SGM Journals
Research Article

Characterization of the haem-uptake system of the equine pathogen Streptococcus equi subsp. equi

Mary Meehan, Fiona M. Burke, Susan Macken, Peter Owen

  • Department of Microbiology, Moyne Institute of Preventive Medicine, Trinity College Dublin, Dublin 2, Ireland
  • Correspondence
    Mary Meehan
    meehanmary{at}yahoo.co.uk

    • Microbiology 2010; 156(6):1824 · https://doi.org/10.1099/mic.0.036087-0

    View at publisher PubMed

    Abstract

    Streptococcus equi possesses a haem-uptake system homologous to that of Streptococcus pyogenes and Streptococcus zooepidemicus. The system consists of two ligand-binding proteins (Shr and Shp) and proteins (HtsA–C) with homology to an ABC transporter. The haem-uptake system of S. equi differs from that of S. pyogenes and S. zooepidemicus in that Shr is truncated by two-thirds. This study focused on the SeShr, SeShp and SeHtsA proteins of S. equi. Analysis of shr, shp and shphtsA knockout mutants showed that all three proteins were expressed in vitro and that expression was upregulated under conditions of iron limitation. SeShr possesses no membrane-/cell wall-spanning sequences and was shown to be secreted. Both SeShp and SeHtsA were confirmed to be envelope-associated. Recombinant SeShp and SeHtsA proteins have been previously shown to bind haem and SeHtsA could capture haem from SeShp. This report extends these studies and shows that both SeShp and SeHtsA can sequester haem from haemoglobin but not from haemoglobin–haptoglobin complexes. Like full-length Shr, SeShr possesses haemoglobin and haemoglobin–haptoglobin binding ability but unlike full-length Shr, it lacks haem- or fibronectin-binding capabilities. Analysis of SeShr truncates showed that residues within and upstream of the near transporter (NEAT) domain are required for this ligand binding. Structural predictions suggest that truncation of NEAT1 in SeShr accounts for its impaired ability to bind haem. Haem and haemoglobin restored to almost normal the impaired growth rates of wild-type S. equi cultured under iron-limiting conditions. However, no difference in the growth rates of wild-type and mutants could be detected under the in vitro growth conditions tested.

    • Supplementary material, describing detailed construction of shr, shp and htsA insertion mutations, a supplementary figure, showing schematic representations of the secondary structures of the NEAT domains of IsdC-haem, IsdH-NEAT3 and IsdA and the predicted structures of the NEAT1 domains of SzShr and SeShr modelled on the structures of the three Isd proteins utilizing the Phyre server, and two supplementary tables, listing strains and plasmids used to construct S. equi mutants and PCR primers used to construct S. equi mutants and recombinant proteins, are available with the online version of this paper.

    Edited by: H. Ingmer

    INTRODUCTION

    Limitation of iron is an important innate host defence mechanism and strategies to overcome this limitation and compete with the host are crucial for the survival of pathogenic bacteria. The majority of iron in the host is in the form of haem (Fe–protoporphyrin complex) bound by several host proteins, the principal one being haemoglobin (Hb). Not surprisingly, pathogenic bacteria have evolved numerous acquisition systems to take advantage of this iron-rich reservoir. Common to both Gram-positive and Gram-negative pathogens are ABC transporters, which transport haem through the plasma membrane. Gram-negative bacteria have outer membrane receptors that acquire haem from heamoproteins/haemophores and transfer it to the ABC transporters via the periplasm. Gram-positive bacteria possess receptors that transfer haem to the ABC transporters (reviewed by Genco & Dixon, 2001; Wandersman & Delepelaire, 2004; Wilks & Burkhard, 2007). In the Staphylococcus aureus Isd system, nine proteins (IsdA–I) constitute the capture and transport pathway (Mazmanian et al., 2003). IsdB and IsdH appear to remove haem from Hb and transfer it directly (or via IsdA) to IsdC. IsdC then transfers bound haem to the ABC transporter (IsdD–F). Once in the cytoplasm, haem is degraded by monooxygenases, IsdG and IsdI (Maresso & Schneewind, 2006; Muryoi et al., 2008; Reniere et al., 2007; Zhu et al., 2008b). The haem-acquisition machinery of Streptococcus pyogenes parallels that of Staph. aureus and consists of Shr which binds Hb, haemoglobin–haptoglobin (Hb–Hp) complexes and haem. Shr can transfer haem to a second protein, Shp which in turn transfers it to HtsA, the lipoprotein component of the ABC transporter (HtsA–C also termed SiaA–C; Bates et al., 2003; Lei et al., 2002, 2003; Zhu et al., 2008a). The four Isd receptor proteins, Shr and Shp possess one or more copies of the structurally conserved but sequence divergent near iron transporter (NEAT) domain (Andrade et al., 2002). Three-dimensional and other studies of several NEAT domains have shown that they possess a common β-sandwich immunoglobin-like fold (Aranda et al., 2007; Dryla et al., 2007; Grigg et al., 2007a; Pilpa et al., 2006; Sharp et al., 2007; Villareal et al., 2008; Watanabe et al., 2008) and can be categorized on the basis of their sequence similarities and ligand-binding profiles. Some iron-regulated NEAT-containing proteins such as IsdA and Shr are multifunctional and possess adhesive/survival functions in addition to haem acquisition (Clarke et al., 2004, 2009; Fisher et al., 2008).

    Streptococcus equi subsp. equi (herein referred to as S. equi) is the causative agent of strangles, one of the most feared diseases of horses. Strangles is a highly contagious disease of the upper respiratory tract characterized by abscessation of lymph nodes in the head and neck (Timoney, 1993, 2004). S. equi is considered to have evolved from S. equi subsp. zooepidemicus (herein referred to as S. zooepidemicus; Chanter et al., 1997), an opportunistic mucosal commensal which can cause serious diseases in a range of animals and, occasionally, humans (Timoney, 2004).

    Several putative virulence factors and protective antigens of S. equi have been identified. These include numerous ligand-binding surface-associated and secreted proteins, several superantigens and the hyaluronic acid capsule (Harrington et al., 2002; Holden et al., 2009; Timoney, 2004). Recently, S. equi homologues of the S. pyogenes Shp (designated SeShp) and HtsA (designated SeHtsA) have been described. Recombinant SeShp and SeHtsA proteins were shown to bind haem and SeHtsA to capture haem from SeShp (Nygaard et al., 2006b). Here, we demonstrate that S. equi expresses these two proteins and also a truncated Shr homologue (designated SeShr) under conditions of iron-limitation. We describe the cloning and overexpression in Escherichia coli of SeShr, SeShp and SeHtsA recombinant fusion proteins and use these proteins to extend studies of their ligand-binding capabilities.

    METHODS

    Growth conditions, bacterial strains and plasmids.

    S. equi was grown on Todd–Hewitt Yeast Extract (THYE) supplemented where appropriate with 100–600 μM 2,2′ dipyridyl (DIP), 10 mM NTA, 5 mM ferric citrate or streptonigrin (0.05 μM to 1.6 μM). THYE with NTA (THYE+NTA; Eichenbaum et al., 1996) was supplemented with 500 μM MgCl2, MnCl2 and CaCl2 and 125 μM ZnCl2. E. coli strains BL21 (DE3) and TB1, used for protein expression, were grown in Luria–Bertani (LB) broth or M9 medium. Bacterial strains and plasmids used in construction of S. equi mutants are shown in Supplementary Table S1, available with the online version of this paper.

    Expression and purification of recombinant SeShr, SeShp and SeHtsA proteins.

