CD69 Regulates Type I IFN-Induced Tolerogenic Signals to Mucosal CD4 T Cells That Attenuate Their Colitogenic Potential

The intestinal immune system is continuously exposed to large amounts of potential beneficial or harmful Ags derived from constituents of the commensal flora and ingested food Ags (1–3). In healthy individuals, rare inflammatory immune responses against these Ags are observed although T cell reactivity against commensal-derived Ags have been described (4). These Ags induce an Ag-specific state of systemic immunologic unresponsiveness, a phenomenon called oral tolerance (5). Breakdown of oral tolerance is associated with severe immunopathology as observed in patients with inflammatory bowel disease (6, 7), in which Th1 and Th17 reactivity are a major driving force in the disease process (8, 9). T cells in the inflamed mucosa are characterized by high expression of the activation Ag CD69 (10). The activation Ag CD69 (encoded in the NK C-type lectin gene cluster) is a type II transmembrane protein of the C-type lectin family (11–13). In contrast to other members of the NK C-type lectin cluster, CD69 plays apparently no role in NK cell recognition of target cells (13). CD69 inhibited T cell egress from the thymus, as demonstrated by transgenic overexpression studies (14, 15). CD69 is also involved in regulation of T and B lymphocyte egress from secondary lymphoid tissues by blocking the spingosine 1-phosphate 1 receptor (16). Recent work in mice indicates a role of CD69 in the regulation of arthritis (17, 18), asthma (19, 20), myocarditis (21), pathogen clearance (22), and tumor immunity (23, 24). CD69 surface expression is upregulated by lymphocytes at mucosal sites such as intraepithelial and lamina propria (LP) lymphocytes (25, 26). Its function within the mucosal immune system is largely unknown. CD4⁺CD69⁺ T cells inhibit the secretion of proinflammatory cytokines and proliferation of CD4⁺CD69⁻ T cells in a process partially dependent on TGF-β1 (27, 28). Expression of CD69 is strongly induced by type I IFNs (IFN-I) (16). Protective effects of IFN-I have been shown in colitis models and clinical trials with ulcerative colitis patients (29–33).

We show in this study that the intestinal microflora induces expression of CD69 by CD4 T cells. The oral challenge of naive TCR-transgenic animals with cognate Ag induced CD69 expression by CD4 T lymphocytes. These CD69⁺ CD4 T cells were characterized by high latency-associated peptide (LAP)/TGF-β1 expression. Genetic deletion of CD69 attenuated secretion of the suppressive cytokine TGF-β1, whereas increased production of the inflammatory cytokines IFN-γ, TNF-α, and IL-21 by CD4 T cells. CD69-deficient CD4 T cells showed lower potential to become Foxp3⁺ regulatory T (Treg) cells in vivo and in vitro. Transfer of CD69-deficient T cells into RAG^−/− hosts induced a severe colitis. CD69^−/− cells showed impaired IFN-β1 induction by polyinosinic-polycytidylic acid [poly (I:C)], and beneficial effects of poly (I:C) treatment could not be observed in RAG^−/− hosts after transfer of CD69^−/− or IFNAR^−/− CD4 T cells.

Inbred C57BL/6J (B6) mice, RAG^−/− (RAG^tm1Mom) mice, CD69^−/− mice (17, 19), IFNAR^−/− mice (34), and transgenic OT-II, OT-II × RAG^−/− (35), and OT-II × CD69^−/− mice were bred and kept under specific pathogen-free (SPF) conditions in the animal facility of Ulm University (Ulm, Germany). Germ-free (GF) B6 mice were screened weekly for viral, bacterial, and fungal contamination. Female and male mice were used at 6–12 wk of age. All animal experiments were performed according to the guidelines of the local Animal Use and Care Committee and the National Animal Welfare Law.

Mice were treated with ampicillin (catalog no. A6140-5G; Sigma-Aldrich, Steinheim, Germany) 1 g/l added in water flasks and an antibiotic mixture consisting of 5 mg/ml vancomycin (catalog no. 861987-1G; Sigma-Aldrich), 10 mg/ml neomycin (catalog no. N6386-5G; Sigma-Aldrich), and 10 mg/ml metronidazol (catalog no. M3761-5G; Sigma-Aldrich) diluted in distilled H₂O. Gavage volume of 10 μl/g body weight was delivered intragastrically every 12 h. Fresh antibiotic mixture was mixed at every feeding point, and ampicillin in fresh water was renewed every seventh day. Mice subjected to this protocol for 17 d displayed the properties of GF mice as previously reported (36), and CD69 expression by spleen, mesenteric lymph nodes (MLN), small intestinal LP (siLP), and colonic LP (cLP) CD4 T cells was analyzed after 18 d of treatment.

