A 4'-Phosphopantetheinyl Transferase Mediates Non-Ribosomal Peptide Synthetase Activation in Aspergillus fumigatus

A4’-Phosphopan e heinyl T ans e ase Media es Non-Ribosomal Pep ide Syn he ase Ac i a ion in Aspe gillus umiga us Clai e Ne ille, Alan Mu phy, Ke in Ka anagh, and Sean Doyle*[a] In oduc ion Aspe gillus umiga us is a signi ican human pa hogen, pa icu- la ly in immunocomp omised indi iduals.[1] Th ee o ms o aspe gillosis a e ecognised clinically: sap ophy ic, alle gic and in asi e. In asi e aspe gillosis (IA) is he mos se ious o m o disease as i in ol es he in asion o iable issue and may p o- duce a mo ali y a e o app oxima ely 90%.[2] IA has eme ged as an impo an disease in ecen decades due o he use o agg essi e immunosupp essi e he apy causing p olonged neu openia in he ea men o cance and leukaemia. Despi e agg essi e an i ungal chemo he apy, dea h due o IA usually esul s 7–14 days a e diagnosis.[3] Ampho e icin B, azoles, such as i aconazole, and an eme g- ing a ay o inhibi o s o cell-wall biosyn hesis play a key ole in comba ing ungal in ec ion. Howe e , such ea men is no always e ec i e, al hough expensi e, and he e is a clea e- qui emen o he iden i ica ion o new ungal a ge s and an i- ungal he apies.[4] The nea a ailabili y o he comple e A. u- miga us genome sequence[5] (h p://www. ig .o g) has bo h s imula ed signi ican in e es in, and acili a ed he sea ch o , such no el ungal a ge s. Non- ibosomal pep ide (NRP), polyke ide (PK), a y acid and lysine biosyn heses in many bac e ia and ungi equi e pos - ansla ional phosphopan e heinyla ion o key apo-enzymes o acili a e me aboli e p oduc ion. 4’-phosphopan e heinyl ans- e ases (4’-PPTase) ca alyse he ans e o 4’-phosphopan e- heine om coenzyme A (CoA) o a conse ed se ine esidue wi hin he cogna e apo-enzyme (ei he apo-NRP o PK syn he- ase o a-aminoadipa e educ ase (AAR)). These 4’-phospho- pan e heinyl p os he ic g oups ancho ac i a ed in e media es and acili a e hei mo emen be ween ac i e si es wi hin he mul ienzyme complex.[6] In lysine biosyn hesis, he 4’-PPTases se e o ac i a e AAR, which in u n can con e a-amino- adipa e o a-aminoadipic semialdehyde, which is an essen ial s ep in lysine biosyn hesis in S. ce e isiae.[7] S p- ype 4’-PPTases a e hose p ima ily esponsible o NRPS and PKS phosphopan e heinyla ion, in addi ion o playing a key ole in AAR ac i a ion. Recen wo k[8,9] sugges ed he exis- ence o an S p- ype 4’-PPTase gene in A. umiga us, which may encode a p o ein o 39.6 kDa (359 amino acids) and con ain iden ical conse ed sequence mo i s o he npgA gene in A. ni- dulans. Obe egge e al.[10] ha e subsequen ly shown ha npgA gene exp ession is essen ial o he p oduc ion o he side o- pho es e ic ocin and iace yl usa inine, and may also be e- qui ed o a numbe o o he biosyn he ic pa hways, including lysine and ungal-pigmen biosyn hesis. Al hough A. umiga us is known o p oduce a ange o low- molecula -weigh me aboli es (e.g., side opho es and oxins), many o which a e likely o be o med by ei he polyke ide o NRP syn hesis, li le e o has been di ec ed owa ds he in es- iga ion o NRP syn hesis. Haas[11] has indica ed he p esence o an NRPS gene in A. umiga us, bu e ealed no u he de ails ega ding gene sequence o o ganisa ion. I is well known om s udies in o he p oka yo ic and ungal species ha NRP syn hesis occu s in an ATP-dependen manne h ough he ac- [a] C. Ne ille, D . A. Mu phy, D . K. Ka anagh, D . S. Doyle Na ional Ins i u e o Cellula Bio echnology Depa men o Biology, Na ional Uni e si y o I eland Maynoo h, Co. Kilda e (I eland) Fax: (+353)1-708-3845 E-mail: [email protected] Aspe gillus umiga us is a signi ican human pa hogen. Non- i- bosomal pep ide (NRP) syn hesis is hough o be esponsible o a signi ican p opo ion o oxin and side opho e p oduc ion in he o ganism. Fu he mo e, i has been shown ha 4’-phospho- pan e heinyla ion is equi ed o he ac i a ion o key enzymes in ol ed in non- ibosomal pep ide syn hesis in o he species. He e we epo he cloning, ecombinan exp ession and unc- ional cha ac e isa ion o a 4’-phosphopan e heinyl ans e ase om A. umiga us and he iden i ica ion o an a ypical NRP syn- he ase (A pes1), spanning 14.3 kb. Phylogene ic analysis has shown ha he NRP syn he ase exhibi s g ea es iden i y o NRP syn he ases om Me a hizium anisolpiae (PesA) and Al e na ia b assicae (Ab ePsy1). No he n hyb idisa ion and RT-PCR analysis ha e con i med ha bo h genes a e exp essed in A. umiga us.A 120 kDa agmen o he A. umiga us NRP syn he ase, con ain- ing a pu a i e hiola ion domain, was cloned and exp essed in he baculo i us exp ession sys em. De ec ion o a 4’-phosphopan- e heinyla ed pep ide (SFSAMK) om his p o ein, by MALDI-TOF mass spec ome ic analysis a e coincuba ion o he 4’-phos- phopan e heinyl ans e ase wi h he ecombinan NRP syn he- ase agmen and ace yl CoA, con i ms ha i is compe en o play a ole in NRP syn he ase ac i a ion in A. umiga us. The 4’-phosphopan e heinyl ans e ase also ac i a es, by 4’-phos- phopan e heinyla ion, ecombinan a-aminoadipa e educ ase (Lys2p) om Candida albicans, a key enzyme in ol ed in lysine biosyn hesis. ChemBioChem 2005, 6, 679 – 685 DOI: 10.1002/cbic.200400147  2005 Wiley-VCH Ve lag GmbH & Co. KGaA, Weinheim 679 i i y o mul i unc ional enzymes known as non- ibosomal pep- ide syn he ases (NRPS) and ha he o ganisa ion o modules wi hin hese NRPS di ec ly in luences he o de o he amino acids in he esul an pep ide p oduc s.