Proc. Natl. Acad. Sci. USAVol. 74, No. 11, pp. 4900-4904, November 1977Biochemistry

Sequence of an oligonucleotide derived from the 3' end of eachof the four brome mosaic viral RNAs

(aminoacylation/tRNA-like structure)

RANJIT DASGUPTA AND PAUL KAESBERGBiophysics Laboratory of the Graduate School and Biochemistry Department of the College of Agricultural and Life Sciences, University of Wisconsin, Madison,Wisconsin 53706

Communicated by Heinz Fraenkel-Conrat, August 29, 1977

ABSTRACT A -3'-terminal oligonucleotide fragment, 161bases long, can be obtained from each of the four brome mosaicvirus BNAs by means of nuclease digestion. Like the four intactbrome mosaic virus RNAs, each fragment accepts tyrosine in-a reaction catalyzed by wheat germ aminoacyl-tRNA synthetase.The complete nucleotide sequence of the RNA 4 fragment hasbeen determined by use of standard radiochemical methods.Comparative data for the fragments from RNAs 1, 2, and 3 showthat-the have nearly the same sequence as the RNA 4 fragment.The eight bases adjacent to the 3' terminus of the RNA 4 frag-ment are identical in sequence to the eight terminal bases oftyrosine tRNA from Torula utilis and eleven interior bases areidentical in sequence to eleven bases encompassing the anti-codon region of tyrosine tRNA from Saccharomyces cerevisiae,T. utilis, and Escherichia coli. Nevertheless, reasonable base-pairing schemes yield, at best, a distorted cloverleaf secondarystructure.

Nucleotide sequence analysis of the extremities of plant viralnucleic acids has provided substantial information regardingtheir structure and their function as messengers. For example,the 5'- termini of some of these RNAs, in common with mosteukaiyotic messengers, contain the "cap" structure m7GpppGp.(For an extensive review, see ref. 1.) The efficiency of transla-tion is dependent on the presen'ce of cap (2-4). For bromemosaic virus (BMV) RNA 4, the monocistronic messenger forBMV coat protein, the initiation codon is only 10 nucleotidesfrom the capped 5' end and this short sequence, together withthe cap, constitutes an effective ribosome binding site (5).iThe 3' termini of some plant viral RNAs have the unique

property among messengers that they can quantitatively acceptan amino acid in a reaction catalyzed by aminoacyl'tRNAsynthetase. Thus, turnip yellow mosaic virus RNA (6) and eggplant mosaic virus RNA (7) can be charged with valine; tobaccomosaic virus RNA can be charged with histidine (8). Each ofthe four BMV RNAs can be charged with tyrosine (9). Somepicornaviral RNAs can be charged, albeit inefficiently andprobably only after they have been fragmented (10, 11). Theproperty of chargeability implies that these RNAs have atRNA-like structure and possibly also a tRNA-like function (12).A secondary structure, somewhat like that of tRNA, is com-patible with the sequences determined for the 3'-end regionsof turnip yellow mosaic virus RNA and possibly egg plantmosaic virus RNA (13, 14). The nucleotide sequence at the 3'-terminal region of tobacco mosaic virus RNA seems less ame-nable to folding into a cloverleaf secondary structure (15).

Although the BMV RNAs can be aminoacylated with tyro-sine, this amino acid is not donated to nascent peptides uponin vitro translation (16).> However, integrity of the 3' end is

necessary for infectivity of BMV RNA (17). Thus, like their 5'ends, the 3' ends of these viral RNAs have a distinctive structureand presumably an important, although unknown, function.We have reported that partial hydrolysis of each of the four

BMV RNAs with RNase T1 releases a 3'-terminal fragmentabout 160 nucleotides long-and that this fragment is virtuallyas efficient in accepting tyrosine as the intact BMV RNAs (18).The present paper reports the complete nucleotide sequenceof the fragment derived from RNA 4 and gives data indicatingthat the fragments from the other three RNAs have nearly thesame sequence. They are each 161 bases long, and we designatethem Q161 of BMV RNAs 1, 2, 3, and 4.

