- Email: [email protected]
Isolation of tRNA isoacceptors by affinity chromatography on immobilized bacterial elongation factor Tu
ANALYTICAL
BIOCHEMISTRY
136, 161-167 (1984)
Isolation of tRNA lsoacceptors by Affinity Chromatography on Immobilized Bacterial Elongation Factor Tu’ KARL-HEINZDERWENSKUS,WOLFGANGFISCHER,ANDMATHIASSPRINZL Lehrstuhl
ftir
Biochemie der Universittit D-8580 Bayreuth, Federal
Bayreuth, Universitiitsstrasse Republic of Germany
30,
Received April 27, 1983 Elongation factor Tu from Escherichia coli or Thermus thermophilus was immobilized on cyanogen bromide-activated Sepharose. Immobilized elongation factor Tu . GDP could be converted to Tu. GTP, which is able to bind aminoacyl-tRNA. When bulk tRNA from E. coli, baker’s yeast, or bovine liver was aminoacylated by one amino acid only, the resulting aminoacyltRNA could be separated in one step from the rest of the tRNA using an affinity column of immobilized elongation factor Tu . GTP. Specific tRNA isoacceptors can be isolated in amounts sufficient for gel electrophoretic analysis, sequence determinations, and hybridization experiments. KEY WORDS: tRNA isoacceptors; aminoacyl-tRNA; immobilized protein; affinity chromatography; elongation factor Tu; two-dimensional gel electrophoresis.
Isolation of pure tRNA species for sequence analysis and other biochemical investigations is still tedious. Besides conventional column chromatographic procedures for the purification of non-aminoacylated tRNA species ( l3) various techniques for the isolation of aminoacyl-tRNA isoacceptors have been developed. These techniques mostly rely on the specificity of the enzymatic aminoacylation reaction combined with chemical derivatization of the free a-amino group of the attached amino acid. If aminoacylation of unfractionated tRNA is performed in the presence of one amino acid only, the corresponding tRNA isoacceptors will provide free a-amino groups for subsequent modifications. Gillam et al. modified such aminoacyl-tRNAs by introducing aromatic moieties and isolated the modified aminoacyl-tRNAs by chromatography on benzoylated DEAEcellulose (4). Another method uses a bifunctional reagent to attach the acylated tRNA species via their a-amino groups to a sulfhydryl-substituted Sepharose 4B (5). Variable ’ This publication is dedicated to the 60th birthday of Prof. Dr. F. Cramer.
yields and side reactions of the chemical derivatization are the main disadvantages of these methods (6,7). Separation of the aminoacyl-tRNAs from uncharged tRNA without further derivatization was also reported. The majority of these methods are confined to the isolation of only one (8,9) or a few ( 10) particular isoacceptor sets and are lacking general applicability. The method using affinity columns containing dihydroxyboryl residues (11) is limited by the fact that the borate complex at the cis-diol groups of the uncharged tRNA is unstable, especially at neutral and acidic conditions necessary to preserve an integrity of the aminoacyl-tRNA ester bond. Recently, we reported on the l&and-binding properties of the immobilized elongation factor Tu from Thermus thermophilus (12). In the present work we describe the use of the immobilized EF-Tu2 from T. thermophilus and from Escherichia coli for the isolation of isoaccepting tRNAs. For this purpose the natural high affinity of bacterial EF-Tu . GTP for * Abbreviations used: EF-Tu, elongation factor Tu; Hepes, 4-(2-Hydroxyethyl)-l-piperazineethanesulfonic acid. 161
0003-2697184 $3.00 Copyright Q 1984 by Academic Press Inc. All rights of reproduction in any form reserved.
162
DERWENSKUS,
FISCHER, AND SPRINZL
Sepharose 4B using the procedure suggested by the manufacturer (18). EF-Tu . GDP or EFTu . GTP was dialyzed against a buffer containing 100 mM sodium borate, pH 8.3, 500 mM NaCI, 10 mM MgS04, 1 mrvr dithiothreitol, and 10 PM GDP or GTP, respectively; MATERIALS AND METHODS 1O-20 mg of the protein/ml gel was used. Two volumes of the dialyzed protein solution were Pyruvate kinase (EC 2.7. I .40) from rabbit added to 1 vol of CNBr-activated Sepharose muscle, phosphoenolpyruvate, GDP, GTP, 4B equilibrated with the same buffer, and the ATP, and unfractionated tRNA from E. coli coupling reaction was performed for 2 h at MRE 600 and from baker’s yeast were ob- room temperature with mixing. Care was tained from Boehringer (Mannheim, FRG). taken to avoid air bubbles in the reaction mixSepharose 4B activated with CNBr was a ture which could oxidize the protein. The gel commercial product from Pharmacia (Uppwas pelleted by centrifugation and incubated sala, Sweden) and cellulose filter discs, type 3 with a buffer of 200 IIIM glycine, pH 8.0, conMM, were obtained from Whatman (Maidtaining 300 I’IIM NaCl, 10 mM MgS04, 1 mM stone, U. K.). All radioactively