NAC domain function and transcriptional control ... - Wiley Online Library

The Plant Journal (2011) 68, 1104–1114

doi: 10.1111/j.1365-313X.2011.04764.x

NAC domain function and transcriptional control of a secondary cell wall master switch Huanzhong Wang, Qiao Zhao, Fang Chen, Mingyi Wang and Richard A. Dixon* Plant Biology Division, Samuel Roberts Noble Foundation, 2510 Sam Noble Parkway, Ardmore, OK 73401, USA Received 15 August 2011; accepted 23 August 2011; published online 10 October 2011. * For correspondence (fax 580 224 6692; e-mail: [email protected]).

SUMMARY NAC domain transcription factors act as master switches for secondary cell wall thickening, but how they exert their function and how their expression is regulated remains unclear. Here we identify a loss-of-function point mutation in the NST1 gene of Medicago truncatula. The nst1-3 mutant shows no lignification in interfascicular fibers, as previously seen in tnt1 transposon insertion alleles. However, the C fi A transversion, which causes a T94K mutation in the NST1 protein, leads to increased NST1 expression. Introduction of the same mutation into the Arabidopsis homolog SND1 causes both protein mislocalization and loss of target DNA binding, with a resultant inability to trans-activate downstream secondary wall synthesis genes. Furthermore, transactivation assays show that the expression of SND1 operates under positive feedback control from itself, and SND1 was shown to bind directly to a conserved motif in its own promoter, located within a recently described 19-bp secondary wall NAC binding element. Three MYB transcription factors downstream of SND1, one of which is directly regulated by SND1, exert negative regulation on SND1 promoter activity. Our results identify a conserved amino acid critical for NST1/SND1 function, and show that the expression of the NAC master switch itself is under both positive (autoregulatory) and negative control. Keywords: NAC domain, transcription factor, secondary cell wall, transcriptional regulation, autoregulation.

INTRODUCTION Plant secondary cell walls, which are composed primarily of cellulose, lignin and hemicelluloses, provide mechanical strength to tracheary elements and fibers, and hydrophobicity to water and nutrient transporting vessels. Secondary walls also have specialized functions associated with pollen release from anthers, seed shattering in pods and fiber elongation in seed trichomes (Mitsuda et al., 2005; Lee et al., 2007; Mummenhoff et al., 2009). The genes involved in cellulose, xylan and lignin synthesis must be coordinately expressed during secondary wall biosynthesis. In Arabidopsis, several closely related NAM, ATAF and CUC (NAC) transcription factors (TFs) have been shown to be key regulators of secondary wall biosynthesis: these are NAC SECONDARY WALL THICKENING PROMOTING FACTOR 1 (NST1), NST2/NST3/SECONDARY WALL-ASSOCIATED NAC DOMAIN PROTEIN 1 (SND1), VASCULAR-RELATED NAC-DOMAIN 6 (VND6) and VND7 (Mitsuda et al., 2005; Zhong et al., 2006; Mitsuda et al., 2007; Ohashi-Ito et al., 2010; Yamaguchi et al., 2011). Recently, a single NST gene was cloned from the model legume 1104

Medicago truncatula by forward genetic screening, and was shown to function as a master switch for secondary wall synthesis in various tissues (Zhao et al., 2010). MYB TFs have also been identified as important for secondary wall biosynthesis, and act downstream of NST genes (Zhong et al., 2007; Ko et al., 2009; McCarthy et al., 2009; Zhou et al., 2009). Among these MYB TFs, MYB46 was shown to be a direct target of NST3/SND1 in vitro and in vivo, and is regarded as a master switch that turns on the expression of many of the genes responsible for the biosynthesis of cellulose, xylan and lignin (Zhong et al., 2007; Ko et al., 2009). The NAC TFs feature a conserved NAC domain, and are encoded by a large gene family in Arabidopsis. Phylogenetic analysis shows that functionally related members are clustered in the same clade (Shen et al., 2009). Although the importance of NAC TFs in controlling secondary wall thickening has been recognized, our understanding of the function(s) of the NAC domain and how NAC genes/proteins are regulated is still limited. Studies on the Arabidopsis NAC1 gene showed that the NAC domain functions in ª 2011 The Authors The Plant Journal ª 2011 Blackwell Publishing Ltd

NST1 function and regulation 1105 nuclear localization, DNA binding and also protein–protein interactions (Xie et al., 2000). The NAC domain is about 190 amino acids long, which is more than half of the full-length protein, and is composed of five subdomains. It is still unclear which subdomain is responsible for DNA binding, and no nuclear localization motif has been identified. By mapping the SND1 binding sequence using an electrophoretic mobility shift assay (EMSA) and transactivation analysis, it has recently been demonstrated that SND1, together with other secondary wall regulatory NACs, including VND6, VND7, NST1 and NST2, bind to a 19-bp consensus sequence designated as the secondary wall NAC binding element (SNBE) (Zhong et al., 2010b). However, this element as defined is quite degenerate, and is therefore very common in plant genomes; different NAC genes may bind to different versions, or other factors may be involved in determining binding specificity for this site. In the present study we report the identification and characterization of a loss-of-function point mutation (T94K) of the MtNST1 gene of M. truncatula, and the further analysis of the effects of the same mutation in the homologous AtSND1 gene of Arabidopsis. Surprisingly, NST1 transcript levels are increased in the point mutation plants. The AtSND1 protein harboring the T94K mutation lost the ability to trans-activate downstream secondary wall synthesis genes. Localization and DNA binding analysis showed that the point mutation causes both mislocalization and lack of DNA binding. Trans-activation analyses using a dual luciferase system and EMSA analyses showed that the SND1 protein can bind to its own promoter to activate its transcription, and that expression of SND1 is also under negative control of MYB TFs. The SND1 binding site in the SND1 promoter is a more stringent variant of the previously identified SNBE element.

