Biometals (2014) 27:143–153 DOI 10.1007/s10534-013-9695-2

Demonstration of the functional role of conserved Glu-Arg residues in the Staphylococcus aureus ferrichrome transporter Enrique D. Vine´s • Craig D. Speziali David E. Heinrichs

•

Received: 3 December 2013 / Accepted: 12 December 2013 / Published online: 22 December 2013 Ó Springer Science+Business Media New York 2013

Abstract The features that govern the interaction of ligand binding proteins with membrane permeases of cognate ABC transporters are largely unknown. Using sequence alignments and structural modeling based on the structure of the Escherichia coli BtuCD vitamin B12 transporter, we identified six conserved basic residues in the permease, comprised of FhuB and FhuG proteins, in the ferrichrome transporter of Staphylococcus aureus. Using alanine-scanning mutagenesis we demonstrate that two of these residues, FhuB Arg-71 and FhuG Arg-61, play a more dominant role in transporter function than FhuB Arg-74 and Arg-311, and FhuG Arg-64 and Lys-306. Moreover, we show that at positions 71 and 61 in FhuB and FhuG, respectively, arginine cannot be substituted for lysine without loss of transporter function. Previously, our laboratory demonstrated the importance of conserved acidic residues in the ferrichrome binding protein, FhuD2. Taken together, these results support the hypothesis that Glu-Arg salt bridges are critical for the interaction of the ligand binding protein with the transmembrane domains FhuB and FhuG. This hypothesis was further studied by ‘‘charge swapping’’ E. D. Vine´s  C. D. Speziali  D. E. Heinrichs (&) Department of Microbiology and Immunology, University of Western Ontario, London, ON N6A 5C1, Canada e-mail: [email protected] D. E. Heinrichs The Centre for Human Immunology, University of Western Ontario, London, ON N6A 5C1, Canada

experiments whereby we constructed a S. aureus strain expressing FhuD2 with conserved residues Glu-97 and Glu-231 replaced by Arg and FhuB and FhuG with conserved basic residues Arg-71 and Arg-61, respectively, replaced by Glu. A strain containing this combination of substitutions restored partial function to the ferrichrome transporter. The results provide a direct demonstration of the functional importance of conserved basic residues on the extracellular surface of the ferrichrome permease in the Gram-positive bacterium S. aureus. Keywords ABC transporter  Iron-siderophore  Fhu  Protein interaction  Salt bridge

Introduction Iron is a critical micronutrient for virtually all microorganisms with only a few exceptions being described (Weinberg 1997; Posey and Gherardini 2000). Despite its abundance on Earth, the amount of iron available to biological systems is well below the concentration required to support growth, because of the extremely low solubility of ferric hydroxides formed under aerobic conditions and neutral pH (Chipperfield and Ratledge 2000). Microbes have, therefore, developed specialized means to acquire this element, including the production of small iron scavenging molecules termed siderophores that,

123

144

together with cognate ATP binding cassette transporters, provide an efficient means for bacteria to obtain iron from the environment (Neilands 1995; Winkelmann 2002; Hider and Kong 2010; Chu et al. 2010). ABC transporters allow for the accumulation of a solute against a concentration gradient by coupling the energy of ATP hydrolysis to the transport process. Typically, an ABC transporter is comprised of three functional units: (i) two transmembrane proteins, a homo- or heterodimer, that provide a channel across the membrane; (ii) two cytoplasmic ATP binding domains that provide the energy required for transport; and (iii) a solute binding protein that interacts with the transmembrane proteins to deliver the transport substrate (Higgins and Linton 2001; Hollenstein et al. 2007; Locher 2009). The Gram-positive bacterium Staphylococcus aureus is an important human pathogen and a leading cause of nosocomially-acquired infections. It is capable of causing a wide array of infections ranging from relatively mild (e.g. food poisoning, impetigo) to more severe (e.g. endocarditis, osteomyelitis, toxic shock syndrome) (Crossley and Archer 1997), and is becoming an increasing threat to health care due to the emergence of multidrug resistant strains both in the hospital and community setting (Chambers 2005; Diep and Otto 2008; Baldan et al. 2009). The S. aureus genome encodes several known and predicted ABC transporters involved in siderophore transport including sirABC (Heinrichs et al. 1999; Dale et al. 2004), sstABCD (Morrissey et al. 2000; Beasley et al. 2011), and the ferric hydroxamate uptake (fhu) system which is involved in the transport of ferrichrome, in addition to other hydroxamate-type siderophores (Beasley and Heinrichs 2010). The Fhu system is comprised of the fhuCBG operon encoding the components of an ABC transporter (FhuB and FhuG are transmembrane permeases and FhuC is an ATPase) (Sebulsky et al. 2000; Speziali et al. 2006) and the separately encoded fhuD1 and fhuD2, each coding for a ferric-hydroxamate siderophore binding lipoprotein (Sebulsky et al. 2003, 2004). FhuD2 is expressed in S. aureus-infected tissues and vaccination of mice with FhuD2 elicited protective immunity against S. aureus challenge in mouse models of infection (Mishra et al. 2012). Further investigations showed that both apoFhuD2 and FhuD2 complexed to hydroxamate siderophores were equally protective (Mariotti et al. 2013).

