Identification Of Colorectal Cancer Protein Biomarker Based On ITRAQ

Background and AimsColorectal cancer(CRC) is the third most commonly diagnosed cancer in males andthe second in females, with over1.2million new reported cases and608,700deathcases estimated to have occurred in2008according to WHO report. Incidence rateswere unbalanced distributed worldwide.The highest incidence rates are found inAustralia and New Zealand, Europe, and North America,whereas the lowest rates arefound in Africa and South-Central Asia. Rates are substantially higher in males thanin females.In China the incidence and mortality of CRC is rapidly increasing in recent years,especially in big cities,mainly due to the raising of people’s living standards andchanging of people’s diets.Datas showed CRC is the third most commonly diagnosedcancer in the whole population of over1.3billion, the fourth in males and the third infemales, with over0.39million new cancer cases and0.19million deaths estimated tohave occurred in2009.Several researches showed modifiable risk factors for colorectal cancer includesmoking, physical inactivity, overweight and obesity, red and processed meatconsumption, and excessive alcohol consumption. CRC develops through a complexprogressions of genetic and epigenetic abnormalities that transform normal epitheliuminto dysplastic epithelium, which may ultimately progress into invasive and metastaticdisease. Extensive molecular characterization of colon tumors at the DNA and RNAlevel has shaped our current knowledge about the molecular mechanisms thatinvolved in tumorigenesis,including DNA sequence mutations, DNA copy numberchanges, promoter hypermethylation, altered miRNA expression and so on, whichultimately affecting the spectrum of proteins expressed in CRC.Despite of the advances in surgery-based comprehensive treatments for CRC, theoverall outcomes and survival terms of patients have not significantly improved or prolonged over the last few decades sinace most of the patients are diagnosed atadvanced stages.Therefore,diagnosis at early stage is considered as a crucial andpromising approach to reduce mortality and improve the prognosis.Reports haveshowed that the prognosis of CRC patients range dramaticly with five year survivalrates of more than90%for localized CRC (stage I) and only about10%for CRC thatmetastasized to distant organs (stage IV).Although considerable attentions and effortshave been spend on CRC, the reduced incidences of CRC and/or prolonged survivalterms of CRC patients were not gained during recent years, mostly due to the fact thatmost CRC patients were diagnosed at advanced stages. In fact,30–40%of patientshave regionally advanced or metastatic disease on presentation, which cannot becured by surgery alone. In addition, more than half of the patients develop recurrenceand die of the disease.It’s widely recommended to screen CRC in population as wellas interference at early stages,by which the survival terms of CRC patients could beprolonged.All these screening and early stages interference attempts are based onidentification of protein biomarkers with superior diagnosis effects.Although biomarkers can adopt many molecular forms, proteins are major players indisease, response and recovery. Therefore, many efforts have been made to identifyprotein biomarkers.ITRAQ is short for isobaric tags for relative and absolute quantitation,developed byApplied Biosystems Incorporation of USA. The iTRAQ is based on differentialisotopic labeling of proteins or peptides which are derived from two cell states (e.g.,healthy and tumor cells) with either light or heavy tags. This technique could label4components to the most in identification experiments simultaneously and had shownhuge advantages in throughput,sensitivity and quantitative precision to overcomesome drawbacks of traditional proteomic techniques and represented a powerfulalternative.In the present work, iTRAQ was combined with two-dimensional liquidchromatography (2D-LC) and MS (Qstar-ESI, Applied Biosystems) to identifypotential biomarkers for CRC by two steps of comparison of protein profiles.First,weseperately compared protein profiles of CRC tissues and paired normal mucosal aswell as those of pre-surgical serum and post-surgical serum.Then the proteinsdifferentially quantificated in both tissue and serum sections were sorted out forfurther verification. The disregulated proteins were verified further by western-blot and immunohistochemistry（IHC）. The present study not only evaluated thesensitivity and specificity of the new potential biomarkers but also analysis thecorrelation and regression of the patients’ clinical characteristic and the new potentialbiomarkers.Our work compared the disregulated proteins of CRC tissues and serumsto sort the protein biomarkers with accordant differential expression,this stragetycontributes to the reliablity of our results.MethodsTissue and serum specimens for iTRAQ were obtained from the same patient whounderwent colon cancer surgery in our hospital. CRC tissue and paired normalmucosal tissue(located at least5cm away from the tumor margin) were obtainedduring surgical resection. After excision, sample tissues were flash frozen at80°Cand stored until use.Serum samples were obtained1day before surgery and twoweeks after surgery. Peripheral venous blood samples(10ml) were taken byvenipuncture and allowed to clot.Then the serums were collected and stored at-80℃until use.The iTRAQ labeled samples were fractionated by a SCX column and RP-LC followedby ESI-MS/MS analysis.Protein identification and quantification for the iTRAQexperiment were performed with Mascot version2.3.01(Matrix Science Inc,USA).The Scaffold version3.0(Proteome Software Inc,Portland) in Mascot software wasused for peptide identification and isoform specific quantification.The sub-cellular location and function of the identified proteins were elucidated bythe UniProt knowledgebase (Swiss-Prot/TrEMBL,www.expasy.org) and the GeneOntology(GO) database(http://www.geneontology.org/).CRC tissue and paired normalmucosal tissue samples used in iTRAQ exprement and other3cases of CRC tissuesand paired normal