|Home | About | Journals | Submit | Contact Us | Français|
Genetic studies in the tuberculosis mouse model have suggested that mycobacterial metal efflux systems, such as the P1B4-ATPase CtpD, are important for pathogenesis. The specificity for substrate metals largely determines the function of these ATPases; however, various substrates have been reported for bacterial and plant P1B4-ATPases leaving their function uncertain. Here we describe the functional role of the CtpD protein of Mycobacterium smegmatis. An M. smegmatis mutant strain lacking the ctpD gene was hypersensitive to Co2+ and Ni2+ and accumulated these metals in the cytoplasm. ctpD transcription was induced by both Co2+ and superoxide stress. Biochemical characterization of heterologously expressed, affinity purified CtpD showed that this ATPase is activated by Co2+, Ni2+ and to a lesser extend Zn2+ (20% of maximum activity). The protein was also able to bind one Co2+, Ni2+ or Zn2+ to its transmembrane transport site. These observations indicate that CtpD is important for Co2+ and Ni2+ homeostasis in M. smegmatis, and that M. tuberculosis CtpD ortholog could be involved in metal detoxification and resisting cellular oxidative stress by modulating the intracellular concentration of these metals.
Mycobacterium tuberculosis (Mtb) is the causative agent of tuberculosis (Harries & Dye, 2006). Considering the emergence of drug resistant strains and the consequent need or novel therapeutic targets, it is critical to better understand the physiology of this pathogen. The maintenance of metal homeostasis is one of the fundamental challenges that pathogens face during host infection. Like all organisms, Mycobacteria depend on transition metals like Co2+, Ni2+, Mn2+, Fe2+, Zn2+ and Cu+ to maintain cellular functions such as respiration, redox homeostasis and transcription (Fraústro da Silva & Williams, 2001). At the same time, intracellular excess of these ions can be very toxic and represents a major threat for cell survival inside the phagosomal compartment of the macrophage (Wagner et al., 2005; White et al., 2009; Botella et al., 2011). High concentrations of metal ions in the phagosomal compartment have the potential to generate damaging free radicals that can react with lipids, proteins and DNA. Alternatively, excess of metal ions can displace by competition, required metals at enzyme catalytic sites (Ranquet et al., 2007; Osman & Cavet, 2008).
Bacteria have evolved finely tuned mechanisms to preserve transition metal homeostasis. Among these, P1B-typeATPases1 are transporters critical for controlling cytoplasmic metal levels and delivering metals to nascent metalloproteins found in the plasma membrane and periplasm (Raimunda et al., 2011; Argüello et al., 2011). The presence of a high number of genes encoding P1B-ATPases in pathogenic/symbiotic bacterial organisms suggests an important role for these transporters in virulence (Argüello, 2003; Argüello et al., 2011). P1B-ATPases are distinguished by a homologous core protein structure and common mechanism of transport (Argüello, 2003; Argüello et al., 2011). These proteins are found in all kingdoms of life and are divided into phylogenetically distinct subgroups. The members of each phylogenetic cluster share similar metal specificity and conserved sets of metal coordinating amino acids in three transmembrane segments (TMs) flanking the ATP binding and hydrolysis domains (Argüello, 2003) (Fig. 1A).
Genetic studies have suggested that a P1B4-ATPase, ctpD, is important for M. tuberculosis growth in the mouse model (Sassetti & Rubin, 2003; Joshi et al., 2006). The P1B4-subgroup includes proteins only present in bacteria and plants. Interestingly, M. tuberculosis and some other members of its genus (M. bovis, M. vanbaalenii, M. gilvum) are among the few organisms that possess two genes encoding for these ATPases (Table S1). Their function in cell physiology and their contribution to virulence is not understood. While it is clear that the specificity for substrate metals is a determinant factor of their function, the experimental evidence is quite conflicting in this regard.
