The mouth, the anus, and the blastopore—open questions about questionable openings

Gastrulation is one of the major events during the embryogenesis of an animal. In addition to the for- mation of the germ layers it is often the time when the future axial properties and digestive open- ings become apparent, and it is not surprising that this event plays an important role in hypotheses regarding metazoan evolution. A major difference between these theories concerns the structure of the alimentary canal and the relationship of its openings to the blastopore of the last common bilaterian ancestor. Here we review competing theories of bilaterian evolution and evaluate their plausibility in the light of recent insights into meta- zoan phylogeny and development.


4.1 Gastrulation as a process determining multiple body plan characteristics

The evolution of an internal germ layer enabled the compartmentalization of the body of multicellular animals (Metazoa) into a digestive layer (endo- derm) and an outer layer, the integument (ecto- derm). The developmental process that separates the inner from the outer cell populations is called gastrulation. During gastrulation, cells or cell layers are internalized and later form the digestive epithe- lium and often also germ cells. Bilaterians have in addition a third germ layer, the mesoderm, which either separates from the endoderm or ingresses as independent precursors. Gastrulation not only generates distinct cell types, it also establishes the organismal axes from a pre-existing animal–vegetal embryonic axis. One can thus picture gastrulation as
a fundamental event in every organism as it deter- mines the main body features during development. Consequently, gastrulation plays a major role in hypotheses regarding the transition from a radi- ally symmetrical diploblastic animal to a bilaterally symmetrical triploblastic organism with meso- derm, the last common bilaterian ancestor.


4.2 The blastopore as the site of internalization

In most animals the site of internalization of gas- trulating cells is limited to a specific area, the blas- topore. The fate of this site is often not only the area of germ layer specification, it sometimes becomes the connection of the endodermal digestive cavity to the ectoderm and thus to the animal’s environ- ment. This connection is usually called either the
‘mouth’ if the organism has only a single opening to the gut or if it is the anterior opening of a through gut, or the ‘anus’ when it corresponds to the pos- terior opening and functions as a site of excretion. Grobben (1908) subdivided the bilaterian clade into the taxa Protostomia and Deuterostomia based on the fate of the blastopore becoming either the mouth or the anus. Grobben claimed that in most proto- stomes (‘first mouth’) the blastopore becomes the mouth, and the anus is formed secondarily at a dif- ferent site, later during embryogenesis. In contrast to protostomes, in deuterostome embryos the site of gastrulation becomes the anus while the mouth is formed at a different site in the animal hemi- sphere of the embryo (Figure 4.1). The blastopore becoming the anus appears to be ancestral for the


33

34 AN I M AL EV O L UTI O N



I. II.
M


A
M M



A A
Protostomy Amphistomy



M M
M
variable (see Table 4.1) and the reconstruction of the ancestral type of gastrulation is further hin- dered by intrataxon variation within larger clades such as annelids, nematodes, arthropods, and nemerteans (Table 4.1). The variability of the rela- tionship between the blastopore and its possible future fates—mouth or anus—has led Ulrich (1951) to dismiss the term Protostomia and rename it Gastroneuralia, based on the ventral localization of nerve cords in these taxa. However, as it is not clear if a ventrally centralized nervous system is part of the ground pattern of the Protostomia, it is premature to rename this clade.


A


A
Deuterostomy
4.3 The fate of the blastopore and its role in scenarios of bilaterian evolution

The significance of transitions from the cnidarians


Figure 4.1 Protostomy, deuterostomy, and amphistomy and evolutionary scenarios about the direction of change. In protostomy, the blastopore (bright in all figures) is displaced from posterior to antero-ventral by morphogenetic movements (bright arrow), and gives rise to the mouth (M). The anus (A), if present, is formed at a different site at the posterior end of the embryo. In deuterostomy
the blastopore stays at the posterior site of the embryo and either closes or gives rise to the anus. The mouth is formed at a different site from the former blastopore in the anterior of the embryo. In amphistomy, the blastopore elongates and closes laterally with both ends giving rise to the mouth and anus. One scenario (I) for
the evolution of deuterostomy is using the protostomy as ancestral, proposing the abbreviation of the anterior movement of the blastopore (dotted bright arrow) by a molecular separation of the mouth determination from the blastopore (intermediate stage small). The other scenario (II) claims the amphistomy as being ancestral, from which deuterostomy and protostomy evolved by closure of the blastopore from the posterior to anterior direction (protostomy) or from the anterior to posterior direction (deuterostomy).



