Evolution of defective-interfering double-stranded RNAs of the yeast killer virus.

0022-538X/79/11-0692/05$02.00/0

Evolution of

Defective-Interfering Double-Stranded

RNAs of

the Yeast Killer

Virus

WILLIAM P. KANE,* DENNIS F.PIETRAS,t AND JEREMY A. BRUENN

Division of Cell and Molecular Biology, State University of New YorkatBuffalo,Buffalo, New York14260 Received for publication 26 June 1979

We havecharacterized by

T,

fingerprint analysisseveral defectiveinterfering

(DI) double-stranded RNAs of thesimpleyeastvirusScV.Acommonsequence

ofabout 0.5 to 0.6 kilobasepairs, including both 3' termini of the parental RNA, was presentin eachDIRNA.Several DI RNAs had novel

T,

oligonucleotidesnot

presentin theirparental RNA.

Saccharomyces~cerevisiae has asimple

dou-ble-stranded (ds) RNA virus (ScV) with two,

separately encapsidated

dsRNA's

(4,

10). Like

other fungal viruses (17) ScV exhibits only a

latent infection. The larger viral dsRNA (L) encodes the major capsid protein (14). The

smaller (M) encodes a toxin

("killer factor")

lethal to strains without ScV-M (S. cerevisiae

virus-M-containing) particles (1, 16; K. A. Bos-tian et al., Cell, in press). L and M have no

detectable sequencehomology (2), and ScV-M particles are dependent onScV-L

particles

for the synthesis of their capsid

protein

(8, 11). Deletion mutations of M result in defective-in-terfering (DI) particles containing fragments of

M(ScV-S particles) (2,7;thiswork). ScV-M and ScV-S

particles

areequally

dependent

on

prod-uctsof thenucleargenomefor their

replication,

products not necessary for the replication of ScV-L particles (19, 20). ScV-S RNAs are de-rived from M by internal deletion, sometimes followed by tandemduplication (2, 7), andthey diverge in sequence from their parental RNA

(2),asdothe RNAs of animal virusesreplicated inpersistent infections(13).

In the present work we utilize two-dimen-sionalgel electrophoresis of

T,

oligonucleotides

tocharacterizeseveral S RNAs, including three independently derived S RNAs from parental

strains with differentgeneticbackgrounds. The dsRNA's werephenolextracted andpurifiedby

CF11chromatography and agarose gel electro-phoresis (2, 3). The sizes of dsRNA's were

de-termined by electron microscopy and by

CH3HgOH agarose denaturing gel electropho-resis (2). The two new S dsRNA's examined wereS14,of0.78kb(kilobase pairs), from strain L014 (M. Vodkin); and S733, of 0.77 kb, from

strain K733(R.Wickner). The S dsRNA'swere present inScVparticleswith a density inCsCl

tPresent address:Departmentof ViralOncology,Roswell Park Memorial Institute, Buffalo,NY 14263.

of 1.36 g/ml. ScV-Lparticles had a density of

1.42, and ScV-M particles hadadensity of1.37

g/ml.

We havefingerprinted four of these dsRNA's: Sl(a) from strain T132B NK3; Sl(b) from strain T132B NK1;S14, and S733. Thesearecompared with Mfrom strain T158D SK and witha

pre-viously published

T,

fingerprint of S1 from strain T132B NK3, Sl(c). Uniformly

32P04-3-labeled dsRNA waspurified and digested with RNase

T,

aspreviously described (2, 3), and the products were separated by two-dimensional polyacrylamide gel electrophoresis (5, 6). The resolution of the two-dimensional gelsystem is

greater thanthat of the cellulose acetate-hom-ochromatography system previously used (2). We have adoptedanew numbering systemfor both S and M. Figures 1 and2depict only the resolvable large

T,

oligonucleotides,

and Table

1gives the pancreatic digestion products of those large

T,

oligonucleotides

ofM presentinone or moreS dsRNA's.

