I've been busy — I'm teaching genetics this term, and usually the first two thirds of the course is trivial to prepare for — we're covering Mendelian genetics, and the early stuff is material the students have seen before and are at least generally familiar with the concepts, and all I have to do is cover them a little deeper and with a stronger quantitative component. That's relatively easy.
The last part of the course, though, is where we start moving into uncharted waters for them, and every year I have to rethink how I'm going to cover the non-Mendelian concepts, and sometimes my ideas work well, and sometimes they don't. If I teach it for another 20 years, I'll eventually reach the point where every lecture has been honed into a comprehensible ideal. At least that's my dream.
Anyway, one of the subjects we're covering in the next lecture or two is imprinting, and I know from past experience that this can cause mental meltdowns in my students. This makes no sense if you're used to thinking in Punnett squares! So I've been reworking this little corner of the class, and as long as I'm putting together a ground-up tutorial on the subject, I thought I might as well put it on the web. So here you are, a basic introduction to imprinting.
Here's a somewhat hypothetical situation. You have two genes, A and B, which have different roles in the two sexes. In females, A is inactivated by modification of the DNA — methyl groups are added to the DNA strand to prevent transcription. I've drawn it below with a little pink box to indicate the female pattern of methylation. The female still has two active copies of the B gene.
The male has the complementary pattern. He uses the A gene, but turns off the B gene with methyl groups. In my cartoon, I've drawn a little blue box to show the male pattern of methylation.
Now the happy couple produces gametes, eggs and sperm. The female produces egg cells that still have the female pattern of methylation—gene A is turned off. The male produces sperm that have gene B turned off. And then they get together at fertilization to produce a zygote or embryo which has a mixed pattern of methylation: on one chromosome, A is inactivated, and on the other chromosome, B is inactivated. This means that the zygote still has one functioning A and B gene…it has both working.
What's described above is the normal state of affairs. Now it may be that the zygote needs both A and B genes functional — there could be functions at the sexually indifferent stage of early development that demand that both genes be operational. Later, when the sexes differentiate, then the embryo may change the pattern of methylation to that seen in one sex or the other. So, for instance, if the zygote happened to be male, it would remove the little pink methylation block to turn on both A genes, and put a blue methylation block on the B gene.
So far, so good, I hope. Now let's look at an unusual situation.
This diagram is just like the one above, with one change. The male has a deletion on one chromosome, marked in red, that destroys both the A and B genes on that chromosome. He's fine, though: he still has a working copy of the A gene on the other chromosome, and because he's male, he had the B gene shut off anyway. It's all good, until he starts producing sperm…and half of his sperm will carry the deletion, and therefore will not contribute either an A or B gene to his progeny.
With some genes, this is OK, because he can count on his female partner to provide good copies of the A and B genes, making up for his shortcoming. In this case, though, there's a problem: yes, the mother provides an A and B gene, but the A gene is methylated or imprinted, and doesn't do anything in the embryo! This can be big trouble if the A gene provides something essential for early development.
There's another way this problem can come up that doesn't involve a deletion—it's called uniparental disomy.
In this case, the mother has a nondisjunction, an error in meiosis that produces an egg with two instead of one copy of the chromosome. This is bad news, because when fertilized by a normal sperm, the zygote will be trisomic for that chromosome. Trisomies are deleterious; the only viable autosomal trisomy in humans is trisomy 21, also known as Down syndrome, so this accident is most likely to result in a spontaneous abortion and loss of the embryo. However, sometimes in the cases of trisomies, cells will 'kick out' the extra chromosome and develop normally. That's a useful mechanism for rescuing the embryo from a meiotic error, but in this case it has an additional problem.
What if the extra chromosome that is booted out is the one from the male?
Now the zygote is left with the right number of chromosomes (it is disomic), but they both come from one parent (uniparental), and have the pattern of imprinting of that parent. In this example, both A genes are maternally imprinted and blocked from expression, so the embryo is effectively lacking A.
Does this ever happen in the real world? Yes. In humans we've got one clear example, called Prader-Willi syndrome. There is a series of genes on chromosome 15 that have different patterns of methylation in males and females, and as diagrammed above, in normal cases the zygote receives both the male and female pattern of methylation, and all is well. If a zygote receives only the female pattern, either because of a deletion on the father's chromosome or because of uniparental disomy, it will develop abnormally, exhibiting a range of common effects that include early hyperphagia (they can't stop eating), obesity, and learning disabilities.
There is also a complementary syndrome. What if the deletion is on the maternal chromosome, so only the paternal pattern of gene expression is found in the zygote? It's exactly the same genetics, but the sole difference is in whether the trait is passed on from the mother, or from the father. In this case, the child would have Angelman syndrome, cause by an absence of necessary maternal gene products. Angelman syndrome is much more severe, causing large developmental delays, seizures, and ataxia, or jerky movements (also, curiously, Angelman kids are often also cheerful, happy people).
The important point here is that you don't just inherit genes: you also inherit a history of modification of those genes. Your chromosomes are modified by passage through your mother and your father, and normally, that pattern isn't visible because the modifications complement each other to produce progeny with a functional mixture of active and inactive genes. In some cases, though, heritable or developmental changes to the chromosomes allow us to peek through and see the effect of imprinting.
- Log in to post comments
Came for the atheism, stayed for the science.
Thanks PZ
Came for the science; then rejected atheism.
More limitations coming from evolution the Designer.
My, it's inscrutable, in a curiously deluded and sinister manner.
Glen D
http://tinyurl.com/6mb592
I think that that sounds like a good explanation. The real-world examples really help. Thank you for the informative post!
Wow!
My neice has Angelmans, and I had wondered the root of it. This is the clearest explanation I have ever heard. Thanks!
This is a really nice explanation, especially the explanation of why it might be important to have both genes active up to a certain point. I found it easy enough to follow, and liked the use of human genetic problems for an example.
Nice. Genetics over lunch and General Relativity before work. Good day.
That's good. First the admission that your thinking processes are either blocked by prejudice, or incapable of coming to an intelligent conclusion. Then should come a good education, and the practice of thinking. And someday, you might be able to think rationally.
Then again, if you're Facilis and/or Eric, improvement seems a dim, dim hope.
Glen D
http://tinyurl.com/6mb592
Sorry PZ, I'm going to have to disagree with you. Even given a strict definition of "viable" such as "can produce offspring" which rules out most of them, Triple X is certainly viable.
I can understand why you'd want to simplify things for class, but I think you still need to be accurate.
Now you've made me go an insert the word "autosomal".
This isn't the simple version for class. This is the simple version for the world that hasn't taken my class.
Null_Hypothesis: "Came for the science; then rejected atheism."
You "rejected" atheism because of this blog? That's amazing! Somehow I tend to think you had always accepted God and while you may have questioned your belief there was always belief. Or else you wouldn't use the phrase "rejected atheism" becuase you would understand how silly that sounds.
@PZ #10
Fair enough. :)
Thanks for the breakdown PZ, but makes me a little paranoid of all the things that could happen if I ever have children. I'll have to take a look at how common these syndromes are.
PZ, I honestly don't like your explanation!
I think you may have simplified to the point it is no longer correct.
I find figure one especially troublesome. It makes it seem as though imprinting is epigenetically homozygous in adult cells, and is maintained through gametogenesis. While the details are still fuzzy, it seems that the consensus in the field is that the imprinting is wiped clean and reset somehow during gametogenesis.
My understanding was that the male/female imprints can be maintained into adulthood. What is the basis of saying the imprinting patterns goes homozygous in adults after development? Rather than being maintained as epigenetically heterozygous?
Also, I would rather see the IGF2 pathway story used as an example than disease. Imprinting allows males to upregulate the IGF2 pathway while females downregulate it (the probable result of an evolutionary arms race over fitness). Removing the female imprints involved results in too-large embryos, removing the male imprints results in runts. Maintaining imprinting results in balance due to the heterozygosity at the epigenetic level.
Did you make the figures yourself?
Not only online HTML validation, but online proof-reading too :o)
This is simplified. Imagine revising the cartoons so that the imprinting is cleared (as it is) and reset at different stages in development -- it would be unwieldy.
Oh the anger..... it must bother you so. This is a topic about imprinting, so I won't further taint it with philosophy and religious crap.
Wouldn’t you expect the adult female and adult male to also have a maternally and paternally inherited imprinting pattern, just like the zygote?
Okay, I figure it was simplified, but there is still the problem of the adults being epistatically homozygous for the imprinting!
At least for some genes, imprinting is maintained as heterozygous through adulthood. It even seems like loss of heterozygosity for imprinting may be related to cancer and other diseases.
I would rather see a revised form of your figure one, showing heterozygously imprinted adults, and setting of the imprint one way or the other at the arrows indicating gametogenesis.
I don't think it makes the story more complex to teach that the imprint is set during gametogenesis. As it stands I think it is misleading, maybe even wrong.
pink sasquatch : Oh please consider poor little people like me who never took biology even as an elective in undergrad. I'd like to understand what you've described, but at the moment it's a bit beyond me. That's the trouble with simplification I think... but you have to start some where to have something to add further and more accurate explanations to.
blueelm @20
Sorry, I wasn't gearing my comments to the non-geneticist - I was using specific lingo to get across to PZ what I see as a problem in his explanation.
Basically, in figure one I think the adult chromosome diagrams should both look like the zygote chromosome diagram. As it stands now I think it is incorrect.
Hopefully that clears up some of my lingo-infested criticism.
Simplification is fine, unless it is wrong!
Makes sense to me and gets across the core concept in a manner that would continue to make sense were I to print out the pictures and come back to them later, without your words explaining.
Apparently it's simplified a lot according to the other posters but I felt that it gets across that point you concluded with quite clearly.
