Cd²⁺ extrusion by P-type Cd²⁺-ATPase of Staphylococcus aureus 17810R via energy-dependent Cd²⁺/H⁺ exchange mechanism

Abstract

Cd²⁺ is highly toxic to Staphylococcus aureus since it blocks dithiols in cytoplasmic 2-oxoglutarate dehydrogenase complex (ODHC) participating in energy conservation process. However, S. aureus 17810R is Cd²⁺-resistant due to possession of cadA-coded Cd²⁺ efflux system, recognized here as P-type Cd²⁺-ATPase. This Cd²⁺ pump utilizing cellular energy—ATP, ∆μ ⁺_H (electrochemical proton potential) and respiratory protons, extrudes Cd²⁺ from cytoplasm to protect dithiols in ODHC, but the mechanism of Cd²⁺ extrusion remains unknown. Here we propose that two Cd²⁺ taken up by strain 17810R via Mn²⁺ uniporter down membrane potential (∆ψ) generated during glutamate oxidation in 100 mM phosphate buffer (high P_iB) are trapped probably by high affinity sites in cytoplasmic domain of Cd²⁺-ATPase, forming SCdS. This stops Cd²⁺ transport towards dithiols in ODHC, allowing undisturbed NADH production, its oxidation and energy conservation, while ATP could change orientation of SCdS towards facing transmembrane channel. Now, increased number of P_i-dependent protons pumped electrogenically via respiratory chain and countertransported through the channel down ∆ψ, extrude two trapped cytoplasmic Cd²⁺, which move to low affinity sites, being then extruded into extracellular space via ∆ψ-dependent Cd²⁺/H⁺ exchange. In 1 mM phosphate buffer (low P_iB), external Cd²⁺ competing with decreased number of P_i-dependent protons, binds to ψ_s of Cd²⁺-ATPase channel, enters cytoplasm through the channel down ∆ψ via Cd²⁺/Cd²⁺ exchange and blocks dithiols in ODHC. However, Mg²⁺ pretreatment preventing external Cd²⁺ countertransport through the channel down ∆ψ, allowed undisturbed NADH production, its oxidation and extrusion of two cytoplasmic Cd²⁺ via Cd²⁺/H⁺ exchange, despite low P_iB.

Introduction

Cadmium is highly toxic to living organisms, since it blocks sulphhydryl groups in essential proteins (Vallee and Ulmer 1972; Moulis and Thevenod 2010). Some bacteria carry plasmid-linked cadA gene (Novick and Roth 1968; Dyke et al. 1970) conferring Cd²⁺ resistance expressed as decreased ¹⁰⁹Cd uptake (Chopra 1975; Tynecka et al. 1975; Silver et al. 1982). Tynecka et al. (1981a, 1981b) were the first to report that the highly decreased ¹⁰⁹Cd uptake by growing cells of Cd²⁺-resistant Staphylococcus aureus 17810R was due to pH gradient (∆pH)-dependent, nigericin-sensitive cadA-coded Cd²⁺ efflux system.

Subsequently, Silver and coworkers (Nucifora et al. 1989; Silver et al. 1989) showed that the cadA gene from staphylococcal plasmid pI258 coded the P-type Cd²⁺-ATPase, belonging to family of membrane-bound, cation-translocating pumps found in eukaryotes and prokaryotes. These pumps located across the membrane maintain homeostasis of essential cations (e.g. Mg²⁺, Ca²⁺, K⁺, Na⁺) or protons (Apell 2003; Kühlbrandt 2004; Pedersen 2007), and confer resistance to heavy metals (e.g. Cd²⁺, Zn²⁺, Cu²⁺) (Rosen 2002; Nies 2003; Kühlbrandt 2004; Silver and Phung 2005; Argüello et al. 2007, 2011). The best characterized is the P-type Ca²⁺-ATPase of sarcoplasmic reticulum (SR) for which detailed biochemical and biophysical data (Apell 2003; Toyoshima 2008) and about 50 crystal structures are available (Toyoshima et al. 2013). However, it is still controversial, how ATP energy is transduced to vectorial Ca²⁺ movement (Scarborough 2003; Toyoshima 2009).

According to sequencing data by Silver and coworkers (Nucifora et al. 1989; Silver et al. 1989), the four cysteine residues present in staphylococcal CadA protein are essential for Cd²⁺-ATPase activity: the conserved Cys23X₂Cys26 in cytoplasmic domain—a possible high affinity Cd²⁺ binding site, and in conserved Cys371ProCys373 inside transmembrane channel, involved probably in Cd²⁺ extrusion. The CysX₂Cys motif is related to copper-binding region in Cu²⁺-ATPases (Fan and Rosen 2002) and to mercury-binding region in proteins involved in Hg²⁺ resistance (Barkay et al. 2003). According to Tsai et al. (1992), staphylococcal P-type Cd²⁺-ATPase requires only ATP. Here is shown, that the cadA-coded Cd²⁺ efflux system in Cd²⁺-resistant S. aureus 17810R (Tynecka et al. Tynecka et al. 1981a, 1981b; Tynecka and Szcześniak 1991) is a P-type Cd²⁺-ATPase requiring: ATP, electrochemical proton potential (∆μ ⁺_H ), high phosphate buffer (P_iB) and P_i-dependent protons or Mg²⁺. The mechanism of Cd²⁺ extrusion by this staphylococcal Cd²⁺-ATPase is proposed.

