Both NS3 and NS4A Are Required for Proteolytic ... - Europe PMC

1 downloads 0 Views 2MB Size Report
Fax: 39-6-910-. 93225. Electronic mail address: [email protected]. the case for the homologous NS3 protein of flaviviruses (9) and p80 of pestiviruses (45).
Vol. 68, No. 6

JOURNAL OF VIROLOGY, June 1994, p. 3753-3760

0022-538X/94/$04.00+0 Copyright © 1994, American Society for Microbiology

Both NS3 and NS4A Are Required for Proteolytic Processing of Hepatitis C Virus Nonstructural Proteins CRISTINA FAILLA, LICIA TOMEI, AND RAFFAELE DE FRANCESCO* Istituto di Ricerche di Biologia Molecolare "P. Angeletti"-Pomezia, 00040 Rome, Italy

Received 20 December 1993/Accepted 11 March 1994

The proteolytic cleavages at the NS3-NS4A, NS4A-NS4B, NS4B-NS5A, and NS5A-NS5B junctions of hepatitis C virus (HCV) polyprotein are effected by the virus-encoded serine protease contained within NS3. Using transient expression in HeLa cells of cDNA fragments that code for regions of the HCV polyprotein, we studied whether viral functions other than NS3 are required for proteolytic processing at these sites. We found that, in addition to NS3, a C-terminal 33-amino-acid sequence of the NS4A protein is required for cleavage at the NS3-NS4A and NS4B-NS5A sites and that it accelerates the rate of cleavage at the NS5A-NS5B junction. In addition, we show that NS4A can activate the NS3 protease when supplied in trans. Our data suggest that HCV NS4A may be the functional analog of flavivirus NS2B and pestivirus plO proteins.

Hepatitis C virus (HCV) is the major etiological agent of posttransfusion and sporadic community-acquired non-A, non-B hepatitis (11, 31). In most cases, HCV causes a persistent liver infection that eventually develops into cirrhosis or hepatocellular carcinoma (10). HCV contains a positive-stranded linear RNA genome of about 9.5 kb in length (12, 29, 41). On the basis of their genomic structure and virion properties, HCV, the pestivirus, and the flavivirus have been classified as three genera in the family Flaviviridae (19). As in the case of flavi- and pestiviruses, HCV is a small, enveloped virus, whose genomic RNA includes a single open reading frame encoding a precursor polyprotein that is cleaved co- or posttranslationally into mature viral polypeptides. The structural proteins are encoded by the N-terminal portion of the genome, whereas the remaining part codes for the nonstructural proteins, which are believed to be components of the viral replication machinery (for a recent review on the molecular biology of HCV, see reference 34). The map of the HCV polyprotein has been obtained by cell-free and cell culture expression studies (23, 24, 42). The gene order that has emerged from these studies is 5'-C-E1-E2NS2-NS3-NS4A-NS4B-NSSA-NSSB-3'. C is a basic protein that binds RNA (39a), and it is postulated to be the nucleocapsid protein. El and E2 are putative envelope glycoproteins. NS2 through NS5B represent the viral nonstructural proteins and could be functional analogs of the corresponding proteins of pesti- and flaviviruses (7, 13). Processing of the polyprotein requires both host and viral proteases. The signal peptidase of endoplasmic reticulum is responsible for generating the N termini of El, E2, and possibly NS2 (24). Whereas it has been recently established that a novel virus-encoded metalloprotease is responsible for the cleavage between NS2 and NS3 (21, 25), we and others have identified NS3 as the protease responsible for cleavages at the NS3-NS4A, NS4A-NS4B, NS4B-NSSA, and NS5ANS5B junctions (3, 15, 22, 42). Analysis of the amino acid sequence of the NS3 protein of HCV suggested that this viral protein contains a trypsin-like serine protease domain that functions in the processing of the viral polyprotein (35), as is

the case for the homologous NS3 protein of flaviviruses (9) and p80 of pestiviruses (45). The active site of serine proteases invariably contains three residues, histidine, aspartate, and serine, that maintain the same relative spatial position in all the known structures of these enzymes. These residues constitute the catalytic triad: the histidine imidazole group abstracts the alcohol proton of the serine and transfers it to the amine-leaving group, while the aspartate stabilizes the positive charge that develops on the histidine in the intermediate complex (36). The histidine 1083, aspartate 1107, and serine 1165 residues (numbered according to their location in the polyprotein of BK strain HCV [41]), found in the N-terminal domain of the NS3 protein, are highly conserved among all HCV strains sequenced so far and have been proposed to constitute the catalytic triad of the HCV protease (4, 7, 35). Consistent with this prediction, when the presumptive catalytic serine 1165 of the polyprotein was mutated to alanine, processing at the NS3-NS4A, NS4A-NS4B, NS4B-NS5A, and NS5A-NS5B sites was abolished (22, 42). Cleavage at three of the four sites of the mutated polyprotein could be restored by supplying functional NS3 in trans, while processing at NS3-NS4A occurred only in cis (42). This finding indicated that the cleavage of the latter site is an intramolecular event. Interestingly, the NS3-NS4A cleavage site differs from the others in that the processing takes place between a threonine and a serine residue, whereas the remainder are cleaved between a cysteine and a serine (22, 37). In flavivirus, NS3 is necessary, but not sufficient, for processing at the nonstructural cleavage sites: in addition to the catalytic NS3 domain, NS2B is required for proteolytic activity (1, 6, 18, 33, 38). A stable physical interaction between NS3 and NS2B has been demonstrated (1, 18). In the case of the other two genera of the family Flavivinidae, pestiviruses and HCV, it has been observed that the respective viral serine proteases, p80 and NS3, can function independently of all the regions of the polyprotein that are present upstream of the proteases themselves (22, 42, 45). However, in the case of bovine viral diarrhea virus (BVDV), a member of the pestivirus genus, a region of the polyprotein outside p80 may be necessary to effect cleavage by the viral serine protease at the p58-p75 site (45). This region has tentatively been mapped to a portion of the polyprotein (13, 45) that corresponds to HCV NS4B. We initiated a study aimed to elucidate whether viral

* Corresponding author. Phone: 39-6-910-93221. Fax: 39-6-91093225. Electronic mail address: [email protected].

