Eur. J . Biochem. 102, 573-582 (1979)

The 5-S RNA Binding Protein from Yeast (Saccharomyces cerevisiae) Ribosomes Evolution of the Eukaryotic 5-S RNA Binding Protein

Ross N. NAZAR, Makoto YAGUCHI, Gordon E. WILLICK, C. Fernand ROLLIN, and Camille ROY

Department of Botany and Genetics, University of Guelph, and Division of Biological Sciences, National Research Council of Canada, Ottawa

(Received February R/June 14, 1979)

The ribonucleoprotein complex between 5-S RNA and its binding protein (5-S RNA . protein complex) of yeast ribosomes was released from 60-S subunits with 25 mM EDTA and the protein component was purified by chromatography on DEAE-cellulose. This protein, designated YL3 (MI = 36000 on dodecylsulfate gels), was relatively insoluble in neutral solutions (pH 4-9) and migrated as one of four acidic 60-S subunit proteins when analyzed by the Kaltschmidt and Wittman two-dimensional gel system. Amino acid analyses indicated lower amounts of lysine and arginine than most ribosomal proteins. Sequence homology was observed in the N terminus of YL3, and two prokaryotic 5-S RNA binding proteins, EL 18 from Escherichia coliand HL 13 from Halobacterium cutirubrum : Ala'-Phe2-Gln3-Lys4-AspS-Ala6-Lys7- Ser8- Ser9-Ala'o-Tyr''- Ser'2-Ser'3-Arg'4-Phe's- Gln'6-Tyr'7-Pro's-Phe'9-Arg20-Arg21-Arg22-Arg23-G1~24-Gly2s-Ly~26-Thr27-A~p28-Tyr29-Tyr3s ;of particular interest was homology in the cluster of basic residues (18-23). Since the protein contained one methionine residue it could be split into two fragments, CN1 (MI = 24700) and CN2 ( M , = 11 300) by CNBr treatment; the larger fragment originated from the N terminus. The N-terminal amino acid sequence of CN2 shared a limited sequence homology with an internal portion of a second 5-S RNA binding protein from E. coli, EL5, and, based also on the molecular weights of the proteins and studies on the protein binding sites in 5-S RNAs, a model for the evolution of the eukaryotic 5-S RNA binding protein is suggested in which a fusion of the prokaryotic sequences may have occurred. Unlike the native 5-S RNA . protein complex, a variety of RNAs interacted with the smaller CN2 fragment to form homogeneous ribonucleoprotein complexes; the results suggest that the CN1 fragment may confer specificity on the natural 5-S RNA-protein interaction.

Ribonucleoprotein complexes between 5-S RNAs and their binding protein(s) (5-S RNA . protein) have been released from both prokaryotic and eu- karyotic ribosomes [l -61. While all contain a single RNA molecule of nearly identical molecular weight, they vary widely in the number and size of their protein components. Bacterial complexes contain at least two proteins [2,3] and, depending on the organism or method of preparation, a third protein has been observed [3,7]. All of the eukaryotic com- plexes which have been examined contain only one somewhat larger protein component [l, 41.

