Intrinsic and Induced Regulation of the Age-Associated Onset of Spontaneous Experimental Autoimmune Encephalomyelitis¹

Multiple sclerosis (MS)³ is considered to be an immune-mediated disease of the CNS characterized by perivascular CD4⁺ T cell and mononuclear cell infiltration (1, 2, 3, 4). The critical role of CD4⁺ T cells in the autoimmune pathogenesis of MS is based on multiple lines of evidence. Activated myelin-reactive CD4⁺ T cells are found in CNS lesions (5). Several CD4⁺ T cell-targeted therapies inhibit the progression of MS (5). Genomic studies have confirmed the associations of the HLA class II with MS susceptibility (6, 7). Lastly, experimental autoimmune encephalomyelitis (EAE), an animal model of MS, is induced by priming with MHC class II-restricted myelin peptides in CFA and can be transferred solely by CD4⁺ T cells, providing direct definitive evidence that CD4⁺ T cells can initiate immune-mediated CNS demyelinating disease (8, 9, 10, 11). Collectively, these observations support an important pathogenic role for CD4⁺ T cells in mediating myelin damage in MS.

Transgenic (Tg) mice expressing TCRs specific for myelin basic protein (MBP) or proteolipid protein (PLP) have been used to study spontaneous EAE (sEAE) without the requirement for immunization with myelin peptide and adjuvant. However, sEAE does not develop in MBP_Ac1–11-specific TCR Tg mice housed in specific pathogen-free (SPF) conditions (12). In the presence of nontolerant MBP-reactive T cells, regulatory T cells (Tregs) expressing endogenous α-chain contribute to the prevention of sEAE in MBP TCR Tg mice (13). These data indicate that the mere presence of self-specific CD4⁺ T cells does not ensure development of autoimmune disease, and suggest, but do not prove, that the balance of self-reactive pathogenic T cells and Tregs determine whether the mice will develop spontaneous autoimmune disease.

A protective role for Tregs has been reported in multiple autoimmune diseases. CD4⁺CD25⁺ Tregs (natural Tregs; nTregs) are the best-documented Treg population. Depletion of this cell population in neonatal mice results in the spontaneous induction of spontaneous autoimmune diseases (14, 15, 16, 17, 18, 19, 20). For example, multiorgan autoimmune diseases (such as thyroiditis, gastritis, insulitis, et al.) are induced in BALB/c athymic nude mice receiving syngeneic CD4⁺ T cells depleted of CD25⁺ T cells. Reconstitution of CD4⁺CD25⁺ T cells within a limited period of time after the transfer of CD25⁻ T cells inhibited disease development (18). Similarly, CD4⁺CD25⁺ T cells from NOD mice delayed/prevented development of spontaneous diabetes in CD28-deficient NOD mice (21). It is noteworthy, however, that regulation of self-reactive T cells likely involves multiple populations of immunoregulatory T cells, including nTregs, IL-10-producing Tregs (T_R1), and TGF-β-producing Tregs (T_H3) (22, 23, 24, 25). Using an adoptive transfer system using MBP TCR Tg cells, Cabbage and colleagues (26) show that endogenous Tregs can induce the differentiation of naive MBP Tg T cells into a unique tolerized state in the presence of nonactivated APCs. These MBP-specific regulatory Tg T cells produced IL-10 and TGF-β, which inhibited disease induced by activated APCs. Furthermore, it was shown that both the initial differentiation and subsequent tolerant state required the presence of endogenous Tregs. These data demonstrate that nTregs along with induced Tregs (iTregs), including T_R1 and T_H3, play an important role in maintaining self-tolerance.

