Proc. Natl. Acad. Sci. USAVol. 93, pp. 11131-11136, October 1996Microbiology

Synthetic transcripts of double-stranded Birnavirus genomeare infectious

(infectious bursal disease virus/double-stranded RNA virus/infectious RNA/reverse genetics)

EGBERT MUNDT*t AND VIKRAM N. VAKHARIAt§*Federal Research Center for Virus Disease of Animals, Friedrich-Loeffler-Institutes, Institute of Molecular and Cellular Virology, D-17498 Insel Riems,Germany; and tCenter for Agricultural Biotechnology, University of Maryland Biotechnology Institute and Virginia-Maryland Regional College of VeterinaryMedicine, University of Maryland, College Park, MD 20742

Communicated by Edwin D. Kilbourne, New York Medical College, Valhalla, NY July 24, 1996 (received for review February 29, 1996)

ABSTRACT We have developed a system for generation ofinfectious bursal disease virus (IBDV), a segmented double-stranded RNA virus of the Birnaviridae family, with the use ofsynthetic transcripts derived from cloned cDNA. Independentfull-length cDNA clones were constructed that contained theentire coding and noncoding regions of RNA segments A andB of two distinguishable IBDV strains of serotype I. SegmentA encodes all of the structural (VP2, VP4, and VP3) andnonstructural (VP5) proteins, whereas segment B encodes theRNA-dependent RNA polymerase (VP1). Synthetic RNAs ofboth segments were produced by in vitro transcription oflinearized plasmids with T7 RNA polymerase. Transfection ofVero cells with combined plus-sense transcripts of bothsegments generated infectious virus as early as 36 hr aftertransfection. The infectivity and specificity of the recoveredchimeric virus was ascertained by the appearance of cyto-pathic effect in chicken embryo cells, by immunofluorescencestaining of infected Vero cells with rabbit anti-IBDV serum,and by nucleotide sequence analysis of the recovered virus,respectively. In addition, transfectant viruses containing ge-netically tagged sequences in either segment A or segment Bof IBDV were generated to confirm the feasibility of thissystem. The development of a reverse genetics system fordouble-stranded RNA viruses will greatly facilitate studies ofthe regulation of viral gene expression, pathogenesis, anddesign of a new generation of live vaccines.

Infectious bursal disease virus (IBDV), a member of theBirnaviridae family, is the causative agent of a highly immu-nosuppressive disease in young chickens (1). Two distinctserotypes of IBDV, designated as serotype I and serotype II,have been identified. The IBDV genome consists of twosegments of double-stranded (ds)RNA that vary from 2827(segment B) to 3261 (segment A) nucleotide base pairs (2).The larger segment A encodes a polyprotein that is cleaved byautoproteolysis to form mature viral proteins VP2, VP3, andVP4 (3). VP2 and VP3 are the major structural proteins of thevirion. VP2 is the major host-protective immunogen of IBDVand contains the antigenic regions responsible for the induc-tion of neutralizing antibodies (4). A second open readingframe (ORF), preceding and partially overlapping the polypro-tein gene, encodes a protein (VP5) of unknown function thatis present in IBDV-infected cells (5). The smaller segment Bencodes VP1, a 90-kDa multifunctional protein with polymer-ase and capping enzyme activities (6, 7).Although the nucleotide sequences for genome segments A

and B of various IBDV strains have been published, it was onlyrecently that the complete 5'- and 3'-noncoding sequences ofboth segments were determined. The 5'-noncoding region ofIBDV segments A and B contains a consensus sequence of 32

nucleotides. The 3'-noncoding terminal sequences of bothsegments are unrelated, but they are conserved among IBDVstrains of the same serotype (2). These terminii may containsequences important in packaging and in the regulation ofIBDV gene expression, as demonstrated for other dsRNA-containing viruses such as mamamlian and plant reovirusesand rotaviruses (8-10).

In recent years, a number of infectious animal RNA viruseshave been generated from cloned cDNA using transcriptsproduced by DNA-dependent RNA polymerase (11). Forexample, poliovirus (a plus-stranded RNA virus), influenzavirus (a segmented negative-stranded RNA virus), and rabiesvirus (a nonsegmented negative-stranded RNA virus) all wererecovered from cloned cDNAs of their respective genomes(12-14). For reovirus, it was shown that transfection of cellswith a combination of single-stranded (ss)RNA, dsRNA, andin vitro translated reovirus products generated infectious re-ovirus when complemented with a helper virus of a differentserotype (15). However, to date, there is no report of arecovered infectious virus with a segmented dsRNA genomederived from synthetic RNAs only.

