Proc. Nat. Acad. Sci. USA Vol. 72, No. 4, pp. 1364-1367, April 1975
Metabolites Influence Control of Lysine Transfer Ribonucleic Acid Synthetase Formation in Escherichia coli K-12 (E. coli mutant/alanine, leucine peptides/in vivo tRNA charging/antibody neutralization assay)
IRVIN N. HIRSHFIELD, FU-MEEI YEH, AND LAURA E. SAWYER The John Collins Laboratories of the Huntington Memorial Hospital of Harvard University at the Massachusetts General Hospital, Boston, Mass. 02114
Communicated by Paul C. Zamecnik, January 27, 1976
Media. Cells were grown in minimal medium (5) or in medium supplemented with amino acids, vitamins, purines, and pyrimidines (6) as described previously (3). Mutagenesis and Selection of Mutant. Cells were mutagenized with nitrosoguanidine according to CerdA-Olmedo et al. (7), except that thymine starvation was not utilized. Mutants were selected by their resistance to the lysine analog thiosine, 50 ,g/ml (1). Preparation of Enzyme Extracts and Assay of Lysyl-tRNA Synthetase. These were done as previously described (1). Protein Determination. Protein was determined by the method of Lowry et al. (8), with crystalline bovine serum albumin (Sigma Chemical Co.) as a standard. Determination of the In Vivo Charging of Lysine tRNA. The in vivo charging of lysine tRNA was measured according to the procedure of Folk and Berg (9) as modified by Lewis and Ames (10). Antibody Preparation. Lysyl-tRNA synthetase from wildtype grown in the presence of 20 mM L-alanine was purified by the procedure of Waldenstrom (11). This gave an enzyme which was approximately 90% pure (Hirshfield and Yeh, unpublished results). This enzyme preparation was purified to homogeneity by preparative gel electrophoresis (Hirshfield and Yeh, unpublished results.) Each of two male rabbits was inoculated via the footpads with 50 jzg of the electrophoretically purified enzyme. A booster of 50 ,g of the same protein was administered 4 weeks later. Three weeks later a second booster of 300 MAg of the 90% purified protein was administered, and the serum from the rabbits was collected by cardiac puncture over a three-day period beginning 8 days later. The globulin fraction of the serum was concentrated by ammonium sulfate precipitation (12). Immunological Procedures. Two-dimensional immunodiffusion gels were run according to the procedure of Ouchterlony (13). The antibody neutralization assays followed the procedure of Pollock (14). The antibody neutralization assays were performed in two ways. In the first procedure crude extracts of the wild-type or mutant strain (15-20 ug of protein) were incubated in pH 7.25 Tris-maleate buffer (0.05 M) at 250 for 30 min with various amounts of 1/20 diluted antibody. In a second method the crude extract protein was added to the preincubation mixture so that the number of units of lysyl-tRNA synthetase were equal even
A mutant of E. coli K-12 has been isolated ABSTRACT which has only 1-3% of the wild-type lysyl-tRNA synthetase activity [L-lysine:tRNA ligase (AMP forming), EC 6.1.1.61. Additions of 20 mM L-alanine or 6 mM leucine dipeptides to the culture medium can restore the activity of lysyl-tRNA synthetase in the mutant strain to the wildtype level. Experiments on the in vivo charging of lysine tRNA in the mutant show that in the absence of the metabolites lysine tRNA is charged 15-23%. Upon the addition of 3 mM L-leucyl-L-alanine to the medium the lysyl-tRNA synthetase activity increases 25-fold and the in vivo charging of lysine tRNA returns to the wild-type level. Experiments with antibody against lysyl-tRNA synthetase show that the stimulation of lysyl-tRNA synthetase activity by the metabolites is the result of new protein synthesis.
