Dietary gluten alters the balance of pro-inflammatory and anti-inflammatory cytokines in T cells of BALB/c mice

Abstract

Several studies have documented that dietary modifications influence the development of type 1 diabetes. However, little is known about the interplay of dietary components and the penetration of diabetes incidence. In this study we tested if wheat gluten is able to induce differences in the cytokine pattern of Foxp3⁺ regulatory T cells, as well as Foxp3⁻ T cells, isolated from intestinal mucosal lymphoid tissue and non-mucosal lymphoid compartments in BALB/c mice. The gluten-containing standard diet markedly changed the cytokine expression within Foxp3⁻ T cells, in all lymphoid organs tested, towards a higher expression of pro-inflammatory interferon-γ (IFN-γ), interleukin-17 (IL-17) and IL-2. In Foxp3⁺ regulatory T cells, gluten ingestion resulted in a mucosal increase in IL-17 and IL-2 and an overall increase in IFN-γ and IL-4. The gluten-free diet induced an anti-inflammatory cytokine profile with higher proportion of transforming growth factor-β (TGF-β)⁺ Foxp3⁻ T cells in all tested lymphoid tissues and higher IL-10 expression within non-T cells in spleen, and a tendency towards a mucosal increase in TGF-β⁺ Foxp3⁺ regulatory T cells. Our data shows that the gluten-containing standard diet modifies the cytokine pattern of both Foxp3⁻ T cells and Foxp3⁺ regulatory T cells towards a more inflammatory cytokine profile. This immune profile may contribute to the higher type 1 diabetes incidence associated with gluten intake.

Introduction

Type 1 diabetes (T1D) and coeliac disease are both diseases of autoimmune origin. In T1D tolerance to beta cell antigens is broken, leading to expansion of autoreactive T cells that attack pancreatic beta cells with subsequent loss of insulin production. In addition to being a disease-initiating factor in coeliac disease, there is evidence that gluten affects the development of T1D. A gluten-free (GF) diet is found to have a protective effect on the development of disease, as observed in the non-obese diabetic (NOD) mice where a GF diet prevents diabetes.¹ Studies have found that antibody production against a wheat-storage globulin (Glb1) was higher in patients with T1D compared with controls² and newly diagnosed patients with T1D have enhanced T-cell reactivity to gluten.³ The highest incidence of diabetes in experimental animal models of T1D is found in animals on a wheat-based diet,⁴ and a cereal-based diet promoted T1D development in two animal models of T1D, the NOD mice and the biobreeding rat.^5,6 The diabetogenic potential of gluten seems to be dependent on the time of gluten introduction, both in biobreeding rats⁷ and NOD mice, as delayed exposure to dietary gluten in newborn NOD mice was sufficient to reduce later development of T1D.⁸ This is consistent with human studies where early introduction of gluten-containing cereals to children increases the risk of developing islet autoantibodies.⁹

Evidence points towards involvement of the intestinal immune system in the development of T1D. In NOD mice the lymphocytes infiltrating the islets express the gut-associated lymphoid tissue-specific integrin α₄/β₇,¹⁰ and mesenteric lymphocytes transfer diabetes from NOD mice to non-diabetic mice, which indicates that diabetogenic T cells are activated in the intestinal compartment before infiltrating the pancreas.¹¹ Moreover, gastrointestinal changes (increased intestinal permeability and enteropathy) are found in both human T1D patients¹² and in animal models of T1D.¹³

Few studies have been performed to clarify how dietary gluten affects the immune system. NOD mice on a wheat-based diet have a T helper type 1 (Th1) cytokine bias in the gut,¹⁴ and in biobreeding rats a cereal-based diet induces a Th1 cytokine profile in pancreas-infiltrating cells.¹⁵ Moreover, gluten has been shown to directly activate dendritic cells.^16,17 Hence, evidence points towards a role for gluten in T1D, perhaps through a complex interplay between gluten, predisposing genes and mucosal immunity. A better understanding of the effect of gluten on the mucosal immune system may help to explain the observed link between gluten intake and the development of T1D.

