Induction of an outer membrane protein of 78 kDa in Vibrio vulnificus cultured in the presence of desferrioxamine B under iron-limiting conditions

Abstract

Although Vibrio vulnificus is known to be able to utilize ferrioxamine B as an iron source, its outer membrane receptor remains to be determined. In this study, we found that V. vulnificus expressed a new outer membrane protein of 78 kDa when grown in the presence of desferrioxamine B under iron-limiting conditions. The desferrioxamine B-dependent iron uptake was only observed in bacterial cells expressing this protein. Furthermore, non-denaturing polyacrylamide gel electrophoresis followed by autoradiography of the outer membrane preparation containing the 78-kDa protein preincubated with [⁵⁵Fe]ferrioxamine B provided a single radioactive band in which the 78-kDa outer membrane protein was present as the major component. These lines of evidence suggest that the inducible 78-kDa protein may serve as the cell-surface receptor for ferrioxamine B.

Siderophore receptor induction, Ferrioxamine B, Desferal, Vibrio vulnificus

1 Introduction

Iron is an essential element for the growth of most bacteria, but low levels of free iron both in the environment and mammalian hosts make it difficult for bacteria to sustain their growth. Many bacteria have adapted to this iron limitation by synthesizing small excreted molecules called siderophores which bind ferric ion and deliver it to the bacterial cell via specific cell-surface receptors[1]. Since in vivo growth of bacteria depends upon their ability to acquire iron from the host, the siderophore-mediated iron uptake has been suggested as one of the factors determining bacterial virulence[2]. In addition to their endogenous siderophores, some bacterial species such as Escherichia coli[2], Salmonella typhimurium[3], Pseudomonas aeruginosa[4], Yersinia enterocolitica[5], and Neisseria gonorrhoeae[6] are able to utilize exogenous siderophores produced by other bacteria or fungi. Bacteria do not solely utilize the cognate siderophores, whether or not they produce their own siderophores.

Vibrio vulnificus, a marine bacterium which is capable of causing lethal septicemia or wound infections[7], has been shown to produce vulnibactin as a cognate catecholate siderophore[8] and utilize heme as a sole source of iron[9]. Recently, the outer membrane receptors for ferric vulnibactin and heme have been characterized at the gene level [10,11]. In addition, it has long been recognized that this species can utilize desferrioxamine B, which is structurally dissimilar to vulnibactin, as an exogenous siderophore[12]. However, nothing is known about the components involved in its iron uptake and transport. It is well established that the transport of ferric siderophores involves the binding of the ligand to a specific outer membrane receptor as the first step[1]. We, therefore, hypothesized that a ligand-specific outer membrane receptor would be expressed for transport of ferrioxamine B in V. vulnificus. In this paper the induction of an outer membrane protein of 78 kDa is described in V. vulnificus cultured in the presence of desferrioxamine B. In addition, possible involvement of this protein as the receptor for the siderophore is explored.

2 Materials and methods

2.1 Bacterial strains and siderophores

The clinical isolates of V. vulnificus biotype 1, M2799 and L180, were used in this study. Desferrioxamine B mesylate (Desferal) and ferrichrome were purchased from Sigma, and vibrioferrin was isolated from the spent culture supernatant of Vibrio parahaemolyticus as previously described[13].

2.2 Growth promotion assays

Stationary-phase cells of V. vulnificus were diluted to an OD₆₆₀ of 0.05 with fresh Luria–Bertani broth (LB; 1% Difco tryptone, 0.5% Difco yeast extract, 2% NaCl) containing 250 μM 2,2′-dipyridyl (DPD) and the siderophore to be tested. Such a concentration of DPD was adopted, because there was no growth unless a usable exogenous siderophore was added to the medium. Cultures were shaken (200 rpm) at 37°C, and the cell density at OD₆₆₀ was determined at regular intervals.

