**2.2 Glycosyltransferases involved in CPSA biosynthesis**

*Bioactive Compounds - Biosynthesis, Characterization and Applications*

Within the CPSA biosynthesis locus, there is a predicted aminotransferase gene, *wcfR*, and a hexose phosphate initiating transferase, *wcfS*, but no predicted dehydratase was found. However, a gene encoding a potential dehydratase, *ungD2*, has been identified elsewhere in the *B. fragilis* genome. When this gene was knocked out by Coyne *et al*, they found out that, synthesis of the seven out of the eight capsular polysaccharides was stopped. Initial studies with UngD2 and WcfR did not show any promise in the synthesis of AADGal. Hence a previously well characterized dehydratase, PglF [34], from *Campylobacter jejuni* was used to provide the substrate needed for WcfR function. The coupling of these enzymes together led to the production of AADGal (**Figure 6**) [35, 36]. This also points to the notion that depending on homology alone for functional assignment of genes, is not always right, and

wet lab results are needed to confirm the function of the gene product.

phosphate analogue to find out the function of WcfS (**Figure 7**) [33].

The synthesized UDP-AADGal was further used as a potential substrate for WcfS, identified as the initiating hexose phosphate transferase. Studies done by Mostafavi et al. demonstrated that WcfS was indeed the initiating hexose phosphate transferase, which lead to the formation of the bactoprenyl linked monosaccharide

As mentioned previously, assembly of the polysaccharides in bacterial cells is done on a C55 bactoprenyl anchor. It is produced by the condensation of farnesyl diphosphate (FPP) to eight units of isopentenyl diphosphate (IPP), done by the enzyme undecaprenyl diphosphate synthase (UPPS). A major drawback of using this compound in in vitro assays is that, it does not have easily distinguishable chromophores associated to it, hence very few rapid assays are available to detect and quantify the activity of enzymes associated with polysaccharide synthesis. To circumvent this problem, the Troutman lab developed fluorescent analogues of the native bactoprenyl, which are easily traceable [25, 37]. Assays done using these analogues take a short time to reveal valuable information about the enzymes when compared to traditional assays, which follow the more tedious route of using radioactive labeled substrates. Mostafavi et al. used a p-nitroaniline bactoprenyl

**110**

**Figure 7.**

**Figure 6.**

*Biosynthesis of AADGal.*

(**Figure 7**) [33].

*Biosynthesis of bactoprenyl linked monosaccharide.*

Cell surface polysaccharides are nothing but complex carbohydrates. They play important roles in a number of biological processes such as cell growth, cell to cell interactions, immune response, and inflammation. The polysaccharides are synthesized by a class of enzymes known as glycosyltransferases [38]. Glycosyltransferases are an enzyme superfamily responsible for the attachment of carbohydrate moieties to a wide array of acceptors that include nucleic acids, polysaccharides, proteins, lipids, and carbohydrates. The majority of glycosyltransferases are sugar nucleotide-dependent enzymes, and utilize nucleoside diphosphate sugars (NDP-sugars) as donors for the glycosidic bond formation. In other cases, the sugar donors can also be lipid phosphates and unsubstituted phosphate [39].

The glycosyltransferases have been classified by sequence homology into 96 families in the Carbohydrate Active enZyme database (CAZy), each of which catalyze the reaction as shown in **Figure 8** [40]. Chain elongation of the oligosaccharide units in complex carbohydrates is achieved by the addition of monosaccharide units through the action of different glycosyltransferases in a specific sequence. The CAZy database provides a highly powerful predictive tool, as the structural fold and mechanism of action are invariant in most of the families [22]. Therefore, where the structure and mechanism of a glycosyltransferase member for a given family has been reported, some assumptions about other members of the family can be made. Substrate specificity, however, is more difficult to predict, and requires experimental characterization of individual glycosyltransferases.