3753

3754

FAILLA ET AL.

functions other than NS3 are required for the proteolytic processing of the nonstructural region of the HCV polyprotein. Using transient expression in HeLa cells of cDNA fragments that encode regions of the HCV polyprotein, we analyzed the cleavage activity of the NS3 serine protease on each individual cleavage site. In this paper, we report that, in addition to NS3, a C-terminal domain of the NS4A protein is also required for efficient cleavage at the NS3-NS4A, NS4A-NS4B, NS4BNS5A, and NS5A-NS5B junctions. In addition, we also show that NS4A can act in trans to affect cleavage at these sites.

MATERLALS AND METHODS Cells and virus. HeLa cells, originally obtained from the American Type Culture Collection, were grown in Dulbecco's modified Eagle's medium containing 10% fetal calf serum (FCS). Vaccinia virus vTF7-3 (20) was grown in RK13 cells cultured in minimum essential medium containing 10% FCS. Construction of recombinant plasmids. HCV cDNA clones of the HCV-BK isolate were used for all plasmid constructions (41). DNA fragments derived from HCV cDNA were inserted downstream of the 5' untranslated region of encephalomyocarditis virus and under a T7 promoter in the pCite-1 vector (Novagen), in the appropriate translational reading frame and followed by a termination codon. All the plasmids described below were obtained with standard recombinant DNA technology (39). Plasmid pCite(NS3) contains the portion of the HCV genome between nucleotides 3351 and 5177 and therefore encodes the portion of the viral polyprotein between amino acids 1007 and 1615. Construction of this plasmid has been described earlier (42). pCite(NS4AB) bears a cDNA fragment corresponding to nucleotides 5276 to 6225 of the HCV genome (amino acid residues 1649 to 1964 of the polyprotein). pCite(NS5AB) contains the portion of the HCV genome between nucleotides 6224 and 9400 (amino acid residues 1965 to 3010 of the polyprotein). pCite(NS4B/5A) was obtained by cloning in pCite-1 a cDNA fragment containing nucleotides 5653 to 7472. This construct encodes the portion of the polyprotein between amino acid residues 1775 and 2380. pCite(NS3-AP) is a derivative of pCite(SX) (42) which contains HCV cDNA from nucleotides 3305 to 5373 (amino acid residues 992 to 1680). A fragment derived from the pBC12/RSV/AP vector and coding for placental alkaline phosphatase (5) was inserted downstream of this region, with care taken to maintain the correct reading frame. pCite(NS3/4A) contains a cDNA fragment comprising the region between nucleotides 3305 and 5465 of HCV cDNA. The resulting construct codes for the portion of the HCV polyprotein between amino acids 992 and 1711. pCite(NS3/4A:S , 165-A) encodes the same polypeptide, except that the catalytic serine (S-1165) of NS3 has been replaced with an alanine. pCite(NS3/4AAC) contains a cDNA fragment comprising the region between nucleotides 3305 and 5373 of HCV cDNA. The resulting construct codes therefore for the portion of the HCV polyprotein between amino acids 992 and 1680. pCite(NS4A) bears a cDNA fragment spanning nucleotides 5276 to 5465 of the HCV genome (amino acid residues 1649 to 1711). pCite(NS4ABAN) contains the HCV sequence from nucleotide 5367 to nucleotide 6225 (amino acid residues 1679 to 1964). pCite(NS3AC) is a 3-deletion product of pCite(NS3) (42)

J. VIROLv.

that contains HCV cDNA from nucleotide 3351 to nucleotide 4043 (amino acid residues 1007 to 1237). Preparation of labelled extracts. HeLa cells seeded at a density of 4 x 105 cells per 6-cm plate were infected with vaccinia virus vTF7-3 at a multiplicity of 5 PFU per cell as previously described (42). After adsorption for 30 min at 37°C, 3 ml of Dulbecco's modified Eagle's medium supplemented with 10% FCS was added. Cells were incubated for an additional 30 min at 37°C. Recombinant plasmid DNA was precipitated by the calcium phosphate method (39) and added directly to each plate. When required, pUC19 plasmid DNA was used as carrier DNA for transfection. At 4 h posttransfection, the medium was replaced with minimum essential medium lacking methionine (GIBCO) and the cells were starved for 1 h at 37°C. Cells were then radiolabelled for 3 h with 400 jLCi of Tran35S-label (ICN) in 2 ml of minimum essential medium lacking methionine and supplemented with 2% dialyzed FCS. Cells were harvested and resuspended in IPB150 (20 mM Tris-HCl [pH 8], 150 mM NaCl, 1% Triton) supplemented with 1 mM phenylmethylsulfonyl fluoride, 1 mM EDTA, and 1 mM dithiothreitol. Immunoprecipitations were conducted on denatured extracts with HCV region-specific rabbit polyclonal antisera as described previously (37, 42). Analytical gel electrophoresis of the immunoprecipitated products was performed as described previously (37). When necessary, quantitation of radioactive protein bands was carried out by densitometric scanning of the autoradiograms with the program Image (version 1.22) on a Macintosh LC475 personal computer equipped with a Sony CCD videocamera. RESULTS

Expression plasmids and method of analysis. Previous studies of HCV polyprotein processing indicated that NS3 is a serine protease involved in the maturation of HCV nonstructural proteins (3, 22, 42). Although it has been demonstrated that NS3 is absolutely required for the cleavage at the NS3NS4A, NS4A-NS4B, NS4B-NSSA, and NS5A-NS5B sites, no evidence that the same polypeptide is per se sufficient for cleavage, either in vitro or in vivo, has been provided. In order to address this matter, we constructed the series of expression plasmids described in Fig. 1. Appropriate fragments of the HCV genome were cloned downstream of the T7 bacteriophage promoter in the pCite-1 vector. This expression vector contains the encephalomyocarditis virus internal ribosome entry site (27), which promotes efficient and CAPindependent translation of mRNA transcribed from the T7 promoter (16). In order to obtain transient expression of the various portions of the HCV polyprotein, HeLa cells were infected with vTF7-3, a recombinant vaccinia virus that supports the synthesis of bacteriophage T7 polymerase in the cytoplasm of infected cells (20), and were then transfected with the appropriate plasmids. Cell proteins were metabolically labelled with [35S]methionine as described in Materials and Methods. Viral proteins and processing products were then identified by immunoprecipitation of denatured whole-cell extracts with region-specific polyclonal antibodies to HCV NS3, NS4, and NSSA (37, 42) and on the basis of their molecular weights. It must be pointed out, however, that in the absence of N-terminal sequence analysis of each processing product we cannot rule out that new cleavage sites have been generated in the genetically altered NS3 protease substrates. NS4A is required for the NS3-NS4A cleavage. NS3 generates the N terminus of NS4A by cleaving itself off the polyprotein via an intermolecular reaction (in cis) (42). We initially wanted

