A New Type of Asymmetrically Acting β-Carotene Ketolase Is Required for the Synthesis of Echinenone in the Cyanobacterium Synechocystis sp. PCC 6803

A New Type o Asymme ically Ac ing b -Ca o ene Ke olase Is Requi ed o he Syn hesis o Echinenone in he Cyanobac e ium Synechocys is sp. PCC 6803* (Recei ed o publica ion, No embe 27, 1996, Janua y 23, 1997) Blanca Fe na´ndez-Gonza´lez, Ge ha d Sandmann‡, and Agus ı´n Vioque§ F om he Ins i u o de Bioquı´mica Vege al y Fo osı´n esis, Uni e sidad de Se illa-CSIC, Ame ico Vespucio s/n, 41092 Se illa, Spain and he ‡Bo anisches Ins i u , Biosyn hesis G oup, Pos ach 111932, D-60054 F ank u , Ge many We ha e isola ed, based on he knowledge o he com- ple e genomic sequence o he cyanobac e ium Synecho- cys is sp. PCC 6803, an open eading ame (sl 0088) simila o known bac e ial ca o ene desa u ases and ha e analyzed he unc ion o he encoded p o ein. Su - p isingly, his p o ein has no de ec able desa u ase ac- i i y wi h phy oene, hyd oxyneu ospo ene, o z -ca o- ene as subs a es, bu is a he a b -ca o ene ke olase ha ac s asymme ically in oducing a ke o g oup on only one o he wo b -ionone ings o b -ca o ene o gen- e a e echinenone. This is in con as o he so a cha - ac e ized b -ca o ene ke olases ha ac symme ically, p oducing he di-ke o ca o enoid can haxan hin om b -ca o ene wi hou signi ican accumula ion o echi- nenone. We ha e designa ed his new gene c O. The unc ion o he c O gene p oduc has been demon- s a ed by 1) he biosyn hesis o echinenone when he c O gene is exp essed in an Esche ichia coli s ain able o accumula e b -ca o ene, 2) he in i o biosyn hesis o echinenone om b -ca o ene wi h cell ee ex ac s om E. coli cells ha exp ess he c O gene, and 3) he ab- sence o echinenone in a Synechocys is s ain in which he c O gene has been inse ionally inac i a ed. The p ima y s uc u e o he Synechocys is asymme ic ke- olase bea s no simila i y wi h he known b -ca o ene ke olases. c O is no equi ed o no mal g ow h unde s anda d o high ligh condi ions, nei he is he pho o- syn he ic ac i i y o he c O-de icien s ain a ec ed. Ca o enoids a e pigmen s ha a e syn hesized by all pho o- syn he ic o ganisms as essen ial componen s o he pho osyn- he ic appa a us (1). They a e also syn hesized in he e o ophic g owing bac e ia and ungi. They pa icipa e in ligh collec ion in pho osyn he ic o ganisms and play a p o ec i e ole agains oxida ion damage induced by s ong oxidan s p oduced in he pho osyn he ic memb anes upon illumina ion (2). In plan s and cyanobac e ia, ca o enoids a e syn hesized by a simila pa hway, om he C 40 p ecu so phy oene (Fig. 1). Phy oene is con e ed o z -ca o ene by wo sequen ial desa u- a ions ca alyzed by he enzyme phy oene desa u ase. z -Ca o- ene is u he oxidized by z -ca o ene desa u ase o gi e lyco- pene, which is con e ed o b -ca o ene by lycopene cyclase ( e iewed in Re s. 3–5). Phy oene desa u ase om cyanobac e ia and plan s, which in oduce only wo double bonds o p oduce z -ca o ene, ha e no homology wi h he bac e ial ype phy oene desa u ases, which in oduce h ee o ou double bonds, p oducing neu ospo ene o lycopene, espec i ely (6). The e o e a second enzyme ( z - ca o ene desa u ase) is equi ed in cyanobac e ia and plan s o he p oduc ion o lycopene om z -ca o ene. z -Ca o ene de- sa u ase om plan s is homologous o he plan and cyanobac- e ial ype phy oene desa u ases (7). Su p isingly, he only cyanobac e ial z -ca o ene desa u ase iden i ied, om Anabaena sp. PCC 7120, is homologous o he bac e ial ype phy oene desa u ases (8). I has been sugges ed ha plan and bac e ial desa u ases ha e independen e olu iona y o igins and ha in cyanobac e ia he bac e