(Received for publication, August 16, 1979, and in revised form, April 7, 1980)

Transcription

1 THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 255, No. 15. Issue of Aumst IO, pp , 19Ho Prlnted m U.S.A. The Coupling with Polypeptide Synthesis of the GTPase Activity Dependent on Elongation Factor G* (Received for publication, August 16, 1979, and in revised form, April 7, 1980) Gianni ChinaliS and Andrea Parmeggianig From the Gesellschaft fur Molekularbiologische Forschung, Abteilung Biochemie, Braunschweig-Stoeckheim, Federal Republic of Germany, Cattedra di Chimica, II" Facolta di Medicina e Chirurgia, Universita di Napoli, I NaDoli, Italy. and Laboratoire de Biochimie. Laboratoire Associe No. 240 du Centre National de la Recherche Scientifipue, Ecke Polytechnique, F Palaiseau Cedex, France The coupling with polypeptide synthesis of the ribo- formed peptidyl-trna from the A site to the peptidyl site (P some-elongation factor G (EF-G)-dependent GTPase ac- site). In both cases, the hydrolysis of GTP is involved in the tivity was studied in a highly purified system with well characterized NH&l-washed ribosomes which were from 55 to 67% active in poly(u)-directed polyphenylalanine synthesis. The lowest stoichiometries of total GTP hydrolysis to polyphenylalanine incorporation (2.4 to 2.8) were observed at concentrations of MgC1, (4 to 6.5 m) slightly lower than the M&' optimum for polyphenylalanine synthesis (7 to 8 m~), in a system containing 80 m~ NKCl or KCl. For minimal stoichiometry, the concentration of EF-G should be rate-limiting, whereas that of EF-T (EF-Tu-EF-Ts) and amino- acyl-trna should be in excess, since the coupling of the EF-GGTPase activity depends on ribosomes in pretranslocative state. Under this condition, the apparent K,,, values for GTP of GTPase activity and polyphenylalanine synthesis are identical, and they are about an order of magnitude lower than the K,,, of the ribosome- EF-G-dependent GTPase activity uncoupled from polypeptide synthesis. The stoichiometry was calculated without the usual correction for GTP hydrolysis obtained in the same system lacking elongation factor T or aminoacyl-trna. Such a correction causes overestimation of the uncoupled EF-G GTPase activity still present in the complete system, leading to artificially low stoichiometric values. During elongation of the polypeptide chain, the interaction between the ribosome and the elongation factors Tu (EF-Tu) and G (EF-G) is in either case accompanied by the hydrolysis of GTP (for reviews see Refs. 1 to 5). The EF-Tu-dependent reaction is associated with the enzymatic binding of arninoacyl-trna to the ribosomal acceptor site (A site), while that dependent on EF-G' follows the translocation of the newly * This work supportedin partbygrants of the Deutsche Forschungsgemeinschaft (Pa 106) and the Delegation Cenerale a la Recherche Scientifque et Technique (No and ). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. $ Present address, Cattedra di Chimica, 11" Facolta di Medicina e Chirurgia, Universita di Napoli, Napoli, Italy. fj Present address, Laboratoire de Biochimie, Laboratoire Associe No. 240 du C.N.R.S., Ecole Polytechnique, F Palaiseau Cedex, France. 'The abbreviationsusedare: EF-G, elongationfactor G; EF-T, elongation factor T (the complex formed by elongation factors TU and Ts); A site, aminoacyl site; P site, peptidyl site; trnaphe, trna species specific for phenylalanine release of the factor from the ribosome. The two reactions can also occur uncoupled from the elongation process by omitting one of the two factors. In these partial systems, the GTPase activity dependent on EF-G displays a turnover which is higher than that of EF-Tu by 2 to 3 orders of magnitude, the latter activity remaining more tightly associated with the enzymatic binding of aminoacyl-trna to the ribosome even in the absence of elongation. The property of EF-G to support a high GTPase activity uncoupled from polypeptide synthesis and also from codon-anticodon interaction indicates that it is principally this factor which is responsible for the excess of GTPase activity over amino acid incorporation, normally observed in protein-synthesizing systems. Already in 1964, when the EF-Tu-dependent GTPase reaction had yet to be discovered, Conway and Lipmann (6) reported that in uitro the GTPase activity proceeds at a much faster rate than polypeptide synthesis and that the apparent K,,, for GTP of the latter reaction is lower by 1 order of magnitude. The uncoupling of