This blog is no longer being updated. I've moved on to The Accidental Weblog. Hope to see you there.

Wednesday, December 15, 2004

A living missing link and its ten sex chromosomes

(More from the 'cool things in Nature' file. A fairly good commentary accompanies the correspondence concerning this research--it's out in today's issue.)

We've known for a while that monotremes are an interesting bunch. As they diverged from the rest of us mammalian types quite some time ago (210 million years or so back), taking a close look at them occasionally turns up features intermediate between the rest of mammalia and some of our more distant relatives--particularly reptiles (the most obvious and notorious being, probably, that monotremes lay eggs). Talkorigins has a FAQ on the general subject (pertaining specifically, as is usual in that venue, to monotremes' ramifications for a certain rather notorious pseudosience). Beyond this, as a long-diverged group, monotremes have also got some pretty odd characteristics apparently all their own. As Talkorigins puts it, they're far from 'living fossils', as they're probably about as distinct from the common ancestor as are the rest of modern mammals.

The useful thing from the point of view of working out mammalian natural history is: monotremes and the rest of us are occasionally distinct from that ancestor in different ways. And because they occasionally preserve different traits inherited from the common ancestor than do the rest of the living members of mammalia, monotremes sometimes tell us things about that ancestor--and thus about our relationship to our more distant relatives--that we didn't know before.

What might well prove to be just such an oddity came to light recently. Gutzner et al report in a letter this week to Nature (and in an earlier paper for the Proceedings of the National Academy of Sciences) that they've unravelled a number of critical aspects of the platypus sex chromosomes. This is the same research previewed in The Scientist some weeks back.

Gutzner et al's research shows that sex chromosomes in the platypus prove to be as odd as the creature that contains them. In most mammals (including ourselves and all our nearer relatives), males have an X and a Y chromosome, and females have two X chromosomes. The sex of offspring is determined by which component the male parent contributes: if the sperm that fertilizes the ovum carries a Y, the child is male; if an X, female (the female parent always contributes an X, since those are all she has). A parallel pattern exists in birds (which are believed to have evolved from dinosaurs some 150 million years ago--around 70 million years after mammals split from a reptile-like group called the therapsids)--though in birds it's the females that have the two different types--a Z chromosome and a W, whereas the males have two Z chromosomes; thus, which chromosome is contained in the egg prior to fertilization determines the sex of the offspring.

Platypuses, it turns out--and this is part of the fresh discovery--actually don't do it either way--or not exactly. Or, you could say, they do have a sort of XX, XY pattern similar to the rest of mammalia, except that it's spread over ten--yes ten--chromosomes.

In detail: platypus nuclei, according to Gutzner et al's recent work, contain no less than ten types of sex chromosomes. Five of the sex chromosomes only appear in the males. The other five appear in both males and females, but female somatic cells have two copies of each, and the males just have one copy of each.

Or, to use nomenclature parallel to what works for the rest of mammalia: a female platypus has the sex chromosomes X1, X1, X2, X2, X3, X3, X4, X4, X5, X5, while the males have X1, Y1, X2, Y2, X3, Y3, X4, Y4, X5, Y5. The female gets one of each of her types of X chromosomes from one parent, and one of each from the other, while the male gets all his X chromosomes from the female parent, and all the Y chromosomes from the male parent.

As you might expect, the meiotic method ensuring a sperm cell gets either just the five X chromosomes, or just the five Ys is likely to make meiosis a bit more complicated in monotreme males. The evidence suggests that the odd chain-like structures previously noticed in platypus male reproductive cells are the key to the trick--homologous structures at the ends of each of the relevant sex chromosomes ensure that the Xs and Ys join up into convenient accordion structures of alternating Xs and Ys, prior to being sorted and packed as groups of Xs or Ys into the sperm.

All of this is what's odd about platypuses specifically--though structures like this have been noticed in a few other species (though no other vertebrates). What's interesting from the missing link point of view is this: the platypus sex chromosomes have apparent similarities both to the mammalian X chromosome and to the avian Z chromosome.

That's interesting because it suggests that the mammalian and avian sex chromosomes are related. The hypothetical lineage might look a bit like this if the data lead where they seem to: the platypus X and Y chromosomes are descended from the same ancestral chromosomes as are mammalian X and Y chromosomes, and those chromosomes have a common ancestor chromosome pair (probably) with the avian Z and W chromosomes--and that ancestral chromosome pair (proto-XY, proto-ZW) probably acted as the sex chromosomes in a common ancestor to therapsids and dinosaurs--and thus of birds. Given our current ideas about who diverged when, this ancestor chromosome pair would have to have been around in the Permian, since this is when the synapsids (ancestors to the mammals) and diapsids (ancestors to the dinosaurs) diverged. I'd go further and say that X and Z specifically have a common ancestor (implying that so do Y and W), but I'm not sure you can go that far on this much.

This contradicts certain other evidence and previous theories--see Fridolfsson et al, for a paper presenting evidence for independent evolution of sex chromosomes in mammals and birds. This school of thought held that both mammalian X,Y sex chromosomes and avian W,Z sex chromosomes evolved from autosome pairs (non-sex-specific chromosome pairs, in which each member of the pair is the same size, and contains the same loci--though not necessarily the same alleles), but held that the differentiation of the avian sex chromosomes from their autosomal precursors and the differentiation of the mammalian sex chromosomes were separate events.