[12] Modules wi hin NRP syn he ases a e composed o domains including adenyla ion, hiola ion (pep idyl ca ie p o ein) and condensa ion do- mains.[12] (O he domain ypes can also be p esen and a e e- iewed elsewhe e.[13]) Each domain has a speci ic unc ion: he adenyla ion domain ecognises and adenyla es, by ATP hyd ol- ysis, a subs a e amino acid, which is in u n ans e ed o he 4’-phosphopan e heine p os he ic g oup o he hiola ion domain by co alen a achmen o an a ailable SH g oup on he 4’-phosphopan e heine moie y. Ul ima ely, he co alen a - achmen o he amino acid o a p oximal amino acid occu s h ough he condensa ion domain ollowed by ei he sequen- ial ans e wi hin he NRPS modula sys em o elease h ough hioes e ase ac i i y o yield unc ional, non- iboso- mally syn hesised pep ides. To da e, no de ailed analysis o ei he NRPS genes o gene exp ession om A. umiga us has been o hcoming, despi e he eme gence o he o ganism as a signi ican human pa hogen. In an e o o in es iga e he ole o he pu a i e 4’-PPTase gene in, and elucida e he non- ibosomal biosyn he - ic capaci y o , A. umiga us, we ha e unde aken he isola ion o a 4’-PPTase gene, ca ied ou e alua ion o 4’-PPTase gene exp ession and exp essed he 4’- PPTase gene in a euka yo ic exp ession sys em. Sec- ondly, we ha e iden i ied and e alua ed he exp es- sion o an NRPS open eading ame wi hin he A. u- miga us genome. Finally, we p esen da a ha sup- po he hypo hesis ha he A. umiga us 4’-PPTase is esponsible o he unc ional ac i a ion o an NRPS apo-enzyme h ough 4’-phosphopan e heinyla ion. Resul s Phylogene ic analysis o 4’-PPTase and NRP syn he ase genes Da a mining o he nea -comple e A. umiga us genome wi h known sequences enabled he iden i ica ion o wo open ead- ing ames (o ) po en ially in ol ed in non- ibosomal pep ide syn hesis. The en i e coding egion (1.12 kb) o a 4’-phosphopan e- heinyl ans e ase (4’-PPTase) was iden i ied, cloned and se- quenced (Genbank accession numbe AY607103). Phylogene ic analysis shows ha he A. umiga us 4’-PPTase is an o hologue o he A. nidulans npgA/c wA gene (Genbank accession numbe AF198117). The A. umiga us 4’-PPTase was also used o BLAST (basic local alignmen sea ch ool) he Na ional Cen e o Bio echnological In o ma ion (NCBI) da abase (h p:// www.ncbi.nlm.nih.go ) in o de o de e mine he simila i y and iden i y o his sequence o o he known 4’-PPTases. Highes simila i y le els a e seen o A. nidulans npgA gene which is known o be in ol ed in penicillin biosyn hesis (55% iden i y; 67% simila i y) while he 4’-PPTase displays lowe le els o iden i y (25–32%) and simila i y (37–49%) o o he known 4’- PPTases (da a no shown). A pa ial o co esponding o a 3.76 kb egion o a pu a i e A. umiga us NRP syn he ase ( e med A pes1TCA) was also iden i ied, sequenced o con i m iden i y (Genbank Acc No AY607101 and AY607102) and cloned. This 3.76 kb o ( e med A pes1TCA) was ound o encode a pu a i e hiola ion, AMP- binding and condensa ion domain, simila o hose o o he non- ibosomal pep ide syn he ases. Fu he in silico analysis o he nea -comple e A. umiga us genome sugges ed ha A pes1TCA was loca ed wi hin a la ge NRP syn he ase gene (14.3 kb; e med A pes1). A pes1 clus e s o a clade ha also con ains known NRP syn he ases possibly in ol ed in des uxin and side opho e o ma ion in o he ungi,[14,15] and he p o ein is shown o be homologous o he NRP syn he ase (Ab epsy1) o Al e na ia b assicae (58% simila i y and 38 % iden i y), which is p oposed o be in ol ed in he pa hogenici y o ha o gan- ism. Simila i y o A pes1 o o he known NRP syn he ases anges om 40–56 % and iden i y anges om 24–37 % (da a no shown). Conse ed domain analysis o A pes1 indica es he p esence o ou adenyla ion (A), h ee condensa ion (C) and wo hiola ion (T) domains wi hin he gene '(Figu e 1). Fu he - mo e, he domain a chi ec u e o A pes1 indica es ha i is a ype o nonlinea ( ype C) NRP syn he ase, as de ined by a leas one unusual a angemen o co e A, C o T domains.[13] The double condensa ion domains may allow he o ma ion o cyclic pep ides o , i p esen a he end o he sequence, may be in ol ed in chain e mina ion. The AMP-binding domains o NRP syn he ases a e highly conse ed and may he e o e ep esen he speci ici y code o he pa icula NRP syn he ase, as i is his domain ha ecog- nises and binds he subs a e amino acid.