key: cord-0689238-j32352te
authors: Lu, Xiaotao; Lu, Yiqi; Denison, Mark R.
title: Intracellular andin Vitro-Translated 27-kDa Proteins Contain the 3C-like Proteinase Activity of the Coronavirus MHV-A59
date: 1996-08-15
journal: Virology
DOI: 10.1006/viro.1996.0434
sha: 3bf6e190b007abebec86eeb942d87340d72eeb59
doc_id: 689238
cord_uid: j32352te

Abstract The coronavirus mouse hepatitis virus-A59 (MHV-A59) encodes a serine-like proteinase (3C-like proteinase or 3CLpro) in ORF 1a of gene 1 between nucleotides 10209 and 11114. We previously have demonstrated that proteins expressedin vitrofrom a cDNA clone of the 3CLpro region possess proteinase activity, and that the proteinase is able to cleave substratein trans.We sought to determine if the 27-kDain vitrocleavage product (p27) was an active form of the 3CLpro and whether this was consistent with the 3CLpro expressed in virus-infected cells. Antibodies directed against the 3CLpro domain detected 27-kDa MHV proteinsin vitroand in MHV-A59-infected cells. The 27-kDa proteins were able to cleave substratein transwithout other protein cofactors or supplemental membranes, and the p27 proteinase activity was retained after purification by immunoprecipitation and gel electrophoresis. When p27 was expressedin vitrowith portions of the amino- and carboxy-terminal flanking domains (MP1 and MP2), p27 was not liberated byciscleavage. The proteolytic activity of the 27-kDa proteins was inhibited by a variety of cysteine and serine proteinase inhibitors, and was eliminated by the cysteine proteinase inhibitor E64d. These results indicate that the 27-kDa protein is a mature proteinase in MHV-A59-infected cells, and that appropriate processing of this molecule occursin vitro.

of p28 (Baker et al., 1993; Dong and Baker, 1994; Hughes et al., 1995) . The papain-like proteinase has The coronavirus mouse hepatitis virus-A59 (MHVnot been identified in MHV-infected cells; however, A59) contains a 32-kb single-stranded RNA genome based on immune precipitation of MHV proteins from of positive polarity. Intracellular replication of MHV is infected cells using specific antisera, it is likely that initiated by translation of gene 1 of the genome RNA this activity is contained in either the 50-or the 240into a polyprotein with an estimated mass of greater kDa MHV gene 1 proteins (Denison et al., 1992 ; Bonilla than 750 kDa. The gene 1 polyprotein is a fusion proet al., 1995) . tein that is assumed to be the result of translation We have identified a second MHV proteinase that through two out-of-frame open reading frames, ORF is encoded between nucleotides 10209 and 11114 1a and ORF 1b, by a ribosomal frameshift (Pachuk et from the 5 end of gene 1 (Fig. 1) . In vitro translation al., 1989; Lee et al., 1991) . The gene 1 polyprotein is of a cDNA clone encoding the predicted MHV-A59 3Cthought to incorporate all the activities necessary for like proteinase (3CLpro) and portions of the flanking MHV RNA replication, including three proteinases that domains (MP1 and MP2) results in expression of a are predicted to mediate all the maturational cleavproteinase and liberates a 27-kDa protein (p27) (Lu et ages of the polyprotein (Lee et al., 1991; Gorbalenya al., 1995) . p27 is cleaved in trans at an amino-terminal et al., 1991; Gorbalenya and Koonin, 1993) . It is known glutamine -serine (QS 3333-4 ) dipeptide. Although mutathat proteolytic processing of the gene 1 polyprotein genesis of predicted catalytic cysteine 3478 or histiis required for MHV replication, and that inhibition of dine 3374 residues within the 3CLpro domain abolished proteolytic activity also results in rapid shutoff of MHV proteolytic activity of proteins expressed in vitro, it RNA synthesis (Kim et al., 1995) .

