VIROLOGY

177.281-288

Minus-Strand

(1990)

RNA Synthesis by the Segmented Double-Stranded Requires Continuous Protein Synthesis NIKOS PAGRATlS* *Committee

AND

RNA Bacteriophage

46

HELEN R. REVEL?’

on Developmental Biology and tDepartment of Molecular Genetics and Cell Biology, University of Chicago, 920 East 58th Street, Chicago, Illinois 60637 Received

August

2 1, 1989; accepted

March

20, 1990

Bacteriophage $6 contains three dsRNA chromosomes. Strand-separating agarose gels were used to study plusand minus-strand synthesis in viva and the effect of protein synthesis inhibitors. Analysis of 46 RNA synthesis shows low levels of all three dsRNAs and ssRNAs at 10 min, increasing label uptake into all RNAs except the large message from 20 to 60 min, and a greater abundance of medium and small messages than large mRNAs at late times. Isoconformers of the small message are synthesized throughout infection. Northern analysis suggests that large messages made early may persist to direct continuing translation of L-segment-encoded transcription and replication proteins. The time course of 66 minus-strand RNA synthesis in viva, in the absence of background.label in host RNAs, is reported for the first time. Label in minus strands is detected only after heat denaturation of RNA samples and appears sequentially in the small, medium, and large strands beginning at 20 min. At both early and late times, chloramphenicol arrests minus-strand synthesis rapidly and all three mRNAs accumulate. The results are consistent with the reovirus asynchronous model for dsRNA viral replication: plus ssRNAs made first are used as templates for minus-strand synthesis. They also indicate that replication protein(s) acts stoichiometrically. o 199oAcademic PRBS, I~C.

INTRODUCTION

phase of reovirus replication (minus-strand synthesis) must be tightly coupled with assembly of subviral particles (reviewed by Zarbl and Millard, 1983). These data indicate that sorting and packaging of RNA genome segments, in addition to template recognition and the initiation of minus-strand synthesis, must proceed at the level of single plus strands. It has been suggested that $6 employs a similar replication and packaging strategy (Coplin et a/., 1975; Mindich and Bamford, 1988). Coplin et al. (1975) observed that ssRNAs, made via phage R17 RI-like intermediates, were chased into dsRNA. Early addition of chloramphenicol reduced label uptake into dsRNA two- to threefold and altered the pattern of ssRNA synthesis so that label increased in both the large mRNA and its RI-like precursor. They concluded that $6 replication is analogous to that of reoviruses in that complementary strands are made at different times: mRNA transcripts are made first and serve as templates for minus-strand synthesis to form duplex molecules. Unlike reoviruses, however, the initial transcriptional phase of replication was not conservative. The effects of chloramphenicol on 46 replication were not understood. The authors could not determine whether chloramphenicol reduced dsRNA synthesis because it blocked synthesis of proteins required for minus-strand synthesis or it blocked formation of regulatory proteins that normally repress synthesis of the large plus strands.

Bacteriophage 46 contains three segments of dsRNA in a subviral procapsid composed of four early proteins that are required for the normal pattern of in viva RNA synthesis (Sinclair and Mindich, 1976; Mindich et a/., 1976; Rimon and Haselkorn, 1978b; Bamford and Mindich, 1980; Revel et al., 1986; Pagratis, unpublished observations; reviewed in Mindich and Bamford, 1988). Final virion assembly occurs with addition of a nucleocapsid coat protein and membrane envelopment. This phage shares with the segmented dsRNA reoviruses an ability to package multiple chromosomes into a single structure. It is possible that genome packaging and replication are coordinated but the molecular mechanisms are currently unknown. Replication of dsRNA viral genomes must involve the synthesis of both plus and minus strands to produce duplex progeny molecules. It has been shown that reoviruses synthesize complementary strands asynchronously: plus strands are made first, by conservative transcription of the genome, and these transcripts then serve as templates for minus-strand synthesis (Schonberg et a/., 1971; Silverstein et a/., 1976). Because cycloheximide severely inhibits label incorporation into dsRNA (Watanabe et a/., 1967), and because dsRNA synthesis has been shown to proceed within nascent subviral particles (Zweerink, 1974), the final ’

To

whom

requests

for

reprints

should

be addressed. 281

0042-6822/90

$3.00

CopyrIght 0 1990 by Academic Press. Inc. All rights of reproduction I” any form reserved.

