INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Avian infectious bronchitis virus (IBV), the prototype of the family of enveloped viruses Coronaviridae, is now categorized under the newly established order Nidovirales, which comprises the families Coronaviridae and Arteriviridae. The name Nidovirales came from the Latin nidus, for nest, referring to the 3'-coterminal "nested" set of subgenomic mRNAs produced during viral infection (33). Two overlapping open reading frames (ORFs), 1a and 1b, are located at the 5'-end unique region of the genome and are expected to encode polyproteins of 441 and 300 kDa, respectively (2). By a unique frameshifting mechanism, the downstream ORF 1b is expressed as a fusion protein of 741 kDa with ORF 1a (3, 4). The 441-kDa 1a and 741-kDa 1a/1b polyproteins are proteolytically processed into smaller mature products required for RNA synthesis and other aspects of viral replication.

Nucleotide sequence analyses of coronaviruses from three different antigenic groups have predicted the presence of three proteinases (2, 7, 9, 10, 16), namely, two overlapping papain-like proteinase domains (PLPDs) and a picornavirus 3C-like proteinase domain (3CLP). To date, both PLPD-1 and 3CLP have been demonstrated to be active, and their catalytic properties have been characterized. PLPD-1 is involved in proteolytic cleavage of the N-terminal region of the 1a and 1a/1b polyproteins. For mouse hepatitis virus (MHV), it was shown that PLPD-1 mediates cleavage at Gly²⁴⁷-Val²⁴⁸ and Ala⁸³²-Gly⁸³³ dipeptide bonds to release p28 and p65, respectively (1, 6, 13). The Cys¹¹³⁷ and His¹²⁸⁸ residues were defined as the catalytic dyad of this proteinase (1, 6, 13). For human coronavirus (HCV) 229E, it was reported that PLPD-1 is required for releasing p9 upon cleavage at the Gly¹¹¹-Asn¹¹² dipeptide bond, and the Cys¹¹³⁷ and His¹²⁰⁵ residues were identified as its catalytic dyad (11). No experimental evidence has so far shown that PLPD-2 is functionally active for both viruses. Multiple proteinase domains have also been described for arteriviruses. For example, nsp1, nsp2, and nsp4 of equine arteritis virus were shown to have proteolytic activities (32).

In our previous reports, IBV PLPD-1 was shown to be responsible for cleavage of the 1a and 1a/1b polyproteins at the Gly⁶⁷³-Gly⁶⁷⁴ dipeptide bond to release an 87-kDa N-terminal cleavage product (17, 21). The Cys¹²⁷⁴ and His¹⁴³⁷ residues were revealed to be the catalytic dyad of this proteinase (17). In this paper, we report yet another cleavage event performed by PLPD-1. Our results show that PLPD-1 is responsible for the cleavage of the product encoded by nucleotides 2548 to 8865 (which is located between the 87-kDa protein and 3CLP) (Fig. 1) to release a 195-kDa N-terminal product and a 41-kDa C-terminal product. Both cleavage products were identified in IBV-infected cells. Systematic deletion and mutagenesis studies map the second cleavage site to the Gly²²⁶⁵-Gly²²⁶⁶ scissile bond. No further cleavage activity mediated by either PLPD-1 or 3CLP was found to occur before the region encoding 3CLP, suggesting that both the 195- and 41-kDa proteins may be mature cleavage products.

View larger version (19K): [in this window] [in a new window]   FIG. 1.   Diagram of the structure of the 5'-terminal 9-kb region of ORF 1a, illustrating the location of the 87-kDa protein, the 195-kDa protein with the two putative overlapping PLPDs, the 41-kDa glycoprotein, and 3CLP. The Gly673Gly674, Gly2265Gly2266, and Gln2779Ser2780 scissile bonds, the Asn2313 N-linked glycosylation site, and the catalytic residues of PLPD-1 (Cys1274 and His1437) are indicated. Also shown are the IBV sequence recognized by antisera V59, V54, V46, and anti-3C and the stretches of acidic and hydrophobic residues encoded by ORF1a.

Characterization of the two cleavage products showed that the 195-kDa, PLPD-1-containing protein possesses both cis- and trans-cleavage activities at the Gly²²⁶⁵-Gly²²⁶⁶ dipeptide bond when expressed in intact cells. Only cis-cleavage activity was detected when the protein was expressed in vitro in rabbit reticulocyte lysates. The 41-kDa protein was characterized as an N-linked glycoprotein, and mutagenesis studies confirmed that the putative Asn²³¹³ residue is the actual N-linked glycosylation site.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Virus and cells. The egg-adapted Beaudette strain of IBV (ATCC VR-22) was obtained from the American Type Culture Collection and was adapted to Vero cells as described previously (23). Virus stocks were prepared by infecting Vero cells at a multiplicity of infection of approximately 0.1 PFU/cell and incubating them at 37°C in 5% CO₂ for 48 h. The virus titer was determined by plaque assay on Vero cells.

