Pseudomonas fluorescens CHA0 Produces Enantio-pyochelin, the Optical Antipode of the Pseudomonas aeruginosa Siderophore Pyochelin*

The siderophore pyochelin is made by a thiotemplate mechanism from salicylate and two molecules of cysteine. In Pseudomonas aeruginosa, the first cysteine residue is converted to its D-isoform during thiazoline ring formation whereas the second cysteine remains in its L-configuration, thus determining the stereochemistry of the two interconvertible pyochelin diastereoisomers as 4 ′R, 2 ″R, 4 ″R (pyochelin I) and 4 ′R, 2 ″S, 4 ″R (pyochelin II). Pseudomonas fluorescens CHA0 was found to make a different stereoisomeric mixture, which promoted growth under iron limitation in strain CHA0 and induced the expression of its biosynthetic genes, but was not recognized as a siderophore and signaling molecule by P. aeruginosa. Reciprocally, pyochelin promoted growth and induced pyochelin gene expression in P. aeruginosa, but was not functional in P. fluorescens. The structure of the CHA0 siderophore was determined by mass spectrometry, thin-layer chromatography, NMR, polarimetry, and chiral HPLC as enantio-pyochelin, the optical antipode of the P. aeruginosa siderophore pyochelin. Enantio-pyochelin was chemically synthesized and confirmed to be active in CHA0. Its potential biosynthetic pathway in CHA0 is discussed.

Iron is a cofactor for many redox-dependent enzymes and thus essential for most living organisms including bacteria. But despite its abundance on earth, iron is not freely available to microorganisms under aerobic conditions, as it forms poorly soluble ferric hydroxides in the environment or is tightly bound to transport and storage proteins in mammalian hosts. To acquire iron, bacteria have evolved sophisticated strategies, the most common of which is the production of iron-chelating molecules termed siderophores (1,2). Under iron-limiting growth conditions these molecules are secreted to the envi-ronment where they chelate ferric iron and deliver it to the bacterial cytoplasm via specific membrane-associated receptor proteins.
The siderophore pyochelin, which is the focus of this study, was isolated in the late 1970s from iron-deficient cultures of Pseudomonas aeruginosa ATCC 15692 (strain PAO1), and its chemical structure was established as 2-(2-o-hydroxyphenyl-2thiazolin-4-yl)-3-methylthiazolidine-4-carboxylic acid (3)(4)(5)(6). Pyochelin forms a 2:1 complex with ferric iron but despite recent physiochemical and crystallographic data, which suggest it to be a tetradentate ligand, the structure of the biologically relevant ferric complex is still unclear (7)(8)(9). Pyochelin has three chiral centers at the C4Ј, C2Љ, and C4Љ positions and is extracted from P. aeruginosa PAO1 as a mixture of two interconvertible diastereoisomers whose absolute configuration was determined as 4ЈR,2ЉR,4ЉR (pyochelin I) and 4ЈR,2ЉS,4ЉR (pyochelin II) ( Fig. 1 and Refs. 10,11). The presence of iron(III) and zinc(II) were shown to induce a shift from pyochelin II to pyochelin I by converting the S configuration to the R configuration at the chiral center C2Љ (12,13). Similar metal-induced shifts may occur in the other diastereoisomer pairs (Fig. 1).
Pyochelin is a condensation product of salicylate and two cysteinyl residues, its biosynthesis in P. aeruginosa requires proteins encoded by the two divergent operons pchDCBA and pchEFGHI (14 -17). The PchA and PchB enzymes convert chorismate into salicylate (18,19), which is then coupled to the cysteine moieties by a thiotemplate mechanism involving the salicylate adenylating enzyme PchD, the peptide synthetases PchE and PchF, and the reductase PchG (15,20,21). Specialized tailoring domains in PchE (E domain) and PchF (MT domain) are responsible for the epimerization of the L-cysteinyl to the D-cysteinyl residue during formation of the thiazoline ring (22) and for methylation of the nitrogen in the thiazolidine ring (20), respectively. The expression of the pyochelin biosynthetic genes is strongly induced by pyochelin (14) and depends on the AraC-type regulator PchR (23,24).