In an effort to develop a reverse genetics system for IBDV,we constructed two independent full-length cDNA clones thatcontain segment A of serotype I strain D78 and segment B ofstrain P2, respectively. Furthermore, these cDNA clones weremodified by oligonucleotide-directed mutagenesis to generatethe tagged sequences in each segment. Synthetic RNAs ofsegments A and B were produced by in vitro transcriptionreactions on linearized plasmids with T7 RNA polymerase.Transcripts of these segments, either untreated or treated withDNase or RNase, were evaluated for the generation of infec-tious virus by transfecting Vero cells.

MATERIALS AND METHODS

Viruses and Cells. Two serotype I strains of IBDV, theattenuated P2 strain from Germany and the vaccine strain D78(Intervet, Millsboro, DE), were propagated in chicken embryocells (CEC) and purified as described (2, 16). Vero cells weregrown in M199 medium supplemented with 5% fetal calfserum and used for transfection experiments. Further propa-gation of the recovered virus and immunofluorescence studieswere carried out in Vero cells as described (5). For plaqueassay, monolayers of secondary CEC were prepared and usedas described (17).

Abbreviations: ds, double stranded; ss, single stranded; CEC, chickenembryo cells; CPE, cytopathic effect; RT, reverse transcription; IBDV,infectious bursal disease virus.tVisiting scientist at: University of Maryland, College Park, MD20742.§To whom reprint requests should be addressed at: Center forAgricultural Biotechnology, College of Veterinary Medicine, Uni-versity of Maryland, College Park, MD 20742.

11131

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement" inaccordance with 18 U.S.C. §1734 solely to indicate this fact.

11132 Microbiology: Mundt and Vakharia

Construction of Full-Length cDNA Clones of IBDV Ge-nome. Full-length cDNA clones of IBDV segments A and Bwere independently prepared. The cDNA clones containingthe entire coding region ofRNA segment A of strain D78 wereprepared using standard cloning procedures and methods as

described (16). By comparing the D78 terminal sequences withrecently published terminal sequences of other IBDV strains(2), it was observed that D78 cDNA clones lacked the con-

served first 17 nucleotides at the 5' end and the last 10 nucleotidesat the 3' ends. Therefore, to construct a full-length cDNA cloneof segment A, two primer pairs (A5'-D78, A5-IPD78, andA3'D78, A3-4PD78) were synthesized and used for PCR ampli-fication (Table 1). The DNA segments were amplified accordingto the protocol of the supplier (New England Biolabs) usingDeepVent polymerase. Amplified fragments were cloned into theEcoRI site of a pCRII vector (Invitrogen) to obtain plasmidspCRD78A5' and pCRD78A3'. Each plasmid was digested withEcoRI and SalI, and the resultant fragments were ligated intoEcoRI digested-pUC19, to obtain plasmid pUC19FLAD78. Thisplasmid contains a full-length cDNA copy of segment A, whichencodes all of the structural proteins (VP2, VP4, and VP3), aswell as the nonstructural protein VP5 (Fig. 1).To obtain cDNA clones of segment B of P2 strain, two

primer pairs (B5'-P2, B5-IPP2, and B3'-P2, B3-IPP2) were

designed according to published sequences and used forreverse transcription (RT)-PCR amplification (see Table 1).Using genomic dsRNA as template, cDNA fragments were

synthesized and amplified according to the supplier's protocol(Perkin-Elmer). Amplified fragments were blunt-end ligatedinto SmaI-cleaved pBS vector (Stratagene) to obtain clonespBSP2B5' and pBSP2B3'. To construct a full-length clone ofsegment B, the 5'-end fragment of plasmid pBSP2B5' was firstsubcloned between EcoRI and PstI sites of the pUC18 vectorto obtain pUCP2B5'. Then the 3'-end fragment of plasmidpBSP2B3' was inserted between the unique BglII and PstI sitesof plasmid pUCP2B5' to obtain a full-length plasmidpUC18FLBP2, which encodes VP1 protein (Fig. 1). PlasmidspUC18FLBP2 and pUC19FLAD78 were completely se-

quenced by using the sequenase DNA sequencing system(United States Biochemical), and the sequence data wasanalyzed using either DNASIS (Pharmacia) or PC/GENE (Intel-liGenetics) software. The integrity of the full-length constructswere tested by in vitro transcription and the translation-coupled reticulocyte lysate system using T7 RNA polymerase(Promega).To introduce the sequence tags into segments A and B of