deuterostomes, but the situation in the protostomes is unclear due to its extreme variability in differ- ent forms (Table 4.1) and because the phylogenetic relationships of several key taxa are not known. In some cases a large maternal investment of yolk (e.g. hexapods, cephalopods, and onychophorans) influ- ences the gastrulation process, similar to what is found in most amniote embryos.
However, even in embryos with small quan- tities of yolk, the fate of the blastopore is highly
to the Bilateria, which possess an antero-posterior and a dorsoventral axis, might be one of the big- gest controversies in zoology over the last centur- ies. In cnidarians and ctenophores, two groups that diverged prior to the origin of the Bilateria, the blas- topore gives rise to a single opening (mouth–anus) of the gastric digestive cavity, while in many bilat- erians two openings are present, both of which can be formed independently from the blastopore. The scenarios under debate for explaining this transi- tion differ in various details, such as the presence of a coelom or a larval stage. However, the forma- tion of a through gut and the fate of the blastopore to the openings are central to understanding bilat- erian body plan evolution.


4.3.1 The gastraea theory and the ancestral lateral closure of a slit like blastopore

The gastraea theory (Haeckel, 1872, 1874) proposes a hypothetical metazoan ancestor similar to the gastrula stage of recent animals. A fundamental assumption in the future variations of Haeckel’s theme of the gastraea, for example the bilaterogas- traea theory (Jägersten, 1955) and the trochaea the- ory (Nielsen and Nørrevang, 1985), is the extension of the ventral blastopore along the antero-posterior body axis. In these theories, a lateral closure of this elongated blastopore, which stays open at its ends,

TH E M O UTH , TH E AN U S , AN D TH E B L A S T O POR E 35


Table 4.1 Blastopore fates in bilaterian taxa

Taxon Fate of blastopore References

Gastrotricha Protostomy (Lepidodermella, Turbanella) Lepidodermella (Sacks, 1955), Turbanella (Teuchert, 1968)

Nematoda Protostomy (Tobrilus, Ascaris), blastopore closure (Tylenchida)
Tobrilus (Schierenberg, 2005), Ascaris (Boveri, 1899),
Tylenchida (Malakhov, 1994)
Nematomorpha Deuterostomy (Paragordius) Paragordius (Montgomery, 1904)

Mollusca Protostomy (Crepidula, Patella), deuterostomy (Viviparus)
Crepidula (Conklin, 1897; Hejnol et al., 2007), Patella
(Dictus and Damen, 1997), Viviparus (Dautert, 1929)
Priapulida Blastopore closure? Priapulus (Wennberg et al., 2008) Kinorhyncha ?
Loricifera ?
Platyhelminthes Protostomy (Planocera, Hoploplana) Planocera (Surface, 1907), Hoploplana (Boyer et al., 1998) Gnathostomulida ?
Rotifera Protostomy (Asplanchna, Calidina, Philodina) Asplanchna (Lechner, 1966), Calidina (Zelinka, 1891),
Philodina (AH, unpublished) Entoprocta Protostomy (Pedicellina) Pedicellina (Marcus, 1939)

Nemertea Protostomy (Procephalotrix), deuterostomy
(Drepanophorus)
Annelida Protostomy (Polygordius), deuterostomy
(Eunice), blastopore closure (Capitella)
Procephalotrix (Iwata, 1985), Depranophorus (Friedrich,
1979)
Polygordius (Woltereck, 1904), Eunice (Åkesson, 1967),
Capitella (Eisig, 1898)
Sipunculida Protostomy (Phascolosoma) Phascolosoma (Gerould, 1906) Cycliophora ?
Phoronida Protostomy (Phoronopsis) Phoronopsis (Rattenbury, 1954)
Brachiopoda Protostomy (Terebratulina) Terebratulina (Conklin, 1902)

Onychophora Mouth and anus form at different site of blastopore (Peripatopsis)
Peripatopsis (Manton, 1949)
Tardigrada Protostomy (Thulinia) Thulinia (Hejnol and Schnabel, 2005)

