Biogenesis of the Pyoverdin Chromophore M. Böckmann, K. Taraz and H. Budzikiewicz* Institut für Organische Chemie der Universität zu Köln, Greinstr. 4. D-50939 Köln, Germany Z. Naturforsch. 52c, 3 1 9 -3 2 4 (1997); received April 17, 1997 Pseudomonas, Siderophores, Pyoverdin, Biogenesis, l:,N-Labelling After growing Pseudomonas aeruginosa in the presence of 2,4-[4-15N]-diaminobutyric acid (Dab) its incorporation into the quinoline chromophore of the pyoverdin produced by this bacterium could be shown by mass and NMR-spectroscopic techniques. In combination with earlier results it can thus be stated that the precursor of the chromophore is a condensation product of L-Dab and D-Phe or D-Tyr. A synthesis for [4-15N]-Dab is described. Introduction mophore, as there are ferribactins (2), 5,6- When growing in a surrounding lacking a suffi­ dihydropyoverdins (3) and their 7-sulfonic acids cient quantity of soluble iron many microorga­ (4) (Schröder et al., 1995). 2 comprises all struc­ 1 nisms give off so-called siderophores, i.e. sub­ tural elements for the formation of by ring clo­ stances with a high complexing constant for Fe3+. sure and introduction of a second hydroxyl group The typical siderophores of the so-called fluores­ and also explains the formation of isopyoverdins cent group of the genus Pseudomonas, the pyover- (5) (Jacques et al., 1995). For the ring closure a dins (Budzikiewicz, 1993) comprise a peptide mechanism corresponding to a Bucherer reaction leading to 3 can be assumed. chain consisting of 6 to 12 partially modified d - and L-amino acids bound with its N-terminus to Feeding experiments with isotope labelled Phe the chromophore (lS)-5-amino-2,3-dihydro-8,9-di- or Tyr established their incorporation into 1 for hydroxy-l//-pyrimido-[l,2a]quinoline-l-carboxylic different Pseudomonas spp. (Maksimova et al., acid (1). The peptide chain is probably synthesized 1992; Stintzi and Meyer, 1993; Novak-Thompson through a multi-enzyme thiotemplate mechanism and Gould, 1994). However, 3,4-dihydroxy-Phe involving peptide synthetases. They activate the (DOPA) was not accepted in agreement with the constituent amino acids as their adenylates presence of only one hydroxyl group in 2. Appa­ 1 (Georges and Meyer, 1995; Menhart and Viswana- rently the second hydroxyl group of (and the 6 tha, 1990) and are probably also responsible for third one in the hypothetical intermediate - note the L/D-isomeration (Kreil, 1994) as had been the formation of 7 by a genetically modified non- shown for the peptide antibiotics (Kleinkauf and fluorescent Pseudomonas aeruginosa: Longerich et Döhren, 1987). al., 1993) is introduced at a later stage. Condensa­ Regarding the biosynthesis of the chromophore tion of D-Tyr and L-Dab would lead directly to 2 two schemes were advanced. The first (Fig. 2) is (similar condensation products of L-Dab with based on the isolation of compounds accompany­ other amino acids were also observed in the pep­ ing the pyoverdins in the culture broth which pos­ tide chain of pyoverdins, (e.g., Demange et sess the same peptide chain but differ in the chro- al., 1990; Gipp et al., 1991). Based on the observation that for non-fluores- cent mutants of Pseudomonas putida M whose py­ rimidine biosynthesis was blocked at different Abbreviations: Common amino acids, 3-letter code; stages the production of the pyoverdin was re­ FAB-MS, fast atom bombardement mass spectrometry; stored by auxotrophic dihydroorotate, it was con­ TMS, tetramethylsilane; DSS, 2,2-dimethyl-5-silapentan- cluded that the latter (and, therefore, Asp rather 5-sulfonate. than Dab) is a precursor of the tetrahydropyrimi- * Part LXX of the series „Bacterial Constituents“. For part LXIX see Budzikiewiczet al., 1997. dine part of 1 and that the carboxyl group of Phe Reprint requests to Prof. Dr. H. Budzikiewicz. is lost. The reaction sequence depicted in Fig. 5 Telefax: +49-221-470-5057. was proposed (Maksimova et al., 1993; Blazhevich 0939-5075/97/0500-0319 $ 06.00 © 1997 Verlag der Zeitschrift für Naturforschung. All rights reserved. D 320 M. Böckmann et al. ■Biogenesis of the Pyoverdin Chromophore and Maksimova, 1994). In order to clarify this Synthesis o f 2,4-[4-I5N]-diaminobutyric acid point [4-15N]-Dab was synthesized and added to dihydrochloride the culture medium of Pseudomonas aeruginosa Methyl 2,4-[4-1:iN]-diphthalimidobutyrate: ATCC 15692 (Briskot et al., 1989; Demange