Similarity in the Size and Number of Ribosomal

Transcription

1 JOURNAL OF BACTERIOLOGY, Aug. 1972, p Copyright i 1972 American Society for Microbiology Vol. 111, No. 2 Printed in U.S.A. Similarity in the Size and Number of Ribosomal Proteins from Different Prokaryotes TUNG-TIEN SUN, THOMAS A. BICKLE, AND ROBERT R. TRAUT Department of Biological Chemistry, School of Medicine, University of California, Davis, California Received for publication 3 April 1972 The ribosomal proteins from nine species of prokaryotes have been compared by polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate. The stained gels were scanned spectrophotometrically, the weight and number average molecular weights were calculated, and the detailed distribution of the proteins as a function of molecular weight was determined. By all of these criteria, the ribosomal proteins from all the species closely resembled each other, despite differences in the pattern of protein bands by conventional disc-gel electrophoresis. Therefore, it is suggested that the structural requirements for the assembly of ribosomal subunits have imposed limitations on the evolution of the ribosomal proteins and that their size has been highly conserved. The ribosomes from most, if not all, prokaryotes thus far examined are structurally similar. The large and small subunits have sedimentation coefficients of approximately 50 and 30S (21), contain 5, 16, and 23S ribonucleic acid (RNA) species (14, 21), and contain similar amounts of protein (2, 8, 19). We have postulated that the requirement that ribosomal proteins interact with 23 and 16S RNA to produce only 50 and 30S particles should place constraints on both the size and the number of proteins that can fit together to make the particle (6). This consideration leads to the prediction that the size of ribosomal proteins should be conserved during evolution, whereas sequence and charge, at least insofar as they do not affect the function of the proteins, should be less highly conserved. Until now detailed molecular weight studies of ribosomal proteins have been reported only for Escherichia coli (7, 12, 24). We have examined the ribosomal proteins of nine bacterial species representing three orders by electrophoresis at ph 4.5 in buffers containing urea and at ph 7.2 in buffers containing sodium dodecyl sulfate. The first type of gel separates proteins on the basis of both size and charge, whereas the second separates them solely on the basis of molecular weight. The results show that the weight and number average molecular weights (and therefore the number of ribosomal proteins) as well as the assortment of the ribosomal proteins into specific molecular weight classes is similar for all these species. 474 No such similarity was seen for the patterns of bands on urea gels, indicating that charge has been less highly conserved than size. The results indicate the generality in other prokaryotes of the type of size distribution first observed for E. coli ribosomal proteins. By contrast, eukaryotic ribosomal proteins are larger and more numerous (6, 11, 26). MATERIALS AND METHODS Strains. The nine bacterial species listed in Table 1 were selected from the orders of the Pseudomonadales, Eubacteriales, and Actinomycetales. Mycobacterium phlei was obtained from General Biochemicals Corp., Chagrin Falls, Ohio. Azotobacter vinelandii strain OP was obtained from P. W. Wilson, Univ. of Wisconsin, Madison. Other strains were obtained from the culture collection of the Department of Bacteriology, Univ. of California, Davis. Growth of cells. E. coli, Proteus vulgaris, and Salmonella typhimurium were grown in 1% yeast extract, 1% glucose, 2.16% K2HPO4, 1.7% KH2PO4. A. vinelandii OP was grown as described by P. W. Wilson in 2% sucrose, 0.08% K2HPO4, 0.02% KH2PO4 with an appropriate supplement of trace metals. Bacillus subtilis was grown in 2.5% tryptone, 2% yeast extract, 0.3% Na2HPO4, 3% glucose; Bacillus stearothermophilus was grown in 0.8% nutrient broth, 0.5% glucose, 0.1% Casamino Acids. M. phlei was purchased from General Biochemicals Corp. Pseudomonas testosteroni was grown in 1% yeast extract, 0.1% (NH4)2HPO4, and 0.2% KH2PO4 and was kindly.given to us by W. Benisek. All cells were grown at 37 C except B. stearothermophilus which was grown at 55 C. Cells were har-

