Science Ama Series: I’m Nicholas Staropoli, director of the Epigenetics Literacy Project, here to discuss the controversial field of epigenetics that some say is revolutionizing human health and our understanding of how inheritance and evolution work. AMA!

Abstract

Hi Reddit! I’m Nicholas Staropoli a science journalist, director of the Epigenetics Literacy Project and a writer/editor for the Genetic Literacy Project—two initiatives of the 501c3 non-profit Science Literacy Project. Both sites cover the intersection of science , media and policy. I also write a weekly report--Epigenetics Around the Web--debunking claims about epigenetics and its impacts. The media, and even many scientists, just don’t understand how epigenetics works—and how little we know about its mysteries.

Scientists have known for decades that epigenetic modifications to DNA—such as methylation—play a vital role in activating or turning off gene expression. Over the last decade some studies have shown that epigenetic changes may be driven by environmental factors. The most notable study found that rats licked as pups (a sign of maternal nurturing) dealt with stress as adults better than those which were not. Epigenetic changes on a gene involved with stress tolerance appeared to underlie this difference. The study also reported that the changes were heritable. It inspired a lot of pseudoscience. Many in the media hypothesized these results broadly to humans, suggesting that nurturing could radically alter our genes—and those changes could be passed on to future generations.

Numerous studies since have found links between environmental triggers (e.g. severe trauma and diet) and changes in gene behavior, which led to alterations in overall health. But most studies have been performed on rodents, and the actual meaningfulness to humans remains speculative.

While the science is intriguing, the media, alternative health purveyors and some clinicians have overhyped what we actually know about epigenetics, behavior and disease. Frequently, the public is told that ‘epigenetics means you can control your genes’—a claim that is both untrue and dangerous to human health.

I’ll answer questions from 1-3 pm EST — Ask Me Anything!

Edit: Hi everyone! Thanks so much for your questions, the puppy is overdue for a walk so I need to step away. I hope you all learned something about epigenetics and health, but if there's one message to take away it's this: the modern understanding of epigenetics is new and controversial, outside of maybe cancer, there aren't too many definitive answers. So don't get ahead of the science! Be patient and don't over-hype!

And if you want to stay up on the latest epigenetics news—as well as background about the microbiome and endocrine disruptors—check out our site the Epigenetics Literacy Project https://epigeneticsliteracyproject.org/. Follow us on Twitter (@EpigeneLiteracy) and like our Facebook page.

As a Biology teacher, I used to talk about Lamarck as a counterpoint to Darwin - acquired inheritance as a wrong idea to better illustrate what natural selection is not. In recent years I've begun walking this back as we just begin scratching the surface of epigenetics. Can you weigh in on this? How should this work impact what students learn about Lamarck?

dbo340

First of all, I’d like to thank you for your work as a biology teacher (assuming high school if you’re talking about evolution and Darwin vs Lamarck). I think we can all agree scientific literacy (not just epigenetic literacy) is in a disaster state in the US. To me that’s largely because we aren’t getting science education (particularly evolution but other topics like genetics, health and physiology) right in schools. So, people grow up with a poor understanding of science and then are incapable of making educated decisions as adults—whether those decisions be voting practices or just how they generally take care of themselves. So kudos for all your hard work.

To your question, I would advise you to change nothing in your approach. No legitimate scientist believes that epigenetics disproves Darwinian natural selection or our current understanding of evolution. In fact, if you look at the Royal Society meeting on this topic that took place in November 2016 (New Trends in Evolutionary Biology) nobody was attempting to repeal our current view of evolution the so-called ‘Modern Synthesis.’ These scientists were trying to add the idea that sometimes environmentally induced acquired characteristics are inherited and therefore could affect evolution (i.e. transgenerational epigenetic inheritance).

It’s important to note that this isn’t Lamarckianism as Lamarck didn’t think the environment could directly affect evolution only “use or disuse” affected an organism. So in some ways, Lamarack and epigenetics are at odds with each other.

