Reconstitution of Bacillus stearothermophilus 50 S ribosomal subunits from purified molecular components.

Bacillus stearothermophilus 50 S ribosomal subunits have been reconstituted from a mixture of purified RNA and protein components. The protein fraction of 50 S subunits was separated into 27 components by a combination of various methods including ion exchange and gel filtration chromatography. The individual proteins showed single bands in a variety of polyacrylamide gel electrophoresis systems, and nearly all showed single spots on two-dimensional polyacrylamide gels. The molecular weights of the proteins were determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. An equimolar mixture of the purified proteins was combined with 23 S RNA and 5 S RNA to reconstitute active 50 S subunits by the procedure of Nomura and Erdmann (Nomura, M., and Erdmann, V. A. (1970) Nature 226, 1214-1218). Reconstituted 52 S subunits containing purified proteins were slightly more active than subunits reconstituted with an unfractionated total protein extract in poly(U)-dependent polyphenylalanine synthesis and showed comparable activity in various assays for ribosomal function. The reconstitution proceeded more rapidly with the mixture of purified proteins than with the total protein extract. Reconstituted 50 S subunits containing purified proteins co-sedimented with native 50 S subunits on sucrose gradients and had a similar protein compsoition. Initial experiments on the roles of the individual proteins in ribosomal structure and function were performed. B. stearothermophilus protein 13 was extracted from 50 S subunits under the same conditions as escherichia coli L7/L12, and the extraction had a similar effect on ribosomal function. When single proteins were omitted from reconstitution mixtures, in most cases the reconstituted 50 S subunits showed decreased activity in polypheylalanine synthesis.

Bacillus stearothermophilus 50 S ribosomal subunits have been reconstituted from a mixture of purified RNA and protein components. The protein fraction of 50 S subunits was separated into 27 components by a combination of various methods including ion exchange and gel filtration chromatography. The individual proteins showed single bands in a variety of polyacrylamide gel electrophoresis systems, and nearly all showed single spots on two-dimensional polyacrylamide gels. The molecular weights of the proteins were determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. An equimolar mixture of the purified proteins was combined with 23 S RNA and 5 S RNA to reconstitute active 50 S subunits by the procedure of Nomura and Erdmann (Nomura, M., and Erdmann, V. A. (1970) Nature 226, 1214-1218). Reconstituted 50 S subunits containing purified proteins were slightly more active than subunits reconstituted with an unfractionated total protein extract in poly(U)-dependent polyphenylalanine synthesis and showed comparable activity in various assays for ribosomal function. The reconstitution proceeded more rapidly with the mixture of purified proteins than with the total protein extract. Reconstituted 50 S subunits containing purified proteins co-sedimented with native 50 S subunits on sucrose gradients and had a similar protein composition. Initial experiments on the roles of the individual proteins in ribosomal structure and function were performed. B. stearothermophilus protein 13 was extracted from 50 S subunits under the same conditions as Escherichia coli L7/L12, and the extraction had a similar effect on ribosomal function. When single proteins were omitted from reconstitution mixtures, in most cases the reconstituted 50 S subunits showed decreased activity in polyphenylalanine synthesis.
The demonstration that 30 S ribosomal subunits from Escherichia coli can be reconstituted from 16 S RNA and a mixture of 21 ribosomal proteins (1) showed that all the information needed for the assembly of 30 S subunits is contained in the structures of their molecular components. Subsequently, all of the 21 proteins have been purified (for a review, see Ref. 2), and it has been demonstrated that 30 S subunits can be reconstituted from a mixture of all purified components (3). This system has been successfully exploited to study the roles of the individual RNA and protein molecules in the structure, assembly, and function of the 30 S subunit (for a review, see Ref. 4).
