Plant Science 158 (2000) 163–171

Cloning and molecular analysis of two new sesquiterpene cyclasesfrom Artemisia annua L.�

Els Van Geldrea, Isabel De Pauwa, Dirk Inzeb, Marc Van Montagub,Elfride Van den Eeckhouta,*

a Laboratory for Pharmaceutical Biotechnology, Uni6ersity of Ghent, Harelbekestraat 72, 9000 Ghent, Belgiumb Laboratory for Genetics, Uni6ersity of Ghent, K. L. Ledeganckstraat 35, 9000 Ghent, Belgium

Received 5 April 2000; received in revised form 15 June 2000; accepted 15 June 2000

Abstract

Artemisia annua L. is the only source of artemisinin, a new promising antimalarial drug (Qinghaosu Antimalarial CoordinatingResearch Group, Chin. Med. J. 92 (1979) 811). Our efforts are focused on the overproduction of this valuable medicine by geneticengineered A. annua plants. Therefore, we decided to isolate the gene(s) encoding sesquiterpene cyclase(s) in A. annua as a firststep in improving artemisinin yield. Four partial genomic clones, gASC21, gASC22, gASC23 and gASC24, were isolated throughpolymerase chain reaction (PCR) with degenerated primers based on homologous boxes present in sesquiterpene cyclases fromdivergent sources. Intron–exon organisation of those partial genomic clones was analysed and it was shown that A. annuacontains a gene family for sesquiterpene cyclases. Based on gASC21, gASC22, gASC23 and gASC24 sequences, the full-lengthcDNA clones cASC34 and cASC125 were subsequently isolated by rapid amplification of cDNA ends PCR. The derived aminoacid sequences of both full-length clones show high homology with sesquiterpene cyclases from plants. Reverse transcription-PCRanalysis revealed transient and tissue specific expression patterns for cASC34 and cASC125, in contrast to the constitutivelyexpressed 8-epicedrol synthase, a previously reported sesquiterpene cyclase from A. annua. Both cASC34 and cASC125 could onlybe detected in flowering plants when artemisinin concentration is at highest. © 2000 Elsevier Science Ireland Ltd. All rightsreserved.

Keywords: Terpene cyclases; Artemisinin; Malaria

www.elsevier.com/locate/plantsci

1. Introduction

Artemisia annua L., a member of the Composi-tae family, has been used for decennies in thetreatment of fever and malaria [1]. Analysis of theactive compounds responsible for these pharmaco-logical activities led to the discovery of a sesquiter-pene lacton, called artemisinin or qinhaosu [1].This molecule possesses a different mode of actionin comparison with existing antimalarial drugs(e.g. quinine) and is active against chloroquineresistant forms of Plasmodium falciparum [2].Based on this secondary plant metabolite, newsemi-synthetic drugs for the treatment of malaria,e.g. artemether, arteether and artesunate, are pro-duced [1]. Unfortunately, only very low amounts

Abbre6iations: A, adenine; aa, amino acid(s); bp, base pair(s); C,cytosine; CAD, cadinene synthase; cDNA, DNA complementary toRNA; dNTP, deoxyribonucleotide triphosphate; FPP, farnesylpy-rophosphate; G, guanine; M, A/C; PCR, polymerase chain reaction;pI, isoelectric point; R, A/G; RT-PCR, reverse transcription-poly-merase chain reaction; RACE-PCR, rapid amplification of cDNAends polymerase chain reaction; S, C/G; Taq, Thermus AquaticusPolymerase; TEAS, 5-epi-aristolochene synthase; T, thymidine; VS,vetispiradiene synthase; Y, C/T.� The nucleotide sequence data reported in this article have been

submitted to the GenBank/EMBL Database under Accesssion Num-bers AJ271792, AJ271793, AJ276412, AJ276413, AJ276414,AJ276415.

* Corresponding author. Tel.: +32-9-2648053; fax: +32-9-2206688.

E-mail address: elfriede.vandeneeckhout@rug.ac.be (E. Van denEeckhout).

0168-9452/00/$ - see front matter © 2000 Elsevier Science Ireland Ltd. All rights reserved.