Mutants of Escherichia coli with low lysyl-tRNA synthetase [rlysine:tRNA ligase (AMP forming), EC 6.1.1.6] activity, 5-50% of wild-type, have previously been isolated in this laboratory (1). The lysyl-tRNA synthetase (LRS) from these strains appears to be altered, as it is characteristically more thermostable than the wild-type enzyme (2). The activity of LRS in these mutants can be restored to normal by including L-alanine + D-fructose, or L-leucine dipeptides in the growth medium (3, 7). These compounds can also stimulate the activity of wild-type LRS 2- to 3-fold as well as arginyl- and methionyl-tRNA synthetases (1, 3, 4). The stimulation of LRS activity by the metabolites appears to be the result of new protein synthesis (1, 3, 4). In this report experiments with a new mutant of this type are discussed. This mutant has lower activity of LRS (1-3% of wild type) than the others of this type. Studies on the in vivo charging of lysine tRNA indicate that the control of LRS synthesis in at least this strain is independent of transfer RNA but dependent upon the noncognate metabolites, alanine or leucine peptides. MATERIALS AND METHODS
Strains. The wild-type strain in these experiments is the E. coli K-12 strain AT2092 (Be-, Arg-, tIis-, Phe-, purF-(aden), strR, F-). The mutant IH2017 derived from this strain is resistant to the lysine analog thiosine in addition to the above characteristics. Reagents. Chemical compounds were of highest quality from commercial institutions. Uniformly " C-labeled L-amino acids were obtained from New England Nuclear Corp., Boston, Mass., or Schwarz/Mann, Orangeburg, N.Y. Abbreviation: LRS, lysyl-tRNA synthetase. 1364
Proc. Nat. Acad. Sci. USA 72
(1975)
though the amount of total crude extract protein differed. In both of these methods the total volume of the incubation mixture was 100 pl. After incubation of the enzyme with the antibody, an aliquot of the mixture was added to a reaction mixture for assaying lysyl-tRNA synthetase activity in order to measure the remaining enzyme activity. RESULTS The lysyl-tRNA synthetase mutant IH2017 described in this report was derived from strain AT2092 by nitrosoguanidine mutagenesis. The mutant was selected as a thiosine-resistant strain on 2% agar plates containing supplemented minimal medium with 50 ,sg/ml of thiosine (S-0-aminoethylcysteine). No lysine was present in the medium. Unlike other lysyltRNA synthetase mutants of this type which were lysine bradytrophs [leaky for lysine biosynthesis (1)], mutant IH2017 is prototrophic for lysine biosynthesis, and we have found that the lysine analog thiosine is just as effective in inhibiting growth of lysine prototrophs as bradytrophs. The selection procedure and its rationale are described in greater detail in ref. 1. Mutant IH2017 has only 1-3% of the wild-type lysyltRNA synthetase (LRS) activity in minimal medium or supplemented minimal medium, which is even lower than that of the mutants of this type previously isolated (1). The generation time of IH2017 in supplemented minimal medium with or without lysine is usually 80-100 min, compared to 35-45 min for wild-type. The LRS from strain IH2017 is apparently an altered protein, as it is more thermostable than the wild-type enzyme. This is characteristic of LRS from this type of mutant strain. Data from heat inactivation studies with the LRS from this mutant are not shown, but are quite similar to those of LRS from other mutants of this type (see ref. 2). The activity of LRS can be enhanced in the mutant by the inclusion of L-alanine or L-leucine dipeptides in the growth medium as with other mutants of this type (3, 4). The optimal concentration for L-alanine is 20 mM and for L-leucine dipeptides, 6 mM. The stimulation of LRS activity by these two metabolites is variable, but its usually in the range of 25- to 50fold; it is sometimes as high as 50- to 70-fold. With IH2017 D-fructose has no effect on LRS activity either alone or with L-alanine (3). L-Leucine also has an effect on LRS activity in the mutant but only 15-fold at a concentration of 40 mM. The stimulation of LRS activity by optimal and suboptimal concentrations of L-alanine and L-leucyl-ileucine is shown in Table 1. Optimal concentrations of each metabolite are not additive but suboptimal concentrations are. This indicates that these two metabolites either interact with two different receptors which influence a common process, or act at two different subsites on the same receptor (presumably a protein). Since mutant IH2017 has a slow growth rate and a very low level of LRS activity, it was of interest to examine the in vivo level of charged lysine tRNA. It was of particular interest to determine if the in vivo level of charged lysine tRNA was low, since in other studies a low in vivo level of charging of a tRNA has been associated with increased levels of the cognate synthetase (15, 16). The data in Table 2 demonstrate that low activity of LRS in the mutant is associated with low levels of in vivo charging of lysine tRNA (15-23%). The level of in vivo charging of lysine tRNA in the wild-type varied from 80 to 100% in our experiments. The addition of 3 mM Lleucyl-Lalanine to the growth medium had three effects on the mutant.
Metabolite Control of Lysyl-tRNA Synthetase
1365
TABLE 1. Optimal and suboptimal concentrations of L-alanine and L-leucyl-L-leucine stimulate lysyl-tRNA synthetase activity in IH2017 grown in supplemented minimal medium LysyltRNA Optimal syntheAdditionto concentra- tase, Strain medium tion, mM units/mg AT2092 8.6 IH2017 0.15 IH2017 Ala 20 8.5 IH2017 Leu-Leu 6 8.7 I112017 Ala + Leu-Leu 20 + 6 9.3
SubLysyloptimal tRNA concen- synthetration, tase, mM units/mg 9.5 0.25 10 2.6 3 7.7 10 + 3
11.4
The cells were grown to an ODs,0 of 0.3 (log phase). All additions were made prior to inoculation with the strains. The cells were grown at least four generations under all conditions. Lysyl-tRNA synthetase activity was assayed as described in ref 1.
The activity of lysyl-tRNA synthetase increased 25-fold, the in vivo charged level of lysine tRNA increased to the wild-type level, and the growth rate of the strain doubled. Thus, the slow growth of the mutant appears to be due to insufficient availability of lysine in the form of lysyl-tRNA for protein syntheS1S.
