RNA-dependent RNA polymerase isolated from cowpea chlorotic mottle virus-infected cowpeas is specific for bromoviral RNA

VIROLOGY

132,53-60

(1984)

RNA-Dependent RNA Polymerase Isolated from Cowpea Chlorotic Mottle Virus-Infected Cowpeas Is Specific for Bromoviral RNA W. A. MILLER’ Biophysics

Laboratory

of the Gradude

AND

T. C. HALL

School, University

of Wisconsin, Ma&xm,

Wisconsin 55706

Received May 26, 1985; accepted September 27, 198s An RNA-dependent RNA polymerase activity capable of synthesizing full length doublestranded RNA products only in the presence of bromoviral RNA templates has been isolated from cowpea cblorotic mottle virus (CCMV)-infected cowpeas. No comparable discrete products were obtained when a nonbromoviral (cowpea mosaic virus) RNA was used as template. Heterodisperse, ribonuclease-sensitive products were obtained in reactions catalyzed by similar extracts from mock-inoculated (uninfected) plants in the presence of added CCMV RNA. The extraction procedure was identical to that used for obtaining virus-specific polymerase from brome mosaic virus-infected barley (Bujarski, Hardy, Miller, and Hall, ViroZogg 119,465473,1982). Comparison of the activities of the barley- and cowpea-derived enzymes provides further evidence for our contention that a virus-encoded polypeptide is an integral component of these replicases.

closely related. It follows that if a polymerase can be isolated from CCMV-infected leaves that is similar to BMV polymerase, but dissimilar to the endogenous cowpea polymerase, then a viral origin is strongly indicated. Moreover, if the replicase is host encoded, it would be expected to copy the RNA of any virus that infects cowpeas. To evaluate these alternatives, we prepared extracts from mock-inoculated and CCMV-infected cowpeas and from BMV-infected barley using the same procedure (Bujarski et aL, 1982). The template preferences of the various extracts were compared, and their reaction products analyzed electrophoretically.

INTRODUCTION

Several laboratories have reported the presence of RNA-dependent RNA polymerase activity in healthy plants, including cowpeas (White and Dawson, 1978; Ikegami and Fraenkel-Conrat, 1978,1980, Dorssers et al, 1982,1983). The role of such activity in the healthy plant is unclear, but its presence has led to the suggestion that the host plant supplies all of the enzyme(s) necessary for the replication of viral RNA upon infection (Fraenkel-Conrat, 1979; 1983). By contrast, properties of RNA-dependent RNA polymerase from brome mosaic virus (BMV) -infected barley, such as its high preference for BMV RNA over other viral RNAs as template, the presence of a viral protein in the extract, and the lack of detectable template-specific activity in uninfected plants, imply that virally encoded gene products make an essential contribution to replicase (Hardy et al, 1979; Bujarski et aZ., 1982). Cowpea chlorotic mottle virus (CCMV) is closely related to BMV, but the host plants in which they replicate are not 1 Author addressed.

to whom requests for reprints

MATERIALS

AND

METHODS

Mattiak. Uridine 5’-[Lu-32P]triphosphate, (triethylammonium salt, 410 Ci/mmol) was purchased from the Radiochemical Centre, Amersham, England, [5,6-3H]uridine 5’ triphosphate (tetrasodium salt, 40 Ci/mmol) was purchased from New England Nuclear (Boston, Mass.), and micrococcal nuclease was purchased from P-L Biochemicals, Milwaukee, Wisconsin. Cowpea seeds (Vigna sinensis, “black eye”)

should be

53

0042-6822/84 $3.00 Copyright All rights

0 1984 by Academic Press. Inc. of reproduction in any form reserved

54

MILLER

AND

were obtained from L. L. Olds Seed Co., Madison, Wisconsin.