these two reactions has greatly hindered a precise determination of the stoichiometry between peptide bond formation and GTP hydrolysis in the in uitro system (7). Only recently, the work of Modolell et al. (8) has supplied experimental evidence for the theoretically expected stoichiometry of 2 GTP molecules hydrolyzed per each new peptide bond formed in a system which utilized purified, endogenous polysomes. This value was approximated, however, after a large correction for uncoupled EF-G GTPase activity. So far, neither the conditions which favor the coupling of the GTPase activity with polypeptide synthesis, nor the role played in this phenomenon by the functional heterogeneity of ribosomes have been investigated thoroughly. Recently, we have described the important role of monovalent and divalent cations in the regulation of the uncoupled ribosome-ef-g GTPase reaction (9-12). In the light of this experience, we have investigated the role of EF-Tu, trna, ribosomes, and ionic environment in coupling the EF-G-dependent GTPase reaction with peptide bond formation. For the first time, a stoichiometry of total GTP hydrolyzed versus amino acid incorporation close to the theoretically expected value of two has been obtained without any uncoupled EF-G GTPase activity. correction for EXPERIMENTAL PROCEDURES Materials-All reagents and biological components not quoted in this section were as already reported (13, 14). GTP was freed from traces of GDP by chromatography on DEAEcellulose (15) andlongchain poly(u) (M, 2 60,000, Ref. 13) was isolated on Sephadex G-75 gel filtration. Particular care was devoted to the control of the specific activity of ~-['~C]phenylalanine and [y-

2 7456 Coupling of EF-G GTPase Activity with Polypeptide Synthesis 32P]GTP, the latter prepared as described (14). The trnaph' and the Phe-tRNAPh', 75 to 90% pure (13), were freed from salts and nucleotides by gel filtration on Sephadex G-25 columns equilibrated with 1 mm MgCIz and 1 mm potassium acetate, ph 5.2. EF-G and elongation factor T (EF-Tu-EF-Ts) from Escherichia coli BT2' or A19 were electrophoretically homogeneous (16-18). One microgram of EF-G and EF-T corresponds to 12 and 15 pmol, respectively (19-21). Two or three times NRC1-washed ribosomes from E. coli were isolated as described (14). The ribosomes used showed a very low endogenous GTPase activity (less than 2 mol of GTP hydrolyzed/mol of ribosomes/h at 30 C) and displayed high activity in nonenzymatic binding of trnaphe. In the presence of saturating amounts of the various trnaphe forms, 60 to 70% of the 70 S ribosomes formed a complex with trnaphe at 13 to 20 mm MgC12, as judged by analysis on sucrose gradients in the presence of 5 mm MgCL (13). One AWJ unit of ribosomes was taken to represent 25 pmol (14). Assays-The details of the systems are given in the legends to tables and figures. The standard buffer was: 25 mm Tris-HC1, ph 7.8, 1 mm dithiothreitol, and 0.3 to 1.0% glycerol. The last component was carried over with ribosomes and elongation factors. The poly(u)-directed polyphenylalanine synthesis was measured as incorporation of ~-['~C]phenylalanine into hot trichloroacetic acidinsoluble material (13). The reaction mixtures containing all the components except EF-G were incubated for 10 min at 30 C to allow complete formation of the Phe-tRNAPhe.poly(U).ribosome complex. EF-G was then added, and after 10 min of additional incubation at 3OoC, phenylalanine incorporation was measured. The GTPase activity was measured as liberation of "P, as described (14). In the stoichiometry experiments, polyphenylalanine synthesis and GTP hydrolysis were assayed in two portions of the same reaction mixture. Both activities were measured under conditions giving linear kinetics. Blanks without EF-G were carried out in parallel and subtracted. The final percentage of glycerol carried over with ribosomes was 0.3 to 1%. Selection and Characterization of Ribosomes-The selection of the most active ribosome preparations was done as follows. Ribosomes were incubated at 30 C in a complete in uitro system for polyphenylalanine synthesis including a Phe-tRNA regeneration system consisting of trnaph', ATP, phenylalanyl-trna synthetase, and unlabeled phenylalanine. The amino acid was present in a limiting amount (8 mol/mol of ribosomes) which corresponds to 20 to 30% of the maximum incorporation obtainable by the system in the presence of excess substrate. The phenylalanine incorporation was run until completion at which point all