Sugar Phosphate:Sugar Transphosphorylation and Exchange Group Translocation Catalyzed by the Enzyme II Complexes of the Bacterial Phosphoenolpyruvate:Sugar Phosphotransferase System*

The bacterial phosphoenolpyruvate:sugar phosphotransferase system catalyzes the concomitant transport and phosphorylation of many sugars. Sugar phosphorylation with phosphoenolpyruvate requires several soluble proteins as well as the integral membrane Enzyme II complex. We have demonstrated transfer of phosphate from sugar-P to ‘%-sugar in extracts of Salmonella typhimurium, Escherichia coli, and Staphylococcus aureus. Transfer occurred specifically from glucose-6-P to glucose, mannose, or methyl cu-glucoside; from mannitol-1-P to mannitol; from sorbitol-6-P to sorbitol; and in S. aureus from galactose-6-P to methyl @thiogalactoside. The stoichiometry, phosphoryl donor specificities, and phosphoryl acceptor specificities of the transphosphorylation reactions were determined. Mutant analyses, induction and competitive


The bacterial
phosphoenolpyruvate:sugar phosphotransferase system catalyzes the concomitant transport and phosphorylation of many sugars. Sugar phosphorylation with phosphoenolpyruvate requires several soluble proteins as well as the integral membrane Enzyme II complex. We have demonstrated transfer of phosphate from sugar-P to '%-sugar in extracts of Salmonella typhimurium, Escherichia coli, and Staphylococcus aureus. Transfer occurred specifically from glucose-6-P to glucose, mannose, or methyl cu-glucoside; from mannitol-1-P to mannitol; from sorbitol-6-P to sorbitol; and in S. aureus from galactose-6-P to methyl @thiogalactoside.
The stoichiometry, phosphoryl donor specificities, and phosphoryl acceptor specificities of the transphosphorylation reactions were determined. Mutant analyses, induction studies, and competitive inhibition studies established that each transphosphorylation reaction was catalyzed by the Enzyme II complex specific for the substrate sugar. Extraction of soluble and peripheral proteins from the membrane with butanol and urea did not diminish activity, and full activity was observed in membranes prepared from a strain which lacked the soluble proteins of the phosphotransferase system due to deletion of the pts and err genes.
In vitro kinetic studies showed that the pH optima for sugar phosphate:sugar transphosphorylation were more acidic than the corresponding phosphoenolpyruvate-dependent sugar phosphorylation reactions. Optimal reaction rates were observed when the concentration of substrate sugar phosphate was high (>lO mu) and the concentration of substrate sugar was low (400 PM).
Substrate inhibition was a characteristic of the transphosphorylation reaction. A dependence on divalent cation was noted, but this dependence was less pronounced than that for phosphoenolpyruvate-dependent sugar phosphorylation. Studies with intact bacteria showed that under appropriate conditions sodium arsenate, which reduced the intracel-* This study was supported by National Science Foundation Grant BMS 73-00802 A01 and Public Health Service Grant 1 ROl CA 165521-01Al MBY. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "Wuertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
$ Materials -['4C1Mannitol-l-P was synthesized by phosphorylation of P4C]mannose with ATP and crystalline yeast hexokinase and subsequent quantitative reduction with NaBH, (12). The reduced product was purified by paper chromatography employing ethyl acetateacetic acid:formic acid:water (18:3:1:4) as solvent (12, 13). Glucose-6-32P was synthesized by phosphorylation of glucose with [A-02P]ATP and hexokinase followed by paper chromatographic separation of the radioactive products. 