Preparation

of RNA-dependent

RNA

polvmerase extracts. RNA-dependent RNA polymerase was prepared from barley as described in Bujarski et aL (1982). Briefly, infected leaves were ground and the pellet obtained by high speed centrifugation was treated with the detergent dodecyl-/3-Dmaltoside. Subsequent ultracentrifugation yielded a pellet which was resuspended and subjected to centrifugation through a sucrose pad to obtain the final replicase pellet which was suspended in 50 mM Tris-HCl (pH 8.0), 10 mM Mg(OAc)z, 0.1% dodecylP-D-maltoside, 10 mM dithiothreitol, and 1 mM each of ATP, CTP, and GTP. To prepare CCMV polymerase, primary leaves of 9day-old cowpeas were inoculated with CCMV; 5 days after inoculation, the young trifoliate leaves were used for RNA-dependent RNA polymerase extraction by the above procedure. To obtain “mock extracts,” virus was omitted from the inoculum and the extracts were prepared by the above procedure. Except where otherwise indicated, polymerase extracts were pretreated with micrococcal nuclease in the presence of calcium to digest endogenous RNA as described by Miller and Hall (1983). EGTA was then added to the extracts to terminate nuclease digestion before they were added to the polymerase reaction.

RNA-dependent RNA polymerase reaction and product analysis. RNA-dependent RNA polymerase activity was assayed by monitoring the incorporation of [a-32P]UTP into an acid-precipitable form. The reaction mixture (50 ~1) contained 2 PM [a32P]UTP, 200 pg/ml actinomycin D, viral RNA, and 40 ~1 polymerase extract, unless otherwise indicated. The mixture was incubated for 30 min at 30”, after which 40 ~1 was spotted onto trichloroacetic acid-sodium pyrophosphate-saturated discs which were washed and monitored for radioactivity as described previously (Hardy et al, 1979). Radioactivity detected at zero time (-10,000 cpm) was subtracted. To characterize the products, the polymerase reaction was scaled up to contain

HALL

160 ~1 polymerase, 35 PM [3H]UTP (20 Ci/ mmol), or 13 PM [cF~~P]UTP (67 Ci/mmol), 200 pg/ml actinomycin D, and 20 pg viral RNA to a total volume of 190 ~1. After incubation at 30” for 60 min, the reaction was stopped by the addition of 1 ml TNE+ SDS buffer (0.1 M Tris-Cl, pH 7.2, 0.1 M NaCl, 0.01 M EDTA, and 0.5% SDS). RNA was extracted from the mixture with phenol followed by ethanol precipitation. A fraction of the RNA obtained was digested with ribonucleases (RNases) A (10 pg/ml) and Tl (10 U/ml) in 2X SSC (0.3 M NaCl, 0.03 M sodium citrate) as described (Hardy et aZ., 1979). The volumes of the RNA products loaded onto each lane of a gel were adjusted so as to give similar levels of acid precipitable radioactivity (10,000 cpm). Electrophoresis was through a 2.4% acrylamide-0.45% agarose gel at 25 mA for 110 min (Peacock and Dingman, 1968). The gel was impregnated in 1 M sodium salicylate for 20 min prior to fluorography (Chamberlain, 1979). Other method-s. Protein content of the polymerase extracts was determined using a Protein Assay Kit purchased from BioRad Laboratories, Richmond, California. CCMV RNA components were separated by linear-log sucrose gradient centrifugation (Kiberstis et a& 1981) and found to be pure by ethidium bromide staining of the RNAs after acrylamide-agarose gel electrophoresis as described above. RESULTS

Properties of the cowpea extracts. In the absence of micrococcal nuclease treatment, reactions containing extracts from either mock-inoculated or CCMV-infected cowpeas incorporated significant amounts of radioactivity into an acid-precipitable form (Table 1). However, neither extract showed stimulation of activity when CCMV RNA was added to the reaction. In contrast, extracts prepared by the same procedure from BMV-infected barley showed significant template dependence and higher activity than did those from uninfected barley, which had almost no activity. Treatment of the preparations with micrococcal nuclease to digest RNA and DNA that co-

CCMV

RNA

POLYMERASE

TABLE

55

1

EFFECT OF MI~FLOCOCCALNUCLEASE TREATMENT ON INCORPORATION OF [8epjUMP FPJ’UMP

Source of extract Mock-inoculated

Nuclease treatment

cowpeas +

CCMV-infected

cowpeas

+

Mock-inoculated

barley

+

BMV-infected

barley

+

-RNA (w4

incorporated

+RNA (cpm)

+RNA/-RNA

Minimum specific activity (pmol/w)