Either way, proponents of newer evolutionary ideas still have a long way to go before we add transgenerational epigenetic inheritance to the theory of evolution. First, some of the opinions given at the conference follow ‘wouldn’t it be nice’ or ‘wouldn’t it make sense’ lines of reason. For example, ‘wouldn’t it make sense if that because it takes a long time in humans for an advantageous trait to spread for there to be a faster mechanism of evolution too.’ But that’s not how evolution works because wouldn’t it make sense for us to fly because we eat birds and it’d be easier to catch them. This is a really great takedown of the idea, the aforementioned, conference and some coverage surrounding it: http://scienceblogs.com/pharyngula/2016/12/04/you-dont-get-to-revise-evolutionary-theory-until-you-understand-evolutionary-theory/

Second, transgenerational epigenetic inheritance still has a long road ahead of itself to be proven as scientific fact in humans/other animals. Epigenetic reprogramming eliminates or alters almost the entire human embryo’s epigenome. No one has shown how epigenetic modifications withstand that process. Plus these markings are so transient it seems unlikely a modification would stick around long enough to be passed on directly for hundreds of generations. All the studies that show epigenetic inheritance over generations deal with static environment (see unrealistic), a recent study similar to the rat licking one tried exposing these rodents to a more dynamic environment and those modifications associated with stress disappeared quickly.

TLDR Don’t change a thing and let the science become more settled before altering your course.


I went to a talk by Prof Nessa Carey, senior director at Pfizer on this. She said that the ethics surrounding clinical trials with epigenetic drugs is really complicated. It might get approved for studies in people with terminal illnesses, but she's doubtful it will be approved for chronic diseases like diabetes or chronic heart disease.

What are your thoughts on this? How do you see epigenetic medication being used?

emplah

I see epigenetic medicine being used primarily to treat cancers because we can confidently say epigenetic changes can flip that switch from healthy to a cancerous cell. Actually, the handful of epigenetic-based cancer treatments that are already on the market have done pretty well. From a recent piece in The Scientist "For example, Decitabine, for example, a DNA methyltransferase inhibitor approved by the FDA in 2006 for the treatment of myelodysplastic syndrome, a precursor to leukemia—has achieved complete remissions in nearly half of patients studied." http://www.the-scientist.com/?articles.view/articleNo/49062/title/Evaluating-Epigenome-Targeting-Cancer-Therapies/&utm_campaign=NEWSLETTER_TS_The-Scientist-Daily_2016&utm_source=hs_email&utm_medium=email&utm_content=49877180&_hsenc=p2ANqtz-_VgYm-meicQXJy_Xe00LKEoYw29J8sYaydkbcabNQ3oLxyR3o51eMIvHwhGQzfCdvVVWxDZg4kz8ynyVgtBbAvCOGbXQ&_hsmi=49877180

There will soon be a lot more on the market too. Numerous experts have said they see a big role for epigenetic therapies in the future to play complimentary roles to other treatments (like immunotherapies).

Outside of cancer, many researchers are trying to work on epigenetic-based therapies for probably every disease out there but from what I know the only non-cancer therapy that is far enough along to talk about is a drug for Alzheimer's disease that is in phase II clinical trial. It’s called ORY-2001.

I think as far as ethical concerns; I think it’s very similar to gene therapy. If the drug is packaged or designed in a way that it only targets one cell type (i.e. the affected cell) which isn’t germline cells, then any potential transgenerational effects would be inconsequential. That’s probably why she said no to diabetes because that’s a disease that affects every cell’s ability to process insulin. And to that extent, I’d probably agree with her because the drug would affect sperm or eggs and could cause transgenerational effects.

But I do think that’s different than the transgenerational epigenetic inheritance a lot of people believe is happening. What I’d call that scenario is an untoward side effect on reproductive cells.


Have studies been performed on epigenetics in plants, and were these traits also heritable?

