Eur. J. Biochem. 247, 526-534 (1997) 0 FEBS 1997
The pipecolate-incorporating enzyme for the biosynthesis of the immunosuppressant rapamycin Nucleotide sequence analysis, disruption and heterologous expression of rapP from Streptomyces hygroscopicus Ariane KONIG', Torsten SCHWECKE', Istvin MOLNAR', Giinter A. BOHM', Philip A. S. LOWDEN2,James STAUNTON* and Peter F. LEADLAY' ' Department of Biochemistry and Cambridge Centre for Molecular Recognition, University of Cambridge, Cambridge, United Kingdom University Chemical Laboratory and Cambridge Centre for Molecular Recognition, University of Cambridge, Cambridge, United Kingdom (Received 4 March 1997) - EJB 97 032312
An open reading frame (rapP) encoding the putative pipecolate-incorporating enzyme (PIE) has been identified in the gene cluster for the biosynthesis of rapamycin i n Streptomyces hygroscopicus. Conserved amino acid sequence motifs for ATP binding, ATP hydrolysis, adenylate formation, and 4'-phosphopantetheine attachment were identified by sequence comparison with authentic peptide synthetases. Disruption of rupP by phage insertion abolished rapamycin production in S. hygroscopicus, and the production of the antibiotic was specifically restored upon loss of the inserted phage by a second recombination event. rupP was expressed in both Escherichia coli and Streptomyces coelicolor, and recombinant PIE was purified to homogeneity from both hosts. Although low-level incorporation of [ I4C]B-alanine into recombinant PIE isolated from E. coli was detected, formation of the covalent acylenzyme intermediate could only be shown with the PIE from S. coelicolor, suggesting that while the recombinant PIE from S. coelicolor was phosphopantetheinylated,only a minor proportion of the recombinant enzyme from E. coli was post-translationally modified. Keywords: rapamycin ; pipecolate-incorporating enzyme; Streptomyces hygroscopicus ; non-ribosomal peptide synthetase; polyketide biosynthesis.
Rapamycin is a polyketide macrolide produced by Streptonzyces hygroscopicus, which first attracted attention because of its anti-fungal activity [I]. Recently it was discovered to be a powerful immunosuppressant, like the structurally related FKS06 and immunomycin produced by Streptomyces tsukuhaensis and S. hygroscopicus var. ascomyceticus, respectively (Fig. 1). All three molecules inhibit T-cell activation at concentrations 10- 100 times lower than that required by cyclosporin A, the agent currently favoured for the prevention of graft rejection following transplantation 121. The polyketide backbone of rapamycin is synthesised by head-to-tail condensation of a total of seven propionate and seven acetate units to a shikimate-derived cyclohexane carboxylic starter unit. The 14 chain-extension cycles are carried out by type-I polyketide synthase (PKS) multienzymes, in a process clearly analogous to erythromycin biosynthesis [3, 41. The imino acid pipecolate is condensed in an amide linkage onto the last acetate of the polyketide backbone. Cheng et al. reported on a 150 % increase in the production of rapamycin by S. hygroscopicus i n a chemically defined medium upon the addition of Llysine [51. Incorporation studies using ~ - [ ' ~ C ] l y s i nand e L['Hlpipecolate support a pathway in which L-lysine is converted to pipecolate by a deaminative cyclisation, and free pipecolate Corrr..sp,ondenreto I? F. Leadlay, Department of Biochemistry, Unio f Cambridge, CB2 lQW, UK F m : +44 1223 333 345. E-mail:
[email protected] Abbreviations, PIE, pipecolate-incorporating enzyme : ACP, acyl car-
versity
rier protein.
is directly incorporated into rapamycin [6]. The pipecolate-incorporating enzyme (PIE) is thought to act by a mechanism analogous to that proposed for non-ribosomal peptide synthetases (reviewed i n 7-9). In non-ribosomal peptide biosynthesis, amino acids are first activated through the formation of an aminoacyl adenylate, and covalently linked through a thioester bond to a phosphopanthetheine prosthetic group attached to an acyl-carrier-protein(ACP)-like domain of the synthetase before condensation takes place [lo]. The first evidence that pipecolate incorporation into macrolides occurs via a similar mechanism was obtained through the identification and purification of an enzyme from the immunomycin producer S. hygroscopicus var. ascomyceticus that covalently binds pipecolate, and hydrolyses ATP in a pipecolate-dependent fashion [l I]. The pipecolate-incorporating enzyme (PIE) is thought to catalyse both the transfer of the polyketide chain to an amino acid acceptor, and the final cyclisation of the macrolactone. The final steps in the biosynthesis of rapamycin are thought to be modifications of the 31-member lactone ring by hydroxylation and 0-methylation. The gene cluster governing rapamycin biosynthesis in S. hygroscopicus has been cloned and sequenced [12] and most of the activities necessary for the biosynthesis of the polyketide chain [12-141 as well as a number of genes governing specific hydroxylation or methylation [15] of the macrolide ring of rapamycin could be readily inferred from the DNA sequence. One gene, rapP, situated between two divergently transcribed giant open reading frames encoding polyketide synthase multienzymes, had significant similarity to genes of authentic non-ribosoma1 peptide synthetases, and was therefore thought to encode
-
Konig et al. ( E m J. Biochern. 247) OH
>24
I
3
b,
Rapamycin
5-2--m3
28 I
R
/J
FK506
7'
Immunomycin
Fig. 1. Structures of the related immunosuppressant macrolides rapamycin, FK 506 and immunomycin.
the activity responsible for the insertion of pipecolate into rapamycin. This report presents an analysis of the functional organisation of PIE from S. hygroscopicus as inferred from the deduced amino acid sequence, the disruption of rapP in S. hygroscopicus, the purification to homogeneity of recombinant PIE from Escherichia coli and Streptomyces coelicolor, and the determination of the extent to which the recombinant enzymes have undergone appropriate post-translational modification by attachment of 4'phosphopantetheine. In contrast to a previous report, where a PIE analog from the ascomycin producer S. hygroscopicur var. ascomyceticus was found to be a dimer under native conditions [I 1J , the recombinant PIE of rapamycin appeared to be a monomer as judged by gel filtration.