labeled subdithiothreitol, and 10 pM GDP under the same stances came from Amersham-Buchler conditions as above to block residual reactive (Braunschweig, FRG). [14C]Threonine and groups of the matrix. After filling it into ap[‘4C]tyrosine had specific activities of 226 and propriate columns the matrix was washed al57 Ci/mol, respectively. For the measurement ternately with 100 mM sodium borate, pH of filter-bound radioactivity the scintillation 9.5, 500 mM NaCl, 10 mM MgS04, 1 mM fluid Rotiszint 11 from Roth (Karlsruhe, FRG) dithiothreitol, and 20 pM GDP buffer and with and the Beckman LS 7500 scintillation coun- the same buffer except that sodium borate was ter were used. EF-Tu . GDP from E. coli was replaced by 100 Tris/HCl, pH 6.9. For storage, isolated according to Leberman et al. (13) and the columns were equilibrated with 50 mM had a specific activity of 22,000 units/mg.j Hepes, pH 7.6, 50 mM NH,Cl, 50 IIIM KCl, EF-Tu . GDP from T. thermophilus was pu10 mM MgC12, 1 mM dithiothreitol, and 1 rified by a described procedure (14) up to a mM GTP (buffer A). The efficiency of the couspecific activity of 12,800 units/mg. For the pling reaction was calculated from the absorcatalysis of the aminoacylation reactions the bance of the supernatant according to EhresS 100 supernatant from E. coli, highly purified mann et al. (19). tyrosyl-tRNA synthetase (EC 6.1.1.1) from Conversion of EF-Tu * GDP to EFbaker’s yeast (sp act 15000 units/mg ( 15)), or Tu . GTP. Matrix-bound EF-Tu . GDP was the partially purified threonyl-tRNA synthe- converted to EF-Tu . GTP by the pyruvate kitase (EC 6.1.1.3) (sp act 2.2 units/mg) from nase activation procedure (12). Alternatively, bovine liver were used. The purification of the the immobilized EF-Tu. GDP from T. therliver enzyme was accomplished by the pro- mophilus could be converted to EF-Tu * GTP cedure of Roe ( 16) including a DEAE-cellulose by washing the column with buffer A for about step (17) and a gel filtration on Ultrogel AcA- 2 h at room temperature. 34. Bulk tRNA from bovine liver was isolated Deacylation of tRNA. When non-commeraccording to (16). cial bulk tRNA preparations were used, a deImmobilization of EF-Tu. Elongation fac- acylation step was introduced. Bulk tRNA was tors Tu - GDP or Tu - GTP from T. thermoincubated in 100 mM Tris/HCl, pH 7.5, 32.5 philus as well as from E. coli were coupled to mM CuS04 for 10 min at 37’C in order to hydrolyze the aminoacyl-tRNAs (20). The re3 One unit of EF-Tu is defined as the amount of protein which binds 1 pmol of GDP. moval of CuS04 could be accomplished by aminoacyl-tRNA is utilized. The method is generally applicable for the isolation of any given group of tRNA isoacceptors and does not require chemical modification of the aminoacyl-tRNA.
CHROMATOGRAPHIC
ISOLATION
several ethanol precipitations in the presence of EDTA. Aminoacylation oftMA. Bulk tRNA from E. coli, baker’s yeast, or bovine liver was aminoacylated in a reaction mixture containing 50 mM Hepes, pH 7.5, 50 mM KCl, 50 mM NH&l, 10 mM MgS04, 10 mM ATP, 5 mM dithiothreitol, lo-50 AZ60units4 of bulk tRNA, 15-22 PM 14C-labeled amino acid, and a sufficient amount of the homologous synthetase preparation in a total volume of 400 ~1. After incubation for 20 min at 37°C this reaction mixture was used without further purification for affinity chromatography. Ajjinity chromatography. A constant flow rate during the chromatography was maintained by a Varioperpex peristaltic pump from LKB (Bromma, Sweden). The affinity chromatography with immobilized EF-Tu from T. thermophilus was performed at room temperature, whereas in the case of the immobilized EF-Tu from E. coli it was done at 4°C because of the thermal instability of this protein (2 1). After application of the aminoacyltRNA sample, the ternary complex formation was allowed to take place during an incubation for 10 min at 30°C. If the concentration of total bulk tRNA did not exceed 300 PM, no efficient interaction of uncharged tRNA with matrix-bound EF-Tu . GTP could be detected. The column was washed with 5 ml of buffer B (10 mM Hepes, pH 7.6, 10 mM NH,Cl, 10 mM KCl, 10 mM MgCl*, 1 mM dithiothreitol, 50 PM GTP), 5 ml of buffer C (50 mM Hepes, pH 7.6, 150 mM KCl, 50 IrIM NH,Cl, 10 mM MgC&, 1 mM dithiothreitol, 50 PM GTP), and 8 ml of buffer D (100 mM sodium borate, pH 8.1, 1 M NaCl, 10 tnM MgC&, 1 mM dithiothreitol, 50 PM GTP). The elution profiles were monitored by determining the radioactivity precipitable in 5% aqueous trichloroacetic acid of appropriate aliquots. As an alternative to the high salt elution, matrix-bound aminoacyl4 One AzM) unit is the quantity of material contained in 1 ml of a solution, which has an absorbance of I at 260 nm, when measured in a I-cm-pathlength cell. 1 AzM) unit of tRNA corresponds to I .5 nmol.
