Long cold exposure induces transcriptional and biochemical remodelling of xylem secondary cell wall in Eucalyptus

Long cold exposure induces transcriptional and biochemical remodelling of xylem secondary cell... Abstract Although eucalypts are the most planted hardwood trees worldwide, the majority of them are frost sensitive. The recent creation of frost-tolerant hybrids such as Eucalyptus gundal plants (E. gunnii × E. dalrympleana hybrids), now enables the development of industrial plantations in northern countries. Our objective was to evaluate the impact of cold on the wood structure and composition of these hybrids, and on the biosynthetic and regulatory processes controlling their secondary cell-wall (SCW) formation. We used an integrated approach combining histology, biochemical characterization and transcriptomic profiling as well as gene co-expression analyses to investigate xylem tissues from Eucalyptus hybrids exposed to cold conditions. Chilling temperatures triggered the deposition of thicker and more lignified xylem cell walls as well as regulation at the transcriptional level of SCW genes. Most genes involved in lignin biosynthesis, except those specifically dedicated to syringyl unit biosynthesis, were up-regulated. The construction of a co-expression network enabled the identification of both known and potential new SCW transcription factors, induced by cold stress. These regulators at the crossroads between cold signalling and SCW formation are promising candidates for functional studies since they may contribute to the tolerance of E. gunnii × E. dalrympleana hybrids to cold. Introduction Eucalypts are the most planted hardwood trees worldwide because of their very rapid growth, exceptional wood quality and wide adaptability (Myburg et al. 2007). Two mains species are used in industrial plantations: the subtropical Eucalyptus grandis and the temperate Eucalyptus globulus, which are among the leading sources of wood biomass for pulpwood and timber, and are considered promising feedstocks for the production of second-generation biofuels. Although eucalypts are planted in >100 countries over six continents, the extension of industrial plantations to Northern hemisphere countries has been hampered by the fact that most commercial species and derived hybrids are very sensitive to frost (Booth 1990). This frost sensitivity is partially explained by the fact that eucalypts are evergreen trees that do not cope with frost by avoidance as many woody plant species from temperate countries do. The increasing demand for wood led the French ‘Institut Technologique FCBA’ to select fast-growing and frost-resistant hybrids between Eucalyptus gunnii, a species originating from the Central Plateau of Tasmania (Potts et al. 2001), and Eucalyptus dalrympleana, known under the generic name of Eucalyptus gundal (Navarro et al. 2009). The genes responsible for the high frost tolerance of E. gunnii have not yet been identified; however, a comparative study of the dehydration-responsive element binding (DREB) family of transcription factors (TFs) in E. grandis and E. gunnii suggested that additional copies and/or higher basal expression of CBF/DREB1 account, at least in part, for the ability of E. gunnii to adapt to cold conditions (Nguyen et al. 2017). Wood, or secondary xylem, is produced by the activity of an internal meristem, the vascular cambium, through a complex differentiation process leading to highly specialized xylem cells characterized by thick, lignified secondary cell walls (SCWs) (Plomion et al. 2001). Secondary cell walls are mainly composed of cellulose, hemicelluloses and lignin, in proportions of ~2:1:1, crosslinked into a supramolecular network, thereby creating a highly resistant barrier. The phenolic polymer lignin is mainly responsible for the recalcitrance of SCWs to degradation, which is a major obstacle for industrial end-uses of wood such as for pulp and paper manufacture or bioethanol production (VanAcker et al. 2013). In angiosperms, lignin is a complex tridimensional polymer containing mainly guaiacyl (G) and syringyl (S) units (Vanholme et al. 2010) that are connected by an array of inter-monomeric linkages, most commonly labile β-O-4 bonds. A high S/G ratio, which is associated with a high proportion of β-O-4 bonds, is, for instance, beneficial for lignin depolymerization during the kraft pulping process (Ventorim et al. 2014). The lignin content and composition of wood varies greatly within and between species, between developmental stages, and in response to environmental cues (Plomion et al. 2001). In eucalypts, for instance, the structure and content of lignin is affected by the nitrogen status of the plant (Camargo et al. 2014). The gene families encoding the 11 enzymes involved in biosynthesis of lignin monomeric units, or monolignols, have been studied at the genome level in E. grandis (Carocha et al. 2015). Lignin biosynthesis starts with the deamination of phenylalanine and leads to the production of hydroxycinnamoyl-CoA esters. This part of the pathway is called ‘general phenylpropanoid metabolism’ since hydroxycinnamoyl-CoA esters are also precursors of a wide range of end products including flavonoids, anthocyanins and condensed tannins, which vary according to species, cell types and in response to environmental signals (Dixon and Paiva 1995). Based on comparative phylogeny and patterns of expression, a total of 38 genes were identified as bona fide members of the 11 gene families, of which 25 are likely involved in developmental lignification belonging to the so-called ‘core vascular lignification toolbox’ (Carocha et al. 2015). In addition, Carocha et al. (2015) pointed out several members of the cinnamyl alcohol dehydrogenase (CAD), caffeoyl-CoA 3-O-methyltransferase (CCoAOMT) and caffeic acid O-methyltransferase (COMT) superfamilies called ‘CAD-like’, ‘CCoAOMT-like’ and ‘COMT-like’ as strongly expressed in xylem, suggesting possible roles in xylogenesis, although their functions are still unknown. Our knowledge of how SCW biosynthesis is regulated at the transcriptional level comes mainly from studies of the herbaceous plant Arabidopsis. In this model plant, a complex hierarchical network of TFs of the NAC and R2R3-MYB families act as first- and second-level master switches, respectively, to regulate a battery of downstream TFs (most of them also belonging to the MYB family) and SCW biosynthesis genes (Grima-Pettenati et al. 2012, Hussey et al. 2013, Nakano et al. 2015, Zhong and Ye 2015). Remarkably, a recent study demonstrated that abiotic stresses such as high salinity or iron deprivation can co-opt the xylem regulatory network in Arabidopsis, i.e., distinct stresses are able to modulate the expression of regulators of SCW biosynthesis to potentially promote functional adaptation of the plant to the stress conditions (Taylor-Teeples et al. 2015). In eucalypts, two master regulators of SCW deposition, EgMYB2 (Goicoechea et al. 2005) and EgMYB1 (Legay et al. 2007, 2010), have been characterized functionally. Potential orthologues of Arabidopsis SCW regulators in the E. grandis genome (Myburg et al. 2014) have also been identified by comprehensive genome-wide surveys of the R2R3-MYB and NAC TF families (Hussey et al. 2015, Soler et al. 2015). Importantly, members of subgroups present neither in Arabidopsis nor in rice were also identified, some of which are preferentially expressed in vascular cambium suggesting that they may be novel genes that regulate xylem differentiation (Soler et al. 2015). Exposure to chilling is known to increase plants’ ability to later withstand freezing temperatures. This process of cold acclimatization induces complex physiological and biochemical changes. Among them, increased cell-wall rigidity might lead to higher cell resistance to dehydration due to freezing (reviewed in LeGall et al. 2015). In some cases, lignin synthesis is enhanced during cold acclimatization and this may contribute to cell strengthening, thereby preventing cell damage and collapse (Domon et al. 2013). Most of these studies were performed on crop plants. By contrast, very few studies have investigated the effect of cold acclimatization on the SCWs of forest trees, despite the fact that cold is one of the major abiotic constraints affecting perennial plant development in many countries (Teulières et al. 2007). To the best of our knowledge, there is only one study of trees (Hausman et al. 2000), which reported an increase in the lignin content of shoots of young poplar trees exposed to 10 °C for 2 weeks. To evaluate the effects of low temperature on the SCW properties of Eucalyptus wood and to identify the metabolic and regulatory genes involved in SCW biosynthesis under cold stress, our study combined morphological analyses of xylem with biochemical characterization and transcriptomic profiling in young E. gundal hybrids submitted to chilling temperatures for up to 7 weeks and in adult trees grown in field conditions over three seasons in the southwest of France. Moreover, we performed gene co-expression network analyses on expression profiling data, which identified new potential regulators at the crossroad between SCW synthesis and response to cold stress. These regulators are promising candidates for future studies since they are potentially promoting functional adaptation to cold conditions. Materials and methods Plant material and cold treatment Eucalyptus gundal plants (E. gunnii × E. dalrympleana hybrids) obtained by in vitro propagation of cuttings were provided to us by the Institut Technologique FCBA (Pierroton, France). Six months after rooting, 36 plants were cultivated in a growth chamber with long day photoperiod (16 h light/8 h dark) at 25 °C/22 °C (light/dark) and 80% relative humidity. They were first exposed to a short acclimatization period of 3 days at 12 °C/8 °C (light/dark) before 46 days of cold treatment at 8 °C/4 °C (light/dark). Samples were collected before the cold treatment (0 day, control). Further samples were collected at 2 and 15 days after the start of cold treatment. Final samples were collected at the end of the experiment (46 days). Each sampling was performed on 6–10 independent plants. The lower part of the main stem of each plant was collected; 1 cm was kept in ethanol 80% (v/v) for histochemical analyses, the rest was frozen in liquid nitrogen, milled using a Mixer Mill MM 400 (Retsch, Haan, Germany) and the resulting powders were kept at −80 °C for later use. Samples from adult plants (>10 years old) were taken from the FCBA experimental plantation in southwest France (43°21′21.0′N 1°14′27.0′E). A set of 12 E. gundal (four independent pools of three individuals) were sampled over three seasons (see Figure S1 available as Supplementary Data at Tree Physiology online): autumn 2014 (mean temperatures 17 ± 6 °C), winter 2015 (mean temperatures 5 ± 5 °C) and summer 2015 (mean temperatures 22 ± 6 °C). After removing the bark at a height of ~150 cm, samples of xylem were collected by scraping using a single-edge, sharp blade and the samples were processed as above for the young plants. For wood structure observations, micro-core samples were collected by using a 5-mm diameter punch and were immediately fixed in 80% ethanol until use. RNA isolation Total RNA was extracted from the milled powder samples produced from the stem fragments of young plants and scraped xylem for adult trees as described by Southerton et al. (1998). Contaminating DNA was removed by using the TURBO DNA-free Kit (Ambion, Austin, TX, USA), according to the manufacturer’s instructions. RNA quantity and quality were checked by using a NanoDrop ND-1000 spectrophotometer (Thermo Scientific, Waltham, MA, USA) and 1% agarose gels. DNA contamination was evaluated by PCR using the housekeeping gene Tubulin (see Table S1 available as Supplementary Data at Tree Physiology Online). Quantitative RT-PCR First strand cDNAs were synthesized from 1 μg of total RNA using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystem, Foster City, CA, USA) according to the manufacturer’s instructions. The primer pairs were designed by using QuantPrime online software (http://www.quantprime.de/, Arvidsson et al. 2008) and are listed in Table S1 available as Supplementary Data at Tree Physiology online. Transcript abundance was assessed by using high-throughput microfluidic qPCR (96.96 Dynamic Array Fluidigm®, BioMark, Fluidigm Corporation, San Francisco, CA, USA) according to the manufacturer’s protocol and by conventional RT-qPCR using 1 μl of 1/10 diluted cDNA and Power SYBR® Green PCR Master Mix (Applied Biosystems), as described previously (Cassan-Wang et al. 2012). Expression profiles were obtained by the E−ΔΔCt method (Pfaffl 2001) using primers efficiency calculated through LinRegPCR (Ruijter et al. 2009). Expression was normalized against reference genes PP2A1, PP2A3, IDH, SAND and EF1a (Cassan-Wang et al. 2012). Biochemical analyses of SCWs Biochemical analyses of young trees were performed on bark-free stem samples, dried for 48 h at 60 °C and milled with a Mixer Mill MM 400 (Retsch). For adult trees, scraped xylem samples were ground in liquid nitrogen and freeze-dried for 48 h. Extractives content analyses and Analytical Pyrolysis (Py-GC/FID) applied to extractive-free xylem residues (EXRs) were performed as described in Lepikson-Neto et al. (2013). At least two runs were performed per sample. Pyrolysis products were identified as previously described (Alves et al. 2009). Lignin content was evaluated by the Klason method as described by Méchin et al. (2014). Lignin structure was analysed by thioacidolysis. For each sample, thioacidolysis was performed as described by Méchin et al. (2014), on 10 mg of EXR heated for 4 h and at 100 °C (oil bath) with 7 ml of a mix prepared with 10 ml of ethanethiol, 2.5 ml of BF3 etherate (Fisher Scientific, Waltham, MA, USA) and then adjusted the final volume of 100 ml with dioxane (Fisher Scientific). Lignin monomers were then extracted with 7 ml of dichloromethane (Fisher Scientific). Extracts were trimethylsilylated by incubating with 50 μl of N,O-bis(trimethylsilyl)trifluoroacetamide (Sigma, St. Louis, MO, USA) and 5 μl of ACS-grade pyridine (Sigma) for 30 min at 60 °C. These samples were then analysed in a Triple Quadrupole GC-MS system (Thermo Scientific) fitted with an autosampler, a splitless injector (280 °C), and an iontrap mass spectrometer operating in the electronic impact mode with a source at 220 °C, an interface at 280 °C and a 50–650 m/z scanning range. The column was a ZB5-MSi column (Phénomenex, Torrance, CA, USA) operated in the temperature program mode (from 45 to 180 °C at +30 °C min–1, then from 180 to 260 °C at +3 °C min−1), with helium carrier gas at a 1.5 ml min−1 flow rate. S and G peaks were identified using m/z (respectively, 399 and 369) and integrated using Xcalibur© software. Microscopy and histochemistry Tissues fixed in 80% ethanol were cut into thin sections (20 μm) by using a Vibratome VTS1000 (Leica, Wetzlar, Germany) and a sapphire vibratome knife (Delaware Diamond Knives, Wilmington, DE, USA). Lignin and cellulose were stained, respectively, with a mix of Safranin (0.1%) and Astra Blue (1%) according to the protocol of Vazquez-Cooz and Meyer (2009). Samples were also dehydrated in a EM CPD300 automated critical point dryer (Leica) using CO2 as a transitional fluid. When dry, the samples were coated with nickel (5 nm) in a EM MED020 vacuum coating system (Leica) and analysed in a scanning electron microscope (ESEM Quanta 250 FEG; FEI, Merignac, France) at 5 kV. In parallel, small fragments were embedded in LR White resin (London Resin Company Ltd, London, UK) by immersion in four consecutive baths (24 h) of increasing concentrations of resin (33, 50, 66 and 100%). Semi-thin sections (1 μm) were obtained by using a Histo Diamond Knife (DiATOME, Biel, Switzerland) on an ultramicrotome Reichert UltraCutE (Leica Microsystems, Rueil Malmaison, France). Sections were immediately mounted on glass slides and warmed to 60 °C for 5 min. For tissue morphology observation, sections were stained with 1% (w:v) aqueous Toluidine Blue O for 10 min at room temperature, rinsed in distilled water, dried at 60 °C and mounted under coverslips in mounting medium (Richard-Allan Scientific, San Diego, CA, USA). Semi-automatic image acquisition was performed using a Nanozoomer C9600-12 (Hamamatsu, Hamamatsu City, Japan) in bright field at ×40 magnification. Images were exported from raw data by using NDPview 2.3.1 and subsequently treated with ImageJ software (V1.5) to determine cell number, lumen and total cell-wall surface areas, thus allowing the calculation of mean cell density, average lumen diameter and average cell-wall thickness. Cells with diameters superior to 20 μm were considered as vessels. Average cell-wall thickness was determined on >15,000 cells in 7–20 stem sections for each tree. Co-expression network analyses Expression data from 146 genes related to the MYB family, NAC family and enzymes involved in SCW deposition were obtained from our previous studies (Carocha et al. 2015, Hussey et al. 2015, Soler et al. 2015). In these works, transcript levels were assessed for 22 parameters, including various eucalypts species, tissues, stages of development and growth conditions (see Table S2 available as Supplementary Data at Tree Physiology Online). Pearson pairwise correlation matrix was generated using ‘rcorr’ function of ‘Hmisc’ package with R software. The P-values obtained were adjusted using ‘p.adjust’ function with Bonferroni correction method. We eliminated correlations with an adjusted P-value >0.05 and calculated distances=(1−(correlation value)2)between the remaining genes. Distances were used to graphically represent gene correlation network using Cytoscape© software (force-directed layout). Results Analyses of xylem structure in plants submitted to low temperatures To evaluate the effects of prolonged cold treatment on xylem structure, we first compared thin sections of stems from young Eucalyptus hybrids grown for 7 weeks (46 days) at 4 °C with similar control sections harvested before cold exposure (i.e., grown at 25 °C). In the controls, the newly formed xylem cells immediately below the cambium formed a differentiating zone characterized by a gradual increase in cell size and cell-wall thickness (Figure 1a). The same observations were done in a control experiment performed with young trees grown during 46 days at 25 °C (see Figure S2a and b