An epigenetic panic?

 Epigenetics is a hot topic. DNA now has six bases rather than four and there are modified versions of the proteins that wrap DNA into chromosomes. Should we be developing epigenotoxicity testing strategies?

It has been known for more than a century that exposure to some chemicals can cause cancers, and it is 60 years since the double helical structure of DNA was discovered. The chromosomes are now counted and the human genome is sequenced: every A, G, C and T has its place. There is the maturing science of genotoxicity with aspirations to mechanistically link chemical exposure to the breakage or mis-segregation of chromosomes, and the alteration of DNA sequences which can alter gene expression and risk the development of tumours. There are regulatory bodies that require batteries of tests to detect such agents.

But things have changed. First there were reports of heritable gene silencing and gene activation, without apparent change to DNA sequence. And then methyl-cytosine and subsequently hydroxymethyl-cytosine residues were found within genome sequences – they had appeared as cytosines using regular sequencing methods, but modified methods now allow them to be distinguished almost routinely. Over a similar period, the methods for studying the proteins that wrap DNA into chromatin and chromosomes – the histones - were becoming more sophisticated, and acetylation (and other variants) of these were described. It became clear that both the modified DNA bases and the modifications to histones are faithfully copied during chromosome replication, and that they can produce heritable effects on gene expression. This new information, ‘above’ the genome, has given us the new science of ‘epi’genetics, produced a new generation of anticancer drug targets, and might even allow the estimation of a person’s age from their DNA¹.

Changes in gene regulation are fundamental to the development of cancers, so should we be concerned that there might be epigenetic carcinogens that are not identified by the current batteries of regulatory genotoxicity testing? There are already recognised non-genotoxic origins of non-genotoxic carcinogenesis, such as regenerative hyperplasia, hormonal induction of cell proliferation, nuclear hormone receptor activation. However, whilst induction of cell division by any mechanism will increase chances of mutations arising simply through rare errors in DNA replication, it is clear that there are chemicals which can influence epigenetic status, so by implication there would seem to be a case to identify such properties in novel compounds for which human exposure is intended or might be anticipated. In order to develop tests it is important to understand what the mechanisms of epigenotoxicity might be and by extension, what an epimutagenicity assay might/should measure?

One might expect an epimutagen to affect the enzymes that conserve epigenetic status – for example the enzymes which conserve the methylation and hydroxymethylation status of C residues, or the acetylation of histones. The genes encoding these enzymes could of course be subject to genotoxic mutagenesis – and conventional genotoxicity assays already identify mutagens – even if the subsequent events were caused by changes to epigenetic status. However, the repair of DNA damage requires the unwrapping of chromatin to allow access for the enzymes of repair, as well as the removal of damaged DNA. DNA repair must also allow for the resetting of epigenetic status, so it is conceivable that there are epimutagens that interfere with the enzymes that restore epigenetic status – such as cytosine modification or histone modification etc. In this case an increase in genotoxic stress might lead to an increase in epigenetic change. In fact there might already be an early precedent for this. In 2002, Catherine Klein from New York University School of Medicine published a study using a reporter gene to study Chromium (VI) induced mutations. As well as the expected spectrum of DNA mutations, there were reporter mutants with unchanged DNA sequences. However in such cases, a subset of DNA cutting enzymes (“restriction endonucleases”) which are sensitive to DNA methylation status produced different patterns of DNA cleavage². In other words, it appears that Chromium (VI) can produce both mutagenic and epimutagenic changes. Are there ‘pure’ epimutagens? This is not yet clear.

There is one further aspect to the epigenetics story, recently discussed by Farooq Ahmed in Nature⁴ which emerged from three tragic periods in recent history: the Dutch ‘hunger winter’ famine caused by a blockade by the German army in 1944; the Great Chinese Famine (1958-1961) and the Biafra Famine (1968–1970). In all three cases, children conceived during the famine were unsurprisingly underweight at birth. Compared with siblings conceived before or after the famine many of these children had increased incidence of obesity and/or schizophrenia. Subsequent DNA sequence analysis of these survivors showed significant differences in the level of cytosine methylation in genes linked to these conditions, compared with the average. It is clear that our genomes are not as hard-wired as we might think.

Thus, we have a new element in genetics and a potential new hazard to investigate. Inevitably this has landed on the agendas of regulators, international workshops – and the general public. In 2009, Jay Goodman and colleagues, summarised an International Life Sciences Institute workshop, concluding that “…employing a model system for studying epigenetics is a long-term initiative that still requires a substantial amount of basic research, model system development, and validation. There is no short-term patch here. Fundamental work needs to be performed before we can even consider implementing epigenetic models into testing for regulatory decision making.”³

More recently John M. Greally wrote an OECD review considering the role of epigenetics in endocrine disruption⁵. Within that field it was probably a small molecule called Bisphenol A (BPA) which brought epigenetics into the wider public domain, particularly in the USA. It is widely used in the manufacture of polycarbonate plastic goods including water bottles, and as a consequence it can be found at very low (nanomolar) concentrations in humans as well as in the environment. It has been found to demonstrate estrogen mimetic properties, which raised developmental concerns, and in 2009 the majority of U.S. manufacturers of baby bottles and infant feeding cups ceased BPA use. It was subsequently banned from such use in the EU.

In order to understand epigenetics we need fast, inexpensive and reliable test methods to identify epigenetic affecters. Even though modified genome high-throughput sequencing methods are effective, they are still too expensive for routine screening. The author’s laboratory has just started work on a UK government-funded Biomedical Catalyst (www.innovateuk.org) study to discover whether it might be feasible to develop a simpler test for agents which affect DNA methylation status.

In the meantime, we should avoid epigenetic panic.  It is as well to reflect on the first and most important lesson in toxicology, given to us by the 15^th century scientist Paracelsus. It is most frequently paraphrased as “The Dose Makes the Poison”.  Its truth is evident from the observation that you need salt in your diet, but drinking sea water can kill you.

Author: Prof Richard Walmsley, Gentronix Ltd and University of Manchester

References

1: Fraga M F and Esteller M, 2007, Trends Genet. 23 413–418 2. Klein et al., 2002, Env. Health Perspectives 110 739-743 3. Ahmed, F., 2010, Nature,  468, S20 DOI:doi:10.1038/468S20a 4. Goodman et al., Toxicological Sciences 116(2), 375–381 (2010) 5. Greally http://www.oecd.org/chemicalsafety/testing/48435503.pdf