Materials and methods

Bacterial strains and culture conditions

Cd²⁺-resistant S. aureus 17810R, carrying cadA gene on penicillinase plasmid pII17810 (Shalita et al. 1980), was described previously (Tynecka et al. 1981a, 1981b). Experiments were performed at 37 °C using early exponential phase cells grown aerobically in 3 % nutrient broth and suspended in 100 mM potassium/sodium phosphate buffer, pH 7 (P_iB). Cell suspensions were vigorously aerated for 3 h at 37 °C without exogenous electron donor to deprive cells of endogenous energy reserves (Tynecka and Malm 1995; Tynecka et al. 2001). Next, cells were suspended in P_iB of various concentrations, depending on the experiment, at a density of 0.2 mg dry weight/ml and preincubated with 10 mM glutamate for 10 min at 37 °C (glutamate oxidizing cells). In some experiments, cells were suspended in other buffers: 100 mM triethanolamine/phosphate, pH 7, 100 mM Tris/HCl, pH 7.2 or 100 mM MOPS/NaOH, pH 7. Cd²⁺-sensitive variant strain S. aureus 17810S lacking cadA gene, also described previously (Tynecka et al. 1981a, 1981b), was used in some experiments as a control organism.

Reagents

Inhibitors: 2-heptyl-4-hydroxyquinoline N-oxide (HQNO) and dicyclohexylcarbodiimide (DCCD), and ionophores: valinomycin, nigericin or carbonyl cyanide m-chlorophenyl hydrazone (CCCP) were from Sigma (St. Louis, MO). The following radiolabeled compounds were used: ¹⁰⁹Cd (carrier-free) or sodium [U-¹⁴C]glutamate (7.4 GBq/mmol)—from Amersham, UK, ⁸⁶RbCl (1.075 GBq/mmol), sodium [¹⁴C]benzoate (407 MBq/mmol), [³H]inulin (3.7 GBq/mmol) or [γ-³²P]ATP (111 TBq/mmol)—from NEN™ Life Science Products (Boston, MA), while ³²P_i—inorganic orthophosphate (740 MBq/mmol)—from the Institute of Nuclear Research, Świerk, Poland.

Uptake experiments

Uptake of ¹⁰⁹Cd at 10 μM (as CdCl₂) by glutamate oxidizing cells of strain 17810R and strain 17810S was assayed by filtration procedure, as described previously (Tynecka et al. 1981a, 1981b). These cells suspended in 100 or 1 mM P_iB were preincubated at 37 °C for 10 min, with appropriate compounds: MgCl₂, MnCl₂ or ionophores—nigericin, valinomycin + KCl or CCCP, depending on the experiment, before addition of 10 μM ¹⁰⁹CdCl₂. In order to determine K _m and V _max of ¹⁰⁹Cd uptake in strain 17810R, the initial influx rate of ¹⁰⁹Cd uptake in 1 mM P_iB within 1 min at various CdCl₂ concentrations was measured.

Uptake of ³²P_i (inorganic orthophosphate) by glutamate oxidizing cells of strain 17810R suspended in 100 or 1 mM P_iB was assayed by filtration procedure, as described previously (Tynecka and Szcześniak 1991).

Assay of ¹⁰⁹ Cd efflux

¹⁰⁹Cd efflux was assayed by filtration procedure, as described previously (Tynecka et al. 1981b). ¹⁰⁹Cd efflux from washed, glutamate oxidizing cells of strain 17810R was performed after cell preincubation for 20 min with 10 μM ¹⁰⁹CdCl₂ in 1 mM P_iB. After removal of external ¹⁰⁹CdCl₂ by cell washing at 4 °C, cells were resuspended in 100 mM P_iB or in other buffers: 100 mM triethanolamine/phosphate, pH 7, 100 mM Tris/HCl, pH 7.2 or 100 mM MOPS/NaOH, pH 7. DCCD and ionophores: CCCP, valinomycin + KCl or nigericin were added, depending on the experiment. In each experiment 10 mM glutamate was added and all suspensions were prewarmed to 37 °C, before ¹⁰⁹Cd efflux was measured. ¹⁰⁹Cd efflux from unwashed, glutamate oxidizing cells was also performed after cell preincubation for 20 min with 10 μM ¹⁰⁹CdCl₂ in 1 mM P_iB. Then, P_iB concentration was increased at steady-state from 1 mM to 100 mM without cell washing, before ¹⁰⁹Cd efflux was measured. Appropriate compounds: MgCl₂ or ionophores: CCCP, valinomycin + KCl or nigericin were added at steady-state, depending on the experiment.

Assay of ¹⁰⁹Cd content in subcellular fractions

¹⁰⁹Cd distribution among subcellular fractions obtained from glutamate oxidizing cells of strain 17810R and strain 17810S preloaded with 10 μM ¹⁰⁹CdCl₂ in 100 or 1 mM P_iB and preincubated at 37 °C with appropriate compounds: MgCl₂ or ionophores—CCCP, valinomycin + KCl or nigericin, depending on the experiment, was determined according to Tynecka et al. (2001).

Assay of enzyme activity

Activity of 2-oxoglutarate dehydrogenase complex (ODHC) was measured in the cytoplasmic fraction obtained from glutamate oxidizing cells of strain 17810R suspended in 100 or 1 mM P_iB and preincubated at 37 °C with CdCl₂, according to the method described previously (Tynecka and Malm 1996).

Assay of membrane potential (∆ψ) and pH gradient (∆pH)

The values of ∆ψ and ∆pH in glutamate oxidizing cells of strain 17810R suspended in 100 or 1 mM P_iB were determined by a filtration procedure from the steady-state distribution of 100 μM ⁸⁶Rb in the presence of 10 μM valinomycin or of 20 μM sodium [¹⁴C]benzoate, respectively, as described previously (Tynecka et al. 1999). [³H]inulin served as a marker for extracellular water.