Early genetic studies of the cyanobacteria Synechocystis PCC 6803 showed that a mutant lacking the single P1B4-ATPase coding gene, coaT, led to both Co2+ sensitivity and the accumulation of this metal in the cytoplasm (Rutherford et al., 1999). Thus, a function as a Co2+ transporter was proposed for this protein. Genetic and biochemical characterization of Cupriavidus metallidurans CzcP suggested a secondary role in Zn2+ homeostasis, perhaps masked by the presence of three Zn2+ transporting P1B2-ATPases in this organism (Scherer & Nies, 2009). In vitro, CzcP showed higher transport rates for Cd2+ and Zn2+ than for Co2+. However, the differential transcriptional regulation of various transporters by Zn2+ may render CzcP nonessential for Zn2+-homeostasis in vivo (Scherer & Nies, 2009). Studies of metal accumulation by Bacilus subtilis zosA mutant strains suggested yet an alternative role as Zn2+ importer for this P1B4-ATPase (Gaballa & Helmann, 2002). zosA is part of the perR regulon, as its expression is induced by H2O2. This, together with H2O2 and diamide sensitivity shown by the zosA mutant strain, suggested a putative role of P1B4-ATPases in defense against reactive oxygen species (ROS).
It has been proposed that Arabidopsis thaliana HMA1, also a P1B4-ATPase located in the chloroplast membrane, is involved in supplying Cu+/2+ to the chloroplast Cu/Zn-Superoxide dismutase (Seigneurin-Berny et al., 2006). Metal sensitivity and accumulation assays in AtHMA1 expressing yeast, as well as ATPase activity determinations in purified chloroplast envelope membranes overexpressing HMA1, support the idea that HMA1 is a Cu-ATPase. However, based on yeast mutants functional complementation and Ca2+ transport assays, Moreno et al. have proposed that Ca2+ is the main substrate transported by HMA1 (Moreno et al., 2008). In parallel determinations, Cd2+ and to a lesser extent Ca2+, Co2+, Cu+ and Zn2+ stimulated the associated ATPase activity. Finally, Kim et al. described a Zn2+ sensitive phenotype in A. thaliana hma1 knockouts and proposed a functional role as a plastidic Zn2+ exporter for this enzyme (Kim et al., 2009).
In summary, there is a lack of consensus on the metal specificity P1B4-ATPases and little information of their role in bacteria. Considering the putative importance of CtpD for M. tubeculosis virulence, we characterized the function of the homologous M. smegmatis CtpD in metal homeostasis. Both proteins display the structural features common to this group: six TMs and the motifs SPC and HEG[S/G]T in TM4 and TM6, respectively (Fig. 1A). It is assumed that these invariant amino acids constitute the transmembrane metal binding site (TM-MBS) participating in metal coordination during transport (Argüello, 2003; Seigneurin-Berny et al., 2006). Our results indicate that the CtpD is a Co2+-ATPase that is required for maintaining Co2+ homeostasis in Mycobacteria. However, like other P1B-ATPases, it can accept other substrates such as Zn2+ and Ni2+, when tested in vitro or under stress conditions.
Metal transport specificity in P1B-ATPases is the consequence of metal coordination by the conserved amino acids forming TM-MBS (Argüello, 2003; Seigneurin-Berny et al., 2006; González-Guerrero & Argüello, 2008). However, the presence of N-terminal metal binding sites (N-MBD) has also been proposed as determining factor (Borrelly et al., 2004). Toward verifying the presence of these elements in Mycobacteria proteins, we compared them to 129 sequences in the P1B4-subgroup retrieved from ~600 bacterial genomes (Fig 1B). Despite sharing 48–66 % similarity, the previously characterized P1B4-ATPases (HMA1, CoaT, CzcP, ZosA) clustered in separate phylogenetic branches. The early divergence of these genes might explain the partially alternative functions that have been observed (see above), although the specific alignment of their metal binding TM4, TM5 and TM6 showed a much higher degree of identity (80–90%). P1B4-ATPases do not have archetypical cytosolic metal binding domains (MBDs). However, a subset of them (22%) showed longer N-terminal sequences containing a C[V/L/I/F]HYD motif and a number of His residues that could participate in metal binding. The mycobacterial P1B4-ATPases are located in a distinct branch containing no functionally characterized proteins. The Mtb gene most related to M. smegmatis CtpD is also annotated as CtpD (Rv1469), although we found that the Mtb CtpJ (Rv3743) gene is also highly related. A further indication that the CtpD and CtpJ proteins may serve similar function is the presence of the Co2+ responsive transcription factor nmtR up-stream of both M. smegmatis CtpD and Mtb CtpJ (Cavet et al., 2002).
