In Vivo and in Vitro Analyses of Single-amino Acid Variants of the Salmonella enterica Phosphotransacetylase Enzyme Provide Insights into the Function of Its N-terminal Domain*

The function of the N-terminal domain (∼350 residues) of the Pta (phosphotransacetylase) enzyme of Salmonella enterica is unclear. Results from in vivo genetic and in vitro studies suggest that the N-terminal domain of Pta is a sensor for NADH and pyruvate. We isolated 10 single-amino acid variants of Pta that, unlike the wild-type protein, supported growth of a strain of S. enterica devoid of Acs (acetyl-CoA synthetase; AMP-forming) activity on 10 mm acetate. All mutations were mapped within the N-terminal domain of the protein. Kinetic analyses of the wild type and three variant Pta proteins showed that two of the variant proteins were faster enzymes (kcat 2.5–3-fold > kcat PtaWT. Results from sedimentation equilibrium experiments are consistent with PtaWT being a trimer. Pta variants formed more hexamer than the PtaWT protein. NADH inhibited PtaWT activity by inducing a conformational change detectable by limited trypsin proteolysis; NADH did not inhibit variant protein PtaR252H. Pyruvate stimulated PtaWT activity, and its effect was potentiated in the variants, being most pronounced on PtaR252H.

synthesis and energy generation (27). Likewise, Ac-P is an important phosphorylating agent used by the cell to modulate gene expression (28 -31).
Surveys of sequenced microbial genomes reveal two classes of Pta enzymes. Class I enzymes (Pta I ) are ϳ350 residues in length, whereas class II enzymes (Pta II ) are twice as long as Pta I (ϳ700 residues) (26,32). The pduL gene of S. enterica was recently shown to encode an additional, evolutionarily distinct class of phosphotransacetylase enzymes (33).
Pta I enzymes share end-to-end homology with the C-terminal domain of Pta II enzymes; hence, it is inferred that the active site of Pta II enzymes is located within their C-terminal domain.
This assignment leaves open the question of what the function of the N-terminal domain of Pta II might be.
Earlier biochemical characterizations of the Pta enzyme of Escherichia coli (a class II enzyme) showed that its forward Ac-CoA-forming activity was allosterically regulated by nucleotides, including ADP, ATP, and NADH (negative effectors), and by pyruvate (positive effector) (34). These authors hypothesized that the N-terminal of the Pta II enzyme had a regulatory role. However, this idea was not pursued any further.
We set out to investigate the function of the N-terminal domain of the Pta II enzyme of S. enterica (encoded by pta ϩ ). Our initial approach was genetic. A positive selection was used to isolate alleles of the pta gene that encoded variant proteins with improved catalytic properties. We isolated 10 singleamino acid variants, all of which had the mutation within the N-terminal domain of the protein. We report here the kinetic analysis of three variant Pta proteins and discuss the implications of these findings.

Microbiological Techniques: Bacterial Strains and Growth Conditions
All strains used in this study were derivatives of S. enterica serovar Typhimurium strain LT2 (hereafter referred to as S. enterica). The genotypes of bacterial strains and plasmids used in this work are listed in supplemental Table 4. S. enterica strains were grown in no-carbon essential minimal medium (35) supplemented with potassium acetate (10 mM), MgSO 4 (1 mM), L-methionine (0.5 mM), and trace metals (36). Nutrient broth (NB) (37) and lysogeny broth (LB) (38,39) were used as rich media to cultivate S. enterica and E. coli strains, respectively. Working concentrations of antibiotics were 100 g/ml for ampicillin (Ap) and 20 g/ml for tetracycline (Tc) in rich medium and 2 g/ml for tetracycline in minimal medium. Growth behavior was analyzed using a 96-well microtiter dish (BD Biosciences) format using a computer-controlled Ultra Microplate Reader (Bio-Tek Instruments) equipped with the KC4 software package. The temperature of the incubation chamber was set at 37°C. Each well of the plate contained 195 l of fresh minimal medium supplemented with 10 mM potassium acetate, pH 7. Each well was inoculated with 5 l of a culture of S. enterica grown in NB medium for ϳ24 h. Growth was monitored as the increase in the optical density at 650 nm (OD 650 ). Cultures were shaken for 870 s between readings, pausing for 10 s before each reading. Each experiment consisted of seven replicates per strain, and the experiment was performed at least four times from four independent sets of cultures to ensure reproducibility and statistical significance.

Isolation of pta Alleles Encoding Variant Pta Proteins That Support Growth on Low Acetate
Localized mutagenesis (43) was performed to isolate chromosomal pta alleles encoding variant Pta proteins that would support growth on 10 mM acetate. Briefly, phage was propagated on strain JE7807, which harbored a mini-Tn10 element (44) in an open reading frame proximal to pta (i.e. open reading frame stm2340).
Phage P22 was concentrated by centrifugation at 39,191 ϫ g for 2 h at 4°C using a Beckman Coulter Avanti J-25I centrifuge equipped with a JA-25.50 rotor. Hydroxylamine mutagenesis of the phage was monitored by a plaque assay (37) using strain TR6583 as an indicator strain. Recipient strain JE4312 (⌬acs) was transduced to Tc R using mutagenized phage as donor. Tc R recombinants were replica-printed onto minimal medium plates containing 10 mM acetate as the sole source of carbon and energy, tetracycline, and the calcium chelator EGTA. Tc R recombinants that grew on 10 mM acetate were purified on the selective minimal medium and reconstructed using P22 phage grown on them as donor and strain JE4312 as recipient.