RESULTS Isolation of a mutant lacking lignification in fibers In a screen to identify mutants with altered secondary wall structure and changed lignin accumulation pattern in M. truncatula, several mutants with reduced lignification in interfascicular and vascular tissues were isolated (Zhao et al., 2010). One of these mutants, NF2225, showed a similar phenotype to the Mtnst1 mutant in which the NST1 gene was disrupted by a Tnt1 retrotransposon insertion. UV microscopy of stem cross sections of young (fourth) internodes revealed no differences in lignin autofluorsecence pattern between the NF2225 line and the wild type, as shown in Figure 1(a,f). However, as the internodes mature, the wildtype plants accumulate increasing levels of lignin in the interfascicular and vascular bundle fibers, whereas no lignin is observed in these tissues in NF2225 mutant plants, even in the very mature eigth internode (Figure 1). Except for the lack of lignin accumulation in specific cell types, the mutant appeared identical to the wild type, with normal height and no obvious growth phenotypes. The NF2225 phenotype is caused by a point mutation in the MtNST1 coding sequence As the NF2225 mutant plants showed a very similar phenotype to Mtnst1, we checked the possibility of a Tnt1 retrotransposon insertion in the MtNST1 gene. Using Tnt1 and MtNST1 gene-specific primers, we could not detect any insertion in the MtNST1 gene in the mutant plants. We then checked the expression of MtNST1 by RT-PCR. Compared with the R108 wild-type plants, expression of MtNST1 in NF2225 appeared higher, whereas the Tnt1 insertion line nst1-2 has no detectable transcript (Figure 2a). The NST1 PCR product from the mutant was the same size as that from

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

(i)

(j)

Figure 1. Stem cross sections showing the lack of lignification in interfascicular tissue of the nst1-3 mutant. Fluorescence microscopy of cross sections of stems of the wild type (a–e) and nst1-3 (f–j). The blue color represents autofluorescence from the lignified vascular bundles and interfaciscular cells. The fourth (a, f), fifth (b, g), sixth (c, h), seventh (d, i), and eigth (e, j) internodes from the wild type and mutant were compared. Scale bars: 20 lm, and all pictures are taken at the same magnification.

ª 2011 The Authors The Plant Journal ª 2011 Blackwell Publishing Ltd, The Plant Journal, (2011), 68, 1104–1114

1106 Huanzhong Wang et al.

(a)

CK

(b)

nst1-3 nst1-2

T N R A T A A G F W 268-ACCAATCGCGCTACGGCTGCAGGGTTCTGG 94K ACCAATCGCGCTAaGGCTGCAGGGTTCTGG

MtNST1 AcƟn

DdeI

(c)

M

WT

HE

HO 655 bp 465 bp 190 bp

(d) AtSND1/NST3 AtNST1 AtNST2 MtNST1

(e)

nst1-3

nst1-3 nst1-2

Figure 2. Characterization of the Mtnst1 point mutation. (a) Expression of MtNST1 transcripts in control, nst1-3 and nst1-2. Actin was used as control. (b) Diagram showing the position of the mutation (wild-type sequence above). Numbers indicate the position from the start codon in the coding sequence. The underlined letters represent the Dde1 cleavage site. The changed amino acid is marked in red. (c) Gel picture showing the segregation of the CAPs marker. M, size marker; WT, wild-type plant; HE, heterozygous plant; HO, homozygous plant. (d) Alignment showing part of the NAC domains of NST genes in Medicago truncatula and Arabidopsis (Thompson et al., 1994). (e) Fluorescence microscopy of the sixth internodes of nst1-3 (left) and double heterozygotes of nst1-3 nst1-2. Scale bars: 20 lm.

the wild type, further confirming that there was no Tnt1 insertion in the MtNST1 gene. We reasoned that the phenotype of NF2225 might be caused by disruption of a gene that functions in the same pathway with MtNST1, and therefore performed a microarray analysis to reveal candidate downregulated genes. The results confirmed overexpression of MtNST1 in the mutant (by approximately between three and fivefold, as compared with an approximately 20-fold reduction in the transposon insertion lines), but most of the downregulated genes were the same as those previously observed from the microarray analysis of Mtnst1 transposon insertion lines (Table S1). Out of the 188 genes that were downregulated by at least twofold in NF2225 compared with the wild-type control, 157 (83.5%) were also downregulated by at least twofold in either nst1-1 or nst1-2 transposon insertion lines (Table S1). To further exclude the possibility that the phenotype was caused by disruption of MtNST1 expression, we sequenced the MtNST1 genomic sequence from line NF2225. We found two mutations in the coding sequence: one located in the first intron and the other in the second exon. The CfiA point

mutation in the second exon, resulting in a T94 fi K amino acid change, introduced a DdeI cleavage site that could be used as a cleaved amplified polymorphic sequence (CAPS) marker (Figure 2b,c). This CAPS marker co-segregated with the mutant phenotype in both R0 and R1 segregating populations. The point mutation was present in the NAC domain of the protein (Figure 2d). To further confirm that the point mutation was responsible for the mutant phenotype, we made a cross between NF2225 and the Tnt1 insertion allele nst1-2 (NF2911). As expected, the F1 plants showed the mutant phenotype (Figure 2e), and genotyping indicated the presence of both mutations. Because the point mutation:transposon insertion double heterozygote gave the altered lignification phenotype, we can conclude that the phenotype of NF2225 is caused by the point mutation. MtNST1 belongs to the NAC-c subfamily of NAC domain TFs, and the single T94K amino acid substitution causes loss of function. Multiple sequence alignment confirmed that T94 is conserved among all but one of the 130 homologous proteins in the NAC-c clade from 11 different species (Shen et al., 2009). The T94K mutation eliminates the trans-activation activity of Arabidopsis SND1 Arabidopsis AtSND1 is homologous to MtNST1, and has been shown to activate the secondary cell wall synthesis program. To check whether the T94K mutation may also cause the loss of function of AtSND1, we introduced the point mutation into AtSND1 by PCR and analyzed the transactivation activity of the corresponding recombinant protein using a dual luciferase system. As reported previously (Zhong et al., 2006; Ko et al., 2009), the wild-type SND1 expression construct could activate the promoters of a number of cellulose, xylan and lignin biosynthesis genes. However, the T94K mutant could barely activate any of the promoters (Figure 3a,b), indicating that T94 is critical for SND1 function. The T94K mutation affects the subcellular localization and DNA binding activity of SND1 To determine how the point mutation affects SND1 function, we first examined protein localization using protein–reporter fusions. Both N-terminal and C-terminal green fluorescent protein (GFP) fusion constructs were tested to eliminate the potential influence of the fusion, and no difference was observed between the two types of fusion (Figure S1). The wild-type AtSND1 fusion protein was localized to the nucleus as previously reported (Zhong et al., 2006). A large portion of the mutated AtSND1-GFP signal was localized to the nucleus, but signal was also observed in the cytoplasm (Figure S1). The NAC domain has been reported to be responsible for the nuclear localization of some NAC TFs (Xie et al., 2000), and the T94K mutation is located in the NAC domain. To test


NST1 function and regulation 1107

(a)

Reporter

DNA (Figure 5b, lane 4). The T94K mutated SND1 protein, however, totally lost binding ability (Figure 5b, lane 5).

