X Inactivation and Epigenetics
September 14, 2013 2:50 AM   Subscribe

X inactivation is a type of gene dosage compensation. In humans, the sex chromosomes X and Y determine the sex of an individual - females have two X chromosomes (XX), males have one X and one Y chromosome (XY). All of the genes on the Y chromosome are required in male development, while the genes on the X chromosome are needed for both male and female development. Because females receive two X chromosomes, they inherit two copies of many of the genes that are needed for normal function. Extra copies of genes or chromosomes can affect normal development. An example is Down's syndrome, which is caused by an extra copy of part or all of chromosome 21. In female mammals, a process called X inactivation has evolved to compensate for the extra X chromosome. In X inactivation, each cell 'switches off' one of its X chromosomes, chosen at random, to ensure the correct number of genes are expressed, and to prevent abnormal development.
Here is a helpful eleven minute description of what it is and why it's important by Etsuko Uno and metafilter's own Drew Berry in a fucking gorgeous Goodsell-esque 3D animation.

There have been two huge posts about animations like this previously, one more academically focused and one more commercially focused previously.
posted by Blasdelb (34 comments total) 31 users marked this as a favorite
Never trust a biologist who says the word random.

Among other things, color blindness fails to express in women. If either X is normal, that's generally what's expressed (tetrachromats being an exception). That requires something to choose.
posted by effugas at 3:01 AM on September 14, 2013 [2 favorites]

There are also a few other points they said 'this is the case' rather than the perhaps more precise 'this is always the case except for the times its not.' For example the sex chromosomes of an individual always determine the sex of an individual, except for the times they don't. These kinds of simplifications have a complicated place in pedagogy as they are not quite accurate to our full understanding of complex systems that may be helped by them, but do allow for the best possible more shallow understandings that can lead to deeper ones.
posted by Blasdelb at 3:12 AM on September 14, 2013 [1 favorite]

Agreed, given that the alternative would be several thousand pages of background information, I'm OK with the use of random above, as well as the other statements that have been simplified.

I mean, that's how science works. It's a series of progressively longer factually incorrect conceptual explanations until you finally get to the truth; but by that point everyone else has gotten bored and wandered off and you publish your seminal work in dejected isolation, at best, to be recognized decades after your death.
posted by hobo gitano de queretaro at 3:15 AM on September 14, 2013 [10 favorites]

Personally, my go to weasel word is arbitrarily. It's provocative. No one knows what it means.

It gets the people going.
posted by hobo gitano de queretaro at 3:17 AM on September 14, 2013 [3 favorites]

The problem I have is that I keep seeing the use of "random", not when there are occasional exceptions, but when there's fairly obvious non-random behavior going on. And of course there's the polar opposite, when actual random noise is identified as signal (see that great paper on using a fMRI on a dead fish).

Science is hard.
posted by effugas at 3:20 AM on September 14, 2013 [1 favorite]

(Also, can we really not individually sequence two X chromosomes in a female?)
posted by effugas at 3:21 AM on September 14, 2013 [1 favorite]

Weirdie weirdo.

I just read about X inactivation in a book a few days ago. Me being me (there's a reason data is part of my preferred online identity), I had to go read up on it. A few days sooner would have made things easier.

Que sera, sera.
posted by Samizdata at 3:27 AM on September 14, 2013 [1 favorite]

"(Also, can we really not individually sequence two X chromosomes in a female?)"

At this point it would be more difficult, and pointless, to sequence them independently of each other.

The great poster on using a fMRI on a dead fish
posted by Blasdelb at 3:32 AM on September 14, 2013 [1 favorite]


No question it'd be difficult, but pointless?
posted by effugas at 3:41 AM on September 14, 2013

Wrote my preliminary exam for my phd on this. Totally fascinating stuff.
posted by sciencegeek at 3:48 AM on September 14, 2013

Never trust a biologist who says the word random.

Among other things, color blindness fails to express in women. If either X is normal, that's generally what's expressed (tetrachromats being an exception). That requires something to choose.

Do you have a source on this, effugas? What I was taught and what I've always taught my students is that women who are heterozygous for colorblindness therefore have both functioning and non-functioning cones in their eyes, but because on average half of their cones are functioning, they see color well enough to not realize a difference. It is random (or arbitrary if you prefer), just like a tortoiseshell cat's spots, but in this case we just don't see a clear phenotypic expression. Is this not how X-inactivation actually works? Have I been lying to my intro students for years?
posted by hydropsyche at 6:26 AM on September 14, 2013 [4 favorites]

Among other things, color blindness fails to express in women. If either X is normal, that's generally what's expressed (tetrachromats being an exception). That requires something to choose.