polyphenylalanine chains were located in the ribosomal P site, as shown by their response to puromycin (22). The percentage of active ribosomes (55 to 67% in the best preparations) was calculated from the molar ratio between the ["HJpuromycin incorporated into trichloroacetic acid-insoluble material and the amount of 70 S ribosomes. The puromycin reaction was performed after gel filtration on Sepharose 4B to free the ribosome complexes from all other components of the system for polyphenylalanine synthesis. A second portion of these ribosome complexes was sedimented on an analytical sucrose density gradient in the presence of 5 mm MgCla (13), to determine the percentage of ribosomes unable to bind trna (20 to 30%). The ribosomes active in polyphenylalanine synthesis corresponded therefore to 70 to 80% of the ribosomes able to bind trna. The residual portion (20 to 30%) very probably consisted of ribosomes lacking peptidyltransferase activity as indicated by the observation that an equivalent amount of acetylphenylalanyltrnaph' bound to ribosomes at 10 mm MgCL was unreactive to puromycin even in the presence of EF-G and GTP (23). Other Methods-Protein concentration was determined by the method of Lowry et al. (24) using crystalline bovine serum albumin as standard. Radioactivity was determined by liquid scintillation spectrometry as described (13). RESULTS Conditions Regulating the Coupling of GTPase Activity with Polypeptide Synthesis-In this section we describe the experimental conditions which favor the coupling of GTP hydrolysis with amino acid incorporation, in a highly purified and well characterized system for poly(u)-directed polyphenylalanine synthesis. The stoichiometric values, in all cases, were obtained by subtracting the GTP hydrolyzed in a pardel system lacking EF-G, and therefore not corrected for the EF- G-dependent GTPase activity uncoupled from polypeptide synthesis. This activity represents by far the largest part of uncoupled GTPases present in a highly purified system (6). The importance to apply this procedure is shown in a later section. In Fig. 1, we report a typical experiment in which the stoichiometry of GTP hydrolysis versus amino acid polymerization was investigated as a function of MgCL concentration and under conditions of linear dependence of phenylalanine incorporation on EF-G concentration. Ribosomes highly active in polypeptide synthesis (about 65%, see "Experimental Procedures") were used. The concentration of EF-T was saturating. In this system, in the presence of 80 mm NRCl (Fig. la), the lowest stoichiometric ratios were observed between 4 and 6.5 mm MgClr and corresponded to 2.4 to 2.8 mol of GTP hydrolyzed/mol of amino acid incorporated. At higher MgCl:! concentrations, GTP hydrolysis became progressively uncoupled from polyphenylalanine synthesis. Similar results were obtained by replacing NHiCl with KC1 (Fig. 1B). In the presence of the latter cation, the lower activity in polyphenylalanine synthesis was accompanied by a comparably lower GTPase activity. In experiments not illustrated similar stoichiometries were also found when mixed monovalent cations (30 m~ NHlCl plus 30 mm KC1) were used or the NH4Cl concentration was raised to 240 m ~ In. the latter case, the MgCL optimum was slightly higher (8 mm uersus 7 m ~ and ) the maximum rate of polyphenylalanine synthesis was 25% of that observed at 80 mm NH4C1. Substitution of the Phe-tRNAPhe-regenerating system with precharged Phe-tRNAPhe did not affect the results. By contrast, with less active ribosomes, minimal stoichiometries higher than 3 were observed. In all cases, the lowest stoichiometric values were constantly observed at MgC12 concentrations slightly below the optimum for amino acid incorporation. 80 rnm NH~+ FIG. 1. Stoichiometry of GTP hydrolysis versus amino acid incorporation 0 during poly(u)-directed polyphenylalanine synthesis as a function of MgC12 concentration. The reaction mixtures contained, in of standard buffer, 20 (A) or 40 pmol (B) ribosomes, 10 (A) or 18 (B) pmol of EF-G, 60 pmol of EF- T, 100 pmol of trnaph", 10 pg of poly(u), 4 pg of phenylalanyl-trna synthetase, 700 pmol of ['4C]phenylalanine (specific activity 50 Ci/ mol), 0.2 nm ATP, 0.04 mm [y-"p]gtp (specific activity 95 Ci/mol), 80 mm NHEl (A) or KC1 (B), and varying concentrations of MgCL as indicated. The GTP hydrolyzed and the polyphenylalanine incorporated were determined on 60-~1 aliquots. Values on abscissa refer to the total volume (150 pl) of the reaction mixture.

