Gene expression patterns underlying changes in xylem structure and function in response to increased nitrogen availability in hybrid poplar



    Corresponding author
    1. Department of Renewable Resources, University of Alberta, Edmonton, Alberta, Canada T6G 2E3
      L. Plavcová. Fax: +1 780 492 1767; e-mail:
    Search for more papers by this author

    1. Department of Renewable Resources, University of Alberta, Edmonton, Alberta, Canada T6G 2E3
    Search for more papers by this author

    1. Department of Renewable Resources, University of Alberta, Edmonton, Alberta, Canada T6G 2E3
    Search for more papers by this author
    • Present address: Institut de Recherche en Biologie Végétale (IRBV), Montreal, QC H1X 2B2, Canada.


    1. Department of Botany, University of British Columbia, Vancouver, British Columbia, Canada V6T 1Z4
    Search for more papers by this author

    1. Department of Botany, University of British Columbia, Vancouver, British Columbia, Canada V6T 1Z4
    Search for more papers by this author

L. Plavcová. Fax: +1 780 492 1767; e-mail:


Nitrogen availability has a strong influence on plant growth and development. In this study, we examined the effect of nitrogen availability on xylogenesis in hybrid poplar (Populus trichocarpa × deltoides H11-11). Saplings of hybrid poplar were fertilized for 33 d with either high or adequate levels of ammonium nitrate. We observed enhanced radial growth, wider vessels and fibres and thinner fibre walls in the secondary xylem of high N relative to adequate N plants. These anatomical differences translated into altered hydraulic properties with xylem being more transport efficient but also more vulnerable to drought-induced cavitation in high N plants. The changes in xylem structure and function were associated with differences in gene expression as revealed by the transcriptome analysis of the developing xylem region. We found 388 genes differentially expressed (fold change ±1.5, P-value ≤ 0.05), including a number of genes putatively involved in nitrogen and carbohydrate metabolism and various aspects of xylem cell differentiation. Several genes encoding known transcriptional regulators of secondary cell wall deposition were down-regulated in high N plants, corresponding with thinner secondary cell walls in these plants. The results of this study provide us with gene candidates potentially affecting xylem hydraulic and structural traits.


Wood (i.e. secondary xylem produced by cambial activity) of poplars represents an important raw material of great economic value (Balatinecz, Kretschmann & Leclercq 2001). Poplar biomass also provides a promising bioenergy feedstock that could help to reduce our dependency on fossil fuels (Sannigrahi, Ragauskas & Tuskan 2010). From a biological perspective, wood serves three main functions that are fundamental for plant growth. These functions are (1) long-distance transport of water and nutrients from roots to transpiring leaves, (2) providing mechanical support to the plant body and (3) storage of carbohydrates, water and various other specialized compounds. In a typical hardwood such as poplar, these three functions are divided between three different cell types – vessel elements, fibres and living parenchyma.

Vessel elements and fibres represent 85–90% (vol/vol) of mature wood in poplar (Mellerowicz et al. 2001). Hydraulic and mechanical properties of wood are closely associated with the physical structure of these cells. For instance, xylem hydraulic conductivity is proportional to the vessel diameter to the fourth power as predicted by the Hagen–Poiseuille equation (Tyree & Zimmermann 2002), and mechanical parameters such as modulus of rupture have been linked with wood density and fibre lumen diameters (Woodrum, Ewers & Telewski 2003; Pratt et al. 2007; Onoda, Richards & Westoby 2010). Moreover, hydraulic and mechanical functions appear to be closely integrated as strong mechanical support may be required to prevent implosion of xylem conduits under high xylem tension (Hacke et al. 2001; Jacobsen et al. 2005; Pittermann et al. 2006; Coleman et al. 2008). Furthermore, wood density and fibre length together with the chemical composition of wood are critical factors that determine its material properties and hence its suitability for a specific end use in the wood-processing industry.

Wood structure is established during xylogenesis as a result of cambial activity. Wood is formed through a series of precisely regulated developmental steps that include cell division, cell expansion, secondary cell wall deposition and programmed cell death (Samuels, Kaneda & Rensing 2006). Many genes influencing xylem differentiation in poplar have been recently identified, including several key regulatory and structural genes (e.g. Aspeborg et al. 2005; Groover et al. 2010; Zhong et al. 2011). The process of xylogenesis and the resulting xylem phenotype are strongly affected by environmental conditions such as water (Arend & Fromm 2007), nutrient (Lautner et al. 2007; Hacke et al. 2010) and light availability (Plavcová, Hacke & Sperry 2011). The developmental programme giving rise to a specific physiological and anatomical xylem phenotype is underpinned by changes in gene expression as demonstrated by recent studies describing transcriptional changes in developing xylem of poplars subjected to drought (Berta et al. 2010) and high salinity (Janz et al. 2012). However, more research is needed to better understand the molecular mechanisms underlying xylem phenotypic plasticity as it is likely that expression of different genes is altered by different environmental triggers.

In this study, we used nitrogen (N) fertilization to perturb the xylem phenotype of hybrid poplar (Populus trichocarpa × deltoides, clone H11-11) saplings and to investigate corresponding changes in gene expression. Nitrogen fertilization has a profound effect on poplar growth and development including xylogenesis (Harvey & van den Driessche 1999; Pitre, Cooke & Mackay 2007a; Pitre et al. 2007b; Hacke et al. 2010). The influence of nitrogen supply on the expression of selected genes has been evaluated in poplar leaves, roots, phloem and bulk xylem (Cooke et al. 2003; Cooke, Martin & Davis 2005; Ehlting et al. 2007; Hacke et al. 2010). To our knowledge, there is only one genome-wide study focused on the expression of nitrogen availability-related genes in the cambial region of poplar (Pitre et al. 2010), and this study was specifically designed to compare the effects of nitrogen fertilization and stem leaning on wood formation. In contrast, our study was designed to explore changes in gene expression that may underlie traits related to xylem water transport. Our goal was to identify candidate genes that may be linked with increased radial growth, wide vessel diameters and decreased wood density that we expected to be differentially expressed in poplars growing under high N availability.


