Magnesium Transport in Salmonella typhimurium

Transcription

1 THE JOURNAL OF BIOLOGICAL CHEMISTRY by The American Society for Biochemistry and Molecular Biology, Inc. Vol. 266, No. 2, Issue of January 15, pp ,1991 Printed in U. S. A. Magnesium Transport in Salmonella typhimurium REGULATION OF mgta AND mgtb EXPRESSION* (Received for publication, August 16, 1990) Marshall D. SnavelyS, Stephen A. Gravina, ToBun T. Cheung, Charles G. Millerj, and Michael E. Maguiren From the Department of Pharmacology and the Department of Molecular Biology and Microbiology, School of Medicine, Case Western Reserve University, Cleveland, Ohio Salmonella typhimurium contains three distinct transport systems (CorA, MgtA, and MgtB) that move Mg2+ across the cytoplasmic membrane. Mutant strains containing only one of these three systems have been constructed and used to study each system isolation. in Characterization of these systems has been hampered, however, by the need to use "Mg2+, a relatively unavailable, extremely expensive, and short lived radio- isotope. This paper reports that e3ni2+ is transported into the cell by all three of the S. typhimurium Mg2+ transport systems. In a strain deficient in all three systems, uptake of e3ni2+ was undetectable under the conditions used. Comparison of 63Ni2+ uptake kinetics and inhibition of 63Ni2+ transport by other divalent cations suggest that Ni2+ can be used as an analog of Mg2+ in the study of these three transport systems. Using s3ni2+ to measure uptake, the effect of Mg2+ levels in the growth medium on transport by each system was tested. Transport by the CorA system was unaffected by changes in the amount of Mg2+ in the growth medium. In contrast, uptake via MgtA and MgtB was significantly increased in cells grown in 10 p~ extracellular Mg2+ compared to cells grown in 10 mm Mg2+. The increases in uptake were the result of increases in VmaX without change in K,. This result suggests that, in low Mg2+ medium, cells contained vious studies (1, 4) have identified Co2+ and Ni2+ as competihigher levels of the transporters. Production of B-ga- tive inhibitors of M e influx through all three S. typhimulactosidase from mgta::lacz and mgtb::lacz but not rium transport systems. However, Co2+ is transported only by cora::lacz fusions was also increased when cells were the CorA system (1). This report documents the use of 63Ni2' grown in low extracellular concentrations of Mg2+ in- as an analog of "M2+ in studies of M e accumulation by S. dicating that the regulation occurs at the level of tran- typhimurium and demonstrates that extracellular divalent scription. Expression of &galactosidase was also inhib- cation concentration regulates transcription of mgta and ited by the addition of other divalent cations including mgtb but not cora. Ca2+ and Mn2+. Regulation of transcription from the mgta and mgtb promoters was similar over the range MATERIALS AND METHODS of extracellular Mg2+ concentrations from 10 pm to 10 mm. At 1 p ~ however,, transcription from the mgtb Media and Reagents-Growth of cells was performed in N medium promoter, as measured by &galactosidase levels in a (1) containing 0.1% casamino acids and 38 mm glycerol or 22 mm glucose as carbon source. 63Ni2+ (1.9 Ci/mmol) was purchased from mgtb::lacz transcriptional fusion strain, was in- Amersham. o-nitrophenylthiogalactoside for p-galactosidase assay creased over 800-fold, and Ca2+ could no longer inhibit was purchased from Sigma. transcription effectively. In contrast, growth at 1 pm Strains and Plasmids-Strains "252 (AleuBCD485 extracellular Mg2+ increased transcription from the rngtc1o::mudj cora45:mudj), "278 (AleuBCD485 mgta2l::mudj mgta promoter only about 30-fold and Ca2+ could still cora45::mudj), and "328 (AleuBCD485 mgta27:mudj mgtinhibit this increase. These results suggest that at least B1630:TnlOAcam) each harbor chromosomal insertion mutations in two of the three Mg2+ transport loci (2, 4). In these strains, uptake of *This research was supported by National Institutes of Health Mg2+ is therefore solely dependent on the single remaining