Determining both the sugar donor and acceptor for a glycosyltransferase of unknown function can be challenging, and it is one of the reasons there are significantly fewer well characterized isoprenoid linked sugar glycosyltransferases when compared to the glycosyltransferases responsible for synthesizing disaccharides or the oligosaccharides [40]. The less reports on isoprenoid linked sugar transferases can be attributed to the fact that, a high throughput method has not yet been developed which will enable for faster characterization. Another challenge in characterizing the glycosyltransferases is the availability of rare sugars, as most of the bacterial polysaccharides contain rare sugars. Rare sugars, such as rhamnose or fucose, may provide the bacterial polysaccharides with additional biological properties compared to those composed of more common sugar monomers [23, 41]. Rare sugars are monosaccharides that are not commonly found in nature, in comparison to D-glucose, D-galactose, D-fructose, D-xylose, D-ribose, and L-arabinose which are more abundant [23]. Moreover, the traditional methods like radioisoptopic

**Figure 8.** *General reaction scheme for a glycosyltransferase (GTs).*

labelling, thin-layer chromatography (TLC) used to characterize the glycosyltransferase, often tends to be tedious and challenging in tracking the product.

Glycosyltransferases catalyze glycosidic bond formation with either overall retention or inversion of anomeric configuration when compared to the stereochemistry in the sugar donor (**Figure 9**). Inverting glycosyltransferases are generally believed to proceed via a single displacement SN2 mechanism with concomitant nucleophilic attack by the acceptor at the anomeric carbon, facilitated by proton transfer to the catalytic base, and leaving group departure [22]. Structural data have shown that several inverting glycosyltransferases, contain no obvious candidate catalytic base indicating these enzymes use an alternative mechanism [38, 39].

The reaction coordinate employed by retaining glycosyltransferases has been much debated, and it could be possible the mechanism is not conserved for all retaining enzymes. One possibility is a double displacement mechanism via a covalent mechanism, analogous to that used by glycoside hydrolases [22]. A report by Soya et al. provided mass spectrometry evidence for the formation of a covalent intermediate between the donor substrate and a cysteine, which had been substituted for the candidate catalytic nucleophile, on two retaining glycosyltransferases [42]. The more favored mechanism in the field is an SN1 or SN1-like mechanism, which involves interaction between the leaving group and attacking nucleophile on the same face. This mechanism is supported by kinetic isotope effect studies to analyze the structure of the transition state and by computational modeling [38, 39].

The CPSA gene locus has three genes, wcfQ, wcfP and wcfN, that putatively encode for glycosyltransferases [29, 30]. Each of these glycosyltransferases is expected to transfer a sugar moiety to the bactoprenyl linked monosaccharide, the disaccharide and the trisaccharide. Based on the CAZy database, and homology studies, WcfQ and WcfN are hypothesized to belong to the glycosyltransferase superfamily A, which follows the inverting mechanism in the sugar transfer. Whereas WcfP is proposed to belong to the glycosyltransferase superfamily B, which follows the retaining mechanism [40].

WcfQ, identified as the first glycosyltransferase, transfers galactose to the isoprenoid linked monosaccharide, even though it was observed by authors that, WcfQ could also transfer glucose to the bactoprenyl linked monosaccharide. This is because WcfQ required glucose in much excess when compared to galactose. It was also found out that even though WcfP had the capability of transferring galactose, WcfQ was more efficient in it, hence it was identified as the galactosyltransferase

#### **Figure 9.**

*Glycosyltransferases catalyze glycosyl group transfer with either inversion or retention of the anomeric stereochemistry with respect to the donor sugar.*

**113**

*Biosynthesis of the Immunomodulatory Molecule Capsular Polysaccharide…*

in the CPSA biosynthetic pathway. Moreover, based on the Carbohydrate-Active enZYmes (CAZY) database the WcfQ sequence matched the GT\_2 family of glycosyltransferases which invert the configuration of the anomeric carbon of the donor, while WcfP was similar to a GT\_4 family glycosyltransferase, which retains the anomeric stereo-configuration of the donating sugar [43, 44]. The published structure of the CPSA tetrasaccharide unit suggests that the linkage should be in a beta configuration [27]. This supported the conclusion that WcfQ is the protein responsible for introducing galactose, and that it introduces the sugar in the appro-

As stated before, WcfP is related to the GT\_4 family of proteins suggesting that it is a retaining glycosyltransferase, it was therefore more likely that WcfP catalyzed UDP-GalNAc transfer to the galactose, but it was not known if it transferred UDP-GalNAc to the unpyruvylated disaccharide or the pyruvylated disaccharide. Both WcfN and WcfP were analyzed with the pyruvylated and the unpyruvylated disaccharides, it was demonstrated that WcfP transfers only UDP-GalNAc to the