I don't know about you, pals, but from now on I'm going to ignore Null_Hypothesis for my own mental health.
PZ, thanks for the lesson! I just took my bio final yesterday, though... I can't impress my professor with my new imprinting knowledge now.
Well, I felt like I learned something right up until Pink Sasquatch crushed my sense of minimal accomplishment. Are you telling me it takes more than surfing teh Internets to understand biology? What a disappointment...
That was very well written, very well explained. In fact, I'm going to have to bag some of the approach for a tutorial next term. Credit where it's due ;-)
It is clear when explaining science, even to an undergrad class, that it often helps to strip it of the minutiae of detail such that the general premise can be conveyed. For those interested in the topic, it is then a matter of building upon this foundation, supplementing the initial model with the evidence of the greater detail that one would invariably expect in any biological system.
Genetic imprinting is an absolutely fascinating subject, and the data on other epigenetic processes in everything from bacteria to humans are hugely exciting.
Firefox + Greasemonkey + killfile
Man, I don't remember imprinting being anything like that. I'm really going to have to reread Heretics of Dune.
wow, censorship ... I guess freedom of speech is under attack from all parties. If you don't mind, I'd like to use this topic as an opportunity to learn about epigenesis and imprinting to further my understanding of the fascinating world we are a part of. All hail Wikipedia!!!
hehe, that was a typo. Should have read epigenetics, not epigenesis. Kind of funny.
So far all I understand is this:
XX and XY hook up, get jiggy, and most often, what is predicted to happen, happens. However, at random intervals, there is an epic fail and then shit happens. Some failures are understood to cause specific, and possibly negative results, but is this also one part of the mutation process related to evolution?
PZ, maybe you need to expand on this more, since I was lead to understand that this is a very important concept for embryonic development as well as for mature living which I learned about when reading "Microcosm: E. coli And The New Science of Life" By Carl Zimmer.
Way to not understand the difference between being ignored and being censored. I too will reject the Null Hypothesis.
#11, there are people who discover PZ while going through an atheistic phase, and don't find all the answers here either.
Being mildly theistic can be much for comforting in daily life than PZ's fundamentalist atheist brand. And if only because Xmas is so much easier to explain to kids as "birth of Jesus" than "shopping mania".
How does one be a fundamentalist atheist?
Hey, Null. Maybe you should use Wikipedia to look up the definition of the word "censorship." If someone decides not to read your crappy post, that ain't "censorship." That's "smart."
Wow, comprehension fail.
Maybe it would help if instead of seeing the "Adult" cells as belonging to a somatic line, we're actually looking at cells in the adult in the midst of oogenesis or spermatogenesis.
"I too will reject the Null Hypothesis."
Haha, so you accept H1? What would that be?
Wow what a compliment. You so cannot stand to have your assumptions challenged by observations, that you too would prefer to bury it out of sight. I think I'm on to something!
killfile totally doesn't work for me. All it does is put the gumbies in the background of the comment, but it doesn't remove the comment. If it actually removed dumb comments like it said it was supposed to, then I'd recommend it to others, but...
I am jealous of you people with firefox.
Mu: Eh, I guess. Why on earth would they be looking for all the answers here though!? Seriously. Then again, I came here for hot beard action and squid porn.
@Null Hypothesis:
What is your objection?
Null: "I think I'm on to something!"
No, really you're not. You just aren't clever enough that people find you worth arguing with so they'd just as soon ignore you. I do this with my grandmother often, but then again she has alzheimers. What's your excuse?
It is free... Or are you at work and locked into IE?
I thought I understood the explanation, but is there a one sentence definition of imprinting? Something like, "Imprinting is the process whereby genes are expressed only from either the maternal or paternal line." Or am I misunderstanding? The whole rest of the post seems an attempt to explain how this can mess up occasionally.
Someone complains about coming for science, then argues about everything but -- and subsequently hijacks the comments of a layperson who would just like to understand a basic concept...please see comment #30.
Maybe make explicit in the diagrams that these are adult germ line cells?
revBDC: yep. I can't download it at work. They get really angry when we do that and won't fix any other problems that occur on our computers so long as the offending sofware is there. I don't want another rebuild! I shouldn't complain. Company X lets us post on blogs at least!
I think that I'd be a bigger fan of this explanation if it was a degree or three more sophisticated/complex. Even for the general public, there is such a thing as oversimplifying a bit too much. We're not here because we're the general general public - we're here because we're the general nerdy public, after all.
"What's your excuse?"
I had a mean daddy. Scientists are discovering that it changes the development of the brain. I guess I'm a freak.
Yeah I understand. I am a member of "they" or rather I am the king of "they" at my company.
"Maybe make explicit in the diagrams that these are adult germ line cells?"
That would make enough sense for me. At first I might not understand why that was significant, but later I would have absorbed passively the idea that it might be. Good teaching method there.
So let me see if I understand this correctly. Are you saying that, through imprinting, mutations which developed in parents during their own embryo phase, but which are not represented in their complete gene line, can be passed on to their offspring? Is this something like Lamarck's inheritance of acquired characteristics at the genetic level, or is it more mundane and procedural than that?
well I've enjoyed this friendly banter, but I am now seriously going to get down to the serious business of learning about imprinting.
To Alice, according to Wikipedia, "In mammals, genomic imprinting describes the processes involved in introducing functional inequality between two parental alleles of a gene."
pink sasquatch@21: Simplification is fine, unless it is wrong!
All simplification is wrong. In fact, all models are wrong. It's just the degree to which they are wrong that matters. :-)
Sorry. The statistician has to get his jabs in when he can :-) Remember Monty Python? "It's only a model!"
Excellent! So simple even a physicist can understand it!
A suggestion, unless I'm misunderstanding -
If the graphic showed a piece actually separate from the chromosome to demonstrate the deletion, it might be clearer what's going on. It is that kind of deletion, rather than a deactivating point deletion, right-?
rBDC: LOL! I work for a different part of the larger body that "they" also work within. So we're allies!!!
blueelm, you can get a lot of interesting points on the intellectual side of atheism in this blog if you filter out the polemic bashing of creotards. I consider that part the entertainment.
"they" are often too busy stopping stupid people from doing stupid things to be able to allow not-so-stupid people to do un-stupid things like install firefox.
Also, if they allowed the not-so-stupid people to do moderately unstupid things, the stupids would cry afoul and demand to download SpamGenerator XXXThousand and the Spyware-o-Rama Toolbar
Stogoe:
Interesting! I've never seen that behaviour or noticed it reported before. What browser version are you running, and have you poked into the killfile script code itself?
Oh — and how does it behave on other blogs? I'm pretty sure that the gumbies are defined via some CSS which is specific to Pharyngula.
Actually, these aren't mutations. Think of them as padlocks. They don't change the DNA base pairs, they prevent the genes from being used. (Yes there's a chemical change, but it is fully reversible - take off the methyl group and you have the original DNA) So yeah, it's just something that's mundane and procedural.
Hi PZ
I am colourblind, a protonope. This means I don't process red light at all. This creates all sort of colour confusions, for example, I can't tell the blue and the pink apart. (That's weird, I know...) My suggestion is to not necessarily rely on the colours to identify what you intend to discriminate. The labels are good, a highly visible key is good, a background-other-than-white is good. Just please don't assume because they are different colours to most people that there isn't someone like me in your class who can't see the difference.
I know from my own experience, that I'm especially prone to not distinguishing colours especially on high contrast backgrounds like projected slides.
You may have students with colour deficient vision like mine or a related deficiency.
That is my one and only technical criticism, and it's one I wish I could have shared with all my professors down the years. I put it here for your consideration and I hope it helps.
RH
Will there be actual mating involved in the lab component to the lecture?
"Oh the anger..... it must bother you so. "
I've often wondered if trolls actually think they're making people angry by posting something stupid, or if they hope to make people angry by calling them angry when they've shown absolutely no signs of being angry.
Either way, it's a huge waste of effort and time.
well, that was understandable, even with the in-comment corrections :-)
however, just like another poster said, this stuff makes me paranoid about breeding. bu the time i might be ready to do so, I'd probably be too scared of all the shit that can go wrong to actually go through with it :-p
@PZ #37
It would help. It might also be worth (depending on how much development they've received so far) explaining that there's a LOT of new DNA methylation which occurs in the blastocyst. DNA methylation itself is reversible, with the exception of those in the Imprinted genes... which behave (largely) as you describe them.
My Post-doc is looking at where those methyl groups come from in blastocysts.
at convo regarding firefox
i cannot download at work either... so i have firefox, photoshop, and some random programs installed onto a jump drive that i just leave plugged in. photoshop runs a little slow but firefox runs perfect(bonus is that i can go to the sites they have blocked on IE)
Really? How is "we give each other presents to make each other happy" harder to explain than — deep breath — "The Creator of the Universe, who apparently decided to hide all evidence of his existence, decided to forgive humans for violating rules which he made up, so he sacrificed his son, who was really himself, to himself, in a way which left at most the tiniest bits of historical evidence, and which people always describe as 'dying for our sins', even though at other times those same people hate the idea of punishing the innocent, and even though it wasn't really 'dying' anyway, so much as it was a really bad weekend, because the son whom the father sacrificed to himself is now up in Heaven deciding who is good and bad, and today we celebrate the birth of that son-who-is-really-also-the-father (on a day which some guys picked because they wanted to take advantage of the celebrations which were already happening this time of year) by telling each other stories which are really amalgams of the weird and contradictory tales handed down in various books which were written decades after the events they claim to describe (the oldest surviving copies of which date to decades later still), stories which feature stars leading people around like GPS systems and a whole lot of murdered babies."
@62
Wait, I could be confused myself, but I was under the impression the original comment was about deletion (fig 2), and this is describing Methylation (fig 1). But I could have misunderstood the ?...
