Differentially Expressed Protein Database And The Methodology Research On The Characterization Of Phosphorylation Stoichiometry In A Large-scale

Proteomics,as an up-growing science,investigates proteins and their interaction by systematic and dynamical perspectives.It takes mass-spectrometry as the key technology and is supported by bioinformatics to identify,quantify and located the components of proteome with high-throughput and comprehensive ways.Our Ph.D researches focus on following two aspects:1)Constructing related databases,data standard,functional annotation system for experimental data management in proteomics;2)Developing methods for characterization of protein phosphorylation stoichiometry in a large scale. This thesis has two parts to summarize my researches of the both aspects.The first part introduces the construction of the differentially expressed protein database(DEPD).Comparative proteomics typically aims to systematically compare global protein expression profile between two cellular statuses(such as:disease and normal)and focuses on quantitative changes that occur as a function of disease or treatment conditions.Everyday,a wealth of experimental results about differentially expressed proteins(DEPs)are published and distributed among unformatted published literatures.Although the data is very valuable for finding biomarkers and drug target,little efforts have been taken to integrate them as a unified data resource.And the online database about DEPs is still unavailable.The reality obstructs scientists from getting reliable and valuable information of comparative proteomics rapidly.For resolving the problem,we developed a relational database(Differentially expressed database,DEPD)to store these data. Till now,we have extracted more than 3,000 DEPs information from about 104 comparative proteomics references manually.These data were annotated reasonably with comprehensive information about biological, environmental and methodological contexts which make the data interpreted correctly.For unifying the data structure of comparative proteomics,we designed a data model(CPXS 0.1).It was defined as a minimum required data unit for describing a comparative proteomics experiment.It has been applied to all entries of DEPD.To facilitate data query and submitting,a user-friendly web interface was designed to provide comprehensive tools for protein functional analysis.All of data in DEPD can be downloaded freely from the website http://protchem.hunnu.edu.cn/depd.The second part introduces a new method to characterize protein phosphorylated stoichiometry in large scale.It is well known that the protein phosphorylation research is always a hot but difficult topic in the proteomics area.Most of previous researches about it mainly focus on the following issues:1)The identification of phosphorylated proteins, peptides and sites;2)Finding and validating kinases,phosphoryases and their ligands as well as the relationship between them;3)The quantification and dynamics of phosphorylated proteins and peptides. However,the phosphorylation stoichiometry(the ratio of phosphorylation)of protein,as an important biological question for phosphorylation process,was usually omitted by these researches. Furthermore,recently few researches about it are usually limited by a small number of targets in vitro.So developing a robust,large scale method to measure protein phosphorylation stoichiometry in vivo is a key step to further proteome phosphorylation research.I and my colleagues took the nucleus and plasma of Hela cell as the sample, enriched target proteins by density gradient centrifugation and SDS-PAGE.After in Gel trypsin digestion,the peptide mixture was spited into two halves in which one of them was dephosphorylated and labeled by ¹⁸O water.Using the Bruker Ultraflex TOF-TOF mass spectrometry instrument,we get the phosphorylation stoichiometry data by comparing the intensity of peptide peaks in the two samples.For developing the methodology,I have fully studied the methods about preprocess of mass spectrometry raw data and ¹⁸O/¹⁶O labeling quantification.The house coding software named MSExplorer was designed to facilitate the data manages and process.The results of the preparing experiments confirmed the reliability of these algorithms and the feasibility of experimental process.Till now,using LC-MALDI strategy,we have identified 141 proteins and get phosphorylation stoichiometry data of 35 phosphorylated peptides.And most of them are newly discovered.