the primary structure the acidic phosphoprotein p2 from ...9190 primary structure of ribosomal...

THE JOURNAL OF Blo~ocrcu CHEMISTRY Vol. 257, No. 15, Issue of August 10, pp. 9189-9197,1982 Printed in U.S.A.

The Primary Structure of the Acidic Phosphoprotein P2 from Rat Liver 60 S Ribosomal Subunits COMPARISON WITH RIBOSOMAL ‘A’ PROTEINS FROM OTHER SPECIES*

(Received for publication, January 4, 1982)

Alan LinS, Brigitte Wittmann-Liebold& James McNallyS, and Ira G. Wool$! From the $Department of Biochemistry, The University of Chicago, Chicago, Illinois 60637 and the §Max-Planck-Institut fur Molekulare Genetik, Abteilung Wittmann, Berlin-Dahlem, West Germany

The primary structure of rat liver ribosomal protein P2 was deduced from the sequence of the peptides. Ten peptides were obtained by cleavage of P2 with trypsin. The peptides, which accounted for the 111 residues of P2, were isolated by high voltage electrophoresis and chromatography on cellulose thin layer sheets, and the partial or complete sequence was determined by micro- manual or solid phase procedures using .I-N,N-dimethylaminoazobenzene 4’-isothiocyanate and phenyliso- thiocyanate. In a similar manner, the sequence of 14 peptic peptides was determined. The sequence of the NHz-terminal 30 residues of P2 was obtained by automatic Edman degradation in a sequenator. The order- ing of the tryptic peptides was aided by determination of the partial or complete sequence of fragments generated with chymotrypsin, or Anillaria mellea protease, or by secondary cleavage of peptic peptides with trypsin. The carboxyl-terminal sequence was obtained from a cyanogen bromide fragment and from hydrolysis with carboxypeptidase. The sequence of protein P3 was also determined. P3 differs from P2 only in that it lacks the carboxyl-terminal 8 residues, and hence, it is likely to be a proteolytic product of P2. Rat liver ribosomal protein P2 is homologous with yeast YP Al, with Artemia salina eL12, and with Halobacterium cutirubrum L20. It is likely that rat liver P2 is also homologous with the prokaryotic ribosomal “A” proteins, Escherichia coli L7/L12, Micrococcus lysodeikticus MAl, and Bacillus subtilis L9, but that during evolution, a transposition of a portion of the molecule occurred.

Information on the chemistry of the constituents is necessary for a determination of the structure of ribosomes; and knowledge of the structure of the particle is, in turn, required if molecular organization is to be related to function in protein synthesis. Eukaryotic (rat liver) ribosomes have some 70 to 80 different proteins(1). The proteins have been isolated and characterized (l), and the NHz-terminal sequences of some have been determined (2). But eukaryotic ribosomes have of the order of 16,000 amino acids, and to determine the sequence of each of the proteins would be a monumental task, a task whose achievement is not imminent. Clearly, for the present, the determination of the covalent structure of eukaryotic ribosomal proteins will have to be limited to those which have

* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

7 Recipient of National Institutes of Health Grants GM-21769 and CA-19265.

been established to be especially important for function, or to those which might shed light on ribosome evolution. The acidic phosphoprotein P2 from rat liver 60 S ribosomal subunits qualified on both counts, and its primary structure has been determined.

The large subunit of all ribosomes that have been examined contains acidic proteins similar to Escherichia coli L7/L12 (3, 4). (Protein L7 differs from L12 only in that its NH2- terminal serine is acetylated.) The proteins, collectively des- ignated “A” proteins, have unique chemical and physical properties. First, they are acidic, which is unusual for ribosomal proteins. Second, they have an exceptional amino acid composition: more than 20 mol % alanine, few aromatic amino acids, only one or two arginines, and generally no cysteine or tryptophan. There is a large hydrophobic region near the center of the molecule, and the A proteins may have more than 70% helical content (5, 6). In bacteria, the proteins are present in four copies (7); whether that is the case in eukaryotic ribosomes is not known. Finally, the A proteins provide at least part of the binding site for the initiation, elongation, and termination factors involved in protein synthesis (8).

EXPERIMENTAL PROCEDURES’

1 Portions of this paper (including “Experimental Procedures,” part of “Results,” and Figs. 1-4 and 6) are presented in miniprint at the end of this paper. The abbreviations used are: DABITC, 4-N,N- dimethylaminoazobenzene 4”isothiocyanate; DABTH, 4-N,N-dimethylaminoazobenzene 4’-isothiohydantoin; PITC, phenylisothio- cyanate; PTH, phenylthiohydantoin; TP60, the total proteins of the 60 S ribosomal subunit. Miniprint is easily read with the aid of a standard magnifying glass. Full size photocopies are available from the Journal of Biological Chemistry, 9650 Rockville Pike, Bethesda, MD 20814. Request Document No. 82M-0004, cite authors, and include a check or money order for $4.80 per set of photocopies. Full size photocopies are also included in the microfrlm edition of the Journal that is available from Waverly Press.

