key: cord-0921827-uoei747i
authors: Leibowitz, Julian L.; Perlman, Stanley; Weinstock, George; Devries, James R.; Budzilowicz, Carol; Weissemann, Jane M.; Weiss, Susan R.
title: Detection of a murine coronavirus nonstructural protein encoded in a downstream open reading frame
date: 1988-05-31
journal: Virology
DOI: 10.1016/0042-6822(88)90631-9
sha: de24628bf8551727b4b9462aa5afb5bb517c6118
doc_id: 921827
cord_uid: uoei747i

Abstract Mouse hepatitis virus (MHV) gene 5 contains two open reading frames. We have expressed the second open reading frame of this gene (gene 5 ORF 2) in an Escherichia coli expression system. This system utilized a plasmid which contained the promoter and the first 36 codons of the recA gene fused in frame with the MHV gene 5 ORF 2, which is fused in turn to the β-galactosidase gene. The protein product of this gene fusion was used to raise antibody to gene 5 ORF 2. The specificity of the antibody was verified by immunoprecipitation of the in vitro transcribed and translated protein product of gene 5 ORF 2. The second reading frame of MHV gene 5 was shown to be expressed during the course of infection by immunocytochemistry and radioimmunoprecipitation using the antibody raised against the E. coli fusion protein and by two-dimensional gel electrophoresis.

The coronaviruses are a group of RNA viruses with genomes of positive polarity which are about 27 kb in length (Boursnell et a/., 1987; Lai and Stohlman, 1978; Leibowitz et a/., 1981) . Infected cells contain five or six subgenomic mRNAs, depending upon the virus studied, and an RNA species which is indistinguishable from genome RNA in structure (Leibowitz et al., 1981; Stern and Kennedy, 1980a, b; Lai et al., 1981) . Structural studies have established that the subgenomic mRNAs make up a "nested set" with coterminal 3' ends (Leibowitz et a/., 1981; Stern and Kennedy, 1980a, b; Lai et a/., 1981; Cheley et a/., 1981; Weiss and Leibowitz, 1983; Spaan et al., 1982) and contain a leader sequence of approximately 72 bases at their 5' termini (Spaan et a/., 1983; Lai et a/., 1983 Lai et a/., , 1984 . This is shown schematically in Fig. 1 . The leader sequence is identical to the sequence present at the 5'terminus of the virion RNA and contains a short sequence homologous to sequences in the corresponding regions preceding each gene in the genomic RNA.

In vitro translation studies of purified populations of mouse hepatitis virus mRNAs have demonstrated that the MHV nucleocapsid (N), transmembrane (El), and peplomer (E2) proteins are encoded by RNA 7, RNA 6, and RNA 3, respectively (Leibowitz et a/., 1982; Siddell, 1983) . RNA 2 encodes a 30-to 35-kDa nonstructural protein (Leibowitz et al., 1982; Siddell, 1983) . The coding assignments for RNAs 4 and 5 have been harder to determine, in part because of difficulties in resolving these two species. In cellfree translation studies, a 14-kDa nonstructural protein was synthesized in response to a mixture of these two mRNAs (Leibowitz et al., 1982; Siddell, 1983) . These functional studies of MHV mRNA have led to the hypothesis that the most 5' sequences of each mRNA, not present in smaller mRNA species, contain the coding sequences utilized during infection. This hypothesis has allowed the construction of a tentative genetic map of MHV, shown in Fig. 1 .

