Identification of four complementary RNA species in Akabane virus-infected cells.

(1)

0022-538X/83/090452-11$02.OO/O

Identification

of Four

Complementary

RNA

Species

in

Akabane

Virus-Infected

Cells

ASITK.PATTNAIKANDG. ABRAHAM*

School of Science,Griffith University,Nathan, 4111,Australia

Received 8 March1983/Accepted 26May1983

The analysis ofRNA extracted from purified Akabane virusdemonstratedthe

presence of three size classes of single-stranded RNAs with sedimentation coefficients of 31S (large, L), 26S (medium, M), and 13S (small, S). Molecular

weights of theseRNAspecies were estimated to be 2.15 x

106,

1.5 x

106,

and 0.48

x 106 for the L, M, and S RNAs, respectively. Hybridization analysis involving

viral genomic RNA and RNA fromvirus-infectedcellsresulted in the

identifica-tion of four virus-specific cRNA species in infected cells. These cRNAs were

found tobenonpolyadenylated by their inabilityto bind to

oligodeoxythymidyl-ate-cellulose. KineticanalysisofcRNAsynthesis in infected cells at various times

postinfection suggested that cRNA synthesis could be detected as early as 2 h

postinfection and that maximal synthesis occurred at 4 to 6 hpostinfection. The

RNAssynthesized in infectedcells could be partially resolved by sucrose density

gradient centrifugation. TheRNA fraction thatcosedimented with the S segment

of viral genomic RNA yielded two duplex RNA species when hybridized with

viral genomicRNA,suggestingthepresence of two small cRNA species. Specific

hybridization with individual viralgenomic RNAs confirmed that two species of

cRNA are coded by the S RNA segment. Analysis of cRNA synthesis in the

presenceof the protein synthesis inhibitors cycloheximideandpuromycin

indicat-ed that cycloheximide completely inhibited virus-specific RNA synthesis early

and late in infection, whereas a very low level of synthesis occurred in the

presence of puromycin. The inhibitory effects of these drugs were found to be reversible when the drugs were washed from the cells. It is concluded that continuedprotein synthesis is required forcRNA synthesis to proceed in Akabane

virus-infected cells.

Akabane virus is a member of the genus

Bunyavirus and family Bunyaviridae (7). All

bunyaviruses sofar characterized contain three

unique segments ofsingle-stranded RNA (large

[LI, medium

[Ml,

and small [S]) of negative

polaritywithatotalmolecularweightof

approx-imately 4 x 106 to 6 x 106 (3). These RNAs

reside within the virion as three helical and

circular nucleocapsids, being complexed with

multiple copies of a nucleocapsid protein (N)

and afewcopies ofalargeprotein(L), believed

tobe atranscriptase component(5). Two

exter-nalglycoproteins(G1 andG2)form thespikeson

the viralenvelope;G1 protein containsasitefor

binding to cellular receptors and also functions

astheviralhemagglutinin,whereasthe function

ofG2 still remains to be demonstrated (15). A recentreport(25)alsosuggests the presenceofa

fifth virion-associated polypeptide of molecular

weight16,000 (16K). In addition to these

struc-tural proteins, the viral genome also codes for

several nonstructural proteins (12).

Recombina-tion studies (4, 13,14)showthat the small viral

RNA contains coding information for the N

protein, whereas the MRNA segmentcontains

theinformation for both G1 and G2.Itis

there-fore inferred that theLproteinis codedbythe L

RNAsegment.

Although more than five virus-specific

pro-teins have been identified, the mechanism of

bunyavirus gene expression in infected cells

remains to be explained. It is not known how many mRNAs mediate the expression of three

genome segments into more than five

virus-specificpolypeptides. Using duplexRNA

analy-ses, Cash et al. (9) identified the presence of

three cRNA species in snowshoe hare virus-infected cells but could only detect messenger activity in the smallest of the species. By

frac-tionatingRNAfrom Uukuniemivirus(amember

of the familyBunyaviridae)-infected cells in

su-crose gradients, Ulmanen et al. (26) detected

fourvirus-specificRNAs, twoofwhich directed

the synthesis of virus-specific polypeptides in

vitro. The mRNA species that cosedimented

withthe M RNA segmentdirected thesynthesis

452

on November 10, 2019 by guest

http://jvi.asm.org/

(2)

ofa11OK-molecular-weight protein (foundtobe theprecursor toGl andG2).Inaddition, a 12S

RNA species directed the synthesis of the N

protein anda30K nonstructuralprotein.

This communicationreportstheidentification of four nonpolyadenylated cRNA species

syn-thesized in Akabane virus-infected cells and the

sensitivity of theirsynthesistotheprotein

syn-thesisinhibitors cycloheximide andpuromycin.

MATERIALS ANDMETHODS

Cellsand virus stock. Akabane virus(strain R7947)

wasobtained from the Queensland Institute of Medical

Research, Brisbane, Australia, and was cloned by

plaque isolation twice before use. Vero cells were

grown in minimal essential medium containing 5%

fetalcalfserum. Virus stocks wereusually prepared by infecting confluent monolayers of Vero cells in roller bottles at amultiplicity of infection (MOI) of

0.01 PFUpercell. High-titer stockswereprepared by directly pelleting virus from the clarified culture fluids andresuspending it in phosphate-buffered saline con-taining 0.2% bovine serum albumin and 0.001% DEAE-dextran.

Growth ofvfrusandpreparationof vRNA. Subcon-fluentmonolayercultures of Vero cells in roller bottles

wereinfected with Akabane virusatanMOI of 0.001

PFU percell. After 1 hof virus adsorption at room

temperature,inoculawereremoved andreplaced with virus growth medium and incubatedat37°C for about 35 h. Virus wasconcentrated from clarified

superna-tant by polyethylene glycol precipitation (22) or by direct pelleting and purification by centrifugation in potassium tartrate-glycerol gradients (22).After repel-leting the virus, viral genomic RNA (vRNA) was extracted with sodium dodecyl sulfate (SDS) and phenol (1).

