Nowadays it is assumed that an innate immunity to gluten plays a key role in the development of coeliac disease (CD).1 This innate response, mediated by interleukin (IL) 15 and elicited by “toxic peptides”, like the 19‐mer, through a DQ2‐independent mechanism, induces epithelial stress and reprogrammes intraepithelial lymphocytes into natural killer (NK)‐like cells2 leading to enterocyte apoptosis and an increase in epithelium permeability. Thus, immunodominant peptides, like the 33‐mer, can reach the lamina propria to trigger adaptive immunity. However, although an innate specific response in CD has been reported,3 no differential factors between patients with and without CD have been described controlling the innate immune response. Thus, since the toxic 19‐mer elicits its harmful effect through a DQ2‐independent mechanism, we hypothesise that the innate response is common in patients with and without CD, whereas the adaptive response is exclusive of susceptible patients with CD.
To test the hypothesis, biopsy cultures were taken from at least three patients with CD who are on a gluten‐free diet (GFD) and three patients without CD (table 11).). Biopsy specimens were challenged with crude gliadin and the gliadin synthetic 19‐mer and deaminated 33‐mer peptides after discarding the presence of lipopolysaccharide in all the cases. This was carried out at 100 μg/ml for only 3 h to imitate what are considered the normal timing and concentration in the gut after a normal meal. All biopsy specimens were then washed and cultured for another 21 h in new clean culture medium to determine whether an innate stimulus is reflected by an adaptive response. Each sample cultured in basal medium constituted an internal control. Innate immune mediators IL15 and nitrites were determined by western blot in the biopsy protein extract and by a Griess reagent system in the 3 h supernatants respectively. mRNA levels of adaptive immunity mediators like signal transducers and activators of transcription (STAT) 1, STAT3, tumour necrosis factor α, interferon (IFN) γ, IL23 (p19), IL27 (p28) and IL12 (p35) were determined by real‐time polymerase chain reaction using β actine levels as housekeeping.
All patients with and without CD on GFD who were challenged with the gliadin solution produced IL15 when compared with the basal culture (fig 1A1A).). Moreover, the IL15‐mediated response in patients without CD was also triggered by the toxic 19‐mer gliadin peptide (three of six) and, especially, by the 33‐mer gliadin peptide (five of six). Importantly, none of the basal cultures produced this cytokine and, although not expected, the “non‐toxic” immunodominant 33‐mer was also able to induce an innate response. Interestingly, this IL15 response was also confirmed by western blot in the supernantant of one GFD patient with CD and three patients without CD, who were on GFD (fig 1B1B),), therefore, discarding an intracellular and non‐biologically active IL15 response in patients without CD. We also found an increase in nitrite in the gliadin‐challenged patients with CD who were on a GFD, although not in patients without CD. In a similar way, as expected, after the biopsy mRNA isolation, adaptive mediators (STAT1, STAT3, IFNγ) were only modified in GFD patients with CD. Finally, basal GFD‐CD samples showed an 80‐fold increase in IFNγ mRNA levels compared with non‐CD basal samples (p value 0.002) and a slightly higher production of nitrites (p value 0.052).
We consider that, to our knowledge, this is the first time that an IL15‐mediated innate response to gliadin and gliadin peptides is described in individuals without CD, as well as an IL15‐mediated innate response to the “non‐toxic” deaminated immunodominant 33‐mer peptide.
The data obtained in this pilot study support the hypothesis that gluten elicits its harmful effect, throughout an IL15 innate immune response, on all the individuals. This innate response is found in both patients with and without CD, although the triggering of an adaptive response is CD specific. We propose that somehow patients with CD need to be DQ2 and also have a lower threshold for triggering an adaptive TH1 response. This lower threshold could be mediated by the higher basal levels of immune mediators, like IFNγ mRNA, found in patients with CD, a defect in the CD permeability or even a higher IL15‐sensitive response under the same stimulus, which might be mediated by a higher density of IL15 receptor in patients with CD.4