Abstract
Numerous strategies allowing bacteria to detect and respond to oxidative conditions depend on the cell redox state. Here we examined the ability of Lactococcus lactis to survive aerobically in the presence of the reducing agent dithiothreitol (DTT), which would be expected to modify the cell redox state and disable the oxidative stress response. DTT inhibited L. lactis growth at 37 °C in aerobic conditions, but not in anaerobiosis. Mutants selected as DTT resistant all mapped to the pstFEDCBA locus, encoding a high-affinity phosphate transporter. Transcription of pstFEDCBA and a downstream putative regulator of stress response, phoU, was deregulated in a pstA strain, but amounts of major oxidative stress proteins were unchanged. As metals participate in oxygen radical formation, we compared metal sensitivity of wild-type and pstA strains. The pstA mutant showed approximately 100-fold increased resistance to copper and zinc. Furthermore, copper or zinc addition exacerbated the sensitivity of a wild-type L. lactis strain to DTT. Inactivation of pstA conferred a more general resistance to oxidative stress, alleviating the oxygen- and thermo-sensitivity of a clpP mutant. This study establishes a role for the pst locus in metal homeostasis, suggesting that pst inactivation lowers intracellular reactivity of copper and zinc, which would limit bacterial sensitivity to oxygen.
Edited by: D. J. Jamieson
INTRODUCTION
Oxygen and derivative radical oxygen species (ROS) such as superoxide anion radicals, hydrogen peroxide (H2O2), or hydroxyl radicals are toxic for cells. In the reducing environment of the cytosol, they act as oxidative agents and their accumulation may lead to growth arrest in the absence of cell protection. Detoxification enzymes (e.g. catalase, superoxide dismutase; Imlay, 2008), which remove oxygen or ROS, constitute a major means of escape from oxidative stress conditions. Oxygen toxicity is intimately associated with metals. Reduced metals, in particular iron and copper, can convert H2O2 to highly toxic hydroxyl radicals (Fenton reaction). However, other metals can increase oxidative stress independently of the Fenton reaction. For example, zinc is not redox-active by itself but interacts with redox-active cysteines in bacterial proteins (such as Hsp333; Graf & Jakob, 2002). It has been shown that zinc deficiency or overload can lead to oxidative stress, whereas intermediate cellular concentrations of zinc have an antioxidant effect (Maret, 2006).
Oxidative stress is also exacerbated by high temperature. Increased temperature triggers an oxidative burst of superoxide, which causes a profound loss of Escherichia coli viability (Benov & Fridovich, 1995). Thus, a connection between response to oxidative stress and heat shock is expected. As such, defects in Clp proteins, chaperones or RecA dampen resistance to oxidative stress in bacteria (Duwat et al., 1995; Frees et al., 2003; Robertson et al., 2002).
Bacteria are programmed to coordinate numerous strategies to respond to damage induced by the presence of oxygen. Interestingly, stress sensors rely on reactivity between oxygen or metal and the sensor itself. Oxidation of cysteines in disulfide-bond-containing proteins (RSH→RSSR), redox balance of cofactors (e.g. iron–sulfur clusters), or metal binding, allows regulators to be active and thereby govern transcriptional expression of genes under their control. Note that one stress protein may belong to several regulons, and thus respond to several stress conditions. This is exemplified by SodA, which is induced by metal, heat, and oxidative or acid stress (Budin-Verneuil et al., 2005).
Lactococcus lactis is a Gram-positive mesophilic bacterium of industrial interest, related to its uses in cheese production (O'Connor, 2007), and recently in novel bioprotein delivery strategies (see Steidler & Rottiers, 2006). Several L. lactis defence proteins have been characterized, and its genome contains stress protein homologues, including RecA (Duwat et al., 1995), Mn-SodA (Sanders et al., 1995), TrmA (Turner et al., 2007), the thioredoxin reductase–thioredoxin system (Vido et al., 2005) and a glutathione reductase (Li et al., 2003), although L. lactis reportedly does not synthetize glutathione. L. lactis, which is generally known for its fermenting capacities, can also undergo aerobic respiration when haem is provided (Duwat et al., 2001; Gaudu et al., 2002, 2003; Vido et al., 2004); in this condition, respiration chain activity eliminates oxygen and decreases the occurrence of ROS (Rezaiki et al., 2004). Analyses of the L. lactis genome revealed that conserved CXXC motifs are present in numerous L. lactis ORFs encoding repair or stress response proteins, suggesting that these factors might be modulated by the oxidative state of the cytoplasm.
