Metabolic Changes in Response to Cell Cycle Synchronization and Treatment with Chemotherapeutic Drugs: Targeted and Non-Targeted Mass Spectrometric Analysis of Human Cell Lines

Abstract

In life sciences, the rise of the omics-era the recent decades has greatly contributed to increased understanding of biological systems. The latest addition to the omics group is metabolomics ? the science that aims at identifying and quantifying the entirety of low-molecular weight compounds in a biological system, also known as metabolites. Metabolomics is frequently employed in medical fields of research to enhance our knowledge of causes and pathogenesis of disease. In oncology studies, the cell cycle has become a popular target in anti-cancer therapy, as disruption of normal cell cycle regulation is often closely linked to cause and progression of cancer. Metabolic profiling of a synchronized cell culture, i.e. that all the cells reside in the same phase of the cell cycle, will enable the study of metabolic changes as a function of cell cycle. The work presented here has evaluated two methods of synchronizing human cell lines, i.e. serum deprivation and double thymidine block (DTB). It was found that the prostate cancer cell line DU145 could not be synchronized, whereas HaCaT (human keratinocytes) cells were successfully synchronized by both serum deprivation and DTB. To study changes in the metabolome as a function of cell cycle, both targeted metabolic profiling by gas chromatography coupled to a triple quadrupole mass spectrometer (GC-QqQ-MS) and non-targeted analysis by hydrophilic interaction and reverse phase liquid chromatography coupled to quadrupole-time-of-flight mass spectrometer (HILIC and RP LC-QTOF) of serum deprived HaCaT cells was performed. Principal component analysis (PCA) of the results from both analyses showed clustering of biological replicas residing in the same phase, as well as distinct separation of unsynchronized samples from synchronized. Some cell cycle phase-dependent concentration changes of metabolites were also observed. However, as no distinct trends were detected, it was concluded that the metabolome, when reflected by the chosen set of metabolites, does not change extensively from G1 to S to G2-M phase of the cell cycle. A pilot experiment to study dynamic stress-response in DU145, HaCaT and HEK293 (human embryonic kidney) cells induced by the chemotherapeutic drugs cisplatin and ATX-1 was performed. GC-QqQ-MS and HILIC and RP LC-QTOF gave coinciding results: DU145 and HaCaT were observed to cluster based on type of drug used, whereas HEK293 exhibited time-based clustering. This work presents findings that may contribute to future studies of metabolite pools and stress-induced metabolic changes in human cells.