Structural elucidation of novel biomarkers of known metabolic disorders based on multistage fragmentation mass spectra.

Specific diagnostic markers are the key to effective diagnosis and treatment of inborn errors of metabolism (IEM). Untargeted metabolomics allows for the identification of potential novel diagnostic biomarkers. Current separation techniques coupled to high-resolution mass spectrometry provide a powerful tool for structural elucidation of unknown compounds in complex biological matrices. This is a proof-of-concept study testing this methodology to determine the molecular structure of as yet uncharacterized m/z signals that were significantly increased in plasma samples from patients with phenylketonuria and 3-hydroxy-3-methylglutaryl-CoA lyase deficiency. A hybrid linear ion trap-orbitrap high resolution mass spectrometer, capable of multistage fragmentation, was used to acquire accurate masses and product ion spectra of the uncharacterized m/z signals. In order to determine the molecular structures, spectral databases were searched and fragmentation prediction software was used. This approach enabled structural elucidation of novel compounds potentially useful as biomarkers in diagnostics and follow-up of IEM patients. Two new conjugates, glutamyl-glutamyl-phenylalanine and phenylalanine-hexose, were identified in plasma of phenylketonuria patients. These novel markers showed high inter-patient variation and did not correlate to phenylalanine levels, illustrating their potential added value for follow-up. As novel biomarkers for 3-hydroxy-3-methylglutaryl-CoA lyase deficiency, three positional isomers of 3-methylglutaconyl carnitine could be detected in patient plasma. Our results highlight the applicability of current accurate mass multistage fragmentation techniques for structural elucidation of unknown metabolites in human biofluids, offering an unprecedented opportunity to gain further biochemical insights in known inborn errors of metabolism by enabling high confidence identification of novel biomarkers.