Poplar PdMYB221 is involved in the direct and indirect regulation of secondary wall biosynthesis during wood formation

Wood is formed by the successive addition of secondary xylem, which consists of cells with a conspicuously thickened secondary wall composed mainly of cellulose, xylan and lignin. Currently, few transcription factors involved in the direct regulation of secondary wall biosynthesis have been characterized in tree species. Here, we show that PdMYB221, a poplar ortholog of the Arabidopsis R2R3-MYB transcription factor AtMYB4, directly regulates secondary wall biosynthesis during wood formation. PdMYB221 is predominantly expressed in cells of developing wood, and the protein it encodes localizes to the nucleus and acts as a transcriptional repressor. Ectopic expression of PdMYB221 resulted in reduced cell wall thicknesses of fibers and vessels in Arabidopsis inflorescence stems. The amounts of cellulose, xylose, and lignin were decreased and the expression of key genes synthesizing the three components was suppressed in PdMYB221 overexpression plants. Transcriptional activation assays showed that PdMYB221 repressed the promoters of poplar PdCESA7/8, PdGT47C, PdCOMT2 and PdCCR1. Electrophoretic mobility shift assays revealed that PdMYB221 bound directly to the PdCESA8, PdGT47C, and PdCOMT2 promoters. Together, our results suggest that PdMYB221 may be involved in the negative regulation of secondary wall formation through the direct and indirect suppression of the gene expression of secondary wall biosynthesis.

master switches regulating secondary wall formation. Of them, PtrMYB2, PtrMYB3, PtrMYB20 and PtrMYB21 are direct targets of PtrWNDs and share similar functions to their Arabidopsis orthologs MYB46 and MYB83 6,7 . These PtrMYBs positively regulate secondary wall formation and have the abilities of activating secondary wall biosynthetic pathways. Interestingly, poplar PtrMYB152, an ortholog of Arabidopsis MYB43, is not induced by PtrWND2B overexpression but functions in the regulation of secondary wall biosynthesis, suggesting the presence of a PtrWND2B-independent pathway governing secondary wall biosynthesis 8 . Our recent study shows that PdC3H17 and PdC3H18 are direct targets of PdMYB3 and PdMYB21 and function as positive regulators in the differentiation of vascular cambium and secondary wall thickening in poplar 9 . Overexpression of PdC3H17 or PdC3H18 in poplar activates the expression of genes involved in the biosynthesis of cellulose, xylan and lignin.
Compared with many indirect regulators identified, few genes involved in the direct regulation of secondary wall biosynthesis during wood formation have been characterized in tree species. Only three studies in Arabidopsis and switchgrass to date show that transcription factors can directly control secondary wall biosynthesis. Arabidopsis MYB46 functions as second-level master switch that regulates many secondary wall-associated genes 10 . It has recently been shown that MYB46 directly activates the expression of three cellulose synthetic genes CESA4, CESA7 and CESA8 by binding to the M46RE sequences of their promoters 11 . Arabidopsis MYB58, MYB63 and switchgrass PvMYB4, three lignin-specific regulators, have been shown to have the abilities of binding to the AC elements of the promoters of some lignin synthetic genes to directly activate the expression of these genes [12][13][14] . Bioinformatic analysis of the promoters of all lignin biosynthetic genes in Arabidopsis indicated that such AC elements are present in the majority of the genes except for the promoters of ferulate 5-hydroxylase (F5H) 15 . It is supposed that the AC elements may be specific for the regulation of lignin biosynthetic genes 15,16 . However, little is known whether the AC elements are involved in the regulation of gene expression of cellulose and xylan biosynthesis.
Arabidopsis MYB4, a transcriptional repressor, has previously been shown to negatively regulate the accumulation of the UV-protectant compound sinapoylmalate by repressing the expression of the gene encoding the phenylpropanoid enzyme cinnamate 4-hydroxylase 17 . Recently, several studies demonstrate that MYB4 may be involved in the regulation of secondary wall biosynthesis. It can not only be activated by MYB46, a second-level master regulator of secondary wall biosynthesis 18 , but also suppress the expression of the first-level master regulator SND1/NST3 through a feedback loop 19,20 . In addition, PvMYB4, the switchgrass ortholog of Arabidopsis MYB4, has been shown to be a negative regulator of lignin biosynthesis 14 .