32P-labeled vRNAwasprepared similarlyfrom virus

growninphosphate-free mediumcontaining1 mCi of

32Pi per ml. The vRNA segments were resolved by

electrophoresis ina3%polyacrylamide gel (120Vfor 16 h) containing 6 M urea and90 mM Tris-EDTA-boratebuffer, pH8.3. The individual RNA segments

were detected by autoradiography, excised, eluted from the gel (in 10 mM Tris-hydrochloride, pH 7.5, 1 mMEDTA,0.5%SDS,0.5 MNaCI),and precipitat-ed with ethanol.

Preparation of RNA from infected cells. Confluent monolayersof Vero cellswereinfected with Akabane virusatausualMOIofapproximately7. After 1 h of virusadsorption at room temperature, inocula were

removed and the monolayerwaswashed with warm phosphate-buffered saline and incubated in the pres-ence of Hanks balanced salt solution supplemented with15 mMHEPES (N-2-hydroxyethylpiperazine-N'-2-ethanesulfonicacid) adjustedtopH7.4. Actinomy-cin D (2 sLg/ml), [5-3H]uridine (25 to 50 pCi/ml), cycloheximide (100 pLg/ml),andpuromycin (100iLg/ml) wereaddedatappropriate times postinfection accord-ingtotheprotocolofeachexperiment. After comple-tion of the labeling period, cell monolayers were

washedwith cold phosphate-buffered saline and dis-solved in buffercontaining 0.5%SDS, and the RNA

wasextractedwithphenolandprecipitatedwith

etha-nol. Polyadenylated and nonpolyadenylated RNAs were separated by oligodeoxythymidylate [oligo(dT)]-cellulosechromatography (1).

RNA-RNA hybridization and electrophoresis. La-beled cRNAsfrom infected cells were hybridized with anexcessof unlabeled vRNA (17). The hybrid mole-cules thus formed were digested with single strand-specific nuclease Si (1,000 U/ml) at 37°C for 2 h and the residual double-stranded RNAs (dsRNAs) recov-eredby ethanol precipitation were analyzed by elec-trophoresis on a 4% polyacrylamide gel at 40V for 15 to 20 h (17). Thegel was processed for fluorography (8), and the radioactive bands were detected by expos-ing the dried gel to X-ray film at -70°C (19).

To determine the excess of unlabeled vRNA re-quired by hybridization, fixed amounts of [3H]cRNA from the virus-infected cells were hybridized with increasing amounts of unlabeled vRNA. The trichloro-aceticacid-precipitable radioactivity in portions of the hybridized RNAs before and after S1 nuclease diges-tion was determined (Table 1). This analysis indicated that saturation levels of hybridization were achieved by usingatleast6FxgofpartiallypurifiedvRNA.

Sucrosedensity gradient analysis.Labeled RNAs for sucrose density gradient analysis were dissolved in water, denatured by heating at 100°C for 1 min,

immediately chilled, and adjusted to 10 mM Tris-hydrochloride (pH 7.5)-100mMNaCl,1 mM EDTA-0.1% SDS. RNA samples were then analyzed by sedimentation through 15 to 30%o (wt/vol) sucrose density gradients (prepared in the same buffer) in a BeckmanSW41 rotor (25,000 rpm, 16 h,20°C). Frac-tions (0.4 ml) werecollected from the bottom of the gradient, and trichloroacetic acid-precipitable radioac-tivity in portions of each fraction was determined. Where appropriate, RNAs were recovered from the gradient by ethanol precipitation of the pooled peak fractions.

Materials.Cycloheximide, puromycin, and nuclease

Siwereobtained fromSigmaChemicalCo., oligo(dT)-cellulose from Collaborative Research Inc., and

[5-3H]-uridine from the Radiochemical Centre, Amer-sham, England. ActinomycinDwas agiftfrom Merck

Sharp & Dohme, Australia. [3H]uridine-labeled

[image:2.492.257.451.529.655.2]

dsRNAfrom Wallal virus was agift from P. J. Walker, Queensland Institute of Medical Research, Brisbane, Australia.

TABLE 1. Determination of thesaturatinglevelof unlabeled vRNArequiredforhybridizationwith

[3H]cRNA

Amt(.g)of unlabeled vRNA

hybridizedwith[3HJcRNA %Radioactivity resistantto from 2 x106cells Si nucleasedigestiona

0 32

1 41

2 50

4 67

6 82

8 83

10 82

aDetermined by precipitating aportion of the

hy-bridizedRNAsbefore and afterS1 nucleasedigestion.

http://jvi.asm.org/

(3)

454

RESULTS

Sucrose density gradient analysis of vRNA.

[3H]uridine-labeled RNA was extracted from

purified virus, and the denatured productswere

analyzed by sedimentation inanSDS-containing sucrose density gradient. A profile of the

acid-precipitable radioactivity in the gradient frac-tions revealed three discrete peaks (Fig. la) representing RNA species with sedimentation coefficients of 31, 26, and 13 for theL, M,and S RNAs, respectively. Molecular weights of2.15

x 106, 1.5 x 106 and 0.48 x 106 for these three

RNA species were calculated. The single-stranded nature of the RNA was shownby its sensitivity to prior digestion with pancreatic RNase (Fig. la). These data confirmed that Akabane virus RNA shows the characteristic properties of RNA fromamember ofthe

Bunya-viridae.

Identification of cRNA species in infected cells. RNA species synthesized in mock-infected or

Akabane virus-infected cells were labeled with [3H]uridine from 2to 8 h postinfection (p.i.) in thepresenceofactinomycinD. Total RNAwas

extracted andanalyzed by density centrifugation in SDS-containing sucrose density gradients

(Fig. lb). Incorporation of radioactivity into mock-infected cell RNA was considerably less

than into infected cellRNA andwas

concentrat-edatthetopof the gradient. Two of the major peaks of radioactivity in the infected cell RNA sample corresponded with the sedimentation positions of the M and S vRNA segments(Fig. la), whereas a shoulder of radioactive material

(fractions 6to10) correspondedtothe sedimen-tationposition of the L vRNA segment.