MATERIALS AND METHODS Materials. All molecular biology enzymes and reagents were from commercial sources. Thiostrepton and viomycin were gifts from Bristol-Myers Squibb and Pfizer, respectively. [y"PIATP (5000 Ci/mmol) and ~-[U-'~C]proline (50 mCi/mmol) were supplied by Amersham, ['*P]pyrophosphate (1 mCilmmo1) was from NEN and [1-14C]P-alanine(54.5 mCi/mmol) from NEN. DL-[ 1-"C]Pipecolic acid (1 mCi/mmol) was synthesised from 2-bromopyridine and ['"Clbarium carbonate (Sigma) according to a method adapted from that of Hemscheidt and Spenser [16].
527
Bacterial strains, plasmids and culture conditions. The rapamycin producer S. hygroscopicus NRRL 5491 and its derivatives were maintained on SY agar (Soluble starch 15 g/l, yeast extract 1 g/l, K,HPO, 1 g/l, MgSO, . 7 H,O 1 g/l, NaCl 3 g/l, Tes 30 mM, pH 7.4, agar 15 g/l), and cultivated in Tryptic Soy Broth with 10 g/1 glucose, pH 6.0, supplemented with I 0 mg/ml viomycin if required. S. coelicolor CH999 [I71 and Streptomyces lividans 511326 and their derivatives were cultivated i n YEME [I81 or tap water medium (5 g/l glucose, 10 g/l sucrose, 5 g/l tryptone, 2.5 g/l yeast extract, 36 mg/l EDTA, pH 7.1), supplemented with 50 (agar) or 10 mg/ml (liquid cultures) thiostrepton as required. Liquid cultures were grown at 30°C in roundbottomed flasks with shaking at 200-250 rpm. Infection with the att- actinophage KC515 [18] and its derivative @PIE (present work) were done on solid DNA medium supplemented with 10 mM MgSO, and 8 mM Ca(NO,), [18]. E. coli TG1 504 rec0: :Tn5 was obtained from Dr P. Oliver (Department of Genetics, University of Cambridge), E. coli BL21 DE3 [plyss] [19] was from N. Scott (Department of Biochemistry, University of Cambridge), E. coli ET12567 [20, 211 and E. coli SJ16 panD2- zad-220::TnlO [22] were gifts from Dr T. MacNeil (Merck Research Laboratories, Rahway, New Jersey) and Dr J. E. Cronan (University of Illinois), respectively. E. coli strains were grown in 2TY medium [23], supplemented with ampicillin (100 mg/ml) or chloramphenicol (30 mg/ml) as required. The plasmids described in. this work are derivatives of pUC119 [24], pT7-7 and pGP1-2 [25], or pRM52 [26]. Recombinant DNA methods. Standard techniques for DNA manipulation were performed as described i n Sambrook et al. [23j. Polymerase chain reactions (PCR) were performed using a programmable Robocycler Gradient 40 (Stratagene), in reaction mixtures containing Pfu buffer, 20 pmol of each primer, 250 nmol of each dNTP, 100 ng template DNA and 10 o/o (by vol.) dimethylsulfoxide in a total volume of 100 p1. The DNA was denatured by heating to 95°C for 5 min, and the primers were annealed at 48°C while 1 U Pfu DNA polymerase (Stratagene) was added. After 25 thermal cycles (denaturation at 95 "C for 1.5 min ; annealing at 48 "C for 1 min ; and extension at 72 "C for 3 min), the reaction mixture was incubated at 72°C for 7 min to promote end filling of the PCR product. All PCR products were verified by sequencing on both strands. Double-stranded DNA cyclo-sequencing reactions were analysed on an ABI 373A sequencer. Computer analysis of DNA sequences were performed using the programs from the University of Wisconsin Genetics Computer Group [27] and the Staden package [28] of programs. Transformation and transfection of Streptomyces protoplasts, and infection of S. Iividans JI1326 with KC51.5 and @PIE, were performed as described by Hopwood et al. [IS]. Infection of S. hygroscopicus NRRL 5491 with @PIE was done according to Lomovskaya et al. [29] on DNA plates supplemented with MgSO, and Ca(NO&. Lysogens were selected by overlaying the plates with 50 mg/ml (final concentration) viomycin 24 h postinfection. Strains that had undergone a second recombination event deleting the integrated phage were identified by selecting viomycin sensitive isolates after three rounds of non-selective growth and sporulation on SY plates. The insertion and subsequent loss of the phage were demonstrated by genomic Southern hybridization. Construction of the PIE expression plasmids and @PIE. A 4807-bp MscI fragment of 3.-D [12], containing all but the first 529 bp of rupP, was ligated into pUC119 linearised with H i n d 1 to create plasmid pP1. The 5' end of rapP was amplified by PCR so as to create a NdeI site overlapping the start codon, and this fragment was cloned blunt-ended into HincII-cut pUC119 to create pP2. A modified rapP casette with the engi-
528
Kiinig et al. (Eur. b. Biochem. 247)
neered NdeI site was constructed by subcloning the insert of pP2 into pP1 via the vector-based PstI site and the NcoI site that naturally occurs at position 1049 in rupP. This cassette was further subcloned into pT7-7 to create the E. coli expression plasmid pT7-PIE, and into pRM52 to create the Sfreproinyces expression vector pRM-PIE. @PIE was constructed by end-repairing and inserting a 2658-bp BaniHl-Hind111 fragment, entirely internal to the rupP open reading frame, into the unique PvuII site of KC515. Purification of recombinant PIE from E. coli BL21 DE3 [pLysS, pT7-PIEI. 800 ml 2TY me.dium, supplemented with 25 mg/ml timentin (1 : 15 mixture of the [j-lactamase inhibitor clavulanic acid and the /j-lactam antibiotic ticarcillin, Beecham Research Laboratories) in a 2-1 flask, were inoculated with 40 ml of a seed culture grown i n 2TY medium (supplemented with 100 mg/ml ampicillin) at 37°C to an A,,, of 0.4. The main culture was grown to an A,,,,, of 2.0 at 30°C. After cooling the cells to 23°C for 1 h expression was induced by the addition of isopropylthio-/j-D-galactosideto a final concentration of 20 mg/ ml. The culture was incubated with shaking for a further 4 h at 23°C. The following steps of the purification procedure were carried out at 4°C. The cells were harvested by centrifugation for 10 min at 4000 rpm in a Sorvall GS3 rotor and were resuspended in 3 vol. 100 mM TricineNaOH, pH 8.0, 10 mM EDTA and I0 % glycerol (by vol.) (buffer A) containing 50 mg/ml lysozyme, 50 mg/ml DNAse I, 50 mg/ml RNAse A and 20 