ial ype desa u ase has been es ic ed o he las wo dehyd ogena ion s eps, while he plan ype desa u ase ook o e he unc ion o he i s wo desa u a ion s eps o phy oene. In Synechocys is no z -ca o ene desa u ase has been iden i ied so a . The ca o enoid pa e n o cyanobac e ia is e y di e se (4), bu cyanobac e ia a e no able o syn hesize a -ca o ene and i s de i a i es as well as epoxy ca o enoids. In Synechocys is he majo ca o enoids accumula ed a e b -ca o ene, myxoxan ho- phyll, zeaxan hin, and echinenone (9, 10). The biosyn he ic pa hway o hose ca o enoids is ou lined in Fig. 1. None o he enzymes in ol ed in he biosyn hesis o myxoxan hophyll, ze- axan hin, and echinenone om b -ca o ene has been cha ac e - ized, no ha e hei genes been cloned in cyanobac e ia. The biosyn hesis o zeaxan hin equi es an hyd oxylase ha has been al eady cha ac e ized in some bac e ia (11, 12). The bio- syn hesis o myxoxan hophyll equi es se e al uniden i ied en- zymes, and some o hem migh be equi alen o he ones used in Rhodobac e o he syn hesis o sphe oidenone (13). Syn- echocys is is one o he e y ew species ha accumula es echi- nenone in subs an ial amoun s (9). The biosyn hesis o echi- nenone equi es a ke olase ha a acks only one o he b - ings o b -ca o ene. Genes coding o b -ca o ene ke olases we e cloned and sequenced om wo di e en bac e ia (14) and he g een alga Haema ococcus (15, 16). Howe e , all hese ela ed enzymes ca alyze he simul aneous in oduc ion o a ke o g oup in o posi ion 4 o e e y ionone ing a each end o he molecule, yielding he symme ically di-ke o ca o enoid can- haxan hin. Echinenone, wi h only one ke o g oup, is ound as a mino in e media e o he eac ion (17, 18). In he Synechocys is genomic sequence (19) an ORF 1 (sl 0088) has been ound wi h signi ican homology o bac e ial * This wo k was suppo ed in pa by Di eccio´n Gene al de In es i- gacio´n Cien ı´ ica y Te´cnica (G an PB94-1444), Jun a de Andalucı´a (no. CVI 0215), and by he Deu sche Fo schungsgemeinscha and he Bundesminis e ium u¨ Bildung, Wissenscha , Fo schung und Tech- nologie, BMBF. The cos s o publica ion o his a icle we e de ayed in pa by he paymen o page cha ges. This a icle mus he e o e be he eby ma ked “ad e isemen ” in acco dance wi h 18 U.S.C. Sec ion 1734 solely o indica e his ac . § To whom co espondence should be add essed: Ins i u o de Bio- quı´mica Vege al y Fo osı´n esis, Cen o de In es igaciones Isla de la Ca uja, Ame ico Vespucio s/n, 41092 Se illa, Spain. Tel.: 34-5- 4489519; Fax: 34-5-4460065; E-mail: [email p o ec ed]. 1 The abb e ia ions used a e: ORF, open eading ame; HPLC, high pe o mance liquid ch oma og aphy; m E, mic oeins ein. THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 272, No. 15, Issue o Ap il 11, pp. 9728–9733, 1997 © 1997 by The Ame ican Socie y o Biochemis y and Molecula Biology, Inc. P in ed in U.S.A. This pape is a ailable on line a h p://www-jbc.s an o d.edu/jbc/9728 This is an Open Access a icle unde he CC BY license. ype ca o ene desa u ases. Because i was possible ha his p o ein was he Synechocys is z -ca o ene desa u ase, we de- cided o unde ake he cloning o he co esponding gene and he cha ac e iza ion o i s unc ion. Su p isingly sl 0088 does no show any desa u ase ac i i y, bu is a he a b -ca o ene ke olase ha eac s asymme ically on i s subs a e. We ha e demons a ed i s unc ion by unc ional complemen a ion in Esche ichia coli,byi sin i o ac i i y, and by inac i a ion o he co esponding gene in Synechocys is. This new b -ca o ene ke olase om Synechocys is is un e- la ed in s uc u e o he symme ically ac ing ke olases cha - ac e ized in algae and bac e ia and ca alyzes a di e en eac- ion. The