Pyochelin has also been isolated from other pseudomonads and closely related bacteria (25)(26)(27)(28)(29) although the configuration of the three chiral centers was not always determined. Interestingly, we discovered that the PchE peptide synthetase of the sequenced Pseudomonas fluorescens strain Pf-5 (30) lacks the epimerase domain required to generate the R configuration at the asymmetric center C4Ј, which prompted us to analyze the pyochelin cluster of the closely related P. fluorescens strain CHA0 and to elucidate the structure of the molecule that it specified. Here we report that CHA0 and Pf-5 synthesize the enantiomer of pyochelin for which we propose the name enantio-pyochelin ( Fig. 1). We show that enantio-pyochelin promotes growth under iron-limiting conditions and induces the expression of its biosynthetic genes in CHA0, but is not recognized as a siderophore and signaling molecule by P. aeruginosa.

EXPERIMENTAL PROCEDURES
Chemicals-Metal-free silica was obtained by gentle stirring of Merck Geduran Kieselgel Si 60 (40 -63 m) with 1 N HCl at 25°C for 12 h. The resulting suspension was filtered and washed several times with MilliQ water until the pH of the filtrate was between 5 and 6. The silica was then dried, first under reduced pressure and afterward in an oven at 110°C for 48 h.
Media and Growth Conditions-Bacteria were routinely grown on nutrient agar and in nutrient yeast broth (31) at 37°C (P. aeruginosa and Escherichia coli) or 30°C (P. fluorescens). For (enantio-)pyochelin production and green fluorescent protein (GFP) 2 reporter assays, strains were cultivated in the complex medium GGP (32) in which iron is present but not imme-diately accessible (probably because it is bound to proteins and peptides), thus inducing the expression of (enantio-)pyochelin biosynthesis and uptake genes. As siderophore-negative mutants grow well in GGP, the minimal medium M9 (33) with 0.5% glycerol as a carbon source was used for siderophore utilization assays. Iron limitation was achieved in this medium by adding the iron chelator 2,2Ј-dipyridyl at 500 M. Antibiotics were added to the growth media at the following concentrations: tetracycline 25 g ml Ϫ1 for E. coli and 125 g ml Ϫ1 for Pseudomonas. To counterselect E. coli donor cells in gene replacement experiments, chloramphenicol was used at a concentration of 10 g ml Ϫ1 ; mutant enrichment was performed with tetracycline at a final concentration of 20 g ml Ϫ1 and carbenicillin (for P. aeruginosa) or cycloserine (for P. fluorescens) at final concentrations of 2 mg ml Ϫ1 and 50 mg ml Ϫ1 , respectively.
DNA Manipulation and Sequencing-Small-and large-scale preparations of plasmid DNA were made with the QIAprep Spin Miniprep kit (Qiagen, Inc.) and Jetstar kit (Genomed GmbH), respectively. DNA fragments were purified from agarose gels with the Geneclean II kit (Bio 101, La Jolla, CA) or the MinElute and QIAquick Gel Extraction kits from Qiagen (Qiagen, Inc.). DNA manipulations were performed according to standard procedures (33). Transformation of E. coli, P. aeruginosa, and P. fluorescens was carried out by electroporation (34). All constructs involving PCR techniques were verified by sequence analysis. Sequencing was performed with the BigDye Terminator Cycle Sequencing Kit and an ABI-PRISM 373 automatic sequencer (Applied Biosystems) or was carried out commercially. The DNA sequence of the P. fluorescens CHA0 pchR-pchDHIEFKCBA genomic region was deposited at GenBank TM under accession number EU088199. Sequences were compared with those from P. aeruginosa PAO1 and P. fluorescens Pf-5. Data base searches were conducted at NCBI using BLAST algorithms.
The pyoverdine and pyochelin-negative P. aeruginosa mutant PAO6399 was constructed with the previously described suicide plasmid pME7152 (39), which was mobilized from E. coli S17-1 (40) to the pchE deletion mutant PAO6310 (14) and chromosomally integrated with selection for tetracycline resistance. Excision of the vector was obtained by enrich- 2 The abbreviation used is: GFP, green fluorescent protein.