Assuming that all 5-S RNA . protein complexes have a common role in ribosome structure or function (e.g. tRNA binding [8] or subunit association [9]).

~~~

This paper is issued as NRCC No. 17715

heterogeneity in the protein components poses an interesting inconsistency. If the differences are not due simply to differences in extraction, it is attractive to postulate that they may result from gene fusion during the course of evolution. Studies on the molecular weights of the proteins and on protein binding sites in 5-S RNAs both suggest the latter. For example, the molecular weights of the three protein components (EL5, EL18, EL25) in the Escherichia coli 5-S RNA . protein complex, as determined from their primary amino acid sequences [lo- 121, are 20172, 12 770 and 10 695, respectively, a total molecular weight of 43637. The molecular weights of the two equivalent proteins (HL13 and HL19) form Halo- bacterium cutirubrum [6], as determined by sedimen- tation equilibrium, are 18 700 and 18 000 respectively a total molecular weight of 36700; and the single

574 Yeast 5-S R N A Binding Protein

protein from yeast [4] or rat liver [I] ribosomes, as determined by gel electrophoresis in sodium dodecyl sulfate, is about 38000. Despite differences in the number of protein components, the total molecular weight is similar in each case. Furthermore, studies on primary protein binding sites in the 5-S RNAs indicate that this region is common to all 5-S RNAs. All three 5-S RNA binding proteins from E. coli [ l ] appear to interact with approximately the same region of the 5-S RNA molecule as the two protein components in the H. cutirubrum complex [13] or as the single protein molecule in the yeast complex [14]. Again, despite differences in the number of proteins, there appears to be equivalence in the total protein content.

In this study, the structure of a 5-S RNA binding protein from a eukaryote, Saccharomyces cerevisiae, was examined. The results indicate that this protein bears a sequencial homology with its prokaryotic counterparts and further suggest that this single protein may be equivalent to the two or more 5-S RNA binding proteins of prokaryotic ribosomes.

MATERIALS AND METHODS

Growth Conditions and Isolation of Ribosomal Subunits

Saccharomyces cerevisiae, strain S288C, grown aerobically in 0.3 %yeast extract, 0.5 ”/, Bacto-peptone, and 2 glucose at 28 “C, were harvested in late log phase. Ribosomes and ribosomal subunits were pre- pared essentially as described by Torano et al. [15] with minor modification. Briefly, 200 g of cells were suspended in 300 ml of 10 mM KCl, 5 mM magnesium acetate, 10 mM 2-mercaptoethanol, and 10 mM Tris- HCl, pH 7.5, 1 vol. glass beads (0.1 mm) was added to the suspension and the cells were disrupted for 20 min in an Eppenbach colloid mill. The cell homo- genate was centrifuged at 10000 x g for 20 min and then the supernatant was centrifuged at 35000 x g for 20 min. Ribosomes were collected from the super- natant by centrifugation at 105000 x g for 2 h. Ribo- somal subunits were prepared by suspending the ribosomal pellet in 0.8 M KCl, 12 mM magnesium acetate, 10 mM 2-mercaptoethanol, and 50 mM Tris- HCl, pH 7.5 (100- 200A260 units/ml) and fractionated on a 10-30”/;; linear sucrose gradient at 37000 rev./ min for 4 h in a Beckman B-4 rotor. The 60-S fraction was collected by centrifugation at 250000 x g for 18 h.

Extraction and Purification of’the 5-S RNA . Protein Complex and the Binding Protein

The 5-S RNA . protein complex was released from 60-S subunits [ l ] by suspending the pellet in 25 mM EDTA, pH 7.0 (100 A260 units/ml); the ribosomal

core was subsequently removed by centrifugation at 226000 x g for 18 h. In some experiments the ribonu- cleoprotein complex was further purified by centrif- ugation on a 5 -20 ”/, linear sucrose gradient [l] . The 5-S RNA binding protein was prepared from the 5-S RNA . protein complex by applying the supernatant (100-200 A260 units/ml) to a column (1.5 x 2.0 cm) of DEAE-cellulose (Whatman DE 52). Minor contaminants were washed away with 20 ml of 0.15 M KCl, 10 mM Tris-HC1, pH 7.5, and the protein was eluted as a fine precipitate with 20 rnl of freshly pre- pared 0.3 M KCl, 6 M urea, 10 mM Tris-HC1, pH 7.5. One drop of conc. HCl was added to dissolve the protein and to prevent cyanate formation, the eluant was dialyzed exhaustively against distilled water and the protein was recovered by freeze-drying.

Identification and Characterization of the 5-S R N A . Protein Complex

The RNA . protein complex was identified and its purity was assessed by electrophoresis on an 8 % polyacrylamide slab gel at pH 8.3 [16]. For studies on the effect of pH on complex stability the complex solution (25 mM EDTA, pH 7.0) was titrated to the appropriate pH with acetic acid or sodium hydroxide and analyzed by gel electrophoresis as indicated above. For studies on the reassociation of ribonucleoprotein complexes the components were incubated under appropriate conditions before analysis by gel electro- phoresis.

Identification and Characterization of the 5-S RNA Binding Protein

The 5-S RNA binding protein was identified and characterized by polyacrylamide gel electrophoresis. Ribosomal proteins were extracted from ribosomal subunits by the magnesium/acetic acid method of Hardy et al. 1171 and analyzed by two-dimensional polyacrylamide gel electrophoresis as described by Kaltschmidt and Wittmann [18]. Molecular weights were determined by sodium dodecyl sulfate/poly- acrylamide gel electrophoresis [19]. Ribonuclease A chymotrypsin, ovalbumin and bovine serum albumin were used as the calibration proteins.

Protein Sequence Analysis

The primary amino acid sequence was examined by amino acid analysis and partial sequence de- termination. Amino acid compositions were de- termined on a Durrum D-500 amino acid analyzer; the protein fractions were hydrolyzed for 20 or 70 h at 110 ‘C in 6 M HCI. Tryptophan was estimated by both the method of Edelhoch [20] using 6 M guanidine

HCl and the method of Liu and Chang [21]. Cysteine

R. N. Nazar, M. Yaguchi, G. E. Willick, C. F. Rollin, and C. Roy 575

A a b C

B d e f

RNA * protein

5.8-S RNA 5-S RNA

YL3

Fig. 1. Electrophoretic unulyses of the 5-S R N A . protein complex isolutedjrom S . cerevisiae 60-S ribosomal subunits. (A) Electrophoresis of the complex on an 8 % polyacrylamide gel a t pH 8.3. Aliquots were mixed with equal volumes of dodecylsulfate extraction buffer (a) or 25 mM EDTA, pH 7.0 (b) and applied directly to the gel as indicated. The gel was stained with methylene blue; whole 60-S subunit RNA was included as a marker (c). (B) Electrophoresis of the complex on a 10% sodium dodecyl sulphate/polyacrylamide gel. Aliquots of the 5-S RNA . protein complex (e), proteins extracted from whole 6 0 3 subunits (d), or the 6 0 3 subunit core after EDTA extraction (f) were applied directly to the gel as indicated. The gel was stained with Coomassie blue