All four S dsRNA'sare clearly derived from M. Sl(a), -(b), and -(c) havea minimum of19

large

T,

oligonucleotides of M; S14 and S733 have 17. Each dsRNA has both of the 3'

T,

oligonucleotides

of M(oligonucleotides5and13

or54of S and43and13 or 44ofM). Therefore, allappear to be theresultof internal delection

ofM,followedbytandem

duplication

in the case of

S1

(2, 7). Threeof the fourS

dsRNA's

have

large

T, oligonucleotides

notpresent in M. Of

those novel

T,

oligonucleotides previously ob-served in S1 [2, shown in Sl(c)], Sl(b) has

preservedallthree(no. 3, 8, and 20) and added twomore(thetwo

T,

oligonucleotidesinspot4).

S1(a)

has lost

all

three.

S1(b)

haslostoneof the

21

T,

oligonucleotides

originally derivedfrom M

[no.44ofS1(a) or

Sl(c)].

S14 is

slightly

largerthan S3 (50 to 100 base

pairs), but it clearlyarose from M in an event very similar to that whichgeneratedS3. S14has

692

on November 10, 2019 by guest

http://jvi.asm.org/

NOTES 693

3 , s

5?

24 54

23-2 0

5.3

3 40

9 44 3

9>

g

M

exc

5,

24 C,o D5 2'

Kb "'-\46 t4 5

31022K~)

Q014

S()

89 K7j---

4h

13 S a

J5Q~-16 2

43\ 556

22:~

C

2

iC C

2'.

.%--n 34

,-' 2-; -3 -,- \2 2

I 3BBP /S|r,

-,11 I !,

435 5 _,

.-k2

./, Xc

44

7...

-

S1

(b)

2nd

1st --t

FIG. 1. Two-dimensional gel electrophoresis ofT, digests of32PO4.3-labeleddsRNA's.RefertoTable1for

thepancreaticanalysis oftheT, oligonucleotides. Thepositions ofthebromophenolblue(BB)andthexylene cyanol(XC)markerdyesareshownforeachgeL T1 oligonucleotides ofMthatarepresent inone or moreS dsRNA'sareshaded.(Sl(a)is Sl dsRNAfromstrainT132B NK3.Sl(b)isSl dsRNAfromstrain T132BNKI.

17 ofthe 21 large

T,

oligonucleotides from M

presentin Sl(c). It has both of the 3'T1 oligo-nucleotides of M. In contradistinction to Si, it hasnodetectable large

T,

oligonucleotides not

derivedfrom M.S733,ontheotherhand,hasat

least threelarge

T,

oligonucleotidesnotpresent

inM (no. 56, 62, and 65). S733 is similar in size

toS3,preservesboth 3' ends ofM, and contains

16 of the 21 large

T,

oligonucleotides derived from M by S3. S733 has also been observedto

form tandem duplicates, that is, some clonal

isolates of strain K733 produce an S RNA of

twice the usual sizewiththesame

T,

fingerprint.

S733hasoneadditionallarge

T,

oligonucleotide (no. 63) derived from M (no. 32) that is not present in any other S dsRNA so far finger-printed.

The existenceofthesame3'terminionM, S1,

S33, and S14 has been confirmed by 3' end

labelingandsequenceanalysis (V. Brennanand

J. Bruenn, unpublished data). The T1 finger-prints of in vivo labeled RNA have

demon-strated theheterogeneityof atleast one3'end

ofM;

T,

oligonucleotides 13 and 44 of M have

different mobilities but apparently have the

same pancreaticdigestion products (Fig. 1 and

2and Table1). This would be consistentwitha

change in the 3' terminal nucleoside, which is not detected in the pancreatic digestion

prod-ucts.Most of the S dsRNA's have the 3' terminal

T,

oligonucleotide with the mobilityof13, but

S733hasonewith themobility of 44(no. 54 of

S733). Sl(a) has both, like M. The other 3' terminus istentatively assignedtoT1 oligonucle-otide no. 5 of S (no. 43 of M). Thepancreatic products andmobilityof thisoligonucleotideare

26

)5

.-,.2 5 2' -.6 ,

2C -.,-) -

1-"I 1 4

2 .- "-/

A

'l-'-/

r-1; 4

,--/ -7,

on November 10, 2019 by guest

http://jvi.asm.org/

[image:2.507.60.452.71.417.2]

694 NOTES

.