IBDV, plasmids pUC19FLAD78mut and pUC18FLBP2mutwere constructed by oligonucleotide-directed mutagenesis,using specific primer pairs and PCR amplification of their

respective cDNA templates. To construct plasmidpUC19FLAD78mut, three primer pairs [RsrIIF, NheA(-);NheA(+), SpeA(-); and SpeA(+), SaclIR; see Table 1] wereused to amplify the DNA fragments of 428, 655, and 623 bp,respectively. These fragments were combined and subse-quently reamplified by PCR using the flanking primers (RsrIIFand SacIIR) to produce a 17Q6-bp fragment. This fragment wascloned into a pCRII vector to obtain plasmid pCRNhe-Spe.This plasmid was digested with RsrII and Sacll enzymes, andthe resulting 1557-bp fragment was subcloned into uniqueRsrII and SacII sites of plasmid pUC19FLAD78. Finally, amutant plasmid of segmentAwas obtained, which contains theunique NheI and SpeI restriction sites (nucleotide positions 545and 1180, respectively). Similarly, plasmid pUC18FLBP2 was

modified by PCR using an oligonucleotide primer containingthree silent mutations. After amplification, the resulting frag-ment was digested with KpnI and BglII and cloned intoKpnI-BglII-cleaved pUC18FLBP2. A mutant plasmid of seg-ment B (pUC18FLBP2mut) was obtained, which contains thesequence tag (mutations at nucleotide positions 1770, 1773,and 1776).

Transcription and Transfection of Synthetic RNAs. Plas-mids pUC19FLAD78, pUC18FLBP2, and their mutant cDNAclones were digested with BsrGI and PstI enzymes (Fig. 1),respectively, and used as templates for in vitro transcriptionwith T7 RNA polymerase (Promega). Briefly, restriction en-zyme cleavage assays were adjusted to 0.5% SDS and incu-bated with proteinase K (0.5 mg/ml) for 1 hr at 37°C. Thelinearized DNA templates (-3 ,tg) were recovered afterethanol precipitation, and were added separately to a tran-scription reaction mixture (50 ,l) containing 40 mM Tris HCl(pH 7.9), 10 mM NaCl, 6 mM MgCl2, 2 mM spermidine, 0.5mM ATP, 0.5 mM CTP, 0.5 mM UTP, 0.1 mM GTP, 0.25 mMcap analog [m7G(5')ppp(5')G], 120 units RNasin, 150 units T7RNA polymerase (Promega), and incubated at 37°C for 1 hr.Synthetic RNA transcripts were purified by phenol/chloroform extraction and ethanol precipitation. As controls,the transcription products were treated with either DNase or

RNase (Promega) before the purification step.Vero cells were grown to 80% confluency in 60-mm dishes

and washed once with phosphate-buffered saline (PBS). Threemilliliters of OPTI-MEM I (GIBCO/BRL) were added to themonolayers, and the cells were incubated at 37°C for 1 hr in a

CO2 incubator. Simultaneously, 0.15 ml of OPTI-MEM I wasincubated with 12.5 jig of Lipofectin reagent (GIBCO/BRL)for 45 min in a polystyrene tube at room temperature. Syn-thetic RNA transcripts of both segments resuspended in 0.15ml of diethyl pyrocarbonate-treated water, were added to theOPTI-MEM/Lipofectin mixture, mixed gently, and incubated

Table 1. Oligonucleotides used for the construction of full-length cDNA clones of IBDV genomic segments A and B

NucleotideNucleotide sequence Orientation Name no.

TAATACGACTCACTATAGGATACGATCGGTCTGACCCCGGGGGAGTCA + A5'-D78 1-31TGTACAGGGGACCCGCGAACGGATCCAATT - A3'-D78 3237-3261CGTCGACTACGGGATTCTGG - A5-IPD78 1711-1730AGTCGACGGGATTCTTGCTT + A3-IPD78 1723-1742ATGACAAACCTGCAAGAT + RsrIIF 131-148CTGACAGATGCTAGCTACAATGGG - + NheA(+) 536-559GTCCCGTCACACTAGTGGCCTA + SpeA(+) 1170-1191CCTCTCTTAACACGCAGTCG - SacIIR 1774-1793AGAGAATTCTAATACGACTCACTATAGGATACGATGGGTCTGAC + B5'-P2 1-18CGATCTGCTGCAGGGGGCCCCCGCAGGCGAAGG - B3'-P2 2807-2827CTTGAGACTCTTGTTCTCTACTCC - B5-IPP2 1915-1938ATACAGCAAAGATCTCGGG + B3-IPP2 1839-1857

Composition and location of the oligonucleotide primers used for cloning. T7 promoter sequences are marked with italictype, the virus specific sequences are underlined, and the restriction sites marked in boldface type. Orientation of thevirus-specific sequence of the primer is shown for sense (+) and antisense (-). The positions where the primers bind(nucleotide number) are according to the published sequences of P2 strain (2).

