Velvet worms in a prettier light

I bumped into Mayer et al. (2010) while hunting for reagents to use in an experiment I’m planning. The article is about segmentation (sort of), so I had to have a closer look, and man. Those pictures. Fluorescence and a good microscope can do wonders. This is what a velvet worm looks like in normal light (whole body shot of an unspecified peripatid by Geoff Gallice, Wikipedia, and portrait of Euperipatoides rowelli by András Keszei via EOL):

I think they are adorable and cuddly the way they are (apart from that hunting with slime bit), but they look simply gorgeous if you stick some glowing antibodies to them and start playing with a confocal ‘scope.

This is a fairly late-stage embryo of E. rowelli, the same guy waving its chubby leggies at you in the right-hand photo above. The green dots are cells that were copying their DNA when the baby was killed (all of the pictures below are from Mayer et al., of course):

These are younger embryos of the same species, with all their DNA labelled in blue and dividing cells labelled in red:

And these are embryos of another species from the same family as the unidentified guy from Wikipedia (colours are the same as above):

Seriously, there is something about the mystical glow of these images that always gets me. I think you could make almost anything look beautiful with a fluorescent marker and the right equipment. I know aesthetic appeal isn’t the primary aim of scientific imaging, but damn. Look at those alien creatures glowing with the light of the unknown.

In case you wondered, the point of the paper is that velvet worms lack a posterior growth zone. That means that when they develop their numerous segments, there isn’t a well-defined pool of cells at the rear of the embryo that divide to generate segment material. As you can see in all the red glow, cell division happens evenly all over the place. Why is this significant? Well, posterior growth zones were thought to be one of the characteristics that segmented animals might have inherited from their common ancestor. But Mayer and colleagues point out that the existence of a PGZ in the arthropod ancestor is dubious at best, and velvet worms (one of the closest living relatives of arthropods) also lack one, so maybe it’s kind of wrong to use the PGZ as an argument for the common ancestry of segmentation.

(There, that’s the science in a nutshell. Now I’ll just go back and admire the pretty glowy pictures some more :D)
***

Reference:

Mayer G et al. (2010) Growth patterns in Onychophora (velvet worms): lack of a localised posterior proliferation zone. BMC Evolutionary Biology 10:339

News bites

Just quickly before I completely forget about these…

(1) Common ancestry of segmentation: back-and-forth-and-back-and-forth

Seaver EC et al.(2012) Expression of the pair-rule gene homologs runt, Pax3/7, even-skipped-1 and even-skipped-2 during larval and juvenile development of the polychaete annelid Capitella teleta does not support a role in segmentation. EvoDevo 3:8

I’ve made throwaway mentions of segmentation before. The conundrum about segmentation is whether (or rather, to what extent) it is homologous in the three “eusegmented” phyla, arthropods, annelids and chordates. It arises because all three phyla are separated from the others by many lineages that aren’t usually considered segmented – yet the three share some tantalising similarities. People have been trying to solve the question by comparing the genetic mechanisms generating the segments in each group, with mixed results. One of the papers in my previous news bite post was about the similarity of segmentation in arthropods and vertebrates. Now, here’s one for the differences between arthropods and the wormies. (You can’t say I’m not fair :-P) The genes listed in the study’s title were originally described in the fruit fly Drosophila melanogaster, one of the best studied animals in developmental biology (and, like, every other area of biology). There, they have an interesting role in that each of them helps define every other body segment. IIRC, Pax3/7 (known in flies as paired) and even-skipped are for even-numbered segments, runt is for the odd ones. Now, segmentation in Drosophila is (to put it mildly) fucking weird, but if memory serves, several of these pair-rule genes have been confirmed to play similar roles in less eccentric arthropods. Elaine Seaver and colleagues looked at their expression in their favourite worm (this guy. Seaver’s group obviously didn’t pick it for its beauty :-P), and they found that they were active in… nothing resembling a two-segment pattern. Or anything segment-related. The more genetic studies come out, the more complicated the whole segmentation issue is looking…

(2) Someone found the cause of the Cambrian explosion. (Again.)

