Lipid and Metabolite Profiling Mass Spectrometry of Human Glioblastoma Using Mouse Xenograft Model

Soo Jung Ha, Purdue University

Abstract

Glioblastoma (GBM) is the most common and malignant form of human primary brain tumor. Due to its extremely malignant and invasive nature, the mean survival period of GBM patients is only 12 to 15 months from diagnosis. GBM patients rely on current treatment options including surgery, radiation and chemotherapy. Unfortunately, these treatment options fail to improve quality of patients’ lives. Studies have shown that altered lipids and metabolites in GBM may play a critical role in chemo-resistance, cancer cell initiation and progression regulated by oncogenic signaling pathways. This research utilized multiple reaction monitoring (MRM) profiling mass spectrometry (MS) and desorption electrospray ionization mass spectrometry (DESI-MS) imaging MS to investigate altered lipid and metabolite expression in GBM tissues using mouse xenograft models. Two commonly used mouse xenograft models, orthotopic and subcutaneous models were transplanted with human GBM cell lines. Intracerebral implants of the human GBM cell lines were performed in the right cerebral hemisphere in NOD.Cg-PrkdcscidIL2rgtm1Wjl/ Sz (NSG) mice to obtain tumor tissues. Subcutaneous implants were performed on the flank site of the mice and tumors were extracted following sufficient tumor growth. These tumor and control tissues were applied to MRM-profiling by flow injection electrospray (ESI) MS, which provides a platform for screening analysis of lipidomes directly from crude extracts of biological samples.MRM-profiling has benefits of using neutral loss (NL) and precursor ion (PREC) scan mode to profile diverse classes of lipids and metabolites based on the lipid chemical structure, is common in ‘building blocks’. Therefore, diverse lipid classes present characteristic fragment ions. MRM- profiling approach also simplifies the isolation of relevant lipids by avoiding the necessity of internal standards, chromatographic separation and, by simplifying data and metabolite structural analysis, workflow. After MRM-profiling analysis was performed, DESI-MS experiment was followed. Ambient ionization DESI was developed to detect and identify the abundance of various lipids and metabolites to distinguish tumor sites in clinical setting. The tissues were cryosectioned for lipid and metabolite profiling by DESI-MS imaging. Complementary to MRM-profiling, DESI was able to visualize the location and relative ion abundances of a small number of lipids in whole tissue. In order to visualize the data, we utilized supervised statistical analysis methods (principal component analysis -PCA), univariate statistics and receiver operating characteristics (ROC) curve, and machine learning algorithms to show and visualize the different expression of lipid and metabolite profiles. Our findings from MRM-profiling MS indicate that some phosphotidylcholine (PC) and sphingomyelin (SM) lipids were significantly differentiated between GBM10 brain tumors and control, and GBM brain tumors and subcutaneous tumors. In contrast, In contrast, in-vivo exposure to the standard-of-care drug temozolomide (TMZ) resulted in significant changes in the level of some acylcarnitines. In addition, DESI-MS imaging mapped metabolites such as N-acetyl-L- glutamine and glutamate, as well as the lipid phosphotidylserine PS(40:0), which were discriminant between tumor and normal brain tissues. Understanding of metabolic reprogramming and differences between GBM and normal tissues, and drug effect on tumors in orthotopic and subcutaneous models might provide knowledge on the therapeutic targets for GBM.

Degree

Ph.D.

Advisors

Clase, Purdue University.

Subject Area

Bioengineering|Biology

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server
.

Share

COinS