2.3 Preparation and sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) analysis of outer membrane proteins

For the preparation of outer membrane fractions, V. vulnificus was grown at 37°C to the late-exponential phase of culture in LB broth with or without 200 μM DPD. In some experiments, desferrioxamine B was added to LB broth at a final concentration of 10 μM. Cells thus obtained were sonicated and 1% sodium N-lauroylsarcosinate (Sigma)-insoluble outer membrane proteins were isolated as previously described[14]. The protein concentration was determined by the Lowry method. Outer membrane preparations were mixed with an equal volume of the sample loading buffer (0.5 M Tris–HCl, pH 6.8, containing 2% SDS, 5% mercaptoethanol, 0.05% bromophenol blue and 30% glycerol), and heated for 5 min prior to electrophoresis. SDS–PAGE was carried out using 8% acrylamide in the running gel and the proteins were stained with Coomassie blue.

2.4 Determination of N-terminal amino acid sequences of outer membrane proteins

The proteins separated by SDS–PAGE were electroblotted to a prewetted polyvinylidene difluoride membrane (ProBlott, Applied Biosystems) using a Trans-Blot semi-dry electrophoretic transfer cell (Bio-Rad) essentially as described by Towbin et al.[15], and stained with Coomassie blue. After the membrane was rinsed several times with distilled water, the protein bands to be examined were excised from the membrane and the resulting fragments were air-dried. The N-terminal amino acid sequence was determined by automated Edman degradation with a Model 491 protein sequencer (Applied Biosystems) equipped with an online Model 120A PTH-amino acid analyzer.

2.5 Iron uptake assay

One hundred and five microliters of 1 mM FeCl₃ and 6.5 μl of the ⁵⁵FeCl₃ commercial solution (370 MBq ml⁻¹, NEN) were incubated with 41 μl of 30 μM desferrioxamine B at 37°C for 1 h with gentle shaking. The 10-fold molar excess of desferrioxamine B to ferric ion was used to ensure that all the added iron was coordinated. The [⁵⁵Fe]ferrioxamine B solution thus prepared was diluted with the uptake buffer (100 mM Tris–HCl, pH 7.5, containing 0.025% sodium succinate, 2% NaCl, and 10 mM MgCl₂) to make a final Fe³⁺ concentration equal to 5 μM. The M2799 strain was grown under the above conditions, and the cell pellet was washed three times with the uptake buffer containing 100 μM ethylenediamine-di-(o-hydroxyphenyl)acetic acid (Sigma) to remove contaminating iron and then suspended in the uptake buffer at an OD₆₆₀ of 0.2. The cell suspension (3.5 ml) was preincubated in a conical polypropylene tube for 10 min at 37°C with agitation prior to addition of an equal volume of the [⁵⁵Fe]ferrioxamine B solution. A sample (1.0 ml) was taken at regular intervals, rapidly filtered onto a membrane filter (Millipore HAWP, 0.45 μm), and washed twice with 5 ml of the uptake buffer. The filters were dried, and the amount of ⁵⁵Fe retained on the filters was measured by liquid scintillation counting (tritium channel). The non-specific radioactivity of the zero-time sample was subtracted from each value.

2.6 Non-denaturing PAGE followed by autoradiography of the outer membrane proteins preincubated with [⁵⁵Fe]ferrioxamine B