Plant material

Seedlings of hybrid poplar (P. trichocarpa × deltoides H11-11) were produced from rooted cuttings and maintained in a growth chamber under the following conditions: 16/8 h day/night cycle, 24/18 °C day/night temperature, c. 75% daytime relative humidity, photosynthetically active radiation of c. 400 µmol m−2 s−1. During the initial phase of sapling establishment, plants were kept in 6 L pots filled with a standard gardening soil and fertilized once a week with 500 mL of N-P-K: 20-20-20 fertilizer (1 g L−1 dilution; Plant Products Co. Ltd., Brampton, Ontario, Canada). After 6 weeks, when the saplings were c. 70 cm in height, plants were randomly assigned to either adequate or high nitrogen (N) treatments. The fertilizer was applied every other day as either 0.75 mm or 7.5 mm NH4NO3 in 0.5 × Hocking's complete nutrient solution (Hocking 1971) for adequate and high N plants, respectively. To avoid drought stress, plants were irrigated with tap water on the days when the fertilizer was not applied. The fertilization treatment was applied for 33 d after which plants were harvested. The total amount of N added to each plant over the course of experimental treatments was 0.12 and 1.2 g for adequate N and high N plants, respectively. Final height and stem basal diameter (Dstem) of each sapling were assessed with a measuring tape and a caliper. Leaf area (AL) was measured with an area meter (LI-3100, Li-Cor, Lincoln, NE, USA). Stem segments c. 25 cm in length were excised 5 cm above the root collar, placed in a dark plastic bag with a wet paper towel and stored at 4 °C until hydraulic measurements were conducted. The same stem segments were later used for anatomical measurements. For RNA extraction, developing xylem tissue was collected from 50 cm long stem segments distally adjacent to the segments used for hydraulic and anatomical measurements. In both treatments, the secondary xylem was well developed and no obvious differences in stem maturation were apparent in this region of the stem. First, the bark was peeled from the stem and discarded. To obtain the tissue for the microarray analysis, the developing secondary xylem tissue was scraped from the exposed surface of the woody core using a fresh razor blade until resistance could be felt and the colour of the scraping changed, indicating that fully mature xylem cells were reached. The scraped tissue was immediately frozen in liquid nitrogen and stored at −80 °C until used for RNA extraction. Careful preliminary observations using a light microscope revealed that the bark separates from the stem in the cambial region. Thus, the tissue scraped from the outermost layer of the exposed xylem contained newly formed expanding xylem cells and cells that underwent secondary wall thickening. Although this sampling method did not allow sampling-specific developmental stages, it has been widely used to analyse transcript and protein levels in developing secondary xylem (Gray-Mitsumune et al. 2004; Berta et al. 2010; Song et al. 2011).

Xylem anatomy

Xylem anatomy was analysed in stem segments sampled at a height of 10–20 cm above the root collar using light microscopy as described previously (Plavcováet al. 2011). Five or six individual stems were measured for each treatment. The cross-sectional surface of stems was photographed with a digital camera attached to a stereomicroscope (MS5; Leica Microsystems, Wetzlar, Germany) at 10–16× magnification. Xylem cross-sectional area (AX), excluding pith and bark, was measured with image analysis software (ImagePro Plus version 6.1, Media Cybernetics, Silver Spring, MD, USA). Vessel lumen diameters were measured in two radial sectors on stem cross sections prepared with a sliding microtome and stained with toluidine blue. Between 300 and 400 vessels observed at 200× magnification were measured from each stem. Since fibres are much narrower than vessels, resin-embedded samples and 400× magnification were used to produce reliable measurements of fibre lumen diameters. 200–300 fibres from at least three different randomly selected areas were measured for each stem. In addition, measurements of fibre double wall thickness were conducted on at least 80 fibre pairs. Only fibres that did not have a conspicuous gelatinous layer were selected for measurements of lumen diameters and double wall thickness. Vessel element and fibre length measurements were conducted on macerated xylem tissue. At least 100 individual cells were photographed under 100× magnification and measured for each stem. Vessel length was assessed using the silicone injection method (Sperry, Hacke & Wheeler 2005). Finally, wood density was measured by the water displacement method. Debarked stem segments c. 2 cm in length were longitudinally split with a razor blade and the pith was removed. Samples were submersed in a beaker of water on a balance to measure the wood fresh volume. Samples were then oven-dried at 70 °C for 48 h and weighted. Wood density was expressed as dry weight per fresh volume. For statistical analysis, an independent two-sample t-test was used to compare the differences in means between the treatments. Prior to the analysis, normality and homogeneity of variances were graphically checked.

Hydraulic measurements

Stem hydraulic conductivity and vulnerability to cavitation were measured following the standard methodology described in detail in Hacke et al. (2010). Stem segments 14.2 cm in length were flushed with standard measuring solution (20 mm KCl +1 mm CaCl2 filtered at 0.2 µm) for 20 min at 50 kPa. Subsequently, the gravity-driven flow through the segments was recorded using an electronic balance (CP225, Sartorius, Göttingen, Germany) interfaced with a computer. The value of maximal conductivity was normalized by xylem cross-sectional area (AX) to calculate the xylem-specific hydraulic conductivity (KS). After measuring maximal conductivity, stem segments were fixed into a custom-built centrifuge rotor and spun to progressively more negative pressures. Hydraulic conductivity was measured after each pressure increment. The percentage loss of hydraulic conductivity was plotted against the corresponding xylem pressure to generate vulnerability curves. Data points were fitted with a Weibull function and the xylem pressure corresponding to 50% loss of conductivity (P50) was calculated for each segment.

Microarray analysis

Total RNA was extracted from six individual poplar saplings per treatment using the hexadecyltrimethylammonium bromide (CTAB) extraction protocol of Chang, Puryear & Cairney (1993). Total RNA quality for each individual sample was assessed with an Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA, USA) prior to microarray analysis. Samples with an RNA integrity number value of greater than or equal to 8.0 were deemed to be acceptable for microarray analysis. Samples were prepared following NimbleGen's Arrays User Guide (Gene Expression Analysis version 3.2; Roche NimbleGen Inc., Madison, WI, USA). 10 µg of each total RNA was converted to double-stranded (ds) cDNA with the Invitrogen SuperScript Double-Stranded cDNA Synthesis Kit (Invitrogen, Carlsbad, CA, USA), 1 µg of each ds cDNA was fluorescently labelled using the NimbleGen One-Color DNA Labeling Kit, and 4 µg of each Cy3-labelled sample was hybridized on Roche NimbleGen poplar gene expression microarrays (Design Name Populus 135K EXP HX12 090828). These arrays target 55 794 gene models predicted in the P. trichocarpa genome with each gene model represented by three unique 60-mer probes. Arrays were scanned at 5 µm resolution with a Molecular Devices GenePix 4200AL scanner (Molecular Devices LLC, Sunnyvale, CA, USA). NimbleScan version 2.5 was used for quantitation and robust multichip average (RMA) normalization of data which included quantile normalization and background subtraction. Agilent's GeneSpring 7.3.1 was used to analyse the normalized data. To find significantly differentially regulated genes, fold changes between the compared groups and P-values gained from t-test between the same groups were calculated. The P-values were further corrected for multiple testing (MTC) using the method of Benjamini & Hochberg (1995). The t-tests were performed on normalized data that had been log transformed and the variances were not assumed to be equal between sample groups. The genes with MTC P-value ≤ 0.05 and fold change ±1.5 were considered to be significantly differentially regulated. The robustness of the selected cut-off criteria was validated with quantitative real-time PCR (qRT-PCR) analysis. The gene model names and annotation is based on the Phytozome v2.2 version of the Populus genome (