When you said that the study on rats who had been licked as babies shown that the lower stress was a heritable trait, how many generations was that seen in?

What other epigenetics modifications can be made besides DNA methylation, and can DNA methylation be "undone" in the same organism by some "opposite" environmental factor at some point in he organisms life?

Pokedoob

you ask 4 questions so I numbered:

1) Yea, so I actually intentionally set up this AMA to focus just on human implications because the implications and the scientific understanding in plants are drastically different. For humans we are still working in a mess of data, weak methodology and misconceptions, but for plants there is a much clearer picture (we may someday soon sponsor another AMA just on plant--specifically crop—epigenetics). This piece in the scientist is really great on the topic--http://www.the-scientist.com/?articles.view/articleNo/48087/title/Plants--Epigenetic-Secrets/

But briefly, the idea that epigenetic changes acquired in one generation due to an environmental trigger such as drought can be passed on is widely accepted. To the point that a number of researchers are working to improve crop breeding via this method particularly because these changes don’t require a change to the gene sequence—if you follow the GMO debate that’s a big freaking deal.

A few highlights: There’s a lot of work on this topic with rice. Just this week researchers at University of Texas released a pilot study explaining how epigenetic modifications could improve cotton. There’s even a company Epicrop that’s working on this on a few crops (e.g. soybeans) based largely on the research of Sally Mackenzie from Nebraska. This is a story I wrote on epigenetically modified organisms: https://epigeneticsliteracyproject.org/epigenetically-modified-organisms-the-coming-epo-farming-and-food-revolution/

2) That study I think only claimed a few generations. Others have stated 10 generations and recently a roundworm study claimed 14. Nothing has been proven yet for humans and the fact that human embryos go through epigenetic reprogramming (removal and tweaking of the entire epigenome) makes me question the applicability to humans.

3) Methylation is just one although it’s the most commonly studied. There’s modifications to histones and chromatin too. Non-coding RNA is an epigenetic process. (cue shameless plug) Our ELP page’s Epigenetics University’ is a great resource to learn about these: https://epigeneticsliteracyproject.org/epigenetics-university/

4) And yes the are reversible. For example a recent study in rats found that a positive experience can remove any epigenetic modifications from a stressful one. And you can imagine this is a big deal for trying to translate this science to your health.


How are techniques to describe or "profile" someone's epigenetic state coming along? Do we even know what to look for?

couchmaster518

A great question and I think really one of the most promising areas where epigenetics probably has a translatable role to human health.

There’s this big debate surrounding whether epigenetic modifications on their own cause diseases or if there is some other genetic or environmental cause which causes the disease and the epigenetic marks are a consequence of the disease. But either way understanding how patterns of epigenetic modifications form on and around the genome during diseases could lead to some really effective diagnostic tests and even just monitoring general health.

The major thing to point to in this arena is Horvath’s clock (you can read a longer description of it here: https://epigeneticsliteracyproject.org/cellular-clock-may-help-improve-cancer-treatment-forensic-science/).

Horvath, a UCLA researcher, has created a DNA methylation based algorithm that looks at epigenetic profiles to determine a ‘cellular age’ and if that age is significantly older than your chronological age there’s an association with numerous health maladies: cancer, higher mortality risk, infection, HIV, Alzheimer’s. A life insurance company is actually using this to determine policy worth, which personally I think is wrong because we still don’t know a lot about these modifications to be making serious financial decisions based of them (read more here: https://epigeneticsliteracyproject.org/blog/epigenetics-around-web-spinach-turnips-likely-wont-protect-air-pollution/).

Outside of that, several liquid biopsies (for those unfamiliar with the term: http://www.roche.com/research_and_development/what_we_are_working_on/oncology/liquid-biopsy.htm) based on epigenetic profiles of tumors have shown some promise in accurately diagnosing and staging of tumors. Epigenetic profiling of tumors has also shown promise in determining how well a specific cancer will respond to a specific treatment.