Sequencing studies of molecular cDNA clones derived from either MHV mRNA or virion RNA have, in general, confirmed the assignment of the MHV structural genes suggested by in vitro translation studies (Armstrong et a/., , 1984 Siddell, 1983, 1985) . Sequence analyses of mRNA cDNA clones corresponding to the 5' portions of RNA 4 of MHV, strain JHM, demonstrate that the unique 5' terminus of RNA 4 contains an open reading frame encoding a 15.4-kDa protein . This is similar in size to the 14-kDa nonvirion protein observed in MHV-infected cells (Leibowitz et al., 1982; Rottier eta/., 1982; Siddell, 1983) and sequencing experiments Budzrlowrcz and Weiss, 1987) , assuming that each mRNA is translated from its 5' unique sequences. A map of the portion of clone 9344 whrch contains gene 5 is shown at the bottom of the figure. ORF 1 and ORF 2 are depicted as solid dark Irnes. The posrtions of their initiating AUG codons and the restriction sites relevant to thts work are indicated. of 13 and 9.6 kDa (Budzilowicz and Weiss, 1987) . Similar results have been obtained in sequencing studies of cDNA clones representing the genome of MHV, strain JHM (Skinner er al., 1985) . The question of which of the two open reading frames present in MHV RNA 5 is expressed in infected cells is not addressed by the above data. Recently Budzilowicz and Weiss (1987) have demonstrated that RNA synthesized in vitro from pGEM recombinant vectors containing these open reading frames can be translated in vitro into polypeptides corresponding in size to both reading frames, although the downstream open reading frame is translated preferentially.

In this work, we show that the downstream open reading frame within RNA 5 is expressed in MHV-infected cells.

The origin and growth of the 17CI-1 and L-2 cell lines have been described (Sturman and Takemoto, 1972; Rothfels er al., 1959; Leibowitz et a/., 1981) . The origin and growth of MHV-A59 has been described (Sturman and Takemoto, 1972; Leibowitz et a/., 1981) .

The isolation and characterization of the plasmid 9344 has been described previously (Budzilowicz et a/., 1985) . This plasmid carries an 1.8-kbp cDNA clone of MHV-A59 spanning the 3' portion of gene 4, all of gene 5, gene 6 (encoding the El protein), and the 5' portion of gene 7 (encoding the nucleocapsid protein).

The plasmids pGE372 (6.58 kbp) and its derivatives are shown in Fig. 2 . Plasmid pGE372 is a derivative of the plasmid pBR322 in which the tet gene (between the EcoRl and Aval sites in pBR322) has been replaced by a fusion of the E. co/i recA and /acZ genes. The fusion contains the regulatory region from recA as well as the first 36 codons of the structural gene. The recA sequence is followed by /acZ, encoding P-galactosidase. The /acZ sequence is missing its promoter, translation start site, and first eight codons. However, the /acZ sequence is fused in frame to recA, such that a RecA-/3-galactosidase hybrid protein is produced. A BamHl site is located between recA and /acZ in pGE372. Plasmid pGE374 is analogous to pGE372 except that the spacer sequence between recA and laczdiffers such that the /acZ reading frame is +2 with respect to the recA reading frame. Consequently a hybrid protein is not produced. 

The entire 1.8-kb MHV-specific insert was excised from the plasmid 9344 by digestion with Pstl and purified by gel electrophoresis. This fragment was digested with TaoI and EcoRV and the resulting Taql-EcoRV fragment of MHV gene 5 containing ORF 2 (254 bases) was purified by agarose gel electrophoresis. The Taql end was repaired using the Klenow fragment of DNA polymerase. Octameric BarnHI linkers were then ligated to this fragment, and after digestion with BarnHI, it was subcloned into the BarnHI site of pBR322. This subcloned fragment was then excised from pBR322 with BarnHI, purified by gel electrophoresis, and ligated into the BarnHI site of pGE372. This yields a construct in which the recA promoter and the first 108 bases of the coding sequence of the recA gene are fused in frame to MHV gene 5 ORF 2, which is, in turn, fused in frame with the /acZ gene. This construction is shown schematically in Fig. 2 .

Large-scale growth and purification of plasmids was performed as described by Clewell and Helinski (1972) . Small-scale preparations of plasmid (~20 ml) were made using the procedure of Birboim and Doty (1979) . Transformation of HBl 01 was by the method of Hanahan (1983) ; transformation of MC1061 was by Dagert and Ehrlich's (1979) modification of the calcium chloride procedure. Ligations and other manipulations of plasmid DNA were essentially as described by Maniatis et al. (1982) . Computer analyses were performed using the programs provided by Dr. Charles Lawrence, Baylor College of Medicine.