6-0-Tosylgalactose and the (Y and p anomers of 6-0-tosylmethylgalactoside were generously supplied by Dr. J. B. Hays (14). Other compounds were prepared as described (1) or were obtained commercially at the highest purity available. Sugars were of the D configuration and glycosides were pyranosides unless otherwise indicated. Bacterial Strains -Bacteria were grown as described in previous communications (15)(16)(17). Most of the strains employed in the present studies are listed in Tables I and II and have been described (15 18). SaZmonelZa typhimurium strain LJ144, which carries the F'198 episome and exhibits 5-fold elevated activities of Enzyme I, HPr, and the Enzyme III"" (18), was used as a source of the soluble proteins of the PTS. These proteins were partially purified as described previously (7,19). Preparation of Enzyme II Complexes -Cells were routinely disrupted by passage through a French pressure cell at 10,000 p.s.i. A second procedure, used for preparation of butanol/urea-extracted membranes (8) from Escherichia coli and S. typhimurium cells, involved treatment with lysozyme before passage through a French press. Washed cells (2 to 4 g, wet weight) were suspended in 20 ml of 25 rnM Tris/HCl buffer, pH 8.0, containing 0.5 rnM dithiothreitol, 10 rnM EDTA, and 2 mg of lysozyme. The suspension was stirred at room temperature for 30 min and the following additions were made: MgCl, (15 mM, final concentration), DNase (0.2 mg), and RNase (0.1 mg). After an additional 10 min, EDTA was added to a final concentration of 20 rnM and the suspension was centrifuged at 20,000 rpm for 30 min in the SS34 rotor. The membrane pellet was resuspended in 5 ml of 25 mM Tris/HCl, pH 7.4, containing 0.5 mM dithiothreitol for passage through a French pressure cell at 10,000 p.s.i. and was subsequently sedimented by centrifugation at 200,000 x g for 2 h.
Membranes, prepared by the lysozyme-French press method, were extracted with urea and butanol by a modification of the procedure of Kundig and Roseman (8). Washed membranes from 2 to 4 g (wet weight) of cells were resuspended in 20 ml of 25 mM Tris/ HCl buffer, pH 7.4, in a 50-ml Erlenmeyer flask. Crystalline urea (IO g) was added with slow stirring at room temperature, and after complete dissolution of the urea, 0.8 ml of 1-butanol was added. The preparation was chilled to 0" and was rotated at this temperature for 1 h at 60 rpm. Particulate material was separated from the supernatant fraction by centrifugation (3 h at 200,000 x g). The resuspended membranes were dialyzed twice against 2 liters of 25 rnM Tris/HCl buffer, pH 7.4, containing 1 mM EDTA and 0.5 mM dithiothreitol.
Enzyme II activities for mannitol, methyl Lu-glucoside, and mannose were recovered quantitatively with a P-fold increase in specific activity. These activities were stable for at least a week at 0" and indefinitely at -60". Cell-free extracts were prepared from StaphyZococcus aureus strains after disruption of cells by exposure to lysostaphin and subsequent sonication as described by Simoni et al. (20). Cell debris was removed by low speed centrifugation, and the particulate fraction was separated from the soluble proteins by centrifugation at 200,000 x g for 2 h. Membranes were washed once before assay.
Assay for Sugar Phosphorylation in Vitro -The standard assay conditions for sugar-P and phosphoenolpyruvate-dependent "'Csugar phosphorylation were similar to those described previously (1, 17). Unless otherwise indicated, the reaction mixture contained 50 rnM potassium phosphate buffer, pH 7.4, 10 mM MgCl,, 20 mM KF 1 rnM dithiothreitol, 14C-sugar at 100 PM for methyl a-glucoside, or 50 PM for all other sugars, and the phosphoryl donor at 5 mM for phosphoenolpyruvate or 10 mM for sugar phosphates. The final volume was 250 ~1. Phosphoenolpyruvate-dependent reactions were measured in the presence of a saturating amount of the soluble enzymes of the PTS with from 0.5 to 10 pg of membrane protein/ tube. The transphosphorylation reactions were measured in the absence of the soluble proteins of the PTS with 50 to 1000 pg of membrane protein/tube. All assays were conducted at 37". Isolation of Enzyme II"" Mutants -A positive selection procedure described previously (16,21) was used for the isolation of mutants which were defective for the Enzyme II"l' complex of the PTS. The parental strain used in these studies was S. typhimurium strain LJ49 (ppc-201 manA54).
