Energetics of glycylglycine transport in Escherichia coli

Escherichia coli, an Intestinal Microorganism, as a Biosensor for Quantification of Amino Acid Bioavailability

In animal diets optimal amino acid quantities and balance among amino acids is of great nutritional importance. Essential amino acid deficiencies have negative impacts on animal physiology, most often expressed in sub-optimal body weight gains. Over supplementation of diets with amino acids is costly and can increase the nitrogen emissions from animals. Although in vivo animal assays for quantification of amino acid bioavailability are well established, Escherichia coli-based bioassays are viable potential alternatives in terms of accuracy, cost, and time input. E. coli inhabits the gastrointestinal tract and although more abundant in colon, a relatively high titer of E. coli can also be isolated from the small intestine, where primary absorption of amino acids and peptides occur. After feed proteins are digested, liberated amino acids and small peptides are assimilated by both the small intestine and E. coli. The similar pattern of uptake is a necessary prerequisite to establish E. coli cells as accurate amino acid biosensors. In fact, amino acid transporters in both intestinal and E. coli cells are stereospecific, delivering only the respective biological l-forms. The presence of free amino- and carboxyl groups is critical for amino acid and dipeptide transport in both biological subjects. Di-, tri- and tetrapeptides can enter enterocytes; likewise only di-, tri- and tetrapeptides support E. coli growth. These similarities in addition to the well known bacterial genetics make E. coli an optimal bioassay microorganism for the assessment of nutritionally available amino acids in feeds.

Transport and hydrolysis of peptides by microorganisms

The structural specificities of the dipeptide and oligopeptide permeases of E. coli are briefly reviewed and related to the requirements found for other microorganisms. New, quick, sensitive methods for studying peptide transport are described, based on the following: (i) peptide-dependent incorporation of free radioactive amino acid into newly synthesized protein by a double amino acid auxotroph, (ii) colorimetric assay of peptide-dependent enzyme synthesis by an amino acid auxotroph, (iii) dansyl fingerprint technique. These approaches provide information on peptide binding affinity to a permease and rates of peptide uptake and amino acid efflux. Among current and future research areas considered are: the influence of the pKb of the N-terminal amino group on transport, generality of peptide transport in microorganisms, energy coupling and regulation, involvement of binding proteins, and the 'smugglin' concept. Peptide hydrolysis, and nutritional ultilization of peptides, by microorganisms are briefly discussed.

Method for enhancing solubility of the expressed recombinant proteins in Escherichia coli

The production of correctly folded protein in Escherichia coli is often challenging because of aggregation of the overexpressed protein into inclusion bodies. Although a number of general and protein-specific techniques are available, their effectiveness varies widely. We report a novel method for enhancing the solubility of overexpressed proteins. Presence of a dipeptide, glycylglycine, in the range of 100 mM to 1 M in the medium was found to significantly enhance the solubility (up to 170-fold) of the expressed proteins. The method has been validated using mycobacterial proteins, resulting in improved solubilization, which were otherwise difficult to express as soluble proteins in E. coli. This method can also be used to enhance the solubility of other heterologous recombinant proteins expressed in a bacterial system.

Synthesis of a selenomethionine peptide and a preliminary study of transport into Escherichia coli monitored by high-performance liquid chromatography

The tripeptide Gly-SeMet-Gly has been synthesized by a combination of solution and solid-phase methods. Increase in weight of the resin was very nearly theoretical, and purification was straightforward. Its absorption was compared to that of the corresponding peptide, Gly-Met-Gly, in E. coli using HPLC ion-exchange separation and fluorometric determination of the disappearance of peptides in the culture medium and the appearance of methionine and selenomethionine in the same culture medium. As E. coli are not known to possess extracellular peptidases, and in fact have been shown to possess transport systems for peptides, this absorption is interpreted as transport of the peptide through the cell wall and membrane into the cytoplasm, endohydrolysis of the peptide, and efflux of the peptides' amino acids. Uptake of both peptides was approximately equal, but was slowed when both peptides were present simultaneously.