C CaMV 35SPro 3 S

Effector

SND1/SND1m Cell wall genes Pro Firefly luciferase

Control

2500 2000

50

CESA8 IRX8 IRX9

1500 1000 500 0

CK

SND1

SND1mu

Self-activation of SND1

Renilla luciferase

Rela ve expression level

Rela ve expression level

(b)

CaMV 35SPro

CESA7 PAL4

CCoAOMT

CK

SND1

40 30 20 10 0

SND1mu

Figure 3. The T94K mutation in SND1 causes a loss of trans-activation of secondary cell wall synthesis genes. (a) Diagram showing the constructs used in the transactivation assay. (b) Activation of six secondary cell wall synthesis promoters by SND1 and its point mutant. CES, cellulose synthase; irx, irregular xylem; PAL, L-phenylalanine ammonia lyase; CCoAOMT, caffeoyl coenzyme A 3-O-methyltransferase. The relative expression level represents the expression of these cell wall synthetic genes compared with the expression of the ubiquitin gene. Bars represent SDs from three independent biological replicates.

if the NAC domain is responsible for the nuclear localization of SND1 and to further assess the effect of the point mutation on NAC domain function, we constructed an NAC domain-YFP fusion consisting of 191 amino acids that included the entire NAC domain of SND1 (subdomains A–E) (Shen et al., 2009). The wild-type NAC domain-YFP was localized exclusively to the nucleus, whereas the NACmYFP, which contains the point mutation, again showed both nuclear and cytoplasmic localization (Figure 4). These results indicate that the point mutation somehow affects the nuclear localization function of the NAC domain, but this may not be the main reason for the loss of function, as most of the signal is still in the nucleus. We then tested the DNA binding ability of wild-type and mutated SND1 proteins. In a previous report, SND1 was shown to trans-activate the MYB46 promoter, to which it can directly bind (Zhong et al., 2007). We first confirmed the trans-activation of the MYB46 promoter by SND1 in the dual luciferase system (Figure 5a), and showed that the T94K mutation reduced the activation by approximately 85%. To investigate this phenomenon at the level of promoter binding, wild-type and mutant NAC domain polypeptides were expressed in Escherichia coli as fusions to maltosebinding protein (MBP), and the proteins were purified for the DNA binding assay (Figure S2). The same promoter fragment of MYB46 as used for the trans-activation assay was then used in EMSAs for the direct demonstration of promoter binding. The direct binding of SND1 to this fragment was confirmed by adding unlabeled competitor

Microarray analysis had indicated that the expression of MtNST1 was upregulated more than threefold in the nst1-3 mutant background compared with the control plants (Table S1), suggesting that the expression of MtNST1 may somehow be autoregulated. To directly address this possibility, we used the trans-activation assay with Arabidopsis SND1. Surprisingly, the SND1 promoter was activated more than 18-fold when protoplasts were co-transfected with 35S:SND1, but this trans-activation activity was reduced to around twofold for the T94K mutation (Figure 6a,b). To address whether the ability of SND1 to activate its own expression is likely to be a direct phenomenon, we investigated whether SND1 bound to its own promoter by EMSA analysis. The SND1 promoter sequences used for the dual luciferase assay spanned 1052 bp upstream of the start codon. For EMSA analysis, we cloned three overlapping promoter fragments by PCR, designated as P1, P2 and P3, encompassing this region of the 5¢-upstream sequence (Figure 6c). Both P2 and P3 fragments showed a clear gel shift when incubated with SND1 that was abolished by adding a competitor DNA fragment. These results indicate that the SND1 protein can bind directly to its own promoter, and that there may be more than one binding site in the promoter sequence. The SND1 T95K mutant protein was unable to bind to the P2 fragment (Figure S3). To further narrow down the promoter region of SND1 needed for SND1 binding, we focused on the P2 fragment and synthesized nine short overlapping fragments encompassing this sequence (Figure 6d). In competition experiments, we found that P2-5 and P2-9 could compete with the biotin labeled full-length P2 probe (Figure 6e). The P2-5 fragment is shared by P2 and P3. To test if P2-5 is also responsible for the binding of SND1 to the P3 fragment, we checked for competition of P2-5 with a biotin-labeled P3 probe. P2-5 did indeed compete with P3, and was the only fragment that could do so (Figure S4), indicating that P2-5 contained the binding sequence for SND1 in the SND1 promoter. P2-9 is shorter than P2-5, and can also compete with the P2 fragment. To determine the critical nucleotides required for SND1 binding, we compared the P2-9 fragment with the published binding sequence of SND1 in the MYB46 promoter (Figure 7a) (Zhong et al., 2007). A consensus sequence, TATACXTTXXXXATGA, was found between the two. To check if all the nucleotides in the consensus sequence are critical we generated a series of mutants (M1–M5; Figure 7b), and tested these as competitors. Only M1 competed as strongly as the wild-type sequence, indicating that the first two nucleotides are not critical. We


1108 Huanzhong Wang et al. (a)

35S:NAC-YFP

(b)

(c)

(d)

35S:NACm-YFP

(e)

(f)

Figure 4. The T94K point mutation affects the subcellular localization of SND1. Nuclear localization of SND1 (NAC-YFP) (a–c) and of SND1 with the T94K mutation (NACm-YFP) (d–f) revealed by transient expression via leaf infiltration in Nicotiana benthamiana. (a, d) YFP signal detected by confocal microscopy of the infiltrated N. benthamiana leaf. (b, e) Light microscopy of the infiltrated N. benthamiana leaf. (c, f) Overlapping images of (a) and (b), and of (d) and (e), respectively.

Rela ve luciferase ac vity

(a)

ProMYB46:Firefly luciferasey

25

therefore conclude that the critical nucleotides required for SND1 binding are TACXTTXXXXATGA.