was not known if the 27-kDa protein was the active Three proteinases have been predicted to be enproteinase. coded within the MHV gene 1; two of the proteinases In this study we sought to determine if the active have been experimentally confirmed by in vitro expres-3CLpro in vitro and in virus-infected cells was the 27sion and activity. A papain-like proteinase, encoded kDa protein previously identified in vitro. Antisera spebetween nucleotides 3559 and 4049 from the 5 end cific to the 3CLpro domain detected 27-kDa (p27) proof gene 1, is known to cleave at the carboxy terminus teins both in vitro and in cells. Gel-purified p27 proteins from cells or in vitro translation reactions were able to cleave substrate in trans, whereas the in vitro-p.i. During proteinase-inhibition experiments, proteinase inhibitors were added to the media 1 hr before the addition of radiolabel. During pulse-label and pulse-chase experiments, synchronization of translation was achieved by the addition of NaCl (200 mM) to the media for 30 min at 5.5 hr p.i. (Denison et al., 1992) . The high-salt medium was replaced with isotonic medium containing [ 35 S]Met, and polypeptides were labeled for various times from 15 to 120 min. The infected monolayers were then harvested or alternatively chased with excess cold methionine for an additional 90 min.

We raised antisera directed against the MHV 3CLpro by immunizing rabbits with a keyhole limpet hemocyanin (KLH)-linked oligopeptide or a fusion protein within the 3CLpro region of gene 1 (Fig. 1) . A synthetic 19-aminoacid oligopeptide ( 3426 -LQNPNTPKYSFGVVKPGET-3444 ) was dissolved in phosphate-buffered saline and coupled to KLH (Sigma) using glutaraldehyde. The fusion protein spanned residues 3239 to 3476 of the ORF 1a polypro- tein, which included 95 amino acids upstream of the antibodies. (A) The MP-1/3CLpro/MP-2 coding region is shown both 3CLpro amino-terminal serine and extended to within 2 by nucleotide (nt) number from the 5 end of the genome, as well as amino acids of the essential cysteine 3478 . The plasmid by amino acid residue number from the amino terminus of the gene 1 pGEM-K/B was constructed by subcloning a 713-nucleopolyprotein, after the sequence of Bonilla et al. (1994) . Numbers repretide KpnI-BamHI restriction fragment of ORF 1a from nt sent amino acid residue number unless preceded by ''nt.'' The 3CLproencoding construct pGpro encodes residues 3239-3687. The B3 anti-9922 to 10635 into pGEM 3Zf 0 . A 735-nucleotide KpnIserum was induced against the fusion protein spanning residues 3239-PstI restriction fragment was isolated from pGEM, ligated 3476. The SP9 antiserum was raised against a synthetic peptide as to KpnI-PstI-digested pQE-30 plasmid (Qiagen) and

shown. Locations of essential His and Cys residues are shown, as used to transform Escherichia coli M15(pREP4) cells. instructions. Antisera directed against the oligopeptide and fusion protein were obtained from Cocalico. The antisera were named Sp9 (oligopeptide) and B3 (fusion pro-The proteolytic activity of p27 was inhibited by proteintein). ase inhibitors, with complete inhibition occurring with the cysteine proteinase inhibitor E64d. We were there-Immunoprecipitation and electrophoresis fore able to confirm that the p27 protein detected in virus-infected cells and in vitro was an active form of Immunoprecipitation (IP) of MHV proteins was perthe 3C-like proteinase.