Total spleen cells of B6, CD69^−/−, or IFNAR^−/− mice were isolated and stained for CD3, CD4, and CD45RB with PE-conjugated mAb binding CD3ε 145-2C11 (catalog no. 12-0031-83; eBioscience, Frankfurt, Germany), FITC-conjugated mAb binding CD4 GK1.5 (catalog no. 11-0041-86; eBioscience), and biotinylated mAb binding CD45RB 16A (catalog no. 553093; BD Biosciences, Heidelberg, Germany), followed by the second-step reagent Streptavidin-Peridinin Chlorophyll-a Protein-Cy5.5 (catalog no. 45-4317-80; eBioscience). CD45RB^high CD4 T cells from the CD3⁺ CD4⁺ population were enriched using the FACSAria system (BD Biosciences) to a purity >95%. Purified CD45RB^high CD4 T cells were injected i.p. (3 × 10⁵ cells/mouse) into RAG^−/− mice. In some experiments, hosts were i.p. treated twice a week with 20 μg poly (I:C) (catalog no. 27-4729-01, Amersham, Freiburg, Germany). The weight of transplanted mice and their clinical condition were monitored twice weekly. Tissue samples for histopathological examination were taken from the large intestine, fixed in neutral-buffered formalin, embedded in paraffin, sectioned on a microtome, mounted on slides, and stained with H&E. Histology of the large intestine was categorized as normal (score 0); mild colitis (score 1), with few inflammatory cells in the cLP, stroma edema, and a slight reduction of goblet cells; moderate colitis (score 2), with an intense inflammatory infiltration of the LP, hyperplasia of crypts, and a marked reduction of goblet cells; or severe colitis (score 3), with a spillover of leukocytes beyond the mucosa into deeper layers of the colonic wall, complete loss of goblet cells, distortion of the mucosal architecture, erosions or ulcerations, and crypt abscesses as previously published (9, 35).

Mice received intragastrically 1 mg OVA protein (catalog no. A5253, grade II; Sigma-Aldrich) dissolved in 100 μl PBS daily five times as previously reported (37). Control mice received PBS only. Seven days after the last treatment, mice were immunized s.c. in the base of the tail with 50 μl/mouse OVA protein and 50 μl/mouse oligodeoxynucleotide (ODN) 1826 (Thermo Fisher Scientific, Ulm, Germany) emulsified in 50 μl/mouse IFA (catalog no. F5506; Sigma-Aldrich) containing 50 μl/mouse PBS. One week later, spleen cells were isolated from the treated mice and restimulated with OVA_323–339 peptide ISQAVHAAHAEINEAGR (ISQAV) (catalog no. OR266830; Thermo Fisher Scientific). After 72 h, supernatants were collected, and IFN-γ concentration was determined by ELISA. Some mice were challenged by s.c. injection of 50 μg OVA protein in 12.5 μl PBS into the right ear pinna 1 wk after immunization. As a control, 12.5 μl only PBS was injected into the left ear pinna of the same mice. Ear swelling was measured before injection and daily for 3 d after injection. OVA-specific ear swelling was calculated as the following: (right ear thickness − left ear thickness)_48h − (right ear thickness − left ear thickness)_0h, as 48 h after injection was the peak of swelling.

CD4 T cells were isolated from the cLP, siLP, spleen, and MLN of B6, CD69^−/−, IFNAR^−/−, RAG^−/−, OT-II, OT-II × RAG^−/−, and OT-II × CD69^−/− animals.

Single-cell suspensions were aseptically prepared from spleen and MLN, washed, and resuspended in PBS supplemented with 1% FCS.

Segments of the small intestine or colon were washed with PBS to remove debris and mucous. The epithelium was removed by incubation at 37°C for 10–15 min under gentle shaking with 1 mM DTT (and 1 mM EDTA for colon tissue) in 25 ml PBS supplemented with 1% FCS. The remaining tissue was washed in PBS to remove residual epithelial cells, and the supernatants were discarded. Intestinal tissues were cut into 2 × 2-mm pieces and digested by incubation with 0.25 mg/ml collagenase type VIII from Clostrodium histolyticum (catalog no. C-2139; Sigma-Aldrich) for 30–45 min at 37°C in RPMI under shaking. Supernatants were collected from which LP lymphocytes were pelleted. LP lymphocytes were resuspended in RPMI medium containing 35% Percoll (density 1.124 g/ml; catalog no. L-6145; Biochrome, Berlin, Germany). This cell suspension was overlaid onto 70% Percoll and centrifuged for 20 min at 750 × g. Viable cells at the 35/70% Percoll interface were collected and washed twice.

The following reagents and mAbs from eBioscience were used: FITC-conjugated mAb binding CD4 GK1.5 (catalog no. 11-0041-86), allophycocyanin-conjugated mAb binding CD4 GK1.5 (catalog no. 17-0041-83); PE-conjugated mAb binding CD69 H1.2F3 (catalog no. 12-0691-82), IL-21R eBio4A9 (catalog no. 12-1219), biotinylated mAb binding CD25 PC61.5 (catalog no. 13-0251-81), and CD69 H1.2F3 (catalog no. 13-0691-81). From BD Biosciences, the following biotinylated mAbs were used: anti-CD103 M290 (catalog no. 557493), anti-CD122 TM-β1 (catalog no. 559884), anti–IFN-γR GR20 (catalog no. 550482), and anti-ICOS 7E.17G9 (catalog no. 552145) as well as PE-conjugated anti–IL-10R1 1B1.3a (catalog no. 559914). The biotinylated mAbs binding TGF-βRII (catalog no. BAF532) and APC-conjugated anti-LAP (TGF-β1) 27232 (catalog no. FAB2463A) were purchased from R&D Systems. As a second-step reagent, PerCP-Cy5.5–conjugated streptavidin (catalog no. 45-4317-80; eBioscience) was used.