Mechanism and energetics of dipeptide transport in membrane vesicles of Lactococcus lactis

Alanyl-alpha-glutamate transport has been studied in Lactococcus lactis ML3 cells and in membrane vesicles fused with liposomes containing beefheart cytochrome c oxidase as a proton-motive-force-generating system. The uptake of Ala-Glu observed in de-energized cells can be stimulated 26-fold upon addition of lactose. No intracellular dipeptide pool could be detected in intact cells. In fused membranes, a 40-fold accumulation of Ala-Glu was observed in response to a proton motive force. Addition of ionophores and uncouplers resulted in a rapid efflux of the accumulated dipeptide, indicating that Ala-Glu accumulation is directly coupled to the proton motive force as a driving force. Ala-Glu uptake is an electrogenic process and the dipeptide is transported in symport with two protons. In both fused membranes and intact cells the same affinity constant (0.70 mM) for Ala-Glu uptake was found. Accumulated Ala-Glu is exchangeable with externally added alanyl-glutamate, glutamyl-glutamate, and leucyl-leucine, while no exchange occurred upon addition of the amino acid glutamate or alanine. These results indicate that the Ala-Glu transport system has a broad substrate specificity.

Binding specificity of the periplasmic oligopeptide-binding protein from Escherichia coli

The structural properties required for the binding of peptide substrates to the Escherichia coli periplasmic protein involved in oligopeptide transport were surveyed by measuring the ability of different peptides to compete for binding in an equilibrium dialysis assay with the tripeptide Ala-Phe-[3H]Gly. The protein specifically bound oligopeptides and failed to bind amino acids or dipeptides. Acetylation of the peptide amino terminus of (Ala)3 severely impaired binding, whereas esterification of the carboxyl terminus significantly reduced but did not completely eliminate binding. Peptides composed of L-amino acids competed more effectively than did peptides containing D-residues or glycine. Experiments with a series of alanyl peptide homologs demonstrated a decrease in competitive ability with increasing chain length beyond tripeptide. Competition studies with tripeptide homologs indicated that a wide variety of amino acyl side chains were tolerated by the periplasmic protein, but side-chain composition did affect binding. Fluorescence emission data suggested that this periplasmic protein possesses more than one substrate-binding site capable of distinguishing peptides on the basis of amino acyl side chains.

Peptide chemotaxis in E. coli involves the Tap signal transducer and the dipeptide permease

Bacterial chemotaxis provides a simple model system for the more complex sensory responses of multicellular eukaryotic organisms. In Escherichia coli, methylation and demethylation of four related membrane proteins, the methyl-accepting chemotaxis proteins (or MCPs), is central to chemotactic sensing and signal transduction. Three of these proteins, Tar, Tsr and Trg, have been assigned specific roles in chemotaxis. However, the role of the fourth MCP, Tap, has remained obscure. We demonstrate here that Tap functions as a conventional signal transducer, enabling the cell to respond chemotactically to dipeptides. This provides the first evidence of specific bacterial chemotaxis towards peptides. Peptide taxis requires the function of a periplasmic component of the dipeptide permease. This protein represents the first example of a periplasmic chemoreceptor that does not have a sugar substrate.

Activity of a peptidyl prodrug, alafosfalin, against anaerobic bacteria

Alafosfalin, an antibacterial phosphonodipeptide requiring peptide transport for activity, was tested for activity against clinical strains of anaerobic bacteria in peptide-free Roche Sensitivity Test Medium no. 5 agar. It was active against Bacteroides spp., Fusobacterium nucleatum, and Clostridium perfringens but not against Clostridium difficile. Alafosfalin activity was antagonized by appropriate peptides. Synergy was obtained with other cell wall-active antibiotics.

Genetic analysis of Escherichia coli oligopeptide transport mutants

The composition of the outer membrane channels formed by the OmpF and OmpC porins is important in peptide permeation, and elimination of these proteins from the Escherichia coli outer membrane results in a cell in which the primary means for peptide permeation through this cell structure has been lost. E. coli peptide transport mutants which harbor defects in genes other than the ompF/ompC genes have been isolated on the basis of their resistance to toxic tripeptides. The genetic defects carried by these oligopeptide permease-negative (Opp-) strains were found to map in two distinct chromosomal locations. One opp locus was trp linked and mapped to the interval between att phi 80 and galU. Complementation studies with F'123 opp derivatives indicated that this peptide transport locus resembles that characterized in Salmonella typhimurium as a tetracistronic operon (B. G. Hogarth and C. F. Higgins, J. Bacteriol. 153:1548-1551, 1983). The second opp locus, which we have designated oppE, was mapped to the interval between dnaC and hsd at 98.5 min on the E. coli chromosome. The differences in peptide utilization, sensitivity and resistance to toxic peptides, and the L-[U-14C]alanyl-L-alanyl-L-alanine transport properties observed with these Opp-E. coli strains demonstrated that the transport systems encoded by the trp-linked opp genes and by the oppE gene(s) have different substrate preferences. Mutants harboring defects in both peptide transport loci defined in this study would not grow on nutritional peptides except for tri-L-methionine, were totally resistant to toxic peptides, and would not actively transport L-[U-14C]alanyl-L-alanyl-L-alanine.