2 VOL. 111, 1972 CONSERVATION OF SIZE OF RIBOSOMAL PROTEINS 475 TABLE 1. Bacteria used in comparison of ribosomal proteins Genus and species Order Family G + C strainm Proteus vulgaris Eubacteriales Enterobacteriaceae 38 Salmonella typhimurium Eubacteriales Enterobacteriaceae 52 Escherichia coli Eubacteriales Enterobacteriaceae 50 Bacillus stearothermophilus Eubacteriales Bacillaceae 45 + Bacillus subtilis Eubacteriales Bacillaceae 42 + Azotobacter vinelandii Eubacteriales Azotobacteriaceae 58 Pseudomonas testosteroni Pseudomonadales Pseudomonadaceae 60 Mycobacterium phlei Actinomycetales Mycobacteriaceae 68 + Micrococcus luteus Eubacteriales Micrococcaceae 72 + a G + C = molar percentage of guanine plus cytosine in DNA. The values are from Hill (9). According to Bergey's Manual of Determinative Bacteriology (7th ed., 1957). vested in mid or late logarithmic phase, washed with 0.01 M tris(hydroxymethyl)aminomethane (Tris), ph 7.5, 0.01 M Mg acetate (TM buffer) to remove contaminating growth medium and stored at -70 C. Preparation of ribosomes and supernatant fraction. Frozen cells were ground with alumina with the addition of deoxyribonuclease according to published methods (23) and were extracted with TM buffer containing M,B-mercaptoethanol. All operations were carried out at 4 to 8 C. Alumina and cell debris were removed by low-speed centrifugation, and the supernatant fluids were further centrifuged at 18,000 rev/min for 30 min in a Sorvall SS34 rotor. The upper four-fifths of each tube was carefully removed and centrifuged at 60,000 rev/min for 2 hr in a Spinco 65 rotor. The upper two-thirds of the final ribosome-free supernatant liquid was removed and constituted the S-100 supernatant fraction. The ribosome pellets were gently rinsed with TM buffer and then dissolved in 0.01 M Tris, ph 7.4, 0.01 M MgCl2, 0.1 M NH4Cl (TMN I buffer) to make the final concentration approximately equal to 20 to 30 mg of ribosomes per ml. The M. phlei ribosomes were suspended in TM buffer instead of TMN I buffer. Dissociation of subunits. For all bacterial strains except M. phlei, 0.01 M Tris, ph 7.4, M MgC12, 0.1 M NH4Cl (TMN II) was used to dissociate ribosomal subunits. Analytical sucrose gradients were used to show that ribosomes from all strains except M. phlei gave 50 and 30S particles in an equimolar ratio in TMN II buffer. The ribosomes from M. phlei were unstable in TMN II buffer but dissociated into 50 and 30S subunits in TMN I buffer. Ribosomal subunits were prepared by dialyzing the crude 70S ribosomes against a buffer appropriate for dissociation for 16 hr, followed by centrifugation in linear 7 to 25% sucrose density gradients in the same buffer for 10 hr at 26,000 rev/min in a Spinco SW27 rotor. In general, 3-ml samples of a 3 to 4 mg/ml ribosome preparation were applied to each gradient. The purity of the isolated subunits was analyzed in small sucrose density gradients, and if necessary a second cycle of purification was carried out so that the contamination of one subunit by the other was less than 3%. The purified 30S subunits from Azotobacter were found to dimerize to form a band sedimenting at about 50S when analyzed in TMN II buffer. However, the faster sedimenting component was shown to contain only 16S RNA and to have a protein composition identical to the 30S component. The purity of the 50S particles was established by the absence of both 16S RNA and characteristic 30S proteins. Disc-gel electrophoresis and gel scanning. Procedures for gel electrophoresis in sodium dodecyl sulfate (SDS) and at ph 4.5 in urea have been described previously (24). The SDS gels contained 10% acrylamide and 0.27% bisacrylamide, and the ph 4.5 urea gels contained 7.5% acrylamide with 0.2% bisacrylamide. Gels were stained with Coomassie brilliant blue (R-250) and scanned at 600 nm with a Gilford spectrophotometer. The staining intensity of Coomassie blue has been shown to be proportional to the quantity of protein present between 0.5 and 10 jig per component (5, 6); the amounts of ribosomal protein samples were chosen so that all bands fell within the linear range. Calculation of the weight and number average molecular weight. The molecular weight (m,) corresponding to each migration distance on the SDS gel was estimated by calculating a calibration curve for a series of standard proteins (6, 27). The amount of material (cl) at each distance was estimated by the intensity of staining with Coomassie brilliant blue. These values for each of the points throughout the scan were used to calculate the number and weight average molecular weights according to the equations Mn = (ZcA)/(Zcj/m1) and Mw = 2c1m1/2ci. In most cases two or more gels were analyzed for each subunit; results agreed within 5% and the average values were used. RESULTS Gel electrophoresis of ribosomal proteins. The ribosomal proteins from both subunits of each of nine species investigated were analyzed by gel electrophoresis in urea at ph 4.5. The results are shown in Fig. 1. For each subunit the gels are arranged so that closely related species are adjacent (Proteus, Salmonella, and Escherichia; B. stearothermophilus

3 476 UREA ph 4.5 SUN, BICKLE, AND TRAUT J. BACTERIOL. SOS......o 50 S 30S P Y. S.t. E.c. B1st B.sL A.v. Pt. U.p. MW PtY S.t. E.c. B.st. BIst.A.w. Pt. M.p. P.v. S.t. Cc. I.st. B.su A.v. Pt. U.p. U.I. Sitd. *...: S..t, E... t.b. sua.v. P.t. M- td FIG. 1. Gel analysis of ribosomal proteins from different bacterial species. a, ph 4.5 urea gels of the 50S ribosomal proteins. b, SDS gels of the 50S proteins. c, ph 4.5 urea gels of 30S ribosomal proteins. d, SDS gels of 30S proteins. The protein standards used to estabolish the molecular weight scale shown at the right of b and d were phosphorylase a (94,000), bovine serum albumin (67,500), ovalbumin (45,000), carboxypeptidase a (34,000), carbonic anhydrase (29,000), chymotrypsinogen (25,700),,3-lactoglobulin (18,400), and ribonuclease S (11,400). Abbreviations used: P.v. = P. vulgaris, S.t. = S. typhimurium, E.c. = E. coli, B.st. = B. stearothermophilus, B.su. = B. subtilis, A.v. = A. vinelandii, P.t. = P. testosteroni, M.p. = M. phlei, stds. = standards. The gels were prepared as described in Materials and Methods. and B. subtilis). Several points are clear from inspection of the urea gel patterns. The proteins from both subunits of the various species show distinct differences in the band patterns, and this is true even for very closely related species. SDS gels of all the ribosomal proteins are also shown in Fig. 1. The pattern of bands of 50S ribosomal proteins are very similar for all the species, indicating that the molecular weights of the various proteins are very similar. This similarity will be demonstrated quantitatively in a later section. The principal differences that are observable are seen in the two Bacillus species, which have three distinct bands of molecular weight greater than 40,000 not present in the other species. The pattern of the 30S ribosomal proteins also shows uniformity on SDS gels, but in this case the similarity is not so striking as for the 50S subunits. The protein pattern of the subunits from the related species E. coli, P. vulgaris, and S. typhimurium are almost identical, in contrast to the urea gel analysis which showed definite differences. The patterns for the two Bacillus species do however differ in detail from each other. All species have in common a large proportion of small proteins with molecular weights of 15,000 or less. A component of molecular weight approximately 70,000 is present in seven of the species but is missing from M. phlei and B. stearothermophilus. Average molecular weights of prokaryotic ribosomal proteins and the number of proteins per ribosomal subunit. The weight and number average molecular weights of the total

4 VOL. 111, 1972 CONSERVATION OF SIZE OF RIBOSOMAL PROTEINS 477 ribosomal proteins from each subunit of the nine species studied are presented in Table 2. In addition, for six of the species the supernatant proteins free from ribosomal protein were analyzed. The number of moles of protein per mole of subunit has been calculated from the average molecular weights by assuming that each subunit contains the same number of daltons of total protein as the corresponding subunit from E. coli (Table 2). The average molecular weights of all the prokaryotic ribosomal proteins from both the large and small subunits are similar. These proteins are smaller than the corresponding supernatant proteins, and the ribosomal proteins as a class are among the smallest in the cell. All of the prokaryotic ribosomal subunits contain approximately the same number of moles of protein