e o e we ha e named i s gene c O o di e en ia e i om he symme ically ac ing ke olase genes (c W). EXPERIMENTAL PROCEDURES Recombinan DNA Techniques—All manipula ions we e pe o med by s anda d me hods (20) o as ecommended by he manu ac u e s. Synechocys is cells we e g own in BG11 medium (21) and DNA ex- ac ed as desc ibed (22). Sou he n blo was pe o med by s anda d p ocedu es (20). Cloning o b -Ca o ene Ke olase om Synechocys is 6803—ORF sl 0088 was cloned by polyme ase chain eac ion based on he a ailable Synechocys is genomic sequence (19). The o wa d p ime (59-AACA- GAGAATTCATCACCACCGATGTTGTC-39) con ains an EcoRI si e (un- de lined) and o e laps he beginning o he coding sequence. The e- e se p ime (59-AACAGAGGATCCTTACCAAAAACGACGTTG-39) con ains a BamHI si e (unde lined) and o e laps he 39-end o he gene. A polyme ase chain eac ion p oduc o he expec ed size was pu i ied and ea ed wi h EcoRI and BamHI and cloned in he polylinke o pT c99A (Pha macia Bio ech Inc.) o gene a e plasmid pTRCRT-O. In his plasmid exp ession o he c O gene is con olled by he inducible pT c p omo e o he ec o . The amino end o he expec ed ecombinan p o ein has wo addi ional amino acids (MEFITTD. . . e sus MITTD. . .). Inac i a ion o c O—To inac i a e he c O gene, a 1.3-kilobase pai HincII agmen con aining a kanamycin esis ance gene (np ) om Tn5 (23) was used o eplace an in e nal ClaI agmen o he c O gene. A plasmid wi h he np gene inse ed in he opposi e o ien a ion o he c O ORF was used o ans o m Synechocys is sp. PCC 6803 wild ype s ain and cells we e pla ed on kanamycin-con aining pla es. T ans o - man s we e g own on he same medium o se e al seg ega ion ounds. Seg ega ion was checked by Sou he n blo using as a p obe an in e nal Bs XI agmen om he c O gene. Complemen a ion in E. coli—Plasmid pTRCRT-O was in oduced in E. coli JM101 con aining di e en plasmids depending on he unc ion o be es ed (see Table I). Plasmid pACCRT-EB (24) con ains he genes c E and c B om E winia. Cells ha ca y his plasmid accumula e phy oene. Plasmid pACCRT-EBP (25) con ains he genes c E and c B om E winia and he c P gene om Synechococcus, and in plasmid pACCRT-EBI Rc he c P gene is eplaced by c I om Rhodobac e capsula us (26). Cells ha ca y hese plasmids accumula e z -ca o ene and neu ospo ene, espec i ely. Plasmid pACCAR16Dc X (14) con ains he genes c E,c B,c I, and c Y om E winia. Cells ha ca y his plasmid accumula e b -ca o ene. These ou plasmids a e all de i a i es o ec o pACYC184 ca ying an o igin o eplica ion compa ible wi h he o igin o pTRCRT-O. Ano he ec o used o complemen a ion was pRKCRT-C ha is de i ed om pRK404 and ca ies he c C gene om R. capsula us coding o a neu ospo ene hyd a ase. The ans o man s we e cul i a ed a 28 °C in LB medium con aining ampicillin (100 m g/ml), chlo amphenicol (50 m g/ml), and/o e acycline (50 m g/ml) ac- co ding o he plasmids p esen . Cells we e g own o 48 h in he p esence o 2 mMisop opyl- b -D- hiogalac opy anoside be o e ha es ing o ca o enoid analysis. Ca o enoid Analysis—Ca o enoids we e ex ac ed om eeze-d ied E. coli o Synechocys is 6803 cells wi h me hanol con aining 6% KOH by hea ing o 15 min a 60 °C. A e pa i ioning in o 10% e he in pe ol, he uppe phase was collec ed, he sol en wad e apo a ed, and he FIG.1.Ca o enoid biosyn hesis pa hway. The main ca o enoids accumula ed in Synechocys is 6803 a e boxed. The pa hway om g -ca o ene o myxoxan hophyll equi es se e al uncha ac e ized s eps. Asymme ically Ac ing b -Ca o ene Ke olase in Synechocys is 9729 esidual ca o enoids we e esuspended in ace one. Sepa a ion and quan i ica ion o ca o enoids we e done by HPLC wi h a Nucleosil-C 18 3 m column and ace