Six different types of cancer (i.e., breast, lung, colorectal, esophageal, gastric, and thyroid cancer) have high rates of incidence or mortality worldwide. It has been shown that activation of de novo lipogenesis is an early and common event in the cancer microenvironment. In this study, we performed lipid imaging and profiling for 134 tissue samples from six different types of cancer using matrix-assisted laser desorption/ionization-Fourier transform ion cyclotron resonance mass spectrometry, with 2,5-dihydroxybenzoic acid and 1,8-bis(dimethyl-amino)naphthalene as matrices in the positive and negative ion modes, respectively. Multivariate statistical analysis coupled with lipid distribution images revealed that significantly increased levels of monounsaturated fatty acids and monounsaturated phosphatidylcholines relative to polyunsaturated fatty acids and polyunsaturated phosphatidylcholines were observed in the cancer microenvironment compared with the adjacent normal tissue. The immunohistochemical assay indicated that fatty acid synthase, stearoyl-CoA desaturase-1, and choline kinase α were up-regulated in the cancer microenvironment compared with the adjacent normal tissue. Our findings suggest that de novo lipogenesis was activated in six types of cancer to promote a biosynthesis of lipids with monounsaturated acyl chains and to suppress a biosynthesis of polyunsaturated lipids in the cancer microenvironment.

Lipid composition in cell membrane is closely associated with cell characteristics. Here, matrix-assisted laser desorption/ionization- Fourier transform ion cyclotron resonance mass spectrometry was employed to in situ determine membrane components of human mammary epithelial cells (MCF-10 A) and six different breast cancer cell lines (i.e., BT-20, MCF-7, SK-BR-3, MDA-MB-231, MDA-MB-157 and MDA-MB-361) without any lipid extraction and separation. Partial least-square discriminant analysis indicated that changes in the levels of these membrane lipids were closely correlated with the types of breast cell lines. Elevated levels of polyunsaturated lipids in MCF-10 A cells relative to six breast cancer cells and in BT-20 cells relative to other breast cancer cell lines were detected. The Western blotting assays indicated that the expression of five lipogenesis-related enzymes (i.e., fatty acid synthase 1(FASN1), stearoyl-CoA desaturase 1 (SCD1), stearoyl-CoA desaturase 5 (SCD5), choline kinase α (CKα) and sphingomyelin synthase 1) was associated with the types of the breast cells and that the SCD1 level in MCF-7 cells was significantly increased relative to other breast cell lines. Our findings suggest that elevated expression levels of FASN1, SCD1, SCD5 and CKα may closely correlated with enhanced levels of saturated and monounsaturated lipids in breast cancer cell lines.

Brain lipid homoeostasis is critical during recovery process after traumatic brain injury (TBI). In this study, we integrated liquid extraction and electrosonic spray ionization technology to develop an ionization device coupled with a Fourier transform ion cyclotron resonance mass spectrometer for imaging of docosahexaenoic acid (DHA)-containing lipids on rat brain tissues. The ion images of the brain tissue sections from the normal rats and the rats after TBI at acute phase (0 and 1 day) and chronic phase (3, 5, and 7 days) were obtained. The imaging results indicate that the levels of DHA and lyso-phosphatidylethanolamine (22:6) in the injury area of TBI rats increased significantly at the acute phase and subsequently decreased at the chronic phase. But the levels of DHA-containing phospholipids including phosphatidylethanolamine (PE)(P-18:0/22:6), PE(18:0/22:6), and phosphatidylserine (18:0/22:6) decreased at the acute phase and gradually increased at the chronic phase in the injury area accompanied by the morphogenesis and wound healing. These findings indicate that the DHA may participate in the recovery process of brain injury. This is the first report to in situ detect the changes in the levels of DHA and DHA-containing lipids in the TBI model.