Ah, those pesky sponges again. Although their lives are rather low on action, these strange animals have found lots of other ways to fascinate me. Borrowed skeletons, mysterious lost Hox genes, wonderful alien shapes and perhaps the oldest fossils that might be animals – sponges have no trouble supplying us with stories.
Sponges are also something of a headscratcher for evolutionary biologists. These days it’s generally agreed that they can be divided into four major groups, the glass sponges (hexactinellids), the demosponges, the calcareous sponges and the homoscleromorphs. My impression is that the close relationship of the first two is also well-established. However, to this day biologists haven’t quite managed to agree whether the three are one lineage to the exclusion of other animals, or two or three separate lineages with some of them being closer to non-sponge animals. The trees below illustrate two possibilities:
This is kind of important if you want to know what kind of organism gave rise to the diversity of animals. If sponges are paraphyletic (some sponges closer to non-sponges, as in the right-hand tree), then the mother of all animals was likely sponge-like, sitting on the seafloor and driving water through its porous body to capture food. In such scenarios, two or three of the deepest branchings of the animal tree run into sponges on one side. The simplest explanation for that is that the ancestors at these branching points were themselves sponge-like.
If, however, sponges are monophyletic (their last common ancestor only gave rise to sponges, as in the left-hand tree), then the last common ancestor of animals immediately branched into sponges and non-sponges, whose living descendants are very different from sponges. Suddenly, guessing what Mummy Metazoa might have been like becomes much harder.
The deepest phylogeny of animals is difficult. We are talking about lineages that diverged over 600 million years ago even by conservative estimates (Peterson et al., 2004), and it’s also likely that their early divergences followed each other in rapid succession. That combination is depressingly good at eroding the useful information in gene and protein sequences (Rokas and Carroll, 2006). So what can a phylogeneticist do?
Picking your genes carefully is one solution – use as many as you can to maximise the information you can glean from them, but don’t use genes that evolve so fast their “information” is basically all noise. Another option is to use so-called rare genomic changes. These are things like gaining and losing bits of genes or insertions of parasitic DNA.
Their advantage is that they are unlikely to occur twice in the same way. If two animals have the same viral sequence between genes A and B, it’s far more likely to indicate relatedness as opposed to chance similarity than having the same letter at position 138 in the sequence of gene A. The principle of parsimony (choose the simplest explanation) is a shitty way of interpreting sequence similarity because there’s a high chance of any given change occurring more than once. It works much better for such unlikely events as gaining a virus in the same spot.
microRNAs look like a pretty good source of rare genomic changes. They are small RNAs encoded in the genome, and they play crucial roles in gene regulation in most animals. Their sequences evolve extraordinarily slowly, so it’s relatively easy to identify them across species despite their tiny size. There’s loads and loads of them – miRbase, the microRNA database, lists 1600 different miRNA genes for humans, yielding over 2000 mature miRNAs after processing. The miRNAs in our genome include everything from ancient types with origins in the mists of the Precambrian to young sequences confined to our close relatives. On top of all that, they are thought to be very difficult to lose. All in all, perfect phylogenetic markers.
Perfect for some cases, that is. Their presences and absences may paint a coherent evolutionary picture for most animals, but don’t ask them about sponges. Robinson et al. (2013) tried…
In their introduction stands this depressing summary of current animal miRNA lore, based on many non-sponge genomes plus that of the demosponge Amphimedon queenslandica:
“None of the thousands of miRNAs thus far discovered in eumetazoans are present in the genome [of] A. queenslandica and none of the eight silicisponge‐specific miRNAs have been described in any eumetazoan (or any other eukaryotic group for that matter).”
(Eumetazoans = all animals except sponges and the Blob; silicisponges = glass sponges + demosponges)
However, that’s only two of the four sponge groups. What about the other two? Are they any more helpful? Might they have silicisponge-like repertoires, supporting sponge monophyly? Or might they be hiding some “eumetazoan” miRNAs, arguing for one history or another involving sponge paraphyly? This is what the authors wanted to find out.
They looked for miRNAs by collecting and sequencing small RNAs from calcareous and homoscleromorph sponges. Two species of the former and one of the latter also have genome projects going, which allowed the researchers to verify the RNAs they found as bona fide miRNAs (miRNA genes have a particular structure that doesn’t all show in the mature RNA product) as well as look for the protein components of the editing machinery miRNAs need to reach maturity.
Well, the calcarean genomes certainly contained genes for miRNA-processing enzymes, which is a good sign that they also have miRNAs somewhere. So what do those look like?
Overall, the results of Robinson et al.‘s search are a bit disappointing. They used strict criteria to identify miRNAs, since there are plenty of other kinds of small RNA molecules floating around doing stuff in animal cells. According to these criteria, only one miRNA was confidently identified in the calcareous sponges. This was present in both Sycon and Leucosolenia, but the niggardly bastards didn’t share it with either silicisponges or other animals. Leucosolenia may have a second one. A bunch of eumetazoan-like sequences also showed up, but these were probably contamination from actual eumetazoans, since the Sycon genome, which was obtained from squeaky clean lab-grown sponges, had none of them.
Oscarella only yielded two possible miRNAs, neither of which was known from anything else (including the other homoscleromorphs in this study!) Worse, they couldn’t even find the two processing enzymes in the Oscarella genome – they only recovered a small fragment of one. Maybe the genome sequence is just incomplete, which wouldn’t be very surprising. Then again, maybe Oscarella genuinely doesn’t have a functioning miRNA system, and that could be quite interesting.
Either way, now we know something about microRNAs in all the great sponge lineages. It doesn’t look like they’re going to help us sort out deep animal phylogeny, but maybe the very absence of similarities is telling us something. The reason many microRNAs are so conserved in other animals is that they play important roles in fundamental developmental processes, such as specifying cell types. If sponges aren’t really fussed about keeping them, then maybe their development just doesn’t depend heavily on miRNAs. So what do they do with theirs? Why the difference? Questions, questions…
(Incidentally, here’s yet another reason not to just look at one species and make sweeping claims about “sponges”. Here is also a reason to thank the gods for next generation sequencing. Without the ability to quickly and cheaply [for certain values of “cheap”] sequence tons of DNA and RNA from any creature you fancy, half of the story of animal evolution would be hidden in undeciphered strings of DNA in animals too few people care about for a sequencing project. Yay for technology, yay for diversity!)
[P.S.: there’s so much I’ve wanted to write about and didn’t recently. Since I came back from my Christmas break, I’ve spent most of my time buried in work-related literature. Reading more literature was the last thing I wanted to do with my free time. I’m almost done with that, though. No promises, but I’m almost done ;)]
Peterson KJ et al. (2004) Estimating metazoan divergence times with a molecular clock. PNAS 101:6536-65451
Robinson JM et al. (2013) The identification of microRNAs in calcisponges: independent evolution of microRNAs in basal metazoans. Journal of Experimental Zoology B, advance online publication available 24/01/2013, doi: 10.1002/jez.b.22485
Rokas A, Carroll SB (2006) Bushes in the tree of life. PLoS Biology 4:e352