To study the functional role of CtpD in M. smegmatis, we assessed the in vitro fitness of a ΔctpD deletion mutant in the presence of putative metal substrates of PIB4-ATPases. The ΔctpD strain was more sensitive to Co2+ and Ni2+ than the wild type cells (Fig. 2A and B). The wild type phenotypes were restored, albeit partially for Ni2+, by complementation with the plasmid pMV306 harboring the gene ctpD under the regulation of its native promoter. Since Zn2+ and Cu2+, Ca2+, and Mn2+ are putative substrates of P1B4-ATPases (Gaballa & Helmann, 2002; Scherer & Nies, 2009; Kim et al., 2009; Seigneurin-Berny et al., 2006; Moreno et al., 2008), the tolerance of the mutant strain to these metals was tested. No difference between WT and ΔctpD strains was detected when Zn2+ or Cu2+ were added to the media (Fig. 2C–D). Similarly, no sensitivity to high Ca2+ or Mn2+ in the media was observed (not shown), suggesting that these would not be in vivo substrates of this transporter. It could be postulated that other redundant Zn2+-ATPases (PIB2-ATPases) might mask the CtpD Zn2+-ATPase activity, although M. smegamtis genome encodes no predicted PIB2-ATPases. On the other hand, the cation diffuson facilitator (CDF) transporter ZitA appears responsible for Zn2+ homeostasis in M. smegmatis (Grover & Sharma, 2006). To further assess a possible role of CtpD in Zn2+ homeostasis a double deletion ΔctpD ΔzitA mutant strain was constructed and its Zn2+ sensitivity was compared to that of a ΔzitA strain. No differences were observed, suggesting that CtpD does not contribute to Zn2+ efflux (Fig. 2C and inset). Taking into account the proposed role of ZosA in ROS detoxification, the effect of the superoxide generator, paraquat, and H2O2 were tested (Gaballa & Helmann, 2002). No effect of these stressors nor of a reactive nitrogen species generator (NO2Na at pH 5.5) was observed (Fig. S3).
Several metals were tested as modulators of ctpD transcription (Fig. 3A). We found that Co2+ induced ctpD expression by more than twenty fold. This effect was very specific to Co2+, as no induction of ctpD mRNA was observed upon treatment with Ni, Cu, Zn, Ca, or Mn. Induction of ctpD by superoxide generators was also tested. Paraquat treatment led to 5-fold increase in the expression of ctpD; while H2O2 did not elicit a response (Fig 3B).
The results presented above suggested a functional role of CtpD in Co2+/Ni2+ detoxification. To test this hypothesis, exponentially growing cells were incubated for 2 h in presence of sub-lethal concentrations of Co2+ and Ni2+. Accumulation of other metals previously postulated as substrates of P1B4-ATPases like Zn2+, Cu2+, Ca2+ and Mn2+ were also investigated. Consistent with the Co2+ and Ni2+ sensitive phenotype of the ΔctpD strain, only accumulation of Co2+ and Ni2+ was observed in this mutant (Fig. 4A). Moreover, the complemented strain showed metal levels comparable to those observed in WT cells. Similar results were obtained at higher concentrations of Co2+ (200 μM). However in this case, significant complementation was not observed (not shown).
PIB-ATPases couple transmembrane metal transport to ATP hydrolysis (Argüello et al., 2007). Consequently, assessment of ATPase activity reports on the capability to transport candidate substrates. To measure its enzymatic activity, CtpD was expressed in Escherichia coli, solubilized, and affinity purified following well-established protocols (Mandal et al., 2002). However, the purified enzyme showed constitutive activation; i.e. high ATPase activity in the absence of added metals. A similar phenomenon is apparent when human Cu+-ATPases are heterologously expressed (Tsivkovskii et al., 2002). To remove possible activating metals, CtpD was treated with the divalent metal chelators, EDTA and tetrathiomolybdate (TTM), before ATPase determinations. The treatment led to a metal responsive enzyme. Figure 5 shows the activation of CtpD ATPase by Co2+, Ni2+ and a lesser extent Zn2+ (Fig. 5). It is notable, that while the K1/2 for activation are comparable to those observed for P1B-ATPases (Liu et al., 2006; Scherer & Nies, 2009), the Vmax for Co2+ or Ni2+ activation was at least an order of magnitude lower than those of Cu+- or Zn2+-ATPases (Liu et al., 2006; González-Guerrero et al., 2010).