Identification of the Causative Mutations in pta Alleles
Strains carrying pta alleles encoding Pta variants capable of supporting growth on low acetate (10 mM) were grown overnight in 2 ml of NB. A 10 Ϫ1 dilution of cells was boiled for 5 min at 95°C, and debris was removed by centrifugation for 1 min at 18,000 ϫ g in a Microfuge 18 centrifuge (Beckman-Coulter). The resulting DNA preparation was used as template during PCR amplification using forward primer 5Ј-TGTAACCCGG-GCCCAAAAGACTGTAACGA-3Ј, reverse primer 5Ј-TCAC-CTCTAGACCTGACAAGGCGTTCAC-3Ј, TripleMaster polymerase (Eppendorf). The resulting 2.2-kb fragment was purified using a QiaQuick gel extraction kit (Qiagen) and was used as template for DNA sequencing using nonradioactive Big-Dye (ABI PRISM) protocols (University of Wisconsin-Madison Biotechnology Center).

Plasmid Constructions
Primers used to generate site-directed mutant alleles are shown in supplemental Table 5. The Internet-based program PrimerX (available on the World Wide Web at bioinformatics. org/primerx/) was used to help design mutagenic primers.
Plasmid pPTA21-Allele pta ϩ from S. enterica was amplified from strain TR6583 using the forward primer 5Ј-TGTAACC-CGGGCCCAAAAGACTGTAACGA-3Ј and reverse primer 5Ј-TCACCTCTAGACCTGACAAGGCGTTCAC-3Ј. The underlined bases denote the engineered 5Ј SmaI and 3Ј XbaI restriction sites, respectively. The 2.2-kb DNA fragment was bluntended, phosphorylated, and ligated into the multiple cloning site (MCS) of plasmid pCC1 (Epicenter). The presence and orientation of the insert (opposing the PT 7 promoter) was verified by restriction analysis. This plasmid (named pPTA3) was cut with restriction enzymes SmaI and XbaI. The released pta ϩ fragment was extracted from the gel and ligated into the same sites of plasmid pBAD30 (45), yielding plasmid pPTA11. The sequence encoding N-terminal domain amino residues 1-401 was amplified from pPTA11 using the forward primer 5Ј-GAGG-ATAAACCATGGCCCGTATTATTATGCTG-3Ј and reverse primer 5Ј-ATTTCCCGGGTTTACGCGCCAGCTCAGTCA-3Ј. Underlined bases denote the engineered 5Ј NcoI and 3Ј SmaI restriction sites, respectively. Construction of the 5Ј NcoI site incorporated the amino acid substitution S2A. The 1.1-kb DNA fragment was A-tailed with Vent (exo-) (New England Biolabs), gel-extracted, and ligated into the MCS of pGEM-T-Easy (Promega). The presence of the insert was confirmed by restriction enzyme analysis, and this plasmid was named pPTA19. The 1.1-kb DNA containing the Pta N-terminal domain coding sequence was released from pPTA19 using NcoI and SmaI. Following gel extraction, the fragment was ligated into the same restriction sites of cloning vector pTYB4 (New England Biolabs). The resulting plasmid encoded the N-terminal 401 amino acid residues of Pta fused to a chitin binding domain tag at the C terminus of the protein. We used restriction enzyme analysis and DNA sequencing to verify that the plasmid was constructed correctly. The resulting plasmid was 8.6 kb long and was named pPTA21.
Plasmid pPTA69-The pta ϩ gene was amplified from pPTA11 using the forward primer 5Ј-GTAACGAAAGAGGA-GCTAGCATGTCCCGTA-3Ј and the reverse primer 5Ј-TCA-CCTCTAGACCTGACAAGGCGTTCAC-3Ј. The underlined bases denote the engineered 5Ј NheI and 3Ј XbaI restriction sites, respectively. The 2.2-kb DNA fragment was A-tailed with Vent (exo-), gel-extracted, and ligated into the MCS of plasmid pGEM-T-Easy, opposite to the direction of transcription of P lacZ . This 2.2-kb pta ϩ DNA fragment was released by cutting with restriction enzyme NheI and was ligated into the same dephosphorylated site of plasmid pKLD37. The resulting plasmid encoded a Pta protein fused to a recombinant tobacco etch virus (rTEV)-protease-cleavable His 6 tag at the N terminus of the protein. Restriction enzyme analysis and DNA sequencing were used to verify the orientation of pta ϩ . The resulting plasmid was 7.7-kb long and was named pPTA69.

