Proc. Nat. Acad. Sci. USA
Vol. 72, No. 11, pp. 4332-4336, November 1975 Biochemistry
Identification of a region in 23S rRNA located at the peptidyl transferase center (photoaffinity labeling/dipeptidyl-tRNA analog/antibiotics/23S rRNA/18S RNA fragment)
N. SONENBERG*, M. WILCHEKt, AND A. ZAMIR* *
Biochemistry and t Biophysics Departments, The Weizmann Institute of Science, Rehovot, Israel
Communicated by David Shemin, August 20,1975
ABSTRACT A pihotolyzable derivative of dipeptidyltRNA, p-azido-N-tBoc-Phe{3H]Phe-tRNA, bound reversibly to 70S ribosomes in the presence of poly(U), becomes, when irradiated, covalently attached to components of the 50S ribosomal subunit. Most of the reaction occurs within the 23S RNA, while ribosomal proteins are only weakly labeled. Reversible binding as well as the covalent reaction are reduced in the absence of poly(U) or in the presence of several antibiotics specific for the 50S ribosoma subunit. There are apparently few sites (or perhaps a single site) of reaction on the 23S RNA that are exclusively located within the 2000 nucleotides from the 3' terminus of the molecule. Part of the 23S RNA within this region must therefore be closely associated with the peptidyl transferase center of the ribosome.
40,000 rpm at 40 for 225 min. Gradients of type (b) were scanned continuously for UV absorption, the fractions containing 50S subunits were pooled, Mg acetate concentration was adjusted to 10 mM, and the subunits were precipitated by adding 0.7 volumes of ethanol. The ribosomal pellet was dissolved in and dialyzed against 1 mM Mg acetate, 10 mM Tris-HCl (pH 7.4). Reversible Binding of (AP)Phe-tRNA [(AP) = p-azido-NtBoc-Phe-] to 70S Ribosomes. The reaction mixture in 30 mM Mg acetate, 0.15 M NH4Cl, 50 mM Tris-HCl (pH 7.4), contained per ml: 120 Mg of poly(U), 640 pmol of (AP)[H3]Phe-tRNA ([3H]Phe, from Schwarz, 6 Ci/mmol, unless otherwise stated) synthesized as described in the Results, and 1.4 mg of 70S ribosomes. Reversible binding was assayed in 50 Ml mixtures and irradiated mixtures were in the volumes indicated in the legends to the figures. All mixtures were incubated at 370 for 20 min (referred to then as binding complexes), and the extent of reversible binding was determined by digestion with 2 ,g/ml of RNase A followed by acid precipitation (8). Millipore filtration could not be used because of the tendency of (AP)Phe-tRNA to adsorb on nitrocellulose filters. All operations were carried out under red light. Irradiation. Binding complexes were irradiated with a high pressure mercury lamp (Hanovia, 450 W) at an average distance of 8 cm, for 5 min at room temperature. Wavelengths below 250 nm were eliminated by a Corex filter. Gel Electrophoresis. rRNA was resolved by electrophoresis on a composite 1.7% acrylamide, 0.5% agarose gel (9). Bands were located by staining with Stains All (10). The gel was cut into 2 mm slices, which were incubated with 0.4 ml of NCS solubilizer (Amersham/Searle) for 3 hr at room temperature. Radioactivity was determined following the addition of scintillation fluid. Materials. Chloramphenicol, erythromycin, and puromycin were commercial products. The following antibiotics were kindly provided as gifts: lincomycin from Dr. G. B. Whitefield, Upjohn, Kalamazoo, Mich.; vernamycin A from Dr. H. L. Ennis, Roche Institute for Molecular Biology, Nutley, N.J.; and thiostrepton from Ms. B. Stearns, the Squibb Institute for Medical Research, Princeton, N.J. RNase A and RNase T1 were from Calbioche~m.
The identification of ribosomal components located within functional sites of the ribosome is of major significance in the understanding of the mechanism of action of the organelle. The great complexity of ribosome structure makes the method of affinity labeling most suitable for this purpose and studies employing, for the most part, ligands carrying electrophilic groups provided valuable information on the functional significance of specific ribosomal proteins (1). However, little is known yet on the role of the nucleic acid components of the ribosome in the construction of specific functional sites. Although several affinity labeling studies, mostly with photoreactable ligands, have indicated binding to ribosomal RNA (2-4), no detailed analysis of the reaction has been carried out to date. The present experiments show that a photoreactable derivative of dipeptidyl-tRNA, p-azido-N-tBoc-Phe-PhetRNA, reversibly bound in the presence of poly(U) to 70S ribosomes, attaches covalently, on irradiation, mostly to the 23S RNA of the 50S subunit. The site of binding has been further characterized and located within the 18S fragment formed on mild nucleolytic cleavage of 50S subunits. This region of the RNA must therefore include sequences in close proximity to the peptidyl transferase center. MATERIALS AND METHODS Ribosomes. High salt washed 70S ribosomes from Escherichia coli MRE 600 were prepared as described by Miskin and Zamir (5). The ribosomes were dissociated into subunits by overnight dialysis against 1 mM Mg acetate, 0.1 M NH4C1, and 20 mM Tris-HCl (pH 7.4). Subunits were -resolved by (a) layering 0.32 mg of dissociated ribosomes in 100 Ml on a convex exponential sucrose gradient (6, 7) in 1 mM Mg acetate, 10 mM Tris-HCl (pH 7.4), and centrifuging in the Spinco SW 50.1 rotor at 50,000 rpm at 40 for 100 min, or by (b) layering 0.8-2 mg of dissociated ribosomes in 0.60.8 ml on a linear 15-30% sucrose gradient in the same buffer as in (a) and centrifuging in the Spinco rotor SW 41 at
RESULTS
The photolyzable derivative of dipeptidyl-tRNA used in this study, p-azido-N-tBoc-Phe-Phe-tRNA, was synthesized by coupling, according to Lapidot et al. (11), the N-hydroxysuccinimide ester of p-azido-N-tBoc-phenylalanine (12) with [3H]Phe-tRNA (nonfractionated). In a standard preparation 3500 pmol of [3H]Phe-tRNA were dissolved in 3.0 ml of 0.2 M triethanolamine-HCl (pH 8.0) or 0.2 M N-2-hydroxyethylpiperazine-N'-2 ethanesulfonic acid (Hepes) (pH
Abbreviation: (AP), p-azido-N-tBoc-Phe-. 4332
Proc. Nat. Acad. Sci. USA 72 (1975)
Biochemistry: Sonenberg et al.
