Magnesium transport in Escherichia coli.

Escherichia coli has a specitic active transport system for magnesium which can be studied with the radioisotope 28Mg. The accumulation of 2*Mg is temperature-dependent and inhibited by cyanide, dinitrophenol, and m-chlorophenyl carbonylcyanide hydrazone. The initial rate of uptake of SMg shows typical saturation kinetics with a & = 18 /LM magnesium and a V,,,,, = 0.56 pmole per min per 1012 cells at 25” ix$Tris medium. Neither calcium nor strontium compete with, inhibit, or stimulate the transport of 28Mg. However, manganese shows complex effects on celhrlar magnesium metabolism. (a) Manganese appears to be a competitive inhibitor for the magnesium accumulation system with a K; = 0.5 nm manganese in Tris. (b) The addition of high (5 to 10 m&r) manganese causes the rapid loss of *KMg from previously loaded cells and a net loss of magnesium, as if manganese were displacing internal bound magnesium which then leaves the cells via the magnesium transport system.

magnesium have been studied. (i) Manganese is a competitive inhibitor of the uptake of magnesium by the cells with a Kc approximately 25 times higher than the K,,, for magnesium under comparable conditions. (ii) Once high manganese concentrations accumulate within the cell via the magnesium transport system, they cause the rapid loss of cellular magnesium, presumably by exchanging with ribosome-bound magnesium and raising the intracellular concentration of free magnesium which then leaves the cell by specific transport system(s).
After the first draft of this paper was prepared, we exchanged manuscripts with Nelson and Kennedy2 who have pursued a very similar line of investigation and with largely similar conclusions. The main differences between the results of the two studies are in the kinetic constants describing the phenomena and, when we have used Mn* salts to compete with and displace cellular magnesium, Nelson and Kennedy2 have used Co*.

EXPERIMENTAL PROCEDURE
Bacteria-Several substrains of E. coli B and K-12 were used interchangeably without noticeable effects on results.
Me&u-Two basic media were used: a low magnesium (35 PM Mg* by atomic absorption spectroscopy) dilute tryptone broth containing 4 g per liter of Difco Bacto-tryptone and 2.5 g per liter of NaCl and a Tris-glucose minimal medium containing 0.12 M Tris-HCl (pH 7.0), 80 mu NaCl, 20 mu KCl, 20 mu NH&I, 3 mu NazSOe, 0.64 mM KHzPOd, 0.201, glucose, 2 PM FeCh, and MgClz as desired, generally 1 mu for high magnesium growth conditions. %Mg-%Mg (t* = 21.3 hours) was obtained from the Hot Laboratory Division, Brookhaven National Laboratory, Upton, New York. Since our earlier report (l), the method of producing %Mg has been changed by the Brookhaven Laboratory (5) and current specific activities are about 200 PCi of 2*Mg-%A1 per 0.2 mg of Mgzf at time of shipment. This represents a lo-fold improvement and allows us to add as little as 1 PM radioactive magnesium and still obtain sufhcient incorporation to count with ease.
The general procedures for growth of cells, incorporation of radioactivity, and filtration have been previously described (1, 6-8). Briefly, for influx experiments cells were grown at 37" in nonradioactive growth medium and distributed in small flasks at room temperature.
Inhibitors of uptake or nonradioactive 570 Magnesium Transport in E. coli Vol. 246,No. 3 salts were added, followed by the addition of SMg. One-milliliter samples were filtered through 25-mm Millipore HA filters and washed.
The dried filters were glued on planchets and counted in a Nuclear-Chicago low background gas flow counter. 1. Specificity of the magnesium transport system. A, Tris-glucose-grown Escherichia coli K-12 were centrifuged and resuspended at 1.2 X 108 cells per ml in Tris-glucose medium without added magnesium but containing 25 pg per ml of chlorampheni-co1 to stop further growth.
Aliquots, 15 ml, were distributed into small flasks in a 25" shaking water bath and z*Mg (0.1 PCi per ml; 10 PM) was added at time zero to a control flask and to flasks to which 1 mM MgCL, MnCL, CaCb, or SrCL was added as indicated at -2 min.
Samples were filtered and washed as described under "Experimental Procedure." B, tryptone broth-grown K-12 cells were distributed at 20" at 2.3 X lo8 cells per ml. No chloramphenicol was added. MgC12 or MnCle was added as indicated either 2 min before or 30 min after the addition of 28Mg (0.15 pCi per ml; 10fiM).