9189

9190 Primary Structure of Ribosomal Protein P2

Amtno Acld Analyses - Samp le of pmte ln I5 nmal) or peptlde (1-2.5 "mol) were hgdrolyzed In 50 y l of 5.7 M HCI, containrng 0.02% 0-maeaptoetMnd and 0.01 % phenol, at 110 C for 20 h Ifw omtam Rlso for 48 and 72 hl. The hvdmlvsls t u D e r were flushed with nitroeen and evacuated

Isolation and Characterization of Rat Liver Acidic Ribo- somal Proteins PI, P2, and P3-The proteins were extracted from rat liver 60 S ribosomal subunits with either 67% (9) or 10% acetic acid in 3 mM Tris-HC1 containing 33 mM magne- sium acetate. The 10% acetic acid extract which contained the acidic phosphoproteins P1, P2, and P3 was resolved by chromatography on carboxymethylcellulose with a linear gradient of 0 to 0.1 M LiCl at pH 4.2 (Fig. 1). Protein P3, which had not been identified before, was characterized by two-dimensional polyacrylamide gel electrophoresis in urea (Fig. 2 ) and by one- dimensional electrophoresis in polyacrylamide containing sodium dodecyl sulfate (Fig. 3 ) . The latter gel was scanned at 540 nm, and no contamination was detectable in the P3 preparation.

Primary Structure of Ribosomal Protein P2 9191

0.3 1 I

PllP2 0 2 -

~ LJ- j P - 4 P3

0 A

0 1 - "-" ""

"" -0 05

P2 "" "_"" "

" _ - " ""

_ - - 0

0 50 100 150 200

FRACTION Fig. 1. Chromatography, on carboxymethylcellulose, of acidic PrOteiN from 60 S ribcsomal

sUbunlts - Thc pmteim were extracted from ret liver 60 S riboSOmal subunits with 10% acetic R F i d d 19 mg were epDlied to B column (1.6 x 2.5 ern) of cBlboxymethylcellUloe and eluted Wlth D 1.2 Iller llncar gradbent of 0 lo 0.1 M LiCl at pH 4.2. Fractions ( 6 mll were colleeled. and the 8bSDrplion 111 280 nm was determined. The amteim in the frsetims were idenlifred bv polyscrylamlde gel eleclmphorcris.

0'

0 P:*.

P3 *

P3 (I ugl was anslped alme M the right side of the second-dimension gel lorigin st the top on Fig. 2. Two-dimensional palyaarylsrnide gel electropho?esis of protein P3. The isolated protein

the extreme right) for asesment of 1u purity. and on the left sde (origin st the top in the center) wtlh B rnlxture or m d l e 60 S ribmomd subunit protans ( 5 up) f a asislance in lhe ~denuftcat!on. Electrophorers was from right to left In the furst dimenam and from top to bottom m t h e Second. The aigin is marked 2.

The fractions containing proteins P1 and P2 (Fig. 1) were resolved further either by chromatography on DEAE-cellulose (9) or by preparative cellulose acetate gel electrophoresis. The purified proteins P1 and P2 had no appreciable contamination (Fig. 3).

The amino acid composition of the three proteins (PI, P2, and P3) was determined (Table I). There are definite differences in the composition of the three proteins; for example, only P1 has cysteine, threonine, and histidine. However, each is characterized by a large number of alanine residues (about 22), few arginine residues (one or two), and no tryptophan.'

Ribosomal proteins P1, P2, and P3 were digested with trypsin, and the peptides were separated by high voltage electrophoresis and chromatography on thin layer sheets (Fig. 4). Proteins P2 and P3 had very similar tryptic peptide maps; the only difference was that P3 lacked peptides T10 and TlOa, and instead had TlOb. It was established later that TlOa differed from T10 only in that the methionine in the peptide was oxidized. We determined by hydrolysis with carboxypeptidase A that tryptic peptide T10 from P2 has a carboxyl- terminal aspartic acid (rather than lysine or arginine). Thus, it was likely to be derived from the carboxyl terminus of the protein. Peptide TlOb obtained from P3 contained lysine and aspartic acid (Table 11), but the lysine was at the NHP terminus. Therefore, TlOb could have been derived from the carboxyl terminus of P3. The amino acid composition of peptides T10 and TlOb was determined (Table 11). We found that eight

'Whether PI has tryptophan was not determined (Table I).

"c

"

P3. The anslysls was of IO ug of ecidlc proteim fmm the 60 S ribmomd SubUnll (A60). 1.5 ug Fig. 3. Polyacrylamide Eel electrophoresis. in sodium &decy1 sulfate, of proteim P1. P2, and

3 the purlfled p t e i m (PI, P2. and PJ). and 20 ug of 60 S riboSOmal wbunll prolelnc (TP6O).