PII: S 0 1 68 -9452 (00 )00322 -8

E. Van Geldre et al. / Plant Science 158 (2000) 163–171164

of artemisinin can be isolated from the aerialparts of A. annua (between 0.01 and 0.6% dryweight) [3–5]. The limited supply of this valu-able compound from A. annua L. prompted re-searchers to develop alternate means ofproduction. Many attempts have been under-taken to overproduce artemisinin through cell ortissue cultures [6–9]. However, so far, only verylimited success has been obtained. Also, totalchemical synthesis of artemisinin was revealed tobe not economically interesting due to stereo-chemical problems [10]. Therefore, all efforts arefocused on the overproduction of this valuablecompound by genetic engineered A. annuaplants. Such plants can be obtained through ge-netic improvement of the biosynthetical pathwayof artemisinin by sense and/or antisense tech-niques [11]. An Agrobacterium tumefaciens-medi-ated transformation procedure has already beenestablished in our laboratory [11–13]. In thebiosynthesis of artemisinin, very little is knownabout enzymes responsible for the production ofthe different intermediates. At one of the impor-tant branchpoints farnesylpyrophosphate (FPP)is cyclisised by the action of a sesquiterpene cy-clase to amorpha-4,11-diene, which leads to se-quentially to artemisinic acid, dihydroartemisinicacid and artemisinin by a series of oxidationsteps and nonenzymatic conversions [3,14–17](Fig. 1). Thus, the cyclase reaction establishes an

important stereochemical framework upon whichall other chemical modifications take place [17].Several publications report on the fact that thiscyclisation step is a putative regulatory point,probably rate limiting [17,18]. The accumulationof artemisinic acid and dihydroartemisinic acidin absence of any intermediates en route fromFPP also supports this hypothesis [14]. Farnesyldiphosphate synthase, which produces FPP, hasalready been cloned from A. annua [19], but thisenzymatic reaction is probably not rate limitingfor the production of artemisinin.

Therefore, we decided to isolate the gene(s)encoding sesquiterpene cyclase(s) in A. annua asa first step in improving artemisinin yield. In1992, tobacco 5-epi-aristolochene synthase(TEAS) was cloned from elicitor-treated tobaccocell cultures [20]. Based on the TEAS sequenceinformation, vetispiradiene synthase fromHyoscyamus muticus [17] and (+ )-delta-cadinenesynthase from Gossypium arboreum [21] werecloned. During the course of this work, the iso-lation of 8-epicedrol synthase from A. annua wasreported [22,23]. The role of 8-epicedrol synthasein A. annua and artemisinin biosynthesis is un-certain, since 8-epicedrol cannot be detected bygas chromatography–mass spectrometry in A.annua extracts and the conversion of epicedrol tothe putative artemisinin bioprecursor artemisinicacid would require rearrangements that do notseem plausible [22,23].

Fig. 1. Proposed biosynthetic pathway for artemisinin in A. annua L. (adapted from Refs. [14,27]).

E. Van Geldre et al. / Plant Science 158 (2000) 163–171 165

This paper reports the isolation of two othermembers of the sesquiterpene cyclase gene familyfrom A. annua. Both full-length isolated genes arenew potential candidates involved in the rate-limit-ing cyclisation step of artemisinin biosynthesis,thereby enabling a molecular approach for im-provement of artemisinin production.

2. Materials and methods

2.1. Materials

Restriction endonucleases were purchased fromBoehringer Mannheim (Mannheim, Germany) andBiolabs (New England). Taq DNA polymerasewas purchased from Perkin Elmer (Norwalk, CT,USA). ABI PRISM™ Ready Reaction Dye De-oxy™ Terminator Cycle Sequencing Kit and ABIPRISM™ dRhodamine Terminator Cycle Se-quencing Ready Reaction Kit with AmpliTaq®

DNA Polymerase FS were from Applied Biosys-tems. Dynabeads mRNA Direct Kit was fromDynal (Oslo, Norway). The SMART RACE PCRKit was from Clontech (Palo Alto, CA, USA).Superscript II Reverse Transcriptase was fromGibco BRL (Life Technologies). Gene Images ran-dom priming labeling module and CDP-star detec-tion module was from Amersham Life Science(Little Chalfont, Buckinghamshire, UK). Es-cherichia coli strain JM109 was used as bacterialhost for pGem-T-Easy recombinant plasmids.Wizard SV DNA extraction, dNTPs, SP6 Promo-tor primer, Reverse Transcription System andpGem-T-Easy vector were purchased fromPromega (Madison, USA). M13/pUC universalsequencing primer 17-mer was from BoehringerMannheim.

2.2. Plant materials

Seeds of A. annua of West Virginia origin werekindly provided by the Walter Reed Army Insti-tute of Research (Washington). Plants were grownin an experimental greenhouse using Hg- and Na-vapor lamps 16 h/day and at a temperature of22°C and a relative humidity of 40%.