et al., Methyl 4-bromo-2-phthalimidobutyrate (12.65 g) 1990) and pyoverdin D (=PaA) (succinic acid side (synthesized according to Logusch, 1986), potassi- chain) was isolated. um[15N]phthalimide (5 g) and dimethylformamide (80 ml) were kept at 100 °C for 24 hours. After cooling to room temperature, the reaction mixture Material and Methods was treated with aqueous acetic acid ( 0.1 m, Spectroscopy 150 ml) and the resulting mixture was extracted with three 50 ml portions of CHC13. The CHC13 Mass: Finnigan MAT 900 ST (FINNIGAN, Bre­ was evaporated under reduced pressure. The pro­ men), matrix: Thioglycerol/dithioethanol. duct was chromatographed on silicagel with ether: NMR: Bruker AM 300 (BRUKER, Karlsruhe). hexane 1:1 (v:v). In doing so the unreacted part of For 13C-experiments samples of 20 mg and for [15N]phthalimide could be recovered. Yield 7.40 g 15N-experiments samples of 50 mg were dissolved (70% ), m.p. 160 °C. in 0.6 ml 0.1 m phosphate buffer (pH 4.3), brought 2,4-[4-15N]-diaminobutyric acid dihydro­ to dryness and redissolved in 0.6 ml D 20/H20 1:9 chloride: (v/v). 13C chemical shifts are given relatively to A mixture of methyl 2,4-[4-l;,N]-diphthalimido- TMS with the internal standard DSS using the cor­ butyrate (7.4 g), conc. hydrochloric acid (35 ml) relation ö(TMS) = ö(DSS)-1.61 ppm. For I:,N ex­ and glacial acetic acid (15 ml) was heated to periments 15NH4C1 was used as external standard. 100 °C for 36 hours. After the reaction mixture The chemical shifts are given in relation to was set aside at -18 °C overnight, phthalic acid was CH315N 0 2 using the relation 6(CH 315N 0 2) = filtered of and the filtrate was evaporated to dry­ 6 (15NH4Cl)-352.9 ppm. ness. Absolute ethanol was added to the residue and after two days 2,4-[4- 1:,N]-diaminobutyric acid dihydrochloride crystallized from the solution. Bacterial Growth The product was collected and washed with ace­ Pseudomonas aeruginosa was grown in 250 ml tone and ethanol. Yield 1.81 g (51 %). 15N-NMR: culture medium in 500 ml Erlenmeyer flasks with -342.5 ppm (sample: 20 mg, 0.6 ml D 20/H20 1:9 passiv aeration, rotary shaking (100 rpm) and (v/v), pH 1.5). light. The bacteria were grown in two steps. The first 500 ml culture medium consisted of 6.5 g Na- Results D-gluconate, 2.0 g KH 2P 04, 2.5 g (NH4)2S 0 4 and [15N]pyoverdin D isolated from the bacteria 0.25 g M gS04-7 H 20 , pH adjusted to 7.1 with grown with [4-15N]-Dab showed in FAB mass NaOH. After 36 hours 1.5 1 culture medium con­ spectrometry a prominent ion [M+H]+ at m/z 1335 taining 19.5 g Na-D-gluconate, 6 g KH2P 04, 0.75 g shifted by 1 u with respect to unlabelled pyoverdin M gS04-7H 20 , and 0.5 g D/L-[4- 15N]-Dab were D. The incorporation of labelled Dab into the added and the bacteria grown for further 48 hours. Isolation and purification is described elsewhere (Demange et al., 1990). Chemicals HO. 9 R: dicarboxylic acid For the experiments comercial chemicals of p.a. side chain quality were used. The potassiumf'^Njphthalimide was of 98% isotopic purity (C A M BRID G E ISO ­ TOPE LABORATORIES, Cambridge MAI 1 USA). Fig. 1. Structure of pyoverdins. M. Böckmann et al. ■Biogenesis of the Pyoverdin Chromophore 321 HOOC,„ peptide peptide N H , N H , peptide peptide 2H R: dicarboxylic acid side chain pyoverdin 1 Fig. 2. Proposed biogenesis of pyoverdine starting from L-Dab and D-Tyr. HO NH, NHCHO Fig. 3. Proposed biogenesis of pseudoverdin. R: dicarboxylic acid side chain Fig. 4. Proposed biogen­ esis of isopyoverdins. 322 M. Böckmann et al. ■Biogenesis of the Pyoverdin Chromophore HOOC\ ^ \ ^ ° H N \ ^ .N H O H NH, HOOC^^-^^OH T T H O O C ^ ^ ^ O HOOC ^NH ^NH OCX NH-, oa NHi oa”“NB, HOOC HOOC NL »N HO'Y ^ v/ N\^N NHi HONH, 3 1 Fig. 5. Proposed biogenesis of the pyoverdin chromophore 1 starting from dihydroorotic acid and Phe. /C ^ V HN " C t t , H N C H H i N O t H 2 I I I I II 0=C /C—coo o=c v / C — C O O " 0= C / C — coo N o N H N H H H H M dihydroorotate cai b am y la sp a ra g a te o ro ta te O /C ^ H N C H /C^ H N C H 0=C^ /CH I II c —COO N o=cv ^c—coo O — P O C H 2 o N p r p p H+ CO, H v A orotate- orotidylate- phosphoribosyl- decarboxylase transferase O H O H O H O H Fig. 6. Role of dihydroorotate and aspartate in the pyrimidine biosynthesis. M. Böckmann et al. ■Biogenesis of the Pyoverdin Chromophore 323 HOOC n h 2 I5n h 2 HO N ' + n'/ H COOH Fig. 7. r15Nlpyoverdin isolated after growing IX a NHi HO NHR with [4- N]-Dab.