In this study, we investigated the role of PdMYB221, a poplar ortholog of Arabidopsis MYB4, in secondary wall formation, and determined its targets involved in cellulose, xylan and lignin biosynthesis. We showed that ectopic expression of PdMYB221 in Arabidopsis led to a reduction in secondary cell wall thicknesses of fibers and vessels. The amounts of cellulose, xylan, and lignin were decreased and the expression of key genes synthesizing the three components was suppressed in transgenic plants. Furthermore, we demonstrated that PdMYB221 suppressed the expression of the poplar PdCESA7/8, PdGT47C, PdCOMT2 or PdCCR1 promoters and had the abilities of binding to the AC elements of the PdCESA8, PdGT47C, and PdCOMT2 promoters. Our results suggest that poplar PdMYB221 may be involved in the negative regulation of secondary wall biosynthesis through the direct and indirect suppression of the expression of secondary wall biosynthetic genes.

Results
PdMYB221 is highly conserved in a range of plant species. The Populus trichocarpa genome contains 192 R2R3-MYB family members, which form 81gene pairs 21,22 . Of them, PdMYB156 and PdMYB221 are a pair of paralogous genes, with 88% sequence identity over their full lengths and 97% identity over their R2R3-MYB domains. Both proteins contain the potential repression motifs and are the orthologs of Arabidopsis AtMYB4 (Fig. 1A), consistent with the previous phylogenetic analysis 21 . Thirteen orthologs of PdMYB156 and PdMYB221 were identified in eight representative plant species, including green algae, moss, spike moss, Arabidopsis, rice, sorghum, alfalfa and switchgrass (Fig. 1B). Phylogenetic analysis revealed that the fifteen R2R3-MYB proteins including PdMYB156/221 have high bootstrap percentage values and are therefore relatively conserved. By contrast, PdMYB156 and PdMYB221 fell into the group containing Arabidopsis AtMYB4 and alfalfa Medtr4g073420.1, suggesting that these proteins are more evolutionary conserved within dicots than in monocots, moss or green algae. PdMYB221 and its paralog PdMYB156 are predominantly expressed in fibers and vessels of poplar stems. qRT-PCR was first performed to investigate expression patterns of PdMYB156 and PdMYB221 in six tissues of poplar. The results showed that both genes were expressed in all the tissues tested, with highest expression in xylem and phloem of the stems ( Fig. 2A). In situ hybridization analysis of stem sections indicated that PdMYB156 and PdMYB221 were highly expressed in xylary fiber, vessel, and phloem fiber cells, albeit at differing levels ( Fig. 2C,D). No hybridization signal was observed in wood cells in the control sections hybridized with the PdMYB221 sense probe (Fig. 2B). The similar result was obtained with the PdMYB156 sense probe (data not shown). The preferential expression in developing secondary xylem (i.e. wood-forming tissues) and similar expression patterns of the two genes suggest that they may have similar functions in wood formation. It appeared that PdMYB221 was more highly expressed than PdMYB156 in the woody tissues. Therefore, PdMYB221 was selected for further functional characterization in this study. PdMYB221 localizes to the nucleus and is a transcriptional repressor. The subcellular localization of PdMYB221 was examined using a tobacco leaf transient expression system 23 . Figure 3A showed that the PdMYB221:GFP fusion protein was colocalized to DAPI-staining nuclei, indicating that PdMYB221 encodes a nuclear-localized protein. Previous study showed that Arabidopsis MYB4 acts as a transcriptional repressor 17 . We examined whether poplar PdMYB221 functions as transcriptional repressor using the GAL4-DNA-binding-domain (GAL4BD) and its binding sites (GAL4(4X)-D1-3(4X)-GUS)-based protoplast transient expression system 24 . Figure 3B indicated that PdMYB221 caused repression of the reporter gene expression by a similar amount (≈ 50%) conferred by a known repressor protein, HOS15 24 , compared with BD alone (set to 1). The negative control, ARF5M, activated GUS expression 3.1-fold compared with BD only. Therefore, PdMYB221 has the ability to repress the transcription of its target genes.
