Ascorbic acid (AsA) biosynthesis in plants predominantly occurs via a pathway with d-mannose and l-galactose as intermediates. One alternative pathway for AsA synthesis, which is similar to the biosynthesis route in mammals, is controversially discussed for plants. Here, myo-inositol is cleaved to glucuronic acid and then converted via l-gulonate to AsA. In contrast to animals, plants have an effective recycling pathway for glucuronic acid, being a competitor for the metabolic rate. Recycling involves a phosphorylation at C1 by the enzyme glucuronokinase. Two previously described T-DNA insertion lines in the gene coding for glucuronokinase1 show wild type-like expression levels of the mRNA in our experiments and do not accumulate glucuronic acid in labelling experiments disproving that these lines are true knockouts. As suitable T-DNA insertion lines were not available, we generated frameshift mutations in the major expressed isoform glucuronokinase1 (At3g01640) to potentially redirect metabolites to AsA. However, radiotracer experiments with H-myo-inositol revealed that the mutants in glucuronokinase1 accumulate only glucuronic acid and incorporate less metabolite into cell wall polymers. AsA was not labelled, suggesting that Arabidopsis cannot efficiently use glucuronic acid for AsA biosynthesis. All four mutants in glucuronokinase as well as the wild type have the same level of AsA in leaves.

These two authors contributed equally. SUMMARYPlant cell wall polymers are synthesized by glycosyltransferases using nucleotide sugars as substrates. Most UDP-sugars are synthesized from UDP-glucose via de novo pathways but salvage pathways work in parallel to recycle sugars, which have been released during cell wall polymer and glycoprotein turnover. Here we report on the cloning and biochemical analysis of two arabinokinases in Arabidopsis. Arabinokinase is a 100 kDa protein located in the cytosol with a putative N-terminal glycosyltransferase domain and a C-terminal sugar-1-kinase domain. This unique structure is highly conserved in the plant kingdom. Arabinokinase has a high affinity for L-arabinose, which is the only sugar substrate of this GHMP (galactose; homoserine; mevalonate; phosphomevalonate) kinase. Plants that were knocked-out for arabinokinase and the previously described ara1-1 mutant were characterized. The ARA1-1 mutant form of the enzyme carries a point mutation in an a-helix. The mutation is close to the substrate binding site and changes the K m value for arabinose from 80 lM in the wild type to 17 000 lM in ARA1-1. The previous arabinose toxicity explanation is challenged by knockout plants in arabinokinase that accumulate higher levels of arabinose but do not show signs of arabinose toxicity. Analysis of marker genes from sugar signalling pathways (SnRK1 and Tor) suggest that ara1-1 misinterprets its carbon energy status. Although glucose is present in ara1-1 similar to wild type levels, it constitutively changes gene expression as typically found in wild type plants only under starvation conditions. Furthermore, ara1-1 shows increased expression of marker genes for programmed cell death as found in other lesion mimic mutants.

We report the characterization of the dimeric protein AB21 from Agaricus bisporus, one of the most commonly and widely consumed mushrooms in the world. The protein shares no significant sequence similarity with any protein of known function, and it is the first characterized member of its protein family. The coding sequence of the ab21 gene was determined and the protein was expressed in E. coli in a recombinant form. We demonstrated a high thermal and pH stability of AB21 and proved the weak affinity of the protein to divalent ions of some transition metals (nickel, zinc, cadmium, and cobalt). The reported crystallographic structure exhibits an interesting rod-like helical bundle fold with structural similarity to bacterial toxins of the ClyA superfamily. By immunostaining, we demonstrated an abundance of AB21 in the fruiting bodies of A. bisporus.