- Top of page
- Supporting Information
During the life cycles of conifer trees, such as maritime pine (Pinus pinaster Ait.), large quantities of carbon skeletons are irreversibly immobilized in the wood. In energetic terms this is an expensive process, in which carbon from photosynthesis is channelled through the shikimate pathway for the biosynthesis of phenylpropanoids. This crucial metabolic pathway is finely regulated, primarily through transcriptional control, and because phenylalanine is the precursor for phenylpropanoid biosynthesis, the precise regulation of phenylalanine synthesis and use should occur simultaneously. The promoters of three genes encoding the enzymes prephenate aminotransferase (PAT), phenylalanine ammonia lyase (PAL) and glutamine synthetase (GS1b) contain AC elements involved in the transcriptional activation mediated by R2R3-Myb factors. We have examined the capacity of the R2R3-Myb transcription factors Myb1, Myb4 and Myb8 to co-regulate the expression of PAT, PAL and GS1b. Only Myb8 was able to activate the transcription of the three genes. Moreover, the expression of this transcription factor is higher in lignified tissues, in which a high demand for phenylpropanoids exits. In a gain-of-function experiment, we have shown that Myb8 can specifically bind a well-conserved eight-nucleotide-long AC-II element in the promoter regions of PAT, PAL and GS1b, thereby activating their expression. Our results show that Myb8 regulates the expression of these genes involved in phenylalanine metabolism, which is required for channelling photosynthetic carbon to promote wood formation. The co-localization of PAT, PAL, GS1b and MYB8 transcripts in vascular cells further supports this conclusion.
- Top of page
- Supporting Information
Wood is traditionally among the most important commercial products because of the high demand that exits for its derivatives, such as paper, fibres and building materials. Recent developments in biotechnology have inspired novel applications for this traditional raw material, such as its use for biofuel production, and these demands have greatly increased the pressure to improve forest productivity.
Wood (secondary xylem) formation represents a clear example of cell differentiation that is controlled by a wide range of internal and external factors. The regulation of the process itself involves the coordinated expression of various genes involved in cell wall synthesis, cell differentiation and programmed cell death. Owing to the diversity of gene families and the plasticity of wood formation, this complex process is difficult to study (Plomion et al., 2001).
Trees, including conifers, divert large quantities of carbon into the biosynthesis of phenylpropanoids, particularly to generate lignin, an important constituent of wood (Wagner et al., 2012). After cellulose, lignin is the most abundant organic compound on Earth, and comprises 30% of the plant biomass (Higuchi, 1990; Boerjan et al., 2003). Because lignin biosynthesis is one of the most energy demanding biosynthetic pathways in plants (Amthor, 2003), it is precisely regulated. Consequently, tree trunks act as powerful sinks in which considerable percentages of the carbon and energy derived from photosynthesis are immobilised.
Although lignin and other phenolic compounds do not contain nitrogen, phenylalanine metabolism is required to channel photosynthesis-derived carbon to phenylpropanoid biosynthesis (Figure 1). Recent genetic evidence regarding the pathway itself and the genes involved has indicated that the arogenate pathway is the predominant route for phenylalanine biosynthesis in plants (Maeda and Dudareva, 2012). Chorismate, the final product of the shikimate pathway, generates prephenate, which is subsequently aminated to arogenate in a reaction that is catalysed by prephenate aminotransferase (PAT). Until recently, the identity of the gene that encodes this enzyme was unkown. Graindorge et al. (2010) and Maeda et al. (2011) independently discovered that PAT activity was housed by a previously described prokaryotic-type plastidic aspartate aminotransferase (de la Torre et al., 2006, 2009).
Figure 1. The channelling of photosynthetic carbon for phenylpropanoid biosynthesis and the associated nitrogen metabolism. The central role of the phenylalanine metabolism is highlighted. Critical steps in the pathway are indicated, including amination, deamination and nitrogen recycling. The enzymes PAT, PAL and GS1b are circled.