Purification of Recombinant Proteins
Proteins were visualized after SDS-PAGE (46) and Coomassie Blue staining (47). Protein quantification was performed using the method of Bradford (48). Purity was assessed by band densitometry using a computer-controlled Fotodyne imaging system with Foto/Analyst version 5.00 software (Fotodyne Incorporated) for image acquisition and TotalLab version 2005 software for analysis (Nonlinear Dynamics). Reported activities were adjusted to reflect enzyme purity.
Overproduction and Purification of Recombinant His 6 -Pta Proteins-E. coli strain BL21(DE3) was used as the overexpression strain. Following transformation of E. coli strain BL21(DE3) with the appropriate plasmid, a 5-ml culture was grown overnight at 30°C and was used to inoculate 500 ml (1:100, v/v) of LB supplemented with ampicillin. Cultures were grown to an OD 650 of ϳ0.6 -0.8 and placed on ice for at least 10 min. pta gene expression was induced by the addition of isopropyl-␤-D-thiogalactopyranoside (250 M), followed by overnight incubation (ϳ16 h) at 30°C with shaking (ϳ175 rpm). Cells were harvested by centrifugation for 10 min at 7,354 ϫ g at 4°C in a Beckman-Coulter Avanti J25-I centrifuge equipped with a JA-25.50 rotor. Cell pellets were resuspended in 30 ml of cold His-Bind buffer (buffer A) (Tris-HCl (20 mM, pH 7.9, at 4°C) containing NaCl (500 mM), imidazole (5 mM), and phenylmethylsulfonyl fluoride (1 mM)) and passed thrice through a chilled French pressure cell at 1.05 ϫ 10 6 kilopascals. Cell debris was removed by centrifugation at 39,191 ϫ g for 30 min at 4°C. Wild-type and variant Pta proteins were purified by Ni 2ϩ affinity chromatography on His-Bind Quick 900 cartridges (Novagen). Cartridges were equilibrated and developed with ice-cold buffers according to the manufacturer's instructions. Target proteins were first dialyzed at 4°C against buffer B (Tris-HCl (50 mM, pH 7.5, at 4°C) containing NaCl (200 mM) and EDTA (10 mM)) for 2 h, followed by an overnight dialysis against buffer C (Tris-HCl buffer (50 mM, pH 7.5, at 4°C) containing NaCl (200 mM) and 10% (v/v) glycerol) and a second 2-h dialysis against buffer C. Dialyzed proteins were flash-frozen into liquid nitrogen in ϳ50-l pellets. Proteins were stored at Ϫ80°C until used. The average purity of protein preparations was ϳ70%.
When necessary, protein production was scaled up to 8-liter cultures. Cell paste was resuspended in 80 ml of buffer A, cells were broken by French pressing, and cell extracts were manipulated as above except that extracts were loaded onto a His-TrapFF 5-ml column connected to a computer-controlled Ä KTA fast protein liquid chromatography system. Unbound proteins were eluted off of the column after extensive washing with buffer A (Ն50 ml). A 30-ml wash with 94% buffer A and 6% buffer D (Tris-HCl (20 mM, pH 7.9, at 4°C) containing NaCl (500 mM) and imidazole (1 M)) was applied to the column prior to application of a 50-ml linear gradient (6 -100%) of buffer D. A single protein peak containing His 6 -Pta protein was recorded at ϳ25% buffer D. Fractions containing His 6 -Pta were pooled and dialyzed against buffer B (containing 250 mM NaCl) for 2 h. Samples were dialyzed overnight (ϳ16 h), and again for 2 h against buffer C (containing 250 mM NaCl). The concentration of protein was determined before freezing and storage at Ϫ80°C. Protein purity was Ն90%.
Cleavage of the Hexahistidine Tag of His 6 -Pta Proteins by rTEV Protease-We used rTEV protease isolated in house (Ͼ60% homogeneity; data not shown) as described elsewhere (see, on the World Wide Web, www.cf.ac.uk/biosi/staff/ ehrmann/tools/TEVprot.html). Extended incubation of His 6 -Pta proteins with rTEV protease (2 days at 4°C) was needed to remove the His 6 tag of 40% of the Pta proteins. After elution of His 6 -Pta proteins from the affinity column, they were mixed with rTEV (1:50 rTEV:His 6 -Pta (w/w)) and extensively dialyzed, first against buffer E (Tris-HCl (50 mM, pH 8.0, at 4°C) containing NaCl (250 mM) and EDTA (10 mM)) for 2 h, followed by a 4-h dialysis against buffer F (buffer E containing EDTA at 0.5 mM). His 6 -Pta/rTEV mixtures were further dialyzed overnight against buffer G (buffer F plus dithiothreitol (3 mM)). Dialysis was then performed using buffer H (Tris-HCl (50 mM, pH 8.0, at 4°C) containing 0.5 mM EDTA, 250 mM NaCl, and dithiothreitol (2 mM) for 8 h, with a second overnight dialysis against the same buffer.
After cleavage, the protein mixtures were dialyzed into buffer A and passed over the His-Trap FF 5-ml column following the protocol described above. Protein that did not bind to the column was analyzed by 8% SDS-PAGE. Fractions containing Pta protein were pooled and dialyzed overnight against buffer B (containing 250 mM NaCl). A second dialysis was performed twice against buffer C (containing 250 mM NaCl) for at least 2 h each. Tagless Pta protein was quantified via Bradford assay, frozen in liquid nitrogen, and stored at Ϫ80°C until used.
Overproduction and Purification of the N-terminal Domain of S. enterica Pta II Enzyme-The N-terminal domain of the S. enterica Pta II enzyme (residues 1-401) was overproduced in E. coli strain BL21(DE3) and purified using the IMPACT system (New England Biolabs). After transforming plasmid pPTA21 pta ϩ into the host, a single colony was used to inoculate 10 ml of LB broth supplemented with Ap and grown overnight at 30°C. The culture was subcultured 1:100 (v/v) into 1 liter of LB supplemented with Ap. Cultures were grown to an OD 650 of ϳ0.6 -0.8 and placed on ice for 10 min. pta gene expression was induced by the addition of isopropyl-␤-D-thiogalactopyranoside (250 M), followed by overnight incubation (ϳ16 h) at 15°C. Cells were harvested by centrifugation for 10 min at 10,543 ϫ g at 4°C in a Beckman-Coulter Avanti J-20 XPI centrifuge equipped with a JLA-8.100 rotor. The cell pellet was resuspended in 35 ml of cold buffer I (HEPES (20 mM, pH 7.9, at 4°C) containing NaCl (500 mM), EDTA (1 mM)) and passed three times through a chilled French pressure cell at 1.05 ϫ 10 6 kilopascals. Cell debris was removed by centrifugation at 39,191 ϫ g for 30 min at 4°C. Extract was loaded onto a chitin column (20-ml bed volume) equilibrated with buffer I and developed according to the manufacturer's instructions with a flow rate of 200 ml h Ϫ1 . Overnight, on-column cleavage of the tag was achieved with DL-dithiothreitol (dithiothreitol, 50 mM). Fractions containing Pta N-terminal domain purified to Ͼ97% homogeneity (data not shown) were pooled and dialyzed against buffer I lacking EDTA and at pH 7.5 (4°C) overnight. Protein was dialyzed extensively against buffer J (Tris-HCl (10 mM, pH 7.5, at 4°C) containing 10% glycerol (v/v)). Protein was quantified and flash-frozen as ϳ50-l drops in liquid nitrogen before storage at Ϫ80°C.

