Metazoan Evolution: Some Animals Are More Equal Than Others Dispatch

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

Metazoan Evolution: Some Animals Dispatch are More Equal than Others

1,2 1 Florian Raible and Detlev Arendt the vertebrate lineage, we have been misled by the rapid rate of molecular evolution, with large gene losses, of the invertebrate model species. Comparison of newly available sequence data Kortschak et al. [4] analysed 1400 EST clusters from facilitates reconstruction of the gene inventory of the a basal metazoan, the coral Acropora millepora —a Urbilateria, the last common ancestors of flies, nema- cnidarian of the Anthozoa — which they compared to todes and humans. The most surprising outcome is the gene inventories of man, Drosophila and C. that human genes seem to be closer to the bilaterian elegans. The advantage of choosing Acropora is that roots than previously assumed. it represents an ‘evolutionary outgroup’ to the Bilateria — that is, the evolutionary line leading to corals branched off the metazoan tree before the Urbilateria It is a truism that the plausibility of an evolutionary infer- came into existence (Figure 1). But corals are still fairly ence increases with the amount of data on which it is complex Metazoa, much closer to the bilaterian roots based, and the ever-quickening provision of full genome than another frequently used evolutionary outgroup sequences is providing a huge amount of grist for the for genome comparisons, the budding yeast Saccha- evolutionary biologist’s mill. Genome data are now romyces cerevisiae, a unicellular fungus. The outgroup available for man, mouse, fish, tunicates, nematodes status implies that any gene that Acropora shares with and flies, and their comparison shows that a large pro- any of the bilaterians necessarily formed part of the portion of genes are shared across all the Bilateria – the urbilaterian gene inventory — even if it is absent from animals with bilateral symmetry. But such comparisons all but one of the extant bilaterian groups. This allows also, of course, reveal phylogenetically relevant differ- unambiguous determination of gene loss events along ences. What were the molecular changes that accom- the divergent bilaterian evolutionary lines. panied the evolution of the major bilaterian branches — The most striking finding of this study [4] is a strong the vertebrates in particular — from their last common asymmetry in the frequency of gene loss across the ancestors, the Urbilateria [1] (Figure 1)? Bilateria. Out of a set of Acropora genes that are present While it is clear that the duplication of existing genes in at least one of the compared bilaterian genomes, 12% has played a major role in vertebrate evolution (for appear to be shared exclusively with humans, while only example, see [2]), the contribution of novel genes to 1% of sequences are shared with just Drosophila or C. the rise of the vertebrate lineage remains ill-defined. elegans (Figure 2). The sequences shared with humans Current estimates from our chordate relative, the tuni- include genes that so far appeared to be vertebrate cate Ciona intestinalis, claim that as many as one sixth innovations, such as tumorhead [7] or churchill [8]. This of its genes could represent evolutionary innovations implies that, from a random sample of urbilaterian shared only with the vertebrates [3]. But a series of genes, more than one in ten has been lost during the recent studies [4–6], one published very recently in evolution of Drosophila and C. elegans, as opposed to Current Biology [4], reveal that many of the supposedly only 1% gene loss in the vertebrate lineage. vertebrate/chordate-specific genes instead have a long Importantly, this asymmetry in gene loss is support- evolutionary history. ed by another recent comparative study [5]. Krylov et Genes found in only one of the sequenced genomes, al. [5] conducted a systematic survey of the presence but not in others, have naturally been considered evolu- or absence of orthologous genes across six fully tionary novelties. This follows the parsimonious view sequenced eukaryote genomes — those of man, that a gene that is found, for example, in the human Drosophila, C. elegans, S. cerevisiae, the fission yeast genome but not in that of the fruitfly Drosophila or Schizosaccharomyces pombe and the pathogen nematode Caenorhabditis elegans, should have arisen Encephalitozoon cuniculi. They find that among these, on the evolutionary line leading to humans (red square humans have the lowest rate of gene loss. Like that of in Figure 1). New studies, however, indicate that this Kortschak et al. [4], this study revealed a loss of view is too simplistic; it seems, rather, that we still around 15% of genes along the evolutionary lineage possess many genes that were lost on the lineages from Urbilateria to C. elegans or Drosophila, which leading to Drosophila and C. elegans, and can be traced compares to 5% gene loss for the human lineage. back to pre-bilaterian times or even stem-line meta- (Note that differences in gene loss rates between the zoans (blue square in Figure 1). A more complete sam- two studies might be related to the different reference pling of transcriptomes across the Metazoa, in the form collections they used.) of ‘expressed sequence tag’ (EST) collections, indicates Gene loss, however, is only the most extreme case that, in assessing potential evolutionary innovations on of gene modification. In line with the strong asymmetry in the frequencies of gene loss, Kortschak et al. [4] also European Molecular Biology Laboratory, 1Developmental found that the human gene sequences generally show Biology Programme, 2Computational Biology Programme, a much higher overall similarity to the corresponding Meyerhofstr. 1, 69012 Heidelberg, Germany. E-mail: sequences of the Acropora outgroup. Sequence simi- [email protected] larity can be measured by the ‘E-value’ in an analysis Current Biology R107

