Requirement for GTP in the Initiation Process on Reticulocyte Ribosomes

Transcription

1 Proc. Nat. Acad. Sci. USA Vol. 68, No 9, pp , September 1971 Requirement for GTP in the Initiation Process on Reticulocyte Ribosomes and Ribosomal Subunits (GDPCP/Met-tRNAF/initiation factors/elongation factors/fusidic acid) D. A. SHAFRITZ, D. G. LAYCOCK, R. G. CRYSTAL, AND W. F. ANDERSON Section on Molecular Hematology, National Heart and Lung Institute, National Institutes of Health, Bethesda, Maryland Communicated by Robert W. Berliner, July 13, 1971 ABSTRACT The requirement for GTP in the initiation process on reticulocyte ribosomes and ribosomal subunits has been examined by studying Met-tRNAr binding, ribosome-dependent [y-82p]gtp hydrolysis, and peptide-bond formation with puromycin. Met-tRNAr binding can be obtained with the methylene analogue, 5'-guanylylmethylene diphosphonate, as well as GTP, and it is not inhibited by fusidic acid or several other inhibitors of protein synthesis. This reaction can be performed with the 40S subunit and has the same requirements as the Met-tRNAF-binding reaction with washed ribosomes. Ribosome-dependent [y-82pjgtp hydrolysis can be obtained with the initiation factor M2A using either washed ribosomes or the 40S subunit. This reaction is also not significantly inhibited by fusidic acid. Peptide-bond formation between puromycin and Met-tRNAF, however, is inhibited by fusidic acid, and does not occur if the methylene analogue of GTP is substituted for GTP. These data suggest that the binding of the initiator trna to the 40S subunit does not require the hydrolysis of GTP, but that at least one GTP hydrolysis event must occur after Met-tRNAF binding in order for the first peptide bond to be formed. The mechanism of protein synthesis in mammals, just as in bacteria, involves a special recognition process in order to initiate the synthesis of polypeptide chains. This process utilizes a distinct group of protein factors, called initiation factors (1-4), and a distinct species of Met-tRNA, MettRNAF, (5, 6). In order to understand the detailed mechanism of initiation, it is helpful to examine the roles for each separate component in the individual sequential reactions that compose the initiation process. The present studies investigate the GTP requirements for Met-tRNAF binding and for peptide-bond formation on reticulocyte ribosomes or ribosomal subunits programmed with the artificial template AUG. MATERIALS AND METHODS The preparation of washed reticulocyte ribosomes, elongation factors T, and T2, initiation factors M1 and M2(AB), and acylated reticulocyte trna was as reported (4, 7-9), except that M1 was purified through phosphocellulose as well as DEAE-cellulose and Sephadex G-150 chromatography. The separation of M2 into two components, M2A (which elutes at the void volume on Sephadex G-200 chromatography) and M2B (which has a molecular weight of 20,000-40,000), will be reported elsewhere. Fusidic acid and aurintricarboxylic acid were the gifts of H. Weissbach and A. Grollman, respectively. The assays for aminoacyl-trna binding, peptide-bond Abbreviation: GDPCP, 5'-guanylyl-methylene diphosphonate. formation with puromycin, and cell-free poly(u)-dependent protein synthesis were also as reported (4, 7, 8). Preparation of ribosomal subunits Reticulocyte 40S and 60S ribosomal subunits were prepared by zonal