Gábor Transform-Based Deconvolution and Quantitative Analysis Methods for Electrospray Mass Spectrometry of Intact Biomolecules

Abstract

Advancements in high-resolution mass spectrometry (HRMS) have enabled the analysis of complex biological samples with minimal preparation, a critical advantage for tissue imaging and therapeutic drug monitoring (TDM), particularly as personalized medicine shifts toward biomolecule-based therapies. However, the resulting data-rich mass spectra (abundances of detected ions as a function of their mass-to-charge ratios (m/z)) can be challenging to interpret due to the charging process of electrospray ionization (ESI), the large number of analytes detected, and the complexity of the matrix. ESI, the most common method for introducing solution-phase samples into a mass spectrometer, preserves biomolecule structure but distributes a single mass across multiple charge states. To simplify HRMS data and facilitate biomolecular characterization, mass spectrometry (MS) deconvolution algorithms identify charge states and recombine related peaks into a single mass. Most modern MS deconvolution algorithms effectively fit the mass spectrum with a hypothetical mass distribution; conversely, the Prell group’s iFAMS software uses Fourier transform (FT) and Gábor transform (GT) to create a highly specific notch filter to isolate and deconvolve mass spectral signal. FT and GT are invertible functions used to detect periodic signal which can manifest in MS from a variety of sources, such as a varying number of neutrons in isotopes or sugars in glycoforms or multiple adductions of sodium or potassium ions. Like most other MS deconvolution algorithms, iFAMS software has primarily been applied to complex biomolecule characterization, and despite the potential major advantage of increased signal-to-noise from combining charge state signal, application of MS deconvolution algorithms to biomolecule quantitation has been limited due to a lack of algorithm objectivity, among other reasons. This dissertation presents the first application of GT for biomolecule quantitation via charge deconvolution of ESI-HRMS data. Developments in iFAMS software presented in this dissertation (“iFAMS Quant) improve spectral deconvolution and objectivity and introduce a platform to quantitate directly from iFAMS deconvolved mass spectra. Quantitation capabilities are validated against a Mayo Clinic method for t-mAb quantitation in serum, demonstrating comparable accuracy while also identifying chemical interferents. Additionally, an automatic pixel-specific GT-based deconvolution algorithm is introduced, further increasing peak capacity and objectivity. Finally, future applications of and improvements to these GT-based algorithms are discussed. This dissertation includes previously published and unpublished co-authored material.