Proc. Nati. Acad. Sci. USA Vol. 83, pp. 63-66, January 1986 Biochemistry

Infectious RNA derived by transcription from cloned cDNA copies of the genomic RNA of an insect (black beetle virus//viral gene expression/in vitro transcription/Drosophila melanogaster) BIMALENDU DASMAHAPATRA, RANJIT DASGUPTA, KEITH SAUNDERS, BERNARD SELLING, THOMAS GALLAGHER, AND PAUL KAESBERG Biophysics Laboratory and Biochemistry Department, University of Wisconsin, Madison, WI 53706 Communicated by Robert H. Burris, August 30, 1985

ABSTRACT RNA transcripts of cloned cDNA of the MATERIALS AND METHODS genomic of BBV (black beetle virus) are infectious to We used BBV W17 (7), a viral strain that is highly cytolytic cultured cells of Drosophila melanogaster. Individual tran- to cultured cells of D. melanogaster. scripts had approximately 10% of the infectivity of the corre- Synthesis of Full-Length DNA Copies of BBV RNAs and sponding authentic virion RNA. Progeny virus resulting from Their Cloning into Transcription Vectors. RNAs were re- transcript infection was phenotypically indistinguishable from verse-transcribed into cDNA and converted to the double- the progenitor virus used to generate the original cDNA forms stranded form with reverse transcriptase (12). Specific as judged by sucrose density gradient sedimentation, specific oligonucleotides, used as primers, had extra nonviral bases at infectivity, plaque morphology, and serology. Although the their 5' ends, which resulted in the emergence of unique transcript RNAs used to produce this virus had 20 nonviral restriction sites at each end ofthe double-stranded DNA (Pst bases headed by a capping group at their 5' termini, these 20 I at the 5' end and Xba I at the 3' end). Full-length bases were absent in the progeny viral RNAs. The cDNA forms, double-stranded DNA copies ofBBV RNA1 and RNA2 were and therefore the resulting transcript RNAs, should be readily inserted into the Sma I site of the multicopy plasmid pUC13 modifiable by the techniques of recombinant DNA technology and were cloned in Escherichia coli JM 101. Complete BBV both for viral studies and for the insertion offoreign genes into DNA inserts were excised from pUC13 recombinant plas- the viral and thus into the host cytoplasm. mids and were inserted into transcription vectors pSP64 or pPM1 and cloned. The methods ofrecombinant DNA technology can be ofgreat In Vitro Transcription ofCloned BBV DNA. Selected pSP64 recombinant plasmids were cleaved with the restriction utility for studies of RNA . Such methods are intrin- enzyme Xba I, and the resulting linear DNA templates (20 sically available for , which have DNA as an ,ug/ml) were transcribed with SP6 RNA polymerase as intermediate in their synthesis. The methods have now described by Konarska et al. (13). Generally, 500 kLM become applicable to RNA phage QB (1) and to poliovirus (2), guanosine (5') triphospho(5')guanosine [G(5')ppp(5')G] or whose cDNAs have been shown to be infectious, and to the 7-methyguanosine derivative m7G(5')ppp(5')G was in- brome mosaic virus (3), for which infectious RNA has been cluded in the reaction mixture to provide capped transcripts. made by transcription from cloned viral cDNA. Infectious pPM1 recombinant plasmid DNA, linearized with Xba I, was cDNA and infectious transcript RNAs are also known for the transcribed with E. coli RNA polymerase as described by plant pathogens potato spindle tubor viriod (4) and hop stunt Ahlquist and Janda (12). viriod (5). To our knowledge there have not been published Clones were named according to the format pxBySPz or reports ofinfectious transcript RNAs for any insect or animal pxByPMz, where x = 1 or 2 indicates derivation from RNA1 virus. or RNA2, y is the isolate number of the pUC13 recombinant We show here that RNA transcripts derived from DNA plasmid containing the BBV insert, and z indicates the copies of the genomic RNAs of BBV are