EgMYB1, an R2R3 MYB transcription factor from eucalyptus negatively regulates secondary cell wall formation in Arabidopsis and poplar


Author for correspondence:
J. Grima-Pettenati
Tel: +33 562 193513


  • The eucalyptus R2R3 transcription factor, EgMYB1 contains an active repressor motif in the regulatory domain of the predicted protein. It is preferentially expressed in differentiating xylem and is capable of repressing the transcription of two key lignin genes in vivo.
  • In order to investigate in planta the role of this putative transcriptional repressor of the lignin biosynthetic pathway, we overexpressed the EgMYB1 gene in Arabidopsis and poplar.
  • Expression of EgMYB1 produced similar phenotypes in both species, with stronger effects in transgenic Arabidopsis plants than in poplar. Vascular development was altered in overexpressors showing fewer lignified fibres (in phloem and interfascicular zones in poplar and Arabidopsis, respectively) and reduced secondary wall thickening. Klason lignin content was moderately but significantly reduced in both species. Decreased transcript accumulation was observed for genes involved in the biosynthesis of lignins, cellulose and xylan, the three main polymers of secondary cell walls. Transcriptomic profiles of transgenic poplars were reminiscent of those reported when lignin biosynthetic genes are disrupted.
  • Together, these results strongly suggest that EgMYB1 is a repressor of secondary wall formation and provide new opportunities to dissect the transcriptional regulation of secondary wall biosynthesis.


Plant scientists have devoted considerable efforts to understanding the molecular bases of wood formation with the aim of increasing the diverse industrial uses of this important bioresource (pulp and paper, lumber, biofuels and others) by modifying the chemical composition of the wood cell walls (for reviews see Boerjan et al., 2003; Mellerowicz & Sundberg, 2008; Vanholme et al., 2008; Weng et al., 2008). Wood is a complex vascular tissue containing specialized cells types such as tracheary elements, fibres and tracheids derived from a secondary meristem called the vascular cambium (Plomion et al., 2001; Ye, 2002). These cells form thick secondary cell walls (SWs) typically composed of three main polymers: cellulose, hemicelluloses and lignin (Plomion et al., 2001; Boerjan et al., 2003; Mellerowicz & Sundberg, 2008). Formation and deposition of these components in SWs require fine temporal and spatial regulation. Transcriptional regulators are expected to precisely coordinate the expression of hundreds of genes participating in this process (Sivadon & Grima-Pettenati, 2004; Demura & Fukuda, 2007; Zhong & Ye, 2007).

Targeted transcriptome profiling experiments performed in Arabidopsis and in poplar (Hertzberg et al., 2001; Oh et al., 2003; Schrader et al., 2004; Ehlting et al., 2005; Kubo et al., 2005) have highlighted the stage-specific transcriptional regulation of lignin and cellulose biosynthetic genes during development and the importance of transcriptional regulators such as members of the R2R3 MYB, bHLH, bZIP, HD-ZIP and NAC-domain transcription factors families (NACs). Functional analyses of some of these transcription factors have provided insights into the complex network of transcriptional regulatory pathways involved in SW biosynthesis (reviewed in Baucher et al., 2007; Demura & Fukuda, 2007; Zhong & Ye, 2007; Zhong et al., 2008). Recently, a hierarchical network of transcription factors has been proposed to control SW formation in Arabidopsis. In this network SECONDARY WALL-ASSOCIATED NAC DOMAIN 1 protein (SND1/NST3) and its functional homologues (NST1 and NST2, vessel-specificVND6 and VND7) are master switches that turn on a subset of transcription factors (SND3, MYB46, MYB103 and KNAT7 (knotted1-like homeodomain protein) in different cell types, which, in turn, activate the SW biosynthetic pathways (Zhong et al., 2008). Based on functional analyses carried out in conifers, it was proposed that this pathway may be largely conserved across the entire plant kingdom (Bomal et al., 2008).

In addition to NACs, several MYB transcription factors were shown to be important regulators of SW formation. We have previously shown that the Eucalyptus gunnii EgMYB2 protein is able to bind to the promoters of the EgCCR (cinnamoyl coA reductase) and EgCAD2 (cinnamyl alcohol dehydrogenase) genes and activate their transcription (Goicoechea et al., 2005). Overexpression of EgMYB2 in tobacco was associated with significantly thicker xylem SWs as well as a modified lignin profile consistent with increased transcript abundance of most of the lignin biosynthesis genes. These data suggest that EgMYB2 is a positive regulator for both lignin biosynthesis and SW formation. Similarly, constitutive overexpression of AtMYB46, the putative EgMYB2 orthologue in Arabidopsis, has been shown to be associated with ectopic lignification, SW thickening and activation of lignin and other SW genes (Zhong et al., 2007; Ko et al., 2009). Interestingly, transient transcriptional activation analyses in Arabidopsis protoplasts indicated that AtMYB46 overexpression also activates other MYB factors, thereby underlining the complexity of the regulatory network involved (Ko et al., 2009). Other R2R3 MYB transcription factors were also shown to bind to AC elements and/or regulate the biosynthesis of phenylpropanoid and derived-products including lignins (Patzlaff et al., 2003). More recently, AtMYB58 and AtMYB63 were shown to function as specific transcriptional activators of lignin biosynthesis in Arabidopsis (Zhou et al., 2009).