    Appropriate DNA fragments were amplified using specific primers (see Supplementary Table S2, available with the online version of this paper) and cloned into pET16b (Novagen), pGEX-KG (T. J. Foster, Trinity College Dublin) or pMAL-c2X (New England Biolabs) plasmids. His-tagged proteins, glutathione S-transferase (GST)-fusion proteins and MBP-fusion proteins were purified from E. coli by Ni-affinity chromatography, GSTrap FF 5 ml columns and amylase agarose (GE Healthcare and New England Biolabs), respectively. Proteins were dialysed against 20 mM Tris/HCl, pH 8.0, concentrated by ultrafiltration and stored at −70 °C. Where appropriate, His-tags were removed with Factor Xa.

    Construction of S. equi mutants.

    Utilizing the temperature-sensitive broad-host-range plasmid pG+host9, insertion mutants of shr, shp and htsA were generated by replacing S. equi DNA with the ΩKm2 interposon (Maguin et al., 1996; Meehan et al., 2001; Perez-Casal et al., 1991; Prentki & Krisch, 1984). For S. equi shr : : ΩΚm2, a central 127 bp fragment (nucleotides 619–745) was replaced. For S. equi shp : : ΩΚm2, the N-terminal 791 bp of shp and 232 bp upstream of shp was replaced and for S. equi shphtsA : : ΩΚm2, an additional 894 bp of htsA was replaced (see Supplementary Tables S1 and S2 and supplementary material for full methods).

    Biochemical procedures.

    UV-visible absorbance spectra of proteins in 20 mM Tris/HCl, pH 8.0, were recorded on a Shimadzu UV-160PC spectrophotometer. The haem content of proteins was estimated from reduced pyridine haemochrome spectra using an ϵ418.5 of 191.5 mM−1 cm−1 (Smith, 1975). Haem transfer was examined by recording the spectral shifts following co-incubation of proteins at 24 °C. Protein concentrations were estimated by a modification (Dulley & Grieve, 1975) of the Lowry method using BSA as a standard. Bacterial envelopes from S. equi strains, grown under iron-rich and iron-restricted conditions, were isolated as previously described and exhibited very similar protein profiles on SDS-PAGE gels to those previously observed (Meehan et al., 1998). For in vitro expression studies, equal protein loadings of envelope preparations were analysed by Western immunoblotting.

    In silico sequence and 3D homology modelling of NEAT1 domains of Shr, SeShr and SzShr.

    Amino acid sequence alignments and 3D modelling were performed following submission of NEAT1 domain sequences of Shr, SeShr and SzShr proteins to the Phyre webserver (Kelley & Sternberg, 2009).

    Hb–agarose binding experiments.

    Binding of SeShr and SzShr to Hb–agarose was performed as follows. Triton X-100-solubilized cell envelopes (150 μg), purified SeShr (20 ng) or culture supernatants (obtained from cultures normalized to OD600 0.8) in THYE were incubated for 1.5 h with 45 μl Hb–agarose (Sigma Aldrich) in 0.5 M NaCl and 0.05 % (v/v) Tween 20. The agarose was then washed in PBS (Dulbecco A; Oxoid) containing 0.05 % (v/v) Tween-20 (PBS-T), boiled in Laemmli sample buffer (Laemmli, 1970) and supernatants were analysed by Western immunoblotting. Haem capture from Hb agarose was performed by incubating 10 μM GST-SeShp, GST-SeHtsA (exhibiting 1.2–1.3 % haem saturation) or GST in PBS containing 0.5 M NaCl with Hb agarose at 37 °C for 2 h and then recording the UV-visible spectra of the supernatants.

    GST pull-down assays.

    Glutathione–agarose (10 μl; GE Healthcare) was incubated at room temperature for 45 min with 17 μM GST or GST fusion proteins, washed three times in PBS and finally resuspended in PBS-T. His-tagged or untagged proteins were then added to a final concentration of 0.8 μM and incubation continued for 2 h. Agarose beads were washed three times with PBS-T, resuspended in Laemmli sample buffer and analysed by SDS-PAGE and Western immunoblotting.

    ELISA.

    To examine Hb, Hp and Hb–Hp binding, 96-well plates were coated with recombinant SeShr, SeShp, SeHtsA and IsdH proteins (10 μg ml−1). Following blocking with 5 % (w/v) BSA, plates were probed with doubling dilutions of either human Hb, Hp (Sigma) or Hb–Hp complexes (prepared by mixing equimolar amounts of Hb and Hp for 1 h at room temperature). Molarities of haemoglobin were calculated from the molecular mass (34 000) of an α1β1 dimer. Ligand binding was detected using anti-Hb (1 : 2000; Biogenesis) or anti-Hp (1 : 20 000; Sigma) monoclonal antibodies followed by 1 : 2000 horseradish peroxidase (HRP)-labelled anti-mouse IgG (Sigma). Staphylococcal IsdH/HarA which binds Hb, Hp and Hb–Hp complexes was a gift from Dr Naoko Yanagisawa (Trinity College Dublin) and was used as positive control. Recombinant fibrinogen-binding protein (FgBP1) of S. equi was used as a negative control (Meehan et al., 1998).

    To examine binding to immobilized Hb, plates were coated with Hb (10 μg ml−1) and then probed with doubling dilutions of recombinant proteins. Binding was detected with anti-SeShr (1 : 80 000), anti-SeShp (1 : 4000) and anti-SeHtsA (1 : 16 000) followed by HRP-labelled goat anti-rabbit IgG (1 : 2000; MP Biomedicals). To examine binding to fibronectin (Fn), plates were coated overnight with GST-SeShr, GST or FnBPA476–838 of Staph. aureus (20 μg ml−1) followed by blocking with 5 % (w/v) dried skimmed milk and probing with doubling dilutions of Fn (100–0 μg ml−1). Binding was detected with rabbit anti-Fn antibodies (Sigma) followed by HRP-labelled goat anti-rabbit IgG. ELISA tests were developed using 3,3′ 5,5′-tetramethylbenzidine and A450 was measured. Dissociation constants (Kds) of binding curves were calculated using Graphpad Prism software.

    Western immunoblotting and affinity blotting experiments.

    Antibodies to His-tagged recombinant SeShr, SeShp and SeHtsA were raised and partially purified as described previously (Owen, 1985). For Western immunoblotting, membranes were blocked with 5 % (w/v) dried skimmed milk, probed with anti-SeShr, anti-SeHtsA or anti-SeShp as appropriate followed by incubation with HRP-labelled goat anti-rabbit IgG (Dako). For dot blots, doubling dilutions of purified proteins were dotted onto nitrocellulose and then probed with 63 nM Hb or Hb–Hp followed by either 1 : 5000 anti-Hp or 1 : 3000 anti-Hb mouse monoclonal antibodies and 1 : 2000 HRP-labelled anti-mouse IgG. Probes were detected by chemiluminescence using ChemiGLO (Cell Signalling Technology) and visualized by exposure to X-ray film.

