Dissertation > Medicine, health > Oncology > Gastrointestinal Cancer > Intestinal neoplasms

Studies on Proteomic from Serum in Colorectal Cancer by SELDI-TOF Mass Spectrometry

Author ZhouZhiYong
Tutor CaoJiaQing
School Nanchang University
Course Surgery
Keywords Colorectal cancer Biological Markers Protein fingerprint Surfaceenhanced laser desorption ionization-time of flight-mass spectrometry (SLEDI–TOF-MS)
CLC R735.3
Type Master's thesis
Year 2012
Downloads 38
Quotes 0
Download Dissertation

Objective: Discussion colorectal cancer patients and healthy subjects serumprotein fingerprint wave differences, and then analyzes the differences between serumprotein molecules expression and its content. The difference from colorectal cancerscreening serum related proteins, establish protein database, constructing thecolorectal cancer diagnosis model and for early diagnosis of colorectal cancer providea kind of effective method.Methods: The application of protein fingerprint technique (SELDI-TOF-MS)analysis of colorectal cancer and healthy subjects serum samples, using IMAC3(Immobilized Metal Affinity Capture, metal affinity surfaces) chip.Collecting in May2010to May2011in Nanchang university second affiliated hospital gastrointestinalsurgery in hospital patients with colorectal cancer early morning hollow venous blood,both by endoscopic surgery or histopathological diagnosed, radiation and chemoth-erapy and other treatment without intervention, a total of134patients, the averageage of53years old, male for76cases, the female is58cases; The healthy subjects: atotal of87patients who kindergarden, collect in our health medical center, theaverage age for49years old, men52cases, female35patients, blood samples arefrom the early morning hollow state patients, no effect on the protein content inserum of other related diseases.The training set:74cases of patients with colorectalcancer [Dukes′A(n=8)、Dukes′B (n=18)、Dukes′C(n=31)、Dukes′D(n=17)],48healthy subjects; validation sets:60in patients with colorectal cancer [Dukes′A (n=7)、Dukes′B(n=24)、Dukes′C(n=17)、Dukes′D(n=12)], in39healthy subjects.Thefirst application of SELDI-TOF-MS detection of mixed control serum, usingsoftware to obtain mixed control serum patterns were analyzed and calculated thedifference, with the coefficient of variation (CV) value, set the CV <15%.Usingcase-control study, using protein fingerprint reading apparatus PBSII-C analysistraining set of serum protein samples, obtained the training set compared studysample protein content.Application of Biomarker software of Wizard Protein Chipsoftware treatment to obtain the protein molecular weight, M/Z, content data using the t test or rank test was used for statistical analysis, calculating the average value, Pvalue, SD.Biomarker Pattern software using Biomarker Wizard software to analyzethe data using a decision tree model for data mining, screening of colorectal cancerearly diagnosis of specific biomarkers for diagnosis of colorectal cancer, constructionmodel.After the construction of diagnosis model, using the validation set of serumsamples, for the blind test, in order to check the validity of the model.Results: Analysis of mixed control serum protein mass spectrum, CV12.20%<15%, reproducibility is good, chip different experimental error can be accepted, M/Z2000below the peak filter, avoid possible interference peaks.Application of SELDIsoftware in M/Z,2000-M/Z50000between216protein peaks weredetected.Comparison of patients with colorectal cancer and healthy subjects proteinpeaks, identified78protein peaks differences exist, in the serum of patients withcolorectal cancer in a sample of26protein peaks cut,52up-regulated.In78thedifference of protein peaks, screening out of12protein peaks: M/Z8132, M/Z,4002M/Z4070, M/Z,8351M/Z6341, M/Z,4472M/Z8446, M/Z,11092M/Z4875, M7571/Z,4435M/Z, M/Z5994(P <0.001), with statistical significance,we constructed a decision tree analysis model.With M/Z8132, M/Z4002has thebest results, for patients with colorectal cancer and healthy subjects have differentialsignificance.This diagnostic model for detecting colorectal cancer sensitivity of95%,specificity of94.87%, accuracy of94.95%.Conclusion: SELDI-TOF-MS profiling of serum proteins combined withbioinformatics tools can be applied to accurately differentiate patients with colorectalcancer from healthy subjects. The high sensitivity and specificity achieved bySLEDI-TOF-MS show great potential for the early detection of colorectal cancer.

Related Dissertations
More Dissertations