In this study, we examined the ability of L. lactis to survive aerobically in the presence of dithiothreitol (DTT), a membrane-diffusible thiol reducing agent. DTT would be expected to modify the cell redox state and disable the oxidative stress response. Mutants conferring resistance to DTT were found to map exclusively to genes in the pstFEDCBA locus (referred to as the pst locus) involved in phosphate uptake. Our results indicate that pstA inactivation impacts on copper and zinc toxicity. We explain these effects by the modified homeostasis of metals which may limit bacterial sensitivity to oxygen.
METHODS
Bacterial strains, plasmids, and general growth conditions.
Bacterial strains and plasmids are listed in Supplementary Table S1, available with the online version of this paper. L. lactis strain MG1363 (wild-type, WT), or strain MG1363 carrying low-copy plasmid pIL252 conferring EryR (referred to as WT-Ery) was used as control. Details of strain constructions are described below. L. lactis strains were grown at 30 °C under static growth conditions in M17 liquid medium (Difco) supplemented with 1 % glucose (GM), or on solid GM that contained 1.5 % agar. M17 modified medium (GMm), containing 1 % glucose, but lacking β-glycerophosphate and beef extract (and therefore essentially phosphate-free), and buffered to pH 7 with 200 mM MOPS was used as indicated. Plates were incubated aerobically or under anaerobic conditions in jars (containing the ‘GENbox anaer’ generator for culture of anaerobic bacteria, bioMérieux). Erythromycin (Ery) was used at 2.5 or 1 μg ml−1. DTT was added to solid medium at stated concentrations. Potassium phosphate (20 mM) was added to solid or liquid medium where indicated. Growth curves for the WT and pstA : : Ery strains were obtained by diluting overnight cultures 1/1000 in GMm medium supplemented or not with 20 mM K2HPO4 and monitoring growth for 24 h using a plate reader (EL808, BioTek Instruments).
DTT sensitivity tests.
To determine the level of L. lactis sensitivity to DTT, the WT strain was grown in GM to OD600 0.05, then shifted to 37 °C for 1 h under static growth conditions, and finally plated aerobically on GM containing different DTT concentrations at 35 °C or 37 °C. Selection conditions were determined by finding the lowest DTT concentrations that led to near-total mortality of WT-Ery in the presence of Ery (1 μg ml−1), as seen by the absence of c.f.u. after 48 h: 30 mM DTT at 35 °C or 25 mM DTT at 37 °C.
DTT resistance of the pstA : : Ery mutant was assessed by growing mutant and control strains to OD600 0.2 under static conditions in GMm containing 20 mM K2HPO4 and 1 μg Ery ml−1 at 30 °C, and then plating at 37 °C on GMm plates containing 20 mM K2HPO4 and 35 mM DTT. C.f.u. were assessed after 48 h incubation.
DTTR mutant isolation by insertional mutagenesis.
MG1363 containing the thermosensitive pGhost9 : : ISS1 was grown at 30 °C to OD600 0.05, shifted to 37 °C for 1 h, and then plated at 35 °C or 37 °C on GM plates containing Ery (2.5 μg ml−1), and either 30 mM (at 35 °C) or 25 mM (at 37 °C) DTT. Plates were incubated aerobically for 48 h. The occurrence of DTT-resistant mutants was around 0.1 % of the total number of EryR mutants obtained without DTT selection. Colonies that appeared after 48 h were restreaked and insertions were characterized. pGhost9 was excised from the chromosome, leaving a single copy of ISS1 (Maguin et al., 1996) to obtain stable mutants.
Protein expression
1D gel electrophoresis.
Protein expression levels were assessed by growing the WT and pstA : : ISS1 strains to OD600 0.4 under static conditions in GM. Cultures were then shifted at 37 °C with aeration (230 r.p.m.) in the presence or absence of 30 mM DTT, for 1 h 30 min. Cells were then collected for protein extraction (Guillot et al., 2003) and proteins were subjected to denaturing 1D gel electrophoresis following standard protocols.
2D gel electrophoresis.
Protein expression levels were assessed by growing the WT-Ery and pstA : : Ery strains to OD600 0.2 under static conditions in GMm containing 20 mM K2HPO4 and 1 μg Ery ml−1. Cultures were then shifted to 37 °C with aeration (230 r.p.m.) in the presence or absence of 30 mM DTT, until they reached OD600 1.5 (approx. 2 h). Cells were then collected for protein extraction (Guillot et al., 2003). 2D gels (Vido et al., 2004) were repeated twice.
