Characterization of vesicular stomatitis virus replicating complexes isolated in renografin gradients

VIROLOGY

99,

'75-83 (1979)

Characterization

VIRGINIA

of Vesicular Stomatitis Isolated in Renografin

M. HILL,

Department of Cellular,

CHRISTIAN Viral,

Virus Replicating Gradients

Complexes

C. SIMONSEN AND DONALD F. SUMMERS

and Molecular Biology, College of Medicine, Salt Lake City, Utah 84132

University

of Utah,

Accepted July 20, 1979

Newly synthesized products of replication and transcription from vesicular stomatitis virus (VSU)-infected HeLa cells have been separated on Renografin gradients and characterized. The replicative products have been identified by the following criteria: (1) sucrosevelocity gradients and methyl-mercury agarose gel analyses show RNA of sizes between 18 S and genome size 42 S. (2) This RNA is predominantly minus stranded by hybridization analyses. (3) A majority of the nascent RNA molecules of less than half genome size from replicative complexes are complementary to the 3’ end of the VSV plus strand, i.e., the L protein cistron. These short nascent RNA species are nuclease resistant and thus are presumed to be contained in ribonucleoprotein complexes.

direct evidence as to the nature of the template is not available. Vesicular stomatitis virus (VSV) is an During the VSV infectious cycle, at a time animal virus with an enveloped negative when VSV-specific RNA synthesis is single-strand RNA. Contained in the maximal, about 90% of the RNA species genome, whose molecular weight is about being synthesized in the infected cell are the 3.6 x 106, is the information necessary to products of transcription, i.e., the five code for five virus-specific proteins: G and species of VSV mRNAs (Wertz and Levine, M, the envelope associated proteins; and N, 19’73; Batt-Humphries et al., 1979). The NS, and L, those proteins located in the remaining 10% represents the full length virion ribonucleoprotein complex (Wagner, (42 S) RNA molecules, both plus and minus 1975). The L and NS proteins are com- strands. Early in infection the ratio of newly ponents of the RNA-dependent RNA made full length plus strand to minus strand transcriptase which, using the RNA ge- is about 40:60, while later in infection the nome-N protein complex as a template, ratio changes to 20:80 (Wertz, 1978; Simonsynthesizes the five VSV-specific messen- senet al., 1979a). The total intracellular pool ger RNA (mRNA) species in viwo and in of RNP particles should contain transcribvitro (Banerjee et al., 19’77; Emerson and ing complexes, RNP particles containing Yu, 19’75; Naito and Ishihama, 1976). only minus strands which are to be assemReplication of the VSV genome (synthesis bled into virus particles, and replicating of virion-length plus and minus stranded complexes synthesizing 42 S RNA moleRNA species) occurs in viva, and it has been cules (if the replicating complexes are presumed that the template was a ribo- RNPs). nucleoprotein (RNP) complex similar to Transcribing complexes consist of a that used for transcription (Wertz, 1978; minus strand template (42 S) upon which Soria et al., 1974). Studies on the mecha- the mRNA species are synthesized. Replinism of VSV replication have been ham- cating complexes should contain either a pered by the lack of an in vitro system plus or minus stranded 42 S RNA molecule, capable of synthesizing virion RNA, and each serving as a template for complementary genome sized RNA. 1 To whom reprint requests should be addressed. INTRODUCTION

75

0042-6822/79/150075-09$02.00/O Cspegbt All rights

b 19’79 by Academic Press, Inc. of reproduction in any form reserved.

‘76

HILL,

SIMONSEN,

Using short exposures to radiolabeled RNA precursors, we have preferentially labeled nascent transcriptive and replicative products and have separated them on Renografin gradients. We have characterized the replicating complexes; the data indicate that the templates are RNP complexes. MATERIALS