MATERIALS AND METHODSMaterials. T1, T2, and U2 RNases were obtained from

Calbiochem. Pancreatic ribonuclease, snake venom phospho-diesterase, and bacterial alkaline phosphatase were obtainedfrom Worthington Biochemical Corp. Nuclease P1 was ob-tained from Yamasa Shoyu Co., Ltd. (Tokyo, Japan). T4 poly-nucleotide kinase was obtained from P-L Biochemicals, ['y-32P]ATP, at a specific activity of 1500-2000 Ci/mmol, wasobtained from Amersham/Searle Corp. Nuclease S1, purifiedby the method of Vogt (19), was a gift from James E. Dahlbergof the University of Wisconsin. Thin-layer plates (CEL 300DEAE or CEL 300 DEAE/HR-2/15) were purchased fromMacherey-Nagel and Co.

Preparation of BMV RNA and Fragment Q161. The pro-cedures for growth and radioactive labeling of BMV and iso-lation and fractionation of its RNAs have been described(20-22). Partial digestion of the individual BMV RNAs withRNase T1 and subsequent isolation of Q161 by electrophoresison polyacrylamide gels have been described (18). Except whereexplicitly noted, all our descriptions refer to Q161 isolatedseparately from each of the four RNAs.

Sequence Analyses. Complete digestion of 0161 with eitherT1 RNase or pancreatic ribonuclease, separation of the resultingoligonucleotides by two-dimensional electrophoresis or byelectrophoresis-homochromatography, and further analysis ofthese oligonucleotides were according to the methods of Sangerand his colleagues (23, 24).The larger oligonucleotides, especially those rich iii pyri-

midines, were also analyzed by the wandering spot procedureas described by Silberklang et al. (13). Oligonucleotides werelabeled at their 5' ends with 32P by use of [y-32P]ATP andpolynucleotide kinase according to the procedures of Simseket al. (25).T1 and pancreatic ribonuclease oligonucleotide catalogs were

obtained for a variety of large and small pieces of Q161, espe-cially those having overlapping sequences. To obtain large

Abbreviation: BMV, brome mosaic virus.

4900

The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby marked"advertisement" in accordance with 18 U. S. C. §1734 solely to indicatethis fact.

Dow

nloa

ded

by g

uest

on

July

7, 2

021

Proc. Natl. Acad. Sci. USA 74 (1977) 4901

,S P21 P19 +P20O

VP

P15P16 *P14 1PI2

P12SPIN PIO

* * P8*7

* P6

P5 P4T2

*PIPi!

FIG. 1. Autoradiograms of 1161 digests. All oligonucleotides wereanalyzed by pancreatic, T1, U2, and snake venom (preceded by al-kaline phosphatase) ribonucleases and some also by the wanderingspot technique. (Left) Two-dimensional fractionation of a completeT1 RNase digest of 1161. The first dimension was electrophoresis oncellulose acetate at pH 3.5; the second dimension was homochroma-tography on a DEAE-cellulose (DEAE/HR 2/15) thin-layer plate. Thedotted circle marked B indicates the position of the blue dye. Notethat the oligonucleotide A-C-C-AOH (spot T8) is readily identifiablefrom its unusual position on the fingerprint. The spots T1-T26 andtheir relative molar yield are: T1, Gp, 8.4; T2, C-Gp, 1.2; T3, A-Gp,3.8; T4, C-A-Gp, 0.8; T5, A-A-Gp, 0.9; T6, C-A-A-Gp, 1.1; T7, A-C-A-C-Gp, 0.7; T8, A-C-C-AOH, 0.8; T9, U-Gp, 4.0; T10, C-A-U-Gp, 1.0;Til, U-A-C-C-Gp, 0.8; T12, U-A-C-A-Gp, plus T13, C-A-U-A-Gp,2.0; T14, A-A-U-Gp, 1.0; T15, U-U-Gp, 1.1; T16, C-U-U-Gp, plus T17,U-U-C-Gp, 2.0; T18, U-C-U-A-Gp, 1.2; T19, A-A-A-A-A-C-A-C-U-Gp, 0.9; T20, A-A-C-C-C-U-U-A-Gp, 0.9; T21, A-C-C-U-C-U-U-A-C-A-AGp, 1.0; T22, U-A-A-A-U-C-U-C-U-A-A-A-A-Gp, 1.1; T23,C-U-C-U-C-U-U-C-Gp, 0.9; T24, C-C-U-U-U-Gp, 1.1; T25, U-C-U-U-A-Gp, 0.8; and T26, U-U-A-C-U-C-U-U-U-Gp, 1.0.

