The Hepatitis C Virus Nonstructural Protein 4B Is an Integral Endoplasmic Reticulum Membrane Protein

Virology 284, 70–81 (2001) doi:10.1006/viro.2001.0873, available online at http://www.idealibrary.com on

Thomas Hu¨gle,* Frauke Fehrmann,†,1 Elke Bieck,* Michinori Kohara,‡ Hans-Georg Kra¨usslich,†,2 Charles M. Rice,§,3 Hubert E. Blum,* and Darius Moradpour*,4

*Department of Medicine II, University of Freiburg, D-79106 Freiburg, Germany; †Heinrich-Pette-Institute, D-20251 Hamburg, Germany; ‡Department of Microbiology and Cell Biology, The Tokyo Metropolitan Institute of Medical Science, Tokyo 113, Japan; and §Department of Molecular Microbiology, Washington University School of Medicine, St. Louis, Missouri 63110-1093 Received December 14, 2000; returned to author for revision January 12, 2001; accepted February 14, 2001

The hepatitis C virus (HCV) nonstructural protein 4B (NS4B) is a relatively hydrophobic 27-kDa protein of unknown function. A tetracycline-regulated gene expression system, a novel monoclonal antibody, and in vitro transcription–translation were employed to investigate the subcellular localization and to characterize the membrane association of this viral protein. When expressed individually or in the context of the entire HCV polyprotein, NS4B was localized in the endoplasmic reticulum (ER), as shown by subcellular fractionation, immunofluorescence analyses, and double-label confocal laser scanning microscopy. In this compartment NS4B colocalized with the other HCV nonstructural proteins. Association of NS4B with the ER membrane occurred cotranslationally, presumably via engagement of the signal recognition particle by an internal signal sequence. In membrane extraction and proteinase protection assays NS4B displayed properties of a cytoplasmically oriented integral membrane protein. Taken together, our findings suggest that NS4B is a component of a membrane-associated cytoplasmic HCV replication complex. An efficient replication system will be essential to further define the role of NS4B in the viral life cycle. © 2001 Academic Press Key Words: hepatitis C virus; nonstructural proteins; endoplasmic reticulum; tetracycline-regulated gene expression system; replication complex; in vitro transcription–translation.

INTRODUCTION viral proteases to yield the mature structural and nonstructural proteins (Fig. 1A). The structural proteins core, The hepatitis C virus (HCV) is a major cause of chronic E1, and E2 are released from the polyprotein precursor hepatitis, liver cirrhosis, and hepatocellular carcinoma worldwide (National Institutes of Health, 1997; European by the endoplasmic reticulum (ER) signal peptidase of Association for the Study of the Liver, 1999). HCV has the host cell. The nonstructural (NS) proteins include two been classified in the genus Hepacivirus within the fam- viral proteases, a RNA helicase located in the carboxy- ily Flaviviridae, which includes the classical flaviviruses, terminal region of NS3, the NS4A polypeptide, the NS4B such as yellow fever virus and the animal pestiviruses and NS5A proteins, and a RNA-dependent RNA polymer- (Rice, 1996; van Regenmortel et al., 2000). HCV contains ase represented by NS5B. Cleavage of the polyprotein a single-stranded, positive-sense RNA genome of ap- precursor at the NS2–NS3 junction is accomplished in proximately 9600 nucleotides (nt) in length that encodes cis by an autoprotease encoded by NS2 and the amin- a polyprotein precursor of about 3000 amino acids (aa) oterminal one-third of NS3. A distinct serine protease (see Bartenschlager and Lohmann, 2000, and Reed and located in the amino-terminal one-third of NS3 is respon- Rice, 2000, for recent reviews). The polyprotein precursor sible for the downstream cleavages in the nonstructural is co- and posttranslationally processed by cellular and region. The NS4A polypeptide serves as a cofactor for the NS3 serine protease and is incorporated as an integral component into the enzyme core. NS5A is a serine 1 Present address: Department of Microbiology and Immunology, phosphoprotein of unknown function. Northwestern University Medical School, Morton 6-693, 303 East Chi- NS4B, a relatively hydrophobic protein of 27 kDa, is the cago Avenue, Chicago, IL 60611. 2 Present address: Department of Virology, University of Heidelberg, least characterized HCV protein. The NS4B proteins of Im Neuenheimer Feld 324, D-69120 Heidelberg, Germany. HCV, pestiviruses, and flaviviruses are similar in size, aa 3 Present address: Center for the Study of Hepatitis C, Rockefeller composition, and hydrophobic properties. No function, University, Box 64, 1230 York Ave, New York, NY 10021. however, has yet been ascribed to NS4B in any of these 4 To whom correspondence and reprint requests should be ad- related viruses. Interestingly, it has recently been shown dressed at the Department of Medicine II, University Hospital Freiburg, Hugstetter Strasse 55, D-79106 Freiburg, Germany. Fax: ϩ49 761 270 that hyperphosphorylation of the HCV NS5A protein de- 3610. E-mail: [email protected]. pends on the presence of the nonstructural proteins 3,

4A, and 4B (Koch and Bartenschlager, 1999; Nedder- An immunoblot analysis of UNS4Bcon-4 cells cultured mann et al., 1999). NS4B appears to be directly involved for 24 h in the presence or absence of tetracycline is in the modulation of NS5A hyperphosphorylation, since shown in Fig. 1B. No NS4B reactivity was observed when two deletions in NS4B that do not affect polyprotein the cells were cultured in the presence of tetracycline. By processing and basal NS5A phosphorylation abolish contrast, a single NS4B-specific band of approximately NS5A hyperphosphorylation (Koch and Bartenschlager, 27 kDa was observed in cells cultured in the absence of 1999). This observation, together with the demonstration tetracycline. This illustrates the tight regulation of NS4B of a physical interaction between NS4A (and, by conse- expression in this cell line. quence, also NS3) and a NS4B–5A cleavage substrate Interestingly, high-level expression of NS4B affected (Lin et al., 1997), suggests that NS4B associates with the viability and growth of UNS4Bcon-4 cells seeded at low other nonstructural proteins to form a replication com- confluence. Colony formation efficiency assays did not plex. In this context, it has recently been shown in the yield any colonies when these cells were seeded at low related flavivirus Kunjin that NS4B is essential for viral confluence and subsequently cultured in the absence of replication and functions only in cis (Khromykh et al., tetracycline. By contrast, the expected number of colo- 2000). Intriguingly, HCV NS4B has recently been reported nies formed when these cells were cultured in the pres- to transform NIH3T3 cells in cooperation with the Ha-ras ence of 1 ␮g/ml tetracycline (i.e., when NS4B expression oncogene (Park et al., 2000). was completely repressed) or when expression was ti- As a first step toward determining the function of HCV trated to low levels in the presence of 0.01 ␮g/ml tetra- NS4B, we investigated its subcellular localization and cycline (data not shown). Consistent with the dose-de- characterized its membrane association. For this pur- pendent cytotoxic effect, colony formation was not af- pose, continuous human cell lines inducibly expressing fected in UNS4Bcon-52 cells, which produce lower NS4B either individually or in the context of the entire amounts of NS4B. HCV polyprotein were established using a tetracycline- The subcellular localization of NS4B was investigated regulated gene expression system, and monoclonal an- by indirect immunofluorescence microscopy. An immu- tibodies (mAbs) were generated by hybridoma fusion. nofluorescence analysis of UNS4Bcon-4 cells stained Subcellular fractionation and double-label immunofluo- with the mAb 4b-52 is shown in Fig. 1C. No immunore- rescence analyses revealed that NS4B was localized in activity was detected when the cells were cultured in the the ER, where it colocalized with the other HCV nonstruc- presence of tetracycline (0 h). NS4B expression became tural proteins. In vitro transcription–translation experi- clearly detectable 6 h following tetracycline withdrawal ments performed in the presence or absence of micro- and increased to reach a steady-state level after 24 h. At somal membranes revealed cotranslational targeting to this time point, mAb 4b-52 revealed a reticular staining the ER membrane, where NS4B displayed properties of a pattern, which surrounded the nucleus, extended cytoplasmically oriented integral membrane protein. through the cytoplasm, and appeared to include the nuclear membrane. No nuclear or plasma membrane RESULTS staining was observed. An analogous staining pattern was found in UNS4Bcon-52 cells (data not illustrated). In NS4B mAb and tetracycline-regulated cell lines UHCVcon-57 cells, which inducibly express NS4B in the

