Notice: Wiley Online Library will be unavailable on Saturday 30th July 2016 from 08:00-11:00 BST / 03:00-06:00 EST / 15:00-18:00 SGT for essential maintenance. Apologies for the inconvenience.
•Throughout their lifetimes, plants must coordinate the regulation of various facets of growth and development. Previous evidence has suggested that the Arabidopsis thaliana R2R3-MYB, AtMYB61, might function as a coordinate regulator of multiple aspects of plant resource allocation.
•Using a combination of cell biology, transcriptome analysis and biochemistry, in conjunction with gain-of-function and loss-of-function genetics, the role of AtMYB61 in conditioning resource allocation throughout the plant life cycle was explored.
•In keeping with its role as a regulator of resource allocation, AtMYB61 is expressed in sink tissues, notably xylem, roots and developing seeds. Loss of AtMYB61 function decreases xylem formation, induces qualitative changes in xylem cell structure and decreases lateral root formation; in contrast, gain of AtMYB61 function has the opposite effect on these traits. AtMYB61 coordinates a small network of downstream target genes, which contain a motif in their upstream regulatory regions that is bound by AtMYB61, and AtMYB61 activates transcription from this same motif. Loss-of-function analysis supports the hypothesis that AtMYB61 targets play roles in shaping subsets of AtMYB61-related phenotypes.
•Taken together, these findings suggest that AtMYB61 links the transcriptional control of multiple aspects of plant resource allocation.
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.
AtMYB61 (At1g09540), which encodes a member of the Arabidopsis thaliana R2R3-MYB family of transcription factors, is a gene that potentially controls resource acquisition and allocation. AtMYB61 expression was both sufficient and necessary to bring about reductions in stomatal aperture with consequent effects on gas exchange (Liang et al., 2005). Analysis of loss-of-function atmyb61 mutants showed that AtMYB61 was also necessary for the deposition of seed coat mucilage (Penfield et al., 2001). Other experiments have revealed that AtMYB61 plays a role in the control of lignification and photomorphogenesis (Newman et al., 2004; Dubos et al., 2005).
The link between the involvement of AtMYB61 in the control of stomatal aperture, seed coat mucilage deposition, lignification and photomorphogenesis is not obvious. We show here that AtMYB61 orchestrates changes in transcriptome activity that modify plant resource allocation. Together with our previous results, these new data support the hypothesis that AtMYB61 functions to control both resource acquisition through stomata, as well as resource allocation, largely into nonrecoverable carbon sinks, throughout plant growth and development.
Materials and Methods
Plant material, seed sterilization and growth conditions
All wild-type (WT) and mutant Arabidopsis thaliana (L.) Heynh seeds were in the Columbia-0 background. Plants over-expressing AtMYB61 under the control of the Cauliflower Mosaic Virus 35S promoter (35S::MYB61) were as described previously (Newman et al., 2004; Liang et al., 2005). Similarly, AtMYB61 loss-of-function mutants (atmyb61) have been described previously (Penfield et al., 2001; Liang et al., 2005). Three independently transformed 35S::MYB61 lines and minimally two loss-of-function alleles (atmyb61-2 and atmyb61-5) were used in all experiments, and results are representative. T-DNA insertional mutant lines corresponding to either AtMYB61 or putative downstream targets of AtMYB61 were obtained from the Arabidopsis Biological Resource Center (ABRC) (Alonso et al., 2003). Homozygous T-DNA lines were obtained by PCR screening using the left border A T-DNA primer and a right border gene-specific primer. Insertion sites were sequenced for all mutants to verify insertional mutagenesis (data not shown), and quantitative PCR was conducted to show that the mutants were loss-of-function (data not shown).
For root length and leaf vasculature assays, seeds were liquid sterilized with 10% commercial bleach and 0.1% Triton X-100 solution for 10 min, and then rinsed five times with autoclaved deionised water. Seeds were sown in Petri plates on sterile semi-solid Murashige and Skoog (MS) 0.8% agarose medium (supplemented with 30 mM sucrose and Gamborg’s vitamins) prepared as described previously (Newman et al., 2004; Liang et al., 2005), and then cold stratified for 3 d. Petri plates were then placed in a tissue culture chamber at 21°C under 16 h of light (150 μmol m−2 s−2) and 8 h of dark for 14 d unless stated otherwise.
For primary bolt and hypocotyl analyses, seeds were germinated and plants were grown on soil. Seeds were sown on dampened soil and then cold stratified for 3 d before placement in a growth chamber at 21°C with a regime of 12 h of light (120 μmol m−2 s−2) and 12 h of dark. This growth regime is referred to as short-day conditions herein.