123

Biometals (2014) 27:143–153

The structure of the Escherichia coli BtuCD vitamin B12 ABC transporter (Locher et al. 2002) and the corresponding periplasmic binding protein BtuF (Borths et al. 2002; Karpowich et al. 2003) have been reported and, more recently, a structure of the BtuCD–BtuF complex was solved (Hvorup et al. 2007). Combined, the structures revealed negatively charged ‘‘knobs’’ on each lobe of BtuF and corresponding basic ‘‘pockets’’ at the periplasmic surface of BtuC, suggesting that salt bridges form between these regions of the proteins that aid in mediating an interaction between the solute binding protein and the ABC transporter. The structure of the complex revealed that BtuF residues Glu72 and Glu202 were interacting with Arg56 on each of the two BtuC proteins in the complex (Hvorup et al. 2007). These acidic and basic residues are conserved across ironsiderophore/heme/cobalamin transporters from various organisms. In the S. aureus ferrichrome-binding protein FhuD2, we previously demonstrated that alanine substitution of the residues in equivalent positions to BtuF Glu72 and Glu202 (Glu97 and Glu231 in FhuD2), resulted in a loss of transport function without loss of FhuD2 solute binding ability (Sebulsky et al. 2003), which is in agreement with crystal structure data showing their positions are at a distance from the substrate binding pocket (Mariotti et al. 2013). In this study, we have used alanine substitution mutagenesis to investigate the biological role of the six conserved basic residues in the FhuB and FhuG proteins in the S. aureus ferrichrome transporter. We demonstrate that mutation of two of the six residues, FhuB Arg-71 and FhuG Arg-61, had the most significant effect on ferrichrome transport in S. aureus, implicating them as the residues most likely to form salt bridges with FhuD2 Glu97 and Glu231.

Materials and methods Bacterial strains, plasmids, and growth conditions Escherichia coli DH5a (Promega) was used for all cloning procedures and was routinely grown in Luria– Bertani broth (Difco). Staphylococcus aureus RN4220 is a restriction deficient strain of S. aureus capable of accepting foreign DNA and fhuCBG or the combined fhuCBG fhuD1 fhuD2 null derivatives of this strain,

Biometals (2014) 27:143–153

H1068 (RN4220 fhuCBG::ermB) (Speziali et al. 2006) or H1110 (RN4220 fhuCBG::ermB fhuD2::Tet fhuD1::Km) (this study), were used to assess the impact of mutations in the FhuC2BG–FhuD2 transport system (expressed from pFhuCBG, pMTS37, and their derivatives, see below) on ferrichrome utilization. Phage transduction using phage 80 was used as described elsewhere (Novick 1991), to mobilize fhuD2::Tet and fhuD1::Km into the H1068 background, thus creating strain H1110, a strain devoid of all components of the ferric-hydroxamate transport system. S. aureus was routinely grown in tryptic soy broth (Difco), except for experiments assessing iron acquisition when Tris– minimal succinate (TMS) (Sebulsky et al. 2004) was used as the iron-restricted media. To further restrict the level of free iron available in TMS the chelator ethylene diamine-di(o-hydroxyphenol acetic acid) (EDDHA) was added as indicated. Where necessary, ampicillin (100 lg/mL) for the growth of E. coli or chloramphenicol (5 lg/mL), tetracycline (4 lg/mL) or kanamycin (50 lg/mL) for the growth of S. aureus was added to media. All incubations were carried out at 37 °C. Ironfree water for the preparation of media and solutions was obtained using a Milli-Q water filtration system (Millipore). The construction of pFhuCBG (Sebulsky et al. 2000) and pMTS37 (Sebulsky and Heinrichs 2001) have been described. Creation of site-directed mutations in fhuB, fhuG, and fhuD2 For all mutations, the QuikChangeTM (Stratagene) site-directed mutagenesis procedure was used as directed by the manufacturer. Plasmid pFhuCBG was used as a template to create mutations in fhuB and fhuG, and plasmid pMTS37 was used to create mutations in fhuD2. Plasmids harbouring multiple site-directed mutations were constructed by incorporation of multiple mutations within oligonucleotides or via sequential mutation reactions. All mutations were confirmed by sequencing before introduction, by electroporation, of mutated plasmids into either strain H1068 or H1110. Construction of fhuCBGFLAG Staphylococcus aureus RN6390 chromosomal DNA was used a template to amplify a fragment of the fhuCBG operon, including 680 bp of fhuB and the