The current study focused on PLP_139–151-specific 5B6 TCR Tg (5B6 Tg) mice on the highly EAE-susceptible SJL background. Unlike MBP peptide-specific TCR Tg mice that developed sEAE only when raised in a conventional animal facility or when crossed to RAG 1-deficient (RAG^−/−) background to remove Tregs, 5B6 Tg mice on a RAG^+/+ background develop sEAE at a high incidence under barrier SPF conditions (27). The 5B6 Tg SJL mice thus provide a valuable model to determine endogenous mechanisms regulating spontaneous development of autoimmune disease in susceptible mice with an intact immune system. We show that initial development of clinical disease correlates with the appearance of hyperactivated CD4⁺ T cells in the cervical lymph nodes (cLN), but not spleens, of clinically affected 5B6 Tg mice compared with age-matched healthy controls and with infiltration of CD4⁺ T cells into the lumbar spinal cord. Interestingly, nTregs from 80-day-old 5B6 Tg mice exhibit decreased suppressive capacity compared with 40-day-old mice, and depletion/inactivation of Tregs using anti-CD25 mAb treatment between 30 and 40 days of age led to a significant increase in disease incidence and severity. Furthermore, PLP_139–151-activated splenocytes from 48-day-old and 116-day-old 5B6 Tg mice produce significantly less IL-10 than 27-day-old Tg, suggesting an age-associated decrease in iTreg/T_R1 activity. Lastly, induction of peripheral immune tolerance using peptide-pulsed, ethylcarbodiimide (ECDI)-fixed APCs (Ag-splenocytes; Ag-SP) in 30–40-day-old 5B6 mice prevented or delayed onset of disease concomitant with increased production of IL-10 and induction of Foxp3⁺ Tregs from CD4⁺CD25⁻ 5B6 progenitors. Collectively, these findings suggest that multiple regulatory mechanisms contribute to the regulation of the age-associated sEAE in 5B6 Tg mice.

PLP_139–151-specific 5B6 TCR Tg mice were obtained from Dr. Vijay Kuchroo (Harvard Medical School, Boston, MA) and crossed to SJL CD90.1 congenic mice. All mice used in this study were female 5B6 Tgs. The mice were housed under specific pathogen-free conditions in the Northwestern University Center for Comparative Medicine Barrier Facility. Paralyzed animals were provided easier access to food and water. All protocols were approved by Northwestern University Animal Care and Use Committee.

PLP_139–151 (HSLGKWLGHPDKF) and OVA_323–339 (ISQAVHAAHAEINEAGR) were purchased from Genemed Synthesis.

Mice were observed for clinical symptoms of sEAE three times a week. Mice were scored on a scale of 0–5 as follows: 0 = no abnormality; 1 = limp tail or hind limb weakness; 2 = limp tail and hind limb weakness; 3 = partial hind limb paralysis; 4 = complete hind limb paralysis; 5 = moribund. The data are plotted as the mean daily clinical score for all animals in a particular experimental group.

CNS immunohistological evaluation was performed as previously described (28). In brief, brain and spinal cord were removed by dissection from anesthetized, perfused mice. Blocks of brain halves and 2 to 3 mm sections of spinal cord were frozen in OCT (Miles Laboratories) in liquid nitrogen and stored at −80°C. The 6 μM thick cross-sections from spinal cord and 10 μM cross or sagittal sections from brain were stained using tyramide signal amplification direct kit (NEN) according to manufacturer’s instructions. Slides were examined using a Leica DM500B fluorescent microscope and images captured using the SPOT RT camera (Diagnostic Instruments) and SPOT imaging software. Three serial sections from each fragment per examined mouse were analyzed at ×2 and ×40 magnification.

DTH responses were measured using a 24-h ear-swelling system as previously described (28). The increase in ear thickness was determined 24 h after ear challenge by injecting 10 μg of respective peptide (in 10 μl of saline) into the dorsal surface of the ear. Results are expressed in units of 10⁻⁴ inches ± SEM.

Single-cell suspensions were blocked for 10 to 15 min with anti-CD16/32 before staining with a fluorescently tagged Ab-mixture directed against surface markers CD4 (RM4–5) and CD25 (PC61) (BD Pharmingen or eBioscience). Intracellular Foxp3 was stained using eBioscience Foxp3 staining buffer set and Abs according to the manufacturer’s instruction. Data were acquired on an LSR II cytometer (BD Biosciences) and analyzed with FlowJo software (Tree Star).