To demonstratethevirus-specific and

comple-mentary nature ofthe RNA synthesis, we

hy-bridized [3H]uridine-labeled RNAs from either mock- or virus-infected cells with unlabeled

vRNA.Preliminary experiments had shown that saturation levels ofhybridizationwereachieved

instandardizedexperiments using approximate-ly 7 pug of crude vRNA extracted from partialapproximate-ly purified virus. After removal of unhybridized

[image:3.492.50.243.61.562.2]

BOTTOM FRACTION NUMBER TOP

FIG. 1. (a)Sucrosedensitygradient centrifugation ofRNA frompurifiedAkabanevirus.3H-labeledRNA from purified virus was denatured and sedimented

througha15to30% SDS-sucrosedensitygradient as

described inthetext. Fractions(0.4 ml)were

collect-ed, andtrichloroacetic acid-precipitable radioactivity in eachfraction wasdetermined. Symbols:.0, RNA

from purified virus; 0, RNA from purified virus

digested withpancreatic RNase(20 ,ug/ml in 10 mM Tris-hydrochloride [pH 7.5],0.3 MNaClat37°Cfor 30 min). Arrows indicate the sedimentation position of 3H-labeled 35S poliovirus RNA and 28S and 18S

rRNAs. (b) Analysis of RNA from Akabane

virus-infectedormock-infectedcells. The virus-infectedor

mock-infected cells were labeled with [3H]uridine from 2to8 hp.i.in thepresenceof 2,ugofactinomycin Dperml(addedat0 hp.i.). Total labeled RNAwas

extracted from each cultureandanalyzed by centrifu-gation on15 to 30% SDS-sucrose densitygradients. Fractions(0.4 ml)werecollected and the

trichloroace-ticacid-precipitable radioactivity inportions of each fractionwasdetermined. Symbols: 0,RNA from the

virus-infected cells; 0, RNA from mock-infected

cells. w

0

z

ar

r-J. VIROL.

http://jvi.asm.org/

(4)

A B C D E F G b

<1

_n _lU

-*

<d2

A B C

nr-,

Da

2

<3 < 4

m-.'fl 10,I

o-u.no

MO sw-O "

urn

3 . J4

[image:4.492.68.433.76.388.2]

4 *ow

FIG. 2. Polyacrylamide gel electrophoresis of virus-specific dsRNA hybrids. (a) Total RNA from mock-infected (lane A)orvirus-infected (lanesB andC) cellslabeled with [3H]uridine at 2 to8 hp.i. wereeither hybridized with unlabeled vRNAs(lanes A and B) or self-annealed (lane C),digestedwith nuclease Si, and

analyzedbyelectrophoresison a4%polyacrylamide gelasdescribedin thetext.Lane Drepresents the dsRNAs produced byhybridizationof[3H]uridine-labeled vRNAswith unlabeled RNAsfrom virus-infected cellsat8 h

p.i. Afterelectrophoresis, the radioactive bands were detected byfluorography. Arrowheads with arbitrary

numbers represent thefourvirus-specificdsRNAhybrids.(b)Infectedormock-infected cultureswereincubated inpresenceof actinomycinD at0 hp.i. and labeled with[3H]uridinefor 2-hperiodsatvarioustimesstartingat0 hp.i.The RNAswerethenhybridizedwithunlabeled vRNAs, and theduplexesformedwereanalyzedasabove. Duplexes in lanes B to G werederived from infected cells labeledat0,2, 4, 6,8,and 10 hp.i., respectively. Lane Acorrespondsto amock-infected culture labeled similarly.

RNAs by digestion with Si nuclease, the

dou-ble-strandedhybridmolecules were resolvedby

polyacrylamide gel electrophoresis (Fig. 2a).

Four radioactive bands representing four

dsRNA hybrids (arbitrarily numbered 1 to 4) were detected in extracts from virus-infected

cells (lane B) but not from mock-infected cells

(lane A). When a portion of3H-labeled RNA

from infected cells was self-annealed, digested

with nuclease Si, and analyzed similarly, the

samefourhybridRNAs weredetected(Fig. 2a,

laneC). This result indicatedthat RNA of both

polarities was synthesized in infectedcells

dur-ing the period 2 to 8 h p.i. To confirm that

hybridization to vRNA was occurring in these

experiments, thereverse procedureof

hybridiz-ing[3H]uridine-labeledvRNAtounlabeledRNA

from infected cells was done (Fig. 2a, lane D).

FourhybridRNAswith thesame

electrophoret-icmobilitiesasseenin lane B were detected. All

fourcRNAspeciesininfected cellswere shown

to be cytoplasmic rather than nuclear in their

location (data notshown).

Codingorigin of cRNA species. The results in

Fig. 2demonstrated the presence of four cRNA species in the virus-infected cells. To identify

thecoding origins ofthese cRNAs, vRNA

uni-formly labeled with 32p was prepared, and the

three single-stranded RNA segments were

re-solvedbypolyacrylamidegelelectrophoresis(as

VOL.47,1983

v0 -* 1

W*

S.

lw on November 10, 2019 by guest

http://jvi.asm.org/

(5)

456 PATTNAIK AND ABRAHAM

A

B

C

D

4-

U-FIG. 3. Coding origin of the induced (

cies.Uniformlylabeled (32P]vRNA(totaloi

segments)washybridizedwithan excesso cRNAs from virus-infected cells and ar

electrophoresis as in Fig. 2 and detecte radiography. Hybrids were produced [32P]vRNA (lane A), L RNA segment (l RNAsegment(lane C),and S RNAsegmel

[3H]uridinewasadded for 2-h periodsatvarious times from 0 to 10 h p.i. Labeled RNAs were

extracted,

hybridized

withan excess of

vRNA,

and analyzed (Fig. 2b)asdescribedabove (Fig.

