Amphioxus and the evolution of the chordate genome

Blogging on Peer-Reviewed Research

This is an amphioxus, a cephalochordate or lancelet. It's been stained to increase contrast; in life, they are pale, almost transparent.

amphioxus.jpg

It looks rather fish-like, or rather, much like a larval fish, with it's repeated blocks of muscle arranged along a stream-lined form, and a notochord, or elastic rod that forms a central axis for efficient lateral motion of the tail…and it has a true tail that extends beyond the anus. Look closely at the front end, though: this is no vertebrate.

amphioxus_closeup.jpg

It's not much of a head. The notochord extends all the way to the front of the animal (in us vertebrates, it only reaches up as far as the base of the hindbrain); there's no obvious brain, only the continuation of the spinal cord; there isn't even a face, just an open hole fringed with tentacles. This animal collects small microorganisms in coastal waters, gulping them down and passing them back to the gill slits, which aren't actually part of gills, but are components of a branchial net that allows water to filter through while trapping food particles. It's a good living — they lounge about in large numbers on tropical beaches, sucking down liquids and any passing food, much like American tourists.

These animals have fascinated biologists for well over a century. They seem so primitive, with a mixture of features that are clearly similar to those of modern vertebrates, yet at the same time lacking significant elements. Could they be relics of the ancestral chordate condition? A new paper is out that discusses in detail the structure of the amphioxus genome, which reveals unifying elements that tell us much about the last common ancestor of all chordates.

But first, a little question that needs to be resolved. This is an ascidian, a member of the subphylum urochordata.

ascidian-amphioxus.jpg

Like us and like amphioxus, it also has a central notochord, and it also has segmental blocks of muscle and a dorsal nerve cord, and in this case you can also see that it has an expansion of the front of the nerve cord into something that looks rather brain-like. Ascidians are also more diverse and successful than the cephalochordates.

A major difference, though, is that amphioxus maintains its chordate morphology throughout its life — the picture at the top of this page is an adult. Ascidians, on the other hand, only look like tadpoles in their larval stages, which is only a dispersal form. They swim away from their parent and settle down on a solid surface and throw away their tail and brain and squat permanently as a sessile filter feeder called a sea squirt. This opens a debate about chordate ancestry. Did we 1) evolve from an ascidian-like animal, that maintained it's larval stage throughout life and refined its form to become something like an amphioxus, which then evolved into the vertebrates? Or was 2) an amphioxus-like animal at the base of the family tree, and the ascidians are the weird cousins who went off and evolved a specialized adult form?

I've discussed ascidian evo-devo before, and the consensus so far is that the second model is most likely, and ascidians are a highly derived descendant of the last common ancestor of all chordates, which have gone off in some radical and very interesting directions. The current paper on the amphioxus genome confirms that, and in particular reveals an interesting fact: while the ascidian genome has undergone some major remodeling that makes it very different from our own, the amphioxus genome seems to have conserved many ancestral features and may very well be a good proxy for examining the genetics of the last common chordate ancestor.

Before we move on to the genetics, though, look at the diagram above. At the bottom are two embryos at the 8-cell stage, with the prospective fates of some regions marked: cells that form in the area with the horizontal hatching will become notochord, and those with the vertical hatching will become the neural tube. The two are very similar; the one on the left is an ascidian embryo, and the one on the right is an amphioxus embryo. Both exhibit a pattern of mosaic development, in which the fates of regions of the egg are mapped out by a pre-pattern of molecular determinants. Vertebrates do not do this, or do so to a much lesser degree; we have regulative development, in which early cells are much more plastic, and gradually and progressively determine their role in the embryo. This suggests, though, that mosaicism may be an ancestral property of the chordate lineage, and we are the weirdos who have departed far from the traditional pattern in this regard.

Let's also have a brief video interlude. Below is a time-lapse movie of amphioxus development. You'll see the 8-cell stage illustrated above, and then it flies through a series of transformations into a hollow ball, folding in a portion of the structure to make a cup-shape organism, then elongating to make a kind of canoe-shaped beast, which then differentiates the various muscles and tissues to make an embryonic amphioxus. Isn't it lovely?

