CA2330086A1 - Chimeras of hepatitis c virus and bovine viral diarrhea virus - Google Patents

Abstract

Disclosed is a polynucleotide comprising a chimeric viral RNA which contains: a 5' nontranslated region (5' NTR), an open reading frame (ORF) region, and a 3' nontranslated region (3' NTR) wherein at least one of said regions is chimeric. The chimeric region comprises a first nucleotide sequence from a pestivirus in operable linkage with a heterologous nucleotide sequence. The chimeric viral RNA is replication-competent. Preferably the pestivirus sequence is from a bovine viral diarrhea virus and the heterologous nucleotide sequence is from a hepatitis C virus. Also disclosed are a method for identifying compounds having antiviral activity agaisnt hepatitis C virus, a genetically-engineered chimeric RNA virus and a vaccine against bovine viral diarrhea virus.

Description

Chimeras of Hepatitis C Virus and Bovine Viral Diarrhea Virus Reference to Government Grant This invention was made with government support under a grant from the National Institutes of Health, grant numbers PHS CA57973 and AI40034. The government has certain rights in this invention.
Related Applications This application claims priority to, and incorporates herein in its entirety, U.S.
60/082,964 filed April 24, 1998..
Background of the Invention ( 1 ) Field of the Invention This invention relates generally to the development of therapies for treating hepatitis C virus (HCV) and bovine viral diarrhea virus (BVDV) and more particularly to the identification of such therapies using chimeric viruses comprising a genomic sequence derived from HCV and bovine viral diarrhea virus (BVDV).

(2) Description of the Related Art The Flavivirdae is an important family of human and animal RNA viral pathogens (Rice, CM. 1996. Flavivlrdae: The viruses and their replication. In: Fields BN, Knipe DM, Howley PM., eds. Fields virology. Philadelphia: Lippincott-Raven Publishers.
pp. 931-960.) The three currently recognized genera of the Flavivirdae family exhibit distinct differences in transmission, host range, and pathogenesis. For example, members of the classical flavivirus genus, such as yellow fever virus and dengue virus, are typically transmitted to vertebrate hosts via arthropod vectors and cause acute self limiting disease (Monath TP, Heinz FX.
1996. Flaviviruses. In: Fields BN, Knipe DM, Howiey PM., eds. Fields virology.
New York:
Raven Press. pp. 961-1034). The pestiviruses, such as bovine viral diarrhea virus (BVDV) and classical swine fever virus (CSFV), cause economically important livestock disease and are spread by direct contact or the fecal-oral route (Thiel et al., 1996.
Pestiviruses. In: Fields BN, Knipe DM, Howley PM., eds. Fields virology. New York: Raven Press. pp.
1059-1073).
The most recently characterized Flavivirdae genus is the hepacivirus genus, the sole member of which is the common and exclusively human pathogen, hepatitis C virus (HCV). HCV is transmitted by contaminated blood or blood products and is the most common agent of non-A, non-B hepatitis, affecting more that 1% of the population worldwide (Houghton, 1996.
Hepatitis C viruses. In: Fields BN, Knipe DM, Howley PM., eds. Fields virology.
Philadelphia: Lippincott-Raven Publishers. pp. 1035-1058.). Unlike flavivirus and pestivirus infections, which are usually eliminated by host immune response, chronic HCV
infections are common and can cause mild to severe liver disease including cancer.
Despite these differences, members of the Flavivirdae family share common structural features and gene expression strategies. Virus particles consist of a lipid bilayer envelope with embedded transmembrane glycoproteins surrounding a protein-RNA
nucleocapsid. Genome RNAs are single-stranded of positive polarity, and function as the sole mRNA species for translation of a single long open reading frame (ORF). This ORF is translated into a polyprotein which is processed by cellular and viral proteases into mature viral proteins. Structural proteins destined for incorporation into virus particles are encoded in the N-terminal portion of the polyprotein, while the nonstructural proteins which form components of the viral RNA replicase are encoded in the remainder.
Replication of the Flavivirdae RNA genome occurs via synthesis of a full-length negative-strand intermediate and is asymmetric, favoring synthesis of positive-strand RNAs.
However, little is known about the details of this process. For all three genera of the Flavivirdae family, full-length functional cDNA clones have been constructed and RNAs transcribed from these cDNA templates are infectious. For flaviviruses and pestiviruses, mutagenesis of these clones and efficient RNA transfection of permissive cell cultures provides a means of probing the role of cis RNA elements and viral proteins in replicase assembly and function. Such analyses are not yet possible for HCV since this virus is unable to replicate efficiently in cell culture.
Like many other RNA viruses, it is believed the 5' and 3' terminal sequences of the Flavivirdae contain conserved cis-elements important for translation, RNA
replication, and packaging (Bukh et al., Proc. Nutl. Acad. Sci. USA 89:4942-4946, 1992; Deng et al., Nucleic Acids Res. 21:1949-1957, 1993; Cahour et al., Virol. 207:68-76, 1995;
Kolykhalov et al., J.
Virol. 70:3363-3371, 1996; Men et al., J. Yirol. 70:3930-3937, 1996; Tanaka et al., J. Virol.
70:3307-3312, 1996; Huang HV. 1997. Evolution of the alphavirus promoter and the cis-acting sequences of RNA viruses. In: Saluzzo J-F, Dodet B. eds. Factors in the emergence of arbovirus disesases. Paris: Elsevier Press, pp. 65-79; Mandl et al., J. Yirol.
72:2132-2140, 1998). The 5' nontranslated region (NTR) functions initially at the level of translation.
Similar to most cellular mRNAs, flavivirus genome RNAs are translated in a cap-dependent manner. These RNAs contain a S' cap structure that is presumably added by virus-encoded RNA triphosphatases, guanylyl-, and methyl-transferases (Rice, 1996, supra).
In contrast, the translational strategy employed by pestiviruses and HCV is more similar to that of the picornaviruses. These RNAs appear to be uncapped and contain long S' NTRs with cis RNA
elements that function as internal ribosome entry sites (IRES) for translation initiation at the polyprotein AUG (Lemon et al., Semin. Virol. 8:274-288, 1997).
The 5' NTRs of HCV and BVDV have a similar structural and functional organization despite containing only short stretches of high sequence identity (Wang et al., Curr. Top.
Microbiol Immunol. 203:99-115, 1995; Lemon et al., 1997, supra). The IRES
within each NT'R is located at the 3' end of the NTR at a position proximal to the AUG
initiation codon of the ORF. Although the 5' terminal sequence of each of these viruses is apparently not required for IRES function (Rijnbrand et al., FEBSLett 365:115-119, 1995;
Honda et al., Virology. 222:31-42, 1996; Rijnbrand et al., J. Virol. 71:451-457, 1997), these sequences are highly conserved among different strains of HCV (Bukh et al., Proc. Natl.
Acad. Sci.
USA:89:4942-4946, 1992) or BVDV (Deng et al., 1993, supra), suggesting they play other roles in viral replication. For example, sequences in the 5' NTR may be required for regulating translation versus initiation of negative-strand RNA synthesis.
Such regulation could occur by direct interaction of S' and 3' RNA elements or indirectly, via RNA-protein interactions. Sequences in the 5' NTR may also modulate packaging versus translation.
Finally, sequences complementary to the 5' NTR, which are located at the 3' end of negative-strand RNA, are likely to function in the initiation of positive-strand RNA
synthesis.
The HCV 3' NTR contains an internal polypyrimidine tract followed by a highly conserved sequence of 98 bases at the 3' terminus, which has been shown to be required for replication of HCV (U.S. Application Serial No. 08/811,566).
Further elucidation of the role of sequences in the HCV 5' and 3' NTRs has been hampered by the inefficient replication of HCV in cell culture. This aspect of HCV biology also makes it difficult to identify and test possible antiviral compounds for activity against HCV. Thus, a need exists for a system which facilitates investigation of HCV
replication and therapeutic approaches to control HCV infections.
Summary of the Invention Briefly, therefore, the present invention provides novel compositions and methods for studying HCV replication which are based on the discovery that chimeras of HCV
and BVDV
genomic sequences can be constructed that are able to replicate in cell culture. The BVDV-specific sequence provides the chimeric viral nucleic acid with the ability to replicate in cell culture, while the HCV-specific sequence allows the chimeric viral nucleic acid to be used to screen possible compounds for anti-viral activity against HCV. It is believed that similar replication-competent chimeras can be constructed from HCV and other pestiviruses.
Thus, in one embodiment, the present invention provides a novel, chimeric viral RNA
in which at least one of the 5' NTR; ORF and 3' NTR regions is chimeric and comprises a nucleotide sequence from the corresponding region of a pestivirus in operable linkage with a nucleotide sequence from the corresponding region of an hepatitis C virus (HCV). The chimeric viral RNA is replication-competent. In preferred embodiments, the pestivirus is BVDV.
In other embodiments, the invention provides a polynucleotide comprising a DNA-dependent promoter operably linked to a cDNA of a chimeric viral RNA as described above and cells transiently transfected or stably transformed with the polynucleotide. In some embodiments the cDNA may encode a dominant selectable marker or an assayable reporter.
In yet another embodiment, the invention provides a method for identifying compounds having anti-HCV activity. The method comprises providing a first cell containing a chimeric viral nucleic acid derived from HCV and a pestivirus as described above and a second cell containing the pestivirus, and then comparing the replication efficiency of the chimeric viral nucleic acid in the presence and absence of a test compound to the replication efficiency of the pestivirus in the presence and absence of the test compound, wherein a greater reduction in compound-induced replication efficiency of the chimeric viral nucleic acid than the pestivirus indicates the compound has anti-HCV activity.
The invention also provides a genetically-engineered virus which comprises a chimeric viral nucleic acid derived from HCV and a pestivirus as described above. In one embodiment the genetically-engineered virus comprises virus particles containing at least one HCV structural protein and is useful in a vaccine against HCV. In another embodiment, the genetically-engineered virus is attenuated as compared to the pestivirus and is useful as a vaccine against the pestivirus.
In a still further embodiment, the invention provides a replication-competent BVDV
vector expressing a heterologous sequence. The BVDV vector comprises the BVDV
sequences encoding the BVDV replication machinery. In some embodiments, the replication-competent BVDV vector expresses an antigen and is useful as a vaccine.
Brief Description of the Drawings Figure 1 is a schematic representation of the 5' NTRs of BVDV, HCV, and EMCV
showing the position of the start codons of the ORF, and the boxes indicating the canonical IRES elements.