It has been proposed that the relative affinities of the TM-MBS for the various transition metals are major determinants of metal transport specificity (Borrelly et al., 2004; Argüello et al., 2007; Argüello et al., 2012). In order to test this hypothesis, first the metal binding capability of CtpD, lacking the His-tag used for purification, was determined by incubating the protein in the presence of excess Cu+, Zn2+, Co2+, Ni2+ and subsequent removal of free metals. In all cases a 1:1 molar ratio (bound metal: CtpD) was observed (Table 1). This stoichiometry indicated the existence of one TM-MBS in this P1B4-ATPase, and facilitated the direct assessment of metal dissociation constants for Co2+, Ni2+ and Zn2+. Using competition assays in the presence of mag-fura-2 (Fig. 6A–C), we found that CtpD appears to bind Zn2+ with higher affinity than Co2+ and Ni2+ (Table 1).
The genome of Mtb encodes two metal transporting P1B4-ATPases, CtpJ and CtpD. Genetic screens for mutants that are attenuated for growth in animal models of TB indicate that CtpD may represent a significant virulence factor (Sassetti & Rubin, 2003; Joshi et al., 2006). The results obtained through the characterization of the M. smegmatis ortholog suggest that CtpD is a Co2+/Ni2+ transporting ATPase that contributes to metal homeostasis, and perhaps ROS resistance, by driving Co2+ and Ni2+ export.
We have proposed that PIB-ATPases sharing invariant residues in core TMs have a similar metal specificity (Argüello, 2003; Argüello et al., 2007). Despite having similar TM-MBS residues, previous studies have proposed that different members of the P1B4-subgroup transport a variety of metals; Co2+ (Rutherford et al., 1999), Cu+ (Seigneurin-Berny et al., 2006; Moreno et al., 2008), Ca2+ (Moreno et al., 2008), and Zn2+ (Moreno et al., 2008; Scherer & Nies, 2009). These observations could be rationalized by considering the influence of second spheres of coordination. This factor may be particularly important for in vitro assays where ions do not access the transport site via chaperone/complexing delivery molecules (Argüello et al., 2012)(Argüello et al., 2011). Interestingly, all characterized members of the P1B4-ATPases subgroup are located in different phylogenic branches, opening the possibility that non-coordinating residues close to the TM-MBS might vary among branches and thus confer slightly different selectivity. Consequently our characterization of M. smegmatis CtpD was necessary to assess the role of these enzymes in Mycobacteria. Our in vitro assays suggest that the enzyme can bind and be activated by Co2+, Ni2+, and Zn2+. Co2+ and Ni2+ appear to be the preferred substrates showing lower K1/2 and significantly higher Vmax values compared to Zn2+; even though, this last ion has a much lower Kd. Characterization of Cu+-, Cu2+ and Zn2+-ATPases (P1B1-3-ATPases) has shown a complicated relation between TM-MBS metal binding affinities and ATPase turnover rates (Mandal et al., 2002; Liu et al., 2006; Mana-Capelli et al., 2003)(González-Guerrero et al., 2008). It is apparent that the rate of transport is dependent on the protein capability transition from E1-metal bound state to the E2 metal free conformation, with the implicit metal release rate limiting step (Mandal et al., 2002). As a result, metals that are easily released, rather than tightly bound, are transported more rapidly. On another hand, free metal Kd values might not represent selectivity in vivo since transition metals gain access to the TM-MBS via delivery by a metallo-chaperone or sequestering molecule (Argüello et al., 2012). Importantly, the phenotype of ΔctpD mutant cells shows that Co2+ and Ni2+ are physiological substrates of this protein. Three distinct assays support this conclusion: metal tolerance, metal accumulation, and gene induction. CtpD does not appear to be a factor in Zn2+ homeostasis, even when the main contributor to tolerance, ZitA is inactivated. As with other ATPases, it is possible that Zn2+ is a substrate in vitro but not in vivo (Argüello et al., 2007). Our study on the functional role of CtpD in M. smegmatis supports previous findings in the cyanobacterial organism Synechosistis PCC6803 (Rutherford et al., 1999) and suggest that P1B4-ATPases are major participants in bacterial Co2+ homeostasis. The sensitivity of the CtpD-deficient bacteria to high levels of Ni2+ and the intracellular accumulation of this metal in the mutant cells, allow us to extend the specificity of this sub-group to include Ni2+ as transported substrate (Co2+/Ni2+-ATPase).
