The dispute between Georges Cuvier and Etienne Geoffroy St. Hilaire, two of the most famous scientists of their time, in Paris in 1830 has long been recognised as an important episode in the history of the life sciences. Geoffrey claimed that the body plan (which he called “archetype”) of vertebrates corresponded to an insect or earthworm lying on their back. Cuvier, who strongly objected to any notion of organic evolution and claimed that lifeforms did not evolve over time, was generally regarded as the winner of the debate. Moreover, Geoffroy had to endure the mockery of caricaturists for his conjecture about “the worm that turned”. “The worm that turned” is still a widely used expression and was even chosen as the title for a popular sketch show aired on the BBC in 1980. However, due to the successes of modern evo-devo research – “hoxology” as it was called by the late American evolutionary researcher Stephen Jay Gould – Geoffroy’s idea about the dorsoventral inversion of body axes between vertebrates and arthropods is receiving considerable belated attention. 

Edward Lewis carried out groundbreaking experiments with Drosophila fruit flies that led to the observation that the body patterning along the head-tail axis is the result of the activity of a complex of Hox genes. Subsequent studies showed that the same gene cluster also determines the head-tail axis in mice (see article: “The discovery of homeotic genes”). This can only be explained on the basis that the longitudinal body axes of fruit flies and mice are homologous structures that have been conserved during evolution and hence have a common ancestor, although the internal organisation (body plans) of vertebrates and arthropods differ fundamentally. 

Fruit flies belong to the phylum of arthropods (insects, crayfish, spiders, etc.), which, along with the phylum of annelids (e.g. earthworms), molluscs (snails, shells, squid) and many other invertebrates, make up the Protostomia, a clade of animals in which the nervous system is located on the ventral side, i.e. on the side of the foregut (mouth – anus). Vertebrates (mice, humans), echinoderms (sea urchins, starfish, etc.) and some other phylums belong to the deuterostomes in whom the neural tube is located on the dorsal side of the digestive tract. Protostomes and deuterostomes are subtaxons of the Bilateria which are animals with bilateral symmetry (animals with a front and back end, and upside and downside; they differ from the radially symmetric animals of the Radiata – freshwater polyps, jellyfish, corals). 

Around 540 million years ago (Early Cambrian), there were animals that can be differentiated into protostomes (e.g. arthropods) and deuterostomes (e.g. the Haikouichthys shown in the photo). It is believed that the bilaterians’ hypothetical last common ancestors, the urbilaterians (also referred to as protostome-deuterostome ancestor, PDA) most likely lived more than 600 million years ago.

In Drosophila fruit flies, the Hox genes are arranged one after another on the chromosome in clusters (gene cassettes). Mice and humans have around 40 Hox genes within 4 Hox clusters located on four different chromosomes. While it was previously assumed that fruit flies reflect the original situation and that the complex vertebrate pattern developed later, new investigations support the idea that the opposite is in fact the case. 

Dr. Detlev Arendt from the Centre for Organismal Studies at Universität Heidelberg (previously at the European Molecular Biology Laboratory in Heidelberg) has shown that the marine ragworm Platynereis dumerilii (which belongs to the annelids, protostomes) has a Hox pattern that is similar to that of vertebrates. Along with the results obtained from a number of comparative genomic investigations using representative organisms, it can be concluded that also the urbilaterians possessed a complex Hox pattern, which is more similar to that of vertebrates than to that of arthropods and nematodes in which many Hox genes have been lost during evolution. It seems as if earlier reconstructions of bilaterian phylogeny did not take into account the loss of genes.

But what led to the divergence of urbilaterians into protostomes and deuterosomes? When he was student at the University of Freiburg, Arendt became interested in Geoffroy St. Hilaire’s work and started to use molecular biology data as supporting evidence for Geoffroy’s theory (Arendt D., Nübler-Jung K., 1994: Inversion of dorsoventral axis? Nature 371, 26). A number of genes involved in the development of the nervous system are expressed in similar though inversely positioned domains in Drosophila and Xenopus (claw frog). The tinman gene, which is also a popular tool in evo-devo research, is expressed very early during early embryonic development in the region of the future heart, ventrally in vertebrates (claw frog, mouse) and dorsally in arthropods (fruit fly).

A conserved cassette of two antagonistic genes, which have been shown to be inverted in fruit flies relative to frogs, is the most convincing evidence so far for substantiating the assumption that the dorsoventral axis was inverted. Drosophila expresses a gene (Dpp, decapentaplegic) on the dorsal side of the blastula; in Xenopus, the homologus gene – BMP-4 (bone morphogenetic protein 4) – is expressed on the ventral side of the blastula. The protein products of these two genes are signalling molecules for specific growth factors. The antagonistic genes are called sog (expressed on the ventral side) in Drosophila and chordin in Xenopus (expressed on the dorsal side). These two genes are also homologous. These findings are of such high importance because this gene pair seems to determine the dorsoventral axis in a similar way as the Hox genes determine segmentation along the head-tail axis. Dpp (BMP-4) and sog (chordin) are not involved the differentiation of a specific cell type or in the development of a specific organ; instead, they are evo-devo genes par excellence used for investigating general body organisation. However, nothing is known about why inversion of the genes and dorsoventral axis occurred in the Precambrian urbilaterians; though there is much speculation. Richard Dawkins (in: The Ancestor’s Tale, 2004) pointed out that there are still animals that swim “upside down”, i.e. belly-up, including protosomes such as brine shrimp (Artemia), backswimmers (Notonectidae, a group of water bugs) and deuterostomes such as catfish (Synodontis nigriventis). Of course, nobody believes that the genes have been inverted in these animals. Is it not possible that given the millions of years over which evolution has taken place that altered ways of life might also have resulted in different body plans? However, with such speculation we have almost come back to where we started, with Geoffroy St. Hilaire and the “worm that turned”.