282

PAGRATIS

Subsequent studies have both clarified and confused the understanding of 46 in viva RNA synthesis. Experiments in several laboratories have documented that transcription occurs by a semiconservative strand displacement mechanism (Emori et al., 1980; Usala et a/., 1980; Van Etten et a/., 1980). A consequence is that uracil label can be incorporated directly into dsRNA by plus-strand synthesis alone. Experiments with chloramphenicol, however, raised questions with regard to the hypothesis that plus strands are templates for minus-strand synthesis. While the effects of addition of chloramphenicol at early times were confirmed, late addition of the antibiotic had no major effect on label uptake into dsRNA (Sinclair and Mindich, 1976; Rimon and Haselkorn, 1978a). An additional obser-vation that chloramphenicol blocks degradation of ssRNAs led to the suggestion that in the absence of the drug the chase of label from message RNAs into dsRNA might occur indirectly by degradation of ssRNAs to a nucleotide pool and resynthesis (Rimon and Haselkorn, 1978a). To provide a rational basis for further investigation of 46 replication and assembly we have reexamined 46 RNA synthesis in viva using strand-separating agarose gels to monitor the time course of plus-, minus-, and double-stranded RNA synthesis and to determine how this is affected by inhibitors of protein synthesis. The results confirm and refine the previous suggestion that 46 selects plus ssRNAs as templates for minus-strand synthesis and they show that replication protein(s) acts stoichiometrically. MATERIALS Phage, bacteria,

AND

METHODS

and media

The propagation of wild-type 46 (Vidaver eta/., 1973) on Pseudomonas phaseolicola HBl OY cells on LB medium or supplemented M9 minimal medium, as well as virion and RNA purification, was as described (Pagratis and Revel, 1990). Enzymes,

plasmids,

and chemicals

T7 RNA polymerase was from Boehringer-Mannheim. pT7/T3 plasmids carrying $6 cDNA fragments for synthesis of riboprobes have been described (Pagratis and Revel, 1990). [5,6-3H]Uracil (46 Ci/mmol) was purchased from Amersham, [a-32P]UTP (800 Ci/ mmol) was from New England Nuclear, and ultrapure agarose was a product of Bethesda Research Laboratories. Ribonucleotides and RNasin were from Pharmacia. Rifampicin, chloramphenicol, tetracycline, and streptomycin were Sigma products.

AND

REVEL

Agarose

gel electrophoresis

TBE and citrate agarose gels were as described (Pagratis and Revel, 1990) except that RNA samples were run rapidly into the gels at 10 V/cm for 10 min. Electrophoresis times or voltages that deviate from standard conditions are noted in the figure legends. Carrier 46 RNA was omitted when native ssRNAs were to be analyzed by Northern blot hybridization. Northern

analysis

Synthesis of 32P-labeled riboprobes from pT7/T3-$6 cDNA recombinant plasmids by T7 RNA polymerase and Northern blot hybridization was as previously described (Pagratis and Revel, 1990). In vivo 46 RNA synthesis $6-infected HBlOY cells in supplemented M9 medium were pulsed for 5 min at intervals with [3H]uracil (30 &i/ml) in the presence of rifampicin (200 pg/ml) and RNAs were extracted as previously described (Pagratis and Revel, 1990). Uracil label and rifampicin were omitted for Northern analysis. Details of experiments with protein synthesis inhibitors are given in the figure legends. RESULTS Synthesis

