Regulation of EAE by spontaneously generated IL‐10‐secreting regulatory T cells in HLA‐DR15/TCR.Ob1A12 double transgenic mice

Summary

Humanized double transgenic mice express both HLA‐DR15 (the MHC gene linked to MS) and TCR.Ob1A12 from a multiple sclerosis patient (that recognizes MBP85‐99 presented by HLA‐DR15), yet they fail to develop autoimmune encephalomyelitis quickly, although 5–10% develop disease at 12 months. These mice were found to express large numbers of IL‐10‐secreting splenocytes as early as 4 weeks of age. These regulatory T cells appeared spontaneously without prior immunization with the autoantigen MBP85‐99. They were of murine origin and had a cytokine secretion profile and surface phenotype similar to that reported for Tr1 cells. Notably, the frequency of disease appeared to increase at 14 months. The diseased mice had small spleens which averaged 47 mg, while the remaining non‐diseased mice in our colony killed at ages 14–15 months had splenocytes that averaged 80 mg (ranging from 47–130 mg). Thus, the appearance of disease was associated with diminution in numbers of IL‐10‐secreting regulatory T cells with age.

HLA‐DR15/TCR.Ob1A12 human double trangenic mice, express larges numbers of IL‐10 secreting splenocytes that appeared spontaneously without prior immunization with the autoantigen MBP85‐99.

Abbreviations

EAE

experimental autoimmune encephalomyelitis

GITR

glucocorticoid‐induced TNF receptor

IL‐10

interleukin‐10

MHC

major histocompatibility complex

TCR

T‐cell receptor

Introduction

Experimental autoimmune encephalomyelitis (EAE) can be induced in different strains of mice by autoantigens or peptides derived from them. ¹ The peptides that induce disease are specific for each strain and related to the MHC protein expressed by that strain and the relevant autoantigen. These mice have been studied extensively as models for the human disease multiple sclerosis that is genetically strongly associated with the MHC protein HLA‐DR15 (DRA/DRB1*1501). The pathogenic TCR. Ob1A12 cDNA, identified in a T‐cell line obtained from the patient OB, recognized myelin basic protein or the peptide MBP85‐99 derived from it in the context of HLA‐DR15 to induce cytolysis and disease. A double transgenic mouse was constructed using the HLA‐DRA/HLA‐DRB1*1501 genes together with the cDNAs for TCR.Ob1A12. ² This double transgenic mouse did not quickly develop EAE spontaneously despite the presence of the appropriate MHC protein, the endogenous autoantigen and the transgenic pathogenic TCR, although disease could be induced rapidly by immunization with MBP85‐99. However, EAE also appeared spontaneously when these mice reached 12 months at a frequency of 8–10%. A second strain of the same mice was generated in another laboratory and had the same phenotype. ³ Further investigation of this phenomenon, particularly of the failure to develop spontaneous EAE early, is reported in the present paper.

Materials and methods

Peptides

MBP85‐99 or MBP 84–102 and copolymers were synthesized as described. ⁴ , ⁵

Mice

Humanized double tg mice expressing both HLA‐DR15 (DRA/DRB1*1501) and TCR from the MS patient Ob in the Aβ⁰ background were generated by standard techniques as described. ² C57BL/6 (B6) mice were purchased from Jackson Laboratory. Mice were maintained according to the Guidelines of the Committee on Animal Care of Harvard University and the Committee on Care and Use of Laboratory Animal Resources, National Research Council.