Construction of GFP Reporters-Translational fusions of pchD (CHA0) and pchE (PAO) to the gfp gene carried by the vector pPROBE-TTЈ (43) were constructed by overlap extension PCR as follows. A 0.4-kb PCR fragment 1 was amplified from CHA0 chromosomal DNA using the primers PB1 (ACG-TGAATTCATGGCGAACTCCCTGTGG) and PB9 (GCTGT-CTCCTGATGTTTTTTACG) and a 0.23-kb PCR fragment 2 was amplified from pPROBE-TTЈ using the primers PB11 (AAAAAACATCAGGAGACAGCATGAGTAAAGGAGAA-GAAC) and PB4 (GCCGTTTCATATGATCTGGG). Fragments 1 and 2, which are complementary to each other over a length of 20 nucleotides, were mixed in equimolar amounts, elongated during 10 PCR cycles with Taq polymerase and dNTPs and subsequently amplified during 20 PCR cycles with primers PB1 and PB4. The resulting PCR fragment C was cleaved with EcoRI and NdeI and ligated to pPROBE-TTЈ cut with the same enzymes. This yielded the pchD CHA0 Ϫ gfp reporter construct pME7585. The pchE PAO Ϫ gfp reporter plasmid pME7588 was constructed in a similar way. PCR fragment 1 (0.34 kb) was amplified from chromosomal DNA of PAO1 using the primers pchEgfp1 (ACGTGAATTCCTGCAGGAA-TACCGCCTG) and pchEgfp2 (GGGGGCTCCCTAGGGCA-AGC) and PCR fragment 2 (0.23 kb) was amplified from pPROBE-TTЈ with primers pchEgfp3 (GCTTGCCCTAG-GGAGCCCCCATGAGTAAAGGAGAAGAAC) and PB4. Subsequent overlap extension PCR yielded a 0.57-kb fragment, which was cleaved with EcoRI and NdeI and cloned into pPROBE-TTЈ.
Extraction and Purification of (Enantio-) pyochelin-Enantio-pyochelin was extracted and purified from CHA400 as follows. The cell-free supernatant from a 2-day culture grown in GGP medium was acidified to pH 1-2 and extracted with 1 volume ethyl acetate. For thin-layer chromatography, mass spectrometry, and polarimetric measurements, this extract was washed successively with 0.5 N HCl and MilliQ water. The organic phase was dried over Na 2 SO 4 , and filtered. Solvents were evaporated under reduced pressure. The resulting orange oily residue was purified by chromatography on a metal-free silica gel column eluted with a 90:10 CH 2 Cl 2 /acetone mixture. Enantio-pyochelin was obtained as a pale yellow powder after evaporation of solvents under reduced pressure.
For use in NMR experiments, chiral HPLC, and bioassays, the crude ethyl acetate extract was dried by evaporation, dissolved in methanol and further purified by HPLC using a VP 250/10 Nucleosil 100-7 C18 preparative column (Macherey-Nagel) and a Waters TM HPLC system equipped with a 2487 Dual Absorbance Detector. Aliquots (300 l) were injected into the HPLC system and separated at room temperature using an isocratic gradient consisting of 70% Solvent A (H 2 O ϩ 0.1% trifluoroacetic acid) and 30% Solvent B (95% acetonitrile ϩ 0.1% trifluoroacetic acid) and a flow rate of 1 ml/min. Two peaks with maximal absorbance at 235 and 260 nm (absorbance of the thiazoline ring) were collected and dried by evaporation. These peaks with retention times of 16 and 24 min, correspond to the two diastereoisomers of enantio-pyochelin as determined by NMR, chiral HPLC, and bioassays (see "Results"). Extraction and purification of pyochelin from P. aeruginosa was done in a similar way and has been previously described (14).
GFP Reporter Assays-Pseudomonas strains carrying translational gfp fusions were grown in 96-well black microtiter plates (Greiner bio-one) with a flat transparent bottom. Each well contained 200 l of GGP medium and was inoculated with 3 l of a bacterial preculture grown overnight in the same medium. HPLC-purified pyochelin or enantio-pyochelin was dissolved in methanol to a concentration of 10 mM (determined by the CAS assay (42)) and added to the growth medium at the concentrations indicated. Microtiter plates were incubated at 37°C (P. aeruginosa) or 30°C (P. fluorescens) with orbital shaking at 500 rpm. At each given time point growth (A 600 ) and green fluorescence (excitation at 480 nm and emission at 520 nm) were measured from triplicate cultures using a Fluostar fluorescence microplate reader (BMG Lab Technologies). For each individual measurement the green fluorescence value was divided by the respective A 600 value giving the specific fluorescence of cells expressed as relative fluorescence units. The green fluorescence of cells carrying the empty vector pPROBE-TTЈ was determined for background fluorescence correction.
