ZOOLOGYZoology 107 (2004) 75–86

ARTICLE IN PRESS

*Correspondin

E-mail addres

0944-2006/$ - see

doi:10.1016/j.zoo

www.elsevier-deutschland.de/zool

Basiepidermal nervous system in Nemertoderma westbladi(Nemertodermatida): GYIRFamide immunoreactivity

Olga I. Raikovaa, Maria Reuterb,*, Margaretha K.S. Gustafssonb,Aaron G. Maulec, David W. Haltonc, Ulf Jondeliusd

aZoological Institute, Russian Academy of Sciences, St. Petersburg, RussiabDepartment of Biology, (Abo Akademi University, Artillerigatan 6, FIN-20520 (Abo, FinlandcParasitology Research Group, School of Biology and Biochemistry, the Queen’s University of Belfast, Belfast, UKdEvolutionary Biology Centre, Uppsala University, Uppsala, Sweden

Received 13 October 2003; accepted 23 December 2003

Abstract

The Nemertodermatida are a small group of microscopic marine worms. Recent molecular studies havedemonstrated that they are likely to be the earliest extant bilaterian animals. What was the nervous system (NS) ofa bilaterian ancestor like? In order to answer that question, the NS of Nemertoderma westbladi was investigated bymeans of indirect immunofluorescence technique and confocal scanning laser microscopy. The antibodies to aflatworm neuropeptide GYIRFamide were used in combination with anti-serotonin antibodies and phalloidin-TRITCstaining. The immunostaining revealed an entirely basiepidermal NS. A ring lying outside the body wall musculature atthe level of the statocyst forms the only centralisation, the ‘‘brain’’. No stomatogastric NS has been observed. TheGYIRFamide immunoreactive part of the ‘‘brain’’ is formed of loosely packed nerve fibres with multiple smallneurones and a few large ones. The peptidergic and aminergic patterns of the NS do not correspond to each other: theformer is more developed on the ventral side, the latter is more pronounced on the dorsal side. A pair of GYIRFamideimmunoreactive nerve cords innervates the ventral side of the animal, the mouth and the male genital opening. Thenemertodermatids studied to-date display no common NS pattern. Possible synapomorphies of the Acoelomorpha arediscussed. The study demonstrates that the nemertodermatid NS possesses a number of plesiomorphic features andappears more primitive than the NS in other worms, except the Xenoturbellida. The bilaterian ancestor supposedlypossessed only a basiepidermal nerve net and had no centralised brain-like structures and no stomatogastric NS.r 2004 Elsevier GmbH. All rights reserved.

Keywords: Nemertodermatida; Nervous system; Immunocytochemistry; GYIRFamide; Bilaterian ancestor

Introduction

The Nemertodermatida are a small group of micro-scopic marine worms. In the summer of 1927, OttoSteinb .ock and Erich Reisinger extracted a single smallworm from mud dredged off the east coast of Green-land. On this specimen Steinb .ock (1930-1931) based his

g author.

s: mreuter@ra.abo.fi (M. Reuter).

front matter r 2004 Elsevier GmbH. All rights reserved.

l.2003.12.002

description of Nemertoderma bathycola. Westblad(1937) gave a more thorough description of what heassumed were more mature specimens of N. bathycola.These animals turned out to be a distinct species, laternamed Nemertoderma westbladi by Steinb .ock (1938). Athird nemertodermatid, Meara stichopi, was foundsymbiotic in the intestine of the holothurian Stichopus

tremulus by Westblad (1949). The Nemertodermatidanow comprise nine species (Sterrer, 1998). The nemer-todermatids are easily recognisable by the presence

ARTICLE IN PRESSO.I. Raikova et al. / Zoology 107 (2004) 75–8676

of a statocyst with two statoliths and few parietalcells.

The phylogenetic position of the Nemertodermatidahas been discussed in detail by Lundin (2000) andLundin and Sterrer (2001). The Nemertodermatida wereformerly included into the Acoela within the phylumPlatyhelminthes (Steinb .ock, 1930-1931; Westblad, 1937,1949). Later, Karling (1940, 1974) classified them in theseparate taxon Nemertodermatida, chiefly owing to thepresence of a distinct gut lumen. Karling commented onthe anatomical simplicity of the nemertodermatids andconsidered them to be closest to the ‘‘turbellarianarchetype’’. The uniflagellar sperm structure, unusualfor flatworms, further supported this view (Tyler andRieger, 1975, 1977; Rieger et al., 1991). Similaritiesbetween Nemertodermatida and Acoela in epidermalciliation (Hendelberg and Hedlund, 1974; Tyler andRieger, 1977; Lundin, 1997), intestine organisation(Tyler and Rieger, 1977), glandular and sensorystructures (Ehlers, 1992) and the near absence ofextracellular matrix were used by Tyler and Rieger(1977) to argue for a sister-group relationship of the twotaxa. Later, Ehlers (1985) united the Acoela and theNemertodermatida in the taxon Acoelomorpha Ehlers,1984, within the Platyhelminthes.

Present morphological and ultrastructural evidencefavours the view that the Nemertodermatida and Acoelaare sister taxa, with the Nemertodermatida being themore plesiomorphic taxon of the Acoelomorpha (Tylerand Rieger, 1975, 1977; Ehlers, 1985; Smith et al., 1986;Lundin and Hendelberg, 1996; Lundin, 1997, 2000;Lundin and Sterrer, 2001). Surprisingly, the two groupshave been assigned widely separate positions in phylo-genetic hypotheses of the Platyhelminthes based onmolecular data (Carranza et al., 1997; Littlewood et al.,1999; Ruiz-Trillo et al., 1999). In these studies nemerto-dermatids fell within the bulk of rhabditophoranPlatyhelminthes, and separate from the Acoela, whichbranched as the first bilaterians. Quite recently, Jonde-lius et al. (2002) using 18S rDNA and mitochondrialnucleotide sequences from five nemertodermatids andmany other bilaterian species have re-evaluated thephylogenetic position of the Nemertodermatida. Theanalysis strongly supports a basal position within theBilateria for the Nemertodermatida as a sister group toall other bilaterian species except the Acoela. Despitethe basal position of both Nemertodermatida andAcoela, the clade Acoelomorpha was not retrieved(Jondelius et al., 2002). Phylogenetic analysis of myosinheavy chain type II sequences retrieved the Acoelomor-pha clade, though with weak support, and confirmedtheir basal position within the Bilateria (Ruiz-Trilloet al., 2002). Analysis of combined large and smallsubunit ribosomal RNA sequences published recentlyby Telford et al. (2003) support a basal position of theNemertodermatida. Therefore, though there still re-

mains some conflict of molecular data and morpholo-gical characters, there is little doubt that either theAcoela or the Nemertodermatida are the earliest extantbilaterian animals.

