Gluten-free diet affects faecal small non-coding RNA profiles and microbiome composition in celiac disease supporting a host-gut microbiota crosstalk

The current treatment for Celiac Disease (CD) is adhering to a gluten-free diet (GFD), although its long-term molecular effects are still undescribed. New molecular features detectable in faecal samples may improve and facilitate non-invasive clinical management of CD on GFD. For this purpose, faecal small non-coding RNAs (sncRNAs) and gut microbiome profiles were concomitantly explored in CD subjects in relation to strict (or not) GFD adherence over time. In the present observational study, we performed small RNA and shotgun metagenomic sequencing in stool from 63 treated CD (tCD) subjects and 66 sex- and age-matched healthy controls. tCD included 51 individuals on strict GFD and with negative transglutaminase (TG) serology (tCD-TG-) and 12 symptomatic with not strict/short-time of GFD adherence and positive TG serology (tCD-TG+). Samples from additional 40 adult healthy individuals and from a cohort of 19 untreated paediatric CD subjects and 19 sex/age matched controls were analyzed to further test the outcomes. Several miRNA, other sncRNA (piRNA and tRNA) and microbiota profiles were altered in tCD subjects(adj.p<0.05). Findings were validated in one external group of controls. In tCD-TG-, GFD duration correlated with five miRNA levels (p<0.05): for miR-4533-3p and miR-2681-3p, the longer the diet adherence, the less the expression differed from controls. tCD-TG+ and untreated paediatric CD patients showed a similar miRNA dysregulation. Immune-response, trans-membrane transport and cell death pathways were enriched in targets of identified miRNAs. Bifidobacterium longum, Ruminococcus bicirculans and Haemophilus parainfluenzae abundances shifted (adj. p<0.05) with a progressive reduction of denitrification pathways with GFD length. Integrative analysis highlighted 121 miRNA-bacterial relationships (adj.p<0.05). Specific faecal sncRNA and microbial patterns characterise CD subjects on GFD, reflecting either the long-term effects or the gut inflammatory status, in case of a not strict/short-time adherence. Our findings suggest novel host-microbial interplays and could help the discovery of biomarkers for the clinical monitoring of GFD over time. Hundred and thirty-two subjects were enrolled by the Italian Institute for Genomic Medicine (IIGM), the Gastroenterology Unit of Ospedale Mauriziano Umberto I and the Gastroenterology outpatient clinic of San Giovanni Antica Sede, all in Turin, Italy. The cohort included tCD (n=63), untreated CD (i.e., recruited before starting the GFD; n=3) and healthy volunteers with no dietary restrictions (n=66). Inclusion criteria for CD subjects were: disease diagnosis with histological confirmation; for all: no use of antibiotics and other drugs in the month before sampling, age >18 and no concomitant gastrointestinal disorders (e.g., allergies/intolerances, diabetes, inflammatory bowel diseases, or tumours). The TG IgA levels were measured in serum samples of all subjects who provided blood for the present study, unless not already clinically tested in concomitance with the recruitment. All participants signed informed consent to participate in the study and received two questionnaires from the European Prospective Investigation into Cancer and Nutrition (EPIC) study (one on dietary habits and another, sex-specific, about lifestyle habits) and a disposable container for stool collection. Small RNA sequencing and shotgun metagenomic sequencing was performed on all fecal samples collected.