infectious to isolation number of either a pSP64 or a pPM1 recombinant cultured cells of Drosophila melanogaster; thus this virus, carrying the insert BBV DNA in the correct orientation. also, is modifiable by DNA methods. Infectivity Assays. Drosophila melanogaster cells (5 x 106) Black beetle virus (BBV) is an insect virus of the family in 80 Al ofPNKC buffer (35 mM Pipes/100 mM NaCl/10 mM Nodaviridae. Its genome consists of two single-stranded, KCl, 1 mM CaCl2/400 iLg of DEAE dextran per ml, pH 6.0), messenger-sense RNAs contained in a single virion (6, 7). were transfected (10) by addition of various amounts of Virion RNA1 (3106 bases) (8) codes for protein A (involved transcribed or virion RNAs in 20 ,ul ofPNKC buffer. After 10 in viral RNA synthesis) and protein B (function unknown), min at room temperature, cells (200 to 2 x 106) from each transfection mixture were added to untreated Drosophila whose cistron is silent. The protein B cistron is expressed by cells to give a total of 4 x 106 cells in 5 ml of NPKA buffer means ofa subgenomic messenger, RNA3 (389 bases) (9, 10), (25 mM Pipes, pH 6.75/100 mM NaCl/10 mM KCl/0.1% which is not encapsidated. Virion RNA2 (1399 bases) (11) bovine serum albumin). Cells were mixed and poured into encodes the virion coat protein precursor, a, which is 6-cm tissue culture dishes and, after 2-3 days at 26°C, proteolytically processed into the coat proteins /3 and y. Both infectious centers were counted (7). RNA1 and RNA2 have a 5'-terminal capping group and a blocking moiety, presumably a protein (8, 11), at their 3' terminus. RESULTS DNA from pSP64 clones was cleaved with Xba I and transcribed with SP6 RNA polymerase to produce complete The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact. Abbreviations: BBV, black beetle virus; pfu, plaque-forming unit(s). 63 Downloaded by guest on September 29, 2021 64 Biochemistry: Dasmahapatra et al. Proc. Natl. Acad. Sci. USA 83 (1986)

copies of BBV . All such transcripts have the same Infectivity of Transcript RNA1 Plus Transcript RNA2. The four additional nonviral nucleotides at their 3' ends and the crucial test ofbiological activity is of course the infectivity of same 20 additional nonviral nucleotides at their 5' ends transcript RNA1 plus transcript RNA2 in the absence of any headed by a capping group, or in some experiments, by a authentic RNA. When cells were transfected with a mixture methylated capping group or a triphosphate. In several of transcripts that we had identified as infectious by the experiments we cloned full-length BBV insert DNAs in the procedures described above, plaque assays consistently transcription vector pPM1 and used E. coli RNA polymerase indicated infectivity under conditions that had been shown to to provide transcripts whose 5' termini are identical in be optimal for authentic RNAs. Infectivity required the sequence to authentic virion RNA. Table 1 shows the presence of both transcript RNA1 and transcript RNA2. We terminal structure of the various transcripts. were unable to detect infectivity with Xba I-linearized Infectivity of Transcript RNA1 Plus Authentic RNA2 and of p1B9SP DNA plus p2B1OSP DNA. The RNA input required Transcript RNA2 Plus Authentic RNA1. In order to initially for the maximal number of plaque-forming centers was identify infectious transcripts and to quantify their infectiv- always higher for transcript RNA than for virion RNA (Table ity, preparations of individual transcripts were combined 4). Fig. LA shows the results obtained when the input of with preparations of cognate RNAs derived from virions, and transcript RNA1 was varied and the input oftranscript RNA2 these mixtures were assayed for infectivity. To serve as a was held constant at 0.2 jig. Fig. 1B shows the results baseline for judging infectivity of the transcript RNAs, we obtained when the input of transcript RNA2 was varied and first determined the