Overexpression of PdMYB221 affects secondary cell wall development in the Arabidopsis stems. In order to gain insights into the biological roles of PdMYB221, the PdMYB221 gene was overexpressed under the control of the 35S promoter in wild-type Arabidopsis. At least 27 T 1 transgenic plants were generated and confirmed by RT-PCR (Fig. 4A). Phenotypic analyses were carried out on homozygous T 3 plants recovered from three independent transformants with high transcriptional levels. Compared with the wild type, transgenic lines overexpressing PdMYB221 had little difference before four weeks, but were no longer able to remain upright when they reached heights of >15 cm, possibly because of the loss of secondary walls in the cells of the stems (Fig. 4B). Observation of cross sections of the basal inflorescence stems of six-week-old plants revealed that the cell wall thicknesses of fibers and vessels were thinner in the transgenic plants than in the wild type ( Fig. 4C-J). Quantitative analysis showed that the cell-wall thicknesses of interfascicular fibers, xylary fibers, and vessels were decreased by 32%, 16% and 15%, respectively, in the transgenic lines, compared with the wild type (Fig. 4K). It is notable that PdMYB221 overexpression lines were often found to have partially deformed vessels, probably because of the weakening of the vessels' secondary walls (Fig. 4F). It is known that the change of wall thickness may reflect differences in cell-wall composition. We thus investigated the alteration of the cellulose, xylan and lignin contents in the stems of PdMYB221 overexpression plants. Calcofluor staining of cellulose, immunolabeling of xylan with the monoclonal antibody LM10 and phloroglucinol-HCl staining of lignin revealed the presence of very weak signals in fiber and vessel cells of PdMYB221 overexpression plants, while abundant cellulose, xylan and lignin were observed in those of the wild type (Fig. 5). Quantitative analysis of the cell-wall composition revealed that the amounts of cellulose, xylose and lignin were decreased by 10%, 26% and 28%, respectively, in the transgenic plants, compared with the wild type (Table 1). These results suggest that poplar PdMYB221 may negatively regulate secondary cell wall biosynthesis in Arabidopsis.

Overexpression of PdMYB221 affects the expression of secondary wall biosynthetic genes and secondary wall-associated transcription factors in Arabidopsis. Given that overexpression
of PdMYB221 led to visible alteration in secondary cell wall formation in Arabidopsis stems, we used qRT-PCR to examine the expression of the related genes in PdMYB221 overexpression lines. Expression of three cellulose synthetic genes (CESA4, CESA7, and CESA8), three xylan synthetic genes (FAR8, IRX8 and IRX9), and nine lignin synthetic genes (CCOMT1, COMT1, C3H1, HTC, 4CL1, CAD5, C4H, PAL1 and CCR1) was suppressed in PdMYB221 overexpression lines (Fig. 6A), consistent with the reduction of the cellulose, xylose, and lignin contents in transgenic plants (Table 1). This suggests that PdMYB221 may negatively regulate secondary wall biosynthesis by repressing these secondary wall biosynthetic genes in Arabidopsis. Detection of the expression of eight secondary wall-associated transcription factors indicated that SND1, NST1, VND6, and MYB85 were suppressed while MYB75 was induced in PdMYB221 overexpression lines (Fig. 6B). However, the expression of MYB43, MYB58, and MYB63 in transgenic lines did not differ from the expression levels detected for these genes in the wild type. Interestingly, the expression of three programmed cell death-associated genes XCP1, XCP2 and XND1 was found to be induced in PdMYB221 overexpression lines, suggesting that PdMYB221 may be involved in programmed cell death.
The PdCESA7/8, PdGT47C, PdCOMT2 and PdCCR1 promoters are repressed by PdMYB221 and PdMYB221 can bind to some of these promoters in vitro. Since the expression of a subset of  secondary wall biosynthetic genes was suppressed in PdMYB221 overexpression plants, we employed transcriptional activation assays in Arabidopsis mesophyll protoplasts to investigate whether their poplar counterparts' promoters could be repressed by PdMYB221. The PdCESA7, PdCESA8, PdGT47C, PdCOMT2, and PdCCR1 promoters were amplified from poplar gDNAs. The five genes were selected, because their Arabidopsis counterparts CESA7, CESA8, IRX8, COMT1 and CCR1 had relatively lower expression than other genes in PdMYB221 overexpression plants (Fig. 6A). The reporter and effector constructs were co-transfected into Arabidopsis leaf protoplasts, and subsequent assays of GUS activity in the transfected protoplasts showed that PdMYB221 was able to repress the expression of PdCESA7, PdCESA8, PdGT47C, PdCOMT2, and PdCCR1 (Fig. 7A,B).