The 50 S subunit of Bacillus stearothermophilus can also be reconstituted from 23 S RNA, 5 S RNA, and a mixture of ribosomal proteins (5). This system has been used to study the mechanism of the in vitro assembly reaction (6) RNA (7)(8)(9)(10)(11)(12)(13)(14), the role of 23 S RNA methylation in resistance to lincomycin and spiramycin (15), and the role of protein "L3" (16). However, the system has not been useful for examining the roles of all the individual proteins of the 50 S subunit, since these proteins have not been available in purified form. We now report resolution of the protein fraction from B. stearothermophilus 50 S subunits into 27 purified components. We further show that these purified proteins can be combined with 23 S RNA and 5 S RNA to yield active 50 S subunits, and that the reconstitution system using all purified components is as efficient as the system in which an unfractionated mixture of proteins is used. Finally, we report initial experiments on the roles of the individual ribosomal proteins in polypeptide synthesis. The mutant has been used merely to provide a genetic marker to identify the strain, and the mutation has been shown to reside in a 30 S ribosomal protein. ' Presumably, the 50 S subunits are identical with those of strain 799 which has been used in the previous studies (5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16)(17)(18).

Buffers
The growth of the bacteria, the preparation of the ribosomes, and the zonal separation of ribosomal subunits have been described previously (17). The 70 S ribosomes were used as a source of 23 S RNA, since the RNA thus obtained is usually more active in reconstitution than RNA extracted from 50 S subunits (18). Then, the 70 S ribosomes were extracted with 4 M urea and 2 M LiCl, and the precipitated RNA was further extracted with 4.5 M urea/O.5 M Mg(OAc), pH 2.0, to remove residual protein, as described previously (16). The pelleted RNA was redissolved in 0.05 M Tris-Cl, pH 7.4. This RNA was free of protein, as evidenced by the absence of stained bands when RNA samples were digested with RNase and analyzed by polyacrylamide gel electrophoresis in the "standard" system (see below). The 23 S RNA was separated from 16 S RNA and 5 S RNA either by sedimentation in 5 to 20% sucrose gradients in a Beckman type SW27 rotor or by gel filtration on a column of Bio-Gel A-5m in the same buffer. The stacking gel is identical with that used in the "standard gel" system. The electrode buffer contains 14 g of glycine and 1.5 ml of acetic acid per liter. Samples containing 5 to 10 pg of each protein in 50 to 200 ~1 are mixed with an equal volume of stacking gel solution, 10 pl of 2.mercaptoethanol, and 5 pl of 0.1% pyronin B and applied to the tops of the gels. Electrophoresis is performed with the cathode at the bottom at an initial current of 1.5 mA/gel until the dye bands have passed through the stacking gel and then at a current of 3 mA/gel until the dye bands are 1 cm from the bottoms of the tubes. The gels are stained in Amido black as described above. above.
Sodium  Fig. lb shows a similar pattern for a different preparation of 50 S subunits, with electrophoresis in 8% acrylamide in the first dimension and shorter electrophoresis times in both dimensions so that Proteins 37 and 38 remain on the gel. The numbers were assigned simply by numbering the spots on a similar pattern several years ago in this laboratory. The nomenclature appeared in a previous publication (12).* Two points should be made. First, the electrophoresis pattern is different from that of Escherichia coli 50 S subunits, and no attempt was made to label analogous proteins from the two species with the same number. For example, it has been shown that B. stearothermophilus Protein 3 is immunologically related to E. coli L2 (36). Second, the numbering of spots from 1 to 38 does not imply the existence of 38 proteins. The numbers listed correspond to the proteins that have been purified in the present study. Several numbers in the original numbering system correspond to faint spots which have not been observed by us or to spots which we do not regard as distinct proteins but rather as derivatives of proteins present in other spots. The individual cases will be taken up under "Discussion".
The gel in Fig. lb also shows the presence of several small acidic spots near the top of the gel. These spots vary in intensity from one sample to the next but are always faint. Similar spots are seen in E. coli 30 S and 50 S subunits. Presumably they represent large supernatant proteins which adhere to ribosomal subunits.
