IL-4, IL-10, IL-13, and TGF-β from an Altered Peptide Ligand-Specific Th2 Cell Clone Down-Regulate Adoptive Transfer of Experimental Autoimmune Encephalomyelitis

Experimental autoimmune encephalomyelitis (EAE) is induced in the SJL/J mouse by adoptive transfer of activated proteolipid protein peptide (PLP) 139-151-specific Th1 cells. T cells responding to altered peptide ligands (APL) of PLP, previously shown to induce Th2 differentiation and regulate disease in PLP-immunized mice, do not transfer EAE. However, the exact mechanism of disease regulation by APL-specific T cells has not been elucidated. In this report, we show that 1F1, a Th2 clone specific for an APL of PLP139-151 can prevent adoptive transfer of EAE when cocultured with PLP-encephalitogenic spleen cells (PLP-spleen). Cytokines from activated 1F1 cells were detected by hybridization of mRNA to oligonucleotide arrays (DNA chip) and by ELISA. The Th2 cytokines found to be present at the highest protein and mRNA levels were evaluated for their role in suppression of adoptive transfer of EAE from PLP-activated spleen cell cultures. Abs to individual cytokines in 1F1 PLP-spleen cocultures suggested that IL-4, IL-13, and TGF-β played a significant role in suppressing EAE. Abs to the combination of IL-4, IL-10, IL-13, and TGF-β completely neutralized the protective effect of 1F1. Addition of Th2 cytokines to PLP-spleen cultures showed that IL-13 and TGF-β were each individually effective and low levels of IL-4 synergized with IL-13 to inhibit disease transfer. IL-5, IL-9, and IL-10 had little or no effect whereas GM-CSF slightly enhanced EAE. Our results demonstrate that Th2 cytokines derived from APL-specific Th2 cells can effectively down-regulate the encephalitogenic potential of PLP-spleen cells if present during their reactivation in culture.

xperimental autoimmune encephalomyelitis (EAE), 2 an animal model of multiple sclerosis, results from immunization with CNS myelin proteins or peptides under Th1 conditions or by adoptive transfer of CNS Ag-specific CD4 ϩ Th1 cells (1)(2)(3). Recent experiments in transgenic models of EAE have shown that the context of encephalitogenic T cells can influence disease outcome. Transgenic mice that express a TCR specific for the myelin Ag MBP only rarely develop spontaneous EAE. However, when these mice were backcrossed to RAG-1-deficient mice, a high proportion of the mice developed spontaneous EAE. In addition, adoptive transfer of CD4 ϩ T cells from non-TCR-transgenic mice into MBP-TCRtransgenic mice on the RAG-1-deficient background prevented spontaneous autoimmune disease (4). These and other data indicate that disease induction and severity can be regulated by the presence of peripheral CD4 ϩ T cells.
A good candidate for regulatory cells that may modulate disease severity and onset are CD4 ϩ Th2 cells. Th2 cytokines can clearly antagonize the function and differentiation of Th1 cells (5,6), and IL-4 and TGF-␤ (7), and IL- 10 (8) can be detected in the CNS in animals recovering from EAE. Additionally, neutralization or the absence of the Th2 cytokines IL-4 and IL-10 can result in more severe disease (9 -12). Attempts at regulating disease however with Th2 cells have met with mixed results. Although Th2 clones have been reported to successfully prevent EAE when given at the same time as Ag (13,14), adoptive transfer of CNS-Ag-specific Th2 cells either at the same time or following transfer with committed encephalitogenic Th1 cells fails to inhibit disease (15,16) or may only partially eliminate disease (17). A recent study showed that Th2 cells may also induce EAE when transferred into T cell-deficient animals (16).