AND METHODS

Cells, virus, infection,

and radiolabeling

of VSV RNA. HeLa S3 cells and VSV Indiana, were grown as previously described (Hunt and Summers, 1976). Short-term labeling with [3H]uridine and [3H]adenosine (New England Nuclear) of infected cells was carried out as detailed in the figure legends. RNA extractions. Appropriate samples from the Renografin gradients were pelleted and either resuspended in NETS buffer (100 mM NaCl, 1 m&f EDTA, 10 mM Tris-HCl pH 7.4, and 0.2% SDS) for further analyses on NETS-sucrose gradients or phenol extracted and then ethanol precipitated for RNA-RNA hybridization experiments or methyl-mercury agarose gel electrophoretic analyses. VSV-specific mRNA was isolated from infected cells at 4.5 hr post infection (p.i.) and purified through oligo (dT) cellulose which removed most of the rRNA. The RNA was then centrifuged in a sucrose gradient and the greater than 28 S and less than 18 S RNAs were pooled separately. This process was repeated twice; the greater than 28 S RNA pool was used as a source of 31 S mRNA and the less than 18 S RNA as a source of the 12-18 S mRNAs. The OD 260 of each preparation was recorded, although the bulk of the RNA in each preparation was 28 S or 18 S rRNA and therefore the amount of unlabeled RNA added to the hybridization mixtures reflects a mixture of the VSV mRNAs (31 or 12-18 S), cell mRNAs, and rRNA. VSV minus strand 42 S RNA was purified from phenolextracted virions and ethanol precipitated. Methyl-mercury agarose gel electrophoresis. The method for methyl-mercury

gels was a modification of that described by Bailey and Davidson (1976) using horizontal slab gels of 1% agarose and 5 mil4 methyl-

AND SUMMERS

mercury hydroxide (Alfa Products). The electrode buffer was recirculated at 80 ml/ hr; the samples were subjected to electrophoresis at 75 V, 45 mA for 7-9 hr. After electrophoresis, the gel was fixed in 10% acetic acid and dehydrated in methanol by two immersions of 1 hr each. At this step, the slab gel was dried down to less than 0.5 mM thickness under vacuum and the thin slab was soaked in a solution of 10% (w/w) 2,5-diphenyloxazole (PPO) in methanol for 3 hr at room temperature and rinsed in water for 15 min to precipitate the fluor. After blotting the excess H20, the gel was sealed in plastic wrap and subjected to autoradiography with Kodak X-Omat film at -70”. All operations were performed under a hood because of the toxic and volatile nature of methyl mercury. RESULTS

Presumably, by pulse-labeling VSVinfected cells with radioactive RNA precursors for short periods of time, much of the radioactivity would be incorporated into RNA molecules being synthesized by transcriptive and replicative complexes. If one could identify RNP particles associated with radiolabeled nascent RNA species of heterogeneous sizes, 42 S and smaller, and with a predominance of minus-strand sense, then one could assume that these structures were involved in replication. We pulse-labeled VSV-infected cells at 5 hr p.i., a time when the rate of VSV replication is maximal (Simonsen et al., 1979a) and analyzed the cytoplasmic RNP species on Renografin gradients. Figure 1A shows three predominant peaks of radioactivity from a VSV-infected cell cytoplasm, labeled 3 to 5 hr p.i., on a Renografin gradient. A similar pattern of radioactivity is seen after a 5-min pulse at 5 hr p.i., except that peak I (fractions 15-20) is more heterogeneous (Fig. 1B). In some experiments this 5-min pulse peak I resolved into two or more distinct peaks of radioactivity. When the entire peak I from the 2-hr label was pooled, the RNA released with SDS and analyzed on NETS-sucrose gradients, it is clearly shown (Fig. 2A) that the radiolabeled peak contained full length (42 S)