The unique position of the group in the phylogeny ofthe Bilateria justifies a thorough examination ofnemertodermatid morphology. The animals are likelyto possess some of the characters of the extinct ancestorof all bilaterian animals.

The organisation of the nervous system (NS) hasalways been considered a character of phylogeneticvalue (Haszprunar, 1996). The conserved neural char-acters identified in the brains of a variety of inverte-brates have been used to reconstruct phylogeneticrelationships (Strausfeld, 1998). Traditionally, manyauthors believed that the NS and the construction ofthe brain in the Acoela (and the Nemertodermatida)represented the earliest form of the bilaterian NS andthat they were inherited from a common Platyhelminth-like ancestor (Hyman, 1951; Beklemischev, 1963; Rei-singer, 1972; Ivanov and Mamkaev, 1973). Alterna-tively, some authors regarded the acoelomorph NS assecondarily reduced (Ax, 1996). The Platyhelminthes(including Catenulida and Rhabditophora) are charac-terised by a NS consisting of a bilobed ganglionic brainand a more or less regular orthogon: a ladder oflongitudinal cords interconnected by transversal com-missures. One pair of cords (ventral or lateral) is thestrongest. They are called ‘‘main cords’’ (Reuter andGustafsson, 2000). In addition, a developed stomato-gastric NS is observed (Reuter et al., 1998; Reuter andHalton, 2001). The ganglia and the nerve cords showstrong aminergic and peptidergic immunostaining, whilethe stomatogastric NS in all flatworms is mainlypeptidergic (Reuter and Halton, 2001). The Acoelomor-pha do not possess the same NS pattern (Raikova et al.,1998, 2000a, 2001; Reuter et al., 2001a, b), no aminergicor peptidergic brain ganglion has been observed and noimmunoreaction associated with the digestive parench-yma (no stomatogastric system).

At present, immunocytochemical methods (ICC) arewidely used to reveal different neuronal signal sub-stances (neuronally based molecules involved in inter-cellular communication), and to localise thecorresponding neurones in the nervous system ofinvertebrates (Reuter and Halton, 2001). Antibodies toneuronal signal substances such as 5-HT, FMRFamideand GYIRFamide have been shown to selectivelyimmunostain neurones in a wide array of helminthspecies (for reviews see Halton et al., 1994; Reuter andGustafsson, 2000) as well as in primitive bilaterians suchas acoels (Reuter et al., 2001a, b; Raikova et al., 2004),xenoturbellids (Raikova et al., 2000b) and nemertoder-matids (Raikova et al., 2000a).

5-HT (5-hydroxytryptamine), commonly known asserotonin, belongs to the biogenic amines, signalling

ARTICLE IN PRESSO.I. Raikova et al. / Zoology 107 (2004) 75–86 77

molecules with established neurotransmitter and neuro-hormonal roles. Serotonin is an evolutionarily con-served neurotransmitter, found in both invertebratesand vertebrates, and involved in locomotor andbehavioural roles. In mammals it occurs in both thecentral and peripheral nervous systems and is implicatedin a variety of physiological tasks, including learningand memory. 5-HT appears to be the dominantbiogenic amine in all of the flatworm taxa examined,and it serves a variety of functions, most notably that ofmyoexcitatory neurotransmission (Reuter and Halton,2001) and stimulation of adenylate cyclase (Mansour,1984). Serotonin also controls ciliary locomotionin turbellarians (Welsh and Williams, 1970; Sakharov,1989).

Neuropeptides are now known to be of ancient originwithin the nervous systems of invertebrates. Theregulatory peptides FMRFamide and GYIRFamideare members of the family of RFamide peptides (FaRP),all sharing the same C terminal RFamide sequence.Originally identified in mollusc ganglia (Price andGreenberg, 1977), the FaRP family of peptides havebeen found in every major metazoan phylum, fromcoelenterates to chordates (e.g. Schneider and Taghert,1988; Grimmelikhuijzen and Westfall, 1995; Aarnisaloand Panula, 1995). GYIRFamide has been first isolatedfrom an acidified ethanolic extract of the tricladturbellarians, Dugesia tigrina and Bdelloura candida

(Johnston et al., 1995, 1996). GYIRFamide has beenshown to induce motility in the liver fluke F. hepatica,increasing both the frequency and amplitude of musclecontractions (Graham et al., 1997), and to be morepotent than FMRFamide (Johnston et al., 1996).

We have previously studied 5-HT and FMRFamideimmunoreactivity in two species of Nemertodermatida(Raikova et al., 2000a). In M. stichopi (Nemertoderma-tida), two loose submuscular longitudinal bundles of5-HT immunoreactive (-IR) fibres and a basiepidermalnerve net were observed. Strands of FMRFamideimmunoreactive fibres following the 5-HT-IR fibrebundles were described. In N. westbladi a symmetricalring-shaped brain-like structure composed of 5-HT-IRcommissural fibres associated with a few cell bodies wasobserved. Only few data were obtained on the peptider-gic part of the NS using the anti-FMRFamide antibody(Raikova et al., 2000a). This study provides furtherinsight into the nemertodermatid NS structure usingantibodies against the flatworm neuropeptide GYIRFa-mide, which were proved to reveal peptidergic fibres inacoelomorph worms much better than anti-FMRFa-mide antibodies. To obtain a better picture of spatialrelationships between different parts of the NS, and therespective position of the NS and the body wallmusculature, the anti-GYIRFamide antibodies wereused in combination with anti-serotonin antibodiesand phalloidin-TRITC staining.