It is possible that the selection is skewed in certain organs. This, along with the mosaicism, would cause the color blindness to be much milder, so low detection is also possible. There are around 0.4% of women with color blindness, while the frequency for males is 8%.
posted by francesca too at 6:57 AM on September 14, 2013 [1 favorite]

Interesting. Unfortunately, the external sources on the Wikipedia Skewed X-Inactivation article are either broken or behind a paywall (I can look at those at work on Monday). Is there a theory regarding the process that could cause a skew in different organs? If X chromosomes with the colorblind allele are preferentially inactivated in the eyes, wouldn't that likely have other phenotypic effects since the X chromosome carries plenty of information other than colorblindness?
posted by hydropsyche at 7:11 AM on September 14, 2013

A good article on skewed inactivation. You will have to click each page.
posted by francesca too at 7:40 AM on September 14, 2013 [1 favorite]

That's really cool stuff, francesca too. I found the sections about MZ female twins and the families with strong skewing especially interesting.
posted by hydropsyche at 7:51 AM on September 14, 2013

This is very interesting. I do remember my high school biology teacher enjoying telling the class how all women were calicos, in that smug high school teacher, "This will blow your minds" kind of way.

Layperson questions: The inactive X isn't really fully inactive, though, right? It seems interesting to wonder how much an inactivated X chromosome is like a Y chromosome (or rather, like a third kind of chromosome more similar to a Y in function than an active X). Wikipedia is telling me about pseudoautosomal regions; if X-inactivation was a total shutdown of one X chromosome, wouldn't that mean that in females the PAR is only singly expressed, while in males it's doubly expressed? Do those regions actually stay expressed in an inactive X? Sex chromosomes are crazy.

Does inactivation happen in other species with autosomes? Some DNA decides, hey, let's randomly disable one of this particular pair of chromosomes from now on, and see if this fucks everything up. It'll be cool, we can put different kinds of genes there that have more complicated kinds of expression. Blah blah recessive blah mutations blah blah. What, do you hate fun now?
posted by bleep-blop at 8:09 AM on September 14, 2013

Woo, now everyone can understand my profile!
posted by maryr at 8:53 AM on September 14, 2013 [2 favorites]

bleep-blop, it's been a while since genetics class for me, but IIRC Drosophila males over-express their single X chromosome to match the pair of X's in females.

Without doing too much research, here's what Wikipedia says on sex determination in fruit flies:
Determination of sex in Drosophila occurs by the X:A ratio of X chromosomes to autosomes, not because of the presence of a Y chromosome as in human sex determination. Although the Y chromosome is entirely heterochromatic, it contains at least 16 genes, many of which are thought to have male-related functions.
So I think drosophila females are XXY and makes are XY and use the X twice as much. But again, it's been a while.
posted by maryr at 9:04 AM on September 14, 2013

Oops, re-reading I see that bleep-blop was actually asking about autosomes. Sorry, enjoy the potentially incorrect fruit fly gender trivia anyway.

And my "random" word of choice is stochastic. Sounds v. educated.
posted by maryr at 9:06 AM on September 14, 2013

swats away bio-nerds

Back to the video...

which was indescribably awesome.
posted by rhombus at 9:10 AM on September 14, 2013 [1 favorite]

*swarms back*

You can't get rid of us that easily. Turns out vinegar attracts more flies that honey.
posted by maryr at 9:19 AM on September 14, 2013 [4 favorites]

Very nice - thanks for posting. When I was in graduate school in the 1980s, epigenetics wasn't a respectable concept (Lysenkoism had poisoned that well) and now it's pretty much dogma. Since I missed out on a broad academic-style review of epigenetics it's nice to get caught up a little here on MeFi, especially in pleasant little infotainment nuggets like this.

On another note, I liked how the molecules constantly wiggled and flapped around from Brownian motion. Many animations show them fairly static, which gives a false impression of purposeful directed action, but here they just kind of bounced around until they hit something they stuck to. Random!

got a problem with that, effugas?
posted by Quietgal at 10:00 AM on September 14, 2013 [1 favorite]

The sound effects were hilarious. I especially liked the DNA methylation plinks.
posted by localroger at 12:15 PM on September 14, 2013 [2 favorites]

Ha. I'm still waiting for a post on uniparental disomy. Really weird stuff!
posted by francesca too at 3:29 PM on September 14, 2013


Tetrachromats have dual expression going on, for sure. I don't know of any evidence that 1/10 females have degraded color vision in any form, so you either have preferential inactivation, or the cells that do express are the ones that are hooked up to the rest of the system, etc. Lots of possibilities actually, I'll look into this more.

One interesting thing is that color blindness is not a single disorder, and the few color blind females I have interviewed tend to have stranger variants.
posted by effugas at 1:12 PM on September 15, 2013

(Also, tetrachromats are pretty rare.)
posted by effugas at 1:12 PM on September 15, 2013

Sorry, yes, I should have speciified red-green colorblindness in everything I said, because that's the most common form and the one we use as an example constantly in freshman biology.

