Interleukin 10 Secretion and Impaired Effector Function of Major Histocompatibility Complex Class II–restricted T Cells Anergized In Vivo

The TCR-HA transgenic mice expressing a TCR-α/β specific for peptides 111–119 from influenza HA presented by I-E^d have been previously described (6, 26). Mice expressing HA (IG-HA) have the HA transgene under control of the Ig-κ promoter and enhancer elements and are on BALB/c background (6). TCR-HA × IG-HA double transgenic mice were bred in our animal facilities. TCR-HA mice on recombination activating gene (RAG)-2^−/− background have been described (25). INS-HA mice, already described (27), were backcrossed onto the RAG^−/− background and were bred onto the H-2^d haplotype. Transgene expression was determined by PCR. Animal care was in accordance with institutional guidelines.

In vitro assays were performed in complete IMDM as previously described (6). Responder cells (2 × 10⁵/well) were isolated from the spleen of TCR-HA × IG-HA or TCR-HA mice and cultured in the presence of 5 × 10⁵ total irradiated (2,200 rads) BALB/c splenocytes and different concentrations of the HA peptide. Responder cells were also stimulated with coated anti-CD3 at 10 μg/ml or with coated 6.5 mAb at 100 μg/ml in absence or presence of IL-2 (30 UI/ml). ³H incorporation was measured over the last 18 h of a 66-h culture.

The following mAbs were used for staining: F23.1-biotinylated (specific for the TCR-β chain of HA-reactive TCR), 5(6)-carboxyfluorescein-N-hydroxysuccinimide ester (FLUOS)-labeled 6.5 clonotypic mAb, and SA-PE (Southern Biotechnology, Birmingham, AL). All stainings were done in 96-well plates (5 × 10⁵ cells/well) in 20 μl of mAb in PBS plus 2% FCS plus 0.1% sodium azide for 20 min on ice. Between first and second step reagents, cells were washed in PBS plus 2% FCS plus 0.1% sodium azide, as was done after the last step. Data were analyzed on a FACScan^® (Beckton Dickinson, Mountain View, CA), using Lysis II software.

For cell sorting experiments, splenocytes were depleted of surface Ig⁺ cells using Dynabeads (Dynal, Oslo, Norway), and cells were subsequently stained with F23.1 and 6.5 mAbs. F23.1⁺6.5^hi cells were sorted with a FACS^® Vantage.

Sorted F23.1⁺6.5^hi spleen cells from TCR-HA × IG-HA mice, TCR-HA mice, or total lymph node and spleen cells from TCR-HA mice on RAG^−/− background were washed in serum-free IMDM and were injected into the lateral tail vein of RAG^−/− INS-HA^+/+ mice. Recipients were tested for urine glucose levels every day and diabetes was confirmed by blood glucose levels.

Pancreata were quick frozen in O.C.T. compound, and 5-μm sections were obtained and fixed in acetone for 10 min. For immunohistochemistry, primary antibodies consisting of biotin-conjugated rat anti–mouse CD4 (clone GK1.5) and rat anti– mouse CD8 (clone 53-67.2) were applied on sections for 45 min at the appropriate dilution. After washing in PBS, slides were incubated with a biotinylated anti–rat κ antibody (Immunotech, Marseille, France) for 30 min, washed again, and finally incubated with streptavidin-peroxidase (Vector Laboratories, Inc., Burlingame, CA). Color reaction was revealed with 3-amino-9-ethyl-carbazole (AEC; Sigma Chemical Co., St. Louis, MO), and the slides were counterstained with hematoxylin. For the insulin staining, a guinea pig anti–porcine insulin antibody (Dako, Glostrup, Denmark) and a peroxidase-conjugated anti–guinea pig antibody (Dako) were used as primary and secondary antibodies, respectively.

