Molecular Plant Advance Access originally published online on August 5, 2009
Molecular Plant 2009 2(5):840-850; doi:10.1093/mp/ssp056
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Plant Cell Wall Matrix Polysaccharide Biosynthesis
a Crop Genetics Research and Development, Pioneer Hi-Bred International, Inc., A DuPont Company, 7300 NW 62nd Avenue, Johnston, IA 50131, USA
b Department of Biotechnology, Indian Institute of Technology, Roorkee-247667, Uttarakhand, India
1 To whom correspondence should be addressed. E-mail kanwarpal.dhugga{at}pioneer.com, fax +1-515-334-4729, tel. +1-515-270-3951.
 |
Abstract |
|---|
TOP
Abstract
INTRODUCTIONThe wall of an expanding plant cell consists primarily of cellulose microfibrils embedded in a matrix of hemicellulosic and pectic polysaccharides along with small amounts of structural and enzymatic proteins. Matrix polysaccharides are synthesized in the Golgi and exported to the cell wall by exocytosis, where they intercalate among cellulose microfibrils, which are made at the plasma membrane and directly deposited into the cell wall. Involvement of Golgi glucan synthesis in auxin-induced cell expansion has long been recognized; however, only recently have the genes corresponding to glucan synthases been identified. Biochemical purification was unsuccessful because of the labile nature and very low abundance of these enzymes. Mutational genetics also proved fruitless. Expression of candidate genes identified through gene expression profiling or comparative genomics in heterologous systems followed by functional characterization has been relatively successful. Several genes from the cellulose synthase-like (Csl) family have been found to be involved in the synthesis of various hemicellulosic glycans. The usefulness of this approach, however, is limited to those enzymes that probably do not form complexes consisting of unrelated proteins. Nonconventional approaches will continue to incrementally unravel the mechanisms of Golgi polysaccharide biosynthesis.
Key Words: Biofuels • cell wall • Csl • Golgi • hemicellulose • membrane biochemistry
Received for publication May 4, 2009. Accepted for publication July 6, 2009.
|
INTRODUCTION |
|---|
Primary cell wall differs from secondary wall by its ability to yield to turgor pressure, allowing expansion growth to occur (Ray et al., 1972). A growing plant cell wall consists of cellulose microfibrils embedded in a matrix of hemicellulosic and pectic polysaccharides. In addition, small amounts of structural and enzymatic proteins are present. Secondary wall, deposition of which begins as cell expansion approaches cessation, is rich in cellulose and lignin. Cell wall of grasses contains only small amounts of pectin and xyloglucan, both of which are major constituents of primary wall in most non-grass plants (Carpita, 1996). The role of these two polysaccharides is mostly taken up by glucuronoarabinoxylan in grasses (Figure 1).
|
Cellulose is made at the plasma membrane by an enzyme complex and is deposited directly into the cell wall in a directional manner (Somerville, 2006). Another polysaccharide, β-1,3-glucan or callose that is made at the plasma membrane, is generally a wound-response polysaccharide but is also present in specialized cells or cell parts, such as plasmodesmata, sieve plate pores, pollen mother cells, and pollen tubes. Callose is also transiently deposited in the cell plate during cell division. The remaining matrix polysaccharides are first made in the Golgi and then exported to the cell wall by exocytosis, where they intercalate among cellulose microfibrils to form a composite wall (Northcote and Pickett-Heaps, 1966; Ray et al., 1976).
Unlike other areas of plant sciences, progress on molecular understanding of cell wall synthesis has been slow. For example, the first gene involved in wall polysaccharide synthesis, a cellulose synthase (CesA), was cloned in the mid-1990s (Pear et al., 1996; Arioli et al., 1998). Availability of a plant CesA sequence allowed annotation of a large number of genes with relatively low sequence similarity as CesA-like (Csl) (Pear et al., 1996; Richmond and Somerville, 2000). Sequences in the Csl class are further divided into subgroups A–H and J. Groups B and G are specific to non-grass species, whereas groups F, H, and J are found only in grasses (Hazen et al., 2002; Fincher, 2009). Mutational genetics failed to assign function to any of the Csl genes, although it did imply that several non-Csl genes were involved in matrix polysaccharide synthesis (Lerouxel et al., 2006; Zhong et al., 2008).
Pectin biosynthesis has been the topic of several recent reviews (Scheller et al., 2007; Mohnen, 2008). In this review, we focus on the biosynthesis of non-pectic polysaccharides of the wall matrix with particular emphasis on backbone formation. Areas discussed are a brief history of the Golgi polysaccharide synthesis in plants and its relevance to plant growth, polysaccharide backbone synthesis, backbone substitution with other sugars, sidedness of the catalytic site of polysaccharide synthases, mechanism of glycan chain elongation, disparity between the amount of the enzyme protein and the product it makes, and industrial applications of matrix polysaccharide biosynthesis.
|
GOLGI CELL WALL SYNTHESIS AND CELL EXPANSION |
|---|
Cell expansion has been formulated as the difference between turgor pressure and yield threshold of the wall (Ray et al., 1972). At a given turgor, rheological properties of the wall determine whether a cell expands or not. Rearrangement of wall polymers through hydrolysis, transglycosylation, and disruption of hydrogen bonds is believed to loosen the wall, making it amenable to extension under turgor pressure (Ray et al., 1972; Cosgrove, 1999). Synthesis and deposition of new wall polymers, although not required for cell expansion, are necessary to maintain wall thickness and integrity (Ray, 1987).
Release of xylose during auxin-induced cell expansion implied a role for xyloglucan in cell wall alteration, even though this phenomenon was independent of cell expansion itself (Labavitch and Ray, 1974a, 1974b). An auxin-inducible downward shift in molecular mass of pre-existing and newly synthesized xyloglucan molecules was associated with cell expansion (Talbott and Ray, 1992). Molecular mass of xyloglucan first decreased and then increased upon auxin-induced rapid cell expansion, suggesting modulation of wall properties by this polysaccharide (Talbott and Ray, 1992). The increase in molecular mass occurred apparently to strengthen the cell wall so as to slow down the rapid expansion that initially accompanied a reduction in molecular mass of xyloglucan.