    RESULTS

    S. equi haem-uptake system

    The S. equi and S. zooepidemicus genomes (Holden et al., 2009) possess a highly homologous 10 gene cluster, the first five genes of which resemble the Hts haem-uptake system of S. pyogenes (Fig. 1a⇓; Bates et al., 2003; Lei et al., 2002). shr, the first gene in the cluster, is truncated in S. equi compared with its S. pyogenes and S. zooepidemicus homologues due to a single base (A/T) pair deletion at position 1327 of the shr gene, which results in a frame-shift and a nonsense codon 54 bp downstream. Shr (Mr 142 704) in S. pyogenes and its homologue SzShr (Mr 138 516) in S. zooepidemicus are large proteins, both possessing a predicted N-terminal signal sequence followed by two NEAT domains encompassing a leucine-rich repeat (LRR) region and terminating in a stretch of 19 hydrophobic aa likely to be involved in membrane spanning followed by a short charged tail (Fig. 1b⇓). In contrast, the S. equi homologue (termed SeShr) possesses only the N-terminal 460 aa (Mr 51 439). Although there is strong homology (96 %) throughout SeShr and SzShr, the final 18 aa show considerable divergence in sequence due to the frame shift mutation. SeShr terminates in a cysteine residue and a truncated NEAT domain (Fig. 1c⇓). The NEAT domains of Shr, SzShr and SeShr exhibit strong homology. Thus, the NEAT1 domain of Shr exhibits 80 % and 82 % identity with SeShr (excluding the final divergent 18 aa) and SzShr, respectively. The NEAT2 domains of Shr and SzShr exhibit 73 % identity.

    Figure image not available in archive
    Fig. 1.

    Schematic representations of the haem-uptake gene cluster of S. equi and the SeShr protein and alignment of Shr and Isd NEAT domains. (a) The haem-uptake gene cluster of S. equi. The percentage identities between proteins encoded by the S. equi, S. pyogenes and S. zooepidemicus genes are shown. shr* indicates the pseudogene generated as a result of a base pair deletion (arrow). (b) Comparisons of Shr of S. pyogenes and SeShr of S. equi. The positions of the NEAT domains, leucine-rich repeat (LRR) region, signal sequence (S) and membrane-spanning region (M) are indicated. Below the diagram are the positions of recombinant SeShr proteins used in this study. (c) Alignment of the NEAT1 domains of SeShr (S. equi), SzShr (S. zooepidemicus) and Shr (S. pyogenes) with the NEAT domains of IsdA, IsdC and IsdH-NEAT3 using structural predictions from Phyre. Tyr and Ser residues conserved among all proteins are highlighted in grey. In the Isd proteins, residues in bold type have been shown to interact with haem. The predicted positions of the eight β sheets for SzShr are indicated by unbroken arrows. Broken arrows indicate the predicted positions of β6 and β7 of SeShr.

    Shr is followed by ORFs encoding the 293 residue haem-binding protein SeShp and three components of an ABC transporter, viz. a ligand-binding lipoprotein (SeHtsA), a permease (SeHtsB) and an ATPase (SeHtsC). The sequences of SeShp (residues 30–261) and SeHtsA (residues 21–294), determined in this study for strain TW, are identical to those determined for strain SEM1 (Nygaard et al., 2006b) and strain 4047 (Holden et al., 2009). The N-termini of Shp, SeShp and SzShp show most homology and are likely to possess the same immunoglobulin-like fold as that of Shp (Aranda et al., 2007; Holden et al., 2009; Nygaard et al., 2006b). The HtsA proteins exhibit strong (40 %) identity with the staphylococcal IsdE protein which has been shown to possess a bi-lobed structure characteristic of class III periplasmic binding proteins (Grigg et al., 2007b). Other genes within the cluster encode proteins with homology to permease/ATP-binding export proteins and proteins involved in cobalt transport (Bates et al., 2003).

    In silico sequence and structural homology modelling of SzShr and SeShr NEAT domains

    Shr binds a broad spectrum of ligands (Bates et al., 2003; Fisher et al., 2008; Zhu et al., 2008a). However, the binding domains of these ligands have not been localized. Sequence comparisons and in silico structural homology modelling was performed using the PHYRE website (Fig. 1c⇑ and Supplementary Fig. S1, available with the online version of this paper; Kelley & Sternberg, 2009) to determine the likely effect of the truncated NEAT on structure and function. Results showed that SzShr, SeShr and Shr modelled best (100 % precision) on the structures of the haem-binding proteins IsdA-NEAT, IsdC-NEAT and IsdH-NEAT3, producing a similar model regardless of the template protein; these were predicted to possess an immunoglobin-like β-sandwich fold (Fig. 1c⇑ and Supplementary Fig. S1). The predicted structural models of the two full-length Shr-NEAT1 domains showed that they possess several identical or similar residues to those that constitute the haem-binding pocket of the Isd proteins. These include the key metal-co-ordinating Tyr residue in β8 and a Ser residue in the 310 helix (linking β1b and β2), conserved in all Isd proteins, and hydrophobic resides in β4 and β7 (Fig. 1c⇑ and Supplementary Fig. S1). A lower precision was obtained when Shp (95 and 40 % precision for SzShr and SeShr, respectively) and IsdH-NEAT1 (80 and 15 % precision for SzShr and SeShr, respectively) were used a templates. Shp binds haem by bis-methionyl co-ordination of iron, contrasting with the Tyr co-ordination of the Isd proteins (Aranda et al., 2007). IsdH-NEAT1 binds Hb and Hb–Hp through an YYHFF motif in the β1–β2 loop, a motif not present in the haem-binding Isd proteins or Shr (Pilpa et al., 2009).

    Significantly, due to truncation and a lack of conservation in the final 18 aa, the region of SeShr encompassing β6 to β8 differs from that of Shr/SzShr. SeShr lacks all of the residues which constitute β8, including the co-ordinating Tyr. Furthermore, residues constituting β6 and β7 are predicted to be shifted in SeShr (see Fig. 1c⇑). Key Val/Ile residues which form the binding pocket are in the β7 sheet of the Isd proteins and are conserved in Shr/SzShr (Fig. 1c⇑). SeShr possesses a shorter β7 sheet containing two bulkier hydrophobic residues Trp and Phe (Fig. 1c⇑).

    These data suggest that the NEAT1 domains of SzShr/Shr are responsible for the haem-binding function and, given the significant difference in the C terminus of SeShr, that this protein has impaired haem-binding ability. Certainly, mutation of the co-ordinating Tyr resulted in almost complete abolition of haem binding in IsdA and significant spectral changes in IsdH-NEAT3 (Grigg et al., 2007a; Watanabe et al., 2008).

    Expression, purification and haem-binding properties of recombinant SeShr, SeShp and SeHtsA proteins

    Recombinant decaHis and GST fusion proteins, lacking predicted signal sequences and putative transmembrane region (SeShp only), were purified to homogeneity. SDS-PAGE analysis revealed that all purified recombinant proteins migrated at their predicted Mrs, except His-tagged SeShr which migrated predominantly as a dimer under non-reducing conditions (Fig. 2a⇓).

    Figure image not available in archive
    Fig. 2.

    SDS-PAGE of purified His-tagged recombinant proteins (a) and analysis of GST fusion proteins binding to SeShp during in vitro pull-down assays (b, c). (a) His10-tagged recombinant proteins SeShr (SeShr+, lanes 1 and 5), SeShr NEAT− (SeShr−, lanes 2 and 6), SeShp (lanes 3 and 7) and SeHtsA (lanes 4 and 8). Samples were boiled in Laemmli sample buffer with (+mer; lanes 1–4) or without (−mer; lanes 5–8) 2-mercaptoethanol. The positions of the His-tagged recombinant proteins are indicated to the left and the positions and molecular masses (in kDa) of marker proteins are indicated to the right. (b, c) GST-SeHtsA (lanes 1 and 4), GST-SeShr (lanes 2 and 5) or GST (lanes 3 and 6) were firstly immobilized on glutathione agarose. The agarose beads were then incubated with either control PBS (lanes 1–3) or with soluble His10-SeShp (lanes 4–6), washed, boiled in Laemmli sample buffer and analysed by SDS-PAGE (b). Bound SeShp was detected by Western immunoblotting using anti-SeShp antibodies (c). Similar results were obtained when SeShp without a histidine tag was used (data not shown). Lane 7, purified His10-SeShp. The positions of the GST fusion proteins are indicated to the left and that of His10-SeShp is indicated to the right.

