Hepatitis C Virus Protease

This invention relates to the molecular biology and virology of the hepatitis C virus (HCV). More specifically, this invention relates to a novel protease produced by HCV, methods of expression, recombinant protease, protease mutants, and inhibitors of HCV protease.

Non-A, Non-B hepatitis (NANBH) is a transmissible disease (or family of diseases) that is believed to be virally induced, and is distinguishable from other forms of virus-associated liver disease, such as those caused by hepatitis A virus (HAV), hepatitis B virus (HBV), delta hepatitis virus (HDV), cytomegalovirus (CMV) or Epstein-Barr virus (EBV). Epidemiologic evidence suggests that there may be three types of NANBH: the water-borne epidemic type; the blood or needle associated type; and the sporadically occurring (community acquired) type. However, the number of causative agents is unknown. Recently, however, a new viral species, hepatitis C virus (HCV) has been identified as the primary (if not only) cause of blood-associated NANBH (BB-NANBH). See for example, PCT WO 89/04669; and U.S. patent application Ser. No. 7/456,637, filed 21 Dec. 1989, incorporated herein by reference. Hepatitis C appears to be the major form of transfusion-associated hepatitis in a number of countries, including the United States and Japan. There is also evidence implicating HCV in induction of hepatocellular carcinoma. Thus, a need exists for an effective method for treating HCV infection: currently, there is none.

Many viruses, including adenoviruses, baculoviruses, comoviruses, picomaviruses, retroviruses, and togaviruses, rely on specific, virally-encoded proteases for processing polypeptides from their initial translated form into mature, active proteins. In the case of picornaviruses, all of the viral proteins are believed to arise from cleavage of a single polyprotein (B. D. Korant, CRC Crit Rev Biotech (1988) 8:149-57).

S. Pichuantes et al, in "Viral Proteinases As Targets For Chemotherapy" (Cold Spring Harbor Laboratory Press, 1989) pp. 215-22, disclosed expression of a viral protease found in HIV-1. The HIV protease was obtained in the form of a fusion protein, by fusing DNA encoding an HIV protease precursor to DNA encoding human superoxide dismutase (hSOD), and expressing the product in E. coli. Transformed cells expressed products of 36 and 10 kDa (corresponding to the hSOD-protease fusion protein and the protease alone), suggesting that the protease was expressed in a form capable of autocatalytic proteolysis.

T. J. McQuade et al, Science (1990) 247:454-56 disclosed preparation of a peptide mimic capable of specifically inhibiting the HIV-1 protease. In HIV, the protease is believed responsible for cleavage of the initial p55 gag precursor transcript into the core structural proteins (p17, p24, p8, and p7). Adding 1 .mu.M inhibitor to HIV-infected peripheral blood lymphocytes in culture reduced the concentration of processed HIV p24 by about 70%. Viral maturation and levels of infectious virus were reduced by the protease inhibitor.

We have now invented recombinant HCV protease, HCV protease fusion proteins, truncated and altered HCV proteases, cloning and expression vectors therefore, and methods for identifying antiviral agents effective for treating HCV.

FIG. 1 shows the sequence of HCV protease (SEQ ID NO: 69 and SEQ ID NO: 70).

FIG. 2 shows the polynucleotide sequence and deduced amino acid (SEQ ID NO: 71 and SEQ ID NO: 72) sequence of the clone C20c.

FIG. 3 shows the polynucleotide sequence and deduced amino acid sequence of the clone C26d. (SEQ ID NO: 73 and SEQ ID NO: 74).

FIG. 4 shows the polynucleotide sequence and deduced amino acid sequence of the clone C8h (SEQ ID NO: 75 and SEQ ID NO: 76).

FIG. 5 shows the polynucleotide sequence and deduced amino acid sequence of the clone C7f (SEQ ID NO: 77 and SEQ ID NO: 78).

FIG. 6 shows the polynucleotide sequence and deduced amino acid sequence of the clone C31 (SEQ ID NO: 79 and SEQ ID NO: 80).

FIG. 7 shows the polynucleotide sequence and deduced amino acid sequence of the clone C35 (SEQ ID NO: 81 and SEQ ID NO: 82).

FIG. 8 shows the polynucleotide sequence and deduced amino acid sequence of the clone C33c (SEQ ID NO: 83 and SEQ ID NO: 84).

FIG. 9 schematically illustrates assembly of the vector C7fC20cC300C200.

FIG. 10 shows the sequence of vector cflSODp600 SEQ ID NO: 85 and SEQ ID NO: 86).

The terms "Hepatitis C Virus" and "HCV" refer to the vital species that is the major etiological agent of BB-NANBH, the prototype isolate of which is identified in PCT WO89/046699; EPO publication 318,216; U.S. Ser. No. 7/355,008, filed 18 May 1989; and U.S. Ser. No. 7/456,637, the disclosures of which are incorporated herein by reference. "HCV" as used herein includes the pathogenic strains capable of causing hepatitis C, and attenuated strains or defective interfering particles derived therefrom. The HCV genome is comprised of RNA. It is known that RNA-containing viruses have relatively high rates of spontaneous mutation, reportedly on the order of 10.sup.-3 to 10.sup.-4 per incorporated nucleotide Wields & Knipe, "Fundamental Virology" (1986, Raven Press, N.Y.)). As heterogencity and fluidity of genotype are inherent characteristics of RNA viruses, there will be multiple strains/isolates, which may be virulent or avirulent, within the HCV species.

Information on several different strains/isolates of HCV is disclosed herein, particularly strain or isolate CDC/HCVI (also called HCV1). Information from one strain or isolate, such as a partial genomic sequence, is sufficient to allow those skilled in the art using standard techniques to isolate new strains/isolates and to identify whether such new strains/isolates are HCV. For example, several different strains/isolates are described below. These strains, which were obtained from a number of human sera (and from different geographical areas), were isolated utilizing the information from the genomic sequence of HCV1.

The information provided herein suggests that HCV may be distantly related to the flaviviridae. The Flavivirus family contains a large number of viruses which are small, enveloped pathogens of man. The morphology and composition of Flavivirus particles are known, and are discussed in M. A. Brinton, in "The Viruses: The Togaviridae And Flaviviridac" (Series eds. Fraenkel-Conrat and Wagner, vol. eds. Schlesinger and Schlesinger, Plenum Press, 1986), pp. 327-374. Generally, with respect to morphology, Flaviviruses contain a central nucleocapsid surrounded by a lipid bilayer. Virions are spherical and have a diameter of about 40-50 nm. Their cores are about 25-30 nm in diameter. Along the outer surface of the virion envelope are projections measuring about 5-10 nm in length with terminal knobs about 2 nm in diameter. Typical examples of the family include Yellow Fever virus, West Nile virus, and Dengue Fever virus. They possess positive-stranded RNA genomes (about 11,000 nucleotides) that are slightly larger than that of HCV and encode a polyprotein precursor of about 3500 amino acids. Individual viral proteins are cleaved from this precursor polypeptide.

The genome of HCV appears to be single-stranded RNA containing about 10,000 nucleotides. The genome is positive-stranded, and possesses a continuous translational open reading frame (ORF) that encodes a polyprotein of about 3,000 amino acids. In the ORF, the structural proteins appear to be encoded in approximately the first quarter of the N-terminal region, with the majority of the polyprotein attributed to non-structural proteins. When compared with all known viral sequences, small but significant co-linear homologics are observed with the nonstructural proteins of the Flavivirus family, and with the pestiviruses (which are now also considered to be part of the Flavivirus family).