2.3. Isolation of nucleic acids

Genomic DNA was isolated from plant leaves

according to the extraction procedure from Staceyand Isaac [24]. Plasmid DNA samples were pre-pared by an alkaline lysis method with the WizardDNA Purification System (Promega) according tothe manufacturer’s instructions. Total RNA wasisolated with the Trizol reagent according the pro-cedures of Life Technologies. mRNA could beisolated with Dynabeads from Dynal (Oslo,Norway).

2.4. Oligonucleotide primer design

Comparison of the amino acid sequences fromsesquiterpene cyclases isolated from divergentsources Nicotiana tabacum, H. muticus and G.arboreum, revealed highly conserved boxes forwhich degenerate primers were designed (PILEUPand LINEUP from GCG package). These boxeswere used to design degenerate primers by back-translation from the respective amino acid se-quences. At places of three- and fourfold baseredundancy, inosines were incorporated to preventhigh degeneracy of the primers. One forwardprimer (5%-TAYCAYTTYGARATIGARGA-3%)and one reverse primer (5%-GARTAYTGIG-GTTCRAARTAIACICC-3%) (R=A, G; Y=C,T) were thus selected with respective anealing tem-peratures of 62°C and 68°C calculated with[2°*(A+T)+4°*(C+G)].

2.5. Polymerase chain reactions

The primers were subsequently used in a HotStart polymerase chain reaction (PCR) on ge-nomic DNA from A. annua. (10 mM Tris–HCl(pH 8.3), 25°C, 50 mM KCl, 200 mM dNTP, 1.2 UTaq polymerase and 20 pmol of each primer in atotal volume of 50 ml; concentrations of MgCl2ranged from 1.5 to 4.5 mM). Cycling conditionswere set as follows: 94°C for 1 min, 50°C for 1 minand 72°C for 1 min for a total of 35 cycles. ThisPCR resulted in a predominant 900 base pair (bp)product as detected by agarose gel electrophoresisand ethidium bromide staining. This product wassequenced after TA-cloning in pGem-T-Easy Vec-tor System (Promega) according to the manufac-turer’s instructions. Recombinant plasmids wereisolated after amplification in E. coli JM109 andanalysed for insert by blue/white screening andPCR.

E. Van Geldre et al. / Plant Science 158 (2000) 163–171166

2.6. Nucleotide sequencing and analysis of nucleicacids

Double-stranded forms of template DNA wereused in dideoxynucleotide cycle sequencing reac-tions. Both sense and antisense strands of theTA-cloned PCR products were sequenced withvector-specific primers (M13/pUC universal se-quencing primer 17-mer or SP6 Promotor primer)or ABI PRISM™ Ready Reaction Dye Deoxy™Terminator Cycle Sequencing Kit and ABIPRISM™ dRhodamine Terminator Cycle Sequenc-ing Ready Reaction Kit with AmpliTaq® DNAPolymerase FS (Applied Biosystems) using an auto-mated fluorescence-based system ABI373 orABI310 (Applied Biosystems), respectively. DNAsequences and the deduced amino acid sequenceswere analyzed and aligned with related sequencesusing programs available through the GeneticsComputer Group package (University of Wiscon-sin).

2.7. Southern blot

Twenty micrograms of genomic DNA were blot-ted onto a+charged nylon membrane (BoehringerMannheim) according to standard procedures [25].Prehybridisation and hybridisation were performedat 55°C. The probe, fluoresceine labeled, consistedof the partially purified cDNA sequence alreadyisolated in A. annua (Gene Images Random priminglabeling and CDP-star detection).

2.8. RNA analysis

In order to isolate corresponding cDNA se-quences, new degenerate primers based on thepartial genomic sequences were designed. One for-ward (5%-AAGCMCTTCSCTYTGGTTCMGA-3%)and one reverse primer (5%-CCAAAARTAR-CMTTCAACYATTCT-5%) were selected with theprogram OLIGO™ (M=A or C; S=C or G; Y=Cor T). Dnase-treated RNA isolated from leaves andflowers harvested on different stages during plantdevelopment (from young cotyledons to floweringplants) was denatured for 2 min at 90°C andsubsequently transcribed to cDNA with AMV Re-verse Transcriptase during 60 min at 42°C. Thisreaction could be used immediately for PCR with20 pmol of both primers and 2.5 U Taq in 10 mMTris–HCl (pH 8.3), 25°C, 50 mM KCl and 200 mM

