A Recombinant Classical Swine Fever Virus Stably Expresses a Marker Gene

Abstract

The gene coding for bacterial chloramphenicol acetyltransferase (CAT) was inserted in frame into the viral N^pro gene of the full-length cDNA clone pA187-1 of the classical swine fever virus (CSFV) strain Alfort/187. RNA transcribed in vitro from the resulting plasmid was transfected into SK-6 porcine kidney cells. Infectious progeny virus vA187-CAT recovered from transfected cells had growth characteristics indistinguishable from those of parental virus vA187-1. In cells infected with vA187-CAT the predicted fusion protein, CAT-N^pro, was detected, and it retained the enzymatic activities of both CAT and N^pro. The CAT gene remained stably inserted in the viral genome after 10 virus passages. Thus, marker virus vA187-CAT represents a useful tool for quantitative analysis of viral replication and gene expression.

Classical swine fever virus (CSFV) is a small, enveloped virus with a positive-stranded RNA genome 12.3 kb in length. The genome is composed of a 5′ untranslated region (5′ UTR) containing an internal ribosomal entry site (IRES), a single large open reading frame (ORF) encoding the viral polyprotein, and a 3′ UTR (10, 16–18). Together with bovine viral diarrhea virus (BVDV) and border disease virus, CSFV forms the genus Pestivirus within the family Flaviviridae (25, 28). Several infectious cDNA clones of pestiviruses have been constructed (9, 11, 13, 22, 27). To allow differentiation of the helper virus genome from engineered genomes in complementation experiments and to quantify the effects of defined mutations on viral replication and gene expression, we decided to insert a reporter gene into the genome of CSFV strain Alfort/187 (2) via cDNA clone pA187-1 (22). Several BVDV isolates having inserts of cellular sequences in their genomes and/or duplications of viral sequences of up to 4 kb have been described (4, 8, 10), suggesting that the length of the genomic RNA is not a critical factor, either for replication or for packaging of these genomes. The bacterial chloramphenicol acetyltransferase (CAT) gene, which has been used successfully as a reporter gene in other RNA viruses (1, 6, 7, 15, 19, 23), was chosen because of its small size and because of the possibility of quantifying protein expression by measuring CAT enzyme activity.

Generation of vA187-CAT.

The CAT gene was amplified by PCR from a plasmid derived from pSV2CAT (26) with primers CAT-SmaL (5′-TCCCCCGGGGAGAAAAAAATCACTGGATAT-3′) and CAT-SmaR (5′-TCCCCCGGGCGCCCCGCCCTGCCACTC-3′), each containing SmaI restriction sites (underlined). The translation start codon of the CAT gene was replaced by GGG, and stop codon TAA was replaced by CCC. These triplets represented the ends of the PCR fragment after cleavage with SmaI. Plasmid pAC-A187/82-778 was derived from pACNR1180 (22) and contains an EagI-BspDI fragment corresponding to nucleotides (nt) 82 to 778 of the CSFV Alfort/187 genome. The TfiI site located at nt 383 served for insertion of the PCR-derived CAT gene. From the resulting plasmid, the EagI-BspDI fragment was recovered and used to replace the corresponding sequence in pA187-1 to give pA187-CAT. By DNA sequencing, the CAT gene was confirmed to have been inserted in frame at nt 386 of the viral cDNA, 9 nt downstream of translation initiation codon ATG (Fig. 1) and to be flanked by the sequence AAT. This triplet corresponds to the fourth triplet of the viral ORF and was duplicated by treatment of the TfiI-cleaved plasmid DNA with Klenow enzyme (Fig. 1). Infectious RNA was obtained by runoff transcription from SrfI-linearized pA187-CAT and was transfected into SK-6 cells as described before (22). After 30 h the cell culture supernatant was collected and passaged twice on SK-6 cells to obtain a virus stock (P2 stock) which had a titer of 10^7.7 50% tissue culture infective doses (TCID₅₀)/ml. The presence of the CAT gene in the viral genome was demonstrated by reverse transcription-PCR (RT-PCR) performed with RNA extracted from this virus stock (Fig. 4A and B, lane 2). Viral RNA extraction from the cell culture supernatant, RT, and PCR were performed as described before (12). Primer pairs used were either CAT-specific primers CAT-SmaL and CAT-SmaR or CSFV-specific primers Pest 1, corresponding to nt 310 to 329, and N^pro-R1, complementary to nt 877 to 856 of the genome. For detection of the CAT protein, infected cells were lysed with the hypotonic buffer contained in the CAT enzyme-linked immunosorbent assay (ELISA) kit provided by Boehringer. The hypotonic lysates also served to demonstrate the enzymatic activity of the virus-expressed CAT protein in a [¹⁴C]chloramphenicol assay (26) (data not shown). Attempts to establish a CAT-specific in situ staining of vA187-CAT-infected SK-6 cells failed. The commercially available CAT staining set (Boehringer) proved to be unsuitable, as only weak signals were obtained in cells infected with vA187-CAT. An indirect immunoperoxidase assay, which was carried out with the antibodies from the CAT ELISA kit, produced a high background which masked the expected specific staining.

