Of the Same Glycine Residue by Arginine and Glutamic Acid Residues

Home , Amino acid replacement, Glutamic acid, Glycine, Glycine (data page)

VOL. 48, 1962 GENETICS: HENNING AND YANOFSKY 1497

of their data with ours indicates that the phenomenon described herein is most likely due to the combination of ma-i + and ry+ subunits. We have also observed reactivation of "inactive" xanthine dehydrogenase of wild-type Drosophila. II Forrest, H. S., E. W. Hanly, and J. M. Lagowski, Genetics, 56, 1455 (1961). 12Horowitz, N., personal communication. 13 Hadorn, E., and I. Schwink, Nature, 177, 940 (1956); Glassman, E., Dros. Inform. Service, 31, 121 (1957). '4Glassman, E., and W. Pinkerton, Science, 131, 1810 (1960).

AMINO ACID REPLACEMENTS ASSOCIATED WITH REVERSION AND RECOMBINA TION WITHIN THE A GENE* BY ULF HENNINGt AND CHARLES YANOFSKY

DEPARTMENT OF BIOLOGICAL SCIENCES, STANFORD UNIVERSITY, STANFORD, CALIFORNIA Communicated by V. C. Twitty, July 26, 1962 Studies with the A protein of the tryptophan synthetase of Escherichia coli have established that mutationally altered sites that are located at or near the same position in the A gene lead to amino acid substitutions in the same tryptic peptide of the A protein." 2 It has also been shown that two mutants, strains A23 and A46, with alterations extremely close to one another on the genetic map, form A proteins that are distinguishable from the wild-type A protein by the replacement of the same glycine residue by arginine and glutamic acid residues, respectively.1 2 It was concluded from these findings that the mutational alteration characteristic of each of the mutants involved a different nucleotide in the same amino acid coding unit in the A gene. 1 2 Mutants A23 and A46 revert spontaneously to tryptophan independence. Some of the strains obtained, termed full revertants, are phenotypically indistinguishable from the wild type, while others, termed partial revertants, grow slowly without tryptophan supplementation. The mutations resulting in partial reversion may occur at the same site as the original mutation (primary site reversion), or at a second site in the A gene.3' 4 The present report is concerned with a study of a primary site partial revertant and several full revertants of mutant A46, and full revertants derived from two iden- tical mutants, A23 and A28. The results obtained demonstrate that any one of four different amino acids can occupy a certain position in the A protein and the protein will be functional. It is also shown that recombinants obtained from crosses between some of the strains form A proteins with different amino acids at a particular position in the A protein than the amino acids present at the same position in the A proteins of the parental strains. Materials and Methods.-Bacterial strains: Mutants A23, A28, and A46, obtained by penicillin selection6 following ultraviolet irradiation of E. coli K-12, have been described in detail elsewhere." 2, 7 The full and partial revertants studied were spontaneous revertants.3, 5 Genetic procedures: The methods employed for the preparation of transducing lysates of phage Plkc and for transduction with this phage have been described Downloaded by guest on September 28, 2021 1498 GENETICS: HENNING AND YANOFSKY PROC. N. A. S.

by Lennox.8 In order to virtually exclude the carry-over of transducing phage with the wild-type tryptophan region, all lysates were prepared with phage that was grown on the donor strain. The use of cys- tryp- recipients for recombina- tional analyses has been discussed previously.7 Isolation of peptides and amino acid analyses: The procedures used for the isolation of the A protein,9 as well as the methods used for the digestion of the protein with trypsin and chymotrypsin,1' the isolation of peptides,2"1' and the analysis of peptides have been described previously." 2 Results. Characteristics of strain A46PR9 and the full revertants derived from mutants A46, A23, and A28: The characteristics of the revertant strains examined in this study are summarized in Table 1. With the exception of partial revertant A46PR9, the strains are indistinguishable from the wild type. As reported previously,3 the A protein of strain A46PR9 is only about 1/1,000 as active as the wild- type strain in the physiologically important reaction, the conversion of indoleglycerol phosphate to tryptophan. The A proteins of all the revertants studied TABLE 1 CHARACTERISTICS OF REVERTANTS OF MUTANTS A23 AND A46 Generation Accumulation Specific activity time, of of Strain Colony size* minutes* indoleglycerol A protein: wild type - 70-72 - 2 A46 auxotroph auxotroph +t 31t A46PR9 2/3 of wild type 84 + 91 A46FR1 = wild type 70-72 - 2.2 A46FR2 = wild type 70-72 - 2.1 A23 auxotroph auxotroph + t 23 t A23FR1 = wild type 70 - 2.1 A23FR2 = wild type 70 - 1.9 A28FR1 = wild type 70 - 1.9 * On glucose-minimal medium at 370C. t Grown on limiting amounts of indole (2.5 jig/nml). I In the conversion of indole to tryptophan.