The gel pattern also reveals faint spots to the upper left and lower right of Protein 30. These spots were always present but were always quite faint, and both underwent co-electrophoresis with 30 S proteins, the upper spot appearing in the region of B. stearothermophilus S15 and S17, and the lower spot appearing at the position of B we attached the prefix "L" to the protein number, e.g. Ll, L2, etc., in analogy to the nomenclature used for Escherichio coli ribosomal proteins (32). A different numbering system, also using the L prefix, has appeared, based on the two-dimensional gel pattern for Bacillus subtilis (33). In this publication we refer to the proteins by numbers alone. We feel that a tentative numbering system should be adopted until the structural and functional correspondence of each of these proteins with one of the E. coli 50 S proteins has been established, and then an L number should be assigned on the basis of this correspondence.
For absent in most preparations of 50 S subunits. No other faint spots reproducibly appeared in the gel patterns. The electrophoresis pattern of 50 S subunits was similar whether the 70 S ribosomes used as the starting material were washed with sucrose and 0.5 M KC1 or simply obtained as a crude pellet from an S30 extract. Therefore, no 50 S ribosomal proteins were removed in substantial quantities by the washing procedure. Often the spots due to supernatant proteins referred to in the last paragraph were of lesser intensity in subunits derived from salt-washed ribosomes.
One-dimensional Polyacrylamide Gel Electrophoresis- Fig.  2 shows an electrophoresis pattern of RNase-digested 50 S subunits in the "standard gel system" (see "Methods").
The proteins are resolved into many bands. Some of the bands contain two proteins, and one band, the very dark band in the middle of the pattern, contains six different proteins. The group of bands near the top of the gel contains several different large acidic proteins, among them, Protein 1. (Identification of the proteins in each band was done in the course of the present study by the two-dimensional gel electrophoresis of purified or partially purified protein preparations; see below.) In some cases, the proteins undergoing co-electrophoresis were well separated on primary phosphocellulose columns, and resolution of the proteins on one-dimensional gels was not necessary for the subsequent purification. In other cases, proteins undergoing co-electrophoresis were not resolved in the initial stages of the purification.
Since it was too laborious to perform two-dimensional electrophoresis routinely on column fractions, we devised one-dimensional gel systems similar to the first and second dimensions of the two-dimensional gels to analyze column fractions in these cases.
Such gels are shown in Fig. 3 all three of the urea gel systems, were resolved in sodium dodecyl sulfate gels.
Protein Purification-A schematic diagram of the initial stages of the purification is shown in Fig. 4. The procedure followed methods which have been described earlier (17)  The supernatant (600 ml) from the. first extraction was dialyzed overnight against 3.5 liters of Buffer VI, and a small amount of precipitate formed inside the dialysis bag. This precipitate was removed by centrifugation, and the solution was again dialyzed against Buffer VI containing 0.25 M KCl. The precipitate was dissolved in 7.5 M urea containing 10 mM phosphoric acid and was dialyzed against Buffer VI containing 0.3 M KCl. It was found to contain about 75 mg of protein consisting of roughly equal amounts of Proteins 206, 25, and 37 or 38 (or both). Most of these proteins remained in solution. This precipitation occurred to varying extents in subsequent experiments. Sometimes, the KC1 concentration could be lowered to 0.005 M without any precipitation of proteins, and on other occasions even more extensive precipitation of these three proteins occurred if the KC1 concentration was decreased below 0.2 M. We do not know what conditions affect the amount of precipitation.
The protein solution was divided into two equal parts, and each was applied to a phosphocellulose column (  Appropriate fractions from these columns were pooled and concentrated as described under "Methods," and the proteins corresponding to the various bands on one-dimensional gels were identified by two-dimensional polyacrylamide gel electrophoresis. A few of the proteins, labeled nr on the elution patterns, corresponded to the large non-ribosomal supernatant proteins referred to earlier. The elution profiles also revealed small amounts of proteins which did not correspond to any spot on the two-dimensional gel pattern, presumably 30 S proteins or non-ribosomal contaminants. These are labeled (30 S) on the elution profiles.
