identity (about 40%) to the HCV protein (Muerhoff et al.,. 1995). .... amino acid identity, colons denote ..... Trivandrum, India: Transworld Research Network.
Journal of General Virology (2000), 81, 2183–2188. Printed in Great Britain ..........................................................................................................................................................................................................
SHORT COMMUNICATION
Processing of GB virus B non-structural proteins in cultured cells requires both NS3 protease and NS4A cofactor Andrea Sbardellati,1 Elisa Scarselli,1 Viviana Amati,1† Sabrina Falcinelli,1† Alexander S. Kekule! 2 and Cinzia Traboni1 1 2
Istituto di Ricerche di Biologia Molecolare P. Angeletti (IRBM), Via Pontina Km 30.600, 00040 Pomezia (Roma), Italy Institut fu$ r Medizinische Mikrobiologie, Martin-Luther-Universita$ t Halle-Wittenberg, Magdeburger Str. 6, D-06097 Halle (Saale), Germany
The identification of antivirals and vaccines against hepatitis C virus (HCV) infection is hampered by the lack of convenient animal models. The need to develop surrogate models has recently drawn attention to GB virus B (GBV-B), which produces hepatitis in small primates. In a previous study in vitro, it was shown that GBV-B NS3 protease shares substrate specificity with the HCV enzyme, known to be crucial for virus replication. In this report, GBV-B NS3 activity on GBV-B precursor proteins has been analysed in a cell-based system. It is shown that mature protein products are obtained that are compatible with the cleavage sites proposed on the basis of sequence homology with HCV and that GBV-B NS4A protein is required as a cofactor for optimal enzymatic activity. Experiments in vitro supported by a structural model mapped the region of NS4A that interacts with NS3 and showed that the GBV-B cofactor cannot be substituted for by its HCV analogue.
GB virus B (GBV-B) is a recently discovered flavivirus that causes hepatitis in small New World monkeys of the genus Saguinus (Schlauder et al., 1995 a, b ; Traboni et al., 1999) and shows striking similarities to hepatitis C virus (HCV) (De Francesco, 1999 ; Rice, 1996), which is responsible for a very severe and widespread form of hepatitis (WHO, 1997 ; NIH, 1997). These observations encouraged us to establish a surrogate animal model of HCV infection using tamarins infected by GBV-B or chimeric GBV-B\HCV viruses. The need for such an indirect animal model arises from the extremely Author for correspondence : Cinzia Traboni. Fax j39 06 91 09 32 25. e-mail traboni!irbm.it † Present address : Istituto Dermopatico dell’Immacolata (IDI), IRCCS, Via dei Monti di Creta 104, 00167 Roma, Italy.
0001-6973 # 2000 SGM
narrow host range of HCV, infecting only humans and chimpanzees (Farci et al., 1992 a, b). In order to investigate the extent to which HCV and GBV-B are similar and whether any of their genes can be shuffled to construct chimeric viruses, several studies have been performed recently on the GBV-B genome and its protein products (Bukh et al., 1999 ; Grace et al., 1999 ; Rijnbrand et al., 2000 ; Sbardellati et al., 1999 ; Scarselli et al., 1997 ; Zhong et al., 1999). A key molecule in the life-cycle of HCV is the NS3 protein, which shows different enzymatic activities organized in two domains. The N-terminal domain is responsible for the proteolytic cleavages that lead to the mature non-structural proteins, whereas the C-terminal domain shows helicase and NTPase activity (De Francesco, 1999). A similar molecule is encoded by the GBV-B genome that shows limited amino acid identity (about 40 %) to the HCV protein (Muerhoff et al., 1995). However, a high level of conservation of specific sequence motifs is present in both the protease (Scarselli et al., 1997) and helicase (Muerhoff et al., 1995 ; Zhong et al., 1999) domains. In a previous study, we described the purification of an active recombinant GBV-B NS3 protease domain and demonstrated that this enzyme is able to cleave HCV substrates in the form of synthetic peptides or in vitro-translated proteins. We also showed that the addition of an HCV NS4A activator peptide, Pep4A – , did not modify the activity of the GBV#" $% B enzyme (Scarselli et al., 1997). In order to characterize further the substrate specificity and cofactor requirements of GBV-B NS3 protease under more physiological conditions, we have analysed the full-length NS3 activity by transient expression in a hepatoma cell line. Experiments were performed both by transfecting plasmids that spanned NS3 and its cognate substrates (in cis) and by cotransfecting plasmids that encoded NS3 and its substrates separately (in trans). Experiments were also performed in vitro to confirm the identification and specificity of the cofactor. The constructs used in this study, outlined in Fig. 1 (a), were produced by RT–PCR amplification of specific portions of the GBV-B genome from total RNA of infected tamarin sera, as described previously (Sbardellati et al., 1999). PCR products were cloned in the pCite-2b vector downstream of the T7 CBID