TABLE I Amino acid composition ofthe acidic rihosomalproteins PI, P2,

and P3 The amino acid composition of the proteins was derived either (A)

from analvsis of a hvdrolvsate of the protein or (B) from the sequence P1 P2 P3

A A B A B Amino acid -

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

2 0 0 7 5

3.3 0.4 0 7.7 8.7 9

1 1 1

8.8 5.4 5 15.0 11.8 12 23.0 21.5 22 8.7 8.9 9 1.4 1 .7 2 5.6 5.5 5 8.8 8.3 8 0.9 1.7 2 2.2 2.2 2 1.2 0.4 0 6.2 8.5 9 1 .0 1.9 2

0 0 (125) 1 1 1

12.7 11.7

15.9 12.4

0 0 6 5

0 0 8.4 9

10 1

5.4 5 10.2 10 22.3 22 8.6 9 1 .o 1 4.8 5 7.2 7 1.8 2 0 0 0 0 9.4 9 1.7 2 0 0

103

11.3

11.8

eleclrophoresis and thin-layer chromalopaphy on cellulose sheets. The p ~ o t e m were dlgested wtth Fig. 4. Separation of the tryptic peptides of proteins PI. P2, and P3 by hiKh-VollaKe

amino acids that constitute the difference between the two peptides, i.e. one aspartic acid, one glutamic acid; two glycine, one methionine, one leucine, and two phenylalanine, account for the difference in the amino acid Composition of P2 and P3

9192 Primary Structure of Ribosomal Protein P2

TABLE I1 Amino acid composition of tryptic peptide TI0 from protein P2 and

of TlOb from protein P3 T10-P2 TIOb-P3

Aspartic acid 2.1 (2) 1.2 (1) Glutamic acid 0.8 (1) Glycine 2.3 (2) Methionine 0.9 (1) Leucine 0.8 (1) Phenylalanine 1.7 (2) Lysine 1.1 (1) 0.9 (1) Residues 10 2

(Table I). Thus, P3 appeared to differ from P2 in that it lacked eight residues at the carboxyl terminus. The supposition was later confirmed. The NH2-terminal sequences of P2 and P3 were identical, establishing that the difference in the sequence of the two proteins must lie at the carboxyl terminus.

The tryptic peptide map of P1 is in some respects similar to that of P2 and P3. There are differences, however, especially in that there are many more minor spots in the map of P1 (Fig. 4). It was not possible to correlate peptides from P1 with those from either P2 or P3 from a determination of the amino acid compositions, nor was it possible to determine the NH2- terminal sequence of P1, which may be blocked. It is very likely that P1 is related to P2, since a monoclonal antibody to the latter reacts with the former (34). However, the nature of the relationship remains to be determined.

DISCUSSION

The Primary Structure of Ribosomal Protein P2-The primary structure of rat liver ribosomal protein P2 was deduced from the sequence of the peptides (Fig. 5). All of the tryptic peptides and all of the peptic peptides were purified, and the complete or partial sequence was determined at least twice. All of the tryptic peptides could be aligned from the sequence of the peptic peptides. The sequence of the fist 30 NH2-terminal residues of protein P2 was obtained by automatic Edman degradation in a sequenator. That sequence confirmed the assignment of amino acids in peptides T1, T2, T3, and T4 and their alignment. The order of other peptides was substantiated from the sequence of fragments generated by cleavage with chymotrypsin (Fig. 6). The sequence of the tryptic and peptic peptides was sufficient to provide all of the primary s h c t u r e of P2 except in a small region encompassing residues 105-109. The carboxyl-terminal sequence and the sequence of residues 105-109 were obtained from a cyanogen bromide peptide (CN2), and the bridge between T7 and T8 was established from a partial sequence of an Armillaria mellea protease peptide, AmA. Finally, the carboxyl-terminal amino acids (- - -Phe-Asp) were identified after carboxypeptidase treatment.

The amino acid composition of P2 determined by analysis after hydrolysis and the number of residues of each amino acid calculated from the sequence are in excellent agreement (Table I). P2 has 111 amino acid residues; the molecular weight calculated from the sequence is 13,252, reasonably close to that of 15,200 obtained by electrophoresis in gels containing sodium dodecyl sulfate (9).

Protein P2 contains no threonine, histidine, tryptophan, or cysteine, and only two arginine residues. On the other hand, the protein has 22 alanine residues. The alanine-rich region, which is a characteristic of all A proteins, extends in P2 from residue 64 through residue 92 (Fig. 5). Of these 29 residues, 15 are alanine, and of the remaining 14, four are glycine, four are proline, three are valine, and three are serine. Thus, that region is very hydrophobic. The glycine and proline residues

give the region flexibility. The predictions of secondary structure suggest that the alanine-rich region is a helix, and this may explain why the alanine residues in the sequence are not cleaved by chymotrypsin. Cleavage was obtained in the alanine-rich region only by pepsin in the sequence - - -Val-Ala- Val---(residues 72-74), and then only after 18 h of incuba- tion. The carboxyl-terminal region of protein P2, on the other hand, is rich in acidic residues. Of 19 carboxyl-terminal amino acids (residues 93 through lll), nine are either aspartic or glutamic acid. Of the four aromatic amino acids in P2, the two tyrosine residues are near the NHz terminus (residues 3 and 7), and the two phenylalanine residues are near the carboxyl terminus (residues 107 and 110). The functional significance of the distribution of amino acids is not known.

Ribosomal protein P2 can be phosphorylated to form at least three derivatives, P2a, P2b, and P ~ c , which contain, respectively, one, two, or three (or some multiple thereof) phosphate residues (9). Neither the identity of the phosphorylated amino acids nor their position in the sequence is known. It was not possible to make those determinations during the course of analysis of the primary structure, since it is likely that P2 itself has no phosphoamino acid (9).