Cells were isolated, washed with PBS supplemented with 0.3% BSA and 0.1% sodium azide, and stained extracellular with FITC-conjugated anti-CD4 mAb binding CD4 GK1.5 (catalog no. 11-0041-86; eBioscience) as described above. Surface-stained cells were fixed with 4% paraformaldehyde for 15 min in the dark, then pelleted and resuspended in 200 μl Fixation/Permeabilization buffer (catalog no. 00-5123-43; eBioscience). After overnight incubation at 4°C in the dark, permeabilized cells were washed two times with 1× Permeabilization buffer (catalog no. 00-8333-56; eBioscience) and incubated for 15 min, 4°C, with mAb 2.4G2 directed against the FcγRIII/II CD16/CD32 (0.5 ng mAb/10⁶ cells) for preventing nonspecific binding of the Abs. Cells were pelleted and, after discarding the supernatant, incubated for 30 min at 4°C in the dark with 0.5 ng/10⁶ cells of the following PE-conjugated Abs: anti-Foxp3 FJK-16s (catalog no. 12-5773-80; eBioscience), anti–T-bet eBio4B10 (catalog no. 12-5825-80; eBioscience), anti–GATA-3 TWAJ (catalog no. 12-9966-42; eBioscience), and anti-RORγt AFKJS-9 (catalog no. 12-6988-82; eBioscience). PE-conjugated Abs were diluted in 1× Permeabilization buffer. Stained cells were washed once with 1× Permeabilization buffer, once with PBS supplemented with 0.3% BSA and 0.1% sodium azide, resuspended in PBS supplemented with 0.3% BSA and 0.1% sodium azide, and analyzed on an FACSCalibur (BD Biosciences).

Cells were washed twice in PBS/0.3% w/v BSA supplemented with 0.1% w/v sodium azide. Nonspecific binding of Abs to FcRs was blocked by preincubation of cells with mAb 2.4G2 directed against the FcγRIII/II CD16/CD32 (0.5 ng mAb/10⁶ cells). Cells were washed and incubated with 0.5 ng/10⁶ cells relevant mAb for 20 min at 4°C and washed again twice. In most experiments, cells were subsequently incubated with a second-step reagent for 20 min at 4°C. Four-color flow cytometry (FCM) analyses were performed using an FACSCalibur (BD Biosciences). The forward narrow-angle light scatter was used as an additional parameter to facilitate the exclusion of dead cells and aggregated cell clumps. Data were analyzed using FCS Express V3 software.

Total spleen, siLP, or cLP cells of B6 or CD69^−/− mice were cultured in 200 μl medium containing RPMI, 10% FCS, 1% Penicillin/Streptomycin (catalog no. P11-010; PAA Laboratories, Pasching, Austria) with addition of anti-CD3/anti-CD28 dynabeads (catalog no. 11452D, Invitrogen, Karlsruhe, Germany) at a cell/bead ratio of 10:1. After 24 h at 37°C, 5% CO₂, the concentration of IFN-γ, TNF-α, and TGF-β1 was measured in the supernatants by ELISA. CD4⁺ T cells were enriched via a MACS isolation kit (catalog no. 130-090-860; Miltenyi Biotec, Bergisch Gladbach, Germany) from total spleen cells of B6 or CD69^−/− mice and cultured in 500 μl medium containing RPMI, 10% FCS, and 1% Penicillin/Streptomycin with addition of anti-CD3/anti-CD28 dynabeads at a cell/bead ratio of 10:1. After overnight incubation, anti-CD69 H1.2F3 (catalog no. 13-0691-81, eBioscience) Ab was added (1:200 Ab dilution finally), and after another 24 h incubation, goat anti-Armenian hamster IgG (catalog no. 127-165-160; Dianova, Hamburg, Germany) Ab was added (1:200 Ab dilution finally) to cross-link the anti-CD69 Abs. In some experiments, CD4-enriched spleen cells (from B6, CD69^−/−, or IFNAR^−/− mice) were cultured at 37°C, 5% CO₂ for 48 h in 500 μl medium containing RPMI, 10% FBS, and 1% Penicillin/Streptomycin with or without addition of 200 μg poly (I:C) per well. In one experiment, 1 × 10⁶ enriched B6 CD4⁺ T cells from spleen were mixed with 9 × 10⁶ CD4⁻CD69^−/− spleen cells or 1 × 10⁶ CD4⁺ T cells from spleen of CD69^−/− mixed with 9 × 10⁶ CD4⁻ cells from spleen of B6. These mixed cells were than cultured for 48 h in 500 μl medium containing RPMI, 10% FBS, and 1% Penicillin/Streptomycin with or without addition of 200 μg poly (I:C) per well. Assays for all reactions were performed in five repeats.

Naive CD4⁺CD25⁻ T cells were enriched via MACS from a total cell population of B6 or CD69^−/− spleen cells. These cells were cocultured with CD11c⁺ dendritic cells (DC) from the spleen of B6 mice (enriched via CD11c MicroBeads, catalog no. 130-052-001; Miltenyi Biotec). T cells and DC were mixed in the ratio 2:1 (8 × 10⁴ T cells:4 × 10⁴ DC) and cultured together with T cell-activating anti-CD3/anti-CD28 dynabeads (T cell/bead ratio 10:1) in 200 μl final/well medium RPMI, 10% FBS, and 1% penicillin/streptomycin. To simulate conditions that favor Foxp3⁺ Treg cell induction, 2000 U/ml IL-2 (catalog no. 402-ML; R&D Systems) and 5 ng/ml TGF-β1 (catalog no. 240-B; R&D Systems) were added to the cell culture. After 5 d culture, cells were washed with PBS supplemented with 0.3% BSA and 0.1% sodium azide, stained for surface CD4 with allophycocyanin-conjugated mAb binding CD4 GK1.5 (catalog no. 17-0041-83; eBioscience), intracellular for Foxp3 with FITC-conjugated anti-Foxp3 Ab FJK-16s (catalog no. 11-5773-82; eBioscience), and analyzed with FCM. The control Foxp3 expression in both populations of naive CD4 T cells was measured by FCM before the culture.