Arthropoda Protostomy (Cyprideis), deuterostomy
(Meganyctiphanes)
Cyprideis (Weygoldt, 1960), Meganyctiphanes
(Alwes and Scholtz, 2004)
Chaetognatha Deuterostomy (after blastopore closure) Sagitta (Hertwig, 1880) Xenoturbella ?
Hemichordata Deuterostomy (Balanoglossus) Balanoglossus (Heider, 1909) Echinodermata Deuterostomy (Synapta) Synapta (Selenka, 1876) Cephalochordata Deuterostomy (Branchiostoma) Branchiostoma (Cerfontaine, 1906)

? Indicates unknown



gives rise to a mouth and anus in the bilaterally symmetrical ancestor with a through gut. This pro- cess has been termed ‘amphistomy’ by Arendt and Nübler-Jung (1997). The same concept was the foun- dation of Remane’s enterocoely theory (Remane,
1950), which begins with a cnidarian polyp trans- forming into an ‘ur-bilaterian’ by stretching along its directive axis. Remane’s theory assumes the simultaneous evolution of the mouth and anus, and predicts that the coeloms of this hypothetical ances- tor are formed from the common gastric pouches of an anthozoan-like polyp (Remane, 1950). These
theories thus predict a rather morphologically com- plex bilaterian ancestor, which is consistent with what has also been proposed on the basis of similar- ities of the expression of some developmental genes (Carroll et al., 2001; Arendt, 2004). The two common themes of these theories are: (1) the simultaneous evolution of the mouth and anus, and (2) that the blastopore gives rise to both openings in the com- mon ancestor. Many authors suggest that a slit-like blastopore is ancestral for the Bilateria and argue that a variety of extant animals such as onycho- phorans, polychaetes, insects and some nematodes

36 AN I M AL EV O L UTI O N



appear to show this kind of gastrulation (Sedgwick,
1884; Arendt and Nübler-Jung, 1997; Nielsen, 2001;
Malakhov, 2004). The most elaborate explanation
of the evolution of gastrulation processes which
supports the gastraea-based theories is delivered
by Arendt and Nübler-Jung (1997). These authors
explain the evolution of deuterostomy by a closure
of the slit-like blastopore from anterior to poster-
ior, leaving an opening, which becomes the anus
(Figure 4.1). Accordingly, protostomy evolved by a
closure of the blastopore from posterior to anterior
leaving a mouth open (Figure 4.1). Indeed, some
bilaterians show a ventrally elongated blastopore
that follows this pattern (e.g. Polygordius).


4.3.2 The acoeloid-planuloid theory and the ancestrality of protostomy

Competing with ‘gastraea’-based theories is a dif- ferent view that does not require the simultaneous evolution of mouth and anus to establish bilateral- ity. The starting points of these hypotheses were pioneered by von Graff (1891), who proposed a paedomorphic planula larva, similar to that of recent cnidarians, which adopted a benthic life- style and flattened along one body axis, giving rise to the bilateral symmetry of the Bilateria (Hyman,
1951; Salvini-Plawen, 1978). The authors suggest that the former posterior blastopore—which gives rise to the mouth—is shifted to the ventral body side, to facilitate uptake of food from the ventral sur- face. This condition is represented by extant acoel and nemertodermatid flatworms, which have only a ventral opening to their digestive cavity. A pos- terior position of the mouth is found in the nearly radially symmetrical acoel Diopisthoporus, which is thought to reflect the ancestral planula-like condi- tion (Beklemishev, 1969; Reisinger, 1970) of some feeding anthozoan planula larvae (Widersten,
1973).
According to these acoeloid-planuloid theories,
the last common bilaterian ancestor was a rather
simple benthic, probably meiofaunal, worm lack-
ing a through-gut, coeloms, and excretory organs
(Hejnol and Martindale, 2008b).
How do proponents of the acoeloid-planuloid
theory explain the variation of the gastrulation
types in the Bilateria and how this is related to the
evolution of the anus? The most thorough thoughts about this problem are presented by Salvini-Plawen (1978, 1980). He points out that in many spiralian embryos the vegetal (=posterior) blastopore gets displaced into the antero-ventral direction (e.g. in molluscs, annelids, nemerteans, and polyclad flatworms), and thus recapitulates the evolution- ary process (protostomy). The anus is thought to have evolved independently from the blastopore in multiple lineages, which is also reflected by the late developmental formation of the anus in many pro- tostomes. The deuterostome condition is explained with an evolutionary ‘abbreviation’ of the anter- ior movement of the blastopore (Figure 4.1). The mouth, instead of moving anteriorly in the form of the blastopore, is immediately formed at its final location (Beklemishev, 1969; Salvini-Plawen, 1980) and the blastopore either closes completely, such as in chaetognaths and nemerteans, or stays open and forms the anus, as in nematomorphs, deuteros- tomes, and several crustaceans (Table 4.1).


