Promotion of autoimmune diabetes by cereal diet in the presence or absence of microbes associated with gut immune activation, regulatory imbalance, and altered cathelicidin antimicrobial Peptide

Abstract

We are exposed to millions of microbial and dietary antigens via the gastrointestinal tract, which likely play a key role in type 1 diabetes (T1D). We differentiated the effects of these two major environmental factors on gut immunity and T1D. Diabetes-prone BioBreeding (BBdp) rats were housed in specific pathogen-free (SPF) or germ-free (GF) conditions and weaned onto diabetes-promoting cereal diets or a protective low-antigen hydrolyzed casein (HC) diet, and T1D incidence was monitored. Fecal microbiota 16S rRNA genes, immune cell distribution, and gene expression in the jejunum were analyzed. T1D was highest in cereal-SPF (65%) and cereal-GF rats (53%) but inhibited and delayed in HC-fed counterparts. Nearly all HC-GF rats remained diabetes-free, whereas HC-fed SPF rats were less protected (7 vs. 29%). Bacterial communities differed in SPF rats fed cereal compared with HC. Cereal-SPF rats displayed increased gut CD3(+) and CD8α(+) lymphocytes, ratio of Ifng to Il4 mRNA, and Lck expression, indicating T-cell activation. The ratio of CD3(+) T cells expressing the Treg marker Foxp3(+) was highest in HC-GF and lowest in cereal-SPF rats. Resident CD163(+) M2 macrophages were increased in HC-protected rats. The cathelicidin antimicrobial peptide (Camp) gene was upregulated in the jejunum of HC diet-protected rats, and CAMP(+) cells colocalized with CD163. A cereal diet was a stronger promoter of T1D than gut microbes in association with impaired gut immune homeostasis.

FIG. 1.

Feeding a cereal diet promotes T1D in BBdp rats in the presence or absence of microbes—diabetes incidence, islets, and bacterial communities. BBdp rats were maintained under SPF or GF conditions, fed an HC diet or cereal diet, and monitored for T1D development. (A) Kaplan-Meier plot shows percentages of nondiabetic rats: HC-GF group (n = 27), HC-SPF group (n = 48), cereal-GF group (n = 17), and cereal-SPF group (n = 48). (B) Islet number and β-cell mass are shown in asymptomatic 130-day BBdp rats fed an HC diet (open bars) or cereal diet (hatched bars) and housed under GF or SPF conditions. Data represent mean ± SD; n = 5–7/group. P values obtained using ANOVA, followed by least significant difference post hoc test. (C) Nonmetric multidimensional scaling diagram shows microbial community distribution in BBc rats fed cereal (open triangles), BBdp rats fed cereal (filled circles), or HC (open circles). The figure shows the relative statistical differences between bacterial communities derived from each sample on an arbitrary two-dimensional surface. Similar community profiles are closer together in two-dimensional space. (A high-quality color representation of this figure is available in the online issue.)

FIG. 2.

Influence of diet and microbes on gut T cells. BBdp rats were maintained under SPF or GF conditions and fed an HC diet (open bars) or cereal diet (hatched bars). (A) Representative image displays CD3⁺ (arrowheads) IEL and LPL in BBdp jejunum (bar = 20 μm); number of CD3⁺ LPL/mm² mucosa and CD3⁺ IEL/100 epithelial cells in 130-day asymptomatic BBdp rat jejunum. (B) Representative image displays CD8α⁺ cells (arrowheads) in BBdp jejunum (bar = 20 μm); number of CD8α⁺ LPL/mm² mucosa and CD8α⁺ IEL/100 epithelial cells. (C) Representative image displays CD4⁺ cells (arrowheads) (bar = 20 μm); number of CD4⁺ LPL/mm² mucosa and CD4⁺ IEL/100 epithelial cells. Data represent mean ± SD; n = 5–7 per group. P values obtained using ANOVA, followed by least significant differences post hoc test. (A high-quality color representation of this figure is available in the online issue.)

FIG. 3.

Regulatory Foxp3⁺ T cells. (A) Representative image displaying Foxp3⁺ cells (arrowheads) in BBdp jejunum (bar = 20 μm). (B) Number of total Foxp3⁺ cells/mm² mucosa. Number of Foxp3⁺ cells/mm² mucosa normalized to the number of CD3⁺ (C) and normalized to CD4⁺ LPL/mm² mucosa in BBdp rat jejunum (D). Data represent mean ± SD; n = 7–9 rats per group. P values obtained using ANOVA, followed by least significant differences post hoc test. (A high-quality color representation of this figure is available in the online issue.)

FIG. 4.