available as Supplementary Data at Tree Physiology Online). After 46 days of cold exposure, we observed an increase in SCW thickness in the xylem cells located within the first four to eight layers close to the cambium initials (Figure 1b). This was confirmed by using scanning electron microscopy (Figure 1c and d), which showed that in plants exposed to cold for 46 days, several layers of cells in the early differentiating xylem zone (close to the cambium) had thick SCWs comparable to those observed in the mature xylem zone. On the contrary, the early differentiating xylem cells of control plants had thin primary cell walls similar to those of cambium initials. It is worth noticing that, during cold exposure, young trees continue to undergo apical and radial growth. Therefore, the xylem cells next to cambium, which exhibit thicker and more lignified cell wall, are supposed to be newly generated (see Figure S2c–f available as Supplementary Data at Tree Physiology Online). Likewise, in adult Eucalyptus trees grown in forest plantations, the SCWs were thicker in autumn (average temperature 17 ± 6 °C) and in winter (average temperature 5 ± 5 °C) than in summer (average temperature 22 ± 6 °C) (Figure 1e–g and see Figure S1 available as Supplementary Data at Tree Physiology Online). We used ImageJ software to quantify the impact of low temperature on SCW thickness in xylem (Figure 1h and i) and found a significant increase in SCW thickness (average for fibres and vessels) in young trees exposed to cold for 46 days when compared with the controls (+0.5 μm), and between winter and summer in adult trees (+1.5 μm). We also measured fibre density (number of cells per mm2) and the diameter of the vessel lumens in the xylem tissues of young and adult trees (see Figure S3 available as Supplementary Data at Tree Physiology Online). For both, we observed a significant reduction of the fibre lumen diameter in cold conditions, in agreement with the increase of SCW thickness, but no significant difference in fibre cell density. Figure 1. View largeDownload slide Cold increases the thickness of xylem SCWs. (a and b) Bright field microscopy of transverse sections of xylem from the stems of young control plants grown at 25 °C (a) and similar plants exposed to cold for 46 days (b). (c and d) Scanning electron microscopy (×1000 magnification) of sections as in (a and b). Insets show ×4000 magnifications of xylem cells located immediately below the cambium. (e–g) Transverse sections of xylem from adult trees growing in field conditions. Xylem was collected from the main trunk in summer (e), autumn (f) and winter (g). (h and i) Histograms showing the SCW thickness (average of fibres and vessels) in xylem of young plants before and after 2, 15 and 46 days of cold (h) and in xylem of adult Eucalyptus trees in summer, autumn and winter (i). For each time point, >15,000 measurements were performed by using ImageJ software on 7–20 stem sections from independent trees. Significant differences compared to the control (0 day) for young trees and compared to samples collected in summer for adult trees are indicated by asterisks, where *P < 0.05, **P < 0.005 and ***P < 0.001 according to Student’s t-test. X, xylem; P, phloem; Ca, cambial zone. 0 d, control; 2 d, 2 days; 15 d, 15 days; 46 d, 46 days of cold treatment. S, summer; A, autumn; W, winter. Figure 1. View largeDownload slide Cold increases the thickness of xylem SCWs. (a and b) Bright field microscopy of transverse sections of xylem from the stems of young control plants grown at 25 °C (a) and similar plants exposed to cold for 46 days (b). (c and d) Scanning electron microscopy (×1000 magnification) of sections as in (a and b). Insets show ×4000 magnifications of xylem cells located immediately below the cambium. (e–g) Transverse sections of xylem from adult trees growing in field conditions. Xylem was collected from the main trunk in summer (e), autumn (f) and winter (g). (h and i) Histograms showing the SCW thickness (average of fibres and vessels) in xylem of young plants before and after 2, 15 and 46 days of cold (h) and in xylem of adult Eucalyptus trees in summer, autumn and winter (i). For each time point, >15,000 measurements were performed by using ImageJ software on 7–20 stem sections from independent trees. Significant differences compared to the control (0 day) for young trees and compared to samples collected in summer for adult trees are indicated by asterisks, where *P < 0.05, **P < 0.005 and ***P < 0.001 according to Student’s t-test. X, xylem; P, phloem; Ca, cambial zone. 0 d, control; 2 d, 2 days; 15 d, 15 days; 46 d, 46 days of cold treatment. S, summer; A, autumn; W, winter. Effects of cold on lignin content in differentiating xylem We used histochemical staining to characterize further the effects of cold on xylem cells (Figure 2). Safranin–Astra Blue stains blue those cell walls rich in cellulose, such as those of cambium initials, and stains red the lignified cell walls, such as those of mature xylem cells. In young control trees (grown at 25 °C), several layers of immature xylem cells in close proximity to cambium cells presented thin, weakly lignified SCWs stained both in red and blue (Figure 2a). After 46 days of cold exposure, xylem cells in the corresponding zone had thicker cell walls and more intense red staining, indicating higher lignin content (Figure 2b). Similar differences were observed between adult tree samples collected in summer (Figure 2c) and those collected in autumn (Figure 2d) and winter (Figure 2e), in which Safranin stained red the SCWs of xylem cells located immediately below the cambium. Figure 2. View largeDownload slide Cold enhances lignin deposition in xylem SCW. (a–e) Safranin–Astra Blue double staining of transverse sections from the stems of young plants (a, control; b, 46 days of cold exposure) and adult trees (c, summer; d, autumn; e, winter). Phenolic compounds are stained red and polysaccharides blue. X, xylem; Ca, cambium; P, phloem. (f and g) Klason lignin content in bark-stripped stems of young plants and in xylem scrapings of adult trees. Significant differences from the control (0 day) for young plants and from samples collected in summer for adult trees are indicated as in Figure 1 (n = 6–8). Abbreviations as in Figure 1. Figure 2. View largeDownload slide Cold enhances lignin deposition in xylem SCW. (a–e) Safranin–Astra Blue double staining of transverse sections from the stems of young plants (a, control; b, 46 days of cold exposure) and adult trees (c, summer; d, autumn; e, winter). Phenolic compounds are stained red and polysaccharides blue. X, xylem; Ca, cambium; P, phloem. (f and g) Klason lignin content in bark-stripped stems of young plants and in xylem scrapings of adult trees. Significant differences from the control (0 day) for young plants and from samples collected in summer for adult trees are indicated as in Figure 1 (n = 6–8). Abbreviations as in Figure 1. An increase in the insoluble lignin (called Klason lignin) content of 1.5% (Figure 2f) and 1.9% (Figure 2g) was found by analysis of the samples from young and adult eucalypts submitted to cold. In addition to this moderate but statistically significant increase in lignin content, which we confirmed by analytical pyrolysis (Table 1), the amount of ethanol-extractible compounds increased fivefold in xylem samples from cold-treated young plants when compared with controls (Table 1). Table 1. Analyses of SCW composition in xylem. Wood composition was assessed in young control plants (0 day), in similar plants after 2 days, 15 days or 46 days in the cold and in adult trees in summer, autumn and winter. Ethanol extractives, Klason acid-insoluble lignin and pyrolysis lignin content are expressed in % of cell-wall mass. S, G and S+G β-O-4 yields were obtained by thioacidolysis. They are expressed in μmoles per gram of cell wall (G and S yield) or μmoles per gram of Klason lignin (for S+G yield KL). Bold values represent significant differences (Student’s t-test, n = 6–8; a, P < 0.05; b, P < 0.01; c, P < 0.001) compared with controls (0 day, for young plants and Summer for adult plants); –, not analysed. Sample type  EtOH extractives  Lignin content  S/G ratio  Klason  Pyrolysis  G yield(β-O-4)  S yield(β-O-4)  S+G yield KL  Pyrolysis  Thioacidolysis  0 day  1.43 ± 0.74  19.3 ± 1.3  24.6 ± 1.1  167 ± 26  814 ± 152  5070 ± 1237  1.34 ± 0.1  4.9 ± 0.5  2 days  3.91 ± 0.65c  20.3 ± 0.3  24.6 ± 0.2  183 ± 31  850 ± 83  5091 ± 614  1.36 ± 0.04  4.7 ± 0.3  15 days  6.46 ± 0.63c  20 ± 0.8  23.9 ± 0.9  152 ± 31  742 ± 109  4471 ± 513  1.45 ± 0.02  5 ± 1  46 days  7.66 ± 1.73c  20.8 ± 1.3a  25.8 ± 1a  179 ± 39  859 ± 156  4997 ± 956  1.34 ± 0.06  4.8 ± 0.4  Summer  –  21.6 ± 0.97  –  93 ± 30  387 ± 120  2386 ± 610  –  4.11 ± 0.8  Autumn  –  23.6 ± 0.9c  –  136 ± 21c  492 ± 94a  2260 ± 641  –  3.38 ± 0.57b  Winter  –  23.5 ± 1.1c  –  127 ± 22c  461 ± 118  2158 ± 946  –  3.48 ± 0.6a  Sample type  EtOH extractives  Lignin content  S/G ratio  Klason  Pyrolysis  G yield(β-O-4)  S yield(β-O-4)  S+G yield KL  Pyrolysis  Thioacidolysis  0 day  1.43 ± 0.74  19.3 ± 1.3  24.6 ± 1.1  167 ± 26  814 ± 152  5070 ± 1237  1.34 ± 0.1  4.9 ± 0.5  2 days  3.91 ± 0.65c  20.3 ± 0.3  24.6 ± 0.2  183 ± 31  850 ± 83  5091 ± 614  1.36 ± 0.04  4.7 ± 0.3  15 days  6.46 ± 0.63c  20 ± 0.8  23.9 ± 0.9  152 ± 31  742 ± 109  4471 ± 513  1.45 ± 0.02  5 ± 1  46 days  7.66 ± 1.73c  20.8 ± 1.3a  25.8 ± 1a  179 ± 39  859 ± 156  4997 ± 956  1.34 ± 0.06  4.8 ± 0.4  Summer  –  21.6 ± 0.97  –  93 ± 30  387 ± 120  2386 ± 610  –  4.11 ± 0.8  Autumn  –  23.6 ± 0.9c  –  136 ± 21c  492 ± 94a  2260 ± 641  –  3.38 ± 0.57b  Winter  –  23.5 ± 1.1c  –  127 ± 22c  461 ± 118  2158 ± 946  –  3.48 ± 0.6a  We analysed the lignin structure by thioacidolysis, which specifically targets the G and S units involved in labile β-O-4 linkages (Méchin et al. 2014). Thioacidolysis yields expressed per gram of Klason lignin, which reflect the overall amounts of lignin monomers linked by β-O-4 bounds, were not affected by cold in both young plants and adult trees (Table 1). The S/G ratio was not significantly altered in young trees but was significantly reduced in adult trees: in autumn and winter, G units increased 46 and 37%, respectively, when compared with the summer level and S units were 27 and 19% higher, respectively, in autumn and winter than in summer. Finally, the cP/cH ratio, which indicates the relative proportion of hemicelluloses (pentosanes) compared with cellulose (hexosanes) (Lepikson-Neto et al. 2013), significantly increased in young plants after 46 days of cold treatment (from 25.7 ± 2.7 in 0 day samples to 29.6 ± 3.6 in 46 days samples, t-test, n = 10, P < 0.05; see Figure S4 available as Supplementary Data at Tree Physiology Online). This suggests that the deposition of hemicelluloses could be cold-induced in Eucalyptus as reported in annual plants (LeGall et al. 2015). Transcriptional regulation of genes related to SCW biosynthesis We performed RT-qPCR analyses on young Eucalyptus trees sampled at 2, 15 and 46 days after exposure to cold and compared with controls (0 days) (see Table S1 available as Supplementary Data at Tree Physiology at Tree Physiology Online and Figure 3). A twofold change of the transcript level ratio (between cold conditions and control) was defined as a threshold above which a gene was considered differentially expressed in response to cold. We first analysed the transcript level of a member of the DREB/CBF TF gene family, EgCBF14, a marker of cold stress known to be highly expressed in E. gunnii stems (Cao et al. 2015); EgCBF14 was induced 300-fold after 2 days of cold treatment and stayed highly expressed after 46 days when compared with controls (see Table S1 available as Supplementary Data at Tree Physiology Online). Figure 3. View large Download slide Cold induces genes of the lignin biosynthesis pathway in Eucalyptus. (a) The transcript levels of the genes encoding the 11-enzymatic steps involved in the monolignol biosynthetic pathway were analysed by RT-qPCR in response to cold. They include the 17 genes (marked in bold) identified as being involved in developmental lignification, belonging to the ‘core vascular lignin toolbox’ (Carocha et al. 2015). In addition, three close orthologues of Arabidopsis laccases 11, 4 and 17, were analysed. Because they represent a minor pathway in woody angiosperms, the enzymatic steps leading to hydroxyphenyl (H units) were not represented. Green and black arrows represent enzymatic steps induced or not induced by cold, respectively. The heat map illustrates the gene expression ratios in young plants exposed to cold for 2, 15 and 46 days compared with controls. The white to green colour scale represents log10 of expression ratio (2 , 15 or 46 days versus control). (b) The heat map represents, for each enzymatic step, the transcript abundance of each gene versus the total transcript production of the gene family expressed as percentage before and after cold treatment. The white to red colour scale represents 0 to 100% of transcript contribution. Expression values are provided in Table S1 available as Supplementary Data at Tree Physiology Online. Genes short names were given according to Carocha et al. (2015): PAL, phenylalanine ammonialyase; C4H, cinnamate-4-hydroxylase; 4CL, 4-hydroxycinnamate CoA ligase; HCT, hydroxycinnamoyl transferase; C3H, p-coumarate 3-hydroxylase; CCR, cinnamoyl CoA reductase; CCoAOMT, caffeoyl-CoA 3-O-methyltransferase; F5H, ferulate 5-hydroxylase; COMT, caffeic acid O-methyltransferase; CAD, cinnamyl alcohol dehydrogenase. Figure 3. View large Download slide Cold induces genes of the lignin biosynthesis pathway in Eucalyptus. (a) The transcript levels of the genes encoding the 11-enzymatic steps involved in the monolignol biosynthetic pathway were analysed by RT-qPCR in response to cold. They include the 17 genes (marked in bold) identified as being involved in developmental lignification, belonging to the ‘core vascular lignin toolbox’ (Carocha et al. 2015). In addition, three close orthologues of Arabidopsis laccases 11, 4 and 17, were analysed. Because they represent a minor pathway in woody angiosperms, the enzymatic steps leading to hydroxyphenyl (H units) were not represented. Green and black arrows represent enzymatic steps induced or not induced by cold, respectively. The heat map illustrates the gene expression ratios in young plants exposed to cold for 2, 15 and 46 days compared with controls. The white to green colour scale represents log10 of expression ratio (2 , 15 or 46 days versus control). (b) The heat map represents, for each enzymatic step, the transcript abundance of each gene versus the total transcript production of the gene family expressed as percentage before and after cold treatment. The white to red colour scale represents 0 to 100% of transcript contribution. Expression values are provided in Table S1 available as Supplementary Data at Tree Physiology Online. Genes short names were given according to Carocha et al. (2015): PAL, phenylalanine ammonialyase; C4H, cinnamate-4-hydroxylase; 4CL, 4-hydroxycinnamate CoA ligase; HCT, hydroxycinnamoyl transferase; C3H, p-coumarate 3-hydroxylase; CCR, cinnamoyl CoA reductase; CCoAOMT, caffeoyl-CoA 3-O-methyltransferase; F5H, ferulate 5-hydroxylase; COMT, caffeic acid O-methyltransferase; CAD, cinnamyl alcohol dehydrogenase. We then analysed the effect of cold on the transcript levels of genes involved in the biosynthesis of SCW polymers, i.e., lignin, cellulose and hemicelluloses (see Table S1 available as Supplementary Data at Tree Physiology Online). Among the 11 phenylpropanoid gene families involved in the monolignol biosynthesis pathway, the vast majority of the members identified as being recruited for lignin biosynthesis during development belonging to the so-called ‘core vascular lignification toolbox’ (Carocha et al. 2015) were significantly induced by cold, although moderately (marked in bold in see Table S1 available as Supplementary Data at Tree Physiology Online and Figure 3a). Of these genes, the two most highly induced were the two members of the C4H family, EgC4H1 (24.5-fold induction) and EgC4H2 (7.5-fold induction); the latter was already known to be inducible by abiotic stress responses in eucalypts (Carocha et al. 2015). It is worth noting that EgF5H1 and EgCOMT1, which encode enzymes catalysing the two key steps in S unit biosynthesis, were not significantly induced in response to cold (see Table S1 available as Supplementary Data at Tree Physiology Online and Figure 3a). The levels of induction of the other family members in response to cold were in general much higher than those of the ‘core vascular lignification toolbox’ genes. This was mainly due to their lowest expression in xylem in control conditions (see Table S1 available as Supplementary Data at Tree Physiology Online and Figure 3a). For instance, in the PAL family, EgPAL1 thought to be involved in anthocyanin production was induced 10-fold and EgPAL8, whose role is unknown, was induced up to 330-fold (see Table S1 available as Supplementary Data at Tree Physiology Online). In the COMT family, except EgCOMT1, all the genes were induced from 12-fold up to 618-fold (for EgCOMT6) (see Table S1 available as Supplementary Data at Tree Physiology Online). For each enzymatic step, we compared before and after cold treatment, the transcript abundance of each gene to the total transcript production of the family expressed as percentage (Figure 3b). This comparison showed that in most cases, main transcript contributions were still attributed to the ‘core vascular lignification toolbox’ genes. The enzymes encoded by the latter are likely keeping prominent roles in lignin biosynthesis in response to cold temperatures. However, genes encoding EgPAL8, EgHCT2 and EgCOMT6, which are not part of this ‘lignification toolbox’, were dramatically induced in response to cold and reached expression levels similar to the prominent ‘lignification toolbox’ genes EgPAL3, EgHCT4 and EgCOMT1, respectively. Two Eucalyptus laccases orthologues of AtLac11 and AtLac4 (involved the lignin polymerization step in Arabidopsis (Zhao et al. 2013) were induced by cold ninefold and threefold, respectively (see Table S1 available as Supplementary Data at Tree Physiology Online)). Considering polysaccharides biosynthesis, two of the four cellulose synthases genes analysed (EgCesA2 and EgCesA6) (Ranik and Myburg 2006) were induced fivefold and twofold, respectively, whereas