formed as previously described (Denison et al., 1992) . Briefly, whole-cell lysates of MHV-A59-infected DBT cells MATERIALS AND METHODS (equivalent to 10 6 cells) or in vitro translation products (5 1 10 6 TCA-precipitable counts) were diluted into 1 ml of Infection of DBT cells with MHV-A59 radioimmunoprecipitation assay (RIPA) buffer to obtain a final concentration of SDS of 0.1%. Polyclonal antisera, DBT cells were infected with MHV-A59 at a multiplicity of infection (m.o.i.) of 20 PFU/cell in DMEM 2% FCS (Deni-4 ml of PMSF (phenylmethylsulfonyl fluoride; Sigma) (20 mg/ml), and 30 ml of preswollen protein A-Sepharose son et al., 1992) . Methionine-free DMEM (Gibco) containing 2% FCS and 10 mg/ml actinomycin D (Sigma) were beads (Sigma) were added directly to the reaction mix. After rocking for 2 hr at 4Њ, the supernatant was removed added to the cells at 3 hr postinfection (p.i.). Intracellular proteins were labeled from 6 to 8 hr p.i. with 200 mCi/ml and protein A bead/antibody/antigen complexes were washed four times with alternating high (1 M NaCl) and [ 35 S]methionine (Expre 35 S 35 S protein labeling mix; Du-Pont, NEN), unless otherwise indicated. During experi-low (100 mM NaCl) salt RIPA buffer. The beads were boiled in 50 ml of 21 Laemmli buffer for 5 min before ments to determine the time of appearance of p27, proteins were labeled for 2-hr intervals between 2 and 12 hr electrophoresis on a 5-18% gradient SDS-polyacryl-amide gel (Laemmli, 1970) . Radiolabeled proteins were visualized by fluorography.

Recombinant plasmids were transcribed and translated in vitro, as previously described (Lu et al., 1995) . Standard reactions were performed in a total volume of 25 ml, with 0.5 mg of plasmid DNA and 20 mCi of [ 35 S]Met (Amersham) for 90 min at 30Њ. Reactions were terminated by addition of 21 Laemmli buffer or quick freezing in dry ice/ethanol and storage at 070Њ.

DBT cells were infected with MHV-A59 at an m.o.i. of 20 PFU/cell. At 6 to 8 hr postinfection, nonradiolabeled whole-cell lysates (4 1 10 6 cells) were immunoprecipitated with antiserum Sp9 or preimmune serum from the same rabbit. The pGpro construct, which expresses an active proteinase, was translated in vitro without radiolabel for 1 hr, and translation products were immunoprecipitated with antiserum Sp9. Immunoprecipitation products were separated on a 5-18% gradient SDS-polyacrylamide gel that was prerun with 0.1 mM thioglycolate in the upper running buffer (100 V, 15 min). The 27-and 32-kDa bands from pGpro translation were located by comparison with prestained protein markers and radiolabeled pGpro translation products, and were excised from the gel. The 27-kDa bands from IP of infected cell lysates inhibited, proteinase inhibitors such as 2 mM PMSF, 100 mg/ml E64c (Sigma), or 1 mM ZnCl 2 (zinc chloride; Sigma) were added into the mixtures containing radiolabeled additional 120 min at 30Њ. Diluted lysates were concen-pGproG41 and cold 27-kDa protein.

trated using a Centricon-10 concentrator (Amicon, Inc.). Concentrated samples were electrophoresed on 5-18% gradient SDS-polyacrylamide gels, the gel was treated Cis cleavage assay with DMSO/PPO, and the proteins were visualized by fluorography. pGpro was expressed in the presence of [ 35 S]Met for 40 min in the combined transcription/translation reaction.

Identification of 27-kDa proteins in vitro and in MHVbuffer (50 mM Tris-acetate, pH 7.5, 5 mM magnesium A59-infected cells acetate, 5 mM dithiothreitol, 0.2 mM disodium ethylenediamine tetraacetate, and 100 mM potassium acetate) in Antisera Sp9 and B3 were used to immunoprecipitate MHV-A59 gene 1 proteins from in vitro translation prod-serial dilutions from 1:2 to 1:320, and incubated for an ucts of pGpro and from MHV-infected cells. Preimmune with excess unlabeled methionine for an additional 90 min (variable pulse/constant chase). Infected and mock-sera, uninfected cell lysates, and antibodies directed against other regions of the gene 1 polyprotein were infected cell lysates were immunoprecipitated with antiserum Sp9. A third set of infected cell monolayers was used as controls for specificity. Major processed polypeptides of 32 and 27 kDa were expressed from pGpro radiolabeled for 60 min, and translation was terminated with cycloheximide followed by additional incubation for in vitro, as previously described (Lu et al., 1995) . Both the 32-and the 27-kDa proteins were detected by B3 up to 3 hr (pulse-chase). The variable pulse/constant chase experiment (Fig. and Sp9 antisera (Fig. 2) . Both antisera also precipitated products of greater than 200 kDa that were not detected 3B) demonstrated that radiolabel was incorporated into p27 between 60 and 75 min after initiation of translation. by preimmune sera. Since the maximum possible size of the expression construct was 48 kDa, these bands