Spleen cell isolation and culture

Humanized mice were immunized s.c., each with 200 μg MBP85‐99 emulsified in CFA. Ten days later, spleen cells were collected and purified using MACS MicroBeads and magnetic separation columns (Miltenyi Biotec Auburn, CA, USA). Splenic CD4⁺ T cells were isolated from splenocytes and purified using CD4⁺ T Cell Isolation Kit by MACS (Miltenyi Biotec Auburn, CA, USA) according to the manufacturer’s protocol. Then, splenic Vβ2⁺CD4⁺ T cells were isolated from CD4⁺ T cell by successively using an anti‐TCRVβ2 FITC antibody (Immunotech Burlingame, CA, USA) and purified using Anti‐FITC MicroBeads Kit by MACS (Miltenyi Biotec Auburn, CA, USA) according to the manufacturer’s protocol. The Vβ2^‐CD4⁺ T cells were obtained by negative selection after magnetic separation of Vβ2⁺CD4⁺ T cells. Cells were cultured in the presence of antigen‐presenting cells (APC) (irradiated splenocytes) and stimulated with MPB 84–102 (10 μg/ml) or unstimulated for 4 days. After that, cells were used for further experiments.

FACS analysis

Isolated splenocytes of double transgenic mice, Vβ2^‐ CD4⁺ T cells and Vβ2⁺ CD4⁺ T, isolated as described above, were unstimulated or stimulated with MBP84‐102 (10 μg/ml) for 4 days, and then, the cells were collected and stained with antibodies specific for mouse CD3, CD4, CD25, CD49b, CD69, GITR, LAG3 (BioLegend San Diego, CA, USA) and anti‐TCRVβ2 (Immunotech Burlingame, CA, USA) according to the manufacturer’s protocol. The cells were analysed in a FACScan flow cytometer using Cell Quest software (BD Bioscience San Jose, CA, USA).

Cytokine analysis

Splenocytes, splenic Vβ2⁺ CD4⁺ T cells and splenic Vβ2^‐ CD4⁺ T cells were harvested from mice and isolated as described above. Then, these cells were unstimulated or stimulated with MBP84‐102 (10 μg/ml) in the presence of APC (irradiated splenocytes) for 4 days. Supernatant was collected and analysed for cytokine production, IL‐4, IL‐5, IL‐10, IL‐17 and IFN γ, by ELISA (BioLegend San Diego, CA, USA ) according to the manufacturer’s protocol.

Results

Generation of IL‐10‐secreting T cells in double transgenic mice in the absence of immunization with MBP85‐99

Initially, published experiments ⁴ , ⁵ , ⁶ showing that immunization with MBP85‐99 (or alternatively MBP84‐102) of 8‐week‐old double transgenic mice induced EAE within 12 days in both male and female mice were reproduced, although no disease appeared without immunization. Induction was effectively blocked by administration of either amino acid copolymer polyYFAK (an alternative to Copaxone, polyYEAK, that is used therapeutically) or peptide 15mer J5 (the sequence of which was based on the motif for binding of Copaxone to HLA‐DR15). At the low dose of 30 micrograms per day for two weeks, J5 was particularly effective. Moreover, both MBP84‐102 and YFAK induced secretion of copious amounts of IL‐10 (1500–2500 pg/10⁶ cells under the conditions of assay) from either splenocytes or splenic CD4⁺ T cells isolated from these preimmunized mice. At the dose employed in vitro, J5 (10 μg/ml) produced much less IL‐10.

Next, as a control, IL‐10 secretion from splenocytes or splenic CD4⁺ T cells of double transgenic mice in response to MBP84‐102 was measured in the absence of prior immunization with the peptide or copolymer. Surprisingly, IL‐10 was produced in an amount identical to that found in mice preimmunized with MBP84‐102 (~1500 pg/10⁶ cells) (Fig. 1A,B). These experiments were carried out with 8‐week‐old double transgenic mice but the same result was obtained using 4‐week‐old mice (Fig. 1C). Thus, the CD4 + T cells producing IL‐10 were generated in utero or early in the post‐partum period.

Figure 1.

IL‐10 secretion from splenocytes of double transgenic mice in response to MBP 85‐99. (A) 8‐week‐old mice without immunization. (B) 8‐week‐old mice preimmunized with MBP 85‐99. (C) 4‐week‐old mice without immunization. *P < 0·05; **P < 0·01; and ***P < 0·005.