of 46 plus and minus RNA

Although previous investigators have analyzed the time course of $16 RNA synthesis in vivo it was necessary for us to establish the pattern of 46 RNA synthesis on our new agarose gel systems. Accordingly, $6 RNA samples, pulse labeled and isolated at various times after infection, were electrophoresed on strand-separating agarose gels, either before or after heat denaturation, as shown in Fig. 1. The autoradiograms of rifampicin-treated uninfected controls (U+) show that all labeled bands from infected cells are 46 RNA species, except for a band indicated as “DNA.” The pattern of native RNA (panel I) closely resembles that seen in previous studies using polyacrylamide-agarose gels (Coplin et al,, 1975, 1976; Rimon and Haselkorn, 1978a; Emori et a/., 1982). Specifically all mRNAs and dsRNAs are labeled at low levels early in infection. Label incorporation into all RNAs except large message increases rapidly from about 20 to 60 min postinfection and then levels off. At late times dsRNAs and the medium and small mRNAs are abundant. The data also show that a newly identified isoconformer of the small message (s’) (Pagratis and Revel, 1990) is synthesized throughout infection. No free minus strands were detected in this analysis of native RNA samples.

MINUS-STRAND

I

1 2

II

3

4

5

I 2 3

6

7

4 5

8

6 7

9

IO Ii

8

RNA

12 13 14

9 IO IL 12 13 14

A

III

SYNTHESIS

234

2

BY BACTERIOPHAGE

2

3

4

34

5

6

7 8

91011121314

4

5

6

23456

B 12

3

2

2 3

283

d6

3

4

4

5 6

5

6 7

7

8 9

8

10 11 I2 13

9 10 II 12 13

C 2

34

s

2 3 4

5

6

7 8

9 IO I1

12 13

St23s Time point after infection

FIG. 1. Time course of in viva $6 RNA synthesis analyzed on strand separating agarose gels. RNAs from &6-infected HBl OY. pulse labeled at various times with [3H]uracil (10 Ci/pl) for 5 min following a 1 0-min pretreatment with rifampicin (200 pg/ml), were extracted and electrophoresed before or after heat denaturation (see Materials and Methods). Panel 1: native RNA with added carrier RNA (0.2 pgllane) electrophoresed on a TBE gel at 4.6 V/cm for 5 hr-(A) ethidium stain, (B. C, D) autoradiograms exposed for 336,48, and 12 hr, respectively. Pane///: denatured RNA without carrier RNA electrophoresed on a TBE gel at 5.2 V/cm for 5.5 hr-(A) ethidium stain, (B, C, D) autoradiograms exposed for 336, 48, and 12 hr, respectively. Pane/ //I: denatured RNA with added carrier RNA (0.9 pg/lane) electrophoresed on a citrate gel at 7.1 V/cm for 4.5 hr-(A) ethidium stain, (B, C) autoradiograms exposed for 72 and 12 hr, respectively. N and D are native and denatured virion RNA; U+ and U- are RNA samples from uninfected cells labeled in the presence and absence of rifampicin; L, M, and S are the large, medium, and small 66 chromosomes; I, m, s, and s’are ssRNA species derived from L, M, and S by heat denaturation or by transcription; + indicates the RNA strand has the same sequence as &~6 mRNAs; - indicates the RNA strand has a sequence complementary to 66 mRNAs; DNA, 23 S, and 16 S are host DNA and ribosomal RNAs respectively. Arrows indicate positions of I, m. s, and s’ mRNAs.

The pattern of label incorporation in heat-denatured RNA samples from +6-infected cells is shown in Fig. 1, panels II and III. Plus strands migrate faster than minus

strands on TBE gels (panel II) but are retarded on citrate gels (panel Ill). Only five bands are seen because L strands are not resolved with TBE and S strands are

284

PAGRATIS

AND

REVEL

superimposed on citrate gels (Pagratis and Revel, 1990). At 10 min, only plus strands are labeled. Label in minus strands appears first at 20 min. Then label increases in both plus and minus strands until 60 min and levels off. The data from these gels suggest that initiation of synthesis of minus strands corresponding to the different segments may be asynchronous or uncoordinated. At a given exposure, s- is more intense than m- (panel II: C, lanes 4 and 5; D, lanes 5 and 6), and m- is more intense than I- (panel III: B, lanes 4 and 5; C, lanes 5 and 6). At late times the higher amount of label in m+ and s+ strands compared to the corresponding minus strands reflects the excess of medium and small mRNAs seen in the analysis of native RNA. Equal label incorporation into the I+ and I- strands correlates with the absence of detectable label in the large message (see Fig. 1, panel I). Steady-state