Total RNA was isolated from F23.1⁺6.5^hi sorted splenocytes (2 × 10⁵ cells) from TCR-HA single transgenic and TCR-HA × IG-HA double transgenic mice using RNeasy Mini Kit (Qiagen, Hilden, Germany). Reverse transcription was carried out using Superscript II reverse transcriptase (RT; GIBCO BRL, Gaithersburg, MD). The cytokine polycompetitor plasmid pQRS was used to quantitate amounts of transcripts for IL-4, IL-10, IFN-γ, and the constitutively expressed hypoxanthine guanine phosphoribosyltransferase (HPRT), using the primers and PCR conditions as described previously (28). PCR products were analyzed by electrophoresis on 2% agarose gels and visualized by ethidium bromide staining under UV illumination, and then photographed (665 film; Polaroid, St. Albans, Hertfordshire, UK). Image densitometric analysis was performed using National Institutes of Health Image 1.61 software (SCR, Bethesda, MD) by integrating the volume in individual amplicons. After subtraction of background values, the density ratio of the competitor band to the target mRNA was determined and relative amounts of cytokine mRNA were calculated based on the starting amount of competitor.

The enzyme-linked immunospot (ELISPOT) assay was performed as previously described (29). In brief, F23.1⁺6.5^hi sorted splenocytes from TCR-HA single transgenic and TCR-HA × IG-HA double transgenic mice were transferred by threefold dilutions (starting at 10⁴ cells/well) in duplicate to 96-well microtiter plates (Millipore, Bedford, MA) that had been coated with the capturing mAb to IL-10 (JES5-2A5; PharMingen, San Diego, CA). After 20 h, cells were removed and spot-forming cells visualized with biotinylated detecting mAb (SXC-1; Pharmingen) and avidin peroxidase in conjunction with 3-amino-9-ethylcarbazole (Sigma Chemical Co.) substrate. Spots were counted under a dissecting microscope, and the frequency of antigen-specific cells was determined in both types of mice.

As described previously (6), mice that express a transgenic TCR specific for peptides 111–119 from influenza HA (TCR-HA), as well as influenza HA under the control of the Ig-κ promoter (IG-HA) contained CD4⁺ T cells in periphery that expressed high levels of the transgenic TCR (detected by the idiotypic 6.5 antibody). The proportion of these cells, which increased with the age of these mice, exhibited activation markers (like CD69), but did not proliferate when stimulated with different concentrations of antigen in vitro (Fig. 1). Such cells did not respond when stimulated with the clonotypic antibody, even after addition of exogenous IL-2. In the following, we have further characterized these cells that appear to behave differently from in vitro anergized cells.

To determine whether the lack of proliferation in vitro was reflected by a lack of effector function in vivo, it was tested whether anergic 6.5^hi cells could cause diabetes after transfer into rearrangement-deficient RAG^−/− mice expressing HA under the control of the rat insulin promoter (INS-HA). In this system, transfer of 10⁴ or 10⁵ 6.5⁺ T cells from TCR-HA single transgenic mice on the RAG^−/− background or of 10⁵ 6.5^hi cells sorted from TCR-HA single transgenic mice induced diabetes at days 13, 9, and 10, respectively, after transfer (Table 1). The injected mice died within 12 to 16 d after transfer. In contrast, 10⁵ 6.5^hi sorted cells from TCR-HA × IG-HA double transgenic mice did not induce diabetes with the same kinetics or the same severity, since recipients only became diabetic 16 d after transfer of cells (Table 1) and survived for >40 d after transfer. The significant delay in the onset of diabetes in the RAG^−/− INS-HA mice that received anergic cells was not due to the fact that these cells could not repopulate the host or could not home to the pancreas, as T cells were detected both in the lymph nodes and in the pancreas 11 d after transfer (Fig. 2). Histological examination of the pancreas of recipients of naive versus anergic 6.5^hi cells revealed striking differences; in the former, infiltrating cells were detected in both the exocrine and endocrine (islets) tissue, the islets were completely disrupted, and at the day of diabetes onset, insulin-producing cells were no longer detectable. On the contrary, the distribution of anergic 6.5^hi cells was restricted to the islets and insulin-producing cells were still detectable after diabetes onset (Fig. 3). Thus, the impaired in vitro response of the anergic HA-specific cells was reflected by an impaired effector function in vivo that may be caused by a defect in in vivo proliferation and/or effector function in the pancreas. The precise reason for the late development of diabetes is unclear. We cannot exclude the possibility that among the 6.5^hi cells from TCR-HA × IG-HA double transgenic mice, there are recent thymic emigrants that have not yet been rendered anergic and that are causing the slow and mild diabetes observed in the recipients.