44,140 15,160

31,764 21,056

0.7 1.4

14.8 9.9

100,639 12,420

96,988 52,978

1.0 4.3

43.6 23.8

2,612 518

7,020 0

2.7 0

3.4 0

70,729 6,506

226,318 138,194

3.2 21.2

113.5 69.3

Note. All extracts were prepared and assayed as described in Materials and Methods, except that micrococcal nuclease treatment was omitted where indicated. Reaction mixtures contained 10 &i [82p]vTP (79 Ci/mmol), and either no RNA, 12 pg CCMV RNA (cowpea extracts), or 12 fig BMV RNA (barley extracts). Minimum specific activity was defined as picomoles of nucleotides incorporated per milligram of protein in the presence of added RNA, assuming no isotopic dilution by endogenous UTP. The increase in incorporation that was stimulated by virus RNA corresponded approximately to 0.0034 and 0.011% of the added template mass for the nuclease-treated extracts from inoculated cowpeas and barley, respectively.

purified with the polymerase greatly increased the RNA dependence of the extracts from CCMV- and BMV-infected plants, but only slightly increased the RNA dependence of the extract from mock-inoculated cowpeas. Micrococcal nuclease treatment rendered the mock-inoculated barley extract devoid of activity. The incorporation of [32P]UMP by extracts from infected plants was reduced after nuclease treatment because endogenous RNA, which had contributed to incorporation of radioactivity, was extensively degraded (Miller and Hall, 1983). To further characterize the activity of cowpea extracts, the effects of inhibitors on incorporation of radioactivity were studied. Inhibitors specific for DNA-dependent RNA polymerases (actinomycin D, rifampicin, and cy-amanitin) and polynucleotide phosphorylases (orthophosphate) did not inhibit incorporation, but pyrophosphate inhibited the enzymes to varying degrees. BMV and CCMV polymerases were markedly inhibited by 2 mM pyrophosphate (as expected for RNA polymerases), retaining 0 and 29%) respectively, of

their original activities, whereas the extract from mock-inoculated cowpeas retained 60% of its activity. The requirement for nucleotide triphosphates (NTPs) was also examined. The extracts from mockinoculated and CCMV-infected leaves were essentially inactive when all three unlabeled nucleotide triphosphates were omitted (data not shown). This ruled out the possibility that terminal uridylyl transferase, which has been reported in cowpeas (Zabel et aL, 1981), was contributing to [32P]UMP incorporation. In summary, the extract from CCMV-infected cowpeas appears to have an RNA-dependent RNA polymerase activity that differs from the activity present in extracts from mock-inoculated cowpeas prepared by identical procedures, but further experiments are required to rule out a role for the latter extract in CCMV RNA replication. Characterization of polymerase reaction products. Strong evidence that the RNAdependent RNA polymerase activity observed in extracts from infected plants is involved in the process of viral replication occurring in nature is provided by the

MILLER

56

demonstration of full length doublestranded (ds) RNAs as the major reaction products. 3H-Labeled RNA products synthesized by preparations from mock-inoculated and CCMV-infected cowpeas using added CCMV RNA as template, were extracted and analyzed electrophoretically. The products of the extract from CCMVinfected leaves consisted of three prominent dsRNAs that comigrated with doublestranded replicative forms (RFs) of CCMV RNAs 2, 3, and 4 (Fig. 1). Two differences between in vitro reaction products and RFs found in viva were that RNA 1 was copied

123458

7

8

FIG. 1. Fluorography of ‘H-labeled RNA synthesized by various polymerase extracts. RNA products extracted from 199~1 reactions were ribonuclease treated, and analyzed electrophoretically as described in Materials and Methods. Lane 1: FHjdsRNA extracted from CCMV-infected cowpea leaves labeled with [aH]uridine as described by Dawson et al (1976). The RNA components to which the dsRNAs correspond are indicated at left. Lanes 2-5: Products of untreated (lanes 2 and 3) and micrococcal nucleasetreated (lanes 4,5) RNA-dependent RNA polymerase extracted from CCMV-infected cowpeas using 25 pg CCMV RNA as template. Lanes 3 and 5 represent RNA products that were RNase treated after extraction from the polymerase mixture. Lane 6: Products of the extract from mock-inoculated cowpeas. Lanes ‘7 and 8: Products of micrococcal nucleasetreated BMV RNA-dependent RNA polymersse using 25 Mg added BMV RNA as template. Products in lane 8 were ribonuclease treated.

