INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

In Salmonella typhimurium, there are three transport systems for Mg²⁺, CorA, MgtA, and MgtB. The corA gene is constitutively expressed, and the encoded 40-kDa protein product mediates both the influx and efflux of Mg²⁺ (10, 11, 23). The mgtA and mgtCB loci encode putative P-type ATPases (10, 13, 22). Although found in prokaryotes, MgtA and MgtB share greater similarity to the eukaryotic P-type ATPases, in particular to the sarcoplasmic reticulum Ca²⁺-ATPases, than to other prokaryotic P-type ATPases. Both loci are under control of the two-component PhoPQ regulatory system. PhoQ is a membrane sensor-kinase that at low Mg²⁺ concentrations phosphorylates PhoP to activate a variety of genes including mgtA and the mgtCB operon (8, 14, 15). Recent evidence has suggested, however, that there may be an additional mechanism for transcription of the mgtA locus (28). MgtA and MgtB, unlike CorA, mediate only the influx of magnesium and are produced under Mg²⁺-limiting conditions.

The mgtCB locus encodes two proteins. MgtC is a hydrophobic protein with a predicted molecular mass of 22.5 kDa (25, 28). MgtB, the P-type ATPase, has a molecular mass of 102 kDa (22, 25). Although MgtB exhibits extensive sequence homology to members of the P-type ATPase superfamily, MgtC does not exhibit significant sequence homology to any proteins of known function in the current databases. Recently, the mgtCB locus has been identified as part of a new pathogenicity island, SPI-3 (3). Pathogenicity islands are segments of DNA containing virulence genes that are found in pathogenic organisms but absent from phylogenetically related nonpathogenic organisms. The mgtC gene is not necessary for S. typhimurium invasion or short-term survival in epithelial or macrophage cells (24, 28). It is, however, essential for long-term survival within the macrophage and for virulence in mice (3). A functional allele of mgtA is not required for survival within the macrophage, while inactivation of mgtB has only a small effect on survival. Since addition of 25 mM Mg²⁺ to the macrophage growth medium rescued the ability of an mgtC mutant strain to survive long-term within the macrophage, it was suggested that MgtC may be an additional Mg²⁺ transporter in S. typhimurium (3).

In addition to a low Mg²⁺ concentration, S. typhimurium bacteria that invade epithelial and macrophage cells are also exposed to acid pH (1, 6, 7, 18). Exposure to acid also regulates gene expression of mgtA and mgtCB (2, 28). At pH 7.4, suspension of cells in N minimal medium without added Mg²⁺ induces transcription of mgtA and mgtCB more than 1,000-fold. At pH 5.2, low concentrations of Mg²⁺ fail to induce mgtA, but mgtCB induction is decreased only about twofold. If cells are adapted to pH 5.2 prior to Mg²⁺ starvation, there is no difference in mgtA or mgtCB induction properties compared to exposure at pH 7.4. Interestingly, a functional corA allele had a significant effect on transcription of these two loci. In strains lacking CorA and grown overnight at pH 5.2, induction of transcription at low Mg²⁺ levels was significantly diminished for both the mgtA and mgtCB loci. These data suggest that there is an acid component to induction of mgtA and mgtCB at low Mg²⁺ and that a functional corA allele is required for an optimal response to low Mg²⁺. In this context, it is of interest that Bearson et al. (2) have recently presented evidence that PhoPQ regulation of these loci is also responsive to acid.

To elucidate MgtC's role in a low-Mg²⁺ environment, we compared growth and Mg²⁺ transport of a strain dependent on the mgtCB operon for Mg²⁺ uptake with a strain with an intact mgtC gene but with mgtB inactivated. In Mg²⁺-starved cells, although Western blot analysis indicated that MgtC is produced and maintained, an mgtC⁺ strain exhibited no detectable uptake of ⁶³Ni²⁺ or ⁵⁷Co²⁺. This strain could grow at pH 7.4 in the absence of added Mg²⁺ only after a 24-h lag period and could not grow regardless of the level of extracellular Mg²⁺ at pH 5.2. We interpret these results to indicate that MgtC is not a functional Mg²⁺ transporter and that its expression does not facilitate survival at low pH.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Bacterial strains and plasmid construction. The molecular biology methods used were those of Sambrook et al. (21). Bacterial plasmids and strains used in this study are listed in Table 1. To create a mgtC⁺-only plasmid, pDGK1 (11a) containing mgtCB in a pBS II SK+ (Stratagene, La Jolla, Calif.) vector was digested with HindIII. The 5.1-kb fragment was religated together to give pMBM19. To create a mgtB⁺-only plasmid, pDGK1 was digested with NsiI and Bsu36I, the ends were filled in with Klenow fragment, and the 7.3-kb fragment was religated to give pMBM34.