Thanks for posting this. My education in the sciences is not very extensive, so every little bit helps.
Ah, ok; thanks W. Kevin Vicklund.
(God sends the "wise men" a dream to warn them to get their asses out of town, but he doesn't do anything to stay Herod's soldiers or safeguard the innocent children? All those babies and toddlers had to die just to demonstrate the seriousness of the situation? Dude! Bethlehem was like totally a town of redshirts!)
"All models are wrong, but some are useful."
Huh! And all this time I thought imprinting had something to do with geese getting obsessed with Konrad Lorenz!
"John, when people thought the earth was flat, they were wrong. When people thought the earth was spherical, they were wrong. But if you think that thinking the earth is spherical is just as wrong as thinking the earth is flat, then your view is wronger than both of them put together."
— Isaac Asimov
Thanks, PZ. This article is the best explanation of my nephew's trisomy 21. The little guy wasn't viable for long, but he was amazing.
The science articles don't often get as many comments as the rest do, and I haven't expressed my appreciation for them as much as I should have. I will now, as commenting on all articles is actually easier once logged in on the new system. Thanks, PZ.
Null_Hypothesis, you are an idiot. You popped a provocative comment in on #2, completely unprovoked, accused a person who commented on it of having anger issues, sprayed around some poor-quality snark, then said you were wanting to focus on the science. Way to hypocrite, you null-brained religieuse.
PZ, let me ask you a specific question or two that may clear up some of my SIWOTI feelings towards your post:
You have an example of good-for-male and good-for-female genes being selectively and homozygously turned off or on in opposite patterns in males or females as a result of imprinting. What are you basing this on? Is there a single real example of this?
My understanding is that imprinting accomplishes the selective expression of EITHER the male OR the female allele in BOTH sexes - both during development and in some adult tissues. This is hypothesized to be the evolutionary outcome of different fitness interests for males and females expressed through their progeny. Imprinted genes are biallelically expressed in some tissue/developmental timepoints, but I hadn't heard that they were biallelically turned off in one sex and turned on in the other via a parental imprinting mechanism.
Can you give one example of a gene that is homozygously turned off in one sex, and on in the other, by imprinting?
In other words, one example where imprinting produces sexually-dimorphic expression of a gene?
Richard Healy -> Have you ever tried the Protanope tools here? The RG2BY (probably the one that would work for you) and RG2MG links on that page you can right-click and bookmark; when you run into an image, viewing the image on its own web page and clicking the bookmark will throw it through the tool's processor.
I didn't know that pink and blue would be confusable colours in protanope vision. Vischeck was such a neat site for trying to check out how web sites look to colour-blind people, but I felt so bad for taxing their server by looking at ICanHazCheeseburger that I haven't been back :)
Blake:
I use IE6 (I know, right!?) at work, and Firefox 3.x at home. Pharyngula is the only place where I have killfile on my home computer.
So imprinting does give us males an enduring and essential part in reproduction. The sci-fi scenario of the women getting rid of us guys and reproducing parthenogenetically won't work. Thanks, imprinting.
Whew!
But just tell it to the whiptail lizards.
You're over-analyzing. The chromosomes also don't do it with little pink boxes for the girls, and little blue ones for the boys. The key thing here is to get across the basic concept, of genes that are differentially activated/inactivated in sperm and egg. This has been the hard idea to get across to students.
Otherwise, yes, these processes are much more dynamic than I can illustrate in a couple of quick cartoons.
Sorry, PZ, I'm not over-analyzing, and after a quick brush-up on the subject, I'll say your description is flat-out wrong.
"You have two genes, A and B, which have different roles in the two sexes."
WRONG. This is not a characteristic of imprinted genes.
"In females, A is inactivated by modification of the DNA — methyl groups are added to the DNA strand to prevent transcription. I've drawn it below with a little pink box to indicate the female pattern of methylation. The female still has two active copies of the B gene.
The male has the complementary pattern. He uses the A gene, but turns off the B gene with methyl groups."
WRONG. Imprinting does not act in a biallelic, sexually dimorphic manner as you describe.
"Later, when the sexes differentiate, then the embryo may change the pattern of methylation to that seen in one sex or the other. So, for instance, if the zygote happened to be male, it would remove the little pink methylation block to turn on both A genes, and put a blue methylation block on the B gene."
WRONG. This does not happen. Imprinting does not act in a biallelic, sexually dimorphic manner as you describe.
My criticisms have nothing to do with how "dynamic" the processes are, or your cartoon representations: they have to do with the very basic facts you are presenting being completely untrue.
Imprinting accomplishes the selective expression of EITHER the male OR the female allele of a given gene in BOTH sexes, NOT sexually dimorphic expression of a gene.
Honestly, I am unsure that you understand imprinting yourself given your description of it. PLEASE do not teach these falsehoods to your students, or leave them unchanged here - others have mentioned that they will use them for educational material elsewhere.
I suggest the following as a decent (though possibly dated) review:
Nature Reviews Genetics 2, 21-32 (January 2001)
Genomic imprinting: parental influence on the genome
Wolf Reik & Jörn Walter
Hmmm. When did you install the killfile script? The latest version is applicable to many more blogs; it'd be interesting to see how it behaves on your computer.
Re 79
Ritchie> I'm not familiar with that site in particular, but I am registered with Colbindor (referenced at the bottom of the page) which has some *very* useful tools and articles on colour vision deficiencies and I've used to filters on Vischeck (for the first time EVER I saw the numbers on the Ishihara plates - it was a thrill, I don't mind admitting!)
The in-browser or image up-loader correctional aids work by boosting the spectral channels, which alters the hues of each colour but makes distinguishing colours of different hues possible. It's a kind of messy fix to the problem because it generates a kind of false-colour image that helps to distinguish different colours easier by altering the contrast (which is the problem) but the colours are themselves changed, so it's kind of useless to refer to them as blue and pink anymore - even if the distinction didn't mean much before. That's because filters absorb some spectral frequencies, those that pass through appear brighter and it is the distinction in contrast between the brighter and darker colours which then distinguishes them though still not by hue.
What I mean is my actual colour space, my spectrum is reduced. A protanomolous person is less sensitive, in my case it's shorter. I can and will confuse hues of all colours within a range because more of my colours are similar. And I will tend to distinguish colours using other cues like context and especially brightness and luminosity. Isolated colours (on, for instance, slides) where there is no context, the luminosity is the same, and my reduced spectrum just plays havoc with how sensitive I am to the colour differences and how visible they are.
On PZ's 1st slide, I can follow the example *because* they are labelled but the colours aren't distinct. Although I appreciate for everyone else this must sound mad.
because ...er red-green-colourblindness - doesn't that mean you only confuse red with green? and the answer is no. I can confuse pink and blue - and here's why:
For a dichromat like me, every possible colour I can match is a result of matching just two of the primary colours of light rather than the normal three. If you plot a graph for the colour space of visible light (like the CIE 1931 Chromaticity diagram) it's possible to draw on that graph straight lines which represent all the points along which the colours will look the same, that is will have the same hue and saturation.(These are the pseudoisochromatic colours that form the basis of the colour vision tests)
Light is of course subtractive when it comes to colour, and there are only three primary colours, so all hues are made of a mixture and combination of these . A pink object - a case in point - is reflecting both red and blue light. (try googling an RGB slide and making pink mixing amounts of red and blue) then remember that I do not have any long wavelength (red) cones so am not sensitive to red light, and suddenly it makes sense why I can (very) easily confuse pink and blue.
Colblinor has some especially good explanations of colourblindness and it's effects. I recommend it highly.
You left out the rather crucial sentence before my description.
I am not describing any real genes. There is no A and B gene. There are no pink and blue blocks. There is no comparably simple situation in human genetics. I know the real situation is much more complex than what I've described here.However, you can't simply throw something on the level of the Reik and Walter paper at undergraduates cold -- you have to get them to understand the general concept first, then you build to real world complexity.This isn't so hard. It's how genetics has always worked — we also know that relatively few genes operate with the clear simplicity of Mendel's rules, but we still teach them those simple cases first.
PZ, that's a ridiculous sidestep and you know it. How would you react if I said it was fine to teach Intelligent Design in public schools in great detail as long as preface it with "here's a somewhat hypothetical situation"? "Here's a somewhat hypothetical situation", DNA is single-stranded everyone, since you are incapable of understanding the complexity of double-stranded DNA... Just plain ridiculous.
(And I'm not asking you to throw the Reik and Walter paper at undergrads, but I do think that figure 3 of their paper is easier to understand than your figure 1.)
My point has nothing to with complexity, it has to do with you teaching imprinting falsely.
- Imprinting DOES NOT result in sexually-dimorphic expression of genes.
- Imprinting DOES result in the selective expression of either the male or the female allele of a given gene in both sexes.
I'm not sure why you think the latter is necessarily more complex than former, but that shouldn't matter in any case when one of them is FALSE, and the other is TRUE.
If something is too complex for students to grasp at a given level, we shouldn't teach it wrong just for the sake of teaching it.
I honestly don't understand what your issue is on this point, other than perhaps you don't feel like revising your tutorial. You should not teach students things you know are false. Your description of imprinting is wrong. Not oversimplified, but WRONG. It would be better for you to not teach it at all!
In any case, it is very frustrating. Why not revise it?
Pink Sasquatch, what is your experience in teaching introductory courses?
Imprinting and color blindness, two interesting topics for the price of one.
Okay PZ, let me ask another question:
Why mention the (incorrect) sexually dimorphic expression phase at all?
By doing so, you are ADDING complexity to your description, not simplifying it. You are also making your explanation untrue.
Look again at figure 3 of Reik & Walter and compare it to your explanation. You've added an unnecessary, incorrect step in your description.