Mice were perfused and CNS mononuclear cells were isolated as described (50). Cells were activated for 5 h with 5 ng/ml phorbol 12-myristate 13-acetate and 500 ng/ml ionomycin in the presence of GolgiStop (BD Biosciences), followed by staining with LIVE/DEAD fixable dye (Molecular Probes; Invitrogen) and the fluorescently tagged Ab against CD4 (RM4–5). Intracellular IL-17 and IFN-γ were stained according to the eBioscience Intracellular Cytokine Staining Protocol.

Sorted CD4⁺ responders were labeled with CFSE. A fixed number of responder T cells was cultured with titrated numbers of CD4⁺CD25⁺ T cells in the presence of PLP_139–151 and irradiated splenocytes as APC. Cells were cultured for 72 h at 37°C in HL-1 medium (BioWhittaker) supplemented with 50 μM 2-ME, 2 mM l-glutamine, 100 U/ml penicillin, and 100 μg/ml streptomycin. Proliferation was determined by the dilution of CFSE. Discrimination of responders and Tregs was based on staining with CD90.1 and CD90.2.

Spleen cell suspensions were cultured as described above. Triplicate wells were stimulated with the indicated doses of Ag and incubated for 72 h. Supernatants were collected and cytokine production was tested using the Beadlyte mouse multicytokine detection system (Millipore) and analyzed using Luminex 100 IS software.

Single cell suspensions of RBC-free splenocytes were coupled with PLP_139–151 or OVA_323–339 peptides using 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide HCl (ECDI; Calbiochem-Behring) as previously described (29). A total of 5 × 10⁷ Ag-coupled splenocytes were injected i.v. into the lateral tail veins of recipient 5B6 Tg mice at 30–40 days of age.

Comparisons of cytokine production and DTH responses between the various groups were analyzed by unpaired Student’s t test. Disease incidence was compared via χ² analysis using Fisher’s exact test.

A total of 73 female 5B6 Tg mice housed in SPF conditions were monitored for the development of clinical signs of EAE for 160 days (Fig. 1,A). Over 80% of these mice developed sEAE. No signs of EAE were observed in any animals under 42 days of age, and very few mice over the age of 100 days developed clinical disease. We refer to mice under 42 days of age as young mice and mice older than 42 days of age as old mice. The disease first presented with the loss of tail tone and extended to weakness of hind limbs. In a few cases, hind limb paralysis developed, but remarkably the disease symptoms were generally mild and none of the mice reached a moribund state (score 5). Unlike PLP_139–151/CFA-induced EAE in wild-type SJL mice, no sustained remissions or overt relapses were observed (Fig. 1 B). In a few cases some improvement was observed; however, it never went beyond improvement of more than one clinical grade.