Proc. Natl. Acad. Sci. USA 93 (1996)

Proc. Natl. Acad. Sci. USA 93 (1996) 11133

GAATTCGGCTTTAATACGACTCACTATAGGATACGATCGGTCTGAC P T-TP2ACCTAGGCAGCGGTCCCC GTAGACAGGGTG-CCCTTAAGCCGAAATTATGCTGAGTGATATCCTATGCTAGCCAGACTG TTAACCTAGGCAAGCGCCCAGGGGACAG TTTCGGCTTAAG

Transcription_ O

TBarGI

EcoRI

TTGCATGCCTGCA GGGGGCCCCCGCAGGCGAAG LTCGTATCCTATAGTGAGTCGTATTAGAATTCAACGTACGG ACGTCCCCCGGGGGCGTCCGCTTC I I AGCATAGGATATCACTCAGCATAATCTTA&G

Pst 1 Transcription EcoRI

FIG. 1. Schematic diagram of cDNA constructs used for synthesis of plus-sense ssRNAs of IBDV with T7 RNA polymerase. ConstructpUC19FLAD78 contains the full-length cDNA of segment A of IBDV strain D78. Segment A of IBDV encodes the polyprotein (VP2-VP4-VP3)and the recently identified VP5 protein. Plasmid pUC18FLBP2 contains the cDNA of segment B of strain P2, which encodes the RNA-dependentRNA polymerase (VP1). Virus specific sequences are underlined and the T7 promoter sequences are italicized. Restriction sites are shown inboldface type and identified. The cleavage sites of the linearized plasmids are shown by vertical arrows and the transcription directions are markedby horizontal arrows.

on ice for 5 min. After removing the OPTI-MEM from themonolayers in the 60-mm dishes and replacing it with a fresh1.5 ml of OPTI-MEM, the nucleic acid-containing mixture wasadded drop-wise to the Vero cells and swirled gently. After 2hr of incubation at 37°C, the mixture was replaced with M199medium containing 5% FCS (without rinsing the cells), and thecells were further incubated at 37°C for desired time intervals.

Identification of Generated IBDV. CEC were infected withthe supernatant from Vero cells transfected with transcripts ofeither pUC18FLAD78, pUC18FLBP2, and/or their mutantplasmids. About 16 hr after infection, the whole cell nucleicacids were isolated as described (2). Specific primers ofsegments A and B were used for RT of genomic RNA derivedfrom these transfectant viruses. Following RT, the reactionproducts were amplified by PCR using specific primer pairs ofsegments A and B. Resulting PCR fragments were eithercloned and sequenced as described before, or digested withappropriate restriction enzymes to identify the tagged sequences.

Immunofluorescence. Vero cells, grown on coverslips to80% confluency, were infected with the supernatants derivedfrom transfected Vero cells (after freeze thawing) and incu-bated at 37°C for 2 days. The cells were then washed, fixed withacetone, and treated with polyclonal rabbit anti-IBDV serum.After washing, the cells were treated with fluorescein labeledgoat anti-rabbit antibody (Kirkegaard & Perry Laboratories)and examined by fluorescence microscopy.Plaque Assay. Monolayers of secondary CEC, grown in

60-mm dishes, were inoculated with the supernatants derivedfrom transfected Vero cells. One hour after infection, the cellswere washed once with PBS, and overlaid with 0.8% nobleAgar (Difco) containing 10% tryptose phosphate broth, 2%FCS, 0.112% NaHCO3, 103 units penicillin, 103 ,tg/ml strep-

tomycin, 0.25 ,tg/ml fungizone, 0.005% neutral red, and0.0015% phenol red. The cells were incubated at 37°C for 2-3days until plaques could be observed and counted (17).

RESULTS

Construction of Full-Length cDNA Clones of IBDV Ge-nome. To develop a reverse genetics system for the dsRNAvirus IBDV, two independent cDNA clones were constructedthat contain segment A of strain D78 and segment B of strainP2 (Fig. 1). Each plasmid encoded either the precursor ofstructural proteins (VP2, VP4, VP3) and VP5 or only VP1protein (RNA-dependent RNA polymerase). PlasmidpUC18FLBP2, upon digestion with PstI and transcription invitro by T7 RNA polymerase, would yield RNA containing thecorrect 5' and 3' ends. Similarly, upon digestion with BsrGIand transcription, plasmid pUC19FLAD78 would yield RNAcontaining the correct 5' end, but with an additional fournucleotides at the 3' end. Coupled transcription and transla-tion of the above plasmids in a rabbit reticulocyte systemyielded protein products that were correctly processed. Theseproducts comigrated with the marker IBDV proteins afterfractionation on SDS/polyacrylamide gel and autoradiogra-phy (data not shown).