In the pioneering era of cell wall synthesis and its role in cell expansion, two separate enzymes that made β-glucan from UDP-glucose were identified (Ray et al., 1969; Van Der Woude et al., 1974; Ray, 1979). One of them, β-1,4-glucan synthase, had a low Km for the substrate and required divalent cations for activity whereas the other, β-1,3-glucan synthase, had a high Km and did not require any cations for its activity but was later found to be strongly activated by trace amounts of calcium (Kauss, 1987). These two activities could be separated on a sucrose density gradient. The first activity in the gradient, which was associated with the lighter Golgi fraction, had a high affinity for UDP-glucose and made β-1,4-glucan. It was named glucan synthase-I (GS-I). GS-II, in contrast, migrated farther into the density gradient, was associated with the relatively denser plasma membrane fraction, had a low affinity for the substrate, and made β-1,3 glucan. GS-II is also referred to as callose synthase. GS-I and GS-II have since been used as biochemical markers for Golgi and plasma membrane, respectively.
The product of GS-I was believed to form the backbone of xyloglucan (Ray, 1980). This enzyme, activity of which was induced by auxin, provided a target to understand cell expansion at the molecular level. Attempts at its purification, however, failed (see next section).
|
IDENTIFICATION OF GOLGI POLYSACCHARIDE SYNTHASES THROUGH BIOCHEMICAL APPROACHES |
|---|
Mutational genetics has been successfully used to assign function to a number of genes related to cellulose formation (Somerville, 2006). Some of these genes, upon being inactivated, impart a constitutive or conditional phenotype that can be scored and mapped. The same approach proved unsuccessful in identifying genes involved in the formation of β-glycans that constitute the backbones of hemicellulosic polysaccharides (Lerouxel et al., 2006). However, before mutant screening in Arabidopsis became routine, various groups attempted biochemical purification and substrate labeling to identify polypeptides for cell wall polysaccharide synthases (Henry and Stone, 1982; Read and Delmer, 1987; Dhugga et al., 1991; Meikle et al., 1991; Buckeridge et al., 1999). Detergent-solubilization of GS-I from the Golgi-enriched fraction followed by partial purification did not yield any specific polypeptide(s) that could be microsequenced for subsequent gene cloning (K.S. Dhugga, unpublished results). In retrospect, the outcome was not surprising, as the polypeptides for these enzymes constitute a very low proportion of the Golgi proteins, which, combined with their labile nature, makes them formidable candidates for biochemical purification (for more detail, see the section Low Abundance of Polypeptides for Polysaccharide Synthases).
The alternative approach, covalent labeling of the proteins with radiolabeled substrate, proved to be relatively promising, at least initially, as it allowed identification of a protein of 
40 kDa that became covalently labeled under GS-I assay conditions (Dhugga et al., 1991). The label turned over upon incubation of the pre-labeled protein with excess unlabeled substrate, suggesting that it had enzymatic activity. This protein, named Reversibly Glycosylated Protein (RGP), was autoglycosylated, as the purified protein could also be labeled with various UDP-sugars (Dhugga et al., 1997). RGP was specifically localized to the trans-Golgi compartment (Figure 2). However, this protein was membrane-peripheral and became glycosylated by a single sugar, suggesting that it was not, by itself, sufficient to catalyze product formation and might associate with other Golgi transmembrane proteins to form a functional enzyme complex (Singh et al., 1995; Dhugga et al., 1997).
|
Determining the exact function of RGP proved to be challenging (Singh et al., 1995; Dhugga et al., 1997; Delgado et al., 1998; Porchia et al., 2002; Drakakaki et al., 2006). Recently, Konishi et al. (2007) have shown it to catalyze the inter-conversion of pyranose and furanose forms of UDP-arabinose (Konishi et al., 2007). Lack of inhibition of arabinose mutase activity upon coincubation with high concentrations (>10 mM) of UDP-glucose suggests that the mutase activity is at least partially independent of glycosylation of RGP (T. Ishii, personal communication). Isolated RGP becomes saturably autoglycosylated at micromolar concentrations by nearly all the UDP-sugars and sub-millimolar amounts of each substrate are sufficient to displace the sugar already on the protein (Dhugga et al., 1997). This leaves open the possibility that RGP may carry out more than one function (Dhugga et al., 1997). Another interesting question to address is how RGP localizes to Golgi after being synthesized in the cytosol, as it lacks an apparent signal sequence that would target it through the endomembrane system (Figure 2). RGP has also been localized to plasmodesmatal channels but its function there also remains unknown (Sagi et al., 2005).
|
IDENTIFICATION OF CANDIDATE GENES AND FUNCTIONAL CHARACTERIZATION |
|---|
Success in identifying a Golgi polysaccharide synthase was achieved through identification of a candidate gene via transcriptional profiling followed by functional expression in a heterologous system (Dhugga et al., 2004). Like starch in cereal grains, Golgi-synthesized cell wall polysaccharides serve as storage carbohydrates in the seeds of certain legumes (Reid, 1993). For example, on a dry mass basis, greater than 90% (w/w) of the endosperm of fenugreek (Trigonella foenum graecum L.) and guar (Cyamopsis tetragnoloba L.) seeds consists of Golgi-synthesized galactomannan (Dhugga et al., 2004). Galactomannan consists of a β-1,4-mannan backbone to which galactosyl residues are attached through an
-1,6 linkage (Figure 1C). Mannan synthase, the enzyme that forms the β-1,4-linked backbone of galactomannan, can be assayed in vitro using membrane particles derived from developing endosperm (Edwards et al., 1989). Like GS-I, it has a high affinity for the substrate, GDP-mannose, and requires divalent cations for activity. Unlike GS-I, however, it is relatively stable on ice (K.S. Dhugga, unpublished data). Yet, attempts at purification of the detergent-solubilized activity were unsuccessful (Dhugga et al., 2004). With the underlying hypothesis that the gene expression of mannan synthase coincided with its activity, several cDNA libraries constructed from developing guar endosperm were subjected to random sequencing. In the resulting expressed sequence tag database, sequences for a particular Csl gene, which belonged to the A group, were most abundant at a developmental stage that corresponded to the peak mannan synthase activity (Dhugga et al., 2004). This gene, named mannan synthase (ManS), was specifically expressed in the endosperm. Soybean somatic embryos are routinely used as surrogates for developing seed to test gene function. Membrane particles prepared from soybean somatic embryos do not incorporate significant amounts of mannose into polymeric form from GDP-mannose, the substrate of ManS, thus providing a system with a clean background to test the function of the candidate ManS gene. Somatic embryos transformed with the guar ManS gene exhibited substantial ManS activity, which was correlated with the level of expression of the gene, demonstrating that the candidate gene indeed coded for the ManS activity (Dhugga et al., 2004). Upon density gradient centrifugation, the ManS activity localized to the Golgi compartment (Figure 3). Even a mammalian Golgi glycosyltransferase was targeted to the correct compartment when expressed in plant cells (Wee et al., 1998).
|
A few members of the CslA group from Arabidopsis were later found to possess (gluco)mannan synthase activity (Liepman et al., 2005). The two genes that exhibited activity were closest to ManS, while the others formed a separate clade each for rice and Arabidopsis (Dhugga et al., 2004). Glucomannan constitutes a minor proportion of the plant cell wall. The fact that a majority of the members within the CslA group did not have (gluco)mannan synthase activity points to functional diversity among closely related genes as well as the pitfalls of relying solely on sequence homology to assign specific functions to genes (Liepman et al., 2005).