For efflux experiments, the cells were allowed to accumulate radioactivity for several generations during growth at 37", centrifuged, and resuspended in nonradioactive medium. A series of aliquots was distributed at room temperature (in order to reduce the rate of exchange).
Samples were filtered and either (a) both the unwashed filter and a dried fraction of the filtrate were counted in the gas flow counter to determine the distribution of radioactivity, or (b) in more recent experiments the filters were washed once with 5 ml of nonradioactive medium and the fraction of radioactivity remaining on the filters was normalized by setting the first filtered sample at 100% (within an experiment replicate samples filtered at the same time varied by less than 10% from their mean value).
The advantages of the first procedure are that the actual distribution of radioactivity is measured and that small changes in the filtrate value can be determined accurately when most radioactivity remains in the cells.
The advantages of the second procedure are that there are half as many samples to count and that washing removes the 8% or so of the radioactivity which remains on the filters when 1 ml of cell-free radioactive medium is filtered but not washed.
Washing the filters increases the resolution when there is little radioactivity left in the cells. "K experiments were identical with the %Mg experiments except for the replacement of radioactive magnesium with &K purchased from International Chemical and Nuclear Corporation, Irvine, California.
Cell numbers were determined with a Petroff-Hauser counting chamber and a phase contrast microscope.
An atomic absorption attachment to a Zeiss PMQ II spectrophotometer was used to measure total magnesium in cells grown in different external concentrations of magnesium after centrifugation and washing with water ( Fig. 4). CCCP3 was purchased from Calbiochem, and dinitrophenol from Sigma. Chloramphenicol was the gift of Parke, Davis, and Company.
All other chemicals were analytical grade and special "spectral" grade salts were not used.

RESULTS
Uptake of Radioactive Magnesium-Some general properties of the magnesium transport system in E. coli are shown in the experiment in Fig. 1A. Cells were grown in Tris-glucose plus 1 mM magnesium, centrifuged, and resuspended in magnesiumfree medium containing the inhibitor of protein synthesis, chloramphenicol.' When 10 PM %Mg was added, the time course of magnesium uptake showed approximately hyperbolic kinetics, reaching an equilibrium with a little more than 10% of the radioactivity taken up by the cells. Since 1.2 x 108 cells per ml occupy 0.01% (v/v) of the space, this represents 10 + 0.01 = lOOO-fold concentration of Mg2f. Whether the intracellular magnesium is bound or free is discussed below, but it can be seen in Fig. 1A that the addition of 1 mM nonradioactive magnesium during the course of the uptake experiment results in the loss from the cells of at least 75% of the already accumulated BMg.
Nonradioactive magnesium (1 InM) added prior to 2*Mg competes for the uptake of %Mg, suggesting that the mag-* The abbreviation used is: CCCP, m-chlorophenyl carbonylcyanide hydrazone.
4 In control experiments whose data are not shown, we determined that both the initial rate of z*Mg uptake and the equilibrium concentration of z8Mg within the cells did not vary in the presence or absence of chloramphenicol. Tryptone broth-grown K-12 cells were centrifuged and resuspended in either fresh tryptone broth or Tris-glucose medium. *sMg was added and the rate of uptake was measured for 30 to 60 min at 20". The initial linear rate of uptake was graphed by the method of Lineweaver and Burk (9).
nesium uptake system can be saturated. Finally, in Fig. lA, we can see the specificity of the magnesium uptake system of E. coli, since 1 mu manganese, calcium, and strontium do not inhibit %Mg uptake. The effect of manganese on magnesium uptake is complex: over the first 10 min there is a small inhibition of uptake, but during the subsequent 50 min there is a small stimulation by manganese of magnesium uptake. This manganesestimulated uptake of BMg is very reproducible (1) and our tentative explanation for it follows other experiments under "Discussion." The experiment shown in Fig. 1A was carried out with E. coli K-12 in Tris-glucose medium, whereas our previously reported work (1) was with E. coli B in a tryptone broth with twice the concentrations of NaCl and tryptone as that used in our current experiments. We have never seen significant differences between the K-12 and B strains with regard to %Mg metabolism. There are differences associated with medium changes as seen in Figs. 1 and 2. In Fig. lB, the effects of adding nonradioactive magnesium and manganese on BMg uptake and retention in tryptone broth are shown. When 1.5 mM MgZf was added to the cells accumulating BMg in tryptone, there was an immediate cessation of uptake but no loss of previously accumulated BMg as seen in Fig. 1A after the addition of 1 mM Mg2f. Manganese at 5 mM does not stimulate magnesium uptake but is almost as inhibitory as 0.5 InM nonradioactive magnesium.