To begin to identify potential mechanisms of pathogenesis of sEAE in 5B6 Tg mice, histopathological analysis was performed on cerebellum, brain stem, and spinal cord tissues from five to six representative Tg mice with or without clinical disease symptoms, respectively. Spinal cords were divided into upper cervical, lower cervical, upper thoracic, middle thoracic, lower thoracic, upper lumbar, and lower lumbar sections. Three serial sections were analyzed from each CNS region. CD4⁺ T cells (red) were observed throughout the whole CNS from cerebellum to lumbar spinal cord in mice with clinical symptoms (Fig. 2,A). The T cell infiltration mainly affected white matter. Interestingly, CD4⁺ T cells were also found in some of the Tg mice without clinical signs (Fig. 2,A) as early as 24 days of age (data not shown). The infiltration in healthy mice mainly affected the upper spinal cord, cerebellum, and brain stem, but very few or no CD4⁺ cells were found in lumbar spinal cord, whereas the lumbar spinal cord was heavily infiltrated in mice displaying clinical disease (Fig. 2, A and B). Overall, the numbers of CNS-infiltrating CD4⁺ T cells were significantly greater in mice with clinical symptoms than healthy mice (Fig. 2,C). To determine the cytokine profiles of CD4⁺ T cells in the CNS during disease pathogenesis, we isolated CNS cells from the pooled brain and spinal cord of clinically affected mice for analysis of IL-17 and IFN-γ production. Upon in vitro PMA/ionomycin stimulation, CNS-infiltrating CD4⁺ T cells isolated from diseased animals exhibited a relatively equal distribution of Th17 and Th1 cells (Fig. 2 D) with a Th17/Th1 ratio of 0.95 in this particular experiment. Our preliminary data show similar 1:1 Th17/Th1 ratio in infiltrating CD4⁺ T cells isolated separately from the brain and spinal cord of clinically affected mice. Taken together, these data indicate that clinical symptoms of sEAE correlate with the CD4⁺ T cell infiltration in lumbar spinal cord, and it is likely that both Th17 and Th1 cells are involved in the pathogenesis.

Peptide-specific in vivo DTH and in vitro recall responses were used to determine whether the onset of clinical symptoms of sEAE in clinically affected mice correlated with increased functional activation of PLP_139–151-specific T cells and the anatomic site of the primary activation of the Tg T cells. Compared with age-matched healthy Tg mice, animals with clinical disease displayed enhanced DTH responses to ear challenge with PLP_139–151 (Fig. 3,A), CD4⁺ T cell CNS infiltration (Fig. 2,C), and enhanced production of both IFN-γ and IL-17 in both the spleen and cLNs in response to in vitro stimulation with PLP_139–151 (Fig. 3,B) indicating enhanced development of both effector Th1 and Th17 cells. To identify where the Tg T cells were initially activated, we measured the recall proliferation by CFSE dilution of CD4⁺ T cells from spleens and cLNs of nonsymptomatic controls in comparison to mice with recent onset disease (Fig. 3,C). As anticipated, robust proliferation of CD4⁺ T cells was observed in both groups of mice with no detectable difference of proliferation of spleen CD4⁺ T cells. However, CD4⁺ T cells from the cLNs of mice with recent onset sEAE proliferated more robustly than mice without disease symptoms, although no significant difference of the cLN cell numbers was observed (Fig. 3 D). In summary, these data suggest that mice with recent onset clinical sEAE have more Ag-experienced CD4⁺ T cells than mice without disease symptoms, and these activated CD4⁺ T cells appear to be enriched in cLNs, a site previously associated with CNS Ag drainage (30, 31, 32, 33).

Lafaille et al. (13) reported that MBP TCR Tg mice on RAG^+/+ background did not develop spontaneous EAE unless Tregs were removed by crossing these mice onto a RAG^−/− background. In contrast, 5B6 TCR Tg mice on a RAG^+/+ background develop sEAE in an age-dependent manner (Ref. 27 ; Fig. 1), but when crossed to the RAG^−/− background they develop such severe disease that the line cannot be maintained (27). We thus wished to determine whether development of spontaneous clinical disease in 5B6 TCR^+/+ was associated with an age-related deficiency in the numbers and/or suppressive function of Treg cells. Although we found that 5B6 Tg mice generally have a lower percentage of Foxp3⁺CD4⁺ Tregs as compared with wild-type SJL controls, the numbers of Tregs in the spleens and cLNs of 5B6 Tg mice did not significantly vary with age, nor were the numbers different when comparing clinically affected mice vs nonaffected controls of a similar age (Fig. 4,A). Although sorted CD4⁺CD25⁺ T cells from 5B6 Tg mice at various ages were found to suppress the proliferation of effector CD4⁺CD25⁻ 5B6 T cells at high Treg-Teff ratios (where Teff is effector T cell) (data not shown), the suppressive function of sort purified splenic CD4⁺CD25⁺ T cells was decreased in older mice as illustrated by the significantly reduced ability of Tregs from 80-day-old as compared with 40-day-old mice to suppress CFSE dilution at a 1:1 Treg-Teff ratio (Fig. 4,B). Purification of the nTregs was equivalent in the two age groups (Fig. 4 C).