20

The T94K mutation eliminates binding of SND1 to the MYB32 promoter

15 10 5 0

CK

(b) SND1 SND1m MBP Compe tor

SND1

SND1mu

MYB46-promoter

– – – –

– – + –

+ – – –

+ – – +

– + – –

– + – +

Shi ed complex

Sequences 3 kb upstream of the translation start site of each annotated gene in the TAIR9 database (ftp://ftp.arabidopsis. org/home/tair/Sequences/blast_datasets/TAIR9_blastsets/ database) were searched for the presence of the TAC XTTXXXXATGA motif. This element is present in 647 promoters in the Arabidopsis genome (Table S2), including the promoter of MYB32, which is a known target for regulation by SND1 (Zhong et al., 2010b). EMSA using a 362-bp fragment of the MYB32 promoter sequence including this binding motif clearly showed direct binding of SND1 to the fragment, but SND1m showed no binding (Figure 8a). Negative regulation of the SND1 promoter by downstream MYB TFs

Free probe

Figure 5. The T94K mutation in SND1 results in reduced trans-activation activity. (a) Trans-activation of the MYB46 promoter by SND1 and the T94K mutated version SND1mu. Bars represent SDs from three independent measurements. (b) Electrophoretic mobility shift assay (EMSA), showing loss of binding of SND1mu to the MYB46 promoter DNA fragment.

The protein sequence of MYB32, together with that of MYB7, shares high similarity with MYB4, which has been shown to be a transcriptional repressor, and all three of these MYBs share a conserved EAR motif (Jin et al., 2000). These three MYB transcription factors can be directly activated by MYB46 (Ko et al., 2009), which is a direct target of SND1 (Zhong et al., 2006, 2007). Furthermore, MYB20 and MYB52 can also be activated by SND1 and MYB46 overexpression (Zhong et al., 2006, 2007). As shown above, SND1 directly binds to its own promoter to activate its expression, indicating the existence of a positive feedback loop. To adapt to



(a)

(b)

24 ProSND1

20

Effector CaMV 35SPro

16

SND1/SND1mu

12 Reporter

8

SND1 Pro

Firefly luciferase

4 0

(c)

P1 (-1052- -535)

CK

SND1

SND1mu

P3 (-323-- ATG)

P2 (-563- -255)

SND1 MBP Comp

Shi ed complex

Free probes

(d)

P2-1

P2-2

P2-3

P2-4

P2-5

P2 P2-6

(e)

P2 P2-1

P2-7 P2-2

P2-8 P2-3

P2-4

P2-5

P2-9 P2-6

P2-7 P2-8

P2-9

Shi ed complex Free Probe

Figure 6. SND1 can activate and directly bind to its own promoter. (a) Diagram showing the constructs used in the trans-activation assay. (b) Activation of the SND1 promoter by overexpression of SND1, but not the T94K mutated version (SND1mu). Bars represent SDs from three independent measurements. (c) Electrophoretic mobility shift assay (EMSA), showing the binding of SND1 to the SND1 promoter. P1, P2 and P3 are overlapping promoter fragments, and their relative distances to the start codon are indicated. (d) Diagram showing the relative positions of the short fragments on P2 that were used for the competition analyses in (e). (e) EMSA results showing that P2-5 and P2-9 can compete with the biotin-labeled SND1 promoter probe.

changing environmental and developmental contexts, a negative feedback mechanism will also be required to reset the cell wall synthesis pathways. To test if the above MYB TFs affect the expression of SND1, we performed further trans-activation assays using the dual luciferase system. We cloned the five MYB TFs into overexpression constructs and individually co-transfected them into protoplasts along with the SND1 promoter: luciferase reporter construct. MYB4, MYB7 and MYB32 all repressed the expression of SND1, but MYB20 and MYB52 had no effect (Figure 8b,c). Thus, MYB4, MYB7 and MYB32 may be negative regulators of SND1 expression. To test the repression of SND1 expression in planta, we transformed wild-type plants with a MYB32 over-expression construct driven by the CaMV 35S promoter. The transgenic plants

grew smaller than the wild type, and their leaves curved downwards (Figure 8d,e). Overexpression of MYB32 and the corresponding repression of SND1 expression were observed in stem tissues of two independent transgenic lines, as determined by quantitative reverse transcription (qRT)-PCR (Figure 8g,h). Taken together, the in vitro and in vivo data support the hypothesis that SND1 expression is itself under negative feedback control by its downstream negative regulators. DISCUSSION Secondary cell walls are critical for plant development, and their synthesis is under elegant transcriptional regulation (Zhong and Ye, 2007). NAC domain TFs have been shown to function as master switches in several species by activating


1110 Huanzhong Wang et al.

(a)

SND1 P2-9 TTCCGTTTTATACATTACACATGACATGACATTACACGTC ProMYB46 P6-6 GTATACCTTGTGAATGAAGAAACT Consensus: TATACXTTXXXXATGA

(b)

WT---TTCCGTTTTATACATTACACATGACATGACATTACACGTC M1---TTCCGTTTCTTACATTACACATGACATGACATTACACGTC M2---TTCCGTTTTACTCATTACACATGACATGACATTACACGTC M3---TTCCGTTTTATAGGTTACACATGACATGACATTACACGTC M4--- TTCCGTTTTATACAGGACACATGACATGACATTACACGTC M5--- TTCCGTTTTATACATTACACCGTCCATGACATTACACGTC

(c)

NC

WT

M1

M2

M3

M4

M5

Shi ed complex

Free probe

mutation leads to both protein mislocalization and loss of DNA binding. It is likely that the loss of binding is the major reason for the loss-of-function phenotype, as a significant proportion of the mutant protein is still transported to the nucleus. Studies on the crystal structure of the NAC domain of Arabidopsis ANAC found that the domain forms a twisted b-sheet, surrounded by a few helical elements that bind the DNA (Ernst et al., 2004). The mechanisms underlying several loss-of-function point mutation alleles in the CUC1 NAC gene have been discussed based on the NAC domain structure (Olsen et al., 2005). Although the T94 amino acid is conserved in the NAC-c subfamily, it is not found in all NAC domain TFs. At the corresponding position of the ANAC protein, the amino acid A replaces T. Because there is currently no protein structure for NAC-c subfamily members, it is still not clear how the T94K mutation affects DNA binding and protein localization. The effects could reflect loss of specific binding properties, or might also be nonspecific consequences of protein destabilization. Regulation of SND1 expression