OF tRNA ISOACCEPTORS
163
tRNA can be specifically eluted by 1 ml of buffer A containing 45 PM soluble EFTu. GTP from E. coli, which competes with the immobilized elongation factor for the aminoacyl-tRNAs ( 12). Two-dimensional gel electrophoresis. For electrophoretic analysis appropriate fractions from the affinity column eluate containing 25200 pmol isoaccepting tRNA were desalted by dialysis against water, concentrated, and subjected to electrophoresis. The conditions of the two-dimensional gel electrophoresis were essentially those of Fradin et al. (22) except that the electric field strength was 13.5 V/cm in the first and 17.5 V/cm in the second dimensional runs. The RNA was detected after staining in “Stains all,” Serva (Heidelberg, FRG) (23). RESULTS
Elongation factor Tu from E. coli or T. thermophilus was immobilized on Sepharose 4B. About 95% of the T. thermophilus protein and about 85% of the E. coli protein could be covalently coupled to the matrix. The immobilized elongation factors have the same specificity in ligand binding as the soluble proteins (12). For the binding of aminoacyltRNA the elongation factor Tu * GDP has to be converted to Tu - GTP. This activation can be achieved either by incubation with pyruvate kinase in the presence of phosphoenolpyruvate and GTP or, alternatively, by a direct exchange of GDP for GTP. In the case of the EF-Tu from T. thermophilus the direct exchange is the more convenient procedure, since the dissociation constants of both nucleotides are similar (24). On the other hand EF-Tu from E. coli is known to have a considerably higher affinity for GDP as compared to GTP (21), thus the pyruvate kinase activation is more suitable here. Aminoacyl-tRNAs bind to EF-Tu * GTP with dissociation constants varying between 10e6 and lo-’ M. The interaction of uncharged tRNA with EF-Tu . GTP is much less efficient, KD > 10e4 M (2 1). Thus these large differences
164
DERWENSKUS,
FISCHER,
in the affinity can be used for a selective binding to immobilized EF-Tu - GTP of aminoacyl-tRNAs present in a mixture of nonfractionated tRNA. In Fig. 1 a typical elution profile of an affinity chromatography on immobilized EF-Tu . GTP from T. thermophilus is shown. In this experiment bulk tRNA from bovine liver has been aminoacylated in the presence of [‘4C]threonine. [ 14C]ThreonyltRNATh’ (59%) was retained on the column and could be eluted with the buffer of high ionic strength. In order to eliminate interactions of uncharged tRNA with the matrix an elution with a buffer of intermediate ionic strength was always included in the washing procedure. Usually during this step a small portion of the aminoacyl-tRNA was also eluted. The purification of tRNAs can be demonstrated by two-dimensional gel electrophoretie analysis of appropriate fractions. The tRNA isoacceptor patterns obtained by affinity chromatography of threonyl-tRNAsn” present in bulk tRNA of E. coli and bovine liver are shown in Figs. 2a and 2c, respectively. The isolation of tyrosyl-tRNAsTY from bulk tRNA
FIG. 1. Isolation of Thr-tRNAn” isoacceptors from bovine liver bulk tRNA on a column of immobilized EFTueGTP from T. thermophilus. Unfmctionated tRNA from bovine liver was aminoacylated with [%]threonine as described under Materials and Methods. The aminoacylation mixture containing 5 10 pmol [“C]Thr-tRNAn” was applied onto a column of 0.75 ml corresponding to 13.1 mg immobilized EF-Tu .GTP (sp act 2600 U/m&. From the [‘4c]Thr-tRNAn” eluting from the column, 18% came with buffer B, 23% with buffer C, and 59% with buffer D.
AND
SPRINZL
of yeast is shown in Fig. 2b. The comparison with the corresponding bulk tRNA patterns (left) clearly demonstrates the high efficiency of the method. Both immobilized elongation factors from E. coli (Fig. 2b) and from T. thermophilus (Figs. 2a, 2c) were successfully employed for the isolation of specific tRNAs. In more than 50 different experiments with both immobilized elongation factors, a high specificity with respect to aminoacyl-tRNA binding could be demonstrated independent on the source of the bulk tRNA or the nature of the amino acid. In control experiments where either the amino acid or the aminoacyl-tRNA synthetase were omitted in the aminoacylation mixture, subsequent affinity chromatography did not result in efficient accumulation of tRNA species on the column. The activities of the matrix-bound elongation factors were calculated from their capacity to bind aminoacyl-tRNA released by buffer D. From the amount of soluble protein which was active in the GDP-exchange reaction and initially used for the immobilization, about 1% of the E. coli EF-Tu and up to 5% of the T. thermophilus EF-Tu were active in aminoacyl-tRNA binding in the matrixbound state. The immobilized elongation factor from T. thermophilus could be used at room temperature for more than 2 years without any detectable loss of activity whereas the immobilized E. coli EF-Tu showed a decline in the binding capacity for aminoacylated tRNA after about 6 months, even when operated at 4°C. DISCUSSION
Affinity chromatography of aminoacyltRNA on immobilized bacterial EF-Tu * GTP represents a method for rapid and efficient isolation of isoaccepting tRNAs. The method is based on the specificity of the aminoacylation reaction combined with the selectivity of EF-Tu . GTP for the binding of aminoacyltRNAs. In contrast to previously described
CHROMATOCRAPHIC
ISOLATION
OF tRNA
ISOACCEPTORS
165
first dimension I
FIG. 2. Analysis of the affinity chromatographically isolated tRNA isoacceptors by two-dimensional polyacrylamide gel electrophoresis. tRNAs eluted with buffer D from the affinity columns (right), corresponding to bulk tRNA (left). (a) Thr-tRNAn” from E. coli bulk tRNA, (b) Tyr-tRNATF from yeast bulk tRNA; (c) Thr-tRNATh’ from bovine liver bulk tRNA. The purification was achieved on immobilized EF-Tu from T. thermophilus (a) and (c) or from E. coli (b).
procedures, additional chemical modification steps, which may cause undesirable inactivation of tRNAs and side reactions (6,7) are not required. A crucial step in the described isolation protocol is the aminoacylation reaction. The presence of undesired amino acids in the crude synthetase preparations or in the commercial amino acids can be misleading. Especially by employing crude synthetase preparations from mammalian sources, nondialyzable, synthe-
tase-bound amino acids may cause failures. This can be circumvented by determination of the minimal amount of synthetase needed for catalysis of the aminoacylation reaction and employment of a large excess of the desired amino acid. Conditions of the aminoacylation reaction favoring misaminoacylation should also be avoided. On the other hand, this affinity chromatography can be a potentially useful method for the investigation of the conditions leading to misaminoacylation.
