JOURNAL OF BACTERIOLOGY, Apr. 1977, p. 429-440 Copyright C 1977 American Society for Microbiology
Vol. 130, No. 1 Printed in U.S.A.
Escherichia coli Mutants Deficient in the Aspartate and Aromatic Amino Acid Aminotransferases DAVID H. GELFAND1 *
ROBERT A. STEINBERG2
AND
From the Laboratory of Gordon M. Tomkins, Department of Biochemistry and Biophysics, University of California, San Francisco, California 94143
Received for publication 20 September 1976
Two new mutations are described which, together, eliminate essentially all the aminotransferase activity required for de novo biosynthesis of tyrosine, phenylalanine, and aspartic acid in a K-12 strain of Escherichia coli. One mutation, designated tyrB, lies at about 80 min on the E. coli map and inactivates the "tyrosine-repressible" tyrosine/phenylalanine aminotransferase. The second mutation, aspC, maps at about 20 min and inactivates a nonrepressible aspartate aminotransferase that also has activity on the aromatic amino acids. In ilvE- strains, which lack the branched-chain amino acid aminotransferase, the presence of either the tyrosine-repressible aminotransferase or the aspartate aminotransferase is sufficient for growth in the absence of exogenous tyrosine, phenylalanine, or aspartate; the tyrosine-repressible enzyme is also active in leucine biosynthesis. The ilvE gene product alone can reverse a phenylalanine requirement. Biochemical studies on extracts of strains carrying combinations of these aminotransferase mutations confirm the existence of two distinct enzymes with overlapping specificities for the a-keto acid analogues of tyrosine, phenylalanine, and aspartate. These enzymes can be distinguished by electrophoretic mobilities, by kinetic parameters using various substrates, and by a difference in tyrosine repressibility. In extracts of an ilvE- tyrB- aspC- triple mutant, no aminotransferase activity for the a-keto acids of tyrosine, phenylalanine, or aspartate could be detected.
The terminal reaction in the biosynthesis of many amino acids, including tyrosine and phenylalanine in Escherichia coli, involves the transfer of an amino group from glutamic acid (or another suitable donor) to the appropriate a-keto acid precursor. Although mutations in the ilvE gene define a single aminotransferase with primary responsibility for synthesis of the branched-chain amino acids (2, 27, 30), there has been some confusion over the total number and specificities of enzymes involved in aromatic amino acid synthesis. After isolation of a crude "transaminase A" fraction enriched for aminotransferase activity on the aromatic amino acids and aspartate (27), there have been reports of distinguishable tyrosine and phenylalanine aminotransferases (28) and of separable aromatic amino acid and aspartate aminotransferases (6). These have led at least one group of investigators to conclude that at least three enzymes must comprise "transaminase A." The present studies demonstrate that in an ilvE- strain of E. coli the aspartate and aroI Present address: Cetus Corporation, Berkeley, CA 94710. 2 Present address: Department of Microbiology, University of California, San Francisco, CA 94143.
matic amino acid aminotransferases reside in two enzymes with distinguishable, but overlapping, activities that can be selectively elimi-
nated by mutation. MATERIALS AND METHODS Bacterial strains. The strains used in this study are all derivatives of E. coli K-12. Their designations and characteristics are listed in Table 1. Media and culture methods. Minimal medium was medium E of Vogel and Bonner (31) supplemented with 100 ,tg each of asparagine, glutamic and 50 to 100 ,ug of required amino acids per ml. Minimal liquid cultures of strains carrying the tyrB- and aspC- mutations were further supplemented with 100 u each of asparagine, glutamic acid, and glutamine per ml; 0.25% succinate; 0.25% malate; and 0.1% a-ketoglutarate. These additions increased the growth rates of these strains. L broth contained 1% tryptone (Difco), 0.5% yeast extract (Difco), 0.2% glucose, and 0.5% NaCl at pH 7.3. Broth cultures of tyrB- aspC- mutants were supplemented with 50 ,ug of tyrosine and phenylalanine per ml. Plates were solidified with 1.5% agar (Difco). Liquid cultures were grown at 37°C in a New Brunswick gyratory shaker, and densities were monitored by observing the increase in absorbance at 660 nm. Overnight cultures were prepared from 429
