Plant Golgi cell wall synthesis: From genes to enzyme activities

The cell wall provides mechanical support to the plant cells, and the surface wall, which is directly exposed to the extracellular environment, acts as a primary barrier against pathogen attack, mechanical injury, and other environmental stresses. Biophysical properties of the wall in conjunction with the cell turgor pressure determine the rate of cell expansion during primary growth (1). Polysaccharides, the most dominant fraction of the wall, consist of cellulose microfibrils that are embedded in a matrix of hemicellulose and pectin. Matrix polysaccharides in plants are made in the Golgi cisternae and then exported to the wall by exocytosis (2, 3). Cellulose is made at the plasma membrane by a cellulose synthase complex and directly deposited into the cell wall. Whereas the wall chemical composition has been well characterized and many of the enzyme activities for polysaccharide synthases have been biochemically assayed (4, 5), true to its name, the cell wall proved impregnable to the molecular understanding of its synthesis until relatively recently (6). In this issue of PNAS, Liepman et al. (7) present their work on overcoming this barrier by functionally expressing several of the plant genes encoding a wall matrix polysaccharide synthase in a non-plant eukaryotic host, cultured Drosophila cells.

Several of the enzyme activities for the wall polysaccharide synthases, including the one the work of Liepman et al. (7) is based on, were identified nearly half a century ago by the pioneering work of Hassid's group (4, 8–10). Preston used the term “stagnation” to describe the state of research on cell wall structure during the whole of the eighteenth century (11). Whereas the intervening period between the classic work of Hassid's group and others (2, 3, 12) and the isolation of the first plant gene for cellulose synthase (CesA) was not quite like that with regard to the area of cell wall synthesis, the attempts at biochemical purification of polysaccharide synthases with the goal of obtaining protein sequence followed by gene cloning were met with limited success (13–15). Even in bacteria, it was not until 1990 that a breakthrough was achieved when the gene for CesA was isolated (16). A full 6 years elapsed until the isolation of the plant CesA gene, which was identified through EST analysis of developing cotton fibers with the help of conserved motifs that were common to the bacterial CesA and other polymerizing β-glycosyltransferases (6, 17). Annotation of a large number of sequences in the public and private EST databases as being either CesA or CesA-like (Csl) followed once the cotton CesA sequence became available (18). Complete genome sequencing revealed the presence of 29 Csl genes in Arabidopsis and 37 in rice (18–20). When compared to the slow progress made in the previous decades, the genomic approach took less than 4 years, leading to the isolation of the complete gene family (6, 18). Now, the problem is different: What do all of these genes do?

The proteins corresponding to the wall polysaccharide synthases are very low in abundance.

The Csl sequences were postulated to participate in Golgi polysaccharide synthesis based on their similarity to CesA; however, their exact function has remained largely obscure (18, 19). Mutational genetics approach, which proved useful in the identification of several of the CesA genes linked to cellulose formation, was not of much help in associating the biochemical activities with any of the Csl genes despite the fact that mutations in some of them caused a discernible phenotype (20). By functionally expressing some of the Csl genes from Arabidopsis in Drosophila cells, Liepman et al. (7) have bypassed both the biochemical purification and mutational genetics approaches. The main reasons for success were a robust enzyme assay, the availability of the analytical tools to analyze the product made by the expressed genes, and the dearth or absence of interfering, background enzyme activity in the Drosophila cells. In addition to providing evidence for the involvement of specific genes in the formation of a hemicellulosic polysaccharide, β-glucomannan, the work of Liepman et al. (7) enables a system that could prove useful in assigning function to at least some of the remaining genes in this class. Although heterologous systems involving Xenopus oocytes or yeast cells have also been used successfully to test the function of plant genes, the scale of protein production for polysaccharide synthase assays and the lack of those endogenous glycan synthase activities that might interfere with the plant enzymes make the Drosophila cell system better suited to the study of these types of enzymes (21, 22). Being eukaryotic, it comes fully equipped with the endomembrane system for the processing and potential targeting of the polytopic membrane proteins such as polysaccharide synthases to the appropriate subcellular compartments. Furthermore, this system could open up an avenue for reconstituting functional enzyme complexes consisting of multiple proteins.