< 2 2a). Two of the virus-specific RNAs were

de-tected as early as 2 h p.i. (lane B) by this

procedure. By 4 h p.i. all four duplex RNAs

were readily detected, and RNA synthesis

reached a peak at 4 to 6 h p.i. After 8 h, the synthesis of the largesthybrid species fell below detectable levels, whereas the synthesis of the other three hybrids continued but in reduced

amounts.These dataindicate thattemporal

con-trol of the synthesis of cRNA species of Aka-bane virus occurs during infection. In similar

experimental conditions, the time of maximal

< 3 virus-specific protein synthesis coincided with < 4 the time ofmaximal cRNAsynthesis (data not

shown).

The molecular weights of the RNA hybrid speciesweredeterminedby comparison with the

known values of the dsRNAofWallal orbivirus (16). Table 2 shows the molecular weights of thesehybrids, their corresponding single-strand-ed components, and the calculated coding

po-tential, after translation, of possible proteins. Capacity of cRNAs to bind to oligo(dT)-cellu-lose. Todetermine whether the cRNAspecies in Akabane virus-infected cells could beseparated intopopulations of polyadenylated or

nonpolya-denylated molecules, fractionationwas

attempt-ed by oligo(dT)-cellulose chromatography. [3H]uridine-labeled RNA was extracted from

cRNA spe- infected cells at either early or late times after

rindividual infection, and the polyadenylated RNA species

funlabeled were separated from the nonpolyadenylated

dalyzed

by RNA. When actinomycin D (2 ,ug/ml) was

in-withd

total

cluded

during

the

labeling

period

(4 to8 h

p.i.),

lane B), M less than 4% oftheacid-precipitable radioactiv-nt(laneD). ity was found to bind specifically to and elute

from oligo(dT)-cellulose. In contrast, a similar

described above). The recovered individual RNA segments were each hybridized with an

excess ofunlabeled cRNA from infected cells,

digested with Si nuclease, and analyzed by polyacrylamide gel electrophoresis. The unfrac-tionated[32P]vRNAuponhybridization produced the fourtypical dsRNA hybrids (Fig. 3, lane A). The L RNA upon hybridization produced

hy-brid1, whereas the M RNAgaverisetohybrid 2 (lanes B and C,respectively). Incontrast, the S RNAyielded bothhybrid 3 and hybrid4(lane D)

uponhybridization.

Kinetics of cRNA synthesis. To follow the kinetics of synthesis of cRNA species after infection, we infected parallel cultures at an

MOI of 7 PFU per cell and incubated in the

[image:5.492.74.220.78.405.2]

presence of 2 ,ug of actinomycin D per ml.

TABLE 2. Molecular weights of Akabane virus-specificcRNAs andtheir coding potentials

Molwt(x106) Approximate HybridRNA Corresponding coding

species dsRNA0 single-stranded potential

RNA(x0)

1 3.8 1.9 210

2 2.76 1.38 153

3 0.25 0.125 14

4 0.21 0.105 12

aMolecularweights of the dsRNAs were calculated

by using the double-stranded RNAs of Wallal orbi-virus (16) as molecular weight standards.

bThecoding potential of each of the cRNAspecies wascalculated as one-ninth of their respective molecu-larweights.

http://jvi.asm.org/

[image:5.492.257.447.539.622.2]

(6)

fractionation of host cell RNA labeled in the

absenceofactinomycinshowed that 35% of the

incorporated radioactivity was in a polyadeny-lated form.

Evidence that virusspecificcRNAs were pre-sent in the fraction that failed to bind to

oli-go(dT)-cellulose was provided by hybridization

experiments. Polyadenylated and

nonpolyaden-ylated RNAs from infected cells labeled either

early (0to 4 h p.i.)or late (4to 8 h p.i.) were

isolated and hybridized withan excess ofvRNA

and analyzed as described above. The

virus-specific dsRNA hybrids could only be seen in

the lanescontaining products derived from

non-polyadenylated RNAs (Fig. 4, lanes A and C).

HybridRNA 1 species couldbe detected early

afterinfection by longer exposure of the gelto

X-ray film (lane A), but was clearly present

when labelingwasdone atthe timeof maximal

RNA synthesis (lane C). The broad band of

radioactivityseeninthelowerpartof the gels is

aresidual hostcomponent asactinomycinD was

not included in this experiment. Preliminary

experiments to test the sensitivity of this

fluo-rography method showed that if virus-specific

radioactivity in the polyadenylated samples had

exceeded 2% of that in the nonpolyadenylated

samples then discrete bands would have been

visible in lanes B and D (Fig. 4). Additional

experiments were done to try to detect

virus-specific polyadenylated RNA species in

Aka-bane virus-infected cells by analyses in sucrose

density gradients. When the small amount of

polyadentylated

RNA thatcould be isolatedwas analyzed asdescribed inFig. lb,noradioactive

peaks corresponding to viral RNAs could be

detected(datanotshown). Thus, by either

meth-od ofdetection, no virus-specific

polyadenylat-ed RNA species could be found in Akabane virus-infected cells.

Virus-specificRNAsynthesis in the presence of

protein synthesis inhibitors. Results of

self-an-nealingof RNAs from infected cells(Fig.2a) and thepartialresistance of suchintracellularRNAs

to digestion with pancreatic RNase (data not shown) indicated that RNAs of both polarities

werebeing synthesizedeven atearlytimes after

infection. Ithasbeen well established (2, 20, 23)

with othernegative-strand viruses that inhibitors

ofprotein synthesis prevent the translation of

mRNA synthesized early in infection and so

restrict viralRNA synthesis to that ofonly the

primary type. In such conditions, only cRNAs

thatcorrespond to viral mRNAsare produced.

Theclassical protein synthesis inhibitors

cyclo-heximide and puromycin were used to treat

Akabane virus-infected cells from the time of

infection.

[3H]uridine

wasadded at 2 h p.i. for a

further 4 h. After extraction and denaturation,

the labeled RNAs were analyzed by sucrose

A

B

C

.