Figure 1. Gene birth and loss in the Occurrence evolution of metazoan genomes. in sequenced A Ð Ð + Ð genomes (A) A pattern frequently found in the sequenced metazoan genomes: a gene is present in the human genome (+) but absent from those of C. elegans, Drosophila and Ciona (–). (B) A simplified Platynereis Aplysia metazoan evolutionary tree, illustrating Drosophila two contrasting scenarios that could H. sapiens account for this occurrence pattern. In the Vertebrate Ciona C. elegans Annelida evolutionary first, the gene emerges on the vertebrate Secondary novelty branch (red square) and persists in the gene loss Mollusca Arthropoda vertebrate lineage (red dashed line). In the Vertebrata Tunicata second, the gene emerges at the base of Nematoda Lopho- Acropora the Metazoa (blue square) and persists in trochozoa many branches (blue lines), but is lost Ecdysozoa from the insect, nematode and ascidian ? Deuterostomia Anthozoa lineages (empty squares). The presence Urbilateria of the gene in EST collections from the Cnidaria cnidarian Acropora [4] or lophotrochozoans Platynereis and Aplysia [6] would decide in favour of the second sce- nario. Note that the branching of nematodes off the bilaterian tree is currently Metazoan evolutionary debated [9,15]. B novelty Current Biology using the ‘BLAST’ algorithm: among the Acropora Lophotrochozoa and vertebrates necessarily existed genes for which homologs exist not only in man, but in Urbilateria. also in Drosophila or C. elegans, 36% have vertebrate These findings imply that the Urbilateria were BLAST ‘hits’ with thousand-fold higher E-values than genetically more complex than previously thought to any fly or nematode hit, and only 7% show the reverse. be the case. But what does this tell us about the Taken together, these new analyses of gene loss complexity of urbilaterian development, anatomy or frequencies and of sequence divergences suggest physiology? Genes do not evolve on hold: whenever a that the human genome — and thus those of the gene appeared on the animal evolutionary tree, it was entire vertebrate lineage — has diverged much less functional. This ancestral function should be close to from the ancestral genome of our urbilaterian ances- the consensus function present in today’s animals that tors than have the Drosophila and C. elegans have retained that gene. The comparative study of genomes (Figure 2). This is consistent with previous gene functions across Bilateria therefore provides a findings that Drosophila and C. elegans have high means to reconstruct an ancient evolutionary state. rates of molecular evolution (for example, see [9,10]), Thornton et al. [6] report an instructive example of but the great impact of this trend on the overall how comparative analysis of vertebrate and lophotro- divergence of flies and nematodes from the bilaterian chozoan sequences helps reconstruct ancestral gene consensus had not been appreciated. functions; it also gives some insight into the evolution of Unlike flies and nematodes, other invertebrates seem hormone systems. The authors found an orthologue of to have retained more of the ancestral urbilaterian gene the vertebrate estrogen receptor gene in the genome of inventory. There is increasing evidence that, as in verte- the Californian sea hare Aplysia californica (Mollusca). brates, gene loss and sequence divergence occurred at This receptor gene was not found in the sequenced a relatively low rate in the Lophotrochozoa, the third genomes of Drosophila and C. elegans, and as it also major branch of the Bilateria [11] (Figure 1). Lophotro- seems not to exist in the tunicate Ciona, it had been chozoan EST collections, including those from the pla- regarded a vertebrate innovation. From sequence align- narian Schmidtea [12] and the polychaete annelid ments, Thornton et al. [6] reconstructed a hypothetical Platynereis dumerilii (our unpublished data) are bringing ancestral steroid receptor, which they showed in func- to light genes that are either more conserved with their tional assays behaves as an estrogen-inducible tran- vertebrate homologues than with those of Drosophila scriptional regulator; the putative ancestral protein thus and C. elegans, or that have even been completely lost closely resembles the estrogen receptor of today’s ver- from the latter genomes. tebrates. This has the interesting implication that estro- Systematic screens have revealed Platynereis gen and its cognate steroid receptor presumably orthologues of members of POU homeobox or Wnt already existed in Urbilateria, probably being involved in gene families that had been considered ‘vertebrate- the reproduction of these animals. It was then kept in specific’ (K. Tessmar-Raible and D. A., unpublished the vertebrate (and lophotrochozoan) lines of evolution, data, and [13]). And the first noggin orthologue found but was lost in insects and nematodes. outside vertebrates has been described for the Comparative analyses of genome sequences, EST planarian Dugesia [14]. Notably, as can be deduced collections, and particular gene functions across the from the branching pattern of the bilaterian evolu- Metazoa are thus providing a clearer picture of urbila- tionary tree in Figure 1, any gene shared between terian biology. This is also enabling us to make a more Dispatch R108