Analysis of the promoter sequences of these five genes showed that PdCESA8, PdGT47C, and PdCOMT2 contained putative AC element consensus sequences ACC(A/T)A(C/A)C (Table S1), which have been shown to be the binding sites of lignin-specific regulatory genes 15 . We here used electrophoretic mobility shift assays (EMSA) to investigate whether PdMYB221 could bind directly to the PdCESA8, PdGT47C, and PdCOMT2 promoters. Recombinant HIS-PdMYB221 fusion protein was expressed in Escherichia coli and purified for use in the EMSA. As shown in Fig. 7B, the recombinant PdMYB221 was able to bind the PdCESA8, PdGT47C, and PdCOMT2 promoter fragments and cause mobility shifts. Addition of unlabeled PdCESA8, PdGT47C, or PdCOMT2 promoter fragments competed with the binding in a dose-dependent manner, indicating that the binding of PdMYB221 to the PdCESA8, PdGT47C, and PdCOMT2 promoters is specific. These results indicate that PdMYB221 has the capacity to directly suppress the expression of PdCESA8, PdGT47C, and PdCOMT2 in vitro. Alcohol insoluble residues (AIR) were prepared from the stems of six-week-old plants and determined by HPLC. The results are given as means ± SD (mg g −1 AIR) of three independent assays. Asterisks indicate significant differences (t test at P < 0.05) when compared with the wild type (n = 3).

Discussion
Many transcription factors including R2R3-MYB family members have been shown to function as positive regulators of wood formation in tree species [5][6][7][8]25 . By contrast, only several negative regulators have been characterized in tree species. Of them, EgMYB1, an ortholog of Arabidopsis MYB46 in Eucalyptus, negatively regulates secondary cell wall formation in Arabidopsis and poplar 26 . Poplar PoptrKNAT7 can rescue the knat7 phenotype when ectopically expressed in Arabidopsis, suggesting its potential negative role in secondary wall biosynthesis 27 . In this study, we provide evidence showing that poplar PdMYB221 may be involved in the negative regulation of secondary wall formation in Arabidopsis. The PdMYB221 gene is predominantly expressed in fibers and vessels of poplar stems (Fig. 2), suggesting its potential role in wood formation. The preferential expression of PdMYB221 in developing secondary xylem is consistent with its binding activity for the xylem expression-associated AC elements 28 . Transcriptional activity assays showed that PdMYB221 suppressed GUS expression in a protoplast transient express system (Fig. 3B), indicating that it is a transcription repressor. Phenotypic analysis revealed that overexpression of PdMYB221 decreased cell wall thickness of fibers and vessels in Arabidopsis inflorescence stems (Fig. 4). Consistent with this observation, the amounts of cellulose, xylose, and lignin were decreased in the stems of transgenic plants, compared with wild type (Table 1). This corresponded to the lower rate of cell wall polymer biosynthesis in transgenic lines. It is possible that the missing mass was not generated in transgenic lines. Alternatively, some intermediary compounds derived from cell wall polymer biosynthesis abnormally accumulated in PdMYB221 overexpression cells, which would partially compensate for the loss of cell wall mass. Consistently, overexpression of switchgrass miR156 has been shown to accumulate more soluble compounds due to the significant decrease in cell wall mass 29 . The detailed mechanism needs to be investigated in future. Together, these results suggest that poplar PdMYB221 may be a negative regulator of secondary wall biosynthesis during wood formation. Given the fundamental importance of secondary walls in vascular plants, it is plausible that these plants contain MYB activators and repressors for fine-tuning regulation of secondary wall formation.
