JOURNAL OF VIROLOGY, June 1969, p. 599-64 Vol. 3, No. 6 Copyright 1969 American Society for Microbiology Printed in U.S.A. Sindbis Virus-induced Viral Ribonucleic Acid Polymerasel T. SREEVALSAN' AND FAY HOH YIN Central Research Department, Experimental Station, E. I. du Pont de Nemours and Company, Wilmington, Delaware 19898 Received for publication 11 March 1969 A cytoplasmic structure containing the viral ribonucleic acid (RNA) polymerase has been isolated by sucrose density centrifugation from cells infected with Sindbis virus. Uninfected cells did not contain any such structure. Preliminary experiments indicated that the structure may be associated with membranes. This structure incorporated 3H-guanosine triphosphate in vitro in the absence of added template. The RNA synthesied in vitro by the enyme consisted of single-stranded 4S RNA, the ribonuclease-resistant replicative form, and possibly the replicative intermediate form of viral RNA. The products formed in vitro by the enyme are identical in sedimentation rates to those formed in the infected cells in vivo. The isolation of a virus-specific ribonucleic acid (RNA) polymerase from cells infected with arboviruses, a group of RNA-containing animal viruses, has been reported (6, 7). However, only one of these reports describes the nature of the product synthesied by the enyme in vitro. Martin and Sonnabend (7) demonstrated that the product synthesied by their preparation consisted of only the ribonuclease-resistant RNA; repeated attempts failed to demonstrate the synthesis of the single-stranded RNA. This communication describes some preliminary results regarding the isolation of a particulate fraction of cytoplasm from chick embryo (CE) cells infected with Sindbis virus, an arbovirus. This fraction was found to contain viral RNA polymerase activity. The products synthesied in vitro by the enyme-containing structure consisted of (i) single-stranded viral RNA, (ii) the replicative form, and possibly (iii) the replicative intermediate of the Sindbis virus RNA. MATERIALS AND METHODS Cells and virus. Primary cultures of CE cells were prepared and grown in Eagle medium with 3% calf serum (1). The source and preparation of Sindbis virus and the methods used to infect cells and to label viral RNA were identical to those described previously (1). The CE cultures were incubated with 3 ml of Eagle medium containing actinomycin D (5,g/ml) for 2 hr at 37 C and were then infected with the virus. Preparation of cytoplasmic extract. The infected cells were harvested in the cold and pelleted. The cell pellet was suspended in 2 ml of cold buffer consisting of.1 M tris(hydroxymethyl)aminomethane (Tris) at ph 7.2 with 1-4 M sodium ethylenediaminetetraacetate (EDTA). The cells were allowed to swell for 3 min at 4 C, and then they were ruptured by using 18 strokes in a tight-fitting stainless-steel Dounce homogenier. Unbroken cells and nuclei were removed from the disrupted cell suspension by centrifugation at 8 X g for 5 min. Centrifugation of cytoplasmic extract. The sucrose density gradient used for analysis of cytoplasmic extract was formed in a Spinco SW25.1 tube by successively layering 2.5 ml of 5%, 2.5 ml of 45%, 2.5 ml of 43.5%, and 2 ml of a linear gradient of 4 to 15% sucrose in the same buffer used for lysing the cells. Centrifugation was at 6, X g for 2 hr at 4 C. The sucrose concentration in each fraction of the gradient was calculated from its refractive index. The refractive index was determined by using a Bausch & Lomb refractometer. Enyme assay. An.2-ml amount of enyme was used for assaying polymerase activity. It was incubated at 37 C for 3 min with.2 ml of a mixture containing the following constituents: 1,uc of 3H-guanosine triphosphate (GTP), specific activity, 1.3 c/ mmoles;.5,moles of the 5'-triphosphates of adenine, cytodine, and