The recent paper in Nature by Putnam and others describes the completion of the draft genome sequence for the amphioxus species, Branchiostoma floridae. The genome is about 520 megabases long (a sixth the size of ours) and contains about 22,000 protein coding regions, or genes, and so is about the same size as ours. All this data means that a better phylogenetic tree can be assembled, which is shown below and which supports the conclusion that amphioxus diverged early in chordate history, followed by a later split between the urochordates and the vertebrates.

chordate_clade.jpg
(click for larger image)

Deuterostome phylogeny. Bayesian phylogenetic tree of deuterostome relationships with branch length proportional to the number of expected substitutions per amino acid position, using a concatenated alignment of 1,090 genes. The scale bar represents 0.05 expected substitutions per site in the aligned regions. Long branches for sea squirt and larvacean indicate high levels of amino acid substitution. This tree topology was observed in 100% of sampled trees. Numbers in red indicate bootstrap support under maximum likelihood. Unlabelled nodes were constrained.

Look at the magnitude of the substitutions in the urochordates! If you want to argue that any chordate group is "more evolved" than any other, you'd have to hand the title to those somewhat obscure organisms — they've been shedding genes and reorganizing and adapting in some amazing ways.

Now for the tricky part, the analysis of synteny. If you're not familiar with the concept, you might want to review yesterday's basic summary of synteny, but it is basically a search for conserved clusters of genes. In this work, they used the logic of synteny to work out the existence of 17 ancestral Chordate Linkage Groups (CLGs). A CLG may roughly compare to a single chromosome in the chordate last common ancestor — it's a set of genes that are associated with one another in multiple species, and are thought to represent the retention of an ancient organization.

Here's the wonderfully complicated summary diagram.

amph_hum_synteny.jpg
(click for larger image)

Quadruple conserved synteny. Partitioning of the human chromosomes into segments with defined patterns of conserved synteny to amphioxus (B. floridae) scaffolds. Numbers 1–17 at the top represent the 17 reconstructed ancestral chordate linkage groups, and letters a–d represent the four products resulting from two rounds of genome duplication. Coloured bars are segments of the human genome, shown grouped by ancestral linkage group (above), and in context of the human chromosomes (below).

First, along the top they have reconstructed the 17 ancestral CLGs, and they've color-coded them: for instance, the genetic contents of CLG #8 (or ancestral chromosome #8, if you'd rather think that way) has been deduced from a comparison of human and amphioxus genomic data, and all those genes have been colored yellow. Below the ancestral CLGs is a diagram of the human chromosome set, from chromosome 1 to 22, with the X chromosome, and homologs to each of the genes in CLG #8 have also been colored yellow. You can see that ancestral chromosome 8 has, over the 500+ million years since, broken up and scattered into human chromosomes 2, 4, 5, 11, 13, and X.

I have to qualify this image a bit, though: don't get the impression that big chunks of the ancestral chromosomes have survived perfectly intact for half a billion years! What's illustrated here is a pattern of macro-synteny — what it says is that within a swathe of a particular color, we can find many, but not all genes that can be mapped to a common linkage group. If we did a more fine-grained synteny diagram, where we only mapped continuous blocks of 3 or more genes that were the same, we'd have a much messier, more salt-and-pepper sort of picture.

Within a region of a single color, there has been much scrambling of the local gene order. There have also been additions of individual new genes in either the amphioxus or human genome, and losses of genes. Also, not all genes are included in this summary: approximately 60% of the human genes that have amphioxus orthologs are found in these linkage groups. What each bit of color means is that this is a region that is enriched for a common set of shared genes between amphioxus and human beings, but it has also been leavened with variation and scrambled about internally. What has been preserved is not a literal duplicate of an ancient chromosome, but a segment that retains enough signal above the noise of ages of slow rearrangement that we can detect its affinities.