Together, these reports clearly show that different transcription factors, including MYB proteins are involved in the temporal and spatial control of SW formation. However, the above factors appear to function as transcriptional activators. In the case of a tightly-regulated process such as SW formation it is possible that transcriptional repressors might also be required for precise control of gene expression. There is strong evidence that some MYB factors can act as repressors. For example, the overexpression of the snapdragon AmMYB308 and AmMYB330 genes in tobacco downregulated phenylpropanoid and lignin biosynthetic genes leading to reduction in lignin accumulation (Tamagnone et al., 1998). Similarly, AtMYB4, the proposed Arabidopsis orthologue of AmMYB308, has been shown to downregulate C4H gene expression and a knockout mutant showed increased amounts of sinapate esters, resulting in better tolerance to UV-B (Jin et al., 2000). In maize, it has been proposed that the R2R3 MYB genes ZmMYB31 and ZmMYB42 are negative regulators with complementary roles in lignin and phenylpropanoid metabolism (Fornaléet al., 2006). ZmMYB42 overexpression was shown to affect the cell wall structure, composition and degradability in Arabidopsis (Sonbol et al., 2009).

Eucalyptus trees represent one of the main sources of wood worldwide and are widely used in industrial plantations. In order to learn more about wood formation in this species our group has initiated a programme to identify and characterize regulator genes involved in this process. We identified an R2R3 E. gunnii MYB, EgMYB1 preferentially expressed in E. gunnii xylem (Legay et al., 2007). EgMYB1 recombinant protein was shown to specifically bind the MYB-binding sites present in the cis-regulatory regions of the E. gunnii CCR and CAD promoters. In addition, in vivo transient coexpression in tobacco leaves with either EgCCR::GUS or EgCAD2::GUS reporter constructs showed that EgMYB1 can repress both EgCAD2 and EgCCR promoter activities suggesting that it could act as a negative regulator of lignin biosynthesis gene expression in planta (Legay et al., 2007).

Eucalyptus is still difficult to transform and so we decided to investigate the role of EgMYB1 in planta in two model systems: Arabidopsis thaliana, an herbaceous annual and Populus tremula × Populus alba, a woody perennial. Our findings (this paper) in both model species are consistent with a role of EgMYB1 in transcriptional repression of lignin biosynthesis and SW formation.

Materials and Methods

Plant material and transformation

Transgenic Arabidopsis and poplar plants overexpressing EgMYB1 were produced by transformation with the EgMYB1 OE construct, a pJR1 vector carrying the EgMYB1 cDNA under the control of the CaMV 35S promoter (Legay et al., 2007). Control plants contained the empty pJR1 vector.

Wild-type A. thaliana plants (ecotype Wassilewskija, WS) were transformed by infiltration with Agrobacterium tumefaciens strain GV3101 carrying the pJR1 vectors that were introduced by using the freeze–thaw procedure. T1 transformants were selected by germination on sterile Murashige and Skoog (MS) medium (Murashige & Skoog, 1962) containing kanamycin (50 μg ml−1). Transformation was confirmed by PCR on leaf disks. Expression of the transgene was confirmed by reverse-transcription polymerase chain reaction (RT-PCR) using the SuperScript first-strand synthesis kit for RT-PCR (Invitrogen) performed on RNA extracted from 3-wk-old transgenic seedlings. Phenotypic characterization was performed on homozygous T4 transgenic plants with a single T-DNA insertion locus, as estimated by the Mendelian segregation of the kanamycin-resistant NptII gene. For growth and developmental analyses, plants were germinated and grown in a glasshouse at 23°C under short-day conditions (8 h light : 16 h dark) for 7 wk before transfer to long-day conditions (16 h light : 8 h dark) for 2 wk in order to induce inflorescence stem formation, as previously described (Chaffey et al., 2002). Plants were then transferred back to short-day conditions for a further 2 wk before harvest.

Poplar plants (hybrid poplar clone INRA 717 1B4 (Populus tremula × P. alba) were transformed by co-cultivation (Leple et al., 1992) using hygromycin instead of kanamycin (Vigneault et al., 2007). Seven hygromycin-resistant lines (independent transformation events) accumulating EgMYB1 transcripts (assessed by qRT-PCR) were propagated in vitro by rooted cuttings, acclimatized in a mist chamber for 15 d and transferred to a glasshouse with a photoperiod and average temperatures of 16 h light at 24°C and 8 h dark at 20°C. The plants were grown in the glasshouse for 5 wk in a fully randomized experimental design, with four blocks each containing one of the transgenic lines as well as three control plants. The developmental stage of poplar stem internodes was determined using the leaf plastochron index (LPI; Larson & Isebrands, 1971) and LPI 0 was defined as the first developing leaf that had at least a half-expanded lamina and was > 2 cm.

Microscopy and histochemistry

Short sections (2–3 mm) of Arabidopsis inflorescence stems were fixed (4 h, 1% glutaraldehyde, 4% paraformaldehyde, 50 mM sodium phosphate buffer, pH 7.2), before dehydration in a graded ethanol series and embedding in JB4 resin (Polysciences, Eppelheim, Germany) according to the manufacturer’s instructions. Semithin sections (5–10 μm) were cut with a glass knife on a microtome (RM 6025, Leica, Wetzlar, Germany). Poplar stem sections (3–4 mm, LPI 35) were fixed (48 h, 2% glutaraldehyde, 2% paraformaldehyde, 0.1 M Cacodylate buffer, pH 7.2; 1 mM CaCl2, 1% sucrose), before dehydration in a graded ethanol series. 40 μm-thick sections of samples conserved in 100% ethanol were made using a microtome (Reichert-Jung 2040, Leica, Wetzlar, Germany).

