Molecular cloning and characterization of the ferric hydroxamate uptake (fhu) operon in Actinobacillus pleuropneumoniae

Leonie G. Mikael1, Peter D. Pawelek2, Josée Labrie1, Marc Sirois3, James W. Coulton2 and Mario Jacques1

Groupe de Recherche sur les Maladies Infectieuses du Porc, Département de Pathologie et Microbiologie, Faculté de Médecine Vétérinaire, Université de Montréal, Saint-Hyacinthe, Québec, CanadaJ2S 7C61
Department of Microbiology and Immunology, McGill University, Montréal, Québec, CanadaH3A 2B42
Département de Chimie-Biologie, Université du Québec à Trois-Rivières, Trois-Rivières, Québec, CanadaG9A 5H73

Author for correspondence: Mario Jacques. Tel: +1 450 773 8521 ext. 8348. Fax: +1 450 778 8108. e-mail: jacqum{at}medvet.umontreal.ca


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The bacterium Actinobacillus pleuropneumoniae, a swine pathogen, utilizes ferrichrome as an iron source. This study details the molecular cloning and sequencing of the genes involved in the uptake of this hydroxamate siderophore. Four ferric hydroxamate uptake (fhu) genes, fhuC, fhuD, fhuB and fhuA, were identified in a single operon, and these were found to encode proteins homologous to proteins of the fhu systems of several bacteria, including Escherichia coli. The fhuA gene encodes the 77 kDa outer-membrane protein (OMP) FhuA, the receptor for ferrichrome. FhuD is the 35·6 kDa periplasmic protein responsible for the translocation of ferric hydroxamate from the outer to the inner membrane. FhuC (28·5 kDa) and FhuB (69·4 kDa) are cytoplasmic-membrane-associated proteins that are components of an ABC transporter which internalizes the ferric hydroxamate. Reference strains of A. pleuropneumoniae that represented serotypes 1 to 12 of this organism all tested positive for the four fhu genes. When A. pleuropneumoniae FhuA was affinity-tagged with hexahistidine at its amino terminus and expressed in an E. coli host, the recombinant protein reacted with an mAb against E. coli FhuA, as well as with a polyclonal pig serum raised against an A. pleuropneumoniae infection. Hence, the authors conclude that fhuA is expressed in vivo by A. pleuropneumoniae. Three-dimensional modelling of the OMP FhuA was achieved by threading it to the X-ray crystallographic structure of the homologous protein in E. coli. FhuA from A. pleuropneumoniae was found to have the same overall fold as its E. coli homologue, i.e. it possesses an N-terminal cork domain followed by a C-terminal ß-barrel domain and displays 11 extracellular loops and 10 periplasmic turns.

Keywords: outer-membrane protein, siderophore transport

Abbreviations: DIG, digoxigenin; fhu, ferric hydroxamate uptake; OMP, outer-membrane protein; phoA, alkaline phosphatase; TbpA, transferrin-binding protein A; TbpB, transferrin-binding protein B

The GenBank accession number for the sequence of the fhuCDBA operon of Actinobacillus pleuropneumoniae serotype 1 reference strain 4074 described in this study is AF351135.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Actinobacillus pleuropneumoniae is the aetiological agent of porcine pleuropneumonia, a highly contagious respiratory disease with major economic implications for the swine industry worldwide (Tascón et al., 1996 ). Infection by A. pleuropneumoniae is a multifactorial process that is governed by many virulence factors which act alone or in concert to establish the pathogen in the porcine host. Iron has long been associated with bacterial virulence, either as a requirement for bacterial growth or by acting as an environmental signal that regulates the expression of other virulence factors (Braun, 2001 ). The scarce bioavailability of iron (10-18 M) at concentrations lower than are required for most bacteria to grow (10-6 to 10-8 M) necessitates that bacteria utilize mechanisms for high-affinity iron acquisition.

A. pleuropneumoniae is capable of using haemoglobin, haemin-containing compounds and porcine transferrin as sources of iron for its growth (Bélanger et al., 1995 ; Deneer & Potter, 1989 ). In addition, it can produce haemolysins, toxins that belong to the RTX (repeats-in-toxin) group of proteins (Schaller et al., 1999 ). All of these factors may contribute to the virulence of this bacterium. A. pleuropneumoniae also responds to iron-restricted conditions by inducing the synthesis of a specific subset of outer-membrane proteins (OMPs) (Deneer & Potter, 1989 ; Niven et al., 1989 ; Soltes & MacInnes, 1994 ; M. Archambault, personal communication), including two membrane-bound transferrin-specific receptors called TbpA and TbpB (Gerlach et al., 1992a , b ; Gonzalez et al., 1995 ). Although it was tempting to speculate that one of these proteins might serve as a receptor for a siderophore, preliminary bioassays by Deneer & Potter (1989) did not demonstrate any siderophore production in A. pleuropneumoniae. However, it was suggested that A. pleuropneumoniae might obtain iron in vivo directly from host sources in a manner similar to that of Neisseria species, which apparently also do not produce siderophores (Mickelson et al., 1982 ; West & Sparling, 1985 ). Niven et al. (1989) were unable to detect hydroxamate and catecholate siderophores in culture supernatants of A. pleuropneumoniae grown under iron-restricted conditions. By contrast, when Diarra et al. (1996) tested the ability of all serotypes of A. pleuropneumoniae to use different exogenous sources of iron (specifically catecholates and hydroxamates), growth promotion assays showed that all of the A. pleuropneumoniae strains tested (with the exception of one field strain of serotype 5) were capable of using ferrichrome as a growth-promoting substance under iron-limited conditions. They also demonstrated that A. pleuropneumoniae strain 87-682 of serotype 1 and strain 2245 of serotype 5 secreted an iron chelator into the culture medium in response to iron stress. However, this potential A. pleuropneumoniae siderophore had a structure that did not conform to that defined by the well-characterized assay for catechols established by Arnow (1937) or the assay of Csaky (1948) for hydroxamates. It is worth noting that some bacteria are known to use siderophores that are produced by other micro-organisms; hence, these bacteria must have the necessary receptors for the assimilation of different siderophores.

Several fungi, including Ustilago sphaerogena, synthesize ferrichrome, a hydroxamate siderophore. The ferric hydroxamate uptake (fhu) system in Escherichia coli is well recognized as one of the paradigms for siderophore transport (Braun, 1995 ; Coulton et al., 1983 ; Locher et al., 1998 ). The E. coli fhu system consists of four genes, designated fhuA, fhuC, fhuD and fhuB, which are arranged in one operon at minute 3 of the linkage map (Fecker & Braun, 1984 ) and transcribed clockwise in the same order. fhuA encodes the multifunctional OMP FhuA (79 kDa) that acts in E. coli as the ferrichrome-iron receptor as well as the receptor for phages T1, T5, {phi}80 and UC-1, for the bacterial toxin colicin M and for some antibiotics, such as albomycin (a structural analogue of ferrichrome) and rifamycin CGP 4832 (Ferguson et al., 2001a ). FhuA is a key player in the fhu system, as it is specific for Fe3+-ferrichrome and functions as a ligand-specific gated channel (Ferguson et al., 1998a ). The elucidation of the high-resolution X-ray crystallographic structure of FhuA from E. coli (Ferguson et al., 1998a ; Locher et al., 1998 ) provided a major advance in the understanding of some of the structure–function relationships of this protein.

The other proteins of the fhu system, namely FhuD, FhuC and FhuB, are also essential to its function. Periplasmic FhuD (31 kDa) and cytoplasmic-membrane-associated FhuC (29 kDa) and FhuB (41 kDa) are proteins necessary for the transport of ferrichrome and other Fe3+-hydroxamate compounds (Fe3+-aerobactin, Fe3+-coprogen) from the periplasm, across the cytoplasmic membrane into the cytoplasm (Braun et al., 1991 ; Coulton et al., 1987 ; Mademidis et al., 1997 ). The protein complex TonB–ExbB–ExbD (Günter & Braun, 1990 ; Postle, 1993 ) provides energy for this process.