Regulatory T cells (Tregs) play an important role in the control of peripheral potentially autoreactive T cells, that have escaped central deletion in thymus. Autoimmune diseases may develop through an altered balance between self-reactive T cells and Tregs.¹⁸ The Tregs include a number of different T-cell subsets, including the forkhead box p3 transcription factor (Foxp3) expressing cells. Foxp3-expressing Tregs are a constitutively occurring T-cell sub-population whose deficiency is associated with development of autoimmune diseases, and that are involved in the control of the overall adaptive immune response both in health and disease.¹⁹ Foxp3⁺ Tregs are able to promote bystander suppression and to block an inflammatory response by regulation of various cell types, including CD4⁺ and CD8⁺ T cells, B cells and antigen-presenting cells.²⁰

Foxp3⁺ Tregs mediate their suppression through cell contact-dependent mechanisms, by secretion of immunosuppressive cytokines such as transforming growth factor-β (TGF-β), interleukin-10 (IL-10) and IL-35 or by competing with T effector (Teff) cells for cytokines, e.g. IL-2, that are necessary for survival and expansion of both Tregs and Teff cells.²⁰ Despite the fact that Foxp3⁺ Tregs are considered to be a constitutive, regulatory T-cell population, it has become evident that this subset is not always stable. When exposed to appropriate cytokine signals, a small fraction of Foxp3⁺ CD4⁺ T cells retains plasticity to differentiate into Teff cells, with the potential to produce inflammatory cytokines such as interferon-γ (IFN-γ) and so lose their regulatory activity.^21,22 Recently, IL-17-producing Foxp3⁺ cells have been identified in human peripheral blood, tonsils²³ and in small intestinal lamina propria.²⁴ Interleukin-17 is usually secreted by Th17 cells, a subpopulation of CD4⁺ T cells, which are instrumental in the response against microbial infection, but Th17 cells are also potent inducers of tissue inflammation and associated with organ-specific autoimmunity.²⁵ Considerable evidence exist in animals as well as humans for the importance of IFN-γ and IL-17 in the development and progression of autoimmune diseases, and for TGF-β and IL-10 as key mediators in maintaining T-cell tolerance to self by regulating differentiation and homeostasis of Tregs and Teff cells.^26,27

The aim of the present study was to clarify the effect of the gluten-containing standard (STD) diet on the cytokine pattern in Foxp3⁻ and Foxp3⁺ T cells isolated from mucosal lymphoid tissues: Peyer's patches (PP), mesenteric lymph nodes (MLN) and pancreatic lymph nodes (PLN), and from non-mucosal lymphoid compartments: spleen (S) and inguinal lymph nodes (ILN), and to compare these results with similar experiments using mice receiving the diabetes-protective GF diet. The experiments were performed using fully immunocompetent BALB/c mice. Gluten-induced changes in the cytokine pattern of T cells may have an important impact on the development of autoimmune disease in susceptive individuals because the overall effect of an inflammatory response is determined by the balance between anti-inflammatory and pro-inflammatory cytokines.

Materials and methods

Animals/diets

Timed pregnant BALB/cA BomTac mice were purchased from Taconic Europe A/S, Ejby, Denmark. Mice were kept in specific pathogen-free animal facilities with free access to water and food, the temperature was 22 ± 2° and light was switched off from 18.00 to 06.00, with air changed 16 times per hour, and humidity maintained at 55 ± 10%. Mice with first-generation female offspring were divided into two groups 7 days after delivery. One group of mice received the gluten-containing STD altromin diet and the other group of mice received the GF modified altromin diet (Altromin, Lage, Germany), both diets have previously been used at the Bartholin Institute to study the effect of a GF diet on diabetes incidence in NOD mice.^1,28 First-generation female offspring (n = 6), in each dietary group, were used in the study when 6 weeks old. Both experimental diets were nutritionally adequate with a similar level of protein, amino acids, minerals, vitamins and trace elements, only the protein source differed between the two diets. The weight of the mice was followed and growth curves were comparable between the two groups of mice. The exact composition of the STD and the GF diets is given in earlier papers.^1,28 The animal experiments were carried out with approval from The National Animal Experimental Board, and experiments were performed in accordance with international guidelines for the care and use of laboratory animals.