was determined after converting to cysteic acid [22]. The N-terminal sequence of the total protein or of digestion fragments was determined by automatic degradation with a Beckman model 890C protein sequenator with a 0.5 M quadrol program (no. 122974). The thiazolinone or phenylthiohydantoin derivatives were hydrolyzed with 6 M HCl in the presence or absence of 0.1 % SnC12 [23] at 130 "C for 20 h, and the amino acids formed were analyzed with Durrum D-500 amino acid analyzer. The identification of some phenylthiohydantoin derivatives (Asp, Asn, Glu, Gln) was made by thin-layer chromatography on silica gel plates [24,25]. Large fragments of the 5-S RNA binding protein were produced by treating the protein with cyanogen bromide [26] and purified by gel filtration on a column (150 x 1.5 cm) of Se- phacryl S-200 (Pharmacia Fine Chemicals AB, Swe- den) eluted with 5 acetic acid in 6 M urea at room temperature at a flow rate of 7 ml/h. Appropriate fractions were pooled, dialyzed exhaustively against distilled water, and freeze-dried. The amino acid composition and N-terminal sequences of each frag- ment were determined as described above.

RESULTS

As has previously been observed with rat liver ribosomes [l], a 5-S . RNA protein complex could be effectively released from yeast ribosomes or 60-S subunits with 25 mM EDTA (Fig. 1). Similarly, after dissociation with sodium dodecyl sulfate (Fig. l), the complex was found to contain only two compo-

nents, 5-S ribosomal RNA and a single ribosomal protein (YL3). As estimated by dodecyl sulfate gel electrophoresis, the protein component had a molec- ular weight ofabout 36000, apparently the third largest protein component in the 60-S subunit (Fig.1). In preliminary experiments the complex was purified by removing the ribosomal core with a single centrif- ugation step (see Materials and Methods) and re- purified by sucrose gradient centrifugation. The second centrifugation was subsequently found to be un- necessary provided the initial centrifugation was sufficiently long and in most experiments the ribo- nucleoprotein was purified by the abbreviated pro- cedure. Traces of free 5-S RNA were always present (Fig, 1); however, dissociation to free RNA could be minimized by maintaining a relatively high RNA concentration (> 5 ,4260 units/ml).

The 5-S RNA binding protein (YL3) was routinely purified from the ribonucleoprotein complex by a single chromatographic step on DEAE-cellulose. The protein was dissociated from the complex and eluted with 0.3 M KCl, 6 M urea, 10 mM Tris-HC1, pH 7.5, leaving behind the 5-S RNA component. Unexpect- edly, because of its very low solubility, the protein was recovered as a fine precipitate in the eluent; it could be solubilized in larger volumes of eluting buffer or by reducing the pH to 3.5. Because neutral urea solutions contain cyanate which can block the amino termini of proteins, the pH was routinely reduced immediately after elution and the protein was therefore fully dissolved. During subsequent dialysis the protein was again frequently observed to precipitate out of solution in distilled water and was

576 Yeast 5-S RNA Binding Protein

A B C

Fig. 2. IdentSfftution of'thr 5-S RNA-binding protein from S . cerevisiae 60-S ribo.romnl sirbunit.,. using two-rhrnsionul gel rlrttrophora~is. The total protein fraction from the 5-S RNA . protein complex (B), whole 60-S ribosomal subunits (C) or the 6 0 3 core (A) was fractionated by two-dimensional gel electrophoresis using the Kaltschmidt-Wittmann system [18]. A typical 6 'x polyacrylamide first-dimersion gel (pH 8.6) is also shown (D) for the 5-S RNA binding protein (YL3)

difficult to dissolve in neutral buffers in the absence of denaturing agents such as urea or sodium dodecyl sulfate. 5-S RNA could also be recovered from the DEAE-cellulose column with 0.5- 1.0 M KCI, 6 M urea, 10 mM Tris-HCI, pH 7.5; this was found to be a very convenient method for preparing unlabeled RNA.

When the RNA binding protein was identified by the standard two-dimensional electrophoresis system of Kaltschmidt and Wittmann [18], two striking fea- tures were noted. As shown in Fig.2, the binding protein (YL3) was relatively acidic; in the first di- mension it migrated towards the anode. Only three other major 60-S subunit proteins from yeast ribosomes migrate towards the anode: YL44 and YL45 which migrate off the gel under the electrophoretic conditions used [27] and YL24 [27] which migrated faster than YL3 in both dimensions (Fig. 2). Further- more, as shown in the one-dimensional insert, the 5-S RNA binding protein was largely insoluble in the first dimension buffer and most of the protein remained trapped at the origin. As a result, this protein appeared as a streaking minor component in two-dimensional analyses of yeast ribosomal proteins and has previously been considered as non-ribosomal [27]. As shown in Fig. 2, this minor protein observed in two-dimensional analyses of 6 0 3 subunit proteins clearly corresponds in migration to the purified yeast 5-S RNA binding

protein and is absent from the core when the 5-S RNA . protein complex is released with EDTA. Also, as observed with sodium dodecyl sulfate gel electro- phoresis, this protein is the third slowest migrating protein in the second dimension ; accordingly, we have labeled it YL3.

The low solubility of YL3 at neutral pH was observed to correlate in an interesting fashion with the pH dependence of the 5-S RNA . protein complex. As shown in Fig. 3, the complex is most stable in the neutral pH range and is fully and irreversibly denatured at pH 3.5, the pH at which the protein becomes generally soluble. Attempts to reform the RNA . pro- tein complex from purified components have been largely unsuccessful. 5-S RNA, pretreated at neutral or acid pH, readily exchanged into the ribonucleo- protein complex (e.g. Fig.4) but purified YL3 ag- gregates at neutral pH, even when dissolved with urea and dialyzed into 25 mM EDTA, pH 7.0. The protein aggregates interact with the 32P-labeled 5-S RNA but, because the aggregates did not enter the gel, only a very small amount of radioactivity was present as a smeared band at the 5-S RNA . protein complex migration position (results not shown). These results indicate that the solubilization of the RNA binding protein at acid pH denatures the protein in a relatively irreversible fashion and caution must be taken when studies are conducted on the reasso-

R. N . Nazar, M. Yaguchi, G. E. Willick, C. F. Rollin, and C. Roy 577

0 2.0 4.0 6.0 8.0 10.0

PH

Fig.3. EfJkct q f p H on the 5-S RNA .protein complex,from S . cere- visiae. Aliquots, dissolved i n 25 mM EDTA, pH 7.0, were titrated to the appropriate pH with acetic acid or sodium hydroxide and applied directly to an 8 '%, polyacrylamide gel for electrophoresis at pH 8.3. After staining with inethylene blue, the gels were scanned at 570 nm in a Gilford model 240 spectrophotometer equipped with a model 2140 linear transport. The amount of RNA . protein at each pH is expressed as a percentage of the total absorbance (free 5-S RNA + RNA . protein)

CN2 , RNA - protein A B C I A B C

Table 1, Amino acid composition of the 5-S R N A binding protein from yeast (YL3) and two,frugments (CNl and CN2) ohtnined h j cyanogen bromide cleavage The results are averages of two determinations; approximate number of amino acid residues based on the molar percentage is given in parenthesis

Amino acid Amount in ~ ~ ~- .~

Y L3 C N I CN2

mo1/100 mol

Aspartic acid Threonine Serine Glutamic acid Proline Glycine Alanine Valine Methionine Isoleucine Leucine Tyrosine Phenylalanine Histidine Lysine Arginine Cysteine Tryptophan