FIG. 1.

Schematic representation of the genome of vA187-CAT. The authentic genes of the CSFV genome are shown as open boxes, and the CAT gene is shown as a shaded box. The lower part of the figure shows the insertion site and flanking sequences of the CAT gene in detail. The nucleotide numbers correspond to the numbering of the CSFV Alfort/187 genome and the CAT gene sequence.

FIG. 4.

RT-PCR of viral RNA after 10 passages of vA187-CAT. Viral RNA was extracted from cell culture supernatants and RT was carried out with primer HR3. (A) PCR with primers CAT-SmaL and CAT-SmaR resulting in a CAT gene-specific product of 672 bp; (B) PCR with primers Pest 1 and N^pro-R1 flanking the site of insertion of the marker gene. The lengths of the specific products are 1,231 and 568 bp for vA187-CAT and vA187-1, respectively. Lanes 1, vA187-CAT, P2 stock virus; lanes 2, vA187-CAT, virus passage 10; lanes 3, vA187-CAT, persistently infected cells, passage 10; lanes 4, vA187-1; lanes 5, mock-infected SK-6 cells; lanes 6, pA187-CAT DNA; lanes 7, pA187-1 DNA.

Growth and expression characteristics of vA187-CAT.

The replication characteristics of vA187-CAT were compared to those of the parent, cDNA-derived virus vA187-1. SK-6 cells were infected at a multiplicity of infection (MOI) of 5 for comparison of the one-step growth curves, and at MOIs of 0.5 and 0.005 for comparison of the rates of secondary infection. The kinetics of the virus titer as well as of CAT and E2 expression were determined (Fig. 2). An increase of the virus titer was first detected between 6 and 12 h post infection (p.i.) for both vA187-1 and vA187-CAT (Fig. 2A). Maximal titers reached were dependent on the MOI but not on the virus type. Similarly, the kinetics of CAT expression, as measured by a CAT ELISA of hypotonic lysates of infected cells, were found to be dependent on the MOI of the infecting virus (Fig. 2B). As expected, lysates from cells infected with vA187-1 were negative in the CAT ELISA. An E2 capture ELISA was established to allow measurement of the expression of an authentic viral gene product. The hypotonic cell lysates were incubated on microtiter plates coated with serum which had been obtained from a rabbit immunized with baculovirus-expressed and affinity-purified E2 protein of CSFV Alfort/187 (20, 21). This protein also served as a quantitative standard in the E2 ELISA. Captured E2 protein was detected with monoclonal antibody HC/TC 26 (3) and antimouse horseradish peroxidase conjugate (DAKO) and quantified by optical density measurement after reaction with ABTS (2,2′-azinobis[3-ethylbenzthiazolinesulfonic acid]; Boehringer). For hypotonic lysates of cells infected with vA187-CAT, the kinetics of E2 and of CAT expression did not differ significantly. Furthermore, at 48 h p.i. similar E2 concentrations were detected for vA187-1 and vA187-CAT at all three MOIs (Fig. 2B).

FIG. 2.

Comparative growth kinetics of vA187-1 and vA187-CAT. A total of 2 × 10⁶ SK-6 cells were infected at the indicated MOIs. (A) Virus titers in the cleared supernatant after freezing and thawing of infected cells. The values indicated at time zero represent the titers of the respective inoculates. (B) Expression of CAT and E2 proteins as measured by CAT and E2 ELISAs. The cells were lysed in 1 ml of hypotonic buffer and tested in duplicate. The y axis is logarithmic for better comparison with the corresponding virus titers shown in panel A. The detection limit of the CAT ELISA (0.1 ng/ml) is shown as a dashed line.