centrifugation, using a modification of the procedure of Falvey and Staehelin (10); both the ribosomal preincubation and the deoxycholate detergent steps were eliminated. Crude polysomes were suspended in standard sucrose solution [0.25 M sucrose-1 mm dithiothreitol-0.1 mm EDTA, (ph 7.0)] at a final concentration of 200 A260 units/ml, and 4 M KCl was added to a final concentration of 0.5 M. The suspension was mixed gently in an ice bucket for min and placed over a 15-30% exponential sucrose gradient buffered - with TKM [10 mm Tris HC1, (ph 7.5)-40 mm KCI-3 mm MgCl2] in a Beckman Ti-14 zonal rotor. The material was centrifuged at 45,000 rpm for 4 hr and unloaded under constant UV-monitoring while 10-ml fractions were collected. The fractions containing 40S and 60S material were pooled, pelleted at 258,000 X g for 3.5 hr, and suspended in standard sucrose-tkm. Materials from two separate zonal centrifugations for each subunit were pooled and centrifuged in the Ti- 14 zonal rotor as described above, after which time only the peak fractions for the 40S and 60S subunit were collected, centrifuged, and resuspended in standard sucrose-tkm to a final concentration of 200 A260 units/ml. Small aliquots were stored in liquid nitrogen until use. Ribosome-dependent GTP hydrolysis The ribosome- and factor-dependent hydrolysis of [932P]GTP was measured by a modification of the assay procedure reported by Conway and Lipmann for the bacterial system (11). Incubations, in a total volume of 100 IAI, were performed at 370C for 10 min and contained 20 mm Tris HCl (ph 7.5), 100 mm KC1, 5 mm MgCl2, 1 mm dithiothreitol, 1.2 A260 units of reticulocyte ribosomes twice washed with 0.5 M KCl (or 40S and/or 60S ribosomal subunits), 200 pmol of ['y-'2p] GTP (2,000-4,000 cpm/pmol, International Chemical and Nuclear Corp.), and factors as indicated in Table 6. The reaction was stopped by the addition of 0.5 ml of 0.02 M silicotungstic acid in 0.02 M H2S04, followed by 1.2 ml of 1 mm potassium phosphate buffer (ph 8.0), and 0.5 ml of 5%0 ammonium molybdate in 4 M H2S04. The [,y_32p]phosphomolybdate complex was extracted into 2 ml of isobutanol: benzene 1:1, and 1 ml of the organic layer was assayed for radioactivity in 10 ml of the scintillation fiuid described by Bray (12). 2246

2 Proc. Nat. Acad. Sci. USA 68 (1971) TABLE 1. Energetics of Met-tRNA binding to washed reticulocyte ribosomes Met- MettRNAF trnam Additions or (5 mm Additions or (10mM deletions Mg++) deletions Mg++) A. Complete system 0.95 B. Complete sys (Ml + M2AB) tem (T,) - Ml T, M2AB Ribosomes Ribosomes AUG Poly(AUG) GTP fusidic GDP 0.10 acid + GDPCP GTP GDPCP + + GDPCP 0.89 fusidic acid fusidic acid fusidic acid 0.89 Incubations, in a total volume of 50 pl, were performed at 230C for 2.5 min and contained 20 mm Tris.HCl (ph 7.5), 100 mm KCl, 0.2 mm GTP (or 0.25 mm GDP or 0.25 mm GDPCP, as indicated), 1 mm dithiothreitol, and 1.2 A260 units of washed reticulocyte ribosomes. A. In addition, the reaction mixture contained 5 mm Mg++, 8.0 pmol of rabbit reticulocyte [3H]MettRNAF (14.1% charged, 1600 cpm/pmol), 0.15 A20 units of AUG triplet, 0.76,g Ml protein, 12.5 pxg M2(A.B) protein, and 0.6 mm fusidic acid, as indicated. Under the conditions used, no significant incorporation of Met-tRNAF into protein or oligopeptide was detected either by hot-trichloroacetic acid precipitation or paper chromatography in