To induce secondary growth in the hypocotyl, the primary inflorescence and secondary inflorescences were continually removed from plants grown under short days for 10 wk (Chaffey et al., 2002). Hypocotyl sections were fixed, coated or stained for transmission electron microscopy, scanning electron microscopy and bright- and dark-field microscopy, respectively.
RNA isolation and quantitative PCR
In order to verify that the insertionally mutagenized mutants identified as above were loss-of-function mutants, transcript accumulation corresponding to the mutagenized gene was determined in the mutants. Primary and secondary inflorescences were excised with a scalpel and immediately frozen in liquid nitrogen. Approximately 1 g (fresh weight) of ground tissue was used per RNA extraction. TRIzol reagent (Invitrogen) was used following the manufacturer’s recommendations. The RNA pellet was dissolved in 30 μl diethylpyrocarbonate (DEPC)-treated water. RNA quantity and purity were analysed using a NanoDrop spectrophotometer (Thermo Scientific, Wilmington, DE, USA), and RNA integrity was assessed by loading 1 μg of RNA onto a 1% agarose 0.5X TBE (Tris-borate-EDTA) gel. First-strand cDNA was generated using 5 μg total RNA with oligo dT primer with SuperScript II (Invitrogen). Standard curves, quantitative PCR and melt curves were conducted with a Bio-Rad Chromo4 Real-Time PCR detector using Sybr-Green florescent dye (Bio-Rad). In order to avoid the generation of a reverse transcription-polymerase chain reaction (RT-PCR) amplicon from genomic DNA, primers were designed so that the 3’-end of at least one of the primers spanned an intron splice site. The relative mRNA levels were determined by normalizing the PCR threshold cycle number of each gene with that of tubulin-4 reference gene (At1g04820). Primer sequences and quantitative PCR amplification conditions are available on request.
β-Glucuronidase (GUS) analysis of gene expression
Histochemical localization of GUS was used to analyse the expression of AtMYB61 as described previously (Newman et al., 2004), making use of transgenic plants harbouring a reporter comprising a translational fusion of AtMYB61 regulatory sequences and the uidA reporter gene (Supporting Information Fig. S1). Primary inflorescences and whole leaves were incubated for 24 h. Hypocotyls were incubated for 5 h. Seedlings were incubated for 48 h. After incubation, tissues were rinsed with water and fixed with 3 : 1 ethanol : acetic acid. Samples were mounted in water on a glass slide and viewed with an Olympus SZX16 microscope under bright field. Images were captured with a QImaging MicroPublisher 3.3RTV digital camera utilizing QCapture version 2.7 software (Olympus, Center Valley, PA, USA).
Root length analysis
Sterile seeds were sown on square Petri plates on semi-solid MS medium and oriented horizontally to allow seedling root growth to take place on the face of the semi-solid medium surface. After 14 or 21 d in the growth chamber, roots were imaged with a Canon EF-S 18-55 digital camera (Canon, Mississauga, ON, Canada). Primary root length and total lateral root length were measured using ImageJ 1.38x (Collins, 2007).
Leaf vasculature assays
Sterile seeds were germinated and the resulting seedlings were grown in a 24-well plate on semi-solid MS medium, using the growth day conditions described previously. After 7 d, seedlings were destained with 100% ethanol. Seedlings were then rehydrated with water, placed on a glass slide and cotyledons were viewed with an Olympus SZX16 microscope under dark field. Images were captured with a QImaging MicroPublisher 3.3RTV digital camera utilizing QCapture version 2.7 software. Leaf vasculature formation was assessed by binning each cotyledon into one of five categories based on the number of vasculature loops (two, two and a half, three, three and a half, four loops).
Primary bolt staining with toluidine blue
Plants were grown in soil under short-day conditions. Once the primary inflorescence reached a length of c. 26 cm, the physical attributes of the plant were recorded and hand sections were made at the base (c. 1 mm). The sections were stained with toluidine blue (stains lignified xylem and parenchyma cells blue, and phloem and pith cells pink–purple). The sections were treated for 2 min with agitation in 0.02% toluidine blue, rinsed with water and mounted in water. Sections were viewed with an Olympus SZX16 microscope under dark field. Images were captured with a QImaging MicroPublisher 3.3RTV digital camera utilizing QCapture version 2.7 software.
Secondary thickened hypocotyls stained with phloroglucinol
Secondary thickened hypocotyl sections (c. 1 mm) were stained with phloroglucinol as described previously (Newman et al., 2004). Sections were viewed with an Olympus SZX16 microscope under both bright and dark field. Images were captured with a QImaging MicroPublisher 3.3RTV digital camera utilizing QCapture version 2.7 software. Measurements employed to calculate the area of xylem and area of phloem were calculated using ImageJ 1.38x (Collins, 2007).