A. Sbardellati and others
Fig. 1. (a) Schematic representation of the GBV-B constructs used in this study. The organization of the virus polyprotein is shown at the top. Vertical lines indicate the putative boundaries between the different non-structural proteins. Numbers indicate the positions of the N-terminal amino acids of individual proteins. For NS5B, the C-terminal residue number is also shown. Numbers refer to amino acids residues of the polyprotein deduced from the GBV-B genome sequence (accession no. U22304). The portions of the GBV-B polyprotein expressed by the different constructs are shown below. (b) GBV-B nonstructural protein processing (NS3 transfected in cis). Vaccinia virus-infected cells were transfected with the construct NS3–3hUTR/pCite and labelled with [35S]methionine. SDS-denatured lysates were immunoprecipitated with antisera to specific proteins : lane 1, NS3 ; 2, NS4B ; 3, NS5A ; 4, NS5B. Positions of the relevant mature immunoprecipitation products are indicated ; putative intermediate precursors are indicated by open arrows. (c) Western blot of proteins produced by transfecting vaccinia virus-infected cells with either NS3/pCite (lane 1) or NS3–3hUTR (lane 2). The filter was incubated with anti-NS5A rabbit antiserum and the colour was developed by using an anti-rabbit IgG alkaline phosphatase-conjugated secondary antibody. The band corresponding to NS5A is indicated. (d ) GBV-B non-structural protein processing (NS3 transfected in trans). Vaccinia virus-infected cells were transfected either with the construct NS4B–3hUTR/pCite (lanes 1–3, 8–13) or with NS4A–4B/pCite (lanes 4–7) and co-transfected with either NS3/pCite (lanes 1, 5, 8 and 11), NS3–4A/pCite (lanes 2, 6, 9 and 12) or HCV-NS3–4A/pCite (lanes 3, 7, 10, and 13) or with no NS3-encoding construct (lane 4). Cells were labelled with
CBIE
GBV-B non-structural protein processing
polymerase promoter. The construct NS3–3hUTR, which spanned the region encoding NS3, NS4A, NS4B, NS5A and NS5B and part of the 3h UTR, was used to transfect Hep3B hepatoma cells infected with recombinant vaccinia virus encoding T7 polymerase as described previously (Failla et al., 1994). Transfected cells were metabolically labelled with [$&S]methionine and proteins were extracted and immunoprecipitated as described previously (Failla et al., 1994). Specific GBV-B proteins were immunoprecipitated by using rabbit antisera raised against recombinant purified GBV-B antigens NS3 (Scarselli et al., 1997), NS5BdeltaC23 (L. Tomei and C. Traboni, unpublished), NS4B–His and NS5A–His (S. Falcinelli, unpublished) and analysed by SDS–PAGE and autoradiography. The results (Fig. 1 b) show that proteins corresponding to products of the size expected on the basis of the proposed cleavage sites of the polyprotein (Scarselli et al., 1997) were immunoprecipitated by the individual antisera against GBV-B NS4B, NS5A and NS5B. The background observed in this experiment is probably due to the use of polyclonal sera, some of which also perform poorly in immunoprecipitation. In order to gain further information about the specificity of the protein products visualized after immunoprecipitation, we subjected the protein extract to Western blotting with the anti-NS5A serum as the source of primary antibody. The results obtained with this technique display a significantly increased signal-tobackground ratio (Fig. 1 c) and confirm the interpretation of the results shown in Fig. 1 (b). Unfortunately, only the antisera to NS5A and NS3 (data not shown) worked as Western blot reagents. The experiments described above indicate that mature protein products were obtained from a construct encoding both NS3 and its cognate substrates upon expression in mammalian cells. They do not, however, provide any indication of the involvement of a