[16] In gene al, each adenyla ion domain con ains an eigh - esidue, nonlinea se- quence mo i ha is esponsible o binding he subs a e amino acid. The ou adenyla ion domains o A pes1 we e aligned wi h he phenylalanine-ac i a ing adenyla ion domain o G sA[17] o iden i y each eigh - esidue mo i . The eigh - esi- due mo i o each domain was hen used o in e oga e a da- abase o domains wi h assigned speci ici y, a ailable a h p:// aynam.chm.jhu.edu/ ~n ps/ by using BLAST analysis. F om his alignmen , he pu a i e amino acid subs a e o each adenyla- Figu e 1. Schema ic diag am showing he domain a chi ec u e o A. umiga us pes1 (A pes1), a 14312 bp pu a i e non- ibosomal pep ide syn he ase egion. A: adenyla ion do- mains, C: condensa ion domains and T: hiola ion domains. The adenyla ion domains occu be ween nucleo ides (amino acids (aa)) 1–912 (1–304), 4326–5505 (1442–1835), 7437–8664 (2479–2888) and 10710–11 919 (3570–3973). The condensa ion domains occu be ween nu- cleo ides (aa) 1485–2376 (495–792), 2919–3783 (973–1261) and 9336–10 161 (3112–3387). The hiola ion domains encompass nucleo ides (aa) 1251–1446 (417–482) and 9018–9213 (3006–3071). The 3760 bp egion ( e med A pes1TCA) was cloned, sequenced and exp essed in he baculo i us exp ession sys em. 680  2005 Wiley-VCH Ve lag GmbH & Co. KGaA, Weinheim www.chembiochem.o g ChemBioChem 2005, 6, 679 – 685 S. Doyle e al. ion domain was de e mined. This analysis indica es ha adenyla ion-domain mo i s 1 (DAMVAYLS), 2 (DVKSVGAV), 3 (DAMFAGGI) and 4 (DVKSVGAV) encode C, T, V/I and T/D, espec i ely. Exp ession analysis o 4’-PPTase and A pes1 4’-PPTase and A pes1 gene exp ession we e bo h assessed by No he n analysis. In addi ion, A pes1 exp ession was e alua ed by RT-PCR. 4’-PPTase exp ession is e iden a all ime poin s du ing A. umiga us g ow h om =24 o 96 h (Figu e 2). The obse ed band is app oxima ely 2 kb; his sugges s ha he 4’- PPTase ansc ip is signi ican ly longe han he gene i sel . A pes1 exp ession was only de ec able a 24, 48 and 72 h by No he n analysis (mRNA size: ca. 15 kb). A pes1 exp ession a 48 ( e y weak) and 72 h was e iden on RT-PCR analysis, and, despi e epea ed a emp s, no A pes1 exp ession could be de- ec ed a 96 h; his sugges s ha A pes1 gene exp ession is down egula ed unde he A. umiga us cul u e condi ions em- ployed o hese analyses. The lack o de ec ion o A pes1 ex- p ession by RT-PCR a 24 h is sugges i e o a g adual up- egu- la ion in A pes1 gene exp ession. Al e na i ely, his obse a ion could be due o he di e en ial sensi i i y o RT-PCR and No he n analysis o he de ec ion A pes1 exp ession a low le els o ansc ip . RNA loading was used as a con ol o he amoun s o RNA used o No he n analysis and cDNA syn he- sis. In addi ion, he p esence o genomic DNA was excluded by bo h DNAse ea men o isola ed RNA p io o RT-PCR and analysis o he size di e ence be ween he genomic and cDNA amplicon o a gs A gene[18] (Genbank accession numbe AF425746) om which in ons ha e been excised (da a no shown). Baculo i us exp ession o ecombinan 4’-PPTase and pa ial NRPS (A pes1TCA) Bo h 4’-PPTase- and A pes1TCA-encoding egions we e inde- penden ly ans e ed in o a baculo i us ans e ec o (pBlue- Bac4.5) in o de o gene a e ecombinan baculo i uses, each encoding 4’-PPTase o A pes1TCA, by co ans ec ion wi h linea - ised baculo i us. A. umiga us 4’-PPTase was pu i ied om e- combinan baculo i us-in ec ed insec cells by single-s ep Ni- NTA aga ose a ini y ch oma og aphy o yield a soluble p o ein o 42 kDa as judged by SDS-PAGE analysis (Figu e 3). The yield o eleased and soluble 4’-PPTase was calcula ed o be 0.4 mg pe 108cells, al hough i was clea om subsequen analysis ha signi ican enzyme emained wi hin in ec ed Spodop e a ugipe da 9 (S 9) insec cells (da a no shown). MALDI-TOF mass spec ome y was used o e i y he iden i y o he e- combinan 4’-PPTase (ac ual/ heo e ical pep ides de ec ed: 8:106 (7.5 %), which ep esen s 16% sequence co e age). A. u- miga us pes1TCA was hen pu i ied om ecombinan baculo i- us-in ec ed insec cells by di e en ial ex ac ion and ex ensi e washing o yield a p o ein o 120 kDa as judged by SDS-PAGE analysis (Figu e 3). Lowe M p o ein bands p esen in pu i ied A pes1TCA a app oxima ely 38 and 62 kDa, (Figu e 3, lane 3) we e eac i e wi h abbi an ise a p e iously aised agains an E. coli-exp essed condensa ion domain o A pes1TCA and ep e- sen p o eoly ic agmen s o he pa en al p o ein (da a no shown). The yield o ecombinan A pes1TCA was calcula ed o be 2–3 mg pe 108cells. Al hough a His6a ini y ag had been enginee ed in o he DNA con- s uc , i was no possible o de ec A pes1TCA on Wes e n blo s wi h mu ine monoclonal an i-His6an ibody (da a no shown). Consequen ly, exp es- sion o ecombinan A pes1TCA was con i med by bo h MALDI- TOF mass spec ome y (ac ual/ heo e ical pep ides de ec ed: 23:287 (8%) ep esen ing 25 % sequence co e age) and abbi an ise a p e iously aised agains an E. coli-exp essed condensa- ion domain also encoded by A - pes1TCA exp essed in insec cells (da a no shown). Func ional analysis o 4’-phos- phopan e heinyl ans e ase Func ional 4’-PPTase ac i i y was ini ially in es iga ed by using an