In homology studies, WcfN was predicted to be a member of the GT\_2 family, whose members have been identified to transfer furanose residues. WcfN was also hypothesized to be an inverting transferase, which inverts the stereochemistry of the anomeric carbon. Since the linkage between the third and the fourth sugar in the tetrasaccharide repeat unit is in the beta configuration, WcfN fitted the role of being the last glycosyltransferase. WcfN was found to transfer the galactofuranose to the trisaccharide, hence completing the mapping of the pathway of synthesis of

Polysaccharides composed of furanosyl residues are important constituents of many bacteria, protozoa, fungi, plants and archaebacteria [46, 47]. The furanosyl constituents have also been identified in glycopeptides, glycolipids as well as nucleotide sugars. D-Galactose is by far the most widespread hexose in the furanose form in naturally occurring polysaccharides, and the most impressive examples of these glycans are encountered in mycobacteria [48–50]. Galactofuranose, (Galf), which is thermodynamically less stable than galactose, is essential for the viability of several pathogenic species of bacteria and protozoa. It is absent in this form in mammalian cell structure, hence the biochemical pathways by which galactofuranose containing glycans are assembled have been attractive sites for drug action [47, 51]. This potential has led to an increased interest in the synthesis of molecules containing galactofuranose residues, and their subsequent use in studies directed towards understanding of the enzymes that process these residues and the identifi-

The enzyme UDP-galactopyranose mutase is central to galactofuranose metabolism. Most organisms cannot use exogenous galactofuranose, and UDPgalactofuranose appears to be the biological source of galactofuranose residues in polysaccharides [46]. The major structural component of the *Mycobacterium tuberculosis* cell wall contains a galactan chain of approximately thirty-five galactofuranose units, and the biosynthesis of the galactan is essential for viability [47]. The O-antigens of both *Escherichia coli* and *Klebsiella pneumoniae* contain galactofuranose as a component of lipopolysaccharide [47]. Several galactofuranose containing glycoconjugates have been found in *Trypanosoma cruzi*, the causative agent of Chagas disease, including glycoinositolphospholipids, lipopeptidophosphoglycans and mucin-like proteins. The galactomannan of *Aspergillus fumigatus* also contains galactofuranose, and this polysaccharide is used for clinical detection

*DOI: http://dx.doi.org/10.5772/intechopen.96937*

priate beta configuration [45].

pyruvylated disaccharide.

the tetrasaccharide.

**2.3 WcfM as the galactopyranosemutase**

cation of potential inhibitors of these pathways [46].

*Biosynthesis of the Immunomodulatory Molecule Capsular Polysaccharide… DOI: http://dx.doi.org/10.5772/intechopen.96937*

*Bioactive Compounds - Biosynthesis, Characterization and Applications*

which follows the retaining mechanism [40].

labelling, thin-layer chromatography (TLC) used to characterize the glycosyltrans-

Glycosyltransferases catalyze glycosidic bond formation with either overall retention or inversion of anomeric configuration when compared to the stereochemistry in the sugar donor (**Figure 9**). Inverting glycosyltransferases are generally believed to proceed via a single displacement SN2 mechanism with concomitant nucleophilic attack by the acceptor at the anomeric carbon, facilitated by proton transfer to the catalytic base, and leaving group departure [22]. Structural data have shown that several inverting glycosyltransferases, contain no obvious candidate catalytic base indicating these enzymes use an alternative mechanism [38, 39]. The reaction coordinate employed by retaining glycosyltransferases has been much debated, and it could be possible the mechanism is not conserved for all retaining enzymes. One possibility is a double displacement mechanism via a covalent mechanism, analogous to that used by glycoside hydrolases [22]. A report by Soya et al. provided mass spectrometry evidence for the formation of a covalent intermediate between the donor substrate and a cysteine, which had been substituted for the candidate catalytic nucleophile, on two retaining glycosyltransferases [42]. The more favored mechanism in the field is an SN1 or SN1-like mechanism, which involves interaction between the leaving group and attacking nucleophile on the same face. This mechanism is supported by kinetic isotope effect studies to analyze the structure of the transition state and by computational modeling [38, 39]. The CPSA gene locus has three genes, wcfQ, wcfP and wcfN, that putatively encode for glycosyltransferases [29, 30]. Each of these glycosyltransferases is expected to transfer a sugar moiety to the bactoprenyl linked monosaccharide, the disaccharide and the trisaccharide. Based on the CAZy database, and homology studies, WcfQ and WcfN are hypothesized to belong to the glycosyltransferase superfamily A, which follows the inverting mechanism in the sugar transfer. Whereas WcfP is proposed to belong to the glycosyltransferase superfamily B,