This strain lacked phosphoenolpyruvate carboxylase (16,21) as well as the Enzyme IP'"" complex of the PTS which exhibits specificity toward glucose, mannose, glucosamine, and fructose (16,22,23). It could not utilize sugars in minimal medium (due to the ppc-201 mutation) and fermented glucose in complex medium in a process which depended exclusively on the activity of the Enzyme II"" (due to the manA mutation) (16). Enzyme II"" mutants were isolated and characterized a5follows. Strain LJ49 was plated on minimal citrate plus glucose medium together with a crystal of N-methyl N'-nitro N-nitrosoguanidine (16), and mutants capable of growth on this medium were clonally isolated. Clones which were incapable of glucose fermentation on eosin-methylene blue fermentation agar (BBL) (16) were obtained and tested for the fermentation of various sugars. The mutants fell into two classes: 90% were deficient for glucose fermentation but were normal when tested for the utilization of sugars such as mannitol and fructose. These mutants proved to be defective for the Enzyme 11"'~ complex. About 10% of the mutants isolated had lost the ability to ferment several sugars, exhibited about 3-fold elevated Enzyme IIGIC activity in in vitro assays or employing toluenized cells with the glucose 6-phosphate-dependent transphosphorylation assay (see below) and lacked Enzyme I of the PTS. This procedure can therefore be used for the isolation of ptsI mutants.
The former mutants were further characterized with respect to (a) phosphoenolpyruvate: ['4Clmethyl a-glucoside phosphotransferase activity; (b) glucose 6-phosphate:P4Clmethyl a-glucoside phosphotransferase activity; and (c) ['*CImethyl ol-glucoside transport activity. Mutants were screened for the first two reactions employing toluenized cells (24), and the results were frequently confirmed by assay of isolated membranes for Enzyme II activity in vitro bee above). Transport rates were measured as described below. Assay for Sugar Phosphorylation in Toluenized Cells -For assay of phosphotransferase activities in permeabilized mutant strains of S. typhimurium, cells were first grown in Medium 63 containing nutrient broth and 0.4% galactose to the early stationary growth phase, harvested, washed three times with 10 rnM Tris/HCl buffer, pH 7.5, and resuspended to a cell density of 0.6 mg, dry weight/ml. For the determination of the phosphoenolpyruvate-dependent rate of sugar phosphorylation, each assay tube contained in a final volume of 200 ~1: 30 pg, dry weight, of cells, 100 FM [14Clmethyl aglucoside (specific activity = 5 &i/pmol), 5 rnM dithiothreitol, 50 rnM potassium phosphate buffer, pH 7.4,30 mM KF, 10 mM MgCl,, 5 rnM phosphoenolpyruvate, and 1 ~1 of toluene. For the assay of glucose 6-phosphate-dependent [i4C]methyl ol-glucoside phosphorylation, a similar reaction mixture was used with the following modifications: (a) the cell density was increased to 90 pg (dry weightYtube; (b) the concentration of KF was 10 mM; (c) phosphoenolpyruvate was replaced by 10 mM glucose 6-phosphate; and (d) 1 ~1 of 50% toluene in ethanol was employed. The reaction mixtures were vigorously mixed with a Vortex mixer at 0" and then incubated at 37" for 30 min.
[i4C]Methyl ol-glucoside 6-phosphate was determined by the anion exchange method described previously (7). These conditions were found to give optimal rates of phosphorylation which were proportional to cell density and time. Transport of W-Sugars -Uptake of "'C-sugars by intact bacterial cells, and efflux of nonmetabolizable sugars from the cells was conducted as described previously (16) and as specified in the figure and table legends, All assays were conducted in a modified Medium 63 (21) at 37".