Induction of Th2 responses concordant with the Th1 response has been shown to successfully suppress EAE. Expression of Th2 cytokines by activated bystander cells, at the induction of disease, will effectively inhibit the Th1 response by immune deviation, where the cytokine profile is shifted from an encephalitogenic Th1 type to Th2 to alter disease severity (18). SJL mice that were preimmunized with keyhole limpet hemocyanin in IFA to generate a memory Th2 population did not develop EAE when challenged with keyhole limpet hemocyanin and encephalitogenic peptide (18). In another system, immunization of mice with LR, an APL of PLP peptide 139-151, protected mice from EAE induced with PLP peptide as well as unrelated encephalitogenic peptides. The T cell population that is induced by LR is of a Th2/Th0 phenotype, cross-reacts with PLP 139-151, and can down-regulate actively induced EAE through bystander effects (19,20).
We have used an adoptive transfer model of EAE to examine the potential contributions of Th2 cytokines in the regulation of Th1mediated disease. We show here for the first time that a Th2 cell responding to an APL can regulate EAE mediated by an encephalitogenic Th1 population. In these experiments, primed encephalitogenic lines are used to transfer disease after coincubation with a Th2 clone specific for LR. We have systematically examined the effects of Th2 cytokines, individually and in combinations, from this clone for their effects on disease. Our data indicate that particular combinations of Th2 cytokines are most effective at modulating disease. Interestingly, IL-4 and IL-13 were synergistic in suppression of disease, suggesting that effects mediated through the IL-4 receptor on T cells are complemented by effects on other cell types expressing the IL-4/IL-13 receptor.

Cell lines and clones
The LR 1F1 T cell clone was generated from lymph node cells from LR and PLP-immunized mice as described (20) and is maintained in clone medium. It does not cross-react with PLP 139-151 (20). 1F1 cells are restimulated every 3 wk in proliferation medium with 10 M LR peptide and 10:1 irradiated (3000 rads) male SJL spleen cells.

1F1 conditioned medium
Supernatants were harvested from 1F1 cells cultured at 10 6 cells/ml with 10 7 irradiated spleen cells and 10 M LR peptide after 72 h and tested in ELISAs. Control conditioned media were generated from cultures containing no LR peptide (nonactivated 1F1) and no 1F1 cells (peptide pulsed spleen cells).

DNA chip analysis
One microgram of 1F1 cytoplasmic RNA was extracted from 4-and 16-h anti-TCR-stimulated cells (1 g H57-597 PharMingen 01300D/ml coated tissue culture dishes). 1F1 cells (10 7 /ml) and poly(A ϩ ) RNA were isolated, amplified, and converted to cDNA and then to biotin-labeled RNA in an in vitro transcription reaction as described by Lockhart et al. (22). Samples were hybridized to 114 gene arrays (Affymetrix, Santa Clara, CA) in a self-contained flow cell and imaged with a scanning confocal microscope also as described (22). Data are expressed as the relative message frequency per 10 6 .

Adoptive transfer of EAE
Spleen cells from female SJL mice immunized for 8 -10 days with 100 g PLP 139-151 in CFA injected s.c. were harvested and cultured at 3 ϫ 10 6 cells/ml with 10 M PLP in proliferation medium (PLP-spleen). A total of 20 U/ml recombinant COS human IL-2 was added on day 2 of culture. Cultures were harvested after 3 days and washed, and 5 ϫ 10 6 -1 ϫ 10 7 cells were injected i.p. into naive 8 -12-wk-old SJL females. Some experiments were designed to test the effect of including 1F1 Th2 cells in the culture. 1F1/PLP-spleen cell cocultures typically included 3 ϫ 10 5 resting 1F1 cells/ml and 10 M LR peptide added at the same time as the PLPspleen cells. To neutralize Th2 cytokines in 1F1 cocultures, 10 g/ml of the indicated neutralizing Ab was added at the start of culture. A25Fc (21) at 2 g/ml was included in some cultures to neutralize IL13. Control cultures without 1F1 cells contained PLP-spleen cells with PLP peptide and the indicated Ab. Culture supernatants from PLP-spleen cultures and 1F1 cocultures were harvested after 3 days before adoptive transfer to measure cytokines. Mice were scored for EAE using the following scale: 1, limp tail; 2, unsteady gait; 3, hind limb paralysis; 4, hind and forelimb paralysis; and 5, death.