4.4 Recent progress in molecular systematics and developmental biology and their impact on
the problem

Both competing scenarios differ in their assump- tion about which type of gastrulation is ancestral for the Bilateria. In gastraea-based theories the lat- eral closure of the blastopore—or ‘amphistomy’— with the simultaneous evolution of the orifices would deliver the state from which the diversity of gastrulation can be derived. In the acoeloid-planu- loid theory, protostomy is ancestral, including an independent evolution of an anal opening. Since both hypotheses have their roots in a time when the metazoan phylogeny was speculative, a proper phylogenetic framework is required to determine the direction of evolutionary change.


4.5 A new animal phylogeny

Recent progress in molecular biology, com- puter technology and the development of new phylogenetic reconstruction algorithms have improved the ability to determine animal rela- tionships with the use of molecular data (Philippe

TH E M O UTH , TH E AN U S , AN D TH E B L A S T O POR E 37



and Telford, 2006; Dunn et al., 2008). In addition to the seminal publication of Aguinaldo et al., (1997) which established the subdivision of the Bilateria into three large clades, Ecdysozoa, Lophotrochozoa, and Deuterostomia, further resolution of metazoan relationships has been accomplished by increased taxon sampling and the use of phylogenomic approaches (Dunn et al., 2008). A major result per- tinent to understanding the role of gastrulation in body plan evolution is the placement of the acoels as the sister group to the remaining Bilateria (Ruiz- Trillo et al., 1999; Baguñà and Riutort, 2004). Their position has been corroborated by multiple inde- pendent molecular approaches (see Telford et al.,
2003, for example). Our current understanding places the nemertodermatids as sister to Bilateria, and Acoela as sister to that group, thus breaking the monophyly of the Acoelomorpha (Jondelius et al., 2002; Wallberg et al., 2007). Having the acoel flatworms at the base of the Bilateria has important implications for our understanding of the evolution of organ systems (Hejnol and Martindale, 2008b). The similarity of the body plan of acoels and nemer- todermatids, both possessing only one opening to their digestive system and an orthogonal nervous system and lacking a through gut and nephridia, clearly support a simple acoeloid bilaterian ances- tor, which was previously proposed on the basis of morphological data (Hyman, 1951; Salvini-Plawen,
1978; Haszprunar, 1996a). These data do not sup- port the gastraea theory of Haeckel or the trans- formation of a coelom bearing ur-bilaterian from a sessile cnidarian polyp (Remane, 1950).


4.6 ‘Amphistomy’—a common theme in bilaterian development?