    UV-visible spectral analysis revealed that SeShp and SeHtsA proteins possessed absorption spectra characteristic of haemoproteins with Soret peaks of 427 nm and 406 nm, respectively (Fig. 3a⇓; see also Nygaard et al., 2006b). The level of haem saturation, determined from spectra of their pyridine haemochromes, was 7–11 % and 1.2–1.3 % when purified from E. coli cultured in enriched LB medium and minimal medium, respectively. The difference probably reflects a reduced level of accumulated haem when grown in minimal media. In marked contrast with analysis of recombinant full-length Shr (Zhu et al., 2008a) and with the above proteins, E. coli cells expressing His-SeShr, GST-SeShr or GST were not red in colour and purified proteins did not exhibit spectra characteristic of haemoproteins, suggesting that SeShr binds negligible levels of haem (Fig. 3a⇓).

    Figure image not available in archive
    Fig. 3.

    UV-visible absorption spectra of purified GST-fusion proteins and the spectral shifts observed following incubation of GST-SeShp and GST-SeHtsA with Hb agarose, Hb and Hb–Hp complexes. (a) GST fusion proteins (25 μM; possessing 7–11 % haem) and GST. (b) Spectra of supernatants obtained following incubation of 10 μM GST, GST–SeShp and GST–SeHtsA with Hb agarose (Hb). (c, d) Spectra of purified Hb and spectral shifts obtained following an 8 h incubation of 25 μM GST-SeShp (c) and GST-SeHtsA (d) with soluble Hb or Hb–Hp. GST fusion proteins (possessing 1.2 % haem) purified from E. coli grown in minimal media were used for spectral shift experiments in (b)–(d).

    Ligand-binding properties of SeShr, SeShp and SeHts

    Both ELISA tests and affinity dot blots were performed to investigate the Hb-binding properties of the (truncated) SeShr (Fig. 4⇓). ELISA tests showed that SeShr could bind both Hb and Hb–Hp complexes (but not human Hp) in a dose-dependent manner and that Hb–Hp (Kd 0.02±0.004 μM) was bound more efficiently than Hb (Kd 0.11±0.04 μM) alone (Fig. 4a⇓). Similar ELISA tests showed that SeShp and SeHtsA bound soluble Hb (Kds of 2.3±0.3 μM and 1.5±0.5 μM, respectively) to a significantly weaker extent than SeShr (Fig. 4b⇓) and binding to Hb was inhibited by Hp (data not shown). Binding to Hb was not affected by mammalian source, since each bacterial protein bound immobilized human, equine, bovine and porcine Hb with equal efficiency (data not shown).

    Figure image not available in archive
    Fig. 4.

    ELISA and dot blot analysis of SeShp, SeHtsA, SeShr, GST and FnBPA476–838 proteins binding to Hb, Hp, Hp–Hb and Fn. Binding of Hb, Hp and Hb–Hp complexes to immobilized untagged SeShr (a) and binding of Hb and Hp to immobilized untagged SeShp and SeHtsA (b) was detected using anti-Hb and anti-Hp monoclonal antibodies. (c) Hb–Hp-affinity dot blot of recombinant SeShr, SeShr NEAT−, MBP–NEAT domain fusion (MBP+NEAT) and MBP alone. The first well in each row of doubling dilutions contains 180 pmol recombinant protein. (d) Binding of immobilized GST-SeShr, GST and FnBPA476–838 to solubilized Fn. Binding was detected using anti-Fn antibody. The means±sd of three experiments are shown.

    The above results provide evidence that the N-terminal region encompassing the NEAT1 domain of Shr proteins possesses Hb and Hb–Hp binding abilities. To investigate whether the (truncated) NEAT domain of SeShr was involved in Hb and/or Hb–Hp binding, a recombinant truncated SeShr protein (residues 19–358) lacking the predicted NEAT domain (termed Shr NEAT−) and a recombinant MBP–NEAT domain fusion protein were constructed. The construction of the latter was necessary since recombinant (untagged) NEAT domain (residues 359–460) was insoluble. Semiquantitative dot blot assays revealed that (i) deletion of the predicted NEAT domain reduced Hb and Hb–Hp binding by about 16-fold but did not totally eliminate binding and (ii) the MBP-NEAT fusion protein did not bind either ligand (Fig. 4c⇑). ELISA tests confirmed that SeShr NEAT− possesses reduced binding to Hb–Hp and Hb, and possessed Kds of 0.2±0.05 μM and 0.18±.03 μM for Hb and Hb–Hp, respectively. Together, these results suggest that residues both within and upstream of the NEAT domain of SeShr are required for full wild-type Hb binding.

    Fisher et al. (2008) recently showed that Shr additionally binds to Fn and laminin. To investigate the Fn-binding ability of SeShr, ELISA tests were performed. These tests showed that recombinant GST-SeShr did not bind to Fn significantly better than the non-Fn-binding GST protein. The control staphylococcal recombinant protein FnBPA476–838 exhibited significant Fn binding (Fig. 4d⇑; Fitzgerald et al., 2006).

    SeShp and SeHtsA can capture haem from Hb but not from Hb–Hp complexes

    Previous studies have shown that Shp of S. pyogenes can scavenge haem when incubated with Hb (Liu & Lei, 2005). The ability of HtsA to capture haem from Hb has not been reported to date. Two lines of experimentation were used to convincingly show that both SeShp and SeHtsA can capture haem from Hb. The first involved spectral analysis of supernatant fractions obtained following co-incubation of GST-fusion proteins and Hb agarose (Fig. 3b⇑). It should be noted that only a very small absorbance peak at 406 nm was observed for control supernatants when Hb agarose was incubated alone or with GST and that neither (apo) SeShp nor (apo) SeHtsA exhibits significant absorbance in the 400–600 nm range. Significantly, following incubation with Hb agarose, SeShp and SeHtsA gained major Soret peaks at 421 and 414.5, respectively, plus minor peaks at 530 – 560 nm (Fig. 3b⇑). Notably, however, unlike the ability of Shr to bind Hb-agarose (Fig. 6⇓), no binding of SeShp or SeHtsA to Hb agarose could be detected (data not shown). Analysis of spectral shifts obtained following co-incubation of soluble Hb with the GST fusion proteins for several hours gave similar spectral shifts to those obtained with the Hb agarose (compare Fig. 3b⇑ with Fig. 3c and d⇑). The 406 nm Soret peak of soluble Hb exhibited a significant shift to give major Soret peaks at 428 nm and 411 nm when incubated with SeShp or SeHtsA, respectively (Fig. 3c and d⇑). However, co-incubation of soluble Hb with either GST or GST-SeShr resulted in a Soret band that remained unaltered in position (406 nm) or intensity (data not shown).

