Perspective Inversion of the chordate body axis: Are there alternatives? J. Gerhart* Department of Molecular and Cell Biology, University of California, Berkeley, CA 94720-3200 One major morphological difference between chordates and annelids or arthropods is the opposite orientation of the nerve cord and heart. A long-standing proposal is that the chordate axis evolved by inverting the body of an ancestor with the annelid͞arthropod orientation. However, the data can also be explained by a common ancestor with diffuse dorsoventral organization, followed by oppositely directed condensation of the nerve cord and relocation of the heart in the two lines. n 1822 Geoffroy St. Hilaire suggested nificance of this ancestor is great in meta- sally) (12). The common ancestor’s com- Ithat arthropods have the inverse dorso- zoan evolution, wrapped up as it is with plexity is also implied by complex antero- ventral organization of chordates. After the origins of bilateral animals. Modifica- posterior expression domains (emx, otx, acceptance of Darwin’s descent with mod- tion of a radially symmetric or biradial Hox genes) in the chordate and arthropod ification, Dohrn (1) proposed in 1875 that body plan into a bilateral body plan with a nervous systems (8), and by expression the last common ancestor of these groups dorsoventral axis is thought to have in- similarities in segmentation, eyes, and was an annelid worm-like animal with a volved many changes, including appear- limbs (13, 14). ventral nerve cord, a dorsal heart, and a ance of mesoderm and coeloms, conver- Furthermore, the embryonic ectoderm circulatory system with blood flowing an- sion of a closed gut to a through-gut, in flies and the embryonic mesoderm in teriorly in dorsal vessels. This orientation cephalization, and centralization of the frogs segregate into two dorsoventral do- was retained by evolving members of the nerve cord from a diffuse net. These mod- mains, inversely related in the two cases 25–30 protostome phyla (such as annelids ifications presumably occurred in steps (3, 10). Cells of one domain secrete cer- ␤ and arthropods) whereas, within the deu- between the time of the first simple bilat- tain type transforming growth factor terostomes (chordates, hemichordates, eral animal, variously called the ‘‘urbilat- signals (bone morphogenetic proteins 2, 4, and echinoderms), an ancestor of chor- eria’’ (3) or ‘‘bilaterogastrea’’ (4), and the and 7 in chordates; Dpp in Drosophila), dates inverted its body, sideways over, but time of the last common ancestor of mod- and cells of the other do not but do secrete PERSPECTIVE retained the same relative organ place- ern bilateral animals. Because the fossil antagonists of these signals (Chordin in ments. The nerve cord was now dorsal, record and modern phylogenies give few chordates; Sog in Drosophila). These the heart ventral, and the blood flowed clues (see below) about the ancestor, it genes seem particularly significant be- anteriorly in ventral vessels (Fig. 1). In would be significant if comparative mo- cause their domains may serve as dorso- chordates, axial muscle blocks were dor- lecular and anatomical data could illumi- ventral compartments of the body axis solateral and visceral mesoderm was ven- nate its character. (like Hox compartments of the anteropos- trolateral whereas protostomes had the op- terior axis). Hence, they might reveal posite arrangement. In the chordate line, Recent Evidence for Inversion. Modern pro- global body organization rather than spe- SPECIAL FEATURE the mouth eventually formed on the new ponents of inversion, who have taken gene cific cell type differentiation or local organ ventral side and vanished from the old lo- expression data as evidence, include development. The molecular data support cation. Thus, in the amended inversion hy- Arendt and Nu¨bler-Jung (5–9), De Rob- the assertions from morphology that nerve cords, hearts, and dorsoventral or- pothesis, the last common ancestor already ertis and Sasai (2), and Holley and Fer- ganization of all modern bilateral animals had a complex differentiated dorsoventral guson (10). Data are striking on the sim- derive from organs and organization of a axis. After protostomes and deuterostomes ilar but inverted domains of expression of complex common ancestor, with body diverged, members of each group would several orthologous genes in embryos of inversion in one of the two lines. have added organs along this axis. In chor- chordates (mostly frogs and mice) and When did inversion occur? Deutero- dates, for example, besides a dorsal hollow protostomes (mostly Drosophila). In the stomes currently include chordates, hemi- nerve cord and ventral heart, there evolved nerve cord, three genes, vnd͞Nk2, ind͞ ͞ chordates, and echinoderms (15, 16). Be- a dorsal notochord, ventrolateral gill slits, Gsh1,2, and Msh Msx1,3, are expressed in cause the body axis of echinoderms is a ventral endostyle in the pharynx, and a three columns of nerve cells aligned me- greatly (and