THE JOURNAL OF BIOLOGTCAL CHEMISTRY Yol. 236, No. 6, June 1961 Piinled in U.S.A. The Enzymic Synthesis of Amino Acyl Derivatives of Ribonucleic Acid II. THE PREPARATIOK OF LEUCYL-, VALYL-, IsomucYr,-, AND METHIOYYL RIBONUCLEIC ACID SYNTHETASES FROM ESCHERICHIA COLI* FRED H. BERGM.4NNJ7 PAUL BERG,$ AND hl. DIECKMANN~ From the Department of Microbiology, Washiizgton Unicersity Xchool of Medicine, St. Louis 10, Missouri (Received for publication, November 8, 1960) The enzymic synthesis of amino acyl ribonucleic acid deriva- tives appears to take place in two steps, both of which are cata- lyzed by a single enzyme (1-4). The first step involves the reaction of adenosine triphosphate, an amino acid, and an enzyme specific for that amino acid, resulting in the formation of an en- zyme-amino acyl adenylate complex and free inorganic pyro- phosphate. In the subsequent reaction, the amino acyl moiety is transferred to the 2'- or 3'-hydroxyl group of a terminal adenylic acid in a specific ribonucleic acid chain. The present paper describes the isolation and characterization of the en- zymes that catalyze the formation of leucyl-, valyl-, isoleucyl-, and methionyl ribonucleic acid. EXPERIXIENTAL PROCEDURE Escherichia coli strain B or ML30 was grown in a medium con- taining 1.1 % KeHP04, 0.85% KH2POa,O.G% Difco yeast extract, and 1% glucose. The cells were grown with vigorous aeration and harvested near the end of the period of esponcntial growth. The cells were washed with a solution containing 0.5% NaCl and 0.5% KCI, recentrifuged, and stored at -15O.I Cells stored for several months in this way did not lose significant amounts of enzymatic activity. Both the DL- and L-forms of the amino acids used were obt,ained from the California Foundation for Biochemical Research (CFP grade), or from Nutritional Biochemicals Corporation. ATP was purchased as the crystalline sodium salt from the Sigma Chemical Company. DEAE-cellulose, type 40, was pur- chased from The Brov-n Company, and it was converted to the hydroxyl form before equilibration with the appropriate buffers. Superbrite glass beads, 200 p average diameter, were obtained from the Minnesota Mining and Manufacturing Company, and they were washed, before use with 1 N HC1 and then with water. 1'3*-labelcd sodium pyrophosphate was prepared by heating * This investigation was supported by grant funds from the National Institutes of Health of the Unitcd States Public 1-Iedth Service. t Postdoctoral Research Fellow of the National Institutes of Health, United States Public Health Service; present address, Department of Biochemistry, Brandeis University, Waltham, Massachusetts. 1 Present address, Department of Biochemistry, Stanford Uni- versity School of Medicine, Palo Alto, California. 1 We are deeply grateful to Dr. A. Hanson of the Grain Process- ing Corporation for his help in obtaining E. coli cells grown in this way. NaHP3204 for 1 hour at 400" (5) and after separation from any remaining P3I inorganic phosphate by chromatography on a Dowex 1-C1- column (6), it was diluted with carrier sodium pyrophosphate to a final specific activity of lo4 to lo5 c.p.m. per pmole. Protein determinations were carried out by the method of Lowry et al. (7). Paper chromatographic separation of isoleucine and valine was carried out with a mixture of pyridine, iso-amyl alcohol, and H20 (35:35:30) (8). The activity of each of the amino acyl RNA synthetases was assayed by the amino acid-dependent eschange of ATP and PPi32 as described earlier (9). Potassium fluoride (0.01 hi), which WI'RS not inhibitory to the ATP-PP;32 reaction, was routinely added to the reaction mixtures to inhibit any inorganic pyrophospha- tase. One unit of enzyme activity is equivalent to the incor- poration of 1 pmole of PP,3* into ATP in 15 minutes. As already pointed out (9), the initial rate of incorporation of PYi3' into ATP was equivalent to the initial rate of amino acyl adenylate formation. Fig. 1 shows that for each assay the rate of the ATP-PPi32 exchange was proportional to thc amount of each amino acyl RNA synthetase preparation used. This propor- tionality was observed in the range of 0.02 to 0.3 pmole of ATP32 formation in 15 minutes (0.02 to 