Notice: Wiley Online Library will be unavailable on Saturday 27th February from 09:00-14:00 GMT / 04:00-09:00 EST / 17:00-22:00 SGT for essential maintenance. Apologies for the inconvenience.
Wood harvested from trees is one of the most widely utilized natural materials on our planet. Recent environmental issues have prompted an increase in the demand for wood, especially as a cost-effective and renewable resource for industry and energy, so it is important to understand the process of wood formation. In the present study, we focused on poplar (Populus trichocarpa) NAC domain protein genes which are homologous to well-known Arabidopsis transcription factors regulating the differentiation of xylem vessels and fiber cells. From phylogenetic analysis, we isolated 16 poplar NAC domain protein genes, and named them PtVNS (VND-, NST/SND- and SMB-related proteins) genes. Expression analysis revealed that 12 PtVNS (also called PtrWND) genes including both VND and NST groups were expressed in developing xylem tissue and phloem fiber, whereas in primary xylem vessels, only PtVNS/PtrWND genes of the VND group were expressed. By using the post-translational induction system of Arabidopsis VND7, a master regulator of xylem vessel element differentiation, many poplar genes functioning in xylem vessel differentiation downstream from NAC domain protein genes were identified. Transient expression assays showed the variation in PtVNS/PtrWND transactivation activity toward downstream genes, even between duplicate gene pairs. Furthermore, overexpression of PtVNS/PtrWND genes induced ectopic secondary wall thickening in poplar leaves as well as in Arabidopsis seedlings with different levels of induction efficiency according to the gene. These results suggest that wood formation in poplar is regulated by cooperative functions of the NAC domain proteins.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
For centuries, wood (secondary xylem) has been used for pulp and timber production and the manufacture of secondary wood products. Today, woody biomass is also expected to be utilized as a sustainable and carbon-neutral resource for bioenergy. Thus, a thorough understanding of wood formation (xylogenesis) has become very important for improving the quantity and quality of wood produced. The formation of secondary xylem is a well-ordered developmental process that includes cell division, cell expansion, secondary wall deposition, lignification and programmed cell death (Plomion et al., 2001). Moreover, in the case of angiosperm trees, the wood consists of several types of cell, including vessel elements, fibers and parenchyma (Evert, 2006), which makes it difficult to elucidate the molecular regulation of wood formation.
Considerable effort has been devoted to identifying the regulatory genes for wood formation. In particular, recent advances in large-scale expressed sequence tag (EST) sequencing (for loblolly pine: Allona et al., 1998 and Lorenz and Dean, 2002; for poplar: Sterky et al., 1998; for spruce: Ralph et al., 2008; for eucalyptus: Rengel et al., 2009) and in global gene expression analysis using microarrays (Hertzberg et al., 2010; Whetten et al., 2001) have greatly expanded our understanding of gene expression profiles related to wood formation. The obtained sequence information, including the genomic sequence of poplar (Populus trichocarpa; Tuskan et al., 2006), has accelerated the identification of genes involved in the biosynthesis of cellulose, xylan, glucomannan, and lignin in trees (Boerjan et al., 2003; Mellerowicz and Sundberg, 2008). In addition, comparative genomics has revealed that the xylem transcriptomes of vascular plants are more highly conserved than the total transcriptomes, suggesting the existence of a common ancestral xylem transcriptome (Li et al., 2010). These findings seem to point to evolutionarily conserved regulatory mechanisms for transcriptomes.
A major portion of the transcriptomes would be directly governed by the functions of transcription factors. Among the transcription factors, extensive research has focused on R2R3-MYBs involved in wood formation, especially in lignin biosynthesis. The accumulated data indicate coordinated regulation of lignin biosynthesis genes by MYB proteins, probably by acting on AC elements that are found in the promoters of many lignin biosynthesis genes (Rogers and Campbell, 2004). The involvement of MYB proteins in the regulation of lignin biosynthesis genes has been reported in a wide range of plant species, especially in such trees as eucalyptus (Eucalyptus gunnii; Goicoechea et al., 2005), loblolly pine (Pinus taeda; Patzlaff et al., 2003), white spruce (Picea glauca; Bedon et al., 2007) and poplar (Wilkins et al., 2009; McCarthy et al., 2010), suggesting that the functions of specific transcription factors have been conserved in the control of gene expression for wood formation.