The pulse-label (A) showed that p27 was cleaved between 60 and 75 min after addition of radiolabel. Densito-probably represented aggregates. No proteins were detected by the unrelated B4 antiserum. Thus it appeared metric analysis of the bands showed no significant difference between the amount of p27 detected in the pulse-that both B3 and Sp9 were specific for the 3CLpro region and were able to detect p27.

label or pulse-chase translations at 60 and 75 min (data not shown), indicating that the time of incorporation of From infected cell lysates, both antisera precipitated a virus-specific protein that had the same apparent mass radiolabel into the polypeptide and the time of cleavage were closely spaced. Thus p27 appeared to be cleaved (27 kDa) as the in vitro cleavage product. The intracellular p27 was not detected in mock-infected cell lysates, nor soon after synthesis in infected cells. In contrast, the cleavage of p27 during in vitro translation of the pGpro was it precipitated from infected cells by preimmune serum. Multiple larger polypeptides were precipitated by construct has been shown to be delayed for 30 to 60 min after incorporation of radiolabel (Lu et al., 1995) . The both B3 and Sp9; however, p27 was the only protein uniquely detected by immune sera in infected cells. The reason for the difference in the rates of cleavage in vitro and in cells is not known. results demonstrated that p27 was a specific product of the 3CLpro domain both in vitro and in virus-infected

In the pulse-chase experiment (Fig. 3C) , after p27 was cleaved, the amount detected by immunoprecipitation re-cells.

We next sought to determine the kinetics of p27 syn-mained constant for at least 2 hr. Since it appeared that p27 was rapidly processed after synthesis, this sug-thesis in virus-infected cells. Our studies of other MHV gene 1 products have shown that they are first detected gested that p27 was relatively resistant to degradation in infected cells. at 4 to 6 hr postinfection and are present in larger amounts as the infection progresses, suggestive of accu-We could not identify precursors to p27 in any of these experiments. It is possible that p27 was rapidly cleaved mulation or increasing rates of synthesis due to amplification of genome late in infection. The B3 antiserum was from the growing polyprotein with no intermediate precursors. However, both Sp9 and B3 precipitated a large used to immunoprecipitate whole-cell lysates of infected and mock-infected cells labeled with [ 35 S]Met at different number of other proteins that may have obscured p27containing precursors. We are currently using the Sp9 times p.i. p27 was first detected at 6-8 hr postinfection, (Fig. 2) . This pattern was consistent with the earliest peptide to obtain mAbs to the 3CLpro domain, that can be used to further investigate possible p27 precursors. appearance of all other identified MHV gene 1 proteins (Denison et al., 1992) . The p27 protein from lysates of 1 1 10 5 cells was easily detected by fluorography of SDS Inhibition of p27 processing in cells gels within 12 hr. An additional faint band of approximately 27 kDa was visible during every labeling period We next investigated whether p27 processing in cells could be blocked by proteinase inhibitors. Infected DBT from 0 to 12 hr, from B3-precipitated, mock-infected cell lysates and in infected cells precipitated by preimmune cells were radiolabeled in the presence of leupeptin, PMSF, E64d, N-ethylmaleimide (NEM), or ZnCl 2 , and cell lysates serum, suggesting that it was probably a cell protein that comigrated with the MHV p27.