only one of the eight genes involved in hemicelluloses biosynthesis—the closest Eucalyptus orthologue of AtIRX10 (GUT2, a member of GT47 family, Wu et al. 2009)—was induced more than fourfold by cold (see Table S1 available as Supplementary Data at Tree Physiology Online). Xylem SCW transcriptional regulators induced by cold A recent study in Arabidopsis demonstrated that abiotic stresses (high salinity and iron deprivation) can induce genes that regulate biosynthesis of the SCW (Taylor-Teeples et al. 2015). To investigate whether the response to cold in Eucalyptus might also induce genes that regulate SCW biosynthesis, we selected 25 close orthologues of the Arabidopsis TFs involved in the three-levels hierarchical SCW regulatory network (Figure 4), building on the works of Hussey et al. (2013) and Soler et al. (2015) and we analysed their transcript levels in young trees exposed to cold (Figure 4). Fourteen of the 25 genes were differentially expressed following cold exposure: seven were up-regulated and seven were down-regulated (Figure 4 and see Table S1 available as Supplementary Data at Tree Physiology Online). Four genes presented a non-significant twofold induction or repression. The differentially regulated genes were found mainly in the lower third level of the SCW regulatory network (Level 3). These TFs act directly on the structural genes involved in the biosynthesis of the SCW polymers. Six of the up-regulated genes are positive regulators of SCW deposition (i.e., orthologues of AtMYB20, AtMYB54, AtMYB58, AtMYB63, AtMYB69 and AtMYB79), whereas three of the down-regulated genes are negative regulators of SCW deposition (i.e., orthologues of AtMYB7, AtMYB75 and AtKNAT7). Only 4 of the 14 genes significantly affected by cold are involved in the upper part of the regulatory network (Levels 1 and 2). Of these, three were repressed (orthologues of AtMYB83, AtMYB55 and AtSND1) and one was induced (one orthologue of AtSND3). Figure 4. View large Download slide Secondary cell wall-related MYB and NAC TFs are regulated by cold to promote SCW deposition. (a) The 25 genes represented are the closest eucalypts orthologues of Arabidopsis NACs and MYBs TFs involved in the control of SCW deposition (Hussey et al. 2013). For each gene, heat maps represent expression ratios after 2, 15 and 46 days of cold exposure normalized by control (0 day). The red to green colour scale represents log2 of expression ratios (available in Table S1 available as Supplementary Data at Tree Physiology Online). Gene names in green indicate those induced more than twofold by cold and those in red indicate those repressed by more than twofold. The genes marked with asterisks indicate those whose transcription was significantly up- or down-regulated at two or more time points (15 d, 15 days; n = 6, P < 0.05, according to Student’s t-test). (b) The genes are grouped into the three-level hierarchical network of TFs that regulates xylem cell differentiation and deposition of lignin, cellulose and hemicelluloses in SCWs of Arabidopsis (adapted from Hussey et al. 2013). The putative orthologues of Arabidopsis NAC and MYB genes in eucalypts are indicated in brackets below their Arabidopsis counterparts. Arrows indicate positive regulation and the blocked line indicates negative regulation. Figure 4. View large Download slide Secondary cell wall-related MYB and NAC TFs are regulated by cold to promote SCW deposition. (a) The 25 genes represented are the closest eucalypts orthologues of Arabidopsis NACs and MYBs TFs involved in the control of SCW deposition (Hussey et al. 2013). For each gene, heat maps represent expression ratios after 2, 15 and 46 days of cold exposure normalized by control (0 day). The red to green colour scale represents log2 of expression ratios (available in Table S1 available as Supplementary Data at Tree Physiology Online). Gene names in green indicate those induced more than twofold by cold and those in red indicate those repressed by more than twofold. The genes marked with asterisks indicate those whose transcription was significantly up- or down-regulated at two or more time points (15 d, 15 days; n = 6, P < 0.05, according to Student’s t-test). (b) The genes are grouped into the three-level hierarchical network of TFs that regulates xylem cell differentiation and deposition of lignin, cellulose and hemicelluloses in SCWs of Arabidopsis (adapted from Hussey et al. 2013). The putative orthologues of Arabidopsis NAC and MYB genes in eucalypts are indicated in brackets below their Arabidopsis counterparts. Arrows indicate positive regulation and the blocked line indicates negative regulation. In addition to the orthologues of known regulators of SCW in Arabidopsis, potential new SCW regulators have been pointed out by the works of Hussey et al. (2015) and Soler et al. (2015). To identify the most promising candidate regulators among these eucalypt genes and to evaluate their responses to cold, we used co-expression analyses that have been shown to be useful to identify new genes involved in SCW-related synthesis (Ruprecht and Persson 2012). We first built a gene co-expression network based on transcript profiling of 146 candidate SCW-related genes in a large panel of eucalypt organs and tissues, stages of development and/or environmental conditions previously described in Cassan-Wang et al. (2012), Camargo et al. (2014), Hussey et al. (2015), Soler et al. (2015) and Carocha et al. (2015). This set of genes comprised lignin and cellulose biosynthesis genes, as well as known and as yet uncharacterized NAC and MYB TFs. Pearson’s correlation coefficient is commonly used to estimate transcriptional coordination between two genes, suggesting their involvement in the same biological process (Aoki et al. 2007). We thus generated a Pearson correlation matrix to build a co-expression network in which 132 of the 146 genes (nodes) were connected by 592 edges representing significant co-expression relationships (see Table S2 available as Supplementary Data at Tree Physiology Online). These correlations were represented graphically in a network regrouping 93 of the 132 genes (Figure 5). The remaining 39 genes were grouped in smaller networks (see Figure S5 available as Supplementary Data at Tree Physiology Online). Because we focused our interest in woody tissues, we set up node sizes to be proportional to the expression ratio between woody and non-woody tissues. Edge lengths are inversely proportional to the absolute value of correlation for each gene pair comparison, so that genes in close vicinity are considered as strongly correlated. This approach led to three clusters of genes, which were clearly separated and in which almost all the significant correlations detected were positive (Figure 5). We used a hierarchical clustering approach to investigate the tissue preferential expression of the genes contained in each of these clusters (see Figure S6 available as Supplementary Data at Tree Physiology Online). Based on this, we called the upper cluster ‘xylem cluster’ because it contained 47 genes preferentially expressed in xylem and related to SCW deposition, i.e., lignin toolbox genes, cellulose synthases and orthologues of known TFs that regulate SCW biosynthesis (Figure 5). Interestingly, it also contained uncharacterized TFs. The intermediate or ‘cambium cluster’ contained 17 genes expressed preferentially in vascular cambium, mainly composed of TFs of unknown function (with no orthologues in Arabidopsis). The lower ‘non-woody tissues’ cluster contained 29 genes; of these, 11 encoded enzymes of unknown function (COMT_like, CAD_like, HCT_like) and 18 were unknown TFs. We chose a subset of 23 uncharacterized TFs, contained both in the ‘xylem’ and the ‘cambium’ clusters, to investigate their responses to cold temperatures by measuring their transcript levels in cold acclimated young plants (squares, Figure 5). Eleven of these genes were up-regulated by cold and three were down-regulated. Those of the ‘xylem cluster’ were being either up- or down-regulated by cold from twofold to eightfold. The 10 TFs of the ‘cambium cluster’ regulated by cold were all induced, some being dramatically up-regulated such as EgMYB64 (>5000-fold) and EgMYB59 (>100-fold). Figure 5. View large Download slide Correlation network analyses highlight xylem and cambium clusters enriched in cold responsive TFs. Transcript profiling of 146 genes encoding members of the MYB, NAC families of TFs as well as enzymes involved in SCW biosynthesis in a panel of 22 conditions were obtained from previous studies in eucalypts (see Table S2 available as Supplementary Data at Tree Physiology Online) (Cassan-Wang et al. 2012, Carocha et al. 2015, Hussey et al. 2015, Soler et al. 2015). Significant pairwise Pearson correlation values (Bonferroni adjusted, P < 0.05) were used to draw a co-expression network. A short link between two genes represents a strong correlation, explained by similar expression profiles in the 22 conditions analysed. Grey and red edges represent respectively positive and negative correlations. Gene names and accession numbers are listed in Table S2 available as Supplementary Data at Tree Physiology Online. Node sizes are proportional to the expression ratio between woody and non-woody tissues (see Table S2 available as Supplementary Data at Tree Physiology Online). Circles represent genes related to SCW formation (enzymes or TFs), squares represent new, uncharacterized TFs. The main network shown here is composed of 93 genes (nodes) linked by 520 relationships (edges); small isolated sub-networks regrouping 39 genes are provided in Figure S5 available as Supplementary Data at Tree Physiology Online. The three clusters of the main network were named according to the tissue preferential expression of their gene members (‘xylem’, ‘cambium’ or ‘non-woody tissues’; see Figure S6 available as Supplementary Data at Tree Physiology Online). Transcript levels were measured in young eucalypts plants submitted to cold (2, 15 and 46 days at 4 °C) and compared with control (25 °C, 0 days). Node colour gradient (red to green) represents gene expression ratios between cold and control conditions (see Table S1 available as Supplementary Data at Tree Physiology Online; ratio = Coldmean(15 days/46 days)/Control(0 days)). Green and red colours represent gene induction or repression in response to cold, respectively. Grey nodes represent genes with no expression data. Figure 5. View large Download slide Correlation network analyses highlight xylem and cambium clusters enriched in cold responsive TFs. Transcript profiling of 146 genes encoding members of the MYB, NAC families of TFs as well as enzymes involved in SCW biosynthesis in a panel of 22 conditions were obtained from previous studies in eucalypts (see Table S2 available as Supplementary Data at Tree Physiology Online) (Cassan-Wang et al. 2012, Carocha et al. 2015, Hussey et al. 2015, Soler et al. 2015). Significant pairwise Pearson correlation values (Bonferroni adjusted, P < 0.05) were used to draw a co-expression network. A short link between two genes represents a strong correlation, explained by similar expression profiles in the 22 conditions analysed. Grey and red edges represent respectively positive and negative correlations. Gene names and accession numbers are listed in Table S2 available as Supplementary Data at Tree Physiology Online. Node sizes are proportional to the expression ratio between woody and non-woody tissues (see Table S2 available as Supplementary Data at Tree Physiology Online). Circles represent genes related to SCW formation (enzymes or TFs), squares represent new, uncharacterized TFs. The main network shown here is composed of 93 genes (nodes) linked by 520 relationships (edges); small isolated sub-networks regrouping 39 genes are provided in Figure S5 available as Supplementary Data at Tree Physiology Online. The three clusters of the main network were named according to the tissue preferential expression of their gene members (‘xylem’, ‘cambium’ or ‘non-woody tissues’; see Figure S6 available as Supplementary Data at Tree Physiology Online). Transcript levels were measured in young eucalypts plants submitted to cold (2, 15 and 46 days at 4 °C) and compared with control (25 °C, 0 days). Node colour gradient (red to green) represents gene expression ratios between cold and control conditions (see Table S1 available as Supplementary Data at Tree Physiology Online; ratio = Coldmean(15 days/46 days)/Control(0 days)). Green and red colours represent gene induction or repression in response to cold, respectively. Grey nodes represent genes with no expression data. One way to verify the accuracy of the co-expression clusters is to investigate whether orthologous genes are also co-expressed in other species (Ruprecht and Persson 2012). We extracted, for instance, the subset of genes co-expressed with the Eucalyptus orthologue (EgNAC61) of the fist level master regulator AtSND1 (Figure 6a). The EgNAC61 node-vicinity network included second and third levels TFs of the hierarchical SCW regulating network (see Figure 5) as well as lignin and cellulose biosynthesis genes (Figure 6a). The Arabidopsis orthologues of all these genes are also co-expressed with AtSND1, which regulates their transcription levels (Ruprecht et al. 2011, Yao et al. 2012) thereby strengthening the validity of the co-expression network. Interestingly, the EgNAC61 node-vicinity network also contained new, as yet uncharacterized TFs, such as EgNAC64 and EgMYB137, which have no orthologues in Arabidopsis, and which were both significantly induced by cold (see Table S1 available as Supplementary Data at Tree Physiology Online). The co-expression network of EgNAC64 is composed of 19 genes mostly SCW-related genes, more than that of EgNAC61 (AtSND1) (Figure 6b). Remarkably, it contains EgNAC61 itself, suggesting that EgNAC64 could also be a SCW master regulator. This led us to investigate more precisely the phylogenetic relationships between Arabidopsis and Eucalyptus NAC sequences (Hussey et al. 2015). We found that EgNAC64 belongs to the NAC subgroup II and could be a co-orthologue of AtSND3, a second-level SCW master regulator in Arabidopsis whose closest co-orthologues are four Eucalyptus genes duplicated in tandem. We then extracted the co-expression network of EgMYB137, which has no close orthologues in Arabidopsis (Soler et al. 2015). It contains nine genes of which five are SCW regulators, and one is a lignin biosynthesis gene CCoAOMT1 supporting a possible involvement in SCW regulation in response to stress (Figure 6c). It also includes two as yet uncharacterized TFs. Finally, we extracted the sub-network of EgCOMT6, (Figure 6d), the most strongly induced COMT gene in response to cold. EgCOMT6 which belongs to the cambium cluster, presents a high degree of connectivity with nine uncharacterized MYB TFs, which form the core of a module enriched in stress-induced woody-specific genes preferentially expressed in cambium. Figure 6. View largeDownload slide Xylem and cambium sub-networks reveal promising new candidate genes. Sub-networks were generated by extracting the genes directly correlated to (a) EgNAC61 (AtSND1), (b) EgNAC64, (c) EgMYB137 or (d) EgCOMT6. For legend, see Figure 5. Yellow nodes represent the ‘hub gene’ correlated to all the others. Figure 6. View largeDownload slide Xylem and cambium sub-networks reveal promising new candidate genes. Sub-networks were generated by extracting the genes directly correlated to (a) EgNAC61 (AtSND1), (b) EgNAC64, (c) EgMYB137 or (d) EgCOMT6. For legend, see Figure 5. Yellow nodes represent the ‘hub gene’ correlated to all the others. Discussion Cell-wall adjustment under abiotic stress is an important process in plant adaptation to cope with non-optimal growth conditions (LeGall et al. 2015). Cell wall changes induced by cold have been studied extensively in crop plants and mostly in their leaves, but they have been only rarely investigated in trees and never in non-dormant trees such as eucalypts. Here, we report that long exposure to cold increases SCW thickening in the developing xylem cells of the cold-resistant hybrid E. gundal. This was observed in young trees and to an even greater extent in adult trees. These thicker SCWs were enriched in Safranin-stained, mostly phenolic compounds. On one hand, both Klason lignin analysis and analytical pyrolysis suggest that at least part of these phenolic compounds were incorporated into lignin polymers, resulting in more lignin in the SCWs of plants submitted to cold conditions. On the other hand, the observed fivefold increase in ethanol-extractible compounds suggests that part of these phenolics stayed soluble or loosely linked to the cell walls. As suggested previously for Miscanthus and wheat (Olenichenko and Zagoskina 2005, Domon et al. 2013), these compounds might increase cell-wall rigidity by crosslinking with cell-wall polymers and/or might contribute to xylem cell protection from reactive oxygen species. The greater lignin content in the xylem SCWs of cold-treated eucalypts is in good agreement with the up-regulation of most genes belonging to the ‘core vascular lignification toolbox’. In young trees, all lignin genes whose products are involved in G unit biosynthesis were induced by cold, whereas genes dedicated to S unit biosynthesis (encoding EgF5H1 and EgCOMT1, Anterola and Lewis 2002) were not differentially expressed. This differential expression of genes involved in G and S unit biosynthesis in response to cold, which might result in relatively more G units than S units, may explain the significant decrease in the S/G ratio observed in adult trees in autumn and winter compared with that in summer. In general, the genes most highly induced by cold were those not involved in developmental lignification. These stress-inducible genes, which are often poorly expressed in xylem in control conditions, encode isoforms (EgPAL8, EgHCT1 and EgHCT2, EgCCR2, EgCOMT2–EgCOMT6), whose activities might lead to the increased soluble phenolics content and/or to the increased lignin content. Of the SCW polysaccharides, we observed a higher ratio of pentosanes (cP) to hexosanes (cH) in xylem samples from young plants exposed to cold than in control plants, suggesting that more hemicelluloses than cellulose might be incorporated into the SCWs in cold conditions. However, at the transcript gene level, we observed very significant inductions by cold of both a SCW-specific cellulose