To further define the role of nTregs in the regulation of sEAE, we determined whether in vivo depletion of CD25⁺ T cells in young 5B6 Tg mice would affect the incidence or severity of sEAE. A total of 30–40-day-old mice were treated with monoclonal anti-CD25 Ab (7D4) for 7 treatments at two days interval and monitored for development of clinical EAE for an additional 30 days. Fig. 4 D shows mice treated with anti-CD25 Ab have higher incidence (6/10 = 60%) than the mice treated with control IgM (1/9 = 11%, p < 0.05). Taken together, these findings indicate that nTregs from 5B6 Tg mice play a functional role in suppressing disease in young mice, but that their suppressive capacity declines with age.

We were also interested in determining whether the age-associated development of sEAE in the 5B6 Tg mice was associated with a change in the pattern of peptide-induced cytokine production. Spleen cells isolated from the 5B6 Tg mice at 27, 48, and 116 days of age were analyzed for the production of IL-10, IL-4, IL-5, IL-6, TNF-α, IFN-γ, (Fig. 5), and TGF-β (data not shown) in response to in vitro stimulation with PLP_139–151. Lymphocytes from individual mice were analyzed to avoid pooling cells from mice having preclinical sEAE with cells from healthy mice. No conclusion can be drawn regarding TGF-β production due to the inconsistency from experiment to experiment (data not shown). No significant differences in TNF-α and IFN-γ production between the different groups (Fig. 5, E and F) was observed, whereas IL-10, IL-4, IL-5, and IL-6 production was generally higher in young 27-day-old mice vs mice at 48 and 116 days of age (Fig. 5, A–D). Interestingly, IL-10 production was consistently significantly higher in the younger mice. Thus regulatory cytokine production, especially IL-10, is decreased as mice age and exhibits an enhanced incidence of clinical disease. In a separate experiment, splenocytes from 5B6 Tg mice at 29, 55, and 97 days of age were similarly analyzed for the production of IL-17, and 97-day-old mice were found to produce significantly more peptide-induced IL-17 than nonclinically affected younger mice (Fig. 5 G). This indicates that a pathologic population of PLP_139–151-specific T cells arises as regulation is lost.

The above findings indicate that the functional decline of nTreg activity as well as peptide-induced IL-10 production in 5B6 Tg mice correlates with the age-related development of clinical sEAE. We next tested whether development of spontaneous disease could be prevented or delayed by the induction of peptide-specific tolerance induced by the i.v. injection of PLP_139–151-pulsed, ECDI-fixed syngeneic splenocytes (Ag-SP), which we have shown can successfully prevent and treat both actively induced and adoptive EAE in conventional mice (34, 35, 36, 37, 38, 39, 40, 41, 42, 43). The 5B6 mice were tolerized with PLP_139–151-SP or OVA_323–339-SP between 30 and 40 days of age and observed for development of clinical disease for an additional 30–40 days. Tolerization significantly reduced peptide-specific DTH responses (Fig. 6,A) and delayed onset of clinical sEAE (Fig. 6,B). Furthermore, splenocytes from PLP_139–151-SP tolerized 5B6 mice produced significantly more IL-10 than OVA_323–339-SP tolerized mice upon peptide recall in vitro (Fig. 6 C). Again, the TGF-β results were inconsistent from experiment to experiment and no conclusions could be drawn.