Coculture with a Th2 clone regulates adoptive transfer of EAE
Typically, immunization of SJL mice with the CNS peptide PLP 139-151 in CFA results in an inflammatory and demyelinating disease characterized by ascending paralysis and CNS mononuclear cell infiltrates (23)(24)(25). Adoptive transfer of lymph node or splenic T cells derived from mice primed with PLP 139-151 also results in induction of disease in naive mice. Previous work has shown that coimmunization of mice with PLP 139-151 and LR peptide, an altered peptide derived from PLP 139-151 in which the TCR contact residues Trp 144 and His 147 are changed to Leu 144 and Arg 147 , results in disease suppression (20). These results suggested that T cells primed by immunization with the LR peptide act to regulate disease.
LR 1F1 (1F1) is a T cell clone derived from SJL mice immunized with the LR peptide (20). To determine whether this T cell clone could regulate disease induced by primed encephalitogenic T cells, we compared the encephalitogenic potential of T cell populations from mice immunized with PLP 139-151 with similar populations cocultured and coactivated with 1F1 T cells. Immune spleen cells from mice immunized 8 -10 days earlier with PLP 139-151 were cocultured for 3 days with resting 1F1 T cells and the appropriate peptides. After 3 days, cells from cocultures were adoptively transferred into naive SJL mice, and the mice were monitored for disease. As shown in Fig. 1, mice that received cells from cultures containing only PLP 139-151 immune spleen cells activated with their corresponding peptide (PLP-spleen/PLP) developed EAE, whereas cultures containing both PLP-spleen cells and 10% 1F1 cells with their corresponding peptides (PLP-spleen/PLP ϩ 1F1/LR) did not transfer disease. LR peptide itself was unable to affect the disease transferred by PLP-spleen cells since cells from cultures containing both PLP and LR peptides transferred disease as well as PLP-stimulated spleen cells alone. Additionally PLP-spleen/PLP cultures that included 10% 1F1 cells without the LR peptide also transferred EAE, suggesting that the 1F1 cells required activation with APL to suppress transfer of disease by PLP-spleen cells. Transfer of 1F1 cells activated with normal spleen cells and LR peptide did not result in EAE.

1F1 decreases Th1 cytokines in cocultures
We examined the levels of Th1 and Th2 cytokines in the conditioned medium from cocultures containing increasing concentrations of 1F1 cells. Cells from these activated cocultures were then transferred to naive mice to evaluate their potential to transfer disease. As shown in Table I, cultures containing as few as 2% 1F1 cells with the PLP-spleen cells transferred less severe EAE than the PLP-spleen cells alone. Disease transfer from cocultures containing 20% 1F1 cells was completely inhibited. Cytokine analysis of coculture supernatant revealed 5-fold lower levels of the Th1 cytokines IL-2 and IFN-␥, and increasing levels of Th2 cytokines in proportion to the number of 1F1 cells included in the culture. Th2 CYTOKINES INHIBIT ADOPTIVE TRANSFER OF EAE To determine whether this inhibitory effect was mediated by a soluble factor or cell-cell contact, we established a transwell system in which PLP-spleen cells and 1F1 cells were separated by a 0.4-m membrane. Transwell cocultures in which PLP-spleen cells and 10% 1F1 cells were separated by a 0.4-m membrane did not transfer EAE when both peptides were included in the culture, whereas control cultures without LR peptide did transfer disease (data not shown), suggesting that the 1F1 inhibition is mediated by soluble factors.

