A virus-induced molecular mimicry model of multiple sclerosis (original) (raw)

Mice infected with PLP139-TMEV develop early-onset inflammatory demyelinating disease. A 30–amino acid sequence (PLP130-159) encompassing the native encephalitogenic PLP139-151 peptide was inserted into the leader portion of a cDNA construct of a nonpersisting, mutant BeAn strain of TMEV that had a ClaI restriction site added and 23 amino acids deleted (ΔClaI-BeAn) (Figure 1a). In addition, recombinant PLP139-151 epitope mimic viruses were constructed (Figure 1b) that had amino acid substitutions at either the primary (W144A) or secondary (H147A) T cell receptor (TCR) recognition sites (23, 24). As expected, SJL mice infected intracerebrally with the wild-type BeAn strain of TMEV (WT BeAn) developed clinical signs of demyelinating disease around day 30–35 PI and exhibited a chronic-progressive demyelinating disease (Figure 2, a and b). Clinical signs begin with an abnormal gait and worsen to paralysis and often death. Mice infected with the ΔClaI-BeAn virus, the nonpersisting parental construct for the recombinant viruses, did not develop any signs of autoimmune demyelinating disease. Interestingly, insertion of a 30-mer encompassing a non-self OVA323-339 epitope led to clinical disease, albeit with a delayed onset (approximately 50 days PI) compared with the wild-type BeAn TMEV strain, indicating that reintroduction of a nonspecific 30-mer into the leader of the ΔClaI-BeAn parental strain restores the ability of the virus to persist in vivo. Most significantly, mice infected with PLP139-BeAn virus developed severe clinical signs at very early times, i.e. 7–10 days PI. A similar early-onset disease course was observed in mice infected with the H147A PLP139-BeAn mimic virus. In contrast, mice infected with the W144A PLP139-BeAn, containing an amino acid substitution at the primary TCR contact residue, exhibited a delayed disease course similar to that of the virus containing the non-self OVA323-339 epitope. Virus load in the brains and spinal cords was similar up to 28 days after infection with any of these strains (data not shown), demonstrating that the recombinant viruses were able to persist in the CNS at wild-type levels during the early stages of infection. Spinal cords from mice sacrificed at 14 days PI with PLP139-BeAn virus had areas of infiltrating immune cells and signs of demyelination (Figure 2d), while mice infected with OVA323-BeAn virus had no signs of infiltration or demyelination (Figure 2c). Histology in the spinal cords of mice infected with H147A PLP139-BeAn was similar to that observed in PLP139-BeAn infected mice, containing infiltrating cells and displaying moderate demyelination, while mice infected with W144A PLP139-BeAn virus did not show histologic signs of disease at 14 days PI (data not shown). Thus, mice infected with viruses containing either the native encephalitogenic PLP139-151 epitope or the H147A altered peptide ligand (APL) developed early-onset demyelinating disease, while the W144A PLP139-BeAn virus APL was unable to initiate the early-onset demyelination.

SJL mice infected with TMEV containing native PLP139-151 sequence or sequenFigure 2

SJL mice infected with TMEV containing native PLP139-151 sequence or sequence mimics develop early demyelinating disease. Separate groups of SJL mice (n = 10) were injected intracerebrally with 3 × 107 PFU of the indicated recombinant viruses as well as wild-type virus (WT BeAn) or the parental construct (ΔClaI-BeAn). The mean clinical score (a) based on a grading scale of 0–5 and the percentage of mice showing clinical signs of demyelination (b) were determined at varying times after infection. Results are representative of at least three separate experiments. At 14 days PI, mice infected with OVA323-BeAn (c) or PLP139-BeAn (d) were anesthetized and sacrificed by total body perfusion through the left ventricle using chilled 3% glutaraldehyde in PBS, pH 7.3. Spinal cords were cut into 1-μm-thick segments and post-fixed in OsO4, dehydrated, and embedded in Epon. Toluidine blue–stained sections from ten segments per mouse were read and scored blind.