To determine whether the protective effects of PLP_139–151-SP treatment on clinical disease development were associated with activation and/or generation of peptide-specific Tregs, the effects of i.v. administration of PLP_139–151-SP vs OVA_323–339-SP on both bulk CD90.1⁺CD4⁺ and FACS-sorted CD90.1⁺CD25⁻ 5B6 TCR Tg T cells were determined following adoptive transfer into wild-type CD90.2 SJL mice. Interestingly, peptide-specific tolerance resulted in an increase of 7-fold and 2-fold, respectively, in the percentage of CD90.1⁺CD25^lowFoxp3⁺ and CD90.1⁺CD25^highFoxp3⁺ CD4⁺ T cells in wild-type SJL recipients of bulk donor 5B6 CD90.1⁺CD4⁺Tg T cells 3 days after treatment (Fig. 7,A). PLP_139–151-SP tolerization also induced expression of Foxp3 in recipients of purified naive donor CD4⁺CD25⁻ Tg T cells (Fig. 7, B and E) and CD25 expression on these donor CD90.1⁺Foxp3⁺ T cells was lower than on recipient CD90.2⁺ Foxp3⁺ T cells (Fig. 7, C and D), indicating conversion/activation of an adaptive Foxp3⁺ iTreg by Ag-SP tolerance. Collectively, the data indicate that Ag-SP tolerance inhibits induction of sEAE by inducing various Treg subsets including adaptive and IL-10-producing Foxp3⁺ Tregs.

It has been previously reported that 5B6 TCR Tg mice specific for PLP_139–151 on a conventional RAG^+/+ background developed a high incidence of sEAE when raised in SPF conditions (27). In the current study, we extended the previous study by describing the immunological changes correlating with disease onset and determining that spontaneous disease development was associated with the age-related decline in several intrinsic regulatory mechanisms involved in maintenance of self-tolerance.

Two major observations pertaining to the maintenance of tolerance are made in this study. First, CD4⁺CD25⁺ Treg cells from 80-day-old 5B6 Tg mice were found to be functionally less suppressive than Treg cells from isolated 40-day-old mice (Fig. 4,B) correlating with the age-dependent increase in onset of clinical sEAE. In confirmation of a protective role of nTregs in young 5B6 Tg mice, we found that their inactivation by treatment of young animals with anti-CD25 led to a significant increase in disease incidence and severity (Fig. 4,C), whereas their induction following tolerization with PLP_139–151-SP led to a significant delay in disease onset (Fig. 6,A). This is consistent with earlier reports showing an age-related decline in suppressive function of CD4⁺CD25⁺ Tregs in 16 wk-old as compared with 8 wk-old NOD mice corresponding with development of spontaneous type 1 diabetes (24, 44). The age-associated functional deficiency in suppressive function was demonstrated despite the observation that there was no observable difference in the frequency of CD25⁺Foxp3⁺ cells between young vs old 5B6 mice or between clinically affected vs unaffected mice of the same age (Fig. 4,A). Second, splenic T cells from 48- and 116-day-old 5B6 mice produced significantly less IL-10 upon peptide recall (Fig. 5,A) than 27-day-old Tg mice, a time point before onset of sEAE. IL-10 has been reported to play an important role in maintaining the nonencephalitogenic phenotype of the autoreactive T cells (45). As the Tr1 regulatory cell population is a major source of IL-10 (25), this observation is consistent with an important potential role for Tr1 cells in maintaining self-tolerance in young 5B6 mice. In support for a protective role of IL-10 in preventing disease onset, we found that tolerization of young 5B6 mice with PLP_139–151-SP that led to a significant delay in disease onset (Fig. 6,B) was associated with enhanced production of IL-10 (Fig. 6 C). In addition, we previously reported that tolerization of mice with pre-existing PLP_139–151-induced relapsing-EAE with PLP_139–151-SP at peak of acute disease significantly increased the levels of production of the anti-inflammatory cytokines TGF-β and/or IL-10 in CD4⁺ T cells recovered from both the periphery and the CNS upon peptide restimulation in vitro (46). Together, these data support the hypothesis that the regulation of the age-related induction of sEAE in 5B6 Tg SJL mice is highly complex and likely involves a heterogeneous population of immunoregulatory T cells whose activity can be enhanced by tolerogenic administration of cognate peptide in the form of PLP_139–151-pulsed, ECDI-fixed syngeneic splenocytes (PLP_139–151-SP).