Bacterial cultures were harvested by centrifugation and resuspended in a small volume of 10 mll/l Tris, pH 7.4, 10 mlLl NaCI, and 1.5 mlLl MgCl*. The bacterial suspension was sonicated for three 20-set bursts, mixed with an equal volume of 2X SDS-PAGE sample buffer (Maizel, 1971 ) and immersed in a boiling water bath for 5 min. The extract was then clarified in either a microfuge or a table top centrifuge.

Bacterial extracts were prepared in sample buffer and electrophoresed on 8% polyacrylamide gels as described by Maize1 (1971) . The resolved proteins were transferred to nitrocellulose at 250 mA for 16 hr as described by Towbin et al, (1979) . The transferred proteins were either stained with 0.1% amido black or tested for the presence of RecA sequences with a rabbit anti-RecA antibody. Blots were blocked over-night at 4" with 3% nonfat dry milk in Tris saline, 1 mM PMSF, washed with Tris saline, reacted for 2 hr with 1:lOOO dilution of anti-RecA antibody in Tris saline, washed three times with Tris saline containing 0.05% Tween 20, reacted with a 1 :lOOO dilution of peroxidase-conjugated goat anti-rabbit lg (Cappel) for 2 hr, washed three times with Tris saline, 0.05% Tween 20, and the bound antibody was visualized by incubation for no more than 10 min with 0.06% 4-chloronaphthol plus 0.02% hydrogen peroxide.

A 20-ml culture was inoculated with a single colony carrying the desired plasmid and grown to an ODBoo of 0.2. These cells were then inoculated into a mass culture (400-1000 ml) which was incubated until an ODeoO of 0.2 was reached. Mitomycin C was added to the cells at a concentration of 1 pg/ml to induce the recA promoter and the cultures were incubated for an additional 3 hr. Induction with mitomycin C was necessary since the presence of MHV inserts dramatically decreased the amount of the tribrid protein synthesized relative to the parental RecA-LacZ fusions (data not shown). The cells were harvested and protein extracts were prepared as described above. The tribrid proteins were resolved by electrophoresis on 8% polyacrylamide gels, located by Coomassie blue staining of strips cut from the ends of the preparative gel, electroeluted as described previously (Welch et al., 1981) and quantitated. Between 50 and 100 pg of protein was homogenized with complete Freund's adjuvant and injected subcutaneously in two NZW rabbits. Rabbits were subsequently boosted at approximately 4-week intervals by injection of antigen in incomplete Freund's adjuvant. Antisera were initially evaluated by immunoblotting against bacterial extracts which had been resolved by polyacrylamide gel electrophoresis and transferred to nitrocellulose membranes. These extracts were prepared from bacteria carrying the plasmids pGEA59.G5.2, pGE372, and pGE374. At least three immunizations were required before activity against the appropriate bands on the Western blots was observed. This assay merely measured the reactivity of the sera with the tribrid protein and does not distinguish between reactivity to RecA, LacZ, and MHV determinants.

The MHV-A59 specific insert present in plasmid 9344 was excised with Pstl and inserted into the Pstl site of pGEM-1 as described previously (Budzilowicz and Weiss, 1987) . For the synthesis of RNA representing ORF 2, the MHV-specific insert was excised from plasmid 9344 and digested with Taql, and the 1103nucleotide

TaqllPstl fragment was subcloned into pGEM-2 which had been digested with Accl and Pstl. The resulting plasmid was digested with HindIll and RNA was synthesized using the SP6 polymerase. All transcription reactions were carried out as described by Krieg and Melton (1984) .

Approximately 1 pg of in vitro transcribed RNA was translated in a wheat germ cell-free extract (Amersham) as described previously (Budzilowicz and Weiss, 1987) . Translation products were analyzed on 8-l 6% gradient gels (Maizel, 1971) and processed for fluorography with En3Hance (New England Nuclear). lmmunocytochemistry Cells infected with MHV-A59 or mock-infected were stained at 24 hr postinfection for ORF 2 antigens by the immunoperoxidase method using an ABC kit (Vector Labs). Cells were counter-stained with hematoxylin and photographed.