mg/ml trypsin inhibitor. Typically, about 8 g of cells were obtained from 1.6 1 of cell culture. The cells were passed twice through a French pressure cell. Unbroken cells and the insoluble material were removed by centrifugation (1 8 000 rpm, 1 h, SS34 rotor). DNA, together with contaminating proteins was precipitated by addition of streptomycin sulphate (Sigma) to 1 % (mass/vol.) and 107 g/I solid ammonium sulphate. Further solid ammonium sulphate was added to the supernatant (177 g/l), and the precipitated protein was collected by centrifugation and stored at -70°C. A Pharmacia fast protein liquid chromatography (FPLC) apparatus fitted with a pair of P-500 pumps and a GP-250 gradient programmer was used with a phenyl-Sepharose High Performance HiLoad (26 m m X 10 cm, Pharmacia) column that had been equilibratcd in buffer A with 0.4 M ammonium sulphate. The PIE-containing ammonium sulphate pellet was resuspended in 20 ml of the above buffer. The solution was centrifuged at 12000 g for 10 min at 4°C to remove particulate matter and applied to the column. The column was washed with two column volumes of buffer A with 0.4 M ammonium sulphate, before applying a gradient from 0.4 M to 0 M ammonium sulphate over 80 ml, at a flow rate of 2 ml/min. Eight 5-ml peak fractions containing recombinant PIE, as judged by SDS/PAGE, were pooled, diluted threefold with buffer A and applied to a MonoQ HR column (10 mmXlO cm, Pharmacia). The column was washed with 40 nil of buffer A with 0.15 M sodium chloride, before a shallow gradient from 0.15 M to 0.32 M sodium chloride over 40 ml was applied. PIE protein was eluted at around 0.26 M sodium chloride. Purification of recombinant PIE from S. coelicolor LpRM52-PIEl. 400 ml YEME medium supplemented with 10 mg/ml thiostrepton, in a 2-1 flask, were inoculated with 20 ml of a seed culture grown in YEME medium (supplemented with 10 mg/ml thiostrepton) at 28°C to mid-log phase. The main culture was grown for 6-7 days. The following steps of the purification procedure were carried out at 4°C. The cells were harvested by centrifugation for 20min at 8000rpm in a Sorvall GS3 rotor and were resuspended in 3 vol. buffer A containing 50 mg/ml lysozyme, 50 mg/ml DNAseI, 50 mg/ml RNAseA and 20 mg/ml trypsin inhibitor. 1.6 1 of cell culture yielded typically
about 30 g of cells. Purification of the recombinant protein was essentially as described above for E. coli. Protein methods. Protein concentration was determined by the dye-binding method of Bradford 1301. ''C-labelled proteins were resolved by SDS/PAGE, the gels were soaked in Amplify fluorographic reagent (Amersham International) for 15- 30 min, and a phosphor imager screen was exposed to the dried gel. The data were collected using a Phosphor Imager 425 (Molecular Dynamics) and analysed using the software provided. To test the solubility of the recombinant PIE, cells were broken using the French press, and a 250-ml portion of the lysate was centrifuged for 3 min at 13200 rpm. The supernatant was centrifuged for 1 h at 60000 rpm in a TLlOO ultracentrifuge equipped with a TLA-200.1 rotor. The pellet was washed with buffer A, resuspended in 250 p1 sample loading buffer, and analysed using SDSPAGE. For N-terminal sequence analysis 100-200 pmol of the target protein was resolved by PAGE, and electrophoretically transferred to ProBlott (ABI) membrane by the method of Matsudaira [31]. Bands were excised and sequenced using an Applied Biosystems model 470A pulsed-liquid protein sequencer fitted with an on-line model 120A analyser for detection of phenylthiohydantoin amino acids. Assay of enzyme activity. The formation of aminoacyl adenylate was measured performing an ATP/["P]PP, exchange reaction as described by Nielsen et al. [ I l l . A11 incubations contained 10 mM MgCI,, 2 mM ATP, 1 mM dithiothreitol, 0.1 mM EDTA, 1 mM tetrasodium pyrophosphate (approximately 100000 cpm), 5 mM amino acid and 50 pl enzyme solution in a total volume of 100 pl. The reaction mixtures were incubated at room temperature for 15 min, during which time the rate of incorporation of ["PIPP, into ATP was linear. The reaction was terminated by the addition of 400 pl of a 1 % suspension of activated charcoal, which was collected by filtration through glassfibre filters. Thioester formation was measured as described by Keller 1321 with the following modifications. Purified PIE was incubated with ''C-labelled pipecolate in the presence of 10 mM MgCI,, 2 mM ATP, 1 mM dithiothreitol, 0.1 mM EDTA and 50 mM Tricine, pH 8.0, for 20 min at 30°C. The protein was precipitated with I0 % trichloroacetic acid, washed three times with 5 % trichloroacetic acid and once with ethanol. For thioester analysis the dried pellet was redissolved in 100 p1 formic acid. Formic and performic acid treatment and thin-layer chromatography on silica gel with butanol/acetic acid/water, 4: 1: 1 (by vol.) were performed as described by Keller [32]. A phosphor imager screen was exposed to the dried silica plate. The data were collected using a Phosphor Imager 425 (Molecular Dynamics) and analysed using the software provided. Molecular mass determination. Native molecular mass was determined by gel-filtration on Superose 12 (Pharmacia) and Superdex S200 (Pharmacia) in buffer A supplemented with 100 mM NaCI. Ferritin (440 kDa), catalase (232 kDa), aldolase (158 kDa), and bovine serum albumin (67 kDa) were used as standards. Blue Dextran 2000 (2 MDa) was used to determine the exclusion volume of both columns. Rapamycin production assay. S. hygroscopicus and its derivatives were cultivated for 60-72 h in 100 ml liquid medium as described above. Methanolic extracts of mycelia were analysed on a Hewlett Packard HP 1090 HPLC system with a diode array ultraviolet detector set at 277 nm, fitted with a Beckman Ultrasphere RP165 (4.6 mmX 25 cm) reverse-phase column, eluted using a linear gradient of acetonitrile in water increasing from 60 % (by vol.) acetonitrile to 100 % (by vol.) acetonitrile over 30 min.