3 Coupling of EF-G GTPase Activity with Polypeptide Synthesis mm Mg mm Mg2+ 80 mm NH,' 80mM K+ 1 /[GTP], mm" FIG. 2. Double reciprocal plots of the rate of GTP hydrolysis (0) and amino acid incorporation m) versus GTP concentration. The reaction mixtures contained, in 150 pi of standard buffer, 18 pmol (A) or 36 pmol (B) of ribosomes, 10 pmol (A) or 20 pmol (B) of EF-G, 80 pmol of EF-T, 150 pmol of trnaph", 10 pg of phenylalanyl synthetase, 1 nmol of ["Clphenylalanine (specific activity 50 Ci/mol), 0.4 mm ATP, [y-'*p]gtp (specific activity from 20 to 260 Ci/mol) as indicated, 80 mm NHaCI (A) or KC1 (B), and 6.5 mm MgCL. The assay conditions were as in Fig. 1. Velocities are expressed in nanomoles of GTP hydrolyzed and polyphenylalanine synthesized per 10 min at 30 C calculated for the total volume of the reaction mixture. TABLE I Determination of the stoichiometry of total GTP hydrolysis versus phenylalanine incorporation in the presence of different ratelimiting concentrations of EF-G The 150-pl reaction mixtures contained, in standard buffer, 17 pmol of ribosomes, EF-G as indicated, 80 pmol of EF-T, 110 pmol of trnaph', 10 pg of poly(u), 4 pg of phenylalanyl-trna synthetase, 620 pmol of ["Clphenylalanine (513 Ci/mol), 0.2 mm ATP, 12 PM [y- "'PIGTP (160 Ci/mol), 80 m~ NHIC1, and 4.5 mm (Experiment I) or 6.5 mm MgClz (Experiment 11). The assay conditions were as described under "Experimental Procedures." The mount of GTP hydrolyzed and phenylalanine incorporated refers to the total volume of the reaction mixture. ~~ ~ Experi- EF-G GTP hy-. Stoichiometry" ment added drolyzed nine lncorwrated B (A) C (A) C (B) pmol pmol pmol I A B C A B C To calculate the stoichiometry the amounts of GTP hydrolyzed and phenylalanine incorporated in the experiments within brackets were taken as blanks and subtracted. To achieve minimum stoichiometry (2.4), the presence of saturating amounts of Phe-tRNAPh" and EF-T, well as as the presence of limiting amounts of EF-G were a prerequisite. Higher concentrations of EF-T (tested up &fold to the optimal amount) did not further improve the stoichiometric values. Affinity for GTP of Polyphenylalanine Synthesis and GTPase Activity-The coupling of GTPase activity and amino acid incorporation in our system was examined by determining the apparent Michaelis constants of the two reactions under conditions of minimum stoichiometry. Therefore, the experiments in Fig. 2 were performed at 6.5 mm MgCl' and 80 m~ NH4CI (Panel A) or KC1 (Panel B) under conditions similar to those in Fig. 1. As illustrated by the double reciprocal plots, GTPase activity and phenylalanine incorporation show an identical affinity for GTP, which is 7 p~ with NH&l and 15 /AM with KC1. These values are much lower than those we determined in the absence of EF-Tu, under otherwise the same conditions (K,,, = 70 to 90 p~ at 80 mm NH&I or KCI, experiments not illustrated). It is important to note that under coupling conditions the K', of the GTPase activity in the complete system can be considered to represent that of EF-G. The other GTPase activity present in the system, that of EF-Tu, whose affinity for GTP is at least 1 order of magnitude higher in the absence of EF-G,' should display an apparent value identical with that of the EF-G GTPase activity. In fact, in a highly coupled system, the turnover of the EF-Tu GTPase activity associated with the enzymatic binding of aminoacyl-trna to the ribosome depends on the rate at which ribosomes with empty A site (post-translocative complexes) are generated by the translocation process associated with the coupled EF-G GTPase which thus limits the EF-Tu-dependent reaction. Our results are in line with the observations of Conway and Lipmann (6), who reported in 1964 that the apparent K, for GTP of polyphenylalanine synthesis (16 PM) was about 1 order of magnitude lower than that (180 p ~ of ) the GTPase activity. Indeed, their experimental conditions support a large excess of uncoupled GTPase activity. Correction for GTPase Activity Uncoupled from Polypeptide Synthesis-A precise estimation of the