Transmission electron microscopy
Segments (1 cm) of primary inflorescence stems from stage 6.30 plants, or secondary thickened hypocotyls, were fixed in 2% glutaraldehyde in 0.1 M Sorensen’s phosphate buffer (pH 7.4) for 72 h at room temperature, and postfixed in 1% osmium tetroxide in 0.1 M phosphate buffer for 1 h in the dark. Samples were then dehydrated through an ascending graded series of ethanol (30%, 50%, 70%, 80%, 90%, 100%), infiltrated with Spurr’s epoxy resin and polymerized overnight at 65°C. Semi-thin sections (0.5–1 μm) were cut using glass knives on a Leica EM UC6 ultramicrotome (Leica, Allendale, NJ, USA), stained with 0.1% toluidine blue and 0.025% methylene blue, and examined by light microscopy to determine the quality of fixation and orientation of the samples. Ultrathin sections (60–90 nm) were cut using a diamond knife, stained with 3% uranyl acetate in 50% methanol, poststained with Reynold’s lead citrate and examined using a Hitachi H7000 transmission electron microscope (Hitachi, Mississauga, ON, Canada) operated at 75 kV. Pictures were taken using Kodak 4489 electron microscope film, and negatives were scanned using an Epson Perfection 1680 scanner (Epson, Markham, ON, Canada) at 1200 dpi.
Scanning electron microscopy
Segments (1 cm) of secondary thickened hypocotyls were excised, mounted on aluminium stubs using double-sided carbon tape, sputter coated with gold–palladium using a Bal-Tec SCD 050 (Leica Microsystems) and examined with a Hitachi S-2500 scanning electron microscope (Hitachi High Technologies Canada, Inc., Mississauga, ON, Canada) at 20 kV. Digital images were acquired using Quartz PCI (frame grabber and software), (Quartz PCI, Vancouver, BC, Canada).
Seed and seed coat mucilage microscopy
For ruthenium red staining, seeds were either shaken in water for 90 min and then in 0.01% (w/v) ruthenium red for 1 h, or prehydrated with shaking in 0.05 M EDTA for 90 min followed by ruthenium red stain, as indicated. For the seeds stained with shaking, samples were rinsed in dH2O before visualization. Negative staining with India ink was performed by placing seeds in a 1 : 100 dilution of India ink in water. Seeds were observed on a Leica MZ-16F stereomicroscope and imaged with a Micropublisher 3.3 camera (QImaging) operated via Openlab 5 (Perkin Elmer, Woodbridge, ON, Canada).
Developing seeds were prepared for bright-field microscopy, sectioned and stained with toluidine blue O as described in Western et al. (2001). Samples were examined using a Leica DM 6000B compound microscope and images were captured with a QImaging Retiga CCD camera operated through Openlab. Scanning electron microscopy of dry seeds was performed as described in Western et al. (2001) using a Hitachi S4700 field emission scanning electron microscope.
Fibre quality analysis
Secondary thickened hypocotyls were subjected to fibre quality analysis according to published methods (Chaffey et al., 2002). Fibre quality analysis enables the determination of cell types liberated from secondary xylem following maceration, by documenting cell lengths, widths and frequencies as suspended cells pass through a flow chamber.
Cell wall chemical analysis
Primary inflorescence stems were harvested from growth stage 6.30 seedlings that had been grown under short-day conditions. All plants were grown until the inflorescence stems were an equivalent length (26 cm), and were then cut at 0.5 cm from the base of the stem (adjacent to the rosette). Stems were oven dried at 60°C until constant weight was achieved (c. 3 d). The cell wall chemistry of dried stems was determined as described previously (Patzlaff et al., 2003; Rogers et al., 2005b).
Total RNA was extracted using described methods (Newman et al., 2004) from 6-d-old Arabidopsis seedlings grown in the dark in liquid MS medium as described above. Each pool of RNA was derived from hundreds of seedlings. Three biological replicates were collected for each condition (genotype × sucrose presence/absence) for RNA extraction. As there were three genotypes (WT, atmyb61, 35S::MYB61) and two conditions (presence and absence of sucrose), there were 18 RNA samples in total. The quality of the total RNA was assessed using an Agilent Bioanalyser (Agilent, Mississauga, ON, Canada) at the Genomic Arabidopsis Resource Network (GARNet) microarray facility at the Nottingham Arabidopsis Stock Centre. Hybridization to the 18 Affymetrix GeneChip Arabidopsis ATH1 Arrays (Affymetrix, Santa Clara, CA, USA), scanning of the hybridized arrays and raw data collection were performed at the GARNet facility at the Nottingham Arabidopsis Stock Centre according to standard Affymetrix protocols (http://affymetrix.com). The data for the RNA quality control, the raw data for the triplicated microarray experiments and the detailed description of the MIAME (minimum information about a microarray experiment)-compliant experimental conditions are publicly available at: http://ssbdjc2.nottingham.ac.uk/narrays/experimentpage.pl?experimentid=14.