Surface and intracellular cytokine staining and flow cytometry

Single-cell suspensions were prepared from S, ILN, MLN, PP and PLN from mice in each diet group. Cells from each organ were pooled because of the limited number of cells obtained from PLN. Cell suspensions were separated into two groups, one for stimulation with ionomycin (2 μg/ml) and PMA (50 ng/ml) for 4 hr in the presence of Golgi-stop, the rest of the cells remained unstimulated as controls. The cells were kept in a tissue culture incubator at 37°. Then cells were pre-incubated with Fc block CD16/CD32 (2.4G2, IgG2b; BD Biosciences, Franklin Lakes, NJ) and stained with peridinin chlorophyll protein-conjugated rat anti-mouse CD3 monoclonal antibody (17A2; IgG2b; BD Biosciences). Cell pellets were resuspended with fixation/permeabilization solution (Mouse regulatory T cell Staining kit, cat. no 88–8111; eBioscience, San Diego, CA) following the manufacturer's protocol. Fc block CD16/CD32 was added and the labelled antibodies were added to the cells, and stained for 30 min at 4°. We used phycoerythrin-conjugated rat anti-mouse Foxp3 monoclonal antibody (FJK-16s, IgG2a; eBioscience), Alexa Fluor 488-conjugated rat anti-mouse IL-2 (JES6-5H4, IgG2b; BD Biosciences), Alexa Fluor 488-conjugated rat anti-mouse IL-4 (11B11, IgG1; BD Biosciences), FITC-conjugated rat anti-mouse IL-10 (JES5-16E3, IgG2b), Alexa Fluor 488-conjugated rat anti-mouse IFN-γ (XMG1.2; IgG1κ; BD Biosciences), biotin-conjugated rat anti-mouse TGF-β (A75-3; IgG2a, κ; BD Biosciences), streptavidin (cat.no 349024; BD Biosciences) and Alexa Fluor 488-conjugated rat anti-mouse IL-17 (17B7; IgG2a,κ; eBioscience). Isotype control antibodies were used to determine the amount of non-specific binding, and propidium iodide was used to exclude dead cells. Cells were analysed by flow cytometry using a FACSscan (BD Biosciences), and data were analysed using cellquest software (BD Biosciences). Cytokine expression analysis was performed by gating according to the CD3 and Foxp3 parameters. The results of the study are from the PMA/ionomycin-stimulated cells because no, or very few, cytokine-positive cells were found in unstimulated controls.

Data analysis

Statistical analyses were performed using ncss 2004/gess statistical software, Kaysville, UT. Differences between the diets were determined using a one-sample t-test. We tested diet (STD versus GF) -induced differences between non-mucosal (S, ILN) and mucosal (MLN, PP, PLN) lymphoid compartments to determine whether dietary gluten affects the balance of pro-inflammatory versus anti-inflammatory cytokines in Foxp3⁺ Tregs as well as Foxp3⁻ T cells at systemic lymphoid organs or only locally; at mucosal lymphoid organs. Based on the concept of the common mucosal immune system²⁹ we investigated various mucosal lymphoid organs; MLN, PP, PLN. Differences at P < 0·05 were considered statistically significant. Only data showing substantial diet-induced differences are shown in the Results section. Graphic presentation of the results was performed using the Sigma Plot 9.0 (Systat Software, San Jose, CA) and graphpad prism version 5 software (Graphpad Software, La Jolla, CA).

Results

The proportion of IL-2⁺ T cells was increased in mice receiving the STD diet

The IL-2 expression within Foxp3⁺ T cells was significantly higher (P < 0·05) in mucosal lymphoid tissues (MLN, PP, PLN) in mice receiving the STD compared with the GF diet (Fig. 1). The proportion of IL-2⁺ Foxp3⁺ T cells was on average 21% higher in mucosal lymphatic tissue in mice receiving the STD diet than in mice on the GF diet. No diet-induced difference was found in non-mucosal lymphoid compartments (S, ILN). When evaluating the proportion of IL-2⁺ Foxp3⁻ T cells, we found an average of 17% more IL-2⁺ Foxp3⁻ T cells in mice on the STD diet, both in non-mucosal (S, ILN) and mucosal (MLN, PP, PLN) lymphoid compartments (P < 0·01).

Figure 1.

Intracellular staining and FACS analysis of interleukin-2 (IL-2)⁺ T cells. (a) Representative plot of lymphocyte and the subsequent CD3⁺ gate. (b) The bars represent the proportion of IL-2⁺ cells (gated on 100% Foxp3⁺CD3⁺) in mice receiving the standard (STD) diet compared with the gluten-free (GF) diet. We found significant (P < 0·05) differences between the diets in mucosal lymphoid tissues. Histogram from mesenteric lymph nodes (MLN), showing the proportion of IL-2⁺ Foxp3⁺ T cells in STD mice compared with GF mice. (c) Significant (P < 0·01) diet-induced changes in the proportion of IL-2⁺ Foxp3⁻ T cells were found, in all tested lymphoid compartments. Spleen (S); inguinal lymph nodes (ILN); MLN; Peyer's patches (PP) and pancreatic lymph nodes (PLN). Histograms from MLN showing differences in IL-2⁺ Foxp3⁻ T cells according to diet. Pools of lymphocytes from six mice in each group were used. Black bars: STD diet. Grey bars: GF diet. *P < 0·05. **P < 0·01.