430
J. BACTERIOL.
GELFAND AND STEINBERG TABLE 1. List of bacterials trains and their characteristics
Strain JC7623
Source
Genotype
F- thr-1 leu-6 thi-1 proA2 argE3 hisG4 lacYl galK2 ara-14 xyl-5 mtl-l str-31 tsx-33 supE44 recB21 recC22 sbcB15 XHfr Hayes (P01) thi gal hsdS (Xdgal) HB156 thi ilvEl2 CU2 KLF42/KL253 tyrA2 thi-1 pyrD34 his-68 trp-45 recAl galK35 str-118 xr X-/F142 (tyrA+) Hfr(PO2A) met gltC8 tamA rel-1 (X+) YM811 Hfr(P044) sbcB15 JC9223 Hfr(P03) ilv metBl rel-1 JC8563 F- thi metE46 proA2 trp-3 his-4 galK2 mtl-l AB1976 ara-9 lacYl or Z4 malAl Tlr T6r Smr Ar xHfr(P012) thr-1 leu-6 thi-1 lacZ4 str-8 XAB312 As in JC7623, hsdS KH31
DG14 DG15 DG19 DG21 DG27
As in KH31, tyrAl6 As in DG14, hppT29 As in DG15, ilvE12 As in DG19, tyrB507 As in DG21, tyrA+ F-
As in DG27, aspC13 DG30 As in DG19, tyrA+ DG31 As in DG30, tyrB+ DG34 As in DG30, ilvE+ DG44 As in DG30, ilvE+ metE46 DG45 As in DG30, ampA7 DG48 As in JC8563, metB+ DG51 a NTG, Nitrosoguanidine mutagenesis.
single colonies in minimal media, diluted 1:100, and grown to mid-log phase before use. For enzyme assays and gel electrophoresis, cultures were harvested at 3 x 108 to 4 x 108 cells/ml, chilled, centrifuged, and washed with 0.9% NaCl; cell pellets were stored frozen at -20°C. Chemicals. Chemicals were obtained commercially and used without further purification. Glutamate dehydrogenase, malate dehydrogenase, phenylpyruvate (PP), oxaloacetic acid (OAA), phenazine methosulfate, and nitroblue tetrazolium were purchased from Sigma Chemical Co. p-Hydroxyphenylpyruvate (pHPP) was obtained from Calbiochem. Reagents for polyacrylamide gel electrophoresis were purchased from Bio-Rad Laboratories. Phage stocks and transductions. R-17 was prepared on strain AB312 and T7+ was prepared on E. coli B, using standard techniques (25, 29). Plbt (14) stocks were grown on plates, using a multiplicity of infection of 0.01 as described by Helling (15). Transductions were performed in L broth using mid-logphase recipient cells and a multiplicity of infection of 0.1 to 0.3 (15). Mutagenesis. Bacteria were mutagenized to 10 to 50% survival with nitrosoguanidine by the procedure of Adelberg et al. (1). After mutagenesis, the cells were washed and grown overnight in minimal media permissive for the desired mutant. Efficiency of mutagenesis was ascertained by determining the frequency of valine resistance (1), T7 resistance (9), or reversion to his+ or arg+. /3-
