Lipid quantification method using FTIR spectroscopy applied on cancer cell extracts

• FTIR rapid method to provide quantitative evaluation of the various lipid classes
• Partial least square (PLS) models based on FTIR spectra
• Quantification of lipids in complex mixtures such as cell extracts
• The lipid class distribution in cells exposed to anticancer drugs

Reprogramming energy metabolism constitutes one of the hallmarks of cancer. Changes in lipid composition of cell membranes also appear early in carcinogenesis. Quantification of various molecules such as lipids evidences the modifications in the metabolism of tumour cells and can serve as potential markers for cancer diagnosis and treatment. Fourier Transform Infrared (FTIR) spectroscopy is a powerful tool used for the detection and characterization of various types of molecules. This technique remains an attractive approach as it is cheap (equipment and reagents), does not require high grade solvents or expensive internal standards, equipment is widely available in standard laboratories and the method is robust and suitable for routine analyses. In this work we established partial least square (PLS) models based on FTIR spectra able to quantify lipids in complex mixtures such as cell extracts. In the first part, we attempted to build PLS models with FTIR spectra of 53 mixtures of 8 well-characterized pure lipids. Second, the PLS models were verified using FTIR spectra of mixtures that did not contribute to the calibration. The third step was the validation of the models on lipid cell extracts. In order to obtain reference values for cell extracts, high performance liquid chromatography was carried out by AVANTI. The lipid distribution were globally similar with both techniques, PLS models and chromatography. Finally, the models were applied to determine the lipid composition of cells exposed to four treatments. We could not evidence significant changes in the lipid composition of cell extracts after treatment, in terms of polar head groups. However, the models established in this study appear reliable and could be applied for high throughput measurements. This article is part of a Special Issue entitled Tools to study lipid functions.