Identification and characterization of a functional L-gulono-1,4-lactone oxidase in Arabidopsis thaliana.

摘要

Vitamin C (ascorbate, AsA) is essential for plant and animal health. We cannot synthesize AsA on our own and therefore we must consume it from fruits and vegetables. Ascorbate plays multiple roles in plant physiology, and increasing AsA content in plants remains challenging. There are four biosynthetic AsA pathways in plants: the D-mannose/L-galactose, L-gulose, D-galacturonate, and myo-inositol routes. The AsA metabolic network is not completely understood. The inositol pathway has four enzyme-catalyzed steps, and our group has made significant progress on characterizing the first three enzymes. This work focuses on L-gulono-1,4-lactone oxidase (GulLO), the last enzyme in the inositol pathway to AsA. GulLO belongs to the aldonolactone oxidoreductases protein family and properties of these enzymes are here reviewed in detail. There are seven putative GulLO genes in Arabidopsis, and bioinformatic analyses narrowed down our interest into two: AtGulLO3 (At5g11540) and AtGulLO5 (At2g46740). A Japanese group over-expressed these enzymes in tobacco cells and that lead to elevated AsA content after substrate ( L-gulono-1,4-lactone, L-GulL) feeding. However the specificity of the proteins was not studied. Here we characterized the recombinant AtGulLO5 in detail and found it to be an exclusive dehydrogenase and specific only to L-GulL. To our knowledge, this constitutes the characterization of the first true plant GulLO enzyme. Overexpression of AtGulLO3 or AtGulLO5 in N. benthamiana (transient) or A. thaliana (stable) did not lead to increased foliar AsA content. However, our evidence indicates that AtGulLO3 and AtGulLO5 undergo a post-transcriptional regulation. AtGulLO5 may need an effector molecule for its increased catalytic efficiency while AtGulLO3 is short-lived. We propose that GulLO activity is limiting for AsA synthesis in plants. The proposed post-transcriptional regulation may be present only in whole plants but not in cell cultures. We also propose that the isoform from rat (RnGulLO) likely uses L-GulL as a substrate in previously generated transgenic plants. These findings suggest that the L-GulL pool may be used independent of the L-galactono-1,4-lactone (L-GalL) pool. The potential of using these new findings to generate plants with increased AsA content in the future is discussed.