Lipid profiling using Raman and a modified support vector machine algorithm

Area of Science:

• Cell Biology
• Biochemistry
• Spectroscopy

Background:

• Lipid droplets are essential organelles involved in cellular lipid metabolism, composed of a phospholipid membrane and a core of triglycerides and sterol esters.
• Fatty acids are crucial for phospholipid synthesis, signaling pathways, and energy storage via triglyceride formation.
• Accurate, non-invasive methods for profiling and quantifying cellular lipids, particularly in cancer, are currently lacking.

Purpose of the Study:

• To evaluate the efficacy of Raman spectroscopy as a non-invasive tool for analyzing lipid droplet composition.
• To accurately determine changes in fatty acid and cholesterol content within prostate cancer cells.
• To develop and validate a novel analytical method for lipid profiling in cancer biology.

Main Methods:

• Utilized Raman micro-spectroscopy to analyze lipid droplets in prostate cancer cells.
• Employed a modified least squares fitting (LSF) routine incorporating a support vector machine (SVM) algorithm.
• Identified highly discriminatory wavenumbers specific to different fatty acids for precise quantification.

Main Results:

• Demonstrated high accuracy in determining fatty acid and cholesterol composition changes in lipid droplets.
• Successfully profiled lipid composition in prostate cancer cells treated with various fatty acids.
• Validated the potential of the developed LSF-Raman spectroscopy method for lipid analysis.

Conclusions:

• Raman micro-spectroscopy, enhanced by a novel LSF-SVM approach, offers a powerful non-invasive method for lipid droplet analysis.
• This technique can accurately profile and measure fatty acids and cholesterol, crucial for understanding cancer biology.
• The developed methodology holds significant promise for advancing non-invasive lipid profiling in cancer research.