Figure 2 shows a schematic representation of BVDV and HCV chimeras, plaque phenotypes, reticulocyte translation efficiencies relative to parental BVDV, specific infectivities in MDBK cells, titers at 24 and 48 h post-transfection (or 72 h, as indicated), and an indication of whether pseudorevertants arose with results from BVDV, 5'HCV, BVDV+HCV, and BVDV+HCVdeIB3 chimeras shown in Fig. 2A and results from BVDV+HCVde1B2B3, BVDV+HCVde1B1B2B3, BVDV+HCVde1B2B3H1, and BVDV+HCVde1B2B3H1H2 shown in Fig. 2B, where N.D. means not determined.
Figure 3 illustrates the in vitro translation efficiency of BVDV RNA or chimeras showing bar graphs of the amount of IVv"°, the N-terminal protein in the BVDV ORF, expressed by the various constntcts.
Figure 4 illustrates a schematic representation of EMCV chimeras, plaque phenotypes, reticulocyte translation efficiencies relative to parental BVDV, specific infectivities in MDBK cells, titers at 24 and 48 h post-transfection (or 72 h, as indicated), and an indication of whether pseudorevertants arose.
Figure 5 illustrates a pseudorevertant analyses showing in (Fig. SA) the relative positions of mutations detected within the plaque-purified variants of passaged BVDV+HCVdeIB 1B2B3, S'EMCV, and 5'HCV, and in (Fig. SB) the 5' terminal sequences of pseudorevertants of BVDV+HCVdeIB 1B2B3, 5'EMCV, and 5'HCV. Novel nucleotides or sequences are shown in bold upper case type. Pseudorevertants are numbered and designated by the suffix ".R". The upper case sequence in BVDV+HCVde1B1B2B3 and BVDV+HCVde1B1B2B3.R1 is a remnant of downstream BVDV 5' NTR sequences and was created during the cloning procedures.
Figure 6 illustrates the construction of derivatives of 5'HCV designed to contain 5' termini corresponding to the sequence detected within the three analyzed pseudorevertants.
Fig. 6A shows the 5' terminal sequence of the 5'HCV derivatives with the suffix (orig) designating a derivative containing the~inal 5' terminal sequence of the pseudorevertant;
the suffix (cons) designating a derivative containing the consensus tetranucleotide sequence 5'-GUAU at the same position; and novel sequences shown in bold upper case type. Fig. 6B
shows plaque phenotypes, reticulocyte translation efficiencies relative to parental BVDV, specific infectivities in MDBK cells, and titers at 24 and 48 h post-transfection are indicated.
Figure 7 illustrates a single step growth curve for various chimeric constructs showing released virus titers measured by performing plaque assays on MDBK
cells transfected with various constructs.
Figure 8 illustrates replication of BVDV RNA or chimeric derivatives in transfected 3 S MDBK cells. Equal numbers of MDBK cells (~ 8 x 106) were electroporated with 5 D g of WO 99/55366 PG"T/US99/08850 each in vitro synthesized RNA. MDBK cells were also transfected with infectious yellow fever 17D and Sindbis RNAs to provide molecular mass markers. One fifth of the transfected cells were seeded on 35-mm dishes and incubated in D-MEM supplemented with 10%
horse serum for 6 h at 37°C. The media were then replaced with 1 ml of fresh media containing 2 g/ml of actinomycin D and 40 Ci/ml of 3H-uridine. Incubations were continued for 10 h at 37°C. RNAs were isolated as described in Materials and Methods, and 1/4 of the samples was denatured in glyoxal and loaded on an agarose gel. (A) Autoradiograph of the dried gel.
Only the portion of the gel containing the genomic RNAs is shown. (B) Amount of radioactivity contained within the displayed fragments as determined by scintillation counting. BVDV, lane 1; 5'HCV, lane 2; BVDV+HCVdelB2B3, lane 3;
BVDV+HCVdelB2B3H1, lane 4; 5'HCV.Rlorig, lane 5; 5'HCV.Rlcons, lane 6;
5'HCV.R3orig, lane 7; 5'HCV.R3cons, lane 8; 5'HCV.R2orig, lane 9;
5'HCV.R2cons, lane 10;
yellow fever 17D, lane 11; Sindbis, lane 12; non-transfected MDBK cells, lane 13. The experiments shown is one of two repetitions which yielded similar results.
Figure 9 illustrates the genetic map of plasmid pACNR/BUD.
Figure 10 illustrates the sequence of low copy number plasmid pACNR/BVDV
NADL (circular) harboring the functional cDNA of cytopathic BVDV NADL
(positive sense cDNA 5' to 3 ; nt 1-12578.
Figure 11 illustrates the sequence of infectious BVDV NADL (positive sense cDNA
5' to 3').
Figure 12 illustrates the sequence of infectious non-cytopathic BVDV NADL
lacking cIns (positive sense cDNA S' to 3').
Figure 13 illustrates the sequence adapted HCV 5' NTR from 5'HCV/Rl.cons (positive sense cDNA 5' to 3 ; only the sequence from the 5' base to the ATG
initiating the polyprotein is shown).
Figure 14 illustrates the sequence of adapted HCV 5' NTR from 5'HCV/Rl.orig (positive sense cDNA S' to 3 ; only the sequence from the 5' base to the ATG
initiating the polyprotein is shown).
Figure 15 illustrates the sequence of adapted HCV 5'NTR from 5'HCV/R2.cons (positive sense cDNA 5' to 3 ; only the sequence from the 5'' base to the ATG
initiating the polyprotein is shown).
Figure 16 illustrates the sequence of adapted HCV 5' NTR from 5'HCV/R2.orig (positive sense cNDA 5' to 3 ; only the sequence from the 5' base to the ATG
initiating the polyprotein is shown).

Figure 17 illustrates the sequence of adapted HCV S' NTR from 5'HCV/R3.cons (positive sense cDNA 5' to 3 ; only the sequence from the 5'base to the ATG
initiating the polyprotein is shown).
Figure 18 illustrates the sequence of adapted HCV 5'NTR from 5'HCV/R3.orig (positive sense cDNA 5' to 3'; only the sequence from the 5' base to the ATG
initiating the polyprotein is shown).
Figure 19 illustrates the sequence of prototype HCV-BVDV chimera from pNADL/5'HR3.orig/3'H3'B with the adapted HCV 5'NTR from 5'HCV/R3.orig and tandem 3' NTR elements from HCV followed by BVDV (positive sense cDNA 5' to 3') as discussed in Example 5.
Figure 20 illustrates various deletions of the poly U track in the 3'NTR HCV
sequence of BVDV/HCV chimera p5H-3H33.
Figure 2lillustrates the schematic representation of functional HCV/-BVDV
chimera from pCBV/p7.
Figure 22 illustrates the sequence of functional HCV-BVDV chimera from pCBV/p7 (positive sense cDNA 5' to 3').
Figure 23 illustrates the schematic representation of a HCVBVDV chimera with selectable marker.
Figure 24 illustrates the sequence of functional HCV-BVDV chimera from pCBV/p7/IRES-pac expressing a dominant selectable marker conferring resistance to puromycin (positive sense cDNA 5' to 3').
Figure 25 illustrates the schematic representation of a bicistronic HCVBVDV
chimera.
Figure 26 illustrates the sequence of functional bicistronic chimera expressing the entire HCV structural region derived from plasmid pNADLBI#41BCV str (positive sense cDNA 5' to 3') Description of the Preferred Embodiments In accordance with the present invention, the inventors herein have succeeded in generating HCV-BVDV chimeric RNAs which are replication competent. Such chimeras are useful in screening compounds in vitro for antiviral activity against HCV. In addition, it is believed that in vivo replication of HCV-BVDV chimeras according to the invention may be attenuated as compared to wild-type BVDV and thus may be useful in vaccinating animals against BVDV. It is also believed that the HCV chimeric structures described herein for BVDV are applicable to other pestiviruses.

In the context of this disclosure, the following terms will be defined as follows unless otherwise indicated:
"Cis-acting sequences" means the nucleotide sequences from an RNA virus genome that are necessary for recognition of the genomic RNA by specific proteins) of the RNA
virus or host cell that carry out replication, transcription, translation or packaging of the genome.
"Genetically-engineered virus" means any virus whose genome is different than that of a wild-type virus due to a human-made deletion, insertion, or substitution of one or more nucleotides to the wild-type viral genome.
"Infectious" when used to describe a virus means the virus is capable of entering cells and initiating a virus replication cycle, whether or not this leads to the production of new RNA virus particles.
"Nucleotide sequence" as used herein refers to DNA and the corresponding RNA
sequence where relevant. It will be understood that sequences shown in the Figures are DNA
versions of the RNA sequence and that chimeric molecules of the invention may comprises RNA molecules or cDNA copies of such RNA molecules.
"Replication-competent" as applied to a chimeric HCV-pestivirus RNA means the RNA is capable of RNA-dependent replication in at least one cell type that supports replication of the wild-type parental pestivirus. The number of replicated RNA
molecules produced by an HCV-pestivirus chimeric RNA of the invention is at least 10-fold higher than the limit of detection, which is typically 10 to 100 molecules. More preferably, chimeric RNA production by the HCV-pestivirus chimeric RNA is at least 102 to 103-fold higher than the detection limit. The replication-competent chimeric RNA replicates at an efficiency that is preferably, at least 0.001 %, more preferably, at least 0.01 %, more preferably, at least 0.1 %, more preferably, at least 1%, more preferably at least 10% and most preferably at least 50%
up to 90% that of the parental pestivirus in the same cell type.
"Transfected cell" means a cell containing an exogenously introduced nucleic acid molecule, and includes cells that are transiently transfected with the exogenous nucleic acid.
"Transformed cell" or "stably transformed cell" means a cell containing an exogenously introduced nucleic acid molecule which is present in the cytoplasm or nucleus of the cell and may be stably integrated into the chromosomal DNA of the cell.
"Virus" means a virion, virus particle or a viral genome.
A chimeric viral RNA according to the invention is designed to comprise a 5' NTR, an ORF, and a 3' NTR, at least one of which is a chimeric region containing two operably linked nucleotide sequences that are from the same region of a pestivirus and an HCV.

