The science of hope: an interview with Randy Jirtle

You have had an extraordinary 50-year research journey, ranging from the physical to the biological sciences, which has been well documented on your geneimprint website [1]. Looking back at your career, what has been your favorite highlight?

My scientific career is different from that of most epigenetic scientists because I started in nuclear engineering. My undergraduate studies were in physics, mathematics and nuclear engineering. I transitioned from nuclear engineering to radiation biology when I entered graduate school. Then, in mid-career, my research changed again from radiation biology and cancer research to epigenetics. Thus, to pick one highlight is not easy since I have a large variety of options from which to choose! That said, my favorite studies were in epigenetics when we determined the time at which imprinted genes evolved [2].

Genomic imprinting is functionally defined as monoallelic expression in a parent-of-origin-dependent manner. This process is regulated epigenetically, and gene silencing is caused by DNA methylation and histone modifications that result in only one parental copy being expressed. Our studies showed that imprinting evolved about 150 million years ago in a common ancestor that gave rise to both marsupials and eutherian mammals. Interestingly, we also showed that the repertoire of imprinted genes varies among species. These findings support the intriguing postulate that once this novel form of gene regulation evolved in therian mammals, it was used to speciate.

In the 2009 Epigenomics editorial [3], I defined the human imprintome as the cis-acting regulatory elements that result in the genome-wide, monoallelic expression of imprinted genes in people. Moreover, we have recently defined 1488 regions in the human genome that are potentially involved in imprint gene regulation (submitted for publication). By comparing the human imprintome with those of different species, the bigger evolutionary biology question about what role genomic imprinting plays in mammalian speciation can now be systematically addressed. This would be a wonderful extension of our early evolutionary biology studies.

Your group's 2003 research paper titled ‘Transposable elements: targets for early nutritional effects on epigenetic gene regulation’, also known as the Agouti Mouse Study, started the age of epigenetic environmental research [4]. How well do you think our understanding of disease susceptibility has progressed?

Before we published this paper in 2003, the possibility that epigenetic changes were mechanistically involved in the developmental origins of health and disease was not considered. By contrast, since its publication, I don't think there is a paper that doesn't suggest or provide additional evidence that this is indeed the mechanism for developmental origins of health and disease. Thus, our understanding of disease susceptibility has improved dramatically since the publication of this paper. We now know that susceptibility to adult behavioral and metabolic disorders starts at or near the time of fertilization. Moreover, even though we are all on different epigenetic disease susceptibility trajectories that were established in the womb, ultimate disease formation depends, in part, on environmental exposures and how we live our lives.

I think our determination of the human imprintome and Rob Waterland's identification of human metastable epialleles [5] will further enhance the field of human environmental epigenomics research. It is possible that this new epigenetic regulatory information may even allow us to more accurately predict an individual's epigenetic biological clock [6] and possibly even reverse the epigenetic aging process. I am looking forward to continuing to do research in environmental epigenomics because I see a bright future for it.

The radiation hormesis hypothesis has been challenged throughout the last century. Since the publication of your paper titled ‘Adaptive radiation-induced epigenetic alterations mitigated by antioxidants’ in 2013 [7], what has been the impact of these findings?

I think of the DNA in a cell as being comparable to the hardware of a computer, whereas the epigenome is the software that instructs the genes when, where and how to work. I view a cell as functioning like a programmable computer. Thus, chemical and physical exposures affect not only the hardware but also the software components in a cell, thereby altering its overall biological response. The radiation epigenetics study you refer to demonstrated this in a surprising way.

The phenomenon of hormesis is characterized by beneficial effects at low-dose environmental exposures and toxic effects at high doses. Low-dose radiation is defined as being ≤10 cGy. To put this into perspective, the dose received from a CT scan is around 1 cGy, and that for a chest x-ray is about 100-fold less. Thus, patients receive significantly lower doses of radiation during diagnosis than while undergoing radiation therapy.

I began research at a time when biological risk of ionizing radiation was defined by the linear no-threshold model, which is still the case. This risk assessment model postulates that there is no dose of ionizing radiation that is safe. This model would be correct if radiation affected only the genome, but importantly, it also modifies the epigenome.

