BMP shuttling and the evolution of animal bilaterality

The evolution of bilateral body symmetry marked a major transition in the life history of animals allowing for more complex and diverse body plans. Bilaterality is a name-giving trait of Bilateria (insects, mollusks, worms and vertebrates i.e. all animals with a head-tail and back-belly axes), but also in the sea anemones and corals. The latter two belong to a group Cnidaria, an evolutionary sister of Bilateria, which also encompasses animals with a radial body symmetry: jellyfish, hydras and the like. It is thus still unclear whether bilaterality evolved independently in Cnidaria and in Bilateria or whether their last common ancestor was already a bilaterally symmetric animal, and radially symmetric cnidarians lost bilaterality later. Intriguingly, the back- belly axis in Bilateria and the directive axis in bilaterally symmetric sea anemones and corals are regulated by the same molecular mechanism: the BMP signaling. BMPs are secreted proteins, which can be bound and prevented from activating their receptors by another protein, Chordin. Chordin, however, is not a simple BMP inhibitor, because it can also diffuse with the BMP bound to it over long distances and then release a functional BMP ligand far away from the Chordin source. This process of Chordin-assisted BMP signaling known as BMP shuttling is crucial for the regulation of the back-belly axis in frogs and flies but also of the directive axis in sea anemones. In contrast, in jellyfish and other radially symmetric cnidarians studied to-date the genes coding either for Chordin or for the main BMP protein termed BMP2/4 appear to be lost. It is tempting to speculate that the loss of these genes or the lack of functionality of the Chordin- mediated BMP shuttling might have caused the loss of bilaterality in radially symmetric cnidarians. To test this, we will first check by sequencing and analyzing the genomes of the representatives of two uncharacterized groups of radially symmetric cnidarians whether Chordin and BMP2/4 are really missing in all groups of radially symmetric cnidarians. Then, we will analyze the capacity of cnidarian Chordin proteins to shuttle BMP ligands using fruit flies, in which their own Chordin gene was switched off, as a testbed. If we show that the sea anemone Chordin is indeed capable of BMP shuttling and that crucial shuttling components are indeed absent in the radially symmetric cnidarians, it will be a strong argument in favor of moving the emergence of animal bilaterality some 200 million years back in time and placing it prior to the cnidarian-bilaterian split.

BMP signaling is responsible for patterning secondary body axes in Bilateria and in bilaterally symmetric Cnidaria (corals, sea anemones). Radially symmetric Cnidaria (jellyfish, hydroids) also have all intracellular BMP signaling components, however, they often lose either the main BMP ligand BMP2/4 or the key secreted BMP antagonist Chordin. Since radially symmetric cnidarians did not lose BMP signaling machinery, it was obvious that BMP signaling function cannot be restricted to the generation of the second axis however, it was unclear what the other functions might be. Moreover, Chordin has a unique capacity of "shuttling" BMP ligands in Bilateria, i.e. repressing BMP signaling locally but facilitating BMP diffusion and promoting BMP signaling at a distance. We proposed that this Chordin-mediated "shuttling" mechanism may be responsible for the BMP-dependent bilateral symmetry of sea anemones and, possibly, for the bilateral symmetry of the cnidarian-bilaterian ancestor. In this FWF project, we addressed two questions: i) is BMP shuttling really the mechanism of Chordin action in the sea anemone Nematostella, and ii) what does BMP signaling do when it does not pattern a body axis. To answer the first question, after testing several unsuccessful approaches, we established an experimental system in which we could generate local sources of Chordin and different BMP ligands in a genetic background lacking Chordin and one or more BMPs. Using this "local source" system, we proved that Chordin indeed shuttles BMP ligands in the Nematostella embryo. To address the second question, we used an antibody recognizing the activated form of the BMP signaling effector SMAD1/5 and created a whole-body atlas of the BMP signaling activity in the adult Nematostella polyp, i.e. at the stage when no secondary axis patterning is taking place any longer. Then we correlated it with the single-cell transcriptomic data we generated from this life stage. Using single-cell data, we showed that a broad range of cell types can receive BMP signals from a much smaller number of emitting cell types. Using a combination of immunohistochemistry and transgenic reporter line analyses, we could identify specific cell populations emitting and receiving BMP signals. Among the most likely possible functions of BMP signaling not related to axial patterning we identified its involvement in tentacle formation, gameto- or gonadogenesis and differentiation of specific neural cell types. Analysis of BMP signaling in radially symmetric cnidarians Aurelia and Tripedalia suggest that the latter two roles of BMP signaling may be conserved throughout Cnidaria and potentially between Cnidaria and Bilateria.