Microarray analyses were performed using 200 ng total RNA as starting material and 5.5 μg ssDNA per hybridization (GeneChip Fluidics Station 450; Affymetrix, Santa Clara, CA). The total RNAs were amplified and labeled following the Whole Transcript Sense Target Labeling Assay (http://www.affymetrix.com). Labeled ssDNA was hybridized to Mouse Gene 1.0 ST Affymetrix GeneChip arrays (Affymetrix). The chips were scanned with an Affymetrix GeneChip Scanner 3000 and subsequent images analyzed using Affymetrix Expression Console Software (Affymetrix). Gene expression was than analyzed using the Affymetrix GeneChip Mouse Gene 1.0 ST Array platform (Affymetrix). Gene expression microarray files (Affymetrix .CEL files) were generated using the GCOS software (Affymetrix). Statistical analyses were carried out using R (v. 2.12.1; R-Development-Core-Team). Arrays have been normalized using robust multiple-array average (38). Expression data were analyzed using Bioconductor oligo package for R (39). Differentially expressed genes were determined by the shrinkage T-statistic (40). Functional enrichment analysis (gene ontology category enrichment and Kyoto Encyclopedia of Genes and Genomes Pathway enrichment) was determined using the hypergeometric test statistic with all genes on the Affymetrix GeneChip Mouse Gene 1.0 ST Array (Affymetrix) as a reference set. Multiple comparison results were controlled by maintaining a false discovery rate (FDR) <0.05 (41).

RNA was prepared from frozen colon tissue or pelleted CD4 T cells using the RNAeasy mini kit (catalog no. 74904; Qiagen, Hilden, Germany). Contaminating genomic DNA was eliminated from samples by treatment with RNAse-free DNAse I (catalog no. 1010395; Qiagen). A total of 2 μg RNA isolated from tissue or 200 ng RNA isolated from CD4 T cells was reverse transcribed with SuperScript II Reverse Transcriptase (catalog no. 18064-014; Invitrogen) using random primers (catalog no. 48190-011; Invitrogen) according to the manufacturer’s instructions. SYBR Green qPCR Master mix (catalog no. PA-012-12; SABiosciences, Qiagen) was used for amplification and detection. Real-time PCR reactions were performed using the 7500 Fast Real-Time PCR System (Applied Biosystems, Darmstadt, Germany) and the following conditions: 50°C for 2 min, repeat 1; 95°C for 10 min, repeat 1; 95°C for 15 s, 60°C for 1 min, repeats 40; 95°C for 15 s, 60°C for 1 min, 95°C for 15 s, and 60°C for 15 s, repeat 1. β-actin PCR signals were used to equalize cDNA amounts between preparations. The following primers were used: β-actin (catalog no. PPM02945A; SABiosciences, Qiagen); IFN-γ (catalog no. PPM03121A; SABiosciences); IL-21 (catalog no. PPM03761E; SABiosciences); IFN-β1 (catalog no. PPM03594B; SABiosciences); TLR3 (catalog no. PPM04216B; SABiosciences); TGF-β1 (forward) 5′-GTA CAG CAA GGT CCT TGC CCT-3′; and TGF-β1 (reverse) 5′-TAG TAG ACG ATG GGC AGT GGC-3′ (Thermo Scientific). Expected product length was: for β-actin, 154 bp; for IFN-γ, 95 bp; for IL-21, 189 bp; for IFN-β1, 52 bp; for TLR3, 184 bp; and for TGF-β1, 102 bp. In some experiments, RT-PCR products were visualized by agarose gel electrophoresis. A 10-cm length 2% agarose (catalog no. A9539-500G; Sigma-Aldrich) gel was used. Electrophoresis was performed in 1× TAE buffer, 8 V/cm, 45 min. Gene Ruler Plus (catalog no. SM1332; Fermentas, St. Leon-Rot, Germany) was used to identify the band length. Samples were visualized by adding ethidium bromide 0.07% solution (catalog no. A2273, 0015; AppliChem, Darmstad, Germany) in gel and using the GeneGenius System (Syngene, Frankfurt, Germany) after electrophoresis. Images were obtained by GeneSnap image acquisition software.

Cytokines in supernatants and blood serums were detected by a conventional double-sandwich ELISA. The following mAbs (from BD Biosciences) were used for detection and capture: mAb R4-6A2 (catalog no. 551216) and biotinylated mAb XMG1.2 (catalog no. 554410) for IFN-γ, mAb TC11-18H10 (catalog no. 555068), and biotinylated mAb TC11-8H4.1 (catalog no. 555067) for IL-17A and mAb TN3-19.12 (catalog no. 557516) and biotinylated mAb C1150-14 (catalog no. 557432) for TNF-α. TGF-β1 concentration was measured using the human/mouse TGF-β1 ELISA Ready-SET-Go Kit (catalog no. 88-7344; eBioscience). Extinction was measured at 405/490 nm on a TECAN microplate-ELISA reader using EasyWin software (both from Tecan, Wetzlar, Germany).