145

complete fhuG gene, using primers fhuBG fwd (50 CATTTGCTTTAGCATTAACATATGCAGTTTTA CC-30 ) and fhuG-FLAG (50 -GTCAGATCTTTACT ATTTATCGTCGTCATCTTTGTAGTCTACATTT TTCGTTTTGTACA-30 ); primer fhuG-FLAG includes the sequence encoding for the FLAG epitope fused in frame with the 30 end of fhuG and a BglII site (underlined). Digestion of the PCR amplicon with XbaI, a recognition site present within 120 bp of the amplicon, allowed for the cloning of the fhuBGFLAG construct into the XbaI and EcoRV sites of pBAD24 (Guzman et al. 1995) to generate plasmid pEV62. Plasmid pEV62 was digested with HpaI, site present within fhuB, and BglII, and the 1.5-kb fragment containing the end of fhuB and the complete fhuG-FLAG was cloned into pFhuCBG digested with HpaI and BglII, site present 300 bp downstream of fhuG in pFhuCBG, generating plasmid pEV64 (fhuCBG-FLAG). In similar fashion, a plasmid expressing point mutations FhuB R71A and FhuG R61A, also tagged with a C-terminal FLAG epitope in FhuG, was constructed. Bacterial growth curves Growth curves were performed to assess ferrichrome utilization. Briefly, overnight cultures of the strains to be tested were grown in liquid TMS media, then diluted to an OD600 of 0.01 in fresh TMS media containing 0.2 lM EDDHA and supplemented with 5 lM ferrichrome. Cultures were then incubated at 37 °C in 100 well plates, 300 lL per well with medium shaking for 24 h and growth was monitored as OD600 every 30 min using a Bioscreen C automated microbiology reader (Growth Curves USA, Piscataway, NJ, USA). Ferrichrome transport assays Transport assays were performed as described (Sebulsky et al. 2000), with modifications. Briefly, S. aureus cells were grown overnight in TMS minimal medium, diluted to a 20-mL volume of the same medium supplemented with 0.2 lM EDDHA and grown to an OD600 of 1.0. Cells were harvested by filtering 5 mL of culture on 0.45-lm-pore-size membrane filters (SuporÒ-450, Pall Corp.), washed twice with 5 mL of 0.9 % NaCl, and resuspended in 5 mL of the same solution. Cells were incubated for 10 min at 37 °C

123

146

Biometals (2014) 27:143–153

with shaking before the assay. A mixture containing 500 lM ferrichrome, 16 lM 55FeCl3 and 2 lM nitrilotriacetic acid was prepared at least 30 min before the assay; 7.5 lL of this mixture was incubated with 750 lL of cells and 200 lL aliquots were removed at intervals, filtered onto 0.45-lm-pore-size membrane filters (SuporÒ-450, Pall Corp.), and washed twice with 10 ml of 100 mM LiCl. The membranes were allowed to dry and were counted in scintillation fluid by using the tritium channel of a scintillation system LS 6500 (Beckman). Molecular modeling of FhuC2BG A ClustalW alignment (Thompson et al. 1994) was performed to align the amino acid sequences of FhuB and FhuG with BtuC, the sequence of FhuC with BtuD, and the sequence of FhuD2 with BtuF. The sequences were submitted to Swiss-Model for optimization. The resulting models of FhuD2, FhuB, FhuG, and FhuC were aligned onto the structure of the BtuCDF complex (Hvorup et al. 2007), thus providing insight into possible interactions occurring within the FhuC2BG–FhuD2 transporter. Membrane preparation and detection of FhuD2 and FhuGFLAG proteins Strains expressing FhuCBGFLAG from plasmids pEV64 and pCS75FLAG were grown overnight at 37 °C and subcultured in 500 mL of TMS containing 200 lM dipyridyl. Growth was continued with aeration for 24 h. Cells were then collected by centrifugation and resuspended in 15 mL of 25 % sucrose in 25 mM HEPES pH 7.4; 10 lg/mL of lysostaphin were added to the cell suspension and the cells were incubated at 37 °C for 1 h and then lysed by several passages through a French press cell at 10,000 psi. Debris and unbroken cells were precipitated by centrifugation at 27,0009g for 15 min, and the clear lysate was layered on a 60 % (w/v) sucrose cushion (25 mM HEPES pH 7.4) followed by a 2 h centrifugation at 270,0009g. The membrane fraction, collected from the interface of the sucrose cushion, was mixed with protein tracking dye and proteins were separated in 12 % SDS-polyacrylamide gels. Proteins were transferred to a nitrocellulose membrane according to standard procedures, and the nitrocellulose membrane was incubated with the FLAG M2 monoclonal antibody (Sigma). Alexa Fluor 680 goat