butanol:acetic acid:water 4:1:1. B. In addition, the reaction mixture contained 10 mm Mg++, 8.0 pmol of rabbit reticulocyte [3H]Met-tRNAM [4.7% charged, 1600 cpm/pmol), 0.3 Amoo units of poly(a,u,g), 4.25 pg of T1 protein and 0.6 mm fusidic acid, as indicated. Under these conditions, a portion of the radioactivity remained at the origin as determined by paper chromatography, but little radioactivity was precipitated by hot-trichloroacetic acid. For further details of the methods, see ref. 4. RESULTS Binding of Met-tRNAF to reticulocyte ribosomes The requirements for binding of Met-tRNAF and MettRNAm reticulocyte ribosomes are shown in Table 1. As reported (4), there is a strict requirement for GTP, as well as M1 and M2, for the binding of Met-tRNAF to ribosomes at 5 mm Mg++. Under the conditions used, GDPCP (5'-guanylylmethylenediphosphonate) can substitute for GTP, whereas GDP is inactive (Table 1A). Fusidic acid, an inhibitor of protein synthesis that blocks T2-dependent translocation, does not significantly inhibit Met-tRNAF binding. Similar results are found for Ti-dependent binding of Met- trnam to ribosomes at 10 mm Mg++ (Table 1B). Since the trna-binding reaction for bacterial initiation at low Mg++ concentration is specific for fmet-trnaf, and requires both initiation factors and GTP, these parameters were retested for fmet-trnaf binding in the mammalian system. We previously reported that mammalian fmettrnaf binding to washed reticulocyte ribosomes can be obtained with factor M1 (4). This reaction is dependent on exogenous AUG template, can be performed with either GTP and Initiation in Reticulocyte Ribosomes Mg9 CONC. (mm) FIG. 1. Ml-dependent binding of [3H]fMet-tRNAF to washed reticulocyte ribosomes as a function of Mg++ concentration. Incubations, in a total volume of 50 ul, were performed for 2 min at 23 C under conditions as noted in Table 1, except that 10 pmol reticulocyte [3HlfMet-tRNAF (>98% charged) were used, along with 1.5 pg of Ml protein. Escherichia coli or reticulocyte fmet-trnaf, and has a Mg++ concentration optimum of 10 mm, as shown in Fig. 1. In contrast to Met-tRNAF binding, fmet-trnaf binding to reticulocyte ribosomes does not require GTP (Table 2). In the poly(u) system, N-acetyl-Phe-tRNA binding with M1 is also totally independent of GTP (Table 2), whereas PhetRNA binding with either T1 or M1 is clearly GTP dependent. For the binding studies reported here, M1 has been purified more than 300-fold (Table 3), but this does not exclude the possibility that the factor responsible for fmet-trnaf (or N-acetyl-Phe-tRNA) binding to ribosomes is distinct from M1 and has not yet been separated by the purification steps TABLE 2. GTP requirement for aminoacyl-rna binding to washed reticulocyte ribosomes N-acetyl- [3H]fMet- [14C]Phe- [14C]Phe- Additions to trnaf trna trna ribosomes (10 mm Mg++) (5 mm Mg++) (5 mm Mg++) T, + GTP T, - GTP M1 + GTP Ml - GTP Incubations were performed for 3 min at 23 C under conditions similar to those described in Table 1. In each instance, 8.0 pmol of substrate was used: E. coli B [14C]Phe-tRNA ( % charged, 750 cpm/pmol), N-acetyl-[l"CjPhe-tRNA (>98% acetylated, otherwise the same as ["4C]Phe-tRNA), [3H]Met-tRNAF (14.1% charged, 1600 cpm/pmol), [3H]fMet-tRNAF (>98% formylated, otherwise the same as [3H]Met-tRNAF). T, (4.25 pg protein) and Ml (0.76 pug protein)were used in saturating amounts.