Phosphorylation assay

Membrane fraction of strain 17810R was obtained according to the procedure described previously (Tynecka and Malm 1996; Tynecka et al. 2001). Phosphorylation assay was performed as described elsewhere (Tsai and Lynn Linet 1993) with some modifications. To 200 μl of membrane fraction (2.4 mg protein/ml), 2 μl of 1.2 mM EDTA were added, followed by incubation for 10 min, and then 2 μl of 50 μM CdCl₂ or equivalent volume of deionized water, followed by incubation for 5 min. The reaction was started by addition of 10 μCi of [γ-³²P]ATP and 2 μl of 0.8 M MgCl₂. The reaction mixture was incubated at room temperature, then the reaction was stopped after 60 s by addition of equivalent volume of ice-cold 20 % TCA. After 10 min incubation on ice, the membranes were collected by centrifugation (14,000 rpm, 5 min). In order to assay the effect of alkali or hydroxylamine, the pellets were incubated with 100 μl of 0.5 M KOH for 5 min on ice or with 200 μl of 0.1 M sodium acetate containing 260 mM hydroxylamine for 10 min at room temperature. After incubation, equivalent volume of ice-cold 10 % TCA was added. In each case, the collected pellets were washed with water and then twice with 50 mM H₃PO₄/NaOH, pH 2.4. Then, the pellets were dissolved in 10 % SDS at 100 °C and suspended in a standard sample buffer used for acidic SDS-PAGE, as described elsewhere (Fairbanks and Avruch 1972). Gels were run at 40 mA for 4–5 h at room temperature. After electrophoresis, autoradiography of the dried gels was performed at 4 °C for 48 h.

Reproducibility of results

The experimental data shown in each figure are the mean ± SD from at least three independent experiments.

Results

Highly decreased ¹⁰⁹Cd accumulation in Cd²⁺-resistant S. aureus 17810R oxidizing glutamate in 100 mM phosphate buffer, pH 7 (high P_iB).

First, membrane proteins of S. aureus 17810R harbouring cadA gene were phosphorylated by [γ-³²P]ATP (Fig. 1). The protein band of about 100 kDa was strongly phosphorylated, when Cd²⁺ was present. Intensity of this band was decreased by alkali or hydroxylamine, which is typical for phosphoenzyme intermediate of P-type ATPases (Tsai and Lynn Linet 1993). This suggests that the band strongly phosphorylated in strain 17810R in the presence of Cd²⁺ (Fig. 1) may correspond to CadA protein, having also molecular weight of about 80 kDa (Nucifora et al. 1989; Tsai and Lynn Linet 1993).

Fig. 1

Phosphorylation of membrane proteins in S. aureus 17810R by [³²P]ATP. Lane 1—membrane proteins + 50 μM Cd²⁺, lane 2—membrane proteins + 50 μM Cd²⁺ + 260 mM hydroxylamine, lane 3—membrane proteins + 50 μM Cd²⁺ + 0.5 M KOH, lane 4—membrane proteins without Cd²⁺, the molecular mass markers are also presented. A position of CadA protein is indicated by an arrow

Cd²⁺-resistant S. aureus 17810R took up only 0.5 ± 0.15 nmol ¹⁰⁹Cd/mg dry wt (Fig. 2a) and accumulated in cytoplasm merely 0.37 ± 0.1 nmol ¹⁰⁹Cd/mg protein (Fig. 2b). Under similar conditions, the Cd²⁺-sensitive variant strain S. aureus 17810S lacking cadA gene, took up 20 ± 1.2 nmol ¹⁰⁹Cd/mg dry wt (Fig. 2a) and accumulated in cytoplasm 21 ± 1.5 nmol ¹⁰⁹Cd/mg protein (Fig. 2b) down ∆ψ (membrane potential) via high affinity Mn²⁺ uniporter sensitive to Mn²⁺ or valinomycin + K⁺ (Fig. 2a). As was already reported (Tynecka et al. 1981a, 1981b; Tynecka and Malm 1995, 1996: Tynecka et al. 1989), two Cd²⁺ accumulated by strain 17810S in a transport cycle, blocked vicinal dithiols in dihydrolipoate and dihydrolipoate dehydrogenase in the cytoplasmic 2-oxoglutarate dehydrogenase complex (ODHC) in Krebs cycle located in the first energy coupling site of respiratory chain (Tynecka et al. 1999). These dithiols are the only Cd²⁺-sensitive targets in glutamate-linked energy conservation system in strain 17810S; their blocking stopped endogenous NADH production, its oxidation via respiratory chain, generation of electrochemical proton potential (∆μ ⁺_H ) and consequently ∆μ ⁺_H -dependent processes without direct blocking of solute transporters and ATP synthase (Tynecka and Malm 1995, 1996; Tynecka et al. 1989, 2001).