In order to detect lignified cell walls, Arabidopsis and poplar sections were stained with the Weisner reagent (phloroglucinol-HCl), or viewed under UV light. Observations were made using a Nikon Eclipse TS100 light microscope (Arabidopsis, Nikon, Amstelveen, Netherlands), or an Axioskop light microscope (Zeiss, Le Pecq, France) and Pixera Pro 150E5 camera (poplar, Pixera, San Jose, CA, USA). In addition, Arabidopsis sections were stained with Toluidine Blue-O (TBO, 0.5% w : v) in order to observe both lignified and nonlignified tissues.

Cell wall thicknesses were measured using image-pro plus software (MediaCybernetics, Bethesda, MD, USA) on semithin stem sections performed on three to eight independent replicates and stained either with phloroglucinol or TBO. More than 200 measurements were made for Arabidopsis and poplar xylem cell walls. Between 12 (in the particular case of Arabidopsis overexpressors lacking interfascicular fibres) and 50 measurements were made for phloem fibres. For both species, the double cell wall thickness was measured and then halved to obtain an estimate of single cell wall thickness for a particular cell type. In all cases, pairs of similar cell types (e.g. vessel and vessel, fibre and fibre) were selected.

Lignin extraction and analysis

One- and two-month-old A. thaliana inflorescence stems of two control and two transgenic lines (lines 5 and 9; 20–80 pooled stems per sample) were harvested, ground in liquid nitrogen and then lyophilized. For poplar, the basal part of the stem (LPI25-34) was harvested from two controls and two transgenic lines (lines 1 and 6; 4 biological replicates per sample). After bark removal, the wood tissue was placed separately in a well-aired cupboard and left to dry several days. Dried tissues were processed in a Cyclotec sample mill (FOSS, Hillerød, Denmark) and ground to a fine powder. Klason lignin content was determined in pre-extracted tissues as previously described (Dence, 1992).

RNA isolation

For poplar RNA extraction, wood and bark samples were harvested from each of the poplar genotypes (transgenics and control) based on the developmental stage determined using the LPI from all internodes included between LPI15 and LPI34. Bark was first removed from the stem and immediately frozen in liquid nitrogen. The remainder of the stems (secondary xylem) was collected identically and stored separately. Tissue samples were ground in liquid nitrogen using a mixer mill (MM300 engine; Retsch, Newtown, PA, USA) and stored at −80°C until the extraction. Total RNA was extracted according to Chang et al. (1993). Total RNA from 2 g of Arabidopsis inflorescence stems was isolated from frozen tissues ground manually with sand using the Extract-All kit (Eurobio, Remaining traces of DNA were removed with DNAse I (Invitrogen) and DNAse-treated RNAs were immediately processed on RNA Easy Clean Up columns (Qiagen). RNA quality and quantity was checked with a RNA 600 Nano kit on a BioAnalyser 2100 (Agilent, and a Nanodrop ND-1000 Spectrophotometer (Labtech,

Microarray RNA profiling in poplar

A poplar 3.4 K cDNA microarray was used for analyses, and is described in the Supporting Information Notes S1 and Table S4. Experiments were carried out on bark and wood of three independent biological replicates (line 6 vs control) using two-dye hybridizations with dye-swaps. One microgram of RNA for each sample was amplified using the SuperScript Indirect RNA Amplification System (Invitrogen) and 5 μg of RNA were labelled with Alexa Fluor 555 and 647 dyes (Invitrogen), for use in dye-swap experiments. Microarray slides were prehybridized (2 h at 42°C) in a solution containing 5× standard saline citrate (SSC), 0.1% sodium dodecyl sulphate (SDS), 0.02% BSA (m : v), 0.01% herring sperm DNA (m : v) and 50% formamide. The slides were then washed twice in 0.1× SSC, once in water, rinsed in 2-propanol, and finally dried by centrifugation. The labelled targets (3.5 μl) were mixed with 52.5 μl of hybridization solution containing 5 × SSC, 0.1% SDS, 0.01% herring sperm DNA (m : v) and 50% deionized formamide. The mixture was heat-denatured for 4 min at 95°C and cooled for 5 min on ice before hybridization with the microarray. The microarray was then covered with a LifterSlip (Erie Scientific Company, Portsmouth, NH, USA) and placed in a hybridization chamber (Corning, Lowell, MA, USA) and incubated (12 h at 45°C) in a hybridization oven (Shel Laboratories, Cornelius, OR, USA). After hybridization, the slides were iteratively washed for 15 min in 2× SSC + 0.5% SDS, 0.5× SSC + 0.1% SDS, and 0.1× SSC solutions at 45°C.

Slides were scanned using a ScanArray Express scanner (Packard BioScience, Meriden, CT, USA) and image files were analyzed using QuantArray software (Packard Bio-Science). Scan intensities were comparable between each set of slides for a given hybridization. Data analysis were carried out using bioconductor packages ( in r (Venables et al., 2008). Data processing, quality assessment, normalization, and statistical analyses were carried out as described by Pavy et al. (2008). Median foreground intensity minus median background intensity was used for the statistical analysis. Briefly, data quality was assessed using the marray and olin packages, and by assessment of within- and between-slide Pearson correlation (r) coefficients calculated both from raw and normalized data. Background-corrected intensities were normalized using a composite method with the functions maNorm2D and maNormLoess in the marray package (Dudoit & Fridlyand, 2002). We identified differentially expressed sequences with the limma package (Release 2.0.7). P-values were adjusted for multiple testing by using the false discovery rate (FDR) approach of Benjamini & Hochberg (1995). Visualization of the data used the mev program from the TM4 suite (Saeed et al., 2003). Data reported in the article are log2 ratios of Alexa Fluor 555/647. Genes were considered as differentially expressed based on P-value < 10−4 and a log2-fold difference > ± 0.5. Similar statistical cut-off criteria (P-values and M) were previously shown to be robust, based on quantitative RT-PCR validation (Pavy et al., 2008).