Transient expression of IBV sequence in Cos-7 cells by using a vaccinia virus-T7 expression system. IBV sequences were placed under the control of a T7 promoter and transiently expressed in mammalian cells, using the system described by Fuerst et al. (8). Briefly, 60 to 80% confluent monolayers of Cos-7 cells grown on 35-mm-diameter dishes (Falcon) were infected at 10 PFU/cell with a recombinant vaccinia virus (vTF7-3) which expresses bacteriophage T7 RNA polymerase. The cells were then transfected with 2.5 µg of plasmid DNA (purified by Qiagen plasmid Midi kits) mixed with DOTAP liposomal transfection reagent according to the instructions of the manufacturer (Boehringer Mannheim). The transfection efficiency achieved by this method was about 25 to 35%, as determined by monitoring the expression of a reporter plasmid encoding green fluorescent protein from jellyfish Aequorea victoria (5). After incubation at 37°C in 5% CO₂ for 5 h, the cells were washed twice with methionine-free medium and labeled with 25 µCi of [³⁵S]methionine per ml. The radiolabeled cells were then harvested at 18 h posttransfection.

PCR. Complementary DNA templates for PCR were prepared from purified IBV virion RNA by using a first-strand cDNA synthesis kit (Boehringer Mannheim). Amplification of template DNAs with appropriate primers were performed with Pfu DNA polymerase (Stratagene) under the standard buffer conditions with 2 mM MgCl₂. The reaction conditions used were 30 cycles of 95°C for 45 s, x°C for 45 s, and 72°C for x min. The annealing temperature (x°C) and the extension time (x min) were subjected to adjustments according to the melting temperature of the primers used and the length of PCR fragments synthesized.

Site-directed mutagenesis. Site-directed mutagenesis was carried out by two rounds of PCR and two pairs of primers, as previously described (22).

Radioimmunoprecipitation. IBV-infected Vero cells and plasmid-transfected Cos-7 cells were lysed with radioimmunoprecipitation assay buffer (50 mM Tris-HCl [pH 7.5], 150 mM NaCl, 1% sodium deoxycholate, 1% NP-40, 0.1% sodium dodecyl sulfate [SDS]) and precleared by centrifugation at 12,000 rpm for 5 min at 4°C in a microcentrifuge. Immunoprecipitation with region-specific rabbit polyclonal and anti-T7 monoclonal antibodies (Novagen) were carried out as previously described (19).

Cell-free transcription and translation. Plasmid DNAs were expressed in rabbit reticulocyte lysates by using a transcription-coupled translation (TnT) system according to the instructions of the manufacturer (Promega). Briefly, 1 µg of plasmid DNA was included in a 50-µl reaction mix, in the presence or absence of 5 µl of canine pancreatic microsomal membranes. The reaction mix was incubated at 30°C for 90 min in the presence of 50 µCi of [³⁵S]methionine per ml.

SDS-PAGE. Polyacrylamide gel electrophoresis (PAGE) of viral polypeptides was performed on SDS-12.5% polyacrylamide gels (15). The ³⁵S-labeled polypeptides were detected by autoradiography of the dried gels.

Endo H treatment of immunoprecipitated proteins. Endoglycosidase H (endo H) treatment (Boehringer Mannheim) of samples was performed as described by Machamer et al. (24). Briefly, the immunoprecipitated viral proteins were released from the Sepharose beads with immobilized Staphylococcus protein A by boiling in phosphate-buffered saline for 4 min. The supernatant was transferred to a fresh tube containing an equal volume of 2× endo H buffer (50 mM sodium citrate, 0.2% SDS, 100 mM 2-mercaptoethanol [pH 5.5]) and enzyme (40 U/mg) and incubated overnight at room temperature.