1F1 cells secrete high levels of Th2 cytokines upon activation
To identify candidate proteins made by 1F1 cells that might regulate EAE, we used oligonucleotide arrays to examine the expression of 114 different genes known to be expressed in T cells, including cytokine-and lymphocyte-specific genes. RNA was isolated from resting 1F1 cells and from 1F1 cells that had been activated for 4 and 16 h with plate-bound anti-TCR Ab. The RNA was amplified, labeled, and hybridized to the oligonucleotide array as described above (see Materials and Methods). Message levels for 27 cytokines and 6 chemokines were examined. As depicted in Fig. 2, mRNA levels for several cytokines were significantly upregulated, 3-fold or greater, compared with t ϭ 0 at the 4-h time point. The mRNA for several genes of interest was increased from one to two copies per 10 6 messages at t ϭ 0 to several hundred copies per 10 6 messages at 4 h: IL-4, 906; IL-10, 608; IL- 13,1944; and GM-CSF, 846. TGF-␤ was increased 8-fold from 52 to 401 copies per 10 6 messages. Expression of macrophage-inflammatory protein1␣ and macrophage-inflammatory protein1␤ was also induced by activation. We also examined the expression of genes encoding 28 cytokine receptors and 42 lymphocyte cell surface proteins or adhesion molecules. Message levels for the cell surface receptors IL-2R␣, ␤, and ␥, TCR-␤, 4-1BB, and 4-1BBL were also increased upon activation. P-selection glycoprotein ligand-1 expression was rapidly down-modulated after activation.
To confirm that detection of messenger RNA by oligonucleotide arrays correlated with protein expression, ELISAs were conducted for cytokines detected by our analysis. ELISA results of 72-h culture supernatants from 1F1 cells activated with LR peptide and APC are shown in Fig. 3. The 1F1 cells consume both IL-2 and IL-4 during culture. All cytokines with the exception of IL-2 (which was added to all cultures on day 2) and IL-4 (18 ng/ml at 48 h) were maximally expressed in the supernatant at 72 h. The presence of each predicted cytokine that we could measure by an ELISA was confirmed. Resting 1F1 cells did not make detectable levels of cytokines.

IL-4 and IL-13 contribute to inhibition of EAE by the Th2 clone 1F1
To examine the role of Th2 cytokines in the modulation of the adoptive transfer model of EAE by 1F1 cells, neutralizing Abs or reagents to IL-4, IL-10, or IL-13 were included in the 1F1/PLPspleen cocultures as indicated in Fig. 4. ELISA (data not shown) confirmed neutralization of each cytokine in the coculture supernatant. Neutralization of either IL-4 or IL-13 alone only partially abrogated the protective effect of the 1F1 cells. In each case, the disease observed was less severe than that of the activated spleen-PLP controls, suggesting that other factors made by 1F1 could also contribute to the suppression of EAE. Anti-IL-10 did not neutralize any of the protective effects of 1F1 cells and, interestingly, anti-IL-10 in combination with anti-IL-4 appeared to counteract the previously observed neutralization effect of anti-IL-4 alone. Animals transferred with PLP-spleen/1F1/LR cocultures containing combinations of anti-IL-4, anti-IL-10, and anti-IL-13 or anti-IL-4 and anti-IL-13 had significantly more disease than mice receiving untreated cocultures. Cultures containing all three Abs however transferred slightly less disease (less severity and delayed onset) compared with PLP-spleen control cultures not containing 1F1 cells, suggesting that not all down-regulatory effects had been neutralized. Interestingly, inclusion of all three Abs in control PLPspleen/PLP cultures resulted in the transfer of significantly less

Role of TGF-␤ in suppression of EAE by Th2 clone 1F1
TGF-␤ was detected both in activated 1F1 supernatants (670 pg/ ml) and at low levels of mRNA by DNA chip analysis. As shown in Fig. 5, addition of anti-TGF-␤ to PLP-spleen-1F1 cocultures partially neutralized the protective effect of 1F1, suggesting that TGF-␤ plays a role in suppressing transfer of EAE by coculture with the 1F1 cells. Since neutralization of IL-13 alone (Fig. 4) removed a significant amount of 1F1 suppression, we tested whether the combination of anti-IL-13 and anti-TGF-␤ would be additive. Interestingly, neutralization of the combination of IL-13 and TGF-␤ was no more effective than neutralization of either cytokine alone. This effect was also observed with the combination of anti-IL-4 and anti-TGF-␤ (data not shown). Neutralization of the four cytokines (IL-4, IL-13, IL-10, and TGF-␤) however not only abrogated the protective effect of the 1F1 cells but actually exacerbated disease severity when compared with PLP-spleen cells cultured without 1F1 cells.