Growth curves. N minimal medium (10) was routinely supplemented with 0.4% (vol/vol) glucose as a carbon source and 0.1% (vol/vol) Casamino Acids. Strains were grown in 5 ml of N minimal medium supplemented with 10 mM Mg²⁺ and the appropriate antibiotics overnight at 37°C. Cell pellets were washed three times in N minimal medium without added Mg²⁺ and then resuspended in 5 ml of N minimal medium without added Mg²⁺. The volume was adjusted slightly to give the same optical density at 600 nm (OD₆₀₀) for each strain. To 100 ml of N minimal medium, 500 µl of the cell sample was added, and then 7 ml of this mixture was added to each one of 10 test tubes. Ten milliliters was reserved and 200 µl of 1 M Mg²⁺ was added to give a 20 mM Mg²⁺ solution. Seven milliliters of this solution was added to the first test tube (already containing 7 ml of the cell culture), and the sample was mixed thoroughly to give a [Mg²⁺] of 10 mM. Serial dilutions of Mg²⁺ were made by removing 7 ml of culture from the first test tube and adding it to the second test tube, etc., until the final test tube contained 10 µM Mg²⁺. The tubes were incubated at 37°C, and the OD₆₀₀ was measured at various time points. For comparison of growth at neutral and acid pH, the 100 mM Tris-Cl buffer in N minimal medium was replaced with 50 mM Na-HEPES plus 50 mM Na-MES (morpholineethanesulfonic acid), and the pH was adjusted appropriately.

Transport of ⁵⁷Co²⁺ and ⁶³Ni²⁺. Strains were grown in 5 ml of N minimal medium supplemented with 10 mM Mg²⁺ and the appropriate antibiotics for 4 h at 37°C. This culture was used to inoculate 200 ml of N minimal medium containing 10 mM Mg²⁺ and grown overnight at 37°C. The cells were centrifuged at 1,000 × g for 10 min and washed three times with the same volume of N minimal medium without added Mg²⁺. Cells were resuspended in 500 ml of N minimal medium without added Mg²⁺ and grown at 37°C. At various time points, 25 ml of culture was centrifuged at 1,000 × g and washed once with N minimal medium without added Mg²⁺. Cells were resuspended in N minimal medium to give an OD₆₀₀ of 1.0 for use in the transport assay. The transport assay was performed as described previously (25).

Polyacrylamide gel electrophoresis. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis was carried out by using the buffers of Laemmli (12) and included 4.5% stacking and 12.5 or 7.5% running gels. Gels were electroblotted onto nitrocellulose, probed with the appropriate antibodies, and visualized by using donkey anti-rabbit horseradish peroxidase and enhanced chemiluminescence detection reagents (Amersham, Arlington Heights, Ill.). Antipeptide antibodies against the N-terminal 15 amino acids of MgtB and the C-terminal 16 residues of MgtC were prepared by Quality Controlled Biochemicals (Hopkinton, Mass.) and raised in rabbits. The internal control was an unknown protein that cross-reacted similarly with both the preimmune serum and serum containing the anti-MgtC antibodies. The gels shown have been scanned in Adobe Photoshop. Contrast has not been adjusted.

Disk diffusion assays for cation sensitivity. Strains were grown overnight at 37°C in Luria-Bertani (LB) broth supplemented with 10 mM Mg²⁺. Cells were washed twice with an equal amount of LB. The cell pellet was resuspended in 5 ml of LB, and 100 µl was spread onto an LB plate. A 6-mm-diameter Whatman filter paper disk was placed in the center of the plate and loaded with 15 µl of 1 M cation solution (CoCl₂, NiCl₂, ZnCl₂, MnCl₂, CaCl₂, and SrCl₂). The plates were incubated at 37°C for 24 or 48 h. Sensitivity was measured as the area of the clear growth inhibition ring minus the 18-mm² area of the disk.

