Resistance to Celiac Disease in Humanized HLA-DR3-DQ2-Transgenic Mice Expressing Specific Anti-Gliadin CD4+ T Cells

Andrea L. de Kauwe, Zhenjun Chen, Robert P. Anderson, Catherine L. Keech, Jason D. Price, Odilia Wijburg, David C. Jackson, Jodi Ladhams, Janette Allison and James McCluskey
J Immunol June 15, 2009, 182 (12) 7440-7450; DOI: https://doi.org/10.4049/jimmunol.0900233

Abstract

Celiac disease is a chronic inflammatory enteropathy caused by cellular immunity to dietary gluten. More than 90% of patients carry HLA-DQ2 encoded by HLA-DQA1*05 and DQB1*02, and gluten-specific CD4+ T cells from intestinal biopsies of these patients are HLA-DQ2-restricted, produce Th1 cytokines and preferentially recognize gluten peptides deamidated by tissue transglutaminase. We generated mice lacking murine MHC class II genes that are transgenic for human CD4 and the autoimmunity and celiac disease-associated HLA-DR3-DQ2 haplotype. Immunization with the α-gliadin 17-mer that incorporates the overlapping DQ2-α-I and DQ2-α-II epitopes immunodominant in human celiac disease generates peptide-specific HLA-DQ2-restricted CD4+ T cells. When exposed to dietary gluten, naive or gliadin-primed mice do not develop pathology. Coincident introduction of dietary gluten and intestinal inflammation resulted in low-penetrance enteropathy and tissue transglutaminase-specific IgA. Two further strains of transgenic mice expressing HLA-DR3-DQ2 and human CD4, one with a NOD background and another TCR transgenic having over 90% of CD4+ T cells specific for the DQ2-α-II epitope with a Th1 phenotype, were also healthy when consuming gluten. These humanized mouse models indicate that gluten ingestion can be tolerated without intestinal pathology even when HLA-DQ2-restricted CD4+ T cell immunity to gluten is established, thereby implicating additional factors in controlling the penetrance of celiac disease.

Celiac disease (CD)4 is a common chronic inflammatory disorder of the small intestine affecting ∼1% of the world’s population (1). It is precipitated by an inappropriate immune response against gluten proteins from ingested wheat, rye, barley, or sometimes oats. These dietary Ags are normally tolerated by the immune system, but in patients with CD the gluten-induced immunity characteristically results in villous atrophy (partial or total), crypt hyperplasia, and the accumulation of various leukocyte populations in both the lamina propria and epithelium (2). IgA against the autoantigen tissue transglutaminase (tTG) is frequently associated with untreated CD but disappears with gluten exclusion (3).

More than 90% of patients possess HLA-DQ2, encoded by DQA1*05 and DQB1*02 (4), while gluten-specific intestinal CD4+ T cells from HLA-DQ2+ patients are almost exclusively restricted by DQ2 (5). The epitopes recognized by patient-derived intestinal T cell clones and lines raised against gluten often correspond to gluten sequences selectively deamidated by tTG, rich in proline and glutamine, and which are relatively resistant to human intestinal proteases (6, 7, 8, 9, 10).

Despite these major advances and recent discovery of several non-HLA genetic associations in CD (11, 12), it remains unclear why >90% of HLA-DQ2+ individuals establish and maintain immune tolerance to gluten. The absence of an animal model of CD dependent on HLA-DQ2-restricted, gluten-specific CD4+ T cells has limited the opportunities to study early events in disease development.

Although only 5% of patients with CD possess HLA-DQ8 without having HLA-DQ2 (4), gluten immunopathology has only been reported in HLA-DQ8- transgenic (Tg) mice. Immunization of such animals elicits gluten-specific T cells with a regulatory profile (producing IL-6, TGF-β, and IL-10), with no autoantibody production and no enteropathy (13). Interestingly, breeding to the autoimmune-prone NOD background appears to render the HLA-DQ8 Tg mouse susceptible to gluten sensitivity since immunization with gluten and coadministration of pertussis toxin provoked a lesion similar to the cutaneous manifestation of CD, dermatitis herpetiformis, in 17% of the mice (14). But unlike dermatitis herpetiformis, neither enteropathy nor tTG-specific autoantibody production was detected in any of the animals.

Our laboratory has previously generated Tg mice that express the HLA-DR3-DQ2 haplotype (spanning HLA-DRA to DQB2) on a partial murine class II knockout background (Aβo/o) (15). These mice demonstrate cell-specific regulation of the HLA-DR3 and DQ2 gene products, with fully functional CD4+ and CD8+ T cells (15). In the present study, we have bred mice with the HLA-DR3-DQ2 transgene to a complete class II knockout background (MHCIIΔ/Δ) (16), introduced human CD4 (hCD4) (17), and also generated a HLA-DQ2-restricted, gliadin-specific CD4+ TCR Tg mouse to examine the pathological effect of dietary gluten on these mice.

Materials and Methods

Generation and use of hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice

Human CD4 Tg mice (hCD4 Tg; B6 × DBA/2 background) (17) and mice lacking the murine MHC class II locus (MHCIIΔ/Δ; B6 × 129 background) (16) were individually backcrossed to B6 for six to eight generations. MHCIIΔ/Δ mice were subsequently crossed with hCD4 Tg mice or HLA-DR3-DQ2 haplotype Tg mice (15), creating hCD4.MHCII+/− mice and DR3-DQ2.MHCII+/− mice. These double transgenic lines were mated to generate F1 mice expressing HLA-DR3, DQ2, and hCD4 in the complete absence of endogenous MHCII (hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice). The HLA-DR3-DQ2 haplotype was bred to homozygosity by intercrossing the hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice; all mice were hCD4+. Generation of hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice on a NOD background was undertaken in the same manner after each of the initial founder lines were backcrossed to NOD for eight generations.

Mouse housing and diet

Mice were housed under specific pathogen-free conditions and after weaning were maintained on gluten-free food pellets (Pig Research Institute, Victorian Institute of Animal Health, Werribee, Australia). Gluten-containing food was standard laboratory mouse chow (The Walter and Eliza Hall Institute breeder cubes; Barastoc, Ridley Agriproducts) that contained wheat. Upon introduction of dietary gluten, mouse cages were also supplemented thrice weekly with commercial white bread (∼12 g/mouse or ∼1 g of gluten/mouse). Female breeders were maintained on standard (gluten-containing) food. Unless otherwise specified, mice were used at 6–10 wk of age.

Abs and flow cytometry

mAbs specific for HLA class II β-chain (Rm5.112), HLA-DQ (SPV-L3), and TCR Vβ8.2 (F23.2) were prepared in our laboratory and conjugated with FITC or biotin. Fluorochrome-labeled Abs that recognize B220 (RA3-6B2), CD11c (HL3), F4/80 (BM8), mouse CD4 (mCD4; GK1.5), hCD4 (SK3), CD8α (53-6.7), I-Ab (AF6-120.1), TCR Vβ2 (B20.6), TCR Vβ3 (KJ25), TCR Vβ4 (KT4), TCR Vβ5.1/5.2 (MR9-4), TCR Vβ6 (RR4-7), TCR Vβ7 (TR310), TCR Vβ8.1/8.2 (MR5-2), TCR Vβ8.3 (1B3.3), TCR Vβ9 (MR10-2), TCR Vβ10b (B21.5), TCR Vβ11 (RR3-15), TCR Vβ12 (MR11-1), TCR Vβ13 (MR12-3), TCR Vβ14 (14-2), TCR Vβ17a (KJ23), mouse CD3 (145-2C11), TCR Vα8.3 (KT50), TCR Vβ8 (F23.1), and CD25 (PC61) were purchased from BD Biosciences Pharmingen. Intracellular Foxp3 staining was performed using an allophycocyanin anti-mouse/rat Foxp3 staining set (clone FJK-16s; eBioscience) according to the manufacturer’s instructions. Stained cells were acquired using a FACSCalibur (BD Biosciences) and analyzed with FlowJo software (Tree Star).

Antigens

Synthetic peptides were purchased from Mimotopes. They were as follows: α-gliadin 17-mer A-gliadin57–73 (57QLQPFPQPQLPYPQPQS73), deamidated α-gliadin 17-mer A-gliadin57–73 Q65E (57QLQPFPQPELPYPQPQS73), EBNA2 (280TVFYNIPPMPL290 from EBV), TPO (632IDVWLGGLAENFLP645 from human thyroid peroxidase), and MT65kDa3–13 (3KTIAYDEEARR13 from the Mycobacterium tuberculosis 65-kDa heat shock protein). The combitope (QLQPFPQPELPYPQPFPQQPEQPFPQPEQPFPWQP), deamidated α-gliadin 33-mer (56LQLQPFPQPELPYPQPELPYPQPELPYPQPQPF88), deamidated gda09 (206SGEGSFQPSQE216 from α-gliadin), deamidated DQ2-γ-I (139PQQPQQSFPEQQRP152 from γ-gliadin), deamidated ω-gliadin peptide (QPEQPFPQPEQPFPWQP), lysine-substituted scan of A-gliadin57–73 Q65E (17 peptides in total), and six overlapping 15-mers containing Gly-Ser-Gly-flanked 9-mer sequences from A-gliadin57–73 Q65E (i.e., GSGLQPFPQPELGSG through to GSGQPELPYPQPGSG, underscores indicate the core epitope) were purchased from Research Genetics and Pepscan Presto. Mass spectroscopy and HPLC were used to confirm the authenticity and >70% purity of the peptide. The A-gliadin57–73 Q65E lipopeptide was synthesized as previously described (18). Gliadin (G-3375; Sigma-Aldrich) was digested with α-chymotrypsin (G-3142; Sigma-Aldrich) and deamidated with guinea pig liver transglutaminase (T-5398; Sigma-Aldrich) as previously described (19). Crude gluten (G-5004; Sigma-Aldrich) was used for oral gavage and immunization, where ∼20 mg of crude gluten was suspended in 100 μl of peanut oil for gavage and 100 μg was emulsified in 50 μl of CFA for immunization.