dNTP. The reaction conditions were 94°C during 30s, 53°C during 30 s and 72°C during 1 min for a totalof 40 cycles. For detection of specific transcripts byreverse transcriptase (RT)-PCR exact primers (5%-GGCTATAAACTTGCGCTAGTAG-3% and 5%-CCGGATCCTCATGTAATGATGGCATC-3% for8-epicedrol synthase; 5%-CCTCACCAACGATG-TAGAAG-3% and 5%-A-GGTAGATTGTTTGGG-ACATC-3% for cASC34; 5%-TCTGTATGAAG-CGGCATTTATG-3% and 5%-GTCTCGAACATA-AGGAAGCTTA-3% for cAS-C125) were designedon three isolated sequences and used for PCR withthe following cycling conditions: 1 min at 94°C, 1min at 60°C and 1 min at 72°C for 30 cycles.Repeating the experiment twice with newly isolatedRNA from fresh plant samples compensated thelack of an internal constitutively expressed control.

2.9. Rapid amplification of cDNA ends PCR

Both 5% rapid amplification of cDNA ends(RACE) and 3%RACE PCR were performed withgene-specific primers designed against partial se-quences obtained through RT-PCR with theSMART-RACE PCR kit (Clontech) according tothe manufacturer’s instructions. Primer sequenceswere as follows: for cASC32, 5%RACE 5%-AGGTA-GATTGTTTGGGACATC-3% and nested primer5%-CTCAAATATGTTGCCTCGTACAGC-3%; 3%-RACE 5%-CCTCACCAACGATGTAGAAG-3%;for cASC125, the primers used for 5%-RACE were5%-GTCTCGAACATAAGGAAGCTTA-3% andnested primer 5%-CCCTCATAAATGCCGCTTC-ATACAG-3%, and 3%-RACE was performed with5%-TCTGTATGAAGCGGCATTTATG-3%.

3. Results and discussion

3.1. Cloning of partial genomic clones gASC21,gASC22, gASC23 and gASC24

Molecular comparisons of plant terpene cyclaseshave revealed a striking level of sequence similarity[17,26,27]. After comparison of the amino acidsequences of sesquiterpene cyclases isolated from N.tabacum [20], G. arboreum [21] and H. muticus [17],different degenerate primers were subsequentlydesigned against conserved boxes highlighted bythis comparison. RT-PCR with these primerson mRNA isolated from young A. annua

E. Van Geldre et al. / Plant Science 158 (2000) 163–171 167

Table 1Identities (%) and similarities (%) (in parentheses) obtained through the BLASTX/BEAUTY search for gASC21–gASC24,cASC34 and cASC125 with corresponding amino acid sequences of sesquiterpene cyclases from divergent sourcesa

CAD1A VS GSTEAS 8-epicedrol CAD VS1synthase

G. arboreum H. muticus L. esculentumN. tabacum A. annua G. hirsutum S. tuberosum51 (68) 46 (62) 56 (74) 60 (72)gASC21 50 (71)47 (63) 49 (65)54 (71) 50 (67) 53 (72)50 (68) 70 (82)gASC22 53 (70) 50 (67)

40 (58)gASC23 40 (61) 40 (59) 40 (61) 53 (67) 39 (61) 39 (60)56 (68)gASC24 64 (82) 57 (72) 59 (79) 71 (78) 62 (81) 58 (73)

47 (66) 43 (65) 45 (65)42 (63) 57 (76)cASC34 47 (65) 45 (64)cASC125 40 (60) 42 (61) 40 (61) 41 (62) 49 (68) 41 (61) 40 (61)

a CAD, Cadinene synthase; VS, vetispiradiene synthase.

L. plants remained unsuccessful (data not shown).Therefore, analysis of genomic DNA has beenperformed. This resulted in a specific PCR productthat appeared to be a mixture of four differentgenomic fragments (gASC21, gASC22, gASC23and gASC24), respectively, 841, 821, 788 and 841bp in length. Alignment of the deduced amino acidsequences for these four different partial genomicclones revealed high identities and similarities withother cloned sesquiterpene cyclases, as shown inTable 1. These results clearly suggested that wecloned partial genomic sequences encoding differ-ent sesquiterpene cyclases from Artemisia annua(see Table 2 for identities of four partial clones onnucleotide level). Moreover, the fact that fourdifferent but clearly homologous sequences havebeen cloned suggested that we are dealing with acomplex gene family of sesquiterpene cyclases inA. annua L., which has also been confirmed by aSouthern blot (data not shown). This was notsurprising since H. muticus contains a small genefamily of approximately six to eight genes [17],and 12–15 epi-aristolochene synthase like genesare present in the tobacco genome [20]. From G.arboreum, several cadinene synthase genes havealready been cloned, indicating that this organismalso contains a gene family for sesquiterpene cy-clase [21]. Solanum tuberosum appears also to con-tain different veti-spiradiene synthase genes [28].