0

...-.

S

0

,.

_.

0 lb

P

FIG. 2. T1fingerprints ofS733and S14arecomparedwiththesameMT,fingerprintsasthose inFig. 1. Notation isasexplainedinFig.1.

consistentwiththe3'sequenceanalysis

(Bren-nan and Bruenn, unpublished data). S733 ap-pearstohavetwopossible3'

T,

oligonucleotides

atthis end (no. 5 and 61) with closely related

sequences.

Among the fourDIRNAsanalyzed, ofwhich threeareindependentlyderived,there isa core

ofsome12large T1 oligonucleotidesfrom Mthat are present ineach S (Table 1),includingboth

3'

T,

oligonucleotides of M. Each of the S dsRNA's thus appears to be the result ofthe internal deletionofM,as suggested for S3 (7).

Furthermore, thisdeletion seemsto occurin a

roughly similar place in eachcase.The

complex-ity of theoriginalS dsRNAderived from Mcan, in eachcase,beapproximatedby the expression

C=

N/(0.24)(1-0.24)n-1,

whereNis thenumber

of T1 oligonucleotides of sizen orlarger and 0.24

isthefrequency of G residues (2). This figure is

about0.65 to 0.80 kb in eachcase,whichagrees well with the size ofS3,S14,andS733(0.7to 0.8

kb). Thecoresequencecorrespondstoabout 0.5 to 0.6 kb. Therealso appearsto beasize

con-straintin the formation of DIRNAs,since there

areonlythe two size classes: that of S3 and that

ofSi, the latter beingatandem duplicationof

the former. At leastsomeanimal virus DI RNAs

arealsogeneratedbydeletion(15).

Asimple modelcanaccountfor both deletion

andtandem duplication. ScV RNA replication appearstobe similar to that of reovirus (9, 12, 18; J. D. Welschand M.J. Leibowitz, personal communication). Ifa transcriptase copying M

dissociated from its template, reassociated at

some internal sequence, and then completed synthesis, a deletionwould result. If the

tran-scriptase completed the synthesis of the viral

plus strand and dissociated from its template

-.0

--. 1-

on November 10, 2019 by guest

http://jvi.asm.org/

[image:3.507.64.458.74.434.2]

TABLE 1. T,oligonucleotides of M and defective interfering dsRNA 's S1(C) NR + + S14 S733 + _ + +

+ + + 43

+ + + 9

NR + - 6

+ + + 55

+ + - 30

+ + + 27

+ + + 52

+ + - 13

+ - - 45

+ - + 37

+ + + 33

+ - + 34

+ + + (53)

+ + + 51

+ + _ 40

+ + + 25

NR - + (54)