An enigmatic finding was that the proteins corresponding to the wall polysaccharide synthases are very low in abundance (13, 14). How is it then that such a low level of an enzyme can make a product that constitutes a great majority of the tissue, like galactomannan in guar seeds where it constitutes >90% of the endosperm (23)? Apparently, because the polysaccharide product is accumulated over a period spanning the development of the organ without being turned over, not much enzyme protein may be needed. Another reason could be that these enzymes are activated in vivo by as-yet-unidentified activators. For example, callose synthase is strongly activated by calcium and polyamines, but no such activators except for divalent cations, which are generally required for the activity, are known for the other polysaccharide synthases (24). These types of problems, along with the labile nature of these enzymes, were perhaps the reasons for slow progress in enzyme purification (13, 14, 23, 25). Alternative approaches, such as the one used by Liepman et al. (7), have thus proved more effective in deciphering the function of the Csl genes (23).

Another significant outcome of this work is evidence that at least some of the Csl sequences can be autonomously functional (7, 23). It was long believed that polysaccharide synthases consisted of large complexes. This concept arose partly from the complexity of the cellulose synthase system in plants and partly from biochemical purification studies of the other polysaccharide synthases (13–15). It is still remotely possible, but unlikely, that some other proteins are acquired from the insect cells to make a functional glucomannan synthase complex.

That some of the Csl enzymes might need multiple proteins to form functional complexes is also suggested by the work of Liepman et al. (7). For example, two genes, CslE1 and CslH1, also showed detectable expression in the Drosophila cells, but neither exhibited any of the assayed enzyme activities (7). Was it because of a lack of a proper acceptor in the reaction mixture or of other polypeptides needed to form a functional complex that these expressed proteins did not exhibit any assayable activity? The latter scenario appears more likely because the polysaccharide synthases are able to catalyze the respective reactions in the absence of any added acceptor even upon a high level of purification (14, 20). Now that Liepman et al. (7) have demonstrated that the Csl genes can be functionally expressed in Drosophila cells, it might become possible to identify other proteins that might be needed to form functional polysaccharide synthase complexes.

Interestingly, all three CslA sequences that show different levels of glucomannan synthase activity reported by Liepman et al. (7) are closely related to each other (18, 23). However, even among these, the highly variable levels of enzyme activity raise the question whether they all catalyze the same reaction in vivo as in vitro. This concern warrants some caution in labeling all of the CslA sequences as being involved in glucomannan formation pending the demonstration of activity in at least some of the other, more divergent family members, e.g., CslA10 and CslA15.

Other than the role the matrix polysaccharides play in cell expansion, they have many industrial applications (23, 26). For example, jellies and jams are derived mainly from the pectic fraction. Xyloglucan and glucuronoarabinoxylan are commercial gums. The most commonly used plant gum, galactomannan, is mostly derived from the seeds of guar or carob (23). A major hemicellulosic constituent in the walls of monocot plants, glucuronoarabinoxylan, is known to adversely affect digestibility in monogastric animals, such as poultry and swine; thus, a reduction in its content could have significant agricultural and industrial implications. Likewise, mixed-linked glucan in the wall of barley and rye grains interferes with digestibility in monogastric animals as well as with the process of brewing. Clearly, identification of the genes that make these polysaccharides has enormous potential in industrial applications.

The availability of the complete sets of Csl sequences from Arabidopsis and rice, and now the ability to screen them for specific biochemical functions, constitute a fertile and productive mix for cell wall researchers (7, 18, 19). The main cell wall β-linked glycans that occur as either chains or backbones of branched polysaccharides are glucan, glucomannan, mixed-linked glucan, and xylan (5). In addition, β-linked galactan occurs as a part of pectin. A separate class of genes, referred to as glucan synthase-like, is believed to be responsible for the formation of callose, a β-1,3-linked glucan (27). Mixed-linked glucan is specific to grasses, and the remaining polysaccharides occur in both the monocot and dicot walls, although their relative levels may vary (5, 26).

The occurrence of conserved motifs that are common to the polymerizing β-glycosyltransferases in the Csl sequences implies their involvement in β-glycan formation (17). Arabidopsis and rice each have six different classes of the Csl genes, although there is an indication that the CslD class may be involved in cellulose formation (18, 19, 28). Taking this into account, the number of Csl classes begins to approach the number of β-linked glycans in plant cell walls. In vitro assays for all of the mentioned β-glycan-forming enzymes are available and could be exploited to screen the rest of the Csl genes for function assignment. The efforts of Liepman et al. (7) have helped breach the molecular wall of matrix polysaccharide synthases, and exciting times lay ahead for researchers interested in cell wall biosynthesis.