4333
Table 1. Reversible binding of (AP)Phe-tRNA to 70S ribosomes
(AP)Phe-tRNA Exp.
Binding mixture
bound (cpm)
1
Complete Complete - poly(U) Complete + 1 mM lincomycin Complete + 0.1 mM chloramphenicol Complete + 8.6 ,uM vernamycin A Complete Complete + 1 mM erythromycin Complete Complete + 0.1 mM thiostrepton
12,812 3,000
I
5,206
cl
2
3
9
EOL
Ct
6,920
3,240 10,149 8,583 16,548 5
11,258
Binding mixtures and assay conditions were as described in Materials and Methods. The mixture in Exp. 2 included 8.3% methanol and in Exp. 3, 6.6% dimethylsulfoxide.
8.0), and 600 mg of the N-hydroxysuccinimide ester of pazido-N-tBoc-phenylalanine in 30 ml of freshly distilled dimethylsulfoxide were added. The mixture was incubated for 2 hr at 300, 3.7 ml of 50% dichloroacetic acid were added, and, following 2 hr in the cold, the mixture was centrifuged and the pellet was washed once with dimethylformamide and twice with ethanol. The dried pellet was finally dissolved in H20 and titrated to pH 7.0 with Tris.HCI. The preparation contained at most 5% of nonreacted Phe-tRNA as determined by paper electrophoresis following alkaline hydrolysis. (AP)Phe-tRNA was tested for its ability to bind reversibly to ribosomes in response to poly(U), and in the presence of several antibiotics. The results (Table 1) indicate that the binding of the synthetic analog is about 4-fold higher in the presence of poly(U) than in its absence. The extent of poly(U)-dependent binding is comparable to that of PhetRNA under similar conditions (not shown). As seen below, the binding in the absence of poly(U) does not give rise, on irradiation, to appreciable covalent binding. The table also indicates that several antibiotics known to interact specifically with the 50S subunit suppress to different degrees the reversible binding. The most effective inhibitor is vernamycin A. Lincomycin, chloramphenicol, thiostrepton, and erythromycin are progressively less inhibitory. Binding of (AP)Phe-tRNA appears to be predominantly to the peptidyl donor (P) site, since nearly 70% of the bound material is releasable by puromycin, as determined by the amount of product extracted into ethyl acetate. These findings and the dependence on poly(U) demonstrate the specific nature of (AP)Phe-tRNA binding to the ribosomes. To induce covalent binding, a complex of (AP)Phe-tRNA and 70S ribosomes was irradiated as described, and irreversible attachment to the ribosomes was assessed by digesting the mixture with RNase A in the presence of EDTA so as to release all radioactive material associated with tRNA that had not reacted irreversibly with the ribosome (2). The results, although pointing to the existence of an irreversible reaction induced by irradiation, did not allow an accurate quantitive estimate, since a significant amount of acid-insoluble radioactivity was noticed on irradiation of (AP)PhetRNA alone. Both quantitative and qualitative characteriza-
25 20 10 15 Fraction number
FIG. 1. Distribution of label in proteins of isolated ribosomal subunits. One hundred fifty microliters of a binding complex or of a binding mixture without poly(U) were irradiated, and the ribosomes were dissociated int6 subunits. Fifty microliters of each sample were mixed with 250 ,g of untreated dissociated ribosomes in 50 ql and the subunits were fractionated by method a (Materials and Methods). Ten drop fractions of the gradients were collected, diluted two-fold with H20, and UV absorption was determined. Radioactivity associated with the ribosomal proteins was determined following digestion of the fractions with RNase A and RNase T1 according to Czernilofsky et al. (13).
tion of the reaction with the ribosome had therefore to rely on the analysis of resolved ribosomal components that reacted with (AP)Phe-tRNA, as the side-product did not interfere with the analyses of ribosomal RNA and proteins. To analyze for reacted ribosomal proteins, irradiated ribosomal complexes were dissociated at low Mg2+ concentration and resolved by sucrose gradient centrifugation (Fig. 1). Gradient fractions were digested exhaustively by RNase (13) so as to degrade all the RNA present, and radioactivity attached to the protein was determined by acid precipitation. The results show that some reaction has taken place with 50S subunit proteins but hardly any with proteins of the small ribosomal subunit. The reaction with 50S subunit proteins is much higher in the presence of poly(U) than in its absence, but altogether was too low to allow the identification of the specific proteins involved. Irradiated complexes were also analyzed for possible reaction with ribosomal RNA. Binding complexes and controls without poly(U) were, following irradiation, dissociated into their protein and RNA components, the RNA fraction was
resolved by gel electrophoresis, and the distribution of radioactivity along the gel was determined. It can be seen (Fig. 2) that significant radioactivity is associated with the 23S RNA of ribosomes irradiated in the presence of poly(U) but is nearly missing in the sample irradiated in its absence. No radioactivity co-migrates with 16S or 5S RNA but some radioactivity associated with the 4S material is evident in both samples and possibly originates in some nonreacted (AP)Phe-tRNA that has precipitated together with the ribosomal RNA. The analysis of the reaction products of (AP)Phe-tRNA with ribosomes thus indicates that both protein and RNA of the 50S subunit react with the photolyzable reagent, but the reaction with the rRNA is about 6-fold higher than the reaction with the ribosomal proteins. Altogether, about 10% of the (AP)Phe-tRNA originally bound to the ribosome in the
4334
Biochemistry: Sonenberg et al.
Proc. Nat. Acad. Sci. USA 72 (1975)
E
I -
poiy(U)
0 N
200
300 -,
100
200 E 20
40
60 80 Slice Number
120
FIG. 2. Analysis of rRNA from irradiated binding complexes. Eight hundred microliters of irradiated binding complex or of an irradiated mixture without poly(U) were each combined with 1 mg of carrier 70S ribosomes in 7 ol, mixed with an equal volume of 6 M LiCl in 8 M urea, and left at 00 for 48 hr. The RNA collected by centrifugation was dissolved in and dialyzed against 10 mM Tris-HCl (pH 7.4). Aliquots containing 1 A260 of RNA were analyzed by gel electrophoresis for 2 hr.
