Trapped: Developing an unbiased approach to identify epoxide hydrolase substrates in vivo

Abstract

Epoxide hydrolases (EHs) of the α/β hydrolase fold enzyme family hydrolyze epoxides to the corresponding vicinal diols. In mammals, epoxides are mainly formed within the body through epoxidation of xenobiotic or endogenous substrates by cytochrome P450-dependent monooxygenases (CYPs). Two of the five known mammalian EHs are well characterized. The microsomal epoxide hydrolase (mEH) is primarily involved in detoxification of carcinogenic epoxides derived from xenobiotic compounds and the soluble epoxide hydrolase (sEH) primarily regulates endogenous signaling epoxides derived from fatty acids. Although some substrates are known for the epoxide hydrolase 3 (EH3), its function remains unclear. For the epoxide hydrolase 4 (EH4) and mesoderm specific transcript (MEST) no substrates are known, but knockout of MEST in mice leads to growth retardation and a behavioral phenotype. Another EH with activity in the human body is the CFTR inhibitory factor (Cif), a virulence factor secreted by the opportunistic pathogen Pseudomonas aeruginosa. Cif reduces the surface expression of the chloride channel CFTR in human airway endothelial cells, resulting in increased mucus viscosity facilitating bacterial colonization. The enzyme activity of Cif is crucial for the effect on CFTR but the molecular target remains to be identified. In this project, we aimed to identify the physiologically relevant substrate(s) of the aforementioned EHs. For the bacterial EH Cif, a substrate screening was performed using recombinantly expressed enzyme, but kinetic analysis suggested that the found substrates 14,15-epoxyeicosatrienoic acid (14,15-EET), 17,18-epoxyeicosatetraenoic acid, and 19,20-epoxydocosapentaenoic acid are not relevant in vivo. To identify relevant substrates in vivo, an unbiased approach was developed that takes advantage of the characteristic two-step reaction mechanism of α/β hydrolase fold EHs. By introducing a point mutation, trapping mutants of the six EHs were constructed which are able to nucleophilically attack and bind their substrates but cannot perform the second hydrolytic step, thus trapping their substrates with a covalent ester bond. Using adeno-associated viruses (AAVs), these trapping mutants were expressed in mice, where they should encounter and bind their substrates in a physiologic environment. For expression in peripheral organs of mice, the serotype AAV-rh10 was used, while the serotype AAV-PHP.B was used to target the brain after the intravenous injection of the virus into the tail vein. Successful expression in the liver was confirmed for all trapping EHs except EH3, and in the brain for mEH, sEH and EH4 by western blot analysis. Mass spectrometry was used to identify the substrates that were trapped by the enzymes. For this, mouse tissue was lyzed and the virally expressed trapping mutants were enriched using His-tag affinity chromatography. After digestion with trypsin, the samples were analyzed on a TripleTOF mass spectrometer using SWATH, a data independent scan mode which provides time-resolved recording of the fragment ions of all precursors. Since the sequence and therefore the mass of the peptide which carries the trapped substrate is known, the mass of the substrate can be determined. With mass spectrometry we confirmed the presence of virally expressed mEH, sEH, Cif and MEST in the liver and mEH and sEH in the brain by detecting peptides specific for these enzymes. However, no substrate modified peptide was detected for any enzyme in any of the tissues. As proof of principle, the peptide-substrate complex of recombinantly expressed sEH incubated with 14,15-EET was detected in an in vitro trapping experiment confirming that the ester intermediate is stable enough for detection by mass spectrometry. Analysis of the sample preparation procedure with synthetic peptides revealed a selective and substantial loss of the sEH, EH3, and EH4 peptides. This was probably the main problem in the analysis of the sEH expressing tissue. For mEH and MEST, only the unmodified peptide was detected, indicating that the trapping mutants were not able to bind their substrates, probably due to protein misfolding. EH3, EH4, and Cif peptides were likely not detected due to their low tissue concentrations, probably caused by a shorter half-life of these proteins. Nevertheless, successful trapping and analysis was shown in vitro and required measures to improve the method were identified. Establishment of the in vivo trapping approach and identification of EH substrates could bring fundamental insights into the role of EHs in mammalian physiology and new strategies for therapeutic intervention. Unexpectedly, mice expressing trapping mEH in the brain developed a striking trembling phenotype which was milder in mEH KO mice and absent when wild type instead of trapping mEH was expressed. Further analysis is required but preliminary experiments suggest a loss of dopaminergic cells in the substantia nigra causing a Parkinson’s disease-like pathology Weniger anzeigen