Chemical-Gene-Disease Interaction Analysis Reveals Overlaps Between Autism and Cancer

Thursday, May 11, 2017: 12:00 PM-1:40 PM
Golden Gate Ballroom (Marriott Marquis Hotel)
Y. Wen1,2 and M. R. Herbert1,2, (1)Neurology, Massachusetts General Hospital, Charlestown, MA, (2)Harvard Medical School, Boston, MA

Published research increasingly supports the contribution of environmental factors such as toxicants and nutrients to Autism Spectrum Disorders (ASDs). Finding associations between chemicals and ASDs can provide a starting point for investigation of possible mechanisms by which these substances may contribute to ASDs and their features.


We set out to generate informatics-based data linking autism-associated genes and autism-associated chemicals and substances in order to provide insight into the underlying molecular means by which environmental factors influence ASDs.


We used a well-established chemical-gene-disease database, Comparative Toxicogenomics Database (CTD), as our data source to investigate gene-chemical interactions in relation to ASD-associated chemicals. 1) We used the MeSH term “Autism Spectrum Disorder” as a search term under the “Disease” category in CTD. Under the “Chemical” sub-category for ASD, we chose the chemicals that have a curated association to the disorder, which means that the chemicals have been reported to be related to ASDs by what CTD categorizes as “direct evidence.” 2) Then, for each ASD associated chemical, CTD’s “top 10 interacting genes” were used in further analysis. Some chemicals had less than 10 interacting genes, in which case all genes noted in this CTD category were used. 3) We pooled all the top genes for every selected chemical and used the Gene Set Analyzer tool provided within CTD to perform a gene set enrichment analysis to identify enriched pathways and gene ontologies. 4) Some of the genes interacted with more than one ASD substance; we tabulated the number of substances with which each gene interacted and ranked the genes by this number.


16 substances were found to be associated with ASDs by presently available direct evidence: Valproic Acid, Arsenic, 5-Methylcytosine, 5-hydroxymethylcytosine, Testosterone, Serotonin, Air Pollutants, Lead, Zinc, 8-oxo-7-hydrodeoxyguanosine, Triiodothyronine, Cadmium, Magnesium, Mercury, Aluminum, and Androstenedione. A total 109 genes that interact with these chemicals were investigated. The enrichment analysis found that these genes were enriched for pathways “Pathways in Cancer” (p-value 5.03e-16) from pathway database KEGG; and “Metabolism” (p-value 1.01e-18), and “Signal Transduction” (p-value 6.57e-15) from pathway database REACTOME. The Gene Ontology enrichment analysis found that these 109 genes were enriched for “Response to Chemical” (Biological Process, 4.92e-62) and “Binding” (Molecular Function 4.04e-33). The most impacted genes with most number of interactions with ASD chemicals/substances were: a) CASP3, CYP1A1, HMOX1, MT1, PTGS2, and TNF, each with 4 interactions; and b) CAT, and MAPK3, each with 3 interactions.


The gene enrichment findings for KEGG “Pathways in Cancer” and for REACTOME “Metabolism” suggest overlaps between ASDs and both cancer and metabolism. All of the most impacted genes listed above are also associated with cancers. The ASD-cancer overlap, while previously reported, is derived here from chemical-gene interactions—a new source. The enriched GO categories “Response to Chemical” and “Binding” – together with the enriched REACTOME pathway “Signal Transduction” – are cellular reactions that process “environmental information.” Our results suggest that information processing procedures may play a mechanistically crucial role in environmental contributions to ASDs.

See more of: Genetics
See more of: Genetics