J . Med. Microbiol. - Vol. 36 (1992), 184-189
01992 The Pathological Society of Great Britain and Ireland
Isolation of a novel siderophore from Pseudornonas cepacia P. A. SOKOL,
C. J. LEWIS and J. J. DENNIS
Department of Microbiology and Infectious Diseases, University of Calgary Health Sciences Centre, Calgary, Alberta, Canada TZN 4N 1
Summary. A novel iron-binding compound was identified in ethyl acetate extracts of the supernates from Pseudornonas cepacia cultures. This compound, named azurechelin, was produced by 88% of P . cepacia strains isolated from the respiratory tract. Production of azurechelin was regulated by the iron concentration in the culture medium. Azurechelin enhanced the growth of P . cepacia in a medium containing transferrin 200 mg/L. Azurechelin released iron from transferrin in an equilibrium dialysis assay, suggesting that it could compete with transferrin for iron. Azurechelin could also stimulate iron uptake by P . cepacia. This siderophore appeared to have a novel structure with neither the typical characteristics of catechol nor of hydroxamate compounds.
Introduction
Hospital for Sick Children, Toronto, Ontario, Canada and J. D. Klinger, Rainbow Babies’ and Children’s Hospital, Cleveland, Ohio, USA and have been described * Strains Pc275c and Pc7lOm produced both pyochelin and azurechelin. Strains K30-6 and H1724-1 produced only azurechelin and strain H1721 did not produce detectable levels of either siderophore.
Siderophore-mediated iron acquisition has been correlated with the ability of various bacteria to The fact that establish and maintain infection. many bacterial species produce more than one siderophore is also indicative of the importance of iron in bacterial proliferation. Pseudornonas cepacia can cause respiratory infections in cystic fibrosis patients.’ In some cases, these P . cepacia infections have been correlated with a deterioration in clinical status. In a previous study, we reported a correlation between production of the siderophore pyochelin by P . cepacia isolates from cystic fibrosis patients and morbidity and mortality in these patients.6 We have also shown that exogeneously supplied pyochelin can enhance the virulence of pyochelin-non-producing P . cepacia strains in a chronic pulmonary infection model in rats.’ While performing these studies, we identified another ironbinding compound present in ethyl acetate extracts of the culture supernates from several strains. In the present study we describe the production, purification and preliminary characterisation of this iron-binding compound, which we have given the trivial name azurechelin because of its blue fluorescence under ultraviolet (UV) light. 1v
39
Culture conditions For the detection of azurechelin, cultures were grown in 50-ml amounts of Casamino Acids Medium (CAA; Difco) 0.5% which was deferrated as described previously.6 Assays to assess the effects of azurechelin and pyochelin on the growth of P . cepacia were performed in M9 minimal medium supplemented with glucose 0.5%, 10 mM NaHC03 and transferrin (Sigma) 200 mg/L (Transferrin Medium).6Production of azurechelin was also determined in M9 medium with 20 mM sodium succinate or sodium citrate instead of glucose as the carbon s o ~ r c e The . ~ average iron content of the CAA and M9 media was approximately 3-5 p ~All . glassware was washed with acid to remove iron. All reagents were made with water purified by the Milli-Q System (Millipore, Missisauga, Ontario, Canada).
Pyochelin pur fication
Materials and methods Strains
Pyochelin was purified from P . aeruginosa strain P A 0 as described previously.6**
The P . cepacia isolates from the sputum of cystic fibrosis patients were a gift from C. L. Prober, the
Azurechelin purification Azurechelin was purified by essentially the same method as was used for pyochelin. Cultures (500 ml)
-
Received 1 1 March 1991; revised version accepted 28 May 1991. 184
SIDEROPHORE FROM P . CEPACZA
were grown in CAA medium for 24 h at 30°C. The cells were removed by centrifugation at 10 000 g. The supernates were collected, adjusted to pH 2.0 with HCl, and extracted with ethyl acetate (2 vol:5 vol supernate). The extracts were concentrated to dryness by rotary evaporation then dissolved in 1-ml amounts of methanol and chromatographed on a thin layer of ) Scientific) with chloSilica Gel G (250 p ~ (Mandel roform :acetic acid :ethanol 90 : 5 : 2.5 as the mobile phase. The fluorescent blue band visible under UV light was scraped from the plate and eluted into methanol. The methanolic extracts were rechromatographed and developed with chloroform :acetic :ethanol 90 :5 : 5. The azurechelin band (R,=O.SO) was detected by its fluorescence, eluted into methanol, and stored at - 20°C. In some purifications, azurechelin was subjected to gel filtration chromatography on a Sephadex G-10 column (1-5x 30 cm) with aqueous methanol 10%v/v as the mobile phase instead of to a second thin layer chromatography step. For most experiments, azurechelin was used within 2 days of purification. To quantify azurechelin, methanol extracts were filtered through 0 . 2 2 - p ~filters (Millipore) and dried under vacuum into pre-weighed acid-washed vials. Azurechelin was also quantified in culture supernates by measuring the absorbance at 310 nm (the absorbance maximum for azurechelin) and determining the concentration from a standard curve for purified azurechelin.
Assay for azurechelinproduction Strains were grown in 50 ml of CAA medium for 24 h. The supernates were processed according to the purification procedure above, through to the first thinlayer chromatogram. Azurechelin was identified by its relative mobility (R,= 0-8), its blue fluorescence under UV light, and its red-brown reaction when sprayed with 0.1 M FeCl, in 0.1 M HCl. Purified azurechelin was chromatographed as a control.
Chemical characterisation Absorption spectra for purified azurechelin and pyochelin were obtained with a Model 80 spectrophotometer (Pye-Unicam, Cambridge, Cambs.). Siderophore activity was determined by the Chrome Azurol S assay." This assay is based on the ability of a siderophore to remove iron from a dye-containing complex. Purified azurechelin was analysed for catechol structures by the assays of Arnow12 and Rioux et aZ.13 Bound and free hydroxylamine compounds were analysed by the method of Csaky.l4
Growthpromotion assays Growth assays were performed as described previously.6 The cultures were inoculated at an initial density of lo5 cfu/ml in Transferrin Medium. Azure-
185
chelin or pyochelin, dissolved in ethanol, was added to the medium to a final concentration of 10 mg/L. Equivalent volumes of ethanol were added to the control cultures. Growth was monitored by measuring the increase in A600with time.