were immunoprecipitated with Sp9 antiserum. Both serine and cysteine proteinase inhibitors affected proteolytic processing of p27 in DBT cells during MHV infection (Fig. 4) . p27 processing and stability in MHV-infected cells Leupeptin, PMSF, and NEM all consistently inhibited p27 processing between 50 and 85%, whereas E64d inhibited We next sought to determine whether p27-containing precursors could be identified, and to assess the stability processing 100%. E64d has been previously shown to inhibit MHV RNA synthesis and virus replication in DBT cells at of p27 following cleavage. Intracellular gene 1 translation was synchronized by pretreatment of cells with hyper-the same concentration used in this experiment (Kim et al.,  1995) . There was no detectable inhibition of p27 processing tonic NaCl, and proteins were radiolabeled for intervals from 15 to 120 min (Fig. 3) . At each time point one plate in the presence of 0.1 mM ZnCl 2 . This concentration was the highest that could be used without cellular cytotoxicity, was harvested (pulse-label), and another was chased Molecular mass markers are shown to the right and the location of p27 is indicated next to each gel. (C) Pulse-chase: following synchronization, proteins were radiolabeled for 60 min. Incorporation was terminated by addition of cycloheximide and excess cold methionine, and samples were chased for the time in minutes shown above the gel. Proteins were immunoprecipitated with Sp9 antiserum and separated on a 5-18% SDS gel. The location of p27 and the location of marker proteins are shown. Inf, infected cells; pre, Sp9 preimmune rabbit serum; Mock, mock-infected cell lysates.

but was much lower than the concentrations used (1-2 proteolytic activity. Nonradiolabeled p27 proteins from in vitro translation of the construct expressing active pro-mM) to inhibit MHV gene 1 proteinases in vitro (Denison et teinase (pGpro) and from virus-infected cells were exal., 1992) . No p27 precursors were detected in the presence cised from the SDS-PAGE gels and incubated with IVT of these inhibitors.

lysates containing radiolabled proteins translated from the pGproG41 mutant (His41-Gly). PGproG41 lacks pro-Trans proteolytic activity of p27 from IVT and infected teinase activity but is still a cleavage substrate for the cells trans active 3CLpro in vitro, resulting in the appearance The MHV 3CLpro has been predicted to include amino of new p27 molecules (Lu et al., 1995) . The pGproG41 acids from S 3334 to Q 3635 , with an estimated mass of 34 translation reactions were terminated by the addition of kDa (Lee et al., 1991) . In contrast, we have consistently RNAse, cycloheximide, and excess cold methionine bedetected proteins with an estimated mass of 27 kDa, fore incubation with the p27-containing gel slices. As both from infected cells and from in vitro translation (IVT) controls, pGproG41 translation lysates were incubated products of pGpro; we have also shown that the amino with gel fragments lacking protein bands, as well as with terminus of in vitro-translated p27 is the QS dipeptide gel slices containing 27-kDa region proteins precipitated predicted to be the cleavage site of the 3CLpro. Therefrom MHV-infected cells by preimmune sera. fore, we sought to determine whether the 27-kDa pro-