SND binding mo f: TACXTTXXXXATGA

Figure 7. Identification of the binding motif in the SND1 promoter. (a) Comparison of P2-9 with the fragment from the MYB46 promoter that is responsible for SND1 binding. (b) Diagram showing the mutagenesis of different nucleotides within the binding region. (c) EMSA results showing competition of the different mutant oligonucleotides in (b) with the biotin-labeled P2-5 fragment. NC, no competitor; WT, wild-type version of P2-9.

different layers of downstream secondary wall-related TFs (Zhong et al., 2006; Mitsuda et al., 2007; Zhong et al., 2008; Zhao et al., 2010). NAC TFs are potential targets for genetic modification of feedstock secondary wall composition to meet the reduced recalcitrance requirements for liquid biofuel production. Although NAC TFs are proposed to be transcriptional activators, overexpression of SND1 in wildtype Arabidopsis resulted in thinner walls in vascular and interfascicular fibers, suggesting that control of secondary wall formation is more complex than is suggested by current models (Zhong et al., 2006; Ko et al., 2007). The present study provides insight into the functions of NAC domains and the mechanisms underlying NAC gene regulation, findings that may enhance the use of these genes for the development of bioenergy crops. A conserved amino acid is critical for NAC domain function The nst1-3 mutant of M. truncatula shows the same phenotype as tnt1 insertion mutants of MtNST1, indicating a total loss of NST1 function, and this was confirmed by the loss of in vitro trans-activation activity of AtSND1 harboring the T94K mutation. T94 of MtNST1 is conserved in the NAC-c subfamily of NAC domain TFs (Shen et al., 2009), and its

Because SND1 is a master TF for controlling secondary wall synthesis, its expression must be under strict developmental regulation (Demura and Ye, 2010; Zhong et al., 2010a). We found that SND1 can bind directly to its own promoter to selfactivate its expression, and that MYB TFs under the direct control of MYB46 can suppress the expression of SND1. The in vitro trans-activation and EMSA analysis provided valuable information for the assessment of SND1 function, but in vitro data should be interpreted with caution. We therefore further confirmed the repression of SND1 by MYB32 in vivo using transgenic studies. The results confirm that expression of SND1 is under both positive and negative regulation. The fact that SND1 binds to sequences within its own promoter and that of MYB46 allowed us to determine the critical nucleotides required for binding. We identified the binding motif TACXTTXXXXATGA by mutagenesis and EMSA. This motif is different from the tracheary-elementregulating (TERE) cis-element, which has been associated with the control of secondary wall formation and the programmed cell death of vascular elements (Pyo et al., 2007). Searching promoter regions of all Arabidopsis genes revealed that MYB32 may be a direct target of SND1, as recently shown by Zhong et al. (2010b), and this was confirmed by EMSA. Consistent with the binding results, we detected 2.15 times induction of the MYB32 promoter by SND1 in a trans-activation assay (data not shown), and the expression of MYB32 is downregulated in the nst1:nst3 double mutant background (Mitsuda et al., 2005; Zhong et al., 2006; Mitsuda et al., 2007; Yamaguchi et al., 2008). Further analyses of other genes that harbor the SND1 binding motif may result in new insights into the transcriptional control of secondary wall biosynthesis.



(a)

MYB32-promoter

(b) Effector

SND1 SND1m

Reporter

MBP Compe tor

(c)

CaMV 35SPro MYB s SND1 Pro

1.2

Firefly luciferase ProSND1

1 0.8 Complex

0.6 0.4 0.2 0

Free probe

(d)

(e)

(f) 10

(g)

1.2

8 0.8

6 4

0.4 2 0

0 CK OE-1 OE-2

CK OE-1 OE-2

Figure 8. SND1 binds to the MYB32 promoter, and three MYB transcription factors negatively regulate SND1 promoter activity. (a) Electrophoretic mobility shift assay (EMSA) showing the direct binding of SND1 to the MYB32 promoter. The arrowhead indicates a non-specific band. (b) Diagram showing the constructs used in the trans-activation assay. (c) SND1 promoter activity following co-transfection with different MYB TFs. Bars represent SDs from three independent biological replicates. (d) Representative MYB32 overexpression transgenic plant (upper part) and a wild-type seedling (lower part, picture was taken 3 weeks after germination). (e) Leaves of wild-type plant (left) and MYB32 overexpression plant (right). (f, g) Quantitative RT-PCR results showing the overexpression of MYB32 (f) and the downregulation of SND1 in stem tissues of transgenic plants (g). Bars represent SDs from three independent measurements.

Zhong et al. have recently identified a 19-bp secondary wall NAC binding element termed SNBE (T/A)NN(C/T)(T/C/ G)TNNNNNNNA(A/C)GN(A/C/T) (A/T), which bears significant tolerance to nucleotide substitutions. Statistically, the SNBE sequence could be present once in every 1.8 kb (Zhong et al., 2010b), which is basically almost in every gene promoter in Arabidopsis. The core motif we propose is located in the 19-bp SNBE sequence, but is more stringent than the SNBE motif, and is found in 647 gene promoters in Arabidopsis. Among the genes that contain this element are MYB46 and MYB32, as well as four other genes, At1g655570, At2g14620, At2g21300 and At5g47000, all of which were previously identified as direct targets of SND1 (Zhong et al., 2010b). The SND1 binding site in the SND1 promoter possesses our core motif, but this does not totally match the proposed SNBE element. There is therefore a possibility that some NAC domain protein binding sites may differ from either our core SND1 binding motif or the less stringent SNBE element, and that many or most of the genes containing the TACXTTXXXXATGA motif (Table S2) are not regulated by NAC transcription factors.