Equilibrium dialysis assay Release of iron from transferrin was investigated by the method described by Sriyosachati and Cox" with the following modifications. Apotransferrin was prepared by dialysing transferrin (20 g/L ; Sigma) against 1 L of 0 . 2 ~ sodium citrate-sodium acetate buffer, pH 4.5. Portions of the apotransferrin were brought to 100% saturation by the procedure of Simonson et aZ.16 Excess 59Fe was removed by extensive dialysis against 40 mM Tris, 20 mM sodium bicarbonate buffer pH 7.4 until the amount of 59Fein the dialysate remained at a minimal constant level. Reactions were performed in 20 mM morpholine propane sulphonic acid (MOPS), 20 mM pyrophosphate. The pH was adjusted to 5.0,6.0 or 7-4with 1 M HC1. 59Fe-transferrin was added to this buffer to a concentration of 12 mg/L. Azurechelin was added to some reactions at a final concentration of 25 mg/L. Apotransferrin was included at 100 mg/L. This mixture was placed in dialysis tubing (6.2 mm diameter, mol. wt 10 000 exclusion limit) and dialysed against 2ml of the same buffer containing apotransferrin (100 mg/L). Portions of the dialysate were removed at intervals and the amount of 59Fe released into the dialysate was determined by measuring radioactivity in a gamma counter (Compugamma LKB Instruments, Inc. Rockville, MD, USA). The percentage of the available iron released was calculated relative to the initial counts in the reaction mixture.
Iron uptake assays Cultures were grown in M9-glucose medium to a density of lo8 cfu/ml, centrifuged, washed once, and resuspended in fresh medium. Uptake reactions were initiated by the addition of 100 pg (1 pCi) of 59FeC1, (Amersham Corpn., Arlington Heights, IL, USA) and either 1.0 pg of pyochelin or 4 pg of azurechelin in a total volume of 1OOpl. One-ml samples of these reaction mixtures were removed at 10-min intervals and filtered through cellulose acetate 0.2 pm pore filters (Sartorius GmbH, Germany). The amount of "Fe accumulated was determined as previously described6 in a gamma counter (LKB).
Results Production of azurechelin When ethyl acetate extracts of the supernates from cultures of P.cepacia were chromatographed on Silica Gel G plates, two fluorescent bands were visible under
I86
P. A. SOKOL, C. J. LEWIS AND J. J. DENNIS
UV light, both of which turned red-brown when sprayed with 0.1 M FeC1,. A yellow-green fluorescent band at an Rfof 0.4 has previously been reported to be pyocheh6 A blue fluorescent band, which turned brown with FeCl,, was also visible and had an Rf of 0.8. This latter spot also turned brown when sprayed with 0.1 M ferric ammonium sulphate, ferric citrate, and ferric ammonium citrate in 0.1 M HC1. Forty-six (88%) out of 52 respiratory isolates of P . cepacia from patients with cystic fibrosis produced this compound at concentrations detectable by thin layer chromatography. Eight of the 46 positive strains did not produce detectable pyochelin, so this compound could not be a degradation product of pyochelin. We have given this compound the trivial name azurechelin because of its fluorescent blue colour and to prevent confusion with cepabactin, a siderophore produced by some strains of P . cepacia and described by Meyer et a!.l 7 To determine the optimal medium for production of azurechelin, the yields from strains Pc275c and K30-6 were compared in four different media (table I). Yields were highest in CAA medium, which, therefore, was used for subsequent studies. The effect of iron on production of azurechelin was examined for three strains of P . cepacia grown in CAA medium (table 11). Production of azurechelin in strains Pc275c and H1724-1 was highly iron-regulated. In strain Pc7 10, iron concentrationsof 50-1 00 C(Mwere required to repress azurechelin expression. Thus, although expression of this compound was iron-regulated,there was some inter-strain variation in the effect of iron on its yield.
Table I. Comparison of azurechelin production by P . cepacia strains in different culture media Azurechelin (mg/L) from strain
Medium
Pc275c
K30-6
5.9 1.1 1.9 0.3
4.6 1.0 1.4 0
CAA M9 +glucose 0.5% M9 + 20 mM succinate M9 + 20 mM citrate
Table 11. Effect of iron concentration on production of azurechelin by P . cepacia ~
~~
~
Azurechelin (mg/L)/ ODeooUnit at FeCI, concentrations*
Strain no.
1
Pc275c H1724-1 Pc7 10m
10.8 6.8 3.8
3.4 3.4 3.6
1.2 3.2 3.2
1.1 0.7 2.4
1.0 0-2 0.7
*Final concentration of FeC1, added to the CAA medium.
iron-free azurechelin dissolved in methanol. The absorption spectrum for iron-free pyochelin was obtained for comparative purposes. Azurechelin had absorbance maxima at 240 and 310 nm (fig. 1). By contrast, pyochelin gave peaks at 210,250 and 310 nm (fig. 1 p 9 The Chrome Azurol S (CAS) assay is a universal assay for siderophores, exploiting their ability to release iron from the Chrome Azurol S/iron/hexadeyltrimethyl-ammonium bromide complex. * Purified azurechelin removed iron from the dye complex in the CAS assay and changed the colour of the solution
Characterisation of azurechelin Azurechelin was purified by preparative thin-layer chromatography on Silica Gel G or by gel filtration on Sephadex G-10. UV-visible spectra were obtained for
3-51 3.02.52.0 1-5-
1.00.5y
ih
I
I
. L
..
A-
q
0 2i 00 0 2 2 0 2 4 0 2 6 0 2 8 0 3 0 0 3 2 0 3 4 0 3 6 0 3 8 0 4 0 0 4 2 0 Y
1
I
I
I
I
I
I
Wavelength (nm)
a)
Fig. 1. Comparison of absorption spectra for purified azurechelin and pyochelin ( 0 ) 50 ; pg of azurechelin or pyochelin were dissolved in methanol and scanned from 190 to 800 nm at 10-nm intervals. Absorbance between 190 and 400 nrn is shown; the values remained at a background level between 400 and 800 nm.
