Date of Award
1-1-2023
Language
English
Document Type
Dissertation
Degree Name
Doctor of Philosophy (PhD)
College/School/Department
Department of Chemistry
Content Description
1 online resource (xxiv, 334 pages) : illustrations (some color)
Dissertation/Thesis Chair
Rabi A Musah
Committee Members
Jeremy I Feldblyum, Mehmet Yigit, A. John Dane
Keywords
Disease diagnosis, Earwax, Imaging mass spectrometry, Mass spectrometry of complex matrices, Small molecule analysis, Wood, Mass spectrometry, Molecular diagnosis, Ménière's disease
Subject Categories
Chemistry
Abstract
In this work, the newly developed technique laser ablation direct analysis in real – time high-resolution imaging mass spectrometry (LADI-MS) and the more established methods direct analysis in real time – high-resolution mass spectrometry (DART-HRMS) and two-dimensional gas chromatography – mass spectrometry (GC×GC-MS), were used to interrogate challenging plant- and human-derived complex matrices that have proven difficult to analyze by conventional methods, in order to extract heretofore inaccessible information that can be applied to the understanding of wood systems biology, ecophysiology, dendrochronology and natural product isolation of biologically relevant compounds (on the plant side), and explore the potential for discovery of the presence of molecular markers of a rare otolaryngological disorder, Ménière’s disease (on the human disease diagnostics side). LADI-MS analysis of the small molecule spatial distributions within wood transects of Entandrophragma candollei Harms (named Kosipo), Millettia laurentii De Wild (named Wenge) and Pericopsis elata (Harms) Meeuwen (named Afrormosia), which are all of economic and forensic importance, revealed for the first time several compounds that were associated with specific anatomical features including glycerin, furfural, pentanoic acid, benzaldehyde, benzyl alcohol and benzoic acid. The greatest diversity of compounds were found withing the vessels and parenchyma, and to a lesser extent in the wood fibers. It was also revealed that for certain compounds, their distributions were varied in the heartwood in comparison to the sapwood (i.e., benzoic acid and furfural had lower levels in the sapwood). The approach illustrates how plant ecophysiology investigations can be facilitated, and demonstrates how studies that reveal not only the association between chemical make-up and environmental factors, but also how the possible mechanisms for the synthesis and trafficking of small molecules within specialized tissues can be conducted. GC-MS analysis of the economically important Voacanga africiana seeds revealed the presence of 31 newly reported compounds in the seeds. The spatial distributions of a number of these molecules were mapped using LADI-MS. It was found that fatty acids were centralized in the embryo while alkaloids spanned the endosperm. These findings illustrate that V. africana seeds could be used as a renewable resource for biologically relevant molecules, and the knowledge of their locales could assist in the development of protocols for their isolation. On the application of mass spectrometry to disease diagnosis, GC×GC-MS was applied in an unconventional way to study the highly intractable and challenging to analyze earwax matrix, as a means by which to reveal the presence of Ménière’s disease, an incurable vestibular disorder that is challenging to diagnose. Analysis of earwax from non-Ménière’s disease donors generated chromatograms in the form of visually interpretable contour plots that revealed a range of compound classes present, including alkanes, alkenes, fatty acids, esters, cholesterol esters and triglycerides. While these classes had previously been detected by other methods, GC×GC-MS facilitated the resolution of peaks that co-eluted by 1D-GC, thereby providing the opportunity to detect and identify several analytes for the first time, including 1,6,10,14,18,22-tetracosahexaen-3-ol, 2,6,10,15,19,23-hexamethyl, lathosterol, C28H48O, cholest-4-en-3-one, lanost-8-en-3-ol (3β), lanosterol, cholesteryl myristate, cholesteryl heptadecanoate and cholesteryl stearate. In all, 39 compound were identified, 30 more than had been previously reported for analysis of unsaponified earwax. The characterization of earwax from Ménière’s disease donors contrasted with that of non-Ménière’s disease donors in a number of ways, including the number of peaks detected by DART-HRMS and relative abundance of a number of compounds including fatty acids, alkanes and esters in GC-MS analyses. Application to the DART-HRMS-derived chemical profiles of the machine learning technique random forest (RF), facilitated determination of m/z values that enabled discrimination between the two sample types. GC-MS facilitated identification of these compounds showed them to be cis-9-hexadecenoic acid, cis-10-heptadecenoic acid and cis-9-octadecenoic acid. While all three molecules were detected in both Ménière’s disease and non-Ménière’s disease samples, the relative amounts in each sample type differed markedly. In non-Ménière’s disease samples, the average concentrations were 7.89 µg/mL, 0.87 µg/mL and 4.94 µg/mL, while in the Ménière’s disease samples, the average amounts were 1.70 µg/mL, 0.13 µg/mL, and 2.07 µg/mL. It was the differences in the relative concentrations of this subset of compounds that enabled accurate prediction of the presence of Ménière’s disease from analysis of the earwax samples. Thus, the prediction ability of the RF model using external validation samples showed it to be 100% accurate in predicting whether or not the earwax donor had Ménière’s disease. Overall, it has been demonstrated in this work how some challenging matrices can be interrogated in a manner that unveils attributes and characteristics of materials that can be leveraged for forensic, economic, research and disease diagnosis benefits.
Recommended Citation
Coon, Allix Marie, "Development of new mass spectrometric approaches for biomarker discovery: application to complex biological matrices for tissue characterization and disease diagnosis" (2023). Legacy Theses & Dissertations (2009 - 2024). 3102.
https://scholarsarchive.library.albany.edu/legacy-etd/3102