Figs. 5 to 7 reveal that several of the proteins were obtained partially or completely in pure form from these initial phosphocellulose and DEAE-cellulose columns, but most of the proteins required further purification. This was achieved, whenever possible, by gel filtration on Sephadex columns and in a few cases by chromatography on phosphocellulose at pH 8.5. A summary of the secondary column procedures is shown in Fig. 8. We encountered considerable difficulties in the separation of Proteins 8, 11, and 13 from the pass-through of the DEAE-cellulose column. Therefore, the following procedures were used to purify these three proteins. The 50 S subunits were first extracted with ethanol in the presence of 0 experiments, the procedure just described (that is, prior extraction of Protein 13, subsequent extraction of the other proteins with urea/LiCl, and chromatography on a single phosphocellulose column with a gradient beginning at 0.003 M KCl) appears to be a better, more straightforward procedure and avoids the difficulties encountered in the first procedure in separating Proteins 8, 11, and 13.
Purity and Zdentity of Proteins-The purity of the individual protein preparations was established by polyacrylamide gel electrophoresis with the use of the three systems described under "Methods" as well as sodium dodecyl sulfate gels and two-dimensional polyacrylamide gels. Gel patterns of the purified proteins using the "standard gel system" are shown in Fig. 9. The gels demonstrate that nearly all of the proteins are at least 95% pure, with the possible exception of Protein 8. This protein contains perhaps 10% Protein 11. Fig. 10 shows electrophoresis patterns for Proteins 8,13,22,23,26,28,30, and 32, with the use of the appropriate gel systems in order to show that these proteins are free of contaminants which would appear in the same band in the standard gel system. In addition, all of the proteins showed single bands on sodium dodecyl sulfate gels except for Protein 8 which contained a minor band of Protein 11 (see above). The identity of each protein was established by two-dimensional polyacrylamide gel electrophoresis. All proteins showed single spots on twodimensional gels, with the exception of Proteins 8, 13,30,32,37, and 38. As discussed below, we feel that the multiple spots present in these cases represent derivatives of a single protein.
The yields of the purified proteins were quite variable, about 60 mg in the best cases and as low as 10 mg in the worst, compared to a theoretical yield of about 60 mg per 10,000 molecular weight.
Molecular Weights of Proteins-The molecular weights of the proteins were determined by polyacrylamide gel electrophoresis in sodium dodecyl sulfate solutions, as described under "Methods." The results are presented in Table I. The distribution of molecular weights is similar to that of the E. coli 50 S ribosomal proteins (2), except that E. coli has no protein corresponding to Protein 1 with a molecular weight of 5.3 x 10'.
Reconstitution of 50 S Subunits Using Purified Proteins-An equimolar mixture of the purified proteins was obtained by mixing together amounts of the individual proteins proportional to their molecular weights, with the following exceptions. The mixture contained 2.5 mol of Protein 13 per mol of each of the other proteins, since E. coli 50 S subunits contain 2 to 3 copies of the analogous protein L7/L12 (39). Also, the mixture contained a 3-fold excess of Protein 20b since preliminary experiments showed that this amount was necessary in order to obtain maximal activity in reconstitution experiments. This could be due to a partial inactivation of the Protein 2Ob preparation during the purification. Also, Protein 37 was omitted since preliminary experiments showed that the presence of this protein did not stimulate, and often slightly inhibited, activity in reconstitution experiments. The purified protein mixture was dialyzed overnight against Buffer IV.
This mixture of purified proteins was compared with a total protein extract, prepared by extraction of 50 S subunits as described under "Methods," for its ability to form 50 S subunits when incubated with 23 S RNA and 5 S RNA under reconstitution conditions. Fig. 11 shows titration curves for FIG. 9. Electrophoresis of purified proteins in the standard gel system. Since the various gels were run at different times, the relative positions of the proteins are not always the same as in Fig. 2 this experiment, increasing amounts of each protein mixture were added to a constant amount of 23 S RNA and 5 S RNA, reconstitution was performed, and aliquots of the reconstitution mixtures were assayed directly for their activity in poly(U)-dependent polyphenylalanine synthesis. Fig. 11 shows that the mixture of purified proteins was actually slightly more active in reconstitution than the total protein extract, and that a plateau value was reached at a slightly lower molar ratio of proteins to RNA, about 1.5 instead of about 1.8. The fact that the activity plateaus at a molar ratio of proteins to RNA greater than 1.0 could be due to errors in estimating protein concentrations or molecular weights or partial inactivation of some of the proteins. This experiment demonstrates that the mixture of purified proteins does not lack any component present in the total protein ext,act which is essential for ribosome activity in poly(U)-dependent polyphenylalanine synthesis.  12 compares the kinetics of reconstitution using the two protein preparations, each at a concentration of 2.0 mol/mol of RNA, a value in the plateau regions of the titration curves of Fig. 10 13 shows that the bulk (about 60%) of the reconstituted 50 S subunits co-sedimented with native 50 S subunits. In addition, reconstituted 50 S subunits showed a somewhat greater tendency to aggregate than native 50 S subunits. A small amount of slowly sedimenting material, probably due to degradation of the 23 S RNA, was also present.