Quantitative real time RT-PCR

First-strand cDNA synthesis was accomplished using 1 μg of total RNA using Superscript III RetroTranscriptase (Invitrogen). First-strand cDNAs were purified with QIAquick purification kit for PCR products (Qiagen) and eluted in a final volume of 100 μl. Real-time PCR was performed on a ABI 7900HT fast real-time PCR system (Applied Biosystems, Foster City, CA, USA) using the Power SYBR Green PCR Master Mix (Applied Biosystems) following the manufacturer’s procedure. Gene specific primer pairs were designed using the oligo explorer 1.1.0 software ( and are provided as (Table S2). AtACT2 (actin2) and PtaCDC2 (cell division cycle 2) genes served as reference genes for A. thaliana and poplar gene expression studies, respectively. The specificity of primer annealing was examined by monitoring the dissociation curve when real-time PCR reactions were completed. Quantitative RT-PCR efficiencies were calculated using the linregpcr V7.5 program (Ramakers et al., 2003). Quantification of gene expression relative to reference genes and calibrator tissues was determined using the 2−ΔCT method (Pfaffl, 2001).


EgMYB1 overexpression affects plant growth and development similarly in Arabidopsis and poplar

In order to gain insights into the biological role of EgMYB1, we overexpressed EgMYB1 under the control of the 35S CaMV promoter in two model plants: the model plant A. thaliana (ecotype WS) and the woody model plant poplar (P. tremula × alba), clone 717-1B4. Twenty transgenic Arabidopsis and 20 transgenic poplar lines were selected by PCR and the steady-state EgMYB1 mRNA levels assessed in each independent line (data not shown). Lines showing the highest EgMYB1 transcript levels (EgMYB1 OE) were retained for further study.

When compared with control plants (transformed by the vector alone), EgMYB1 OE Arabidopsis seeds germinated 2–3 d later than control seeds under identical conditions. All EgMYB1 OE plants showed shorter inflorescences (Fig. 1a,b), smaller rosettes (Fig. 1c) and rosette leaves, as well as altered leaf morphology (Fig. 1d). The height difference between control and transgenic inflorescence stems was most noticeable at 1-month (Fig. 2a) but disappeared after 2 months. The EgMYB1 OE plants developed thinner inflorescence stems (c. 50% reduction) both in the 1-month-old and 2-month-old plants (Fig. 2b).

Figure 1.

 Phenotype of Arabidopsis thaliana (a–d) and poplar (e–g) transgenic lines overexpressing EgMYB1. (a) One-month-old Arabidopsis control plant, (b) 1-month-old EgMYB1 OE, (c) leaf rosette control (left) and EgMYB1 OE (right), (d) leaf size and morphology of EgMYB1 OE (left) and control (right). (e) Control (left) and EgMYB1 OE (right) poplar plants after 5-wk growth in glasshouse, (f) leaf sizes at 4 leaf plastochron index (LPI), control (upper) and EgMYB1 OE (lower) photos (g) detail of cup-shaped leaves of EgMYB1 OE poplar plants.

Figure 2.

 Effect of EgMYB1 OE on stem height and diameter growth of Arabidopsis thaliana and poplar transgenic lines. Closed bars, control lines (Arabidopsis and poplar); tinted bars, Arabidopsis EgMyb1 OE; tinted cross-hatched bars, EgMyb1 OE poplar lines. (a) Stem height (inflorescence) of Arabidopsis at 1 month and 2 months, and poplar after 5 wk growth in the glasshouse. (b) Basal diameter of Arabidopsis inflorescences at 1 month and 2 months and of poplar stems after 5 wk growth in the glasshouse. Mean and standard deviations from 4 and 2 lines of Arabidopsis and poplar, respectively. Student t-test: *, < 0.05; **, < 0.005.

Several of the EgMYB1 OE poplar lines developed phenotypes similar to those observed in Arabidopsis, but only after plants were transferred to the glasshouse (no differences were observed during the in vitro growth phase). The most obvious phenotype was a decreased height (Fig. 1e). The transgenic lines with the most severe reduction in height (lines 1 and 6) were 20–25% shorter than the controls throughout the 5 wk of growth in glasshouse (Fig. 2a) and had a decreased diameter of 23%, at 5 wk (Fig. 2b). The leaves were smaller and curled upward in a cup shape (Fig. 1f,g).

EgMYB1 overexpression reduces secondary cell wall thickening and lignification in both Arabidopsis and poplar stems

Examination of EgMYB1 OE Arabidopsis inflorescence stem cross-sections stained with phloroglucinol-HCl revealed profound changes in the lignification pattern (Fig. 3a–d). The absence of red staining in the interfascicular zone of 1-month old EgMYB1 OE plants compared with the control (Fig. 3a,b) indicated the absence of lignified fibres in this zone. In 2-month-old plants, a weak and discontinuous red staining (Fig. 3d) suggested that only very limited lignification had occurred in the interfascicular zone. Secondary xylem tissue in the vascular bundle regions of 2-month-old EgMYB1 OE plants (Fig. 3d) was also substantially reduced or even absent compared with controls (Fig. 3c). The phloem fibres of EgMYB1 OE plants also showed significantly less staining than control plants (Fig. 3c,d).

Figure 3.

