Cellulose Synthesis Takes the CesA Train

The presence of a cell wall is the principal feature that distinguishes plants from animals and imparts many of the characteristic gross morphological features of plants. Far from being a static tissue that merely provides support and acts as a barrier to invading pests and pathogens, the cell wall also is recognized as a dynamic partner of living plant cells that plays important roles in the absorption, transport, and secretion of substances throughout plant growth and development. The principal component of plant cell walls is cellulose, a fibrous polymer consisting of linear chains of β-(1,4)-linked glucose molecules. These ribbon-like glucan chains crystallize to form microfibrils that impart the characteristic flexible strength—similar to that of an equivalent thickness of steel—of cellulose.

Cellulose is synthesized in higher plants by large multimeric plasma membrane–bound complexes that form rosette structures at the ends of microfibrils (Brown, 1996). Despite the prominence and importance of cellulose in plants, it was not until the mid to late 1990s that the catalytic activity of these complexes was linked definitively to cellulose synthase (CesA) proteins (Arioli et al., 1998; Kimura et al., 1999) and CesA genes were identified in a number of plant species, including cotton, rice, and Arabidopsis (Pear et al., 1996; Arioli et al., 1998; Delmer, 1999). It is now known that plants contain multiple CesA proteins; for example, the CesA gene family in Arabidopsis contains 10 members, CesA1 to CesA10 (Richmond and Somerville, 2000), which show different patterns of expression and different functional characteristics.

The CesA nomenclature, outlined by Delmer (1999), is the standard nomenclature for cellulose synthase genes. However, many CesA genes have additional names based on mutational and other analyses, and it can be instructive to retain the use of the original name in addition to the CesA name because it emphasizes the functional genetics studies that have been performed. For example, Simon Turner and colleagues isolated a number of irregular xylem (irx) mutants, which are characterized by collapsed xylem resulting from a decrease in the cellulose content of secondary cell walls. Three IRX genes were cloned and found to encode CesA proteins: IRX1/CesA8, IRX3/CesA7, and IRX5/CesA4, all of which are essential for secondary cell wall biosynthesis (Taylor et al., 2003).

The irx mutants are not to be confused with isoxaben-resistant (ixr) mutants, which were isolated based on resistance to the herbicide isoxaben, an inhibitor of cellulose biosynthesis. Two IXR genes have been cloned, and these encode the CesA proteins IXR1/CesA3 and IXR2/CesA6, which are essential for primary cell wall biosynthesis (Scheible et al., 2001; Desprez et al., 2002). CesA3 and CesA6 also have been cloned from the Arabidopsis mutants constitutive expression of vegetative storage proteins (Ellis et al., 2002) and procruste1 (Fagard et al., 2000), respectively. The radial swelling1 (rsw1) mutant led to the isolation of RSW1/CesA1, which also is required for primary cell wall biosynthesis (Arioli et al., 1998).

Together, the work on cesA mutants suggests that three different CesA proteins interact as subunits within a cellulose synthase complex: CesA1, CesA3, and CesA6 form the complex in primary cell wall biosynthesis, whereas CesA4, CesA7, and CesA8 form the complex in secondary cell walls. Some of the properties of cellulose in secondary walls, which have many more β-glucan chains per microfibril and a higher overall cellulose content than primary walls, may be related in part to different functional properties of individual CesA subunits. There is no evidence that other proteins interact directly with CesA subunits within the CesA complex, but this remains a possibility.

In this issue of The Plant Cell, Gardiner et al. (pages 1740–1748) confirm previous work showing that IRX1/CesA8, IRX3/CesA7, and IRX5/CesA4 are all required for the assembly of a functional cellulose synthase complex during secondary wall formation in developing xylem tissue. The current work further shows that the correct localization of IRX proteins to regions of secondary cell wall deposition requires the presence of cortical microtubules, but actin filaments are unlikely to play a major role. In addition, the use of green fluorescent protein (GFP)–tagged CesA and time-lapse confocal imaging suggested a rapid reorientation of the CesA complex along the length of protoxylem filaments during secondary cell wall deposition. This observation is sure to stimulate further research into the mechanism of action and regulation of the CesA complex.

Gardiner et al. first conducted localization studies using antibodies specific for each of the three IRX proteins in wild-type and irx mutant plants. These studies showed that the IRX proteins colocalize with microtubule bands that mark the site of secondary wall deposition in developing xylem (Figure 1). In the absence of any one of the three subunits, the other two remain within the cell and do not migrate to the microtubules, demonstrating that all three proteins are necessary for the formation of a functional CesA complex. This finding provides strong support for the notion that three different CesA subunits interact together to form the CesA complex. Furthermore, localization of all three proteins was normal in an irx1 mutant that produces altered protein carrying a mutation in the CesA catalytic region, suggesting that it is the presence of the CesA subunits, and not catalytic activity, that is required for protein localization and the formation of the CesA complex.

Figure 1.

Localization of Microtubules and the CesA Protein IRX1 in Developing Xylem Vessels.