Pestivirus-specific sequences useful in the invention can be taken from the appropriate genomic region of any cytopathic or noncytopathic type I or type II BVDV
isolate, classical swine fever virus (CSFV) isolate, or border disease viral isolate. For a list of pestiviruses , see Thiel, H.-J., P. G. W. Plagemann, and V. Moennig. 1996. Pestiviruses, p.
1059-1073. In B. N. Fields, D. M. Knipe and P. M. Howley (ed.), Fields Virology. Raven Press, New York.
HCV-specific sequences can be taken from any strain or isolate of HCV, including but not limited to HCV-l, HCV-la, HC'V-lb, HCV-lc, HCV-2a, HCV-2b, HCV-2c, HCV-3a .
Preferably, the parental pestivirus is a cytopathic strain of BVDV and the parental HCV strain is HCV-1.
The pestivirus- and HCV-specific sequences are operably linked in the chimeric region, meaning the sequences are arranged such that the resulting chimeric structure is functional in the context of replication of the pestivirus. For example, in one preferred embodiment the chimeric viral RNA comprises a chimeric 5' NTR which comprises a BVDV-specific 5' terminal sequence of 5'-(G/A)UAU and an IRES derived from HCV, with the ORF and the 3' NTR consisting of a sequence from the same regions of BVDV.
The BVDV-specific sequences at the 5' terminus and in the ORF and 3' NTR are chosen such that they are functional in the context of BVDV, meaning the chimeric viral RNA
expresses the replication machinery of BVDV and this replication machinery is capable of replicating the chimeric RNA. In addition, translation of the BVDV ORF in the chimeric viral RNA is dependent upon a functional HCV IRES. The presence of a functional HCV IRES in this chimera allows the chimera to be used to screen for compounds that target the HCV IRES and thereby inhibit translation of the BVDV ORF as well as replication of the chimeric virus.
Such compounds would be expected to also inhibit translation of the ORF in a wild-type HCV
and consequently inhibit HCV replication.
Compounds that could be screened for anti-HCV activity using this and other HCV-BVDV 5' NTR chimeras include but are not limited to antisense RNAs, RNA decoys that bind proteins involved in recognition of the HCV-specific sequences, ribozymes, and small molecule inhibitors of critical RNA-protein interactions. The use of such substances for therapeutic applications are known in the art. See, e.g., Amarzguioui M, et al., "Hammerhead ribozyme design and application." Cell Mol Life Sci. 1998 Nov;54(11):1175-202;
Welch PJ, et al., "Expression of ribozymes in gene transfer systems to modulate target RNA levels.", Curr Opin Biotechnol. 1998 Oct;9(5):486-96; Bramlage B, et al. "Designing ribozymes for the inhibition of gene expression."; Trends Biotechnol. 1998 Oct;16(10):434-8;
Gewirtz AM, et al. "Nucleic acid therapeutics: state of the art and future prospects.";
Blood. 1998 Aug 1;92(3):712-36; Altman S., "RNase P in research and therapy." Biotechnology (N
Y). 1995 Apr;13(4):327-9; Flanagan WM., "Antisense comes of age."; Cancer Metastasis Rev. 1998 Jun; 17(2):169-76; Agrawal S, et al., "Antisense therapeutics." Curr Opin Chem Biol. 1998 Aug;2(4):519-28; Caselmann WH, et al., "Synthetic antisense oligodeoxynucleotides as potential drugs against hepatitis C." Intervirology 1997;40(5-6):394-9;
Neckers LM., S "Oligodeoxynucleotide inhibitors of function: mRNA and protein interactions." Cancer J Sci Am. 1998 May;4 Suppl 1:535-42; Agrawal S, et al. "Mixed backbone oligonucleotides:
improvement in oligonucleotide-induced toxicity in vivo." Antisense Nucleic Acid Drug Dev.
1998 Apr;8(2):135-9; Crooke ST., "An overview of progress in antisense therapeutics."
Antisense Nucleic Acid Drug Dev. 1998 Apr;B(2):115-22; Fraisier C, et al., "High level 10 inhibition of HIV replication with combination RNA decoys expressed from an HIV-Tat inducible vector."; Gene Ther. 1998 Dec;S(12):1665-76; Gervaix A, et al. "Gene therapy targeting peripheral blood CD34+ hematopoietic stem cells of HIV-infected individuals."
Hum Gene Ther. 1997 Dec 10;8(18):2229-38; Nakaya T, et al. "Inhibition of HIV-replication by targeting the Rev protein." Leukemia 1997 Apr; l l Suppl 3:134-7; Nakaya T, et al. "Decoy approach using RNA-DNA chimera oligonucleotides to inhibit the regulatory function of human immunodeficiency virus type 1 Rev protein." Antimicrob Agents Chemother. 1997 Feb;41(2):319-25; Smith C, et al. "Transient protection of human T-cells from human immunodeficiency virus type 1 infection by transduction with adeno-associated viral vectors which express RNA decoys." Antiviral Res. 1996 Oct;32(2):99-115;
Bahner I, et al. "Transduction of human CD34+ hematopoietic progenitor cells by a retroviral vector expressing an RRE decoy inhibits human immunodeficiency virus type 1 replication in myelomonocytic cells produced in long-term culture." J Virol. 1996 Ju1;70(7):4352-60; Lee SW, et al. "Inhibition of human immunodeficiency virus type 1 in human T cells by a potent Rev response element decoy consisting of the 13-nucleotide minimal Rev-binding domain." J
Virol. 1994 Dec;68(12):8254-64; Lisziewicz J, et al. "Inhibition of human imrnunodeficiency virus type 1 replication by regulated expression of a polymeric Tat activation response RNA
decoy as a strategy for gene therapy in AIDS." Proc Natl Acad Sci USA. 1993 Sep 1;90(17):8000-4; Bevec D, et al. "Inhibition of human immunodeficiency virus type 1 replication in human T cells by retroviral-mediated gene transfer of a dominant-negative Rev trans-activator." Proc Natl Acad Sci USA. 1992 Oct 15;89(20):9870-4.
It is contemplated that a number of replication-competent chimeric structures can be made that allow the function of various HCV sequence elements and proteins to be studied and targeted in drug screening assays. For example, the invention includes replication-competent HCV-pestivirus chimeras having a chimeric ORF. One such chimeric ORF
is one comprising an HCV sequence encoding the structural proteins and a pestivirus sequence WO 99/55366 PCTlUS99/08850 encoding the nonstructural proteins. It is believed that upon introduction into a cell, such a HCV-BVDV ORF chimera will produce HCV-like virus particles that will be released from the cell and capable of infecting cells normally infected by wild-type HCV, i.e., cells expressing an HCV receptor such as human CDBI. Such ORF chimeras would be useful to screen compounds for drugs that inhibit formation, release or entry of HCV
particles. In addition, ORF chimeras that produce virus particles containing at least one HCV structural protein would be useful as vaccines against HCV. Other ORF chimeras contemplated by the invention include, for example, chimeras comprising a pestivirus sequence encoding structural proteins and an HCV sequence encoding one or more nonstructural proteins such as the NS3 protease, NS4A cofactor, NSSA phosphoprotein/interferon resistance determinant and/or the NSSB polymerase. Replication of such ORF chimeras would be dependent upon the function of the HCV nonstructural proteins) and these ORF chimeras could be used to screen for drugs that target the HCV nonstructural proteins) as well as to screen for and map potential drug resistance mutations in HCV nonstructural proteins. In addition, HCV-pestivirus ORF chimeras could be useful for developing alternative in vivo animal models for HCV replication and HCV-associated hepatocellular carcinoma to evaluate antivirals and anti-tumor agents.
The, invention also provides replication-competent HCV-pestivirus chimeras having a chimeric 3' NTR which contains one or more conserved elements of the HCV 3' NTR. Such 3' NTR chimeras would be useful for screening or evaluating compounds targeted against the HCV 3' NTR. Compounds that could be screened include antisense RNA molecules, ribozymes and small molecule inhibitors of critical RNA-protein interactions.
Chle 3' NTR
chimera according to the invention comprises a BVDV 5' NTR, BVDV ORF and a chimeric 3' NTR which consists of an HC'.V-specific sequence derived from the HCV 3' NTR
immediately followed by a BVDV 3' NTR. The HCV-specific 3' NTR that allows for replication in the context of BVDV has a deletion in the 3' NTR poly (U) tract but has all the other HCV 3' NTR elements, including the 98 by 3' terminal conserved element.
HCV-pesdvirus chimeras included within the scope of the invention include those comprising combinations of chimeric regions, i.e., 5' NTR and ORF chimeras; 5' NTR and 3' NTR chimeras; ORF and 3' NTR chimeras; and chimeric RNAs in which each of the S' NTR, ORF and 3' NTR regions comprise an HCV sequence operably linked to a pestivirus sequence.
T'he invention also provides chimeric RNAs having two ORFs, or bicistronic HCV-pestivirus chimeras. Bicistronic chimeras contemplated by the invention include structures in which the first ORF contains one or more HCV genes and is followed by a second IRES

operably linked to a second ORF encoding the pestivirus replicase machinery.
It is also contemplated the first ORF may encode a heterologous sequence such as an antigen.
It is believed that many HCV-pestivirus chimeras of the invention will be attenuated as compared to the parental wild-type pestivirus. Such attenuated chimeric RNA
genomes would be candidate vaccines in the form of live-attenuated virus particles or as RNA or cDNA "genetic" vaccines.
The invention also includes vaccines against HCV which comprise an immunogenically-effective amount of HCV-pestivirus particles or nucleic acid.
Anti-HCV
vaccines comprising virus particles should preferably contain one or more HCV
structural proteins.
The therapeutic or pharmaceutical compositions of the present invention can be administered by any suitable route known in the art including for example by injection such as intraperitoneal, intravenous, subcutaneous, intramuscular, transdermal, intrathecal or intracerebral injection. Administration can be either rapid as by injection or over a period of time as by slow infusion or administration of slow release formulation.
Compositions according to the invention can be employed in the form of pharmaceutical or veterinary preparations. Such preparations are made in a manner well known in the pharmaceutical and veterinary arts. One preferred preparation utilizes a vehicle of physiological saline solution, but it is contemplated that other pharmaceutically acceptable carriers such as physiological concentrations of other non-toxic salts, five percent aqueous glucose solution, sterile water or the like may also be used. It may also be desirable that a suitable buffer be present in the composition. Such solutions can, if desired, be lyophilized and stored in a sterile ampoule ready for reconstitution by the addition of sterile water for ready injection. The primary solvent can be aqueous or alternatively non-aqueous.
The carrier can also contain other pharmaceutically-acceptable excipients for modifying or maintaining the pH, osmolarity, viscosity, clarity, color, sterility, stability, rate of dissolution, or odor of the formulation. Similarly, the carrier may contain still other pharmaceutically-acceptable excipients for modifying or maintaining release or absorption or penetration across the blood-brain barrier. Such excipients are those substances usually and customarily employed to formulate dosages for parenteral administration in either unit dosage or mufti-dose form or for direct infusion into the cerebrospinal fluid by continuous or periodic infusion.
It is also contemplated that certain formulations containing a chimeric virus according to the invention are to be administered orally. Such formulations are preferably encapsulated and formulated with suitable carriers in solid dosage forms. Some examples of suitable carriers, excipients, and diluents include lactose, dextrose, sucrose, sorbitol, mannitol, starches, gum acacia, calcium phosphate, alginates, calcium silicate, microcrystalline cellulose, polyvinylpyrrolidone, cellulose, gelatin, syrup, methyl cellulose, methyl- and propylhydroxybenzoates, talc, magnesium, stearate, water, mineral oil, and the like. The formulations can additionally include lubricating agents, wetting agents, emulsifying and suspending agents, preserving agents, sweetening agents or flavoring agents.
The compositions may be formulated so as to provide rapid, sustained, or delayed release of the active ingredients after administi~ation to the patient by employing procedures well known in the art. The formulations can also contain substances that diminish proteolytic degradation and promote absorption such as, for example, surface active agents.
The specific dose is calculated according to the approximate body weight or body surface area of the patient or the volume of body space to be occupied. The dose will also be calculated dependent upon the particular route of administration selected.
Such calculations can be made without undue experimentation by one skilled in the art. Exact dosages are determined in conjunction with standard dose-response studies. It will be understood that the amount of the composition actually administered will be determined by a practitioner, in the light of the relevant circumstances including the condition or conditions to be treated, the choice of composition to be administered, the age, weight, and response of the individual patient, the seventy of the patient's symptoms, and the chosen route of administration. Dose administration can be repeated depending upon the pharmacokinetic parameters of the dosage formulation and the route of administration used.
Replication-competent HCV-pestiviruses are generated by choosing the HCV
function or sequence element desired to be studied. The HCV sequence can be obtained from a plasmid clone of a partial or full HCV genome using PCR to amplify a target region containing the desired sequence or by restriction enzyme digestion. The HCV
fragment is then inserted into the desired location of a clone of the pestivirus genome using standard techniques. Desired portions of the pestivirus genome may be deleted before or after addition of the HCV fragment. The recombinant genome is then transfected into a cell that supports replication of the parental pestivirus genome and their ability to replicate using standard assays. For example, replication can be assessed by virus-induced cytopathic effect; plaque formation; detection of viral antigens and/or viral RNA accumulation; and by plaque assay measuring released infectious virus. The inventors herein have found that the BVDV RNA
replication machinery works in many cell types, including bovine, hamster, mouse and human cells. It has also been reported that BVDV RNAs can amplify in other cell types including human hepatoma lines and hepatocytes (Behrens SE, et al., J Virol. 1998 Mar;72(3):2364-72).

