Gregg Morin

SARS-CoV-2 exploits the activity of surface-expressed human proteases to achieve maturation of its Spike (S) protein and undergo conformational changes that enable virus-host membrane fusion and efficient infection of the airways. Early in the COVID-19 pandemic, Transmembrane Protease, Serine-2 (TMPRSS2) was recognized as a key proteolytic driver of SARS-CoV-2 infection of the human aerodigestive tract, analogous to its role in SARS-CoV-1 cleavage activation, but no TMPRSS2-specific drugs had been developed. This lack of progress was in part due to challenges procuring a recombinant source of the active protease. This thesis describes the first production and purification of fully matured TMPRSS2 protease through protein engineering efforts targeting its zymogen activation motif, leading to its first crystal structures and biochemical interrogations of its interaction with recombinant SARS-CoV-2 S protein substrates. The closely related TMPRSS3 and -4 proteases, which belong to the same Type II Transmembrane Serine Protease (TTSP) family as TMPRSS2, were similarly produced and biochemical investigations revealed that matured TMPRSS3 efficiently cleaves the S protein, potentially explaining SARS-CoV-2 tropisms in cells expressing TMPRSS3 and not TMPRSS2. A platform to identify potent and selective TMPRSS-targeted antivirals was developed using the natural TMPRSS protease inhibitor protein, hepatocyte growth factor activator inhibitor-2 (HAI-2) as a novel scaffold. The protease engineering and inhibition strategies established here may enable future interrogations of TTSP pathobiology in vitro and in vivo and rationally select the most critical proteases for viral pathobiology to disable with potent and selective pharmacological inhibitors, improving our pandemic preparedness against respiratory viruses exploiting this entry pathway.

View record

Bottom-up proteomics characterizes proteins from the analysis of peptides released from proteolysis. In this process, proteins are digested, and the mass spectrometry (MS) information acquired from the resulting peptides is used to infer the identity and quantity of the proteins. Various methodologies have been established for the bottom-up approach to achieve high-throughput analysis of proteins. However, the performance of bottom-up proteomics, such as protein coverage, reproducibility, and efficiency still needs to be improved by modifying the existing methods. Chapter 1 provided a general introduction to the field of proteomics and the main techniques used in this thesis. In Chapter 2, seven lysis protocols were compared and three of them were selected as best for global proteomics since they had more proteins detected, with less variation. The selected protocols could increase protein coverage and reproducibility in future studies. Chapter 3 introduced one of the limitations that the data-dependant acquisition (DDA) mode has. Although DDA is widely used in proteomics, the abundance-dependent sampling of DDA often leads to the irreproducible or non-detection of low-abundance peptides. In this chapter, a method termed Isobaric Peptide Doping (isoDoping), which utilizes synthesized peptides in conjugation with tandem mass tags (TMT) was developed to improve the detection of targeted low-abundance peptides within a large-scale global proteomics experiment. In Chapter 4, four different methods were compared to characterize the surfaceome of multiple myeloma (MM) cell lines. Among the methods, global proteome profiling was demonstrated promising for MM since it had low sample consumption, more identified membrane proteins, and the ability to compare the protein expression between MM cells and negative controls. The method could help to find potential therapeutic targets for the MM. In Chapter 5, a method was established to estimate the overall peptide concentration by using a quality control (QC) run after the peptide extraction and before the TMT labeling. The method could improve the normalization of sample amounts for a TMT-based global proteome profiling experiment, thus improving the quantification accuracy.

View record

High-throughput phenotype-based screening of large libraries of compounds without known targets can identify small molecules that elicit a desired cellular response, but additional approaches are required to find and characterize their targets and mechanisms of action. Through such a screen, the novel compound LCS3 was previously identified that selectively kills lung adenocarcinoma (LUAD) cells, but its mechanism of action remained unknown. This thesis used gene expression profiling to elucidate the cellular responses of LUAD cells to LCS3. I demonstrated that LCS3 induces NRF2 pathway activation and oxidative stress through the generation of reactive oxygen species in sensitive LUAD cell lines. I then developed and applied a thermal proteome profiling (TPP) approach and identified the disulfide reductases GSR and TXNRD1 as LCS3 targets. Through enzymatic assays using purified protein, I confirmed that LCS3 inhibits disulfide reductase activity through a reversible and uncompetitive mechanism. The results demonstrated that LCS3-sensitive LUAD cells are correspondingly sensitive to the synergistic inhibition of glutathione and thioredoxin pathways, suggesting a mechanistic overlap in cell death induced by LCS3 and lethality arising from disulfide reductase inhibition. I established that challenging resistant cells with oxidative stress increases reliance on the glutathione and thioredoxin pathways and sensitizes cells to LCS3 and dual disulfide reductase inhibition. Finally, a genome-wide CRISPR-Cas9 knockout screen identified the loss of NQO1 as a mechanism of LCS3 resistance. Together, this work shines light on the mechanism of action of LCS3 and demonstrates the potential utility of disulfide reductase inhibition in lung cancer. This work also highlights the ability of TPP to uncover novel targets of novel small molecules identified by high-throughput screens.