AND

HALL

poorly in vitro and that RF of RNA 4 was present in lower levels in vivo. These differences were also evident in a comparison of BMV RNA-dependent RNA polymerase reaction products with the RFs found in vivo (Hardy et al, 1979). The reason for this is unknown. The RNA products of the CCMV extract are resistant to ribonuclease in 2~ SSC which indicates that they are double stranded. The minor low molecular weight bands that can be discerned are probably due to degradation of the full length template RNAs, giving some fragments that can be copied because they contain the same 3’ end as the intact RNA (J. Bujarski, personal communication). Micrococcal nuclease treatment of the extract before the polymerase reaction did not alter its ability to synthesize full length dsRNA products (Fig. 1, lanes 2-5). This result strongly suggests that the CCMV polymerase is capable of initiating replication, as was shown in the case of BMV RNA-dependent RNA polymerase (Miller and Hall, 1983). In contrast to the above observations, the yield of RNA products obtained in reactions containing extracts from mock-inoculated cowpeas was very low. Equal amounts of reaction products, in terms of acid-precipitable radioactivity, were loaded on to each lane of the gel shown in Fig. 1. This required five times the volume (10 ~1) of the products synthesized in the presence of extracts from mock-inoculated cowpeas than that (2 ~1) for the extracts from CCMV-infected cowpeas. The few products that were discernable were about the sizes of ribosomal RNAs and were single stranded. No products the size of CCMV RFs were seen, even after prolonged exposure of the gel to film. When the products of the extract from mock-inoculated leaves were RNase treated no acid-precipitable radioactivity remained, precluding electrophoretic analysis. These data show that the extract from mock-inoculated cowpeas had no ability to synthesize RF RNAs, as would be expected of a viral replicase. The reason for the low level of radioactivity that was incorporated into ribosomal RNA in these experiments is unclear. The yield of 3H-labeled products synthesized by the

CCMV

RNA

57

POLYMERASE

micrococcal nuclease-treated extract from mock-inoculated leaves was too low to permit electrophoretic analysis. For comparison, 3H-labeled products of reactions containing BMV RNA-dependent RNA polymerase were run on the gel beside those synthesized by the cowpea extracts. The products synthesized by the extract from CCMV-infected leaves looked very similar in distribution to those synthesized by the BMV extract, but those synthesized by the mock extract did not. In conclusion, only extracts from plants infected with virus show any RNA-dependent RNA polymerase (presumptive replicase) activity. Template spetificitg of CCMV RNA polgmerase. The incorporation of radioactivity by extracts from mock-inoculated and CCMV-infected cowpeas in the presence of various viral RNA templates is shown in Table 2. The extract from CCMV-infected plants was much more dependent on template for activity, and was about three times as active as the extract from mockinoculated cowpeas. By this assay, neither the extracts from mock-inoculated nor those from CCMV-infected cowpeas appeared to use CCMV RNA preferentially as a template. To determine the significance of the radioactivity incorporated in the experiments shown in Table 2, the products synthesized by CCMV polymerase in the presence of nonhomologous viral RNAs were analyzed. CCMV RNA polymerase was found to copy BMV RNA just as well as it copied CCMV RNA, as demonstrated by the full length

FIG. 2. Radiolabeled products of micrococcal nuclease-treated CCMV RNA polymerase to which 25 pg unfractionated CCMV RNA (lanes 1 and 2), BMV RNA (lanes 3 and 4), CPMV RNA (lanes 7 and 8), purified CCMV RNA 3 (lanes 9 and lo), or CCMV RNA 4 (lanes 11 and 12) was added as template. Products of BMV RNA polymerase using BMV RNA as template are shown in lanes 5 and 6. Products were ‘H-labeled in lanes 1-8 and q-labeled in lanes 9-12. Even-numbered lanes contain products that were ribonuclease-treated as described in Materials and Methods. Arrows at left indicate the positions (for lanes 1-8) of CCMV and BMV RFs extracted from tissue including BMV RF 4 (B4) and CCMV RF 4 (C4). Arrows labeled M and B indicate the positions expected for full length RFs of CPMV RNAs M and B. Arrows at right indicate positions of CCMV RFs 3 and 4 for lanes 9-12

products shown in Fig. 2. The RF of BMV RNA 4 (B4) migrated more slowly than did that of CCMV RNA 4 (C4), allowing the reaction products to be distinguished. These data are interesting because they

