JOURNAL OF INTERFERON RESEARCH 9:23-34 (1989)Mary Ann Liebert, Inc., Publishers

Possible Involvement of Virus-Induced ProteinKinase in the Antiviral State Induced with

Interferon-7 against Sindbis Virus

SAKURA SAITO

ABSTRACT

The kinetics of induction of the antiviral state against two RNA viruses, vesicular stoma-titis virus (VSV) and Sindbis virus, by human interferons (IFNs)-o-, -ß, and -7 was mea-sured and compared with that of 2,5-oligoadenylate (2-5A) synthetase and protein ki-nase in cells treated with IFNs. Both enzymes were induced in similar time courses andthe induction by IFN-7 was slower than IFN-a or ß. The time course of the induction ofantiviral state against VSV almost paralleled with that of the enzyme induction by eachIFN species. In contrast, the induction of antiviral state against Sindbis virus with IFN-7was as fast as that induced with IFN-a or ß, in spite of the slower enzyme induction byIFN-7. The addition of actinomycin D at the time of virus challenge did not substantiallyaffect the induction of the antiviral state against VSV, but markedly retarded the estab-lishment of IFN-7-induced antiviral state against Sindbis virus. These results suggest thatthe antiviral machinery against VSV is induced solely by IFN during the pretreatment,but the one against Sindbis virus involves additional cellular component(s) inducedshortly after virus infection, especially in the case of IFN-7. Sindbis virus, but not VSV,induced a cellular double-stranded (ds) RNA-dependent protein kinase at an early stageof virus replication. The kinase appeared to phosphorylate the same protein as IFN-in-duced kinase in the IFN-7-treated and Sindbis virus-infected cells, leading to an increasedphosphorylation level. These results are consistent with the idea that the Sindbis virus-in-duced protein kinase may be involved in the IFN-7-induced antiviral state against Sindbisvirus.

INTRODUCTION

Treatment of animal cells with specific interferons (IFNs) renders them resistant to a widespectrum of viruses. However, the molecular basis of antiviral activity of IFNs may have some

variations depending on virus species with which the effect of IFN is tested. In fact, there have beenmany reports dealing with differential sensitivity of viruses to IFN as well as the differential sensi-tivities of a virus to different IFN species.(1-2>

Department of Measles Virus, National Institute of Health of Japan, 4-7-1, Gakuen, Musashimurayama,Tokyo 190-12, Japan.

23

SAITO

A systematic study comparing the kinetics of establishment of antiviral state induced by IFNs-a,-ß, and -7 against different viral species would help understanding possible differences in the modeof action of three IFN species as well as the involvement of a virus-specific machinery, if any, inantiviral action of IFN.

In this study, vesicular stomatitis virus (VSV), with which differential sensitivity to IFN-a andIFN-7 has been noted,(2> and Sindbis virus, which has been used routinely in our laboratory for as-

saying IFN titer, were chosen as challenge viruses, and kinetics of induction of antiviral state withthree IFN species in FL cells were compared. The induction kinetics of 2',5'-oligoadenylate (2-5A)synthetase and double-stranded (ds) RNA-dependent protein kinase, both of which are induced byIFNs and assumed to be partly responsible for the inhibition of protein synthesis,'3*" were also de-termined.

MATERIALS AND METHODS

Cells and Viruses: Human amnion FL cell lines were grown in Eagle's minimum essential me-dium (MEM) containing 100 U/ml of penicillin and 100 pg/ml of streptomycin and supplementedwith 5% calf serum. The cells were inoculated to 4 ml of the medium in 36-mm plastic plates at a

density of 6 x lOVml and cultured for 1 day. VSV and Sindbis virus were grown and titrated byplaque formation in primary chick embryo cell cultures. The antiviral state induced with IFN-a, -ß,and -7 was assessed by the reduction of virus yield. The virus yield was measured after one-stepgrowth as follows: each virus was infected at multiplicity of infection (moi) of 5. After 1 h adsorp-tion of virus, the cell monolayer was washed three times with 2 ml of Eagle's MEM, replenishedwith 2 ml of Eagle's MEM containing 2% calf serum, and then incubated at 37°C, for an additional11 h. At the end of incubation for 12 h altogether, culture plates were frozen and thawed threetimes and then titrated for each virus by the plaque-forming method. The antiviral state is ex-

pressed in terms of logarithms of the ratio of the viral yields between cells untreated and treatedwith IFNs (log viral yield ratio).