SIDEROPHORE FROM P. CEPACIA
from blue to orange. Culture supernates quickly changed the colour of the CAS solution, but purified azurechelin required the addition of 5-sulphosalicylic acid, which acts as a shuttle between the dye complex and the siderophore. * When TLC plates were sprayed with Chrome Azurol S solution, the azurechelincontaining spots turned pink, as has been reported for other siderophores, including aerobactin, enterochelin and rhizobactin. * Most siderophores are either of the catechol or hydroxamate class. However, azurechelin gave negative results in assays for the detection of catecholcontaining siderophores. l4 It did not react, either, with a “phenolate detection” spray (50 mM FeC13, 50 mM potassium ferricyanide,l o which normally turns blue when sprayed on a TLC plate containing phenolic compounds. Although not expected to be a hydroxamate siderophore, since it was extractable with ethyl acetate, azurechelin was tested for hydroxylamine by the Csaky assay,I5 with negative results. Azurechelin appeared unstable. It lost iron-binding activity after a few days when stored in a dried form. When stored in ethanol at - 20°C, the iron-binding activity decreased by approximately 25% within 7 days. 1 3 9
Eflects of azurechelin on growth and iron transport Previously, we have shown that pyochelin can enhance the growth rate of P . cepacia strains in a medium containing transferrin.6 To determine if azurechelin also could increase the growth rate of P . cepacia, the effects of azurechelin and pyochelin on the growth of three strains of P . cepacia were determined. Strain Pc275c produced both pyochelin
lo[
t
A
187
and azurechelin, whereas strains K45- 1 and H 1721 did not produce detectable levels of either siderophore. Cultures were grown for 24 h in M9 medium with 200 PM ethylene diamine-N-N’-diacetic acid to starve the bacteria for iron, then diluted into Transferrin Medium at an initial density of 104-105 cfu/ml. The cultures were divided into three flasks and pyochelin (10 mg/L) or azurechelin (10 mg/L) was added to two of the cultures. Growth was monitored by measuring A600 at intervals for up to 4 days (fig. 2). Both pyochelin and azurechelin dramatically increased the growth rates of the siderophore-negative strains, K45-I and H1721, in the Transferrin Medium. Although the effect on strain Pc275c was not as dramatic, both pyochelin and azurechelin shortened the lag phase for this organism. Pyochelin increased the growth rates of strains Pc275 and H1721 to a greater extent than did azurechelin (fig. 2B, C). These data suggest that azurechelin may be effective in removing iron from transferrin. To investigate this aspect further, the ability of azurechelin to remove iron from transferrin was examined in an equilibrium dialysis assay (table 111). Removal of iron from transferrin was determined at pH 7.4, 6.0 and 5.0. Azurechelin was effective in releasing iron at both pH 6.0 and 5.0 but not at pH 7.4. At pH 6.0, equilibrium was reached within 24 h. These data suggest that azurechelin can compete with transferrin for iron. To confirm that P . cepacia strains could use azure chelin to transport iron, their ability to accumulate ferri-azurechechelin was determined and the uptakes of ferri-azurechelin and ferri-pyochelin were compared (fig. 3). Strains Pc275c and H1721 were grown in M9 minimal medium, in which pyochelin production is not detectable and azurechelin production is
’9 I
E C
Fig. 2. Effects of azurechelin and pyochelin on growth of P. cepacia strains in M9 medium containing transferrin 200 mg/L: A, strain K45-1; B, strain H1721;C, strain Pc275c; a,no siderophore added; 0 ,azurechelin 10 mg/L added; m, pyochelin 10 mg/L added.
188
P. A. SOKOL, C. J. LEWIS A N D J. J. DENNIS
Table 111. Iron-binding activity of azurechelin in equilibrium dialysis Percentage of 59Fereleased from 59[Fe]-transferrin Buffer pH
7.4 6.0 5.0 6.0 6-0 6.0
Time (h)
24 24 24 2 24 120
without azurechelin
with azurechelin
0 0 0 1-5 5.6 7.5
0 56.8 50.5
5.3 44.6 49.7
59[Fe]-labelledtransferrin was prepared by the method of Simonson et The assay consisted of 20 pg of azurechelin on one side of the dialysis membrane and s9[Fe]-transferrin (6 mg/L, 100% saturated) on the opposite side. Apotransferrin (100 mg/L) was included on both sides with 100 PM MOPS buffer, 20 m M sodium bicarbonate and 10 mM pyophosphate as the diluent; c. 56 000 cpm of 59Fewere available for equilibrium in the first experiment; 34 500 cpm in the second.
30
-
at 210, 250 and 310nm, whereas those of pyochelin are at 240 and 310 nm and those of cepabactin are 25 reported to be 330 and 440 nm.” Moreover azurechelin reacted only with FeC13; whereas pyochelin n 20and cepabactin react also with the phenolate detection spray reagent. Azurechelin promoted the growth of P. cepacia strains in a medium containing transferrin. This effect was most dramatic in the two siderophore-negative strains examined (K45-1 and H1721; fig. 2). One strain, K45-1, was able to grow in the iron-limited 0 2 4 6 8 1 0 1 2 1 4 1 6 medium without the addition of azurechelin or pyochelin although the lag phase was quite long. This Time (min) suggests that there may be an alternative iron Fig. 3. Uptake of s9Fe by P.cepueiu strains H1721 ( 0 , O )and acquisition mechanism in this strain. This strain may Pc275c (W, 0). Uptake assays were initiated by the addition of produce the siderophore cepabactin, although this either s9[Fe]-azurechelin(closed symbols) or s9[Fe]-pyochelin(open possibility was not investigated. Another possibility is symbols); I-ml samples were removed at intervals and the amount of s9Feaccumulated was determined. that this strain produces a protease which we have shown degrades transferrin (Sokol and Lewis, unpublished observations).This degradation may relieve the iron limitation. very low (table I). When the cultures reached a density In an equilibrium dialysis assay, azurechelin reof lo8 cfu/ml, the cells were washed and either 59[Fe]moved radiolabelled iron from transferrin at pH 6.0 pyochelin or 59[Fe]-azurechelin was added to the and 5.0, though not at pH 7-4. Sriyosachati and cultures and the amount of 59Feaccumulated by the Cox’ previously reported similar pH-dependent iron cells was determined at selected intervals, Both strains accumulated 59[Fe]-pyochelinand 59[Fe]-azurechelin. sequestration results for pyochelin and pyoverdin from P. aeruginosa. Therefore, azurechelin appears similar Strain H 1721, which did not produce detectable levels of either siderophores, initially bound more 59[Fe]- to these other Pseudomonas siderophores in that it is more effective at acidic pH. Acid pH levels have been siderophore than did strain Pc275c, which produced reported in inflammatory exudates. 