Application Of Proteomics Techniques In Diagnosing And Screening Colorectal Cancer And Esophageal Carcinoma

Background and ObjectiveColorectal cancer and esophageal carcinoma are common malignant tumors in the world. Colorectal cancer(CRC), includes neoplasms of colon and rectum, is a much common malignant tumor which do great harm to people's health. The mobidity of CRC is rising year by year in China along with the improvement of people's living standard and alteration of life style. The esophageal carcinoma(ESCC) is one of the most common malignant tumors in China. The average age of patients is over 40 years old. It is one of the main malignant tumors with high mortality in digestive system in China. Although there were some progress in diagnosis and therapy, the total treatment effectiveness of these two cancers did not show any significant improvement and the 5-year survival rate after radical operation was still low. In China, since there is disproportion of economic development and health education among different areas and population, early diagnosis and early therapy are always important for cancer patients.Currently, there are some difficulties in early diagnosis and screening of CRC and ESCC. For ESCC, net-balloon examination is a valid method for mass screening in high-morbidity area. But many patients feel painful and are hard to accept it. There are other techniques such as endoscope, barium swallow, radio isotope, CT, MRI, PET, EUS and tumor marker. For CRC, there are many techniques for diagnosis and screening such as fecal occult blood tests(FOBT), flexible sigmoidoscopy(FS), double contrast barium enema(DCBE), CT, MRI, PET, colonoscopy, genetic screening, and so on. They all have their shortcomings and not satisfactory on the whole. So it's eagerly to develop new technique to screen and diagnose them early, mini-traumaly, economically and easily accepted by individuals with high sensitivity and specificity.The development of proteomics leads to a new period for tumor research. The proteomics is the temporal cell or tissue-specific protein complement of the genome, encompassing all proteins expressed at any given time. Proteomics is an emerging scientific field that involves the identification, characterization, and quantification of proteins in whole cells, tissues, or body fluids. The proteome ultimately dictates the function of the cell, and therefore dictates phenotype. The proteome undergoes dynamic changes as it continuously responds to autocrine, paracrine, and endocrine factors, . blood-borne mediators, temperature, drug treatment, and developing disease over time. While increasing interest for identification of tissue markers and for providing data for analysis, proteomics gives the unique opportunity to develop blood-based biomarkers to be used for diagnosis, prognosis, and therapy modulation. Tumor cell produces distinct proteins in the development of disease, and result in different proteomes between cancer and nontumorous individuals or among diffiernt stages. This change is potential for tumor research.Surface enhanced laser desorption/ionization-time of flight-mass spectrometry (SELDI-TOF-MS) and isobaric tags for relative, absolute quantification (iTRAQ) and mass spectra (MS) are applied in our study. SELDI-TOF-MS is one of the key proteomics platform. It's mainly consisted of proteinchip, proteinchip reader and bioinformatics software. The proteinchip can bind many kinds of proteins in sample, and is more sensitive than conventional technique. SELDI-TOF-MS has increased in popularity in the examination of clinical samples, can be used to detcet different samples directly without complicated sample preparation. SELDI-TOF-MS has been widely used in cancer research of ovary, prostate, esophagus, liver, colon and rectum. As sporadic moderately differentiated adenocarcinoma is the mainly pathological type of colorectal cancer, we used SELDI-TOF-MS and Immobilized Metal Affinity Capture(IMAC30-Cu2+) chips and corresponding software to detect and analyze serumal proteome of sporadic moderately differentiated CRC and nontumorous individuals, and ESCC and nontumorous individuals, expecting to identify the characteristic protein and establish the pattern-matching algorithms of CRC and ESCC by serum protein profiling, which is potential to develop new technique of diagnosing CRC and ESCC. The iTRAQ reagents are a set of isobaric reagents which are amine specific and allow for the identification and quantification of up to four different samples simultaneously. The amine specificity of these reagents makes most peptides in a sample amenable to this labeling strategy with no loss of information from samples involving post-translational modifications, such as the scrutiny of signal transduction pathways that often involve phosphorylation phenomena. In addition, the multiplexing capacity of these reagents allows for information replication within certain LC-MS/MS experimental regimes, providing additional statistical validation within any given experiment. The