Gluten-sensitization protocols

To assess the effect of immunization, 6-wk-old hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice were immunized s.c. with 80 μg of A-gliadin57–73 Q65E (n = 7), 80 μg of chymotrypsin-digested gliadin treated with tTG (n = 6) or 20 nmol of A-gliadin57–73 lipopeptide administered intranasally (n = 6). Following immunization, the diet was switched to gluten-containing food. At 8 wk, mice were boosted with the same immunogen in IFA. At 30 wk of age, all animals were killed and intestinal tissue was examined. To assess the effect of prior administration of a proinflammatory mediator, hCD4.DR3-DQ2.MHCIIΔ/Δ Tg and C57BL/6 mice were injected i.p. with pertussis toxin (200 ng/mouse; Sigma-Aldrich) and immunized 7 days later with 80 μg of chymotrypsin-digested gliadin or OVA in the presence of 200 ng of pertussis toxin (emulsified in CFA, delivered i.p.). On day 10, gliadin-immunized Tg mice (n = 14), gliadin-immunized B6 mice (n = 19), and OVA-immunized Tg mice (n = 11) had their diet switched to gluten-containing food. A separate group of gliadin-immunized Tg mice (n = 13) were maintained on gluten-free food. Eighteen weeks later, all mice were killed and intestinal tissue was examined. Experiments with hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice on a NOD background and HH8.TCR Tg mice were performed in a similar manner to those described above, but with smaller group sizes when using TCR Tg animals. Where Salmonella typhimurium infections were used to induce intestinal inflammation, mice were infected 14 days after immunization as previously described (20); dietary gluten was introduced 5 days later and all animals were culled after 25 wk. Serum was collected from all experimental animals at regular intervals after exposure to dietary gluten, with the final sample taken at death.

In vitro T cell proliferation assays

C57BL/6 mice or hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice (n = 4) were immunized with 80 μg of synthetic peptide emulsified 1:1 (v/v) in CFA (Difco) administered s.c. into the tail base and hind footpad. Eight days postimmunization, spleen and draining lymph nodes were harvested and the T cell population was enriched by panning with anti-IgM plus anti-IgG (100 μg/ml; M8644-1MG, M8645-1MG; Sigma-Aldrich). Cells were incubated for 1 h at 4°C (1 × 107 cells/ml) and unbound cells were removed. T cell-enriched lymphocytes typically contained ≤5% B cells. Enriched T cell suspensions were cultured with syngeneic APCs (splenocytes) and medium alone, synthetic peptide or Con A as previously described (15). Proliferation was assayed by [3H]thymidine incorporation as described elsewhere (15). Results are expressed in cpm or as a stimulation index, with the mean of each triplicate plotted and error bars representing the SD.

T cell hybridomas

CD4+ T cells were isolated and enriched from the spleen and draining lymph nodes of hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice (n = 5) following immunization with 80 μg of A-gliadin57–73 Q65E peptide and boosting 2 wk later (same peptide in IFA, s.c.). These cells were used to generate BW5147:T cell hybridomas as previously described (21). Activation of T hybridomas was assayed by IL-2 bioassay as previously reported (22). The IL-2-dependent cell line CTLL-2 (ATCC TIB-214) was maintained with 20–100 U/ml recombinant human IL-2.

Generation of gliadin-specific TCR Tg mice

Genomic DNA encoding the TCR β-chain and cDNA encoding the TCR α-chain were prepared from HH8.2 hybridomas and cloned into the p3A9cβTCR (23) and pES4 (24) plasmids, respectively. Vector sequences were removed by restriction enzyme digestion and constructs were coinjected into fertilized hCD4.DR3-DQ2.MHCIIΔ/Δ oocytes. Oocytes were transferred into surrogate mothers (F1 of CBA × C57BL/6) and pups were screened by PCR with primers for the transgenes. Two TCR Tg founders were identified by PCR and confirmed by flow cytometry and two TCR Tg lines were established by backcrossing to hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice.

In vivo proliferation using adoptive transfer of CFSE-labeled cells

B cell-depleted splenocytes from unimmunized HH8.TCR Tg mice were labeled with CFSE (2.5 μM; Molecular Probes) for 10 min at 37°C. Cells were washed and adoptively transferred (i.v.) into hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice (0.5–1 × 107 cells/mouse). Recipient mice were subsequently either immunized with 40 μg of A-gliadin57–73 Q65E or TPO peptide in the hind footpad or gavaged daily for 3 days with A-gliadin57–73 Q65E (2 mg/day) or vehicle alone. Three days following adoptive transfer and immunization/gavage, cells from the draining and nondraining lymph nodes were stained with Abs specific for TCRVα8.3, TCRVβ8, and murine CD4 and analyzed by flow cytometry.

Cytokine detection

Cells from the spleen, mesenteric lymph nodes (MLNs), and Peyer’s patches of unimmunized C57BL/6 and HH8TCR Tg mice were stimulated with 2 μM A-gliadin57–73 peptide (native or modified Q65E), 2 μg/ml Con A, or medium alone. After 48 h, supernatants were collected and incubated with cytokine detection reagents from a BD Biosciences Cytometric Bead Array Mouse Th1/Th2 Cytokine Kit (no. 551287) according to the manufacturer’s instructions. Samples were examined by flow cytometry and data were analyzed using FCAP Array software version 1.0 (BD Biosciences) and FlowJo software.

ELISAs

Anti-gliadin IgA ELISA.

Crude gliadin (G-3375; Sigma-Aldrich) was dissolved in 8 M urea, 0.1 M NaH2PO4, and 10 mM Tris (pH 8.0) at 5 mg/ml and filtered. Nunc Maxisorp immunoplates were coated overnight with 1 μg/well dissolved gliadin in carbonate-bicarbonate coating buffer (pH 9.6) at 4°C. Plates were washed (five times) with PBST (PBS and 0.05% Tween 20); the last wash was incubated for 10 min to block plate. Mouse sera or human sera from celiac patients were diluted in PBST and added to plates in duplicate. Plates were washed (five times) before addition of alkaline phosphatase-labeled anti-mouse IgA (A-4937; Sigma-Aldrich) or anti-human IgA (A-3063; Sigma-Aldrich). Following washing, bound Ab was detected with 1 mg/ml p-nitrophenyl phosphate substrate (S0942; Sigma-Aldrich) in diethanolamine buffer (pH 9.8).

Anti-tTG IgA ELISA.

Plates were coated with 0.5 μg/well guinea pig liver tTG (T-5398; Sigma-Aldrich) in 50 mM Tris-HCl, 150 mM NaCl, and 5 mM CaCl2 (pH 7.5) for 2 h at 37°C. Other steps are identical to the anti-gliadin IgA ELISA, except that the wash buffer contained 50 mM Tris-HCl, 150 mM NaCl, 10 mM EDTA, and 0.1% Tween 20. All plates were read at OD405 nm and reactivity of mouse sera was determined using a positive cutoff value of ≥3 SD above the mean obtained from pretreatment sera.

Histology

The proximal 9–10 cm of mouse small intestine (duodenum and jejunum) was dissected into 3-cm segments, cut longitudinally, and fixed in 10% neutral-buffered formalin; distal intestine (ileum, 3 cm) was also collected from many animals. Paraffin-embedded sections (4 μm) were H&E stained and assessed by two histological parameters: number of intraepithelial lymphocytes (IELs) per 100 enterocytes and the ratio of villus height compared with crypt depth. Both parameters were measured from ≥5 well-orientated villi and the median was graphed. Assessment occurred in a blinded manner.

Statistical analysis

The Kruskal-Wallis test was used to compare the histological data between more than two treatment groups. The comparison of only two treatment groups was performed using the Mann-Whitney U test. Tests were conducted using the median value of the data set.

Results

HLA-DQ2 molecules are expressed and functional in Tg mice

Human class II expression was first confirmed on various splenic populations of leukocytes from hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice and B6 controls. DQ2 expression was demonstrated on B cells, macrophages, and dendritic cells from hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice, but not on the same cells of B6 mice (Fig. 1⇓A). Unlike activated human T cells, expression of DQ2 was not seen on either naive (Fig. 1⇓A) or activated (data not shown) T cells from Tg or non-Tg mice. Murine class II was not expressed on Tg leukocytes (data not shown).