529
Konig et al. ( E M J. Biochem. 247)
A
I 0
1
I
I
I
I
500
I
I
I
I
I I 1000
I
I
I
I1 1542
aa
phosphopantetheine attachement site Motif
KLM
NO
AB
peptidebond formation
DEFG H I J K L M
C
adenylate formation
B
6
N
peptidebond formation
L-YREL----N--A--L L-YAEL----N--A--L
B V-----LKAG--P-Dp V-----LKRG--Y-PI
e
AY--YTSG-TG-PKGV AY--YTSG-TG-PKAV
3
N-YGP-E N-YGP-E
B
GEL---G-G--RGY-N-P-LT-E-F AQV---G-G--RGY-A-P-LT-Q-F
L' Y-TGD--RWL Y-SGD--RRL
C
E--GR-D-QVK-RG-RIE-GEIE E--OR-D-QVK-RG-RTE-GEIE
1%
LP--M-P LP--M-P
t
LT-NGK-----LP LT-NGK-----LP
2
GGHSL--M GGHSL- - S
K
YP-V--Q-RM AP-S--Q-QL
L
L--RHE-L I - -RHE-L V--RHE-L
I DMHHII-DG-S-- I MLHHIA-DG-S--V TVHHIA-DG-S--I N
LSX-GQ-DI--G-P-AGR LSH-AR2DV--G-P-ANR LSG-D**DI--A-P-ANR
0
GMFVNT-LA-R GMFVNT-VL-G
Fig.2. Sequence motifs common to peptide synthetases and PIE. (A) Schematic diagram of PIE showing the relative positioning of the conserved sequence motifs. Several conserved sequence motifs were identified in non-ribosomal peptide synthetdses (reviewed in [39]). (B) Conserved sequence motifs A - 0 are shown in the top row for each motif. The corresponding sequence(s) in PIE are shown beneath.
RESULTS AND DISCUSSION Sequence analysis. Sequence comparison of the deduced amino acid sequence of rapP to those of known peptide synthetases, including those for the biosynthesis of the cyclic peptide antibiotics gramicidin S [33, 341 and tyrocidine [35] in Bacillus brevis, the cyclo-lipoheptapeptide surfactin in Bacillus subtilis [35-381, and the cyclic undecapeptide cyclosporin A in Tolypocludium niveum [39] reveals the presence of highly conserved amino acid sequence motifs (Fig. 2). These motifs of six to ten amino acids appear in a defined order and in roughly the same positions in peptide synthetase modules [9, 401. The motifs A-J were shown to have functions in amino acid activation, ATP and phosphopantetheine binding [41]. The core sequence LGGXS (motif J) found at the C-terminus of the adenylation domain of every peptide synthetase module resembles the 4'-phosphopantetheine--binding site of acyl carrier proteins in fatty acid and polyketide synthases [lo], and constitutes the site for covalent attachment of the amino acid via the 4'-phosphopantetheine
prosthetic group. The amino acid sequence LGGHS with the covalent attachment point Ser is found at amino acid position 1033 of PIE (Fig. 2A). Conserved sequences further C-terminal from the cofactor attachment site (motifs K, L and M) have been termed spacer motifs [37, 38, 421. Recently, motif M (HHXXXDGXS) was proposed to function in epimerisation and peptide-bond formation [40, 431. The amino acid sequence HHIAGDGWS is situated at position 1226 of PIE, 193 amino acids C-terminal of the phosphopantetheine attachment site. Part of motif M bears significant similarity to a conserved sequence (HHXXXDG) in the active sites of chloramphenicol acetyltransferases and dihydrolipoamide acyltransferases (Fig. 2B, [43]). Lactonisation is a common mechanism for release of cyclic peptides and polyketides from their synthesising enzymes. Non-ribosomal peptide synthetases, including gramicidin S synthetase 2, ACV synthetase and surfactin synthetase 2, harbour a thioesterase domain at their C-terminus for substrate release 134, 37, 441, as do most
Konig et al. (ELK J. Biochern. 247)
530
r
ATP Ho%
2
PPI
0dUP
1
AMP
2
___)
PIE
RAp&B :::::::::::i::;i:: ...............
077 ......... ........
-5
3
0
14
Fig. 3. Incorporation of pipecolate into rapamycin by PIE. Pipecolate is adenylated (1) and subsequently bound to the 4'-phosphopantetheine arm in a thioester linkage (2). PIE is then thought to dock onto the C-terminal ACP domain of R A P S , for the transfer of the polyketide chain to the amino group of pipecolate (3), where cyclisation takes place (4). The domain structure shown fur PIE is inferred from the sequence analysis reported here and analyses of the protein structure of peptide synthetases [9, 52-54]
of the macrolide polyketide synthases, for example the multienzyines for the biosynthesis of erythromycin [3],avermectin 1201, and oleandomycin [45].However, neither PIE nor the rapamycin polyketide synthase contain a thioesterase motif, nor is there an activity with similarities to discrete thioesterase enzymes in the rup cluster. Therefore the transfer motif at the C-terminal part of PIE at position 1226 may be involved in macrolactone formation. A remarkable feature of PIE is the similarity of the N-terminal 470 amino acids to the C-terminal 470 amino acids, including the spacer motifs K, L and M. The presence of a second copy of the transfer motif M (HHXXXDGXS, positions 167177) at the N-terminus of PIE suggests its involvement in the transfer of the polyketide chain from the terminal ACP of the rapamycin polyketide synthase onto the pipecolyl-phosphopantetheinyl thioester on PIE. The proline-activating domain of the gramicidin S synthetase 2 also harbours a copy of motif M at its N-terminus, which has been proposed to be involved in the transfer of the activated and racemised D-phenylalanine from the gramicidin S synthetase 1. Accordingly, a 115-kDa N-terminal
proteolytic fragment of grainicidin S syiithetase 2 was shown to accept ''C-labelled phenylalanine from gramicidin S synthetase 1 in the absence of proline [46]. A similar transfer motif was also found near the start of o m 6 of the pksX locus of B. subtilis [47]. The deduced amino acid sequence of this open reading frame shows both peptide synthetase and polyketide synthase functions. The location of the motif M at the N-terminus of the orJX6 gene product suggests that it could be involved in the transfer of an (unidentified) intermediate assembled by other activities to the activated amino acid held on the N-terminal peptide synthetase module of the o f l 6 gene product. Based on the sequence analysis, a model for the mechanism of catalysis of PIE can be proposed (Fig. 3). First, pipecolate would be activated as an aminoacyl adenylate with the concomitant hydrolysis of ATP. Subsequently, the activated amino acid is linked through a thioester bond to a phosphopantetheine moiety attached to Ser1033 of PIE to form a covalent intermediate. It is not yet clear at which stage of biosynthesis pipecolate is incorporated into the macrolide chain. If head-to-tail growth were to be continued for this heterologous building block, acylation of
Konig et al.
