Molecular mechanisms involved in the side effects of fatty acid amide hydrolase inhibitors: a structural phenomics approach to proteome-wide cellular off-target deconvolution and disease association

Abstract

Fatty acid amide hydrolase (FAAH) is a promising therapeutic target for the treatment of pain and CNS disorders. However, the development of potent and safe FAAH inhibitors is hindered by their off-target mediated side effect that leads to brain cell death. Its physiological off-targets and their associations with phenotypes may not be characterized using existing experimental and computational techniques as these methods fail to have sufficient proteome coverage and/or ignore native biological assemblies (BAs; i.e., protein quaternary structures). To understand the mechanisms of the side effects from FAAH inhibitors and other drugs, we develop a novel structural phenomics approach to identifying the physiological off-targets binding profile in the cellular context and on a structural proteome scale, and investigate the roles of these off-targets in impacting human physiology and pathology using text mining-based phenomics analysis. Using this integrative approach, we discover that FAAH inhibitors may bind to the dimerization interface of NMDA receptor (NMDAR) and several other BAs, and thus disrupt their cellular functions. Specifically, the malfunction of the NMDAR is associated with a wide spectrum of brain disorders that are directly related to the observed side effects of FAAH inhibitors. This finding is consistent with the existing literature, and provides testable hypotheses for investigating the molecular origin of the side effects of FAAH inhibitors. Thus, the in silico method proposed here, which can for the first time predict proteome-wide drug interactions with cellular BAs and link BA–ligand interaction with clinical outcomes, can be valuable in off-target screening. The development and application of such methods will accelerate the development of more safe and effective therapeutics.

ANTENNA, a Multi-Rank, Multi-Layered Recommender System for Inferring Reliable Drug-Gene-Disease Associations: Repurposing Diazoxide as a Targeted Anti-Cancer Therapy

“…Disease-disease similarity is required for tREMAP to infer chemical-disease associations and can be calculated using distributed word representations [19]. In this work, we do not infer the chemical-disease association directly using tREMAP, since only less than 0.4% of chemicals have observed associations with one or more diseases.…”

Endocannabinoid System Components as Potential Biomarkers in Psychiatry

“…Apart from the widely explored role of CB1r, it should be noted that in recent years an increasing emphasis is being placed on the design of strategies to regulate endogenous cannabinoid tone, through inhibitors of the degradation or reuptake of eCBs. Since this approach provided negative results, particularly regarding the inhibition of FAAH (373), current trends focused on the combination of different mechanisms of action to enhance the endocannabinoid tone (374). Moreover, the pharmacological modulation of CB2r has also attracted much attention given its safety profile and the wide range of properties attributed to it, including mood and cognitive regulation.…”

Multi-scale predictive modeling discovers Ibudilast as a polypharmacological agent to improve hippocampal-dependent spatial learning and memory and mitigate plaque and tangle pathology in a transgenic rat model of Alzheimer’s disease

“…The details of methods were published elsewhere 28 . In brief, disease, gene, and chemical name entries were recognized using DNorm v0.0.6 50 , Gimli v1.0.2 51 , and tmChem 52 , respectively.…”