synthase gene EgCesA2 (Eucgr.A01324; Myburg et al. 2014), and of EgIRX10, which encodes a GT47 glycosyl transferase involved in chain elongation of the most abundant hemicelluloses, i.e., xylan. In Arabidopsis, a mutant deficient for IRX10 and IRX10_like proteins showed a dramatic loss of SCW deposition associated with reduced xylan backbone polymerization and a complete loss of glucuronic acid side chains (Brown et al. 2009). Based on these results, we hypothesize that induction of EgIRX10 in Eucalyptus xylem in response to cold increases SCW deposition and may account, at least in part, for the observed increased SCW thickness in plants exposed to cold. In summary, we observed increased SCW thickness likely resulting from enhanced deposition of lignin and possibly of xylan in E. gundal developing xylem cells submitted to cold conditions. An increased thickness of primary walls has been frequently reported in crop plants in response to various abiotic stresses and is considered as one of the main common mechanisms of stress responses in cell-wall architecture that could enable functional adaptation to abiotic stress. Increased lignin and/or hemicelluloses content is proposed to modulate cell-wall rigidity to prevent cell damage due to freezing (LeGall et al. 2015). Our findings are consistent with a model in which reinforcement of SCWs could be one mechanism by which perennial plants cope with unfavourable conditions such as cold and might help them adapt to frost. The content and composition of lignin polymers in SCWs seem to be modulated by transcriptional regulation of genes encoding enzymes involved in developmental lignification and of stress-responsive genes such as EgCOMT6. We cannot exclude, however, that such genes could be responsible for the synthesis of other phenylpropanoid compounds in response to cold. EgCOMT6, for instance is tightly co-expressed with cambium woody-specific MYB TFs whose role in SCW regulation based on co-expression networks is not obvious. Recently, one of these TFs, EgMYB88, has been functionally characterized in poplar and shown to control the biosynthesis of phenylpropanoid-derived secondary metabolites including lignin (Soler et al. 2016). The E. grandis orthologues of Arabidopsis TFs belonging to the three-levels SCW hierarchical network established for the model plant were, for a large majority, positioned in the ‘xylem co-expresssion cluster’. Cold temperatures triggered a clear regulation pattern for the most downstream TFs, downregulation of the negative regulators and induction of the positive regulators, in good agreement with the increase of SCW thickness and lignin content observed in response to cold. The ‘xylem cluster’ also included the ‘core vascular lignin toolbox’ genes and SCW-specific cellulose synthases. Co-expression between transcriptional regulators and effector genes ensuring SCW deposition has already been described and shown to be consistent across species (Ruprecht and Persson 2012). Indeed, the EgNAC61 (AtSND1) sub-network included the closest Arabidopsis orthologues of TFs controlling SCW deposition, as well as cellulose and lignin related genes, known to be co-expressed and transcriptionally regulated by AtSND1 (Zhong et al. 2006, Yao et al. 2012), and is, therefore, highly conserved between Eucalyptus and Arabidopsis. The analysis of the EgNAC64 network highlighted a hub position of this TF and suggested that it could be a functional orthologue of AtSND3 given its phylogenetic proximity even if it is not the closest Eucalyptus orthologue. EgMYB137 is a new uncharacterized TF, co-expressed with EgNAC61 (AtSND1) and other orthologues of Arabidopsis SCW-master regulators, like EgMYB2 (AtMYB46) or EgNAC49 (AtNST1/2). EgNAC64 and EgMYB137 are induced by cold stress and might lead, at least in part, to the phenylpropanoid/lignin increase and the S/G ratio decrease triggered by chilling temperatures. Notably, the use of a woody plant allowed a physical separation of differentiating xylem from vascular cambium (Cassan-Wang et al. 2012) and thus enabled to highlight in our co-expression analysis a ‘cambium co-expression cluster’ that obviously cannot be characterized in Arabidopsis. This cluster contains mostly yet uncharacterized MYB TFs, belonging to woody-preferential clades, not present in Arabidopsis and many of them are cold-induced. Twelve of them are also correlated to stress-inducible genes of the phenylpropanoid pathway, like EgCOMT6 or EgPAL8 and could be involved in lignin deposition or soluble phenolics biosynthesis. In conclusion, this work provides a strong foundation to decipher the function of new TF candidates at the crosstalk between cold stress and wood tissue formation. This is particularly relevant in Eucalyptus, which, as a woody evergreen perennial tree, must ensure its fitness and adaptability to challenging environmental cues. This study also provides a global view of the consequences of cold temperatures on wood formation in Eucalyptus. The observed increase in the amount of lignin and the modified lignin composition (lower S/G ratio) may contribute to greater tolerance of cold in E. gunnii × E. dalrympleana hybrids, as in crop plants, but these SCW modifications may also greatly affect the properties of the wood and consequently its industrial uses. In this respect, the identification of regulators that act both in the developmental patterning of wood tissues and in the abiotic stress response will be extremely useful for both wood engineering and breeding strategies as stressed by Mashkina and Butorina (2003) and Harfouche et al. (2011). Supplementary Data Supplementary Data for this article are available at Tree Physiology Online. Acknowledgments The authors acknowledge F. Melun (FCBA) for his help with the identification of field-grown eucalypt clones in Longages (France), R. Simões (CEF) for the extraction of xylem samples, B. Savelli and H. San Clemente (LRSV) for their help with bioinformatics, A. Gauthier and A. Desplat for their technical help during their Master internships, the Genotoul Bioinformatics Platform Toulouse Midi-Pyrenees for computing and storage resources and the TRI-Genotoul platform for microscopic analyses. The authors are grateful to Prof. C. Lapierre (Institut JP Bourgin, France) and Dr Y. Barrière (INRA Lusignan, France) for helpful discussions and advice on lignin analyses. The authors acknowledge Centre National pour la Recherche Scientifique (CNRS), the University Paul Sabatier Toulouse III, the Agence Nationale de la Recherche (ANR), the Fundação para a Ciência e a Tecnologia (FCT) and the Ministère de l'Education Nationale, de l'Enseignement Supérieur et de la Recherche for the support of this work. Funding This work was supported by the Centre National pour la Recherche Scientifique (CNRS), the University Paul Sabatier Toulouse III (UPS, IDEX UNITI, project EuCoWood), the TREEFORJOULES project (ANR-2010-KBBE-007-01 and FCT, P-KBBE/AGR_GPL/0001/2010) the French Laboratory of Excellence project ‘TULIP’ (ANR-10-LABX-41; ANR-11-IDEX-0002-02). R.P. was supported by a PhD grant from the Ministère de l’Education Nationale, de l’Enseignement Supérieur et de la Recherche. References Alves A, Gierlinger N, Schwanninger M, Rodrigues J ( 2009) Analytical pyrolysis as a direct method to determine the lignin content in wood. Part 3. Evaluation of species-specific and tissue-specific differences in softwood lignin composition using principal component analysis. J Anal Appl Pyrolysis  85: 30– 37. Google Scholar CrossRef Search ADS   Anterola AM, Lewis NG ( 2002) Trends in lignin modification: a comprehensive analysis of the effects of genetic manipulations/mutations on lignification and vascular integrity. Phytochemistry  61: 221– 294. Google Scholar CrossRef Search ADS PubMed  Aoki K, Ogata Y, Shibata D ( 2007) Approaches for extracting practical information from gene co-expression networks in plant biology. Plant Cell Physiol  48: 381– 390. Google Scholar CrossRef Search ADS PubMed  Arvidsson S, Kwasniewski M, Riaño-Pachón DM, Mueller-Roeber B ( 2008) QuantPrime—a flexible tool for reliable high-throughput primer design for quantitative PCR. BMC Bioinformatics  9: 465. Google Scholar CrossRef Search ADS PubMed  Booth TH ( 1990) Mapping regions climatically suitable for particular tree species at the global scale. For Ecol Manage  36: 47– 60. Google Scholar CrossRef Search ADS   Brown DM, Zhang Z, Stephens E, Dupree P, Turner SR ( 2009) Characterization of IRX10 and IRX10-like reveals an essential role in glucuronoxylan biosynthesis in Arabidopsis. Plant J  57: 732– 746. Google Scholar CrossRef Search ADS PubMed  Camargo ELO, Nascimento LC, Soler M et al.  . ( 2014) Contrasting nitrogen fertilization treatments impact xylem gene expression and secondary cell wall lignification in Eucalyptus. BMC Plant Biol  14: 256. Google Scholar CrossRef Search ADS PubMed  Cao PB, Azar S, SanClemente H, Mounet F, Dunand C, Marque G, Marque C, Teulières C ( 2015) Genome-wide analysis of the AP2/ERF family in Eucalyptus grandis: an intriguing over-representation of stress-responsive. PLoS One  10: 1– 22. Carocha V, Soler M, Hefer C, Cassan-Wang H, Fevereiro P, Myburg AA, Paiva JAP, Grima-Pettenati J ( 2015) Genome-wide analysis of the lignin toolbox of Eucalyptus grandis. New Phytol  206: 1297– 1313. Google Scholar CrossRef Search ADS PubMed  Cassan-Wang H, Soler M, Yu H et al.  . ( 2012) Reference genes for high-throughput quantitative reverse transcription-PCR analysis of gene expression in organs and tissues of Eucalyptus grown in various environmental conditions. Plant Cell Physiol  53: 2101– 2116. Google Scholar CrossRef Search ADS PubMed  Dixon RA, Paiva NL ( 1995) Stress-induced phenylpropanoid metabolism. Plant Cell  7: 1085– 1097. Google Scholar CrossRef Search ADS PubMed  Domon JM, Baldwin L, Acket S et al.  . ( 2013) Cell wall compositional modifications of Miscanthus ecotypes in response to cold acclimation. Phytochemistry  85: 51– 61. Google Scholar CrossRef Search ADS PubMed  Goicoechea M, Lacombe E, Legay S et al.  . ( 2005) EgMYB2, a new transcriptional activator from Eucalyptus xylem, regulates secondary cell wall formation and lignin biosynthesis. Plant J  43: 553– 567. Google Scholar CrossRef Search ADS PubMed  Grima-Pettenati J, Soler M, Camargo ELO, Wang H ( 2012) Transcriptional regulation of the lignin biosynthetic pathway revisited: new players and insights. In: Jouanin L, Lapierre C (eds) Advances in botanical research: lignins biosynthesis, biodegradation and bioengineering . Elsevier, San Diego, pp 173– 218. Google Scholar CrossRef Search ADS   Harfouche A, Meilan R, Altman A ( 2011) Tree genetic engineering and applications to sustainable forestry and biomass production. Trends Biotechnol  29: 9– 17. Google Scholar CrossRef Search ADS PubMed  Hausman JF, Evers D, Thiellement H, Jouve L ( 2000) Compared responses of poplar cuttings and in vitro raised shoots to short-term chilling treatments. Plant Cell Rep  19: 954– 960. Google Scholar CrossRef Search ADS   Hussey SG, Mizrachi E, Creux NM, Myburg AA ( 2013) Navigating the transcriptional roadmap regulating plant secondary cell wall deposition. Front Plant Sci  4: 1– 21. Google Scholar CrossRef Search ADS PubMed  Hussey SG, Saidi MN, Hefer CA, Myburg AA, Grima-Pettenati J ( 2015) Structural, evolutionary and functional analysis of the NAC domain protein family in Eucalyptus. New Phytol  206: 1337– 1350. Google Scholar CrossRef Search ADS PubMed  LeGall H, Philippe F, Domon J, Gillet F, Pelloux J, Rayon C ( 2015) Cell wall metabolism in response to abiotic stress. Plants  4: 112– 166. Google Scholar CrossRef Search ADS PubMed  Legay S, Lacombe E, Goicoechea M, Brière C, Séguin A, Mackay J, Grima-Pettenati J ( 2007) Molecular characterization of EgMYB1, a putative transcriptional repressor of the lignin biosynthetic pathway. Plant Sci  173: 542– 549. Google Scholar CrossRef Search ADS   Legay S, Sivadon P, Blervacq A-S et al.  . ( 2010) EgMYB1, an R2R3 MYB transcription factor from eucalyptus negatively regulates secondary cell wall formation in Arabidopsis and poplar. New Phytol  188: 774– 786. Google Scholar CrossRef Search ADS PubMed  Lepikson-Neto J, Alves A, Simões R et al.  . ( 2013) Flavonoid supplementation reduces the extractive content and increases the syringyl/guaiacyl ratio in Eucalyptus grandis × Eucalyptus urophylla hybrid trees. BioResources  8: 1747– 1757. Google Scholar CrossRef Search ADS   Mashkina OS, Butorina AK ( 2003) Genetic engineering of forest woody plants. Genetika  39: 309– 317. Google Scholar PubMed  Méchin V, Laluc A, Legée F, Cezard L, Denoue D, Barrie Y, Lapierre C ( 2014) Impact of the brown-midrib bm5 mutation on maize lignins. J Agric Food Chem  62: 5102– 5107. Google Scholar CrossRef Search ADS PubMed  Myburg A, Potts B, Marques CM, Kirst M, Gion J, Grattapaglia D, Grima-Pettenati J ( 2007) Eucalyptus. In: Kole CE (ed) Genome mapping and molecular breeding in plants . Springer, New York, NY, USA, pp 115– 160. Myburg AA, Grattapaglia D, Tuskan GA et al.  . ( 2014) The genome of Eucalyptus grandis. Nature  510: 356– 362. Google Scholar CrossRef Search ADS PubMed  Nakano Y, Yamaguchi M, Endo H, Rejab NA, Ohtani M ( 2015) NAC-MYB-based transcriptional regulation of secondary cell wall biosynthesis in land plants. Front Plant Sci  6: 288. Google Scholar CrossRef Search ADS PubMed  Navarro M, Marque G, Ayax C, Keller G, Borges JP, Marque C, Teulieres C ( 2009) Complementary regulation of four Eucalyptus CBF genes under various cold conditions. J Exp Bot  60: 2713– 2724. Google Scholar CrossRef Search ADS PubMed  Nguyen H, Cao PB, San Clemente H, Ployet R, Mounet F, Harvengt L, Ladouce N, Marque C, Teulieres C ( 2017) Special trends of CBF and DREB2 groups in Eucalyptus gunnii vs Eucalyptus grandis suggest that CBF are master players in the trade-off between growth and stress resistance. Physiol Plant  159: 445– 467. Google Scholar CrossRef Search ADS PubMed  Olenichenko NA, Zagoskina NV ( 2005) Response of winter wheat to cold: production of phenolic compounds and L-phenylalanine ammonia lyase activity. Appl Biochem Microbiol  41: 600– 603. Google Scholar CrossRef Search ADS   Pfaffl MW ( 2001) A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res  29: 2002– 2007. Google Scholar CrossRef Search ADS   Plomion C, Leprovost G, Stokes A ( 2001) Wood formation in trees. Plant Physiol  127: 1513– 1523. Google Scholar CrossRef Search ADS PubMed  Potts BM, Potts WC, Kantvilas G ( 2001) The miena cider gum, Eucalyptus Gunnii subsp. Divaricata (Myrtaceae): a taxon in rapid decline. Pap Proc Roy Soc Tas  135: 57– 61. Ranik M, Myburg AA ( 2006) Six new cellulose synthase genes from Eucalyptus are associated with primary and secondary cell wall biosynthesis. Tree Physiol  5: 545– 556. Google Scholar CrossRef Search ADS   Ruijter JM, Ramakers C, Hoogaars WMH, Karlen Y, Bakker O, Van Den Hoff MJB, Moorman AFM ( 2009) Amplification efficiency: linking baseline and bias in the analysis of quantitative PCR data. Nucleic Acids Res  37: e45. Google Scholar CrossRef Search ADS PubMed  Ruprecht C, Persson S ( 2012) Co-expression of cell wall-related genes: new tools and insights. Front Plant Sci  3: 1– 7. Google Scholar CrossRef Search ADS PubMed  Ruprecht C, Mutwil M, Saxe F, Eder M, Nikoloski Z, Persson S ( 2011) Large-scale co-expression approach to dissect secondary cell wall formation across plant species. Front Plant Sci  1: 2– 23. Soler M, Camargo ELO, Carocha V et al.  . ( 2015) The Eucalyptus grandis R2R3-MYB transcription factor family: evidence for woody growth-related evolution and function. New Phytol  206: 1364– 1377. Google Scholar CrossRef Search ADS PubMed  Soler M, Plasencia A, Lepikson-Neto J et al.  . ( 2016) The woody-preferential gene EgMYB88 regulates the biosynthesis of phenylpropanoid-derived compounds in wood. Front Plant Sci  7: 1422. Google Scholar CrossRef Search ADS PubMed  Southerton SG, Marshall H, Mouradov A, Teasdale RD ( 1998) Eucalypt MADS-box genes expressed in developing flowers. Plant Physiol  118: 365– 372. Google Scholar CrossRef Search ADS PubMed  Taylor-Teeples M, Lin L, de Lucas M et al.  . ( 2015) An Arabidopsis gene regulatory network for secondary cell wall synthesis. Nature  517: 571– 575. Google Scholar CrossRef Search ADS PubMed  Teulières C, Marque C, Nagata T, Lörz, Wildhom JM ( 2007) Eucalyptus. In: Pua EC, Davey MR (eds) Biotechnology in agriculture and forestry . Springer Verlag, Berlin, Germany, pp 387– 406. VanAcker R, Vanholme R, Storme V, Mortimer JC, Dupree P, Boerjan W ( 2013) Lignin biosynthesis perturbations affect secondary cell wall composition and saccharification yield in Arabidopsis thaliana. Biotechnol Biofuels  6: 46. Google Scholar CrossRef Search ADS PubMed  Vanholme R, Demedts B, Morreel K, Ralph J, Boerjan W ( 2010) Lignin biosynthesis and structure. Plant Physiol  153: 895– 905. Google Scholar CrossRef Search ADS PubMed  Vazquez-Cooz I, Meyer RW ( 2009) A differential staining method to identify lignified and unlignified tissues. Biotech Histochem  77: 277– 282. Google Scholar CrossRef Search ADS   Ventorim G, Alves AF, Penna LS, Francis RC ( 2014) Effect of S/G ratio on kraft pulping and ecf bleaching of some poplars and eucalyptus. Cellul Chem Technol  48: 365– 373. Wu AM, Rihouey C, Seveno M, Hörnblad E, Singh SK, Matsunaga T, Ishii T, Lerouge P, Marchant A ( 2009) The Arabidopsis IRX10 and IRX10-LIKE glycosyltransferases are critical for glucuronoxylan biosynthesis during secondary cell wall formation. Plant J  57: 718– 731. Google Scholar CrossRef Search ADS PubMed  Yao DX, Wei Q, Xu WY, Syrenne RD, Yuan JS, Su Z ( 2012) Comparative genomic analysis of NAC transcriptional factors to dissect the regulatory mechanisms for cell wall biosynthesis. BMC Bioinformatics  13: 12. Google Scholar CrossRef Search ADS PubMed  Zhao Q, Nakashima J, Chen F et al.  . ( 2013) Laccase is necessary and nonredundant with peroxidase for lignin polymerization during vascular development in Arabidopsis. Plant Cell  25: 3976– 3987. Google Scholar CrossRef Search ADS PubMed  Zhong R, Ye ZH ( 2015) Secondary cell walls: biosynthesis, patterned deposition and transcriptional regulation. Plant Cell Physiol  56: 195– 214. Google Scholar CrossRef Search ADS PubMed  Zhong R, Demura T, 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. Google Scholar CrossRef Search ADS PubMed  Author notes handling Editor Janice Cooke © The Author 2017. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Tree Physiology Oxford University Press