Analytical Procedures -Paper electrophoresis of sugar phosphates was conducted with two buffer systems, pyridyl acetate, pH 6.5, and sodium tetraborate, pH 10. The solvent systems used for descending paper chromatographic analyses were 1-butanol:pyridine:water, 10:3:3 (Solvent A) and ethyl acetate:acetic acid:formic acid:water, 18:3:1:4 (Solvent B) (12, 13). Compounds were located on the chromatograms employing the silver nitrate-sodium hydroxide reagent (25), and radioactive compounds were located by scintillation counting after the chromatographic strips were cut into l-inch segments. Relative mobilities in Solvent B of ATP, glucose-6-P, methyl (Yglucoside-6-P, glucose, inorganic phosphate, and methyl a-glucoside were 0.4, 1.0, 2.0, 2.6, 3.5, and 3.9 in that order.  (Table I). Mutational loss of the entire soluble energy-coupling system of the PTS, i.e. Enzyme I, HPr, and the Enzyme III"'" (26), resulted in the complete loss of phosphoenolpyruvate-dependent sugar phosphorylation, but did not lower either activity, providing that the soluble enzymes were added to the assay mixture for the phosphoenolpyruvate-dependent reactions. By contrast, genetic loss of a specific Enzyme II reduced both reactions in parallel (Table  I). Thus, the glu-19 mutation, which reduced the activity of the glucose.Enzyme II complex, lowered both reactions to 20% of the wild type activity.
The manAl mutation, which affected the Enzyme II complex for the phosphorylation of glucose, mannose, glucosamine, and fructose (161, drastically reduced the phosphorylation of mannose in both reactions, and the mtl-61 and stl-113 mutations, known to abolish mannitol and sorbitol Enzyme II activities (16), simultaneously affected the two reactions for the appropriate substrates. In other experiments (not shown) the specific Enzyme II mutations did not affect the phosphorylation of sugars which were not substrates of the affected enzyme when either Penolpyruvate or sugar phosphate was the phosphoryl donor. These results are consistent with the conclusion that the Enzyme II complexes are responsible for both the P-enolpyruvate-dependent and the sugar-P-dependent reactions.
Dependence of transphosphorylation reactions on Enzymes II of phosphotransferase system of Salmonella typhimurium The bacterial strains indicated below were grown in Medium 63 containing 1% m-lactate and the sugar indicated in the table at a. concentration of 0.4%. Cells were harvested in the early stationary phase of growth, washed three times with Medium 63, resuspended in 10 mM potassium phosphate buffer containing 1 mM dithiothreitol, and ruptured by passage through a French pressure cell at 10,000 p.s.i. Cell debris was removed by low speed centrifugation, and washed membranes were prepared by repeated centrifugation at 200,000 x g and resuspension in phosphate/dithiothreitol buffer. W-Sugar phosphorylation was measured as described under "Experimental Procedures" with the i4C-sugar concentration at 50 j&M, the phosphoenolpyruvate concentration at 5 mM, and the sugar-P concentration at 10 mM. Sugar-P:sugar transphosphorylation rates were measured with from 0.1 to 1 mg of membrane protein/tube (0.25 ml). Phosphoenolpyruvate-dependent sugar phosphorylation was measured with from 1 to 25 pg of membrane protein and 250 pg of the soluble proteins from an extract of S. typhimurium strain LJ144 (18). Rates of "C-sugar phosphorylation were linear with time and amount of membrane nrotein added. In a more extensive genetic study of Enzyme IIoic function, 80 mutants which were defective for Enzyme IF'" (glu mutants) were isolated (see "Experimental Procedures"). Each of these mutants was assayed for the three known catalytic activities thought to be associated with the Enzyme IIG1' complex: (a) phosphoenolpyruvate:methyl Lu-glucoside phosphotransferase activity; (b) glucose 6-phosphate:methyl a-glucoside transphosphorylation; and (c) methyl a-glucoside transport. Most of the mutants assayed had quantitatively lost all three of the activities studied; some of the mutants had coordinately lost these activities to an incomplete degree; and only two mutants were isolated which exhibited marked differential loss of one or two of these activities without a concomitant decrease in the activity of the other reaction(s). The results suggest that a close relationship (and possibly an interdependence) exists between the different Enzyme II-catalyzed reactions.  Table II. As can be seen, growth of cells in the presence of glucose enhanced the rates of transphosphorylation with methyl a-glucoside and mannose as phosphoryl acceptors without enhancing the rate of mannitol or sorbitol phosphorylation.