While molecular phylogenetic results support the acoeloid-planuloid theory, the recently improved resolution of the metazoan relationships with the use of the phylogenomic approach has implications for accepting the ‘amphistomy’ hypothesis (Dunn et al., 2008). If one assumes gastrulation with a slit-like blastopore is in the ground pattern of the protostomes and deuterostomes (together named Nephrozoa after Jondelius et al., 2002), one would expect a broad distribution of a lateral closure of a slit-like blastopore in the Bilateria (Arendt and
Nübler-Jung, 1997). The cases for which amphistomy are most commonly cited are onychophorans, the polychaete Polygordius, and the nematode Pontonema (e.g. Arendt and Nübler-Jung, 1997; Nielsen, 2001). In drawings from the early research on the ony- chophoran Peripatus capensis (Figure 4.2a), it indeed seems that a large extended blastopore closes lat- erally and both ends stay open and give rise to the mouth and anus (Balfour, 1883; Sedgwick, 1885). In contrast, Kennel (1885) draws a different picture for Peripatus edwardsii, showing that the opening that gives rise to the mouth and anus is separate from the blastopore, which is positioned more posteriorly (Figure 4.2a). The most thorough analysis of ony- chophoran gastrulation (Manton, 1949) corroborates Kennel’s findings for several onychophoran species and describes the immigration of the mesoderm and germ cells at the posterior blastopore which is never in contact with either mouth or anus. The syncytial development of the yolky onychophoran embryos seems to be a rather derived adaptation to their terrestrial lifestyle, as is the case in other ter- restrial arthropods (e.g. hexapods and myriapods), and does not represent an example of an ‘amphis- tomic’ type of gastrulation. Another taxon often referred to as being ‘amphistomic’ is the polychaete annelids. The traditional example is the description of the development of Polygordius (Woltereck, 1904). A close examination of the original work shows that the extended blastopore first closes laterally, but instead of leaving both ends open, only the anterior edge gives rise to the mouth (Figure 4.2b). The anus is formed later in development one cell row posterior (Figure 4.2b) to the former ‘seam’ of the blastopore (Woltereck, 1904). Thus, the develop- ment of Polygordius follows a protostomic pattern rather than amphistomy. Even if a polychaete can be shown to possess a prototypical amphistomy form of gastrulation, it is difficult to be sure that it is a plesiomorphic character, since we observe a high variation in gastrulation patterns in poly- chaete annelids. Both protostomy (Mead, 1897) and deuterostomy (Åkesson, 1967), have been described in polychaetes as well as in other trochozoan taxa including nemerteans (Friedrich, 1979; Iwata, 1985) and molluscs (Fioroni, 1979; see Table 4.1).
A similar variation of gastrulation is found in
the third taxon for which amphistomy has been


Figure 4.2 Original drawings of the development of the onychophoran Peripatus and the annelid Polygordius. (a) The drawing of Sedgwick (left) shows the supposed ‘amphistomic’ gastrulation in the onychophoran Peripatus capensis. On the right the drawing from the study of Kennel on Peripatus edwarsii, which shows the blastopore (bl.) separate from the secondary opening, which gives rise to the mouth (m.) and anus (a.). (b) Original drawings of the gastrulation of Polygordius, a supposed amphistomic polychaete. The drawing on the left shows the closure of the blastopore (dark black line in the original) leaving the mouth open (protostomy). The hindgut (hg) forms at a different site from the blastopore (arrow).



described, the Nematoda (Malakhov, 1994). Apart from a clear case of blastocoelic protostomy in Tobrilus (Schierenberg, 2005) the site of immigra- tion of the two entoderm cells becomes the future mouth in most nematodes, which is separate from the later immigration site of the mesodermal pre- cursors which are descendants from different lineages (Schierenberg, 2006).
The developmental stage for which amphistomic gastrulation has been described in Pontonema is much later than the immigration of the E precur- sors, which form the endoderm—a process which is usually referred to as gastrulation in nematodes (Schnabel et al., 1997; Sulston et al., 1983; Voronov and Panchin, 1998).
Again, even if the cursory study of Pontonema is correct, it is not clear what type of gastrulation the last common ancestor of nematodes had, because the sister group of nematodes, the Nematomorpha (Dunn et al., 2008) gastrulate by deuterostomy (Montgomery, 1904; Inoue, 1958). Tardigrades, which form the sister group to the Arthropoda and Onychophora (Dunn et al., 2008), gastrulate by immigration of mesodermal and endodermal precursor cells and show protostomy (Hejnol and Schnabel, 2005), with no evidence of enterocoely as has been assumed by early investigators (Marcus,
1929).
If amphistomy was an ancestral pattern, giv- ing rise to both deuterostomy and protostomy, it must have been lost independently in nearly every larger protostome clade. Taken thus, the developmental diversity of bilaterians described today gives little support for either amphistomic gastrulation or Haeckel’s gastraea. Both concepts seem to deliver a feasible evolutionary scenario for the human mind but are not represented in living organisms.


4.7 ‘Abbreviated’ protostomy as a model for the variability of gastrulation?