The protein composition of the reconstituted 50 S subunits peptidyltransferase, EF-G-dependent GTPase, EF-Tu-dependent GTPase, and ability to protect tRNA from ribonuclease digestion in a 70 S complex. This last assay (41) is a measure of the ability of 50 S subunits to associate with 30 S subunits to form functional 70 S couples. In this experiment, the particles were concentrated by pelleting before assay. We Finally, the reconstituted particles were assayed for their activity in various partial activities of protein synthesis, found that pelleting the reconstituted particles did not cause TABLE II Activities of reconstituted 50 S subunits in various assays Reconstitution mixtures contained 5.5 A,,, units of 23 S RNA, 0.33 concentration of 12.5 A pI0 units/ml and assayed as described under A 1o0 unit of 5 S RNA, and 11.0 A (,0 equivalents of proteins in 1.8 ml of "Methods." The assays contained the following amounts of particles: Buffer V. The purified protein mixture did not contain the proteins 1 polyphenylalanine synthesis, 0. 25  any decrease in activity in polyphenylalanine synthesis, and actually caused a slight increase. The other assays were performed only with pelleted particles. The results show that 50 S subunits reconstituted with the mixture of purified proteins had activities comparable to 50 S subunits reconstituted with the total protein extract in all the assays, slightly higher activity in polyphenylalanine synthesis, EF-G-dependent GTPase, and EF-T-dependent GTPase, and slightly lower activity in peptidyltransferase and Phe-tRNA protection. The values for percentage of reconstitution compared to native 50 S subunits, as examined by the various assays, were comparable, except that the reconstitution of peptidyltransferase activity was less efficient than the reconstitution of other ribosomal activities. ' Many reconstitution experiments have been done using purified 50 S components. The efficiency of reconstitution determined by the activity in poly(U)-dependent polyphenylalanine synthesis varied from about 20% to as high as 50% with various preparations of 23 S RNA (33% in Table II). Failure to get 100% efficiency is at least partly due to incomplete reconstitution of physically intact 50 S particles (cf- Fig. 13). Breakdown of some 23 S RNA appears to take place during reconstitution to a variable extent. Ethanol/~~,CE Extraction Experiments-Hamel et al. (31) showed that precipitation of E. coli 50 S subunits with ethanol in the presence of 0.5 M NH&l released a protein which was essential for polyphenylalanine synthesis and specifically for all reactions involving elongation factors G and T. This protein was identified as E. coli L7/L12 (42-44). We performed a similar experiment with B. stearother,raophilus 50 S subunits to see if they contained a protein with similar properties. 50 S subunits were extracted by the procedure of Hamel et ai. (31), and the extracted protein was isolated. Two-dimensional polyacrylamide gel electrophoresis (not shown) revealed that both of the spots labeled 13 (see Fig. 1) were completely removed by the extraction procedure. The relative intensities of these spots vary among different samples of subunits. This particular sample had a large amount of the left hand spot and a smaller amount of the right hand spot. The protein extracted 'In previous studies, using only total protein extract, the reconstitution of peptidyltransferase activity was as efficient as the reconstitution of other ribosomal activities (7,16). The reason for this difference between the present study and past work is unknown.
by the ethanol treatment showed only the left hand spot. Reconstitution experiments with these preparations are presented in Table III. The extraction procedure led to a severe loss of polyphenylalanine-synthesizing activity and a complete loss of EF-G-dependent GTPase activity. Both activities were nearly completely restored by adding back the extracted protein.