166
DERWENSKUS,
FISCHER, AND SPRINZL
If the aminoacylation is performed with a crude aminoacyl-tRNA synthetase or with a S 100 supernatant, a deproteinization step should be included after aminoacylation in order to avoid a competition of the soluble EF-Tu. GTP from the crude enzyme preparation with the immobilized one. Furthermore, the proteases present in crude enzymes and capable of degrading the bound EFTu. GTP will be removed. For a successful isolation of aminoacyltRNA from a large excess of uncharged tRNA by the described method the selective binding of the aminoacylated tRNA by immobilized EF-Tu * GTP is the second crucial step. Since the differences in the dissociation constants between tRNA and aminoacyl-tRNA for this interaction are dependent on ionic strength (25), the choice of the salt concentrations, where these differences are maximal, is important. It proved to be useful to form ternary complexes at low ionic strength, although at these conditions some non-aminoacylated tRNA remained bound to the column also. This uncharged tRNA could be completely removed by the subsequent elution with a buffer of intermediate ionic strength (buffer C). The purification effect of this step compensates the disadvantage that a small portion of the aminoacylated tRNAs is usually also eluted with buffer C (Fig. 1). The analysis of the fractions eluted with buffer C revealed that some uncharged tRNA species were retarded more efficiently than others. On the other hand, in some cases particular aminoacyltRNAs from an isoacceptor set are bound less efficiently as compared to other isoacceptors. Therefore, the isoacceptor pattern obtained from the analysis of the high salt portion of the eluate does not necessarily reflect their quantitative distribution in the unfractionated tRNA. As an alternative to the high salt elution, the release of matrix-bound aminoacyltRNA could be accomplished employing soluble EF-Tu . GTP in buffer A, which competes with the immobilized factor for charged tRNA. As shown in Fig. 2 the affinity chromatography on matrix-bound bacterial elongation
factor Tu was successfully used for the isolation of specific tRNAs from prokaryotes (E. co/i, Fig. 2a) as well as from eukaryotes (yeast, bovine liver, Figs. 2b, 2~). Satisfactory results have also been obtained isolating specific tRNA from the slime mold Dictyostelium discoideum (W. Bertling, Erlangen, personal communication) and from bovine lymphocytes, thus manifesting the generality of this technique. Another application of the method is the determination of the extent of charged tRNAs in crude cellular extracts. No other direct method for this purpose is available at the present time. Furthermore, in our laboratory the described affinity technique was employed for the deprivation of bulk tRNA for one group of specific isoacceptors. ACKNOWLEDGMENTS We thank Dr. S. Chladek for comments to the manuscript, Dr. H. G. Faulhammer and Dr. K. Gulewicz for the gifts of purified proteins, and Mrs. B. Wagner for technical assistance. This work was supported by the Deutsche Forschungsgemeinschaft (Sp 243/1-l). REFERENCES 1. Gillam, I., Millward, S., Blew, D., van Tigerstrom, M., Wimmer, E., and Tener, G. M. (1967) Biochemistry 6, 3043-3056. 2. Holmes, W. M., Hurd, R. E., Reid, B. R., Rimerman, R. A., and Hatfield, G. W. ( 1975) Proc. Natl. Acad. Sci. USA 72, 1068-1071. 3. Pearson, R. L., Weiss, J. F., and Kelmers, A. D. ( 197 1) Biochim. Biophys. Acta 228, 770-774. 4. Gillam, I., Blew, D., Warrington, R. C., von TigerStrom, M., and Tener, G. M. (1968) Biochemistry 7,3459-3468.
5. Goss, D. J., and Parkhurst, L. J. (1978) J. Biol. Chem. 253, 7804-7806. 6.
7. 8. 9. 10.
Roy, K. L., Bloom, A., and Sdll, D. ( 197 1) in Procedures in Nucleic Acid Research (Cantoni, G. L., and Davies, D. R., eds.), Vol. 2, pp. 524-541, Harper and Row, New York. Friedmann, S. (1972) Biochemistry 11, 3435-3443. Zekavat, S. Y., James, H. L., Morrison, J. C., Whybrew, W. D., and Bucovaz, E. T. (1972) Biochemistry 11, 4344-4347. Denburg, J., and DeLuca, M. (1970) Proc. Nat/. Acad. Sci. USA 67, 1057-1062. Maxwell, I. H., Wimmer, E., and Tener, G. M. (1968) Biochemistry 7, 2629-2634.
CHROMATOGRAPHIC
ISOLATION
11. MC Cutchan, T. F., Gilham, P. T., and Soll, D. (1975) Nucl. Acids
Res. 2, 853-864.
12. Fischer, W., Derwenskus, K.-H., and Sprinzl, M. (1982) Eur. J. Biochem. 125, 143-149. 13. Leberman, R., Antonsson, B., Giovanelli, R., Guariguana, R., Schumann, R., and Wittinghofer, A. (1980) Anal. B&hem. 104, 29-36. 14. Gulewicz, K., Faulhammer, H. G., and Sprinzl, M. (1981) Eur. J. Biochem. 121, 155-162. 15. Faulhammer, H. G., and Cramer, F. (1977) Eur. J. Biochem. 15, 56 l-570. 16. Roe, B. A. (1975) Nucl. Acids Rex 2, 21-42. 17. Kisselev, L. L., Favorova, 0. O., and Kovaleva, G. K. (1979) in Methods in Enzymology (Moldave, K., and Grossman, L., eds.), Vol. 59, pp. 234-257, Academic Press, New York. 18. Pharmacia Fine Chemicals (1979) Affinity Chromatography, Principles and Methods.
OF tRNA ISOACCEPTORS
167
19. Ehresmann, B., Imbault, P., and Weil. J. H. (1973) Anal.
Biochem.
54, 454-463.
20. Schofield, P., and Zamecnik, P. C. (1968) Biochim. Biophys. Acta 155, 410-416. 21. Miller, D. L., and Weissbach, H. (1977) in Molecular Mechanisms of Protein Biosynthesis (Weissbach, H., and Pestka, S., eds.), pp. 323-411, Academic Press, New York. 22. Fradin, A., Gruhl, H., and Feldmann, H. (1975) FEBS Lett. 50, 185-189. 23. Dahlberg, A. E., Dingmann, C. W., and Peacock, A. C. (1969) J. Mol. Biol. 41, 139-147. 24. Arai, K., Arai, N., Nakamura, S., Oshima, T., and Kaziro, Y. (1978) Eur. J. Biochem. 92, 521-531. 25. Pingoud, A., Urbanke, C., Krauss, G., Peters, F., and Maass, G. (1977) Eur. J. Biochem. 78, 403-409.