transport Grant GM The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. $ Recipient of a postdoctoral fellowship from the Northeast Ohio Affiliate, American Heart Association during this work. Current address: AMGEN, Amgen Center, Thousand Oaks, CA n To whom correspondence and reprint requests should dressed. be ad- 824 two distinct mechanisms are responsible for regulation of the mgta and mgtb transcription in response to extracellular cation concentration. Salmonella typhimurium has three transport systems, CorA, MgtA, and MgtB, that move M e across the membrane (1, 2). These three systems can be distinguished genetically and on the basis of distinct transport properties (1-4). The CorA transport system is constitutively expressed, mediates both influx and efflux of M$+, and appears to be the primary M$+ transport system (1, 4). In contrast, the MgtA and MgtB systems mediate M$+ influx only, and transport decreases with increasing concentrations of M e in the growth medium (4). The accompanying report (5) demonstrates that MgtB is a P-type ATPase (6, 7) with greater homology to the Ca2+ ATPases of yeast and mammalian sarcoplasmic reticulum than to P-type ATPases of other prokaryotes. Characterization of the CorA, MgtA, and MgtB systems is hampered by high cost (>$2O,OOO/mCi), short half-life (21 h), and lack of routine availability of "M$+. It has therefore been of interest to find an analog of "Mg2+ to facilitate the study of these interesting and important transporters. Pre- system. "328 harbors the intact CorA system, "252 harbors MgtA, and "278 harbors MgtB. "281 (AleuBCD485 cora45::mudj mgta2l::mudj mgtcl0:mudj zjh1628::tnloacam) contains insertions in all three Mg2+ transport loci (2, 4). This strain requires 100 mm Mg2+ in the medium for optimal growth and does not accumulate "M$+ (4). "199 (AleuBCD485 cora45::mudj), "638 (AleuBCD485 mgtael::mudj), and "599 (AleuBCD485 mgtbl6:mudj) carry chromosomal MudJ insertions (8) in cora, mgta, and mgtb, respectively. The MudJ insertions in these strains

2 0 "252 lmgta*l "278 (MgtB') Transport Magnesium in S. typhimurium 825 I e Minutes FIG. 1. Time courses of s3ni2' uptake by the MgtA and MgtB To confirm that 'j3ni2+ was being accumulated by the CorA, Mgz+ transport systems. 63Niz+ uptake was measured at 37 "C in MgtA, and MgtB systems, competition for 63Ni2+ uptake by strains "252 (mgta+, 0) and "258 (mgtb+, U) and "281 (A) other divalent cations was examined. Mg2+ was a potent and as described under "Materials and Methods" in the absence of M e and at a Ni2+ concentration of 100 PM. M e concentration during competitive inhibitor of 63Ni2+ uptake by all three Mg2+ transgrowth was 30 pm. port systems. Ki values for Mg2+ inhibition of 63Ni2+ uptake (Table 11) were comparable to K, values for Mg2+ derived from "Mg2+ transport experiments (1, 4). Other divalent are oriented so that lacz is transcribed from the M e transport gene's promoters. Previously, the mutation mgtci0:mudj was designated cations showed distinct rank orders of potency for each uptake mgtbi0:mudj. From the sequence data in the accompanying report, system in competing for 63Ni2+ uptake (Table 11). The rank this insertion is not within the mgtb open reading frame but is orders were Mg2+ > Co'+ > Mn'+ > Ca2+ for the CorA system, upstream of mgtb in a new gene now designated mgtc, hence the Co2+ = M? > Mn'+ > Ca2+ for the MgtA system, and Mg2+ correction in the name. = Co2+ > Mn2+ >> Ca2+ for the MgtB system. The rank orders Manipulation of DNA-DNA manipulations were performed as observed for each transport system were similar for inhibition described in Sambrook et al. (9). Assays-Accumulation of 63Ni'+ was assayed as described (1,4) for of 63Ni2+ uptake and "M$+ uptake. Some minor differences uptake of "Me except that 0.1% casamino acids were added to the in rank order were observed among cations with similar N medium (1). Casamino acids have been found to improve the affinities, but the overall orders were comparable for both growth of all strains at low M e concentrations.' Transport assays 'j3ni2+ and "M$+ uptake. The only discrepancy was Mn2+ were typically conducted using cells grown in glucose as a carbon inhibition of uptake via the MgtA system, where Mn2+ was a source in order to be comparable to previous studies with "Me. competitive inhibitor of 63Ni2+ uptake. In contrast, Mn2+ Glycerol was used as a carbon source assays because it was observed that expression of the MgtA and MgtB inhibited only 35%of "Mg2+ uptake by MgtA (4). Overall, systems was enhanced under these conditions. In control experiments, these results indicate that 63Ni2+ is transported by each of the the effects of changing the extracellular M$+ concentration on three Mg'+ transport systems in S. typhimurium. expression were the same regardless of whether the Regulation of Cation Uptake by the CorA, MgtA, and MgtB cells were grown in medium containing glycerol or glucose. Systems-Previous studies demonstrated that "Mg2+ uptake All transport assays were performed at 37 "C. Time of incubation in S. typhimurium is influenced by the concentration of M$+ in kinetic experiments varied with the system being assayed, 10 min for the CorA system and 30 min for the MgtA and MgtB systems. in the growth medium (1, 4). To characterize this effect, Uptake was linear for the period indicated in all cases. The amount of fi3ni2+ retained on the filters was quantitated in a Beckman LS7000 TABLE I scintillation counter with a counting efficiency of 60%. Kinetics of fi3ni2+ uptake by Mg2+ transport systems (&Galactosidase was assayed as described previously (3). Inclusion of S. typhimurium of up to 10 mm M e in the assay medium had no effect on the Strains dependent on the CorA (MM328), MgtA (MM252), or observed levels activity. MgtB ("278) M e transport systems were grown in N medium containing the indicated concentrations of Mg2+ to an OD,, nm of RESULTS about 0.4. Cells were then washed twice in ice-cold N medium without added M e and assayed for 63Ni2' uptake after resuspension to a 63Ni2+ as an Analog of "M$+"Strains harboring a single final ODW nm of 0.2. Uptake of 63Ni2+ was determined as described intact M e uptake system were examined for uptake of 63Ni2+. under "Materials and Methods." K,,, (see text) and VmaX values were Time course experiments demonstrated that 63Ni2+ was accumulated by all three Mg2+ transport systems in S. typhimurium (Fig. 1 and data not shown). Uptake of 63Ni2+ by the CorA system was linear for 10 min (data not shown) while uptake by both the MgtA and MgtB systems was linear for 60 min (Fig. 1). Accumulation of 63Ni2+ by strain "281 which harbors insertions in all three M e transport systems was very low and at the limit of detection by our assay. The maximal rate of uptake observed in "281 was less than 0.1 pmol min" IO8 cells" (Fig. I), was unaffected by the addition of the protonophore m-chlorocarbonylcyanide phenylhydra- M. D. Snavely and M. E. Maguire, unpublished observations. zone (data not shown), and was not further characterized. Since 63Ni2+ uptake in a strain lacking all three M F transport systems was essentially zero, uptake in "328, "252, and "278 was therefore due to transport of Ni2+ by the single Mg2+ transport system present in these strains and not by another cation transporting system. Kinetic experiments yielded distinct K, and VmaX values for 63Ni2+ uptake by each of the three M$+ transport systems. The apparent K, values for 63Ni2+ uptake in cells grown in medium containing 500 pm M$+ were 237 f 72 pm (n = 5) for the CorA system, 5.1 f 1.0 pm (n = 7) for the MgtA system, and 1.7 f 0.8 p~ (n = 15) for the MgtB system. These values agreed well with Ki values derived from Ni2+ inhibition of uptake by each transport system (4). When determined using cells grown in medium containing 5 or 500 pm M$+, the V,,, values for the three systems gave the same rank order of transport capacity as was observed for uptake of "Me (4), that is, CorA >> MgtA > MgtB (Table I). calculated from the concentration dependence of 'j3ni2+ uptake using 6 to 8 Ni2+ concentrations. The values are averages of from two to seven experiments for each growth condition except for the CorA system at mm MgZ+ which represents a single experiment. ND = not determined. [Mg2+] in Vmax of 6'3Ni2+ uptake growth medium CorA MgtA MgtB mm pmol min" I@ cell." ND ND