The host cell range for a particular chimera will be dependent upon the properties of that chimera as empirically determined.
As described below, some chimeras do not replicate stably as indicated by heterogeneity in the size of plaques produced by the chimeric virus. Upon passage, pseudorevertants can frequently be isolated that are capable of stable replication. Such pseudorevertants will have one or more deletions or base substitutions in the HCV and/or pestivirus sequences. Information derived from these gain-of function mutations can be used to define the elements necessary for generating stable, replication-competent chimeras of HCV and a pestivirus.
The invention provides a method for screening compounds for antiviral activity against HCV. The method involves comparing a test compound's effect on replication of a chimeric HCV-pestivirus RNA molecule as described above with the compound's effect on replication of the parental pestivirus. Compounds which have a greater effect on replication of the chimeric virus than the pestivirus are likely directed against the HCV
portion of the I S chimera. Typically, the method is performed by providing duplicate cell cultures containing a chimeric viral RNA which is replication-competent in that cell, treating one of the culture with the test compound, and then measuring the replication efficiency of the chimeric RNA in both cultures. Any effect induced by the compound is compared against the compound's effect on replication of the parental pestivirus in cells of the same type.
This control assay is preferably performed at the same time using the same culture conditions.
The cells used in the screening assay can be prepared by transiently transfecting the cells with the desired chimeric RNA molecule as described below.
Alternatively, it is contemplated that the chimeric RNA molecule can be constitutively expressed in the cell by transfecting the cell with a polynucleotide comprising a cDNA of the chimeric RNA operably linked to a DNA-dependent promoter. The chimeric cDNA may include a selectable marker.
which would allow for selection of cells expressing the chimeric RNA. It is also envisioned the selectable marker could be a dominant marker that allows selection of cells expressing chimeras having adaptive mutations or selection of cells permissive for virus replication (Frolov et al., J. Virol. 73:3854-:3865, 1999). It is also contemplated the cDNA could express a reporter gene that could be assayed to measure RNA replication.
Alternatively, chimeric virus particles are incubated with a cell permissive for infection by the pestivirus in the presence or absence of the test compound and then replication of the chimeric virus is measured and compared to the replication of the parental pestivirus incubated with the same cell type in the presence or absence of the test compound.