'-D

9

M

t

r---<3

[image:6.492.282.422.72.376.2]

4

FIG. 4. Electrophoretic analysis of dsRNAs formed by hybridization ofpolyadenylated or non-polyadenylated RNAs from the virus-infected cells. [3H]uridine-labeledRNAfrom the virus-infectedcells

at0 to 4 h (lanes A and B) or4to8 h(lanes C and D) werefractionated into polyadenylated and nonpolya-denylated RNAs by oligo(dT)-cellulose chromatogra-phy and hybridized to vRNA and analyzed as de-scribed inthelegend toFig. 2. Lanes A and C contain

duplexes derived from nonpolyadenylated RNAs, whereas lanes B and D contain duplexesfrom poly-adenylated fractions.

density gradientcentrifugation (Fig. 5). The

gra-dients containing mock-infected cell RNAs or

infected cell RNAs(in the absence of

cyclohexi-mideorpuromycin) produced profiles similarto

that seen inFig. lb.However,the RNAprofiles from cultures treated with either of the drugs

failed to reveal any radioactive peaks

corre-spondingtothesedimentationpositions ofthe L

andMvRNAsegments.Minorpeaks of

radioac-tivitywere seenin thesegradients,

correspond-ingtothesedimentationposition ofthe S vRNA

segment, which was also the region to which

some labeled host RNAs sedimented (see the

mock-infected profile, Fig. 5). These results

indicatedthat RNA synthesis in thepresenceof

http://jvi.asm.org/

(7)

0

10

5-v 5 lo is 20 25 30

BOTTOM FRACTION NUMBER TOP

FIG. 5. Sucrosedensitygradient analysisof RNA fromvirus-infectedormock-infectedcells in the

pres-ence of cycloheximide or puromycin. Infected or mock-infected cultureswereincubated inthepresence

ofactinomycinDand eithercycloheximide (100 ,ug/ml)

orpuromycin (100 ,ug/ml)from the time ofinfection. RNAslabeled with[3H]uridinefrom2to6hp.i.were extracted andanalyzedbycentrifugationon15to30% SDS-sucrose density gradients as described in the

legend to Fig. 1. RNAfrommock-infected cells (0)

and RNA from Akabane virus-infected cells either directly (0)orin thepresenceofcycloheximide (O)or

puromycin(U).

either of these inhibitors was below detectable

levels in thisanalysis. Even if the MOI in such experiments was increased to 50 PFU percell,

no RNA synthesis was detectable in cells in

which protein synthesis had been stopped by these inhibitors (see below).

To determine whether the radiolabeled RNA species inFig. 5 were virus specific, the RNAs

infourregions of the gradients (numbered I, II,

III, and IV) wererecoveredby ethanol

precipi-tation and hybridized with an excess of

unla-beled vRNA. When theresulting dsRNAs after S1 nuclease digestion were analyzed by

poly-acrylamide gel electrophoresis, virus-specific RNAscouldbedemonstrated only in fractions I

to III of thegradient containing RNA from the control infected cells (data not shown). No vi-rus-specific productsweredetected in gradients

containing RNA from either cycloheximide- or

puromycin-treatedcells (data not shown), which

confirmed that primary RNA transcription in

such treated cellswas belowdetectable levels.

Sincetheinhibitoryeffects ofpuromycin and

cycloheximide on protein synthesis are fully

reversible, itwasofinteresttodetermine

wheth-erRNAsynthesis intreatedinfected cells would

recommence if the drugs were removed. To

make thedetection ofanyprimary transcription

in such cellsmorelikely,anMOI of50PFUper

cell was used. Cycloheximide and puromycin

were added to four cultures at the time of

infection. At 1.75 h p.i., the drugs were

thor-oughly washed from two ofthecultures over a

period of10 min. All four cultures (together with

anuntreated control culture) were labeled with

[3H]uridine

from2 to 6 h p.i. The labeled RNA

from each culture was hybridized with vRNA

andanalyzed as described above (Fig. 6a). No

dsRNAs were seen when cycloheximide was

retained intheculture (lane B), butafaint band

corresponding to the fourth viral hybrid RNA

could be seenin the corresponding

puromycin-treated sample (lane D). When the inhibitors

werewashedfromthecultures, cRNA synthesis was restored as shown by the presence of the

typicalpatternof four dsRNAs in lanes C andE.

However,theinhibitory effect of cycloheximide

could not be completely reversed within the

labeling periodused asjudgedby acomparison

of the intensities oftheradioactive bands (lanes

A and C).

Since in the presence of cycloheximide and

puromycin primary transcription wasibelow

de-tectablelevels,anexperimentwasdonetoshow

theireffectsonsecondary(amplified)RNA

tran-scription when syntheticratesaremuchhigher.

Five cell cultures were infected (at the usual

MOIof7PFU percell) and incubated for4h to

allowRNA synthesisto reachitsmaximalrate.

Cycloheximide and puromycin were added to

two cultures each and then washed from two

culturesat5.75 hp.i.

[3H]uridine

wasaddedfor

a further 4 h before preparing and analyzing

samples as described forFig. 6a.

Results observed(Fig.6b) indicated that both

cycloheximide and puromycin had a dramatic

inhibitory effect on secondary RNA

transcrip-tion also(lanes B and D) and that thisinhibition

couldbe readilyreversed bywashingthedrugs

from the cellcultures(lanes C and E). However, theeffectofpuromycin onsecondary

transcrip-tion was not complete, as afaint band of

radio-activity correspondingto hybridRNA 4 canbe

seen in lane D, Fig. 6b. The sensitivity ofthe

fluorography procedure used couldjust detect

bands containing 2% of the radioactivity of

control bands(lane A),andsoanestimateofthe

effect ofpuromycin on RNA synthesis in this

experimentis that itisat least90% complete.

http://jvi.asm.org/

[image:7.492.48.239.55.331.2]

(8)