reversible complex became, on irradiation, covalently bound to the ribosome. The binding (80-90%) was to 23S RNA. It should be noted that a similar analysis of nonirradiated complexes did not reveal any radioactivity associated with isolated ribosomal components. Further study of the reaction centered on the ribosomal RNA, clearly the major site of attachment. The dependence on poly(U) points to the specificity of the reaction, as it indicates that the covalently bound material must have originated in the reversibly bound (AP)Phe-tRNA. Furthermore, the labeling of the 23S RNA is reduced when the binding mixture includes any one of the antibiotics shown above to inhibit the reversible binding (data not shown). The effectiveness of the different compounds tested in suppressing the covalent binding is similar but not identical to their effect on the reversible binding. It thus appears that sites within the 23S RNA are easily accessible for reaction with peptidyltRNA functionally bound to the ribosome. To estimate the number of reaction sites, the 23S RNA of the irradiated complexes was digested exhaustively with RNase T1 and the distribution of radioactivity among the oligonucleotide fragments was determined by DEAE-cellulose chromatography (Fig. 3). Prior to nucleolytic digestion the RNA preparation was incubated with 1.66 M Tris-HCl (pH 8.0) to hydrolyze the bond between the dipeptidyl moiety and tRNA (14). This should eliminate apparent heterogeneity in the distribution, due to partial hydrolysis of this linkage that might take place during the digestion and fractionation steps. The results indicate the presence of several radioactive oligonucleotide fractions, whose elution pattern does not overlap that of the bulk UV absorbing material. The shift in eluting position is most probably due to the modifying group introduced in the photochemical reaction. One peak eluting between the peaks including the pentaand hexanucleotides is, however, significantly more radioactive than the others. These results, which were reproduced in several different analyses, do not allow a definite conclusion as to the number of reaction sites, but nevertheless suggest that the reaction is limited to a few sites or perhaps a single major site in the 23S RNA. The following experiment supports the conclusion that the
I
a
Fraction number
FIG. 3. Distribution of label in an RNase T1 digest of 23S RNA. Four hundred micrograms of 50S subunits in 200 gl isolated from an irradiated binding complex were dissociated into protein and RNA components by LiCl/urea as described in the legend to Fig. 2. The RNA was dissolved in and dialyzed against H20, lyophilyzed, and redissolved in 20 ul of 1.66 M Tris-HCl (pH 8.0). The mixture was incubated at 37° for 90 min to hydrolyze peptidebound tRNA (14). Six hundred micrograms of carrier 23S RNA and 140 units of RNase T1 were added and the mixture, in a final volume of 300 M1l, was incubated for 3 hr at 37°. The digest was diluted 3-fold with 7 M urea in 10 mM Tris-HCl (pH 7.5) (equilibration buffer) and applied to a DEAE-cellulose column (0.4 X 40 cm) prepared according to Katz et al. (15) that was eluted with a linear 0-0.4 M gradient of NaCl in equilibration buffer. Fractions of 1.9 ml were collected and their UV absorption and radioactivity (in Instagel, Packard) were determined.
label is not randomly distributed, and furthermore identifies the part of the 23S RNA molecule within which the labeled sequence(s) is located. The experiment is based on the observation that mild nucleolytic digestion of 50S subunits results in the fragmentation of the 23S RNA into two unequal fragments of 13 S and 18 S (16, 17) containing the 5' and 3' termini of the molecule, respectively (18, 19). The analysis by gel electrophoresis of the RNA from nuclease-treated 50S subunits of irradiated complexes is shown in Fig. 4. The analysis of the starting material that has not been exposed to nuclease is shown for comparison. It is seen that of the two fragments produced by the nuclease, the label is exclusively associated with the larger, 18S fragment. Some radioactivity at the 18S position is also evident in the nondigested samples and presumably arises from endogenous nucleolytic cleavage. It thus appears that the site(s) on the 23S RNA closely related to the peptidyl transferase center is located in the fragment at the 3' terminus of the 23S RNA. DISCUSSION Functional and structural studies of the ribosome have so far centered mostly on the roles played by the ribosomal proteins. Various lines of evidence point, however, to the major significance of ribosomal RNA components as structural backbones of ribosomal particles (20), involved in their function (21) and specificity (22). For a detailed view of the roles of rRNA it is necessary to identify regions within the molecules that are involved in specific interactions and functions. Analyses of this nature have so far concentrated on ribosomal protein binding sites on rRNA, while the scarcity of infor-
Proc. Nat. Acad. Sci. USA 72 (1975)
Biochemistry: Sonenberg et al.
0 1
E
Slice number
FIG. 4. Distribution of label in two large fragments of 23S RNA. Sixty micrograms of 50S subunits isolated from irradiated binding complexes in 50 ,g of 5 mM Mg acetate, 5 mM Tris-HCl were treated with 0.02 ,g/ml of RNase A according to Hartman et al. (17) for 50 min at 00. The RNA was released with 0.2% Na dodecyl sulfate, an equal volume of phenol was added, and the aqueous phase was analyzed by gel electrophoresis for 3 hr. (a) RNasetreated; (b) untreated control [half the amount of (a)].
