Virus Attenuation after Deletion of the Cytomegalovirus Fc Receptor Gene Is Not due to Antibody Control

Cells and virus.

Mouse NIH 3T3 cells (ATCC CRL1658) were grown in Dulbecco’s modified Eagle’s medium supplemented with 10% (vol/vol) newborn calf serum. A immortalized cell line of mouse fibroblasts (B12 cells) (7) was grown in minimum essential medium with 10% fetal calf serum. Third-passage mouse embryo fibroblasts, prepared from BALB/cJ mice and grown in minimum essential medium supplemented with 10% fetal calf serum, were used for virus infection. The Smith strain (ATCC VR-194) of MCMV was used as a tissue culture-grown virus.

Recombinant plasmids.

Plasmid cloning was done by standard methods (26). To generate the recombination plasmid pJAR4, a lacZ cassette driven by the Rous sarcoma virus promoter from plasmid pATLacZ (a gift of G. Darai, University of Heidelberg, Heidelberg, Germany) was inserted between the BssHII and BglII sites of the phagemid pREG16. This plasmid contains a 2.4-kb BamHI-BamHI subfragment of the MCMV HindIII J fragment encompassing the fcr-1 gene (38).

To flank the lacZ marker with loxP sites (35), a polylinker (5′-CTAGGAAGCTTGATATCGAATTCCTGCAGATCTG-3′) was inserted into a XbaI site between tandem loxP sites of plasmid pP4 (a gift from T. Boehm, DKFZ, Heidelberg, Germany) and then the lacZ cassette was cloned between the HindIII and BglII sites. The loxP-flanked lacZ cassette was released by digestion with NotI and BamHI and inserted into the BssHII and BglII sites of plasmid pREG16 by using a BssHII-NotI adapter (5′-GGCCACATGCCGATGG-3′), resulting in the recombination plasmid pIC4.

The Kunkel method of site-directed mutagenesis (22) was used to introduce a silent mutation into the fcr-1 gene. To delete the XbaI site of the pREG16 phagemid, single-stranded DNA was purified from phages produced in the ung Escherichia coli strain CJ236. The oligonucleotide 5′-GTCTCTCTGGACCTCTC-3′ was annealed to the single-stranded DNA, and the complementary strand was synthesized by T7 DNA polymerase. After electroporation of the double-stranded phagemids into ung⁺ E. coli, mutagenized phagemids were identified by the absence of the XbaI site. The correct alteration was confirmed by sequencing of the mutated region.

Generation of a recombinase cre⁺ cell line (N2).

A plasmid designated pneocre2 was generated by introducing a XhoI fragment containing the recombinase cre gene from pMC-Cre (16; a gift from K. Rajewsky, Cologne, Germany) into the XhoI site of the plasmid pUC21neo, which contains a neomycin resistance gene driven by the simian virus 40 enhancer-promoter. The cre recombinase-encoding plasmid pneocre2 (10 μg) was transfected into NIH 3T3 cells, and G418 resistant clones were isolated and tested for cre recombinase activity.

Construction of recombinant MCMV.

The linearized recombination plasmid (30 fmol) and intact viral DNA (1.5 μg) were cotransfected into NIH 3T3 fibroblasts by the calcium phosphate precipitation technique (27). Recombinant viruses were isolated from the resulting virus progeny by at least five rounds of limiting dilution followed by 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside (X-Gal) staining. The lacZ mutants were generated by a single passage through the recombinase cre⁺ cell line (N2) and purified by limiting dilution after X-Gal color screening.

DNA analysis.

MCMV DNA was prepared from infected NIH 3T3 cells. Five days after infection, the cells were rinsed with phosphate-buffered saline, overlaid with lysis buffer (100 mM NaCl, 10 mM Tris [pH 8.0], 10 mM EDTA, 0.1% sodium dodecyl sulfate, 0.25 mg of proteinase K per ml), and incubated for 3 h at 37°C. Lysates were extracted twice with an equal volume of phenol and then once with phenol-chloroform-isoamyl alcohol (25:24:1) and chloroform-isoamyl alcohol (24:1). After addition of 2 volumes of isopropanol, the DNA was spooled on a glass rod, dried, and resuspended in 10 mM Tris–1 mM EDTA (pH 7.5).