Mice developing demyelinating disease display PLP139-151–specific CD4+ T cell responses. The infected mice were next analyzed to determine the temporal progression of antigen-specific T cell responses generated during the infection. Mice infected with the wild-type strain of TMEV developed initial CD4+ DTH and T cell proliferative response to the immunodominant virus epitope, VP2 70-86, within 14 days after infection (Figure 3, a and b), which persisted through day 28 (Figure 3c) and day 90 (Figure 3d). As previously reported (22, 25), myelin-specific CD4+ T cell responses in mice infected with wild-type TMEV are first observed to the immunodominant PLP139-151 epitope beginning around 50 days PI and eventually spread to other myelin epitopes, including PLP56-70 and PLP178-191, by 120–150 days PI. This is confirmed by the presence of PLP139-151–specific splenic T cell proliferative responses at 90 days PI (Figure 3d), but not at day 14 (Figure 3b) or 28 (Figure 3c), in these mice. In contrast, mice infected with the PLP139-BeAn virus developed early CD4+ T cell responses to both VP2 70-86 and the virus-encoded PLP139-151 epitope by 14 days PI (Figure 3, a and b), which each persisted through day 90 (Figure 3, c and d). This indicates that the PLP139-151 peptide could be processed from the 30-mer insertion within the virus leader polyprotein and that the early-onset demyelinating disease is temporally related to the autoimmune response. Interestingly, mice infected with ΔClaI-BeAn virus had only virus-specific CD4+ T cell responses early after infection (Figure 3, a and b) and these responses waned by 90 days PI (Figure 3d), indicative of the failure of this variant to persist in the CNS (data not shown). Infection with OVA323-BeAn expressing the non-self antigen resulted in a CD4+ T cell response pattern similar to that observed in wild-type virus infection, where PLP139-151 responses were observed only at day 90 PI.

SJL mice infected with TMEV containing the native PLP139-151 sequence develFigure 3

SJL mice infected with TMEV containing the native PLP139-151 sequence develop early myelin-specific CD4+ T cell responses. SJL mice were infected with PLP139-BeAn, wild-type BeAn, OVA323-BeAn, or ΔClaI-BeAn. (a) At 14 days PI, the mice were ear-challenged with 5 μg PLP139-151 or TMEV VP2 70-86 and in vivo DTH responses were determined 24 hours after challenge. **DTH responses significantly higher than those in mock infected controls; P ≤ 0.01. Splenic CD4+ T cell proliferative responses from 2–3 mice per group were determined at 14 days (b), 28 days (c), and 90 days PI (d) to the immunodominant TMEV epitope, VP2 70-86, and to a panel of myelin epitopes — PLP139-151, PLP56-70, and PLP178-191. T cell proliferation was determined at 96 hours and results expressed as mean cpm ± SEM of triplicate cultures. *Stimulation index (cpm + antigen)/(cpm – antigen) ≥ 3. Results are representative of three separate experiments.

CD4+ T cell responses in mice infected with the H147A and W144A viruses differed significantly. Although infection with either virus induced early and sustained responses to VP2 70-86, only mice infected with the H147A PLP139-BeAn virus, which developed early-onset demyelination (Figure 2, a and b), developed CD4+ T cell responses cross-reactive with PLP139-151 at early times following infection (Figure 4, a and b). Interestingly, PLP178-191–specific T cell responses were observed at 90 days PI in mice infected with either PLP139-BeAn or the H147A APL virus (Figure 3d and Figure 4d). Therefore, infection with TMEV expressing a self myelin peptide with an amino acid substitution at the secondary, but not the primary, TCR contact residue can initiate a CD4+ T cell response cross-reactive with the native PLP139-151 epitope. In addition, as PLP178-191–specific responses were observed at 28 and 90 days PI, it appears that sustained tissue damage initiated by virus-induced molecular mimicry can subsequently lead to epitope spreading to additional PLP epitopes following the release of endogenous epitopes.

SJL mice infected with TMEV containing the H147A, but not the W144A, PLP139Figure 4

SJL mice infected with TMEV containing the H147A, but not the W144A, PLP139-151 APL develop early cross-reactive T cell responses to native PLP139-151. SJL mice were infected with PLP139-BeAn, H147A PLP139-BeAn, or W144A PLP139-BeAn. (a) At 14 days PI, the mice were ear-challenged with 5 μg PLP139-151 or VP2 70-86 and in vivo DTH responses were determined at 24 hours after challenge. **DTH responses were significantly higher than those in mock-infected controls; P ≤ 0.01. Splenic CD4+ T cell proliferative responses from 2–3 mice per group were determined at 14 days (b), 28 days (c), and 90 days PI (d) to the immunodominant TMEV epitope, VP2 70-86, and to a panel of self myelin epitopes — PLP139-151, PLP56-70, and PLP178-191. T cell proliferation was determined at 96 hours and results expressed as mean cpm ± SEM of triplicate cultures. *Stimulation index (cpm + antigen)/(cpm – antigen) ≥ 3. Results are representative of three separate experiments.