(Right) Two-dimensional fractionation of a complete pancreaticribonuclease digest ofQ161. The oligonucleotides were separated byelectrophoresis in the first dimension on cellulose acetate at pH 3.5and in the second dimension on DEAE-cellulose paper in 7% formicacid. The spots Pl-P21 and their molar yield are: P1, Up, 28.7; P2,Cp, 16.6; P3, A-Cp, 7.5; P4, G-Cp, 5.2; P5, A-Up, 1.9; P6, A-G-Cp, 0.9;P7, A-G-A-Cp, 1.0; P8, A-A-A-Up, 0.9; P9, G-Up, 5.0; P10, G-A-A-A-A-A-Cp, 0.7: P11, A-G-Up, 1.2; P12, A-A-G-Up, 1.1; P13, A-G-A-A-Up, 1.0; P14, G-G-G-Cp, 0.9; P15, A-A-A-A-G-A-G-A-Cp, 0.9; P16,A-G-G-Up, 1.0; P17, A-A-G-A-G:-Up, 1.2; P18, G-G-A-A-G-A-A-Cp,0.9; P19, G-A-G-A-G-Up, 0.9; P20, G-G-G-Up, 0.9; P21, A-G-G-G-G-Up, 0.5.

overlapping pieces, 0161 was digested with 20 ng of pancreaticribonuclease or 400 ng of T1 RNase per 200 jig of RNA. Theincubation was at 00 for 20 min. To obtain comparativelyshorter pieces (10-30 nucleotides), Q161 was digested with T1or pancreatic ribonuclease at an enzyme to substrate ratio of1:100. The incubation was at 40 for 10 min. The 0161 fragmentwas tested for the presence. of abnormal bases by two-dimen-sional thin-layer chromatography on cellulose as described byNishimura (26). Conditions for digesting 0161 with nucleaseS1 were similar to those of Rushizky and Mozejko (27).

RESULTSLimited digestion of BMV RNA 4 with T1 RNase and subse-quent fractionation by electrophoresis on polyacrylamide gelsproduces one major band and many minor bands correspondingto fragments of various lengths (18). The major band fragment,designated 0161, is the only one produced quantitatively, andit can be obtained readily in pure form. On complete digestion

with RNase T1, this fragment gives rise to the 3'-terminal oli-gonucleotide A-C-C-AoH, indicating that 0161 is cleaved fromthe 3' end of RNA 4.Sequence Analysis of RNase T1 and -Pancreatic Ribonu-

clease Oligonucleotides of RNA 4 (161. Fragment 0161 ofBMV RNA 4, uniformly labeled with 32p (specific activity 109dpm/mg of RNA), was digested to completion with RNase T1and the products were separated and characterized. A two-dimensional separation by cellulose acetate electrophoresis andDEAE-cellulose thin-layer homochromatography is shown inFig. 1 left. All products were well resolved except the isomersT12 and T13 and T16 and T17. Separation of similar qualityis obtainable by two-dimensional electrophoresis (18). Fig. 1right illustrates complete separation by two-dimensionalelectrophoresis of the products of digestion of 0161 by pan-creatic ribonuclease. The sequences of most of the pancreaticribonuclease and of the short T1 RNase oligomerswere deter-mined by analysis with the complementary RNase T2 RNase,snake venom phosphodiesterase, and U2 RNase. The oligomersfor which such analyses were not definitive were examined, inaddition, by the wandering spot procedure; these includedT17-T26 and P15, P17, P18, and P21. Some typical wanderingspot analyses, those of oligomers T21, T24, T26, and P18, areshown in Fig. 2. Fragment 0161 of BMV RNA 4 was digestedwith a mixture of ribonucleases and the products were analyzedfor the presence of unusual bases. No spotswere detected otherthan those of the common nucleotides A, C, G, and U.