3 2248 Biochemistry: Shafritz et al. Proc. Nat. Acad. Sci. USA 68 (1971) TABLE 4. Effect of antibiotic inhibitors on MttRNAF binding and on polyphenylalanine synthesis LgM2A, Protein pg Ml protein FIG. 2. M factor-dependent binding of Met-tRNAF to the reticulocyte 40S ribosomal subunit. Incubations, in a total volume of 50 jul, were performed at 230C for 2.5 min as noted in Table 1. For these studies, 0.19 A260 units of 40S subunit were used. A. Ml, 0.76 Mg protein. B. M2(AB), 12.5,g protein. used. It is also interesting to note that, for both Met-tRNAF and Phe-tRNA, binding is greater with the N-blocked than with the free aminoacyl-trna derivative (see Tables 1 and 2). Certain substances reported to be specific inhibitors of mammalian initiation, i.e., aurintricarboxylic acid (13), pactomycin (14), and NaF (15) do not inhibit AUG-dependent Met-tRNAF binding at the concentrations shown (Table 4A). Higher concentrations of these substances partially inhibit the binding reaction. At the same, or even lower, concentration of inhibitors, however, there is almost complete cessation of poly(u)-dependent protein synthesis at low Mg++ concentration (Table 4B). Puromycin and tetracycline do not inhibit reticulocyte Met-tRNAF binding, and the increase in radioactivity with tetracycline is not dependent on either ribosomes or initiation factors. TABLE 3. Purification of initiation factor Ml Sp. Stage of Purification Vol. (ml) Protein (mg) Units Act. (U/mg) Yield % 1. Ribosomal wash Ammonium sulfate (35-65%) DEAEcellulose Phosphocellulose Sephadex G Activity for Ml was determined in the presence of saturating M2(AB) by [8H]Met-tRNAF binding. A unit of Ml activity was defined as the amount required to bind 1.0 pmol of Met-tRNAF to 1.0 A260 unit of twice=washed reticulocyte ribosomes in 2.0 min. The ammonium sulfate fraction gave high blanks for binding in the absence of added ribosomes, so that no calculation of units or specific activity was made for this step. Met- [I4C]PhetRNAF trna polymerized Inhibitor (5 mm Mg++) (5 mm Mg++) None Aurintricarboxylic acid (5 ug) Pactamycin(2X1O-4M) NaF (5 X 10-3 M) Puromycin (6 X 10-4 M) Tetracycline (3 X 1O-4 M) Incubations for Met-tRNAF binding were performed as described in Table 1. Incubations for polyphenylalanine synthesis were performed in a total volume of 50 Mul at 370C for 2 min. Incorporation of [14C]phe-tRNA into protein was determined by hot trichloroacetic-acid precipitation. Further details of this assay are given in ref. 7. Ml, 0.76 jg protein; M2, 12.5 jg protein; T1, 4.25 Mg protein and T2, 1.6,ug protein, were all present in saturating amounts. Binding of Met-tRNAF occurs predominantly with the 408 subunit, as shown in Table 5, and has all the characteristics previously demonstrated for washed ribosomes including separate requirements for M1 and M2(AB), as shown in Fig. 2. Under the conditions tested, binding is not stimulated significantly by the 60S subunit when an Aeso ratio of 60S to 408 subunits from 1.0 to 3.7 was tested. At the higher amounts of 60S subunit there was, in fact, a decrease in Met-tRNAF binding; raising (or lowering) the Mg++ concentration did not increase 60S dependence (data not shown). With MettRNAm, optimal activity is obtained at 8 mm Mg++ in the presence of Ti, but only when both the 40S and 60S subunits TABLE 5. Binding of Met-tRNA to reticulocyte ribosomal subunits Additions or deletions 40S 60S 40S + 60S Met-RNAF (5 mm Mg++) A. Ml M2AB M1 + M2AB AUG GTP T, Met-tRNAm (8 mm Mg++) B. Ml+M2A,B T, poly(aug) GTP O. 17 Assay conditions were as described in Table 1, except that 0.19 A260 units of 408 subunits and/or 0.35 A2.0 units of 60S subunits were substituted for washed ribosomes. A range of 60S subunits from 0.20 A260 units to 0.70 A26. units was used. Nonenzymatic binding blanks with subunits in the absence of added factors were less than 0.08 pmol in each instance.