A one-way ANOVA test (for nonparametric data) and a t test for two unequal variances were used. A p value <0.05 was considered statistically significant.

We determined the fraction of CD4 T cells in spleen, MLN, and intestinal LP of B6 mice that expressed on the surface the activation marker CD69. Approximately 10% of the CD4 T cells from the spleen, 17% of CD4 T cells from the MLN, and 50% of CD4 T cells from the intestinal LP expressed CD69 in B6 mice were raised and kept under SPF conditions (Fig. 1A). The fraction of CD69⁺ cells is hence increased substantially in the intestinal LP CD4 T cell population. This was observed in SPF but not syngeneic GF B6 mice, in which the fraction of CD69⁺ cells was reduced, especially in the intestinal LP CD4 T cell population (Fig. 1B). The fraction of CD69⁺ CD4 T cells in spleen, MLN, and particularly the intestinal LP was also reduced in SPF OT-II mice, although a substantial fraction of the transgenic TCR-expressing CD4 T cells did express CD69 in these mice (Fig. 1A), confirming a previous report (42). Compared to SPF B6 mice, the CD69⁺ CD4 T cell fraction in OT-II × RAG^−/− mice was reduced in spleen, MLN, and siLP but not in cLP (i.e., a 10-fold reduction in CD69⁺ CD4 T cell population was seen in spleen, MLN, and siLP) (Fig. 1A). The treatment of OT-II and OT-II × RAG^−/− animals with an antibiotic mixture led to the reduction of CD69⁺ CD4 T cell numbers as compared with SPF OT-II and OT-II × RAG^−/− animals (Fig. 1B). These data indicate that the endogenous microflora have a major impact on the CD69 surface expression by CD4 T cells in all tissues tested.

We fed 1 mg OVA (in 100 μl PBS) to SPF B6 and (age- and sex-matched) OT-II × RAG^−/− mice twice (on consecutive days) by oral gavage. Control animals were gavaged with PBS only. Tested at 24 h after the feeding, SPF B6 mice challenged with OVA showed no change in CD69 and CD25 surface expression on CD4 T cells from the intestinal LP, spleen, and MLN (Fig. 2A). In contrast, oral challenge of SPF OT-II × RAG^−/− mice with OVA (using the same protocol) induced CD69 surface expression by CD4 T cells from the siLP, spleen, and MLN, but not cLP, where the CD4 T cells were expressing CD69 even before the OVA gavage (Fig. 2B). Following oral OVA challenge, 40–50% of all CD4 T cells from the intestinal LP and spleen of challenged mice expressed CD69 on the surface (Fig. 2B). Convincing induction of CD69 expression by CD4 T cells was seen in the MLN of OVA-gavaged OT-II × RAG^−/− animals, in which a clear population of ∼20% of all CD4 T cells was detected to be CD69⁺ (Fig. 2B). Interestingly, these specifically activated CD4 T cells did not express the activation marker CD25 that renders them responsive to IL-2 and allows their clonal expansion and survival after specific challenge (Fig. 2B). Hence, CD4 T cells respond to specific priming by an oral Ag challenge by expressing a CD69⁺CD25⁻ surface phenotype.

We compared the phenotype of CD69⁺ versus CD69⁻ CD4 T cells from the siLP of OT-II × RAG^−/− mice recently challenged orally by OVA in intracellular and surface FCM analyses. As expected, most intestinal LP CD69⁻ CD4 T cells are naive and do not express the master regulators T-bet, GATA-3, or RORγt [required for Th1, Th2, or Th17 differentiation, respectively (43–45)]. Almost 20% of the intestinal LP CD69⁻ CD4 T cells were Foxp3⁺ Treg cells. A fraction (of ∼20%) of the intestinal LP CD69⁻ CD4 T cells expressed the IFN-γR (binding IFN-γ), CD122 (binding IL-2/IL-15), TGF-βRII (binding TGF-β), the TGF-β–regulated integrin CD103, and IL-21R1. A fraction (of ∼11%) expressed the LAP associated with TGF-β1. The cytokine receptor IL-10R and the costimulatory molecule ICOS were not expressed by LP CD69⁻ CD4 T cells. In contrast, none of the transcription factors—T-bet, GATA-3, or RORγt—or the cytokine receptors—IFN-γR, CD122, TGF-βRII, IL-21R1, IL-10R, the integrin CD103, or the costimulatory molecule ICOS—were detected in activated intestinal LP CD69⁺ CD4 T cells (Fig. 2C). A fraction (∼40%) of the activated intestinal LP CD69⁺ CD4 T cells expressed LAP/TGF-β1. Hence, CD69⁺ CD25/CD122⁻ CD4 T cells recently activated in the intestinal LP by specific Ag challenge are not committed to the development of a proinflammatory phenotype and are largely unresponsive to key cytokines. Activated CD69⁺CD25/CD122⁻ LP CD4 T cells expressed LAP/TGF-β1, indicating that these cells could serve as regulatory cells.