I

D

1

E 12

3 4 5

6 78

9 1011

levels of in viva 46 RNAs

The absence of pulse label in I mRNA at late times is paradoxical because L-encoded early proteins are synthesized throughout infection (Sinclair et a/., 1975; Rimon and Haselkorn, 1978a; Pagratis and Revel, unpublished data). To explore the possibilities that intact early I mRNA remains complexed with ribosomes or that subgenomic I mRNA could provide the template for late translation, we determined the amount and integrity of $16 mRNAs by Northern blot hybridization. Figure 2 (panels B and C, lanes l-l 0) shows that, in addition to the small and medium messages, free full-length I mRNA is available throughout infection and gradually accumulates with time. Subgenomic species of the large message were not detected with minus riboprobes. No free minus strands were detected with plus riboprobes (Fig. 2, panels B and C, lanes 12-20; panel E, lanes l-10). However, the conclusion that minus strands are present only in dsRNA is not completely warranted since small amounts of minus strands could rapidly anneal with excess plus strands after phenol treatment (Weissmann et al., 1968). Effect of chloramphenicol

FIG. 2. Northern analysis of total RNA extracted from @infected HBl OY at various times. Native RNA samples from @g-infected cells grown in the absence of rifampicin or uracil label were electrophoresed without carrier RNA, transferred to GeneScreenPlus membranes, and hybridized with mixtures of riboprobes corresponding to the large, medium, and small $16 RNA segments having the polarities shown under panels C and E. (A) TBE agarose gels electrophoresed at 5.4 V/cm for4 hr and stained with ethidium; (B, C) autoradiograms of membranes derived from the TBE gel (A) hybridized to minus and plus 96 riboprobes and exposed for 12 and 24 hr, respectively. (D)

on 46 RNA synthesis

The time course of label uptake into 46 RNAs following addition of chloramphenicol either early or late in development is shown in Fig. 3. Several exposures of the autoradiograms are presented to emphasize the quantitative differences that depend upon the time of

Citrate gel electrophoresed at 5.3 V/cm for 4 hr and stained with ethidium. (E) Autoradiogram of a membrane derived from the citrate gel (D) hybridized to plus 46 riboprobes and exposed for 12 hr. Symbols as described in Fig. 1.

