The Isolation of Eukaryotic Ribosomal Proteins THE PURIFICATION AND CHARACTERIZATION

The proteins of the small subunit of rat liver ribosomes were separated into five groups by stepwise elution from carboxymethylcellulose with LiCl at pH 6.5 (Collatz, E., Lin, A., Stöffler, G., Tsurugi, K., and Wool, I.G., (1976) J. Biol. Chem. 251, 1808-1816). From the several groups, 12 proteins (S2,S3, S4, S5, S6, S7, S8, S9, S13, S23/S24, S27, and S28) wereisolated by ion exchange chromatography on carboxymethylcellulose, by chromatography on sulfopropyl-Sephadex, and by gel filtration through Sephadex G-75. The amount of protein obtained varied from 1 to 9 mg depending on the number of steps required for the preparation; several proteins had no detectable contamination and the impurities in the others were no greater than 9%. The molecular weight of the proteins was estimated by polyazrylamide gel electrophoresis in sodium dodecyl sulfate; the amino acid composition was determined.

, and the protein was extracted from the 40 S subparticles with 67Yo acetic acid, 10 mM Tris-HCl, 33 rnM magnesium acetate (14,15). The proteins were precipitated with 9 volumes of acetone (4). The precipitate was left overnight at 4", then collected by centrifugation (45 min at 1000 x g); it was dissolved in 8 M urea that had been treated with charcoal (14) and dialyzed overnight against at least 10 volumes of 8 M urea. The concentration of protein was determined (16) using bovine serum albumin dissolved in 8 M urea as standard.
Dithiothreitol was added to the ribosomal protein solution to a final concentration of 10 rnM before it was stored at 20". Because we suspected that the proteins, when stored in urea, were being altered by carbamylation we have lately employed a different procedure.
After precipitation with acetone (see above) the proteins are dissolved in 10% acetic acid and dialyzed overnight against 10 volumes of 2%) acetic acid and stored at 20 with LiCl at pH 6.5 (10). The 16 proteins contained in the groups B40 and C40 (Sl, S2, S3, S3', S4, S5, S7, SlO, S14, S14', S15', S16, S17, S19, S20, and S28) were pooled; 350 mg of the mixture were bound to a column of carboxymethylcellulose and the protein was eluted with a linear gradient of 0 to 0.22 M LiCl (Fig. 1). there were in addition four poorly resolved peaks (Fig. 4). The proteins contained in the peaks were determined by two-dimensional polyacrylamide gel electrophoresis.
The descending portion of peak IV (designated ZVa and set off by arrows in Fig. 4) contained only S28 (Figs. 2 and 3); 9 mg of the protein were obtained (Table  I) Table  I).
The 11 proteins in D40 (S6, S8, S9, Sll, S13, S14, S15, S18, S23/S24, S25, and S26) were applied to a column of carboxymethylcellulose and eluted with a linear gradient of 0.15 to 0.32 M LiCl (Fig.  5). The individual proteins eluted over a relatively broad range of salt concentrations; there were few sharp distinct peaks and no fractions contained single proteins (Fig. 5). The fractions (numbers 250 to 340 in Fig. 5) containing S23/S24 and S6 were pooled, concentrated, and resolved by filtration through a column of Sephadex G-75 (Fig. 6). The identity and purity of the proteins were assessed (Figs. 2 and  3). It is not certain whether S23 and S24 are a single protein or two distinct polypeptides. They have never been completely resolved by two-dimensional polyacrylamide gel electrophoresis; however, the configuration of the spot had originally suggested it was two proteins and it was so designated (2). The protein we have isolated forms a single spot on two-dimensional polyacrylamide gel electrophoresis (Fig. 2), and a single band (albeit a diffuse one) after electrophoresis in one dimension in 8%, 10% (Fig. 3) Fig. 5) containing S13 and smaller amounts of S16, S18, and S23/S24 were pooled and S13 was isolated (Figs. 2 and 3) from the mixture by Sephadex G-75 filtration (results not shown). Proteins S9, Sll, S13, S15, S16, and S18 eluted together from carboxymethylcellulose in a broad peak (around fraction numbers 100 to 190 in Fig. 5). S9 was resolved from that group of proteins (Figs. 2 and 3) by rechromatography on sulfopropyl-Sephadex (results not shown). The proteins (S6, S9, S13, and S23/S24) that were purified from D40 required two procedures and only 1 to 2.6 mg were obtained ( Table I). The isolated proteins (4 to 7 pg) were analyzed by electrophoresis in polyacrylamide gels containing sodium dodecyl sulfate. The extent of the contamination was estimated by scanning the gels (Fig. 3) at 540 nm; the part of the total absorption that did not derive from the main band (in a sample shown to contain only one spot after two-dimensional polyacrylamide gel electrophoresis, Fig. 2 three proteins, S8, S27, and S30; of those, S8 and S27 eluted in a single peak (Fig. 7) when chromatographed on carboxymethylcellulose; elution was with a linear gradient of 0.25 to 0.75 M LiCl. S8 and S27 were resolved by filtration through Sephadex G-75 (Fig. 8). The purity of the proteins was assessed (Figs. 2 and 3); 2.4 and 1.2 mg were obtained (Table I). Protein S27 generally forms a series of satellite spots when TP40 is analyzed by two-dimensional polyacrylamide gel electrophoresis (2), presumably because of chemical modification during preparation of the ribosomal protein. When TP80 is analyzed only a single S27 spot is displayed on the electropherogram (2). The purified S27 formed a diffuse or incompletely separated double spot after electrophoresis in two dimensions (Fig. 2). However, the protein formed only a single band, although a broad one, after one-dimensional electrophoresis in a series of gel systems: in 8%, lo%, 12% (Fig. 3), or 15% polyacrylamide in sodium dodecyl sulfate; and in 6%, lo%, or 18% polyacrylamide in urea at pH 4.5 (results not shown).
The isolated proteins were analyzed by electrophoresis in acrylamide gels containing sodium dodecyl sulfate (Fig. 3); the purity of the proteins was estimated by scanning the same gels at 540 nm (Table I). Several proteins (S2, S3, S6, S23/S24, and S28) had no detectable contamination; the impurities in the others were no greater than 9%.
The isolated proteins had a stronger tendency to aggregate than did TP40 or fractions containing a small number of different proteins. Aggregation was probably due to disulfide bond formation, since the aggregates seemed to be dispersed by vigorous treatment with dithiothreitol. That treatment was important before reaction of the proteins with sodium dodecyl sulfate and electrophoresis to determine their purity and molecular weight. Some of the purified proteins, particularly S6, S9, S13, and S23/S24, suffered an alteration in their migration during two-dimensional polyacrylamide gel electrophoresis.
The new position occupied by the proteins, which did not differ greatly from the usual zone, appeared the result of a more negative (or less positive) charge on the protein. We assume, without evidence, that some chemical modification, perhaps carbamylation, occurred during isolation or storage of the protein. For that reason the proteins are now kept frozen in 2% acetic acid rather than 8 M urea (see "Experimental Procedures"). Molecular Weight-The molecular weight of the purified proteins was determined by electrophoresis in acrylamide gels containing sodium dodecyl sulfate (Table II).
There is general agreement that the small subunit of eukaryotic ribosomes contains about 30 proteins (1). We have isolated 12 of the proteins. There is some difficulty in correlating our results with those reported before, because of differences in methods and in nomenclature. The molecular weights reported for individual proteins are not consistent, although, there tends to be agreement on the mean values for the 30 proteins. The number average molecular weight and the range of molecular weights that have been reported are: Terao and Ogata (8), 24,600 (9,200 to 67,000); Westermann and Bielka (9), 19,209 (9,700 to 30,600); Lin and Wool (7), 25,400 (10,000 to 44,000); Howard et al. (5), 25,000 (8,000 to 39,000); Terao and Ogata (6), 23,000 (10,000 to 38,000). Terao and Ogata (8) and Westermann and Bielka (9) determined the molecular weight The easiest correlation is of the present results with those of Lin and Wool (7), since the nomenclature is the same and the determinations were in the same laboratory.
The analyses carried out on purified proteins in the present instance give molecular weights which are about 20% less than reported before (7) for the same proteins. 3 We have now determined ' An exception is the values for S28 which are the same in the two studies.

