Role of DNA methylation in imprint maintenance in differentiated human cells

Summary
The phenomenon of genomic imprinting lies at the heart of a significant number of childhood disease syndromes, including disorders such as Beckwith-Weidemann Syndrome (BWS), Prader-Willi Syndrome (PWS) and Transient neonatal diabetes mellitus (TNDM) type I. Each of these syndromes involves a gene which is imprinted, that is to say is normally only active from one of the two chromosomes in the cell, either the one inherited from the father via the sperm, or the one inherited from the mother via the egg, the genes thus being "imprinted" with a memory of which parent they came from. Patients with imprinted disease syndromes have increased risks for further problems such as diabetes (TNDM) or Wilms' tumour (BWS) in later life as well. The mechanism by which one copy of a gene becomes inactivated, while its identical partner in the same cell nucleus remains active, is still under active research. We know that chemical tags added to the DNA, called DNA methylation, are involved in marking the inactive copy of the imprinted gene, but it is not known whether this is sufficient in all cases to turn off the gene, or whether the proteins associated with the DNA, called histones, are more important. While such mechanistic questions can be addressed in experimental animals such as mice, human studies have either used cells from patients, where it is not possible to tell if the changes seen in the DNA methylation or the histones are causative, or have manipulated DNA methylation levels using drugs in cancer cell lines, which are not genetically stable and thus a poor model.
We have developed human cell lines which are non-cancerous and genetically stable, with normal imprinting, and then selectively reduced the levels of DNA methylation only. These cells appear to show loss of imprinting of the genes which cause BWS and PWS and most likely all other imprinted regions, but are unaffected at non-imprinted regions. They have altered growth characteristics and appearance, in line with theories that imprinted genes have important roles in regulating growth and nutrition. They thus represent a unique resource for exploring the importance of DNA methylation versus histone status in regulating imprinted genes and by cross-comparing to cells from BWS patients and others, it should be possible to establish which is more likely to be causative. The cells are a potentially valuable tool for determining the side-effects of pharmacological inhibitors of DNA methylation currently being used in the clinic. We can also use the cells to look for previously unrecognised imprinted genes, which may be linked to additional disorders. Additionally we can use the cells to carry out further mechanistic work, including attempting to identify the factors involved in establishing imprints, which would in turn be candidates for causative agents in these diseases.

Imprinted genes are exemplars of epigenetic control and are highly dependent on DNA methylation for their regulation. Abberant imprints are the cause of a number of human growth and neurodevelopmental syndromes and are a frequent occurrence in cancer. Despite this, functional assessment of the role of the maintenance methyltransferase in regulating imprints in humans has been lacking. By using stably integrated small hairpin RNA targeting the DNMT1 gene in hTERT-immortalised cells we appear to have generated the first normal human cell lines lacking imprinting at both paternally and maternally methylated ICR due to targeted loss of methylation. These provide a unique resource for examining the effects of loss of maintenance methylation on imprinted genes. We aim to characterise these cells more fully using genome-wide assays of DNA methylation and gene expression, a bank of targeted pyrosequencing assays designed specifically to interrogate the imprinted loci and the use of chromatin immunoprecipitation. This should greatly inform the clinical and research community as to the effects of loss of maintenance methylation in normal cells and help to establish if imprinted disorders are primarily caused by chromatin or DNA methylation changes. Additionally, we will use these cells to benchmark 5'-aza-2'-deoxcytidine (Aza) treatment with the effects of DNMT1 depletion in a normal cell line, of direct clinical relevance for the treatment of myelodysplastic syndromes and potentially for other disorders. Further hypomorphic cell lines will be created to determine if cells from different tissues share a common response to loss of maintenance methylation: we will also attempt to re-establish methylation at an imprinted locus in these and other imprint-free cells by rescue experiments. Finally, the experiments may identity novel imprinted regions in humans.

Academics in the areas of epigenetics and stem cell biology both UK and world-wide will benefit from this research by greater knowledge of the role of maintenance methylation in immortalised human cells and the consequences of removing it. Measuring the sensitivity of imprinted genes to disruption will be of benefit to clinicians carrying out IVF treatments and may help to inform policy makers with regards to IVF safety. Identification of new imprinted genes, or of new control elements or genes involved in imprinted disorders will be of benefit to the clinician and potentially to patients, who could benefit from potential new tests for loss of imprinting. Likewise, better understanding of the effects of demethylating drug treatments has clinical relevance to patients with myelodysplastic syndromes who could benefit from improved clinical management in the short term and potentially from better drugs with more specific targeting in the long term. Policy-makers at NICE may be able to use the knowledge gained to better inform clinical reccomendations and drug manufacturers benefit from improved drugs and better market share. The modified hTERT lines may be of commercial value. The general public may see improvements in clinical care and we aim through our proposed outreach plans to increase the public's understanding of imprinting.