Lefebvre on epigenetics
Nov. 10th, 2008 01:00 pm![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
guslThis Saturday I skipped Grisman to see Louis Lefebvre talk about epigenetics. Being a total layman, I found it very accessible and enjoyable, even if I was somewhat puzzled about the basic facts.
Rough summary / questions:
* epigenetics can mean many different things. Etymologically, "epi-" is like "beyond".
* one kind is CG-methylation (aka "Ç" if you think of methyl as cedille) which renders genes inactive.
It is a kind of statefulness that can get passed on during transcription (due to methyl's tendency to cross the DNA), but apparently not meiosis(?), related to inactivation of genes (IIRC, also X-inactivation in females)
* in the zygote, the methylation state gets reset (methyl goes away?); human cloning is a bad idea because this resetting doesn't happen, leading to severe malformations. Of course, this raises the question of how methylation state could possibly be inheritable.
* diseased genes can be passed for many generations, e.g. down a paternal line, and only manifest themselves once they pass this gene on to a daughter, who passes it on to a child, who shows the diseased phenotype (or vice-versa: invisible down a maternal line until they have a son who has a child). I wonder what kind of mechanism could explain this.
* I found it a bit surprising that the zygote is formed before the two gametes' DNA get together (duh!), meaning there's no attempt to check how well this individual sperm matches this individual egg. This lack of selection implies that the zygotes formed are "really" "random". Though I wonder if sperm DNA is related to sperm phenotype, leading to a selection there.
* You can observe methylation state as a green light, using rats bred with genes for bioluminescence! There were two types of females (presumably, due to them having two Xs), one showing no light, one showing a dim green light. All males showed a bright green light. I don't understand this.
* The presenter himself was colorblind. So was his father. But of a very different type, inherited from his mother.
* Someone asked if epigenetics was bringing back the idea of Lamarckian inheritance. I didn't understand his answer.
Sigal and her former evolutionary psych professor (Cauffrey?) were there, and we had an interesting geeky conversation (even if largely beyond me, due to my lack of background).
Rough summary / questions:
* epigenetics can mean many different things. Etymologically, "epi-" is like "beyond".
* one kind is CG-methylation (aka "Ç" if you think of methyl as cedille) which renders genes inactive.
It is a kind of statefulness that can get passed on during transcription (due to methyl's tendency to cross the DNA), but apparently not meiosis(?), related to inactivation of genes (IIRC, also X-inactivation in females)
* in the zygote, the methylation state gets reset (methyl goes away?); human cloning is a bad idea because this resetting doesn't happen, leading to severe malformations. Of course, this raises the question of how methylation state could possibly be inheritable.
* diseased genes can be passed for many generations, e.g. down a paternal line, and only manifest themselves once they pass this gene on to a daughter, who passes it on to a child, who shows the diseased phenotype (or vice-versa: invisible down a maternal line until they have a son who has a child). I wonder what kind of mechanism could explain this.
* I found it a bit surprising that the zygote is formed before the two gametes' DNA get together (duh!), meaning there's no attempt to check how well this individual sperm matches this individual egg. This lack of selection implies that the zygotes formed are "really" "random". Though I wonder if sperm DNA is related to sperm phenotype, leading to a selection there.
* You can observe methylation state as a green light, using rats bred with genes for bioluminescence! There were two types of females (presumably, due to them having two Xs), one showing no light, one showing a dim green light. All males showed a bright green light. I don't understand this.
* The presenter himself was colorblind. So was his father. But of a very different type, inherited from his mother.
* Someone asked if epigenetics was bringing back the idea of Lamarckian inheritance. I didn't understand his answer.
Sigal and her former evolutionary psych professor (Cauffrey?) were there, and we had an interesting geeky conversation (even if largely beyond me, due to my lack of background).
![[identity profile]](https://www.dreamwidth.org/img/silk/identity/openid.png){kind=small}
(no subject)
Date: 2008-11-10 10:34 pm (UTC)Aren't those cells already mature, and therefore useless in embryonic development?
Doesn't the body gradually get rid of such cells as foreign?
(no subject)
Date: 2008-11-10 11:00 pm (UTC)* one kind is CG-methylation (aka "Ç" if you think of methyl as cedille) which renders genes inactive.
usually it's written CpG. that p indicates there's a phosphate between them, so they're on the same strand (with the C on the 5' end and G on the 3' end), rather than a C and G on opposite strands base-pairing.