Combinations are more effective than individual Th2 cytokines
To determine whether individual Th2 cytokines suppress adoptive transfer of EAE, each cytokine was added separately to cultures of PLP-spleen cells at the approximate concentrations that were measured in cultures containing activated 1F1 cells that did not transfer disease. As shown in Fig. 6, both IL-4 (25 ng/ml) and TGF-␤ alone were able to significantly inhibit EAE, whereas IL-13 (100 ng/ml) was only partially inhibitory. Addition of IL-9 (10 ng/ml) or IL-10 (50 ng/ml) did not significantly affect the disease course whereas GM-CSF appeared to slightly enhance the severity of EAE. Fig. 7 shows that the combination of a suboptimal dose of IL-4 (5 ng/ml) with either IL-10 or IL-13 was synergistic in suppressing disease and was as effective as the combination including the four cytokines IL-4, IL-10, IL-13, and TGF-␤. Levels of the cytokines IL-2, IL-3, IL-4, IL-10, IL-17, GM-CSF, IFN-␥, and TNF-␣ from supernatants of cytokine-treated cultures were measured by  ELISA and compared with PLP-spleen/PLP control cultures. Table II shows each measured cytokine expressed as a percentage of the control culture. The single cytokines which had the greatest effect in modulating transfer of disease, 1.0 ng/ml TGF-␤, 25 ng/ml IL-4, 300 ng/ml IL-13, and IL-4 plus IL-13, also contained significantly lower levels of IFN-␥ compared with control PLP-spleen cultures. Cells from these cultures also transferred significantly less disease (Figs. 6 and 7). IL-4 and IL-10 were present at low levels in some cultures, but none of the cytokine treatments significantly altered these levels from controls (data not shown). TGF-␤ suppressed the levels of all measured cytokines with the exception of IL-17, whereas IL-4-treated cultures had increased levels of IL-2, IL-3, GM-CSF, and TNF-␣ and decreased levels of IL-17. Although the combination of IL-4 and IL-13 was more effective in suppressing transfer of disease than IL-13 alone, both cultures showed reduced levels of IFN-␥ but only the IL-4 and IL-13 combination showed reduced levels of IL-17.

Discussion
Disease course in EAE is determined by the number and activation state of encephalitogenic T cells, as well as the cellular context in which the T cells are activated (15)(16)(17). Th2 cells have been pro-posed as a regulatory cell type in EAE, and introduction of Th2 cytokines or cells during the initial priming phases of the immune response can result in disease suppression (11,13,18,19,20). However, Khoruts et al. (15) found that adoptive transfer of Agspecific polarized Th2 populations did not inhibit EAE induced with Th1 cells. In addition, EAE was induced after adoptive transfer of Th2 cells into RAG-1-deficient mice (16). Thus, the precise mechanisms by which Th2 cells and cytokines can modulate disease remain to be defined. Immunization of mice with analogues of encephalitogenic peptides such as LR can protect mice from actively induced EAE. A previous report shows that LR induces a cross-reactive Th2/Th0 T cell population which protects mice from EAE in a bystander fashion (20). The experiments presented in this report demonstrate that a Th2 cell clone, responding exclusively to an APL of a known encephalitogenic peptide, can regulate EAE mediated by an encephalitogenic PLP 139-151 -specific Th1 population. We have conducted experiments to evaluate the effects of individual Th2 cytokines and combinations of Th2 cytokines made by a LR-responsive clone on primed, encephalitogenic T cells and EAE. We have used an adoptive transfer model of EAE induced with PLP 139-151 in the SJL mouse to examine the potential role of Th2 cells and cytokines in modulating the disease induced by PLPprimed T cells. We demonstrated that coculture of a Th2 cell line, 1F1, with PLP-primed encephalitogenic Th1 cells prevented adoptive transfer of EAE. Disease prevention was dependent upon activation of the Th2 cells, since cocultures containing 1F1 cells without the activating LR peptide transferred EAE. The inhibition by the 1F1 cells was due to soluble bystander effects since the 1F1 cells only responded to a non-cross-reactive analogue peptide included in the culture and were still inhibitory if activated on separate sides of a transwell membrane (data not shown).
