5 ms·

Is there a good, real-world example of how we think epigenetics works? I've seen a lot of mentions that "epigenetics _proves_ that intergenerational trauma is r

Is there a good, real-world example of how we think epigenetics works? I've seen a lot of mentions that "epigenetics _proves_ that intergenerational trauma is real!" -- and at least to my eyes this feels like a stretch of the science. For instance, why are epigenetic effects inherited? Why is it not the case that in environment, the parents experience one environment and so some genes are expressed. (ie, epigenetics happens based on that environment) and then the children are in the a different environment, and so different genes are expressed. Do I have a wholly-incorrect understanding of epigenetics? Is there a clear and simple example of an epigenetic mechanism?

ch4s3 2y ago
It seems pretty clear for example that being an alcoholic can cause DNA methylation in your gametes such that your offspring have a higher risk above and beyond regular genetic factors to become an alcoholic. However that is at a population level and there are complicating factors. Inherited trauma is totally unsupported but up or down regulated gene expression based on environmental factors before your conception are supported by some compelling evidence.
- jvanderbot 2y ago
  I'm trying on a mental model re: Epigenetics. How does this look: Your genes are a recipe book, not a blueprint. What recipe is chosen is based somewhat on environmental effects. When you reproduce, you shuffle together the parents book + the earmarks on the pages.
  - ch4s3 2y ago
    I'm not sure if that totally fits. There's a kind of signal amplification or de-amplification happening.
    - jvanderbot 2y ago
      Yeah I was using dogear as a way of signalling "prefer this page"
      - ch4s3 2y ago
        Yeah, I assumed that, but down regulating isn't well captured in that metaphor.
- margalabargala 2y ago
  > Inherited trauma is totally unsupported This is only true if you take a pretty narrow definition of the word "trauma". One example would be an ancestor experiencing famine, who then passes down epigenetic markers have marked effects on their offspring in terms of health, longevity, and behavior. Here's a Nature article on it: https://www.nature.com/articles/s41418-023-01159-4 https://www.nature.com/articles/s41418-023-01159-4 > "Collectively, these discoveries infer the existence of “a heritable memory of starvation/hunger”."
  - ch4s3 2y ago
    The Dutch Hunger cohort study is highly suspect. However, the idea that someone is literally starving to death just before or during gestation having some knock on genetic effects is not all that surprising. But again, the phrase "inherited trauma", as it's typically used is bunk. Also the phrase "a heritable memory" is kind of nonsense even with respect to epigenetic changes. Again what we're talking about with epigenetics is up or down regulation of your existing genes, not some new information being encoded.
    - margalabargala 2y ago
      I found the research about the Chinese Suihua cohort to be more compelling. I fully agree that someone starving to death right around the time of pregnancy could have some knock on effects; this is discussing changes in the following generation, the ones that are tens of years removed from the starvation event. > Also the phrase "a heritable memory" is kind of nonsense even with respect to epigenetic changes. Again what we're talking about with epigenetics is up or down regulation of your existing genes, not some new information being encoded. I wonder if you are missing the forest for the trees here somewhat. We're talking about the up/down regulation of specific genes based on specific methylation patterns in DNA that are present only because of some event that occurred. It's not memory in the sense of "I remember what I had for breakfast this morning", but it's memory in the sense of "these methylation patterns exist because their genesis was triggered by a specific event in the last few generations". That IS new information being encoded, just in a more fuzzy way than with neurons or computer memory. I would need more convincing to think the phrases you mention are nonsense/bunk. Could you perhaps articulate the definitions you're using for "trauma" and "memory"? I think we are using different definitions, as it seems clear what is meant by the phrases as used. EDIT: what issue do you take with the Dutch hunger study? I hadn't heard criticism about it before, and I just googled "dutch hunger cohort study criticism" and found nothing.
gmm1990 2y ago
just look up dna methylation
- everdrive 2y ago
  Thank you, I will do so.
biophysboy 2y ago
The basic idea is that genome modifications can be inherited with the genome. Methylation is one example; histone mods are another. If you want specific examples, I'd recommend focusing on a specific type of modification and searching "fill-in-the-blank inheritance", to avoid the more dramatic pop science.
directevolve 2y ago
I'm in a comp bio PhD studying epigenetics in development and disease. Epigenetics can be thought of as the set of biochemical mechanisms other than the linear sequence of DNA that store and respond to information to control cell behavior. This is how normal development from a single zygote to an adult organism is controlled, why cells vary so much despite having (mutations/mosaicism aside) the same DNA, how cells respond to changing biochemical conditions, and where cancer comes from. The classic example of epigenetics - of stable information transfer between cell generations not based on DNA sequence - is methylation of cytosines. Here's the simplified version. As you know, the DNA is based on ACTG. When you have a CG, the C can be methylated, which means an extra carbon is attached at a specific position on the molecule. That extra carbon gets 'copied' when the DNA is duplicated during mitosis and gets passed to the daughter cells. Yet it can be added or erased during the cell's lifetime. It typically works to block proteins responsible for transcribing genes into mRNA from binding to the gene, so the gene is silenced when it is methylated. This gives a way to transmit patterns of gene expression across generations of cells. This mark is extremely well-studied and understanding everything about DNA methylation is beyond this humble PhD student. But there are also extremely mechanisms for erasing and/or resetting certain aspects of cell state during mitosis or meiosis, including RNA, the proteome, and methylation state. We can't take it for granted that just because there's a mechanism by which biochemical information could be transmitted that it actually is (or that it isn't). A common source for the idea of intergenerational trauma being supported by epigenetic evidence is the Dutch hunger winter [1]. The idea is that an experience that parents had in utero seems to have had an effect on the physical development of their children, in a way that's specific to parental sex. I'm willing to say "sure, that effect could be real and there could be an epigenetic explanation." But AFAIK we don't know specifically what that hypothetical mechanism would have been. Note that the bod(y|ies) of the mother and child are interchanging material and engaging in biochemical crosstalk constantly during gestation, so there is abundant opportunity for environmental influences on the mother to have lifelong impacts on the child. Since the child's body will in turn engage in the same 'dialog' with the child's own children during fertilization and gestation, it's entirely possible for environmental influences on generation 1 to impact generation 3 as mediated by generation 2 via an in utero exposure. All that said, I have a lot of uncertainty about what our priors ought to be on how important such hypothetical mechanisms might be. I think that in the coming decades, we'll make huge progress in understanding development and disease within an individual's body through an epigenetic lens. We will probably also see some interesting controlled experiments on "epigenetic inheritance" using short-lived model organisms, such as fruit flies or worms. Possibly we will collect enough epigenetic information in large-scale biobanks over 20-40 years that we can get a more detailed understanding of possible epigenetic inheritance mechanisms in humans, but these assays are currently quite expensive, and moreover there's no single measure applicable to an individual person, since the whole point of epigenetics is that the state varies across cell and tissue type. [1] https://www.ohsu.edu/school-of-medicine/moore-institute/dutch-famine-birth-cohort https://www.ohsu.edu/school-of-medicine/moore-institute/dutc...