Previous work on mutants of this type has indicated that the increase in the activity of LRS elicited by metabolites is due to an increase in the amount of the enzyme (1, 3, 7). If so, this would indicate that the insufficiency of lysyl-tRNA in vivo in strain IH2017 is due to a deficiency in the amount of LRS. In order to corroborate the results of these previous experiments which used indirect methods to demonstrate changes in the amount of protein, a more direct approach, an immunological approach, has been taken. Antibody was prepared against LRS purified from wild type grown in the presence of 20 mM L-alanine. Two-dimensional immunodiffusion in agar of the anti-LRS antibody against LRS purified from wild-type grown in the presence and absence of 20 mM L-alanine, and from IH2017 grown with 20 mM L-alanine reveals a line of identity with no spurs (Fig. 1). The absence of spurs indicates that the mutant and wild-type TABLE 2. Charging of lysine tRNA in vivo upon growth in supplemented minimal medium
Strain AT2092 IH2017 IH2017 AT2092 IH2017
Generation time, min 45 120 63 40 100
%tRNA Addition to medium 3 mM Leu-Ala -
Chrig
Charging Lys Leu 98 23 79 98 15
89 100
LysyltRNA synthetase, units/mg 9.5 0.11 2.5 9.6 0.10
Cells were grown to an OD,%o of 0. 3 (log phase). Additions were made prior to inoculation to cells and the cells were grown at least four generations under all conditions. The in vivo charging was measured by the procedure of Folk and Berg (9) as modified by Lewis and Ames (10). Lysyl-tRNA synthetase activity was assayed according to ref. 1.
1366
Proc. Nat. Acad. Sci. USA 72 (1.975)
Genetics: Hirshfield et al.
FIG. 1. Two-dimensional immunodiffusion gels of purified lysyl-tRNA synthetase from wild-type cells grown with 20 mM L-alanine (A), wild type without alanine (B), and mutant IH2017 with alanine (C). The wells contain 8-10 ,ug of purified ptotein, and the center well 95 ,gg of antiserum protein.
enzymes are immunologically very similar. A second fainter precipitin line can be seen emanating from the wells containing wild-type LRS purified from medium with alanine (wells marked A). This second line lies between the wells marked A and the major precipitin band, but closer to the latter. This line most likely represents a second form of the enzyme. Recent experiments (Hirshfield, Yeh, and Zamecnik, un-
published) indicate that LRS is composed of two subunit forms, but one subunit form is present in large quantities only when wild type is grown in the presence of alanine. The antibody neutralization tests corroborate other results from this laboratory that the change in activity of LRS is due to a difference in the amount of enzyme protein present (Fig. 2). When the specific activities of wild-type and mutant LRS are similar (the mutant activity was enhanced 25- to 30-fold by alanine), the initial slopes in the neutralization assays are similar, indicating the presence of equal amounts of enzyme protein per unit of enzyme activity. As the specific activity of the mutant enzyme decreases the slope becomes flatter, which indicates that less enzyme protein is present. In a second set of experiments, neutralization assays were run on LRS from wild-type grown with and without 20 mM alanine. The activity of LRS from the former was 2.5 times higher than the latter, and the LRS from the alanine-grown cells was diluted 2.5 times so that the units of LRS in the assay were approximately equal. In Fig. 3 it can be seen that the initial slope of the neutralization curve is the same for each, demonstrating the presence of an equal amount of enzyme protein per unit of enzyme activity. Since the LRS from wild type grown with alanine was diluted 2.5-fold, it must be present in 2.5 times the amount of LRS from the control culture in the total crude extract protein. DISCUSSION Several laboratories have been working on the regulation of aminoacyl-tRNA synthetases. Originally Nass and Neidhardt (17), and later Williams and Neidhardt (81) proposed a mechanism for aminoacyl-tRNA synthetase regulation which is analogous to repression of amino-acid biosynthesis. According to their model a decrease in the amount of an amino-acid or
12 40
3.0
2.5 K,41
It4,)". '4
2.0
bc
4 A v C
1.5
i
ZZ
1.0
5
40
15
0.5
,u/ OF ioANr/SE-RUM
FIG. 2. Antibody neutralization assay of lysyl-tRNA synthetase from wild type and mutant IH2017. The wild-type strain was grown in supplemented minimal medium in the absence of alanine. The mutant st~rain was grown in the above medium in the absence or presence of various concentrations of L-alanine or L-leucylglycine. In these assays 15-20 /Ag of crude extract protein were incubated at pH 7.25 with various amounts of 1/20 diluted antiserum as described in Materials and Methods. Upon completion of the incubation the remaining lysyl-tRNA synthetase activity was assayed at pH 7.25 (see ref. 1). O. wild-type; A, mu'tant enzyme stimulated 25- to 30-fold; A, mutant enzyme stimulated 10-fold;
*, mutant enzyme,
no
stimulation.
5
10
15
20
,J/ OF ;.r2o ANTISERUM FIG. 3. Antibody neutralization assay of LRS from wild type grown with and without 20 mM L-alanine. The protein was diluted so that approximately the same number of enzyme units were present in the incubation with 1/20 diluted antiserum. The remainder of the experiment was carried out as in Fig. 2. 0, LRS from wild type without alanine, specific activity 7.5 units/mg; *, LRS from wild type with alanine, specific activity 19.6 units/mg.