IFNs and Antibodies: Natural human IFN-/3 (sp. act., 10s IU/mg protein) was supplied by ToreyCo. Natural human IFN-a and IFN-7, produced by Green Cross Co., had a specific activity of 106IU/mg protein, and 106 U/mg protein, respectively. Recombinant human IFN-a (rIFN-a) pro-duced by Takeda Chemical Industries Ltd., rIFN-/3 produced by Torey Co., and rIFN-7 producedby Suntory Ltd. had specific activities of 2 x 108 IU/mg protein, 10s IU/mg protein, and 107units/mg protein, respectively. The titers of IFNs were determined by the method described by Ko-hase et a/.,'51 which involves dye-uptake after challenge of Sindbis virus to IFN-treated FL cells.Anti-IFN-/3 rabbit globulin, made by Torey Co., and anti-IFN-a sheep globulin (NIH Cat. No.G-026-502-568) were kindly donated by Torey Co. and by Dr. Tsukui of Central Blood Center,The Japanese Red Cross Society, respectively.

Assay of2-5A Synthetase: Cells (2.4 x 106) were lysed with 1 ml of solution A (20 mM HEPES,pH 7.5, 5 mMMgCL., 25 mM KC1, 1 mM dithiothreitol, 10% glycerol) containing 0.5% NP-40 at0°C for 10 min.1" The supernatant obtained by centrifugation at 2000 x g for 5 min was stored as

cytoplasmic fraction. The activity of 2-5A synthetase was measured by the method of Ball'7' withmodification as follows. Ten-microliter aliquots of the cytoplasmic or nuclear extract were appliedto a column of poly(I):poly(C)-agarose (PL-Biochemicals) (30 pi), washed with 400 pi of solutionA and then incubated at 37°C in the presence of [3H]ATP (10 pi, 1 mM, 0.2 mCi/ml). After incu-bation for 2 h, the column was eluted with 10 pi of water. The eluate was heated at 80°C for 2 minand then digested with phosphatase (Sigma Type III). An aliquot (10 pi) of the eluate was appliedto a polyethyleneimine (PEI)-cellulose thin-layer plate (Merck). Incorporation of [3H]adenosine to2-5A was measured by the method of Ball.'71 The enzyme activity was expressed as percent conver-sion of ATP into 2-5A."3'

24

HIGH SENSITIVITY OF SINDBIS TO IFN-7

Assay of Protein Kinase: The activity of protein kinase in ribosomal salt wash (RSW) fractionsprepared from untreated and IFN-treated cells was measured essentially according to the methoddescribed by Samuel."» FL cells (2.4 x 106) were lysed with 200 ¡à of solution B (20 mM HEPES,pH 7.5; 150 mM KC1, 4 mM Mg(OAc)2, 1 mM DTT, 0.5% NP-40) and centrifuged for 10 min at10,000 x g. The supernatant, the S-10 fraction, was ultracentrifuged at 120,000 x g 60 min to geta ribosomal fraction. The ribosome pellet was suspended in solution B containing 0.5 MKC1, soni-cated for 60 s, stirred for 20 min, and centrifuged for 1 h at 120,000 x g. The resulting supernatantsolution was dialyzed against solution B containing 150 mM KC1 for 6 h. The dialyzed RSW frac-tions were stored at -80°C. The standard reaction mixture (50 ju.1) contained solution B (30 pi), 10ul of RSW fraction, 5 pi of 100 /iM ATP containing 5 pCi of [7-"P]ATP, and 5 /tl of 100 pg/mlpoly(I):poly(C). Samples were analyzed by SDS-polyacrylamide gel electrophoresis after incubationfor 20 min at 30° C. The phosphorylated 80K protein was detected by autoradiography after elec-trophoresis, and its optical density was monitored by a video densitometer (model 620, Bio-Rad).The activity of protein kinase was expressed in arbitrary units by the area under the 80K peak on