uridine (ATP, CTP, and UTP, respectively); 2,moles of Tris buffer (ph 8.); 2,moles of MgC12;.5 jgmoles of phosphoenolpyruvate;.1 mg of phosphoenolpyruvate kinase; 2,ug of actinomycin D; and 7,moles of 2-mercaptoethanol. The reaction was terminated by adding.5 ml of ice-cold.1 M Na4P O7 and 2 ml of.5 N perchloric acid. The pre- 1 Contribution no. 1556 from the Central Research Department, E. I. du Pont de Nemours and Co. cipitate was collected on a filter, washed exhaustively 2Present address: Department of Microbiology, Georgetown five times with cold 1% trichloroacetic acid, and University Medical School, Washington, D.C. 27. counted in a scintillation counter. 599
6 SREEVALSAN AND YIN J. VIROL. Assay for protein. This was determined by the method of Lowry et al. (5) with correction for the content of sucrose in the samples (3). Analysis of the RNA synthesied in vitro. The enyme fraction was incubated with an equal volume of reaction mixture and 3H-GTP at 37 C. At the end of the desired incubation period, sodium dodecyl sulfate (SDS) was added to the reaction mixture to a final concentration of 1%; the mixture was then incubated for an additional 1 min. The RNA was precipitated with 2.5 volumes of alcohol. The precipitate was dissolved in buffer containing.1 M Tris,.1 M NaCl,.1 M EDTA, and.5% SDS and centrifuged directly on a 15 to 3% sucrose gradient, made of the same buffer, in an SW5 rotor at 1, X g for 2.5 hr. RNA was treated with pancreatic ribonuclease at a concentration of 1,ug/ml for 1 min at 37 C in the presence of.3 M NaCl. SDS present in the sample of RNA was removed prior to incubation with the enyme by precipitation with potassium chloride at 4 C. Viral RNA was precipitated with lithium chloride by methods similar to those described by Plagemann and Swim (9). Reagents. Actinomycin D was purchased from Mann Research Laboratory, Inc., New York. N.Y.; 3H-uridine (24 c/mmoles) was purchased from Nuclear-Chicago Corp., Des Plaines, Ill. 3H-GTP (1.3 c/mmoles) and unlabeled ATP, CTP, and UTP were purchased from Schwar BioResearch, Inc., Orangeburg, N.Y. Bovine pancreatic ribonuclease, pyruvate kinase (rabbit muscle), and 2-phosphoenolpyruvate were purchased from Calbiochem, Los Angeles, Calif. RESULTS Localiation of the site of viral RNA synthesis in infected cells. A cytoplasmic extract was prepared from cells that had been exposed to 3H-uridine for 5 min at 5 hr after infection, the time of maximal viral RNA synthesis. The cytoplasmic extract was analyed by sedimentation on a sucrose density gradient (Fig. 1). The acid-precipitable radioactivity shows a very broad distribution. There are several faster-sedimenting structures besides the peak of radioactivity sedimenting slightly slower than the ribosomes. The latter peak of radioactivity was similar to the 65S peak reported earlier (1). The peaks at 3 to 4% sucrose concentration contained about 4% of the total radioactivity. The following experiment was performed to determine whether these structures were virus-specific polyribosomes or a viral RNA replication complex as described for polio and Mengo viruses (4, 8). The cytoplasmic extract was prepared from infected cells and divided into three equal parts. Part of the cytoplasmic extract was incubated with either.5% sodium deoxycholate or pancreatic ribonuclease (1,g/ml) before centrifugation on sucrose density gradients (Fig. 2). Even after incubation with ribonuclease, cli x U 'ci L-) 25 =: 2O 15F le 5 F 1 2 3 FIG. 1. Sucrose density gradient analysis of cytoplasmic extract from cells infected with Sindbis virus. At 5 hr after virus was added to CE cultures which had been previously incubated with actinomycin D, 2 mc of 3H-uridine (2 clmmoles) was added to 5 X 17 cells. The cells were allowed to incorporate radioactivity for 5 min, at the end of which they