FIGURE 1.
FIGURE 1.

HLA-DR3-DQ2 molecules are expressed and functional in Tg mice. A, Splenocytes from C57BL/6 or hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice were assessed by FACS for HLA-DQ2 expression by direct immunofluorescence using the mAb SPV-L3. Representative histograms show cells from hCD4.DR3-DQ2.MHCIIΔ/Δ Tg (shaded) and C57BL/6 (bold line), gated, respectively, on B220+, F4/80+, CD11c+, CD8+, or CD4+ cells. B, Dot plots of FACS staining for mouse and hCD4 on splenocytes from C57BL/6, DR3-DQ2.MHCIIΔ/Δ Tg, or hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice. All plots are gated on lymphocytes with the percentage of lymphocytes shown in the relevant quadrants. C and D, Proliferation of B cell-depleted LN cells from hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice immunized with 80 μg of A-gliadin57–73 Q65E. T cell responses recalled in vitro with (C) A-gliadin57–73 Q65 or Q65E or (D) 5 μM of control peptides as follows: DQ2-restricted aa 632–645 from thyroid peroxidase (TPO), DQ8-restricted deamidated aa 206–216 from α-gliadin (gda09), DQ2-restricted deamidated aa 139–152 from γ-gliadin (DQ2-γ-I), or Con A (2 μg/ml). DC, Dendritic cell.

To demonstrate functionality of hCD4, the proportion of peripheral CD4+ T cells was compared in DR3-DQ2.MHCIIΔ/Δ Tg mice with and without coexpression of hCD4 (hCD4.DR3-DQ2.MHCIIΔ/Δ Tg vs DR3-DQ2.MHCIIΔ/Δ Tg). The proportion of normal B6 splenic lymphocytes expressing mCD4 was 20% (Fig. 1⇑B). In contrast, in HLA class II Tg mice lacking hCD4 (DR3-DQ2.MHCIIΔ/Δ Tg), only 13% of lymphocytes expressed mCD4 (Fig. 1⇑B), similar to the numbers in Tg mice homozygous for the DR3-DQ2 haplotype on an Abo/o background (15). However, expression of the hCD4 transgene boosted CD4+ T cell numbers to 18% of splenic lymphocytes in the hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice (Fig. 1⇑B), indicating the importance of the specific interaction between human CD4 and the human class II molecules. Interestingly, TCR Vβ usage was not significantly different between Tg mice with or without hCD4 (data not shown).

To confirm that the HLA molecules and CD4+ T cells were functional on the MHCIIΔ/Δ and hCD4 Tg background, CD4+ T cell responses were examined in the hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice. Two well-characterized human T cell epitopes were used to immunize Tg and non-Tg animals (n ≥5 mice/epitope): MT65kDa3–13 is a M. tuberculosis heat shock protein-derived peptide that binds exclusively to HLA-DR3 (DRB1*0301) and is recognized by T cells from tuberculoid leprosy patients (25); Α-gliadin57–73 Q65E is a deamidated gluten peptide which includes two overlapping HLA-DQ2-restricted epitopes, DQ2-α-I (PFPQPELPY) and DQ2-α-II (PQPELPYPQ), that account for approximately one-half the gluten-specific T cells present in blood after in vivo wheat challenge in patients with CD (19, 26, 27). Eight days after immunization, T cells from draining lymph nodes were examined for their ability to proliferate in vitro in the presence of MT65kDa3–13, Α-gliadin57–73 Q65E, or a panel of irrelevant HLA-DR- or DQ-restricted peptide controls. A specific proliferative response to MT65kDa3–13 was observed in T cells isolated from hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice previously immunized with this peptide, but not by T cells from B6 mice similarly immunized (data not shown). Likewise, T cells from hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice responded specifically to in vitro recall with Α-gliadin57–73 Q65E in a dose-dependent manner (Fig. 1⇑C). The native version of the Α-gliadin57–73 peptide triggered a very weak T cell response at high concentrations of peptide (>5 μM; Fig. 1⇑C), but irrelevant HLA-DQ-binding peptides were nonstimulatory (Fig. 1⇑D). Together, the data indicate that both the human class II molecules and the CD4+ T cells support Ag-specific immunity in hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice.

Low penetrance of celiac pathology in gluten-immunized DR3-DQ2 Tg mice

Given that T cells from hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice are reactive to the immunodominant gluten peptide in HLA-DQ2+ patients with CD (19), we hypothesized that these animals might develop hypersensitivity to gluten. The hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice are healthy while consuming gluten: histology of the small and large intestine, body weight, and stools were normal; Abs against gliadin or tTG were not detectable. However, prior activation of T cells by immunizing the hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice with gluten, chymotrypsin-digested gliadin, or various preparations of Α-gliadin57–73 Q65E did not induce hypersensitivity to dietary gluten (data not shown).

Pertussis toxin has been used for many years as an adjuvant to promote the induction of organ-specific autoimmunity when codelivered with self-Ags (28). Therefore, non-Tg B6 and hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice were injected with pertussis toxin i.p. (day 0) and immunized 7 days later with chymotrypsin-digested gliadin or OVA i.p. in the presence of pertussis toxin and CFA. On day 10, one group of gliadin-immunized Tg mice (n = 14) had their diet switched to gluten-containing food, while a second group of gliadin-immunized Tg mice (n = 13) were maintained on gluten-free food. A group of OVA-immunized Tg mice (n = 11) and a group of gliadin-immunized non-Tg controls (n = 19) were also introduced to dietary gluten on day 10. Eighteen weeks after commencement of gluten-containing food, all mice were killed and intestinal samples were collected for histological analysis. Serum was collected at six time points throughout the course of the experiment and tested for anti-gliadin and anti-tTG Abs.

As shown in Fig. 2⇓A, 2 of 14 animals from the gliadin-immunized plus gluten-fed group developed tTG-specific IgA Abs; Fig. 2⇓A shows only data from day 61 but both animals also had tTG-specific IgA in their serum from day 38. End-point titer analysis revealed that one of these animals was producing high-titer anti-tTG IgA (1:1200), while the IgA autoantibodies for the second mouse titrated out at 1:300 (data not shown). High-titer anti-tTG IgA autoantibodies were not detected in any other animal (Fig. 2⇓A) at any time point (data not shown). Analysis of sera for the presence of anti-gliadin IgA demonstrated consistent anti-gliadin reactivity in only a single Tg mouse (from the gliadin-immunized/gluten-free group), but interestingly in 10 of 19 gliadin-immunized B6 controls (Fig. 2⇓B). Histological examination of small intestinal tissue found no evidence of intraepithelial lymphocytosis or architectural changes; however, all three groups of Tg mice demonstrated significantly higher IEL counts compared with non-Tg mice irrespective of exposure to dietary gluten (data not shown). In addition, the Tg mouse producing high-titer tTG-specific IgA autoantibodies had an elevated IEL count of 32 IELs/100 enterocytes compared with the determined upper limit of normal (20 IELs/100 enterocytes; data not shown). Mucosa from the jejunum of this mouse also demonstrated elongated crypts (suggestive of increased proliferative activity) and increased cellular content in the lamina propria, but villous structure remained intact (Fig. 2⇓C). A 3-cm piece of distal intestine (ileum) was also collected from many animals for examination and no signs of enteropathy were observed in these tissues (data not shown). Another Tg animal from the gliadin-immunized/gluten-free group also demonstrated an elevated IEL count, elongated crypts, and a higher cellular density in the lamina propria (data not shown); both of these Tg mice were clinically scored as having mild enteropathy. A second experiment conducted in a similar manner (but without priming of the system with pertussis toxin before immunization) found that coinjecting with pertussis toxin resulted in one of eight gliadin-immunized plus gluten-fed Tg mice developing high-titer tTG-specific IgA autoantibodies and mild enteropathy (data not shown).

FIGURE 2.
FIGURE 2.

Gluten exposure in a proinflammatory context triggers mild low-penetrance gluten-induced pathology in HLA-DR3-DQ2 Tg mice. C57BL/6 or hCD4.DR3-DQ2.MHCIIΔ/Δ mice were immunized in the presence of pertussis toxin and then transferred to a gluten-containing diet. A and B, Serum was collected before treatment and gluten challenge, and again at various intervals after commencement of dietary gluten, and tested for IgA reactivity to (A) tTG or (B) crude gliadin by ELISA. Day 61 serum is representative of all intervals after day 38. Reactivity was determined using a positive cutoff value of ≥3 SD above mean of pretreatment sera (dotted line). Sera from untreated patients with CD were used as a positive control for ELISA, obtaining OD of >0.9. C, Selection of H & E-stained small intestinal sections showing healthy tissue and mild enteropathy observed in a small proportion of gluten challenged mice; original magnification, ×100 (left) and ×200 (right).