Rabbit Polyclonal Antibodies against the Pta N-terminal Domain and Western Blot Analysis
Polyclonal antibodies were elicited by subcutaneous injection of 270 g of purified N-terminal domain as primary antigen into a New Zealand White female rabbit (Laboratory Animal Resources, University of Wisconsin, Madison, WI). Subsequent antigen boosts were performed with ϳ250 g of protein.
Western blots were performed using standard protocols (49) with a 1:5,000 dilution of antiserum as primary antibody and a 1:70,000 dilution of Immunopure donkey antirabbit immunoglobulin G (IgG, heavy and light chains) conjugated to horseradish peroxidase (Pierce) as secondary antibody. Signal was detected using the SuperSignal West Dura trial chemiluminescence kit (Pierce) and a computer-controlled Cyclone phosphor imager furnished with OptiQuant version 4.00 imaging software (Packard Instruments).

Phosphotransacetylase Activity Assays
Pta in vitro activity assays of the forward reaction (Reaction 3, Ac-CoA-forming direction) were performed by monitoring the formation of the thioester bond as the increase in absorbance at 233 nm. The amount of product formed was quantified using the extinction coefficient (⑀) of 5.55 mM Ϫ1 cm Ϫ1 (26,32). The only modification to the reported protocol was the final volume of the reaction mixture, which was increased to 1 ml; dithiothreitol was not required for Pta activity.
Assays of the back reaction (Reaction 3, Ac-P-forming direction) monitored the release of free CoASH as the increase in absorbance at 412 nm observed when the thiolate anion of 5,5Јdithiobis(2-nitrobenzoic acid) (⑀ 412 ϭ 13,600 M Ϫ1 cm Ϫ1 ) was formed upon reaction of 5,5Ј-dithiobis(2-nitrobenzoic acid) and the sulfhydryl group of CoA (50 -53). Reaction mixtures contained Tris-HCl buffer (50 mM, pH 7.5, at 37°C), NH 4 Cl or KCl, Pta protein, and one substrate (plus allosteric effector, if added). In experiments that involved allosteric effectors, we used HEPES buffer (50 mM, pH 7.5) containing 100 mM KCl to prevent destruction of pyruvate by Tris buffer (54). Preincubation times were 1 min for the forward reaction and 2 min for the back reaction. Reactions were initiated by the addition of the second substrate. Progress of the reactions was monitored using a computer-controlled PerkinElmer Lambda 40 spectrophotometer (PerkinElmer Life Sciences) furnished with the UV KinLab software package.

Lactate Dehydrogenase (LDH) Activity Assays
LDH activity was determined by monitoring the oxidation of reduced nicotinamide adenine dinucleotide (␤-NADH) (decrease in absorbance at 340 nm; ⑀ 340 ϭ 6.3 mM Ϫ1 cm Ϫ1 ) (55). Lyophilized rabbit muscle lactate LDH (Calbiochem) was used as a positive control. Briefly, LDH was dissolved in HEPES buffer (50 mM, pH 7.5, at 37°C) containing KCl (100 mM) and mixed with glycerol to 10% (v/v). The protein concentration was determined, and the resulting preparation was stored in 100-l samples at Ϫ80°C until used. Reactions were incubated for 2 min with LDH or His 6 -Pta proteins, buffer, and pyruvate (0.7 mol); reactions were initiated by the addition of ␤-NADH (0.2 mol).

Kinetic Analysis
Reactions were performed under conditions that ensured that the Pta enzyme did not consume more than 10% of the substrate. Pseudo-first order kinetic parameters were calculated using Prism version 4.0a software (GraphPad Software), fitting the data to Equation 1, where V o is initial velocity, [S] is substrate concentration, V is maximum velocity, and K m is the Michaelis constant (56).
For Ac-P titrations that exhibited positive cooperativity, data were fitted to Equation 2, where h represents the Hill constant and K 0.5 is the concentration of substrate yielding half-maximal velocity to account for positive cooperativity.
K i and K a values were calculated by plotting V o Ϫ1 against effector concentration; data were fitted to Equation 3, where E represents the effector concentration. In the absence of effec- When determining K i for NADH, K e ϭ K i , and when determining K a for pyruvate, K e ϭ K a . 4