From the sequence of protein P2, one can make predictions concerning features of the secondary structure (37, 38). We used four separate algorithms (39-44) to calculate the residues likely to be in helices, in extended regions, in turns, or in random coils (Fig. 7). The most prominent feature is the great amount of secondary structure that the programs suggest for rat liver P2, as is the case for E.coZi L7/L12 (5). However, the long stretch of helices in the NHz-terminal region of E . coli L12 (residues 8-61) is interrupted in P2 by several turns which are almost equidistant from each other. There are, moreover, a number of regions (residues 23-27, 54-58, 72-76, 105-111) where the results are ambiguous in designating a helical or extended conformation. However, there is a consensus in the predictions that the acidic carboxyl-terminal regions in both E. coli L12 (residues 103-120, Ref. 5) and in rat liver P2 (residues 88-104) are helical. Although E . coli L7/L12 (5) and rat liver P2 have some common features, one cannot say that there is a striking similarity in their predicted secondary structure.

Determination of the Sequence of Ribosomal Protein P3- Comparison of the tryptic peptide maps of proteins P2 and P3 (Fig. 4) indicated that the two proteins are very similar; indeed, the only difference was in peptide T10 of P2 and peptide TlOb of P3. This impression was reinforced when we determined that the NH2-terminal 20 residues of P3 are the same as in P2. Since T10 was known to be the carboxyl- terminal peptide of P2 (see above), it seemed likely that P3 differed from P2 only in that portion of the molecule. To determine whether this was the case, we isolated and se- quenced the tryptic peptides from P3. Peptides T1 through T9 (residues 1-101) have the same sequence as the corresponding peptides in P2. The sequence of the tryptic peptide TlOb from protein P3 is - - -Lys-Asp. The difference in the amino acid composition of TlOb from P3 and of T10 from P2 accounts for the difference (eight residues) in the amino acid composition of the two proteins. Thus, P3 differs from P2 in that it lacks the carboxyl-terminal eight amino acid residues. P3 has a molecular weight of 12,100. It is assumed that P3 is a fragment generated from P2 during isolation and purifica- tion. There is no evidence that P3 is a component of ribosomes. It is worth noting that P3 is soluble in water, whereas P2 is not. Indeed, P2 is poorly soluble even in 6 M urea or 2% acetic acid. The hydrophobic carboxyl-terminal octapeptide thus appears to affect the solubility of the entire molecule. More- over, since this sequence is highly conserved (see later), it may


I 5 IO 15 20 25 Met-Arg-Tyr-Val-Ala-Ser-Tyr-Leu-Leu-Alo-Ala-Leu-Gly-Gly-Asn-Ser-Asn-Pro-Ser-Ala-Lys-Asp-Ile-Ala-Lys-

T TI T2 T 3

J 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 " 7 7 7 " 7 - + 4 + + + + + - 3 + + + + + + + + +

Ps PSI Pa2

P X " " ~ " ~ T ~ 7 7 7 7 7 7 7 7 7 7 7 7 7 7 -

- 7 7 7 7 7 7 7 7 7 7 7 7 7 7 PrZ-TZ 7777777777777""

Pa2-T3

Ch ChI Ch2 Ch3 " 7 7 7 7 7 7 7 7 x307

7 7 7 7 7

7 "7 7 7 -7 7 7 7 7 7 7 7 7 7 7 "2 7 7 "7 7 "2 -7 "7 --, --, "~,""""""""-\""~"-~""\~", x ) 35 4 0 45 50

I~e-Leu-Asp-Ser-Val-Gly-Ile-Glu-Alo-Asp-Asp-Glu-Arg-Lys-Leu-Asn-Lys-Val-Ile-Ser-Glu-Leu-Asn-Gly-Lys- T T 4 T 5

" " 7 7 7 7 7 7 7 7 7 7 ~ " 7 ~ 7 " 7 " 7 "

T 6

Ps PS3 Pa4 p55 77

p56

" 7 7 7 - 7 7 7 - 7 7 7 7 " 7 7 "

Ps2 - T4 7 7 7 7 - 7

Ch - Ch4 Ch5 7 7 7 7 7 777

6 0 65 70 75

T T7 Asn-Ile-Glu-Asp-Val-Ile-Ala-G~n-Gly-Val-Gly-Lys~eu-Ala-Ser-Val~ra-Ala-Gly-Gly-Ala-Val-Ala-Val-Ser-

7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 "

T 8

"~"++"+++++-+"3 Ps PS7 Pa8 777777777777777777777777-

P.9

Ch

Am 77

C h6 7 7 7 7 7 7 7 7 7

7 7 7 7 7 7 7

pun -A

80 85 90 95 100 Ala- Ala-Pro-Gly-Ala-Ala-Ala-Pro-Ala-Ala-Gly-Ser-Ala-~a-Ala-Ala-Ala-Glu-Glu-Lys-Glu-Glu-SerClu-Glu-

Ps 7 7 7 7 7 7 - 7 7 - 7 7 7 - 7 7 7 7 - 7 7 7 7 7 7 7 7

pa10 Pfi

Ch Am

I05 I IO

T Lys-Lys-Asp-Glu-Met-Gly-Phe-Gly-Leu-Phe-Asp.