To increase the incidence and severity of enteropathy in the hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice, the hCD4 transgene, the HLA-DR3-DQ2 haplotype, and the MHCIIΔ/Δ knockout were individually backcrossed to the autoimmune-prone NOD strain for eight generations. The resulting Tg NOD strains were subsequently intercrossed to produce congenic hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice on a NOD background. Nevertheless, this genetic background did not accentuate the gluten-induced phenotype in the hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice (data not shown). Similarly, infection with Salmonella typhimurium, a natural enteric bacterial pathogen for rodents, coincidentally with gluten exposure, did not enhance the incidence or severity of gluten hypersensitivity in the hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice. A small number of animals (∼10%) were found to be producing tTG-specific IgA and also had an elevated IEL count, but villous structure, body weight, and stools remained normal (data not shown).

Isolation of a murine CD4+ T hybridoma specific for the immunodominant CD-associated DQ2-α-II epitope

One obstacle to the development of gluten-induced disease in the hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice may be a low frequency of gluten-specific T cells. We generated gliadin-specific CD4+ T cell hybridomas from hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice: animals were immunized with Α-gliadin57–73 Q65E and peptide-specific CD4+ T cell clones were fused to a TCR-negative murine thymoma cell line, BW5147, which does not express CD4 or CD8. A proportion of the actively growing hybridomas were found to be mCD4+, hCD4+, and TCR+ (CD3+); Fig. 3⇓A shows representative flow cytometric data of one such fusion (DB4). Homogeneous cell populations that demonstrated the highest expression of CD3 and CD4 were tested for gliadin specificity by stimulation with irradiated hCD4.DR3-DQ2.MHCIIΔ/Δ Tg splenocytes and Α-gliadin57–73 Q65E or an irrelevant HLA-DQ2-binding peptide (29). A number of hybridoma lines were found to respond specifically to Α-gliadin57–73 Q65E, producing IL-2 in a dose-dependent and HLA-DQ-dependent manner (Fig. 3⇓, B and C, shows the response of a representative hybridoma line).

FIGURE 3.
FIGURE 3.

Isolation and characterization of a murine HLA-DQ2-restricted, gliadin-specific CD4+ T cell hybridoma. Following fusion of immune T cells, hybridomas were phenotyped and assayed for their Ag specificity. A, FACS histograms of cell surface mouse CD3 (mCD3), mCD4, and hCD4 expression on the BW5147 fusion partner and the representative T hybridoma DB4. B, DB4 T hybridoma cells were cultured with irradiated hCD4.DR3-DQ2.MHCIIΔ/Δ Tg splenocytes and Α-gliadin57–73 Q65E or TPO; after 24 h, supernatants were added to IL-2-dependent CTLLs and their proliferation was assayed after 24 h. C, Activation of DB4 cells incubated with Α-gliadin57–73 Q65E in the presence or absence of Ab against HLA-DQ or HLA-DR (5 μg/ml) assayed as above. D, Activation of the DB4 hybridomas cultured with Α-gliadin57–73 Q65E and a panel of irradiated B-LCLs expressing DQ2 (α1*0501, β1*0201/2) encoded in cis- (•) or trans (▴) or DQ8 (α1*0301, β1*0302). Dotted line indicates the EC50 for the two observed responses measured by IL-2 bioassay as above.

The HLA-DQ2 molecule carried by the majority of CD patients is DQα1*05/DQβ1*02 (DQ2.5). This heterodimer can be encoded in cis- or in trans. A second DQ2 molecule (DQ2.2), comprised of DQα1*0201/DQβ1*02, is rarely associated with CD without DQ2.5 (4). Although possession of the DQA1*05 and DQB1*02 alleles confers the primary increased risk of CD, studies have shown that this risk is increased four to six times by homozygosity for the cis-haplotype or a second DQB1*02 allele on the other haplotype (30, 31). To verify the functional effect of DQ2 gene dosage on T cell recognition (and further confirm the MHC restriction element for the hybridomas), one T hybridoma line (DB4) was cultured with Α-gliadin57–73 Q65E and an irradiated panel of B-lymphoblastoid cell lines (B-LCLs) expressing different DQα and DQβ chains (Fig. 3⇑D). Hybridomas incubated with DQ2.5 homozygous B-LCLs (9022 cells) or B-LCLs expressing DQA1*05 and DQB1*02 in trans (T177 cells) responded to gliadin peptide in a dose-dependent manner, whereas neither DQ2.2, DQ7, nor DQ8 homozygous B-LCLs activated the T hybridomas (Fig. 3⇑D). Importantly, the hybridomas were most efficiently activated by DQ2.5 homozygous B-LCLs (see EC50 for the dose-response curves; Fig. 3⇑D). This finding is consistent with human T cell clones isolated from intestinal lines raised against deamidated gliadin (32).

A second T hybridoma (HH8; with specificity similar to DB4) coexpressed TCR Vα8.3 and TCR Vβ8.2 (data not shown). T cells using this combination of TCRαβ chains can be tracked using mAbs; therefore, the HH8 TCR was selected to generate TCR Tg mice. To identify which of the two epitopes in A-gliadin57–73 Q65E (DQ2-α-I or DQ2-α-II) was recognized by the HH8 TCR, the hybridoma was cultured with irradiated hCD4.DR3-DQ2.MHCIIΔ/Δ Tg splenocytes and 15-mer peptides spanning a 9-residue core sequence from Α-gliadin57–73 Q65E. The hybridoma responded to 15-mers composed of Α-gliadin residues 61–69 (GSG-FPQPELPYP-GSG), 62–70 (GSG-PQPELPYPQ-GSG), and 63–71 (GSG-QPELPYPQ-GSG); peptides containing residues 58–66, 59–67, or 60–68 did not stimulate a response (Fig. 4⇓A). The fine specificity of the HH8 TCR was further studied by examining responsiveness of the hybridoma to lysine (K) substitutions at each of the 17 aa positions of Α-gliadin57–73 Q65E. The HH8 hybridoma did not tolerate any lysine substitution from P62 through to Q72 (Fig. 4⇓B). Thus, the specificity of the HH8 hybridoma was consistent with the human CD-associated DQ2-α-II gliadin epitope (PQPELPYPQ) (33).

FIGURE 4.
FIGURE 4.

The fine specificity of murine gliadin-specific CD4+ T cell hybridoma. HH8 resembles the human T cell response toward the dominant DQ2-α-II epitope in patients with CD. HH8 T hybridoma cells were cultured with irradiated hCD4.DR3-DQ2.MHCIIΔ/Δ Tg splenocytes and (A) Α-gliadin57–73 Q65E or 15-mers spanning a nine-residue core sequence of the DQ2-α-II peptide or (B) lysine substitutions of each amino acid in Α-gliadin57–73 Q65E. Supernatants were added to IL-2-dependent CTLLs in a 24-h proliferation assay. Bars represent the mean value obtained in a triplicate experiment; error bars represent the SD.

Gliadin-specific CD4+ T cells are present and populate the periphery in TCR Tg mice

A gliadin-specific TCR Tg line was generated on the hCD4.DR3-DQ2.MHCIIΔ/Δ background using αβTCR genes of the HH8 T hybridoma. The TCRα and TCRβ transgenes cosegregated and flow cytometric analysis indicated that ∼93% of the thymocytes were CD4+ and 39% were mature CD4+CD8− cells (Fig. 5⇓). In the periphery of the TCR Tg mice, 24% of splenic lymphocytes and 53% of MLN lymphocytes were CD4+ (both human and mouse; Fig. 5⇓). Importantly, 97% of peripheral CD4+ T cells expressed the transgene-encoded TCR (Fig. 5⇓). These data indicate normal thymic development and peripheral export of T cells expressing the TCRα and β transgenes on the hCD4.DR3-DQ2.MHCIIΔ/Δ background. In contrast, development of the CD8+ T cell compartment is affected by the TCR transgenes as the CD8+CD4− subset is dramatically reduced in both the periphery and thymus (Fig. 5⇓).

FIGURE 5.
FIGURE 5.

HH8 TCR Tg mice develop single-positive CD4+ T cells expressing the Tg TCR. Representative dot plots of FACS costaining for Vβ8 and Vα8.3 expressed by HH8. Staining for mCD4 and hCD4 is also shown on lymphocytes from thymus, spleen, and MLN of TCR Tg mice. Top and middle rows are gated on lymphocytes; bottom row is gated on mouse and human CD4+ lymphocytes.

Gliadin-specific Tg CD4+ T cells are functional in vitro and in vivo

To confirm the specificity of the CD4+ T cells in the TCR Tg mouse and determine whether they were functionally responsive to Ag, splenocytes from these animals were depleted of B cells and examined for their ability to proliferate in vitro in the presence of Α-gliadin57–73 Q65E. Ag dose-dependent proliferation of the T cells occurred in response to the DQ2-α-II epitope (Fig. 6⇓A) and fine mapping revealed dependence on the same core residues as observed with the parental HH8 hybridoma (Fig. 4⇑B and data not shown).

FIGURE 6.
FIGURE 6.