123

Fig. 1 a Amino acid sequence alignment of ABC transmembrane domains (permeases) from E. coli and S. aureus members of the iron-siderophore/heme/cobalamin superfamily. Boxes indicate positions of conserved basic amino acids. b Ribbon diagrams of the molecular model of S. aureus FhuC2BG– FhuD2, highlighting the positions of the conserved acidic residues in FhuD2 and the conserved basic residues in FhuB and FhuG. Figure was generated using PyMol (DeLano Scientific)

anti-mouse IgG (Molecular Probes) was used as secondary antibody. Detection was performed by infrared imaging, using the Odyssey Infrared Imager (Li-Cor Biosciences).

Results Conserved basic residues are present in S. aureus FhuB and FhuG. A ClustalW alignment of E. coli and

Biometals (2014) 27:143–153

S. aureus members of the iron-siderophore/heme/ cobalamin superfamily of ABC transporters identified three basic residues at conserved locations in each protein (Fig. 1a). The alignment between BtuC and FhuB and FhuG from S. aureus showed identity of 28 % and an overall similarity of 50 %. For the S. aureus ferrichrome transporter FhuBG, the conserved basic residues are Arg71, Arg74 and Arg311 in FhuB, and Arg61, Arg64 and Lys306 in FhuG; these residues align with Arg56, Arg59, and Arg295 in BtuC and are also conserved across a much wider range of ABC transmembrane domains in Gram-negative bacteria, as described previously by Borths et al. (2002). The arrangement of these residues in the three dimensional structure of the BtuC transporter orients them with the negatively charged residues in the solute binding protein BtuF. Indeed, the crystal structure of the BtuF– BtuCD complex confirmed an interaction between Glu72 and Glu202 in BtuF with Arg56 in BtuC (Hvorup et al. 2007). The sequences of FhuD2, FhuB, FhuG and FhuC were threaded through the coordinates of the Btu complex structure to create the FhuC2BGD2 model depicted in Fig. 1b; this model allowed insight into the potential arrangement of the conserved basic residues in S. aureus FhuB and FhuG, indicating that residues Arg71, Arg74, and Arg311 in FhuB, and Arg61, Arg64, and Lys306 in FhuG, are arranged similarly to the basic pockets formed by Arg56, Arg59, and Arg295 in the BtuC homodimer. Alanine mutations in conserved positively charged residues in FhuB and FhuG negatively impact on ferrichrome transport. To test the biological importance of the conserved basic residues in FhuB and FhuG we used site-directed mutagenesis to make site-specific mutations in each of the residues. We used a fhuCBG deletion mutant, H1068, that, in contrast to its parent strain RN4220, is unable to use ferrichrome as a source of iron (Speziali et al. 2006). FhuC, FhuB and FhuG proteins, along with derivatives of FhuB and FhuG carrying Ala substitutions, were expressed in H1068 from plasmid pFhuCBG; pFhuCBG complements the ferrichrome transport defect of H1068 (Speziali et al. 2006). With mutations constructed, we assessed mutants for both ferrichrome transport and ferrichrome-dependent growth in iron-restricted media. Upon mutation to Ala of each of the six highly conserved basic residues in FhuBG, we observed that none resulted in significant impairment of bacterial growth in ferrichrome-containing media (Fig. 2a). However when the strain

147

Fig. 2 Effect of single alanine substitution in conserved basic amino acids on ferrichrome-dependent growth of S. aureus (a) and ferrichrome transport (b). Results are presented as mean ± SD of three replicates

expressing the same point mutations was analyzed for ferrichrome transport, all point mutations led to a noticeable defect in ferrichrome transport, causing a decrease in transport rate ranging from around 33 % for FhuG R64A, to over 95 % for FhuB R71A after 5 min of the transport experiment (Fig 2b). We next proceeded to combine individual Ala substitutions to generate FhuB and FhuG proteins that contained double mutation combinations. When strains expressing these constructs were tested as above, we observed that S. aureus expressing FhuB Arg71Ala Arg74Ala was significantly attenuated for ferrichrome-dependent growth (Fig. 3a) and, in agreement, this strain showed the largest defect in ferrichrome transport, with a decrease around 98 % compared to wild type at 5 min of transport (Fig. 3b).

