Abstract
OBJECTIVE: The aim of this study was to explore the lipidomic profiles of the CAL-27 human tongue cancer cell line and the human oral keratinocyte (HOK) cell line. METHODS: The lipidomic differences between the CAL-27 and the HOK cell lines were investigated using non-targeted high-performance liquid chromatography-mass spectrometry lipidomic analysis. The resulting data were then further mined via bioinformatics analysis technology and metabolic pathway analysis was conducted in order to map the most affected metabolites and pathways in the two cell lines. RESULTS: A total of 711 lipids were identified, including 403 glycerophospholipids (GPs), 147 glycerolipids, and 161 sphingolipids. Comparison of the enhanced MS (EMS) spectra of the two cell lines in positive and negative ionization modes showed the lipid compositions of HOK and CAL-27 cells to be similar. The expressions of most GP species in CAL-27 cells showed an increasing trend as compared with HOK, whereas a significant increase in phosphatidylcholine was observed (p < 0.05). Significant differences in the lipid composition between CAL-27 and HOK cells were shown as a heatmap. Through principal component analysis and orthogonal partial least squares discriminant analysis, noticeably clear separation trends and satisfactory clustering trends between groups of HOK and CAL-27 cells were identified. The numbers of specific lipid metabolites that could distinguish CAL-27 from HOK in positive and negative modes were 100 and 248, respectively. GP metabolism was the most significantly altered lipid metabolic pathway, with 4 metabolites differentially expressed in 39 hit products. CONCLUSION: This study demonstrated the potential of using untargeted mass spectra and bioinformatics analysis to describe the lipid profiles of HOK and CAL-27 cells.