VSV REPLICATING

I

I 20

COMPLEXES

77

I 40 FRACTION NUMBER

FIG. 1. Renografin gradients of VSV-infected cytoplasmic extract. HeLa cells (2 x 1O8)in 20 ml MEM (serum free) were infected with VSV at 10 PFU/cell at 37”. At 1 hr pi., serum and actinomycin D were added to concentrations of 5% and 1 pg/ml, respectively. At 3 hr p.i. the culture was divided in half and radiolabeled as described below: (A) Two-hour pulse: 0.5 mCi [3H]adenosine and 0.5 mCi [3H]uridine were added and the infection continued for 2 hr. (B) Five-minute pulse: The infection was continued for 2 more hr at which time (5 hr p.i.) 5 mCi [3H]uridine and 5 mCi [3H]adenosine, made isotonic with 10x Earle’s salts and warmed to 37”C, were added for 5 min. At 5 hr PI, both cultures were stopped by pouring the infected cells over frozen MEM and centrifuged immediately for 5 min at 1000rpm, using an IEC 267 rotor. The cells were washed once in Earle’s salts and resuspended in 2-4 ml cold NT buffer (0.1 M NaCl, 0.01 M Tris-HCl pH 7.4), NP40 (Shell) was added to a concentration of 1% to lyse the cells. After 5 min on ice, the nuclei were pelleted by centrifugation at 1500rpm (IEC 267 rotor) for 3 min. The supernatant was removed and the nuclei resuspended in l-2 ml of NT buffer. A mixture of NP40 and sodium deoxycholate was added to a final concentration of 0.3 and 0.2%, respectively. The nuclei were again pelleted, and the original and nuclear-wash supernatants were combined and layered onto Renografin gradients (Renografin 76, Squibb). Preformed linear gradients were made in 36-ml cellulose nitrate tubes and contained 15-47% (v/v) Renografin in NT buffer. Gradients overlayed with cytoplasmic extracts were centrifuged at 23,000 rpm for 16 hr at 4” in a SW27 rotor (Beckman). Gradients were collected in 60 fractions by pumping from the tube bottom. Renogratin is precipitable by trichloroacetic acid, perchloric acid, and cetyltrimethylammonium bromide, and therefore radioactivity was assayed by adding 50 ~1samples directly to a toluene-based scintillation fluid (New England Nuclear). Sedimentation is from right to left in this and all subsequent figures. The radioactivity present in the very bottom of the gradient in Fig. 1B represents trapped soluble radiolabel and is not recoverable by centrifugation.

RNA species which cosedimented with virion 42 S RNA marker and essentially no smaller mRNA-sized RNA species were found in this peak. However, peak I from the gradient of 5-min labeled cytoplasm (Fig. 1B) clearly contains a heterogeneous population of radiolabeled RNA molecules with sizes from 42 S down to 18 S (Fig. 2B). Figure 3B is the portion of a Renografin gradient of a 5-min pulse-labeled VSV infected cytoplasm which contains peak I. Individual fractions of the radioactive peak shown in 3B (this peak corresponds to peak I, Fig. 1B) were analyzed in NETSsucrose gradients. The slower sedimenting

fractions from the Renografin gradient seem to contain shorter nascent RNA molecules (Figs. 2D and E). To rule out the possibility that the observed spectrum of labeled RNA species obtained from the peak (fractions 15-25) in Fig. 3B were largely or partly due to double-stranded species (Wertz, 1978), fractions from this Renografin gradient were subjected to electrophoresis on denaturing methyl-mercury agarose gels (Fig. 3A). It can be seen (Fig. 3A, lanes B, C, and D) that the three fractions selected from the slower sedimenting shoulder of the peak (Fig. 3B, fractions 18-26) contained

HILL, SIMONSEN, AND SUMMERS

42s t

28s t

1% t

'51'E'

42s t

28s t

IBS t

FRACTION NUMBER

FIG. 2. Velocity sedimentation in sucrose gradients of [“HIRNA extracted from Renografin gradients. Samples from Renografin gradients were diluted in NT buffer, pelleted for 2 hr at 48,000 rpm in a SW50.1 rotor, and resuspended in NETS buffer with 1% SDS. [“‘C]Uridine-labeled VSV virion (42 S) RNA and HeLa cell 28 and 18 S ribosomal RNA were added as markers. The samples were centrifuged on sucrose (15-30% w/w) NETS gradients for 16 hr at 20,000 rpm at 22” in a SW41 rotor. Gradients were collected by puncturing the bottom of the tubes; l&drop fractions were dripped onto GFB Whatman filters which were washed successively with 10% tricholoroacetic acid (TCA), two 5% TCA rinses, and dehydrated with 95% ethanol. The filters were then assayed for radioactivity. In each case approximately 20% of the pools obtained from Fig. 1 were analyzed. (A) Pool of fractions 12-18 from Renografin gradient 2 hr label shown in Fig. 1A. (B) Pool of fractions 12-22 of the 5-min pulse shown in Fig. 1B. (C) Peak, fraction 16 of the 5-min pulse shown in Fig. 3B. (D) Middle shoulder, fraction 18 of the 5-min pulse shown in Fig. 3B. (E) Light shoulder, fractions 20-21 of the .5-minpulse shown in Fig. 3B. [Yluridine-labeled marker RNA (0). [3H]Uridine-labeled RNA from Renografin gradients (01.