Material and methods

In July 2000, about 50 specimens of N. westbladi

Steinb .ock, 1938, were collected at a depth of 30–50mfrom muddy bottoms in the vicinity of KristinebergMarine Research Station at the Gullmar Fjord on theSwedish west coast. Collected worms were fixed inStefanini’s fixative (2% paraformaldehyde and 15%picric acid in 0.1M Na-phosphate buffer) at pH 7.6.They were stored for several days in fixative, then rinsedfor 24–48 h in 0.1M Na-phosphate buffer (pH 7.6)containing 20% sucrose and processed as whole-mountsin Eppendorf tubes. Prior to staining, the animals wereimmersed in phosphate-buffer saline containing 0.2%Triton X-100 (PBS-T). Non-specific antigens wereblocked with 2% bovine serum albumin (BSA) inPBS-T. Double-stainings were carried out according tothe indirect immunofluorescence method of Coons et al.(1955). Incubations were performed with a mixture ofguinea-pig antiserum against the flatworm (B. candida)neuropeptide GYIRFamide raised in guinea pigs (seeJohnston et al., 1995, 1996) and of rabbit anti-5-HT(INCSTAR) antiserum. Both primary antisera werediluted 1:400. As compared with our previous experi-ments (Raikova et al., 2000a), the incubation time wasnow increased to about 4–7 days at 10�C and the wormswere incubated in Eppendorf tubes on a shaker. Afterincubation with the primary antibodies, the animalswere rinsed 3� 5min in PBS-T and incubated for 1–2 hat room temperature with FITC-labelled goat anti-guinea pig (Cappel) and swine anti-rabbit TRITC(DAKO) secondary antibodies (dilution 1:30). Afterrinsing 3� 5min in PBS, the animals were mounted in50% glycerol–PBS and stored in the dark at –20�C.

Stainings with anti-GYIRFamide antibody only, fol-lowed by a 2 h incubation in phalloidin-TRITC (Sigma)diluted 1:100 in PBS were also performed.

As a control for specificity, some animals wereincubated for a week in PBS-T solution alone, withoutthe primary antibodies, then the secondary antibodieswere applied in a usual way. No staining of the nervoussystem or the statoliths was obtained in controls.However, in about a third of the animals dorsal glandcells were brightly stained by FITC-labelled goat anti-guinea pig antibodies. The same cells were also stainedby swine anti-rabbit TRITC antibodies, but to a lesserdegree. The cells stained by the secondary antibodieswere situated in the upper layers of the thick epidermis,at the dorsal Side of the animal, starting at about 1/4 ofthe body length from the anterior end and reaching allthe way to the posterior end (Fig. 4a). The staining ofthe dorsal glands did not obstruct the view on the‘‘brain’’ region of the animals, or on the longitudinalnerves which were situated deeper in the basiepidermalposition, but prevented us from observing the sensorycells, which are not described in the present study.

ARTICLE IN PRESSO.I. Raikova et al. / Zoology 107 (2004) 75–8678

The animals were examined in a confocal scanning lasermicroscope (CSLM) LEICA TCS 4D. Some specimenswere viewed from the dorsal side, some from the ventralside. Usually 15–25 optical sections 1–3mm thick wereobtained while scanning through the specimen. Theprojection option was used to make reconstructions fromseries of adjacent optical sections in a series (thus resultingin a thicker optical section). Figs. 2c–e, described in thecaptions as ‘‘general views’’ are projections from all theoptical sections in a series made throughout the specimenfrom the dorsal to the ventral aspect or vice versa. The filesobtained (with initial resolution of 512 pixels) wereprocessed with Adobe Photoshop 7.0 software. Only thecommands ‘‘crop’’, ‘‘image size’’, ‘‘mode’’ and ‘‘levels ofgrey’’ were used to avoid any distortion of the informationcontent of the image.

Fig. 1. Schematic diagram of the organisation of the

GYIRFamide-IR elements in the NS of the anterior end of

the body. View from the ventral side showing a ring-shaped

‘‘brain’’ consisting of a concentrated nerve net with numerous

small neurones (arrowheads). Note the presence of large

neurones (arrows) associated with the roots of ventral and

lateral longitudinal nerve cords. All the NS elements lie within

the epidermis. af, anterior nerve fibres; ep, epidermal layer; f,

nerve fibres; l, lateral longitudinal nerve cords; nr, nerve ring;

s, statocyst; v, ventral longitudinal nerve cords.

Results

N. westbladi is a small worm about 0.4–0.7mm in size.At the anterior end there are frontal glands and astatocyst with two statoliths, while at the posterior endthere is a male copulatory organ. The mouth is ventral;in adult specimens the mouth opening disappears,making it rather difficult to recognise the dorsal andthe ventral sides of the worm. However, two featureshelp in orientating the animals: in adults the statocyst isshifted slightly towards the dorsal side and the eggs arelocated dorsally.

Anti-GYIRFamide and anti-5HT immunostainingsreveal a NS composed of a ring-shaped anteriorcentralisation (the ‘‘brain’’), and two pairs of nervecords (ventral and lateral) starting from it in the caudaldirection (Figs. 1 and 2). Double stainings of nerves andmuscles show that all nerve elements lie basiepidermally,outside the body wall musculature (Figs. 1 and 2f). Thebasiepidermal ‘‘brain’’ forms a ring at a level slightlyposterior to the statocyst. Surprisingly, the stones of the

Fig. 2. Confocal scanning laser microscope images of double-stained

red (a–f); 5-HT pattern is shown in green (a–e); phalloidin-stained

projection of optical sections taken from the level of the ventral epid

extensive development of GYIRFamide-IR fibres on the ventral side

part of the ring (arrows). (b) Dorsal side of the same specimen, 12mdorsal to the statocyst to the level of the dorsal epidermis. Note th

nerve fibres going from the ‘‘brain’’ in anterior and posterior directi

optical sections. Note the presence of ventral and lateral longitudin

associated with the ventral cord is visible (arrow). (d) General view o

that 5-HT-IR ‘‘brain’’ rings are stronger dorsally and more comp

ventrally. (e) General view of the ‘‘brain’’, 40mm thick projection o

Note also that the ventral longitudinal nerve cords are composed o

some 5-HT-IR fibres as well. (f) Projection, 37 mm thick, of optical se

of the statocyst. Note that the GYIRFamide-IR ‘‘brain’’ ring lies wi

with phalloidin. af, anterior nerve fibres; ep, epidermal layer; f, ner

statocyst; v, ventral longitudinal nerve cords. Scale bars, 50mm.

statocyst are sometimes stained by GYIRFamide anti-body, therefore the statocyst is easily detected on opticalsections (Figs. 2a, e and f). The peptidergic, GYIRFa-mide-IR, part of the NS is more pronounced than theaminergic, 5-HT-IR, one. No cross-reactivity of thesetwo neuronal signal substances has been observed.