have the same electrophoretic mobility as the wild-type protein in polyacrylamide gel"2 at pH 9.5.13 The A46 protein is more negatively charged under the same conditions. This behavior is consistent with the amino acid substitution in this protein, glycine to glutamic acid. The charge of the A23 protein could not be determined in polyacrylamide gel, presumably because of its extreme lability.'3 The A23 protein would be expected to be more positively charged than the wild- type protein. The mobilities of the revertant proteins suggest that in each case a charged amino acid was replaced by an amino acid with a side chain that is not charged at pH 9.5. Examination of the A46PR9, A46FR1, and A4f6FR2 tryptic peptides corresponding to the tryptic peptide altered in mutant A46: Tryptic peptide TP3 was isolated from digests of the A46PR9, A46FR1, and A46FR2 proteins. The amino acid composition of these three peptides is compared with the composition of the corresponding wild-type and A46 peptides in Table 2. The analyses show that one glutamic acid or glutamine residue present in the A46 peptide is replaced by a valine residue in the peptide from A46PR9, by an alanine residue in the peptide from A46FR1, and by a glycine residue in the peptide isolated frcm A46FR2. As described previously,2 chymotryptic hydrolysis of peptide TP3 gives three peptides, designated Downloaded by guest on September 28, 2021 VOL. 48, 1962 GENETICS: HENNING AND YANOFSKY 1499

TP3C1, TP3C2, and TP3C3. These - three peptides were isolated from chy- O . Oj O motryptic digests of TP3 from the partial revertant and the two full revertants of A46. The amino acid analyses of these R+g R. peptides (Table 2) showed that in all cases peptide TP3C1 contained the amino - acid substitution. Since both glutamic coRoO° acid and glutamine residues were present ¢a in TP3C1 from the A protein of mutant 0 A46, it was not possible to decide from ¢ the analyses alone whether the glutamic W e A acid residue or the glutamine residue was >- ¢ E replaced in the peptides from the revert- - eq ,^ V_ oooo0 0 C eq ants. It had been deduced" 2 from pre- : ..H - vious experiments that the glycine in g, TP3C1 from the wild-type protein was Z ¢ >, t replaced by arginine in peptide TP3C1 00° --4 °s 4

from strain A23. Since the glycine in the be wild-type peptide is also replaced by glutamic acid in peptide TP3C1 from mutant 2 0 A46, it was assumed that glycine, argin- v) ine, and glutamic acid occupy the same w position in these different proteins. Ex- Q a) periments using carboxypeptidase diges- ¢¢d tion of the TP3C1 peptide from the vari- z ous proteins have now verified this as- COO. sumption.15 The same studies have Eq , shown that the valine, alanine, and gly- a cine residues of the A46PR9, A46FR1, v Q"coE, 0-4co- o0o co aco 0 3 and A46FR2 peptides occupy the same * position, in each case replacing the glu- °

tamic acid residue of the A46 peptide evN _ q (Fig. 1). Examination of the A23FR1, A23FR2, oa)O

and A28FR1 chymotryptic peptides cor- E ID s, W.0 0 responding to the chymotryptic peptide ct "am*0 altered in mutant A23: Peptide TP3C1 S.0 is present in chymotryptic digests of the -- Cq-q -4 - a) A protein and thus can be isolated with- 0Eq Go<-

out prior digestion of the protein with CD

trypsin. This peptide was isolated from B C Cq-4 M -- - chymotryptic digests of the A proteins of '50 0 3 the three full revertants derived from 0 °,t. strains A23 and A28. The results of the 0 *~~0a)~~~ 0 'v-C)~ a) amino acid analyses of the three peptides E Q08° C0)~ .0 Z- a are given in Table 2. In the A protein of W >b* F. Downloaded by guest on September 28, 2021 1500 GENETICS: HENNING AND YANOFSKY PROC. N. A. S.