A schematic alignment of possible regions of a flaviviral polyprotein (using Yellow Fever Virus as an example), and of a putative polyprotein encoded in the major ORF of the HCV genome, is shown in FIG. 1. Possible domains of the HCV polyprotein are indicated in the figure. The Yellow Fever Virus polyprotein contains, from the amino terminus to the carboxy terminus, the nucleocapsid protein (C), the matrix protein (M), the envelope protein 0), and the non-structural proteins 1, 2 (a+b), 3, 4 (a+b), and 5 (NS1, NS2, NS3, NS4, and NS5). Based upon the putative amino acids encoded in the nucleotide sequence of HCV1, a small domain at the content of basic residues to the nucleocapsid protein (C) found at the N-terminus of flaviviral polyproteins. The non-structural proteins 2,3,4, and 5 (NS2-5) of HCV and of yellow fever virus (YFV) appear to have counterparts of similar size and hydropathicity, although the amino acid sequences diverge. However, the region of HCV which would correspond to the regions of YFV polyprotein which contains the M, E, and NS1 protein not only differs in sequence, but also appears to be quite different in size and hydropathicity. Thus, while certain domains of the HCV genome may be referred to herein as, for example, NS1, or NS2, it should be understood that these designations are for convenience of reference only; there may be considerable differences between the HCV family and flaviviruses that have yet to be appreciated.

Due to the evolutionary relationship of the strains or isolates of HCV, putative HCV strains and isolates are identifiable by their homology at the polypeptide level. With respect to the isolates disclosed herein, new HCV strains or isolates are expected to be at least about 40% homologous, some more than about 70% homologous, and some even more than about 80% homologous: some may be more than about 90% homologous at the polypeptide level. The techniques for determining amino acid sequence homology are known in the art. For example, the amino acid sequence may be determined directly and compared to the sequences provided herein. Alternatively the nucleotide sequence of the genomic material of the putative HCV may be determined (usually via a eDNA intermediate), the amino acid sequence encoded therein can be determined, and the corresponding regions compared.

The term "HCV protease" refers to an enzyme derived from HCV which exhibits proteolytic activity, specifically the polypeptide encoded in the NS3 domain of the HCV genome. At least one strain of HCV contains a protease believed to be substantially encoded by or within the following sequence: ##STR1##

The above N and C termini are putative, the actual termini being defined by expression and processing in an appropriate host of a DNA construct encoding the entire NS3 domain. It is understood that this sequence may vary from strain to strain, as RNA viruses like HCV are known to exhibit a great deal of variation. Further, the actual N and C termini may vary, as the protease is cleaved from a precursor polyprotein: variations in the protease amino acid sequence can result in cleavage from the polyprotein at different points. Thus, the amino- and carboxy-termini may differ from strain to strain of HCV. The first amino acid shown above corresponds to residue 60 in FIG. 1. However, the minimum sequence necessary for activity can be determined by routine methods. The sequence may be truncated at either end by treating an appropriate expression vector with an exonuclease (after cleavage at the 5' or 3' end of the coding sequence) to remove any desired number of base pairs. The resulting coding polynucleotide is then expressed and the sequence determined. In this manner the activity of the resulting product may be correlated with the amino acid sequence: a limited series of such experiments (removing progressively greater numbers of base pairs) determines the minimum internal sequence necessary for protease activity. We have found that the sequence may be substantially truncated, particularly at the carboxy terminus, apparently with full retention of protease activity. It is presently believed that a portion of the protein at the carboxy terminus may exhibit helicase activity. However, helicase activity is not required of the HCV proteases of the invention. The amino terminus may also be truncated to a degree without loss of protease activity.

The amino acids underlined above are believed to be the residues necessary for catalytic activity, based on sequence homology to putative flavivirus serine proteases. Table 1 shows the alignment of the three serine protease catalytic residues for HCV protease and the protease obtained from Yellow Fever Virus, West Nile Fever virus, Murray Valley Fever virus, and Kunjin virus. Although the other four flavivirus protease sequences exhibit higher homology with each other than with HCV, a degree of homology is still observed with HCV. This homology, however, was not sufficient for indication by currently available alignment software. The indicated amino acids are numbered His.sub.79, Asp.sub.103, and Ser.sub.161 in the sequence listed above (His.sub.139, Asps.sub.163, and Ser.sub.221 in FIG. 1 ).

Alternatively, one can make catalytic residue assignments based on structural homology. Table 2 shows alignment of HCV with against the catalytic sites of several well-characterized serine proteases based on structural considerations: protease A from Streptomyces griseus, .alpha.-lytic protease, bovine trypsin, chymotrypsin, and elastase (M. James et al, Can J Biochem (1978) 56:396). Again, a degree of homology is observed. The HCV residues identified are numbered His.sub.79, Asps.sub.125, and Ser.sub.161 in the sequence listed above.

The most direct manner to verify the residues essential to the active site is to replace each residue individually with a residue of equivalent stearic size. This is easily accomplished by site-specific mutagenesis and similar methods known in the art. If replacement of a particular residue with a residue of equivalent size results in loss of activity, the essential nature of the replaced residue is confirmed.

"HCV protease analogs" refer to polypeptides which vary from the full length protease sequence by deletion, alteration and/or addition to the amino acid sequence of the native protease. HCV protease analogs include the truncated proteases described above, as well as HCV protease muteins and fusion proteins comprising HCV protease, truncated protease, or protease muteins. Alterations to form HCV protease muteins are preferably conservative amino acid substitutions, in which an amino acid is replaced with another naturally-occurring amino acid of similar character. For example, the following substitutions are considered "conservative": ##STR2## Nonconservative changes are generally substitutions of one of the above amino acids with an amino acid from a different group (e.g., substituting Asn for Glu), or substituting Cys, Met, His, or Pro for any of the above amino acids. Substitutions involving common amino acids are conveniently performed by site specific mutagenesis of an expression vector encoding the desired protein, and subsequent expression of the altered form. One may also alter amino acids by synthetic or semi-synthetic methods. For example, one may convert cysteine or serine residues to selenocysteine by appropriate chemical treatment of the isolated protein. Alternatively, one may incorporate uncommon amino acids in standard in vitro protein synthetic methods. Typically, the total number of residues changed, deleted or added to the native sequence in the muteins will be no more than about 20, preferably no more than about 10, and most preferably no more than about 5.

The term fusion protein generally refers to a polypeptide comprising an amino acid sequence drawn from two or more individual proteins. In the present invention, "fusion protein" is used to denote a polypeptide comprising the HCV protease, truncate, mutein or a functional portion thereof, fused to a non-HCV protein or polypeptide ("fusion partner"). Fusion proteins are most conveniently produced by expression of a fused gene, which encodes a portion of one polypeptide at the 5' end and a portion of a different polypeptide at the 3' end, where the different portions are joined in one reading frame which may be expressed in a suitable host. It is presently preferred (although not required) to position the HCV protease or analog at the carboxy terminus of the fusion protein, and to employ a functional enzyme fragment at the amino terminus. As the HCV protease is normally expressed within a large polyprotein, it is not expected to include cell transport signals (e.g., export or secretion signals). Suitable functional enzyme fragments are those polypeptides which exhibit a quantifiable activity when expressed fused to the HCV protease. Exemplary enzymes include, without limitation, .beta.-galactosidase (.beta.-gal), .beta.-lactamase, horseradish peroxidase (HRP), glucose oxidase (GO), human superoxide dismutase (hSOD), urease, and the like. These enzymes are convenient because the amount of fusion protein produced can be quantified by means of simple colorimetric assays. Alternatively, one may employ antigenie proteins or fragments, to permit simple detection and quantification of fusion proteins using antibodies specific for the fusion parmer. The presently preferred fusion parmer is hSOD.