The participation of B. subtlis ZosA and A. thaliana HMA1 in ROS detoxification has been observed (Gaballa & Helmann, 2002; Seigneurin-Berny et al., 2006). Similarly, the up-regulation of ctpD by the superoxide generator, paraquat, further supports an important role for these ATPases in the control of redox stress. Considering the high level of noxious transition metals and ROS in the phagosome and that Mtb has two genes coding for P1B4-ATPases, it is tempting to speculate that these transporters might participate in both functions, directly in metal homeostasis via ion transport and ROS detoxification by contributing to metal uploading into secreted redox-regulating enzymes. Similar roles have already been proposed for homologous Cu+-ATPases (Argüello et al., 2011).
M. smegmatis mc2155 (WT) and derived strains were grown in 7H9 liquid media, or 7H10 agar plates (BD-Difco), supplemented with 0.5% bovine serum albumin, 0.2% Dextrose and 0.05% Tween 80 (ADT). Bacterial strains, plasmids and PCR products were obtained using standard molecular biology techniques (Sambrook et al., 1989). The M. smegmatis ΔctpD mutant strain was constructed by replacement of the coding sequence of ctpD (locus tag MSMEG_5403) with a hygromycin resistance cassette. This was amplified from plasmid pKM342 using primers Smeg5403-F and Smeg5403-R (Table S2) containing the 50 bp of 5′- and 3′-flanking regions of ctpD. The resulting amplicon served as template for a PCR using primers MSMEG_5403-F and MSMEG_5403-R. The product containing the hygromycin cassette bordered by ~125 bp of the 5′- and 3′-flanking region of ctpD was transformed into electrocompetent M. smegmatis mc2155 cells harboring plasmid pJV53 and induced with 0.02% acetimide for 8 h (van Kessel & Hatfull, 2007). Transformants were selected on 50 μg ml−1 hygromycin, 7H10-ADT. Double crossover was verified by PCR using primers matching inside (primers intSmeg5403-For and intSmeg5403-Rev) and outside (Smeg5403-V1 and Smeg5403-V2) locus SMEG_5403 (Fig. S1). Mutants ΔzitA (MSMEG_0750) and ΔctpD ΔzitA were constructed following the same procedure using primers Smeg0750-F, Smeg0750-R, MSMEG_0750-F and MSMEG_0750-R (Table S2). To obtain the double mutant ΔctpD ΔzitA the hygromycin cassette in strain ΔctpD was cleaned-up using the plasmid pCre-Zeo-sacB which contains Cre-recombinase to resolve loxP sites flanking the cassette.
To complement the ΔctpD strain, ctpD and upstream promoter elements (−250 bp) were amplified by PCR with primers For-5XbaI-P-5403 and Rev-3EcoRI-P-5403, which introduce XbaI and EcoRI sites at the fragment termini. The digested amplicon was cloned into vector pMV306, creating plasmid pMV306-PctpD. This was electroporated into ΔctpD strain and transformants were selected on 25 μg ml−1 kanamycin, 7H10-ADT and verified by PCR using primers matching inside the gene (Fig. S1B).
Liquid 7H9-ADT cultures were inoculated at OD600 of 0.1 from late exponential phase cultures and supplemented with the desired metal concentration as indicated in the figures. Cells we incubated for 40 h and OD600 measured.
15 ml of liquid 7H9-ADT cultures in late exponential phase were supplemented with either 100 nM CoCl2, 100 nM NiCl2, 200 μM CuCl2, 1 mM ZnCl2, 1 mM CaCl2 or 1 mM MnCl2 and incubated for 2 h. After this incubation, protein concentration was determined, cells harvested, and washed with 0.9% NaCl. Pellets were acid digested with 0.5 ml NO3H (trace metal grade) for 1 h at 80°C and then overnight at 20°C. Digestions were stopped by addition of 0.1 ml of 30% H2O2 and dilution to 10 ml with water. Metal contents in digested samples were measured by furnace atomic absorption spectroscopy (AAS) (Varian SpectrAA 880/GTA 100, Santa Clara, CA).