protease cofactor, which is known to be required by the NS3 protease of HCV (Failla et al., 1994). Moreover, the detection of a mature GBV-B protein product corresponding to NS4A was hampered by the lack of a suitable specific antibody. To test the hypothesis that GBVB NS3 is activated by its cognate NS4A protein, we compared the activity of the full-length NS3 protein in the presence or absence of NS4A by co-transfecting constructs that encoded substrates with constructs that encoded either NS3 or NS3–4A. A tentative identification of the NS4A protein boundaries was achieved by comparing the complete polyprotein sequences of GBV-B and HCV (using programs of the Genetics Computer Group package, Madison, WI, USA). In spite of a very low
similarity at the level of NS4A protein, the homology was high enough in the sequences flanking NS4A, that is the NS3 helicase region and the boundary between NS4A and NS4B. We constructed two plasmids, NS3\pCite and NS3–4A\pCite (see Fig. 1 a), on the basis of this analysis. Each of the two plasmids was introduced into Hep3B cells infected with recombinant vaccinia virus encoding T7 polymerase in combination with one of the plasmids described below that encoded GBV-B substrates. Specific labelled protein products were immunoprecipitated and analysed as described previously (Failla et al., 1994). NS4B–3hUTR\pCite was chosen to detect the cleavage between NS4B and NS5A and between NS5A and NS5B ; NS4A–4B\pCite was used to detect the cleavage between NS4A and NS4B. The experiment demonstrated that mature NS3 protein was produced when either the NS3\pCite or NS3–4A\pCite construct was transfected (Fig. 1 d, lanes 1 and 2). Unfortunately, the lack of specific reagents for the NS4A moiety did not allow us to discriminate between the NS3 and NS3–4A proteins, which have a very small size difference, and hence to ascertain whether cis cleavage occurred between NS3 and NS4A. However, specific NS5A and NS5B products were clearly detectable when NS4B–3hUTR\pCite was used as a substrate and NS3–4A\pCite as an enzyme source (Fig. 1 d, lanes 9 and 12). They were instead not distinguishable from the background when the NS3\pCite construct was used (Fig. 1 d, lanes 8 and 11), as well as when no protease construct was cotransfected (data not shown). The use of the construct NS4A–4B\pCite as a substrate to visualize the NS4B product gave a more complex result. A specifically immunoprecipitated NS4B product, clearly missing in control experiments without the enzyme source (Fig. 1 d, lane 4), was obtained even in the absence of the NS4A gene in the NS3 construct (Fig. 1 d, lane 5), although it was more abundant with NS3–4A\pCite as an enzyme source (Fig. 1 d, lane 6). This was expected, since an enhancing activity may be exerted by an NS4A protein either encoded in cis (as in the NS3–4A\pCite construct) or supplied in trans, i.e. the NS4A moiety of the NS4A–4B precursor and the free NS4A obtained eventually by cleavage of the precursor. Nonetheless, this part of the experiment is an important internal control that shows that the NS3\pCite (without NS4A) construct produces an enzyme that, by itself or upon activation by NS4A, is able to display proteolytic activity. This implies that the lack of cleavage products when other substrates were used cannot be attributed to a lack of activity of the enzyme produced by the NS3\pCite plasmid.
[35S]methionine and SDS-denatured lysates were immunoprecipitated with antisera to specific proteins : αNS3, αNS4B, αNS5A and αNS5B indicate antisera against GBV-B NS3 (lanes 1–2), HCV NS3 (lane 3), GBV-B NS4B (lanes 4–7), GBV-B NS5A (lanes 8–10) and GBV-B NS5B (lanes 11–13). The antisera as well as the constructs used in co-transfection are indicated below the autoradiograms. Positions of the relevant mature immunoprecipitation products are indicated. Different exposures of the autoradiogram were used depending on the relative immunoprecipitation capability of the individual antisera, anti-HCV NS3 and anti-GBV-B NS3 being much more efficient than the others.
CBIF
A. Sbardellati and others
Fig. 2. Three-dimensional model of GBVB NS3 protease domain (blue ribbon) interacting with the central region of NS4A (yellow). Lateral chains of NS4A residues and of relevant NS3 residues are represented. Below, a sequence alignment is shown between NS4A of HCV (strain BK) and GBV-B. Bars indicate amino acid identity, colons denote conservative replacements.
