Genetic and environmental influences contribute to Crohn’s disease (CD) development. However, the relationship between these two factors is unclear, limiting insights into CD pathophysiology, as well as interventions to prevent disease.

To address this knowledge gap, we performed a longitudinal study of entire multiplex families to identify relatives discordant between their genetic risk for disease, as calculated by an IBD polygenic risk score (PRS), and their CD status. Fecal specimens were collected quarterly (up to 32 months) for assessment of intestinal inflammation by fecal calprotectin assays and gut microbiome composition by 16S sequencing. We evaluated 8 candidate IBD PRS, and the best performing score across all families was used to rank each subject by genetic risk. Untargeted fecal metabolomic analysis was performed on a single specimen chosen from each of the 5 subjects that best represented the following categories: CD+/high PRS, CD-/high PRS, CD+/low PRS, and CD-/low PRS. To avoid potential confounders, we excluded the following specimens: elevated calprotectin, collected < 3 months since antibiotic use, or from pediatric subjects, who may still develop CD. From these 20 samples, 12 fecal specimens (3 from each category) were each gavaged into 5 germ-free IL-10 deficient mice (male, SvEv129, 7.5-12 weeks old). As a control, 4 other mice were gavaged with a fecal slurry from age- and sex-matched Murine Pathogen Free 129S6/SvEvTac mice. Mice were maintained in isolator cages with sterile food, water, and bedding. A score evaluating weight loss, pellet consistency, and fecal blood was used to measure disease activity over time.

We studied 396 fecal specimens from 52 subjects across four multigenerational families. We found low microbiome diversity and an altered microbial composition based on 16S sequencing of 12 CD-affected subjects compared to unaffected relatives. Among subjects with CD, we observed an inverse correlation between microbiome diversity and PRS that included CD-relevant genetic variants. There was no relationship between PRS and microbiome diversity among unaffected relatives or with a PRS using ulcerative colitis-specific genetic variants. Among relatives with high PRS, those protected from disease (i.e., CD-/high PRS) had high relative abundance of multiple fecal metabolites (Figure 1). Germ-free IL-10 deficient mice gavaged with these fecal specimens showed reduced disease activity - on par with mice treated with human inocula from those with no CD or low PRS (Figure 2). In contrast, mice colonized with CD+/high PRS inocula displayed high disease activity immediately following gavage.

In conclusion, our findings suggest the presence of a CD-protective factor within the gut microbiome of unaffected relatives with high PRS that is transferable and that mitigates disease activity in an IBD mouse model.

<b>Figure 1. Fecal small molecules that characterize Crohn’s disease risk in multiplex families.</b> Ten subjects with high genetic risk for CD, half with disease and half without, underwent untargeted metabolomic analysis by hybrid liquid chromatography-mass spectrometry. For each subject, the single optimal specimen was identified from longitudinal sampling to avoid sampling near antibiotic use or with intestinal inflammation (determined by the biomarker, fecal calprotectin). Unsupervised clustering analysis of the 460 differentially abundant metabolites (p<0.05) sorts subjects precisely by CD status. CD protection is characterized overall by a relative abundance of numerous metabolites.

Figure 1. Fecal small molecules that characterize Crohn’s disease risk in multiplex families. Ten subjects with high genetic risk for CD, half with disease and half without, underwent untargeted metabolomic analysis by hybrid liquid chromatography-mass spectrometry. For each subject, the single optimal specimen was identified from longitudinal sampling to avoid sampling near antibiotic use or with intestinal inflammation (determined by the biomarker, fecal calprotectin). Unsupervised clustering analysis of the 460 differentially abundant metabolites (p<0.05) sorts subjects precisely by CD status. CD protection is characterized overall by a relative abundance of numerous metabolites.

<b>Figure 2. Disease activity index (DAI) over time by different inocula.</b> DAI differs based on treatment inocula (p=0.001) as determined by two-way ANOVA testing. Tukey’s multiple comparison test shows that this difference is driven primarily by the disease induced by the CD, high PRS inocula compared to other inoculation groups (p=0.001-0.012). DAI in the CD-protected group (No CD, high PRS) is on par with mice treated with No CD and low PRS inocula.

Figure 2. Disease activity index (DAI) over time by different inocula. DAI differs based on treatment inocula (p=0.001) as determined by two-way ANOVA testing. Tukey’s multiple comparison test shows that this difference is driven primarily by the disease induced by the CD, high PRS inocula compared to other inoculation groups (p=0.001-0.012). DAI in the CD-protected group (No CD, high PRS) is on par with mice treated with No CD and low PRS inocula.

Martin Blaser

Appearances