were chilled to 4 C in I min. The cells were harvested in the cold. Cytoplasmic extract was prepared anid analyed on a preformed sucrose gradient. Centrifugation was performed with ani S W25.1 rotor at 6, X g for 2 hrat 4 C. One-milliliter fractions were collected and used for anialying radioactivity and sucrose concentration. Trichloroacetic acidinsoluble radioactivity in each fraction was determined. Sucrose concenitration in each fraction was calculated from its refractive index. () 3H-uridinie; () coiicelltration of sucrose. about 25%o of the total radioactivity was still associated with the faster-sedimenting structures. However, after incubation with deoxycholate, only about 5% of the total radioactivity was still associated with similar structure. These results indicate that, unlike polyribosomes, the fastersedimenting structure was only partially destroyed by ribonuclease. It may be attached to membranes as indicated by its sensitivity to sodium deoxycholate. Properties of the RNA polymerase induced by Sindbis virus. Since the above structure contained nascent viral RNA, it might also be expected to contain viral RNA polymerase. The following experiments were performed to localie the viral polymerase. Cytoplasmic extracts from infected cells were prepared and analyed in sucrose density gradients as described above. Fractions from the sucrose gradients were assayed for polymerase activity by using 3H-GTP. Figure 3 5 4 w 3 U) cr. 2 D 1n 1 8
VOL. 3, 1969 SINDBIS VIRUS-INDUCED VIRAL RNA POLYMERASE 61 I x. ~ w -a 25 2 151 such activity was found in the cytoplasm of uninfected cells similarly analyed (Table 1). Some properties of the enyme obtained from infected cells are summaried in Table 1. In general, the properties are similar to those of other RNA polymerases reported. The activity of the enyme was only partially sensitive to digestion with ribonuclease in the reaction mixture. The enyme incorporated a maximal amount of 3H- GTP when the Tris buffer used in the assay mixture was at ph 8.. There was an absolute requirement for magnesium, its optimal concentration being 2 Atmoles/.4 ml. However, substitution of manganese for magnesium inhibited enymatic activity by about 8%. The enyme incorporated 3H-GTP nearly linearly with time up to 3 min, and thereafter the rate slowly declined. On the average, 1 mg of enyme incorporated 75 pmoles of GTP in 6 min. Our method of isolating the RNA polymerase from the infected cells differed from that of Martin and Sonnabend (7) mainly in one respect. This was the centrifugation of the cytoplasmic extract on a sucrose density gradient. The results 1 2 3 FIG. 2. Effect of ribonuclease and sodium deoxycholate on the cytoplasmic extract from infected cells. Conditions for preparation of the extract and its centrifugation are identical to those in Fig. 1. (a) 3H-uridine-labeled infected cytoplasmic extract; (O3) 3H-uridine-labeled infected cytoplasmic extract incubated with 1 sgg of ribonucleaselml at 37 Cfor 1 min in the presence of.15 M sodium chloride; (O) 3H-uridine-labeled infected cytoplasmic extract incubated at 4 C for 1 min with.5% deoxycholate. shows the results of one such experiment. The fractions 7 to 12 contained most of the enymatic activity. This indicated that most of the polymerase activity in the cytoplasmic extract resided in the fractions containing the faster-sedimenting structures as described in Fig. 1 and Fig. 2. No 4 8 12 16 2 24 FIG. 3. Viral RNA polymerase activity of the cytoplasmic extract from cells infected with Sindbis virus. Cytoplasmic extract was prepared by methods and conditions similar to those described in Fig. 1. The infected cells did not receive any radioactive label. Fractions (1.2 ml) were collected from the sucrose density gradients;.2 ml of each fraction was assayed after 3 min of incubation at 37 C.