oni ile/me hanol/2-p opanol (85:10:5, / ) as eluen a a low a e o 1 ml/min. Spec a we e eco ded on-line a he elu ion peaks wi h a Kon on 440 pho odiode a ay de ec o . Fo iden i ica ion o ca o enoids, au hen ic s anda ds isola ed and iden i ied p e iously (9) we e used. In Vi o b -Ca o ene Ke olase Assay—F eshly ha es ed cells o JM101 ca ying pCAR16Dc X, which syn hesize b -ca o ene, and o JM101 ca ying pTRCRT-O, in which he ke olase gene is exp essed, we e esuspended in 50 mMT is-HCl, pH 7.5, and dis up ed in a F ench p essu e cell a 95 MPa. Bo h homogena es we e ea ed wi h DNase (5 m g/ml) o 15 min on ice. One se ed as a sou ce o subs a e and he o he as a sou ce o enzyme. The assay consis ed o 250 m l o each ex ac and 2 mMNADPH. Incuba ion was o 15 h a 30 °C. The equi alen o 600 ng o b -ca o ene was p o ided by he pCAR16Dc X ex ac . The eac ion was e mina ed by addi ion o 2.5 ml o me hanol. A e hea ing o 60 °C, he ca o enoids we e ex ac ed and sepa a ed by HPLC as desc ibed abo e. De e mina ion o Pho osyn he ic Ac i i y—Synechocys is 6803 wild ype and c O 2 s ains we e cul u ed a 30 °C in BG11 and bubbled wi h a con inuous s eam o 1.5% ( / ) CO 2 in ai unde cons an illumina ion (50 m Em 22 s 21 ). Fo he c O 2 s ain, he medium was supplemen ed wi h kanamycin o a inal concen a ion o 25 m g/ml. When he cul u es eached a densi y o 0.2 abso bance uni a 580 nm, hey we e di ided in wo hal es, one was main ained a 50 m Em 22 s 21 and he o he was illumina ed a 250 m Em 22 s 21 . Oxygen e olu ion was pe iodically de e mined a 30 °C upon i adia ion wi h sa u a ing whi e ligh by using a Cla k- ype oxygen elec ode. Chlo ophyll was de e mined as desc ibed p e iously (27). RESULTS Ou in e es in he biosyn he ic pa hway o ca o enoids in cyanobac e ia led us o iden i y he genes coding o he en- zymes phy oene desa u ase (c P) and phy oene syn hase om Synechocys is (28, 29) and z -ca o ene desa u ase (c Q) om Anabaena sp. PCC 7120 (25). In he Synechocys is genome da a base an ORF (sl 0088) wi h signi ican homology o bac e ial ype desa u ase (c I) genes has been iden i ied. As he Anabaena c Q gene is homologous o c I a he han o he plan o cyanobac e ial c P genes (8), we decided o explo e he possibili y ha sl 0088 was he Synechocys is z -ca o ene desa u ase. Cloning and Func ional Exp ession o c O—The open ead- ing ame iden i ied as sl 0088 in he Synechocys is genome was cloned by polyme ase chain eac ion and in oduced in an exp ession ec o o E. coli o gene a e plasmid pTRCRT-O. E. coli cells ca ying plasmid pTRCRT-O syn hesized a p o ein o he expec ed size (59 kDa) (no shown) upon induc ion wi h isop opyl- b -D- hiogalac opy anoside. To es he unc ion o he exp essed p o ein, plasmid pTRCRT-O was in oduced in E. coli in combina ion wi h se e al plasmids ha allow he accu- mula ion o di e en ca o enoids (Table I). In E. coli cells ha accumula e phy oene (con aining pACCRT-EB), 1-hyd oxyneu- ospo ene (con aining pACCRT-EBI Rc and pRKCRT-C), and z -ca o ene (con aining pACCRT-EBP), no desa u ase ac i i y could be de ec ed when he c O gene p oduc was exp essed. These esul s sugges ha he c O gene p oduc is no a ca o ene desa u ase, as a e he enzymes encoded by c I and c D, no is i ela ed o he z -ca o ene desa u ase gene om Anabaena, because no de ec able desa u a ion p oduc s om phy oene, z -ca o ene, o 1-hyd oxyneu ospo ene accumula ed in E. coli cells ha exp ess c O. Al e na i ely c O could be a ca o ene desa u ase in Synechocys is ha canno be unc ion- ally exp essed in E. coli, e en hough he p o ein is syn hesized o ha equi es addi ional co ac o s no p esen in E. coli. Howe e , when pTRCRT-O was in oduced in E. coli