Why? It only complicates things.
Blake, I don't bother with most of the stuff, which is why I will never be asked to give a theological discourse at church. I'm one of those cherry-picking-the positive-skip-the-nonsense(especially everything BC) guys. Luckily, my specific brand of superstition is pretty easy, as long as I go for my crackers and don't kill anyone.
Count me as another interested, nerdy, but non-science-trained type who appreciates getting the simple explanation as a baseline.
Until a child has the ABCs figured out, reading (in English) is impossible. Don't ask us simpletons to read a chapter book when we're still working out that C sometimes sounds like CAT and other times like CELL.
@ #88 Nerd of Redhead
I do have experience teaching undgrads, but it shouldn't take any experience teaching to come to the conclusion that knowingly teaching something wrong is acceptable. It is better to not teach it all, or reserve it for higher-level classes.
PZ's version of imprinting here is wrong by definition. Check NCBI bookshelf for any number of definitions of imprinting from standard textbooks. None of them will include the weird (incorrect) extra step that PZ has added to the process in order to make it more 'simple'.
@ #92 chgo_liz
PZ isn't teaching the ABCs in this post. He's adding a few extra letters to the alphabet that no one else uses. It's only complicating the issue.
That's why I find it so frustrating.
Pink Sasquatch, as an ex college teacher I strongly disagree with you. Some concepts are best learned in increments. The students need to have the proper background, sometimes in other disciplines, in order to understand the subtleties of the concepts. If I introduced NMR spectra in introductory organic, I didn't expect the students to learn third order coupling as sophomores. That can wait.
pink sasquatch:
Ahem.
So, you're saying that PZ either is knowingly teaching something wrong or isn't doing his job properly. This based on a blog post to a general audience.
I think that's an extraordinary claim.
No, I've stripped away a great many complications to present this as if the male/female differences in imprinting were constant -- we all know this is not the case. However, I'm not really willing to get into erasure and resetting of methylation states and epigenotype spreading at this level of discussion. The point here is to introduce a very general idea, how deletions can expose different patterns of gene expression on maternal and paternal chromosomes. I am not talking about developmental patterns at all. That's it. It's a cartoon of the actual genetics, OK?
It's nice that you mentioned the Reik and Walter article. In class, right after I introduce the concept by way of the admittedly greatly simplified version shown here, I bring up figure 1 from that paper and we talk about the actual genes involved, and I try to explain that there are many more complexities involved. That's about where it has to be left, because there are only 3 more weeks in the term, and I still have to cover chromosome structure, a little bit of genomics, and some developmental genetics in that time...and the obvious reality here is that everything has to be abridged.
Pink Sasquatch:
Fortunately (for everyone) I deleted my first post to you. RIght or not, the imperious hand-wringing tone is... off-putting. (There - no hair curling, spittle flecked invectives this time) & Unless you're being facetious, lose the exclamation marks or your posts might be as well be typed in Comic Sans font with a screened Monty Python character in the background. Have a good day!
No wonder why my high school history classes never got past World War II.
Well, keep in mind that Pink Sasquatch is right — I'm really compressing the whole process a lot to try and get straight to the core principles. If you're closer to the subject, it hurts to see important bits whittled away.
There's not much I can do to satisfy everyone. Pedagogy is often about compromise.
Definitely.
Sounds like a good lesson to me. I'm a senior in hs now (not off to UMM, sorry, PZ), took AP Bio last year, and with that background it pretty much makes sense to me this way.
Sounds like this is for a decently basic level course, right?
Mu (#91):
Theologians do that too. In some sense, the tradition is as old as Christianity itself: the first guy we know of who tried to establish what the Christian canon should be was Marcion of Sinope (c. 85–160 CE), whose Bible consisted only of a few Pauline epistles and an edited version of Luke.
PZ (#100):
No. Not even in the face of Armageddon. Never compromise. Hehn.
PZ, maybe I'm not getting across my key point to you. I'm really not trying to come off as an unsatisfiable asshole.
You say, "It's a cartoon of the actual genetics, OK?"
And I say, "Okay, but the cartoon is wrong, OK?"
Imprinting never goes to biallelic sexually dimorphic expression like you state.
This is an EXTRA step that you've added. PLEASE let me know that you understand this. You've added an EXTRA step to the process/cartoon.
I don't care about the developmental details, erasure or resetting. I wouldn't teach that either. I care about the genetics. And the genetics, as you've presented them, are wrong.
Here is the extra step you've added:
"Later, when the sexes differentiate, then the embryo may change the pattern of methylation to that seen in one sex or the other. So, for instance, if the zygote happened to be male, it would remove the little pink methylation block to turn on both A genes, and put a blue methylation block on the B gene."
This is not part of imprinting, and seems to me to complicate things. You are not "stripping away the complications", you are adding some. And they are wrong.
I hope you understand this.
I'm with Pink Sasquatch on this one. It's a disservice to couple one correct concept, "maternal and paternal chromosomes can have different patterns of gene expression, which we discover through deletions" with one more obvious, broader-ranging, simpler, and incorrect concept, "these genes act differently in males and females." How are you going to explain deleterious effects from uniparental disomy while your diagram clearly shows that two pink boxes are BAD for the embryo, but FINE in the mother of said embryo?
Any of these students that go on in genetics, or encounter a disease affected by imprinting, are going to remember "different in males in females" when they ought to be remembering "different when FROM males than when FROM females". IMHO, that is not so hard to explain. I suggest:
An imprinted gene is a gene whose alleles have different levels of activity depending on whether the allele came from the sperm or the egg that produced the animal.
Here's another non-science explanation from keepkidshealthy:
"In children with Prader Willi Syndrome, this chromosomal deletion is derived from the father. If the same microdeletion were derived from the mother, then it would cause Angelman syndrome instead. The process where a chromosomal abnormality can cause different syndromes depending on if it is inherited from the mother or father is called genetic imprinting."
No I'm not a professor; I'm just a genetics M.S. who has TA'd introductory genetics and adores imprinting.
@Nerd of Redhead #95
I agree with your post entirely. We don't disagree at all.
I have a problem with an incorrect part of PZ's description, and not his oversimplification of it. He is, in fact, adding an extra, incorrect step that complicates the process. His description is wrong because of this.
This doesn't happen; it is wrong:
"Later, when the sexes differentiate, then the embryo may change the pattern of methylation to that seen in one sex or the other. So, for instance, if the zygote happened to be male, it would remove the little pink methylation block to turn on both A genes, and put a blue methylation block on the B gene."
Apparently I'm not communicating well here...
@ Pteryxx #105:
I'm happy you showed up. I was beginning to feel like a lone kook.
Pink Sasquatch, you remind of one of my old colleagues who was terrible at teaching P-chem to the juniors. He expected them to grasp concepts that were easy for him to see, but difficult for the students. He tended to grade harshly since the students did not pick up on the subtleties. And he was not thought well of by the students.
Sometimes one should just introduce a concept without going heavily into details that might confuse the students, even if it may be scientifically inaccurate due to subtleties.
I'm certain PZ has a better idea of what his students are capable of learning than you.
blueelm:
Try http://portableapps.com/ for all your Firefox-on-a-(USB)-stick needs.
Pink Sasquatch:
I admit I am not a teacher and I don't know that much college level biology, but I think I am following you pretty well. To me what PZ is doing sound almost exactly like how some concepts in Comp. Sci. are taught (for very good reason).
Stuff is sometimes vastly simplified, or sometimes even taught in an actually incorrect form in order to get the concepts across. For example, we are first taught ascii and straight binary counting, with no mention of ISO 8859, unicode/UTF-8 or binary coded decimal, two's complement, or IEEE floating point. The freshmen learning binary counting and ascii would freak out and abandon the program if you started with FP, two's compliment and Unicode.
This is true, your point isn't coming across at all well. What I'm trying to get across in those few lines is the idea that the pattern of methylation is not fixed, but will change in the early weeks of embryogenesis—that the pattern inherited from the parents will be erased and reset to something different. The only reason I'm reusing my little pink and blue boxes is that they don't add an unnecessary new pattern to the picture; I could have drawn them with a green box in a different location, and explained that this will be changed yet again when this embryo matures and begins gametogenesis and starts slapping the blue/pink boxes on DNA in the germ line, but trust me...tossing in a whole bunch of novel methylation patterns and multiple steps to the long, long process is going to add nothing but confusion. That has to be saved for later, not on the first day students get exposed to the idea.But otherwise, this is simply a cartoon version of erasure and alteration of methylation patterns as the embryo differentiates, as Reik and Walter explain it.
The only thing I've done in my description is limited myself to only using two possible methylation patterns. I also tell the students this is a gross simplification. I am not about to get into a discussion of every nuance and detail in this kind of class, or with this particular internet audience. You're free to do so yourself, and that's actually one of the purposes of these comments — so that more informed readers can expand on the stuff I only skim over. Where you're going awry is in not simply trying to explain more cool stuff about this topic, but instead seem to be freaking out over a pedagogical necessity.Build on the rough framework I've thrown up in this post, don't throw a tantrum and try to throw it all out. That doesn't help with understanding at all.
Another kook here. I think this post does not really avoid talking about erasure and resetting methylation states, rather it implies that they occur during embryonic development (when the embryo is already multicellular - how would that work exactly?) I don't think it would make the explanation more complicated to put the erasure step in the right place.
PZ:
But it's not because the embryo needs both genes during development and then switches both chromosomes to the "female" or "male" position during sex differentation. That wouldn't be imprinting, it would be something else. The pattern of methylation in the somatic cells of the embryo is fixed. The change occurs only in the germ line of the embryo.
I used to teach celestial navigation. Just about the first thing I'd tell my students is "Forget Copernicus, for our purposes the Earth is fixed in the center of the universe and the sky rotates around the Earth." This completely false concept makes celestial navigation easier to understand and makes the math a whole lot simpler.