123

148

Biometals (2014) 27:143–153

Fig. 3 Effect of double alanine substitution in conserved basic amino acids on ferrichrome-dependent growth of S. aureus (a) and ferrichrome transport (b). Results are presented as mean ± SD of three replicates

Fig. 4 Effect of simultaneous alanine substitution in FhuB R71 and FhuG R61, and FhuB R74 and R311 and FhuG R64 and K306 on ferrichrome-dependent growth (a) and ferrichrome transport (b). Results are presented as mean ± SD of three replicates

Staphylococcus aureus expressing FhuB Arg71Ala Arg311Ala showed a slight defect in growth (Fig. 3a), which correlated with a decrease in ferrichrome transport around a 92 % (Fig. 3b). Staphylococcus aureus expressing FhuB Arg74Ala Arg311Ala showed a 70 % decrease in ferrichrome transport (Fig. 3b), although it was not impaired for ferrichrome-dependent growth (Fig. 3a) suggesting that, of the three conserved basic residues in FhuB, Arg71 is most critical for transporter function. Double point mutant combinations in the FhuG protein did not seem to affect ferrichrome-dependent growth (Fig. 3a). However, as observed for the single alanine substitutions, and the FhuB double mutants, all combinations of double alanine substitutions in FhuG had some

impact on ferrichrome transport, ranging from 48 to 92 % reduction. The combination of Arg61Ala and Arg64Ala showed the most drastic inhibition of ferrichrome transport (Fig. 3b).

123

FhuB Arg71 and FhuG Arg61 are critical for ferrichrome transporter function Our preceding results suggested that, of the six conserved basic residues in the S. aureus ferrichrome transporter, FhuB Arg71 and FhuG Arg61 were the most critical for transporter function. According to our molecular modeling, the sidechains of FhuB Arg71 and FhuG Arg61 are situated closest to the surface of the respective proteins and therefore most likely to

Biometals (2014) 27:143–153

Fig. 5 FhuB R71 or FhuG R61 alone are not sufficient to mediate ferrichrome-dependent growth (a) or ferrichrome transport (b). Results are presented as mean ± SD of three replicates

interact with negative charges on the solute binding protein. In light of this, we next tested the H1068 strain expressing FhuB Arg71Ala FhuG Arg61Ala and observed that the strain was completely abrogated for growth on ferrichrome as a sole source of iron (Fig. 4a). The strain showed a 98 % reduction in ferrichrome-iron transport when compared to wild type (Fig. 4b). To further investigate their relative importance, we constructed a strain expressing FhuB with Ala substitutions in Arg74 and Arg311, in addition to Ala substitutions in Arg64 and Lys306 in FhuG (i.e. leaving FhuB Arg71 and FhuG Arg61 intact). This strain did not demonstrate any significant impairment in its ability to grow on ferrichrome as a sole iron source (Fig. 4a), yet did have an impairment in ferrichrome transport to about 14 % of that of the

149

wild type (Fig 4b). In agreement with this and preceding data with single and double mutants, a reduction in transport to even 10 % of wildtype is not enough to cause an observable defect in ferrichromedependent growth in our culture conditions. Taken together, these data show that of the six conserved basic residues in the S. aureus ferrichrome transporter, FhuB Arg71 and FhuG Arg61 together are sufficient for transporter function that will lead to growth using ferrichrome as a sole iron source. We next addressed the question of whether the transporter could function with only one out of the six basic residues unmutated. To do this we constructed Ala substitutions at five out of six positions, leaving only either FhuB Arg71 or FhuG Arg61 intact. In Fig. 5a, we show that both S. aureus H1068 expressing FhuB Arg74Ala Arg311Ala FhuG Arg61Ala Arg64Ala Lys306Ala (i.e. only FhuB Arg71 intact) and H1068 expressing FhuB Arg71Ala Arg74Ala Arg311Ala FhuG Arg64Ala Lys306Ala (i.e. only FhuG Arg61 intact) were severely impaired in ferrichrome-dependent growth. Accordingly, both strains showed a reduction in ferrichrome-iron transport to approximately 2 % of wildtype levels, indicating that FhuB Arg71 or FhuG Arg61 by themselves are not sufficient to mediate sufficient ferrichrome transport that would lead to appreciable growth on ferrichrome as the sole iron source. To rule out the possibility that mutations in two of the most critical basic residues could affect the expression and/or localization of the transporter in the membrane, we constructed a plasmid that contains the full fhuCBG operon, but that expresses FhuG fused to a FLAG epitope at its C-terminal end, as well as a plasmid that expresses point mutations FhuB Arg71Ala and FhuG Arg61Ala with a C-terminal FLAG epitope in FhuG. Anti-FLAG antibodies were used in Western blots to follow expression of the FhuGFLAG protein in membrane fractions of strains carrying each plasmid. The results demonstrate that FhuG is expressed in the membrane (Fig. 6, lane 1) and that Ala mutations in FhuB Arg71 and FhuG Arg 61, expressed from the same plasmid, do not alter the levels of FhuG expression in the membrane of these cells (Fig. 6, lane 2). FhuB Arg71 and FhuG Arg61 cannot be substituted for Lys without loss of transporter function. After demonstrating the dominant role played by FhuB Arg71 and FhuG Arg61 in ferrichrome transport, we