cells ha accumula e b -ca o ene due o he p esence o pACCRTD16c X, wo addi ional ca o enoids we e obse ed, echinenone and can- haxan hin in lowe amoun s (Fig. 2). This esul indica es ha c O is a b -ca o ene ke olase. P e iously iden i ied b -ca o ene ke olases p oduce he accumula ion o can haxan hin wi h li - le o none o echinenone. Howe e , c O p oduces he accumu- la ion o much highe amoun s o echinenone han i does o can haxan hin. This co ela es wi h he si ua ion in Synecho- cys is ha accumula es echinenone a he han can haxan hin. In Vi o Ke olase Ac i i y o he c O Gene P oduc —We ha e con i med he in i o complemen a ion esul s by showing ha he c O gene p oduc can syn hesize echinenone om b -ca o- ene in i o. Cell homogena es om E. coli cells ha exp ess c O can con e b -ca o ene o echinenone a a con e sion a e o 12% in he p esence o NADPH (Table II). Some o ma ion o he di-ke o de i a i e can haxan hin (2% con e sion) was also obse ed. This indica es ha he speci ici y o b -ca o ene o he c ude enzyme is no absolu e. The eac ion p oduc s echi- nenone and can haxan hin we e posi i ely iden i ied no only by co-ch oma og aphy wi h au hen ic s anda ds bu also by hei abso bance spec a (Fig. 3). NADPH was equi ed o ac i i y. Inac i a ion o he c O Gene—To es he unc ion o c O in Synechocys is, we ha e inac i a ed he c O gene by inse ion o a kanamycin esis ance case e (Fig. 4A). The inac i a ed copy o c O could be comple ely seg ega ed as es ed by Sou h- e n blo (Fig. 4B), and cells lacking a unc ional c O gene seem o be iable. The ca o enoid composi ion o he c O-de icien FIG.2. Echinenone biosyn hesis in E. coli by he c O gene p oduc . HPLC analysis o ca o enoid pigmen s ex ac ed om E. coli JM101 cells ca ying plasmid pACCAR16Dc X (A) and plasmids pACCAR16Dc X 1pTRCRT-O (B). The ca o enoids iden i ied we e can haxan hin (1), echinenone (2), and b -ca o ene (3). TABLE I Ca o enoids accumula ed in E. coli in he p esence o pTRCRT-O in combina ion wi h se e al plasmids ha con ain ca o enoid biosyn hesis genes Ca o enoids we e ex ac ed om E. coli cells ca ying he plasmids indica ed and analyzed as desc ibed unde “Expe imen al P ocedu es.” Plasmid(s) pTRCRT-O Ca o enoids accumula ed pACCRT-EB 2Phy oene pACCRT-EB 1Phy oene pACCRT-EBP 2 z -Ca o ene pACCRT-EBP 1 z -Ca o ene pACCRT-EBIRc 1pRKCRT-C 2Hyd oxyneu ospo ene pACCRT-EBIRc 1pRKCRT-C 1Hyd oxyneu ospo ene pACCAR16Dc X 2 b -Ca o ene pACCAR16Dc X 1Echinenone, can haxan hin, and b -ca o ene Asymme ically Ac ing b -Ca o ene Ke olase in Synechocys is9730 s ain (c O 2 ) was compa ed wi h ha o wild ype Synecho- cys is (Fig. 5). Myxoxan hophyll, zeaxan hin, echinenone, and b -ca o ene we e he main ca o enoids iden i ied in he wild ype s ain. In he mu an s ain echinenone was absen , while he e we e wild ype le els o he o he ca o enoids. No can- haxan hin was de ec ed ei he in he wild ype o in he c O 2 s ain, e en when a di e en HPLC sys em op imized o he sepa a ion o can haxan hin om zeaxan hin (18) was used. This esul indica es ha c O is equi ed o echinenone bio- syn hesis in Synechocys is and, oge he wi h he p e ious esul s, clea ly demons a es ha c O is a b -ca o ene ke olase ha ac s asymme ically o p oduce echinenone as he main p oduc a he ha o can haxan hin. Cha ac e iza ion o he c O 2 S ain—The p e iously de- sc ibed esul s show ha we ha e been able o comple ely seg ega e a s ain lacking c O, and ha his s ain is de icien in echinenone biosyn hesis. To s udy wha could be he unc- ional ole o echinenone, we ha e analyzed he beha io o he c O 2 s ain when incuba ed unde ei he low (50 m Em 22 s 21 ) o high (250 m Em 22 s 21 ) ligh in ensi y. Bo h wild ype and c O 2 