the densitometer tracings. The molecular weight 80K was estimated by coelectrophoresis of 14Cmethylated proteins (Amersham CFA 626). Its 80K band was significantly higher than the 67Kband, which was phosphorylated in the extract from IFN-treated L929 cells.

RESULTSInduction of antiviral state

The time course of induction of the antiviral state with three IFN species against VSV and Sind-bis virus was compared (Fig. 1). FL cell cultures were incubated for varying time periods with 1,000units of each IFN species. That dose was sufficient to induce the maximal level of antiviral state

against respective viruses. The yield reduction of each virus was determined as described in Meth-ods. IFN-7 higher than 1,000 units did not raise the level of antiviral state and did not acceleratethe time course of induction of antiviral state against VSV (data not shown). The induction of anti-viral state against VSV with IFN-7 was markedly slow as compared with IFNs-a and -ß and itsmaximal level was low. It took more than 20 h for the maximal antiviral state to be induced. Incontrast, the antiviral state against Sindbis virus was induced rapidly in almost indistinguishabletime courses with either one of the three IFN species; thus a full antiviral state was induced in 3-6h. Three rIFN species also induced antiviral states against these viruses in the similar time courses

as natural IFN species (data not shown).

Inductions of 2-5A synthetase and protein kinase

Activity of 2-5A synthetase was measured in parallel in similar experiments with those shown inFig. 1 using the three IFN species. The time course of 2-5A synthetase induction in cytoplasm with1,000 units of IFNs up to 24 h is shown in Fig. 2A. The increase in 2-5A synthetase in the nucleusfollowed a similar time course to that in the cytoplasm (data not shown). Induction of the enzymeby IFN-/3 and IFN-a became detectable after 6 h and 9 h, respectively, and increased steadily there-after. On the other hand, IFN-7 induced only a very low level of the enzyme even after 24 h.

IFNs also induce a protein kinase which is activated by dsRNA to phosphorylate an 80K proteinin FL cells. Time course of induction of protein kinase was also examined (Fig. 2B). In this case,too, IFN-7 was much less effective in the enzyme induction as compared with IFN-a and IFN-18,and the enzyme induction by IFN-7 continued during 24 h. The rIFN species also induced proteinkinase in similar time courses as natural IFN species (data not shown). It can be seen that the induc-tion of the antiviral states was completed by 24 h with all IFN species (Fig. 1). It may be that more

enzymes eventually are synthesized in the cells treated by IFN-a or IFN-/3 than are necessary to in-hibit viral growth maximally. The differential effectiveness of three IFN species in inducing these

25

SAITO

-o

fi>,

CO3

Time ( h )FIG. 1. Time course of the development of antiviral state against VSV and Sindbis virus. FL cellswere incubated for various periods as indicated with 1,000 units of each IFN. After pretreatmentwith IFNs, VSV and Sindbis virus were infected at moi 5 and the viral yield was measured. Theyield of VSV and Sindbis virus in control cells were 10* ° and 106 2 pfu/ml, respectively.

enzymes as shown in Figs. 2, A and B, roughly parallels their relative effectiveness in the inductionof antiviral state against VSV (Fig. 1). This is consistent with the idea that 2-5A and protein kinasesystems may be involved intimately in the inhibition of the VSV replication. On the other hand, therapid induction of antiviral state against Sindbis virus with IFN-7 suggested that a yet unidentifiedfactor other than the IFN-induced enzymes might be involved in the establishment of the antiviralstate against this virus.