0.3 unit). RESULTS Crude extracts of E. coli catalyze the formation of the amino acyl adenylate derivatives of leucine, valine, isoleucine, and methionine, as well as the format,ion of the corresponding amino acyl RNA derivahives (10). The same initia.1 fractionation pro- cedures served for the purification of thc valyl-, isoleucyl-, and lcueyl RNA synthetases, and eventual separation of these activ- ities was achieved by chromatography on a DEAE-cellulose column (11). The methionyl RNA synthetase was obtained by another procedure. Isolation of Balyl-, Isoleucyl-, and LeuczJl RNA Synthetases- Unless otherwise stated, all operations were carried out at 0-5". A mixture of 100 g wet weight of E. coli cells, 100 ml of 0.025 IYI Tris buffer, pH 8.0, and 300 g of glass beads (200 p) was stirred in a stainless steel Waring Blendor run at about 90 volts. If the temperature of the mixture approached 5O, stirring was stopped and the material was cooled in an ice bath to 0". After a total of 15 minutes of stirring, 300 ml of Tris buffer were added, and the mixture was stirred at low speed for an addhional 2 to 3 minutes. After thc beads had settled, the supernatant fluid was 1735 1736 Enzymic Synthesis of Amino Acyl RNA Derivatives. 11 Vol. 236, No. 6 Fro. 1. The rate of ATP-PPis* exchange as a function of the amount of isoleucyl- (A), vaIyl- (B), leucyl- (C), and methionyl RNA synthetases (D) added. The reaction mixture contained, in a volume of 1.0 ml, 100 prnoles of Tris buffer, p1-I 8.0; 5 pmoles of hlgC12; 5 pmoles of potassiuni fluoride; 2 pmoles of ATP; 2 pinoles of P32-sodium pyrophosphate containing t.o lo5 c.p.m. per pmole; and 2pmoles of the appropriate amino acid. The incu- bation was carried out at 37" for 15 minutes and the amount of radioactivity in tlie ATP was determined as previously described (9). decanted, and the beads were washed twice with 150 nil aliquots of buffer. The first supernatant fluid and washings wcre coni- hined and centrifuged at 20,000 x g in a Serval1 SS-1 centrifuge for 1 hour and the precipitate discarded. The total volume of extract was usually 600 to 700 ml. For each 100 nil of cell extract, 5 ml of 1 M PIlnCls were added with stirring, and after 20 minutes, the mixture was centrifuged for 15 to 20 minutes as described above. To every 100 ml of the supcrnatant fluid were added 31 g of ammonium sulfate, and aftcr 20 minutes, the mixture was centrifuged at 30,000 X g for 1 hour and the precipitate discarded. To the supernntant fluid mere added 12 g of ammonium sulfa.te, and after the mixture was stirred for 20 minutes, it was centrifuged at 30,000 X g for 15 minutes. The precipitnte obtained from 100 g of cells was dissolved in GO to 65 ml of 0.02 M phosphate buffer, pII 7.5, and dialyzed for 5 hours against the same buffer. The dialyzed solution was clarified by centrifugation (ammonium sulfate frac- tion 1). To every 100 ml of nminoriiurn sulfate Fraction 1 were added 16 ml of 0.5 M potassium phosphate buffer, pH 7.5. The solution was incubated at 30" for 1$ hours, chilled to O", and then cen- trifuged to remove any insoluble material. By this procedure most of the nuclcic acid in ammonium sulfate Fraction 1 is con- verted to an acid-soluble form, facilitating the further purification of the cnzymes. The breakdown of the nucleic acids results from the action of a ribonucleasc present in the ribosomal particles (12). To each volume of the supernatant fluid were added 0.4 volume of 0.05 M ATP, pH 7.1, 0.2 volume of 0.1 M L-valine, 0.2 volume of 0.1 M MgC12, 0.14 volume of 0.05 M KF, and 0.06 volume of 0.5 M K2HPOa. The mixture was heated at 55" for 45 minutes, and after cooling to 0", the insoluble niaterial was removed by cent,rifugation. To each volunic of the heated su- pernatant fluid, 1.6 volumes of ammonium sulfate solution, snturated at 4", were added with stirring ovcr a 10-minute period. The precipit,ate was removed by centrifugation at 30,000 X g for 10 minutes and dissolved in 20 ml of 0.02 M phosphate buffer, pH 7.5, containing 0.05 N 2-mercaptoethanol. This solution was dialyzed against the same buffer for 2 to 3 hours, centrifugcd, and any insoluble material was discarded (ammonium sulfate Fraction 