62 SREEVALSAN AND YIN J. VIROL. TABLE 1. Effect of Mg++, nucleotides, mercaptoethanol, ATP generating system, actinomycin D, and ribonuclease on incorporating 3H- GTP by viral polymerase Reaction system Specificb Complete assay... 23,25 Complete assay Mg... 1,83 ATP... 2,46 CTP... 1,62 UTP... 4,919 ATP, CTP, UTP... 1,61 Mercaptoethanol... 22,154 Phosphoenolpyruvate and phosphoenolpyruvate kinase....... 18,337 Actinomycin D... 23,12 Complete assay +.5 jug of ribonuclease per.4 ml of reaction mixture... 8,)53 Uninfected, complete assay... 1,37 a Enyme was fraction 13 obtained from experiment shown in Fig. 3. b Counts per minute of 3H-GTP incorporated per milligram of protein per 3 min. (Table 2) show that the average of the specific enymatic activity present in the fractions from the sucrose density gradient was about fourfold higher than the specific enymatic activity present in the 2-min sediment at 13, X g prepared according to Martin and Sonnabend (7) from the same number of infected cells. Furthermore, the specific enymatic activity in fractions of the sucrose gradient was 1-fold more than that of the original cytoplasmic extract. Characteriation of the product synthesied by the enyme. The following series of experiments was undertaken to determine whether the cell-free preparations of polymerase could synthesie RNA resembling the RNA synthesied in infected cells in vivo. The RNA extracted from Sindbis virus has a sedimentation coefficient of 4S, whereas the infected cells contain 26S and 2S types of viral RNA in addition to 4S type. The 2S type of viral RNA is the ribonuclease-resistant form of viral RNA (1). RNA was isolated from a reaction mixture after the enyme had been incubated with labeled GTP for 6 min. About 1% of the labeled RNA sedimented on a sucrose density gradient with a sedimentation coefficient of 4S (Fig. 4). Most of the remaining radioactivity was contained in a peak sedimenting slightly faster than the 18S ribosomal RNA. A portion of the RNA was also incubated with ribonuclease prior to centrifugation. The data shown in Fig. 4 demonstrate that about 4% of the RNA synthesied was resistant to ribonuclease and most of the ribonuclease-resistant material sedimented with a coefficient of 2S. Therefore, the enyme in all likelihood was synthesiing viral RNA in vitro similar to viral RNA made in vivo. Purified viral-specific double-stranded RNA from infected cells is soluble in 2 M NaCl or LiCl, TABLE 2. Comparison of the RNA polymerase activity obtained from the infected cells by different methods Method of isolation Specific Total activitya activity Original cytoplasmic extract... 2,61 17,22 Cytoplasmic extract, fractionated on sucrose density gradient... 21,17b 89,97 2 min sediment at 13, X g prepared according to Martin and Sonnabend (7)... 5,735 33,27 a Counts per minute of 3H-GTP incorporated per milligram of protein per 3 min. b The value represents the average of the specific enymatic activity in fractions 7 to 16 of Fig. 3. 4 28S 18S 9 Is \ ' p3 U. a.2- I-. S~~~~~~b 1 2 3 FIG. 4. Sedimentation analysis of RNA synthesied by the enyme. One milliliter of enyme from fraction 13 obtained from the experiment similar to that shown in Fig. 3 was incubated with an equal volume of reaction mixture and 5 uc of 3H-GTP at 37 C; after 1 hr, the RNA was extracted. One-half of the solution was centrifuged directly on a 15 to 3% sucrose gradient. The other half was digested with I ;g of ribonuclease/ml at 37 Cfor 1 min prior to centrifugation. Centrifugation was done with an S W5 rotor at 1, X g for 2.5 hr. () 3H-GTP; () digested with ribonuclease.
VOL. 3, 1969 SINDBIS VIRUS-INDUCED VIRAL RNA POLYMERASE LICI INSOLUBLE LiCI SOLUBLE 63 I- U - C-) : ID 1 2 1 2 FIG. 5. Lithium chloride precipitation of a 3-min product of 3H-GTP incorporation with the enyme. A 2-ml amount ofenymefromfraction 13 obtainedfrom an experiment similar to that shown in Fig. 3 was incubated with an equal volume of reaction mixture and 1 pc of 3H-GTP for 3 min at 37 C. The RNA was obtained and precipitated with lithium chloride. The salt-precipitable RNA and -soluble RNA were subjected to digestion with ribonuclease and sedimentation analysis as described in Fig. 4. () Without ribonuclease; (O) after digestion with ribonuclease. whereas viral-specific single-stranded RNA is precipitated under these conditions. Doublestranded RNA that possesses a single-stranded component, on the other hand, behaves like single-stranded RNA during precipitation with salt (2, 9). The following experiment was undertaken to investigate the salt solubility of RNA synthesied by the enyme in vitro. Labeled GTP was incorporated by the enyme for 3 min, and the