What Dr Bernal, when a graduate student in my lab, showed in her 2013 paper was that not only are low doses of ionizing radiation not detrimental but they also actually induce positive adaptive responses through alterations in the epigenome. Moreover, she showed that the addition of antioxidants can negate this radiation-induced positive hormetic effect.

In summary, low-dose radiation affects biological risk principally by modifying the epigenome, whereas high-dose effects are mediated primarily through damage to the genome. The resulting dose–response curve is J-shaped rather than linear, as would be predicted by the linear no-threshold model. Although our findings have not been embraced as readily as I had hoped, their importance will be better understood as scientists continue to study the role of epigenetics in mediating the biological responses to low doses of radiation.

At an epigenetics conference in 2018, you mentioned that young scientists will lead the way in this new era of research. What advice do you have for early career scientists in epigenetics?

At the beginning of my career, scientists were told that if they worked in the field of epigenetics, particularly in cancer research, they would have no career. This was because the dogma at the time was that cancer was solely a genetic disease. Although we now know this belief is incorrect, biases against this field of investigation persist. A career in epigenetic research also brings investigators into a variety of biological disciplines, potentially making it more difficult to become established as an expert in a single field of research. In some cases, this can also negatively impact one's ability to get funding.

Nevertheless, I would encourage any young scientist interested in getting into epigenetics research to go for it! This research is incredibly important; the field is massive in scope; and the probability of making significant, novel discoveries is phenomenal. Additionally, you can study epigenomics at any level, from the biochemistry involved in epigenetic programming to human epidemiology, which is what my colleague, Dr Hoyo, is currently doing at North Carolina State University. But to succeed, you are going to have to have a relatively thick skin and be persistent because there are still scientists who are not convinced that epigenetics is biologically important.

In your group's 2019 research paper, ‘Cadmium exposure and MEG3 methylation differences between Whites and African Americans in the NEST cohort’, you present the first evidence linking ethnic-specific hypermethylation and non-occupational cadmium exposure [8]. Should there be cause for concern over the carcinogenic effects of exposure to this environmental toxin, particularly in African–American communities?

Epidemiology studies can only demonstrate significant associations, not cause and effect; however, it appears from the evidence presented in this paper that African–Americans may develop diseases more frequently through changes in the epigenome than through genetic mutations. This indicates that environmental effects on the epigenome may potentially play a bigger role in the formation of diseases and disorders in the African–American population than in people of European ancestry.

This is an important issue to continue studying. If true, even though all populations develop diseases and behavioral disorders, the epigenetic and/or genetic paths that lead there may be different. Consequently, the methods needed to diagnose, prevent and treat these pathologies most effectively may also need to vary among human populations.

As briefly mentioned in your 2009 editorial, ‘Epigenome: the program for human health and disease’, the definition of epigenetics has changed time and time again [3]. How can we improve the clarity of epigenetic nomenclature?

Unfortunately, the epigenetic nomenclature is not easily controlled. As I previously said, we coined and precisely defined ‘imprintome’ as the regulatory elements that control the monoallelic expression of imprinted genes genome-wide. We subsequently expanded on this concept [9] and have recently experimentally identified the human imprintome (submitted for publication). Nevertheless, ‘imprintome’ is currently also being misused to define the repertoire of imprinted genes, not just their regulatory elements.

In 1998, every scientist in the world who attended the first international epigenetics and genomic imprinting meeting in the United States fit into one small conference room in Durham, NC [1]. Since then, the field of epigenetics has grown exponentially, and today thousands of scientists actively study epigenetics from multiple perspectives. This makes it even more important for scientists to use epigenetic terms in a rigorous manner!

A plethora of research has emerged using the agouti mouse model, as highlighted in your 2014 editorial, ‘The agouti mouse: a biosensor for environmental epigenomics studies investigating the developmental origins of health and disease’ [10]. Why is it so important to exercise caution when doing epigenetic research?

I discuss in this editorial how to use the agouti mouse model appropriately to do epigenetic studies. This is a powerful model system for studying the effects of early developmental exposures on epigenetic-induced disease susceptibility; however, it is sensitive to confounding factors as subtle as components in the food, which can alter the biological responses to environmental factors being tested.

I wrote this editorial because many people entering the field of epigenetics are not necessarily trained in the proper use of epigenetic techniques and model systems. Thus, I felt it was necessary to write about how to utilize the agouti model optimally so the results can be appropriately interpreted.