C TP3Cl C

11 2-5 6 7 8 9 -TyrjAsp-(Pro2, Ala2)-Leu-G1uNH2-Gly-Phe Gly- UGG 4 (wild type)

UAG UGC -Glu- -Arg- (A46) (A23, A28) I X/IX UUG UCG UGG UGG UUC UGG -Val- -Ala- -Gly- -Gly- -Ser- -Gly- (A46PR9) (A46FR1) (A46FR2) (A23FR1) (A23FR2) (A28FR1) FIG. 1.-The amino acid substitutions at position 8 in peptide TP3Cl and the corresponding RNA coding triplets.". 20 C indicates bonds attacked by chymotrypsin. The nucleotide sequence for glutamic acid has been assigned arbitrarily. The other sequences are based on the assumption that each mutational change involves a single nucleotide substitution. Partial amino acid sequence data is based on unpublished studies of Carlton and Yanofsky. strains A23FR1 and A28FR1 the arginine residue characteristic of the A23 or A28 peptides is replaced by a glycine residue, while the arginine is replaced by serine in the peptide obtained from the protein of A23FR2. Carboxypeptidase treatment of TP3C1 from these strains has shown'5 that the two glycine residues and the serine residue are located in the same position as the arginine residue in the A23 peptide, i.e., in the same position as the glycine residue in the wild-type peptide (Fig. 1). Genetic analyses: It was reported previously that tryptophan-independent colonies were obtained at a low frequency in transduction crosses between mutants A23 and A46.1 7 The partial revertant A46PR9 can also be employed as a "mutant" in transduction experiments if the plating medium contains 5-methyl tryptophan (0.5 ,ug/ml). This supplement inhibits the growth of A46PR9 colonies while having no effect on the growth of wild-type colonies. Transduction crosses were performed in all possible combinations with strains A46PR9, A23, and A46 (Table 3). Recombinants were obtained in the following transductions: A46 -. A23, A23 -o A46 and A46PR9 --- A23. The absence of recombinants in the other crosses cannot be considered conclusive in view of the low frequency of appearance of recombinants in the transductions that did yield recombinants. Because of the low recombination frequencies in these crosses it was essential to exclude reversion as a source of tryptophan-independent colonies. The low reversion rates of mutants A23 and A46 (< 10-8) and the use of cys- tryp- double mu- tants7 virtually eliminate reversion as a factor to consider in crosses between these strains. Strain A46PR9 presents a more serious problem, however, because it is not possible to determine its reversion rate by plating on minimal medium. Fur- thermore, it is not possible to estimate the reversion rate of A46PR9 by plating on agar supplemented with 5-methyl tryptophan since the rate of spontaneous mutation to 5-methyl tryptophan resistance by a change outside the A gene"'6 1' is very high, about 10-6. In order to overcome these difficulties the reversion of A46PR9 Downloaded by guest on September 28, 2021 VOL. 48, 1962 GENETICS: HENNING AND YANOFSKY 1501

