COMMENTARY

The molecular ancestry of segmentation mechanisms

E. M. De Robertis1 Howard Hughes Medical Institute and Department of Biological Chemistry, University of California, Los Angeles, CA 90095-1662

n 1830 a very important debate on pair-rule, and segment polarity genes that stripes of Delta and hairy are cyclical and natural history took place in the encode transcription factors (7). Most move rhythmically from the post- French Academy of Sciences. As re- other insects, including the cockroach, erior to the anterior growth zone every told in enjoyable detail by T. A. Ap- develop from a short embryonic germ- time a new segment is formed in this in- pelI (1), the adversaries were Georges Cu- band within a much larger egg. New sect (4). vier and Etienne Geoffroy Saint-Hilaire. metameres are added sequentially through These were pre-Darwinian times—the the proliferation of a posterior growth A Requirement for Notch Origin of Species was published in zone. This sequential addition of In the vertebrates, the cycling behavior of 1859—so their arguments sound anti- metameres resembles segmentation in the chick hairy has been known since the quated today, but their reverberations verterbrate posterior paraxial mesoderm. landmark experiment of the Pourquie´ among biological ideas have continued to After Periplaneta segment borders are group (8), in which the paraxial meso- the present day. Cuvier, the discoverer of formed (and marked by a stripe of the derm was bisected. One half was fixed extinction, held the view that anat- and the other incubated for variable times, omy was determined by the functional revealing posterior-to-anterior waves or purpose of each organ, which he called The Notch pathway cycles of expression with the same period- the ‘‘conditions of existence.’’ Geoffroy, icity as somite formation. The hairy gene the discoverer of ‘‘analogies’’ (now called is required for was discovered in Drosophila as a pair homologies), held that animal anatomy rule gene (7). However, Notch signaling had a ‘‘unity of plan’’ upon which thou- segmentation in plays no role in Drosophila segmentation, sands of variations were imposed. He whereas in the vertebrates it is essential compared vertebrates to arthropods, argu- both arthropods for the segmentation clock (9). ing that they shared antero–posterior In the cockroach, RNAi is very effec- (A–P) characteristics, such as head, tho- and vertebrates. tive when injected into the abdominal cav- rax, and abdomen, as well as dorsal– ity of virgin females. Many other ventral (D–V) landmarks such as a CNS, arthropods, such as beetles, moths, crick- , and heart, except homeobox gene engrailed), the mesoderm ets, and spiders, also show potent systemic that a D-V inversion of the body plan had separates from the ectoderm forming effects of parental RNAi. In Drosophila, occurred. Geoffroy even had the temerity somites with an internal cavity. Formation however, the effects of RNAi are limited of comparing the segmentation of arthro- of mesodermal somites represents an ad- to the cells in which it is expressed (this is pods to that of vertebrates. Cuvier pre- ditional similarity with vertebrate believed to be due to the loss of a gene sented the better arguments and was con- metamerism. In contrast, Drosophila seg- encoding an siRNA transporter in the cell sidered the victor of the debate. However, mentation takes place exclusively in the membrane). modern evolution and development stud- ectodermal layer. The greatest advantage In Periplaneta, parental injection of ies have discovered that common genetic of using Periplaneta is that the female lays RNAi for Notch or hairy strongly inhib- networks pattern the A–P and D–V axes daily clutches of eggs encapsulated in an ited segmentation in the posterior region of very distantly related bilateral oo¨theca. Eggs are laid in a gradient of age of the , causing truncations. Notch (2, 3). This realization has provided a from one end of the oo¨theca to the other, is a transmembrane receptor that requires measure of molecular support for Geof- representing a time series that differs by froy’s unity of plan hypothesis and for the less than one segment (a new segment is the proteolytic cleavage and nuclear trans- idea that the last common ancestor of the formed every6hinthecockroach). Fur- location of its intracellular domain for invertebrate and vertebrate lineages was a thermore, there are two rows of signaling. The last step after activation of Notch by Delta ligand is the cleavage of rather complex animal, called Urbilateria in each oo¨theca, so that one side can be ␥ (which means primeval bilateral animal), fixed and the other cultured further. the transmembrane domain by -Secre- which predated the Through detailed in situ hybridization tase, an intramembranous protease for that took place 535–525 million years ago studies, Pueyo et al. (4) observed waves of which an effective chemical inhibitor, (3). In this issue of PNAS, Pueyo et al. (4) expression of Delta, a gene that encodes a called DAPT, exists. Both Notch-RNAi make an important contribution to the transmembrane protein that binds to and and immersion in DAPT inhibited seg- question of whether common mechanisms activates the receptor Notch in adjoining mentation and the cycling behavior of of segmentation are shared by insects and cells in the posterior growth zone. Notch hairy (4). vertebrates (5, 6). and its target gene hairy (a basic-loop– The evolutionary conservation of the helix transcription factor) were also ex- segmentation clock mechanism is the im- The Importance of Being pressed, leading portant contribution made by this study in a Cockroach Embryo to the activation of a band of the Periplaneta embryos (4). The new results The key to studying in detail the expres- homeobox gene engrailed once each new should be compared with previous work sion of genes known to regulate verte- segment is formed. Only one band was in the spider embryo, in which it was also brate segmentation was the use of a seen for Delta, Notch, and hairy, with their particularly favorable embryo, the cock- order of expression in a posterior-to-ante- roach Periplaneta americana. Drosophila rior progression consistent with the view Author contributions: E.M.D.R. wrote the paper. has a long germ-band type of develop- that their sequential activation leads to The author declares no conflict of interest. ment, in which the embryo develops very the expression of engrailed and a new seg- See companion article on page 16614. rapidly and all segments develop almost ment. The most important new observa- 1E-mail: [email protected]. simultaneously under the control of gap, tion reported in this study is that the © 2008 by The National Academy of Sciences of the USA