NS4A, A COFACTOR OF HCV NS3 PROTEASE

VOL. 68, 1994

A

XI&

5'

B

B

A

XI&

m m

c

E1 E2 190 383

il. '

NS2 NS3 810 1027

3755

3'

NS5

NS4 1658 1712 1973

2420

pCite(NS3)

pCite(NS3AC) pCite(NS3/4AAC) pCite(NS3/4A)

pCite(NS3-AP) pCite(NS4A)

pCite(NS4AB) pCite(NS4ABAN)

pCite(NS4B/5A)

pCite(NS5AB) pCite(NS3/4A:Scat-A)

Icat.A

FIG. 1. Schematic representation of the recombinant expression plasmids used in this study. The organization of viral polyprotein is shown at the top. The vertical bars indicate the boundaries between the different proteins. The numbers at the top of each bar indicate the position of the N-terminal amino acid of the following protein within the viral polyprotein. The HCV polyprotein portions expressed by the different constructs are shown below; the names given to the corresponding plasmids are shown on the left. The rectangle marked AP indicates a portion of placental alkaline phosphatase.

to study what viral structures are required for this intramolecular cleavage to occur. As a first step towards this goal, we analyzed the protein products of pCite(NS3/4A), a plasmid predicted to drive the expression of a polypeptide spanning the active protease domain of NS3 and ending at the C terminus of NS4A (Fig. 1). As shown in Fig. 2A, transient expression of this plasmid in HeLa cells yielded proteins of 70 kDa (lane 1) and 6 kDa (lane 2) that immunoprecipitated with anti-NS3 and anti-NS4 antisera, respectively. The NS3 protein generated by this construct migrates anomalously in comparison with wild-type NS3 because it contains a short stretch of unprocessed NS2 at its N terminus (42). The result of the experiment just described indicates that the NS3-dependent cleavage between NS3 and NS4A occurs efficiently in the absence of any other viral proteins. We then wished to determine what the minimal portion of NS4A required for this cis cleavage to occur is. To this end, we constructed pCite(NS3/4AAC), in which a termination codon was inserted after the 23 N-terminal amino acids of NS4A (Fig. 1). This latter plasmid yielded a protein, recognized by the anti-NS3 antisera, that corresponds in molecular weight to an uncleaved precursor (data not shown). However, the very small difference between the molecular weight of this latter protein precursor and that of mature NS3 made it very difficult to establish whether proteolytic cleavage was completely or only partially abolished. In order to resolve the possible cleavage products from the unprocessed precursor on sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), we decided to append a heterologous sequence to the N-terminal 23 residues of NS4A. A cDNA fragment encoding a soluble form of a placental alkaline phosphatase (5) was thus fused in frame at the 3' end of the coding region of the pCite(NS3/ 4AAC) construct (Fig. 1). Transfection with the resulting plasmid, pCite(NS3-AP), resulted in the accumulation of a 130-kDa uncleaved protein precursor that was recognized by the anti-NS3 antiserum (Fig. 2B, lane 1).

114~~~~+

K A+/-

NS3/4A

N53-AP

B

A

; DA}

(kDA)

MW

(kDA)

ct3A4

.. 90_ 6 9- S

a3

NS4A

a4 a3

a_-

NS3-AP

90-

NS3

69-

45 -

45-

3 0-

30-

d

NS3

14-

14 -

NS4A

NS4A

1 2 3 1 2 FIG. 2. Processing in cis at the NS3-NS4A junction. Vaccinia virus vT7F-3-infected HeLa cells were transfected as described in Materials and Methods. Cells were labelled with [5S]methionine for 3 h, and SDS-denatured lysates were immunoprecipitated with antisera to specific regions of HCV polyprotein. The positions of the relevant HCV proteins immunoprecipitated with anti-HCV antibodies are indicated. a3 and a4 indicate antisera against NS3 and NS4, respectively. MW, molecular weight standards. (A) Lysates from cells transfected with pCite(NS3/4A) were immunoprecipitated with anti-NS3 (lane 1) and anti-NS4 (lane 2). The immunoprecipitated products were resolved on an SDS-14% PAGE. (B) Lysates from cells transfected with pCite(NS3-AP) alone (lane 1) or cotransfected with pCite(NS4A) (lanes 2 and 3) were immunoprecipitated with anti-NS3 (lanes 1 and 3) and anti-NS4 (lane 2). The immunoprecipitated products were resolved on an SDS-14% PAGE.

3756

FAILLA ET AL.

J. VIROL.

A

B NS5AB

2 \SS,: ;S

NS4ABAN

V

_

NS4ABAN

NS3 MW ax4 (kDA)

VS4BI5A

+/- E NS4A

a3 a4

aY3 a4

NS5AB

NtS4B/5A

NS3

MW a5A

aS3 nSA

(kDA)

NS3

9-

6 9

NS3

-

45-

%

N53 -

_

NS4B/5A

+1 -

NS3

NS4A

NS4A

MW a5A a3 a5A cx3 a aSA (kD A -NS5AB WM NS5AB

Y3 ut4 a5A

9g0-

906

NS4B/5A

69-

N53 -

NS3 _

NS3

-

- NS5A

NS4B/5A

NS5A

- NS5A

45 -

45-

30-

_

NS4ABAN

-

-

30-

30

14

1 4-

-

NS4ABAN

NS4B

-

1 4-

NS4A

NS4A

NS4A 1

1

2 3

2

3

4

5

6

2

3

4

5

6

4 5

FIG. 3. Processing of the HCV polyprotein substrate in trans. (A) Plasmid pCite(NS4ABAN) was transfected alone (lane 1), with pCite(NS3) (lanes 2 and 3), or with both pCite(NS3) and pCite(NS4A) (lanes 4 and 5). Cell lysates were immunoprecipitated with anti-NS3 (lanes 2 and 4) and anti-NS4 (lanes 1, 3, and 5). (B) Plasmid pCite(NS4B/5A) was transfected alone (lane 1), with pCite(NS3) (lanes 2 and 3), or with both pCite(NS3) and pCite(NS4A) (lanes 4, 5, and 6). Cell lysates were immunoprecipitated with anti-NS3 (lanes 2 and 4), anti-NS4 (lane 5), and anti-NS5A (lanes 1 and 3). (C) Plasmid pCite(NS5AB) was transfected alone (lane 1), with pCite(NS3) (lanes 2 and 3), or with both pCite(NS3) and pCite(NS4A) (lanes 4, 5, and 6). Cell lysates were immunoprecipitated with anti-NS3 (lanes 2 and 4), anti-NS4 (lane 5), and anti-NS5A (lanes 1, 3, and 6). c3, c4, and ox5A indicate antisera against NS3, NS4, and NS5A, respectively. The positions of the relevant HCV proteins immunoprecipitated with anti-HCV antibodies are indicated. MW, molecular weight standards.