Moreover, when the CD4⁺ T cells were separated into those expressing human Vβ2⁺ (that identified the pathogenic human cDNA used in construction of the double transgenic mouse) and Vβ2^‐ murine T cells, the endogenous Vβ2^‐ T cells produced IL‐10 (1500 μg/10⁶ cells) on stimulation, but the transgenic Vβ2⁺ T cells did not (Fig. 2A,B). The production of IL‐5 and IL‐4 in these experiments was also increased, but both were very low (IL‐5, ~300 µg/10⁶ cells; IL‐4, ~80 µg/10⁶ cells) (Fig. 3). The Vβ2^‐ CD4⁺ T cells also produced relatively large amounts of INFγ and much smaller amounts of IL‐17, but the amount was not increased by stimulation with MBP84‐002 (Fig. 3). Because the IL‐10 secretion from murine splenocytes was somewhat higher than that from the isolated Vβ2^‐ CD4⁺ T cells, other cells from splenocytes were examined in the same manner. CD11c⁺, CD11b⁺ splenocytes (dendritic cells and macrophages) and CD19⁺ B cells failed to produce IL‐10 under the same conditions.

Figure 2.

Secretion of IL‐10 from splenocytes, Vβ2 + CD4+ T cells and Vβ2‐ CD4 + T cells of both 4‐week‐old (A) and 8‐week‐old (B) double transgenic mice in response to MBP 85‐99. ns: not significant; *P < 0·05; **P < 0·01; and ***P < 0·005.

Figure 3.

Secretion of additional cytokines. Note that the y‐axes have different scales in the various panels. *P < 0·05; **P < 0·01; and ***P < 0·005.

Additional mouse strains were used for the same experiments. Double transgenic mice are in the C57BL/6J (B6) background. CD4⁺ T cells from the parental B6 mice, as well as those from SJL mice, failed to produce IL‐10 in response to MBP84‐102. Importantly, moreover, CD4⁺ splenic T cells from HLA‐DR15 single transgenic mice (in the absence of the transgenic TCR and in the same B6 background as the double transgenic mice) also produced no detectable amount of IL‐10 and neither IL‐4 nor IL‐5 (Fig. 4).

Figure 4.

Cytokine secretion from splenocytes of single transgenic mice in response to MBP 85‐99. *P < 0·05; **P < 0·01; and ***P < 0·005.

Phenotype of IL‐10‐secreting T cells induced in vitro by MBP85‐99

An extensive literature on IL‐10‐secreting T cells from both mice and humans, called Tr1 cells, has been contributed by Roncarolo and her colleagues. ⁷ , ⁸ These cells are characterized particularly by the presence of both CD49b and LAG3 ⁷ and also express a variety of costimilatory molecules when activated, including CD40L, CD69, CD28 and PD1 and the regulatory factors GITR, OX40 and TNFRSF9. They have a unique cytokine profile secreting IL‐10, TGFβ, IL‐5 and interferon γ: IL‐4 and IL‐2 were either absent or were secreted at very low levels.

Freshly isolated splenocytes of 8‐week‐old double transgenic mice stimulated with MBP 85‐99 contained 18% LAG3⁺ CD49b⁺ and 27·5% LAG3⁺, CD49b^‐ CD4⁺ T cells as compared to 0·21% and 0·24% for unstimulated control mice. The increases were exclusively in the Vβ2^‐ preparation of CD4⁺ T cells (Fig. 5A). The unstimulated CD4⁺ T cells had low expression for CD25, FOXp3 and CD69. However, treatment of splenocytes with MBP85‐99 or MBP84‐102 in vitro simulated a large expansion to 39% and 37% CD25⁺ CD69^‐ T cells, but only small numbers of double or single positive Vβ2⁺ T cells. Parental and single positive HLA‐DR15 transgenic B6 mice showed no similar phenotype, that is, no expansion of cells expressing any of the above‐referenced proteins (Fig. 5B). A large increase in the level of glucocorticoid‐induced TNF receptor (GITR) that costimulates IL‐10 secretion on splenocytes on Vβ2⁺ T cells ⁸ was also observed (Fig. 5C).

Figure 5.