A NS4B-specific mAb of the IgG1 isotype, designated context of the entire HCV polyprotein, overall expression 4b-52, was obtained by hybridoma fusion after immuni- levels were lower, but the cytoplasmic reticular staining zation of BALB/c mice with recombinant NS4B protein was very similar to that observed in UNS4Bcon-4 and produced in Escherichia coli. This mAb was found to UNS4Bcon-52 cells (Fig. 3, right panel). Taken together, function well in ELISA and indirect immunofluorescence the NS4B staining pattern observed in these cell lines applications, but not in immunoblot analyses, suggesting was typical of a membrane-associated protein and that it recognizes a conformational epitope. highly suggestive of a localization of the protein in the A tetracycline-regulated gene expression system was ER. used to generate cell lines inducibly expressing NS4B. Subcellular fractionation experiments were performed Screening of 60 antibiotic double-resistant clones resulting to confirm the membrane association of NS4B suggested from the transfection of the tetracycline-controlled transac- by the immunofluorescence data. For this purpose, cells tivator (tTA)-expressing founder cell line UTA-6 with the were lysed in a hypotonic buffer and separated into construct pUHDNS4Bcon allowed the isolation of two nuclear, mitochondrial, microsomal, and cytosolic frac- tightly regulated cell lines, designated UNS4Bcon-4 (high- tions by differential centrifugation (Fig. 1D). When equal level expression) and UNS4Bcon-52 (medium-level expres- amounts of protein from each fraction were analyzed by sion). These cell lines were maintained in continuous cul- immunoblot, NS4B was detected only in the membrane- ture for more than 12 months and over 50 passages with containing fractions, i.e., the nuclear pellet [which con- stable characteristics and without loss of the tightly regu- tains the outer nuclear membrane (contiguous with the lated NS4B expression. ER) and membranes adsorbed to the nucleus], the mito- 72 HU¨ GLE ET AL.

FIG. 1. Tetracycline-regulated expression and subcellular localization of HCV NS4B. (A) Expression cassettes present in UNS4Bcon and UHCVcon cells. Included HCV aa positions are indicated below. The genetic organization and polyprotein processing of HCV are schematically illustrated at top. Diamonds denote cleavages of the HCV polyprotein precursor by the ER signal peptidase and arrows indicate cleavages by the NS2-3 and NS3 proteases. Asterisks indicate glycosylation of the envelope proteins. (B) Tightly regulated expression of NS4B in UNS4Bcon cells. UNS4Bcon-4 cells were cultured for 24 h in the presence (ϩ) or absence (Ϫ) of tetracycline (tet) and subsequently processed for Western blot analysis. Per lane, approximately 60 ␮g protein was separated by 12% SDS–PAGE and analyzed by immunoblot using a pool of sera from 10 patients with chronic hepatitis C at a dilution of 1:500. Molecular weight standards in kDa are indicated on the left. (C) Subcellular localization of NS4B. UNS4Bcon-4 cells were cultured in the presence (0 h) or for 6, 12, and 24 h in the absence of tetracycline and subsequently processed for indirect immunofluorescence microscopy using mAb 4b-52 as described under Materials and Methods. (D) Subcellular fractionation. UNS4Bcon-4 and UGFP-9.22 cells were cultured for 48 h in the absence of tetracycline and subsequently subjected to subcellular fractionation as described under Materials and Methods. Per lane, 60 ␮g protein was separated by 15% SDS–PAGE and analyzed by immunoblot using the polyclonal rabbit antiserum WU151 against NS4B or the mAb JL-8 against GFP, respectively. nuc, nuclear; mit, mitochondrial; mic, microsomal; cyt, cytosolic fraction. chondrial pellet [which contains ER membranes that in ization of NS4B in the ER. Double-label confocal laser U-2 OS cells are often wrapped around mitochondria scanning microscopy with antibodies to cellular marker (Moradpour et al., 1996)], and the microsomal pellet. proteins was performed to explore the subcellular local- However, NS4B was not detected in the cytosolic frac- ization of NS4B in more detail. As shown in Fig. 2, NS4B tion, which contains soluble proteins. By contrast, green colocalized perfectly with protein disulfide isomerase fluorescent protein (GFP), a protein that is diffusely dis- (PDI), a marker specific for the ER. The NS4B staining tributed within cells, was found in all fractions (Fig. 1D, pattern observed in these cells was different, however, lower panel). Taken together, these results clearly dem- from that revealed by antibodies directed against ERGIC- onstrate the membrane association of NS4B. 53, a marker of the ER-to-Golgi intermediate compartment, and mannosidase (Man) II, a marker of the Golgi NS4B is localized in the ER apparatus. The possibility of an association of a minor The staining pattern and the subcellular fractionation proportion of NS4B with these related compartments can data shown in Fig. 1 were highly suggestive of a local- not be excluded by this technique. Taken together, how- HCV NS4B PROTEIN 73