Oligomeric State Analysis of Pta
Gel Filtration-Gel filtration was performed using a Ä KTA fast protein liquid chromatography Explorer system equipped with a Superdex 200 HR 10/30 column (Amersham Biosciences). The column was equilibrated with sodium phosphate buffer (50 mM, pH 7.4, at 4°C) containing NaCl (150 mM). Pta protein (ϳ100 g) was applied onto the column, which was developed isocratically at a rate of 0.5 ml min Ϫ1 . Molecular weight calibration was performed by using bovine thyroglobulin (670 kDa), bovine ␥-globulin (158 kDa), chicken ovalbumin (44 kDa), horse myoglobin (17 kDa), and vitamin B 12 (1.35 kDa), all components of the Bio-Rad gel filtration standards kit. The exclusion limit of the column matrix was defined using blue dextran.
Equilibrium Sedimentation-Experiments were performed at the Biophysics Instrumentation Facility (University of Wisconsin, Madison, WI) in a refrigerated (4°C) computer-controlled XL-A analytical ultracentrifuge (Beckman-Coulter). Tagless Pta WT protein was extensively dialyzed against HEPES buffer (50 mM, pH 7.5, at 4°C) containing KCl (100 mM) and then diluted to A 278 of ϳ0.8, ϳ0.4, and ϳ0.2 using the final dialysate as diluent. Run speeds ranged from 3000 to 11,000 rpm with a run at 32,000 rpm to deplete out macromolecular species at the conclusion of the experiment. A reversal was performed at 5400 rpm to investigate slowly changing aggregation over time. Data were analyzed using software developed in house.

Analysis of the NADH-induced Conformation Change of Pta
Limited trypsin proteolysis of His 6 -Pta proteins was performed as described (57), except diphenylcarbamyl chloridetreated trypsin (Sigma) was used in 20-l reactions performed in 50 mM Tris-HCl buffer (pH 7.5, at 24°C). Reactions were allowed to proceed after a 10-min preincubation with the effector molecule present at the indicated concentration. Protein composition of the reaction mixture was analyzed by 12% SDS-PAGE. Tryptic fragments were excised from the gel and subjected to in-gel tryptic digestion followed by MALDI-TOF mass spectrometry (University of Wisconsin-Madison Biotechnology Center). Peptide fingerprint data were analyzed using Mascot (available on the World Wide Web at www.matrixscience. com/) with the following modifications: carbamidomethyl (ation) of (cysteine), deamidation of asparagine and glutamine, and oxidation of methionine.

Single-amino Acid Variants of Pta Support Growth of S. enterica on Low Acetate
Growth of strains of S. enterica lacking functional Acs protein was not supported by low concentrations of acetate (Յ10 mM) in the medium (Fig. 1, compare acs ϩ pta ϩ versus acs pta ϩ ). We isolated 10 S. enterica strains carrying chromosomal alleles of pta that encoded variant Pta proteins capable of supporting growth on 10 mM acetate in the absence of a functional Acs protein. Alleles of pta encoding these variants were obtained after hydroxylamine mutagenesis of a P22 lysate grown on strain JE7807 (prpC114::MudJ pta ϩ stm2340::Tn10d(tet ϩ ). Mutagenized P22 phage was used as donor in a cross with strain JE4312 (⌬acs1231) as recipient. Tetracycline-resistant transductants were screened for the ability to grow on minimal medium containing 10 mM acetate as sole source of carbon and energy. Colonies displaying robust growth were freed of phage, P22 phage was grown on strains that grew on 10 mM acetate, and the resulting phage lysate was used as donor in a cross with strain JE4312 (⌬acs1231) as recipient. Transductants that were resistant to tetracycline and grew on 10 mM acetate were further analyzed. Analysis of pta coding sequences in 10 strains identified the following single-amino acid changes: G31D, R113H, G140S, P177S, T184I, R252H, R252C, G273D, M294I, and G296S. Growth behavior analysis of four strains demonstrated that variant Pta proteins could support growth on 10 mM acetate compared with the acs pta ϩ strain. Strains harboring mutant pta alleles achieved similar final optical densities in liquid culture as the strain containing a functional Acs protein but reached these densities at a faster rate (Fig. 1, alleles  pta104-pta107).

Gene Dosage Effect
To determine whether high levels of variant proteins had any deleterious effect on growth on 10 mM acetate, five mutant pta alleles were individually cloned into an overexpression vector. Resulting plasmids (supplemental Table 4) were introduced into strain JE8086 (⌬acs pta recA supplemental Table 4), and growth on 10 mM acetate was assessed subculturing from overnight cultures grown in LB broth supplemented with Ap. Alleles pta104 -pta107 supported growth of strain JE4312 on 10 mM acetate, but allele pta108 (encodes Pta R113H ) did not. Interestingly, allele pta108 did support growth on 50 mM acetate. No additional mutations were present in pta108 or in the promoter region, as determined by DNA sequencing (data not shown). Variant Pta R113H was not isolated or analyzed in vitro.

Optimal in Vitro Conditions for Pta Activity
To understand the effect of the mutations described above on Pta activity, it was necessary to kinetically characterize the wild type enzyme first. For this purpose, the pta ϩ allele was cloned into plasmid pKLD37, fusing a tobacco etch virus-cleavable hexahistidine tag to the N terminus of the Pta polypeptide. This plasmid was moved into E. coli strain BL21(DE3), the pta ϩ gene was overexpressed, and the His 6 -Pta WT protein was overproduced and purified (Ն70% homogeneity; data not shown). Activity assays with MES, Tris-HCl, and CHES buffers identified a pH optimum of 7.5 for Pta ( Fig. 2A).
Inclusion of salts in the reaction mixture indicated that NH 4 Cl was most efficient at stimulating Pta enzyme activity when present at 40 mM (3-fold; Fig. 2B, squares). No significant increase in activity was observed at NH 4 Cl Ն 40 mM. KCl also stimulated Pta activity but to a slightly lesser degree (2.5-fold; Fig. 2B, triangles), whereas NaCl did not stimulate Pta activity at all (Fig. 2B, circles).