2). Although the heating procedure was designed to specifically protect the valyl RNA synthetase, about 50y0 of the activity with isoleucine and leucine remained after this treatment. Be- cause there a,ppeared to be considerable denaturation oE protein and tlicreforc some purification of cadi activity, the procedure was routinely used. The dialyzed ammonium sulfate Fraction 2 was adsorbed to a DEAE-cellulose column (3.8 cm2 x 16 cni) which had previously been equilibrated with 0.02 M phosphate buffer, pH 7.5, contain- ing 0.05 M 2-mercaptoethanol. A linear gradient of decreasing pH and increasing phosphate concentration was used to elute the enzymes. The nixing chamber contained 550 ml of 0.02 M phosphate buffer, pH 7.5, whereas the reservoir contained 550 ml of 0.25 M phosphate buffer, pH 6.5; both solvents contained 0.05 M 2-rr,ercaptoetlianol. Tlie flow rate was maintained at 50 to 60 in1 per hour and fractions of approsimately 25 ml were collected. The isoleucyl-, valyl-, and leucyl RN,4 synthetases were eluted after about 400, 600, and 800 nil of buffer, respec- tively, had passed through the column (Fig. 2). The major portion of cach peak was contained in 75 to 100 1x1. The frac- tions containing the major portion of each peak mere pooled and the enzymes were concentrated by addition of solid ammonium sulfate to 0.85 saturation. After 15 minutes, the precipitat,e was centrifuged at 30,000 x g for 1 hour and then dissolved in 2 to 3 ml of 0.02 M phosphate buffer, pH 7.5, containing M reduced glutathionc. The individual cnzymes were dialyzed overnight against about 100 volumes of the same buffer, and then against fresh buffcr for another 3 to 4 hours. Any precipitate which formed during the dialysis was removed by centrifugation, and the enzynics were stored at -15". Isolation of Methionyl RNA Synthetase-Extract from 100 g of E. coli was prepared as described above, and then dialyzed over- night against 20 liters of 0.01 M Tris buffer, pII 8.0. For each volume of dialyzed extract, 0.9 volume of 0.05 M Tris buffer, pH 7.5, and 0.1 volume of 2 M potassium phosphate buffer, pH 7.2, were added. This mixture was incubated at 30" for 2.5 to 3 hours, and after it was chilled to O", the small amount of precipi- tate was removed by centrifugation. As previously pointed out, this procedure facilitated the subsequent fractionation steps by depolymerizing the nucleic acids present in the extract. To each 100 ml of this solution were added 21.1 g of ammonium sul- 70 - 60 - _I 50- , 40- ?30- t ? 20- - z o b- U 10- 250 500 750 1000 0 VOLUME,ML. FIG. 2. Chromatographic separation of isoleucyl-, valyl- and leucyl RNA synthetases on a DEAE-cellulose column. See text for complete details. June 1961 Fractions F. H. Bergmann, P. Berg, and M. Dieckmann Volume 1737 23;:; uds/ nts protein TABLE I Purijication oJ valyl-, isuleucyl-, and leucyl RNA synthetases Jrom E. coli Total units' 11,400 6,850 G , 0GO 3,420 2,850 units/ nrg protein 1.3 3.4 lG 166 Protein concen- tration Crude extract. ................ Ammonium sulfate 1.. ............. Ammonium sulfate 2 ............. DEAE-cellulose fractions?, ........ Valyl RhTA synthetase. ........... nig/nrl 13.4 16.4 17.1 5.2 7.2 4.5 635 105 22 3.3 Val: ' RNA synthetase Fraction Dialyzed crude extract ..... Ammonium sulfate 1.. ...... .4mmonium sulfate 2.. ...... Cy gel eluate ............. Methioriyl RNA synthetase. Total units' Protein Volume concen- tration __- tnl mg fml 425 13.7 172 6.2 72 5.7 100 1.4 5 3.0 Specific activity Isoleucyl RNA synthetase I Leucyl RNA synthetase Total units' 16,600 11,550 5,320 3,770 3,120 Specific activity tmits/ mg protein 2.7 7.0 14 120 * Corrected for activity of isoleucyl RNA synthetase with valine. (The true valine-dependent Al'P-PPiTp exchange a,ctivity is ? The actual recovery of each enzyme from the coliinin was: valyl RNA synthetase, 4860 units; isoleucyl RNA synthetase, 46SO The values shown in the table represent the amounts in the most active fractions which equal to the valine-dependent ATP-PPi32 exchange activity minus 0.55 of the isoleucine-dependent ATP-PPi32 exchange activity.) units; leucyl IZN.4 synthctasc, 3875 units. were pooled and used in the ammonium sulfate concentration