RESUMEN
Folate deficiency in the critical developmental period has been repeatedly associated with an increased risk of Autism spectrum disorders (ASD), but the key pathophysiological mechanism has not yet been identified. In this work, we focused on identifying genes whose defect has similar consequences to folate depletion in the metabolic network. Within the Flux Balance Analysis (FBA) framework, we developed a method of blocked metabolites that allowed us to define the metabolic consequences of various gene defects and folate depletion. We identified six genes (GART, PFAS, PPAT, PAICS, ATIC, and ADSL) whose blocking results in nearly the same effect in the metabolic network as folate depletion. All of these genes form the purine biosynthetic pathway. We found that, just like folate depletion, the blockade of any of the six genes mentioned above results in a blockage of purine metabolism. We hypothesize that this can lead to decreased adenosine triphosphate (ATP) and subsequently, an S-adenosyl methionine (SAM) pool in neurons in the case of rapid cell division. Based on our results, we consider the methylation defect to be a potential cause of ASD, due to the depletion of purine, and consequently S-adenosyl methionine (SAM), biosynthesis.
RESUMEN
Recently, cerebral folate deficiency (CFD) was suggested to be involved in the pathogenesis of autism spectrum disorders (ASD). However, the exact role of folate metabolism in the pathogenesis of ASD, identification of underlying pathogenic mechanisms and impaired metabolic pathways remain unexplained. The aim of our study was to develop and test a novel, unbiased, bioinformatics approach in order to identify links between ASD and disturbed cerebral metabolism by focusing on abnormal folate metabolism, which could foster patient stratification and novel therapeutic interventions. An unbiased, automatable, computational workflow interaction model was developed using available data from public databases. The interaction network model of ASD-associated genes with known cerebral expression and function (SFARI) and metabolic networks (MetScape), including connections to known metabolic substrates, metabolites and cofactors involving folates, was established. Intersection of bioinformatically created networks resulted in a limited amount of interaction modules pointing to common disturbed metabolic pathways, linking ASD to CFD. Two independent interaction modules (comprising three pathways) covering enzymes encoded by ASD-related genes and folate cofactors utilizing enzymes were generated. Module 1 suggested possible interference of CFD with serine and lysine metabolism, while module 2 identified correlations with purine metabolism and inosine monophosphate production. Since our approach was primarily conceived as a proof of principle, further amendments of the presented initial model are necessary to obtain additional actionable outcomes. Our modelling strategy identified not only previously known interactions supported by evidence-based analyses, but also novel plausible interactions, which could be validated in subsequent functional and/or clinical studies. Autism Res 2017, 10: 1424-1435. © 2017 International Society for Autism Research, Wiley Periodicals, Inc.