FIG. 2. NS4B is localized in the ER. UNS4Bcon-4 cells were cultured for 24 h in the absence of tetracycline and subsequently processed for double-immunolabeling with antibodies against the indicated cellular marker proteins and mAb 4b-52 or, in the case of double-labeling with the mAb G1/93 against ERGIC-53, with the polyclonal antiserum WU151 against NS4B. Bound primary antibodies were revealed with FITC- and TXR-conjugated secondary antibodies. Slides were analyzed by confocal laser scanning microscopy as described under Materials and Methods. Horizontal sections taken through the center of the nuclei are shown. Images recorded in red (TXR) and green (FITC) channels are presented separately on the left and on the right, respectively, and merged images are shown in the middle. ever, the results clearly demonstrate that NS4B is local- mous replication of subgenomic HCV RNA (Lohmann et ized predominantly in the ER. al., 1999). To further explore the role of NS4B we com- As found previously by us (Moradpour et al., 1996, pared the localization of NS4B with that of the other HCV 1998a,b) and others (Precious et al., 1995) using the nonstructural proteins. In a first set of experiments (Fig. tetracycline-regulated gene expression system, there 3, left panels), UNS4Bcon-4 cells cultured in the absence was some heterogeneity in expression levels among of tetracycline were transiently transfected with cyto- individual cells of a given monoclonal cell line. This megalovirus promoter-driven expression constructs cod- inherent feature of the expression system explains the ing for the HCV nonstructural proteins 3–4A, 5A, and 5B. observation that not all cells stained with antibodies Subsequently, cells were processed for double-label im- against marker proteins stained for NS4B in the double- munofluorescence with a polyclonal rabbit antiserum immunolabeling experiments. against NS4B and murine mAbs against NS3, NS4A, NS5A, or NS5B. In a second set of experiments (Fig. 3, NS4B colocalizes with the other HCV nonstructural right panels), double-immunolabeling was performed in proteins UHCVcon-57 cells, which inducibly express the entire The nonstructural proteins NS3 through NS5B repre- open reading frame derived from a functional HCV con- sent an independent module sufficient for the autono- sensus cDNA. As shown in Fig. 3, NS4B colocalized with 74 HU¨ GLE ET AL.

FIG. 3. NS4B colocalizes with the other HCV nonstructural proteins. In the left panels, UNS4Bcon-4 cells cultured in the absence of tetracycline were transiently transfected with the CMV promotor-driven expression construct pCMVNS3–4A (NS3, NS4A), pCMVNS5Acon (NS5A), or pCMVNS5Bcon (NS5B). Twenty-four hours later cells were processed for double-immunolabeling as described under Materials and Methods. In the right panels, UHCVcon-57 cells were cultured for 24 h in the absence of tetracycline, followed by double-immunolabeling. The polyclonal rabbit antiserum WU151 against NS4B and murine mAbs 1B6, 8N, 11H, and 5B-12B7 against NS3, NS4A, NS5A, and NS5B, respectively, were used in both panels. Bound primary antibodies were revealed with FITC- and TXR-conjugated secondary antibodies. Slides were viewed with filter sets for TXR and FITC. the other nonstructural proteins under both experimental distribution of these proteins was not influenced by the conditions. These observations do not necessarily imply, coexpression of NS4B (data not shown). NS4B, therefore, but raise the possibility, that NS4B is a component of the even if interacting in some way with NS3 and NS5B in the HCV replication complex. Furthermore, these experi- context of a replication complex, was unable to direct ments demonstrate that the subcellular localization of these proteins to the ER. NS4B is not influenced by the coexpression of other HCV proteins. NS4B per se, therefore, contains the signals Association of NS4B with microsomal membranes in necessary for ER targeting. vitro We have previously shown that NS3 expressed alone was diffusely distributed in the cytoplasm and nucleus In vitro transcription–translation and membrane sedi- while coexpression of NS4A in cis or in trans directed mentation analyses were performed to further character- NS3 to the ER membrane (Wo¨lk et al., 2000). Moreover, a ize the membrane association of NS4B. To this end, carboxy-terminal truncated NS5B protein was recently NS4B was translated in a coupled rabbit reticulocyte found to be redistributed from the ER to the nucleus, lysate system in the presence or absence of microsomal where it accumulated in nucleoli (Moradpour et al., un- membranes. Subsequently, membrane-associated mate- published data). Based on these observations, we per- rial was separated by centrifugation and NS4B was formed cotransfection experiments of NS4B with NS3 or quantified in both fractions. As a control, MIA2, which carboxy-terminal truncated NS5B constructs, to examine encodes the Gag, Gag-PR, and Gag-PR-Pol polyproteins if interactions of NS4B with these nonstructural proteins of the murine intracisternal A-type particle (IAP) MIA14, may direct them to the ER. However, the subcellular was used. The IAP Gag protein contains an amino- HCV NS4B PROTEIN 75

eukaryotic cells, ER transport of secreted and integral membrane proteins is generally mediated by a signal sequence that is recognized by the signal recognition particle (SRP). The SRP interacts with the signal sequence of nascent polypeptide chains during translation and directs the translation complex to the ER membrane. By consequence, SRP-mediated ER transport occurs only cotranslationally, while SRP-independent transport and binding of hydrophobic proteins should also occur posttranslationally. To distinguish between these two possi- bilities, therefore, microsomal membranes were added FIG. 4. Association of NS4B with microsomal membranes in vitro. In to the reaction either during or after completion of in vitro vitro transcription–translation reactions of pCMVNS4Bcon (left panel) and pTM1-MIA2 (right panel) were performed in the presence or ab- transcription–translation. Puromycin was added to the sence of microsomal membranes as indicated. Subsequently, mem- reaction mixture in the posttranslational setting to stop brane sedimentation analyses were performed as described under translation and to ascertain that polypeptides were re- Materials and Methods. Supernatant (S) and pellet fractions (P) were leased from ribosomes. As shown in Fig. 5, 47% of NS4B applied in equivalent amounts and separated by 12% SDS–PAGE. was found in the pellet fraction in the cotranslational [35S]Methionine-labeled translation products were detected by autoradiography. Quantitation was performed as described under Materials reaction. By contrast, only 9% was found in the pellet and Methods and values expressed in % are given at the top. Molecular when the membranes were added posttranslationally. weight standards in kDa are indicated at left. This observation was confirmed in three independent experiments with mean values of 40% of the NS4B protein pelleted in the cotranslational and 8% in the post- terminal signal sequence that directs these polyproteins translational setting. MIA2 Gag polyproteins, which are specifically to the ER membrane in transfected cells directed to the ER via the SRP (Fehrmann et al., manu- (Welker et al., 1997) and in vitro (Fehrmann et al., manuscript in preparation), served as a control for cotransla- script in preparation). The results of a typical experiment tional ER membrane targeting (Fig. 5, middle panel). The are shown in Fig. 4. When NS4B was translated in the MIA4 construct, in which the amino-terminal signal se- absence of microsomal membranes only 26% was subsequently found in the pellet. By contrast, 68% of the NS4B protein sedimented when translation was performed in the presence of microsomal membranes. These results demonstrate that membrane association of NS4B occurs also in vitro. Similarly, MIA2 associated efficiently with microsomal membranes (8 vs 79% pelleted in reactions performed in the absence or presence of microsomal membranes, respectively). Relatively more NS4B compared to MIA2 was found in the pellet fraction of in vitro transcription–translation reactions performed in the absence of microsomal membranes. This is likely a consequence of some aggregation due to nonspecific hydrophobic interactions. This tendency of NS4B could be reduced by the addition of 0.01% digitonin to the translation reactions (Fig. 5 and data not shown). A minor, slightly faster migrating band was observed in the FIG. 5. Association of NS4B with the ER membrane occurs cotrans- NS4B translations irrespective of the presence or ab- lationally. In vitro transcription–translation reactions of pCMVNS4Bcon (top panel), pTM1-MIA2 (middle panel), and pTM1-MIA4 (bottom panel) sence of microsomal membranes (Fig. 4). This smaller were performed in the presence (co) or absence of microsomal mem- product, therefore, is likely a consequence of leaky scan- branes (post and Ϫ). Digitonin at a concentration of 0.01% was present ning and internal translation initiation at two ATGs lo- in all translation reactions to reduce nonspecific aggregation and cated 10 and 11 codons downstream from the engi- sedimentation of NS4B. After 1 h, translation was stopped by the neered translation initiation codon. addition of puromycin to 1.25 mM and microsomal membranes were added to a set of the reactions without membranes (post) and incubated for one additional hour. Subsequently, membrane sedimentation Membrane association of NS4B occurs analyses were performed as described under Materials and Methods. cotranslationally Supernatant (S) and pellet fractions (P) were applied in equivalent amounts and separated by 12% SDS–PAGE. [35S]Methionine-labeled To gain insight into the mechanism of membrane as- translation products were detected by autoradiography. Quantitation sociation of NS4B we examined whether membrane tar- was performed as described under Materials and Methods and values geting of NS4B occurred co- or posttranslationally. In expressed in % are given at the bottom of each panel. 76 HU¨ GLE ET AL.

control, membranes were disrupted with 0.5% Triton X-100. Under these conditions, 78% of the protein could be extracted into the supernatant fraction.