step. fate, and after stirring for 10 minutes thc solution was centri- fuged. To the supernatant fluid were added 6.3 g of ainmoniuni sulfate, and aftcr 10 minutes, the precipitate was removed by centrifugation and dissolved in about 20 ml of 0.02 M potassium phosphate buffer, pH 7.2 ammonium sulfate Fraction 1). To each 100 ml of ammonium sulfate Fraction 1, 18.3 g of animonium sulfatc and then 6.3 nil of 1.8 M acetic acid were added. After about 5 minutes, the precipitate was removed by centrifugation, and 4.1 g of ammonium sulfate were added to the supernatant fluid; the mixture was again centrifuged, and the precipitate was dissolved in 40 nil of 0.02 M potassium phos- phate buffer, pH 6.5 (ammonium sulfate Fraction 2). To each 100 ml of ammonium sulfate Fraction 2 were added 250 ml of cold water, and then 25 nil of alumina Cr gel (15 mg dry weight per ml). After 10 minutes, the mixture was ccntri- fuged at 10,000 x g, the gel was washed with about 250 ml of cold water, and the washings were discarded. The enzymc was eluted from the gel p-ith 150 ml of 0.1 M potassium phosphate buffer, pH 7.0, and this fraction was dialyzed overnight against 50 to 100 volumes of 0.02 ai potassium phosphate buffer, pH 7.5 (Cy eluatc). Of the dialyzed Cy eluate fraction, 100 in1 were adsorbed to a DEAE-cellulose column (3.2 cni2 x 7 cm) which had previously been equilibrated with 0.02 M potassium phosphate buffer, pH 7.5. The enzyme was eluted with a linear gradient of increasing ionic strength produced with a mixing chamber with 300 ml of 0.07 M potassium phosphate buffer, pH 7.0, and a reservoir of 300 ml of 0.2 M pot,assium phosphate buffer, pH 7.0. The flow rate was inaintaiiied at 50 to 60 in1 per hour, and 100 ml frac- tions were collected. The major portion of thc cnzyme was eluted in about 75 ml (Fractions 17 to 25). These frac.tions were pooled, and the enzyme mas concentrated by precipitation with ammonium sulfate as described for the other enzymes. The pellet was dissolved in 5 ml of 0.02 M phosphate buffer, p1-I 7.5; any insoluble material mas removed by centrifugation. The en- zyme solution was stored at -15". The data relative to the purification of the valyl-, isoleucyl-, leucyl-, and methionyl RKA synthetases are suiiimarized in Tables I arid 11. Sta.bility of Purijied Enzyme Fractions-Preparations of the most purified valyl- and isoleucyl RNA synthetases were rca- sonably stable when kept frozen at -15" in the presence of reduced glutathione M) for periods as long as 3 to 4 months. The loss in activity ranged froin 10 to 30%, and these prepara- tions could not be react,ivated by the addition of 2-mercapto- ethanol, reduced glutathione, or cysteine. In the absence of reduced glutathione, approximately 95 76 of the initial activity was lost after 1 month, but such preparations could be reactivated to about 50% of the initial level by the addition of 10-3 M 2-mer- captoethanol to the assay mixture. The purified preparations of niethionyl RNA synthetase showed little or no loss in activit,y on storage at -16", and no stimulation of the activity with glutathione or 2-mereaptoethanol was ever observed. In general, the purified leucyl RNA synthetase preparation was the most labile on storage. After several weeks at -15", there mas a marked dependence of the activity on 2-mercsptoethanol (or glutathione). With some aged and completely inactive prep- arations up to 90% of the activity could be restored by the addi- tion of the 2-mercaptocthanol. This stimulation of the activity by 2-mercaptoethanol was found with Tris, glycylglycine, /3 ,/3- dimethylglutarate, and imidazole buffers, but there was marked inhibition by 2-mereaptoethanol when cacodylate buffer \vas used. Requirements for B 1'P-PPi32 Exchange Reaction-With each of the purified enzymes there was 1 % or less of the activity of the. complete system if either the amino acid, A@++, ATP, or the TABLE I1 Pur@cci,tion of methiany1 RNA synthetase from E. coli Total units 7245 3870 2304 1710 1363 Specific activity anils/nzg protei,t 1.2 3.8 5.6 12.2 91 1738 L-Valine .._...., ,,.... D-Valine ... . . , . _. . . . . . . L-Isoleucine . , . . . . . . D-Isoleucine . . . . . . . . . . . 