PZ I see another glaring mistake.
There is absolutely no bacon in that post.
As one who is severely unhabituated to the fiendish complexity of developmental biology or genetics and just about all science dealing with them thar animicules and their innards, I certainly couldn't say whether whatever I'm reading is "wrong".
But what I read in PZ's "ground-up tutorial" here made pretty good and consistent sense to me. I read through it again. I still groked it.
An introductory tome to ease the transition from Mandelian to non-Mandelian concepts. (PZ says so).
From the "relatively easy" to something potentially confusing enough for "mental meltdown". I girded my mental loins. I got through it. I didn't see any inconsistencies.
Despite a few treacherous spots it made good sense every step of the way. Then again, I'm a dunce, and I always manage to find it treacherous. I have to concentrate hard and go over and over and reread repeatedly whenever I'm reading about things that boast a metabolism and the ability to reproduce. Just the phrase "organic chemistry" makes me wince. I had to look up "Punnett squares". Had no idea what the hell that was. Had to find out. But I have an unusually stubborn nature (or is it masochism?) and I refuse to accept defeat. If I start it, I must finish it. I just want to understand.
I'm slow, but that also gives me a good idea of why technical material is difficult for people who aren't nearly as stubborn as I am. But just because I'm slow doesn't mean I can't detect a false argument or an inconsistent one. I didn't see any in PZ's piece. It made sense to me, from top to bottom. I could TELL he was employing a hypothetical exercise. HE SAYS SO.
A means of using a hypothetical in order to introduce a transition to another concept, in this case, non-Mendelian genetics in terms of an exercise.
Many mathematicians and maths instructors are familiar with the tactic. They do it all the time. It may not be pretty, but it's practically useful. It really works. The exercise doesn't have to be right. All it has to do is introduce the tools in a consistent way, step by step, so one can see how it might work, stepwise. (Even musicians know the value of exercise).
With mangy screwy formalisms, you don't throw students into the deep-end of the quicksand without arming them first by building a few new neural connections (dendrites, right?) with which to cope. You give them a chance to grasp a foothold, a purchase, a ladder of some kind by which, rung by rung, step by step they can confidently reach the next restful ausicht and be able to gasp, with a comprehension that apparently comes from nowhere, "AHA!!!"
It clicks because of the scaffolding that was erected. Never mind that the final building doesn't look like the scaffold. The scaffold isn't "wrong" just because it doesn't exactly match the looks of the building. The scaffold has nothing to do with the building other than to have fostered the building's construction. Without the scaffolding, the building would never exist. Once the scaffold is removed, the building stands. That's what counts.
PZ concludes in his last paragraph what his "EXERCISE" was all about. There's nothing "wrong" here. It's a valid introduction that sets things up for the final assault towards the "AHA!!!" - WHEN, it comes to actually understanding this business about "imprinting". I'm sure PZ has those next chapters too.
Even though I've adored it since I was a child - I LOVE biology - he's once again made me think AND comprehend what was formerly an impeneterably complicated concept that is (to me, anyway) a frontier of a field which has often been as difficult for me to digest as tar.
pink sasquatch. PZ isn't teaching undergrads here. He's just giving some ropes of exercise to dopes like me. I certainly don't dispute your contentions, but I'm pretty sure you don't have much of a handle on PZ's actual intentions here either. You insist PZ must be accurate with the business of imprinting, but as much of a dunce as I am, even I can plainly see that's NOT the whole point of his post. He mentions it as a point of departure, saying, in effect, "in order to see how this works, try thinking about this exercise". It will help. It's not always a question of immediate accuracy. It's just a question of whether people ever get a chance to understand a relatively complicated thing in the first place.
The difference between a good teacher and a lousy one is that the good one knows how to ease me into it by exercise like this, that shows me I actually CAN float. The lousy one throws me directly into the deep end of the quicksand. There's actually no real TEACHING in that. No transition whatsoever. All of a sudden one finds oneself sinking in cold quicksand, and nothing is learned...except maybe something idiotic like, "difficult science can't be taught to slow-witted people" or, "I'll never understand this stuff because teacher thinks I'm a dimwit".
Try and relax. Practicing the chops, even if they're a bit off the mark is a Good Thing. It CAN lead to undertsanding the accurate parts, just by suggesting the differences.
Or are you suggesting that scientific instruction - even in an informal setting such as this - is an inherently dogmatic enterprise, where an audience (or students) cannot be trusted to have the wit to find out more for themselves after having their curiosity piqued? If you think so, than you would probably consider people like me even dumber than we already know we are. Such an attitude would make you a false teacher: there'd be no point to teaching OR making sure that everything was absolutely accurate, then, would there?
Nerd of Redhead @ #108
Again, I don't disagree with what you are saying, and I'm not worried about subtleties/details/simplification in the way that people seem to think.
I don't think you get why I'm frustrated, but since you have a chemistry background I'll put it to you this way:
This isn't comparable to teaching NMR without teaching third order coupling. This is comparable to teaching NMR by describing how time of flight mass spectrometry works.
"Build on the rough framework I've thrown up in this post, don't throw a tantrum and try to throw it all out. "
But even the rough framework is off. I also didn't try to throw it all out - I suggested you remove all the incorrect stuff you added about sexually dimorphic expression, which is extraneous and has nothing to do with imprinting. It simply does not belong there, and adds complication and confusion. I'm not sure why that is so hard to understand. The rest of your post would be fine. I'm also not sure why you are so defensive of your blue and pink boxes. I have no problems with those, either.
Honestly, you've made a number of false statements in the post that make me wonder if your understanding of imprinting is correct.
"What I'm trying to get across in those few lines is the idea that the pattern of methylation is not fixed, but will change in the early weeks of embryogenesis—that the pattern inherited from the parents will be erased and reset to something different."
If this is what you are trying to get across, then you are indeed misunderstanding imprinting.
I am not saying that facts in your post are wrong because it is oversimplified.
I am saying that facts in your post are wrong because they are wrong (at least in the context of imprinting).
You write: "Later, when the sexes differentiate, then the embryo may change the pattern of methylation to that seen in one sex or the other. So, for instance, if the zygote happened to be male, it would remove the little pink methylation block to turn on both A genes, and put a blue methylation block on the B gene."
This has nothing to do with imprinting and will not be found in any textbook under imprinting.
"Where you're going awry is in not simply trying to explain more cool stuff about this topic, but instead seem to be freaking out over a pedagogical necessity."
Sorry I'm not adding "cool stuff" or that I seem to be a pedagogue. Really I'm "freaking out" because it sounds as if you plan to teach some weird hypothetical situation about sexually dimorphic expression and call it "imprinting". That's just not right.
What textbook are you using, out of curiosity?
PZ, your description of imprinting is wrong. Period. It is not wrong because of simplication, it is wrong because you added extra complications that have nothing to do with imprinting. It is not hard to correct, and I'm not sure why you are so stubbornly resistant to the idea.
Sorry, this was wrong:
The pattern of methylation in the somatic cells of the embryo is fixed.
Should have said that the imprinting pattern in the somatic cells of the embryo is fixed. The methylation itself is not fixed. As the Reik and Walter paper says, one of the outstanding questions is:
How are imprints maintained when there is genome-wide active and passive demethylation in the early embryo?
But it is maintained, somehow, otherwise we wouldn't have things like the Prader-Willi syndrome. Right?
Well, I, for one, both enjoyed the original post and the ensuing conversation. Both have taught me quite a bit. Thanks moth to PZ and pink sasquatch.
I have to admit that without PZ's refresher, I wouldn't have remembered enough to get the sasquatch's point.
Eukaryotes are weird.
Pink Sasquatch, you are being unnecessarily pedantic. As one who has taught remedial chemistry to a graduate level course, if your audience can't understand what you are trying to teach, you shouldn't be teaching that way. My third order argument is correct. This needs to be simplified so PZ's students can understand the basic concept. You are so wrapped up in the correct science that the total concept will blow over the student's heads. Been there, done that, revised notes.
Im not a specialist in this,but this abstract from an article in Nature does seem to suggest that PZ's simplified explanation is at least not incorrect.
http://www.nature.com/ng/journal/v9/n4/abs/ng0495-395.html
Pink Sasquatch - Just noting the original objection*:
then
Perhaps you'd care to correct it, since it's not hard to do - what edits should be made to the post to get it not wrong?
--
* and, what proportion of the general public would follow that? ;)
I could have taught third order NMR coupling in introduction to Organic Chemistry, but if it took me two weeks to get there, and the organic group decided at most 2 days on NMR, it would throw the whole years syllabus off. You can't always go into gory details and get where you must to hand students off to the next course if you fail to cover the expected material. Sometimes you sacrifice depth to get breadth.
Good point. You don't like that paragraph? Rewrite it. I'm willing to patch in a revision, as long as you don't turn it into a complicated mish-mash.
To Nerd of Redhead and John Morales:
Sorry if I'm speaking over your heads, but in this thread I'm not speaking to you, I'm not speaking to undergrads, and I'm not speaking to the general public. I'm speaking to PZ Myers, who should be able to grasp the terminology since he is qualified to teach the material.
I have no problem with simplifying material or leaving out extensive details.
Your repeated hounding on these points is off-base, since I don't disagree with you.
for starters (sorry for butting in pink sasquatch), the normal "Adults" all should look like the "Zygote" in Figure 1. This was pointed out early on by pink sasquatch and Matt in #18.
To PZ @#123 who says, "Good point. You don't like that paragraph? Rewrite it. I'm willing to patch in a revision, as long as you don't turn it into a complicated mish-mash."