TI0 - . .- 77" " --, --, --7 ", "~ - -, - -, ", ", - .7

Ps Pa12 Pal3 PSI4 7 7 7 7 7 7 7 7 7 7 7

- Micro-monuol DABITC/PITC

4 Solrd-phose DABITC/PITC Automatic Edman Degradotlon-Sequenotor -". - ", Resldues not determmed by automatlc

CN Edman

CN2 - Corboxypeptldase 7 7 7 7 7 7

----> Donsyl-Edman Degradotlon FIG. 5. The covalent structure of rat liver ribosomal protein P2. T, tryptic peptides; Ps, peptic peptides; Ch, chymotryptic peptides;

Am, A. mellea protease peptide; CN, cyanogen bromide peptides.

well play an important role in the structure or the function, or liver P2 was compared to that of the other A proteins whose both, of P2. sequence has been determined: A. salina eL12 ( L I ~ ) , ~ yeast

Ribosomal protein p.2 with the Sequences of the Ribosomal serine rather than threonine and that the unknown residues at A Proteins from Other Species-The primary structure of rat positions 20 and 34 are threonine and cysteine, respectively.

Comparison O f the Amino Acid Sequence O f Rat Liver We have been informed by W. MGuer that residues 13 and 19 me


I A t

C.

.T

0 '

0 a c h 6

.or,g,n

Ps2-14

D

O I q m

J Iig. 6. Sepnmtim of DeDtides. A. Protein P2 WBS digested with pepln, and the peptides

weFe separated by high-voltage electr$horhoresis 8nd thin-layer chromatography on cellulase sheets. E. Peptic peptide Ps2 was digested wlth trypsin, and the secondary peptldes were sepamted by thin-layer chromatography on cellulase sheets. C. Protein PZ was digested wlth chymotrypsin, and the peptldes Were sepnrated as in 4. E. Trotem PZ was digested Wlth Arrnillana rneUea protease, end the peptides were separefed 8s in 4.

(Saccharomyces cereuisiae) YP A1 (46), Halobacterium cutirubrum L20 (47), Bacillus subtilis L9 (48), Micrococcus lysodeikticus MA1 (49), and E . coli L7/L12 (3). It is obvious merely by inspection that the eukaryotic A proteins (rat, A. salina, and yeast) are homologous; that is, they are derived from a common ancestral gene. Rat liver P2 and A. salina eL12 have 44 identical residues (40% homology), i.e. the same amino acid occurs at the same place in the sequence. The number of identical residues in rat liver P2 and yeast YP A1 is 41 (37%), and for A. salina eL12 and yeast YP Al, that number is 33 (30%). If two or three gaps are introduced into the sequences (as in Fig. 8),the number of identities and the homology are even greater. The number of identical residues for P2 and eL12 is 68 (61% homology), that for P2 and YP A1 is 52 (47%), and that for eL12 and YP A1 is 49 (44%). It may be only a coincidence, but it is striking that each of the gaps (at positions 39,63,95, and 96) is opposite a lysine residue, as if that amino acid had been deleted (Fig. 8).

There is no obvious similarity in the sequence of eukaryotic and prokaryotic A proteins, at least not upon inspection. It seemed important to make a decision concerning homology of eukaryotic and prokaryotic A proteins on a less arbitrary basis, For that reason, we compared the primary structure of the several proteins by using the computer program RELATE (32) (Table 111). This comparison confirmed that the eukaryotic ribosomal A proteins are homologous and that the A protein from the archebacterium H . cutirubrum is homologous with the eukaryotic proteins, but not with those of prokaryotes. That conclusion had been reached before (4,50) without application of formal criteria. The finding of greatest interest concerns the comparison of eukaryot,ic and prokaryotic A proteins. At least some of the eukaryotic A proteins are related to some of the prokaryotic A proteins.

To authenticate the results obtained with RELATE, we used the computer program ALIGN (32) to compare the proteins (Table IV). The results obtained with ALIGN confirmed (i) that the eukaryotic proteins are homologous, (ii)

that the prokaryotic proteins are homologous, and (Si) that H . cutirubrum L20 is homologous with the eukaryotic A proteins and not with the prokaryotes. However, the results differ in an important respect: they provide no evidence of homology of eukaryotic and prokaryotic A proteins. The results obtained with ALIGN accord with previous conclu- sions (4,50), but are in contrast with results which we obtained with RELATE.

A possible explanation for these conflicting results was that the methods of comparison employed by the two programs differed significantly. RELATE compares every fragment of a given length in one protein (20 amino acids in our case) with every fragment of the same length in the other protein. Fragment position within the sequence is irrelevant. ALIGN, on the other hand, makes a residue by residue comparison from the NH2 to carboxyl terminus. The only degree of freedom permitted is the insertion of gaps. As a consequence, if the eukaryotic sequences had been rearranged relative to the prokaryotic ones, then RELATE might still detect homologies while ALIGN would not.

One way of determining a putative sequence rearrangement is to examine the raw data from RELATE. The program lists the positions in the two sequences for which high scores have been obtained in the fragment comparisons and determines the distance in amino acids between any pair of fragments which yield a high score. Repeated appearance of a specific distance among the high scores might suggest that a conserved region in one protein had shifted a corresponding number of amino acids relative to the equivalent region in the other protein. If many displacements had taken place during evolution, it would be difficult to reconst,ruct them only on the basis of the data from RELATE. However, less complicated rearrangements like a transposition could conceivably be de- tected.

