46 二機能性環化酵素遺伝子の相同性に基づく新規テルペノイドのゲノムマイニング (口頭発表の部)

抄録

Bifunctional terpene synthases (BFTSs) are attractive target for genome mining because they yield di/sesterterpenes with unique polycyclic structures. Bioinformatics analysis of N-terminal terpene cyclase domains of BFTSs allowed us to categorize them into 6 clades. Interestingly, all functionally characterized BFTSs, such as PaFS, PaPS, and AcOS, were classified into Clade B. Taken that those BFTSs catalyze cyclization via carbocation intermediate with 5-11 bicyclic ring system, it was proposed that the amino acid sequence reflects the initial cyclization mode, which is most likely related to the initial conformation of a linear prenyl diphosphate. In this study, we examined genome mining focusing on BFTSs classified into Clade A. Heterologous expression of five BFTS genes into Aspergillus oryzae or Escherichia coli allowed us to isolate structurally related compounds, three sesterterpene hydrocarbons and two sesterterpene alcohols. Computational analysis of the cyclization mechanism of NfSS, which produced a most elaborated sesterterpene alcohol, revealed two kinetically and thermodynamically favorable pathways. The cyclization mechanism was supported by the analysis of in vivo and in vitro enzymatic reactions with isotopically labeled precursors. Interestingly, the structures of carbocation intermediates proposed in one of the computed pathway were in good agreement with those of isolated compounds, suggesting that functionally characterized BFTSs follow the cyclization mechanism. Additionally, co-expression of those BFTSs with modification enzyme genes enabled us to isolate oxidative products.