Artemisinin degradation

In connection with studies on anthnalarial compounds, simpler mimics of artemisinin based on substituted 1,2,4-trioxepanes were examined. Examples include the 1,2,4-trioxepanes 152, 153 and 154, with the seven-membered ring being made by acid catalysed condensation of the appropriate ketone with the hydroxy hydroperoxide 151. Unfortunately the 1,2,5-trioxepanes were not active as antimalarials in vitro (up to 1000 nM) probably due to their resistance to Fe(II)-mediated degradation <06BMCL6124>. 

Posner and coworkers proposed that the highly electrophilic epoxide could not be isolated due to its inherent instability but was a potent alkylating agent responsible for parasite death. However, Wu and coworkers have isolated a small quantity (1-2% yield) of the epoxide 16 in their iron(II) degradations of artemisinin using iron(II) sulphate in aqueous acetonitrile s. These authors conclude that it is not the active killing species since any external nucleophiles would have to compete with the in-built nucleophile (the OH moiety). Moreover, Avery and coworkers also concluded that the epoxide could not... 

Wu and coworkers provided the first direct evidence for the formation of a secondary radical intermediate from artemisinin . By degrading artemisinin in the presence of a spintrapping agent, 2-methyl-2-nitrosopropane, they were able to observe an ESR spectrum characteristic of a secondary radical . In addition, they have recently reported isolation of a cysteine-artemisinin adduct derived from the C4 secondary radical, vide infra. ... 

The results prompted Jefford and coworkers to re-examine the iron(II) degradation of artemisinin in aqueous acetonitrile with iron(II) chloride (Scheme 10), a system they suggested was closer to the physiological conditions than iron(II) bromide in THF. They reported that iron(II) chloride catalysed isomerization of artemisinin to afford the same products identified by Posner (13 and 21), except that deoxyartemisinin 3 was not observed. When the reaction was carried out in the presence of cyclohexene, none of the expected epoxide was produced, which suggested (in sharp contrast to Posner s results) that a high-valent metal oxo species was not involved. 

Robert and Meunier have isolated and characterized covalent adducts of metallopor-phyrins with artemisinin. Initial attempts using an iron(II) heme model were inconclusive, since the strong paramagnetism of the iron did not allow detailed characterization of the adduct and attempts at demetallation resulted in its degradation. Therefore, manganese(II) tetraphenylporphyrin (TPP) was used since it required milder demetallation conditions and the adduct 50a was formed. The suggested mechanism for the alkylation involves a pendent ethyl radical 50 (Scheme 13). 

More recently, Wu and coworkers proposed that heme may not be the only iron containing species capable of initiating degradation of the peroxide bond but free rron(II) could form complexes with amino acids and be greatly activated . They examined the iron(II) degradation of artemisinin in the presence of non-heme iron chelates such as... 

When artemisinin was degraded in the presence of cysteine, they observed that the reaction was rapid and, in addition to the usual products vide supra), an aldehyde 54 was formed. They concluded that 54 could only arise from a pendent ethyl radical 50 (Scheme 15) and therefore provided unequivocal evidence for its involvement as a killing species . 

Artemisinin (qinghaosu) antimalarial activity, 1309, 1313 biological targets, 1311-13 ESR spin-trapping agents, 1291 Feai) degradation, 1283, 1293, 1295-6 heme adducts, 1298, 1311-12 heterolyhc cleavage of peroxide bond, 1301-2, 1309... 

Acid degradation of artemisinin (9) in either methanol or ethanol gives a mixture of compounds which includes the 1,2,4-trioxanes (164a,b). Both compounds (164a,b) were as active in vitro as artemisinin (9) <90H(3l)l0ll>. 

Besides artemisinin more than 150 natural peroxides are known in nature. The presence of the typical peroxide functions is not related to one natural product group and occurs as cychc and acyclic peroxides in terpenoids, polyketides, phenolics and also alkaloids. The most stable are cychc peroxides, even under harsh conditions and artemisinin is a nice example of this. Artemisinin can be boiled or treated with sodiiun borohydride without degradation of the peroxide function. In contrast, acychc peroxides are rather unstable, form hydrogen peroxides and are easily broken by metals or bases. 

---