MINUS-STRAND

RNA SYNTHESIS BY BACTERIOPHAGE

addition of the drug. In the analysis of native RNAs (Fig. 3, panels A-C), the pattern of RNA synthesis with a 15min pulse, shown in control lanes 1-6, is very similar to the results obtained with a 5-min pulse shown in Fig. 1 (panel 1) with the exception that discrete degradation products are more apparent with the longer pulse. This degradation of ssRNA is significantly reduced after addition of chloramphenicol at either 30 or 50 min postinfection (Fig. 3, A-C, lanes 7-l 6) as previously reported (Rimon and Haselkorn, 1978a). We find that early addition of chloramphenicol, at 30 min, severely reduces label uptake into all the dsRNAs and also into the medium and small mRNAs, in agreement with the results of Coplin et al. (1975). mRNA synthesis continues at a low but increasing rate and all three mRNAs have accumulated to the same extent at 30 min (Fig. 3, panel C, lanes 7-l 0). When chloramphenicol is added late in infection, at 50 min, we also detect a reduction of label incorporation into all dsRNA species (Fig. 3, panel A). However, now label incorporation into m and s mRNAs is not reduced. Synthesis of all ssRNAs continues and messages accumulate with time. lsoconformers of the small message appear to accumulate preferentially (more apparent on gels not shown). The results, which confirm and extend the observations of Coplin et al. (1975), show that chloramphenicol has a similar effect both early and late in development: label incorporation into dsRNAs is reduced and the pattern of transcripts is altered. Analysis of heat-denatured RNA samples isolated at various times after early or late addition of chloramphenicol is shown in Fig. 3 (panels D-G). The major result is that minus-strand synthesis is turned off rapidly (within 3 min) at both times. On TBE gels (panels DF) m- synthesis is detected at all times in the untreated controls (lanes l-6) but is reduced 80-85% when chloramphenicol and [3H]uracil are added simultaneously (lanes 7 and 1 l), and is absent at later labeling times (lanes 8-10 and 12-14). Unfortunately, it is difficult to evaluate s- strand synthesis because of a compression effect of abundant cold 23 S ribosomal RNA in the relevant region of the gel. However, electrophoresis of denatured RNA samples on a citrate gel (panel G) shows that I- strand synthesis also ceases rapidly after chloramphenicol addition. Other protein synthesis inhibitors mimic the effect of chloramphenicol Since chloramphenicol could interfere with minusstrand synthesis either by inhibiting synthesis of required replication proteins or by rendering message RNAs inaccessible as replication templates by fixing them to ribosomes (Dresden and Hoagland, 1966) we

66

285

also tested tetracycline and streptomycin. These antibiotics do not fix mRNA on ribosomes (Cundliffe, 1967; Luzzatto et a/., 1968; Wallace and Davis, 1973) and therefore should not affect the availability of ssRNAs as replication templates. The data in Fig. 4 show that both tetracycline and streptomycin added at 30 min mimic the action of chloramphenicol: minus-strand synthesis is shut off, overall RNA synthesis is reduced, and all mRNAs are visibly labeled in a uracil pulse at 60 min (gel of native RNA not shown). This result favors the conclusion that continuing protein synthesis is required for 46 replication. DISCUSSION We have reexamined the in viva time course of bacteriophage 46 RNA synthesis, with particular emphasis on minus strands and the effects of inhibitors of protein synthesis, to provide a rational basis for the future study of phage replication and early assembly of RNAfilled subviral particles. Our data are consistent with the asynchronous replication model proposed by Coplin et al. (1975) which suggests that mRNAs, made by strand displacement transcription, serve as templates for $6 minus-strand synthesis to form duplex molecules, The analogy with reoviruses has been extended to include coupling of minus-strand synthesis with ongoing protein synthesis. Protein synthesis inhibitors rapidly shut off 46 minus-strand synthesis and this is accompanied by an altered pattern of RNA transcripts. We believe that the altered pattern represents accumulation of plus-strand replication templates and is the result rather than the cause of the replication block. A firm conclusion is that some or all 46 replication proteins must act stoichiometrically and that replication, sorting, and packaging of $6, as for reoviruses, must involve recognition of single plus strands. This could initially involve interaction with procapsids composed of all the early proteins (Mindich and Bamford, 1988) or with a complex composed of a subset of these proteins, A brief overview of 46 RNA synthesis in viva is as follows. At 10 min postinfection a uracil pulse labels only plus strands present in all three dsRNAs and as mRNAs. Newly synthesized minus strands are first detected at 20 min and are present only in association with plus strands in dsRNA. Label may appear sequentially in the small, medium, and large minus strands. In parallel with label entering the three dsRNAs and the medium and small mRNAs, label in minus strands increases until about 60 min and then levels off. While free medium and small messages are abundant at late times the majority of newly synthesized large plus strands are found in dsRNAs.