For Southern blot analysis, approximately 1 μg of viral DNA was digested with restriction enzymes, separated by gel electrophoresis, blotted to positive-charged nylon membrane (Qiagen, Hilden, Germany), and hybridized with a digoxigenin-labeled probe. The bound probe was detected with an alkaline phosphatase-coupled anti-digoxigenin antibody and visualized by a chemiluminiscence method (Boehringer, Mannheim, Germany).

Flow cytometry.

For detection of FcR on the cell surface, MCMV-infected cells (16 h postinfection [p.i.]) were detached from the plastic with PBS–2 mM EDTA and incubated for 30 min with murine IgG (Dianova, Hamburg, Germany) at 100 μg/ml. Fluorescein isothiocyanate (FITC)-conjugated goat anti-mouse IgG F(ab′)₂ (Dianova) served as a secondary reagent.

Immunoprecipitation.

B12 cells were infected either with wild-type MCMV or with FcR recombinants at a multiplicity of infection of 20. At 16 h p.i., the cells were pulse-labelled for 45 min with 150 μCi of [³⁵S]methionine (Amersham, Braunschweig, Germany) per ml in methionine-free minimal essential medium supplemented with 5% dialyzed fetal calf serum. Labelled cells were lysed in a buffer containing 10 mM Tris-HCl (pH 7.4), 100 mM NaCl, 1% Nonidet P-40, 1 mM EDTA, and 1 mM phenylmethylsulfonyl fluoride. Cytoplasmic extracts were precleared twice with protein A-coupled Sepharose (Pharmacia, Uppsala, Sweden) and precipitated with murine IgG-coated protein A-Sepharose. Bound complexes were eluted by heating at 96°C for 5 min in reducing sodium dodecyl sulfate-polyacrylamide gel electrophoresis sample buffer and analyzed by polyacrylamide gel electrophoresis (9% polyacrylamide).

Animals and infection conditions.

Newborn BALB/c mice (H-2^d), B-cell-deficient mice (μMT/μMT), and their heterozygous littermates (μMT/+) (H-2^b) (20) from the breeding colony at the Medical Faculty, University of Rijeka, were injected intraperitoneally with 10³ PFU of either wild-type or recombinant MCMV. Homozygous and heterozygous B-cell-deficient animals were selected by an enzyme-linked immunosorbent assay for the presence of the IgM in sera, as described previously (18).

Adult, immunocompetent BALB/c mice were subcutaneously injected with 10⁵ PFU of wild-type or recombinant viruses.

C57BL/6 mice (6 to 8 weeks old) were depleted of immune cells and injected by either the intravenous or subcutaneous (hind footpad) route with 10⁵ PFU of wild-type or mutant virus. The depletion was performed by weekly injection of 1 mg of purified monoclonal antibodies to CD4 (YTS 191.1.2), CD8 (YTS 169.4.2) (6), and NK1.1 (PK136) (21).

Adult B-cell-deficient mice and their heterozygous littermates were depleted of CD4⁺, CD8⁺, and NK1.1⁺ cells and additionally treated with cyclophosphamide (150 mg/kg).

The animals were sacrificed at the indicated time p.i. Virus titers in organ homogenates were determined on mouse embryo fibroblasts by a standard plaque assay (29).

Plasma membrane expression of the MCMV FcR.

In previous experiments, the intracellular location of the FcR function was used to identify the gene (38). To fulfill an immunomodulatory function during infection, the FcR activity has to be present at the cell membrane. Therefore, we tested the surface expression of the viral FcR under various conditions. We found that trypsin addition to cells for removal of adherent cells destroyed the plasma membrane expression of the FcR activity. In the absence of trypsin, however, the binding of monomeric murine IgG can be demonstrated on the surface of infected fibroblasts (Fig. 1). Thus, the FcR activity is located at the site of potential immunological function.