We have been studying Arabidopsis (Arabidopsis thaliana) NAC domain protein genes, VASCULAR-RELATED NAC-DOMAIN1 (VND1)-VND7, which were isolated as upregulated transcription factor genes during transdifferentiation into tracheary elements in an induction system using Arabidopsis suspension-cultured cells (Kubo et al., 2005). In particular, VND7 is considered to play a crucial role in xylem vessel formation, since it is expressed in all types of differentiating vessels and its overexpression effectively induced the transdifferentiation of various cells into xylem vessel cells (Kubo et al., 2005; Yamaguchi et al., 2008, 2010a). In the Arabidopsis genome, the Class IIB NAC transcription factor comprises the VND family and two other gene families (Jensen et al., 2010). One family includes NAC SECONDARY WALL THICKENING PROMOTING FACTOR1 (NST1), NST2 and NST3/SECONDARY WALL-ASSOCIATED NAC DOMAIN PROTEIN1 (SND1), and the other contains the SOMBRERO (SMB), BEARSKIN1 (BRN1) and BRN2 genes. NST1 and NST3/SND1 are the key differentiation regulators of Arabidopsis fiber cells (Zhong et al., 2006, 2007; Mitsuda et al., 2007). SMB, BRN1 and BRN2 are involved in regulating the cellular maturation of root cap (Willemsen et al., 2008; Bennett et al., 2010). Interestingly, ectopic secondary cell wall thickening is induced by the overexpression of these Class IIB NAC transcription factor genes (Kubo et al., 2005; Mitsuda et al., 2005, 2007; Zhong et al., 2007; Bennett et al., 2010; Yamaguchi and Demura, 2010; Yamaguchi et al., 2010a). Thus, these genes could evolve from a common ancestral gene having the capacity to regulate secondary cell wall thickening (Zhong et al., 2010c). This led us to assume that similar NAC domain proteins also play important roles in wood formation.
Here, we focus on 16 P. trichocarpa NAC domain protein genes that we named as PtVNS (P. trichocarpa VND-, NST/SND-, SMB-related proteins) genes, which are homologous to the VND, NST and SMB genes of Arabidopsis. Recently, Zhong et al. (2010a) reported on the functional analysis of some of them, which were named the PtrWND (for wood-associated NAC domain transcription factor) genes, and revealed that some PtVNS/PtrWND genes are expressed in the developing xylem and activate several genes expected to be associated with wood formation. Complementation analysis using the Arabidopsis nst1 nst3 double mutant showed that some of the poplar NAC proteins can activate the entire secondary wall biosynthesis program in Arabidopsis, suggesting a conserved function of NAC proteins for regulating secondary cell wall thickening (Zhong et al., 2010a,c). In the present paper, we report a detailed analysis of all the poplar PtVNS/PtrWND genes, demonstrating the variation in PtVNS/PtrWND expression and activity, which suggests coordinated regulation of wood formation by multiple NAC domain proteins in poplar.
Results and Discussion
Twelve PtVNS genes expressed in xylem tissue and shoots
The genomic information about P. trichocarpa (Tuskan et al., 2006) allowed us to identify poplar NAC domain protein genes homologous in sequence to the VND, NST and SMB genes of Arabidopsis. Recently, 163 NAC domain protein genes have been identified via a comprehensive informatics study (Hu et al., 2010). We isolated 16 poplar genes encoding proteins homologous to AtVND, AtNST and AtSMB among others, and some of them have been reported as 12 PtrWND genes (Table S1 in Supporting Information; Mitsuda et al., 2007; Zhong et al., 2010a; Hu et al., 2010). Phylogenetic tree analysis placed them into three distinct groups: the VND group, the NST group and the SMB group (Figure 1a; Mitsuda et al., 2007; Zhong et al., 2010a; Hu et al., 2010). Based on this result, here we propose a revised nomenclature for this NAC domain protein family, the VNS (VND-, NST/SND-, SMB-related proteins) family, which is applicable to genes homologous to Arabidopsis VND, NST and SMB genes with different functions from many plant species. Then we named the poplar NAC domain protein genes PtVNS and numbered them according to their location in the phylogenetic tree: PtVNS01–08 are included in the VND group, PtVNS09–12 are in the NST group, and the remainder are in the SMB group (for details on the gene identities see Table S1). Following the idea of multiple gene duplication processes, including a whole-genome duplication event in the Populus lineage (Tuskan et al., 2006), the PtVNS/PtrWND genes were arrayed in pairs (Figure 1a; Zhong et al., 2010a). The branch shape of the VND group genes appeared to be complex, although the other groups showed a simple correspondence relationship between the Arabidopsis and poplar genes. Despite this complexity, AtVND7, PtVNS07/PtrWND6A and PtVNS08/PtrWND6B formed a well-separated branch (bootstrap value of 100%) in the VND group, which suggests an evolutionarily conserved specificity, and probable importance, of the functions of AtVND7, PtVNS07/PtrWND6A and PtVNS08/PtrWND6B (Figure 1a).