A

-e ~- ,

B C D E a

A B C _D _E _b

<1

_ <2 gEE

<3

_- 4

FIG. 6. Effect of cycloheximide and puromycin on RNA synthesis in virus-infected cells. (a) Cultures infected with Akabane virus at a multiplicity of infection of 50 PFU per cell were treated with either

cycloheximide(100,ug/ml)orpuromycin (100,ug/ml)at0hp.i.At 1.75 hp.i., drugsfromone setof cultureswere washed out and thecellswerelabeled with[3H]uridinefrom 2to6hp.i.The RNAswereextracted, hybridized

with unlabeled vRNAs, and analyzed by polyacrylamide gel electrophoresis. Duplexes were derivedfrom infected cells (lane A), in the presence ofcycloheximideorpuromycin (lanes BandD,respectively), orwhen cycloheximide and puromycin were washed out (lanes C and E, respectively). (b) Similaranalysis of the duplexes when cell cultures were infected at an MOI of7 PFU percell were treated withcycloheximide or

puromycinat4hp.i.,and thedrugswerewashedout at5.75 hp.i.RNAwaslabeledwith[3H]uridinefrom6 to 10 hp.i. Duplexeswerefrom infected cells(laneA) in the presence ofcycloheximideorpuromycin (lanesB andD,

respectively)orwhencycloheximide andpuromycinwerewashedout(lanesC and E,respectively).

DISCUSSION

Analysis ofthe RNA contained in Akabane

virus particles indicated that it was typical of

that found for other bunyaviruses in that the

genome RNA was single stranded and in three

segments, although the measured sizes of these

segments wereslightlysmaller thanthose found

for other members of the family (5). However,

Akabane virus was found to differ from the

generalpropertiesshownby negative-strand

vi-ruses in that no polyadenylated cRNA species

could be detected in infected cells, and

estab-lished secondary RNAtranscription was

sensi-tivetotheaction of inhibitors ofprotein

synthe-sis. In addition, four species of cRNA were

detected inAkabane virus-infectedcells, which

differs from thethreefoundpreviouslyfor

snow-shoe hare bunyavirus(9).

The synthesis of

[3H]uridine-labeled

cRNAin

Akabane virus-infected cells was followed by

hybridizationwith unlabeledvRNAandanalysis

of the 3H-labeled hybrid species so formed by

gelelectrophoresis. Thiswasnecessarybecause

the usual experimental procedures used with

negative-strand virusestoanalyze only primary RNAtranscripts resulted in undetectable levels of[3H]uridine incorporation. Kinetic studies of cRNAsynthesis atvarious times afterinfection

at an MOI of7 PFUper cell (Fig. 2b) showed

that detectablesynthesiswasestablishedas

ear-lyas 2h p.i. Atthis time, only RNA hybrids 2

and 4 were visible(Fig. 2b,lane B). Synthesisof cRNAs increased as the infection proceeded,

reaching amaximal rateat4 to 6 h p.i. At this

time, all four hybrid species representing the

fourcRNA species were detected readily (lane

D). The synthesis ofthe cRNAspecies leading

to the formation of hybrid 1 was the first to

decrease to below detectable levels (lane E),

whereas synthesis ofthe cRNA speciesleading

totheformation ofhybrid3persisted alongwith

gI-4

<1 <2

w

M

<3

< 4

http://jvi.asm.org/

[image:8.492.57.446.73.336.2]

(9)

460 PATTNAIK AND ABRAHAM

hybrids2and 4untilafter 10 h p.i. These results

show that although all four cRNA species

reached their time of maximalsynthesis together

(4 to 6hp.i.),there waslimited temporal control

overthe synthesis of individual cRNA species.

Toconfirm that the fourcRNA species being

labeled in these experiments truly represented

viral cRNA species, the alternative experiment

inwhich[3H]uridine-labeledvRNAwas

hybrid-ized with unlabeled infected cell RNAs was

done andanalyzed similarly. Againfourspecies

of double-stranded hybrid molecules that

showed the same electrophoretic mobilities as

those in theconverseexperiment(Fig. 2a, lanes

BandD) werefound. This result contrasted with

the results of Cash et al. (9) who found only

three such hybrid species in analyses of

snow-shoe harevirus-infected cells. Although the

sin-gle-strand-specific S1 nuclease was used

rou-tinely inourexperimentstodigest unhybridized

(single-stranded) molecules, the same four

hy-brid RNA species could be produced if the

RNases T1 or T2 were used instead for the

digestionstep(datanot shown).

MolecularweightsofthehybriddsRNAs and

the corresponding cRNAs were determined by

comparison with the genome RNAs of the

de-scribedorbivirus, Wallal (16). Thederived sizes

of the cRNAspecies (Table1)weresmallerthan

the values measured for the vRNAs, indicating

that none represented full-length transcripts of

thegenomeRNAs. Thecodingpotentialsof the

cRNAs(calculatedasone-ninthof the molecular

weight of the cRNA)werefoundtobe

approxi-mately210 x 103, 153 x

103,

or14x 103daltons

respectivelyfor the L, M,orStranscripts (Table

1). Either of the cRNAs represented in hybrid

species 1 and 2appear tohavesufficient coding

information for either the L protein (121 x 10

daltons)ortheG1 andG2proteins together(140

X 103 daltons, total). In contrast, the cRNAs presentinhybridspecies 3 and4do notappear

tobelargeenoughtocode for either the LorG1

plus G2 proteins. Within the approximations

used for calculating the coding potentials of

these cRNAs, it seems possible that one of

cRNA species present in either hybrid 3 or 4

codes for the N protein. This conclusion has

been supported by preliminary in vitro

transla-tion ofthe RNA species in region III, Fig. 5,

which hasresulted in thetranslation ofonly the

polypeptidecorrespondingtotheNprotein

(un-published results). HybridRNAspecies3and4

are similar in size and may have overlapping

sequences since hybridization shows that they

areboth derived from segment S ofviral RNA

(Fig. 3). Ifso,thisinformation is consistentwith

thereportofBishop and co-workers (6)that the S segments of bunyaviruses have sequences

that allow for the potential expression oftwo

gene products by using overlapping reading

frames. Additional nonstructural proteins found

in variousbunyavirus-infectedcells (12, 25, 26),

including Akabane virus(21),could conceivably

betranslated from the remainingcRNAspecies.