~~~~~

8.61 (27) 7.78 (17) 7.14 (23) 8.22 (18) 5.30 (17) 5.60 (12)

13.70 (43) 12.20 (26) 3.66 (12) 3.57 (8) 7.07 (22) 9.03 (19)

10.85 (34) 8.80 (19) 4.56 (15) 5.46 (12) 0.26 (1) 0 4.37 (14) 4.47 (10) 8.11 (26) 8.94 (19) 5.18 (16) 4.04 (9) 4.63 (15) 4.38 (9) 2.2 (7) 2.50 (5) 6.94 (22) 6.59 (14) 7.40 (24) 8.42 (18) 0 0 0 0

~

10.01 (10) 5.09 ( 5 ) 5.99 (6)

15.67 (16) 2.67 (3) 4.81 (5)

15.08 (15) 3.61 (4) 0 4.39 ( 5 ) 7.34 (7) 3.47 (4) 3.98 (4) 1.52 (2)

10.27 (10) 6.11 (6) 0 0

M , 36000 (318) 24700 (215) 11300 (102)

Origin

RNA protein 5.8 - S RNA

5-S RNA

ciation of this type of RNA . protein complex, parti- cularly when assayed by membrane filter binding

In an attempt to establish the relationship of the single eukaryotic 5-S RNA binding protein to the multiple binding proteins in prokaryotic complexes, partial amino acid sequence analyses wereconducted on YL3. As observed in Fig.2, YL3 is a relatively acidic protein when compared to the three protein com- ponents in the E. coli 5-S RNA . protein complex [18]. This observation was confirmed by the amino acid composition (Table 1) which indicated an acidic/basic residue ratio of about 1.6 depending on the total amide content. Further, it was interesting to note that only one methionine residue was present, and no tryptophan or cysteic acid were detected by any of the methods used. About 10 mg protein were used for a sequence run to determine the first 30 residues as shown

(e% 126 1).

Fig.4. Forma/ion ofrihonucleoprotein complexes with the 5-S RNA- binding protein or fragment CN2 f rom S. cerevisiae. 32P-labeled H . cutiruhrum 5-S RNA (A), S. cerevisiae 5-S RNA (B), or S. cerr- visiae 5 . 8 3 RNA (C) was incubated for 30 min in 25 of 25 m M EDTA, pH 7.0, containing0.6 mg/ml of 5-S RNA . protein complex or 1 mg/ml of the CN2 fragment at 0 "C and applied to an 8 poly- acrylamide gel for electrophoretic analysis. The positions of 5-s RNA, 5 . 8 3 RNA, and the 5-S RNA . protein complex after me- thylene blue staining are indicated along with the origin and the bromophenol blue marker dye. The autoradiograph shows the

in Table 2, When this sequence (Fig, 5 ) was compared with the known primary amino acid sequences of the three E. coli 5-s RNA binding proteins [lo- 121 and the partially determined sequences for the two protein components in the equivalent complex from H . u t i -

ribosomes [6], a sequence homology was observed with EL18 and HL13, respectively. Of particular note Was the Cluster Of basic residues at about residue 22. distribution of '*P-Iabeled components