I think the theory of simple mosaicism and female heterozygotes with red-green colorblindness is nothing to do with tetrochromats and everything to do with the old college freshmen "what if my blue is your red?" discussion. If a woman who is heterozygous for red-green colorblindness is able to distinguish the colors because of her approximately 50% normal cones, but doesn't see the same thing as someone with 100% normal cones, we might not necessarily be able to detect that. She could probably pass the vision test at the DMV, and unless her dad is colorblind, it may never have even occured to her that she might be.

But it sounds like you know much much more about colorblindness than I do (I'm an ecologist who has to teach genetics to freshmen--I pity the poor geneticists who have to teach ecology to freshmen), so please let me know if that's not the current understanding.
posted by hydropsyche at 2:48 PM on September 15, 2013


I wrote an app called DanKam that does color blindness correction (shift all hues to a hue that's seen correctly), then went on a bit of a tear studying the genomics behind the syndrome. Trichromatic color vision itself is a mutation; OPN1LW and OPN1MW ("red" and "green") are basically the same gene. The status of five base pairs basically describes the spectral peak of the opsin emitted, and most of the difference comes down to a single nucleotide.

There's only about 26nm of spectral peak difference between the two opsins, and 21nm or so is ascribed to a single nucleotide.

Of course, five nucleotides being shifted creates quite a few opportunities for variation. And there do seem to be occasional other issues that cause the entire opsin not to emit at all. The key is that I've seen no evidence for 1 of 10 females having even slight issues, which is what you'd expect if half the greens were nonfunctional. I have seen evidence of what happens if both X's express, which is tetrachromaticism. And that's really quite rare.
posted by effugas at 4:12 PM on September 15, 2013 [1 favorite]

I just googled your app. That's really amazing. I will definitely do some digging into the literature to try to figure out if the best current explanation is that heterozygote women have preferential skewing in X-inactivation. From your description, it sounds like that could be what's happening. In which case, I guess the only red-green colorblind women are likely homozygotes for the mutant allele (0.4% of the population is the number we throw around).
posted by hydropsyche at 4:37 PM on September 15, 2013


If we can't even sequence X's independently, I can't see how we'd be able to experimentally validate random X inactivation. That being said, you could still get skewed activation if all the incorrectly activated cells underwent apoptosis or something like that.
posted by effugas at 5:17 PM on September 15, 2013

Wikipedia suggests that the discovery that one X chromosome was inactive was thanks to noticing that one X chromosome appeared like those of the other autosomes but the second was condensed - the second is the Barr body. I presume that this work was done with karyotypes. Mary Lyons suggested that the Barr body was inactive based on coat expression in mice (and the process of X-inactivation is now called lyonization!) and was verified using a metabolic pathway gene (and a deficient copy of it) in blood cells. So sequencing the chromosomes was certainly not needed for the original discovery. You can probably use some nice Southern and Northern blotting techniques to verify without full sequencing too.

Actually, if I think about it, in this day and age you might be able to somewhat prove X-inactivation by cloning. Huh. I wonder - do cloned mammals only express from whichever X was active in the somatic cell the DNA was derived from? So if you cloned a tortoiseshell cat, would you get only black or orange kittens? Or does the oocyte you put the DNA in contain (or retain) the enzymes needed to reverse the inactivation (as apparently happens in germ line cells, but I don't when)? Sorry, this is all just me spitballing based on Wikipedia links and the two pages of the review francesca linked above.

Fun fact: Apparently, in marsupials, the father's X chromosome is always the one inactivated. Thanks, Wikipedia!
posted by maryr at 7:41 PM on September 15, 2013

PS: That review also notes a family where a mutation in X inactivation was discovered thanks to prevelence of X linked diseases. Mutants are awesome and super useful.
posted by maryr at 7:43 PM on September 15, 2013

PPS: Pure genetics: A great place to visit, but I wouldn't want to live there.
posted by maryr at 7:46 PM on September 15, 2013

effugas-- that's sort of what I was thinking about myself when you first mentioned colorblindness--that perhaps the cells that didn't correctly express their opsins didn't fully mature, or didn't become cones, etc. But I'm pretty much at a loss to really explain it, and a cursory lit search didn't help me. I'm actually planning on asking a prof tomorrow.

maryr-- Part of making induced pluripotent stem cells (iPSCs) is reverting all the methylation patterns on the chromatin, un-shutting-down the shut-down gene expression. (It's not just the Barr body; various sections of chromatin get more-or-less permanently shut down over the course of development.) So a calico cat clone would still be a calico, but would have a completely different arbitrary pattern, because 1) the arbitrary X shutdown would happen anew and 2) pigment cell migration isn't completely deterministic... Even in animals that don't have this kind of coloration, clones come out looking a bit different.
posted by Made of Star Stuff at 4:20 AM on September 16, 2013 [1 favorite]

« Older The Bible as fanwank and flamewars   |   Cosmic infection Newer »

This thread has been archived and is closed to new comments