Still, pretty cool, I think. Evolutionary change hasn't completely swamped out our origins, only made them fainter and trickier to discern.

There is one other significant detail to this illustration. If you look at the ancestral CLGs, you see that each has more than one colored bar descending from them — there's four for each. What's going on here? For each gene (more or less) identified in amphioxus, there tend to be multiple genes in the human. When aligning them, you'll get multiple roughly equivalent sets of genes lining up with each amphioxus group. What this reveals is that our history is marked by two rounds of whole genome duplications after we branched away from amphioxus. This is also very cool; it might explain some of the greater diversity we see in later chordates, that they have more potential for new combinations and variations of genes.

You might recall that I mentioned that humans and amphioxus have approximately the same number of genes, about 20,000, and be wondering how that jibes with the evidence that there were 2 rounds of duplications. Shouldn't we have four times as many genes? The answer is no. When a gene is duplicated, more often than not is that one of the copies will accumulate errors and drift away into nonexistence.

Imagine that we have a stretch of chromosome that has these genes on it:

A-B-C-D

A duplication occurs; now we have two regions:

A-B-C-D and A'-B'-C'-D'

Errors take out some of the copies, so we end up with this:

A-.-C-. and .-B'-.-D'

Now, you see, we've got a result where the organism hasn't gained or lost anything overall — it still has an A, B, C, and D gene — we haven't had a doubling of the total gene numbers, we still have four genes. If we look at the synteny, we'd see one stretch of chromosome, A-.-C-., that lines up with the ancestral A-B-C-D, and we'd also see a second region, .-B'-.-D', that also lines up with it — there'd be two matches. Even though we haven't seen a significant net increase in gene numbers, we still have a doubling of the number of syntenic regions.

That's what this paper is reporting: no major increases in the number of genes, but we are seeing the vestiges of that ancient doubling. Furthermore, the losses aren't entirely random. There has been a preferential retention of duplicate genes involved in signal transduction, transcriptional regulation, neuronal activity, and developmental processes, exactly what we'd expect to see if new avenues in developmental and behavioral complexity were being explored by our ancestors. So what probably happened is that there was a fairly abrupt duplication event in our history that was largely neutral in its initial effect, and was gradually pared back down, but that some of the duplications generated new capabilities that were promoted by natural selection.

Before our ancestors were even fish, they were Cambrian and pre-Cambrian filter-feeding torpedos, muscular tubes with springy tails that darted through clouds of bacteria and plankton in the seas, harvesting microorganisms in a branchial net, and that probably resembled modern amphioxus more than anything else. That was our primitive niche, motile grazers on suspended particles. Where we vertebrates came from was an accidental set of duplications that created new combinations of genes and new phenotypic variants, some of which had adaptive advantages.

Don't get too cocky and think that surges in complexity are always a good thing, and lucky for us that we are the outcome. Evolution is blind and sometimes the fortunate result can go the other way: look to the ascidians, which, rather than expanding their genetic library, have been successful by streamlining their genomes, chucking out unnecessary elaborations and juggling a smaller suite of genes around in more dramatic ways. There are many directions evolution can take, and there's no a priori reason to think our particular path is the most powerful.


Putnam NH, Butts T, Ferrier DE, Furlong RF, Hellsten U, Kawashima T, Robinson-Rechavi M, Shoguchi E, Terry A, Yu JK, Benito-Gutiérrez EL, Dubchak I, Garcia-Fernàndez J, Gibson-Brown JJ, Grigoriev IV, Horton AC, de Jong PJ, Jurka J, Kapitonov VV, Kohara Y, Kuroki Y, Lindquist E, Lucas S, Osoegawa K, Pennacchio LA, Salamov AA, Satou Y, Sauka-Spengler T, Schmutz J, Shin-I T, Toyoda A, Bronner-Fraser M, Fujiyama A, Holland LZ, Holland PW, Satoh N, Rokhsar DS (2008) The amphioxus genome and the evolution of the chordate karyotype. Nature 453(7198):1064-71.