evolutionary lines. Vertebrates, lophotrochozoans and A 44 anthozoans are a good choice for such comparative 36 evolutionary research, because they appear to share a surprisingly large part of the ancestral gene inventory that has been lost in other groups. In a certain sense, therefore, these animals, like some of those on Orwell’s Animal Farm, are more equal than others, and thus should be most revealing about our complex past. 12 References 7 1. De Robertis, E.M., and Sasai, Y. (1996). A common plan for dorsoventral patterning in Bilateria. Nature 380, 37-40. 1 2. Levine, M., and Tjian, R. (2003). Transcription regulation and animal diversity. Nature 424, 147-151. only in more more only in 3. Dehal, P., Satou, Y., Campbell, R.K., Chapman, J., Degnan, B., De related related Tomaso, A., Davidson, B., Di Gregorio, A., Gelpke, M., Goodstein, to to D.M., et al. (2002). The draft genome of Ciona intestinalis: insights into chordate and vertebrate origins. Science 298, 2157-2167. 4. Kortschak, R.D., Samuel, G., Saint, R., and Miller, D.J. (2003). EST C.elegans H. sapiens analysis of the cnidarian Acropora millepora reveals extensive gene loss and rapid sequence divergence in the model invertebrates. Dr osophila Curr. Biol. 13, 2190-2195. 5. Krylov, D.M., Wolf, Y.I., Rogozin, I.B., and Koonin, E.V. (2003). Gene B loss, protein sequence divergence, gene dispensability, expression level, and interactivity are correlated in eukaryotic evolution. Genome Res. 13, 2229-2235. 6. Thornton, J.W., Need, E., and Crews, D. (2003). Resurrecting the ancestral steroid receptor: ancient origin of estrogen signaling. Science 301, 1714-1717. 7. Wu, C.F., Nakamura, H., Chan, A.P., Zhou, Y.H., Cao, T., Kuang, J., Gong, S.G., He, G., and Etkin, L.D. (2001). Tumorhead, a Xenopus gene product that inhibits neural differentiation through regulation of proliferation. Development 128, 3381-3393. 8. Sheng, G., dos Reis, M., and Stern, C.D. (2003). Churchill, a zinc finger transcriptional activator, regulates the transition between gastrulation and neurulation. Cell 115, 603-613. 9. Aguinaldo, A.M., Turbeville, J.M., Linford, L.S., Rivera, M.C., Garey, J.R., Raff, R.A., and Lake, J.A. (1997). Evidence for a clade of nema- Relative todes, arthropods and other moulting animals. Nature 387, 489-493. frequencies 10. Ranz, J.M., Casals, F., and Ruiz, A. (2001). How malleable is the of derived/ eukaryotic genome? Extreme rate of chromosomal rearrangement lost genes in the genus Drosophila. Genome Res. 11, 230-239. 11. Tessmar-Raible, K., and Arendt, D. (2003). Emerging systems: between vertebrates and arthropods, the Lophotrochozoa. Curr. Urbilateria Opin. Genet. Dev. 13, 331-340. 12. Sanchez Alvarado, A., Newmark, P.A., Robb, S.M., and Juste, R. (2002). The Schmidtea mediterranea database as a molecular resource for studying platyhelminthes, stem cells and regeneration. Development 129, 5659-5665. 13. Prud'homme, B., Lartillot, N., Balavoine, G., Adoutte, A., Vervoort, Current Biology M. (2002). Phylogenetic analysis of the Wnt gene family. Insights from lophotrochozoan members. Curr. Biol. 12, 1395-1400. Figure 2. Variable rates of gene loss and sequence evolution in 14. Ogawa, K., Ishihara, S., Saito, Y., Mineta, K., Nakazawa, M., Ikeo, K., different bilaterian lineages. Gojobori, T., Watanabe, K., Agata, K. (2002). Induction of a noggin- like gene by ectopic DV interaction during planarian regeneration. (A) Asymmetry in gene loss and sequence divergence in human, Dev. Biol. 250, 59-70. nematode and fly. Results of a systematic evaluation of bilater- 15. Blair, J.E., Ikeo, K., Gojobori, T., Hedges, S.B. (2002). The evolu- ian BLAST hits obtained with Acropora EST sequences. 36% of tionary position of nematodes. BMC Evol. Biol. 2, 7. genes are significantly more related to human than C. elegans and Drosophila orthologues (light yellow bar), while 12% are only found in humans (yellow bar); in contrast, only 7% are more related to C. elegans or Drosophila orthologues (light red bar), and 1% only found there (red bar). 44% of the sequences show comparable similarity to human, Drosophila and C. elegans orthologues (grey bar). Relative similarity and absence were deduced from differences in BLAST hit E-values: hits for which E-values were 1000-fold lower in one group were ranked to be more related to that group; absence was inferred from E-values above 10–6 in the corresponding group (modified from [4]). (B) As Acropora EST sequences represent an outgroup for bilaterian sequence evolution, these comparisons imply that the human genome has diverged significantly less from the urbilaterian genome than have those of C. elegans and Drosophila, as judged by their rate of gene loss and gene divergence. reliable estimate of the quality and amount of evolutionary change that occurred along the descending