as the corresponding subunit from E. coli. Distribution of molecular weights. Figure 2 shows by means of histograms the weight and mole fractions of the ribosomal proteins from the different subunits plotted in molecular weight classes of 5,000 between 10,000 and 80,000. There is a characteristic distribution common to all of the ribosomal protein samples. For all of them there is a peak in the distribution between 10,000 and 15,000 molecular weight, a cutoff below 10,000, and decreasing amounts of material above 20,000, so that at least 45% of the material is less than 15,000 and 70% is less than 20,000 molecular weight. This distribution was noted earlier for E. coli and has been confirmed by isolating and characterizing the individual ribosomal proteins (24). In sharp contrast, the supernatant proteins, also shown in Fig. 2, are evenly distributed between 10,000 and 80,000 molecular weight. DISCUSSION Several types of studies comparing different bacterial ribosomal proteins have been made previously: (i) Analysis on polyacrylamide gels and on carboxymethyl cellulose columns. Similarities in the proteins of closely related species were observed but few obvious correlations could be made between the ribosomal proteins of more distantly related species (18, 20). These separations depend on both size and amino acid composition in one case (electrophoresis) and primarily on composition in the second case (chromatography), and it is difficult to infer any relationship between proteins from two different kinds of ribosomes except when they are identical. (ii) Analysis of immunological cross-reaction. Antisera against several purified E. coli ribosomal proteins were tested for reaction with the ribosomal proteins from other species (28). Only within the Enterobacteriaceae was significant cross-reaction observed. Immunological methods are thus useful tools for investigating closely related species but are capable of providing only negative information for TABLE 2. Molecular weights and number of moles of ribosomal proteins per mole of particlea 30S subunit 50S subunit Supernatant proteins No. of No. of Strain moles moles Mn Mw, per Mn M, per Mn Mw mole of mole of particle particle Proteus vulgaris 18,000 25, ,000 19, ,400 45,000 Bacillus subtilis 14,500 18, ,000 23, ,000 46,500 Bacillus stearothermophilus 14,300 17, ,300 24, ,000 45,000 Escherichia coli 16,000 21, ,000 21, ,000 41,000 Salmonella typhimurium 16,000 21, ,000 19, ,000 46,000 Azotobacter vinelandii 17,000 22, ,000 21, ,400 30,800 Pseudomonas testosteroni 18,000 25, ,000 21, Mycobacterium phlei 15,500 17, ,000 20, Micrococcus luteus 17,000 26, ,000 23, a Weight and number average molecular weights were calculated from SDS gel data (7). Moles of protein per mole of subunit were calculated with the following assumptions. The subunits of the other strains were assumed to contain the same amount of protein as E. coli (330,000 daltons for the 30S particle; 600,000 daltons for the 50S particle), and the maximum and minimum number of moles of protein per particle were calculated by dividing the total daltons of protein per particle by either the number (M.,) or weight (Mw) average molecular weights.

5 478 SUN, BICKLE, AND TRAUT J. BACTERIOL. LJ Q-- 1C1 G MOLECULAR WEIGHT X 10-3 FIG. 2. Size distribution of ribosomal proteins and supernatant fractions. The histogram is derived from SDS gel data by the methods described in the text. Filled bars: percent protein in each molecular weight class of 5,000 on a weight basis. Empty bars: percent protein in each molecular weight class on a molar basis. The standard proteins used to establish the molecular weight scale are listed in the legend of Fig. 1. more distantly related species. (iii) Functional analysis. The ribosomal proteins and 16S RNA from several unrelated bacteria have been shown to be interchangeable in the reconstruction of active 30S particles from RNA and protein (17). These results are indicative of certain common structural features in both the RNA and protein components of these different species. The results presented here on the characterization of ribosomal proteins by SDS gel electrophoresis indicate that the size and number of the ribosomal proteins have been conserved, despite differences in primary structure indicated by altered electrophoretic