None of the RNAs synthesized in Akabane

virus-infected cells wasfound to be

polyadeny-lated, basedontheirinabilityto bindspecifically

tooligo(dT)-cellulose. The RNA synthesizedat

earlytimes after infection and presumably

con-taining the products of primary transcription

failed to bind to oligo(dT)-cellulose in amounts

sufficienttoanalyze asbeing virus specific (Fig.

4). In contrast, RNA from the fraction which

failedtobindtooligo(dT)-cellulose could readily

bedemonstratedto contain virus-specific

mole-cules. Similar results were obtained if RNA

synthesized atlater times (undoubtedly

includ-ingtheproductsof secondary transcription)was

used for the analysis (Fig. 4). If the MOI in

experiments of this type was increased to

ap-proximately 50 PFU per cell to increase the

probability of detecting primary RNA

tran-scripts, the results were unchanged (data not

shown). It could be argued that the level of

polyadenylated RNA synthesis was toolow to

detect in theseexperiments. However,thedata

in Fig. 4 combined with measurements ofthe

sensitivity of the fluorography detection

proce-dureindicate that the synthesis of virus-specific

nonpolyadenylated RNA was at least 30 times

higher than thatof polyadenylatedRNAin

Aka-bane virus-infected cells. Such a situation is

unique among the negative-strand viruses in

general, but is consistent with work reported previously for another bunyavirus, Uukuniemi

virus (26). However, it conflicts withtheresults

of Cash et al. (9), who were able to show that

both polyadenylated and nonpolyadenylated

RNAfractions from snowshoe hare

virus-infect-edcells could be translated in vitroto produce

thenucleocapsidprotein N. Preliminary studies

in our laboratory with in vitro translation

sys-tems suggestmRNAactivity is associated only withnonpolyadenylatedRNAin Akabane

virus-infected cells, thus supporting our contention

that in this system nopolyadenylatedRNAs can be detected.

It is accepted that the first new synthetic

event in cells infected with a negative-strand

virus isthe production ofmRNA by the

virus-associatedRNApolymerase. Thisprimary

tran-scription is independent of protein synthesis.

However, the change to secondary (amplified)

transcriptionis dependenton theexpression of

viralproteins, and this transitioncanbe

prevent-ed by the use of inhibitors ofprotein synthesis (10, 11, 28). Thus, cycloheximide and puromy-cin were used to treat Akabane virus-infected

cells from the time of infection to examine

http://jvi.asm.org/

(10)

primary RNAtranscription products. When

ana-lyzedeither bysedimentation inSDS-containing

sucrose density gradients or by hybridization

plus gelelectrophoresis, novirus-specific RNA

productscould be detected (Fig. 5and6).

Simi-lar results were obtained if the inhibitors were

not added untilthe time of maximal RNA

syn-thesis, when secondary transcription was

un-doubtedly occurring. This was a surprising

re-sult as secondary transcription with other

negative-strandviruses hasnotbeen reportedto

bedependentuponcontinuing protein synthesis

(23, 24). Similar effects were observed if an

alternate protein synthesis inhibitor, emetine,

was used (data not shown). Studies on the

primary transcription of bunyaviruses in the

presence ofcycloheximide orpuromycin have

been reported previously. In snowshoe hare

virus-infected cellsat ahigh MOI, low levels of

RNAsynthesis could be detected by

hybridiza-tion in thepresenceofcycloheximidebutnotin

thepresence ofpuromycin (27). Similar results

alsowereobtainedbyusingmutantsof this virus

(27). RNA synthesis in Bunyamwera

virus-in-fected cells in thepresenceofcycloheximide has

been examined(18) andagain[3H]uridine

incor-poration levels were only marginally above

background levels, and the products were not

showntobecomplementary tovRNA.Thus, in

bothof thesereportsand the resultspresentedin

thiscommunication, primaryRNAtranscription

by bunyaviruses in the absence ofprotein

syn-thesis hasnotbeen demonstrated conclusively.

Thesynthesis of protein andRNAin Akabane

virus-infected cells appearedtobelinked.Since

theinhibitory effects of cycloheximideand

puro-mycin on protein synthesis can be readily

re-versedby thewashing of treated cells, itwasof

interest to see whether secondary RNA

tran-scription was affected similarly. The results in

Fig. 6aandb show thatsecondarytranscription

can be readily established after the removal of

either inhibitor (Fig. 6a) and can recommence

andapproach formersyntheticratesif

interrupt-ed by the action of these drugs (Fig. 6b).

Al-though it is not known whether the effects of

cycloheximideandpuromycinonAkabaneviral

RNA synthesis are direct orindirect, it seems

that the latteralternative ispossible because of

thecorrespondencebetween theeffects on

pro-teinand RNA synthesis and the reversible

na-ture ofeach (Fig. 6b). The conclusion follows

thatsecondaryRNAtranscription (and perhaps

primarytranscription as well) in Akabane

virus-infectedcells isdependentoncontinuingprotein

synthesis.

ACKNOWLEDGMENTS

WethankCecilia Devlin for excellent technical assistance. This work was supported by grants from the National

Health and Medical Research Council ofAustraliaand the Australian Meat ResearchCommittee.

LITERATURE CITED

1. Abraham, G. 1979. Theeffect of ultraviolet radiationon

the primary transcription of influenza virus messenger RNAs.Virology 97:177-182.

2. Bean,W.J.,andR. W.Simpson. 1973.Primary transcrip-tionof the influenza virus genome in permissive cells. Virology56:646-651.