 Microscopic analyses of stems from control and EgMYB1 OE Arabidopsis thaliana plants. (a–d) General view of stem vascular tissues stained by phloroglucinol-HCl in basal inflorescence stem transverse sections of: 1-month-old (a) and 2-month-old (c) controls, 1-month-old (b) and 2-month-old (d) EgMYB1 OE plants. (e,f) General view of stem vascular tissues stained by Toluidine blue-O in basal inflorescence stem transverse sections of 2-month-old: control (e) and EgMYB1 OE (f) plants. (g,h) Higher magnification view of primary xylem stained with phloroglucinol-HCl in basal inflorescence stem transverse sections from 1-month-old control (g) and EgMYB1 OE (h) plants. Arrowheads indicate collapsed xylem cells (irx phenotype). Co, cortex; Fi, interfascicular fibres; Pf, phloem fibres; Pi, pith; V, xylem vessels; Xp, primary xylem; Xs, secondary xylem. Bars, (a–d) 20 μm, (e,f) 10 μm, (g,h) 20 μm.

Toluidine blue-O staining of 2-month-old A. thaliana stem cross-sections (Fig. 3e,f) was consistent with the lack of secondary xylem tissue in both interfascicular and fascicular regions. These histochemical results indicated that EgMYB1 overexpression negatively affected not only lignification, but also secondary xylem tissue formation as a whole. Further examination showed collapsed primary xylem vessel elements in cross-sections from 2-month-old EgMYB1 OE inflorescence stems (Fig. 3h). No similar cells were observed in comparable sections from control plants (Fig. 3g), suggesting that EgMYB1 overexpression is associated with reduced mechanical resistance of xylem cell walls.

Quantitative determinations of cell wall thickness in Arabidopsis, illustrated by a frequency graph distribution (Fig. 4), showed a reduction in SW thickening in xylem (Fig. 4a), as well as in phloem fibres (Fig. 4b) of transgenics compared with corresponding controls. On average, cell wall thickness was reduced by 42% and 71%, for xylem and phloem fibres, respectively.

Figure 4.

 Effect of EgMYB1 overexpression on the thickness of xylem and phloem fibre cell walls in Arabidopsis thaliana and poplar. Frequency graphs comparing cell wall thickness distribution (percentage of measured cell walls in different size classes) between controls (open bars) and EgMYB1 overexpressing plants (closed bars) within different classes. Arabidopsis thaliana xylem (a), phloem fibres (b) and poplar xylem (c) and phloem fibres (d). Mean and standard deviations for Arabidopsis cell wall thicknesses were 0.69 ± 0.08 μm (EgMYB1 OE) and 1.2 ± 0.08 μm (control) for xylem and 0.91 ± 1.18 μm (EgMYB1 OE) and 3.13 ± 0.52 μm (control) for phloem fibres. Mean and standard deviations for poplar cell wall thicknesses were 0.9 ± 0.2 μm (EgMYB1 OE) and for 1.8 ± 0.25 μm (control) for xylem and 3.8 ± 0.4 μm (EgMYB1 OE) and 4.7 ± 0.6 μm (control) for fibers. In both species, the difference between the control and transgenic populations is highly significant for all tissues (Student t-test < 0.05).

Histological observations of stem tissues from transgenic and control poplar plants indicated that EgMYB1 overexpression resulted in similar modifications as those observed in A. thaliana. Lignin autofluorescence revealed consistently fewer phloem fibres in EgMYB1 OE stem cross-sections compared with controls (Fig. 5a,b). Both phloem and xylem fibres of EgMYB1 OE poplars appeared to be less heavily lignified than those of controls, as suggested by the weaker phloroglucinol staining (Fig. 5c–f) and the less intense lignin autofluorescence (Fig. 5g,h). On average, xylem and phloem fibre cell walls were 50% and 20% thinner than controls, respectively (Fig. 4c,d).

Figure 5.

 Microscopic analyses of basal stems from EgMYB1 OE poplar plants. (a,b) General view of tissues under UV in transverse sections from basal stem regions of control (a) and EgMYB1 OE (b). Autofluorescence indicates lignified/phenolic-containing tissues. Phloroglucinol-HCl staining of phloem fibres (c,d) and secondary xylem (e,f) from control (c,e) and EgMYB1 OE (d,f) plants. Lignin autofluorescence of control (g) and EgMYB1 OE (h) observed through a confocal microscope. All images are from transverse sections of basal regions of plants after 5 wk growth in glasshouse (leaf plastochron index, LPI = 35). Magnifications: (a, b) ×7.5; (c–h) ×50. Pf, phloem fibers; Xf, xylem fibres; Xv, xylem vessels.

Together, these results indicated that the overexpression of EgMYB1 reduces SW thickening and lignin deposition in both A. thaliana and poplar. The extent of secondary xylem formation was also greatly reduced in Arabidopsis.

In order to quantify lignin modifications, we determined Klason lignin content in the stems of A. thaliana and poplar control and transgenic plants (Table 1). The results show that lignin accumulation (expressed as a percentage of cell wall residue) was significantly decreased (approx. 11%) in stems of both 1-month-old and 2-month-old Arabidopsis plants as well as in poplar wood. For both species, the lignin monomer yield and composition (S : G ratio) determined by thioacidolysis was unchanged (data not shown).

Table 1. Acid-insoluble lignin content (Klason lignin) of stems from overexpressing EgMYB1 Arabidopsis thaliana (1-month-old and 2-month-old plants) and poplar transgenic lines (5 wk growth in glasshouse)
1 month2 months
  1. Measures were done on two Arabidopsis (lines 5, 9) and poplar (lines 1, 6) transgenic lines, respectively. Values presented are mean and standard deviation, all EgMYB1 OE values are significantly different from control values (Student t-test: < 0.01%).