Detection of the transcriptional levels of secondary wall-associated genes using qRT-PCR showed that overexpression of PdMYB221 in Arabidopsis suppressed the expression of fifteen secondary wall biosynthetic genes and four transcription factors (SND1, NST1, VND6, and MYB85) (Fig. 6). It is known that the four transcription factors positively regulate secondary wall biosynthesis 20,[30][31][32] . It is possible that PdMYB221 suppresses the expression of these secondary wall biosynthetic genes through the negative regulation of the four transcription factors. Further, SND1, NST1 or VND6 act as top master switches controlling secondary wall biosynthesis in Arabidopsis 20,30,31 , and SND1 expression is feedback suppressed by its downstream MYB4 gene 19 . Our results suggest that PdMYB221, an ortholog of MYB4, may suppress the expression of poplar orthologs of SND1, NST1 and VND6 through a similar negative feedback loop. Further experimental confirmation needs to be performed to better understand this. Our results also indicated that overexpression of PdMYB221 in Arabidopsis induced the expression of MYB75, a negative regulator of secondary wall formation 33 . Since PdMYB221 functions as a transcription repressor, one possible explanation for our result is that overexpression of PdMYB221 may repress a repressor of MYB75 in Arabidopsis. In addition to the indirect role of PdMYB221 in controlling secondary wall biosynthesis, several lines of evidence showed that PdMYB221 may directly suppress the expression of secondary wall biosynthetic genes in poplar. The PdCESA8, PdGT47C and PdCOMT2 genes are the key genes synthesizing cellulose, xylan and lignin, respectively, in poplar. Their promoters contain the putative AC element sequences (Table S1). We used EMSA to show that the three promoters can be bound by PdMYB221 in vitro (Fig. 7B). Moreover, the PdCESA8, PdGT47C and PdCOMT2 promoters can be suppressed by PdMYB221 in Arabidopsis mesophyll protoplasts (Fig. 7A). It should be noted that we cannot exclude the possibility that the poplar orthologs of other detected nine Arabidopsis lignin synthetic genes except for F5H may be also directly regulated by PdMYB221, because the expression of the nine genes was suppressed in PdMYB221 overexpression plants (Fig. 6A) and the promoters of their poplar orthologs contain the AC elements (Table S1). Further studies are required to test this supposition. Our current results appear to be different from the finding of Jin et al. (2000). These authors showed that Arabidopsis AtMYB4, the ortholog of poplar PdMYB221, negatively regulates the accumulation of the UV-protectant compound sinapoylmalate, principally through suppressing C4H expression 17 . A possible reason for the difference of their lignin-associated targets is dose-dependent selection of target genes by PdMYB221 and AtMYB4 17 .
Our data on poplar PdMYB221, combined with other recent studies of poplar genes encoding orthologs of Arabidopsis NAC and MYB transcription factors in the secondary wall regulatory network support the idea that the transcription regulatory network governing secondary cell wall biosynthesis is largely conserved in poplar and Arabidopsis (Fig. 8).
In summary, this study demonstrated that poplar PdMYB221 may be involved in the negative regulation of secondary wall biosynthesis through the direct and indirect suppression of the expression of key genes synthesizing cellulose, xylan and lignin. Our results mark another step toward the dissection of the molecular network that regulates secondary wall formation for genetic modification of poplar.

Materials and Methods
Plant materials and growth conditions. Poplar (Populus deltoides cv nanlin895) and Arabidopsis (Arabidopsis thaliana, Col-0) were grown in a greenhouse under standard conditions (temperature Scientific RepoRts | 5:12240 | DOi: 10.1038/srep12240 range 20-24 °C with a 16 h day from 06:00 h to 22:00 h and 55% relative humidity). For tissue-specific gene expression pattern analysis, leaves, roots, shoots, cortex, xylem and phloem were collected from 1.5-m-tall poplar plants. Samples were frozen immediately in liquid nitrogen and ground to powder in a cold mortar before subsequent RNA isolation.

RT-PCR and quantitative RT-PCR (qRT-PCR).
Total RNA isolation and first strand cDNA synthesis were performed as described previously 34 . Semi-quantitative RT-PCR was performed with gene-specific primers (Table S2) to detect the expression of PdMYB221 in wild-type and transgenic Arabidopsis plants. ACTIN2 (At3G18780) was used as an internal control. qRT-PCR assays were conducted on a LightCycler ® 480 Detection System (Roche) with TransStart Green qPCR superMix (TransGen Biotech) following the manufacturer's instructions. Baseline and threshold cycles (C t ) were determined with the 2 nd maximum derivative method using LightCycler ® 480 Software release 1.5.0. Quantification of gene expression relative to the reference genes ACTIN2 and PdUBQ10 (BU879229) was determined using the 2 −ΔCT method 35 . Data presented represent the average of three biological replicates.