5 9 These siderboth siderophores. However, rates of uptake were ophores may be most effective at obtaining iron from very similar for both strains. transferrin in sites of inflammation. Both azurechelin and pyochelin promoted iron uptake by P. cepacia. Strain H1721, which did not Discussion produce detectable levels of siderophores, bound more 59[Fe]-azurechelin and 59[Fe]-pyochelin than did We investigated the characteristics of a novel strain Pc275c, which produced both siderophores. It siderophore produced by respiratory isolates of P . is possible that strain H1721 may produce larger cepacia. This compound, given the trivial name amounts of the receptors for these siderophores since azurechelin, had properties common to other sideroit would be deficient at acquiring iron as compared to phores. In particular, production of azurechelin was siderophore-positive strains. This has not yet been regulated by the iron concentration of the culture investigated. Although the rates of pyochelin and medium. Azurechelin was distinct from the other azurechelin uptake appeared similar for these strains, siderophoresof P . cepacia, having absorbance maxima
SIDEROPHORE FROM P . CEPACIA
189
direct comparison is difficult because there may be a difference in expression of the respective siderophore receptors in the M9 medium. There are three general classes of siderophores, hydroxamates, catechols and miscellaneous.1 9 7 2o Azurechelin does not have the characteristics of either the catechol or hydroxamate siderophores, giving negative results in the standard assays for these compounds. ' 3-1 Catechol siderophores usually have three absorption maxima, at 320nm, 250nm and 210 nm, with the most prominent occurring at 210 nrn.l9 However, azurechelin had two peaks of approximately equal prominence around 240 nm and 3 10 nm. Until its structure is determined, azurechelin must be placed in the miscellaneous class of siderophores described by Neilands. l 9 Rhizobiurn rneliloti also produces a siderophore, rhizobactin, which has been characterised structurally and which is neither a catechol nor a hydroxymate.*l Siderophores in the miscellaneous class can be assayed only by functional
or biological techniques' such as the Chrome Azurol S assay, which was used here. Azurechelin required a shuttle reagent, 5-sulphosalicylic acid, to accelerate the reaction. Similar properties are reported for rhizobactin.' * It is interesting that strains of P . cepacia express at least three siderophore-mediated iron-transport systems, pyochelin, cepabactin, and azurechelin. Although previous studies have implicated pyochelin as a virulence factor in P . cepacia respiratory infections,6T the contribution of azurechelin and cepabactin to the pathogenesis of P. cepacia infections has not yet been determined. It would be of interest to compare mutants deficient in each siderophore systems to determine their importance to iron acquisition and pathogenicity .
References
12. Arnow LE. Colorimetric determination of the components of 3,4-dihydroxyphenylalanine-tyrosinemixtures. J Biol Chem 1937; 118: 531-537. 13. Rioux C, Jordan DC, Rattray JBM. Colorimetric determination of catechol siderophores in microbial cultures. Anal Biochem 1983; 133: 163-169. 14. Csiky 2. On the estimation of bound hydroxylamine in biological materials. Acta Chem Scand 1948; 2 : 45M54. 15. Sriyosachati S, Cox CD. Siderophore-mediated iron acquisition from transferrin by Pseudomonas aeruginosa. Infect Immun 1986; 52: 885-891. 16. Simonson C, Brener D, DeVoe IW. Expression of a highaffinity mechanism for acquisition for transferrin iron by Neisseria meningitidis. Infect Immun 1982; 36: 107-1 13. 17. Meyer J-M, Hohnadel, D, HallC F. Cepabactin from Pseudomonas cepacia, a new type of siderophore. J Gen Microbiol 1989; 135: 1479-1487. 18. Menkin V. The role of the hydrogen ion concentration and the cytology of an exudate. Glycolysis in inflammation. Some aspects of the chemistry of exudates. In: Ryan EJ (ed) Biochemical mechanisms in inflammation. Springfield, IL, Charles C Thomas. 1956: 66-103. 19. Neilands JB. Methodology of siderophores. Structure and Bonding 1984; 58 : 1-24. 20. Neilands JB. Microbial iron compounds. Annu Rev Biochem 1981; 50: 715-731. 21. Smith MJ, Shoolery JN, Schwyn B, Holden I, Neilands JB. Rhizobactin, a structurally novel siderophore from Rhizobium meliloti. J A m Chem Soc 1985; 107: 1739-1743.
1. Payne SM, Finkelstein RA. The critical role of iron in hostbacterial interactions. J Clin Invest 1978; 61: 1428-1440. 2. Crosa JH. The relationship of plasmid-mediated iron transport and bacterial virulence. Annu Rev Microbiol 1984; 38: 6989. 3. Neilands JB. Microbial envelope proteins related to iron. Annu Rev Microbioll982; 36: 285-309. 4. Barclay R. The role of iron in infection. Med Lab Sci 1985; 42: 166177. 5. Isles A, Maclusky I, Corey R et al. Pseudomonas cepacia infection in cystic fibrosis: an emerging problem. J Pediatr 1984; 104: 206210. 6. Sokol PA. Production and utilization of pyochelin by clinical isolates of Pseudomonas cepacia. J Clin Microbioll986; 23: 560-562. 7. Sokol PA, Woods DE. Effect of pyochelin on Pseudomonas cepacia respiratory infections. Microb Pathog 1988;5 : 197205. 8. McKevitt AI, Woods DE. Characterization of Pseudomonas cepacia isolates from patients with cystic fibrosis. J Clin Microbioll984; 19: 291-293. 9. Cox CD, Graham R. Isolation of an iron-binding compound from Pseudomonas aeruginosa. J Bacterioll979; 137: 357364. 10. Cox CD. Iron uptake with ferripyochelin and ferric citrate by Pseudomonas aeruginosa. J Bacterioll980; 142 : 581-587. 11. Schwyn B, Neilands JB. Universal chemical assay for the detection and determination of siderophores. Anal Biochem 1987; 160: 47-56.
'
These studies were supported by the Canadian Cystic Fibrosis Foundation. P.A.S is an Alberta Heritage Foundation for Medical Research Scholar.
