Induction of antigen-specific tolerance by oral administration of Lactococcus lactis delivered immunodominant DQ8-restricted gliadin peptide in sensitized nonobese diabetic Abo Dq8 transgenic mice

Abstract

Active delivery of recombinant autoantigens or allergens at the intestinal mucosa by genetically modified Lactococcus lactis (LL) provides a novel therapeutic approach for the induction of tolerance. Celiac disease is associated with either HLA-DQ2- or HLA-DQ8-restricted responses to specific antigenic epitopes of gliadin, and may be treated by induction of Ag-specific tolerance. We investigated whether oral administration of LL-delivered DQ8-specific gliadin epitope induces Ag-specific tolerance. LL was engineered to secrete a deamidated DQ8 gliadin epitope (LL-eDQ8d) and the induction of Ag-specific tolerance was studied in NOD AB degrees DQ8 transgenic mice. Tolerance was assessed by delayed-type hypersensitivity reaction, cytokine measurements, eDQ8d-specific proliferation, and regulatory T cell analysis. Oral administration of LL-eDQ8d induced suppression of local and systemic DQ8-restricted T cell responses in NOD AB degrees DQ8 transgenic mice. Treatment resulted in an Ag-specific decrease of the proliferative capacity of inguinal lymph node (ILN) cells and lamina propria cells. Production of IL-10 and TGF-beta and a significant induction of Foxp3(+) regulatory T cells were associated with the eDQ8d-specific suppression induced by LL-eDQ8d. These data provide support for the development of effective therapeutic approaches for gluten-sensitive disorders using orally administered Ag-secreting LL. Such treatments may be effective even in the setting of established hypersensitivity.

FIGURE 1

Lactococcus lactis secreting eDQ8d (LL-eDQ8d). A PCR amplified eDQ8d cDNA fragment was ligated in the L. lactis specific pT1NX vector (pT1eDQ8d), and subsequently transformed in MG1363 strains producing designated Lactococcus lactis secreting eDQ8d (LL-eDQ8d).

FIGURE 2

L. lactis -derived eDQ8d exhibits bioactivity as it induces proliferation of human HLA DQ8 T cell clones with no e-tag interference. (A) Human DQ8 T cells were derived as described in the material and methods. LL-eDQ8d was grown overnight and diluted 1:50, 4 and 6 hours thereafter supernatant was collected. Cells were stimulated with 1, 3 or 10 μl supernatant of a LL-eDQ8d culture. (B) For detection purposes an e-tag was attached. To exclude any possible interference of the e-tag, NOD AB° DQ8 transgenic mice were immunized by subcutaneous injection of 100 μg DQ8d with or without e-tag in CFA at day 1. At day 7, mice baseline ear-thickness was measured and mice were challenged with 10 μg DQ8d with or without e-tag, corresponding to the peptide used for the immunization, in 10 μl saline in the auricle of the ear. DTH responses were expressed as the difference in ear-thickness 24 h after the peptide injection minus the ear-thickness before injection. Data represent mean (± SEM) increase in ear thickness of 1 experiment including 6 mice per group.

FIGURE 3

Mucosal delivery of eDQ8d epitopes by L. lactis significantly decreases the DQ8d-induced DTH response and proliferative capacity of bulk spleen and inguinal lymph node cells and lamina propria cells. NOD AB° DQ8 transgenic mice were immunized by s.c. injection of 100 μg eDQ8d in CFA at day 1. Mice were orally treated with LL-eDQ8d or LL-pT1NX at days 1–10. Control mice received BM9. At day 10, mice were challenged with 10 μg eDQ8d in 10 μl saline in the auricle of the ear. DTH responses are expressed as the mean (±SEM) increase in ear thickness from baseline, 24 hours after injection (A). After the DTH measurements, spleens, inguinal lymph nodes, and lamina propria cells of the BM9 (control), LL-pT1NX and LL-eDQ8d groups were isolated and ex vivo stimulated with 50 μg/ml eDQ8d peptide or 50 μg/ml irrelevant peptide (irrel) (white bars: 3b–3d). eDQ8d-specific proliferative response of bulk splenocytes (P=0.048) (B) and inguinal lymph node cells (P=0.002) (C) and lamina propria cells (P=0.044) (D) were studied by thymidine incorporation, expressed as the mean (±SEM) cpm. Cytokine measurements in the supernatant of splenocytes (3e-3f) and inguinal lymph node cells (3g) were performed 24 hours after ex vivo eDQ8d (black bars) or irrelevant peptide (irrel)(white bars) stimulation. Results represent the mean (±SEM) of cytokine secretion in pg/ml for at least two individual experiments including 6 mice in each group. ▼ indicates not detected (below detection threshold).

FIGURE 4

Decreased splenic eDQ8d-specific proliferation depends on IL-10 and TGF-β. Mice were fed BM9, LL-pT1NX and LL-eDQ8d and DTH measurements were performed as described above. We next investigated the functional importance of cytokines such as TGF-β, IL-10, TGF-β in combination with IL-10 and LAP on the eDQ8d-specific splenic proliferative response using neutralizing antibodies. To do so, bulk spleen cells of LL-eDQ8d treated mice were isolated and 5 × 10⁵ cells were stimulated ex vivo with 50 μg/ml eDQ8d with or without neutralizing antibodies. Proliferative responses are expressed as percentage of proliferation compared to the BM9 treated group. Results are representative of two individual experiments with 6 mice each per group.

FIGURE 5

LL-eDQ8d treatment significantly increases splenic and GALT Foxp3 expression. At day 11, spleens and gut associated lymph node tissue (GALT) of mice treated with BM9 (blue), LL-pT1NX (yellow) or LL-eDQ8d (pink) were isolated and stained for CD4, CD25 and intracellular Foxp3. Flow cytometry was performed on the splenic CD4+CD25+ (A), CD4+CD25− (B) and GALT CD4+CD25− (C) subpopulations. Data represent 1 experiment with 6 mice per group.