Effects of an epitope-specific CD8+ T-cell response on murine coronavirus central nervous system disease: protection from virus replication and antigen spread and selection of epitope escape mutants

Abstract

Both CD4(+) and CD8(+) T cells are required for clearance of the murine coronavirus mouse hepatitis virus (MHV) during acute infection. We investigated the effects of an epitope-specific CD8(+) T-cell response on acute infection of MHV, strain A59, in the murine CNS. Mice with CD8(+) T cells specific for gp33-41 (an H-2D(b)-restricted CD8(+) T-cell epitope derived from lymphocytic choriomeningitis glycoprotein) were infected with a recombinant MHV-A59, also expressing gp33-41, as a fusion protein with enhanced green fluorescent protein (EGFP). By 5 days postinfection, these mice showed significantly (approximately 20-fold) lower titers of infectious virus in the brain compared to control mice. Furthermore mice with gp33-41-specific CD8(+) cells exhibited much reduced levels of viral antigen in the brain as measured by immunohistochemistry using an antibody directed against viral nucleocapsid. More than 90% of the viruses recovered from brain lysates of such protected mice, at 5 days postinfection, had lost the ability to express EGFP and had deletions in their genomes encompassing EGFP and gp33-41. In addition, genomes of viruses from about half the plaques that retained the EGFP gene had mutations within the gp33-41 epitope. On the other hand, gp33-41-specific cells failed to protect perforin-deficient mice from infection by the recombinant MHV expressing gp33, indicating that perforin-mediated mechanisms were needed. Virus recovered from perforin-deficient mice did not exhibit loss of EGFP expression and the gp33-41 epitope. These observations suggest that the cytotoxic T-cell response to gp33-41 exerts a strong immune pressure that quickly selects epitope escape mutants to gp33-41.

FIG. 1.

Recombinant viruses encoding EGFP and gp33. A schematic diagram of the genome of MHV-A59 is shown. Targeted recombination was used to replace gene 4 of MHV-A59 with the EGFP gene (RA59-gfp) or with a fragment of ORF4a, followed by gp33 and the EGFP gene (RA59-gfp/gp33), as described in the text. SalI and NotI sites were used to clone the gfp or ORF4a/gp33/gfp fragments into the pMH54 plasmids to generate synthetic RNA for recombination.

FIG. 2.

Epitope-specific CD8⁺ T-cell response of RA59-gfp/gp33-infected mice to gp33 and S598, as measured by intracellular IFN-γ expression. Four-week-old B6 mice were infected either i.p. with 10⁴ PFU (A) or i.c. with 2 × 10³ PFU (B). At either 8 (A) or 7 (B) days postinfection, mice were sacrificed, and mononuclear cells were harvested from the spleen and brain. IFN-γ assays were carried out to determine the percentage of CD8⁺ T cells specific for S598 or gp33 as indicated. Data shown are derived from pooled samples from three to five mice.

FIG. 3.

Schematic for protection experiment. Four-week-old mice were inoculated with rLm (10⁴ CFU) expressing gp33 (rLm-gp33) or np118 (rLm-np118). Two weeks later mice were infected with 10⁵ PFU RA59-gfp or RA59-gfp/gp33. Five days later mice were sacrificed. Virus was titrated from the brains and livers, and epitope-specific T cells were quantified from the spleens.

FIG. 4.

