The 4th Wellcome Trust Epigenomics of Common Diseases conference in Cambridge brought together leading scientists to discuss the latest topics in epigenomics. Obviously, we are in the middle of the epigenome-wide association studies (EWAS) era. But the results of these studies need to be interpreted carefully, taking into account many confounding factors to properly distinguish cause and effect. Therefore, there is a big push for “functional genome-wide association studies (GWAS)” that look at the genome and the transcriptome, both of which are essential to study the epigenome. Another big question that was discussed during the conference is whether environmentally induced epigenetic changes can be inherited transgenerational and, if so, what is the importance and consequence of transgenerational epigenetic inheritance?
We listened to many exciting talks during the conference, trying to extract an answer how to interpret genome-wide studies and transgenerational inheritance.
Epigenetic modifications: their function an role in cancer
Tony Kouzarides, Gurdon Institute, Cambridge, UK
Tony Kouzarides is interested in assessing the efficacy and mechanism of action of therapeutic compounds and in uncovering new epigenetic targets for therapeutic intervention. Post-translation modifications of histones regulate all DNA-templated processes, including replication, transcription and repair. These modifications function as platforms for specific effector proteins. Therefore the lab of Tony Kouzarides tries to identify small molecule inhibitors and new histone modifications to target to disrupt cancer pathways. Recently, his laboratory described a new modification pathway: methylation of glutamines, which regulates rDNA transcription. They also demonstrated citrullination as a modification involved in decondensation of chromatin during pluripotency.
Epimatch: harnessing epigenetics for hematopoietic stem cell transplantation
Dirk Paul, University College London, UK
Dirk Paul derived a donor-specific DNA methylation signature that may predict the incidence of severe graft-versus-host disease in matched recipients. For this study genome-wide DNA methylation patterns using Illumina Infinium HumanMethylation 450K BeadChips were performed in HLA-matched donor-recipient siblings pairs. He identified seven methylation variable positions between donors matched with recipients developing no/mild GvHD and those matched with severe GvHD. Is this putative donor-specific DNA methylation signature the first predictive biomarker for severe graft-versus-host-disease?
Interpreting the results of epigenome-wide association studies
John Greally, Albert Einstein College of Medicine, USA
EWAS of human disease are becoming increasingly common. A lot of positive associations have been reported, often with only modest DNA methylation changes, enriched at specific loci, genes and pathways. In EWAS it is usually taken that any differences observed between two groups of individuals (such as those with and without a particular disease) indicate changes in the epigenetic regulatory mechanisms in the cells sampled from those individuals and that the epigenetic association provides insight into a possible pathogenetic mechanism. John Greally highlighted that this study approach could be misleading for a number of reasons. Besides age-related changes in DNA methylation profiles, accounting for cellular heterogeneity is critical in EWAS. Moreover the influence of DNA sequence variability upon 5(h)mC variability is estimated to differ between 22% and 80%. Therefore, John Greally calls for the functional GWAS era since when an EWAS is performed in isolation, it is difficult to assess whether epigenetic changes are indeed occurring in the cells studies. To test this definitively, a more multifaceted project design is needed that includes genotyping and transcriptional studies. The qualitative characteristics of data from sequencing-based assays can also help with the interpretation of the results obtained. Superseding the EWAS approach with a functional GWAS design is the next step in more effective and insightful studies of the role of the genome in human phenotypes and disease.
Mendelian randomization: applications and limitations in epigenetic studies
Caroline Relton, Newcastle University & University of Bristol, UK
Caroline Relton introduced an analytic approach for distinguishing causal from non-causal associations: Mendelian randomization. This is an instrumental variable analysis that uses a genetic variable to determine how a phenotype influences another (intermediate) phenotype. The technique relies on the principle that genotypes being randomly determined at conception, are generally not susceptible to reverse causation bias and confounding and can therefore be used to assess the causal nature of exposure-disease associations.
Transgenerational Epigenomics Debate
Panel: George Davey-Smith, Brian Dias, Anne Ferguson-Smith, Bas Heijmans, Marcus Pembrey
This panel discussion centered on the idea that epigenetics is revealing that we might be more than just the sum of our genes. That what we eat, the air we breathe, or even the emotions we feel may influence not only our own genes but also those of our descendants? The environment can certainly influence gene expression and can lead to disease, and mammalian experiments provide evidence of male line transgenerational effects on health and development. Fundamental questions remain regarding how much transgenerational inheritance takes place and which impact it might have on organisms. The leading scientists recognize the importance of our current environment and the impact of genetic variation on epigenetic phenotype. Therefore it is very difficult to design the ideal experiment: ideally you have a very strong phenotype you can replicate in genetically identical organisms … In the end ethical implications raise; what will for example be the impact on parental responsibility and feelings of guilt? But after all on top of our epigenetic landscape, we all experience ‘cultural inheritance’ and learn from our ancestors. Knowledge, traditions, wisdom and lifestyle are passed from grandparents to their children and grandchildren.
Epigenomics and the environment – can pollutants reprogram our health?
Andrea Baccarelli, Harvard School of Public Health, Boston, MA, USA
Andrea Baccarelli presents evidence from human environmental studies indicating that epigenetic alterations may mediate, or at least help to predict, effects caused by toxic exposures. He provides evidence of particulate-matter induced DNA hypomethylation and its correlation to blood pressure. Exposure to fine concentrated ambient particles (CAPs) lowered Alu methylation and coarse CAPs exposure lowered TLR4 methylation. Both fine and coarse CAPs exposure increased systolic blood pressure. As adjacent CpG sites are often highly correlated with each other, Andrea Baccarelli also presented the Adjacent Site Clustering (A-clustering) algorithm to detect sets of neighboring CpG sites that are correlated with each other.
The epigenetic clock and biological age
Steve Horvath, University of California, Los Angeles, USA
Several recent studies have shown that age-dependent CpG signatures can be defined independently of sex, tissue type, disease state and array platform. Steve Horvath used a collection of publicly available DNA methylation data sets for the development of a multi-tissue predictor of age. Its astonishing accuracy allows DNA methylation age to be used as a biomarker for addressing a host of questions in developmental biology, cancer and aging research. Interventions used for creating induced pluripotent stem cells reset the epigenetic clock to zero, DNA methylation age correlates with cell passage number and finally it gives rise to a highly heritable measure of age acceleration. The epigenetic clock described here is likely to become a valuable addition to the telomere clock.
In the end, we still don’t know whether transgenerational inheritance exists. It’s clearly that our current environment is more important and that our genome plays a very significant role in the whole story. That’s why we are called to undertake functional (epi)genome-wide studies, taking into account the genome and the transcriptome. In our interpretations of these genome-wide studies, it is highly important to distinguish causal from non-causal associations.
** Thanks to Anne Rochtus at the University of Leuven, Belgium for providing this coverage.