Here, we report that the genome of A. pleuropneumoniae contains an operon with genes homologous to those of the E. coli fhu system, albeit in a different gene order. We also studied the distribution of these fhu genes among the different serotypes of A. pleuropneumoniae and the expression of the gene encoding the OMP receptor FhuA. The structural similarities between FhuA of E. coli and FhuA of A. pleuropneumoniae were deduced by three-dimensional modelling.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacterial strains, plasmids and growth conditions.
Strains of E. coli and the plasmids used in this study are listed in Table 1. The A. pleuropneumoniae reference strains (laboratory stock) used in this study were: serotype 1, strain 4074; serotype 2, strain 4226; serotype 3, strain 1421; serotype 4, strain 1462; serotype 5, strain K-17; serotype 6, strain FEMO; serotype 7, strain WF83; serotype 8, strain 405; serotype 9, strain 13261; serotype 10, strain 13039; serotype 11, strain 56153; serotype 12, strain 8329/85. These reference strains were grown on brain–heart infusion agar (Difco Laboratories) supplemented with 15 µg NAD+ ml-1. E. coli BL21(DE3), E. coli CC118 and E. coli DH5{alpha} were grown on Luria–Bertani (LB) Miller medium, whereas E. coli XL-1 Blue MRF' was grown in NZYCM broth (Gibco-BRL) supplemented with 12·5 µg tetracycline ml-1. The phage excision strain of E. coli, XLOLR, was grown in LB broth supplemented with 12·5 µg tetracycline ml-1. Growth of the phagemid pBK-CMV was carried out for selection on LB agar supplemented with 50 µg kanamycin ml-1. Strains harbouring derivatives of the pET30a+ expression vector were grown on LB agar supplemented with 30 µg kanamycin ml-1; those harbouring pGEM derivatives required 50 µg ampicillin ml-1 to be added to the medium. Blue/white colony selection was achieved by the addition of 40 µg X-Gal ml-1 and 1 mM IPTG to the medium. The addition of 50 µg deferrated ethylenediamine dihydroxyphenyl acetic acid ml-1 (Sigma) to the growth medium depleted the iron available to the bacteria. Ferrichrome (Sigma), used as an exogenous source of iron, was either spotted onto the deferrated medium or incorporated into the top agar.


View this table:
[in this window]
[in a new window]
 
Table 1. Bacterial strains and plasmids

 
Construction of an A. pleuropneumoniae DNA-signal-sequence library.
To screen for exported proteins in A. pleuropneumoniae (Sirois et al., 2001 ), chromosomal DNA from A. pleuropneumoniae serotype 1 strain 4074 was digested with Sau3AI and then separated by agarose-gel electrophoresis. DNA fragments of 500–1500 bp in size were excised from the gel, and the DNA was extracted using a QIAquick Gel Extraction Kit (Qiagen). The vector (pHRM104; Pearce et al., 1993 ) containing the truncated alkaline phosphatase gene ('phoA) was digested with BamHI, dephosphorylated with shrimp alkaline phosphatase and ligated with the A. pleuropneumoniae Sau3AI fragments. Cells from E. coli phoA mutant strain CC118 were transformed with the ligation mixture by electroporation and then incubated in LB Miller broth for 1 h. The cells were then plated onto LB agar plates containing 500 µg erythromycin ml-1 and 50 µg 5-bromo-4-chloro-3-indolyl phosphate (XP) ml-1. The translocation of PhoA across the bacterial inner membrane results in the hydrolysis of XP and the development of the blue colony phenotype. This indicates that the fusion proteins were derived from plasmids containing an A. pleuropneumoniae DNA sequence harbouring a promoter, a translational start site and a functional signal sequence (Mintz & Fives-Taylor, 1999 ). Restriction and modification enzymes were purchased from Amersham Pharmacia Biotech and Roche Diagnostics, and were used according to the manufacturers’ instructions.

Genetic techniques.
Plasmids from the PhoA-positive colonies were isolated using the Plasmid QIAprep Spin Miniprep Kit (Qiagen). The A. pleuropneumoniae DNA insert was sequenced using the oligonucleotide primer foA (Table 2), which hybridizes to the negative strand of phoA. Subsequent primers (F7 and R7; Table 2) were designed from within the plasmid sequence of the positive clone of relevance to this study. An oligonucleotide primer (FA1; Table 2) based on a region of promoter sequences for fhuA from E. coli was also designed. PCR was carried out with standard conditions and varying annealing temperatures, depending on the sequence of the primers used. When the expected PCR product was larger than 5·0 kb, the Expand Long Template PCR System 1 (Roche Diagnostics) was used in the reaction, rather than Taq polymerase. DNA sequencing of the PCR product was performed at the DNA Sequencing Core Facility of the University of Maine, by using an ABI model 373A stretch DNA sequencer (Applied Biosystems). Single-stranded synthetic oligonucleotides (Table 2) were synthesized by BioCorp. Various PCR products were then digoxigenin (DIG)-labelled using the DIG DNA Labelling and Detection Kit (Roche Diagnostics) and used as DNA hybridization probes in Southern blots and plaque-lift assays. Chromosomal DNA was extracted by the method of Pitcher et al. (1989) . For Southern-blotting experiments, the genomic DNA of the A. pleuropneumoniae serotypes was digested with different enzymes, run on a 0·7% agarose gel and then transferred to positively-charged nylon membranes (Ausubel et al., 1998 ). Conditions of high stringency were applied. Hybridization of the DIG-labelled DNA probes was detected by using phosphatase-labelled anti-DIG antibodies and revealed colorimetrically with the use of NBT-BCIP as substrate.


View this table:
[in this window]
[in a new window]
 
Table 2. Sequences of primers used in this study

 
Construction and screening of an A. pleuropneumoniae phage bank.
A library of A. pleuropneumoniae serotype 1 strain 4074 DNA was constructed (Vézina et al., 1997 ) in the {lambda}Zap Express phage vector (Stratagene), following the manufacturer’s instructions. Briefly, A. pleuropneumoniae genomic DNA was partially digested with Sau3AI and the resulting DNA fragments were separated by agarose-gel electrophoresis. The DNA fragments corresponding to approximately 7 to 12 kb in length were excised from the gel and ligated with BamHI-digested {lambda}Zap Express arms and packaged in vitro with Stratagene Gold II packaging extracts. The E. coli strains used were XL-1 Blue MRF' as the host and E. coli XLOLR as the excision plating strain. The library was screened with the probes designed for Southern blotting, which were DIG-labelled, and the positive plaques were again detected with phosphatase-labelled anti-DIG antibodies with NBT/BCIP as the substrate. The plaques showing a strong reaction were purified by successive rounds of screening. After three rounds of screening, the pBK-CMV phagemid of the positive clones was excised using ExAssist Helper Phage (Stratagene). Finally, the plasmids were digested with restriction enzymes XbaI and SacI. These enzymes are known to cut once each within the multiple-cloning site of the vector, and thereby evaluate the size(s) of the insert(s). The dideoxynucleotide sequencing reactions were carried out with universal primers T3 and T7 (Table 2). To complete walking the entire sequence of the plasmid insert(s), internal primers (5T3W1, 5T7W1, 5T3W2, 5T7W2, 5T7W2.5, 5T3W3, 5T7W3, 5T3W4, 5T7W4 and 5T3W5; Table 2) were designed that were based on the sequence information obtained.

Analysis of sequence homology and protein localization.
DNA sequence analysis was carried out with the aid of programs from University of Wisconsin Genetics Computer Group Software (Devereux et al., 1984 ) using the National Center for BiotechnoIogy Information database (http://www.ncbi.nlm.nih.gov). The multiple-sequence alignment of proteins employed the CLUSTAL alignment algorithm, provided by the Baylor College of Medicine Search Launcher (http://dot.imgen.bcm.tmc.edu:9331/multi-align/multi-align.htm). Prediction of protein localization and cleavage sites for signal sequences was carried out by applying the method of Nakai & Kanehisa (1991) .