Generation of recombinant MCMV.

To analyze the role of the MCMV FcR in virus-host interactions, we constructed several FcR deletion mutants (illustrated in Fig. 2a). The recombinant virus ΔMS94.4 was generated by homologous recombination with the recombination plasmid pJAR4. In this plasmid, the 1.35-kb BssHII-BglII fragment of the fcr-1 gene, including the promoter and the coding region for first 416 amino acids of the open reading frame, was deleted and replaced by the E. coli lacZ gene. The recombinant virus was isolated by five rounds of limiting dilution prior to generation of virus stocks.

To generate a revertant virus, we reinserted the fcr-1 gene in the ΔMS94.4 deletion mutant. To be able to differentiate between wild-type and revertant rMS95.9, we introduced a silent mutation into the coding sequence of the fcr-1 gene. The introduced mutation (A → G) eliminated a single restriction site (XbaI) and changed the restriction pattern of the rescued virus. The resulting mutant is easily distinguishable from the wild-type virus (Fig. 2c).

Since the fcr-1 deletion mutant carries the foreign gene lacZ, a revertant to wild-type properties would not formally distinguish biological effects caused by the deletion of the fcr-1 gene from effects caused by the insertion of lacZ into that gene position. We therefore created a second recombination plasmid, pIC4. In this plasmid, the lacZ gene is flanked by loxP sites (32). The deletion mutant ΔMC95.16 was isolated as above and served two purposes. It was used as a second fcr-1 deletion mutant prepared independently and was also used to create a deletion mutant without a marker gene. To this end, the ΔMC95.16 recombinant was passaged in the recombinase cre⁺ cell line N2. As expected, the marker gene was efficiently excised from the viral genome and the recombinant virus ΔMC95.15, without the lacZ gene, was generated.

Southern blot analyses of the isolated and plaque-cloned viruses confirmed that the recombination occurred at the expected position (Fig. 2b). In all deletion mutants, the original HindIII J fragment (8.2 kb) was replaced by new HindIII fragments. The sizes of the fragments corresponded to the sizes predicted from the distribution of HindIII sites in the MCMV genome and in the recombination plasmids. In the ΔMS94.4 recombinant, new HindIII fragments of 1.7 and 9.7 kb were present, while in ΔMC95.15 a 6.9-kb fragment was found and in ΔMC95.16 5.1- and 6.2-kb fragments were found. The difference between the fragment pattern of wild-type and rescued virus rMS95.9 after XbaI digestion is shown in Fig. 2c. As expected, the XbaI N (4.1 kb) and G (13.0 kb) fragments, visible in the wild-type virus, were replaced by a single 17.1-kb fragment in rMS95.9. The comparison of the HindIII, EcoRI, and XbaI restriction fragment patterns of the recombinants with those of wild-type MCMV confirmed that the recombinant viruses were free of detectable deletions or insertions in any other region of the viral genome (data not shown).

Expression of MCMV FcR in wild-type, deletion mutant, and rescued virus.

To verify that the recombinant viruses with the deletion of the fcr-1 gene do not express a protein with FcR properties, proteins from MCMV-infected and [³⁵S]methionine-labelled cells were precipitated with mouse IgG. In cells infected with wild-type MCMV and with fcr-1 rescued virus, proteins with molecular masses of 86 to 88 kDa and 105 kDa, corresponding to the MCMV FcR (38), were detected (Fig. 3). These proteins were absent in cells infected with ΔMS94.4, ΔMC95.15, and ΔMC95.16, confirming the correct deletion of the FcR-encoding gene. In addition, it indicated the absence of any other proteins with FcR properties encoded by MCMV whose function could be disguised in the wild-type virus by the fcr-1 gene product.

In vitro growth of MCMV mutants.