www.pnas.org͞cgi͞doi͞10.1073͞pnas.0808774105 PNAS ͉ October 28, 2008 ͉ vol. 105 ͉ no. 43 ͉ 16411–16412 Downloaded by guest on October 3, 2021 larval phase, because many phyla of the two main branches of the bilateral animals (Fig. 1)—such as for example annelids, mollusks, hemichordates, and echino- derms—have retained the ancestral pri- mary pelagic larval phase (14, 15). These larval forms share many anatomi- cal similarities (Fig. 1). However, the ancestral pelagic larval phase of the life cycle has been lost repeatedly in many marine animal lineages (16). It is interesting to reflect on how con- tingent the future body plans of most ani- mals that surround us today were on the survival of the lineage. Present-day fauna would probably look very different if the lineage of this com- plex ancestor had become extinct. The finding of cycling of the Notch pathway in Periplaneta and vertebrates certainly Fig. 1. Diagram showing the two great branches of the . (proto, first; stomo, strengthens the idea of a segmented com- mouth) develop the mouth near the initial site of the blastopore. (deutero, second) develop the anus close to the blastopore and the mouth is perforated secondarily. Urbilateria is the mon ancestor. It should be mentioned hypothetical last common ancestor of the two bilaterian branches and is depicted here as a complex that other authors are of the opinion that segmented bottom-dwelling (benthic) animal. Many extant and phyla have the last common ancestor could not have primary larval forms that are anatomically very similar to each other and are therefore considered been a complex creature because of the ancestral (14–16). Shown here is a trochophore-like larva having ciliary bands that collect food by beating lack of a record (17). As Pueyo et al. in opposite directions, an apical eye and an apical ciliary tuft. These larvae are free-swimming (pelagic) and (4) discuss, it always remains a possibility have a planktonic existence until they settle in the ocean bottom for their adult life phase. Ectoderm is that the Notch pathway was recruited by shown in green, CNS in blue, eye in black, and endoderm in red, with its openings in yellow. rather than by de- scent from a common ancestor. For exam- ple, it could be that Notch was selected Notch Delta found that and parental not seen in vertebrate somites. However, because it was particularly well suited for RNAi caused segmental malformations it is worth keeping in mind that a classical the establishment of boundaries. However, (10, 11). Therefore, strong evidence that arthropod segmentation gene does display the rhythmic segmentation clock would the Notch pathway was required for seg- the expected metameric expression pat- have had to be constructed independently mentation in both arthropods and verte- tern in the primitive chordate amphioxus. as well. brates existed previously. However, the There are many genes that oscillate Evolutionary Implications earlier work in the spider embryo had rhythmically during vertebrate somite for- been unable to document the existence of The fact that an insect and the vertebrates mation, in particular components of the a cycling mechanism in the posterior share rhythmic cycling of the Notch path- Wnt, FGF/MAPK, and Notch pathways growth zone, a demonstration for which way genes raises the question of whether (18). The ancestry of segmentation will be the cockroach oo¨theca was better suited. the Urbilaterian ancestor was a segmented resolved, and perhaps Geoffroy vindi- The fact that spiders and a primitive in- animal (Fig. 1). If this were the case, it cated, when additional genes are found to sect require Notch for segmentation sug- would have very important evolutionary play conserved roles in segment formation gests that this mechanism was ancestral to implications. Thirty-five animal phyla ex- in multiple phyla. In the ideal case, new the arthropod lineage. When discussing ist, of which 30 correspond to bilaterians homologies might be found between the the ancestry of segmentation, mention (13). A segmented ancestor would repre- segmentation of vertebrates and that of should also be made of an earlier study in sent a bottom-dwelling (called benthic) Drosophila, the species that initiated the the cephalochordate amphioxus, in which ancestor with a body suited for an exis- molecular dissection of metamerism (7). the first eight somites had a stripe of en- tence burrowing in the ocean floor. This grailed in the posterior half (12). It has body plan would correspond to the adult ACKNOWLEDGMENTS. I thank V. Taelman and been argued that this discovery affected form of the animal. It is very likely that J.-L. Plouhinec for drawing Fig. 1. This work was supported by National Institutes of Health Grant only 8 of a total of 50 somites formed in Urbilateria had a complex life cycle with RO1 HD21502-22 and the Howard Hughes Medical amphioxus and that engrailed bands are an initial ciliated free-swimming (pelagic) Institute.