H2O2 sensitivity tests.
The WT-Ery and pstA : : Ery strains were grown for 24 h in GMm supplemented with 20 mM K2HPO4 and 1 μg Ery ml−1. For oxidative challenge, cultures were diluted 10-fold and incubated with 10 mM H2O2 for 1 h. H2O2 was then removed by addition of bovine catalase (10 U ml−1, Sigma), and c.f.u. were determined on GMm supplemented with 20 mM K2HPO4. Reported results correspond to the mean of three assays.
Construction of the pstA : : Ery mutant and the pstA : : Ery clpP double mutant.
Single-stranded DNA oligonucleotide primers 5′-TTG ACC GCA AGG ACA CG-3′ and 5′-CAT TAC GAA TGT GCT GG-3′ were used to PCR-amplify a pstA internal 550 bp fragment from the MG1363 chromosome. DNA fragments were treated with T4 and Klenow polymerases and cloned into SmaI-linearized pRV300 (Leloup et al., 1997), giving pPstAint (Supplementary Table S1). The chromosomal pstA gene was then inactivated by single-crossover recombination. To do this, L. lactis MG1363 electrocompetent cells were transformed with pPstAint, selecting on GM plates containing Ery. The same strategy was used to construct a clpP pstA double mutant, starting from the markerless clpP strain (Frees & Ingmer, 1999). Strain constructions were verified by Southern hybridization using pRV300 containing the pstA fragment as probe.
Metal sensitivity assays.
The WT-Ery and pstA : : Ery strains were grown overnight at 30 °C in GM, and dilutions were spotted (5 μl of 10−1 to 10−5 dilutions) on GM plates containing or not different metals, as indicated in the text. Plates were incubated for 24 h at 30 °C and photographed.
To test the effects of metal addition on DTT toxicity, dilutions (5 μl spots) of overnight WT-Ery and pstA : : Ery cultures (grown at 30 °C in GM) were deposited on solid GM containing or not 0.1 mM Cu, 1 mM Zn and 10 mM DTT. All plates were incubated at 37 °C for 48 h and then photographed.
RESULTS AND DISCUSSION
L. lactis is sensitive to DTT under aerobic, but not anaerobic growth conditions at elevated temperature
To explore the role of thiol stress in L. lactis survival, we examined MG1363 (WT) survival in the presence of different concentrations of the thiol reducing agent DTT added to solid medium in aerobic conditions. We expected that DTT addition during aerobic growth might prevent cells from responding to oxidative stress; DTT might also cause cell damage by reducing oxygen to form ROS. We found that L. lactis plating efficiency was unaffected by concentrations of up to 50 mM DTT at 30 °C (data not shown). However, at 37 °C the number of c.f.u. was reduced at least 103-fold in the presence 25 mM DTT (the optimum growth temperature of L. lactis is around 30 °C). This result suggested that the combination of DTT and high temperature is toxic for aerobically grown lactococci.
The contribution of oxygen to mortality due to DTT was evaluated by comparing plating efficiencies of WT cells grown in the presence of DTT, in aerobic versus anaerobic conditions (Table 1⇓). In contrast to the >103-fold drop in viability in aerobic conditions, full viability was observed after incubation in anaerobic conditions at 37 °C. Oxygen is thus an important factor in DTT-mediated toxicity at high temperature.
DTT resistance of the pstA : : Ery mutant in aerobic conditions at high temperature
WT-Ery and pstA : : Ery strains were grown under static liquid growth conditions in GMm containing 20 mM K2HPO4 to mid-exponential phase (OD600 0.2–0.3). Dilutions were plated on solid GMm medium containing 20 mM K2HPO4, with or without 30 mM DTT. C.f.u. were determined after 48 h incubation. Results are the means of three independent experiments.
A proteomic approach was used to examine expression levels of known oxidative stress response proteins of aerobic 37 °C cultures of L. lactis grown without or with DTT. In the presence of DTT at 37 °C, levels of SodA, thioredoxin reductase and the peroxide-detoxifying enzyme alkylhydroperoxide reductase (encoded by sodA, trxB1, and ahpC, respectively) were clearly decreased compared to those observed in cells grown in the same conditions without DTT (Fig. 1⇓). In the case of sodA, lower expression in cultures grown with DTT was also confirmed at the transcriptional level by Northern blotting (data not shown). These results are in keeping with our initial hypothesis, that DTT impairs the cellular response to oxidative stress. They are also supported by a study in E. coli indicating that another thiol agent, homocysteine, has a damping effect on oxidative stress response (Fraser et al., 2006). We did not observe appreciable differences in the amounts of heat-shock proteins DnaK and GroEL, despite the elevated growth temperature; this may reflect the reported transient nature of the heat-shock response (Arnau et al., 1996). These results suggest that DTT disables the normal oxidative stress response of L. lactis.
