Search

P046 Crohn Disease (CD) untargeted fecal metabolome shows strong correlation with specific salivary associated microbial taxa, and with disease activity

Levhar, N.(1)*;Hadar, R.(1);Braun, T.(1);Efroni, G.(1);Abramovich, I.(2);Gottlieb, E.(2);Granot, M.(1);Neuman, S.(3);Selinger, L.(3);Picard, O.(3);Yavzori, M.(3);Lahat, A.(3);Eliakim, R.(3);Weiss, B.(1);Kopylov, U.(3);Ben-Horin, S.(3);Amir, A.(1);Haberman Ziv, Y.(1);

(1)Sheba Medical Center, Pediatric Gastroenterology, Ramat Gan, Israel;(2)Technion Institute, The Ruth and Bruce Rappaport- Faculty of Medicine, Haifa, Israel;(3)Sheba Medical Center, Gastroenterology, Ramat Gan, Israel;

Background

The fecal metabolome is comprised from derivatives of the gut microbiome, the host, the environment, and the diet. Thus, it is a central output and effector that is linked with Crohn disease (CD). We aimed to characterize metabolomics profiles that are associated with CD and link metabolites with gut microbial diversity and composition.

Methods

Fecal metabolomics was performed using liquid chromatography coupled Mass Spectrometer (LC/MS). Microbiome was defined by 16Sseq. Multivariate analyses in MaAsLin2 identified differentially abundant metabolites and these were used for pathways enrichment analyses using Metacyc (PMID: 31586394). Hierarchical All-against-All association (HAllA; PMID: 35758795) was used to identify significant relationships between fecal metabolites and microbial amplicon sequence variants (ASVs).

Results

The cohort included 294 samples from 85 subjects (mean age 36, 51% males). 250 samples were from CD patients (n=59) and 44 samples were from controls (n=26). 545 metabolites were included in our predefined library and were subsequently analyzed. Multivariate analyses identified 86 increased and 18 decreased metabolites in CD, after controlling for age and gender (FDR<0.25). These included reduction in 9 long-chain fatty acids and increase of standard and post-translational modified amino-acids. Enriched pathways in CD included amino-acids biosynthesis and degradation, electron carrier biosynthesis, guanine and guanosine salvage, and indole-3-acetate inactivation (FDR<0.1). Within the CD comparison, 29 metabolites were higher in active and 97 in quiescent CD. Acetylated amino-acids, dopamine, glucuronolactone and decanoic acid were all reduced in active CD. These metabolites and others were also significantly correlated with fecal calprotectin, microbial alpha diversity (Faith’s PD) and CD dysbiosis index (PMID: 24629344) (FDR<0.25; Fig. 1).  To correlate fecal metabolomics with the gut microbial dataset, we applied the HAllA analysis (n= 231) that indicated 7,184 significant associations (FDR<0.05; Fig. 2). An interesting positive correlation was noted between a group of bacteria enriched with saliva bacteria (s_parvula, g_Actinomyces, g_Streptococcus, etc.) and metabolites like fructose, stachyose, glucose, raffinose etc., that are involved in stachyose metabolism (cluster 16).

Conclusion

CD patients have distinct fecal metabolic signature which is also correlated with disease activity. Furthermore, we identified significant interactions between specific fecal metabolites and gut microbial ASVs enriched with saliva bacteria that were previously linked with CD. More research is needed to elucidate the factors governing this newly discovered link.