Fig. 2

¹⁰⁹Cd uptake and its distribution in subcellular fractions in S. aureus 17810R oxidizing glutamate in 100 mM phosphate buffer, pH 7 (high P_iB) or in 1 mM phosphate buffer, pH 7 (low P_iB). In some experiments Cd²⁺-sensitive variant strain S. aureus 17810S was used. (a) Uptake of ¹⁰⁹Cd in high P_iB: control cells of strain 17810R (filled circles), cells of strain 17810R preincubated with 0.5 μM nigericin (filled triangles), control cells of strain 17810S (filled squares), cells of strain 17810S preincubated with 100 μM Mn²⁺ (empty circles) or 5 μM valinomycin + 50 mM K⁺ (empty triangles). (b) Distribution of ¹⁰⁹Cd in subcellular fractions of strain 17810R and strain 17810S in high P_iB. (c) Uptake of ¹⁰⁹Cd in low P_iB: control cells of strain 17810R (filled circles), cells of strain 17810R preincubated with 100 μM Mn²⁺ (filled diamonds), 1 mM Mg²⁺ (filled squares) or 5 mM Mg²⁺ (filled triangles). (d) Distribution of ¹⁰⁹Cd in subcellular fractions of strain 17810R in low P_iB

The Cd²⁺-resistant strain 17810R did not accumulate Cd²⁺ (Fig. 2a, b), although cells of strain 17810R and 17810S oxidizing glutamate generated ∆μ ⁺_H of similar value expressed as protonmotive force (∆p) of about −191 ± 5 mV. Data in Fig. 2a, b suggest that two Cd²⁺ transported by strain 17810R via Mn²⁺ uniporter down ∆ψ of −161 ± 5 mV were extruded by Cd²⁺ efflux system described by Tynecka et al. 1981a, 1981b, which was recognized here as a P-type Cd²⁺-ATPase. Cd²⁺ extrusion by this Cd²⁺ pump via Cd²⁺/H⁺ exchange before reaching Cd²⁺-sensitive targets—dithiols in ODHC, allowed undisturbed NADH production (5.4 ± 0.6 nmol NADH/min/mg protein), and consequently its oxidation via respiratory chain, ∆μ ⁺_H generation and energy conservation (data not shown), rendering host cells Cd²⁺-resistant.

Since nigericin, collapsing ∆pH but stimulating ∆ψ, did not increase Cd²⁺ uptake by strain 17810R (Fig. 2a), this suggested that Cd²⁺ extrusion by Cd²⁺-ATPase from glutamate oxidizing cells was not energized by ΔpH. According to chemiosmotic principles (Mitchell 1966), the enhanced transport of inorganic phosphate (P_i) by strain 17810R via H⁺/³²P_i symport consuming ΔpH (Tynecka and Szcześniak 1991), could stimulate generation of membrane potential (∆ψ). It is probable that according to Rosenberg and Friedberg (1984) the H⁺/P_i cotransport by strain 17810R in high P_iB could result in phosphate polymerization and accumulation of additional protons in cytoplasm. We suggest that these P_i-dependent protons pumped electrogenically via respiratory chain could return through the transmembrane channel of Cd²⁺-ATPase down ∆ψ and extruded two cytoplasmic Cd²⁺ into extracellular space via ∆ψ-dependent Cd²⁺/H⁺ exchange. This was confirmed by dependence of Cd²⁺ extrusion on high P_iB, since below 25 mM P_iB, linear Cd²⁺ uptake by strain 17810R insensitive to Mn²⁺ was observed (data not shown). To explain the mechanism of Cd²⁺ extrusion by P-type Cd²⁺-ATPase in high P_i, ¹⁰⁹Cd uptake by strain 17810R was first characterized in 1 mM P_iB and then requirements for its net extrusion were studied.

Uptake of ¹⁰⁹Cd by Cd²⁺-resistant S. aureus 17810R oxidizing glutamate in 1 mM phosphate buffer, pH 7 (low P_iB).

The markedly decreased ³²P_i uptake by strain 17810R from 350 nmol ³²P_i/mg dry wt/20 min in high P_iB to 150 nmol ³²P_i/mg dry wt/20 min in low P_iB could result in decreased number of P_i-dependent protons pumped electrogenically via respiratory chain. Under these conditions, strain 17810R took up 10 ± 1.3 nmol ¹⁰⁹Cd/mg dry wt, insensitive to Mn²⁺ (Fig. 2c); about 8 ± 0.9 nmol Cd²⁺/mg protein were found in cell wall and membrane and similar amount of ¹⁰⁹Cd—in cytoplasm (Fig. 2d), which was only about half of that accumulated by strain 17810S in high P_iB (Fig. 2b). We suggest that in low P_iB the external Cd²⁺ could compete with decreased number of P_i-dependent protons for entry into cytoplasm down ∆ψ through transmembrane channel of Cd²⁺-ATPase. Therefore, the first cytoplasmic Cd²⁺ could be extruded from strain 17810R via exchange with external Cd²⁺ via Cd²⁺/Cd²⁺ exchange, while the second cytoplasmic Cd²⁺ was absent, suggesting its net extrusion. External ¹⁰⁹Cd uptake in low P_iB showed linear dependence on Cd²⁺ concentration (data not shown) and high K _m = 112 ± 2.3 μM and V _max = 9.1 ± 1.2 nmol Cd²⁺/mg dry wt/min, suggesting that Cd²⁺-ATPase channel may function now as low affinity second pathway transporting external Cd²⁺ down ∆ψ instead of protons towards Cd²⁺–sensitive targets—dithiols in ODHC.