Transcription, Transfection, and Generation of a ChimericInfectious Virus. Plus-sense transcripts of serotype I IBDVsegments A and B were synthesized separately in vitro with T7RNA polymerase, using linearized full-length cDNA plasmidsas templates (Fig. 2). Although two species ofRNA transcriptswere observed for segment B on a neutral gel (lanes 1 and 5),fractionation of these samples on a denaturing gel yielded onlyone transcript-specific band (data not shown). To show that

EcoRI

Microbiology: Mundt and Vakharia

11134 Microbiology: Mundt and Vakharia

1 2 3 4 5 6 M kbp

21.0

1.6

0.83

0.56

FIG. 2. Analysis of the transcription reaction products that were

used for transfection of Vero cells. Synthetic RNAs transcribed in vitro

using T7 RNA polymerase and linearized plasmids pUC19FLAD78

containing the cDNA of segment A of IBDV strain D78 (lanes 2,4, and

6) and pUC18FLBP2 containing the cDNA of segment B of strain P2

(lanes 1, 3, and 5), respectively. After transcription, the reaction

mixtures were either treated with DNase (lanes 1 and 2), RNase (lanes3 and 4), or left untreated (lanes 5 and 6). The reaction products (2 LIi)were analyzed on 1% agarose gel. A DNA, digested with HindllllEcoRI, was used as a marker (lane M).

plus-sense RNA transcripts of both segments are needed for

the generation of infectious virus, the transcription mixtures

were incubated with different nucleases, as shown in Fig. 2.

Synthetic RNAs recovered after treatment of the transcription

products with DNase (lanes 1 + 2), RNase (lanes 3 + 4), or

without treatment (lanes 5 + 6) were used for the transfection

of Vero cells. As mock control, Lipofectin alone was used. Five

days after transfection, cytopathic effect (CPE) was onlyvisible in Vero cells transfected with combined transcripts of

untreated or DNase-treated transcription products, but not

with RNase-treated transcription mixtures or mock-

transfected control. No CPE was detected when Vero cells

were transfected with RNA of only segment A or segment B

(data not shown). These results show that replication of IBDV

ensued after transfection of Vero cells with plus-sense ssRNAs

of both segments of IBDV. To verify that the agent causing the

CPE in Vero cells was indeed IBDV, transfected Vero cells

were freeze-thawed, and cell-free supernatants were used to

infect CEC or Vero cells. CEC infected with the supernatants

derived from Vero transfected cells of untreated or DNase-

treated transcription mixtures produced CPE in 1 day after

inoculation (Table 2). However, no CPE could be detected in

CEC with the supernatants from either transfected Vero cells

of RNase-treated transcription mixtures, untreated segments

A or B transcription mixtures, or mock-transfected Vero cells,

even 5 days after infection. Similarly, when Vero cells on

coverslips were infected with the same supernatants as de-

scribed above and examined by immunofluorescence stainingafter 2 days, only supernatants derived from transfected Vero

cells of untreated or DNase-treated transcription mixtures

gave a positive immunofluorescence signal (Table 2).Recovery of Transfectant Virus. To determine the time

point for the recovery of infectious virus, Vero cells were

Table 2. Generation of infectious IBDV from synthetic RNAs ofsegments A and B

Material transfected CPE Immunofluorescence

ssRNA A + B, DNase treated + +ssRNA A + B, RNase treated - -ssRNA A + B, untreated + +ssRNA A, untreatedssRNA B, untreatedLipofectin only

Vero cells were transfected with synthetic RNAs of segments A andB, derived from transcription reactions, that were either untreated ortreated with DNase or RNase. After 5 days, the supernatants werecollected, clarified by centrifugation, and analyzed for the presence ofvirus. The infectivity of the recovered virus was detected in CEC bythe appearance of CPE 1-2 days after inoculation. The specificity ofthe recovered virus was determined by immunofluorescence stainingof infected Vero cells with rabbit anti-IBDV serum.

transfected with combined RNA transcripts of segments A andB. At 4, 8, 16, 24, 36, and 48 hr after transfection, thesupernatants were examined for the presence of transfectantvirus by infectivity and plaque assays, as shown in Table 3. Ourresults indicate that the virus could be recovered as early as 36hr after transfection. Virus titer was 2.3 x 102 plaque-formingunits/ml, and appeared to drop in samples obtained later than48 hr after transfection. However, after a second passage inCEC, the rescued chimeric virus grew to titers comparable tothose of "natural" serotype I IBDVs.