EgMYB1 OE #113.49 ± 0.1813.95 ± 0.0119.46 ± 0.64
EgMYB1 OE #213.78 ± 0.1114.3 ± 0.0320.23 ± 0.21
Control15.35 ± 0.3315.91 ± 0.2922 ± 0.23

EgMYB1 overexpression modifies gene expression in poplar

A custom poplar cDNA microarray corresponding to 3400 unique poplar sequences (Notes S1) was used to compare gene expression profiles in wood and bark tissues from the EgMYB1 OE poplar with those of wild-type plants. Significant differential transcript accumulation was detected for 541 sequences (> 6% of array sequences) in bark, wood or both tissue types. A total of 341 transcripts accumulated differentially (65% upregulated, 35% downregulated) in wood and 260 transcripts accumulated differentially (51% upregulated, 49% downregulated) in bark. Sixty transcripts were differentially expressed (DE; 36 upregulated, 24 downregulated) in both wood and bark. All sequences were classified into functional groups according to GO (Gene Ontology) terms (Notes S1).

The most upregulated transcripts among those common to both tissues encoded a bark storage protein (Fold Change, FC = 3.5), whereas a peroxidase precursor and an unknown sequence were the most downregulated sequences. A prephenate dehydrogenase homologue (chorismate mutase), a rate-limiting enzyme of phenylalanine biosynthesis, was also downregulated in both wood and bark.

More cell wall biogenesis-related transcripts were downregulated in bark than in xylem. For example, all of the sequences related to lignin biosynthesis on the array were downregulated in bark. Phenylalanine ammonia lyase (PAL) and 4-coumarate CoA ligase (4CL) were among the most strongly downregulated sequences (FC = −3.4 for both). Several other transcripts encoding enzymes of the anthocyanin metabolism, including putative dihydroflavonol reductase (DFR), chalcone synthase (CHS) and flavonol-3-O-glucosyltransferase, were also downregulated in bark tissues (Table S1).

Several other DE transcripts encoded putative sugar and starch metabolism enzymes, possibly linked to cell wall carbohydrate metabolism. The major cell wall carbohydrates are synthesized from cellular hexose sugar pools, that is, glucose and fructose. Transcripts encoding starch synthesis enzymes such as the UDP-glucose:protein transglucosylase (UPTG1) and the starch branching enzymes I and II were downregulated in both wood and bark, whereas transcripts coding for starch breakdown enzymes such as β-amylase and β-glucosidase accumulated in the wood. Similarly, transcripts of UDP-glucuronate decarboxylase and UDP-glucose-6-dehydrogenase, two enzymes involved in the production of UDP-xylose (the building blocks of hemicellulose; Bindschedler et al., 2005) were downregulated in the wood tissue. Sequences encoding fructose-1,6-bisphosphatase and fructose-1,6-biphosphate aldolase, involved in the production of fructose-6-phophate, were upregulated in both wood and bark.

Many transcripts encoding putative light-regulated proteins belonging to photosystem I and II or binding to the light harvesting complex Chla,b were upregulated in wood, and to a lesser extent, in bark (Table S1). Few of the DE transcripts encoded transcription factors and only two were found in both tissues and corresponded to a putative basic helix–loop–helix (bHLH) and a basic leucine zipper (bZIP) (Table S1). The ‘no hit’ and unknown classes were highly represented among DE transcripts relative to most other classes, particularly in bark where they accounted for 32% of the DE transcript set.

EgMYB1 overexpression modifies cell-wall-related gene expression in Arabidopsis and poplar

In light of the overlapping SW phenotypes of EgMYB1 OE poplar and Arabidopsis plants, we used quantitative RT-PCR to compare the impact of EgMYB1 overexpression on cell wall related gene expression in both species. In addition, this also allowed us to investigate the effect on genes not represented on the poplar microarray. We specifically targeted genes involved in the biosynthesis of the three major SW polymers, that is, lignin, cellulose, xylan with gene specific primers. For multigenic families, isoforms putatively implicated in wood formation were selected (Table S2). Our results (Fig. 6a) show that EgMYB1 overexpression in poplar bark downregulates all genes except PtaC4H1 and PtaCesA3. In poplar wood, the cellulose and xylan genes tested were downregulated, as well as the following lignin genes: PtaPAL2, PtaC4H2, PtaC3H1, PtaF5H, PtaCCoAOMT1, PtaCCR2 and PtaCAD1. These findings are in good agreement with the poplar microarray data, and with the higher accumulation of EgMYB1 transcripts in bark compared with wood (data not shown). In Arabidopsis, EgMYB1 overexpression downregulated all cell wall related genes tested to a greater extent than in poplar (Fig. 6b). In both species, the most heavily downregulated gene was PAL2 coding the phenylpropanoid entry-point enzyme.

Figure 6.