External Cd²⁺ accumulated by strain 17810R in low P_iB blocked dithiols in ODHC, which stopped NADH production (from 5.4 ± 0.6 to 0.2 ± 0.1 nmol NADH/min/mg protein) and consequently its oxidation via respiratory chain, but ∆μ ⁺_H generation was unaffected (Δp = −210 ± 4 mV). This suggests that according to Mitchell (1966), Cd²⁺ toxicity to cell respiration could result in conversion of the reversible F_oF₁-ATP synthase into hydrolytic direction, which working now as Cd²⁺-insensitive, anaerobic proton pump—F_oF₁-ATPase (Tynecka et al. 1990), could continue ∆μ ⁺_H generation. We suggest that Δψ of −195 ± 4 mV could energize transport of the second cytoplasmic Cd²⁺ via Mn²⁺ uniporter, while ΔpH of 15 ± 2 mV could support its extrusion via Cd²⁺/H⁺ exchange, as confirmed by absence of the second Cd²⁺ in cytoplasm (Fig. 2c, d). Thus, Cd²⁺-ATPase extruded in low P_iB also two cytoplasmic Cd²⁺, but only external Cd²⁺ reached dithiols in ODHC through the channel via Cd²⁺/Cd²⁺ exchange, disturbing energy conservation and Cd²⁺ resistance of strain 17810R.

We also considered in strain 17810R a controversial problem– existence of low affinity sites on external surface of P-type ATPases (McIntosh 2000; Apell 2003; Scarborough 2003; Toyoshima 2009). First, Silver and coworkers (Nucifora et al. 1989, Silver 1996) recognized during sequencing studies some negatively charged amino acid residues (Glu, Asp) on extracellular surface of CadA protein. It is known (Williams 1978; Barber 1980) that such residues create at physiological pH the surface potentials (ψ_s) on biological membranes, protected by cations of various protective abilities (Mg²⁺ > Ca²⁺ > K⁺ > Na⁺), depending on their concentration and/or affinity. We suggest that Cd²⁺-ATPase channel in strain 17810R may also possess two negatively charged residues forming surface potential (ψ_s) functioning as low affinity sites to which protons or external Cd²⁺ may bind before entering the channel, but this depends on P_iB concentration.

According to Fig. 2c, 1 mM Mg²⁺ prevented external Cd²⁺ uptake by strain 17810R in 50 %, while 5 mM Mg²⁺ (further called Mg²⁺) stopped it. These data confirm existence of low affinity ψ_s sites on extracellular surface of Cd²⁺-ATPase channel in strain 17810R. Protection of ψ_s by Mg²⁺ against external Cd²⁺ binding and its countertransport through the channel towards dithiols in ODHC allowed undisturbed energy conservation and Cd²⁺ resistance. In contrast, Cd²⁺ uptake by strain 17810S was Mg²⁺-insensitive (data not shown), suggesting ψ_s absence in Mn²⁺ uniporter.

The ionophore studies in low P_iB showed that nigericin, collapsing ΔpH, doubled Cd²⁺ uptake by strain 17810R (Fig. 3a). Probably, by stopping ΔpH-dependent efflux of the second cytoplasmic Cd²⁺ energized by the reversed Cd²⁺-insensitive F_oF₁-ATPase, nigericin could unmask Δψ-dependent Cd²⁺ transport via Mn²⁺ uniporter, sensitive to Mn²⁺ (Fig. 3a). Now, strain 17810R accumulating in cytoplasm two Cd²⁺ down Δψ—via transmembrane channel (Cd²⁺/Cd²⁺ exchange) and via Mn²⁺ uniporter (Fig. 3a, b), became Cd²⁺-sensitive, like strain 17810S (Fig. 2a, b). However, Mg²⁺ pretreatment of strain 17810R before nigericin addition, prevented external Cd²⁺ binding to ψ_s of Cd²⁺-ATPase and also stopped Cd²⁺ countertransport through the channel, rendering host cells Cd²⁺-resistant, despite low P_iB (Fig. 3a).

Fig. 3

Effects of ionophores on ¹⁰⁹Cd uptake and its distribution in subcellular fractions in S. aureus 17810R oxidizing glutamate in 1 mM phosphate buffer, pH 7 (low P_iB) with or without 5 mM Mg²⁺. (a) ¹⁰⁹Cd uptake by control cells (filled circles) or cells preincubated with 0.5 μM nigericin (filled squares), 0.5 μM nigericin + 100 μM Mn²⁺ (empty triangles) or 5 mM Mg²⁺ + 0.5 μM nigericin (filled diamonds), cells preincubated with 5 μM valinomycin + 50 mM K⁺ (empty triangles) or 5 mM Mg²⁺ + 5 μM valinomycin + 50 mM K⁺ (empty diamonds), cells preincubated with 10 μM CCCP (empty circles) or 5 mM Mg²⁺ + 10 μM CCCP (empty squares). Distribution of ¹⁰⁹Cd in subcellular fractions after cell preincubation with 0.5 μM nigericin (b), 5 μM valinomycin + 50 mM K⁺ (c) or 10 μM CCCP (d)

Valinomycin + K⁺ collapsing Δψ, did not affect external Cd²⁺ uptake by strain 17810R (Fig. 3a), although Δψ-dependent Cd²⁺ transport via Mn²⁺ uniporter into cytoplasm of strain 17810S was stopped by this ionophore (Fig. 2a). Therefore, valinomycin-insensitive Cd²⁺ uptake by strain 17810R may represent only Δψ-independent external Cd²⁺ binding to cell wall and only to one ψ_s site of Cd²⁺-ATPase, prevented by Mg²⁺ (Fig. 3a, c), while external Cd²⁺ binding to the second ψ_s site was probably prevented by protons countertransported down unaffected ΔpH.