Expression analysis revealed that the PtVNS01-12/PtrWND genes are expressed in xylem tissues; however, the expression level differs according to the gene. In contrast, the expression of PtVNS13-16, which are included in the SMB group, was not observed in xylem (Figure 1b; Zhong et al., 2010a). The PtVNS01-12/PtrWND genes are also expressed in shoot, although several PtVNS/PtrWND genes, such as PtVNS04/PtrWND4B, PtVNS05/PtrWND3A, PtVNS09/PtrWND2A and PtVNS10/PtrWND2B, showed a lower level of expression compared with that in xylem tissues (Figure 1b). Since AtSMB, AtBRN1 and AtBRN2 genes are known to be expressed and function in the root tip region (Willemsen et al., 2008; Bennett et al., 2010), it is supposed that PtVNS genes of the SMB group are expressed and function in the root tissue (Zhong et al., 2010a). As our interest is in wood formation, we further examined the PtVNS01-12/PtrWND genes. Detailed expression analysis of specific tissues isolated from the primary shoots showed that the expression pattern was similar between paired genes, such as PtVNS01/PtrWND5A and PtVNS02/PtrWND5B; however, expression of PtVNS03/PtrWND4A and PtVNS04/PtrWND4B differed depending on the organ (Figure 1b,c). This might be an indication of divergence in expression between the paired genes after gene duplication.
Different expression patterns by gene groups were detected in primary xylem vessels, but not in xylem tissues
In order to obtain more information about gene expression related to xylem tissue, in situ hybridization analysis was carried out at various stages of development of the stems. The results for the PtVNS07/PtrWND6A and PtVNS11/PtrWND1B genes, which are most similar to AtVND7 and AtNST3/SND1, respectively, indicated that after the secondary growth both genes are expressed in developing xylem tissue including vessels and fibers, and in the phloem fiber cells (Figure 2e,f,h,i; Zhong et al., 2010a). Similar patterns were detected using reporter genes, PtVNS/PtrWND promoter::GUS chimeric genes (Figure S1a). These results suggest that PtVNS/PtrWND genes of the VND and NST groups contribute to the formation of xylem tissue and phloem fiber.
Notably, we found the difference in gene expression in the early developmental stages of stems, i.e. before secondary growth. PtVNS07/PtrWND6A expression was clearly observed in primary xylem vessels (Figure 2d; black arrows), although there was no apparent signal of PtVNS11/PtrWND1B in primary xylem vessels (Figure 2g; black arrowheads). Similarly, PtVNS07/PtrWND6A was expressed in the primary xylem vessels of petioles, in which no secondary growth occurs, whereas expression of PtVNS11/PtrWND1B was not detected in primary xylem (Figure 3b,c). We confirmed the same patterns for other PtVNS/PtrWND genes (Figure S1b), suggesting that only PtVNS/PtrWND genes of the VND group function in primary xylem vessel formation. In the inflorescence stems of Arabidopsis, AtVND7 is expressed in the xylem vessels and AtNST3/SND1 is preferentially expressed in the fiber cells (Zhong et al., 2006; Mitsuda et al., 2007; Yamaguchi et al., 2010b). Our findings of PtVNS/PtrWND expression in primary xylem vessels would parallel the different expression patterns of AtVND and AtNST genes in the inflorescence stems.
The accumulation of PtVNS/PtrWND mRNA was also detected not only in xylem tissues but also in parenchyma cells of stems (Figures 2 and 3). The results of RT-PCR analysis showed that PtVNS01–12/PtrWND genes are expressed in shoot apices and leaves, where no secondary growth occurs (Figure 1). This suggested the involvement of PtVNS/PtrWND genes in biological processes other than xylem formation.