NS4B is cytoplasmically oriented in the ER membrane Proteinase protection experiments were performed to determine the orientation of NS4B in the ER membrane. As shown in Fig. 7, treatment of in vitro transcription– translation reactions performed in the presence of microsomal membranes with proteinase K resulted in the complete disappearance of the NS4B signal. Thus, the protease sensitivity of NS4B indicates that the majority of FIG. 6. NS4B is an integral membrane protein. In vitro transcription– translation reactions of pCMVNS4Bcon were performed in the pres- the protein is localized on the cytoplasmic side of the ER ence of microsomal membranes. Subsequently, reaction mixtures were membrane. In these experiments, preprolactin (ppl) was centrifuged for 15 min at 12,000 g to sediment microsomal membranes used as a control for the integrity of microsomal mem- containing associated NS4B protein. The supernatants were removed branes. Ppl is directed to the ER membrane by interac- and the pellets were resuspended in NTE buffer, 1 M NaCl, 100 mM tion of its signal sequence with the SRP. Signal sequence sodium carbonate, pH 11.5, 4 M urea, or 0.5% Triton X-100 and incubated for 20 min at 4°C. Subsequently, membrane sedimentation anal- cleavage is performed by the signal peptidase located at yses were performed as described under Materials and Methods. the luminal side of the ER membrane, followed by re- Supernatant (S) and pellet fractions (P) were applied in equivalent lease of prolactin (pl) into the ER lumen. As shown in Fig. 35 amounts and separated by 12% SDS–PAGE. [ S]Methionine-labeled 7, pl was protected from proteinase K digestion in the translation products were detected by autoradiography. Quantitation presence of microsomal membranes. It became acces- was performed as described under Materials and Methods and values expressed in % are given at the bottom and depicted as bars. White sible to proteolysis only after disruption of the mem- bars represent supernatant, while dark gray bars represent pellet branes by detergent treatment. Because potentially ex- fractions. isting transmembrane or luminal fragments of NS4B may be very small and may lack methionine residues, we repeated these experiments using [14C]leucine instead of quence has been replaced by the membrane targeting [35S]methionine (12.3% of NS4B aa are leucine vs 2.3% signal of the Src protein, represented a control for post- that are methionine) as a radioactive label, followed by translational ER targeting (Fig. 5, bottom panel). Mem- analyses on high-resolution 10 to 16.5% tricine gels. brane association of this construct is mediated by my- However, even under these more sensitive experimental ristoylation of the engineered Src protein domain. Taken conditions we observed no protected fragments (data together, these results demonstrate that NS4B is co- not shown). translationally targeted to the ER membrane.

NS4B is an integral membrane protein Membrane extraction experiments were performed to characterize the nature of the association of NS4B with the ER membrane. To this end, NS4B was translated in vitro in the presence of microsomal membranes and the pellet fraction was subsequently subjected to differential extraction methods. High-salt extraction (1 M NaCl) is expected to shield charges and weaken ionic interactions that bind peripheral proteins to membranes either directly or indirectly through other membrane proteins (Kretzschmar et al., 1996). Treatment with 100 mM sodium carbonate, pH 11.5, should release peripheral proteins by transforming microsomes into membrane sheets (Fujiki et al., 1982). However, as shown in Fig. 6, NS4B FIG. 7. NS4B is cytoplasmically oriented on the ER membrane. In remained associated with microsomal membranes un- vitro transcription–translation reactions of pCMVNS4Bcon and pTM1- der both conditions. Sixty percent of the protein re- ppl were performed in the presence of microsomal membranes, fol- ␮ mained membrane-associated even under strongly de- lowed by digestion with 50 g/ml proteinase K (PK) for1hat0°Cinthe absence or presence of 0.1% Triton X-100 as indicated at the top. naturing conditions, i.e., treatment with 4 M urea. NS4B, Aliquots of each reaction were analyzed by 12% SDS–PAGE and visu- therefore, is tightly associated with the ER membrane alized by autoradiography. Molecular weight standards in kDa are and behaves as an integral membrane protein. As a indicated on the left. ppl, preprolactin; pl, prolactin. HCV NS4B PROTEIN 77