Expression of recombinant histidine-tagged FhuA.
We elected to amplify the coding sequences (amino acids 1–668) of mature FhuA using primers AFor and ARev (Table 2); these primers were designed to maintain the reading frame of the fhuA gene. The PCR product was purified using a PCR Purification Kit (Qiagen), ligated into the pGEM-T Easy Vector (Promega) and then transformed into E. coli XL-1 Blue cells. The recombinant plasmid, pLM201, was digested with NotI, and the resulting NotI fragment was cloned directionally and in-frame into the NotI site of the pET30a+ expression vector (Novagen); this was done to append a hexahistidine tag to the amino terminus of FhuA. Plasmid pLM202, carrying the fhuA sequence for mature FhuA with a hexahistidine tag at its amino terminus, was transformed into E. coli DH5{alpha}. It was then extracted from DH5{alpha}, purified and the orientation and reading frame of the fhuA gene were verified by DNA sequencing. pLM202 was then transformed into E. coli BL21(DE3) (Novagen), the recommended host background for the expression of recombinant proteins in pET vectors. To express the histidine-tagged fusion protein, and following the manufacturer’s suggested procedure, cells containing the recombinant plasmid were grown in broth culture containing kanamycin and induced with IPTG (0·4, 1·0, 1·5 and 3·0 mM). Various times (2, 4, 6, 8 and 24 h) and temperatures (37 and 25 °C) were tested for the optimal expression of the fusion protein. Whole-cell protein samples were separated by SDS-PAGE following standard procedures; the proteins within the gel were then either stained with Coomassie blue or transferred to a nitrocellulose membrane for immunoblotting (Harlow & Lane, 1999 ).

Western blotting.
After the proteins had been transferred, the nitrocellulose membranes were blocked for 1 h with a solution of 1% BSA and then incubated overnight at 4 °C with a mouse-anti-hexahistidine mAb (Roche), a mouse mAb against E. coli FhuA (mAb Fhu6.1) (Moeck et al., 1995 ) or a convalescent serum obtained from a pig that had been experimentally infected with A. pleuropneumoniae serotype 1 (Jacques et al., 1996 ). The secondary antibodies used were either a goat-anti-mouse IgG+IgM (heavy+light) or an anti-swine IgG–horseradish peroxidase conjugate (Jackson ImmunoResearch Laboratories). Reactions were revealed by the addition of 4-chloro-1-naphthol and H2O2 (Sigma) to the membranes (Harlow & Lane, 1999 ).

Homology model for FhuA of A. pleuropneumoniae.
Using the sequence data for the gene encoding A. pleuropneumoniae FhuA, the predicted amino-acid sequence (residues 1–673 of the mature protein) of this protein was submitted to the JIGSAW 3D Protein Homology Modelling Server (http://www.bmm.icnet.uk/servers/3djigsaw) (Bates & Sternberg, 1999 ). The JIGSAW server returned the E. coli FhuA structure, PDB code 1qfg (Ferguson et al., 2000 ), as the only successful structural template within its sample space. Visual inspection of this homology model revealed a number of discontinuities within the extracellular- and periplasmic-loop regions, namely C{alpha} pairs in which the distances were too great to form a covalent bond. To better model the loops of A. pleuropneumoniae FhuA, anchor regions flanking these discontinuities were submitted to the CODA server (http://www-cryst.bioc.cam.ac.uk/coda/coda.html) (Deane & Blundell, 2001 ). CODA models amino-acid sequences of proposed loops by searching against known structures of loop regions within other proteins. The CODA server was able to model all loop regions within the A. pleuropneumoniae FhuA sequence that the JIGSAW server could not. The acceptable r.m.s. (root mean square) deviation was <1·00 . The loops and anchor regions modelled by the CODA server are as follows, 158–166, 183–191, 228–236, 291–299, 301–308, 405–412, 538–546, 564–572, 597–604, 610–617 and 630–637. This numbering is based on the amino acid residues of the FhuA sequence of A. pleuropneumoniae.


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Cloning of the fhu operon in A. pleuropneumoniae
By using a truncated gene for alkaline phosphatase ('phoA) that lacks a functional signal sequence, a system was developed that identifies genes encoding exported proteins (Manoil & Beckwith, 1985 ). This system was modified by Sirois et al. (2001) for use with A. pleuropneumoniae. Gene fusions between the coding region of a heterologous signal sequence plus sequences from a normally exported protein and 'phoA may result in the expression of PhoA activity. PhoA-positive colonies are indicative of fusion proteins derived from plasmids containing a foreign DNA insert harbouring a promoter, a translational start site and a functional signal sequence (Mintz & Fives-Taylor, 1999 ).

When our A. pleuropneumoniae signal-sequence library, representing approximately 8250 individual colonies in the phoA-negative E. coli strain CC118, was screened for the blue colony phenotype on medium containing 5-bromo-4-chloro-3-indolyl phosphate, 95 colonies were found to be PhoA-positive. One plasmid (pI-25) from the PhoA-positive colonies contained a 375 bp A. pleuropneumoniae insert that was relevant to our objectives. BLASTX analysis of this fragment showed that it displayed 36% identity with and 57% similarity to the Rhizobium leguminosarum FhuD protein (amino acids 23–80, out of a total of 301). The 375 bp fragment also showed 35% identity with and 48% similarity to the E. coli FhuD protein (amino acids 28–80, out of a total of 296). To obtain the full fhuD sequence from the chromosome of A. pleuropneumoniae serotype 1, oligonucleotide primers (F7 and R7; Table 2) were designed from the 5' and 3' ends of the sequence in pI-25. These were paired with a primer (FA1; Table 2) that was based on a region of promoter sequences for fhuA from E. coli. A PCR product of 1063 bp was obtained with the primer pair FA1/R7. The deduced amino acids of this amplicon displayed an uninterrupted ORF for a sequence encoding FhuD; this ORF had 27% identity with the protein and 44% similarity to the gene in E. coli.

To isolate larger genomic fragments from A. pleuropneumoniae that contained fhuD and its neighbouring sequences, we used a genomic library of A. pleuropneumoniae serotype 1 strain 4074 DNA made in the {lambda}ZAP Express phage; the fragments in this library were approximately 7–12 kb in size. As a screening tool, we labelled the 1063 bp PCR product (i.e. the fhuD sequence) with DIG. From 10 plaques that gave a positive signal with the fhuD probe, one was subjected to three rounds of purification; the plasmid (pLM101) of this positive clone was then excised. Restriction analysis of pLM101 (Fig. 1) revealed an insert of 8·0 kb. The entire insert was then sequenced with universal primers T3 and T7, which annealed with the vector, and then with internal primers designed from the sequence walking. The total sequence information displayed four different ORFs corresponding to the genes fhuC, fhuD, fhuB and fhuA which appear in the fhu operons of several bacteria, including that of E. coli (Table 3).



View larger version (6K):
[in this window]
[in a new window]
 
Fig. 1. Map of the A. pleuropneumoniae serotype 1 reference strain 4074 fhu gene cluster, which is homologous to the fhuACDB cluster of E. coli. Numbers indicate the base-pairs of the insert within pLM101. ORFs designating the genes fhuC, fhuD, fhuB and fhuA are marked by thick solid lines. The direction of transcription is indicated by an arrow.

