Metabolites were tentatively identified using Mass Hunter Qualitative Analysis software (Agilent Technologies) and an in-house database comprising data from the Human Metabolome Database (HMDB), Lipid Maps, and Metlin as previously described65 (link). The software annotates compounds based on isotope ratios, accurate mass, chemical formulae, and scores. Elements for molecular formula generation were C, H, N, O, S, and P. A 10 ppm mass error cut-off was used with a neutral mass range up to 1700 Da and positive ions selected as H+, Na+, K+, and NH4+. Database identifications were limited to the 10 best matches based on score, charge state was limited to a maximum of two. All identifications are Metabolomics Standard Initiative level 2 based on the proposed minimum reporting standards16 (link). If no annotation through database searches was possible, molecular formulae were generated. If no formula could be generated, compounds were represented by a compound number and their retention time. Compounds represented by a formula or compound number and retention time are referred to as “unannotated” compounds. Only annotated compounds were subjected to more specific analyses; changes in unannotated compounds are only discussed for completeness and to reflect patterns of changes. MPP was used for summarization and visualization of data. Annotated compounds and their normalized abundance values were exported to Excel 2010 (Microsoft Corporation, Redmond, WA) for visualization and organization. For organization of data in Venn diagrams, a freely available online tool provided by the University of Gent, Belgium, was used66 .
Results are discussed in regard to changes in specific metabolites and metabolite groups as well as pathways assigned by MetaboAnalyst software. Annotated compounds were assigned to one of the 19 following groups: (1) Bile acids and bile acid metabolism intermediates, (2) Carbohydrates and sugars, (3) Carnitines, (4) Ceramides, glucosylceramides, and ceramide phosphoinositols, (5) Cholesterol, cholesterol esters, and intermediates, (6) Diacylglycerols, (7) Gangliosides, (8) Nucleosides, nucleotides, purine, and pyrimidine metabolism (9) Organic acids and derivatives, (10) Other, (11) Peptides, (12) Phosphatidic acids, (13) Phosphatidylcholines, (14) Phosphatidylethanolamines and phosphatidylceramides, (15) Phosphatidylinositols, (16) Phosphatidylserines, (17) Sphingomyelins, (18) Triacylglycerols, or (19) Vitamin D and derivatives. Categories were assigned dependent on HMDB super classes, classes, and subclasses or manually in case no HMDB entry was available (e.g. vitamin D metabolites). Certain categories were specified further dependent on the nature of the metabolites to enhance conciseness of designations, e.g. carnitines and phosphatidylcholines. Metabolites listed under “Other” (10) belong to the following categories with less than 3 compounds per category: amines (Palmitoleoyl-EA), aldehydes (4-aminobutyraldehyde), steroids and steroid derivates (12alpha-methylpregna-4,9(11)-diene-3,20-dione, tetrahydrodeoxycorticosterone), monoglycerols (MG(18:0e/0:0/0:0), MG(20:2(11Z,14Z)/0:0/0:0)), glycerophosphoglycerols (PG(22:6(4Z,7Z,10Z,13Z,16Z,19Z)/22:6(4Z,7Z,10Z,13Z,16Z,19Z))), metabolites of sphingosine (N,N-dimethylsphingosine), phytosterols (plant-derived, present in small amounts in humans and animals; 22:0-Glc-stigmasterol, 22:2-Glc-stigmasterol), quaternary ammonium salts/alkaloids and derivatives (Neurine), phosphatidylglycerolphosphate (PGP(18:1(11Z)/20:3(8Z,11Z,14Z)), PGP(18:1(11Z)/22:5(7Z,10Z,13Z,16Z,19Z))), intermediates in fatty acid metabolism (2E-tetradecenoyl-CoA), fatty amides/N-acyl amides ((R)-(16,16-dimethyldocosa-cis-5,8,11,14-tetraenoyl)-1′-hydroxy-2′-propylamine). Pathway enrichment analysis was conducted based on KEGG identifiers using MetaboAnalyst19 (link). For pathway coverage compounds with KEGG ID C00107 (Dipeptide) were marked as C00012 (Peptide). Pathway analysis showed significance for selected pathways only; discussed are significantly changed pathways as well as pathways with possible relation to epileptogenesis despite their p-values > 0.05 (not significantly changed).
Free full text: Click here