DISCUSSION teins were expressed individually (Moradpour et al., unpublished data). This is in contrast to NS3, which is NS4B is the least understood of all HCV proteins. No directed to the ER via interaction with its cofactor function has yet been ascribed to this protein in any of polypeptide NS4A (Wo¨lk et al., 2000). the related Flaviviridae family members. In the present In the ER, NS4B colocalized with the other HCV non- initial characterization, therefore, we investigated the structural proteins. This observation suggests that NS4B subcellular localization and examined the membrane is a component of the HCV replication complex. This association of HCV NS4B. A novel mAb, designated 4b- notion is supported by recent data indicating that NS4B, 52, and cell lines inducibly expressing NS4B were gen- together with NS3 and NS4A, is implicated in the mod- erated for this purpose. Investigation of the viral life cycle ulation of NS5A hyperphosphorylation (Koch and Barten- has been limited by the lack of reproducible cell culture schlager, 1999; Neddermann et al., 1999) and by the systems permissive for HCV infection and replication. In demonstration of a physical interaction between NS4A this context, the tetracycline-regulated cell lines de- and a NS4B–5A cleavage substrate (Lin et al., 1997). scribed here represent a well-defined and highly repro- Formation of a membrane-associated replication com- ducible model system to study structural and functional plex is a characteristic feature of positive-strand RNA properties of NS4B. viruses and has been characterized in some detail, e.g., Interestingly, inducible expression of NS4B in tetracy- in the case of poliovirus (Bienz et al., 1992; Egger et al., cline-regulated cell lines revealed a dose-dependent cy- 1996, 2000; Suhy et al., 2000) and the flavivirus Kunjin totoxic effect. This effect was seen only at high expres- (Westaway et al., 1997b; Mackenzie et al., 1999; Khro- sion levels and under low density seeding conditions, mykh et al., 2000). The mechanisms of membrane asso- when cells are particularly vulnerable. We previously ciation and the protein–protein interactions involved in made similar observations in U-2 OS-derived cell lines the formation of the HCV replication complex, however, inducibly expressing the entire HCV polyprotein (Morad- are poorly understood. A subtly regulated and highly pour et al., 1998b) or the structural region (Moradpour et complex scenario is likely, not the least in view of recent al., 1998c). However, under the same experimental con- data on membrane targeting of the HCV NS3–4A com- ditions no cytotoxicity was observed for the core protein plex by the NS4A polypeptide (Wo¨lk et al., 2000) and the (Moradpour et al., 1996) or the nonstructural proteins 4A, conformational changes of this complex predicted for 5A, and 5B (Moradpour et al., unpublished data). NS4B, cis- and trans-processing events (Yao et al., 1999). therefore, appears to be more toxic compared to the The membrane association of NS4B was further ex- other nonstructural HCV proteins. Intriguingly, recent plored by in vitro transcription–translation experiments work has shown that mutations in NS4B can attenuate performed in the presence or absence of microsomal pestivirus cytopathogenicity (Qu and Rice, submitted for membranes isolated from canine pancreas. An earlier publication). Clearly, the relevance of these findings for study suggested that most HCV nonstructural proteins the viral life cycle and the pathogenesis of hepatitis C are more or less membrane-associated (Hijikata et al., needs to be investigated further. 1993). NS4B, however, was not specifically addressed in Immunofluorescence analyses using the mAb 4b-52 that study. Our results demonstrate a clear membrane and subcellular fractionation experiments suggested association also in vitro. About 40–70% of NS4B protein that NS4B is associated with the ER. This was confirmed sedimented after in vitro transcription–translation in the by double-label confocal laser scanning microscopy, in presence of microsomal membranes, while only 5–30% which NS4B was found to colocalize with the ER-resident was pelleted, presumably as a result of nonspecific hy- protein PDI. Preliminary data on the subcellular localiza- drophobic interactions, when microsomal membranes tion has thus far been reported only for a GFP–NS4B were omitted from the reaction. fusion protein (Kim et al., 1999) and for a FLAG-tagged In mammalian cells, membrane and secreted proteins NS4B protein (Park et al., 2000). In both reports, the are directed first to the ER membrane. Protein transport engineered proteins were localized in the cytoplasm, but to the ER is most commonly mediated by signal se- the subcellular compartment targeted by these proteins quences that are cotranslationally bound by the SRP and was not further explored. Interestingly, NS4B of the re- targeted to the docking protein and the Sec61p complex lated flavivirus Kunjin has been reported to translocate to on the ER membrane (Walter and Johnson, 1994; Matlack the nucleus during viral replication (Westaway et al., et al., 1998). This complex forms a channel through which 1997a). In our studies, however, we never observed a polypeptides are translocated and released into the ER nuclear staining of HCV NS4B. Localization of NS4B in lumen as soluble proteins or inserted into the membrane the ER was independent of the coexpression of other as transmembrane proteins (Schatz and Dobberstein, HCV proteins, indicating the presence of endogenous 1996; Matlack et al., 1998). Posttranslational membrane signals for ER targeting and membrane anchorage. Sim- targeting is common in yeast (Deshaies et al., 1991; ilarly, we recently found that NS4A, NS5A, and NS5B Panzner et al., 1995) and bacteria (Schatz and Beckwith, were targeted to the ER when these nonstructural pro- 1990; Wickner and Leonard, 1996), but few examples 78 HU¨ GLE ET AL. have been described for mammalian cells (Anderson et 3011con (Kolykhalov et al., 1997) using sense primer al., 1983; Mu¨ller and Zimmermann, 1987; Kutay et al., 5Ј-GAGAATTCACCATGTCTCAGCACTTACCGTACATCGA- 1995). Posttranslational membrane targeting does not GCAAGGG-3Ј (EcoRI site underlined, engineered trans- involve the SRP. The cotranslational membrane associ- lation initiation codon in boldface) and reverse primer ation of NS4B, therefore, suggests an SRP-dependent 5Ј-GCTGTCTAGATTAGCATGGAGTGGTACACTCCGAGC- mechanism of membrane targeting. Since NS4B does TTATCC-3Ј (XbaI site underlined, engineered ochre stop not possess a classical amino-terminal signal sequence, codon in boldface). The amplification product was we propose that one or more of the internal hydrophobic cloned into the EcoRI-XbaI sites of pUHD10-3 (Gossen sequences function as signal anchor sequences. The and Bujard, 1992) and pcDNA3.1 (Invitrogen, San Diego, hydrophobicity profiles of NS4B are similar among flavi- CA) to yield the expression constructs pUHDNS4Bcon viruses, with a conserved hydrophobic plateau in the and pCMVNS4Bcon, respectively. pUHDNS4Bcon allows center and two short hydrophobic domains at the car- expression of NS4B under the transcriptional control of a boxy terminus. In this context, we found that deletion of tTA-dependent promoter (Fig. 1A). pCMVNS4Bcon allows the carboxy-terminal 96 aa, including two relatively con- both eukaryotic expression from a CMV promoter and in served highly hydrophobic domains, did not alter the vitro transcription from a T7 RNA polymerase promotor. subcellular distribution of NS4B (data not illustrated). It is pCMVNS3–4A (Wo¨lk et al., 2000), pCMVNS5Acon, and possible, therefore, that one or multiple determinants pCMVNS5Bcon (Moradpour et al., unpublished data) al- within the long hydrophobic domain in the center of low the CMV promotor-driven expression of the HCV NS4B promote membrane association. NS3–4A complex, NS5A, or NS5B, respectively. In membrane extraction and proteinase protection as- pTM1 (Moss et al., 1990) allows high-level in vitro says NS4B behaved as a cytoplasmically oriented inte- expression from a T7 RNA polymerase promotor fol- gral membrane protein. However, no protected trans- lowed by the encephalomyocarditis virus internal ribo- membrane or luminal fragments were detected after ra- some entry site. pTM1-MIA2 contains almost the entire dioactive labeling with [35S]methionine or [14C]leucine. gag-pro-pol region (nt 594 to 4102) of the murine IAP This was somewhat unexpected since in the case of the MIA14 (Fehrmann et al., 1997; Mietz et al., 1987). Plas- related yellow fever virus NS4B has been shown to span mids pTM1-MIA4 and pTM1-ppl will be described in the ER membrane at least once with multiple membrane- detail elsewhere (Fehrmann et al., manuscript in prepa- associated hydrophobic segments and cytoplasmic ration). Briefly, pTM1-MIA4 was derived from pL-MIA4 loops (Lin et al., 1993). Furthermore, membrane topology (Welker et al., 1997) and allows expression of a MIA14 models predicted by PredictProtein (on the World Wide gag gene of which the first 28 codons have been re- Web at URL www.embl-heidelberg.de) or TMHMM (URL: placed by the membrane-targeting signal of the Src pro- www.cbs.dtu.dk) and TopPred2 (URL: www.biokemi. tein. pTM1-ppl contains the entire coding region of bo- su.se) featured six or four internal transmembrane do- vine ppl derived from pSPB4 (Schlenstedt et al., 1992). mains, respectively. On the other hand, observations similar to ours, with a complete sensitivity to proteinase Tetracycline-regulated cell lines digestion, have been reported for NS4B of the related Tetracycline-regulated cell lines were generated as West Nile, Kunjin, and Dengue viruses (Wengler et al., previously described (Moradpour et al., 1996, 1998b,c; 1990; Cauchi et al., 1991; Westaway et al., 1997a). A Wo¨lk et al., 2000). In brief, the constitutively tTA-express- refined membrane topology model of HCV NS4B will be ing, U-2 OS human osteosarcoma-derived founder cell the subject of further experimentation. line UTA-6 (Englert et al., 1995) was cotransfected with In summary, we have shown that the HCV NS4B pro- pUHDNS4Bcon and pBabepuro (Morgenstern and Land, tein is tightly associated with the ER membrane, where it 1990). G418 and puromycin double-resistant clones were behaved as a cytoplasmically oriented integral mem- isolated and screened for tightly regulated NS4B expres- brane protein and colocalized with the other nonstruc- sion by indirect immunofluorescence microscopy and tural proteins. The data suggest that NS4B is a compo- immunoblot. UHCVcon-57 cells, which inducibly express nent of a membrane-associated cytoplasmic HCV repli- the entire open reading frame derived from a functional cation complex. An efficient replication system will be HCV H strain consensus cDNA (Kolykhalov et al., 1997), essential to further define the role of NS4B in the viral life and UGFP-9.22 cells, which inducibly express an en- cycle. hanced GFP derived from pEGFP-N1 (Clontech, Palo Alto, CA), will be described in detail elsewhere (Morad- MATERIALS AND METHODS pour et al., manuscript in preparation). Expression constructs Antibodies A fragment comprising nt 5475 to 6258 (aa 1712 to 1972) of a functional HCV H strain consensus cDNA NS4B derived from a genotype 1b HCV cDNA was (genotype 1a) was amplified by PCR from pBRTM/HCV1- expressed in E. coli and used as an antigen for the HCV NS4B PROTEIN 79