A. J. Clark H. Boyer H. Umbarger via S. Artz B. Ames Y. A. A. A.
Halpern via H. Umbarger J. Clark J. Clark J. Clark
A. J. Clark Exconjugant of HB156 and JC7623 Thr+ Leu+ Smr (X-) NTGa from KH31 Spontaneous low-level pHPP+ from DG14 ilvEl? transduced from CU2 into DG15 NTG from DG19 (pHPP- Tyr-) KLF42/KL253 x DG21 (pHPP+ Tyr+ F- recombinant) NTG from DG27 tyrA+ transduced from CU2 into DG19 tyrB+ transduced from DG19 into DG30 ilvE+ transduced from YM811 into DG30 metE46 transduced from AB1976 into DG30 Spontaneous Apr from DG30 metB+ transduced from CU2 into JC8563
Ampicillin selection. Overnight cultures in minimal media were diluted to 107 cells/ml and grown to about 1 x 108 to 2 x 108 cells/ml. Cells were subjected to ampicillin selection (25), washed twice with 0.9% NaCl, and either plated on supplemented minimal plates or transferred into fresh medium for overnight growth and a second cycle of ampicillin enrichment. A single cycle gave 103- to 104-fold enrichment, whereas two cycles gave an overall enrichment of about 104- to 105-fold. Colonies were replicated onto selective plates with sterile toothpicks; candidate mutants were streaked out two times, and clones were grown overnight in supplemented minimal media to check spontaneous revertant accumulation. Episome transfer and Hfr matings. Matings with F' or Hfr donors were performed in broth suspension at 37°C, using mid-log-phase cells at a donor-torecipient ratio of 0.1 to 0.2. F- recombinants from F' matings arose spontaneously during growth in selective media and were identified by their sensitivity to T7 phage and resistance to R17. Aminotransferase assays. Extracts were prepared by thawing cell pellets in buffer containing 25 mM NaKPO4, pH 7.0, 0.1 mM ethylenediaminetetraacetate, 0.2 mM pyridoxal phosphate, 0.2 mM dithiothreitol, and 5% (vol/vol) glycerol; sonically treating for 30 s (in three 10-s bursts); and centrifuging for 30 min at 30,000 x g. Suitable dilutions of the supernatant fractions were assayed. Tyrosine aminotransferase assays measuring the
VOL. 130, 1977
AMINOTRANSFERASE MUTANTS OF E. COLI
conversion of tyrosine to pHPP were performed by using a modification (13) of the procedure of Diamondstone (10). Assays in the reverse direction used glutamate dehydrogenase to reconvert the a-ketoglutarate generated in the reaction to glutamic acid, with the concomitant oxidation of nicotinamide adenine dinucleotide phosphate, reduced form (NADPH). The course of oxidation of NADPH was monitored by the change in absorbancy at 340 nm, using a Gilford recording spectrophotometer. Assays were performed at 26.5 to 27.5°C and contained, in 1 ml: 50 ,umol of tris(hydroxymethyl)aminomethane (Tris)-hydrochloride, pH 8.0; 50 umol of NH4Cl; 0.2 mg of NADPH; 0.66 mg of NaN3; 0.25 ,umol of pyridoxal phosphate; 10 j,mol of glutamic acid; 1 to 2 U of glutamate dehydrogenase; and a-keto acid substrates as indicated in figures and tables. Initial velocities (V) were determined from the steepest portions of the reaction curves and converted to micromoles of NADPH oxidized per minute, using a molar extinction coefficient of 6,200. For isoleucine aminotransferase assays, a-keto-,-methylvalerate was used. Extract concentrations were chosen which gave linear initial reactions at both high and low substrate concentrations. Oxaloacetate solutions were freshly prepared for use, and their concentrations were determined by using malate dehydrogenase (24). Stock solutions of pHPP and PP were stored frozen at -20°C, and their concentrations were determined by absorbancies at 331 and 320 nm, respectively, after treatment with alkali (8, 13). Polyacrylamide gel electrophoresis and aminotransferase activity staining. Cell pellets were thawed in gel sample buffer (25 mM Tris-PO4, pH 6.9; 0.2 mM pyridoxal phosphate; 0.5 mM dithiothreitol; 0.2 mM ethylenediaminetetraacetate; 10% [vol/ vol] glycerol), sonically treated, and centrifuged as above. Extract protein concentrations were adjusted to 2 to 3 mg/ml, and 35 to 45 ,ig was loaded in wells of a polyacrylamide slab gel (8.5% acrylamide; 0.227% bisacrylamide; 0.375 M Tris-hydrochloride, pH 8.3; 0.05% ammonim persulfate; 0.033% TEMED [N,N,N',N'-tetramethylethylenediamine]) that had been prerun for 1 h at 45 mA in 0.375 M Trishydrochloride, pH 8.3, to remove the persulfate. Electrophoresis was for about 3 h at 275 V at 4°C in 1 mM Tris-76.7 mM glycine, pH 8.3. After electrophoresis, the gel was washed briefly with 0.1 M KPO4, pH 7.5, and stained in the dark for 45 to 60 min at 37°C. The staining mixture contained 12.5 mM aketoglutarate; 0.2 mM pyridoxal phosphate; 0.6 mM nitroblue tetrazolium; 0.098 mM phenazine methosulfate; 0.1 M KPO4, pH 7.5; and 30 mM phenylalanine or 5.5 mM tyrosine. Both pHPP and PP were effective in spontaneous dye reduction (R. A. Steinberg, unpublished observation), but staining for isoleucine or aspartate aminotransferase activities required the addition of NAD+ and glutamate dehydrogenase. Control gels were stained without ketoglutarate or without amino acid substrate to confirm the identities of stained bands as aminotransferases. Protein determinations. Protein concentrations were determined by the method of Lowry et al. (18), using Pentax crystallized bovine serum albumin as a standard. a-
