A Multi-Omic Analysis of Peripheral Blood from Children with ASD and Ileocolonic Inflammation

Background: Gastrointestinal (GI) symptoms are a common co-occurring medical issue in children with autism spectrum disorder (ASD). We have previously described unique GI mucosal biomarkers specific for ASD-associated ileocolitis in children. It is not yet known whether unique biomarkers are also present in the blood of these individuals. Identification of a validated blood-based biomarker of ASD-associated ileocolitis would allow for earlier identification of co-morbid GI disease and earlier GI intervention in affected patients. Moreover, it would provide insight into the relevant genes and metabolic pathways in ASD-associated ileocolitis.

Objectives: In an effort to enable a more complete understanding of the biology that underlies GI inflammation in children with ASD, the goal of these studies was to use an integrated omics approach to evaluate blood-based gene expression and serum metabolite relative abundance in GI-symptomatic children with ASD that have a demonstrated histologic ileocolitis.

Methods: The study cohort was comprised of whole blood and serum from 22 children with ASD who were undergoing clinically-indicated ileocolonoscopy for chronic GI symptoms, and 24 non-ASD (typically developing, TD) children undergoing ileocolonoscopy for a variety of GI symptoms. All children with ASD had histologic inflammation of the ileum, colon, or both. The TD controls used for this study were selected based on absence of histologic inflammation anywhere in the GI tract and absence of a neurodevelopmental disorder. Molecular profiling in peripheral blood (transcriptome) and serum (metabolome) from children with ASD (and ileocolitis) and TD children (without ileocolitis) was performed to identify differentially expressed transcripts and metabolite abundance levels that may serve as a proxy for GI inflammation.

Results: Differential gene expression analysis (using whole genome microarray) identified a large number of both up- and downregulated transcripts. The significantly upregulated transcripts in ASD were enriched for pathways including ECM-receptor interaction, intestinal immune network, fatty acid biosynthesis, hematopoiesis, serotonin transporter activity, platelet degranulation, platelet activation, signaling and aggregation, and cytokine signaling. The significantly downregulated transcripts in ASD were enriched for pathways such as arachidonic acid metabolism, linoleic acid metabolism, extracellular vesicle-mediated signaling, NOD pathway, aryl hydrocarbon receptor pathways, and oxidative ethanol degradation. From the metabolomics data, fatty acid metabolism (acyl carnitines), phenylalanine and tyrosine metabolism were uniquely associated with ASD, whereas steroidal metabolism and xanthine metabolism were uniquely associated with TD controls. Integration of transcript and metabolite data revealed enrichment of caffeine metabolism, purine nucleotide metabolism, and glucose alanine cycle.

Conclusions: On one hand, the gene-expression signatures revealed molecular signaling pathways, while on the other hand the metabolomics data revealed the metabolic biosynthetic and catabolic routes of ASD. Both datasets point to dysregulation of fatty acid and lipid metabolism spanning the two omics layers. Such complementary multi-omics efforts have proven useful to understand the disease mechanisms in greater detail and to facilitate rapid hypothesis generation in the context of ASD.