a-AAR-ac i a ion assay p e ious- ly desc ibed o he assessmen o C. albicans Lys5p, which ex- Figu e 2. Gene-exp ession analysis o 4’-PPTase and A pes1 in A. umiga us. No he n hyb idisa ion (60 mg o al RNA pe well) was used o assess 4’-PPTase and A pes1 exp ession in ungal cul u es anging om 24–96 h pos -inocula- ion. In addi ion, A pes1 exp ession was e alua ed by RT-PCR (1 mg o al RNA pe cDNA syn hesis eac ion) o e an iden ical ime pe iod. Cul u e s a iona y phase was eached a 72 h. The 4’-PPTase is cons i u i ely exp essed a all ime poin s e alua ed. A Pes1 exp ession, as judged by No he n hyb idisa ion is obse ed a 24, 48 and 72 h only. RT-PCR analysis o A pes1 exp ession also de ec s ansc ip s a 48 and 72 h. RNA om each isola ion ime poin was applied equally o No he n hyb idisa ion by using ibosomal RNA as loading con ol. ChemBioChem 2005, 6, 679 – 685 www.chembiochem.o g  2005 Wiley-VCH Ve lag GmbH & Co. KGaA, Weinheim 681 Pep ide Syn he ase Ac i a ion in A. umiga us hibi s 4’-PPTase ac i i y.[7] Figu e 4 clea ly shows ha AAR ac i - i y is de ec able in he p esence o ecombinan Lys2p and C. albicans CLD2 (Lys2p mu an ) ex ac , which con ains a unc- ional 4’-PPTase. When AAR ac i i y is measu ed a e p io in- cuba ion o he ecombinan C. albicans Lys2p wi h he pu a- i e A. umiga us 4’-PPTase, a signi ican 4.5- old enhancemen o AAR ac i i y is e iden (21.3% o posi i e con ol CLD2 ac i - i y). This obse a ion con i ms ha he ecombinan 4’-PPTase can ac i a e C. albicans Lys2p in he p esence o CoA and sug- ges s a possible ole o 4’-PPTase in lysine biosyn hesis in A. umiga us. App op ia e compensa ion o backg ound ab- so bance (due o CLD2, 4’-PPTase and Lys2p) has been made in each case, as i was obse ed ha p o ein addi ion con ib- u ed o backg ound signals. NRPS 4’-phosphopan e heinyla ion by 4’-PPTase The capaci y o ecombinan 4’-PPTase o ac i a e non- iboso- mal pep ide syn he ases was in es iga ed by coincuba ion o 4’-PPTase, CoA and ecombinan A pes1TCA, which encodes a pu a i e hiola ion domain (A pes1 aa esidues: 3006–3071), ollowed by MALDI-TOF mass spec ome ic analysis o he e- ac ion mix u e o de ec e idence o 4’-phosphopan e heinyla- ion. Codiges ion o he eac ion mix u e wi h bo h ypsin and V8 p o ease was used as ypsin-only diges ion gene a ed pep- ides om he A pes1TCA hiola ion domain ha we e oo la ge o eliable analysis (da a no shown). In addi ion, solu ion- phase enzyma ic diges ion was employed o op imise pep ide eco e y p io o MALDI-TOF mass spec ome y. I can be seen om Figu e 5 ha a pep ide wi h a monoiso opic m/z alue o 1009.1, ep esen ing 4’-phosphopan e heinyla ed SFSAMK ( he- o e ical m/z=1009.3), is de ec ed by his analysis, he eby con- i ming ha ecombinan A. umiga us 4’-PPTase bo h ecognis- es and 4’-phosphopan e heinyla es he hiola ion domain p es- en in A pes1. In he absence o ei he CoA o ecombinan A pes1TCA, no modi ied pep ide is de ec ed a he p edic ed m/z a io. Discussion He e we epo he cloning and unc ional exp ession o he ull-leng h open eading ame encoding an Aspe gillus umiga- us 4’-phosphopan e heinyl ans e ase in he baculo i us-ex- p ession sys em. In addi ion, an NRP syn he ase in A. umiga us (A pes1), which exhibi s unusual domain a chi ec u e showing cha ac e is ics o a nonlinea NRP syn he ase, has been iden i- ied om A. umiga us con ig 4944. A egion o his NRP syn- he ase (3.7 kb; A pes1TCA) was exp essed and ound o unde - go phosphopan e heinyla ion o he holo-enzyme o m ollow- ing ac i a ion by 4’-PPTase in he p esence o CoA; his sup- po s he hypo hesis ha he 4’-PPTase unc ions o ac i a e an NRP syn he ase by 4’-phosphopan e heinyla ion. The A. umiga us 4’-PPTase ha has been in es iga ed is iden ical o ha p e iously no ed by Moo z e al. and Keszen- man-Pe ey a e al.[8,9] Al hough bo h p e ious epo s specula - ed on he unc ion o he 4’-PPTase gene, by compa ison o he A. nidulans o hologue (npgA), de ini e con i ma ion ha he A. umiga us 4’-PPTase was exp essed in i o and unc- ioned as hypo hesised, in i o, has no been a ailable un il now. F om phylogene ic analysis, he A. umiga us 4’-PPTase was con i med o be an o hologue o he A. nidulans 4’- Figu e 3. A) SDS-PAGE analysis o ecombinan 4’-PPTase and A pes1TCA om A. umiga us. Lane M: molecula -weigh ma ke s, lane 1: unin ec ed insec cell lysa e (2.5 105cell equi alen s), lane 2: A pes1TCA ecombinan baculo i us-in ec ed S 9cells a an MOI 15, lane 3: pu i ied A pes1TCA, lane 4: 4’-PPTase ecombinan bacu- lo i us-in ec ed S 9cells a an MOI 15 and lane 5: a ini y-pu i ied 4’-PPTase. In lane 3, he p o ein bands e iden a 38 and 62 kDa we e eac i e wi h an ise a aised agains he condensa ion domain in A pes1TCA (da a no shown). B) Wes e n blo analysis o ecombinan 4’-PPTase p obed wi h an i-(His)6monoclonal an i- body. Lane M: molecula -weigh ma ke s, lane 1: 4’-PPTase ecombinan baculo i us-in ec ed S 9cells a an MOI 15 and lane 2: a ini y-pu i ied 4’-PPTase. Figu e 4. a-aminoadipa e educ ase (AAR) ac