BIOCHEMISTRY
136, 161-167 (1984)
Isolation of tRNA lsoacceptors by Affinity Chromatography on Immobilized Bacterial Elongation Factor Tu’ KARL-HEINZDERWENSKUS,WOLFGANGFISCHER,ANDMATHIASSPRINZL Lehrstuhl
ftir
Biochemie der Universittit D-8580 Bayreuth, Federal
Bayreuth, Universitiitsstrasse Republic of Germany
30,
Received April 27, 1983 Elongation factor Tu from Escherichia coli or Thermus thermophilus was immobilized on cyanogen bromide-activated Sepharose. Immobilized elongation factor Tu . GDP could be converted to Tu. GTP, which is able to bind aminoacyl-tRNA. When bulk tRNA from E. coli, baker’s yeast, or bovine liver was aminoacylated by one amino acid only, the resulting aminoacyltRNA could be separated in one step from the rest of the tRNA using an affinity column of immobilized elongation factor Tu . GTP. Specific tRNA isoacceptors can be isolated in amounts sufficient for gel electrophoretic analysis, sequence determinations, and hybridization experiments. KEY WORDS: tRNA isoacceptors; aminoacyl-tRNA; immobilized protein; affinity chromatography; elongation factor Tu; two-dimensional gel electrophoresis.
Isolation of pure tRNA species for sequence analysis and other biochemical investigations is still tedious. Besides conventional column chromatographic procedures for the purification of non-aminoacylated tRNA species ( l3) various techniques for the isolation of aminoacyl-tRNA isoacceptors have been developed. These techniques mostly rely on the specificity of the enzymatic aminoacylation reaction combined with chemical derivatization of the free a-amino group of the attached amino acid. If aminoacylation of unfractionated tRNA is performed in the presence of one amino acid only, the corresponding tRNA isoacceptors will provide free a-amino groups for subsequent modifications. Gillam et al. modified such aminoacyl-tRNAs by introducing aromatic moieties and isolated the modified aminoacyl-tRNAs by chromatography on benzoylated DEAEcellulose (4). Another method uses a bifunctional reagent to attach the acylated tRNA species via their a-amino groups to a sulfhydryl-substituted Sepharose 4B (5). Variable ’ This publication is dedicated to the 60th birthday of Prof. Dr. F. Cramer.
yields and side reactions of the chemical derivatization are the main disadvantages of these methods (6,7). Separation of the aminoacyl-tRNAs from uncharged tRNA without further derivatization was also reported. The majority of these methods are confined to the isolation of only one (8,9) or a few ( 10) particular isoacceptor sets and are lacking general applicability. The method using affinity columns containing dihydroxyboryl residues (11) is limited by the fact that the borate complex at the cis-diol groups of the uncharged tRNA is unstable, especially at neutral and acidic conditions necessary to preserve an integrity of the aminoacyl-tRNA ester bond. Recently, we reported on the l&and-binding properties of the immobilized elongation factor Tu from Thermus thermophilus (12). In the present work we describe the use of the immobilized EF-Tu2 from T. thermophilus and from Escherichia coli for the isolation of isoaccepting tRNAs. For this purpose the natural high affinity of bacterial EF-Tu . GTP for * Abbreviations used: EF-Tu, elongation factor Tu; Hepes, 4-(2-Hydroxyethyl)-l-piperazineethanesulfonic acid. 161
0003-2697184 $3.00 Copyright Q 1984 by Academic Press Inc. All rights of reproduction in any form reserved.
162
DERWENSKUS,
FISCHER, AND SPRINZL
Sepharose 4B using the procedure suggested by the manufacturer (18). EF-Tu . GDP or EFTu . GTP was dialyzed against a buffer containing 100 mM sodium borate, pH 8.3, 500 mM NaCI, 10 mM MgS04, 1 mrvr dithiothreitol, and 10 PM GDP or GTP, respectively; MATERIALS AND METHODS 1O-20 mg of the protein/ml gel was used. Two volumes of the dialyzed protein solution were Pyruvate kinase (EC 2.7. I .40) from rabbit added to 1 vol of CNBr-activated Sepharose muscle, phosphoenolpyruvate, GDP, GTP, 4B equilibrated with the same buffer, and the ATP, and unfractionated tRNA from E. coli coupling reaction was performed for 2 h at MRE 600 and from baker’s yeast were ob- room temperature with mixing. Care was tained from Boehringer (Mannheim, FRG). taken to avoid air bubbles in the reaction mixSepharose 4B activated with CNBr was a ture which could oxidize the protein. The gel commercial product from Pharmacia (Uppwas pelleted by centrifugation and incubated sala, Sweden) and cellulose filter discs, type 3 with a buffer of 200 IIIM glycine, pH 8.0, conMM, were obtained from Whatman (Maidtaining 300 I’IIM NaCl, 10 mM MgS04, 1 mM stone, U. K.). All radioactively labeled subdithiothreitol, and 10 pM GDP under the same stances came from