1,-Leucine. . . . . . . . . . D-Leucine . . . . , . , . . . . . . IJ-Methionine .. . . . . r,-Threonine . . , . . . . Enzymic Synthesis of Amino Acyl RNA Derivatives. 11 100 <1 1 < 1 <1 28 , enzyme mas omitted. The addition of 2-mercaptoethanol to the reaction mixture usually had a stimulatory effect. Leucyl RNA formation showed the greatest effect being activated 3 to 5 times in thc presence of 2-mercaptoethanol. In the case of the valyl and isoleucyl RNA Synthetases, the activation was between 1.5 and 3 times depending upon the age of the prepara- tion. Effect of pH and Type of Buffer on Rate of ATP-PPi32 Ex- change-Although each of the isolated enzymes displayed little variation in activity in the pH range of 7 to 9 there were marked differences with various buffers. For example, with the valyl RNA synthetase, maximal activity with Tris or glycine buffers was observed between pH 8.5 to 9.0, whereas with Cacodylate or glycylglycine buffer, the maximal rate, which was about 10% higher than those found with Tris and glycine buffer, was found between pH 7.0 and 8.0. In the case of the isoleucyl RNA synthetase, the activity was masimnl at pH 8.5 (glycine or Tris buffers). There was approxi- mately 50% of thc activity at pH 9.5 (glycine buffer), whereas at pH 6.6 (cacodylate buffer) there was 55% of the maximal ac- tivity. With the leucyl RKA synthetase there was little difference in the activity in the pH range of 6 to 9 with Tris, glycine, glycyl- glycine, or @, @-dimethylglutarate buffers. The methionyl RSA synthetase had the same activity bc- tween pH 6.0 and 8.8 and was approximately 40 and 30% as active at pK 5.3 (cacodylate buffer) and pH 9.3 (glycine buffer), respectively. Determinuhion oj I<, Values for ATP und Amino AcidsThe K,,, values for each of the amino acids were estimated by the Linewaver-Burk (13) plot from data obtained in the standard ATl'-PPiS2 exchange assay (Fig. 3). These are: L-isoleucine, 5 x M; L-valine, 1 X M; L-leucine, 5.6 X M; and mnethionine, 2.4 X hi. The I<,,, values for ATP in the formation of va.ly1-, leucyl-, and met,hionyI RNA under the conditions already described (2) are 2.3 x 10-5 M, 1.3 x IO-` M, and 8.5 x M, respectively. Specificity of Isolated Amino Acyl RNA Synthetases in Formation of Amino Acyl Adenylates (A) Nuturally Occurring Amino AcidsTable 111 shows the results of experiments in which each of the isolated enzyme 5 10 15- 15- 200 600 1000 Vol. 236, No. 6 TABLE 111 Specificity of valyl-, isoleucyl-, leucyl-, and melhionyl RNA synthstuses for amino acyl adenylate formation The results have been expressed with the specifically activated amino acid as 100 and all others relative to that figure, so that results with differcnt preparations carried out at different times could be tabulated. Each amino acid was tested under the con- ditions already described at a final concentration of L-amino acid of 2 X M. The results with DL- or L-amino acids were the same escept for certain instances noted in the test. Amino acid tested* I %t synthetase Isoleucyl Leucyl RNA RNA synthetase synthetasc Methionyl RNA synthetase 44-56 (1 100 <1 2-5 5-10 <1 4 <1 100 <1 2 <1 <1 <1 2 100 <1 * Experiments not included in the table with the r>-isomers of phenylalanine, tyrosine, tryptophan, lysine, serine, alanine, ar- ginine, histidine, cysteine, and glycine with each of the enzymes shown gave values less than 5% of that found with the appro- priate amino acid. L-Proline, L-glutamic, and aspartic acids were tested only with the methionyl RNA synthetase and found to have less than 1% the activity of L-methionine. preparations was tested for its ability to catalyzc the ATP-PPi32 exchange reaction in the presence of the different naturally occur- ring amino acids. Aside from the reaction of L-valine with the isoleucyl RNA syntheta.se, and L-threonine with the valyl RNA synthetase, both of which will be discussed subsequently, each of the enzymes was found to show high specificity for a single amino acid. It should be pointed out that in certain instances anomalous results were obtained with certain commercially ob- tained samples of L-amino acids. For example, several samples of L-leucine gave significant amounts of exchange with the methionyl