Consider two proteins, each 120 amino acids long. Suppose that protein I1 evolved from protein I by transposition of the last 30 residues of 1 from the carboxyl to the NH2 terminus (Fig. 9). RELATE would then find homologies between seg- ments in region A (1-90 in 1 and 31-120 in II), as well as in region B (91-120 in I and 1-30 in 11). For the homology in region A, the program would compute a distance of 30 amino acids separating the conserved domains (i.e. protein I1 position - protein I position). For the homology in region B, the distance would be -90 amino acids (again I1 - I). Among the high scores from RELATE, there should be a pair of distances, one positive and one negative, the sum of whose absolute magnitudes is the length of the protein.

We now inspected the RELATE data for the comparison of eukaryotic and prokaryotic sequences. We found that, in each case, the distances which yielded high scores included a pair that met the criterion for a transposition (Table V). The sum of each pair of distances lies between 111 and 120, the lengths of the eukaryotic and prokaryotic proteins. Furthermore, homologous fragments were clustered rather than scattered ran- domly. The distances (28 to 35) came from fragment comparisons encompassing residues 1-60 in the prokaryotic A proteins and 33-92 in the eukaryotic proteins. Similarly, the dis,ance (-81 to -86) came from fragment comparisons rang- ing from positions 90-120 in the prokaryotes and 5-35 in the eukaryotes. These results, then, are what would be predicted for a transposition (Fig. 9).

TO test the prediction of a transposition, we repeated the analysis with ALIGN in the following manner: for each comparison with a prokaryotic protein, the eukaryotic sequence was partially duplicated so that the first 68 residues were added to the carboxyl terminus. Since these residues were still present at the NH2 terminus, the total sequence length was


0 *"l J - I l l

I

Y I I I

~ O K

w m a I l l

n a w I l l

0 2 0 I l l

I l l

w w w I l l

H H a I l l

( 3 ( 3 ( 3

I l l

u v >

a x w I l l

N N

9 196 Primary Structure of Ribosomal Protein P2

TABLE 111 Comparison of the sequences of ribosomal A proteins of different

species obtained with the computer program RELATE The proteins compared are E . coli L7/L12, M. lysodeikticus (M.

lyso.) MPI, B. subtilis (B. sub.) L9, H. cutirubrum (H. cut.) L20, rat liver P2, yeast (S. cerevisiae) YP A l , A. salina (A. sal.) eL12. For the comparison, the fragment length was 20 residues, and 100 random comparisons were made. The values are the number of standard deviations of the real score above the random score; three standard deviations correspond to p < 0.001.

E. coli M. lyso. B. sub. 2;. Rat Yeast 2;. E . coli X M. lyso. 11.9 X B. sub. 11.8 12.8 X H. cut. 1.0 2.2 1.6 X Rat 3.8 4.0 3.8 8.2 X Yeast 2.8 3.2 4.1 6.1 11.2 X A. sal. 1.5 3.4 3.0 3.1 13.0 12.0 X

831 120

1 A I

FIG. 9. A scheme for the transposition of a portion of ribosomal A proteins.

TABLE IV Comparison of the sequences of ribosomal A proteins of different

species obtained with the computer program ALIGN The proteins compared and abbreviations are given in Table 111.

For the comparison, the penalty was 40 and the bias 40. The values are the number of standard deviations of the real score above the random score; three standard deviations correspond top < 0.001.

E . coli M . lyso. B . sub. 5;. Rat Yeast ,",t, E . coli X M. lyso. 25.5 X B. sub. 19.7 27.8 X H. cut. -0.8 -0.1 -0.7 X Rat 0.1 0.0 -0.8 6.7 X Yeast -0.7 -0.4 -0.7 7.6 22.4 X A. sal. -1.3 -0.2 -1.8 3.7 21.0 19.9 X

now 179 amino acids. The ALIGN program was run so that no gaps would be inserted into either sequence. As a result, the prokaryotic protein was forced to line up en masse against some contiguous segment of the partially duplicated eukaryotic sequence. In each case, the optimal alignment was with the fiist residue of the prokaryotic protein 30-45 amino acids downstream from the NH, terminus of the eukaryotic protein. Thus, when given the option, ALIGN chose to match an intact prokaryotic sequence with an intact eukaryotic sequence by transposing the first 30-45 residues to the carboxyl terminus.

Since this finding was consistent with the data from RE- LATE, we next tried to localize the site of transpotion by using ALIGN. The preceding results and the distance data (Table V) indicated that the site was in the region between amino acids 30 and 45 for the eukaryotic proteins and perhaps beyond residue 20 for the halophile H. cutirubrum. To determine the precise site, we used ALIGN to compare the prokaryotic sequences to all possible transposed sequences suggested by the limits given above. Thus, for rat liver P2, the modified protein starting at position 30 and extending to position 111,

followed by residues 1-29 (i.e. (30-111, 1-29)), was compared to the three prokaryotic protein sequences. The same comparison was undertaken for rat liver P2 (31-111, 1-30) and (32-111,1-31) etc., up to (50-111,l-49). The most convenient procedure in this special case, where a given set of artificially modified eukaryotic ribosomal proteins is compared with a particular prokaryotic protein sequence, is to use the so-called "real score" (the number obtained for a particular alignment by summing of the scores for the residue by residue comparison recorded in the mutation data matrix). For each eukaryotic ribosomal A protein, the real score rose when the transposition was of 30 residues from the NH2 to the carboxyl terminus, peaked when the transposition was of 36 amino acids, and then fell off. Moreover, the maximum score was always near, if not at the same position (amino acid 36). From this pattern, the site of the conjectured transposition can be predicted with some confidence.