Effect of actinomycin D on the antiviral state

It is noteworthy that the induction of antiviral state against Sindbis virus by IFN-7 was muchfaster and more extensive than that of 2-5A synthetase and dsRNA-dependent protein kinase,which are typical IFN-induced proteins. To understand the mechanism by which Sindbis virus isrendered highly sensitive to IFN-7, the possibility of the involvement of a factor induced by viralinfection was examined. Transcription of cellular genes after virus infection was arrested by addingactinomycin D (5 pg/ml) for 1 h during virus adsorption. Figure 3, A and B, shows the effects ofactinomycin D on the establishment of antiviral state by IFN-7 against VSV and Sindbis virus, re-

spectively.

26

HIGH SENSITIVITY OF SINDBIS TO IFN-7

( nun XjDj;iqjv )u|8j.0Jd »09 1-0 uo!j.t>iÄJoqdsoiJd CQ

'_: E

( diV ¿0 U0ISJ8AU03 % )X;|AU3D aSD^eqUlAs VQ'2

zJ>'S -

aieo p.-"V,

-—1

"S-ocues

-O3

»h í<rv cu*-* «5i*. ^ c/ll¡ O U

¿^ &•g O CU

ä * Ses es««fe.-*i

.

SWS1 -<'S

.5 es

aj HH

VI fO1 1 v^n o JD

esC

a'53olH&

ccu

-Oecuacu-a

1

<Zoíc/l aT3ces

.•O — cu

_e Pu ftHH r_^ eu

sO O è3»O 33 .y C-3 2'55-. cu oo a Syoes y ^¿

T3 «- ©eu C 00

s g ° 2 &S-t-» ^ oy «» 'ses cij cej

cu C?«1 oI" °-H-i« 2 §

cuC/l

-2J3c

Si:5.9

-5* « Pu -°

Sí«-S s »3 g_

*-* r-! H

13 U.S»Bu-Si111 2.5 ü ö o2 y SS eseSjÇ- cu

•3 < U «

eu cu "3 cu

§ B c5« 8"S g3B C O -5O B u. !"

•C » O- 3S fc >s

a •E-í <¿¿ o111. .£?'

£3«3

U PU eu 73 2E .9 1 § £

27

SAITO

Time ( h )FIG. 3. The effect of actinomycin D on the development of antiviral states with IFN-7 againstVSV and Sindbis virus. Time courses of induction of antiviral states against VSV (A) and Sindbisvirus (B) in the absence (•) or presence (O) of actinomycin D were compared. FL cells were treatedwith IFNs and infected with each virus under the same conditions as in Fig. 1. Actinomycin D (5pg/ml) was added for 1 h during virus adsorption and then removed. The yields of VSV and Sind-bis virus in the absence of actinomycin D were 10" and 106 " pfu/ml, respectively, and those in itspresence were 10" and 10" pfu/ml, respectively.

The time course of the induction of antiviral state against VSV was not influenced by the addi-tion of actinomycin D at all. The response to IFN-a and IFN-/3 also was not affected by actinomy-cin D (data not shown). In marked contrast, the antiviral state against Sindbis virus was profoundlyaffected by actinomycin D. Thus, the antiviral state induced by IFN-7 reached the maximum levelwithin 3 h in the absence of actinomycin D, while in the presence of actinomycin D it took 18 h forthe maximum level to be attained (Fig. 3B). The time course of the induction of antiviral state byIFN-a and IFN-0 also lagged behind that of the control by 2-3 h in the presence of actinomycin D(data not shown). The results suggest that the rapid induction of antiviral state against Sindbis virusby IFNs requires a de novo synthesis of an additional cellular macromolecule triggered by virus in-fection. Figure 4 shows that actinomycin D was most effective in inhibiting the development ofantiviral state against Sindbis virus when added at the same time with infection. If actinomycin Dwas added later, its effectiveness was decreased rapidly, and by 4 h after infection its effect on theIFN-7-induced antiviral state became negligible. This result strongly suggests that not only the IFN-induced transcription of antiviral protein genes but also the transcription of yet unidentified cellu-lar genes, which is triggered by virus infection and occurs early in infection, is necessary at least forIFN-7 to build up an efficient antiviral state against Sindbis virus. The results of experiments usingactinomycin D after short pretreatment by IFN-7 cannot exclude a possible involvement of un-known factor(s) that is induced by IFN-7 and turns over rapidly. However, the results that pro-