    To extend these studies, the effect of Hp on haem transfer from Hb to either SeShp or SeHtsA was tested in spectral-shift assays utilizing Hb–Hp as haem donor. The results of these studies showed that in contrast with the significant spectral shifts observed following incubation with Hb, only a relatively slight spectral shift occurred (Fig. 3c and d⇑) when either SeShp or SeHtsA was incubated with Hb–Hp, suggesting that SeShp and SeHtsA can efficiently sequester haem from Hb but not from Hb–Hp complexes. The inability to bind to or scavenge haem from Hb–Hp, the slow rate of haem transfer from Hb to the SeShp and SeHtsA and the low efficiency of binding of either protein to Hb lends further support for the prediction (Zhu et al., 2008a) that haem transfer from Hb to Shp is not by direct acquisition but occurs indirectly following diffusion of haem from Hb.

    Recombinant Shr can acquire haem from E. coli haemoproteins and directly transfers bound haem to Shp (Zhu et al., 2008a). However, haem transfer from Hb to full-length Shr has yet to be documented. Given the ability of full-length Shr to bind haem, Hb and Hb–Hp and transfer bound haem to Shp, the proposed role for Shr in haem acquisition would be to accelerate the rate of haem transfer from Hb/Hb–Hp to Shp (Zhu et al., 2008a). Despite repeated efforts, no in vitro SeShr-dependent promotion of haem transfer could be detected following incubation of SeShr with SeShp or SeHtsA together with MetHb or Hb–Hp complexes (data not shown). Furthermore, pull-down assays confirmed, as expected, a direct interaction between SeShp and GST-SeHtsA. However, a similar interaction between SeShp and GST-SeShr or control GST protein was not evident (Fig. 2b and c⇑).

    SeShr, SeShp and SeHtsA are expressed in vitro and growth of iron-restricted S. equi is enhanced by haemoglobin

    To facilitate in vitro expression studies, insertion knockout mutants of shr, shp and shphtsA were generated. Western immunoblotting of cell envelopes showed that expression of SeShp and SeHtsA was significantly increased in S. equi grown under iron-limiting conditions (THYE supplemented with either DIP or NTA) when compared with S. equi grown under iron-rich conditions (Fig. 5a⇓, lanes 1, 4 and 7 and Fig. 5b⇓, lanes 1 and 4). Importantly, neither SeShp nor SeHtsA was expressed in shp : : ΩΚm2 or shphtsA : : ΩΚm2 mutants (Fig. 5a⇓, lanes 2–3, 5–6 and 8–9 and Fig. 5b⇓, lanes 2–3 and 5–6). These results thus validate the knockout mutations, show that SeShp and SeHtsA expression is iron-regulated in wild-type cells and also provide evidence that shp and htsA genes are co-transcribed, since inactivation of shp abolished expression not only of SeShp but also of SeHtsA (Fig. 5a⇓, lanes 3, 6 and 9 and Fig. 5b⇓, lanes 3 and 6). The S. equi shr mutant showed the same low level expression of SeHtsA when cultured in THYE as wild-type S. equi (data not shown) and SeHtsA expression was enhanced in the shr mutant grown under iron-limiting conditions (Fig. 5a⇓, lane 10).

    Figure image not available in archive
    Fig. 5.

    In vitro expression of SeHtsA and SeShp. (a) Cell envelopes (30 μg) prepared from S. equi wild-type (WT, lanes 1, 4 and 7), S. equi shp : : ΩKm (lanes 2, 5 and 8), S. equi shphtsA : : Ωkm (lanes 3, 6 and 9) and S. equi shr : : ΩKm (lane 10) grown in THYE (lanes 1–3), THYE+400 μM DIP (lanes 4–6) and THYE+NTA (lanes 7–10). Lane 11, purified His-tagged SeHtsA. (b) Cell envelopes (15 μg) were prepared from S. equi wild-type (lanes 1 and 4), S. equi shp : : Ωkm (lanes 2 and 5) and S. equi shphtsA : : Ωkm (lanes 3 and 6) grown in THYE (lanes 1–3) and THYE+NTA (lanes 4–6). Lane 7, purified His10-tagged SeShp. Samples were analysed by SDS-PAGE and Western immunoblotting using anti-SeHtsA (a) or anti-SeShp (b) antiserum. The positions of SeHtsA, His10-tagged SeHtsA, SeShp and His10-tagged SeShp are indicated.

    Shr/SzShr possesses a predicted membrane-spanning domain, and localization studies have recently shown that Shr is cell-membrane-associated and is also found in culture supernatants (Fisher et al., 2008; Bates et al., 2003). Sequence analysis of pre-SeShr indicates that the protein possesses an N-terminal signal sequence but, unlike pre-Shr/SzShr, that it lacks a C-terminal membrane-spanning sequence and is therefore predicted to be secreted. To confirm this prediction, expression of SeShr in envelope and supernatant fractions was assessed by a sensitive Hb-agarose pull-down assay (Fig. 6b⇓, lane 6). Results showed that (i) SeShr is present in culture supernatants (Fig. 6a⇓, lanes 5, 8, 11) but cannot be detected in cell envelopes (Fig. 6b⇓, lanes 3–4) of wild-type S. equi even at loadings 10 times higher than the equivalent loading of culture supernatants; (ii) SeShr is expressed at low levels following culture in iron-rich conditions but that expression increases following iron restriction (Fig. 6a⇓, compare lane 2 with lanes 5, 8 and 11); and (iii) SeShr is not expressed in shr : : ΩΚm2 (Fig. 6a⇓, lanes 3, 7, 10 and 13) but is expressed in the shphtsA double mutant (Fig. 6a⇓, lanes 6, 9 and 12). Sequence analysis predicts that SzShr of S. zooepidemicus is not truncated (Fig. 1c⇑) and that it possesses a membrane-spanning domain. In accordance with these predictions, a high-molecular-mass cross-reactive protein (apparent Mr 140 000) could be detected following Western immunoblotting analysis of detergent-extracted S. zooepidemicus cell envelopes but not in analogous S. equi cell envelope extracts (Fig. 6b⇓, compare lanes 1–2 with 3–4).

    Figure image not available in archive
    Fig. 6.

    In vitro expression of SeShr. (a) Purified His10-tagged SeShr (lane 1). Culture supernatants of S. equi wild-type (WT, lanes 2, 5, 8 and 11), S. equi shphtsA : : ΩKm (lanes 3, 6, 9 and 12) and S. equi shr : : ΩKm (lanes 4, 7, 10 and 13) grown overnight in THYE+ferric citrate (lanes 2–4), THYE (lanes 5–7), THYE+400 μM DIP (lanes 8–10) and THYE+NTA (lanes 11–13) were incubated with Hb–agarose. The position of His10-SeShr and the positions of SeShr and an SeShr degradation product [SeShr(d)] occasionally seen in S. equi grown in THYE, DIP and NTA are indicated. Recombinant SeShr proteins also exhibited some evidence of proteolytic degradation (see Figs 2 and 6b). (b) Triton X-100-extracted cell envelopes prepared from S. zooepidemicus (lanes 1 and 2) and S. equi (lanes 3 and 4) grown in either THYE (lanes 1 and 3) or THYE+NTA (lanes 2 and 4). Cell envelopes (lanes 1–4) and His-tagged SeShr (lane 6) were incubated with Hb agarose. Lane 5 contains purified His10-tagged SeShr only. Following incubation, Hb agarose beads were boiled in Laemmli sample buffer and analysed by SDS-PAGE and Western immunoblotting using anti-SeShr antibodies. The positions of His10-SeShr and SzShr from S. zooepidemicus are indicated.