Download figure to PowerPoint
The phenylpropane skeleton required for lignin biosynthesis is provided by the deamination of phenylalanine in the reaction catalysed by the enzyme phenylalanine ammonia-lyase (PAL), producing a molecule of cinnamic acid and a molecule of ammonium (Figure 1; Higuchi, 1997). This reaction is quantitatively important in trees because lignin biosynthesis is required for wood formation, and it releases large quantities of ammonium (Humphreys and Chapple, 2002).
The sequential action of the enzymes glutamine synthetase (GS) and glutamate synthase (GOGAT) is crucial for ammonium re-assimilation and glutamate biosynthesis (Figure 1). Nitrogen recycling can satisfy the demand for nitrogen produced during vegetative growth, and it therefore contributes to efficient nitrogen use (Cantón et al., 2005). In conifers, two genes encoding distinct GS isoforms have been described (Ávila et al., 2001). One of these genes, GS1b, is specifically expressed in vascular bundles, and is potentially involved in the recycling of ammonia derived from phenylalanine during lignin synthesis (Cánovas et al., 2007).
It is well known that the levels of lignin biosynthesis enzymes are primarily regulated at the transcriptional level, and that the presence of AC elements in the regulatory regions of these genes is necessary and sufficient to drive localized xylem expression (Lacombe et al., 2000; Wagner et al., 2012). The R2R3Myb transcription factor family is one of the largest transcription factor families in plants, with more than 120 members in Arabidopsis (Riechmann et al., 2000). Some of these genes are involved in the regulation of lignin biosynthesis genes (Stracke et al., 2001). It has been shown that R2R3Myb factors are able to bind the AC elements present in the promoter regions of genes encoding lignin biosynthesis enzymes to activate transcription (Patzlaff et al., 2003a,b). In fact, the presence of AC elements in these gene promoters is almost universal, and suggests coordinated regulation of the set of genes involved in the process (Rogers and Campbell, 2004).
An efficient and coordinated pathway for the amination of prephenate and the deamination of phenylalanine should be operative in lignifying cells to provide phenylalanine for lignin biosynthesis, and to re-assimilate ammonium because this cyclical re-assimilation maintains the efficiency of nitrogen use in these long-lived organisms (Figure 1). We hypothesized that one way to ensure efficient photosynthetic carbon channelling for lignin and other phenylpropanoid biosynthesis, together with nitrogen recycling, would be to couple both processes in time and in space by transcriptionally regulating the genes involved in phenylalanine biosynthesis and use.
The experiments described in this article attempt to test this hypothesis. To this end, we have isolated the promoter region of the three genes involved in the phenylalanine pathway in Pinus pinaster: PAL, GS1b and PAT. We have conducted both in vitro and in vivo studies using three different Myb transcription factors: PtMyb1, PtMyb4 from Pinus taeda and PpMyb8 from P. pinaster. We have studied the possible coupling in space and time of gene products for the operative co-regulation of both processes in pine trees, and have proven that Myb8 is a potential candidate to be the transcriptional regulator of phenylalanine metabolism in P. pinaster vascular cells.
- Top of page
- Supporting Information
The understanding of the transcription regulatory network associated with key processes in conifer trees is crucial for future applications in tree improvement and sustainable forest management. In the present work, the transcriptional regulation of the phenylalanine pathway was studied in P. pinaster. We isolated the promoters of three genes involved in phenylalanine metabolism in maritime pine: PAL, GS1b and PAT (Figure 1). The structure of the PAL regulatory region in P. pinaster was found to be similar to the previously described PAL promoter in P. taeda (Osakabe et al., 2009); however, according to the primary structure of the full-length cDNA, the TSS was mapped to a location 187 bp upstream from the ATG. In contrast, the TSS in the P. taeda PAL regulatory region was positioned 22 bp upstream of the ATG. The structures of the GS1b promoters in P. pinaster and P. sylvestris (Gómez-Maldonado et al., 2004a) were almost identical, with only two single nucleotide polymorphisms.