Brain-like structure

Double staining of the same specimen (Figs. 2a–e)reveals that the ‘‘brain’’ is composed of the wide ring ofloosely packed GYIRFamide-IR fibres lying alongside

specimens of N. westbladi. GYIRFamide pattern is shown in

muscles are shown in green (f). (a) Ventral side, 18mm thick

ermis to the level of the statocyst, including the latter. Note the

and the presence of large neurones associated with the ventral

m thick projection of optical sections taken from the level just

e predominance of 5-HT-IR fibres dorsally and the numerous

ons. (c) General view of the ‘‘brain’’, 35 mm thick projection of

al nerve cords starting from the ‘‘brain’’ ring. A large neurone

f the ‘‘brain’’, 24 mm thick projection of optical sections. Note

act than the GYIRFamide-IR ring, which is more developed

f optical sections. Note the anterior GYIRFamide-IR fibres.

f GYIRFamide-IR fibres only, while the lateral cords contain

ctions taken from the level of the ventral epidermis to the level

thin the epidermis, completely outside the muscle layer, stained

ve fibres; l, lateral longitudinal nerve cords; nr, nerve ring; s,

ARTICLE IN PRESSO.I. Raikova et al. / Zoology 107 (2004) 75–86 79

ARTICLE IN PRESSO.I. Raikova et al. / Zoology 107 (2004) 75–8680

the two more condensed 5HT-IR rings (anterior andposterior ones). The examination of serial opticalsections (Figs. 2a and b) shows that the GYIRFamide-IR ‘‘brain’’ fibres appear to be more prominent on theventral side of the animal (Fig. 2a), while the 5-HT-IRrings are stronger dorsally and laterally (Fig. 2b). Thepart of the ‘‘brain’’ ring that lies between the roots of theventral cords is the widest and it is predominantlycomposed of loose strands of GYIRFamide-IR fibres(Figs. 1, 2a and d). The ventral part of the ‘‘brain’’ ringhas a net-like appearance. The dorsal part of theGYIRFamide-IR ‘‘brain’’ ring seemingly contains al-most the same number of fibres as the ventral part of it,but dorsally the GYIRFamide-IR fibres form a narrowcompact band lying just in front of the more posterior5-HT-IR ring (Figs. 1 and 2b).

Several GYIRFamide-IR nerve fibres start from the‘‘brain’’ in an anterior direction, presumably innervatingthe sensory area around the openings of the frontalglands (Figs. 1, 2e and 3a). The nerves are arranged likethe ribs of a dome at the apical pole of the animal. Theanterior end is also innervated by a curtain of thin 5-HT-IR fibres starting from the dorsal part of theanterior 5-HT-IR ‘‘brain’’ ring (Fig. 2b). Dorsally,numerous thin 5-HT fibres, starting from the moreposterior ‘‘brain’’ ring in a caudal direction, form acontinuous layer at the base of the dorsal epidermis(Fig. 2b).

Nerve cords

Two pairs of longitudinal nerve cords start caudallyfrom the ‘‘brain’’ (Figs. 1, 2a and c–f). The lateral cordscontain several GYIRFamide-IR fibres intermingledwith 5-HT-IR fibres originating from the lateral parts ofthe lower 5-HT-IR ‘‘brain’’ ring (Figs. 2b, d and e). Thelateral cords could be followed only a short distancefrom the ‘‘brain’’, sometimes up to a quarter of the bodylength (Figs. 2a–e and 4a). The ventral cords containloosely packed GYIRFamide-IR fibres, but no 5-HT-IRones (Figs. 1, 2a, c and e). The proximal parts of theventral cords have a net-like structure (Figs. 2a, c–f and3b). In young animals the ventral cords reach theposterior end of the body and form an extensive fibrenetwork around, and posterior to, the mouth opening(Figs. 4a and b). In adults the mouth openingdisappears, and fibre networks were observed only inconnection with the male copulatory organ (see below).

Nerve cells and fibres

GYIRFamide and 5-HT immunoreactivity was ob-served in different sets of nerve cells and fibres. TheGYIRFamide immunoreactivity pattern is characterisedby the presence of several large neurones lying in pairs at

the bases of longitudinal nerve cords (Figs. 1, 2a and 3b)or associated with the ventral cords (Figs. 1 and 2c).These neurones seem to be unipolar (or bipolar) cellsabout 12–13 mm in diameter (Fig. 3b).

Very numerous small GYIRFamide-IR neurones liealong all the GYIRFamide-IR nerve fibres eitherincorporated in the ‘‘brain’’ ring or in the longitudinalcords (Figs. 1, 3a and c). These neurones are mostlybipolar, seldom multipolar cells, about 2–4 mm indiameter. By contrast, no neurones could be detectedin the 5-HT-IR ‘‘brain’’ ring, only varicosities could beseen (Fig. 3d). Some 5-HT-IR neurones, 4–5 mm indiameter, probably sensory cells, occur at the end ofnerve fibres going from the 5-HT-IR ‘‘brain’’ ring in ananterior direction (Fig. 2b).

Innervation of the male copulatory organ

The male copulatory organ is composed of theseminal vesicle opening through the male pore into theejaculatory duct (Fig. 5). The latter is a shallowinvagination, lined with epidermis, situated at theposterior end of the body. The ventral longitudinalnerve cords send branches to the male pore (Figs. 4c, dand 5). GYIRFamide-IR nerve fibres were observedwithin the epidermis of the ejaculatory duct (Figs. 4cand 5). The nerve fibres innervating the male openingare associated with relatively large (about 8 mm)unipolar GYIRFamide-IR neurones (Figs. 4c, d and 5).

Discussion

The NS of N. westbladi has been thoroughly describedby Westblad (1937). According to his light microscopicobservations, N. westbladi has a nerve layer at the baseof the epidermis, outside the musculature, which isthickened at the anterior end of the body forming a‘‘brain’’, shaped like a broad ring, thicker on the dorsalside. Westblad commented on its being entirely intrae-pithelial. A nerve layer in the anterior end looks like aneuropile, while at the posterior end it becomes a diffuseplexus. This description perfectly fits the patternobserved in the present study by immunocytochemicalmethods.

In addition to the basiepidermal NS in Nemertoder-

ma, Westblad (1937) and Meixner (1938) observed asmall group of ganglionic cells adjacent to the statocyst,lying in the parenchyma below the muscles. Thepresence of the ganglion adjacent to the statocyst isconfirmed by electron microscopic observations madeon other nemertodermatid species, Nemertoderma sp.(Ehlers, 1985, Fig. 60) and Flagellophora apelti (Tyler,2001). Neither 5-HT nor GYIRFamide innervation ofthe statocyst has been observed in this study. Therefore,

ARTICLE IN PRESS

Fig. 3. Details of the NS structure in N. westbladi. (a) Dorsal side, 14mm thick projection of optical sections showing the dorsal part

of GYIRFamide-IR ‘‘brain’’ ring and the anterior nerve fibres starting from it. (b) Ventral side of the GYIRFamide-IR ‘‘brain’’

ring, 16 mm thick projection. Note large paired IR neurone cell bodies lying at the base of longitudinal nerve cords (arrows) and

occurring also along the anterior nerve fibres. The GYIRFamide-IR ‘‘brain’’ fibres are more loosely packed on the ventral side.