Inhibition of replication can be measured in many ways, including assaying for the reduction of virus-induced cytopathic effect; inhibition of plaque formation, reduced production of viral antigens as detected by immunofluoresence assay; reduced viral RNA
accumulation; reduction in released infectious virus from treated and untreated control and 5 chimera samples using a plaque assay. In addition, it is contemplated that a cell line that is designed for pestivirus-specific transactivation of a reporter gene could be used directly or in lieu of a plaque assay. The reporter gene is operably linked to a promoter that is activated upon infection by the chimeric virus and production of the viral transactivator protein.
Preferred embodiments of the invention are described in the following examples.
10 Other embodiments within the scope of the claims herein will be apparent to one skilled in the art from consideration of the specification or practice of the invention as disclosed herein. It is intended that the specification, together with the examples, be considered exemplary only, with the scope and spirit of the invention being indicated by the claims which follow the examples.
15 Example 1 This example illustrates the construction and analysis of S' HCV-BVDV chimeras as reported in detail in Frolov et al. (RNA 4:1418-1435, 1998) which is incorporated in its entirety by reference. A functional clone of BVDV (Mendez et al., J. Virol.
72:4737-4745, i 998) was used to construct and characterize a series of 5' NTR chimeras with sequences derived from HCV and the picornavirus, encephalomyocarditis virus (EMCV). The results help to define the requirements of a functional BVDV S' NTR and provide replication-competent BVDV-HCV chimeras dependent on a functional HCV IRES.
Example 2 This example illustrates the construction of chimeras for expressing additional functional portions of the HCV genome by addition of further HCV sequence downstream from the functional or adapted HCV 5'NTR chimeras fused in-frame to the BVDV
ORF.
One such construct (Figure 21) involves fusion of HCV sequences to BVDV
sequences in the p7 protein coding region (at a convenient BseRI restriction site). Both HCV
and BVDV encode a p7 protein that is located immediately downstream of the E2 protein.
The p7 protein is a small hydrophobic protein of unknown function. pCBV/p7 consists of the first 79 bases of the BVDV 5'NT'R encoding stem loop structure B 1' and B 1, followed by the entire HCV 5'N'TR, the entire HCV structural protein coding region and the first 36 amino acids of HCV p7 fused to the C-terminal 31 amino acids of BVDV p7. The fused p7 gene is followed by the remainder of the BVDV ORF including the entire nonstructural region and the BVDV 3' NTR. Transfection of MDBK cells with the RNA corresponding to this sequence (Figure 22) leads to replication of the chimeric RNA and production of the expected HCV and BVDV polyprotein cleavage products. Variations on this strategy are envisioned in which all or part of the HCV polyprotein and cis elements important for RNA
packaging can be expressed in viable chimeras. In addition the BVDV replicase regions for either cytopathic or non-cytopathic pestiviruses (like NADL cIns-) can be used. Transfection of cells permissive for HCV particle, assembly, release and reinfection with this chimeric RNA can be used to make HCV-like particles. These particles and this infection system can be used (i) to screen for specific inhibitors of HCV particle, assembly, release and reinfection, (ii) for identifying antibodies capable of neutralizing HCV infectivity and (iii) as live or inactivated vaccines. Furthermore, this embodiment of the invention demonstrates that the BVDV RNA
replication machinery can be used for expression of heterologous RNA and polypeptide sequences and can be used as a vehicle for RNA or DNA "genetic" vaccination in which the BVDV replicase amplifies the level of antigen expression by cytoplasmic RNA-dependent replication.
Example 3 This example illustrates chimeric RNA's that are modified to express dominant selectable markers, assayable markers or FACS sortable markers.
Such variants can be used to select for chimeras capable of replication in particular cell types, or to screen for cell types that are permissive for replication of the chimeric RNA.
Selectable markers include, but are not limited to, the genes encoding puromycin resistance (puromycin N-acetyl transferase; PAC), neomycin resistance, blasticidin resistance, hygromycin resistance, etc. Assayable markers include, but are not limited to, the genes encoding B-galactosidase, luciferase, B-glucuronidase, ete. Easily sortable molecules include single chain antibodies, cell surface markers, and non-toxic protein markers like green fluorescent protein. In a specific example (Figures 23 and 24), the RNA
encoded by pCBV/p7 was modified to include a cassette at the beginning of the BVDV 3'N'1R
that is comprised of the EMCV IRES driving the gene encoding PAC. This chimeric RNA
can replicate, expresses PAC and confers resistance to puromycin resistance. This property can be used to select for variants of the chimera that are capable of noncytopathic replication in desired cells type and also provides a means of showing that cells harbor a functional chimeric RNA. Desired variants can be identified, cloned and further characterized as described in Example 1. Of note, is that this location in the BVDV genome and this strategy for expressing heterologous genes may also be applied to using infectious attenuated pestiviruses as gene expression vectors and as chimeric live vaccines against other animal pathogens.
Example 4 This example illustrates the use of the bicistronic strategy as an alternative to the in-frame fusions described in Example 2.
A specific example is shown in Figure 25 and its sequence as Figure 26. In this bicistronic chimera, the 5' sequences are identical to that of pCBV/p7 except that the HCV
ORF continues to include the first 246 amino acids of NS4B. The HCV sequence is followed by the EMCV IRES fused to BVDV Npro, the N-terminal 10 as of BVDV C, the C-terminal 19 as of C, 9 N-terminal amino acids of Ems, 48 C-terminal amino acids of E2 and the remainder of the BVDV NADL ORF and 3' NTR. The constructed BVDV ORF encodes a functional BVDV RNA replicase. The deletions in the N-terminal portion of this ORF were designed to preserve proper membrane topology and processing of the replicase.
The bicistronic chimeric RNA can replicate upon transfection of permissive BVDV
host cells.
Example 5 This example illustrates 3'NTR chimeras. Although initial attempts to recover viable chimeric viruses in which the BVDV 3'NTR was completely replaced by that of HCV were unsuccessful, a strategy similar to that detailed in Example 1 has produced chimeras that harbor the conserved elements of the HCV 3'NTR. An initial tandem 3'NTR
construct was made in which the HCV 3'NTR was engineered to follow the BVDV ORF. The complete BVDV 3'NTR was position 3' to the HCV 3' NTR after a short heterologous sequence. This sequence of this parental construct, which replicated poorly, is shown in Figure 19 RNAs transcribed from this plasmid were of low specific infectivity suggesting that revertants or pseudorevertants might have arisen. Indeed isolation and sequence analysis of several independent plaque-forming variants revealed that deletions in the HCV poly U
tract of various lengths had occurred. These revertant sequences are shown in Figure 20. When these altered HCV 3'NTRs were reconstituted into the original tandem 3' NTR parent, they gave rise to plaque forming RNA transcripts of high specific infectivity, demonstrating that these alterations restored the ability of the chimeric RNA to replicate. Large deletions in the U tract gave rise to virus with more robust replication and larger plaques while stably maintaining the conserved HCV 3'NTR 98-base element and the polypyrimidine "transition"
region. Such chimeric viruses can now be used to screen and evaluate antisense, ribozyme, and other therapeutics targeted against this conserved HCV RNA element that is essential for replication.
Materials and Methods Plasmid Constructs pACNR/BVDV NADL was previously described (Mendez et al., 1998, supra).
pBVDV is a derivative of pACNR/BVDV NADL which contains a G-~T transversion at nt 14994 that creates an Xba I site upstream of the T7 promoter (T. Myers & C.M.
Rice, unpubl.). To facilitate construction of the chimeras, subclones were created.
First, two fragments were isolated by PCR amplification of p90/HCVFLIongpU (Kolykhalov et al., Science 277:570-574, 1997) with primers #498 (5'-TGTACATGGCACGTGCCAGCCCC) and #498 (5'-GATCAACTCCATGGTGCACGGTCT) and pBVDV with primers #481 (5'-AGACCGTGCACCATGGAGTTGATC) and #482 (5'-1 S CGTTTCACACATGGATCCCTCCTC). These two fragments were digested with ApaL I
and ligated to produce a fragment containing a fusion of the HCV 5' NTR to the BVDV ORF.
This fragment was digested with SacI and ligated into pGEM3Zf(-) which had been digested with Sma I and Sac I to produce the subclone pGEM498-Sacl. Next, a fragment containing the BVDV 5' NTR was synthesirxd by PCR amplification of pBVDV with primers #183 (5'-TTTTCTAGATAATACGACTC:ACTATAGTATACGAGAATTAGAAA.AGGCACTCG) and #480 (S'-GGGGGCTGGCACGTGCCATGTACA). This fragment was digested with Xba I and BsrG I and ligated into pGEM498-SacI digested with the same two enzymes, to create the plasmid pGEMXbaI-Sacl. pGemXbal-Sacl contains a tandem fusion of the BVDV
5' NTR, the HCV 5' NTR, and the 5' portion of the BVDV IV'"° gene.
pBVDV + HCV was created by digesting pGEMXbaI-SacI with Xba I and Sac I and ligating the fragment into pBVDV digested with the same two enzymes, and as such pBVDV + HCV contains the promoter, followed by the entire 385-nt 5' NTR of BVDV, a GT dinucleotide (nt 386-387), the entire 341-nt 5' NTR of HCV' (nt 388-728), and the sequence of the BVDV
NADL strain including the ORF and 3' NTR. Derivatives of pBVDV + HCV containing deletions within the BVDV 5' NTR and/or the HC:V 5' NTR were created in the subclone pGEMXbaI-Sacl, as described below, prior to ligation into Sba I- and Sac I-digested pBVDV. For making deletions, restrictions sites with non-compatible protruding ends were treated with the Klenow fragment of DNA polymerase I prior to ligation. For creation of pBVDV +
HCVdeIB3 (deletion of nt 174-374, inclusive), pGEMXbaI-Sacl was digested with A,~1 II and BsrG I. For pBVDV + HCVde1B2B3 (deletion of nt 67-374), pGEMXbaI-Sacl was digested with Avr II and BsrG I. For pBVDV + HCVdeIB1B2B3 (deletion of nt 33-374), pGEMXbaI-Sacl was digested with Snag I and BsrG I. For pBVDV + HCVde1B2B3H1 (deletion of nt 67-3396), pGEMXbaI-Sacl was digested with Avr II and Xcm I. For pBVDV +
HCVde1B2B3H1H2 (deletion of nt 67-513), pGEMXbaI-Sacl was digested with AVR II
and S Bsg I. For pBVDV + HCVde1B2B3H3 (deletion of nt 67-374, 518-704), subclone pGEMXbaI-SacidelB2B3 was digested with Sma I. p5'HCV was created by digesting p90/HCVliongpU with Xba I and Nru I and ligating the fragment into pBVDV + HCV
digested with the same two enzynnes.
The EMCV plasmid, pECg, was provided by Ann Palmenberg and is described elsewhere (Hahn et al., J. Virol b9:2697-2699, 1995). p5'EMCV contains the entire 710 nt of the S' NTR of EMCV, followed by the open reading frame of BVDV and the 3' NTR.
One extra G residue was added between the T7 promoter and the first nucleotide of the EMCV S' NTR to facilitate efficient in vitro transcription. Convenient restriction sites within the BVDV 5' NTR or the EMCV 5' NTR were used to create additional chimeras. Sites with noncompatible protruding ends were treated with the Klenow fragment of DNA
polymerase I
prior to ligation. For example, the plasmid pBVDV + EMCVdeIA contains nt 1-378 of BVDV 5' NTR fused with nt 45-710 of EMCV (the BsrG I site of BVDV ligated to the EcoR
V site of EMCV), pBVDV + EMCVdeIB3A contains nt 1-173 of BVDV fused with nt 45-of EMCV (the AJI II site of BVDV ligated to the EcoR V site of EMCV). pBVDV +
EMCVdelB2B3A contains nt 1-66 of BVDV fused with nt 45-710 of EMCV (the Avr II
site of BVDV ligated to the EcoR V site of EMCV). pBVDV + EMCVdelB3ABC contains nt 173 of BVDV fused with nt 161-710 of EMCV (the Afl II site of BVDV ligated to the Psp1405 site of EMCV). pBVDV + EMCVde1B2B3ABC nt 1-66 of BVDV fused with nt 161-710 of EMCV (the Avr II site of BVDV ligated to the Psp1406 site of EMCV).
pBVDV
+ EMCVdeIB3A-H contains nt 1-101 of BVDV fused with nt 289-710 of EMCV (the lVhe I
site of BVDV ligated to the Avr II site of EMCV). pBVDV + EMCVdelB2B3A-H
contains nt 1-62 of BVDV fused with nt 289-710 of EMCV (the Avr II site of BVDV ligated to the Avr II site of EMCV). The schematics of the chimeric 5' NTRs are presented in Figures 2 and 4.
All other heterologous 5' NTRs used in the study were generated by PCR using an oligonucleotide complementary to nt256-272 of the HCV 5' NTR and primers containing the sequence of the Xba I restriction site followed by the T7 promoter, the heterologous sequences found in sequenced pseudorevertants, or sequences corresponding to different regions of the HCV 5' NTR. All the fragments were subcloned into the plasmid, pRS2 (a derivative of pUCl9), sequenced, and recloned into the p5'HCV plasmid by replacing the fragment between the XBa I site located upstream of the T7 promoter and the Nhe I site (nt 249-254) in the 5' NTR of HCV.
Cell cultures MDBK cells were obtained from M. Collett (ViroPharma, Inc.) and BT cells were S obtained from the American Type Culture Collection (Rockville, Maryland).
Cells were grown in Dulbecco's modified Eagle medium (D-MEM) suppiernented with 10% horse serum and sodium pyruvate.
Transcriptions and transfections All the designed plasmids , including pBVDV and the chimeric derivatives, were 10 digested to completion with Sda I (Sse83871), purified by phenol extraction, precipitated by ethanol, and dissolved in water. The transcription reactions were performed sin the T7 Megascript kit (AMBION) using the conditions recommended by the manufacturer.
Reactions were incubated at 37°C for 1 h, and 3H-UTP was added to the reaction to quantify the RNA synthesis. The quality of the synthesized RNAs was checked by agarose gel 15 electrophoresis, and samples containing 50-60% of full-length RNA were used for electroporations and in vitro translations. The reaction mixtures were aliquoted and stored at -70°C prior to electroporation or in vitro translations.
Transfection was performed by electroporation of MDBK cells using previously described conditions (Mendez et al., 1998, supra). Two micrograms of in vitro synthesized 20 RNA, corresponding to approximately 1 p, g of the full-length transcript, were used per electroporation. In standard experiments, ten-fold dilutions of electroporated cells were seeded in 6-well tissue culture plates containing 5 x 105 naive MDBK cells per well. After 1 h of incubation at 37°C in an 5°/m COz incubator, cells were overlaid with 3 ml of 0.6% LE
Sea Kem agarose (FMC Bioproducts) containing minimal essential medium supplemented with 5% horse serum. Plaques were stained with crystal violet after 3 days incubation at 37°C. The rest of the transfected cells was seeded into 100-mm dishes and incubated for approximately 48 h or until cytopathic effect was observed in virtually all cells. Samples of the media were taken at 24 and 48 h, and virus titers were determined as described above and previously (Mendez et al., 1998, supra).
Analysis of the 5' ends of viral genomes Sequencing of the 5' ends of selected variants of BVDV was performed on plaque-purified viruses. Plaques were typically isolated from the agarose overlay without staining with neutral red. Virus was eluted in 1 ml of D-MEM/10% horse serum for several hours and was used to infect 5 x 105 MDBK cells in 35-mm dishes. After 1 h of virus adsorption of 37 °C, an additional 1 ml of D-MEM/10% horse serum was added to the dishes, and incubation was continued for 36-48 h until cytopathic effect was observed in virtually all cells.
Fifty microliters of harvested viral stocks were clarified by low speed centrifugation, and viral RNAs were isolated by TRIzoI reagent {Gibco-BRL) using the protocol recommended by the manufacturer. Sequencing of the 5' termini was performed using an oligonucleotide/cDNA-ligation strategy described elsewhere (Troutt et al., Proc. Natl. Acad.
Sci. USA 89:9823-9825, 1992). The primer S 1 (5'-GTCGTTTCACACATGGATCC), complementary to nt 710-729 of the BVDV genome, was used for cDNA synthesis. A
phosphorylated oligonucleotide tag (5'-GACTGTTGTGGCCTGCAGGGCCGAATT) with an amino group on the 3' terminus was ligated to the first strand cDNA {Troutt et al., 1992, supra). One tenth of this reaction mixture was used for PCR amplification. The primers for PCR amplification were as follows: primer A (5'-GCCCTGCAGGCCACAACAGTC), complementary to the tag; primer B (5'-TCAGGCAGTACCACAA) complementary to nt 281-296 of the HCV 5' NTR; and primer C (5'-GGAATGCTCGTCAAGAAGACAG), complementary to nt 268-289 of the EMCV 5' NTR. The primer pairs of A + B or A
+ C
were used for analysis of the pseudorevertants of 5'HCV and BVDV + HCVdeIB
1B2B3 or 5'EMCV, respectively. For the 5'HCV pseudorevertants, one tenth of the ligation mixture was used for an additional PCR reaction. This fragment was synthesized using primer S 1, describe above, and a primer corresponding to nt 147-175 of the HCV genome.
Fragments were purified by agarose gel electrophoresis and cloned into the plasmid pRS2.
Multiple independent clones were sequenced by the standard dideoxy-mediated chain termination methods using the Sequenase version 2.0 DNA Sequencing Kit (USB).
Cell-free translation Cell-free translation reactions were performed in reticulocyte extracts (Promega) using conditions recommended by the manufacture. Usually 0.1-1 pg of the same in vitro synthesized RNAs used in transfection experiments were used in 25 p,l translation reactions.
After 45 min of incubation at 30 °C, 2 pl were dissolved in 10 p,l of sample buffer, and those samples were analyzed by sodium dodecyl sulfate PAGE. Labeled proteins were visualized by autoradiography of the dried gel. The efficiency of translation was measured using phosphorimager analysis (Molecular Dynamics) by comparing the radioactivity in the band corresponding to the Np'° protein. In preliminary experiments, an eightfold increase in incorporation was observed for translation of 4 p,g versus 0.4 p,g BVDV
transcript RNA.
Quantitative data were obtained from reactions using subsaturating (0.4 p,g) amounts of BVDV or BVDV chimera transcript RNAs.