mucosal tissues containing20μg of total proteins were tested bywestern blot.Two proteins from the biomarker panel analysis were selected for IHCverification using an independent and larger sample set(80cases).Immunopositivestaining was evaluated in five areas.Sections were scored as positive if epithelial cellsshowed immunopositivity in the cytoplasm, plasma membrane, and/or nucleus whenjudged independently by two scorers who were blinded to the clinical outcome.Aquantitative score was performed by estimating the percentage of immuno-positivestained cells:0, negative;1,<10%positive cells;2,11%to50%positive cells;3,≥51%positive cells. Meanwhile, the intensity of staining was scored by evaluating the average staining intensity of the positive cells (0, negative;1, weak;2, moderate; and3, strong). Finally, a total score(ranging from0to6) was obtained by adding thequantitative score and the intensity score for each of the160sections.Statisticalanalyses were carried out with SPSS19.0software (SPSS, Chicago, IL, USA).Wilcoxon signed-rank test was used to compare qualitative variables of IHC resultsbetween CRC tissues and paired normal mucosal tissues. Pearson χ2or correctionalχ2tests were used to analyze the correlation between the differential expression ofindicated proteins and patients’ clinicopathologic characteristics.Logistic regressionwas used to analysis risk factors that could effect expressions of potential proteinbiomarkers. Receiver operating characteristic (ROC) curves were used to determinethe diagnostic values of the markers.ResultsiTRAQ agents were used in the present study to label digested peptides for comparingprotein expressions between tissues of CRC and matched normal mucosal as well asthat between serums of pre-and post-surgical. Here, up-or down-regulation is statedas protein expression in CRC tissues relative to paired normal mucosa or inpre-surgical serums relative to post-surgical serums.A total of568proteins(391differentially expressed) were identified in tissues and198proteins(102differentiallyexpressed) in serums. Using the cut-off value of1.5-fold for up-regulation anddown-regulation, a total of187proteins(73up-regulated and114down-regulated) and44proteins(19up-regulated and25down-regulated) were identified as differentiallyexpressed in tissues and serums,seperately.13proteins differentially expressed in bothtissue and serum sections were sorted out for further analysis.UniProt knowledge base (Swiss-Prot/TrEMBL) and the Gene Ontology (GO)database annotated that10of the13differential proteins were located in more thanone cellular component or organelles. In brief,38.5%,7.7%,69.2%,69.2%,23.1%,53.8%and53.8%of the13differential proteins were located inthe organelle, organelle part,extracellular region,extracellular regionpart,macromolecular complex, cell and cell part,respectively. We analyzed the proteinfunction annotated by the UniProt knowledgebase (Swiss-Prot/TrEMBL) and the GOdatabase.The following biological processes and molecular functions were involvedmore frequently:binding (84.6%), biological regulation(76.9%), cellular process(69.2%), metabolic process(69.2%), and response to stimulus (84.6%). In total,84.6% differential proteins were involved in more than one molecular function andbiological process.To validate our proteomic results, we selected4proteins, with the accordantdifferential expression of tissues and serums,out of13differentially expressedproteins and directly assessed their levels of expression by westernblot.Prenylcysteine oxidase1and peroxiredoxin-4were validated to be up-regulatedin4CRC cases.Immunohistochemical verifications of prenylcysteine oxidase1andperoxiredoxin-4were tested in80cases of primary human CRC tissues and theirpaired normal mucosa.The scores of prenylcysteine oxidase1and peroxiredoxin-4between CRC tissues and normal mucosa show statistically significant differences.(Wilcoxon signed-ranks test, P<0.001for prenylcysteine oxidase1, P<0.001forperoxiredoxin-4). Compared with normal cells, prenylcysteine oxidase1andperoxiredoxin-4were verified to be up-regulated in85%and71.25%of the CRCcases, respectively. As the results of Pearson χ2or correctional χ2tests, theup-regulated expression of peroxiredoxin-4was found to be correlated with lymphnodes invasions and TNM stages.Interestingly, logistic correlation tests showed thatdifferential expression of both prenylcysteine oxidase1and peroxiredoxin-4betweenCRCs and normal mucosa samples were correlated with lymph nodes invasions withOR of3.386(95%Wald Confidence Limits,1.474-8.243) and2.858(95%WaldConfidence Limits,1.219-6.701),respectively.Prenylcysteine oxidase1expressionshowed cytoplasm expression,which was detected in95%of CRC samples and60%of paired normal mucosa samples.Peroxiredoxin-4showed cytoplasm andextracellular expression,which was detected in96.25%of the CRC samples and80%of matched normal mucosa tissues.Cut-off values for prenylcysteine oxidase1andperoxiredoxin-4were stastically analyzed and determined as score3and4,respectively, to distinguish CRC from paired normal mucosa by comparing the sumsof sensitivity and specificity of various diagnosis criterias; these criterias were set as I,II, III, IV, V, VI and VII (scores≥0,1,2,3,4,5and6,respectively). Relateddiagnostic values for CRC detection using prenylcysteine oxidase1andperoxiredoxin-4were shown in Table4. The areas under the curve (AUC) of thereceiver operator characteristic (ROC) curves of prenylcysteine oxidase1andperoxiredoxin-4were89.4and84.2, respectively. These results showed moderatediagnostic accuracies for prenylcysteine oxidase1and peroxiredoxin-4. ConclusionsWe performed the iTRAQ-based CRC proteomic research and identified a total of391differentially expressed proteins in tissues and102differentially expressedproteins in serums. Prenylcysteine oxidase1and peroxiredoxin-4were validated to beup-regulated in CRC compared with paired normal mucosa by western blot andimmunohistochemistry. Our results suggest that prenylcysteine oxidase1andperoxiredoxin-4are potential diagnostic or prognosis biomarkers for CRCs, and theymay play an important role in tumorigenesis and development of CRCs.