3 826 Transport Magnesium TABLE I1 Divalent cation inhibition of 63Ni + uptake through the Mg2+ transport systems of S. typhimurium Uptake was assayed as described under Material and Methods and in the legend to Table I using strains dependent on the CorA (MM328), MgtA (MM252), or MgtB ( 278) Mg2+ transport systems. Cells were grown in medium containing 5 p~ M2+. The values presented for the MgtA and MgtB Mg2+ transport systems are the averages of two experiments while those for CorA are for at least three experiments in which increasing concentrations of the inhibitory cation were added at a fixed Ni2+ concentration of 100 p~ (CorA) or 4 pm (MgtA and MgtB). The Ki values are calculated from the cation concentration giving 50% inhibition (ICbo) and the K,,, values for Ni2+ uptake (see text) according to the formula K, = ICbo*(K,/ (K, + [Ni +])J. MgtB Cation MgtA CorA Kc fim Mg Ca2+ Mn + 20, , , CO TABLE 111 Effect of Mg2+ concentration in the growth medium on 6: Ni2+ uptake by Mg2+ transport systems Uptake was assayed as described under Material and Methods and in the legend to Table I using strains dependent on the CorA (MM328), MgtA (MM252), or MgtB ( 278) Mg + transport systems. The M e concentration in the growth medium was 10 p~ or 10 mm as indicated. The final concentration of Ni2+ in the assay medium was 50 pm. MgtB [Me1 MgtA CorA M e transport system pmol min 108 cells 10 pm mm uptake of fi Ni2+ was examined in strains dependent on a single functional Mgz+ transport system after growth in N medium containing 10 p~ or 10 mm Mg2+ (Table 111). The initial rate of uptake through the CorA system did not change appreciably with the change in M e concentration in the growth medium; however, the initial uptake rate increased severalfold for both the MgtA and MgtB transport systems when cells were grown in medium containing 10 pm M$+ compared to cells grown in medium containing 10 mm M$+. This suggested that an increased Mg2+ concentration in the growth medium could repress expression of MgtA and MgtB but not the CorA M F transport systems. This hypothesis was investigated by determination of the effect of extracellular divalent cations on the kinetics of uptake by each system and on the transcription of each locus. The apparent K,,, values for 63Ni2+ uptake by the CorA and MgtB transport systems did not change as a function of extracellular Mg2+ concentrations between 5 p~ and 10 mm. That for the MgtA system appeared to increase slightly in cells grown in medium containing less than 50 p~ Mg, from 5.1 k 1.0 pm (n = 7) to 10.7 k 1.4 pm (n = 5). Similarly, the V,,,, of the CorA system was affected minimally by the extracellular Mg2+ concentration, changing only about 2-fold over a 2000-fold range of extracellular M$+ concentration (Table I). In contrast, both the MgtA and MgtB systems showed marked repression of cation uptake as the Mg2+ concentration in which the cells were grown increased (Table I). The V,,, for 63Ni2+ uptake by MgtA decreased over 20-fold as the concentration of Mg + in the growth medium was raised from 0.5 pm to 10 mm. The VmaX of the MgtB system decreased in S. typhimurium 50-fold. The Mg2+ concentration in the growth medium affected V,,, with a similar dose-response curve for either MgtA or MgtB. Experiments using either MgS04 or MgC12 gave identical results (data not shown). The repression of transport capacity without a concomitant change in K,,, for both MgtA and MgtB suggests that regulation of these systems occurs by changing the number of transporters rather than by altering the properties of existing transporters. Cation Regulation of Transcription-Operon fusions of cora,mgta, and mgtb to lac2 were used to determine the effect of Mg2+ concentration in the growth medium on expression of these genes. Strains 199, 638, and 599, which harbor cora::mudj, mgta::mudj, and mgtb::mudj fusions, respectively, were grown in