 

View this table:
[in this window]
[in a new window]
 
Table 3. Description of the genes in the fhu operon of A. pleuropneumoniae serotype 1 reference strain 4074, with the proteins they encode and their identities with homologous proteins from other bacteria

 
Sequence of the fhuC region
The nucleotide sequence encoding fhuC contains a single ORF that extends from nucleotide 892 to nucleotide 1656 (Fig. 1 and Table 3). The proposed translation start site is GTG (valine), which is known to act as a start codon in rare instances. The termination codon is TGA. BLASTX analysis of this ORF showed that it had identity with the ATP-binding protein FhuC of several bacteria (Table 3). FhuC is a cytoplasmic protein that belongs to the family of transporters that require binding proteins located within the periplasmic space (Linton & Higgins, 1998 ). It is a hydrophilic but membrane-associated protein that contains two domains which are typical for nucleotide-binding proteins (Burkhardt & Braun, 1987 ; Coulton et al., 1987 ; Köster & Braun, 1986 , 1989 ). Protein prediction analysis (Nakai & Kanehisa, 1991 ) of the FhuC homologue in A. pleuropneumoniae characterized this protein as having an ATP/GTP-binding-site motif A (GHNGSGKS, amino acids 34–41) and an ABC-transporter-family signature (LSGGERSRIWLAMLL, amino acids 138–152).

Sequence of the fhuD region
In A. pleuropneumoniae, the last codon (TGA) of fhuC overlaps with the initiation codon (ATG) of the following gene fhuD, which extends from nucleotide 1656 to nucleotide 2609 (Fig. 1 and Table 3) – this genetic organization matches the overlap of the fhuC and fhuD genes in E. coli (Coulton et al., 1987 ). This stretch of nucleotide sequence displays an uninterrupted ORF that terminates with TAA and which encodes a protein whose deduced amino-acid sequence reveals a protein with identity to the FhuD proteins of several bacteria (Table 3). FhuD is a periplasmic protein that is responsible for transporting ferrichrome from the FhuA receptor in the outer membrane to the FhuB protein in the cytoplasmic membrane (Köster & Braun, 1989 ; Moeck et al., 1996 ). Using software (Nakai & Kanehisa, 1991 ) for the prediction of signal-sequence-cleavage sites, a possible cleavage site was identified and is proposed to be situated at amino acid 47 for FhuD of A. pleuropneumoniae; however, the program failed to characterize this region as an N-terminal signal sequence that is usually displayed by proteins destined for export into the periplasm or outer membrane.

Sequence of the fhuB region
The initiation codon (ATG) for fhuB is situated 8 bp upstream of the termination codon TAA of the preceding gene fhuD in the fhu operon of A. pleuropneumoniae (Fig. 1). The ORF for FhuB extends from nucleotide 2603 to nucleotide 4555 (Fig. 1 and Table 3) and terminates with a TAA stop codon. The deduced amino-acid sequence of fhuB encodes a protein that displays identity with the FhuB protein of several bacteria (Table 3). Software for the prediction of protein localization (Nakai & Kanehisa, 1991 ) identified the FhuB homologue in A. pleuropneumoniae as being a cytoplasmic-membrane protein with 19 membrane-spanning regions, as compared to the 16 membrane-spanning regions that were predicted for E. coli. The FhuB homologue was also characterized as having an ABC-transporter-family signature sequence (IASGDPRANQLITWT, amino acids 489–503).

Sequence of the fhuA region
The last gene in the A. pleuropneumoniae fhu operon is an ORF that commences 46 bp downstream of fhuB, stretching from nucleotide 4602 to nucleotide 6689 (Fig. 1) and terminating with a TAA stop codon. The deduced amino-acid sequence of this ORF encodes a protein which displays identity with the FhuA proteins of several bacteria (Table 3). Software analysis for the prediction of protein localization (Nakai & Kanehisa, 1991 ) identified the putative A. pleuropneumoniae FhuA homologue as an OMP with 11 extracellular loops and 10 periplasmic turns, as well as a cleavable N-terminal signal sequence characteristic of proteins destined for export to the outer membrane. A potential cleavage site for the signal sequence was detected between amino-acid residues 23 and 24. The predicted molecular mass of the mature protein is 74750 Da. A multiple-sequence alignment generated using CLUSTAL W allowed a comparison of the amino termini of three known OMPs of A. pleuropneumoniae, TbpA (Gonzalez et al., 1995 ), FhuA (this study) and haemoglobin-binding protein, HgbA (A. Khamessan, personal communication). A stretch of 6 aa residues (shown in bold) that may act as a TonB box is located within the first 13 aa of these three target sequences (TbpA, EQAVQLNDVYVTG; FhuA, QETAVLDEVSVVS; HgbA, QEQMQLDTVIVKD). In other Gram-negative bacteria, the TonB box serves as a site of physical interaction between some outer-membrane receptors and TonB, a protein that delivers the proton motive force of the cytoplasmic membrane to the outer membrane (Moeck & Coulton, 1998 ; Postle, 1993 ).

Characterization of the fhu chromosomal locus in A. pleuropneumoniae serotype 1
In contrast to the fhu operon in E. coli in which fhuA is at the 5' end, fhuA in A. pleuropneumoniae is preceded by the genes fhuC, fhuD and fhuB. Upstream of fhuC in A. pleuropneumoniae is an incomplete ORF in the opposite direction, separated from the 5' start of fhuC by a 244 bp intergenic region and encoding a protein homologous to E. coli YaaH. This region contains promoter sites for the fhu genes, as well as a potential Fur-binding site for their regulation. A putative TAATTA box at -10 (nucleotides 854–859, Fig. 1) and a Shine–Dalgarno sequence (GGAG; nucleotides 883–886, Fig. 1) were also found in this stretch, along with a consensus sequence (TTTAA) for the -35 region (nucleotides 835–839, Fig. 1). Both the -35 and Shine–Dalgarno sequences match the consensus sequence proposed for promoter regions in A. pleuropneumoniae (Doree & Mulks, 2001 ). The spacing between the -10 and -35 regions was 14 bp, a value which falls within the range of 13 and 16 bp proposed for this region in A. pleuropneumoniae (Doree & Mulks, 2001 ). In E. coli, the fur gene product is a negative regulator of iron-dependent genes and the consensus sequence for the Fur box is GATAATGATAATCATTATC. In A. pleuropneumoniae the region upstream of fhuC (nucleotides 833–851) and the stretch of 46 non-coding base-pairs that falls between fhuB and fhuA (nucleotides 4598–4616) both demonstrated putative Fur boxes – nine out of 19 of the nucleotides were conserved upstream of fhuC, whereas 10 out of 19 of the nucleotides were conserved upstream of fhuA.

Analysis of the DNA sequence immediately downstream of the A. pleuropneumoniae fhuA gene shows that this region displays similarity with a hypothetical protein from Neisseria meningitidis Z2491 (NMA0986; Parkhill et al., 2000 ), which is transcribed in the same direction as the fhu genes. Four-hundred base-pairs downstream of fhuA a homologue to a Haemophilus influenzae protein is encoded (phospho-ribosyl-aminoamidazole-succinocarboxamide synthase), which is divergently transcribed.

Distribution of the fhu genes among A. pleuropneumoniae serotype reference strains
To determine if the fhu operon was unique to the A. pleuropneumoniae serotype 1 reference strain or if it was widely distributed among the different serotypes of this organism, samples of DNA from reference strains representing serotypes 2–12 of A. pleuropneumoniae were investigated by PCR. PCR amplification was carried out with primers 5T3W1 and 5T7W2 (Table 2), which flank the fhu region in A. pleuropneumoniae serotype 1, and the Expand Long Template PCR System 1. Reference strains from serotypes 2, 6, 8, 9, 10, 11 and 12 all showed (Fig. 2) the expected 6·0 kb PCR product, identical to the one observed for serotype 1. Using primers 5T3W1 and 5T7W2 for PCR, some serotype strains yielded a negative PCR result for the entire fhu operon. To investigate these serotype strains further, amplification of the fhu operon was attempted with primers internal to this operon (CFor and ARev; Table 2). Using primers CFor and ARev, the expected PCR product of 5·7 kb in size was obtained for the serotype 3 strain, confirming that the fhu operon is present in this serotype and with an arrangement of fhu genes similar to that seen in the serotype 1 strain. As the reference strains from serotypes 4, 5 and 7 produced a negative PCR result when CFor and ARev were used as primers, these serotypes were investigated with respect to the individual genes within the fhu operon both by PCR and Southern blotting. Pairs of primers internal to each gene of the fhu operon were used for PCR (AFor/ARev for fhuA, BFor/BRev for fhuB, CFor/CRev for fhuC and DFor/DRev for fhuD; Table 2). To elucidate the arrangement of the fhu genes in serotypes 4, 5 and 7, we performed PCR amplifications using various combinations of primers. For the Southern-blot analyses, EcoRI-digested genomic DNA from these serotypes was tested for hybridization to a DIG-labelled PCR product for each gene. Reference strains from serotypes 4, 5 and 7 tested positive for the presence of the fhuA, fhuB, fhuC and fhuD genes both in PCR and Southern-blot analyses, although the conditions of stringency had to be lowered to obtain a PCR amplicon for the fhuA and fhuD genes and a positive hybridization signal for fhuD – this may be due to some dissimilarity in the fhuD and fhuA gene sequences between serotype 1 and serotypes 4, 5 and 7.



