In Vitro Synthesis of Cellulose from Plant Extracts

The β-1,3-Glucan Synthase and Lessons from in vitro β-1,3-Glucan Synthesis
To understand the biochemical machinery required for cellulose synthesis in plants, it is necessary to demonstrate in vitro synthesis of cellulose using plant extracts. Unfortunately, much to the dismay of most researchers studying cellulose biosynthesis, the major in vitro polysaccharide product synthesized from plant extracts using UDP-glucose as the precursor was and is still found to be callose, the β-1,3-glucan first reported from mung bean extracts by Feingold and colleagues in 1958 (Feingold et al., 1958). Observing the synthesis of this polysaccharide in place of cellulose has been both frustrating and invigorating as it brings up a number of very interesting questions, many of which have not been fully answered.

During normal development, cellulose is found in all plant cells, whereas callose generally is synthesized in response to wounding, physiological stress, or infection, and is a component of the cell plate in dividing cells apart from being present in specialized cells. As such, enzymes for synthesis of this polysaccharide are not expected to be active most of the time. The general explanation to account for the large amount of in vitro synthesis of callose as opposed to cellulose using plant extracts is that this occurs in response to the wounding or stress of the cells during cell breakage. Using antibodies against β-1,4-glucan synthase and β-1,3-glucan synthase, Nakashima et al. (2003) recently demonstrated that the activation of β-1,3-glucan synthase upon wounding may be dependent on the degradation of β-1,4-glucan synthases by specific proteases (Nakashima et al., 2003). However, under appropriate conditions in the presence of UDP-glucose, plant extracts synthesize both callose and cellulose, and the optimal conditions required for synthesis of these two polysaccharides have been shown to be only slightly different. Whether the same enzyme catalyzes the synthesis of both callose and cellulose has been debated for a number of years, but so far no conclusive evidence is available in support of either the one enzyme-two polysaccharides or the one enzyme-one polysaccharide synthesis with respect to these two polysaccharides. Although it has been possible to separate the major cellulose-synthesizing and callose synthesizing activities by native gel electrophoresis, the polypeptide composition in these two fractions could not be completely analyzed (Kudlicka and Brown, 1997). Interestingly, relatively much more is known about the identity of the catalytic subunit of cellulose synthase as compared to the nature of the catalytic subunit of callose synthase. This is true, in spite of the fact that genes required for synthesis of β-1,3-glucans have been identified in yeast, and similar genes have been identified in a number of plants (Cui et al., 2001; Doblin et al., 2001; Hong et al., 2001; Li et al., 2003). Surprisingly, the proteins encoded by these genes do not show similarity to any known glycosyltransferase, much less the cellulose synthases. These proteins are classified as 1,3-β-D-glucan synthases and have been placed in family 48 of glycosyltransferases (http://afmb. cnrs-mrs.fr/CAZY/). In plants, genes encoding this protein form a gene family, and in Arabidopsis 10 members are identified in this gene family.

Since synthesis of β-1,3-glucans occurs much more readily when plant extracts are used in vitro, many more studies have reported on characterization of the conditions for β-1,3-glucan synthase activity and its purification from a variety of plants. As an example, optimal conditions for in vitro synthesis of β-1,3 glucan from Arabidopsis were defined by the presence in the reaction mixture of 50 mM 3-(N-morpholino) propanesulfonic acid (MOPS) buffer, pH 6.8, 1 mM UDPglucose, 8 mM Ca2+, and 20 mM cellobiose (Lai-Kee-Him et al., 2001). Similar conditions, in the presence or absence of Mg2+ in the reaction mix, have also been shown to be optimal for the synthesis of cellulose using plant extracts (Colombani et al., 2004). Since both callose synthase and cellulose synthase are membrane proteins, the choice and concentration of detergents used during extraction of the proteins have been found to be very crucial in obtaining high specific activity of both callose synthase and cellulose synthase from plant extracts. Incorporating a variety of techniques, Dhugga and Ray (1994) purified the β-1,3-glucan synthase activity 5,500-fold from pea homogenates and found two polypeptides that copurified with the enzyme activity (Dhugga and Ray, 1994). Unfortunately, the identity of these proteins could not be determined, although one of these polypeptides was shown to bind to UDP-glucose. In related sets of experiments, Kudlicka and Brown (1997) demonstrated separation of the callose synthase and cellulose synthase activities in digitonin-solubilized mung bean membranes using gel electrophoresis under nondenaturing conditions (Kudlicka and Brown, 1997). The polypeptide composition in the two fractions was analyzed by SDS-PAGE, and while three similar sized polypeptides were observed in both activities, polypeptides unique to each activity were also observed. However, the characterization of these polypeptides did not provide any further information regarding the similarities or differences between the two enzyme activities. As mentioned in this section, many of the studies for in vitro synthesis of callose were applicable to in vitro synthesis of cellulose using plant extracts. Interestingly, conclusive demonstrations of cellulose synthesis in vitro using plant extracts had to do more with utilizing a greater variety of techniques for product characterization than with development of novel assay methods.