Oxidative stress response protein levels are significantly decreased in the presence of DTT. Protein extracts of WT L. lactis grown with and without DTT were subjected to 2D gel proteome analysis. Levels of known oxidative stress response proteins SodA, TrxB1 and AhpC are decreased in cells grown in the presence of DTT. The FbaA protein was used as reference to compare total protein amounts on the gels. Two independent experiments were performed; a representative gel is shown.
DTT-resistant insertional mutants all map in the pst locus
We selected for transposition insertion mutants that could grow aerobically in the presence of DTT. Two independent pG+host9 : : ISS1 insertional mutageneses of MG1363, performed in the presence of DTT at 35 °C or 37 °C, resulted in the selection of 20 mutant colonies. DTT-resistant phenotypes were reconfirmed on purified mutants, and insertion sites were identified by chromosomal junction cloning and DNA sequencing. All 20 insertions were independent, and all interrupted the pst locus, comprising the pstFEDCBA genes. The precise insertion points of 10 mutants were determined by sequencing the chromosomal junctions: they all mapped in pstA, pstB, pstC, pstD and pstE (Fig. 2⇓). This locus shows strong similarity to the E. coli pst locus, which encodes a high-affinity phosphate carrier (Wanner, 1996). The pst locus is followed by phoU, encoding a homologue of the E. coli PhoU protein. In E. coli PhoU is a negative regulator of the pho regulon (Wanner, 1996; Steed & Wanner, 1993) and was shown to be a global negative regulator that shuts down several hundred genes involved in bacterial metabolism (Li & Zhang, 2007).
DTT-resistant mutants all map in genes of the pst locus. Sites of pGhost9 : : ISS1 insertions in the pst locus are shown by open triangles and putative promoters by arrows. L. lactis pstE (also called pstS: Rallu et al., 2000), and pstF are both homologues of E. coli pstS. L. lactis pstC and pstD are homologues of E. coli pstA and pstC respectively. L. lactis pstA and pstB are homologues of E. coli pstB.
Stabilized excision mutants of two of the initial pG+host9 : : ISS1 DTT-resistant isolates corresponding to insertions in pstA and pstE were constructed, and the DTT-resistant phenotype at 37 °C was confirmed on both. To avoid possible secondary mutations during selection, an independent pstA strain containing an erythromycin resistance marker was constructed by single-crossover mutation (pstA : : Ery). This strain showed DTT resistance (Table 1⇑), indicating that the observed phenotype was not due to a secondary event during selection.
Physiological characterization of the pstA : : Ery mutant
To confirm that the pst locus was involved in phosphate transport in L. lactis, growth of the WT strain containing an erythromycin resistance marker (WT-Ery) and the pstA : : Ery strain was compared in phosphate-limiting medium (GMm, see Methods) and phosphate-rich medium (GMm supplemented with 20 mM K2HPO4) (Fig. 3⇓). Under phosphate-limiting conditions, the pstA : : Ery mutant showed a lower growth rate than the WT-Ery strain. In contrast, pstA : : Ery grew nearly as well as WT-Ery in phosphate-rich medium, probably due to activity of the low-affinity phosphate transport system predicted from the genome sequence. These results are in keeping with a role for the pst locus in phosphate uptake at low-phosphate concentrations in L. lactis.
Phosphate addition alleviates growth lag of the DTT-resistant pstA : : Ery mutant. Overnight cultures of WT and pstA : : Ery strains were diluted 1/1000 and growth in GMm (phosphate free) or GMm plus 20 mM K2HPO4 was monitored for 24 h. Results are means of at least two experiments.
Rallu et al. (2000) reported that acid stress resistance of a pstE mutant (also called pstS) of L. lactis was due to decreased intracellular phosphate concentration. We therefore examined the influence of phosphate depletion on DTT resistance. We compared WT-Ery and pstA : : Ery cultures for survival in aerobic conditions in the presence of DTT on GMm plates containing or not 20 mM phosphate (Table 2⇓). The pstA mutant remained highly DTT resistant in the presence of excess phosphate. It is thus likely that a low cytoplasmic phosphate concentration is not the cause of DTT resistance of the pstA mutant.
