N12001 and N12005: Detection of CpG island methylation in faecal DNA

Study Duration: March 2001 to March 2002

Contractor: University of Newcastle

Colorectal carcinoma (CRC) is the second most common cause of death from cancer in the UK. In contrast, in some developing countries of Asia and Africa, the age-standardised incidence of CRC is less than 10% of that in Europe. Overall, approximately 80% of colorectal cancer in the UK is thought to be due to dietary factors that remain largely unidentified, and which exert their effects over several decades. To facilitate research on the dietary prevention of CRC in the UK, a scoping study for the FSA Colonic Health Programme (N12) called for research to identify biological changes marking the earliest stages of the disease process that leads eventually to colorectal cancer, and to design novel ways to measure these changes and study their modification by diet.

Rationale and Objectives
Much is known about the changes that occur in cells during the onset of CRC. The most important mechanisms involve damage to genes that control the rate at which cells divide, their normal functions, or their ability to undergo self-destruction. Such damage may involve mutations, which destroy the signal sent by the gene, or gene-silencing, so that its signal is no longer sent. Certain genes are tumour suppressors, which normally prevent the development of cancer. Silencing of such genes therefore removes a brake on cancer development. The main mechanism of gene-silencing is a chemical change to the backbone of the DNA molecule called methylation. The location at which such methylation occurs is called a “CpG island” because it is a region of DNA in which adjacent pairs of the bases cytosine and guanine occur with unusual frequency.

So-called aberrant methylation of CpG islands has recently been recognised as a key feature of many kinds of cancer cells. Genes are methylated in the fully developed tumour, but the process of aberrant methylation begins in the normal tissue years before the appearance of the disease.

Methylation can be measured using minute quantities of DNA. The cells that line the inside of the intestine are constantly shed into the bowel and voided with the faeces. Any DNA that survives this process will carry the same methylation pattern as the cells from which it came, and this will be the same as the cells that are continuing to divide in the intestinal lining. In this short (12-month) study we tested the feasibility of detecting aberrant DNA methylation using faecal DNA.

After obtaining ethical approval, volunteers were recruited from patients attending the gastroenterology outpatient and surgical lists of a large general hospital. Samples of the cells lining the intestine were obtained from consenting patients during routine procedures for the examination of the bowel using a flexible endoscope, or at surgery for cancer. Patients later donated samples of stool passed under normal conditions at home.

All the samples were examined in the laboratory and compared with the diagnosis of the patient from whom it was obtained. The patients remained anonymous to the scientists conducting the analysis, and the diagnostic information was not available to the scientists until the analysis was complete.

The researchers used highly sensitive molecular biological techniques called Methylation-specific PCR (MSP) and Combined bisulphite restriction analysis (COBRA) to detect the presence of methylation, and measure its extent in the tissue samples, and they adapted them to work in the highly contaminated DNA samples obtained from human faeces. They then validated the methods and applied them to a selection of genes known to be involved in colorectal cancer (ESR1, MLH1, APC, MGMT, p16 and HPP1).

We have shown for the first time that DNA methylation detectable in the cells lining the intestine is also measurable in very small faecal samples. The COBRA technique allows us also to quantify the results, so that comparisons can be made between test-groups in the population, and the results can be analysed statistically. The level of CpG methylation we have observed in the mucosal tissue is broadly consistent with other recent publications, but the application of these techniques to faecal samples has never been described in the scientific literature before. From now on we believe it will become very much easier to study the mechanism of CpG methylation because it will no longer be necessary to obtain samples of tissue in a hospital clinic. We have also shown that the level of methylation differs considerably between one gene and another. Since the techniques we are using are fast, and reasonably cheap, and the quantities of DNA required are small, it will be possible to study many genes, including entirely new ones, simultaneously, and compare the patterns of methylation in groups eating different diets or exposed to a range of other environmental factors.

These findings provide three new opportunities, all of which need to be tested in a much large number of volunteers. First, since genes become methylated in healthy people with no detectable cancer, it should become possible to characterise individuals and populations in such a way that we can determine their future susceptibility to CRC a long time in advance. Secondly we will be able to design entirely new experiments to enable us to understand how gene methylation comes about, and how it relates to diet. Thirdly, these approaches will eventually enable us to quantify the effects of food components such as folate, other B vitamins and alcohol, on gene methylation in the colonic mucosa, and to determine which dietary manoeuvres may be effective in reversing aberrant gene methylation.

Contact: Peter Sanderson
Tel: 020 7276 8920
Email: peter.sanderson@foodstandards.gsi.gov.uk