Microarray qRT-PCR validation

One microgram of total RNA was treated with DNase I (Invitrogen) and used for first strand cDNA synthesis using oligo(dT)23VN (IDT, Coralville, IA, USA) and SuperScript II reverse transcriptase (Invitrogen) according to the manufacturer's protocols. Four potential reference genes were identified by screening the microarray data for cDNAs whose signals remained apparently unchanged (fold difference ratios between 0.97 and 1.05). Eight candidate genes were also selected in the microarray and included in the validation assay. Gene-specific qRT-PCR primers were designed mainly in the 3′ untranslated (UTR) region using Primer Express v3 (Applied Biosystems, Foster City, CA, USA; Supporting Information Table S1). For each gene, PCR efficiency (E) was determined from a four-point cDNA serial dilution, according to E = 10 [−1/slope]. The stability of the gene expression profile of the four potential reference genes was evaluated in three biological replicates for each of the two fertilization treatments. The two most stable reference genes were selected for the qRT-PCR assay [protein phosphatase 2A (PP2A), POPTR_0010s13760; and yellow-leaf-specific gene 8 (YLS8), POPTR_0007s07660]. RT-PCR was performed on a 7900 HT Fast Real-Time PCR system (Applied Biosystems). Assays were carried out in 384-well plates. Three biological replicates, each with three technical replicates, were assayed for high N and adequate N nitrogen treatments. A negative control (no cDNA template) was included for every gene. PCR was carried out in a final volume of 10 µL including a final concentration of 20 ng of cDNA, 0.4 µm of each primer (IDT), 1× master mix containing 0.2 mm dNTPs, 0.3 U Platinum Taq polymerase (Invitrogen), 0.25× SYBR Green and 0.1× ROX. PCR conditions were as previously described (Almeida-Rodriguez et al. 2010). Samples were subjected to auto Ct (cycle threshold) for analysis, and dissociation curves were verified for each of the genes. Changes in gene expression of nine target genes (including YLS8) were calculated according to Pfaffl (2001), relative to the reference gene PP2A.


Nitrogen availability affects xylem structure and function in hybrid poplar

We exposed clonally propagated hybrid poplar ramets (P. trichocarpa × deltoides, clone H11-11) to either adequate or high levels of ammonium nitrate. As expected, nitrogen fertilization enhanced growth of poplar saplings in both height and girth (Table 1). High N availability influenced not only the amount but also the structure of the secondary xylem produced. Substantial differences in xylem cell dimensions and wall reinforcement were identified when adequate versus high N plants were compared (Table 2, Fig. 1). More specifically, vessel and fibre lumens were significantly wider (Fig. 2) and the average vessel element length was higher in high N plants (Table 2). In contrast, the mean vessel length was not significantly different (Table 2). However, like vessel element length, a slight tendency towards longer vessels was identified in high N plants (Supporting Information Fig. S1b), suggesting that there may be a subtle effect. Similarly, the average fibre length did not significantly differ between the treatments (Table 2); however, a decrease in the proportion of fibres longer than 700 µm was apparent in high N plants (Supporting Information Fig. S1c). Lower fibre length in high N-treated plants has been previously reported as one of the hallmarks of nitrogen fertilization in this hybrid poplar genotype (Pitre et al. 2007a, 2010). These studies have also reported increased occurrence of fibres with thick secondary cell walls resembling a gelatinous layer typical for the reaction wood in high N-treated plants of this genotype. While we noticed the gelatinous fibres in our plant material as well, we found such layers in plants treated with both high N and adequate N. Furthermore, the gelatinous fibres were found only in certain regions of cross sections usually forming a distinct band around the stem. In regions where fibres lacked the gelatinous layer, the regular secondary cell wall was thinner in high N in comparison with adequate N plants (Fig. 1, Table 2). In agreement with these results, lower secondary cell wall thickness has been reported in three poplar species subjected to nitrogen fertilization (Luo et al. 2005). Thus, while nitrogen fertilization may stimulate gelatinous layer production (Pitre et al. 2007a, 2010), it also negatively affects the deposition of the regular secondary cell wall. Increased vessel and fibre lumen diameters and lower wall reinforcement in fibres translated into lower wood density in high N plants (Table 2).

Table 1.  Growth characteristics of poplar saplings receiving 0.75 mm (adequate N) versus 7.5 mm (high N) levels of ammonium nitrate
 Adequate NHigh N
  1. Parameters shown are final height of saplings (height), stem diameter at 10 cm above the root collar (Dstem), xylem cross-sectional area at ∼10 cm above the root collar (AX) and plant total leaf area (AL). Mean ± SE (n = 8). Results of independent two-sample t-tests, testing for differences between adequate and high N plants, are indicated (*P ≤ 0.05, **P ≤ 0.01).

Height (cm)160 ± 2.3171.3 ± 2.2**
D stem (mm)8.9 ± 0.29.5 ± 0.1*
A X (mm2)40.3 ± 1.351.3 ± 1.8**
A L (m2)0.63 ± 0.030.99 ± 0.03**
Table 2.  Xylem characteristics of poplar saplings receiving 0.75 mm (adequate N) versus 7.5 mm (high N) levels of ammonium nitrate
 Adequate NHigh N
  1. Mean ± SE (n = 5–6). Results of independent two-sample t-tests, testing for differences between adequate and high N plants, are indicated (*P ≤ 0.05; **P ≤ 0.01; ns, non-significant).

Vessel lumen D (µm)38.9 ± 0.242.6 ± 0.7**
Vessel element length (µm)234 ± 2255 ± 4**
Vessel length (cm)5.2 ± 0.25.8 ± 0.2 ns
Fibre lumen D (µm)9.5 ± 0.411.2 ± 0.5*
Fibre length (µm)579 ± 9561 ± 7 ns
Fibre double wall thickness (µm)3.3 ± 0.12.7 ± 0.2*
Wood density (g cm−3)0.37 ± 0.010.31 ± 0.01**
Figure 1.

Cross sections of stem xylem from (a) adequate N and (b) high N plants. High N plants had wider fibres and vessels and thinner fibre walls than adequate N plants. Scale bar = 20 µm.

Figure 2.

Frequency histograms of (a) vessel and (b) fibre lumen diameters for adequate N (open bars) and high N (black bars) plants. Means and SE (n = 5–6).