TABLE

2

TEMPLATE SPECIFICITY OF EXTRACTS FROM COWPEAS RNA added

Mock-inoculated cowpeas

cpm with added RNA/ cpm with no added RNA

CCMV-infected cowpeas”

cpm with added RNA/ cpm with no added RNA

None

18,405

-

11,880

-

CCMV CPMV BMV

18,205 18,110 20,145

1.0 1.0 1.1

78,220 60,330 90,545

6.6 5.2 7.6

Note. The specific activity of $PlUTP supplied was 83 Ci/mmol. Where present, 12 pg of RNA was added as template. a Values are given as cpm of [=P]UMP incorporated for 40-pl samples of 50-pl reactions.

58

MILLER

AND

provide a contrast between BMV RNA polymerase which showed very low stimulation of activity on addition of CCMV RNA (Miller and Hall, 1983), and CCMV polymerase which did not discriminate against BMV RNA. The absence of any CCMV RF 4 among the products of CCMV polymerase when BMV RNA was added confirms that the endogenous CCMV RNA in the CCMV polymerase extract was completely degraded by pretreatment of the extract with micrococcal nuclease. This is supported by the use of sucrose gradientseparated CCMV RNAs as template. When purified CCMV RNAs 3 or 4 were added to the polymerase reaction, the only products obtained were the double-stranded forms corresponding to the added RNA (Fig. 2, lanes 9-12). These observations support the notion that the extract from CCMV-infected leaves copied the RNA submitted to it and can initiate polymerization on added template. Although CCMV RNA-dependent RNA polymerase was able to copy bromoviral RNA, the products of the reaction when CPMV RNA was added as template were strikingly different. In this case only heterodisperse products of less than full length were synthesized (Fig. 2, lanes 7 and 8), most of which ran off the gel. These products were single stranded as evidenced by their degradation by ribonuclease. No bands remotely resembling full length double-stranded CPMV RNAs could be detected. The possibility that the preparation of CPMV RNA used here was degraded or contaminated, preventing it from functioning as template, had been ruled out by its ability to be translated into functional proteins in a reticulocyte in vitro translation system (A. Palmenberg, personal communication). Such dsRNAs would be expected to migrate to the positions indicated in Fig. 2 (arrows at right of lane 8). Thus, despite the observation that CCMV polymerase incorporated nearly as much radioactivity when CPMV RNA was added as when CCMV RNA was added (Table 2), the differences among the products (Fig. 2) show that CCMV RNA polymerase could not copy CPMV RNA.

HALL

DISCUSSION

The importance of product analysis in interpretation of replication experiments. Although little or no [32P]UMP-incorporating activity was detected in extracts from mock-inoculated barley, a significant level was found in reactions containing extracts from mock-inoculated cowpeas (Table 1). However, the lack of sensitivity to pyrophosphate suggests that only part of the acid-precipitable radioactivity was due to a polymerase-like activity. Moreover, the product analysis depicted in Fig. 1 demonstrates that whereas the doublestranded viral RF RNAs are readily detected in reactions catalyzed by enzyme from virus-infected eowpeas, the extract from mock-inoculated cowpeas was incapable of synthesizing double-stranded products of any size. It follows from these observations that electrophoretic analysis of the RNA products synthesized in a putative replicase reaction is vital to the interpretation of the results in terms of a role in viral replication.

Ir&wnm for a virus-m&d wntributti to replicate. In a recent review (Hall et al, 1982) we have questioned previous assertions of the similarities of polymerase activities extracted from healthy and virally infected plants. The differences in such activities described here for the CCMV enzyme are in agreement with the comparisons made by Dorssers et aZ. (1983) for extracts from healthy and cowpea mosaic virus (CPMV) -infected cowpeas. These investigators concluded that the enzyme from uninfected cowpeas was incapable of viral RNA replication. RNA-dependent RNA polymerase activity has been reported to reside in the soluble fraction of extracts from uninfected cowpeas, while the majority of the activity is membrane bound in extracts from infected cowpeas (Fraenkel-Conrat, 1983; White and Dawson, 1978). These observations define a distinguishing feature between the enzymes which alone would be difficult to reconcile with the view that viral replicase is only quantitatively different from host polymerase. In fact, White and Dawson (1978)

