| Liver is the biggest organ in human body, which plays many key roles in life process, such as metabolism of materials, production and diversion of energy etc, as well as the popular host infected by other pathogenies. The knowledge on liver proteome could help us to explore not only the basic physiology function of liver, but also the potential mechanisms of liver diseases. One of the major tasks of proteomics is to disclose all proteins expressed in a cell or tissue. Although recent advancement in proteomics techniques, such as the combination of multidimensional liquid chromatography with tandem mass spectrometry makes this object more realizable, there are still big challenges confronted to comprehensively identify all the proteins targeted in a bio-sample, because of the complexity of liver proteome and wide range of protein concentrate. In chapter 2, a LC-MS platform was set up and optimized to per-separate protein and peptide. RPLC was used to reduce the complexity of human liver sample in protein level, and offline SCX-RPLC strategy in peptide level. Then, the platform was used to comprehensively identify the expression of human liver proteome. To obtain the most comprehensive human liver proteome, six times of SCX-RPLC MS/MS run were carried out, and more than 6,000,000 MS/MS spectra were collected. The MS/MS was interpreted by SEQUEST software, and a total of 24311 proteins corresponding to 13150 noreduntant proteins were successfully identified, in which 7001 proteins were identified by two or more peptides, which could be considered as a high confident dataset. The relationship between abundance of protein and their detecting frequency by mass spectrometer were also analyzed, which showed that the reduplicate SCX-RPLC MS/MS run could enhance the low abundance protein identification. Altogether, the results indicated that replicate RPLC/MS/MS strategy was a high throughput, sensitive analytical method for large scale profiling of human liver proteome.In large scale proteomics research, the confidence of protein identification is pivotal.In chapter 3, the relationship between mass accuracy and confidence of identification were studied using a novel LTQ-FTICR mass spectrometer with three different MS scan modes: LTQ with normal scan; FT with or without SIM scan. The resolution, mass accuracy and distribution of Xcorr and deltaCn value under different mass accuracy were compared and discussed. Our result indicated that the confidence of identification was affected by the mass accuracy of mass spectrometer and mass error set in database search. The comparison between SEQUEST and Mascot indicated that these two algorithms have similar yield for high accurate MS data, and more than 70% of the identified protein was overlapped by using these two searching softwares. Based on those analysis, a total of 4898 human liver proteins corresponding to 2640 protein groups with less than 1% of false positive and high mass accuracy were obtained.Single nucleotide polymorphisms (SNPs) are the most common form of variety genetic in human genome, which are an important source of interindividual human variation, including disease susceptibility and drug sensitivity. In chapter 4, we propose a blue print to discovery nsSNP from MS/MS data for the first time. A total of 42 nsSNPs in human liver were annotated by our method, among them, 5 peptides were interpreted by both mutation and wild type. The functional effect of each nsSNP was predicted by SIFT, and 2 nsSNPs were damaged to proteins. Our works represent a novel strategy to identify nsSNP in protein level based on mass spectrometry data.In chapter 5, based on the developed LC-MS/MS platform, a label free quantitative proteomics method was set up using yeast as model. The correlating relationship between concentrate of protein and spectra counts was confirmed, and the algorithms to normalize spectra counts ratio were compared. The method was used to biomarkers discovery in HepG2 and HepG2-HBx cell lines.
|
PDF Full Text Request