infectivity of combinations of authentic the input oftranscript RNA1 was held constant at 1.2 ,g (i.e., virion RNA1 and RNA2. Maximal infectivity was obtained approximately the optimal value found in Fig. LA). It is when 5 x 106 cells were transfected with a mixture of 0.5 ,ug evident from Fig. 1B that the highest infectivity is obtained at of virion RNA1 and 0.2 ,ug of virion RNA2. Under these approximately equal molarity oftranscript RNA1 and RNA2. conditions approximately 5% of the transfected cells formed At equimolarity, maximal infectivity is about 1% of that of plaques. authentic RNAs and is obtained at a transcript RNA con- Preparations of transcript RNA1 were assayed together centration approximately 3 times that required for optimal with virion RNA2 at concentrations optimal for virion RNAs. infectivity with authentic RNAs (Fig. 1C). Table 2 shows the infectivity of three transcripts from the The virus arising from the combination RNA1(plB9SP) same DNA [capped transcript p1B13SP, methylated capped and RNA2(p2BlOSP), designated BBV K1, has been studied transcript p1B13SP(M), and uncapped transcript p1B13SP- in detail. BBV K1 is phenotypically indistinguishable from (ppp)] and another transcript, p1B9SP, relative to virion BBV W17, the progenitor virus from which it was derived, as RNA1. The p1B13SP and p1B13SP(M) transcripts were each we describe below. approximately 10% as infectious as virion RNA1. The uncap- Plaque Analyses. Assay of W17 virions, W17 virion RNAs, ped transcript p1B13SP(ppp) had approximately 0.1% of the or W17 RNA1 plus any of our infectious transcript RNA2s infectivity ofvirion RNA1. Table 2 indicates infectivity at the resulted in mostly large (2-4 mm) clear plaques. Assay of level of 0.03% when no RNA1 was added. This infectivity is transcript RNA1 plus virion or transcript RNA2 resulted in due to the presence of a small admixture of virion RNA1 in small (0.5 mm) clear plaques. Nevertheless, assay of K1 our preparations of virion RNA2 (10). virions or K1 virion RNAs resulted in large plaques indis- Similarly, several preparations of transcript RNA2 were tinguishable in morphology from those of W17 virions. We assayed together with virion RNA1 at concentrations optimal judge that the plaque morphology engendered by RNA1 for virion RNAs. Table 3 shows that two RNA2 transcripts transcripts is a transient consequence of their terminal independently derived from the pSP64 vector and one de- structure. However, we do not rule out the possibility that rived from the pPM1 vector had equal infectivity, about 8% viral sequence changes were introduced in the course of of that obtained with virion RNA2. Their similar infectivity construction oftranscript RNA1 and that these changes were suggests that under these conditions the additional 20 reversed by mutation during the proliferation of K1. nonviral bases on the 5' ends of the transcripts derived from Virion and Virion RNA Comparisons. The virions ofK1 and pSP64 did not have a strong negative influence on infectivity. its progenitor virus, W17 sedimented identically in sucrose We have not yet checked the validity of this conclusion for gradients. Their specific infectivities were similar [220 and pPM1-based RNA1 transcripts. Again it may be noted that 290 particles required per plaque-forming unit (pfu) for W17 the virion RNA1 preparation is slightly infectious in the and K1, respectively]. Antiserum raised against wild-type absence of added RNA2 because of the presence of a small BBV neutralized both viruses. Their virion RNA1 and RNA2 admixture of virion RNA2. migrated identically upon agarose gel electrophoresis. Their