WcfQ, identified as the first glycosyltransferase, transfers galactose to the isoprenoid linked monosaccharide, even though it was observed by authors that, WcfQ could also transfer glucose to the bactoprenyl linked monosaccharide. This is because WcfQ required glucose in much excess when compared to galactose. It was also found out that even though WcfP had the capability of transferring galactose, WcfQ was more efficient in it, hence it was identified as the galactosyltransferase

*Glycosyltransferases catalyze glycosyl group transfer with either inversion or retention of the anomeric* 

ferase, often tends to be tedious and challenging in tracking the product.

**112**

**Figure 9.**

*stereochemistry with respect to the donor sugar.*

in the CPSA biosynthetic pathway. Moreover, based on the Carbohydrate-Active enZYmes (CAZY) database the WcfQ sequence matched the GT\_2 family of glycosyltransferases which invert the configuration of the anomeric carbon of the donor, while WcfP was similar to a GT\_4 family glycosyltransferase, which retains the anomeric stereo-configuration of the donating sugar [43, 44]. The published structure of the CPSA tetrasaccharide unit suggests that the linkage should be in a beta configuration [27]. This supported the conclusion that WcfQ is the protein responsible for introducing galactose, and that it introduces the sugar in the appropriate beta configuration [45].

As stated before, WcfP is related to the GT\_4 family of proteins suggesting that it is a retaining glycosyltransferase, it was therefore more likely that WcfP catalyzed UDP-GalNAc transfer to the galactose, but it was not known if it transferred UDP-GalNAc to the unpyruvylated disaccharide or the pyruvylated disaccharide. Both WcfN and WcfP were analyzed with the pyruvylated and the unpyruvylated disaccharides, it was demonstrated that WcfP transfers only UDP-GalNAc to the pyruvylated disaccharide.

In homology studies, WcfN was predicted to be a member of the GT\_2 family, whose members have been identified to transfer furanose residues. WcfN was also hypothesized to be an inverting transferase, which inverts the stereochemistry of the anomeric carbon. Since the linkage between the third and the fourth sugar in the tetrasaccharide repeat unit is in the beta configuration, WcfN fitted the role of being the last glycosyltransferase. WcfN was found to transfer the galactofuranose to the trisaccharide, hence completing the mapping of the pathway of synthesis of the tetrasaccharide.

#### **2.3 WcfM as the galactopyranosemutase**

Polysaccharides composed of furanosyl residues are important constituents of many bacteria, protozoa, fungi, plants and archaebacteria [46, 47]. The furanosyl constituents have also been identified in glycopeptides, glycolipids as well as nucleotide sugars. D-Galactose is by far the most widespread hexose in the furanose form in naturally occurring polysaccharides, and the most impressive examples of these glycans are encountered in mycobacteria [48–50]. Galactofuranose, (Galf), which is thermodynamically less stable than galactose, is essential for the viability of several pathogenic species of bacteria and protozoa. It is absent in this form in mammalian cell structure, hence the biochemical pathways by which galactofuranose containing glycans are assembled have been attractive sites for drug action [47, 51]. This potential has led to an increased interest in the synthesis of molecules containing galactofuranose residues, and their subsequent use in studies directed towards understanding of the enzymes that process these residues and the identification of potential inhibitors of these pathways [46].