i a ion by ecombinan 4’-PPTase om A. umiga us. Ac i a ion o C. albicans Lys2p (apo-AAR), by a con ol ex- ac (CLD2) con aining a C. albicans 4’-PPTase, esul ed in a 21- old inc ease in obse ed AAR ac i i y. Enhancemen o AAR ac i i y (4.5- old) is e iden a e p io incuba ion o he C. albicans Lys2p wi h he A. umiga us 4’-PPTase and ep esen s 21.3 % o posi i e con ol CLD2-induced ac i i y. All eac ions we e ca ied ou in duplica e and in he p esence o 200 mmace yl CoA. 682  2005 Wiley-VCH Ve lag GmbH & Co. KGaA, Weinheim www.chembiochem.o g ChemBioChem 2005, 6, 679 – 685 S. Doyle e al. PPTase; his sugges ed common ances y, and i also showed simila i y o a Bacillus sub ilis 4’-PPTase. Mo eo e , A. umiga us 4’-PPTase exhibi s 25% iden i y (37% simila i y) o he human 4’-PPTase (Genbank accession numbe BC015470) and sha es some signi ican egions o homology (WxLKExxxK) as p e i- ously no ed.[9] In mos o ganisms he e is an indi idual 4’-PPTase o each unc ion, howe e he possible p esence o only one 4’-PPTase in A. umiga us is no unique, since ecen wo k has iden i ied and cha ac e ised a human 4’-PPTase ha appea s o exhibi a b oad speci ici y o all 4’-phosphopan e heinyla ion eac ions including human (apo-ACP domain o cy osolic a y acid syn- he ase (FAS), mi ochond ial ACP, a-aminoadipa e semialde- hyde dehyd ogenase ac i a ion (lysine ca abolism)) and non- human (B. sub ilis ACP-A (in ol ed in a y acid syn hesis) and B. b e is y ocidine syn he ase) apo-enzymes[19,20] A new ype o 4’-PPTase was also iden i ied in Pseudomonas ae uginosa ha showed associa ion wi h a y acid syn hesis and side o- pho e me abolism.[19] No he n analysis o A. umiga us 4’-PPTase exp ession indi- ca es ha he gene is cons i u i ely exp essed; his s ongly sugges s ha p o ein 4’-phosphopan e heinyla ion is equi ed o bo h p ima y- and seconda y-me aboli e p oduc ion in A. umiga us. Indeed, he pu a i e ole o 4’-PPTase in lysine biosyn hesis (see below), and i s ole in NRP-syn he ase ac i a- ion is compa ible wi h his obse a ion. To da e, cha ac e isa ion o yeas and ungal 4’-PPTases has p ima ily ocussed on gene ic complemen a ion s udies in ei he S. ce e isiae[8] o A. nidulans[9] as opposed o di ec bio- chemical s udies. Guo and Bha acha jee ha e de eloped an AAR-ac i a ion assay o assess 4’-PPTase ac i i y in i o.[7] Thus, he po en ial ole o A. umiga us 4’-PPTase in lysine biosyn hesis was in es iga ed. The in i o ac i- a ion o C. albicans Lys2p (Apo-AAR) by A. umiga us 4’-PPTase sugges s ha he enzyme may play a ole in lysine biosyn hesis. Al hough equi alen AAR ac i - i y (4.5- old ac i a ion) in compa ison o he posi i e con ol (22- old ac i a ion) was no e iden , his could be accoun ed o by di e ences be ween co e 1, 2 and 3 conse ed mo i s[7] in C. albicans Lys5p (4’- PPTase) and hose p esen in he A. umiga us 4’- PPTase. Mo eo e , phylogene ic analysis con i ms ha he 4’-PPTase encoded by C. albicans clus e s o a sepa a e clade o ha o he A. umiga us 4’-PPTase. A numbe o g oups ha e epo ed he applica ion o p o ein mass spec ome y o he de e mina ion o apo-enzyme 4’-phosphopan e heinyla ion.[20,22] In hese epo s, he subs a e p o eins exhibi ed mo- lecula masses in he o de o 11–38 kDa; howe e , in he case o baculo i us-exp essed A pes1TCA (120 kDa), he e was conce n ha p o ein mass spec- ome y would no exhibi he equi ed esolu ion o con i m 4’-phosphopan e heinyla ion. Consequen ly, he pos -phosphopan e heinyla ion eac ion mix u e was enzyma ically diges ed (wi h ypsin and V8 p o- ease) p io o pep ide mass inge p in ing. As he ex- pec ed pep ide con aining he conse ed se ine (S) wi hin he hiola ion domain (SFSAMK) had m/z=670.3 Da, pos - ansla ional modi ica ion by he 4’-PPTase was expec ed o add an addi ional 339 Da and esul in a pep ide o m/z= 1009.3 Da. Ou da a con i m he p esence o he 1009.3 Da pep ide ollowing coincuba ion o CoA, 4’-PPTase and A pes1T- CA unde op imal condi ions. The con i ma ion ha 4’-PPTase can 4’-phosphopan e heinyla e a hiola ion domain o an NRP syn he ase is highly signi ican as i s ongly indica es ha a leas one NRP syn he ase encoded by he A. umiga us genome can be ac i a ed, and he eby pa icipa e in non- ibo- somal pep ide syn hesis in he o ganism. Un il now, no mecha- nis ic e idence o NRP syn hesis in A. umiga us has been o h- coming, despi e he exis ence o mul iple side opho es and o he low-molecula -weigh me aboli es. Al hough i is unlikely ha A pes1 encodes a side opho e (see below), we hypo he- sise ha 4’-PPTase exp ession and unc ionali y will also be es- sen ial o side opho e p oduc ion in A. umiga us and so may ep esen a key d ug a ge in con olling side opho e p oduc- ion wi h a iew o minimising o ganism i ulence.