CCMV

RNA

POLYMERASE

59

concluded that the host polymerase and barley stripe mosaic virus (BSMV) infects viral replicase are indeed different enzymes the same host as BMV, and has a tRNAeven though the extracts used in their ex- like 3’ terminus which can be tyrosylated periments showed little template depen- like BMV RNA (Agranovsky et ul, 1981; dence. We found no polymerase activity in Loesch-Fries and Hall, 1982), it is a poor the soluble fraction of extracts from cow- template for BMV polymerase (Miller and peas, but instead detected a high level of Hall, 1983). In our hands, CCMV polymer[32P]UMP-incorporating activity that was ase extracted from cowpeas is neither as determined to be due to a terminal trans- active nor as template dependent as is BMV RNA polymerase isolated by similar proferase (data not shown) perhaps similar to that described by Zabel et al. (1981). It cedures from barley. However, it is noteis possible that the replicase that generates worthy that application of the extraction viral RNA in naturally infected plants procedure optimized for the isolation of consists of more than one subunit. One of BMV polymerase from barley to a comthese could be a host-supplied RNA-de- pletely unrelated plant yields substantial pendent RNA polymerase, which, by com- RNA-dependent RNA polymerase activity, bining with viral subunit(s) that are per- especially in view of the numerous host haps required for initiation and specificity proteins and membranes present in the exof replication, becomes an integral part of tract (Bujarski et al, 1982). The ability of CCMV polymerase to use the viral replicase (see Hall et a& 1982). BMV RNA as template but not vice versa The results presented here do not distinguish between this alternative and the may reflect the in vivo situation. Watts possibility that the entire replicase enzyme and Dawson (1980) reported that mixed inis encoded by the viral genome. fections of tobacco protoplasts with BMV and CCMV resulted in dominance by BMV The observation that CCMV RNA-dependent RNA polymerase did not copy due to events occuring early in infection. CPMV RNA also suggests that a host en- They suggested that “BMV RNA may be zyme alone cannot replicate viral RNA. able to compete for both replicases,” which CPMV also infects cowpeas and should is in accord with our results. The inability of RNA-dependent RNA polymerases from therefore be a good template if a host-encoded polymerase is responsible for rep- bromovirus-infected plants to copy RNA lication of its RNA genome. It is possible of nonbromoviruses that infect the same that the tRNA-like structure at the 3’ end hosts, and the inability of the extracts from mock-inoculated plants to replicate viral of all bromoviral RNAs (Ahlquist et al, RNA, strongly suggest that at least one 1981a) acts as a primer for bromoviral (-) viral gene product is an essential compostrand RNA synthesis, and that a primer forming double-stranded regions must be nent of the viral replicase. added for the replication of CPMV RNA ACKNOWLEDGMENTS which does not have a tRNA-like 3’ end. This research was supported by NIH Grant AI However, experiments in which oligo(U) 11572. We thank T. L. German, L. S. Loesch-Fries, was added as a primer in attempts to pro- J. J. Bujarski, P. A. Kiberstis, and S. F. Hardy for mote the synthesis of RNA complementary valuable technical advice and A. C. Palmenberg and to CPMV RNA (which is polyadenylated P. Kaesberg for helpful discussion. We are grateful at its 3’ end) yielded heterodisperse prod- to P. Kaesberg for generously providing laboratory ucts like those shown in Fig. 2 (lanes ‘7and space and facilities. 8). The same treatment had no effect on REFERENCES the product profile of bromoviral RNA (data not shown) which contains an interAGRANOVSKY, A. A., DOLJA. V. V., GORBULEV. V. G., nal poly(A) sequence in RNA 3 (Ahlquist Koztov, Yu. V., and ATABEKOV, J. G. (1981). Amiet al, 1981b). Similar template specificity noacylation of barley stripe virus RNA: Polyadenylate-containing RNA has a 3’-terminal tyrosinecan be observed in the case of BMV RNAaccepting structure. virdogy 113, 174-187. dependent RNA polymerase. Even though

60

MILLER

AND HALL

AHLQUIST, P., DASGUPTA, R., and KAESBERG, P. (1981a).Near identity of 3’RNA secondary structure in bromoviruses and cucumber mosaic virus. Cell 23,183-189. AHLQUIST,P., LUCKOW,V., and KAESBERG,P. (1981b). Complete nucleotide sequence of brome mosaic virus RNA 3. J. Md Biol 153, 23-28. BUJARSKIJ. J., HARDY,S. F., MILLER,W. A., and HALL, T. C. (1982). Use of dodecyl-@-D-maltoside in the purification and stabilization of RNA polymerase from brome mosaic virus-infected barley. Virology 119.465-473. CHAMBERLAIN,J. P. (1979). Fluorographic detection of radioactivity in polyacrylamide gels with the water-soluble fluor, sodium salicylate. Anal Biochem. 98,132-135.