These experiments demonstrate that B. stearothermophilus Protein 13 is extracted from 50 S subunits under conditions similar to those employed for E. coli L7/L12, and that the extraction has similar effects on ribosomal activity.
Incidentally, we have found that ethanol precipitation of 50 S subunits from concentrated sucrose solutions a!so results in the removal of Protein 13. Apparently, sucrose and NH&l have similar effects in facilitating the extraction of this protein. About 80% loss of polyphenylalanine-synthesizing activity occurs when 50 S subunits are precipitated from solutions containing 38% (w/w) sucrose (the solution used in zonal centrifugation), while no loss of activity occurs if the sucrose is diluted to a concentration of less than 10% (w/w) before the addition of ethanol.
Single-component Omission Experiments-In order to determine whether the purified proteins are individually required for the activity of reconstituted 50 S subunits, reconstitution was performed with 23 S RNA, 5 S RNA, and mixtures of proteins in which a single protein was omitted in each case. These mixtures contained all of the purified proteins except Protein 37, for the reasons mentioned above. In addition, early experiments indicated that Protein 1 had no effect on the activity of reconstituted particles, except that in a few experiments the addition of this protein led to inhibition of activity as high as 30%. Therefore, Protein 1 was also omitted in the later experiments. The reconstituted particles were assayed for their activity in poly(U)-dependent polyphenylalanine synthesis. The results are summarized in Table IV. They indicate that the various proteins are required to different extents for ribosomal activity. The proteins could be divided into several groups according to the activity of "50 S subunits" reconstituted in their absence: (a) strongly required (3 to 25%), 3, 4, 5, 6, 10, 13, 18, ZOb, and 25; (b) moderately required (32 to 58%), 2,8,11,21,22,23, and 29; (c) weakly required (65 to 84%), 20a, 24, 26, 34, and 38; (d) not required (91 to lOO%), 16, 28, 30, and 32; and (e) not required and sometimes inhibitory, 1 and 37 (data not shown). Although the original numbering system designated 38 spots on the two-dimensional gel pattern, some of the spots originally numbered are very faint spots which have not been observed consistently, and in certain cases, multiple spots are probably derivatives of a single protein. These spots are as follows. Spot 7-The original pattern showed a faint spot, 7, to the left of Spot 8. This spot has not been observed in the present study. The fainter spot to the right of Spot 8 is seen only when the first dimension gel contains 4% acrylamide and, therefore, was not numbered in the original nomenclature. The purified Protein 8 preparation contains a minor amount of this spot in addition to Spot 8. No column chromatographic procedure that we employed succeeded in separating Protein 8 into two components, and a single band was observed in both the "standard" urea gel system and on sodium dodecyl sulfate gels. Treatment of total 50 S proteins with performic acid resulted in the disappearance of the double spot pattern and the appearance of a single spot which appeared to be at the same position as Spot 8.' Therefore, we regard these two spots, Spot 8 and the right hand spot, as oxidized and reduced derivatives of a single As mentioned above, all these spots are extracted from ribosomes by the ethanol/NH&l procedure used for extracting E. coli L7/L12. The proteins in the right hand and left hand spots chromatographed together on all columns that we employed. As in the case of Protein 8, treatment of total 50 S proteins with performic acid resulted in the disappearance of the right hand spots and increased intensity of the left hand spots. Hence, as with Protein 8, it appears that the right and left hand spots are reduced and oxidized derivatives of each other. In many samples, both the right and left hand regions for Protein 13 appear as a pair of spots vertically separated by a small distance (see Fig. lb, Fig. 14). Although the gel of Protein 13, shown in Fig. 10 , and no data establishing the uniqueness of all 33 is often greater with aged preparations of 50 S proteins. The proteins has appeared. We have purified 28 proteins from B. proteins in these two spots chromatographed together on stearothermophilus 50 S subunits. One of the B. stearotherphosphocellulose, and a single band was observed in the mophilus proteins, Protein 1, is larger than any of the prostandard gel system and on sodium dodecyl sulfate gels. teins present in the E. coli 50 S subunit and is dispensable Therefore, we regard these two spots as derivatives of a single for ribosomal activity if not inhibitory. Therefore, Protein 1 protein.