The STD diet increases the proportion of IL-4⁺ Foxp3⁺ T cells both in non-mucosal lymphoid compartments and in mucosal lymphoid tissues

We found an overall (S, ILN, MLN, PP, PLN) increase (P < 0·01) in IL-4⁺ Foxp3⁺ T cells, in mice on the STD diet compared with mice on the GF diet (Fig. 2). The proportion of IL-4⁺ Foxp3⁺ T cells was on average 26% higher in mice receiving the STD compared with mice on the GF diet. No substantial differences between the diets were found in IL-4⁺ Foxp3^– T cells (data not shown).

Figure 2.

Intracellular staining and FACS analysis of interleukin-4 (IL-4)⁺ T cells. The bars represent the proportion of IL-4⁺ cells (gated on 100% Foxp3⁺ CD3⁺), in mice receiving the standard (STD) diet versus the gluten-free (GF) diet. We found a significantly (P < 0·01) higher proportion of IL-4⁺ Foxp3⁺ T cells in STD mice in all isolated lymphoid tissues. Histogram from pancreatic lymph nodes (PLN), is shown. The indicated values represent the percentage of IL-4⁺ Foxp3⁺ T cells. Spleen (S); inguinal lymph nodes (ILN); mesenteric lymph nodes (MLN); Peyer's patches (PP) and PLN. Pools of lymphocytes from six mice in each group were used. Black bars: STD diet. Grey bars: GF diet. **P < 0·01.

Increased proportion of IL-10⁺ non-T cells in spleens from mice on the GF diet

Intracellular staining for IL-10⁺ Foxp3⁺ T cells revealed that these cells were present both in non-mucosal and mucosal-associated lymphoid tissue, varying from 5·7% in S to 16·8% in PP (gated on 100% Foxp3⁺ CD3⁺) (Fig. 3). Diet-induced differences in this cell population were not found. Moving the gate to non-T cells (CD3⁻ cells) we found no significant differences in IL-10⁺ CD3⁻ cells, according to diet, when comparing any difference between the mucosal lymphoid tissues (MLN, PP, PLN) and the non-mucosal lymphoid compartments (S, ILN). Only a marked difference between the two diets in IL-10⁺ CD3⁻ cells was observed in S. Mice on the STD diet were found to have a low proportion (0·7%) of IL-10⁺ CD3⁻ cells, whereas mice on the GF diet had a > 100% increase in these cells (1·43%). We found no diet induced differences in IL-10⁺ Foxp3⁻ T cells (data not shown).

Figure 3.

Intracellular staining and FACS analysis of interleukin-10 (IL-10)⁺ cells. (a) The proportion of IL-10⁺ Foxp3⁺ T cells in mice on the standard (STD) versus the gluten-free (GF) diet. No significant diet-induced changes were found in this cell population. (b) When gating on 100% CD3⁻ cells, we found no significant diet-induced changes when comparing non-mucosal lymphoid tissues with mucosal-associated lymphoid tissues. Although a higher proportion of IL-10⁺ non-T cells was observed in spleen (S). (c) Dot plots showing the CD3⁻ gate and the proportion of IL-10⁺ CD3⁻ cells isolated from S in mice receiving the STD diet versus the GF diet. Spleen; inguinal lymph nodes (ILN); mesenteric lymph nodes (MLN); Peyer's patches (PP) and pancreatic lymph nodes (PLN). Pools of lymphocytes from six mice in each group were used. Black bars: STD diet. Grey bars: GF diet.

Higher proportion of TGF-β⁺ T cells in mice on the GF diet

The proportion of TGF-β⁺ Foxp3⁺ T cells was between 3·75%, in ILN in mice on the STD diet and 8·07% in PLN in the GF mice (Fig. 4). Diet-induced differences were not observed in either S, ILN or MLN. However, mice on the GF diet had a tendency towards a higher proportion of TGF-β⁺ Foxp3⁺ T cells in PP (5·2% in STD versus 8·1% in GF) and in PLN (4·0% in STD versus 8·1% in GF). Interestingly, we found a consistently higher proportion (P < 0·001) of TGF-β⁺ Foxp3⁻ T cells in mice receiving the GF diet. The GF mice had more than twice as many TGF-β⁺ Foxp3⁻ T cells in all isolated lymphoid organs (S, ILN, MLN, PP, PLN). The highest difference between the diets was found in PP (0·3% versus 0·7%) in mice fed the STD and GF diets.