286

PAGRATIS

AND

REVEL

I?,

B

-I 23

c L-I-MS-

FIG. 3. Autoradiograms of the time course of 46 RNA synthesis following addition of chloramphenicol early samples were labeled in rifampicin-treated @infected HBlOY at 0. 10, 20, and 30 min after CAM addition Specifically, for the initial zero time points at 30 and 50 min, l-ml samples were removed from a 20-ml master

and late in +6 development. RNA at 30 and 50 min postinfection. culture of @infected cells at 20

MINUS-STRAND 1

23

4

5

RNA

SYNTHESIS

6

II-m+-

06+ rif

te~5opghnl str4cxlp&l Exposure in hours

+

+ +

+ -

time

++ ++

+

-

120 I20 120 9

23s

+

-

-

-

-

-

120120

FIG. 4. Autoradiograms of I#J~ RNA synthesis after treatment with tetracycline or streptomycin. @6-infected HBl OY cells were treated with tetracycline (ret) (50 pglml) or streptomycin (sfr) (400 @g/ml) at 30 min postinfection. (3H]Uracil labeling was from 60 to 75 min following a 1 0-min pretreatment with rifampicin (200 pg/ml) at 30 min. RNA samples were heat denatured and electrophoresed on a TBE gel at 6.1 V/cm for 5.0 hr. + and - designations below the lanes indicate the presence and absence of $6. rifampicin, tetracycline, and streptomycin. Symbols as described in Fig. 1.

The labeling pattern at 10 min reflects the known semiconservative strand displacement mechanism of $6 transcription. It is consistent with a need to synthesize new proteins to complete an asynchronous replication process analogous to that of reoviruses, i.e., minus-strand synthesis on mRNA templates, a model suggested by the observation that ssRNAs were chased into dsRNAs (Coplin et al., 1975). Staggered initiation of labeling of different minus strands could be due to a selective mechanism that allows 46, like reovirus (Zweerink, 1974) to distinguish between different mRNA templates. Alternatively, the apparent temporal difference might simply reflect the unequal molar concentration of mRNA templates present at 20 min, re-

BY BACTERIOPHAGE

66

287

ported to be about 1 1:4: 1 for small, medium, and large messages (Coplin et al., 1975). If the minus-strand phase of 46 replication is really analogous to that of reoviruses it should also be coupled to protein synthesis. In this case inhibitors of protein synthesis would be expected to block minusstrand synthesis at both early and late times in development. This prediction is borne out by our experiments. At both times, chloramphenicol rapidly shuts off minusstrand synthesis with the consequence that label incorporation into all dsRNAs is reduced. The dsRNA molecules continue to be labeled in plus strands, however, by semiconservative transcription, and mRNAs, most notably the large message, appear to accumulate, as previously observed (Coplin et al., 1975; Sinclair and Mindich, 1976). Accumulation of mRNAs in the presence of chloramphenicol could be due either to the block in minusstrand synthesis or to stabilization of ssRNAs, or both. For the large message we believe this is due mainly to inhibition of minus-strand synthesis for the following reason. When chloramphenicol is present, compared to its absence, the increase in intensity of the I message RNA band (Fig. 3, A and B, compare lanes 5, 6 and 13, 14) is much greater than the increase in intensity of the large plus-strand band of denatured RNA samples (Fig. 3, G, compare lanes 5, 6 and 13, 14). Since the latter difference in intensities is a measure of RNA stabilization, this discrepancy can best be understood as due to an inhibition of minus-strand synthesis with a resultant accumulation of precursor template I mRNA molecules. Although a smaller fraction of the medium and small messages is expected to serve as replication templates, because of their greater abundance and the roughly equal labeling of the dsRNAs (Fig. l), we believe that accumulation of these mRNAs is caused also, in part, by the block in minus-strand synthesis. Of novel interest is the preferential accumulation of newly identified isoconformers of the small message when