RNA synthetase ranging from 10 to 50% of that found with L-methionine, whereas other samples of the same amino acid were inert. Similar observations were made with samples of isoleucine and valine with the leucyl RNA syn- thetase. W'e believc that such anomalous results are due to contamination of these L-amino acid preparations by other L-amino acids. Since the K, values for the amino acids by their respective enzymes (see above) were relatively low and because each amino acid preparation was routinely tested at a concen- tration of 2 x SI (with respect to the L-isomer), it is clear, for examplc, that an impurity of 1% of L-methionine in the L-leucine preparation or of L-leucine in the L-isoleucine and L-val- ine preparations would be sufficient to account for a significant exchange reaction. Moreover, since the isotope exchange is not accompanied by any net change in concentrations of the react- ants, the trace impurity functions catalytically. Support for this conclusion comes from the fact that synthetically prepared DL-amino acids were found to be uniformly inactive. This was not due to inhibition by the D-isomer since mixing equal amounts of the o-amino acid with the active L-amino acid samples gave no decrease in the rate of reaction. These findings are being stressed inasmuch as several investigators have tested the spec- June 1961 30 0 100 15 F. H. Bergrnann, P. Berg, and M. Dieckmann 1739 ificity of these enzymes by Ihe ATP-PPi3? exchange reaction (14, 15) with commercially obtained L-amino acid preparations. (B) Evidence for Formation of L-Valyl Adenylate by L-Isoleucyl Synthetase-During the analysis of the enzyme fractions eluted from the DEAE-cellulose column, two separate peaks of valyl adenylate-forming activity were found (Fig. 2). One of these (Peak 1) was found to react with L-isoleucine as well as L-valine. Moreover, the &io of activity with isolcucinc and valine was constant (1.72 to 1.82) in all fractions of the peak (see Fig. 2). The materkl in this peak accountcd for ovcr 85% of the isoleucine-dependent ATP-PP,32 exchange activity of am- monium sulfate Fraction 2 and about 30 % of thc valinc-dcpend- ent activity. The finding of two separable activities for valyl adenylate formation and the constant ratio of activity with isoleucine and valine in the first peak suggested the existence of a single enzyme capable of forming both isoleucyl- and valyl adenylate. The following lines of evidence support this hy- pothesis. 1. The activity observed in the prcscnce of equimolar but saturating amounts of isoleucine and valine are not additive but rather equal to the activity with isoleucine alone. This would be predicted on the basis of thc 100-fold difference in K,,, values for the two amino acids as mentioned below. 2. The ratio of activities with isoleucine and valine remains constant during inactivation and reactivation of the enzyme. In one preparation in which 2-mercaptoethanol was omitted from the eluting solutions, approximately SOY, of the activity was lost whether tested with isoleucine or with valinc as sub- strate. Moreover, incubation of the partially inactivated en- zyme preparation with 2-mcrcaptoethanol resulted in equal res- toration of the two activities. 3. The possibility of contamination of the valine samples with a trace of isoleucine as the basis for t.he observed activity with valine can be eliminated. With several different samples of L-valine or m-valine, the ratio of the two activities (with isoleu- cine and valine) did not vary significantly. When a mixture of L-isoleucine and L-valine was chromatographed with a solvent which separates the two amino acids (8), the material recovered from successive 0.5-cm strips cut perpendicularly to the line of solvent flow gave two separate peaks of activity when tested in licu of the amino acid in the standard assay system. The posi- tion of these two peaks corresponded to that for the isoleucine and valine. 