Molecular and structural characterization of the candidate enzymes responsible for tartaric acid synthesis in the grapevine

Abstract

L-tartaric acid (TA) is accumulated to be the dominant organic acid in the grape (Vitis vinifera) berries and remains metabolically stable during berry ripening and the wine making process. It confers a low pH and a “sharp” flavor to wine, affecting many quality aspects such as colour, taste, microbial stability and aging potential. Exogenous TA is widely used as a flavorant and an antioxidant additive in the food and wine industry. The metabolic intermediates involved in the TA synthesis pathway in grapevine have been well characterized. The primary pathway utilizes L-ascorbic acid (Vitamin C) as the precursor which is sequentially converted to 2-keto-L-gulonic acid (2KLG), L-idonic acid (IA), 5-keto-D-gluconic acid (5KGA) and TA by several biochemical reactions. Only one candidate enzyme, V. vinifera L-idonate-5-dehydrogenase (LIDH, EC 1.1.1.264), involved in this pathway has been characterized so far. It catalyzes the inter-conversion of IA and 5KGA, the rate-limiting step of the TA biosynthesis pathway. In addition, another enzyme, a putative 2-keto-L-gulonate reductase (VV2KR), which is assumed to be responsible for the reduction of 2KLG to IA, has also been reported. However, further enzymatic characterization of this enzyme is still needed to validate its metabolic function in grapevine. The present study aims to investigate the molecular and structural characteristics of these two proteins, thereby improving our understanding on TA biosynthesis in grapevine. LIDH is highly homologous to sorbitol dehydrogenase (SDH, EC 1.1.1.14). We employed a computational approach to show that LIDH originates from the functional divergence of SDH in grapevine. We demonstrated that core eudicot (the strongly supported monophyletic group following the early-diverging eudicot lineages (Soltis, Senters et al. 2003)) SDHs have evolved into two distinctive phylogenetic lineages (Class I and Class II) due to positive natural selection after tandem gene duplication in the common ancestor of core eudicot plants. LIDH was identified as a Class II SDH. While the Class I SDH gene is universally conserved, Class II SDH genes are retained only in some plants including V. vinifera, Solanum tuberosum, Theobroma cacao. The distribution of SDH genes among plant genomes showed a positive correlation between the occurrence of Class II SDH genes and the capacity for TA synthesis. These results provided new insights into the TA biosynthesis pathway from the evolutionary perspective. Protein modeling analyses revealed three amino acid substitutions at the catalytic sites between Class I and Class II SDHs. These three amino acid changes are suspected to be responsible for the unique enzymatic profile of LIDH. To confirm this hypothesis, Class I and Class II SDH genes from V. vinifera, T. cacao and S. tuberosum were cloned in this study. Site-mutations of the three amino acids were performed to assess their impact on enzyme substrate specificity. Recombinant protein of both wild-type and mutant SDHs from the above plants were expressed and purified. Enzyme kinetic tests confirmed that LIDH has a preference for L-idonate over D-sorbitol as its substrate. The Class I SDH from T. cacao demonstrated the highest activity with D-sorbitol but could hardly utilize L-idonate. The recombinant protein of the other SDHs could not be purified in this study due to troubles in protein induction and purification. Despite this, the preliminary results showed a significant enzymatic divergence between Class I and Class II SDHs, supporting the putative role of Class II SDHs in the TA biosynthesis pathway. In addition, a complete enzymatic characterization of the recombinant VV2KR was performed. VV2KR has the highest substrate specificity with D-glyoxylate, followed by hydroxypyruvate and 2KLG. We showed that VV2KR could catalyze the reduction of 2KLG to L-idonic acid effectively using NADPH as the preferred coenzyme. We went further to crystallize the recombinant VV2KR to determine its 3-dimmentional structure. The ligand-free crystal structure of VV2KR was solved to a resolution of 2.1 Å. VV2KR has the highest similarity with Coleus blumei hydroxypyruvate/glyoxylate reductase (CbHPR) and Arabidopsis thaliana hydroxypyruvate/glyoxylate reductase (AtHPR) isoform 2 (~75% amino acid sequence identity). The VV2KR monomer structure adopts the typical folding of the D-isomer 2-hydroxyacid dehydrogenase (2KDH) proteins and has the highest structural homology with the CbHPR structure (main-chain atom Root Mean Square Deviation of 0.76 Å; main-chain atom include all atoms in the peptide chain except those from the R-group, i.e. the side-chain). It represents the second plant HPR structure being determined to date. Structural comparison of VV2KR and CbHPR with other HPR structures revealed some unique structural features for plant HPRs. The commonly accepted substrates for HPRs (D-glyoxylate, hydroxypyruvate, pyruvate) and the 2KLG intermediate from TA pathway were docked to the VV2KR structure by a computational method. The potential biological function of VV2KR in the TA biosynthesis pathway in grapevine was discussed. In summary, we focused on the molecular and structural characterization of the candidate enzymes responsible for TA biosynthesis in grapevine. The results of the present study highlighted the effects of enzyme structural variations on their in-vivo biological functions and significantly extended our understanding on the molecular mechanism of the TA biosynthesis pathway in grapevine.