(c–d) Detail of the ‘‘brain’’ ring in a double stained animal, 5.5 mm thick projection of optical sections showing the absence of cross-

reactivity of GYIRFamide and 5-HT. (c) Anti-GYIRFamide immunostaining. Note the presence of numerous small IR neurones.

(d) Same place, anti-5-HT immunostaining. Instead of nerve cells, only numerous varicosities can be observed. A halo of 5-HT-IR

fibres surrounds the ‘‘brain’’ ring. af, anterior nerve fibres; ep, epidermal layer; f, nerve fibres; l, lateral longitudinal nerve cords; nr,

nerve ring; s, statocyst; v, ventral longitudinal nerve cords. Scale bars, 50 mm.

O.I. Raikova et al. / Zoology 107 (2004) 75–86 81

the statocyst must be innervated by nerve fibres withsome other neuronal signal substances, which accountsfor the discrepancies between NS patterns as revealed byimmunocytochemistry and traditional histologicalstains. Exactly the same situation occurs with the brain

ganglion adjacent to the statocyst in acoels: though ithas been repeatedly visualized by traditional histologicalstainings and electron microscopy, it is revealed neitherby anti-GYIRFamide, nor by anti-FMRFamide, nor byanti-5-HT antibodies (Raikova et al., 2004). The acoels

ARTICLE IN PRESS

Fig. 4. GYIRFamide pattern in the caudal end of N. westbladi. (a) Ventro-lateral view of a young animal, 16mm thick projection of

optical sections showing the course of GYIRFamide-IR ventral longitudinal cords. The lateral cords could be followed only up to a

quarter of the body length. The dorsal glands are non-specifically stained by secondary antibodies. (b) Detail of the posterior end

of the same specimen, 9mm thick projection of optical sections through the ventral epidermis showing an intricate network of

GYIRFamide-IR fibres interconnecting the ventral cords. (c) Posterior end of an animal, general view, 16mm thick projection of

optical sections. Note that fibres coming from the longitudinal nerve cords lie in the epidermis of the male ejaculatory duct and end

at terminal groups of neurones (arrows) lying close to the male pore. (d) Projection of optical sections, 6 mm thick, showing several

nerve fibres surrounding the male pore. dg, dorsal glands; ed, male ejaculatory duct; ep, epidermal layer; f, nerve fibres; l, lateral

longitudinal nerve cords; m, mouth; mp, male pore; n, neurone; nr, nerve ring; s, statocyst; sv, seminal vesicle; v, ventral longitudinal

nerve cords. Scale bars, 50mm.

O.I. Raikova et al. / Zoology 107 (2004) 75–8682

and the nemertodermatids seem to share the same(synapomorphic or symplesiomorphic?) character: astatocyst ganglion with unknown neuronal signalsubstance(s). The part of the brain in nemertodermatids

and acoels connected with the statocyst has beentraditionally described as ‘‘endonal brain’’ (Steinb .ock,1930-1931, 1938; Westblad, 1937, 1948; Reisinger, 1972;Ivanov and Mamkaev, 1973) and was supposed to be

ARTICLE IN PRESS

Fig. 5. Schematic diagram of the organisation of the

GYIRFamide-IR elements innervating the male copulatory

organ at the posterior end of the body. The nerve fibres (f)

associated with the ventral longitudinal nerve cords lie within

the epidermis (ep) of the male ejaculatory duct (ed). Note the

presence of large unipolar neurones (n) innervating the male

pore (mp) that leads to the seminal vesicle (sv).

O.I. Raikova et al. / Zoology 107 (2004) 75–86 83

homologous with the cerebral ganglion in the Platyhel-minthes. However, the homology seems questionable, asthe cerebral ganglion in the Platyhelminthes, unlike thestatocyst ganglion in the Acoelomorpha, is alwaysrevealed by the above-mentioned immunostainings.

Peculiar is the fact that the statoliths of the statocystare often stained by the anti-GYIRFamide antibody. Insome cases the stones are stained quite strongly,especially on the surface (Figs. 2e and f), in other casesthey are stained weakly (Fig. 2a) or not at all (Figs. 2cand d). No staining of the statocyst stones has beenobserved in controls (where the animals were incubatedwith secondary antibodies only); neither were thestatoliths stained with anti-5-HT antibody or phalloidin.Therefore, unusual as it may seem, one must assumethat the stones of the statocyst display some sort ofspecific GYIRFamide-like immunoreactivity. It is inter-esting to note that in a number of acoels treatedaccording to the same procedure the statocysts nevershowed any staining, even after incubation with anti-GYIRFamide antibody for over a week (Raikovaet al., 2004).

Our data confirm Westblad’s (1937) observation onthe absence of a stomatogastric NS. Juvenile specimensof N. westbladi have a developed gut with a ventralmouth opening. However, no innervation of the gut hasbeen visualised. As peptidergic immunoreactivity char-acterises the stomatogastric NS of rhabditophoranflatworms, the absence of peptidergic immunoreactivityin the gut of N. westbladi is striking. A similar situationhas been described in Xenoturbella westbladi (Raikovaet al., 2000b). The absence of a stomatogastric NSwould argue against the origin of the Nemertodermatida(as well as the Xenoturbellida and the Acoela) fromanimals with a developed gut and a stomatogastric NS.

The 5-HT-IR part of the NS in N. westbladi is moredeveloped dorsally, while the GYIRFamide-IR part is

more developed ventrally. The predominance of thedorsal part of the NS seems to be characteristic of theAcoela (Raikova et al., 1998; Reuter et al., 2001a, b;Raikova et al., 2004). In the Platyhelminthes (Catenu-lida and Rhabditophora) studied to date, dorsal cordsare never the strongest. The main cords (MCs) arelateral in Catenulida and mostly ventral (or ventro-lateral) in Rhabditophora (see Reuter and Gustafsson,2000; Reuter and Halton, 2001). This observation is inaccordance with a separate origin of the Acoela andthe Platyhelminthes. As to the Nemertodermatida, inM. stichopi the main submuscular nerve bundles withintermingled 5TH-IR and FMRFamide-IR fibres occu-py a lateral position (Raikova et al., 2000a), while inN. westbladi the single pair of GYIRFamide-IR cordslies within the epidermis on the ventral side. However,the 5-HT-IR pattern is more pronounced on the dorsalside in both species of the Nemertodermatida studied todate. This feature may constitute a synapomorphy of theAcoelomorpha, but other nemertodermatid speciesshould first be checked in this respect.