We went on to define the mechanism by which the 1F1 cells mediate disease suppression and have systematically examined the role of cytokines produced by these cells. Both gene chip analysis and ELISA identified potential immunoregulatory cytokines expressed and secreted by the 1F1 cells. By neutralizing and addback experiments, our results showed that several Th2 cytokines made by 1F1 cells were capable of suppressing the adoptive transfer of EAE. We found that IL-4, IL-13, and TGF-␤ were critical for the observed down-modulation of disease by 1F1 Th2 cells.
We found that IL-10 alone did not play a major role in downregulating the induction of disease by adoptive transfer. The described role for IL-10 in the regulation of EAE is complex, having both exacerbating and protective functions depending on the timing of IL-10 administration and, most likely, the types of other cytokines that are present. IL-10 has potent down-regulatory effects during the priming and induction phases of the response to autoantigen (26 -28) as well as in the recovery phase. If administered at the same time as PLP, IL-10 delayed disease onset and severity (27). Mice transgenic for IL-10 were resistant to the induction of EAE (9), suggesting that encephalitogenic cells do not develop in the presence of IL-10. In later stages of disease, IL-10 mRNA expression is increased during the recovery phase (8,10) and IL-10-deficient mice develop a more chronic disease following immunization with myelin oligodendrocyte glycoprotein (9,29), both suggesting that IL-10 is required for recovery from EAE. In adoptive transfer models however IL-10 is ineffective in inhibiting disease. IL-10 pretreatment of lymph node cells sensitized to MBP before adoptive transfer had no effect on their ability to transfer disease (12). Similarly, treatment of mice with IL-10 at various intervals after adoptive transfer showed no or a worsening effect on clinical signs of EAE. These results suggest that IL-10 alone has little or no effect on primed Th1 cells or established disease. As shown by the experiments described in this report, addition of as much as 50 ng/ml IL-10 had little or no effect on disease transferred by activated T cells. We have also been unable to inhibit proliferation of PLP-spleen cells to PLP peptide in vitro (data not shown).
Neutralization of IL-10 in the 1F1 cocultures suggested that this cytokine alone does not play a significant role in suppressing disease and that additional cytokines from 1F1 are involved in mediating suppression. Moreover, 1F1 cocultures with IL-10 present (treated with anti-IL-4 and anti IL-13) transferred more severe disease than cultures without IL-10, implying an enhancing role for IL-10. Removal of only IL-4 and IL-10 from the cocultures (leaving IL-13 and TGF-␤) was ineffective in abrogating suppression whereas removal of IL-4 alone (leaving IL-10, IL-13, and TGF-␤) resulted in disease, again suggesting that IL-10 contributes to enhanced encephalitogenicity. In control PLP-spleen cultures as well, neutralization of IL-10 alone resulted in less severe disease and neutralization of the combination of IL-10, IL-4, and IL-13 resulted in significantly less disease, suggesting that IL-10 in combination with other cytokines could enhance encephalitogenicity. These results from the experiments with Abs to IL-10 suggest that the suppressive effects of IL-4 and IL-13 in 1F1 cocultures are more profound in the absence of IL-10. The direct cytokine addback experiments demonstrate that IL-10 alone neither suppresses disease transfer nor significantly enhances EAE when included at the 10-ng/ml level, but can suppress disease when included with a low level of IL-4. The effect of IL-10 appears to be complex and may enhance or suppress Th1 effector function depending upon the nature of the cytokine milieu.