Bioinformatic analyses to identify AtMYB61 targets
To identify AtMYB61 targets, a two-stage complete transcriptome analysis was undertaken. In the first stage, publicly available, complete Arabidopsis transcriptome microarray data were used to identify those genes sharing the same transcript abundance profile as AtMYB61 across multiple stages of development. Genes were identified whose transcript abundance profiles had a Pearson correlation coefficient > 0.8 when compared with the transcript abundance of AtMYB61 across the 66 microarrays comprising the AtGenExpress ‘Developmental Baseline’ dataset (http://web.uni-frankfurt.de/fb15/botanik/mcb/AFGN/atgenex.htm, http://bar.utoronto.ca/ntools/cgi-bin/ntools_expression_angler.cgi). The 58 genes identified in this manner (Supporting Information Table S1) should fit into one of two categories: (1) genes regulated in parallel with AtMYB61, and (2) genes regulated by AtMYB61. To select genes in the latter category, a second stage of analysis was undertaken.
The second stage of transcriptome analysis identified genes whose transcript abundance was influenced by the presence or absence of AtMYB61. A complete transcriptome microarray dataset was generated using WT, myb61 and 35S::MYB61 grown at a time point and under conditions that allow the comparison of the impact of AtMYB61 on transcriptome activity (seedlings grown in the dark in the absence or presence of sucrose). In this dataset, genes that are either direct or indirect targets of AtMYB61 should have reduced transcript abundance in atmyb61 mutants and elevated expression in 35S::MYB61 over-expressing plants in comparison with WT. Using these criteria to generate a ‘bait’ transcript abundance profile for use in the Expression Angler co-expression tool, 31 genes were identified that had a Pearson correlation coefficient > 0.8 across the 18 microarrays in the ‘MYB61 dataset’ (Fig. S2, Table S2). Groups of genes with transcript abundance profiles that had high Pearson correlation coefficients relative to AtMYB61 were identified using Expression Angler (http://bar.utoronto.ca/ntools/cgi-bin/ntools_expression_angler.cgi) (Toufighi et al., 2005). The calculations of the Pearson correlation coefficients were based on raw expression values across all 18 GeneChips generated in our study. Both gene lists, generated from the microarray and Expression Angler analyses, were then compared using Venn Selector (http://bar.utoronto.ca/ntools/cgi-bin/ntools_venn_selector.cgi). Three genes were identified in the intersection set.
The 5’ noncoding sequences (1000 bp) for the three genes with Pearson correlation coefficients > 0.75 in both datasets were obtained by bulk download from The Arabidopsis Information Resource (TAIR, http://www.arabidopsis.org/tools/bulk/sequences/index.jsp). Over-represented sequence motifs in the 5’ noncoding sequences were identified using option 2 of Promomer (http://bar.utoronto.ca/ntools/cgi-bin/BAR_Promomer.cgi) (Toufighi et al., 2005) with the following parameters: base pairs in the element = 6, minimum percentage of genes in which the identified element should occur = 75. Bootstrap analysis (n = 1000) with a randomized dataset allows the validation of the significance of an over-represented motif within the sequences being queried.
Electrophoretic mobility shift assay (EMSA)
Recombinant AtMYB61 protein was produced in Escherichia coli using the coding sequence cloned in frame into the NdeI and BamHI sites of the pET15b vector (Novagen, EMD Millipore, Mississauga, ON, Canada). Recombinant AtMYB61 protein was produced, extracted and affinity purified as described previously for pine MYB proteins (Patzlaff et al., 2003). EMSA conditions were exactly as described previously (Patzlaff et al., 2003; Gomez-Maldonado et al., 2004), but using recombinant AtMYB61 protein instead of pine MYB protein.
Transcriptional activation assay
Transcriptional activation assays using yeast were performed as described previously (Patzlaff et al., 2003), but substituting the AtMYB61 coding sequence instead of the pine MYB sequences.
Results and Discussion
AtMYB61 influences multiple stages of plant growth and development
The role of AtMYB61 in growth and development was investigated by comparing gain-of-function mutants (35S::MYB61), loss-of-function mutants (myb61) and WT plants (Fig. 1). Insertionally mutagenized AtMYB61 loss-of-function mutants (atmyb61) germinated early, developed slowly, and showed delayed flowering and delayed senescence relative to WT plants. Conversely, plants that constitutively over-expressed AtMYB61 (35S::MYB61) germinated slowly, but developed quickly, and flowered and senesced more rapidly than WT plants. On the basis of these findings, AtMYB61 appeared to play an important role in shaping plant growth and in the rate of progression through the life cycle.