Identification of Sequence Tags. To demonstrate the utilityof the reverse genetics system, two recombinant IBDVs weregenerated, which contain sequence tags in segments A and B,respectively. In plasmid pUC19FLAD78mut, the IBDV nucle-otide sequence was altered by oligonucleotide-directed mu-tagenesis to create unique NheI and SpeI restriction sites in theVP2 gene. Synthetic transcripts of this modified segment A ofstrain D78 and segment B of strain P2 were then used totransfect Vero cells. To verify that these sites were present inthe recovered virus, the genomic RNA was purified from thevirions and analyzed after RT-PCR or PCR amplificationsusing segment A specific primers. The presence of the genetictag in the transfectant virus was confirmed by digestion of thePCR products with the appropriate restriction enzymes, asshown in Fig. 3. Sequences from mock-infected Vero cells andtransfectant viruses (with and without the genetic tag insegment A) were reverse transcribed and amplified in parallel.An 1184-bp fragment was obtained from both transfectantviruses (lanes 5 and 6) but not from Vero cells (lane 4). Noproduct was obtained when reverse transcriptase was omittedfrom the reactions before PCR (lanes 1-3), indicating that thePCR product was derived from RNA, not from contaminatingDNA. After digestion with NheI and Spel, expected fragmentsof 769 bp and 415 bp (lane 8), and 1049 bp (lane 10; 135 bp not

Table 3. Recovery of virus at various times after transfection

Time in hoursafter transfection CPE Immunofluorescence pfu/ml

4 - - 08 - - 0

16 - - 024 - - 036 + + 2.3 x 10248 + + 6.0 x 101

Vero cells were transfected with synthetic RNAs of segments A andB as described. The infectivity and specificity of the recovered viruswas detected by CPE in CEC and immunofluorescence staining inVero cells, respectively. Monolayers of secondary CEC were used forplaque assay after inoculating the cells with the supernatants derivedfrom transfected Vero cells. Approximate titer of the virus wascalculated as plaque forming units per ml (pfu/ml).

Proc. Natl. Acad. Sci. USA 93 (1996)

Proc. Natl. Acad. Sci. USA 93 (1996) 11135

-RT +Rr''II Nhel

NI 1 2 3 4 5 6 7 8

Spel9

9 10 Xlkbp

21.0

5.0

2.01.61.3

BP2.., ,..

............ ........ . . ..._ .. *-

:Ft.

.....::

._w..

............ ,....;,....

jtt4Stit ...............

_.

_ii P.t

vi_t::!.::-

itid:J$..i...$.

*........"::.::.:

0.980.83

0.56S.g

A* C G T

FIG. 3. Analysis of the RT-PCR products for identification of thesequence tags in segment A of the transfectant viruses. Genomic RNAisolated from transfectant viruses was amplified by RT-PCR usingsegment A-specific primers 5'-ATGACAAACCTGCAAGAT-3' (nu-cleotide positions 131-148) and 5'-CATGGCTCCTGGGT-CAAATCG-3' (nucleotide positions 1295-1315); the products were

analyzed on 1% agarose. An 1184-bp fragment was obtained from bothsamples containing transfectant viruses (lanes 5 and 6), but not fromthe Vero cells (lane 4) or the controls in which reverse transcriptasewas omitted from the reaction (lanes 1-3). Purified RT-PCR frag-ments, derived from the transfectant viruses, were digested with NheIand SpeI restriction enzymes, as indicated (lanes 7-10). Only the DNAfragment derived from the tagged virus was able to be digested, thusverifying the presence of these two restriction sites (lanes 8 and 10),whereas the one derived from the control transfectant virus remainsundigested (lanes 7 and 9). A DNA, digested with HindIll/EcoRI, was

used as a marker (lane M).

shown) were obtained from the tagged-viral RNA segment A,whereas the PCR product derived from the unmodified viralRNA remained undigested (lanes 7 and 9).