s ains ha e simila g ow h a es a low and a high ligh in ensi y. The pho osyn he ic a e (as measu ed by O 2 e olu- ion) o he c O 2 s ain is also simila o ha o wild ype a ei he low o high ligh in ensi y. The e o e incuba ion a high ligh in ensi y does no e eal any signi ican di e ences be- ween bo h s ains ega ding he measu ed pa ame e s. DISCUSSION The cyanobac e ium Synechocys is sp. PCC 6803 is he o ga- nism o choice o gene ical s udies. I is na u ally ans o m- able, and genes can be easily inac i a ed by a ge ing and homologous ecombina ion. Fu he mo e, he comple e se- quence o he Synechocys is genome is a ailable (30). Se e al genes o he ca o enoid biosyn hesis pa hway ha e been iden- i ied in Synechocys is. The a ailabili y o he comple e genome sequence gi es us he oppo uni y o sea ch o he so a uniden i ied genes o his pa hway in a a ional way. Among he open eading ames p esen in he Synechocys is genome, he e is one wi h signi ican homology o bac e ial ype phy- oene desa u ases (sl 0088). The only cyanobac e ial z -ca o ene desa u ase cha ac e ized so a is homologous o bac e ial phy- FIG.3.Abso bance spec a o he eac ion p oduc s o med in i o by he c O gene p oduc om b -ca o ene. A, can haxan hin; B, echinenone. FIG.4.Dis up ion o he c O gene. A, s a egy o dis up ion o c O by inse ion o he np gene. B, Sou he n blo o Synechocys is genomic DNA om he wild ype s ain (lane 1)o hec O 2 s ain (lane 2) diges ed wi h Bs XI and p obed wi h he in e nal Bs XI agmen o c O. Size ma ke s (kilobase pai s) a e indica ed on he le . FIG.5.The c O gene is equi ed o echinenone biosyn hesis in Synechocys is.HPLC analysis o ca o enoids pigmen s ex ac ed om Synechocys is 6803 wild ype s ain (A)o c O 2 s ain (B). The ca o enoids iden i ied we e myxoxan hophyll (1), zeaxan hin (2), b -c yp oxan hin (3), echinenone (4), and b -ca o ene (5). TABLE II In i o con e sion o b -ca o ene by homogena es om E. coli cells exp essing c O In i o ke olase assays we e ca ied ou as desc ibed unde “Expe - imen al P ocedu es” wi h ex ac s om E. coli cells ca ying plasmid pACCAR16Dc X alone o wi h a mix u e o he abo e wi h an ex ac o E. coli cells ca ying plasmid pTRCRT-O. Ex ac s Ca o enoid composi ion a e eac ion b -Ca o ene Echinenone Can haxan hin ng pACCARD16c X 510 0 0 pACCARD16c X 1 pTRCRT-O 429 62 9 Asymme ically Ac ing b -Ca o ene Ke olase in Synechocys is 9731 oene desa u ases a he han o plan ype phy oene desa u- ases, he e o e i was easonable o expec sl 0088 o be he Synechocys is z -ca o ene desa u ase. The esul s p esen ed in his wo k clea ly indica e ha sl 0088 is a b -ca o ene ke olase. This ac illus a es he isk o assuming unc ions o un- known ORFs based solely on weak homologies and indica es ha unc ional s udies a e equi ed o con i m he ole o he expec ed p o ein p oduc . The Synechocys is mu an s ain lacking c O does no accu- mula e echinenone, bu has no mal le els o b -ca o ene. The e- o e c O is no equi ed o he desa u a ion s eps necessa y o p oduce b -ca o ene, excluding ha c O is he unc ional z -ca - o ene desa u ase in Synechocys is. I should be no ed ha c O is closely linked and in he same o ien a ion as se e al o he genes, possibly o ganized as an ope on. The e o e inac i a ion o c O could a ec he exp ession o hese o he genes ha could accoun o he pheno ype obse ed. Howe e , he unc- ion o hose o he genes is hypo he ically un ela ed o ca o e- noid biosyn hesis (as deduced om sequence homologies), and some o hem a e expec ed o be essen ial. In any case a pola e ec on genes placed downs eam o c O could no accoun o he in i o ac i i y o he c O gene p oduc and he comple- men a ion obse ed in E. coli. As he plo in Fig. 6 clea ly shows, c O is homologous o bac e ial ype phy oene desa u ases as exempli ied by he E - winia he bicola enzyme. The mechanism o ca o ene ke olases is no known in de ail, bu he symme ically ac ing enzyme is mos likely a dioxygenase. 