RNA species of sizes less than 42 S and of progressively smaller sizes. The peak fractions contained 42 S RNA species and RNA with sizes ranging from 42 S down to smaller than 18 S (Fig. 3A, lane A). Also, treatment of fractions from NETS-sucrose gradients, similar to that shown in Fig. 2B, with pancreatic RNase revealed no RNaseresistant species throughout the gradient, further indicating an absence of doublestranded RNA species in the RNA extracted from Renografin peak I (data not shown). It should be noted that a distinct band migrating at “26 S,” which is minus stranded (see below), is seen in the methyl-mercury gel (Fig. 3A, lanes A and 0. The nature of

this band is unclear at this time. It is of interest that an in vivo transcribing system utilizing cytoplasmic VSV RNPs synthesizes an RNA species of similar size (BattHumphries et al., 1979). Furthermore, multiple discrete bands smaller than 26 S are seen in Fig. 3A, lane A. Since at 5 hr p.i. in the replicative cycle 80% of the replicating templates should be synthesizing minus strand sequences, we tested by RNA-RNA hybridization to see if the polarity of the putative nascent RNA molecules isolated in Renografin gradients was predominantly negative stranded. As shown in Fig. 4A, under our conditions of hybridization, we could protect 100% of labeled virion 42 S RNA with a mixture of

VSV REPLICATING

COMPLEXES

79

42s

-

1% -

I

1

IO

FRACTION

20 NUMBER

I

30

FIG. 3. Methyl-mercury agarose gel analysis of the nascent RNA molecules extracted from a Renografin gradient. (A) Methyl-mercury agarose slab gel. Lane (A): RNA extracted from fractions 15-17 in 3B. (B): RNA extracted from fractions 18-19 in 3B. (C): RNA extracted from fractions 20-22 in 3B. (D): RNA extracted from fractions 24-26 in 3B. The recovery of RNA was not quantitative in this experiment. (B) Renografin gradient of a cytoplasmic extract from VSV-infected cells pulsed 5 min at 5 hr p.i. The conditions for separation were identical to those of Fig. 1B.

all five VSV mRNA species, and could protect about 50% of the virion RNA with mRNA enriched for either the 31 S L-protein mRNA (complementary to the 5’-half of the VSV genome) or the 12-18 S mRNA species (G, M, NS, and N mRNA species) (Banerjee et al., 1976). When the peak fraction from peak I (see Fig. 1B) in a Renografin gradient containing a 5-min pulse-labeled infected cell extract was hybridized to an excess of cold total VSV mRNA or to excess cold virion RNA (Fig. 4B), it can be seen that almost 100% of the RNA species could be protected with the total mRNA, whereas the virion RNA afforded only 15% protection. This indicated that approximately 15% of the labeled

nascent RNA was (+> strand and that about 85% was (-> strand. These values agree with previous reports on the percentages of (+) and (-) full length RNA in the cell at this time (Soria et al., 1974). Due to the presence of unlabeled template strands in peak I, addition of the unlabeled mRNA resulted in 100% protection and not 85% protection of the radiolabeled RNA in Fig. 4B, as might otherwise be expected. If the shorter nascent strands (see lanes C and D, Fig. 3A) were primarily minus strands which were less than 50% completed, they should be mainly composed of sequences from the L protein cistron which is known to be located at the 5’ end of the virion RNA genome (Banerjee et al., 19’7’7;

HILL,

80

SIMONSEN,

Ball and White, 19’76). We tested this hypothesis by separating the 31 S L mRNA from the 12-18 S VSV mRNAs and annealing these two mRNA size classes to RNA obtained from the pooled light shoulder of a 5-min pulse-labeled peak I from a Renografin gradient. Figure 4C shows that the nascent RNA was protected by the sequences enriched for the 31 S mRNA species to a much greater degree than by