The practice of the present invention generally employs conventional techniques of molecular biology, microbiology, recombinant DNA, and immunology, which are within the skill of the art. Such techniques are explained fully in the literature. See for example J. Sambrook et al, "Molecular Cloning; A Laboratory Manual (1989); "DNA Cloning", Vol. I and II (D. N Glover ed. 1985); "Oligonucleotide Synthesis" (M. J. Gait cd, 1984); "Nucleic Acid Hybridization" (B. D. Hames & S. J. Higgins eds. 1984); "Transcription And Translation" (B. D. Hames & S. J. Higgins eds. 1984); "Animal Cell Culture" (R. I. Freshney ed. 1986); "Immobilized Cells And Enzymes" (IRL Press, 1986); B. Perbal, "A Practical Guide To Molecular Cloning" (1984); the series, "Methods In Enzymology" (Academic Press, Inc.); "Gene Transfer Vectors For Mammalian Cells" (J. H. Miller and M. P. Calos eds. 1987, Cold Spring Harbor Laboratory); Meth Enzymol (1987) 154 and 155 (Wu and Grossman, and Wu, eds., respectively); Mayer & Walker, eds. (1987), "Immunochemical Methods In Cell And Molecular Biology" (Academic Press, London); Scopes, and "Handbook Of Experimental Immunology", volumes I-IV (Weir and Blackwell, eds, 1986).

Both prokaryotic and cukaryotic host cells are useful for expressing desired coding sequences when appropriate control sequences compatible with the designated host are used. Among prokaryotic hosts, E. coli is most frequently used. Expression control sequences for prokaryotes include promoters, optionally containing operator portions, and ribosome binding sites. Transfer vectors compatible with prokaryotic hosts are commonly derived from, for example, pBR322, a plasmid containing operons conferring ampicillin and tetracycline resistance, and the various pUC vectors, which also contain sequences conferring antibiotic resistance markers. These plasmids are commercially available. The markers may be used to obtain successful transformants by selection. Commonly used prokaryotic control sequences include the .beta.-lactamase (penicillinase) and lactose promoter systems (Chang et al, Nature (1977) 198:1056), the tryptophan (trp) promoter system (Goeddel et al, Nuc Acids Res (1980) 8:4057) and the lambda-derived P.sub.L promoter and N gene ribosome binding site (Shimatake et al, Nature (1981 ) 292:128) and the hybrid tac promoter (De Boer et al, Proc Nat Acad Sci USA (1983) 292:128) derived from sequences of the trp and lac UV5 promoters. The foregoing systems are particularly compatible with E. coli; if desired, other prokaryotic hosts such as strains of Bacillus or Pseudomonas may be used, with corresponding control sequences.

Eukaryotic hosts include without limitation yeast and mammalian cells in culture systems. Yeast expression hosts include Saccharomyces, Klebsiella, Picia, and the like. Saccharomyces cerevisiae and Saccharomyces carlsbergensis and K. lactis are the most commonly used yeast hosts, and are convenient fungal hosts. Yeast-compatible vectors carry markers which permit selection of successful transformants by conferring prototrophy to auxotrophic routants or resistance to heavy metals on wild-type strains. Yeast compatible vectors may employ the 2 .mu.origin of replication (Broach et al, Meth Enzymol (1983) 101:307), the combination of CEN3 and ARS1 or other means for assuring replication, such as sequences which will result in incorporation of an appropriate fragment into the host cell genome. Control sequences for yeast vectors am known in the an and include promoters for the synthesis of glycolytic enzymes (Hess et al, J Adv Enzyme Reg (1968) 7:149; Holland et al, Biochem (1978), 17:4900), including the promoter for 3-phosphoglycerate kinase (R. Hitzeman et al, J Biol Chem (1980) 255:2073). Terminators may also be included, such as those derived from the enolase gene (Holland, J Biol Chem (1981) 256:1385). Particularly useful control systems are those which comprise the glyceraldehyde-3 phosphate dehydrogenase (GAPDH) promoter or alcohol dehydrogenase (ADH) regulatable promoter, terminators also derived from GAPDH, and if secretion is desired, a leader sequence derived from yeast .alpha.-factor (see U.S. Pat. No. 4,870,008, incorporated herein by reference).

A presently preferred expression system employs the ubiquitin leader as the fusion panner. Copending application U.S. Ser. No. 7/390,599 filed 7 Aug. 1989 disclosed vectors for high expression of yeast ubiquitin fusion proteins. Yeast ubiquitin provides a 76 amino acid polypeptide which is automatically cleaved from the fused protein upon expression. The ubiquitin amino acid sequence is as follows: ##STR3##

See also Ozkaynak et al, Nature (1984) 312:663-66. Polynucleotides encoding the ubiquitin polypeptide may be synthesized by standard methods, for example following the technique of Barret al, J Biol Chem (1988) 268:1671-78 using an Applied Biosystem 380A DNA synthesizer. Using appropriate linkers, the ubiquitin gene may be inserted into a suitable vector and ligated to a sequence encoding the HCV protease or a fragment thereof.

In addition, the transcriptional regulatory region and the transcriptional initiation region which are operably linked may be such that they are not naturally associated in the wild-type organism. These systems are described in detail in EPO 120,551, published Oct. 3, 1984; EPO 116,201, published Aug. 22, 1984; and EPO 164,556, published Dec. 18, 1985, all of which are commonly owned with the present invention, and are hereby incorporated herein by reference in full.

Mammalian cell lines available as hosts for expression are known in the an and include many immortalized cell lines available from the American Type Culture Collection (ATCC), including HeLa cells, Chinese hamster ovary (CHO) cells, baby hamster kidney (BHK) cells, and a number of other cell lines. Suitable promoters for mammalian cells are also known in the an and include viral promoters such as that from Simian Virus 40 (SV40) (Fiers et al, Nature (1978) 273:113), Rous sarcoma virus (RSV), adenovirus (ADV), and bovine papilloma virus (BPV). Mammalian cells may also require terminator sequences and poly-A addition sequences. Enhancer sequences which increase expression may also be included, and sequences which promote amplification of the gene may also be desirable (for example methotrexate resistance genes). These sequences are known in the art.

Vectors suitable for replication in mammalian cells are known in the art, and may include viral replicons, or sequences which insure integration of the appropriate sequences encoding HCV epitopes into the host genome. For example, another vector used to express foreign DNA is Vaccinia virus. In this case the heterologous DNA is inserted into the Vaccinia genome. Techniques for the insertion of foreign DNA into the vaccinia virus genome are known in the art, and may utilize, for example, homologous recombination. The heterologous DNA is generally inserted into a gene which is non-essential to the virus, for example, the thymidine kinase gene (tk), which also provides a selectable marker. Plasmid vectors that greatly facilitate the construction of recombinant viruses have been described (see, for example, Mackett et al, J Virol (1984) 49:857; Chakrabarti et al, Mol Cell Biol (1985) 5:3403; Moss, in GENE TRANSFER VECTORS FOR MAMMALIAN CELLS (Miller and Calos, eds., Cold Spring Harbor Laboratory, N.Y., 1987), p. 10). Expression of the HCV polypeptide then occurs in cells or animals which are infected with the live recombinant vaccinia virus.

In order to detect whether or not the HCV polypeptide is expressed from the vaccinia vector, BSC1 cells may be infected with the recombinant vector and grown on microscope slides under conditions which allow expression. The cells may then be acetone-fixed, and immunofluorcscence assays performed using serum which is known to contain anti-HCV antibodies to a polypeptide(s) encoded in the region of the HCV genome from which the HCV segment in the recombinant expression vector was derived.