Table 1. Terminal sequences of RNA1 and RNA2 RNA1 W17 vRNA1 7mGpppGUUUU ...... AGGU (protein) p1B13SP GpppGaauacaagcuugggcugcaGUUUU ...... AGGUcuagOH p1B13SP(M) M7GpppGaauacaagcuugggcugcaGUUUU ...... AGGUcuagOH p1B13SP(ppp) pppGaauacaagcuugggcugcaGUUUU ...... AGGUcuagOH K1 vRNA1 M7GpppGUUUU ...... (protein) RNA 2 W17 vRNA2 m7GpppGUAAA ...... AGGU (protein) p2B1OSP GpppGaauacaagcuugggcugcaGUAAA ...... AGGUcuagOH p2B1OPM8 m7GpppGUAAA ...... AGGUcuagOH K1 vRNA2 m7GpppGUAAA ...... (protein) Terminal sequences of the RNAs we have used are shown, with capital letters signifying viral nucleotides and lower-case letters signifying nonviral nucleotides. The existence of a protein at the 3' termini of the virion RNAs is an inference from the fact that the terminus is blocked and that, in wild-type BBV RNA, the block is partially removable by protease. p1B13SP(M) refers to a transcript having a methylated cap; plB13SP-(ppp) refers to a transcript having a noncapped 5'-terminal triphosphate. vRNA1 and vRNA2, virion RNA1 and RNA2. Downloaded by guest on September 29, 2021 Biochemistry: Dasmahapatra et al. Proc. Natl. Acad. Sci. USA 83 (1986) 65 Table 2. Infectivity of transcript RNA1 in the presence Table 4. Infectivity of transcript RNA1 (plB9SP) plus transcript of virion RNA2 RNA2 (p2B1OSP) pfu x 10-3 per pfu per 5 x Relative RNA1 5 x 106 cells Relative infectivity RNA1, ,ug RNA2, Ag 106 cells infectivity

None 0.063 0.03 0.5 None 0 0 p1B13SP 18.8 10.0 None 0.2 0 0 p1B13SP(M) 18.3 10.0 0.5 0.2 115 0.07 p1B13SP(ppp) 0.200 0.1 0.75 0.2 160 0.09 p1B9SP 18.0 10.0 1.0 0.2 283 0.16 vRNA1 188 100 0.75 0.3 210 0.12 Drosophila cells (5 x 106) in 100 1.l ofbuffer were transfected with 1.0 0.5 450 0.25 0.5 1Lg ofthe RNA1 preparations in conjunction with 0.2 ,ug of virion RNA2. After transfection, productively infected cells were scored by 0.5 (vRNA1) 0.2 (vRNA2) 175 x 103 100 plaque assay. RNA infectivities are normalized to the virion RNA1 Drosophila cells were transfected with transcript RNA1 plus (vRNA1) plus RNA2 combination. p1B13SP(M) refers to a transcript transcript RNA2 at the indicated concentrations. Productively in- having a methylated cap. p1B13SP(ppp) refers to a transcript having fected cells were scored as described in Table 2. In later experiments, a noncapped 5'-terminal triphosphate. higher infectivity was obtained (see Fig. 1). vRNA1 and vRNA2, virion RNA1 and RNA2. virion proteins a and 8 migrated identically upon NaDodSO4/polyacrylamide gel electrophoresis. consequence of structural differences at their 5' and 3' W17 and K1 virion RNA1 and RNA2 each have a 5'- termini. Nevertheless, they should provide an effective terminal cap and a 3'-terminal blocking group. Sequence vehicle for introduction of defined sequence changes into the analysis (Fig. 2) demonstrates that the additional 20 nonviral viral genome by the methods of recombinant DNA technol- bases present at the 5' ends of the transcript RNAs are not ogy. Moreover, the fact that the RNA1 transcripts are reproduced in the K1 virion RNAs. Except for base 15 in K1 functional allows study of their replication in the complete RNA1, both RNA1 and RNA2 of the K1 virion RNAs have absence of RNA2. Previously, studies of transfection with 5'-terminal sequences identical to their W17 counterparts. RNA1 alone were limited by the difficulty of completely Position 15 in K1 RNA1 and the corresponding position in eliminating RNA2, since even a minute admixture of virion transcript RNA1 (position 35) is a uridine residue while W17 RNA2 led to completion of the infectious process (10). RNA1 has a cytidine residue, indicating a transitional muta- The RNA1 terminal structures may be involved in the viral tion presumably introduced during cloning. Thus, 5' nonviral infectious cycle at both the level of translation and of nucleic sequences, existent in transcripts, are not reproduced in acid replication. RNA1 codes for protein A, which is essen- progeny viral RNA, but internal viral sequence changes introduced into transcripts are reproduced in the progeny virus they generate. We have not yet sequenced the 3' A proximal regions of these RNAs, but we have determined 400 that the 3' termini are resistant to addition of pCp with T4 RNA ligase, suggesting that the termini are blocked as are the W17 RNA 3' termini. 200 Thus by our present criteria, virus K1 and its progenitor virus W17 are phenotypically indistinguishable but have at least one genotypic difference. 2 4 0 B )o DISCUSSION C.) 200 0