The effects of nonradioactive magnesium and manganese on the uptake of %Mg were carefully measured over a series of concentrations in both Tris-glucose and broth media. The rates of uptake of %Mg were followed by collecting samples on filters over 30-to 60-min periods. The initial rate of uptake of z*Mg was plotted according to Lineweaver and Burk (9). Brothgrown cells were studied either in broth or after centrifugation and resuspension in Tris-glucose. Both the apparent K, and V max for 2*Mg uptake are medium-dependent (Fig. 2). The Tryptone broth-grown cells were centrifuged and resuspended in Tris-glucose medium without added magnesium. In order, nonradioactive magnesium, MnC& as indicated, and ZsMg were rapidly added to the cells and the uptake of zsMg was followed over a 20-min period at 22". The initial rate of uptake was graphed as in Fig. 2. apparent K, is 18 pM magnesium in Tris-glucose and 31 PM in tryptone broth. The V,,= is 0.56 pmoles per min. 1012 cells in Tris-glucose and 0.20 pmoles per min. 1Or2 cells in broth. Although values vary slightly from experiment to experiment the direction of the differences is reproducible and the Km is higher and the VmrtX is lower in tryptone broth than in Tris-glucose.
We next addressed the nature of the inhibition of magnesium uptake by manganese seen in Fig. 1 and asked whether this inhibition showed a competitive or noncompetitive relationship with magnesium concentration.
As seen in Fig. 3, the inhibition of magnesium uptake by manganese follows classical competitive inhibition kinetics with a Ki of 0.5 mM manganese in Tris-glucase or 2%fold higher than the K,,, for magnesium. In tryptone broth for which the K,,, for magnesium is a-fold higher, the inhibition by manganese also appears to be competitive with a K; of about 2 mM.
We have also characterized the magnesium transport system as being energy-dependent, inhibited by cyanide, dinitrophenol, (1) and CCCP,6 and dependent on temperature (1).
Cellular Magnesium Content as Function of External Mgzf-Although the initial rate of uptake of %Mg is dependent on external magnesium (Fig. 2), whether intracellular magnesium varies with extracellular magnesium has been the subject of some contention. Therefore, we measured total cellular Mg* by atomic absorption spectroscopy under conditions similar to our b Unpublished data. E. coli was grownin tryptone brothsupplemented with added MgCl* at 37" for 3 hours. After centrifuging the cells and washing with water twice, the total cellular magnesium was determined by atomic absorption spectroscopy. The results from two independent experiments are shown.
In the first experiment the cells were grown to 1.0 X 109 per ml ; in the second experiment the cells grew to 1.4 X 10' per ml and, in the unsupplemented tryptone broth, 70% of the Mg* was taken up by the cells during growth.
In the second experiment, samples of cells grown in low magnesium broth were exposed for 30 min at 25" to higher magnesium concentrations ("step up") prior to centrifugation Constitutivity of the magnesium transport system ., 0 20 40 60 minutes at 24°C FIG. 5. Constitutivity of the magnesium transport system. Cells were grown to 5.8 X lo* per ml in trypt,one broth supplemented with added MgCls; 100 pg per ml of chloramphenicol were added and the cells were harvested by centrifugation, washed twice in low magnesium tryptone broth, and again resuspended in low magnesium (35 PM) tryptone broth. **Mg was added and uptake was followed by filtration at 24'. Chloramphenicol was present throughout the centrifugations and the exposure to 28Mg. The uptake of 10% of the z*Mg corresponds to 6 pmoles per 1012 cells.