Intracellular labeling with radioactive amino acids, immunoprecipitation, and two-dimensional gel electrophoresis L-2 cells were infected with MHV-A59 (multiplicity of infection, 10) and labeled at the times indicated in the figure captions with either [35S]methionine or [35S]cysteine in methionine-or cysteine-depleted medium, respectively.

For radioimmunoprecipitation, labeled cells were washed with phosphate-buffered saline and lysed in NET (0.1 M NaCI, 0.01 M Tris, pH 7.4, 0.001 M EDTA) with 0.5% NP-40. Approximately 1 O6 cpm was reacted with either preimmune or immune antibody overnight at 4'. Antigen-antibody complexes were bound to protein A-Sepharose (Pharmacia) and purified away from unbound, labeled protein by extensive washing with PBS, pH 8.6, containing 0.1% bovine serum albumin, 0.1% SDS, 0.1% NP-40, and 0.1% sodium azide. Complexes were dissociated by boiling in Laemmli buffer (Laemmli, 1970) and the labeled products were analyzed by electrophoresis through polyacrylamide gels (Maizel, 1971) . Gels were prepared for autoradiography as described previously (Denison and Perlman, 1986) . In vitro translation products were diluted into 10 mM phosphate, 0.4 M NaCI, 1% aprotinin (Sigma), 0.5% NP-40, pH 7.4, and bound for 2 hr at 4" to preformed antibody-protein A-Sepharose complexes. The resulting antigen-antibody-protein A-Sepharose complexes were then washed as described above for cell lysates.

For two-dimensional gel electrophoresis, cells were directly lysed in lysis buffer as described previously (O'Farrell, 1975; Denison and Perlman, 1987) . Samples were analyzed by two-dimensional gel electrophoresis as described previously, except that the first dimension was an isoelectric focusing gel containing ampholines 3.5-l 0 and 5-7 at final concentrations of 0.4 and 1.60/o, respectively, and the second dimension was a 15% SDS-polyacrylamide gel.

Sequence analysis of the plasmid 9344 revealed a TaoI site (position 412 using the coordinate system of Budzilowicz and Weiss, 1987) at the junction of the two open reading frames in MHV-A59 gene 5. Cleavage of 9344 at this site and at the Ndel and EcoRV sites at positions 88 and 664, respectively, allows the convenient separation and subsequent separate expression of the two open reading frames.

The TagI-EcoRV fragment of MHV gene 5 containing ORF 2 was inserted into the BarnHI site of pGE372 as illustrated in Fig. 2 . The sequences of the vector and the MHV insert predict that those inserts which are ligated in the correct orientation will be in frame with the N-terminal portion of the recA gene and also preserve the reading frame of the /acZgene. Therefore bacteria containing this construct will synthesize a tribrid protein consisting of the first 36 amino acids of the RecA protein, the amino acids encoded by the second open reading frame of MHV gene 5, and the 1015 C-terminal amino acids of fl-galactosidase. Bacteria containing this plasmid will form blue colonies in the presence of an appropriate chromagen such as X-Gal. Bacteria containing plasmids with the MHV insert in the opposite orientation will not express P-galactosidase activity due to encountering a termination codon within the MHV insert. pGE372 is an in frame recA-/acZ fusion and thus transformants containing this plasmid will also be P-galactosidase positive. Therefore we selected colonies containing the correct construction by the presence of &galactosidase activity and the presence of the MHV insert as detected by colony hybridization.

Several of these colonies were analyzed further. Small-scale plasmid preps were digested with BarnHI, electrophoresed on an agarose gel, and blotted to nitrocellulose.