The changes in xylem structure were paralleled by differences in hydraulic properties, which have important implications for plant water use. High N plants with larger vessel diameters transported water more efficiently, as measured by average xylem area-specific conductivity (Fig. 3a), than adequate N plants with narrower vessel diameters. More efficient water transport is correlated with the necessity to sustain the larger areas of transpiring leaves in high N-treated plants (Table 1). However, high N plants were more vulnerable to cavitation than adequate N plants, as shown by the less-negative P50 values in high N plants (Fig. 3b). Thus, although more efficient, water transport in high N plants was more prone to dysfunction under drought conditions. This finding is in agreement with previous work (Harvey & van den Driessche 1999; Hacke et al. 2010).

Figure 3.

Average (a) xylem area-specific conductivity (KS) and (b) xylem pressure causing 50% loss of conductivity (P50) of stem segments from adequate N (open bars) and high N (black bars) plants. Error bars are SE (n = 8). * indicates the means are significantly different at P ≤ 0.05; ** indicates significant differences at P ≤ 0.01 (independent two-sample t-test).

Nitrogen availability evokes transcriptional changes in developing xylem

To investigate the molecular basis for the developmental changes leading to the xylem phenotypes described above, we carried out gene expression profiling of developing xylem isolated from high N- and adequate N-treated plants. RNA was isolated from developing secondary xylem, exposed by peeling bark from stem segments adjacent to segments used for hydraulic and anatomical measurements, and hybridized to NimbleGen poplar full genome microarrays. Six biological replicates per treatment were analysed. Expression of 49 476 gene models was detected; of those 388 non-redundant genes showed statistically significant (P ≤ 0.05) 1.5-fold or greater changes in transcript abundance between the treatments. Out of these genes, 243 were up- and 145 down-regulated (Supporting Information Table S2). We used quantitative qRT-PCR to independently assay expression of nine selected genes, which confirmed the reliability of the microarray analysis (Supporting Information Fig. S2).

Microarray results suggest that extensive remodelling of the poplar transcriptome accompanies changes in xylem development and secondary cell wall deposition associated with increased nitrogen availability. The potential functional roles played by differentially regulated genes in xylem development were further analysed based on the annotations of the differentially regulated poplar genes or their presumed Arabidopsis orthologs. Genes of particular interest that might be associated with various aspects of stem growth and development including xylem and secondary wall development are presented in Table 3 and discussed below.