431
RESULTS Strategy for mutant isolation. The studies reported in this publication were initiated in an effort to obtain a tyrosine aminotransferasenegative strain of E. coli suitable for interspecific gene transfer experiments. The strategy adopted for isolation of the aminotransferase mutant was to first construct a tyrosine auxotroph that could substitute pHPP for its tyrosine requirement, and then to select cells that could no longer utilize this a-keto acid precursor of tyrosine. Strain KH31 (see Table 1) was chosen as a parental strain, since it carried a large number of markers to facilitate mutant mapping and since its recB, recC, sbcB, and hsdS mutations rendered it amenable to eventual deoxyribonucleic acid transformation studies (5, 7; R. Helling, personal communication). Selection of a pHPP-dependent mutant. Strain DG14, a tyrA- mutant of strain KH31, was selected as a low-reverting simple tyrosine auxotroph after nitrosoguanidine mutagenesis. Consistent with previous findings on tyrosine auxotrophs (27), strain DG14 had a normal (repressed) level of tyrosine aminotransferase activity (Table 2) but grew very poorly on pHPP. Whereas 20 to 25 ,ug of tyrosine per ml was sufficient for optimal growth (generation time, -1 h) of strain DG14 in liquid culture, 100 ,ug of pHPP per ml allowed only slow growth (generation time, -7 h), and 20 ,.tg of pHPP per ml was insufficient for any detectable growth. To obtain a tyrA - derivative that could utilize pHPP effliciently, plates prespread with 1 mg of pHPP were seeded with 107 to 108 DG14 cells that had been washed thoroughly after overnight growth in minimal medium containing 50 jug of tyrosine per ml. The plates were TABLE 2. Specific activity of tyrosine aminotransferasea and isoleucine aminotransferase in several strains
Strain
Supplementb
Sp act (nmol/min per mg of extract protein) Isoleucine Tyrosine aminotrans- aminotransferase
350 CU2 335 JC7623 320 KH31 + 103 KH31 100 + DG14 131 + DG19 87 DG21 + a Tyrosine aminotransferase activities termined by the Diamondstone reaction. bWhere indicated, strains were grown supplemented with tyrosine at 20 ,ug/ml. e Not determined.
ferase
Vol. 130, No. 1 Printed in U.S.A.
Escherichia coli Mutants Deficient in the Aspartate and Aromatic Amino Acid Aminotransferases DAVID H. GELFAND1 *
ROBERT A. STEINBERG2
AND
From the Laboratory of Gordon M. Tomkins, Department of Biochemistry and Biophysics, University of California, San Francisco, California 94143
Received for publication 20 September 1976
Two new mutations are described which, together, eliminate essentially all the aminotransferase activity required for de novo biosynthesis of tyrosine, phenylalanine, and aspartic acid in a K-12 strain of Escherichia coli. One mutation, designated tyrB, lies at about 80 min on the E. coli map and inactivates the "tyrosine-repressible" tyrosine/phenylalanine aminotransferase. The second mutation, aspC, maps at about 20 min and inactivates a nonrepressible aspartate aminotransferase that also has activity on the aromatic amino acids. In ilvE- strains, which lack the branched-chain amino acid aminotransferase, the presence of either the tyrosine-repressible aminotransferase or the aspartate aminotransferase is sufficient for growth in the absence of exogenous tyrosine, phenylalanine, or aspartate; the tyrosine-repressible enzyme is also active in leucine biosynthesis. The ilvE gene product alone can reverse a phenylalanine requirement. Biochemical studies on extracts of strains carrying combinations of these aminotransferase mutations confirm the existence of two distinct enzymes with overlapping specificities for the a-keto acid analogues of tyrosine, phenylalanine, and aspartate. These enzymes can be distinguished by electrophoretic mobilities, by kinetic parameters using various substrates, and by a difference in tyrosine repressibility. In extracts of an ilvE- tyrB- aspC- triple mutant, no aminotransferase activity for the a-keto acids of tyrosine, phenylalanine, or aspartate could be detected.