In situ hybridization. The 144-bp 5′ UTR of PdMYB156 and 141-bp 3′ UTR of PdMYB221 were separately used for the synthesis of digoxigeninlabeled antisense and sense RNA probes with DIG RNA Labeling mix (Roche). The third internodes of the stems of 1.5-m poplar plants growing in a greenhouse were sectioned (8 mm thickness), fixed in 4% glutaraldehyde and embedded in paraffin for in situ mRNA localization according to our previous method 9 . Stem sections (10 μ m) were cut on a rotary Leica RM 6025 microtome, mounted onto Superfrost Plus glass slides (Thermo Fisher, Waltham, USA) and hybridized with digoxygenin-labeled PdMYB156 and PdMYB221 antisense or sense RNA probes. The hybridization signals were detected by incubating with alkaline phosphatase -conjugated antibodies against Figure 8. Diagram of the transcriptional networks regulating secondary wall biosynthesis in Arabidopsis and poplar. Arabidopsis NAC transcription factors (e.g., SND1, NST1, and VND6) function as the first-level master switches that directly activate the MYB46 and MYB83 targets 20,30,31 , which act as second-level master switches activating MYB4, MYB85 and MYB75 indirectly and in turn control secondary wall biosyntehsis 10,11 . Meanwhile, SND1, NST1, VND6, MYB46, and MYB83 have been shown to activate directly the expression of secondary wall biosynthetic genes. Further, SND1 expression is feedback suppressed by the MYB4 gene 19 . Poplar orthologs (PtrWND and PtrMYB) of these Arabidopsis regulators constitute a complicated network controlling secondary wall biosynthesis similar to that of Arabidopsis. Note that many additional AtMYB4/PdMYB221 -regulated transcription factors are not shown here. Arrows, positive regulation; line with block, negative regulation. Solid and dashed lines separately indicate known and non-verified targets. Question mark indicates a potential regulator relationship but unknown direct targets.
Scientific RepoRts | 5:12240 | DOi: 10.1038/srep12240 digoxigenin and evaluating the subsequent color development with alkaline phosphatase substrates. The images were captured with an Olympus X51 light microscope and processed with Adobe Photoshop v7.0. Subcellular localization of PdMYB221. The PdMYB221 coding region was amplified from stems of P. deltoides cv. nanlin895 and ligated between the CaMV 35S promoter and the 35S terminator in pK7FWG2 (Invitrogen). The resulting construct encoded fusion protein with PdMYB221 located at the C-terminus and GFP at the N-terminus. Infiltration of tobacco leaves with GFP-tagged PdMYB221 was performed as described previously 23 . After a 3-d post-infiltration period, transfected leaves were examined for green fluorescence signal using an Olympus FluoView FV1000 confocal microscope equipped with 488-nm argon laser.
Overexpression of PdMYB221 in Arabidopsis. The full-length PdMYB221 cDNA was ligated downstream of the 35S promoter in the pEARLY100 vector (Invitrogen) to generate the overexpression construct. Five-week-old Arabidopsis plants were used for transformation via Agrobacterium tumefaciens (GV3101) following the floral dip method 36 . T 0 transgenic plants were selected on basta (50 mg/L). At least three homologous T 3 transgenic lines were used for examination of phenotypes.
Microscopy and histochemistry. Basal stems (5-6 mm) of six-week-old Arabidopsis plants were fixed in 2.5% glutaraldehyde in phosphate buffer (pH 7.2), and embedded in resin (SPI-Chem.). Sections (1 μ m thick) were cut with a Leica EM UC6 microtome and stained with 0.05% toluidine blue for light microscopy (Olympus DX51). For transmission electron microscopy, 70-nm-thick sections were cut with a microtome, stained with TBO, and observed using a Hitachi H-7650 electron microscope. Wall thickness was measured in metaxylem vessels and in interfascicular fibers next to the endodermis. At least three transgenic plants were examined.
Stems sections (1 μ m thick) were stained for cellulose with 0.01% Calcofluor White and observed with a UV fluorescence microscope. Under the conditions used, only secondary walls exhibited brilliant fluorescence. Sections (50 μ m thick) of stems were stained with phloroglucinol-HCl for lignin, which was shown as bright red color. For examination of xylan in secondary walls, 1-μ m-thick sections were probed with LM10 monoclonal antibody, which binds to 4-O-methylglucuronoxylan 37 , and detected with fluorescein isothiocyanate-conjugated secondary antibodies. The fluorescence-labeled xylan signals were visualized with an Olympus DX51 light microscope.

Analysis of cell wall composition.