Another possible synapomorphy of the Acoelomor-pha is the occurrence of large (12–13 mm in diameter)neurones in the GYIRFamide-IR pattern. InN. westbladi, large paired GYIRFamide-IR neuroneswere found to be associated with the basal parts of thelongitudinal cords. Similar neurones have been de-scribed in a number of acoels (Reuter et al., 2001a, b;Raikova et al., 2001, 2003), and their number andposition seems to be a character of phylogenetic value(Raikova et al., 2003).

Peculiar is the apparent absence of neuronal cellbodies in the 5-HT-IR ‘‘brain’’ ring of N. westbladi.Only varicosities could be seen. It is possible that 5-HTimmunoreactivity does not occur in neuronal cell bodiesand is confined to axonal and synaptic regions. This iscommonly seen in higher animals where 5-HT isrecycled and not transported from the soma to thesynapse. A similar situation occurs in the 5-HT-IR partof the NS of the Acoela, where mostly varicosities arevisible along the cords, but in acoels a few 5-HT-IRneuronal cell bodies could be seen associated with thebrain (Reuter et al., 2001a, b; Raikova et al., 2004).On the other hand, in another nemertodermatid,M. stichopi, large 5-HT-IR neuronal cell bodies (about10 mm in diameter) occur in the basiepidermal nerve netand at the beginning of parenchymal bundles of nervefibres (Raikova et al., 2000a).

In N. westbladi the GYIRFamide-IR fibres within the‘‘brain’’ ring and the longitudinal cords are very looselypacked. Actually, the peptidergic NS pattern is com-posed of an irregular nerve net with numerous smallneurones. It shows a much lower degree of concentra-tion than the NS in acoels and rhabditophoran flat-worms does. Only the NS of X. westbladi is even lesscentralized: it consists only of a thick basiepidermal

ARTICLE IN PRESSO.I. Raikova et al. / Zoology 107 (2004) 75–8684

nerve net (Raikova et al., 2000b). Xenoturbella wasassigned widely different positions on the bilaterian treeranging from the sister-group to the Bilateria (Ehlersand Sopott-Ehlers, 1997a, b), to a bivalve mollusc(Nor!en and Jondelius, 1997) and even to a deuteros-tome, related to hemichordates and echinoderms (Rei-singer, 1960; Bourlat et al., 2003). The idea thatXenoturbella could represent a model of ancestry ofboth deuterostomes and protostomes (Gee, 2003) seemschallenging.

The major centralisation of nerves in N. westbladi isshaped like a ring. Here a question arises: has such aring been a part of the nervous system of thehypothetical common bilaterian ancestor? If we considerXenoturbella to be genuinely primitive, as its morphol-ogy suggests, then we should not expect the ancestor tohave any anterior nerve centralisations. The absence ofan anterior brain-like structure in M. stichopi (Nemer-todermatida) further supports this idea (Raikova et al.,2000a). On the other hand, one might find someevidence to the contrary. A ring-shaped ‘‘brain’’structure occurs in another nemertodermatid, F. apelti

(Tyler, 2001) as well as in a number of acoels (see Ivanovand Mamkaev, 1973, for references), including thosebelieved to be more basal, like Diopisthoporus (West-blad, 1940; Tyler, 2001; and our unpublished observa-tions on 5-HT immunoreactivity in D. longitubus).A ring-shaped brain is also observed in more basalrepresentatives of derived families of acoels, forexample, in Paraphanostoma cycloposthium within theChildiidae (Raikova et al., 2004). Therefore, if we acceptthe monophyly of the Acoelomorpha, we may supposethat the ring-shaped anterior nerve centralisation was afeature of the acoelomorph ancestor.

The NS patterns in nemertodermatids studied to dateare very different from each other. Acoels also showconsiderable variety in the form of their nervous system(Reuter et al., 2001a, b; Raikova et al., 2004). The NSmust have undergone independent evolution withinthese groups, acquiring specific apomorphic features.It seems undeniable that the NS in the Acoela andparticularly in the Nemertodermatida still displays anumber of plesiomorphic features, in common withthe Xenoturbellida and with the hypothetical commonbilaterian ancestor. These features are: basiepidermalposition of the NS and low degree of nerve concentra-tion, the NS resembling a loosely packed nerve net.The bilaterian ancestor supposedly possessed only abasiepidermal nerve net. The nerve net probablycontained intermingled peptidergic and aminergicfibres and neurones, but no neurones and fibreswith co-existing neuronal signal substances. The bilater-ian ancestor may have possessed a gut, but nostomatogastric NS.

The new insight in the NS structure obtained byimmunocytochemical methods is in line with the basal

position of the Nemertodermatida and the Acoela in thebilaterian tree.

Acknowledgements

Thanks are extended to the staff of the marinebiological station at Kristineberg (Sweden) for theirhelp with collecting the material. Financial support wasreceived from the Wenner-Gren Foundation grant toO. Raikova and from the Russian Basic ResearchFoundation grant 02-04-48583. We thank the ResearchInstitute of the (Abo Akademi University Foundation,Svenska Kulturfonden, Magnus Ehrnrooths foundationfor grants to M. Gustafsson and the Swedish ResearchCouncil for a grant to U. Jondelius.

References

Aarnisalo, A.A., Panula, P., 1995. Neuropeptide FF contain-

ing efferent projections from the media hypothalamus

of rat: A Phaseolus vulgaris leucoagglutinin study. Neu-

roscience 65, 175–192.

Ax, P., 1996. Multicellular Animals. A New Approach to the

Phylogenetic Order in Nature. Springer, Berlin.

Beklemischev, V.N., 1963. On the relationship of the

Turbellaria to the other groups of the animal kingdom.

In: Dougherty, E.C. (Ed.), The Lower Metazoa. University

of California Press, Berkeley, pp. 234–244.

Bourlat, S.J., Nielsen, C., Lockyer, A.E., Littlewood, D.T.J.,

Telford, M.J., 2003. Xenoturbella is a deuterostome that

eats molluscs. Nature 424, 925–928.

Carranza, S., Baguna, J., Ruitort, M., 1997. Are the

Platyhelminthes a monophyletic group? An assessment

using 18S rDNA sequences. Mol. Biol. Evol. 14, 485–497.

