Xyloglucan endo-transglycosylase-mediated xyloglucan rearrangements in developing wood of hybrid aspen

Abstract

Xyloglucan endo-transglycosylases (XETs) encoded by xyloglucan endo-transglycosylases/hydrolase (XTH) genes modify the xyloglucan-cellulose framework of plant cell walls, thereby regulating their expansion and strength. To evaluate the importance of XET in wood development, we studied xyloglucan dynamics and XTH gene expression in developing wood and modified XET activity in hybrid aspen (Populus tremula × tremuloides) by overexpressing PtxtXET16-34. We show that developmental modifications during xylem differentiation include changes from loosely to tightly bound forms of xyloglucan and increases in the abundance of fucosylated xyloglucan epitope recognized by the CCRC-M1 antibody. We found that at least 16 Populus XTH genes, all likely encoding XETs, are expressed in developing wood. Five genes were highly and ubiquitously expressed, whereas PtxtXET16-34 was expressed more weakly but specifically in developing wood. Transgenic up-regulation of XET activity induced changes in cell wall xyloglucan, but its effects were dependent on developmental stage. For instance, XET overexpression increased abundance of the CCRC-M1 epitope in cambial cells and xylem cells in early stages of differentiation but not in mature xylem. Correspondingly, an increase in tightly bound xyloglucan content was observed in primary-walled xylem but a decrease was seen in secondary-walled xylem. Thus, in young xylem cells, XET activity limits xyloglucan incorporation into the tightly bound wall network but removes it from cell walls in older cells. XET overexpression promoted vessel element growth but not fiber expansion. We suggest that the amount of nascent xyloglucan relative to XET is an important determinant of whether XET strengthens or loosens the cell wall.

Figure 1.

Distribution of XTH genes expressed in the wood-forming tissues of Populus plants within the XTH gene family. The phylogenetic tree presents predicted protein sequences for the XTH family of P. trichocarpa, numbered according to Geisler-Lee et al. (2006), and the gene models used are listed in Supplemental Table S1. Arabidopsis XTH proteins, numbered according to Yokoyama and Nishitani (2001), and NtXET1 of tobacco (Herbers et al., 2001) are included for comparison. The subfamily IIIA represents putative XEH proteins (Baumann et al., 2007).

Figure 2.

Expression profiles of wood-expressed XTH genes in the Populus stem, determined with a gene-specific macroarray. A, Cross-section through a P. tremula stem showing the sampled tissue fractions. B, XTH gene signal intensity in relation to the ubiquitin signal in the indicated stem tissue fractions and other tissues for comparison. Values shown are means ± se; n = 3 technical replicates. The same expression patterns were observed in samples from two independent trees, and representative data are shown. C, Distribution of the expression of selected XTH genes among different vegetative tissues. For each gene, the sum of expression in all tissues is set to 1. Genes with very low expression levels (greater than 20% of ubiquitin expression) are not included. Xylem 1 indicates primary-walled developing xylem comprising the vascular cambium and radial expansion zone, and xylem 2 indicates secondary-walled developing xylem.

Figure 3.

XET activity in developing xylem of PtxtXET16-34-OE lines and wild type (WT) plants. XET activity was measured as incorporation of [1-³H]XLLGol into xyloglucan. Means ± ranges from duplicate analyses of pooled samples from 10 trees are shown. Xylem 1 indicates primary-walled developing xylem comprising vascular cambium and the radial expansion zone, and xylem 2 indicates secondary-walled developing xylem.

Figure 4.

PtxtXET16-34 overexpression stimulates vessel element, but not fiber, radial expansion. Line designation is as in Figure 3. Error bars represent least square mean se values calculated for the ANOVA model. The effect of genotype was significant only for dimensions of the vessel elements, for which the P values of the contrasts between the transgenic lines and the wild type (WT) are given above the bars.

Figure 5.

Molecular size distribution of the hemicellulose fraction of polysaccharides from xylem 1 tissues of OE and wild-type (WT) lines (pooled samples from five to six trees per line) extracted with 24% KOH and analyzed by size exclusion chromatography. Polysaccharides were detected by monitoring the refractive index (RI) of the eluate (A), and xyloglucan concentrations in different fractions were determined by iodine staining (B). Line designation is as in Figure 3. The molecular masses of dextran standards in kD are shown above the graph.

Figure 6.

Molecular composition of hemicellulose fractions extracted from primary- and secondary-walled xylem tissues (denoted xylem 1 and 2, respectively) of OE and wild-type lines. The analyzed lines include two OE lines (2-2 and 5-2), a transgenic line with the same construct but having a wild-type level of PtxtXE16-34 expression (7-1; compare with Supplemental Fig. S1), and two independent samples of the wild type (WT 1 and WT 2). Data represent averages obtained from two technical replicates of pooled samples from three to six trees per line, sequentially extracted with EDTA, followed by 4% KOH and 24% KOH, and analyzed by methanolysis followed by trimethylsilyl derivatization and gas-liquid chromatography. Line designation is as in Supplemental Figure S1. Asterisks indicate significant differences between the transgenic lines and the wild type according to the ANOVA test with a contrast (P ≤ 0.05).

Figure 7.

Semiquantitative transmission electron microscopy analysis of xyloglucan labeling in situ with the monoclonal antibody CCRC-M1. Ultrathin sections of the wood-forming tissues of two trees representing PtxtXET16-34-OE lines (2-1 and 5-2) and two wild-type (WT) trees were probed with CCRC-M1 antibody, and the gold particles were counted along 3 μm of the common cell wall between pairs of adjacent cells in the cambial zone (CZ), radial expansion zone (RE), early secondary wall formation zone (SW), and nearly mature xylem (MX). For each cell type, both radial (R) and tangential (T) walls were examined. In RE, SW, and MX zones, labeling in primary and secondary wall layers of common fiber-fiber (FF), fiber-vessel (FV), and vessel-vessel (VV) walls was separately scored. P values indicate the significance of the genotype (wild-type versus transgenic) on the measured variables according to the ANOVA; NS indicates nonsignificant. Significances of other factors analyzed are given in the ANOVA table. DF, Degrees of freedom.

Figure 8.

Distribution of CCRC-M1-labeled xyloglucan (red dots) in cell walls along the wood development gradient in the wild type (WT) and transgenic PtxtXET16-34-OE lines for which quantitative labeling data are presented in Figure 7. CM, Compound middle lamella; F, fiber; S1 and S2, successive secondary wall layers; V, vessel element. Other labels are as in Figure 7. Except for the cambial zone, all presented walls are radial. Bar = 1 μm.