In agreement with previously reported experiments (10, ll), the Mg* content of E. coli varies by less than a factor of 2 (15 to 28 pmoles/1012 cells) when the external magnesium concentration is varied by a factor of 2500 ( Fig. 4). Data from two experiments are shown in Fig. 4 and the cellular Mg"+ content shows the same less than a %-fold increase when extracellular magnesium during growth was increased from 0.1 mM to 0.1 M in both experiments. In the second experiment two additional points can be seen. (a) Cells grown in broth containing 35 PM Mg* to concentrations as high as 1.5 X log per ml have a lower Mg* content per cell both because the cells become smaller toward the end of growth and because of the onset of magnesium starvation with more than 70% of the available magnesium in the cells. (b) When such low magnesium cells are exposed to "step up" conditions of higher concentrations of magnesium for 30 min at room temperature (the conditions of the radioactive experiments in Figs.  1 to 3), the intracellular Mg* does not change as a function of extracellular Mg*. Constitutitity of Magnesium Transport System-Concentrations of magnesium above 0.15 M are growth inhibitory in our dilute tryptone broth, but this is already 6 times the total internal concentration from Fig. 4 (25 hmoles in 1012 cells which occupy about 1 ml of space and are 75 to 80% water). Since the cells can be grown in higher concentrations than are needed intracellularly, we could next ask whether the magnesium transport system is synthesized constitutively during growth in high magnesium or whether the magnesium transport system is repressible and only synthesized under conditions in which it is needed to raise the intracellular magnesium to 25 mM. Some carbohydrate transport systems in E. coli are constitutive and others repressible (12, 13). Cells were grown in low and high magnesium, centrifuged, washed, and resuspended in low magnesium broth, and their ability to concentrate BMg was measured. As seen in Fig. 5, 28Mg uptake is independent of the growth medium, showing that the transport system is constitutive. Chloramphenicol was added during the centrifuging and exposure to %Mg in the experiment in Fig. 5 to prevent the synthesis of any repressible transport system. Another similar experiment in the absence of chloramphenicol showed comparable results. The cells which had been grown in 100 mM Mg* show about a 50% inhibition in the rate of uptake of %Mg in Fig. 5. This inhibition was not reproducible and may be due to a partial repression of the magnesium transport system, but it is more likely to be due to the carry over through the washing process of nonradioactive magnesium. The carry over of less than 1 part in 1000 of the nonradioactive 100 rnM magnesium would be sufficient to account for this result.
Experiments with BMg E&r-Ideally, if the magnesium transport system is mediating two-compartment exchange kinetics between osmotically free external and internal magnesium pools and the internal magnesium concentration is independent of the external concentration, we could monitor the system either by measuring (a) uptake or accumulation or by measuring (b) efflux or turnover exchange, with completely consistent results and kinetic parameters. The limits of this simplistic approach are seen in Figs. 6 and 7 where the specificity of the efflux of %Mg was tested to compare with the experiments in Figs. l-3. The results of these two types of experiments are rather similar: calcium neither affects the uptake of magnesium nor does it influence the loss of radioactive magnesium from the cells (Fig. 6). Increasing the magnesium from 35 PM (near the K,,, for uptake) to 10 mu increases the rate of efflux about a factor of 2 (Fig. S), as would be expected since the rate of magnesium uptake at 10 mM magnesium should be about twice the rate at 35 PM magnesium. However, 5 mu Mn*, which is a poorer competitor than Mg2+ for wMg uptake, causes a rapid Cells were grown in broth for 3 hours at 37" to 5 X 108 per ml in the presence of z8Mg. After centrifugation and resuspension in fresh nonradioactive broth, 15-ml aliquots were distributed into four flasks. Samples, 1 ml, were filtered at 5-or lo-min intervals and washed once with 5 ml of tryptone broth. The control had no salts added to the resuspension broth; 10 mM MgC12 or CaC& or 5 mu MnClz was added as indicated to the other three aliquots 2 min after the first samples were filtered. loss of %Mg from the cells (Fig. 6). In a nonradioactive experiment in which magnesium was measured by atomic absorption spectroscopy, the addition of 10 mM Mn* caused the net loss of 76% of the cellular magnesium. Next we tested how the rate of loss of 28Mg would vary as a function of external concentration and whether this rate could be used to determine a half-saturation concentration for efflux much as we have measured K,,, and KS values for influx. The rate of efflux was determined by following the time course of efflux as in Fig. 6 for 