Having demonstrated the high CD69 expression by CD4 T cells at mucosal surfaces, we analyzed by microarray experiment the CD69-dependent gene expression by comparing CD4 T cells from CD69-deficient animals, CD69-activated CD4 T cells and CD4 T cells from B6 animals. When CD4 T cells from B6 animals were compared with those from CD69^−/− animals, 2472 genes were differentially regulated. The comparison of B6 CD4 T cells with CD69-activated CD4 T cells yielded 2641 differentially regulated genes. The comparison of CD4 T cells from CD69^−/− animals with CD69-activated CD4 T cells yielded 3259 differentially regulated genes. IL-21, TNF-α, and IFN-γ were downregulated, and the TGF-β3 was upregulated in CD69-activated CD4 T cells (Table I). The microarray data sets have been uploaded in the Gene Expression Omnibus (accession number GSE27706).

To map the CD69-dependent pathways, we checked for enrichment of our microarray data sets in the gene ontology category and Kyoto Encyclopedia of Genes and Genomes pathways. These pathways included cytokine–cytokine receptor interaction, chemokine signaling pathway, TLR signaling pathway, nucleotide-binding oligomerization domain-like receptor signaling pathway, and the TGF-β signaling pathway (p value from hypergeometric test [rawP] was 7.24e-05, and p value adjusted by the multiple test adjustment [adjP] was 0.0004, respectively) (data not shown). Spleen CD4 T cells from B6 or CD69^−/− mice were then isolated and stimulated with anti-CD3/anti-CD28 Ab-coated beads in the presence or absence of a cross-linked anti-CD69 Ab. CD69 ligation enhanced TGF-β1 transcript levels expressed by B6 CD4 T cells (measured by quantitative RT-PCR [qRT-PCR]) but decreased IFN-γ and IL-21 transcript levels (Fig. 3A–C). Activation of CD69 also increased the amount of TGF-β1 released into supernatants by stimulated B6 CD4 T cells but decreased IFN-γ secretion (Fig. 3D, 3E). No differences in IL-2 and IL-17A transcript levels or secreted protein by CD4 T cells with ligated or unligated CD69 were detectable (data not shown).

Furthermore, (CD3/CD28 ligation) stimulated cells from the spleen, siLP, and cLP of CD69^−/− B6 mice produced significantly increased IFN-γ and TNF-α but decreased TGF-β1 levels when compared with cells from (age- and sex-matched) B6 mice (Fig. 3F–H). IL-21 transcript level was increased in the small intestine and colon tissue of nontreated CD69^−/− mice compared with B6 mice (Fig. 3I). CD69 deficiency hence results in the reduction of induced TGF-β1 expression by CD4 T cells that is associated with enhanced IFN-γ, TNF-α, and IL-21 production.

Because CD69 influences the secretion of cytokines, we then analyzed the numbers of Foxp3⁺ CD4 Treg cells in CD69^−/− animals. The fraction of Foxp3⁺ CD4 Treg cells was reduced in the MLN, siLP, and cLP CD4 T cell population from nontreated CD69^−/− when compared with B6 mice (Fig. 4A). In the absence of TCR signaling in nontreated OT-II animals, reduced LAP/TGF-β1 expression was observed in isolates from the siLP of OT-II × CD69^−/− mice as compared with OT-II mice (Supplemental Fig. 1). Feeding with OVA protein induced significantly higher fraction and cell number of Foxp3⁺ Treg cells among spleen, MLN, siLP, and cLP CD4 T cells of OT-II mice compared with OT-II × CD69^−/− mice (Fig. 4B, 4C). This effect could be due to reduced TGF-β1 level in the absence of CD69, and we tested if exogenous addition of TGF-β1 can restore the ability of CD69^−/− CD4 T cells to become Foxp3⁺ Treg. Naive CD4⁺CD25⁻ B6 or CD69^−/− cells were cultured in Treg-polarizing conditions, with B6 CD11c⁺ DC and exogenous IL-2 and TGF-β1 for 5 d. Although a significant fraction of CD69^−/− CD4 T cells induced Foxp3 expression (∼45%), this was still significantly reduced as compared with the fraction of B6 CD4 T cells expressing Foxp3 (∼55%) (Fig. 4D, 4E). These data are showing that other factors that reduced TGF-β1 production are influencing reduced Foxp3⁺ Treg cell induction in the absence of CD69.

Because CD69 deficiency in transgenic animals is associated with reduced numbers of Foxp3⁺ Treg cells and an increased secretion of IFN-γ and TNF-α by CD69^−/− CD4 T cells, CD69^−/− animals were monitored for >6 mo for the occurrence of clinical and histopathological signs of a spontaneous colitis. However, clinical and histopathological signs for the development of a spontaneous colitis could not be observed in our animal facility (data not shown). We then transferred B6 or CD69^−/− CD45RB^high CD4 T cells into RAG^−/− hosts to test if CD69 has an influence on the severity of CD4 T cell-induced transfer colitis. We monitored the clinical course, percent loss of body weight, and histological severity of the progressive and lethal colitis. Transfer of CD45RB^high CD4 T cells from CD69^−/− donors induced a colitis that was more severe than the colitis induced by transfer of an equal number of CD45RB^high CD4 T cells from B6 donors. This was evident by accelerated body weight loss (Fig. 5A), more severe histopathology (Fig. 5B–E), and early lethality. The transfer of CD69^−/− CD45RB^high CD4 T cells into RAG^−/− hosts induced a higher rise in IL-17A, IFN-γ, and TNF-α serum levels when compared with transfers of B6 CD45RB^high CD4 T cells (Fig. 5F–H). These data point to a critical role of CD69 in regulating mucosal immune responses in inflammatory conditions.