Atomic absorption. A 1.0-ml volume of cells was layered over 0.3 ml of a 2:1 mixture of dibutyl and dioctyl phthalate and spun in a microcentrifuge for 30 s. The supernatant was carefully aspirated, the sides of the tube were swabbed with a cotton-tipped applicator, and 0.1 ml of 1.0 N HNO₃ was added. The sealed tubes were batch sonicated for 30 s, and the divalent cation content was measured by atomic absorption as described previously (20).

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Requirement of mgtCB for growth. As controls to confirm previous results (3) that mgtC is necessary for optimal growth at low extracellular Mg²⁺ concentrations, cells lacking a functional mgtCB operon (MM196 [mgtC mgtB]) or lacking mgtB (MM197 [mgtC⁺ mgtB]) were grown in N minimal medium with high or low Mg²⁺. Strains grown in 10 mM Mg²⁺ exhibited a similar OD₆₀₀ in the presence or absence of mgtC (Fig. 1A). When cells were grown in 10 µM Mg²⁺ (Fig. 1B), the mgtC mgtB mutant reached an OD₆₀₀ considerably lower than that of the wild-type strain. The mgtC⁺ mgtB mutant grew to a higher OD₆₀₀ than the mgtC mgtB mutant but less than that of the wild type. These results are similar to those observed by Blanc-Potard and Groisman (3), suggesting that both genes of the mgtCB locus are necessary for optimal growth at low Mg²⁺ concentrations.

View larger version (14K): [in this window] [in a new window]   FIG. 1.   Growth curves of strains possessing or lacking a functional mgtC. Strains MM1442 (wild-type LT2) (), MM281 (corA mgtA mgtC mgtB) (), MM196 (mgtC mgtB) (), and MM197 (mgtC+ mgtB) () were grown overnight in N minimal medium supplemented with 10 mM or 10 µM Mg2+. Cultures were washed three times with N minimal medium without added Mg2+ and resuspended in a volume to give the same initial optical density (OD600). The samples were diluted 1:200 in N medium containing various Mg2+ concentrations and incubated at 37°C. Aliquots were removed at the indicated times, and the OD600 was measured. Data similar to that at 10 µM Mg2+ were obtained at Mg2+ concentrations below 0.5 mM.

View larger version (12K): [in this window] [in a new window]   FIG. 2.   Growth analysis of strains grown in the presence of absence of mgtB. Cultures of MM1490 (mgtC+ mgtB+) () and MM1665 (mgtC+ mgtB) () were grown overnight in N minimal medium supplemented with 10 mM or 70 µM Mg2+. Cultures were processed as described in the legend to Fig. 1 and grown at various Mg2+ concentrations at 37°C. Aliquots were removed at the indicated times, and the OD600 was measured. Growth similar to that seen at 70 µM Mg2+ was seen at other Mg2+ concentrations below 500 µM.

View larger version (135K): [in this window] [in a new window]   FIG. 3.   Growth phenotype of the strain carrying only mgtC. LB plates were streaked with the following strains and incubated at 37°C for 24 h. Strains: 1, LT2 (wild type); 2, MM281 (corA mgtA mgtC mgtB); 3, MM1490 (mgtC+ mgtB+); 4, MM1665 (mgtC+ mgtB); 5, MM1733 (mgtC mgtB+); 6, MM1542 (mgtC+ mgtB[D379A]) The plate was scanned and contrast was adjusted by using Adobe Photoshop 4.0.

Absence of MgtC in strains containing MgtB. Western blot analysis was performed on mgtC⁺ mgtB⁺ strains versus mgtC⁺ mgtB strains to determine if MgtC protein was being produced. MM1665 (mgtC⁺ mgtB) expressed significant quantities of MgtC by 2 h after initiation of Mg²⁺ starvation, and MgtC was still detectable after 48 h (Fig. 4B). Surprisingly, no significant quantities of MgtC were detected in the MM1490 strain carrying the intact mgtCB operon (Fig. 4A) or in the wild-type LT2 strain (data not shown) after Mg²⁺ starvation. MgtB, which is encoded by the second gene of the operon, is expressed after Mg²⁺ starvation in MM1490 (Fig. 4C) and LT2 but not MM1665 (data not shown).