1. Appel TA (1987) The Cuvier-Geoffroy Debate (Oxford 8. Palmeirim I, Henrique D, Ish-Horowicz D, Pourquie´O ment-polarity homolog in Drosophila. Development Univ Press, Oxford). (1997) Avian hairy gene expression identifies a molec- 124:1723–1732. 2. Carroll S (2005) Endless forms most beautiful: The new ular clock linked to vertebrate segmentation and somi- 13. Valentine JW (2004). On the origin of phyla. (Univ of science of Evo Devo (Norton, New York). togenesis. Cell 91:639–648. Chicago Press, Chicago). 3. De Robertis EM (2008) Evo-Devo: Variations on ances- 9. Dale JK, et al. (2003) Periodic notch inhibition by lunatic 14. Ja¨ gersten G (1972) Evolution of the Metazoan tral themes. Cell 132:185–195. fringe underlies the chick segmentation clock. Nature Life Cycle: A Comprehensive theory (Academic, 4. Pueyo JI, Lanfear R, Couso JP (2008) Ancestral notch- 421:275–278. London). mediated segmentation revealed in the cockroach 10. Stollewerk A, Schoppmeier M, Damen WGM (2003) 15. Nielsen C (1998) Origin and evolution of animal life Periplaneta americana. Proc Natl Acad Sci USA Involvement of Notch and Delta genes in spider seg- cycles. Biol Rev 73:125–155. 105:16614–16619. mentation. Nature 423:863–865. 16. Gilbert SF, Raunio AM, eds (1997) Embryology: Con- 5. Kimmel CB (1996) Was Urbilateria segmented? Trends 11. Damen WG (2007) Evolutionary conservation and di- structing the (Sinauer, Sunderland, MA) Genet 12:320–331. vergence of the segmentation process in arthropods. 17. Erwin DH, Davidson EH (2002) The last common 6. Patel NH (2003) The ancestry of segmentation. Dev Cell Dev Dyn 236:1379–1391. bilaterian ancestor. Development 129:3021– 5:2–4. 12. Holland LZ, Kene M, Williams NA, Holland ND (1997) 3032. 7. Nu¨sslein-Volhard C, Wieschaus E (1980) Mutations af- Sequence and embryonic expression of the am- 18. Deqe´ant ML, et al. (2006) A complex oscillating net- fecting segment number and polarity in Drosophila. phioxus engrailed gene (AmphiEn): The metameric work of signaling genes underlies the mouse segmen- Nature 287:795–801. pattern of transcription resembles that of its seg- tation clock. Science 314:1595–1598.

16412 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0808774105 De Robertis Downloaded by guest on October 3, 2021