A series of artemisinin-related endoperoxides was tested for cytotoxicity to Ehrlich ascites tumor (EAT) cells using the microculture tetrazolium (MTT) assay. Artemisinin [1] had an IC50 value of 29.8 microM. Derivatives of dihydroartemisinin [2], being developed as antimalarial drugs (artemether [3], arteether [4], sodium artesunate [5], artelinic acid [6], and sodium artelinate [7]), exhibited a somewhat more potent cytotoxicity. Their IC50 values ranged from 12.2 to 19.9 microM. The presence of an exocyclic methylene fused to the lactone ring, as for artemisitene [9], led to higher cytotoxicity than 1. From the two epimeric 11-hydroxyartemisinin derivatives, the R form 12 showed a considerably higher cytotoxicity than the S form 13. Opening of the lactone ring of 1 dramatically reduced the cytotoxicity. The ether dimer 8 of 2 was the most potent cytotoxic agent, its IC50 being 1.4 microM. The variations in cytotoxicity between the structurally related compounds mostly correlated well with the theoretical capacity of radical formation and stabilization. In some cases lipophilicity or the presence of an electrophilic moiety seemed to have a determinant influence on cytotoxicity. The artemisinin-related endoperoxides showed cytotoxicity to EAT cells at higher concentrations than those needed for in vitro antimalarial activity, as reported in the literature.

From aseptically grown Artembia annua plantlets, shoot cultures were initiated. Using different concentrations of auxine, cytokinine and sucrose, a suitable culture medium was developed, with respect to the growth of the shoots and their artemisinin accumulation. Nitrate concentration and conductivity appeared to be suitable growth parameters. The artemisinin content was measured gas chromatographically. The shoot cultures were maintained in the developed standard medium, consisting of a half concentration of MS-salts with vitamins, 0.2 mg 1 1 BAP, 0.05 mg 1-1 NAA and 1% sucrose. The growth of the shoots and the artemisinin content remained stable for a longer period. They showed considerable photosynthetic activity and generally contained ca. 0.08% artemisinin on a dry weight basis. The highest artemisinin content found was 0.16% in the above mentioned standard medium, but also on the same medium with 0.5% sucrose. Attempts were made to further improve the artemisinin production by varying the medium composition through addition of gibberellic acid or casein hydrolysate; by omitting plant growth regulators; by precursor feeding, i.e. mevalonic acid; by influencing the biosynthesis routing through inhibition of the sterol synthesis by miconazole, naftifine or terbinafine; by changing gene expression with 5-azacytidine or colchicine; and by elicitation, using cellulase, chitosan, glutathione or nigeran. Enhanced artemisinin production was found with 10 mg 1-1 gibberellic acid, 0.5 g 1 -~ casein hydrolysate, 10 mg 1-1 or 20 mg 1-1 naftifine. Relative increases of 154%, 169%, 140% and 120% were found, respectively. Other additions caused the growth to cease and the artemisinin contents to drop.

The sesquiterpenoid artemisinin (3) and its biosynthetic precursors arteannuic acid (1), arteannuin B (2) and artemisitene (4) can be separated and identified by combined gas chromatography/mass spectrometry both as a mixture of reference standards as well as in extracts of Artemisia annua L. From this study a generally applicable gas chromatographic method has been developed for the analysis of these sesquiterpenoids in A. annua. Analysis of plant parts revealed that the highest sesquiterpenoid concentrations were present in the leaves, buds and small green stems, while the main stem contained only small amounts. Low levels of arteannuic acid and arteannuin B were present in the side‐roots. No sesquiterpenoids could be detected in the main root and none of the plant parts accumulated detectable quantities of artemisitene.

The volatile constituents of Artemisia annua L. plants, grown in the field in The Netherlands from seeds of Chinese and Vietnamese origin, were investigated using GC and GC–MS (EI, NICI) analysis. The plants grown from Chinese seeds contained 4.0% (v/w) essential oil on a dry weight (DW) basis, those from Vietnamese seeds, 1.4% (v/w). More than forty compounds were identified. The principal component of the Chinese oil was artemisia ketone (63.9%); other major constituents included artemisia alcohol (7.5%), myrcene (5.1%), α‐guaiene (4.7%) and camphor (3.3%). In the Vietnamese oil the main components were camphor (21.8%) and germacrene‐D (18.3%); other important constituents were β‐caryophyllene (5.6%), trans‐β‐farnesene (3.8%) and 1,8‐cineole (3.1%). In the Vietnamese variety the terpenoid biosynthesis proceeded further towards sesquiterpenes, whereas in the Chinese predominantly monoterpenes were formed. This was also reflected in the artemisinin contents, found in dichloromethane extracts of the herbaceous plant material: 0.17% (DW) in Chinese and 1.00% (DW) in Vietnamese plants. We suggest that the differences in the essential oil composition may be ascribed to the existence of different A. annua chemotypes. In a dichloromethane extract of the roots several non‐volatile sesquiterpenes were found. The main constituent was a new compound, which was identified tentatively as arteannuic alcohol.

The phenylpropanoid derived lignan podophyllotoxin, occurring inPodophyllum species, is used as a starting compound for the chemical synthesis of the antitumour agents etoposide (VP-16-213) and teniposide (VM-26). At present, the availability of this lignan becomes increasingly limited. As an alternative source, cell cultures originating fromPodophyllum hexandrum Royle were initiated. Analysis of the cell extracts using different HPLC systems as well as TLC, indicated the presence of podophyllotoxin. After prepurification of the extracts by means of ITLC, the identity was confirmed by mass spectrometric analysis. Dark-grown cultures accumulated considerable higher amounts of podophyllotoxin in comparison with the light-grown cultures.