Similarly, the IBDV nucleotide sequence in segment B wasaltered by oligonucleotide-directed mutagenesis to introducethree silent mutations at nucleotide positions 1770 (G -> C),1773 (T -- C), and 1776 (G -- C) in the VP1 gene. Vero cells

were transfected with the synthetic transcripts derived fromplasmids pUC18FLBP2mut and pUC19FLAD78, and thetransfectant virus was recovered as described. Simultaneously,the unmodified transfectant virus was also recovered. Aftertwo passages in CEC, the genomic RNA of these viruses wascloned by RT-PCR using a specific primer pair of segment B,and then sequenced. No product was obtained when reversetranscriptase was omitted from the reactions before PCR,indicating that the derived RT-PCR product was not the resultof contaminating DNA (data not shown). Sequence analysis ofthe cloned fragment of transfectant viruses exhibits the ex-pected nucleotide mutations, as shown in Fig. 4. These resultsunequivocally show that the sequence tags are present in thegenomic RNA of the recovered viruses.

DISCUSSIONIn this report, we demonstrate for the first time, to the best ofour knowledge, that synthetic transcripts derived from cloned

BP2mut^_.

,W,j,¢ . . . i ... . £.. ,¢, .. ,_:

......

.........* $i-ii-- ... ,,,,B MjiS.'j i

_ ... ..._S. XSWF .. #

.__: .. ....... :................. ... <

.. ,pj.}.*.. ..it_St

*:

.:_

tii,,:::.>}8!E.. <,. t.4b:., t. .:-:':.. ......... ..

t .. :::

:io!w.*,.,. Xbt:

Ct::

:=. .....

it,p,,jUni..L........ .. :: ''..la, wr

*t, t':_w.vX'.' fi::Ct;: . jj#

_:

_

A Cz G TFIG. 4. Autoradiogram showing the nucleotide sequences of

cloned RT-PCR fragments from segment B of the unmodifed (BP2)and modified (BP2mut) transfectant viruses. Silent mutations atnucleotide positions 1770 (G -* C), 1773 (T -> C), and 1776 (G -> C)are indicated by arrows.

DNA, corresponding to the entire genome of a segmenteddsRNA animal virus, can give rise to a replicating virus. Therecovery of infectious virus after transfecting cells with syn-thetic plus-sense RNAs, derived from cloned cDNA of a viruswith dsRNA genome (IBDV), completes the quest of gener-ating reverse infectious systems for RNA viruses. A number ofinvestigators have generated infectious RNA viruses fromcloned cDNA using synthetic transcripts (11). Racaniello andBaltimore were first to rescue poliovirus, a plus-stranded RNAvirus, using cloned cDNA (12). Later, van der Werf et al. (18)generated infectious poliovirus using synthetic RNA producedby T7 RNA polymerase on a cloned cDNA template. Enamiet al. (13) rescued influenza virus, a segmented negative-stranded RNA virus; and Schnell et al. (14) generated rabiesvirus, a nonsegmented negative-stranded RNA virus, fromcloned cDNAs of their respective genomes. Chen et al. (19)demonstrated that the electroporation of fungal spheroplastswith synthetic plus-sense RNA transcripts, which correspondto the nonsegmented dsRNA hypovirus, an uncapsidatedfungal virus, yields mycelia that contain cytoplasmic-replicating dsRNA. Roner et al. (15) developed an infectioussystem for a segmented dsRNA reovirus by transfecting cellswith a combination of ssRNA, dsRNA, in vitro translatedreovirus products, and complemented with a helper virus ofdifferent serotype. The resulting virus was discriminated fromthe helper virus by plaque assay. However, in this system, theuse of a helper virus was necessary. In contrast, the describedreverse genetics system of IBDV does not require a helpervirus or other viral proteins. Transfection of cells with plus-sense RNAs of both segments was sufficient to generateinfectious virus (IBDV). In this regard, the system was com-parable to other rescue systems of plus-stranded poliovirus anddouble-stranded hypovirus (18, 19). The fate of the additionalone and four nucleotides, respectively, transcribed at the 3' endof segment A, was not determined. Obviously, this did not

Microbiology: Mundt and Vakharia

11136 Microbiology: Mundt and Vakharia

prevent the replication of the viral dsRNA. Similar effects havebeen observed in plus-stranded RNA viruses by differentinvestigators (1 1).