and 40 min and treated immediately with rifampicin (200 pg/ml); then chloramphenicol (100 rglml) and [3H]uracil (10 &i/ml) were added at 30 and 50 min. For the remaining points 4-ml samples of infected cells were removed from the master culture at 30 and 50 min and treated with CAM immediately. At 1 0-min intervals (at 30. 40, 50 min, and at 50, 60, 70 min) rifampicin was added, followed 10 min later by [3H]uracil. For controls without CAM 1 -ml samples of the master culture were treated with rifampicin at 1 0-min intervals beginning at 20 min and [3H]uracil was added 10 min later. In all cases cells were harvested after a 15.min labeling period and extracted as described under Materials and Methods. (A, B, and C) Native RNA samples electrophoresed on a TBE gel at 4.5 V/cm for 8 hr; autoradiograms are 12., 72-, and 28%hr exposures, respectively. (D) Denatured RNA samples electrophoresed on a TBE gel at 6.1 V/cm for 5.5 hr and stained with ethidium. (E, F) Autoradiograms of the denatured RNA(D) exposed for 12 and 72 hr, respectively. (G) Denatured RNA samples electrophoresed on a citrate gel at 6.1 V/cm for 4 hr; the autoradiogram was exposed 12 hr. Symbols as described in Fig. 1. “d” marks the position of 66 RNA degradation products. Duplicate experiments gave consistent results. In this figure we present the results for native RNA (A-C) from one experiment and the denatured samples (D-G) from a repetition experiment because of gel quality. In D, E, and F, lanes 15 and 16 are RNA samples from uninfected cells with and without rifampicin. These were not heat denatured prior to electrophoresis. Ribosomal RNAs are labeled 23 S and 16 S. The intense bands near the origin are host DNA.

288

PAGRATIS

chloramphenicol is present. This may suggest a role for these molecules in recognition, sorting, or packaging. Our experiments also addressed the question why I mRNA is poorly detected with pulse label at late times despite the fact that L-segment-encoded proteins are synthesized throughout infection (Sinclair et al., 1975; Pagratis, unpublished). The strand separating gels provided an explanation. Transcription of the L segment is not reduced. Rather, with the onset of minus-strand synthesis, the vast majority of the newly synthesized I transcripts are immediately sequestered in replication complexes. Northern analysis showed ample amounts of I mRNA at late times, excluded subgenomic I mRNAs as translation templates, and suggested that early I messages may be stabilized by association with polysomes. Finally, the rate-limiting step in assembly of transcription-competent particles may not be replication. We draw this conclusion because chloramphenicol does not cause transcription to level off with the same kinetics as it shuts off replication. The amount of label incorporated into ssRNAs increases with time in the presence of drug, without a corresponding decrease in the amount of RNA degradation products (Fig. 3, panels A-C). If we exclude a direct effect of chloramphenicol on the enzymatic properties of $6 RNA polymerase and we assume that the rate of transcription is related to the intracellular concentration of transcribing par-ticles, then this phenomenon can be explained only by postulating a slow conformational change that converts complexes with replicated dsRNA to transcribing particles.

ACKNOWLEDGMENTS We thank Robert Haselkorn for critical review This work was supported by NIH Grant GM-35268

of the manuscript. to H.R.R.

REFERENCES BAMFORD. D. H.. and MINDICH, L. (1980). Electron microscopy of cells infected with nonsense mutants of bacteriophage $6. Virology 107,222-228. COPLIN, D. L., VAN ETTEN, J. L., KOSKI, R. k., and VIDAVER, A. K. (1975). Intermediates in the biosynthesis of double-stranded ribonucleic acids of bacteriophage 46. Proc. Nat/. Acad. Sci. USA 72, 849853. COPLIN, D. L., VAN ETTEN, J. L., and VIDAVER, A. K. (1976). Synthesis of bacteriophage I#J~ double-stranded ribonuclelc acid. /. Gen. Viral. 33, 509-512. CUNDLIFFE, E. (1967). Antibiotics and polyribosomes: Chlortetracycline and polyribosomes of Bacillus megaterium. Mol. Pharmacol. 3,401-411.