Oligomeric State of Native Pta WT
We performed gel filtration and equilibrium sedimentation studies to gain insights into the oligomeric state of native and variant forms of Pta enzyme from S. enterica. Results of equilibrium sedimentation experiments suggested an average minimum species of a trimer above a protein concentration of 0.9 M (Fig. 3). A two-species model was required to fit the data (below 0.9 absorbance units), since deviation from linearity in natural log plots suggested aggregation in a loading concentrationdependent manner (Fig. 3, radial positions Ͼ50.5 cm 2 ). Reversing the gradient indicated a lower average absorbance and a  Optimal pH and salt for Pta activity. Profiles were generated for the forward (Ac-CoA-forming) reaction, using CoASH (0.6 mol) and Ac-P (6 mol). A, pH profile with activities normalized to pH 7.5, the optimal pH in the forward reaction. B, effect of various salts on phosphotransacetylase activity; Tris-HCl (50 mM, pH 7.5, at 37°C) was used as buffer. Data are presented as -fold change over no added salt. higher ratio of the observed mass to calculated Pta II polypeptide mass (M w /M s ), consistent with a loss of mass due to aggregation (data not shown). In contrast, the behavior of Pta on a gel permeation column was consistent with a mass of ϳ490 kDa, suggesting a hexameric protein (supplemental Fig. 8). It is possible that native Pta is a dimer of trimers.

Kinetic Analysis of His 6 -Pta WT
Forward Reaction (Ac-CoA-forming)-The His 6 -Pta WT enzyme displayed measurably higher affinity for CoASH (K m ϭ 162 M) than for Ac-P (K 0.5 ϭ 1 mM) ( Table 1), and the enzyme was very active (V max ϭ 142.2 mol min Ϫ1 mg Ϫ1 ) ( Table 1). The forward reaction displayed positive cooperativity with respect to Ac-P concentration. No phosphotransacetylase activity was detected in a preparation of protein purified in a strain harboring the empty pKLD37 vector.
Back Reaction (Ac-P-forming)-The back reaction was slower, with product (CoASH) release occurring at a rate ϳ7-fold slower than that of the forward reaction (V max ϭ 20.6 mol min Ϫ1 mg Ϫ1 ; Table 2). Under the conditions used, the His 6 -Pta WT enzyme displayed higher affinity for Ac-CoA (K m ϭ 329.3 M) than for P i (1.5 mM) ( Table 2). Turnover number (k cat ) and catalytic efficiency (k cat /K m ) values were calculated for the forward and back reactions and served as points of reference for comparisons with variant Pta enzymes. No phosphotransacetylase activity was detected in a preparation of protein purified in a strain harboring the empty pKLD37 vector.

Pta R252H , Pta G273D , and Pta M294I Variants Are Faster than the Pta WT Enzyme
To investigate the effect of changes in residues Met 294 , Gly 273 , and Arg 252 on Pta activity, pta alleles encoding Pta R252H , Pta G273D , and Pta M294I variants were introduced into plasmid pPTA69 by site-directed mutagenesis. The resulting His 6 -tagged proteins supported growth on low acetate (10 mM) of an acs pta recA strain (data not shown); the same strain carrying the empty vector or the plasmid containing the pta ϩ allele did not grow on acetate. Once the phenotype was confirmed, each plasmid was transformed into E. coli strain BL21(DE3) for overexpression. Under the conditions used, the purification of Pta WT and variant proteins yielded 0.6 -2 mg of Pta protein from 500 ml of culture, with a calculated purity of Ն70% (data not shown).
In the forward (Ac-CoA-forming) ( Table 1) and back (Ac-Pforming) reaction (Table 2), variant proteins Pta R252H and Pta G273D were 2-3-fold faster than Pta WT , whereas Pta M294I was slightly slower (0.7-fold) than Pta WT . In the forward and back reactions, variant and wild-type Pta proteins displayed similar affinities for their substrates (Tables 1 and 2).

NADH Inhibits Pta Activity by Changing Its Conformation
To learn more about how NADH inhibited Pta activity, His 6 -Pta WT protein was subjected to limited trypsin proteolysis, and changes in its structure were assessed by SDS-PAGE followed by MALDI-TOF mass spectrometry. Exposure of Pta proteins to trypsin in the presence of NADH resulted in appearance of two prominent fragments as a function of NADH concentration. The size of each fragment and their location within the wild-type protein was established by MALDI-TOF mass spectrometry. One fragment was the result of tryptic cleavage after residues Arg 33 and Arg 415 (42 kDa), and the second fragment resulted from cleavage after Lys 70 and Arg 415 (38 kDa) (data not shown). It should be noted that the 38-kDa peptide ran abnormally during SDS-PAGE analysis (supplemental Fig. 9). Polyclonal antibodies against the N-terminal domain of Pta reacted with both fragments. Although the trypsin cleavage site needed to generate the 38-kDa fragment (site A) was accessible in the absence of NADH, the trypsin cleavage site for the generation of the 42-kDa fragment (site B) was detected only when the concentration of NADH was 0.2-0.5 mM (Fig. 4 (closed inverted  triangles), supplemental Fig. 9A).
Similar experiments were performed with variant Pta proteins. (Fig. 4, supplemental Fig. 9). Results from experiments with variant Pta R252H showed that the mutation affected exposure of site B, showing Ͼ2-fold decrease in the band intensity ratio when the concentration of NADH reached 5 mM. (Fig. 4, circles). The effect of the M294I or G273D substitution on the ␤-NADH-induced  conformational change was not significantly different than the Pta WT protein (Fig. 4, open triangles and diamonds, respectively).

