Academia.eduAcademia.edu

Neuroanatomy of

Journal of Morphology

The position of Tardigrada in the animal tree of life is a subject that has received much attention, but still remains controversial. Whereas some think tardigrades should be categorized as cycloneuralians, most authors argue in favor of a phylogenetic position within Panarthropoda as a sister group to Arthropoda or Arthropoda + Onychophora. Thus far, neither molecular nor morphological investigations have provided conclusive results as to the tardigrade sister group relationships. In this article, we present a detailed description of the nervous system of the eutardigrade Halobiotus crispae, using immunostainings, confocal laser scanning microscopy, and computer‐aided three‐dimensional reconstructions supported by transmission electron microscopy. We report details regarding the structure of the brain as well as the ganglia of the ventral nerve cord. In contrast to the newest investigation, we find transverse commissures in the ventral ganglia, and our data suggest that the brain i...

JOURNAL OF MORPHOLOGY 273:1227–1245 (2012) Neuroanatomy of Halobiotus crispae (Eutardigrada: Hypsibiidae): Tardigrade Brain Structure Supports the Clade Panarthropoda Dennis K. Persson,1,2* Kenneth A. Halberg,2 Aslak Jørgensen,3 Nadja Møbjerg,2 and Reinhardt M. Kristensen1 1 Invertebrate Department, Natural History Museum of Denmark, University of Copenhagen, Universitetsparken 15, DK-2100 Copenhagen Ø, Denmark 2 Department of Biology, University of Copenhagen, Universitetsparken 13, DK-2100 Copenhagen Ø, Denmark 3 Laboratory of Molecular Systematics, Natural History Museum of Denmark, University of Copenhagen, Sølvgade 83, DK-1307 Copenhagen K, Denmark ABSTRACT The position of Tardigrada in the animal tree of life is a subject that has received much attention, but still remains controversial. Whereas some think tardigrades should be categorized as cycloneuralians, most authors argue in favor of a phylogenetic position within Panarthropoda as a sister group to Arthropoda or Arthropoda 1 Onychophora. Thus far, neither molecular nor morphological investigations have provided conclusive results as to the tardigrade sister group relationships. In this article, we present a detailed description of the nervous system of the eutardigrade Halobiotus crispae, using immunostainings, confocal laser scanning microscopy, and computer-aided three-dimensional reconstructions supported by transmission electron microscopy. We report details regarding the structure of the brain as well as the ganglia of the ventral nerve cord. In contrast to the newest investigation, we find transverse commissures in the ventral ganglia, and our data suggest that the brain is partitioned into at least three lobes. Additionally, we can confirm the existence of a subpharyngeal ganglion previously called subesophagal ganglion. According to our results, the original suggestion of a brain comprised of at least three parts cannot be rejected, and the data presented supports a sister group relationship of Tardigrada to 1) Arthropoda or 2) Onychophora or 3) Arthropoda 1 Onychophora. J. Morphol. 273:1227–1245, 2012. Ó 2012 Wiley Periodicals, Inc. KEY WORDS: neuroanatomy; tardigrade; phylogeny INTRODUCTION Tardigrades are small invertebrates predominantly found in mosses and lichens but also on marine macroalgae and between sand grains (Bertolani, 1982; McInnes, in press). They exhibit many autapomorphic features and were categorized as a phylum, Tardigrada, by Ramazzotti and Maucci (1983). When subjected to adverse environmental conditions, they may endure by means of i) active regulating processes or ii) entering diapause or a stress tolerant state called cryptobiosis (Møbjerg et al., 2011). Cryptobiosis enables tardigrades to survive severe physical stress, and conÓ 2012 WILEY PERIODICALS, INC. sequently, they occupy some of the most inhospitable habitats (Renaud-Mornant, 1975; Dastych and Kristensen, 1995; Pugh and McInnes, 1998; Wright, 2001; Møbjerg et al., 2007; Halberg et al., 2009b; Persson et al., 2011). The position of Tardigrada in the animal tree of life is a subject that has received much attention but remains controversial. Ever since their discovery in the middle of the 18th century, their position in the tree of life has been debated, and affiliations to Rotifera, Pentastomida, Onychophora, Nematoda, and Arthropoda have been suggested (Dujardin, 1851; Plate, 1889; Marcus, 1929; Crowe et al., 1970; Dewel and Clark, 1973; Ramsköld and Hou, 1991). Most evidence suggests one of the two large, species-rich and economically important groups Nematoda or Arthropoda as the closest relatives to Tardigrada. Therefore, the position of tardigrades has been at the front of the ongoing discussion on metazoan systematics. The two most supported theories are that tardigrades either belong to Cycloneuralia, as a close relative to nematodes (Crowe et al., 1970, Dewel and Clark, 1973; Ruppert and Barnes, 1994), or to Panarthropoda, along with arthropods and onychophorans (Baccetti and Rosati, 1971; Bussers and Jeuniaux, 1973; Contract grant sponsor: Danish Carlsberg Foundation; Contract grant sponsor: Danish Natural Science Research Council; Contract grant sponsor: National Science Foundation under the AToL program. *Correspondence to: Dennis Krog Persson, Invertebrate Department, Natural History Museum of Denmark, University of Copenhagen, Universitetsparken 15, DK-2100 Copenhagen Ø, Denmark. E-mail: [email protected] Received 16 November 2011; Revised 26 April 2012; Accepted 27 May 2012 Published online 18 July 2012 in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/jmor.20054 1228 D.K. PERSSON ET AL. Greven, 1982; Kristensen, 1976, 1981; Møbjerg and Dahl, 1996; Nielsen, 2001). Molecular investigations seem to support both views (Garey et al., 1996; Giribet et al., 1996; Aguinaldo et al., 1997; Garey, 2001; Mallatt et al., 2004; Dunn et al., 2008; Rota-Stabelli et al., 2010; Campbell et al., 2011), whereas morphological data more often support the view that tardigrades are related to onychophorans/ arthropods (Schmidt-Rhaesa et al., 1998; Budd, 2001). Importantly, recent papers emphasize that the molecular data are not conclusive as to whether tardigrades are more closely related to arthropods and onychophorans or to the nematodes and nematomorphs (Dunn et al., 2008; Edgecombe, 2010; Edgecombe et al., 2011; Campbell et al., 2011). Phylogenomic analyses have yielded different results depending on which substitution model that was used to analyze the data (Dunn et al., 2008). In addition, morphological investigations have generated contradictory results, inferring an inconsistency in the phylogenetic analyses (Dewel and Clark 1973; Kristensen and Higgins, 1984a,b; Dewel et al., 1993; Dewel and Dewel, 1996; Zantke et al., 2008). This is in part due to the fact that tardigrades possess morphological characters relating to both nematodes and arthropods. As an example, they possess a muscular myoepithelial triradiate pharynx (Ruppert, 1982; Ruppert and Barnes, 1994), which is also seen in nematodes and loriciferans (Kristensen, 2003), and conversely have lobed cerebral ganglia connected with a ladder-type chain of ventral trunk ganglia, much like an arthropod nervous system. For review of some of the inconsistencies, see Greven (1982). The architecture of the nervous system, and in particular, the brain of Tardigrada, has often been emphasized as important for phylogenetic analysis attempting to elucidate the relation of the group to other phyla, however, a general consensus has yet to be reached (Marcus, 1929; Dewel and Dewel, 1996; Zantke et al., 2008). Even the most recent investigation on the tardigrade nervous system, using confocal laser scanning microscopy, immunocytochemical staining, and three-dimensional (3D) reconstruction, did not provide explicit conclusion toward the phylogenetic relationship between tardigrades and arthropods or nematodes (Zantke et al., 2008). Earlier investigations of tardigrade cephalic sense organs and their innervations have led to suggestions of a three-lobed brain (Kristensen and Higgins, 1984a,b; Dewel and Dewel, 1996; Wiederhöft and Greven, 1996). Furthermore, the ventral nerve cord has been described to comprise paired