The data described thus far suggest that an intact NS4A protein might be necessary for the NS3 protease to cleave at the NS3-NS4A site. There are essentially two possible explanations for this observation: either the integrity of NS4A is required to maintain the appropriate conformation at the NS3-NS4A junction, or NS4A acts as a cofactor for the NS3 protease activity. Whereas it would be difficult to prove the former hypothesis, we reasoned that if the latter were the case, then we might be able to observe cleavage at the NS3-NS4A boundary of the NS3/alkaline phosphatase chimera providing intact NS4A in trans. We therefore transfected, together with the pCite(NS3-AP) construct, a plasmid driving the expression of intact NS4A [pCite(NS4A), Fig. 1]. The cotransfection resulted in the production of the expected 6-kDa NS4A protein (Fig. 2B, lane 2). More noticeably, in addition to some uncleaved precursor, a protein corresponding in size to the expected cleavage product (70 kDa; see above) was specifically recognized by the anti-NS3 antisera (Fig. 2B, lane 3). These results indicate that NS3 and NS4A are both required for cleavage at the NS3-NS4A junction and that NS4A is capable of acting in trans to effect cleavage at this site. NS4A is also an effector of cleavage by NS3 at the NS4BNS5A and NS5A-NS5B junctions. In order to establish whether NS4A is involved as a cofactor of NS3 protease activity on the remaining cleavage sites, we constructed pCite(NS4ABAN), pCite(NS4B/5A), and pCite(NS5AB) (Fig. 1). Each plasmid was then transfected alone or cotransfected with pCite(NS3) alone or with the combination of pCite(NS3) and pCite(NS4A). Transfection with plasmid pCite(NS4ABAN) results in the synthesis of a protein of 30 kDa that is recognized by the anti-NS4 antiserum (Fig. 3A, lane 1). This protein contains the C-terminal 33 residues of NS4A and most of NS4B (Fig. 1).

Coexpression of the NS3 protein (Fig. 3A, lanes 2 and 3) resulted in the partial proteolytic conversion of such a 30-kDa protein into a smaller product showing electrophoretic mobility similar to that of the authentic NS4B (26 kDa [42]). When

NS4A was coexpressed in combination with NS3 (Fig. 3A, lane 4), we could observe a reproducible increase in the efficiency of precursor processing (from -65 to -90%), in addition to the 6-kDa protein encoded by the cotransfected pCite(NS4A) plasmid (Fig. 3A, lane 5). When we transfected plasmid pCite(NS4B/5A), the expected 68-kDa protein product was detected by immunoprecipitation with anti-NS5A antiserum (Fig. 3B, lane 1). This protein contains most of the NS4B and NS5A sequences (Fig. 1). Interestingly, when NS3 was coexpressed with this protein (Fig. 3B, lane 2), the same 68-kDa product was immunoprecipitated by the anti-NS5A antiserum and we found no evidence of proteolytic cleavage (Fig. 3B, lane 3). However, when NS3 was expressed in combination with NS4A (Fig. 3B, lanes 4 and 5), we could observe the occurrence of partial cleavage of the 68-kDa protein and the production of a 46-kDa protein product that was recognized by the anti-NS5A antisera (Fig. 3B, lane 6). It should be pointed out here that our polyclonal anti-NS4 antiserum recognizes NS4A far more efficiently than NS4B. In particular, the portion of NS4B encoded by the pCite(NS4B/5A) construct appears not to be immunoprecipitated, either when fused to NS5A or when cleaved off the precursor (Fig. 3B, lane 5; see also Fig. 4, lane 4, and Fig. 5A, lanes 2 and 5). Last, we analyzed the activity of the NS3 protease on an HCV polyprotein portion containing the cleavage site between NS5A and NS5B. Plasmid pCite(NS5AB) supported the expression of a 115-kDa protein that was specifically recognized by the anti-NS5A antiserum (Fig. 3C, lane 1). When NS3 was