Amersham-Buchler conditions as above to block residual reactive (Braunschweig, FRG). [14C]Threonine and groups of the matrix. After filling it into ap[‘4C]tyrosine had specific activities of 226 and propriate columns the matrix was washed al57 Ci/mol, respectively. For the measurement ternately with 100 mM sodium borate, pH of filter-bound radioactivity the scintillation 9.5, 500 mM NaCl, 10 mM MgS04, 1 mM fluid Rotiszint 11 from Roth (Karlsruhe, FRG) dithiothreitol, and 20 pM GDP buffer and with and the Beckman LS 7500 scintillation coun- the same buffer except that sodium borate was ter were used. EF-Tu . GDP from E. coli was replaced by 100 Tris/HCl, pH 6.9. For storage, isolated according to Leberman et al. (13) and the columns were equilibrated with 50 mM had a specific activity of 22,000 units/mg.j Hepes, pH 7.6, 50 mM NH,Cl, 50 IIIM KCl, EF-Tu . GDP from T. thermophilus was pu10 mM MgC12, 1 mM dithiothreitol, and 1 rified by a described procedure (14) up to a mM GTP (buffer A). The efficiency of the couspecific activity of 12,800 units/mg. For the pling reaction was calculated from the absorcatalysis of the aminoacylation reactions the bance of the supernatant according to EhresS 100 supernatant from E. coli, highly purified mann et al. (19). tyrosyl-tRNA synthetase (EC 6.1.1.1) from Conversion of EF-Tu * GDP to EFbaker’s yeast (sp act 15000 units/mg ( 15)), or Tu . GTP. Matrix-bound EF-Tu . GDP was the partially purified threonyl-tRNA synthe- converted to EF-Tu . GTP by the pyruvate kitase (EC 6.1.1.3) (sp act 2.2 units/mg) from nase activation procedure (12). Alternatively, bovine liver were used. The purification of the the immobilized EF-Tu. GDP from T. therliver enzyme was accomplished by the pro- mophilus could be converted to EF-Tu * GTP cedure of Roe ( 16) including a DEAE-cellulose by washing the column with buffer A for about step (17) and a gel filtration on Ultrogel AcA- 2 h at room temperature. 34. Bulk tRNA from bovine liver was isolated Deacylation of tRNA. When non-commeraccording to (16). cial bulk tRNA preparations were used, a deImmobilization of EF-Tu. Elongation fac- acylation step was introduced. Bulk tRNA was tors Tu - GDP or Tu - GTP from T. thermoincubated in 100 mM Tris/HCl, pH 7.5, 32.5 philus as well as from E. coli were coupled to mM CuS04 for 10 min at 37’C in order to hydrolyze the aminoacyl-tRNAs (20). The re3 One unit of EF-Tu is defined as the amount of protein which binds 1 pmol of GDP. moval of CuS04 could be accomplished by aminoacyl-tRNA is utilized. The method is generally applicable for the isolation of any given group of tRNA isoacceptors and does not require chemical modification of the aminoacyl-tRNA.
CHROMATOGRAPHIC
ISOLATION
several ethanol precipitations in the presence of EDTA. Aminoacylation oftMA. Bulk tRNA from E. coli, baker’s yeast, or bovine liver was aminoacylated in a reaction mixture containing 50 mM Hepes, pH 7.5, 50 mM KCl, 50 mM NH&l, 10 mM MgS04, 10 mM ATP, 5 mM dithiothreitol, lo-50 AZ60units4 of bulk tRNA, 15-22 PM 14C-labeled amino acid, and a sufficient amount of the homologous synthetase preparation in a total volume of 400 ~1. After incubation for 20 min at 37°C this reaction mixture was used without further purification for affinity chromatography. Ajjinity chromatography. A constant flow rate during the chromatography was maintained by a Varioperpex peristaltic pump from LKB (Bromma, Sweden). The affinity chromatography with immobilized EF-Tu from T. thermophilus was performed at room temperature, whereas in the case of the immobilized EF-Tu from E. coli it was done at 4°C because of the thermal instability of this protein (2 1). After application of the aminoacyltRNA sample, the ternary complex formation was allowed to take place during an incubation for 10 min at 30°C. If the concentration of total bulk tRNA did not exceed 300 PM, no efficient interaction of uncharged tRNA with matrix-bound EF-Tu . GTP could be detected. The column was washed with 5 ml of buffer B (10 mM Hepes, pH 7.6, 10 mM NH,Cl, 10 mM KCl, 10 mM MgCl*, 1 mM dithiothreitol, 50 PM GTP), 5 ml of buffer C (50 mM Hepes, pH 7.6, 150 mM KCl, 50 IrIM NH,Cl, 10 mM MgC&, 1 mM dithiothreitol, 50 PM GTP), and 8 ml of buffer D (100 mM sodium borate, pH 8.1, 1 M NaCl, 10 tnM MgC&, 1 mM dithiothreitol, 50 PM GTP). The elution profiles were monitored by determining the radioactivity precipitable in 5% aqueous trichloroacetic acid of appropriate aliquots. As an alternative to the high salt elution, matrix-bound aminoacyl4 One AzM) unit is the quantity of material contained in 1 ml of a solution, which has an absorbance of I at 260 nm, when measured in a I-cm-pathlength cell. 1 AzM) unit of tRNA corresponds to I .5 nmol.