segmental ganglia interconnected by longitudinal connectives, and connected intrasegmentally by transverse commissures giving the appearance of a rope-ladder like organization (Kristensen, 1982). This structuring of the central nervous system (CNS) is similar to that observed in arthropods (Scholtz, 2002; Müller, 2006; Scholtz and Journal of Morphology Edgecombe, 2006) and would therefore strengthen the hypothesis that Tardigrada is closely related to Arthropoda. However, the results obtained by Zantke et al. (2008) on the eutardigrade Macrobiotus hufelandi C.A.S. Schultze (1833) showed a three lobed brain, but did not support the existence of three segments in the head. Their argument against three brain segments is founded on a suggested hypothetical model based on their data, in which it seems all brain commissures are located in the hypothetical deutocerebrum. Also, they find no connective from the hypothetical tritocerebrum to the first ventral trunk ganglion and transverse commissures of the ventral trunk ganglia were not observed. Studies on tardigrade development support the sister group relationship of Tardigrada and Arthropoda 1 Onychophora (Hejnol and Schnabel, 2006), although it was not determined whether or not a third brain lobe is present. In addition, the nature of the subpharyngeal ganglion was questioned as it seemed to be an outgrowth of the brain. Due to these differences between earlier morphological investigations, we found it is necessary to reinvestigate tardigrade neuroanatomy. A clarification of whether or not certain structures exist is needed, and a detailed neuroanatomical investigation can furthermore provide possible evidence for the sister group relationship of Tardigrada. Here, we provide a detailed description of the nervous system of the marine eutardigrade Halobiotus crispae Kristensen (1982) based on immunocytochemical staining, confocal laser scanning microscopy, and computer-aided 3D reconstructions, supported by transmission electron microscopy. Specifically, the brain and the ventral ganglia receive much attention, as these structures are of phylogenetic importance. Our investigation expands on the current knowledge of tardigrade neuroanatomy, and our data are interpreted according to existing data and theories. MATERIALS AND METHODS Specimens of the tardigrade Halobiotus crispae were sampled at Vellerup Vig, in the Isefjord. Animals were extracted according to methods previously described (see Kristensen, 1982; Eibye-Jacobsen, 1997; Møbjerg and Dahl, 1996, Møbjerg et al., 2007; Halberg et al., 2009a,b; Halberg and Møbjerg, 2012). Live tardigrades were stored at 48C in seawater (20%) from the locality and supplied with substrate. The animal examined with transmission electron microscopy was collected intertidally at the type locality Nipissat Bay, Disko Island, West Greenland. Relaxation and Fixation Specimens of H. crispae were stretched in freshwater and subsequently relaxed using CO2-enriched water to prevent muscle contractions during fixation. The CO2-enriched water was applied drop by drop until the animals were completely passive. Immediately after relaxation, the specimens were fixed in 4% paraformaldehyde in 0.1 mol l21 phosphate buffered saline (PBS) NEUROANATOMY OF Halobiotus crispae 21 (for 500 ml of a 53 concentrated stock solution of 0.5 mol l PBS: 33.48 g Na2HPO42H2O, 7.93 g NaH2PO4H2O, pH 7.2–7.4) for 60 min at room temperature, and washed at least 83 10–15 min in washing buffer (0.1 mol l21 PBS with 0.1% NaN3). Permeabilization and Blocking After fixation, the animals were perforated using a very fine needle to facilitate penetration of the antibodies, followed by 90 s of sonication in a sonication bath at 30 W (Branson, B-1200 E1). Subsequently, the tardigrades were transferred to a blocking and permeabilizing solution consisting of 0.1 mol l21 PBS with 1% Triton X-100, 0.1% NaN3 and normal goat serum (NGS, 6% working concentration; Sigma cat. no. G9023), and incubated over night at 48C. Staining For immunolabeling, the primary antibody used was antiacetylated a-tubulin at a concentration of 1:250 (antimouse, monoclonal; Sigma, cat. no. T6793). Specimens were incubated with primary antibody on a shaker (celloshaker variospeed) at 48C for 48 h, and subsequently rinsed with washing buffer for at least 63 20–30 min and finally over night. Specimens were then incubated with secondary antibody (Alexa flour 488 goatantimouse, Invitrogen cat. no. A11001) at a concentration of 1:200 for 48 h on a shaker (celloshaker variospeed) at 48C. Antibodies were diluted with PBS with 0.1% NaN3 and NGS. During the last half of the incubation period with secondary antibody, 40 ,6-diamidino-2-phenylindole (5 lg/ml, DAPI; Invitrogen cat. no. D21490) was added for nuclear staining. After incubation, the animals were rinsed thoroughly in washing buffer for at least 63 20–30 min followed by an overnight wash. A total of 58 animals were used. A complementary immunolabeling with serotonin was performed following the same protocol as described above, using antiserotonin 5-HT 1:200 (antirabbit, polyclonal; Sigma, cat. no. S1561). Negative controls were performed during the initial experiments. The controls were performed by omitting the primary antibody from the staining procedure. The controls showed no fluorescence other than the autofluorescence of cuticular structures. Mounting and Image Acquisition Prior to mounting, specimens were treated with a glycerol series. This was done to prevent any deformation of the animals, due to the density and viscosity of the mounting media. The glycerol was applied at the start of the third buffer rinsing (see above), one drop at a time, starting with 5% glycerol. A few drops were applied and a small amount of liquid was simultaneously removed, until all the rinsing liquid had been replaced with 5% glycerol. This process was repeated with 10% followed by 25–100% glycerol. Then, the specimens were mounted on coverslips in Flouromount mounting medium (Southern Biotechnology Associates, Birmingham, AL). Using an Erwin loop, specimens were placed in droplets of the mounting medium and carefully manipulated with a fine needle into the desired position and the coverslips were sealed. Image acquisition was performed on a Leica DM RXE 6 TL microscope equipped with a Leica TCS SP2 AOBS confocal laser scanning unit, using the 488-nm line of an argon/crypton laser for antibody detection and UV laser for DAPI. A maximum projection of the image series was used and processed in the 3D image program IMARIS (Bitplane, Zurich, Switzerland). Transmission Electron Microscopy 8 specimens were used for transmission electron microscopy. The specimens were collected in Nipissat Bay, Disko Island, West Greenland (N 698 25.9340 , E 548 10.7680 ) in April 1979 1229 (type material: Kristensen, 1982), and fixed with a trialdehyde solution followed by postfixation with 1% OsO4. Ultrathin sections The specimens were cut with a diamond knife. For details on the preparation of transmission electron microscopy sections see Kristensen (1976). Regarding the nomenclature, we follow that used by Marcus (1929), also implemented by Zantke et al. (2008), and to some extent Kristensen (1982). Common neuroanatomical terms are in accordance with Richter et al. (2010). RESULTS H. crispae showed pronounced immunoreactivity against acetylated a-tubulin, particularly in the brain as well as the longitudinal nerve cords and ventral ganglia (Fig. 1). The acetylated a-tubulin immunoreactions combined with nuclear labeling with DAPI produced detailed images of the nervous system. In the following description, the CNS, comprising the brain, ventral longitudinal nerve cords, and ventral ganglia, will be described separately from the peripheral nervous system (PNS), which is comprised of every nervous structure outside the CNS. The sensory areas described as papilla cephalica (Figs. 1 and 2; pc) and temporalia (Fig. 2A,B; t. see Kristensen, 1982) are very closely associated with the brain and will be treated as part of the CNS. The nervous system of H. crispae has a clear segmental organization, with four paired ventral trunk ganglia each connected with two leg ganglia and two lateral neurons. Each of the lateral neurons in turn communicates with a dorsal