The results obtained during in vitro translation of teins we identified in cells and in vitro were an active pGpro and pGproG41 were consistent with our previous form of the 3CLpro. study (Fig. 5) (Lu et al., 1995) : pGpro translation resulted in processing of a prominent 27-kDa band that was ab-We used an in vitro trans cleavage assay to assess p27 gel slices containing the p27 proteinase before addition of substrate, processing was decreased or abolished (lanes 6-11). PMSF completely inhibited p27 cleavage activity, compared with the partial inhibition seen in virusinfected cells. E64c completely inhibited cleavage activity of p27, similar to the result obtained in virus-infected cells. E64c is the active form of E64 to which E64d is converted on entry into cells. The lowest concentration of E64c used (100 mg/ml or 300 mM) was 25% of that required to inhibit cleavage of p27 in cells, indicating that the proteinase was more sensitive to inhibition by E64 than previously thought. Zinc chloride inhibited proteolytic activity of p27 isolated from virus-infected cells, but did not inhibit activity of p27 isolated from in vitro translation products of pGpro. The reason for this was not clear, but was compatible with the less consistent results of ZnCl 2 inhibition seen in virus-infected cells. drophobic domains, as previously described (Lu et al., 1995) . Following 40 min of translation in vitro, the translasent from the translation reaction containing the tion lysates were diluted from 2-to 320-fold with dilution pGproG41 mutant construct alone (lane 1). However, buffer and incubated an additional 2 hr (Fig. 6) . Cleavage when the pGproG41 translation products were incubated of p27 was observed only in the undiluted and 2-foldwith the gel-purified MHV p27 proteins immunoprecipidiluted samples, demonstrating that cis cleavage did not tated from cells or from IVT lysates containing pGpro occur in the in vitro system using a partial construct of translation products, new molecules of radiolabeled p27 gene 1. It is still possible that cis cleavage may be a were generated (lanes 3 and 4) . Since only the pGproG41 mechanism of cleavage of the 3CLpro early in infection, translation products were labeled, the new p27 molecules were the result of trans cleavage by the nonlabeled p27 proteins in the gel slices. In contrast, when pGproG41 IVT products were incubated with gel fragments containing the 27-kDa region proteins obtained by immunoprecipitation of infected cells by preimmune serum, no cleavage of p27 was observed (lane 2). Together these results demonstrated that the proteolytic activity was specific to the 27-kDa protein precipitated When proteinase inhibitors were incubated with the active proteinases. In addition, purification of the 27-kDa proteins by immunoprecipitation and SDS-polyacrylamide gel electrophoresis did not eliminate their ability to process other molecules of p27 in trans. Cellular proteinases involved in hemostasis have been shown to retain or regain activity after SDS gel electrophoresis FIG. 6 . Dilution analysis of p27 proteinase activity. pGpro was trans- (Wagner et al., 1985; Tans et al., 1989) . However, there is lated as described under Materials and Methods. Translation was a paucity of analogous studies with the viral proteinases. halted at 40 min and lysates were diluted as shown above the figure.

Proteinase activities from hepatitis C virus and yellow Undil., undiluted lysate; 320, 320-fold dilution. Samples were incubated fever virus have been demonstrated in cell lysates.

an additional 120 min prior to electrophoresis. The location of p27 is shown to the left and the location of protein markers to the right. Amberg et al. reported trans cleavage activity of the yellow fever virus NS2B-3 in lysates of virus-infected BHK cells, using an in vitro-translated substrate (Amberg et but these results suggest that trans cleavage is the major al., 1994). The trans cleavage activity of the hepatitis C mechanism by which the polyprotein is processed by the virus NS3 serine proteinase has been demonstrated in 3CLpro.