VOL. 68, 1994

coexpressed with this latter protein, which contains the NS5ANS5B junction (Fig. 3C, lane 2), we observed partial proteolytic cleavage and production of a polypeptide with the electrophoretic mobility and immunological properties expected for the mature NS5A (Fig. 3C, lane 3). This finding indicates that the cleavage between NS5A and NS5B can occur in the absence of NS4A. However, when NS4A was coexpressed with NS3 and the NS5A-NS5B precursor (Fig. 3C, lanes 4 and 5), the cleavage efficiency increased substantially, since an NS5ANS5B precursor was no longer observable (Fig. 3C, lane 6). NS4A retains its activity in trans when uncleaved from NS3. In a previous paper, we reported that the mutation of the catalytic serine of NS3 within the full-length HCV polyprotein abrogated processing at the NS3-NS4A, NS4A-NS4B, NS4BNS5A, and NS5A-NS5B sites (42). Wild-type NS3 provided in trans could effectively restore cleavage at the NS4A-NS4B, NS4B-NS5A, and NS5A-NS5B junctions but was unable to cleave between NS3 and NS4A. This finding suggested that this latter cleavage can only occur in cis, i.e., intramolecularly. As a consequence, mature NS4A was not detectable by immunoprecipitation in our trans-complementation experiment. These results are in apparent contradiction with the data presented here: if the NS3 protease needs to be activated by NS4A, how could processing at several sites be restored by complementing the mutated polyprotein with NS3 alone and not with NS4A? A possible explanation of this paradox is that NS4A is still capable of acting as an effector of the NS3 protease in trans, even when uncleaved from the C terminus of a nonfunctional NS3 protein. To test this hypothesis, we carried out the experiments described below. We inserted a mutation in plasmid pCite(NS3/4A) that resulted in the replacement of the catalytic serine of NS3 with an alanine. The mutated construct is called pCite(NS3/4A: S1 65-A) (Fig. 1). When transfected in vTF7-3-infected HeLa cells, this plasmid drives the expression of a 75-kDa polypeptide containing both the mutated NS3 and NS4A (Fig. 4, lanes 3 and 4). We cotransfected pCite(NS3/4A:Sl 165-A) in combination with pCite(NS4B/5A) and with pCite(NS3AC). The latter plasmid (Fig. 1) encodes a C-terminal deletion mutant of NS3 that retains full proteolytic activity (16a). The results of the experiments are shown in Fig. 4. Anti-NS3 antiserum immunoprecipitated two major species, of 75 kDa and 29 kDa (Fig. 4, lane 3), corresponding respectively to the uncleaved NS3-NS4A precursor bearing the mutation in the catalytic serine of NS3 and to the active C-terminal deletion mutant of NS3. As expected, only the former protein was recognized by the anti-NS4 antiserum (Fig. 4, lane 4). Most interestingly, immunoprecipitation with antiNS5A antiserum revealed that the NS4B-NS5A protein had undergone proteolytic cleavage (Fig. 4, lane 5) nearly as efficiently as in the experiment shown in Fig. 3B, lane 6. These results suggest that NS4A does not need to be cleaved off the NS3 protein in order to act as an effector of other NS3 molecules in trans. The C-terminal portion of NS4A is sufficient to act as an effector of NS3 proteolytic activity. The results described thus far indicate that NS4A is indispensable for NS3 cleavage at the NS3-NS4A and NS4B-NS5A sites. In addition, NS4A augments significantly the rate of proteolysis at the NS5A-NS5B cleavage site. Conversely, the rate of cleavage of the protein substrate encoded by the pCite(NS4ABAN) plasmid, which contains the NS4A-NS4B junction but only the 33 C-terminal residues of NS4A, seems to be little affected by the presence of NS4A. There are two possible interpretations of this experimental evidence: either the rate at which the NS4A-NS4B cleavage site of the HCV polyprotein is processed by NS3 does

NS4A, A COFACTOR OF HCV NS3 PROTEASE

NS4B/5A

NS4B/5A NS3AC

MW (kDA)

~

a3 cxSA4

90 _

6 9-

;NS3iC

3757

+

NS34A/Scat EA a3 a4 a54

NS4B/5A

NS34A/Scat

-A

_ NBSA

-w NS5A

45 -

30-

14-

1

2

3 4 5

FIG. 4. NS4A retains its activity when uncleaved from NS3. Plasmid pCite(NS4B/5A) was transfected with pCite(NS3AC) (lanes 1 and 2) or with both pCite(NS3AC) and pCite(NS3/4A:S, 65-A) (lanes 3 to 5). Cell lysates were immunoprecipitated with anti-NS3 (lanes 1 and 3), anti-NS4 (lane 4), and anti NS5A (lanes 2 and 5). 03, a4, and a.5A indicate antisera against NS3, NS4, and NS5A, respectively. The positions of the relevant HCV proteins immunoprecipitated with anti-HCV antibodies are indicated. MW, molecular weight standards.

not depend on the presence of NS4A, or the C-terminal portion of NS4A that is contained in the protein encoded by the pCite(NS4ABAN) plasmid is sufficient to act as an effector of the NS3 proteolytic activity. In order to address this issue, we performed the experiments described below. As described earlier, pCite(NS4B/5A) codes for a protein precursor whose cleavage is strictly dependent on the presence of the NS4A protein. We reasoned that if the C-terminal 33 amino acids of NS4A were indeed sufficient to modulate the proteolytic activity of NS3, then the coexpression of the polypeptide encoded by the pCite(NS4ABAN) plasmid with NS3 should result in the cleavage of the NS4B-NS5A junction. The properties of the protein encoded by pCite(NS4ABAN) as an effector of NS3 were thus compared with those of the polypeptide encoded by plasmid pCite(NS4AB). This latter plasmid contains, in addition to the NS4B region common to both constructs, the entire NS4A coding region (Fig. 1). The results of an experiment in which pCite(NS4B/5A) was cotransfected with both pCite(NS3) and pCite(NS4AB) are shown in Fig. SA, lanes 1 to 3. When the NS3 protein (Fig. 5A, lane 1) is expressed in combination with the polypeptide encoded by pCite(NS4AB), the latter was partially cleaved to yield the two expected fragments of 26 and 6 kDa (Fig. SA, lane 2). Moreover, the NS4B-NS5A protein precursor was also efficiently processed (Fig. SA, lane 3). Lanes 4 to 6 of Fig. 5A show the results of similar cotransfection experiments, but in which pCite(NS4AB) was replaced by pCite(NS4ABzAN). The expression of the NS3 protease (Fig. 5A, lane 4) in combination with the products of the pCite(NS4ABAN) construct (Fig. 5A, lane 5; see also Fig. 3A, lane 3) resulted in the cleavage of the polypeptide containing the NS4B-NS5A junction (Fig. 5A, lane 6). This last result suggests that a domain of NS4A contained within the 33 most C-terminal amino acids of the protein may be sufficient to modulate the activity of the NS3

protease.

3758

J. VIROL.

FAILLA ET AL.

A

NS4B/5A

NS4B/5A

-+ NS3i03NS\A NS4AB a3 a4 mLSA

MW

(kDA) 90

69

S3 + NS3

a5A

-

-

NS3

_ NS4B/5A

NS3

_

_ NS5A

45-

-

-

NS4B

4

NS4B/5A

NSSA

NS4AB

30 -

1

NS4BN NS4ABN

-

-

NS4A

1 2 3

B

1660

1670

4

1680

1690

5

6

1700

1710

Hydrophobic

Kite-

Doolittle

.:

Hydropathv plot

0

Iydrophilic

-, -

-3

FIG. 5. The C-terminal domain of NS4A is sufficient to activate NS3. (A) Plasmid pCite(NS4B/5A) was transfected with pCite(NS3) and with either pCite(NS4AB) (lanes I to 3) or with pCite(NS4ABAN) (lanes 4 to 6). Cell lysates were immunoprecipitated with anti-NS3 (lanes 1 and 4), anti-NS4 (lanes 2 and 5), and anti-NS5A (lanes 3 and 6). c3, a4, and otSA indicate antisera against NS3, NS4, and NS5A, respectively. The positions of the relevant HCV proteins immunoprecipitated with anti-HCV antibodies are indicated. MW, molecular weight standards. (B) Hydropathy plot of NS4A. The average hydrophobicity index was obtained with the algorithm of Kyte and Doolittle with a search length of 8 amino acids (32).