Cytokine gene expression. Frozen jejunum samples were obtained from BBdp rats maintained under SPF or GF conditions and fed an HC diet (open bars) or a cereal diet (hatched bars). RNA was isolated, and gene expression analyses were performed by using reverse transcription, followed by quantitative RT-PCR. The ΔC_T value was obtained by subtracting the Ct value of β-actin from the gene of interest, and the triplicate ΔC_T values were averaged for each animal. The ΔC_T value of the standard Wistar Furth (WF) rat was subtracted from each animal to obtain the ΔΔC_T value. Results are shown as relative amounts 2^-ΔΔCT using an age-matched cereal-fed WF rat as a standard; genes of interest were normalized to expression of β-actin. The average value obtained from the HC-GF group was standardized to 1 and averages from other groups were standardized by the same factor. Data represent mean ± SEM (boxes) ± SD (whiskers); n = 5–7 per group. P values obtained using ANOVA, followed by least significant differences post hoc test.

FIG. 5.

CD163⁺ M2 macrophages in BBdp jejunum. BBdp rats were maintained under SPF or GF conditions and fed an HC diet (open bars) or cereal diet (hatched bars). (A) Representative image displays CD163⁺ cells (arrowheads) in BBdp jejunum (bar = 20 μm). (B) Number of CD163⁺ cells/mm² mucosa in jejunum of 130-day control BBc, 130-day asymptomatic BBdp, and ∼100–130-day overt diabetic BBdp rats, n = 6–8. (C) Number of CD163⁺ cells/mm² mucosa in 130-day asymptomatic BBdp rat jejunum. Data represent mean ± SD; n = 7–9 rats per group. Gene expression of Cd163 (D) and Hmox1 (E) was analyzed using quantitative RT-PCR. Data represent mean ± SEM (boxes) ± SD (whiskers); n = 5–7 per group. (B–E) P values obtained using ANOVA, followed by least significant differences post hoc test. (F) Image of CD163⁺ cells (arrowheads) in gut biopsy from a non-T1D control subject (10-year-old male) with gastroesophageal reflux disease (bar = 25 μm). (G) Number of CD163⁺ cells/mm² mucosal area in the jejunum of control subjects (filled blue circles; n = 14) or patients with T1D (red triangles; n = 8). Data represent mean ± SD. P value obtained using Mann-Whitney U test. (A high-quality color representation of this figure is available in the online issue.)

FIG. 6.

Gene expression profiling of jejunum identifies Camp upregulation in HC diet-protected rats associated with CD163⁺ M2 macrophages. BBdp rats were maintained under SPF conditions and fed an HC diet or a cereal diet and screened for a panel of innate and adaptive immune factors by PCR array analysis. (A) Results presented as a volcano plot of gene expression in HC-fed rats relative to cereal-fed rats; circled candidates were HC downregulated (green circle) or HC upregulated (red circles) at least twofold, with P values < 0.05. P values were obtained using the Student t test. (B) Highlighted significant results from the screen indicating Camp and Il1f6 as significant HC upregulated genes; Lck as a significant HC downregulated gene. (C) Representative images (left panels) display CAMP⁺ cells (arrowheads) in lamina propria and epithelium of 130-day asymptomatic BBdp jejunum (bar = 20 µm), followed by a double immunofluorescence confocal microscopy image displaying CAMP⁺ cell (red arrowhead) adjacent to CD68⁺ macrophage (green arrowhead) in lamina propria of a 130-day asymptomatic BBdp rat (bar = 5 µm). CAMP⁺ cells (arrowheads) in sterile 130-day GF adult BBdp jejunum and BBdp embryonic gut (right panels). (D) Confocal microscopy image displays multiple CAMP⁺ cells (Cy3/magenta arrowhead) in the lamina propria of 130-day asymptomatic GF BBdp jejunum colocalizing with both CD14 (Alexa 488/green arrowhead) and CD163 (Cy5/red arrowhead); nuclei labeled with Hoechst (bars = 5 µm). The yellow arrowheads designate CAMP⁺CD14⁺CD163⁺ cells. (A high-quality color representation of this figure is available in the online issue.)

FIG. 7.

Enteropathy in BBdp rat is not attributable to diabetes risk MHC or Gimap5 mutation. Jejunum sections were from 60–80-day non–diabetes-prone congenic rats (Wistar Furth, BBc, ACI.1^u, and ACI.1^u.lyp/lyp) and BBdp rats fed cereal diets. Hematoxylin and eosin–stained gut sections displayed crypt hyperplasia, villus atrophy (flattened, shorter), and immune cell infiltration in BBdp rats but not in the non–diabetes-prone rats. All the non–diabetes-prone rats have the diabetes risk MHC, and the ACI.1^u.lyp/lyp rats also have the Gimap5 (lymphopenia) gene but none displays enteropathy, suggesting these genes are not involved in enteropathy and gut inflammation.(A high-quality color representation of this figure is available in the online issue.)