Construction of plasmids. Plasmid pORF1a3, which covers nucleotides 2545 to 9786, was constructed by three steps of cloning. First, a reverse transcription (RT)-PCR fragment covering nucleotides 6870 to 9786 was generated from viral RNA, using primer XIANG-35 for RT and primers LN-1 and XIANG-35 for PCR. This fragment was digested with NcoI and BamHI, which were introduced by the two PCR primers, and ligated into NcoI- and BamHI-digested pKTO (19), giving rise to pIBV1a7. Second, an RT-PCR fragment covering nucleotides 3827 to 8885 was made from viral RNA, using primer LN-6 for RT and primers AD-2 and LN-6 for PCR. This fragment was digested with NcoI and SnaBI, which digest the IBV sequence at nucleotide positions 3831 and 7999, respectively, and ligated into NcoI- and SnaBI-digested pIBV1a7, giving rise to pIBV1a6(ext). Finally, a PCR fragment covering nucleotides 2548 to 5759 was generated using pIBV1a2 (17) as a template and LKP-14 and AD-3 as primers. This fragment was digested with BamHI and MluI and ligated into BglII- and MluI-digested pIBV1a6(ext), giving rise to pORF1a3. BamHI site was introduced by primer LKP-14. MluI digests the IBV sequence at nucleotide position 3996, and BglII cuts pIBV1a6(ext) at the position 16 nucleotides upstream of the IBV sequence. The construction and nucleotide sequence of pORF1a3 were confirmed by automated sequencing.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Expression of the IBV sequence from nucleotides 2545 to 9786. In a previous report, we demonstrated that the first cleavage product of the 1a and 1a/1b polyproteins is an 87-kDa protein encoded by ORF 1a between nucleotides 529 and 2547. To study the processing events occurring downstream of the 87-kDa protein, plasmid pORF1a3 was constructed to cover nucleotides 2545 to 9786. As shown in Fig. 2a, both PLPDs and 3CLP were included in this construct. Transient expression of pORF1a3 in Cos-7 cells using the vaccinia virus-T7 expression system resulted in the immunoprecipitation of two polypeptides of approximately 33 and 195 kDa (Fig. 2b, lanes 2 and 4). The 33-kDa protein, which represents 3CLP (27), was immunoprecipitated with anti-3C antiserum (lane 2), and the 195-kDa protein was detected by immunoprecipitation with antiserum V46, which was raised against the IBV sequence encoded by nucleotides 4864 to 5576. To investigate if the 195-kDa protein represents the N-terminal cleavage product of polyprotein encoded by this construct, two deletion constructs, pORF1a31 and pORF1a4, were made. Plasmid pORF1a31 contains a C-terminal deletion of the IBV sequence coding for 3CLP (nucleotides 8866 to 9786), and pORF1a4 contains a UAA termination codon inserted at nucleotide position 7952 (Fig. 2a). Expression of these plasmids in Cos-7 cells resulted in the detection of a 195-kDa protein which comigrates with the 195-kDa protein expressed from pORF1a3 (Fig. 2b, lanes 4 to 6). Interestingly, a smeary band of approximately 41 kDa and several smaller protein species migrating around 20 to 25 kDa were detected by immunoprecipitation with V46 from the expression of pORF1a31 and pORF1a4, respectively (lanes 5 and 6). As the same 195-kDa protein was detected from all three constructs and the only difference among them lies in the 3'-terminal region, it was suggested that a novel cleavage event might occur, resulting in the release of the N-terminal 195-kDa protein and a C-terminal cleavage product. The detection of the potential C-terminal cleavage protein by antiserum V46 may indicate the interaction between the N- and C-terminal cleavage product; the reason for failing to detect the 41-kDa protein from the expression of pORF1a3 with antiserum V46 is unclear (lane 4).

View larger version (23K): [in this window] [in a new window]   FIG. 2.   (a) Diagram showing the IBV sequence contained in plasmids pORF1a3, pORF1a31, pORF1a4, and pORF1a3(M). (b) Analysis of transiently expressed ORF 1a products from cells transfected with pORF1a3(M), pORF1a3, pORF1a31, and pORF1a4. Plasmid DNAs were transiently expressed in Cos-7 cells, using the vaccinia virus-T7 expression system. Cells were labeled with [35S]methionine, lysates were prepared, and polypeptides were immunoprecipitated with anti-3C and V46. Gel electrophoresis of the polypeptides was performed on an SDS-12.5% polyacrylamide gel, and polypeptides were detected by fluorography. Numbers indicate molecular masses in kilodaltons.

In vitro translation and processing of the polyprotein encoded by nucleotides 2545 to 8865. To confirm that a novel cleavage event did occur in the region between the 87-kDa protein and 3CLP and to investigate if the cleavage activity could be detected in vitro as described for MHV and HCV 229E (1, 6, 11, 13), plasmids pORF1a31 and pORF1a4 were expressed in in vitro time course experiments using the TnT reticulocyte lysate system. For each reaction, 10-fold unlabeled methionine was added to the reaction mixture after incubation for 90 min in the presence of 50 µCi of [³⁵S]methionine per ml. A prolonged pulse time was required to allow labeling of the polypeptide towards the C terminus. As can be seen, expression of pORF1a31 led to the detection of a polypeptide with an apparent molecular mass of 39 kDa after incubation for 90 min; the product was increased toward the end of the time course (Fig. 3a). Similarly, expression of pORF1a4 led to the detection of a small protein of approximately 20 kDa (Fig. 3b). This product was first seen at the 60-min time point and was increased over time (Fig. 3b). The appearance of the 195-kDa protein, which was more clearly observed from gels with lighter exposure (data not shown), coincided with that of the 39- and 20-kDa proteins (Fig. 3). These results clearly showed the occurrence of a novel cleavage event, resulting in the release of the N-terminal 195-kDa protein and a C-terminal product. We noted that the C-terminal cleavage products detected in vitro migrate more rapidly than those detected in intact cells. This reflects the posttranslational modification of the cleavage products and will be addressed later.

View larger version (61K): [in this window] [in a new window]   FIG. 3.   Time course analysis of polypeptides synthesized in rabbit reticulocyte lysates from pORF1a31 (a) and pORF1a4 (b). A 10-fold excess of unlabeled methionine was added to the in vitro translation reaction mixture after incubation at 30°C for 90 min, and aliquots were taken from the reaction mixture after incubation for 30, 60, 90, 120, and 240 min. [35S]methionine-labeled translation products were separated on an SDS-15% polyacrylamide gel and detected by fluorography. HMW, high-molecular-mass markers (numbers indicate kilodaltons).