Peters SE & Gaines RR (2012) Formation of the ‘Great Unconformity’ as a trigger for the Cambrian explosion. Nature 484:363-366

The Cambrian explosion is probably not what you think it is (no, all animal phyla didn’t just suddenly pop into existence fully formed ;)). Nevertheless, the (relatively) quick rise of animals – particularly animals with hard parts – beginning in the Early Cambrian is still odd enough to fascinate generations of palaeontologists, evolutionary biologists and geologists. The list of proposed causes is pretty long by this point (and believe me, I really really would like to go into them once… but, uh. Huge, dauntingly huge topic). Explanations range from denying the need for an explanation through pinning it on oxygen levels, ice ages, developmental genetics, predation, biomineralisation, even the evolution of eyes, and, working from memory, I probably left some more out of that list. Peters and Gaines’ preferred explanation seems to be geological and ecological: they suggest that a combination of lots of erosion/weathering on land, and a subsequent rise in sea levels, led to large new shallow seas that were chock full of dissolved minerals. New habitats to conquer + widely available minerals = an explosion of new animals with mineralised hard parts. This study is a nice two-in-one: it purports to explain not only the Cambrian explosion, but also the conspicuous gap in the geological record that separates Cambrian from Precambrian rocks in many places.

News bites

Since work is quite frantic lately, and my attention span has gone on holiday, I’ve decided to do something I haven’t done before and just say a few words about papers that caught my interest today without actually reading them. Each of these is probably worth a full-blown meandering of its own, but I know I wouldn’t ever get to them at this rate. Better read their abstracts and give some quick thoughts than let them sink unnoticed into the murk of my “papers” folder!

(1) How many genomes do we (not) have?

Reference: Ali Abbasi A & Hanif H (2012) Phylogenetic history of paralogous gene quartets on  human chromosomes 1, 2, 8 and 20 provides no evidence in favor of the vertebrate octoploidy hypothesis. Molecular Phylogenetics and Evolution, in press (doi: 10.1016/j.ympev.2012.02.028)

(How many papers have authors with alliterating names? :D)

In the circles I move in, it’s pretty much canon that the ancestors of living vertebrates doubled their entire genomes twice. It’s still debated exactly when these duplications occurred, but few people doubt that they did. This so-called 2R hypothesis is supported by things like our possession of several (quite often, four) copies of genes that are singletons in our closest living relatives (read: lancelets*), and more importantly, that whole big chunks of lancelet chromosomes can be matched to chunks of four different vertebrate (mainly, human) chromosomes. Genes that are close to one another in lancelets are often also close together in vertebrates.

The relationship is not perfect – in well over 500 million years of evolution, genes inevitably get lost and bits of chromosome scrambled. And, thus, there is always room to question the 2R scenario, which is what this paper clearly does. They propose that those four-gene families originated at all sorts of different times, from small local duplications and rearrangements. If they are right, this is a very important result. It basically uproots every bit of speculation ever proposed on how the genome duplications contributed to the evolution of vertebrates, which, far as I can tell, is a hell of a lot of speculation. Not having read the whole paper, I would still put my money on 2R, but who knows what the future holds? Maybe we are facing a minor paradigm shift?

(2) The segmentation clock also ticks in insects!

Reference: Sarrazin AF et al. (2012) A segmentation clock with two-segment periodicity in insects. Science, advance online publication (doi: 10.1126/science.1218256)

The evolutionary history of segmentation is one of my random interests, and from my point of view, the above is a good reason to squee in a most fangirlish way. Segmentation is the construction of a body from repeating units. In its purest form, which isn’t that common in modern animals, the animal is essentially made up of identical repeated blocks containing a copy of each key organ like kidneys, nerve centres, limbs and muscles. (Even in the most perfectly segmented creatures, head and tail ends form something of an exception. Ragworms make a nice example.) More commonly, only some components are repeated, and they are repeated with slight differences along the body. Vertebrates’ spine and associated muscles are a good example, and so are the defining traits of arthropods, their jointed exoskeletons equipped with repeated pairs of appendages.