Figure 4.

Intracellular staining and FACS analysis of transforming growth factor-β (TGF-β)⁺ T cells. (a) Bars show the proportion of TGF-β⁺ Foxp3⁺ T cells. Mice on the gluten-free (GF) diet have a tendency towards a higher proportion of TGF-β⁺ Foxp3⁺ T cells in Peyer's patches (PP) and pancreatic lymph nodes (PLN). Histograms show the diet-induced differences in PLN. (b) The distribution of TGF-β⁺ Foxp3⁻ T cells. A significantly (P < 0·001) higher proportion was found in all lymphoid organs isolated from mice on the GF diet. Histogram showing diet-induced differences in PLN. Spleen (S); inguinal lymph nodes (ILN); mesenteric lymph nodes (MLN); PP and PLN. Pools of lymphocytes from six mice in each group were used. Black bars: STD diet. Grey bars: GF diet. ***P < 0·001.

Marked increase in IFN-γ⁺ cells in mice on the STD diet

The proportion of IFN-γ⁺ Foxp3⁺ T cells was increased (P < 0·01) in mice receiving the STD diet compared with the GF diet, in all isolated lymphoid organs (S, ILN, MLN, PP, PLN) (Fig. 5). The diet-induced differences were highest in MLN and PLN. In PLN from mice on the STD we found 12·18% cells being IFN-γ⁺ compared with 6·98% IFN-γ⁺ cells in the PLN isolated from GF mice, when gated on 100% Foxp3⁺ CD3⁺. In Foxp3⁻ T cells, we also found an increase in IFN-γ⁺ cells in STD mice (P < 0·01). The average proportion of IFN-γ⁺ Foxp3⁻ T cells in all lymphoid organs was 36% higher in STD mice than in the GF mice. Furthermore, we investigated differences in IFN-γ expression within non-T cells. The MLN and PLN had a tendency towards a higher proportion of IFN-γ⁺ non-T cells, although the higher proportion was not significant when calculating on an overall mucosal (MLN, PP, PLN) difference. In MLN, mice on the STD diet had 2·35% IFNγ⁺ CD3⁻ cells compared with 1·7% in mice receiving the GF diet. The same pattern was found in PLN with 2·67% IFN-γ⁺ CD3⁻ cells in STD mice and 1·1% in GF mice.

Figure 5.

Intracellular staining and FACS analysis of interferon-γ (IFN-γ)⁺ T cells. (a) Bars show the proportion of IFN-γ⁺ Foxp3⁺ T cells with a significant (P < 0·01) increase in mice on the standard (STD) diet compared with the gluten-free (GF) diet in all isolated lymphoid tissues. Dot plot of isolated cells from pancreatic lymph nodes (PLN). (b) Bars show a diet-induced increase (P < 0·01) in IFN-γ⁺ Foxp3⁻ T cells in all tested lymphoid organs. Dot plots of IFN-γ⁺ Foxp3⁻ T cells isolated from PLN. (c) Bars represent the proportion of IFN-γ⁺ non-T cells. Dot plots of IFN-γ⁺ non-T cells in PLN. Spleen (S); inguinal lymph nodes (ILN); mesenteric lymph nodes (MLN); Peyer's patches (PP) and PLN. Pools of lymphocytes from six mice in each group were used. Black bars: STD diet. Grey bars: GF diet. **P < 0·01.

Increased proportion of IL-17⁺ T cells in mice on the STD diet

The gluten-containing STD diet significantly increased (P < 0·05) the proportion of IL-17⁺ CD3⁺ cells in both non-mucosal and mucosal lymphoid compartments (S, ILN, MLN, PP, PLN) (Fig. 6). The overall increase of mean was 60% between the diets, with the highest diet-induced difference in PP and PLN. Surprisingly, we found IL-17⁺ Foxp3⁺ T cells in all organs. The lowest proportion was found in PP in mice on the GF diet, with 2·3% IL-17⁺ Foxp3⁺ T cells and the highest proportion in MLN in mice receiving the STD diet with 7·9% IL-17⁺ Foxp3⁺ T cells (gated on 100% Foxp3⁺ CD3⁺ cells). Gluten induced a significant (P < 0·05) increase in IL-17⁺ Foxp3⁺ T cells in mucosal lymphoid tissues (MLN, PP, PLN), but no significant difference between the diets was found in non-mucosal lymphoid compartments (S, ILN). The proportion of IL-17⁺ Foxp3⁺ T cells was on average 39% higher in mucosal lymphoid tissue in mice receiving the STD diet than in mice on the GF diet. When analysing the effect of the two diets we found no substantial diet-induced differences in IL-17⁺ Foxp3⁻ T cells (data not shown).