CD4+ T cells from naive TCR Tg mice respond to the DQ2-α-II gliadin peptide both in vitro and in vivo. A, Naive splenocytes from TCR Tg mice were depleted of B cells and cultured with irradiated hCD4.DR3-DQ2.MHCIIΔ/Δ splenocytes and Α-gliadin57–73 Q65E; cells were assessed for proliferation by incorporation of [3H]thymidine during the last 24 h of a 72-h culture. B, Naive B cell-depleted splenocytes from TCR Tg mice were CFSE labeled and transferred into hCD4.DR3-DQ2.MHCIIΔ/Δ mice immunized with 40 μg of A-gliadin57–73 Q65E emulsified in CFA. On day 3, cells from draining and nondraining lymph nodes were analyzed by FACS for mCD4, TCR Vα8.3, and TCR Vβ8. Dot plots are gated on lymphocytes.

To assess whether the Tg CD4+ T cells were functional in vivo, purified T cells were labeled with the cytoplasmic dye CFSE and transferred into hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice that were then immunized with Α-gliadin57–73 Q65E or control TPO peptide in CFA. Three days after transfer and immunization, draining and nondraining lymph nodes were removed and cell proliferation was analyzed by flow cytometry. The transferred T cells proliferated in the draining lymph nodes of mice immunized with Α-gliadin57–73 Q65E, with seven rounds of division observed; proliferation was not seen in cells from the nondraining lymph nodes of these mice (Fig. 6⇑B). Similarly, transgene-encoded T cells proliferated in the draining nodes of mice immunized with native Α-gliadin57–73, chymotrypsin-digested gliadin, or whole crude gluten, but the proportion of cells dividing was lower than that observed with synthetic Α-gliadin57–73 Q65E (data not shown). T cell proliferation was not observed in TPO-immunized mice (data not shown). Furthermore, CD4+ T cells from the TCR Tg possessed a Th1-type effector phenotype. On stimulation of CD4+ T cells from the TCR Tg mice with Α-gliadin57–73 Q65E (and Con A), there was prominent secretion of IFN-γ and TNF-α and to a lesser extent IL-2, but relatively much less IL-4 or IL-5 (Fig. 7⇓).

FIGURE 7.
FIGURE 7.

Gliadin-specific T cells from HH8 TCR Tg mice produce Th1-type cytokines upon encounter with cognate peptide containing the DQ2-α-II epitopes. Cells from the spleen, MLNs, and Peyer’s patches (PP) of (A) C57BL/6 mice or (B) TCR Tg mice were cultured with medium alone, 2 μM Α-gliadin57–73 Q65E, or 10 μg/ml Con A. After 48 h, supernatants were added to a BD Biosciences Cytometric Bead Array Mouse Th1/Th2 Cytokine Kit and analyzed by FACS. Dot clusters in each plot represent IL-2, IL-4, IL-5, IFN-γ, and TNF-α (top to bottom). Clusters to the right of the dotted line indicate cytokine production above background levels.

Taken together, these results demonstrate that CD4+ T cells in the TCR Tg mice are gliadin peptide-specific and functionally capable of responding to Ag in vitro and in vivo. Importantly, the gliadin-specific T cells secrete Th1-type cytokines upon encounter with gliadin peptide; such cytokines are proposed to be instrumental in the small intestinal damage associated with CD (34).

Gliadin peptide is detectable by T cells in draining lymphoid tissue following mucosal delivery

To determine whether immunogenic gliadin peptide is absorbed from the healthy murine gut, TCR Tg splenocytes were labeled with CFSE and transferred as reporter cells into hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice. Following transfer, hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice on the gluten-free diet were gavaged daily with Α-gliadin57–73 Q65E or vehicle only and, on day 3, MLN and Peyer’s patch cells were isolated and examined for proliferating gliadin-specific T cells by flow cytometry. Transferred T cells were present in both the Peyer’s patches and MLNs of peptide-gavaged and vehicle-gavaged mice, but it was only in the gliadin peptide-gavaged mice that the transferred T cells were proliferating (Fig. 8⇓A). T cell proliferation was not as marked as observed in draining lymph nodes of mice s.c. immunized with Α-gliadin57–73 Q65E (Fig. 6⇑B).

FIGURE 8.
FIGURE 8.

Limited proliferation of gliadin-specific T cells in gut-draining lymphoid tissues following exposure to oral peptide is not associated with increased Foxp3+ T cells. A, Naive B cell-depleted splenocytes from TCR Tg mice were CFSE-labeled and transferred into hCD4.DR3-DQ2.MHCIIΔ/Δ mice undergoing a 3-day gavage with 2 mg/day Α-gliadin57–73 Q65E suspended in peanut oil or peanut oil only (vehicle only). On day 3, cells from MLNs and Peyer’s patches were analyzed by FACS for CFSE fluorescence and expression of mCD4, Vβ8, and Vα8.3. Dot plots are gated on lymphocytes. B, Dot plots of FACS staining for mCD4, TCR Vβ8, TCR Vα8.3, and intracellular Foxp3 on cells from spleen, MLNs, Peyer’s patches, and peripheral lymph nodes (PLN) of gluten-free diet or 10-day gluten-fed TCR Tg mice. Plots are gated on CD4+ lymphocytes and are representative of eight mice.

Foxp3 expression by CD4+CD25+Foxp3+ T cells following gluten ingestion

Given that hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice remain healthy on gluten-containing diet despite being able to absorb immunogenic gliadin peptide, we investigated whether regulatory T cells might be induced by gluten ingestion. However, the proportion of T cells expressing CD4+CD25+Foxp3+ was no different in the spleen, MLN, Peyer’s patch, and axillary lymph node (peripheral lymph node) from B6 mice, TCR Tg mice on gluten-free food, and TCR Tg mice given gluten-containing food for 10 days (Fig. 8⇑B). In addition, distinguishing gliadin-specific CD4+ T cells from the small proportion of CD4+ T cells expressing an endogenously encoded TCR (non-Tg T cells) revealed no significant difference (Fig. 8⇑B).

Tg mice expressing gliadin-specific T cells remain healthy while consuming gluten

Experimental protocols used earlier in the hCD4.DR3-DQ2.MHCIIΔ/Δ mice to induce enteropathy, including immunization with gliadin peptide/protein before gluten gavage with coadministration of pertussis toxin or coincident infection with S. typhimurium, did not induce tTG-specific IgA or any discernable signs of intestinal pathology in the TCR Tg mice (data not shown). Notably, an increase in IEL numbers was not observed in the TCR Tg animals following any intervention. In fact, these animals appeared to have a lower IEL count than observed in B6 controls and “healthy” hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice. The lower number of IELs in the TCR Tg mice is probably attributable to the impact of the prearranged TCRα and TCRβ transgenes on CD8 T cell development and possibly also γδ T cell development, as these populations are reduced in the periphery and are known to comprise a significant proportion of intestinal IELs.

Discussion

HLA-DQ2 or DQ8 is necessary but not sufficient for development of CD (2). Early introduction of gluten while infants are not breast-fed, frequent episodes of gastroenteritis before age 6 mo, homozygosity for HLA-DQ2, and a first-degree relative having CD are among the strongest factors predicting development of CD (2, 31, 35, 36). Presumably, these factors influence the priming, expansion and/or maintenance of HLA-DQ2-restricted CD4+ Th1 cells specific for deamidated gluten found in established CD.

In human CD, enteropathy probably evolves gradually from intraepithelial lymphocytosis to subtotal villous atrophy, and in clinical practice serum tTG IgA may precede intestinal damage (37, 38). Autoantibodies specific for tTG typically appear in serum for the first time between the ages of 1 and 7 years, but can be transient or even reappear later in childhood (39). These IgA class autoantibodies disappear with gluten exclusion and are widely considered to be dependent on uptake of tTG-gluten peptide complexes by tTG-specific B cells receiving help from gluten-specific CD4 T cells (40, 41). Hence, contemporary hypotheses accounting for the development of CD invoke gene dose of HLA-DQ2, various genes regulating the activation threshold for CD4+ T cells, costimulation and danger signals provided by intestinal infection and/or innate immunity (triggered by gluten itself), mucosal transglutaminase activity, intestinal permeability, the dose of gluten ingested, and the unusual chemical properties of gluten (12, 42, 43).

Although this model of CD induction is plausible, supporting evidence is observational; many aspects would be testable in a valid animal model. To model the CD4+ T cell-mediated pathology of CD, we introduced the HLA-DR3-DQ2 haplotype and the hCD4 gene into mice lacking murine MHCII genes (MHCIIΔ/Δ). These mice Tg for hCD4.DR3-DQ2.MHCIIΔ/Δ express functional HLA-DR3 and DQ2 molecules that mediate thymic selection generating Ag-specific CD4+ T cell immunity. Although about one in five Europeans homozygous for HLA-DQ2 develops CD (31), homozygous and heterozygous hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice do not spontaneously develop hypersensitivity to gluten. However, standard immunization protocols do produce HLA-DQ2-restricted CD4+ T cells in hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice that preferentially recognize the deamidated α-gliadin 17-mer immunodominant in human CD (19). Nevertheless, we have found that preexisting DQ2-restricted CD4+ T cell immunity to deamidated α-gliadin 17-mer in hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice is not sufficient for gluten ingestion to cause measurable intestinal pathology.