28

HIGH SENSITIVITY OF SINDBIS TO IFN-7

100

«>•

CO3t_'>

O50

cg

00 1 2 3 4 5

Time of actinomycinD addition( h after infection )

FIG. 4. Effect of actinomycin D added at various times after infection with Sindbis virus on thedevelopment of antiviral state with IFN-7. FL cells were treated for 3 h by 1,000 units of IFN-7 andthen were infected with Sindbis virus at moi 5. Actinomycin D (5 pg/ml) was added for 1 h at vari-ous times after infection as indicated and viral yield was determined. The inhibition of viral yieldwas estimated by comparing the viral yield in the presence and absence of actinomycin D. The viralyields in the IFN-7-treated and untreated cells were 105 3 and 10" pfu/ml, respectively.

longed pretreatment by IFN-7 decreased the effect of actinomycin D on the antiviral state seem toeliminate that possibility.

IFN could be a candidate for the factor responsible for the above-mentioned effect, since its syn-thesis is triggered by viral infection. However, IFN activities could not be detected in the mediacollected from cell cultures infected by Sindbis virus until 8 h postinfection, even though it was re-

leased from Sindbis virus-infected cells 18 h postinfection (Table 1). Moreover, the induction kinet-ics of antiviral state against Sindbis virus was not influenced at all, even if anti-IFN-a and anti-IFN-/3 antisera in amount sufficient to neutralize 1,000 units of each IFN were added to the media(data not shown). These results indicate that the rapid and efficient protection by IFN against Sind-bis virus is not mediated by IFN released from the virus infected cells.

Enzyme levels in IFN-treated and virus-infected cellsIn searching for other cellular factor(s) that render Sindbis virus highly sensitive to IFN, the

levels of 2-5A synthetase and dsRNA-dependent protein kinase were examined in the cells whichwere treated with IFN and infected with viruses.

The activities of 2-5A synthetase were very low in cells treated with IFN for 3 h and were not af-fected significantly by viral infection (data not shown).

Figure 5 shows the activities of dsRNA-dependent protein kinase in the cells that were treatedwith or without IFN for 3 h and then infected with either VSV or Sindbis virus. Cell extracts were

29

SAITO

Table 1. IFN Released into the Media from theSindbis Virus-Infected Cells with or

without IFN Pretreatment

IFN 8 h 18 hpretreatment* postinfection postinfection

40b2

<2<2

aFL cells were treated for 3 h with 1,000 units of eachIFN before those were infected with Sindbis virus at moi 5.

bPotency of IFN (units) released into the media.

prepared at 5 h postinfection. IFN-7 did not induce a significant level of protein kinase in unin-fected cells under this condition (Fig. 5, lane 4). The infection with VSV for some reason slightlydecreased the activities of the enzyme induced with every IFN species. On the other hand, it is note-worthy that the infection of Sindbis virus induced a protein kinase (subsequently referred to as

PK-S) which phosphorylates 80K protein even without IFN treatment. The enhancement of proteinkinase by Sindbis virus infection also was obvious in extracts of cells treated with IFN-7 (Fig. 5B),but was not readily decernible in IFN-a- or IFN-/3-treated cells in which a substantial amount of theenzyme was induced even without viral infection.