Phenotype of T cells induced by MBP85‐99. (A) CD49b, LAG3 expression. (B) CD69, CD25 expression. (C) GITR expression.

The effect of ageing on development of EAE and generation of IL‐10‐secreting T cells

Twenty‐eight double transgenic mice in our colony were available beyond 12 months of age. Of these, four mice developed EAE in 8–14 months and progressed to a score of ~ 3 at which time they were killed in accordance with Harvard IACUC policy. Their spleens were obviously small (average 47 mg) as compared to the remaining 14‐month‐old mice who did not have EAE (average 80 mg) (Fig. 6). The total number of cells in the animals who had EAE was also correspondingly reduced as compared to 14‐month‐old normal mice. However, the number of mice available was too small to reach statistical significance. Based on an equivalent number of cells, the residual splenocytes secreted an amount of IL‐10 (1700 pg/10⁶ cells) that was the same as those of the around 12‐month‐old normal mice. Examination of secretion of additional cytokines again revealed very little IL‐4 and modest amounts of IL‐5 in all groups. Again, with respect to proinflammatory cytokines, relatively large amounts of INF were secreted by splenocytes of untreated double transgenic mice and the amount was only modestly increased by addition of MBP84‐102 (around 1200 pg/10⁶ cells without MPB85‐99‐2000 pg/10⁶ cells with MBP85‐99). Only extremely small amounts of IL‐17 were found. With regard to phenotype, all groups exhibited enhanced expression of CD49b and LAG3 on stimulation of splenocytes with MBP84‐102. Similarly, enhanced expression of CD25 and CD69 was observed in all groups. Thus, no difference was found between CD4 T cells from 8‐week‐old mice, from 52‐week‐old mice without EAE and from 52‐week‐old mice with EAE, except for the reduction of the number of cells in the latter group. As the colony was being eliminated, we were unable to repeat the experiment which would require at least 18 months. These preliminary results are included for the interest of future investigators.

Figure 6.

Weight of splenocytes of aged double transgenic mice with or without spontaneous EAE. *P < 0·05; **P < 0·01; and ***P < 0·005.

Discussion

Splenocytes from double transgenic mice ² displayed an anti‐inflammatory response to administration of MBP84‐102, characterized by secretion of IL‐10 rather than a proinflammatory response. ² , ³ These mice did not develop EAE. However, at about 12 months, EAE did develop spontaneously in 4 of the 28 mice that reached this age, a low percentage similar to that previously reported. These mice had very small spleens, although the residual T cells secreted normal amounts of IL‐10 per cell on stimulation with MBP84‐102. We suggest that the reduction in cells that secreted anti‐inflammatory cytokines may lead to spontaneous EAE in these mice, although with the small number of animals available, statistical significance was not possible. Phenotypically, the IL‐10‐secreting T cells had all the characteristics of Tr1 cells, as previously described ⁹ , that is they were LAG3 and CD49b double positive as well as expressing both CD25 and CD69 as activation markers. They also secreted both interferon gamma and small amounts of IL‐5, but virtually no IL‐4. These cells appeared as early as 4 weeks of age spontaneously and may have been present at birth. Their appearance required the presence of a T‐cell receptor specific for MBP85‐99 since splenocytes from single transgenic HLA‐DR15 mice did not secrete IL‐10 either spontaneously or in response to MBP85‐99. Thus, in this study, as in an earlier study using retrogenic mice, ¹⁰ we hypothesized that some feature of the TCR, possibly a peptide derived by its proteolysis, may be required for the generation of these IL‐10‐secreting regulatory T cells.

A variant of double transgenic mice (called line 7) has been described in which EAE appears spontaneously in all mice at 8–10 weeks of age. ¹¹ These mutant mice are therefore very difficult to maintain as a colony. The present experiments raise the question whether the defect in the variant is an inability to generate IL‐10‐secreting regulatory T cells.

Disclosures

All authors declare that they have no Conflicts of Interest.

Data availability statement

Not applicable, all available data are included in the manuscript.