Spectrophotometric determination of affinities of peptides for their transport systems in Escherichia coli

The use of novel synthetic peptides to measure peptide transport by spectrophotometric means is described. These peptides contain glycine residues alpha-substituted with thiophenol and are recognized as substrates by both peptide transport systems and intracellular peptidases of Escherichia coli (Kingsbury et al., Gilvarg, C., Proc. Natl. Acad. Sci. U.S.A. 81:4573-4576, 1984). Transport and peptidase cleavage results in the intracellular release of thiophenol, which exits rapidly from the cell. The release of thiophenol from these peptides by cell suspensions can be measured with Ellman sulfhydryl reagent [5,5'-dithiobis(2-nitrobenzoic acid)] and provides a direct determination of the rate of peptide transport. The reductions in thiophenol release from these peptides resulting from the addition of peptide competitors enable the affinities of the competitors for their transport systems to be determined. By this method, it is shown that the dipeptide transport system is more restrictive with respect to changes in the amino acid sidechains of its substrates than those of the oligopeptide transport system.

Periplasmic protein associated with the oligopeptide permeases of Salmonella typhimurium and Escherichia coli

A periplasmic protein essential for the function of the oligopeptide transport system of Salmonella typhimurium was identified. This protein, encoded by the oppA gene, is one of the most abundant proteins in the periplasm and, with an apparent molecular weight of 52,000, is considerably larger than any other known periplasmic transport component. A similarly abundant periplasmic protein forms part of the oligopeptide transport system of Escherichia coli.

Genetic organization of the oligopeptide permease (opp) locus of Salmonella typhimurium and Escherichia coli

The opp locus of Salmonella typhimurium and Escherichia coli encodes the oligopeptide permease. In this study we showed by complementation analysis that the locus in both species consists of four separate genes, which we named oppA, oppB, oppC, and oppD. These genes are organized as a single operon. The direction of transcription was determined.

Genetic map of the opp (Oligopeptide permease) locus of Salmonella typhimurium

The uptake of peptides by Salmonella typhimurium is mediated by three apparently independent transport systems. One of these systems, the oligopeptide permease, is encoded by a genetic locus (opp) which has been mapped at 34 min on the S. typhimurium chromosomal map. We accurately mapped the location of opp by cotransduction frequencies and by deletion analysis and show that the gene order for this region of the chromosome is cysB-trp-tonB-opp-galU-tdk. All opp mutants, independently isolated by a variety of means, mapped at this one locus, between tonB and galU. Spontaneous and transposon Tn10-generated deletions were used to construct a fine-structure genetic map of opp. Evidence is presented which indicates that opp covers a 5- to 6-kb segment of DNA and is therefore likely to consist of more than one gene.

Ammonium and methylammonium transport by the nitrogen-fixing bacterium Azotobacter vinelandii

Azotobacter vinelandii, grown with NH4+ as nitrogen source, was shown to possess an active transport system which can take up NH4+ against a concentration gradient of 58-fold. The properties of the NH4+ uptake system were investigated with the NH4+ analog CH3NH3+. The use of this analog was justified on the basis of the conclusion that the uptake of NH4+ and CH3NH3 involves a common binding site, as shown by the competitive inhibition of CH3NH3+ uptake by NH4+ (Ki approximately 3 microM). A Lineweaver-Burk plot for CH3NH3+ uptake revealed a biphasic curve, suggesting the existence of two CH3NH3+ (NH4+) uptake systems with apparent Km's for CH3NH3+ equal to 61 microM and 661 microM. The uptake of CH3NH3+ was inhibited by arsenate, as well as by cyanide or carbonyl cyanide-m-chlorophenyl hydrazone, indicating that phosphate bond energy is required.