Table 3.  Genes differentially expressed in the developing xylem of hybrid poplar growing under high N versus adequate N availability. The poplar gene names indicated in the brackets were obtained from (1) Phytozome (, (2) Zhong et al. (2011), (3) Gupta & Sankararamakrishnan (2009). Fold change values represent a ratio between the normalized averaged values of high N relative to adequate N plants. Multiple testing corrections adjusted P-values of t-tests comparing adequate and high N plants are indicated
Populus trichocarpa Arabidopsis thaliana homologous geneTAIR descriptionFold change P-value
Phytozome v2.0 gene name
Amino acid transport    
 POPTR_0008s03620 At2g39130 Transmembrane amino acid transporter family protein4.750.014
 POPTR_0006s18790 At5g57685 (GDU3) Glutamine dumper 32.350.021
 POPTR_0008s12400 At2g01170 (BAT1) Bidirectional amino acid transporter 11.690.031
 POPTR_0001s36330 (Pt-LHT1.2)1 At5g40780 (LHT1) Lysine histidine transporter 1−1.870.037
 POPTR_0002s07960 (PtrAAP5)1 At1g77380 (AAP3) Amino acid permease 3−3.100.041
Organic acid metabolism    
 POPTR_0008s11330 At1g68750 (PPC4) Phosphoenolpyruvate carboxylase 42.520.015
 POPTR_0001s13510 (Pt-G6PD.2)1 At5g13110 (G6PD2) Glucose-6-phosphate dehydrogenase 22.510.032
 POPTR_0001s16300 (Pt-ALAAT1.1)1 At1g72330 (ALAAT2) Alanine aminotransferase 21.820.037
 POPTR_0013s00660 At1g09420 (G6PD4) Glucose-6-phosphate dehydrogenase 41.690.028
 POPTR_0007s14250 At4g37870 (PCK1) Phosphoenolpyruvate carboxykinase 11.680.047
 POPTR_0006s25280 At5g25880 (NADP-ME3) NADP-malic enzyme 31.570.032
Carbohydrate metabolism    
 POPTR_0006s06460 At5g20250 (DIN10) Raffinose synthase family protein5.210.018
 POPTR_0011s00480 At1g11580 (PMEPCRA) Methylesterase PCR A3.840.017
 POPTR_0007s05670 At5g65140 Haloacid dehalogenase-like hydrolase (HAD) superfamily protein/trehalose-phosphatase family protein3.420.029
 POPTR_0014s12000 (Pt-ATCSLD5.1)1 At1g02730 (CSLD5) Cellulose synthase-like D52.880.015
 POPTR_0002s21700 At2g05790 O-Glycosyl hydrolases family 17 protein2.740.030
 POPTR_0002s20130 (Pt-ATCSLD5.2)1 At1g02730 (CSLD5) Cellulose synthase-like D52.460.045
 POPTR_0009s01210 (Pt-XTR8.1)1 At2g36870 (XTH32) Xyloglucan endotransglucosylase/hydrolase 322.460.023
 POPTR_0008s20870 At5g18670 (BMY3) Beta-amylase 32.420.029
 POPTR_0008s13200 (Pt-CEL1.3)1 At1g70710 (GH9B1) Glycosyl hydrolase 9B12.070.042
 POPTR_0008s09380 (Pt-HIUHASE.1)1 At1g26560 (BGLU40) Beta glucosidase 402.060.023
 POPTR_0003s07040 (Pt-PE3.4)1 At3g14310 (PME3) Pectin methylesterase 32.050.033
 POPTR_0004s03830 At5g08370 (AGAL2) Alpha-galactosidase 22.010.019
 POPTR_0006s08030 At5g58090 O-Glycosyl hydrolases family 17 protein1.990.023
 POPTR_0010s13560 (Pt-XYL1.1)1 At1g68560 (XYL1) Alpha-xylosidase 11.980.011
 POPTR_0011s09660 At5g55180 O-Glycosyl hydrolases family 17 protein1.940.048
 POPTR_0019s09740 At1g71380 (CEL3) Cellulase 31.860.032
 POPTR_0002s20340 (Pt-SUS2.2)1 At4g02280 (SUS3) Sucrose synthase 31.790.049
 POPTR_0010s26170 At5g08380 (AGAL1) Alpha-galactosidase 11.750.046
 POPTR_0002s09450 At1g78060 Glycosyl hydrolase family protein1.730.022
 POPTR_0008s10090 At3g06500 Plant neutral invertase family protein1.690.043
 POPTR_0018s09560 (Pt-HXK1.1)1 At4g29130 (HXK1) Hexokinase 11.690.014
 POPTR_0018s14730 At5g58090 O-Glycosyl hydrolases family 17 protein1.650.041
 POPTR_0013s05620 At1g64760 O-Glycosyl hydrolases family 17 protein1.610.021
 POPTR_0010s11510 At1g68020 (ATTPS6) UDP-Glycosyltransferase/trehalose-phosphatase family protein1.580.015
 POPTR_0008s13590 At1g68020 (ATTPS6) UDP-Glycosyltransferase/trehalose-phosphatase family protein1.560.042
 POPTR_0003s13940 At1g64390 (GH9C2) Glycosyl hydrolase 9C21.500.033
 POPTR_0005s06280 At3g18660 (PGSIP1) Plant glycogenin-like starch initiation protein 1−1.550.023
 POPTR_0004s06840 At3g16520 (UGT88A1) UDP-glucosyl transferase 88A1−1.620.023
 POPTR_0006s11040 At5g01930 (MAN6) Glycosyl hydrolase superfamily protein−1.760.029
 POPTR_0004s06910 At3g16520 (UGT88A1) UDP-glucosyl transferase 88A1−1.790.009
 POPTR_0016s02640 At3g10740 (ASD1) Alpha-L-arabinofuranosidase 1−1.790.009
 POPTR_0016s14550 At5g01930 (MAN6) Glycosyl hydrolase superfamily protein−1.850.023
 POPTR_0002s11090 At1g09890 Rhamnogalacturonate lyase family protein−1.950.041
 POPTR_0016s02620 At3g10740 (ASD1) Alpha-L-arabinofuranosidase 1−1.950.040
 POPTR_0001s08750 At1g11260 (STP1) Sugar transporter 1−2.020.025
 POPTR_0002s02550 At1g19940 (GH9B5) Glycosyl hydrolase 9B5−2.020.025
 POPTR_0016s02690 At3g10740 (ASD1) Alpha-L-arabinofuranosidase 1−2.040.011
 POPTR_0005s27680 At4g35300 (TMT2) Tonoplast monosaccharide transporter2−2.070.047
Fasciclin-like arabinogalactan    
 POPTR_0014s16610 At4g12730 (FLA2) FASCICLIN-like arabinogalactan 22.070.010
 POPTR_0073s00210 At5g06390 (FLA17) FASCICLIN-like arabinogalactan protein 17 precursor−1.670.047
 POPTR_0004s16370 At5g03260 (LAC11) Laccase 112.650.026
 POPTR_0011s12090 At5g60020 (LAC17) Laccase 17−1.670.045
 POPTR_0011s12100 At5g60020 (LAC17) Laccase 17−1.910.040
 POPTR_0009s03940 At2g29130 (LAC2) Laccase 2−4.980.045
Transcription factor    
 POPTR_0009s01110 At5g22920 CHY-type/CTCHY-type/RING-type Zinc finger protein3.370.035
 POPTR_0009s09270 At3g49940 (LBD38) LOB domain-containing protein 383.210.019
 POPTR_0003s12240 At1g63100 GRAS family transcription factor2.520.028
 POPTR_0001s08850 At1g63100 GRAS family transcription factor2.50.009
 POPTR_0017s06590 At4g26400 RING/U-box superfamily protein2.380.011
 POPTR_0018s10510 At5g57660 (COL5) CONSTANS-like 52.350.036
 POPTR_0014s11940 At5g49300 (GATA16) GATA transcription factor 161.820.029
 POPTR_0005s05470 (Pt-BZO2.3)1 At5g28770 (BZO2H3) bZIP transcription factor family protein1.690.040
 POPTR_0005s20890 (Pt-SCL1.1)1 At1g21450 (SCL1) SCARECROW-like 1−1.610.021
 POPTR_0017s12740 (PtrZF1)2 At1g26610 C2H2-like zinc finger protein−1.630.041
 POPTR_0018s06790 (PtrNAC150)2 At4g29230 (NAC075) NAC domain containing protein 75−1.650.028
 POPTR_0006s27570 At5g25220 (KNAT3) KNOTTED1-like homeobox gene 3−1.680.032
 POPTR_0011s05740 (PtrNAC105)2 At4g28500 (NAC073) NAC domain containing protein 73−1.720.023
 POPTR_0003s11120 At2g30590 (WRKY21) WRKY DNA-binding protein 21−1.730.022
 POPTR_0006s04770 (Pt-HSFB3.1)1 At2g41690 (HSFB3) Heat shock transcription factor B3−1.730.011
 POPTR_0010s22320 At2g28550 (RAP2.7) Related to AP2.7−1.830.017
 POPTR_0013s08040 At2g40470 (LBD15) LOB domain-containing protein 15−1.830.041
 POPTR_0009s16590 At1g08320 (TGA9) bZIP transcription factor family protein−1.890.041
 POPTR_0004s04900 (PtrNAC157)2 At4g28500 (NAC073) NAC domain containing protein 73−1.910.033
 POPTR_0005s14860 At2g01275 RING/FYVE/PHD zinc finger superfamily protein−2.010.018
 POPTR_0009s03240 (Pt-MYB111.1)1 At3g46130 (MYB48) myb domain protein 48−2.280.038
 POPTR_0006s21640 At5g06510 (NF-YA10) Nuclear factor Y, subunit A10−2.910.010
 POPTR_0016s06860 At5g06510 (NF-YA10) Nuclear factor Y, subunit A10−4.120.022
Cell cycle    
 POPTR_0001s27890 At2g26760 (CYCB1;4) Cyclin B1;43.260.041
 POPTR_0009s16730 At1g76310 (CYCB2;4) Cyclin B2;42.870.047
 POPTR_0019s04670 (Pt-APY1.2)1 At5g18280 (APY2) Apyrase 25.970.023
 POPTR_0001s19180 At4g25150 HAD superfamily, subfamily IIIB acid phosphatase5.220.020
 POPTR_0012s09940 (Pt-PPD3.2)1 At5g50400 (PAP27) Purple acid phosphatase 273.820.023
 POPTR_0004s16720 (Pt-PAP.2)1 At2g16430 (PAP10) Purple acid phosphatase 101.640.018
 POPTR_0010s21700 (PtTIP1;3)3 At4g01470 (TIP1;3) Tonoplast intrinsic protein 1;32.260.047
 POPTR_0008s05050 (PtTIP1;4)3 At4g01470 (TIP1;3) Tonoplast intrinsic protein 1;32.250.041
 POPTR_0001s18730 (PtTIP2;1)3 At3g16240 (DELTA-TIP) Delta tonoplast integral protein1.870.029
 POPTR_0019s04640 (PtSIP1;2)3 At3g04090 (SIP1A) Small and basic intrinsic protein 1A−1.700.033
 POPTR_0001s45920 (PtNIP3;3)3 At4g10380 (NIP5;1) NOD26-like intrinsic protein 5;1−1.800.040
 POPTR_0006s09920 (PtPIP1;4)3 At4g00430 (PIP1;4) Plasma membrane intrinsic protein 1;4−3.270.040
Programmed cell death    
 POPTR_0008s03480 At3g10410 (SCPL49) Serine carboxypeptidase-like 49−1.530.032
 POPTR_0002s00720 At1g20850 (XCP2) Xylem cysteine peptidase 2−1.690.043
 POPTR_0014s07050 At4g00230 (XSP1) Xylem serine peptidase 1−1.850.030
 POPTR_0001s13140 At1g28110 (SCPL45) Serine carboxypeptidase-like 45−1.990.040
 POPTR_0002s15330 At4g00230 (XSP1) Xylem serine peptidase 1−2.560.023