You know, earlier I actually considered sitting down and spending some time rewriting your post and rearranging your figures to try to get my point across. And then I thought:
"Fuck that. Let PZ write his own damned lecture."
It should be enough for you to realize that your description of imprinting is wrong, or at least misleading, to change it yourself. You are the one responsible for it.
Good night.
To Rorschach #120: I think that paper is describing a special case in which imprinting is not reset... but I only skimmed it. I was working on this:
To John Morales #121:
"Perhaps you'd care to correct it, since it's not hard to do - what edits should be made to the post to get it not wrong?"
Please see below for my suggested edits to PZ's original material.
-------------
Here's a somewhat hypothetical situation. You have two genes, A and B, which have different roles in the embryo. As the ovary produces eggs, A is inactivated by modification of the DNA — methyl groups are added to the DNA strand to prevent transcription. I've drawn it below with a little pink box to indicate the maternal pattern of methylation. The prospective germ cell still has two active copies of the B gene.
The male germ cells have the complementary pattern. His sperm use the A gene, but turn off the B gene with methyl groups. In my cartoon, I've drawn a little blue box to show the paternal pattern of methylation.
Image 1 edit
Now the happy couple mixes their gametes, eggs and sperm. The female produces egg cells that have the maternal pattern of methylation—gene A is turned off. The male produces sperm that have gene B turned off. And then they get together at fertilization to produce a zygote or embryo which has a mixed pattern of methylation: on one chromosome, A is inactivated, and on the other chromosome, B is inactivated. This means that the zygote still has one functioning A and B gene…it has both working, just like the parents.
So far, so good, I hope. Now let's look at an unusual situation.
This diagram is just like the one above, with one change. Some of the male's sperm have a deletion on one chromosome, marked in red, that destroys both the A and B genes on that chromosome. If a sperm carrying the deletion fertilizes an egg, it will not contribute either an A or B gene to his progeny.
Image 2 edit
With some genes, this is OK, because he can count on his female partner to provide good copies of the A and B genes, making up for his shortcoming. In this case, though, there's a problem: yes, the mother provides an A and B gene, but the A gene is methylated or imprinted, and doesn't do anything in the embryo! This can be big trouble if the A gene provides something essential for early development.
There's another way this problem can come up that doesn't involve a deletion—it's called uniparental disomy.
In this case, the mother has had a nondisjunction, an error in meiosis that produces an egg with two instead of one copy of the chromosome. This is bad news, because when fertilized by a normal sperm, the zygote will be trisomic for that chromosome. Trisomies are deleterious; the only viable autosomal trisomy in humans is trisomy 21, also known as Down syndrome, so this accident is most likely to result in a spontaneous abortion and loss of the embryo. However, sometimes in the cases of trisomies, cells will 'kick out' the extra chromosome and develop normally. That's a useful mechanism for rescuing the embryo from a meiotic error, but in this case it has an additional problem.
What if the extra chromosome that is booted out is the one from the male?
Image 3 edit
Now the zygote is left with the right number of chromosomes (it is disomic), but they both come from one parent (uniparental), and have the pattern of imprinting of that parent. In this example, both A genes are maternally imprinted and blocked from expression, so the embryo is effectively lacking A.
Does this ever happen in the real world? Yes. In humans we've got one clear example, called Prader-Willi syndrome. There is a series of genes on chromosome 15 that have different patterns of methylation in sperm and eggs, and as diagrammed above, in normal cases the zygote receives both the maternal and paternal pattern of methylation, and all is well. If a zygote receives only the maternal pattern, either because of a deletion on the father's chromosome or because of uniparental disomy, it will develop abnormally, exhibiting a range of common effects that include early hyperphagia (they can't stop eating), obesity, and learning disabilities.
There is also a complementary syndrome. What if the deletion is on the maternal chromosome, so only the paternal pattern of gene expression is found in the zygote? It's exactly the same genetics, but the sole difference is in whether the trait is passed on from the mother, or from the father. In this case, the child would have Angelman syndrome, cause by an absence of necessary maternal gene products. Angelman syndrome is much more severe, causing large developmental delays, seizures, and ataxia, or jerky movements (also, curiously, Angelman kids are often also cheerful, happy people).
The important point here is that you don't just inherit genes: you also inherit a history of modification of those genes. Your chromosomes are modified by passage through your mother and your father, and normally, that pattern isn't visible because the modifications complement each other to produce progeny with a functional mixture of active and inactive genes. In some cases, though, heritable or developmental changes to the chromosomes allow us to peek through and see the effect of imprinting.
---------
Pteryxx speaking again. I made minor edits, with the exception of one paragraph:
"What's described above is the normal state of affairs. Now it may be that the zygote needs both A and B genes (to have both copies) functional — there could be functions at the sexually indifferent stage of early development that demand that both genes (alleles) be operational. Later, when the sexes differentiate, then the embryo may change the pattern of methylation to that seen in one sex or the other. So, for instance, if the zygote happened to be male, it would remove the little pink methylation block to turn on both A genes, and put a blue methylation block on the B gene."
Ptx says: As far as I know, embryos don't activate both copies of imprinted genes in their somatic tissues, and imprinting patterns don't change during sex differentiation, again in somatic tissues. At least in general this doesn't happen; there may be special cases in depth. Therefore I suggest striking this paragraph entirely. Otherwise, I put my suggested edits in parentheses.
Thank you all for your patience and attention. -Peace, Pteryxx
Who needs religion when the real world is this awesome?
pink sasquatch,
I hope I'm not being presumptuous when I think you're wrong. You're addressing PZ, but speaking to all his readers.
Sure, you're way over my head in the biology.
But I'm a regular reader, as are thousands following this, and can follow discourse. Many will be like me, the general public.
Anyway, PZ made you an offer that would be the envy of many. Why not take advantage?
I have two points. First, I think this presentation could be made very effective if you take your concluding paragraph and modify it into an introductory paragraph. Still keep it as the final word on the presentation, but just add it to the beginning as well. In other words, tell your students what it is you're going to tell them, tell it to them, and then tell them what you just told them. (English and Humanity departments do have some use.)
Second, you talk about two sets of chromosomes coming from one parent. I'm wondering if this can give any insight into the recorded instances of asexual reproduction by sharks. I'm in over my head with my legs dangling helplessly on this one to go any further (...sorry), but it was the first thing that popped into my mind.
... and I just saw Pteryxx's response. Thank you.
<engage lurk mode>
*sigh*
this conversation ended up muddling me up so much that i had to look up imprinting to get what this conversation was about.
and I can't figure out why people are so freaked out over the sexually dimorphic thing, when apparently that's exactly what imprinting does in some insects...?
now if we're specifically talking about mammals, then apparently this has nothing to do with the gender of the embryo, but everything with which gene we're talking about. as I understand it, some genes are expressed only if they come from the father, while others are only expressed if they come from the mother. there doesn't seem to be a "choice" in this matter however. a particular gene with a particular imprint will always express the gene from the same parent. Right?
anyway, maybe the sexual dimorphism really isn't the best way to explain this. Maybe this would be clearer (assuming I got the concept right, anyway):
I'd have Gene A be mother-dependent, and Gene B father-dependent.
In Image 1 the adults would then have equally methylated genes (but differently, depending on the allele). then I'd have a middle step where one allele from each parent goes into the embryo, but the methylation disappears. and then I'd have a third step, in which they methylate, and then look exactly like the parents' set.
In Image 2, I'd have the father have a deletion on the unexpressed gene B(i.e. the one from the mother). then the embryo inherits that gene B(and one from the mother). then comes the methylation phase, and the mother's gene gets blocked, but there's no gene B from the father, so there's no expression!
In Image 3, I'd say how the expulsion of the entire male allele resulted in both the mother's Genes B being blocked by methylation, because that's a father=dependent gene and there's no "father" here!
Personally I'd find it easier to understand that way than with the sexual dimorphism. but maybe i got it all wrong anyway, and maybe I'm already introducing things I know that PZ's students don't know, so what do other readers think of my version?
note: I've only ever had High-School biology, so it's not as if I'm one of those scientists getting hung up on details. I just find the explanation easier without the dimorphism, but with showing that there's a deletion phase.
To John Morales: Thanks very much for your thanks. It helps reassure me that I didn't just do something monumentally stupid... or at least not universally so. *inclines head*
ah, while I was typing my epic, #127 did a more science-y version which is similar to what i wrote.
You ask this after PZ's snotty reply? Frankly, I don't get why so many people felt the need to be so dismissive of pink sasquatch. But thanks Pteryxx for stepping up, I think his/her version makes more sense.
In case you still can't see anything out of place with the original explanation, consider this part:
Was he born an adult? :)
Windy, I guess one person's 'snotty' is another's 'magnanimous'. Whatever, I already said it's above my head.
@136: It might depend on whether you think about it as an offer to revise a blog post, or a tutorial for a class. If the latter, it may come across as "I'm willing to let you revise my homework, as long as you don't turn it into a complicated mish-mash" - not that magnanimous ;)
I don't like this post. Too orthodox. Where is the lamarckian stuff?
;o)
Pink Sasquatch et al,
have you heard of the concept of 'lies-to-children'?
There's a wikipedia article on it here: http://en.wikipedia.org/wiki/Lie-to-children
Here's what the people who coined it give as their explanation, which is far more succint than I just wasted 10 minutes writing and deleting:
"A lie-to-children is a statement that is false, but which nevertheless leads the child's mind towards a more accurate explanation, one that the child will only be able to appreciate if it has been primed with the lie"
It's the latter part of that statement (the part I've emboldened) that is most important. You give someone a cartoon so they get the bigger picture then later explain why the cartoon is actually wrong about a key point, so they understand the broad concept and then the subtleties and nuances.
PZ is teaching imprinting 101 here, so I think that it's fair to use a lie-to-children for a small part of it to get the concept across.
