MultiOmics Integrative Analysis for Delineating Genetic Variants in Metabolic (Dysfunction) Associated Steatotic Liver Diseases(MASLD)

Abstract

Metabolic (Dysfunction) Associated Steatotic Liver Diseases (MASLD) is the leading cause of end stage liver disease (ESLD) in the United States. Moreover, the management of this disease comes at an excessive financial burden on the health sector. Unfortunately, due to very diverse and complex combination of environmental, metabolic and genetic factors that underlie individual’s progression of this disease, it has been very challenging for researchers in the field to develop an FDA approved drug for its treatment. In this thesis, using a precision medicine approach, we developed a multiomics integrative platform to delineate the phenotypic implication of high risk genetic variant associated with the development of MASLD. First, we developed a tool to automate and fast track analysis of raw metabolomics data generated from our Liquid Chromatography/Mass Spectrometry (LC/MS) instrument. The R-based package called Raw Global Metabolomics Data Analyzer (RODA) is the first of its kind that incorporates state of the art algorithms for post-acquisition/annotation analysis of LC/MS data into one single package. RODA takes as input user-generated text files of information that characterized individual experiment(s) and generates outputs that summarize the major findings from the experiment(s). RODA is user-friendly and its agnostic to the computational acumen of its user. To dissect the phenotypic implication of the presence of PNPLA3-I148M genetic variants in the hepatocytes of the liver and how this leads to the development of MASLD, we isolated and screen hepatocytes from donated human livers for the presence of the PNPLA3-I148M variant. Using these isolated hepatocytes from both the variants and the control samples, we performed metabolomics and transcriptomic analysis. Next, we incorporate the transcriptomics data into genome-scale model of human metabolism to generate a propensity score that reflects the activities of all the reactions per sample. The propensity score was then correlated with the metabolomics data to identify pathways or reactions with therapeutic target potentials. Our results show that PNPLA3-I148M hepatocytes, in response to the onset of cellular stress, upregulate endogenous levels of reduced glutathione but have a significantly reduced expression of GPX4. They also exhibit intracellular accumulation of primary bile acids and increased peroxisomal beta oxidation. Our experimental validation confirms this observation and further suggests the predisposition of the PNPLA3-I148M hepatocytes to ferroptosis. Moreover, targeting molecules critical to ferroptosis susceptibility in PNPLA3-I148M hepatocytes rescued and improved their resistance to development of MASLD. Furthermore, to show that gene-editing technology can be utilized to study rare but significantly risky MASLD-associated genetic variant, TM6SF2-E167K. We perform lipid profiling and transcriptomics analysis to characterize hepatocytes-differentiated gene-edited induced pluripotent stem cells (iHeps TM6SF2-E167K). Our analysis shows that the iHeps model recapitulates the known phenotype of TM6SF2-E167K variants in humans. Lastly, we developed a pipeline for spatial multiomics data integration. As proof of concept, the tool was used to study the effect of glutamine deprivation in a murine model of high grade serous ovarian cancer (HGSOC). Our pipeline was utilized to integrate the spatial metabolomics and the Imaging mass cytometry data to perform cellular neighborhood analysis and metabolic profiling of the same neighborhood. We found some noteworthy significantly different cellular neighborhood with distinct molecular profile in the treated group versus the control and these findings will be subjected to biological validation Overall, our work reflects the power of multiomics data integration to dissect metabolic associated disease and pinpoint targets for therapy development.