Siderophore-mediated Growth Promotion-Utilization of pyochelin or enantio-pyochelin as a siderophore was measured with liquid growth assays. Microtiter wells containing 200 l of M9-glycerol minimal medium with or without the iron chelator 2,2Ј-dipyridyl at 500 M were inoculated with 3 l of precultures grown overnight in M9-glycerol medium. HPLC-purified pyochelin or enantio-pyochelin was added to the growth medium at 20 M. Growth (A 600 ) at 500 rpm was recorded by a microplate reader (BMG Lab Technologies).
Thin-layer Chromatography (TLC)-Analytical TLC was performed with Merck TLC silica gel 60F 254 on aluminum sheets using n-butyl alcohol/water/acetic acid 4:1:1 (v/v/v) as the mobile phase. Compounds were detected either by fluorescence at 365 nm or by spraying the TLC sheet with a solution of FeCl 3 in MeOH. Pictures were recorded with a Kodak DX7590 digital camera and processed using the Kodak EasyShare Picture Editor.
Mass Spectrometry-Electrospray mass spectrometry experiments were performed on a microTOF LC from Brucker Daltonics.
Polarimetry-Pyochelin and enantio-pyochelin samples were dissolved in acetone at a concentration of 10 mg ml Ϫ1 and optical activities were determined at 20°C in a Hellma glass cell (100-mm length) using a Perkin Elmer Model 341 polarimeter.
Separation of Pyochelin and Enantio-pyochelin by Chiral HPLC-Samples (10 g in 20 l methanol) of pyochelin or enantio-pyochelin were injected into a Merck Hitachi HPLC system equipped with a L-7450A Diode Array Detector, and separated on a Daicel Chiralcel OD-H analytical column at room temperature using an isocratic gradient consisting of 95% Solvent A (heptane ϩ 0.1% trifluoroacetic acid) and 5% Solvent B (ethanol ϩ 0.1% trifluoroacetic acid) and a flow rate of 1 ml/min. Pyochelin and enantio-pyochelin were detected by their absorption at 254 nm.
Synthesis and Purification of Pyochelin, Neopyochelin, and Enantio-pyochelin-Pyochelin and neopyochelin were synthesized using protocols published previously (44,45). The synthesis of enantio-pyochelin was inspired by the same protocol and started with the condensation of 2-hydroxybenzonitrile with L-cysteine in a buffered hydromethanolic medium (46). The resulting thiazoline, recrystallized from a mixture of ethanol and n-hexane, was then converted into the corresponding Weinreb amide, and the latter was reduced to an aldehyde with lithium aluminum hydride (44,45). The aldehyde was condensed with N-methyl-D-cysteine prepared from D-cysteine according to the procedure of Blondeau et al. (47). At this stage, the synthesis of enantio-pyochelin (and also that of pyochelin and neopyochelin), leads to a mixture of four stereoisomers (11,44,45). Among them, the two naturally occuring diastereoisomers of pyochelin or enantio-pyochelin were separated from the two "neo" isomers by chromatography on a metal-free silica gel column eluted with a gradient of acetone in CH 2 Cl 2 (from 5:95 to 30:70). The resulting yellow oily residue was converted into a pale yellow powder by precipitation in a mixture of acetone and n-hexane followed by evaporation under reduced pressure.

RESULTS
Comparison of the Pyochelin Biosynthetic Genes from P. aeruginosa with Their Counterparts in P. fluorescens-Putative pyochelin biosynthetic (pch) genes have been located in the recently sequenced genome of P. fluorescens strain Pf-5 (30). We sequenced the corresponding genes from the closely related strain CHA0. Comparison of the deduced amino acid sequences revealed identities of 98 to 99% between the pch genes in the two strains (not shown). When these genes were compared with the well-characterized pch genes of P. aeruginosa PAO1 (14 -17), three major differences were discovered. Firstly, the genomic organization of the pch genes is different (Fig. 2). In PAO1 the pch genes are organized in two operons flanking the regulatory gene pchR. In Pf-5 and CHA0 all pch genes seem to be contained within one operon. Second, an epimerization domain is absent from PchE Pf-5/CHA0 . Third, there is no homology between PchG PAO1 and its presumed counterpart in Pf-5 and CHA0 (named here PchK, see Fig. 2). A motif search indicated that PchK could have a reductase and/or epimerase function.