Figure 6.

Intracellular staining and FACS analysis of interleukin-17 (IL-17)⁺ T cells. (a) Increased (P < 0·05) proportion of IL-17⁺ T cells in mice receiving the standard (STD) diet compared with the gluten-free (GF) diet in all tested lymphoid tissues. (b) Dot plots of isolated IL-17⁺ T cells from spleen (S) and pancreatic lymph nodes (PLN). (c) The proportion of IL-17⁺ Foxp3⁺ T cells. With a mucosal increase (P < 0·05) in IL-17⁺ Foxp3⁺ T cells in mice receiving the STD diet. (d) Histograms from S and PLN showing the proportion of IL-17⁺ Foxp3⁺ T cells. Spleen; inguinal lymph nodes (ILN); mesenteric lymph nodes (MLN); Peyer's patches (PP) and PLN. Pools of lymphocytes from six mice in each group were used. Black bars: STD diet. Grey bars: GF diet. *P < 0·05.

Dietary gluten alters the balance of pro-inflammatory and anti-inflammatory cytokines in T cells

The gluten-containing STD diet induces a pro-inflammatory (IL-17, IFN-γ) cytokine pattern in Foxp3⁻ T cells in all lymphoid tissues tested, and a decreased level of anti-inflammatory cytokines (IL-10, TGF-β) in ILN, MLN, PP, PLN (Fig. 7). Furthermore, dietary gluten induces a pro-inflammatory (IL-17, IFN-γ) cytokine pattern in Foxp3⁺ T cells and an increased level of IL-2 and IL-4 in all lymphoid tissues tested, and a decreased level in anti-inflammatory cytokines (IL-10, TGF-β) in S, PP and PLN.

Figure 7.

The overall gluten-induced changes in the cytokine pattern in Foxp3⁻ and Foxp3⁺ T cells. (a) Foxp3⁻ T cells. Ratio of cytokine positive Foxp3⁻ T cells, from mice on the standard (STD) diet relative to the gluten-free (GF) diet. Dietary gluten induces a pro-inflammatory [interleukin-17 (IL-17), interferon-γ (IFN-γ)] cytokine pattern in Foxp3⁻ T cells in all lymphoid tissues tested, and a deceased level of anti-inflammatory [IL-10, transforming growth factor-β (TGF-β)] in inguinal lymph node (ILN), mesenteric lymph node (MLN), Peyer's patches (PP), and pancreatic lymph node (PLN). (b) Foxp3⁺ T cells. Ratio of cytokine-positive Foxp3⁺ T cells, from mice on the STD diet relative to the GF diet. Dietary gluten induces a pro-inflammatory (IL-17, IFN-γ) cytokine pattern in Foxp3⁺ T cells and an increased level of IL-2 and IL-4 in all lymphoid tissues tested, and a decreased level in anti-inflammatory (IL-10, TGF-β) in spleen (S), Peyer's patch (PP) and pancreatic lymph nodes (PLN). Inguinal lymph nodes (ILN).

Discussion

The present study shows that gluten intake changes the cytokine pattern in Foxp3⁺ and Foxp3⁻ T cells towards a pro-inflammatory cytokine profile. The protective role of a GF diet¹ is not associated with changes in the number of the Foxp3⁺ Tregs³⁰ but could be the result of an increase in the expression of immunosuppressive cytokines such as TGF-β and, to a lesser extent, IL-10. Conversely, the STD, gluten-containing diet increased the level of potent pro-inflammatory (IL-17, IFN-γ) cytokines, as well as of IL-2 and IL-4.

The exact mechanism by which dietary gluten influences cytokine profiles in Foxp3⁺ and Foxp3⁻ T cells is not known. We have previously documented that the diabetes-preventive GF diet decreases the number of caecal bacteria and is associated with quantitative changes in the composition of the bacterial flora in NOD mice.³¹ Whether the gluten-induced changes in cytokine profiles are mediated directly or via intestinal microflora have not yet been clarified by gnotobiotic germ-free experiments.