DTT resistance of pstA : : Ery is independent of phosphate concentration
WT and pstA : : Ery strains were grown under static liquid growth conditions in GMm with or without 20 mM K2HPO4, to mid-exponential phase (OD600 0.2–0.3). Dilutions were plated on solid GMm medium with or without 20 mM K2HPO4, and 30 mM DTT. C.f.u. were determined after 48 h incubation. Results are the means of three independent experiments.
Mutants in the pst locus were reported as being resistant to H2O2 (Rallu et al., 2000). We checked whether a pstA mutant was resistant to oxidative stress independently of the presence of DTT. The ability of WT-Ery versus pstA : : Ery cells to survive an H2O2 shock at 30 °C was examined in GMm liquid medium containing 20 mM phosphate. The pstA : : Ery mutant showed fivefold greater survival than the control strain upon 1 h exposure to 10 mM H2O2 in stationary phase. Thus, pstA inactivation results in greater resistance to H2O2-provoked oxidative stress.
Stress response protein expression is not induced by pstA inactivation
We compared protein expression in response to DTT treatment of WT and pstA strains grown aerobically at 37 °C, by 1- and 2D gel electrophoresis. Proteins were identified by MALDI-TOF mass spectrometry. We observed a strong expression of PstF, a putative lipoprotein component of the high-affinity phosphate transport system, which was also confirmed by transcriptional analysis (see supplementary material). A similar finding was recently reported in a proteomic study of lactococcal pstS mutant (Budin-Verneuil et al., 2007). However, other than this difference, protein expression in the pstA : : Ery mutant and WT-Ery strains was similar, in both normal and DTT-stress conditions (data not shown). In particular, in the presence of DTT, the low levels of oxidative stress response proteins SodA, TrxB1 and AhpC remained low in both WT and pstA : : Ery strains. We conclude that the pstA mutation has little, if any, impact on the induction of several characterized oxidative stress response genes.
The pstA : : Ery mutant is resistant to copper and zinc
Studies in Saccharomyces cerevisiae and E. coli suggested that a phosphate transporter could be implicated in transport of metals such as manganese, zinc and cobalt (Jensen et al., 2003; van Veen et al., 1994). Metals catalyse ROS production (Schutzendubel & Polle, 2002; Teitzel et al., 2006) and disulfide bond formation (Hiniker et al., 2005; Maret, 2006). It was thus possible that pst inactivation confers oxygen resistance in the presence of DTT by affecting metal availability. We compared the metal sensitivities of L. lactis WT-Ery and pstA : : Ery strains. The pstA : : Ery mutant showed approximately 100-fold greater survival in the presence of copper and zinc compared to the control strain (Fig. 4⇓). No differences were observed in the presence of nickel, cobalt, iron or manganese (data not shown). These results suggest that pstA has an impact on copper and zinc homeostasis.
The pstA mutation confers resistance to zinc and copper. Tenfold dilutions of WT-Ery and pstA : : Ery overnight cultures were spotted on GM plates containing or not 0.5 mM copper chloride or 7 mM zinc chloride. Plates were incubated 24 h at 30 °C and photographed. ud, undiluted.
DTT sensitivity of the WT strain is exacerbated by copper or zinc
We hypothesized that greater copper and zinc tolerance of the pstA strain could be implicated in DTT resistance. The effect of copper or zinc addition on the sensitivity of the WT and pstA : : Ery strains to DTT at 37 °C was tested on solid medium. In these tests, we lowered the amount of DTT to 10 mM (from 25 mM above). The presence of 0.1 mM copper or 1 mM zinc alone had no effect on survival of either the WT or the pstA : : Ery strain. However, the combination of each metal and DTT had a marked effect on the WT, compared to the pstA : : Ery strain (Fig. 5⇓). These results indicate that DTT toxicity is exacerbated in the presence of copper and zinc in the WT strain. We propose that a pstA mutation would result in altered homeostasis of free metal loads in the cell. In the case of copper, decreased intracellular free metal levels would result in lower ROS formation, thereby improving survival in the presence of DTT. In the case of zinc, which is not directly redox-active by itself, we assume that the pst mutation results in a modified balance between redox-active cysteines and zinc–cysteine complexes in proteins (Maret, 2006), which would limit DTT toxicity. A role of pst in metal homeostasis would not require induction of oxidative stress response proteins, in keeping with our observations.