(6tr:
53 1
J. Biochrnr. 247)
the pipecolate amino group by the activated polyketide chain would occur by the nucleophilic attack of the imino nitrogen electron pair of pipecolate on the thioester of the activated polyketide chain bound to the RAPS3 ACP. The resulting peptidol attached to PIE would then cyclise in a lactonisation reaction, possibly aided by the transfer motif at the C-terminus of PIE. Mutagenesis experiments are currently under way to test the proposed functions of the individual sequence motifs of PIE.
kDa
Disruption of rapP. S. hygroscopicus NRRL5491 produces several polyketide metabolites, and different gene clusters containing polyketide synthase and also peptide synthetase genes have been cloned and partially characterized from this strain (Konig, A. and Molnir, I. unpublished results; see also [28]). The rapamycin biosynthetic gene cluster had originally been cloned based on the similarity of the rapamycin polyketide synthase to that of erythromycin, and identified by the rigorous correlation of the predicted polyketide synthase, peptide synthetase, cyclodeaminase, methyltransferase and cytochrome P-450 hydroxylase activities with those necessary for the biosynthesis of rapamycin [12]. Recently, Lomovskaya et al. [29] created rapamycin-non-producers by deleting a region of the S. hygroscopicus chromosome harbouring the predicted rapamycin polyketide synthase and rapP. This experimental strategy, however, did not allow the authors to exclude the possible effects exerted on rapamycin production by any unintentional and undetected second site mutations, or the consequences of genomic instability that may have been caused by creating a large (78.4 kb) deletion on the S. hygroscopicus chromosome. To provide conclusive proof for the involvement of rupP in rapamycin biosynthesis, we disrupted this gene by insertional inactivation (data not shown). A derivative (@PIE) of the attachment site-deleted actinophage vector KC515 was used to lysogenise S. hygroscopicus through homologous recombination via a cloned internal fragment of rupR Viomycin-resistant colonies were isolated and shown to carry @PIE inserted into rupP by genomic Southern hybridizations (data not shown). These isolates did not produce detectable amounts of rapamycin, although production of another polyketide metabolite, elaiophyllin 1481, was undisturbed, making pleiotropic inhibition of polyketide production unlikely. To exclude the possibility that the loss of rapamycin production is the consequence of undetected second-site mutations, viomycinsensitive derivatives were isolated from the lysogens, and were shown by genomic Southern hybridizations to have lost the inserted phage through a second homologous recombination event, regenerating rupP. These viomycin-sensitive derivatives all produced rapamycin at a level indistinguishable from that of the wild-type S. hygroscopicus.
97
-
66
-
45
-
29
-
Expression of rapP in E. coli and S. coelicolor. PIE expression vectors containing a modified version of rupP where an NdeI site was created to overlap the start codon were constructed using the E. coli plasmid pT7-7 1251 and a modified version of the Streptomyces expression vector pRM5 [17], pRM52 [26], as described in Materials and Methods section. E. coli BL21 DE3 [pLysS, pT7-PIEI and S. coelicolor CH999 [pRM-PIE] both produced PIE in substantial amounts, as judged by SDSPAGE. Purification of recombinant PIE from E. coli and S. coelicolor. E. coli BL21 DE3 [pLysS, pT7-PIEI and S. coelicolor CH999 [pRM-PIE] were cultivated and the recombinant PIE proteins were purified as described in Materials and Methods, by differential precipitation with ammonium sulphate, hydrophobic-interaction chromatography and anion-exchange chromatography on a Mono-Q column. Enzyme activity was measured by the formation of labelled ATP from ["PIPP, in the
205
1
2
3
4
-
116-
Fig. 4. Purification of recombinant PIE from E. coli. SDS/PAGE on a 4 % to 15 % gradient gel showing each individual stage of purification, stained with Coomassie blue. Lane M shows high-molecular-mass markers (Sigma). Lane 1, soluble fraction of induced E. coli BL21 [pLysS, pT7-PIE1, contains 16 mg total protein. Lane 2, ammonium sulphate cut (20-50 % saturation), contains 12 mg total protein. Lane 3, pooled fractions of a phenyl Sepharose HR 16/10 column (Pharmacia), contains 9 mg total protein. Lane 4, pooled fractions after anion-exchange chromatography on a Mono Q 10/10 column (Pharmacia), contains 4 m g total protein.
presence of the substrate amino acid. Fig. 4 shows the SDSI PAGE analysis of samples from the various stages of purification from E. coli, and Table 1 compares the degree of purification achieved at each step. Identical protocols were used for purification of the enzyme from both hosts, to yield approximately the same amount of pure protein as judged by SDS/PAGE. The N-terminal amino acid sequences of the recombinant PIE proteins isolated from E. coli as well as from S. coelicolor were determined as described in Materials and Methods, and were found to be in agreement with the deduced amino acid sequence. The N-terminal methionine was removed from the expressed proteins in both hosts, which is in agreement with the findings of Flinta et al. 1491 that in prokaryotes the N-terminal methionine is usually post-translationally removed if followed by an alanine.
Post-translationalmodification of recombinant PIE in E. coli and S. coelicolor. E. coli BL21 DE3 [pLysS, pT7-PIEI was cultivated as described in Materials and Methods, but in a tenfolddiluted medium containing an excess of 5 pCi ['"CJp-alanine. The cleared lysate was analysed by SDS/PAGE. Prolonged exposure of the gel during autoradiography revealed a band at 25 kDa, corresponding to the host-specific fatty acid synthase ACP, and a faint new band at 170 kDa, implying that a small part of the heterologously expressed PIE is phosphopantetheiny-
Kiinig et al. ( E m J. Biochem. 247)
532
Table 1. Purification of PIE from E. coli and S. coelicolor. PIE was purified to homogeneity from 8 g of E. coli or 30 g of S. coelicolor cells (wet mass). To determine the specific activity of the protein solution at each step of the purification, 11 mg total protein of E. coli (I) or S. coelicolor (11) was analysed by an ATP/[”P]PP, exchange assay as described in Materials and Methods.