Purification of Eukaryotic
Ribosomal Proteins 4671 (results not shown) that the difference is due to the earlier analysis (7) having been of proteins contained in acrylamide gel plugs rather than on proteins eluted from the gel. A number of the proteins we have isolated can be correlated from the two-dimensional polyacrylamide electropherograms, with the reticulocyte ribosomal proteins analyzed by Howard et al. (5); in general our results and theirs (5) are in good agreement. In contrast our values seem about 10% less than those reported by Terao and Ogata (6).
Amino Acid Composition-The amino acid composition of the isolated proteins was determined (Table III). In calculating the composition only the amino acids listed in the table were taken into account; no allowance was made for tryptophan or cysteine, which were not determined, or for the additional ninhydrin-reactive compounds which were occasionally encountered. The chromatogram of the hydrolysate of protein S2 displayed an extra peak that eluted at 70 min, 18 s, somewhat before arginine (74 min, 28 s). Protein S28 had an additional peak at 58 min, 53 s, just ahead of lysine (59 min, 56 s). S27 had a similar pre-lysine peak, but it was far less prominent (in S28 the pre-lysine peak was larger than lysine) and occurred only if the acid hydrolysis had been for 72 hours. The identities of the pre-arginine (S2) and pre-lysine (S28) material have not yet been determined.
While the general pattern was similar for the several proteins (S28 excepted), the exact composition for each was unique. Protein S28 had a particularly high content of serine (15.7 mol%) and glycine (22 mol%), and strikingly low amounts, for ribosomal proteins, of lysine (2.3 mol%) and arginine (1.8 mol%).
There is only one report (9) of the amino acid composition of individual proteins from the small subunit of eukaryotic ribosomes, and the purity of the proteins analyzed was not established. We have attempted to compare our results with that report (9) using the two-dimensional electrophoretograms (which were done under somewhat dissimilar conditions) to correlate the proteins. Each of the 12 proteins for which we have the amino acid composition (Table III) was compared, using the method of Metzger et al. (20) as applied by Kalt-Schmidt et al. (21), to several possible corresponding proteins analyzed by Westermann and Bielka (9). If' one uses as criteria position on a two-dimensional electropherogram and molecular weight, than none of the proteins have the same amino acid composition; one pair of proteins, S3 in our nomenclature (2) and S5 in theirs (9), have similar compositions, but their molecular weights are 30,400 and 22,000, respectively, and they migrate to different places on the gels.
Protein S6 is the only rat liver ribosomal protein that is phosphorylated in uivo (22); there are at least five forms containing increasing numbers of phosphorserine residues. S6 was found to have a molecular weight of 31,000 and to contain 5 mol% of serine, about 14 residues/molecule. Acknowleclgments-We are grateful to Mr. Walter MacKinlay and Mrs. E. Loraine Karran for technical assistance.
Note Added in Proof-After submission of this manuscript we discovered that Martini and Gould (23) had published a comprehensive study of the molecular weight of the ribosomal proteins from several vertebrate species (but not of the proteins from the small subunit of rat liver ribosomes); the analysis was by electrophoresis with the use of sodium dodecyl sul-