4 Proc. Nat. Acad. Sci. USA (1971) were used. As with washed ribosomes, enzymatic binding of Met-tRNAM requires GTP. Ribosome-dependent GTP hydrolysis Ribosome-dependent[_y-32P]GTP hydrolysis has been studied in bacterial systems to study the energy-requiring steps in the initiation of protein synthesis (16-18). The ability of M2A to hydrolyze [7-32P]GTP in both the coupled and uncoupled reactions is shown in Table 6. These reactions require the use of ribosomes twice washed with 0.5 M KCl to reduce nonenzymatic hydrolysis blanks. Activity for M2A occurs predominantly with the 40S subunit, and stimulation by the 60S subunit is no greater than the additive effect; hydrolysis is not significantly inhibited by fusidic acid. T2-dependent GTP hydrolysis occurs with the 60S subunit, and is significantly increased by the addition of the 40S subunit. In contrast to M2A activity, T2-dependent hydrolysis is markedly inhibited by fusidic acid. Hydrolysis of [-y-32p]atp also occurs with M2A in the presence of washed ribosomes or the 40S subunit; this activity has not been obtained with T2. Peptide-bond formation with puromycin We have previously reported that peptide-bond formation between puromycin and Met-tRNAF is dependent on Ml and M2(AB), and is strongly stimulated by elongation factor T, (8). As shown in Table 7, the puromycin reactions with Met-tRNAF, Met-tRNAM, and Phe-tRNA all are partially inhibited by fusidic acid. This is true in spite of the great variation in the requirement for added T2 in these reactions, TABLE 6. GTP and A TP hydrolysis by reticulocyte ribosomes and ribosomal subunits Washed Additions ribosomes 40S 60S 40S + 60S A Pi released from [ -y32p] GTP A. None 6.0 <0.5 <0.5 <0.5 M2A Fusidic acid M2A + AUG + Met-tRNAF Fusidic acid T Fusidic acid A Pi released from [y.32p] ATP B. None M2A T Incubations, in a total volume of 100 jal, were performed at 37 C for 10 min, and contained 20 mm Tris*HCl (ph 7.5), 100 mm KCl, 5 mm MgCl2, 1 mm dithiothreitol, and 200 pmol [-y-32p]gtp (2,000-4,000 cpm/pmol) or 200 pmol [y-32p]atp (1,500-3,000 cpm/pmol) both from International Chemical and Nuclear Corp. M2A, 6.0 jg protein, T2, 1.6 lg protein, 0.15 A260 units AUG, pmol Met-tRNA, 0.6 mm fusidic acid, 1.2 A260 units of ribosomes twice washed with 0.5 M KC1, 0.19 A260 units of 40S subunits, and/or 0.35 A260 units of 60S subunits were added as indicated. For [-y-32p]gtp, a background of 3.2 pmol in the absence of ribosomes, subunits, or factors and additional blank values of 5.8 pmol with M2A alone or 0.2 pmol with T2 alone were subtracted from each point. For [y-32pjatp, similar, but slightly higher, blanks were obtained. GTP and Initiation in Reticulocyte Ribosomes 2249 TABLE 7. Effect of GDPCP and fusidic acid on reticulocyte puromycin-peptide release Addition to reaction mixture Met-tRNAF Met-tRNAM Phe-tRNA GTP GTP + fusidic acid GDPCP Incubations, in a total volume of 50 pl, were performed at 300C for 10 min, and contained 20 mm Tris- HCl (ph 7.5), 150 mm KCI, either 3 or 6 mm Mg++, 0.5 mm GTP or 0.5 mm GDPCP, 1 mm dithiothreitol, 0.5 mm puromycin (neutralized to ph 7.0), 1.2 A260 units of washed ribosomes, 15 pmol [14C]Phe-tRNA, 10 pmol [3H]Met-tRNAF, 10 pmol ['4C]fMet-tRNAF, or 10 pmol [3H]Met-tRNAm, 0.76 sg M1 protein, 12 jsg M2(A, B) protein, 4.25,g T1 protein, and 0.6 mm fusidic. acid, where indicated. Details of the assay procedure are given in ref. 8. i.e., none for Met-tRNAF, partial for Met-tRNAm, and almost complete for Phe-tRNA (8). GDPCP cannot substitute for GTP in either the Met or the Phe system (Table 7). When fmet-trnaf is used as substrate, fusidic acid still partially inhibits and GDPCP will substitute for GTP to a limited extent, if at all. The puromycin reaction with MettRNAF at low Mg++ concentration has also been performed with ribosomal subunits and appears to require both the 408 and 60S components (data not shown). A summary of the characteristics of Met-tRNA binding, ribosome-dependent ['y-82p]gtp hydrolysis, and