Fully automatable multidimensional liquid chromatography with online tandem mass spectrometry for proteomics and glycoproteomics

Abstract

This dissertation reports the development of novel, fully automatable, online multidimensional liquid chromatography (MDLC) technologies and methodologies to accelerate proteomics and glycomics mapping from complex biological samples. Chapter 2 reports the development of an online two-dimensional (2D) liquid chromatography (LC) system—with separations based on hydrophilic interactions in the first dimension and low-pH reversed-phase (RP) separation (peptide hydrophobicity) in the second—that operated with high resolution and orthogonality. This hydrophilic interaction liquid chromatography (HILIC) –RP platform featured an RP trap column plus a mixing loop before the first dimension to facilitate direct aqueous sample loading; an additional sample loop plus a strong cation exchange (SCX) trap column was implemented to circumvent the problem of solvent incompatibility between the two columns. The performance of this system was benchmarked through analysis of the proteome of Saccharomyces cerevisiae, resulting in the identification of more than 2000 proteins with abundances spanning from 40 to 〖10〗^6 copies/cell.

Chapter 3 reports a novel online three-dimensional (3D) HILIC-SCX-RP coupled with porous graphitic carbon (PGC) LC platform derived from the HILIC-RP design, featuring additional SCX fractionations, operating through a charge-centric separation mechanism, to extend the separation efficiency and platform orthogonality; the PGC column was integrated to recapture non-retained hydrophilic analytes for concomitant analyses of both hydrophilic and hydrophobic analytes within the same sample injection event. This integrated technology exhibited superior performance for the proteomics analyses of the total lysate of primary cerebellar granule neurons (CGNs) and cynomolgus monkey brain tissue, with enhanced protein and proteome coverage. One of the most comprehensive CGNs proteome to date was characterized: in total, 2201 proteins and 16,937 unique peptides. This 3D HILIC-SCX-RP/PGC system allowed the first detailed and simultaneous N-glycomics and N-glycoproteomics analyses of cynomolgus monkey plasma, establishing a glycan library containing 122 proposed N-glycans with confirmed complementary sites of N-glycosylation; 38 N-glycolylneuraminic acid (NeuGc)–containing N-glycans were also verified through tandem mass spectrometry for the first time.

Finally, Chapter 4 describes the first online 2D PGC-RP LC system with dual sample traps that allowed the implementation of shotgun proteomics and glycomics analyses using less-sophisticated instrumentation. The PGC-platform operated through a mixed mode of mechanisms for peptide separation, taking advantage of both planar contact area–based interactions and hydrophobicity, allowing elimination of the aforementioned RP trap column, mixing loop, and related switching valves for sample loading; thus, this system could be readily assembled on a commercially available MDLC system with minimal modifications. The dual-trap column configuration was adopted, offering desirable high-throughput with almost no idle time for sample fractionation, trapping, or desalting. This 2D PGC-RP technology performed well, as judged by the results of proteomics and glycoproteomics analyses of cerebellar granule neurons lysates and cynomolgus monkey plasma. A comparison of the HILIC-SCX-RP and PGC-RP analyses in Chapters 2–4 revealed that these technologies identified primarily different peptides, consistent with their different mechanisms of separation. Thus, HILIC-SCX-RP and PGC-RP technologies offer a potent combination of protein coverage, efficiency, reproducibility, and sensitivity for large-scale proteomics applications.