OF tRNA ISOACCEPTORS
163
tRNA can be specifically eluted by 1 ml of buffer A containing 45 PM soluble EFTu. GTP from E. coli, which competes with the immobilized elongation factor for the aminoacyl-tRNAs ( 12). Two-dimensional gel electrophoresis. For electrophoretic analysis appropriate fractions from the affinity column eluate containing 25200 pmol isoaccepting tRNA were desalted by dialysis against water, concentrated, and subjected to electrophoresis. The conditions of the two-dimensional gel electrophoresis were essentially those of Fradin et al. (22) except that the electric field strength was 13.5 V/cm in the first and 17.5 V/cm in the second dimensional runs. The RNA was detected after staining in “Stains all,” Serva (Heidelberg, FRG) (23). RESULTS
Elongation factor Tu from E. coli or T. thermophilus was immobilized on Sepharose 4B. About 95% of the T. thermophilus protein and about 85% of the E. coli protein could be covalently coupled to the matrix. The immobilized elongation factors have the same specificity in ligand binding as the soluble proteins (12). For the binding of aminoacyltRNA the elongation factor Tu * GDP has to be converted to Tu - GTP. This activation can be achieved either by incubation with pyruvate kinase in the presence of phosphoenolpyruvate and GTP or, alternatively, by a direct exchange of GDP for GTP. In the case of the EF-Tu from T. thermophilus the direct exchange is the more convenient procedure, since the dissociation constants of both nucleotides are similar (24). On the other hand EF-Tu from E. coli is known to have a considerably higher affinity for GDP as compared to GTP (21), thus the pyruvate kinase activation is more suitable here. Aminoacyl-tRNAs bind to EF-Tu * GTP with dissociation constants varying between 10e6 and lo-’ M. The interaction of uncharged tRNA with EF-Tu . GTP is much less efficient, KD > 10e4 M (2 1). Thus these large differences
164
DERWENSKUS,
FISCHER,
in the affinity can be used for a selective binding to immobilized EF-Tu - GTP of aminoacyl-tRNAs present in a mixture of nonfractionated tRNA. In Fig. 1 a typical elution profile of an affinity chromatography on immobilized EF-Tu . GTP from T. thermophilus is shown. In this experiment bulk tRNA from bovine liver has been aminoacylated in the presence of [‘4C]threonine. [ 14C]ThreonyltRNATh’ (59%) was retained on the column and could be eluted with the buffer of high ionic strength. In order to eliminate interactions of uncharged tRNA with the matrix an elution with a buffer of intermediate ionic strength was always included in the washing procedure. Usually during this step a small portion of the aminoacyl-tRNA was also eluted. The purification of tRNAs can be demonstrated by two-dimensional gel electrophoretie analysis of appropriate fractions. The tRNA isoacceptor patterns obtained by affinity chromatography of threonyl-tRNAsn” present in bulk tRNA of E. coli and bovine liver are shown in Figs. 2a and 2c, respectively. The isolation of tyrosyl-tRNAsTY from bulk tRNA
FIG. 1. Isolation of Thr-tRNAn” isoacceptors from bovine liver bulk tRNA on a column of immobilized EFTueGTP from T. thermophilus. Unfmctionated tRNA from bovine liver was aminoacylated with [%]threonine as described under Materials and Methods. The aminoacylation mixture containing 5 10 pmol [“C]Thr-tRNAn” was applied onto a column of 0.75 ml corresponding to 13.1 mg immobilized EF-Tu .GTP (sp act 2600 U/m&. From the [‘4c]Thr-tRNAn” eluting from the column, 18% came with buffer B, 23% with buffer C, and 59% with buffer D.
AND
SPRINZL
of yeast is shown in Fig. 2b. The comparison with the corresponding bulk tRNA patterns (left) clearly demonstrates the high efficiency of the method. Both immobilized elongation factors from E. coli (Fig. 2b) and from T. thermophilus (Figs. 2a, 2c) were successfully employed for the isolation of specific tRNAs. In more than 50 different experiments with both immobilized elongation factors, a high specificity with respect to aminoacyl-tRNA binding could be demonstrated independent on the source of the bulk tRNA or the nature of the amino acid. In control experiments where either the amino acid or the aminoacyl-tRNA synthetase were omitted in the aminoacylation mixture, subsequent affinity chromatography did not result in efficient accumulation of tRNA species on the column. The activities of the matrix-bound elongation factors were calculated from their capacity to bind aminoacyl-tRNA released by buffer D. From the amount of soluble protein which was active in the GDP-exchange reaction and initially used for the immobilization, about 1% of the E. coli EF-Tu and up to 5% of the T. thermophilus EF-Tu were active in aminoacyl-tRNA binding in the matrixbound state. The immobilized elongation factor from T. thermophilus could be used at room temperature for more than 2 years without any detectable loss of activity whereas the immobilized E. coli EF-Tu showed a decline in the binding capacity for aminoacylated tRNA after about 6 months, even when operated at 4°C. DISCUSSION
Affinity chromatography of aminoacyltRNA on immobilized bacterial EF-Tu * GTP represents a method for rapid and efficient isolation of isoaccepting tRNAs. The method is based on the specificity of the aminoacylation reaction combined with the selectivity of EF-Tu . GTP for the binding of aminoacyltRNAs. In contrast to previously described
CHROMATOCRAPHIC
ISOLATION
OF tRNA
ISOACCEPTORS
165
first dimension I
FIG. 2. Analysis of the affinity chromatographically isolated tRNA isoacceptors by two-dimensional polyacrylamide gel electrophoresis. tRNAs eluted with buffer D from the affinity columns (right), corresponding to bulk tRNA (left). (a) Thr-tRNAn” from E. coli bulk tRNA, (b) Tyr-tRNATF from yeast bulk tRNA; (c) Thr-tRNATh’ from bovine liver bulk tRNA. The purification was achieved on immobilized EF-Tu from T. thermophilus (a) and (c) or from E. coli (b).
procedures, additional chemical modification steps, which may cause undesirable inactivation of tRNAs and side reactions (6,7) are not required. A crucial step in the described isolation protocol is the aminoacylation reaction. The presence of undesired amino acids in the crude synthetase preparations or in the commercial amino acids can be misleading. Especially by employing crude synthetase preparations from mammalian sources, nondialyzable, synthe-
tase-bound amino acids may cause failures. This can be circumvented by determination of the minimal amount of synthetase needed for catalysis of the aminoacylation reaction and employment of a large excess of the desired amino acid. Conditions of the aminoacylation reaction favoring misaminoacylation should also be avoided. On the other hand, this affinity chromatography can be a potentially useful method for the investigation of the conditions leading to misaminoacylation.