Tolerization with PLP_139–151-SP resulted in a 7-fold and 2-fold expansion vs OVA_323–339-SP injected controls, respectively, in the percentage of Foxp3⁺CD25^low and Foxp3⁺CD25^high T cells in recipients of bulk 5B6 CD4⁺ T cells (Fig. 7,A) within 62 h following treatment. This rapid increase in Treg percentage indicates the generation of adaptive Tregs. In addition, PLP_139–151-SP vs OVA_323–339-SP tolerization of wild-type recipients of purified CD90.1⁺CD4⁺CD25⁻ 5B6 TCR Tg cells led to the generation of a significant population of Foxp3⁺ T cells, the majority of which expressed a CD25^low phenotype distinguishing them from CD25^high nTregs (Fig. 7, B–E). This CD4⁺Foxp3⁺CD25^low phenotype is a characteristic of adaptive Tregs that were recently reported to be induced from peripheral CD4⁺Foxp3⁺ precursors in CD28-deficient NOD mice by nonmitogenic anti-CD3 immunotherapy (47). Thus, tolerization with PLP_139–151-SP apparently leads to the induction of multiple regulatory mechanisms including some expansion of CD4⁺CD25^highFoxp3⁺ nTregs and the conversion of naive CD4⁺CD25⁻ 5B6 T cells to CD4⁺CD25^lowFoxp3⁺ adaptive Tregs, as well as inducing enhanced IL-10 production (Fig. 6) with the net outcome of significantly inhibiting/delaying onset of sEAE in 5B6 TCR Tg mice.

Another key observation made in this study is that cLN T cells, but not spleen T cells, from 5B6 Tg SJL mice with recent-onset sEAE are hyperresponsive to peptide stimulation compared with cLN T cells from mice without disease symptoms (Fig. 3 C). The initial presence of hyperresponsive CD4⁺ T cells in cLNs supports the hypothesis that this lymphoid organ is the primary site for the peripheral activation of T cells specific to an immunodominant myelin proteinduring the development of sEAE. This finding is consistent with prior studies indicating that the cLNs are lymphatic sites draining some areas of the CNS (30, 31, 32, 33), and may suggest that initial activation of myelin-specific Tg T cells occurs as the result of leakage of myelin proteins from the CNS. Similar results have been found in the MBP_Ac1–11 TCR Tg mouse system wherein activated T cells are first noted in the cLNs preceding onset of clinical disease in 7–8-wk-old mice and surgical removal of the cLNs can delay the age of disease onset (54). The failure to identify peptide-specific hyperresponsive T cells in the spleen speaks against the possibility that the activated cells in the cLNs were initially activated in the CNS and simply “leaked” into the peripheral circulation through the compromised blood-brain barrier. In situ tolerance has been reported to occur in T cells entering the CNS during noninflammatory conditions. Indeed, CNS T cells from MBP TCR Tg mice without disease symptoms fail to respond to Ag, whereas T cells in the cLN and other non-CNS draining LNs proliferated robustly (48). In addition, CNS cells from asymptomatic MBP TCR Tg mice are able to suppress the proliferation of peripheral MBP-specific T cells in vitro. Other experiments have demonstrated that the interaction between neurons and T cells could convert encephalitogenic T cells to regulatory T cells, which presumably suppress activation of encephalitogenic T cells under noninflammatory conditions (49). These data suggest that the noninflamed CNS may normally be a “tolerogenic” milieu in the steady state. In contrast, in the inflamed CNS during progressive relapsing EAE in wild-type SJL mice, the CNS serves as the primary site where naive T cells specific for endogenously released spread epitopes become activated (50, 51).