Increasing Cellulose Synthase Activity in vitro and Utilizing More Techniques for Product Characterization
Techniques to identify and characterize the cellulose product have played a crucial role in determining cellulose synthesis in vitro. Interestingly, many of the criteria used by Glaser in 1958 for in vitro cellulose production using bacterial extracts are still used for characterizing the cellulose product and determining the cellulose synthase activity, namely incorporation of 14C-glucose from UDP-14C-glucose into a hot alkali-insoluble fraction (Glaser, 1958). The product was further characterized by acid hydrolysis and/or enzymatic analysis using cellulases. Although less than 1% of the glucose fromUDP-glucose was incorporated into the alkali-insoluble fraction in the in vitro reaction, the product was characterized as cellulose.

A major breakthrough in understanding cellulose biosynthesis in A. xylinum and increasing cellulose synthase activity in bacterial extracts came with the identification of cyclic di-guanosine monophosphate (c-di-GMP) as an allosteric activator of cellulose synthase (Ross et al., 1986). This nucleotide is now recognized to be a regulator of many more bacterial functions than previously thought (D’Argenio and Miller, 2004). The addition of c-di-GMP in reaction mixtures using bacterial extracts led to a remarkable increase in incorporation of glucose from UDP-glucose into a cellulose product.

In another development, the in vitro product using bacterial extracts for the first time was visualized by electron microscopy, and this product was shown to bind to gold-labeled cellobiohydrolase providing evidence that this product is cellulose (Lin et al., 1985). The in vitro product obtained using A. xylinum inner membrane was furthermore shown to be cellulose II (Bureau and Brown, 1987). The capability to synthesize large amounts of the in vitro product was crucial in performing X-ray diffraction, sugar analysis, linkage analysis and molecular weight analysis to demonstrate conclusively that the product was cellulose (Bureau and Brown, 1987).

Many of these techniques were later utilized by Okuda et al. (1993) using cotton fiber extracts to demonstrate the in vitro production of cellulose II (Okuda et al., 1993). Additionally, the incorporation of glucose from UDP-glucose into an Updegraff reagent-resistant fraction was included to be a stricter criterion for the cellulose product. Although no activator comparable to c-di-GMP was identified for activation of the cellulose synthase fromplant tissues, a number of nucleotides were found to increase the in vitro cellulose synthase activity (Li and Brown, 1993). Overall, the success in demonstrating cellulose synthesis in vitro is ascribed to the choice of plant tissue (cotton fibers), method of extraction, and the ability to synthesize large amounts of the in vitro product for characterization. Although cellulose was synthesized in vitro using plant extracts, the major product was still β-1,3 glucan, and this could be distinguished from cellulose using electron microscopy.

In later studies, using a variety of detergents, Kudlicka et al. (1995) was able to demonstrate not only an increase in the amount of cellulose synthesized in vitro, but also the production of cellulose I using plant extracts (Kudlicka et al., 1995). Lai-Kee-Him et al. (2002) used detergent solubilized microsomal fractions from suspension-cultured cells of blackberry (Rubus fruticosus) for in vitro cellulose synthesis (Lai-Kee-Him et al., 2002). These investigators found that the detergents Brij 58 and taurocholate were effective in solubilizing membrane proteins that allowed synthesis of both cellulose and callose given UDP-glucose as the substrate. Roughly 20% of the in vitro product was cellulose with taurocholate as the detergent, and no Mg2+ was required. The cellulose product was characterized by methylation analysis, electron microscopy, electron and X-ray synchrotron diffractions, and resistance to Updegraff reagent. Cellulose microfibrils were obtained in vitro, and they had the same dimensions as microfibrils isolated from primary cell walls. However, the cellulose diffracted as cellulose IVI, a disorganized form of cellulose I that is formed when the fibrillar material contains crystalline domains that are too narrow or too disorganized to be considered real cellulose I crystals (Lai-Kee-Him et al., 2002).

In related studies, but using immunoaffinity purified cellulose synthase from mung bean hypocotyls, Laosinchai (2002) also demonstrated the in vitro synthesis of cellulose microfibrils (Laosinchai, 2002).