AtMYB61 is expressed throughout plant growth and development in sink tissues
Sink tissues are characterized as sites of carbon deposition, where sugars are transported for integration into structural or storage molecules (Roitsch, 1999; Paul & Foyer, 2001). AtMYB61 gene regulatory sequences directed reporter gene expression in multiple sink tissues (Fig. 2). The gene regulatory sequences necessary to accurately report on AtMYB61 expression included 2000 bp of 5’ noncoding sequences and both introns contained within the AtMYB61 intragenic region (Fig. S1). The inclusion of the introns was necessary for reporter expression that fulfilled the criterion of being responsive to the previously reported action of sugar, which induced MYB61 transcript accumulation (Penfield et al., 2001).
As reflected by the localization of GUS reporter activity, AtMYB61 was expressed in cells destined to become xylem vessels in all tissues examined, including the inflorescence stem, leaves, hypocotyls and roots (Fig. 2). Prominent expression was also observed in the roots at the site of emergence of lateral roots, as well as the elongation zone (Fig. 2). Expression was also observed in seeds throughout development, from shortly after pollination of the ovule, until the seed was completely mature. Expression in the seed did not appear to be confined to any particular cell type, but appeared to be uniform throughout the developing seed (Fig. 2). Altogether, the results suggest that AtMYB61 is most prominently expressed in tissues that might best be collectively described as sinks – developing vasculature, roots and seeds.
AtMYB61 shapes resource allocation related to xylem vessel formation
Consistent with AtMYB61 expression in differentiating xylem vessels, gain and loss of AtMYB61 function had an impact on both quantitative and qualitative aspects of xylem vessel formation (Fig. 3). In the inflorescence stem, atmyb61 mutants had fewer xylem vessels relative to WT plants, whereas 35S::MYB61 plants had more xylem vessels than WT plants (Fig. 3). In atmyb61 mutants, there were distinctly fewer small lumen protoxylem vessels and fewer larger diameter metaxylem cells, whereas gain-of-function mutants had a greater abundance of both cell types. Transmission electron microscopy of sectioned inflorescence stems confirmed these findings, and also revealed that the secondary cell walls characteristic of xylem vessels were poorly formed in the atmyb61 loss-of-function mutants (Fig. 3). Xylem vessels in atmyb61 loss-of-function mutants were not only less abundant, but were also irregularly shaped, had thinner cell walls and sometimes retained cytoplasm, suggesting that they had not undergone the final stage of xylogenesis, autolysis.
Staining of the lignified xylem vessels in hypocotyls was less intense in atmyb61 loss-of-function mutants relative to WT plants (Fig. 3). In 35S::MYB61 gain-of-function mutants, Wiesner reagent staining encompassed a greater number of hypocotyl cell files relative to WT plants, and was not limited to xylem vessels, but instead extended to adjacent cells, which were ectopically lignified (Fig. 3). That is, the introduction of AtMYB61 function in these cells was sufficient for them to acquire the cell wall characteristics of xylem vessels.
In keeping with a role in promoting the formation of xylem, secondary thickened hypocotyls of atmyb61 loss-of-function mutants contained fewer cells with secondary cell walls relative to the equivalent WT hypocotyls (Fig. 3). By contrast, gain of AtMYB61 function in secondary thickened hypocotyls resulted in a greater diameter of xylem cell files in comparison with equivalent WT hypocotyls (Fig. 3). Consistent with this, atmyb61 mutants had fewer differentiated xylem vessels, and any xylem vessels that were present had thinner cell walls relative to equivalent WT hypocotyls (Fig. 3). By contrast, the 35S::MYB61 mutants had more xylem vessels (Fig. 3).
Fibre quality analysis determined the relative proportions of the three major cell types in hypocotyl secondary xylem: differentiated xylem vessels and fibres, and incompletely differentiated fusiform cambial cells (Fig. S3). Fibre quality analysis enables the determination of cell types liberated from secondary xylem following maceration by documenting cell lengths, widths and frequencies as suspended cells pass through a flow chamber. Fibre quality analysis of macerated hypocotyl secondary xylem revealed that atmyb61 had a fraction of the amount of detectable xylem cells per unit of hypocotyl in comparison with WT plants, whereas 35S::MYB61 had more fibre quality analysis-detected xylem cells in an equivalent portion of hypocotyl relative to WT plants (Fig. S3). The secondary xylem of atmyb61 mutants characteristically had very few differentiated vessel or fibre cells and, as a proportion of overall cells detected, was dominated by fusiform cambial cells relative to WT plants. However, 35S::MYB61 secondary xylem had more xylem cells overall in comparison with WT plants. Thus, AtMYB61 activity appeared to promote the differentiation of meristematic fusiform cambial cells to generate fully differentiated vessels or fibres.
