Mechanisms of active transport in isolated bacterial membrane vesicles. 8. The transport of amino acids by membranes prepared from Escherichia coli.

Abstract The studies presented in this paper demonstrate that the transport of 16 amino acids by membrane vesicles prepared from Escherichia coli ML 308–225 exhibits properties similar to those found previously for respiration-linked β-galactoside transport. Concentrative uptake of all 16 amino acids is stimulated maximally by d-lactate and by the artificial electron donor system ascorbate-phenazine methosulfate, and to a lesser extent by succinate, l-lactate, dl-α-hydroxybutyrate, and NADH. Competitive uptake and displacement studies indicate that the vesicles possess at least nine distinct amino acid uptake systems with the following substrate specificities: proline, lysine, glycine-alanine, serine-threonine, glutamic acid-aspartic acid, phenylalanine-tyrosine-tryptophan, histidine, leucine-isoleucine-valine, cysteine. In the presence of ascorbate-phenazine methosulfate, the initial rates of lysine, alanine, serine, glutamic acid, tyrosine, leucine, cysteine, and lactose transport by membrane vesicles are 42 to 102% of those exhibited by appropriately treated whole cells. Initial rates of d-lactate-dependent transport as a function of external amino acid concentration yield simple hyperbolic kinetics for all but four amino acids: histidine, leucine, isoleucine, and valine yield biphasic velocity curves. Apparent Km values determined for the amino acids are in general agreement with results reported for intact cells. Steady state levels of d-lactate-dependent amino acid accumulation vary with temperature. These levels represent a balance between rates of influx and efflux, which can be shifted readily from one steady state level to another by raising or lowering the temperature. Different amino acid transport systems exhibit different steady state temperature optima, ranging from 20°–35°. In contrast, temperature optima for the initial rate of uptake occur at the same temperature (45°) with nearly all of the transport systems, and are similar to the temperature optimum for d-lactic dehydrogenase. There is no relationship between rates of oxidation of electron donors by the membranes (NADH g d-lactate g succinate g dl-α-hydroxybutyrate g l-lactate) and the ability of these compounds to stimulate amino acid uptake. These results indicate that the site of energy coupling between respiration and amino acid transport lies between the primary dehydrogenase and cytochrome b1. Furthermore, the relative effectiveness of these electron donors in supporting uptake varies widely among the various transport systems. The d-lactate-coupled concentrative uptake of proline, lysine, alanine, threonine, aspartic acid, tyrosine, histidine, leucine, and cysteine is inhibited by anaerobiosis, by the electron transfer inhibitors cyanide, 2-heptyl-4-hydroxyquinoline-N-oxide, amytal, and oxamate, and by p-chloromercuribenzoate; inhibition by the latter compound is reversed by dithiothreitol. Furthermore, anaerobiosis, cyanide, 2-heptyl-4-hydroxyquinoline-N-oxide, and amytal cause rapid efflux of the same amino acids from the intravesicular pool. Oxamate and p-chloromercuribenzoate, however, produce little or no efflux of accumulated amino acids. Moreover, p-chloromercuribenzoate blocks tyrosine exchange and cyanide-induced tyrosine efflux; in each case inhibition is reversed by dithiothreitol. Finally, cyanide-induced efflux of proline exhibits a Km for intramembranal proline which is 350 times higher than the Km determined for d-lactate-dependent uptake of external proline, whereas the Vmax values are the same for both processes. The findings presented are consistent with the conceptual model suggested for d-lactic dehydrogenase-coupled β-galactoside transport.


SUMMARY
The membrane-bound D-lactate dehydrogenase of Escherichia coli ML 308-225 has been solubilized and purified to homogeneity.
The enzyme has a molecular weight of 75,000 f 7% and contains approximately 1 mole of flavin adenine dinucleotide per mole of enzyme. Its pH optimum before exposure to Triton X-100 is '7.5 to 8.0; after exposure it is 8.0 to 9.5.
The purified enzyme has a K, value of 1.4 X 10e3 M for D-lactate and 3.0 X 1O-2 M for L-lactate before exposure to Triton X-100; after exposure its K, values are 0.5 X 10W3 M and 2.1 X 10m2 M for the same substrates, respectively.
Active transport of a wide variety of metabolites by membrane vesicles of Bscherichia coli is coupled primarily to the oxidation of D-lactate (l-14) or reduced phenazine methosulfate (5-7, 14 15). Other oxidizable substrates, such as succinate, L-lactate, a-hydrosybutyrate, and NADH also support transport, but not nearly as effectively as n-lactate or reduced phenazine methosulfate.
The membrane-bound D-lacta,te dehydrogenase catalyzes the conversion of n-lactate to pyruvate in the absence of NAD or NADH (1, 2, 4, 7), has a characteristic flavin absorbance in partially purified preparations (3), and is inactivated by treatment m-ith 2-hydroxy+butynoic acid (7). The site of energycoupling between D-lactate dehydrogenase and active transport lies between the primary and dehydrogenase and cytochrome 61, above 80% of the membrane-bound flavin (3,4,15), and recent evidence (16) suggests that an electron transfer-coupling component may mediate the transfer of electrons between the respiratory chain and the "carriers."