 Gene expression analyses of cell wall related genes in overexpressing EgMYB1 poplar (a) and Arabidopsis thaliana (b) transgenic plants. Transcript accumulation was quantified by quantitative real-time polymerase chain reaction (qRT-PCR). Relative expression levels (± SD) were calculated using PtaCDC2 and AtACT2 as references for poplar and Arabidopsis, respectively. Wild-type pJR1 transgenic plants were used as calibrator. ‘*’ indicate gene expression ratios significantly different from 1 (Student t-test, < 0.05). (a) Poplar data were obtained on three individual biological replicates and two RT-PCR replicates from either wood (light tinted bars) or bark (dark tinted bars) of EgMYB1 OE transgenic line 6. (b) Arabidopsis data were obtained from three RT-PCR replicates from inflorescence stems undergoing secondary xylem formation. Open bars, EgMYB1 OE (line 5); closed bars, EgMYB1 OE (line 9). Key: Populus tremula × alba (Pta), Arabidopsis thaliana (At), cellulose synthase (CesA), glycosyl transferase 8 (GT8D) Fragile fibre 8 (FRA8), Irregular xylem (IRX), phenylalanine ammonia lyase (PAL), trans-cinnamate 4-hydroxylase (C4H), 4-coumarate:coA ligase (4CL), hydroxycinnamoyl-coA:shikimate/quinate hydroxycinnamoyltransferase (HCT), p-coumarate 3-hydroxylase (C3H), ferulate 5-hydroxylase (F5H), caffeoyl-coA 3-O-methyltransferase (CCoAOMT), caffeic acid O-methyltransferase (COMT), cinnamoyl-coA reductase (CCR), Cinnamyl alcohol dehydrogenase (CAD). Accession numbers of all genes can be found in the Supporting Information, Table S2.

Our results led us to hypothesize that EgMYB1 could function as a negative regulator of lignin biosynthesis and the expression of other cell-wall associated genes. In silico analyses revealed that the promoter regions of the great majority of these candidate genes contained MYB binding sites (MBSIIG) (Table S3).


Wood is the most abundant biomass on earth. It is mainly composed of SWs whose biosynthesis requires the coordinated transcriptional regulation of hundreds of genes to synthesize the main SW polymers, that is, cellulose, xylan and lignins. Although the SW biosynthetic pathway has been well characterized biochemically and genetically, our current knowledge of the signals and transcriptional regulators that are responsible for controlling the SW biosynthetic program is still limited. In previous studies, we have functionally characterized the promoters of two key lignin genes (EgCCR and EgCAD2) from eucalyptus (Lacombe et al., 2000; Lauvergeat et al., 2002) and provided functional evidence showing that the coordinated vascular expression of these two genes is mediated through MYB transcription factors (Rahantamalala et al., 2010).

We have also shown that EgMYB1, an R2R3 MYB transcription factor is preferentially expressed in secondary xylem from stems and roots of eucalyptus and binds specifically to the MBSIIG sites located in the promoters of the EgCAD2 and EgCCR lignin biosynthetic genes. The EgMYB1 protein sequence harbours an active repressor motif in the regulatory domain and is able to repress the activity of both EgCCR and EgCAD2 promoters in vivo, as shown by transient assays performed in tobacco leaves (Legay et al., 2007). These data led us to hypothesize that EgMYB1 acts as a transcriptional repressor of lignin biosynthesis genes. In order to test this hypothesis and to further investigate the role of EgMYB1 we compared the effects of EgMYB1 overexpression in both Arabidopsis and poplar. The data reported here support the hypothesis that EgMYB1 acts as a negative regulator of SW formation in vascular tissue, the first example to our knowledge of such a repressor in highly lignified woody tissues.

EgMYB1 overexpression in both Arabidopsis and poplar affected growth and development by reducing stem height and diameter, leaf size and altering leaf shape. It also reduced lignin content by a modest but consistent percentage of total SW weight. Similar modifications in growth and development have been observed in different plant species downregulated for lignin biosynthetic genes. For example, PAL down-regulated tobacco plants had reduced height, cup-shaped leaves and decreased lignin content (Elkind et al., 1990). Similarly, CCR downregulated plants were shorter, had smaller leaves with altered morphology and contained less acid-insoluble lignin (Piquemal et al., 1998; Goujon et al., 2003; Leple et al., 2007). These observations suggest that changes in lignin content/phenylpropanoid metabolism can often have an important effect on plant growth and development.

Histochemical analyses indicated that stem vascular tissue SWs were significantly thinner in both Arabidopsis and poplar EgMYB1-overexpressing plants. In agreement with these observations, quantitative RT-PCR analysis in both species indicated that EgMYB1 overexpression downregulated three different SW CesA genes (CesA4/IRX5, CesA7/IRX3, CesA8/IRX1: Taylor et al., 2004) and two glycosyltransferases (IRX8: Pena et al., 2007; IRX7/FRA8: Zhong et al., 2005) associated with xylan biosynthesis, in addition to a number of lignin biosynthetic genes. The negative impact on overall SW thickening could explain the apparent discrepancy between the moderate decrease in Klason lignin (11% relative to cell wall weight) and the marked effect on lignin staining visualized by histochemistry (particularly dramatic in Arabidopsis interfascicular zones). Together, these results could suggest that EgMYB1 functions as a negative regulator of the SW developmental program and not just as a negative regulator of lignification.

Interestingly, we recently reported (Goicoechea et al., 2005) that overexpression of the transcriptional activator EgMYB2 in tobacco plants had the opposite effect (i.e. thicker SW and slight increase of lignin content). Similar observations have been made for AtMYB46, the closest Arabidopsis orthologue of EgMYB2. Whereas dominant repression of AtMYB46 caused a drastic reduction of SW thickening in fibres and vessels, its overexpression led to an activation of the biosynthetic pathways of lignin as well as those of cellulose and xylan (Zhong et al., 2007; Ko et al., 2009). Thus AtMYB46 (and most likely EgMYB2) is considered as a key master switch activating the SW developmental programme through coordinated regulation of the biosynthetic pathways of all three major SW components (Zhong et al., 2007; Ko et al., 2009). The fact that EgMYB1 overexpression resulted in the downregulation of SW genes in both Arabidopsis and poplar suggests that it could also play a central role in SW formation in higher plants, at least in dicots. Interestingly, the dominant repression form used for AtMYB46 by Zhong et al. (2007) corresponding to a fusion between AtMYB46 and the EAR domain is structurally quite similar to EgMYB1 (R2R3-MYB containing an EAR domain).