All of the specific products immunoprecipitated in this cotransfection experiment were also detectable when an HCV NS3–4A\pCite construct was used in place of the GBV-B homologous plasmid (Fig. 1 d, lanes 3, 7, 10, 13). This indicates that the HCV enzyme is capable of processing GBV-B substrates, thus confirming that the two proteases share substrate specificity, as we have shown already for the purified protease domains tested on in vitro-translated HCV substrates (Scarselli et al., 1997). In these transfection experiments, in which a full-length form of the enzyme was produced, NS3 was apparently active only in the presence of the cofactor, at least for the NS4B–5A and NS5A–5B cleavage sites, in contrast to what occurs with the protease domain when tested on HCV substrates (Scarselli et al., 1997) and on the predicted GBV-B NS4A–4B cleavage site peptide (A. Sbardellati, unpublished). This is not surprising, since it is known that the HCV NS3 protease domain shows baseline NS4A-independent activity whereas the cofactor is essential for the full-length enzyme (Gallinari et al., 1998, 1999 ; Hamatake et al., 1996 ; Steinku$ hler et al., 1996 a). In order to identify the core region of NS4A that makes contact with NS3 and is essential for the enzyme enhancement, we constructed a structural homology model of GBV-B protease by using the program Insight II (Molecular Simulation Inc.) (Dayringer et al., 1986) and manual refinement (Amati & Tramontano, 1997). The model is based on the similarity of the primary sequences of the HCV and GBV-B NS3 protease domains (about 30 % identity) and on information about the three-dimensional structure of the HCV NS3 protease domain complexed with an HCV NS4A cofactor peptide (Kim et al., 1996 ; Yan et al., 1998). HCV NS4A participates in the CBIG
formation of the eight-β-strand sheet in the N-terminal part of NS3. Some mutation-sensitive NS4A residues, Val-23, Ile-25, Ile-29 and Leu-31 (Bartenschlager et al., 1995 ; Lin et al., 1995 ; Shimizu et al., 1996), protrude into hydrophobic holes of the NS3 structure, contributing to the formation of the stabilized core of the complex (Kim et al., 1996 ; Yan et al., 1998). By comparing the sequences of HCV and GBV-B NS4A proteins, it appears that the homology in the region of the activator peptide is not greater than in the rest of the molecule (see Fig. 2). Even the four critical hydrophobic residues of the HCV cofactor are not conserved in the GBV-B NS4A protein. However, the alignment before and after that region suggests the correspondence shown in Fig. 2. In this hypothesis, the hydrophobic character of the interaction with the protease would be conserved, although different local alignment obviously cannot be excluded. Correspondingly, in the predicted structure of the GBV-B NS3 protease, the hydrophobicity of the residues forming the pockets in which NS4A residues are positioned is maintained (Amati, 1999). We have then modelled the GBV-B NS4A putative core peptide into the model of NS3 protease to check its structural compatibility (Fig. 2). The peptide is well inserted into the NS3 structure ; non-favourable contacts between amino acids and discontinuities in the structure are absent. In order to test the prediction about the identity of GBVB NS4A experimentally, we synthesized the peptide FGATCVRRCWSITS (GBV-B Pep4A – ) and tested it for activation ## $& of the GBV-B protease. Purified GBV-B protease domain, prepared as described previously (Scarselli et al., 1997), was used to cleave an in vitro-translated NS4A–4B precursor, produced from the construct described in Fig. 1, in the presence
GBV-B non-structural protein processing
Fig. 3. Effect of GBV-B NS4A peptide on the activity of NS3 protease. In vitro-translated GBV-B NS4A–4B precursor was incubated with either the GBV-B or HCV NS3 protease domain in the absence of NS4A cofactor or in the presence of homologous or heterologous NS4A peptide. Protease : no protease (lane 1), GBV-B protease (2–4) or HCV protease (5–7). NS4A peptide : no NS4A peptide (lanes 1, 2 and 5), GBV-B Pep4A22–35 (3 and 6) or HCV Pep4A21–34 (4 and 7). The band corresponding to NS4B is indicated by an arrow.