could be a non-ribosomal protein. Although we have not Spot 33-Spot 33 is the faint spot to the right of Spot 32. rigorously established the purity of each of our protein prep-Spot 33 is always present, but always faint. Again, the proteins arations by protein chemical analysis, it seems most likely in these two spots were purified together, and single bands were that the number of proteins in the B. stearothermophilus 50 S observed on the standard and sodium dodecyl sulfate poly-subunit is less than 33. More experiments are required to acrylamide gels. Therefore, Spot 33 is probably a derivative of establish whether this apparent difference in the number of 50 Protein 32.
S proteins between the two bacterial species is real or is a result Spot 35-Spot 35 was a bulge on the right side of Protein 34 of an error in the determination of the number of proteins for which we have not observed in this study.
one or both of the species. Spot 36-This was a spot to the left of Proteins 37 and 38. This spot is only occasionally seen on the gel pattern and is Reconstitution of 50 S Subunits with Purified Proteins very faint when present. It may represent a derivatized form of This is the first demonstration of the total reconstitution of Protein 38 (see above and Fig. 14).
50 S subunits of any species from purified molecular compo-Spots 37 and 38- Fig. 5  second report (51) giving further details and explanations for in the "standard" gel system, although usually only one band the variability of results with their system. Recently Nierhaus was observed. However, "Protein 38" stimulated the activity of and Dohme (52) reported the total reconstitution of E. coli 50 S reconstituted subunits, while "Protein 37" did not. Furthersubunits by a different approach. These studies employed more, preliminary amino acid analyses showed significant unfractionated mixtures of E. coli 50 S subunit proteins. Total differences in the amino acid compositions of purified Proteins reconstitution of active E. coli 50 S subunits with the use of 37 and 38. Yet a mixture of the two proteins gave single bands purified proteins has not yet been demonstrated. in both the "standard" gel system and on sodium dodecyl The reconstitution using all purified components demonsulfate gels. We do not have enough evidence at present to strates that the reconstitution of B. stearothermophilus 50 S decide whether Proteins 37 and 38 are one protein or two. subunits in vitro does not require any components from 30 S Further studies on these proteins will be required to settle this subunits or any non-ribosomal factors. The 23 S RNA and 5 S question.
RNA used in the reconstitution experiments were free of 16 S It should be pointed out that the occurrence of oxidized and RNA, transfer RNA, or other contaminants, as evidenced by reduced forms of a single protein which give rise to two distinct the presence of single bands on polyacrylamide gels (data not spots on two-dimensional gels has already been documented in shown). For E. coli there is genetic evidence that 50 S subunit the case of the E. coli 30 S subunit Protein S12 and S17 (3,47). assembly in uiuo may be facilitated by 30 S subunits or some In summary, then, the evidence available at the present time components thereof (53). There are additional reports suggestsuggests that each of the proteins which we have isolated and ing that a non-ribosomal "maturation factor" may play a role identified is a single component and, with the possible exception of Proteins 37 and 38, a unique species. The purity and uniqueness of the proteins can be established rigorously only OS. Mizushima, V. Erdmann, and M. Nomura, unpublished experimencs.
in 50 S subunit assembly both in uivo and in vitro (54, 55). Since our protein preparations were not all completely pure, and small amounts of 30 S subunit or non-ribosomal components may have been present in small amounts in reconstitution mixtures, we cannot completely exclude the possibility that small amounts of these impurities, acting catalytically rather than stoichiometrically, are necessary for the reconstitution. We also cannot rule out the possibility that such non-50 S subunit components may play a role in the in uiuo assembly of 50 S subunits which is not observed in vitro. However, it is reasonable to conclude that for B. stearothermophilus, active 50 S subunits can be assembled in a system containing only components found in native 50 S subunits, and that the participation of other components is not obligatory.
