Dr Anthony W. Smith, Department of Pharmacy and Pharmacology, University of Bath, Claverton Down, Bath, BA2 7AY, UK (e-mail: email@example.com).
It has previously been shown that myo-inositol hexakisphosphate (myo-InsP6) mediates iron transport into Pseudomonas aeruginosa and overcomes iron-dependent growth inhibition. In this study, the iron transport properties of myo-inositol trisphosphate and tetrakisphosphate regio-isomers were studied. Pseudomonas aeruginosa accumulated iron (III) at similar rates whether complexed with myo-Ins(1,2,3)P3 or myo-InsP6. Iron accumulation from other compounds, notably d/lmyo-Ins(1,2,4,5)P4 and another inositol trisphosphate regio-isomer, d-myo-Ins(1,4,5)P3, was dramatically increased. Iron transport profiles from myo-InsP6 into mutants lacking the outer membrane porins oprF, oprD and oprP were similar to the wild-type, indicating that these porins are not involved in the transport process. The rates of reduction of iron (III) to iron (II) complexed to any of the compounds by a Ps. aeruginosa cell lysate were similar, suggesting that a reductive mechanism is not the rate-determining step.
Most Gram-negative micro-organisms take up physiologically essential iron by synthesizing and utilizing high affinity ferric iron chelators called siderophores. The ferric-siderophore complexes formed with environmental iron are recognized by proteins in the outer membrane that initiate their active transport into the cell. A number of micro-organisms bypass the requirement to synthesize their own siderophores by transporting iron complexed with exogenous chelators. Examples include adaptation to use siderophores produced by other species, such as ferrichrome, coprogen, and rhodotorulic acid transport by Escherichia coli ( Braun et al. 1991 ), and enterobactin transport by Pseudomonas aeruginosa ( Poole et al. 1990 ). Other compounds, such as ferric dicitrate, can also be used ( Harding & Royt 1990; Van Hove et al. 1991 ).
In order to gain a better understanding of the iron transport process mediated by inositol polyphosphates in Ps. aeruginosa, several inositol tetrakis- and trisphosphate regio-isomers have been studied. Iron transport properties in mutants lacking outer membrane porins have also been compared, and the rate of reduction of iron (III) complexed with each inositol polyphosphate by a Ps. aeruginosa cell lysate has been measured.
Materials and methods
Bacterial strains and culture conditions
Pseudomonas aeruginosa PAO1 (ATCC 15692) was from this laboratory’s collection. Pseudomonas aeruginosa H729 (oprD, met) ( Trias et al. 1989 ), H636 (oprF, met) ( Woodruff & Hancock 1988) and H576 (oprP, met) ( Poole & Hancock 1986) were the kind gift of R.E.W. Hancock (University of British Columbia, Vancouver). Strains were cultured in an iron-deficient succinate minimal medium (K2HPO4 6 g l−1, KH2PO4 3 g l−1 (NH4)2SO4 1 g l−1, MgSO4.7H2O 0·2 g l−1, sodium succinate 4 g l−1, pH 7·0) ( Meyer & Abdallah 1978), supplemented with 1 mmol l−1 methionine as necessary.
55Fe (III)-Inositol polyphosphate transport studies
Cells were grown in succinate medium in an orbital shaking incubator at 37 °C for 15 h and harvested by centrifugation at 4 °C. The cells were washed twice in 100 mmol l−1N-morpholinepropanesulphonic acid (MOPS pH 7·0) containing 60 μmol l−1 glucose, suspended to an O.D. at 470 nm of 2·0 (corresponding to 2 × 109 cells ml−1) and equilibrated at 37 °C for 15 min prior to transport studies. No observable increase in O.D. at 470 nm was detected during this period. The Fe-inositol polyphosphate complex was formed at 37 °C, 15 min prior to assay, as 200 μmol l−1 inositol polyphosphate and 400 nmol l−1 55FeCl3 (1·91 GBq mg−1, Amersham) in 1 ml MOPS buffer. The addition of this labelled complex to the cell suspension resulted in final concentrations of 100 μmol l−1 inositol polyphosphate and 200 nmol l−1 55FeCl3 (500 : 1 ratio). Iron transport was assayed by withdrawing duplicate 200 μl samples and filtering them through 0·2 μm pore size cellulose acetate membrane filters. The membrane filters were washed twice with 10 ml volumes of saline and allowed to air dry. The activity retained on the membrane filters was determined by scintillation counting on the 3H channel.