+ + + 31

- - + 44

+ + + 30

- - 56

-- - 61

-- - 62

-- - 63 32

- - 65

-Pancreatic RNAse products (AG)(AU)5(AC)5C2U7 (A2C)(AG)(AU)2(AC)3C4U3 (AG)(AU)4(AC)2C2U4 (A3C)(A2U)(AG)(AU)(AC)2GC4U6 (AU)2-3C3U810 (A2U)(A2C)2(AG)(AU)(AC)2C (A2U)(AG)(AC)C6U6 (A2U)(A2C)(AG)(AU)2(AC)CU3 (A2U)(AC)2GU2C (AU)2(AC)2GC2U3 (A2U)(AG)(AU)2(AC)CU3 (A3U)(AU)(AC)GU4 (A3C)(AU)(AC)2C (A2G)(AU)(AC)2C3 (AC)4GC3U2 (AU)3(AC)2GC5U2 (A2U)(AU)2(AC)GC4U2 (AU)3(AC)GC2U5 (AU)2GC3U8 (AU)2GC2U4 (A2G)(A2C)(AC)3C3 (A2U)(A2C)(AU)2(AC)GCU3 (A2U)(A2C)2(AG)(AU)(AC)2C (A2U)(AG)(AU)(AC)3CU3 (A3C)(AU)(AC)2C (A3C)(A2U)(AG)C2U3 (A2C)(AG)(AC)3C (AU)3(AC)C3U8 (A3C)(A2C)(AC)3G (AU)4(AC)GC3U3 (A3C)(A2C)(AG)(AC)2C

aT1 oligonucleotides are numbered asinFig. 1 and2. Symbols: +, present; -, absent;NR,notresolved;

( ), pancreatic RNAse productsnottested.Sl(c) is Si dsRNA analyzedon10April1977from T132BNK3 (2).

Sl(b) is S1 dsRNA analyzedon1June1978from T132B NK1.Sl(a)isSi dsRNAanalyzedon1December 1978

from strain T132B NK3.S14isS14 dsRNA from strain L014. S733 is S733 dsRNA from strainK733,andMis M dsRNA fromstrainT158D SK.

withoutreleasing the plus strand, the

transcrip-tasecould then associate withasecond molecule

of S present in the particle (or with the same

molecule ofS)and continueplus-strand synthe-sis, making atandemly duplicated plus strand.

In both deletion and duplication, a replicase

wouldsynthesize the virus minus strand. Since tandem duplication of L-, M-, and Sl-sized DI RNAsis notobserved,thepresenceoftwo mol-ecules of a dsRNA in one particle may be a

prerequisite for tandem duplication. Two mole-cules of L (4.5 kb)-, M (1.8 kb)-, orSl-sized DI

RNAs (1.5 kb) may not be packaged easily, whereas two molecules of S3-sized DI RNAs (0.75 kb) may.

Theappearance of novel T1 oligonucleotides

in the S dsRNA's and the disappearance of M

T,

oligonucleotides originallypresentare

indic-ative of the accumulation ofsequencevariants ofthe DI RNAs. All of the DI RNAsexamined (including S14)areheterogeneousatthe 3' ends (Brennan and Bruenn, unpublished data), in

supportof thisproposition. Similar changes have beenobserved in DI RNAs of vesicular

stoma-titis virus(13). Constantsequencechange is only

tobeexpected in the absence of selective

pres-sure, and the S RNAs do not code for any

proteinsnecessaryformaintenance of the virus.

We thank S. Kahn andJ. Campagna forexcellenttechnical assistance,Loren Fieldand Herb Thomas for electron

micros-copy, and G. Fink, M. Vodkin, and R. Wickner foryeast

strains. This research wassupported byNIH grant 5 RO 1

GM22200toJ. Bruenn.

LITERATURE CTED

1. Bevan,E.A.,and M.Makower. 1963. Thephysiological basis of the killercharacterin yeast,p.202-203.In S.J. Geerts (ed.), Geneticstoday,vol. 1. Macmillan,The Hague.

2. Bruenn, J., and W. Kane. 1978. Relatedness of the double-stranded RNAs present in yeast virus-like

par-ticles.J. Virol. 26:762-772.