M. smegmatis mc2155 cells in exponential phase were supplemented with 500 μM CoCl2, 500 μM NiCl2, 500 μM CuCl2, 1 mM ZnCl2, 1 mM CaCl2 or 1mM MnCl2 and incubated for 2 h. Cells were harvested, resuspended in 1 ml TRIzol reagent (Invitrogen, Carlsbad, CA), and disrupted using 0.5-ml zirconium beads (0.1-mm diameter) in a cell disrupter (FastPrep FP120, Qbiogene, Carlsbad, CA). RNA pellets were air dried and redissolved in 50 μl diethyl pyrocarbonate-treated ultrapure water. Remaining DNA was removed with RNeasy minikit and an on-column DNase I kit (Qiagen, Valencia, CA). The RNA samples (5μg) were used as templates for cDNA synthesis with random primers and SuperScript III reverse transcriptase (Invitrogen, Carlsbad, CA). Quantitative reverse transcription-PCR (qRT-PCR) was performed with iQ SYBR green supermix (Bio-Rad Laboratories, Hercules, CA). For transcript analysis of ctpD, qRT-PCR was performed with primers q5403-F and q5403-R. Primers qSMEG-2758sigA-F and qSMEG-2758sigA-R were used to amplify the RNA polymerase sigma factor (sigA), as an internal reference. Real-time cycler conditions used have been previously described (González-Guerrero et al., 2010). Determinations were carried out with RNA extracted from three independent biological samples, with the threshold cycle (Ct) determined in triplicate. The relative levels of transcription were calculated by using the 2−ΔΔCt method (Livak & Schmittgen, 2001). The mock reverse transcription reactions, containing RNA and all reagents except reverse transcriptase, confirmed that the results obtained were not due to contaminating genomic DNAs (data not shown).
cDNA encoding M. smegmatis CtpD was amplified using genomic DNA as template and primers Smeg_5403For and Rev-TEV-54. This introduced a Tobacco etch virus (TEV) protease site coding sequence at the amplicon 3′ end. The PCR product was cloned into pBAD-TOPO/His vector. cDNA sequence was confirmed by automated DNA sequence analysis. This construct was introduced into E. coli Top10 cells (Invitrogen) carrying a plasmid encoding for rare tRNAs (tRNA argAGA/AGG and tRNA ileAUA). Cells were grown at 37 °C in ZYP-505 media supplemented with 0.05% arabinose, 100 μg ml−1 ampicillin, 50 μg ml−1 kanamycin (Studier, 2005). Cells were harvested at 24 h post inoculation, washed with 25 mM Tris, pH 7.0, 100 mM KCl and stored at −70°C. Protein purification was carried as previously described (Mandal et al., 2002). Briefly, cells were disrupted in a French press and membranes were isolated by centrifugation. Membranes (3 mg ml−1 protein) were treated with 0.75% dodecyl-β-D-maltoside (DDM) (Calbiochem) 25 mM Tris, pH 8.0, 100 mM sucrose, 500 mM NaCl, 1 mM phenylmethylsulfonyl fluoride. The solubilized membrane protein suspension was cleared by centrifugation at 163,000 × g for 1 h. CtpD was affinity purified using Ni2+-nitrilotriacetic acid (Ni-NTA) resin (Qiagen, Valencia, CA). After washing the resin with 5–20 mM imidazole, 50 mM Hepes, pH 7.4, 0.05% DDM, the protein was eluted with 200 mM imidazole in buffer C (50 mM Hepes, pH 7.4, 200 mM NaCl, 0.01% DDM) and imidazole was removed by buffer exchange (buffer C) using Ultra-30 Centricon (Millipore, Billerica, MA) filtration device. The (His)6-tag was removed from the (His)6-CtpD fusion protein by treatment with a (His)6-tagged TEV protease (Rosadini et al., 2011) at 5 CtpD:1 TEV-(His)6 weight ratio for 1 h at 22°C in buffer C plus 1 mM Tris(2-carboxyethyl)phosphine (TCEP) and asolectin 0.01%. TEV-His was removed by affinity purification with Ni-NTA resin. Protein concentration was determined by Bradford (Bradford, 1976). All purification procedures were carried out at 0–4 °C, and no special precautions were taken to prevent enzyme oxidation. Purity was assessed by Coomassie brilliant blue staining of overloaded SDS-PAGE gels and by immunostaining Western blots with rabbit anti-(His)6 polyclonal primary antibody (GenScript, Piscataway, NJ) and goat anti-rabbit IgG secondary antibody (horseradish peroxidase conjugate; GenScript, Piscataway, NJ) (Fig. S2). Enzyme used for ATPase determination was incubated for 45 min at RT with 0.5 mM EDTA and 0.5 mM TTM in buffer C at 1 mg ml−1 protein concentration. Chelators were removed by buffer exchange (buffer C) using Ultra-30 Centricon (Millipore, Billerica, MA) filtration device.