View larger version (62K):
[in this window]
[in a new window]
 
Fig. 2. PCR analysis of the fhu operon in A. pleuropneumoniae serotype reference strains. The amplifications were done by using primers (5T3W1 and 5T7W2) flanking the fhu operon and chromosomal DNA from the A. pleuropneumoniae reference strains as templates. Lane M, DNA ladder (kb); the numbering of the other lanes relates to the serotypes outlined in Methods. The arrow points to the 6·0 kb mark.

 
Our results show that all of the genes of the fhu operon are present in all of the reference strains that represent serotypes 1–12 of A. pleuropneumoniae, and that they all seem to be organized in a similar arrangement. However, the regions flanking the fhu operon are dissimilar in serotypes 3, 4, 5 and 7 compared to the regions flanking the fhu operon in serotypes 1, 2, 6, 8, 9, 10, 11 and 12, as shown by the negative PCR result obtained for these four serotypes when using primers (5T3W1 and 5T7W2) that flanked this region.

Protein expression and Western blotting
The fhuA gene of the A. pleuropneumoniae serotype 1 strain was cloned into the expression vector pET30a+ to yield pLM202 (Table 1). pLM202 was then transformed into E. coli BL21 cells. The E. coli BL21(pLM202) cells were grown under various conditions to optimize the expression of recombinant FhuA with the amino-terminal histidine tag, i.e. we tested different concentrations of IPTG and varied the parameters of time and temperature. Whole-cell samples of E. coli BL21(pLM202) were run on 8 or 12·5% SDS-polyacrylamide gels and the gels were subsequently stained with Coomassie blue. A protein of approximately 79 kDa in size showed the highest level of expression at 25 °C (Fig. 3a). Induction with IPTG concentrations higher than 1·0 mM and induction periods longer than 6 h did not further enhance the expression of recombinant FhuA. The proteins from the gels were then transferred to a nitrocellulose membrane and blotted with an anti-hexahistidine mAb which reacted specifically with the 79 kDa protein, namely A. pleuropneumoniae FhuA with an N-terminal histidine tag (Fig. 3b). We then tested mAb Fhu6.1 (Moeck et al., 1995 ) (Fig. 3c), which recognizes a linear epitope between amino acids 241 and 281 of E. coli FhuA, against the 79 kDa protein. For both Western blots, a purified FhuA–hexahistidine of E. coli (Moeck et al., 1996 ) served as a positive control and uninduced E. coli BL21(pLM202) cells served as a negative control. We also tested a polyclonal antiserum taken from a pig infected experimentally with A. pleuropneumoniae serotype 1 and were able to show that the 79 kDa protein also reacted with this immune serum (Fig. 3d).



View larger version (37K):
[in this window]
[in a new window]
 
Fig. 3. SDS-PAGE (a) and Western blot (b–d) analyses of whole-cell lysates of E. coli BL21 cells expressing recombinant FhuA from A. pleuropneumoniae. Lanes: 1, E. coli BL21(pLM202) with no IPTG added; 2, E. coli BL21(pLM202) plus 1 mM IPTG, with 4 h incubation at 25 °C; 3, E. coli BL21(pLM202) plus 1 mM IPTG, with 24 h incubation at 25 °C; 4, E. coli BL21(pHX405) expressing the FhuA–hexahistidine protein. The gel shown in: (a) was stained with Coomassie blue; (b) was immunoblotted with an mAb against hexahistidine; (c) was immunoblotted with mAb Fhu6.1 directed against E. coli FhuA (Moeck et al., 1995 ); and (d) was immunoblotted with convalescent serum taken from a pig that had been infected experimentally with A. pleuropneumoniae serotype 1.

 
Homology model for FhuA of A. pleuropneumoniae and comparison with the structure of FhuA of E. coli
The JIGSAW and CODA servers were used to generate a composite homology model for FhuA of A. pleuropneumoniae. Following alignment of the amino-acid sequences of the FhuA proteins of A. pleuropneumoniae and E. coli (Fig. 4), the model (Fig. 5a) for the FhuA of A. pleuropneumoniae generated by the JIGSAW server showed 11 extracellular loops (L1–L11) and 10 periplasmic turns within this protein. These numbers of extracellular loops and periplasmic turns are consistent with the E. coli FhuA structure, as is the presence of 22 ß-strands. FhuA of A. pleuropneumoniae has the same overall fold as FhuA of E. coli. Both proteins possess two domains (Fig. 5b). In the A. pleuropneumoniae FhuA sequence, these two domains are represented by an N-terminal cork domain (residues 1–134) followed by a C-terminal ß-barrel domain (residues 135–673). There were two gaps in the model: the first occurred between residues 74 and 79 in the cork domain and the second occurred between residues 375 and 391 in the ß-barrel domain. The second gap overlapped the hexahistidine tag and flanking linker regions (a total of 11 aa) that were inserted into the sequence to facilitate affinity purification of the E. coli recombinant FhuA (Ferguson et al., 1998b ). It is reasonable to assume that the A. pleuropneumoniae model structure should not be threaded against this region, since it is not present in the wild-type A. pleuropneumoniae fhuA gene. We hypothesize that residues 375–391 in the A. pleuropneumoniae FhuA structure would be predominantly located within extracellular loop L5, with a short C-terminal stretch being contained within the adjacent ß-strand (Fig. 5a).



View larger version (38K):
[in this window]
[in a new window]
 
Fig. 4. Structure-based alignment of the A. pleuropneumoniae FhuA sequence relative to that of its template for homology modelling, E. coli FhuA. The numbering of both proteins is based on their mature, processed forms. The N termini (prior to the first gap) of the A. pleuropneumoniae and E. coli sequences are aligned relative to the putative TonB box (LDEVSV, enclosed text) of the A. pleuropneumoniae sequence, which is homologous to the E. coli TonB box. The structural alignment used by the JIGSAW server to produce the homology model begins at position 13 of the A. pleuropneumoniae FhuA protein. Residues corresponding to the recombinant hexahistidine tag and linker regions in the E. coli FhuA primary sequence are indicated in bold; these residues are not included in the 714 aa of the mature, processed FhuA protein from E. coli. Underlined residues in the A. pleuropneumoniae sequence are not present in the homology model for FhuA of this organism. Secondary-structure elements denoted by letters prefixed by an {alpha} or a ß correspond to those found within the cork domain of FhuA from E. coli; numbered secondary-structure elements prefixed by a ß correspond to those found in the ß-barrel domain of FhuA from E. coli; the boundaries for each of the 22 ß-strands of the ß-barrel domain are shown as solid lines below the alignments.