Further indications that the purified system does not lack any important component come from examining the properties of the reconstituted 50 S subunits. A stoichiometric mixture of the purified proteins was slightly more efficient than total protein extracts of 50 S subunits in the reconstitution of active 50 S subunits upon incubation with 23 S RNA and 5 S RNA under reconstitution conditions. The reconstituted 50 S subunits were active in poly(U)-dependent polyphenylalanine synthesis, association with 30 S subunits (Phe tRNA protection assay), EF-G-dependent GTPase, peptidyltransferase, and EF-T-dependent GTPase, although the reconstitution of peptidyltransferase activity was somewhat less efficient than the other ribosomal activities which were examined.5 The rate of reconstitution was faster with the mixture of purified proteins than with the total protein extract. Most of the reconstituted "50 S subunits" sedimented at the same rate as native 50 S subunits and had a similar protein composition. These experiments demonstrate that our mixture of purified proteins contains all the components necessary for 50 S subunit activity and for efficient reconstitution of 50 S subunits very similar in physical, chemical, and functional properties t.o native 50 S subunits.
It should be pointed out that none of the assays used to test the functional activities of 50 S subunits, including polyphenylalanine synthesis, requires the polypeptide chain termination function, and a defect in termination activity would not have been detected in our experiments. However, our present understanding of the termination function suggests that it is performed by the same components of the ribosome which are involved in the peptidyltransferase reaction (56, 57). Hence the demonstration that reconstituted 50 S subunits are active in the peptidyltransferase reaction probably indicates that they are also active in chain termination.
Roles of Individual 50 S Subunit Proteins in Protein Synthesis-The reconstitution system demonstrated in this report should be useful in studying the roles of the individual proteins in the structure and function of 50 S subunits. To demonstrate the usefulness of the system and to initiate studies on the roles of the individual components in protein synthesis, we have determined the activities of "50 S subunits" reconstituted in the absence of individual proteins in poly(U)-dependent polyphenylalanine synthesis. The results show that in most cases the resultant particles were less active than particles reconstituted from mixtures containing all the purified proteins. Such a decrease in activity indicates that the omitted protein: (a) is essential for 50 S subunit assembly, (b) is directly involved in a 50 S function, (c) is indirectly involved in ribosome function by maintaining an active center of the ribosome in an active conformation, or (d) is required for both assembly and function of 50 S subunits (cf. Ref. 4). Also, there are several proteins whose omission from reconstitution mixtures has virtually no effect on the activity of the reconstituted particles. This does not necessarily mean that these proteins play no role in the reconstruction or activity of 50 S subunits. For example, the E. coli 30 S protein S16 is not required for ribosomal activity, but the reconstitution reaction proceeds more slowly in its absence (40). The protein S18 is not required for activity when reconstituted particles are assayed directly, but when the reconstituted particles are isolated before assay, their activity is lower because Proteins Sll and S21, which require the presence of S18 for stable binding, are lost during the isolation procedure (40). The proteins which were not required for activity in the experiments reported in Table III might be of the type represented by E. coli S16 or S18. Alternatively, these proteins might be required for some other ribosome functions which are not involved in poly(U)dependent polyphenylalanine synthesis. In order to test the specific roles of individual proteins, it will be necessary to prepare 50 S subunits reconstituted in the absence of single proteins, to measure their sedimentation coefficients, determine their protein and RNA compositions, and measure their activity not only in overall polyphenylalanine synthesis but in each of the partial reactions of polypeptide synthesis, such as peptidyltransferase or EF-G-dependent GTPase. Such experiments are now in progress. Also, it should be possible to correlate the B. stearothermophilus proteins with their more extensively studied E. coli counterparts by a combination of substitution experiments and immunological techniques, as was done for the 30 S subunit proteins (34). In short, the reconstitution of B. stearothermophilus 50 S subunits from purified molecular components should prove to be a useful technique for studying the structure and function of 50 S ribosomal subunits.