Isolation of poplar genes involved in xylem vessel formation by post-translational induction of AtVND7
We previously reported that the overexpression of AtVND6 or AtVND7 induces xylem vessel transdifferentiation not only in Arabidopsis but also in poplar (Kubo et al., 2005; Yamaguchi et al., 2010a). In particular, the post-translational induction system of AtVND7 should provide useful information about the poplar genes involved in xylem vessel formation, because of its effective and synchronous induction (Yamaguchi et al., 2010a). When AtVND7 function was induced by the application of dexamethasone (DEX), a strong synthetic glucocorticoid, in the transgenic 35S::VND7:VP16:GR poplar, transdifferentiation of xylem vessel elements was observed in the region around the leaf veins (Figure S2; Yamaguchi et al., 2010a). Using this system, we performed microarray analysis to isolate poplar genes functioning in xylem vessel differentiation, especially downstream from NAC domain proteins.
Table 1. Genes upregulated by the induction of AtVND7 in poplar
Protein ID (JGI)
At homologous gene
In a case where several probes corresponding to the same gene were found in the list of probes upregulated by more than 10-fold (Table S2), only the most highly increased probe was selected for entry in this table. For computing the fold changes between the dexamethasone-treated sample and the control, the normalized signal values were used.
aProtein IDs shown with superscript represent the genes used for transient expression assays.
Cysteine proteinase (XCP1)
5.49 × 10−4
Cysteine proteinase (XCP1)
Cysteine proteinase (XCP2)
4.55 × 10−5
Cysteine peptidase (XCP2)
6.36 × 10−5
Papain-like cysteine protease
3.63 × 10−4
Aspartyl protease family protein
3.47 × 10−4
3.02 × 10−5
Apoptosis regulator Bcl-2
6.53 × 10−5
Polygalacturonase (pectinase) family protein
Polygalacturonase (pectinase) family protein
4.41 × 10−5
Cellulase family protein
9.19 × 10−4
6.61 × 10−5
2.62 × 10−5
1.31 × 10−5
Cellulose synthase genes (CSLA)
1.24 × 10−4
3.31 × 10−4
Phenylcoumaran benzylic ether reductase 1
1.41 × 10−4
Fasciclin-like arabinogalactan protein
2.03 × 10−5
Fasciclin-like arabinogalactan protein
1.68 × 10−4
8.86 × 10−4
2.54 × 10−4
Tubulin beta chain 4
Tubulin beta chain 16
3.95 × 10−4
Tubulin beta chain 13
Leucine-rich repeat transmembrane protein kinase
Leucine-rich repeat family protein
1.92 × 10−4
Leucine-rich repeat transmembrane protein kinase
2.45 × 10−4
Leucine-rich repeat receptor kinase
1.22 × 10−4
Leucine-rich repeat transmembrane protein kinase
5.76 × 10−4
Protein kinase family protein
Protein kinase family protein
4.56 × 10−5
bHLH protein family
9.18 × 10−4
Dof-type zinc finger
WRKY-type transcription factor
1.62 × 10−4
bHLH protein family
GATA family protein
Zinc finger (C2H2 type) family protein
bHLH transcription factor
2.21 × 10−4
Zinc finger (C3HC4-type RING finger) protein
3.23 × 10−4
Zinc finger (C3HC4-type RING finger) protein
8.77 × 10−4
PtVNS/PtrWND transactivation activity varies according to the gene
Based on the information obtained from microarray analysis, we tested PtVNS/PtrWND transactivation activity toward downstream genes. Several genes were selected from the list of poplar genes induced by AtVND7 (Table 1), such as those encoding XCP1, polygalacturonase, peroxidase, laccase and MYB transcription factors (MYB020 and MYB021; McCarthy et al., 2010). For cellulose synthase genes, we targeted four PtCesA genes, PtCesA7, -8, -17 and -18, which have been reported as highly expressed in xylem tissue (Suzuki et al., 2006). Transient expression assay using poplar leaves showed that PtVNS/PtrWND transactivation activity differed from gene to gene (Figures 4 and 5). Among the PtVNS/PtrWND genes of the VND group, PtVND07/PtrWND6A and/or PtVND08/PtrWND6B tended to show higher activity, except for MYB021 and PtCesA18. In Arabidopsis, several differences in downstream genes between AtVND and AtNST have been reported, e.g. genes participating in programmed cell death are only upregulated by AtVND (Zhong et al., 2006, 2010b; Ohashi-Ito et al., 2010; Yamaguchi et al., 2011). In the case of poplar genes, such divergence in target genes did not seem as stringent as in the Arabidopsis genes, because PtVNS/PtrWND genes of the NST group seemed to activate PtXCP1 gene transcription to some extent (Figure 4b). On the other hand, we still found a difference in preference for target genes by gene groups. For instance, VND group genes show higher activity for genes encoding XCP1, polygalacturonase and peroxidase, whereas NST group genes effectively activate the MYBs and PtCesAs genes (Figures 4 and 5). Thus, we speculate that the functions of VND group genes and NST group genes are partially diversified for wood formation.