In contrast to IL-10, we observed that activation of PLP-specific spleen cells in the presence of IL-4 inhibited adoptive transfer of EAE (Fig. 6). These data are in agreement with studies in the  (30). IL-4 plays a less significant role in regulating the induction of EAE and in recovery from actively induced EAE. IL-4-transgenic mice develop slightly more severe EAE with similar recovery times compared with nontransgenic littermates, and IL-4-deficient mice do not develop more severe EAE when compared with wild-type controls (9, 10).
IL-4 was particularly effective in inhibition of adoptive transfer of EAE in combination with other Th2 cytokines. We observed that IL-4 synergized with IL-10 to inhibit adoptive transfer of EAE  in experiments using suboptimal concentrations of IL-4, even though IL-10 alone had no effect on disease severity. Surprisingly, neutralization of IL-10 and IL-4 in the 1F1 cocultures had no effect on disease severity, although neutralization of IL-4 alone partially abrogated the suppressive effects of the 1F1 cells. Interference between the suppressive effects mediated by IL-4 and IL-10 was also seen in the direct disease model by Nagelkerken et al. (27). In this experiment IL-4 had no effect on the severity of EAE when injected at the same time as PLP peptide in SJL mice, but abrogated the inhibitory effects of IL-10 when administered with IL-10. In our hands, IL-4, unlike IL-13 or TGF-␤ alone, also significantly reduced the levels of IL-17 when compared with control PLPspleen cultures. IL-17 has recently been reported to stimulate the production of IL-1 and TNF-␣ from macrophages, suggesting a role for IL-17 in promoting an inflammatory response (31).
We found that murine IL-13 can inhibit the adoptive transfer of EAE in PLP-induced disease in SJL mice. Additionally, neutralization of IL-13 in the 1F1-spleen cell cocultures demonstrated a role for IL-13 in suppressing disease transfer. In agreement with our data, expression of human IL-13 by a hybridoma cell line adoptively transferred into rats results in suppression of clinical signs of EAE induced by immunization with MBP-CFA (32). The results in the report show that IL-13 synergized with IL-4 to inhibit adoptive transfer of EAE, although both IL-4 and IL-13 were able to independently inhibit adoptive transfer of EAE at high concentrations.
The mechanism for suppression by IL-13 is not clear, although it is likely to be functioning through APC in the coculture. IL-13 has been reported to exert both pro-and anti-inflammatory effects on monocytes. IL-13 up-regulates the capacity of monocytes to present Ag, resulting in increased proliferation of T cells to Ag (33). IL-13 can have a priming immunostimulatory effect on monocytes if added before an inflammatory stimulus and suppressive effect when added after (34). In our hands, proliferation of PLP-specific lymph node or spleen populations in the presence of PLP peptide over a wide range of IL-13 concentrations was not inhibited and was enhanced at higher concentrations of IL-13 (data not shown).
IL-4 and IL-13 use a shared receptor on monocytes and other APC (35). Both IL-4 and IL-13 can antagonize the effects of IFN-␥ and down-regulate inflammatory cytokines (33). We observed decreased levels of IFN-␥ in cocultures containing 1F1 cells and in PLP-spleen cultures containing the combination of IL-4, IL-13, and IL-10, suggesting that IFN-␥ made by the Th1 population was down-regulated in these cultures. Both IL-13 alone and IL-4 alone also decreased IFN-␥ levels in these studies. In contrast, we observed increased levels of TNF-␣ in PLP-spleen activation cultures containing IL-13 alone or IL-4 and IL-13 in combination.
IL-4 exhibits comparable anti-inflammatory effects to IL-13, but IL-4 and IL-13 do not exhibit synergistic effects on monocytes (33). IL-4 unlike IL-13 can enhance proliferation of activated T cells (33), and the synergy we observe between IL-13 and IL-4 may be due to the actions of IL-13 and IL-4 on distinct APC and T cell populations, respectively.
