UDP-Glc dehydrogenase (UDPGDH, UGD) catalyzes the C-6 oxidation of UDP-Glc in the presence of two molecules of NAD+ to UDP-GlcA and 2NADH. The enzyme is well

characterized from numerous organisms and crystal structures are available. In plants, the gene encoding this activity was first described in soybean (460), and soon after, as genome sequence became available, numerous UDPGDH gene isoforms that share high amino acid sequence similarity to each other were identified in Arabidopsis (461), poplar (462), tobacco (463), maize (464), and Dunaliella (465). In Arabidopsis, four UDPGDH isozymes are present and, likewise, multiple UDPGDH isoforms are found in the genome of every plant species that has been sequenced. The role of multiple UDPGDH isoforms and their specific activities is of great interest since recent work in sugar cane revealed that CTP — glucose and TDP-glucose were oxidized as well, although at low rates compared to UDP-Glc (466). The current knowledge indicates that UDPGDH is cytosolic and its activity is strongly inhibited by a very low level of UDP-Xyl. This suggests that flux of UDP-Glc (hexose) to UDP-pentoses can be feedback-regulated, in part, at the enzyme level.

The source of UDP-GlcA (i. e., from myo-inositol or UDP-Glc) for wall synthesis was elegantly determined by a genetic approach. Two UDP-Glc dehydrogenase mutants were identified in maize, one in isoform A and the other in isoform B of UDPGDH. Polysaccha­rides isolated from isoform A mutant had lower Ara/Gal and Xyl/Gal ratios when compared with wildtype, indicating the importance of UDPGDH in directly providing major flux of NDP-sugars to wall polysaccharides (464). On the other hand, the lack in “wall alteration” in the isoform B mutant, could suggest either that different UDPGDH isoforms function in a different metabolic pathways, for example in the formation of glycosides, or that isoform B contributes to the synthesis of a glycan structure that is present at such low levels that small alterations in the amount of GlcA was undetected. RG-I, for example, consists of a small amount of GlcA residue (only 1% of total sugars). Similarly, GlcA residues are found in relatively low amounts in RG-II and xylan, representing 3 and 5%, respectively, of total sugars in those polysaccharides.

In 1996, Robertson and coworkers (467) reported that alcohol dehydrogenase (ADH) from Phaseolus vulgaris can convert UDP-Glc to UDP-GlcA. This bifunctional activity of ADH to oxidize both ethanol and UDP-Glc opened a debate in the literature. A major difficulty in interpreting this result is that Robertson and coworkers (467) used a spectrophotometer assay which measures NADH formation, not UDP-GlcA directly. Recently, the tobacco ADH homologous genes, cloned and expressed in bacteria, had activity on both ethanol and UDP — Glc. But again, the UDPGDH assay was determined by spectrophotometer assays (463). Whether recombinant ADH catalyzed the formation, UDP-GlcA was not explicitly con­firmed. Further kinetic work with pure ADH must be carried out to claim the dual specificity of this enzyme. Only if knockouts of all ADH genes are obtained can it be definitely deter­mined whether ADH is a bifunctional enzyme. Maize mutants lacking ADH1 and ADH2 iso­forms, however, had no affect on sugar composition of hemicellulose (464). If only two ADH isoforms exist in maize ADH, this would suggest that ADH is not a bifunctional enzyme.