30 to 60 min and then determining the rate of loss of BMg for the exponential loss region of the curve (generally the initial rate after MnClz or uMgC12 was added) as the rate constant k in the equation 28Mgo,/28Mgo = e --kt. The results, shown in Fig. 7, allow the estimation of a half-saturation concentration for *Mg efflux in tryptone broth with 0.1 rnr+r added Mg* of about 2 to 5 rnM Mnz+, which is in satisfactory accord with the Ri for Mnl+ in broth as a competitive inhibitor of %Mg uptake. Two additional points are apparent. (a) Increasing the external Mg* over a range of concentrations 1009fold higher than the K,,, for the wMg accumulation system causes an increasingly rapid rate of loss of 28Mg without any apparent maximum rate or plateau. (a) Even growth inhibitory concentrations of Mg* (0.25 M) do not cause as rapid a loss of SMg from the cells as do growth inhibitory concentrations of Mn* (10 n-@. We have not directly determined whether manganese is a substrate for as well as a competitive inhibitor of the magnesium accumulation system. Nevertheless, as shown in Figs. 8 and 9, the Mn*-induced loss of %Mg is energy-dependent and Mg* protects the cells against this effect of Mn* in a way consistent with the hypothesis that Mn* is entering the cell via the magnesium transport system. Cellular exchange of radioactive 28Mg for nonradioactive magnesium is inhibited by energy poisons such as cyanide, dinitrophenol, and aside (1, 2, 7). As can be seen in Fig. 8 the addition of dinitrophenol or CCCP by itself does not cause %Mg leakage, but rather promotes the retention of cellular radioactivity.6 Similarly, the manganese-induced loss of BMg is also inhibited by energy poisons including cyanide, dinitrophenol (Fig. 8A), and CCCP (Fig. 8B). This suggests that the manganese must enter the cell via an energy-dependent system or that intracellular manganese-magnesium exchange is energy-dependent (or both).
Formaldehyde has been reported to prevent loss of smalI molecules from the cells by sealing or "tanning" the surface transport carriers (14). The effects of adding formaldehyde on the manganese-induced loss of radioactive magnesium are seen in Fig. 9A. Formaldehyde treatment completely eliminates manganese-induced magnesium loss. In order to compare transport-mediated loss of ?Mg with passive leakage, sufficient toluene (1% v/v> was added to cause passive loss of 2eMg at a rate comparable to the manganese-induced loss. Formaldehyde does slow the rate of Ioss of ?Mg in toIuene-treated cehs, but it does not eliminate the loss as with manganese-treated cells. To pursue still further our working hypothesis that the manganese is entering the cell via a transport system and then causing the exchange loss of %Mg, we addressed the question as to which transport system(s) might be involved.
If the manganese enters the cells via the normal magnesium transport system, then added nonradioactive magnesium should competitively inhibit the manganese-stimulated loss of %Mg, as is in fact seen in so as to be able to distinguish specific effects on the magnesium transport system from more general effects on cell permeability and osmotic regulation.
Typical results with 42K-labeled cells in dilute tryptone broth are seen in Fig. 10. Whereas 1 mM Mnz+ has a discernible effect on the uptake of %Mg and 10 mM Mn2+ inhibits magnesium uptake more than 90% (Figs. 1 and 3), there is no effect of 1 mM Mn2+ on &K uptake by E. coli and 10 rnM Mn"t-(which is growth inhibitory) slightly stimulates 42K uptake over the period of 1 hour (Fig. lOA) A, uptake in tryptone broth. As in Fig. lB, except that 42K (0.1 @i per ml; 1.5 PM) was added 2 min after the addition of MnCL B, efflux or loss. As in Fig. 6, except that the cells were grown with radioactive 4zK.
(turnover) of previously loaded 42K from the cells (Fig. 10B). The manganese-induced loss of radioactive potassium does not result in a net loss of potassium since uptake is also stimulated (Fig. lOA)-it is the rate of turnover or exchange of intracellular for extracellular potassium which is increased by high manganese. The manganese-stimulated loss of &K is inhibited by energy poisons such as CCCP (Fig. 1OB). CCCP causes a very rapid loss of about 30% of the cellular &K followed by a slower rate of loss of the remaining @K. It is during this slower phase of &K loss or turnover that neither 10 mM K+ nor Mn* increases the rate of loss of &K (Fig. 10B). 100 80 60 50 Following the logic used in the experiment shown in Fig. 9B, 40 that Mg2+ g might compete for Mn2+ uptake via the normal 2 magnesium transport system and thereby prevent Mn2f-induced g loss of radioisotope from the cells, we asked whether Mg2f could g 3o prevent the manganese-stimulated loss of 42K. As seen in Fig. 11, high magnesium does protect the cells against manganese-induced loss of &K, and furthermore this effect of magnesium is specific in that high magnesium does not affect the rate of &K loss (turn-20 over) stimulated by 25 mM nonradioactive K+.
