010; Jensen et al. 2010; Brown et al. 2011), three laccase (IRX12/laccase4, laccase10 and laccase11) involved in lignin biosynthesis (Brown et al. 2005), two cytoskeleton-related genes (Myosin5, microtubule-associated protein) (Kaneda et al. 2010; Pesquet et al. 2010), and two protein kinases (At1g24030 and At1g03920). In addition, the list includes two DUF579 (At1g33800 and At4g09990), which are homologous to two xylan biosynthesis genes IRX15 and IRX15-L (Jensen et al. 2010; Brown et al. 2011). This result suggests that MYB46 may directly regulate not only transcription factors but also structural genes of secondary wall biosynthesis as Zhong et al. (2010) have shown with ANAC012/SND1. Furthermore, most of the putative direct targets of MYB46 were co-expressed with secondary wall-associated cellulose synthases (AtCesA4, AtCesA7 and AtCesA8), and up-regulated by as well as co-expressed with ANAC012, a direct upstream regulator of MYB46 (Table 2).

Table 2 Putative direct targets of MYB46
Full size table

Discussion

Secondary wall biosynthesis is one of the most important biological processes on Earth because secondary walls constitute the majority of lignocellulosic biomass, which is of primary importance to humans as feedstock for biomaterials and as an environmentally cost-effective renewable source of energy. Recent studies have demonstrated that MYB46, one of the direct targets of ANAC012, functions as a master regulator of secondary wall biosynthesis (Zhong et al. 2007; Ko et al. 2009). However, the mechanism by which MYB46 activates the transcription of downstream target genes is largely unknown. Identification of MYB46 target sequences in the promoter of downstream genes will facilitate our efforts to understand the mechanism.

A recent report described that MYB46 strongly binds to the promoter region of AtC3H14, a CCCH-type zinc finger protein (Ko et al. 2009). To identify the cis-acting regulatory motif recognized by MYB46, we used the promoter sequence of AtC3H14 gene. First, we successfully confirmed the strong interaction between MYB46 and the promoter sequence (pro_AtC3H14-S2; Fig. S1). And then, a series of competitive EMSAs with various deletions of the promoter were carried out to narrow down the binding region to a 14-nucleotide motif (Pro_AtC3H14-core; Figs. 1, 2, 3). Subsequent EMSAs with a series of point mutations within the sequence identified an eight-nucleotide motif, RKTWGGTR, as a MYB46-responsive cis-regulatory element (M46RE) (Fig. 3). The in vivo activation analyses indicated that M46RE was both necessary and sufficient for MYB46-responsive transcription (Figs. 4, 5). Furthermore, we demonstrated that MYB46 directly binds to the AtC3H14 promoter in planta using a chromatin immunoprecipitation analysis (Fig. 6). We also examined the expression level of AtC3H14 in a T-DNA knockout mutant myb46 (SALK_100993C) and found no changes in its expression (Fig. S3). This was expected because, in addition to MYB46, there are other upstream regulators (e.g., MYB83 and ANAC012 of AtC3H14 (Ko et al. 2009; Kim, Ko, and Han, unpublished result).

It appears that M46RE is highly enriched in the promoters of MYB46-regulated secondary wall biosynthesis genes (Table 1). A promoter sequence analysis using Patmatch (TAIR) revealed that 43.5% of the 22,591 on the ATH1 genome array have M46RE in their promoters. However, the frequency dramatically increases among the genes up-regulated by MYB46, especially the genes involved in secondary wall biosynthesis (Table 1). For example, all of the five CesA/CSL genes (AtCesA4, AtCesA7, AtCesA8, AtCslA9, and AtCslB2) upregulated by MYB46 have multiple M46REs in their promoter regions. Seven of the nine MYB46-induced xylan biosynthesis genes have M46REs in their promoter region. Furthermore, 21 of the 23 lignin biosynthesis genes that are upregulated by MYB46 have multiple M46REs in their promoter regions.