Comparison of intron–exon organisation ofgASC21, gASC22, gASC23 and gASC24 with cor-responding sequences of TEAS from N. tabacumand vetispiradiene synthase from H. muticus re-vealed that intron positions (Fig. 2) and sizes (notshown) of the A. annua clones were consistent withearlier reported results [17,20]. This correlates withthe fact that plant sesquiterpene cyclases have

highly conserved intron–exon organisation of ge-nomic DNA [17]. This conservation can even beextended to the diterpene cyclases [29].

3.2. Detection of sesquiterpene cyclase transcripts

Based upon the four partial genomic sesquiter-pene cyclase sequences, a new set of degeneratedprimers was designed for RT-PCR. mRNA iso-lated from leaves and flowers of A. annua L. plantsat different moments of development served as atemplate for RT-PCR. We detected a specificproduct both in leaves and flowers of floweringplants (data not shown). Also, in Fragaria 6escaL., a partial sequence encoding sesquiterpene cy-clase was isolated along with seven other ripening-induced cDNAs by differential screening of acDNA library [30]. The author states that theabsence of a detectable signal in commercialstrawberry either suggests that sesquiterpene cy-clase does not play an important role in the ripen-ing of the fruits or that the transcript levels are toolow in commercial strawberries [30]. It is possiblethat this phenomenon also takes place in A. annua,and that sesquiterpene cyclase transcripts are onlydetectable in flowering plants, when expressionlevels are highest.

Table 2Identities (%) for gASC21, gASC22, gASC23 and gASC24 atthe nucleotide level

gASC22 gASC23 gASC24

51 45gASC21 8943gASC22 51

gASC23 45

E. Van Geldre et al. / Plant Science 158 (2000) 163–171168

Fig. 2. Alignment of sesquiterpene cyclases isolated from divergent plant sources (PILEUP GCG package, University ofWisconsin). cASC34 and cASC125, new sesquiterpene cyclases isolated from A. annua ; gASC21–gASC24, corresponding AAsequences deduced from respective partial genomic sequences gASC21–gASC24; Aa, A. annua 8-epicedrol synthase; Hmut, H.muticus VS; Ntab, N. tabacum TEAS; Garb, G. arboreum CAD. Conserved amino acid sequences used for primer design aremarked with a frame. �, Positions of introns in partial genomic sequences gASC21–gASC24; �, introns in TEAS and VS. Theshading colors of the residues correspond to the level of similarity: black, 100% similarity; dark gray, ]80% similarity; light gray,]60% similarity.

E. Van Geldre et al. / Plant Science 158 (2000) 163–171 169

Two partial cDNA sequences were obtained(cASC34-p and cASC125-p), of 465 and 462 bp,respectively. Comparison of the deduced aminoacid sequences revealed that gASC22 andcASC34-p were identical. The corresponding ge-nomic sequence of cASC125-p, however, has notyet been isolated. cASC125-p showed only identi-ties between 49 and 61% and similarities between62 and 77% with the four partial genomic se-quences. Identities and similarities with sesquiter-pene cyclases cloned from N. tabacum, H. muticus,G. arboreum, Gossypium hirsutum, S. tuberosumand Lycopersicon esculentum were comparablewith those obtained for the gASC21 to gASC24(identities ranging from 45 to 8% and 46 to 58%;similarities between 65 and 81% and 65 and 78%,for cASC34-p and cASC125-p, respectively).These values clearly indicate that cASC34-p andcASC125-p are partial cDNA sequences encodingtwo different members of the A. annua gene familyfor sesquiterpene cyclases. The fact that, forcASC125-p, the corresponding genomic sequenceescaped isolation thus far can be due to the highdegeneracy of the primers used for PCR on ge-nomic A. annua DNA. Out of ten partial genomicsequences isolated and analyzed, we identified fourdifferent sesquiterpene cyclases. In addition, itcould also be possible that not all members of thesesquiterpene cyclase gene family are active at theflowering time of A. annua or that some membersof the gene family stay below detection limits. It isnot clear at which moment of plant developmentthe artemisinin biosynthetical pathway is acti-vated. Studies on influences of elicitors onartemisinin production have been established, butthe influence on enzymes related to this biosynthe-sis has not been analyzed. Analysis of artemisininyield estimated at different stages of developmentreveals a positive correlation between plant ageand artemisinin yields, without any peak produc-tion at any moment of the plant development[3,31–33]. This probably means that enzymes re-lated to artemisinin biosynthesis are active duringa large part of plant development, but probablystay below detection limits at mRNA level. RNAtranscripts of previously isolated sesquiterpene cy-clases from N. tabacum, H. muticus and G. ar-boreum could only be detected after induction byelicitor treatment [17,20,21], and UV treatmentwas necessary to detect sesquiterpene cyclaseprotein in pepper [26].