Coons, A.H., Leduc, E.H., Conolly, J.M., 1955. Studies on

antibody production. I. A method for the histochemical

demonstration of specific antibody and its application to a

study of the hyperimmune rabbit. J. Exp. Med. 102,

49–60.

Ehlers, U., 1985. Das phylogenetische System der Plathel-

minthes. G. Fischer, Stuttgart.

Ehlers, U., 1992. Frontal glandular and sensory structures in

Nemertoderma (Nemertodermatida) and Paratomella

(Acoela): ultrastructure and phylogenetic implications for

the monophyly of the euplathelminthes (plathelminthes).

Zoomorphology 112, 227–236.

Ehlers, U., Sopott-Ehlers, B., 1997. Ultrastructure of epider-

mal musculature of Xenoturbella bocki, the adelphotaxon of

the Bilateria. Zoomorphology 117, 71–79.

Gee, H., 2003. You aren’t what you eat. Nature 424, 885–886.

Graham, M.K., Fairweather, I., McGeown, J.G., 1997. The

effects of FaRPs on the motility of isolated muscle strips

from the liver fluke, Fasciola hepatica. Parasitology 114,

455–465.

Grimmelikhuijzen, C.J., Westfall, J.A., 1995. The nervous

systems of cnidarians. In: Breidbach, O., Kutsch, W. (Eds.),

ARTICLE IN PRESSO.I. Raikova et al. / Zoology 107 (2004) 75–86 85

The Nervous Systems of Invertebrates — A Comparative

Approach. Experientia 72, Birkhauser. Verlag, Basel, pp. 7–24.

Halton, D.W., Shaw, C., Maule, A.G., Smart, D., 1994.

Regulatory peptides in helminth parasites. Adv. Parasitol.

34, 163–227.

Haszprunar, G., 1996. Plathelminthes and Plathelminthomor-

pha—paraphyletic taxa. J. Zool. Syst. Evol. 34, 41–48.

Hendelberg, J., Hedlund, K.O., 1974. On the morphology of

the epidermal ciliary rootlet system of the acoelous

turbellarian Childia groenlandica. Zoon 2, 13–24.

Hyman, L.H., 1951. The Invertebrates. II. Platyhelminthes

and Rhynchocoela. The Acoelomate Bilateria. McGraw-

Hill, New York.

Ivanov, A.V., Mamkaev, Y.V., 1973. Turbellaria, Their

Origin and Evolution. Phylogenetic Considerations. Nau-

ka, Leningrad (in Russian).

Johnston, R.N., Shaw, C., Halton, D.W., Verhaert, P.,

Bagun"a, J., 1995. GYIRFamide: a novel FMRFamide-

related peptide (FaRP) from the triclad turbellarian,

Dugesia tigrina. Biochem. Biophys. Res. Commun. 209,

689–697.

Johnston, R.N., Shaw, C., Halton, D.W., Verhaert, P., Blair,

K.L., Brennan, G.P., Price, D.A., Anderson, P.A.V., 1996.

Isolation, localization and bioactivity of the FMRFamide-

related neuropeptides GYIRFamide and YIRFamide from

the marine turbellarian Bdelloura candida. J. Neurochem.

67, 814–821.

Jondelius, U., Ruiz-Trillo, I., Bagun"a, J., Ruitort, M., 2002.

The Nemertodermatida are basal bilaterians and not

members of the Platyhelminthes. Zool. Scr. 31, 201–215.

Karling, T.G., 1940. Zur Morphologie und Systematik der

Alloeocoela Cumulata and Rhabditophora Lecithophora

(Turbellaria). Acta Zool. Fenn. 26, 1–160.

Karling, T.G., 1974. On the anatomy and affinities of the

turbellarian orders. In: Riser, N.W., Morse, M.P. (Eds.),

Biology of the Turbellaria. McGraw-Hill, New York,

pp. 1–16.

Littlewood, D.T.J., Rohde, K., Clough, K.A., 1999. The

interrelationships of all major groups of Platyhelminthes:

phylogenetic evidence from morphology and molecules.

J. Linn. Soc. 66, 75–114.

Lundin, K., 1997. Comparative ultrastructure of the epidermal

ciliary rootlets and associated structures in species of the

Nemertodermatida and Acoela (Plathelminthes). Zoomor-

phology 117, 81–92.

Lundin, K., 2000. Phylogeny of the Nemertodermatida

(Acoelomorpha, Platyhelminthes). A cladistic analysis.

Zool. Scr. 29, 65–74.

Lundin, K., Hendelberg, J., 1996. Degenerating epidermal

bodies (‘‘pulsatile bodies’’) in Meara stichopi (Plathel-

minthes, Nemertodermatida). Zoomorphology 116, 1–5.

Lundin, K., Sterrer, W., 2001. The Nemertodermatida. In:

Littlewood, D.T.J., Bray, R.A. (Eds.), Interrelationships of

the Platyhelminthes. Taylor & Francis, London, pp. 24–27.

Mansour, T.E., 1984. Serotonin receptors in parasitic worms.

Adv. Parasitol. 23, 1–36.

Meixner, J., 1938. Turbellaria (Strudelw .urmer). I. Allgemeiner

Teil. In: Grimpe, G., Wagler, E., Remane, A. (Eds.), Die

Tierwelt der Nord- und Ostsee. 33/IVb Lief., Akad. Verlag,

Leipzig, pp. 1–146.

Nor!en, M., Jondelius, U., 1997. Xenoturbella’s molluscan

relatives. Nature 390, 31–32.

Price, D.A., Greenberg, M.J., 1977. Structure of a molluscan

cardioexcitatory neuropeptide. Science 197, 670–671.

Raikova, O.I., Reuter, M., Kotikova, E., Gustafsson, M.K.S.,

1998. A commissural brain! The pattern of 5-HT immunor-

eactivity in Acoela (Plathelminthes). Zoomorphology 118,

69–77.

Raikova, O.I., Reuter, M., Jondelius, U., Gustafsson, M.K.S.,

2000a. The brain of Nemertodermatida. As revealed by

anti-5-HT and anti-FMRFamide immunostainings. Tissue

Cell 32, 358–365.

Raikova, O.I., Reuter, M., Jondelius, U., Gustafsson, M.K.S.,

2000b. An immunocytochemical and ultrastructural study

of the nervous and muscular systems of Xenoturbella

westbladi (Bilateria inc. Sed.). Zoomorphology 120,

107–118.

