孔洞石墨碳管柱液相層析串連式質譜分析技術之研發與其在分離與結構鑑定神經膠質母細胞瘤之醣脂體醣鏈之應用

Abstract

Mass-spectrometry (MS)-based glycomics aims to map all the molecular glycan species under specific physiological or pathological conditions. Glycosphingolipids (GSLs) are found to be involved in many biological processes. Overexpression of GSLs in different cancers has been observed widely, including in glioblastoma multiforme (GBM). The conventional MS method used for GSLs detection is mostly MALDI. However, the poor capacity for separation of isomeric molecules of MALDI is not sufficient for desired complete glycomic mapping. In this thesis, we developed a porous graphitized carbon (PGC)-LC-MS based workflow to systematically analyze all glycans released from GSLs in cells. By using acidic solvent system in positive mode, six GM1 (Hex3HexNAc1NeuAc1) standards were mixed together and then loaded onto PGC-LC-MS for separation. Under optimized gradient condition, the six standards were completely separated and their structures were also confirmed by CID-MS/MS. After platform being established, GSLs glycans of DBTRG GBM cells were applied to this platform for the isomer separation. The GM1 isomers (Hex3HexNAc1NeuAc) of DBTRG cells were separated, identified and quantified for the first time. Moreover, the whole GSLs of DBTRG cells could be separated and analyzed on the PGC-LC-MS system. Minor species, such as globo-series molecules, were also detected and several other isomeric glycans were identified. Next, this method was used to compare the changed expression of GSLs between DBTRG parental cells and stem-like neurospheres. The increased level of sialylated and decreased level of neutral GSLs were detected in neurospheres. Finally, the platform was further applied to profile and compare the intact GSLs of other GBM cells. The short GSLs usually have fewer isomers because of the simple molecular structures. On the other side, long GSLs usually have more isomeric ions because of their complicated structures. There are some unusual isomeric glycans being detected, such as Gb5/iGb5 distribution in each kind of cells examined. In short, we here developed a useful PGC-LC-MS platform for isomeric GSLs separation by LC and identification by MS/MS and it can be applied to more biological samples in the future.