CCCP also doubled ¹⁰⁹Cd uptake by strain 17810R (Fig. 3a), although Cd²⁺ accumulation in cytoplasm was stopped (Fig. 3d), since Δψ for Cd²⁺ transport through the channel was blocked by CCCP. This means that two external Cd²⁺ could bind without energy to two ψ_s sites in Cd²⁺-ATPase channel (Fig. 3d), prevented by Mg²⁺ (Fig. 3a). These CCCP data strongly confirm existence of two low affinity ψ_s sites in Cd²⁺-ATPase channel of strain 17810R.

Restoration of Cd²⁺ resistance in S. aureus 17810R by extrusion of Cd²⁺ preaccumulated in 1 mM phosphate buffer, pH 7 (low P_iB).

Cd²⁺-preloaded cells of strain 17810R in low P_i were washed and resuspended in high P_iB. Since in these Cd²⁺-poisoned cells of strain 17810R the NADH production was blocked and consequently its oxidation, therefore ∆μ ⁺_H generation and Cd²⁺ extrusion via Cd²⁺/H⁺ exchange were also stopped. Therefore, under such conditions only the Cd²⁺-insensitive proton pump—the reversed F_oF₁-ATPase could provide protons for Cd²⁺-ATPase to start Cd²⁺ extrusion from dithiols in ODHC. We suggest that these protons could bind easily to ψ_s of Cd²⁺-ATPase channel in washed cells, since there was no extracellular Cd²⁺ to compete. Finally, protons countertransported through the channel down Δψ displaced Cd²⁺ from dithiols in ODHC, which was evidenced by undisturbed ODHC activity (5.6 ± 0.8 nmol NADH/min/mg protein).

Cd²⁺ extrusion was inhibited in 50 % by DCCD, blocking H⁺ channel of F_oF₁-ATPase and also by CCCP or valinomycin + K⁺ collapsing Δψ (Fig. 4b). This suggests that Cd²⁺ could be removed from cell wall and ψ_s of strain 17810R without energy, but Cd²⁺ extrusion from dithiols in ODHC requiring H⁺ and Δψ, was stopped as evidenced by inhibited ODHC activity with all three compounds (from 1.4 ± 0.2 to 1.5 ± 0.3 nmol NADH/min/mg protein). Only nigericin collapsing ΔpH, allowed Cd²⁺ extrusion down undisturbed Δψ (Fig. 4b). Since CCCP or valinomycin + K⁺ prevented Δψ-dependent proton countertransport through the channel and also stopped Cd²⁺ extrusion, this suggests that dithiols in ODHC (S^–S⁻) may function in glutamate-linked energy conservation process probably as Δψ generation site.

Fig. 4

¹⁰⁹Cd efflux from washed cells of S. aureus 17810R preloaded with ¹⁰⁹Cd in 1 mM phosphate buffer, pH 7 (low P_iB + glutamate. (a) Cd²⁺-preloaded, washed cells were resuspended in 1 mM P_iB + glutamate (filled circles) or in 100 mM P_iB + glutamate (filled squares). (b) Cd²⁺-preloaded, washed cells were resuspended in 1 mM P_iB (filled circles) or in 100 mM P_iB + 10 μM CCCP (filled squares), 100 mM P_iB + 5 μM valinomycin + 50 mM K⁺ (filled triangles), 100 mM P_iB + 100 μM DCCD (empty diamonds) or 100 mM 100 mM P_iB + 0.5 μM nigericin (empty triangles), to each suspension glutamate was added. (c) Cd²⁺-preloaded, washed cells were resuspended in 100 mM Tris/HCl, pH 7.2 (filled squares), 100 mM MOPS/NaOH, pH 7 (empty triangles) or in 100 mM triethanolamine/phosphate, pH 7 (filled triangles), to each buffer glutamate was added. (d) Cd²⁺-preloaded, washed cells were suspended in 1 mM P_iB (filled circles) or in 100 mM triethanolamine/phosphate buffer + 10 μM CCCP (filled triangles), 100 mM triethanolamine/phosphate buffer + 5 μM valinomycin + 50 mM K⁺ (empty triangles) or 100 mM triethanolamine/phosphate buffer + 0.5 μM nigericin (filled squares), to each suspension glutamate was added

According to Fig. 4c, other 100 mM buffers containing glutamate-Tris/HCl, pH 7.2 or MOPS/NaOH, pH 7.0, did not initiate Cd²⁺ extrusion from washed cells of strain 17810R. Only 100 mM triethanolamine/phosphate buffer pH 7, triggered total Cd²⁺ efflux (Fig. 4c) sensitive to CCCP or valinomycin + K⁺ in 50 %, but insensitive to nigericin (Fig. 4d). These data strongly confirm requirement of high P_i and of P_i-dependent protons for net Cd²⁺ extrusion by Cd²⁺-ATPase.

Cd²⁺ efflux triggered by 100 mM P_iB from unwashed cells of strain 17810R was incomplete (Fig. 5a), since only Cd²⁺ from cell wall and ψ_s could be released without energy, but not Cd²⁺ from cytoplasm requiring protons and Δψ, as evidenced by blocked ODHC activity (1.4 ± 0.3 nmol NADH/min/mg protein). Only high P_iB plus Mg²⁺—preventing external Cd²⁺ countertransport down Δψ through the channel, allowed Cd²⁺ extrusion from dithiols in ODHC via Cd²⁺/H⁺ exchange (Fig. 5a), as evidenced by unblocked ODHC activity (5.5 ± 0.8 nmol NADH/min/mg protein). This Cd²⁺ efflux was equally affected by ionophores (Fig. 5b), as that from washed cells (Fig. 4b).