These blots were then probed for the presence of MHV sequences using a random primed cDNA probe prepared from MHV-A59 genomic RNA. The results of this analysis for the clone used in all subsequent experiments, designated pGEA59.G5.2, are shown in Fig. 3 . This clone carries a plasmid containing an insert of the predicted size which strongly hybridized with the MHV probe. Protein extracts were prepared from this bacteria carrying this plasmid, unmodified pGE372, or pGE374 as described under Materials and Methods. These extracts were then electrophoresed on 8% polyacrylamide gels and the resolved proteins were visualized by Coomassie blue staining. A band representing the putative RecA-MHV-LacZ tribrid protein was tentatively identified by comparison with pGE372, which directed the synthesis of the RecA-LacZ fusion protein, and with pGE374, which did not since the RecA-LacZ fusion in this vector was out of frame. No difference in electrophoretic mobility could be demonstrated between the putative tribrid protein and the RecA-LacZ fusion protein synthesized in bacteria bearing the plasmid pGE372 (data not shown). To demonstrate that synthesis of the putative RecA-ORF 2-LacZ fusion protein was initiating at the desired site, we performed a Western blot analysis of protein extracts of pGE372, pGE374, and clone pGEA59.G5.2 (Fig. 3B) . Inspection of the amido black stained filter allows the identification of a protein present in extracts of pGE372 and pGEA59.G5.2 and missing from pGE374. The apparent molecular weight of this protein was about 120 kDa, the predicted molecular weight of the RecA-LacZ fusion present in pGE372. An identical filter was probed for the pres- lanes 1 and 4) . pGE372 (lanes 2 and 5) and pGE374 (lanes 3 and 6) . Bacterial extracts were prepared, electrophoresed on an 8% SDS-polyacrylamide gel, and transferred to nitrocellulose as described under Materials and Methods. Lanes l-3 were stained with amido black. Lanes 4-6 were stained with rabbit anti-RecA antibody used at a dilution of 1: 1000. The positions of molecular weight markers are indicated on the left. ence of RecA antigen using a rabbit anti-RecA antibody. As shown in Fig. 3B , the 120-kDa protein synthesized by pGE372 and pGEA59.G5.2, but absent in extracts of bacteria containing pGE374, reacted with the anti-RecA antibody. Thus we concluded that the plasmid pGEA59.G5.2 contained the desired MHV sequences inserted into the BarnHI site between the recA and /acZ sequences, conferred B-galactosidase activity on bacteria containing it, and directed the synthesis of a protein which reacts with anti-RecA antisera and therefore presumably contains RecA sequences at its N-terminus. This led us to conclude that the putative tribrid protein did in fact express the second open reading frame of MHV-A59 gene 5.

Antisera to the gene 5 ORF 2 containing tribrid protein were prepared and initially characterized by immunoblotting against bacterial extracts containing the immunogen used to raise these antibodies. This test did not distinguish anti-MHV antibodies from anti-RecA or anti+-galactosidase antibodies. Therefore when sera demonstrated reactivity in this assay they were then tested by radioimmunoprecipitation of the MHV-specific products encoded in gene 5. Transcripts of MHV-A59 gene 5 ORF 2 were produced in vitro from subclones of this gene in the vector pGEM-1 as described previously (Budzilowicz and Weiss, 1987) and translated in a wheat germ cell-free system. The in vitro translation products were immunoprecipitated with hyperimmune and preimmune sera. The resulting precipitates were then analyzed by SDS-polyacrylamide gel electrophoresis. As shown in Fig. 4 , the sera raised against the tribrid protein containing the second open reading frame react with the 9.6-kDa cell-free translation product corresponding to that reading frame.

The product of the second open reading frame contained in MHV gene 5 could be demonstrated by immunoperoxidase staining of infected cells, by immunoprecipitation and analysis of products by one-dimensional gel electrophoresis, and by two-dimensional gel electrophoresis. As shown in Figure 56 , MHV-infected cells were stained by the immunoperoxidase procedure with sera raised against the tribrid protein containing ORF 2. These same sera failed to react with uninfected cells (Fig. 5C ). Preimmune sera from the same rabbit did not stain infected cells (A).