Recent experiments
(17) demonstrate that guanidine extracts from wild type vesicles containing n-lactate dehydrogenase activity are able to reconstitute D-lactate-dependent oxygen consumption and active transport in membrane vesicles from an E. coli mutant deficient in D-lactate dehydrogenase.
In the present report, the purification of D-lactate dehydrogenase from E. coZi ML 308-225 is described.
It is demonstrated that the purified, homogeneous enzyme is a flavoprotein containing FAD, and several of its properties are characterized.
In subsequent reports structure-function relationships of the purified enzyme with regard to reconstitution of active transport will be examined in detail.

MATERIALS AND METHODS
All chemicals were obtained from commercial sources and were used without further purification.
u-lactate and L-lact,ate n-ere lithium salts obtained from Calbiochem. Growth of the Bacteria-E.
Standard Enzyme Assays-Two assays were used to monitor enzyme activity.
The first, the dichlorophenolindophenol assay, was monitored at 600 mp using a Beckman model DU spectrophotometer equipped with a Gilford cuvette changer. Absorbance decreases were monitored for a 3-min period at 25'; cuvettes with a l-cm light path were used.
The assay mixture contained the following in a final volume of 1 .O ml: 100 mM potassium phosphate, pH 7.8; 0.2 ml of a freshly prepared solution of dichlorophenolindophenol containing 2 mg per ml; 10 rnhl Dlactate; and an appropriate amount of enzyme. The second assay measured t,he reduction of nitroblue tetrazolium dye or &ITT' at 570 rnp (18,19). Absorbance increases were monitored as in the first assay, i.e., for a 3-min period at 25" using l-cm light path cuvettes.
The assay mixtures contained the same components as the dichlorophenolindophenol assay with the exception that 60 pg per ml of phenazine methosulfate and 30 pg per ml of the tetrazolium dye replaced dichlorophenolindophenol.
One unit of enzyme activity in the dichlorophenolindophenol assay is defined as a decrease in absorbance at 600 rnk of 1.0 optical density unit per min.
One unit of enzyme activity in the tetrazolium dye assay is defined as the amount of enzyme that results in the reduction of 1 pmole of MTT per min, using an e value of 17 mm-l cm-l (18,19). Specific activity is expressed as units per mg of protein.
Protein was measured calorimetrically (20) with crystalline bovine serum albumin as a standard.
All fractions for protein analyses were first precipitated with 10% trichloroacetic acid and solubilized by heating in 0.1 M sodium hydroxide for 30 min. Polyacrylamide Gel Eleclrophoresis-Polyacrylamide gel electrophoresis in the absence of sodium dodecyl sulfate was carried out using a pH 9.5 system (21). After electrophoresis, each gel was halved longitudinally with a mechanical slicing device and one-half was stained for protein with Coomassie blue (22) while the other half was stained for enzyme activity.
The gel halves subjected to enzyme staining were rinsed with 0.1 M potassium phosphate, pH 7.8, before incubating them in the assay mixture; incubations were carried out in the dark.
Analytical disc gels containin, v sodium dodecyl sulfate were performed according to the method of Neville (23) ; enzyme solutions to be run on these gels were prepared as described (23).
UltracentrQkgafion-Sedimentation equilibrium experiments were performed at 4-6" in a Beckman model E ultracentrifuge equipped with a temperature control unit and an electronic speed control.
Experiments were conducted at 25" in a double sector, artificial boundary cell of the capillary spill type; Rayleigh optics were used to evaluate the concent,ration changes. protein was prepared for these experiments by dialysis at 2" against at least three successive portions of 0.1 in potassium phosphate, pH 7.5, containing 1 mM mercaptoethanol.