123

150

Biometals (2014) 27:143–153

Fig. 6 Western blot, using anti-FLAG antibody, of membrane preparations from S. aureus strains expressing wild type FhuB, FhuGFLAG (lane 1) and FhuB Arg71Ala FhuGFLAG Arg61Ala (lane 2). Molecular weight markers are shown at the left (lane M)

Fig. 8 Effect of ‘‘charge swapping’’ substitutions on growth of S. aureus on ferrichrome as a sole iron source (a) and ferrichrome transport (b). Results are presented as mean ± SD of three replicates

Fig. 7 Arginine cannot be substituted for lysine at positions 71 (FhuB) and 61 (FhuG) without loss of ferrichrome-dependent growth (a) or ferrichrome transporter function (b). Results are presented as mean ± SD of three replicates

were interested to study whether it was the charge or side chain length of these two residues that was the determining factor in the proper function of the transporter. In Fig. 7a, we demonstrate that H1068 expressing FhuB Arg71Lys and FhuG Arg61Lys is

123

severely attenuated for ferrichrome-dependent growth. Accordingly, the strain is also drastically impaired for ferrichrome transport, showing less than 2 % transport compared to that of the wild type strain (Fig 7b). Together, these results indicate that arginine residues, and not simply the positive charge, at these two positions are vital for transporter function. Charge swapping supports the importance of GluArg salt bridges between FhuD2 and FhuBG. On the basis of crystal structure data on the E. coli cobalamin transporter (Borths et al. 2002; Karpowich et al. 2003; Hvorup et al. 2007), as well as data from our previous work (Sebulsky et al. 2003) and this study, it is likely

Biometals (2014) 27:143–153

that Glu-Arg salt bridges mediate an interaction between FhuD2 (or FhuD1) and FhuBG. To address this further, we constructed a series of ‘‘charge swapped’’ S. aureus strains that expressed either (i) FhuD2 Glu97Arg, Glu231Arg (with wildtype FhuB and FhuG), (ii) FhuB Arg71Glu and FhuG Arg61Glu (with wildtype FhuD2), or (iii) FhuD2 Glu97Arg, Glu231Arg and FhuB Arg71Glu and FhuG Arg61Glu. Plasmids expressing the various mutants described in this section were introduced into strain H1110 (RN4220 fhuD1::Tet, fhuD2::Kan, fhuCBG::Erm). Figure 8a shows that strains expressing either FhuD2 Glu97Arg Glu231Arg in combination with wild type FhuB and FhuG, or wildtype FhuD2 with FhuBArg71Glu and FhuG Arg61Glu are unable to grow when provided ferrichrome as a sole iron source. In both cases, like charges are present on both the ferrichrome binding protein (FhuD2) and the permease (FhuBG). However, growth on ferrichrome is partially restored in the strain expressing the ‘swapped charges’ (i.e. Glu to Arg on FhuD2 in positions 97 and 231, and Arg to Glu on FhuB and FhuG in positions 71 and 61, respectively). As shown in Fig. 8b, strains expressing like charges in both binding protein and permease show very poor ferrichrome transport (i.e. less than 5 % of wildtype levels). The strain expressing the ‘swapped charges’, on the other hand, shows higher transport ability than the strains expressing the like charges in the transporter, in agreement with the bacterial growth data from Fig. 8a.

Discussion Previous work from our lab on the characterization of the ferric hydroxamate uptake system in S. aureus led us to identify residues Glu97 and Glu231 in the receptor lipoprotein FhuD2 that are critical for siderophore transport, without affecting binding of the substrate (Sebulsky et al. 2003). Here we provide more insight into the transport of ferrichrome, by identifying conserved basic residues in the ferric-hydroxamate permeases FhuB and FhuG of S. aureus. According to our molecular model of FhuC2BG that was based on the crystal structure of the BtuDC2 ABC transporter from E. coli (see Fig. 1b), these conserved residues come together to form a basic pocket at the surface of each permease protein. Conservation of these residues in permeases from ABC transporters in both Gram-