The ORF 2 product synthesized in the cell-free translation system migrated as a single spot when analyzed by two-dimensional gel electrophoresis (Fig.  6A ). When infected cell lysates labeled at late times p.i. 9.6 K abc FIG. 4. lmmunoprecipitatron of in vitro synthesized ORF 2 product. ORF 2 RNA, transcribed from a recombinant pGEM vector, was translated in a wheat germ cell-free system. The [3H]leucine-labeled protein products were immunoprecipitated with preimmune serum (lane b) or anti-ORF 2 serum (lane c). Lane a shows all of the translation products before immunoprecipitation.

Electrophoresis was on 8-l 6% gradient gels. with [35S]methionine were analyzed by the same technique, a spot corresponding to the cell-free product was readily detectable (Fig. 6D ). This protein comigrated with the cell-free product, as shown in Fig. 6E , in which the cell-free product and infected cell lysate were mixed prior to analysis. The ORF 2 protein was not present in uninfected cells, as shown in Figs. 6B and 6C.

The identity of the in viva and cell-free products was confirmed using the antibody directed against the ORF 2 product. For these experiments, infected and uninfected cells were labeled with [35S]cysteine and reacted with the anti-ORF 2 antibody. Precipitated labeled protein was isolated and analyzed by one-dimensional gel electrophoresis, as shown in Fig. 7 . The anti-ORF 2 antibody reacted with a protein from infected cells (lane 4) which comigrated with the ORF 2 cell-free product (lane 5). This protein was not present in uninfected cells (lane 2) and was not detected when cells were labeled at early times (3-4 hr) after infection (lane 3). This protein could not be precipitated from infected cells with preimmune sera. A protein with the same approximate mobility could be detected in infected (lane 6) but not uninfected cell lysates without immune precipitation (data not shown).

In this paper we have constructed a plasmid encoding a fusion protein containing the second open read-ing frame of MHVA59 gene 5. We have used the tribrid protein synthesized by E. co/i carrying this plasmid to raise antisera to the protein encoded by this open reading frame. This antiserum was then used to dem- FIG. 5 . Detection of ORF 2 protein product by rmmunoperoxrdase staining. Infected and mock-infected cells were deposited on slides by cytocentrifugation and fixed by Immersion in 5% paraformaldehyde for 5 min. Slides were starned by the ABC lmmunoperoxrdase technrque. (A) Infected cells stained with prermmune serum. (5) Infected cells stained wrth anti-ORF 2 serum (C) Uninfected cells stained wrth antr-ORF 2 serum. onstrate that the second open reading frame was translationally active during infection with MHV, resulting in the synthesis of a protein which could be demonstrated by immunocytochemistry. Furthermore, a protein which corresponds to the in vitro product of the second open reading frame, ORF 2, could be demonstrated by two-dimensional gel electrophoresis of infected cell extracts and by immunoprecipitation using anti-ORF 2 antibody. The use of antibodies to a protein derived in part from a MHV cDNA clone ensures that the 9.6-kDa protein identified by radioimmunoprecipitation is virus-encoded.

Our results are consistent with those of Skinner ef a/. (1985) who showed that a 9-to 1 0-kDa polypeptide was present in MHV-JHM-infected cells but not in uninfected controls. Cell-free translation of size fractionated RNA isolated from infected cells was consistent with this protein being encoded in gene 5, although it could not be definitively determined that this protein was virus-encoded, in part due to the unavailability of specific antisera. Eukaryotic mRNAs, in general, are monocistronic and initiate protein synthesis at the most 5' AUG. Although the utilization of a downstream open reading frame, as we have documented for MHV gene 5, is unusual, it is not unique. Other examples of RNA viruses which utilize a downstream reading frame are reovirus (Ernst and Shatkin, 1985) vesicular stomatitis virus (Herman, 1986) and Sendai virus (Curran et a/., 1986) . In these cases the upstream reading frame is also utilized. At present there are no data indicating that the gene 5 upstream reading frame is expressed in MHV-infected cells.