PS, you are missing our point every time. I suspect PZ understands the material. But he has to decide how he can present it in a cursory way to his students so they get the basic concept. This may mean that the material is "dumbed down" and, to an expert, become "wrong". I see you are arguing he is "wrong", but then, I see no evidence he can add the total explanation, which may take days for the students to get, into a 20 minute presentation.
You need to write how you would present the material in those 20 minutes. Then you need to do the reality check of can PZ's students understand and learn the material.
@ Pteryxx #127
Many thanks for putting that together - hopefully it will be helpful to people. Yours appears to me to be correct and internally consistent.
I let myself be a bit too put-off by some of PZ's comments (whether or not he intended them to be off-putting) to sit down with a cool head and put something like that together...
Thanks again.
I know next to nothing about biology since I've not taken more than the college (high school here in America) level classes, but that made complete sense to me. I can see how it would get complicated if one actually knew more or read more into it, but thankfully for my work day and the need to concentrate once I get there, I don't.
The comments above, especially those that add more detail, make me want to pull out a biology book and actually read it.
Thanks for the brief lesson on genetics!
This is not a case of "dumbing down" or oversimplifying the material - I would not have a problem with that.
PZ's described something other than imprinting, in order to explain imprinting. It was wrong by definition.
Any student that tried to cross-reference PZ's explanation to a textbook would be hopelessly confused. If those students move on to higher classes or research projects they would need to unlearn and relearn the concept.
I realize that if your first exposure to imprinting was PZ's post, it might seem to make sense, but it is not even internally consistent, as windy and others have pointed out. Part what of is frustrating me is all of the comments thanking PZ for their understanding of imprinting, which is likely wrong if based upon that post.
Here is my "Jefferson Bible" version of the first ~four paragraphs of PZ's description - all I've done is delete the invented fantasy details that complicate and falsify the story:
___
"Here's a somewhat hypothetical situation. You have two genes, A and B. The female produces egg cells that have the female pattern of methylation—gene A is turned off. The male produces sperm that have gene B turned off. And then they get together at fertilization to produce a zygote or embryo which has a mixed pattern of methylation: on one chromosome, A is inactivated, and on the other chromosome, B is inactivated. This means that the zygote still has one functioning A and B gene…it has both working.
Figure 1 edit: http://img144.imageshack.us/img144/9108/imprint1.jpg
What's described above is the normal state of affairs. So far, so good, I hope. Now let's look at an unusual situation..."
___
That is actually far simpler, far shorter than PZ's explanation. True it could use a bit of a lead-in and polishing, but at least it is correct, and could serve as an initial exposure to the idea of imprinting. The rest of the post could follow much as-is from this simplified intro.
I'm tired of people claiming I'm complicating the issue, when I'm trying to simplify it.
PS:
No, I'm sure you you are able to simplify it. However, from your tone, I wonder if you wear dentures.
@ E.V. #144
I see that all you've contributed to this conversation is complaining about my tone. Good job.
I'm sure many a colleague has contemplated knocking your teeth down your throat PS.
Strange but true story about imprinting...
When I was a special ed teacher, I had a child with Angelman Syndrome and another child with Prader-Willie Syndrome. They come from the same genetic defect, but one is the result of imprinting in the mother and the other from imprinting in the father. The two syndromes manifest totally different phenotypes as well. They are both pretty rare... so what are the odds they'd both end up in my class!
Pteryxx has the better explanation. The problem with the original post is that it indicates that there is expression of both copies of an imprinted gene in one gender-females will express both alleles of gene B, males will express both copies of gene A. This is not true-for a given imprinted gene, only one allele is expressed, and its the same allele in both sexes. That is, only the allele of gene A you got from your Dad is expressed, and that's true whether you're a boy or a girl. Likewise, only the allele of gene B you received from Mom is expressed, and that's true whether you are a boy or girl.
In a sense, then, for imprinted genes, we're functionally heterozygous-even though no mutations in DNA sequence have occurred, we're only going to express one allele of such a gene, so we better hope it's a good one.
Nitpick correction to Chris Thorpe:
"In a sense, then, for imprinted genes, we're functionally hemizygous-even though no mutations in DNA sequence have occurred, we're only going to express one allele of such a gene, so we better hope it's a good one."
That's exactly what hemizygous means: only one useful copy when we normally have two. It doesn't just apply to imprinting; male mammals, having only one X chromosome, are hemizygous for most of the genes on the X. Females then inactivate one X-chromosome in all of their cells so they don't get an unnecessary double dose of all the X genes.
While imprinting inactivation is parent-specific and happens in the germ line, X-inactivation is random and happens in the somatic tissues during embryonic development. The random inactivation results in female mammals made up of patches of tissue with one or the other X inactive: mosaicism. This pattern is famously visible in the coats of calico cats, which have two different alleles of a coat color gene on their X chromosomes. The orange coat gene is active in the orange patches, and the black coat gene in the black patches on the same cat. Males only have one X, one copy of the gene, and one coat color. Hence coat color is an "X-linked" trait.
...Whoops, pardon my reflex lecture. This tends to happen when I'm having too much fun with a topic, and/or avoiding doing my statistics homework. *blush*
E.V., what the fuck, man?!
For those who think this is justified as a teaching tool. Imagine if I taught Mendelian genetics like this:
Generation 1:
100 green peas AA + 100 pink peas Aa
Then I explained to you that pink behaves as a deleterious recessive. You might go "wait a minute, that's not consistent with your original situation!" Is that all right, as long as it's a "toy model"?
To PS: Heh, and thanks for your thanks. I'm a bit embarrassed though, because I started to edit PZ's material back around post #121, before the pedagogy war really got going. So it seems my OCD'ing just happened to make my intervention look timely. Such is the internet.
"If I find anything that may be to the good of the patients, be it either in Galen or Paracelsus, yea, Turk, Jew, or any other infidel, I will not refuse it but be thankful to God for the same."
-Unknown 17th-century English surgeon
And if parental imprints were erased and replaced with new ones during embryonic development, wouldn't you expect similar mosaicism in expression of imprinted genes? When in fact, mosaicism seems to be rare in these genes?
Also, maternal and paternal imprinting can be detected in adults using methylation analysis (here's an example of diagnosis of Prader-Willi and Angelman), which shouldn't work if the imprints had been erased in the embryo!
Windy:
PS states
It's hard to stomach correction when the bearer is pedantic, overbearing and smug. I'm sure this is PS's personal style - to correct everyone without tact or diplomacy in a public forum; in short Pink Sasquatch is no Sastra.
( I would not be surprised if he/she has probably had the locks to his/her office or car super glued a few times or other such payback pranks.)
Hey windy -
Yes, some imprinting is maintained through adulthood. It seems like there may be roles for loss-of-imprinting in adult diseases, such as some cancers.
This is also where the language can get complicated, since the methylation signature does not equate to imprinted expression. For example, a gene may have imprinted expression as a result of methylation, but only in certain cell type at certain times. That same gene, with the same methylation signature, may show biallelic expression elsewhere. (That's my understanding.) It likely has to do with different sites/effects of methylation on different regulatory elements for each gene.
So - the methylation that underlies imprinting may be maintained throughout adulthood, even though the gene in question no longer displays imprinted expression.
Hopefully that makes sense.
Oh, and thanks for the back-up vs. E.V. I know I've let myself get exasperated in this discussion, but he was adding nothing but nastiness to this thread. I'm choosing to ignore him.
@pteryxx. I typically think of hemizygous as referring specifically to a case where an individual only has one allele for a given gene-as when one copy is missing due to a deletion. That's not the case here-both copies are present, but only one is expressed. At any rate, given the discussion above about pedagogy and simplification, the last thing I wanted to do is introduce MORE terminology!
I take your point about X chromosome inactivation-I teach that before the section on imprinting, and that seems to help the students understand the whole idea of regulation of gene expression by chromatin modification, and sets up some nice contrasts with imprinting that they can grasp (whole chromosome vs. small # of genes, females only vs. both genders, random vs. parent-origin-specific, etc.). My students are actually kind of jazzed by the whole topic-its like a surprise plot twist coming after all the straight Mendelian inheritance stuff.
@Chris Thorpe: Would the proper term then be "functionally hemizygous" instead of "functionally heterozygous"? Because "heterozygous" does not imply that one of the two alleles is doing nothing - see blood types, for instance, where both alleles are expressed.
I'm not trying to make a point here, just having fun playing let's-review-genetics-party.
How is it pedantic and smug to say that something is wrong, when that's your honest opinion, and the other person doesn't seem to be getting it? And don't we always say to the religionists that "I don't like your tone" is not a good excuse? And don't you think that the second response you quoted might have been affected by PZ's own tone?
Seriously, would it be OK to excuse New Scientist's mangling of Darwin by saying "it's just a cartoon version of what happened, to get people interested in the science! It's simplified so you can't say it's wrong! And you are being pedantic and overbearing by insisting that we change it!"
To E.V.:
"It's hard to stomach correction when the bearer is pedantic, overbearing and smug."
*gently* That's all very true. Whoever is speaking may be annoying, insulting, snotty, whatever.
...But are they right?
I have the greatest respect for PZ Dr. Paul Zachary Myers and his contributions to science, reason, and this community. That's all the more reason that I should speak up when I see him making a mistake, and offer the best and clearest correction that I'm able.
I hope I've been polite, if overly enthusiastic. But regardless of how you think of me personally, or anyone else in this discussion, the correction should stand on its own merits. Isn't that why we're here?
Pterryxx, I've been corrected many times. I rarely resent it when it comes with tact, however I tend to find ways of evening the score when someone is intentionally humiliating me or abusing the point. PS seems more cluelessly pedantic and bratty than openly hostile.