Long cold exposure induces transcriptional and biochemical remodelling of xylem secondary cell wall in Eucalyptus

Loading next page...
 
/lp/ou_press/long-cold-exposure-induces-transcriptional-and-biochemical-remodelling-4f0582WY2d
Publisher
Oxford University Press
Copyright
© The Author 2017. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com
ISSN
0829-318X
eISSN
1758-4469
D.O.I.
10.1093/treephys/tpx062
Publisher site
See Article on Publisher Site

Abstract

Abstract Although eucalypts are the most planted hardwood trees worldwide, the majority of them are frost sensitive. The recent creation of frost-tolerant hybrids such as Eucalyptus gundal plants (E. gunnii × E. dalrympleana hybrids), now enables the development of industrial plantations in northern countries. Our objective was to evaluate the impact of cold on the wood structure and composition of these hybrids, and on the biosynthetic and regulatory processes controlling their secondary cell-wall (SCW) formation. We used an integrated approach combining histology, biochemical characterization and transcriptomic profiling as well as gene co-expression analyses to investigate xylem tissues from Eucalyptus hybrids exposed to cold conditions. Chilling temperatures triggered the deposition of thicker and more lignified xylem cell walls as well as regulation at the transcriptional level of SCW genes. Most genes involved in lignin biosynthesis, except those specifically dedicated to syringyl unit biosynthesis, were up-regulated. The construction of a co-expression network enabled the identification of both known and potential new SCW transcription factors, induced by cold stress. These regulators at the crossroads between cold signalling and SCW formation are promising candidates for functional studies since they may contribute to the tolerance of E. gunnii × E. dalrympleana hybrids to cold. Introduction Eucalypts are the most planted hardwood trees worldwide because of their very rapid growth, exceptional wood quality and wide adaptability (Myburg et al. 2007). Two mains species are used in industrial plantations: the subtropical Eucalyptus grandis and the temperate Eucalyptus globulus, which are among the leading sources of wood biomass for pulpwood and timber, and are considered promising feedstocks for the production of second-generation biofuels. Although eucalypts are planted in >100 countries over six continents, the extension of industrial plantations to Northern hemisphere countries has been hampered by the fact that most commercial species and derived hybrids are very sensitive to frost (Booth 1990). This frost sensitivity is partially explained by the fact that eucalypts are evergreen trees that do not cope with frost by avoidance as many woody plant species from temperate countries do. The increasing demand for wood led the French ‘Institut Technologique FCBA’ to select fast-growing and frost-resistant hybrids between Eucalyptus gunnii, a species originating from the Central Plateau of Tasmania (Potts et al. 2001), and Eucalyptus dalrympleana, known under the generic name of Eucalyptus gundal (Navarro et al. 2009). The genes responsible for the high frost tolerance of E. gunnii have not yet been identified; however, a comparative study of the dehydration-responsive element binding (DREB) family of transcription factors (TFs) in E. grandis and E. gunnii suggested that additional copies and/or higher basal expression of CBF/DREB1 account, at least in part, for the ability of E. gunnii to adapt to cold conditions (Nguyen et al. 2017). Wood, or secondary xylem, is produced by the activity of an internal meristem, the vascular cambium, through a complex differentiation process leading to highly specialized xylem cells characterized by thick, lignified secondary cell walls (SCWs) (Plomion et al. 2001). Secondary cell walls are mainly composed of cellulose, hemicelluloses and lignin, in proportions of ~2:1:1, crosslinked into a supramolecular network, thereby creating a highly resistant barrier. The phenolic polymer lignin is mainly responsible for the recalcitrance of SCWs to degradation, which is a major obstacle for industrial end-uses of wood such as for pulp and paper manufacture or bioethanol production (VanAcker et al. 2013). In angiosperms, lignin is a complex tridimensional polymer containing mainly guaiacyl (G) and syringyl (S) units (Vanholme et al. 2010) that are connected by an array of inter-monomeric linkages, most commonly labile β-O-4 bonds. A high S/G ratio, which is associated with a high proportion of β-O-4 bonds, is, for instance, beneficial for lignin depolymerization during the kraft pulping process (Ventorim et al. 2014). The lignin content and composition of wood varies greatly within and between species, between developmental stages, and in response to environmental cues (Plomion et al. 2001). In eucalypts, for instance, the structure and content of lignin is affected by the nitrogen status of the plant (Camargo et al. 2014). The gene families encoding the 11 enzymes involved in biosynthesis of lignin monomeric units, or monolignols, have been studied at the genome level in E. grandis (Carocha et al. 2015). Lignin biosynthesis starts with the deamination of phenylalanine and leads to the production of hydroxycinnamoyl-CoA esters. This part of the pathway is called ‘general phenylpropanoid metabolism’ since hydroxycinnamoyl-CoA esters are also precursors of a wide range of end products including flavonoids, anthocyanins and condensed tannins, which vary according to species, cell types and in response to environmental signals (Dixon and Paiva 1995). Based on comparative phylogeny and patterns of expression, a total of 38 genes were identified as bona fide members of the 11 gene families, of which 25 are likely involved in developmental lignification belonging to the so-called ‘core vascular lignification toolbox’ (Carocha et al. 2015). In addition, Carocha et al. (2015) pointed out several members of the cinnamyl alcohol dehydrogenase (CAD), caffeoyl-CoA 3-O-methyltransferase (CCoAOMT) and caffeic acid O-methyltransferase (COMT) superfamilies called ‘CAD-like’, ‘CCoAOMT-like’ and ‘COMT-like’ as strongly expressed in xylem, suggesting possible roles in xylogenesis, although their functions are still unknown. Our knowledge of how SCW biosynthesis is regulated at the transcriptional level comes mainly from studies of the herbaceous plant Arabidopsis. In this model plant, a complex hierarchical network of TFs of the NAC and R2R3-MYB families act as first- and second-level master switches, respectively, to regulate a battery of downstream TFs (most of them also belonging to the MYB family) and SCW biosynthesis genes (Grima-Pettenati et al. 2012, Hussey et al. 2013, Nakano et al. 2015, Zhong and Ye 2015). Remarkably, a recent study demonstrated that abiotic stresses such as high salinity or iron deprivation can co-opt the xylem regulatory network in Arabidopsis, i.e., distinct stresses are able to modulate the expression of regulators of SCW biosynthesis to potentially promote functional adaptation of the plant to the stress conditions (Taylor-Teeples et al. 2015). In eucalypts, two master regulators of SCW deposition, EgMYB2 (Goicoechea et al. 2005) and EgMYB1 (Legay et al. 2007, 2010), have been characterized functionally. Potential orthologues of Arabidopsis SCW regulators in the E. grandis genome (Myburg et al. 2014) have also been identified by comprehensive genome-wide surveys of the R2R3-MYB and NAC TF families (Hussey et al. 2015, Soler et al. 2015). Importantly, members of subgroups present neither in Arabidopsis nor in rice were also identified, some of which are preferentially expressed in vascular cambium suggesting that they may be novel genes that regulate xylem differentiation (Soler et al. 2015). Exposure to chilling is known to increase plants’ ability to later withstand freezing temperatures. This process of cold acclimatization induces complex physiological and biochemical changes. Among them, increased cell-wall rigidity might lead to higher cell resistance to dehydration due to freezing (reviewed in LeGall et al. 2015). In some cases, lignin synthesis is enhanced during cold acclimatization and this may contribute to cell strengthening, thereby preventing cell damage and collapse (Domon et al. 2013). Most of these studies were performed on crop plants. By contrast, very few studies have investigated the effect of cold acclimatization on the SCWs of forest trees, despite the fact that cold is one of the major abiotic constraints affecting perennial plant development in many countries (Teulières et al. 2007). To the best of our knowledge, there is only one study of trees (Hausman et al. 2000), which reported an increase in the lignin content of shoots of young poplar trees exposed to 10 °C for 2 weeks. To evaluate the effects of low temperature on the SCW properties of Eucalyptus wood and to identify the metabolic and regulatory genes involved in SCW biosynthesis under cold stress, our study combined morphological analyses of xylem with biochemical characterization and transcriptomic profiling in young E. gundal hybrids submitted to chilling temperatures for up to 7 weeks and in adult trees grown in field conditions over three seasons in the southwest of France. Moreover, we performed gene co-expression network analyses on expression profiling data, which identified new potential regulators at the crossroad between SCW synthesis and response to cold stress. These regulators are promising candidates for future studies since they are potentially promoting functional adaptation to cold conditions. Materials and methods Plant material and cold treatment Eucalyptus gundal plants (E. gunnii × E. dalrympleana hybrids) obtained by in vitro propagation of cuttings were provided to us by the Institut Technologique FCBA (Pierroton, France). Six months after rooting, 36 plants were cultivated in a growth chamber with long day photoperiod (16 h light/8 h dark) at 25 °C/22 °C (light/dark) and 80% relative humidity. They were first exposed to a short acclimatization period of 3 days at 12 °C/8 °C (light/dark) before 46 days of cold treatment at 8 °C/4 °C (light/dark). Samples were collected before the cold treatment (0 day, control). Further samples were collected at 2 and 15 days after the start of cold treatment. Final samples were collected at the end of the experiment (46 days). Each sampling was performed on 6–10 independent plants. The lower part of the main stem of each plant was collected; 1 cm was kept in ethanol 80% (v/v) for histochemical analyses, the rest was frozen in liquid nitrogen, milled using a Mixer Mill MM 400 (Retsch, Haan, Germany) and the resulting powders were kept at −80 °C for later use. Samples from adult plants (>10 years old) were taken from the FCBA experimental plantation in southwest France (43°21′21.0′N 1°14′27.0′E). A set of 12 E. gundal (four independent pools of three individuals) were sampled over three seasons (see Figure S1 available as Supplementary Data at Tree Physiology online): autumn 2014 (mean temperatures 17 ± 6 °C), winter 2015 (mean temperatures 5 ± 5 °C) and summer 2015 (mean temperatures 22 ± 6 °C). After removing the bark at a height of ~150 cm, samples of xylem were collected by scraping using a single-edge, sharp blade and the samples were processed as above for the young plants. For wood structure observations, micro-core samples were collected by using a 5-mm diameter punch and were immediately fixed in 80% ethanol until use. RNA isolation Total RNA was extracted from the milled powder samples produced from the stem fragments of young plants and scraped xylem for adult trees as described by Southerton et al. (1998). Contaminating DNA was removed by using the TURBO DNA-free Kit (Ambion, Austin, TX, USA), according to the manufacturer’s instructions. RNA quantity and quality were checked by using a NanoDrop ND-1000 spectrophotometer (Thermo Scientific, Waltham, MA, USA) and 1% agarose gels. DNA contamination was evaluated by PCR using the housekeeping gene Tubulin (see Table S1 available as Supplementary Data at Tree Physiology Online). Quantitative RT-PCR First strand cDNAs were synthesized from 1 μg of total RNA using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystem, Foster City, CA, USA) according to the manufacturer’s instructions. The primer pairs were designed by using QuantPrime online software (http://www.quantprime.de/, Arvidsson et al. 2008) and are listed in Table S1 available as Supplementary Data at Tree Physiology online. Transcript abundance was assessed by using high-throughput microfluidic qPCR (96.96 Dynamic Array Fluidigm®, BioMark, Fluidigm Corporation, San Francisco, CA, USA) according to the manufacturer’s protocol and by conventional RT-qPCR using 1 μl of 1/10 diluted cDNA and Power SYBR® Green PCR Master Mix (Applied Biosystems), as described previously (Cassan-Wang et al. 2012). Expression profiles were obtained by the E−ΔΔCt method (Pfaffl 2001) using primers efficiency calculated through LinRegPCR (Ruijter et al. 2009). Expression was normalized against reference genes PP2A1, PP2A3, IDH, SAND and EF1a (Cassan-Wang et al. 2012). Biochemical analyses of SCWs Biochemical analyses of young trees were performed on bark-free stem samples, dried for 48 h at 60 °C and milled with a Mixer Mill MM 400 (Retsch). For adult trees, scraped xylem samples were ground in liquid nitrogen and freeze-dried for 48 h. Extractives content analyses and Analytical Pyrolysis (Py-GC/FID) applied to extractive-free xylem residues (EXRs) were performed as described in Lepikson-Neto et al. (2013). At least two runs were performed per sample. Pyrolysis products were identified as previously described (Alves et al. 2009). Lignin content was evaluated by the Klason method as described by Méchin et al. (2014). Lignin structure was analysed by thioacidolysis. For each sample, thioacidolysis was performed as described by Méchin et al. (2014), on 10 mg of EXR heated for 4 h and at 100 °C (oil bath) with 7 ml of a mix prepared with 10 ml of ethanethiol, 2.5 ml of BF3 etherate (Fisher Scientific, Waltham, MA, USA) and then adjusted the final volume of 100 ml with dioxane (Fisher Scientific). Lignin monomers were then extracted with 7 ml of dichloromethane (Fisher Scientific). Extracts were trimethylsilylated by incubating with 50 μl of N,O-bis(trimethylsilyl)trifluoroacetamide (Sigma, St. Louis, MO, USA) and 5 μl of ACS-grade pyridine (Sigma) for 30 min at 60 °C. These samples were then analysed in a Triple Quadrupole GC-MS system (Thermo Scientific) fitted with an autosampler, a splitless injector (280 °C), and an iontrap mass spectrometer operating in the electronic impact mode with a source at 220 °C, an interface at 280 °C and a 50–650 m/z scanning range. The column was a ZB5-MSi column (Phénomenex, Torrance, CA, USA) operated in the temperature program mode (from 45 to 180 °C at +30 °C min–1, then from 180 to 260 °C at +3 °C min−1), with helium carrier gas at a 1.5 ml min−1 flow rate. S and G peaks were identified using m/z (respectively, 399 and 369) and integrated using Xcalibur© software. Microscopy and histochemistry Tissues fixed in 80% ethanol were cut into thin sections (20 μm) by using a Vibratome VTS1000 (Leica, Wetzlar, Germany) and a sapphire vibratome knife (Delaware Diamond Knives, Wilmington, DE, USA). Lignin and cellulose were stained, respectively, with a mix of Safranin (0.1%) and Astra Blue (1%) according to the protocol of Vazquez-Cooz and Meyer (2009). Samples were also dehydrated in a EM CPD300 automated critical point dryer (Leica) using CO2 as a transitional fluid. When dry, the samples were coated with nickel (5 nm) in a EM MED020 vacuum coating system (Leica) and analysed in a scanning electron microscope (ESEM Quanta 250 FEG; FEI, Merignac, France) at 5 kV. In parallel, small fragments were embedded in LR White resin (London Resin Company Ltd, London, UK) by immersion in four consecutive baths (24 h) of increasing concentrations of resin (33, 50, 66 and 100%). Semi-thin sections (1 μm) were obtained by using a Histo Diamond Knife (DiATOME, Biel, Switzerland) on an ultramicrotome Reichert UltraCutE (Leica Microsystems, Rueil Malmaison, France). Sections were immediately mounted on glass slides and warmed to 60 °C for 5 min. For tissue morphology observation, sections were stained with 1% (w:v) aqueous Toluidine Blue O for 10 min at room temperature, rinsed in distilled water, dried at 60 °C and mounted under coverslips in mounting medium (Richard-Allan Scientific, San Diego, CA, USA). Semi-automatic image acquisition was performed using a Nanozoomer C9600-12 (Hamamatsu, Hamamatsu City, Japan) in bright field at ×40 magnification. Images were exported from raw data by using NDPview 2.3.1 and subsequently treated with ImageJ software (V1.5) to determine cell number, lumen and total cell-wall surface areas, thus allowing the calculation of mean cell density, average lumen diameter and average cell-wall thickness. Cells with diameters superior to 20 μm were considered as vessels. Average cell-wall thickness was determined on >15,000 cells in 7–20 stem sections for each tree. Co-expression network analyses Expression data from 146 genes related to the MYB family, NAC family and enzymes involved in SCW deposition were obtained from our previous studies (Carocha et al. 2015, Hussey et al. 2015, Soler et al. 2015). In these works, transcript levels were assessed for 22 parameters, including various eucalypts species, tissues, stages of development and growth conditions (see Table S2 available as Supplementary Data at Tree Physiology Online). Pearson pairwise correlation matrix was generated using ‘rcorr’ function of ‘Hmisc’ package with R software. The P-values obtained were adjusted using ‘p.adjust’ function with Bonferroni correction method. We eliminated correlations with an adjusted P-value >0.05 and calculated distances=(1−(correlation value)2)between the remaining genes. Distances were used to graphically represent gene correlation network using Cytoscape© software (force-directed layout). Results Analyses of xylem structure in plants submitted to low temperatures To evaluate the effects of prolonged cold treatment on xylem structure, we first compared thin sections of stems from young Eucalyptus hybrids grown for 7 weeks (46 days) at 4 °C with similar control sections harvested before cold exposure (i.e., grown at 25 °C). In the controls, the newly formed xylem cells immediately below the cambium formed a differentiating zone characterized by a gradual increase in cell size and cell-wall thickness (Figure 1a). The same observations were done in a control experiment performed with young trees grown during 46 days at 25 °C (see Figure S2a and b available as Supplementary Data at Tree Physiology Online). After 46 days of cold exposure, we observed