DISCUSSION The evidence presented here shows that, in our transient expression system, the protease domain of HCV NS3 requires the participation of an additional virus-encoded protein, NS4A, in order to efficiently cleave the substrates contained within the HCV polyprotein. NS4A can act as an effector of the NS3 proteolytic activity when supplied in trans, and furthermore a 33-amino-acid domain located at the C terminus of NS4A seems to be sufficient to complement the NS3 activity. The requirement for a two-component protease thus seems to be common to all members of the Flaviviridae (see below). In flaviviruses, NS2B is required to facilitate efficient proteolytic processing of the nonstructural region of the viral polyprotein by NS3 (1, 6, 18, 33, 38). In contrast to HCV NS4A, flavivirus NS2B is located at the N terminus of NS3. In spite of their different topological arrangements, HCV NS4A and flavivirus NS2B share several features: both proteins are required as effectors of protease activity, can function in cis or in trans, and are liberated from their respective catalytic domains by an intramolecular cleavage. A 40-amino-acid segment of NS2B conserved among all flaviviruses has been shown

to be sufficient to trans-complement the NS3 proteolytic activity in dengue (17) and yellow fever (8) viruses. A hydropathy plot of the NS2B protein shows that this 40-amino-acid segment is a hydrophilic domain surrounded by two hydrophobic regions (17). Similarly, the 33-amino-acid region of HCV NS4A, identified as being sufficient for the activation of the NS3 protease, spans a hydrophilic, C-terminal domain and is preceded by a very hydrophobic N-terminal portion (Fig. 5B). The genetic organization and the processing of the nonstructural region of the polyprotein of BVDV, a member of the pestivirus genus, parallels closely that of the corresponding region of HCV: the p80 serine protease, which is the functional homolog of NS3, has been shown by Wiskerchen and Collett (45) to be required for processing of the nonstructural proteins. Cleavage by this protease activity at the N and C termini of p80 itself occurs intramolecularly (in cis). Cleavage at the remaining nonstructural sites can occur in trans. The region corresponding to NS5 of flavivirus is processed, as in HCV, to two distinct polypeptides, p58 and p75. The segment of the polyprotein following p80, equivalent to NS4, generates a small N-terminal hydrophobic protein, piO, and possibly a 32-kDa protein that has not yet been characterized in detail. It is currently thought that p58 and p75 are the functional homologs of HCV NS5A and NS5B, whereas plO and the uncharacterized 32-kDa factor correspond to HCV NS4A and NS4B polypeptides, respectively. Interestingly, p80 protease activity alone was shown to be insufficient for the cleavage of a substrate containing the p58-p75 junction. Conversely, expression of a construct encoding, in addition to the entire p80 protein, plO and part of the 32-kDa protein, led to efficient cleavage of the same site. In view of these data and of the evidence presented here, we suggest that BVDV plO is necessary to complement the p80 serine protease activity and therefore is the functional homolog of HCV NS4A. plO had previously been disregarded as a BVDV protease cofactor (13, 45). This conclusion was based on the observation that processing of a polyprotein that had been mutated in the catalytic serine of p80 was rescued at several sites by trans-complementation with a protein containing an intact protease domain. In this experiment, plO was not generated, but the corresponding sequence remained uncleaved from the mutant p80 (45); this finding indicated that the p80-plO cleavage can only occur in cis and was taken as an argument to rule out the involvement of plO as a possible p80 cofactor required for cleavage at the remaining sites. However, it has emerged from the experiment described in Fig. 4 that HCV NS4A can complement the NS3 protease activity in trans even when uncleaved from an NS3 molecule that has been mutated in the catalytic serine. It is therefore also likely, by analogy, that BVDV plO does not need to be released from a mutated p80 in order to act in trans as an effector of functional protease molecules.

The mechanism by which HCV NS4A activates the NS3contained serine protease is not yet clear. The two proteins could associate in a complex to form an active heterodimeric protease. Indeed, a stable complex between NS3 and a protein containing NS4A and part of NS4B has been observed in vitro (25). A similar complex has been observed in flaviviruses between NS3 and NS2B (1, 8). A related possibility is that NS4A is required to target NS3 to the correct cellular compartment for maximal protease activity. Alternatively, NS4A could act as a virus-encoded chaperonin aiding the correct folding of the protease domain (14). In this regard, it is of interest that several bacterial serine proteases, including ax-lytic protease and subtilisin, possess propeptides that share several properties with HCV NS4A and flavivirus NS2B: they are necessary for protease activity, can activate the protease either

NS4A, A COFACTOR OF HCV NS3 PROTEASE

VOL. 68, 1994

in cis or in trans, and are cleaved off the protease domain by intramolecular proteolytic reactions (26, 40, 47). These propeptides are required to assist the folding of their protease counterpart, and when folding has been completed the propeptide becomes dispensable and no longer required for proteolytic activity (2, 47). Alternatively, NS4A may act as a chaperonin that facilitates the interaction of NS3 with the substrate, for example by locally unfolding the polyprotein precursor, in order to make the cleavage site accessible to the protease. In other viral systems, examples in which dimerization is required for protease activity (30) or in which the protease activity is modulated by interaction with additional polypeptides exist. In cowpea mosaic virus, the specificity of the 24-kDa protease for trans cleavage of the capsid precursor is altered by the virus-encoded 32-kDa protein (43). The 3C and 3CD proteases encoded by poliovirus differ in cleavage specificity for capsid precursor, the specificity being determined by the interaction of the 3D region with the substrate (28, 46). Moreover, the adenovirus cysteine protease has been recently shown to be activated via the formation of a disulfide bridge with an 11-amino-acid peptide derived from the viral protein pVI (44). Further studies on the structures and the functions of both HCV NS3 and NS4A, as well as of their interactions with the other components of the viral polyprotein, are required to solve the issue of how the virus-encoded serine protease is activated by a second viral polypeptide. These studies will eventually shed light on the temporal and spatial regulation of HCV gene expression. ACKNOWLEDGMENTS