There is a difference in going to someone's territory with many people around and saying, "pardon me, but could we talk about X, because I think an error has been made" vs. "You don't know what the fuck you're talking about - you are plainly WRONG! You have no business teaching this!"
Now my reaction to this poses danger of being misinterpreted as Nisbetian framing, but it's simply a question of the etiquette of initial engagement. Had PS started with more diplomacy and tact and PZ become rude and defensive, then descending to bitchy snark may have been overlooked (and perhaps welcomed), but PZ attempted to rationally clarify his stance and then conceded that PS's basic argument was correct. PS started at smug and elevated to asshole within 5 posts. It's hard to hear the message when the messenger is overbearingly pompous and pretentious from the start, I don't care how many degrees they have.
I don't care how right someone is, get in my face or needlessly humiliate me or someone in my presence, just know there will be consequences.
Shorter:
The argument was supposed to be rational and academic but it was immediately an emotional argument. I chose to shoot the messenger, he/she can receive credit for being right posthumously.
E.V.@#159
You're making it hard to ignore you.
"Had PS started with more diplomacy and tact and PZ become rude and defensive, then descending to bitchy snark may have been overlooked (and perhaps welcomed), but PZ attempted to rationally clarify his stance and then conceded that PS's basic argument was correct."
You apparently need to reread the thread. My first posts were diplomatic, and tried to get PZ to clarify what he meant.
Please see my first three posts, all diplomatic, to PZ: #14, #19, and #78.
PZ responded to each of my posts with short responses that did not cover any of my direct questions or criticism.
After PZ condescendingly dismissed me in #82 by informing me that there weren't really little pink and blue boxes on the chromosomes, I decided to take a more direct approach in my fourth post and tell him that he was WRONG on a few important points.
PZ continued to dismiss and sidestep, and started making borderline personal attacks. Now if this had happened in some thread about a cracker, or Jebus' face in a seat cushion, I wouldn't have cared much. But this is PZ letting us sample his lecture for his students, and his description is demonstrably wrong, and it is STILL unclear if he's going to teach the incorrect version to his students.
This caused me frustration and exasperation, and I elevated my tone. While some misrepresented or misunderstood what I was trying to get across, others agreed that PZ was incorrect. When PZ finally wrote, "you're right, it's wrong, and you should fix it for me," that was a huge slap in the face - more to his students and readers than to me.
You don't have your story straight, and YOU decided that my emphasis was emotional, even when it was just EMPHASIS.
Now you are regaling us with your bizarre little fantasies of everyone hating me and retaliating against me with violence or childish pranks in real life. I don't care what you think, but you continue to disparage my character and wish me harm even though I ignore you, and I don't like that. Again, you have added nothing interesting or constructive to this thread, and honestly, I don't see what offended you so much that you can't stop interrupting it.
But I do know that you hate CAPSLOCK AND EXCLAMATION POINTS!!!!!!!!!!!!
/rant
PZ, I'm not sure if you are still paying attention to this thread, but it would be nice to know that you saw Pteryxx's hard work at comment #127 rewriting your description.
If that seems too long and complex, I'd recommend something along the lines of my "Jefferson Bible" version of your description, at comment #143.
Thanks.
Peace, PS. Funny how a little hyperbole can ignite overreaction, huh? I'll give you props though, you're great at the drama queen game.
You misinterpreted him, it was you can (if you feel that strongly about it) not you should. That nuance makes a world of difference.
My gut reaction to reading and re-reading your posts is still the same -
and the word I've been looking for all day is petulant.
However , you deserve credit for correcting and fine tuning the errors. Peace.
do you mean post #100? That still seems to misinterpret pink sasquatch as objecting to the compression of the subject, rather than problems in the compressed explanation.
"But I try not to think with my gut" said Carl Sagan ;)
"You misinterpreted him, it was you can (if you feel that strongly about it) not you should. That nuance makes a world of difference."
No, it doesn't. PZ put up an educational post for the world, and perhaps his students, and is shown that is incorrect or at least so misleading as to be incorrect.
His first response shouldn't be, "oh well, anonymous-blog-follower, you can fix it if you want."
It should be, "oh crap, I'd better fix that right away."
_____
And no matter how many times you wish me "peace," I'll remember your violent fantasies about me expressed earlier, and stick with my conclusion that you are an asshole.
@pteryxx. I don't know if there is a 'proper' term when using an analogy to get across the concept here. In a colloquial sense,"heterozygous" at a locus (at least in my field) implies one wild-type allele and one loss-of-function allele. So the analogy here is imprinted allele=loss of function allele (in terms of function). Using hemizygous gets the same point across, as long as your audience knows what hemizygous means.
Yeah, Pink? Funny, you never denied that you were the victim of any pranks due to your "personal style" and your inability to do anything but over-intellectualize. So sad. Imagine the fantasies I'm having now, my impotent little Bubisch...
E.V.:
Again, WTF? Maybe that's because everyone understands you made that nonsense up, and it's not worth answering any more than "have you stopped beating your wife yet?" is?
No, pink sasquatch has been pointing out a rather basic misunderstanding. What is this, bizarro-thread?
I can understand PZ reacting less than cordially to criticism of his post, when he doesn't see where the problem is, but you are behaving like a complete ass! Take this from someone who normally appreciates your comments, or feel free to invent some nonsense about me.
@Chris: Thanks, I was assuming the general case that "heterozygous" meant "two different alleles" for your description.
@ps: Heh, yeah, would've been.
-------------
General suggestion: That the 'winner' of a flamewar shall be the party that does not post last. Kind of like reverse capture-the-flag. Therefore, I declare myself loser of this thread.
*solemnly hoists the battered red flag over a silent, abandoned and debris-strewn field*
-Peace, Pteryxx
Okay Windy, you win. Perhaps I was wrong about Pink Sasquatch. There was just a little schadenfreude in yanking PS's chain after the "Over your head" remark to Nerd and the {forgive the pun} punch line to the denture set up came out way too harshly when it was simply intended to be," If you speak to your colleagues like this, don't be surprised when your coffee tastes funny."
So, Pink Sasquatch: No more razzing from me.
OK, back to the topic of the post... Pteryxx's suggested images are good, but I don't think a figure which includes the somatic line necessarily has to be too complicated.
This figure is similar to figure 3 in Reik & Walter, but a little bit easier to follow, IMO. It was reproduced from figure 1 in this paper, which is also pretty good.
So, IIUC, there is no general erasure of imprints in the embryo, only in the embryonic germ cell lineage. I don't think it adds unnecessary frills to the explanation to say that imprints are not erased in the developing embryo, since it seems to be the distinguishing feature of imprinted genes! (see: fig 4 in Reik & Walter).
Did that make sense or am I out to lunch?
Makes sense to me; in general, the figure you cite is more accurate (and thanks for linking to the original paper that's been mentioned repeatedly, so I can now find it ; )
For simplicity's sake though, I think it's unnecessary to state that "imprints are NOT erased in the embryo" when it's even more straightforward not to mention erasure outright at all. In both diagrams, the one you cite and my edited version of PZ's originals, the pattern of methylation is reset in the germ cells: the parents have the same pattern of complementary imprints as the zygote.
I was trying to stick to PZ's original material as closely as possible, and also I was in a hurry, heh. My edited figures don't specifically say that resetting of imprints happens in the embryonic germ cells, or even make any specific distinction between germ and somatic cells. However, INM very HO, that distinction should be mentioned in teaching the material. I say that because even a very basic mention of genetics must have covered mitosis and meiosis, and therefore the students should have the concept of a germ cell lineage in which different genetic processes occur than in somatic cells: e.g. "Mitosis happens in every cell division except in germ cells, which undergo meiosis so they can have variety."
However, I also noted PZ's avoidance of the terms "somatic" and "germline", that the term "gametes" is mentioned only once, and that he uses extensively "male/female", less often "mother/father", and the term "maternal" is used only twice and "paternal" once. If I make the very large assumption that his blog post is word-for-word similar to his lecture material, then it seems odd to me to avoid these rather basic-to-genetics terms which the students must have been exposed to earlier in the course. Maybe the blog post was written for an audience less knowledgeable about genetics than his college students.
In any case, even leaving the blog text completely aside, PZ's original diagrams are incorrect. They show the male and female adult as carrying twinned imprints specific to their sex, when each adult actually carries complementary imprints with one maternal and one paternal copy. I think this is a more important distinction than specifying that resetting of imprints occurs during the embryonic phase.
*corrects self* Meiosis, so they can have variety and end up with one copy of each chromosome instead of two. Pardon me, forest for trees.
Actually my links were to a couple of newer reviews - this is the image from Reik&Walter.
Right, it doesn't need to be mentioned out of the blue, but if there's going to be some discussion of how "the pattern of methylation is not fixed, but will change in the early weeks of embryogenesis", then one should add "...except for imprinted genes [in somatic cells]"
And, one thing that might be useful to explain even in a simple tutorial like this is that methylation is not something that's limited to imprinting, imprinted genes are just the limited number of genes that are methylated differently depending on the parent's sex. Otherwise, I think it's easy to get the impression that methylation = imprinting.
To be fair, in #37 PZ suggested viewing them as adult germ line cells. But I don't think that would have been very helpful - it seems not to make intuitive sense and the original text did not support that interpretation. Your new diagrams and text would incorporate both suggestions - show complementary imprints in adults, and then the change in the germ line.
Another ironic thing I noticed afterwards is that the original diagrams basically work exactly like Mendelian crosses in Punnett squares (except each only shows one of the possible outcomes) - the pink and blue boxes behave just like they were alleles of A and B. The only deviation from standard Mendelian behavior is the nondisjunction, not anything related to imprinting. If the goal is to explain what's special about imprinting, it seems counterproductive to leave the change in the imprints outside of the picture.