neuron. The brain appears to be composed of three paired lobes each containing a commissure. The brain is connected with a subpharyngeal ganglion (g0) by a pair of connectives from the ventrolateral lobes. The subpharyngeal ganglion is in turn connected by a ventral longitudinal nerve cord to the first ventral trunk ganglion. The Central Nervous System The brain. The gross structure of the brain of H. crispae consists of 11 nerve cell clusters containing the soma of the nerve cells, and an elaborate network of nerve fibers and fiber bundles connecting the clusters (Figs. 2–6). The number of nerve cells and cells associated with these in the head region accounts for approximately one-quarter to one-third of all the cells in the animal (see review by Møbjerg et al., 2011). The interpretation of the number of brain clusters depends on which structures are considered to be part of the brain as well as on the interpretation of what comprises a cluster. Here, we define a brain cluster as a distinct group of cells or cells forming part of a lobe. The overall brain structure of H. crispae comprises two lateral outer lobes, two inner lobes, a median ganglion, and two ventrolateral lobes (Fig. 2). The laterally located outer brain lobes are comprised of two clusters, which is also true for the two inner brain lobes. The median ganglion, situated Journal of Morphology 1230 D.K. PERSSON ET AL. Fig. 1. Overview of the nervous system of Halobiotus crispae showing immunoreactivity against antiacetylated a-tubulin, maximum projection. (A) Double labeling with antiacetylated a-tubulin and DAPI. Lateral view, showing the strong innervation of the papilla cephalica (pc) from the brain, as well as the arrangement of the PNS. (B) Dorsal and (C) ventral view, respectively, giving an overview of the CNS with brain and ventral ganglia as well as the arrangement of the PNS. cl, cloacal neurons; gI-IV, ventral ganglion I-IV; dne, dorsal neuron; dn, dorsal nerve; lgg, leg ganglion; lne, lateral neuron; lln, lateral longitudinal nerve; ln, lateral nerve; pc, nerves of papilla cephalica. Journal of Morphology NEUROANATOMY OF Halobiotus crispae 1231 Fig. 2. Halobiotus crispae, immunoreactivity against antiacetylated a-tubulin in the brain viewed from four different angles. (A) Dorsal view showing the five nerves of the papilla cephalica (n7, n8, n9, n10a, n10b) as well as the head nerves n12 and n13. (B) Frontal view revealing the preoral commissure (prcm) just ventral to the dorsal commissure. Maximum projection. White arrowheads indicate conspicuous immunoreactive nerves of the inner lobes. (C) Lateral and (D) anterolateral view, both showing the 3D arrangement of the outer and inner lobes. Normal shading, Imaris. (E, F) Frontal view showing immunoreaction in the area of the ventrolateral lobe and the ventral commissure, maximum projection and iso-surface rendering, respectively. dc, dorsal commissure; ic, inner connective; il.acl, inner lobe anterior cluster; il, inner lobe; il.pcl, inner lobe posterior cluster; mg, median ganglion; oc, outer connective; ol, outer lobe; pc, nerves of papilla cephalica; prcm, preoral commissure; t, temporalia; vc, ventral commissure; vll, ventrolateral lobe. Double arrows indicate orientation of the image. Anterior and posterior is indicated with a and p, respectively, whereas d and v indicate dorsal and ventral, respectively. Journal of Morphology 1232 D.K. PERSSON ET AL. Fig. 3. Halobiotus crispae, close-up of the brain, with individual stack images showing specific details. (A) Dorsal view of a maximum-projection. Here, the connection of the brain to the nervous system of the trunk is revealed as well as the stylet commissure (stcm) which is connected with the stylets. (B–E) Individual sections from dorsal to ventral. (B) Notice the innervations of the temporalia (t), as well as the conspicuous immunoreactive areas of the inner lobes, lateral to the dorsal commissure (white arrowheads). The n13 nerve is clearly visible in the anteriodorsal part of the head. (C) Going further ventral reveals the large nerve bundle from the outer lobes, which become part of the dorsal commissure (white arrowheads). (D) This section shows the postoral commissure (pocm) and the connectives (co) to the ventral part of the brain. Also, notice the buccal neurons (bn) dorsal to the mouth opening. (E) The mouth opening (mo) is innervated by several nerves connected with the ventral part of the brain (white arrowheads). In the same focal plane as the mouth opening, the ventral commissure (vc) is clearly discernible. bn, buccal neurons; co, connective; dc, dorsal connective; ey, eye; gI, first ventral ganglion, ic, inner connective; ln, lateral nerve; mo, mouth opening; oc, outer connective; pc, nerves of papilla cephalica; pocm, post oral commissure; t, temporalia; vc, ventral commissure. Journal of Morphology NEUROANATOMY OF Halobiotus crispae Fig. 4. Halobiotus crispae, close-up of the mouth opening, showing the external and internal structures of the buccal lamella. (A) Scanning electron micrograph revealing the external morphology of the six buccal lamella (asterisk) surrounding the mouth opening. (B) Transmission electron micrograph of a section through the buccal lamella shows the internal structure of the buccal lamella and the surrounding tissue. (C) Antiacetylated a-tubulin immunoreaction in the mouth opening. ci, cilia. Scale bars are 15 lm. between the inner lobes, consists of a single cluster of cells and innervates the forehead via the nerve n13, whereas the ventrolateral lobe comprises one cluster on each side of the buccal tube (Figs. 2 and 3). The mouth is innervated by nerves from both ventrolateral and inner lobes (Figs. 3E and 4; se 1233 below). The brain contains a commissure between each of the paired brain lobes and a smaller commissure in connection with the innervations of the stylets. In total, we identified four commissures: the large dorsal commissure connects the inner lobes, the preoral commissure connects the outer lobes, the ventral commissure connects the ventrolateral lobes and the stylet commissure innervates the stylets and stylet supports (Figs. 2,3, 5 and 6). A pair of connectives extending from the ventrolateral lobes, connects the brain to the central nervous system of the trunk through a subpharyngeal ganglion (Fig. 7-8). The subpharyngeal ganglion communicates with the first ventral trunk ganglion through the ventral longitudinal nerve cord. In addition, so-called outer and inner connectives extend from respectively the outer and ventrolateral lobes, connecting the brain to the first ventral trunk ganglion (see below). The lateral outer brain lobes constitute the largest and most conspicuous part of the brain. These large lobes are elongated and positioned dorsolaterally to the buccal tube and extend caudally; they are referred to as the outer lobes (Figs. 2, 3, 6, 9, and 10; ol) and they contain the eye spots (Fig. 3A1D; ey). The outer lobes can be divided into an anterior and a posterior cluster (Figs. 2C and 6; ol.acl, ol.pcl). The second pair of lobes is closer to the median plane of the animal and is referred to as the inner lobes (Figs. 2, 5, 6, and 10; il). The inner lobes also seem to be composed of two nerve cell clusters, a posterior cluster and an anterior cluster (Figs. 2A1C and 5; il.pcl, il.acl). Both pairs of brain lobes are positioned dorsal to the buccal tube, and the two ‘‘hemispheres’’ are interconnected by a massive bundle of nerve fibers containing the preoral and predorsal commissures (Fig. 2A1B; prcm, dc). On each side of the dorsal commissure small, highly immunoreactive areas of the inner lobes are clearly visible (Fig. 2A,B; white arrowheads), which seem to be connected with nerves running through the dorsal commissure. A very pronounced nerve runs between the anterior cluster of the inner lobe and the dorsal commissure (Fig. 5; n4, see also Zantke et al., 2008). The anterior clusters of the inner lobes receive nerve extensions from the sensory area of the anterior part of the head, known as the papilla cephalica (Figs. 1–315). A total of five nerves can be distinguished in the papilla cephalica named n7, n8, n9, n10a, and n10b (Figs. 2A1C and 5), of these only n9 originates from the outer lobe, whereas the others