to the wild-type-like state was determined indirectly. It was reasoned that since the methyl tryptophan resistance locus was not transduced jointly with the A gene, it would be possible by transduction to distinguish between reversion in the A gene and mutation at the methyl tryptophan resistance locus. A suspension of A46PR9 cells was plated on a large number of plates containing 5-methyl tryptophan at a level that would inhibit A46PR9 but not wild type. The colonies obtained were scrubbed from the surface of the plates with a spreader and this cell suspension was used to prepare a lysate. This lysate was added to a suspension of A46 cells and the cells were plated on minimal agar and minimal agar supplemented with methyl tryptophan. The number of colonies obtained on the methyl tryptophan plates was less than 1/500th of the number of colonies on the minimal plates. Since the methyl tryptophan resistance locus is not transduced jointly with the A gene, the colonies on the two types of plates represent A46PR9 plus wild-type-like revertants (on minimal) and wild-type-like revertants alone (on methyl tryptophan plates). The A46PR9 colonies on the minimal plates must have been derived from bacteria transduced by phage grown on A46PR9 carrying the methyl tryptophan resistance locus. Other experiments have shown that there is no loss of wild-type-like recombinants when the plating medium is supplemented with 5-methyl tryptophan. Since less than 1/500th of the colonies were wild type- like it appears that the reversion rate of A46PR9 to the wild-type-like state is roughly the same as that of mutants A23 and A46. TABLE 3 RECOMBINATION TESTS % recombination$ donor recipient cys + transductants* his - cys + tryp + his - cys + tryp + (his+ cys +) (his - cys ) observed correctedt transductants cys + corrected A46PR9 A46 951,000 570,000 2§ <0.00018 A46PR9 A23 1,010,000 606,000 411 0.00066 A46PR9 A46PR9 453,000 271,800 0 A46 A46PR9 727,000 436,200 0 <0.00023 A46 A23 791,000 474,600 3 0.00063 A46 A46 443,000 265,800 0 A23 A46PR9 761,000 456,800 0 <0.00022 A23 A46 345,000 207,000 1 0.00048 A23 A23 523,000 313,800 0 * Sum of 3 experiments for each combination listed in column 1 and 2. t Corrected for the number of cys+ transductants which could have undergone recombination in the tryp region. cys and tryp markers are transduced jointly at frequencies of 0.53-0.66;7 0.6 was used as the cor- rection factor in this table. $ In order to obtain the "true" recombination frequencies these values should be multiplied by a factor of 2 since approximately 50% of the recombinants are lost in large-scale experiments. This factor is relatively con- stant and was determined with appropriate control transductions in each series of experiments. § These two colonies are not recombinants but are A46PR9's that are resistant to 5-methyl tryptophan. Two of these accumulate indoleglycerol. All of the recombinants listed in Table 3 were picked, purified, and tested for indoleglycerol accumulation. As indicated in Table 3, four accumulators were found. The two accumulators from the transduction A46PR9 -p- A46 were shown to be methyl tryptophan-resistant A46PR9 and were not investigated further. The two accumulators from the A46PR9 -- A23 transduction were found to be recombinants at the A46PR9 and A23 sites. The accumulation was shown to be due to an alteration at some other site, presumably within the A gene. This alteration has no apparent effect on the growth of these strains on low levels of methyl tryptophan, but strains with this alteration accumulate indoleglycerol. It has since Downloaded by guest on September 28, 2021 1502 GENETICS: HENNING AND YANOFSKY PROC. N. A. S.

been established that this second alteration was introduced into the recombinants from the A46PR9 donor stock that was employed in these crosses. Apparently this second mutational alteration occurred in the A46PR9 stock at some time during the two-year period since it was originally isolated. Examination of peptide TP3C1 from the A proteins of recombinants: The wild- type-like recombinants recovered from the transductions listed in Table 3 are designated according to their origin, e.g., 46PR9/23Rel refers to recombinant 1 from the transduction A46PR9 -> A23. To date, chymotryptic peptide TP3C1 has been isolated from five of the eight wild-type-like recombinants listed in Table 3. Each recombinant peptide has been found to contain an amino acid at position 8 other than those present at this position in the peptides of the parental strains (Table 4). TABLE 4 MOLAR RATIOS* OF CONSTITUENT AMINO ACIDS IN CORRESPONDING PEPTIDES Peptide TPC31 from Recombinant strains Amino Parental strainst 46PR9/23 46PR9/23 46PR9/23 46/23 46/23 acid A46 A23 A46PR9 Rel Re2 Re3j Rel Re3 Arginine 1 Aspartic 1 1 1 1.00 0.90 0.95 0.98 0.96 Serine§ 0.93 0.99 Glutamic 2 1 1 1.05 1.30 1.16 1.02 1.05 Proline 2 2 2 2.04 2.18 1.83 2.00 1.98 Glycine 1.19 1.02 1.06 Alanine 2 2 2 2.04 1.97 1.98 1.95 1.95 Valine 1 Leucine 1 1 1 0.99 0.94 0.98 0.96 1.03 Phenylalanine 1 1 1 0.95 0.86 0.90 0.90 0.97 * Calculation of molar ratios according to Hirs et al.14 t Given as number of amino acid residues present. t Accumulates indoleglycerol. § Corrected for partial destruction upon acid hydrolysis. Discussion.-The results presented in this paper demonstrate that six different amino acids can occupy the same position in the A protein of tryptophan synthetase. The amino acids alanine and serine can apparently replace glycine without any de- tectable effect on the ability of the A protein to function in tryptophan formation. However, extensive kinetic and stability studies must be performed before it can be concluded that these substitutions have no effect on the stability or catalytic activity of the protein. When valine occupies the same position in the A protein, the protein is only slightly active in the physiologically important reaction, the conversion of indoleglycerol phosphate to tryptophan. These findings suggest that there may be certain spatial requirements at this position in the protein, if the protein is to be catalytically active. Charged amino acids at this position also have a profound effect; the proteins with arginine and glutamic acid at this position are enzymatically inactive. When the location of peptide TP3C1 in the native protein is known, as well as its relation to the active site of the protein, it will be of interest to re-evaluate these substitutions. The results of these studies have considerable bearing on the interpretation of reversion experiments. Revertants with glycine, alanine, and serine A proteins are phenotypically indistinguishable, while strains with the valine protein are very similar to the wild-type strain in growth rate. In most reversion studies all of these strains would be scored as wild-type revertants and assumed to have under- Downloaded by guest on September 28, 2021 VOL. 48, 1962 GENETICS: HENNING AND YANOFSKY 1503