an increase in SCW thickness in the xylem cells located within the first four to eight layers close to the cambium initials (Figure 1b). This was confirmed by using scanning electron microscopy (Figure 1c and d), which showed that in plants exposed to cold for 46 days, several layers of cells in the early differentiating xylem zone (close to the cambium) had thick SCWs comparable to those observed in the mature xylem zone. On the contrary, the early differentiating xylem cells of control plants had thin primary cell walls similar to those of cambium initials. It is worth noticing that, during cold exposure, young trees continue to undergo apical and radial growth. Therefore, the xylem cells next to cambium, which exhibit thicker and more lignified cell wall, are supposed to be newly generated (see Figure S2c–f available as Supplementary Data at Tree Physiology Online). Likewise, in adult Eucalyptus trees grown in forest plantations, the SCWs were thicker in autumn (average temperature 17 ± 6 °C) and in winter (average temperature 5 ± 5 °C) than in summer (average temperature 22 ± 6 °C) (Figure 1e–g and see Figure S1 available as Supplementary Data at Tree Physiology Online). We used ImageJ software to quantify the impact of low temperature on SCW thickness in xylem (Figure 1h and i) and found a significant increase in SCW thickness (average for fibres and vessels) in young trees exposed to cold for 46 days when compared with the controls (+0.5 μm), and between winter and summer in adult trees (+1.5 μm). We also measured fibre density (number of cells per mm2) and the diameter of the vessel lumens in the xylem tissues of young and adult trees (see Figure S3 available as Supplementary Data at Tree Physiology Online). For both, we observed a significant reduction of the fibre lumen diameter in cold conditions, in agreement with the increase of SCW thickness, but no significant difference in fibre cell density. Figure 1. View largeDownload slide Cold increases the thickness of xylem SCWs. (a and b) Bright field microscopy of transverse sections of xylem from the stems of young control plants grown at 25 °C (a) and similar plants exposed to cold for 46 days (b). (c and d) Scanning electron microscopy (×1000 magnification) of sections as in (a and b). Insets show ×4000 magnifications of xylem cells located immediately below the cambium. (e–g) Transverse sections of xylem from adult trees growing in field conditions. Xylem was collected from the main trunk in summer (e), autumn (f) and winter (g). (h and i) Histograms showing the SCW thickness (average of fibres and vessels) in xylem of young plants before and after 2, 15 and 46 days of cold (h) and in xylem of adult Eucalyptus trees in summer, autumn and winter (i). For each time point, >15,000 measurements were performed by using ImageJ software on 7–20 stem sections from independent trees. Significant differences compared to the control (0 day) for young trees and compared to samples collected in summer for adult trees are indicated by asterisks, where *P < 0.05, **P < 0.005 and ***P < 0.001 according to Student’s t-test. X, xylem; P, phloem; Ca, cambial zone. 0 d, control; 2 d, 2 days; 15 d, 15 days; 46 d, 46 days of cold treatment. S, summer; A, autumn; W, winter. Figure 1. View largeDownload slide Cold increases the thickness of xylem SCWs. (a and b) Bright field microscopy of transverse sections of xylem from the stems of young control plants grown at 25 °C (a) and similar plants exposed to cold for 46 days (b). (c and d) Scanning electron microscopy (×1000 magnification) of sections as in (a and b). Insets show ×4000 magnifications of xylem cells located immediately below the cambium. (e–g) Transverse sections of xylem from adult trees growing in field conditions. Xylem was collected from the main trunk in summer (e), autumn (f) and winter (g). (h and i) Histograms showing the SCW thickness (average of fibres and vessels) in xylem of young plants before and after 2, 15 and 46 days of cold (h) and in xylem of adult Eucalyptus trees in summer, autumn and winter (i). For each time point, >15,000 measurements were performed by using ImageJ software on 7–20 stem sections from independent trees. Significant differences compared to the control (0 day) for young trees and compared to samples collected in summer for adult trees are indicated by asterisks, where *P < 0.05, **P < 0.005 and ***P < 0.001 according to Student’s t-test. X, xylem; P, phloem; Ca, cambial zone. 0 d, control; 2 d, 2 days; 15 d, 15 days; 46 d, 46 days of cold treatment. S, summer; A, autumn; W, winter. Effects of cold on lignin content in differentiating xylem We used histochemical staining to characterize further the effects of cold on xylem cells (Figure 2). Safranin–Astra Blue stains blue those cell walls rich in cellulose, such as those of cambium initials, and stains red the lignified cell walls, such as those of mature xylem cells. In young control trees (grown at 25 °C), several layers of immature xylem cells in close proximity to cambium cells presented thin, weakly lignified SCWs stained both in red and blue (Figure 2a). After 46 days of cold exposure, xylem cells in the corresponding zone had thicker cell walls and more intense red staining, indicating higher lignin content (Figure 2b). Similar differences were observed between adult tree samples collected in summer (Figure 2c) and those collected in autumn (Figure 2d) and winter (Figure 2e), in which Safranin stained red the SCWs of xylem cells located immediately below the cambium. Figure 2. View largeDownload slide Cold enhances lignin deposition in xylem SCW. (a–e) Safranin–Astra Blue double staining of transverse sections from the stems of young plants (a, control; b, 46 days of cold exposure) and adult trees (c, summer; d, autumn; e, winter). Phenolic compounds are stained red and polysaccharides blue. X, xylem; Ca, cambium; P, phloem. (f and g) Klason lignin content in bark-stripped stems of young plants and in xylem scrapings of adult trees. Significant differences from the control (0 day) for young plants and from samples collected in summer for adult trees are indicated as in Figure 1 (n = 6–8). Abbreviations as in Figure 1. Figure 2. View largeDownload slide Cold enhances lignin deposition in xylem SCW. (a–e) Safranin–Astra Blue double staining of transverse sections from the stems of young plants (a, control; b, 46 days of cold exposure) and adult trees (c, summer; d, autumn; e, winter). Phenolic compounds are stained red and polysaccharides blue. X, xylem; Ca, cambium; P, phloem. (f and g) Klason lignin content in bark-stripped stems of young plants and in xylem scrapings of adult trees. Significant differences from the control (0 day) for young plants and from samples collected in summer for adult trees are indicated as in Figure 1 (n = 6–8). Abbreviations as in Figure 1. An increase in the insoluble lignin (called Klason lignin) content of 1.5% (Figure 2f) and 1.9% (Figure 2g) was found by analysis of the samples from young and adult eucalypts submitted to cold. In addition to this moderate but statistically significant increase in lignin content, which we confirmed by analytical pyrolysis (Table 1), the amount of ethanol-extractible compounds increased fivefold in xylem samples from cold-treated young plants when compared with controls (Table 1). Table 1. Analyses of SCW composition in xylem. Wood composition was assessed in young control plants (0 day), in similar plants after 2 days, 15 days or 46 days in the cold and in adult trees in summer, autumn and winter. Ethanol extractives, Klason acid-insoluble lignin and pyrolysis lignin content are expressed in % of cell-wall mass. S, G and S+G β-O-4 yields were obtained by thioacidolysis. They are expressed in μmoles per gram of cell wall (G and S yield) or μmoles per gram of Klason lignin (for S+G yield KL). Bold values represent significant differences (Student’s t-test, n = 6–8; a, P < 0.05; b, P < 0.01; c, P < 0.001) compared with controls (0 day, for young plants and Summer for adult plants); –, not analysed. Sample type  EtOH extractives  Lignin content  S/G ratio  Klason  Pyrolysis  G yield(β-O-4)  S yield(β-O-4)  S+G yield KL  Pyrolysis  Thioacidolysis  0 day  1.43 ± 0.74  19.3 ± 1.3  24.6 ± 1.1  167 ± 26  814 ± 152  5070 ± 1237  1.34 ± 0.1  4.9 ± 0.5  2 days  3.91 ± 0.65c  20.3 ± 0.3  24.6 ± 0.2  183 ± 31  850 ± 83  5091 ± 614  1.36 ± 0.04  4.7 ± 0.3  15 days  6.46 ± 0.63c  20 ± 0.8  23.9 ± 0.9  152 ± 31  742 ± 109  4471 ± 513  1.45 ± 0.02  5 ± 1  46 days  7.66 ± 1.73c  20.8 ± 1.3a  25.8 ± 1a  179 ± 39  859 ± 156  4997 ± 956  1.34 ± 0.06  4.8 ± 0.4  Summer  –  21.6 ± 0.97  –  93 ± 30  387 ± 120  2386 ± 610  –  4.11 ± 0.8  Autumn  –  23.6 ± 0.9c  –  136 ± 21c  492 ± 94a  2260 ± 641  –  3.38 ± 0.57b  Winter  –  23.5 ± 1.1c  –  127 ± 22c  461 ± 118  2158 ± 946  –  3.48 ± 0.6a  Sample type  EtOH extractives  Lignin content  S/G ratio  Klason  Pyrolysis  G yield(β-O-4)  S yield(β-O-4)  S+G yield KL  Pyrolysis  Thioacidolysis  0 day  1.43 ± 0.74  19.3 ± 1.3  24.6 ± 1.1  167 ± 26  814 ± 152  5070 ± 1237  1.34 ± 0.1  4.9 ± 0.5  2 days  3.91 ± 0.65c  20.3 ± 0.3  24.6 ± 0.2  183 ± 31  850 ± 83  5091 ± 614  1.36 ± 0.04  4.7 ± 0.3  15 days  6.46 ± 0.63c  20 ± 0.8  23.9 ± 0.9  152 ± 31  742 ± 109  4471 ± 513  1.45 ± 0.02  5 ± 1  46 days  7.66 ± 1.73c  20.8 ± 1.3a  25.8 ± 1a  179 ± 39  859 ± 156  4997 ± 956  1.34 ± 0.06  4.8 ± 0.4  Summer  –  21.6 ± 0.97  –  93 ± 30  387 ± 120  2386 ± 610  –  4.11 ± 0.8  Autumn  –  23.6 ± 0.9c  –  136 ± 21c  492 ± 94a  2260 ± 641  –  3.38 ± 0.57b  Winter  –  23.5 ± 1.1c  –  127 ± 22c  461 ± 118  2158 ± 946  –  3.48 ± 0.6a  We analysed the lignin structure by thioacidolysis, which specifically targets the G and S units involved in labile β-O-4 linkages (Méchin et al. 2014). Thioacidolysis yields expressed per gram of Klason lignin, which reflect the overall amounts of lignin monomers linked by β-O-4 bounds, were not affected by cold in both young plants and adult trees (Table 1). The S/G ratio was not significantly altered in young trees but was significantly reduced in adult trees: in autumn and winter, G units increased 46 and 37%, respectively, when compared with the summer level and S units were 27 and 19% higher, respectively, in autumn and winter than in summer. Finally, the cP/cH ratio, which indicates the relative proportion of hemicelluloses (pentosanes) compared with cellulose (hexosanes) (Lepikson-Neto et al. 2013), significantly increased in young plants after 46 days of cold treatment (from 25.7 ± 2.7 in 0 day samples to 29.6 ± 3.6 in 46 days samples, t-test, n = 10, P < 0.05; see Figure S4 available as Supplementary Data at Tree Physiology Online). This suggests that the deposition of hemicelluloses could be cold-induced in Eucalyptus as reported in annual plants (LeGall et al. 2015). Transcriptional regulation of genes related to SCW biosynthesis We performed RT-qPCR analyses on young Eucalyptus trees sampled at 2, 15 and 46 days after exposure to cold and compared with controls (0 days) (see Table S1 available as Supplementary Data at Tree Physiology at Tree Physiology Online and Figure 3). A twofold change of the transcript level ratio (between cold conditions and control) was defined as a threshold above which a gene was considered differentially expressed in response to cold. We first analysed the transcript level of a member of the DREB/CBF TF gene family, EgCBF14, a marker of cold stress known to be highly expressed in E. gunnii stems (Cao et al. 2015); EgCBF14 was induced 300-fold after 2 days of cold treatment and stayed highly expressed after 46 days when compared with controls (see Table S1 available as Supplementary Data at Tree Physiology Online). Figure 3. View large Download slide Cold induces genes of the lignin biosynthesis pathway in Eucalyptus. (a) The transcript levels of the genes encoding the 11-enzymatic steps involved in the monolignol biosynthetic pathway were analysed by RT-qPCR in response to cold. They include the 17 genes (marked in bold) identified as being involved in developmental lignification, belonging to the ‘core vascular lignin toolbox’ (Carocha et al. 2015). In addition, three close orthologues of Arabidopsis laccases 11, 4 and 17, were analysed. Because they represent a minor pathway in woody angiosperms, the enzymatic steps leading to hydroxyphenyl (H units) were not represented. Green and black arrows represent enzymatic steps induced or not induced by cold, respectively. The heat map illustrates the gene expression ratios in young plants exposed to cold for 2, 15 and 46 days compared with controls. The white to green colour scale represents log10 of expression ratio (2 , 15 or 46 days versus control). (b) The heat map represents, for each enzymatic step, the transcript abundance of each gene versus the total transcript production of the gene family expressed as percentage before and after cold treatment. The white to red colour scale represents 0 to 100% of transcript contribution. Expression values are provided in Table S1 available as Supplementary Data at Tree Physiology Online. Genes short names were given according to Carocha et al. (2015): PAL, phenylalanine ammonialyase; C4H, cinnamate-4-hydroxylase; 4CL, 4-hydroxycinnamate CoA ligase; HCT, hydroxycinnamoyl transferase; C3H, p-coumarate 3-hydroxylase; CCR, cinnamoyl CoA reductase; CCoAOMT, caffeoyl-CoA 3-O-methyltransferase; F5H, ferulate 5-hydroxylase; COMT, caffeic acid O-methyltransferase; CAD, cinnamyl alcohol dehydrogenase. Figure 3. View large Download slide Cold induces genes of the lignin biosynthesis pathway in Eucalyptus. (a) The transcript levels of the genes encoding the 11-enzymatic steps involved in the monolignol biosynthetic pathway were analysed by RT-qPCR in response to cold. They include the 17 genes (marked in bold) identified as being involved in developmental lignification, belonging to the ‘core vascular lignin toolbox’ (Carocha et al. 2015). In addition, three close orthologues of Arabidopsis laccases 11, 4 and 17, were analysed. Because they represent a minor pathway in woody angiosperms, the enzymatic steps leading to hydroxyphenyl (H units) were not represented. Green and black arrows represent enzymatic steps induced or not induced by cold, respectively. The heat map illustrates the gene expression ratios in young plants exposed to cold for 2, 15 and 46 days compared with controls. The white to green colour scale represents log10 of expression ratio (2 , 15 or 46 days versus control). (b) The heat map represents, for each enzymatic step, the transcript abundance of each gene versus the total transcript production of the gene family expressed as percentage before and after cold treatment. The white to red colour scale represents 0 to 100% of transcript contribution. Expression values are provided in Table S1 available as Supplementary Data at Tree Physiology Online. Genes short names were given according to Carocha et al. (2015): PAL, phenylalanine ammonialyase; C4H, cinnamate-4-hydroxylase; 4CL, 4-hydroxycinnamate CoA ligase; HCT, hydroxycinnamoyl transferase; C3H, p-coumarate 3-hydroxylase; CCR, cinnamoyl CoA reductase; CCoAOMT, caffeoyl-CoA 3-O-methyltransferase; F5H, ferulate 5-hydroxylase; COMT, caffeic acid O-methyltransferase; CAD, cinnamyl alcohol dehydrogenase. We then analysed the effect of cold on the transcript levels of genes involved in the biosynthesis of SCW polymers, i.e., lignin, cellulose and hemicelluloses (see Table S1 available as Supplementary Data at Tree Physiology Online). Among the 11 phenylpropanoid gene families involved in the monolignol biosynthesis pathway, the vast majority of the members identified as being recruited for lignin biosynthesis during development belonging to the so-called ‘core vascular lignification toolbox’ (Carocha et al. 2015) were significantly induced by cold, although moderately (marked in bold in see Table S1 available as Supplementary Data at Tree Physiology Online and Figure 3a). Of these genes, the two most highly induced were the two members of the C4H family, EgC4H1 (24.5-fold induction) and EgC4H2 (7.5-fold induction); the latter was already known to be inducible by abiotic stress responses in eucalypts (Carocha et al. 2015). It is worth noting that EgF5H1 and EgCOMT1, which encode enzymes catalysing the two key steps in S unit biosynthesis, were not significantly induced in response to cold (see Table S1 available as Supplementary Data at Tree Physiology Online and Figure 3a). The levels of induction of the other family members in response to cold were in general much higher than those of the ‘core vascular lignification toolbox’ genes. This was mainly due to their lowest expression in xylem in control conditions (see Table S1 available as Supplementary Data at Tree Physiology Online and Figure 3a). For instance, in the PAL family, EgPAL1 thought to be involved in anthocyanin production was induced 10-fold and EgPAL8, whose role is unknown, was induced up to 330-fold (see Table S1 available as Supplementary Data at Tree Physiology Online). In the COMT family, except EgCOMT1, all the genes were induced from 12-fold up to 618-fold (for EgCOMT6) (see Table S1 available as Supplementary Data at Tree Physiology Online). For each enzymatic step, we compared before and after cold treatment, the transcript abundance of each gene to the total transcript production of the family expressed as percentage (Figure 3b). This comparison showed that in most cases, main transcript contributions were still attributed to the ‘core vascular lignification toolbox’ genes. The enzymes encoded by the latter are likely keeping prominent roles in lignin biosynthesis in response to cold temperatures. However, genes encoding EgPAL8, EgHCT2 and EgCOMT6, which are not part of this ‘lignification toolbox’, were dramatically induced in response to cold and reached expression levels similar to the prominent ‘lignification toolbox’ genes EgPAL3, EgHCT4 and EgCOMT1, respectively. Two Eucalyptus laccases orthologues of AtLac11 and AtLac4 (involved the lignin polymerization step in Arabidopsis (Zhao et al. 2013) were induced by cold ninefold and threefold, respectively (see Table S1 available as Supplementary Data at Tree Physiology Online)). Considering polysaccharides biosynthesis, two of the four cellulose synthases genes analysed (EgCesA2 and EgCesA6) (Ranik and Myburg 2006) were induced fivefold and twofold, respectively, whereas only one of the eight genes involved in hemicelluloses biosynthesis—the closest Eucalyptus orthologue of AtIRX10 (GUT2, a member of GT47 family, Wu et al. 2009)—was induced more than fourfold by cold (see Table S1 available as Supplementary Data at Tree Physiology Online). Xylem SCW transcriptional regulators induced by cold A recent study in Arabidopsis demonstrated that abiotic stresses (high salinity and iron deprivation) can induce genes that