An aliquot of the outer membrane fraction (25 μg of protein) was incubated with 0.3 ml of 5 μM [⁵⁵Fe]ferrioxamine B at 37°C for 15 min. After centrifugation (20 000×g for 15 min), the precipitate was washed twice with 0.5 ml of 0.1 M Tris–HCl buffer (pH 7.5) containing 0.1 mM ethylenediamine tetraacetic acid (EDTA), suspended in 20 μl of 10 mM Tris–HCl buffer (pH 8.0) containing 10 mM benzamidine, 2% Triton X-100, 5 mM EDTA and 1%n-octyl-β-d-glucoside (Dojindo, Kumamoto, Japan) and then left to stand for 15 min to solubilize the outer membrane proteins, according to the procedure used for solubilization of ferric enterobactin receptor protein, FepA[16]. The supernatant after centrifugation was applied to non-denaturing PAGE with an 8% separating gel, where 0.1% Triton X-100 was used in place of SDS, and the proteins were electrophoresed at 15 mA for 4 h at 4°C, as described elsewhere[17]. The same samples were run in parallel, and one was stained with Coomassie blue and the other was dried for autoradiography. The autoradiogram was prepared by exposing the dried gel to Fuji RX-U film for 5 days at −80°C. In other experiments with cold ferrioxamine B, the Coomassie blue-stained band with the same electrophoretic mobility as the radioactive band was excised from a preparative gel, and the proteins were electro-eluted from the gel placed in a dialysis bag with the SDS–PAGE running buffer. The eluate was concentrated with a Centricon-10 (Amicon) filter for SDS–PAGE analysis.

3 Results and discussion

3.1 Utilization of exogenous siderophores by V. vulnificus

We were interested to examine whether V. vulnificus could utilize different exogenous siderophores for growth. Although V. vulnificus M2799 failed to grow in LB broth containing the synthetic iron chelator DPD (250 μM), the addition of desferrioxamine B to this medium restored the growth (Fig. 1). This was consistent with the observation by Wright et al.[12] that desferrioxamine B was capable of providing iron to V. vulnificus. The result also showed that this siderophore has a stronger affinity for iron than DPD. However, vibrioferrin and ferrichrome failed to restore the growth of strain M2799 under the iron limitation imposed by DPD. This failure indicated a lack of the transport systems for these siderophores, since they were efficient in stimulating growth of V. parahaemolyticus WP1 under the same conditions. A similar growth stimulatory effect of desferrioxamine B was seen in V. vulnificus L180 under the same conditions (data not shown).

1

Growth promotion assay of V. vulnificus M2799 by exogenous siderophores. Each of the siderophores was added at a final concentration of 10 μM to LB broth containing 250 μM DPD. Symbols: ◯, LB medium without DPD (positive control); •, LB medium with 250 μM DPD (negative control); ▵, desferrioxamine B; □, vibrioferrin; █, ferrichrome. A representative example of three experiments is shown.

3.2 Induction of a new outer membrane protein in the presence of desferrioxamine B

It has been reported that V. vulnificus strains usually express at least two major iron-repressible outer membrane proteins in response to iron limitation, although their apparent molecular masses vary from 71 to 88 kDa between strains[18]. The V. vulnificus M2799 also expressed two major iron-repressible outer membrane proteins with apparent molecular masses of 72 and 86 kDa, both of which were completely repressed under iron-sufficient conditions (Fig. 2A, lanes 1 and 2). A faint band of 76 kDa was also detected, but was excluded in this study. Interestingly, a new outer membrane protein of 78 kDa was expressed when strain M2799 was cultured in the presence of desferrioxamine B (10 μM) under iron-limiting conditions (Fig. 2A, lanes 3). Expression of this inducible protein, however, was completely suppressed upon growth under iron-sufficient conditions even if desferrioxamine B was present (Fig. 2A, lane 4), indicating that the induction of this protein requires two signals, iron limitation and the presence of desferrioxamine B. The 78-kDa protein was also induced in strain L180 under the same conditions (Fig. 2B). Growth promotion of V. vulnificus by desferrioxamine B accompanying the concomitant induction of the 78-kDa outer membrane protein strongly suggests that this inducible protein functions as the receptor for ferrioxamine B. However, it is uncertain whether induction of the 78-kDa protein is specific to desferrioxamine B, since the possibility that other siderophores structurally closely related to desferrioxamine B may also induce this putative receptor is not excluded.