ch4s3 2y ago

It seems pretty clear for example that being an alcoholic can cause DNA methylation in your gametes such that your offspring have a higher risk above and beyond regular genetic factors to become an alcoholic. However that is at a population level and there are complicating factors. Inherited trauma is totally unsupported but up or down regulated gene expression based on environmental factors before your conception are supported by some compelling evidence.

jvanderbot 2y ago
I'm trying on a mental model re: Epigenetics. How does this look: Your genes are a recipe book, not a blueprint. What recipe is chosen is based somewhat on environmental effects. When you reproduce, you shuffle together the parents book + the earmarks on the pages.
- ch4s3 2y ago
  I'm not sure if that totally fits. There's a kind of signal amplification or de-amplification happening.
  - jvanderbot 2y ago
    Yeah I was using dogear as a way of signalling "prefer this page"
    - ch4s3 2y ago
      Yeah, I assumed that, but down regulating isn't well captured in that metaphor.
margalabargala 2y ago
> Inherited trauma is totally unsupported This is only true if you take a pretty narrow definition of the word "trauma". One example would be an ancestor experiencing famine, who then passes down epigenetic markers have marked effects on their offspring in terms of health, longevity, and behavior. Here's a Nature article on it: https://www.nature.com/articles/s41418-023-01159-4 https://www.nature.com/articles/s41418-023-01159-4 > "Collectively, these discoveries infer the existence of “a heritable memory of starvation/hunger”."
- ch4s3 2y ago
  The Dutch Hunger cohort study is highly suspect. However, the idea that someone is literally starving to death just before or during gestation having some knock on genetic effects is not all that surprising. But again, the phrase "inherited trauma", as it's typically used is bunk. Also the phrase "a heritable memory" is kind of nonsense even with respect to epigenetic changes. Again what we're talking about with epigenetics is up or down regulation of your existing genes, not some new information being encoded.
  - margalabargala 2y ago
    I found the research about the Chinese Suihua cohort to be more compelling. I fully agree that someone starving to death right around the time of pregnancy could have some knock on effects; this is discussing changes in the following generation, the ones that are tens of years removed from the starvation event. > Also the phrase "a heritable memory" is kind of nonsense even with respect to epigenetic changes. Again what we're talking about with epigenetics is up or down regulation of your existing genes, not some new information being encoded. I wonder if you are missing the forest for the trees here somewhat. We're talking about the up/down regulation of specific genes based on specific methylation patterns in DNA that are present only because of some event that occurred. It's not memory in the sense of "I remember what I had for breakfast this morning", but it's memory in the sense of "these methylation patterns exist because their genesis was triggered by a specific event in the last few generations". That IS new information being encoded, just in a more fuzzy way than with neurons or computer memory. I would need more convincing to think the phrases you mention are nonsense/bunk. Could you perhaps articulate the definitions you're using for "trauma" and "memory"? I think we are using different definitions, as it seems clear what is meant by the phrases as used. EDIT: what issue do you take with the Dutch hunger study? I hadn't heard criticism about it before, and I just googled "dutch hunger cohort study criticism" and found nothing.