Induction of celiac pathology might require activation of local intestinal innate immunity at the time of initial exposure to dietary gluten. Conditions designed to create an inflammatory context, such as infecting with S. typhimurium (an enteric bacterial pathogen for rodents (44)) or prior exposure to a Rhesus strain of rotavirus (data not shown), did not result in consistent signs of celiac pathology, such as increased numbers of IELs or villous changes on histological examination. However, systemic administration of pertussis toxin before immunization with the immunogenic gluten peptide lead to IgA autoantibody production and mild histological changes in the small intestine of a small proportion of hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice.

To explore the effect of other genetic factors, hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice were bred to the autoimmunity-prone NOD background, which in DQ8 Tg mice predisposes to gluten-sensitive cutaneous pathology (14). However, hCD4.DR3-DQ2.MHCIIΔ/Δ Tg NOD mice immunized with gliadin and coinjected with pertussis toxin remained healthy.

Next, we considered whether the magnitude of the CD4+ T cell response to gluten might be inadequate in hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice immunized with gliadin or the α-gliadin 17-mer to cause celiac-like intestinal pathology. Autoantigen-specific CD4+ TCR Tg mice have been used to generate various models of HLA-linked autoimmunity (45, 46, 47, 48). Therefore, we generated gliadin-specific CD4+ TCR Tg mice on the hCD4.DR3-DQ2.MHCIIΔ/Δ background. The transgene-encoded TCR was derived from a DQ2-restricted CD4+ T cell hybridoma specific for the DQ2-α-II epitope commonly recognized by intestinal T cell clones from DQ2+ celiac patients (26, 33). Gliadin-specific CD4+ T cells were thymically selected and exported to the periphery of the TCR Tg mice, where they constituted the vast majority of T cells. By comparison, the maximum frequency of T cells specific for the immunodominant DQ2-α-I and DQ2-α-II epitopes, enumerated by MHC-peptide tetramers, is reported to be no more than 0.4% of the total peripheral blood CD4+ T cell population (49).

TCR Tg T cells proliferated following systemic administration of Α-gliadin57–73 Q65E or chymotrypsin-digested gliadin in vivo and secreted Th1-type cytokines implicated in CD (34) in vitro. But ingestion of gluten or Α-gliadin57–73 Q65E did not trigger intestinal pathology in TCR Tg mice, either spontaneously or induced by immunization strategies that have produced autoimmune disease in other TCR Tg models (45, 46, 48).

The resistance of these mouse models to development of celiac enteropathy is surprising given the increased precursor frequency of gliadin-specific CD4+ T cells engendered in the TCR Tg system. One possibility is that T cells were not exposed to sufficient levels of the gluten peptides to trigger tissue damage. Gavage with deamidated Α-gliadin57–73 (2 mg daily) did stimulate modest T cell proliferation in gut-draining lymph nodes, but this was substantially less than observed in local draining lymph nodes following s.c. administration with one-fiftieth of the dose (0.04 mg). In gluten challenge experiments, mice were fed a substantial amount of gluten (∼1 g/mouse, three times per week) in the form of standard gluten-containing bread. For comparison, spiking the gluten-free diet with 50 mg of gluten daily causes 20% reduction in the villous height:crypt depth ratio in patients with established CD (50). Even though the dose of gluten ingested was substantial, only a modest amount of peptide may have been absorbed due to relatively low permeability or luminal proteolysis; yet permeability would be expected to increase with coexistent infectious enteritis and immunogenic gluten peptides are unusually resistant to rodent and human digestive proteases (10). Mucosal transglutaminase may have been the other factor limiting exposure of T cells to cognate (deamidated) gluten peptides with enhanced binding to HLA-DQ2. Although not directly measured, transglutaminase activity is up-regulated with inflammation (51) and would also be expected to increase with coincident enteric infection, an experimental condition that did not trigger gluten enteropathy. In summary, the experimental conditions assessed would have been expected to result in adequate presentation of immunogenic gluten peptides to reactivate primed T cells. Additionally, both hCD4.DR3-DQ2.MHCIIΔ/Δ Tg mice that received 5–10 million cognate T cells (by adoptive transfer) and unmanipulated TCR Tg mice would be expected to have substantially greater frequencies of T cells specific for immunodominant gluten peptide than observed in human CD (49).

Although mice may have been “tolerized” by accidental exposure to gluten before deliberate immunization with gluten, we think this possibility is unlikely. A more plausible explanation for our findings is that in the absence of mucosal damage, the gastrointestinal mucosa of the mouse is innately resistant to inflammation. Mice deficient in IL-10 or TGF-β are predisposed to intestinal inflammation (52, 53), indicating the importance of cytokine balance in mediating steady-state immunological hyporesponsiveness. PGE2 may be a crucial contributor to intestinal hyporesponsiveness because it is highly expressed by lamina propria mononuclear cells and is known to suppress immune responses (54). Notably too, GALT is a preferential site for the peripheral induction of Foxp3+ regulatory T cells (55). Such regulatory T cells may contribute to immunological homeostasis in the intestine with respect to ingested gluten, but our study found no evidence for an expanded population of Foxp3+ regulatory T cells in the gut-draining lymph nodes of TCR Tg mice following exposure to dietary gluten.

Given the potential for multiple quantitative susceptibility traits to contribute to disease penetrance in DQ2+ individuals, it may be necessary to make backcrosses of the DQ2 and TCR transgenes onto multiple mouse genetic backgrounds to screen for higher disease penetrance. Alternatively, if there were known gene polymorphisms with significant impact on susceptibility to CD, these could be targeted individually as components of an improved animal model of CD.

Taken together, our findings indicate that gluten ingestion in humanized mice expressing functional HLA-DQ2 and possessing a substantial population of CD4+ T cells specific for gluten is not sufficient to cause celiac-like enteropathy. These observations are particularly relevant in the context of an evolving viewpoint that CD may be better defined as a systemic immune response to gluten with fluctuating or occasionally absent gut pathology (56). Accordingly, the model provides a unique opportunity to study both the systemic immune response to gluten, as well as the local gut-associated regulation of gluten-specific T cells relevant to HLA-DQ2-associated CD.

Acknowledgments

We thank Anthony Purcell and Nick Williamson for quality control of synthesized peptides, Prithi Bhatal for assessment of mouse pathology, Maria Kaparakis for assistance with histology, Dianna Starczak for assistance with statistics, and Tom Gordon and Jenny Roland for providing patient sera.

Disclosures

Robert P. Anderson is the inventor of patents relating to diagnostics, therapeutics, and nontoxic gluten based upon knowledge of peptides recognized by T cells in celiac disease. Robert P. Anderson is also involved in the commercialization of these patents as Director, Chief Executive Officer, and is a substantial shareholder in Nexpep Pty Ltd and Nexgrain Pty Ltd.

Footnotes

  • The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

  • ↵1 This work was supported by a National Health and Medical Research Council of Australia Program Grant.

  • ↵2 A.L.d.K. and Z.C. contributed equally to the work.

  • ↵3 Address correspondence and reprint requests to Prof. James McCluskey, Department of Microbiology and Immunology, University of Melbourne, Parkville 3010, Australia. E-mail address: jamesm1{at}unimelb.edu.au

  • ↵4 Abbreviations used in this paper: CD, celiac disease; tTG, tissue transglutaminase; Tg, transgenic; hCD4, human CD4; mCD4, mouse CD4; TPO, thyroid peroxidase; MLN, mesenteric lymph node; MHCII, MHC class II; IEL, intraepithelial lymphocyte; B-LCL, B lymphoblastoid cell line.

  • Received January 22, 2009.
  • Accepted April 17, 2009.