3.3. Cloning of cASC34 and cASC125

Using cASC125-p and cASC34-p as a probe toscreen a cDNA library of A. annua, we did notobtain any clone containing the correspondingfull-length sequence of the respective probe used,but we cloned a full-length sequence encodinganother member of the sesquiterpene cyclase fam-ily (unpublished results). By the time we wereperforming expression experiments with theprotein encoded by this gene, two other researchgroups reported the isolation of 8-epicedrol syn-thase from A. annua [22,23], which appeared to beexactly the same sequence as our full-length clone.But this sesquiterpene cyclase is apparently notinvolved in artemisinin biosynthesis [22,23].

Since we were unable to clone the correspondingfull length sequence of cASC34-p and cASC125-pfrom our cDNA library, 5%RACE and 3%RACEPCR for the full-length cloning of both genes wasperformed.

For cASC34, a sequence of 2026 bp was ob-tained containing 38 bp of the untranslated 5% endwith an ORF encoding a 549 amino acid (aa)protein of 64.2 kDa and a calculated isoelectricpoint (pI) of 5.28. Two AATAAT–poly-adenyla-tion signals were found in the 338 bp untranslated3% sequence (positions, 1876 and 2003 bp).

The cASC125 sequence contained an ORF of1734 bp encoding a 577 aa protein with a molecu-lar weight of 67.4 kDa and a pI of 5.50. Thissequence has 60 bp at the 5% untranslated end andan untranslated 3% sequence of 48 bp. Here, wefound the poly-adenylation signal AATTAA at1820 bp.

Comparison of deduced amino acid sequencesfrom cASC34 and cASC125 with cloned sesquiter-pene cyclases from divergent sources revealed highsimilarities and identities as shown by Table 1 andFig. 2.

3.4. RNA analysis for three sesquiterpene cyclasesfrom A. annua

Fig. 3 shows the expression in leaves and flow-ers of 8-epicedrol synthase, cASC34 and cASC125during plant development. We were able todemonstrate that 8-epicedrol synthase is expressedduring the whole plant development in leaves andflowers, but that cASC125 is only present inmRNA-extracted from flowers. cASC34, on the

E. Van Geldre et al. / Plant Science 158 (2000) 163–171170

Fig. 3. RNA analysis for 8-epicedrol synthase (A), cASC34(B) and cASC125 (C).

artemisinin biosynthesis. The enzymes related tothis biosynthetical pathway are possibly more ac-tively transcribed with as a consequence that theirmRNA transcripts are detectable during this mo-ment of plant development. Further experimentsare necessary, and bacterial expression analysis ofcASC34 and cASC125, in order to determine theirexact enzymatic activity, will be performed in thenear future. On the other hand, if cASC34 orcASC125 do not possess amorpha-4,11-diene ac-tivity, the search for the gene(s) encoding a proteinwith this enzymatic activity will be continued byuse of the partial genomic sequences gASC21,gASC23 and gASC24 reported here.

Acknowledgements

The authors thank Prof. L. Gheysen and allmembers of the Nematode Group (Laboratory ofGenetics, University of Ghent, Belgium) for assis-tance during the course of this work, ChristopheGilot (Laboratory of Genetics, University ofGhent, Belgium) for synthesis of primers, andArdiles Diaz Wilson and Villarroel Raimundo(Laboratory of Genetics, University of Ghent, Bel-gium) for sequencing analysis of cASC34 andcASC125. Prof. Chappell is gratefully acknowl-edged for the gift of pBSK-TEAS, used in apreliminary stage of our experiments, and Dr D.L.Klayman for providing Artemisia annua L. seeds.

References

[1] Qinghaosu Antimalarial Coordinating Research Group,Chin. Med. J. 92 (1979) 811.

[2] S.R. Meshnick, The mode of action of antimalarialendoperoxides, R. Soc. Trop. Med. Hygiene 88 (Suppl.1) (1994) 31–32.

[3] D.R. Vandenberghe, A.N. Vergauwe, M. Van Montagu,E.G. Van den Eeckhout, Simultaneous determination ofartemisinin and its bioprecursors in Artemisia annua, J.Nat. Prod. 58 (1995) 798–803.