There is another side to the question of specificity. Do other transition elements have similar effects on magnesium metabolism to that of manganese? Apparently not: under the conditions of these experiments and when marginally inhibitory concentrations are added, Co*, Fe3+, Ni*, and Cr" do not -> 10 FIG. 11. Protection by magnesium of manganese-induced loss of *2K. As in Fig. lOB E. coli has a highly specific magnesium accumulation system with the usual criteria for active transport.
The uptake of BMg is temperature-dependent (1, 2) and inhibited by a wide range of energy poisons (1, 2). The initial rate of %Mg accumulation shows classical saturation kinetics with a K,,, of 31 PM and a V max of 0.2 pmole per min per 1012 cells in dilute broth. The kinetic parameters for magnesium accumulation are mediumdependent and in Tris-glucose the K, is 18 PM and the V,,, is 0.56 bmole per min per 10'2 cells (Fig. 2). Manganese is a competitive inhibitor of %Mg accumulation with a Ki of about 2 mu in broth and 0.5 mM in Tris-glucose (Fig. 3).
The rates of uptake and loss of magnesium by the cells vary with external concentration so as to maintain a relatively constant cellular magnesium content (Fig. 4) regardless of the external concentration.
For example, at 1 nuu external magnesium in broth, the rate of uptake is near the V,,, of 0.2 pmole per min per 1012 cells. The rate of turnover or efflux of radioactive %Mg is about 0.01 per min (Fig. 7), which with a cellular Mg* content of 25 pmoles per lo12 cells corresponds to 0.25 pmole per min per 10'2 cells. In Tris-glucose medium the cellular magnesium content is about the same (10) but the maximum rate of magnesium uptake (Fig. 2) and turnover2 is higher.
Cellular Magnesium-What fraction of the cellular magnesium is bound or osmotically free is not measured in our experiments, but this question is important to our understanding of the magnesium accumulation system and in particular for an understanding of the effect of manganese on magnesium metabolism. Most of the intracellular magnesium appears to be bound to ribosomes and other intracellular polyanions (see Reference 3 for a discussion of the evidence).
Effects of Manganese on Magnesium Metabolism-The interactions of manganese with cellular magnesium are clearly complex and, in addition to its function as a competitive inhibitor of uptake (Fig. 3), the displacement of %Mg from the cell by manganese must be explained (Figs. 6 to 9). Our current hypothesis for these effects has been developed under "Results" and need be brought together only briefly here. Manganese is not only a competitive inhibitor of magnesium accumulation but appears to enter the cells via this accumulation system (this has not been directly measured in our experiments').
Once within the cells, manganese freely exchanges with ribosomal bound magnesium (3,20,21), increasing the free intracellular concentration of magnesium which then leaves the cells via the magnesium transport system. At inhibitory concentrations of manganese (5 to 10 rn~), the result is a net loss of cellular magnesium without striking changes in cellular permeability toward other "free" intracellular cations such as potassium (Fig. 10). With increasingly higher concentrations of manganese, magnesium involved in membrane stability (22,23) is also displaced, leading to a more 7 The high affinity manganese-specific transport system (18,19) is not likely to be responsible for Mnz+ accumulation in the millimolar range since it has a K, of 0.2 FM and a V,, of only 2 nmoles per min per lOI* cells-1996 times lower than required for effects on cellular magnesium. general breakdown of the cellular permeability barriers (Figs. 10 and 11). This model for the action of manganese on E. coli also accounts for the sigmoidal course of %Mg uptake seen in the presence of 1 mM Mn* ( Fig. 1A and Reference 1): first the inhibition of Mg* uptake is due to direct competition for the transport carrier. Then following net magnesium loss, a type of feedback control results in an increasing rate of magnesium uptake as the cell attempts to maintain a constant cellular magnesium content (10,24). Similar explanations would account for the increased uptake of &K in 10 mM Mn* (Fig. lOA). Control over the cellular rate of sulfate active transport (25) and of sugar uptake in subcellular membrane vesicles (26) has also been described.
The kinetic analysis of magnesium accumulation leads to the prediction of a class of bacterial mutants with an altered magnesium carrier system with a lower affinity for manganese. These will be manganese-resistant in the presence of high magnesium concentrations.
We have isolated a series of manganese-resistant magnesium-dependent mutants of E. coli which appear to fuhill the requirements for carrier mutants with a higher Ki for manganese in the magnesium transport system. The growth, genetics, and magnesium transport characteristics of these strains will be described shortly.8