The ORF 2 polypeptide appears to be expressed in relatively low amounts during the MHV replication cycle (Fig. 7, lanes 6 and 7) . This could be due to regulation at the transcriptional and/or translational level. MHV RNA 5 is one of the less abundant viral 7 . lmmunoprecipitation of gene 5 ORF 2 product from Infected cells. MHV-infected cells and uninfected L-2 cells were resuspended in DMEM lacking cysteine at the times indicated below and labeled for 1 hr with [35S]cysteine. Cytoplasm was then prepared and reacted with antibody as described under Materials and Methods prior to analysis by electrophoresis on 15% SDS-polyacrylamlde gels. Lane 1, infected cell extract (labeled 6.5-7.5 hr p.1.) precipitated with preimmune serum. Lane 2, uninfected cell extract precipitated with anti-ORF 2 serum. Lanes 3 and 4. infected cell extracts labeled at 3.5-4.5 hr p.i. and 6.5-7.5 hr p.i.. respectively, and preclpltated with anti-ORF 2 serum. Lane 5, cell-free translation product of gene 5 ORF 2 gene. Lanes 6 and 7, total cytoplasmic extract of cells labeled at 6.5-7.5 hr p.i. The positions of molecular weight markers are indicated on the left and the El and N viral structural proteins are indicated on the right. Lane 7 was exposed for l/l0 as long as the rest of the gel. , 1981) . In regard to translational regulation, the sequence found near the initiation site of protein synthesis in gene 5 ORF 2, GAAAUGU, is not a good match for the Kozak eukaryotic consensus initiation sequence A(G)CCAUGG (Kozak, 1984) . The most important divergence is at the strongly conserved -3 position. A is present at this position with a frequency of 79%, and G with a frequency of 18% in eukatyotic mRNAs. However, other MHV genes are better matches. Gene 6 (El) and gene 7 (N) have initiation codons embedded in the sequences AUUAUGA and AGGAUGA (Spaan et al., 1983) , respectively. These are sequences which are frequently used by eukary otic mRNAs.

The role of the MHV gene 5 ORF 2 protein in the infected cell is unknown. The coronavirus IBV is predicted to direct the synthesis of two small proteins encoded by overlapping reading frames in a fashion similar to that of MHV gene 5 (Boursnell and Brown, 1984) . One of these is quite hydrophobic and is a likely counterpart to the protein encoded in MHV gene 5 ORF 2. Other families of RNA viruses have recently been demonstrated to encode small hydrophobic proteins as well. Influenza virus induces the synthesis of the nonstructural M* protein, an integral membrane protein (Zebedee et al., 1985) . The paramyxovirus SV5 also encodes a small extremely hydrophobic protein (Hiebert er a/., 1985) . The functions of all of these proteins in viral replication are unknown at present. The approach that we used in this paper to construct a plasmid which directed the synthesis of a tribrid protein suitable for raising antibodies to ORF 2 required knowledge of the sequence of gene 5. We have recently modified our approach to make it more generally useful, even when the sequence of the gene in question is not known. Randomizing the ends of the insert and vector by homopolymer tailing or by nuclease digestion prior to ligation will produce the appropriate gene fusion in about 6% of the transformants obtained. Furthermore, the ability to probe for the presence of the correct N-terminus of the tribrid protein by immunoblot analysis with anti-RecA antibody, the correct C-terminus by the presence of P-galactosidase activity, and the insert by Southern hybridization provides assurance that the correct construction has been achieved.