Nitrogen metabolism

Several poplar genes with a putative function as amino acid transporters were differentially regulated in this study. Amino acids and amides represent the principal long-distance transport form of organic nitrogen. They are abundant in phloem and xylem sap of woody plants (Sauter & Van Cleve 1992; Weber et al. 1998) from which they can be translocated radially into the cambial zone where they are required for protein and lignin biosynthesis. Among the differentially regulated genes encoding amino acid transporters were an uncharacterized amino acid transporter, a homolog of Arabidopsis amino acid exporter GLUTAMINE DUMPER 3 (GDU3) and a homolog of BIDIRECTIONAL AMINO ACID TRANSPORTER 1 (BAT1). The importance of GDU3 and BAT1 genes in vascular tissue physiology is reinforced by the fact that they are highly expressed in the Arabidopsis vasculature (Dundar 2009; Pratelli et al. 2010). Two other amino acid transporter genes, namely genes homologous to Arabidopsis LYSINE HISTIDINE TRANSPORTER 1 (LHT1) and AMINO ACID PERMEASE 3 (AAP3), were down-regulated in high N plants. The specific role of the differentially regulated transporters cannot be elucidated from this study; nonetheles, these genes represent interesting candidates for future research on xylogenesis and nitrogen cycling in poplar. Increased expression during secondary xylem formation in poplar has been already demonstrated for the GDU3 and AAP3 homologs (Dharmawardhana, Brunner & Strauss 2010).

Furthermore, we found several genes involved in organic acid metabolism up-regulated in high N plants. Organic acids such as malate, citrate and α-oxoglutarate are required as carbon skeletons for amino acid synthesis. The genes up-regulated in this study included two genes encoding glucose-6-phosphate dehydrogenase of the pentose phosphate pathway, and several other enzymes (phosphoenolpyruvate carboxylase, phosphoenolpyruvate carboxykinase, alanin aminotransferase and NADP-malic enzyme) related to organic acid metabolism. In agreement with our findings, an increased nitrate supply has been previously shown to induce genes involved in the synthesis of organic acids in Arabidopsis (Scheible et al. 1997; Stitt 1999). The observed patterns in gene expression suggest that due to an increased need for nitrogen assimilation and amino acid synthesis, relatively more carbon is channelled towards the synthesis of organic acids in the high N plants relative to controls. Organic acid metabolism hence represents an important intersection between nitrogen and carbon metabolism.

Carbohydrate metabolism

The microarray analysis revealed extensive changes in the transcription of genes involved in carbohydrate metabolism. Sucrose synthase (Pt-SUS2.2) and plant-neutral invertase genes were up-regulated in high N plants. These enzymes break down sucrose into fructose and glucose monomers (Koch 2004) and are important in carbon partitioning into cellulose biosynthesis during secondary wall develoment (Hauch & Magel 1998; Coleman, Yan & Mansfield 2009). Furthermore, a gene encoding hexokinase (Pt-HXK1.1), an enzyme required for activation of non-phosphorylated sugar monomers, was up-regulated. Thus, these three enzymes could act together to provide free phosphorylated sugar monomers for the biosynthesis of cellulose and/or hemicellulose. Enhanced expression of enzymes involved in sucrose metabolism is expected in fast-growing high N plants (Table 1) considering that the developing xylem acts as a major carbon sink and carbohydrates for the production of new cells are delivered from photosynthetic source organs via the phloem mainly in the form of sucrose.

Three trehalose-phosphatase genes showed increased expression in high N plants. Up-regulation of genes involved in trehalose metabolism in response to increased nitrate availability was previously reported in Arabidopsis (Wang et al. 2003; Scheible et al. 2004), and it has been suggested that trehalose could act as a regulator of the pentose phosphate pathway (Wang et al. 2003). Simple sugars such as trehalose or glucose represent effective signalling molecules, and a pivotal role of sugar-mediated signalling during many developmental processes has been recognized (Rolland, Baena-Gonzalez & Sheen 2006). To our knowledge, the potential role of sugar sensing and signalling during xylogenesis has not yet been investigated. Nevertheless, it might provide a promising avenue for future research considering the complex dynamics of sugars (Schrader & Sauter 2002), hormones (Tuominen et al. 1997) and transcription factors (Du & Groover 2010) and their interplay that occurs in the cambial zone of trees.

The poplar genes potentially involved in metabolism of cell wall polysaccharides have been previously studied (Aspeborg et al. 2005); however, the specific functions of many of them still remain elusive. Only few genes encoding polysaccharide synthases and glycosyl transferases, enzymes responsible for cell wall polysaccharide synthesis, were differentially expressed in our dataset. Among them, two closely related cellulose synthase-like D genes (Pt-ATCSLD5.1, Pt-ATCSLD5.2) both showed more than twofold up-regulation in high N plants. These genes are likely involved in the biosynthesis of xylan (Aspeborg et al. 2005; Bernal et al. 2007), which represents a hemicellulosic polysaccharide that is abundantly present in poplar wood. Furthermore, two putative UDP-glucosyl transferase genes and a homolog to an Arabidopsis gene encoding plant glycogenin-like starch initiation protein 1 (PGSIP1) were down-regulated in high N plants. The protein encoded by poplar PGSIP1 may have a priming function for cell wall polysaccharides (Aspeborg et al. 2005). The fact that only few polysaccharide biosynthetic genes, and especially no genes related to cellulose synthesis, were down-regulated in high N plants is surprising, considering the thin cell wall phenotype of high N plants. However, the faster radial growth in high N plants (Table 1) could lead to thinner cell walls even under constant biosynthetic activity. Alternatively, other processes such as the perturbation of overall carbon partitioning, polysaccharide remodelling and alternations in lignin metabolism can drive the lower cell wall thickness in high N versus adequate N plants.