AtMYB46 was shown to be a direct target of a SW NAC protein, SND1, a master switch which turns on downstream direct target genes SND3, MYB46, MYB103 and KNAT7 in different cell types, which in turn activate the SW biosynthetic pathway. Whether EgMYB1 could be part of a similar SND1-mediated transcriptional network regulating SW synthesis is a testable hypothesis as potential eucalyptus SND1 orthologues were recently identified (Rengel et al., 2009). The very recent release of the Eucalyptus grandis genome will provide access to EgMYB1 promoter sequences and should help elucidate the position of EgMYB1 in a putative regulatory network. Identifying direct targets of EgMYB1 is a tangible goal which could take advantage of experimental systems such as the steroid-receptor based inducible systems described by Zhong et al. (2008) in combination with deep sequencing (Ko et al., 2009) and may help uncover other transcription factors involved in SW formation.

As two important MYB regulators playing apparently opposite roles in SW biogenesis and lignin biosynthesis are both preferentially expressed in eucalyptus xylem, it is possible that dynamic competition between EgMYB1 and EgMYB2 for the same promoters will allow the formation of either a repressing or an activating regulatory complex, thereby providing a sophisticated mechanism for the spatial and temporal control of lignified SW formation. Consistent with this idea, we have recently shown that the EgCAD2 promoter has a complex organization with its cis-elements arranged in two similar modules (containing MYB sites) suggesting that redundancy and mechanisms of cooperation or competition between cis-elements and trans-acting factors might be involved in the regulation of promoter activity (Rahantamalala et al., 2010). In particular, EMSAs with recombinant EgMYB2 and transactivation assays clearly demonstrated that the two MYB sites of the EgCAD2 promoter cooperated for binding and activation and it is possible that one of these sites could bind a repressor MYB factor (Rahantamalala et al., 2010).

The activity of transcription factors can be regulated at the transcriptional or post-transcriptional level, and may involve combinatorial action with other transcription factors. It is known that bHLH proteins are capable of interacting with MYBs (Zimmermann et al., 2004) and the fact that the EgMYB1 sequence contains a conserved motif for bHLH interaction suggests that MYB–bHLH interaction might be necessary for the control of SW synthesis in xylem. Indirect evidence in support of this hypothesis include the preferential accumulation of several bHLH transcripts in eucalyptus xylem, and the close association of putative bHLH binding sites sequences and MYB binding sites in the promoters of many poplar lignin biosynthetic genes (Legay et al., 2007; Table S3).

A number of reports in different species have shown that modifying the expression of a single lignin biosynthesis gene affects the expression of many other genes in various unanticipated pathways, thereby revealing previously unsuspected and uncharacterized interactions between monolignol and other metabolic pathways (Rohde et al., 2004; Sibout et al., 2005; Dauwe et al., 2007; Leple et al., 2007). As EgMYB1 overexpression in poplar also downregulated a number of lignin biosynthetic genes, we decided to analyse the poplar expression profiles in order to investigate the potential interactions between lignin and other metabolic pathways. In agreement with results obtained on CCR-/CAD-downregulated tobacco (Dauwe et al., 2007) transcript accumulation in EgMYB1-overexpressing poplar was altered for genes associated with phenylpropanoid-, starch- and hemicellulose-metabolisms, stress metabolism and light-regulated genes. Of interest was the downregulation of the gene encoding prephenate dehydrogenase/chorismate mutase, involved in the synthesis of phenylalanine and modulation of carbon flux into the phenylpropanoid pathway. However, further work is necessary to determine whether this gene was downregulated because it is a direct target of EgMYB1 or because of indirect feedback mechanisms resulting from the general repression of lignin biosynthesis, as observed in CCR downregulated tobacco (Dauwe et al., 2007).

In conclusion, this study shows that EgMYB1 overexpression leads to consistent molecular phenotypes in both a woody perennial and a herbaceous annual species and provides evidence for the biological role of EgMYB1 as a negative regulator of lignification and SW formation. Based on these new findings, as well as on previous studies, we propose that the combinatorial control of gene expression through the action of positive regulators (such as EgMYB2, Goicoechea et al., 2005) and negative regulators (such as EgMYB1) could provide the necessary flexibility to ensure tight temporal and spatial regulation of SW biosynthesis in vascular tissues. The identification of a repressor of SW development opens new avenues to dissect the complex networks of transcriptional regulators involved in SW biosynthesis, thereby contributing to a better understanding of SW formation in plants, and in particular in woody species of major economic importance.


The authors would like to thank Dr C. Paule (University of Minnesota) for bioinformatics assistance, Dr J. E. K. Cooke (University of Alberta) for EST library and microarray development, Dr D. Tessier and Ms T. Rigby (Biotechnology Research Institute, National Research Council, Montreal QC, Canada) for production of the 3.4K microarray, and J. Stott, Dr R. Holt and Dr M. Marra (Genome Sciences Centre, Vancouver BC, Canada) for sequencing of the poplar expressed sequence tags, Y. Martinez and P. Panegos (UMR 5546) for their help with microscopy, C. Brière (UMR 5546) for help with statistical analyses. SL was supported by grants from the French Ministère de l’Education Nationale, de l’Enseignement Supérieur et de la Recherche and the Arborea project (Genome Canada and Genome Québec) AB gratefully acknowledges the Syrian government for financial support during this work.