Direct determination of the properties of peptide transport systems in Escherichia coli, using a fluorescent-labeling procedure

A direct study of peptide uptake by Escherichia coli was made using a fluorescent procedure. After incubation with the bacteria, peptides remaining in the medium were dansylated, separated chromatographically, and quantitated from their fluorescent intensities and/or from their incorporated radioactivity when tritiated dansyl derivatives were prepared. Peptide uptake was apparently not regulated and proceeded continuously until complete, with the absorbed peptides undergoing rapid intracellular hydrolysis and the excess amino acid residues leaving the cell. Thus, peptide uptake and amino acid exodus occur concurrently. However, peptidase-resistant substrates, e.g. triornithine and glycylsarcosine, which can be similarly estimated in cell extracts, were accumulated about 1,000-fold. The influence of amino acid composition and chain length on rates of transport was assessed. Different strains of E. coli showed variability in their rates of di- and oligopeptide transport. With respect to energy coupling, both the di- and oligopeptide permeases behaved like shock-sensitive transport systems.

Mechanism of autoenergized transport and nature of energy coupling for D-lactate in Escherichia coli

To fully energize the active transport systems of Escherichia coli, it is common practice to preincubate the cells for 10 min with 10 or 20 mM concentration of a compound that can serve as an energy source. This paper shows that the active accumulation of D-lactate can be achieved within 1 min with only 50 micron D-lactate serving as an energy source for its own uptake in starved cells (autoenergization). The cells were strain DL54 which had been induced by growth in the presence of D-lactate. Uninduced cells were not able to show autoenergized D-lactate uptake under these conditions. The induced cells were also able to transport proline in the presence of 100 micron D-lactate as sole energy source. The D-lactate-dependent dehydrogenase activity in inverted French press vesicles was comparable for the induced and uninduced cells. The same was true for respiration of whole cells in the presence of 20 mM D-lactate. However, the Vmax for D-lactate transport of induced cells was six times higher than that of uninduced cells. It appears that a sufficient number of high-affinity carrier molecules in the cytoplasmic membrane are necessary for the autoenergized transport of D-lactate. A similar conclusion was reached for the autoenergized uptake of glycerol-3-phosphate by Escherichia coli strain 7. The active transport of D-lactate is driven by the protonmotive force.

Energy supply for active transport in anaerobically grown Escherichia coli

Escherichia coli K-12, grown under anaerobic conditions with glucose as the sole source of carbon and energy without any terminal electron acceptor added, contains a fumarate reductase system in which electrons are transferred from formate or reduced nicotinamide adenine dinucleotide via menaquinone and cytochromes to fumarate reductase. This fumarate reductase system plays an important role in the metabolic energy supply of E. coli, grown under so-called "glycolytic conditions," as is indicated by the growth yields and maximal growth rates of mutants impaired in electron transfer or adenosine triphosphatase (uncB). In mutants deficient in menaquinone, cytochromes, or fumarate reductase, these values are considerably lower than in mutants deficient in ubiquinone or a functional adenosine triphosphatase. Electron transfer in this fumarate reductase system leads to the generation of a membrane potential, as is indicated by the uptake of the lipophilic cation triphenylmethylphosphonium by membrane vesicles prepared from cytochrome-sufficient and uncB cells. The generation of a proton-motive force by the fumarate reductase system was also demonstrated by the uptake of amino acids under anaerobic conditions in membrane vesicles of cytochrome containing and uncB cells grown under glycolytic conditions. Membrane vesicles of cytochrome-deficient cells failed to accumulate triphenyl-methylphosphonium and amino acids under these conditions, indicating that cytochromes are essential for the generation of a proton-motive force. Using glutamine uptake as an indication of the generation of ATP and proline uptake as an indication of the generation of a proton-motive force, it was demonstrated in whole cells that the proton-motive force is formed by ATP hydrolysis in cytochrome-deficient cells and by electron transfer in the uncB cells. In cytochrome-containing cells it was not possible to distinguish between these two possibilities, but the growth parameters suggest that, under glycolytic conditions, the proton-motive force is generated via electron transfer in the fumarate reductase system rather than via ATP hydrolysis.

Photoaffinity inhibition of dipeptide transport in Escherichia coli

A dipeptide containing a nitrene precursor, glycyl-4-azido-2-nitro-L-phenylalanine, has been synthesized. This compound is a photoaffinity inhibitor of dipeptide transport in E. coli. In the dark, the dipeptide is a reversible inhibitor of glycylglycine uptake by live E. coli W cells. The 14C-labeled compound is a substrate for the transport system, with a Km of 7 micrometer and V max of 5 x 10(3) molecules cell-1 s-1 (compare 9 micrometer and 1 x 10(4) molecules cell-1 s-1, respectively, for the transport of glycylglycine under the same conditions). When intact E. coli cells are photolyzed at approximately 350 nm in the presence of the photolabile dipeptide, their ability to transport either glycylglycine or unphotolyzed glycyl-4-azido-2-nitro-L-phenylalanine is irreversibly inhibited, but their ability to transport arginine is unaffected. The presence of glycylglycine in the medium during photolysis protects the cells against the light-dependent inactivation of dipeptide transport.