151

negative and Gram-positive bacteria suggests a generally-conserved mechanism of transport in this bacterial transporter superfamily. Along with the presence of conserved acidic residues in the receptor lipoprotein FhuD2 this finding lends support to the notion that salt bridges are formed between the ABC transporter binding protein and the permeases, as was suggested by studies on the E. coli ABC transporter BtuFC (Borths et al. 2002) and the FecBCD system in E. coli (Braun and Herrmann 2007). To the best of our knowledge, this study is the first to examine the relative importance of these residues in a transporter in the membrane of a Gram-positive bacterium. Our assays show that, although all six conserved basic residues are required for optimal functioning of the ferrichrome transporter, FhuB Arg71 and FhuG Arg61 seem to make a much greater contribution to the effectiveness of the transporter, as individual substitution of these residues by alanine causes larger decreases in transporter function. A critical role for FhuB Arg71 and FhuG Arg61 in ferrichrome transport was shown in several ways. First, the double mutant FhuB Arg71Ala FhuG Arg61Ala is virtually unable to transport ferrichrome and unable to grow using ferrichrome as sole source of iron. Second, a quadruple mutant FhuB Arg74Ala Arg311Ala FhuG Arg64Ala Lys306Ala (leaving only FhuB Arg71 and FhuG Arg61), although impaired for ferrichrome transport, shows no growth defect in our culture conditions. Third, FhuB Arg71 and FhuG Arg61 cannot be substituted for Lys without loss of transporter function. These results are in agreement with the structural model of FhuBG (based on the BtuC structure) that suggests that these two residues are situated closer to the surface of the FhuB and FhuG proteins, placing them in a more prominent position to interact with the acidic residues in FhuD2. Since the remaining four positively charged residues are somewhat dispensable, we speculate that residues FhuB Arg-74 and Arg-311, and FhuG Arg-64 and Lys-306 have a stabilizing effect on the interactions between FhuB Arg-71 and either, and FhuG Arg-61 and either FhuD2 Glu-97 or Glu-231. The final confirmation for the salt bridges mediating docking of FhuD2 and FhuBG in our system comes from our charge swapping experiment, where reversing of the charges found in FhuD2 and FhuBG leads to partial restoration of both ferrichrome transport and ferrichrome-dependent growth. Together, our results

123

152

lend significant support to the idea that Glu-Arg salt bridges are formed between FhuD2 and FhuBG during the iron-siderophore transport process. These results are in close agreement with the findings of Braun and Herrmann (Braun and Herrmann 2007) who showed that the interaction of the periplasmic citrate binding protein FecB with the permease FecCD in E. coli is also mediated by salt bridges. In that study, similar to our study here, the authors demonstrated that like charges in the conserved regions on the ligand binding protein and permease, which would be predicted to repel an interaction of the proteins, resulted in very poor uptake of iron-citrate. Moreover, they could show disulphide-bridge formation when the Glu residues in FecB were changed to Cys and one of the Arg residues in each of the ‘positive’ pockets of the permease were also changed to Cys. It is interesting that mutations in FhuB seem to, in general, have a greater impact on ferrichrome transport than mutations in FhuG, to the point that none of the mutations in FhuG presented in this study are able to completely impair ferrichrome dependent growth. Interestingly, we observed that when comparing ferrichrome-dependent growth of S. aureus in our culture conditions to the ferrichrome-iron transport assays, we found that growth was unaltered from wildtype-like growth when the transporter functioned to at least 10 % of the level of wildtype. This emphasizes the need to use different experimental methods to fully investigate the effect of mutations on transporter function. Acknowledgments This work was supported by an operating grant from the Canadian Institutes of Health Research (CIHR) to DEH. CDS was supported by a Natural Sciences and Engineering Research Council postgraduate scholarship.

References Baldan R, Tassan Din C, Semeraro G et al (2009) Severe community-onset infections in healthy individuals caused by community-acquired MRSA in an Italian teaching hospital, 2006–2008. J Hosp Infect 72:271–273 Beasley FC, Heinrichs DE (2010) Siderophore-mediated iron acquisition in the staphylococci. J Inorg Biochem 104:282–288 Beasley FC, Marolda CL, Cheung J et al (2011) Staphylococcus aureus transporters Hts, Sir, and Sst capture iron liberated from human transferrin by Staphyloferrin A, Staphyloferrin B, and catecholamine stress hormones, respectively, and contribute to virulence. Infect Immun 79:2345–2355