Source
Stage
crude extract 20-50% (NHJ2S0, phenyl Sepharose Mono Q
E. coli
S. coelicolor
1
2
31.5
8.8 480 152 25.2 7.5
3
4
-
116-
97
-
66
-
45
-
29
-
Total activity
Specific activity
Recovery
Purification
nmol ATP/s
7c
-fold
36
nmol ATP . mg protein ’ s-’ 0.3 0.9 2.0 4.1
100 84 33 19
3 7 14
88.2 51.7 35.6
0.58 2.05 4.75
100 59 40
192 162 63
640 180
crude extract 20-50% (NH,),S04 phenyl Sepharose Mono Q
M
kDa 205
Protein
Fig. 5. Examination of post-translational modification of PIE. To radiolabel 4’-phosphopantetheinylated proteins specifically, E. coli BL21 DE3 [pLysS, pT7-PIEI and E. coli BL21 DE3 [pLysS, pT7-71 as a con-
trol, were grown in 0.2TY medium which was supplemented with 5 pCi [l-’4C]p-alanine1551. SDYPAGE on a 4 c/o to 15 % gradient gel shows the crude cell extracts of E. coli BL21 DE3 [pLysS, pT7-71 in lane I and E. coli BL21DE3 [pLysS, pT7-PIEJ in lane 2. The gel was stained with Coomassie blue. Lanes 3 and 4 are the autoradiographs of lane 1 and 2. In lane 4 a band of a protein of the size of PIE is clearly visible, indicating that a proportion of PIE is 4’-phosphopantetheinylated. This band is not present in the control. Lane M shows high-molecular-mass standards (Sigma).
lated (Fig. 5). Confirmation of this result was sought by labelling the 4‘-phosphopantetheine group of PIE with radioactive pipecolate in the presence of ATP. Autoradiography, however, failed to reveal any detectable radioactive pipecolate covalently bound to trichloroacetic acid-precipitated PIE separated by SDS/PAGE. These results suggest that only a minor part of the recombinant PIE is being phosphopantetheinylated in E. coli. Recent results by Lambalot et al. suggest, there might be a specific phosphopantetheine transferase for every biosynthetic pathway involving acyl carrier proteins [SO]. Pfeifer et al. reported on a low level pantetheinylation of the peptide synthetase component tyrocidine synthetase 1 upon expression in E. coli [40].
1
1
3.5 8.2
In contrast, recombinant PIE produced by S. coelicolor CH999 [pRM-PIE] could readily be labelled with [‘4C]pipecolic acid. The label was not cleaved by treatment with formic acid, but treatment with performic acid under the same conditions released the enzyme-bound radioactivity completely, indicating that as expected pipecolate is covalently bound to the enzyme through a thioester linkage, and consequently the recombinant protein had been phosphopantetheinylated by the Streptomyces host (data not shown). Quantitative analysis of the protein-bound imino acid indicated about 70-80% of PIE to be pantetheinylated. This result is in good agreement with the amount of 60 70 found for another peptide synthetase, &(L-a-aminoadipy1)-Lcysteinyl-D-valine synthetase, purified from its native host Streptomyces clavuligerus [SI 1.
Molecular mass determination. Gel filtration of PIE from both hosts showed the enzyme to behave as a monomer independent of its degree of purification and of the buffer system used (data not shown). This is contrary to the results of Nielsen et al. [I 11 who found the analogous enzyme of the ascomycin producer S. hygroscopicus, var. ascomyceticus to be a dimes under native conditions. Conclusions. Comparison of the deduced amino acid sequence of rapP from S. hygroscopicus with that of authentic peptide synthetases revealed a single amino acid activation domain characterized by the presence of conserved motifs for ATP and cofactor binding, ATP hydrolysis and aminoacyl adenylate formation, and two mutually homologous flanking regions with conserved spacer motifs for peptide-bond formation and substrate and product transfer. These data support a simple model for the incorporation of pipecolate into the macrocycle of rapamycin. rapP was disrupted on the chromosome of S. hygroscopicus by insertion of a recombinant phage, leading to the abolition of rapamycin production. The production of the antibiotic was reestablished in isolates that have lost the inserted phage through a second recombination event. To investigate further the proposed mechanism of ring closure, site-directed mutagenesis followed by in vivo replacement of the altered gene will help to assess the importance of particular amino acid side chains in the catalysis. Mutagenesis of the putative acyl-transfer motifs (HHIAGDGWS), for instance, could help to prove the direct involvement of these residues in catalysis, and could also provide further proof for the formation of an acyl-enzyme intermediate in the transfer reaction. Heterologous expression of rapP in E. coli using the pT7-7 system, and in S. coelicolor using a pRM52-based expression
Konig et al. (Eul: J. Biuchem. 247) vector yielded recombinant proteins of the predicted size and Nterminal sequence. Recombinant PIE isolated from S.coelicolor could b e labelled with ['4C]pipecolate, indicating that the enzyme was post-translationally modified by the insertion of a 4phosphopantheinyl swinging arm. The inability to form the aminoacyl thioester with PIE produced in E. coli is likely to b e attributed to the lack of phosphopantetheinylation. The authors are indebted to K. Chater for supplying organisms and phages, and to C. R. Hutchinson and N. Lomovskaya for sharing their results before publication. We thank J. Lester and K. Pennock for DNA sequencing, and M. Weldon and L. Packman (Cambridge Center for Molecular Recognition) for N-terminal protein sequence determination. This project was supported by a project grant from the Wellcome Trust.