 


View larger version (83K):
[in this window]
[in a new window]
 
Fig. 5. (a) Proposed secondary structure for the ß-barrel domain of the FhuA protein of A. pleuropneumoniae. Squares containing one-letter amino-acid codes represent residues contained within ß-strands; circles containing one-letter amino-acid codes represent residues contained within loop regions. The boundaries between strand and loop regions correspond to those observed by visualization of the three-dimensional homology-based model for FhuA. Residues 375–391 are not present within the homology model, instead their secondary-structure status is predicted from the positions of homologous residues within the E. coli FhuA structure. The approximate position of the outer membrane is shown by parallel double lines. Extracellular loops are numbered L1–L11, starting from residue 135 of the first ß-strand. (b) Proposed tertiary structure (C{alpha} trace, stereo view) for the FhuA protein of A. pleuropneumoniae. The cork domain is represented by a solid line; the ß-barrel domain is represented by a grey line. The positions of loops L3, L4, L6 and L8 are given for reference. The stereo images were generated by using the program MOLSCRIPT (Kraulis, 1991 ).

 

   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacterial virulence is determined by the ability of an organism to compete for essential nutrients (Martinez et al., 1990 ). Because the mammalian host restricts bacterial growth by withholding iron, most bacteria have evolved a diverse series of high-affinity iron-acquisition systems to satisfy their iron requirements. One such system is involved in the synthesis and/or uptake of low-molecular-mass iron chelators, termed siderophores (Braun et al., 1998 ; Neilands, 1995 ). Transporters bind iron chelates with high affinity and mediate their uptake across the outer membrane of Gram-negative bacteria. The proteins required for the uptake of ferric hydroxamates are the products of the genes fhuA, fhuC, fhuD and fhuB where FhuA acts as the receptor for ferrichrome in the outer membrane, and its crystal structure has been determined (Ferguson et al., 1998a ; Locher et al., 1998 ). The energy required to translocate ferric hydroxymates across the outer membrane is derived from the proton motive force of the cytoplasmic membrane, as transduced by the TonB–ExbB–ExbD complex (Braun, 1995 ; Moeck & Coulton, 1998 ).

A. pleuropneumoniae, a Gram-negative bacterium that is an important swine pathogen, is capable of using transferrin, haemoglobin, haemin and exogenous siderophores, including ferric hydroxamates, as sources of iron for growth. The transferrin receptor complex includes TbpA and TbpB in the outer membrane of A. pleuropneumoniae (Deneer & Potter, 1989 ) as well as the exbB and exbD genes upstream of tbpA and tbpB in the same operon (Tonpitak et al., 2000 ). The objectives of the present study were to elucidate the genes and products involved in the uptake of ferric hydroxamates in A. pleuropneumoniae. We successfully cloned the homologues of the E. coli fhuACDB operon from A. pleuropneumoniae. In A. pleuropneumoniae the proteins encoded by this operon are: FhuA, a 77 kDa OMP that acts as the receptor for ferric hydroxamate; FhuD, the periplasmic protein responsible for the translocation of ferric hydroxamate from the outer to the inner membrane; and FhuC and FhuB, cytoplasmic-membrane-associated proteins that are components of an ABC transporter which internalizes ferric hydroxamate. An ABC-transporter-family signature sequence was identified for the FhuC and FhuB proteins of A. pleuropneumoniae. In other Gram-negative bacteria FhuB is a hydrophobic protein that is embedded in the cytoplasmic membrane. It is twice the size of hydrophobic proteins usually found in periplasmic binding transport systems (Linton & Higgins, 1998 ) and displays an internal amino-acid-sequence homology between its N-terminal and C-terminal halves. This internal homology led Köster & Braun (1989) to suggest that fhuB originated from the duplication of an ancestral gene, with the two DNA fragments fusing to form a single gene. The A. pleuropneumoniae FhuB homologue is the same size as its E. coli counterpart, and comparison of the two halves of the fhuB sequence of A. pleuropneumoniae revealed 64% sequence identity, hence suggesting similar ancestry. Southern blot and PCR analyses showed that all of the genes of the fhu operon are present in all of the reference strains tested here, which represented serotypes 1–12 of A. pleuropneumoniae. However, testing for the individual fhuC, fhuD, fhuB and fhuA genes revealed that although they are present individually in the reference strains of serotypes 4, 5 and 7, the sequences of fhuA and fhuD are somewhat different in these serotypes compared to the sequences of these genes in the other nine serotypes studied. Nevertheless, it appears that in all of the serotypes of A. pleuropneumoniae studied, the fhuC, fhuD, fhuB and fhuA genes are arranged in the same order as seen in the serotype 1 reference strain 4074.

The genes involved in the uptake of ferrichrome in A. pleuropneumoniae are arranged differently than their corresponding genes in E. coli. In the latter, the fhu receptor gene fhuA is located upstream of fhuCDB, whereas in A. pleuropneumoniae it is the last gene transcribed in the fhu operon. This difference in gene organization between the two species may reflect differences in the regulation of their iron-transport systems. To explore this possibility, further analyses are needed on the sequences upstream of fhuC and fhuA in A. pleuropneumoniae. The G+C content of the fhu operon in A. pleuropneumoniae is 44 mol%, a value that correlates well with the estimated 43·2 mol% reported (Pohl et al., 1983 ) for the genome of A. pleuropneumoniae. Interestingly, Galindo et al. (2001) have reported that the fhuA gene in the fhuABD operon of Campylobacter jejuni is strikingly GC-rich (65 mol%) compared to the C. jejuni genome (35 mol%).

The predicted primary sequence of FhuA from A. pleuropneumoniae has allowed us to propose a homology-based three-dimensional model for this protein (Fig. 5b). The overall fold of the A. pleuropneumoniae FhuA model resembles that of E. coli FhuA, for which the crystal structure is known (Ferguson et al., 1998a ; Locher et al., 1998 ), with the most significant deviations from the known structure occurring in the extracellular- and periplasmic-loop regions. Overall, the sizes of the extracellular loops of the FhuA model are similar to the corresponding loops in the E. coli structure. However, significant differences between the lengths of two key extracellular-loop regions were observed. Loop L3 in the E. coli structure is 31 residues long compared to 35 residues for the same loop in the A. pleuropneumoniae FhuA model, a difference due to an extension of the loop at its N-terminal end. Furthermore, L4 in the A. pleuropneumoniae FhuA model is considerably longer than the E. coli L4 region (28 residues compared to 20 residues) and does not contain the short ß-strands that are seen in the E. coli FhuA structure. Given that loops L3 and L4 are involved in ligand recognition and uptake (Ferguson et al., 1998a ), it remains to be determined whether these structural variations correspond to differences in function of FhuA from A. pleuropneumoniae relative to the E. coli protein. These differences could also prove to be responsible for the lack of susceptibility of A. pleuropneumoniae to the antibiotics albomycin and rifamycin CGP 4832 and to the bacterial toxin colicin M, all of which use FhuA as a docking site for entry into E. coli bacterial cells.

The X-ray crystallographic structure of E. coli FhuA complexed with ferricrocin indicates that 10 residues of this protein are within 4  of the bound ligand atoms (Ferguson et al., 2001b ). Inspection of the structure-based alignment of A. pleuropneumoniae FhuA with the primary sequence of the E. coli protein (Fig. 4) shows that six out of 10 positions are highly homologous. Absolute conservation is observed for residues Y292 and Y294 of the A. pleuropneumoniae protein, which align with E. coli ligand-binding residues Y313 and Y315. Residues F92 and F224 share the aromatic character of the aligned E. coli residues Y116 and W246. A position of hydrophobicity is maintained at position L364, which aligns with F391 in the E. coli sequence. Finally, S222, which has the potential to hydrogen bond to the siderophore ligand, is aligned with E. coli residue Y244. Interestingly, two of these six residues retain an aromatic character relative to their homologous E. coli ligand-binding residues, yet they do not have the ability to form hydrogen bonds with a siderophore ligand. This suggests either that the mode of ligand binding for the A. pleuropneumoniae protein may have hydrophobic-stacking interactions as a more predominant component or that the hydrophobic component of these residues in the E. coli protein may be more critical to ligand binding than to their ability to form hydrogen bonds. Functional analysis of the ligand-binding ability of a mutant E. coli FhuA protein in which the two aromatic residues have been modified to resemble those in the A. pleuropneumoniae sequence (Y116F, W246F) may resolve this issue.

