Desmetilació genòmica com a biomarcador de predisposició a càncers múltiples

Abstract

Colorectal cancer (CRC) is one of the most frequent cancers in human populations, particularaly in the western countries. Both genetic and epigenetic alterations are germain to CRC develpoment and progression. In a previous collaborative work conducted in Japanese population, it was reported that global genomic methylation of the normal colonic mucosa (NCM) of CRC patients was predictive of of the risk to develop multiple colonic cancers, both synchrounous and metachronous. In that work, global methylation was measured on the LINE--‐1 repetitive elements, which constitute about 17%--‐20% of the human genome and are frequently employed as surrogate marker of global methylation levels. This sequences are mayoritarily methylated in virtually all somatic adult healthy tissues, including the colonic epithelial cells. However, cancer cells have been known for a long time to exhibit global genomic hypomethylaton which affects also LINE--‐1 elements. LINE--‐1 methylation was analyzed by MethyLight, a QPCR--‐ based method capable of quantify the relative proportion of hypomethylated molecules. While the hypomethylation in cancer cells was known, the strong association of genomic demethylation in NCM with multiple cancer risk was novel, and suggested the existence of an epigenetic field for cancerization, i.e. a region of normal colonic mucosa that harbours epigenetic alterations preceding and promoting tumour development. That analysis, however, could not test the extension of that epigentic field because only one normal sample was analyzed in every patient and, in most cases, the distance between the normal biopsy to the tumor location was unreported. Moreover, both the normal and the tumor samples consist of different cell types (cancer cells, epithelial cells, muscle layer) with different epigenomic profiles. This work was aimed to explore the extent of DNA hypomethylation using FFPE samples from multiple--‐CRC patients from the Hospital Germans Trias i Pujol in Catalonia, and frozen samples from single--‐CRC from the CHTN in USA. FFPE samples were microdissected to facilitate ba more precise determination of their cellular content. In addition, in this work two normal samples taken from differnet locations of the colon, plus non--‐cancer cells directly adjacent to the tumor samples have been analyzed. We employed a real--‐time methylation--‐specific PCR (MS--‐QPCR) technique based on that used in the previously published work. The technique has been slighty modified to reduce the cost of the analysis, by substituting the specific fluorescent probes by SYBR--‐Green for the quantification of the amplified products, and including a melting--‐curve analysis to determine the differences between amplicons. Our results demonstrate the suitability of the expermental approach to analyze global genomic hypomethylation in FFPE samples. Tumor cells exhibited significant lower levels of methylation than their non--‐tumoral matching samples, in agreement with previous observations. We also observed a high degree of coordination of the demethylation level in normal samples from the same individual, suggesting that the demethylation is homogeneous along the colon. This coordination might facilitate the estimation of the global genomic demethylation level in the whole colonic tract from a single biopsy