We thank R. Cortese, J. Jiricni, and N. La Monica for critical review, and we thank all IRBM coworkers for helpful discussion. We also thank P. Neuner for oligonucleotide synthesis and Y. Cully for graphics. REFERENCES

1. Arias, C. F., F. Preugschat, and J. H. Strauss. 1993. Dengue 2 virus NS2B and NS3 form a stable complex that can cleave NS3 within the helicase domain. Virology 193:888-899. 2. Baker, D., J. L. Sohl, and D. A. Agard. 1992. A protein-folding reaction under kinetic control. Nature (London) 356:263-265. 3. Bartenschlager, R., L. Ahlborn-Laake, J. Mous, and H. Jacobsen. 1993. Nonstructural protein 3 of the hepatitis C virus encodes a serine-type proteinase required for cleavage at the NS3/4 and NS4/5 junctions. J. Virol. 67:3835-3844. 4. Bazan, J. F., and R. J. Fletterick. 1989. Detection of a trypsin-like serine protease domain in flavivirus and pestivirus. Virology 171:637-639. 5. Berger, J., J. Hauber, R. Hauber, R. Geiger, and B. Cullen. 1988. Secreted placental alkaline phosphatase: a powerful new quantitative indicator of gene expression in eukaryotic cells. Gene 66:1-10. 6. Chambers, T. J., A. Grakoui, and C. M. Rice. 1991. Processing of the yellow fever nonstructural polyprotein: a catalytically active NS3 proteinase domain and NS2B are required for cleavages at dibasic sites. J. Virol. 65:6042-6050. 7. Chambers, T. J., C. Hahn, R. Galler, and C. M. Rice. 1990. Flavivirus genome organization, expression, and replication. Annu. Rev. Microbiol. 44:649-688. 8. Chambers, T. J., A. Nestorowicz, S. M. Amberg, and C. M. Rice. 1993. Mutagenesis of the yellow fever virus NS2B protein: effects on proteolytic processing, NS2B-NS3 complex formation, and viral replication. J. Virol. 67:6797-6807. 9. Chambers, T. J., R. C. Weir, A. Grakoui, D. W. McCourt, F. F. Bazan, R. J. Fletterick, and C. M. Rice. 1990. Evidence that the N-terminal domain of nonstructural protein NS3 from yellow fever virus is a serine protease responsible for site-specific cleavages in the viral polyprotein. Proc. Natl. Acad. Sci. USA 87:8898-8902.

3759

10. Chien, D., Q.-L. Choo, A. Tabrizi, C. Kuo, J. McFarland, K. Berger, C. Lee, J. Shuster, T. Nguyen, D. Moyer, M. M. Tong, S. Furuta, M. Omata, G. Tegtmeyer, H. Alter, E. Schiff, L. Jeffers, M. Houghton, and G. Kuo. 1992. Diagnosis of hepatitis C virus (HCV) infection using an immunodominant chimeric polyprotein to capture circulating antibodies: reevaluation of the role of HCV in liver disease. Proc. Natl. Acad. Sci. USA 89:10011-10015. 11. Choo, Q.-L., G. Kuo, A. J. Weiner, L. R. D. W. Bradley, and M. Houghton. 1989. Isolation of a cDNA clone derived from a blood-borne non-A non-B viral hepatitis genome. Science 244: 359-362. 12. Choo, Q.-I., K. H. Richman, J. H. Han, K. Berger, C. Lee, C. Dong, C. Gallegos, D. Coit, A. Medina-Selby, P. J. Barr, A. J. Weiner, D. W. Bradley, G. Kuo, and M. Houghton. 1991. Genetic organization and diversity of the hepatitis C virus. Proc. Natl. Acad. Sci. USA 88:2451-2455. 13. Collett, M. 1992. Molecular genetics of pestiviruses. Comp. Immunol. Microbiol. Infect. Dis. 15:145-154. 14. Craig, E. 1993. Chaperones: helpers along the pathways to protein folding. Science 260:1902-1904. 15. Eckart, M. R., M. Selby, F. Masiarz, C. Lee, K. Berger, K. Crawford, C. Kuo, G. Kuo, M. Houghton, and Q.-L. Choo. 1993. The hepatitis C virus encodes a serine protease involved in processing of the putative nonstructural proteins from the viral polyprotein precursor. Biochem. Biophys. Res. Commun. 192: 399-406. 16. Elroy-Stein, O., T. R. Fuerst, and B. Moss. 1989. Cap-independent translation of mRNA conferred by encephalomyocarditis virus 5' sequence improves the performance of the vaccinia/bacteriophage T7 hybrid expression system. Proc. Natl. Acad. Sci. USA 86:61266130. 16a.Failla, C., and R. De Francesco. Unpublished data. 17. Falgout, B., R. H. Miller, and C. Lai. 1993. Deletion analysis of dengue virus type 4 nonstructural protein NS2B: identification of a domain required for NS2B-NS3 proteinase activity. J. Virol. 67:2034-2042. 18. Falgout, B., M. Pethel, Y.-M. Zhang, and C.-J. Lai. 1991. Both nonstructural proteins NS2B and NS3 are required for the proteolytic processing of dengue virus nonstructural proteins. J. Virol. 65:2467-2475. 19. Francki, R. I., C. M. Fauquet, D. L. Knudson, and F. Brown. 1991. Classification and nomenclature of viruses. Arch. Virol. Suppl. 2:223-233. 20. Fuerst, T. R., E. G. Niles, F. W. Studier, and B. Moss. 1986. Eukaryotic transient-expression system based on recombinant vaccinia virus that synthesizes bacteriophage T7 RNA polymerase. Proc. Natl. Acad. Sci. USA 83:8122-8126. 21. Grakoui, A., D. W. McCourt, C. Wychowski, S. M. Feinstone, and C. M. Rice. 1993. A second hepatitis C virus-encoded proteinase.

Proc. Natl. Acad. Sci. USA 90:10583-10587. 22. Grakoui, A., D. W. McCourt, C. Wychowski, S. M. Feinstone, and C. M. Rice. 1993. Characterization of the hepatitis C virusencoded serine proteinase: determination of proteinase-dependent polyprotein cleavage sites. J. Virol. 67:2832-2843. 23. Grakoui, A., C. Wychowski, C. Lin, S. M. Feinstone, and C. M.