mation on regions involved in other functions is due primarily to the difficulty in their localization. The specific interaction of affinity labels with rRNA as described in this communication offers now a means of localizing sequences that are located close to, and are presumably involved in, specific ribosomal functional sites. The present experiments employed a photoreactable derivative of dipeptidyl-tRNA to study the peptidyl transferase center of the ribosome. The main advantage in using a photolyzable ligand lies in its broad chemical reactivity, in contrast to
electrophilic ligands that depend for covalent bind-
ing on the presence of properly situated nucleophilic groups. (AP)Phe-tRNA fulfilled the specificity requirements, as it bound reversibly to 70S ribosomes in response to poly(U)
and served as substrate for peptide bond formation. Reversible as well as covalent binding was inhibited by a number of antibiotics that are known to interfere with one or another facet of the activity of 50S ribosomal subunits. The effect of antibiotics on the reversible mRNA-dependent binding of peptidyl-tRNA to 70S ribosomes has not been extensively studied so far. Oen et al. (23) have observed at most a slight inhibition of the reversible binding of N-bromoacetylPhe-tRNA by chloramphenicol, lincomycin, and streptogramin (vernamycin) A. Nevertheless, covalent binding to ribosomal proteins was significantly inhibited in the presence of the antibiotics. Both the dependence on poly(U) and the inhibition by antibiotics indicate that reversibly bound (AP)Phe-tRNA is the only source of the covalently bound material. The reaction must have, therefore, taken place at a specific binding site on the ribosome. Since the reversible binding is characterized (by the puromycin release test) as being predominantly to the P-site, it is highly probable that the covalent linkage occurs mostly within this part of the peptidyl-transferase center. The reduction in irreversible binding in complexes treated with puromycin is in accord with this conclusion. The covalently bound material is distributed between both proteins and 23S RNA components of the 50S subunit, with no reaction observed with the 30S subunit. The reaction with the RNA by far exceeds that with the protein. The
4335
cause for the observed distribution is not clear and it is possible that a single mode of reversible binding of (AP)PhetRNA allows both types of reaction. Alternatively, the binding to protein may originate in a minor mode of reversible binding, e.g., to the acceptor (A) site. The covalent attachment of photolyzable peptidyl-tRNA analogs to rRNA has been reported by several workers (2, 3). Haloacetyl derivatives of puromycin (4) and more recently of Phe-tRNA (24) were also reported to react with 23S RNA. Two other photolyzable peptidyl-tRNA derivatives were reported to react mostly with the 50S subunit proteins LII and L18 (25, 26). The reagent used by Hsiung and Cantor (26) bears a close resemblance to (AP)Phe-tRNA and the cause for the difference in the results is not clear at present. The number of sites on the 23S RNA involved in the reaction has not been definitively determined but the results are consistent with the existence of one major site of reaction. A reaction of N-iodoacetylPhe-tRNA with 70S ribosomes at pH 5.0 has recently been shown also to affect possibly a single site on the 23S RNA (24). However, the conditions of the reaction are far removed from those optimal for the reversible binding of tRNA derivatives or protein biosynthesis in general. A further characterization of the reaction site depended on the specific sensitivity of 50S ribosomal subunits to mild nucleolytic digestion that results in the cleavage of 23S RNA into two large fragments. The results show clearly that the reaction occurs exclusively within the 18S fragment that contains about 2000 nucleotides at the 3' terminus of the molecule. This fragment has recently been shown to include the binding site for L2 (27) a protein shown to be located in the peptidyl transferase center (28-30). Another protein binding to the 18S fragment, L23, has been recently shown to react photochemically with puromycin (31). It thus seems possible that at least some of the proteins implicated in peptidyl transferase activity might be clustered around the region of the 23S RNA that reacts with (AP)Phe-tRNA. It has become known to us, following the completion of these experiments, that a different type of photolyzable peptidyl-tRNA analog has also been found to interact specifically with the part of 23S RNA included in the 18S fragment (32). We are deeply grateful to Dr. B. C. Dudock for his help in oligonucleotide fractionation and to Mr. D. Haik for ribosome preparations. 1. Pongs, O., Nierhaus, K. A., Erdmann, V. A. & Wittmann, H. G. (1974) FEBS Lett. 40 (suppl.) 128-137. 2. Bispink, L. & Matthaei, H. (1973) FEBS Lett. 37,291-294. 3. Girshovich, A. S., Bochkareva, E. S., Kramarov, V. A. & Ovchinnikov, Yu. A. (1974) FEBS Lett. 42,213-217. 4. Greenwell, P., Harris, R. J. & Symons, R. H. (1974) Eur. J.
Biochem. 49,539-554. 5. Miskin, R. & Zamir, A. (1974) J. Mol. Biol. 87, 121-134. 6. Noll, H. (1969) in Techniques in Protein Biosynthesis, eds. Campbell, P. N. & Sargent, J. R. (Academic Press, London), Vol. 2, pp. 101-179. 7. Ginzburg, I., Miskin, R. & Zamir, A. (1973) J. Mol. Biol. 79, 481-494. 8. Pestka, S. (1968) J. Biol. Chem. 243, 4038-4044. 9. Peacock, A. C. & Dingman, C. W. (1968) Biochemistry 7, 668-674. 10. Dahlberg, A. E., Dingman, C. W. & Peacock, A. C. (1969) J.
Mol. Biol. 41,139-147. 11. Lapidot, Y., Eilat, D., Rappoport, S. & de-Groot, N. (1970) Biochem. Biophys. Res. Commun. 190,559-563.
4A36
Biochemistry: Sonenberg et al.
12. Schwyzer, R. & Caviezel, M. (1971) Helv. Chim. Acta 54, 1395-1400. 13. Czernilofsky, A. P. & Kuechler, E. (1972) Biochim. Biophys. Acta 272,667-671. 14. Sarin, P. S. & Zamecnik, P. C. (1964) Biochim. Biophys. Acta 91,653-655. 15. Katz, G. & Dudock, B. C. (1969) J. Biol. Chem. 244, 30623068. 16. Hartman, K. A., Amaya, J. & Schachter, E. M. (1971) Science 170, 171-173. 17. Cahn, F., Schachter, E. M. & Rich, A. (1970) Biochim. Biophys. Acta 209,509-512. 18. Allet, B. & Spahr, B. F. (1971) Eur. J. Biochem. 19,250-255. 19. Saha, B. K. (1974) Biochim. Biophys. Acta 353, 292-300. 20. Held, W. A., Ballou, B., Mizushima, S. & Nomura, M. (1974) J. Biol. Chem. 249,3103-3111. 21. Noller, H. F. & Chaires, J. B. (1972) Proc. Nat. Acad. Sci. USA 69,3115-8118. 22. Held, W. A., Gette, W. R. & Nomura, M. (1974) Biochemistry 13,2115-2122.
Proc. Nat. Acad. Sci. USA 72 (1975) 23. Oen, H., Pellegrini, M. & Cantor, C. R. (1974) FEBS Lett. 45, 218-222. 24. Yukioka, M., Hatayama, T. & Morisawa, S. (1975) Biochim. Biophys. Acta 390, 192-208. 25. Hsiung, N., Reines, S. A. & Cantor, C. R. (1974) J. Mol. Biol. 88,841-855. 26. Hsiung, N. & Cantor, C. R. (1974) Nucleic Acids Res. 1, 1753-1762. 27. Spierer, P., Zimmerman, R. A. & Mackie, G. A. (1975) Eur. J. Biochem. 52, 459-468. 28. Sonenberg, N., Wilchek, M. & Zamir, A. (1973) Proc. Nat. Acad. Sci. USA 70, 1423-1426. 29. Oen, H., Pellegrini, M., Eilat, D. & Cantor, C. R. (1973) Proc. Nat. Acad. Sci. USA 70,2799-2803. 30. Sopori, M., Pellegrini, M., Lengyel, P. & Cantor, C. R. (1974) Biochemistry 13,5432-5439. 31. Cooperman, B. S., Jaynes, E. N., Brunswick, D. J. & Luddy, M. A. (1975) Proc. Nat. Acad. Sci. USA 72,2974-2978. 32. Barta, A., Kuechler, E., Branlant, C., Sriwidada, J., Kvol, A. & Ebel, J. P. (1975) FEBS Lett. 56, 170-174.