Functional significance of AC-II elements in the PAL, GS1b and PAT promoters
The in silico analysis allowed us to identify the presence of AC elements in the promoters of the three genes under study; however, the AC-II class was the only class that was commonly present in the three genes (Figure 2). The relative positions of these AC-II elements relative to their respective TSSs were not conserved. Thus, the elements were found to be distributed along the PAL promoter, covering the proximal and distal portions, whereas they were specifically located in the proximal region of the GS1b and PAT gene promoters. AC conserved sequences are also found in the promoters of genes involved in lignin biosynthesis in angiosperms (Hauffe et al., 1993; Hatton et al., 1995; Raes et al., 2003; Zhou et al., 2009). A recent review compared the interactions between Myb proteins and their target DNA binding sites in plants and animals, and analysed the consensus sequences of the AC elements (Prouse and Campbell, 2012). Most plant Myb proteins display considerable flexibility in recognising their target sites. Thus, Myb1 and Myb4 proteins were able to bind AC-I and AC-II elements present in the PsGS1b promoter; however, the stronger shift was observed when Myb1 bound the AC-II element as the target site (Gomez-Maldonado et al., 2004b). This result was also confirmed in the present study: the Myb1, Myb4 and Myb8 proteins were able to bind PAL Box 2 and Box 4, GS1b Box 1 and PAT Box 1, but they displayed different affinities (Figure 3). However, an in vivo transient expression analysis revealed a different picture, where a particular Myb protein differentially affected the expression of each gene under study. Thus, the PAL promoter was exclusively transactivated by Myb8. In contrast Myb1, Myb4 and Myb8 activated GS1b expression, and Myb4 and Myb8 activated PAT expression (Figure 6). These data indicate that Myb8 was the only Myb transcription factor that could transactivate the PAL, GS1b and PAT genes. All of these findings suggest that besides a common function in monolignol biosynthesis, the individual gene products contribute to other metabolic pathways separately, they are likely to display differential regulatory specificities by the Myb proteins; for example, GS1b in the biosynthesis of glutamine for nitrogen transport in vascular cells (Ávila et al., 2001); and PAT in the biosynthesis of amino acids in the aspartate metabolic pathway (de la Torre et al., 2009).
The accumulation patterns of the PAL, GS1b, PAT and Myb8 transcripts support a role of these genes in the lignified tissues of maritime pine
The comparative expression profiles of genes involved in phenylalanine metabolism revealed high transcript levels of PAL, GS1b and PAT in the stems and roots of young trees, and in the compression wood of adult trees, therefore reflecting an essential role of this metabolic pathway in tissues that are undergoing lignification. Maximal transcript abundances for Myb1, Myb4 and Myb8 were also observed in pine stems, but only Myb8 was almost exclusively expressed in the stems of seedlings and young trees (Figure 4b). Enhanced gene expression levels were also observed for Myb4 and Myb8 in maritime pine compression wood (Figure 4d), which is consistent with the Myb8 expression pattern that was previously reported in white spruce compression wood (Bedon et al., 2007); however, the observed enhancement was greater for Myb8. Increased lignin deposition and the upregulation of genes encoding enzymes of the monolignol biosynthetic pathway in maritime pine compression wood have been recently reported (Villalobos et al., 2012).
Myb4 appears to be a transcriptional activator of genes involved in lignin biosynthesis in loblolly pine, and the overexpression of Myb4 resulted in ectopic lignification in tobacco (Patzlaff et al., 2003b) and Arabidopsis (Newman et al., 2004).
Previous studies have investigated the involvement of PtMyb1 and PtMyb8 in regulating secondary xylem. The authors overexpressed both Myb genes in spruce (Bomal et al., 2008), and their data clearly showed the impact of PtMyb8 in secondary cell wall biogenesis and in lignin deposition during the formation of compression wood; these results were confirmed by the current study. However, Myb1 overexpression was shown to have a lower impact in the biogenesis of secondary xylem, despite its putative role in lignin deposition (Bomal et al., 2008). These findings are consistent with the similar expression levels observed for Myb1 in the compression and opposite wood of maritime pine (Figure 4d) and white spruce (Bedon et al., 2007).