[10, 11] Mo e- o e , he iden i ica ion o only one 4’-PPTase in A. umiga us, o da e, ha may be in ol ed in mul iple 4’-phosphopan e heiny- la ion eac ions, makes selec i e inhibi ion in he absence o concomi an inhibi ion o human 4’-PPTase e en mo e a ac- i e wi h a iew o an i-A. umiga us d ug he apy in humans. Howe e , he iden i ica ion o simila enzymes in A. umiga us canno be comple ely uled ou a p esen . The A pes1 open eading ame, which ex ends o e 14300 bp, encodes he i s NRP syn he ase (p edic ed p o ein ela i e molecula mass: 460 kDa) o be de ini i ely iden i ied in A. umiga us. Fu he mo e, No he n analysis con i ms ha he A pes1 gene is di e en ially exp essed ( ansc ip size: ca. Figu e 5. MALDI TOF MS analysis o in i o phosphopan e heinyla ion by A. umiga us 4’- PPTase. Mass spec um o A pes1TCA-de i ed pep ide mix u e ( ollowing codiges ion wi h yp- sin and V8 p o ease). A speci ic pep ide wi h a monoiso opic m/z alue o 1009.1 is e iden and ep esen s phosphopan e heinyla ed-SFSAMK. In he absence o ei he 4’-PPTase o CoA , du ing he enzyma ic eac ion, no modi ied pep ide is subsequen ly de ec ed a he expec - ed m/z a io. ChemBioChem 2005, 6, 679 – 685 www.chembiochem.o g  2005 Wiley-VCH Ve lag GmbH & Co. KGaA, Weinheim 683 Pep ide Syn he ase Ac i a ion in A. umiga us 15 kb) a 24, 48 and 72 h pos -inocula ion when A. umiga us is g own in liquid cul u e (5%, / FCS in minimal essen ial medium (MEM)). Upon phylogene ic analysis, A pes1 clus e s wi h he NRP syn he ase (Ab ePsy1) om Al e na ia b assi- cae,[15] hus sugges ing o hology. Also g ouped in he same clade is he pesA gene o Me a hizium anisopliae.[14] Un o u- na ely he unc ions o he NRP syn he ase genes o bo h A. b assicae and M. anisopliae ha e ye o be iden i ied, bu a e p oposed o be in ol ed in des uxin biosyn hesis, which is hough o play a ole in he p og ession o plan disease media ed by bo h species.[14, 15] Al hough i has been p oposed ha he NRP syn he ase gene o A. b assicae may play a ole in side opho e biosyn hesis,[15] pa icula ly since he i s adenyla- ion domain o Ab ePsy1 is p edic ed o encode 5-hyd oxyo ni- hine, he clus e ing o A pes1 away om a pu a i e Aspe gillus NRP syn he ase (A. o yzae sid2; Genbank accession numbe AB087617) sugges s ha his may no , in ac , be he case. A u he di e ence be ween bo h A. umiga us and Al e na ia genes is ha while he A pes1 gene ep esen s a comple e exonic egion, he Ab ePsy1 gene con ains se en in ons. Thus, while i is possible ha A pes1 may ha e a simila unc ion o ha o Ab ePsy1, o hology does no gua an ee simila i y o unc ion, which can change o e ime.[23] One s iking obse a ion wi h espec o he s uc u e o A pes1 is he a angemen o he indi idual domains wi hin he p o ein: ATCCAATCA. Such NRPSs a e e e ed o as non- linea ( ype C), as ecen ly desc ibed.[13] The p esence o an ad- enyla ion domain a he C e minus o an NRP syn he ase is somewha unusual and sugges s ha A pes1 may be in ol ed in a mul isubuni NRPS complex wi h ano he syn he ase. In- e es ingly, he p oduc ion o e go alkaloids con aining d-ly- se gic acid in he ungus Cla iceps pu pu ea occu s h ough a clus e o syn he ases, namely LPS1 (370 kDa) and LPS2 (140 kDa). These syn he ases a e encoded by wo genes, cpps2 and cpps1, p esen in a gene clus e . Mechanis ically, i has been shown ha d-lyse gic acid is ini ially bound o LPS2 h ough a hioes e bond and ha his hioes e -bound d-ly- se gic acid is hen ans e ed o LPS1. Following his, h ee condensa ion eac ions in o he d-lyse gyl mono-, di- and, i- nally, ipep ide hioes e occu o esul in he o ma ion o d- lyse gyl ipep ide lac am. This sys em o e go -pep ide o ma- ion was he i s iden i ied ungal-NRP-syn hesis sys em con- sis ing o di e en NRP syn hesis subuni s.[24] O he examples o ype C NRP syn he ases include PesA o Me a hizium anisol- piae and Ab ePsy1 o Al e na ia b assicae.[14, 15] The unc ional assessmen o 4’-PPTase ac i i y and he iden- i ica ion o a no el NRP syn he ase p o ide new insigh s in o non- ibosomal pep ide syn hesis in A. umiga us. In addi ion, we pos ula e ha he 4’-PPTase may also play a ole in lysine p oduc ion wi hin A. umiga us by ac i a ion o key biosyn he - ic enzymes. Al hough no de ini i e ole o he NRP syn he ase, A pes1, has been iden i ied, i may be esponsible o oxin biosyn hesis, and wo k is ongoing in ou labo a o y o iden i y gene unc ion wi hin he ascomyce e. Modula ion o 4’-PPTase exp ession o unc ionali y in A. umiga us may ep esen a no el an i ungal a ge . Expe imen al Sec ion Mic oo ganisms and cul u e media:Aspe gillus umiga us s ain ATCC 26933 (ob ained om he Ame ican Type Cul u e Collec ion, Manasas, VA (USA)) was used o his wo k and was g own a 378C in MEM supplemen ed wi h oe al cal se um (FCS; 5%, / ) o 48 h in o de o acili a e DNA isola ion. Fo gene-exp ession s ud- ies, in which cul u e was ex ended o 96 h, i was obse ed ha s a iona y phase was eached a e 72 h pos -inocula ion. Esche i- chia coli s ains Top10’, XL1Blue (In i ogen, Dublin (I eland)) and BL21 (No agen, No ingham (UK)) we e g own on LB aga (Sigma– Ald ich, Do se (UK)) and LB aga con aining ampicillin (100 mgmL1—chlo amphenicol (34 mgmL1) was also included o BL21 cul u e), whe e app op ia e, a 37 8C o e nigh . Bioin o ma ic analyses, genomic DNA isola ion and PCR cloning: P elimina y A. umiga us sequence da a we e ob ained om The Ins i u e o Genomic Resea ch websi e a h p://www. ig .o g. By using he sequence sea ch acili y a h p://www. ig .o g, he nea - comple e A. umiga us genome was in e oga ed by using ennia in syn he ase (Genbank accession numbe Z18755), wi h a BLAST p o- g am. Genomic DNA was isola ed by c ushing A. umiga us in liquid N2 ollowed by phenol/chlo o o m ex ac ion, washing wi h e hanol (70%, / ) and inal DNA esuspension in T is-HCl (10 mm), EDTA (1 mm; pH 7.4, 300 mL). Genomic DNA (5 mL) was used in all PCR eac ions con aining T is-HCl (10 mm; pH 9.0), KCl (50 mm), T i on X-100 (0.1%, / ), MgCl2(1.5 mm), dNTP (200 mmo each), o wa d and e e se p ime (1.0 mmo each; A pes1-F: 5’-GAGAGG- TACCATGTCGGAAGCAACACG-3’and A pes1-R: 5’-GAGAGGTACCTT- ACCAGTCAGCCTC-3’;o 4’-PPTase-F: 5’-GAGAGGATCCATGGGCTC- TGCACAAAACG-3’and 4’-PPTase-R: 5’-GAGAAAGCTTGGGCTGTTTT- TTTATACAC-3’) and Taq polyme ase (1 uni ) (P omega, Sou hamp- on (UK)) in a o al olume o 49.5 mL. AccuTaq LA DNA polyme ase (Sigma–Ald ich) was used o ampli y a pa ial egion om A pes1 only. A e incuba ion a 958C o 5 min, all 4’-PPTase and A pes1 PCR eac ions we e con inued wi h 35 cycles consis ing o 60 s de- na u a ion (90 s o A pes1) a 958C, p ime annealing a 55 8C o 60 s, ex ension a 728C o 240 s and inally ex ension o 360 s a 728C. PCR-ampli ied DNA was analysed by aga ose gel elec opho- esis whe eby p oduc (10 mL) was elec opho esed on aga ose (1%, w/ ) con aining e hidium b omide (0.5 mgmL1) o 30 min a 100 V. Visualisa ion o he 4’-PPTase- and A pes1-de i ed amplicons was pe o med by using an “Eagle-Eye II” digi al s ill ideo sys em. Once PCR had been success ully ca ied ou , bo h amplicons we e indi idually cloned in o he TOPO cloning ec o and subsequen ly ans e ed in o sepa a e pBlueBac4.5 ec o s (Clon ech, Palo Al o, CA) by using s anda d molecula biology echniques, o exp es- sion in S 9cells. No he n hyb idisa ion and RT-PCR analysis o 4’-PPTase and A pes1 gene exp ession A. umiga us cul u es we e ha es ed a de ined ime poin s (24, 48, 72 and 96 h), mycelia collec ed, washed wi h ice-cold phospha e-bu e ed saline and o al-RNA ex- ac ed by using TRI eagen (Sigma–Ald ich). No he n analyses (60 mg o al RNA pe well) we e ca ied ou by using 32P-labelled egions o A pes1 (NT: 9716–10535) and 4’-PPTase (NT: 1–1123) o de ec ele an ansc ip s acco ding o s anda d p o ocols.[25] RT- PCR was pe o med by ollowing cDNA syn hesis wi h Supe sc ip (P omega; 1 mg RNA pe eac ion) and ele an PCR p ime s (see abo e). Cons uc ion o ecombinan baculo i us encoding 4’-PPTase and A pes1TCA:S 9insec cells we e co- ans ec ed wi h ecombi- nan ans e ec o s and Bsu36 I-linea ised BacPAK-6 baculo i us by using Bac ec in liposomal p epa a ion (Clon ech, Palo Al o, 684  2005 Wiley-VCH Ve lag GmbH & Co. KGaA, Weinheim www.chembiochem.o g ChemBioChem 2005, 6, 679 – 685 S. Doyle e al. CA).[24] Well pla es con aining S 9cells we e in ec ed wi h he co- ans ec ion mix u es, and he TC-100 supe na an was ha es ed a e 5 days’ incuba ion a 278C. Pu e clones o ei he 4’-PPTase- o A pes1TCA-encoding ecombinan baculo i uses we e iden i ied by plaque assay.[26] An igen exp ession and pu i ica ion: Recombinan baculo i uses encoding he 4’-PPTase and A pes1TCA genes we e indi idually used o in ec S 9cells in monolaye cul u e a a mul iplici y o in ec ion o 15. In ec ed cells (5108) we e ha es ed 4 days pos -in ec ion. In each case, SDS-PAGE, Wes e n Blo ing and MALDI-TOF mass spec ome y we e used o con i m he p esence o 4’-PPTase and A pes1TCA. I was obse ed ha while he 4’-PPTase was soluble and could be isola ed om cell pelle s by washing wi h PBS, ol- lowed by Ni chela e a ini y ch oma og aphy by elu ion wi h imida- zole (200 mm), a mo e ex ensi e ex ac ion p ocedu e was equi ed o A pes1TCA pu i ica ion. Fo A pes1TCA isola ion, cells we e lysed in he p esence o p o ease inhibi o s (phenylme hylsul onyl luo- ide (0.1 mm), peps a in (2 mgmL1) and leupep in (2 mgmL1)) by he addi ion o phospha e-bu e ed saline/sodium deoxychola e (0.5%, w/ ) and subjec ed o DNAse (Sigma, Poole (UK)) ea men ( inal concen a ion: 10 mgmL1), and he insoluble pelle was washed ex ensi ely o emo e con amina ing p o eins.[26] Insoluble A pes1TCA was esuspended in guanidinium hiocyana e (0.5– 1.0 mL, 6m), con aining di hio h ei ol (5 mm), a a concen a ion o 3mgmL 1and sequen ially dialysed agains sodium ca bona e (50 mm; pH 9.4) con aining U ea (3m) and sodium ca bona e (50 mm; pH 9.4) o p oduce soluble p o ein. Enzyme assays: AAR ac i a ion by ecombinan 4’-PPTase was de- e mined as desc ibed by Guo e al. and Guo and Bha acha - jee.