Furthermore, we noticed a large difference in activity between paired genes for some cases. For example, PtVNS09/PtrWND2A effectively activated the PtCesA18 gene promoter, although its activity for the PtCesA17 promoter was not significant. In contrast, PtVNS10/PtrWND2B significantly activated PtCesA17 promoter activity, whereas its activity for the PtCesA18 promoter was not significant (Figure 5b). These findings are intriguing, because all paired genes show more than 80% identity at the amino acid level (Table S3). Taken together with our previous data indicating that VND family proteins function as homo- and/or heterodimers (Yamaguchi et al., 2008), considerable variation in transactivation activity can be expected by the combination of PtVNS/PtrWND genes.
PtVNS/PtrWND genes have full potential to induce secondary wall deposition in poplar and Arabidopsis
Finally, we checked PtVNS/PtrWND activity for the induction of secondary cell wall formation. The 12 PtVNS/PtrWND genes with xylem expression were driven under the control of the cauliflower mosaic virus (CaMV) 35S promoter, resulting in ectopic secondary wall formation in the regions around the vascular tissue in transgenic poplar leaves, like in the transgenic 35S::VND7:VP16:GR poplar (Figures 6 and S2; Yamaguchi et al., 2010a). Ectopic secondary wall thickening was also observed in the epidermal cells (Figure 6), which are separate from the vascular cell lines, clearly indicating that all PtVNS/PtrWND genes have the full potential to induce secondary wall formation in poplar.
It has already been demonstrated that the overexpression of PtVNS08/PtrWND6B and PtVNS10/PtrWND2B induces the deposition of a secondary wall, containing ectopic lignin, cellulose and xylan, in Arabidopsis (Zhong et al., 2010a). In order to further examine the properties of PtVNS/PtrWND genes, we carried out overexpression analysis in Arabidopsis seedlings for PtVNS/PtrWND genes. Our data evidently showed that all tested PtVNS/PtrWND genes induced ectopic secondary wall thickening in Arabidopsis seedlings as well as in poplar (Figure 7 and Table 2). Importantly, our investigation also revealed that the induction efficiency for ectopic secondary wall deposition differed between PtVNS/PtrWND genes (Table 2). In Arabidopsis, PtVNS05/PtrWND3A, PtVNS06/PtrWND3B, PtVNS07/PtrWND6A, PtVNS08/PtrWND6B and PtVNS11/PtrWND1B appeared to have high induction activity. We also detected a gap in induction efficiency even between paired genes, especially in the roots (compare the number of seedlings showing secondary wall deposition in root cells between PtVNS07/PtrWND6A and PtVNS08/PtrWND6B, or between PtVNS09/PtrWND2A and PtVNS10/PtrWND2B; Table 2). Transient expression assays demonstrated the variation in PtVNS/PtrWND transactivation activity (Figures 4 and 5). These gaps in induction efficiency would arise from such differences in PtVNS/PtrWND characteristics. In addition, we found an organ-specific difference in induction efficiency. Ectopic secondary wall formation was frequently observed in hypocotyls for any PtVNS/PtrWND genes, compared with the other organs (Table 2). It is known that the secondary xylem-like structure is formed in mature hypocotyls of Arabidopsis (Chaffey et al., 2002). Therefore, we assume that the hypocotyl cells possess innate properties allowing secondary wall formation, and this causes the higher efficiencies in the hypocotyls.
Table 2. Efficiency of secondary cell wall deposition induction by PtVNS overexpression
aNumber of transgenic Arabidopsis seedlings (T1) with ectopic secondary cell wall deposition when we counted 40 14-day-old seedlings in total (n =40).
bTotal number of transgenic Arabidopsis seedlings (T1) with ectopic secondary cell wall deposition in any organ. Percentages indicate the portion of seedlings with ectopic secondary wall. Values >80% are represented by boldface type.