mobility, changed behavior on ion exchange chromatography, and absence of immunological crossreactivity. This general structural similarity is entirely consistent with the functional interchangeability found by Nomura (17). The nine bacterial species used in this study were chosen from the orders of Eubacteriales, Actinomycetales, and Pseudomonadales. Their deoxyribonucleic acid (DNA) guanine plus cytosine composition, a valuable taxonomic criterion, varies from 38 to 72% (9). Five of them are gram-negative and four gram-positive, and two of them are sporeformers. The nine species therefore represent widely divergent evolutionary trends. The structure of the ribosome seems to have been conserved during the evolution of the prokaryotes, and indeed one of the main features that differentiates eukaryotes from prokaryotes is the evolution of a new type of ribosome (19). Ribosomal RNA varies far less with respect to its base composition than does DNA (1, 3, 4). The size of the ribosomal subunits and of the component RNA molecules, and the ratio of RNA to protein is similar for most prokaryotes. The principal conclusion from the experiments described here is that in addition the size and number of the prokaryotic ribosomal proteins have been conserved. Our previous studies on E. coli ribosomal proteins with the methods employed here indicated the presence of 16 to 21 proteins of number average molecular weight 16,000 on the 30S subunit and 29 to 35 proteins of number average molecular weight 17,000 on the 50S subunit (6). These values were in agreement with a number of distinct protein species identified in E. coli ribosomes by direct isolation and chemical characterization. It seems likely that, for other bacteria also, the present estimates for the number of moles of protein per particle will closely approximate the number of different kinds of proteins (10, 12, 24). The present results show that ribosomal proteins of other bacterial species closely resemble those from E. coli in size and number, the shape of the overall molecular weight distribution and, especially for the 50S proteins, in the presence of a common pattem of discrete bands with the same molecular weights. These criteria define a prokaryotic type of ribosomal protein distribution which is distinct from that of eukaryotes. The proteins from eukaryotic ribosomes are both larger and more numerous than those from prokaryotes. For example,

6 VOL. 11l, 1972 CONSERVATION OF SIZE OF RIBOSOMAL PROTEINS 479 mouse plasmocytoma ribosomes contain about 80 proteins of number average molecular weight 28,000 (6), and other types of eukaryotes have been reported to have ribosomal proteins of similar size and number (11, 26). In contrast to this distinctive difference between the ribosomal proteins, the supernatant proteins of prokaryotes and eukaryotes are quite similar in their molecular weight distributions; the molecular weight differences are characteristic of the ribosomal proteins and not of eukaryotic and prokaryotic proteins in general. A more detailed examination of the gels of the different ribosomal proteins raises several additional points. (i) The SDS gel analysis of the 50S ribosomal proteins shows greater similarity than that of the 30S proteins, and this may be related to the greater degree of stoichiometry of the 50S ribosomal proteins (15, 24). (ii) The sensitivity of the urea gel analysis to small differences between species and, indeed, between different strains of the same species (13, 20) combined with the relative insensitivity of the SDS gel pattern to quite major taxonomic differences suggests that polyacrylamide gel analysis of ribosomal proteins could be a valuable taxonomic tool. (iii) A band with the highest molecular weight of any ribosomal protein is found on the 30S subunit of seven of the nine species examined. The protein from E. coli has been examined for its function by two groups of workers; Nomura's group found the protein to be dispensable (16) whereas Kurland's has implicated the protein in the binding of poly(u) and transfer RNA to the ribosome (25). The absence of this protein in M. phlei and B. stearothermophilus raises again the question of its functional role. ACKNOWLEDGMENTS This work was supported by Public Health Service grant GM from the National Institute of General Medical Sciences, General Research Support and Cancer Research Funds from the University of California, and the American Heart Association grant Robert R. Traut is an Established Investigator of the American Heart Association. We thank Robert M. Bock for raising questions which stimulated the investigation reported here. LITERATURE CITED 1. Attardi, G., and F. Amaldi Structure and synthesis of ribosomal RNA. Annu. Rev. Biochem. 