Attempts to explain variations in bilaterian gastru- lation patterns from the point of view of the acoe- loid-planuloid theory are based on the movement of the vegetal blastopore in the antero-ventral direction, which then gives rise to the mouth in protostomic animals (Figure 4.1). In the case of deuterostomy, the mouth forms at a separate site and no such movement can be observed. Salvini- Plawen (1980) explains the multiple independent origins of deuterostomy in several animal lineages with an evolutionary ‘abbreviation’ of the antero- ventral movement by means of the mouth forming directly in the anterior part, the former animal hemisphere of the embryo. This includes a spatial

TH E M O UTH , TH E AN U S , AN D TH E B L A S T O POR E 39



separation of the molecular mechanisms determin- ing the mouth from the site of gastrulation. It is clear from the developmental studies on ctenophores and cnidarians that the mouth and the blastopore have a common origin (Goldstein and Freeman,
1997). In both animals the blastopore gives rise to the single opening of the digestive cavity. Thus, it is not surprising that many genes which have been assigned to gastrulation and foregut development, for example brachyury, goosecoid, and forkhead, are expressed at the cnidarian and ctenophore blast- opore (Scholz and Technau, 2003; Martindale et al.,
2004; Matus et al., 2006a; Yamada et al., 2007). It is difficult, however, to dissect the role of these genes, since the blastopore has overlapping functions in mouth formation, axis determination, and germ layer specification.
It is important to point out that the site of gastru- lation has changed along the animal–vegetal egg axis in the stem lineage of the Bilateria. While cni- darians and ctenophores gastrulate at the animal pole, bilaterians, including acoels, gastrulate at the vegetal pole. Since bilaterians form their mouths at the animal hemisphere, this indicates an ancient separation of the determining factors of mouth formation and site of germ layer specification. The molecular separation of signalling centres might explain the variation of the relationships between mouth and blastopore in the Bilateria by a facili- tation of movements of signalling centres. The absence of molecular data from a broader range of taxa limits detailed conclusions at this time, but in animals in which the mouth is formed at a separate site from the blastopore (deuterostomy) (including asteroids, echinoids, hemichordates, and chaetognaths), the gene brachyury is expressed in both locations, indicating the spatial separation of a former common expression domain at the blastopore (Peterson et al., 1999; Shoguchi et al.,
1999; Takada et al., 2002). Furthermore, recent work indicates that the mouth in protostomes and deu- terostomes—although formed by variable devel- opmental processes—is homologous, based on the shared arrangement of the expression domains of goosecoid and brachyury (Arendt et al., 2001). A spe- cific mouth signalling system which would enable a detailed explanation of how the position of the mouth becomes specified in the Bilateria has not
yet been identified, but the dissociation of the blastopore from the position of the mouth already occurred in the bilaterian stem lineage.
The fate map of the acoel Neochildia fusca shows that the mouth is formed at a site different from the blastopore (Henry et al., 2000) and the mouth of Isodiametra pulchra is formed long after blast- opore closure (Ladurner and Rieger, 2000). Our own studies of brachyury and goosecoid expression in the acoel Convolutriloba longifissura supports the homology of the acoel mouth with the protostome and deuterostome mouth (Hejnol and Martindale,
2008a). The fate map of the acoel Neochildia fusca furthermore shows that the vegetal part of the embryo gets shifted in an antero-ventral direction by the increased proliferation of descendants of micromere 1a versus those of its ventral counter- part micromere 1b (Henry et al., 2000). This mirrors the observations made by fate map studies in spi- ralian taxa, like polyclads (Boyer et al., 1998), mol- luscs (Dictus and Damen, 1997; Hejnol et al., 2007), annelids (Ackermann et al., 2005), and nemerteans (Maslakova et al., 2004a) in which the dorsal side of the embryo proliferates more than the ventral regions (see also discussion in van den Biggelaar et al., 2002). This shifting of the vegetal part of the embryo by proliferation of dorsal cells might be an ancestral bilaterian feature which was lost in the deuterostome lineage where the position of the mouth is specified independently from the blastopore. Unfortunately, the relationship of the blastopore to the primary egg axis has not been investigated in a large number of ecdysozoan taxa and appears to be highly variable and obscured by yolk content.


