An Efficient Chemoenzymatic Synthesis of Dihydroartemisinic Aldehyde

Abstract Artemisinin from the plant Artemisia annua is the most potent pharmaceutical for the treatment of malaria. In the plant, the sesquiterpene cyclase amorphadiene synthase, a cytochrome‐dependent CYP450, and an aldehyde reductase convert farnesyl diphosphate (FDP) into dihydroartemisinic aldehyde (DHAAl), which is a key intermediate in the biosynthesis of artemisinin and a semisynthetic precursor for its chemical synthesis. Here, we report a chemoenzymatic process that is able to deliver DHAAl using only the sesquiterpene synthase from a carefully designed hydroxylated FDP derivative. This process, which reverses the natural order of cyclization of FDP and oxidation of the sesquiterpene hydrocarbon, provides a significant improvement in the synthesis of DHAAl and demonstrates the potential of substrate engineering in the terpene synthase mediated synthesis of high‐value natural products.

The sesquiterpenoid endoperoxide artemisinin (1)i sw idely used as af irst-line treatment for malaria in combination therapy. [1] Although elegant organic syntheses of artemisinin have been published, [2] thew orldwide supply of 1 predominantly relies on extraction from the plant Artemisia annua. [3] Thed emand for artemisinin is mainly from the developing world, which requires the drug to be produced at low cost. Currently the most efficient way to synthesize artemisinin is to combine biosynthesis with chemical steps.C entral to the biosynthesis of 1 (Scheme 1) is the class Is esquiterpene cyclase amorphadiene synthase (ADS), which catalyzes the conversion of (E,E)-farnesyl diphosphate (FDP, 2)i nto amorpha -4,11-diene (3). In this complex reaction cascade, two 6-membered rings,f our stereocentres,a nd two double bonds are formed with exquisite regio-and stereochemical control in one step. [4] Dihydroartemisinic aldehyde (DHAAl, 4)c an be made from amorpha -4,11-diene (3)e ither through at hree-step chemical synthesis or by combining ab iooxidation with two chemical steps. [5] Compound 4 can then be converted into 1 chemically or enzymatically in four well-established steps. [2a, 6] It is noteworthy that an elegant semisynthetic pathway has been developed that uses ADS and five other enzymes in yeast to produce artemisinic acid (5), which is then converted into dihydroartemisinic acid (6)b yt ransition metal-catalyzed hydrogenation. Thep harmaceutical company Sanofi scaled up this process in 2014 but the manufacture was discontinued owing to strong market forces, [7] thus highlighting the need for ecologically friendly, low-cost alternatives for the production of artemisinin.
Herein, we report an ovel chemoenzymatic process that exploits the substrate promiscuity of ADS to convert the hydroxylated FDP analogue 7 into the synthetic intermediate DHAAl (4). In contrast to existing procedures,t his process, which reverses the natural order of cyclization of FDP and oxidation of the sesquiterpene hydrocarbon, significantly shortens the synthesis of dihydroartemisinic aldehyde (4), in that it uses only one enzyme and requires as ingle oxidation step that occurs prior to the ADS-catalyzed cyclization to 4. Thep rocess avoids several redox steps after the cyclization since it bypasses the formation of the intermediate amorphadiene (3)a ltogether (Scheme 1).
Many sesquiterpene synthases display some degree of substrate promiscuity,and are able to convert methylated and fluorinated farnesyl diphosphate analogues into modified terpenoids. [9][10][11] In this work, we investigated for the first time the potential of asesquiterpene synthase to accept ahydroxylated FDP analogue as as ubstrate.I np articular, we asked whether ADS [11b] could convert 12-hydroxyfarnesyl diphosphate (7; [12] Scheme 1) into dihydroartemisinic aldehyde (4) via a1 2-hydroxyamorphyl cation (OH-9;S cheme 3). It is known that a-hydroxylated carbocations such as OH-9 can isomerize under acidic conditions to aldehydes. [13] The electrophilic nature of terpene synthase chemistry combined with the inherent reactivity of a-hydroxylated carbocations such as OH-9 should therefore lead to aldehyde 4.
To further advance this novel chemoenzymatic approach towards the production of artemisinin, the 2:3m ixture of (11R)a nd (11S)-4 was oxidized to the corresponding dihydroartemisinic acids (6)i n9 3% yield (Scheme 6), [2b] and subsequently converted into the corresponding dihydroartemisinic methyl esters (18)i n9 4% yield ( Figure S11). [18] Tr eatment of the esters with LDAu nder kinetic control resulted in a1 :1 mixture of (11R)-and (11S)-methyl esters ( Figure S28). (11R)-dihydroartemisinic methyl ester is the desired intermediate for the synthesis of artemisinin. [19] In conclusion, we have developed an efficient chemoenzymatic route to dihydroartemisinic aldehyde (4), am ajor intermediate in the production of artemisinin (1), the most important drug for the treatment of malaria. We have for the first time shown that hydroxylated FDP analogues can be accepted as substrates by sesquiterpene synthases.Hence our work offers anovel "reversed biosynthetic" approach for the synthesis of functionally diversified hydroxylated terpenoids. Thec hemical synthesis of such products is often difficult owing to the need to make several small rings with high stereo-and regiocontrol. Ther elatively high substrate promiscuity and the templating effect of the active site allow terpene synthases to chaperone unnatural substrates along well-defined reaction paths to specific products with high fidelity.T his design offers ap romising approach for the production of high-value terpenoids and terpene alkaloids [11a]  Scheme 5. ADS-catalyzed production of (11S)-4 from 13-acetoxyfarnesyl diphosphate ( 16)via carbocation 17. Scheme 6. Synthesis of dihydroartemisinic acid methyl ester (18).

Angewandte Chemie
Communications www.angewandte.org that is complementary to conventional synthetic and biosynthetic procedures.