[5,25] B ie ly, es o a ion o AAR ac i i y by 4’-phosphopan e hei- nyla ion in apo-Lys2p om C. albicans is indica i e o 4’-PPTase ac i i y. The e o e, ecombinan A. umiga us 4’-PPTase ac i i y was assessed by incuba ing he enzyme (50 mg pe eac ion) wi h e- combinan Lys2p (100 mg; inal olume 500 mL) ollowed by de ec- ion o AAR ac i i y by de e mina ion o DA460nm due o he con- e sion o dl-a-aminoadipa e o a-aminoadipa e-d-semialdehyde. C. albicans CLD2 lysa e (a Lys2p mu an exp essing C. albicans 4’- PPTase) con aining 1 mg o al p o ein was used as a posi i e con- ol o he eac ion. NRPS 4’-phosphopan e heinyla ion was de e - mined by incuba ing A pes1TCA (100 mg) and 4’-PPTase (10 mg) in he p esence o CoA (100 mm; eac ion olume 100 mL)[26,27] ol- lowed by u ea dena u a ion (5m inal) and in-solu ion double di- ges ion wi h ypsin and S. au eus V8 p o ease p io o MALDI-TOF pep ide mass inge p in ing o de ec phosphopan e heinyla ed pep ide species. MALDI-TOF mass spec ome y: Mass spec ome y was ca ied ou by using an E an MALDI-TOF mass spec ome e (Ame sham Biosciences (Eu ope), F eibu g (Ge many)). P o ein samples o pep- ide mass de e mina ion we e ei he i) sepa a ed by SDS-PAGE, di- ges ed wi h ypsin o ii) ob ained ollowing in-solu ion enzyma ic diges ion and deposi ed (1 mL) wi h a-cyano-4-hyd oxycinnaminic acid (1 mL; ace oni ile in aqueous i luo oace ic acid 5 mg/200 mL 50%, / ) on o mass spec ome y slides and allowed o d y p io o delayed ex ac ion and e lec on TOF analysis a 20 kV. Acknowledgemen s This wo k was unded unde he I ish Highe Educa ion Au ho i- y–P og amme o Resea ch in Thi d Le el Ins i u ions (HEA- PRTLI) Cycle 3. C.N. was a ecipien o a Daniel O’Connell Fellow- ship om NUI Maynoo h. P elimina y sequence da a we e ob- ained om The Ins i u e o Genomic Resea ch websi e a h p:// www. ig .o g. Sequencing o Aspe gillus umiga us was unded by he Na ional Ins i u e o Alle gy and In ec ious Disease U01AI48830 o Da id Denning and William Nie man, he Well- come T us , and Fondo de In es icagiones Sani a ias. Finally we a e g a e ul o P o . J. K. Bha acha jee, Miami Uni e si y, Ohio (USA) o AAR eagen s. Keywo ds: d ug a ge s ·myco oxin ·non- ibosomal pep ide syn he ases ·polyke ides ·side opho es [1] J. L. B ookman, D. W. Denning, Cu . Opin. Mic obiol. 2000,3(5), 468– 474. [2] P. Daly, K. Ka anagh, B . J. Biomed. Sci. 2001,58, 197– 205. [3] D. W. Denning, Clin. In ec . Dis. 1996,23, 608– 615. [4] D. P. Kon oyiannis, G. P. Bodey, Eu . J. Clin. Mic obiol. In ec . Dis. 2002,21- (3), 161– 172. [5] J. E. Mabey, M. J. Ande son, P. F. Giles, C. J. Mille , T. K. A wood, N. W. Pa on, E. Bo nbe g-Baue , G. D. Robson, S. G. Oli e , D. W. Denning, Nucleic Acids Res. 2004,32, 401– 405. [6] R. H. Lambalo , A. M. Geh ing, R. S. Flugel, P. Zube , M. LaCelle, M. A. Ma ahiel, R. Reid, C. Khosla, C. T. Walsh, Chem. Biol. 1996,3, 923–936. [7] S. Guo, J. K. Bha acha jee, FEMS Mic obiol. Le . 2003,224, 261– 267. [8] H. D. Moo z, K. Schç gendo e , M. A. Ma ahiel, FEMS Mic obiol. Le . 2002, 213, 51–57. [9] D. Keszenman-Pe ey a, S. Law ence, M. E. Tw ieg, J. P ice, G. Tu ne , Cu . Gene . 2003,43, 186 –190. [10] H. Obe egge , M. Eisendle, M. Sch e l, S. G aessle, H. Haas, Cu . Gene . 2003,44, 211 –215. [11] H. Haas, Appl. Mic obiol. Bio echnol. 2003,62, 316 –330. [12] H. Kleinkau , H. on Doh en, Eu . J. Biochem. 1996,236, 335–351. [13] H. D. Moo z, D. Schwa ze , M. A. Ma ahiel, ChemBioChem 2002,3, 490– 504. [14] A. M. Bailey, M. J. Ke shaw, B. A. Hun , I. C. Pa e son, A. K. Cha nley, S. E. Reynolds, J. M. Cla kson, Gene 1996,173, 195–197. [15] T. Guilleme e, A. Sellam, P. Simoneau, Cu . Gene . 2004,45, 214–224. [16] H. on Dçh en, U. Kelle , J. Va e , R. Zoche , Chem. Re . 1997,97(7), 2675–2706. [17] T. S achelhaus, H. D. Moo z, M. A. Ma ahiel, Chem. Biol. 1999,6, 493– 505. [18] J. A. F ase , M. A. Da is, M. J. Hynes, Appl. En i on. Mic obiol. 2002,68, 2802–2808. [19] V. P aphanphoj, K. A. Sacks ede , S. J. Gould, G. H. Thomas, M. T. Ge agh- y, Mol. Gene . Me ab. 2001,72, 336 –342. [20] A. K. Joshi, L. Zhang, V. S. Rangan, S. Smi h, J. Biol. Chem. 2003,278, 33142–33 149. [21] R. Finking, J. Solsbache , D. Konz, M. Schobe , A. Scha e , D. Jahn, M. A. Ma ahiel, J. Biol. Chem. 2002,277, 50293–50 302. [22] K. J. Weissman, H. Hong, M. Oliynyk, A. P. Siskos, P. F. Leadlay, ChemBio- Chem 2004,5, 116–125. [23] D. B. Sea ls, Na u e Re . 2003,2, 613– 623. [24] T. Co eia, N. G ammel, I. O el, U. Kelle , P. Tudzynski, Chem. Biol. 2003, 10, 1281– 1292. [25] J. Samb ook, E. F. F i sch, T. Mania is, Molecula Cloning—A Labo a o y Manual, 2nd ed., Cold Sp ing Ha bo Labo a o y P ess, New Yo k, 1989. [26] O. Ennis, A. Co co an, K. Ka anagh, B. P. Mahon, S. Doyle, J. Clin. Vi ol. 2001,22, 55–60. [27] S. Guo, S. A. E ans, M. B. Wilkes, J. K. Bha acha jee, J. Bac e iol. 2001, 183, 7120– 7125. [28] L. Du, B. Shen, Chem. Biol. 1999,6, 507–517. [29] C. Sanchez, L. Du, D. J. Edwa ds, M. D. Toney, B. Shen, Chem. Biol. 2001, 8, 725–738. Recei ed: May 10, 2004 Re ised: Decembe 2, 2004 Published online on Feb ua y 18, 2005 ChemBioChem 2005, 6, 679 – 685 www.chembiochem.o g  2005 Wiley-VCH Ve lag GmbH & Co. KGaA, Weinheim 685 Pep ide Syn he ase Ac i a ion in A. umiga us