Ordering of Oligonucleotides and the Sequence of RNA4 0161. Fragment Q161 of BMV RNA 4 was partially digestedwith T1 or pancreatic ribonuclease and the pieces were sepa-rated by polyacrylamide gel electrophoresis or by celluloseacetate electrophoresis and homochromatography. The purifiedproducts were analyzed as had been 0161 itself. Enough piecesof different chain lengths and with overlapping sequences wereobtained to permit ordering of all Ti and pancreatic oligonu-cleotides except those between residues 70 and 74. We couldnot distinguish between the sequences C-A-U-G-G-G-C-U-U-G-C-A-U-A-Gp and C-A-U-G-C-U-U-G-G-G-C-A-U-A-Gp, since both sequences gave rise to the same products aftera variety of partial digestions with T1 and pancreatic ribonu-clease. This ambiguity was resolved by 32P end-labeling of theTi partial product corresponding to residues 69-74, followedby wandering spot analysis as shown in Fig. 3.

161 150 140A -C-A-C-G-C-A-G-A-C-C- U -C-U-U-A-C-A-A-G-A- G- U-G-U-C-U-A-G-G-U-130 120 110 100G -C-C-U-U-U-G-A-G-A- G -U-U-A-C-U-C-U-U-U- G -C-U-C-U-C-U-U-C-G- G

90 80 70A-A-G-A-A-C-C-C-U- U -A-G-G-G-G-U-U-C-G- U -G-C-A-U-G-G-G-C-U- U -G-

60 50 40C-A-U-A-G-C-A-A- G -U-C-U-U-A-G-A-A-U- G -C-G-U-A-C-C-G-G-G- U -G-U-

30 20 10A-C-A-G-U-U-G- A -A-A-A-A-C-A-C-U-G- U -A-A-A-U-C-U-C-U-A- A -A-A-G-A-G-A-C-C-A-OH

Sequences of 0161 from BMV RNAs 1, 2, and 3. The Tiand pancreatic ribonuclease maps for Q161 from RNAs 1, 2,and 3 were identical to those of RNA 4 except as follows: (a) In0161 from RNAs 1 and 2, spot 11 (U-A-C-C-Gp) was missingand was replaced by spots corresponding to the oligomers U-Gpand U-C-Gp in RNA 1 and U-Gp and C-C-Gp in RNA 2. (ii)A-Cp (spot P3) was reduced in amount while G-Up (spot P9)was increased for RNA 1 and G-Cp (spot P4) was increased forRNA 2. Other possible differences are not precluded, e.g., se-quence isomers that migrate identically both in electrophoresisand homochromatography. Because extensive differences ofthis kind are unlikely and would, furthermore, be unlikely tohave a bearing on our conclusions, we chose, for the present,to ignore these possibilities.

T212@9 OT19* T21 @T20T26 S

T23

T25%T24 T13 *T7

T14#p *T6T16+TI70 Te T5T5T

,#WT15 OT4

T9 *T3 TO

Biochemistry: Dasguipta and Kaesberg

Dow

nloa

ded

by g

uest

on

July

7, 2

021

4902 Biochemistry: Dasgupta and Kaesberg

c

\c

U

._

\G0

-.B_m* G\GA).*A I

IA.r

IAjc

FIG. 2. Autoradiogram of partial snake venom phosphodiesterasedigests (Upper left and right) and partial nuclease P1 digests (Lowerleft and right) of 5'-32P-labeled spots T21 (Upper left), T24 (Upperright), T26 (Lower left), and P18 (Lower right) of Fig. 1. CompleteT1 and pancreatic ribonuclease digests of 0161 were labeled with[,y-32PJATP using polynucleotide kinase and the products were sep-arated. 5'-32P-Labeled spots were then eluted, partially digested withthe above enzymes, and analyzed. The first dimension was electro-phoresis on cellulose acetate, pH 3.5; the second dimension washomochromatography on thin-layer plates made of either pureDEAE-cellulose (CEL 300 DEAE; Upper left and right) or a mixtureof DEAE-cellulose and cellulose in the ratio 2:15 (CEL 300 DEAE/HR2/15; Lower left and right). The homochromatography mixture usedwas a 3% solution of yeastRNA in 7M urea that had been hydrolyzedfor 45 min, prepared according to Barrell (24). The mononueleotideof T26 waS partially trapped in the paper wick.

In all probability RNA 4 (161 and RNA 3(161 are identical,RNA 2 (161 differs only in that base 46 is G, and RNA 1 1161differs in that base 46 is G and base 45 is U.