Protection from recombinant MHV expressing gp33 by immunization with rLm expressing gp33. Four-week-old B6 mice were inoculated with rLm-gp33 or rLm-np118 or not inoculated. Two weeks later they were infected with virus, all as outlined in Fig. 3 legend. (A) Titers of virus in the brains of animals either not inoculated with rLm (lane 1) or inoculated with rLm-np118 (lane 2) or rLm-gp33 (lane 3). The values shown are averages from three experiments, a total of 13 mice inoculated with rLm-np118, 12 mice inoculated with rLm-gp33, and 5 mice that were not inoculated. The horizontal line shows the limit of detection for the assay. The titers from mice inoculated with rLm-gp33 were significantly less than those from mice inoculated with rLm-np118 (P < 0.001, Mann-Whitney U test). (B) Percentages of gp33-specific CD8⁺ T cells in the spleens of infected mice. Splenocytes were isolated from mice; the frequency of CD8⁺ T cells specific for the gp33 epitope was determined by intracellular IFN-γ staining. The numbers in the upper-right corners indicate the percentage of CD8⁺ T cells that are positive for the intracellular IFN-γ stain. Mice were inoculated with rLm-gp33 alone (quadrant 1); RA59-gfp/gp33 alone (quadrant 2); rLm-np118 followed 2 weeks later by RA59gfp/gp33 (quadrant 3); or rLm-gp33 followed 2 weeks later by RA59gfp/gp33 (quadrant 4). The immune response of every animal was examined; sometimes samples were pooled from two or three animals. Data shown are from one representative assay; similar results in all cases were obtained in assays from two more independent experiments.

FIG. 5.

Viral antigen expression in the brains of mice infected with RA59-gfp/gp33 after immunization with rLm-gp33 or rLm-np118. Mice were inoculated with rLm-np118 (A) or rLm-gp33 (B) and 2 weeks later were infected with RA59-gfp/gp33 as described in Fig. 3. Mice were sacrificed at day 5, and sagittal brain sections prepared as described in Materials and Methods. Representative sections (two sections from each of four mice) are shown for each of three regions of the brain.

FIG. 6.

Perforin-deficient (PKO) mice are not protected from RA59-gfp/gp33 after immunization with rLm-gp33. Four-week-old PKO mice were inoculated with rLm-gp33 or rLm-np118 or were not inoculated. Two weeks later they were infected with RA59-gfp/gp33, as in Fig. 4. (A) Titers of virus in the brains of animals either not inoculated with rLm (lane 1) or inoculated with rLm-np118 (lane 2) or rLm-gp33 (lane 3). The values shown are averages from two experiments, a total of five mice inoculated with rLm-np118, nine mice inoculated with rLm-gp33, and three mice not inoculated with rLm. The horizontal line shows the limit of detection for the assay. (B) Percentages of gp33-specific CD8⁺ T cells in the spleens of infected mice. Splenocytes were isolated from mice; the frequency of CD8⁺ T cells specific for the gp33 epitope was determined by intracellular IFN-γ staining. The numbers in the upper-right corners indicate the percentage of CD8⁺ T cells that are positive for the intracellular IFN-γ stain. Mice were inoculated with rLm-np118 alone (quadrant 1), RA59-gfp/gp33 alone (quadrant 2), rLm-np118 followed 2 weeks later by RA59gfp/gp33 (quadrant 3), or rLm-gp33 followed 2 weeks later by RA59gfp/gp33 (quadrant 4). The immune response of every animal was examined; sometimes samples were pooled from two or three animals. Data shown are from one representative assay; similar results in all cases were obtained from two independent experiments.

FIG. 7.

RT-PCR sequence analysis of the EGFP/gp33 region of the genomes of viruses isolated from the CNS of protected and unprotected mice. (A) Schematic diagram of a portion of the genome of RA59-gfp/gp33 including the intergenic sequence (IGS) preceding gene 4, a short region of ORF4a, gp33, and the EGFP gene. Genomic RNAs from viruses recovered from nonfluorescent plaques from brains lysates of mice that were inoculated with rLm-gp33 (B) or rLm-np118 (C) 2 weeks before RA59-gfp/gp33 infection were sequenced through this region using primers FIJ81 and RIJ84 as discussed in Materials and Methods and the text. All (four of four) genomes from protected mice had deletions in gp33 as well as in EGFP (B), while the genome from an unprotected animal had gp33 intact while EGFP was deleted (C).