All naturally produced terpenes are
derived from two universal
C5 diphosphate precursors, dimethylallyl diphosphate (DMAPP) and isopentenyl
diphosphate (IPP). Various prenyl transferases use DMAPP to prenylate
aromatic compounds, while others, in combination with IPP, lead to
the enzymatic formation of geranyl diphosphate (GPP), farnesyl diphosphate
(FPP), geranylgeranyl diphosphate (GGPP), and geranylfarnesyl diphosphate,
the direct precursors of monoterpenes (C10), sesquiterpenes (C15),
diterpenes (C20), and sesterterpenes (C25), respectively. FPP and
GGPP are also the basis for the biosynthesis of triterpenes (steroids)
and tetraterpenes (carotenoids), respectively. Nature has developed
two biosynthetic pathways to produce DMAPP and IPP, the mevalonate
(MEV) pathway and the methylerythritol phosphate (MEP) pathway. Both
use compounds derived from glucose through glycolysis, and 18 enzymes
are involved to generate both DMAPP and IPP. Here, we sought to simplify
biochemical access to these two universal diphosphates using the two
commercially and industrially available C5-OHs, dimethylallyl alcohol
and isopentenol (IOH), as starting substrates, as well as two enzymes,
selected from a diverse choice, able to carry out the double phosphorylation
of these two C5-OHs at room temperature using ATP as a phosphate donor.
The first phosphorylation is performed by a promiscuous acid phosphatase
(AP), used in the reverse reaction mode, whereas the second is performed
by the recently described isopentenyl phosphate kinase (IPK). We show
the interest of this artificial biosynthetic terpene mini-path (TMP)
by testing it in a three-enzyme cascade, leading to the formation
of the cytotoxic prenylated diketopiperazine tryprostatin B (TB) from
chemically synthesized brevianamide F (BF), using FtmPT1 prenyltransferase
as a biocatalyst, in addition to the two previously mentioned kinases.
We first performed the proof of concept of this simplified pathway
in vivo (Escherichia coli), using already
described enzymes, that is, an AP from Salmonella enterica and an IPK from Thermoplasma acidophilum. The complete conversion of BF (3.3 mM, 1 g/L) to TB was obtained
after optimization of culture conditions and process parameters. Following
this first success, we performed a screen in search of highly active
phosphatases and IPKs to develop the TMP in vitro. A highly active
AP from Xanthomonas translucens and
an IPK from Methanococcus vannielii were selected from these screens, allowing the in vitro development
of the TMP. Under optimized conditions, the three-enzyme cascade led
to the total transformation of BF (10 mM, 3.3 g/L) to TB in less than
24 h, establishing the in vitro utility as well as the in vivo utility
of the TMP. The implementation of this biosynthetic TMP offers thus
the possibility to access virtually any terpene structure using two
easily commercially and industrially available compounds in bulk,
either in vivo or in vitro, and is thus a viable alternative to the
natural MEV and MEP pathways for bioaccess to terpenes.