View larger version (52K): [in this window] [in a new window]   FIG. 4.   Effect of mgtB on the detection of MgtC protein in whole cells. Strains MM1490 (A and C) and MM1665 (B) were grown overnight in 10 mM Mg2+, washed three times in N minimal medium without added Mg2+, resuspended in the same medium in the absence of Mg2+, and incubated at 37°C. Cell aliquots were removed at the indicated time points and subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis on 12.5 or 7.5% polyacrylamide gels followed by transfer to nitrocellulose. Proteins were visualized with anti-MgtC (A and B) or anti-MgtB (C) antibodies as described in Materials and Methods. The internal control was a band that reacted with both the preimmune serum and the anti-MgtC antibody.

View larger version (42K): [in this window] [in a new window]   FIG. 5.   MgtC detection in an D379A MgtB mutant. Strain MM1542 was grown and samples were processed as described in the legend to Fig. 4.

Growth of an mgtC⁺ strain at low pH. When S. typhimurium bacteria invade epithelial or macrophage cells, they encounter an environment with a low pH in addition to one low in Mg²⁺ (7). To investigate whether mgtC might be involved in S. typhimurium survival at low pH, MM1665 (mgtC⁺ mgtB) was grown at pH 7.5 or 5.2 in N minimal medium supplemented with various concentrations of Mg²⁺. In the presence of 10 mM Mg²⁺, cells grown at pH 7.4 grew to an OD₆₀₀ similar to strain MM281 (Fig. 6A). At pH 5.2 with 10 mM Mg²⁺, the mgtC⁺ mgtB strain exhibited a markedly increased lag phase and grew to a lower OD₆₀₀. At lower Mg²⁺ concentrations, an mgtC⁺ mgtB strain could not grow to any significant extent at pH 5.2 (Fig. 6B).

View larger version (13K): [in this window] [in a new window]   FIG. 6.   Effects of pH 5.2 and low Mg2+ concentration on growth in the presence of mgtC. MM1665 (, ) and MM281 () were grown overnight in N minimal medium supplemented with 10 mM or 70 µM Mg2+. Cultures were washed with N minimal medium (pH 7.4 or pH 5.2) three times and resuspended in a volume to give the same initial OD600. The samples were diluted 1:200 in N minimal medium at pH 7.2 () or pH 5.2 () containing various Mg2+ concentrations and incubated at 37°C. Aliquots were removed at the various time points, and the OD600 was measured. Growth of MM281 at pH 7.4 was used as a negative control. Data similar to that shown for 70 µM Mg2+ were obtained at Mg2+ concentrations below 500 µM.

View larger version (50K): [in this window] [in a new window]   FIG. 7.   Effect of pH on the detection of MgtC in whole cells. Strains MM1665 (A), MM1490 (B and C), and MM1737 (D) were grown in N minimal medium in the absence of Mg2+ at pH 5.2 (A to C) or pH 7.2 (D). Cell aliquots were removed at the indicated time points for electrophoresis and visualization of MgtC and MgtB proteins was performed as described in the legend to Fig. 4.