Other systems for expression of eukaryotic or viral genomes include insect cells and vectors suitable for use in these cells. These systems are known in the art, and include, for example, insect expression transfer vectors derived from the baculovirus Autographa californica nuclear polyhedrosis virus (AcNPV), which is a helper-independent, viral expression vector. Expression vectors derived from this system usually use the strong vital polyhedrin gene promoter to drive expression of heterologous genes. Currently the most commonly used transfer vector for introducing foreign genes into AcNPV is pac373 (see PCT WO89/046699 and U.S. Ser. No. 7/456,637). Many other vectors known to those of skill in the an have also been designed for improved expression. These include, for example, pVL985 (which alters the polyhedrin start codon from ATG to ATT, and introduces a BamHI cloning site 32 bp downstream from the ATT; See Luckow and Summers, Virol (1989) 17:31). AcNPV transfer vectors for high level expression of nonfused foreign proteins are described in copending applications PCT WO89/046699 and U.S. Ser. No. 7/456,637. A unique BamHI site is located following position -8 with respect to the translation initiation codon ATG of the polyhedrin gone. There are no cleavage sites for SmaI, PstI, BglII, XbaI or SstI. Good expression of nonfused foreign proteins usually requires foreign genes that ideally have a short leader sequence containing suitable translation initiation signals preceding an ATG start signal. The plasmid also contains the polyhedrin polyadenylation signal and the ampicillin-resistance (amp) gene and origin of replication for selection and propagation in E. coli.

Methods for the introduction of heterologous DNA into the desired site in the baculovirus virus are known in the art. (See Summer and Smith, Texas Agricultural Experiment Station Bulletin No. 1555; Smith et al, Mol Cell Biol (1983) 3:2156-2165; and Luckow and Summers, Virol (1989) 17:31). For example, the heterologous DNA can be inserted into a gene such as the polyhedrin gene by homologous recombination, or into a restriction enzyme site engineered into the desired baculovirus gene. The inserted sequences may be those which encode all or varying segments of the polyprotein, or other offs which encode viral polypeptides. For example, the insert could encode the following numbers of amino acid segments from the polyprotein: amino acids 1-1078; amino acids 332-662; amino acids 406-662; amino acids 156-328, and amino acids 199-328.

The signals for post-translational modifications, such as signal peptide cleavage, proteolytic cleavage, and phosphorylation, appear to be recognized by insect cells. The signals required for secretion and nuclear accumulation also appear to be conserved between the invertebrate cells and vertebrate cells. Examples of the signal sequences from vertebrate cells which are effective in invertebrate cells are known in the art, for example, the human interleukin-2 signal (IL2s) which signals for secretion from the cell, is recognized and properly removed in insect cells.

Transformation may be by any known method for introducing polynucleotides into a host cell, including, for example packaging the polynucleotide in a virus and transducing a host cell with the virus, and by direct uptake of the polynucleotide. The transformation procedure used depends upon the host to be transformed. Bacterial transformation by direct uptake generally employs treatment with calcium or rubidium chloride (Cohen, Proc Nat Acad Sci USA (1972) 69:2110; T. Maniatis et al, "Molecular Cloning; A Laboratory Manual" (Cold Spring Harbor Press, Cold Spring Harbor, N.Y., 1982). Yeast transformation by direct uptake may be carried out using the method of Hinnen et al, Proc Nat Acad Sci USA (1978) 75:1929. Mammalian transformations by direct uptake may be conducted using the calcium phosphate precipitation method of Graham and Van der Eb, Virol (1978) 52:546, or the various known modifications thereof. Other methods for introducing recombinant polynucleotides into cells, particularly into mammalian cells, include dextran-mediated transfection, calcium phosphate mediated transfection, polybrene mediated transfection, protoplast fusion, electroporation, encapsulation of the polynucleotide(s) in liposomes, and direct microinjection of the polynucleotides into nuclei.

Vector construction employs techniques which are known in the art. Site-specific DNA cleavage is performed by treating with suitable restriction enzymes under conditions which generally are specified by the manufacturer of these commercially available enzymes. In general, about 1 .mu.g of plasmid or DNA sequence is cleaved by 1 unit of enzyme in about 20 .mu.L buffer solution by incubation for 1-2 hr at 37.degree. C. After incubation with the restriction enzyme, protein is removed by phenol/chloroform extraction and the DNA recovered by precipitation with ethanol. The cleaved fragments may be separated using polyacrylamide or agarose gel electrophoresis techniques, according to the general procedures described in Meth Enzymol (1980) 65:499-560.

Sticky-ended cleavage fragments may be blunt ended using E. coli DNA polymerase I (Klenow fragment) with the appropriate deoxynucleotide triphosphates (dNTPs) present in the mixture. Treatment with S1 nuclease may also be used, resulting in the hydrolysis of any single stranded DNA portions.

Ligations are carried out under standard buffer and temperature conditions using T4 DNA ligase and ATP; sticky end ligations require less ATP and less ligase than blunt end ligations. When vector fragments are used as part of a ligation mixture, the vector fragment is often treated with bacterial alkaline phosphates (BAP) or calf intestinal alkaline phosphatase to remove the 5'-phosphate, thus preventing religation of the vector. Alternatively, restriction enzyme digestion of unwanted fragments can be used to prevent ligation.

Ligation mixtures are transformed into suitable cloning hosts, such as E. coli, and successful transformants selected using the markers incorporated (e.g., antibiotic resistance), and screened for the correct construction.

Synthetic oligonucleotides may be prepared using an automated oligonucleotide synthesizer as described by Warner, DNA (1984) 3:401. If desired, the synthetic strands may be labeled with .sup.32 P by treatment with polynucleotide kinase in the presence of .sup.32 P-ATP under standard reaction conditions.

DNA sequences, including those isolated from cDNA libraries, may be modified by known techniques, for example by site directed mutagenesis (see e.g., Zoller, Nuc Acids Res (1982) 10:6487). Briefly, the DNA to be modified is packaged into phage as a single stranded sequence, and convened to a double stranded DNA with DNA polymerase, using as a primer a synthetic oligonucleotide complementary to the portion of the DNA to be modified, where the desired modification is included in the primer sequence. The resulting double stranded DNA is transformed into a phage-supporting host bacterium. Cultures of the transformed bacteria which contain copies of each strand of the phage are plated in agar to obtain plaques. Theoretically, 50% of the new plaques contain phage having the mutated sequence, and the remaining 50% have the original sequence. Replicates of the plaques are hybridized to labeled synthetic probe at temperatures and conditions which permit hybridization with the correct strand, but not with the unmodified sequence. The sequences which have been identified by hybridization are recovered and cloned.

DNA libraries may be probed using the procedure of Grunstein and Hogness Proc Nat Acad Sci USA (1975) 73:3961. Briefly, in this procedure the DNA to be probed is immobilized on nitrocellulose filters, denatured, and prehybridized with a buffer containing 0-50% formamide, 0.75M NaCl, 75 mM Na citrate, 0.02% (wt/v) each of bovine serum albumin, polyvinylpyrrolidone, and Ficoll.RTM., 50 mM NaH.sub.2 PO.sub.4 (pH 6.5), 0.1% SDS, and 100 .mu.g/mL carrier denatured DNA. The percentage of formamide in the buffer, as well as the time and temperature conditions of the prehybridization and subsequent hybridization steps depend on the stringency required. Oligomeric probes which require lower stringency conditions are generally used with low percentages of formamide, lower temperatures, and longer hybridization times. Probes containing more than 30 or 40 nucleotides, such as those derived from cDNA or genomic sequences generally employ higher temperatures, e.g., about 40.degree.-42.degree. C., and a high percentage formamide, e.g., 50%. Following prehybridization, 5'-.sup.32 P-labeled oligonucleotide probe is added to the buffer, and the filters are incubated in this mixture under hybridization conditions. After washing, the treated filters are subjected to autoradiography to show the location of the hybridized probe; DNA in corresponding locations on the original agar plates is used as the source of the desired DNA.