The structure of E. coli FhuA also possesses a stretch of highly conserved residues along the longitudinal axis of the inner wall of the ß-barrel that is thought to form a ‘staircase’, potentially to facilitate siderophore transfer from the ligand-binding site to the periplasm (Ferguson et al., 1998a , 2001a ). Of these eight staircase residues, four homologous positions are conserved in the A. pleuropneumoniae primary sequence, as indicated by the structure-based alignment in Fig. 4. With a two-residue shift towards the C terminus, R274 and N276 of the A. pleuropneumoniae protein align with E. coli residues R297 and N299, respectively. Absolute conservation is seen with two charged residues, D331 and D352, which align with D358 and D379, respectively, in the E. coli sequence. Position Q397 in the A. pleuropneumoniae protein aligns perfectly with E. coli staircase residue Q431. Position R333 aligns with E. coli staircase residue Q360. Although at this position there is a charge difference between the two proteins, both side-chains are of approximately the same length, suggesting that the role of this staircase residue may be based more on the steric character of the side-chain than on its charge. It is interesting to note that with the exception of the Q397/Q431 pair, the highest degree of homology in the staircase is clustered at its N-terminal end, proximal to the periplasmic end of the ß-barrel domain.

This is, to the best of our knowledge, the first description of a siderophore receptor in A. pleuropneumoniae. In view of the fact that we now have genetic evidence as well as a three-dimensional model for the OMP FhuA of A. pleuropneumoniae, we can now distinguish different molecular mechanisms of siderophore receptors from different bacterial species and compare structural information between these species.


   ACKNOWLEDGEMENTS
 
This work was supported by Strategic Grant 224192 from the Natural Sciences and Engineering Research Council of Canada and by Fonds pour la formation de chercheurs et l’aide à la recherche (99-ER-0214 and 2002-ER-71900) to M.J. and J.W.C. We appreciate the contribution of G. S. Moeck in the early stages of this project. A. Khamessan also provided useful suggestions.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Arnow, L. E. (1937). Colorimetric determination of the compounds 3,4 dihydroxyphenylalanine-tyrosine mixture. J Biol Chem 118, 531-537.

Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Seidman, J. G., Smith, J. A. & Struhl, K. (1998). Current Protocols in Molecular Biology. New York: Wiley Interscience.

Bates, P. A. & Sternberg, M. J. E. (1999). Model building by comparison at CASP3: using expert knowledge and computer automation. Proteins 37 (S3), 47–54.

Bélanger, M., Bégin, C. & Jacques, M. (1995). Lipopolysaccharides of Actinobacillus pleuropneumoniae bind pig hemoglobin. Infect Immun 63, 656-662.[Abstract]

Braun, V. (1995). Energy-coupled transport and signal transduction through the gram-negative outer membrane via TonB–ExbB–ExbD-dependent receptor proteins. FEMS Microbiol Rev 16, 295-307.[Medline]

Braun, V. (2001). Iron uptake mechanisms and their regulation in pathogenic bacteria. Int J Med Microbiol 291, 67–79.[Medline]

Braun, V., Günter, K. & Hantke, K. (1991). Transport of iron across the outer membrane. Biol Met 4, 14-22.[Medline]

Braun, V., Hantke, K. & Köster, W. (1998). Bacterial iron transport: mechanisms, genetics and regulation. In Metal Ions in Biological Systems, pp. 67–145. New York: Marcel Dekker.

Burkhardt, R. & Braun, V. (1987). Nucleotide sequence of the fhuC and fhuD genes involved in iron (III) hydroxamate transport: domains in FhuC homologous to ATP-binding proteins. Mol Gen Genet 209, 49-55.[Medline]

Coulton, J. W., Mason, P. & DuBow, M. S. (1983). Molecular cloning of the ferrichrome-iron receptor of Escherichia coli K-12. J Bacteriol 156, 1315-1321.[Medline]

Coulton, J. W., Mason, P. & Allatt, D. D. (1987). fhuC and fhuD genes for iron (III)-ferrichrome transport into Escherichia coli K-12. J Bacteriol 169, 3844-3849.[Medline]

Csaky, T. Z. (1948). On the estimation of bound hydroxylamine in biological materials. Acta Chem Scand 2, 450-454.

Deane, C. M. & Blundell, T. L. (2001). CODA: a combined algorithm for predicting the structurally variable regions of protein models. Protein Sci 10, 599-612.[Abstract/Free Full Text]

Deneer, H. G. & Potter, A. A. (1989). Effect of iron restriction on the outer membrane proteins of Actinobacillus (Haemophilus) pleuropneumoniae. Infect Immun 57, 798-804.[Medline]

Devereux, J., Haeberli, P. & Smithies, O. (1984). A comprehensive set of sequence analysis programs for the VAX. Nucleic Acids Res 12, 387-395.[Abstract]

Diarra, M. S., Dolence, J. A., Dolence, E. K., Darwish, I., Miller, M. J., Malouin, F. & Jacques, M. (1996). Growth of Actinobacillus pleuropneumoniae is promoted by exogenous hydroxamate and catechol siderophores. Appl Environ Microbiol 62, 853-959.[Abstract]

Doree, S. M. & Mulks, M. H. (2001). Identification of an Actinobacillus pleuropneumoniae consensus promoter structure. J Bacteriol 183, 1983-1989.[Abstract/Free Full Text]

Fecker, L. & Braun, V. (1984). Cloning and expression of the fhu genes involved in iron (III) hydroxamate uptake in Escherichia coli. J Bacteriol 121, 497-503.

Ferguson, A. D., Hofmann, E., Coulton, J. W., Diederichs, K. & Welte, W. (1998a). Siderophore-mediated iron transport: crystal structure of the FhuA with bound lipopolysaccharide. Science 282, 2215-2220.[Abstract/Free Full Text]

Ferguson, A. D., Breed, J., Diederichs, K., Welte, W. & Coulton, J. W. (1998b). An internal affinity-tag for purification and crystallization of the siderophore receptor FhuA, integral outer membrane protein from Escherichia coli K-12. Protein Sci 7, 1636-1638.[Abstract/Free Full Text]

Ferguson, A. D., Welte, W., Hofmann, E., Linder, B., Holst, O., Coulton, J. W. & Diederichs, K. (2000). A conserved structural motif for lipopolysaccharide recognition by procaryotic and eucaryotic proteins. Structure (Camb) 8, 585-592.[Medline]

Ferguson, A. D., Coulton, J. W., Diederichs, K. & Welte, W. (2001a). The ferric hydroxamate uptake receptor FhuA and related TonB-dependent transporters in the outer membrane of gram-negative bacteria. In Handbook of Metalloproteins, pp. 834–849. Chichester: Wiley.

Ferguson, A. D., Ködding, J., Walker, G., Bös, C., Coulton, J. W., Diederichs, K., Braun, V. & Welte, W. (2001b). Active transport of an antibiotic rifamycin derivative by the outer membrane protein FhuA. Structure (Camb) 9, 707-716.[Medline]

Galindo, M. A., Day, W. A., Raphael, B. H. & Joens, L. A. (2001). Cloning and characterization of a Campylobacter jejuni iron-uptake operon. Curr Microbiol 42, 139-143.[Medline]

Gerlach, G. F., Klashinsky, S., Anderson, C., Potter, A. A. & Willson, P. J. (1992a). Characterization of two genes encoding distinct transferrin-binding proteins in different Actinobacillus pleuropneumoniae isolates. Infect Immun 60, 3253-3261.[Abstract]

Gerlach, G. F., Anderson, C., Potter, A. A., Klashinsky, S. & Willson, P. J. (1992b). Cloning and expression of a transferrin-binding protein from Actinobacillus pleuropneumoniae. Infect Immun 60, 892-898.[Abstract]

Gonzalez, G. C., Yu, R.-h., Rosteck, P. R.Jr & Schryvers, A. B. (1995). Sequence, genetic analysis, and expression of Actinobacillus pleuropneumoniae transferrin receptor genes. Microbiology 141, 2405-2416.[Abstract]

Günter, K. & Braun, V. (1990). In vivo evidence for FhuA outer membrane receptor interaction with TonB inner membrane protein of Escherichia coli. FEBS Lett 274, 85-88.[Medline]

Harlow, E. & Lane, D. (1999). Using Antibodies: a Laboratory Manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press.