RNA was then isolated. The total RNA was treated with lithium chloride, and the sedimentation pattern of the salt-insoluble and -soluble RNA was determined on sucrose density gradients (Fig. 5). The RNA present in the salt-soluble portion sedimented as a homogeneous fraction. This probably represents a double-stranded molecule without a single-stranded component. The sedimentation rate of the salt-soluble RNA was more or less unaffected when it was incubated with ribonuclease prior to centrifugation on sucrose gradient. Almost all the single-stranded 4S RNA and about 5% of the total ribonuclease-resistant RNA were precipitated by the salt. After incubation with ribonuclease, the remaining RNA sedimented in a similar fashion to the RNA present in the lithium chloride-supernatant fraction. Similar results were obtained when viral RNA synthesied in vivo was subjected to precipitation with Li Cl. DISCUSSION The results reported here, though preliminary in nature, show that the particulate fraction of the cytoplasm from the infected cells contains a viral-specific polymerase associated with template. This enymatic activity is low or absent in uninfected cells. However, quite unlike that reported by Martin and Sonnabend (7), the enyme described here synthesies single-stranded RNA. The results indicate that the products made by the enyme in vitro resemble the various classes of viral RNA synthesied in vivo in infected cells. The enymatic fraction, prepared according to the method of Martin and Sonnabend (7) from cells infected with Sindbis virus, possessed low specific enymatic activity, and the product synthesied was exclusively ribonuclease-resistant
64 SREEVALSAN AND YIN J. VIROL. RNA. Furthermore, the original cytoplasmic extract possessed low specific enymatic activity. However, by using the fractionation step described here (sucrose density gradient centrifugation), a 1-fold increase of specific enymatic activity of the original cytoplasmic extract can be achieved. Therefore, our preparation of enyme can synthesie single-stranded RNA probably because contaminating ribonuclease is removed by the fractionation of the cytoplasmic extract on sucrose density gradient. The data on the fractionation of the RNA by Li Cl indicate that the single-stranded RNA and the replicative form of viral RNA are synthesied in vitro. The ribonuclease-resistant RNA found in the LiCl precipitate may arise from the replicative intermediate type of RNA. Repeated experiments have shown that the replicative form of viral RNA cannot be precipitated with LiCl under the above conditions (Sreevalsan, unpublished data). Therefore, the ribonuclease-resistant fraction of the RNA present in the LiCl precipitate may not be due to co-precipitation with the replicative form of RNA. Further experiments are needed to clarify whether the replicative intermediate type of RNA is synthesied in the in vitro system. The role of these types of viral RNA in the synthesis of the single-stranded RNA is not clear at present. ACKNOWLEDGMENTS We thank Joan M. Wilson and Eileen M. PoUni for their skilled technical assistance. LITERATURE CITED 1. Dulbecco, R., and M. Vogt. 1954. One-step growth curve of Western equine encephalomyelitis virus grown in vitro and analysis of the virus yields from single cells. J. Exp. Med. 99:183-199. 2. Erikson, R. L. 1966. Fractionation of viral specific ribonucleic acid by cesium sulfate equilibrium density gradient centrifugation. J. Mol. Biol. 18:372-381. 3. Gerhardt, B., and H. Beevers. 1968. Influence of sucrose on protein determination by the Lowry procedure. Anal. Biochem. 24:337-352. 4. Girard, M., D. Baltimore, and J. E. Darnell, Jr. 1967. The poliovirus replication complex-site for synthesis of poliovirus RNA. J. Mol. Biol. 24:59-74. 5. Lowry,. H., N. J. Rosenbrough, A. L. Farr, and R. J. Randall. 1951. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193:265-275. 6. Lust, G. 1966. Alterations of protein synthesis in arbovirus-infected L cells. J. Bacteriol 91:1612-1617. 7. Martin, E. M., and J. A. Sonnabend. 1967. Ribonucleic acid polymerase catalying synthesis of double-stranded arbovirus ribonucleic acid. J. Virol. 1:97-19. 8. Plagemann, P. G. W., and H. E. Swim. 1966. Symposium on replication of viral nucleic acids. III. Replication of mengovirus ribonucleic acid. Bacteriol. Rev. 3:288-38. 9. Plagemann, P. G. W., and H. E. Swim. 1968. Synthesis of ribonucleic acid by mengovirus-induced RNA polymerase in vitro: nature of products and of RNase-resistant intermediate. J. Mol. Biol. 35:13-35. 1. Yin, F. H., and R. Z. Lockart, Jr. 1968. Maturation defects in temperature-sensitive mutants of Sindbis virus. J. Virol. 2: 728-737.