Previous studies have revealed that D-fructose-1,6-biphosphate plays a central role in the biosynthesis of 4-hydroxy-2,5-dimethyl-3(2H)-furanone (HDMF), one of the most important flavor compounds in the strawberry. Furthermore, partially purified protein extracts from the strawberry have been shown to build HDMF from D-fructose-1,6-biphosphate in the presence of NADH. The enzymatic activity could be correlated with the presence of a single polypeptide, which contained two peptide fragments that showed total identity with the protein sequence of a putative quinone oxidoreductase (Fragaria ananassa quinone oxidoreductase) that is ripening induced. Additionally, there were similarities in the expression pattern, the amount of methionine, and the size.
In this study, the enzyme from strawberry and a similar enzyme from tomato were heterologously expressed in E. coli, purified and biochemically characterized. We confirmed, in vitro, that the heterologously expressed enzyme catalyzes the formation of HDMF from D fructose-1,6-biphosphate and NADH. In addition we determined the pH-optimum, co substrate dependencies, inactivation by heat, and the enantiomer selectivity of this enzyme. During the characterization process, it became clear from the results of longer reactions, that D-fructose-1,6-biphosphate was probably not the native substrate. Thus, additional studies on the identification of the precursor of HDMF were initiated.
One option for this precursor was the compound 1-desoxy-2,3-hexodiulose-6-phosphat, which is known to be the precursor of HDMF in yeasts. In this study, we were able to show that it is not the precursor of HDMF in strawberries. To identify the true substrate of the enzyme we synthesized the α, β unsaturated 4-hydroxy-5-methyl-2-methylene-3(2H)-furanone (HMMF). In the presence of NADH, the enzyme reduced the double bond of this compound to form HDMF. The previously unknown compound, HMMF, was synthesized by adding formaldehyde to 4 hydroxy-5-methyl-3(2H)-furanone (HMF). To proof the structure of HMMF by NMR, its reactivity required its derivatisation as a thioether. This derivatisation was also necessary to show that HMMF was indeed the precursor in vivo. The application of 3 mercapto-benzoic-acid (MBA) to strawberries at different ripening stages interrupted the biosynthesis of HDMF because the precursor, HMMF, reacted with the thiol to form a thioether. This thioether was detected in subsequent HPLC-MS-measurements. Additionally the formation of thioether correlated with the ripening stage of the fruit, as do the formation of HDMF, and the expression of the enzyme.
The proof of HMMF by this method was also successful in several other HDMF containing fruits, like the pineapple and tomato. Additionally, we were able to show that, as expected, D-glucose is transformed into HMMF during the biosynthesis of HDMF. The application of 6 13C-D-glucose to strawberries that were later treated with MBA enabled us to confirm the incorporation of the labelled sugar into the furanone part of the thioether.
At first the gene FaQR (Fragaria ananassa quinone oxidoreductase) was classified as a NAD(P)H-quinone oxidoreductase based on sequence-similarities with known QRs. However, a substrate screening with several quinones revealed that the enzyme accepted only 9,10 phenanthrenequinone (PQ), an artificial quinone, as a substrate. Other o quinones, p-quinones, α, β-unsaturated aldehydes, ketones, acids, amides, and flavones were not accepted. Thus it was necessary to synthesize more substrates [such as 2 ethylidene-4-hydroxy-5-methyl-3(2H)-furanone (EDHMF), 4-hydroxy-5-methyl-2-propylidene-3(2H)-furanone (HMPDF); 2 butylidene-4-hydroxy-5-methyl-3(2H)-furanone (BDHMF)] with the same unsaturated double bond as HMMF, but with different substitutions. These compounds were synthesized and characterized by HPLC-MS/MS-, GC-MS-, GC-O- and HPLC-DAD-data for the first time in this study. Additionally, the enzymatically formed products were only partially known, e.g. HDMF formed from HMMF and EHMF (2 (or 5)-Ethyl-4-hydroxy-5 (or 2)-methyl-3(2H)-furanone) formed from EDHMF. These two compounds are important flavor compounds in several foods. Consistent with this, the related structures BDHMF, BHMF, HMPDF, and HMPF revealed similar odor activities. Kinetic data was obtained for EDHMF, HMPDF, BDHMF, and PQ. Due to its high reactivity we were not able to gain kinetic data for HMMF.
This study shows that the enzyme reduces the α, β-unsaturated double bond in those compounds that also have at least one hydrogen atom in the β-position and no second double bond in the trans position. Other double bonds or oxygen nearby doesn’t interfere with the activity. The lack of enzymatic activity with benzil as the substrate shows that FaEO accepts only planar molecules.
Surprisingly, it was not possible to clearly classify the enzyme based solely on biochemical studies. Since we have shown in vivo that HMMF and not PQ is the substrate, we prefer the name FaEO (Fragaria ananassa enone oxidoreductase) to quinone oxidoreductase. This we appropriate, because the two reactions differ not only in their catalytic efficiencies, but also in their optimal pH and reaction mechanism. Using measurements of the stochiometry of radicals and products, we were able to confirm that the reaction mechanism of FaEO depends on whether the available substrate is an unsaturated hydroxyfuranone or a quinone. PQ is transformed into its unstable semiquinone-radical by a one-electron-reduction, whereas, when the unsaturated hydroxyfuranones are available, FaEO catalyzes a two-electron-reduction and forms the saturated compounds.
Since we could detect HMMF in tomato fruit, it wasn’t surprising that we also found a sequence similar to FaEO in a tomato expression database. The protein was heterologously expressed and the enzyme, LeEO (Lycopersicum exculentum enone oxidoreductase), showed similar characteristics as FaEO. The optimal pH and the catalytic performance were slightly higher, and there was a transit peptide at the N-terminus of the sequence.
In this study, by discovering HMMF, the precursor, and FaEO, the responsible enzyme we have uncovered another important step in the biosynthesis of HDMF. Now only the transformation of D-Fructose-1,6-biphosphate to HMMF, which includes a dephosphorylation and a dehydration, remains unknown. Finally, we were able to characterize FaEO as an enzyme able to carry out two different reaction mechanisms depending on the available substrate. The characterization of LeEO, a homologous enzyme from the tomato, confirms these results.
«