Inflorescence stems of six-week-old Arabidopsis plants were used as samples to analyze cell wall composition following the previous procedure 38 . Briefly, Alcohol insoluble residues (AIR) were prepared by treating the powder sequentially with 80% ethanol, 100% ethanol, and acetone. The resulting AIRs were dried under vacuum at 60 °C overnight and hydrolyzed for 2 h at 120 °C with 2 M trifluroacetic acid (TFA). The TFA-released materials were derivatized with 1-phenyl-3-methyl-5-pyrazolone (PMP) and analyzed on a Thermo ODS-2 C18 column (4.6 × 250 mm) connected to a Waters High Performance Liquid Chromatography (HPLC) system with a 2489 UV visible detector at a 245 nm wavelength. Elution solution A was 0.1 M phosphate buffered (pH = 7.0) and elution solution B was acetonitrile. The PMP derivative (10 μ l) was injected, eluted at 1 ml/min with elution A: elution B = 82:18. To determine the cellulose content 39 , TFA-resistant materials were treated with Updegraff reagent (acetic acid/nitric acid/water, 8:1:2 v/v) at 100 °C for 30 min, and the resulting pellets were then completely hydrolyzed with 67% H 2 SO 4 (v/v). The released glucose was measured using a glucose assay kit (Cayman Chemical) using a dehydration factor of 0.9. To determine the lignin content 40 , 3 mg of AIR samples was solubilized by acetyl bromide solution, and 2 M sodium hydroxide and 0.5 M hydroxylamine hydrochloride was then added to stop the reaction. Absorbance at 280 nm was measured using an UV-visible spectrophotometer (VARIAN Cary 50).

Protoplast transformation and transient expression assays. The transcription activity assay was
carried out in the transient-transformed protoplast 24 . The DNA binding domain (BD) from GAL4 was used in the system. The GAL4 BD-PdMYB221 fusion protein can bind to the GAL4 DNA-binding sites of the GUS reporter. Two known proteins HOS15, a transcription suppressor, and ARF5M, an activator, were used as the controls. The GUS reporter containing four upstream GAL4 DNA-binding sites (GAL4 (4X)-D1-3(4X)-GUS) and the luciferase (LUC) reporter were cotransformed with GAL4 BD-PdMYB221 into Arabidopsis protoplasts. The LUC reporter carrying the luciferase gene under control of CaMV 35S promoter was used as the internal control to normalize the data for eliminating variations in the experiment.
For detecting the repressing activity of PdMYB221 to five tested promoters, the promoters of PdCESA7 (1948 bp), PdCESA8 (1446 bp), PdGT47C (2021 bp), PdCCR1 (1758 bp) and PdCOMT2 (1219 bp) were individually ligated upstream of the GUS reporter after removing the 35S promoter in pBI221 to create the reporter constructs. The PdMYB221 coding region was ligated between the 35S promoter and the NOS terminator after removing GUS from the pBI221 vector to create the effector construct. In each experiment, the expression level of the GUS reporter in the protoplasts transfected with the reporter construct alone was used as the control. The data presented are the means of three biological replications.
Scientific RepoRts | 5:12240 | DOi: 10.1038/srep12240 Electrophoretic mobility shift assays (EMSA). The PdMYB221 coding region was fused in frame with HIS in the pET28a vector and expressed in Escherichia coli. The recombinant protein was purified using Ni-NTA Spin Kit (QIAGEN, Germany). The PdCESA8, PdGT47C, and PdCOMT2 promoter fragments used as the probes were amplified with primers labeled with biotin at the 5′ end (Table S2). The EMSA assay was conducted with a LightShift ® Chemiluminescent EMSA Kit (Thermo Fisher) according to the manufacturer's instructions. Briefly, the labeled DNA fragments were incubated for 30 min with 100 ng of the recombinant protein in binding buffer (10 mM Tris, pH 7.5, 50 mM KCL, 1 mM DTT, 2.5% glycerol, 5 mM MgCl2, 0.05% Nonidet P-40 and 100 ng/μ l poly (dl-dC)). The PdMYB221-bound DNA fragments were separated from the unbound fragments by polyacrylamide gel electrophoresis. The DNA was electroblotted onto a nitrocellulose membrane and detected by chemiluminescence.
Statistical analysis. The data in the experiments of measurement of cell wall thickness and composition were subjected to statistical analysis using the Student's t test. The quantitative differences between two groups of data for comparison in all these experiments were shown to be statistically significant ( * P < 0.05; ** P < 0.01).