AND SUMMERS

that of the 12-18 S mRNAs. The presence of unlabeled template strands of both polarities, and the fact that the nascent RNA sample contained some species larger than half-genome size, resulted in the 12- 18 S mRNA probe protecting about 50% of the sample and the 31 S mRNA probe protecting approximately all of the sample. However, it is clearly shown in Fig. 4C that the majority of the short heterogeneous

5 .e s B d gB C. a-” IOOSO-

60-

40-

FIG. 4. Hybridization of pulse-labeled RNA from Renografin gradients. Labeled RNA was phenol extracted and ethanol precipitated from a Renografin gradient similar to that in Fig. 3. Aliquots of the labeled RNA (approximately 5 ng of specific activity 600,000 cpm/pg) were mixed with increasing amounts of unlabeled VSV-specific RNA heat denatured for 1 min at 115”, and hybridized for 3 hr at 73 using the conditions of Kolakofsky (19’76). The radiolabeled RNA is: 3H-labeled 42 S RNA from virions (A); [3H]RNA from a pool of Renografin isolated RNPs similar to that in Fig. 2, Lane A, (B); [3H]RNA from a sample similar to Fig. 3, Lane C, (Cl; [3H]RNA isolated from peak II (D). The unlabeled RNA which was hybridized to each of these samples is: VSV genome 42 S RNA (m); 12-18 and 31 S mRNA mixture (0); 12-18 S mRNA (Cl); 31 S mRNA (0).

VSV REPLICATING

pulse-labeled RNA molecules contained sequences complementary to the L mRNA, and were thus products of replication. All of the Renografin gradients also contained a radiolabeled peak (peak II) at a lesser density than peak I (see Fig. 1). Since at midcycle of infection approximately 90% of the VSV-specific RNA present represents transcripts of virus mRNA and these RNA species are not represented in the Renografin peak I (Figs. 4A-C), we concluded that the VSV transcriptive products must be in peak II or at the top of these Renografin gradients. To test this, peak II was pooled, the RNA was extracted and hybridized. As shown in Fig. 4D, this RNA was essentially 100% plus strand, and therefore likely to be mRNA. Pancreatic ribonuclease treatment has shown that this RNA population was 100% sensitive to RNase, as would be expected for free mRNA or mRNP complexes (Soria et al., 19’74)(data not shown). Methyl-mercury gel analyses have also shown that peak II from a 5-min labeled extract contains RNA molecules heterogeneous in size, ranging from 4 to 20 S (data not shown). Similar

I IO

, 20 FRACTION

COMPLEXES

81

results were obtained with RNA pooled from the very top of these gradients which was shown to be identical to the VSV mRNA species (data not shown). Thus far, the data indicated that peak I in Renografin gradients of cytoplasm from 5min pulses of VSV-infected cells contain nascent minus-stranded RNA molecules. To establish whether virus-specific RNP proteins (L, NS, and/or N) became associated with the growing 42 S RNA strands during their synthesis or after their completion, an infected culture was pulselabeled for 5 min as above, a cytoplasmic extract was prepared and divided into halves, and one half was treated with pancreatic ribonuclease. Both the nuclease treated and untreated samples were then analyzed in Renografin gradients. Figure 5 shows that the labeled RNA species in peak I (fractions 15-30) were completely nuclease resistant, and also shows that a large portion of the radiolabeled material at the top of the gradient (released mRNA species) was nuclease sensitive and could be seen as such, despite the large amount of soluble radioactive uridine and adenosine in

I 30 NUMBER

I 40

1 50

FIG. 5. Ribonuclease-treated radiolabeled cytoplasmic extract fractionated on Renografin gradients. One half of a cytoplasmic extract from VSV-infected cells, pulsed 5 min at 5 hr p.i., was treated with 10 pg/ml of RNase for 30 min at 20”. Both extracts (treated and untreated) were then layered onto Renografin gradients and were centrifuged, using conditions given in Fig. 1. Assays of the radioactivity in the gradients were superimposed in this figure: RNase treated (0); untreated (0).