NST1 expression is paradoxically elevated in the Medicago nst1-3 mutant line. This phenomenon can be explained by the following model. Expression of SND1 is under both positive regulation by its own translation product and under negative regulation by downstream MYB TFs; these effects are presumably balanced until the balance is broken by a mutation or altered environmental factors. However, the change in NST1 transcript can only be detected if the mutation is a point mutation. The elevated transcript level of MtNST1 in the point mutant background also suggests that the positive autoregulation is overridden by negative regulators under normal circumstances. Fine tuning of NST1 expression by balancing activation and repression may enable plants to adapt to ever changing environmental conditions. The increased expression of NST1 transcripts in the point mutation might arise from the action of a yet to be identified upstream signal, which can turn on the expression of NST1 and NST3/SND1. That such a signal exists is supported by the observation of correct temporal and spatial expression of NST1 promoter:GUS and NST3 promoter:GUS constructs


1112 Huanzhong Wang et al. in the nst1:nst3 double mutant of Arabidopsis (Mitsuda et al., 2007). The existence of signals upstream of SND1 is also supported by the analysis of another interfascicular fiber-related mutant, wat1 (walls are thin 1) (Ranocha et al., 2010). WAT1 is a tonoplast protein that functions upstream of SND1: the knock-out of WAT1 affects fiber development, but how this protein regulates SND1 is still not clear. In our model, the upstream signal will turn on the NST1 promoter in the absence of functional NST1 protein, and therefore without trans-activation of MYB46, which itself trans-activates expression of the three MYB TFs that act as negative regulators of SND1.

scanning were conducted according to the manufacturer’s instructions (Affymetrix, http://www.affymetrix.com). The normalization of data was achieved by the robust multi-chip average (RMA) procedure (Irizarry et al., 2003). The presence/absence call for each probe set was obtained from dCHIP (Li and Wong, 2001). Genes significantly differentially expressed between controls and mutants were selected using Associative Analysis (Dozmorov and Centola, 2003). The type-I family-wise error rate was reduced using a Bonferronicorrected P-value threshold of 0.05/n, where n represents the number of genes present on the chip. The false discovery rate was monitored and controlled by Q-value (false discovery rate), calculated using EDGE (extraction of differential gene expression; http:// www.biostat.washington.edu/software/jstorey/edge; Storey and Tibshirani, 2003; Leek et al., 2006).

EXPERIMENTAL PROCEDURES

Protoplast isolation and trans-activation assay

Plant materials and growth conditions

Arabidopsis protoplasts were isolated according to a previously published protocol, with minor modifications (Sheen, 2001; Asai et al., 2002). In brief, leaves from healthy 30-day-old Arabidopsis plants were cut into 0.5–1-mm strips with fresh razor blades. The leaf strips were put into an enzyme solution containing cellulase and macerozyme, then vacuum infiltrated for 5–30 min, followed by digestion for 3 h without shaking in the dark. The protoplasts were filtered with a 35- to 75-lm nylon mesh, collected, and transformed by PEG-mediated transfection. To construct the effector constructs, coding sequences of SND1 and MYB TFs were inserted after the 35S promoter of the Gateway overexpression vector P2GW7 (http:// gateway.psb.ugent.be). To construct the reporter constructs, promoters of SND1, MYB46 and secondary cell wall synthetic genes were cloned to the vector P2GWL7, which was constructed by the ligation of a SacI/SacII cleaved pPGWL7 fragment, including the firefly luciferase reporter gene, to the SacI/SacII cleaved P2GW7 backbone fragment. Primers used for cloning the TF coding and promoter sequences are listed in Table S3 online. Promoter activities are expressed as Firefly LUC/Renilla LUC activities, and normalized to the value obtained from protoplasts transformed with empty effector vector.

A population of 9000 Nicotiana tabacum (tobacco) Tnt1 retrotransposon tagged mutants of M. truncatula (Tadege et al., 2005, 2008) was screened for defects in secondary cell wall formation. Plants were grown at 24C day/20C night, with a 16-h day/8-h night photoperiod (150 lmol m)2 sec)1) and 70–80% relative humidity. The sixth internodes counting from the top of each plant were harvested when the plants had reached around eight internodes, and were stored at )80C. Cross sections (100 lm) of the sixth internodes were cut with a Leica RM 2255 microtome (Leica, http://www. leica.com). Micrographs were taken under a Nikon Micophot-FX system (http://www.nikon.com) with a Nikon DXM 1200 color camera with consistent settings.

Constructs and plant transformation To produce the MYB32 overexpression construct, the coding sequence of MYB32 was cloned to pENTR/D vector and sequenced. The resulting vector was used in an linear recombination (LR) reaction using LR Clonase II (Invitrogen, http://www.invitrogen.com) to insert the coding sequence into the destination vector pB7WG2D (Karimi et al., 2002). Plants were transformed by Agrobacterium-mediated transformation (Clough and Bent, 1998). Seeds were selected on plates supplied with 7.5 mg L)1 glufosinate ammonium (Sigma-Aldrich, http://www.sigmaaldrich.com). Resistant plants were transferred to freshly-prepared soil. From a total of 63 plants, nine showed small stature and were sterile. Two representative mutant plants were used for taking pictures and gene expression analyses. To analyze the subcellular localization of SND1 and SND1(T94K), the coding sequences of both the wild type and the point mutation sequence were cloned as in-frame C-terminal fusions with YFP, and the CaMV 35S promoter was used to drive both constructs. The sequence-confirmed constructs were transformed to AGL1 Agrobacterium cells for transient expression in tobacco leaf epidermal cells by infiltration. The YFP signal was detected by confocal microscopy 24 h after infiltration, and representative pictures were taken for documentation.

Microarray analysis Total RNA samples from fifth to eighth internodes were subjected to Affymetrix microarray analysis. Segregating progeny without loss of lignification in interfascicular tissues were used as controls. RNA was isolated with Tri-reagent according to the manufacturer’s protocol (Invitrogen). RNA was cleaned and concentrated using the RNeasy MinElute Cleanup kit (Qiagen, http://www.qiagen.com). Ten micrograms of purified RNA from three biological replicates was used for microarray analysis. Probe labeling, hybridization and

Protein expression and EMSA To express the recombinant NAC domain of the SND1 protein, a 191 amino acid peptide from the N terminus, including the entire NAC domain, was fused in frame with maltose binding protein (MBP), and expressed in E. coli. The primers used for cloning the NAC domain were NACFw, 5¢-CACCGAATTCATGGCTGATAATAA GGTCAATCTTTCG-3¢, and NACRe, 5¢-ATCTAGATTATAGAGT GATTTTAGGACAATCGTCAATC-3¢. The PCR product was cloned to pENTR/D-TOPO vector (Invitrogen), confirmed by sequencing, and subcloned to pMAL-C2X (New England Biolabs, http://www.neb.com). The resulting construct was transformed to E. coli strain BL21(DE3). To express the peptide, E. coli (100 ml) was grown at 37C until OD600 = 0.6, then 0.5 lM isopropyl-b-D-thiogalactopyranoside (IPTG) was added and the cells were grown for an additional 16 h at 16C. Bacteria were collected by centrifugation at 6000 g for 15 min and stored at )20C until being used for protein isolation and purification using amylose resin. The resulting protein was used for EMSA with the MYB46 or SND1 promoter fragments. The promoter fragments were PCR amplified with one 5¢ biotinlabeled primer and purified using a PCR purification kit (Invitrogen). Primers used for cloning promoter fragments P1–P3 were: P1Fw, 5¢TGGTATTCGCATAACCATCGAATAC-3¢; P1Re, 5¢-TCACACACTCGT TTCATGATCTACG-3¢; P2Fw, 5¢-TGACGTAGATCATGAAACGAGT GTG-3¢; P2Re, 5¢-TGATCATACGTAAGTTCTGACGTAAG-3¢; P3Fw, 5¢ACATTACACATGACATGACATTACACG-3¢; and P3Re, 5¢-TTGACCT