The enzyme UDP-galactopyranose mutase is central to galactofuranose metabolism. Most organisms cannot use exogenous galactofuranose, and UDPgalactofuranose appears to be the biological source of galactofuranose residues in polysaccharides [46]. The major structural component of the *Mycobacterium tuberculosis* cell wall contains a galactan chain of approximately thirty-five galactofuranose units, and the biosynthesis of the galactan is essential for viability [47]. The O-antigens of both *Escherichia coli* and *Klebsiella pneumoniae* contain galactofuranose as a component of lipopolysaccharide [47]. Several galactofuranose containing glycoconjugates have been found in *Trypanosoma cruzi*, the causative agent of Chagas disease, including glycoinositolphospholipids, lipopeptidophosphoglycans and mucin-like proteins. The galactomannan of *Aspergillus fumigatus* also contains galactofuranose, and this polysaccharide is used for clinical detection

of fungal infections. Finally, it is also known that stopping galactofuranose biosynthesis in *Leishmania major* attenuates its virulence [46–48, 51]. The abovementioned pathogenic organisms all use the same building block for synthesizing galactofuranose-containing polysaccharides: uridine diphosphogalactofuranose (UDP-galactofuranose). This sugar nucleotide is produced from UDP-Glcp by the enzymes UDP-Glucose 4-epimerase (generating UDP-Galp,) and UDPgalactopyranose mutase (UGM), which catalyzes the transformation of UDP-Galp to UDP-galactofuranose. The gene encoding UGM was first identified in *E. coli* in 1996, followed shortly by its identification in *K. pneumoniae* and *M. tuberculosis* [48, 49, 52] . More recently, UGM was identified in the eukaryotes *A. fumigatus, Cryptococcus neoformans, L. major* and *T. cruzi*.

In the past several years' major milestones have been achieved, which include an in-depth understanding of the mechanism of UDP-galactopyranose mutase (UGM), the enzyme which produces UDP-galactofuranose, and is the donor species used by galactofuranosyltransferases. A number of methods for the synthesis of galactofuranosides have also been developed [50]. UDP-galactofuranose has also been prepared by a number of approaches, and currently it appears that a chemoenzymatic approach is the most viable method for producing multi-milligram amounts of this important rare sugar [46, 50].

The biosynthetic gene operon of CPSA encodes a wcfM gene, which was found to be homologous to other galactopyranose mutases. It is homologous to two known UDP-galactopyranose mutases, one from *Streptococcus pneumonia* (Cps33fN: 66% identity and 82% similarity) and the other from *E. coli* (59% identity and 79% similarity). The gene encodes a 43 kDa protein with one potential N-terminal transmembrane domain. Like other galactopyranose mutases, the protein is hypothesized to catalyze the reaction as shown in **Figure 10**. The product of WcfM is required for the final step in the synthesis of the CPSA tetrasaccharide repeat unit. The last glycosyltransferase transfers UDP-galactofuranose to the trisaccharide.

#### **2.4 WcfO as the pyruvyltransferase**

Pyruvyltrasferases and pyruvylation have been less studied in prokaryotes, despite a burgeoning evidence of its presence in bacteria. Addition of pyruvate moiety gives a negative charge to the polymer and is utilized by the bacteria in various functions [53]. An example of this is the pyruvylation of ManNAc residue by the enzyme CsaB in the secondary cell wall polymer of *Bacillus anthracis* and *Paenibacillus Alvei* [54, 55]*.* This pyruvylated residue comes in use in anchoring the S-layer proteins in Gram positive bacteria by binding to the SLH domains of the S-layer proteins [56]. Knocking out the CsaB has led to a lethal phenotype, which suggests that, pyruvylation of the secondary cell wall polymer is essential to the

**115**

units [65, 66].

*Biosynthesis of the Immunomodulatory Molecule Capsular Polysaccharide…*

and polysaccharides to their immunological properties [63, 64].

and it apparently utilized both the substrates with equal vigor.