intriguingly) modified, only dorsal postanal tail. None of these is found dial to lateral (11). Also, the netrin gene(s) the hemichordate-chordate comparison in protostomes, which would have added is expressed in the medial floor plate of has been pursued. The worm-like entero- other organs. the chordate nerve cord and the ventral pneust hemichordates differ from chor- After denunciation of it by Cuvier in midline mesectoderm of Drosophila. Be- dates in having a ventral nerve cord, a 1830, the hypothesis was reasserted and cause the similarities seem too complex to protostome-like direction of blood flow, a rejected every 50 years to the present (1). have arisen by evolutionary convergence, liver-like organ protruding dorsally from Recent data on regional gene expression the data imply a complex nerve cord in the the gut (rather than ventrally as in chor- in embryos of fruit flies, frogs, and mice last common ancestor. Results with the dates), a notochord-like pygocord located ͞ have brought new credibility to the hy- heart are also striking. The tinman csk ventrally and posteriorly, a ventral post- pothesis and to deductions about the com- (also called Nkx2-5) gene is expressed very plexity of the last common ancestor of early in the prospective heart region in protostomes and deuterostomes. The sig- chordates (ventrally) and arthropods (dor- *E-mail: [email protected]. PNAS ͉ April 25, 2000 ͉ vol. 97 ͉ no. 9 ͉ 4445–4448 Downloaded by guest on September 30, 2021 new ventral nerve cord in an animal al- ready having a mouth (ventral by defini- tion). Then, inversion occurred later in the chordate line to make it a dorsal cord, and the mouth relocated to the opposite side. Lacalli (21) suggested that a larval ances- tor formed the nerve cord from ciliary rows at a time when it had only one terminal gut opening, and, hence, its dor- soventral axis was ambiguous. Descen- dents on the deuterostome branch perfo- rated a mouth on the side opposite the nerve cord, making it dorsal. Those on the Fig. 1. The inversion hypothesis. (A) An annelid worm, side view. The mouth (m) and nerve cord (dark shading) are ventral. The gut (light shading) is midlevel. Arrows indicate the direction of blood flow. protostome branch formed a mouth on the Inverted, it is a chordate, with the nerve cord dorsal, the gut ventral, and the blood flowing in the opposite same side as the nerve cord, making it direction. A new mouth (stomadeum) and anus (proctodeum) evolve in the chordate. Modified from ref. ventral. Inversion did not really occur 2. (B) The dorsoventral axis in cross section, trunk level. because the mouth arose after the nerve cord. In a related idea (22), a ctenophore- like ancestor with a concentrated nerve anal tail, and a dorsal endostyle-like organ have a mouth separate from the anus, by cord had two opposing openings (anal (Table 1). From these differences, several which to assign dorsoventral orientation? pores) in addition to a terminal gut open- authors (9, 17–19) have concluded that In the absence of intermediate phyla, the ing, also precluding a definition of the enteropneusts have protostome-like orga- possibility of a complex, well organized dorsoventral axis. One of these pores was nization and that inversion must have ancestor cannot be excluded, but it gains retained as a mouth by protostomes and occurred in the chordate line after they no additional support. the other by deuterostomes. These hy- branched off. As noted later, other fea- Critics of inversion have usually favored potheses have in common that the dorso- tures can be used to argue for non- a less complex common ancestor and a ventral axis of the last common ancestor of inversion. different path of chordate evolution. modern bilateral animals was relatively Some of these hypotheses can accommo- undifferentiated and ambiguous, and, Alternatives to Inversion. Recent molecular date the morphological and molecular hence, inversion from one orientation to phylogenies (18S RNA sequences) have data as well as does the inversion hypoth- another is not an issue. As discussed later, deepened the uncertainty about the com- esis. In the Auricularia Hypothesis (20), these alternatives must still rationalize the plexity of the last common ancestor (15, hemichordates and chordates evolved striking inverted gene expression patterns. 16). Bilateral metazoa now fall into three from a bilateral larval ancestor. Through a Related hypotheses involving an ancestor large groups of phyla: the ecdysozoa, lo- series of intermediates, bilateral ciliary such as a bilaterogastrea (4) or a larva-like photrochozoa, and deuterostomes. Deu- rows and the associated nerve net moved bilateral micrometazoan (23) as the plat- terostomes are the sister group of the dorsally, fused at the midline, and sank form for the diversification of bilateral other two protostome groups. In this phy- inside to form a new dorsal cord, without metazoa address early steps of evolving logeny, few, if any, intermediate phyla inversion (paedomorphosis followed).