2

SDS–PAGE analysis of outer membrane fractions of V. vulnificus M2799 (panel A) and L180 (panel B). The organisms were grown to the late-exponential phase in LB broth in the presence (+) or absence (−) of 200 μM DPD or 10 μM desferrioxamine B (DFO), and the outer membrane fractions were prepared for SDS–PAGE analysis as described in the text. Per lane, 50 μg protein was loaded, and proteins were stained with Coomassie blue. Molecular mass markers are indicated on the right in each panel. The N-terminal amino acid sequences of the bands with an arrow were determined.

3.3 N-terminal amino acid sequences of outer membrane proteins

The 78-kDa outer membrane proteins induced in the M2799 and L180 strains provided the same N-terminal amino acid sequence, EQSSIENAQL. However, there was no homology to any known ferrioxamine B receptor, or any other protein. The 86- and 82-kDa proteins in M2799 and L180, respectively, had the same N-terminal sequence, QDAGLFDEVV, which was identical to that of the 77-kDa mature heme receptor, HupA, in V. vulnificus MO6-24[11]. Variation in size assignments of these putative heme receptors from the 77-kDa HupA was unexpected, and is probably due to differences in strains as pointed out by Wright et al.[18] and/or in techniques. Moreover, the N-terminal sequences of the 72-kDa proteins of M2799 and L180 were determined to be QTDNTNSNKK and QTENTNSHKK, respectively, both of which found identity with eight of the first 10 amino acid residues (QTESTNSNKK) of the mature protein of the V. vulnificus MO6-24 vulnibactin receptor, VuuA[10], suggesting that these proteins may be the vulnibactin receptor.

3.4 Desferrioxamine B-mediated iron transport in V. vulnificus

Desferrioxamine B-mediated iron uptake was observed in V. vulnificus M2799, although only in cells which had been cultured in the presence of desferrioxamine B under iron-limiting conditions (Fig. 3). This iron uptake activity was almost completely lost if the cells were incubated at 0°C, suggestive of energy dependence in this iron uptake system. In contrast, no iron uptake activity was found in cells lacking the 78-kDa protein which had been precultured in the absence of desferrioxamine B under iron-limiting conditions. This provided crucial evidence supporting our notion that the 78-kDa outer membrane protein is induced to function as a cell-surface receptor in the initial step of desferrioxamine B-mediated iron uptake.

3

Ferrioxamine B-mediated iron uptake activity in V. vulnificus M2799 cells previously grown in iron-limited LB broth with (◯, ▵) or without (•) ferrioxamine B (10 μM). The uptake mixture contained 2.5 μM [⁵⁵Fe]ferrioxamine B and the cells at an OD₆₆₀ of 0.1, and was incubated at 37°C (◯, •) or 0°C (▵). A representative of three experiments is shown.

3.5 Detection of the [⁵⁵Fe]ferrioxamine B-binding protein in outer membrane fraction

To obtain more convincing evidence that the inducible 78-kDa protein may be the receptor for ferrioxamine B, we examined whether ferrioxamine B can directly bind to this protein. After incubation with [⁵⁵Fe]ferrioxamine B, the outer membrane fractions prepared from cells grown in the presence or absence of desferrioxamine B were analyzed by non-denaturing PAGE followed by autoradiography. Non-denaturing PAGE conditions, essentially the same as those in this study, have been successfully used for detection of the receptor protein binding ferric parabactin[17]. Non-denaturing PAGE revealed a new band only in cells grown in the presence of desferrioxamine B (Fig. 4, lane 2), which corresponded to the radioactive band on the autoradiogram (Fig. 4, lane 4). No radioactive band was detected, when [⁵⁵Fe]ferrioxamine B alone or radioactive mineral iron (⁵⁵Fe³⁺) preincubated with the outer membrane fraction containing the 78-kDa protein was electrophoresed under the same non-denaturing PAGE conditions (data not shown). In addition, the radioactive band disappeared when the outer membrane fraction was subjected to digestion with 0.5 mg ml⁻¹ proteinase K (20 U ml⁻¹; Wako, Osaka) prior to incubation with [⁵⁵Fe]ferrioxamine B (data not shown). Subsequently, the Coomassie blue-stained band (on the preparative non-denaturing gel) corresponding to the radioactive band was electroeluted and analyzed by SDS–PAGE. As shown in Fig. 5, the predominant component in the eluate had the same electrophoretic mobility as the 78-kDa protein. Moreover, this band revealed the N-terminal amino acid sequence to be consistent with that of the 78-kDa protein in Fig. 2A, lane 3, indicating that a single radioactive band only detected in the outer membrane preparation containing the 78-kDa protein exclusively reflects the binding of [⁵⁵Fe]ferrioxamine B to this protein. These data allow us to propose that the inducible 78-kDa protein is consistent with the receptor protein binding ferrioxamine B, and that this binding is representative of the physiological function of the protein.