Sindbis virus-induced protein kinase (PK-S)To investigate the time course of induction of Sindbis virus-induced protein kinase, cells were

lysed, extracted at 3, 5, and 7 h postinfection with Sindbis virus and the activities of protein kinasewere examined (Fig. 6A). Sindbis virus induced a protein kinase (PK-S) already at 3 h postinfec-tion. Furthermore, the enzyme was activated by dsRNA. The addition of poly(I):poly(C) wasneeded because the accumulation of viral dsRNA at 3 h postinfection was not enough to activatethe PK-S. After 5 h postinfection, however, enough viral dsRNA was accumulated in the infectedcells and extracted together with PK-S to activate it. The induction of PK-S was inhibited by addi-tion of actinomycin D (Fig. 6B). The addition of actinomycin D before viral infection was most ef-fective. These results indicate that the PK-S is one of the cellular gene products that are rapidly in-duced on infection with Sindbis virus.

DISCUSSION

IN FL cells, the induction of 2-5A synthetase and protein kinase with IFN-7 was considerablyslower and lower than that with IFN-a and IFN-/3. Efficiency of IFN-7 to inhibit the replication ofVSV was correspondingly lower than IFN-a and IFN-/3, suggesting that antiviral state against VSVand the level of these enzymes are regulated by a similar genetic mechanism. Schulman andRevel"0' also suggested that the expression of both 2-5A synthetase mRNA and the antiviral effectagainst VSV was controlled by a similar mechanism. The present results with VSV are also consis-tent with those of Samuel and Knuston,'2' in that VSV was less sensitive to IFN-7 than to IFN-aand that IFN-7 was a poor inducer of dsRNA-dependent protein kinase. There are very suggesive

<2a <2ß <2

y <2

30

HIGH SENSITIVITY OF SINDBIS TO IFN-7

1 2 3 4 5 6 7 8 9 10 11 12

— 80K

VSV -» L sindbis -J

U = = =

g BFIG. 5. Activities of protein kinase in IFN-treated and virus-infected cells. A. Autoradiogramshowing the effects of IFN treatment and viral infection. FL cells were treated for 3 h with eachIFN (1,000 units) and were infected with an appropriate virus at moi 5. Uninfected control cellsthat were treated similarly with IFN were replenished with fresh medium instead of virus suspen-sion. The extracts were prepared at 5 h postinfection in the same manner and assays were doneusing the same volume of the extracts in the presence of poly(I):poly(C) (1 pg/ml): lane 1, none;lane 2, IFN-a; lane 3, IFN-/3; lane 4, IFN-7; lanes 5-8, same as lanes 1-4 and infected by VSV;lanes 9-12, same as lanes 1-4 and infected by Sindbis virus. B. The extent of phosphorylation of80K protein shown in A is estimated by densitometry and is presented in terms of arbitrary units byusing the peak area of the 80K protein band.

31

SAITO

r°i r3i r5"i r7i T-ADir*ADiPolyhC -+- + - + - + - + - +

FIG. 6. Induction of protein kinase by Sindbis virus infection and the effect of actinomycin D.A. FL cells were infected with Sindbis virus at moi 5. Extracts for enzyme assay were prepared atvarious times after viral infection as indicated. B. FL cells were pretreated for 2 h with actinomy-cin D (5 pg/ml) and then infected with Sindbis virus as in A. Extracts for enzyme assay were pre-pared at 5 h postinfection. Assay of protein kinase was done in the absence and presence of poly(I):poly(C) (1 pg/ml).

reports that the protein kinase plays an important role in antiviral state against VSV, in which a fac-tor that inhibits IFN-mediated dsRNA-dependent protein kinase was described in vaccinia-infectedcells, and reported to rescue VSV from the antiviral state induced with IFN."1'13'

In contrast to VSV, Sindbis virus was as sensitive to IFN-7 as to IFN-a and IFN-/3. The timecourse of the induction of the antiviral state against Sindbis virus was faster than that against VSV.These results are essentially in accordance with those of Dianzani et al.'14' and Rose et al.,'15' whohave also observed differential effectiveness among IFN species and to individual viral species.Development of antiviral state against Sindbis virus far preceded the induction of 2-5A synthetaseand protein kinase, suggesting that these alone are not sufficient to inhibit the replication of Sindbisvirus in IFN-treated cells.