Identification of the proteinase domain required for the cleavage event and demonstration of its trans-cleavage activity. We previously demonstrated that PLPD-1 encoded between nucleotides 4243 and 5576 was responsible for the cleavage of the 87-kDa protein at the Gly⁶⁷³-Gly⁶⁷⁴ dipeptide bond (17), and the catalytic dyad for this activity was the Cys¹²⁷⁴ and His¹⁴³⁷ residues. We then used deletion and mutational analyses to test if either of the two PLPDs is involved in this novel cleavage activity. Two constructs, pORF1a4C¹²⁷⁴-S and pORF1a41, were initially constructed and expressed. pORF1a4C¹²⁷⁴-S is similar to pORF1a4 but contains a mutation of the putative catalytic residue Cys¹²⁷⁴ to a Ser residue; pORF1a41 contains a deletion of nucleotides 4010 to 4610, which encode part of PLPD-1 including the catalytic residue Cys¹²⁷⁴ (Fig. 4a). Expression of these constructs showed no detection of the cleavage product compared with pORF1a4 (Fig. 4b, lanes 1 to 3), suggesting that PLPD-1 may be involved in this cleavage event.

View larger version (23K): [in this window] [in a new window]   FIG. 4.   (a) Diagram showing the IBV sequences present in plasmids pORF1a4, pORF1a4C1274-S, pORF1a41, pORF1a42, and pORF1a32. (b) Analysis of transiently expressed ORF 1a products from transfection of pORF1a4, pORF1a4C1274-S, pORF1a41, and pORF1a42. Plasmid DNAs were transiently expressed in Cos-7 cells, using the vaccinia virus-T7 expression system. Cells were labeled with [35S]methionine, lysates were prepared, and polypeptides were immunoprecipitated with antisera V46 and V54. Gel electrophoresis of the polypeptides was performed on an SDS-12.5% polyacrylamide gel, and polypeptides were detected by fluorography. The upper panel was prepared from a gel exposed for 1 day, and the lower panel was prepared from the same gel exposed for 5 days. Numbers indicate molecular masses in kilodaltons. (c) trans-cleavage assay of PLPD-1 by coexpression of pORF1a32 and pORF1a4C1274-S. Cos-7 cells were transfected with either a single plasmid or two plasmids together and labeled with [35S]methionine. Lysates were prepared, and polypeptides were immunoprecipitated with antiserum V46. Gel electrophoresis of the polypeptides was performed on an SDS-12.5% polyacrylamide gel, and polypeptides were detected by fluorography. Numbers indicate molecular masses in kilodaltons.

Mapping of the second cleavage site to the Gly²²⁶⁵-Gly²²⁶⁶ scissile bond. As shown in Fig. 3b, the actual C-terminal cleavage product of the pORF1a4-encoded product was approximately 20 kDa. This allows us to estimate that the cleavage site is located near nucleotide position 7300. By comparative analysis of the cleavage sites utilized by this and other viral papain-like proteinases (1, 6, 11, 13, 16), one of the following scissile bonds may be the cleavage site: Gly²²⁶⁵-Gly²²⁶⁶, Gly²²⁴⁶-Ala²²⁴⁷, or Gly²²⁴⁷-Val²²⁴⁸. Deletion and site-directed mutagenesis approaches were used to investigate this possibility.

View larger version (1165K): [in this window] [in a new window]   FIG. 5.   (a) Effects of substrate deletions and mutations on the second cleavage activity mediated by PLPD-1. The amino acid sequence flanking the Gly2265-Gly2266 dipeptide bond and the deletions and mutations introduced are outlined. (b) Effects of deletions of the Gly2265, Gly2265-Gly2266, Gly2246-Ala2247, and Gly2237-Val2238 residues on cleavage activity. Plasmid DNAs were transiently expressed in Cos-7 cells, using the vaccinia virus-T7 expression system. [35S]methionine-labeled cell lysates were prepared, and polypeptides were immunoprecipitated with antiserum V46. Gel electrophoresis of the polypeptides was performed on an SDS-12.5% polyacrylamide gel, and polypeptides were detected by fluorography. The upper panel was prepared from a gel exposed for 1 day, and the lower panel was prepared from the same gel exposed for 5 days. Numbers indicate molecular masses in kilodaltons. (c) Effects of mutations of the A2264-N, G2265-A, G2265-N, and G2266-N residues on cleavage activity. The mutants were expressed and analyzed as described for panel b. The upper panel was prepared from a gel exposed for 1 day, and the lower panel was prepared from the same gel exposed for 5 days.

