Proteomic studies of Burkholderia cenocepacia and human breast cancer using two-dimensional liquid separation with mass spectrometry and statistical analysis.

Abstract

Separation and identification of proteins in proteomics have been important and challenging due to the wide dynamic range of proteins. In this thesis, a 2D liquid separation method coupled with mass spectrometry (MS) has been employed to separate intact proteins based upon their isoelectric point and hydrophobicity using chromatofocusing and non-porous silica reversed phase HPLC. Identification of the separated proteins was achieved by intact protein molecular weight, peptide mass fingerprinting (PMF), and tandem MS analysis. These data were combined into a 2D map to represent a proteomic profile with high resolution and reproducibility. The 2D liquid separation method coupled with MS has been applied to one environmental and three clinical <italic>B. cenocepacia</italic> isolates to study bacterial virulence factors. From the comparison of proteomic profiles between four isolates, 38 differentially expressed proteins were identified by PMF and MS/MS sequence analysis. The results indicate that the up-regulated proteins in the clinical isolates are involved in gene translation and bacterial virulence such as transmissibility, resistance, and quorum sensing. Mass maps of one normal (MCF10A) and two malignant (MCF10DCIS.com and MCF10CA2) human breast cell lines were generated by 2D liquid separation method coupled with MS and compared with each other to discover biomarkers of cancer development. 131 differentially expressed proteins were revealed from the direct comparison of three mass maps. Cytokeratin8, cytokeratin18, S114, annexin-1I, and ER-60 are related to the initial stage of cancer malignance. Statistical analyses were performed to interpret large amount of data from these works. The results from hierarchical cluster analysis of mass maps of seven MCF10 cell lines showed the similarity between cell lines based on protein expression patterns. It was confirmed that specific protein expression patterns can discriminate two different cell lines. In order to assign the specific expression patterns two sample <italic>t</italic>-tests were performed for the mass maps. Investigation of the assigned patterns showed that proteins regarded as candidates of breast cancer biomarkers were contained in those patterns. These results have successfully shown that the 2D liquid separation method coupled with MS is a promising technique to separate and to quantitate intact proteins in proteomic analysis.