Enzymes regulate biological processes through the conversion of specific substrates to products. Therefore, of fundamental interest for every enzyme is the elucidation of its natural substrates. Here, we describe a general strategy for identifying endogenous substrates of enzymes by untargeted liquid chromatography-mass spectrometry (LC-MS) analysis of tissue metabolomes from wild-type and enzymeinactivated organisms. We use this method to discover several brain lipids regulated by the mammalian enzyme fatty acid amide hydrolase (FAAH) in ViVo, including known signaling molecules (e.g., the endogenous cannabinoid anandamide) and a novel family of nervous system-enriched natural products, the taurine-conjugated fatty acids. Remarkably, the relative hydrolytic activity that FAAH exhibited for lipid metabolites in Vitro was not predictive of the identity of specific FAAH substrates in ViVo. Thus, global metabolite profiling establishes unanticipated connections between the proteome and metabolome that enable assignment of an enzyme's unique biochemical functions in ViVo.Enzymes are central components of nearly all signal transduction cascades and metabolic pathways. Genome sequencing projects have provided, for the first time, a view of the total set of enzymes expressed in an organism. Additionally, emerging genetic technologies (e.g., RNAi, targeted gene disruption) coupled with cell-and organismbased phenotypic screens enable a genome-wide analysis of the (patho)physiological activities of enzymes (1-3). In contrast, complementary global approaches for elucidating the endogenous biochemical functions of enzymes are lacking, and as a consequence, the substrates of many enzymes remain unknown. Thus, the assignment of natural substrates to enzymes represents a major problem of the postgenomic era, and systematic methods that directly connect the proteome with the metabolome are needed for its solution.The substrate selectivities of enzymes are generally determined in Vitro using purified protein preparations. However, translating these in Vitro findings into a comprehensive understanding of the scope of substrates utilized by enzymes in ViVo is problematic for multiple reasons. First, in the living cell and organism, enzymes do not function in isolation, but rather as parts of large and complex biochemical networks (4). Accordingly, in Vitro assays may fail to account for many potentially competitive metabolic pathways that alter or restrict the substrates utilized by a particular enzyme in ViVo (5). Second, efforts to assign natural substrates to enzymes in Vitro are limited by our current knowledge of cellular biochemistry and, therefore, ill-suited for the discovery of novel metabolites that are regulated by enzymes in ViVo. Finally, many enzymes are subject to posttranslational regulation in ViVo, including covalent modification (e.g., phosphorylation) and protein-protein interactions, which may alter substrate recognition and catalysis (6).We postulated that a more direct route for identifying endoge...