The sequence 5′-CCAACCAC/A-3′ functions as a regulatory element in the presence of Myb8
Promoter activity is generally the result of the combined effects of many cis elements. These sites mediate intercellular or intracellular stimuli, developmental signals or other functions as particular regulators of metabolic pathways. To identify cis elements mediating Myb8 transcriptional activation, a gain-of-function experiment was performed. We demonstrated that an 8-bp conserved sequence is sufficient to mediate the Myb8 response, and is necessary for the transcriptional activation of the PAL, GS1b and PAT genes. Relatively few of the possible plant R2R3Myb DNA targets have been characterized to date (Prouse and Campbell, 2012). In the current work, a consensus DNA target for Myb8 was derived from both in vitro and in vivo assays. To our knowledge, this is the only transient expression experiment conducted in woody protoplasts that demonstrates the specificity of a Myb protein for a cis element localised in three different promoters, and that demonstrates that this sequence is sufficient to activate the expression of a reporter gene under the control of a minimal promoter.
Maritime pine Myb8 co-regulates prephenate amination, phenylalanine deamination and nitrogen recycling
Considering the major contribution of PAT towards phenylalanine biosynthesis, and also the importance of this pathway in trees, we have contrasted the hypothesis of the transcriptional co-regulation of three genes involved in the process. In Figure 8, a metabolic scheme illustrates the pathways in which the three gene products are involved. Because prephenate amination occurs in the plastid, phenylalanine deamination occurs in the cytosol, and nitrogen recycling via the GS/GOGAT cycle involves enzymes that are located in the two subcellular compartments, at least the two following amino acid transporters are necessary 15 to make the route operative: i) a glutamine translocator and ii) a phenylalanine transporter. The presence of a functional glutamine/glutamate translocator in pine cells has been reported recently (Claros et al., 2010), and its specificity and activity facilitate glutamine import into the chloroplast and glutamate export to the cytosol to avoid nitrogen draining from this essential pathway in plants. It is also simultaneously necessary to translocate phenylalanine to the cytosol for phenylpropanoid biosynthesis. The mechanisms that dictate phenylalanine availability in the cytosol of lignifying cells are poorly understood. In any case, the activity of a phenylalanine transporter in the plastid membrane is necessary to satisfy the great demand of phenylalanine in the cytosol of cells undergoing lignification.
Figure 8. Proposed model of co-regulation of PAL, GS1b and PAT by Myb8 in pine cells. The PAL, GS1b and PAT enzymes are compartmentalized in the cytosol and plastids of pine cells. The functionality of the pathway requires specific membrane transporters (black circles). The photosynthetic carbon flux is highlighted with thick arrows channelling carbon from the shikimate pathway to phenylalanine, and finally to lignin. In the process, the level of ammonium produced that is quantitatively important is also indicated with a thick arrow. The three enzymes included in this study are encircled and marked with asterisks.
Download figure to PowerPoint
In this scenario, the simultaneous co-activation of the pathway enzymes would represent a major operational advantage by directing carbon through phenylalanine biosynthesis to promote lignin biosynthesis in different metabolic situations such as regular development or increased lignin production during wood formation.
The role of R2R3Myb transcription factors has been shown to be essential in many aspects of plant secondary metabolism, particularly in the phenylpropanoid metabolism (Stracke et al., 2001). The in vitro and functional interactions described here extend this regulatory role to a critical point at which nitrogen exchange is required to maintain the photosynthetic carbon flux for wood formation in pine trees. Our experimental approaches using different Myb proteins have revealed that despite a possible redundancy in Myb protein function, Myb8 has a prominent role in the transcriptional regulation of essential steps in phenylalanine metabolism. The co-localization studies of gene expression in pine vascular cells further support a role of Myb8 in the coordinated regulation of the phenylalanine pathway. Moreover, a recent study demonstrated that asparagine synthetase expression is downregulated by Myb proteins in maritime pine, which likely prevents the collapse of lignin biosynthesis during wood formation (Canales et al., 2012). In this metabolic context, ammonium assimilated by the GS/GOGAT cycle must be redirected towards arogenate synthesis to maintain aromatic amino acid supplies in lignifying cells.
The maritime pine seeds (Pinus pinaster Ait.) used in all of the experiments were provided by Centro de Recursos Genéticos Forestales ‘El Serranillo’ (Ministerio de Medio Ambiente y Medio Rural y Marino, Spain). The seed germination and growth of the seedlings were performed as previously described (Cánovas et al., 1991).