These latter two rates were specifically enhanced by growth in the presence of mannitol or sorbitol, respectively.
Thus, the inducer specificities for transphosphorylation correspond to the inducer specificities reported previously for the Enzyme II complexes of the PTS (17,27,28,29 (15) were grown in minimal Medium 63 containing the carbon source indicated below at a final concentration of 0.5%. Washed membranes were prepared and assayed for sugar-P:sugar transphosphorylation as described in the legend to Table I. Phosphoryl donors (10 mM) were as follows: glucose-6-P with methyl ol-glucoside and mannose; mannitol-1-P with mannitol; sorbitol-6-P with sorbitol. Corresponding Enzyme II induction data, obtained by assaying the phosphoenolpyruvate-dependent phosphorylation reactions, have been reported (17) Tables I and II. Strains and carbon sources were as follows: methyl a-glucoside and mannose phosphorylation:glucose-grown S. typhimurium strain LJ62; mannitol phosphorylation:mannitol-grown S. typhimurium strain LJ62; sorbitol phosphorylation:sorbitol-grown S. typhimurium strain LJ62; methyl P-thiogalactoside phosphorylation: S. aureus strain C22 grown in staphylococcal broth. For assay of sugar phosphorylation, conducted as described under "Experimental Procedures," the i4C-sugar concentration was 50 UM. and the sugar-P concentration was 10 mM.  and 6-O-tosylmethyl P-galactoside are potent and highly specific competitive inhibitors of Enzyme IF"" function (14). As shown in Table IV, Experiment 2, these analogues exerted similar inhibitory effects on the phosphoenolpyruvate and galactose-6-P-dependent phosphorylation reactions. The phosphor-y1 donor specificity of the sugar-P:methyl p-thiogalactoside transphosphorylation reaction is recorded in Table III. Of the sugar phosphates tested, only galactose-6-P was an effective phosphoryl donor. In S. oureeus, a single pair of sugar-specific proteins (the Enzyme IF and the Enzyme III"") catalyzes the phosphorylation and transport of the hexitols, mannitol and sorbitol (30). In this organism, mannitol is readily taken up from the medium and utilized as a source of carbon, although sorbitol is transported and utilized very slowly. Mannitol, which is not a substrate of the Enzyme IIG1", did not stimulate the release of free sugar from [14C]glucose-6-P.
It can also be seen that high sugar concentrations strongly inhibited the transphosphorylation reactitin (Table V). Corresponding stoichiometric data for the Enzyme IF'-catalyzed transphosphorylation reaction are recorded under Experiment 2 of Table V. The transfer of [32P]phosphate from glucose-6-"*P to methyl a-glucoside was demonstrated employing Solvent B for chromatographic separation of the radioactive products. Employing the conditions outlined in Table V, the formation of a ?lPproduct which co-chromatographed with methyl a-glucoside-6-P was observed when glucose-6-32P, membranes, and nonradioactive methyl a-glucoside were present in the incubation mixture.
Omission of methyl a-glucoside or heat inactivation of the Enzyme IIGic prevented the formation of this compound. Omission of fluoride depressed the amount of product formed, possibly because the residual sugar-P phosphatase activity was strongly inhibited by fluoride under the conditions of the assay.
The effects of pH on the relative rates of several phosphorylation reactions are shown in Fig. 1. Data are included for both the phosphoenolpyruvate-dependent and the sugar-Pdependent reactions. The pH optima for sugar-P:sugar transphosphorylation were lower than those for sugar phosphorylation with phosphoenolpyruvate.
Although some variability was noted with different membrane preparations, each of the Enzyme II-catalyzed reactions showed distinctive pH profiles. Another characteristic feature of the transphosphorylation reactions was the phenomenon of substrate inhibition. Under appropriate conditions, each of the transphosphorylation reactions was inhibited by both the sugar and the sugar phosphate substrates.