Sequence Similarity to tRNAs. The susceptibility of q161to aminoacylation with tyrosine suggests a structural similarityto tyrosine tRNA. Is this resemblance reflected in the (161sequence? Comparison shows that only a limited correspon-dence to authentic tyrosine tRNAs exists. The sequence of theseven and eight bases adjacent to the 3' terminus of (161 isidentical to that of the corresponding bases of Saccharomycescerevstsae and Torula utilis tyrosine tRNA, respectively (28).Also eleven bases in the sequence of (161, residues 14-24, havethe same sequences as eleven bases in the anticodon loop regionof S. cerevisiae, T. utilis, and Escherichia coli tyrosine tRNAexcept that in these tRNAs, three of the bases are modified.However, the eleven corresponding bases are not in sequentially

A- Uc

\G

\G

.!,

-~~~~~~~~~~~~~~~~~~~~~~~~~~~-k1:;".~~~~~~~~~~~~~~~~~~~~~~~~~~~~~Il

2.. 3

oI*I*C/A

4A

FIG. 3. Autoradiogram of a partial nuclease P1 digest on a 5'-32P-labeled RNA fragment obtained by partial T1 RNase digestionof 0161. Partial T1 digestion was done on a mixture of uniformly32P-labeled 0161 (to aid in locating the bands) and unlabeled 0161and the products were separated on 10% polyacrylamide gels. Cor-responding bands were then eluted, labeled with kinase at their 5' end,and purified further by two-dimensional electrophoresis and homo-chromatography.

similar locations with respect to their 3' ends. Thus,24 14 1

n 161 A-C-U-G-U-A-A - A-U-C-U...A-G-A-G-A-C-C-AT. utilis Tyr tRNA . . A-C-U-G-* -A-i6A-A-' -C-U ... .A-G-A-G-A-C-C-A

46 36 1

Other sequence correspondences to Tyr tRNAs are five baseslong or less.

Secondary Structure and Base Pairing. Secondary structureis not obviously revealed by inspection of base sequence. Nev-ertheless some inferences can be drawn by considering struc-tures that maximize the number of Watson-Crick base pairs.For RNA 4 Q161, the secondary structure of highest stabilityaccording to the rules of Tinoco et al. (29) is-shown in Fig. 4.As indicated in that figure, all but one T1 RNase cleavage pointsof high susceptibility are located in the regions lacking basepairs. The single exception is between the C-A bond in positions77-78. A number of essentially equivalent base-pairing schemesexist. For example, the sequence A-A-G-A-G-A in positions 4-9can pair with U-C-U-C-U-U in positions 103-108. X-ray crys-tallography shows that in tRNAs the classical cloverleaf modelaccurately reflects actual base-pairing (although additional baseinteractions exist that are not revealed by cloverleaf folding).For 1161, can a cloverleaf structure be drawn with an accessibleA-C-C-A terminus and a tyrosine anticodon centered on an"anticodon" loop? Both features are believed to be involved inrecognition by the charging enzyme (30). With some alternativefolding and with elimination of some marginal base pairs thesecondary structure of Fig. 4 can be converted to the morecloverleaf-like structure shown in Fig. 5. However, we areunable to construct a secondary structure of high stability thatprovides an "anticodon" loop centered on 1161 bases 14-24.

It is possible, of course, that tertiary interactions, unrevealedby cloverleaf base pairs, determine the charging enzyme rec-ognition features. Regardless of detailed knowledge of tertiarystructure, it is to be expected that an enzyme recognition sitewould be near the surface of a substrate molecule and thus

GA'C

/Cx

_

Proc. Natl. Acad. Sci. USA 74 (1977) 4903

AC145 UCU A C UUA A UCG *U UUA - UGCGu GCCG AUAU GC

N GC z AU"k AUX GCC AGGUGCCUUU U

,.,tG UC CA \ G

A U AG161A 90U CCCAAGAIPA J-1d-Isot

A1c

AGAGA A1AUCUCUA

A it

AUGUCACAA28UACAGUG

G>X AA

G

,C

A47, A GGGUUC G

G UGCGCG

-AU /UAAG

G A55GCUU NkGU \CGUAUAGC

C GA UA66

FIG. 4. Possible secondary structures for RNA 4 Q161 drawn tomaximum base-pairing. Preferred sites of action of T1 and pancreaticRNases are indicated by solid and broken arrows, respectively.