4

Detection of the [⁵⁵Fe]ferrioxamine B-binding protein in the outer membrane franctions of V. vulnificus M2799 by non-denaturing PAGE followed by autoradiography. Outer membrane proteins prepared from cells which had been cultured under iron-limiting conditions in the absence (lanes 1 and 3) or presence (lanes 2 and 4) of 10 μM desferrioxamine B were preincubated with [⁵⁵Fe]ferrioxamine B and the solubilized proteins were separated in duplicate by non-denaturing PAGE with an 8% polyacrylamide gel (25 μg of protein per lane). One was stained with Coomassie blue (lanes 1 and 2) and the other was autoradiographed (lanes 3 and 4). An arrow indicates the band which corresponded to that which bound ⁵⁵Fe.

5

SDS–PAGE of the proteins eluted from the band corresponded with the radioactive band in non-denaturing PAGE. Lane 1, outer membrane protein fraction prepared from cells grown under iron-limiting conditions in the presence of desferrioxamine B; lane 2, proteins electroeluted from the band (marked with an arrow in Fig. 4) with the same electrophoretic mobility as the radiolabeled band. Proteins were stained with Coomassie blue. An arrow indicates the 78-kDa protein.

In this study, utilization of desferrioxamine B by V. vulnificus reported by Wright et al.[12], was further extended by the fact that the 78-kDa outer membrane protein is induced in the presence of desferrioxamine B. Moreover, the experimental results presented suggest that this protein may bind desferrioxamine B in its ferric complex form to enter iron into the cell. Enteric bacteria such as Y. enterocolitica[5], Salmonella enterica[19] and Serratia marcescens[20] have also been shown to utilize desferrioxamine B as an exogenous iron carrier. However, these desferrioxamine B-mediated iron transport systems are quite different from the V. vulnificus system shown in this study. In these enteric bacteria, expression of the receptors required for utilization of desferrioxamine B is constitutive under iron-limiting conditions and independent of the presence of desferrioxamine B. Thus, further studies will be required not only to clarify the mechanism by which V. vulnificus induces the 78-kDa protein in the presence of desferrioxamine B, but also to define the components responsible for the transport of ferrioxamine B into the cytosol.

Although the in vivo significance of exogenous (heterologous) siderophore utilization has not yet been characterized, the existence of these uptake systems implies a finite possibility of encountering the corresponding siderophores in the bacterial environment. Ferrioxamine E and G, structural homologs of ferrioxamine B, have been shown to be produced by Erwinia[21] and Hafnia[22] species, respectively, in Enterobacteriaceae. Therefore, it is not impossible that desferrioxamines secreted by members of the gut commensal flora under some specific conditions may promote survival and proliferation of V. vulnificus and thereby assist its pathogenesis. The evaluation of this possibility will be the subject of further research.

Acknowledgements

The authors thank H. Yamada for determining the N-terminal amino acid sequences. This work was supported by a Grant-in-Aid for Scientific Research (C) from the Ministry of Education, Culture, Sports, Science and Technology, Japan.