Mechanism of energy coupling for transport of deoxycytidine, uridine, uracil, adenine and hypoxanthine in Escherichia coli

The transport processes for uridine, deoxycytidine, uracil, adenine and hypoxanthine require an energy source and are active under anaerobic or aerobic conditions. Inhibitory effects of cyanide, arsenate, carbonylcyanide m-chlorophenylhydrazone, 2,4-dinitrophenol and N,N'-dicyclohexylcarbodiimide on the transport of uridine and deoxycytidine differ from the corresponding effects on the transport of uracil, adenine and hypoxanthine. The nature of these inhibitory effects supports the conclusion that uridine and deoxycytidine transport is energized either by electron transport or by ATP hydrolysis via (Ca2+ + Mg2+)-ATPase. The transport or uracil, adenine and hypoxanthine is dependent upon ATP or some high energy phosphate derivative of ATP, but is independent of (Ca2+ + Mg+)-ATPase and electron transport. Uptake of the ribose moiety of uridine by a mutant of Escherichia coli B, which lacks the transport system for uracil and intact uridine, is neither stimulated by energy sources nor inhibited by various inhibitors of energy metabolism under either aerobic or anaerobic conditions.

Identification of the structural proteins of an ATP-driven potassium transport system in Escherichia coli

The three structural proteins of the ATP-driven Kdp potassium transport system of Escherichia coli [Rhoads, D. B., Waters, F. B. & Epstein, W. (1976) J. Gen. Physiol. 67, 325-341] have been identified and found to be located in the inner membrane. The high-affinity repressible Kdp system in one of four potassium transport systems in E. coli. The Kdp proteins were identified both in growing cells as well as in heavily UV-irradiated cells infected with transducing phages carrying the kdp operon. Although all previously identified ATP-driven transport systems of Gram-negative bacteria have been shown to contain a periplasmic protein component, no evidence was found for such a component or for an outer membrane component of the Kdp system. The molecular weights of the three inner membrane proteins, KdpA, KdpB, and KdpC, were determined to be 47,000, 90,000 and 22,000, respectively.

Transport of [14C]Gly-Pro in a proline peptidase mutant of Salmonella typhimurium

The transport of [14C]Gly-Pro was examined using a mutant of Salmonella typhimurium (strain TN87) deficient in an X-Pro dipeptidase and an X-Pro-Y iminopeptidase. The dipeptide was taken up by one saturable transport system having a Km of 5.3-10(-7)M and a V of 1.4 nmol/mg dry wt cell per min. The uptake of Gly-Pro was not inhibited by amino acids or tripeptides and the transport system exhibited a rather broad side chain specificity for dipeptides. Dipeptides containing hydrophobic residues were the most potent inhibitors of this dipeptide transport system exhibiting Ki values between 10(-8) and 10(-7) M. In contrast, dipeptides containing glycine residues were particularly weak inhibitors. Finally, Gly-Pro was found to be in the intact form inside the cell and was concentrated more than 1000-fold.

Studies on phosphate transport in Escherichia coli. II. Effects of metabolic inhibitors and divalent cations