might be especially vulnerable to nuclease attack. The U-Abond in the "anticodon" region of the structure in Fig. 5 (po-sitions 66-65) is the most susceptible pancreatic ribonucleasepoint in Q161. With an enzyme to substrate ratio of 1:10,000,it was possible to obtain a break only at this point in Ql161,separating the molecule into two parts. The isolated fragmentswere no longer chargeable. However, a molecule with a hiddenbreak at this point, with the halves still noncovalently bound,was chargeable (M. Bastin and P. Kaesberg, unpublished ob-servations), indicating that the structure on both sides of this

145AC

U A

UA 118CG UUA U ACGU GCCG AU AAU GC CGC AU CAU G UUGCU A

C AGGUGCCUUU CGG UAC CGA UAC U^AAUUA CUAAAU 15

161 G UCG A CAAA AUGUCA A

U A UACAGU A90U CCCAAG GU63A GGGUUC U G

G G GGU U C-

G GGC UA AUC U GU G GA CG UC 50

UA GUA

G CC 6AU 65

FIG. 5. Slightly modified secondary structure for RNA 4 Q161drawn to illustrate a similarity to the cloverleaf structure of tRNA.

0

0

>: ~~~~~~~~~~dF> I

b

~~i ~ a C

0 20 40 60 80 100FRACTION NUMBER

FIG. 6. Polyacrylamide gel pattern of RNA 4 Q161 that has beendigested with nuclease S1. The digest was precipitated with ethanol,dissolved in buffer, and then applied to a 10% polyacrylamide gel. Theelectrophoresis was carried out at 3 mA for 2.5 hr, at which time thebromophenol dye had moved 10 cm. The gel was then crushed in aGilson automatic gel crusher. Fraction 1 corresponds to the bottomof the gel.

anticodon-like feature, although not the integrity of the anti-c6don itself, is necessary. This is similar to the situation in sometRNAs in which a hidden break in the anticodon loop does notpreclude aminoacylation (30).

Digestion of Q161 by Si Nuclease. S1 nuclease preferen-tially cleaves in the anticodon loop of tRNAs (31). It is possibleto digest tRNA to yield more than 90% half-molecules, that arestable to 50-fold higher concentration of enzyme (27). We thusundertook a study of the digestion of Q161 to ascertain whethera similar preferential cleavage site existed. (161 was digestedwith nuclease S1 under conditions optimal for cleaving tRNAsin their anticodon loops. After digestion, the products werefractionated on 10% polyacrylamide gels. Fig. 6 shows an ex-periment in which 30,000 cpm of 32P-labeled 2161, togetherwith 50Mg of unlabeled yeast RNA, were digested with 80 unitsof enzyme in 0.3 M NaCl, 0.05 M sodium acetate (pH 5.7) for24 hr at 00 in a 50-,gl volume. Three major bands, designatedI, II, and III, and several minor bands were observed. The RNAfrom each band was eluted and subjected to electrophoresis andhomochromatography after complete digestion with RNase T1.The bands were identified by comparison of their T1 catalogswith those of 161. Band I contains all Ti oligonucleotides; itis undigested 161. Band II contains, in equimolar yield, all Tioligonucleotides on the 5' side of oligonucleotide T13 (C-A-U-A-Gp) and is thus a cleavage product encompassing residues69-161. Band III contains (i) in equimolar amount all Tioligonucleotides on the 3' side of oligonucleotide T13 exceptT8, (ii) T8 in less than equimolar amount, and (iii) small oli-gomers, in less than equimolar amount, which we believe areportions of T13 or T8. Band III is thus heterogeneous; it containsresidues 1-68, but residues 1-4 and 63-68 exist only fractionally.We believe that minor bands a, b, c, and d represent mixturesof SI cleavage products. Bands a and b contained most T1oligonucleotides except C-A-U-A-Gp, A-C-C-AOH, and A-C-A-C-Gp. Bands c and d also contained most T1 oligonucleotides,but those near the 5' and 3' ends of Q1i61 were present in lowamounts. We conclude that S1 nuclease cleaved preferentiallynear residue 66, the "anticodon" region of Fig. 5.