Up-regulation of a plethora of glycoside hydrolase (GH) genes in high N plants was one of the most apparent patterns revealed in this study. Five 1,3-β-D-glucan endohydrolases (GH17 family), two α-galactosidase, three endo-1,4-β-D-glucanase (GH9 family, cellulase), α-xylosidase, β-glucosidase and xyloglucan endotransglucosylase/hydrolase all showed increased transcript abundance in high N plants. High activity of GH enzymes is typical for growing and expanding tissue (Cosgrove 2005). The wide lumen diameters of vessels and fibres in high N plants (Fig. 2) indicate that more intensive cell expansion took place in the cambial region of high N-treated plants. Thus, a role of at least some of these genes in primary cell wall loosening conferring increased cell wall extensibility can be implied. Three α-L-arabinofuranosidases, two glycosyl hydrolases with a putative function of endo-β-mannanases and one endo-1,4-β-D-glucanase (GH9) showed lower expression levels in high N plants. α-L-arabinofuranosidases are most likely involved in modification of the carbohydrate moieties of arabinogalactan proteins (AGPs; Geisler-Lee et al. 2006; Kotake et al. 2006). AGPs in turn are involved in a number of developmental processes and their importance during secondary cell wall deposition has been reported in poplar (Dharmawardhana et al. 2010) and pine (Zhang et al. 2003). In our dataset, two genes encoding fasciclin-like AGPs (FLAs) were differentially expressed, one up- and the other down-regulated. Several FLAs have been previously shown to be involved in tension wood formation (Lafarguette et al. 2004; Andersson-Gunneras et al. 2006). Gelatinous layer formation is affected by nitrogen fertilization (Pitre et al. 2007a); hence, it is possible that the two genes differentially expressed in this study are involved in this process.

Substantial changes in pectin composition occur during xylogenesis (Guglielmino et al. 1997a; Hafren, Daniel & Westermark 2000). We found two pectin methyl esterase (PME) genes up-regulated and one rhamnogalacturonate lyase gene down-regulated in high N plants. PMEs catalyse de-esterification of pectin molecules. In the cambial zone of poplar, PME activity has been localized predominantly in cell corners (Guglielmino et al. 1997b) and a role of PMEs in xylem cell expansion and fibre elongation has been demonstrated (Siedlecka et al. 2008). Alternatively, some of the genes encoding GH and PME enzymes may play a role in the secondary cell wall deposition or production of gelatinous fibres as extensive remodelling involving carbohydrate hydrolysis occurs during the secondary cell wall biosynthesis as well (Aspeborg et al. 2005).

Lignin-related genes

Surprisingly, only few genes with a predicted function in secondary cell wall lignification were differentially regulated in this study. Among the few were three laccase (LAC) genes that were down-regulated in high N plants. These included two homologs of Arabidopsis LAC17 and a closely related homolog of LAC2. LAC17 is known to be directly involved in lignin monomer polymerization, and lac17 mutants have reduced lignin deposition (Berthet et al. 2011). Thus, reduced laccase polymerization activity may be one of the mechanisms underlying reduced lignin deposition in high N-treated plants previously reported (Pitre et al. 2007b).

Transcription factors

A number of transcription factors were differentially expressed in high N versus adequate N plants indicating extensive changes in regulation of gene transcription in response to fertilization. A gene homologous to LATERAL ORGAN BOUNDARY DOMAIN 38 (AtLBD38) showed 3.2-fold up-regulation in high N plants. Expression of AtLBD38 is strongly induced by nitrate addition in Arabidopsis and an important regulatory role for this and two other closely related genes (AtLBD37, AtLBD39) in response to nitrate availability has been reported (Rubin et al. 2009). These three LBD transcription factors act as repressors of many N-responsive genes such as nuclear factor Y subunit A-10 (NF-YA10; Rubin et al. 2009). In agreement, we found two poplar genes homologous to AtNF-YA10 strongly down-regulated in high N plants. At least some members of the LBD family of transcription factors have been recently identified as regulators of secondary growth in poplar (Yordanov, Regan & Busov 2010). A homolog of AtLBD15 that was down-regulated in high N plants in this study could fulfil such function as its paralog has been recently characterized to be part of the transcriptional network involved in secondary xylem patterning (Zhong et al. 2011).

Furthermore, three genes encoding NAC domain transcription factors, PtrNAC157, PtrNAC105 and PtrNAC150, and a zinc finger transcription factor PtrZF1were down-regulated in high N plants. These genes have been implicated as intermediate regulators in a transcriptional network governing secondary cell wall biosynthesis in poplar (Zhong et al. 2011). Thus, their lower transcript abundance could be correlated with the significantly thinner secondary cell walls observed in fibres of high N plants. Work in Arabidopsis and poplar showed that expression of many regulatory and structural genes related to xylem differentiation, including several genes differentially expressed in this study, is under the control of NAC domain transcription factor master regulators such as SND1, NST1, VND6 and VND7 (Kubo et al. 2005; Yamaguchi et al. 2011) and their homologs in poplar (Ohtani et al. 2011; Zhong et al. 2011). These regulators are characterized by the ability to ectopically activate secondary cell wall biosynthesis and deposition in vessels, fibres and other secondarily thickened cells (Yamaguchi et al. 2010; Ohtani et al. 2011). Five genes encoding these master regulators, namely PtVNS01/PtrWND5A, PtVNS03/PtrWND4A, PtVNS04/PtrWND4B, PtVNS06/PtrWND3B, PtVNS10/PtrWND2B (gene names according to Ohtani et al. 2011; Zhong et al. 2011) showed more than 1.5-fold down-regulation in high N plants but did not meet our criteria for significance (P ≤ 0.05). Nevertheless, it is reasonable to expect that even such potentially subtle changes in their expression are biologically significant, considering the central position of these transcription factors in the regulatory cascade controlling secondary cell wall biosynthesis. In addition, down-regulated in high N plants was the poplar homolog of Arabidopsis WRKY21, a gene of undefined function that is regulated by Arabidopsis VND7 in transgenic poplar (Ohtani et al. 2011). A homolog of the Arabidopsis KNOTTED-like gene KNAT3 was also down-regulated in high N plants. While the function of KNAT3 is unknown, this poplar gene is up-regulated in response to osmotic stress treatments (Bae et al. 2010), suggesting that it could play a role in adjustment of xylem to water stress.