Jacques, M., Rioux, S., Paradis, S.-E., Bégin, C. & Gottschalk, M. (1996). Identification of two core types in Actinobacillus pleuropneumoniae representing serotypes 1 to 12. Can J Microbiol 42, 855-858.[Medline]

Köster, W. & Braun, V. (1986). Iron hydroxamate transport of Escherichia coli: nucleotide sequence of the fhuB gene and identification of the protein. Mol Gen Genet 204, 435-442.[Medline]

Köster, W. & Braun, V. (1989). Iron-hydroxamate transport into Escherichia coli K12: localization of FhuD in the periplasm and of FhuB in the cytoplasmic membrane. Mol Gen Genet 217, 233-239.[Medline]

Kraulis, P. (1991). MOLSCRIPT: a program to produce both detailed and schematic plots of protein structures. J Appl Crystallogr 24, 946-950.

Linton, K. J. & Higgins, C. F. (1998). The Escherichia coli ATP-binding cassette (ABC) proteins. Mol Microbiol 28, 5-13.[Medline]

Locher, K. P., Rees, B., Koebnik, R., Mitschler, A., Moulinier, L., Rosenbusch, J. P. & Moras, D. (1998). Transmembrane signaling across the ligand-gated FhuA receptor: crystal structures of free and ferrichrome-bound states reveal allosteric changes. Cell 95, 771-778.[Medline]

Mademidis, A., Killmann, H., Kraas, W., Flechsler, I., Jung, G. & Braun, V. (1997). ATP-dependent ferric-hydroxamate transport system in Escherichia coli: periplasmic FhuD interacts with a periplasmic and with a transmembrane/cytoplasmic region of the integral membrane protein FhuB, as revealed by competitive peptide mapping. Mol Microbiol 26, 1109-1123.[Medline]

Manoil, C. & Beckwith, J. (1985). TnphoA: a transposon probe for protein export signals. Proc Natl Acad Sci USA 82, 8129-8133.[Abstract]

Martinez, J. L., Delgado-Iribarren, A. & Baquero, F. (1990). Mechanisms of iron acquisition and bacterial virulence. FEMS Microbiol Rev 75, 45-56.

Mickelson, P. A., Blackman, E. & Sparling, P. F. (1982). Ability of Neisseria gonorrhoeae, Neisseria meningitidis and commensal Neisseria species to obtain iron from lactoferrin. Infect Immun 35, 915-920.[Medline]

Mintz, K. P. & Fives-Taylor, P. M. (1999). Identification of genes coding for exported proteins of Actinobacillus actinomycetemcomitans. Infect Immun 67, 6217-6220.[Abstract/Free Full Text]

Moeck, G. S. & Coulton, J. W. (1998). TonB-dependent iron acquisition: mechanisms of siderophore-mediated active transport. Mol Microbiol 28, 675-681.[Medline]

Moeck, G. S., Ratcliffe, M. J. H. & Coulton, J. W. (1995). Topological analysis of the Escherichia coli ferrichrome-iron receptor by using monoclonal antibodies. J Bacteriol 177, 6118-6125.[Abstract]

Moeck, G. S., Tawa, P., Xiang, H., Ismail, A. A., Turnbull, J. L. & Coulton, J. W. (1996). Ligand-induced conformational change in the ferrichrome-iron receptor of Escherichia coli K-12. Mol Microbiol 22, 459-471.[Medline]

Moeck, G. S., Coulton, J. W. & Postle, K. (1997). Cell envelope signaling in Escherichia coli. Ligand binding to the ferrichrome-iron receptor FhuA promotes interaction with the energy-transducing protein TonB. J Biol Chem 272, 28391-28397.[Abstract/Free Full Text]

Nakai, K. & Kanehisa, M. (1991). Expert system for predicting protein localization sites in gram-negative bacteria. Proteins 11, 95-110.[Medline]

Neilands, J. B. (1995). Siderophores: structure and function of microbial iron transport compounds. J Biol Chem 270, 26723-26726.[Free Full Text]

Niven, D. F., Donga, J. & Archibald, F. S. (1989). Response of Haemophilus pleuropneumoniae to iron restriction; changes in the outer membrane protein profile and the removal of iron from porcine transferrin. Mol Microbiol 3, 1083-1089.[Medline]

Parkhill, J., Achtman, M., James, K. D. & 25 other authors (2000). Complete DNA sequence of a serogroup A strain of Neisseria meningitidis Z2491. Nature 404, 502–506.[Medline]

Pearce, B. J., Yin, Y. B. & Masure, H. R. (1993). Genetic identification of exported proteins in Streptococcus pneumoniae. Mol Microbiol 9, 1037-1050.[Medline]

Pitcher, D. G., Saunders, N. A. & Owen, R. J. (1989). Rapid extraction of bacterial genomic DNA with guanidium thiocyanate. Lett Appl Microbiol 8, 151-156.

Pohl, S., Bertschinger, H. U., Frederiksen, W. & Mannheim, W. (1983). Transfer of Haemophilus pleuropneumoniae and the Pasteurella haemolytica-like organism causing porcine necrotic pleuropneumonia to the genus Actinobacillus (Actinobacillus pleuropneumoniae comb. nov.) on the basis of phenotypic and deoxyribonucleic acid relatedness. Int J Syst Bacteriol 33, 510-514.

Postle, K. (1993). TonB protein and energy transduction between membranes. J Bioenerg Biomembr 25, 591-601.[Medline]

Schaller, A., Kuhn, R., Kuhnert, P., Nicolet, J., Anderson, T. J., MacInnes, J. I., Segers, R. P. A. M. & Frey, J. (1999). Characterization of apxIVA, a new RTX determinant of Actinobacillus pleuropneumoniae. Microbiology 145, 2105-2116.[Abstract]

Sirois, M., St-Arneault, A., Provencher, J., Perreault, N. & Gagnon, L. A. (2001). Identification of Actinobacillus pleuropneumoniae DNA sequences encoding exported proteins using PhoA. In Abstracts of the 101st General Meeting of the American Society for Microbiology 2001, abstract B-468, p. 146. Washington, DC: American Society for Microbiology.

Soltes, G. A. & MacInnes, J. I. (1994). Regulation of gene expression by the HlyX protein of Actinobacillus pleuropneumoniae. Microbiology 140, 839-845.[Abstract]

Tascón, R. I., Vàsquez-Boland, J. A., Gutierez-Martin, C. B., Rodriguez-Barbosa, J. I. & Rodriguez-Ferri, E. F. (1996). Virulence factors of the swine pathogen Actinobacillus pleuropneumoniae. Microbiologia 12, 171-184.[Medline]

Tonpitak, W., Thiede, S., Oswald, W., Baltes, N. & Gerlach, G. F. (2000). Actinobacillus pleuropneumoniae iron transport: a set of exbBD genes is transcriptionally linked to the tbpB gene and required for utilization of transferrin-bound iron. Infect Immun 68, 1164-1170.[Abstract/Free Full Text]

Vézina, G., Sirois, M., Clairoux, N. & Boissinot, M. (1997). Cloning and characterization of the groE locus from Actinobacillus pleuropneumoniae. FEMS Microbiol Lett 147, 11-16.[Medline]

West, S. E. & Sparling, P. F. (1985). Response of Neisseria gonorrhoeae to iron limitation: alterations in expression of membrane proteins without apparent siderophore production. Infect Immun 47, 388-394.[Medline]

Received 14 January 2002; revised 15 April 2002; accepted 21 May 2002.