Infection with TMEV encoding the H147A APL induces PLP139-151–specific CD4+ Th1 responses. Autoimmune demyelination is associated with CNS infiltration of autoreactive proinflammatory CD4+ Th1 cells (26). We thus analyzed the cytokine profile of the antigen-specific T cell responses at early (28 days) and late (90 days) time points in the infected mice. IL-2 secretion correlated with the virus- and myelin-specific CD4+ T cell proliferation profiles for the mice infected with the different virus constructs (Figure 5a). TMEV VP2 70-86–specific CD4+ T cells in all the infected mice secreted Th1 (IFN-γ and TNF-α), but not Th2 (IL-4), cytokines at early and late times (Figure 5, b–d). PLP139-151–specific CD4+ T cells from mice infected with PLP139-BeAn and H147A PLP139-BeAn secreted IFN-γ and TNF-α at 28 days PI, which was not seen in mice infected with W144A PLP139-BeAn or OVA323-BeAn viruses. However, by 90 days PI, PLP139-151–specific CD4+ T cells from all of the infected groups secreted IFN-γ and TNF-α. Thus, the PLP139-151–specific CD4+ T cells associated with the early-onset demyelinating disease in the mice infected with TMEV encoding either native PLP139-151 or the H147A APL were Th1 cells. IL-4 secretion was not demonstrated in any of the infected groups (Figure 5d).

SJL mice infected with PLP139-BeAn and H147A PLP139-BeAn develop CD4+ Th1 rFigure 5

SJL mice infected with PLP139-BeAn and H147A PLP139-BeAn develop CD4+ Th1 responses to both virus and native PLP139-151 epitopes. Culture supernatants from T cell proliferation assays shown in Figures 2 and 3 were collected at 48 hours and measured for cytokine secretion by ELISA: (a) IL-2, (b) IFN-γ, (c) TNF-α, and (d) IL-4. *Responses significantly above those of PBS-stimulated wells; P ≤ 0.01.

Mice infected with TMEV expressing a bacterial mimic of PLP139-151 develop clinical disease associated with the cross-reactive induction of CD4+ Th1 responses. A previous report by Carrizosa et al. (27) identified mimic peptides in mouse pathogens that cross-reacted with PLP139-151–specific T cell clones derived from SJL mice. We inserted one of these mimic epitope sequences, H. influenzae protease IV 574–586 (HI574-586), which shares only 6 of 13 amino acids with the core PLP139-151 epitope (Figure 1b), into the BeAn viral genome. Significantly, mice infected with HI-BeAn developed early clinical disease following infection (Figure 6a). Clinical disease was somewhat less severe than in mice infected with PLP139-BeAn, and disease onset was also slightly delayed. However, both PLP139-BeAn– and HI-BeAn–infected mice developed clinical signs at a significantly earlier time PI than did WT BeAn–infected mice. HI-BeAn disease was associated with the induction of cross-reactive PLP139-151–specific CD4+ T cells as assessed by T cell proliferation assays (data not shown), and, significantly, CD4+ T cells at 21 days PI from both the PLP139-BeAn– and HI-BeAn–infected mice secreted IFN-γ at equivalent levels upon challenge with the native PLP139-151 epitope (Figure 6b). Additionally, both PLP139-BeAn– and HI-BeAn–infected mice secreted IFN-γ when rechallenged with the immunodominant TMEV VP2 70-86 epitope.

SJL mice infected with TMEV expressing a PLP139-151 molecular mimic from thFigure 6

SJL mice infected with TMEV expressing a PLP139-151 molecular mimic from the protease IV protein of H. influenzae develop clinical disease associated with activation of PLP139-151–specific Th1 cells. Separate groups of mice (n = 8) were injected intracerebrally with 3 × 107 PFU of recombinant virus (PLP139-BeAn or HI-BeAn) or with wild-type BeAn (WT BeAn) and observed at varying times after infection for development of clinical signs (a). Clinical results are representative of two separate experiments. Spleen cells were harvested at 21 days PI and rechallenged with viral peptide (VP270-86), myelin peptides (PLP139-151, PLP178-191), or mimic peptide (HI574-586) and supernatants collected after 48 hours and measured for IFN-γ secretion by ELISA (b). *Responses statistically significant compared with PBS-stimulated wells; P ≤ 0.01.