Cell division, expansion and death

The processes of cell division, cell expansion and programmed cell death and their dynamics are important during xylem development. The increased radial stem growth of high N plants suggests that increased cell proliferation may have occurred in the cambial region, leading to more cambial derivatives. Transcription of several genes encoding important cell cycle regulators was elevated in high N plants. Two cyclin (CYC) genes were strongly up-regulated. At least some poplar CYC genes have been previously shown to have specific expression maxima on the xylem side of the cambium; and hence, have been linked with xylem cell proliferation (Schrader et al. 2004).

Another gene strongly up-regulated in high N plants was an apyrase (Pt-APY1.2). Apyrases displaying nucleoside triphosphate diphosphohydrolytic activity are involved in cell growth regulation and are highly expressed in rapidly growing tissues and/or tissues that accumulate auxin at high levels (Clark & Roux 2011). Interestingly, we also found three genes encoding acid phosphatases strongly up-regulated in high N plants. While acid phosphates are involved in a variety of physiological functions, they are also involved in xylem development. Their specific functions during xylem differentiation are not clear, but putative roles in secretion and resorption of sugars, secondary cell wall deposition (Charvat & Esau 1975), and/or degradation of cellular content (Gahan 1978) have been proposed.

Water uptake is essential to drive cell expansion. Three genes encoding aquaporins of the tonoplast intrinsic protein (TIP) class were up-regulated in high N plants, PtTIP1;3, PtTIP1;4 and PtTIP2;1. In contrast, three other aquaporin genes were down-regulated, a plasma membrane intrinsic protein PtPIP1;4, a small basic intrinsic protein PtSIP1;2 and a nodulin-like intrinsic protein PtNIP3;3 (gene names according to Gupta & Sankararamakrishnan 2009). Expression of PtTIP2;1 and PtSIP1;2 in response to high N availability was previously studied in the bulk xylem of poplar stems (Hacke et al. 2010) and the change in expression was in the same direction as in the present study. Expression of several aquaporin genes has been shown to peak in the radial expansion zone of poplar cambium (Schrader et al. 2004) and it has been speculated that aquaporins may play a role in xylogenesis by facilitating the flow of water into the zone of expanding cells (Groover et al. 2010; Hacke et al. 2010; Almeida-Rodriguez & Hacke 2012). Following this hypothesis, the up-regulation of the three TIPs observed in this study may be linked with the wider vessels in high N plants. Many TIPs and PIPs have been functionally characterized as water channels (Chaumont et al. 1998; Almeida-Rodriguez 2009; Sade et al. 2009; Secchi et al. 2009). In a recent study, an elevated hydraulic conductance of root tips was associated with a decreased expression of PtPIP1;4 suggesting that the PtPIP1;4 may not be a water channel (Almeida-Rodriguez, Hacke & Laur 2011). Similarly, NIPs usually display low water permeability but are permeable to small solutes such as urea and glycerol (Gomes et al. 2009) and the transport specificity of SIPs is largely unknown. Thus, it is reasonable to expect that the down-regulated aquaporins may fulfil roles other than water transport within the developing xylem region.

Programmed cell death (PCD) is a crucial process during xylogenesis, as xylem vessels and fibres are dead and hollow at maturity. Two genes homologous to Arabidopsis XYLEM SERINE PEPTIDASE 1 (XSP1) and a homolog of XYLEM CYSTEIN PEPTIDASE 2 (XCP2), known to be involved in PCD during xylem differentiation in Arabidopsis (Funk et al. 2002) showed significantly lower expression in high N plants. In addition, two other PCD-related genes, homologs of Arabidopsis SERINE CARBOXYPEPTIDASE-LIKE 45 and 49 (SCPL45 and 49), were down-regulated in high N plants indicating that cell maturation and death was suppressed. The down-regulation of genes involved in PCD in high N plants supports previous findings that nitrogen fertilization results in wood with more juvenile characteristics (Pitre et al. 2007b) and supports a recent notion that programmed cell death and secondary cell wall formation are correlated and governed by common regulatory mechanisms involving PtVNS/PtrWND transcription factors (Bollhöner, Prestele & Tuominen 2012).


In this study, we showed that fertilization with ammonium nitrate evokes changes in growth, anatomy and hydraulic properties of secondary xylem in the stems of hybrid poplar (Ptrichocarpa × deltoides, clone H11-11). These anatomical and physiological differences were underpinned by changes in transcription of hundreds of genes in the developing xylem region. Our results revealed that several transcriptional patterns previously observed in Arabidopsis roots and shoots in response to high nitrate availability are also elicited in the developing secondary xylem of poplar. An example of such common responses is increased expression of genes encoding the transcription factor LBD38 and enzymes involved in organic acid and trehalose metabolism. Such comparisons between a short-lived annual herb and a relatively long-lived perennial tree as well as between the different types of tissues suggest that metabolic and regulatory pathways controlled by the corresponding proteins may represent evolutionary conserved aspects of plant responses to high N availability. However, future studies are needed to confirm this finding in a broader range of species and experimental conditions and to better understand the specific roles of these genes and processes in plant responses to high N.

Furthermore, the data presented in this study shed light on molecular mechanisms that underlie the phenotypic plasticity of xylem hydraulic and structural traits in this hybrid poplar genotype. We identified gene candidates that may affect xylem cell dimensions and cell wall thickness, although detailed functional characterization of these genes in poplar is required to corroborate the proposed function. Based on our results, wider lumens of vessels and fibres in high N plants might be linked with increased expression levels of genes encoding cell wall loosening enzymes such as various glycoside hydrolases and genes encoding aquaporins that may facilitate increased water uptake into expanding xylem cells. Our results also suggest that the changes in xylem development and secondary wall deposition in response to N availability may, at least in part, be mediated by the differential expression of several transcription factors that are part of a core transcriptional cascade governed by the recently characterized master regulators of xylem cell differentiation, the NAC domain transcription factors (PtVNS/PtrWND). Future research will reveal if the same genes that were identified in this study underlie changes in xylem phenotype under different environmental conditions, or if distinctly different suits of genes are regulated by various environmental cues, yet resulting in a similar xylem phenotype.


The authors wish to thank Ms Anne Haegert (Laboratory for Advanced Genome Analysis, Prostate Centre, Vancouver) for providing microarray services and Ms Arlene Oatway (Advanced Microscopy Unit, Department of Biological Sciences, University of Alberta) for her help with microscopy work. We also thank Hua Bao (Faculty of Forestry, University of British Columbia) for assistance with bioinformatics and Dr. Michael Deyholos for his comments on an earlier version of the manuscript. U.H. acknowledges support for this research from the Canada Research Chair Program, an Alberta Ingenuity New Faculty Award and an NSERC Discovery grant. This work was also supported by funds to C.J.D. from the Genome British Columbia Applied Genomics Innovation Program.