The terminal reaction in the biosynthesis of many amino acids, including tyrosine and phenylalanine in Escherichia coli, involves the transfer of an amino group from glutamic acid (or another suitable donor) to the appropriate a-keto acid precursor. Although mutations in the ilvE gene define a single aminotransferase with primary responsibility for synthesis of the branched-chain amino acids (2, 27, 30), there has been some confusion over the total number and specificities of enzymes involved in aromatic amino acid synthesis. After isolation of a crude "transaminase A" fraction enriched for aminotransferase activity on the aromatic amino acids and aspartate (27), there have been reports of distinguishable tyrosine and phenylalanine aminotransferases (28) and of separable aromatic amino acid and aspartate aminotransferases (6). These have led at least one group of investigators to conclude that at least three enzymes must comprise "transaminase A." The present studies demonstrate that in an ilvE- strain of E. coli the aspartate and aroI Present address: Cetus Corporation, Berkeley, CA 94710. 2 Present address: Department of Microbiology, University of California, San Francisco, CA 94143.
matic amino acid aminotransferases reside in two enzymes with distinguishable, but overlapping, activities that can be selectively elimi-
nated by mutation. MATERIALS AND METHODS Bacterial strains. The strains used in this study are all derivatives of E. coli K-12. Their designations and characteristics are listed in Table 1. Media and culture methods. Minimal medium was medium E of Vogel and Bonner (31) supplemented with 100 ,tg each of asparagine, glutamic and 50 to 100 ,ug of required amino acids per ml. Minimal liquid cultures of strains carrying the tyrB- and aspC- mutations were further supplemented with 100 u each of asparagine, glutamic acid, and glutamine per ml; 0.25% succinate; 0.25% malate; and 0.1% a-ketoglutarate. These additions increased the growth rates of these strains. L broth contained 1% tryptone (Difco), 0.5% yeast extract (Difco), 0.2% glucose, and 0.5% NaCl at pH 7.3. Broth cultures of tyrB- aspC- mutants were supplemented with 50 ,ug of tyrosine and phenylalanine per ml. Plates were solidified with 1.5% agar (Difco). Liquid cultures were grown at 37°C in a New Brunswick gyratory shaker, and densities were monitored by observing the increase in absorbance at 660 nm. Overnight cultures were prepared from 429
430
J. BACTERIOL.
GELFAND AND STEINBERG TABLE 1. List of bacterials trains and their characteristics
Strain JC7623
Source
Genotype
F- thr-1 leu-6 thi-1 proA2 argE3 hisG4 lacYl galK2 ara-14 xyl-5 mtl-l str-31 tsx-33 supE44 recB21 recC22 sbcB15 XHfr Hayes (P01) thi gal hsdS (Xdgal) HB156 thi ilvEl2 CU2 KLF42/KL253 tyrA2 thi-1 pyrD34 his-68 trp-45 recAl galK35 str-118 xr X-/F142 (tyrA+) Hfr(PO2A) met gltC8 tamA rel-1 (X+) YM811 Hfr(P044) sbcB15 JC9223 Hfr(P03) ilv metBl rel-1 JC8563 F- thi metE46 proA2 trp-3 his-4 galK2 mtl-l AB1976 ara-9 lacYl or Z4 malAl Tlr T6r Smr Ar xHfr(P012) thr-1 leu-6 thi-1 lacZ4 str-8 XAB312 As in JC7623, hsdS KH31
DG14 DG15 DG19 DG21 DG27
As in KH31, tyrAl6 As in DG14, hppT29 As in DG15, ilvE12 As in DG19, tyrB507 As in DG21, tyrA+ F-
As in DG27, aspC13 DG30 As in DG19, tyrA+ DG31 As in DG30, tyrB+ DG34 As in DG30, ilvE+ DG44 As in DG30, ilvE+ metE46 DG45 As in DG30, ampA7 DG48 As in JC8563, metB+ DG51 a NTG, Nitrosoguanidine mutagenesis.