1. Study has been made of the effects of a variety of metabolic inhibitors and divalent cations (Ni2+ and Mn2+), normally after 5 min exposure, on the biphasic uptake of inorganic phosphate (Pi) exhibited by phosphate-deprived cells of Escherichia coli, strains AB3311 (Reeves met-) and CBT302 (a (Ca2+ + Mg2+)-ATPase-deficient mutant). 2. In AB3311 cells cyanide (1-10 mM) produced comparable reductions in phosphate uptake to anaerobiosis, but in both instances significant uptake was maintained. Examination of intracellular Pi concentrations showed that, despite these inhibitions, Pi is still concentrated 130 times compared to 394 times under aerobic conditions. Arsenate (100 muM) and iodoacetate (100 muM pre-exposed 15 min) both abolished anaerobic-supported uptake. Under aerobic conditions the former eliminated primary uptake while the latter reduced both phases of uptake 60%. The uncouplers, dinitrophenol (100-1000 muM) and carbonyl cyanide m-chlorophenyl hydrazone (CCCP) (50muM) produced very significant, but not complete inhibitions of both phases of uptake. Inhibitions by iodoacetate and dinitrophenol were additive while dithiothreitol protected against the effects of 50-250 mum CCCP. N,N'-Dicyclo-hexylcarbodiimide (DCCD), the potent inhibitor of membrane-bound (Ca2+ + Mg2+)-ATPase, at 10(-3) M caused significant inhibitions of aerobic- (approx. 60%) and anaerobic- (approx. 80%) supported uptakes thus suggesting some obligatory requirement for this ATPase. 3. CBT302 cells, like AB3311, supported Pi transport both aerobically and anaerobically. CCCP (50muM) reduced the primary uptake similarly to AB3311 cells, but the secondary uptake was less affected. DCCD (10(-5)-10(-3) M), as expected, showed no effects in contrast to AB3311 cells. 4. In AB3311 cells Ni2+ (10 mM) caused significant but different reductions of secondary (70%) and primary (33%) phases of phosphate uptake. Mn2+ (10 mM) showed a greater differential effect with the primary uptake being minimally affected and the secondary uptake being abolished (97%). Partial relief of these inhibitions by Mg2+ (10 mM), suggested that these ions compete with Mg2+ transport. High voltage electrophoresis studies showed that Ni2+ cause intensification in the labelling from 32Pi (i.e. during Pi uptake) of hexose phosphates and a reduction in the labelling of complex molecules left at the origin. With Mn2+, labelling of fructose 1,6-diphosphate was reduced, the triose phosphate area was intensified and an unknown area (X) was intensely labelled. When Mn2+ was combined with anaerobiosis, phosphate uptake though diminished in rate exceeded after 16 min the plateau level of uptake under aerobic conditions with Mn2+ present.

Energy coupling to active transport in anaerobically grown mutants of Escherichia Coli K12

1. Anaerobic uptake of proline requires either the presence of a coupled Mg2+-stimulated adenosine triphosphatase or anaerobic electron transport. 2. Anaerobic uptake of glutamine does not require anaerobic electron transport even in the absence of a coupled Mg+2-stimulated adenosine triphosphatase. 3. These results support previous suggestions [Berger (1973) Proc. Natl. Acad. Sci. U.S.A. 70, 1514--1518; Berger & Heppel (1974) J. Biol. Chem. 249, 7747-7755; Kobayashi, Kin & Anraku (1974) J. Biochem. (Tokyo) 76, 251-261] that two distinct mechanisms of energy coupling to active transport exist in Escherichia coli in that energization of anaerobic proline uptake requires the 'high-energy membrane state', whereas the energization of anaerobic glutamine uptake does not.

Energy coupling for methionine transport in Escherichia coli

The source of metabolic energy for the accumulation of methionine in cells of Escherichia coli was shown to differ from that for proline uptake. In contrast to proline uptake, methionine accumulation was sensitive to arsenate, and relatively resistant to azide or dinitrophenol. Adenosine triphosphatase mutant strains also differentiated between the two systems, consistent with the conclusion that, although proline uptake is driven directly by the energized membrane state, methionine uptake is not. Methionine transport is similar to that of other osmotic shock-sensitive systems in its direct utilization of adenosine 5'-triphosphate or a related compound as energy source.

Accumulation of arsenate, phosphate, and aspartate by Sreptococcus faecalis

Uptake of arsenate and phosphate by Streptococcus faecalis 9790 is strictly dependent on concurrent energy metabolism and essentially unidirectional. targinine supports uptake only in presence of glycerol or related substances; glycerol is not directly involved in transport but depletes the cellular orthophosphate pool and thus relieves feedback inhibition of transport. Uptake of phosphate and arsenate is stimulated by K+ and by other permeant cations. The results suggest that electroneutrality is preserved by compensatory movement of either H+ or OH minus. Ionophores and N,N'-dicyclohexylcarbodiimide, which prevent establishment of a proton motive force, block the accumulation of thiomethylgalactoside and of threonine but not that of arsenate or phosphate. We conclude that arsenate accumulation requires adenosine 5'-triphosphate but is not driven by the proton-motive force. However, conditions and reagents that lower the cytoplasmic pH do inhibit accumulation of arsenate and phosphate, suggesting that uptake depends on the capacity of the cells to maintain a neutral or alkaline cytoplasm. We therefore propose that phosphate accumulation is an electroneutral exchange for OH driven by adenosine 5'-triphosphate or by a metabolite thereof. Accumulation of aspartate and glutamate also requires adenosine 5'-triphosphate but not the proton-motive force and may involve a similar mechanism.