Resolution of Flavin from Enzyme and Its Quantitative
Deter-m&&ion-The presence of flavin in enzyme preparations was monitored fluorometrically as described by Bessey et al. (24). Measurements were made using either microcuvettes (25) or cuvettes with a l-cm light path in an Aminco-Bowman spectrofluorimeter.
Control solutions were buffers against which the enzyme had been dialyzed and which were subjected to the same resolution procedures. Quantitat,ive determinations were made by comparisons with standard FAD solutions. Flavin was extracted from the enzyme by either acid extraction with 10% trichloroacetic acid (26) or acid ammonium sulfate precipitation (27). For acid extraction, enzyme solutions at O-2" contninirlg 3 mg per ml of bovine serum albumin were mixed with enough 50% trichloroacetic acid to yield a final trichloroacetic acid concentration of 10%. The precipitated protein was removed by centrifugation at 18,000 x g for 10 . , 1 /1 > , / ,I, ,.
.,I 0 At 0", 0.5-ml aliquots of the enzyme solution were rapidly mixed with 0.5 ml of 3 M potassium bromide and 1 ml of 4 M ammonium sulfate containing 0.055 ml of 1.0 N H2S04. The precipitate which formed was rapidly sedimented (1-min centrifugation at 18,000 X g) at 0" and subsequently dissolved in 1.0 ml of 0.1 M potassium phosphate, pH 7.5. The supernatant was assayed for flavin as described above.
Miscellaneous Procedures-Absorption spectra were measured in a Cary model 11 recording spectrophotometer.

Enzyme Pur$ication
A summary of a typical purification is presented in Table I. All steps were carried out at O-4" unless otherwise noted.
Step 1: Crude Eztra&--Frozen cells (150 g) were suspended in 3 times their weight (450 g) of 0.02 M potassium phosphate, pH 7.2, containing 1 rnM mercaptoethanol and 100 mg per liter of phenylmethylsulfonyl fluoride. After thawing at 4", the cells were disrupted by two passages through a Gaulin homogenizer at 12,000 to 13,000 p.s.i.
The extract was centrifuged at 10,000 x g for 10 min and the precipitate was discarded.
Step 2: Ammonium Xulfale PrecipUion-Ammonium sulfate, 17.6 g per 100 ml, was added to the crude extract (620 ml) and allowed to equilibrate for 1 hour with constant stirring. The precipitate was recovered by centrifugation at 30,000 x g for 30 min and was suspended in 0.05 M Tris-chloride, pH 8.0.
Step S: Sodium Perchlorate Extraction-The suspension described in Step 2 (150 ml) was adjusted to 0.02 M n-lactate by addition of 290 mg of lithium n-lactate and enough 8.0 M sodium perchlorate (7.9 ml) was added to yield a final concentration of 0.4 K The suspension was allowed to stand for 20 minutes at O-2" and was centrifuged for 90 min at 24,000 rpm in a Beckman model L ultracentrifuge using an SW 25.2 rotor. The precipitate was discarded.
Step 4: DEAE-cellulose Chromatography-The supernatant from Step 3 was passed through a Sephadex G-25 (medium) column (4 X 40 cm) equilibrated with 0.05 M Tris-chloride, pH  The details of A, B, and C are described in Steps 4, 5, and 6, respectively, of the purification. Enzyme activity was measured using the dichlorophenolindophenol assay.
Because of the Triton X-100, protein in B and C was monitored calorimetrically after trichloroacetic acid precipitation (20). Elution rates were 1 to 2 ml per min; fractions in A, B, and C were 28, 6, and 5 ml, respectively. The black bars denote the fractions pooled for the next step or for the final enzyme concentrate.
7.4, containing 1 mM mercaptoethanol, 0.01 M n-lactate, and 100 mg per liter of phenylmethylsulfonyl fluoride. The protein in the excluded volume was applied to a microgranular DEAE-cellulose column (4 x 50 cm) equilibrated with 0.05 M Tris-chloride, pH 7.4. After washing the column with 750 to 1000 ml of equilibrating buffer, it was developed with a 5000.ml linear gradient of sodium chloride (from 0 to 0.35 M) in equilibrating buffer (Fig. 1A).
Step 5: DEAE-cellulose Chromatography-The fractions from Step 4 which contained the peak of enzyme activity (Fig. IA to 400 ml of equilibrating buffer, a 1,400ml linear gradient from 0 to 0.20 M sodium chloride (in equilibrating buffer) was applied (Fig. IB).
Step 6': DEAE-cellulose Chromatography-The fractions from Step 5 containing the peak of enzyme activity were pooled and concentrated in an Amicon ultrafiltration apparatus using a UM-IO membrane.
The concentrate was dialyzed overnight against 0.02 M potassium phosphate, pH 7.0, containing Ire Triton X-100 and 1 mar mercaptoethanol. The dialysate was applied to a column of microgranular DEAE-cellulose (0.9 x 25 cm) previously equilibrated with 0.02 nz potassium phosphate, pH 7.0, containing 1% Triton X-100. After washing with 200 ml of equilibrating buffer, the column was developed using a linear gradient arranged so that the reservoir contained 500 ml of 0.02 hf potassium phosphate, pH 7.0, 1% Triton X-100, and 0.25 11 sodium chloride, while the mixing chamber contained 500 ml of 0.02 $1 potassium phosphate, pH 7.0 and 1% Triton X-100 (Fig. 1C). The fractions with maximal enzyme activity (Fig. lC, black bar) were concentrated by ultrafiltration using Sartorius colloidion bags. The enzyme was stored in liquid nitrogen and in the presence of 25% glycerol. Under these conditions, the preparation lost less than 10% of its activity in 1 month.