originate from the inner lobe. From the posterior region of the outer lobes, in the vicinity of the eyes, extensive dorsal innervations of the cuticle are clearly visible (Figs. 2A–D, 3A1B, and 5; t). These innervations are more or less similar to the innervations seen in the cuticle of the papilla cephalica and are termed temporalia by Kristensen (1982); the exact number of nerves Journal of Morphology 1234 D.K. PERSSON ET AL. Fig. 5. Halobiotus crispae, antiacetylated a-tubulin immunoreactions of the brain. (A) Maximum projection revealing the n4 nerve. Also, the head nerves n12 and n13 are clearly visible. (B) Normal shading rendering, using the software Imaris, reveal the n13 nerve to originate from the median ganglion (mg). Furthermore, the individual nerves, n7-n10a1b, of the papilla cephalica can be distinguished. dc, dorsal commissure; il.acl, inner lobe anterior cluster; mg, median ganglion; pc, nerves of papilla cephalica; t, temporalia. innervating the area cannot be distinguished. Between the inner lobes, an unpaired triangularshaped median ganglion is located (Figs. 2A1C and 5; mg). From this ganglion, a double nerve cord extends anteriorly above the dorsal commisJournal of Morphology sure, ending in a small circular nerve structure in the cuticle of the forehead. This nerve corresponds with the n13 nerve described in Zantke et al. (2008), and it follows that the very fine paired nerve lateral to n13 is n12 (Figs. 1B,C, 2A–D, 3A– NEUROANATOMY OF Halobiotus crispae 1235 Fig. 6. Halobiotus crispae, individual sections from an image stack, showing immunoreaction to antiacetylated a-tubulin and corresponding DAPI labeling. (A) A large part of the dorsal commissure (dc) originates from the outer lobes (arrows). (B) The DAPI labeling in the same focal plane as A shows the posterior and anterior clusters of the outer lobes, and the posterior cluster of the inner lobes. (C and D) The position of the stylet commissure coincides with the stylet supports (ss). (E and F) In the ventral part of the animal, a ventral commissure is clearly visible (E, vc), which corresponds very well with the position of two ventrolateral lobes revealed with DAPI labeling (F, vll). (G) In the most ventral part of the animal, the antiacetylated a-tubulin staining show that the outer connectives connects to the outer lobes and the inner connectives connects to ventrolateral lobes, which is connected with the subpharyngeal ganglion (g0). bt, buccal tube; dc, dorsal commisure; g0, ventral ganglion; ic, inner connective; il, inner lobe; oc, outer connective; ol, outer lobe; ol.acl, outer lobe anterior cluster; il.pcl, inner lobe posterior cluster; ol.pcl, outer lobe posterior cluster; ss, stylet support; stcm, stylet commissures; stn, stylet nerves; vll, ventrolateral lobe. C, and 5; n12, n13). The outer lobes, inner lobes, the median ganglion, and associated commissures constitute the dorsal part of the brain. Ventral to the buccal tube another commissure is located, the ventral commissure (Figs. 2E1F, 3E, and 6E; vc). Posterior to the ventral commisJournal of Morphology 1236 D.K. PERSSON ET AL. Fig. 7. Halobiotus crispae, lateral view of the brain and ventral ganglia. (A) Immunoreactivity against antiacetylated a-tubulin in the brain showing the longitudinal nerve cord connecting the first ventral trunk ganglion (gI) to the subpharyngeal ganglion (g0). (B) DAPI staining showing the nuclei of gI and g0. (C) Lateral view of the ventral trunk ganglia and the subpharyngeal ganglion (g0). Asterisks indicate possible perikarya. bt, buccal tube; clg, claw gland; g0, subpharyngeal ganglion; gI, first ventral trunk ganglion; ic, inner connective; oc, outer connective; lgg, leg ganglion; lnc, longitudinal nerve cord. sure and slightly dorsal to the buccal tube, a fourth commissure is visible in the area of the stylets (Fig. 3D; stcm). From this stylet commissure, stylet nerves (Fig. 3D; stn) extend anterolaterally into the ventrolateral brain lobes. These lobes consist of a cell cluster on each side of the buccal tube (Figs. 2E1F and 6C–H). From the ventrolateral lobes, several nerves extend anteriorly toward the mouth (Fig. 3E; white arrowheads). In addition, the mouth receives nerves from several nerve cells dorsal to the mouth opening (Fig. 3E; bn) which are connected with the inner lobes. These nerves are all connected with the six buccal lamellae surrounding the mouth opening (Fig. 4; asterisks). Sections of the image stack in the area of the ventral commissure reveal cell clusters around the buccal tube and immediately ventral to it, as well Journal of Morphology as a ganglion-like structure posterior to these clusters, these are termed the ventrolateral lobes (vll) and the subpharyngeal ganglion (g0), respectively (Fig. 6F1H; stippled circles/oval). The ventrolateral lobes are also clearly visible in Figure 2E and F, interconnected by the ventral commissure. Additionally, a transmission electron microscopic image of the same area support the existence of a third pair of brain lobes and also shows the connective extending ventrally from these lobes (Fig. 9). The ventral ganglia and longitudinal nerve cords. From the posterior region of the outer lobes, a double nerve tract extends ventrocaudally, with one lateral nerve (Fig. 3A; ln) branching off into the lateral nervous system and the other connecting to the first ventral trunk ganglion. The latter is referred to as the outer connective (Fig. 3A; NEUROANATOMY OF Halobiotus crispae 1237 ally, serotonin immunoreactivity is shown in these regions as well, which could indicate the presence of perikarya (Fig. 7, asterisks). In addition, we find a fifth ventral ganglion, the subpharyngeal ganglion (g0), which is connected with the ventrolateral brain lobes and the first ventral trunk ganglion (gI) via the longitudinal nerve cord (Fig. 7). The Peripheral Nervous System Fig. 8. Halobiotus crispae, the nervous system of the trunk. Insets show close-up of the ventral ganglia I–IV and arrows indicate the number of commissures. cl, cloacal nerves/neurons; cm, commissure. oc). From the ventrolateral part of the brain extends the inner connective (Fig. 3A; ic); the inner connective is also connected with the first ventral trunk ganglia. These nerve tracts connect the brain with the CNS of the trunk comprised of paired ventral trunk ganglia and longitudinal nerve cords. There are four paired ventral trunk ganglia, each composed of approximately 40 cells that are associated with a corresponding leg pair (Figs. 1 and 8; gI–gIV). The ventral trunk ganglia are intrasegmentally connected by transverse commissures and intersegmentally connected by ventral longitudinal connectives, giving the appearance of a rope-ladder-like arrangement (Figs. 118). In the first three ganglia, we observe two commissures, whereas in the fourth, only one can be observed (Fig. 8; inserts). The longitudinal connectives are collectively referred to as the longitudinal nerve cords (Fig. 1; lnc), and extend through most of the length of the animal, terminating in the fourth ventral trunk ganglion (Fig. 1; gIV). It is not unambiguous whether the longitudinal nerve cords contain perikarya or not; the a-tubulin stainings show some thickenings of the longitudinal nerve cords in the regions between the ventral trunk ganglia (Figs. 1 and 8). Addition- The PNS is mainly comprised of four dorsal and four lateral neurons on each side of the animal (Fig. 1; dne and lne). The dorsal and lateral neurons are connected by dorsal and lateral nerves (Fig. 1A1B; dn, ln). The lateral nerves connect the lateral neurons with the ventral trunk ganglia. In addition, there is a lateral nerve connecting the first lateral neuron to the brain (Fig. 3A; ln). This lateral nerve originates from the outer lobe, propagating in parallel with the outer connective, and connects to the first lateral neuron (Fig. 1A; lne). From the first lateral neuron, nerve projections extend both in an anterior–posterior as well as in a dorsal direction, and are referred to as lateral longitudinal nerves (Fig. 1A–C; lln) and dorsal nerves (Fig. 1A1B; dn), respectively. The lateral nerve extends through most of the length of the animal, connecting the four lateral neurons. The lateral neurons are connected with dorsal neurons (Fig. 