Cellulose, an abundant glucan polymer, is synthesized by organisms belonging to a variety of phylogenetic domains (1). In nature, the majority of plants and the bacterium Komagataeibacter xylinus (formerly Gluconacetobacter xylinus),, synthesize a crystalline form designated cellulose I (2). The glucan chains in cellulose I are arranged parallel to one another and assembled side-by-side forming structures known as microfibrils (3,4). Cellulose I exists in two allomorphs; Iα (triclinic) and cellulose Iβ (monoclinic), which can be differentiated using solid-state 13C NMR (5). The two allomorphs are distinguished from one another by their crystal packing hydrogen bond interactions and molecular conformation, which alter the physical characteristics of the polymer (6). One cellulose microfibril may contain both types of allomorphs; with the Iβ form dominant in higher plants and the Iα form dominant in bacteria and algal species (7). An alternative form of cellulose, known as cellulose II, can be obtained through either industrial mercerization (alkali treatment) or regeneration (solubilization and subsequent recrystallization) (5). In cellulose II, the glucan chains are arranged in an antiparallel conformation (8); each glucose residue has an additional hydrogen bond which makes it a more thermodynamically stable conformation. In contrast to cellulose I, cellulose II is made by very few organisms, such as the gametophyte cells of the marine alga of the genus Halicystis and gram-positive bacteria of the genus Sarcina (1).
In bacteria, cellulose is synthesized as an extracellular polysaccharide via a single polymerization step that uses uridine diphosphoglucose (UDP-glucose) as substrate which undergoes self-assembly at the site of biosynthesis (9). This process is driven by integral membrane glycosyltransferases (GTs) that couple the elongation of the polymer with the translocation of the product outside the cell (10). A single cell of K. xylinus can polymerize up to 200,000 glucose molecules into β–1,4 glucan chains every second (11). The synthesis of bacterial cellulose (BC) is precisely regulated by a diverse number of enzymes and regulatory proteins. Synthesis occurs in the periplasm by cellulose synthesizing complexes (CSCs) or terminal complexes (TCs), which are associated with outer membrane pores organized linearly along the long axis of the cell (12–14). Protofibrils are thought to be assembled into ribbon-shaped microfibrils that elongate while remaining associated with the cell envelope, even during cell division, giving rise to BC or pellicle (14). The polymerization of cellulose is catalyzed by cellulose synthase (CS) without the formation of intermediates (15).
K. xylinus possesses a cellulose synthase operon (bcs operon) composed of three to four genes, depending on the strain or subspecies: bcsAB (or bcsA and bcsB),, bcsC, and bcsD (16). The protein product encoded by bcsA is an integral membrane protein that contains multiple transmembrane (TM) domains; four N-terminal and four C-terminal TM helices separated by an extended intracellular loop between helices 4 and 5 that forms the glycosyltransferase (GT) domain (10) that includes a conserved motif with three spaced aspartates followed by a pentapeptide sequence motif (QxxRW) characteristic among processive GTs like hyaluronan, chitin and alginate synthases (16,17). The GT domain of BcsA contains seven β-strands surrounded by seven α-helices. The narrow channel for cellulose translocation is formed by the TM helices 3 to 8 and appears to accommodate 10 glucose residues of the translocating glucan chain (10). Most inverting GTs require an essential divalent cation for catalysis. The cation is coordinated by the conserved Asp-X-Asp (DXD) motif at the active site to stabilize the nucleotide diphosphate leaving group during glycosyl transfer (10). The activity of BcsA is stimulated by an allosteric regulator, 3’, 5’-cyclic diguanylic acid (c-di-GMP), which interacts with the PilZ domain located at the C-terminus of the BcsA, next to the GT domain (18,19). Binding of the c-di-GMP causes a conformational change in the PilZ domain disrupting a specific Arg-Glu salt bridge, which results in the displacement of a “gating loop” that allows UDP-glucose to bind (19).
In contrast, the product of the bcsB gene is a predominantly β-stranded periplasmic protein with a single transmembrane anchor that interacts with BcsA (10). The BcsB protein contains four domains: two jellyroll domains, which show high similarity to carbohydrate binding domains (CBDs), and two flavodoxin-like folds (10). Although BcsB is essential for catalysis, only the interactions of its C-terminal TM anchor and an amphipathic helix seem to be necessary to stabilize the TM region of BcsA for catalysis (19). In addition, BcsB is thought to be required to guide cellulose across the periplasm toward the outer membrane via the two carbohydrate-binding domains (CBDs) (10). While the functions of BcsC and BcsD have not been fully elucidated, they both influence cellulose crystallinity. For example, BcsC is predicted to form a β-barrel porin located in the outer membrane, which is preceded by a relatively extensive periplasmic domain carrying tetratricopeptide repeats which may be involved in complex assembly (20,21). On the other hand, BcsD, for which the crystal structure has been elucidated, is thought to function as a periplasmic channel for the nascent cellulose strand (22,23). Indeed, the isolation of K. xylinus bcsD mutants, which form smooth colonies and produce reduced amounts of cellulose in culture, suggests that BcsD is involved in cellulose crystallization (9). Recently, Sajadi et al. (24) showed that heterologous expression of bcsD in Escherichia coli increased the crystallinity but did not affect the yield of BC. In addition to the cellulose synthesis operon genes, ancillary genes have been implicated in cellulose biosysnthesis. One of these genes, bcsZ (formerly cmcax), which is located upstream of the cellulose synthase operon, encodes an endo-β–1,4-glucanase (BcsZ, formerly CMCax) which possesses cellulose hydrolyzing activity that is essential for BC production (25). In addition, bcsH (formerly ccpax), also located upstream of the bcs operon, encodes a cellulose complementing protein. The function of BcsH is not well understood, but it has been suggested to be important for cellulose crystallization (26). Mutations in the bcsH locus result in a significant decrease in cellulose production (27). A downstream region of the operon which encodes an enzyme with exo–1,4-β-glucosidase activity (BglAx) towards cellotriose or larger cello-oligosaccharides also leads to significant decrease in cellulose production when mutated (28). The precise role of BglAx is, however, still unknown (27).
In a previous study, we used a chemical genetic approach to identify pellicin, a potent inhibitor of cellulose pellicle formation that exerts its effect extracellularly (29). An important property of pellicin is that it can inhibit cellulose crystallization without disrupting cell growth (29). In addition, pellicin does not inhibit cellulose synthase activity directly since crude membrane preparations from K. xylinus cells grown in pellicin-containing medium show increased cellulose synthesis compared to untreated membrane preparations (29). These and other features of pellicin suggest that it allows normal cellulose II biosynthesis assembly but inhibits the assembly of cellulose I. In order to better understand the mechanism of action of pellicin, we have conducted a forward genetic screen aimed at identifying pellicin-resistant mutants. Our current findings indicate that in K. xylinus, one indirect target for pellicin activity is the BcsA protein.