HILL,

82

SIMONSEN,

the top fractions of the gradient. The fact that the heterogeneous minus-stranded radiolabeled RNA molecules in peak I are ribonuclease resistant, as are mature RNPs extracted from infectious virus particles or from infected cell cytoplasm (Soria et al., 1974) indicates that these partially replicated RNA strands are probably closely associated with VSV protein(s) shortly following their initiation. Recent studies have also shown that these nascent RNP molecules have a density in CsCl gradients identical to that of virion RNP complexes (Simonsen et al., 1979b).We are now examining the protein(s) which rapidly associate with these nascent replicating RNA molecules. DISCUSSION

In this study we have been able to partially isolate and characterize [3H]uridine and [3H]adenosine pulse-labeled RNP complexes from VSV-infected cells. One of the peaks (peak I) in the Renografin gradients contains labeled RNA that is less than 18 S to 42 S in size. Almost all of this radiolabeled RNA is minus stranded. The nascent RNA that is less than 50% full length is complementary to the L mRNA; since replication is initiated at the 3’ end of the plus strand template (the 3’ end of the plus strand contains the L protein cistron) and replication must proceed in the 5’ to 3’ direction on the progeny minus strand, these RNA sequences meet a basic requirement for being partially replicated minus stranded RNA. As replicating RNA strands are initiated and elongated, they become rapidly complexed with protein(s) which render these nascent RNA molecules resistant to nuclease digestion. The data in Fig. 2 and our analyses of these RNA species with respect to nuclease sensitivity clearly show that there are no detectable double-stranded molecules in peak I. Since these nascent RNAs are nuclease resistant and band in CsCl at a density equal to that of virion RNPs, it is likely that the proteins on the newly made RNAs are VSV specific, i.e., L, NS and/or N. Contained in peak I (Figs. 1A and B) are minus stranded particles destined to be-

AND SUMMERS

come mature virions, minus strand and plus strand structures that are replicating, and probably the minus strand transcriptive template RNPs. Our results provide evidence that the template for VSV replication is an RNP complex similar to that involved in VSV transcription. It will be of importance to determine by what mechanism RNAs containing plus and minus strands are shunted to the pool of replicating or transcribing RNPs or to the pathway for virion assembly. The Renografin gradients used to separate the replicating complexes, for unknown reasons, cause transcriptive complexes to dissociate and release their nascent mRNA chains. This points out that a structural difference must exist between replicative and transcriptive RNP complexes. Past observations (Birnie et al., 1973; Chan and Schifler, 19’74;Hutterman and WendlbergerSchieweg, 1976) and recent studies in our laboratory with Renografin gradients (unpublished results) have shown that the basis for separation of protein-nucleic acid complexes in this medium is an approach to equilibrium in that some molecular species such as smaller RNP complexes are still being separated on the basis of velocity. We therefore cannot state with any certainty why peak I, containing 5-min pulselabeled RNA, has full length 42 S RNA molecules plus heterogeneous-sized nascent molecules down to less than 18 S in size, and the slower sedimenting fractions contain nascent RNA chains of progressively smaller sizes. This observation might imply that some templates contain multiple nascent RNA chains which have somehow been initiated synchronously and, therefore, contain one or more RNA chains of a defined size class which are separable in Renografin. It is also possible that some templates contain a single nascent strand or that some of the nascent strands have dissociated from their templates. These results show that the nascent RNA molecules extracted from the replicative templates (see Fig. 3A) seem to occur as discrete size classes with a rather noticeable accumulation of a “26 s” species (Fig. 3A, lanes A and C). This may indicate that the VSV replicase polymerizes at a

VSV REPLICATING

nonuniform rate with a major pause site at about 50% of the genome, i.e., the L cistron. A similar phenomenon has recently been reported for the Qp polymerase (Mills et al., 1978) and had been observed in reverse transcriptase reactions as well (Elstratiades et al., 1976; Haseltine et al., 1976). The finding that intracellular VSV replicating complexes can now be isolated in Renografin gradients, coupled with the recent observation that an in vitro transcribing system derived from a VSVinfected cell extract will synthesize full length (42 S) RNA (Batt-Humphries et al., 1979) now makes it feasible to develop an in vitro VSV replication system with which we can study the precise mechanisms of replication and packaging of newly synthesized 42 S RNA molecules. ACKNOWLEDGMENTS We thank Dr. Daniel Kolakofsky for helpful discussions and advice. This work was supported by Grants NIH AI12316-04 and NSF PCM77-17876AOl. C.S. is a recipient of a University of Utah Graduate Research Fellowship. REFERENCES BAILEY, J. M., and DAVIDSON, W. (1976). Methylmercury as a reversible denaturing agent for agarose gel electrophoresis. Anal. Biochem. 70, 75-85. BALL, L. A., and WHITE, C. N. (1976). Order of transcription of genes of vesicular stomatitis virus. Proc. Nat. Acad. Sci. USA 73, 442-446.