Vol. 72, No. 11, pp. 4332-4336, November 1975 Biochemistry
Identification of a region in 23S rRNA located at the peptidyl transferase center (photoaffinity labeling/dipeptidyl-tRNA analog/antibiotics/23S rRNA/18S RNA fragment)
N. SONENBERG*, M. WILCHEKt, AND A. ZAMIR* *
Biochemistry and t Biophysics Departments, The Weizmann Institute of Science, Rehovot, Israel
Communicated by David Shemin, August 20,1975
ABSTRACT A pihotolyzable derivative of dipeptidyltRNA, p-azido-N-tBoc-Phe{3H]Phe-tRNA, bound reversibly to 70S ribosomes in the presence of poly(U), becomes, when irradiated, covalently attached to components of the 50S ribosomal subunit. Most of the reaction occurs within the 23S RNA, while ribosomal proteins are only weakly labeled. Reversible binding as well as the covalent reaction are reduced in the absence of poly(U) or in the presence of several antibiotics specific for the 50S ribosoma subunit. There are apparently few sites (or perhaps a single site) of reaction on the 23S RNA that are exclusively located within the 2000 nucleotides from the 3' terminus of the molecule. Part of the 23S RNA within this region must therefore be closely associated with the peptidyl transferase center of the ribosome.
40,000 rpm at 40 for 225 min. Gradients of type (b) were scanned continuously for UV absorption, the fractions containing 50S subunits were pooled, Mg acetate concentration was adjusted to 10 mM, and the subunits were precipitated by adding 0.7 volumes of ethanol. The ribosomal pellet was dissolved in and dialyzed against 1 mM Mg acetate, 10 mM Tris-HCl (pH 7.4). Reversible Binding of (AP)Phe-tRNA [(AP) = p-azido-NtBoc-Phe-] to 70S Ribosomes. The reaction mixture in 30 mM Mg acetate, 0.15 M NH4Cl, 50 mM Tris-HCl (pH 7.4), contained per ml: 120 Mg of poly(U), 640 pmol of (AP)[H3]Phe-tRNA ([3H]Phe, from Schwarz, 6 Ci/mmol, unless otherwise stated) synthesized as described in the Results, and 1.4 mg of 70S ribosomes. Reversible binding was assayed in 50 Ml mixtures and irradiated mixtures were in the volumes indicated in the legends to the figures. All mixtures were incubated at 370 for 20 min (referred to then as binding complexes), and the extent of reversible binding was determined by digestion with 2 ,g/ml of RNase A followed by acid precipitation (8). Millipore filtration could not be used because of the tendency of (AP)Phe-tRNA to adsorb on nitrocellulose filters. All operations were carried out under red light. Irradiation. Binding complexes were irradiated with a high pressure mercury lamp (Hanovia, 450 W) at an average distance of 8 cm, for 5 min at room temperature. Wavelengths below 250 nm were eliminated by a Corex filter. Gel Electrophoresis. rRNA was resolved by electrophoresis on a composite 1.7% acrylamide, 0.5% agarose gel (9). Bands were located by staining with Stains All (10). The gel was cut into 2 mm slices, which were incubated with 0.4 ml of NCS solubilizer (Amersham/Searle) for 3 hr at room temperature. Radioactivity was determined following the addition of scintillation fluid. Materials. Chloramphenicol, erythromycin, and puromycin were commercial products. The following antibiotics were kindly provided as gifts: lincomycin from Dr. G. B. Whitefield, Upjohn, Kalamazoo, Mich.; vernamycin A from Dr. H. L. Ennis, Roche Institute for Molecular Biology, Nutley, N.J.; and thiostrepton from Ms. B. Stearns, the Squibb Institute for Medical Research, Princeton, N.J. RNase A and RNase T1 were from Calbioche~m.
The identification of ribosomal components located within functional sites of the ribosome is of major significance in the understanding of the mechanism of action of the organelle. The great complexity of ribosome structure makes the method of affinity labeling most suitable for this purpose and studies employing, for the most part, ligands carrying electrophilic groups provided valuable information on the functional significance of specific ribosomal proteins (1). However, little is known yet on the role of the nucleic acid components of the ribosome in the construction of specific functional sites. Although several affinity labeling studies, mostly with photoreactable ligands, have indicated binding to ribosomal RNA (2-4), no detailed analysis of the reaction has been carried out to date. The present experiments show that a photoreactable derivative of dipeptidyl-tRNA, p-azido-N-tBoc-Phe-PhetRNA, reversibly bound in the presence of poly(U) to 70S ribosomes, attaches covalently, on irradiation, mostly to the 23S RNA of the 50S subunit. The site of binding has been further characterized and located within the 18S fragment formed on mild nucleolytic cleavage of 50S subunits. This region of the RNA must therefore include sequences in close proximity to the peptidyl transferase center. MATERIALS AND METHODS Ribosomes. High salt washed 70S ribosomes from Escherichia coli MRE 600 were prepared as described by Miskin and Zamir (5). The ribosomes were dissociated into subunits by overnight dialysis against 1 mM Mg acetate, 0.1 M NH4C1, and 20 mM Tris-HCl (pH 7.4). Subunits were -resolved by (a) layering 0.32 mg of dissociated ribosomes in 100 Ml on a convex exponential sucrose gradient (6, 7) in 1 mM Mg acetate, 10 mM Tris-HCl (pH 7.4), and centrifuging in the Spinco SW 50.1 rotor at 50,000 rpm at 40 for 100 min, or by (b) layering 0.8-2 mg of dissociated ribosomes in 0.60.8 ml on a linear 15-30% sucrose gradient in the same buffer as in (a) and centrifuging in the Spinco rotor SW 41 at
RESULTS
The photolyzable derivative of dipeptidyl-tRNA used in this study, p-azido-N-tBoc-Phe-Phe-tRNA, was synthesized by coupling, according to Lapidot et al. (11), the N-hydroxysuccinimide ester of p-azido-N-tBoc-phenylalanine (12) with [3H]Phe-tRNA (nonfractionated). In a standard preparation 3500 pmol of [3H]Phe-tRNA were dissolved in 3.0 ml of 0.2 M triethanolamine-HCl (pH 8.0) or 0.2 M N-2-hydroxyethylpiperazine-N'-2 ethanesulfonic acid (Hepes) (pH
Abbreviation: (AP), p-azido-N-tBoc-Phe-. 4332
Proc. Nat. Acad. Sci. USA 72 (1975)
Biochemistry: Sonenberg et al.