Although traditionally it has been thought that arthropod and vertebrate segmentation have independent origins, parts of the genetic machinery are shared between both groups (as well as segmented worms). Various “segmentation genes” are active in distinct stripes in our embryos, marking out future segments even before we can see the segments themselves. In vertebrates, cells periodically switch “segmentation genes” on and off, and this combined with the growth of the embryo produces a dynamic stripey pattern of gene expression. While segments and stripes of gene expression are darn obvious in arthropods, this is the first time anyone has confirmed that some arthropod segmentation genes actually oscillate like their vertebrate counterparts do, as opposed to, say, the cells expressing them moving about. Whether this is a spectacular example of convergent evolution or evidence of a shared ancestral heritage, I couldn’t say, but it’s really cool either way.

(3) Old genes are entrenched, new genes are redundant after all?

Reference: Chen WH et al.(2012) Younger genes are less likely to be essential than older genes, and duplicates are less likely to be essential than singletons of the same age. Molecular Biology and Evolution advance online publication (doi: 10.1093/molbev/mss014)

So, this claims to resolve a conundrum I wasn’t even aware of before. Gene duplication is thought to be important for the evolution of new functions because two copies of a gene mean there is a backup if one of them fails at its original function. Hence, theory goes, duplicate genes are much less restricted in the evolutionary paths they can take. Apparently, studies in mice have contradicted this common wisdom by claiming that duplicate genes are just as likely to be indispensable as genes without backup copies. However, Chen et al. are saying that this is wrong, confounded by gene age. Since new genes are less likely to be essential than old genes (which had more time to evolve interactions with the rest of the genome), and mouse duplicates are on average older than mouse singletons, the two effects end up cancelling out. When they factor in gene age, duplicates are indeed less essential than loners. One of the central tenets of current thinking about (genetic) novelty stays in the ring for another round…

(4) Is reducing complexity easier than increasing it?

Reference: Harjunmaa E et al. (2012) On the difficulty of increasing dental complexity. Nature advance online publication (doi: 10.1038/nature10876)

How complexity increases in evolution is more than a breeding ground for creationist incredulity, it’s also quite interesting for bona fide evolutionary biologists. Looking at the development of mouse teeth, Harjunmaa et al. notice that increases and decreases in the complexity of tooth shapes require different sorts of mess-ups. Simpler-than-normal teeth are common in mutants and easy to make in experiments. More complex teeth – i.e. those with more cusps – are rarely if ever seen in natural mutants. Turns out they are perfectly possible – you just need to manipulate several genetic pathways at the same time to produce a clear result.

Can this be generalised? Is greater complexity usually harder to achieve? When does this apply and when does it not? I’ve recently read papers that explore how complexity increases easily and completely by chance (I have a half-written post about them languishing on my hard drive, FWIW). Are the rules different for different levels of organisation? The aforementioned complexity-by-chance papers analyse the molecular level: one is about the architecture of gene switches, the other about a protein machine. Teeth are pretty large pieces of life with thousands upon thousands of such machines participating in their production. Does that make a real difference, or is what I’m seeing just coincidence? Dunno, but it’s fascinating to think about!

***

Heh, it looks like I took rather bigger “bites” of these news than I planned to. I kind of managed to write the equivalent of a full-blown meandering anyway. The only difference is that I didn’t painstakingly reference this one. I hope that doesn’t mean that half of what I wrote off the top of my head is wrong 😀

***

Note:

*Lancelets are now not considered our closest relatives. Unbelievable as that may seem, that honour goes to sea squirts and their ilk. However, the sea squirt bunch are ridiculously weird in all sorts of respects, and their genomes are jumbled beyond recognition. So… not so great if you want to learn anything about our ancestors.