[4] M.R. Morales, D.J. Charles, J.E. Simon, Seasonal accu-mulation of artemisinin in Artemisia annua L., Acta.Hort. 344 (1993) 416–420.

[5] J.F.S. Ferreira, J. Janick, Immunoquantitative analysisof artemisinin from Artemisia annua using polyclonalantibodies, Phytochemistry 41 (1996) 97–104.

[6] M.S. Nair, N. Acton, D.L. Klayman, Production ofartemisinin in tissue cultures of Artemisia annua, J. Nat.Prod. 49 (1986) 504–507.

other hand, could be detected in leaves of flower-ing plants and flowers. In leaves, cASC34 tran-scripts disappear completely when floweringstopped and seeds are formed. In flowers, bothcASC34 and cASC125 transcripts are still presentwhen flowers faded but also disappear later whenseeds are formed. The fact that, in the RT-PCRwith degenerated primers, transcripts were onlydetected in RNA isolated from flowering plantscan be explained by the high degeneration of theprimers. Primers designed based on gASC21-24were probably not compatible with the 8-epicedrolsynthase sequence and did not amplify the mRNAtranscripts corresponding to this enzyme.

4. Conclusion

Recently, Bouwmeester et al. [14] described anamorpha-4,11-diene synthase activity in leaf ex-tracts of A. annua. Amorpha-4,11-diene is likely tobe the olefinic intermediate in the biosynthesis ofartemisinin, the first step after FPP [14]. The au-thors partially purified and characterised the en-zyme responsible for this activity, but did notisolate the corresponding gene(s). If cASC34 and/or cASC125 possess amorpha-4,11-diene synthaseactivity, this would mean a big step forward in theoverproduction of artemisinin through genetic ma-nipulation. Artemisinin is mainly present in leavesand flowers of A. annua L. plants, with the highestcontent around flowering time [31,32]. Therefore,the fact that cASC34 transcripts are present inthese plant parts at flowering time could be anindication for relationship of this gene to the

E. Van Geldre et al. / Plant Science 158 (2000) 163–171 171

[7] G.D. Brown, Secondary metabolism in tissue culture ofArtemisia annua, J. Nat. Prod. 57 (1994) 975–977.

[8] N.B. Paniego, A.M. Giuletti, Artemisia annua L. dedif-ferentiated and differentiated cultures, Plant Cell TissueOrgan Cult. 36 (1994) 163–168.

[9] A. Whipkey, J.E. Simon, D.J. Charles, J. Janick, In vitroproduction of artemisinin from Artemisia annua L., J.Herbs Spices Medicinal Plants 1 (1992) 15–25.

[10] S.S. Zaman, R.P. Sharma, Some aspects of the chem-istry and biological activiy of artemisinin and relatedantimalarials, Heterocycles 32 (1991) 1593.

[11] A. Vergauwe, E. Van Geldre, D. Inze, M. Van Mon-tagu, E. Van den Eeckhout, Factors influencingAgrobacterium tumefaciens-mediated transformation ofArtemisia annua L., Plant Cell Rep. 18 (1998) 105–110.

[12] A. Vergauwe, C. Cammaert, D. Vandenberghe, C.Genetello, D. Inze, M. Van Montagu, E. Van denEeckhout, Agrobacterium tumefaciens-mediated trans-formation of Artemisia annua L. and regeneration oftransgenic plants, Plant Cell Rep. 15 (1996) 929–933.

[13] A. Vergauwe, E. Van Geldre, D. Inze, M. Van Mon-tagu, E. Van den Eeckhout, The use of amoxicillin andticarcillin in combination with a b-lactamase inhibitor asdecontaminating agents in the Agrobacterium tumefa-ciens-mediated transformation of Artemisia annua L., J.Biotechnol. 52 (1996) 89–95.

[14] H.J. Bouwmeester, T.E. Wallaart, M.H.A. Janssen, B.van Loo, B.J.M. Jansen, M.A. Posthumus, C.O.Schmidt, J.W. De Kraker, W.A. Konig, M.C.R.Franssen, Amorpha-4,11-diene synthase catalyses thefirst probable step in artemisinin biosynthesis, Phyto-chemistry 52 (5) (1999) 843–854.

[15] T.E. Wallaert, Z. Van Uden, H.G.M. Lubberink, H.J.Woerdenbag, N. Pras, W.J. Quax, Isolation and identifi-cation of dihydroartemisinic acid from Artemisia annuaand its possible role in the biosynthesis of artemisinin, J.Nat. Prod. 62 (3) (1999) 430–433.