For the assays, the plants were pooled, and samples were fractionated in the roots, stems and cotyledons. All of the samples were immediately frozen in liquid nitrogen and stored at −80°C until processing. Seedlings of 1.5 cm in length were used for the transient expression protoplast assays. The stems of these plants were used to produce the protoplasts.
Cloning of PpMYB8 cDNA for overexpression and purification
To clone full-length PpMyb8, RNA was extracted from roots of 60-day-old seedlings following the protocol described by Liao et al. (2004). The RNA concentration and purity were determined spectrophotometrically, and only samples with an A260/A280 nm wavelength ratio of 1.9–2.1 and A260/A230 nm wavelength ratio of more than 1.7 were used for the subsequent experiments. We designed two primers based on the P. taeda Myb8 (DQ399057.1) sequence that included ATTP sites in order to obtain GATEWAY-compatible inserts. The primers used for PCR amplification are described in Table S1 as follows: Myb8 ATTP up and Myb8ATTP down. The PCR reaction was performed with the high-fidelity enzyme AccuSure DNA Polymerase from Bioline (http://www.bioline.com), with the following programme: 94°C for 5 min (1 cycle); 94°C for 30 sec, 45°C for 30 sec and 72°C for 2 min (3 cycles); 94°C for 30 sec, 50°C for 45 sec and 72°C for 2 min (35 cycles); and 72°C for 7 min (1 cycle). The blunt-end PCR products were cloned into the pDON201 plasmid and used with the GATEWAY SYSTEM (Invitrogen, http://www.invitrogen.com).
The full-length cDNA encoding PpMyb8 (accession FN868598) was subcloned into the pDEST17 plasmid by recombination with the GATEWAY System (Invitrogen). The regions flanking the insert were sequenced before the plasmid was used for expression. The Escherichia coli strain BL21 (DE3) was used to express the recombinant protein. The bacterial growth and protein induction were performed according to the manufacturer's instructions (Novagen, now EMD Millipore, http://www.emdmillipore.com; see Figure S2a). Recombinant protein production was induced at 37°C for 2 h in the presence of 0.4 mm isopropyl-β-d-thiogalactoside (IPTG). The N-terminus 6XHis-tagged protein was purified by affinity chromatography using an Ni-NTA column (Qiagen, http://www.qiagen.com) under denaturing conditions. The purification was verified by SDS-PAGE analysis and visualization by Coomassie blue staining (Figure S2b).
Isolation of the promoter regions of the PAL, GS1b and PAT genes
The promoter regions of the P. pinaster PAL, GS1b and PAT genes were amplified by PCR walking using the protocol described by Devic et al. (1997) and following the specifications of the Universal Genome Walker kit, as indicated by Clontech (http://www.clontech.com). The transcriptional start site (+1) in the promoter regions was determined using RLM-RACE using the FirstChoice RLM-RACE Kit (Ambion, now Invitrogen, http://www.invitrogen.com). The 5′ RACE oligonucleotides were designed from the P. pinaster cDNA sequences contained in the EuroPine DB (Fernández-Pozo et al., 2011).
Electrophoretic mobility shift assay
The recombinant proteins PpMyb8 (FN868598), PtMyb1 (AY356372) and PtMyb4 (AY35-6371) were produced as previously described (Gómez-Maldonado et al., 2004b). The oligonucleotide sequence probes described in Figure 3 and Figure S3, and containing the AC elements, were generated by annealing complementary oligonucleotides designed to create 5′ overhangs that were filled-in with Klenow DNA polymerase in the presence of [α32P]dCTP for the boxes described in the PAL and GS1b promoters, and [α32P]dTTP for the box described in the PAT promoter; these are detailed in Table S1. At the end of the incubation period, the DNA–protein complexes were analysed by electrophoresis as previously described (Rueda-López et al., 2008). The binding specificity was evaluated using competition experiments with the non-specific competitor polydI-dC (NSC) or treatment with proteinase K (Gómez-Maldonado et al., 2004b).