1; dne) via dorsal nerves (Fig. 1; dn). From all the dorsal ganglia/neurons sensory cilia extends further toward the dorsal side. These modified cilia could very well correspond to the lateral cirri of heterotardigrades. At the fourth ventral trunk ganglion, nerves extend dorsocaudally to the dorsal neuron (Fig. 1; dne) and nerves branch off into the hind legs. The nerves in the hind legs connect to leg ganglia in the distal part of the legs (Fig. 1; lgg), and two nerves extend from these ganglia terminating in a cilia in the proximal part of the legs. In addition, a pair of nerves extends from the fourth ventral trunk ganglion and terminates near the cloaca (Fig. 1A; cl). Leg ganglia are connected with nerves originating from the ventral trunk ganglia. From the second and third ventral trunk ganglion, three nerves n1, n2, and n3 originates (Fig. 1C). The n3 and n2 nerves connect with the anterior and posterior region of the associated leg, with the posterior nerve attaching at the distal part of the leg, and the middle nerve connect to a ganglion in the proximal part of the leg—a leg ganglion (Fig. 1C; lgg). The n1 nerve connect to the lateral neuron. For the first ventral trunk ganglion, it follows that the outer connectives correspond to the n1 nerves of the second and third ventral ganglia, whereas the n2 and n3 nerves exhibit the same pattern in ventral trunk ganglia I–III. For complete overview of Journal of Morphology 1238 D.K. PERSSON ET AL. Fig. 9. Halobiotus crispae, transmission electron micrograph merging of five cross-sections of the brain of H. crispae. Image merging performed in CorelDraw. Notice the ventrolateral lobes (vll) and their connectives to the ventral region. dc, dorsal commissure; bt, buccal tube; co, connectives; ol, outer lobe; stn, stylet nerve; vll, ventrolateral lobe. the described nervous structures and nerves in the present study, as well as comparison to the nervous structures previously described in the literature see table 1. DISCUSSION Marcus (1929) described the nervous system of Macrobiotus hufelandi with impressive detail, exemplified by the fact that most of our findings (using contemporary techniques) in H. crispae confirm what he described. Moreover, our results support most of the descriptions of the tardigrade nervous system by Zantke et al. (2008), albeit we report some new findings together with increased details. In particular, we have obtained greater details on brain structure, and important structures in the ventral ganglia. Noticeably, several of the details uncovered in our study pertain to structures that were declared missing in tardigrades or possibly misinterpreted by Zantke et al. (2008). Journal of Morphology In the nervous system investigation of M. hufelandi by Zantke et al. (2008) a preoral commissure is described in the brain to be positioned just below the dorsal commissure. In addition, they show that a postoral commissure is connected with the first ventral ganglion via the inner connectives, and to the dorsal commissure via the circumbuccal connectives, forming a circumbuccal ring. Our data show a similar preoral commissure positioned just ventral to the dorsal commissure (Fig. 2B, prcm). In addition, we find a commissure ventral to the buccal tube which could be equivalent to the postoral commissure described by Zantke et al. (2008, Fig. 2F, pocm). We choose to term this structure the ventral commissure (Fig. 3E, vc) as it appears more similar to the ventral ring commissure found in the heterotardigrade Echiniscus viridissimus by Dewel and Dewel (1996). In H. crispae, the ventral commissure is not connected with the first ventral trunk ganglia via the inner connectives; more precisely, the ventral commissure connects the ventrolateral lobes NEUROANATOMY OF Halobiotus crispae 1239 Fig. 10. Conceptual drawing constructed on the basis of the data in this study, showing our interpretation of the brain structure of H. crispae. (A) Lateral view. (B) Frontal view. clg, claw gland; co, connective; dc, dorsal commissure; ey, eye; g0, subpharyngeal ganglion; gI, first ventral trunk ganglion; ic, inner connective; il, inner lobe; lgg, leg ganglion; mg, median ganglion; mo, mouth opening; oc, outer connective; ol, outer lobe; pc, papilla cephalica; pb, pharyngeal bulb; st, stylet; t, temporalia; vll, ventrolateral lobe. Arrows indicate the approximate area of the transmission electron microscopical section in figure 9. Journal of Morphology 1240 Journal of Morphology TABLE 1. Table listing the structures of the nervous system of H. crispae described in this study and compared with previous studies of H. crispae or other tardigrades Identified structures and named nerves CNS, head PNS M. hufelandi M. hufelandi H. crispae Constructed ancestral tardigrade Persson et al. (this study) Marcus (1929) Zantke et al. (2008) Kristensen (1982) Nielsen (2001) Inner lobes (il) with anterior and posterior clusters (acl, pcl) Dorsal commissure (dc) Outer lobes (ol) with anterior and posterior clusters (acl, pcl) Preoral commissure (prcm) Median ganglion (mg) n12, n13 n4 Buccal neurons (bn) Ventrolateral lobes (vll) Ventral commissure (vc) Subpharyngeal ganglion (g0) Stylet commissure (stcm) Stylet nerve (stn) Outer connective (oc) Inner connective (ic) Innervation of papilla cephalica (pc, n7, n8, n9, n10a, n10b) Innervation of temporalia (t) Innervation of buccal lamella Four ventral trunk ganglia (gI-IV) Transverse commissures (cm) Longitudinal nerve cords (lnc) Leg ganglia (lgg) Leg nerves (n2+n3) Lateral nerves (n1, ln) Lateral neurons (lne) Dorsal nerves (dn) Dorsal neurons (dne) Lateral longitudinal nerves (lln) Cloacal nerves (cl) Inner lobes Inner lobes (posterior cluster) Protocerebrum Protocerebrum Outer lobes Dorsal commissure Outer lobes (dorsal cluster) Deutocerebrum Deutocerebrum Tritocerebrum Tritocerebrum Subesophageal ganglion Subesophageal ganglion Stylet nerve (s.ne) Outer connective Stylet nerve (s.ne) Outer connective Innervation of papilla cephalica Innervation of temporalia Innervation of papilla cephalica Innervation of temporalia Innervation of buccal lamella Four ventral trunk ganglia Transverse commissures Longitudinal nerve cords Preoral commissure Median ganglion n11, n12n, n13 n4, n5, n6 Postoral commissure? Subesophageal ganglion (I) Postoral commissure? Outer connective Innervation of papilla cephalica Innervation of temporalia Innervation of buccal lamella Four ventral trunk ganglia (IIV) Transverse commissures Longitudinal nerve cords Leg ganglia Leg nerves (n3) Lateral nerves (n1, n2) Lateral neurons Dorsal nerves Dorsal neurons Cloacal nerves Outer connective Inner connective Innervation of papilla cephalica (n7, n8, n9, n10) Four ventral trunk ganglia (gI-IV) Longitudinal nerve cords Leg ganglia n1, n2, n3 Neurophils of the hind legs Four ventral trunk ganglia Transverse commissures Longitudinal nerve cords Leg nerves Sensilla D.K. PERSSON ET AL. CNS, trunk H. crispae NEUROANATOMY OF Halobiotus crispae flanking the buccal tube. It is these lobes that are connected with the first ventral trunk ganglia via the inner connective. When looking at a DAPI staining of the area of the stylet commissure, there is not an actual ganglion (compare Fig. 6C,D) and the commissure is positioned exactly at the stylet supports. Also, nerves extend laterally from the commissure (Figs. 3A1D and 6C; co) maybe connecting to the stylet muscles. So, this seems to be a stylet commissure in connection with the innervations of the stylets. Therefore, we chose the name stylet commissure. In addition, the stylet commissure is not directly connected with the dorsal commissure, but it is actually connected with the ganglion-like structure associated with the ventral commissure. Furthermore, through closer observations of the brain, it is obvious that the stylet commissure is positioned dorsal to the buccal tube, so it should not be confused with the postoral commissure described by Zantke et al. (2008), which is positioned ventral to the buccal tube. The transmission electron microscopical image in Figure 9 and the image stack of the a-tubulin and DAPI staining on Figure 6 supports the immunoreactions seen in Figure 2F, which suggests that the ganglion of the ventral commissure is a paired, lobed structure situated lateral to