A similar approach as for ManS was employed in an attempt to identify a xyloglucan synthase from nasturtium (Tropaeolum majus L.) seeds, which accumulate xyloglucan as a storage polysaccharide (Cocuron et al., 2007). Expression of the candidate nasturtium gene, which clustered with the CslC class, in yeast resulted in β-1,4-glucan deposition in the wall. Xyloglucan synthesis has been reported to require simultaneous actions of glucan synthase and xylosyltransferase enzymes (Hayashi, 1989). Co-expression of a xylosyltransferase along with the candidate CslC gene did not make xyloglucan (Cocuron et al., 2007). It is possible that the two enzymes did not assemble into a xyloglucan synthase complex in yeast; however, another possibility is that CslC actually carries out the GS-I activity, since it can make β-1,4-glucan in the absence of xylosylation (Cocuron et al., 2007).
Although GS-I is stimulated by UDP-xylose in vitro and recent results show that xylose can be added onto an oligomeric acceptor in vitro, its role in xyloglucan formation has been criticized because it can make the backbone in the absence of UDP-xylose (Ray, 1980; Delmer, 1987; Hayashi, 1989; Faik et al., 2002). What is the functional significance of the product of GS-I, then? Adjacent glycosyl residues are rotated by 180° in a β-1,4-glucan chain, forming a flat ribbon. However, the adjacent residues in the chains on the surface of a cellulose microfibril are twisted with respect to each other (Vietor et al., 2002; Jarvis, 2003). One possibility is that the surface glucan chains of the microfibril are not made by the cellulose synthase complex, but by a separate enzyme instead, such as GS-I, and their deposition on a pre-formed microfibril causes the alteration in their structure. To complicate matters, another group has localized CslC protein to the plasma membrane (A. Bacic, personal communication), which would discount its role in GS-I or xyloglucan synthase activities, as they are both localized to Golgi (Ray, 1979, 1980; Hayashi, 1989). The product of CslC, even if made at the plasma membrane, could still coat the cellulose microfibrils along with its ensuing structural alteration.
Quite a different route was taken to determine the function of two groups of genes specific to grasses, CslF and H (Burton et al., 2006; Doblin et al., 2009). Mixed-linked β-glucan (MLG), which accumulates in primary walls and is recycled upon cessation of cell expansion, accumulates terminally in the endosperm wall of cereal grains (Carpita, 1996). MLG content was genetically mapped in barley grains (Han et al., 1995). As the rice genome sequence became available, Burton et al. (2006) took advantage of the synteny among plant genomes in identifying a cluster of rice CslF genes that overlapped with the barley MLG QTL. Expression of rice CslF genes in Arabidopsis, the walls of which do not contain MLG, led to the deposition of this polysaccharide in the cell wall that was detected by an anti-MLG antibody (Burton et al., 2006). Similarly, expression of a barley CslH gene in Arabidopsis resulted in the deposition of MLG in the cell wall, suggesting that CslF and CslH groups of genes are both capable of making MLG (Burton et al., 2006; Doblin et al., 2009). Despite the fact that both these genes were expressed under the control of a constitutive promoter, variable and often patchy distribution of MLG and its accumulation in only some cell types and not others suggest that some type of regulatory control is also exerted by Arabidopsis. It could be an evolutionarily conserved factor among grasses and non-grass species that associates with CslF or CslH to form a functional enzyme and the expression of which determines which cells make MLG.
Because it has not been possible to detect MLG in the Golgi, even though CslF appears to be a Golgi-resident enzyme, an explanation has been advanced that MLG oligomers are made in the Golgi but assembled at the plasma membrane (Fincher, 2009). Golgi vesicles deposit their cargo into the cell wall though exocytosis so the enzyme that assembles oligomers into longer chains is expected to coexist with the oligomers and become active only on vesicle fusion with the plasma membrane. About the only difference in keeping such an enzyme from performing the proposed function would be pH, which is lower by approximately a unit in the apoplast as compared to the Golgi lumen (Kim et al., 1996). Rapid transit of MLG through Golgi could also contribute to the difficulty of detecting it in this sub-cellular compartment.
Candidates for xylan synthase, feruloyl transferase, and arabinosyl transferase were proposed based on a bioinformatics approach whereby gene expression patterns of rice and Arabidopsis were compared across a number of tissue expression libraries (Mitchell et al., 2007). However, functional characterization of these genes remains to be done. The xylan synthase candidate does not possess any membrane domain whereas all the β-glycan synthases identified thus far are known to be polytopic membrane proteins (Dhugga et al., 2004; Burton et al., 2006; Cocuron et al., 2007; Doblin et al., 2009). A number of other groups have used mutational genetics to identify genes encoding putative non-processive glycosyltransferases as affecting xylan content (Pena et al., 2007; Persson et al., 2007; York and O'Neill, 2008; Fincher, 2009). The proteins encoded by these genes may form complexes with other proteins, including a processive xylan synthase, to make a functional enzyme complex (York and O'Neill, 2008).