IFN-Is induce and upregulate CD69 surface expression in many cell types (16), which is confirmed by our data. Injection of 20 μg of the IFN-I inducer poly (I:C) induced CD69 surface expression by CD4 T cells in B6 and OT-II × RAG^−/− animals. After injection of 200 μg poly (I:C), most CD4 T cells (from the spleen, MLN, and intestinal LP) expressed CD69 in B6 or OT-II × RAG^−/− mice but not in (IFN-I–unresponsive) IFNAR^−/− mice within 24 h (Fig. 6). These data indicated that poly (I:C) is a strong inducer of CD69 on CD4 T cells in an IFN-I–dependent manner.

To examine the role of CD69 and IFN-I signals in mucosal CD4 T cell responses, the protocol for the induction of oral tolerance was conducted in B6, OT-II × RAG^−/−, CD69^−/−, and IFNAR^−/− mice. Animals were fed with OVA, rested, and immunized with OVA emulsified in IFA containing CpG⁺ ODN. Splenocytes harvested 7 d postimmunization were restimulated in vitro for 72 h with titrated concentrations of the antigenic A^b-binding OVA_323–339 peptide, and the concentration of specifically induced IFN-γ was measured in supernatants by ELISA. OVA feeding prior to immunization significantly reduced IFN-γ production by spleen cells from B6 and OT-II × RAG^−/− mice, indicating establishment of oral tolerance (Fig. 7A). Checking of delayed-type hypersensitivity (DTH) response in B6 and OT-II × RAG^−/− mice confirmed establishment of tolerance in OVA-fed mice, as these animals showed significantly reduced response to ear challenge with OVA (conducted 7 d after immunization and measured as ear swelling) compared with those fed with PBS (Fig. 7B). These data confirmed a previous report that oral tolerance can be induced in TCR-transgenic animals (46).

We used the same protocol to test if oral tolerance to OVA can be induced in CD69^−/− and IFNAR^−/− B6 mice. The specific IFN-γ response of spleen cells from CD69^−/− and IFNAR^−/− B6 mice fed with OVA, rested, and immunized was not impaired compared to the animals that were PBS fed, rested, and immunized (Fig. 7A). Ear challenge with OVA 7 d after immunization revealed that DTH response in OVA-fed mice of these two mice strains was not altered compared to PBS-fed animals (Fig. 7B). Oral tolerance can hence not be induced either in CD69^−/− or IFNAR^−/− B6 mice, pointing to a critical role of CD69- and IFN-I–mediated signaling in the induction and/or maintenance of oral tolerance.

It is reported that poly (I:C) injection can diminish dextran sulfate sodium colitis (47). We tested if poly (I:C) injections attenuate transfer colitis. The transfer of CD45RB^high CD4 T cells in RAG^−/− hosts results in progressive body weight loss and histopathological signs of colitis (Fig. 8A, 8D). Poly (I:C) injections diminish the severity of transfer colitis in RAG^−/− hosts transplanted with B6 CD45RB^high CD4 T cells as indicated by significant reduced body weight loss, histopathological signs of colitis, and IFN-γ transcript levels in colon (Fig. 8A, 8D, 8E). Compared to nontreated RAG^−/− hosts transplanted with CD69^−/−CD45RB^high CD4 T cells, body weight loss, the severity of histopathological signs of colitis, as well as IFN-γ transcript levels in colon were not reduced in RAG^−/− hosts transplanted with CD69^−/− CD4 CD45RB^high T cells and treated with poly (I:C) (Fig. 8B, 8D, 8E). Body weight loss, histopathological signs of colitis, and IFN-γ transcript levels in colon between poly (I:C)-treated and nontreated RAG^−/− hosts transplanted with IFNAR^−/− CD4 CD45RB^high T cells were not significantly different (Fig. 8C–E). Hence, poly (I:C)-derived protective signals to CD4 T cells are depending on CD69 and IFN-I.

In RAG^−/− animals reconstituted with B6, CD69^−/−, or IFNAR^−/− CD4 T cells, the observed protective effects on colitis development could be a direct effect of poly (I:C) on CD4 T cells. Alternatively, poly (I:C) acts on non-CD4 T cells (i.e., myeloid cells) in reconstituted RAG^−/− animals that elicits protective effects on CD4 T cells in animals with transfer colitis. We first investigated TLR3 expression in B6, CD69^−/−, and IFNAR^−/− cells by qRT-PCR. A significant difference in TLR3 expression among B6, CD69^−/−, and IFNAR^−/− animals was not observed (Fig. 9A, 9B). Myeloid cells (plasmacytoid DC, conventional DC, and macrophages) are the main producers of IFN-I after poly (I:C) challenge (48). We analyzed IFN-β1 expression in lymphocytes from B6, CD69^−/−, and IFNAR^−/− mice stimulated in vitro with poly (I:C). IFN-β1 expression after poly (I:C) stimulation was significantly increased in B6 and IFNAR^−/− cells, but not in CD69-deficient cells (Fig. 9B). We then carried out cocultures with B6 CD4 T cells and CD69^−/− myeloid cells or with CD69^−/− CD4 T cells and B6 myeloid cells. When CD69^−/− CD4 T cells were cultured with myeloid cells from B6 animals, reduced IFN- β1 responses were observed (Fig. 9C). Interactions of CD69 with a yet-unknown ligand (potentially) expressed by myeloid cells may be involved in the observed effects.