AND

REVEL DRESDEN, M. H., and HOAGLAND, M. B. (1966). Polyribosomes of Escherichia co/i. J. Biol. Chem. 242, 1065-l 068. EMORI, Y., IBA, H., and OKADA, Y. (1980). Semi-conservative transcription of double-stranded RNA catalyzed by bacteriophage 46 RNA polymerase. J. Biochem. 88, 1569-l 575. EMORI, Y., IBA, H., and OKADA, Y. (1982). Morphogenetic pathway of bacteriophage $6. J. Mol. Biol. 154, 287-310. LUZZA-0, L., APIRION, D., and SCHLESSINGER, D. (1968). Mechanism of action of streptomycin in E. co/i; Interruption of the ribosome cycle at the initiation of protein synthesis. Proc. Nat/. Acad. Sci. USA 60,873-880. MINDICH, L., and BAMFORD, D. H. (1988). Lipid-containing bacteriophages. In “The Bacteriophages” (R. Calender, Ed.), Vol. 2, pp. 475-520. Plenum, New York. MINDICH, L.. SINCLAIR, J. F., and COHEN, J. (1976). The morphogenesis of bacteriophage 66: Particles formed by nonsense mutants. Virology75,224-231. PAGRATIS, N., and REVEL, H. R. (1990). Detection of bacteriophage $6 minus-strand RNA and novel mRNA isoconformers synthesized in vivo and in vitro, by strand-separating agarose gels. Virology 177, 273-280. REVEL, H. R., EWEN, M. E., BRUSSLAN, J., and PAGRATIS. N. (1986). Generation of cDNA clones of the bacteriophage 46 segmented dsRNA genome: Characterization and expression of L segment clones. Virology 155,402-417. RIMON, A., and HASELKORN, R. (1978a). Transcription and replication of bacteriophage @6. Virology 89,206-2 17. RIMON, A., and HASELKORN, R. (1978b). Temperature sensitive mutants of bacteriophage 46 defective in both transcription and replication. Virology 89, 218-228. SCHONBERG, M.. SILVERSTEIN, S. C., LEVIN. D. H., and Acs, G. (1971). Asynchronous synthesis of the complementary strands of the reovirus genome. Proc. Natl. Acad. Sci. USA 68, 505-508. SILVERSTEIN, S. C., CHRISTMAN, J. K., and Acs, G. (1976). The reovirus replication cycle. Annu. Rev. Biochem. 45, 3755408. SINCLAIR, J. F., and MINDICH, L. (1976). RNAsynthesis during infection with bacteriophage $6. Virology 75, 209-2 17. SINCLAIR, J. F.. TZAGALOFF. A., LEVINE, D., and MINDICH, L. (1975). Proteins of bacteriophage $6. /. Viral. 16, 685-695. USALA, S. J., BROWNSTEIN, B. H., and HASELKORN, R. (1980). Displacement of parental RNA strands during in vitro transcription by bacteriophage 46 nucleocapsids. Cell 19, 855-862. VAN E-EN, 1. L., BURBANK, D. E., CUPPELS. D. A., LANE, L. C., and VIDAVER, A. K. (1980). Semi-conservative synthesis of singlestranded RNA by bacteriophage $6 RNA polymerase. J. Viral. 33, 769-773. VIDAVER, A. K., KOSKI, R. K., and VAN EITEN, J. L. (1973). Bacteriophage 46: A lipid containing virus of Pseudomonas phadeolicola. J. Viral. 11, 799-805. WALLACE, B. J., and DAVIS, B. D. (1973). Cyclic blockade of initiation sites by streptomycin-damaged ribosomes in Escherichia co/i: An explanation for dominance of sensitivity. J. Mol. Biol. 75, 377-390. WATANABE, Y., KUDO, H.. and GRAHAM, A. F. (1967). Selective inhibition of reovirus ribonucleic acid synthesis by cycloheximide. /. Viral. 1, 36-44. WEISSMANN, C., FEIX, G., and SLOR, H. (1968). /n vitro synthesis of phage RNA: The nature of the intermediates. Co/d Spring Harbor Symp. &ant. Biol. 33,83-l 00. ZARBL, H., and MILLWARD, S. (1983). The reovirus multiplication cycle. In “The Reoviridae” (W. K. Joklik, Ed.), pp. 107-l 96. Plenum, New York. ZWEERINK, H. (1974). Multiple forms of ss to dsRNA polymerase activity in reovirus-infected cells. Nature (London) 247, 313-315.