peptidebond formation with puromycin is shown in Table 8. DISCUSSION Since the original reports of (a) a special initiator species of Met-tRNA in higher organisms (5, 6) and (b) the apparent requirement for special initiation factors for cell-free protein synthesis in reticulocytes at low Mg++ concentration (3, 4, 7), TABLE 8. Requirements for the initiation process 1) Met-tRNAF binding to ribosomes: Low Mg++ optimum Requires initiation factors Ml + M2 Can take place on the 40S subunit only GDPCP can replace GTP; fusidic acid does not inhibit Met-tRNAm binding to ribosomes: Higher Mg++ optimum Requires elongation factor T1 Requires both 40S and 60S subunits GDPCP can replace GTP; fusidic acid does not inhibit 2) M2A-dependent GTP hydrolysis: Occurs with the 40S subunit Not inhibited by fusidic acid T2-dependent GTP hydrolysis: Occurs predominantly with the 60S subunit Markedly inhibited by fusidic acid 3) Puromycin-peptide bond formation with Met-tRNAF: Requires initiation factors Ml + M2; stimulated by T1 GDPCP cannot replace GTP; fusidic acid inhibits

5 2250 Biochemistry: Shafritz et al. a large body of evidence has rapidly accumulated to implicate Met-tRNAF as an initiator for many eucaryotic systems (19-32). The primary evidence supporting the initiation properties of Met-tRNAF in hemoglobin synthesis is 2-fold: (a) The reticulocyte initiation factors specifically recognize Met-tRNAF, while they discriminate against other trna species (4) and (b) Met-tRNAF donates methionine to the N- terminal end of the growing hemoglobin chain, while MettRNAM donates methionine into internal positions (19-23,37). We have examined the GTP requirement for Met-tRNAF binding and for peptide-bond formation with puromycin in the AUG-dependent model system. The binding reaction requires a triphosphate derivative of guanylic acid, but since GDPCP can serve this role, hydrolysis of GTP does not occur during the initial binding event, or at least is not an obligatory step; this result is the same as that found in the bacterial system by Thach (33). Met-tRNAF binding can also be performed with just the 40S subunit under conditions where MettRNAM binding requires both the 40S and the 60S subunit. Under the experimental conditions used, the 60S subunit does not appear to stimulate activity significantly. This result does not mean, however, that the 60S subunit does not affect Met-tRNAF binding or does not participate in formation of the final "initiation complex," but merely indicates that such effects were not observed using the nitrocellulose-filter assay technique. In the bacterial system, addition of the larger ribosomal subunit greatly stimulates fmet-trnaf binding (34). In Met-tRNAF binding with MI plus M2, the elongation factors T, and T2 are clearly not required. In contrast to the binding of Met-tRNAF to ribosomes, fmettrnaf binding in the reticulocyte system lacks certain characteristics of initiation; namely, it is optimal at a higher Mg++ concentration, does not require GTP, and does not require the initiation factor M2. Overall, the requirements for the binding of initiator trna in the mammalian system appear to be similar to those of the bacterial system (34), except for the absence of formylation of Met-tRNAF and the absence of a marked stimulation by the 60S subunit. Wool and coworkers (35, 36) and Moldave and coworkers* have reported that certain aminoacyl-trnas are enzymatically to 40S subunits in a mammalian cell-free system. These reactions have been obtained with fractions or factors derived from liver supernatant protein, but that are distinct from T1. Some of the characteristics of these factors in the binding of both Phe-tRNA and Met-tRNA (N-blocked, as well as free) are similar to results we have obtained with crude preparations of reticulocyte M factors. Purification procedures have led to the separation of various types of trna-binding activity (both ribosomal- and nonribosomaldependent) from our Ml and M2 preparations. This