production of mAbs. Spleen cells from female BALB/c Western blot analysis mice immunized with this recombinant protein were Western blot analysis was performed as described fused with the X63-Ag8.653 myeloma cell line (American previously (Moradpour et al., 1996, 1998b). Type Culture Collection, Rockville, MD). Hybridomas were selected and supernatants were screened by Subcellular fractionation ELISA essentially as described (Harlow and Lane, 1988). Hybridomas immunoreactive with recombinant NS4B Subcellular fractionation was performed essentially as protein were cloned at least twice by limiting dilution. described previously (Moradpour et al., 1996). In brief, The polyclonal rabbit antiserum WU151 against NS4B 5 ϫ 10 7 UNS4Bcon-4 or UGFP-9.22 cells cultured for 48 h was produced by immunization of rabbits with a syn- in the absence of tetracycline were dounce-homoge- thetic peptide corresponding to the first 18 aa of geno- nized in a hypotonic buffer containing 10 mM Tris–HCl, type 1a NS4B (SQHLPYIEQGMMLAEQFK; Lin et al., un- pH 7.5, and 2 mM MgCl2, followed by centrifugation at published data). A pool of high-titer anti-HCV-positive 1000 g for 5 min to yield a nuclear pellet. The supernatant sera from 10 patients with chronic hepatitis C was used fraction was adjusted to 0.25 M sucrose and a mitochon- as a source of primary antibodies in some immunoblot drial pellet was obtained by centrifugation at 9000 g for analyses. The NS3-specific mAb 1B6 was described pre- 10 min. Finally, a microsomal pellet was separated from viously (Wo¨lk et al., 2000). mAbs 8N against NS4A and the cytosolic supernatant by centrifugation at 105,000 g for 40 min. 11H against NS5A were kindly provided by Jan Albert Hellings and Winand Habets (Organon Teknika B.V., Box- tel, The Netherlands). mAb 5B-12B7 against NS5B will be In vitro transcription–translation described elsewhere (Moradpour et al., manuscript in In vitro transcription–translation was performed using preparation). A polyclonal rabbit antiserum against PDI the TNT T7 coupled rabbit reticulocyte lysate system was obtained from StressGen (Victoria, BC, Canada). The (Promega, Madison, WI) essentially following the manu- mAb G1/93 against human ERGIC-53 (Schweizer et al., facturer’s recommendations. Reactions were incubated 1988) was kindly provided by Hans-Peter Hauri (Univer- for 90 min at 30°C in the presence of 0.75 mCi/ml sity of Basel, Switzerland). A polyclonal rabbit antiserum [35S]methionine (Amersham, UK) in a volume of 25 ␮l. to Man II (Moremen, 1991) was kindly provided by Kelley Where indicated, 1.5 ␮l canine pancreatic microsomes (a Moremen (University of Georgia, Athens, GA). The mAb gift from Richard Zimmermann, Medizinische Biochemie JL-8 against GFP was purchased from Clontech. und Molekularbiologie, Universita¨t des Saarlandes, Homburg, Germany) and digitonin at a final concentration of 0.01% were added. Initial titration experiments Indirect immunofluorescence and confocal laser indicated that 1.5 ␮l microsomal membranes per 25 ␮l scanning microscopy coupled in vitro transcription–translation reaction was Indirect immunofluorescence microscopy was per- optimal since larger quantities led to a strong reduction formed as described previously (Moradpour et al., 1996, of translation efficiency. ␮ 1998b). In brief, cells grown as monolayers on glass For membrane sedimentation analyses, 15 l NTE coverslips were fixed with 2% paraformaldehyde, perme- buffer (100 mM NaCl, 10 mM Tris–HCl, pH 8.0, 1 mM abilized with 0.05% saponin, and incubated with primary EDTA) was added after completion of the in vitro tran- antibodies in phosphate-buffered saline containing 3% scription–translation reaction, followed by centrifugation bovine serum albumin and 0.05% saponin. Bound pri- at 12,000 g for 15 min. Supernatants were collected and pellets were resuspended in 40 ␮l NTE buffer. Subse- mary antibody was revealed with fluorescein isothiocya- quently, pellet and supernatant fractions were separated nate (FITC)-conjugated goat F(abЈ) fragment to mouse 2 by SDS–PAGE and analyzed by autoradiography. Gels IgG F(abЈ) (Cappel, Durham, NC) or sheep F(abЈ) frag- 2 2 were scanned on a Fuji BAS1000 phosphoimager and ment to rabbit IgG (Boehringer Mannheim, Germany). For analyzed using the Fuji MacBAS V2.4 software. colocalization experiments, a Texas Red (TXR)-conju- For analyses of co- and posttranslational membrane Ј gated sheep F(ab )2 to mouse IgG (ICN/Cappel, Aurora, association, microsomal membranes were added to the OH) was used as a secondary antibody to reveal bound reactions either during or for1hat30°C after completion murine mAbs. Coverslips were mounted in SlowFade of in vitro transcription–translation. In the latter setting, (Molecular Probes, Eugene, OR) and examined with a translation was stopped by 1.25 mM puromycin prior to Zeiss Axiovert photomicroscope equipped with an epi- the addition of microsomal membranes. fluorescence attachment. Confocal laser scanning mi- For membrane extraction experiments, the pellets croscopy was performed using a Zeiss LSM 410 micro- from 15 ␮l in vitro transcription–translation reactions scope and images were processed with the Adobe Pho- performed in the presence of microsomal membranes toshop program, version 3.0.5. were resuspended in 40 ␮l NTE buffer, 1 M NaCl, 100 80 HU¨ GLE ET AL. mM sodium carbonate, pH 11.5, 4 M urea, or 0.5% Triton Gossen, M., and Bujard, H. (1992). Tight control of gene expression in X-100 and incubated for 20 min at 4°C. Subsequently, mammalian cells by tetracycline-responsive promoters. Proc. Natl. membrane sedimentation analyses were performed as Acad. Sci. USA 89, 5547–5551. Harlow, E., and Lane, D. (1988). “Antibodies—A Laboratory Manual.” described above. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. For proteinase protection assays, 1 ␮l proteinase K Hijikata, M., Mizushima, H., Tanji, Y., Komoda, Y., Hirowatari, Y., Akagi, stock solution (500 ␮g/ml 3 final concentration of 50 T., Kato, N., Kimura, K., and Shimotohno, K. (1993). Proteolytic pro- ␮g/ml) was added to a 10-␮l in vitro transcription–trans- cessing and membrane association of putative nonstructural pro- lation reaction containing microsomal membranes. After teins of hepatitis C virus. Proc. Natl. Acad. Sci. USA 90, 10773–10777. Khromykh, A. A., Sedlak, P. L., and Westaway, E. G. (2000). cis- and 30 min incubation at 4°C, proteolysis was terminated by trans-acting elements in flavivirus RNA replication. J. Virol. 