No cleavage at the previously predicted Q²⁵⁸³-G²⁵⁸⁴ dipeptide bond by 3CLP. In a previous study using an in vitro expression and processing system, Tibbles et al. (35) showed that the two predicted Q-S dipeptide bonds (Q²⁷⁷⁹-S²⁷⁸⁰ and Q³⁰⁸⁶-S³⁰⁸⁷) may be the cleavage sites responsible for releasing 3CLP from the 1a and 1a/1b polyproteins. In addition, the Q²⁵⁸³-G²⁵⁸⁴ dipeptide bond encoded by nucleotides 8275 to 8280 was predicted to be a cleavage site of 3CLP (9). As this site is located in the 41-kDa protein-encoding region, mutagenesis studies were carried out to test if cleavage could occur at this position. As outlined in Fig. 6a, expression of pIBV1a71 would result in the detection of a 33-kDa protein representing 3CLP. Mutation of the Q²⁷⁷⁹ to an E was introduced to pIBV1a71 to generate plasmid pIBV1a71Q²⁷⁷⁹-E (Fig. 6a). A 54-kDa protein containing the 33-kDa protein and the upstream region would be immunoprecipitated by anti-3C antiserum (Fig. 6a). As can be seen, expression of pIBV1a71 resulted in the detection of the 33-kDa protein only (Fig. 6b, lane 1). The same product was also detected from expression of pIBV1a71Q²⁷⁷⁹-E (lane 2); in addition, a product of approximately 54-kDa was detected (lane 2). No other cleavage products were observed, suggesting that the Q²⁵⁸³-G²⁵⁸⁴ dipeptide bond may not be used by 3CLP. To explore this possibility further, plasmid pIBV1a72Q²⁷⁷⁹-E, which covers nucleotides 8278 to 9786, was constructed and expressed (Fig. 6a). As expected, expression of this construct resulted in the detection of the 33-kDa protein and a product of approximately 48 kDa (Fig. 6b, lane 3). The 48-kDa protein represents the full-length product encoded by this construct. If cleavage occurred at the Q²⁵⁸³-G²⁵⁸⁴ dipeptide bond, a product comigrating with the 48-kDa protein would be detected from the expression of pIBV1a71Q²⁷⁷⁹-E. These results support the conclusion that no cleavage occurs at this position.

View larger version (35K): [in this window] [in a new window]   FIG. 6.   (a) Diagram showing the IBV sequence contained in plasmids pIBV1a71, pIBV1a71Q2779-E, and pIBV1a72Q2779-E. (b) Analysis of transiently expressed ORF 1a products from cells transfected with pIBV1a71, pIBV1a71Q2779-E, and pIBV1a72Q2779-E. Plasmid DNAs were transiently expressed in Cos-7 cells, using the vaccinia virus-T7 expression system. Cells were labeled with [35S]methionine, lysates were prepared, and polypeptides were immunoprecipitated with anti-3C. Gel electrophoresis of the polypeptides was performed on an SDS-12.5% polyacrylamide gel, and polypeptides were detected by fluorography. Numbers indicate molecular masses in kilodaltons. (c) Time course analysis of the 41-kDa protein by coexpression of pT7P41 and pIBV3C. Cos-7 cells were transfected with either pT7P41 or pT7P41 and pIBV3C together and labeled with [35S]methionine for 4 h. The cells were washed three times with GMEM and were incubated in GMEM before they were harvested 0, 1, 3, and 5 h later. Lysates were prepared, and polypeptides were immunoprecipitated with antisera anti-T7 and anti-3C. Gel electrophoresis of the polypeptides was performed on an SDS-12.5% polyacrylamide gels, and polypeptides were detected by fluorography. Numbers indicate molecular masses in kilodaltons.

N-linked glycosylation of the 41-kDa protein. As presented in Fig. 2b and 3b, migration of the C-terminal cleavage products was different when expressed in vitro and in intact cells, respectively, suggesting that they may be modified posttranslationally. Examination of the amino acid sequence indicates that a potential N-linked glycosylation site (Asn²³¹³-Ser²³¹⁴-Thr²³¹⁵) is encoded by nucleotides 7465 to 7413. Two approaches were used to explore if glycosylation did occur at this position. First, the Asn²³¹³ residue of the putative N-linked glycosylation site (Asn²³¹³-Ser²³¹⁴-Thr²³¹⁵) was mutated to a Gln residue, giving rise to plasmid pORF1a4N²³¹³-Q. When this plasmid was expressed in intact cells, only the 195- and 20-kDa cleavage products were detected (Fig. 7a, lane 2). Both proteins comigrated with the 195- and 20-kDa proteins yielded from the in vitro translation system (lanes 2 and 3). Next, the in vivo-generated products of pORF1a4 were subjected to digestion with endo H. Figure 7b shows that the two upper bands (24- and 25-kDa proteins) were sensitive to endo H treatment (lane 2); the resulting product comigrated with the protein generated from the in vitro-expressed 20-kDa protein (lanes 2 and 3). These results confirm that the C-terminal 20-kDa product is modified by N-linked glycosylation.

View larger version (43K): [in this window] [in a new window]   FIG. 7.   (a) Mutational analysis of the putative N-linked glycosylation site of the 41-kDa protein. Plasmids pORF1a4 and pORF1a4N2313-Q were transiently expressed in Cos-7 cells, using the vaccinia virus-T7 expression system. Cells were labeled with [35S]methionine, lysates were prepared, and polypeptides were immunoprecipitated with antiserum V46. Gel electrophoresis of the in vivo-synthesized polypeptides (lanes 1 and 2) together with the in vitro-synthesized products (lane 3) was performed on an SDS-12.5% polyacrylamide gel. Polypeptides were detected by fluorography. Numbers indicate molecular masses in kilodaltons. (b) Endo H treatment of the C-terminal cleavage products from expression of pORF1a4. Plasmid DNA was transfected into Cos-7 cells, using the vaccinia virus-T7 expression system. The [35S]methionine-labeled cell lysates were prepared, and polypeptides were immunoprecipitated with antiserum V46. The immunoprecipitated viral proteins were analyzed either by gel electrophoresis directly (lane 4) or after incubation with endo H (40 U/mg; lane 2). Incubation of the immunoprecipitated viral proteins with endo H buffer alone (lane 3) and the 20-kDa cleavage product obtained from in vitro expression of pORF1a4 (lane 1) were included as controls. The [35S]methionine-labeled polypeptides were subjected to gel electrophoresis on an SDS-12.5% polyacrylamide gel, and polypeptides were detected by fluorography. (c) Endo H treatment of the in vitro-synthesized 41-kDa protein. Plasmid pP41 was expressed in vitro, using the cell-free transcription-coupled translation system, in the presence or absence of 5 µl of canine pancreatic microsomal membranes. The [35S]methionine-labeled polypeptides were analyzed either by gel electrophoresis directly or after incubation with endo H (40 U/mg) or buffer only.

