Biomarker Development In Colorectal Cancer Based On Dynamic Monitoring Of Tissue Interstitial Fluids From AOM-DSS Mouse Model

The development of serum protein biomarker is important for the prevention, diagnosis, and prognosis of colorectal cancer (CRC). This study attempts to build a new pipeline for developing CRC serum protein biomarker which can solve the problems (can not discover the biomarkers which indicate the early initiation and progression of tumor and the verification of potential biomarkers are inefficient) exist in the traditional methods. We discovered CRC biomarker candidates through quantitative evaluation of the dynamic abundance changes of the proteins in colonic tissue interstitial fluids (TIFs) during the progression of CRC. We hypothesized that there would be great chance to screen CRC serum protein biomarkers out of those TIF proteins which are constantly increased during CRC progression, and then a series of experiments in mouse and human serum samples were conducted to test our hypothesis and evaluate the prospects of the CRC biomarker candidates.We successfully constructed AOM-DSS CRC mouse model and established a method using narrow pH range (pH 3-6) IPG strip for effective TIF peptides separation. Through analysis of the dynamic proteomes of the colonic TIFs from different CRC development periods (Cycle Ⅰ, Cycle Ⅱ, and Cycle Ⅲ), we quantified 584 TIF proteins, of which～66% are theoretical secretory proteins or known serum proteins. We screened 144 differential proteins out of these TIF proteins, and divided them into three groups based on their changing tendencies:constantly increased (n=45), constantly decreased (n=17), and irregular changed (n=82). Then, we quantified 24 constantly changed proteins in 16 mouse TIF samples using multiple reaction monitoring (MRM) technique, and verified 18 proteins (12 constantly increased,6 constantly decreased) showing same trends of change as that quantified by iTRAQ.To test whether the CRC-related TIF proteins have serum biomarker potential, we quantitatively evaluated the abundances of 12 constantly increased TIF proteins in corresponding mouse serum samples using MRM, which showed that comparing with Control group, the abundance of LRG1, TUBB5, and IGJ were significantly up-regulated. Then, to preliminary assess their compatibility of applying in human, we evaluated the abundances of these three proteins in serum samples from 16 CRC patients and 16 healthy people using the same approach. This assay demonstrated that in CRC group, LRG1 and TUBB5 showed a significant increase of abundance, while no significant difference in IGJ abundance was detected. Receiver operating characteristic curve analysis indicated that, the area under the curve of LRG1 and TUBB5 were 0.74 and 0.70, respectively, and the combination of these two indicators improved the detection specificity.To assess the clinical application potentials of biomarker candidates, we measured the absolute concentrations of LRG1 by sandwich ELISA in expanded clinical serum samples, and found that it was significantly up-regulated in CRC group comparing with healthy control group. Furthermore, serum LRG1 could distinguish between healthy controls and CRCs with sensitivity equal to 59.3% and specificity equal to 79.6%, under cut-off value of 34.56μg/mL. These data demonstrated that to a certain extent, LRG1 has an application prospect as CRC serum biomarkers. Then, we compared the expression abundances of LRG1 between colon carcinoma and adjacent epithelia by immunohistochemistry (IHC) assay on tissue microarray of clinical samples. The IHC results demonstrated a significantly up-regulated LRG1 abundance in cancerous tissues compared with paired normal adjacent epithelia, and the abundances were significantly distinct between two primary tumor status (T3 vs. T1), metastasis status (M1 vs. MO), and three different AJCC stages (Stage Ⅳ vs. Stage Ⅰ, Stage Ⅳ vs. Stage Ⅱ).In conclusion, our study demonstrated that it is a novel and effective pipeline for the development of CRC serum protein biomarkers by coupling dynamic and quantitative proteomic analysis toward colonic TIFs from mouse model with targeted verification of biomarker candidates in clinical serum samples. By applying this pipeline, we successfully discovered protein LRG1 and then verified the change of LRG1 abundance within sera of CRC patients and across colon cancer tissues, which makes LRG1 a potential novel serum biomarker for CRC diagnosis and prognosis.