74, 3253– the addition of 2 ␮l 12.5 mM phenylmethylsulfonyl fluo- 3263. ride, followed by SDS–PAGE of the samples. Kim, J.-E., Song, W. K., Chung, K. M., Back, S. H., and Jang, S. K. (1999). Subcellular localization of hepatitis C viral proteins in mammalian ACKNOWLEDGMENTS cells. Arch. Virol. 144, 329–343. Koch, J. O., and Bartenschlager, R. (1999). Modulation of hepatitis C The authors gratefully acknowledge Benno Wo¨lk for help with con- virus NS5A hyperphosphorylation by nonstructural proteins NS3, focal laser scanning microscopy, Christoph Englert for UTA-6 cells, NS4A, and NS4B. J. Virol. 73, 7138–7146. Richard Zimmermann for canine pancreatic microsomal membranes, Kolykhalov, A. A., Agapov, E. V., Blight, K. J., Mihalik, K., Feinstone, S. M., Jan Albert Hellings and Winand Habets for mAbs 8N and 11H, Hans- and Rice, C. M. (1997). Transmission of hepatitis C by intrahepatic Peter Hauri for the mAb G1/93 against ERGIC-53, and Kelley Moremen inoculation with transcribed RNA. Science 277, 570–574. for the antiserum against Man II. Kretzschmar, E., Bui, M., and Rose, J. K. (1996). Membrane association This work was supported by Grant Mo 799/1-2 from the Deutsche of influenza matrix protein does not require specific hydrophobic Forschungsgemeinschaft to D.M. and H.E.B. and by grants from the domains or the viral glycoproteins. Virology 220, 37–45. Public Health Service (CA57973 and AI40034) and the Greenberg Kutay, U., Ahnert-Hilger, G., Hartmann, E., Wiedenmann, B., and Rap- Medical Foundation to C.M.R. oport, T. A. (1995). Transport route for synaptobrevin via a novel pathway of insertion into the endoplasmic reticulum membrane. REFERENCES EMBO J. 14, 217–223. Lin, C., Amberg, S. M., Chambers, T. J., and Rice, C. M. (1993). Cleavage Anderson, D. J., Mostov, K. E., and Blobel, G. (1983). Mechanisms of at a novel site in the NS4A region by the yellow fever virus NS2B-3 integration of de novo-synthesized polypeptides into membranes: proteinase is a prerequisite for processing at the downstream 4A/4B Signal-recognition particle is required for integration into microsomal signalase site. J. Virol. 67, 2327–2335. membranes of calcium ATPase and of lens MP26 but not of cyto- Lin, C., Wu, J.-W., Hsiao, K., and Su, M.-S. (1997). The hepatitis C virus chrome b5. Proc. Natl. Acad. Sci. USA 80, 7249–7253. NS4A protein: Interactions with the NS4B and NS5A proteins. J. Virol. Bartenschlager, R., and Lohmann, V. (2000). Replication of hepatitis C 71, 6465–6471. virus. J. Gen. Virol. 81, 1631–1648. Lohmann, V., Ko¨rner, F., Koch, J.-O., Herian, U., Theilmann, L., and Bienz, K., Egger, D., Pfister, T., and Troxler, M. (1992). Structural and Bartenschlager, R. (1999). Replication of subgenomic hepatitis C functional characterization of the poliovirus replication complex. J. Vi- virus RNAs in a hepatoma cell line. Science 285, 110–113. rol. 66, 2740–2747. Mackenzie, J. M., Jones, M. K., and Westaway, E. G. (1999). Markers for Cauchi, M. R., Henchal, E. A., and Wright, P. J. (1991). The sensitivity of trans-Golgi membranes and the intermediate compartment localize cell-associated dengue virus proteins to trypsin and the detection of to induced membranes with distinct replication functions in flavivi- trypsin-resistant fragments of the nonstructural glycoprotein NS1. rus-infected cells. J. Virol. 73, 9555–9567. Virology 180, 659–667. Matlack, K. E., Mothes, W., and Rapoport, T. A. (1998). Protein translo- Deshaies, R. J., Sanders, S. L., Feldheim, D. A., and Schekman, R. (1991). cation: Tunnel vision. Cell 92, 381–390. Assembly of yeast Sec proteins involved in translocation into the Mietz, J. A., Grossman, Z., Lueders, K. K., and Kuff, E. L. (1987). Nucle- endoplasmic reticulum into a membrane-bound multisubunit com- otide sequence of a complete mouse intracisternal A-particle ge- plex. Nature 349, 806–808. nome: Relationship to known aspects of particle assembly and Egger, D., Pasamontes, L., Bolten, R., Boyko, V., and Bienz, K. (1996). function. J. Virol. 61, 3020–3029. Reversible dissociation of the poliovirus replication complex: Func- Moradpour, D., Englert, C., Wakita, T., and Wands, J. R. (1996). Charac- tions and interactions of its components in viral RNA synthesis. terization of cell lines allowing tightly regulated expression of hep- J. Virol. 70, 8675–8683. atitis C virus core protein. Virology 222, 51–63. Egger, D., Teterina, N., Ehrenfeld, E., and Bienz, K. (2000). Formation of Moradpour, D., Englert, C., and Blum, H. E. (1998a). Independent reg- the poliovirus replication complex requires coupled viral translation, ulation of two separate gene activities in a continuous human cell vesicle production, and viral RNA synthesis. J. Virol. 74, 6570–6580. line. Biol. Chem. 379, 1189–1191. Englert, C., Hou, X., Maheswaran, S., Bennett, P., Ngwu, C., Re, G. G., Moradpour, D., Kary, P., Rice, C. M., and Blum, H. E. (1998b). Continuous Garvin, A. J., Rosner, M. R., and Haber, D. A. (1995). WT1 suppresses human cell lines inducibly expressing hepatitis C virus structural and synthesis of the epidermal growth factor receptor and induces apo- nonstructural proteins. Hepatology 28, 192–201. ptosis. EMBO J. 14, 4662–4675. Moradpour, D., Wakita, T., Wands, J. R., and Blum, H. E. (1998c). Tightly European Association for the Study of the Liver (1999). Consensus regulated expression of the entire hepatitis C virus structural region Statement. J. Hepatol. 30, 956–961. in continuous human cell lines. Biochem. Biophys. Res. Commun. Fehrmann, F., Welker, R., and Kra¨usslich, H. G. (1997). Intracisternal 246, 920–924. A-type particles express their proteinase in a separate reading frame Moremen, K. W., Touster, O., and Robbins, P. W. (1991). Novel purifica- by translational frameshifting, similar to D-type retroviruses. Virology tion of the catalytic domain of Golgi alpha-mannosidase II: Charac- 235, 352–359. terization and comparison with the intact enzyme. J. Biol. Chem. 266, Fujiki, Y., Hubbard, A. L., Fowler, S., and Lazarow, P. B. (1982). Isolation 16876–16885. of intracellular membranes by means of sodium carbonate treatment: Morgenstern, J. P., and Land, H. (1990). Advanced mammalian gene Application to endoplasmic reticulum. J. Cell Biol. 93, 97–102. transfer: High titre retroviral vectors with multiple drug selection HCV NS4B PROTEIN 81