All the β-glycan chain-forming enzymes identified thus far belong to the Csl family (Dhugga et al., 2004; Liepman et al., 2005; Burton et al., 2006; Cocuron et al., 2007; Doblin et al., 2009). Given the similarity of the chemical structures of β-1,4-glycans (glucan, xylan, mannan, galactan), it is reasonable to propose that enzymes from the Csl family are involved in the formation of xylan and galactan as well, particularly when the members of the B, D, E, G, and J groups remain to be characterized. Both dicot and monocot walls contain glucuronoarabinoxylan, which would limit candidate genes for xylan formation to CslD and E groups, as the remaining groups are specific to either grasses or non-grass species (Fincher, 2009). It also remains possible that some of the CslA genes participate in the formation of other β-glycans, including xylan, as only a few of them have been shown to actually possess (gluco)mannan synthase activity (Liepman et al., 2005).
|
SIDE CHAIN ADDING ENZYMES |
|---|
Unlike polysaccharide chain-forming enzymes, glycosyltransferases that add sugars onto the backbone maintain their activity after solubilization in detergents. This allowed fractionation and partial purification of two enzymes, galactosyltransferase (GalT) and fucosyltransferase (FucT). Using different approaches for purification, two groups were able to obtain peptide sequences from the purified fractions that were sufficient to allow cloning of the respective genes (Edwards et al., 1999; Perrin et al., 1999). Fucose-deficient mur2 mutant also later on turned out to be defective in FucT (Vanzin et al., 2002). Another success story of mutational genetics is the identification of a GalT (mur3) that transfers a galactosyl residue onto xylose in xyloglucan (Madson et al., 2003).
After the initial GalT sequence was published (Edwards et al., 1999), a number of genes were annotated in the genomic databases as GalT-like. Arabidopsis primary wall, though rich in xyloglucan, does not contain appreciable amounts of galactomannan. This led Faik et al. (2002) to hypothesize that some of these GalT proteins might actually be xylosyltransferases (XylT) that are involved in the formation of xyloglucan. Both GalT and XylT transfer sugar from the substrate onto a glycan backbone in an -1,6-linkage and both require an acceptor of roughly the same size for their in vitro activity. One of these GalT-like proteins catalyzed the transfer of a xylosyl residue to the glucan oligomer, indicating that it was indeed a xylogucan XylT (Faik et al., 2002).
Success has been limited in identifying molecular components responsible for glucuronoarabinoxylan synthesis, although xylose-deficient mutants have provided some insights (Pena et al., 2007; Persson et al., 2007). None of the transferases that add side groups (arabinose, ferulate, glucuronate) onto the xylan chain has yet been definitively identified.
|
SIDEDNESS OF CATALYTIC SITE OF POLYSACCHARIDE SYNTHASES |
|---|
Whether the catalytic sites of cell wall matrix glucan synthases are in the cytosol or in the Golgi lumen is one of the unresolved questions. Examples thus far suggest that the catalytic sites of Golgi polysaccharide synthases face the cytosol, which would make sense evolutionarily, assuming that the chain-forming enzymes originated from the same primordial sequence and the fact that the catalytic site of cellulose synthase faces the cytosol (Urbanowicz et al., 2004; Zeng and Keegstra, 2008). After all, Csl sequences were initially identified because of their homology to CesA (Richmond and Somerville, 2000). Furthermore, pH gradient across the Golgi cisternal membrane with respect to cytosol, although not as steep, is in the same direction as across the plasma membrane where cellulose is made (Kim et al., 1996). Yet another way to look at it is the extrusion of the glycan chain through the membrane.
Most of the models for glycan extrusion through the membrane are for cellulose synthase where transmembrane (TM) helices of each CesA protein are shown forming a pore that is sufficient for extrusion of a glucan chain through the plasma membrane (Delmer, 1999; Doblin et al., 2002; Kurek et al., 2002). Assuming standard dimensions of an alpha helix (outer diameter of 0.8 nm) and a β-1,4-glucan chain (wider and narrower side-to-side dimensions of 1 and 0.5 nm, respectively) (Kondo et al., 2001), and assuming minimum energy arrangement of helices in the membrane, at least eight helices are required to make a pore wide enough (1.2 nm) for the glucan chain to pass through, not allowing for van der Waals radii around H-bond forming groups. However, if the pore is water-filled, as is expected, through which the glucan chain glides, then a larger number of helices would be needed to widen the pore size to accommodate water molecules. Sixteen helices, for example, from a CesA dimer would form a pore with an internal diameter of 3.3 nm, wide enough to allow water molecules around the glucan chain.
Two polypeptides with their catalytic sites juxtaposed to form a functional enzyme unit solve the problem of having to rotate the chain by 180° after each glucosyl residue addition. Nearly all the predicted Csl proteins from Arabidopsis possess enough TM domains (range 5–8) so that, in a dimeric form, they can form a pore wide enough for glucan chain to pass through (http://crombec.botanik.uni-koeln.de/). Taking into consideration only the strongly predicted TM domains, however, it appears that some members of the CslA, B, and C groups do not possess enough TM helices to allow glycan passage. This leaves open the possibility that the catalytic sites of these enzymes are located in the Golgi lumen and those of the other Csl enzymes in the cytosol. However, the aforementioned speculation is based on predicted TM domains; no experimental evidence is yet available for the number of TM domains for any of the plant processive glycosyltransferases.
|
MECHANISM OF CHAIN ELONGATION |
|---|
A generally accepted mechanism of β-glycan chain elongation is that the synthase does not form a covalent intermediate with the substrate, as sugar from NDP-sugar is transferred directly from an
configuration in the substrate to β configuration in the product (Saxena et al., 1995). Results from a cotton fiber cellulose synthase point to a different mechanism, where developing cotton fibers in the presence of a herbicide made only non-crystalline cellulose (Peng et al., 2001). When cellulose synthase trapped in the product was released by a glucanase treatment followed by sequencing of tryptic peptides, the same peptide containing a variable number of glycosyl residues was independently identified multiple times. This suggests that the sugars on this peptide belonged to the glucan chain, which was not completely digested by the glucanase pretreatment that preceded tryptic digestion (Peng et al., 2001). It is unlikely that the sugars, if they were noncovalently attached to the peptide, would have remained attached through SDS gel electrophoresis. Interestingly, this peptide contained an arginyl residue, which had previously been shown to be conserved between RGP and CesA proteins (Figure 4) (Dhugga et al., 1997). These findings suggest that β-glycosyltransferases, at least some of them, carry out glycosyl transfer for chain elongation through covalent intermediate formation. This would entail two covalent intermediates to confer β configuration on the sugar as the sugar goes from an configuration in the substrate, NDP-sugar, to β in the first covalent intermediate, in the second, and, finally, β in the product.