N medium containing different concentrations of MF. As expected from transport data, transcription from the cora promoter was unaffected by the extracellular M$+ concentration (Fig. 2, inset), while transcription from both the mgta and mgtb promoters increased in parallel about 10-fold as extracellular Mgz+ concentration decreased from 10 mm to 10 ~ L M (Fig. 2, inset). However, regulation of 0-galactosidase production by M F differed dramatically for the two fusions at extracellular concentrations of Mg2+ between 10 and 1 p ~ Expression. from the mgta promoter increased an additional 3- to 4-fold between 10 and 1 p~ extracellular Mg2+ while expression from the mgtb promoter increased over 800-fold (Fig. 2). These results indicate that the increase in M e transport seen in cells grown in low levels of Mg2+ was a result of increased transcription of the genes that code for the transport systems. Cation Specificity of Regulation-The cation specificity of the regulatory system(s) responsible for Mg2+ repression was determined by testing the ability of other divalent cations to inhibit P-galactosidase production from the mgta and mgtb promoters. In these experiments, cells were grown in N medium containing 20 p~ Mg2. The concentrations of added divalent cations were varied over the range from 30 FM to 10 mm, and P-galactosidase levels were determined. Ca2+ (Fig. 3) and Mn2+ (data not shown) repressed transcription from both the mgta and mgtb promoters. Ca2+ and Mn2+ demonstrated similar potencies in expression - Y) c E z 0 In 100 c Log IMg *l (mm1 FIG. 2. Effect of Mg2+ concentration in the growth medium on &galactosidase production from lac2 fusions to the cora, mgta, and mgtb promoters. Cells were grown in N minimal medium with glycerol as a carbon source as described under Materials and Methods. At an ODsoo., of about 0.2, aliquots of cells were removed and immediately frozen for assay of 0-galactosidase as described (3). The inset shows the same data as the larger figure but with the range above 10 p~ M$+ blown up. For all three systems, P- galactosidase activity at 10 mm Mg2+ was between 0.5 and 2.2 units. The increase in transcription was approximately 12-fold at 10 pm Mg2+ for both mgta and mgtb, versus 35-fold for mgta and 900-fold for mgtb at 1 p ~ The. data shown are from a single experiment representative of three similar experiments.

4 Magnesium Transport in S. typhimurium N ICa2+l (mm1 FIG. 3. Ca2+ inhibition of B-galactosidase expression from 1acZ:mgtA and 1acZ:mgtB fusions. Cells were grown as described under "Materials and Methods" and in the legend to Fig. 2 at 20 pm extracellular M e and in the absence (100%) or presence of the indicated Ca'+ activity was measured as described in Fig. 2. For both fusions, 100% activity was approximately 25 units of &galactosidase. A second experiment gave identical results. G o [Mg2+l (I." FIG. 4. Ca2+ inhibition of B-galactosidase expression from 1acZ:mgtA and 1acZ:mgtB.promoters at low- extracellular Mg2+ concentrations. Cells were grown in the presence or absence of 1 mm extracellular Ca'+ at the indicated extracellular Mg2+ concentration as described under "Materials and Methods" and in Fig. 2. A second experiment gave identical results. Values for galactosidase activity were 30 units for the mgta promoter and 1400 units for the mgtb promoter. from both the mgta and mgtb promoters over a range of extracellular M$+ concentrations from 20 PM to 5 mm. At 20 pm extracellular M$+, 50% inhibition of transcription of either mgta or mgtb was observed at a Ca2+ concentration of 200 pm (Fig. 3). Interestingly, when cells were grown in M$+ concentrations less than 5 PM, 1 mm ca2+ was unable to repress the marked increase in expression from the mgtb promoter (Fig. 4), whereas the degree of repression of mgta expression by Ca'+ was unaffected by the extracellular Mg2+ concentration. Ni2+ and Co'+ also appeared to production from both promoters at concentrations of 100 FM (data not shown). However, Ni" and Co'+ are both highly toxic to S. typhimurium grown in minimal medium, making definitive