01992 The Pathological Society of Great Britain and Ireland
Isolation of a novel siderophore from Pseudornonas cepacia P. A. SOKOL,
C. J. LEWIS and J. J. DENNIS
Department of Microbiology and Infectious Diseases, University of Calgary Health Sciences Centre, Calgary, Alberta, Canada TZN 4N 1
Summary. A novel iron-binding compound was identified in ethyl acetate extracts of the supernates from Pseudornonas cepacia cultures. This compound, named azurechelin, was produced by 88% of P . cepacia strains isolated from the respiratory tract. Production of azurechelin was regulated by the iron concentration in the culture medium. Azurechelin enhanced the growth of P . cepacia in a medium containing transferrin 200 mg/L. Azurechelin released iron from transferrin in an equilibrium dialysis assay, suggesting that it could compete with transferrin for iron. Azurechelin could also stimulate iron uptake by P . cepacia. This siderophore appeared to have a novel structure with neither the typical characteristics of catechol nor of hydroxamate compounds.
Introduction
Hospital for Sick Children, Toronto, Ontario, Canada and J. D. Klinger, Rainbow Babies’ and Children’s Hospital, Cleveland, Ohio, USA and have been described * Strains Pc275c and Pc7lOm produced both pyochelin and azurechelin. Strains K30-6 and H1724-1 produced only azurechelin and strain H1721 did not produce detectable levels of either siderophore.
Siderophore-mediated iron acquisition has been correlated with the ability of various bacteria to The fact that establish and maintain infection. many bacterial species produce more than one siderophore is also indicative of the importance of iron in bacterial proliferation. Pseudornonas cepacia can cause respiratory infections in cystic fibrosis patients.’ In some cases, these P . cepacia infections have been correlated with a deterioration in clinical status. In a previous study, we reported a correlation between production of the siderophore pyochelin by P . cepacia isolates from cystic fibrosis patients and morbidity and mortality in these patients.6 We have also shown that exogeneously supplied pyochelin can enhance the virulence of pyochelin-non-producing P . cepacia strains in a chronic pulmonary infection model in rats.’ While performing these studies, we identified another ironbinding compound present in ethyl acetate extracts of the culture supernates from several strains. In the present study we describe the production, purification and preliminary characterisation of this iron-binding compound, which we have given the trivial name azurechelin because of its blue fluorescence under ultraviolet (UV) light. 1v
39
Culture conditions For the detection of azurechelin, cultures were grown in 50-ml amounts of Casamino Acids Medium (CAA; Difco) 0.5% which was deferrated as described previously.6 Assays to assess the effects of azurechelin and pyochelin on the growth of P . cepacia were performed in M9 minimal medium supplemented with glucose 0.5%, 10 mM NaHC03 and transferrin (Sigma) 200 mg/L (Transferrin Medium).6Production of azurechelin was also determined in M9 medium with 20 mM sodium succinate or sodium citrate instead of glucose as the carbon s o ~ r c e The . ~ average iron content of the CAA and M9 media was approximately 3-5 p ~All . glassware was washed with acid to remove iron. All reagents were made with water purified by the Milli-Q System (Millipore, Missisauga, Ontario, Canada).
Pyochelin pur fication
Materials and methods Strains
Pyochelin was purified from P . aeruginosa strain P A 0 as described previously.6**
The P . cepacia isolates from the sputum of cystic fibrosis patients were a gift from C. L. Prober, the
Azurechelin purification Azurechelin was purified by essentially the same method as was used for pyochelin. Cultures (500 ml)
-
Received 1 1 March 1991; revised version accepted 28 May 1991. 184
SIDEROPHORE FROM P . CEPACZA
were grown in CAA medium for 24 h at 30°C. The cells were removed by centrifugation at 10 000 g. The supernates were collected, adjusted to pH 2.0 with HCl, and extracted with ethyl acetate (2 vol:5 vol supernate). The extracts were concentrated to dryness by rotary evaporation then dissolved in 1-ml amounts of methanol and chromatographed on a thin layer of ) Scientific) with chloSilica Gel G (250 p ~ (Mandel roform :acetic acid :ethanol 90 : 5 : 2.5 as the mobile phase. The fluorescent blue band visible under UV light was scraped from the plate and eluted into methanol. The methanolic extracts were rechromatographed and developed with chloroform :acetic :ethanol 90 :5 : 5. The azurechelin band (R,=O.SO) was detected by its fluorescence, eluted into methanol, and stored at - 20°C. In some purifications, azurechelin was subjected to gel filtration chromatography on a Sephadex G-10 column (1-5x 30 cm) with aqueous methanol 10%v/v as the mobile phase instead of to a second thin layer chromatography step. For most experiments, azurechelin was used within 2 days of purification. To quantify azurechelin, methanol extracts were filtered through 0 . 2 2 - p ~filters (Millipore) and dried under vacuum into pre-weighed acid-washed vials. Azurechelin was also quantified in culture supernates by measuring the absorbance at 310 nm (the absorbance maximum for azurechelin) and determining the concentration from a standard curve for purified azurechelin.
Assay for azurechelinproduction Strains were grown in 50 ml of CAA medium for 24 h. The supernates were processed according to the purification procedure above, through to the first thinlayer chromatogram. Azurechelin was identified by its relative mobility (R,= 0-8), its blue fluorescence under UV light, and its red-brown reaction when sprayed with 0.1 M FeCl, in 0.1 M HCl. Purified azurechelin was chromatographed as a control.
Chemical characterisation Absorption spectra for purified azurechelin and pyochelin were obtained with a Model 80 spectrophotometer (Pye-Unicam, Cambridge, Cambs.). Siderophore activity was determined by the Chrome Azurol S assay." This assay is based on the ability of a siderophore to remove iron from a dye-containing complex. Purified azurechelin was analysed for catechol structures by the assays of Arnow12 and Rioux et aZ.13 Bound and free hydroxylamine compounds were analysed by the method of Csaky.l4
Growthpromotion assays Growth assays were performed as described previously.6 The cultures were inoculated at an initial density of lo5 cfu/ml in Transferrin Medium. Azure-
185
chelin or pyochelin, dissolved in ethanol, was added to the medium to a final concentration of 10 mg/L. Equivalent volumes of ethanol were added to the control cultures. Growth was monitored by measuring the increase in A600with time.