Criteria of Purity
When subjected to gel electrophoresis on standard analytical disc gels (21), the enzyme exhibited a single protein band which could be stained with either Coomassic blue or after incubation in an enzyme assay mixture ( Fig. 2A).
The enzyme also migrated as a single protein species on sodium dodecyl sulfate gels ( Fig. 2A) and exhibited linear characteristics in plots of sedimentation equilibrium data (see below). Membrane vesicles (l-17) or enzyme extracts which had not been exposed to Triton X-100 contained only a small amount of this enzyme species and a large amount of activity remained in the stacking gel (Fig. 2B, Gels 1 and 3). After exposure of the vesicles or enzyme extracts to Triton X-100, however, the major portion of the enzyme activity migrated exactly as did the purified enzyme, whereas the enzyme activity previously trapped in the stacking gels disappeared (Fig. 2B, Gels 2 and 4).
Finally, dialysis of the Triton X-loo-treated extract caused the activity to remain in the stacking gel (Gel 5).2

Molecular
Weight-Compared to standards (i.e. albumin dimer, phosphorylase, albumin, catalase, and ovalbumin), the enzyme had a molecular weight of 74,000 5 5,000 when it was electrophoresed in the presence of sodium dodecyl sulfate. Assuming a partial specific volume (8) of 0.73, molecular weights of 73,500 and 76,000 could be calculated from high and low speed sedimentation equilibrium data (Fig. 3). As noted above, the linearity of this plot (Fig. 3) indicated that only one species of protein was present in the homogeneous preparation with respect to molecular weight. Absorption Speclrum-The pure enzyme exhibited a peak of absorption at 440 to 480 rnp (Fig. 4). In the presence of 10 m&I n-lactate or 25 mM L-lactate, the absorption in this area was markedly reduced (Fig. 4, (Fig. 4, ---). Identification of FAD in D-LUCti C Dehydrogenase-The absorption spectrum of the enzyme (Fig. 4)  hydrogenase using a pH 9.5 system (21) in the absence of sodium dodecyl sulfate (Gels 1 and 2). The gel was loaded with 35 pg of enzyme protein.
After running, it was split longitudinally and one-half (Gel 1) was stained with Coomassie blue while the other half was stained for enzyme activity (Gel 2). The anode is at the bottom of the gels. Gel 3 is a polyacrylamide disc gel of purified enzyme using the sodium dodecyl sulfate system of Neville (23).
The gel was loaded with 5 to 10 pg of enzyme protein pretreated as described (23). The gel was stained with Coomassie blue; the anode is at the bottom of the gel. This protein band can be stained for n-lactate-dichlorophenolindophenol reductase activity using nonfixed gels and an incubation mixture analogous to the enzyme assay reaction described under "Materials and Methods." The enzymatically active protein appears as a light band against a blue background.
B, Polyacrylamide disc gels of membrane vesicle preparations disrupted by sonication (Gel 1); membrane vesicles treated with Triton X-100 (Gel 2); the enzyme extract from Step 2 of the purification (Gel 3); the enzyme extract from Step 2 of the purification after exposure to Triton X-100 (Gel 4); and the enzyme extract from Step 2 of the purification after exposure to Triton X-100 and dialysis overnight to remove Triton X-100 (Gel 5). In each case, the pH 9.5 system (21) was used in the absence of sodium dodecyl sulfate, the gels contained 35 to 50 rg of protein, and the gels were stained for enzyme activity as described under "Materials and Methods." The anode is at the bottom of the gels. Membrane vesicles were obtained as described previously (28).
flavin obtained by either acid ammonium sulfate or trichloroacetic acid treatment of the enzyme suggested it was FAD. The flavin extracted from the n-lactate dehydrogenase was able to restore n-amino acid oxidase activity, an FAD-specific function (29). By paper chromatography it migrated in two solvent systems (27) as did authentic FAD and its pH-dependent fluorescence was the same as FAD (data not shown).
The FAD content of the pure enzyme was determined fluorometrically (see "Materials and Methods" and Table II). An average value of 1.21 moles per mole of enzyme using the observed molecular weight of 75,000 for the enzyme was determined.