REFERENCES 1. Vezina, C., Kudelski, A. & Seghal, S. N. (1975) Rapamycin (AY-22, 989), a new antifungal antibiotic. I. Taxonomy of the producing Streptomycete and isolation of the active principal, J . Antibiot. (Tokyo) 28, 721-726. 2. Metcalfe, S. M. & Richards, F. M. (1990) Cyclosporine, FK-506, and rapamycin - some effects on early activation events in serum free, mitogen stimulated mouse spleen cells, Transplantation. (Baltimore) 49, 798-802. 3. Cortes, J., Haydock, S. F., Roberts, G. A., Bevitt, D. J. & Leadlay, P. F. (1990) An unusually large multifunctional polypeptide in the erythromycin-producing polyketide synthase of Saccharopolysporu erythraeu, Nature 348, 176- 178. 4. Donadio, S., Staver, M. J., McAlpine, J. B., Swanson, S. J. & Katz, L. (1991) Modular organization of genes required for complex polyketide biosynthesis, Science 252, 675-679. 5. Cheng, Y. R., Fang, A. & Demain, A. L. (1995) Effect of amino acids on rapamycin biosynthesis by Streptomyces hygroscopicus, ,4ppl. Microbiol. Biotechnol. 43, 1096- 1098. 6 . Paiva, N. L., Demain, A. L. & Roberts, M. F. (1993) The immediate precursor of the nitrogen containing ring of rapamycin is free pipecolic acid, Enz. Microb. Technol. 15, 581-585. 7. Kleinkauf, H., von Dohren, H., Billich, A,, Lawen, A,, Peters, H. & Zocher, R. (1989) Novel microbial products for medicine and agriculture (Demain, A. L., Somkuti, G. A., Hunter-Ceverd, J. C. & Rossmoore, H. W., eds) pp. 49-55, Society for Industrial Microbiology, Washington DC. 8. Marahiel, M. A. (1992) Multidomain enzymes involved in peptide synthesis, FEBS Lett. 307, 40-43. 9. Stachelhaus, T. & Marahiel, M. A. (1995) Modular structure of genes encoding multifunctional peptide synthetases required for nonribosomal peptide synthesis, FEMS Microbiol. Lett. 125, 3 14. 10. Schlumbohm, W., Stein, T., Ullrich, C., Vater, J., Krause, M., Marahiel, M. A,, Kruft, V. & Wittmann-Lieboid, B. (1991) An active serine is involved in covalent substrate amino acid binding at each reaction center of gramicidin-S synthetase, J . Biol. Chem. 266, 23 135-23 141. 11. Nielsen, J. B., Hsu, M. J., Byrne, K. M. & Kaplan, L. (1991) Biosynthesis of the immunosuppressant immunomycin - the enzymology of pipecolate incorporation, Biochemistry 30, 5789-5796. 12. Schwecke, T., Aparicio, J. F., MolnBr, I., Konig, A,, Khaw, L. E., Haydock, S. F., Oliynyk, M., Caffrey, P., CortCs, J., Lester, J. B., Bijhm, G. A,, Staunton, J. & Leadlay, P. F. (1995) The biosynthetic gene cluster for the polyketide immunosuppressant rapamycin, Proc. Nut1 Acad. Sci. USA 92, 7839-7843. 13. Aparicio, J. E, MolnBr, I., Schwecke, T., Konig, A,, Haydock, S. F., Khaw, L. E., Staunton, J. & Leadlay, P. F. (1996) Organization of the biosynthetic gene cluster for rapamycin in Streptomyces hygroscopicus - analysis of the enzymatic domains in the modular polyketide synthase, Gene (Amst.) 169, 9-16. 14. Haydock, S. F., Aparicio, J. F., Molnir, I., Schwecke, T., Khaw, L. E., Konig, A,, Marsden, A. F. A,, Galloway, I. S., Staunton, J. & Leadlay, P. F. (1995) Divergent sequence motifs correlated with the substrate specificity of (methy1)malonyl-CoA: acyl carrier
533
protein transacylase domains in modular polyketide synthases, FEBS Lett. 374, 246-248. IS. Molnir, I., Aparicio, J. F., Haydock, S. F., Khaw, L. E., Schwecke, T., Konig, A,, Staunton, J. & Leadlay, P. F. (1996) Organization of the biosynthetic gene cluster for rapamycin in Strepromyces hygroscopicus - analysis of genes flanking the polyketide synthase, Gene (Amst.) 169, 1-7. 16. Hemscheidt, D. T. & Spenser, I. D. (1993) Biosynthesis of anosmine: incorporation of the intact six-carbon chain of lysine and of pipecolic acid, J. Nut. Prod. 56, 1281-1287. 17. McDaniel, R., Ebert-Khosla, S., Hopwood, D. A. & Khosla, C. (1993) Engineered biosynthesis of novel polyketides, Science 262, 1546- 1550. 18. Hopwood, D. A,, Bibb, M. J., Chater, K. F., Kieser, T., Bmton, C. J., Kieser, H. M., Lydiate, D. J., Smith, C. P., Ward, J. M. & Schrempf, H. (1985) Genetic manipulation of Streptomycetes, a taboraroty manual, John Innes Foundation, Norwich. 19. Doherty, A. J., Connolly, B. A. & Worrall, A. F. (1993) Overproduction of the toxic protein, bovine pancreatic DNAse I, in Escherichia coli using a tightly controlled T7-promoter based vector, Gene (Amst.) 136, 337-340. 20. MacNeil, D. J., Occi, J. L., Gewain, K. M., MacNeil, T., Gibbons, P. H., Ruby, C. L. & Danis, S. J. (1992) Complex organization of the Streptomyces avermitilis genes encoding the avermectin polyketide synthase, Gene 115, 119-125. 21. MacNeil, D. J., Gewain, K. M., Ruby, C. L., Dezeny, G., Gibbons, P. H. & MacNeil, T. (1992) Analysis of Streptomyces avermitilis genes required for avermectin biosynthesis utilizing a novel integration vector, Gene (Amst.) 111, 61-68. 22. Jackowski, S. & Rock, C. 0. (1981) Regulation of coenzyme A biosynthesis, J. Bacteriol. 148, 926-932. 23. Sambrook, J., Fritsch, E. F, & Maniatis, T. (1989) Molecular cloning: a laboratory manual, 2nd edn, Cold Spring Harbor Laboratory Press, Cold Spring Harbor NY. 24. Vieira, J. & Messing, J. (1982) The pUC plasmids, an Ml3mp7derived system for insertional mutagenesis and sequencing with synthetic universal primers, Gene (Amst.) 19, 259-268. 25. Tabor, S. & Richardson, C. C. (1985) A bacteriophageT7 RNA polymerase promoter system for controlled exclusive expression of specific genes, Proc. Nut1 Acad. Sci. USA 82, 1074-1078. 26. Oliynyk, M., Brown, M. J. B., CortCs, J., Staunton, J. & Leadlay, P. F. (1996) A hybrid modular polyketide synthase obtained by domain swapping, Chem. Bid. 3, 833-839. 27. Devereux, J., Haeberli, P. & Smithies, 0. (1984) A comprehensive set of sequence-analysis programs for the VAX, Nucleic Acids Res. 12, 387-395. 28. Staden, R. (1984) A computer program to enter DNA gel reading data into a computer, Nucleic Acids Res. 12, 499-503. 29. Lomovskaya, N., Fonstein, L., Ruan, X., Stassi, D., Katz, L. & Hutchinson, C. R. (1997) Gene disruption and replacement in the rapamycin-producing Streptomyces hygroscopicus strain ATCC 29253, Microbiology (Reading) 143, 875-883. 30. Bradford, M. M. (1976) A rapid and sensitive method for the quantification of microgram quantities of protein utilizing the principle of protein-dye binding, Anal. Biochem. 72, 248-254. 31. Matsudaira, P. (1987) Sequence from picomole quantities of proteins electroblotted onto polyvinylidene difluoride membranes, J. Biol. Chem. 262, 10035-10038. 32. Keller, U. (1987) Actinomycin synthetases - multifunctional enzymes responsible for the synthesis of the peptide chains of actinomycin, J. Biol. Chem. 262, 5852-5856. 33. Kratzschmar, J., Krause, M. & Marahiel, M. A. (1989) Gramicidin-S biosynthesis operon containing the structural genes grsA and grsB has an open reading frame encoding a protein homologous to fatty acid thioesterases, J . Bacteriol. 171, 5422-5429. 34. Turgay, K., Krause. M. & Marahiel, M. A. (1992) 4 homologous domains in the primary structure of grsB are related to domains in a superfamily of adenylate forming enzymes, Mol. Microbiol. 6 , 529-546. 35. Weckermann, R., Furbass, R. & Marahiel, M. A. (1988) Complete nucleotide sequence of the tycA gene coding the tyrocidine synthetase-1 from Bacillus brevis, Nucleic Acids Res. 16, 11 841 11841.