The assay was based on the method described by Halle & Meyer (1992a). Cells from a 4 litre overnight culture grown in succinate minimal medium were harvested by centrifugation (6000 g at 4 °C for 30 min), washed twice in distilled water, suspended in 25 ml buffer A (25 mmol l−1 Tris, 0·1 mol l−1 KCl pH 7·4) and broken by sonication (10 × 30 s pulses with 1 min intervals for cooling). The protein content of the lysate was adjusted to 0·8 mg ml−1 by addition of buffer A. For the reductase assay, 2 ml bacterial lysate were added to a glass spectrophotometer cuvette, followed by addition of inositol polyphosphate and FeCl3, each to a final concentration of 0·2 mmol l−1. Argon was bubbled through the cuvette for 10 min to remove oxygen. The reaction was started by addition of flavin mononucleotide (FMN), Ferrozine (Sigma) and nicotinamide adenine dinucleotide reduced form (NADH) to final concentrations of 0·05 mmol l−1, 0·8 mmol l−1 and 0·15 mmol l−1, respectively. The cuvette was transferred to the spectrophotometer and the reaction followed by measuring the absorbance at 562 nm at room temperature. A continuous stream of argon was bubbled through the reaction mixture. The extinction coefficient for Fe (II)-Ferrozine (28 000 M−1 cm−1) was used to quantify the reaction.
Iron transport studies
It was previously shown that iron (III) complexed by myo-InsP6 is transported into Ps. aeruginosa and overcomes iron-dependent growth inhibition ( Smith et al. 1994 ). In this work, these studies were extended to examine the iron transport properties of a number of inositol tetrakis- and trisphosphates. Figure 1 shows the structures of the inositol polyphosphate molecules used in this study. Each compound chelated iron (III), as evidenced by no precipitated counts in the filter assay. The transport data for iron (III) complexed with inositol tetrakis- and trisphosphates are shown in Figs 2 and 3, respectively. There were substantial differences in the rates of iron transport and accumulation of iron by Ps. aeruginosa from these compounds. For example, for iron complexed with inositol terakisphosphate regio-isomers, the initial rates of transport ranged from 3·05 to 25·00 pmol Fe min−1 per 109 cells for d/l-myo-Ins(1,2,4,6)P4 and d/l-myo-Ins(1,2,4,5)P4, respectively ( Fig. 2). After 30 min, the amount of iron accumulated by Ps. aeruginosa from these two compounds was 9·76 ± 4·53 (mean ± s. e., n= 3) and 95·83 ± 19·93 pmol Fe per 109 cells, respectively ( Fig. 2). Similarly, there was a large variation in the ability of Ps. aeruginosa to transport iron complexed with inositol trisphosphates. The iron transport profile mediated by myo-Ins(1,2,3)P3 was similar to that of myo-Ins6. The initial rate of iron transport was 3·15 pmol Fe min−1 per 109 cells, and 11·61 ± 1·39 pmol Fe per 109 cells had accumulated after 30 min ( Fig. 3). In contrast, an initial rate of transport of 55·56 pmol Fe min−1 per 109 cells, and 106·37 ± 6·57 pmol Fe per 109 cells accumulated after 30 min, were recorded for iron complexed with d/l-myo-Ins(1,4,5)P3 ( Fig. 3). Iron uptake studies mediated by d/l-myo-Ins(1,4,5)P3 were also extended to examine dependence on temperature and energy. Iron uptake by Ps. aeruginosa was inhibited by the metabolic poison carbonyl cyanide m-chlorophenylhydrazone and by incubating the cells at 0 °C (data not shown).
Iron transport into porin-deficient mutants of Ps. aeruginosa
The transport of myo-InsP6-mediated iron (III) into Ps. aeruginosa strains lacking outer membrane porin protein-F OprF ( Woodruff & Hancock 1988), the basic amino acid porin OprD ( Trias et al. 1989 ) and the phosphate porin OprP ( Poole & Hancock 1986) ( Fig. 4) was also investigated. In each case, the iron transport profiles were comparable with the parent strain PAO1.
Reduction of iron (III) complexed with inositol polyphosphates
The rate of reduction of iron (III) complexed with each inositol polyphosphate was compared to explore whether a reductive mechanism could explain the differences in the iron uptake profiles. A lysate of Ps. aeruginosa PAO1 was prepared, it was confirmed that the reduction of Fe (III)-myo-InsP6 was FMN-dependent, and the reduction of iron (III) to iron (II), complexed with each myo-inositol polyphosphate over 30 min, was followed. The rates of reduction are shown in Table 1. In no case was there a correlation between the rate of reduction and the rate of iron transport into Ps. aeruginosa ( Figs 2 and 3).
Table 1. Anaerobic reductase activity of a Pseudomonas aeruginosa PAO1 lysate on iron (III) inositol polyphosphate complexes
Reductase activity (nmol Fe (II)-ferrozine min−1)
The 2 ml reaction comprised bacterial lysate (1·6 mg protein), inositol polyphosphate (0·2 mmol l−1), FeCl3 (0·2 mmol l−1), Ferrozine (0·8 mmol l−1), NADH (0·15 mmol l−1) and FMN (0·05 mmol l−1). Experiments were performed anaerobically to prevent oxidation of FMNH2 formed by the reductase-catalysed reduction of FMN. (The letters refer to the structures in Fig. 1).