    Growth of wild-type S. equi was significantly impaired when cultured under iron-limiting conditions (THYE+DIP) and showed a dose-dependent inhibition of growth in response to increasing concentrations of DIP (Fig. 7⇓). Growth was enhanced in iron-rich (THYE) broth or in THYE+DIP medium supplemented with Hb, BSA : : haem or whole blood (Fig. 7⇓). This provides evidence for the first time, to our knowledge, that S. equi can utilize haem and Hb as iron sources.

    Figure image not available in archive
    Fig. 7.

    Growth of wild-type S. equi under iron-rich and iron-limiting conditions. Wild-type S. equi was grown in THYE and subcultured into THYE, THYE supplemented with 500 μM DIP (THYE+DIP), THYE+DIP supplemented with 30 μM Hb (THYE+DIP+Hb) and THYE+DIP supplemented with 10 μM BSA : haem (THYE+DIP+BSA-Hm). THYE+DIP supplemented with 0.5 % sheep blood showed a similar result to THYE+DIP+Hb (data not shown). Growth curves [time (h) versus OD600] of S. equi grown in THYE and THYE supplemented with 100, 200, 400 and 600 μM DIP are inset. Error bars, sd.

    S. equi shphtsA : : ΩΚm2 showed a slight but consistent decrease in generation time (131±21 min) when cultured in THYE+DIP compared with wild-type S. equi (125±23 min). However, this difference is not statistically significant (P=0.136). Neither was there any significant difference between generation times of mutant and wild-type strains when grown in medium supplemented with Hb, BSA : : haem, whole blood or the iron-dependent antibiotic streptonigrin (data not shown; Yeowell & White, 1982).

    DISCUSSION

    S. equi, S. pyogenes and S. zooepidemicus each possess similar haem-uptake systems (Hts system) containing three ligand-binding proteins. Analyses of the S. equi homologues (SeShr, SeShp and SeHtsA) in this report have extended our overall understanding of this family of proteins. We show for the first time, to our knowledge, by using in vitro expression of SeShr, SeShp and SeHtsA, their iron-regulated expression, providing support for a role in haem/iron transport. The S. equi uptake system differs from the S. zooepidemicus and S. pyogenes systems by possessing a truncated Shr protein which is not envelope-associated, but which can be detected in culture supernatants. This correlates well with sequence data which indicate the presence of an N-terminal signal sequence directing export and the absence of any obvious membrane-/cell-wall-anchoring domain/motif. On this basis, it seems reasonable to conclude that SeShr is secreted.

    S. equi possesses a number of factors which would result in erythrocyte lysis and release of Hb and haem (Harrington et al., 2002; Holden et al., 2009; Timoney, 2004). We showed that S. equi can utilize haem and Hb as iron sources. The lack of significant difference between the growth of wild-type S. equi and mutants of the Hts system may be due to compensation by another iron-uptake systems(s). Indeed, the S. equi genome possesses homologues of the S. pyogenes MtsABC and SuiADEF systems and also expresses a potential siderophore termed equibactin (Hanks et al., 2005; Heather et al., 2008; Holden et al., 2009; Janulczyk et al., 2003).

    The Hts haem-uptake system shows remarkable similarity to the Isd system of Staph. aureus with Shr, Shp and HtsA probably serving roles in haem uptake similar to those of the IsdB/IsdH, IsdA/IsdC and IsdE proteins, respectively (Muryoi et al., 2008; Zhu et al., 2008b). The first important steps in haem acquisition are its direct capture from Hb or Hb–Hp and subsequent transfer to other proteins in the uptake pathway. Full-length Shr is proposed to share that role with the IsdH and IsdB proteins. Although there is no direct evidence as yet that Shr acquires haem from Hb, most available data support this prediction, since Shr binds Hb and Hb–Hp complexes, is predicted to possess a haem-binding pocket similar to that of the Isd proteins and was recently shown to transfer haem to Shp (Bates et al., 2003; Zhu et al., 2008a). However, there are clear differences between the two groups of proteins. The mode of Hb binding is distinct between the Isd and Shr proteins since Shr proteins do not possess the YYHFF motif required for Hb binding of IsdH and IsdB (Pilpa et al., 2009). This report shows that residues within and upstream of the NEAT1 domain are required for Hb binding of SeShr and contrasts with the situation for Isd proteins where purified NEAT domains possess ligand-binding functions. Furthermore, haem acquisition by IsdH and IsdB exploits the fact that Hb and haem bind to distinct NEAT domains on the same protein (Pilpa et al., 2009). However, for the Shr proteins, our ligand-binding experiments and structural modelling suggests that there is some overlap in Hb- and haem-binding regions.

    In the S. equi system, a role for SeShr in haem uptake remains unclear. The protein exhibits significant binding to both Hb and Hb–Hp complexes with greater binding to the latter ligand. However, lack of evidence that SeShr binds haem and facilitates haem transfer from Hb or Hb–Hp to SeShp indicates that SeShr is unlikely to have the same role as Shr/SzShr in haem uptake. SeShp and SeHtsA cannot acquire haem from Hb–Hp complexes. However, the ability to bind and acquire haem from Hb–Hp complexes is physiologically relevant, since MetHb released from lysed erythrocytes is quickly and irreversibly complexed with Hp. Hb–Hp complexes and free Hb are finally cleared from circulation following internalization by CD163 on the surface of macrophages/monocytes (Fabriek et al., 2005). Some pathogens, notably Neisseria meningitidis and Haemophilus influenzae, can scavenge haem as efficiently from Hb–Hp as from Hb alone (Wilks & Burkhard, 2007). A role in haem acquisition cannot be discounted for SeShr, and one potential role for the protein may be to compete with CD163 for Hb–Hp binding, thereby thwarting the clearance of Hb–Hp and permitting increased circulating Hb–Hp from which haem may eventually diffuse.

    Like IsdC, Shp is proposed to be the central conduit in the haem uptake pathway by facilitating haem transfer from Shr to HtsA (Liu & Lei, 2005; Muryoi et al., 2008; Nygaard et al., 2006a, b). Initial direct haem acquisition by Shr would be the most efficient means of acquiring haem from Hb. However, some haem acquisition proteins, notably IsdA and IsdC, can indirectly acquire haem following its diffusion from Hb (Zhu et al., 2008b). We have shown that this is the case for SeShp and SeHtsA. Under conditions of pathogen-induced haemolysis, when concentrations of Hb and haem would be higher than normal, it could be envisaged that SeShp and SeHtsA proteins may acquire enough haem needed to support growth in the absence of a fully functional Shr protein. A model for haem acquisition by S. equi and S .pyogenes/S. zooepidemicus is presented in Fig. 8⇓.

    Figure image not available in archive
    Fig. 8.

    Model for haem uptake in S. equi. Following erythrocyte destruction, possibly by S. equi haemolysin, higher levels of Hb and haem than normal are released. SeShp binds to free haem and/or captures haem from MetHb. Holo-SeShp interacts with and transfers haem to SeHtsA. SeHtsA can also bind haem or capture haem from Hb. Haem is then transported into the cytoplasm using SeHtsBC of the HtsABC transporter. Once released from erythrocytes, free Hb is rapidly complexed with Hp. SeShp or SeHtsA cannot efficiently sequester haem from Hb–Hp. SeShr, which is secreted into the extracellular medium, can bind both Hb and Hb–Hp complexes. However, a role for SeShr in haem transport is unclear at present. In contrast with SeShr, cell-anchored homologues in S. pyogenes and S. zooepidemicus (Shr and SzShr, respectively) can bind haem, Hb and Hb–Hp. Holo-Shr/SzShr transfers bound haem to Shp/SzShp, the proposed central conduit in the transport pathway.