Copper or zinc addition exacerbates DTT sensitivity of WT L. lactis. Dilutions of saturated cultures of WT-Ery (left) and pstA : : Ery (right) strains were spotted on solid GM, containing combinations of the following compounds as indicated in the figure: 10 mM DTT, 0.1 mM CuSO4 (Cu), and 1 mM ZnSO4 (Zn). Plates were incubated for 48 h at 37 °C and then photographed. Each spot shown corresponds to 2.5×104 c.f.u.
The pstA mutation relieves thermosensitivity of a clpP mutation
If pstA inactivation affects free metal availability and lessens DTT sensitivity, we predicted that it might have similar effects in different situations where thiol homeostasis is modified. This expectedly occurs in a clpP mutant. In Bacillus subtilis, a clpP mutant was shown to accumulate thioredoxin reductase (TrxB), thioredoxin (TrxA) and thiol peroxidase (Tpx) at high temperature (Kock et al., 2004). We reasoned that clpP inactivation would thus modulate disulfide bond formation in a way analogous to that seen when DTT is added to a WT culture. As in B. subtilis, clpP is thermosensitive in L. lactis (Frees & Ingmer, 1999). First, we observed that loss of clpP viability at 38 °C is alleviated in anaerobic conditions (Fig. 6A⇓) suggesting that the observed thermosensitivity is due at least in part to impaired oxidative stress response. [In a previous study this phenotype was not reported but tests were performed in conditions where even the WT strain did not grow (Frees et al., 2001)]. Then we asked whether pstA inactivation might rescue a clpP mutant. We constructed a clpP pstA : : Ery double mutant, and examined growth of WT, clpP, and clpP pstA : : Ery strains at 30 °C and 38 °C (Fig. 6B⇓). The pstA mutation fully restored growth of a clpP mutant at 38 °C in the presence of oxygen. These results show that pstA inactivation alleviates both clpP thermosensitivity and WT strain DTT sensitivity.
The pstA mutant allele suppresses clpP oxygen-dependent thermosensitivity. Dilutions of saturated cultures of WT, clpP (A) and clpP pstA : : Ery (B) strains were spotted on solid GM, and incubated at 30 °C or at 38 °C. Plates were incubated for 48 h and then photographed (cultures in B were incubated aerobically). ud, undiluted.
Conclusions
The physiological and genetic responses of L. lactis to DTT stress were examined. Our main findings are that DTT toxicity in lactococci is affected by (i) oxygen, (ii) the availability of metals copper and zinc, and (iii) activity of the pst locus. We suggest that the high-affinity phosphate uptake system encoded by pst genes is, in addition to its role in phosphate transport, involved in copper and zinc homeostasis. Mutants in the pst locus were recently isolated in a screening for tellurite resistance (Turner et al., 2007), suggesting that pst might also affect tellurite homeostasis and possibly that of other metals. Interestingly, results suggestive of metal transport via phosphate uptake systems were previously reported in other species (Alvarez & Jerez, 2004; Beard et al., 2000; van Veen et al., 1994, 1993), and could explain the resistance to metal toxicity. In our mutagenesis, we did not obtain mutants affected in known or putative metal transporters (Gostick et al., 1999; Scott et al., 2000; Turner et al., 2007), nor did we obtain mutants in regulators of such transporters, e.g. ZitR and the FNR-like proteins of L. lactis (Gostick et al., 1999; Llull & Poquet, 2004; Scott et al., 2000). This might suggest that the transporters are redundant or that pst mutation has a pleiotropic role affecting intracellular availability of several metals, which would not be obtained by inactivation of a single transporter.
This study establishes a role for the pst locus in copper and zinc homeostasis in L. lactis. Experiments measuring total intracellular metal loads did not reveal significant differences between WT and pstA strains, even when the growth medium was supplemented with copper or zinc (data not shown). This would suggest that pst does not control total zinc and/or copper pools. To explain this, we speculate that pst may be involved in controlling the pool of free (redox-active) metals; the existence of free and bound iron was shown in E. coli (Keyer & Imlay 1996).
Acknowledgments
We are grateful to our colleagues Isabelle Poquet, Danièle Touati, Gilbert Richarme, Philippe Bouloc, and Marie-Agnès Petit, for valuable discussion and suggestions in the course of this work. This work received financing from the ‘Programme de Microbiologie’, French Ministry of Research, France.