Transport of Ni²⁺ and Co²⁺. MgtC allows growth of S. typhimurium in the absence of added Mg²⁺. This suggests that it may be involved in Mg²⁺ uptake. Blanc-Potard and Groisman (3) postulated MgtC to be a Mg²⁺ transporter based on the ability of a very high (25 mM) concentration of Mg²⁺ to increase intramacrophage survival of a strain of S. typhimurium lacking mgtC. We have previously demonstrated that strain MM281, which lacks functional alleles of mgtC and the three characterized Mg²⁺ transporters, corA, mgtA, and mgtB, requires a high extracellular Mg²⁺ concentration for growth and cannot take up ²⁸Mg²⁺ (25). Unfortunately, this isotope is no longer available; however, corA, mgtA, and mgtB all mediate uptake of ⁶³Ni²⁺ as an alternative substrate (25). In addition, corA (11, 25) and the mgtE Mg²⁺ transporter found in some bacterial species (26, 30) mediate ⁵⁷Co²⁺ uptake. We therefore measured uptake of both of these radioisotopes in a strain carrying only a functional allele of mgtC. To induce mgtC expression, cells were starved for Mg²⁺ for 6.5, 23, or 47 h. The cells contained significant increases in MgtC content indicated by Western blot analysis at each of these time points (data not shown) similar to that shown in Fig. 4. Uptake was measured at multiple time points, since although MgtC protein is expressed by 2 h after Mg²⁺ starvation, the cultures exhibit a 24-h lag phase before achieving significant growth. Nonetheless, despite the presence of substantial amounts of MgtC protein, no ⁶³Ni²⁺ uptake was detectable at any time point (data not shown). With ⁵⁷Co²⁺, an apparent uptake of about 1% that of the wild-type strain could be measured, but this uptake could not be completely inhibited by 100 mM Mg²⁺ (data not shown), suggesting that it was due to nonspecific Co²⁺ binding to the cells.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

The mgtCB operon forms part of an S. typhimurium pathogenicity island, SPI-3. The mgtC gene is essential for virulence in a mouse lethality assay and for long-term survival within a macrophage cell line (3) but is not required for efficient invasion or short-term survival within epithelial or macrophage cultured cell lines (24, 28). Neither MgtB nor MgtA assist in MgtC's role in survival within macrophages. The function of the MgtC protein is unknown, however. Blanc-Potard and Groisman (3) have suggested that MgtC is a fourth Mg²⁺ transporter of S. typhimurium required for acquisition of Mg²⁺ within the macrophage. This hypothesis is based largely on the ability of an S. typhimurium mgtC mutant strain to survive within a macrophage cell line if 25 mM Mg²⁺ is added to the culture medium. Presumably, the high extracellular Mg²⁺ concentration allows the macrophage cell line to increase intracellular Mg²⁺ and thus supply additional Mg²⁺ for intravacuolar growth of S. typhimurium. This scenario seems highly unlikely based on the known properties of Mg²⁺ homeostasis in mammalian cells. While Mg²⁺ exchanges readily in some tissues and cell types, including liver, heart, and adipocytes (4, 5, 9, 17, 19), Mg²⁺ uptake and exchange in most cells, including lymphocytes and muscle cells, is severely restricted. In lymphocytes, Mg²⁺ is compartmented and cannot be fully exchanged with the extracellular medium even after 48 h of incubation (9). Moreover, exposure of intravacuolar S. typhimurium to an increased free Mg²⁺ concentration would presumably require that cytosolic free Mg²⁺ increase markedly upon addition of extracellular Mg²⁺. Cytosolic free Mg²⁺ is tightly controlled, however, and simple addition of even high concentrations of extracellular Mg²⁺ does not increase cytosolic free Mg²⁺ in any cell type examined. Thus, it seems highly unlikely that simple addition of extracellular Mg²⁺ would be able to supply significant additional Mg²⁺ to a bacterium within a eukaryotic cell vacuole.

This interpretation is consistent with the inability of a strain expressing only MgtC to accumulate either ⁶³Ni²⁺ or ⁵⁷Co²⁺. Since one or both of these cations is transported by all known Mg²⁺ transport proteins (25, 26), it is reasonable to infer that MgtC is not a Mg²⁺ transport protein. Nonetheless, since ²⁸Mg²⁺ is no longer available and thus Mg²⁺ uptake cannot be tested directly, a definitive answer to this question must await purification and reconstitution of any transport activity that MgtC might be capable of.