Effects of IFN-a, IFN-/3, and IFN-7 on the induction kinetics of the antiviral state against Sindbisvirus were almost indistinguishable from each other. However, addition of actinomycin D at the timeof virus challenge significantly delayed the induction of the antiviral state against Sindbis virus butnot VSV, especially when IFN-7 was used. The effect of actinomycin D on the action of IFN-7strongly suggests that pretreatment with IFN-7 alone is sufficient for the slow induction of the anti-viral state against VSV, but not for the rapid induction of the full antiviral state against Sindbisvirus. The process of induction of antiviral state with IFN-7 seems to be markedly accelerated by anunidentified cellular gene product whose synthesis is rapidly triggered by Sindbis virus but not VSVinfection. Endogenous IFN that might be released by Sindbis virus-infected cells did not seem to bethe active component.Sindbis virus infection did not affect the activity of 2-5A synthetase either. In contrast, the activityof dsRNA-dependent protein kinase in IFN-7-treated and Sindbis virus-infected cells was signifi-cantly higher than that in IFN-7-treated control cells (Fig. 5). Moreover, Sindbis virus was found tobe able to induce PK-S, even in the absence of IFN treatment. Although the functional role of PK-Sin usual infection by Sindbis virus is unclear at present, it may serve to restrict the viral growth to

32

HIGH SENSITIVITY OF SINDBIS TO IFN-7

some extent in the absence of any external antiviral agent. The fact that the yield of Sindbis virus issignificantly higher in the presence of actinomycin D than in the absence (see legend to Fig. 3) maybe consistent with this idea. However, PK-S alone is insufficient and IFN-induced component isneeded simultaneously for Sindbis virus to be highly sensitive to IFN-7.

The induction of PK-S by Sindbis virus has not been noted previously. It has the following char-acteristics: (i) It appears to phosphorylate the same 80K protein as IFN-induced protein kinase; (ii)it is activated by dsRNA [poly(I):poly(C) or Sindbis viral dsRNA]; (iii) it is a product of a cellulargene and is induced rapidly after virus infection in actinomycin D-sensitive manner. Aujean eta/."61 observed that infection of L929 cells with mengovirus resulted in 2.5-fold-enhanced phos-phorylation of a 67K protein as compared with noninfected cells.

The PK-S activity is induced rapidly after infection with Sindbis virus, and the time course of in-duction agreed well with that of the cellular factor which enhances the sensitivity of Sindbis virus to

IFN-7. Moreover, it is not induced by VSV whose infection does not affect the induction kineticsof the antiviral state by IFN-7. It is unlikely that the differential induction of protein kinase is dueto the differential shut-down of the cellular protein synthesis by VSV and Sindbis virus, since a sig-nificant amount of PK-S was induced by 3 h postinfection, when VSV-induced shut-down had not

yet become obvious. It is tempting to assume that the induction of protein kinase is the rate-limit-ing step in the establishment of the antiviral state against Sindbis virus, and that Sindbis virus-in-duced protein kinase, be it identical or not with the IFN-7-induced one, meets the normal require-ment for the latter and contributes to accelerate the establishment of antiviral state. It would beinteresting to see if analogous factors are also produced and participate in the antiviral state inother cell lines that are briefly treated with IFN-7 and infected with Sindbis virus.

ACKNOWLEDGMENTS

I thank Torey Co., Green Cross Co., and Dr. K. Tsukui for the gifts of IFNs and anti-IFN-glob-ulins, and Dr. S. Kohno, Dr. S. Yamazaki, Dr. M. Kohase, and Dr. M. Kawakita for stimulatingdiscussions and critical reading of the manuscript.