Of note, all tested PtVNS/PtrWND genes gave ectopic secondary wall deposition with helical (protoxylem vessel-like) and/or reticulated (metaxylem vessel-like) patterns in poplar leaves, regardless of the gene group (Figure 6). Similarly, both types of ectopic secondary wall deposition were induced in Arabidopsis, as well as ectopic deposition of secondary wall without an apparent pattern, by overexpression of any PtVNS/PtrWND gene (Figure S3). These findings indicate that all PtVNS/PtrWND genes of the VND and NST groups exhibit, in common, the activity to positively regulate secondary wall thickening. This idea corresponds to the results of expression analysis, showing no apparent difference in expression sites in secondary xylem tissue for all expressed PtVNS/PtrWND genes (Figure 2).
In conclusion, our results provide a vision of the complex and delicate regulation of wood formation by NAC domain protein PtVNS/PtrWND genes in poplar: 12 PtVNS/PtrWND genes redundantly control the differentiation of both vessel cells and fiber cells during xylem tissue formation by adjusting their activity depending on the situation, probably through a change in expression level and/or dimer form. PtVNS/PtrWND genes act not only on the enzyme genes for secondary wall formation and programmed cell death, but also on other transcription factors (Table 1, Figures 5 and 6), suggesting that there is a complex system regulating wood formation.
The results described here also demonstrate the different nature of the regulation system for secondary wall thickening between Arabidopsis and poplar. We recently reported a negative regulator for vessel differentiation, AtVNI2 (Yamaguchi et al., 2010b). In poplar leaves, homologous genes to AtVNI2 were upregulated by AtVND7 overexpression (Table 1), whereas AtVND7 overexpression does not significantly induce AtVNI2 expression in Arabidopsis (Yamaguchi et al., 2010b), suggesting a difference in negative regulation among species. Moreover, expression profiles of PtVNS/PtrWND genes revealed that mRNAs of all expressed PtVNS/PtrWND genes are accumulated in all types of cells in developing xylem, regardless of gene group (Figure 2). In contrast, AtVND and AtNST genes are basically expressed in distinct types of cells, in vessels and fibers, respectively, not only in the inflorescence stems but also in the secondary xylem-like structure found in mature hypocotyls of Arabidopsis (Chaffey et al., 2002; Mitsuda et al., 2007; Yamaguchi et al., 2008). These findings raise the possibility that specialized gene expression for different types of xylem tissue cells by gene groups was acquired in herbaceous Arabidopsis after divergence from a common ancestor of herbaceous and woody plants, or that woody plants developed a woody-plant-specific feature for gene expression patterns in which the NAC protein genes are boundlessly expressed in all types of xylem cells. Future studies on other plant species will answer this question.
Our results also provide information about many poplar genes functioning in xylem cell differentiation (Tables 1 and S2), including unique poplar genes without evident homologous genes in the Arabidopsis genome (Table S2). These should be important clues for understanding the molecular mechanisms for wood formation. Further studies on controlling PtVNS/PtrWND activity for these genes should lead to the control of the quality and quantity of wood cells (Demura and Ye, 2010; Zhong et al., 2010c) to improve the industrial utility of wood.
Plant materials and growth conditions
We used the shoot regions of black cottonwood, P. trichocarpa (poplar), grown in 15 cm high plant pots for PtVNS gene cloning. For expression analysis, P. trichocarpa trees grown in the greenhouse were used. Transgenic poplar plants were generated from hybrid aspen T89 lines (Populus tremula × tremuloides; Nilsson et al., 1992) for overexpression analysis. Propagation and maintenance of transgenic lines was carried out aseptically on medium containing Murashige and Skoog (MS) salt mixture (Sigma-Aldrich, http://www.sigmaaldrich.com/) (pH 5.6) under long-day (LD; 16 h light/8 h dark) conditions at 23°C. The transgenic 35S::AtVND7:VP16:GR poplar is described in Yamaguchi et al. (2010a). We used the Columbia strain of A. thaliana for overexpression analysis. The T1 transgenic seedlings were screened by growing on a medium containing MS salt mixture (Wako Pure Chemical, http://www.wako-chem.co.jp/), 1% (w/v) sucrose, 0.05% (v/v) 2-(N-morpholine)-ethanesulfonic acid (MES)-KOH (pH 5.8), B5 vitamins, and 1.5% (w/v) agar supplemented with 50 μg ml−1 hygromycin at 22°C under continuous light.