Establishment of the xylem vessel network that comprises cotyledon veins was also shaped by AtMYB61 activity. Xylem vessels form stereotypical looped venation patterns in A. thaliana cotyledons (Rolland-Lagan, 2008). Seven days post-germination, WT cotyledons had two or three completely closed loops and two open loops of primary vein (Fig. 4). By contrast, cotyledons of atmyb61 mutants had only two closed loops, and were only just beginning to form additional loops (Fig. 4). 35S::MYB61 gain-of-function mutants had three or four closed loops (Fig. 4). Thus, AtMYB61 appeared to promote xylem vessel formation in photosynthetically active cotyledons.
AtMYB61 shapes resource allocation related to root development
As AtMYB61 regulatory sequences promoted reporter expression in roots (Fig. 2), the effect of AtMYB61 on the root system architecture was examined. Loss of AtMYB61 function resulted in a decrease in primary root growth, and a significant decrease in the formation of lateral roots relative to WT plants (Fig. 5). Conversely, gain of AtMYB61 function promoted root growth, in which more lateral roots were formed in comparison with WT plants (Fig. 5). On the basis of these analyses, AtMYB61 appeared to play a role in shaping the ratio of lateral to primary roots.
AtMYB61 modulates the expression of a specific set of target genes
As a transcription factor, AtMYB61 should exert its control over facets of the plant transpiration stream by modulating the expression of specific target genes (Ptashne & Gann, 1997). To identify such targets, a two-stage complete transcriptome analysis was undertaken, using publicly available microarray datasets in combination with a custom microarray dataset comparing the transcriptomes of WT, atmyb61 and 35S::MYB61 plants (see the Materials and Methods section and Tables S1, S2). Three genes emerged from the sequential filtering of publicly available microarray data and the MYB61-specific microarray data, which were shared in both tiers of data mining. These genes are strong candidates for direct targets of AtMYB61: At1g62990, At2g45220 and At4g26220 (Fig. 6).
The nature of the gene products encoded by the three putative AtMYB61 targets is consistent with their role in xylem development. At1g62990 encodes the homeobox protein AtKNAT7. AtKNAT7 is expressed in xylem fibres, and xylem vessels of AtKNAT7 loss-of-function mutants (irregular xylem11, irx11) have thin, weak cell walls resulting in collapsed vessels (Brown et al., 2005; Zhong et al., 2008). At2g45220 encodes a pectin methylesterase (AtPME), a class of enzymes with demonstrable roles in reconfiguring plant cell wall chemistry (Pelloux et al., 2007). At4g26220 encodes a caffeoyl-CoA O-methyltransferase (AtCCoAOMT7), which, based on the extent of sequence similarity, is probably involved in the genesis of the monolignol precursors used to build the lignin polymer, as do related homologues (Do et al., 2007).
AtMYB61 regulates genes with specific target motifs in their promoters
Consistent with the three genes functioning as downstream targets of AtMYB61, recombinant AtMYB61 protein bound to 300-bp DNA regions residing upstream of the TATA box for each of the putative target genes (Fig. 6). Candidate AtMYB61 binding sites in the gene regulatory regions of these three genes were identified by algorithm-based screening for over-represented motifs in the three DNA sequences. The most over-represented DNA motif in the gene regulatory sequences showed high similarity to canonical R2R3-MYB binding sites known as AC elements (Fig. 6). Three such AC elements were found in each of the upstream regions of AtPME and AtKNAT7, whereas four elements were found in the AtCCoAOMT7 upstream noncoding sequences (Fig. 6, Table S3). Recombinant AtMYB61 bound to this element, but could not bind to a mutated version of the element, confirming that this is the likely target of AtMYB61 binding in these genes (Fig. 6). Non-AC-element-containing DNA could not be bound by AtMYB61 (Fig. 6), nor could it compete with AC elements for AtMYB61 binding (Fig. S4). The AC element was also an effective competitor for recombinant AtMYB61 bound to the 300-bp upstream regulatory sequences (Fig. 6). Moreover, expression of AtMYB61 in yeast transactivated an artificial target gene comprising a tandem repeat of the AC element fused to a yeast minimal promoter, upstream of the reporter β-glucosidase (Fig. 6). These binding data are in accordance with the literature surrounding MYB–DNA interactions (Prouse & Campbell, 2012). Thus, AtMYB61 activity appeared to promote the expression of target genes containing the AC element.