NST1 function and regulation 1113 TATTATCAGCCATTAACG-3¢. Primers used for cloning the MYB32 promoter fragment were: ProMYB32BDF, 5¢-TCGTTGCAGTAC GAAACCATCCAACTTTG-3¢; and PROMYB32BDR, 5¢-TGAAACCA TAATTATGAATGGCGACATGC-3¢. The biotin-labeled DNA fragments were incubated for 20 min with 100 ng of SND1-MBP in binding buffer, according to the protocol for the EMSA kit (Pierce, http://www.piercenet.com). The reaction mixture was then loaded for polyacrylamide gel electrophoresis at 100 V for 1 h until the front reached one-third to the bottom of the gel. The DNA was electroblotted onto nitrocellulose membrane and detected by the chemiluminescent method.

Genome-wide search for promoter elements Promoter sequences (3 kb) were downloaded from TAIR (TAI R9_upstream_3000_translation_start_20090619; ftp://ftp.arabidopsis. org/home/tair/Sequences/blast_datasets/TAIR9_blastsets/database). Then, a PERL script was used to search the SND1 binding motif pattern, TACNTTNNNNATGA, from the 3-kb promoter sequences for all genes.

Real-time PCR Quantitative real-time PCR (qRT-PCR) and the calculation of relative expression were performed as described previously (Wang et al., 2010). In brief, cDNA samples were used for qRT-PCR with technical duplicates. The 10-ll reaction included 2 ll of primers (0.5 lM of each primer), 5 ll of Power Sybr (Applied Biosystems, http:// www.appliedbiosystems.com), 2 ll 1:20 diluted cDNA from the reverse transcription step, and 1 ll of water. qRT-PCR data were analyzed using SDS 2.2.1 (Applied Biosystems). Transcript levels were determined by relative quantification (Pfaffl, 2001) using the Arabidopsis ubiquitin gene as a reference.

ACKNOWLEDGEMENTS We thank Drs Patrick Zhao and Hui Shen for their critical reading of the manuscript. The M. truncatula plants used in this research project, which are jointly owned by the Centre National de la Recherche Scientifique and the Samuel Roberts Noble Foundation, Ardmore, OK, USA, were created through research funded, in part, by grant number 703285 from the National Science Foundation. This work was supported by grants to RAD from the USDA-DOE Bioenergy Feedstock Genomics program and the Oklahoma Department of Energy Bioenergy Center (OBC), and by the Samuel Roberts Noble Foundation.

SUPPORTING INFORMATION Additional Supporting Information may be found in the online version of this article: Figure S1. Localization of full-length SND1-GFP fusions. Figure S2. SDS-PAGE gel analysis showing the proteins used in electrophoretic mobility shift assay (EMSA) analysis. Figure S3. Electrophoretic mobility shift assay (EMSA) showing loss of binding of SND1mu to the SND1 promoter P2 fragment. Figure S4. Competition between P2-5 and the biotin-labeled P3 fragment. Table S1. Genes with expression levels changed more than twofold in nst1-3 mutant plants. Table S2. List of Arabidopsis genes containing the TACXTTXXXXATGA motif in their promoter regions. Table S3. Primers used to clone the transcription factors and promoters for the effector and reporter constructs. Please note: As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer-reviewed and may be re-organized for online

delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

REFERENCES Asai, T., Tena, G., Plotnikova, J., Willmann, M.R., Chiu, W.L., Gomez-Gomez, L., Boller, T., Ausubel, F.M. and Sheen, J. (2002) MAP kinase signalling cascade in Arabidopsis innate immunity. Nature, 415, 977–983. Clough, S.J. and Bent, A.F. (1998) Floral dip: a simplified method for Agrobacterium-mediated transformation of Arabidopsis thaliana. Plant J. 16, 735–743. Demura, T. and Ye, Z.H. (2010) Regulation of plant biomass production. Curr. Opin. Plant Biol. 13, 299–304. Dozmorov, I. and Centola, M. (2003) An associative analysis of gene expression array data. Bioinformatics, 19, 204–211. Ernst, H.A., Olsen, A.N., Larsen, S. and Lo Leggio, L. (2004) Structure of the conserved domain of ANAC, a member of the NAC family of transcription factors. EMBO Rep. 5, 297–303. Irizarry, R.A., Bolstad, B.M., Collin, F., Cope, L.M., Hobbs, B. and Speed, T.P. (2003) Summaries of Affymetrix GeneChip probe level data. Nucleic Acids Res. 31, e15. Jin, H., Cominelli, E., Bailey, P., Parr, A., Mehrtens, F., Jones, J., Tonelli, C., Weisshaar, B. and Martin, C. (2000) Transcriptional repression by AtMYB4 controls production of UV-protecting sunscreens in Arabidopsis. EMBO J. 19, 6150–6161. Karimi, M., Inze, D. and Depicker, A. (2002) GATEWAY vectors for Agrobacterium-mediated plant transformation. Trends Plant Sci. 7, 193–195. Ko, J.H., Yang, S.H., Park, A.H., Lerouxel, O. and Han, K.H. (2007) ANAC012, a member of the plant-specific NAC transcription factor family, negatively regulates xylary fiber development in Arabidopsis thaliana. Plant J. 50, 1035–1048. Ko, J.H., Kim, W.C. and Han, K.H. (2009) Ectopic expression of MYB46 identifies transcriptional regulatory genes involved in secondary wall biosynthesis in Arabidopsis. Plant J. 60, 649–665. Lee, J.J., Woodward, A.W. and Chen, Z.J. (2007) Gene expression changes and early events in cotton fibre development. Ann. Bot. 100, 1391–1401. Leek, J.T., Monsen, E., Dabney, A.R. and Storey, J.D. (2006) EDGE: extraction and analysis of differential gene expression. Bioinformatics, 22, 507–508. Li, C. and Wong, W.H. (2001) Model-based analysis of oligonucleotide arrays: expression index computation and outlier detection. Proc. Natl Acad. Sci. USA, 98, 31–36. McCarthy, R.L., Zhong, R. and Ye, Z.H. (2009) MYB83 is a direct target of SND1 and acts redundantly with MYB46 in the regulation of secondary cell wall biosynthesis in Arabidopsis. Plant Cell Physiol. 50, 1950–1964. Mitsuda, N., Seki, M., Shinozaki, K. and Ohme-Takagi, M. (2005) The NAC transcription factors NST1 and NST2 of Arabidopsis regulate secondary wall thickenings and are required for anther dehiscence. Plant Cell, 17, 2993–3006. Mitsuda, N., Iwase, A., Yamamoto, H., Yoshida, M., Seki, M., Shinozaki, K. and Ohme-Takagi, M. (2007) NAC transcription factors, NST1 and NST3, are key regulators of the formation of secondary walls in woody tissues of Arabidopsis. Plant Cell, 19, 270–280. Mummenhoff, K., Polster, A., Muhlhausen, A. and Theissen, G. (2009) Lepidium as a model system for studying the evolution of fruit development in Brassicaceae. J Exp Bot. 60, 1503–1513. Ohashi-Ito, K., Oda, Y. and Fukuda, H. (2010) Arabidopsis VASCULARRELATED NAC-DOMAIN6 directly regulates the genes that govern programmed cell death and secondary wall formation during xylem differentiation. Plant Cell, 22, 3461–3473. Olsen, A.N., Ernst, H.A., Leggio, L.L. and Skriver, K. (2005) NAC transcription factors: structurally distinct, functionally diverse. Trends Plant Sci. 10, 79– 87. Pfaffl, M.W. (2001) A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res. 29, e45. Pyo, H., Demura, T. and Fukuda, H. (2007) TERE; a novel cis-element responsible for a coordinated expression of genes related to programmed cell death and secondary wall formation during differentiation of tracheary elements. Plant J. 51, 955–965. Ranocha, P., Denance, N., Vanholme, R. et al. (2010) Walls are thin 1 (WAT1), an Arabidopsis homolog of Medicago truncatula NODULIN21, is a


