Biochemistry of Cellulose Biosynthesis in Plants

UDP-Glucose is the Immediate Precursor for Cellulose Synthesis
Although cellulose was characterized as an aggregation of glucose units by Anselme Payen in 1839, it was in 1895 that Tollens proposed that cellulose is a chain of glucose molecules (French, 2000). While the structure of cellulose was being determined and debated, studies on its biosynthesis did not truly begin until the identification of nucleotide sugars, and specifically UDP-glucose as a glucose donor in biosynthetic reactions (Leloir and Cabib, 1953). The transfer of glucose from UDP-glucose to cellulose was first described by Glaser in 1958 using particulate fraction from cell-free extracts of the bacterium Acetobacter xylinum (Glaser, 1958). However, when UDP-glucose was used as the sugar donor in experiments using digitonin-solubilized fractions from various plants, the polysaccharide product obtained in vitro was identified as callose (β-1,3-glucan) instead of cellulose (Feingold et al., 1958). Using particulate extracts from plants, the synthesis of cellulose was reported by Barber and colleagues in 1964, and from their experiments these authors concluded that the sugar donor for synthesis of cellulose was guanosine 5'-diphosphate (GDP)-glucose and not UDP-glucose (Barber et al., 1964). In these experiments, the particulate extracts from plants also allowed synthesis of an alkali-insoluble polysaccharide from GDP-mannose and from a mixture of GDP-glucose and GDP-mannose. Recently, a cellulose synthase-like protein (AtCslA9), identified as a β-glucomannan synthase, has been shown to possess β-mannan synthase, β-glucan synthase, and β-glucomannan synthase activities (Liepman et al., 2005). This β-glucomannan synthase can catalyze the production of β-mannan when supplied with GDP-mannose, a β-glucan when supplied with GDP-glucose or β-glucomannan when supplied with a combination of GDP-glucose and GDP-mannose. It is now clear that in the earlier experiments where GDP-glucose was used as a sugar donor with plant extracts, techniques for characterizing the in vitro products did not allow a clear distinction to be made between the possible β-glucomannan product and cellulose (Barber et al., 1964; Chambers and Elbein, 1970). Moreover, it was felt at the time that synthesis of the major homopolymers of glucose in plants could be regulated by using different nucleotide sugars—UDP-glucose for callose synthesis, adenosine diphosphate (ADP)-glucose for starch synthesis, and GDP-glucose for cellulose synthesis (Barber et al., 1964). We now know that in plants, although ADP-glucose is the precursor for starch synthesis, the precursor for synthesis of callose and cellulose is UDP-glucose. Support for the role of UDP-glucose as a precursor of cellulose in plants came from studies tracing the flow of carbon from glucose to cellulose in developing cotton fibers (Carpita and Delmer, 1981). Evidence for the role of UDP-glucose as the precursor for cellulose synthesis in plants did not come easily, and only a brief historical account is given here to highlight one of the many difficulties encountered in dissecting the mechanism of cellulose synthesis in plants. A detailed account of the early years and the progress that has been made since then is provided by Delmer in a number of excellent review articles (Delmer, 1983, 1999). Suffice it to say that as late as 1983, in one of her reviews Delmer summarized that ‘‘convincing in vitro synthesis of cellulose from UDP-glucose using plant extracts has never been conclusively demonstrated’’ (Delmer, 1983). In plants, UDP-glucose functions as a glucose donor in a number of glucosyl transfer reactions. From genome sequencing, it is now known that plants have the largest number of carbohydrate-modifying enzymes, and consequently UDP-glucose could participate as a glucose donor in many different reactions when unpurified plant extracts are used for in vitro cellulose synthesis (Coutinho et al., 2003). Furthermore in plants, polysaccharides, such as xyloglucan, have a backbone similar to cellulose, and it is important to distinguish the synthesis of these polysaccharides from synthesis of cellulose. Although not much has changed since the early days in the manner in which in vitro cellulose synthesis reactions were performed, a few modifications in the reaction conditions and better product characterization (described later) has allowed conclusive demonstration of in vitro cellulose synthesis from UDP-glucose using extracts from a variety of plants (Colombani et al., 2004; Kudlicka and Brown, 1997; Kudlicka et al., 1995, 1996; Lai-Kee-Him et al., 2002; Okuda et al., 1993; Peng et al., 2002).