Analysis of virus specific RNAs The protocols used for radioactive labeling of virus-specific RNAs are described in the appropriate figure legends. RNAs were isolated from the cells by using TRIzoI reagent as recommended by the manufacturer (Gibco-BRL). After denaturation with glyoxal in dimethylsulfoxide, cellular RNAs were analyzed by electrophoresis in a 1%
agarose gel containing a 10 mM phosphate buffer. Pieces of the dried gel containing the appropriate RNA bands were excised, and their radioactivity measured by liquid scintillation counting.
Results Features of the BVDV, HCV, and EMCV 5' NTRs important for chimera design Schematic representations of the proposed secondary structures of the 5' NTRs of HCV, BVDV, and EMCV are shown, and the location of each IRES is indicated in Figure 1.
EMCV is a member of the cardiovirus genus within the family Picornaviridae.
While not a member of the Flaviviridae, EMCV is similar to HCV and BVDV in that it is a positive-strand RNA virus shown to contain an IRES within its 5' NTR (Jang et al., J.
virol 62:2636-2643, 1988). Based on their proposed secondary structures, the HCV IRES and the BVDV
IRES have been classified as type 3 IRESs, while the EMCV IRES is classified as a type 2 IRES (Lemon & Honda, Siemin. Virol. 8:274-288, 1997). However, these three IRESs as well as IRESs from other members of the Flaviviridae and the Picornaviridae have been proposed to contain a common structural core (Le et al., Yirus Genes 12:135-147, 1996).
The model for the secondary structure of the 341-nt HCV 5' NTR has been refined by enzymatic and chemical analysis of synthetic transcripts (Brown et al., Nucl.
Acids. Res.
20:5041-5045, 1992; Wang et al., J. Virol 68:7301-7307, 1994; Honda et al., RNA 2:955-968, 1996; Lima et al., 1997). This element contains four discreet hairpins (referred to here as H1, H2, H3 and H4) and a pseudoknot at the base of hairpin H3 (Wang et al., 1995).
The secondary structure of the 385-nt BVDV S' NTR has not been as extensively studied, but is proposed to be similar to that of HCV (Brown et al., 1992) with four discrete hairpins (referred to here as B1', B1, B2, and B3) and a pseudoknot at the base of B3 (Rijnbrand et al., 1997). The secondary structure of the longer (>700 nt) EMCV 5' NTR consists of a series of hairpins A-M (Duke et al., 1992; Hoffman & Palmenberg, 1996). Recently, a revised model of the EMCV 5' NTR suggests moderately different secondary structures for the C and G
subregions, and significantly different secondary structures for the I-M
subregion (Palmenberg & Sgro, 1997).
For HCV, H1 is nonessential for 1RES function (Reynolds et al., 1995;
Rijnbrand et al., 1995; Honda et al., 1996b; Reynolds et al., 1996; Kamoshita et al., 1997) and its deletion has actually increased translation efficiency in some analyses (Rijnbrand et al., 1995; Honda et al., 1996b). Most studies have found that hairpin H2 and H3 and the pseudoknot are essential for IRES function (Wang et al., 1993; Rijnbrand et al., 1995; Honda et al., 1996b).
However, two studies indicate that H2 may not be essential (Tsukiyama-Kohara et al., 1992;
Urabe et al., 1997). The 3' boundary of the HCV IRES is more controversial.
The IRES
clearly extends to the AUG initiation codon. However, some studies indicate that sequences affecting the efficiency of translation initiation extend into the ORF
{Reynolds et al., 1995;
Honda et al., 1996a; Honda et al., 1996b; Lu & Wimmer, i 996). By analogy to the HCV
IRES and the related pestivirus (:SFV IRES, the BVDV IRES probably requires hairpins B2 and B3 and the pseudoknot for function, with B 1' and B 1 probably not required for IRES
activity (Poole et al., 1995; Rijnbrand et al., 1997). For EMCV, hairpins H-L
have been shown to be required for IRES function in mono- or dicistronic constructs (Jang & Wimmer, 1990; Duke et al., 1992). The remaining portion of the EMCV 5' NTR is thought to be required for RNA replication or unknown steps in viral replication that are important for pathogenesis (Duke et al., 1990; Martin & Palmenberg, 1996).
Replacement of the BVDV 5' NTR with the HCV 5' NTR results in a large decrease in specific infectivity Since the BVDV 5' NTR and the HCV S' NTR are proposed to have similar RNA
secondary structure and functional organization, an experiment was performed to test whether the BVDV S' NTR could be replaced by the HCV 5' NTR. p5' HCV has an exact replacement of the BVDV 5' NTR with that of HCV (Fig. 2A) while the coding sequence and 3' NTR of p5'HCV are identical to pBVDV" Positioning of the HCV 5' NTR in such a manner was necessary since translation initiation from the HCV IRES begins at or near the AUG start codon (Honda et al., 1996a; Reynolds et al., 1995; Reynolds et al., 1996;
Rijnbrand et al., 1996). The specific infectivity of 5'HCV RNA synthesized in vitro was compared to that of BVDV RNA by transfection of MDBK (bovine kidney) cells (Fig. 2A). The specific infectivity of BVDV RNA was approximately 4 x 106 plaque forming units (PFU)/p.g RNA.
In contrast, the specific infectivity of 5' HCV RNA was near the limit of detection (30-50 PFU/pg RNA) and considerable plaque heterogeneity was apparent. These results suggested that the HCV 5' NTR replacement chimera might be incapable of efficient replication and plaque formation and that the plaque forming virus observed had arisen by secondary mutation(s). Sequence analysis of plaque-purified 5' HCV viruses presented below confirmed that the replicating pool of virus contained such pseudorevertants.

Next, the in vitro translation efficiency of these two RNAs in rabbit reticulocyte extracts was analyzed to test whether the defect in specific infectivity of 5' HCV RNA could be attributed to lower translation efficiency. Although the specific infectivity of S' HCV RNA
was reduced ~5 logs compared to BVDV RNA, its translation efficiency was only slightly reduced, twofold (Fig. 3, lane 1 vs. lane 2). The apparent size of the N-terminal cleavage product, NP'°, was identical for both RNAs, suggesting that translation initiated with the correct AUG. These data are consistent with the hypothesis that the BVDV S' NTR contains signals that are required for a step in replication other than translation which are not present in the 5' HCV chimera.
Given the low specific infectivity of 5' HCV RNA, an experiment was performed to test the effect of placing the BVI)V 5' NTR sequence upstream of the HCV 5' NTR, resulting in tandem BVDV and HCV 5' NTRs (called BVDV + HCV). This arrangement actually decreased translation efficiency (Fig. 3, lane 14 vs. lane 1) yet restored infectivity (Fig. 2A).
The plaques produced by BVDV + HCV were also heterogeneous in size, indicating that this 1 S virus was unstable. Upon passage, RT-PCR analysis indicated that pseudorevertants had indeed arisen in which portions of the BVDV and/or HCV 5' NTRs had been deleted (data not shown). These data show that sequences in the BVDV 5' NTR required for virus replication can function when placed upstream of a functional HCV IRES driving translation of the BVDV polyprotein.
Hairpins B1' and B1 in conjunction with the IiCV IRES are sufficient for stable and efficient BVDV replication The sequences within the BVDV 5' NTR that restored replication in the context of the HCV 5' NTR were mapped using three deletion variants. The deletion BVDV +
HCVdeIB3 removed a large portion of hairpin B3; the deletion within BVDV + HCVde1B2B3 removed hairpins B2 and B3, and the deletion within BVDV + HCVdeIB1B2B3 removed hairpins B1, B2 and B3. The specific infectivities of RNAs from these deletion mutants were near that of BVDV RNA (Fig. 2). Upon passage of these viruses, RT-PCR analyses and sequencing indicated that BVDV + HCV delB3 and BVDV + HCVde1B2B3 were stably propagated and produced homogeneous plaques slightly smaller than those of wild-type BVDV
(data not shown). In contrast, BVDV + HCVdeIB 1B2B3 produced smaller heterogeneous plaques.
Reverse transcription-polymerase chain reaction (RT-PCR) analysis and sequencing indicated that BVDV + HCVde1B1B2B3 underwent a reversion event described in more detail below.
The translation efficiencies of these three RNAs (Fig. 3, lanes 9, 10, and I2) were similar to BVDV + HCV RNA (Fig. 3, lane 14), indicating that the deleted portions (hairpins B1, B2, WO 99/55366 PCT/US99/0$850 and B3) are not required for translation in the BVDV + HCV chimera. These results show that B 1' and B 1 are the minimal elements sufficient for stable replication in conjunction with the HCV 5' NTR.
Having shown that B 1' and B 1 are sufficient for replication in conjunction with the 5 HCV 5' NTR, we next conducted a deletion analysis to determine the sequences within the HCV 5' NTR of BVDV + HCV de1B2B3 required for replication. A large portion of H 1 was deleted in BVDV + HCV de1B2I33H1, while both H1 and H2 were deleted in BVDV +
HCV
delB2B3H1H2. Of these two RNAs, only BVDV + HCV de1B2B3H1 was as infectious as parental BVDV RNA (Fig. 2B). However, the BVDV + HCV delB2B3H1 virus produced 10 smaller plaques than BVDV + HCV delB2B3, indicating that hairpin H1 may augment replication of the chimera. In contrast, BVDV + HCV de1B2B3H1H2 RNA was not infectious (Fig. 2B) and was translated poorly (Fig. 3, lane 11 ). Diminished HCV IRES
activity might be due to deletion of hairpin H2 or juxtaposition of BVDV
hairpins B 1' and B 1 with H3. A third derivative of BVDV + HCV delB2B3, with a Sma I-Sma I deletion 1 S abrogating HCV IRES function by removing H3, was also not infectious (data not shown).
Thus, a 5' NTR consisting of B 1' and B 1 and a functional HCV IRES is sufficient for stable BVDV replication in MDBK cells. Similar results were obtained in BT cells, another BVDV-permissive continuous bovine cell line (data not shown).
20 Replacement of the BVDV 5' NTR with the EMCV 5' NTR
The following experiment was performed to determine whether the BVDV 5' NTR
could be replaced by the 5' NTR of a more phylogenetically distant virus, EMCV. A
derivative of BVDV was created, called 5' EMCV, that contains an exact replacement of the BVDV 5' NTR with the EMCV 5' NTR plus an additional guanosine residue at the 5' terminus 25 for more efficient transcription initiation of T7 polymerase (Fig. 4A). The specific infectivity of 5' EMCV RNA was more than three orders of magnitude lower than BVDV RNA, indicating that it was defective for replication, although its specific infectivity was higher than that of 5' HCV RNA (compare Figs. 4A and 2A). Similar to S' HCV, 5' EMCV
produced heterogeneous plaques, and sequence analysis indicated that pseudorevertants had arisen. The lower specific infectivity of S' EMCV RNA was not likely because of a defect in translation, since the translation efficiency of 5' EMCV RNA was about threefold higher in vitro than that of BVDV RNA (Fig. 3, lane 20 vs. lane 19).
Similar to BVDV + HCV, it was also determined whether the BVDV 5' NTR at the 5' end of the 5' EMCV RNA would increase its specific infectivity. BVDV +
EMCVdeIA (Fig.
4A) contained the entire BVDV S' NTR in tandem with the EMCV 5' NTR lacking a portion of hairpin A. BVDV + EMCVdeIA RNA had a specific infectivity near that of BDVD
RNA
(compare Figs. 4A and 2A) despite having a lower translation efficiency than 5' EMCV (Fig.
3, lane 21 vs. lane 20). Similar to the results with BVDV + HCV, this implicates the added BVDV 5' NTR sequence for a step in viral replication other than translation.
Two derivatives of BVDV + EMCVdeIA that contain deletions of portions of the BDVD 5' NTR but maintain the sequence of B 1' and B 1, BDVD + EMCVdeIB3A and BVDV + EMCVdelB2B3A (Fig.
4A), also were infectious. These derivatives had translation efficiencies near that of the parental BVDV + EMCVdeIA (Fig. 3, compare lanes 15 and 16 with lane 21). This demonstrated that hairpins B 1' and B 1 were sufficient for replication in conjunction with a large portion of the EMCV 5' NTR. Derivatives of BVDV + EMCVdelB3A or BVDV +
EMCVde1B2B3A that contain further deletions of EMCV (BVDV _ EMCVdelB3ABC and BVDV + EMCVde1B2B3ABC in particular) were translated efficiently (Fig. 3, lanes 17 and 1$) and were infectious (Fig. 4B). This indicates that the chimeras did not require putative EMCV RNA replication signals (Martin & Palmenberg, 1996). However, derivatives with 1 S deletions extending into the canonical EMCV IRES were not infectious. For example, BVDV
+ EMCVdelB3A-H and BVDV + EMCVde1B2B3A-H, in which a portion of hairpin H is deleted, were not infectious (Fig. 4B) and were inefficiently translated in vitro (Fig. 3, lanes 22 and 23). It should be noted that all of the BVDV + EMCV chimeras produced plaques of heterogeneous size, indicating some instability.
Relatively simple 5' NTR mutations are observed in adapted pseudorevertants As mentioned previously, BVDV + HCVde1B1B2B3 did not replicate stably as indicated by the heterogeneity in the size of plaques produced by this virus.
Upon passage and selection of medium plaque-producing variants, 5' RACE analysis and sequencing indicated that nt 1-26 had been deleted in the pseudorevertants, removing a large portion of B 1' which was apparently deleterious in the absence of B 1. This deletion results in the 5' terminal sequence 5'GUAUCG which is identical to the first six bases of BVDV
genome RNA (Fig. S) and is repeated at positions 27-32.
Analysis of the passaged 5' EMCV virus indicated that the replicating progeny had also undergone a simple deletion of sequence at the S' end to generate more efficiently replicating variants (Fig. 5). After electroporation, the S' EMCV virus pool was passaged 5 times at a multiplicity of infection of 0.1-1 PFU/cell on MDBK or BT cells, and the 5' termini of three randomly picked plaques were sequenced. For all three plaques selected, nt 2-209 had been deleted, again creating a genome RNA with the 5' terminal tetranucleotide sequence 5'-GUAU.