    Like IsdA, full-length Shr is a broad spectrum iron-regulated adhesin and binds Fn and laminin and promotes bacterial cell adhesion (Fisher et al., 2008). We have shown that SeShr does not bind Fn. This strongly suggests that the Fn-binding region is located in the C-terminal two-thirds of Shr/SzShr and probably in the LRR or NEAT2 domains. S. equi binds less efficiently to Fn than S. zooepidemicus; a factor linked to the increased virulence of S. equi and almost certainly as a result of secretion of SeShr and another truncated Fn-binding protein, FNE (Holden et al., 2009; Lindmark et al., 2001). In contrast with SeShr, FNE retains the ability to bind Fn and has been suggested to compete with the Fn-binding mechanisms of competing pathogens (Holden et al., 2009).

    The Hts system in S. equi highlights common themes observed throughout the genomes of S. equi and S. zooepidemicus in showing evidence of considerable genetic exchange between these two organisms and S. pyogenes, and the presence of a large number of partially deleted/pseudo genes on the S. equi genome. These factors, together with the acquisition of genetic mobile elements, are thought to have facilitated the evolution of S. equi from S. zooepidemicus (Holden et al., 2009). A key event in the evolution of S. equi has been proposed to be the acquisition of an integrative conjugative element encoding the potential siderophore termed equibactin (Heather et al., 2008; Holden et al., 2009). Analysis of Staph. aureus mutants showed that haem acquisition satisfies iron requirements early in the infectious process and that siderophores play a vital role in iron acquisition, later in infection in niches that are low/devoid of haemoproteins (Dale et al., 2004; Skaar et al., 2004). Whether a similar scenario can be envisaged for S. equi remains to be determined. Certainly, the defined mutants in SeShr, SeShp and SeHtsA developed in this study will be invaluable in future studies to unravel the importance of haem/iron acquisition in the pathogenesis and abscess formation of S. equi and to provide insight into its evolution from the less virulent S. zooepidemicus.

    Acknowledgments

    This work was supported by research grant SC/2003/0196 from Enterprise Ireland.