In this study, based on the concept of the common mucosal immune system²⁹ we clarify how dietary gluten influences cytokine pattern of Foxp3⁺ and Foxp3⁻ T cells in mucosal and non-mucosal lymphoid compartments. PLNs, the draining lymph nodes of pancreas, are grouped into the mucosal compartment, because mucosa-associated, β₇-integrin^high lymphocytes accumulate in early phases of islet inflammation in the pancreas of NOD mice.¹⁰ The mucosal character of PLNs is also supported by recent findings³² documenting that B cells in pancreatic lymph nodes of NOD mice express integrin α₄/β₇ and that the integrin α₄β₇/MAdCAM-1 mucosal adhesion pathway is important for their migration to PLN.

Our data provide new insight into the effect of dietary gluten on immune reactivity and have implications for understanding the observed dietary modification of T1D development in NOD mice.¹ Because NOD mice are lymphophenic and display other immune dysbalances, we studied the effect of dietary gluten in fully immunocompetent BALB/c mice. Hence, we observed which immunological changes gluten exposure induces in a healthy animal model at 6 weeks. The age of the animals was chosen to reflect a ‘pre-diabetic time-point’ to allow later comparison with pre-diabetic NOD mice. Our results shed more light on immune status associated with exposure to dietary gluten, that has been reported as an environmental factor in development of T1D, both in animal models^7,8,15 and in humans.⁹ The results are important because they illustrate how a dietary protein directly changes the balance of pro-inflammatory and anti-inflammatory cytokines in T cells.

Multiple studies, both in humans and in mice, propose a deficiency in either the function or the frequency of Tregs,^33,34 as an explanation of development of autoimmunity. Hence it has been tested whether a reduced number of CD4⁺ CD25⁺ T cells are found in patients with T1D compared with controls. The first human study indicated that this was the case,³⁵ but subsequent studies failed to repeat these findings.^36,37 CD4⁺ CD25⁺ T cells from T1D patients have a reduced capacity to suppress proliferation of co-cultured CD4⁺ CD25⁻ T cells, compared with controls.^38,39 This suggests a defective suppressor function in CD4⁺ CD25⁺ T cells in patients with T1D and could be a result of changes in the cytokine pattern of the Tregs, as CD4⁺ CD25⁺/CD4⁺ CD25⁻ co-cultures from T1D patients produced more IFN-γ, and less IL-10 than healthy controls.³⁸ These findings are similar to our present findings in animals receiving the STD diet, supporting the notion that functional changes in Tregs could be involved in the susceptibility for the development of T1D.

The regulatory cytokines IL-10 and TGF-β are of major importance for mediating the suppressive activity of Tregs.⁴⁰ The importance of IL-10 is shown in studies where the dampening of autoaggressive CD8 cells is associated with production of IL-4 and IL-10 by Tregs.⁴¹ This is contrary to studies showing that the regulatory function of IL-10-secreting islet-specific Tregs clones, seems to be cell–cell contact-dependent and independent of IL-10, as blocking of IL-10 or TGF-β had no effect on the ability of these cells to work as suppressors.⁴² Other studies suggest an important suppressive role for TGF-β, based on the association between a reduction in suppression by Tregs with blockade of TGF-β^43,44 and the fact that TGF-β has a broad inhibitory effect on the entire immune system. We observed that the GF diet significantly increases the level of TGF-β in Foxp3⁻ T cells, as well as a tendency towards a mucosal increase of TGF-β in Foxp3⁺ T cells. Moreover, Foxp3⁺ T cells in all lymphoid organs produced IL-10, although we found no diet-induced differences in these cells. The proportion of IL-10⁺ Foxp3⁺ T cells was highest in MLN and PP, as previously described.²⁷ The GF diet increased the proportion of splenic IL-10⁺ non-T cells.

In our recent study we have documented that the gluten-containing STD diet does not change the proportion of Foxp3⁺ T cells in either mucosal or non-mucosal lymphoid organs of BALB/c mice.³⁰ In this present study we documented cytokine changes, which could reflect diet-induced functional changes within the Foxp3⁺ T cells. These findings may suggest that impaired regulatory function of Foxp3⁺ Tregs rather than their absolute number is of importance in progression to autoimmunity (T1D). The importance of gluten exposure combined with a defect in Foxp3⁺ T cells is in line with recently published data showing that sensitization to gliadin and partial systemic depletion of Foxp3⁺ T cells induces insulitis in the non-obese diabetic-DQ8 mice, that do not spontaneously develop diabetes.⁴⁵