single colonies in minimal media, diluted 1:100, and grown to mid-log phase before use. For enzyme assays and gel electrophoresis, cultures were harvested at 3 x 108 to 4 x 108 cells/ml, chilled, centrifuged, and washed with 0.9% NaCl; cell pellets were stored frozen at -20°C. Chemicals. Chemicals were obtained commercially and used without further purification. Glutamate dehydrogenase, malate dehydrogenase, phenylpyruvate (PP), oxaloacetic acid (OAA), phenazine methosulfate, and nitroblue tetrazolium were purchased from Sigma Chemical Co. p-Hydroxyphenylpyruvate (pHPP) was obtained from Calbiochem. Reagents for polyacrylamide gel electrophoresis were purchased from Bio-Rad Laboratories. Phage stocks and transductions. R-17 was prepared on strain AB312 and T7+ was prepared on E. coli B, using standard techniques (25, 29). Plbt (14) stocks were grown on plates, using a multiplicity of infection of 0.01 as described by Helling (15). Transductions were performed in L broth using mid-logphase recipient cells and a multiplicity of infection of 0.1 to 0.3 (15). Mutagenesis. Bacteria were mutagenized to 10 to 50% survival with nitrosoguanidine by the procedure of Adelberg et al. (1). After mutagenesis, the cells were washed and grown overnight in minimal media permissive for the desired mutant. Efficiency of mutagenesis was ascertained by determining the frequency of valine resistance (1), T7 resistance (9), or reversion to his+ or arg+. /3-
A. J. Clark H. Boyer H. Umbarger via S. Artz B. Ames Y. A. A. A.
Halpern via H. Umbarger J. Clark J. Clark J. Clark
A. J. Clark Exconjugant of HB156 and JC7623 Thr+ Leu+ Smr (X-) NTGa from KH31 Spontaneous low-level pHPP+ from DG14 ilvEl? transduced from CU2 into DG15 NTG from DG19 (pHPP- Tyr-) KLF42/KL253 x DG21 (pHPP+ Tyr+ F- recombinant) NTG from DG27 tyrA+ transduced from CU2 into DG19 tyrB+ transduced from DG19 into DG30 ilvE+ transduced from YM811 into DG30 metE46 transduced from AB1976 into DG30 Spontaneous Apr from DG30 metB+ transduced from CU2 into JC8563
Ampicillin selection. Overnight cultures in minimal media were diluted to 107 cells/ml and grown to about 1 x 108 to 2 x 108 cells/ml. Cells were subjected to ampicillin selection (25), washed twice with 0.9% NaCl, and either plated on supplemented minimal plates or transferred into fresh medium for overnight growth and a second cycle of ampicillin enrichment. A single cycle gave 103- to 104-fold enrichment, whereas two cycles gave an overall enrichment of about 104- to 105-fold. Colonies were replicated onto selective plates with sterile toothpicks; candidate mutants were streaked out two times, and clones were grown overnight in supplemented minimal media to check spontaneous revertant accumulation. Episome transfer and Hfr matings. Matings with F' or Hfr donors were performed in broth suspension at 37°C, using mid-log-phase cells at a donor-torecipient ratio of 0.1 to 0.2. F- recombinants from F' matings arose spontaneously during growth in selective media and were identified by their sensitivity to T7 phage and resistance to R17. Aminotransferase assays. Extracts were prepared by thawing cell pellets in buffer containing 25 mM NaKPO4, pH 7.0, 0.1 mM ethylenediaminetetraacetate, 0.2 mM pyridoxal phosphate, 0.2 mM dithiothreitol, and 5% (vol/vol) glycerol; sonically treating for 30 s (in three 10-s bursts); and centrifuging for 30 min at 30,000 x g. Suitable dilutions of the supernatant fractions were assayed. Tyrosine aminotransferase assays measuring the
VOL. 130, 1977
AMINOTRANSFERASE MUTANTS OF E. COLI
conversion of tyrosine to pHPP were performed by using a modification (13) of the procedure of Diamondstone (10). Assays in the reverse direction used glutamate dehydrogenase to reconvert the a-ketoglutarate generated in the reaction to glutamic acid, with the concomitant oxidation of nicotinamide adenine dinucleotide phosphate, reduced form (NADPH). The course of oxidation of NADPH was monitored by the change in absorbancy at 340 nm, using a Gilford recording spectrophotometer. Assays were performed at 26.5 to 27.5°C and contained, in 1 ml: 50 ,umol of tris(hydroxymethyl)aminomethane (Tris)-hydrochloride, pH 8.0; 50 umol of NH4Cl; 0.2 mg of NADPH; 0.66 mg of NaN3; 0.25 ,umol of pyridoxal phosphate; 10 j,mol of glutamic acid; 1 to 2 U of glutamate dehydrogenase; and a-keto acid substrates as indicated in figures and tables. Initial velocities (V) were determined from the steepest portions