From Spectra to Pathways: Assessing Metabolic Function and Nutrient Modulation of Liver In Vitro Systems by LC-MS metabolomics

Induced pluripotent stem cells (iPSCs) are a groundbreaking resource in regenerative medicine and liver research. The potential use of a stable and functional source of hepatocytes has led to the development of protocols for generating human iPSC-derived hepatocyte-like cells (HLCs). However, creating functional hepatic models from iPSCs remains challenging due to their limited metabolic capacity. In Chapter 2, we characterized two human hepatocyte models: HLCs and metabolically active HepG2 (mHepG2) for their drug metabolism activity. Both HLCs and mHepG2 were cultured under a nutrient regimen rich in glycine. With a multi-omics approach, we studied the transcriptome, proteome, and metabolome of 11 drug-relevant cytochrome P450 (CYP) isoenzymes in these cell systems. A liquid chromatography-mass spectrometry (LC-MS)-based metabolomics approach was proposed, using model drugs as isoenzyme reporters, to provide a comprehensive overview of drug metabolism in mHepG2 and HLCs. While HepG2 is known for its limited metabolic capacity, our study positions mHepG2 as a highly efficient cell model, exhibiting activity on 8 drug-metabolizing CYPs. This research confirms that nutritional modulation is a crucial factor in enhancing the metabolic function of in vitro models. Furthermore, HLCs demonstrate extensive CYP coverage at the transcript level and can withstand a diverse range of chemical challenges. After characterizing liver in vitro models for their ability to metabolize xenobiotics, this thesis proposed in Chapter 3 a computational LC-MS/MS workflow for the automatic identification of drug metabolic products, including unknown metabolites. As test compounds for the study, six well-known drugs were administered to primary human hepatocytes (PHHs): amitriptyline, carbamazepine, cyclophosphamide, fipronil, phenytoin, and verapamil. The workflow employs the computational tools BioTransformer and SIRIUS to predict and identify drug metabolite structures. Notably, more than 62% of the drug’s metabolites were traced using BioTransformer in SIRIUS. Metabolic modulation, involving adapting media constituents in cell culture conditions, has demonstrated to be a promising tool for improving in vitro systems and iPSC differentiation. As previous research has shown that glycine-supplemented media enhanced HLCs and HepG2 drug metabolic capacity, we further aimed to reveal why glycine would induce the development of an active xenobiotic machinery. To this end, mapping metabolic intermediates from pathways involving glycine towards xenobiotic metabolism became necessary. Thus, in Chapter 4, we developed an LC-MS semi-targeted pipeline for identifying central carbon metabolism (CCM) intermediates. This analytical workflow facilitated the mapping of metabolic pathways modulated by glycine supplementation. By employing semi-targeted LC-MS-based and stable (13C) isotope-resolved metabolomics in Chapter 5, we observed in HLCs, cultured under high glycine concentration, an increase in oxidative metabolism during differentiation and a boost in several biosynthetic pathways. A glycine-rich environment enhanced collagen, glycogen, and bile acid biosynthesis, as well as one-carbon metabolism and heme biosynthesis, all of which are indicative of a hepatic phenotype. An increase in heme synthesis is linked to enhanced xenobiotic metabolism, as CYPs are heme-containing enzymes. This study highlights the metabolic plasticity of in vitro iPSC models, which undergo metabolic reprogramming in response to changes in nutrient composition. In line with the efforts of improving the metabolic activity of in vitro hepatic models, HLCs, classically cultured as 2D adherent cells, were submitted to an organobody (3D) formation protocol. To finalize this thesis, in Chapter 6, we applied previously developed LC-MS pipelines to analyze the xenobiotic and mitochondrial metabolism of 3D HLCs, embedded within a self-assembled peptide hydrogel. Results were compared to 2D HLCs and benchmarked with 2D PHHs’ response. In summary, this thesis developed analytical LC-MS approaches for studying cell metabolism, enhancing metabolite identification, in vitro model development, and characterization. This work represents an important step forward towards more sustainable and reliable pharmaceutical testing.