Previous studies on the CPSA molecule have revealed it to be effective as a therapeutic molecule, the tetrasaccharide repeat needs to be a polymer of ten repeat units or longer. If shorter than that, it fails to activate the immune system [64, 65]. CPSA operon encodes for a flippase wzx, which takes the repeat unit and flips it from the cytoplasmic space to the periplasmic space, where the polymerase wzy, utilizes the repeat unit and polymerizes it till it reaches a length of approximately 130 repeat

Recent successes in cancer vaccines and in monoclonal antibody cancer immunotherapy have given the impetus towards development of vaccines targeting cancer-associated carbohydrates. The Andreana group have been developing carbohydrate immunogens to elicit a T-cell dependent immune response. CPSA is known to stimulate a strong T-cell mediated response. They have successfully linked CPSA to the tumor-associated carbohydrate antigen (TACA), Sialyl Thomsen-nouveau (STn) and were able to obtain a robust immune response to the antigen [67–71]. They have further reported total synthesis of the CPSA unit in 19 steps with a final yield of 5% [67]. Chemoenzymatic assembly is a faster and scalable approach, that can be used as an alternative or in combination with chemical synthesis. CPSA

**3. Significance of capsular polysaccharide A**

growth and survival of the bacteria [55]. CsaB was recently characterized by the Schaffer group [57]. Including WcfO, a total of three pyruvyltransferases have now been functionally characterized. Pvg1b is from an eukaryote, and whose crystal

Polysaccharides of various prokaryotes are covalently linked with variable combinations of sulfates and pyruvates, for example, *Rhizobium leguminosarum*: 4,6-pyrGalactose and 4,6- pyrGlucose, *Bacillus anthracis*: 4,6-pyrManNAc, and *Xanthomonas campestris*: 4,6-pyrMannose. These modifications provide a highly negative charge of these polysaccharides, which is often essential for function [60]. For example, when the pyruvyltransferase PssM, responsible for the pyruvate modification in the *R. leguminosarum* exopolysaccharide was deleted, the bacterium was found to be ineffective in infecting pea plants to initiate the formation of root nodules. This led to formation of aberrant root nodules, which were unable to fix nitrogen [61, 62]. Moreover, some studies have linked the pyruvic acetals in oligo-

Among the eleven proteins encoded in the CPSA gene operon, one of the genes transcribes a hypothesized pyruvyltransferase based on homology studies performed using pBLAST [31]. There is little sequence similarity to other known proteins with the wcfO gene product. WcfO has very minimal sequence identity to the two characterized pyruvyltransferases Pvg1p from *S. pombe* and PssM from *R. leguminosarum*. The activity of CPSA is dependent on its zwitterionic character in which the –AADGal amino group is positively charged while the pyruvate is negatively charged [16]. Due to the fact that all other sugar modifying enzymes and glycosyltransferases required for CPSA biosynthesis have been located in the CPSA biosynthesis operon, it was proposed by the authors that the wcfO gene product was likely responsible for the pyruvylation modification required for the formation of the second sugar in the CPSA tetrasaccharide repeat unit. WcfO is capable of modifying galactose or glucose when they are linked to the isoprenoid lipid carrier. This points to the direction that, there may be sub-families within the pyruvyltransferase family that utilize different substrates. Kinetic evaluation of WcfO was performed by the authors to test if discriminated between glucose and galactose,

*DOI: http://dx.doi.org/10.5772/intechopen.96937*

structure has been solved [58, 59].

**Figure 10.** *Reaction catalyzed by UGM.*

#### *Biosynthesis of the Immunomodulatory Molecule Capsular Polysaccharide… DOI: http://dx.doi.org/10.5772/intechopen.96937*

*Bioactive Compounds - Biosynthesis, Characterization and Applications*

*Cryptococcus neoformans, L. major* and *T. cruzi*.

amounts of this important rare sugar [46, 50].

**2.4 WcfO as the pyruvyltransferase**

of fungal infections. Finally, it is also known that stopping galactofuranose biosynthesis in *Leishmania major* attenuates its virulence [46–48, 51]. The abovementioned pathogenic organisms all use the same building block for synthesizing galactofuranose-containing polysaccharides: uridine diphosphogalactofuranose (UDP-galactofuranose). This sugar nucleotide is produced from UDP-Glcp by the enzymes UDP-Glucose 4-epimerase (generating UDP-Galp,) and UDP-

galactopyranose mutase (UGM), which catalyzes the transformation of UDP-Galp to UDP-galactofuranose. The gene encoding UGM was first identified in *E. coli* in 1996, followed shortly by its identification in *K. pneumoniae* and *M. tuberculosis* [48, 49, 52] . More recently, UGM was identified in the eukaryotes *A. fumigatus,* 