For routine vector constructions, ligation mixtures are transformed into E. coli strain HB101 or other suitable hosts, and successful transformants selected by antibiotic resistance or other markers. Plasmids from the transformants are then prepared according to the method of Clewell et al, Proc Nat Acad Sci USA (1969) 62:1159, usually following chloramphenicol amplification (Clewell, J Bacteriol (1972) 110:667). The DNA is isolated and analyzed, usually by restriction enzyme analysis and/or sequencing. Sequencing may be performed by the dideoxy method of Sanger et al, Proc Nat Acad Sci USA (1977) 74:5463, as further described by Messing et al, Nuc Acids Res (1981) 9:309, or by the method of Maxam et al, Meth Enzymol (1980) 65:499. Problems with band compression, which are sometimes observed in GC-rich regions, were overcome by use of T-deazoguanosine according to Barret al, Biotechniques (1986) 4:428.

The enzyme-linked immunosorbent assay (ELISA) can be used to measure either antigen or antibody concentrations. This method depends upon conjugation of an enzyme to either an antigen or an antibody, and uses the bound enzyme activity as a quantitative label. To measure antibody, the known antigen is fixed to a solid phase (e.g., a microtiter dish, plastic cup, dipstick, plastic bead, or the like), incubated with test serum dilutions, washed, incubated with anti-immunoglobulin labeled with an enzyme, and washed again. Enzymes suitable for labeling are known in the art, and include, for example, horseradish peroxidase (HRP). Enzyme activity bound to the solid phase is usually measured by adding a specific substrate, and determining product formation or substrate utilization colorimetrically. The enzyme activity bound is a direct function of the amount of antibody bound.

To measure antigen, a known specific antibody is fixed to the solid phase, the test material containing antigen is added, after an incubation the solid phase is washed, and a second enzyme-labeled antibody is added. After washing, substrate is added, and enzyme activity is measured colorimetrically, and related to antigen concentration.

Proteases of the invention may be assayed for activity by cleaving a substrate which provides detectable cleavage products. As the HCV protease is believed to cleave itself from the genomic polyprotein, one can employ this autocatalytic activity both to assay expression of the protein and determine activity. For example, if the protease is joined to its fusion panner so that the HCV protease N-terminal cleavage signal (Arg-Arg) is included, the expression product will cleave itself into fusion parmer and active HCV protease. One may then assay the products, for example by western blot, to verify that the proteins produced correspond in size to the separate fusion panner and protease proteins. It is presently preferred to employ small peptide p-nitrophenyl esters or methylcoumarins, as cleavage may then be followed by spectrophotometric or fluorescent assays. Following the method described by E. D. Matayoshi et al, Science (1990) 247:231-35, one may attach a fluorescent label to one end of the substrate and a quenching molecule to the other end: cleavage is then determined by measuring the resulting increase in fluorescence. If a suitable enzyme or antigen has been employed as the fusion parmer, the quantity of protein produced may easily be determined. Alternatively, one may exclude the HCV protease N-terminal cleavage signal (preventing self-cleavage) and add a separate cleavage substrate, such as a fragment of the HCV NS3 domain including the native processing signal or a synthetic analog.

In the absence of this protease activity, the HCV polyprotein should remain in its unprocessed form, and thus render the virus noninfectious. Thus, the protease is useful for assaying pharmaceutical agents for control of HCV, as compounds which inhibit the protease activity sufficiently will also inhibit viral infectivity. Such inhibitors may take the form of organic compounds, particularly compounds which mimic the cleavage site of HCV recognized by the protease. Three of the putative cleavage sites of the HCV polyprotein have the following amino acid sequences: ##STR4##

These sites are characterized by the presence of two basic amino acids immediately before the cleavage site, and are similar to the cleavage sites recognized by other flavivirus proteases. Thus, suitable protease inhibitors may be prepared which mimic the basic/basic/small neutral motif of the HCV cleavage sites, but substituting a nonlabile linkage for the peptide bond cleaved in the natural substrate. Suitable inhibitors include peptide trifluoromethyl ketones, peptide boronic acids, peptide .alpha.-ketoesters, peptide difluoroketo compounds, peptide aldehydes, peptide diketones, and the like. For example, the peptide aldehyde N-acetyl-phenylalanylglycinaldehyde is a potent inhibitor of the protease papain. One may conveniently prepare and assay large mixtures of peptides using the methods disclosed in U.S. patent application Ser. No. 7/189,318, filed 2 May 1988 (published as PCT WO89/10931), incorporated herein by reference. This application teaches methods for generating mixtures of peptides up to hexapeptides having all possible amino acid sequences, and further teaches assay methods for identifying those peptides capable of binding to proteases.

Other protease inhibitors may be proteins, particularly antibodies and antibody derivatives. Recombinant expression systems may be used to generate quantities of protease sufficient for production of monoclonal antibodies (MAbs) specific for the protease. Suitable antibodies for protease inhibition will bind to the protease in a manner reducing or eliminating the enzymatic activity, typically by obscuring the active site. Suitable MAbs may be used to generate derivatives, such as Fab fragments, chimeric antibodies, altered antibodies, univalent antibodies, and single domain antibodies, using methods known in the art.

Protease inhibitors are screened using methods of the invention, in general, a substrate is employed which mimics the enzyme's natural substrate, but which provides a quantifiable signal when cleaved. The signal is preferably detectable by colorimetric or fluorometric means: however, other methods such as HPLC or silica gel chromatography, GC-MS, nuclear magnetic resonance, and the like may also be useful. After optimum substrate and enzyme concentrations are determined, a candidate protease inhibitor is added to the reaction mixture at a range of concentrations. The assay conditions ideally should resemble the conditions under which the protease is to be inhibited in vivo, i.e., under physiologic pH, temperature, ionic strength, etc. Suitable inhibitors will exhibit strong protease inhibition at concentrations which do not raise toxic side effects in the subject. Inhibitors which compete for binding to the protease active site may require concentrations equal to or greater than the substrate concentration, while inhibitors capable of binding irreversibly to the protease active site may be added in concentrations on the order of the enzyme concentration.

In a presently preferred embodiment, an inactive protease mutein is employed rather than an active enzyme. It has been found that replacing a critical residue within the active site of a protease (e.g., replacing the active site Ser of a serine protease) does not significantly alter the structure of the enzyme, and thus preserves the binding specificity. The altered enzyme still recognizes and binds to its proper substrate, but fails to effect cleavage. Thus, in one method of the invention an inactivated HCV protease is immobilized, and a mixture of candidate inhibitors added.

Inhibitors that closely mimic the enzyme's preferred recognition sequence will compete more successfully for binding than other candidate inhibitors. The poorly-binding candidates may then be separated, and the identity of the strongly-binding inhibitors determined. For example, HCV protease may be prepared substituting Ala for Ser.sub.221 (FIG. 1 ), providing an enzyme capable of binding the HCV protease substrate, but incapable of cleaving it. The resulting protease mutein is then bound to a solid support, for example Sephadex.RTM. beads, and packed into a column. A mixture of candidate protease inhibitors in solution is then passed through the column and tractions collected. The last fractions to elute will contain the strongest-binding compounds, and provide the preferred protease inhibitor candidates.

Protease inhibitors may be administered by a variety of methods, such as intravenously, orally, intramuscularly, intraperitoneally, bronchially, intranasally, and so forth. The preferred route of administration will depend upon the nature of the inhibitor. Inhibitors prepared as organic compounds may often be administered orally (which is generally preferred) if well absorbed. Protein-based inhibitors (such as most antibody derivatives) must generally be administered by parenteral routes.

The examples presented below are provided as a further guide to the practitioner of ordinary skill in the art, and are not to be construed as limiting the invention in any way.

A genomic library of HCV cDNA was prepared as described in PCT WO 89/04669 and U.S. Ser. No. 7/456,637. This library, ATCC accession no. 40394, has been deposited as set forth below.