regulate biosynthesis of the SCW (Taylor-Teeples et al. 2015). To investigate whether the response to cold in Eucalyptus might also induce genes that regulate SCW biosynthesis, we selected 25 close orthologues of the Arabidopsis TFs involved in the three-levels hierarchical SCW regulatory network (Figure 4), building on the works of Hussey et al. (2013) and Soler et al. (2015) and we analysed their transcript levels in young trees exposed to cold (Figure 4). Fourteen of the 25 genes were differentially expressed following cold exposure: seven were up-regulated and seven were down-regulated (Figure 4 and see Table S1 available as Supplementary Data at Tree Physiology Online). Four genes presented a non-significant twofold induction or repression. The differentially regulated genes were found mainly in the lower third level of the SCW regulatory network (Level 3). These TFs act directly on the structural genes involved in the biosynthesis of the SCW polymers. Six of the up-regulated genes are positive regulators of SCW deposition (i.e., orthologues of AtMYB20, AtMYB54, AtMYB58, AtMYB63, AtMYB69 and AtMYB79), whereas three of the down-regulated genes are negative regulators of SCW deposition (i.e., orthologues of AtMYB7, AtMYB75 and AtKNAT7). Only 4 of the 14 genes significantly affected by cold are involved in the upper part of the regulatory network (Levels 1 and 2). Of these, three were repressed (orthologues of AtMYB83, AtMYB55 and AtSND1) and one was induced (one orthologue of AtSND3). Figure 4. View large Download slide Secondary cell wall-related MYB and NAC TFs are regulated by cold to promote SCW deposition. (a) The 25 genes represented are the closest eucalypts orthologues of Arabidopsis NACs and MYBs TFs involved in the control of SCW deposition (Hussey et al. 2013). For each gene, heat maps represent expression ratios after 2, 15 and 46 days of cold exposure normalized by control (0 day). The red to green colour scale represents log2 of expression ratios (available in Table S1 available as Supplementary Data at Tree Physiology Online). Gene names in green indicate those induced more than twofold by cold and those in red indicate those repressed by more than twofold. The genes marked with asterisks indicate those whose transcription was significantly up- or down-regulated at two or more time points (15 d, 15 days; n = 6, P < 0.05, according to Student’s t-test). (b) The genes are grouped into the three-level hierarchical network of TFs that regulates xylem cell differentiation and deposition of lignin, cellulose and hemicelluloses in SCWs of Arabidopsis (adapted from Hussey et al. 2013). The putative orthologues of Arabidopsis NAC and MYB genes in eucalypts are indicated in brackets below their Arabidopsis counterparts. Arrows indicate positive regulation and the blocked line indicates negative regulation. Figure 4. View large Download slide Secondary cell wall-related MYB and NAC TFs are regulated by cold to promote SCW deposition. (a) The 25 genes represented are the closest eucalypts orthologues of Arabidopsis NACs and MYBs TFs involved in the control of SCW deposition (Hussey et al. 2013). For each gene, heat maps represent expression ratios after 2, 15 and 46 days of cold exposure normalized by control (0 day). The red to green colour scale represents log2 of expression ratios (available in Table S1 available as Supplementary Data at Tree Physiology Online). Gene names in green indicate those induced more than twofold by cold and those in red indicate those repressed by more than twofold. The genes marked with asterisks indicate those whose transcription was significantly up- or down-regulated at two or more time points (15 d, 15 days; n = 6, P < 0.05, according to Student’s t-test). (b) The genes are grouped into the three-level hierarchical network of TFs that regulates xylem cell differentiation and deposition of lignin, cellulose and hemicelluloses in SCWs of Arabidopsis (adapted from Hussey et al. 2013). The putative orthologues of Arabidopsis NAC and MYB genes in eucalypts are indicated in brackets below their Arabidopsis counterparts. Arrows indicate positive regulation and the blocked line indicates negative regulation. In addition to the orthologues of known regulators of SCW in Arabidopsis, potential new SCW regulators have been pointed out by the works of Hussey et al. (2015) and Soler et al. (2015). To identify the most promising candidate regulators among these eucalypt genes and to evaluate their responses to cold, we used co-expression analyses that have been shown to be useful to identify new genes involved in SCW-related synthesis (Ruprecht and Persson 2012). We first built a gene co-expression network based on transcript profiling of 146 candidate SCW-related genes in a large panel of eucalypt organs and tissues, stages of development and/or environmental conditions previously described in Cassan-Wang et al. (2012), Camargo et al. (2014), Hussey et al. (2015), Soler et al. (2015) and Carocha et al. (2015). This set of genes comprised lignin and cellulose biosynthesis genes, as well as known and as yet uncharacterized NAC and MYB TFs. Pearson’s correlation coefficient is commonly used to estimate transcriptional coordination between two genes, suggesting their involvement in the same biological process (Aoki et al. 2007). We thus generated a Pearson correlation matrix to build a co-expression network in which 132 of the 146 genes (nodes) were connected by 592 edges representing significant co-expression relationships (see Table S2 available as Supplementary Data at Tree Physiology Online). These correlations were represented graphically in a network regrouping 93 of the 132 genes (Figure 5). The remaining 39 genes were grouped in smaller networks (see Figure S5 available as Supplementary Data at Tree Physiology Online). Because we focused our interest in woody tissues, we set up node sizes to be proportional to the expression ratio between woody and non-woody tissues. Edge lengths are inversely proportional to the absolute value of correlation for each gene pair comparison, so that genes in close vicinity are considered as strongly correlated. This approach led to three clusters of genes, which were clearly separated and in which almost all the significant correlations detected were positive (Figure 5). We used a hierarchical clustering approach to investigate the tissue preferential expression of the genes contained in each of these clusters (see Figure S6 available as Supplementary Data at Tree Physiology Online). Based on this, we called the upper cluster ‘xylem cluster’ because it contained 47 genes preferentially expressed in xylem and related to SCW deposition, i.e., lignin toolbox genes, cellulose synthases and orthologues of known TFs that regulate SCW biosynthesis (Figure 5). Interestingly, it also contained uncharacterized TFs. The intermediate or ‘cambium cluster’ contained 17 genes expressed preferentially in vascular cambium, mainly composed of TFs of unknown function (with no orthologues in Arabidopsis). The lower ‘non-woody tissues’ cluster contained 29 genes; of these, 11 encoded enzymes of unknown function (COMT_like, CAD_like, HCT_like) and 18 were unknown TFs. We chose a subset of 23 uncharacterized TFs, contained both in the ‘xylem’ and the ‘cambium’ clusters, to investigate their responses to cold temperatures by measuring their transcript levels in cold acclimated young plants (squares, Figure 5). Eleven of these genes were up-regulated by cold and three were down-regulated. Those of the ‘xylem cluster’ were being either up- or down-regulated by cold from twofold to eightfold. The 10 TFs of the ‘cambium cluster’ regulated by cold were all induced, some being dramatically up-regulated such as EgMYB64 (>5000-fold) and EgMYB59 (>100-fold). Figure 5. View large Download slide Correlation network analyses highlight xylem and cambium clusters enriched in cold responsive TFs. Transcript profiling of 146 genes encoding members of the MYB, NAC families of TFs as well as enzymes involved in SCW biosynthesis in a panel of 22 conditions were obtained from previous studies in eucalypts (see Table S2 available as Supplementary Data at Tree Physiology Online) (Cassan-Wang et al. 2012, Carocha et al. 2015, Hussey et al. 2015, Soler et al. 2015). Significant pairwise Pearson correlation values (Bonferroni adjusted, P < 0.05) were used to draw a co-expression network. A short link between two genes represents a strong correlation, explained by similar expression profiles in the 22 conditions analysed. Grey and red edges represent respectively positive and negative correlations. Gene names and accession numbers are listed in Table S2 available as Supplementary Data at Tree Physiology Online. Node sizes are proportional to the expression ratio between woody and non-woody tissues (see Table S2 available as Supplementary Data at Tree Physiology Online). Circles represent genes related to SCW formation (enzymes or TFs), squares represent new, uncharacterized TFs. The main network shown here is composed of 93 genes (nodes) linked by 520 relationships (edges); small isolated sub-networks regrouping 39 genes are provided in Figure S5 available as Supplementary Data at Tree Physiology Online. The three clusters of the main network were named according to the tissue preferential expression of their gene members (‘xylem’, ‘cambium’ or ‘non-woody tissues’; see Figure S6 available as Supplementary Data at Tree Physiology Online). Transcript levels were measured in young eucalypts plants submitted to cold (2, 15 and 46 days at 4 °C) and compared with control (25 °C, 0 days). Node colour gradient (red to green) represents gene expression ratios between cold and control conditions (see Table S1 available as Supplementary Data at Tree Physiology Online; ratio = Coldmean(15 days/46 days)/Control(0 days)). Green and red colours represent gene induction or repression in response to cold, respectively. Grey nodes represent genes with no expression data. Figure 5. View large Download slide Correlation network analyses highlight xylem and cambium clusters enriched in cold responsive TFs. Transcript profiling of 146 genes encoding members of the MYB, NAC families of TFs as well as enzymes involved in SCW biosynthesis in a panel of 22 conditions were obtained from previous studies in eucalypts (see Table S2 available as Supplementary Data at Tree Physiology Online) (Cassan-Wang et al. 2012, Carocha et al. 2015, Hussey et al. 2015, Soler et al. 2015). Significant pairwise Pearson correlation values (Bonferroni adjusted, P < 0.05) were used to draw a co-expression network. A short link between two genes represents a strong correlation, explained by similar expression profiles in the 22 conditions analysed. Grey and red edges represent respectively positive and negative correlations. Gene names and accession numbers are listed in Table S2 available as Supplementary Data at Tree Physiology Online. Node sizes are proportional to the expression ratio between woody and non-woody tissues (see Table S2 available as Supplementary Data at Tree Physiology Online). Circles represent genes related to SCW formation (enzymes or TFs), squares represent new, uncharacterized TFs. The main network shown here is composed of 93 genes (nodes) linked by 520 relationships (edges); small isolated sub-networks regrouping 39 genes are provided in Figure S5 available as Supplementary Data at Tree Physiology Online. The three clusters of the main network were named according to the tissue preferential expression of their gene members (‘xylem’, ‘cambium’ or ‘non-woody tissues’; see Figure S6 available as Supplementary Data at Tree Physiology Online). Transcript levels were measured in young eucalypts plants submitted to cold (2, 15 and 46 days at 4 °C) and compared with control (25 °C, 0 days). Node colour gradient (red to green) represents gene expression ratios between cold and control conditions (see Table S1 available as Supplementary Data at Tree Physiology Online; ratio = Coldmean(15 days/46 days)/Control(0 days)). Green and red colours represent gene induction or repression in response to cold, respectively. Grey nodes represent genes with no expression data. One way to verify the accuracy of the co-expression clusters is to investigate whether orthologous genes are also co-expressed in other species (Ruprecht and Persson 2012). We extracted, for instance, the subset of genes co-expressed with the Eucalyptus orthologue (EgNAC61) of the fist level master regulator AtSND1 (Figure 6a). The EgNAC61 node-vicinity network included second and third levels TFs of the hierarchical SCW regulating network (see Figure 5) as well as lignin and cellulose biosynthesis genes (Figure 6a). The Arabidopsis orthologues of all these genes are also co-expressed with AtSND1, which regulates their transcription levels (Ruprecht et al. 2011, Yao et al. 2012) thereby strengthening the validity of the co-expression network. Interestingly, the EgNAC61 node-vicinity network also contained new, as yet uncharacterized TFs, such as EgNAC64 and EgMYB137, which have no orthologues in Arabidopsis, and which were both significantly induced by cold (see Table S1 available as Supplementary Data at Tree Physiology Online). The co-expression network of EgNAC64 is composed of 19 genes mostly SCW-related genes, more than that of EgNAC61 (AtSND1) (Figure 6b). Remarkably, it contains EgNAC61 itself, suggesting that EgNAC64 could also be a SCW master regulator. This led us to investigate more precisely the phylogenetic relationships between Arabidopsis and Eucalyptus NAC sequences (Hussey et al. 2015). We found that EgNAC64 belongs to the NAC subgroup II and could be a co-orthologue of AtSND3, a second-level SCW master regulator in Arabidopsis whose closest co-orthologues are four Eucalyptus genes duplicated in tandem. We then extracted the co-expression network of EgMYB137, which has no close orthologues in Arabidopsis (Soler et al. 2015). It contains nine genes of which five are SCW regulators, and one is a lignin biosynthesis gene CCoAOMT1 supporting a possible involvement in SCW regulation in response to stress (Figure 6c). It also includes two as yet uncharacterized TFs. Finally, we extracted the sub-network of EgCOMT6, (Figure 6d), the most strongly induced COMT gene in response to cold. EgCOMT6 which belongs to the cambium cluster, presents a high degree of connectivity with nine uncharacterized MYB TFs, which form the core of a module enriched in stress-induced woody-specific genes preferentially expressed in cambium. Figure 6. View largeDownload slide Xylem and cambium sub-networks reveal promising new candidate genes. Sub-networks were generated by extracting the genes directly correlated to (a) EgNAC61 (AtSND1), (b) EgNAC64, (c) EgMYB137 or (d) EgCOMT6. For legend, see Figure 5. Yellow nodes represent the ‘hub gene’ correlated to all the others. Figure 6. View largeDownload slide Xylem and cambium sub-networks reveal promising new candidate genes. Sub-networks were generated by extracting the genes directly correlated to (a) EgNAC61 (AtSND1), (b) EgNAC64, (c) EgMYB137 or (d) EgCOMT6. For legend, see Figure 5. Yellow nodes represent the ‘hub gene’ correlated to all the others. Discussion Cell-wall adjustment under abiotic stress is an important process in plant adaptation to cope with non-optimal growth conditions (LeGall et al. 2015). Cell wall changes induced by cold have been studied extensively in crop plants and mostly in their leaves, but they have been only rarely investigated in trees and never in non-dormant trees such as eucalypts. Here, we report that long exposure to cold increases SCW thickening in the developing xylem cells of the cold-resistant hybrid E. gundal. This was observed in young trees and to an even greater extent in adult trees. These thicker SCWs were enriched in Safranin-stained, mostly phenolic compounds. On one hand, both Klason lignin analysis and analytical pyrolysis suggest that at least part of these phenolic compounds were incorporated into lignin polymers, resulting in more lignin in the SCWs of plants submitted to cold conditions. On the other hand, the observed fivefold increase in ethanol-extractible compounds suggests that part of these phenolics stayed soluble or loosely linked to the cell walls. As suggested previously for Miscanthus and wheat (Olenichenko and Zagoskina 2005, Domon et al. 2013), these compounds might increase cell-wall rigidity by crosslinking with cell-wall polymers and/or might contribute to xylem cell protection from reactive oxygen species. The greater lignin content in the xylem SCWs of cold-treated eucalypts is in good agreement with the up-regulation of most genes belonging to the ‘core vascular lignification toolbox’. In young trees, all lignin genes whose products are involved in G unit biosynthesis were induced by cold, whereas genes dedicated to S unit biosynthesis (encoding EgF5H1 and EgCOMT1, Anterola and Lewis 2002) were not differentially expressed. This differential expression of genes involved in G and S unit biosynthesis in response to cold, which might result in relatively more G units than S units, may explain the significant decrease in the S/G ratio observed in adult trees in autumn and winter compared with that in summer. In general, the genes most highly induced by cold were those not involved in developmental lignification. These stress-inducible genes, which are often poorly expressed in xylem in control conditions, encode isoforms (EgPAL8, EgHCT1 and EgHCT2, EgCCR2, EgCOMT2–EgCOMT6), whose activities might lead to the increased soluble phenolics content and/or to the increased lignin content. Of the SCW polysaccharides, we observed a higher ratio of pentosanes (cP) to hexosanes (cH) in xylem samples from young plants exposed to cold than in control plants, suggesting that more hemicelluloses than cellulose might be incorporated into the SCWs in cold conditions. However, at the transcript gene level, we observed very significant inductions by cold of both a SCW-specific cellulose synthase gene EgCesA2 (Eucgr.A01324; Myburg et al. 2014), and of EgIRX10, which encodes a GT47 glycosyl transferase involved in chain elongation of the most abundant hemicelluloses, i.e., xylan. In Arabidopsis, a mutant deficient for IRX10 and IRX10_like proteins showed a dramatic loss of SCW deposition associated with reduced xylan backbone polymerization and a complete loss of glucuronic acid side chains (Brown et al. 2009). Based on these results, we hypothesize that induction of EgIRX10 in Eucalyptus xylem in response to cold increases SCW deposition and may account, at least in part, for the observed increased SCW thickness in plants exposed to cold. In summary, we observed increased SCW thickness likely resulting from enhanced deposition of lignin and possibly of xylan in E. gundal developing