The conclusion that MgtC is not a Mg²⁺ transporter is further consistent with Mg²⁺ effects on the growth of an mgtC⁺ strain carrying mutations in the known Mg²⁺ transporters (corA mgtA mgtB). Substantial amounts of MgtC protein are synthesized in such a strain very early after initiation of Mg²⁺ starvation. Nonetheless, no growth occurs for at least 24 h in N minimal medium (Fig. 2B) or rich medium (Fig. 3). If MgtC were a Mg²⁺ transporter, growth should begin relatively early after Mg²⁺ starvation. Nonetheless, despite data indicating that a strain carrying only mgtC is impaired in growth, given sufficient time, such strains are able to grow at very low Mg²⁺ concentrations. Although this suggests that the cells are able, eventually, to obtain sufficient Mg²⁺ for growth, Mg²⁺ content by atomic absorption analysis is actually slightly decreased in a strain producing only MgtC, again inconsistent with a Mg²⁺ transport function for this protein. Rather, the ability to grow presumably reflects some unknown adaptation to Mg²⁺ deprivation.

Although the cells do not begin to grow until approximately 24 h after Mg²⁺ starvation, MgtC is synthesized very early after initiation of Mg²⁺ starvation, consistent with regulation studies of the mgtCB locus (24, 27-29). Moreover, the amount of MgtC protein did not correlate with growth; it generally remained relatively constant for 48 h after Mg²⁺ starvation was initiated. In sharp contrast, MgtC is virtually undetectable in the presence of a functional MgtB Mg²⁺ transport protein. MgtB could directly interact with MgtC, making it unstable. This would imply some type of protein-protein interaction, perhaps even that MgtC could function as a subunit of MgtB. However, a direct interaction seems unlikely. The mgtB and mgtC genes are not always associated and can occur together or separately within enterobacteria (3). Moreover, the lack of a functional mgtC allele does not affect MgtB membrane insertion or transport properties (29). Nonetheless, the converse conclusion cannot be ruled out. MgtB protein could affect the function of MgtC.

A second possibility is that stability of the mgtC mRNA transcript could be affected by the presence of an intact mgtB open reading frame. There is evidence for differential regulation of protein encoding segments of a polycistronic message. For example, the arsenical resistance operon (arsRABC) of resistance plasmid R773 is transcribed as a single polycistronic mRNA (16). The ArsA and ArsC proteins are produced in amounts much greater than that of the ArsB protein, which is encoded by an intervening segment of the mRNA. Northern blot analysis demonstrated that the initial polycistronic mRNA is processed, yielding messages for ArsA and ArsC. The intervening message encoding ArsB appears to be selectively degraded. This internal instability of the polycistronic mRNA could also be applied to the mgtCB locus. The mgtC-encoded portion of the transcript might be selectively degraded, while the mgtB mRNA species would remain stable and be translated. In the presence of an insertionally inactivated mgtB gene or absence of an mRNA encoding MgtB, the mgtC mRNA might be more stable.

The issue of MgtC expression is made more complicated by the inability of strains dependent on it to grow at acid pH. When S. typhimurium bacteria invade epithelial or macrophage cells, they appear to encounter an environment low in nutrients and pH (1, 7) although pH has not been directly measured. The failure of an mgtC⁺ strain to grow at pH 5.2 (Fig. 6) even though MgtC is produced (Fig. 7) suggests that induction of mgtC transcription and expression of mgtC are not required for survival at low pH within the eukaryotic cell. This is consistent with our observations that strains lacking a functional mgtC have no defect in invasion efficiency or short-term survival in either epithelial-like or macrophage-like cultured cell lines (24). Alternatively, the lack of growth of the "free-living" bacterium at low pH may simply be an indication that such conditions do not fully mimic intravacuolar conditions. Nonetheless, there is evidence that the intravacuolar pH increases with time after bacterial entry into the macrophage (1). Therefore, perhaps the inability of the mgtC⁺ strain to grow at low pH is not a defect at all but simply reflects that the requirement for a functional mgtC occurs late in the invasion process.

MgtC is crucial for S. typhimurium virulence. Its regulation by Mg²⁺ via the two-component PhoPQ regulatory system and its chromosomal association with mgtB, encoding a Mg²⁺ transporter, might suggest that its function involves Mg²⁺. Overall, our data do not support this association. It does not appear to function as a Mg²⁺ transporter. At a pH apparently encountered by the bacterium during the pathogenic process, it cannot support growth. Additional studies will need to address the relationship of MgtC to Mg²⁺ and the actual function of MgtC, both in the free-living bacterium and during bacterial encounters with eukaryotic cells.