stoichiometry requires the correction of total GTP hydrolysis for the amount of GTP hydrolyzed by reactions not associated with the elongation process. These reactions are represented by the uncoupled EF-G and EF-Tu GTPases and by the GTPase activity which to a little extent is always found with NH,Clwashed ribosomes. This activity, whose precise nature is unknown, is associated with ATPase activity and is not involved in the elongation process:' In our experiments, we have corrected for the GTPase activity contaminating the ribosomes and for the EF-Tu GTPase uncoupled from polypeptide synthesis, but not for the uncoupled EF-G GTPase activity which is responsible for most of the GTP hydrolysis not associated with polypeptide synthesis. Our blanks subtract also the GTP hydrolysis associated with the enzymatic binding of PhetRNAPh' to active ribosomes in the absence of elongation. Even when this activity was considered in the calculation of the stoichiometry, its subtraction had little effect on the obtained stoichiometric values because of the small amount of the GTP hydrolysis involved. The correctness of our procedure was verified by measuring GTP hydrolysis and polyphenylalanine synthesis in two reaction mixtures containing E. De Vindittis and A. Parmeggiani, unpublished results, ''I G. Chinali, unpublished results.

4 7458 Coupling of EF-G GTPase Activity with Polypeptide Synthesis TABLE I1 Estimation ofthe uncoupled EF-G GTPase by omitting trnaph' from the complete system for polyphenylalanine synthesis: effect of GTP concentration Assay conditions in the complete system were the same as in the experiment of Fig. 2, Panel B. In the incomplete system, trnaphe was omitted. In the absence of trnaph', no incorporation of [I4C]phenyldanine was observed. PM pmol pmol "A, calculated as molar ratio of total GTP hydrolyzed versus phenylalanin- incorporation in the complete system. B, calculated after correction for the GTP hydrolyzed in the system lacking trnaph'. different rate-limiting amounts of EF-G. As shown in Table I, the rates of both reactions increased linearly with EF-G concentration starting from the background values in the absence of the factor. Therefore, the stoichiometry determined in each reaction mixture after background subtraction was the same as that calculated by difference from the background-uncorrected activities of the two reaction mixtures. For the experiments in Table I, the background values in the absence of EF-G ranged from 7 to 30% of the total GTPase activity. We have decided against correction for uncoupled EF-G GTPase activity usually done by determining the GTP hydrolysis in a parallel system in which EF-Tu or aminoacyltrna were omitted (6, 7), since this correction takes for granted that this hydrolysis represents the uncoupled EF-G GTPase activity in the complete system during polypeptide synthesis. In the experiment reported in Table 11, we show that this assumption is incorrect under our conditions of minimal stoichiometry. As expected from the different affnities for GTP of the coupled and uncoupled GTPase activities, the rate of GTP hydrolysis in the system lacking trnaphe increased more than that of the complete system with increasing substrate concentration to the point that negative stoichiometries were obtained by subtracting the values of the uncoupled system from those of the coupled one. These results indicate that this type of correction cannot. be applied under conditions which couple the EF-G GTPase activity amino acid polymerization. DISCUSSION with Our results show that coupling of the ribosome-ef-g GTPase activity with polypeptide synthesis depends on the coordination of the interaction of EF-G and EF-Tu with the ribosome. To achieve the coupling, use of an appropriate ionic environment, particularly of Mg'+, is required. Moreover, the activity of EF-G in the translocation process has to be ratelimiting with respect to the regeneration of pretranslocative ribosome complexes dependent on the action of EF-Tu which therefore has to be present in saturating amounts. This