4333
Table 1. Reversible binding of (AP)Phe-tRNA to 70S ribosomes
(AP)Phe-tRNA Exp.
Binding mixture
bound (cpm)
1
Complete Complete - poly(U) Complete + 1 mM lincomycin Complete + 0.1 mM chloramphenicol Complete + 8.6 ,uM vernamycin A Complete Complete + 1 mM erythromycin Complete Complete + 0.1 mM thiostrepton
12,812 3,000
I
5,206
cl
2
3
9
EOL
Ct
6,920
3,240 10,149 8,583 16,548 5
11,258
Binding mixtures and assay conditions were as described in Materials and Methods. The mixture in Exp. 2 included 8.3% methanol and in Exp. 3, 6.6% dimethylsulfoxide.
8.0), and 600 mg of the N-hydroxysuccinimide ester of pazido-N-tBoc-phenylalanine in 30 ml of freshly distilled dimethylsulfoxide were added. The mixture was incubated for 2 hr at 300, 3.7 ml of 50% dichloroacetic acid were added, and, following 2 hr in the cold, the mixture was centrifuged and the pellet was washed once with dimethylformamide and twice with ethanol. The dried pellet was finally dissolved in H20 and titrated to pH 7.0 with Tris.HCI. The preparation contained at most 5% of nonreacted Phe-tRNA as determined by paper electrophoresis following alkaline hydrolysis. (AP)Phe-tRNA was tested for its ability to bind reversibly to ribosomes in response to poly(U), and in the presence of several antibiotics. The results (Table 1) indicate that the binding of the synthetic analog is about 4-fold higher in the presence of poly(U) than in its absence. The extent of poly(U)-dependent binding is comparable to that of PhetRNA under similar conditions (not shown). As seen below, the binding in the absence of poly(U) does not give rise, on irradiation, to appreciable covalent binding. The table also indicates that several antibiotics known to interact specifically with the 50S subunit suppress to different degrees the reversible binding. The most effective inhibitor is vernamycin A. Lincomycin, chloramphenicol, thiostrepton, and erythromycin are progressively less inhibitory. Binding of (AP)Phe-tRNA appears to be predominantly to the peptidyl donor (P) site, since nearly 70% of the bound material is releasable by puromycin, as determined by the amount of product extracted into ethyl acetate. These findings and the dependence on poly(U) demonstrate the specific nature of (AP)Phe-tRNA binding to the ribosomes. To induce covalent binding, a complex of (AP)Phe-tRNA and 70S ribosomes was irradiated as described, and irreversible attachment to the ribosomes was assessed by digesting the mixture with RNase A in the presence of EDTA so as to release all radioactive material associated with tRNA that had not reacted irreversibly with the ribosome (2). The results, although pointing to the existence of an irreversible reaction induced by irradiation, did not allow an accurate quantitive estimate, since a significant amount of acid-insoluble radioactivity was noticed on irradiation of (AP)PhetRNA alone. Both quantitative and qualitative characteriza-
25 20 10 15 Fraction number
FIG. 1. Distribution of label in proteins of isolated ribosomal subunits. One hundred fifty microliters of a binding complex or of a binding mixture without poly(U) were irradiated, and the ribosomes were dissociated int6 subunits. Fifty microliters of each sample were mixed with 250 ,g of untreated dissociated ribosomes in 50 ql and the subunits were fractionated by method a (Materials and Methods). Ten drop fractions of the gradients were collected, diluted two-fold with H20, and UV absorption was determined. Radioactivity associated with the ribosomal proteins was determined following digestion of the fractions with RNase A and RNase T1 according to Czernilofsky et al. (13).
tion of the reaction with the ribosome had therefore to rely on the analysis of resolved ribosomal components that reacted with (AP)Phe-tRNA, as the side-product did not interfere with the analyses of ribosomal RNA and proteins. To analyze for reacted ribosomal proteins, irradiated ribosomal complexes were dissociated at low Mg2+ concentration and resolved by sucrose gradient centrifugation (Fig. 1). Gradient fractions were digested exhaustively by RNase (13) so as to degrade all the RNA present, and radioactivity attached to the protein was determined by acid precipitation. The results show that some reaction has taken place with 50S subunit proteins but hardly any with proteins of the small ribosomal subunit. The reaction with 50S subunit proteins is much higher in the presence of poly(U) than in its absence, but altogether was too low to allow the identification of the specific proteins involved. Irradiated complexes were also analyzed for possible reaction with ribosomal RNA. Binding complexes and controls without poly(U) were, following irradiation, dissociated into their protein and RNA components, the RNA fraction was
resolved by gel electrophoresis, and the distribution of radioactivity along the gel was determined. It can be seen (Fig. 2) that significant radioactivity is associated with the 23S RNA of ribosomes irradiated in the presence of poly(U) but is nearly missing in the sample irradiated in its absence. No radioactivity co-migrates with 16S or 5S RNA but some radioactivity associated with the 4S material is evident in both samples and possibly originates in some nonreacted (AP)Phe-tRNA that has precipitated together with the ribosomal RNA. The analysis of the reaction products of (AP)Phe-tRNA with ribosomes thus indicates that both protein and RNA of the 50S subunit react with the photolyzable reagent, but the reaction with the rRNA is about 6-fold higher than the reaction with the ribosomal proteins. Altogether, about 10% of the (AP)Phe-tRNA originally bound to the ribosome in the
4334
Biochemistry: Sonenberg et al.
Proc. Nat. Acad. Sci. USA 72 (1975)
E
I -
poiy(U)
0 N
200
300 -,
100
200 E 20
40
60 80 Slice Number
120
FIG. 2. Analysis of rRNA from irradiated binding complexes. Eight hundred microliters of irradiated binding complex or of an irradiated mixture without poly(U) were each combined with 1 mg of carrier 70S ribosomes in 7 ol, mixed with an equal volume of 6 M LiCl in 8 M urea, and left at 00 for 48 hr. The RNA collected by centrifugation was dissolved in and dialyzed against 10 mM Tris-HCl (pH 7.4). Aliquots containing 1 A260 of RNA were analyzed by gel electrophoresis for 2 hr.