Stability of vA187-CAT.

To analyze the stability of the inserted CAT gene, vA187-CAT (P2 stock) was passaged 10 times in SK-6 cells by serial passage of either virus (virus passage) or persistently infected cells (cell passage). Initially, 5 × 10⁶ SK-6 cells were infected for 1 h at an MOI of 10. Each virus passage was done by inoculation of fresh SK-6 cells with 1/150 volume of the supernatant from the previous culture which had been incubated for 48 to 72 h p.i. Passage of persistently infected SK-6 cells was carried out twice weekly at a splitting rate of 8. In both cases, cell culture supernatants as well as hypotonic cell lysates were collected. In general, the virus titers in supernatants of persistently infected cells were lower (10^5.5 to 10^6.5 TCID₅₀/ml) than those of supernatants of acutely infected cells (10^6.1 to 10^7.9 TCID₅₀/ml), yet considerable fluctuations were observed (Fig. 3A). The hypotonic cell lysates were subjected to a CAT ELISA (Fig. 3B) and to an E2 capture ELISA (Fig. 3C). The concentrations of the respective proteins showed even greater variations than the virus titers. Notably, the CAT concentration decreased with increasing numbers of virus passages (Fig. 3B). RT-PCR was performed with viral RNA extracted from the cell culture supernatants of both virus passage 10 and cell passage 10. In both cases, a 672-bp PCR product specific for the inserted CAT gene could be amplified with primers CAT-SmaL and CAT-SmaR (Fig. 4A, lanes 2 and 3). Similarly, by using CSFV-specific primers Pest 1 and N^pro-R1, which flank the inserted CAT gene, the expected 1,231-bp fragment was obtained (Fig. 4B, lanes 2 and 3), whereas a 568-bp product was amplified from vA187-1 (Fig. 4B, lane 4). Direct cycle sequencing of the CAT inserts of both passage 10 viruses revealed the 100% homology of the CAT gene sequence and of the flanking CSFV sequences (nt 23 to 385 and nt 383 to 400 of Alfort/187) with those of parental plasmid pA187-CAT.

FIG. 3.

Virus titers (A) and CAT (B) and E2 (C) expression during passage of vA187-CAT. For each passage of vA187-CAT (virus passage) and of cells persistently infected with vA187-CAT (cell passage), the cell culture supernatant was titrated and the hypotonic cell lysate was analyzed by CAT and E2 ELISAs. The y axis is logarithmic in panels B and C for better comparison with the corresponding virus titers shown in panel A.

Both vA187-CAT passage 10 viruses were propagated once in fresh cells to obtain high-titer virus stocks (10^7.7 and 10^7.9 TCID₅₀/ml). These virus stocks, as well as vA187-CAT P2 stock (see above) and vA187-1, were used to infect SK-6 cells at an MOI of 10. Thirty hours p.i., cells were lysed in hypotonic buffer. CAT concentrations determined by ELISA were similar (15 to 20 ng/10⁶ cells) for all three vA187-CAT infections. The same extracts were used to perform duplicate sodium dodecyl sulfate-polyacrylamide gel electrophoresis and Western blotting, the products of which were analyzed either with the anti-CAT antibodies used for the CAT ELISA (Fig. 5A) or with rabbit serum R82 (Fig. 5B) directed against baculovirus-expressed CSFV N^pro-C (20). A protein with an apparent molecular mass of approximately 40 kDa corresponding in size to predicted fusion protein CAT-N^pro was detected with both antisera in all three samples derived from cells infected with vA187-CAT (Fig. 5A and B, lanes 3 to 5). By using rabbit antiserum R82, authentic N^pro, which has a predicted molecular mass of 23 kDa, and two additional proteins of 21 and 35 kDa were identified in cells infected with vA187-1 (Fig. 5B, lane 2). These proteins were not found in mock-infected cells, but the 35-kDa protein was also present in cells infected with vA187-CAT (Fig. 5B, lanes 3 to 5). Possibly, it represents unprocessed viral C-E^rns precursor protein, since antiserum R82 is also directed against the viral core protein (13a). Fusion protein CAT-N^pro apparently maintained its autoproteolytic activity since in Western blots no virus-specific proteins with molecular weights higher than that of the fusion protein were detected with either of the antibodies (data not shown). At longer exposure times of the blots, additional proteins with molecular masses ranging from 21 to 37 kDa were detected (Fig. 5A and 5B) but only in cells infected with vA187-CAT. Compared to the intensity of the 40-kDa fusion protein band, the intensities of the 21- and 37-kDa bands were greater in Western blots performed with cell lysates obtained 48 to 96 h after infection, as observed during virus and cell passage (data not shown). These findings suggest that fusion protein CAT-N^pro is degraded late in infection, a hypothesis which is further supported by the reduced CAT concentration in lysates obtained at later times p.i. (data not shown).