the buccal tube. More ventral, we find a cell cluster much like the first ventral trunk ganglion, in very close proximity to the lobes of the ventral commissure (Fig. 6H). The DAPI staining shows a cell structure very similar to the first ventral trunk ganglion, and it is positioned in the same level. Combined with the immunoreaction and complementary DAPI staining in Figure 7, it seems to be a ganglion in the ventral part of the head connected with the first ventral trunk ganglion. Although we cannot determine if these nerve cells have a paired cluster arrangement as is the case for the ventral trunk ganglia, we cautiously hypothesize that it is a subpharyngeal ganglion. The presence of the subpharyngeal ganglion is also supported by a developmental study, in which a subesophageal ganglion (the subpharyngeal ganglion) is observed and suggested to be an outgrowth of the brain (Hejnol and Schnabel, 2005). From this suggestion, it is argued that the subpharyngeal ganglion is part of the brain, which then forms a circumbuccal ring resembling cycloneuralian conditions (Zantke et al., 2008). This is still at a hypothetical stage and actual evidence is needed to clarify if the subpharyngeal ganglion is part of the brain. Nevertheless, if we consider the idea of cephalization of several anterior segments to be true, it would not be surprising if the subpharyngeal ganglion seemed to originate from the same region as the three paired brain lobes. In order to fully understand the nature and origin of the subpharyngeal ganglion, it will be important 1241 to investigate the brain of arthrotardigrades as they are considered to have many plesiomorphic characters. Therefore, they may be closer to the ancestral state in appearance compared to the more specialized eutardigrades. Furthermore, it seems plausible that the lobes of the ventral commissure may be a third paired brain lobe, and it is clearly connected with the subpharyngeal ganglion via two connectives (Fig. 9). The fact that we find two connectives, one from each ventrolateral lobe, suggests that the subpharyngeal ganglion is or may have been a paired cluster like that found in the ventral trunk ganglia. The ventrolateral lobes are also connected with the stylet nerves (Fig. 3A). If we accept the hypothesis of the stylets being internalized claws, that is, internalized leg, this supports the idea of segmentation in the head. Hence, we suggest a brain of at least three parts, and the existence of a subpharyngeal ganglion in tardigrades. The third paired lobe has previously been described in H. crispae by Kristensen (1982) and could be homologs to the arthropod tritocerebrum. In addition, if there is homology between the outer and inner lobes and the protocerebrum and deutocerebrum of arthropods, then, the brain configuration of tardigrades could be interpreted as arthropod-like, as suggested in earlier descriptions (Kristensen and Higgins, 1984a,b; Dewel and Dewel, 1996; Nielsen, 2001, 2011), though further studies on tardigrade development is needed to verify a possible homology. In Figure 6, the subpharyngeal ganglion is indicated by DAPI labeling, but we do not see a distinct immunoreaction for a-tubulin. However, on Figure 7A, which shows antiacetylated a-tubulin staining from the lateral side, we find clear immunoreaction in the subpharyngeal ganglion, as well as in the nerve cord connecting it to the first ventral trunk ganglion. This is also supported by serotonergic immunoreaction in the ventral trunk ganglia as well as g0 (Fig. 7C). One explanation for the lack of immunoreaction in Figure 6 could be failure of antibody recognition, which we have previously encountered with tardigrades. An example is the inner lobes which show immunoreaction in Figure 2C1D, but in Figure 2A no immunoreaction can be observed in the posterior clusters of the inner lobes. The n13 nerve running dorsally from the median lobe (Fig. 5B) terminating in a circular shape in the forehead is interpreted as a rudiment of the median cirrus of Heterotardigrada (see also Zantke et al., 2008). Generally, the median cirrus is present in all Arthrotardigrada (though reduced in some Archechiniscus species), and missing from all Eutardigrada and Echiniscoidea. Assuming that the marine Arthrotardigrada represents the ancestral condition (Renaud-Mornant, 1982; Jørgensen et al., 2010), the presence of a median cirrus within Tardigrada represents the plesiomorphic Journal of Morphology 1242 D.K. PERSSON ET AL. condition. Additionally, connected with the dorsal neurons, we found modified cilia that likely are reduced sensory structures, homologs to the lateral cirri of Heterotardigrada. Consequently, these structures could be important in an intra-phylum phylogenetic perspective. When looking at the ventral longitudinal nerve cords and the ventral trunk ganglia, a distinct segmentation is readily recognizable. Apart from the head, each of the four segments bears a leg and contains a ventral ganglion with peripheral nerves, which innervate the associated leg as well as the lateral and dorsal sides. The ventral trunk ganglia are paired and intersegmentally connected by transverse commissures, as described in the Results section. These commissures were originally described by Marcus (1929); however, their presence in tardigrades was recently questioned (Zantke et al., 2008). Our investigation shows that tardigrades possess intersegmental transverse commissures. Their presence is important to the phylogenetic debate as they are a critical component in the rope-ladder type nervous system. This type of nervous system is not encountered in any cycloneuralians, but is seen in most arthropods (Bullock and Horridge, 1965; Brusca and Brusca, 2002). The PNS primarily comprises the paired nerves n1, n2, and n3, which are forming a repeated pattern in association with the first three ventral ganglia. Only the n1 nerves of the first ventral ganglia differ from the pattern by connecting to the outer lobes of the brain. Also, the nerves associated with the fourth ventral trunk ganglion are arranged in a slightly different pattern. Although the difference in nerve arrangement in the fourth ventral ganglion is linked to the orientation and function of the hind legs, the difference in the n1 nerve in the first ventral trunk ganglion is due to its connection to the brain. We hypothesize that this deviation could be linked to an evolutionary scenario involving cephalization of at least three anterior segments of a tardigrade ancestor. In the light of our results, we propose the following hypothetical model to explain the organization of the tardigrade brain as three segments: The area of the outer lobes, with the eyes and innervations of the temporalia, which we interpret as homologs to the primary clavae and lateral cirri of the heterotardigrades, could be interpreted as the first segment or protocerebrum. This would then correspond to the first segment in arthropods bearing the compound eye, though this structure is not homologs to the tardigrade eye. Also, it would correspond to the first segment in the onychophorans which bears the antenna and the eyes. Although again these structures are not considered homologs to any of the structures found on the first segment in tardigrades. Journal of Morphology The next segment holds the inner lobes (possibly deutocerebrum), which innervates the papilla cephalica, homologs to the secondary clavae and the internal cirri of heterotardigrades (for homology on primary and secondary clava, see also Zantke et al. 2008). Consequently, this would be homologs to the second segment in the head of both arthropods and onychophorans, though again it is not readily possible to homologize between tardigrade sense organs and the modified limps of arthropods and onychophorans. The third segment contains the ventrolateral lobes (possibly tritocerebrum), which innervate the buccal lamella as well as the stylets and stylet supports. As it has been hypothesized that the stylet apparatus may have been formed by internalization of a leg it could be homologs to the second antenna/pedipalp of crustaceans/chelicerates. According to some authors, the tritocerebrum is not present in onychophorans (see Erikson and Budd, 2000; Mayer et al., 2010), if this is true