1114 Huanzhong Wang et al. tonoplast-localized protein required for secondary wall formation in fibers. Plant J. 63, 469–483. Sheen, J. (2001) Signal transduction in maize and Arabidopsis mesophyll protoplasts. Plant Physiol. 127, 1466–1475. Shen, H., Yin, Y., Chen, F., Xu, Y. and Dixon, R. (2009) A bioinformatic analysis of nac genes for plant cell wall development in relation to lignocellulosic bioenergy production. BioEnergy Research, 2, 217–232. Storey, J.D. and Tibshirani, R. (2003) Statistical significance for genomewide studies. Proc. Natl Acad. Sci. USA, 100, 9440–9445. Tadege, M., Ratet, P. and Mysore, K.S. (2005) Insertional mutagenesis: a Swiss Army knife for functional genomics of Medicago truncatula. Trends Plant Sci. 10, 229–235. Tadege, M., Wen, J., He, J. et al. (2008) Large-scale insertional mutagenesis using the Tnt1 retrotransposon in the model legume Medicago truncatula. Plant J. 54, 335–347. Thompson, J.D., Higgins, D.G. and Gibson, T.J. (1994) CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 22, 4673–4680. Wang, H., Avci, U., Nakashima, J., Hahn, M.G., Chen, F. and Dixon, R.A. (2010) Mutation of WRKY transcription factors initiates pith secondary wall formation and increases stem biomass in dicotyledonous plants. Proc. Natl Acad. Sci. USA, 107, 22338–22343. Xie, Q., Frugis, G., Colgan, D. and Chua, N.H. (2000) Arabidopsis NAC1 transduces auxin signal downstream of TIR1 to promote lateral root development. Genes Dev. 14, 3024–3036. Yamaguchi, M., Kubo, M., Fukuda, H. and Demura, T. (2008) Vascular-related NAC-DOMAIN7 is involved in the differentiation of all types of xylem vessels in Arabidopsis roots and shoots. Plant J. 55, 652–664.

Yamaguchi, M., Mitsuda, N., Ohtani, M., Ohme-Takagi, M., Kato, K. and Demura, T. (2011) VASCULAR-RELATED NAC-DOMAIN7 directly regulates the expression of a broad range of genes for xylem vessel formation. Plant J. 66, 579–590. Zhao, Q., Gallego-Giraldo, L., Wang, H., Zeng, Y., Ding, S., Chen, F. and Dixon, R. (2010) A NAC transcription factor orchestrates multiple features of cell wall development in Medicago truncatula. Plant J. 63, 100–114. Zhong, R. and Ye, Z.H. (2007) Regulation of cell wall biosynthesis. Curr. Opin. Plant Biol. 10, 564–572. Zhong, R., Demura, T. and Ye, Z.H. (2006) SND1, a NAC domain transcription factor, is a key regulator of secondary wall synthesis in fibers of Arabidopsis. Plant Cell, 18, 3158–3170. Zhong, R., Richardson, E.A. and Ye, Z.H. (2007) The MYB46 transcription factor is a direct target of SND1 and regulates secondary wall biosynthesis in Arabidopsis. Plant Cell, 19, 2776–2792. Zhong, R., Lee, C., Zhou, J., McCarthy, R.L. and Ye, Z.H. (2008) A battery of transcription factors involved in the regulation of secondary cell wall biosynthesis in Arabidopsis. Plant Cell, 20, 2763–2782. Zhong, R., Lee, C. and Ye, Z.H. (2010a) Evolutionary conservation of the transcriptional network regulating secondary cell wall biosynthesis. Trends Plant Sci. 15, 625–632. Zhong, R., Lee, C. and Ye, Z.H. (2010b) Global analysis of direct targets of secondary wall NAC master switches in Arabidopsis. Mol. Plant, 3, 1087– 1103. Zhou, J., Lee, C., Zhong, R. and Ye, Z.H. (2009) MYB58 and MYB63 are transcriptional activators of the lignin biosynthetic pathway during secondary cell wall formation in Arabidopsis. Plant Cell, 21, 248–266.