In a first attempt to analyze the lack of effector function of the 6.5^hi T cells from the double transgenic mice in some greater detail, the mRNA level of various cytokines was determined by competitive PCR. To this end, 6.5^hi expressing cells from either TCR-HA single or TCR-HA × IG-HA double transgenic mice were sorted, and mRNA levels were compared (Table 2). Although IFN-γ mRNA was not detectable in T cells from both mice, only cells from the double transgenic mice contained low levels of IL-4 mRNA. By far, the most significant difference was observed with regard to IL-10; very high levels of this cytokine mRNA were found in 6.5^hi cells from TCR-HA × IG-HA double transgenic, but not from TCR-HA single transgenic mice (Fig. 4). These results were confirmed with sorted CD4⁺6.5⁺ cells from the single and double transgenic mice. The lack of IL-10 transcripts in 6.5⁺ cells from single TCR transgenic mice was not due to the lack of activation since 6.5⁺ cells transferred to INS-HA recipients did not produce IL-10 mRNA up to a point when the recipient mice became diabetic (not shown).

To determine whether the increased levels of IL-10 mRNA reflected an increased secretion of the protein, we performed ELISPOT assays on 6.5^hi sorted cells from single and double transgenic mice. Fig. 5 clearly confirms that the anergic 6.5^hi cells produced and secreted significantly higher levels of IL-10 than T cells from single transgenic mice.

These data raise the interesting possibility that so called anergic cells, as obtained in vivo through continuous stimulation with antigen, are in fact not really anergic but have assumed the phenotype of regulatory cells. IL-10 has been reported to downregulate costimulation by APCs (30), dendritic cell–driven IFN-γ production by T cells (31), and T cell responses to antigen through inhibition of IL-2 production and IL-2R α chain expression (32, 33). Furthermore, adoptive transfer of autoreactive T cells genetically designed to secrete IL-10 were able to delay the onset of experimental autoimmune encephalomyelitis in mice (34). Finally, a recent study by Groux et al. (35) showed that OVA-specific T cell clones derived from in vitro cultures in the presence of IL-10 and chronic antigenic exposure, were able to prevent T cell–mediated inflammatory bowel disease when cotransferred with CD4⁺CD45RB^hi cells into SCID mice. These cells exerted their influence only upon stimulation with antigen, i.e., when the recipients were fed with OVA.

We show here that IL-10–producing T cells cannot only result from unphysiological maneuvers like genetic alteration or growth in IL-10–containing media, but also develop as a result of chronic antigenic stimulation in vivo. The IL-10 production by anergic T cells is in line with previous observations that repeated superantigen stimulation results in anergy as well as in IL-10 production (36) and with other data showing IL-10 production in SCID patients transplanted with HLA-mismatched hemopoietic stem cells (37). The exact source of IL-10 was not obvious from these experiments and it was not clear whether the unresponsive state of T cells in terms of proliferation and effector function correlated with IL-10 production. The results presented here directly show that T cells unable to proliferate in response to antigen in vitro and with an impaired specific effector function in vivo, can nevertheless produce high levels of IL-10 and therefore may regulate immune response through the release of IL-10.

We are grateful to Carole Zober for excellent technical assistance.