BANERJEE, A. K., ABRAHAM, G., and COLONNO,R. S. (1977). Vesicular stomatitis virus: Mode of transcription. J. Gen. Viral. 34, 1-8. BAIT-HUMPHRIES, S., SIMONSEN, C., and EHRENFELD, E. (1979). Full length viral RNA synthesized in vitro by vesicular stomatitis virus-infected HeLa cell extracts. Virology 96, 88-99 (1979). BIRNIE, G. D., RICKWOOD,D., and HELL, A. (1973). Buoyant densities and hydration of nucleic, acids, proteins, and nucleoprotein complexes in metrizamide. Biochim. Biophys. Acta 453, 176-184. CHAN, R. T. L., and SCHIFFLER, I. E. (1974).

COMPLEXES

83

Isopynic centrifugation of chromatin in Renografin solutions. J. Cell Biol. 61, 780-788. ELSTRATIADES,A., KOFATOS,F. C., MAHAN, A. M., and MANIATIS, T. (1976). Enzymatic in vitro synthesis of globin genes. Cell 7, 279-288. EMERSON,S. U., and Yu, Y. H. (1975). Both NS and L proteins are required for in vitro RNA synthesis by vesicular stomatitis virus. J. Viral. 15, 13481356. HASELTINE, W. A., KLEID, D. G., PANET, A., ROTHERNBERG, E., and BALTIMORE, D. (1976). Ordered transcription of RNA tumor virus genomes. J. Mol. Biol. 106, 109-131. HUNT, L. S., and SUMMERS,D. F. (1976). Association of vesicular stomatitis virus proteins with HeLa cell membranes and released virus. J. Viral. 20, 637645. HUTTERMAN, A., and WENDLBERGER-SCHIEWEG,G. (1976). Studies on metrizamide-protein interactions. Biochim. Biophys. Acta 453, 176-184. KOLAKOFSKY, D. (1976). Isolation and characterization of Sendai virus DI-RNAs. Cell 8, 547-555. MILLS, D. R., DOBKIN, C., and KRAMER, F. R. (1978). Template-determined variable rate of RNA chain elongation. Cell 15, 541-550. NAITO, S., and ISHIHAMA, A. (1976). Function and structure of RNA polymerase from vesicular stomatitis virus. J. Biol. Chem. 251, 4307-4314. SIMONSEN,C., BATT-HUMPHRIES, S., and SUMMERS, D. F. (1979a). RNA synthesis in vesicular stomatitis virus infected cells: In viva regulation of replication. J. Virol. 31, 124-133 (1979). SIMONSEN, C., HILL, V. M., and SUMMERS, D. F. (197913).Further characterization of the replicative complex of vesicular stomatitis virus. J. Viral., 31, 494-505 (1979). SORIA, M., LITTLE, S. P., and HUANG, A. S. (1974). Characterization of vesicular stomatitis nucleocapsids. I. Complementary 40s RNA molecules in nucleocapsids. Virology 61, 270-280. WAGNER, R. R. (1975). Reproduction of rhabdoviruses. In “Comprehensive Virology” (H. FraenkelConrat, and R. R. Wagner, eds.), Vol. 4, pp. l-93. Plenum, New York. WERTZ, G. W. (1978). Isolation of possible replicative intermediate structures from vesicular stomatitis virus-infected cells. Virology 85, 271-285. WERTZ, G. W., and LEVINE, M. (1973). RNA synthesis by vesicular stomatitis virus and a small plaque mutant: Effects of cycloheximide. J. Viral. 12, 253-264.