Improved Liquid Chromatography-Mass Spectrometry Approaches for Characterization of Therapeutic RNA

Abstract

This dissertation focuses on extending and improving current applications of liquid chromatography-mass spectrometry (LC-MS) for therapeutic oligonucleotides (ONs), which continue to grow in both clinical importance and complexity, including a variety of synthetic modifications. These RNA based therapeutics represent state of the art therapeutic drugs capable of treating a wide variety of health problems. The central placement of RNA in an organism’s cellular systems allows oligonucleotide therapeutics to act on virtually all biological processes of the body suggesting possibly limitless potential application. A triethylammoniumbicarbonate (TEAB)/methanol-based LC mobile phase system, free of fluorinated alcohol additives, is shown to allow sufficient chromatographic resolution and mass spectral signal in significantly more environmentally friendly workflows compared with the conventional hexafluoroisopropanol (HFIP) additive, which is both physiologically and environmentally toxic. Additionally, the use of methanol instead of acetonitrile further reduces the environmental impact of these workflows. With this type of chromatography, we address three major challenges in therapeutic oligonucleotide LC-MS analysis, including difficulties with tandem mass spectrometry (MS/MS)-based sequencing of chemically modified ONs, automated annotation of the corresponding LC-MS/MS spectra, and extension to larger nucleic acids. In Chapter 2, we introduce OligoTap, a bioinformatics utility that allows consideration of any type of chemical modification at any ON location. To date no existing open-source software is similarly flexible. Additionally, no other open-source software allows vendor agnostic data processing. The significantly enhanced throughput of OligoTap-based automated annotation of MS/MS spectra allows for a range of charge states and collision induced dissociation (CID) voltages to be assessed for improved sequence coverage. With this approach, we reached 100% sequence coverage for a heavily modified 22mer RNA, including previously refractory 2’ fluorine modifications. This novel report of full sequence coverage using environmentally friendly chemicals sets a new bar for therapeutic oligonucleotide sequencing by mass spectrometry. In Chapter 3, OligoTap is further evaluated with both CID and electron detachment dissociation (EDD) LC-MS/MS data using a larger search space containing isomeric sequence permutation decoys. The complexity of CID fragmentation remains a challenge in confident spectral annotation. This analysis illustrates the possibility of incorrect sequence assignments despite high mass accuracy and resolution. We introduce a novel quantitative approach to evaluate false discovery rates (FDRs) and demonstrate that EDD, under the utilized conditions/assumptions shows lower FDR compared with CID. We further show that EDD provides a twofold improvement in the ability to identify the correct sequence from decoys, suggesting that EDD is able to double the confidence in spectral annotations. This approach will facilitate further optimization of LC-MS/MS workflows. Finally, in Chapter 4, we demonstrate that TEAB/methanol-based LC-MS on a 7 Tesla Fourier transform-ion cyclotron resonance mass spectrometer allows detection and charge state deconvolution of RNA up to ~100 kDa. The wider charge state distributions generated by electrospray ionization from TEAB/methanol compared with HFIP-based solvent systems improve the input of the deconvolution algorithm, which is able to differentiate phosphate loss and iron adduction at this mass range. We also demonstrate isotopic resolution for an ~40 kDa RNA along with resolution of accompanying sodium and potassium adducts. Overall, the presented advances will aid the continuously developing area of LC-MS-based ON analysis for improved characterization of therapeutic oligonucleotides.