    References

    1. Andrade, M. A., Ciccarelli, F. D., Perez-Iratxeta, C. & Bork, P. (2002). NEAT- a domain duplicated in genes near the components of a putative Fe3+ siderophore transporter from Gram-positive pathogenic bacteria. Genome Biol 3, RESEARCH0047
    2. Aranda, R., IV, Worley, C. E., Liu, M., Bitto, E., Cates, M. S., Olson, J. S., Lei, B. & Phillips, G. N., Jr (2007). Bis-methionyl coordination in the crystal structure of the heme-binding domain of the streptococcal cell surface protein Shp. J Mol Biol 374, 374–383.
    3. Bates, C. S., Montaňez, G. E., Woods, C. R., Vincent, R. M. & Eichenbaum, Z. (2003). Identification and characterization of a Streptococcus pyogenes operon involved in binding of hemoproteins and acquisition of iron. Infect Immun 71, 1042–1055.
    4. Chanter, N., Collins, N., Holmes, N., Binn, M. & Mumford, J. (1997). Characterization of the Lancefield group C streptococcus 16S–23S RNA gene intergenic spacer and its potential for identification and sub-specific typing. Epidemiol Infect 118, 125–135.
    5. Clarke, S. R., Witshire, M. D. & Foster, S. J. (2004). IsdA of Staphylococcus aureus is a broad spectrum, iron-regulated adhesin. Mol Microbiol 51, 1509–1519.
    6. Clarke, S. R., Andre, G., Walsh, E. J., Dufrene, Y. F., Foster, T. J. & Foster, S. J. (2009). Iron-regulated surface determinant protein A mediates adhesion of Staphylococcus aureus to human corneocyte envelope proteins. Infect Immun 77, 2408–2416.
    7. Dale, S. E., Doherty-Kirby, A., Lajoie, G. & Heinrichs, D. E. (2004). Role of siderophore biosynthesis in virulence of Staphylococcus aureus: identification and characterization of genes involved in production of a siderophore. Infect Immun 72, 29–37.
    8. Dryla, A., Hoffmann, B., Gelbmann, D., Giefing, C., Hanner, M., Meinke, A., Anderson, A. S., Koppensteiner, W., Konrat, R. & other authors (2007). High-affinity binding of the Staphylococcal HarA protein to haptoglobin and haemoglobin involves a domain with an antiparallel eight-stranded β-barrel fold. J Bacteriol 189, 254–264.
    9. Dulley, J. R. & Grieve, P. A. (1975). A simple technique for eliminating interference by detergents in the Lowry method of protein determination. Anal Biochem 64, 136–141.
    10. Eichenbaum, Z., Muller, E., Morse, S. A. & Scott, J. R. (1996). Acquisition of iron from host proteins by group A streptococcus. Infect Immun 64, 5428–5429.
    11. Fabriek, B. O., Dijkstra, C. D. & van den Berg, T. K. (2005). The macrophage scavenger receptor CD163. Immunobiology 210, 153–160.
    12. Fisher, M., Huang, Y.-S., Li, X., McIver, K. S., Toukoki, C. & Eichenbaum, Z. (2008). Shr is a broad-spectrum surface receptor that contributes to adherence and virulence in group A streptococcus. Infect Immun 76, 5006–5015.
    13. Fitzgerald, J. R., Loughman, A., Keane, F., Brennan, M., Knobel, M., Higgins, J., Visai, L., Speziale, P., Cox, D. & Foster, T. J. (2006). Fibronectin-binding proteins of Staphylococcus aureus mediate activation of human platelets via fibrinogen and fibronectin bridges to integrin GPIIb/IIIa and IgG binding to the FcγRIIa receptor. Mol Microbiol 59, 212–230.
    14. Genco, C. A. & Dixon, D. W. (2001). Emerging strategies in microbial haem capture. Mol Microbiol 39, 1–11.
    15. Grigg, J. C., Vermeiren, C. L., Heindrichs, D. E. & Murphy, M. E. P. (2007a). Haem recognition by a Staphylococcus aureus NEAT domain. Mol Microbiol 63, 139–149.
    16. Grigg, J. C., Vermeiren, C. L., Heindrichs, D. E. & Murphy, M. E. P. (2007b). Heme coordination by Staphylococcus aureus IsdE. J Biol Chem 282, 28815–28822.
    17. Hanks, T. S., Liu, M., McClure, M. J. & Lei, B. (2005). ABC transporter FtsABCD of Streptooccus pyogenes mediates uptake of ferric ferrichrome. BMC Microbiol 5, 62
    18. Harrington, D. J., Sutcliff, I. C. & Chanter, N. (2002). The molecular basis of Streptococcus equi infection and disease. Microbes Infect 4, 501–510.
    19. Heather, Z., Holden, M. T., Steward, K. F., Parkhill, J., Song, L., Challis, G. L., Robsinson, C., Davis-Poynter, N. & Waller, A. S. (2008). A novel streptococcal conjugative element involved in iron acquisition. Mol Microbiol 70, 1274–1292.
    20. Holden, M. T. G., Heather, Z., Paillot, R., Steward, K. F., Webb, K., Ainslie, F., Jourdan, T., Bason, N. C., Holroyd, N. E. & other authors (2009). Genomic evidence for the evolution of Streptococcus equi: restriction, increased virulence, and genetic exchange with human pathogens. PLoS Pathog 5, e1000346
    21. Janulczyk, R., Ricci, S. & Björck, L. (2003). MtsABC is important for manganese and iron transport, oxidative stress resistance, and virulence of Streptococcus pyogenes. Infect Immun 71, 2656–2664.
    22. Kelley, L. A. & Sternberg, M. J. E. (2009). Protein structure prediction on the web: a case study using the Phyre server. Nat Protoc 4, 363–371.
    23. Laemmli, U. K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680–685.
    24. Lei, B., Smoot, L. M., Menning, H. M., Voyich, J. M., Kala, S. V., Deleo, F. R., Reid, S. D. & Musser, J. M. (2002). Identification and characterization of a novel heme-associated cell surface protein made by Streptococcus pyogenes. Infect Immun 70, 4494–4500.
    25. Lei, B., Liu, M. L., Voyich, J. M., Prater, C. I., Kala, S. V., DeLeo, F. R. & Musser, J. M. (2003). Identification and characterization of HtsA, a second heme-binding protein made by Streptococcus pyogenes. Infect Immun 71, 5962–5969.
    26. Lindmark, H., Nilsson, M. & Guss, B. (2001). Comparison of the fibronectin-binding protein FNE from Streptococcus equi subspecies equi with FNZ from S. equi subspecies zooepidemicus reveals a major and conserved difference. Infect Immun 69, 3159–3163.
    27. Liu, M. & Lei, B. (2005). Heme transfer from streptococcal cell surface protein Shp to HtsA of transporter HtsABC. Infect Immun 73, 5086–5092.
    28. Maguin, E., Prévost, H., Ehrlich, S. D. & Gruss, A. (1996). Efficient insertional mutagenesis in lactococci and other Gram-positive bacteria. J Bacteriol 178, 931–935.
    29. Maresso, A. W. & Schneewind, O. (2006). Iron acquisition and transport in Staphylococcus aureus. Biometals 19, 193–203.
    30. Mazmanian, S. K., Skaar, E. P., Gaspar, A. H., Humayun, M., Gornicki, P., Jelenska, J., Joachmiak, A., Missiakas, D. M. & Schneewind, O. (2003). Passage of heme–iron across the envelope of Staphylococcus aureus. Science 299, 906–909.
    31. Meehan, M., Nowlan, P. & Owen, P. (1998). Affinity purification and characterization of a fibrinogen-binding protein complex which protects mice against lethal challenge with Streptococcus equi subsp. equi. Microbiology 144, 993–1003.
    32. Meehan, M., Lynagh, Y., Woods, C. & Owen, P. (2001). The fibrinogen-binding protein (FgBP) of Streptococcus equi subsp. equi additionally binds IgG and contributes to virulence in a mouse model. Microbiology 147, 3311–3322.
    33. Muryoi, N., Tiedemann, M. T., Pluym, M., Cheung, J., Heinrichs, D. E. & Stillman, M. J. (2008). Demonstration of the iron-regulated surface determinant (Isd) heme transfer pathway in Stapylococcus aureus. J Biol Chem 283, 28125–28136.
    34. Nygaard, T. K., Blouin, G. C., Liu, M., Fukumura, M., Olson, J. S., Fabian, M., Dooley, D. M. & Lei, B. (2006a). The mechanism of direct heme transfer from streptococcal cell surface protein Shp to HtsA of the HTsABC transporter. J Biol Chem 281, 20761–20771.
    35. Nygaard, T. K., Liu, M., McClure, M. J. & Lei, B. (2006b). Identification and characterization of the heme-binding proteins SeShp and SeHtsA of Streptococcus equi subsp. equi. BMC Microbiol 6, 82
    36. Owen, P. (1985). Crossed immunoelectrophoresis in the study of outer membrane antigens. In Enterobacterial Surface Antigens: Methods for Molecular Characterization, pp. 207–242. Edited by T. K. K. Korhonen, E. A. Dawes & P. H. Makela. Amsterdam: Elsevier
    37. Perez-Casal, J., Caparon, M. G. & Scott, J. R. (1991). Mry, a trans-acting positive regulator of the M protein gene of Streptococcus pyogenes with similarity to the receptor proteins of two-component regulatory systems. J Bacteriol 173, 2617–2624.
    38. Pilpa, R. M., Fadeev, E. A., Villareal, V. A., Wong, M. L., Phillips, M. & Clubb, R. T. (2006). Solution structure of the NEAT (NEAr Transporter) domain from IdH/HarA: the human haemoglobin receptor in Staphylococcus aureus. J Mol Biol 360, 435–447.
    39. Pilpa, R. M., Robson, S. A., Villareal, V. A., Wong, M. L., Phillips, M. & Clubb, R. T. (2009). Functionally distinct NEAT (NEAr Transporter) domains within the Staphylococcus aureus IsdH/HarA protein extract heme from methemoglobin. J Biol Chem 284, 1166–1176.
    40. Prentki, P. & Krisch, H. M. (1984). In vitro insertional mutagenesis with a selectable DNA fragment. Gene 29, 303–313.
    41. Reniere, M. L., Torres, V. J. & Skaar, E. P. (2007). Intracellular metalloporphyrin metabolism in Staphylococcus aureus. Biometals 20, 333–345.
    42. Sharp, K. H., Schneider, S., Cockayne, A. & Paoli, M. (2007). Crystal structure of the heme–IsdC complex, the central conduit of the Isd iron/heme uptake system in Staphylococcus aureus. J Biol Chem 282, 10625–10631.
    43. Skaar, E. P., Humayun, M., Bae, T., DeBord, K. L. & Schneewind, O. (2004). Iron-source preference of Staphylococcus aureus infections. Science 305, 1626–1628.
    44. Smith, K. M. (1975). Porphyrins and Metalloporphyrins. Amsterdam: Elsevier.
    45. Timoney, J. F. (1993). Strangles. Vet Clin North Am Equine Pract 9, 365–374.
    46. Timoney, J. F. (2004). The pathogenic equine streptococci. Vet Res 35, 397–409.
    47. Villareal, V. A., Pilpa, R. M., Robson, S. A., Fadeev, E. A. & Clubb, R. T. (2008). The IsdC protein from Staphylococcus aureus uses a flexible binding pocket to capture heme. J Biol Chem 283, 31591–31600.
    48. Wandersman, C. & Delepelaire, P. (2004). Bacterial iron sources: from siderophores to hemophores. Annu Rev Microbiol 58, 611–647.
    49. Watanabe, M., Tanaka, Y., Suenaga, A., Kuroda, M., Yao, M., Watanabe, N., Arisaka, F., Ohta, T., Tanaka, I. & Tsumoto, K. (2008). Structural basis for multimeric heme complexation through a specific protein–heme interaction; the case of the third neat domain of IsdH from Staphylococcus aureus. J Biol Chem 283, 28649–28659.
    50. Wilks, A. & Burkhard, K. A. (2007). Heme and virulence: how bacterial pathogens regulate, transport and utilize heme. Nat Prod Rep 24, 511–522.
    51. Yeowell, H. N. & White, J. R. (1982). Iron requirement in the bactericidal mechanism of streptonigrin. Antimicrob Agents Chemother 22, 961–968.
    52. Zhu, H., Liu, M. & Lei, B. (2008a). The surface protein Shr of Streptococcus pyogenes binds heme and transfers it to the streptococcal heme-binding protein Shp. BMC Microbiol 8, 15
    53. Zhu, H., Xie, G., Liu, M., Olson, J. S., Fabian, M., Dooley, D. M. & Lei, B. (2008b). Pathway for heme uptake from human methemoglobin by the iron-regulated surface determinants system of Staphylococcus aureus. J Biol Chem 283, 18450–18460.