FIG. 5.

Western blot analysis of proteins extracted from SK-6 cells infected with vA187-CAT. Specific proteins in hypotonic cell lysates were detected with either anti-CAT antibodies (A) or anti-N^pro rabbit serum (B). Molecular weight markers (in thousands) are shown on the left. Lanes 1, CAT protein; lanes 2 to 6, lysates of SK-6 cells infected with vA187-1 (lanes 2), vA187-CAT, P2 stock virus (lanes 3), vA187-CAT, virus passage 10 (lanes 4), vA187-CAT, persistently infected cells, passage 10 (lanes 5), and lysate of mock-infected SK-6 cells (lanes 6). Arrows indicate the CAT-N^pro fusion (f), CAT (c), and N^pro (n).

Discussion.

The virus vA187-CAT described here represents the first recombinant pestivirus containing a marker gene in its genome. We considered that insertion of the foreign sequence close to the 5′ end of the viral N^pro gene was most likely to allow correct processing and transmembrane transport of the viral polyprotein. We also expected that fusion protein CAT-N^pro would retain the autoprotease activity which is located in the C-terminal half of N^pro (14, 24, 29). The initiation of translation has been shown to occur efficiently for defective genomes of CSFV (9, 12), although the translation start codon is followed directly by the NS3 gene, which has no homology with the N^pro gene. This suggests that the IRES does not extend into the ORF. Furthermore, Rijnbrand et al. (18) have shown for the C strain of CSFV that the IRES function is maintained after insertion of the CAT gene 18 nt downstream of the translation initiation codon.

To detect possible effects of the marker gene on the viral life cycle, we compared the replication kinetics and protein expression of recombinant vA187-CAT and wild-type vA187-1. The data indicate that vA187-CAT is indistinguishable from the parental vA187-1 with respect to the yield of progeny virus, to the expression of viral protein E2, and to the rate of secondary infection, as demonstrated by virus titration and quantification of E2 by ELISA. These findings suggest that the IRES function and the cis-acting signals required for RNA replication and packaging were fully retained in vA187-CAT. Furthermore, the kinetics of CAT and E2 expression closely paralleled each other, indicating that the concentration of CAT protein measured by ELISA reflects the level of viral protein expression (Fig. 2), at least early in infection. Later on, a significant proportion of expressed fusion protein CAT-N^pro apparently was degraded.

The position of the CAT gene at the 5′ end of the ORF certainly favors its genetic stability by negative selection, as any mutation within the CAT gene causing a translation stop or a frameshift would abolish further translation of the viral genome. Virus particles containing such mutant genomes would not be infectious and therefore would be eliminated upon passage. However, taking into account the estimated error rate of viral RNA-dependent RNA polymerases (5), mutations in the CAT gene without negative effects on the translation of the authentic viral ORF were expected to occur after repeated passage of the virus. Surprisingly, no mutations were found in the CAT gene after 10 passages of the virus or of persistently infected cells. Deletion mutants of vA187-CAT are expected to have an advantage during PCR amplification and, therefore, should be detected even if they were present only at a low concentration in the virus population. On the other hand, direct sequencing of RT-PCR products only reveals the predominant sequence in a virus population and thus does not exclude the presence of minor populations of a mutant virus.

Based on the findings that vA187-CAT shows wild-type-like properties in terms of replication and protein expression and is genetically stable, we conclude that the inserted marker gene has no significant effects on any of the viral functions required for cell culture propagation. Therefore, this marker virus should be useful for quantitative analysis of CSFV gene expression and for functional analysis of the CSFV genome. Such analyses are facilitated by the convenience of measuring CAT gene expression either by ELISA or by determination of CAT enzyme activity. We do not know if the CAT gene alters the properties of the virus as a pathogen. This question can only be addressed by experimental infection of pigs with wild-type and CAT marker virus.