Fig. 5

¹⁰⁹Cd efflux from unwashed cells of S. aureus 17810R preloaded with ¹⁰⁹Cd in 1 mM phosphate buffer, pH 7 (low P_iB) + glutamate. (a) At the time indicated by an arrow, P_iB concentration was increased from 1 mM to 100 mM (filled squares), to one portion of suspension 5 mM Mg²⁺ was added (filled triangles), cells suspended in 1 mM P_iB (filled circles). (b) At the time indicated by an arrow, P_iB concentration was increased from 1 mM to 100 mM with the following additions: 5 μM valinomycin + 50 mM K⁺ (empty triangles), 10 μM CCCP (filled squares) or 0.5 μM nigericin (filled triangles), to each suspension 5 mM Mg²⁺ was added; cells suspended in 1 mM P_iB (filled circles)

Discussion

Bacterial Cd²⁺-ATPases belong to the superfamily of P-type ATPases and to the P1 subfamily of soft metal ions pumps (Kühlbrandt 2004). Studies on Cd²⁺-ATPase in S. aureus (Nucifora et al. 1989; Silver et al. 1989; Tsai et al. 2002) and in Listeria monocytogenes (Bal et al. 2003; Wu et al. 2006a) established amino acid sequence of CadA protein, its membrane topology and suggested involvement in Cd²⁺ extrusion of four cysteine residues present in this protein, but the mechanism of how Cd²⁺ is extruded from Cd²⁺-resistant S. aureus remains so far unknown. Also studies on Cd²⁺-ATPase in other microorganisms did not explain this mechanism (Schwager et al. 2012; Schurig-Briccio and Gennis 2012; Chien et al. 2013; Maynaud et al. 2014). As found here, the cadA-coded Cd²⁺ efflux system in S. aureus 17810R described by Tynecka et al. (1981a, 1981b), appeared to be the P-type Cd²⁺-ATPase. Figure 6 presents a proposed scheme for Cd²⁺ extrusion via Cd²⁺/H⁺ exchange mechanism by the native Cd²⁺-ATPase in S. aureus 17810R oxidizing glutamate in high P_iB.

Fig. 6

The proposed mechanism of Cd²⁺ extrusion by P-type Cd²⁺-ATPase from S. aureus 17810R oxidizing glutamate in 100 mM phosphate buffer, pH 7 (high P_iB). Two Cd²⁺ transported via Mn²⁺ uniporter down membrane potential (∆ψ) are trapped by high affinity sites—dithiols located in cytoplasmic domain of CadA protein, recognized by Silver and coworkers (Nucifora et al. 1989; Silver et al. 1989). SCdS formation may trigger phosphorylation of CadA protein by ATP, which changes SCdS orientation from facing cytoplasm into facing transmembrane channel. Cd²⁺ trapping stops Cd²⁺ transport towards dithiols in 2-oxoglutarate dehydrogenase complex (ODHC)—the only Cd²⁺-sensitive targets in glutamate-linked energy conservation system, functioning most likely as ∆ψ generation site S⁻S⁻. This allows undisturbed NADH production, its oxidation and ∆μ ⁺_H generation. Consequently, increased number of P_i-dependent protons pumped electrogenically via respiratory chain in high P_iB compete with external Cd²⁺ and bind to low affinity surface sites (ψ_s) of Cd²⁺-ATPase channel. Finally, protons countertransported down ∆ψ through the channel, extrude two trapped cytoplasmic Cd²⁺ into extracellular space via ∆ψ-dependent Cd²⁺/H⁺ exchange, rendering host cells Cd²⁺-resistant, since the toxic Cd²⁺ could not reach the primary targets—dithiols in ODHC either via Mn²⁺ uniporter or via transmembrane channel

We propose that two Cd²⁺transported by strain 17810R down Δψ via Mn²⁺ uniporter in high P_iB are trapped by high affinity sites—dithiols in cytoplasmic domain of staphylococcal Cd²⁺-ATPase, which were recognized by Silver and coworkers (Nucifora et al. 1989; Silver et al. 1989). We suggest that Cd²⁺ trapping (SCdS) stops its transport towards dithiols in cytoplasmic ODHC, allowing undisturbed NADH production, its oxidation via respiratory chain and ∆μ ⁺_H generation, while ATP could change SCdS orientation from facing cytoplasm towards facing transmembrane channel, probably by a tilting mechanism suggested for Ca²⁺-ATPase (Albers 1967; Post et al. 1972; Higgins and Linton 2001). Finally, increased number of P_i-dependent protons pumped electrogenically via respiratory chain, could compete with external Cd²⁺ for binding to ψ_s of Cd²⁺-ATPase. Then, entering the channel, protons displaced from high affinity sites the two trapped cytoplasmic Cd²⁺, which were transferred through the channel towards low affinity sites ψ_s, being then extruded into extracellular space via Cd²⁺/H⁺ exchange against electrochemical and concentration gradients (Tynecka et al. 1981a, 1981b), rendering host cells Cd²⁺-resistant. In Listeria monocytogenes (Wu et al. 2006b) the cadA-coded Cd²⁺-ATPase also extruded two Cd²⁺.