578 Yeast 5-S RNA Binding Protein

Because there was only a single methionine residue in YL3 (Table l), it was possible to split the protein into two fragments by cyanogen bromide cleavage (Fig. 6). Fortuitously, both fragments were fairly large : the molecular weights were estimated by sodium

Table 2. Sequencer unalysis of YL3 Threonine was identified as cc-aminobutyric acid. TCL = thin- layer chromatography

Residue Digestion wlth 6 M HCI TLC Assigned residue

+ SnClz - SnClz

1 Ser/Ala 2 Phe 3 Glx 4 LY s 5 Asx 6 Ser/Ala 7 LY s 8 Ser/Ala 9 Ser/Ala

10 Ser/Ala 11 TY r 12 Ser/Ala 13 Seri Ala 14 Arg 15 Phe 16 Glx 17 Thr 18 Pro 19 Phe 20 Arg 21 '4% 22 Arg 23 Arg 24 Glx 25 GlY 26 LY s

29 TY r 30 TY r

27 Thr 28 Asx

nmol

83.5 71.7 64.4 39.7 48.4 66.4 32.7 32.3 32.3 54.8 42.0 40.5 35.2 25.7 31.9 35.1 30.0 41 .O 21.7 21.1 22.0 20.5 23.8 21.5 28.8 18.9 12.9 13.7 15.1 18.5

Ala Phe

Lys

Ala Lys Ser Ser Ala TY r Ser Ser

Phe Arg

Ala Phe

Gln Gln LY s

Asp Asp Ala Lys Ser Ser Ala TYr Ser Ser Arg Phe

Gln Gln Thr Pro Phe Arg Arg Arg Arg

Glu Glu GlY Lys Thr

Asp Asp TY r TYr

YL3 C N 2

I EL5

dodecyl sulfate gel electrophoresis as 24700 and 11 300, respectively. When the amino acid sequences of their N termini were determined (Tables 3 and 4), the longer CN1 fragment was found to be the original N terminus of the YL3 protein while the overall sequence was extended by an additional 48 residues representing the N terminus of CN2. This new se- quence was very unusual in its amino acid com- position; 16 (or a third) of the amino acid residues were aspartic or glutamic acid. In total, all but one aspartate residue in CN2 appeared to be located in this sequence. When compared with the known primary amino acid sequences of the three E. coli 5-S RNA binding proteins [lo- 121 some sequence homology was observed with an internal region of EL5 partic- ularly around residue 35 (Fig. 5).

In an effort to identify specific RNA binding regions in YL3, an attempt was made to reconstitute the two cyanogen bromide fragments into ribonucleoprotein complexes with 32Pylabeled yeast 5-S RNA. Both fragments interacted with 5-S RNA but, as found with the whole protein, the CN1 fragment was re- latively insoluble and the ribonucleoprotein which formed was largely an aggregate which did not enter this gel. Interestingly, CN2 interacted efficiently with 5-S RNA and formed a homogeneous complex which migrated in the gel during electrophoresis (Fig. 4). This interaction ; however, appeared not to be specific; the fragment also formed a complex with yeast 5.8-S RNA and H . cutivuhrum 5-S RNA (Fig.4). Unlike yeast 5-S RNA neither of these RNA molecules were able to exchange into the native 5-S RNA . protein complex (Fig.4) so their interaction with CN2 is not physiological or at least is less specific. A greater amount of 32P-labeled yeast 5-S RNA was observed to exchange into unlabeled 5-S RNA . protein com- plex at room temperature but excessive RNA de- gradation was observed with the CN2 fragment and a lower incubation temperature (0 "C) was routinely used (e.g. Fig. 4).

HL13

YL3

EL18

r - - - i r - I I I

AaplPr I I

L- LeulLE

Fig. 5. A comparison of'the partiul unfino ucid seciuance of rhc 5-S RNA-binding proteir?,from S . ccrevisiae with other 5-S RNA-binding prorc~ins. Residues shown within solid boxes are identical; residues shown within dashed boxes may result from a single nucleotide change in their respective genes

R. N. Nazar, M. Yaguchi, G. E. Willick, C. F. Rollin, and C. Roy 579

DISCUSSION

The yeast 5-S RNA binding protein (YL3) is an acidic protein with a molecular weight of about 36 000, apparently the third largest protein component in the 60-S ribosomal subunit. It is relatively insoluble at neutral pH and undergoes a largely irreversible conformational change at acid pH when it becomes soluble. The protein contains about 300 amino acid residues including a methionine residue at which it may be cleaved to give two fragments containing approximately 200 and 100 amino acid residues, respectively. Sequence analyses show that the longer fragment (CN1) originates from the N terminus of the protein molecule while the shorter fragment (CN2) represents the C terminus. Only the shorter fragment could be reconsituted into a homogeneous 5-S RNA

0.3 1 CNI

"0 20 40 60 80 100 120 Tube number

. protein complex but the interaction appeared to be somewhat non-specific since homogeneous complexes could also be formed with yeast 5 .84 RNA and with H. cutirubrum 5-S RNA.