Proc. Nat. Acad. Sci. USA 72
(1975)
aminoacyl-tRNA would lead to an increase in the amount of both the aminoacyl-tRNA synthetase and biosynthetic enzymes for that amino acid. This work has been extended by Williams and coworkers (e.g., 15, 19, 20), and, indeed, they have found that the restriction of the supply of a particular amino acid results in an increase in the activity of the cognate synthetase. In addition they have shown that low levels of in vivo charging of histidine tRNA in a histidyl-tRNA synthetase mutant is associated with an increase in both the level of the histidine biosynthetic enzymes, and histidyl-tRNA synthetase (15). A second mode of regulation has been proposed by Parker and Neidhardt (21) and Parker et al. (22), termed "metabolic regulation." They suggest that at least arginyl- and valyltRNA synthetase formation is regulated similarly to other macromolecular components of the bacterial cell's translational machinery. The fact that the metabolites, alanine and leucine peptides, can modulate the synthesis of LRS suggests a possible link to this model, but at this time a clear relationship has not been established. The mutant IH2017 discussed in this report appears to be a regulatory mutant for LRS. It has a very low level of LRS activity, but unlike the situation in other studies (as ref. 15) the low level of LRS activity is associated with low in vivo charged levels of lysine tRNA. The presence of lysine in the medium does not increase the in vivo charging, but inclusion of a leucine dipeptide increases both the activity of LRS and the in vivo charged level of lysine tRNA. Immunological experiments have confirmed other data (1, 3, 4) that the change in activity of LRS reflects the amount of enzyme protein present. Thus the amount of LRS in this strain is independent of the charging of lysine tRNA, but dependent upon the presence of the metabolites alanine and leucine peptides. These results, in addition to the fact that LRS from the mutant appears to be an altered protein, suggest a model which could account for the normally high level of LRS found in the cell. According to this model LRS is autoregulatory, and regulates its own synthesis by a positive control mechanism. Since the increase in LRS synthesis in these mutants can be blocked by the RNA synthesis inhibitor rifampicin (1), it would seem most likely that the control is transcriptional. Thus, it would be postulated that the mutation in LRS renders it less able to bind to DNA to regulate its own transcription, and that the
Metabolite Control of Lysyl-tRNA Synthetase
1367
metabolites alanine and leucine peptides restore this ability. Until this hypothesis can be tested it cannot be known whether the metabolites are absolutely required for normal maintenance of LRS synthesis or form part of an ancillary control system. We thank Miss Dianne Sanborn for her aid in isolating the mutants, and Ms. Barbara Wilkes, Dr. John Long, and Dr. Alan Aisenberg for their help in producing and isolating anti-lysyltRNA synthetase antibody. This investigation was supported by Grant NP-2Q of the American Cancer Society, and Grant RR05486-12 of the National Institutes of Health. This is publication No. 1474 of the Cancer Commission of Harvard University. 1. Hirshfield, I. N. & Zamecnik, P. C. (1972) Biochim. Biophys. Acta 259, 330-343. 2. Hirshfield, I. N., Tomford, J. W. & Zamecnik, P. C. (1972) Biochim. Biophys. Acta 259, 344-356. 3. Hirshfield, I. N. & Buklad, N. E. (1973) J. Bacteriol. 113, 167-177. 4. Buklad, N. E., Sanborn, D. & Hirshfield, I. N. (1973) J. Bacteriol. 116, 1477-1478. 5. Davis, B. D. & Mingioli, E. S. (1950) J. Bacteriol. 60, 17-28. 6. Novick, R. P. & Maas, W. K. (1961) J. Bacteriol. 81, 236240. 7. Cerd4-Olmedo, E., Hanawalt, P. C. & Guerola, N. (1968) J. Mol. Biol. 33, 705-719. 8. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. (1951) J. Biol. Chem. 246, 7595-7601. 9. Folk, N. B. & Berg, P. (1970) J. Bacteriol. 102, 204-212. 10. Lewis, J. A. & Ames, B. N. (1972) J. Mol. Biol. 66, 131142. 11. Waldenstrom. J. (1968) Eur. J. Biochem. 3, 483-487. 12. Kendall, F. E. (1937) J. Clin. Invest. 16, 921-931. 13. Ouchterlony, 0. (1949) Arkhiv. Kemi I, 43-48. 14. Pollock, M. R. (1956) J. Gen. Microbiol. 14, 90-108. 15. McGinnis, E. & Williams, L. S. (1972) J. Bacteriol. 111, 739744. 16. McGinnis, E., Williams, A. & Williams, L. S. (1974) J. Bacteriol. 119, 554-559. 17. Nass, G. & Neidhardt, F. C. (1967) Biochim. Biophys. Acta 134, 347-359. 18. Williams, L. S. & Neidhardt, F. C. (1967) J. Mol. Biol. 43, 529-550. 19. Archibald, E. R. & Williams, L. S. (1972) J. Bacteriol. 109, 1020-1026. 20. McGinnis, E. & Williams, L. S. (1972) J. Bacteriol. 109, 505511. 21. Parker, J. & Neidhardt, F. C. (1972) Biochem. Biophys. Res. Commun. 49, 495-501. 22. Parker, J., Flashner, M., McKeever, N. G. & Neidhardt, F. C. (1974) J. Biol. Chem. 249, 1044-1053.