it is difficult to homologize between tardigrades and onychophorans with respect to the third brain region. However, if we hypothesize that the cephalization in the Onychophora lineage excluded the ganglion which in tardigrades became the third brain lobe, then, the slime papilla could be homologs to the stylet apparatus. Of course, in order to accept this hypothesis, this needs to be supported by more data—for example, molecular data. Indeed, answers to some of the questions regarding homology may be found in future comparative developmental studies, like those performed by Jager et al. (2006) on Hox gene expression and by Gabriel and Goldstein (2007) on expression patterns of Pax 3/7 and Engrailed homologs. Whether or not tardigrades possess a tritocerebrum is important with regards to their phylogenetic position in relation to Onychophora and Arthropoda. As mentioned above, a tritocerebrum has not been found in the onychophorans. This is because the ganglion innervating the third segment is not part of the brain, but is part of the ventral nerve cord. Consequently, it was hypothesized that the last common ancestor of onychophorans and arthropods possessed a bipartite brain comprised of a protocerebrum and a deutocerebrum (Mayer et al., 2010). As H. crispae possess a third brain lobe, this could then support a clade with Tardigrada and Arthropoda as sister groups within Panarthropoda. The presence of a third brain lobe in tardigrades generates three possible phylogenetic configurations for Panarthropoda, assuming true homology between the brain lobes of Tardigrada and Arthropoda. In one scenario, the ancestor to Panarthropoda could have had three brain lobes, and consequently, the third lobe was lost in Onychophora. If this is true, brain morphology will not NEUROANATOMY OF Halobiotus crispae aid in resolving the phylogeny within Panarthropoda, as all placements of Onychophora are equally parsimonious. In contrast, if the last common ancestor of Panarthropoda had a bipartite brain, the outcome would be that Onychophora could be sister group to Tardigrada 1 Arthropoda, or grouped together with one of them in which case the development of the third brain lobe would be convergent. Hence, the most parsimonious hypothesis, when only considering brain morphology, is that the last common ancestor to Panarthropoda had a bipartite brain and that tritocerebrum was developed in a common ancestor to Tardigrada and Arthropoda. This is of course speculative, and even though molecular data support the inclusion of Tardigrada into Panarthropoda, it does not support a sister group relationship between Tardigrada and Arthropoda (Campbell et al., 2011). As noted above more investigations, especially on tardigrade development, are needed in order to substantiate our suggestions. The stylet apparatus may be internalized claws, as earlier transmission electron microscopical investigations have shown that stylets and claws are formed in the same way (Kristensen, 1976; Nielsen, 2001, 2011; Halberg et al., 2009a) and are similar to the jaws of Onychophora (Storch and Ruhberg, 1993; Mayer and Harzsch, 2007; Mayer et al., 2010). It is also a possibility that the stylets and the supports could be greatly modified mouth appendages. The nervous system is distinctly metameric, consisting of the three-lobed brain, the subpharyngeal ganglion, and the four ventral trunk ganglia. Characteristic to all tardigrades is the large paired outer connective with a small ganglion that connects the outer lobe (protocerebrum) to the first ventral ganglion (Marcus, 1929). However, we also found a thin connective between the ventrolateral lobe (tritocerebrum) and the first ventral trunk ganglion. Our interpretation of the brain structure is summarized in a concept drawing in Figure 10 and represents an integration of all our immunoreactive data from the head region as well as transmission electron microscopy. On the basis of this investigation, we find that tardigrades possess a brain of at least three parts. We hypothesize that the three paired lobes could originate from three head segments and that the subpharyngeal ganglion originated from a fourth segment. Along with the commissures of the ventral trunk ganglia and the segmentation of the body, this leads us to suggest that the tardigrade nervous system structure supports the clade Panarthropoda. The similarities of the tardigrade nervous system toward Cycloneuralia (loriciferans and kinorhynchs, Kristensen, 2003), especially the kinorhynchs, which have segmentally arranged ventral ganglia (Kristensen and Higgins, 1991), have been pointed out by morphologists as being 1243 surprising analogies. However, tardigrades may have some plesiomorphic characters in the nervous system as well as a myoepithelial triradiate pharyngeal bulb, so the mentioned similarities between Tardigrada and Cycloneuralia could be true homologies. ACKNOWLEDGMENT The authors thank Stine Elle for preparing the line art illustrations. LITERATURE CITED Aguinaldo AMA, Turbeville JM, Linford LS, Rivera MC, Garey JR, Raff RA, Lake JA. 1997. Evidence for a clade of nematodes, arthropods and other moulting animals. Nature 387:489– 493. Baccetti B, Rosati F. 1971. Electron microscopy on tardigrades. III. The integument. J Ultrastruct Res 34:214–243. Bertolani R. 1982. Tardigradi (Tardigrada). Guide per il riconoscimento delle specie animali delle acque interne Italiane. Verona, Italy: Consiglio Naz. Delle Richerche 15. pp 1–114. Brusca R, Brusca GJ. 2002. Invertebrates. Sunderland, Massachusetts: Sinauer Associates. pp 1–936. Budd GE. 2001. Tardigrades as ‘Stem-Group Arthropods’: The Evidence from the Cambrian Fauna. Zool Anz 240:265–279. Bussers JC, Jeuniaux CH. 1973. Chitinous cuticle and systematic position of Tardigrada. Biochem Syst 1:77–78. Bullock TH, Horridge GA. 1965. Structure and function in the nervous system of invertebrates. San Franscisco: WH Freemand. Campbell LI, Rota-Stabelli O, Edgecombe GD, Marchioro T, Longhorn SJ, Telford MJ, Philippe H, Rebecchi L, Peterson KJ, Pisani D. 2011. MicroRNAs and suggest that velvet worms are the sister group of Arthropoda. PNAS 108:15920– 15924. Crowe JH, Newell IM, Thomson WW. 1970. Echiniscus viridis (Tardigrada): Fine structure of the cuticle. Trans Am Microsc Soc 89:316–325. Dujardin F. 1851. Sur les Tardigrades et sur une espèce à longs pieds vivant dans l’eau de mer. Ann Sci Nat Zool Sér 3. T 15, 160–166. Dastych H, Kristensen RM. 1995. Echiniscus ehrenbergi sp. n., a new water bear from the Himalayas (Tardigrada). Entomol Mitt zool Mus Hamburg 11:221–230. Dewel RA, Clark WH. 1973. Studies on tardigrades. I: Fine structure of the anterior foregut of Milnesium tardigradum Doyère. Tissue Cell 5:133–146. Dewel RA, Dewel WC. 1996. The brain of Echiniscus viridissimus Peterfi, 1956 (Heterotardigrada): A key to understanding the phylogenetic position of tardigrades and the evolution of the arthropod head. Zool J Linn Soc 116:35–49. Dewel RA, Nelson DR, Dewel WC. 1993. Tardigrada. In: Harrison FW, Rice ME, editors. Microscopic Anatomy of Invertebrates, Volume 12: Onychophora, Chilopoda and Lesser Protostomata. New York: Wiley-Liss, Inc. pp 143–183; Chapter 5. Dunn CW, Hejnol A, Matus DQ, Pang K, Browne WE, Smith SA, Seaver E, Rouse GW, Obst M, Edgecombe GD, Sørensen MV, Haddock SHD, Schmidt-Rhaesa A, Okusu A, Kristensen RM, Wheeler WC, Martindale MQ, Giribet G. 2008. Broad phylogenomic sampling improves resolution of the animal tree of life. Nature 452:665–780. Edgecombe GD. 2010. Arthropod phylogeny: An overview from the perspectives of morphology, molecular data and the fossil record. Arthropod Struct Dev 39:74–87. Edgecombe GD, Giribet G, Dunn CW, Hejnol A, Kristensen RM, Neves,RC, Rouse GW, Worsaae K, Sørensen MV. 2011. Higher-level metazoan relationships: Recent progress and remaining questions. Org Divers Evol 11:151–172. Journal of Morphology 1244 D.K. PERSSON ET AL. Eibye-Jacobsen J. 1997. Development, ultrastructure and function of the pharynx of Halobiotus crispae Kristensen, 1982 (Eutardigrada). Acta Zool (Stock) 78:329–347. Erikson BJ, Budd GE. 2000. Onychophoran cephalic nerves and their bearing on our understanding of head segmentation and stem-group evolution of Arthropoda. Arth Struct Dev 29:197– 209. Gabriel WN, Goldstein B. 