Equilibrium dialysis assay Release of iron from transferrin was investigated by the method described by Sriyosachati and Cox" with the following modifications. Apotransferrin was prepared by dialysing transferrin (20 g/L ; Sigma) against 1 L of 0 . 2 ~ sodium citrate-sodium acetate buffer, pH 4.5. Portions of the apotransferrin were brought to 100% saturation by the procedure of Simonson et aZ.16 Excess 59Fe was removed by extensive dialysis against 40 mM Tris, 20 mM sodium bicarbonate buffer pH 7.4 until the amount of 59Fein the dialysate remained at a minimal constant level. Reactions were performed in 20 mM morpholine propane sulphonic acid (MOPS), 20 mM pyrophosphate. The pH was adjusted to 5.0,6.0 or 7-4with 1 M HC1. 59Fe-transferrin was added to this buffer to a concentration of 12 mg/L. Azurechelin was added to some reactions at a final concentration of 25 mg/L. Apotransferrin was included at 100 mg/L. This mixture was placed in dialysis tubing (6.2 mm diameter, mol. wt 10 000 exclusion limit) and dialysed against 2ml of the same buffer containing apotransferrin (100 mg/L). Portions of the dialysate were removed at intervals and the amount of 59Fe released into the dialysate was determined by measuring radioactivity in a gamma counter (Compugamma LKB Instruments, Inc. Rockville, MD, USA). The percentage of the available iron released was calculated relative to the initial counts in the reaction mixture.
Iron uptake assays Cultures were grown in M9-glucose medium to a density of lo8 cfu/ml, centrifuged, washed once, and resuspended in fresh medium. Uptake reactions were initiated by the addition of 100 pg (1 pCi) of 59FeC1, (Amersham Corpn., Arlington Heights, IL, USA) and either 1.0 pg of pyochelin or 4 pg of azurechelin in a total volume of 1OOpl. One-ml samples of these reaction mixtures were removed at 10-min intervals and filtered through cellulose acetate 0.2 pm pore filters (Sartorius GmbH, Germany). The amount of "Fe accumulated was determined as previously described6 in a gamma counter (LKB).
Results Production of azurechelin When ethyl acetate extracts of the supernates from cultures of P.cepacia were chromatographed on Silica Gel G plates, two fluorescent bands were visible under
I86
P. A. SOKOL, C. J. LEWIS AND J. J. DENNIS
UV light, both of which turned red-brown when sprayed with 0.1 M FeC1,. A yellow-green fluorescent band at an Rfof 0.4 has previously been reported to be pyocheh6 A blue fluorescent band, which turned brown with FeCl,, was also visible and had an Rf of 0.8. This latter spot also turned brown when sprayed with 0.1 M ferric ammonium sulphate, ferric citrate, and ferric ammonium citrate in 0.1 M HC1. Forty-six (88%) out of 52 respiratory isolates of P . cepacia from patients with cystic fibrosis produced this compound at concentrations detectable by thin layer chromatography. Eight of the 46 positive strains did not produce detectable pyochelin, so this compound could not be a degradation product of pyochelin. We have given this compound the trivial name azurechelin because of its fluorescent blue colour and to prevent confusion with cepabactin, a siderophore produced by some strains of P . cepacia and described by Meyer et a!.l 7 To determine the optimal medium for production of azurechelin, the yields from strains Pc275c and K30-6 were compared in four different media (table I). Yields were highest in CAA medium, which, therefore, was used for subsequent studies. The effect of iron on production of azurechelin was examined for three strains of P . cepacia grown in CAA medium (table 11). Production of azurechelin in strains Pc275c and H1724-1 was highly iron-regulated. In strain Pc7 10, iron concentrationsof 50-1 00 C(Mwere required to repress azurechelin expression. Thus, although expression of this compound was iron-regulated,there was some inter-strain variation in the effect of iron on its yield.
Table I. Comparison of azurechelin production by P . cepacia strains in different culture media Azurechelin (mg/L) from strain
Medium
Pc275c
K30-6
5.9 1.1 1.9 0.3
4.6 1.0 1.4 0
CAA M9 +glucose 0.5% M9 + 20 mM succinate M9 + 20 mM citrate
Table 11. Effect of iron concentration on production of azurechelin by P . cepacia ~
~~
~
Azurechelin (mg/L)/ ODeooUnit at FeCI, concentrations*
Strain no.
1
Pc275c H1724-1 Pc7 10m
10.8 6.8 3.8
3.4 3.4 3.6
1.2 3.2 3.2
1.1 0.7 2.4
1.0 0-2 0.7
*Final concentration of FeC1, added to the CAA medium.
iron-free azurechelin dissolved in methanol. The absorption spectrum for iron-free pyochelin was obtained for comparative purposes. Azurechelin had absorbance maxima at 240 and 310 nm (fig. 1). By contrast, pyochelin gave peaks at 210,250 and 310 nm (fig. 1 p 9 The Chrome Azurol S (CAS) assay is a universal assay for siderophores, exploiting their ability to release iron from the Chrome Azurol S/iron/hexadeyltrimethyl-ammonium bromide complex. * Purified azurechelin removed iron from the dye complex in the CAS assay and changed the colour of the solution
Characterisation of azurechelin Azurechelin was purified by preparative thin-layer chromatography on Silica Gel G or by gel filtration on Sephadex G-10. UV-visible spectra were obtained for
3-51 3.02.52.0 1-5-
1.00.5y
ih
I
I
. L
..
A-
q
0 2i 00 0 2 2 0 2 4 0 2 6 0 2 8 0 3 0 0 3 2 0 3 4 0 3 6 0 3 8 0 4 0 0 4 2 0 Y
1
I
I
I
I
I
I
Wavelength (nm)
a)
Fig. 1. Comparison of absorption spectra for purified azurechelin and pyochelin ( 0 ) 50 ; pg of azurechelin or pyochelin were dissolved in methanol and scanned from 190 to 800 nm at 10-nm intervals. Absorbance between 190 and 400 nrn is shown; the values remained at a background level between 400 and 800 nm.