2 Al hough he in i o o ma ion o echinenone by he c O-de i ed ke olase was ca ied ou in cell homogena es, a dependence o he eac ion on NADPH was obse ed. This may be an indica ion o a di e en eac ion mechanism esponsible o he in oduc ion o he ke o g oup, in ol ing a monooxygenase- ype o hyd oxyla ion ollowed by dehyd ogena ion. c O con ains a nucleo ide binding si e a he amino end simila o he one in phy oene desa u ases. This is mos likely he si e o he binding o he co ac o NADPH. The h ee symme ically ac ing b -ca o ene ke olases cha ac- e ized so a , wo om bac e ia and one om a g een algae, a e homologous among hemsel es. They can be aligned, and ou conse ed egions ha e been iden i ied (15). They do no con- ain a nucleo ide binding si e a hei amino end. We ha e ied o iden i y in c O egions o simila i y wi h any o he ou conse ed domains in ke olases bu he e a e no signi ican simila i ies (Fig. 6). The e o e c O is a comple ely di e en enzyme, phylogene ically ela ed o ca o ene desa u ases a he han o ke olases. In addi ion c O has also a di e en mechanism o subs a e ecogni ion, ac ing asymme ically on b -ca o ene o in oduce a ke o g oup on only one o he b -ionone ings. In o he wo ds, c O can use as subs a e mainly b -ca - o ene bu echinenone e y poo ly. The p e iously iden i ied b -ca o ene ke olases do no p oduce he accumula ion o signi - ican amoun s o echinenone. When c O is exp essed in E. coli, abou 10% o he ke oca o- enoid p oduced is can haxan hin. The e o e c O has no a s ic speci ici y o b -ca o ene bu can, a a low a e, in oduce a ke o g oup in he second ionone ing o echinenone o gene - a e can haxan hin. Howe e , his e ec is due o he high le el exp ession o he ke olase and is es ic ed o he he e ologous en i onmen in E. coli, because no signi ican can haxan hin accumula ion is seen in Synechocys is. The a ailabili y o a Synechocys is mu an s ain de icien in he biosyn hesis o echinenone will allow he s udy o he unc- ion o his ca o enoid. I has been sugges ed ha echinenone is loca ed close o he eac ion cen e s in he hylakoid memb ane in a ela ed cyanobac e ium (31). Howe e , ou esul s indica e ha he absence o echinenone has no e ec on he g ow h a e o pho osyn he ic oxygen e olu ion a ei he low o high ligh . The speci ic ole o echinenone in Synechocys is will equi e u he in es iga ion. Ca o ene ke olases and hei genes a e o signi ican bio ech- nological in e es o he p oduc ion o ca o enoids o comme - cial in e es (32). The new b -ca o ene ke olase desc ibed he e has a di e en subs a e speci ici y o he p e iously known ke olases and he e o e p o ides addi ional lexibili y in he design o bio echnological p ocedu es o he p oduc ion o ca o- enoids o po en ial applied in e es . The Synechocys is b -ca - o ene ke olase could be use ul o enginee ing o he biosyn he- sis o no only echinenone, bu also o he asymme ic ca o enoids such as 3-hyd oxyechinenone, 39-hyd oxyechi- nenone, and adonixan hin when exp essed in combina ion wi h a ca o enoid hyd oxylase gene (c Z). REFERENCES 1. B i on, G. (1988) in Plan Pigmen s (Goodwin, T. W., ed) pp 133–182, Academic P ess, London 2. Sie e mann-Ha ms, D. (1987) Physiol. Plan . 69, 561–568 3. A ms ong, G. A. (1994) J. Bac e iol. 176, 4795–4802 4. Hi schbe g, J., and Chamo i z, D. (1994) in The Molecula Biology o Cyanobac e ia (B yan , D. A., ed) pp. 559–579, Kluwe Academic Publishe s, Do d ech 5. Sandmann, G. (1994) Eu . J. Biochem. 