2007. Segmental expression of Pax3/7 and Engrailed homologs in tardigrade development. Dev Genes Evol 217:421–433. Garey JR, Krotec M, Nelson DR, Brooks J. 1996. Molecular analysis supports a Tardigrade-Arthropod association. Invertebr Biol 115:79–88. Garey JR. 2001. Ecdysozoa: The relationship between Cycloneuralia and Panarthropoda. Zool Anz 240:321–330. Giribet G, Carranza S, Baguñà J, Riutort M, Ribera C. 1996. First molecular evidence for the existence of a Tardigrada 1 Arthropoda clade. Mol Biol Evol 13:76–84. Greven H. 1982. Homologues or analogues? A survey of some structural patterns in Tardigrada. In: Nelson DR, editor. Proceedings of the Third International Symposium on the Tardigrada, August 3–6, 1980, Johnson City, Tennessee. Johnson City: East Tennessee State University Press. pp 55–75. Halberg KA, Møbjerg N. 2012. First evidence for epithelial transport in tardigrades: A comparative investigation of organic anion transport. J Exp Biol 215:497–507. Halberg KA, Persson D, Møbjerg N, Wanninger A, Kristensen RM. 2009a. Myoanatomy of the Marine Tardigrade Halobiotus crispae (Eutardigrada: Hypsibiidae). J Morph 270:996–1013. Halberg KA, Persson D, Ramløv H, Westh P, Kristensen RM, Møbjerg N. 2009b. Cyclomorphosis in Tardigrada: Adaptation to environmental constraints. J Exp Biol 212:2803–2811. Hejnol A, Schnabel R. 2005. The eutardigrade Thulinia stephaniae has an indeterminate development and the potential to regulate early blastomere ablations. Development 132:1349–1361. Hejnol A, Schnabel R. 2006. What a couple of dimensions can do for you: Comparative developmental studies using 4D microscopy—examples from tardigrade development. Integr Comp Biol 46:151–161. Jager M, Murienne J, Clabaut C, Deutsch J, Le Guyader H, Manuel M. 2006. Homology of arthropod anterior appendages revealed by Hox gene expression in a sea spider. Nature 441:506–508. Jørgensen A, Faurby S, Hansen JG, Møbjerg N, Kristensen RM. 2010. Molecular phylogeny of Arthrotardigrada (Tardigrada). Mol Phylogenet Evol 54:1006–1015. Kristensen RM. 1976. On the fine structure of Batillipes noerrevangi Kristensen, 1976. 1: Tegument and molting cycle. Zool Anz 197:129–150. Kristensen RM. 1981. Sense organs of two marine arthrotardigrades (Heterotardigrada, Tardigrada). Acta Zool 62:27–41. Kristensen RM. 1982. The first record of cyclomorphosis in Tardigrada based on a new genus and species from Arctic meiobenthos. Z Zool Syst Evol 20:249–270. Kristensen RM. 2003. Comparative morphology: Do the ultrastructural investigations of Loricifera and Tardigrada support the clade Ecdysozoa? The new panorama of animal evolution. Proc 18th Int Congr Zoology, Athens, Greece. pp 467–477. Kristensen RM, Higgins RP. 1984a. A new family of Arthrotardigrada (Tardigrada: Heterotardigrada) from the Atlantic Coast of Florida, U.S.A. Trans Am Microsc Soc 103:295–311. Kristensen RM, Higgins RP. 1984b. Revision of Styraconyx (Tardigrada: Halechiniscidae) with descriptions of two new species from Disko Bay, West Greenland. Smithsonian Contrib Zool 391:1–32. Kristensen RM, Higgins RP. 1991. Kinorhyncha. In: Harrison FW and Ruppert EE, editors. Microscopic Anatomy of Invertebrates, Volume 4: Aschelminthes. New York: Wiley-Liss, Inc. pp 377–404; Chapter 10. Mallatt JM, Garey JR, Shultz JW. 2004. Ecdysozoan phylogeny and Bayesian inference: First use of nearly complete 28S and 18S rRNA gene sequences to classify the arthropods and their kin. Mol Phylogenet Evol 31:178–191. Journal of Morphology Marcus E. 1929. Tardigrada, In: Bronn HG, editor, Klassen und Ordnungen des Tier-reichs. Akademische Verlagsgesellschaft, Leipzig. Vol. 5, Section 4, Part 3:1–609. Mayer G, Hartzsch S. 2007. Immunolocalization of serotonin in Onychophora argues against segmental ganglia being an ancestral feature of arthropods. BMC Evol Biol 7:118, 1–9. Mayer G, Whitington PM, Sunnucks Pflüger HJ. 2010. A revision of brain composition in Onychophora (velvet worms) suggests that the tritocerebrum evolved in arthropods. BMC Evol Biol 10:255, 1–9. McInnes SJ. Tardigrada. In: Marguelis L, Chapman M, editors. Five Kingdoms: An Illustrated Guide to the Phyla of Life on Earth, 4th ed. (in press). Müller MCM. 2006. Polychaete nervous systems: Ground pattern and variations-cLS microscopy and the importance of novel characteristics in phylogenetic analysis. Integr Comp Biol 46:125–133. Møbjerg N, Dahl C. 1996. Studies on the morphology and ultrastructure of the Malpighian tubules of Halobiotus crispae Kristensen, 1982 (Eutardigrada). Zool J Linn Soc 116:85–99. Møbjerg N, Jørgensen A, Eibye-Jacobsen J, Halberg KA, Persson D, Kristensen RM. 2007. New records on cyclomorphosis in the marine eutardigrade Halobiotus crispae. J Limnol 66 (Suppl. 1):132–140. Møbjerg N, Halberg KA, Jørgensen A, Persson D, Bjørn M, Ramløv H, Kristensen RM. 2011. Survival in extreme environments—On the current knowledge of adaptations in tardigrades. Acta Physiol 202:409–420. Nielsen C. 2001. Animal evolution. Interrelationships of the Living Phyla, 2nd ed. Oxford University Press, Oxford. pp 226–231. Nielsen C. 2011. Animal evolution. Interrelationships of the Living Phyla, 3rd ed. Oxford University Press, Oxford. pp 265– 269. Persson D, Halberg KA, Jørgensen A, Ricci C, Møbjerg N, Kristensen RM. 2011. Extreme stress tolerance in tardigrades: Surviving space conditions in low earth orbit. J Zool Syst Evol Res 49:90–97. Plate L. 1889. Beiträge zur Naturgeschichte der Tardigraden. Zool Jb Anat Ontog 3:487–550. Pugh PJA, McInnes SJ. 1998. The origin of Arctic terrestrial and freshwater tardigrades. Polar Biol 19:177–182. Ramazzotti G, Maucci. 1983. Il Phylum Tardigrada. Mem Ist Ital Idrobiol 41:1–1012. Ramsköld L, Hou X. 1991. New early Cambrian animal and onychophoran affinities of enigmatic metazoans. Nature 351: 225–228. Renaud-Mornant J. 1975. Deep-sea Tardigrada from the Meteor Indian Ocean Expedition. ‘Meteor’ Forschungsergebnisse. D 21:54–61. Renaud-Mornant J. 1982. Species diversity in marine Tardigrada. In: Nelson DR, editor. Proceedings of the Third International Symposium on the Tardigrada, August 3–6, 1980, Johnson City, Tennessee. Johnson City: East Tennessee State University Press. pp 149–178. Richter S, Loesel R, Purschke G, Schmidt-Rhaesa A, Sholtz G, Stach T, Vogt L, Wanninger A, Brenneis G, Döring C, Faller S, Fritsch M, Grobe P, Heuer CM, Kaul S, Møller OS, Müller CHG, Rieger V, Rothe BH, Stegner MEJ, Harzsch S. 2010. Invertebrate neurophylogeny: Suggested terms and definitions for a neuroanatomical glossary. Front Zool 7:29, 1–49. Rota-Stabelli O, Kayal E, Gleeson D, Daub J, Boore J, Telford M, Pisani D, Blaxter M, Lavrov D. 2010. Ecdysozoan mitogenomics: Evidence for a common origin of the legged invertebrates, the Panarthropoda. Genome Biol Evol 2:425–440. Ruppert EE. 1982. Comparative ultrastructure of the Gastrotrich pharynx and the evolution of myoepithelial foreguts in Aschelminthes. Zoomorphology 99:181–220. Ruppert EE, Barnes RD. 1994. Invertebrate Zoology, 6th ed. Saunders College Publ. Harcourt Brace and Company, Orlando, Florida. 1056 p. NEUROANATOMY OF Halobiotus crispae Schmidt-Rhaesa A, Bartolomaeus T, Lemburg C, Ehlers U, Garey JR. 1998. The position of the Arthropoda in the phylogenetic system. J Morph 238:263–285. Scholtz G. 2002. The Articulata hypothesis—or what is a segment? Org Divers Evol 2:197–215. Scholtz G, Edgecombe GD. 2006. The evolution of arthropod heads: reconciling morphological, developmental and palaeontological evidence. Dev Genes Evol 216:395–415. Storch V, Ruhberg H. 1993. Chapter 2. Onychophora. In: Harrison FW, Rice ME, editors. Microscopic Anatomy of Invertebrates, Volume 12: Onychophora, Chilopoda and Lesser Protostomata. New York: Wiley-Liss, Inc. pp 11–56. 1245 Wiederhöft H, Greven H. 1996. The cerebral ganglia of Milnesium tardigradum Doyère (Apochela, Tardigrada): Three dimensional reconstruction and notes on their ultrastructure. Zool J Linn Soc (Lond) 116:71–84. Wright JC. 2001. Cryptobiosis 300 years on from van Leuwenhoek: What have we learned about tardigrades? Zool Anz 240:563–582. Zantke J, Wolff C, Scholtz G. 2008. Three-dimensional reconstruction of the central nervous system of M. hufelandi (Eutardigrada, Parachela): implications for the phylogenetic position of Tardigrada. Zoomorphology 127:21– 36. Journal of Morphology