DAWSON,W. O., GERMAN,T. L., and SCHLEGEL,D. E. (1976). Homogenization-resistant and -susceptible components of tobacco mosaic virus replicative form RNA. J. Gen Vird 32, 205-215. DORSSERS, L., ZABEL, P., VAN DER MEER, J., and VAN KAMMEN, A. (1982). Purification of a host-encoded RNA-dependent RNA polymerase from cowpea mosaic virus-infected cowpea leaves. Virology 116, 236-249. DORSSERS,L., VAN DER MEER, J., VAN KAMMEN, A., and ZABEL,P. (1983).The cowpea mosaic virus RNA replication complex and the host-encoded RNA-dependent RNA polymerase complex are functionally different. firology 125, 155-174. EL MANNA, M. M. and BRUENING,G. (1973). Polyadenylate sequences in the ribonucleic acids of cowpea mosaic virus. virdosy 56, 198-206. FRAENKEL-CONRAT,H. (1979). RNA-dependent RNA polymerases of plants. Trends Biochem. Sci 294, 184-186. FRAENKEL-CONRAT,H. (1983). RNA-dependent RNA polymerases of plants. Proc Nat. Acad Sci USA 80,422-424.

HALL, T. C., MILLER, W. A., and BUJARSKI,J. J. (1982). In “Advances in Plant Pathology” (D. Ingram and P. Williams, eds.), Vol 1, Academic Press, London.

HARDY, S. F., GERMAN,T. L., LOESCH-FRIES,L. S., and Hw, T. C. (1979). Highly active templatespecific RNA-dependent RNA polymerase from barley leaves infected with brome mosaic virus. Pram. Nat. Ad

Sci USA 76, 4956-4960.

IKEGAMI, M., and FRAENKEL-CONRAT,H. (1978). The RNA-dependent RNA polymerase of cowpea. FEBS L&t. 96, 197-200. IKEGAMI, M., and FRAENKEL-CONRAT,H. (1980). Lack of specificity of virus-stimulated plant RNA-dependent RNA polymerase. virdogy 100,185-188. KIBERSTIS, P. A., LOESCH-FRIES,L. S., and HALL, T. C. (1981). Viral protein synthesis in barley protoplasts inoculated with native and fractionated brome mosaic virus RNA. Virology 112, 804-808. LOESCH-FRIES,L. S., and HALL, T. C. (1982). In tivo aminoacylation of brome mosaic and barley stripe mosaic virus RNAs. Nature (London) 298.771-773. MILLER, W. A., and HALL, T. C. (1983). Use of micrococcal nuclease in the purification of highly template dependent RNA-dependent RNA polymerase from brome mosaic virus-infected barley. fir125, 236-241. PEACOCK,A. C., and DINGMAN,C. W. (1968). Molecular weight estimation and separation of ribonucleic acid by electrophoresis in agarose-acrylamide composite gels. Biochemistry 7.668-674. WAWS, J. W., and DAWSON,J. R. 0. (1980). Double infection of tobacco protoplasts with brome mosaic virus and cowpea chlorotic mottle virus. Vi105, 501-507. WHITE, J. L., and DAWSON,W. 0. (1978). Characterization of RNA-dependent RNA polymerases in uninfected and cowpea chlorotic mottle virus-infected cowpea leaves: Selective removal of host RNA polymerase from membranes containing CCMV RNA replicase. virdog?l 88, 33-43. ZABEL,P., DORSSERS, L., WERNARS,K., and VAN KAMMEN, A. (1981). Terminal uridylyl transferase of Vigna unguiculati Purification and characterization of an enzyme catalyzing the addition of a single UMP residue to the 3’ end of an RNA primer. N&.&c Acid Res. 9, 2433-2453.