Metabolites Influence Control of Lysine Transfer Ribonucleic Acid Synthetase Formation in Escherichia coli K-12 (E. coli mutant/alanine, leucine peptides/in vivo tRNA charging/antibody neutralization assay)
IRVIN N. HIRSHFIELD, FU-MEEI YEH, AND LAURA E. SAWYER The John Collins Laboratories of the Huntington Memorial Hospital of Harvard University at the Massachusetts General Hospital, Boston, Mass. 02114
Communicated by Paul C. Zamecnik, January 27, 1976
Media. Cells were grown in minimal medium (5) or in medium supplemented with amino acids, vitamins, purines, and pyrimidines (6) as described previously (3). Mutagenesis and Selection of Mutant. Cells were mutagenized with nitrosoguanidine according to CerdA-Olmedo et al. (7), except that thymine starvation was not utilized. Mutants were selected by their resistance to the lysine analog thiosine, 50 ,g/ml (1). Preparation of Enzyme Extracts and Assay of Lysyl-tRNA Synthetase. These were done as previously described (1). Protein Determination. Protein was determined by the method of Lowry et al. (8), with crystalline bovine serum albumin (Sigma Chemical Co.) as a standard. Determination of the In Vivo Charging of Lysine tRNA. The in vivo charging of lysine tRNA was measured according to the procedure of Folk and Berg (9) as modified by Lewis and Ames (10). Antibody Preparation. Lysyl-tRNA synthetase from wildtype grown in the presence of 20 mM L-alanine was purified by the procedure of Waldenstrom (11). This gave an enzyme which was approximately 90% pure (Hirshfield and Yeh, unpublished results). This enzyme preparation was purified to homogeneity by preparative gel electrophoresis (Hirshfield and Yeh, unpublished results.) Each of two male rabbits was inoculated via the footpads with 50 jzg of the electrophoretically purified enzyme. A booster of 50 ,g of the same protein was administered 4 weeks later. Three weeks later a second booster of 300 MAg of the 90% purified protein was administered, and the serum from the rabbits was collected by cardiac puncture over a three-day period beginning 8 days later. The globulin fraction of the serum was concentrated by ammonium sulfate precipitation (12). Immunological Procedures. Two-dimensional immunodiffusion gels were run according to the procedure of Ouchterlony (13). The antibody neutralization assays followed the procedure of Pollock (14). The antibody neutralization assays were performed in two ways. In the first procedure crude extracts of the wild-type or mutant strain (15-20 ug of protein) were incubated in pH 7.25 Tris-maleate buffer (0.05 M) at 250 for 30 min with various amounts of 1/20 diluted antibody. In a second method the crude extract protein was added to the preincubation mixture so that the number of units of lysyl-tRNA synthetase were equal even
A mutant of E. coli K-12 has been isolated ABSTRACT which has only 1-3% of the wild-type lysyl-tRNA synthetase activity [L-lysine:tRNA ligase (AMP forming), EC 6.1.1.61. Additions of 20 mM L-alanine or 6 mM leucine dipeptides to the culture medium can restore the activity of lysyl-tRNA synthetase in the mutant strain to the wildtype level. Experiments on the in vivo charging of lysine tRNA in the mutant show that in the absence of the metabolites lysine tRNA is charged 15-23%. Upon the addition of 3 mM L-leucyl-L-alanine to the medium the lysyl-tRNA synthetase activity increases 25-fold and the in vivo charging of lysine tRNA returns to the wild-type level. Experiments with antibody against lysyl-tRNA synthetase show that the stimulation of lysyl-tRNA synthetase activity by the metabolites is the result of new protein synthesis.