reduces the possibility that EF-G generates an uncoupled GTPase activity by interacting with ribosomes in posttranslocative state. If these conditions are satisfied the stoichiometry between GTP hydrolysis and peptide bond formation should be minimal while the apparent affinity for GTP of the GTPase activity and the amino acid incorporation ought to become identical, as we have proved experimentally in this work. Under optimal conditions, we have observed a stoiochiometry of 2.4 to 2.8 mol of GTP hydrolyzed/mol of phenylalanine incorporated without introducing any correction for uncoupled EF-G GTP hydrolysis. The contribution to this activity of the ribosome inactive in polypeptide synthesis (33 to 45% in our preparations) can be estimated as follows. About twothirds of the inactive ribosomes are represented by ribosomes unable to bind trna and therefore supporting a reduced EF-G GTPase activity due to the lower Mg'+ requirement of the coupled system. Indeed, in experiments to be published elsewhere, we have found that at low Mg2' concentrations (3 to 8 mm) in the absence of trna bound to the ribosomal P site the uncoupled EF-G GTPase activity is strongly reduced (two to three times). Most of the residual inactive ribosomes are those depleted of peptidyltransferase activity, but still able to bind trna. These ribosomes, which should carry PhetRNAPhe at both A and P sites, should support a GTPase activity close to that of ribosomes active in polypeptide synthesis because of their resemblance to the pretranslocative ribosome complex. From these considerations, the ribosomes inactive in polypeptide synthesis should support in the complete system an uncoupled EF-G GTPase which can be esti- mated to account for no more than 20 to 25% of the total GTP hydrolysis. These conclusions, together with our stoichiometric values, indicate that hydrolysis of 2 mol of GTP is required for the incorporation of 1 mol of amino acid. Cabrer et al. (7) also found a value of approximately 2 mol of GTP hydrolyzed/mol of amino acid incorporated, by utilizing NH4C1-washed, endogenous polysomes. This value was obtained after subtraction of the amount of GTP hydrolyzed in an incomplete system lacking EF-Tu or aminoacyl-trna. Uncorrected stoichiometries ranged from 4 to 7 mol of GTP hydrolyzed/moi of amino acid incorporated. Although that correction may represent a reasonable approximation for sys- tems in which most of the EF-G GTPase activity of is uncoupled type, this does not hold for a coupled system. In fact, in the complete system for polypeptide synthesis, the activity of EF-G is regulated by the EF-Tu-dependent binding of aminoacyl-trna to the ribosome, and therefore only the ribosomes inactive in polypeptide synthesis are engaged in uncoupled EF-G GTPase, compared to all the ribosomes in the incomplete system. Moreover, the K, values of the GTPase activity of the coupled and the uncoupled systems are markedly different and therefore the extent of correction increases with GTP concentration. The amount of GTP hydrolyzed in the incomplete system lacking EF-T or aminoacyl-trna may represent a large overestimation of the uncoupled EF-G GTPase activity present in the coupled system as shown by the paradoxically low values of the stoichiometry after this correction (see Table 11). In conclusion, our data indicate that the interaction of the EF-G. GTP complex with the ribosome in the pretranslocative state is the prerequisite for a coupled GTPase reaction. The higher affinity for GTP of this reaction indicates a preferential recognition of the pretranslocation complexes by EF-G. GTP. That the minimum stoichiometry is obtained only when the rate of formation of the pretranslocation complex does not limit the elongation process, suggests that these complexes are specifically needed for triggering the coupled EF-G GTPase reaction. We feel that the definition of the conditions required for coupling the ribosome-ef-g GTPase activity with amino acid incorporation may be useful for further studies on the role of the GTP hydrolysis in polypeptide synthesis.