4. As pointed out in the previous paper (a), although this enzyme produces both isoleucyl- and valyl adenylates, it cata- lyzes the formation of only isoleucyl RNA. Nevertheless, L-Val- ine competitivcly inhibits the reaction linking L-isoleucine to RNA, and the Ki for this inhibition is 3.8 x M, whereas thc K,, for valyl adenylate formation under the same conditions is 3.9 x M. These experiments suggest that whereas the isoleucyl RNA synthetase can form both Lvalyl- and L-isoleucyl adenylates, the enzyme differentiates between the two aniino acyl adenylate derivatives in thc subsequent transfer of the amino acid residue to RNA. (C) Utilization of Bnalogues of Valine by Valyl RNA Synthe- tase-The purified valyl RNA synthetase preparation shows some activity in the ATP-PPia exchange reaction with amino acid derivatives structurally related to valine (Table IV). DL- Threonine, which differs from DL-valine in the replacement of a methyl group by an hydroxyl group, has a 100-fold greater K, T.4BLE Iv Activity of analogues of DL-valine with valyl RNA synthetase All derivatives were tested under the conditions described un- der "Experimentul Proccdure," cxcept that thc substrate con- centrations were varied to determine the IC,,, values. Substrate oL-Valine . . . . . . . . . . . . . . . . . , . . . . . . . . . oL-Threoni ne. . . . , . . . . . , , . . , . . ~L-4llothreonine. . . . . . . . . . . . . . . . . DL-ol-Aniino$-chlorobutyrate. . . . . . . . DL-Allo-a-amino-8-chlorobutyrate. . . . uL-a-Aminobutyrate . . . . . , . . . . . . . . . . . DL-u-Aminoisobutyrate . . . . . . . . . . . . . . oL-B,B-Dimethylcysteine. . . . . . . , . . . . . K,* 'V 1 x 10-4 1.2 x lo-' 3.3 x 10-4 1 x 10-3 3.7 x 10-3 ~~ * The K, values are calculated in each case for the ~-isuiiier. (based on the L-isomer) than valine and the maximal rate is 30% of that found with DL-valine. DL-Allothreonine, in which the hydroxyl group is of opposite configurat.ion to that of m-thrco- nine, is inert. m-a-Amino-P-clilorobutyrate, in which the chlorine atom has the same configuration as the hydroxyl group of threonine, is at saturating conccntra.tions as active as DL- valine, although the K, is slightly higher. DL-Allo-a-amino-0- chlorobutyrate, which corresponds to DL-dlothreonine, is only 15% as active as m-threonine. It is interesting to note that the DL-a-arniiio-P-chlorobut):rtte analogue of DL-threonine is a strong inhibitor of valine incorpo- ration by animal cells in vitro whereas the ~~-allo-a-amino-p- chlorobutyrate is only 2070 as active.2 Substitution of one of the methyl groups of DL-valine by hydrogen, as in DL-a-amino- butyrate, results in an increase of the K, and a decrease in the V,,,. DL-a-hiniiioisobiityrate (norvaline) and DL-~ ,P-di- niethylcysteine (penicillamine) are inert. These data show that the valyl RNA synthetase can differentiate between closely related structures in the formation of t.he enzyme-amino acyl adenylate complexes. Further studies are required to evaluate whether the ardogues which arc converted to the adenylate derivatives are transferred to the valine-specific acceptor KNA chains. DISCUSSION It is now generally accepted that the meclianism of the amino acid-dependent exchange of ATP and PPP is explained by the reversible formation of an enzyme-bound amino acyl adenylate. To date, enzymes highly specific for the formation of L-methio- nyl- (9), L-tryptophanyl- (15), L-tyrosyl- (16), L-alanyl- (17), L-threonyl- (4), L-seryl- (18), D-alanyl- (19) , L-valyl-, L-isoleucyl-, and L-leucyl adenylates have been isolated. A closer inspection of several of these enzymes revealed that they catalyze not only the formation of a specific amino acyl adenylate but also the synthesis of the corresponding amino acyl RNA derivative (24, 17, 19, 20), and therefore may be considered as amino acyl RNA synthetases. Although each enzyme acts stoichiometri- cally in the formation of the enzyme-amino :icy1 adenylate com- plex, it appears to be regenerated during the subsequent syn- thesis of the aniino acyl RNA compound (2). tional Institutes of Health. generous gift of the DL-amino-0-chlorobutyrate derivatives. 