Romero et al. (1998) divided MYB transcription factors into three groups (Group A, B, and C) based on sequence similarity and found that several MYB proteins in Group C, to which MYB46 belongs, bind to a consensus sequence GKTWGGTR, named MBSIIG. Rammírez et al. (2011) reported a similar cis-element (GTTAGGT) from the promoter region of a fungal pathogen-induced type III cell wall-bound peroxidase gene, Ep5C. This cis-element has the consensus sequence of both MBSIIG and M46RE and is recognized by MYB46 (Rammírez et al. 2011). A myb46 mutant showed enhanced induction of Ep5C gene following fungal pathogen infection and subsequent resistance to the pathogen. This observation is interesting because there appears to be no functional redundancy between MYB46 and its close homolog MYB83, which also binds to the M46RE. Whether the GTTAGGT motif is specific to MYB46 and, therefore, not recognized by MYB83 remains to be tested. Moreover, M46RE is complementary to a class of CA-rich regulatory elements named ACI and ACII (i.e., AC element), which are known to mediate transcriptional activation of lignin biosynthetic genes (Hatton et al. 1995; Raes et al. 2003). For example, PtMYB4 and EgMYB2 bind these elements to activate AC element-dependent transcription of lignin biosynthesis genes (Patzlaff et al. 2003; Rahantamalala et al. 2010). It appears that M46RE serve as a cis-acting element responsible for MYB46-mediated regulation of not only lignin biosynthesis but also cellulose and xylan biosynthesis (Tables 1, 2). Recently, Winzell et al. (2010) proposed that a conserved CA-rich motif (CCACCAAC, almost identical to ACII) is responsible for PtxtMYB021 (MYB46-like gene from hybrid aspen)-mediated transcriptional regulation of xylan-active CAZymes GT43A, GT43B and Xyn10A. This fact further supports our hypothesis that M46RE may function as a cis-acting element in MYB46-mediated transcriptional regulation of secondary wall biosynthesis.

One possible utility of M46RE is in the identification of potential direct target genes of MYB46 in a genome-wide promoter sequence analysis. We identified 37 putative direct target genes of MYB46 based their spatio- and temporal expression pattern and the presence of M46RE motif in their promoters (Table 2). This list includes not only downstream transcription factors but also a number of structural genes involved in secondary wall biosynthesis, cell wall modification, and unknown function (Table 2). Secondary wall-associated cellulose synthases (AtCesA4, AtCesA7 and AtCesA8; Taylor et al. 2003) on the list contain M46REs in their promoters, which serve as a binding site for MYB46 (Kim, Ko, and Han, unpublished results). This finding suggests that MYB46 may directly regulate some of the structural genes involved in secondary wall biosynthesis. Similarly, ANAC012/SND1, a direct upstream regulator of MYB46, has been suggested to regulate not only downstream transcription factors but also the structural genes involved in secondary wall biosynthesis (Zhong et al. 2010). However, the direct targets of ANAC012/SND1 that Zhong et al. (2010) identified do not overlap with those of MYB46 except laccase11 (At5g03260), although most of the putative direct targets of MYB46 are up regulated and co-expressed with ANAC-12/SND1 (Table 2). This finding suggests that ANAC012- and MYB46-mediated transcriptional regulation of secondary wall biosynthesis is closely associated but separated, further confirming the multifaceted and complex nature of the transcriptional regulation.

Lignocellulosic biomass, consisting mainly of secondary wall, is gaining research interest as a feedstock for production of bio-based materials and biofuel. However, our understanding of the molecular mechanisms underlying its biosynthesis is limited. In this study, we identified a cis-acting element, M46RE, which plays a critical role in the interaction of MYB46, a master regulator of secondary wall biosynthesis, and its downstream target genes. The findings in this study will facilitate our efforts to better understand the transcriptional regulatory network that controls secondary wall biosynthesis.