Validation of novel colorectal cancer biomarkers derived from animal models of Apc inactivation: analysis of cohorts from the UK and Brazil.

摘要

Colorectal cancer (CRC) is the third and second most common cause of cancer death in men and women, respectively, worldwide. Most deaths result from late diagnosis and the lack of effective treatments for patients with advanced disease. Better biomarkers for early diagnosis, prediction of response to treatment and prognostic determination are therefore urgently needed. In this research, we have assessed several CRC candidate biomarkers that had previously been identified during studies involving animal models of adenomatous polyposis coli (Apc) inactivation (the most common genetic alteration in colorectal carcinogenesis). Our hypothesis was that these candidate proteins would translate into valid biomarkers of human colorectal neoplasia. Therefore, we tested the expression of these candidate biomarkers in tissue and blood samples obtained from patients with colorectal neoplasia as well as healthy controls. Patient cohorts from the UK and Brazil were analysed in this research. Using electronic scoring tools, we assessed the immunohistochemical expression of the candidate proteins in normal colonic mucosa, adjacent non-neoplastic colonic mucosa, colonic adenomas and colorectal cancer samples. Clear differential patterns of expression were observed for nucleosome assembly protein 1 – like 1 (NAP1L1), ribosomal protein L6 (RPL6) and prohibitin (PHB) when comparing cancers and non-malignant tissues. Additionally, NAP1L1 and RPL6 exhibited different expression patterns in low-grade versus high-grade adenomas, thus suggesting that they may play roles in the transition from low-risk to high-risk premalignant lesions. Gene expression studies showed that NAP1L1 and RPL6 were highly expressed in the tumour and the adjacent mucosa from patients with CRC when compared to colonic biopsies obtained from normal control subjects. These results support a role for these genes not only in colorectal carcinogenesis but also in colonic “field cancerisation”. RPL6 silencing resulted in strong inhibition of proliferation in HCT116 colorectal cancer cells. PCR-array studies demonstrated that RPL6 silencing caused up-regulation of BCL2 associated X (BAX) and mutS homolog 2 (MSH2) - protectors against cancer development, and down-regulation of matrix metalloproteases 12 and 13 (MMP-12 and MMP-13) - promoters of cancer progression, supporting the importance of RPL6 in colorectal carcinogenesis. The blood concentrations of NAP1L1 (assessed using a novel in-house electrochemiluminescence immunoassay), RPL6 and PHB (measured using commercial enzyme-linked immunosorbent assay kits) did not show any significant differences in cancer individuals when compared with normal controls and adenoma-bearing individuals. However, several new findings related to the measurement of the concentrations of these proteins in blood-derived fluids were made. A study of a retrospective cohort of CRC patients clearly demonstrated that the immunohistochemical expression of NAP1L1 was related to prognosis. High nuclear expression of NAP1L1 was independently associated with a marked increase in overall survival and 5-year survival estimates. Mortality in this group was 61 to 72% lower when compared with the low-expression group. This difference was however only observed in patients who had late stage disease (stages III and IV). The original contribution of this thesis is the confirmation that the candidate biomarkers derived from animal models of Apc inactivation are also differentially expressed in human CRC samples. The results produced by the various methodologies described suggest that NAP1L1, RPL6 and PHB may be potential novel biomarkers for the early diagnosis of CRC and the identification of high-risk premalignant lesions. Additionally, the association of NAP1L1 expression with the prognosis of CRC patients has not been previously reported and may have a clinical application. Further prospective research assessing larger sample cohorts is now highly recommended in order to confirm these findings.