During the last years, the Th1/Th2 paradigm has expanded, following the discovery of a third subset of Th cells that produce IL-17 (Th17).⁴⁶ They differentiate in response to TGF-β and IL-6, and as potent inducers of tissue inflammation they are associated with the pathogenesis of various human inflammatory conditions and autoimmune diseases,²⁵ including T1D.⁴⁷ The Th17 cells seem more resistant to regulatory control than Th1 and Th2 cells.⁴⁸ In this study we found an overall higher proportion of IL-17⁺ CD3⁺ cells in mice receiving the gluten-containing STD diet, with a surprising increase in mucosal-associated lymphoid tissue (PP and PLN), where the IL-17⁺ CD3⁺ cell population more than doubles. These data are in accordance with our recent study documenting a significantly increased proportion of CD4⁺ IL-17⁺ (Th17) cells in the PLN of BALB/c mice fed an STD diet.³⁰

The detection of IL-17⁺ Foxp3⁺ T cells is in line with recently published data. It has been reported that human peripheral blood and tonsil contain Foxp3⁺ Tregs with the capacity to produce IL-17 upon activation. These cells are possibly generated at mucosal sites during an inflammatory process.²³ Furthermore, the presence of Foxp3⁺ RORγ⁺ (a transcription factor involved in the development of Th17 cells) T cells with the ability to produce IL-17 in the small intestinal lamina propria has been described.²⁴ We found IL-17⁺ Foxp3⁺ T cells in proportions ranging from 2·3% cells in PP to 7·9% in MLN. Interestingly, we found that the proportion of IL-17⁺ Foxp3⁺ T cells was affected by gluten intake, as the STD diet increased the proportion of IL-17⁺ Foxp3⁺ T cells in mucosal lymphoid tissues (MLN, PP, PLN). Hence, gluten intake induces a functional change in Foxp3⁺ Tregs, towards production of highly pro-inflammatory IL-17.

Some Foxp3⁺ T cells analysed were positive for pro-inflammatory cytokines, which supports recent evidence suggesting considerable plasticity of Foxp3⁺ Tregs with respect to their capacity to produce cytokines, such as IL-17²⁴ and IFN-γ.⁴⁹ This is consistent with the notion that Foxp3 may be expressed more widely and transiently than initially conceived.^24,50 Contributing to the diverse picture is the fact that human, activated T cells can express Foxp3 although lacking regulatory function,^51,52 and Foxp3 induction ex vivo does not confer a regulatory phenotype.⁵³ Moreover, several recent studies suggest that there is a loss of Foxp3 expression during inflammatory conditions, and that low Foxp3 expression predisposes for autoimmune diseases in mice susceptible for disease.⁵⁴ Therefore it seems that not all Foxp3⁺ T cells possess a regulatory function, either because of a lower amount, or through unstable Foxp3 expression. It is suggested that some Tregs, under specific inflammatory conditions, become ‘effector-like’ T cells, likely to play an important role in autoimmune diseases.²² This explanation could apply for our findings of gluten-induced pro-inflammatory cytokines in Foxp3⁺ T cells, changes that could affect the ability of Tregs to suppress autologous effector T cells.

The site of Tregs suppression in vivo is not limited to lymphoid organs, and the ability of Tregs to migrate and exert their function in inflamed tissue is important for their function in vivo.⁵⁵ Studies in murine models of T1D indicate that Tregs exert their function within the target organ undergoing autoimmune attack, and also in the associated, draining lymph nodes, so it could be relevant to test whether the observed differences in T cells are also present in pancreatic tissue and in intestinal lamina propria. Furthermore, it is important to notice that the molecular mechanism of suppression used by Tregs seems to differ depending on being present in either lymphoid tissue or non-lymphoid tissue.²⁰

In summary, our study shows that the STD, gluten-containing diet changes the cytokine pattern in both Foxp3⁻ and Foxp3⁺ T cells towards a more pro-inflammatory cytokine profile with higher levels of IL-17, IFN-γ, IL-2 and IL-4. Inversely, a diabetes-protective, gluten-free diet induces an anti-inflammatory cytokine profile with higher IL-10 production by non-T cells in S and a higher proportion of Foxp3⁻ T cells (and a tendency of mucosal increase in Foxp3⁺ T cells) secreting TGF-β.

We suggest that a failure in the development of a proper regulatory immune response to environmental stimuli may contribute to development of T1D in genetically predisposed individuals. Certain environmental stimuli, such as gluten, induce inflammatory changes in T cells that negatively influence mucosal immune homeostasis, which could affect the diabetogenic process in the pancreas towards development of autoimmunity.

Disclosure

The authors declare no conflicts of interest.