gone the same mutational event. Obviously, these strains arose by different mutational events (nucleotide changes) and thus interpretation of reversion rates or reversion events on the basis of phenotypic properties alone would be of question- able value. It is apparent, therefore, that reversion is exceedingly complex when analyzed in detail. It is likely that reversion studies detect all possible single amino acid changes in a protein that will lead to functional activity, whether or not these changes are at the site of the original amino acid substitution in the mutant protein. Amino acid replacements in a series of strains derived from each other or from a common strain by single mutational changes in the same coding unit have considerable bearing on the genetic code. If there is appreciable degeneracy in the code, only by studying amino acid replacements at the same position in a protein could one be certain that the corresponding coding units are related in sequence and composition. Furthermore, series of this type have a high probability of representing single-step changes while the amino acid replacement data from naturally occurring and distinguishable proteins may involve intermediate amino acids in each replacement, e.g., arginine could change to glutamic acid through the intermediate glycine. The finding in the present study that six amino acids can occupy the same position in the protein establishes that at least two nucleotides code for these amino acids. Furthermore, since in TP3C1 from each of the six proteins the amino acids in positions 7 and 9 are the same, glutamine and phenylalanine, respectively,'5 it seems very unlikely that the code is of the overlapping type, unless more than three nucleotides code for each amino acid. Since each of the amino acid substitutions is apparently associated with a single mutational event, it is of interest to compare the amino acid substitutions with the triplet coding units that have been assigned to the amino acids."9 20 It can be seen in Figure 1 that each amino acid substitution is consistent with a single nucleotide change, the other two nucleotides of the coding units remaining unaltered. If the order UAG18 is arbitrarily selected for glutamic acid, then the order of the nucleotides corresponding to alanine must be UCG since a single mutational change results in the replacement of glutamic acid by alanine. The sequence of the same letters, U, C, and G, corresponding to arginine, must be different than the sequence for alanine. Since the glycine to arginine change is a G to C change, the relative order for arginine must be UGC (Fig. 1). This assignment in turn fixes the relative positions of the nucleotides corresponding to serine. Since the observed amino acid substitutions are consistent with the triplet code letters assigned to the six amino acids, it seems likely that the relative positions of the nucleotide in the different coding units can be specified. The order of the nucleotides cannot, of course, be determined by these considerations alone. It is possible to test the assignment of different relative positions to the code letters by analyzing the recombinants of appropriate crosses. As indicated in Table 5, glycine recombinants should be obtained from A46 X A23 crosses if the relative order assigned to the nucleotides is correct. A second recombinant type corresponding to the letters UAC is also expected from this cross but has not been found as yet. Serine and glycine recombinants should be obtained from A46PR9 X A23 crosses. On the other hand, no recombinants should be recovered from the A46PR9 X A46 cross since the difference between the two coding units is at Downloaded by guest on September 28, 2021 1504 GENETICS: HENNING AND YANOFSKY PRoc. N. A. S.