In the past several years' major milestones have been achieved, which include an in-depth understanding of the mechanism of UDP-galactopyranose mutase (UGM), the enzyme which produces UDP-galactofuranose, and is the donor species used by galactofuranosyltransferases. A number of methods for the synthesis of galactofuranosides have also been developed [50]. UDP-galactofuranose has also been prepared by a number of approaches, and currently it appears that a chemoenzymatic approach is the most viable method for producing multi-milligram

The biosynthetic gene operon of CPSA encodes a wcfM gene, which was found to be homologous to other galactopyranose mutases. It is homologous to two known UDP-galactopyranose mutases, one from *Streptococcus pneumonia* (Cps33fN: 66% identity and 82% similarity) and the other from *E. coli* (59% identity and 79% similarity). The gene encodes a 43 kDa protein with one potential N-terminal transmembrane domain. Like other galactopyranose mutases, the protein is hypothesized to catalyze the reaction as shown in **Figure 10**. The product of WcfM is required for the final step in the synthesis of the CPSA tetrasaccharide repeat unit. The last glycosyltransferase transfers UDP-galactofuranose to the trisaccharide.

Pyruvyltrasferases and pyruvylation have been less studied in prokaryotes, despite a burgeoning evidence of its presence in bacteria. Addition of pyruvate moiety gives a negative charge to the polymer and is utilized by the bacteria in various functions [53]. An example of this is the pyruvylation of ManNAc residue by the enzyme CsaB in the secondary cell wall polymer of *Bacillus anthracis* and *Paenibacillus Alvei* [54, 55]*.* This pyruvylated residue comes in use in anchoring the S-layer proteins in Gram positive bacteria by binding to the SLH domains of the S-layer proteins [56]. Knocking out the CsaB has led to a lethal phenotype, which suggests that, pyruvylation of the secondary cell wall polymer is essential to the

**114**

**Figure 10.**

*Reaction catalyzed by UGM.*

growth and survival of the bacteria [55]. CsaB was recently characterized by the Schaffer group [57]. Including WcfO, a total of three pyruvyltransferases have now been functionally characterized. Pvg1b is from an eukaryote, and whose crystal structure has been solved [58, 59].

Polysaccharides of various prokaryotes are covalently linked with variable combinations of sulfates and pyruvates, for example, *Rhizobium leguminosarum*: 4,6-pyrGalactose and 4,6- pyrGlucose, *Bacillus anthracis*: 4,6-pyrManNAc, and *Xanthomonas campestris*: 4,6-pyrMannose. These modifications provide a highly negative charge of these polysaccharides, which is often essential for function [60]. For example, when the pyruvyltransferase PssM, responsible for the pyruvate modification in the *R. leguminosarum* exopolysaccharide was deleted, the bacterium was found to be ineffective in infecting pea plants to initiate the formation of root nodules. This led to formation of aberrant root nodules, which were unable to fix nitrogen [61, 62]. Moreover, some studies have linked the pyruvic acetals in oligoand polysaccharides to their immunological properties [63, 64].

Among the eleven proteins encoded in the CPSA gene operon, one of the genes transcribes a hypothesized pyruvyltransferase based on homology studies performed using pBLAST [31]. There is little sequence similarity to other known proteins with the wcfO gene product. WcfO has very minimal sequence identity to the two characterized pyruvyltransferases Pvg1p from *S. pombe* and PssM from *R. leguminosarum*. The activity of CPSA is dependent on its zwitterionic character in which the –AADGal amino group is positively charged while the pyruvate is negatively charged [16]. Due to the fact that all other sugar modifying enzymes and glycosyltransferases required for CPSA biosynthesis have been located in the CPSA biosynthesis operon, it was proposed by the authors that the wcfO gene product was likely responsible for the pyruvylation modification required for the formation of the second sugar in the CPSA tetrasaccharide repeat unit. WcfO is capable of modifying galactose or glucose when they are linked to the isoprenoid lipid carrier. This points to the direction that, there may be sub-families within the pyruvyltransferase family that utilize different substrates. Kinetic evaluation of WcfO was performed by the authors to test if discriminated between glucose and galactose, and it apparently utilized both the substrates with equal vigor.