The results of sequence analyses appear to offer some insight into the relationship between the single eukaryotic 5-S RNA binding protein and the multiple protein components of the prokaryotic 5-S RNA . protein complex. As indicated earlier, both the si- milarity in the total molecular weight of the protein components and in the RNA binding sites suggest that the largkr eukaryotic protein may represent a fusion of the prokaryotic protein structures and pro- bably the genes. Two aspects of the sequence com- parisons (Fig. 5), the sequence homology and the distribution of acidic residues, appear to support this hypothesis. EL18 and HL13 were previously shown to be equivalent within their respective 5-S RNA . protein complexes by comparisons of their sequences [6] and their effects on the 5-S RNA circular dichroism spectra, ethidium bromide binding and resistance to limited ribonuclease digestion [29]. The comparison in Fig. 5

Table 3. Sequencer analysis Of'CNl Threonine was identified as cc-aminobutyric acid. TCL = thin- layer chromatography

Residue Digestion with 6 M HCI TLC Assigned residue ~~ -

+ SnCIZ - SnClz

B I II m nr

- YL3 - CNI

- CN2

Fig.6. Isolation of two fragments, CNI and CN2, from cyanogen- bromide-treated 5-S RNA-binding protein f rom S. cerevisiae. 15 mg YL3 were treated with cyanogen bromide; the fragments were fractionated by gel filtration on a column of Sephacryl S-200 eluted with 5 % acetic acid in 6 M urea (A). The fractions (2.2 ml each) were scanned at 230 nm and the protein content of each peak was determined using dodecyl sulphate gel electrophoresis (B). Original protein (I), CNBr-treated protein (11), CN1 fraction (111), CN2 frac- ;ion (IV)

1 Ser/Ala 2 Phe 3 Glx 4 Lys 5 Asx 6 Ser/Ala 7 LY s 8 Ser/Ala 9 Ser/Ala

10 Ser/Ala 11 TY r 12 Ser/Ala 13 Ser/Ala 14 Arg 15 Phe 16 Glx 17 Thr 18 Pro 19 Phe 20 Arg 21 Arg 22 Arg 23 Arg

25 GlY 26 LY s

29 TYr 30 TYr

24 Glx

27 Thr 28 Asx

nmol

125.1 Ala 114.0 Phe 100.5 Gln

64.0 Lys

100.2 Ala 46.3 Lys 37.7 Ser 31.5 Ser 50.5 Ala 36.0 Tyr 34.7 Ser 37.8 Ser 18.2 Arg 30.4 Phe 21 .o Gin 22.8 20.1 Pro 17.3 Phe 11.3 Arg 15.2 Arg 33.5 Arg 28.1 Arg 12.3 Glu 15.7 Gly 10.1 Lys 11.7 7.7 ASP 8.0 Tyr 9.9 Tyr

77.8 ASP

580 Yeast 5-S RNA Binding Protein

Table 4. Sequencer analysis of CN2 Threonine was identified as a-aminobytyric acid. TLC = thin- layer chromatography

Residue Digestion with 6 M HCI TLC Assigned ~ ~~~ residue

+ SnC12 - SnClz

1 2 3 4 5 6 7 8 9

10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 4x

Glx Glx Leu Ser/Ala Asx Asx Asx Glx Glx Arg Phe Ser/Ala Glx Leu Phe LY s G ~ Y TY r Leu Ser/Ala Asx Asx Ile Asx Ser/Ala Asx Ser/Ala Leu Glx

Ile TY r Thr Ser/Ala Ser/Ala His Glx Ser/Ala Ile '4% Ser/Ala Asx Pro Ser/Ala Phe

Pro Thr

LY s

nmol

117.3 101.3 115.9 Leu 116.8 Ala 70.4 67.9 64.8 66.7 68.0 31.3 Arg 65.8 Phe 27.3 Ser 55.7 57.1 Leu 57.2 Phe 31.9 Lys

44.1 Tyr 43.0 Leu 42.0 Ala 25.5 28.6 32.4 Ile 22.7 35.2 Ala 21.6 13.2 Ser 23.8 Leu 21.8 15.0 18.0 Ile 18.5 Tyr 10.6 10.5 Ser 19.1 Ala 7.9 His

13.5 15.7 Ala 11.9 Ile 11.5 Arg 13.9 Ala

49.4 Gly

6.8 8.8 Pro

12.0 Ala 12.6 Phe 7.4 Lys 4.8 Pro 2.9

Glu Glu Leu Ala ASP ASP ASP Glu Glu Arg Phe Ser Glu Leu Phe LYS GIY TY r

ASP ASP Ile ASP Ala ASP Ser Leu Glu ASP Ile TY r Thr Ser Ala His Glu ;\ I il I I C

Arg Ala ASP Pro Ala Phe

Pro Thr

Leu Ala

Lys

indicates a significant sequence homology between HL13 and YL3 and a somewhat lower homology between EL18 and YL3. Together with the previous studies it appears reasonable to conclude that the N terminus of YL3 is equivalent to that of HL13 and EL18 (Fig. 7).