Analysis of the 5' HCV progeny indicated that more complicated variants had arisen.
Most small plaque-producing variants were unstable and quickly reverted to medium piaque-producing variants. However, one small plaque-producing variant and two stable medium plaque-producing variants were isolated. 5' terminal sequences of the variants were amplified by rapid amplification of cDNA ends (RACE) and cloned into a plasmid vector, and sequences for several independent colonies were determined. The sequence of three clones of the small plaque-producing virus (5'HCV.RI) contained a deletion of HCV
sequence from nt 1-34 and an addition of the dinucleotides 5'-AU in two clones and S'-GU in the third clone.
This creates a 5' terminus of 5'-(G/A) UAA (Fig. SB), reminiscent of the first three bases of the BVDV genome RNA (5'-GUA). Both medium plaque variants appeared to have arisen by RNA recombination with non-viral sequences (Fig. 5). One medium plaque variant (5' HCV.R2) had deleted the first 21 bases of the HCV sequence and contained instead a heterologous sequence of 22 bases. BLAST searches revealed a perfect match between this sequence and a sequence in a human retina cDNA of unknown function (Tsp509I).
The IS second medium plaque variant (S' HCV.R3) had also undergone a possible recombination event leading to the addition of 12 nt to the 5' end of the HCV sequence.
Given its short length, multiple matches were found in the database with this sequence. As for the small plaque variant, sequencing of multiple clones revealed heterogeneity oat the extreme 5' end, with either G of A identified as the 5' base. Remarkably, for both medium plaque variants, the fused heterologous sequence began with the tetranucelotide sequence 5'-(G/A) UAU (Fig.
SB). For all three variants, sequencing of the entire S' NTR and a portion of the IV'"° coding region revealed only these changes at the 5' termini.
5' NTR sequence changes are sufficient for the pseudorevertant phenotypes To assess the importance of these alterations oat the 5' terminus of the 5' HCV
pseudorevertants, derivatives of 5' HCV were created with the changes determined by 5' RACE (Fig. 6A) and analyzed the specific infectivities of these RNAs (Fig.
6B).
Corresponding to the small plaque variant, a derivative called S' HCV.R1 orig was engineered which contained a 5' NTR consisting of the dinucleotide 5' -GU at the S' terminus of HCV nt 35-341. This results in a 5' terminus consisting of 5'-GUAR. 5'HCV.R1 orig RNA
had a specific infectivity at least four orders of magnitude higher than 5' HCV RNA
(Figs. 6B and 2A). This demonstrates that this 5' NTR structure is sufficient for phenotypic reversion to high specific infectivity. However, small plaques and considerable heterogeneity were observed for 5'HCV.RI orig suggesting that additional mutations may be present in the original small plaque variant.

The engineered derivative S'HCV.R2orig had a 5' NTR consisting of 22 nt of Tsp509I-homologous sequence followed by HCV nt 22-341. Another construct, called 5'HCV.R3orig was made, which has the 12 nt of the other heterologous sequence fused to the intact HCV 5' NTR. Specific infectivities for both these derivatives were essentially the same S as observed for wild type BVDV RNA (2-4 x 106 PFU/~g; Fig. 6B). Transfection with these transcripts produced medium plaques, as observed for the original variants, and this phenotype was stable upon passaging. These results show that the altered 5'NTR
sequences were responsible for the pseudorevertant phenotypes rather than changes elsewhere in their genomes.
Addition of the tetranucleotide sequence 5'-GUAU to the HCV 5' NTR allows efficient BVDV replication For all three 5' HCV variants studied, as well as the BVDV + HCV de1B1B2B3 and S'EMCV pseudorevertants, 5' NTR alterations seemed to involve creation of a three- or four-base "consensus" sequence identical to the S' terminus of BVDV genome RNA. To test the importance of this sequence, as opposed to fused heterologous sequences, we created a set of variants with the BVDV 5' tetranucleotide sequence linked to the HCV 5' NTR or the deletion/recombinant break points identified during sequence analysis of the 5' HCV
pseudorevertants (Fig. 6A). 5' HCV.Rlcons had the tetranucleotide sequence 5'-GUAU fused to HCV nt 35-341. 5'HCV.R2cons had the 5'-GUAU tetranucleotide sequence fused to HCV
nt 22-341. 5'HCV.R3cons contained the tetranucleotide sequence 5'-Guau fused to the intact 5' terminus of the HCV NTR. RNAs from all three of these derivatives had specific infectivities more than five orders of magnitude higher than 5'HCV and comparable to parental BVDV (Fig. 6B).
There were, however, significant differences between the phenotypes of some of these derivatives versus the reconstructed pseudorevertants. As mentioned above, 5'HCV.Rlorig yielded tiny and small plaques and produced low virus yields even after 48 h.
In contrast, the addition of four bases rather than two bases (5'-GUAU vs. 5'-GU) yielded virus with near wild-type plaque morphology (Fig. 6B) and growth Rates (Fig.
7). In the case of the smaller deletion, 5'HCV.R2orig and 5'HCV.R2cons were indistinguishable, suggesting that, other than the 5' four bases, the fused heterologous sequences were dispensable. This 3 S was not he case, however, for the chimera containing the S'-GUAU
tetranucleotide sequence fused to the intact HCV 5' NTR. 5'HCV.R3cons produced small plaques (Fig. 6B) and grew more slowly than S'HCV.R3orig (Fig. 7) suggesting that the sequence/structure of the sequences downstream of the 5' four bases can affect replication efficiency.
The tetranucleotide sequence 5'-GUAU is important for efficient BVDV RNA
accumulation Next, the effects of the different 5' termini on virus-specific RNA
accumulation directly after transfection were analyzed. This allowed a direct comparison between 5'HCV
and the reconstructed pseudorevertants as well as selected BVDV + HCV deletion constructs.
MDBK cells were transfected with in vitro synthesized RNAs and labeled for 10 h beginning at 5 h post-transfection with 3H-UTP in the presence of actinomycin D (Fig.
8). RNA
replication of the 5' HCV chimera was severely impaired to a level below detection (Fig. 8, lane 2). In contrast, every 5' NTR alteration of 5' HCV that increased RNA
specific infectivity and allowed efficient virus growth led to readily detectable viral RNA
accumulation. Addition of B 1' and B 1 to the 5' terminus of the HCV 5' NTR
restored RNA
replication to a level ~50% of that observed for BVDV (BVDV + HCVdelB2B3; Fig.
8, lane 3 vs. lane 1). BVDV + HCVde1B2B3H1 displayed reduced RNA synthesis compared to BVDV + HCVdeIB2B3 (Fig. 8, lane 4 vs. lane 3) perhaps explaining its small plaque phenotype and suggesting a possible positive role for H1 in replication of this chimera.
S'HCV.Rlorig, which had exhibited plaque heterogeneity and slow growth, accumulated less RNA when compared to 5'HCV.Rlcons (Fig. 8, lane 5 vs. lane 6). 5'HCV.R2orig and 5'HCV.R2cons showed similar RNA accumulation (Fig. 8, lane 9 vs. lane 10) consistent with their medium plaque phenotypes; and 5'HCV.R3cons exhibited reduced RNA
synthesis compared to 5'HCV.R3orig (Fig. 8, lane 8 vs, lane 7), consistent with their small-versus medium-plaque phenotypes.
Although these RNA phenotypes acre complex, the most striking result is that addition of the B 1' B 1 hairpins, addition of heterologous 5' sequences terminating with 5'-GUAU or simply fusion of this tetranucleotide sequence with the HCV 5' NTR or short 5' truncations of the HCV 5' NTR all dramatically upreguiated RNA accumulation. This occurred without increasing translation efficiency, at least as measured in a cell-free assay (Fig. 3, compare lanes 3-8 to lane 1), suggesting that these sequences function at the level of RNA replication or stability.