The specific similarities the new research describes: The platypus X chromosomes all bear similarities to one another and to other mammalian X chromosomes, with the X1 bearing the greatest similarity to the X chromosomes found in the rest of mammalia. The platypus X5 chromosome, however, also contains a close relative of the gene DMRT1, found on the Z chromosome in birds. (DMRT1 is a possible sex-determining gene in birds; the theory being that two copies mean you're male; one mean you aren't; it isn't believed it could possibly be doing this in platypuses, however, since it is present in two copies in females in this species, and only one in males).

As mentioned at the top: Nature has two articles this week (the December 16 issue)--a commentary by Laura Carrell ("Chromosome chain makes a link") and the letter from Gutzner et al. See also (as linked above) the earlier article in The Scientist.

Tuesday, December 14, 2004

The revolution will be digitized

Folks who know me know I'm not much for predictions taking the form 'x is a revolutionary event'. Side-effect, I suspect, of growing up in the PR-saturated world in which three blades instead of two on the disposable plastic razor is heralded in breathless terms by the anonymous commercial voiceover guy as a breakthrough of epochal importance. Yeah, yeah, revolutionary, whatever.

That said, okay, this probably does actually qualify as quite the thing, and I'm not just saying so 'cos the guy quoted might be a distant relative:
Google has reached agreements with five of the world’s most celebrated libraries, including Oxford University’s Bodleian, to digitise more than 15 million books and make them freely accessible on the internet, a project that will take six years...
...Ronald Milne, the Bodleian’s acting head librarian, said that in terms of disseminating information, the development was as seminal an event as the invention of the printing press.
“It’s a revolution,” Mr Milne said. Referring to the Bodleian’s own copy of Johann Gutenberg’s Bible, the first real book to be produced using the technique of printing from moveable type which Gutenberg invented in the 1450s, Mr Milne added: “In terms of what Gutenberg’s invention was all about, enabling books to be disseminated cheaply, it is very much comparable to that.”

-- World's leading libraries agree to put books online, The Times Online

A cool thing, to be sure. Though I'm beginning to fear with half-decent libraries available in the browser, I might lose even the 20 minutes remaining of each day I'm not online reading.

Monday, December 13, 2004

Thunderbird follies

So I installed Backport's Thunderbird .deb (the Mozilla folk's mail client) on one of my work machines (I run testing, mostly, and Thunderbird isn't in there). I hadda replace Opera's M2, which, though otherwise wonderful, doesn't have (annoyingly) a 'Forward as attachment' function, which is something I need for reporting spam to SpamCop, so I thought I'd give the new kid on the block a chance.

As seems usual for Mozilla backports stuff, the default theme was wonky, causing nasty crashes in which the error 'BadPixMap' figured prominently. And unlike the Firefox setup, the Thunderbird one wouldn't even start, since, apparently, the offending icon appears by default (the Firefox one only went down if you clicked on certain menus, since the bad icons only got drawn when you did this).

The fix was easy enough. Debian folk, take note, as I can't seem to find this issue documented anywhere:
  1. Get a new theme from the themes page, save it to home.
  2. Become root.
  3. Rename /usr/lib/mozilla-thunderbird/chrome/classic.jar to classic.jar.old or somesuch.
  4. Copy (yes, copy, not rename--you'll need it again) the new theme .jar you saved in home into /usr/lib/mozilla-thunderbird/chrome/classic.jar (you're just covering the .jar with the bad pixmaps with another).
  5. Start Thunderbird. Note that you probably won't get a lot of icons, and it will probably look terrible and be a nuisance to use, since the new theme .jar you just dumped in over classic probably won't map properly to what the system's expecting. You don't care; the object is just to get the damn thing running so you can use the program's built-in theme smarts to replace the bad one properly.
  6. Click cancel when it prompts you to set up an account--this is not something you want to try to work through while you've got no icons.
  7. Select Tools>Options (you will probably need to use Alt-O to get to Options once the Tool menu's down, since you probably can't see the damn menu with the icons broken).
  8. Click themes.
  9. Click 'install new theme', and select the new theme .jar you saved to home.
  10. The name of the new theme should appear in the pane, but you can't click it yet (probably another bug). Just close the Themes window
  11. Select Select Tools>Options again, and select themes.
  12. Select the new theme, and then click OK.
  13. Thunderbird should now tell you you'll see your new theme when you restart. Click ok, and close Thunderbird.
That's it. When you restart Thunderbird, it should look much prettier, and you can set up accounts, do whatever you need to do much less painfully.

A word of caution--there are other themes with issues similar to those Classic presents; I went with RAF's Modern for Thunderbird; it seems to be complete. I've also yet to determine howinhell to get rid of the wonky classic theme, which is probably gonna cause trouble for anyone setting up for multiuser machines; anyone else setting up a new Thunderbird account on such machines is gonna have to go through installing and picking the new theme too. Bit of a pain. But such is life with Backports. The good news: 'Forward as attachment' is present and seems to do what I need it to do.