demonstration of transcriptional repression difficult. Na+ did not inhibit p-galactosidase expression (data not shown). DISCUSSION Use of "Ni*+-The data presented support the use of 63Ni2+ as an analog of '"Mg2+ for transport studies in S. typhimurium. Kinetic results are consistent with the conclusion that 63Ni2+ is taken up via each of the three M$+ transport systems in S. typhimurium. First, for each system, the K, values for uptake of 63Ni2+ agreed with the Ki for Ni2+ inhibition of 'amg2+ uptake. Second, Ni2+ and Mg2+ are competitive inhibitors of each other's uptake regardless of which Mg2+ transport system is present. Third, the ability of other ions to inhibit 'j3ni2+ uptake was consistent with their ability to inhibit '"Mg2+ uptake. Fourth, the rank order of V,,, values for 83NiZ+ uptake determined for the CorA, MgtA, and MgtB systems is identical with that determined from uptake studies using "Mg'+. Fifth, Mg2+ regulation of transport via the CorA, MgtA, and MgtB transport systems was identical whether measured with fiani2+ (this report) or "Mg2+ (4). Finally, a strain lacking all three Mg" transport systems does not transport Ni" under the conditions studied. There are some minor discrepancies in results using 6"Ni2+ and "M$+. The rank order of potencies for cation inhibition, while very similar, are not identical for Ni2+ uersus Mg2+. However, the only differences occur in ordering cations whose affinities are very close. The only major difference between 'j3ni2+ and "Mg2+ uptake in S. typhimurium is in Mn'+ inhibition. Previous experiments with '"Mg'' indicated that Mn'+ could only in- hibit 35% of Mg2+ uptake via the MgtA system (4); however, Mn2+ completely inhibited 63Ni2+ uptake via MgtA. We have no explanation for this difference. Regardless, the weight of the evidence clearly indicates that 63Ni2+ is transported by the three M$+ transport systems of S. typhimurium. This result is important because *'Mg2+ is not routinely available, has a 21-h half-life and is extremely costly. Previous work on nickel transport in prokaryotes is in accord with the conclusion that Ni" can be accumulated via M$+ transport systems. Chemolithotrophic bacteria (10-13) express specific, high affinity nickel transport systems while chemoorganotrophic bacteria (14-17) appear to transport nickel through M$+ transport systems with a relatively lower affinity. For example, specific nickel transport systems in methanogens or acetogens have high affinity for Ni'' and are not selectively inhibited by Mg2+ (10, 11). Regulation of mgta and mgtb Expression-Use of 63Ni2+ has now allowed characterization of the regulation of the MgtA and MgtB transport systems. The results presented here extend previous data (2) suggesting that the concentration of M$+ in the growth medium has little effect on cation transport by the CorA system. In contrast, transport via MgtA and MgtB systems is repressed by the presence of divalent cation in the growth medium. While regulation of Mg2+ transport by M$+ might be expected, repression of mgta and mgtb transcription by Ca'+ is somewhat surprising, and its physiological significance is unknown. Ca'+ does not inhibit Mg'+ uptake via the MgtB system, having a Ki of at least 30 mm, and is thus unlikely to be transported by MgtB (4). Ca'+ does inhibit uptake via MgtA with a K, of 300 p ~ but, even this value is an order of magnitude greater than the K, of MgtA for M F (4). This regulation might be explained if these M$+ transport systems were Mg'+-Ca2+ antiporters. This seems unlikely for MgtB, however, since the accompanying report (5) demonstrates that MgtB is a P-type ATPase (6, 7). Preliminary results indicate that MgtA is also a P-type ATPase.' Unlike eukaryotic cells, Gram-negative bacteria like S. typhimurium have no Ca'+-stimulated ATPase activity, their Ca2+ efflux being modulated by a Na+-Ca'+ exchanger (18, 19). In most respects, the regulation of mgta and mgtb by Ca2+ and Mg2+ is identical, suggesting that the two cations use the M. D. Snavely, S. Gura, C. G. Miller, and M. E. Maguire, unpublished observations.