Interaction between the cleavage products. During the course of this study, we used antiserum V46 (V54 in one case) to immunoprecipitate both the N- and C-terminal cleavage products. As V46 was raised against the IBV sequence encoded by nucleotides 4864 to 5576, the detection of both products by this antiserum may suggest that they are physically associated. To test the specificity of this antiserum and explore the potential interaction between the two products, plasmid pORF1a33, which encodes the 195-kDa protein, and pT7P41 were expressed (Fig. 8a). pT7P41 was used because attempts to raise antiserum against the 41-kDa protein were unsuccessful. As shown in Fig. 8b, when the two constructs were separately expressed, anti-T7 antiserum could precipitate only the 41-kDa protein encoded by pT7P41 (lane 1); no 195-kDa protein was detected (lane 3). Similarly, V46 could precipitate only the 195-kDa protein encoded by pORF1a33 (lane 6); no immunoprecipitation of the 41-kDa protein was observed (lane 4). However, coexpression of the two constructs led to precipitation of both the 195- and 41-kDa proteins by either antiserum (lanes 2 and 5). These results demonstrate that physical interactions of the two cleavage products may occur in intact cells, allowing us to coprecipitate the C-terminal cleavage product with a region-specific antiserum against the N-terminal cleavage product.

View larger version (17K): [in this window] [in a new window]   FIG. 8.   (a) Diagram showing the IBV sequences present in plasmids pORF1a33 and pT7P41. (b) Analysis of transiently expressed proteins from transfection of pORF1a33 and pT7P41. Plasmid DNAs were transiently expressed in Cos-7 cells, using the vaccinia virus-T7 expression system. Cells were labeled with [35S]methionine, lysates were prepared, and polypeptides were immunoprecipitated with anti-T7 or V46 antiserum. Gel electrophoresis of the polypeptides was performed on an SDS-12.5% polyacrylamide gel, and polypeptides were detected by fluorography. Numbers indicate molecular masses in kilodaltons.

Identification of two novel gene products encoded by ORF1a in IBV-infected cells. Confluent monolayers of Vero cells were infected with IBV at a multiplicity of infection of approximately 2 PFU/cell. The cells were labeled with [³⁵S]methionine at 6 h postinfection (p.i.) for 4 h and then harvested (Fig. 9a). Immunoprecipitation of cell lysates with antisera V46 and V54 resulted in the detection of the 41-kDa protein from IBV-infected cells (Fig. 9a, lanes 5 and 6) but not from mock-infected cells (lanes 9 and 10). Meanwhile, immunoprecipitation of the same lysates with anti-3C antiserum led to the detection of the 33-kDa 3CLP in IBV-infected cells (lane 7). However, the expected 195-kDa protein was only marginally, if any, detected by either V46 or V54 (lanes 5 and 6).

View larger version (60K): [in this window] [in a new window]   FIG. 9.   (a) Detection of polypeptides encoded by ORF1a in IBV-infected Vero cells. IBV-infected ("I") and mock-infected ("M") Vero cell monolayers in 35-mm-diameter dishes were labeled with [35S]methionine for 4 h at 6 h p.i. Cell lysates were prepared, and polypeptides were either analyzed directly or immunoprecipitated with antisera anti-IBV, V46, V54, and anti-3C. Gel electrophoresis of the polypeptides was performed on an SDS-12.5% polyacrylamide gel, and polypeptides were detected by fluorography. Numbers indicate molecular masses in kilodaltons. (b) Time course analysis of three ORF 1a-specific polypeptides in IBV-infected Vero cells. Cells were infected with IBV at a multiplicity of infection of approximately 2. After incubation in complete GMEM for 5.5 h, the cells were incubated in methionine-free medium for 0.5 h and were labeled with [35S]methionine (100 µCi/ml) for 4 h. The cells were then washed three times with GMEM and incubated in GMEM until harvested at 10, 12, and 24 h p.i. Lysates were prepared, and polypeptides were immunoprecipitated with antisera V59 and V46. Gel electrophoresis of the polypeptides was performed on an SDS-12.5% polyacrylamide gel, and polypeptides were detected by fluorography. Numbers indicate molecular masses in kilodaltons.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Data presented in this report demonstrate a novel cleavage activity of IBV PLPD-1. Based on sequence comparison, mutagenesis, and deletion studies, the Gly²²⁶⁵-Gly²²⁶⁶ dipeptide bond was defined as most likely the second cleavage site utilized by PLPD-1 to release a 195- and a 41-kDa protein.