(A) The HCV polypeptide encoded within clone 5-1-1 (see Example 1 ) was expressed as a fusion polypeptide with human superoxide dismutase (SOD). This was accomplished by subcloning the clone 5-1-1 eDNA insert into the expression vector pSODCF1 (K. S. Steimer et al, J Virol (1986) 58:9; EPO 138,111) as follows. The SOD/5-1-1 expression vector was transformed into E. coli D1210 cells. These cells, named Cfl/5-1-1 in E. coli, were deposited as set forth below and have an ATCC accession no. of 67967.

First, DNA isolated from pSODCF1 was treated with BamHI and EcoRI, and the following linker was ligated into the linear DNA created by the restriction enzymes:

After cloning, the plasmid containing the insert was isolated.

Plasmid containing the insert was restricted with EcoRI. The HCV eDNA insert in clone 5-1-1 was excised with EcoRI, and ligated into this EcoRI linearized plasmid DNA. The DNA mixture was used to transform E. coli strain D1210 (Sadler et al, Gene (1980) 8:279). Rccombinants with the 5-1-1 eDNA in the correct orientation for expressing the ORF shown in FIG. 1 were identified by restriction mapping and nucleotide sequencing.

Recombinant bacteria from one clone were induced to express the SOD-HCV.sub.5-1-1 polypeptide by growing the bacteria in the presence of IPTG.

Three separate expression vectors, pcflAB, pcflCD, and pcflEF were created by ligating three new linkers, AB, CD, and EF to a BamHI-EcoRI fragment derived by digesting to completion the vector pSODCF1 with EcoRI and BamHI, followed by treatment with alkaline phosphatases. The linkers were created from six oligomers, A, B, C, D, E, and F. Each oligomer was phosphorylated by treatment with kinase in the presence of ATP prior to annealing to its complementary oligomer. The sequences of the synthetic linkers were the following:

Each of the three linkers destroys the original EcoRI site, and creates a new EcoRI site within the linker, but within a different reading frame. Thus, the HCV eDNA EcoRI fragments isolated from the clones, when inserted into the expression vector, were in three different reading frames.

The HCV eDNA fragments in the designated .lambda.gt11 clones were excised by digestion with EcoRI; each fragment was inserted into pcflAB, pcflCD, and pcflEF. These expression constructs were then transformed into D1210 E. coli cells, the transformants cloned, and polypeptides expressed as described in part B below.

(B) Expression products of the indicated HCV cDNAs were tested for antigenicity by direct immunological screening of the colonies, using a modification of the method described in Helfman et al, Proc Nat Acad Sci USA (1983), 80:31. Briefly, the bacteria were plated onto nitrocellulose filters overlaid on ampicillin plates to give approximately 40 colonies per filter. Colonies were replica plated onto nitrocellulose filters, and the replicas were regrown overnight in the presence of 2 mM IPTG and ampicillin. The bacterial colonies were lysed by suspending the nitrocellulose filters for about 15 to 20 min in an atmosphere saturated with CHCl.sub.3 vapor. Each filter then was placed in an individual 100 mm Petri dish containing 10 mL of 50 mM Tris HCl, pH 7.5, 150 mM NaCl, 5 mM MgCl.sub.2, 3% (w/v) BSA, 40 .mu.g/mL lysozyme, and 0.1 .mu.g/mL DNase. The plates were agitated gently for at least 8 hours at room temperature. The filters were rinsed in TBST (50 mM Tris HC1, pH 8.0, 150 mM NaCl, 0.005% Tween.RTM. 20). After incubation, the cell residues were rinsed and incubated for one hour in TBS (TBST without Tween.RTM.) containing 10% sheep serum. The filters were then incubated with pretreated sera in TBS from individuals with NANBH, which included 3 chimpanzees; 8 patients with chronic NANBH whose sera were positive with respect to antibodies to HCV C100-3 polypeptide (also called C100); 8 patients with chronic NANBH whose sera were negative for anti-C100 antibodies; a convalescent patient whose serum was negative for anti-C100 antibodies; and 6 patients with community-acquired NANBH, including one whose sera was strongly positive with respect to anti-C100 antibodies, and one whose sera was marginally positive with respect to anti-C100 antibodies. The sera, diluted in TBS, was protreated by preabsorption with hSOD for at least 30 minutes at 37.degree. C. After incubation, the filters were washed twice for 30 min with TBST. The expressed proteins which bound ,antibodies in the sera were labeled by incubation for 2 hours with .sup.125 I-labeled sheep anti-human antibody. After washing, the filters were washed twice for 30 min with TBST, dried, and autoradiographed.

(A) pBR322-C200

The nucleotide sequences of the HCV cDNAs used below were determined essentially as described above, except that the eDNA excised from these phages were substituted for the cDNA isolated from clone 5-1-1.

Clone C33c was isolated using a hybridization probe having the following sequence:

The sequence of the HCV cDNA in clone C33c is shown in FIG. 8, which also shows the amino acids encoded therein.

Clone 35 was isolated by screening with a synthetic polynucleotide having the sequence:

Approximately 1 in 50,000 clones hybridized with the probe. The polynucleotide and deduced amino acid sequences for C35 are shown in FIG. 7.

Clone C31 is shown in FIG. 6, which also shows the amino acids encoded therein. A C200 cassette was constructed by ligating together a 718 bp fragment obtained by digestion of clone C33c DNA with EcoRI and Hinfl, a 179 bp fragment obtained by digestion of clone C31 DNA with HinfI and BglI, and a 377 bp fragment obtained by digesting clone C35 DNA with BglI and EcoRI. The construct of ligated fragments were inserted into the EcoRI site of pBR322, yielding the plasmid pBR322-C200.

(B) C7f+C20c

Clone 7f was isolated using a probe having the sequence:

The sequence of HCV cDNA in clone 7f and the amino acids encoded therein are shown in FIG. 5.

Clone C20c is isolated using a probe having the following sequence:

The sequence of HCV cDNA in clone C20c, and the amino acids encoded therein are shown in FIG. 2.

Clones 7f and C20c were digested with EcoRI and SfaNI to form 400 bp and 260 bp fragments, respectively. The fragments were then cloned into the EcoRI site of pBR322 to form the vector C7f+C20c, and transformed into HB101 cells.

(C) C300

Clone 8h was isolated using a probe based on the sequence of nucleotides in clone 33c. The nucleotide sequence of the probe was

The sequence of the HCV cDNA in clone 8h, and the amino acids encoded therein, are shown in FIG. 4.

Clone C26d is isolated using a probe having the following sequence:

The sequence and amino acid translation of clone C26d is shown in FIG. 3.

Clones C26d and C33c (see part A above) were transformed into the methylation minus E. coli strain GM48. Clone C26d was digested with EcoRII and DdeI to provide a 100 bp fragment. Clone C33c was digested with EcoRII and EcoRI to provide a 700 bp fragment. Clone C8h was digested with EcoRI and DdeI to provide a 208 bp fragment. These three fragments were then ligated into the EcoRI site of pBR322, and transformed into E. coli HB101, to provide the vector C300.

(D) Preparation of Full Length Clones

A 600 bp fragment was obtained from C7f+C20c by digestion with EcoRI and NaeI, and ligated to a 945 bp NaeI/EcoRI fragment from C300, and the construct inserted into the EcoRI site of pGEM4Z (commercially available from Promega) to form the vector C7fC20cC300.

C7fC20cC300 was digested with NdeI and EcoRI to provide a 892 bp fragment, which was ligated with a 1160 bp fragment obtained by digesting C200 with NdeI and EcoRI. The resulting construct was inserted into the EcoRI site of pBR322 to provide the vector C7fC20cC300C200. Construction of this vector is illustrated schematically in FIG. 9.

