Glycosphingolipidomic investigations of gangliosides and glycosphingolipids in development and disease

Gangliosides, complex sialic acid-containing glycolipids, and other glycosphingolipids are active physiological membrane components with an array of functions in development and disease. Altered profiles are found in many disorders including those of neurological and cancerous aetiologies. Glycosphingolipids are also important for lateral membrane organization, cell communication and as binding sites for extra-cellular components. These lipids have long been implicated as targets in autoimmune diseases such as Guillain-Barré Syndrome (GBS) and Multifocal Motor Neuropathy. In GBS, auto-antibodies bind native membrane gangliosides signalling immune-mediated breakdown of nerves causing acute flaccid paralysis. While the fundamental pathology is understood, differences in clinical presentation, and preference for motor over sensory nerves, have yet to be explained. Understanding the precise nature of native gangliosides, including low abundance species and modifications, is an important first step. Meanwhile genetically engineered mouse models are under development that should increase our understanding of disease pathogenesis. To be truly functional it is essential these models contain a full range of complex and simple glycosphingolipids in the neurological tissue. Mass spectrometry has recently been applied with great effect to lipidomics; the comprehensive profiling of all lipids involved in a system. However, heavily glycosylated, low abundance and chemically unusual lipids such as the gangliosides tend to be neglected in otherwise thorough lipidomic studies. It was the aim here to optimise separation and mass spectrometry methodologies for ganglioside analysis. Workflows were developed for high performance thin layer chromatography (HPTLC) combined with direct imaging mass spectrometry (IMS) detection and identification, and for high performance liquid chromatography (HPLC) with online high resolution mass spectrometry detection and identification with dissociation to confirm structures (MSMS). A range of lipid standards were analysed using this second method to build a database of characteristic ionization behaviour, retention times, and product ion spectra to aid the analysis of unknowns in complex mixtures. Methods were then applied to molecular phenotyping in novel mouse models of GBS, and to glycosphingolipidomics in peripheral sensory and motor nerves. Finally the recently developed technique of imaging mass spectrometry, using matrix assisted laser desorption ionisation (MALDI) and secondary ion mass spectrometry (SIMS) ion sources, was investigated for its capability for direct ganglioside analysis in brain and spinal cord tissue sections. Results are presented below demonstrating the significant benefits of the mass spectrometry-based workflows over more conventional profiling methods as well as comparing and contrasting the two techniques developed here. Limitations and potential areas for future development are debated. Findings from profiling knockout and rescue mouse models and from single nerve glycosphingolipidomics are discussed along with further experiments and directions for these studies. The discovery of a full range of complex gangliosides in neurological tissue from rescue mice, albeit at low levels compared to the wild type, confirmed their molecular usefulness for modelling neurological autoimmune diseases. The sensitivity and reproducibility of the mass spectrometry technique enabled relative quantitation, revealing details into the abundance of different ganglioside species and inclusion of ceramide structures in each mouse type. The ability to detect very low abundance lipids with an additional dimension of structural description also suggested that O-acetylation of the second sialic acid on native disialylated lipids is more prevalent than previously thought. Finally imaging mass spectrometry results are presented. Although sensitivity was limited, both simple and complex gangliosides were detected in spinal cord sections; the first known IMS detection of these lipids outside of the brain. Results also demonstrate the abundance of parallel lipidomic information that can be obtained using these methods. Possible solutions to increasing the sensitivity limit are discussed that may increase IMS usefulness to glycosphingolipid studies in future