of the reaction curves and converted to micromoles of NADPH oxidized per minute, using a molar extinction coefficient of 6,200. For isoleucine aminotransferase assays, a-keto-,-methylvalerate was used. Extract concentrations were chosen which gave linear initial reactions at both high and low substrate concentrations. Oxaloacetate solutions were freshly prepared for use, and their concentrations were determined by using malate dehydrogenase (24). Stock solutions of pHPP and PP were stored frozen at -20°C, and their concentrations were determined by absorbancies at 331 and 320 nm, respectively, after treatment with alkali (8, 13). Polyacrylamide gel electrophoresis and aminotransferase activity staining. Cell pellets were thawed in gel sample buffer (25 mM Tris-PO4, pH 6.9; 0.2 mM pyridoxal phosphate; 0.5 mM dithiothreitol; 0.2 mM ethylenediaminetetraacetate; 10% [vol/ vol] glycerol), sonically treated, and centrifuged as above. Extract protein concentrations were adjusted to 2 to 3 mg/ml, and 35 to 45 ,ig was loaded in wells of a polyacrylamide slab gel (8.5% acrylamide; 0.227% bisacrylamide; 0.375 M Tris-hydrochloride, pH 8.3; 0.05% ammonim persulfate; 0.033% TEMED [N,N,N',N'-tetramethylethylenediamine]) that had been prerun for 1 h at 45 mA in 0.375 M Trishydrochloride, pH 8.3, to remove the persulfate. Electrophoresis was for about 3 h at 275 V at 4°C in 1 mM Tris-76.7 mM glycine, pH 8.3. After electrophoresis, the gel was washed briefly with 0.1 M KPO4, pH 7.5, and stained in the dark for 45 to 60 min at 37°C. The staining mixture contained 12.5 mM aketoglutarate; 0.2 mM pyridoxal phosphate; 0.6 mM nitroblue tetrazolium; 0.098 mM phenazine methosulfate; 0.1 M KPO4, pH 7.5; and 30 mM phenylalanine or 5.5 mM tyrosine. Both pHPP and PP were effective in spontaneous dye reduction (R. A. Steinberg, unpublished observation), but staining for isoleucine or aspartate aminotransferase activities required the addition of NAD+ and glutamate dehydrogenase. Control gels were stained without ketoglutarate or without amino acid substrate to confirm the identities of stained bands as aminotransferases. Protein determinations. Protein concentrations were determined by the method of Lowry et al. (18), using Pentax crystallized bovine serum albumin as a standard. a-
431
RESULTS Strategy for mutant isolation. The studies reported in this publication were initiated in an effort to obtain a tyrosine aminotransferasenegative strain of E. coli suitable for interspecific gene transfer experiments. The strategy adopted for isolation of the aminotransferase mutant was to first construct a tyrosine auxotroph that could substitute pHPP for its tyrosine requirement, and then to select cells that could no longer utilize this a-keto acid precursor of tyrosine. Strain KH31 (see Table 1) was chosen as a parental strain, since it carried a large number of markers to facilitate mutant mapping and since its recB, recC, sbcB, and hsdS mutations rendered it amenable to eventual deoxyribonucleic acid transformation studies (5, 7; R. Helling, personal communication). Selection of a pHPP-dependent mutant. Strain DG14, a tyrA- mutant of strain KH31, was selected as a low-reverting simple tyrosine auxotroph after nitrosoguanidine mutagenesis. Consistent with previous findings on tyrosine auxotrophs (27), strain DG14 had a normal (repressed) level of tyrosine aminotransferase activity (Table 2) but grew very poorly on pHPP. Whereas 20 to 25 ,ug of tyrosine per ml was sufficient for optimal growth (generation time, -1 h) of strain DG14 in liquid culture, 100 ,ug of pHPP per ml allowed only slow growth (generation time, -7 h), and 20 ,.tg of pHPP per ml was insufficient for any detectable growth. To obtain a tyrA - derivative that could utilize pHPP effliciently, plates prespread with 1 mg of pHPP were seeded with 107 to 108 DG14 cells that had been washed thoroughly after overnight growth in minimal medium containing 50 jug of tyrosine per ml. The plates were TABLE 2. Specific activity of tyrosine aminotransferasea and isoleucine aminotransferase in several strains
Strain
Supplementb
Sp act (nmol/min per mg of extract protein) Isoleucine Tyrosine aminotrans- aminotransferase
350 CU2 335 JC7623 320 KH31 + 103 KH31 100 + DG14 131 + DG19 87 DG21 + a Tyrosine aminotransferase activities termined by the Diamondstone reaction. bWhere indicated, strains were grown supplemented with tyrosine at 20 ,ug/ml. e Not determined.
ferase