Mutants of Escherichia coli with low lysyl-tRNA synthetase [rlysine:tRNA ligase (AMP forming), EC 6.1.1.6] activity, 5-50% of wild-type, have previously been isolated in this laboratory (1). The lysyl-tRNA synthetase (LRS) from these strains appears to be altered, as it is characteristically more thermostable than the wild-type enzyme (2). The activity of LRS in these mutants can be restored to normal by including L-alanine + D-fructose, or L-leucine dipeptides in the growth medium (3, 7). These compounds can also stimulate the activity of wild-type LRS 2- to 3-fold as well as arginyl- and methionyl-tRNA synthetases (1, 3, 4). The stimulation of LRS activity by the metabolites appears to be the result of new protein synthesis (1, 3, 4). In this report experiments with a new mutant of this type are discussed. This mutant has lower activity of LRS (1-3% of wild type) than the others of this type. Studies on the in vivo charging of lysine tRNA indicate that the control of LRS synthesis in at least this strain is independent of transfer RNA but dependent upon the noncognate metabolites, alanine or leucine peptides. MATERIALS AND METHODS
Strains. The wild-type strain in these experiments is the E. coli K-12 strain AT2092 (Be-, Arg-, tIis-, Phe-, purF-(aden), strR, F-). The mutant IH2017 derived from this strain is resistant to the lysine analog thiosine in addition to the above characteristics. Reagents. Chemical compounds were of highest quality from commercial institutions. Uniformly " C-labeled L-amino acids were obtained from New England Nuclear Corp., Boston, Mass., or Schwarz/Mann, Orangeburg, N.Y. Abbreviation: LRS, lysyl-tRNA synthetase. 1364
Proc. Nat. Acad. Sci. USA 72
(1975)
though the amount of total crude extract protein differed. In both of these methods the total volume of the incubation mixture was 100 pl. After incubation of the enzyme with the antibody, an aliquot of the mixture was added to a reaction mixture for assaying lysyl-tRNA synthetase activity in order to measure the remaining enzyme activity. RESULTS The lysyl-tRNA synthetase mutant IH2017 described in this report was derived from strain AT2092 by nitrosoguanidine mutagenesis. The mutant was selected as a thiosine-resistant strain on 2% agar plates containing supplemented minimal medium with 50 ,sg/ml of thiosine (S-0-aminoethylcysteine). No lysine was present in the medium. Unlike other lysyltRNA synthetase mutants of this type which were lysine bradytrophs [leaky for lysine biosynthesis (1)], mutant IH2017 is prototrophic for lysine biosynthesis, and we have found that the lysine analog thiosine is just as effective in inhibiting growth of lysine prototrophs as bradytrophs. The selection procedure and its rationale are described in greater detail in ref. 1. Mutant IH2017 has only 1-3% of the wild-type lysyltRNA synthetase (LRS) activity in minimal medium or supplemented minimal medium, which is even lower than that of the mutants of this type previously isolated (1). The generation time of IH2017 in supplemented minimal medium with or without lysine is usually 80-100 min, compared to 35-45 min for wild-type. The LRS from strain IH2017 is apparently an altered protein, as it is more thermostable than the wild-type enzyme. This is characteristic of LRS from this type of mutant strain. Data from heat inactivation studies with the LRS from this mutant are not shown, but are quite similar to those of LRS from other mutants of this type (see ref. 2). The activity of LRS can be enhanced in the mutant by the inclusion of L-alanine or L-leucine dipeptides in the growth medium as with other mutants of this type (3, 4). The optimal concentration for L-alanine is 20 mM and for L-leucine dipeptides, 6 mM. The stimulation of LRS activity by these two metabolites is variable, but its usually in the range of 25- to 50fold; it is sometimes as high as 50- to 70-fold. With IH2017 D-fructose has no effect on LRS activity either alone or with L-alanine (3). L-Leucine also has an effect on LRS activity in the mutant but only 15-fold at a concentration of 40 mM. The stimulation of LRS activity by optimal and suboptimal concentrations of L-alanine and L-leucyl-ileucine is shown in Table 1. Optimal concentrations of each metabolite are not additive but suboptimal concentrations are. This indicates that these two metabolites either interact with two different receptors which influence a common process, or act at two different subsites on the same receptor (presumably a protein). Since mutant IH2017 has a slow growth rate and a very low level of LRS activity, it was of interest to examine the in vivo level of charged lysine tRNA. It was of particular interest to determine if the in vivo level of charged lysine tRNA was low, since in other studies a low in vivo level of charging of a tRNA has been associated with increased levels of the cognate synthetase (15, 16). The data in Table 2 demonstrate that low activity of LRS in the mutant is associated with low levels of in vivo charging of lysine tRNA (15-23%). The level of in vivo charging of lysine tRNA in the wild-type varied from 80 to 100% in our experiments. The addition of 3 mM Lleucyl-Lalanine to the growth medium had three effects on the mutant.
Metabolite Control of Lysyl-tRNA Synthetase
1365
TABLE 1. Optimal and suboptimal concentrations of L-alanine and L-leucyl-L-leucine stimulate lysyl-tRNA synthetase activity in IH2017 grown in supplemented minimal medium LysyltRNA Optimal syntheAdditionto concentra- tase, Strain medium tion, mM units/mg AT2092 8.6 IH2017 0.15 IH2017 Ala 20 8.5 IH2017 Leu-Leu 6 8.7 I112017 Ala + Leu-Leu 20 + 6 9.3
SubLysyloptimal tRNA concen- synthetration, tase, mM units/mg 9.5 0.25 10 2.6 3 7.7 10 + 3
11.4
The cells were grown to an ODs,0 of 0.3 (log phase). All additions were made prior to inoculation with the strains. The cells were grown at least four generations under all conditions. Lysyl-tRNA synthetase activity was assayed as described in ref 1.
The activity of lysyl-tRNA synthetase increased 25-fold, the in vivo charged level of lysine tRNA increased to the wild-type level, and the growth rate of the strain doubled. Thus, the slow growth of the mutant appears to be due to insufficient availability of lysine in the form of lysyl-tRNA for protein syntheS1S.