But I’d also caution that we don’t know for sure what a ‘healthy’ epigenetic profile looks like if it even exists (very similar to the microbiome in that respect) so it’s hard to gauge how beneficial this information will be in general.


I have more of a general question: What role does epigenetics play in diseases that were previously thought to be genetic in nature?

EffectorTCell

For cancer, epigenetics it turns out plays a big role and in some cases epigenetic dysregulation can lead to tumor initiation and growth. This video is an awesome explainer of it: https://www.youtube.com/watch?v=UUM7HiFkDd4

Outside of cancer, I’d say that’s still a matter of debate. A number of diseases are linked to epigenetic modifications (e.g. Alzheimer’s disease) but there’s still a major question as to cause or consequence. In other words, we still haven’t sussed out if these epigenetic changes are actually driving diseases or if other aspects (environmental toxin, genetics) are causing the disease and the epigenetic modifications.

Even if they aren't causing pathologies, these epigenetic modifications could play a big role in diagnosing and monitoring diseases.


Sorry, not my area of knowledge so I'm not sure I'm understanding this concept. I work in remote Australian communities. Is your example saying that they may be genetic changes following on from years of trauma, hence high burden of disease in this population? Is this good news in that advances in genetic manipulation could be a "quick fix" or is it terrible news in that we have caused core damage for future generations?

Canetoad03

Not genetic changes, but epigenetic. so changes to the gene's activity without a change in its sequence. This is a great explainer of the study: http://learn.genetics.utah.edu/content/epigenetics/rats/

I'd say right now for humans it means nothing. There is still a long road ahead to prove that trauma in one generation causes health effects in subsequent ones via epigenetic modifications. For one, no one has yet shown how these modifications get through epigenetic reprogramming of an embryo.


Have there been any experiments involving humans yet? These would all have to be pretty lengthy right? So I cant imagine we'll have a good idea of how epigenetics effects humans for quite some time. What kinds of studies would you like to do, or what kinds of studies are currently in progress?

Sorry for the scattered thoughts, I'm at the tail end of a 12 hour third shift.

Evan_dood

Yea there have been some poor quality (not to be critical more of a statement of the difficulty in overcoming weak instrumentation/methods and strong ethical barriers) that have studied epigenetic changes in humans. This one (https://epigeneticsliteracyproject.org/media-again-mangles-epigenetics-shutting-off-love-hormone-unlikely-to-make-us-less-social/) I wrote about last year look at epigenetic changes on the oxytocin gene in people and linked it to sociability. But a major problem was they looked at the gene in epithelial cells from saliva and not neurons where the gene's activity matters.

Another poor study human study subjected people to air pollution and saw epigenetic changes, but found that those who ate more vitamin b were “more protected” from the pollution. But this study was pretty poor too largely because they only examined 10 people (https://epigeneticsliteracyproject.org/blog/epigenetics-around-web-spinach-turnips-likely-wont-protect-air-pollution/).

As with so many other fields, studying on humans is hard. But in particular it’s very hard with epigenetics because it’s very difficult to conclusively say an effect is epigenetic and not genetic because the processes are so tightly intertwined. Many epigenetic changes are controlled by genes which complicates the matter too. So does the fact we really don’t fully understand just how complicated genetics is.

“What kinds of studies would you like to do”—this is my favorite question I’ve ever been asked. I had to think for a few minutes but here’s my answer:

We need to figure out whether or not epigenetic modifications can get passed epigenetic reprogramming of the embryo and if they can how and under what circumstances it happens.

All the ideas and opinions that epigenetic modifications affect subsequent generations and play a role in evolution are based entirely on the (currently) unsubstantiated belief that these changes magically get survive a process that we know removes or alters almost all of them. If someone can prove how that happens it would be huge. If someone could definitively prove it doesn’t or can’t happen in humans (and I completely understand how hard it is to ‘prove a negative’) that would stop a lot of debates about the field’s significance.