123

Biometals (2014) 27:143–153 Borths EL, Locher KP, Lee AT, Rees DC (2002) The structure of Escherichia coli BtuF and binding to its cognate ATP binding cassette transporter. Proc Natl Acad Sci USA 99:16642–16647 Braun V, Herrmann C (2007) Docking of the periplasmic FecB binding protein to the FecCD transmembrane proteins in the ferric citrate transport system of Escherichia coli. J Bacteriol 189:6913–6918 Chambers HF (2005) Community-associated MRSA: resistance and virulence converge. N Engl J Med 352:1485–1487 Chipperfield JR, Ratledge C (2000) Salicylic acid is not a bacterial siderophore: a theoretical study. Biometals Int J Role Met Ions Biol Biochem Med 13:165–168 Chu BC, Garcia-Herrero A, Johanson TH et al (2010) Siderophore uptake in bacteria and the battle for iron with the host; a bird’s eye view. Biometals 23:601–611 Crossley KB, Archer GL (1997) The staphylococci in human disease. Churchill Livingstone, New York Dale SE, Sebulsky MT, Heinrichs DE (2004) Involvement of SirABC in iron-siderophore import in Staphylococcus aureus. J Bacteriol 186:8356–8362 Diep BA, Otto M (2008) The role of virulence determinants in community-associated MRSA pathogenesis. Trends Microbiol 16:361–369 Guzman L-M, Belin D, Carson MJ, Beckwith J (1995) Tight regulation, modulation, and high-level expression by vectors containing the arabinose pBAD promoter. J Bacteriol 177:4121–4130 Heinrichs JH, Gatlin LE, Kunsch C et al (1999) Identification and characterization of SirA, an iron-regulated protein from Staphylococcus aureus. J Bacteriol 181:1436–1443 Hider RC, Kong X (2010) Chemistry and biology of siderophores. Nat Prod Rep 27:637–657 Higgins CF, Linton KJ (2001) Structural biology. The xyz of ABC transporters. Science 293:1782–1784 Hollenstein K, Frei DC, Locher KP (2007) Structure of an ABC transporter in complex with its binding protein. Nature 446:213–216 Hvorup RN, Goetz BA, Niederer M et al (2007) Asymmetry in the structure of the ABC transporter-binding protein complex BtuCD–BtuF. Science 317:1387–1390 Karpowich NK, Huang HH, Smith PC, Hunt JF (2003) Crystal structures of the BtuF periplasmic-binding protein for vitamin B12 suggest a functionally important reduction in protein mobility upon ligand binding. J Biol Chem 278:8429–8434 Locher KP (2009) Structure and mechanism of ATP-binding cassette transporters. Philos Trans R Soc B Biol Sci 364:239–245 Locher KP, Lee AT, Rees DC (2002) The E. coli BtuCD structure: a framework for ABC transporter architecture and mechanism. Science 296:1091–1098 Mariotti P, Malito E, Biancucci M et al (2013) Structural and functional characterization of the Staphylococcus aureus virulence factor and vaccine candidate FhuD2. Biochem J 449:683–693 Mishra RPN, Mariotti P, Fiaschi L et al (2012) Staphylococcus aureus FhuD2 is involved in the early phase of staphylococcal dissemination and generates protective immunity in mice. J Infect Dis 206:1041–1049 Morrissey JA, Cockayne A, Hill PJ, Williams P (2000) Molecular cloning and analysis of a putative siderophore

Biometals (2014) 27:143–153 ABC transporter from Staphylococcus aureus. Infect Immun 68:6281–6288 Neilands JB (1995) Siderophores: structure and function of microbial iron transport compounds. J Biol Chem 270:26723–26726 Novick RP (1991) Genetic systems in staphylococci. Methods Enzymol 204:587–636 Posey JE, Gherardini FC (2000) Lack of a role for iron in the Lyme disease pathogen. Science 288:1651–1653 Sebulsky MT, Heinrichs DE (2001) Identification and characterization of fhuD1 and fhuD2, two genes involved in ironhydroxamate uptake in Staphylococcus aureus. J Bacteriol 183:4994–5000 Sebulsky MT, Hohnstein D, Hunter MD, Heinrichs DE (2000) Identification and characterization of a membrane permease involved in iron-hydroxamate transport in Staphylococcus aureus. J Bacteriol 182:4394–4400 Sebulsky MT, Shilton BH, Speziali CD, Heinrichs DE (2003) The role of FhuD2 in iron(III)-hydroxamate transport in Staphylococcus aureus. Demonstration that FhuD2 binds iron(III)-hydroxamates but with minimal conformational

153 change and implication of mutations on transport. J Biol Chem 278:49890–49900 Sebulsky MT, Speziali CD, Shilton BH et al (2004) FhuD1, a ferric hydroxamate-binding lipoprotein in Staphylococcus aureus: a case of gene duplication and lateral transfer. J Biol Chem 279:53152–53159 Speziali CD, Dale SE, Henderson JA, Vines ED, Heinrichs DE (2006) Requirement of Staphylococcus aureus ATP-binding cassette ATPase FhuC for iron-restricted growth and evidence that it functions with more than one iron transporter. J Bacteriol 188:2048–2055 Thompson JD, Higgins DG, Gibson TJ (1994) CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res 22:4673–4680 Weinberg ED (1997) The Lactobacillus anomaly: total iron abstinence. Perspect Biol Med 40:578–583 Winkelmann G (2002) Microbial siderophore-mediated transport. Biochem Soc Trans 30:691–696

123