REFERENCES

1. CAYLEY, P.J., DAVIES, J.A., McCULLAGH, K.G., and KERR, I.M. (1984). Activation of theppp(A2'p)nA system in interferon-treated Herpes simplex virus-infected cells and evidence for novel inhib-itors of the ppp(A2p)nA-dependent RNase. Eur. J. Biochem. 143, 165-174.

2. SAMUEL, CE., and KNUSTON, G.S. (1983). Mechanism of interferon action: Human leukocyte andimmune interferons regulate the expression of different genes and induce different antiviral states in hu-man amnion U cells. Virology 130, 474-484.

3. ROBERTS, W.K., HOVANESSIAN, A., BROWN, R.E., CLEMENS, M.J., and KERR, I.M. (1976). In-terferon mediated protein kinase and low-molecular-weight inhibitor of protein synthesis. Nature 264,477-480.

4. HOVANESSIAN, A.G., BROWN, R.E., and KERR, I.M. (1977). Synthesis of low molecular weight in-hibitor of protein synthesis with enzyme from interferon-treated cells. Nature 268, 537-539.

5. KOHASE, M., KOHNO, S., YAMAZAKI, S., SHISHIDO, A., and KONO, R. (1982). Potency stan-dardization of human interferon preparations for clinical trials, in: 77ie Clinical Potential of Interferons.R. Kono and J. Vilcek (eds.). Tokyo: University of Tokyo Press, pp. 299-309.

6. KIMCHI, A., SHULMAN, L., SCHMIDT, A., CHERNAJOVSKY, Y., FRADIN, A., and REVEL, M.(1979). Kinetics of the induction of three translation-regulatory enzymes by interferon. Proc. Nati. Acad.Sei. USA 76, 3208-3212.

7. BALL, L.A. (1979). Induction of 2,5-oligoadenylate synthetase activity and a new protein by chick inter-feron. Virology 94, 282-296.

8. SHIMIZU, N., and SOKAWA, Y. (1979). 2',5'-01igoadenylate synthetase activity in lymphocytes from

33

SAITO

normal mouse. J. Biol. Chem. 254, 12034-12037.9. SAMUEL, C.E. (1981). Procedures for measurement of phosphorylation of ribosome-associated proteinsin interferon-treated cells. Methods Enzymol. 79, 168-178.

10. SHULMAN, L., and REVEL, M. (1980). Interferon-dependent induction of mRNA activity for (2'5>oligo-isoadenylate synthetase. Nature 288, 98-100.

11. RICE, A., and KERR, I.M. (1984). Interferon-mediated, double-stranded RNA-dependent protein kinaseis inhibited in extracts from vaccinia virus-infected cells. J. Virol. 50, 229-236.

12. WHITAKER-DOWLING, P., and YOUNGNER, J.S. (1983). Vaccinia rescue of VSV from interferon-in-duced resistance: Reversal of translation block and inhibition of protein kinase activity. Virology 131,128-136.

13. WHITAKER-DOWLING, P., and YOUNGNER, J.S. (1984). Characterization of a specific kinase in-hibits the interferon-induced protein kinase. Virology 137, 171-181.

14. DIANZANI, I.F., ZUCCA, M., SCUPHAM, A., and GEORIADES, J.A. (1980). Immune and virus in-duced interferons may activate cells by different derepressional mechanisms. Nature 283, 400-402.

15. ROSE, J.M., CROWLY, C.K., and FLEISCHMANN, W.R., Jr. (1985). Evidence that IFN-a/0 inducestwo antiviral states active against different viruses. J. Gen. Virol. 66, 1153-1158.

16. AUJEAN, O., SANCEAU, J., FALCOFF, E., and FALCOFF, R. (1979). Location of enhanced ribonu-clease activity and of a phosphoprotein kinase in interferon-treated mengovirus-infected cells. Virology 92,583-586.

Address reprint requests to:Dr. S. Saito

Department of Measles VirusNational Institute of Japan

4-7-1, GakuenMusashimurayama, Tokyo 190-12, Japan

Received 18 May 1988/Accepted 24 August 1988

34