Rice. 1993. Expression and identification of hepatitis C virus polyprotein cleavage products. J. Virol. 67:1385-1395. 24. Hijikata, M., N. Kato, Y. Ootusyama, M. Nakagawa, and K. Shimotohno. 1991. Gene mapping of the putative structural region

of the hepatitis C virus genome by in vitro processing analysis.

Proc. Natl. Acad. Sci. USA 88:5547-5551. 25. Hijikata, M., H. Mizushima, T. Akagi, S. Mori, N. Kakiuchi, N. Kato, T. Tanaka, K. Kimura, and K. Shimotohno. 1993. Two

distinct proteinase activities required for the processing of a putative nonstructural precursor protein of hepatitis C virus. J. Virol. 67:4665-4675. 26. Ikemura, I., and M. Inouye. 1988. In vitro processing of prosubtilisin in Eschenichia coli. J. Biol. Chem. 263:12959-12963. 27. Jang, S. K., M. V. Davies, R. J. Kaufman, and E. Wimmer. 1989.

Initiation of protein synthesis by internal entry of ribosomes into the 5' nontranslated region of encephalomyocarditis virus RNA in

vivo. J. Virol. 63:1651-1660. 28. Jore, J., B. D. Geus, R. J. Jackson, P. H. Pouwels, and B. E.

3760

FAILLA ET AL.

Enger-Valk. 1988. Poliovirus protein 3CD is the active protease for processing of the precursor protein Pl in vitro. J. Gen. Virol. 69:1627-1636. 29. Kato, M., M. Hijikata, Y. Ootsuyama, M. Nakagawa, S. Ohkoshi, T. Sugimura, and K. Shimotohno. 1990. Molecular cloning of human hepatitis C virus genome from Japanese patients with non-A non-B hepatitis. Proc. Natl. Acad. Sci. USA 87:9524-9528. 30. Katoh, I., Y. Ikawa, and Y. Yoshinaka. 1989. Retrovirus proteinase characterized as a dimeric aspartic proteinase. J. Virol. 63:22262232. 31. Kuo, G., Q.-L. Choo, H. J. Alter, G. L. Gitnick, A. G. Redecker, R. H. Purcell, T. Myamura, J. L. Dienstag, M. J. Alter, C. E. Syevens, G. E. Tagtmeier, F. Bonino, M. Colombo, W.-S. Lee, C. Kuo, K. Berger, J. R. Shister, L. R. Overby, D. W. Bradley, and M. Houghton. 1989. An assay for circulating antibodies to a major etiologic virus of human non-A non-B hepatitis. Science 244:362364. 32. Kyte, J., and R. F. Doolittle. 1982. A single method for displaying the hydropathic character of a protein. J. Mol. Biol. 157:105-132. 33. Lin, C., S. M. Amberg, T. J. Chambers, and C. M. Rice. 1993. Cleavage at a novel site in the NS4A region by the yellow fever virus NS2B-3 proteinase is a prerequisite for processing at the downstream 4A/4B signalase site. J. Virol. 67:2327-2335. 34. Matsuura, Y., and T. Miyamura. 1993. The molecular biology of hepatitis C virus. Semin. Virol. 4:297-304. 35. Miller, R. H., and R. H. Purcell. 1990. Hepatitis C virus shares amino acid sequence similarity with pestiviruses and flaviviruses as well as members of two plantvirus subgroups. Proc. Natl. Acad. Sci. USA 87:2057-2061. 36. Phillips, M. A., and R. J. Fletterick. 1992. Proteases. Curr. Opin. Struct. Biol. 2:713-720. 37. Pizzi, E., A. Tramontano, L. Tomei, N. La Monica, C. Failla, M. Sardana, T. Wood, and R. De Francesco. 1994. Molecular model of the specificity pocket of the hepatitis C virus protease: implications for substrate recognition. Proc. Natl. Acad. Sci. USA 91:888-892.

J. VIROL.

38. Preugschat, F., C.-W. Yao, and J. H. Strauss. 1990. In vitro processing of dengue virus type 2 nonstructural proteins NS2A, NS2B, and NS3. J. Virol. 64:4364-4374. 39. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. 39a.Santolini, E., G. Migliacchio, and N. La Monica. 1994. Biosynthesis and biochemical properties of the hepatitis C virus core protein. J. Virol. 68:3631-3641. 40. Silen, J. L., and D. A. Agard. 1989. The lytic protease pro-region does not require a physical linkage to activate the protease domain in vivo. Nature (London) 341:462-464. 41. Takamizawa, A., C. Mori, I. Fuke, S. Manabe, S. Murakami, J. Fujita, E. Onoshi, T. Andoh, I. Yoshida, and H. Okayama. 1991. Structure and organization of the hepatitis C virus genome isolated from human carriers. J. Virol. 65:1105-1113. 42. Tomei, L., C. Failla, E. Santolini, R. De Francesco, and N. La Monica. 1993. NS3 is a serine protease required for processing of hepatitis C virus polyprotein. J. Virol. 67:4017-4026. 43. Vos, P., J. Verver, M. Jaegle, J. Wellink, A. V. Kammen, and R. Goldbach. 1988. Two viral proteins involved in the proteolytic processing of the cowpea mosaic virus polyproteins. Nucleic Acids Res. 15:1967-1985. 44. Webster, A., R. T. Hay, and G. Kemp. 1993. The adenovirus protease is activated by a virus-coded disulphide-linked peptide. Cell 72:97-104. 45. Wiskerchen, M., and M. S. Collett. 1991. Pestivirus gene expression: protein p80 of bovine viral diarrhea virus is a proteinase involved in polyprotein processing. Virology 184:341-350. 46. Ypma-Wong, M. F., P. G. Dewalt, V. H. Johnson, J. G. Lamb, and B. L. Semler. 1988. Protein 3CD is the major poliovirus proteinase responsible for cleavage of the P1 capsid precursor. Virology 166:265-270. 47. Zhu, X., Y. Ohta, F. Jordan, and M. Inouye. 1989. Pro-sequence of subtilisin can guide the refolding of denatured subtilisin in an intermolecular process. Nature (London) 339:483-484.