Sequence and topology of a model intracellular membrane protein, El glycoproteln. from a coronavirus

Sequence of the nucleocapsid gene from coronavirus MHV-A59

Characterization of leader-related small RNAs In coronavirus-infected cells: Further evidence for leader-primed mechanism of transcription

A rapid alkaline extraction procedure for recombinant plasmid DNA

Sequencing of the coronavirus IBV genomic RNA: A 95-base open reading frame encoded by mRNA gene B

Completion of the sequence of the genome of the coronavlrus avian infectious bronchitis virus

ln wtro synthesis of two polypeptldes from a nonstructural gene of coronavirus mouse hepatitis virus strain A59

Three intergenic regions of coronavirus mouse hepatitis virus strain A59 genome RNA contain a common nucleotlde sequence that IS homologous to the 3'end of the viral mRNA leader sequence. 1. Viral

Intracellular murine hepatitis virus virus-specific RNAs contain common sequences

Effect of growth conditions on the formation of the relaxation complex of supercoiled ColEl deoxyribonucleic acid and protein in E

Ribosome initiation at alternative AUGs on the Sendai virus P/C mRNA

Prolonged incubation in calcium chloride improves the competence of Eschericia co/i cells

Translation and processing of mouse hepatitis virus virion RNA in a cell-free system

Identification of putative polymerase gene product in cells infected with murine coronavirus A59

Reovirus hemagglutin mRNA codes for two polypeptides in overlapping reading frames

Studies on transformation of E. co/i with plasmids

Internal initiation of translation on the vesicular stomatitis virus phosphoprotein mRNA yields a second protein. 1. Viral

Identification and predicted sequence of a previously unrecognized small hydrophobic protein SH, of the paramyxovirus simian virus 5

Compilation and analysis of sequences upstream from the translational start site in eukaryotic mRNAs

Functional messenger RNAs are produced by SP6 in vitro transcription of cloned cDNAs

Cleavage of structural proteins during the assembly of the head of bacteriophage T4

Characterization of leader RNA sequences on the virion and mRNAs of mouse hepatitis virus, a cytoplasmic RNA virus

Mouse hepatitis virus A59: mRNA structure and genetic localization of the sequence divergence from hepatotropic strain MHV-3

Presence of leader sequences in the mRNA of mouse hepatitis virus

The RNA of mouse hepatitis virus

Cell-free translation of murine coronavirus RNA

The virus-specific intracellular RNA species of two murine coronaviruses: MHV-A59 and MHV-JHM

Polyacrylamide gel electrophoresis of viral proteins

Molecular Cloning: A Laboratory Manual

High resolution two-dimensional electrophoresis of proteins

The origin of altered cell lines from mouse, monkey, and man as indicated by chromosome transplantation studies

Viral protein synthesis in mouse hepatitis virus strain A59-infected cells: Effect of tunicamycin

Translation of three mouse hepatitis virus strain A59 subgenomic RNAs in Xenopus laevis oocytes. 1. Viral

Coronavirus JHM: Coding assignments of subgenomic mRNAs

Coronavirus MHV-JHM mRNA 5 has a sequence arrangement which potentially allows translation of a second, downstream open reading frame

Coronavirus JHM nucleotide sequence of the mRNA that encodes nucleocapsid protein

Coding sequence of coronavirus MHV-JHM mRNA 4

Isolation and identification of virus-specific mRNAs in cells infected with mouse hepatitis virus (MHV-A59)

Coronavirus mRNA synthesis involves fusion of non-contiguous sequences

Sequence relationships between the genome and the intracellular RNA species 1, 3, 6, and 7 of mouse hepatitis virus strain A59

Coronavirus multiplication strategy. I. Identification and characterization of virus-specified RNA

Coronavirus multiplication strategy. II. Mapping the avian infectious bronchitis virus intracellular RNA species to the genome

Enhanced growth of a murine coronavirus in transformed mouse cells

Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: Procedures and applications

Characterization of murine coronavirus RNA by hybridization with virus-specific cDNA probes

Amino-terminal sequence analysis of alphavirus polypeptides

Characterization of the influenza virus M2 integral membrane protein and expression at the infected cell surface from cloned cDNA. 1. Viral

The authors acknowledge Davtd Greenstain for performing the immunocytochemistry and Dr. Charles Lawrence and other members of the Baylor Molecular Biology Information Resource for providing software for the computer analysis of sequence data. This work was supported by Public Health Service Grants NS 20834, Al 17418, NS 21954, GM 35427, and NS 24401.