reversible complex became, on irradiation, covalently bound to the ribosome. The binding (80-90%) was to 23S RNA. It should be noted that a similar analysis of nonirradiated complexes did not reveal any radioactivity associated with isolated ribosomal components. Further study of the reaction centered on the ribosomal RNA, clearly the major site of attachment. The dependence on poly(U) points to the specificity of the reaction, as it indicates that the covalently bound material must have originated in the reversibly bound (AP)Phe-tRNA. Furthermore, the labeling of the 23S RNA is reduced when the binding mixture includes any one of the antibiotics shown above to inhibit the reversible binding (data not shown). The effectiveness of the different compounds tested in suppressing the covalent binding is similar but not identical to their effect on the reversible binding. It thus appears that sites within the 23S RNA are easily accessible for reaction with peptidyltRNA functionally bound to the ribosome. To estimate the number of reaction sites, the 23S RNA of the irradiated complexes was digested exhaustively with RNase T1 and the distribution of radioactivity among the oligonucleotide fragments was determined by DEAE-cellulose chromatography (Fig. 3). Prior to nucleolytic digestion the RNA preparation was incubated with 1.66 M Tris-HCl (pH 8.0) to hydrolyze the bond between the dipeptidyl moiety and tRNA (14). This should eliminate apparent heterogeneity in the distribution, due to partial hydrolysis of this linkage that might take place during the digestion and fractionation steps. The results indicate the presence of several radioactive oligonucleotide fractions, whose elution pattern does not overlap that of the bulk UV absorbing material. The shift in eluting position is most probably due to the modifying group introduced in the photochemical reaction. One peak eluting between the peaks including the pentaand hexanucleotides is, however, significantly more radioactive than the others. These results, which were reproduced in several different analyses, do not allow a definite conclusion as to the number of reaction sites, but nevertheless suggest that the reaction is limited to a few sites or perhaps a single major site in the 23S RNA. The following experiment supports the conclusion that the
I
a
Fraction number
FIG. 3. Distribution of label in an RNase T1 digest of 23S RNA. Four hundred micrograms of 50S subunits in 200 gl isolated from an irradiated binding complex were dissociated into protein and RNA components by LiCl/urea as described in the legend to Fig. 2. The RNA was dissolved in and dialyzed against H20, lyophilyzed, and redissolved in 20 ul of 1.66 M Tris-HCl (pH 8.0). The mixture was incubated at 37° for 90 min to hydrolyze peptidebound tRNA (14). Six hundred micrograms of carrier 23S RNA and 140 units of RNase T1 were added and the mixture, in a final volume of 300 M1l, was incubated for 3 hr at 37°. The digest was diluted 3-fold with 7 M urea in 10 mM Tris-HCl (pH 7.5) (equilibration buffer) and applied to a DEAE-cellulose column (0.4 X 40 cm) prepared according to Katz et al. (15) that was eluted with a linear 0-0.4 M gradient of NaCl in equilibration buffer. Fractions of 1.9 ml were collected and their UV absorption and radioactivity (in Instagel, Packard) were determined.
label is not randomly distributed, and furthermore identifies the part of the 23S RNA molecule within which the labeled sequence(s) is located. The experiment is based on the observation that mild nucleolytic digestion of 50S subunits results in the fragmentation of the 23S RNA into two unequal fragments of 13 S and 18 S (16, 17) containing the 5' and 3' termini of the molecule, respectively (18, 19). The analysis by gel electrophoresis of the RNA from nuclease-treated 50S subunits of irradiated complexes is shown in Fig. 4. The analysis of the starting material that has not been exposed to nuclease is shown for comparison. It is seen that of the two fragments produced by the nuclease, the label is exclusively associated with the larger, 18S fragment. Some radioactivity at the 18S position is also evident in the nondigested samples and presumably arises from endogenous nucleolytic cleavage. It thus appears that the site(s) on the 23S RNA closely related to the peptidyl transferase center is located in the fragment at the 3' terminus of the 23S RNA. DISCUSSION Functional and structural studies of the ribosome have so far centered mostly on the roles played by the ribosomal proteins. Various lines of evidence point, however, to the major significance of ribosomal RNA components as structural backbones of ribosomal particles (20), involved in their function (21) and specificity (22). For a detailed view of the roles of rRNA it is necessary to identify regions within the molecules that are involved in specific interactions and functions. Analyses of this nature have so far concentrated on ribosomal protein binding sites on rRNA, while the scarcity of infor-
Proc. Nat. Acad. Sci. USA 72 (1975)
Biochemistry: Sonenberg et al.
0 1
E
Slice number
FIG. 4. Distribution of label in two large fragments of 23S RNA. Sixty micrograms of 50S subunits isolated from irradiated binding complexes in 50 ,g of 5 mM Mg acetate, 5 mM Tris-HCl were treated with 0.02 ,g/ml of RNase A according to Hartman et al. (17) for 50 min at 00. The RNA was released with 0.2% Na dodecyl sulfate, an equal volume of phenol was added, and the aqueous phase was analyzed by gel electrophoresis for 3 hr. (a) RNasetreated; (b) untreated control [half the amount of (a)].