It is a kind of statefulness that can get passed on during transcription (due to methyl's tendency to cross the DNA), but apparently not meiosis(?), related to inactivation of genes (IIRC, also X-inactivation in females)
you mean DNA replication, not transcription (of RNA). It certainly works by turning transcription up or down (usually down with methylation), but transmission all about making new DNA.
* in the zygote, the methylation state gets reset (methyl goes away?); human cloning is a bad idea because this resetting doesn't happen, leading to severe malformations. Of course, this raises the question of how methylation state could possibly be inheritable.
it's heritable because the same sites get methylated during DNA replication. Yes most of it is reset, *some* is not. and severe malformation is not the thing you tend to worry about, since those fetuses tend to spontaneously abort. the problem is with stuff that doesn't quite kill the fetus that's an issue, like "whoops, looks like the immune system's not working so well".
* diseased genes can be passed for many generations, e.g. down a paternal line, and only manifest themselves once they pass this gene on to a daughter, who passes it on to a child, who shows the diseased phenotype (or vice-versa: invisible down a maternal line until they have a son who has a child). I wonder what kind of mechanism could explain this.
here's a really simple one: you make a lot more testosterone than women do, T regulates gene activity related to sex determination (and other stuff) if your screwed up gene is in a pathway that's affected by whether the organism has a lot of T vs a little, then there's your situation.
* I found it a bit surprising that the zygote is formed before the two gametes' DNA get together (duh!), meaning there's no attempt to check how well this individual sperm matches this individual egg. This lack of selection implies that the zygotes formed are "really" "random". Though I wonder if sperm DNA is related to sperm phenotype, leading to a selection there.
some people do hypothesize about sperm phenotype for sure, but to answer the rest, yup, you're right. That DNA recombines later for the next generation's gametes. It's not quite random though, if you've got a bad mix, that stuff just tends to abort spontaneously.
* You can observe methylation state as a green light, using rats bred with genes for bioluminescence! There were two types of females (presumably, due to them having two Xs), one showing no light, one showing a dim green light. All males showed a bright green light. I don't understand this.
so they put the gene on an X. All the X's of a male rat are turned on because they only have one. In females, there's dosage compensation so we turn off half of our X's. I imagine all the rats only had one gene on one X so that the females had half turned off, and males didn't have any turned on.
* The presenter himself was colorblind. So was his father. But of a very different type, inherited from his mother.
about 1/10 men are colorblind, and it's almost always inherited from one's mother. it's unsurprising that he inherited the most common form of color blindness from his mother. I don't know what rare kind of color blindness his dad has, but that's the more interesting one.
* Someone asked if epigenetics was bringing back the idea of Lamarckian inheritance. I didn't understand his answer.
prolly because it's a stupid comparison. in some ways it technically IS bringing back the idea of Lamarckianism, but it's not really at that level yet.
(no subject)
Date: 2008-11-10 11:09 pm (UTC)ah, so a different X is turned off in differents cells. That makes sense for the dim-light females.
but then there's the other type of females, which appeared totally dark...
(no subject)
Date: 2008-11-10 11:12 pm (UTC)(no subject)
Date: 2008-11-10 11:13 pm (UTC)(no subject)
Date: 2008-11-10 11:33 pm (UTC)we do know that fetal cells can break off from the placenta and lodge in maternal tissues like the liver and differentiate into functional cells there and live happily, but that's not the same as what you claim.
(no subject)
Date: 2008-11-10 11:34 pm (UTC)(no subject)
Date: 2008-11-10 11:34 pm (UTC)(no subject)
Date: 2008-11-11 12:24 am (UTC)As for the Lamarckian stuff, I think epigenetics just shows that the situation is more complicated than anyone thought during the Darwinian/Lamarckian dispute. Most heritability is genetic, but there are some sorts of acquired characteristics that can be passed on, and we definitely have to pay attention to that when looking at detailed molecular-level things.
Just as cancer shows that things are more complicated than people thought during the endogenous/exogenous debate about disease. Most diseases are caused by external factors, but some are triggered by purely internal things, and some involve some sort of complex interaction.