5 Acknowledgment-We reading the manuscript. Coupling of EF-G GTPase Activity with Polypeptide Synthesis 7459 wish to thank Dr. G. Sander for critically REFERENCES 1. Lucas-Lenard, J. & Lipmann, F. (1971) Annu. Reu. Biochem. 40, Lucas-Lenard, J. & Beres, L. (1974) in The Enzymes (Boyer, P. D., ed) Vol. 10, pp , Academic Press, New York 3. Brot, N. (1977) in Molecular Mechanisms of Protein Synthesis (Weissbach, H. & Pestka, S., eds) pp , Academic Press, New York 4. Miller, D. L. & Weissbach, H. (1977) in Molecular Mechanisms of Protein Synthesis (Weissbach, H. & Pestka, S. eds) pp , Academic Press, New York 5. Bermek, E. (1978) Prog. Nucleic Acid Res. Mol. Biol. 12, Conway, T. W. & Lipmann, F. (1964) Proc. Natl. Acad. Sci. U. S. A. 52, Nishizuka, Y. & Lipmann, F. (1966) Proc. Natl. Acad. Sci. U. S. A. 55, Cabrer, B., San-Millan, M. J., Vazquez, D. & Modolell, J. (1976) J. Biol. Chem. 251, Voigt, J., Sander, G., Nagel, K. & Parmeggiani, A. (1974) Biochem. Biophys. Res. Commun. 57, Parmeggiani, A,, Sander, G., Marsh, R. C., Voigt, J., Nagel, K. & Chinali, G. (1974) in Lipmann Symposium: Energy, Regulation and Biosynthesis in Molecular Biology (Richter, D., ed) pp , Walter de Gruyter, Berlin 11. Sander, G., Marsh, R. C. & Parmeggiani, A. (1976) Eur. J. Biochem. 61, Sander, G., Parlato, G., Crechet, J.-B., Nagel, K. & Parmeggiani, A. (1978) Eur. J. Biochem. 86, Chinali, G. & Parmeggiani, A. (1973) Eur. J. Biochem. 32, Sander, G., Marsh, R. C., Voigt, J. & Parmeggiani, A. (1975) Biochemistry 14, Sander, G., Marsh, R. C. & Parmeggiani, A. (1972) Biochem. Biophys. Res. Commun. 47, Parmeggiani, A. (1968) Biochem. Biophys. Res. Commun. 30, Parmeggiani, A,, Singer, C. & Gottschalk, E. M. (1971) Methods Enzymol. 20, Chinali, G., Wolf, H. & Parmeggiani, A. (1977) Eur. J. Biochem. 75, Parmeggiani, A. & Gottschalk, E. M. (1969) Biochem. Biophys. Res. Commun. 35, Parmeggiani, A. & Gottschalk, E. M. (1969) Cold Spring Harbor Symp. Quant. Biol. 34, Hachmann, J., Miller, D. L. & Weissbach, H. (1971) Arch. BLOchem. Biophys. 147, Wolf, H., Chinali, G. & Parmeggiani, A. (1974) Proc. Natl. Acad. Sci. U. S. A. 71, Chinali, G. & Parmeggiani, A. (1973) Biochem. Biophys. Res. Commun. 54, Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, H. J. (1951) J. Bid. Chem. 193,