To assess whether the fcr-1 gene affects the growth of virus in cell culture, single-step growth curves of recombinant and wild-type viruses were determined. After infection of NIH 3T3 fibroblasts at a multiplicity of infection of 0.1 PFU per cell, the replication kinetics of recombinants were indistinguishable from that of the wild-type MCMV (Fig. 4). Also, the size and morphology of viral plaques were identical in recombinant and parental viruses, indicating that the fcr-1 gene is dispensable for growth in cultured fibroblasts.

Reduced growth of fcr-1 deletion mutants in newborn mice.

To investigate the influence of the fcr-1 gene deletion on in vivo replication, we infected 6-week-old BALB/c mice with wild-type and fcr-1 deletion mutant viruses. The tissues of mice infected with the fcr-1 deletion mutants contained very low virus titers that made quantitative comparisons difficult (data not shown). Therefore, we studied virus replication in the immunologically immature newborn mice, which are highly susceptible to MCMV infection (30). At 8 days p.i. the wild-type virus reached high titers in the organs of new-born mice (Fig. 5). At the same time, we observed a more than 100-fold-reduced replication of the fcr-1 deletion mutants in the lungs, liver, and spleen. Propagation of the rMS95.9 revertant virus was indistinguishable from that of the wild-type virus, confirming that the attenuation of the ΔMS94.4 recombinant virus is entirely due to the site-specific mutation.

Altered biological properties of a virus mutant can be caused by the deletion or the insertion of a marker gene or both. Reduced growth in salivary glands of the lacZ⁺ MCMV recombinant without any other mutation has been described by Stoddard et al. (36). The authors excluded the immune response to β-galactosidase as a basis of reduced replication, but the mechanism of the observed phenomenon remained unclear. Therefore, the deletion mutant lacking the marker gene was tested. There was no difference between the titers of fcr-1 lacZ⁺ (ΔMS94.4) and fcr-1 lacZ (ΔMC95.15) recombinants (Fig. 5). The ΔMC95.16 recombinant virus, which was used to create the marker-free mutant, reached titers comparable to those of the other two fcr-1 deletion mutants (data not shown). Thus, at least in this case, the lacZ marker gene did not contribute to the attenuation of the mutant.

Growth of the fcr-1 deletion mutant in B-cell-deficient mice.

According to the hypothesis of bipolar bridging (13), the FcR should protect the virus and virus-infected cells from effects mediated through the Fc portion of Igs after binding to specific antigen. We reasoned that if the attenuated phenotype of the fcr-1 deletion mutants was due to the loss of an immunoevasive function mediated by antibody binding, the mutant virus should grow comparably to wild-type MCMV in mutant mice that lack B cells (20). Immunocompetent littermates heterozygous for the μ-chain mutation (μMT/+) severely restricted the growth of the MCMV mutant. The MCMV-specific antibody response of newborn heterozygous animals on day 21 p.i. was comparable to the response of adult mice (data not shown). Notably, the growth of the fcr-1 deletion mutant was inhibited to the same degree in μMT/μMT mice, which completely lack B cells and Igs (Fig. 6). Thus, the capacity to bind the Fc portion of Ig cannot represent the principle which dictates the attenuated phenotype of the fcr-1 deletion mutants.

Reduced growth of fcr-1 deletion mutants in T- and NK-cell-depleted mice.

After depletion of T cells and NK cells, the C57BL/6 mice became increasingly susceptible to MCMV infection. Although the mutant virus showed marginal growth in immunodepleted juvenile mice, the significant difference between the growth of mutant and wild-type viruses and the growth of revertant virus remained evident. This suggested that the viral FcR does not interfere with the function of T cells and NK cells. Another explanation for reduced virulence could be the inability of the mutant virus to spread from the site of primary infection to the target organs. Therefore, we compared the virus titer that viruses reached in different organs after footpad and intravenous infection. The infection of the visceral organs and the salivary gland by the fcr-1 gene deletion mutant occurred at a low rate irrespective of the route of infection (Fig. 7). We therefore concluded that an inability of the mutant virus to spread hematogenously is not likely to explain the data.