Raikova, O.I., Reuter, M., Justine, J.-L., 2001. Contributions

to the phylogeny and systematics of the Acoelomorpha. In:

Littlewood, D.T.J., Bray, R.A. (Eds.), Interrelationships of

the Platyhelminthes. Taylor & Francis, London, pp. 13–23.

Raikova, O.I., Reuter, M., Gustafsson, M.K.S., Maule, A.G.,

Halton, D.W., Jondelius, U., 2004. Evolution of the

nervous system in Paraphanostoma (Acoela). Zool. Scr.

33, 71–88.

Reisinger, E., 1960. Was ist Xenoturbella? Z. Wiss. Zool. 164,

188–198.

Reisinger, E., 1972. Die Evolution des Orthogons der Spiralier

und das Archicoelomatenproblem. Z. Zool. Syst. Evolu-

tionsforsch. 10, 1–43.

Reuter, M., Halton, D.W., 2001. Comparative neurobiology

of Platyhelminthes. In: Littlewood, D.T.J., Bray, R.A.

(Eds.), Interrelationships of the Platyhelminthes. Taylor &

Francis, London, pp. 239–249.

Reuter, M., Gustafsson, M.K.S., 2000. Platyhelminthes. In:

Dorn, A. (Ed.), Reproductive Biology of Invertebrates.

Wiley, Chichester, New York, Weinheim, Brisbane, Singa-

pore, Toronto, pp. 73–103.

Reuter, M., M.antyl.a, K., Gustafsson, M.K.S., 1998. Organi-

zation of the orthogon — main and minor nerve cords.

Hydrobiologia 383, 175–182.

Reuter, M., Raikova, O.I., Gustafsson, M.K.S., 2001a.

Patterns in the nervous and muscle systems in lower

flatworms. Belg. J. Zool. 31, 47–53.

Reuter, M., Raikova, O.I., Jondelius, U., Gustafsson, M.K.S.,

Halton, D.W., Maule, A.G., Shaw, C., 2001b. Organisation

of the nervous system in the Acoela: an immunocytochem-

ical study. Tissue Cell 33, 119–128.

Rieger, R.M., Tyler, S., Smith III, J.P.S., Rieger, G., 1991.

Platyhelminthes: Turbellaria. In: Harrison, F.W., Bogitsh,

B.J. (Eds.), Microscopic Anatomy of Invertebrates, Vol. 3.

Platyhelminthes and Nemertinea. Wiley-Liss, New York,

pp. 7–140.

Ruiz-Trillo, I., Ruitort, M., Littlewood, D.T.J., Herniou,

E.A., Baguna, J., 1999. Acoel flatworms: earliest extant

bilaterian metazoans, not members of platyhelminthes.

Science 283, 1919–1923.

Ruiz-Trillo, I., Paps, J., Loukota, M., Ribera, C., Jondelius,

U., Bagun"a, J., Riutort, M., 2002. A phylogenetic analysis

of myosin heavy chain type II sequences corroborates that

ARTICLE IN PRESSO.I. Raikova et al. / Zoology 107 (2004) 75–8686

Acoela and Nemertodermatida are basal bilaterians. PNAS

99, 11246–11251.

Sakharov, D.A., 1989, Integrative function of serotonin

common to distantly related invertebrate animals. In:

Gustafsson, M.K.S., Reuter, M. (Eds.), The Early Brain.(Abo Akademi Press, (Abo, pp. 73–88.

Schneider, L.E., Taghert, P.H., 1988. Isolation and character-

isation of a Drosophila gene that encodes multiple

neuropeptides related to Phe-Met-Arg-Phe-NH2 (FMRFa-

mide). Proc. Natl. Acad. Sci. 85, 1993–1997.

Smith, J.P.S., Tyler, S., Rieger, R., 1986. Is the Turbellaria

polyphyletic? Hydrobiologia 132, 13–21.

Steinb .ock, O., 1930–1931. Ergebnisse einer von E. Reisinger &

O. Steinb .ock mit Hilfe des Rask-Ørsted Fonds durchgef-

.uhrten Reise in Gr .onland 1926. 2. Nemertoderma bathycola

nov. gen. nov. spec., eine eigenartige Turbellarie aus der

Tiefe der Diskobay; nebst einem Beitrag zur Kenntnis des

Nemertinenepithels. Vidensk. Medd. Dan. Naturhist.

Foren. 90, 47–84.

Steinb .ock, O., 1938. .Uber die Stellung der Gattung Nemerto-

derma Steinb .ock im System der Turbellarien. Acta Soc.

Fauna Flora Fennica 62, 1–28.

Sterrer, W., 1998. New and known Nemertodermatida

(Platyhelminthes-Acoelomorpha)—a revision. Belgian

J. Zool. 128, 55–92.

Strausfeld, N.J., 1998. Crustacean–insect relationships: the use

of brain characters to derive phylogeny amongst segmented

invertebrates. Brain Behav. Evol. 52, 186–206.

Telford, M.J., Lockyer, A.E., Cartwright-Finch, C., Little-

wood, D.T.J., 2003. Combined large and small subunit

ribosomal RNA phylogenies support a basal position

of the acoelomorph flatworms. Proc. R. Soc. Lond. (B)

270, 1077–1083.

Tyler, S., 2001. The early worm: origins and relationships of

the lower flatworms. In: Littlewood, D.T.J., Bray, R.A.

(Eds.), Interrelationships of the Platyhelminthes. Taylor &

Francis, London, pp. 3–12.

Tyler, S., Rieger, R.M., 1975. Uniflagellate spermatozoa in

Nemertoderma (Turbellaria) and their phylogenetic signifi-

cance. Science 188, 730–732.

Tyler, S., Rieger, R.M., 1977. Ultrastructural evidence for the

systematic position of the Nemertodermatida (Turbellaria).

Acta Zool. Fenn. 154, 193–207.

Welsh, J.H., Williams, L.D., 1970. Monoamine-containing

neurones in planaria. J. Comp. Neurol. 138, 103–116.

Westblad, E., 1937. Die Turbellarien-Gattung

Nemertoderma Steinb .ock. Acta Soc. Fauna Flora Fennica

60, 45–89.

Westblad, E., 1940. Studien uber skandinavische Turbellaria

Acoela. II. Ark. Zool. 32A, 1–28.

Westblad, E., 1948. Studien .uber skandinavische Turbellaria

Acoela. V. Ark. Zool. 41A, 1–82.

Westblad, E., 1949. On Meara stichopi (Bock) Westblad, a new

representative of Turbellaria Archoophora. Ark. Zool. 1,

43–57.