Unfortunately, because very little sequence in- formation is available for HL19, the partial sequence of the CN2 fragment can only be compared with the less homologous E. coli proteins. Nevertheless, some sequence homology is observed with an internal portion of EL5 (Fig. 5 ) ; although limited this degree of homology is greatest among the three proteins. Furthermore, among the three proteins, this region (residues 133-178 of EL5) was also unique in its content of acidic residues. Ten (or 25%) of the re- sidues were glutamic or aspartic acid, an unusually high concentration similar to the 33'x observed in the CN2 fragment. Therefore, although they could simply be coincidental, these two observations suggest that the CN2 region of the yeast 5-S RNA binding protein may be structurally equivalent to the EL5 protein. Because EL5 has previously been shown to be struc- turally equivalent to the halophilic HL19 protein [6,29], the CN2 region will also be expected to share sequence homology with HL19. Recent studies on the organization of ribosomal protein genes in E. coli indicate that, although they are not adjacent, both EL5 and EL19 are in the same transcriptional unit [30]. The proximity between these structural genes would simplify the fusion suggested in Fig. 7. Although no sequence homology has yet been observed between EL25 and HL13 or CN1, based on total molecular weight considerations, and since EL18 + E-25 ap- peared to be equivalent to HLI 3 in studies using both chemical and physical techniques to compare 5-S RNA . protein complexes from E. coli and H . cuti ruhvum [6,13,29], a sequence equivalent to EL25 may lie in the spacer sequence between EL1 8 and EL5 shown in Fig. 7.

Two recent studies on the interaction of ribosomal RNAs with ribosomal proteins [31,32] raise the pos- sibility that, although the released 5-S RNA . protein complex from eukaryotes contains only one protein, other ribosomal proteins may interact with 5-S RNA within the whole ribosome. Based on these studies an alternate interpretation of the data exists. While the amino terminus of YL3 is equivalent to EL18 and HL13, EL5 or HL19 may actually correspond to one of the other interacting proteins. Further sequence analyses are clearly required to fully answer this question ; at present only the relationship between YL3 and EL18 or HL13 (Fig. 7) appears certain.

A study by Newberry et al. [38] noted that the N-terminal region of EL18 is very basic and found it to be readily accessible to trypsin digestion in a 5-S RNA . protein complex. The N-terminal sequence of YL3 also contains an unusually high number of basic residues and the sequence homology between Y 13, HL13 and EL13 indicated in Fig.7 is particularly high in the cluster of basic residues at about residue 21. This feature appears to be of sufficient importance to be highly conserved during the evolution of yeast from

R. N. Nazar, M. Yaguchi, G. E. Willick, C. F. Rollin, and C. Roy 58 1

CN I (24,700) CN2 ( I I, 300) S. cerevisioe t -I --

-- Fig. 7. Suggested model f o r the evolutionary rehionship hern,eeti eukaryotic and prokaryotic 5-S RNA-binding proteins. Molecular weights (shown in brackets) for the individual binding proteins from E. coli [ lo , 111 and H . cutirubrum [6] are published values. A sequence equivalent to protein EL25 from E. coli [12] may lie between EL18 and EL5 (see text)

bacteria; possibly it is involved in RNA binding. Because, in E. coli, this region was susceptible to trypsin digestion it is thought not to be involved in 5-S RNA binding and has been postulated to interact with 23-S ribosomal RNA in a complex between 5-S RNA . protein and 23-S RNA [33]. Similar clusters of basic amino acids in protamines have recently been shown to bind along double-helical structures in tRNA [34], and to this extent, a role in 5-S or 26-S ribosomal RNA binding is further supported.

The relatively high degree of sequence homology between the N termini of YL3 and HL13 also raises a recurring question about the evolutionary relation- ship of the extreme halophiles to eukaryotes. Unlike other prokaryotes, extreme halophiles lack a Met- tRNA transformylase [30] and the halophilic cell membrane contains a rhodospin-like protein [33] and a glycoprotein containing the N-glycosidic and O-gly- cosidic linkages common to mammalian glycoproteins [34]. Recently, the N termini of ribosomal ‘A’ proteins have also been shown to have a high sequence homology; seven of the first eleven residues in the ‘A’ protein from S. cerevisiue or Artermia salinu are homologous with H. cutiruhvum (see [35]). Further- more, studies on tRNA charging by heterologous synthetase preparations indicate that tRNA from H . cutirubrum is structurally more closely related to that of yeast than any other prokaryote (U. Kwok and J . T. Wong, unpublished data). It appears that in the search for a prokaryotic ancestor to eukaryotic cells, certainly with respect to translation, the extreme halophile is one organism which deserves considerably more attention.

With respect to the 5-S RNA binding sites in YL3, the ability of the CN2 fragment to form homogenous RNA . protein complexes with several different RNA molecules (Fig. 5 ) poses an interesting contrast to the native protein. If this RNA binding site is available in the native protein, then some structure in the CNI fragment must confer specificity upon the binding site in CN2, restricting the native protein to a specific interaction with yeast 5-S RNA. Recent studies using proton magnetic resonance [36] indicate that a se- lective interaction of carboxylic acid side chains of

proteins with guanine bases may be one way by which proteins recognize base sequences. Accordingly, i t is attractive to postulate that the cluster of acidic residues in the N terminus of CN2 may play such a role. Studies are now being undertaken to identify the region in CN2 which interacts with RNA to examine the effect of CN1 on this interaction.

REFERENCES

1. 2.

3.

4. 5.

6.

7.

8.

9.

10. 11.

12.

13.

14.

15.

16

17.

18

19 20 21

Blobel, G. (1971) Proc. Nu/[ Acud. Sci. U.S.A. 68, 1881 - 1885. Monier, R. (1974) in Ribosomes (Nomura, M., Tissieres. A. Sr

Langyel, P., eds) pp. 141 - 168, Cold Spring Harbor La- boratory, New York.