DISCUSSIONEarlier publications and this study have shown that the fourBMV RNAs can be enzymatically aminoacylated with tyrosine

Biochemistry: Dasgupta and Kaesberg

Dow

nloa

ded

by g

uest

on

July

7, 2

021

4904 Biochemistry: Dasgupta and Kaesberg

in a manner similar to that of tyrosine tRNA, that a highlysusceptible ribonuclease T1 cleavage site exists 161 bases fromtheir 3' terminus, and that in each case the 3'-cleavage product(0161) can be aminoacylated. Moreover, 0161 is relativelyresistant to nuclease digestion; it is obtained in almost quanti-tative yield over a wide range of TI concentrations. The 0161molecules from each of the four RNAs have nearly the samenucleotide sequence. These facts suggest that the 3' end of theBMV RNAs has a structural resemblance to tyrosine tRNA, thatit is a tightly folded structure substantially retaining its con-figuration after cleavage, and (from its sequence conservation)that it plays an important role in the life cycle of the virus.

It was thus gratifying, at least for a short time, that the twolongest sequence identities among S. cerevisiae, T. utilis, andE. coli tyrosine tRNA, namely, their acylating terminus and(ignoring base modifications) their anticodon loop region, ex-isted also in 0161. However, these two regions of sequence wereseparated from each other by 28 bases in the tRNAs but by onlyfour in 0161 and, equally perversely, no manner of stablebase-pairing could produce a structure resembling a tRNAanticodon loop and stem. However, folding of Q161 into acloverleaf-like structure provides a stem and loop structure inwhich anticodon AUA (bases 65-67), rather than AUG (bases19-21) is centered on a loop, as shown in Fig. 5. Moreover, S1nuclease cleaves preferentially at that site just as it does at theanticodon of tRNA. Thus there is an evident region of resem-blance of tertiary structure of 0161 to tRNA but it does notinclude the 11-base-long region of sequence identity. The sig-nificance of the sequence identity is entirely unexplained. Itmay indicate a sequence recognized by the aminoacylatingenzyme, an evolutionary vestige, or coincidence. The problemof recognition of tRNA sites by their aminoacylating enzymesis a complex one and after a variety of studies over a period ofmany years is only partially solved. (For a comprehensive re-view, see ref. 30.) The corresponding problem with chargeableviral messenger RNAs is equally formidable and may not beworth pursuing until more is known about the functional sig-nificance of the charging.Our sequence data for 0161 do not provide immediate in-

sight regarding the function of the noncoding region at the 3'end of the BMV RNAs. Whatever that function may be, it hasresulted in an evolutionary constraint that provides nearlyidentical 0161 sequences for the four BMV RNAs even thoughthe proteins encoded in these RNAs are different.The 0161 regions of the BMV RNAs are not necessarily in-

volved in translation and amino acid transfer even though thisis an important function of tRNA. Indeed, there is suggestiveevidence to the contrary. BMV RNA that has been chemicallymodified to preclude acceptance of tyrosine is still fully capableof serving as a messenger in vitro (16). Certainly, pertinent invivo studies are needed.

Possibly, the 3' ends play an important role in initiation ofviral assembly.

Possibly the structure at the BMV RNA 3' ends is an impor-tant feature of RNA replication. Elongation factors whosenormal function is in translation are needed for the replicationof the RNA of phage QB (32). Tryptophan tRNA serves as aprimer in the synthesis of Rous sarcoma viral RNA (33). Perhapsalso, with viruses such as BMV, structures nominally associatedwith translation have a role in replication.We thank Christopher Sempos and Chris Saris for technical assis-

tance, Dr. G. Altman for computer analysis, and Drs. J. E. Dahlbergand D. Zimmern for helpful discussions. This research was supported

by Grant CA 15613 and Training Grant T32 CA09075 from the Na-tional Cancer Institute, Grant AI 01466 and Research Career AwardAI 21942 from the National Institute of Allergy and Infectious Diseases,and by Contract 1633 from the Biological Division of the Energy Re-search and Development Administration.