According to the proposed concept (Fig. 6), the P_i-dependent protons and Cd²⁺ binding ligands in Cd²⁺-ATPase channel of strain 17810R seem to play vital role in Cd²⁺ extrusion. We suggest that the negative charges exposed in the channel by successive proton movement to ψ_s, CysProCys and CysCys (recognized by Nucifora et al. 1989; Silver et al. 1989; Tsai et al. 2002) can be the driving force for uphill Cd²⁺ pumping through the channel into opposite direction–the extracellular environment. Also Ca²⁺ efflux by Ca²⁺-ATPase from mammalian mitochondria was P_i-dependent (Roos et al. 1980; Nicholls and Akerman 1982; Ligeti and Lukács 1984). Proton requirement for Cd²⁺ extrusion in strain 17810R, is in accord with Scarbourough`s considerations (Scarborough 2003)—that after phosphorylation reaction, something must weaken ion binding site, allowing Ca²⁺ release into extracellular space. However, the role of protons in SR Ca²⁺-ATPase has been controversial for many years (Ueno and Sekine 1981; Levy et al. 1990; Andersen and Vilsen 1995; Karjalainen et al. 2007; Toyoshima 2009; Fibich and Apell 2011; Bublitz et al. 2013).

Our earlier observations (Tynecka et al. 1981a, 1981b) suggested that protons and external Cd²⁺ could compete for entry into cytoplasm through the channel of Cd²⁺-ATPase down Δψ. Our present data confirm that due to decreased number of P_i-dependent protons in low P_iB the external Cd²⁺ could bind to ψ_s of Cd²⁺-ATPase channel and then driven down Δψ through the channel via Cd²⁺/Cd²⁺ exchange, blocks dithiols in ODHC, rendering host cells Cd²⁺-sensitive, like strain 17810S.

However, we found that Mg²⁺ can protect strain 17810R against Cd²⁺ poisoning in low P_iB. According to William’s model (Williams 1978) and our CCCP data, Mg²⁺ can prevent external Cd²⁺ binding to ψ_s and stops its countertransport through the channel towards dithiols in ODHC. In energized cells, Mg²⁺ can be displaced transiently by respiratory protons, but still prevents external Cd²⁺ countertransport. Therefore, even a decreased number of P_i-dependent protons pumped electrogenically during glutamate oxidation in low P_iB, but protected by Mg²⁺, could enter the channel to extrude two trapped cytoplasmic Cd²⁺ via energy-dependent Cd²⁺/H⁺ exchange. Discharge of ∆μ ⁺_H by protons allowed Mg²⁺ return to ψ_s. Such Mg²⁺ oscillation can maintain undisturbed NADH production, its oxidation, energy conservation and Cd²⁺ resistance, despite low P_iB. Also Ca²⁺ influx and Δψ disruption in mitochondria were prevented by Mg²⁺ (Sharikabad et al. 2001; Racay 2008).

Net Cd²⁺ extrusion requires also steady-state thermodynamic equilibrium between activities of two energy-dependent membrane systems—Mn²⁺ uniporter and Cd²⁺-ATPase. Some changes, e.g. increased Cd²⁺ concentration (Tynecka et al. 1981b), alkaline pH (Tynecka et al. 1981a) or decreased P_iB concentration shown here, disturb equilibrium and consequently Cd²⁺ resistance. Therefore, the Cd²⁺-ATPase cooperating with P_i-dependent protons and utilizing cellular energy (ATP and ∆μ ⁺_H ) can protect against Cd²⁺ poisoning the vital dithiols in ODHC.

However, we found that Cd²⁺, which blocked dithiols in ODHC in low P_iB could be also extruded. As was already mentioned, in these Cd²⁺-poisoned cells of strain 17810R, only the Cd²⁺-insensitive, reversed F_oF₁-ATPase could pump protons to start the Cd²⁺ efflux process. Besides, we increased the P_iB concentration to 100 mM (high P_iB) and also inhibited external Cd²⁺ countertransport down Δψ through the channel either by cell washing or by Mg²⁺ pretreatment. In both situations, the countertransport of P_i-dependent protons through the channel was restored, leading to Cd²⁺ displacement from dithiols in ODHC. We suggest that the displaced Cd²⁺ could be trapped by high affinity sites of Cd²⁺-ATPase, forming SCdS, while ATP could change SCdS orientation towards facing transmembrane channel. Now, Cd²⁺ displaced by protons from high affinity sites via Cd²⁺/H⁺ exchange moves towards low affinity sites (ψ_s) of Cd²⁺-ATPase. From here, Cd²⁺ is displaced into extracellular space also via Cd²⁺/H⁺ exchange. Gradual Cd²⁺ extrusion by Cd²⁺-ATPase restored gradually: NADH production, its oxidation, ∆μ ⁺_H generation via respiratory chain, reversal of F_oF₁-ATPase into biosynthetic direction and energy conservation, rendering host cells again Cd²⁺-resistant. DCCD—blocking H⁺ channel of F_oF₁-ATPase or CCCP and valinomycin + K⁺ collapsing Δψ, prevented Cd²⁺ extrusion, confirming the requirement of P_i-dependent protons and of Δψ for the Cd²⁺ efflux process.

To summarize, these studies provide for the first time the novel data on the so far unknown mechanism of Cd²⁺ extrusion by cadA-coded P-type Cd²⁺-ATPase in S. aureus 17810R, oxidizing glutamate in high P_iB. Energy-dependent Cd²⁺ extrusion by this pump via Cd²⁺/H⁺ exchange mechanism renders host cells Cd²⁺-resistant, since the toxic Cd²⁺ could not reach the primary Cd²⁺-sensitive targets—dithiols in ODHC via two routes—Mn²⁺ uniporter or transmembrane channel, allowing undisturbed glutamate-linked energy conservation process. Moreover, the vital role of P_i-dependent protons or Mg²⁺ and of cellular energy (ATP and ∆μ ⁺_H ) in Cd²⁺ extrusion by Cd²⁺-ATPase is underlined.