View record

Coupling the capillary electrophoresis (CE) to mass spectrometry (MS) provides an attractive analytical platform with minimal sample consumption, short analysis time, and high separation efficiency. Small-molecule biomarkers are metabolites associated with biological or pathogenic processes. The comprehensive study of small molecules found within the biological system is known as metabolomics. The work presented herein aimed to fulfill the advantage of CE-MS in analytical efficiency and fill the gaps of current methods by achieving the nonaqueous-CE-MS analysis of isomeric prostaglandins and enabling the peak alignment in CE-based metabolomics with inconsistent flow velocities.Urinary creatinine is commonly measured to normalize urinary metabolite concentrations or to assess renal function. In Chapter 2, multisegment injection-CE-MS was combined with a dilute-and-shoot strategy to analyze urinary creatinine. The diluted urines can be directly analyzed with a total analysis time of 2 min/sample, providing a high-throughput method with reliable quantitative performance.A ureteral stent is widely used to treat or prevent ureteral obstruction, but little is known about the mechanisms that drive indwelling stent-associated pain and discomfort. Chapter 3 presents the first study indicating that prostaglandin E2 (PGE2) may be a biomarker for ureteral stent-associated inflammation. The stent-induced inflammation and the associated elevation of cyclooxygenase-2 (COX-2) expression were observed using porcine models. The level of PGE2, a key metabolite of COX-2 signaling, was found to be significantly higher in urines of stented animals compared to that of controls. Migration time fluctuation strongly affects the peak alignment and identification of unknown compounds, making migration time correction an essential step in large-scale metabolomics. Chapter 4 presents a correction method supporting pressure-assisted CE with varied external pressures. Additionally, the idea of using endogenous metabolites as correction markers instead of adding additional compounds was proposed.Multiple reaction monitoring (MRM)-based metabolomics has the advantages of simple data processing and high quantitative reliability. Nevertheless, the difficulty of using this approach lies in the method construction. Chapter 5 presents a straightforward workflow for developing the MRM-based metabolomics assay. Urine metabolic changes associated with stent-induced inflammation were investigated using this assay, and 13 metabolites were found to have concentrations associated with inflammatory levels.

View record

Capillary electrophoresis (CE) is known for its low sample consumption, high-resolution efficiency as a separation technique, while mass spectrometry (MS) provides unprecedented selectivity and sensitivity as a detector, while adding another dimension in separation. The development of a robust CE-MS interface makes it possible to combine these two technologies for the analysis of many types of compounds. However, the coupling of CE to MS reduces its usable electrolyte ingredients substantially, and many of the popular buffer systems based on phosphate or borate, the MEKC methods dependent on the use of non-volatile surfactants, have to be excluded. As a result, formic acid and acetate buffers are the most widely used background electrolytes (BGEs) in CE-MS. In this work, we used mainly organic solvent systems to expand the choices of BGEs for CE-MS. In chapters 2 and 3, we presented two quantitative 100 % nonaqueous CE-MS methods for the separation and determination of small hydrophobic molecules. Two different BGEs were developed in this part: a basic buffer system for the analysis of negatively charged compounds and an acidic buffer system for the analysis of positively charged compounds. The compounds studied are three anthraquinones extracted from traditional Chinese medicine (TCM), Rhubarb, and six synthesized hydrophobic peptides, respectively. A high-organic-content CE-MS (HOCE) method was developed for the proteomics analysis of envelope proteins. A field amplified sample stacking technique was optimized to improve the concentration sensitivity of CE-MS for samples containing a large number of different analytes. The introduction of methanol into the buffer increased the performance of CE-MS for the detection of hydrophobic peptides in complex proteins digest. A half organic CE-MS method was used for the sequencing of novel mAbs. It was demonstrated that CE-MS/MS provided highly complementary information to LC-MS/MS with much less sample consumption. The last part of this thesis describes the application of a new generation mass spectrometry, timsTOF Pro connected with LC for the site-specific O-glycomics. In TIMS, charged compounds were separated based on their differential sizes and charges, similar to CE. The results suggest that combining CE-MS for PTMs analysis can be even more productive in the future.

View record

No abstract available.