(A) cflSODp600

This vector contains a full-length HCV protease coding sequence fused to a functional hSOD leader. The vector C7fC20cC300C200 was cleaved with EcoRI to provide a 2000 bp fragment, which was then ligated into the EcoRI site of plasmid cflCD (Example 2A). The resulting vector encodes amino acids 1-151 of hSOD, and amino acids 946-1630 of HCV (numbered from the beginning of the polyprotein, corresponding to amino acids 1-686 in FIG. 1 ). The vector was labeled cfl SODp600 (sometimes referred to as P600), and was transformed into E. coli D1210 cells. These cells, ATCC accession no. 68275, were deposited as set forth below.

(B) P190

A truncated SOD-protease fusion polynucleotide was prepared by excising a 600 bp EcoRFNaeI fragment from C7f+C20c, blunting the fragment with Klenow fragment, ligating the blunted fragment into the Klenow-blunted EcoRI site of cflEF (Example 2A). This polynucleotide encodes a fusion protein having amino acids 1-151 of hSOD, and amino acids 1-199 of HCV protease.

(C) P300

A longer truncated SOD-protease fusion polynucleotide was prepared by excising an 892 bp EcoRI/NdeI fragment from C7fC20cC300, blunting the fragment with Klenow fragment, ligating the blunted fragment into the Klenow-blunted EcoRI site of cflEF. This polynucleotide encodes a fusion protein having amino acids 1-151 of hSOD, and amino acids 1-299 of HCV protease.

(D) P500

A longer truncated SOD-protease fusion polynucleotide was prepared by excising a 1550 bp EcoRgEcoRI fragment from C7fC20cC300, and ligating the fragment into the EcoRI site of cfl CD to form P500. This polynucleotide encodes a fusion protein having amino acids 1-151 of hSOD, and amino acids 946-1457 of HCV protease (amino acids 1-513 in FIG. 1).

(E) FLAG/Protease Fusion

This vector contains a full-length HCV protease coding sequence fused to the FLAG sequence, Hopp et al. (1988) Biotechnology 6: 1204-1210. PCR was used to produce a HCV protease gene with special restriction ends for cloning ease. Plasmid p500 was digested with EcoRI and NdeI to yield a 900 bp fragment. This fragment and two primers were used in a polymemse chain reaction to introduce a unique BgllI site at amino acid 1009 and a stop codon with a SalI site at amino acid 1262 of the HCV-1, as shown in FIG. 17 of WO 90/11089, published 4 Oct. 1990. The sequence of the primers is as follows:

and

After 30 cycles of PCR, the reaction was digested with BgllI and SalI, and the 710 bp fragment was isolated. This fragment was annealed and ligated to the following duplex: ##STR5## The duplex encodes the FLAG sequence, and initiator methionine, and a 5' NcoI restriction site. The resulting NcoI/SalI fragment was ligated into a derivative of pCF1.

This construct is then transformed into E. coli D1210 cells and expression of the protease is induced by the addition of IPTG.

The FLAG sequence was fused to the HCV protease to facilitate purification. A calcium dependent monoclonal antibody, which binds to the FLAG encoded peptide, is used to purify the fusion protein without harsh eluting conditions.

(A) E. coli D1210 cells were transformed with cfl SODp600 and grown in Luria broth containing 100 .mu.g/mL ampicillin to an OD of 0.3-0.5. IPTG was then added to a concentration of 2 mM, and the cells cultured to a final OD of 0.9 to 1.3. The cells were then lysed, and the lysate analyzed by Western blot using anti-HCV sera, as described in U.S. Ser. No. 7/456,637.

The results indicated the occurrence of cleavage, as no full length product (theoretical Mr 93 kDa) was evident on the gel. Bands corresponding to the hSOD fusion panner and the separate HCV protease appeared at relative molecular weights of about 34, 53, and 66 kDa. The 34 kDa band corresponds to the hSOD partner (about 20 kDa) with a portion of the NS3 domain, while the 53 and 66 kDa bands correspond to HCV protease with varying degrees of (possibly bacterial) processing.

(B) E. coli D1210 cells were transformed with P500 and grown in Luria broth containing 100 .mu.g/mL ampicillin to an OD of 0.3-0.5. IPTG was then added to a concentration of 2 raM, and the cells cultured to a final OD of 0.8 to 1.0. The cells were then lysed, and the lysate analyzed as described above.

The results again indicated the occurrence of cleavage, as no full length product (theoretical Mr 73 kDa) was evident on the gel. Bands corresponding to the hSOD fusion partner and the truncated HCV protease appeared at molecular weights of about 34 and 45 kDa, respectively.

(C) E. coli D1210 cells were transromped with vectors P300 and P190 and grown as described above.

The results from P300 expression indicated the occurrence of cleavage, as no full length product (theoretical Mr 51 kDa) was evident on the gel. A band corresponding to the hSOD fusion partner appeared at a relative molecular weight of about 34. The corresponding HCV protease band was not visible, as this region of the NS3 domain is not recognized by the sera employed to detect the products. However, appearance of the hSOD band at 34 kDa rather than 51 kDa indicates that cleavage occurred.

The P190 expression product appeared only as the full (encoded) length product without cleavage, forming a band at about 40 kDa, which corresponds to the theoretical molecular weight for the uncleaved product. This may indicate that the minimum essential sequence for HCV protease extends to the region between amino acids 199 and 299.

The HCV protease and fragments expressed in Example 5 may be purified as follows:

The bacterial cells in which the polypeptide was expressed are subjected to osmotic shock and mechanical disruption, the insoluble fraction containing the protease is isolated and subjected to differential extraction with an alkaline-NaCl solution, and the polypeptide in the extract purified by chromatography on columns of S-Sepharose.RTM. and Q-Sepharosc.RTM..

The crude extract resulting from osmotic shock and mechanical disruption is prepared by suspending 1 g of the packed cells in 10 mL of a solution containing 0.02 M Tris HCl, pH 7.5, 10 mM EDTA, 20% sucrose, and incubating for 10 minutes on ice. The cells am then pelleted by centrifugation at 4,000.times.g for 15 min at 4.degree. G. After the supernatant is removed, the cell pellets are resuspended in 10 mL of Buffer A1 (0.01M Tris HCl, pH 7.5, 1 mM EDTA, 14 mM .beta.-mercaptoethanol--".beta.ME"), and incubated on ice for 10 minutes. The cells are again pelleted at 4,000.times.g for 15 minutes at 4.degree. G. Mter removal of the clear supernatant (periplasmic fraction I), the cell pellets are resuspended in Buffer A1, incubated on ice for 10 minutes, and again centrifuged at 4,000.times.g for 15 minutes at 4.degree. G. The clear supernatant (periplasmic fraction II) is removed, and the cell pellet resuspended in 5 mL of Buffer T2 (0.02 M Tris HCl, pH 7.5, 14 mM .beta.ME, 1 mM EDTA, 1 mM PMSF). In order to disrupt the cells, the suspension (5 mL) and 7.5 mL of Dyno-mill lead-free acid washed glass beads (0.10-0.15 nun diameter) (available from Glen-Mills, Inc.) are placed in a Falcon tube and vonexed at top speed for two minutes, followed by cooling for at least 2 min on ice. The vonexing-cooling procedure is repeated another four times. After vortexing, the slurry is filtered through a sintered glass funnel using low suction, the glass beads washed twice with Buffer A2, and the filtrate and washes combined.

The insoluble fraction of the crude extract is collected by centrifugation at 20,000.times.g for 15 min at 4.degree. C., washed twice with 10 mL Buffer A2, and resuspended in 5 mL of MILLI-Q water.

A fraction containing the HCV protease is isolated from the insoluble material by adding to the suspension NaOH (2M) and NaCl (2M) to yield a final concentation of 20 mM each, vortexing the mixture for 1 minute, centrifuging it 20,000.times.g for 20 min at 4.degree. C., and retaining the supernatant.