Previous work on mutants of this type has indicated that the increase in the activity of LRS elicited by metabolites is due to an increase in the amount of the enzyme (1, 3, 7). If so, this would indicate that the insufficiency of lysyl-tRNA in vivo in strain IH2017 is due to a deficiency in the amount of LRS. In order to corroborate the results of these previous experiments which used indirect methods to demonstrate changes in the amount of protein, a more direct approach, an immunological approach, has been taken. Antibody was prepared against LRS purified from wild type grown in the presence of 20 mM L-alanine. Two-dimensional immunodiffusion in agar of the anti-LRS antibody against LRS purified from wild-type grown in the presence and absence of 20 mM L-alanine, and from IH2017 grown with 20 mM L-alanine reveals a line of identity with no spurs (Fig. 1). The absence of spurs indicates that the mutant and wild-type TABLE 2. Charging of lysine tRNA in vivo upon growth in supplemented minimal medium
Strain AT2092 IH2017 IH2017 AT2092 IH2017
Generation time, min 45 120 63 40 100
%tRNA Addition to medium 3 mM Leu-Ala -
Chrig
Charging Lys Leu 98 23 79 98 15
89 100
LysyltRNA synthetase, units/mg 9.5 0.11 2.5 9.6 0.10
Cells were grown to an OD,%o of 0. 3 (log phase). Additions were made prior to inoculation to cells and the cells were grown at least four generations under all conditions. The in vivo charging was measured by the procedure of Folk and Berg (9) as modified by Lewis and Ames (10). Lysyl-tRNA synthetase activity was assayed according to ref. 1.
1366
Proc. Nat. Acad. Sci. USA 72 (1.975)
Genetics: Hirshfield et al.
FIG. 1. Two-dimensional immunodiffusion gels of purified lysyl-tRNA synthetase from wild-type cells grown with 20 mM L-alanine (A), wild type without alanine (B), and mutant IH2017 with alanine (C). The wells contain 8-10 ,ug of purified ptotein, and the center well 95 ,gg of antiserum protein.
enzymes are immunologically very similar. A second fainter precipitin line can be seen emanating from the wells containing wild-type LRS purified from medium with alanine (wells marked A). This second line lies between the wells marked A and the major precipitin band, but closer to the latter. This line most likely represents a second form of the enzyme. Recent experiments (Hirshfield, Yeh, and Zamecnik, un-
published) indicate that LRS is composed of two subunit forms, but one subunit form is present in large quantities only when wild type is grown in the presence of alanine. The antibody neutralization tests corroborate other results from this laboratory that the change in activity of LRS is due to a difference in the amount of enzyme protein present (Fig. 2). When the specific activities of wild-type and mutant LRS are similar (the mutant activity was enhanced 25- to 30-fold by alanine), the initial slopes in the neutralization assays are similar, indicating the presence of equal amounts of enzyme protein per unit of enzyme activity. As the specific activity of the mutant enzyme decreases the slope becomes flatter, which indicates that less enzyme protein is present. In a second set of experiments, neutralization assays were run on LRS from wild-type grown with and without 20 mM alanine. The activity of LRS from the former was 2.5 times higher than the latter, and the LRS from the alanine-grown cells was diluted 2.5 times so that the units of LRS in the assay were approximately equal. In Fig. 3 it can be seen that the initial slope of the neutralization curve is the same for each, demonstrating the presence of an equal amount of enzyme protein per unit of enzyme activity. Since the LRS from wild type grown with alanine was diluted 2.5-fold, it must be present in 2.5 times the amount of LRS from the control culture in the total crude extract protein. DISCUSSION Several laboratories have been working on the regulation of aminoacyl-tRNA synthetases. Originally Nass and Neidhardt (17), and later Williams and Neidhardt (81) proposed a mechanism for aminoacyl-tRNA synthetase regulation which is analogous to repression of amino-acid biosynthesis. According to their model a decrease in the amount of an amino-acid or
12 40
3.0
2.5 K,41
It4,)". '4
2.0
bc
4 A v C
1.5
i
ZZ
1.0
5
40
15
0.5
,u/ OF ioANr/SE-RUM
FIG. 2. Antibody neutralization assay of lysyl-tRNA synthetase from wild type and mutant IH2017. The wild-type strain was grown in supplemented minimal medium in the absence of alanine. The mutant st~rain was grown in the above medium in the absence or presence of various concentrations of L-alanine or L-leucylglycine. In these assays 15-20 /Ag of crude extract protein were incubated at pH 7.25 with various amounts of 1/20 diluted antiserum as described in Materials and Methods. Upon completion of the incubation the remaining lysyl-tRNA synthetase activity was assayed at pH 7.25 (see ref. 1). O. wild-type; A, mu'tant enzyme stimulated 25- to 30-fold; A, mutant enzyme stimulated 10-fold;
*, mutant enzyme,
no
stimulation.
5
10
15
20
,J/ OF ;.r2o ANTISERUM FIG. 3. Antibody neutralization assay of LRS from wild type grown with and without 20 mM L-alanine. The protein was diluted so that approximately the same number of enzyme units were present in the incubation with 1/20 diluted antiserum. The remainder of the experiment was carried out as in Fig. 2. 0, LRS from wild type without alanine, specific activity 7.5 units/mg; *, LRS from wild type with alanine, specific activity 19.6 units/mg.