The epitranscriptome--how much do you know about it? Is it altered by the same stresses as what alters the epigenome? Is the epitranscriptome heritable as well?

xSialicAcidx

The epigenome (at least in the modern understanding) is so new and we still aren't entirely sure how and if the environment affects it or if these modifications are heritable. So to now throw in an even newer field like epitranscriptome and ask those questions I think is jumping too far ahead. I don't think we'll have definitive answers for a long time.


Do epigenetic modifications carry over in generations?

To use your example, is there evidence to suggest that rats that were licked as pups also produce offspring with better stress tolerance?

Kinbaku_enthusiast

Yea so in that example they did see the same epigenetic changes in the descendants which implies they were inherited. As a resource this is a great interactive about the study: http://learn.genetics.utah.edu/content/epigenetics/rats/

Other studies on model organisms have found the modifications carrying on for as many as 14 generations according to one recent study in roundworms.

I think it's important to realize a few things though when trying to translate this very famous study to humans.

First of all only about 30 percent of rodent studies translate to humans. So just because we see this in rodents doesn't automatically mean it has human implications--in fact it's more likely it doesn't.

Second, behaviors like stress are very different in humans than in rats, roundworms and mice. For these animals, it's very instinct based. But we are conscious beings we can overcome stress and make a decision or an active change, something other lower animals can't


Have you found any epigenetic changes in soldiers with PTSD?

theman1119

As with any disease, there may be epigenetic modifications associated with a disease. Other than for cancer, the jury is still out on whether they are cause or consequence. PTSD is no different.


Are there any companies doing anything with epigenetics?

818f150

Yes. The epigenetics-product market is booming.

There are a handful of epigenetic drugs already on the market to treat cancer and there are several in various stages of clinical trials. There's another in phase II of a clinical trial (I believe) for alzheimer's disease which I think is the only one not for cancer.

There's also several detection kits and instruments for research purposes. And in the near future many diseases may have diagnostic kits that are based on epigenetic modifications.

Based on one market report I saw, the market for epigenetic products is supposed to grow 5 times from 2011-2020.


You'll probably not even see this question, but heregoes: I read that stressful environments can cause epigenetic changes, would you expect our genome to reflect the stress of chronic warfare? My grandfather had extensive PTSD. Would his offspring have higher incidence of anxiety and depression?

skipperdog

I see it!

Certainly, what you're outlining is being discussed and researched by scientists. I would say we aren't anywhere close to a consensus answer that PTSD from warfare causes epigenetic changes that are not only inherited but also cause poor health consequences in their descendants. We may not have an actual answer for years or decades...and anyone who tells you that inherited trauma is definitely real is probably trying to sell you something. https://epigeneticsliteracyproject.org/blog/epigenetics-around-web-avoid-craigslist-epigenetics-advice-engineering-superhumans/


Are we gonna see the concept of Designer Babies in action? How long would it take?

I-WantKarma

In terms of designer babies, I don’t think that’s realistic for epigenetics. In other words, you can start methylating (or demethylating) an intelligence-linked genes in an embryo and expect the baby to be brilliant. Epigenetic modifications are transient and in a constant state of flux during early development so I think there’s zero chance an epigenetic modification like that would last.

That be said, some people do believe that someday we’ll have a pill that will target an intelligence-linked gene and increase its activity via epigenetic modifications. Proponents believe this is a nice compromise that allows human enhancement without full on gene editing of humans (where the germline could be affected). But I’d caution this likely won’t happen. The complex traits we’d be targeting are so multifactorial (environment, genes, gene-gene interactions) that we won’t ever be able to make a difference just by increasing a few genes. Again the often transient nature of these markings also means they may not last long enough to have the desired effect.


Additional Assets

License

This article and its reviews are distributed under the terms of the Creative Commons Attribution 4.0 International License, which permits unrestricted use, distribution, and redistribution in any medium, provided that the original author and source are credited.