|
Covalent attachment of the glucan chain at all times to one or the other β-glycosyltransferase polypeptide in the dimer would entail growth of the glucan chain towards the reducing end. In other words, the reducing end of the chain, which is covalently attached to the enzyme, is transferred through a second covalent intermediate onto the non-reducing end of the glycosyl residue that is attached to the second enzyme polypeptide. Repetition of this cycle between the two juxtaposed catalytic sites makes the enzyme processive. Bacterial cellulose has been reported to grow towards the non-reducing end, with the reducing end pointing away from the site of synthesis (Koyama et al., 1997). Another report claims that it grows towards the reducing end (Han and Robyt, 1998). Likewise, hyaluronan, a β-glycan, has been reported to grow towards the reducing end in bacteria and the non-reducing end in mammals (Bodevin-Authelet et al., 2005; Tlapak-Simmons et al., 2005; Prehm, 2006). The arginyl residue that becomes glycosylated in RGP is conserved in CesA and many of the Csl proteins, although, occasionally, it is substituted with lysine (Figure 4). Side chain ionizing groups of both free arginine and lysine have alkaline pKa values (12.5 and 10.5, respectively), allowing for a conservative substitution. The high pKa of the available reactive groups on these amino acids poses a problem in them acting as covalent intermediates in polysaccharide formation. The effective pKa in the catalytic pocket must be lower, as suggested by the optimal pH of the reaction (Dhugga et al., 1997).
|
LOW ABUNDANCE OF POLYPEPTIDES FOR POLYSACCHARIDE SYNTHASES |
|---|
Lack of success with biochemical purification of polysaccharide synthases can be attributed to their instability in isolated membranes, recalcitrance to solubilization in active form in detergents, and very low abundance as a percentage of membrane proteins. Even the enzymes that were relatively stable and could be solubilized in active form proved difficult to purify (Dhugga and Ray, 1994). After the gene for ManS was identified, we compared Western blots with Coomassie-stained gels and could not detect any polypeptide corresponding to the band that was recognized by the anti-ManS antibody (K.S. Dhugga, unpublished results).
With the knowledge that ManS constitutes a very low proportion (assume, for example, 0.02%) of the total membrane protein, has a molecular mass of 60 kDa and a specific activity of 100 pmol min–1 mg–1 membrane protein, and assuming that two polypeptides dimerize to make a functional enzyme, the turnover number is 1 s–1. The turnover number is likely higher in vivo, as the in vitro assay possibly excludes some of the activating effectors of ManS. Assuming a conservative two-fold higher in vivo turnover number, over a 2-week period and at a steady state aforementioned concentration of mannan synthase, a product/enzyme mass ratio of 8 x 105 is achieved. In other words, 1 µg of ManS protein can produce 800 mg of product over this period. The product, galactomannan, accumulates terminally, that is, it is not turned over after it is deposited. The same argument would apply to other cell wall polysaccharide synthases, the products of which are deposited irreversibly.
|
APPLICATIONS OF GOLGI POLYSACCHARIDE BIOSYNTHESIS |
|---|
Hemicellulosic polysaccharides constitute a substantial fraction of a mature cell wall across plant species (Carpita, 1996). In maize grains, glucuronoarabinoxylan accounts for a great majority of the fiber fraction (Hazen et al., 2003; X. Bao, personal communication). This polysaccharide cannot be readily digested by monogastric animals. Furthermore, because it is a gum, it increases the viscosity of the chyme, thereby adversely affecting the digestibility of grain-derived feed. The residue left over from the grain after ethanol distillation is referred to as distillers dried grains with solubles (DDGS). DDGS, which is rich in proteins and glucuronoarabinoxylan, is sold as a feed supplement to the bovine, swine, and poultry industries. Its usefulness in monogastrics (poultry and swine) is limited because of the aforementioned reasons. A reduction in glucuronoarabinoxylan content is thus desirable in increasing the value of DDGS (Dhugga, 2007). Glucuronate, arabinosyl, and acetyl side groups, by steric hindrances, keep the xylan chains of glucuronoarabinoxylan from associating with each other, reducing its crystallinity and keeping it in soluble form. Long stretches of un-substituted xylan would increase the crystallinity of glucuronoarabinoxylan with consequent reduction in its solubility, which might improve the digestibility of DDGS in monogastrics. Acetate is an inhibitor of fermentation and any reduction in its concentration will help improve the quality of biomass for biofuels (Dhugga, 2007). An obvious question is whether these types of changes in wall composition can be tolerated by the plant without undesirable pleiotropic effects. Screening of natural germplasm could provide an indication as to what kind of range there is in nature for any trait of interest. Biotechnology can then be employed to speed up the progress in fixing the desirable variation.
Feruloyl groups in glucuronoarabinoxylan are believed to act as nucleating sites for lignin formation as well as in cross-linking of the wall (Iiyama et al., 1994). Feruloyl transferase remains unknown, as do the transferases for arabinosyl and glucuronosyl sugars, which add respective sugars to the xylan backbone. Once these genes are identified, an obvious target would be direct interference with feruloyl transferase to reduce wall cross-linking. Since the feruloyl group is attached to the arabinosyl residue, a reduction in the content of this sugar in glucuronoarabinoxylan is also expected to lead to the same result. Cellulosic biomass is pretreated by various methods to facilitate its conversion into sugars by hydrolytic enzymes for ethanol production. A reduction in cross-linking could help in cost savings by reducing pretreatment costs. Reduced cross-linking will also enhance digestibility of the silage. A reduction in lignin content through manipulation of monolignol biosynthetic pathway generally has undesirable pleiotropic effects, such as increased lodging, and insect and disease susceptibility (Pedersen et al., 2005). Maintaining lignin content but reducing wall cross-linking might aid in overcoming these negative effects.
MLG, which accumulates terminally in the endosperm walls of small grain cereals, also adversely affects digestibility in monogastric animals and is a cause of beer haze (Carpita, 1996). Conversely, MLG is a natural gum and thus a natural laxative, so it is useful for human health, market size notwithstanding. Identification of the genes involved in its formation has made it possible to explore the alteration of its content and/or structure in the wall for industrial and agronomic applications (Burton et al., 2006; Doblin et al., 2009).