Variants Pta R252H , Pta G273D , and Pta M294I Have Altered Responses to at Least One Allosteric Effector
To determine the effect of the amino acid substitutions on NADH inhibition and pyruvate stimulation, K i and K a values were determined using fast protein liquid chromatography-pu-rified His 6 -Pta WT . To calculate K i and K a values, we plotted the reciprocal of the initial velocity (V o Ϫ1 ) of the reaction versus the concentration of the effector; data were fitted to Equation 3, where E was the effector concentration.
His 6 -Pta WT was found to have an apparent K i of 1.1 Ϯ 0.2 mM for NADH and an apparent K a of 2.6 Ϯ 0.2 mM for pyruvate (Fig. 5,  A and B, respectively). Under conditions of subsaturating levels of substrates (at the K m ), NADH inhibition of the back reaction catalyzed by His 6 -Pta WT was incomplete, with a maximum reached at 70 -80% inhibition (Fig. 6, black bars). Under the same conditions, His 6 -Pta WT enzyme activity in the presence of pyruvate was 120 -140% higher than in the absence of pyruvate (Fig. 6, hatched bars). Pyruvate overcame the inhibition by NADH. Inclusion of pyruvate (30 mM) in the reaction mixture was optimal for counteracting the inhibitory effect of NADH (Figs. 5C and 6, checkered bars).
Comparison of the response of wild-type and variant Pta proteins with allosteric effectors showed that variants Pta G273D and Pta M294I were equally sensitive to NADH inhibition as Pta WT (Fig. 6, black bars). Strikingly, NADH did not inhibit the Pta R252H variant protein.
The stimulatory effect of pyruvate was stronger on the variant proteins than on wild-type protein. Pyruvate stimulated the activity of variant proteins by at least 3-fold over the no-addition control, whereas pyruvate stimulated Pta WT activity 0.2fold over the no-addition control.
Pyruvate counteracted the negative effect of NADH more efficiently in variant proteins (Fig. 6, checkered bars). The strongest effect of pyruvate was measured with the Pta R252H protein in the presence of NADH. Under the same conditions, oxidation of NADH was not detected in the presence of pyruvate for the WT or variant Pta proteins. This indicated that Pta proteins did not have LDH activity and that the effect of pyruvate and NADH was a result of direct allosteric activation and inhibition, respectively. Rabbit muscle LDH was used as the positive control, yielding a specific activity of 531.6 Ϯ 29.1 mol of NADH oxidized min Ϫ1 mg Ϫ1 .

Variants Pta R252H , Pta G273D , and Pta M294I Variants Show Less Aggregation than Pta WT
Results from gel filtration analysis of Pta R252H , Pta G273D , and Pta M294I variant proteins revealed a lower proportion of large  aggregates compared with Pta WT (Table 3). All proteins displayed a retention time consistent with a hexamer (data not shown).

DISCUSSION
Here we have described the isolation and initial biochemical analysis of single-amino acid variants of the class II Pta enzyme of S. enterica. These variant Pta enzymes allow a strain lacking the high affinity Acs enzyme to efficiently use low concentrations of acetate (Յ10 mM) in the medium as a carbon and energy source. Ten single-amino acid changes were mapped to the N-terminal domain of Pta (supplemental Fig. 10). Four mutant alleles were chosen for biochemical analysis, but one of the four variant proteins (Pta G31D ) was too unstable in isolation and hence was not analyzed in vitro.
Together, the evidence presented supports the idea that NADH is an allosteric effector of Pta, that it probably binds to the N-terminal domain of the protein, which exerts its inhibitory effect through a conformational change of the protein, and that the N-terminal domain of Pta has separate binding sites for NADH and pyruvate.

Single-amino Acid Pta Variants with Changes in the N-terminal Domain of the Protein Are More Efficient Enzymes
Differences in substrate affinities (K m values) among variant and wild-type Pta proteins are minor (Tables 1 and 2) and probably not physiologically significant because the intracellular concentrations of Ac-P, Ac-CoA, and P i are far higher than the measured K m values (58 -61), suggesting that under such conditions, the Pta enzyme probably operates at V max when cata-lyzing Ac-P formation. The reported intracellular concentration of CoASH would be at approximately the K m of the enzyme working in the forward direction (Ac-CoA-forming), hence Pta would not operate at V max in this direction.