SIDEROPHORE FROM P. CEPACIA
from blue to orange. Culture supernates quickly changed the colour of the CAS solution, but purified azurechelin required the addition of 5-sulphosalicylic acid, which acts as a shuttle between the dye complex and the siderophore. * When TLC plates were sprayed with Chrome Azurol S solution, the azurechelincontaining spots turned pink, as has been reported for other siderophores, including aerobactin, enterochelin and rhizobactin. * Most siderophores are either of the catechol or hydroxamate class. However, azurechelin gave negative results in assays for the detection of catecholcontaining siderophores. l4 It did not react, either, with a “phenolate detection” spray (50 mM FeC13, 50 mM potassium ferricyanide,l o which normally turns blue when sprayed on a TLC plate containing phenolic compounds. Although not expected to be a hydroxamate siderophore, since it was extractable with ethyl acetate, azurechelin was tested for hydroxylamine by the Csaky assay,I5 with negative results. Azurechelin appeared unstable. It lost iron-binding activity after a few days when stored in a dried form. When stored in ethanol at - 20°C, the iron-binding activity decreased by approximately 25% within 7 days. 1 3 9
Eflects of azurechelin on growth and iron transport Previously, we have shown that pyochelin can enhance the growth rate of P . cepacia strains in a medium containing transferrin.6 To determine if azurechelin also could increase the growth rate of P . cepacia, the effects of azurechelin and pyochelin on the growth of three strains of P . cepacia were determined. Strain Pc275c produced both pyochelin
lo[
t
A
187
and azurechelin, whereas strains K45- 1 and H 1721 did not produce detectable levels of either siderophore. Cultures were grown for 24 h in M9 medium with 200 PM ethylene diamine-N-N’-diacetic acid to starve the bacteria for iron, then diluted into Transferrin Medium at an initial density of 104-105 cfu/ml. The cultures were divided into three flasks and pyochelin (10 mg/L) or azurechelin (10 mg/L) was added to two of the cultures. Growth was monitored by measuring A600 at intervals for up to 4 days (fig. 2). Both pyochelin and azurechelin dramatically increased the growth rates of the siderophore-negative strains, K45-I and H1721, in the Transferrin Medium. Although the effect on strain Pc275c was not as dramatic, both pyochelin and azurechelin shortened the lag phase for this organism. Pyochelin increased the growth rates of strains Pc275 and H1721 to a greater extent than did azurechelin (fig. 2B, C). These data suggest that azurechelin may be effective in removing iron from transferrin. To investigate this aspect further, the ability of azurechelin to remove iron from transferrin was examined in an equilibrium dialysis assay (table 111). Removal of iron from transferrin was determined at pH 7.4, 6.0 and 5.0. Azurechelin was effective in releasing iron at both pH 6.0 and 5.0 but not at pH 7.4. At pH 6.0, equilibrium was reached within 24 h. These data suggest that azurechelin can compete with transferrin for iron. To confirm that P . cepacia strains could use azure chelin to transport iron, their ability to accumulate ferri-azurechechelin was determined and the uptakes of ferri-azurechelin and ferri-pyochelin were compared (fig. 3). Strains Pc275c and H1721 were grown in M9 minimal medium, in which pyochelin production is not detectable and azurechelin production is
’9 I
E C
Fig. 2. Effects of azurechelin and pyochelin on growth of P. cepacia strains in M9 medium containing transferrin 200 mg/L: A, strain K45-1; B, strain H1721;C, strain Pc275c; a,no siderophore added; 0 ,azurechelin 10 mg/L added; m, pyochelin 10 mg/L added.
188
P. A. SOKOL, C. J. LEWIS A N D J. J. DENNIS
Table 111. Iron-binding activity of azurechelin in equilibrium dialysis Percentage of 59Fereleased from 59[Fe]-transferrin Buffer pH
7.4 6.0 5.0 6.0 6-0 6.0
Time (h)
24 24 24 2 24 120
without azurechelin
with azurechelin
0 0 0 1-5 5.6 7.5
0 56.8 50.5
5.3 44.6 49.7
59[Fe]-labelledtransferrin was prepared by the method of Simonson et The assay consisted of 20 pg of azurechelin on one side of the dialysis membrane and s9[Fe]-transferrin (6 mg/L, 100% saturated) on the opposite side. Apotransferrin (100 mg/L) was included on both sides with 100 PM MOPS buffer, 20 m M sodium bicarbonate and 10 mM pyophosphate as the diluent; c. 56 000 cpm of 59Fewere available for equilibrium in the first experiment; 34 500 cpm in the second.
30
-
at 210, 250 and 310nm, whereas those of pyochelin are at 240 and 310 nm and those of cepabactin are 25 reported to be 330 and 440 nm.” Moreover azurechelin reacted only with FeC13; whereas pyochelin n 20and cepabactin react also with the phenolate detection spray reagent. Azurechelin promoted the growth of P. cepacia strains in a medium containing transferrin. This effect was most dramatic in the two siderophore-negative strains examined (K45-1 and H1721; fig. 2). One strain, K45-1, was able to grow in the iron-limited 0 2 4 6 8 1 0 1 2 1 4 1 6 medium without the addition of azurechelin or pyochelin although the lag phase was quite long. This Time (min) suggests that there may be an alternative iron Fig. 3. Uptake of s9Fe by P.cepueiu strains H1721 ( 0 , O )and acquisition mechanism in this strain. This strain may Pc275c (W, 0). Uptake assays were initiated by the addition of produce the siderophore cepabactin, although this either s9[Fe]-azurechelin(closed symbols) or s9[Fe]-pyochelin(open possibility was not investigated. Another possibility is symbols); I-ml samples were removed at intervals and the amount of s9Feaccumulated was determined. that this strain produces a protease which we have shown degrades transferrin (Sokol and Lewis, unpublished observations).This degradation may relieve the iron limitation. very low (table I). When the cultures reached a density In an equilibrium dialysis assay, azurechelin reof lo8 cfu/ml, the cells were washed and either 59[Fe]moved radiolabelled iron from transferrin at pH 6.0 pyochelin or 59[Fe]-azurechelin was added to the and 5.0, though not at pH 7-4. Sriyosachati and cultures and the amount of 59Feaccumulated by the Cox’ previously reported similar pH-dependent iron cells was determined at selected intervals, Both strains accumulated 59[Fe]-pyochelinand 59[Fe]-azurechelin. sequestration results for pyochelin and pyoverdin from P. aeruginosa. Therefore, azurechelin appears similar Strain H 1721, which did not produce detectable levels of either siderophores, initially bound more 59[Fe]- to these other Pseudomonas siderophores in that it is more effective at acidic pH. Acid pH levels have been siderophore than did strain Pc275c, which produced reported in inflammatory exudates. 