Strikingly, evidence suggests that AtKNAT7 is the target of other R2R3-MYB transcription factors and, on that basis, is thought to be a component of a transcriptional network that regulates xylem differentiation (Zhong et al., 2007, 2008). AtKNAT7 appears to function as a common target for several transcriptional networks that are involved in xylem differentiation, including one that involves AtMYB61. As such, AtKNAT7 could be viewed as a regulatory module that is co-opted by several gene regulatory networks.
AtMYB61 regulates genes which themselves contribute to AtMYB61-related phenotypes
To determine whether the putative AtMYB61 targets contribute to any of the xylem-related traits in which AtMYB61 is involved, the phenotypes of the loss-of-function mutants for the target genes (atknat7/irx11, atpme and atccoaomt7) were compared with atmyb61 and WT. Loss-of-function mutations in each of the three target genes generated xylem-related phenotypes that at least partially phenocopied atmyb61 phenotypes. For example, secondary thickening of xylem vessel cell walls was reduced in atknat7/irx11 and atpme mutants relative to WT, like atmyb61 (Fig. 7).
As with atmyb61 mutants, the xylem : phloem ratio was reduced relative to WT in secondary thickened hypocotyls of atknat7/irx11, atpme and atccoaomt7 mutants (Fig. 7). Strikingly, the atknat7/irx11, atpme and atccoaomt7 mutants had far fewer fibre cells and disproportionately more vessel cells relative to WT (Fig. 7). Unlike atmyb61 mutants, the atknat7/irx11, atpme and atccoaomt7 mutants were able to make vessels, and fusiform cambial cells were not the predominant cell type. These findings are in keeping with the hypothesis that AtMYB61 functions upstream of AtKNAT7, AtPME and AtCCoAOMT7, as AtMYB61 activity promotes the differentiation of both vessels and fibres, whereas the differentiation of vessels more prominently occurs in the atknat7/irx11, atpme and atccoaomt7 mutants. This suggests that AtKNAT7, AtPME and AtCCoAOMT7 are involved in pathways governing fibre differentiation in secondary hypocotyl development, whereas AtMYB61 sits upstream of both fibre and vessel differentiation pathways in the development of this anatomical region.
Similar to atmyb61, the development of cotyledon vasculature was altered in atknat7/irx11 and atccoaomt7 mutants relative to WT, such that fewer closed loops of vasculature were formed (Fig. S5). These differences were less pronounced in the atpme mutant, in which cotyledon vascular development was more similar to WT (Fig. S5). This suggests that AtKNAT7 and AtCCoAOMT7, but not AtPME, probably contribute to leaf vascular development by functioning downstream of AtMYB61.
As AtCCoAOMT7 and AtPME encode proteins predicted to play a role in modifying cell wall chemistry (Zhong et al., 1998; Do et al., 2007; Pelloux et al., 2007), the impact of atknat7/irx11, atpme and atccoaomt7 mutations on the chemistry of inflorescence stem cell walls was assessed. The atccoaomt7 mutant shared a very similar cell wall chemical fingerprint to atmyb61, characterized by significant decreases in glucose and xylose moieties (Table S4). This suggests that AtMYB61-mediated changes in cell wall carbohydrate composition may be mediated through AtCCoAOMT7, which is somewhat counterintuitive, given that this enzyme probably plays a role in lignin biosynthesis. It may be that the altered flux of carbon through lignin biosynthesis controlled by AtCCoAOMT7 has an indirect impact on cell wall carbohydrate deposition. Alternatively, although unlikely, it may be that the as yet uncharacterized AtCCoAOMT7 enzyme activity is not involved in lignin biosythesis. The involvement of other lignin biosynthetic enzymes in the modulation of cell wall carbohydrate deposition has been suggested previously (Hu et al., 1999; Wagner et al., 2009). Nevertheless, the data suggest that AtMYB61 is necessary for the deposition of cell wall carbohydrates, presumably into cellulose and/or xyloglucan, respectively, and that this may indirectly require AtCCoAOMT7 activity. AtMYB61 must regulate lignin content through a mechanism other than one of the three downstream targets characterized here, as none of the mutants for the putative downstream target genes showed a significant decrease in lignin content (Table S4).