We have shown that DNA forms of the genomic RNAs of x lOC)o BBV W17 can be transcribed into RNA that is infectious to cultured cells of D. tn melanogaster, yielding virus indistin- 0 guishable from W17. The transcript RNAs are less infectious 0. 0.5 1.0 1.5 2.0 than their W17 virion RNA counterparts, most likely as a

Table 3. Infectivity of transcript RNA2 in the presence of virion RNA1 pfu x 10-3 per RNA2 5 x 106 cells Relative infectivity None 0.15 0.09 p2B1OSP 14 8.3 p2B48SP 13 7.4 2 p2B1OPM8 15 8.6 RNA, ug vRNA2 180 100 FIG. 1. Effect of varying the input of transcript RNA1 (plB9SP) Drosophila cells were transfected with the 0.2Ig of the RNA2 and transcript RNA2 (p2BlOSP). (A) Input of transcript RNA1 was preparations in conjunction with 0.5 /ig ofvirion RNA1. Productively varied, whereas input of transcript RNA2 was 0.2 ,&g. (B) Input of infected cells were scored as described in Table 2. In later experi- transcript RNA2 was varied, whereas input of transcript RNA1 was ments, higher infectivity was obtained (see Fig. 1). vRNA2, virion 1.2 jig. (C) Input of transcript RNA1 and transcript RNA2 was RNA2. equimolar. Downloaded by guest on September 29, 2021 66 Biochemistry: Dasmahapatra et al. Proc. Natl. Acad. Sci. USA 83 (1986) It is clear that it is possible to produce BBV by presenting C -I I I I susceptible Drosophila cells with RNA having appropriate ACGlT ACGlT viral sequences flanked with nonviral structures existent only for ease in construction. Evidently, BBV and its host have a mechanism for identifying the initiation sites for minus- and plus-strand synthesis, even when these sites are not terminal. Thus, for viruses such as BBV, genome modification by means of a DNA intermediate can involve transcription systems that yield RNA transcripts lacking the precise terminal structures that exist on authentic viral RNA. With the availability of a readily modifiable genome, BBV and its host Drosophila comprise a system of enormous potential for elucidating the molecular biology of RNA viruses and for introduction of new proteins into a eukaryotic host. BBV grows to exceptionally high titer in cultured cells of D. melanogaster (17), an organism whose genetic charac- FIG. 2. Sequence analysis of the 5' end of BBV RNA1 transcript teristics are well understood. A reliable plaque assay is p1B9SP (lanes Tr) and RNA2 isolated from virus K1 (lanes K1) and available (7). Crystallography of BBV is well underway (18). its progenitor, W17 (lanes W17). A 32P-5'-end-labeled single-stranded Since RNA1 can replicate independently of RNA2 (10), DNA primer, complementary to bases 70-123 of BBV RNA1, was replicative functions can be studied in the absence of RNA2 used to prime cDNA synthesis from the RNAs in the presence of or with extensively modified RNA2. Similarly, modification dideoxynucleotides A, C, G, or T (14, 15). The arrow points to ofRNA2 can lead to an understanding ofthe functions ofcoat position 15, the site of the base that is different in the RNA1 of K1 protein and the process of assembly. Furthermore, it should and W17. be possible to replicate suitably engineered foreign messen- ger RNAs in the presence of RNA1. tial for viral RNA replication; thus, the RNA1 that initiates Note Added in Proof. Mizutani and Colonno (19) have very recently the infectious process must be translated effectively in order of infectious RNA of human rhinovirus 14. to provide a functional protein A. Therefore, RNA1 tran- reported synthesis type scripts p1B13SP, p1B13SP(M), and to a small extent We thank Prof. Roland R. Rueckert for discussions throughout the