or absence of the peptide. As a control, we used the HCV NS4A activation peptide Pep4A – (Steinku$ hler et al., 1996 b ; #" $% Tomei et al., 1996), shown already to be inactive on GBV-B protease (Scarselli et al., 1997). The results confirmed that GBV-B Pep4A – works as an activator of the GBV-B ## $& protease (Fig. 3, lanes 2 and 3), whereas the HCV peptide does not exert any effect (Fig. 3, lane 4). The reciprocal test was also performed, by challenging the GBV-B substrate with the HCV protease in the absence of activators and in the presence of either of the two NS4A peptides (Fig. 3, lanes 5–7). The HCV protease was able to cleave the GBV-B NS4A–4B substrate, thus confirming the results obtained in co-transfection experiments, and activity was enhanced by its cognate activator peptide, but it was not activated by the GBV-B NS4A peptide. It appears from these and previous results (Scarselli et al., 1997) that, although the HCV and GBV-B proteases are able to cleave both HCV and GBV-B substrates, their cofactors cannot be interchanged. Moreover, HCV and GBV-B Pep4A were ineffective on the purified NS3 protease domain of hepatitis G virus\GB virus C (Leary et al., 1996 ; Linnen et al., 1996), thus reinforcing the conclusion of virus specificity in the interaction between NS3 and its cofactor among flaviviruses (A. Sbardellati, unpublished). In summary, we have demonstrated that GBV-B NS3 is able to process the GBV-B polyprotein in a cellular system, yielding mature proteins the sizes of which are compatible with the predicted cleavage sites of the precursor. NS4A-dependent activation of NS3 was also detected and a peptide spanning NS4A residues 22–35 was proved to be sufficient for cofactor activity in experiments in vitro. The virus specificity of the cofactor was also shown, since activities of both the HCV and GBV-B enzymes were enhanced only by their cognate NS4A peptide. This finding, not totally unexpected on the basis of the
weak homology between the two peptides, is particularly important and will have to be taken into account in the perspective of the construction of chimeric viruses. During the revision process of this manuscript, a paper was published (Butkiewicz et al., 2000) that reported the characterization of the activity of purified full-length GBV-B NS3 protease and its cofactor requirement by using in vitrotranslated GBV-B substrates. That study led to similar conclusions to our own, although some differences were apparent concerning GBV-B NS3 protease activity at the level of specific cleavage sites. Butkiewicz et al. (2000) did not observe cleavage of the NS4B–5A substrate (with full-length NS3 in vitro), whereas we did (in vivo). They obtained complete cleavage for NS4A–4B only in the presence of the cofactor (in vitro with full-length NS3), whereas we noticed a baseline activity even in the absence of the cofactor in experiments in vitro (with NS3 protease domain). These differences might be explained by the different experimental systems used by Butkiewicz et al. (2000) and by us. With the exception of these differences, the data shown by Butkiewicz et al. (2000) are in overall agreement with our results and confirm that NS4A is required as a cofactor for GBV-B NS3 protease and that this cofactor is virus-specific. We thank Riccardo Cortese, Raffaele De Francesco, Paola Gallinari, Petra Neddermann, Christian Steinku$ hler and Anna Tramontano for helpful discussions and Licia Tomei and Ernst Verschoor for providing HCV-specific reagents and S. oedipus tamarin tissues, respectively. We also thank Janet Clench for critical reading of the manuscript and Manuela Emili for artwork.
References Amati, V. (1999). Le interazioni tra proteine : uno studio teorico e applicazioni
pratiche di modellizzazione. PhD thesis. Tor Vergata University, Rome, Italy. Amati, V. & Tramontano, A. (1997). Modelling of the NS3 proteinase of hepatitis C virus : implications for the prediction of the structure of related enzymes. Current Topics in Peptide and Protein Research 2, 153–158. Bartenschlager, R., Lohmann, V., Wilkinson, T. & Koch, J. O. (1995).
Complex formation between the NS3 serine-type proteinase of the hepatitis C virus and NS4A and its importance for polyprotein maturation. Journal of Virology 69, 7519–7528. Bukh, J., Apgar, C. L. & Yanagi, M. (1999). Toward a surrogate model for hepatitis C virus : an infectious molecular clone of the GB virus-B hepatitis agent. Virology 262, 470–478. Butkiewicz, N., Yao, N., Zhong, W., Wright-Minogue, J., Ingravallo, P., Zhang, R., Durkin, J., Standring, D. N., Baroudy, B. M., Sangar, D. V., Lemon, S. M., Lau, J. Y. & Hong, Z. (2000). Virus-specific cofactor
requirement and chimeric hepatitis C virus\GB virus B nonstructural protein 3. Journal of Virology 74, 4291–4301. Dayringer, H. E., Tramontano, A., Sprang, S. R. & Fletterick, R. J. (1986). Interactive program for visualization and modeling of protein,
nucleic acids and small molecules. Journal of Molecular Graphics 4, 82–87. De Francesco, R. (1999). Molecular virology of the hepatitis C virus.