Molecular cloning of PtVNS/PtrWND genes
Based on the publicly accessible information about poplar genomic sequences (US Department of Energy, Joint Genome Institute (JGI), http://www.phytozome.net/poplar), we isolated 16 poplar genes encoding NAC domain proteins as homologs of AtVND genes in silico. When we started this work, full-length cDNA sequences were known for only two genes, PtVNS11/PtrWND1B and PtVNS12/PtrWND1A. Thus, we performed 5′ and/or 3′ rapid amplification of cDNA ends (RACE) analysis for the PtVNS01–10/PtrWND genes. Total RNA was isolated from the shoot region of poplar using Plant RNA Isolation Reagent (Invitrogen, http://www.invitrogen.com/), and then purified by a RNeasy Mini kit (Qiagen, http://www.qiagen.com/). The RACE analysis was carried out using the GeneRacer® kit (Invitrogen), according to the manufacturer’s instructions. The primer sequence information used is described in Table S4. See Appendix S1 for the full-length cDNA sequences obtained. The GenBank accession numbers are given below.
Amplified PtVNS/PtrWND cDNA fragments corresponding to each coding region (for primer sequences see Table S4) were cloned into the pENTR®/D/TOPO vector (Invitrogen) and then integrated into the Gateway® destination vector, pH35GS (Kubo et al., 2005), in which plasmid PtVNS/PtrWND genes are expected to be driven by the CaMV 35S promoter, using LR Clonase® (Invitrogen). These plasmids were used for transient expression analysis (as the effector plasmids) and for overexpression analysis. For transient expression assays, we selected several target genes for constructing the reporter plasmids. After checking the increase in expression levels following the DEX treatment by RT-PCR, we amplified the promoter regions of target genes by PCR with specific primers (Table S4) from genomic DNA isolated from poplar shoots. The resultant DNA fragments were cloned into pENTR®/D/TOPO or pCR®8/GW/TOPO® vector (Invitrogen), and then transferred into pAGL vector using LR Clonase® (Invitrogen) to obtain the reporter plasmids. pAGL vector was constructed from the GAL4UAS:TATA:LUC reporter plasmid described in Yamaguchi et al. (2010b), by converting the regions of the CaMV 35S promoter, GAL4UAS, TATA box and the tobacco mosaic virus Ω sequence to Gateway® Reading Frame Cassette B. For construction of PtVNS09/PtrWND9Apro::GUS and PtVNS10/PtrWND9Bpro::GUS, approximately 2300 bp promoter fragments were amplified and cloned into pGUS_Ent vector (Mitsuda et al., 2007) followed by Gateway LR reaction to transfer the contents to T-DNA vector pBCKK (Mitsuda et al., 2006).
Generation of transgenic plants
The constructed plasmids were electroporated into the Agrobacterium tumefaciens strain GV3101::pMP90. A simplified version of the floral dip method was used for transformation of the Arabidopsis plant (A. thaliana, Columbia strain) (Clough and Bent, 1998). Transgenic poplars were generated using the method described by Eriksson et al. (2000).
Semi-quantitative RT-PCR analysis
For the xylem tissues, 5 mm diameter stems of P. trichocarpa trees grown in the greenhouse were sampled. These were debarked and used for isolation of the xylem tissues. Young shoot tips (first internodes) were harvested from approximately 50 cm long branches of 4-year-old P. trichocarpa trees grown in the greenhouse. Shoot apices, stems, petioles and leaves of the young shoot tips were sampled separately for the detailed expression analysis. The tissue samples were stored immediately in liquid nitrogen until use. Total RNA was isolated from each organ as described above. For semi-quantitative RT-PCR analysis, 2 μg of total RNA was reverse transcribed by SuperScript®II RNase H− Reverse Transcriptase (Invitrogen) with oligo d(T)12–16 primer. Aliquots (1 μl of the reaction solution containing the first-strand cDNA) were used as templates for PCR amplification. Gene-specific primers corresponding to 5′ and 3′ untranslated (UTR) regions were used for PCR experiments (Table S4). The resultants were separated by agarose gel electrophoresis and the gel images were analyzed using the imagequant las 4010 (GE Healthcare, http://www.gehealthcare.com/).