Galactomannan is an industrial gum, the physicochemical properties of which are determined by galactose/mannose ratio, which could be controlled by just two genes, GalT and ManS (Reid, 1993; Reid et al., 2003; Dhugga et al., 2004). ManS is active by itself, that is, it does not require other proteins for the formation of a functional complex (Dhugga et al., 2004; Liepman et al., 2005). This makes it possible to engineer commercial gums with a whole variety of properties by manipulating only two genes, ManS and GalT, not only in the native species, but also in other commercial crops (Dhugga et al., 2004).
|
SUMMARY |
|---|
Much progress has been made in identifying polysaccharide synthases involved in the formation of hemicellulosic polysaccharides in the last decade. Whereas all the genes for the side group-adding enzymes for xyloglucan have been identified, those for the formation of glucuronoarabinoxylan remain elusive. Csl genes have been shown to be involved in the formation of (gluco)mannan (CslA), β-1,4-glucan (CslC), and mixed-linked glucan (CslF and CslH). Functions of the remaining classes of Csl (B, D, E, G, and some members of A) remain to be determined. Although xylan backbone is identical in linkage to cellulose and mannan, non-Csl mechanisms have been proposed for its synthesis. The possibility remains, however, that a Csl protein, likely from the A, D, or E group, actually catalyzes chain formation in association with other protein(s). Why some polysaccharides, such as galactomannan, can be synthesized by just two enzymes and glucuronoarabinoxylan requires a complex is an interesting question to address in terms of both biochemistry and evolution. Whether β-galactan in pectin is also synthesized by one of the Csl proteins remains to be determined. Sidedness of catalytic sites of various hemicellulosic polysaccharide synthases in the Golgi apparatus remains to be determined, although initial indications are that it is in the cytosol, at least for some of them. Discussion on the mechanism of catalysis of β-glycan chain formation is speculative at this stage, mainly because of lack of crystal structure of any of the underlying enzymes. Some of the enzymes, such as mannan synthase, have a simple structure and maintain activity upon solubilization in detergents, thus offering opportunities for structural chemistry. Regardless, recent progress has made it possible to explore molecular mechanisms of polysaccharide synthesis in plants and to alter wall composition for industrial applications.
-
Arioli T, et al. Molecular analysis of cellulose biosynthesis in Arabidopsis. Science (1998) 279:717–720.
Bodevin-Authelet S, Kusche-Gullberg M, Pummill PE, DeAngelis PL, Lindahl U. Biosynthesis of hyaluronan: direction of chain elongation. J. Biol. Chem. (2005) 280:8813–8818.
Buckeridge MS, Vergara CE, Carpita NC. The mechanism of synthesis of a mixed-linkage (1,3), (1,4)β-D-glucan in maize: evidence for multiple sites of glucosyl transfer in the synthase complex. Plant Physiol (1999) 120:1105–1116.
Burton RA, Wilson SM, Hrmova M, Harvey AJ, Shirley NJ, Medhurst A, Stone BA, Newbigin EJ, Bacic A, Fincher GB. Cellulose synthase-like CslF genes mediate the synthesis of cell wall (1,3;1,4)-β-D-glucans. Science (2006) 311:1940–1942.
Carpita NC. Structure and biogenesis of the cell walls of grasses. Ann. Rev. Plant Physiol. Plant Mol. Biol. (1996) 47:445–476.[CrossRef]
Cocuron JC, Lerouxel O, Drakakaki G, Alonso AP, Liepman AH, Keegstra K, Raikhel N, Wilkerson CG. A gene from the cellulose synthase-like C family encodes a β-1,4 glucan synthase. Proc. Natl Acad. Sci. U S A (2007) 104:8550–8555.
Cosgrove DJ. Enzymes and other agents that enhance cell wall extensibility. Ann. Rev. Plant Biol. (1999) 50:391–417.[CrossRef]
Delgado IJ, Wang Z, De RA, Keegstra K, Raikhel N-V. Cloning and characterization of AtRGP1: a reversibly autoglycosylated Arabidopsis protein implicated in cell wall biosynthesis. Plant Physiol (1998) 116:1339–1349.
Delmer D. Cellulose biosynthesis: exciting times for a difficult field of study. Ann. Rev. Plant Physiol. Plant Mol. Biol. (1999) 50:245–276.[CrossRef][Web of Science]
Delmer DP. Cellulose biosynthesis. Ann. Rev. Plant Physiol (1987) 38:259–290.[CrossRef][Web of Science]
Dhugga KS. Maize biomass yield and composition for biofuels. Crop Sci. (2007) 47:2211–2227.
Dhugga KS, Ray PM. Purification of 1,3-β-glucan synthase activity from pea tissue: two polypeptides of 55 kDa and 70 kDa copurify with enzyme activity. Eur. J. Biochem (1994) 220:943–953.[Web of Science][Medline]
Dhugga KS, et al. Guar seed β-mannan synthase is a member of the cellulose synthase super gene family. Science (2004) 303:363–366.
Dhugga KS, Tiwari SC, Ray PM. A reversibly glycosylated polypeptide (RGP1) possibly involved in plant cell wall synthesis: purification, gene cloning and trans-Golgi localization. Proc. Natl Acad. Sci. U S A (1997) 94:7679–7684.
Dhugga KS, Ulvskov P, Gallagher SR, Ray PM. Plant polypeptides reversibly glycosylated by UDP-glucose: possible components of Golgi β-glucan synthase in pea cells. J. Biol. Chem. (1991) 266:21977–21984.
Doblin M, Pettolino F, Wilson S, Campbell R, Burton R, Fincher G, Newbigin E, Bacic A. A barley cellulose synthase-like CSLH gene mediates (1,3;1,4)-β -D-glucan synthesis in transgenic Arabidopsis. Proc. Natl Acad. Sci. U S A (2009) 106:5996–6001.
Doblin MS, Kurek I, Jacob WD, Delmer DP. Cellulose biosynthesis in plants: from genes to rosettes. Plant Cell Physiol (2002) 43:1407–1420.
Drakakaki G, Zabotina O, Delgado I, Robert S, Keegstra K, Raikhel N. Arabidopsis reversibly glycosylated polypeptides 1 and 2 are essential for pollen development. Plant Physiol (2006) 142:1480–1492.
Edwards M, Bulpin PV, Dea ICM, Reid JSG. Biosynthesis of legume-seed galactomannans in vitro: cooperative interactions of a guanosine 5'-diphosphate mannose-linked (1,4)-β-D-mannosyltransferase and a uridine 5'-diphosphate-galactose-linked -D-galactosyltransferase in particulate enzyme preparations from developing endosperms of fenugreek (Trigonella foenum-graecum L.) and guar (Cyamopsis tetragonoloba L. Planta (1989) 178:41–51.[CrossRef][Web of Science]
Edwards ME, Dickson CA, Chengappa S, Sidebottom C, Gidley MJ, Reid JSG. Molecular characterisation of a membrane-bound galactosyltransferase of plant cell wall matrix polysaccharide biosynthesis. Plant J (1999) 19:691–697.[CrossRef][Web of Science][Medline]
Faik A, Price NJ, Raikhel NV, Keegstra K. An Arabidopsis gene encoding an alpha-xylosyltransferase involved in xyloglucan biosynthesis. Proc. Natl Acad. Sci. U S A (2002) 99:7797–7802.