Transfection of plus-sense RNAs from both segments intothe same cell was necessary for the successful recovery ofIBDV. Transfected RNAs of both segments had to be trans-lated by the cellular translation machinery. The polyprotein ofsegment A was presumably processed into VP2, VP3, and VP4proteins, which form the viral capsid. The translated proteinVP1 of segment B probably acted as an RNA-dependent RNApolymerase and transcribed minus-strands from synthetic plus-strands of both segments, and the reaction products formeddsRNA. Recently, Dobos reported that in vitro transcription bythe virion RNA-dependent RNA polymerase of infectiouspancreatic necrosis virus, a prototype virus of the Birnaviridaefamily, is primed by VP1 and then proceeds via an asymmetric,semiconservative, strand-displacement mechanism to synthe-size only plus-strands during replication of the viral genome(20). Our system showed that synthesis of minus strands mustproceed on the plus strands. Whether the resulting transcribedminus-strand RNA serves as a template for the transcription ofplus-strands or not remains the subject of further investigations.To unequivocally prove that the infectious virus (IBDV)

contained in supernatants of transfected cells was indeedderived from the synthetic transcripts, two recombinant vi-ruses were generated containing sequence tags in eithersegment A of strain D78 or segment B of strain P2. Restrictionenzyme digests of the RT-PCR products and sequence analysisof the cloned DNA fragments verified the presence of thesesequence tags in the genomic RNA segments.The recovery of infectious virus (IBDV) demonstrated that

only the plus-strand RNAs of both segments were sufficient toinitiate replication of dsRNA. Thus, results are in agreementwith the general features of reovirus and rotavirus replication,where the plus-strand RNAs serve as a template for thesynthesis of progeny minus-strands to yield dsRNA (21-23).However, the semiconservative strand displacement mecha-nisms proposed by Spies et al. (6) and Dobos (20) could not beexcluded. The development of a reverse genetics system forIBDV will greatly facilitate future studies of gene expression

and pathogenesis, and help in the design of a new generationof live IBDV vaccines.

We thank Drs. Donald L. Nuss, Thomas Mettenleiter, and DieterLutticken for reviewing the manuscript and Kun Yao, Gerard H.Edwards, and Peter K. Savage for technical assistance. This work wassupported by a grant from Intervet International, B.V., Boxmeer, TheNetherlands to V.N.V.

1. Becht, H. (1980) Curr. Top. Microbiol. Immunol. 90, 107-121.2. Mundt, E. & Muiller, H. (1995) Virology 209, 10-18.3. Hudson, P. J., McKern, N. M., Power, B. E. & Azad, A. A. (1986)

Nucleic Acids Res. 14, 5001-5012.4. Azad, A. A., Jagadish, M. N., Brown, M. A. & Hudson, P. J.

(1987) Virology 161, 145-152.5. Mundt, E., Beyer, J. & Muller, H. (1995) J. Gen. Virol. 76,

437-443.6. Spies, U., Muller, H. & Becht, H. (1987) Virus Res. 8, 127-140.7. Spies, U. & Muller, H. (1990) J. Gen. Virol. 71, 977-981.8. Anzola, J. B., Xu, Z., Asamizu, T. & Nuss, D. L. (1987) Proc.

Natl. Acad. Sci. USA 84, 8301-8305.9. Zou, S. & Brown, E. G. (1992) Virology 186, 377-388.

10. Gorziglia, M. I. & Collins, P. L. (1992) Proc. Natl. Acad. Sci. USA89, 5784-5788.

11. Boyer, J. C. & Haenni, A. L. (1994) Virology 198, 415-426.12. Racaniello, V. R. & Baltimore, D. (1981) Science 214, 916-919.13. Enami, M., Luytjes, W., Krystal, M. & Palese, P. (1990) Proc.

Natl. Acad. Sci. USA 87, 3802-3805.14. Schnell, M. J., Mebatsion, T. & Conzelmann, K. K. (1994) EMBO

J. 13, 4195-4205.15. Roner, M. R., Sutphin, L. A. & Joklik, W. K. (1990) Virology 179,

845-852.16. Vakharia, V. N., He, J., Ahamed, B. & Snyder, D. B. (1994) Virus

Res. 31, 265-273.17. Muller, H., Lange, H. & Becht, H. (1986) Virus Res. 4, 297-309.18. van der Werf, S., Bradley, J., Wimmer, E., Studier, F. W. & Dunn,

J. J. (1986) Proc. Natl. Acad. Sci. USA 83, 2330-2334.19. Chen, B., Choi, G. H. & Nuss, D. L. (1994) Science 264, 1762-

1764.20. Dobos, P. (1995) Virology 208, 10-25.21. Schonberg, M., Silverstein, S. C., Levin, D. H. & Acs, G. (1971)

Proc. Natl. Acad. Sci. USA 68, 505-508.22. Patton, J. T. (1986) Virus Res. 6, 217-233.23. Chen, D., Zeng, C. Q.-Y., Wentz, M. J., Gorziglia, M., Estes,

M. K. & Ramig, R. F. (1994) J. Virol. 68, 7030-7039.

Proc. Natl. Acad. Sci. USA 93 (1996)