Biological Activity of the P. fluorescens Siderophore-To characterize the molecule specified by the pchDHIEFKCBA operon, a culture supernatant of the pyoverdine-negative strain CHA400 was extracted with ethyl acetate and the siderophore was purified by HPLC as described under "Experimental Procedures." We first tested its biological activity as a siderophore in growth promotion assays. As shown in Fig. 3, unmodified M9-glycerol medium contained sufficient iron to support growth of the siderophore-negative mutants CHA1085 (Fig.  3A) and PAO6399 (Fig. 3B) but when the medium was amended with the iron chelator 2,2Ј-dipyridyl, these mutants were no longer able to grow. Growth of CHA1085 was restored by the addition of 20 M siderophore purified from P. fluorescens CHA400 whereas equal amounts of HPLC-purified pyochelin from PAO1 had no effect, indicating that pyochelin cannot be used as a siderophore by P. fluorescens (Fig. 3A). Similar results  (30). The sequence of the pchR-pchDHIEFKCBA region in CHA0 is available at GenBank TM under accession number EU088199. Deduced amino acid sequences of P. aeruginosa and P. fluorescens pch genes (gray) are between 35 and 60% identical except for pchG (white) and its presumed counterpart pchK (black), which are not related. Note that pchE is smaller in the P. fluorescens strains than in P. aeruginosa PAO1 due to the absence of an epimerase coding region (boxed in pchE PAO ).  DECEMBER 7, 2007 • VOLUME 282 • NUMBER 49 were obtained in experiments performed with P. aeruginosa (Fig. 3B) where the endogenous siderophore pyochelin was able to promote growth of the pyoverdine-and pyochelin-negative mutant PAO6399, while no growth promotion was observed under iron limitation with the P. fluorescens siderophore.

Enantio-pyochelin, a Siderophore of Pseudomonas fluorescens
We further tested the potential of the P. fluorescens siderophore and of pyochelin for their signaling activities using gfp reporter constructs in which the ATG start codon of the pchE and pchD genes from PAO1 and CHA0, respectively, had been joined to the second codon of gfp. The activities of these translational fusions were tested in pch mutants of P. fluorescens and P. aeruginosa grown with or without siderophore addition in iron-limited GGP medium (Fig. 4). In the absence of any siderophore, pchD expression was very low in CHA1084 but increased in a dose-dependent manner when HPLC-purified P. fluorescens siderophore was added. The strongest response was obtained with a siderophore concentration of 20 M (Fig.  4A) and was not further increased when the siderophore was added at 40 M (not shown). These results demonstrate that the P. fluorescens siderophore is able to induce its own biosynthetic genes by an autoinduction circuit observed previously with pyochelin in P. aeruginosa (14,24). Addition of pyochelin to CHA1084 induced pchD expression to some extent, but the response was about 30-fold lower than that with the P. fluorescens siderophore (Fig. 4A). In a control, pyochelin was very active as a signaling compound in P. aeruginosa, near maximal pchE expression was achieved in PAO6310 with concentrations as low as 4 M (Fig. 4B). In contrast, no response was observed when the P. fluorescens siderophore was added to the growth medium. Taken together, these results demonstrate that the P. fluorescens siderophore is utilized for iron uptake and signaling in its own background, but is clearly not recognized by P. aeruginosa.
Structural Analysis of the P. fluorescens Siderophore-To determine the structure of the P. fluorescens siderophore, a crude preparation from CHA400 was washed and then purified on a metal-free silica gel column and analyzed using electrospray mass spectrometry. The ionic profile obtained presented a peak at m/z 325 characteristic of the pyochelin molecular ion (MϩH ϩ ). To test if the siderophore was neopyochelin as we initially suspected because of the missing epimerase domain in PchE CHA0 , we separated the diastereoisomers by TLC using a ternary solvent system developed for this purpose. As shown in Fig. 5, the two diastereoisomers of the P. fluorescens siderophore did not co-migrate with the two neopyochelin diastereoisomers but had the same frontal retention as synthetic pyochelin I and II. An HPLC-purified sample was therefore analyzed by NMR at 400 MHz, which confirmed that the sample was a mixture of two diastereoisomers, a major and a minor, in a 2:1 ratio. A COSY experiment was used to secure attributions of each proton (Table 1). With the exception of some slight ppm shifts in the signals of H5Ј and H5Љ, the spectrum was superimposable with that obtained under the same conditions for natural and synthetic pyochelin (data not shown). In contrast, no convergences were observed between the proton spectrum of the P. fluorescens siderophore and the spectrum of  neopyochelin. This observation was confirmed using NOESY or ROESY experiments in order to establish the spatial relationship between the protons present at the three chiral centers. For the major isomer, a strong correlation between H2Љ and H4Љ was detected indicating that these hydrogens are on the same side of the molecule. This correlation was absent in the minor isomer suggesting a trans relationship between the two hydrogens. The global correlation profile observed with NOESY experiments is in agreement with the two pyochelin diastereoisomers but not with the diastereoisomers of neopyochelin, thus confirming the results obtained by TLC analysis.
