Abstract
588
Introduction: Carcinogenesis is a multistep and multi-focal process, characterized by stepwise accumulation of genetic and molecular abnormalities. A potential relation between telomere attrition in the early stages of carcinogenesis and the activation of the DNA damage response (DDR) machinery (considered as an anti-cancer barrier) has been proposed. Here we explored telomere length modification and its link with DDR pathway in the multi-step carcinogenic pathway of colorectal carcinoma. Method: Using immunohistochemical analysis, we retrospectively analyzed p-ATM (Ser 1981), p-CHK2 (Thr 68), γ-H2AX (ser139), TRF1, and TRF2 expression in normal, low, high grade dysplasia and invasive carcinoma in colon specimens. Using a Fluorescent In Situ Hybridization (FISH) method we analysed the associated telomere length at each stage of the multi-step progression. Statistical analysis of protein expression patterns and telomere size in relation with the different carcinogenic steps was evaluated by the Wilcoxon test. The correlations were assessed by non parametric Fisher analysis. Results: We demonstrated that telomere drastically shortens between normal and low grade dysplasia (p < 0.001) and between low and high grade dysplasia (p=0.012). In contrast telomeres increase in length in invasive carcinoma to reach similar or longer size as compared the one observed in normal tissue (p=0.007). In parallel we showed telomeric protein expression modification along carcinogenesis. Indeed TRF1 and TRF2 expression was diminished from normal to low grade dysplasia (p=0.004 and p < 0.001) and from low grade to high grade dysplasia (trend with p > 0.05). TRF1 and TRF2 are reexpressed at the invasive stage to levels close to the ones observed in normal tissue (p= 0.053 and p= 0.046). Moreover, TRF1 and TRF2 are statistically correlated between each other and with telomere length along carcinogenesis. ATM, CHK2 and H2AX are not phosphorylated in normal crypts. These proteins become activated between normal and low grade dysplasia (respectively p= 0.001, p= 0.002 and p= 0.03). We also tend to see an increase in this activation between low grade and high grade lesions. The DNA damage pathway is then deactivated between high grade dysplasia and invasive stage (respectively p= 0.03, p= 0.02 and p= 0.37). The activation of these proteins is inversely correlated to telomere length and telomeric proteins. Conclusion: To our knowledge, this is the first report analysing p-ATM (Ser 1981), p-CHK2 (Thr 68), γ-H2AX (ser139), TRF1, TRF2 and telomere length in consecutive slide sections of patient harbouring all histological steps of the colorectal carcinogenic process. Our analyses clearly reflect a correlation between telomere length and telomeric proteins and an inverse correlation with DDR pathway activation, reinforcing a potential telomere role in DDR pathway activation.
99th AACR Annual Meeting-- Apr 12-16, 2008; San Diego, CA