5 828 Magnesium Transport in S. typhimurium that changes in salt concentration are not involved and supports the conclusion that the effect is selective for divalent cations. Physiological Relevance-The significance of transcriptional regulation in response to extracellular cation levels is unclear. Even the importance to S. typhimurium of having three distinct M$+ transport systems has yet be to elucidated although the differential regulation of three distinct cation uptake systems implies physiological relevance. To approach this question, the M$+ uptake capacity was calculated as a MgtB, CorA function of extracellular M$+ concentration using the kinetic 0 t l l l l d 1111 lllj ~ ~ ~ Q 1 1 parameters of Table I. The values for rate of M e uptake by each individual system at a given extracellular Mg2+ concentration were calculated and added together to determine the Mg * in Growth Medium (mm) total capacity for Mg2+ uptake of a wild type cell grown in FIG. 5. Mg2+ transport capacity of S. typhirnurium as a media containing that M$+ concentration (Fig. 5). At high function of Mg2+ Concentration in the growth medium. The M e concentrations, cellular M$+ uptake is dependent total capacity of S. typhimurium Mg2 transport was calculated from : Ni + uptake data as described under Discussion. almost exclusively on the CorA transport system and decreases as the concentration of extracellular M$+ declines from 10 mm to about 30 PM Mg2+. From about 30 PM to about same mechanism for regulation. Likewise, limited data with 5 pm extracellular M$+, the cellular uptake remains roughly Mn2+, Ni, and Co2+ suggest that these cations influence constant due to a relatively balanced decrease in transport mgta and mgtb. Ca2+ appears somewhat less potent than via CorA and an increase in transport by both MgtA and M$+, exhibiting an ICso of about 200 PM for repression of MgtB. However, below 10 PM extracellular MF, the MgtB transcription. Calculation of half-maximal inhibition for Mg2+ system becomes increasingly important and eventually beis somewhat more difficult because of the marked increase in comes the dominant system, even though total Mg2+ uptake expression from the mgtb promoter below 10 PM Mg2+. Con- capacity continues to fall as the extracellular Mg2f concentrasidering only Mg2+ concentrations between 10 FM and 10 mm, tion drops further below the apparent K,,, for uptake. At 50% inhibition by Mg2+ occurs at about 100 PM for both the approximately 5 PM extracellular M$+, the doubling time of mgta and mgtb promoters. The obvious and major difference wild type S. typhimurium begins to increase. It thus appears between the regulation of mgta and mgtb is the remarkable important to the cell to maintain a minimal rate of Mg2+ increase in transcription from the mgtb promoter at Mg2+ uptake. concentrations below 10 PM (Fig. 3). M$+ concentrations This analysis is complicated by the fact that the CorA below 10 pm slow bacterial cell growth; at 1 PM M$+, the transport system mediates both M$+ influx and efflux and doubling time is markedly increased and growth ceases at appears to be the sole Mg2+ efflux system of S. typhimurium. ODsoo, of 0.15 presumably due to depletion of Mg2+ from Moreover, efflux appears to be regulated in such a manner the medium. Thus, the increased expression from the mgtb that no efflux of M e occurs at relatively low extracellular promoter suggests that an additional factor may be involved M$+ concentrations, e.g. 100 PM (4). At higher extracellular in Mg2+ regulation of transcription at Mg2+ concentrations M$+, concentrations, CorA mediates both influx and efflux. low enough to inhibit growth. Interestingly, the data of Fig. 4 Thus, the large CorA-dependent increase in total uptake illustrate that Ca2+ cannot repress the increase in transcrip- capacity suggested in Fig. 5 at M$+ concentrations greater tion from the mgtb promoter at lowm$+ concentrations, than 100 FM (Fig. 5) may not occur due to increased M%+ whereas repression of transcription from the mgta promoter efflux at higher extracellular Mg2+ concentrations. The net