For an assessment of possible evolutionary relationships, it was necessary to consider the distance, in S.D. units, of the real score above the mean random score (Table VI). We now made another comparison between the eukaryotic proteins (rat P2, yeast YP Al, and A. salina eL12) and H. cutirubrium L20 and the proteins from prokaryotic ribosomes (E. coli L7/ L12, M. lysodeikticus MP1, and B. subtilis L9) by using ALIGN. In this instance, however, the 36 residues at the NH2 terminus of the eukaryotic proteins were moved to the carboxyl terminus. The results now were in conformity with those obtained with RELATE in indicating that the eukaryotic and prokaryotic ribosomal A proteins were homologous (Table VI).

In summary, then, the computer data indicate that a transposition occurred during evolution of the ribosomal A proteins. I t is not possible to determine from the results whether it is the present eukaryotic or prokaryotic ribosomal A proteins which most closely approximate the ancestral sequence. It is conceivable that the contemporary eukaryotic proteins evolved from a progenitor without transpositions, and that

TABLE V Distance data for the comparison of the sequences of ribosomal A proteins of different species obtained with the computerprogram

RELATE The proteins compared and abbreviations are given in Table 111.

The values are the pairs of distances that yield high scores when the proteins are compared. The distances are computed from the protein I1 position - protein I position (Fig. 9).

I sequences

E . coli M. lyso. B. sub.

Rat 34, -83 35, -81 34, -85 Yeast 33, -83 34, -81 33, -85 A. sal. 32, -83 33, -81 32, -85 H . cut. 28, -84 29, -82 28, -86

I1 sequences

TABLE vr Comparison of the sequences of ribosomal Aproteins of different

species obtained with the computer program ALIGN The proteins compared and abbreviations are given in Table 111.

For the comparison, the penalty was 100 and the bias 100. The values are the number of standard deviations of the real score above the random score; three standard deviations corresponds t o p < 0.001.

E. coli M. lyso. B. sub.

Rat" 4.3 4.4 3.4 Yeast" 3.8 3.7 5.6 A. sal." 3.5 3.3 2.9 H. cut." 2.8 4.0 3.2

The 36 residues at the NH, terminus were moved to the carboxyl terminus.


the prokaryotes, during their evolution, moved the first 36 residues of this progenitor sequence from the NHI to the carboxyl terminus (the reverse of the scheme shown in Fig. 9) . It is worth noting that a rearrangement during evolution of a portion of the sequence of the ribosomal A proteins, albeit one involving gene duplication, had been suggested before (45).

It would appear, then, that the ribosomal A proteins are among the most ancient and highly conserved of cellular proteins. Clearly, the proteins have been conserved during evolution from an archebacterium (H. cutirubrum) to a fungus (yeast), to an invertebrate (the brine shrimp A. salina), and to a mammal (rat). Our results indicate that the ancestry of P2 goes back even further, and that P2 is related to E. coli L7/L12, to M. lysodeikticus M P 1 , and to B. subtilis L9.

REFERENCES 1. Wool, I. G. (1979) Annu. Rev. Biochem. 48, 719-754 2. Wittmann-Liebold, B., Geissler, A. W., Lin, A., and Wool, I. G.

(1979) J. Supramol. Struct. 12,425-433 3. Terhorst, C., Moller, W., Laursen, R., and Wittmann-Liebold, B.

(1973) Eur. J. Biochem. 34, 138-152 4. Matheson, A. T., Moller, W., Amons, R., and Yaguchi, M. (1980)

in Ribosomes Structure, Function, and Genetics (Chambliss, G., Craven, G. R., Davies, J., Davis, K., Kahan, L., and Nomura, M., eds) pp. 297-332, University Park Press, Baltimore

5. Wittmann-Liebold, B., Robinson, S. M. L., and Dzionara, M. (1977) FEBS Lett. 81, 204-213

6. Luer, C. A,, and Wong, K. P. (1979) Biochemistry 18, 2019-2027 7. Subramanian, A. R. (1975) J. Mol. Biol. 95, 1-8 8. Stoffler, G . (1974) in Ribosomes (Nomura, M., Tissieres, A., and

Lengyel, P. eds) pp. 615-667, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY

9. Tsurugi, K., Collatz, E., Todokoro, K., Ulbrich, N., Lightfoot, H. N., and Wool, I. G. (1978) J. B i d . Chem. 253, 946-955

10. Collatz, E., Lin, A., Stoffler, G., Tsurugi, K., and Wool, I. G. (1976) J. Biol. Chem. 251, 1808-1816

11. Stoffler, G. (1967) Mol. Gen. Genet. 100,374-377 12. Lin, A., Collatz, E., and Wool, I. G. (1976) Mol. Gen. Genet. 144,

13. Laemmli, U. K. (1970) Nature (Lond.) 227,680-685 14. Lewis, W. G., Basford, J. M., and Walton, P. L. (1978) Biochim.

15. Wittmann-Liebold, B., Geissler, A. W., and Marzinzig, E. (1975)

16. Hirs, C. H. W. (1967) Methods Enzymol. 11, 197-199 17. Spiess, J. R., and Chambers, D. C. (1949) Anal. Chem. 21,

18. Edelhoch, H. (1967) Biochemistry 6, 1948-1954 19. Gross, E. (1967) Methods Enzymol. 11, 238-255 20. Vanderckhove, J., and Van Montagu, M. (1974) Eur. J. Biochem.