observation, together with the isolation of Ml and M2 from the ribosomal wash fraction of rabbit liver (unpublished observations), but not from the supernatant, suggests that the various factors reported might be entities distinct from Ml and M2. Subsequent to the (M1 + M2)-directed Met-tRNAF binding to the 40S subunit, the 60S subunit must join the Met-tRNA-mRNA-40S complex, since both the 40S and 60S subunits are required for the synthesis of the N-terminal * Moldave, K., and E. Gasior, Fed. Proc., 30, 1290a (1971). dipeptide analog, methionyl-puromycin (8), and for the synthesis of the natural N-terminal dipeptide of hemoglobin, methionyl-valine (37). Also, GTP, rather than GDPCP, is required for the formation of the initial peptide bond, indicating that hydrolysis probably occurs between initiatortrna binding and peptide-bond formation. Factor M2A hydrolyzes [,y2p]gtp in both coupled and uncoupled reactions, similar to the reaction reported for initiation factor F2 in the E. coli system (16-18). We have also Proc. Nat. Acad. Sci. USA 68 (1971) found that M2A-dependent GTP hydrolysis occurs on the small ribosomal subunit, and is not significantly stimulated by the addition of the large ribosomal subunit. This, together with the ability of M2A to hydrolyze [9y2P]ATP in a ribosomedependent reaction, suggests that there are differences in the procaryotic and eucaryotic systems. Whether the ATP hydrolysis of M2A is related to the liver supernatant factor that is distinct from T, and T2 reported by Maxwell and coworkers (38), which hydrolyzes both GTP and ATP, or whether it is related to the reported ATP requirement for initiation in wheat (39) is not clear. We have not been able to demonstrate an ATP stimulation of the AUG-dependent Met-tRNAF binding reaction or the M factor-dependent poly(u)-mg++ shift. The formation of methionyl-puromycin in the reticulocyte cell-free system, using KCl-washed ribosomes, AUG, MettRNAF, and factors M1, M2(AB), and T1, does not require added T2 (8). However, the inhibition of methionyl-puromycin formation by fusidic acid in concentrations (6 X 10-4 M) known to inhibit all translocase (T2) functions in mammalian systems (38, 40) might implicate factor T2 (presumably contaminating either ribosomes or initiation factors) in methionylpuromycin formation. However, the same concentration of fusidic acid does not inhibit the formation of methionylvaline, the natural N-terminal dipeptide in hemoglobin biosynthesis (unpublished observatiors). As previously reported, the reason for the requirement for T, in the methionylpuromycin reaction remains unclear (8). Therefore, until more complete evidence is obtained, no final conclusions can be reached regarding the possible roles of the elongation factors in the formation of the first peptide bond. We thank Miss Patricia J Hogan and Miss Rebecca L. Rawls for excellent technical assistance. 1. Miller, R. L., and R. Schweet, Arch. Biochem. Biophys., 125, 632 (1968). 2. Shafritz, D. A., P. M. Prichard, J. M. Gilbert, and W. F. Anderson, Biochem. Biophys. Res. Commun., 38, 721 (1970). 3. Prichard, P. M., J. M. Gilbert, D. A. Shafritz, and W. F. Anderson, Nature, 226, 511 (1970). 4. Shafritz. D. A., and W. F. Anderson, Nature, 227, 918 (1970). 5. Smith, A. E., and K. A. Marcker, Nature, 226, 607 (1970). 6. Brown, J. C., and A. E. Smith, Nature, 226, 610 (1970). 7. Shafritz, D. A., and W. F. Anderson, J. Biol. Chem., 245, 5553 (1970). 8. Shafritz, D. A., D. G. Laycock, and W. F. Anderson, Proc. Nat. Acad. Sci. USA, 68, 496 (1971). 9. Gilbert, J. M., and W. F. Anderson, J. Biol. Chem., 245, 2342 (1970). 10. Falvey, A. K., and T. Staehelin, J. Mol. Biol., 53, 1 (1970). 11. Conway, T. W., and F. Lipmann, Proc. Nat. Acad. Sci. USA, 52, 1462 (1964). 12. Bray, G. A., Anal. Biochem., 1, 279 (1960). 13. Stewart, M. L., A. P. Grollman, and M. T. Huang, Proc. Nat. Acad. Sci. USA, 68, 97 (1971). 14. MacDonald, J. S., and I. 11. Goldberg, Biochem. Biophys. Res. Commun, 41, 1 (1970).