Proc. Nat. Acad. Sci. USA 72
(1975)
aminoacyl-tRNA would lead to an increase in the amount of both the aminoacyl-tRNA synthetase and biosynthetic enzymes for that amino acid. This work has been extended by Williams and coworkers (e.g., 15, 19, 20), and, indeed, they have found that the restriction of the supply of a particular amino acid results in an increase in the activity of the cognate synthetase. In addition they have shown that low levels of in vivo charging of histidine tRNA in a histidyl-tRNA synthetase mutant is associated with an increase in both the level of the histidine biosynthetic enzymes, and histidyl-tRNA synthetase (15). A second mode of regulation has been proposed by Parker and Neidhardt (21) and Parker et al. (22), termed "metabolic regulation." They suggest that at least arginyl- and valyltRNA synthetase formation is regulated similarly to other macromolecular components of the bacterial cell's translational machinery. The fact that the metabolites, alanine and leucine peptides, can modulate the synthesis of LRS suggests a possible link to this model, but at this time a clear relationship has not been established. The mutant IH2017 discussed in this report appears to be a regulatory mutant for LRS. It has a very low level of LRS activity, but unlike the situation in other studies (as ref. 15) the low level of LRS activity is associated with low in vivo charged levels of lysine tRNA. The presence of lysine in the medium does not increase the in vivo charging, but inclusion of a leucine dipeptide increases both the activity of LRS and the in vivo charged level of lysine tRNA. Immunological experiments have confirmed other data (1, 3, 4) that the change in activity of LRS reflects the amount of enzyme protein present. Thus the amount of LRS in this strain is independent of the charging of lysine tRNA, but dependent upon the presence of the metabolites alanine and leucine peptides. These results, in addition to the fact that LRS from the mutant appears to be an altered protein, suggest a model which could account for the normally high level of LRS found in the cell. According to this model LRS is autoregulatory, and regulates its own synthesis by a positive control mechanism. Since the increase in LRS synthesis in these mutants can be blocked by the RNA synthesis inhibitor rifampicin (1), it would seem most likely that the control is transcriptional. Thus, it would be postulated that the mutation in LRS renders it less able to bind to DNA to regulate its own transcription, and that the
Metabolite Control of Lysyl-tRNA Synthetase
1367
metabolites alanine and leucine peptides restore this ability. Until this hypothesis can be tested it cannot be known whether the metabolites are absolutely required for normal maintenance of LRS synthesis or form part of an ancillary control system. We thank Miss Dianne Sanborn for her aid in isolating the mutants, and Ms. Barbara Wilkes, Dr. John Long, and Dr. Alan Aisenberg for their help in producing and isolating anti-lysyltRNA synthetase antibody. This investigation was supported by Grant NP-2Q of the American Cancer Society, and Grant RR05486-12 of the National Institutes of Health. This is publication No. 1474 of the Cancer Commission of Harvard University. 1. Hirshfield, I. N. & Zamecnik, P. C. (1972) Biochim. Biophys. Acta 259, 330-343. 2. Hirshfield, I. N., Tomford, J. W. & Zamecnik, P. C. (1972) Biochim. Biophys. Acta 259, 344-356. 3. Hirshfield, I. N. & Buklad, N. E. (1973) J. Bacteriol. 113, 167-177. 4. Buklad, N. E., Sanborn, D. & Hirshfield, I. N. (1973) J. Bacteriol. 116, 1477-1478. 5. Davis, B. D. & Mingioli, E. S. (1950) J. Bacteriol. 60, 17-28. 6. Novick, R. P. & Maas, W. K. (1961) J. Bacteriol. 81, 236240. 7. Cerd4-Olmedo, E., Hanawalt, P. C. & Guerola, N. (1968) J. Mol. Biol. 33, 705-719. 8. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. (1951) J. Biol. Chem. 246, 7595-7601. 9. Folk, N. B. & Berg, P. (1970) J. Bacteriol. 102, 204-212. 10. Lewis, J. A. & Ames, B. N. (1972) J. Mol. Biol. 66, 131142. 11. Waldenstrom. J. (1968) Eur. J. Biochem. 3, 483-487. 12. Kendall, F. E. (1937) J. Clin. Invest. 16, 921-931. 13. Ouchterlony, 0. (1949) Arkhiv. Kemi I, 43-48. 14. Pollock, M. R. (1956) J. Gen. Microbiol. 14, 90-108. 15. McGinnis, E. & Williams, L. S. (1972) J. Bacteriol. 111, 739744. 16. McGinnis, E., Williams, A. & Williams, L. S. (1974) J. Bacteriol. 119, 554-559. 17. Nass, G. & Neidhardt, F. C. (1967) Biochim. Biophys. Acta 134, 347-359. 18. Williams, L. S. & Neidhardt, F. C. (1967) J. Mol. Biol. 43, 529-550. 19. Archibald, E. R. & Williams, L. S. (1972) J. Bacteriol. 109, 1020-1026. 20. McGinnis, E. & Williams, L. S. (1972) J. Bacteriol. 109, 505511. 21. Parker, J. & Neidhardt, F. C. (1972) Biochem. Biophys. Res. Commun. 49, 495-501. 22. Parker, J., Flashner, M., McKeever, N. G. & Neidhardt, F. C. (1974) J. Biol. Chem. 249, 1044-1053.
