Characterization and quantification of urinary metabolic biomarkers for early response to anti-tuberculosis treatment

Abstract

Development of new anti-tuberculosis (TB) therapies remains a major priority to combat this infectious disease and to prevent continued transmission of the causative agent Mycobacterium tuberculosis (Mtb). However, newly developed therapies require large, lengthy clinical trials to determine the number of treatment failures and relapses for evaluation of treatment efficacy. Biomarkers for the prediction of treatment outcome in TB patients at early time points would facilitate movement of new therapies through clinical trials. Previously, liquid chromatography-mass spectrometry (LC-MS) based metabolomics experiments identified potential biomarkers for early response to anti-TB treatment. The research presented in this dissertation involves experiments needed for the progression of these compounds towards a clinically useful biosignature. A major impediment to metabolomics-based biomarker discovery is metabolite identification, as approximately 50% of detectable products do not match structures in existing databases. In concordance with this, several of the potential small molecule biomarkers of anti-TB treatment response lacked structural identification. This research resulted in the structural characterization of three of these compounds as a core 1 O-glycosylated SerLeu peptide, N-acetylisoputreanine, and N1, N12-diacetylspermine. Both the core 1 O-glycosylated SerLeu peptide and N-acetylisoputreanine are novel compounds that had not been previously detected in human urine. Characterization of these compounds indicated a potential alteration of polyamine catabolism and the complement and coagulation pathways during anti-TB treatment. Another key aspect in biomarker discovery is defining the processes involved in formation of potential biomarkers. In order to determine whether these compounds were formed by processes upregulated during active disease, the abundances of these compounds were assessed in active TB patients and household contacts as well as in Mtb infected and uninfected Balb/c mice. The core 1 O-glycosylated SerLeu peptide and N1, N12-diacetylspermine were increased in the urine of index patients demonstrating a potential link between Mtb infection, associated disease pathology, and the formation of these compounds. N-acetylisoputreanine, however, was not increased in TB patient urine or infected mouse tissue indicating that this compound may be formed due to off target drug interactions. These experiments not only provided insights into the mechanisms behind alteration of these compounds during anti-TB treatment, but also highlighted those compounds that may be better biomarkers for anti-TB treatment response. Assessment of these compounds using an independent set of patient samples is needed to validate them as biomarkers for early anti-TB treatment response. Unlike the untargeted experiments used for discovery of potential biomarkers, validation typically employs targeted assays. This research describes the development of a targeted multiple reaction monitoring (MRM) assay which enabled accurate and precise quantitation of compounds previously detected in an untargeted metabolomics experiment. This targeted assay will be used for validation of these compounds in a larger set of patient samples representing a variety of different treatment outcomes. Overall these experiments confirmed the identity of three metabolites that decrease with anti-TB treatment response. Two of these metabolites are novel compounds and their characterization adds to metabolite databases expanding the number of metabolites available to other metabolomics researchers. Assessment of these compounds in samples representative of active TB disease confirmed two of them as promising biomarkers for anti-TB treatment response and highlighted another as a potential result of unintended drug effects. The development of a MRM assay for the quantification of these compounds enables their validation and confirmation as biomarkers of anti-TB treatment response. The work presented in this dissertation describes the advancement of metabolites identified during biomarker discovery towards application in clinical trials.