The Pta R252H Variant Is Different from Other Variants in Its Response to Allosteric Effectors
Pyruvate effect-The increased responsiveness of Pta R252H to pyruvate when NADH is present in the reaction mixture (Fig. 6) suggests that pyruvate and NADH may bind to the protein at different sites. Clearly, the conformation of Pta R252H is more responsive to the synergistic positive effect on the effectors than Pta WT . Interestingly, substitution of residue Arg 252 with cysteine yields a variant protein capable of supporting growth on low acetate. A three-dimensional structure of the Pta protein would help rationalize the effect of this mutation and clarify the role of this important residue.
The bioinformatics analysis by Galperin and Grishin (62) suggested that the N-terminal domain of Pta II evolved from a common ancestor of the family of enzymes that includes dethiobiotin synthase (DTBS), the cell division protein MinD, and the amidotransferases CbiA and CbiP, involved in coenzyme B 12 biosynthesis. To help us visualize the effects of the mutations we have isolated, we placed them within the reported three-dimensional structure of DTBS complexed with ADP, 7-(carboxyamino)-8-aminononanoic acid, and calcium (Protein Data Bank code 1DAF) (63) (supplemental Fig. 11). In the DTBS structure, residue Gly 30 is equivalent to Pta Gly 31 , residue Phe 121 in DTBS is equivalent to Pta Arg 113 , and residue Lys 148 in DTBS is equivalent to Pta Gly 140 . Using the DTBS structure as a model for the N-terminal domain of Pta, we hypothesize that residues Gly 31 and Gly 140 of Pta may affect the positioning of the P-loop, perhaps playing a role in binding NADH. On the other hand, residue Arg 113 may be involved in binding pyruvate, since in the DTBS structure the analogous residue is near the 7-(carboxyamino)-8-aminononanoic acid substrate (supplemental Fig. 11).
It is not clear whether pyruvate exerts its positive effect through a conformational change of the protein. If it does, the change is not as large as the one induced by NADH. Tryptic digestion of Pta WT protein in the presence of pyruvate did not reveal any differences compared with the tryptic digest obtained in the absence of pyruvate (data not shown). The effect of pyruvate can be amplified by concentrating the protein using polyethylene glycol 400. We have measured up to 54% stimulation of Pta WT activity by pyruvate in the presence of 5% polyethylene glycol 400, relative to the condition when polyethylene glycol 400 is present but pyruvate is not (data not shown). The nature of the effect is not known and may be a consequence of molecular crowding. Alternatively, the effect could be the result of reducing water activity through dehydration of the reaction components.
NADH Effect-Unlike the E. coli Pta enzyme (25), we measured positive cooperativity in the forward reaction in the absence of NADH (Table 1, Hill constants). At present, we do not know what the cause of the cooperativity differences between the E. coli and the S. enterica Pta enzyme could be.

The S. enterica Pta Protein Is Likely to Have an Elongated Shape
The large discrepancy between gel filtration data and data obtained from equilibrium sedimentation experiments suggests a somewhat nonglobular shape for the S. enterica Pta protein. This is not surprising, since characterization of the E. coli enzyme also showed a large difference in estimated mass (64). Although the proportions of aggregate and hexamer were significantly different in the variants with respect to the WT protein, each His 6 -Pta protein eluted from the column with a retention time consistent with a hexamer, suggesting that the altered kinetic properties of the variants is not a result of a decrease in oligomeric state. Further insights into the shape and oligomeric state of the Pta II enzyme will be obtained when a crystal structure of the protein becomes available.

Pta I Versus Pta II
The environmental or physiological conditions that dictate whether a prokaryote employs a class I or class II Pta enzyme remain unclear. Perhaps in some habitats the use of nonregulated Pta I enzyme is favored over the allosterically controlled Pta II enzyme, because the concentration of acetate in the environment is low. This idea is supported by reported kinetic analyses of Pta I enzymes, whose substrate affinity constants are at least 1 order of magnitude lower than Pta II enzymes and are also faster than Pta II enzymes (26,65).
Whether a prokaryote uses a Pta I (not allosterically regulated) or a Pta II (allosterically regulated) class enzyme may hinge on the existence of an alternative mechanism for modulating the level of Pta activity (e.g. pta gene expression control). The alluded pta expression control system would use pyruvate as the metabolite that signals the presence of high levels of glucose in the environment.
The Gram-positive bacterium B. subtilis is one example where this level of control of Pta activity does occur. A recent report by Sonenshein and co-workers (66) showed that the global regulatory protein CodY activates ackA ϩ transcription. In this paper, the authors allude to unpublished results of transcription microarray analyses that suggest that CodY also activates pta ϩ gene expression. Three facts are relevant to this discussion. First, in B. subtilis, the pta ϩ gene encodes a Pta I class enzyme not subject to allosteric control. Second, branched-chain amino acid biosynthesis serves as an overflow pathway for pyruvate (67) (i.e. the higher the intracellular level of pyruvate, the more branched-chain amino acids are synthesized), and third, high branched-chain amino acid levels stimulate CodY activity (68), which in turn activates pta ϩ gene expression, increasing the level of Pta I activity in the cell.

Impact of Allosteric Regulation of Pta Activity on S. enterica Physiology
The analysis of Pta II activity reported here reveals the expanded capability of this central metabolic enzyme of prokaryotes. Allosteric control of Pta activity allows the cell to rapidly adjust to changing concentrations of critical metabolites that report the energy state of the cell and the quality of carbon source available in the environment.
In Fig. 7, we depict the role of the Pta protein in S. enterica with our findings incorporated into the model. Pta can generate Ac-CoA for biosynthesis or act in concert with acetate kinase to convert Ac-CoA to Ac-P and acetate with concomitant generation of ATP (21).
Conditions that lead to the accumulation of high levels of reducing power (i.e. high NADH) would inhibit Pta activity, resulting in acetate excretion via PoxB or an as yet unidentified mechanism (69,70). This inhibition would persist until NADH is oxidized and the inhibition is lifted.
Pyruvate may serve as a nutritional signal that carbon is abundant (e.g. excess glucose in the environment), accelerating the Ac-CoA-forming reaction to increase amino acid biosynthesis or the Ac-P-forming reaction to conserve energy while excreting excess carbon as acetate (27). Increasing flux toward Ac-P could also help modulate gene expression by increasing the intracellular level of Ac-P.
Stimulation of activity by pyruvate appears to be the dominant effect, since NADH inhibition can be countered as levels of pyruvate rise intracellularly. Since the allosteric effects affect the forward (Ac-CoA-forming) reaction (34) and the back (Ac-P-forming) reaction (this work), we suggest that that Pta II enzymes sense changes in the carbon and energy status of the cell, triggering physiological changes by altering the level of Ac-P in the cell (23,30,31,71).
Although more efficient variants of Pta II can be obtained by single-amino acid changes in the N-terminal regulatory domain, such changes can lead to somewhat less stable proteins.