Fincher GB. Revolutionary times in our understanding of cell wall biosynthesis and remodeling in the grasses. Plant Physiol (2009) 149:27–37.
Han F, et al. Mapping of β-glucan content and β-glucanase activity loci in barley grain and malt. Theor. Appl. Genet. (1995) 91:921–927.[Web of Science]
Han NS, Robyt JF. The mechanism of Acetobacter xylinum cellulose biosynthesis: direction of chain elongation and the role of lipid pyrophosphate intermediates in the cell membrane. Carbohydr. Res. (1998) 313:125–133.[CrossRef][Web of Science]
Hayashi T. Xyloglucans in the primary cell wall. Ann. Rev. Plant Physiol. Plant Mol. Biol. (1989) 40:139–168.[CrossRef][Web of Science]
Hazen SP, Hawley RM, Davis GL, Henrissat B, Walton JD. Quantitative trait loci and comparative genomics of cereal cell wall composition. Plant Physiol (2003) 132:263–271.
Hazen SP, Scott CJS, Walton JD. Cellulose synthase-like genes of rice. Plant Physiol (2002) 128:336–340.
Henry RJ, Stone BA. Solubilization of β-glucan synthases from the membranes of cultured ryegrass endosperm cells. Biochem. J. (1982) 203:629–636.[Web of Science][Medline]
Iiyama K, Lam Thi Bach T, Stone BA. Covalent cross-links in the cell wall. Plant Physiol (1994) 104:315–320.[Web of Science][Medline]
Jarvis M. Cellulose stacks up. Nature (2003) 426:611–612.[CrossRef][Medline]
Kauss H. Some aspects of calcium-dependent regulation in plant metabolism. Annu. Rev. Plant Physiol (1987) 38:47–72.[CrossRef][Web of Science]
Kim JH, Lingwood CA, Williams DB, Furuya W, Manolson MF, Grinstein S. Dynamic measurement of the pH of the Golgi complex in living cells using retrograde transport of the verotoxin receptor. J. Cell Biol. (1996) 134:1387–1399.
Kondo T, Togawa E, Brown RM Jr. ‘Nematic ordered cellulose’: a concept of glucan chain association. Biomacromol (2001) 2:1324–1330.[CrossRef][Web of Science][Medline]
Konishi T, Takeda T, Miyazaki Y, Ohnishi-Kameyama M, Hayashi T, O'Neill MA, Ishii T. A plant mutase that interconverts UDP-arabinofuranose and UDP-arabinopyranose. Glycobiol. (2007) 17:345–354.
Koyama M, Helbert W, Imai T, Sugiyama J, Henrissat B. Parallel-up structure evidences the molecular directionality during biosynthesis of bacterial cellulose. Proc. Natl Acad. Sci. U S A (1997) 94:9091–9095.
Kurek I, Kawagoe Y, Jacob WD, Doblin M, Delmer D. Dimerization of cotton fiber cellulose synthase catalytic subunits occurs via oxidation of the zinc-binding domains. Proc. Natl Acad. Sci. U S A (2002) 99:11109–11114.
Labavitch JM, Ray PM. Relationship between promotion of xyloglucan metabolism and induction of elongation by indoleacetic acid. Plant Physiol (1974a) 54:499–502.
Labavitch JM, Ray PM. Turnover of cell wall polysaccharides in elongating pea stem segments. Plant Physiol (1974b) 53:669–673.
Lerouxel O, Cavalier DM, Liepman AH, Keegstra K. Biosynthesis of plant cell wall polysaccharides: a complex process. Curr. Opin. Plant Biol. (2006) 9:621–630.[CrossRef][Web of Science][Medline]
Liepman AH, Wilkerson C, Keegstra K. Expression of cellulose synthase-like (Csl) genes in insect cells reveals that CslA family members encode mannan synthases. Proc. Natl Acad. Sci. U S A (2005) 102:2221–2226.
Madson M, Dunand C, Li X, Verma R, Vanzin GF, Caplan J, Shoue DA, Carpita NC, Reiter WD. The MUR3 gene of Arabidopsis encodes a xyloglucan galactosyltransferase that is evolutionarily related to animal exostosins. Plant Cell. (2003) 15:1662–1670.
Meikle PJ, Ng KF, Johnson E, Hoogenraad NJ, Stone BA. The β-glucan synthase from Lolium multiflorum: detergent solubilization, purification using monoclonal antibodies, and photoaffinity labeling with a novel photoreactive pyrimidine analogue of uridine 5'-diphosphoglucose. J. Biol. Chem. (1991) 266:22569–22581.
Mitchell RA, Dupree P, Shewry PR. A novel bioinformatic approach identifies candidate genes for the synthesis and feruloylation of arabinoxylan. Plant Physiol (2007) 144:43–53.
Mohnen D. Pectin structure and biosynthesis. Curr. Opin. Plant Biol. (2008) 11:266–277.[CrossRef][Web of Science][Medline]
Northcote DH, Pickett-Heaps JD. A function of the Golgi apparatus in polysaccharide synthesis and transport in the root-cap cells of wheat. Biochem. J. (1966) 98:159–167.[Web of Science][Medline]
Pear JR, Kawagoe Y, Schreckengost WE, Delmer DP, Stalker DM. Higher plants contain homologs of the bacterial celA genes encoding the catalytic subunit of cellulose synthase. Proc. Natl Acad. Sci. U S A (1996) 93:12637–12642.
Pedersen JF, Vogel KP, Funnell DL. Impact of reduced lignin on plant fitness. Crop Sci. (2005) 45:812–819.
Pena MJ, Zhong R, Richardson EA, O'Neill MA, Darvill AG, York WS, Ye ZH. Arabidopsis irregular xylem8 and irregular xylem9: implications for the complexity of glucuronoxylan biosynthesis. Plant Cell. (2007) 19:549–563.
Peng L, Xiang F, Roberts E, Kawagoe Y, Greve LC, Kreuz K, Delmer DP. The experimental herbicide CGA 325'615 inhibits synthesis of crystalline cellulose and causes accumulation of non-crystalline β-1,4-glucan associated with CesA protein. Plant Physiol (2001) 126:981–992.