As AtMYB61 was first described as a gene influencing seed coat formation, thereby having an impact on the release of seed coat mucilage (Penfield et al., 2001), the involvement of AtKNAT7, AtCCoAOMT7 and AtPME in seed coat formation and mucilage release was examined. Although atccoaomt7 and atpme-1 mutants showed no changes in mucilage release or seed coat morphology, atknat7 mutants were affected both in mucilage amount, as visualized with ruthenium red, and in the morphology of the mucilage secretory cells (Figs S6, S7). When visualized with ruthenium red and India ink, atknat7 mutants resemble atmyb61 mutants which make and release less mucilage than WT. However, when viewed in section, atknat7 mutants do not show obvious changes to the columella height or mucilage pocket size, as seen for atmyb61 or other reduced mucilage mutants (Fig. S6) (Penfield et al., 2001; Western et al., 2001). In addition to the effect on mucilage properties, atknat7 mucilage secretory cells have abnormally thick radial cell walls (Fig. S7), suggesting that production of the secondary cell wall of the columella is aberrant. The data are consistent with AtMYB61 having an impact on seed coat formation, and subsequent mucilage release, through a network that involves AtKNAT7, but, at least as determined at the cellular level, not through AtCCoAOMT7 or AtPME.
AtMYB61 conditions both water retention and resource allocation
The findings support the hypothesis that AtMYB61 regulates all three major components of the plant transpiration stream. AtMYB61 closes stomata, thereby limiting water loss, whilst directing the establishment of water-conducting xylem vessels and elaboration of the root network to better seek, acquire and transport water. Thus, AtMYB61 might function as a pleiotropic regulator that sets the plant in ‘water collection and retention mode’. Consistent with a role in improving water deficit tolerance, 35S::MYB61 gain-of-function plants were clearly more tolerant to water withdrawal relative to WT. Water withdrawal for 21 d, followed by re-watering, was lethal for c. 50% of WT plants (Fig. S8), whereas the majority (> 85%) of 35S::MYB61 plants survived 21 d without watering, and recovered following re-watering.
AtMYB61 may also be thought of as a gene that conditions the allocation of carbon in response to appropriate levels of metabolites. In keeping with this, AtMYB61 expression is shaped by specific metabolites. AtMYB61 expression is increased by sugar, notably sucrose (Penfield et al., 2001; Newman et al., 2004) (Fig. S9), and is diminished by two amino acids implicated in nitrogen partitioning and signalling, glutamate and glycine (Dubos et al., 2005). It is striking that AtMYB61 activity is up-regulated by the most significant product of photosynthesis, sucrose, and is down-regulated by two amino acids that are significant by-products of photorespiration, glutamate and glycine.
It may be that AtMYB61 is poised to respond to the abundance of different carbon skeletons in plants, and thereby modulate carbon acquisition via stomata and carbon allocation in sink tissues. Consistent with this hypothesis, AtMYB61 activity shapes plant development in response to sugar. If grown in the presence of sugars, dark-grown WT seedlings undergo extensive vegetative development that is not seen in the absence of sugar (Roldan et al., 1999). AtMYB61 activity influences this response (Fig. S10). Dark-grown atmyb61 mutants are less sensitive to sugar than WT, whereas 35S::MYB61 plants respond as though they are exposed to greater quantities of sugar. Thus, AtMYB61 conditions sugar sensitivity and the subsequent allocation of resources to development.
These findings suggest that the link between the apparently disparate pleiotropic components that are mediated by AtMYB61 may be the allocation of carbon to sink tissues. That is, AtMYB61 functions to channel carbon to nonrecoverable carbon sinks, including secondary cell walls, xylem, roots and the seed coat. When viewed in this light, one might think of AtMYB61 as a gene that functions under ‘feast’ conditions, allocating carbon to nonrecoverable sinks when conditions are favourable to do so.
We are most grateful to Astrid Patzlaff, Christine Surman and Joan Ouellette for excellent technical assistance. We would like to thank Sarah Pantedelis and Andrej Arsovski for their help with seed coat analysis. We are also grateful for the excellent advice provided by four anonymous reviewers. S.D.M. is a Canada Research Chair in Wood and Fibre Quality. This work was generously supported by funding from the Natural Science and Engineering Research Council of Canada (NSERC) and the Canada Foundation for Innovation (CFI) to S.D.M., by a Canadian Graduate Scholarship (CGSD) from NSERC awarded to O.W., by an NSERC Discovery Grant and the NSERC Green Crops Network to C.J.D., and by funding from the University of Toronto, CFI and NSERC to M.M.C. Research infrastructure was provided by the Centre for Analysis of Genome Evolution and Function at the University of Toronto. Contributions: J.M.R., C.D. and M.M.C. designed the research; J.M.R., C.D., M.B.P., H.H., K-Y.K., M.P., E.L. and T.L.W. performed the research; J.M.R., C.D., M.B.P., M.P., O.W., E.L., C.J.D., T.L.W. and S.D.M. analysed the data; J.M.R., C.D. and M.M.C. wrote the manuscript with editorial assistance from M.B.P., H.H., K-Y.K., M.P., O.W., C.J.D., T.L.W. and S.D.M.