Erdmann, V. A. (1976) Progr. Nucleic Acid Res. Mol. Biol. 18,

Peterman, M. L. (1974) Methods Enzymol. 30, 346 - 349. Terao, K., Takahashi, Y. Sr Ogata, K. (1975) Biocliim. Biopli>x.

Smith, N., Matheson, A. T., Yaguchi, M., Willick, G. E. Sr

Chen-Schmeisser, U. Sr Garrett, R . A. (1977) FEBS Lett. 74.

Erdmann, V. A, , Sprinzl, M. Sr Pongs. 0. (1973) Biochem.

Azad, A. A. Sr Lane, B. G. (1973) ( h n . .J. Biochem. 51. 1669-

Chen, R. Sr Ehrke, G. (1976) FEBS Lett. 69, 240-245. Brosius, J. , Schiltz, E. Sr Chen. R. (1975) FEBS Lrtt. 56.

Dovgas, N. V., Markova, L. F.. Mednikova, T. A, , Vinoku- rov, L. M., Alakhov, Yu. B. Sr Ovchinnikov. Uu., A . (1975)

Nazar, R. N., Willick, G. E. Sr Matheson, A. T. (1979) J . Biol. Chem. 254, 1506-1512.

Yaguchi, M.. Nazar, R. N., Willick, G. E. Rollin, F. Sr Roy, C. (1978) Internutional Cor!ference on Yeast Genetics ur~d Molec- ulur Biology, vol. 9, p. 106.

Torano, A, , Sandoval, A. Sr Heredia, C. F. (1974) A4ethotl.s Enzymol. 30, 254 - 261.

Nazar, R. Sr Matheson, A. T. (1977) J . Biol. C%em. 252, 4256-4261.

Hardy. S. J. S., Kurkland, C. G., Voynow, P. 8 Mora, G. ( I 969) Biochemistry, 8, 2897 - 2905.

Kaltschmidt, E. Sr Wittmann, H. G. (1970) Anal. Biochcm. 36, 401 -412.

Webcr, K. Sr Osborn, M. (1969) J . Biol. Chem. 244,4406-4412. Edelhoch, H. (1967) Biochemistry. 6 , 1948- 1954. Liu, T. Y. Sr Chang, Y. H. (1971) J . Biol. Chrm. 248. 2842-

45 - 90.

Aera, 402, 230-237.

Nazar, R. N. (1978) Eur. J . Biochem. 89. 501 -509.

287-291.

Biophys. Res. Commun. 54, 942 - 947.

1672.

359 - 361.

FEBS Lett. 53, 351 -354.

2848.

582 R. N. Nazar, M. Yaguchi, G. E. Willick, C. F. Rollin, and C. Roy: Yeast 5-S RNA Binding Protein

22. Hirs, C. H. W. (1967) Methods Enzymol. 11, 197-199. 23. Mendez, E. s( Ldi, S. Y. (1975) Anal. Biochem. 68. 47-53. 24. Wittmann-Liebold, B., Geissler, A. W. s( Marzinzig, E. (1975)

25. Chen, R. (1976) Hoppe-Seyler's Z. Physiol. Chem. 357, 873-

26. Gross, E. (1974) Method.7 Enzymol. 11, 238-255. 27. Kruiswijk, T., P h t d , R . J. s( Magcr, W. H. (1978) Eur. J .

28. Zimmermann, J. Sr Erdmann, V. A. (1978) Nucleic Acid Rrs. 5,

29. Nazar, R. N., Willick, G. E. s( Matheson, A. T. (1978) Fed.

30. Jaskunas, S. R., Fallon, A. M. s( Nomurd. M. (1977) J . Biol.

J . Supramol. Struct. 3, 42-447.

886.

Biochem. 83,245 -252.

2267 - 2287.

Pro(,. 37, 2139.

Chem. 252,7323 - 7336.

31. Ulbrich, N. s( Wool, I . G. (1978) J . Bid. Chem. 253. 9049-

32. Metspalu, A., Saarma, M., Villems, R., Ustav, M. Sr Lind,

33. Newberry, V., Brosius, J. Sr Garrett, R. (1978) Nucleic AcidRes.

34. Warrant, R. W. Sr Kim, S.-H. (1978) Nature (Lond.) 271,

35. Matheson, A. T., Yaguchi, M., Nazar, R. N., Visentin, L. P. Sr Willick, G. E. (1978) in Energetics cmd Structure of Halo- philic Microorganisms (Caplan, S..R. s( Ginsburg, M., eds) pp. 481 -501, Elsevier North-Holland Biomed. Press, Am- sterdam.

36. Lancelot, G. s( Helene, C. (1977) Proc. Natl Acad. Sci. U.S.A. 74,4872-4875.

9052.

A . (1978) Eur. J . Biochem. Y I , 73-81.

5, 1753-1766.

130 - 135.

R. M. Nazar, Department of Biochemistry and Genetics, University of Guelph, Guelph, Ontario, Canada N1G 2WL

M. Yaguchi, G. E. Willick, C. F. Rollin, and C. Roy. Division of Biological Sciences, National Research Council of Canada, Ottawa, Ontario, Canada K I A OR6