Representative data are included in Fig. 2 for the Enzyme IIMtl-and Enzyme IIS"-catalyzed reactions. In both cases, high concentrations of sugar were inhibitory, and increasing the concentration of sugar-P partially reversed this effect. The data illustrate the fact that optimal conditions were highly dependent on the concentrations of both substrates. For most systems, a low sugar concentration (400 pM) and a high sugar phosphate concentration (between 10 and 30 mM) gave optimal rates of transphosphorylation. Fig. 3 Table I with  compared with those obtained with methyl P-thiogalactoside (a substrate which binds to the Enzyme IILaC complex with low affinity (6)). The data are plotted both as a function of sugar concentration (Figs. 3,A and C) and as a function of sugar phosphate concentration (Fig. 3, B and D). Employing [YJlactose as the phosphoryl acceptor and galactose-6-P as the phosphoryl donor, inhibition of transphosphorylation was observed at high lactose concentrations (Fig. 3A). The concentration of lactose which gave maximal activity was dependent on the concentration of galactose-6-P, being highest when the concentration of the latter was high. These results suggest that lactose inhibits the transphosphorylation, in part, by competing with galactose-6-P for the sugar-P biding site. When activity was plotted as a function of galactose-6-P concentration (Fig. 3B), hyperbolic kinetics were observed when the lactose concentration was low, but sigmoidal kinetics resulted when the sugar concentration was high. The sigmoidal nature of these curves resembles those noted previously with the mannitol phosphotransferase system of Spirochaeta auruntia (1) and probably resulted from the inhibitory effect of lactose which was overcome when the galactose-6-P concentration was increased. Corresponding data, obtained when methyl P-thiogalactoside served as the phosphoryl acceptor, are reproduced in Fig.  3 In both strains, arsenate did not reduce the initial rates of [Wlmethyl ol-glucoside uptake, although the maximal extent of accumulation was depressed. Employing conditions similar to those used for the uptake experiment depicted in Fig. 4  The experiments were conducted essentially as described in Fig.  5 except that intracellular radioactivity was measured after a 3-min incubation period at 37" in the presence of the nonradioactive sugar indicated below at an extracellular concentration of 50 FM. The release of intracellular radioactive sugar is plotted as a function of the extracellular methyl cY-glucoside concentration in Fig. 6. It can be seen that 10 to 100 pM methyl o-glucoside maximally stimulated the release of intracellular sugar both in the parental and mutant strains. This concentration corresponds approximately to the K, values reported for methyl LYglucoside uptake by E. coli and S. typhimurium cells (7,24,(32)(33)(34)(35). The stimulation of cellular [Wlmethyl a-glucoside release was observed only when sugars of the gluco-and manno-configurations were present in the extracellular medium (Table VI). Other sugars were completely without effect. Release of radioactive sugar from S. typhimurium cells in the presence of extracellular sugar appeared to be representative of the different transport systems studied. Thus, the release of intracellular [Wlisopropyl P-thiogalactoside in the presence of nonradioactive extracellular sugar was observed although this process was relatively slow (data not shown). A number of the substrates of the lactose phosphotransferase system in S. aureus stimulated release of [Wlisopropyl pthiogalactoside, and several metabolizable carbon sources, including glucose and glycerol, were equally effective. Essentially all of the intracellular [Y!lisopropyl P-thiogalactoside was present as the phosphate ester (36, 37).

DISCUSSION
The present report serves to characterize the sugar phosphate:sugar exchange transphosphorylation reactions catalyzed by enzymes present in bacterial extracts. Several lines of evidence supported the conclusion that these reactions were catalyzed by the different Enzyme II complexes of the bacterial phosphotranferase system. First, loss of the catalytic activity of a specific Enzyme II, either as a result of a mutational event or following treatment with a protein reagent, resulted in loss of the corresponding transphosphorylation reaction.
Second, the phosphoryl donor and acceptor specificities corresponded to those expected for known Enzyme II complexes of the PTS. Third, induction studies showed that growth conditions which resulted in enhanced activity of a specific Enzyme II complex (as measured with phosphoenolpyruvate as the phosphoryl donor (17, 28, 29)) also enhanced transphosphorylation activity to a corresponding degree. Fourth, ptsZ mutations enhanced Enzyme IIG1' activity mea-