Development of metabonomics tools for biomarker identification in multiple sclerosis

Date of Completion

January 2010


Health Sciences, Toxicology|Health Sciences, Pharmacology|Chemistry, Analytical




Mass spectrometry and high performance liquid chromatography are commonly used for compound characterization and identification. In the field of metabonomics, these two analytical techniques are often combined to characterize unknown endogenous or exogenous metabolites in complex biological samples. Since the structures of a majority of these metabolites are not identified, the result of most metabonomic studies is a list of m/z values and retention times. However, without knowing actual structures, the biological significance of these 'features' cannot be determined. The process of identifying structures of unknown compounds can be time intensive, costly, and frequently requires the use of multiple orthogonal analytical techniques—this laborious procedure seems insurmountable for long lists of unknowns that must be identified for each study. In addition, the limited sample volume and low concentration of most endogenous analytes frequently make purification and identification by additional instrumentation nearly impossible. This dissertation explores the use of LC/MS techniques and orthogonal structural descriptors for quickly identifying the structures of unknown small molecules. ^ Mass spectral fragmentation profiles are commonly used for structure elucidation. The effectiveness of commercially available mass spectral fragmentation prediction software was evaluated for biomarker identification. This software was most effective at ranking potential structures suggesting that it would be beneficial to apply additional orthogonal descriptor algorithms to exclude incorrect structures. Fragmentation characterization was extended to quantify the energy required to induce fragmentation, termed CE50. CE 50 effectively distinguished between compounds of the same molecular formula and even positional isomers. A preliminary model suggested that CE 50 is predictable. Retention index has been commonly used in GC/MS for structure elucidation and discrimination. A predictive model was developed for HPLC based retention index for an eclectic group of small molecules. ^ Metabonomic analysis of human cerebrospinal fluid from patients with multiple sclerosis produced a list of analytical 'features' for potential biomarkers. The combination of retention index, CE50, and fragmentation pattern prediction excluded on average 65% of the structures in the bins and improved ranking by 43 compounds. This protocol successfully filters through potential structures in a high throughput manner and demonstrates an effective way to handle structure identification from extensive chemical databases. ^