Confocal image showing the localization of α-tubulin (green) and IRX1 (red). The overlapping bands appear orange in the developing protoxylem vessels (center). (Figure courtesy of Simon Turner.)

The authors next used irx3 and irx5 null mutants to examine the involvement of cellulose biosynthesis in microtubule organization. Based on experiments with the cellulose synthesis inhibitor isoxaben, Fisher and Cyr (1998) argued that cellulose biosynthesis is required for microtubule organization and stabilization. These authors suggested that the microtubule/microfibril paradigm, which states that organized cortical microtubules direct the ordered deposition of cellulose in secondary walls, should be updated to include the possibility of a bidirectional flow of information between cellulose microfibrils and cortical microtubules. However, in irx3 and irx5 null mutants, microtubules in developing xylem appear normal, even though the CesA complex does not form in association with the microtubules and there is little or no cellulose synthesized in the secondary wall. Similarly, Sugimoto et al. (2001) found that decreased cellulose biosynthesis in the Arabidopsis CesA1 mutant rsw1 did not lead to abnormal microtubule organization. These results suggest that neither the presence nor the activity of the CesA complex is necessary for the correct formation of microtubules.

In other experiments, Gardiner et al. used the herbicide oryzalin, which causes depolymerization of microtubules, to show that microtubules appear to be required continuously for the maintenance of normal CesA protein localization. In plants treated with oryzalin, IRX3 protein localization was altered rapidly, coincident with microtubule depolymerization. This result appears to contradict the work of Sugimoto et al. (2001), who found that cellulose content (and, by extension, cellulose biosynthesis and CesA protein localization) was normal in root epidermal primary walls of microtubule organization1 (mor1-1) mutant seedlings. The mor1-1 mutation is a temperature-sensitive allele that causes the disassembly of cortical microtubules (Whittington et al., 2001). The results of Sugimoto et al. (2001) support the hypothesis that cellulose microfibrils can self-align in the absence of normal microtubule organization.

Several possible explanations might account for this discrepancy. First, work with inhibitors is fraught with difficulties, and conclusions based on inhibitor experiments ultimately must be confirmed by other means, such as genetic analysis. A similar discrepancy was noted between the inhibitor experiments of Fisher and Cyr (1998) and the genetic analysis of Gardiner et al. with respect to the putative requirement for cellulose biosynthesis in the maintenance of microtubule organization (see above). In both of these cases, the genetic analysis likely carries more weight than the work based on inhibitors. A second possibility is that a requirement of microtubule organization for CesA localization and function might be one of degree; treatment with oryzalin causes the complete depolymerization of microtubules, whereas the mor1-1 mutation results in an abnormal organization but not complete disassembly (Whittington et al., 2001). Finally, there could be differences in primary and secondary cell walls (i.e., differences among the CesA subunits specific to each wall type) with respect to a requirement for microtubule organization. In this regard, it would be interesting to test the localization of the IRX/CesA proteins in developing xylem of mor1-1 mutant plants.

Interestingly, Sugimoto et al. (2001) further suggested that RSW1/CesA1 and MOR1 might function in independent pathways. This conclusion was based on the observation that mor1 rsw1 double mutant plants exhibited an additive phenotype compared with either single mutant, and these mutations would not be expected to produce an additive phenotype if they operated in the same pathway. Sugimoto et al. (2001) also observed little difference in cellulose content between wild-type and mor1 seedlings and concluded that cellulose synthesis was unaffected in the mutant. However, mor1-1 and rsw1-1 are both temperature-sensitive mutants, indicating that they are weak alleles rather than null mutations. In this case, a mor1 rsw1 double mutant could produce an additive effect on phenotype, even if both genes operate in the same or interdependent pathways. The work of Gardiner et al. supports the hypothesis that CesA function depends on microtubule organization, which also suggests dependence on MOR1 activity. An important test would be to compare cellulose content in the mor1 rsw1 double mutant relative to the rsw1 single mutant. If it is lower, this would imply that MOR1 influences cellulose synthesis. No constitutive mutant alleles of MOR1 have been identified, suggesting that null alleles are embryo lethal (Whittington et al., 2001). The use of newer techniques that enable null mutations in putative embryo-lethal genes, such as inducible RNA interference (Guo et al., 2003), might provide answers to some of these questions.

Finally, Gardiner et al. conducted an interesting set of experiments involving confocal microscopy images of developing xylem expressing IRX3-GFP fusion protein taken at 1-min intervals. In these images, regions of bright GFP fluorescence appear to move along the relatively stable background pattern of banded GPF fluorescence that correlates with bands of microtubules in areas of secondary wall deposition. The authors speculate that these dynamic regions of bright fluorescence correspond to specific organelles or specific regions of the endoplasmic reticulum involved in the transport of CesA subunits in and out of regions of secondary wall deposition. This hypothesis might be tested further in similar studies that make use of GFP (and modified GFP for dual marker studies) fused to endoplasmic reticulum markers and other specific cellular marker proteins.