As with IL-13, TGF-␤ can have both proinflammatory and antiinflammatory effects, depending on concentration, the presence of other cytokines, and activation status of the T cell (36). Neutralization of TGF-␤ in the 1F1/PLP-spleen cocultures showed that this cytokine contributed significantly to the inhibition of EAE by 1F1 cells. Addition of as little as 100 pg/ml of TGF-␤ to the culture before adoptive transfer also inhibited development of EAE. Of all of the cytokines made by the 1F1 cells, we found that only TGF-␤ also inhibits the proliferation of PLP-specific LNC or spleen cell populations in vitro (data not shown). Previous reports have de-scribed the in vivo inhibitory role of TGF-␤ in EAE (37)(38)(39)(40) and have shown that addition of anti-TGF-␤ exacerbates EAE, suggesting a regulatory role for endogenous TGF-␤ (41,42). We found that addition of low levels of TGF-␤ alone inhibited adoptive transfer of EAE by PLP-spleen cells and suppressed levels of Th1 cytokines in the culture supernatant. Two additional experiments supported inhibitory roles for TGF-␤ in our adoptive transfer model of EAE. Neutralization of TGF-␤ in cocultures of 1F1 cells with PLP-spleen restored disease transfer and addition of anti-TGF-␤ in combination with Abs to IL-4, IL-10, and IL-13 would sometimes increase disease severity, supporting the concept that TGF-␤ was itself an endogenous regulator of the Th1 population. TGF-␤ in combination with IL-4, IL-10, and IL-13 showed the most profound inhibitory effect on both the transfer of EAE and the production of Th1 cytokines in the culture supernatant, suggesting that the presence of additional Th2 cytokines are important for a maximum inhibitory effect.
Other 1F1 cytokines evaluated in this adoptive transfer model were IL-5, IL-9, and GM-CSF. Not unexpectedly, IL-5, with reported activities on eosinophils and B cells, and IL-9, with activities on mast cells and some T cells (43,44), were found to have negligible effects on the adoptive transfer of EAE in this report. Preliminary results with weakly encephalitogenic PLP-spleen populations however suggest that IL-9 may have a subtle inhibitory effect on adoptive transfer of EAE. We have found that IL-9 enhances proliferation of activated T cell populations, and it is formally possible that the IL-9 treatment allows one to transfer lower numbers of encephalitogenic T cells for disease induction. GM-CSF on the other hand enhances Ag presentation and secretion of inflammatory cytokines in macrophages (45). In this report, addition of a low level of GM-CSF to PLP-spleen cultures had no effect whereas inclusion of higher levels led to a marginal enhancing effect in the adoptive transfer of EAE.
Other cytokines made by 1F1 cells such as IL-6 and IL-3, not evaluated in adoptive transfer experiments presented in this report, have been implicated to play a role in encephalitogenicity. EAE cannot be induced in IL-6-deficient mice (46,47) nor can IL-6deficient myelin oligodendrocyte glycoprotein-specific T cells transfer disease, suggesting that IL-6 is essential for the development of disease (46). IL-3 has also been reported to enhance the growth of encephalitogenic T cells (48), and we have found that IL-3 enhances the proliferation of Th1 cells as well (data not shown).
In summary, our results indicate that the most effective means of modulating EAE by immune deviation results from the induction of certain combinations of Th2 cytokines, which can modulate EAE induced by the adoptive transfer of primed, encephalitogenic T cells. The most effective combinations included TGF-␤, IL-4, and IL-13. This combination of cytokines is similar to that observed in the previously described Th3 subset of regulatory cells (13) and is predicted to have effects on both the APC and T cell populations. Our data support the idea that induction of specific combinations of Th2/Th3 cytokines could be of potential therapeutic benefit, in that suppression of disease mediated by primed, encephalitogenic T cells was observed. Our data also suggest that the variation in results obtained in other studies addressing the role of Th2 cells in modulating EAE may be due to differences in the precise cytokine profiles expressed by these various populations.