1. Shatkin, A. J. (1976) Cell 9,645-653.2. Both, G. W., Banerjee, A. K. & Shatkin, A. J. (1975) Proc. Natl.

Acad. Sci. USA 72, 1189-1193.3. Rose, J. K. & Lodish, H. F. (1976) Nature 262,32-37.4. Shih, D. S., Dasgupta, R. & Kaesberg, P. (1976) J. Virol. 19,

637-642.5. Dasgupta, R., Shih, D. S., Saris, C. & Kaesberg, P. (1975) Nature

256,624-628.6. Pinck, M., Yot, P., Chapeville, F. & Duranton, H. M. (1970)

Nature 226,954-956.7. Pinck, M., Generaux, M. & Duranton, H. (1974) Biochimie 56,

423-428.8. Oberg, B. & Philipson, L. (1972) Biochem Biophys Res. Commun.

48,927-932.9. Hall, T. C., Shih, D. S. & Kaesberg, P. (1972) Biochem. J. 129,

969-976.10. Salomon, R. & Littauer, U. Z. (1974) Nature 249,32-34.11. Lindley, I. J. D. & Stebbing, N. (1977) J. Gen. Virol. 34, 177-

182.12. Haenni, A. L., Prochiantz, A., Bernard, 0. & Chapeville, F. (1973)

Nature 241, 166-168.13. Silberklang, M., Prochiantz, A., Haenni, A. L. & RajBhandary,

U. L. (1977) Eur. J. Biochem. 72,465-478.14. Briand, J. P., Richards, K. E., Bouley, J. P., Witz, J. & Hirth, L.

(1976) Proc. Nati. Acad. Sci. USA 73,737-741.15. Guilley, J., Jonard, G. & Hirth, L. (1975) Proc. Natl. Acad. Sci.

USA 72,864-868.16. Shih, D. S., Kaesberg, P. & Hall, T. C. (1974) Nature 249,

353-355.17. Kohl, R. J. & Hall, T. C. (1977) Proc. NatI. Acad. Sci. USA, 74,

2682-2686.18. Bastin, M., Dasgupta, R., Hall, T. C. & Kaesberg, P. (1976) J. Mol.

Biol. 103, 737-745.19. Vogt, V. M. (1973) Eur. J. Biochem. 33, 192-200.20. Shih, D. S., Lane, L. C. & Kaesberg, P. (1972) J. Mol. Biol. 64,

353-362.21. Bastin, M. & Kaesberg, P. (1975) J. Gen. Virol. 26,321-325.22. Stubbs, J. D. & Kaesberg, P. (1967) Virology 33,385-397.23. Sanger, F., Brownlee, G. G. & Barrell, B. G. (1965) J. Mol. Biol.

13,373-398.24. Barrell, B. G. (1971) in Procedures in Nucleic Acid Research, eds.

Cantoni, G. L. & Davies, D. R., (Harper and Row, New York),Vol. 2, pp. 751-779.

25. Simsek, M., Ziegenmeyer, J., Heckmen, J. & RajBhandary, U.L. (1973) Proc. Natl. Acad. Sci. USA 70, 1041-1045.

26. Nishimura, S. (1972) Prog. Nucleic Acid Res. Mol. Biol. 12,49-85.

27. Rushizky, G. W. & Mozejko, J. H. (1977) Anal. Biochem. 77,562-566.

28. Barrell, B. G. & Clark, B. F. C. (1975) HandbookofNucleic AcidSequences (Joynson-Bruvvers Ltd., Oxford).

29. Tinoco, I., Jr., Borer, P. N., Dengler, B., Levine, M. D., Uhlen-beck, O., Crothers, D. M. & Gralla, J. (1973) Nature New Biol.246,40-41.

30. Chambers, R. W. (1971) Prog. Nucleic Acid Res. Mol. Biol. 11,489-525.

31. Harada, F. & Dahlberg, J. E. (1975) Nucleic Acids Res. 2,865-871.

32. Blumenthal, T., Landers, T. A. & Weber, K. (1972) J. Biol. Chem.249,5801-5808.

33. Faras, A. J., Dahlberg, J. E., Sawyer, R. C., Harada, F., Taylor,J. M., Levinson, W. E., Bishop, J. M. & Goodman, H. M. (1974)J. Virol. 13, 1134-1142.

Proc. Nati. Acad. Sci. USA 74 (1977)

Dow

nloa

ded

by g

uest

on

July

7, 2

021