All the structural analyses performed so far strongly indicated that the P. fluorescens siderophore was identical to pyochelin extracted from P. aeruginosa while all biological tests showed it to be different. To resolve this discrepancy, the optical rotation of both compounds was determined by polarimetric experiments. In acetone, at 20°C, the P. fluorescens siderophore was levorotatory with an [␣] D 20 ϭ Ϫ15°while pyochelin was dextrorotatory with an [␣] D 20 ϭ ϩ7°. Although the two optical rotations were not identical in absolute values, the opposite sign strongly suggested an enantiomeric relationship between the two molecules. This assumption was corroborated by chiral HPLC separation of the two compounds (Fig.  6). Based on these structural investigations we conclude that the siderophore made by P. fluorescens CHA0 is the enantiomer of pyochelin, and we thus name it enantio-pyochelin. The same molecule seems to be made also by strain Pf-5 as verified by chiral HPLC (data not shown).
Chemical Synthesis and Biological Activity of Enantiopyochelin-We chemically synthesized enantio-pyochelin using a four-step procedure adapted from the chemical synthesis of pyochelin (44,45). Identity with the siderophore extracted from P. fluorescens was confirmed by TLC, chiral HPLC, and NMR. Both molecules had the same retention times on TLC (not shown), co-migrated on chiral HPLC (Fig. 6), and had perfectly superimposable high field NMR spectra (not shown). Natural and synthetic enantio-pyochelin were then compared in a bioassay which showed that both compounds were able to activate the pchD CHA0 Ϫ gfp reporter in CHA1084 (Fig. 7).

DISCUSSION
In this work we have shown that P. fluorescens CHA0 produces enantio-pyochelin, the optical antipode of the sid-erophore pyochelin made by P. aeruginosa PAO1. Enantiopyochelin was found to promote growth under iron-limiting conditions in P. fluorescens (Fig. 3A) and to act as an inducer of its own biosynthetic genes (Fig. 4A), thus resembling pyochelin, which is both a siderophore and a signaling molecule in P. aeruginosa (4,14). However, enantio-pyochelin did not promote growth in P. aeruginosa (Fig. 3B) and neither did pyochelin in P. fluorescens (Fig. 3A), suggesting that the (enantio-)pyochelinmediated iron uptake machinery is highly stereospecific in both species. In P. aeruginosa, iron uptake with pyochelin requires the outer membrane receptor FptA and the inner membrane permease FptX, which are encoded by the fptABCX transport operon (39,48,49). This operon is not conserved in P. fluorescens Pf-5. Iron uptake via enantio-pyochelin may instead require the outer membrane receptor encoded by PFL_3498 and adjacent genes specifying a classical ABC transport system (30). Indeed, deletion of the CHA0 homolog of PFL_3498 abolished utilization of enantio-pyochelin for iron uptake (see supplemental Fig. S1). Sequence comparison between the pyochelin receptor of P. aeruginosa and the enantio-pyochelin  Enantio-pyochelin, a Siderophore of Pseudomonas fluorescens DECEMBER 7, 2007 • VOLUME 282 • NUMBER 49 receptor of P. fluorescens showed that the two receptors are not closely related and that the amino acids interacting with pyochelin in the crystallized FptA receptor (7) are not conserved in PFL_3498 and its homolog in CHA0. 3 Which of the three chiral centers of pyochelin is important for receptor binding? A recent study on the binding of pyochelin and structurally related molecules to the FptA receptor suggests that the configuration of the C4Ј chiral center is not important for receptor binding and iron uptake. Docking and in vivo experiments showed binding of pyochelin I, pyochelin II, neopyochelin I, and neopyochelin II to FptA. Moreover, binding properties and iron uptake were not affected by removing the C4Ј chiral center. When both the C4Ј and the C2Љ chiral centers were removed, the molecule still bound to FptA but was no longer able to transport iron (50). From these studies and the results obtained here, which show that enantio-pyochelin cannot be utilized by P. aeruginosa, we suspect that the C4Љ chiral center may be crucial for receptor binding.