Journal of Hepatology 31, 47–53. CBIH
A. Sbardellati and others Failla, C., Tomei, L. & De Francesco, R. (1994). Both NS3 and NS4A
are required for proteolytic processing of hepatitis C virus nonstructural proteins. Journal of Virology 68, 3753–3760. Farci, P., Alter, H. J., Govindarajan, S., Wong, D. C., Engle, R., Lesniewski, R. R., Mushahwar, I. K., Desai, S. M., Miller, R. H., Ogata, N. and others (1992 a). Lack of protective immunity against reinfection
with hepatitis C virus. Science 258, 135–140. Farci, P., London, W. T., Wong, D. C., Dawson, G. J., Vallari, D. S., Engle, R. & Purcell, R. H. (1992 b). The natural history of infection with
hepatitis C virus (HCV) in chimpanzees : comparison of serologic responses measured with first- and second-generation assays and relationship to HCV viremia. Journal of Infectious Diseases 165, 1006–1011. Gallinari, P., Brennan, D., Nardi, C., Brunetti, M., Tomei, L., Steinku$ hler, C. & De Francesco, R. (1998). Multiple enzymatic activities
associated with recombinant NS3 protein of hepatitis C virus. Journal of Virology 72, 6758–6769.
Gallinari, P., Paolini, C., Brennan, D., Nardi, C., Steinku$ hler, C. & De Francesco, R. (1999). Modulation of hepatitis C virus NS3 protease and
helicase activities through the interaction with NS4A. Biochemistry 38, 5620–5632. Grace, K., Gartland, M., Karayiannis, P., McGarvey, M. J. & Clarke, B. (1999). The 5h untranslated region of GB virus B shows functional
similarity to the internal ribosome entry site of hepatitis C virus. Journal of General Virology 80, 2337–2341. Hamatake, R., Wang, H. G., Butcher, J. A., Bifano, M., Clark, G., Hernandez, D., Zhang, S., Racela, J., Standring, D. & Colonno, R. (1996). Establishment of an in vitro assay to characterize hepatitis C
virus NS3–4A protease trans-processing activity. Intervirology 39, 249–258. Kim, J. L., Morgenstern, K. A., Lin, C., Fox, T., Dwyer, M. D., Landro, J. A., Chambers, S. P., Markland, W., Lepre, C. A., O’Malley, E. T., Harbeson, S. L., Rice, C. M., Murcko, M. A., Caron, P. R. & Thomson, J. A. (1996). Crystal structure of the hepatitis C virus NS3 protease
domain complexed with a synthetic NS4A cofactor peptide Cell 87, 343–355. Leary, T. P., Muerhoff, A. S., Simons, J. N., Pilot-Matias, T. J., Erker, J. C., Chalmers, M. L., Schlauder, G. G., Dawson, G. J., Desai, S. M. & Mushahwar, I. K. (1996). Sequence and genomic organization of GBV-
C : a novel member of the flaviviridae associated with human non-A-E hepatitis. Journal of Medical Virology 48, 60–67. Lin, C., Thomson, J. A. & Rice, C. M. (1995). A central region in the hepatitis C virus NS4A protein allows formation of an active NS3-NS4A serine proteinase complex in vivo and in vitro. Journal of Virology 69, 4373–4380.