The data presented in our study indicated that the activation Ag CD69 plays a role in regulating mucosal immune responses in the intestine. CD69 expression is upregulated on intestinal CD4 T cells by the commensal microflora, recognition of specific Ag, or innate stimuli [i.e., poly (I:C) injections]. The upregulation of CD69 surface expression by the commensal microflora may hence require innate and/or TCR-specific stimuli. CD69⁺ CD4 T cells are characterized by LAP/TGF-β1 expression. We confirmed that CD69 activation is associated with an increased TGF-β1 expression and that the lack of this molecule in CD69^−/− animals is associated with CD4 T cells producing proinflammatory cytokines and reduced TGF-β1 (18). The CD4 T cells from spleen, siLP, and cLP showed differential expression of IFN-γ, TNF-α, and TGF-β1 (Fig. 3F–H), but the highest CD69 surface expression was determined on CD4 T cells isolated from intestinal tissues (Fig. 1A), indicating the most important role of this molecule in the mucosal immune system of the gut. Furthermore, CD69 affected the peripheral Foxp3⁺ Treg cell pool, as demonstrated in vivo both in steady-state conditions and after oral Ag challenge of OT-II × CD69^−/− animals. TGF-β1 is known to promote Foxp3⁺ Treg cell generation (49), and we speculated that reduction in Foxp3⁺ Treg cell fraction in CD69^−/− mice was in part due to lack of CD69-induced TGF-β1 production in these animals. Still, exogenous addition of TGF-β1 could not restore the normal potential of naive CD69^−/− CD4 T cells to become Foxp3⁺ Treg cells in vitro. We speculate that CD69 intracellular signaling may interfere with Foxp3 expression (50) or that other cytokines, such as IL-10, are involved in our observed findings (51). Although CD69^−/− animals did not develop a spontaneous colitis in our animal facility, the reconstitution of RAG^−/− hosts with CD69^−/−CD45RB^high CD4 T cells was associated with the accelerated transfer colitis as compared with RAG^−/− hosts transplanted with CD45RB^high CD4 T cells from syngeneic B6 donor mice. In addition, CD69^−/− animals are impaired in fully establishing oral tolerance. The increased expression of IFN-γ and TNF-α, reduced expression of TGF-β1, and reduction in the Foxp3⁺ Treg cell fraction in CD69^−/− mice could contribute to the accelerated transfer colitis development after reconstitution of RAG^−/− hosts with CD69^−/− CD4 T cells. Besides colonization of the host with the commensal flora, CD69 surface expression can be induced by the injection of the IFN-I inducer poly (I:C) (16). Our data confirmed that this effect of poly (I:C) is IFN-I dependent, because IFNAR^−/− animals did not show CD69 upregulation after poly (I:C) injection. However, IFNAR^−/− mice had a high fraction of CD69-expressing CD4 T cells even before treatment (CD69 expression in intestine of IFNAR^−/− mice was comparable with B6 mice), suggesting that stimuli other than IFN-I are inducing CD69 expression in the steady-state conditions. In contrast, CD69^−/− CD4 T cells showed an aberrant IFN-I response after poly (I:C) stimuli. Both CD69- and IFNAR-derived signals are necessary for the establishment of oral tolerance. IFN-I has been reported to have protective effects in colitis models and patients with ulcerative colitis (29–33). The ability of IFN-I to stimulate the production of IL-10 by human T cells and monocytes (52), modulate Th1 and Th2 responses (53, 54), and inhibit the production of IL-13, a key cytokine in the development of ulcerative colitis (30), has been suggested to contribute to the protective effects of IFN-I in colitis models (33). The treatment of dextran sulfate sodium colitis with s.c. injections of poly (I:C) (20 μg/mouse) attenuates the colitis in an IFN-I–dependent manner (47). The protective effects of poly (I:C) on colitis models are dose dependent because injections of high amounts of poly (I:C) can induce mucosal damage (55). IFN-I responses after poly (I:C) injection are modulated by CD69. Poly (I:C) injections did not affect the course of transfer colitis after transfer of CD69^−/− or IFNAR^−/−CD45RB^high CD4 T cells. IFNAR^−/− CD4 T cells are unable to respond to IFN-I, showing that poly (I:C) protects from transfer colitis in an IFN-I–dependent manner.

Because CD69 suppresses spingosine 1-phosphate receptor 1 expression and is hence involved in the regulation of lymphocyte egress (16), the expression of CD69 by T cells may be involved in the arrest of T cells at mucosal sites. In parallel, our data indicated that CD69 is involved in the regulation of IL-21, IFN-γ, TNF-α, and TGF-β1 expression by CD4 T cells. CD69 may hence regulate T cell migration and production of IL-21, IFN-γ, TNF-α, and TGF-β1 by CD4 T cells. The upregulation of the activation Ag CD69 by the commensal flora is not only an indicator of lymphocyte activation but also has significant functional relevance. CD69 may help to control the potential harmful impact of the intestinal microflora to the host by controlling lymphocyte migration and regulating the expression of proinflammatory cytokines and TGF-β1 in CD4 T cells, but further investigations are needed to address the role of CD69 in detail. The activation of CD69 by specific ligands could be a novel option for the treatment of intestinal inflammation.

We thank S. Schmidt for expert gnotobiotic animal breeding, G. Adler and G. von Wichert for continuous support, and Julia Geitner and Nathalie Birth for excellent technical work.

The authors have no financial conflicts of interest.