TABLE 5 AMINO AciD REPLACEMENTS RESULTING FROM INTRA-CODING UNIT RECOMBINATION Amino acids Corresponding Possible Strains at position 8 RNA coding recombinant Corresponding Amino acids crossed of TP3C1 units*. t coding units amino acids found A46PR9/A23 valine/ UUG/UGC UUC serine $ serine arginine UGG glycine t glycine A23/A46 arginine/ UGC/UAG UGG glycine t glycine glutamic UAC histidine? threonine? asparagine? A46PR9/A46 valine/ UUG/UAG None - no recombinants glutamic detected * Relative sequences based on considerations discussed in text (also see Fig. 1). From references 19 and 20. Based on relative sequences given in Fig. 1. the same position. The data obtained are consistent with these expectations. The demonstration that recombinants from two crosses have amino acids at position 8 in peptide TP3C1 other than those present at this position in the parental proteins has two additional important implications. Clearly, recombination can take place within a coding unit, and intra-coding unit recombination can have the same effect as mutation. The authors are greatly indebted to Virginia Horn, Janet Lind, Jean Hale, and John Horan for their excellent technical assistance. * This investigation was supported by grants from the National Science Foundation and the U.S. Public Health Service. t Fulbright grantee 1960-62; present address: Max-Planck-Institut fur Zellchemie, Mfinchen, Germany. 1 Helinski, D. R., and C. Yanofsky, these PROCEEDINGS, 48, 173 (1962). 2 Henning, U., and C. Yanofsky, these PROCEEDINGS, 48, 183 (1962). 3 Yanofsky, C., D. R. Helinski, and B. D. Maling, in Cellular Regulatory Mechanisms, Cold Spring Harbor Symposia on Quantitative Biology, vol. 26, p. 11 (1961). 4Yanofsky, C., U. Henning, D. R. Helinski, and B. C. Carlton, Fed. Proc., in press; D. R. Helinski and C. Yanofsky, in preparation. 5 Allen, M. K., and C. Yanofsky, in preparation. 6Davis, D. B., J. Am. Chem. Soc., 70, 4267 (1948); Lederberg, J., and N. Zinder, J. Amn. Chem. Soc., 70, 4267 (1948); Adelberg, E. A., and J. W. Myers, J. Bacteriol., 65, 348 (1953). 7Maling, B. D., and C. Yanofsky, these PROCEEDINGS, 47, 551 (1961). 8Lennox, E. S., Virology, 1, 190 (1955). 9 Henning, U., D. R. Helinski, F. C. Chao, and C. Yanofsky, J. Biol. Chem., 237, 1523 (1962). 15 Helinski, D. R., and C. Yanofsky, Biochim. Biophys. Acta, in press. 11 Rudloff, V., and G. Braunitzer, Hoppe Seyler's Z. Physiol. Chem., 323, 129 (1961). 12 Ornstein, L., and B. J. Davis, "Disc Electrophoresis," Preprint by Distillation Products Industries, Div. of Eastman Kodak Co. (1961). 13 Henning, U., and C. Yanofsky, in preparation and unpublished observations. 14 Hirs, C. H. W., S. Moore, and W. Stein, J. Biol. Chem., 219, 623 (1956). 15 Carlton, B. C., and C. Yanofsky, in preparation. 16 Cohen, G., and F. Jacob, Compt. rend., 248, 3490 (1959). 17 One locus responsible for 5-methyl tryptophan resistance is located near the threonine locus oi1 the E. coli chromosome.16 18 The letters A, U, C, and G correspond to residues of adenylic, uridylic, cytidilic, and guanylic acids, respectively, in polynucleotides. '9 Matthaei, J. H., 0. W. Jones, R. G. Martin, and M. W. Nirenberg, these PROCEEDINGS, 48, 666 (1962). 20 Speyer, J. F., P. Lengyel, C. Basilio, and S. Ochoa, these PROCEEDINGS, 48, 441 (1962). Downloaded by guest on September 28, 2021