4.8 Has the anal orifice evolved several times convergently?

If the mouth is homologous in all animals and was the earliest opening to the digestive cavity, when did the anal opening evolve and do all anal open- ings share a common ancestry? The outgroups of the Nephrozoa—ctenophores, cnidarians, acoels, and nemertodermatids—as well as several other key bilaterian taxa lack an anal opening. All Platyhelminthes possess only a mouth and the presence of one or more dorsal anal pores in the

40 AN I M AL EV O L UTI O N



branched gut system of some polyclads is a derived character (Ehlers, 1985). Gnathostomulids and some rotifers (e.g. Asplanchna) lack an anus and Xenoturbella, which in the current view forms the sister taxon to the Ambulacraria or to all remaining Deuterostomia (Bourlat et al., 2003, 2006; Perseke et al., 2007; Dunn et al., 2008), also possesses only one opening to the digestive cavity (Westblad, 1949).
This lack of an anus, however, can be inter- preted as either loss from a stem species with a through gut (as seems to be the case for the bra- chiopods and many parasitic forms) or that the anus has evolved later and independently in sev- eral lineages. Although it is not parsimonious that the anus evolved multiple times, its functional advantage and the differences in development and morphology could be evidence for an independent evolutionary origin (Beklemishev, 1969; Schmidt- Rhaesa, 2007).
The anus is morphologically very diverse between the protostome taxa. For example, gastro- trichs do not possess an ectodermal hindgut like most bilaterians; instead the anus is formed by a direct and often temporary connection of the endoderm to the outside (Ruppert, 1991b). Such a temporary anus is also present in Micrognathozoa (Kristensen and Funch, 2000) and in the gnatho- stomulid Haplognathia (Knauss, 1979). Another variation of anal morphology found in many lin- eages is a combined opening of the anus with the gonopore, a so-called cloaca. Cloacas are present in the ecdysozoan nematodes, nematomorphs, tar- digrades, and rotifers. Acoels and nemertoderma- tids lack an anus but have a male gonopore at the posterior end of the body formed by an involution of the ectoderm. Brachyury is expressed in the hatchling of the acoel Convolutriloba longifissura in the posterior ectoderm, which later gives rise to the adult male gonopore (Hejnol and Martindale,
2008a). This might indicate that the anus of some taxa, as in the Platyzoa and Ecdysozoa, might have been derived from a connection between the endo- derm and the ectoderm of the gonoduct. Thus the last common nephrozoan ancestor might have had only an antero-ventral mouth and a posterior male gonopore similar to what is found in acoels and nemertodermatids (Hejnol and Martindale, 2008a). Of the larger animal clades, the Trochozoa show
the highest morphological similarity of their anal openings: most of them possess an ectodermal hindgut, which ends in an opening that is separ- ate from the gonoducts. In most trochozoans the hindgut is formed after the mouth as a secondary ectodermal involution at a site separate from the blastopore. Despite this evidence for the homology of the trochozoan anus, it remains unclear if the anus is homologous in all bilaterians and thus part of the nephrozoan ground pattern. The inde- pendent evolution of the anus might have been a result of the extension of the body length in sev- eral lineages, since a blind gut is mostly present in smaller animals and parasitic forms. However, it is too early to draw a conclusion because import- ant nodes in metazoan phylogeny are still not resolved. It will be important to see if the Platyzoa, a taxon comprising Rotifera, Gnathostomulida, Platyhelminthes, and Gastrotricha, turns out to be a true monophyletic group or a long-branch arte- fact (see Chapter 6).


4.9 Conclusion

We are far from understanding the evolution of gastrulation mechanisms in the different animal lineages. As in phylogenetic systematics where broader taxon sampling helps to reconstruct phyl- ogeny, it is necessary to follow the same approach in comparative developmental biology to answer the questions raised from studies of a handful of arbi- trarily selected animal species. The new molecular approaches to deliver the necessary phylogenetic framework, as well as detailed analyses of the development of more animal taxa using modern cell lineage studies and molecular developmental approaches, will deliver insights into the evolution of the important organ systems.


4.10 Acknowledgements

We thank Tim Littlewood and Max Telford for the invitation to write this chapter and for organiz- ing the Royal Society meeting in London. Kevin Pang critically read the manuscript. The research was supported by the NSF grant AToL Protostome to MQM and by a grant of the German Science Foundation to AH.