It is generally believed that the sequence context determines the recognition of a cleavage site by a proteinase. The initial deletion studies present in this communication demonstrated that among the three double amino acid deletions (Gly²²⁶⁵-Gly²²⁶⁶, Gly²²⁴⁶-Ala²²⁴⁷, and Gly²²³⁷-Val²²³⁸), the only construct that abolished the cleavage activity was the Gly²²⁶⁵-Gly²²⁶⁶ deletion. These results essentially narrowed down the potential candidates for the cleavage sites to the Ala²²⁶⁴-Gly²²⁶⁵ [Val (position 5 {P5})-Glu(P4)-Lys (P3)-Lys (P2)-Ala (P1)-Gly (P1')-Gly (P2')] and Gly²²⁶⁵-Gly²²⁶⁶ [Glu (P5)-Lys (P4)-Lys (P3)-Ala (P2)-Gly (P1)-Gly (P1')-Ile (P2')] dipeptide bonds. Expression of the construct containing a single Gly (Gly²²⁶⁵ or Gly²²⁶⁶) deletion resulted in a drastic drop in the cleavage activity, indicating that Gly²²⁶⁵ may occupy P1. This is because deletion of the putative P1 site will shift all amino acid residues at P2 to P5 and therefore inhibit cleavage activity. For example, an Ala (P2) would occupy the P1 site and a Lys (P3) would be at P2, resulting in a nonconservative substitution of the nonpolar, uncharged Ala residue with the positively charged Lys at P2. This is consistent with the extensive mutagenesis studies reported for PLPD-1 of MHV and HCV 229E (1, 6, 11, 13), showing that the upstream sequences of a cleavage site play a more important role than the downstream sequences. Furthermore, the inhibitory effect observed is unlikely caused by the replacement of Gly with Ala, as proteolysis was permissive at Ala²²⁶⁵-Gly²²⁶⁶ when Gly²²⁶⁵ was mutated to an Ala. On the other hand, if Ala²²⁶⁴ is at P1, deletion of Gly²²⁶⁵ does not cause any change in P1 to P5 of the recognition site. As the next amino acid residue is Gly²²⁶⁶, deletion of Gly²²⁶⁵ shifts Gly²²⁶⁶ to P1', effectively creating the same Ala-Gly dipeptide bond. No obvious inhibitory effect would be expected, because the amino acids at the P1' and P2' sites were less crucial to the cleavage efficiency of the substrate. In fact, mutation of the Gly²²⁶⁶ residue to Asn has little, if any, effect on the proteolytic event. It is therefore established that the Gly²²⁶⁵-Gly²²⁶⁶ dipeptide bond is most likely the second cleavage site of IBV PLPD-1.

The two cleavage activities of IBV PLPD-1 identified so far are on the Gly-Gly dipeptide bond. It is remarkable that the proteinase shows markedly different properties at the two positions. For the second cleavage event, IBV PLPD-1 behaves more like the PLPD-1 of other coronaviruses studied. For example, cleavage of the precursor at this position could occur both in intact cells and in an in vitro expression system, as seen for MHV and HCV 229E (1, 6, 11, 13). Interestingly, cleavage was not significantly affected when Gly²²⁶⁵ (P1) was mutated to an Ala, which creates an Ala-Gly scissile bond similar to the second cleavage site of MHV PLPD-1 (Ala⁸³²-Gly⁸³³) (1). In fact, mutation of the Ala⁸³²-Gly⁸³³ to Gly-Gly dipeptide did not affect the cleavage activity of MHV PLPD-1 (1). A mutation of Gly²²⁶⁶ at the P1' site to Asn creates a dipeptide bond similar to the Gly¹¹¹-Asn¹¹² cleavage site utilized by HCV 229E to release the p9. This mutation was also permissive to proteolytic processing. It seems that as long as the dipeptide bond fit into consensus cleavage sites utilized by other coronavirus PLPDs, cleavage activity was not affected. IBV PLPD-1 could also act on the second cleavage site in a trans-cleavage manner, which would allow us to characterize the proteinase activity in more detail. Recently, Teng et al. (34) used the trans-cleavage activity of MHV PLPD-1 to define the temperature and minimum lengths of both substrate and enzyme required for optimal cleavage efficiency. By contrast, cleavage at the first Gly-Gly dipeptide bond (Gly⁶⁷³-Gly⁶⁷⁴) of IBV occurred only when both the substrate and enzyme were expressed in intact cells and no trans-cleavage activity was detected at the position (17, 21). It would be interesting to investigate whether these differences are dictated simply by the sequence context of the individual cleavage sites or, more sophisticatedly, by the overall folding and/or cellular compartmentalization of the substrates and the enzyme.

Furthermore, the Gly⁶⁷³-Gly⁶⁷⁴ and Gly²²⁶⁵-Gly²²⁶⁶ cleavage sites identified for IBV PLPD-1 are not exactly compatible with the general alignment model [P5-(R, K)XXX(G, A)(G, A, V)-P1'] proposed for coronavirus PLPD (13). For the cleavage sites utilized by PLPD-1 of MHV and HCV 229E (1, 6, 11, 13), P5 is occupied by a conserved basic amino acid, either an Arg or a Lys. But for IBV PLPD-1, instead of a basic amino acid residue, a Val for the first cleavage site and a Glu for the second cleavage site lie at this position. Interestingly, P3 of both IBV cleavage sites is occupied by a Lys. The significance of this residue to the two cleavage activities is under investigation.