NIH (1997). National Institutes of Health Consensus Development Conference Panel statement : management of hepatitis C. Hepatology 26, 2S–10S. Rice, C. M. (1996). Flaviviridae : the viruses and their replication. In Fields Virology, 3rd edn, pp. 931–959. Edited by B. N. Fields, D. M. Knipe & P. M. Howley. Philadelphia : Lippincott–Raven. Rijnbrand, R., Abell, G. & Lemon, S. M. (2000). Mutational analysis of the GB virus B internal ribosome entry site. Journal of Virology 74, 773–783. Sbardellati, A., Scarselli, E., Tomei, L., Kekule! , A. S. & Traboni, C. (1999). Identification of a novel sequence at the 3h end of the GB virus
B genome. Journal of Virology 73, 10546–10550. Scarselli, E., Urbani, A., Sbardellati, A., Tomei, L., De Francesco, R. & Traboni, C. (1997). GB virus B and hepatitis C virus NS3 serine
proteases share substrate specificity. Journal of Virology 71, 4985–4989. Schlauder, G. G., Dawson, G. J., Simons, J. N., Pilot-Matias, T. J., Gutierrez, R. A., Heynen, C. A., Knigge, M. F., Kurpiewski, G. S., Buijk, S. L., Leary, T. P., Muerhoff, A. S., Desai, S. M. & Mushahwar, I. K. (1995 a). Molecular and serologic analysis in the transmission of the GB
hepatitis agents. Journal of Medical Virology 46, 81–90. Schlauder, G. G., Pilot-Matias, T. J., Gabriel, G. S., Simons, J. N., Muerhoff, A. S., Dawson, G. J. & Mushahwar, I. K. (1995 b). Origin of
GB-hepatitis viruses. Lancet 346, 447–448. Shimizu, Y., Yamaji, K., Masuho, Y., Yokota, T., Inoue, H., Sudo, K., Satoh, S. & Shimotohno, K. (1996). Identification of the sequence on
NS4A required for enhanced cleavage of the NS5A\5B site by hepatitis C virus NS3 protease. Journal of Virology 70, 127–132. Steinku$ hler, C., Tomei, L. & De Francesco, R. (1996 a). In vitro activity of hepatitis C virus protease NS3 purified from recombinant baculovirusinfected Sf9 cells. Journal of Biological Chemistry 271, 6367–6373. Steinku$ hler, C., Urbani, A., Tomei, L., Biasiol, G., Sardana, M., Bianchi, E., Pessi, A. & De Francesco, R. (1996 b). Activity of purified hepatitis
C virus protease NS3 on peptide substrates. Journal of Virology 70, 6694–6700. Tomei, L., Failla, C., Vitale, R. L., Bianchi, E. & De Francesco, R. (1996). A central hydrophobic domain of the hepatitis C virus NS4A
protein is necessary and sufficient for the activation of the NS3 protease. Journal of General Virology 77, 1065–1070. Traboni, C., Amati, V., Ceccacci, A., Falcinelli, S., Sbardellati, A. & Scarselli, S. (1999). The GB viruses : a comprehensive survey. In Recent
Research Developments in Virology, part III. Edited by S. G. Pandalai. Trivandrum, India : Transworld Research Network. WHO (1997). Hepatitis C : global prevalence. Weekly Epidemiological Record 72, 341–344. Yan, Y., Li, Y., Munshi, S., Sardana, V., Cole, J. L., Sardana, M., Steinkuehler, C., Tomei, L., De Francesco, R., Kuo, L. C. & Chen, Z. (1998). Complex of NS3 protease and NS4A peptide of BK strain
Linnen, J., Wages, J., Jr, Zhang-Keck, Z. Y., Fry, K. E., Krawczynski, K. Z., Alter, H., Koonin, E., Gallagher, M., Alter, M., Hadziyannis, S., Karayiannis, P., Fung, K., Nakatsuji, Y., Shih, J. W., Young, L., Piatak, M., Jr, Hoover, C., Fernandez, J., Chen, S., Zou, J. C., Morris, T., Hyams, K. C., Ismay, S., Lifson, J. D., Kim, J. P. and others (1996).
hepatitis C virus : a 2n2 A/ resolution structure in a hexagonal crystal form. Protein Science 7, 837–847.
Molecular cloning and disease association of hepatitis G virus : a transfusion-transmissible agent. Science 271, 505–508.
Zhong, W., Ingravallo, P., Wright-Minogue, J., Skelton, A., Uss, A. S., Chase, R., Yao, N., Lau, J. Y. & Hong, Z. (1999). Nucleoside
Muerhoff, A. S., Leary, T. P., Simons, J. N., Pilot-Matias, T. J., Dawson, G. J., Erker, J. C., Chalmers, M. L., Schlauder, G. G., Desai, S. M. & Mushahwar, I. K. (1995). Genomic organization of GB viruses A and B :
triphosphatase and RNA helicase activities associated with GB virus B nonstructural protein 3. Virology 261, 216–226.
two new members of the Flaviviridae associated with GB agent hepatitis. Journal of Virology 69, 5621–5630.
CBII
Received 9 February 2000 ; Accepted 31 May 2000
