Development of multidimensional liquid chromatography approaches for the enhanced qualitative and quantitative shotgun neuroproteomic analyses

Abstract

The immense cellular complexity of the brain poses significant challenges when attempting to elucidate the molecular mechanisms underlying normal and pathological aspects of neuronal function. To facilitate comprehensive and in-depth mapping of proteomes in bottom-up proteomics, the first Section of this Thesis describes the development of advanced separation techniques, employing multidimensional liquid chromatography (MDLC), to minimize sample complexity and overcome the problem of undersampling. Using a high-/low-pH reversed phase–reversed phase (RP–RP) two-dimensional liquid chromatography (2DLC) platform as a building block, two novel online MDLC strategies (Chapters 2 and 3) have been designed and implemented to facilitate high-throughput qualitative and quantitative proteomic analyses; the operation of the online MDLC platform has the attractions of automation and minimal sample loss.

Chapter 2 describes an alternative RP–RP approach to maximize the MDLC performance: through attachment of additional chromatographic dimensions to pre-existing platforms. This approach involved combining the two most prevalent 2DLC platforms, RP–RP and SCX–RP, into a single online platform for peptide separation based on hydrophobicity and charge. This novel MDLC (RP–SCX–RP) platform possessed enhanced peptide and protein identification capabilities with respect to 2D RP–RP, as demonstrated through profiling of the PC12 cell proteome—a widely used neurobiological in vitro model. The most comprehensive proteome profile to date was acquired, consisting of 6345 proteins and 97 309 peptides. The issue of isobaric interference of mass-tagging background contamination was also circumvented when using this platform, thereby improving the accuracy of iTRAQ-based protein quantitation experiments significantly. Chapter 3 demonstrates the first application of multiple fraction concatenation in an online MDLC platform to increase the orthogonality of the separations, as well as alleviate the issues of uneven distribution of the peptides among the first-dimension RP fractions—problems that occur when using the conventional RP–RP platform. This strategy recovered additional peptides, most notably those with acidic and hydrophilic characteristics. A comparison of conventional and concatenation RP–RP 2DLC analyses revealed that the concatenation approach improved the analytical confidence of the normalized spectral abundance factor, a label-free protein quantitation technique using spectral counting.

Applying these developed strategies, in combination with other online MDLC platforms, Chapter 4 discusses qualitative and quantitative proteomic analyses conducted on the cerebral cortex proteome of a Macaca fascicularis ischemic stroke chronic phase model. After application of conservative grouping criteria, 8790 non-redundant proteins were mapped to 4906 protein groups. Of the 2105 quantified proteins, subsequent analysis revealed 31 and 23 differentially expressed proteins in the high and low infarct volume groups, respectively. The dysregulated proteins identified were closely associated with the major tissue injury processes after stroke, including neurogenesis, synaptogenesis, and inflammation. Further protein interaction analysis supported the notion that actin cytoskeleton remodeling may be modulated and involved in determining neuronal cell fate. These analyses provide insight into the long-term injury and recovery processes after stroke, serving as a basis for future studies with the aim of developing potential neurorestorative therapies.