3. Bruenn, J.,and B. Keitz. 1976. The 5' ends of yeast killer factor RNAsarepppGp. Nucleic Acids Res. 3: 2427-2436.

4. Buck,K.W.,P.Lhoas,and B. K. Street. 1973. Virus S1(a) 1 2 5 6 7 9 10 11 12 13 14 15 16 17 18 19 21 43 44 45 54 55 Sl(b) 3 4 + + 8 + + + + + +

on November 10, 2019 by guest

http://jvi.asm.org/

[image:4.507.53.460.83.385.2]

particles in yeast. Biochem. Soc. Trans. 1:1141-1142. 5. DeWachter,R., and W. Fiers. 1972.Preparative two-dimensional polyacrylamide gel electrophoresisof 32P-labeledRNA. Anal. Biochem.49:184-197.

6. Faller,D.V., andN.Hopkins.1977.RNase TI-resistant oligonucleotides of B-tropic murine leukemia virus from BALB/cand five of itsNB-tropicderivatives. J. Virol. 23:188-195.

7.Fried, H. M., andG. Fink. 9978.Electronmicroscopic heteroduplex analysisof"killer"double-strandedRNA speciesfromyeast. Proc. Natl. Acad. Sci. U.S.A. 75: 4224-4228.

8.Harris, M. S. 1978. Virus-like particles and double-stranded RNA from killer and non-killer strains of Saccharomycescerevisiae. Microbios 21:161-176. 9. Hastie, N., V. Brennan, and J. Bruenn. 1978. No

homology between double-stranded RNA and nuclear DNA ofyeast. J.Virol.28:1002-1005.

10. Herring, A. J.,and E. A.Bevan.1974.Virus-like par-ticlesassociated with the double-stranded RNAspecies foundinkillerand sensitive strains oftheyeast Saccha-romycescerevisiae.J. Gen. Virol.22:387-394. 11. Herring, A.J., and E. A. Bevan.1975.Double stranded

RNA containingparticles from the yeast Saccharomy-ces cerevisiae and their relationship to the "killer" character,p. 149-154. InS. Puiseux-Dao (ed.), Molec-ularbiology of nuclear cytoplasmic relationships. Elsev-ierScientific, Amsterdam.

12. Herring,A. J. and E. A. Bevan.1977.Yeast virus-like particles possess a capsid-associated single-stranded RNApolymerase.Nature(London) 268:464-466.

13. Holland,J.J.,E.A.Grabau, C.L.Jones, and B.L. Semler. 1979. Evolution ofmultiplegenomemutations during long-term persistent infection byvesicular

sto-matitis virus.Cell 16:495-504.

14. Hopper, J. E.,K. A.Bostian, L.B.Rowe, andD. J. Tipper. 1977.Translationof theL-speciesdsRNA ge-nome of thekiller-associatedvirus-likeparticlesof Sac-charomycescerevisiae. J. Biol. Chem.252:9010-9017. 15. Nomoto, A., A. Jacobson,Y.F. Lee, J.Dunn, andE. Wimmer. 1979. Defective interfering particles ofpolio virus: mapping of the deletion and evidence that the deletions in the genomesDI(1),(2)and (3) are located in the same region. J. Mol. Biol.128:179-196. 16. Palfree,G.E.,and H.Bussey.1979.Yeast killer toxin:

purification andcharacterization of the protein toxin from Saccharomyces cerevisiae. Eur. J. Biochem.93: 487-493.

17. Saksena,K.N.,and P. A. Lemke. 1978.Virusesinfungi, p. 103-143.H. Frankel-Conratand R. R. Wagner (ed.), Comprehensive virology, vol. 12. Plenum Press, New York.

18. Silverstein,S. C., J. K.Christman,and G. Acs.1976. The reovirus replicative cycle.Ann. Rev.Biochem.45: 375-408.

19. Somers,J. M. 1973. Isolation of suppressive sensitive mutantsfrom killer and neutralstrains of Saccharo-mycescerevisiae. Genetics 74:571-579.

20. Wickner, R. B. 1978. Twenty-six chromosomal genes needed to maintain the killer double-stranded RNA plasmidofSaccharomyces cerevisiae. Genetics 88:419-425.

on November 10, 2019 by guest

http://jvi.asm.org/