The partially purified protease is then purified by SDS-PAGE. The protease may be identified by western blot, and the band excised from the gel. The protease is then eluted from the band, and analyzed to confirm its amino acid sequence. N-terminal sequences may be analyzed using an automated amino acid sequencer, while C-terminal sequences may be analyzed by automated amino acid sequencing of a series of tryptic fragments.

(A) P650 (SOD/Protease Fusion)

This vector contains HCV sequence, which includes the wild-type full-length HCV protease coding sequence, fused at the 5' end to a SOD coding sequence. Two fragments, a 441 bp EcoRI/BglII fragment from clone 11b and a 1471 bp BglIgEcoRI fragment from expression vector P500, were used to reconstruct a wild-type, full-length HCV protease coding sequence. These two fragments were ligated together with an EcoRI digested pS356 vector to produce an expression cassette. The expression cassette encodes the ADH2/GAPDH hybrid yeast promoter, human SOD, the HCV protease, and a GAPDH transcription terminator. The resulting vector was digested with BamHI and a 4052 bp fragment was isolated. This fragment was ligated to the BamHI digested pAB24 vector to produce p650. p650 expresses a polyprotein containing, from its amino terminal end, amino acids 1-154 of hSOD, an oligopeptide -Asn-Leu-Gly-Ile-Arg-, and amino acids 819 to 1458 of HCV-1, as shown in FIG. 17 of WO 90/11089, published Oct. 4, 1990.

Clone 11b was isolated from the genomic library of HCV eDNA, ATCC accession no. 40394, as described above in Example 3A, using a hybridization probe having the following sequence:

This procedure is also described in EPO Pub. No. 318 216, Example IV.A. 17.

The vector pS3EF, which is a pBR322 derivative, contains the ADH2/GAPDH hybrid yeast promoter upstream of the human superoxide dimutase gene, an adaptor, and a downstream yeast effective transcription terminator. A similar expression vector containing these control elements and the superoxide dismutase gene is described in Cousens et al. (1987) Gene 61:265, and in copending application EPO 196,056, published Oct. 1, 1986. pS3EF, however, differs from that in Cousens et al. in that the heterologous proinsulin gene and the immunoglobulin hinge are deleted, and Gln.sub.154 of SOD is followed by an adaptor sequence which contains an EcoRI site. The sequence of the adaptor is:

The EcoRI site facilitates the insertion of hcterologous sequences. Once inserted into pS3EF, a SOD fusion is expressed which contains an oligopeptide that links SOD to the heterologous sequences. pS3EF is exactly the same as pS356 except that pS356 contains a different adaptor. The sequence of the adaptor is shown below:

pS356, ATCC accession no. 67683, is deposited as set forth below.

Plasmid pAB24 is a yeast shuttle vector, which contains pBR322 sequences, the complete 2 .mu. sequence for DNA replication in yeast (Broach (1981) in: Molecular Biology of the Yeast Saccharomyces, Vol. 1, p. 445, Cold spring Harbor Press.) and the yeast LEU.sup.2d gene derived from plasmid pC1/1, described in EPO Pub. No. 116 201. Plasmid pAB24 was constructed by digesting YEp24 with EcoRI and re-ligating the vector to remove the partial 2 micron sequences. The resulting plasmid, YEp24deltaRI, was linearized with ClaI and ligated with the complete 2 micron plasmid which had been linearized with ClaI. The resulting plasmid, pCBou, was then digested with XbaI, and the 8605 bp vector fragment was gel isolated. This isolated XbaI fragment was ligated with a 4460 bp XbaI fragment containing the LEU.sup.2d gene isolated from pC1/1; the orientation of LEU.sup.2d gene is in the same direction as the URA3 gene.

S. cerevisae, 2150-2-3 (pAB24-GAP-env2), accession no. 20827, is deposited with the American Type Culture Collection as set forth below. The plasmid pAB24-GAP-env2 can be recovered from the yeast cells by known techniques. The GAP-env2 expression cassette can be removed by digesting pAB24-GAP-env2 with BamHI. pAB24 is recovered by religating the vector without the BanfftI insert.

p650 was transformed in S. cerevisae strain JSC310, mata, leu2, ura3-52, prbl-1122, pep4-3, prcl-407, cir.degree. : DM15 (g418 resistance). The transformation is as described by Hinnen et al. (1978) Proc Natl Acad Sci USA 75: 1929. The transformed cells were selected on ura- plates with 8% glucose. The plates were incubated at 30.degree. C. for 4-5 days. The tranformants were further selected on leu- plates with 8% glucose putatively for high numbers of the p650 plasmid. Colonies from the leu- plates were inoculated into leu- medium with 3% glucose. These cultures were shaken at 30.degree. C. for 2 days and then diluted 1/20 into YEPD medium with 2% glucose and shaken for 2 more days at 30.degree. C.

S. cerevisae JSC310 contains DM15 DNA, described in EPO Pub. No. 340 986, published 8 Nov. 1989. This DM15 DNA enhances ADH2 regulated expression of heterologous proteins. pDM15, accession no. 40453, is deposited with the American Type Culture Collection as set forth below.

Mature HCV protease is prepared by cleaving vector C7fC20cC300C200 with EcoRI to obtain a 2Kb coding sequence, and inserting the sequence with the appropriate linkers into a ubiquitin expression vector, such as that described in WO 88/02406, published 7 Apr. 1988, or U.S. Ser. No. 7/390,599 filed 7 Aug. 1989, incorporated herein by reference. Mature HCV protease is recovered upon expression of the vector in suitable hosts, particularly yeast. Specifically, the yeast expression protocol described in Example 8 is used to express a ubiquitin/HCV protease vector.

(A) pGEM.RTM.-3Z/Yellow Fever Leader Vector

Four synthetic DNA fragments were annealed and ligated** together to create a EcoRI/SacI Yellow Fever leader, which was ligated to a EcoRI/SacI digested pGEM.RTM.-3Z vector from Promega.RTM.. The sequence of the four fragments are listed below: ##STR6##

For M-vitro translation of the HCV protease, the new pGEM.RTM.-3Z/Yellow Fever leader vector was digested with Banfill and blunted with Klenow.

(B) PvulI Construct from p6000

A clone p6000 was constructed from sequences available from the genomic library of HCV eDNA, ATCC accession no. 40394. The HCV encoding DNA sequence of p6000 is identical to nucleotide -275 to nucleotide 6372 of FIG. 17 of WO 90/11089, published 4 Oct. 1990. p6000 was digested with PvulI, and from the digest, a 2,864 bp fragment was isolated. This 2,864 bp fragment was ligated to the prepared pGEM.RTM.-3Z/Yellow Fever leader vector fragment, described above.

(A) Transcription

The pGEM.RTM.-3Z/Yellow Fever leader/PvulI vector was linearized with XbaI and transcribed using the materials and protocols from Promega's Riboprobe.RTM. Gemini II Core system.

(B) Translation

The RNA produced by the above protocol was translated using Promega's rabbit reticulocyte lysate, minus methionine, canine pancreatic microsomal membranes, as well as, other necessary materials and instructions from Promega.

The following materials were deposited with the American Type Culture Collection (ATCC), 12301 Parklawn Dr., Rockville, Md.:

The above materials have been deposited with the ATCC under the accession numbers indicated. These deposits will be maintained under the terms of the Budapest Treaty on the International Recognition of the Deposit of Microorganisms for purposes of Patent Procedure. These deposits are provided as a convenience to those of skill in the art, and are not an admission that a deposit is required under 35 U.S.C. .sctn.112. The polynucleotide sequences contained in the deposited materials, as well as the amino acid sequence of the polypeptides encoded thereby, are incorporated herein by reference and are controlling in the event of any conflict with the sequences described herein. A license may be required to make, use or sell the deposited materials, and no such license is granted hereby.