[16] T.E. Wallaart, N. Pras, W.J. Quax, Isolation and iden-tification of dihydroartemisinic acid hydroperoxide fromArtemisia annua : a novel biosynthetic precursor ofartemisinin, J. Nat. Prod. 62 (8) (1999) 1160–1162.

[17] K. Back, J. Chappell, Cloning and bacterial expressionof a sesquiterpene cyclase from Hyoscyamus muticus andits molecular comparison to related terpene cyclases, J.Biol. Chem. 13 (1995) 7375–7381.

[18] U. Vogeli, J.W. Freeman, J. Chappell, Purification andcharacterization of an inducible sesquiterpene cyclasefrom elicitor-treated tobacco cell suspension cultures,Plant Physiol. 93 (1990) 182–187.

[19] Y. Matsushita, W. Kang, B.V. Charlwood, Cloning andanalysis of a cDNA encoding farnesyl diphosphate syn-thase from Artemisia annua, Gene 172 (1996) 207–209.

[20] P.J. Facchini, J. Chappel, Gene family for an elicitor-in-duces sesquiterpene cyclase in tobacco, Proc. Natl.Acad. Sci. USA 89 (1992) 11088–11092.

[21] X.Y. Chen, Y. Chen, P. Heinstein, V.J. Davisson,Cloning, expression and characterisation of (+ )-deltacadinene synthase: a catalyst for cotton phytoalexinbiosynthesis, Arch. Biochem. Biophys. 324 (2) (1995)255–266.

[22] L. Hua, S.P.T. Matsuda, The molecular cloning of 8-epicedrol synthase from Artemisia annua, Arch.Biochem. Biophys. 369 (2) (1999) 208–212.

[23] P. Mercke, J. Crock, R. Croteau, P.E. Brodelius,Cloning, expression and characterization of epi-cedrolsynthase, a sesquiterpene cyclase from Artemisia annuaL., Arch. Biochem. Biophys. 369 (2) (1999) 213–222.

[24] J. Stacey, P.G. Isaac, Isolation of DNA from plants, in:P.G. Isaac (Ed.), Methods in Molecular Biology. In:Protocols for Nucleic Acid Analysis by NonradioactiveProbes, vol. 28, Humana Press, Towota, NJ, 1994, pp.9–15.

[25] J. Sambrook, E.F. Fritsch, T. Maniatis, MolecularCloning: A Laboratory Manual, 2nd ed., Cold SpringHarbor Laboratory Press, Cold Spring Harbour, NY,1989.

[26] K. Back, S. He, K.U. Kim, D.H. Shin, Cloning andbacterial expression of sesquiterpene cyclase, a keybranch point enzyme for the synthesis of sesquiterpenoidphytoalexin capsidiol in UV-challenged leaves of Cap-sicum annuum, Plant Cell Physiol. 39 (9) (1998) 899–904.

[27] J. Chappell, Biochemistry and molecular biology of theisoprenoid biosynthetic pathway in plants, Ann. Rev.Plant Physiol. Plant Mol. Biol. 46 (1995) 521–547.

[28] H. Yoshioka, N. Yamada, N. Doke, cDNA cloning ofsesquiterpene cyclase and squalene synthase, and expres-sion of the genes in potato tuber infected with Phytoph-thora infestans, Plant Cell Physiol. 40 (9) (1999)993–998.

[29] J.D.C. Mau, C.A. West, Cloning of casbene synthasecDNA: evidence for conserved structural features amongterpenoid cyclases in plants, Plant Biol. 91 (1994) 8497–8501.

[30] Y.W. Nam, L. Tichit, M. Leperlier, B. Cuerq, I. Marty,J.M. Lelievre, Isolation and characterization of mRNAsdifferentialy expressed during ripening of wild straw-berry (Fragaria 6esca L.) fruits, Plant Mol. Biol. 39(1999) 629–636.

[31] J.F.S. Ferreira, J.E. Simon, J. Janick, Developmentalstudies of Artemisia annua : flowering and artemisininproduction under greenhouse and field conditions,Planta Med. 61 (2) (1995a) 167–170.

[32] J.F.S. Ferreira, J.E. Simon, J. Janick, Relationship ofartemisinin content of tissue cultured, greenhousegrown, and field grown plant of Artemisia annua, J.Planta Med. 61 (1995b) 351–355.

[33] H.J. Woerdenbag, N. Pras, R. Bos, J.F. Visser, H.Hendriks, T.h.M. Malingre, Analysis of artemisinin andrelated sesquiterpoids from Artemisia annua L. by com-bined gas chromatography/mass spectrometry, Phy-tochem. Anal. 2 (1991) 215–219.

.