results presented herein demonstrate a few examples of the wide variety of quantitative information that can be realized when undertaking such experimental approaches. These include temporal analysis of druginduced-protein expression, discovery and elucidation of disease markers, and protein-protein interactions in multi-protein complexes. In our study, iTRAQ and MS are used to discovery different proteins between tumor tissue and normal tissue.MethodsSerum samples were applied to IMAC30-Cu2+protein chips to generate mass pectra by SELDI-TOF-MS. Protein peak identification and clustering were performed using the Biomarker Wizard software.Training group:131 serum samples including CRC(35 cases), colorectal adenoma(CRA)(14 cases), colorectal benign diseases(CBD)(12 cases), ESCC(26 cases) and also 44 healthy individuals(HC) as a control groupTest group:90 serum samples including CRC(30 cases), CRA(10 cases), CBD(10 cases), ESCC(10 cases) and HC(30 cases) were randomly selected at the same period as the training group.Firstly, proteomic spectra of serum samples from 35 CRC and 44 HC were used as a learning mode and classification tree of CRC diagnosis with 3 distinct protein masses was generated by using Biomarker Pattern software. Proteome spectra of serum samples from 14 CRA,12 CBD and 44 HC were used as a learning mode and classification tree of CRA screening with 3 distinct protein masses was generated by using Biomarker Pattern software. Proteome spectra of serum samples from 26 ESCC and 44 HC were used as a learning mode and classification tree of ESCC diagnosis with 3 distinct protein masses was generated by using Biomarker Pattern software.Secondly, tumor tissue from CRC (4 cases), ESCC (4 cases) and norma tissue from intestinal mucosa (4 cases), oesophageal mucosa (4 cases) are analyzed by iTRAQ and LC-MS/MS to find different proteins between CRC tumor tissue and normal intestinal mucosa, ESCC tumor tissue and normal oesophageal mucosa.Thirdly, these classification trees were blindly validated by the protein fingerprinting of the test groups respectively.Fourth, different proteins in tissue are compared with those in serum to find if there are some associations between these two series proteins.ResultsFirstly, via comparative proteomics research of CRC and controls, the software identified an average of 33 mass peaks/spectrum, in which 14 peaks are significantly distinct(P<0.05), and 7 of the identified peaks were used to construct the classification tree of CRC diagnosis. This classification tree separated effectively CRC from HC, achieving a sensitivity and a specificity of 90.91%and 82.86%respectively in test mode. Blinded validation suggested that the total accuracy for prediction of the model was 80%, the positive predictive value was 78.13%, the sensitivity and specificity were 83.33%and 76.67%, respectively. Via comparative proteomics research of CRA, CBD and controls, the software identified an average of 45 mass peaks/spectrum, in which 40 peaks are significantly distinct(.P<0.05), and 3 of the identified peaks were used to construct the classification tree of CRA screening. This classification tree separated effectively CRA from HC and CBD, achieving a sensitivity and a specificity of 71.43%and 96.43% respectively in test mode. Blinded validation suggested that the total accuracy for prediction of the model was 84.62%, the sensitivity and specificity were 60%and 100%, respectively.Via comparative proteomics research of tumor tissue from CRC and normal tissue from intestinal mucosa, the software identified 87 different proteins. However, these proteins don't have any association with those found in serum.Secondly, Via comparative proteomics research of ESCC and controls, the software identified an average of 38 mass peaks/spectrum, in which 18 peaks are significantly distinct(P<0.05), and 3 of the identified peaks were used to construct the classification tree of ESCC diagnosis. This classification tree separated effectively ESCC from HC, achieving a sensitivity and a specificity of 92.31%and 65.91%respectively in test mode. Blinded validation suggested that the total accuracy for prediction of the model was 72.5%, the positive predictive value was 46.67%, the sensitivity and specificity were70%and 73.33%, respectively.Via comparative proteomics research of tumor tissue from ESCC and normal tissue from oesophageal mucosa, the software identified 49 different proteins. However, these proteins don't have any association with those found in serum.Conclusions1. The classification tree of CRC diagnosis, ESCC diagnosis and CRA screening based on protein profillings of IMAC302+proteinchip and SELDI-TOF-MS has been successfully established.2. Cross-validation and blind testing have proved high sensitivity and specificity of the models, which are helpful for CRC and ESCC diagnosis, screening, and classification. 3. Identified different proteins between tumor tissue and normal tissue.4. SELDI-TOF-MS and iTRAQ combined with MS have great promise in the reasearch of tumor proteome.