gmm1990 2y ago

just look up dna methylation

everdrive 2y ago
Thank you, I will do so.

biophysboy 2y ago

The basic idea is that genome modifications can be inherited with the genome. Methylation is one example; histone mods are another. If you want specific examples, I'd recommend focusing on a specific type of modification and searching "fill-in-the-blank inheritance", to avoid the more dramatic pop science.

directevolve 2y ago

I'm in a comp bio PhD studying epigenetics in development and disease. Epigenetics can be thought of as the set of biochemical mechanisms other than the linear sequence of DNA that store and respond to information to control cell behavior. This is how normal development from a single zygote to an adult organism is controlled, why cells vary so much despite having (mutations/mosaicism aside) the same DNA, how cells respond to changing biochemical conditions, and where cancer comes from. The classic example of epigenetics - of stable information transfer between cell generations not based on DNA sequence - is methylation of cytosines. Here's the simplified version. As you know, the DNA is based on ACTG. When you have a CG, the C can be methylated, which means an extra carbon is attached at a specific position on the molecule. That extra carbon gets 'copied' when the DNA is duplicated during mitosis and gets passed to the daughter cells. Yet it can be added or erased during the cell's lifetime. It typically works to block proteins responsible for transcribing genes into mRNA from binding to the gene, so the gene is silenced when it is methylated. This gives a way to transmit patterns of gene expression across generations of cells. This mark is extremely well-studied and understanding everything about DNA methylation is beyond this humble PhD student. But there are also extremely mechanisms for erasing and/or resetting certain aspects of cell state during mitosis or meiosis, including RNA, the proteome, and methylation state. We can't take it for granted that just because there's a mechanism by which biochemical information could be transmitted that it actually is (or that it isn't). A common source for the idea of intergenerational trauma being supported by epigenetic evidence is the Dutch hunger winter [1]. The idea is that an experience that parents had in utero seems to have had an effect on the physical development of their children, in a way that's specific to parental sex. I'm willing to say "sure, that effect could be real and there could be an epigenetic explanation." But AFAIK we don't know specifically what that hypothetical mechanism would have been. Note that the bod(y|ies) of the mother and child are interchanging material and engaging in biochemical crosstalk constantly during gestation, so there is abundant opportunity for environmental influences on the mother to have lifelong impacts on the child. Since the child's body will in turn engage in the same 'dialog' with the child's own children during fertilization and gestation, it's entirely possible for environmental influences on generation 1 to impact generation 3 as mediated by generation 2 via an in utero exposure. All that said, I have a lot of uncertainty about what our priors ought to be on how important such hypothetical mechanisms might be. I think that in the coming decades, we'll make huge progress in understanding development and disease within an individual's body through an epigenetic lens. We will probably also see some interesting controlled experiments on "epigenetic inheritance" using short-lived model organisms, such as fruit flies or worms. Possibly we will collect enough epigenetic information in large-scale biobanks over 20-40 years that we can get a more detailed understanding of possible epigenetic inheritance mechanisms in humans, but these assays are currently quite expensive, and moreover there's no single measure applicable to an individual person, since the whole point of epigenetics is that the state varies across cell and tissue type. [1] https://www.ohsu.edu/school-of-medicine/moore-institute/dutch-famine-birth-cohort https://www.ohsu.edu/school-of-medicine/moore-institute/dutc...