Cleavage at the Gly²²⁶⁵-Gly²²⁶⁶ dipeptide bond would result in the release of the N-terminal 195- and the C-terminal 41-kDa proteins. The 195-kDa protein is therefore the cleavage product immediately downstream of the 87-kDa protein. It contains PLPD-1 and the putative PLPD-2 (17, 21) and has a calculated molecular mass of 180 kDa (1,592 amino acid residues). However, the protein migrates more slowly on the SDS-PAGE system used, probably due to the presence of a stretch of negatively charged residues at its N terminus, as mentioned in our previous report (17). As the two Gly-Gly (Gly⁶⁷³-Gly⁶⁷⁴ and Gly²²⁶⁵-Gly²²⁶⁶) cleavage sites flank the 195-kDa protein, cleavage at these positions effectively results in autoprocessing of the 195-kDa protein. Autoprocessing is so far the only known function of the IBV 195-kDa protein. Characterization of the MHV and HCV PLPD-1 activities showed that the minimal sequences required for functionally active proteinase are 232 (between amino acids 1084 and 1316) and 422 (between amino acid residues 863 and 1285) amino acid residues, respectively (1, 11). More recently, Herold et al. (12) have demonstrated that the HCV PLPD-1 activity is dependent on a unique Zn²⁺-binding finger that connects the two domains of a papain-like fold (12). It is suggested that although the relatively small portion of the proteinase domain is contained in a large protein molecule, the proteinase domain may maintain a unique folding. We do not know whether the bulk of the molecule may play additional, independent functions.

The C-terminal cleavage product migrates at 41 kDa on SDS-PAGE, although it has a calculated molecular mass of 58 kDa (514 amino acids) and has undergone N-linked glycosylation. This is probably due to its high content of hydrophobic residues. No cleavage site at the position equivalent to the IBV Gly²²⁶⁵-Gly²²⁶⁶ dipeptide bond has been identified so far for other coronaviruses. The available evidence indicated that a potential cleavage site might exist at a similar position for MHV-JHM. Schiller et al. (29) have recently reported the identification of two cleavage products of 250 and 210 kDa, which may be encoded by regions corresponding to the IBV sequences coding for the 240- and 195-kDa proteins. Pulse-chase experiments further showed that the 250-kDa precursor may undergo proteolytic processing to release the 210-kDa protein and a protein of 40 kDa (29) strongly suggesting the existence of a cleavage site equivalent to the second cleavage site of IBV identified in this report. The 41-kDa glycoprotein (39 kDa when unglycosylated) would be the equivalent of the 40-kDa protein suggested by Schiller et al. (29).

The 41-kDa protein is the first coronavirus nonstructural protein demonstrated so far to be posttranslationally modified by N-linked glycosylation. Examination of the amino acid sequences showed that the equivalent regions of other coronaviruses also encode potential N-linked glycosylation sites (e.g., MHV contains two, transmissible gastroenteritis virus contains three, and HCV contains one) (7, 10, 16). It is suggested that this modification may be conserved among different groups of coronaviruses. When tagged to green fluorescent protein, the 41-kDa exhibits a staining pattern resembling that of the M protein, a protein shown to be localized to the Golgi apparatus, suggesting that the 41-kDa protein may be also localized to this subcellular compartment (data not shown). N-linked glycosylation of nonstructural proteins was also described for other viral systems. For example, the NS1 protein of flaviviruses was reported to be an N-linked glycoprotein (25).

It is intriguing that the 41-kDa protein is physically associated with the 195-kDa product via its N terminus. In EAV, it was shown that the second and third cleavage products of ORF 1a (corresponding to the 195- and 41-kDa proteins of IBV, respectively) tend to be coimmunoprecipitated (31). As the N-terminal half of the 195-kDa mature product was previously shown to be able to interact with the 87-kDa protein (17), it seems likely that the three products may form a complex. The prime role of this complex formation would be to sequester, by the action of the 41-kDa protein, the proteinase domain-containing 195-kDa protein and the 87-kDa protein to the membranous compartment where the IBV RNA replicates and the virion particles assemble (14). In two recent report, Shi et al. (30) and van der Meer et al. (37) have demonstrated the colocalization and membrane association of several gene 1 products with de novo-synthesized MHV-59 viral RNA in virus-infected cells. When the IBV counterparts of these products were separately expressed, distinct distribution patterns were observed for each protein (D. X. Liu, unpublished observations). It is suggested that the colocalization patterns reported by Shi et al. (30) may reflect the interactions between viral proteins. Such interactions between ORF 1a-encoded products have been demonstrated in arteriviruses (28, 36). They may facilitate the association of the replication complex with cellular membranous compartments (28, 36). Characterization of the multiprotein interaction observed in this study would provide significant insights into the subcellular translocation of coronavirus proteins and the involvement of nonstructural proteins in the viral replication cycle.