166
DERWENSKUS,
FISCHER, AND SPRINZL
If the aminoacylation is performed with a crude aminoacyl-tRNA synthetase or with a S 100 supernatant, a deproteinization step should be included after aminoacylation in order to avoid a competition of the soluble EF-Tu. GTP from the crude enzyme preparation with the immobilized one. Furthermore, the proteases present in crude enzymes and capable of degrading the bound EFTu. GTP will be removed. For a successful isolation of aminoacyltRNA from a large excess of uncharged tRNA by the described method the selective binding of the aminoacylated tRNA by immobilized EF-Tu * GTP is the second crucial step. Since the differences in the dissociation constants between tRNA and aminoacyl-tRNA for this interaction are dependent on ionic strength (25), the choice of the salt concentrations, where these differences are maximal, is important. It proved to be useful to form ternary complexes at low ionic strength, although at these conditions some non-aminoacylated tRNA remained bound to the column also. This uncharged tRNA could be completely removed by the subsequent elution with a buffer of intermediate ionic strength (buffer C). The purification effect of this step compensates the disadvantage that a small portion of the aminoacylated tRNAs is usually also eluted with buffer C (Fig. 1). The analysis of the fractions eluted with buffer C revealed that some uncharged tRNA species were retarded more efficiently than others. On the other hand, in some cases particular aminoacyltRNAs from an isoacceptor set are bound less efficiently as compared to other isoacceptors. Therefore, the isoacceptor pattern obtained from the analysis of the high salt portion of the eluate does not necessarily reflect their quantitative distribution in the unfractionated tRNA. As an alternative to the high salt elution, the release of matrix-bound aminoacyltRNA could be accomplished employing soluble EF-Tu . GTP in buffer A, which competes with the immobilized factor for charged tRNA. As shown in Fig. 2 the affinity chromatography on matrix-bound bacterial elongation
factor Tu was successfully used for the isolation of specific tRNAs from prokaryotes (E. co/i, Fig. 2a) as well as from eukaryotes (yeast, bovine liver, Figs. 2b, 2~). Satisfactory results have also been obtained isolating specific tRNA from the slime mold Dictyostelium discoideum (W. Bertling, Erlangen, personal communication) and from bovine lymphocytes, thus manifesting the generality of this technique. Another application of the method is the determination of the extent of charged tRNAs in crude cellular extracts. No other direct method for this purpose is available at the present time. Furthermore, in our laboratory the described affinity technique was employed for the deprivation of bulk tRNA for one group of specific isoacceptors. ACKNOWLEDGMENTS We thank Dr. S. Chladek for comments to the manuscript, Dr. H. G. Faulhammer and Dr. K. Gulewicz for the gifts of purified proteins, and Mrs. B. Wagner for technical assistance. This work was supported by the Deutsche Forschungsgemeinschaft (Sp 243/1-l). REFERENCES 1. Gillam, I., Millward, S., Blew, D., van Tigerstrom, M., Wimmer, E., and Tener, G. M. (1967) Biochemistry 6, 3043-3056. 2. Holmes, W. M., Hurd, R. E., Reid, B. R., Rimerman, R. A., and Hatfield, G. W. ( 1975) Proc. Natl. Acad. Sci. USA 72, 1068-1071. 3. Pearson, R. L., Weiss, J. F., and Kelmers, A. D. ( 197 1) Biochim. Biophys. Acta 228, 770-774. 4. Gillam, I., Blew, D., Warrington, R. C., von TigerStrom, M., and Tener, G. M. (1968) Biochemistry 7,3459-3468.
5. Goss, D. J., and Parkhurst, L. J. (1978) J. Biol. Chem. 253, 7804-7806. 6.
7. 8. 9. 10.
Roy, K. L., Bloom, A., and Sdll, D. ( 197 1) in Procedures in Nucleic Acid Research (Cantoni, G. L., and Davies, D. R., eds.), Vol. 2, pp. 524-541, Harper and Row, New York. Friedmann, S. (1972) Biochemistry 11, 3435-3443. Zekavat, S. Y., James, H. L., Morrison, J. C., Whybrew, W. D., and Bucovaz, E. T. (1972) Biochemistry 11, 4344-4347. Denburg, J., and DeLuca, M. (1970) Proc. Nat/. Acad. Sci. USA 67, 1057-1062. Maxwell, I. H., Wimmer, E., and Tener, G. M. (1968) Biochemistry 7, 2629-2634.
CHROMATOGRAPHIC
ISOLATION
11. MC Cutchan, T. F., Gilham, P. T., and Soll, D. (1975) Nucl. Acids
Res. 2, 853-864.
12. Fischer, W., Derwenskus, K.-H., and Sprinzl, M. (1982) Eur. J. Biochem. 125, 143-149. 13. Leberman, R., Antonsson, B., Giovanelli, R., Guariguana, R., Schumann, R., and Wittinghofer, A. (1980) Anal. B&hem. 104, 29-36. 14. Gulewicz, K., Faulhammer, H. G., and Sprinzl, M. (1981) Eur. J. Biochem. 121, 155-162. 15. Faulhammer, H. G., and Cramer, F. (1977) Eur. J. Biochem. 15, 56 l-570. 16. Roe, B. A. (1975) Nucl. Acids Rex 2, 21-42. 17. Kisselev, L. L., Favorova, 0. O., and Kovaleva, G. K. (1979) in Methods in Enzymology (Moldave, K., and Grossman, L., eds.), Vol. 59, pp. 234-257, Academic Press, New York. 18. Pharmacia Fine Chemicals (1979) Affinity Chromatography, Principles and Methods.
OF tRNA ISOACCEPTORS
167
19. Ehresmann, B., Imbault, P., and Weil. J. H. (1973) Anal.
Biochem.
54, 454-463.
20. Schofield, P., and Zamecnik, P. C. (1968) Biochim. Biophys. Acta 155, 410-416. 21. Miller, D. L., and Weissbach, H. (1977) in Molecular Mechanisms of Protein Biosynthesis (Weissbach, H., and Pestka, S., eds.), pp. 323-411, Academic Press, New York. 22. Fradin, A., Gruhl, H., and Feldmann, H. (1975) FEBS Lett. 50, 185-189. 23. Dahlberg, A. E., Dingmann, C. W., and Peacock, A. C. (1969) J. Mol. Biol. 41, 139-147. 24. Arai, K., Arai, N., Nakamura, S., Oshima, T., and Kaziro, Y. (1978) Eur. J. Biochem. 92, 521-531. 25. Pingoud, A., Urbanke, C., Krauss, G., Peters, F., and Maass, G. (1977) Eur. J. Biochem. 78, 403-409.