References

  1. ↵
    van Heel, D. A., J. West. 2006. Recent advances in celiac disease. Gut 55: 1037-1046.
    OpenUrlFREE Full Text
  2. ↵
    Sollid, L. M.. 2002. Celiac disease: dissecting a complex inflammatory disorder. Nat. Rev. Immunol. 2: 647-655.
    OpenUrlCrossRefPubMed
  3. ↵
    Dieterich, W., T. Ehnis, M. Bauer, P. Donner, U. Volta, E. O. Riecken, D. Schuppan. 1997. Identification of tissue transglutaminase as the autoantigen of celiac disease. Nat. Med. 3: 797-801.
    OpenUrlCrossRefPubMed
  4. ↵
    Karell, K., A. S. Louka, S. J. Moodie, H. Ascher, F. Clot, L. Greco, P. J. Ciclitira, L. M. Sollid, J. Partanen. 2003. HLA types in celiac disease patients not carrying the DQA1*05-DQB1*02 (DQ2) heterodimer: results from the European Genetics Cluster on Celiac Disease. Hum. Immunol. 64: 469-477.
    OpenUrlCrossRefPubMed
  5. ↵
    Lundin, K. E., H. Scott, T. Hansen, G. Paulsen, T. S. Halstensen, O. Fausa, E. Thorsby, L. M. Sollid. 1993. Gliadin-specific, HLA-DQα1*0501, β1*0201) restricted T cells isolated from the small intestinal mucosa of celiac disease patients. J. Exp. Med. 178: 187-196.
    OpenUrlAbstract/FREE Full Text
  6. ↵
    Molberg, O., S. N. McAdam, R. Korner, H. Quarsten, C. Kristiansen, L. Madsen, L. Fugger, H. Scott, O. Noren, P. Roepstorff, et al 1998. Tissue transglutaminase selectively modifies gliadin peptides that are recognized by gut-derived T cells in celiac disease. Nat. Med. 4: 713-717.
    OpenUrlCrossRefPubMed
  7. ↵
    van de Wal, Y., Y. Kooy, P. van Veelen, S. Pena, L. Mearin, G. Papadopoulos, F. Koning. 1998. Selective deamidation by tissue transglutaminase strongly enhances gliadin-specific T cell reactivity. J. Immunol. 161: 1585-1588.
    OpenUrlAbstract/FREE Full Text
  8. ↵
    Vader, L. W., A. de Ru, Y. van der Wal, Y. M. Kooy, W. Benckhuijsen, M. L. Mearin, J. W. Drijfhout, P. van Veelen, F. Koning. 2002. Specificity of tissue transglutaminase explains cereal toxicity in celiac disease. J. Exp. Med. 195: 643-649.
    OpenUrlAbstract/FREE Full Text
  9. ↵
    Fleckenstein, B., O. Molberg, S. W. Qiao, D. G. Schmid, F. von der Mulbe, K. Elgstoen, G. Jung, L. M. Sollid. 2002. Gliadin T cell epitope selection by tissue transglutaminase in celiac disease: role of enzyme specificity and pH influence on the transamidation versus deamidation process. J. Biol. Chem. 277: 34109-34116.
    OpenUrlAbstract/FREE Full Text
  10. ↵
    Shan, L., O. Molberg, I. Parrot, F. Hausch, F. Filiz, G. M. Gray, L. M. Sollid, C. Khosla. 2002. Structural basis for gluten intolerance in celiac sprue. Science 297: 2275-2279.
    OpenUrlAbstract/FREE Full Text
  11. ↵
    van Heel, D. A., L. Franke, K. A. Hunt, R. Gwilliam, A. Zhernakova, M. Inouye, M. C. Wapenaar, M. C. Barnardo, G. Bethel, G. K. Holmes, et al 2007. A genome-wide association study for celiac disease identifies risk variants in the region harboring IL2 and IL21. Nat. Genet. 39: 827-829.
    OpenUrlCrossRefPubMed
  12. ↵
    Hunt, K. A., A. Zhernakova, G. Turner, G. A. Heap, L. Franke, M. Bruinenberg, J. Romanos, L. C. Dinesen, A. W. Ryan, D. Panesar, et al 2008. Newly identified genetic risk variants for celiac disease related to the immune response. Nat. Genet. 40: 395-402.
    OpenUrlCrossRefPubMed
  13. ↵
    Black, K. E., J. A. Murray, C. S. David. 2002. HLA-DQ determines the response to exogenous wheat proteins: a model of gluten sensitivity in transgenic knockout mice. J. Immunol. 169: 5595-5600.
    OpenUrlAbstract/FREE Full Text
  14. ↵
    Marietta, E., K. Black, M. Camilleri, P. Krause, R. S. Rogers, III, C. David, M. R. Pittelkow, J. A. Murray. 2004. A new model for dermatitis herpetiformis that uses HLA-DQ8 transgenic NOD mice. J. Clin. Invest. 114: 1090-1097.
    OpenUrlCrossRefPubMed
  15. ↵
    Chen, Z., N. Dudek, O. Wijburg, R. Strugnell, L. Brown, G. Deliyannis, D. Jackson, F. Koentgen, T. Gordon, J. McCluskey. 2002. A 320-kilobase artificial chromosome encoding the human HLA DR3-DQ2 MHC haplotype confers HLA restriction in transgenic mice. J. Immunol. 168: 3050-3056.
    OpenUrlAbstract/FREE Full Text
  16. ↵
    Madsen, L., N. Labrecque, J. Engberg, A. Dierich, A. Svejgaard, C. Benoist, D. Mathis, L. Fugger. 1999. Mice lacking all conventional MHC class II genes. Proc. Natl. Acad. Sci. USA 96: 10338-10343.
    OpenUrlAbstract/FREE Full Text
  17. ↵
    Killeen, N., S. Sawada, D. R. Littman. 1993. Regulated expression of human CD4 rescues helper T cell development in mice lacking expression of endogenous CD4. EMBO J. 12: 1547-1553.
    OpenUrlPubMed
  18. ↵
    Jackson, D. C., C. J. Fitzmaurice, L. E. Brown, W. Zeng. 1999. Preparation and properties of totally synthetic immunogens. Vaccine 18: 355-361.
    OpenUrlCrossRefPubMed
  19. ↵
    Anderson, R. P., P. Degano, A. J. Godkin, D. P. Jewell, A. V. Hill. 2000. In vivo antigen challenge in celiac disease identifies a single transglutaminase-modified peptide as the dominant A-gliadin T-cell epitope. Nat. Med. 6: 337-342.
    OpenUrlCrossRefPubMed
  20. ↵
    Price, J. D., K. R. Simpfendorfer, R. R. Mantena, J. Holden, W. R. Heath, N. van Rooijen, R. A. Strugnell, O. L. Wijburg. 2007. γ Interferon-independent effects of interleukin-12 on immunity to Salmonella enterica serovar Typhimurium. Infect. Immun. 75: 5753-5762.
    OpenUrlAbstract/FREE Full Text
  21. ↵
    Chen, W., S. Khilko, J. Fecondo, D. H. Margulies, J. McCluskey. 1994. Determinant selection of MHC class I-restricted antigenic peptides is explained by class I-peptide affinity and is strongly influenced by non-dominant anchor residues. J. Exp. Med. 180: 1471-1483.
    OpenUrlAbstract/FREE Full Text
  22. ↵
    Brooks, A., J. McCluskey. 1993. Class II-restricted presentation of a hen egg lysozyme determinant derived from endogenous antigen sequestered in the cytoplasm or endoplasmic reticulum of the antigen presenting cells. J. Immunol. 150: 3690-3697.
    OpenUrlAbstract
  23. ↵
    Barnden, M. J., J. Allison, W. R. Heath, F. R. Carbone. 1998. Defective TCR expression in transgenic mice constructed using cDNA-based α- and β-chain genes under the control of heterologous regulatory elements. Immunol. Cell Biol. 76: 34-40.
    OpenUrlCrossRefPubMed
  24. ↵
    Kaye, J., N. J. Vasquez, S. M. Hedrick. 1992. Involvement of the same region of the T cell antigen receptor in thymic selection and foreign peptide recognition. J. Immunol. 148: 3342-3353.
    OpenUrlAbstract
  25. ↵
    Van Schooten, W. C., D. G. Elferink, J. Van Embden, D. C. Anderson, R. R. De Vries. 1989. DR3-restricted T cells from different HLA-DR3-positive individuals recognize the same peptide (amino acids 2–12) of the mycobacterial 65-kDa heat-shock protein. Eur. J. Immunol. 19: 2075-2079.
    OpenUrlPubMed
  26. ↵
    Arentz-Hansen, H., R. Korner, O. Molberg, H. Quarsten, W. Vader, Y. M. Kooy, K. E. Lundin, F. Koning, P. Roepstorff, L. M. Sollid, S. N. McAdam. 2000. The intestinal T cell response to α-gliadin in adult celiac disease is focused on a single deamidated glutamine targeted by tissue transglutaminase. J. Exp. Med. 191: 603-612.
    OpenUrlAbstract/FREE Full Text
  27. ↵
    Anderson, R. P., D. A. van Heel, J. A. Tye-Din, D. P. Jewell, A. V. Hill. 2006. Antagonists and non-toxic variants of the dominant wheat gliadin T cell epitope in celiac disease. Gut 55: 485-491.
    OpenUrlAbstract/FREE Full Text
  28. ↵
    Raine, C. S.. 1984. Biology of disease. Analysis of autoimmune demyelination: its impact upon multiple sclerosis. Lab. Invest. 50: 608-635.