p1B13SP(ppp) must have this capability, and, in particular, course ofthis work. We thank Prof. Paul Ahlquist for discussions and they are not fatally flawed because ofthe existence of20 extra for providing his vector, pPM1. We also wish to acknowledge the 5' nonviral bases. These extra bases separate the 5' ends from conscientious technical assistance of Chris Nelson. This research the initiation codon by 58 nucleotides. Transcripts p1B13SP was supported by National Institutes of Health Grants AI1466, and p1B13SP(M) are equally infectious, but transcript A115342, and CA08662 and National Institutes of Health Career p1B13SP(ppp) has comparatively low infectivity. Evidently, Award A121942. the existence of cap is important, but its methylation state is 1. Taniguchi, T., Palmieri, M. & Weissmann, C. (1978) Nature not. Possibly, the principal need for the capping group in (London) 274, 223-228. these assays is to provide a barrier against nucleolytic 2. Racaniello, V. & Baltimore, D. (1981) Science 214, 916-919. degradation. 3. Ahlquist, P., French, R., Janda, M. & Loesch-Fries, S. (1984) The first functional opportunity for the 3' termini of the Proc. Natl. Acad. Sci. USA 81, 7066-7070. 4. Cress, D., Kiefer, M. & Owens, R. (1983) Nucleic Acids Res. incoming RNAs probably exists when an RNA-replicating 11, 6821-6835. complex has been established and synthesis of negative- 5. Ohno, T., Ishikawa, M., Takamatsu, N., Meshi, T., Okada, strand RNA is initiated. All of our transcripts carry four Y., Sano, T. & Shikata, E. (1983) Proc. Jpn. Acad. Ser. B 59, additional nonviral bases, the terminus of which has a free 251-254. hydroxyl group. We conclude that a blocked 3' terminus is 6. Longworth, J. & Carey, G. (1976) J. Gen. Virol. 33, 31-40. not essential for initiation ofinfectivity and that the additional 7. Selling, B. & Rueckert, R. (1984) J. Virol. 51, 251-253. bases do no substantial harm. We are unable to state 8. Dasmahapatra, B., Dasgupta, R., Ghosh, A. & Kaesberg, P. unequivocally that the four bases are not copied. However, (1985) J. Mol. Biol. 182, 183-190. the fact that the eventual virion RNAs have the precisely 9. Guarino, L., Ghosh, A., Dasmahapatra, B., Dasgupta, R. & Kaesberg, P. (1984) Virology 139, 199-203. correct 5'-terminal structure suggests that there is no copying 10. Gallagher, T. M., Friesen, P. D. & Rueckert, R. (1983) J. of the 20 nonviral bases putatively existing at the 3' termini Virol. 46, 481-489. of the RNA negative strands. 11. Dasgupta, R., Ghosh, A., Dasmahapatra, B., Guarino, L. & Since virion protein synthesis is a late event in virus Kaesberg, P. (1984) Nucleic Acids Res. 12, 7215-7223. multiplication (16), input RNA2 need not be translated 12. Ahlquist, P. & Janda, M. (1984) Mol. . Biol. 4, 2876-2882. directly. However, newly synthesized RNA2 must serve as 13. Konarska, M., Padgett, R. & Sharp, P. (1984) Cell 38, 731-736. a template for synthesis of functional messenger. The 5' 14. Zimmern, D. & Kaesberg, P. (1978) Proc. Natl. Acad. Sci. terminus of RNA2 transcript p2B1OPM8 is the same as that USA 75, 4257-4261. of W17 virion RNA2. Thus, its lower infectivity than virion 15. Ahlquist, P., Dasgupta, R. & Kaesberg, P. (1981) Cell 23, 183-189. RNA2 (Table 3) results from its imperfect 3' terminus (or 16. Friesen, P. D. & Rueckert, R. (1984) J. Virol. 49, 116-124. possibly from adventitious sequence changes introduced in 17. Friesen, P., Scotti, P., Longworth, J. & Rueckert, R. (1980) J. its construction). The similarity in infectivity between Virol. 35, 741-747. p2B1OPM8 and p2B1OSP suggests that the 20 nonviral bases 18. Hosur, M. V., Schmidt, T., Tucker, C., Johnson, J. E., Sell- present at the 5' terminus of the latter do not confer an ing, B. & Rueckert, R. R. (1984) Virology 133, 119-127. additional loss in infectivity. 19. Mizutani, S. & Colonno, R. (1985) J. Virol. 56, 628-632. Downloaded by guest on September 29, 2021