In situ hybridization analysis
Poplar explants of stems and petioles were fixed in PBS containing 4% (w/v) paraformaldehyde. Paraffin sections were made from the fixed samples and hybridized with digoxigenin-labeled probes, followed by detection of signals using alkaline phosphatase-conjugated anti-digoxigenin antibody, as described previously (Ohtani et al., 2008). Probes were prepared from the cloned DNA sequences corresponding to the 5′ or 3′ UTR regions of PtVNS genes.
For the DEX treatment, the leaves of 35S::AtVND7:VP16:GR poplar were soaked in water containing 10 μm DEX (Yamaguchi et al., 2010a). Microarray analysis was performed using the GeneChip® Poplar Genome Array (Affymetrix, http://www.affymetrix.com/) on three independent biological replicates from the leaves of 35S::AtVND7:VP16:GR poplar with or without DEX treatment for 24 h. Subsequent procedures of quality control, statistical analysis and filtering were carried out using genespring gx software v11.0.1 (Agilent Technologies, http://www.home.agilent.com/). Probe sets of the GeneChip® Poplar Genome Array were based on content from UniGene Build #6 (16 March 2005), GenBank mRNAs and ESTs for all Populus species (up to April 26, 2005), in addition to the predicted v1.1 gene set from the P. trichocarpa genome (JGI). Moreover, we subjected the transgenic poplar derived from hybrid aspen (P. tremula × tremuloides) to microarray analysis. This made it difficult to distinguish the origin of genes of transcripts giving signals; therefore, we decided to perform data analysis by probe units, without summarizing probe-level data into gene-level data. P-values were calculated for each probe by Welch’s t-test (n = 3), for differences between the treated leaves and the control leaves. We used the Benjamini–Hochberg FDR method for controlling false positives. A corrected P-value cutoff of 0.05 was used to select the regulated genes with the lowest FDR. Fold change values were also computed by GeneSpring GX and we targeted probes that were upregulated by more than 10-fold (Table S2). Microarray data presented in this study were submitted to NCBI GEO (http://www.ncbi.nlm.nih.gov/geo/) and can be retrieved via accession number GSE25304.
Transient expression assays
The reference plasmid containing Renilla reniformis luciferase (Rr-luc) is referenced in Yamaguchi et al. (2010a,b). The PDS-1000/He system (GE Healthcare) was used to bombard the T89 poplar leaves with gold particles (GE Healthcare) coated with a mixture of the reporter, effector and reference plasmid DNA at a ratio of 15:10:12. After overnight incubation, the leaf samples were pulverized in liquid nitrogen, and luciferase activity was assayed using the Dual-Luciferase® Reporter Assay System (Promega, http://www.promega.com/) using a Mithras LB940 reader (Berthold, http://www.bertholdtech.com/).
For histological observation of poplar stems, paraffin sections were obtained by the same method as for in situ hybridization analysis, and were stained with safranin and aniline blue. Explants of poplar and Arabidopsis overexpressors were fixed overnight at −20°C in 90% acetone, and then hydrated. Samples were mounted with a few drops of an 8:1:2 (w/v/v) mixture of chloral hydrate, glycerin and water. Observations were made under a microscope equipped with Nomarski optics (BX51-DIC; Olympus, http://www.olympus.com/). For histochemical detection of GUS activity, samples were treated as described previously (Mitsuda et al., 2007).
We thank Ms Ayumi Ihara, Ms Sachiko Ohyama, Mr Mitsutaka Araki and Mr Hitoshi Endo (RIKEN) for their excellent technical assistance and Ms Toshie Kita, Ms Akiko Sato, Ms Kayo Kitaura and Ms Satsuki Takagi (RIKEN) for their contribution to the propagation of poplar trees. We also thank Ms Tomoko Kuriyama (RIKEN) for her contribution to Arabidopsis transformation and Dr Miho Ikeda (AIST) for her technical advice and assistance in establishing the transgenic poplar. This work was supported in part by RIKEN Biomass Engineering Program and RIKEN Plant Science Center and by a Grant-in-Aid for Scientific Research (grant no 20770041 to MY, no 21770064 to NM and nos 21027031 and 22370020 to TD) from the Ministry of Education, Culture, Sports, Science and Technology of Japan and from the Japan Society for the Promotion of Science.