In this study, we have shown that Ps. aeruginosa can accumulate iron complexed with a number of inositol polyphosphates. The differences in transport properties cannot easily be explained by factors such as molecular weight and size, as all inositol tetrakis- and trisphosphate compounds are regio-isomers, respectively. Rather, we believe it reflects the mechanism by which Fe (III) is complexed. We propose that the 1,2,3 vicinal axial-equatorial-axial trisphosphate motif of myo-InsP6 and myo-Ins(1,2,3)P3 accommodates the six co-ordination sites of iron (III) to form a water-excluding octahedral complex ( Fig. 5). Preliminary molecular modelling studies support this proposal (data not shown). This octahedral co-ordination, which does not have an exchangeable site, is also consistent with our observation that free radical formation is inhibited when iron (III) is complexed with these molecules ( Spiers et al. 1996 ). Clearly, the other compounds in this study also chelated iron (III), even though they do not possess 1,2,3 vicinal phosphate groups. Therefore, groups in other positions can participate in chelation. In most cases, notably d-myo-Ins (1,4,5)P3 and d/l-myo-Ins (1,2,4,5)P4, this resulted in a greater rate and accumulation of iron (III) by Ps. aeruginosa. It is noteworthy that these compounds do not inhibit iron-catalysed free radical formation ( Hirst 1996; Spiers et al. 1996 ), indicating that iron complexed by these molecules can redox cycle. We have been unable to predict 1 : 1 fully occupied octahedral complexes with d-myo-Ins(1,3,4)P3 and d-myo-Ins(1,2,4,6)P4, which mediate similarly low rates of iron accumulation as myo-InsP6 and myo-InsP(1,2,3)P3. However, we cannot rule out the possibility of formation of stable bi- or ter-molecular complexes.
Our attempts to understand the molecular nature of the transport process into Ps. aeruginosa have been frustrating. Using iron complexed with myo-InsP6, we did not detect any differences in iron transport profiles using Ps. aeruginosa mutants lacking OprF, OprD or OprP, indicating that these major porins are not involved. Our low temperature and metabolic inhibitor process also argue against a diffusion process. This is in contrast to some ferri-siderophores in which Meyer (1992) reported reduced accumulation of iron in strains lacking OprF, showing that Ps. aeruginosa can take up siderophore-liganded iron through this porin.
We have shown that Ps. aeruginosa isolated cell membranes have a single class of binding site for myo-InsP6 ( Smith et al. 1994 ). In this present study, no compound, even at a concentration of 100 μmol l−1, was able to displace 1 nmol l−1[3H]-myo-InsP6 ( Hirst 1996), indicating that the interaction with the outer membrane is specific for myo-InsP6. This is perhaps not surprising given that this is the major naturally occurring inositol polyphosphate which Ps. aeruginosa is likely to encounter. In an attempt to identify the transport component, we have screened more than 50 000 transposon insertion mutants but have failed to isolate a single Fe (III)-InsP6 transport-deficient mutant.
We have also examined Ps. aeruginosa reductase activity, but did not demonstrate a rate-determining role. Halle & Meyer (1992a, b) have characterized a ferri-pyoverdine reductase activity from Ps. aeruginosa, which is initiated by the enzyme-catalysed reduction of FMN to FMNH2 by NADH; FMNH2 then mediates the reduction of Fe (III) to Fe (II). We demonstrated that reduction of ferri-inositol polyphosphate, like ferri-pyoverdine, is FMN dependent. However, in no case did the rate of reduction correlate with the rate of iron transport into Ps. aeruginosa, indicating that reductase activity is unlikely to be the rate-determining step. Similarly, our mutagenesis screening strategy should have picked up reductase mutants if this activity was involved in transport.
In summary, we have shown that Ps. aeruginosa can transport iron (III) complexed to a number of inositol polyphosphates. It is likely that it is myo-InsP6 which will be encountered by Ps. aeruginosa in environments such as soil. While some of the other compounds tested do occur naturally, notably the intracellular signalling molecule d-myo-Ins(1,4,5)P3 ( Berridge 1995), it is unlikely that they will be encountered by Ps. aeruginosa, even when causing infections. We do however, believe that they shed light on the manner in which myo-InsP6 chelates iron. We believe that the comparatively slow rate of iron (III) accumulation by Ps. aeruginosa when complexed with myo-InsP6 reflects the ability of this molecule to form an octahedral complex with iron (III) occupying all six co-ordination sites. This is consistent with myo-InsP6 being a biologically ‘safe’ form of stored iron, protecting biological systems from the toxic effects of this element, but one which Ps. aeruginosa is able to exploit. Given its abundance in some environments it will be interesting to determine whether other micro-organisms can utilize iron bound by this molecule.
The authors thank Dr J.M. Gardiner (UMIST) for assistance. This work was supported by grants from the Royal Society (AWS) and the BBSRC for a research studentship (IDS). SF is a Lister Institute fellow. BVLP is a Lister Institute Research Professor and is supported by BBSRC and the Wellcome Trust.