    OpenUrlPubMed
  29. ↵
    Dayan, C. M., M. Londei, A. E. Corcoran, B. Grubeck-Loebenstein, R. F. James, B. Rapoport, M. Feldmann. 1991. Autoantigen recognition by thyroid-infiltrating T cells in Graves disease. Proc. Natl. Acad. Sci. USA 88: 7415-7419.
    OpenUrlAbstract/FREE Full Text
  30. ↵
    Sollid, L. M., G. Markussen, J. Ek, H. Gjerde, F. Vartdal, E. Thorsby. 1989. Evidence for a primary association of celiac disease to a particular HLA-DQ α/β heterodimer. J. Exp. Med. 169: 345-350.
    OpenUrlAbstract/FREE Full Text
  31. ↵
    Ploski, R., J. Ek, E. Thorsby, L. M. Sollid. 1993. On the HLA-DQ(α1*0501, β1*0201)-associated susceptibility in celiac disease: a possible gene dosage effect of DQB1*0201. Tissue Antigens 41: 173-177.
    OpenUrlPubMed
  32. ↵
    Vader, W., D. Stepniak, Y. Kooy, L. Mearin, A. Thompson, J. J. van Rood, L. Spaenij, F. Koning. 2003. The HLA-DQ2 gene dose effect in celiac disease is directly related to the magnitude and breadth of gluten-specific T cell responses. Proc. Natl. Acad. Sci. USA 100: 12390-12395.
    OpenUrlAbstract/FREE Full Text
  33. ↵
    Arentz-Hansen, H., S. N. McAdam, O. Molberg, B. Fleckenstein, K. E. Lundin, T. J. Jorgensen, G. Jung, P. Roepstorff, L. M. Sollid. 2002. Celiac lesion T cells recognize epitopes that cluster in regions of gliadins rich in proline residues. Gastroenterology 123: 803-809.
    OpenUrlCrossRefPubMed
  34. ↵
    Nilsen, E. M., F. L. Jahnsen, K. E. Lundin, F. E. Johansen, O. Fausa, L. M. Sollid, J. Jahnsen, H. Scott, P. Brandtzaeg. 1998. Gluten induces an intestinal cytokine response strongly dominated by interferon γ in patients with celiac disease. Gastroenterology 115: 551-563.
    OpenUrlCrossRefPubMed
  35. ↵
    Ivarsson, A., O. Hernell, H. Stenlund, L. A. Persson. 2002. Breast-feeding protects against celiac disease. Am. J. Clin. Nutr. 75: 914-921.
    OpenUrlAbstract/FREE Full Text
  36. ↵
    Rubio-Tapia, A., C. T. van Dyke, B. D. Lahr, A. R. Zinsmeister, M. El-Youssef, S. B. Moore, M. Bowman, L. J. Burgart, L. J. Melton, III, J. A. Murray. 2008. Predictors of family risk for celiac disease: a population-based study. Clin. Gastroenterol. Hepatol. 6: 983-987.
    OpenUrlCrossRefPubMed
  37. ↵
    Dickey, W., D. F. Hughes, S. A. McMillan. 2005. Patients with serum IgA endomysial antibodies and intact duodenal villi: clinical characteristics and management options. Scand. J. Gastroenterol. 40: 1240-1243.
    OpenUrlCrossRefPubMed
  38. ↵
    Salmi, T. T., P. Collin, J. A. Rvinen, K. Haimila, J. Partanen, K. Laurila, I. R. Korpponay-Szabo, H. Huhtala, T. Reunala, M. Maki, K. Kaukinen. 2006. Immunoglobulin A autoantibodies against transglutaminase 2 in the small intestinal mucosa predict forthcoming coeliac disease. Aliment. Pharmacol. Ther. 24: 541-552.
    OpenUrlCrossRefPubMed
  39. ↵
    Simell, S., A. Kupila, S. Hoppu, A. Hekkala, T. Simell, M. R. Stahlberg, M. Viander, T. Hurme, M. Knip, J. Ilonen, et al 2005. Natural history of transglutaminase autoantibodies and mucosal changes in children carrying HLA-conferred celiac disease susceptibility. Scand. J. Gastroenterol. 40: 1182-1191.
    OpenUrlCrossRefPubMed
  40. ↵
    Dickey, W., D. F. Hughes, S. A. McMillan. 2000. Disappearance of endomysial antibodies in treated celiac disease does not indicate histological recovery. Am. J. Gastroenterol. 95: 712-714.
    OpenUrlCrossRefPubMed
  41. ↵
    Sollid, L. M., O. Molberg, S. McAdam, K. E. Lundin. 1997. Autoantibodies in celiac disease: tissue transglutaminase: guilt by association?. Gut 41: 851-852.
    OpenUrlFREE Full Text
  42. ↵
    Vader, W., D. Stepniak, Y. Kooy, L. Mearin, A. Thompson, J. J. van Rood, L. Spaenij, F. Koning. 2003. The HLA-DQ2 gene dose effect in celiac disease is directly related to the magnitude and breadth of gluten-specific T cell responses. Proc. Natl. Acad. Sci. USA 100: 12390-12395.
    OpenUrlAbstract/FREE Full Text
  43. ↵
    Jabri, B., L. M. Sollid. 2006. Mechanisms of disease: immunopathogenesis of celiac disease. Nat. Clin. Precut. Gastroenterol. Hepatol. 3: 516-525.
    OpenUrl
  44. ↵
    Santos, R. L., S. Zhang, R. M. Stoles, R. A. Kingly, L. G. Adams, A. J. Balmier. 2001. Animal models of Salmonella infections: enteritis versus typhoid fever. Microbes Infect. 3: 1335-1344.
    OpenUrlCrossRefPubMed
  45. ↵
    Goverman, J., A. Woods, L. Larson, L. P. Weiner, L. Hood, D. M. Zaller. 1993. Transgenic mice that express a myelin basic protein-specific T cell receptor develop spontaneous autoimmunity. Cell 72: 551-560.
    OpenUrlCrossRefPubMed
  46. ↵
    Schmidt, D., J. Verdaguer, N. Averill, P. Santamaria. 1997. A mechanism for the major histocompatibility complex-linked resistance to autoimmunity. J. Exp. Med. 186: 1059-1075.
    OpenUrlAbstract/FREE Full Text
  47. ↵
    Verdaguer, J., D. Schmidt, A. Amrani, B. Anderson, N. Averill, P. Santamaria. 1997. Spontaneous autoimmune diabetes in monoclonal T cell nonobese diabetic mice. J. Exp. Med. 186: 1663-1676.
    OpenUrlAbstract/FREE Full Text
  48. ↵
    Osman, G. E., S. Cheunsuk, S. E. Allen, E. Chi, H. D. Liggitt, L. E. Hood, W. C. Ladiges. 1998. Expression of a type II collagen-specific TCR transgene accelerates the onset of arthritis in mice. Int. Immunol. 10: 1613-1622.
    OpenUrlAbstract/FREE Full Text
  49. ↵
    Raki, M., L. E. Fallang, M. Brottveit, E. Bergseng, H. Quarsten, K. E. Lundin, L. M. Sollid. 2007. Tetramer visualization of gut-homing gluten-specific T cells in the peripheral blood of celiac disease patients. Proc. Natl. Acad. Sci. USA 104: 2831-2836.
    OpenUrlAbstract/FREE Full Text
  50. ↵
    Catassi, C., E. Fabiani, G. Iacono, C. D'Agate, R. Francavilla, F. Biagi, U. Volta, S. Accomando, A. Picarelli, I. De Vitis, et al 2007. A prospective, double-blind, placebo-controlled trial to establish a safe gluten threshold for patients with celiac disease. Am. J. Clin. Nutr. 85: 160-166.
    OpenUrlAbstract/FREE Full Text
  51. ↵
    D'Argenio, G., I. Sorrentini, V. Cosenza, A. Gatto, P. Iovino, E. P. D'Armiento, F. Baldassarre, G. Mazzacca. 1992. Serum and tissue transglutaminase correlates with the severity of inflammation in induced colitis in the rat. Scand. J. Gastroenterol. 27: 111-114.
    OpenUrlCrossRefPubMed
  52. ↵
    Kühn, R., D. Löhler;, J. Rennick, K. Rajewsky, W. Müller. 1993. Interleukin-10-deficient mice develop chronic enterocolitis. Cell 75: 263-274.
    OpenUrlCrossRefPubMed
  53. ↵
    Egger, B., F. Procaccino, J. Lakshmanan, M. Reinshagen, P. Hoffmann, A. Patel, W. Reuben, S. Gnanakkan, L. Liu, L. Barajas, V. E. Eysselein. 1997. Mice lacking transforming growth factor α have an increased susceptibility to dextran sulfate-induced colitis. Gastroenterology 113: 825-832.
    OpenUrlCrossRefPubMed
  54. ↵
    Betz, M., B. S. Fox. 1991. Prostaglandin E2 inhibits production of Th1 lymphokines but not of Th2 lymphokines. J. Immunol. 146: 108-113.
    OpenUrlAbstract
  55. ↵
    Sun, C. M., J. A. Hall, R. B. Blank, N. Bouladoux, M. Oukka, J. R. Mora, Y. Belkaid. 2007. Small intestine lamina propria dendritic cells promote de novo generation of Foxp3 Treg cells via retinoic acid. J. Exp. Med. 204: 1775-1785.
    OpenUrlAbstract/FREE Full Text
  56. ↵
    Kaukinen, K., M. Maki, J. Partanen, H. Sievanen, P. Collin. 2001. Celiac disease without villous atrophy: revision of criteria called for. Dig. Dis. Sci. 46: 879-887.
    OpenUrlCrossRefPubMed
Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools

Jump to section