5 9 These siderboth siderophores. However, rates of uptake were ophores may be most effective at obtaining iron from very similar for both strains. transferrin in sites of inflammation. Both azurechelin and pyochelin promoted iron uptake by P. cepacia. Strain H1721, which did not Discussion produce detectable levels of siderophores, bound more 59[Fe]-azurechelin and 59[Fe]-pyochelin than did We investigated the characteristics of a novel strain Pc275c, which produced both siderophores. It siderophore produced by respiratory isolates of P . is possible that strain H1721 may produce larger cepacia. This compound, given the trivial name amounts of the receptors for these siderophores since azurechelin, had properties common to other sideroit would be deficient at acquiring iron as compared to phores. In particular, production of azurechelin was siderophore-positive strains. This has not yet been regulated by the iron concentration of the culture investigated. Although the rates of pyochelin and medium. Azurechelin was distinct from the other azurechelin uptake appeared similar for these strains, siderophoresof P . cepacia, having absorbance maxima
SIDEROPHORE FROM P . CEPACIA
189
direct comparison is difficult because there may be a difference in expression of the respective siderophore receptors in the M9 medium. There are three general classes of siderophores, hydroxamates, catechols and miscellaneous.1 9 7 2o Azurechelin does not have the characteristics of either the catechol or hydroxamate siderophores, giving negative results in the standard assays for these compounds. ' 3-1 Catechol siderophores usually have three absorption maxima, at 320nm, 250nm and 210 nm, with the most prominent occurring at 210 nrn.l9 However, azurechelin had two peaks of approximately equal prominence around 240 nm and 3 10 nm. Until its structure is determined, azurechelin must be placed in the miscellaneous class of siderophores described by Neilands. l 9 Rhizobiurn rneliloti also produces a siderophore, rhizobactin, which has been characterised structurally and which is neither a catechol nor a hydroxymate.*l Siderophores in the miscellaneous class can be assayed only by functional
or biological techniques' such as the Chrome Azurol S assay, which was used here. Azurechelin required a shuttle reagent, 5-sulphosalicylic acid, to accelerate the reaction. Similar properties are reported for rhizobactin.' * It is interesting that strains of P . cepacia express at least three siderophore-mediated iron-transport systems, pyochelin, cepabactin, and azurechelin. Although previous studies have implicated pyochelin as a virulence factor in P . cepacia respiratory infections,6T the contribution of azurechelin and cepabactin to the pathogenesis of P. cepacia infections has not yet been determined. It would be of interest to compare mutants deficient in each siderophore systems to determine their importance to iron acquisition and pathogenicity .
References
12. Arnow LE. Colorimetric determination of the components of 3,4-dihydroxyphenylalanine-tyrosinemixtures. J Biol Chem 1937; 118: 531-537. 13. Rioux C, Jordan DC, Rattray JBM. Colorimetric determination of catechol siderophores in microbial cultures. Anal Biochem 1983; 133: 163-169. 14. Csiky 2. On the estimation of bound hydroxylamine in biological materials. Acta Chem Scand 1948; 2 : 45M54. 15. Sriyosachati S, Cox CD. Siderophore-mediated iron acquisition from transferrin by Pseudomonas aeruginosa. Infect Immun 1986; 52: 885-891. 16. Simonson C, Brener D, DeVoe IW. Expression of a highaffinity mechanism for acquisition for transferrin iron by Neisseria meningitidis. Infect Immun 1982; 36: 107-1 13. 17. Meyer J-M, Hohnadel, D, HallC F. Cepabactin from Pseudomonas cepacia, a new type of siderophore. J Gen Microbiol 1989; 135: 1479-1487. 18. Menkin V. The role of the hydrogen ion concentration and the cytology of an exudate. Glycolysis in inflammation. Some aspects of the chemistry of exudates. In: Ryan EJ (ed) Biochemical mechanisms in inflammation. Springfield, IL, Charles C Thomas. 1956: 66-103. 19. Neilands JB. Methodology of siderophores. Structure and Bonding 1984; 58 : 1-24. 20. Neilands JB. Microbial iron compounds. Annu Rev Biochem 1981; 50: 715-731. 21. Smith MJ, Shoolery JN, Schwyn B, Holden I, Neilands JB. Rhizobactin, a structurally novel siderophore from Rhizobium meliloti. J A m Chem Soc 1985; 107: 1739-1743.
1. Payne SM, Finkelstein RA. The critical role of iron in hostbacterial interactions. J Clin Invest 1978; 61: 1428-1440. 2. Crosa JH. The relationship of plasmid-mediated iron transport and bacterial virulence. Annu Rev Microbiol 1984; 38: 6989. 3. Neilands JB. Microbial envelope proteins related to iron. Annu Rev Microbioll982; 36: 285-309. 4. Barclay R. The role of iron in infection. Med Lab Sci 1985; 42: 166177. 5. Isles A, Maclusky I, Corey R et al. Pseudomonas cepacia infection in cystic fibrosis: an emerging problem. J Pediatr 1984; 104: 206210. 6. Sokol PA. Production and utilization of pyochelin by clinical isolates of Pseudomonas cepacia. J Clin Microbioll986; 23: 560-562. 7. Sokol PA, Woods DE. Effect of pyochelin on Pseudomonas cepacia respiratory infections. Microb Pathog 1988;5 : 197205. 8. McKevitt AI, Woods DE. Characterization of Pseudomonas cepacia isolates from patients with cystic fibrosis. J Clin Microbioll984; 19: 291-293. 9. Cox CD, Graham R. Isolation of an iron-binding compound from Pseudomonas aeruginosa. J Bacterioll979; 137: 357364. 10. Cox CD. Iron uptake with ferripyochelin and ferric citrate by Pseudomonas aeruginosa. J Bacterioll980; 142 : 581-587. 11. Schwyn B, Neilands JB. Universal chemical assay for the detection and determination of siderophores. Anal Biochem 1987; 160: 47-56.
'
These studies were supported by the Canadian Cystic Fibrosis Foundation. P.A.S is an Alberta Heritage Foundation for Medical Research Scholar.