39: Bayley, S. T., and D. J. Kushner The ribosomes of the extremely halophilic bacterium, Halobacterium cutirubrum. J. Mol. Biol. 9: Belozersky, A. N., and A. S. Spirin A correlation between the composition of deoxyribonucleic and ribonucleic acids. Nature (London) 182: Belozersky, A. N., and A. S. Spirin In E. Chargaff, and J. N. Davidson, (ed.), Nucleic acids, vol. 3. Academic Press Inc., New York. 5. Bennett, J., and K. J. Scott Quantitative staining of fraction 1 protein in polyacrylamide gels using Coomassie brilliant blue. Anal. Biochem. 43: Bickle, T. A., and R. R. Traut Differences in size and number of 80S and 70S ribosomal proteins by dodecyl sulfate gel electrophoresis. J. Biol. Chem. 246: Dzionara, M., E. Kaltschmidt, and H. G. Wittman Ribosomal proteins. XIII. Molecular weights of isolated ribosomal proteins of E. coli. Proc. Nat. Acad. Sci. U.S.A. 67: Hess, E. L., and R. Horn Destruction by RNase of 30S and 50S ribosomes from Streptococcus pyogenes. J. Mol. Biol. 10: Hill, L. R An index to deoxyribonucleic acid base compositions of bacterial species. J. Gen. Microbiol. 44: Kaltschmidt, E., and H. G. Wittman Ribosomal proteins. XII. Number of proteins in small and large ribosomal subunits of E. coli as determined by twodimensional gel electrophoresis. Proc. Nat. Acad. Sci. U.S.A. 67: King, H. W. S., H. J. Gould, and J. J. Shearman Molecular weight distribution of proteins in rabbit reticulocyte ribosomal subunits. J. Mol. Biol. 61: Kurland, C. G., P. Voynow, S. J. S. Hardy, L. Randall, and L. Lutter Physical and functional heterogeneity of E. coli ribosomes. Cold Spring Harbor Symp. Quant. Biol. 34: Leboy, P. S., E. C. Cox, and J. G. Flaks The chromosomal site specifying a ribosomal protein in E. coli. Proc. Nat. Acad. Sci. U.S.A. 52: Loening, U. E Molecular weights of ribosomal RNA in relation to evolution. J. Mol. Biol. 38: Mora, G., D. Donner, P. Thammana, L. Lutter, and C. G. Kurland Purification and characterization of 50S ribosomal proteins of E. coli. Mol. Gen. Genet. 112: Nomura, M., S. Mizushima, M. Ozaki, P. Traub, and C. V. Lowry Structure and function of ribosomes and their molecular components. Cold Spring Harbor Symp. Quant. Biol. 34: Nomura, M., P. Traub, and H. Bechman Hybrid 30S ribosomal particles reconstituted from components of different bacterial origins. Nature (London) 219: Otaka, E., T. Itoh, and S. Osawa Ribosomal proteins of bacterial cells: strain and species specificity. J. Mol. Biol. 33: Spirin, A. S., and L. P. Gavrilova The ribosome, p. 28. Springer-Verlag, New York. 20. Sypherd, P. S., D. M. O'Neil, and M. M. Taylor The chemical and genetic structure of bacterial ribosomes. Cold Spring Harbor Symp. Quant. Biol. 34: Taylor, M. M., J. E. Glasgow, and R. Storck Sedimentation coefficients of RNA from 70S and 80S ribosomes. Proc. Nat. Acad. Sci. U.S.A. 57: Taylor, M. M., and R. Storck Uniqueness of bacterial ribosomes. Proc. Nat. Acad. Sci. U.S.A. 52: Tissieres, A., J. D. Watson, D. Schlessinger, and B. R. Hollingworth Ribonucleoprotein particles from E. coli. J. Mol. Biol. 1: Traut, R. R., H. Delius, C. Ahmad-Zadeh, T. A. Bickle, P. Pearson, and A. Tissieres Ribosomal proteins of E. coli: stoichiometry and implication for ri-

7 480 SUN, BICKLE, AND TRAUT J. BACTERIOL. bosome structure. Cold Spring Harbor Symp. Quant. Biol. 34: Van Duin, J., and C. G. Kurland Functional heterogeneity of the 30S ribosomal subunit of E. coli. Mol. Gen. Genet. 109: Warner, J. R The assembly of ribosomes in yeast. J. Biol. Chem. 246: Weber, K., and M. Osborn The reliability of molecular weight determinations by dodecyl sulfate polyacrylamide gels. J. Biol. Chem. 244: Wittmann, H. G., G. Stoffler, E. Kaltschmidt, V. Rudloff, H. G. Janda, M. Dzionara, D. Donner, K. Nierhaus, M. Cech, I. Hindennach, and B. Wittmann Protein-chemical and serological studies on ribosomes of bacteria, yeast and plants. Fed. Eur. Biochem. Soc. Symp. 21:33-46.