Not many naturally occurring siderophore enantiomers have been described so far. To our knowledge, the only other example is rhizoferrin, a siderophore which is made as R,R-rhizoferrin by the fungus Rhizopus (51) and as S,S-rhizoferrin by the soil bacterium Ralstonia pickettii (52). Interestingly, iron uptake was promoted equally well with both enantiomers in R. pickettii, while the fungus showed a certain preference for its own enantiomer. Stereospecificity of iron uptake systems is probably more common in general, and has been observed with several other siderophores such as parabactin, rhodotorulic acid, ferrichrome, and enterobactin, and their corresponding synthetic enantiomers (53)(54)(55)(56)(57)(58).
Stereospecificity is important also for the regulation of the pch genes and is particularly stringent in P. aeruginosa (Fig. 4). In this bacterium, the expression of the two biosynthetic oper-ons pchDCBA and pchEFGHI requires the transcriptional regulator PchR (14). In the presence of pyochelin, PchR binds to a conserved sequence element (PchR-box) in the promoter region of these operons and regulates their transcription (24). A PchR-box is present also in the promoter region of the enantiopyochelin biosynthesis operon pchDHIEFKCBA of P. fluorescens and we have shown here that its expression in CHA0 requires enantio-pyochelin (Fig. 4A). It is very likely therefore that enantio-pyochelin is the effector activating the PchR protein of CHA0. We do not know at this stage if the two PchRs can discriminate between pyochelin and enantio-pyochelin, and the protein domains interacting with the siderophores have not been identified. It cannot be excluded that stereospecificity observed on the level of gene expression is caused by stereospecificity of the two receptors only. At least in P. aeruginosa, the FptA receptor is involved in pyochelin-induced gene expression due to its role in the uptake of the PchR effector (i.e. pyochelin) (39).
The production of enantio-pyochelin by P. fluorescens CHA0 also raises questions regarding its biosynthesis. While the stereochemistry of the thiazoline ring is explained by the lack of the epimerase domain in PchE, it is unclear how the epimerization of the thiazolidine ring occurs. The epimerase domain of PchE PAO was reported to resemble a methylation domain (22). Would it thus be possible that the methylation domain of PchF CHA0 could be bifunctional and carry out methylation and epimerization? Or could PchK be a reductoisomerase as indicated by its sequence motifs? It is interesting to note that PchK is 36% identical to a Streptomyces coelicolor protein encoded by SCO7684 (59). We suspect that this protein could be involved in the biosynthesis of a compound similar in structure to the pyochelin-related antibiotics thiazostatin and watasemycin from Streptomyces sp. TP-A0597 and Streptomyces tolurosus, respectively (60,61). Indeed, the pchK homolog SCO7684 of S. coelicolor is part of a gene cluster resembling the pch clusters of P. aeruginosa and P. fluorescens in many aspects. However, in contrast to enantio-pyochelin where the thiazoline ring and the thiazolidine ring originate from L-and D-cysteine, respectively, both rings of thiazostatin and watasemycin seem to be derived from L-cysteine such that an epimerase function would not be required by their producer strains. Structural analysis will be necessary to determine if the molecule specified by the pch-related gene cluster of S. coelicolor is a stereoisomer of thiazostatin or watasemycin.
Thiazoline rings with a stereochemistry derived from D-cysteine are also present in the pyochelin-related siderophores yersiniabactin and micacocidin (12,63), and it will be interesting to see if nature has evolved stereoisomers of these compounds as well.
In conclusion, we have shown here that P. fluorescens CHA0 and P. aeruginosa PAO1 make different pyochelin enantiomers that cannot be utilized as an iron source in the heterologous background. This strategy is believed to prevent potential competitors occupying the same ecological niche from stealing heterologous iron-siderophore complexes, a concept which is illustrated best by the immense structural variety of pyoverdines and their cognate receptors made by fluorescent pseudomonads (64,65). In contrast to the pyoverdines where struc-3 D. Cobessi, personal communication. tural differences are generated by variations of the peptide chain, the differences between pyochelin and enantio-pyochelin are of stereochemical nature only while the building blocks, e.g. salicylate and cysteine, are not altered. The importance of chirality for biological activity is well known and has been studied extensively with antimicrobial agents (66,67). We expect that stereochemical variations of siderophores may be more frequent in nature than previously assumed, thus forming an additional layer of defense against siderophore piracy.