534
Konig et al. (Eui: J. Biochem. 247)
36. Borchert, S., Patil, S. S. & Marahiel, M. A. (1992) Identification of putative multifunctional peptide synthetase genes using highly conserved oligonucleotide sequences derived from known synthetases, FEMS Microbiol. Lett. 92, 175-180. 37. Cosmina, P., Rodriguez, F., Deferra, F., Grandi, G., Perego, M., Venema, G. & Vansinderen, D. (1993) Sequence and analysis of the genetic locus responsible for surfactin synthesis in Bacillus subtilis, Mol. Microbiol. 8, 821 -831. 38. Fuma, S., Fujishima, Y., Corbell, N., Dsouza, C., Nakano, M. M., Zuber, P. & Yamane, K. (1 993) Nucleotide sequence of 5' portion of SrfA that contains the region required for competence establishment in Bacillus subtilis, Nucl. Acids Res. 21, 93-97. 39. Weber, G., Schorgendorfer, K., Schneider-Scherzer, E. & Leitner, E. (1994) The peptide synthetase catalyzing cyclosporine production in Tolypocladium niveunz is encoded by a giant 45.8-kilobase open reading frame, C L Genet. ~ 26, 120- 125. 40. Pfeifer, E., Pavela-Vrancic, M., von Dohren, H. & Kleinkauf, H. (1995) Characterization of tyrocidine synthetase-l (Tyl ) requirement of posttranslational modification for peptide biosynthesis, Biochemistry 34, 7450-7459. 41. Pavela-Vrancic, M., Pfeifer, E., Schroder, W., von Dohren, H. & Kleinkauf, H. (1994) Identification of the ATP binding site in tyrocidine synthetase-l by selective modification with fluorescein 5'-isothiocyanate, J. Bid. Chem. 269, 14962- 14966. 42. MacCabe, A. P., Riach, M. B. R. & Kinghorn, J. R. (1991) Identification and expression of the ACV synthetase gene, J. Biotechnol. 17, 91 -97. 43. De Crecy-Lagard, V., Marliere, P. & Saurin, W. (1995) Multienzymatic non ribosomal peptide biosynthesis - identification of the functional domains catalyzing peptide elongation and epimerization, C. R. A. Sci. Sex Ill Sci. Vie 318, 927-936. 44. Smith, D. J., Earl, A. J. & Turner, G. (1990) The multifunctional peptide synthetase performing the first step of penicillin biosynthesis in Penicillium chrysogenuin is a 421.073 Dalton protein similar to Bucillus brevis peptide antibiotic synthetases, EMBO J. 9, 2743-2750. 45. Swan, D. G., Rodriguez, A. M., Vilches, C., Mendez, C. & Salas, J. A. (1994) Characterization of a Streptomyces antibioticus gene encoding a type-I polyketide synthase which has an unusual coding sequence, Mol. Gen. Genet. 242, 358-362.
46. Kurotsu, T., Hori, K., Kanda, M. & Saito, Y. (1991) Characterization and location of the L-proline activating fragment from the multifunctional gramicidin-S synthetase-2, J. Biochem. 109, 763-769. 47. Albertini, A. M., Caramori, T., Scoffone, F., Scotti, C. & Galizzi, A. (1995) Sequence around the 159-degrees region of the Bacillus subtilis genome - the PKS-X locus spans 33.6 kb, Microbiol. (Reading), 141, 299-309. 48. Kaiser, H. & Kellerschierlein, W. (1981) Structure elucidation of elaiophylin - spectroscopy and chemical degradation, Helv. Chim. Acta. 64, 407-424. 49. Flinta, C., Persson, B., Jornvall, H. & Von Heijne, G. (1986) Sequence determinants of cytosolic N-terminal protein processing, Eur. J. Biochem. 154, 193- 196. SO. Lambalot, R. H., Gehring, A . M., Flugel, R. S., Zuber, P., Lacelle, M., Marahiel, M. A,, Reid, R., Khosla, C. & Walsh, C. T. (1996) A new enzyme superfamily - the posphopantetheinyl transferases, Chem. Biol. 3, 923-936. 51. Schwecke, T., Aharonowitz, Y., Palissa, H., von Dohren, H., Kleinkauf, H. & van Liempt, H. (1992) Enzymatic characterization of the multifunctional enzyme delta-(L-alpha-aminoadipy1)L-cysteinyl-D-vahe synthetase from Streptomyces clavuligerus, Eur. J. Biochern. 205, 687-694. 52. Skarpeid, H. J., Zimmer, T. L., Shen, B. & von Dohren, H. (1990) The proline activating activity of the multienzyme Gramicidin-S synthetase-2 can be recovered on a 11.5-kDa tryptic fragment, ELK . I . Biochem. 187, 627-633. 53. Skarpeid, H. J., Zimmer, T. L. & von Dohren, H. (1990) On the domain construction of the multienzyme gramicidin-S synthetase-2 - isolation of domains activating valine and leucine, Eul: J . Biochem. 189, 517-522. 54. Stachelhaus, T., Huser, A. & Marahiel, M. A. (1996) Biochemical characterization of peptides carrier protein (PCP), the thiolation domain of multifunctional peptide synthetases, Chem. Biol. 3, 913-921. 55. Gocht, M. & Marahiel, M. A. (1994) Analysis of core sequences in the D-Phe activating domain of the multifunctional peptide synthetase TycA by site directed mutagenesis, J. Bacteriol. 176, 2654- 2662.