Total RNA was isolated following the protocol described by Liao et al. (2004), with minor modifications. The RNA samples were treated with RQ1 RNase-free DNase (Promega, http://www.promega.com) to eliminate any traces of genomic DNA, and cDNA synthesis was performed as described previously (Canales et al., 2012). qPCR was performed on an Mx3000P thermal cycler (Stratagene, now Agilent, http://genomics.agilent.com) with a SYBR Premix ExTaq kit (Takara, http://www.takarabioeurope.com/contact.html) under the following conditions: 95˚C for 30 sec (1 cycle), and then 95˚C for 5 sec, 60˚C for 10 sec and 72˚C for 15 sec (40 cycles). After the final cycle, a melting curve analysis was performed over a temperature range of 60−95°C in 0.5°C increments to verify the reaction specificity. Ten nanograms of reverse transcribed cDNA was used as a template for each reaction. The raw fluorescence data from each reaction were fitted with the MAK2 model, which requires no assumptions regarding the amplification efficiency of a qPCR assay (Boggy and Woolf, 2010). The initial target concentrations (D0 parameter) for each gene were deduced from the MAK2 model using the qPCR package of r environment software (Ritz and Spiess, 2008) and normalized to the geometric mean of the reference genes (actin, 40S ribosomal protein and EF-1α).
Transient expression analysis in pine protoplasts
The procedure was performed essentially as previously described by Gómez-Maldonado et al. (2004b). The full-length promoter seque-nces available were used for the reporter constructs in the initial transactivation assays. The GUS activity levels were assayed in pine stem-derived protoplasts that were co-electroporated with a combination of reporter and effector plasmids at a 1:1 molar ratio.
Construction of a reporter plasmid using a minimal promoter
To generate a reporter construct containing four copies of the consensus binding site of the AC-II element, an oligonucleotide was fused to the 35S minimal promoter into the polylinker of the GUS-encoding plasmid vector pBI221 (Jefferson et al., 1987) in frame with the GUS gene by replacing the full CaMV 35S promoter.
RNA probes for in situ hybridization
Pinus pinaster cDNA fragments of GS1b (391 bp) and PAL (487 bp) were subcloned into the pGEM®-3Zf (+) vector (Promega) and used to make riboprobes. The RNA antisense and sense probes of PAT (509-bp fragment) and the Myb8 transcription factor (618-bp fragment) mRNA transcripts were prepared by a PCR-based technique in which a T7 polymerase promoter sequence (5′-TAATACGACTCACTATAGGG-3′) was introduced at the 5′ ends of the gene-specific primers (Young et al., 1991). The PCR fragments were produced under standard PCR conditions using iProof™ High-Fidelity Master Mix (Bio-Rad, http://www.bio-rad.com), and plasmid DNA containing the cDNA of interest served as the template. The PCR fragments were gel purified using the NucleoSpin® Gel and PCR Clean-up kit (Macherey-Nagel, http://www.mn-net.com), and 500 ng was subsequently used as a DNA template for in vitro transcription by T7 RNA polymerase (Promega), incorporating dig-UTP via the DIG RNA labelling mix (Roche, http://www.roche.com). The template DNA was digested with five units of RQ1 RNAse-free DNase (Promega) in a reaction volume of 50 ml for 15 min at 37°C, and the probe was purified with the NucleoSpin® RNA Clean-Up XS kit (Macherey-Nagel). The yield of the DIG-labelled RNA probe was estimated by comparing the intensity of the sample against defined controls created with DIG-labelled control RNA (Roche).
In situ hybridization
Small pieces (approximately 5 mm3) of pine stems were fixed and paraffin-embedded, essentially as described by Cantón et al. (1999). The embedded tissue was sectioned at 10-mm thickness and affixed to poly-l-lysine-coated glass slides. The probe bound to the sections was immunologically detected using a sheep anti-digoxigenin Fab fragment that was covalently coupled to alkaline phosphatase and NBT/BCIP as chromogenic substrates (Roche). For the signal amplification, we used polyvinyl alcohol (PVA) of high molecular weight (100–120 kDa) in the BCIP/NBT detection system. An Eclipse E-800 microscope (Nikon, http://www.nikon.com) was used for sample visualization and photography.