6 Proc. Nat. Acad. Sci. USA 68 (1971) 15. Ravel, J., R. D. Mosteller, and B. Hardesty, Proc. Nat. Acad. Sci. USA, 56, 701 (1966). 16. Kolakofsky, D., K. F. Dewey, and R. E. Thach, Nature, 223, 694 (1969). 17. Chae, Y. B., R. Mazumder, and S. Ochoa, Proc. Nat. Acad. Sci. USA, 63, 828 (1969). 18. Lelong, J. C., M. Grunberg-Manago, J. Dondon, D. Gros, and F. Gros, Nature, 226, 505 (1970). 19. Bhaduri, S., N. K. Chatterjee, K. K. Bose, and N. K. Gupta, Biochem. Biophys. Res. Commun., 40, 402 (1970). 20. Housman, D., M. Jacobs-Lorena, U. L. Raj Bhandary, and H. F. Lodish, Nature, 227, 913 (1970). 21. Wilson, D. B., and H. Dintzis, Proc. Nat. Acad. Sci. USA, 66, 1282 (1970). 22. Jackson, R., and T. Hunter, Nature, 227, 672 (1970). 23. Yoshida, A., S. Watanabe, and J. Morris, Proc. Nat. Acad. Sci. USA, 67, 1600 (1970). 24. Takeishi, K., T. Sekiya, and T. Ukita, Biochim. Biophys. Acta, 199, 599 (1970). 25. Culp, W., J. Morrisey, and B. Hardesty, Biochem. Biophys. Res. Commun., 40, 777 (1970). 26. Ghosh, K., A. Grishko, and H. P. Ghosh, Biochem. Biophys. Res. Commun., 43, 462 (1971). 27. Kerwar, S. S., C. Spears, and H. Weissbach, Biochem. Biophys. Res. Commun., 41, 78 (1970). GTP and Initiation in Reticulocyte Ribosomes Moon, H. O., J. F. Collins, and E. S. Maxwell, Biochem. Biophys. Res. Commun., 41, 170 (1970). 29. Wigle, D. T., and G. H. Dixon, Nature, 227, 676 (1970). 30. Leis, J. P., and E. B. Keller, Biochem. Biophys. Res. Commun., 40, 416 (1970). 31. Tarrago, A., 0. Monasterio, and J. E. Allende, Biochem. Biophys. Res. Commun., 41, 765 (1970). 32. Marcus, A., D. P. Weeks, J. P. Leis, and E. B. Keller, Proc. Nat. Acad. Sci. USA, 67, 1681 (1970). 33. Thach, S. S., and R. E. Thach, Nature New Biol., 229, 218 (1971). 34. "The Mechanism of Protein Synthesis," Cold Spring Harbor Symp. Quant. Biol., 34 (1969). 35. Leader, D. P., I. G. Wool, and J. J. Castles, Proc. Nat. Acad. Sci. USA, 67, 523 (1970). 36. Wettenhall, R. E. H., D. P. Leader, and I. G. Wool, Biochem. Biophys. Res. Commun., 43, 994 (1971). 37. Crystal, R. G., D. A. Shafritz, P. M. Prichard, and W. F. Anderson, Proc. Nat. Acad. Sci. USA, 68, 1810 (1971). 38. Raeburn, S., J. F. Collins, H. M. Moon, and E. S. Maxwell, J. Biol. Chem., 246, 1041 (1971). 39. Marcus, A., J. Biol. Chem., 245, 955 (1970). 40. Felicetti, L., and F. Lipmann, Arch. Biochem. Biophys., 125, 548 (1968).