WO 99/553bb PCT/US99/08850 Discussion The work presented here helps to define the requirements for a functional BVDV
SNTR. The BVDV-specific 5' NTR sequences required for efficient replication in cell culture are minimal and consist of the 5' terminal sequence, 5'-GUAU. The sequence 5'-5 AUAU, detected for some pseudorevertants, may also be functional but this was not tested for technical reasons. This simple 5'-terminal tetranucleotide sequence, which is conserved among pestivirses (Ruggli et al., 1996; Becher et al., 1998), was shown to function in the context of functional IRES elements derived from the hepacivirus HCV or the picornavirus EMCV. As discussed below, this may indicate that the 5' signals required for BVDV RNA
10 replication are rather simple or that elements in these heterologous IRESs can functionally replace deleted BVDV sequences.
Sequences at the extreme 5' end of BVDV genome RNA could modulate the efficiency of RNA accumulation by affecting RNA stability, translation, promoter efficiency, or some combination of these processes. At this time, we can not distinguish among these 15 possibilities but favor an effect on RNA replication. The complement of the BVDV 5' sequence at the 3' end of the negative-strand RNA presumably functions in the initiation of positive-strand RNA synthesis. 'Thus, AUAC-3' at the 3'terminus fo minus-strand RNA may be important for positive-strand RNA synthesis. Interestingly, for some positive-strand RNA
viruses such as rubella virus (Pugachev & Frey, 1998), flock house virus (Ball, 1994) and 20 turnip crinkle virus (Guar et al., 1997), only minimal cis-acting sequences at the 3' termini of negative-strand RNAs are required positive-strand RNA synthesis. In contrast to the 5' NTR
replacements, we were unable to generate replication-competent BVDV-HCV
replacing that of BVDV (data not shown). This may indicate that the signals within the pestivirus 3' NTR
required for initiation of negative-strand RNA synthesis are more complex and virus specific.
25 Once the replication complex has assembled at the 3' NTR and transversed the RNA during negative-strand synthesis, the requirements of the 5' NTR for initiation of positive-strand synthesis may be minimal.
Although the RNA replication signals within the 5' NTR appear to be rather simple, it is possible that the signals important for RNA replication actually extend into the IRES and 30 are more complicated. For instance, the 5'HCV pseudorevertants were more stable and grew to higher titers than the 5'EMCV counterparts, despite the fact that the 5'EMCV RNAs were translated more efficiently in vitro. This may indicate that the BVDV and HCV
IRESs contain signals important for RNA synthesis that are absent in the EMCV IRES.
It is perhaps not surprising that 5' HCV appeared to recombine with cellular mRNAs to acquire a 5' terminus with the 5' -(G/A) UAU consensus, given that non-cytopathic strains of BVDV can recombine with BVDV RNA or cellular mRNAs to generate cytopathic strains of BVDV (Meyers & Thiel, 1996). Presumably, this recombination event involves template switching during negative-strand RNA synthesis, as observed for polio-virus {Kirkegaard &
Baltimore, 1986). In contrast to 5' HCV, simple deletions of 5' terminal viral sequences could S account for the BVDV + HCVde1B1B2B3 and 5'EMCV pseudorevertants since the tetranucleotide sequence is present in these 5' NTRs upstream of functional IRES elements.
Such deletions could occur by partial degradation of positive-strand template prior to negative-strand synthesis, by premature termination during negative-strand RNA
synthesis, or by degradation of 3' terminal negative-strand sequence after synthesis. It is proposed that 5'HCV was forced to recombine with cellular sequences because HCV does not have an 5'-(G/A) UAU sequence upstream of its IRES. The first occurrence of an (G/A)UAUA
tetranucleotide sequence is at nt 94-97 within hairpin H2, and a 5' deletion extending into this sequence would presumably inactivate or severely impair HCV IRES activity. It is interesting that BVDV + HCVde1B1B2B3 and S'EMCV pseudorevertants were generated at much higher frequency than 5'HCV pseudorevertants. This may indicate that recombination between BVDV and cellular RNAs is a rare event compared to the processes which lead to deletion of terminal viral sequences.
Poliovirus chimeras dependent upon a functional HCV IRES have been reported (Lu & Wimmer, 1996). Interestingly, viable poliovirus chimeras were produced only when HCV
sequences included both the IRES and the N-terminal portion of the HCV ORF.
Nucleotide sequences or structures in the downstream ORF can modulate HCV IRES
translational efficiency (see Reynolds et al., 1995; Honda et al., 1996a) but it was also suggested that the N-terminal portion of the HCV core polypeptide might be involved. In the case of our 5' HCV pseudorevertants, there is no requirement for HCV C protein sequences.
Although the translation efficiency of the HCV IRES in the presence of additional HCV
sequences 3' to the AUG start was not directly assessed, the HCV chimeras and pseudorevertants were translationally active and infectious in the absence of any portion of the HCV
ORF. This indicates that either the HCV IRES does not extend into the HCV ORF or that the BVDV
ORF contains analogous sequence which functions in our 5'HCV chimeras. There is some limited identity between HCV and BVDV within this region. For example, HCV nt and BVDV nt 405-440 are identical at 21 of 36 positions, although identity within this sequence may be attributed to a high adenosine content. It is interesting to note that the luciferase (LUC) and chloramphenicol acetyl transferase (CAT) reporter genes previously used to detect HCV IRES activity (Tsukiyama-Kohara et al., 1992; Wang et al., 1993) also have adenosine- or purine-rich regions in relatively the same position as the HCV ORF and BVDV ORF. It this region is indeed important for IRES activity, this may explain why some have observed that the HCV IRES does not require a portion of the HCV ORF for translation of CAT or LUC (Tsukiyama-Kohara et al., 1992; Wang et al., 1993). Point mutations and insertions within this region of HCV have been shown to reduce HCV IRES
activity in vitro (Honda et al., 1996a,b).
Despite the fact that B 1' and B 1 are conserved among different strains of BVDV and similar hairpins are present in border disease virus and CSFV (Deng & Brock, 1993; Becher et al., 1998), Bl' and B1 were dispensable for BVDV replication, provided that the 5' tetranucleotide sequence 5'-(G/A)UAU remained. This may indicate a role for B
1' and B 1 in viral replication in vivo that we do not observe in cell culture. It will be interesting to test the phenotype of chimeras that lack B I' and B I in vivo to determine if they are attenuated and might serve as useful BVDV vaccines. In this vein, several studies with flaviviruses have demonstrated that alterations in S' NTR or 3' NTR elements can lead to attenuation in vivo (Cahour et al., 1995; Men et a., 1996; Mandl et al., 1998). BVDV chimeras that utilize the HCV or EMCV IRES may also prove to be attenuated simply due to the presence of the heterologous IRES. For poliovirus, it has been shown that differences in IRES
efficiency in different host-cell environments can modulate host range and virulence (Shiroki et al., 1997).
BVDV-HCV chimeras that are dependent on a functional HCV IRES may have another practical application. It may be possible to use these chimeras to screen for anti-HCV
therapeutics that target the HCV IRES. Other researchers have shown antisense oligonucleotide-mediated inhibition of HCV gene expression in hepatocytes by targeting the oligonucleotides to the HCV IRES (Hanecak et al., 1996). It will be of interest to measure the efficacy of antisense oligonucleotides or ribozymes (Lieber et al., 1996) against replicating virus, and these chimeras are more useful than HCV for this purpose since they are able to replicate efficiently in cell culture. BVDV is believed to be a reasonable model of HCV
replication not only because of homology and conserved motifs within the S' NTR but also because of similarities in overall genetic organization (Rice, 1996) and polyprotein processing strategy (Tautz et al., 1997; Xu et al., 1997).
In view of the above, it will be seen that the several advantages of the invention are achieved and other advantageous results attained.
As various changes could be made in the above methods and compositions without departing from the scope of the invention, it is intended that all matter contained in the above description and shown in the accompanying drawings shall be interpreted as illustrative and not in a limiting sense.

All references cited in this specification, including patents and patent applications, are hereby incorporated by reference. The discussion of references herein is intended merely to summarize the assertions made by their authors and no admission is made that any reference constitutes prior art. Applicants reserve the right to challenge the accuracy and pertinency of the cited references.

Claims (23)

What is Claimed is:

1. A polynucleotide comprising a chimeric viral RNA which comprises:
(a) a 5' nontranslated region (5' NTR);
(b) an open reading frame (ORF) region; and (c) a 3' nontranslated region (3' NTR);
wherein at least one of said regions is chimeric and comprises a first nucleotide sequence from a pestivirus in operable linkage with a first nucleotide sequence from an hepatitis C
virus (HCV), and wherein said chimeric viral RNA is replication-competent.

2. The polynucleotide of claim 1, wherein the chimeric region is the 5' NTR
and the first pestivirus nucleotide sequence is from a bovine viral diarrhea virus (BVDV).

3.The polynucleotide of claim 2, wherein the BVDV nucleotide sequence is located at the 5'terminus of the chimeric 5'NTR and comprises 5' RUAU.

4. The polynucleotide of claim 3, wherein the first HCV nucleotide sequence in the chimeric 5' NTR comprises an internal ribosome entry site (IRES).

5.The polynucleotide of claim 4, wherein the ORF and the 3' NTR consist of second and third BVDV sequences.

6. The polynucleotide of claim 5, wherein the 5'terminal sequence comprises 5' GUAU.

7. The polynucleotide of claim 4, wherein the ORF comprises a second HCV
sequence encoding at least one structural protein operably linked to a second BVDV
sequence.

8.The polynucleotide of claim 1, wherein the pestivirus is BVDV and the chimeric region is the 3' NTR.

9. The polynucleotide of claim 8, wherein the first HCV sequence in the chimeric 3' NTR comprises the HCV 98 by 3' terminal element (SEQ ID NO:X) operably linked to the first BVDV sequence.

10. A method for identifying compounds having antiviral activity against hepatitis C virus (HCV) comprising the steps of:
(a) providing a first cell containing a chimeric viral RNA which is replication competent in the cell, the chimeric viral nucleic acid comprising a 5' nontranslated region (5' NTR), an open reading frame (ORF) region; and a 3' nontranslated region (3'NTR);
wherein at least one of said regions is chimeric and comprises a first nucleotide sequence from a pestivirus in operable linkage with a first nucleotide sequence from an hepatitis C
virus (HCV);
(b) providing a second cell containing the pestivirus; and (c) comparing the replication efficiency of the chimeric viral RNA acid in the presence and absence of a test compound to the replication efficiency of the pestivirus in the presence and absence of the test compound, wherein a greater reduction in compound-induced replication efficiency of the chimeric viral RNA than the pestivirus indicates the compound has anti-HCV activity.

11. The method of claim 10, wherein the chimeric region is the 5' NTR and the first pestivirus nucleotide sequence is from a bovine viral diarrhea virus (BVDV).

12. The method of claim 11, wherein the BVDV nucleotide sequence is located at the 5' terminus of the chimeric 5' NTR and comprises 5' RUAU.

13. The method of claim 12, wherein the first HCV nucleotide sequence in the chimeric 5' NTR comprises an internal ribosome entry site (IRES).

14. The method of claim 13, wherein the ORF and the 3' NTR comprise second and third sequences from the BVDV.

15. The method of claim 10, wherein the pestivirus is BVDV and the chimeric region is the 3' NTR.

16. A genetically-engineered virus comprising a chimeric RNA genome which comprises:
(a) a 5' nontranslated region (5' NTR);
(b) an open reading frame (ORF) region; and (c) a 3' nontranslated region (3' NTR);

wherein at least one of said regions is chimeric and comprises a first nucleotide sequence from a pestivirus in operable linkage with a first nucleotide sequence from an hepatitis C
virus (HCV), and wherein said chimeric RNA genome is replication-competent.

17. The genetically-engineered virus of claim 16, wherein the chimeric region is the 5' NTR and the first pestivirus nucleotide sequence is from a bovine viral diarrhea virus (BVDV).

18. The genetically-engineered virus of claim 16, wherein the BVDV nucleotide sequence is located at the 5' terminus of the chimeric 5' NTR and comprises 5' RUAU and the first HCV nucleotide sequence in the chimeric 5' NTR comprises an internal ribosome entry site (IRES).

19. A vaccine against bovine viral diarrhea virus (BVDV) comprising an immunogenically-effective amount of a genetically-engineered virus comprising a chimeric RNA genome having:
(a) a 5' immunogenically region (5' NTR);
(b) an open reading frame (ORF) region; and (c) a 3' nontranslated region (3' NTR);
wherein at least one of said regions is chimeric and comprises a first nucleotide sequence from BVDV in operable linkage with a first nucleotide sequence from an hepatitis C virus (HCV), and wherein the genetically-engineered virus is attenuated as compared to BVDV.

20. The vaccine of claim 19, wherein the chimeric region is the 5' NTR and the BVDV nucleotide sequence is located at the 5' terminus of the chimeric 5' NTR
and comprises 5' RUAU and the first HCV nucleotide sequence in the chimeric 5' NTR
comprises an internal ribosome entry site (IRES).

21. A polynucleotide comprising a chimeric viral RNA which comprises:
(a) a 5' nontranslated region (5' NTR);
(b) an open reading frame (ORF) region; and (c) a 3' nontranslated region (3' NTR);
wherein at least one of said regions is chimeric and comprises a first nucleotide sequence from a pestivirus in operable linkage with a heterologous nucleotide sequence and wherein said chimeric viral RNA is replication-competent.

22. A polynucleotide comprising a DNA-dependent promoter operably linked to a cDNA
of the chimeric viral RNA of claim 1.

23. A cell comprising the polynucleotide of claim 22.