The specific sites of the CNS infiltration of encephalitogenic T cells are major determinants of disease symptoms and severity. Our data clearly shows expression of clinical symptoms of spontaneous EAE correlates with the infiltration of CD4⁺ T cells into the lumbar spinal cord region. Interestingly, significant infiltration of CD4⁺ T cells can be detected in brain and upper spinal cord in both mice with and without disease symptoms. However, only a very few mice without disease symptoms have detectable infiltration in lumbar spinal cord, which is observed in all the mice with symptomatic sEAE (Fig. 2). We were unable to compare the activity of CNS-infiltrating CD4⁺ T cells from mice with sEAE to mice without disease symptoms due to the difficulty of isolating sufficient numbers of CD4⁺ T cells from CNS of nonsymptomatic mice for functional analysis. However, T cells isolated from the CNS of 5B6 mice with recent-onset disease express a highly activated phenotype as indicated by the up-regulation of CD25, CD69, CD44, and CD11a and the down-regulation of CD45RB (data not shown). It is likely that both Th1 and Th17 cells may contribute to the disease pathogenesis based on the observation that both Th17 and Th1 were found at roughly a 1:1 ratio in CNS-infiltrating CD4⁺ T cells in diseased mice (Fig. 2,D), and the observation that spleen and cLN (Fig. 3,B) T cells from mice with disease symptoms produced significantly greater amounts of both IFN-γ and IL-17 than mice without symptoms. Muller et al. (52) reported that the distribution of inflammatory cells correlated with disease symptoms using an atypical model of EAE induced in C3H/HeJ mice. These mice developed two distinct types of EAE, one characterized by ascending paralysis and the other by axial rotatory symptoms. Mice that exhibited only ascending paralysis showed inflammation preferentially in the spinal cord, whereas inflammation presented only in cerebellum and brainstem in mice that exhibited axial rotatory EAE. More recently, it was reported that T cells specific for different epitopes on myelin oligodendrocyte glycoprotein (MOG) generated different ratios of Th17/Th1 cells in two C3H congenic strains (53). T cell infiltration of the brain leading to atypical EAE characterized by proprioception defects, ataxia, spasticity, and hyperreflexivity was evident when Th17 cells were in excess of Th1 cells, whereas more classic EAE, characterized by ascending flaccid hindlimb paralysis and parenchymal inflammatory infiltration in the spinal cord and brain, was seen in mice exhibiting a ∼1:1 Th17/Th1 ratio. These data suggest that T cell specificity may influence sites of inflammation by inducing different T cell effector cells. Our data indicate that infiltration of Th1 and Th17 cells in a 1:1 ratio to the lumbar spinal cord primarily contributes to the pathogenesis of a classic form of sEAE in 5B6 Tg mice, whereas infiltration in cerebellum and brain stem alone is not associated with disease symptoms (Fig. 2). Preferential finding of inflammatory infiltrates in the lumbar spinal cord may be attributed to the specificity and the naturally determined effector function of the Tg T cells (Th17 vs Th1) in our model.

In summary, we propose that the decline of multiple regulatory mechanisms with age contributes to the development of sEAE in 5B6 Tg mice. In the young mice, a balance between pathogenic T cells and a regulatory network including nTregs and induced Tr1 cells protects the animals from onset of clinical CNS disease. However, as suppressive capacity of nTregs and IL-10-producing Tr1 cells declines, the threshold of activation is decreased and proinflammatory PLP_139–151-specific CD4⁺ Th1 and Th17 cells are activated in the cLNs and initiate clinical disease. Our results also indicate that peripheral induction of peptide-specific tolerance can reactivate these two T cell-mediated regulatory networks leading to prolonged protection from development of the spontaneous onset of overt clinical disease.

We thank Dr. Vijay K. Kuchroo for the gift of the 5B6 TCR Tg mice, Samantha Bailey for helpful suggestions and discussions throughout the course of this work, Matt DeGutes and Gwen Goings for technical assistance with immunohistochemistry, and Terra Frederick for critical reading of the manuscript.

The authors have no financial conflict of interest.