44,279-288 21. Wittmann-Liebold, B., and Lehmann, A. (1980) in Me&& in

Peptide and Protein Sequence Analysis (Birr, C., ed) pp. 49-72,

1-9

Biophys. Acta 522,551-560

J. Supramol. Struct. 3,426-447

1249-1266

Elsevier-North Holland Biomedical Press, Amsterdam

Lett. 93, 205-214 22. Chang, J. Y., Brauer, D., and Wittmann-Liebold, B. (1978) FEBS

23. Yang, C. Y. (1979) Hoppe-Seyler’s 2. Physiol. Chem. 360,

24. Chen, R. (1976) Hoppe-Seyler’s 2. Physiol. Chem. 357,873-886 25. Chang, J. Y. (1979) Biochim. Biophys. Acta 578, 188-195 26. Wachter, E., Machleidt, W., Hofner, H., and Otto, J. (1973) FEBS

27. Laursen, R. A., Horn, M. J., and Bonner, A. G. (1972) FEBS Lett.

28. Edman, P., and Begg, G. (1967) Eur. J. Biochem. 1,80-91 29. Wittmann-Liebold, B., Brauer, D., and Dognin, M. J. (1979) in

Solid Phase Methods in Proteins Sequence Analysis (Previero, A., and Coletti-Previero, M. A., eds) pp. 219-232, Elsevier- North Holland Biomedical Press, Amsterdam

30. Wittmann-Liebold, B. (1981) in Chemical Synthesis and Se- quencing of Peptides and Proteins (Liu, T., Schecter, A,, Heinrikson, R., and Condliffe, P., eds) pp. 75-110, Elsevier- North Holland Biomedical Press, Amsterdam

31. Wittmann-Liebold, B., Graffunder, H., and Kohls, H. (1976) Anal. Biochem. 75,621-633

32. Dayhoff, M. 0. (1978) in Atlas ofprotein Sequence andStructure, Vol. 5, Suppl. 3, pp. 1-8, National Biomedical Research Foun- dation, Washington, D. C.

33. Barker, W. C., Ketcham, L. K., and Dayhoff, M. 0. (1978) J . Mol. Evol. 10,265-281

34. Towbin, H., Ramjoue, H. P., Kuster, H., Liverani, D., and Gordon, J. (1982) J. Biol. Chem. in press

35. Amons, R., Van Agthoven, A., Pluijms, W., and Moller, W. (1978) FEBS Lett. 86,282-284

36. Lindemann, H., and Wittmann-Liebold, B. (1979) Hoppe-Seyler’s Z . Physiol. Chem. 358,843-863

37. Dzionara, M., Robinson, S. M. L., and Wittmann-Liebold, B. (1977) J. Supramol. Struct. 7, 191-204

38. Dzionara, M., Robinson, S. M. L., and Wittmann-Liebold, B. (1977) Hoppe-Seyler’s 2. Physiol. Chem. 358, 1003-1019

39. Burgess, A. W., Ponnuswamy, P. K., and Scheraga, H. A. (1974) Isr. J. Chem. 12,239-286

40. Chou, P. Y., and Fasman, G. D. (1974) Biochemistry 13,211-222 41. Chou, P. Y., and Fasman, G. D. (1974) Biochemistry 13, 222-249 42. Chou, P. Y., Adler, A. J., and Fasman, G. D. (1975) J. Mol. Biol.

43. Nagano, K. (1977) J. Mol. Biol. 109, 251-274 44. Robson, B., and Suzuki, E. (1976) J. Mol. Biol. 107, 327-356 45. Amons, R., Pluijms, W., and Moller, W. (1979) FEBS Lett. 104,

46. Itoh, T. (1980) FEBSLett. 114, 119-123 47. Oda, G., S t r ~ m , A. R., Visentin, L. P., and Yaguchi, M. (1974)

48. Itoh, T., and Wittmann-Liebold, B. (1978) FEBSLett. 96,392-394 49. Itoh, T. (1981) FEBS Lett. 127,67-70 50. Itoh, T., Higo, K., Otaka, E., and Osawa, S. (1980) in Genetics

and Evolution of RNA Polymerase, tRNA and Ribosomes (Osawa, S., Ozeki, H., Uchida, H., and Yura, T., eds) pp. 609-624, University of Tokyo Press, Tokyo

1673-1675

Lett. 35.97-102

21, 67-70

96, 29-45

85-89

FEBS Lett. 43, 127-130

the primary structure the acidic phosphoprotein p2 from ...9190 primary structure of ribosomal...

Documents