Perrin RM, DeRocher AE, Bar-Peled M, Zeng W, Norambuena L, Orellana A, Raikhel NV, Keegstra K. Xyloglucan fucosyltransferase, an enzyme involved in plant cell wall biosynthesis. Science (1999) 284:1976–1979.
Persson S, Caffall KH, Freshour G, Hilley MT, Bauer S, Poindexter P, Hahn MG, Mohnen D, Somerville CR. The Arabidopsis irregular xylem8 mutant is deficient in glucuronoxylan and homogalacturonan, which are essential for secondary cell wall integrity. Plant Cell. (2007) 19:237–255.
Porchia AC, Sorensen SO, Scheller HV. Arabinoxylan biosynthesis in wheat: characterization of arabinosyltransferase activity in Golgi membranes. Plant Physiol (2002) 130:432–441.
Prehm P. Biosynthesis of hyaluronan: direction of chain elongation. Biochem. J. (2006) 398:469–473.[CrossRef][Web of Science][Medline]
Ray PM. Maize coleoptile cellular membranes bearing different types of glucan synthetase activity. In: Plant Organelles—Reid E, ed. (1979) Chichester, UK: Halsted Press/John Wiley & Sons. 135–146.
Ray PM. Cooperative action of β-glucan synthetase and UDP-xylose xylosyl transferase of Golgi membranes in the synthesis of xyloglucan-like polysaccharide. Biochim. Biophys. Acta (1980) 629:431–444.[Medline]
Ray PM. Principles of plant cell expansion. In: Physiology of Cell Expansion during Plant Growth—Cosgrove DJ, Knievel DP, eds. (1987) Rockville: American Society of Plant Physiologists. 1–17.
Ray PM, Eisinger WR, Robinson DG. Organelles involved in cell wall polysaccharide formation and transport in pea cells. Ber. Deutsch. Bot. Ges. Bd (1976) 89:121–146.
Ray PM, Green PB, Cleland R. Role of turgor in plant cell growth. Nature (1972) 239:163–&.[CrossRef][Web of Science]
Ray PM, Shininger TL, Ray MM. Isolation of β-glucan synthetase particles from plant cells and identification with Golgi membranes. Proc. Natl Acad. Sci. U S A (1969) 64:605–612.
Read SM, Delmer DP. Inhibition of mung bean UDP-glucose: (1,3)-β-glucan synthase by UDP-pyridoxal: evidence of an active-site amino group. Plant Physiol (1987) 85:1008–1015.
Reid JSG. Seed storage polysaccharides models for the study of cell wall matrix biosynthesis and turnover. J. Cell Biochem. Suppl. 17 (1993) 9–15.
Reid JSG, Edwards ME, Dickson CA, Scott C, Gidley MJ. Tobacco transgenic lines that express fenugreek galactomannan galactosyltransferase constitutively have structurally altered galactomannans in their seed endosperm cell walls. Plant Physiol (2003) 131:1487–1495.
Richmond TA, Somerville CR. The cellulose synthase superfamily. Plant Physiol (2000) 124:495–498.
Sagi G, Katz A, Guenoune-Gelbart D, Epel BL. Class 1 reversibly glycosylated polypeptides are plasmodesmal-associated proteins delivered to plasmodesmata via the Golgi apparatus. Plant Cell. (2005) 17:1788–1800.
Saxena IM, Brown RM, Fevre M, Geremia RA, Henrissat B. Multidomain architecture of β-glycosyl transferases: implications for mechanism of action. J. Bacteriol (1995) 177:1419–1424.
Scheller HV, Jensen JK, Sørensen SO, Harholt J, Geshi N. Biosynthesis of pectin. Physiol. Plant (2007) 129:283–295.[CrossRef]
Singh DG, Lomako J, Lomako WM, Whelan WJ, Meyer HE, Serwe M, Metzger JW. β-Glucosylarginine: a new glucose-protein bond in a self-glucosylating protein from sweet corn. FEBS Lett. (1995) 376:61–64.[CrossRef][Web of Science][Medline]
Somerville C. Cellulose synthesis in higher plants. Ann. Rev. Cell Dev. Biol. (2006) 22:53–78.[CrossRef][Web of Science][Medline]
Talbott LD, Ray PM. Changes in molecular size of previously deposited and newly synthesized pea cell wall matrix polysaccharides: effects of auxin and turgor. Plant Physiol (1992) 98:369–379.
Tlapak-Simmons VL, Baron CA, Gotschall R, Haque D, Canfield WM, Weigel PH. Hyaluronan biosynthesis by class I streptococcal hyaluronan synthases occurs at the reducing end. J. Biol. Chem. (2005) 280:13012–13018.
Urbanowicz BR, Rayon C, Carpita NC. Topology of the maize mixed linkage (1,3), (1,4)-β-D-glucan synthase at the Golgi membrane. Plant Physiol (2004) 134:758–768.
Van Der Woude WJ, Lembi CA, Morre DJ, Kindinger JI, Ordin L. β-Glucan synthetases of plasma membrane and Golgi apparatus from onion stem. Plant Physiol (1974) 54:333–340.
Vanzin GF, Madson M, Carpita NC, Raikhel NV, Keegstra K, Reiter WD. The mur2 mutant of Arabidopsis thaliana lacks fucosylated xyloglucan because of a lesion in fucosyltransferase AtFUT1. Proc. Natl Acad. Sci. U S A (2002) 99:3340–3345.
Vietor RJ, Newman RH, Ha MA, Apperley DC, Jarvis MC. Conformational features of crystal-surface cellulose from higher plants. Plant J (2002) 30:721–731.[CrossRef][Web of Science][Medline]
Wee EGT, Sherrier DJ, Prime TA, Dupree P. Targeting of active sialyltransferase to the plant Golgi apparatus. Plant Cell. (1998) 10:1759–1768.
York WS, O'Neill MA. Biochemical control of xylan biosynthesis: which end is up? Curr. Opin. Plant Biol. (2008) 11:258–265.[CrossRef][Web of Science][Medline]
Zeng W, Keegstra K. AtCSLD2 is an integral Golgi membrane protein with its N-terminus facing the cytosol. Planta (2008) 228:823–838.[CrossRef][Web of Science][Medline]
Zhong R, Lee C, Zhou J, McCarthy RL, Ye ZH. A battery of transcription factors involved in the regulation of secondary cell wall biosynthesis in Arabidopsis. Plant Cell. (2008) 20:2763–2782.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||