2 Private communication from Dr. M. Rabinovita of thc Na- We arc also grateful to him for the 1740 Bnzymic Synthesis of Amino Acyl RNA Derivatives. II Vol. 236, No. 6 The present work ha.s revealed a somewhat anomalous situa- tion with rcspect to the isoleucyl- and valyl RNA synthetases. In the former case, L-valine is utilized at slightly lcss than one- half the rate of isoleucine, and in the latter, L-threonine reacts at about onc-third the rate of L-valine. In both instances, however, the K,, of the less active amino acid is about 100 times higher than that of the "natural" substrate. It seems reasonable to suppose, therefore, that at conditions in vivo there is little competition between amino acids for the appropriate enzyme. Of further interest is the finding that the isoleucyl ltNA syn- thetase will not form valyl RNA even though it makes the enzyme-valyl adenylate (2). It is not yet clear, however, whet,her such discrimination in the second reaction occurs with other synthetases which show low level activity with "unnatu- ral" amino acids and what the basis for this discriniination may be. It niiglit be pointed out that purified amino acyl RNA syn- thctases provide a relatively simple and specific tool for measuring the concentrations of a given L-amino acid in the presence of its D-analogue or in the presence of a mixture of amino acid. The relatively low K, values for t'he amino acids in the ATP-PPP exchange reaction (10-4 to 5 x 10-6 M) makes it possible to measure amino acid concentrations in this range. For exam- ple, Stevens et al. (21), using the methionyl RNA synthetase from yeast, have measured the biosynthesis of L-methionine by liver extracts. SUMMARY Amino acyl ribonucleic acid synthetases, relatively specific for either L-valine, L-leucine, L-isoleucine, or L-methionine, have been purified from extracts of Escherichia coli. The following statemeots summarize some of the pertinent findings made with the various Synthetase preparations. 1. The purified valyl-, isoleucyl-, and leucyl RNA synthe- tases are relatively unstable in the absence of a sulfhydryl- containing compound, whereas the methionyl ribonucleic acid synthetase preparation shows no sulfhydryl requirement. 2. The IC, values of 1 X M (L-valine), 5 x 10-6 iv (isoleucine), 5.6 X M (L-leucine), 2.4 X M (L-methio- nine) were found with the respective enzymes in the formation of the amino acyl adenylates. 3. Specificity studies show t,hat although the leucyl- and methionyl ribonucleic acid synthetases are highly specific for a single naturally occurring amino acid, isoleucyl ribonucleic acid synthetase forms ~-valyt adenylate at about one half the rate of L-isoleucyl adenylate formation, and valyl ribonucleic acid synt,hetase utilizes threonine at about 30 :h the rate of valine. Nevertheless, valine is not converted t.0 valyl ribonucleic acid by the isoleucyl ribonucleic acid synthetase. REFERENCES 1. BERG, P., AND OFENGAND, E. J., Proc. Natl. Acad. Sci. U. S., 2. BERG, P., BERGMAXN, F. H., OFENGAND, E. J., AND DIECK- 3. ALLEN, E. H., GLASSMAN, E., AND SCHWEET, R. S., J. Bid. 44, 78 (1958). MANN, M., J. Biol. Chem., 236, 1726 (1961). Chem., 235, l0Gl (19GO). H. G., AND Acs, G., J. Cell. Comp. Physiol., 54, 75 (1959). 4. LIPMANN, F., HULSSIANN, w. c., HARTMANN, G., BOMAX, 5. PENG, C. H. L., Biochim. et Biophys. Acta, 22, 42 (1956). 0. KORN~~ERG, A,, AND PRICE^, W. E., JR., J. Bid. Chem., 191, 7. LOWRY, 0. H., ROSEBRoUGIr, N. J., FARR, A. L., AND RANDALL, 535 (1051). It. J., J. Biol. Chem., 193, 265 (1951). 8. HEYNS, K., .~ND WaLma, W., 2. physiol. Chem., 287, 15 (1951). 9. BERG, P., J. Biol. Chem., 222, 1025 (1956). 10. BERG, P., AND OFENGAXD, E. J., Proc. Null. Acad. Sci. U. S., 11. PETERSON, E. A., AND SOBER, H. A., J. Am. Chem. Soc., 78, 12. SPAHR, P. F., AND HOLLINGWORTH, B. It., Federation Proc., 13. LINEWEAVER, H., AND BURK, D., J. Am. Chena. Soc., 56, 658 14. CLARK, J. M., JR., J. Biol. Chem., 233, 421 (1958). 44, 78 (1958). 751 (1956). 19, 318 (19GO). (1934). 15. DAVIE, E. W., KONINGSBERGER, v. v., AND LIPMANN, F., 16. SCHWEET, R. S., AND ALLEN, E. H., J. Biol. Chem., 233, 110-1- 17. HOLLEY, R. W., AND GOLDSTEIN, J., J. Biol. Chem., 234, 1765 18. WEBSTER, L. T., JR., AND DAVIE, E. W., Federation Proc., 18, 19. BADDILEY, J., AND NEUHAUS, F. C., Biochim. et Biophys. Acta, 20. WONG, G. K., AND MOLDAVE, K., J. Biol. Chem., 235, 694 Arch. Riochem. Biophys. 65, 21 (1956). (1958). (1959). 348 (1960). 33, 277 (1959). (1960). 21. STEVENS, A,, AND SAKAMI, w., J. BiOl. CheTrZ., 234, 2063 (1959)