223, 7–24 6. Sandmann, G. (1994) J. Plan Physiol. 143, 444–447 7. Alb ech , M., Klein, A., Hugueney, P., Sandmann, G., and Kun z, M. (1995) FEBS Le . 372, 199–202 8. Linden, H., Misawa, N., Sai o, T., and Sandmann, G. (1994) Plan Mol. Biol. 24, 369–379 9. B amley, P. M., and Sandmann, G. (1985) Phy ochemis y 24, 2919–2922 10. Ma ı´nez-Fe´ ez, I., Vioque, A., and Sandmann, G. (1994) Pes ic. Biochem. Physiol. 48, 185–190 11. Misawa, N., Nakagawa, M., Kobayashi, K., Yamano, S., Izawa, Y., Nakamu a, K., and Ha ashima, K. (1990) J. Bac e iol. 172, 6704–6712 12. Misawa, N., Sa omi, Y., Kondo, K., Yokoyama, A., Kajiwa a, S., Sai o, T., Oh ani, T., and Miki, W. (1995) J. Bac e iol. 177, 6575–6584 13. A ms ong, G. A., Albe i, M., Leach, F., and Hea s , J. E. (1989) Mol. Gen. Gene . 216, 254–268 14. Misawa, N., Kajiwa a, S., Kondo, K., Yokoyama, A., Sa omi, Y., Sai o, T., Miki, W., and Oh ani, T. (1995) Biochem. Biophys. Res. Commun. 209, 867–876 2 P. D. F ase , Uni e si y o London, pe sonal communica ion. FIG.6.Simila i y plo s o Synechocys is 6803 b -ca o ene ke o- lase (c O) wi h o he p o eins. The sequence o Synechocys is 6803 b -ca o ene ke olase was compa ed wi h he sequences o Haema ococ- cus plu ialis b -ca o ene ke olase (A), Ag obac e ium au an icum b -ca - o ene ke olase (B), and E winia he bicola phy oene desa u ase (C). Simila i y was calcula ed wi h a window o 21 amino acids and a s ingency o 12 using he Compa e p og am o he GCG package (33). Asymme ically Ac ing b -Ca o ene Ke olase in Synechocys is9732 15. Kajiwa a, S., Kakizono, T., Sa o, T., Kondo, K., Oh ani, T., Nishio, N., Nagai, S., and Misawa, N. (1995) Plan Mol. Biol. 29, 343–352 16. Lo an, T., and Hi schbe g, J. (1995) FEBS Le . 364, 125–128 17. Yokoyama, A., Izumida, H., and Miki, W. (1994) Biosci. Bio echnol. Biochem. 58, 1842–1844 18. B ei enbach, J., Misawa, N., Kajiwa a, S., and Sandmann, G. (1996) FEMS Mic obiol. Le . 140, 241–246 19. Ko ani, H., Tanaka, A., Kaneko, T., Sa o, S., Sugiu a, M., Taba a, S. (1995) DNA Res. 2, 133–142 20. Samb ook, J., F i sch, E. F., and Mania is, T. (1989) Molecula Cloning: A Labo a o y Manual, 2nd Ed., Cold Sp ing Ha bo Labo a o y, Cold Sp ing Ha bo , NY 21. Rippka, R., De uelles, J., Wa e bu y, J. B., He dman, M., and S anie , R. Y. (1979) J. Gen. Mic obiol. 111, 1–61 22. Cai, Y., and Wolk, C. P. (1990) J. Bac e iol. 172, 3138–3145 23. Elhai, J., and Wolk, C. P. (1988) Gene (Ams .)68, 119–138 24. Linden, H., Misawa, N., Chamo i z, D., Pecke , I., Hi schbe g, J., and Sandmann, G. (1991) Z. Na u o sch. 46c, 1045–1051 25. Linden, H., Vioque, A., and Sandmann, G. (1993) FEMS Mic obiol. Le . 106, 99–104 26. Raisig, A., Ba ley, G., Scolnik, P., and Sandmann, G. (1996) J. Biochem. (Tokyo)119, 559–564 27. McKinney, G. (1941) J. Biol. Chem. 140, 315–322 28. Ma ı´nez-Fe´ ez, I. M., and Vioque, A. (1992) Plan Mol. Biol. 18, 981–983 29. Ma ı´nez-Fe´ ez, I., Fe na´ndez-Gonza´lez, B., Sandmann, G., and Vioque, A. (1994) Biochim. Biophys. Ac a 1218, 145–152 30. Kaneko, T., Sa o, S., Ko ani, H., Tanaka, A., Asamizu, E., Nakamu a, Y., Miyajima, N., Hi osawa, M., Sugiu a, M., Sasamo o, S., Kimu a, T., Hosouchi, T., Ma suno, A., Mu aki, A., Nakazaki, N., Na uo, K., Okumu a, S., Shimpo, S., Takeuchi, C., Wada, T., Wa anabe, A., Yamada, M., Yasuda, M., and Taba a, S. (1996) DNA Res. 3, 109–136 31. Nonnengiesse , K., Schus e , A., and Koenig, F. (1996) Bo . Ac a 109, 105–124 32. Nelis, H. J., and De Leenhee , A. P. (1991) J. Appl. Bac e iol. 70, 181–191 33. De e eux, J., Haebe li, H., and Smi hies, O. (1984) Nucleic Acids Res. 12, 387–395 Asymme ically Ac ing b -Ca o ene Ke olase in Synechocys is 9733