mation on regions involved in other functions is due primarily to the difficulty in their localization. The specific interaction of affinity labels with rRNA as described in this communication offers now a means of localizing sequences that are located close to, and are presumably involved in, specific ribosomal functional sites. The present experiments employed a photoreactable derivative of dipeptidyl-tRNA to study the peptidyl transferase center of the ribosome. The main advantage in using a photolyzable ligand lies in its broad chemical reactivity, in contrast to
electrophilic ligands that depend for covalent bind-
ing on the presence of properly situated nucleophilic groups. (AP)Phe-tRNA fulfilled the specificity requirements, as it bound reversibly to 70S ribosomes in response to poly(U)
and served as substrate for peptide bond formation. Reversible as well as covalent binding was inhibited by a number of antibiotics that are known to interfere with one or another facet of the activity of 50S ribosomal subunits. The effect of antibiotics on the reversible mRNA-dependent binding of peptidyl-tRNA to 70S ribosomes has not been extensively studied so far. Oen et al. (23) have observed at most a slight inhibition of the reversible binding of N-bromoacetylPhe-tRNA by chloramphenicol, lincomycin, and streptogramin (vernamycin) A. Nevertheless, covalent binding to ribosomal proteins was significantly inhibited in the presence of the antibiotics. Both the dependence on poly(U) and the inhibition by antibiotics indicate that reversibly bound (AP)Phe-tRNA is the only source of the covalently bound material. The reaction must have, therefore, taken place at a specific binding site on the ribosome. Since the reversible binding is characterized (by the puromycin release test) as being predominantly to the P-site, it is highly probable that the covalent linkage occurs mostly within this part of the peptidyl-transferase center. The reduction in irreversible binding in complexes treated with puromycin is in accord with this conclusion. The covalently bound material is distributed between both proteins and 23S RNA components of the 50S subunit, with no reaction observed with the 30S subunit. The reaction with the RNA by far exceeds that with the protein. The
4335
cause for the observed distribution is not clear and it is possible that a single mode of reversible binding of (AP)PhetRNA allows both types of reaction. Alternatively, the binding to protein may originate in a minor mode of reversible binding, e.g., to the acceptor (A) site. The covalent attachment of photolyzable peptidyl-tRNA analogs to rRNA has been reported by several workers (2, 3). Haloacetyl derivatives of puromycin (4) and more recently of Phe-tRNA (24) were also reported to react with 23S RNA. Two other photolyzable peptidyl-tRNA derivatives were reported to react mostly with the 50S subunit proteins LII and L18 (25, 26). The reagent used by Hsiung and Cantor (26) bears a close resemblance to (AP)Phe-tRNA and the cause for the difference in the results is not clear at present. The number of sites on the 23S RNA involved in the reaction has not been definitively determined but the results are consistent with the existence of one major site of reaction. A reaction of N-iodoacetylPhe-tRNA with 70S ribosomes at pH 5.0 has recently been shown also to affect possibly a single site on the 23S RNA (24). However, the conditions of the reaction are far removed from those optimal for the reversible binding of tRNA derivatives or protein biosynthesis in general. A further characterization of the reaction site depended on the specific sensitivity of 50S ribosomal subunits to mild nucleolytic digestion that results in the cleavage of 23S RNA into two large fragments. The results show clearly that the reaction occurs exclusively within the 18S fragment that contains about 2000 nucleotides at the 3' terminus of the molecule. This fragment has recently been shown to include the binding site for L2 (27) a protein shown to be located in the peptidyl transferase center (28-30). Another protein binding to the 18S fragment, L23, has been recently shown to react photochemically with puromycin (31). It thus seems possible that at least some of the proteins implicated in peptidyl transferase activity might be clustered around the region of the 23S RNA that reacts with (AP)Phe-tRNA. It has become known to us, following the completion of these experiments, that a different type of photolyzable peptidyl-tRNA analog has also been found to interact specifically with the part of 23S RNA included in the 18S fragment (32). We are deeply grateful to Dr. B. C. Dudock for his help in oligonucleotide fractionation and to Mr. D. Haik for ribosome preparations. 1. Pongs, O., Nierhaus, K. A., Erdmann, V. A. & Wittmann, H. G. (1974) FEBS Lett. 40 (suppl.) 128-137. 2. Bispink, L. & Matthaei, H. (1973) FEBS Lett. 37,291-294. 3. Girshovich, A. S., Bochkareva, E. S., Kramarov, V. A. & Ovchinnikov, Yu. A. (1974) FEBS Lett. 42,213-217. 4. Greenwell, P., Harris, R. J. & Symons, R. H. (1974) Eur. J.
Biochem. 49,539-554. 5. Miskin, R. & Zamir, A. (1974) J. Mol. Biol. 87, 121-134. 6. Noll, H. (1969) in Techniques in Protein Biosynthesis, eds. Campbell, P. N. & Sargent, J. R. (Academic Press, London), Vol. 2, pp. 101-179. 7. Ginzburg, I., Miskin, R. & Zamir, A. (1973) J. Mol. Biol. 79, 481-494. 8. Pestka, S. (1968) J. Biol. Chem. 243, 4038-4044. 9. Peacock, A. C. & Dingman, C. W. (1968) Biochemistry 7, 668-674. 10. Dahlberg, A. E., Dingman, C. W. & Peacock, A. C. (1969) J.
Mol. Biol. 41,139-147. 11. Lapidot, Y., Eilat, D., Rappoport, S. & de-Groot, N. (1970) Biochem. Biophys. Res. Commun. 190,559-563.
4A36
Biochemistry: Sonenberg et al.
12. Schwyzer, R. & Caviezel, M. (1971) Helv. Chim. Acta 54, 1395-1400. 13. Czernilofsky, A. P. & Kuechler, E. (1972) Biochim. Biophys. Acta 272,667-671. 14. Sarin, P. S. & Zamecnik, P. C. (1964) Biochim. Biophys. Acta 91,653-655. 15. Katz, G. & Dudock, B. C. (1969) J. Biol. Chem. 244, 30623068. 16. Hartman, K. A., Amaya, J. & Schachter, E. M. (1971) Science 170, 171-173. 17. Cahn, F., Schachter, E. M. & Rich, A. (1970) Biochim. Biophys. Acta 209,509-512. 18. Allet, B. & Spahr, B. F. (1971) Eur. J. Biochem. 19,250-255. 19. Saha, B. K. (1974) Biochim. Biophys. Acta 353, 292-300. 20. Held, W. A., Ballou, B., Mizushima, S. & Nomura, M. (1974) J. Biol. Chem. 249,3103-3111. 21. Noller, H. F. & Chaires, J. B. (1972) Proc. Nat. Acad. Sci. USA 69,3115-8118. 22. Held, W. A., Gette, W. R. & Nomura, M. (1974) Biochemistry 13,2115-2122.
Proc. Nat. Acad. Sci. USA 72 (1975) 23. Oen, H., Pellegrini, M. & Cantor, C. R. (1974) FEBS Lett. 45, 218-222. 24. Yukioka, M., Hatayama, T. & Morisawa, S. (1975) Biochim. Biophys. Acta 390, 192-208. 25. Hsiung, N., Reines, S. A. & Cantor, C. R. (1974) J. Mol. Biol. 88,841-855. 26. Hsiung, N. & Cantor, C. R. (1974) Nucleic Acids Res. 1, 1753-1762. 27. Spierer, P., Zimmerman, R. A. & Mackie, G. A. (1975) Eur. J. Biochem. 52, 459-468. 28. Sonenberg, N., Wilchek, M. & Zamir, A. (1973) Proc. Nat. Acad. Sci. USA 70, 1423-1426. 29. Oen, H., Pellegrini, M., Eilat, D. & Cantor, C. R. (1973) Proc. Nat. Acad. Sci. USA 70,2799-2803. 30. Sopori, M., Pellegrini, M., Lengyel, P. & Cantor, C. R. (1974) Biochemistry 13,5432-5439. 31. Cooperman, B. S., Jaynes, E. N., Brunswick, D. J. & Luddy, M. A. (1975) Proc. Nat. Acad. Sci. USA 72,2974-2978. 32. Barta, A., Kuechler, E., Branlant, C., Sriwidada, J., Kvol, A. & Ebel, J. P. (1975) FEBS Lett. 56, 170-174.