xylem cells submitted to cold conditions. An increased thickness of primary walls has been frequently reported in crop plants in response to various abiotic stresses and is considered as one of the main common mechanisms of stress responses in cell-wall architecture that could enable functional adaptation to abiotic stress. Increased lignin and/or hemicelluloses content is proposed to modulate cell-wall rigidity to prevent cell damage due to freezing (LeGall et al. 2015). Our findings are consistent with a model in which reinforcement of SCWs could be one mechanism by which perennial plants cope with unfavourable conditions such as cold and might help them adapt to frost. The content and composition of lignin polymers in SCWs seem to be modulated by transcriptional regulation of genes encoding enzymes involved in developmental lignification and of stress-responsive genes such as EgCOMT6. We cannot exclude, however, that such genes could be responsible for the synthesis of other phenylpropanoid compounds in response to cold. EgCOMT6, for instance is tightly co-expressed with cambium woody-specific MYB TFs whose role in SCW regulation based on co-expression networks is not obvious. Recently, one of these TFs, EgMYB88, has been functionally characterized in poplar and shown to control the biosynthesis of phenylpropanoid-derived secondary metabolites including lignin (Soler et al. 2016). The E. grandis orthologues of Arabidopsis TFs belonging to the three-levels SCW hierarchical network established for the model plant were, for a large majority, positioned in the ‘xylem co-expresssion cluster’. Cold temperatures triggered a clear regulation pattern for the most downstream TFs, downregulation of the negative regulators and induction of the positive regulators, in good agreement with the increase of SCW thickness and lignin content observed in response to cold. The ‘xylem cluster’ also included the ‘core vascular lignin toolbox’ genes and SCW-specific cellulose synthases. Co-expression between transcriptional regulators and effector genes ensuring SCW deposition has already been described and shown to be consistent across species (Ruprecht and Persson 2012). Indeed, the EgNAC61 (AtSND1) sub-network included the closest Arabidopsis orthologues of TFs controlling SCW deposition, as well as cellulose and lignin related genes, known to be co-expressed and transcriptionally regulated by AtSND1 (Zhong et al. 2006, Yao et al. 2012), and is, therefore, highly conserved between Eucalyptus and Arabidopsis. The analysis of the EgNAC64 network highlighted a hub position of this TF and suggested that it could be a functional orthologue of AtSND3 given its phylogenetic proximity even if it is not the closest Eucalyptus orthologue. EgMYB137 is a new uncharacterized TF, co-expressed with EgNAC61 (AtSND1) and other orthologues of Arabidopsis SCW-master regulators, like EgMYB2 (AtMYB46) or EgNAC49 (AtNST1/2). EgNAC64 and EgMYB137 are induced by cold stress and might lead, at least in part, to the phenylpropanoid/lignin increase and the S/G ratio decrease triggered by chilling temperatures. Notably, the use of a woody plant allowed a physical separation of differentiating xylem from vascular cambium (Cassan-Wang et al. 2012) and thus enabled to highlight in our co-expression analysis a ‘cambium co-expression cluster’ that obviously cannot be characterized in Arabidopsis. This cluster contains mostly yet uncharacterized MYB TFs, belonging to woody-preferential clades, not present in Arabidopsis and many of them are cold-induced. Twelve of them are also correlated to stress-inducible genes of the phenylpropanoid pathway, like EgCOMT6 or EgPAL8 and could be involved in lignin deposition or soluble phenolics biosynthesis. In conclusion, this work provides a strong foundation to decipher the function of new TF candidates at the crosstalk between cold stress and wood tissue formation. This is particularly relevant in Eucalyptus, which, as a woody evergreen perennial tree, must ensure its fitness and adaptability to challenging environmental cues. This study also provides a global view of the consequences of cold temperatures on wood formation in Eucalyptus. The observed increase in the amount of lignin and the modified lignin composition (lower S/G ratio) may contribute to greater tolerance of cold in E. gunnii × E. dalrympleana hybrids, as in crop plants, but these SCW modifications may also greatly affect the properties of the wood and consequently its industrial uses. In this respect, the identification of regulators that act both in the developmental patterning of wood tissues and in the abiotic stress response will be extremely useful for both wood engineering and breeding strategies as stressed by Mashkina and Butorina (2003) and Harfouche et al. (2011). Supplementary Data Supplementary Data for this article are available at Tree Physiology Online. Acknowledgments The authors acknowledge F. Melun (FCBA) for his help with the identification of field-grown eucalypt clones in Longages (France), R. Simões (CEF) for the extraction of xylem samples, B. Savelli and H. San Clemente (LRSV) for their help with bioinformatics, A. Gauthier and A. Desplat for their technical help during their Master internships, the Genotoul Bioinformatics Platform Toulouse Midi-Pyrenees for computing and storage resources and the TRI-Genotoul platform for microscopic analyses. The authors are grateful to Prof. C. Lapierre (Institut JP Bourgin, France) and Dr Y. Barrière (INRA Lusignan, France) for helpful discussions and advice on lignin analyses. The authors acknowledge Centre National pour la Recherche Scientifique (CNRS), the University Paul Sabatier Toulouse III, the Agence Nationale de la Recherche (ANR), the Fundação para a Ciência e a Tecnologia (FCT) and the Ministère de l'Education Nationale, de l'Enseignement Supérieur et de la Recherche for the support of this work. Funding This work was supported by the Centre National pour la Recherche Scientifique (CNRS), the University Paul Sabatier Toulouse III (UPS, IDEX UNITI, project EuCoWood), the TREEFORJOULES project (ANR-2010-KBBE-007-01 and FCT, P-KBBE/AGR_GPL/0001/2010) the French Laboratory of Excellence project ‘TULIP’ (ANR-10-LABX-41; ANR-11-IDEX-0002-02). R.P. was supported by a PhD grant from the Ministère de l’Education Nationale, de l’Enseignement Supérieur et de la Recherche. References Alves A, Gierlinger N, Schwanninger M, Rodrigues J ( 2009) Analytical pyrolysis as a direct method to determine the lignin content in wood. Part 3. Evaluation of species-specific and tissue-specific differences in softwood lignin composition using principal component analysis. J Anal Appl Pyrolysis  85: 30– 37. Google Scholar CrossRef Search ADS   Anterola AM, Lewis NG ( 2002) Trends in lignin modification: a comprehensive analysis of the effects of genetic manipulations/mutations on lignification and vascular integrity. Phytochemistry  61: 221– 294. Google Scholar CrossRef Search ADS PubMed  Aoki K, Ogata Y, Shibata D ( 2007) Approaches for extracting practical information from gene co-expression networks in plant biology. Plant Cell Physiol  48: 381– 390. Google Scholar CrossRef Search ADS PubMed  Arvidsson S, Kwasniewski M, Riaño-Pachón DM, Mueller-Roeber B ( 2008) QuantPrime—a flexible tool for reliable high-throughput primer design for quantitative PCR. BMC Bioinformatics  9: 465. Google Scholar CrossRef Search ADS PubMed  Booth TH ( 1990) Mapping regions climatically suitable for particular tree species at the global scale. For Ecol Manage  36: 47– 60. Google Scholar CrossRef Search ADS   Brown DM, Zhang Z, Stephens E, Dupree P, Turner SR ( 2009) Characterization of IRX10 and IRX10-like reveals an essential role in glucuronoxylan biosynthesis in Arabidopsis. Plant J  57: 732– 746. Google Scholar CrossRef Search ADS PubMed  Camargo ELO, Nascimento LC, Soler M et al.  . ( 2014) Contrasting nitrogen fertilization treatments impact xylem gene expression and secondary cell wall lignification in Eucalyptus. BMC Plant Biol  14: 256. Google Scholar CrossRef Search ADS PubMed  Cao PB, Azar S, SanClemente H, Mounet F, Dunand C, Marque G, Marque C, Teulières C ( 2015) Genome-wide analysis of the AP2/ERF family in Eucalyptus grandis: an intriguing over-representation of stress-responsive. PLoS One  10: 1– 22. Carocha V, Soler M, Hefer C, Cassan-Wang H, Fevereiro P, Myburg AA, Paiva JAP, Grima-Pettenati J ( 2015) Genome-wide analysis of the lignin toolbox of Eucalyptus grandis. New Phytol  206: 1297– 1313. Google Scholar CrossRef Search ADS PubMed  Cassan-Wang H, Soler M, Yu H et al.  . ( 2012) Reference genes for high-throughput quantitative reverse transcription-PCR analysis of gene expression in organs and tissues of Eucalyptus grown in various environmental conditions. Plant Cell Physiol  53: 2101– 2116. Google Scholar CrossRef Search ADS PubMed  Dixon RA, Paiva NL ( 1995) Stress-induced phenylpropanoid metabolism. Plant Cell  7: 1085– 1097. Google Scholar CrossRef Search ADS PubMed  Domon JM, Baldwin L, Acket S et al.  . ( 2013) Cell wall compositional modifications of Miscanthus ecotypes in response to cold acclimation. Phytochemistry  85: 51– 61. Google Scholar CrossRef Search ADS PubMed  Goicoechea M, Lacombe E, Legay S et al.  . ( 2005) EgMYB2, a new transcriptional activator from Eucalyptus xylem, regulates secondary cell wall formation and lignin biosynthesis. Plant J  43: 553– 567. Google Scholar CrossRef Search ADS PubMed  Grima-Pettenati J, Soler M, Camargo ELO, Wang H ( 2012) Transcriptional regulation of the lignin biosynthetic pathway revisited: new players and insights. In: Jouanin L, Lapierre C (eds) Advances in botanical research: lignins biosynthesis, biodegradation and bioengineering . Elsevier, San Diego, pp 173– 218. Google Scholar CrossRef Search ADS   Harfouche A, Meilan R, Altman A ( 2011) Tree genetic engineering and applications to sustainable forestry and biomass production. Trends Biotechnol  29: 9– 17. Google Scholar CrossRef Search ADS PubMed  Hausman JF, Evers D, Thiellement H, Jouve L ( 2000) Compared responses of poplar cuttings and in vitro raised shoots to short-term chilling treatments. Plant Cell Rep  19: 954– 960. Google Scholar CrossRef Search ADS   Hussey SG, Mizrachi E, Creux NM, Myburg AA ( 2013) Navigating the transcriptional roadmap regulating plant secondary cell wall deposition. Front Plant Sci  4: 1– 21. Google Scholar CrossRef Search ADS PubMed  Hussey SG, Saidi MN, Hefer CA, Myburg AA, Grima-Pettenati J ( 2015) Structural, evolutionary and functional analysis of the NAC domain protein family in Eucalyptus. New Phytol  206: 1337– 1350. Google Scholar CrossRef Search ADS PubMed  LeGall H, Philippe F, Domon J, Gillet F, Pelloux J, Rayon C ( 2015) Cell wall metabolism in response to abiotic stress. Plants  4: 112– 166. Google Scholar CrossRef Search ADS PubMed  Legay S, Lacombe E, Goicoechea M, Brière C, Séguin A, Mackay J, Grima-Pettenati J ( 2007) Molecular characterization of EgMYB1, a putative transcriptional repressor of the lignin biosynthetic pathway. Plant Sci  173: 542– 549. Google Scholar CrossRef Search ADS   Legay S, Sivadon P, Blervacq A-S et al.  . ( 2010) EgMYB1, an R2R3 MYB transcription factor from eucalyptus negatively regulates secondary cell wall formation in Arabidopsis and poplar. New Phytol  188: 774– 786. Google Scholar CrossRef Search ADS PubMed  Lepikson-Neto J, Alves A, Simões R et al.  . ( 2013) Flavonoid supplementation reduces the extractive content and increases the syringyl/guaiacyl ratio in Eucalyptus grandis × Eucalyptus urophylla hybrid trees. BioResources  8: 1747– 1757. Google Scholar CrossRef Search ADS   Mashkina OS, Butorina AK ( 2003) Genetic engineering of forest woody plants. Genetika  39: 309– 317. Google Scholar PubMed  Méchin V, Laluc A, Legée F, Cezard L, Denoue D, Barrie Y, Lapierre C ( 2014) Impact of the brown-midrib bm5 mutation on maize lignins. J Agric Food Chem  62: 5102– 5107. Google Scholar CrossRef Search ADS PubMed  Myburg A, Potts B, Marques CM, Kirst M, Gion J, Grattapaglia D, Grima-Pettenati J ( 2007) Eucalyptus. In: Kole CE (ed) Genome mapping and molecular breeding in plants . Springer, New York, NY, USA, pp 115– 160. Myburg AA, Grattapaglia D, Tuskan GA et al.  . ( 2014) The genome of Eucalyptus grandis. Nature  510: 356– 362. Google Scholar CrossRef Search ADS PubMed  Nakano Y, Yamaguchi M, Endo H, Rejab NA, Ohtani M ( 2015) NAC-MYB-based transcriptional regulation of secondary cell wall biosynthesis in land plants. Front Plant Sci  6: 288. Google Scholar CrossRef Search ADS PubMed  Navarro M, Marque G, Ayax C, Keller G, Borges JP, Marque C, Teulieres C ( 2009) Complementary regulation of four Eucalyptus CBF genes under various cold conditions. J Exp Bot  60: 2713– 2724. Google Scholar CrossRef Search ADS PubMed  Nguyen H, Cao PB, San Clemente H, Ployet R, Mounet F, Harvengt L, Ladouce N, Marque C, Teulieres C ( 2017) Special trends of CBF and DREB2 groups in Eucalyptus gunnii vs Eucalyptus grandis suggest that CBF are master players in the trade-off between growth and stress resistance. Physiol Plant  159: 445– 467. Google Scholar CrossRef Search ADS PubMed  Olenichenko NA, Zagoskina NV ( 2005) Response of winter wheat to cold: production of phenolic compounds and L-phenylalanine ammonia lyase activity. Appl Biochem Microbiol  41: 600– 603. Google Scholar CrossRef Search ADS   Pfaffl MW ( 2001) A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res  29: 2002– 2007. Google Scholar CrossRef Search ADS   Plomion C, Leprovost G, Stokes A ( 2001) Wood formation in trees. Plant Physiol  127: 1513– 1523. Google Scholar CrossRef Search ADS PubMed  Potts BM, Potts WC, Kantvilas G ( 2001) The miena cider gum, Eucalyptus Gunnii subsp. Divaricata (Myrtaceae): a taxon in rapid decline. Pap Proc Roy Soc Tas  135: 57– 61. Ranik M, Myburg AA ( 2006) Six new cellulose synthase genes from Eucalyptus are associated with primary and secondary cell wall biosynthesis. Tree Physiol  5: 545– 556. Google Scholar CrossRef Search ADS   Ruijter JM, Ramakers C, Hoogaars WMH, Karlen Y, Bakker O, Van Den Hoff MJB, Moorman AFM ( 2009) Amplification efficiency: linking baseline and bias in the analysis of quantitative PCR data. Nucleic Acids Res  37: e45. Google Scholar CrossRef Search ADS PubMed  Ruprecht C, Persson S ( 2012) Co-expression of cell wall-related genes: new tools and insights. Front Plant Sci  3: 1– 7. Google Scholar CrossRef Search ADS PubMed  Ruprecht C, Mutwil M, Saxe F, Eder M, Nikoloski Z, Persson S ( 2011) Large-scale co-expression approach to dissect secondary cell wall formation across plant species. Front Plant Sci  1: 2– 23. Soler M, Camargo ELO, Carocha V et al.  . ( 2015) The Eucalyptus grandis R2R3-MYB transcription factor family: evidence for woody growth-related evolution and function. New Phytol  206: 1364– 1377. Google Scholar CrossRef Search ADS PubMed  Soler M, Plasencia A, Lepikson-Neto J et al.  . ( 2016) The woody-preferential gene EgMYB88 regulates the biosynthesis of phenylpropanoid-derived compounds in wood. Front Plant Sci  7: 1422. Google Scholar CrossRef Search ADS PubMed  Southerton SG, Marshall H, Mouradov A, Teasdale RD ( 1998) Eucalypt MADS-box genes expressed in developing flowers. Plant Physiol  118: 365– 372. Google Scholar CrossRef Search ADS PubMed  Taylor-Teeples M, Lin L, de Lucas M et al.  . ( 2015) An Arabidopsis gene regulatory network for secondary cell wall synthesis. Nature  517: 571– 575. Google Scholar CrossRef Search ADS PubMed  Teulières C, Marque C, Nagata T, Lörz, Wildhom JM ( 2007) Eucalyptus. In: Pua EC, Davey MR (eds) Biotechnology in agriculture and forestry . Springer Verlag, Berlin, Germany, pp 387– 406. VanAcker R, Vanholme R, Storme V, Mortimer JC, Dupree P, Boerjan W ( 2013) Lignin biosynthesis perturbations affect secondary cell wall composition and saccharification yield in Arabidopsis thaliana. Biotechnol Biofuels  6: 46. Google Scholar CrossRef Search ADS PubMed  Vanholme R, Demedts B, Morreel K, Ralph J, Boerjan W ( 2010) Lignin biosynthesis and structure. Plant Physiol  153: 895– 905. Google Scholar CrossRef Search ADS PubMed  Vazquez-Cooz I, Meyer RW ( 2009) A differential staining method to identify lignified and unlignified tissues. Biotech Histochem  77: 277– 282. Google Scholar CrossRef Search ADS   Ventorim G, Alves AF, Penna LS, Francis RC ( 2014) Effect of S/G ratio on kraft pulping and ecf bleaching of some poplars and eucalyptus. Cellul Chem Technol  48: 365– 373. Wu AM, Rihouey C, Seveno M, Hörnblad E, Singh SK, Matsunaga T, Ishii T, Lerouge P, Marchant A ( 2009) The Arabidopsis IRX10 and IRX10-LIKE glycosyltransferases are critical for glucuronoxylan biosynthesis during secondary cell wall formation. Plant J  57: 718– 731. Google Scholar CrossRef Search ADS PubMed  Yao DX, Wei Q, Xu WY, Syrenne RD, Yuan JS, Su Z ( 2012) Comparative genomic analysis of NAC transcriptional factors to dissect the regulatory mechanisms for cell wall biosynthesis. BMC Bioinformatics  13: 12. Google Scholar CrossRef Search ADS PubMed  Zhao Q, Nakashima J, Chen F et al.  . ( 2013) Laccase is necessary and nonredundant with peroxidase for lignin polymerization during vascular development in Arabidopsis. Plant Cell  25: 3976– 3987. Google Scholar CrossRef Search ADS PubMed  Zhong R, Ye ZH ( 2015) Secondary cell walls: biosynthesis, patterned deposition and transcriptional regulation. Plant Cell Physiol  56: 195– 214. Google Scholar CrossRef Search ADS PubMed  Zhong R, Demura T, 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. Google Scholar CrossRef Search ADS PubMed  Author notes handling Editor Janice Cooke © The Author 2017. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com

Journal

Tree PhysiologyOxford University Press

Published: Mar 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 12 million articles from more than
10,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Unlimited reading

Read as many articles as you need. Full articles with original layout, charts and figures. Read online, from anywhere.

Stay up to date

Keep up with your field with Personalized Recommendations and Follow Journals to get automatic updates.

Organize your research

It’s easy to organize your research with our built-in tools.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

Monthly Plan

  • Read unlimited articles
  • Personalized recommendations
  • No expiration
  • Print 20 pages per month
  • 20% off on PDF purchases
  • Organize your research
  • Get updates on your journals and topic searches

$49/month

Start Free Trial

14-day Free Trial

Best Deal — 39% off

Annual Plan

  • All the features of the Professional Plan, but for 39% off!
  • Billed annually
  • No expiration
  • For the normal price of 10 articles elsewhere, you get one full year of unlimited access to articles.

$588

$360/year

billed annually
Start Free Trial

14-day Free Trial