lysates of HeLa cells. HCV NS3 was expressed in HeLa cells using a recombinant vaccinia virus system, and DISCUSSION cell lysates were incubated with an in vitro-translated substrate (Bouffard et al., 1995) . However, the proteinase We have identified the active 3C-like proteinase molecules expressed from MHV-A59 ORF 1a in vitro and in activity was not purified from the cell lysates in either of these studies. Our approach allowed us to demonstrate virus-infected cells. The proteins have identical apparent masses of 27 kDa, smaller than either the predicted MHV the proteolytic activity of a specific purified protein, something that has not previously been accomplished with a 3CLpro (33 kDa) or the identified human coronavirus 229E 3CLpro (34 kDa). The MHV 3CLpro has been pre-coronavirus proteinase. The p27 protein precipitated from MHV-infected DBT dicted to span amino acids from S 3334 to Q 3635 , and to have a mass of 34 kDa (Gorbalenya et al., 1989a,b ; Lee cells was relatively abundant. We were able to detect p27 following immunoprecipitation of radiolabeled ly-et al., 1991) , similar to that predicted for the HCV 229E 3CLpro. A 34-kDa protein has been detected in 229E-sates from 1 1 10 5 cells by gel exposures of 12 hr. In contrast, the putative HCV 229E 3CLpro was reported to infected cells, using antibodies specific to the 3CLpro region (Ziebuhr et al., 1995) . The 229E 34-kDa protein be present in very small amounts in infected cells, requiring prolonged exposures of 28 days to detect the protein from infected cells has not been shown to possess proteinase activity; however, a 229E 34-kDa protein ex- (Ziebuhr et al., 1995) . It is not known whether these differences reflect the sensitivity of the antisera used or are pressed in E. coli from a cDNA clone encoding the predicted 3CLpro was able to cleave substrate. true indicators of the amounts of the proteinases produced by the two viruses. MHV p27 was detected in cells We have not determined the reason for the discrepancy between the observed and predicted masses of beginning at 6 hr postinfection, similar to the other MHV gene 1 products identified in virus-infected cells. This the MHV 3CLpro. Gel analysis may underestimate the molecular mass of proteins, especially if the protein is also was distinct from the results obtained with the HCV-229E 3CLpro, which was detected at earlier time points, not completely denatured, or if the protein binds a disproportionately large amount of SDS with a concomitant but declined at later times p.i. (Ziebuhr et al., 1995) . Once processed, p27 was resistant to degradation for increase in negative charge. We believe that the p27 proteins were completely denatured, since translation over 2 hr in infected cells and in reticulocyte lysates, suggesting that stability and accumulation of p27 may products were boiled extensively in SDS-containing loading buffer before electrophoresis. Further, boiling the be important for processing of the gene 1 polyprotein. It was surprising that no p27-containing precursors were samples in loading buffer containing 8 M urea, followed by electrophoresis on urea-containing gels, did not alter detected in MHV-infected cells. The B3 antiserum was raised against a fusion protein that extended into the the apparent molecular mass of the 27-kDa proteins (data not shown). We are currently attempting to resolve domain flanking the amino terminus of p27, yet still did not detect precursors. Although it is possible that the this question by identifying the carboxy-terminal cleavage site of p27 and by assessing the size and proteolytic available antisera were unable to bind to intermediate precursors to p27, or that background proteins obscured activity of truncated versions of the predicted full-length 3CLpro.

true precursors, the kinetics of labeling and processing support the conclusion that p27 was rapidly cleaved from Despite the discrepancy in predicted and apparent mass, we have shown that the 27-kDa proteins from cells the nascent polyprotein. The proteolytic activity of p27 was most efficiently in-and in vitro translation lysates were identical in mobility, had the same pattern of antibody detection, and were hibited by derivatives of E64, an irreversible cysteine pro-

after SDS-gel electrophoresis

Chromogenic NS2B-3 proteinase-mediated processing in the yellow fever virus substrate autography: A method for detection, characterization, and structural region: In vitro and in vivo studies

Characterization of a Virol. 67, 6056-6063. human coronavirus (strain 229E) 3C-like proteinase activity

Mouse hepatitis

virus strain A59 RNA polymerase gene ORF 1a: Heterogeneity among teinase inhibitor. E64 was developed as an inhibitor of MHV strains. Virology 198, [736] [737] [738] [739] [740] cellular cysteine proteinases such as calpain, cathepsin Bonilla, P. J., Hughes, S. A., Pinon, J. D., and Weiss, S. R. (1995) . Charac-L, and cysteine proteinases related to papain. Interestterization of the leader papain-like proteinase of MHV-A59: Identificaingly, it has not inhibited 3C proteinases of the picornavition of a new in vitro cleavage site. Virology 209, 489-497. Bouffard, P., Bartenschlager, R., Ahlborn, L. L., Mous, J., Roberts, N., ruses such as 2A and 3C, whereas it has been used and Jacobsen, H. (1995) . An in vitro assay for hepatitis C virus NS3 against papain-like proteinases such as the L proteinase ( Lee et al., 1991) . We have demonstrated the essential Dong, S., and Baker, S. C. (1994) . Determinants of the p28 cleavage nature of the His and Cys residues; however, the corosite recognized by the first papain-like cysteine proteinase of murine navirus 3CLpro domain is predicted to have a much coronavirus. Virology 204, 541-549. Gorbalenya, A., and Koonin, E. (1993) . Comparative analysis of amino-larger number of amino acid residues following the sub-