Maximum metal binding to isolated CtpD was measured as previously described (Eren et al., 2006). Ten micromolar His-less CtpD was incubated for 1 min at 4°C in 50mM HEPES-NaOH, pH 7.5, 200 mM NaCl, 1 mM TCEP and 50 μM of either ZnCl2, CoCl2 or NiCl2. After incubation, excess metal was removed by size exclusion using Sephadex G-25 columns. Eluted protein was acid digested as described above and metal concentrations measured using furnace AAS. Background metal levels control samples lacking protein were <10 % of those observed in CtpD samples.
CtpD-metal binding affinities were determined using the divalent metal binding chromophore mag-fura-2 (Invitrogen, Carlsbad, CA) (Eren et al., 2006). Ten micromolar His-less CtpD and 20 μM mag-fura-2 were titrated with 1 mM Co2+, Ni2+ or Zn2+. Free mag-fura-2 was determined by monitoring OD366, (ε366 of 29,900 M−1 cm−1) (Walkup & Imperiali, 1997). Free Co2+, Ni2+ or Zn2+ concentrations were calculated using the equation [Metal] f =[I.Metal]/[I] f KI, where I is the concentration of mag-fura-2, If is free mag-fura-2, and KI is the empiric dissociation constant of mag-fura-2 for metal (Walkup & Imperiali, 1997). Mag-fura-2 dissociation constants (KI) were 20 nM for Zn2+ (Walkup & Imperiali, 1997), 2.8 μM for Co2+ (Llarrull et al., 2007). Determination of mag-fura-2 KI for Ni2+ was carried under the same conditions of the assay. The indicator was first titrated with Ni2+ and changes in absorbance at 363 nm with added metal were fit to the model I.Metal <−> I+ Metal. A dissociation constant of Ni2+ from the complex with mag-fura-2 of 3.3 μM was obtained. The metal-protein dissociation constant (Kd) and the apparent stoichiometry (n) of CtpD were calculated by fitting the data to the equation v= n [Metal]f/Kd(1+[Metal]f/Kd), where v is the ratio of moles of metal bound to total protein, n is the number of binding sites, and [Metal]f is the free metal concentration (Guo & Giedroc, 1997).
These were performed at 37°C in a medium containing 50mM Tris (pH 7.4 at RT), 3mM MgCl2, 3mM ATP, 0.01% asolectin, 0.01%DDM, 20 mM cysteine, 50 mM NaCl, 2.5mM DTT, and 10–40_μg ml−1 purified protein plus indicated metal concentrations. In ATPase determination testing activation by Co2+, DDT was replaced by TCEP to prevent interference with color reaction. ATPase activity was measured after 10 min incubation and released Pi determined according to Lanzetta et al. (Lanzetta et al., 1979). ATPase activity measured in the absence of metal was subtracted from plotted values. Curves of ATPase activity vs. metal concentrations were fit to v=Vmax[metal]/([metal]+K1/2). The reported standard errors for Vmax and K1/2 are asymptotic standard errors reported by the fitting software KaleidaGraph (Synergy).
P1B4-ATPase protein sequences present in most bacterial genomes available as October 2011 were obtained from the TransportDB (http://www.membranetransport.org) (Ren et al., 2007). Additional sequences were retrieved by BLAST searches at the Comprehensive Microbial Resource (http://cmr.jcvi.org/tigr-scripts/CMR/CmrHomePage.cgi). Plant protein sequences were obtained by BLAST searches at the National Center for Biotechnlogy Information (http://blast.ncbi.nlm.nih.gov/Blast.cgi). The identity of these sequences as P1B4-ATPases was confirmed by the conserved membrane topology and the presence the signature amino acids: SPC motif in TM6 and HEGST in TM8 (Argüello, 2003). Transmembrane segments were determined using SOSUI (Hirokawa et al., 1998) and Top-Pred2 (von Heijne, 1992). Sequences were aligned using ClustalW2 (Thompson et al., 1994) and cladogram visualized with TreeView (Page, 1996).
This work was supported by NIH awards F32A1093049 (J.E.L.), 1R21AI082484 (J.M.A.), AI064282 (CMS) and the Howard Hughes Medical Institute (CMS). We thank K.G. Papavinasasundaram for his valuable technical assistance.
1For simplicity P-type ATPases will be referred as P-ATPases, P1B-ATPases, etc.