Characteristics of th.e Enzyme Reaction
pH Optima-The pH optimum of the purified enzyme was between 7.8 and 8.7 (Fig. 5A).
By comparison, the pH optimum of the enzyme preparation prior to Triton X-100 treatment was 7.2 to 8.0 (Fig. 5B), and the pH optimum for n-lactate oxidation in intact membrane vesicles was 6.2 to 6.8 (Fig. 5C)  Prior to these studies, the enzyme was passed through a Sephadex G-25 (medium grade) column (0.42 X 22 cm) equilibrated in 0.1 M potassium phosphate, pH 7.5. This procedure was vital since the acid ammonium sulfate precipitation was not consistently successful in resolving FAD from the enzyme in the presence of 1% Triton.
Spectra in the presence of D-or L-lactate were taken at least 10 to 15 min after the addition of substrate.
was the finding that Tris-chloride buffers significantly altered the activity of intact membrane preparations (Fig. 5C) and enzyme preparations not exposed to Triton X-100 (Fig. 5B), whereas they had negligible effects on the purified enzyme (Fig.  5A).
Xpeci$city, Kinetics, and Product of Reaction-The enzyme was active with n-lactate but not with n-glycerate, L-glycerate, succinate, malate, n-tartrate, L-tartrate, meso-tartrate, l-propanol, or isopropanol at concentrations as high as 0.1 M. NAD, NADP, FMN, and added FAD had no effect on enzyme activity. L-Lactate was a substrate for the purified enzyme, but the K,,, by guest on March 21, 2020 http://www.jbc.org/ Step 2 enzyme not exposed to Triton X-100 (B), and membrane vesicles (C). Activity was assayed as described under "Materials and Methods" using the dichlorophenolindophenol assay in A and B, and as D-h&tCdependent oxygen uptake (2) in C. value was significantly higher and the Vmaxwas significantly lower than that observed with n-lactate (Table III).
The specificity for n-lactate was also observed with intact membranes and in enzyme preparations not exposed to Triton X-100, and analogous differences in K, values for D-and L-lactate also were detected in these preparations.
Pyruvate was the product of the enzymatic reaction using either n-lactate or L-lactate as substrate.
The purified enzyme was irreversibly inhibited by treatment with 2-hydroxy3-butynoate as described previously for partially purified enzyme preparations (7). Oxamate and oxalate were competitive inhibitors with Ki values of 3.4 and 0.9 PM, respectively.

DISCUSSION
The membrane-bound u-lactate dehydrogenase in E. coli has been purified to homogeneity as judged by gel electrophoresis and equilibrium ultracentrifugation. The enzyme contains approximately 1 mole of F,4D per mole of enzyme and has a molecular weight of approximately 75,000. FAD can be resolved from the enzyme to yield apoenzyme; however, attempts to reactivate the apoenzyme with FAD have been unsuccessful thus far.
This enzyme is different from the pyridine nucleotidedependent n-lactate dehydrogenase previously described by Tarmy and Kaplan (32,33) in that it catalyzes the conversion of n-lactate to pyruvate; it is not dependent upon addition of NAD or NADH; it is not inactivated by sulfhydryl inhibitors; and it is localized on the cytoplasmic membrane.
Since the pyridine nucleotide-dependent enzyme catalyzes the conversion of pyruvate to o-lactate and the flavin-linked enzyme the conversion of n-lactate to pyruvate, it seems likely that the former enzyme generates n-lactate in the intact cell which is then utilized by the membrane-bound enzyme to drive active transport and perhaps other processes.
Alternatively, n-lactate may be produced from methylglyoxal as suggested recently by Cooper and Anderson (34).
The differences in pH optima and K, values before and after exposure to Triton X-100 may represent conformational or structural alterations resulting from environmental changes, i.e. removal of the enzyme from an organized array of proteins and phospholipids in the membrane to a less organized aqueous environment.
Just as likely, t,hese alterations could be the result of tightly bound detergent which could alter the conformation of the enzyme secondary to changes in hydrophobic and hydrogen bond interactions. One final possibility that might account for differences in the K, values before and after exposure to Triton X-100 could be the localization of the enzyme in or on the membrane in its native state. Diffusion barriers and local pH changes causing kinetic differences in membrane-