3. Bishop, D. H. L. 1978.Geneticpotentialofbunyaviruses. Curr.Top. Microbiol. Immunol. 86:1-33.

4. Bishop, D. H. L., B. J. Beaty, and R.E. Shope. 1980. Recombinationand genecoding assignmentsof bunyavi-rusesandarenaviruses.Ann.N.Y.Acad.Sci.354:84-106. 5. Bishop, D. H. L., C. H. Calisher, J. Casals, M. P. Chumakov, S.Y.Gaidamovich, C. Hannoun, D. K. Lvov, I. D. Marshall, N. Oker-Blom, R. F. Pettersson, J. S. Porterfleld, P.K. Russell, R.E. Shope, and E.G. Westaway.1980.Bunyaviridae.Intervirology 14:125-143. 6. Bishop, D. H. L., K. G. Gould,H. Akashi, and C.M. Clerx-van Haaster. 1982. The complete sequence and coding content of snowshoe hare bunyavirus small (S) viral RNA. Nucleic Acids Res. 10:3703-3713.

7. Bishop, D. H. L., andR.E.Shope. 1979.Bunyaviridae,p. 1-156. In H. Fraenkel-Conrat and R. R. Wagner (ed.), Comprehensive virology, vol. 14. Plenum Publishing Corp., New York.

8. Bonner,W.M.,and R. A.Laskey.1974. Afilmdetection method for tritium-labeled proteinsandnucleic acids in polyacrylamide gels. Eur. J. Biochem. 46:83-88. 9. Cash, P.,A.C.Vezz,J. R.Gentsch,and D. H. L.Bishop.

1979.Genomecomplexitiesof the three mRNAspeciesof snowshoe harebunyavirus and invitro translation of S mRNA toviral Npolypeptide.J. Virol.31:685-694. 10. Choppin,P.W., and R.W.Compans. 1975.Reproduction

ofparamyxoviruses, p. 95-178. In H. Fraenkel-Conrat and R.R.Wagner(ed.),Comprehensivevirology, vol.4. PlenumPublishing Corp.,New York.

11. Compans,R.W., andP. W.Choppin.1975.Reproduction oforthomyxoviruses,p.179-252.InH.Fraenkel-Conrat and R. R.Wagner (ed.),Comprehensivevirology, vol.4. PlenumPublishing Corp., New York.

12. Fuller, F., andD. H. L. Bishop. 1982. Identificationof virus-coded nonstructural polypeptides in bunyavirus-infected cells. J. Virol.41:643-648.

13. Gentsch, J. R.,and D. H. L. Bishop. 1978. Small viral RNA segmentofbunyavirusescodesfor viral nucleocap-sidprotein.J.Virol. 28:417-419.

14. Gentsch,J.R., and D.H. L.Bishop. 1979.Aviral RNA segmentofbunyavirusescodes fortwoglycoproteins,Gl and G2. J. Virol. 30:767-770.

15. Gonzalez-Scarano, F.,R. E.Shope,C.E.Calisher,andN. Nathanson. 1982.Characterisation of monoclonal antibod-ies against Gl and Nproteinsof La Crosse and Tahyna, twoCaliforniaserogroupbunyaviruses. Virology 120:42-53.

16.Gorman,B.M.,J.Taylor,P.J. Walker,and P. R.Young. 1978. Isolation of recombinants between related orbivi-ruses.J. Gen.Virol. 41:333-342.

17. Hay,A.J.,B.Lomniczi,A. R.Bellamy,andJ. J.Skehel. 1977.Transcription oftheinfluenza virus genome. Virolo-gy 83:337-355.

18. Kascsak,R.J.,andM.J. Lyons.1977.Bunyamwera virus I.Themolecularcomplexity of the virion RNA. Virology 82:37-47.

19. Laskey,R. A.,and A. D. Mils. 1975. Quantitative film detection of3Hand14Cinpolyacrylamide gels by fluorog-raphy.Eur.J.Biochem. 56:335-341.

20. Marcus, I., D. H. Engelhardt, J. M. Hurt, and M.J. Sekellick. 1971. Interferon action: inhibition of vesicular stomatitis virusRNAsynthesisinducedby virion-bound polymerase. Science 174:593-598.

47,

http://jvi.asm.org/

(11)

21. McPhee, D. A., and A. J. Della-Porta.1980.Biochemical and serological comparisons of Australian bunyaviruses belongingtotheSimbuserogroup, p.93-101. In D. H.L. Bishop and R. W. Compans (ed.), The replication of negative strand viruses, Proceedings of the4th Interna-tional Symposium on Negative Strand Viruses, Virgin Islands. Elsevier/North-Holland, New York.

22. Obijeski,J. F., D. H. L.Bishop,F. A. Murphy, and E. L. Palmer. 1976. Structural proteins of La Crosse virus.J. Virol. 19:985-997.

23. Robinson, W. S. 1971.Sendai virus RNA synthesis and nucleocapsidformation in thepresenceofcycloheximide. Virology 44:494-502.

24. Scholtissek, C., and R. Rott. 1970. Synthesis in vivo of influenza virus plus and minus strand RNA and its prefer-ential inhibitionby antibiotics. Virology 40:989-9%.

25. Short, N. J.,A.D.Meek, and L.Dalgarno.1982.Seven infection-specific polypeptides in BHK cells infected with Bunyamweravirus. J. Virol. 43:840-843.

26. Ulmanen,I., P. Seppala, andR.F. Pettersson. 1981. In vitro translation of Uukuniemivirus-specific RNAs: iden-tification ofanon-structural protein anda precursorto

membraneglycoproteins. J. Virol. 37:72-79.

27. Vezza, A.C.,P. M.Repik,P.Cash,and D. H. L.Bishop. 1979. In vivo transcription and protein synthesis capabili-tiesof bunyaviruses: wild-type snowshoe hare virus and its temperature-sensitivegroupI,groupII,andgroupI/II

mutants.J. Virol. 31:426-436.

28. Wagner, R. R. 1975. Reproduction of rhabdoviruses,p.

1-93. In H.Fraenkel-Conrat and R. R. Wagner (ed.), Com-prehensive virology, vol. 4, Plenum Publishing Corp., New York.

J. VIROL.

http://jvi.asm.org/