remains apparent. Thus, the MgtB syste may interact with rate of cellular Mg2+ accumulation may remain relatively two distinct regulatory elements, only one of which regulates constant over a wide range of extracellular M$+ concentra- MgtA, and neither of which influences CorA. Because of the tions. This implies an even greater dependency of M e transslowed growth at very low Mg2+, however, regulation of tran- port capacity on the MgtB system at very low extracellular scription from the mgtb promoter at such low M$+ concen- Mg2+ concentrations. This hypothesis is difficult to test betrations may not be a Mg2+-specific effect. Under severe Mg2+ cause detailed studies on cation efflux via the CorA system limitation, other physiological changes may be occurring, one have been hampered by the necessity of using s3ni+, or more of which may influence transcription from the mgtb while an excellent analog of Mg2+ for uptake studies, is of promoter. In this case, Mg2+ would only be indirectly respon- little use in the study of efflux. Ni, like Co2+, remains in the sible for the increase in transcription. cell once it has entered, presumably due to tight binding to The fact that several divalent cations are effective in re- intracellular sites. Whether and why the cell needs to mainpressing both the MgtA and the MgtB transport system tain a constant rate of Mg2+ influx will thus be the subject of suggests that a common regulatory system is involved. In future studies. particular, Ca2+ represses transcription of these Mg2+ transport genes, yet it does not potently inhibit and is not trans- REFERENCES ported by either system. It may be that the regulatory ele- 1. Hmiel, S. P., Snavely, M. D., Miller, C. G., and Maguire, M. E. ments that control transcription from the mgta and mgtb (1986) J. Bacteriol. 168, promoters also affect expression of other genes in response to 2. Hmiel, S. P., Snavely, M. D., Florer, J. B., Maguire, M. E., and divalent ion limitation. It is interesting to note that transcrip- Miller, C. G. (1989) J. Bacteriol. 171, tion from a number of genes coding for bacterial toxins is 3. Snavely, M. D., Florer, J. B., Miller, C. G., and Maguire, M. E. (1989) J. Bacteriol. 171, repressed by growth in high concentrations of Mg2+ or Ca2+ 4. Snavely, M. D., Florer, J. B., Miller, C. G., and Maguire, M. E. (20-23). Our assays were performed in N medium which (1989) J. Bacteriol. 171, contains 100 mm Tris. This, coupled with the lack of effect 5. Snavely, M. D., Miller, C. G., and Maguire, M. E. (1991) J. Biol. of Na+ concentrations up to 10 mm on transcription, suggests Chem. 266,

6 6. Pederson, P. L., and Carafoli, E. (1987) Trends Biochem. Sci. 12, Pederson, P. L., and Carafoli, E. (1987) Trends Biochem. Sci. 12, Castilho, B. A., Olfson, P., and Casadaban, M. J. (1984) J. Bacteriol. 158, Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual, 2nd Ed, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY 10. Lundie, L. L., Jr., Yang, H., Heinonen, J. K., Dean, S. I., and Drake, H. L. (1988) J. Bacteriol. 170, Jarrell, K. F., and Sprott, G. D. (1982) J. Bacteriol. 151, Lohmeyer, M., and Friedrich, C. G. (1987) Arch. Microbiol. 149, Robishaw, J. D., Kalman, V. K., Moomaw, C. R., and Slaughter, C. A. (1989) J. Biol. Chem. 264, Bryson, M. F., and Drake, H. L. (1988) J. Bacteriol. 170, Transport Magnesium in S. typhimurium Silver, S., and Lusk, J. E. (1987) in Ion Transport in Prokaryotes (Rosen, B. P., and Silver, S., eds) pp , Academic Press, Orlando, FL 16. Webb, M. (1970) Biochim. Biophys. Acta 222, Willecke, K., Gries, E."., and Oehr, P. (1973) J. Biol. Chem. 248, Gangola, P., and Rosen, B. P. (1987) J. Biol. Chem. 262, Lynn, A. R., and Rosen, B. P. (1987) in Zon Transport in Prokaryotes (Rosen, B. P., and Silver, S., eds) pp , Academic Press, Orlando, FL 20. Kass, E. H. (1989) Reu. Infect. Dis. 11 (suppl.), S167-Sl Iandolo, J. J. (1989) Annu. Reu. Microbiol. 43, Bettelheim, K. A., Dwyer, B. W., Smith, D. L., Goldwater, P. N., and Bourne, A. J. (1989) Med. J. Aust. 151, Miller, J. F., Mekalonos, J. J., and Falkow, S. (1989) Science 243,