markers and a complementary helper-free packaging cell line. Nu- ated with a tubulo-vesicular compartment at the cis-side of the Golgi cleic Acids Res. 18, 3587–3596. apparatus. J. Cell Biol. 107, 1643–1653. Moss, B., Elroy-Stein, O., Mizukami, T., Alexander, W. A., and Fuerst, Suhy, D. A., Giddings, T. H., Jr., and Kirkegaard, K. (2000). Remodeling of T. R. (1990). New mammalian expression vectors. Nature 348, 91–92. the endoplasmic reticulum by poliovirus infection and by individual Mu¨ller, G., and Zimmermann, R. (1987). Import of honeybee prepromelit- viral proteins: An autophagy-like origin for virus-induced vesicles. tin into the endoplasmic reticulum: Structural basis for independence J. Virol. 74, 8953–8965. of SRP and docking protein. EMBO J. 6, 2099–2107. van Regenmortel, M. H. V., Fauquet, C. M., Bishop, D. H. L., Carstens, National Institutes of Health (1997). National Institutes of Health con- E. B., Estes, M. K., Lemon, S. M., Maniloff, J., Mayo, M. A., McGeoch, sensus development conference panel statement: Management of D. J., Pringle, C. R., and Wickner, R. B. (2000). Virus Taxonomy: The hepatitis C. Hepatology 26(Suppl. 1), 2S–10S. VIIth Report of the International Committee on Taxonomy of Viruses. Neddermann, P., Clementi, A., and De Francesco, R. (1999). Hyperphos- Academic Press, San Diego. phorylation of the hepatitis C virus NS5A protein requires an active Walter, P., and Johnson, A. E. (1994). Signal sequence recognition and NS3 protease, NS4A, NS4B, and NS5A encoded on the same protein targeting to the endoplasmic reticulum membrane. Annu. polyprotein. J. Virol. 73, 9984–9991. Rev. Cell. Biol. 10, 87–119. Panzner, S., Dreier, L., Hartmann, E., Kostka, S., and Rapoport, T. A. Welker, R., Janetzko, A., and Kra¨usslich, H. G. (1997). Plasma membrane (1995). Posttranslational protein transport in yeast reconstituted with targeting of chimeric intracisternal A-type particle polyproteins leads to particle release and specific activation of the viral proteinase. a purified complex of Sec proteins and Kar2p. Cell 81, 561–570. J. Virol. 71, 5209–5217. Park, J. S., Yang, J. M., and Min, M. K. (2000). Hepatitis C virus non- Wengler, G., Wengler, G., Nowak, T., and Castle, E. (1990). Description structural protein NS4B transforms NIH3T3 cells in cooperation with of a procedure which allows isolation of viral nonstructural proteins the Ha-ras oncogene. Biochem. Biophys. Res. Commun. 267, 581– from BHK vertebrate cells infected with the West Nile flavivirus in a 587. state which allows their direct chemical characterization. Virology Precious, B., Young, D. F., Bermingham, A., Fearns, R., Ryan, M., and 177, 795–801. Randall, R. E. (1995). Inducible expression of the P, V, and NP genes Westaway, E. G., Khromykh, A. A., Kenney, M. T., Mackenzie, J. M., and of the paramyxovirus simian virus 5 in cell lines and an examination Jones, M. K. (1997a). Proteins C and NS4B of the flavivirus Kunjin of NP-P and NP-V interactions. J. Virol. 69, 8001–8010. translocate independently into the nucleus. Virology 234, 31–41. Reed, K. E., and Rice, C. M. (2000). Overview of hepatitis C virus Westaway, E. G., Mackenzie, J. M., Kenney, M. T., Jones, M. K., and genome structure, polyprotein processing, and protein properties. Khromykh, A. A. (1997b). Ultrastructure of Kunjin virus-infected cells: Curr. Top. Microbiol. Immunol. 242, 55–84. Colocalization of NS1 and NS3 with double-stranded RNA, and of Rice, C. M. (1996). Flaviviridae: The viruses and their replication. In NS2B with NS3, in virus-induced membrane structures. J. Virol. 71, “Fields Virology” (B. N. Fields, D. M. Knipe, and P. M. Howley, Eds.), 6650–6661. third ed. Lippincott–Raven, Philadelphia. Wickner, W., and Leonard, M. R. (1996). Escherichia coli preprotein Schatz, G., and Dobberstein, B. (1996). Common principles of protein translocase. J. Biol. Chem. 271, 29514–29516. translocation across membranes. Science 271, 1519–1526. Wo¨lk, B., Sansonno, D., Kra¨usslich, H.-G., Dammacco, F., Rice, C. M., Schatz, P. J., and Beckwith, J. (1990). Genetic analysis of protein export Blum, H. E., and Moradpour, D. (2000). Subcellular localization, sta- in Escherichia coli. Annu. Rev. Genet. 24, 215–248. bility and trans-cleavage competence of the hepatitis C virus NS3- Schlenstedt, G., Gudmundsson, G. H., Boman, H. G., and Zimmermann, NS4A complex expressed in tetracycline-regulated cell lines. J. Virol. R. (1992). Structural requirements for transport of preprocecropinA 74, 2293–2304. and related presecretory proteins into mammalian microsomes. Yao, N., Reichert, P., Taremi, S. S., Prosise, W. W., and Weber, P. C. J. Biol. Chem. 267, 24328–24332. (1999). Molecular views of viral polyprotein processing revealed by Schweizer, A., Fransen, J. A., Bachi, T., Ginsel, L., and Hauri, H. P. (1988). the crystal structure of the hepatitis C virus bifunctional protease- Identification, by a monoclonal antibody, of a 53-kD protein associ- helicase. Struct. Fold. Des. 7, 1353–1363.