Trapped in freshwater: the internal anatomy of the entoproct Loxosomatoides sirindhornae
© Schwaha et al; licensee BioMed Central Ltd. 2010
Received: 27 October 2009
Accepted: 4 February 2010
Published: 4 February 2010
Skip to main content
© Schwaha et al; licensee BioMed Central Ltd. 2010
Received: 27 October 2009
Accepted: 4 February 2010
Published: 4 February 2010
Entoprocta is a small phylum of tentacle-bearing spiralian lophotrochozoans that comprises mainly marine representatives, with only two known freshwater species. One of them, Loxosomatoides sirindhornae Wood, 2005 was only recently described, and detailed information on its morphology including adaptations to life in freshwater are unknown. We analyzed the internal anatomy of L. sirindhornae using serial semi-thin sections, 3D reconstruction, as well as immunocytochemistry and confocal laserscanning microscopy.
The nephridial system shows high complexity, strikingly similar to that of Urnatella gracilis, the only other known freshwater entoproct. It is composed of 105-120 large flame-bulb terminal organs that occur in the stalk and calyx. In the stalk they terminate in the epidermis, whereas efferent ducts in each terminal organ in the calyx lead to large, paired terminal ducts that fuse close to the central nervous system and open into the atrium by a nephridiopore. Compared to other stolonate entoprocts, L. sirindhornae shows a different stalk-calyx junction by possessing only a single, multicellular canopy instead of a stack of star cells. A sphincter muscle is situated below the diaphragm of the stalk. The remaining musculature is concentrated in the stalk, while the calyx musculature is sparsely developed. The central nervous system is dumbbell-shaped as in basal entoprocts.
The nephridial system probably has mainly osmoregulatory function. Previous studies have shown that L. sirindhornae is unable to cope with higher salinities, suggesting that its adaptation to freshwater has reached such a high degree that it is unable to 'turn off' the nephridial system in higher salinities. The current data available show that the architecture of internal organ systems such as the musculature or the calyx-stalk junction hold more promising information for taxonomic and perhaps even evolutionary inferences in Entoprocta than external characters such as spination. Contrary to previous investigations, the longitudinal calyx musculature of the genus Loxosomatoides should not be classified as generally strong or conspicuous, since its extent and site of insertion differs between species.
Entoprocta or Kamptozoa is a small phylum of sessile, marine, filter-feeding animals which most likely cluster with other spiral cleaving taxa within the protostomian superclade Lophotrochozoa . To date, approximately 180 described species belonging to four families are recognized [2, 3]. The family Loxosomatidae is the most species rich and comprises only solitary forms, while the remaining families Loxokalypodidae, Pedicellinidae and Barentsiidae are colonial. Only two species within the phylum live in freshwater habitats, the widespread Urnatella gracilis (Barentsiidae) and the recently described Loxosomatoides sirindhornae Wood, 2005. This species was discovered in two rivers from central Thailand and forms stolonate colonies with segmented stalks . Since the genus Loxosomatoides belongs to the Pedicellinidae [4, 5], it must be assumed that entoprocts have invaded freshwater at least twice independently in the course of their evolution .
The morphology, ecology and reproduction of Urnatella gracilis has been described in considerable detail [6–11]. As in other invertebrate phyla, freshwater environments have led to specific adaptations of this species such as, for example, the nephridial system, which is composed of numerous large terminal organs found in the calyx and stalk. The terminal organs are followed by thin contorted tubules that in the stalk end directly in the epidermis and in the calyx lead into large terminal ducts on the lateral sides. The terminal ducts meet in the middle below the atrial floor where they exit by a pore . Additional adaptations to freshwater in U. gracilis are formation of hibernacula , which are also found in L. sirindhornae . More adaptations are to be expected and require the investigation of internal structures, particularly the nephridial system. In addition, species differences within the genus Loxosomatoides and the related genus Myosoma are few . Accordingly, this study focuses on the anatomy of L. sirindhornae including possible further adaptations to freshwater as well as new morphological characters applicable to taxonomic inferences.
Specimens of Loxosomatoides sirindhornae, Wood 2005 were collected from ropes dangling into the river Mae Klong close to the city Kanchanaburi, Thailand in February 2009. Pieces of rope with L. sirindhornae were transferred to the laboratories of the Department of Environmental Sciences of the Kasetsart University in Bangkok, where the samples were relaxed with a 1% solution of MgCl2 [see e.g. ]. However, the tentacles remained retracted in most specimens.
Specimens were fixed in a 2% glutaraldehyde solution in 0.01 M sodium-cacodylate buffer with a pH of 7.4 for 1 hour at room temperature. Subsequently, they were rinsed three times in the buffer and removed from the rope with forceps and sharp needles. Samples were afterwards postfixed in 1% osmiumtetroxide in distilled water for 1 hour at room temperature. Dehydration was carried out using acidified dimethoxypropane before embedding in Agar Low Viscosity resin using acetone as intermediate. Serial sectioning of eight specimens was conducted as described by Ruthensteiner  on a Reichert Ultracut S microtome at a sectioning thickness of 1 μm. Sections were stained with toluidine blue. Digital images were captured with a Leica DMRXA microscope equipped with an Evolution MP digital camera or a Nikon Eclipse E800 microscope equipped with a Nikon DS5-U1 camera and edited in Adobe Photoshop CS 2 or 3 (Adobe, San Jose, CA, USA). One complete longitudinal section series was taken for reconstructing the whole animal, while parts of a frontal section series was used for a detailed reconstruction of the nervous system and surrounding tissues such as the nephridial system. For 3D reconstruction, the image stacks were converted to greyscale and reduced in size prior to the import into the visualisation software Amira 4.1 (Mercury Computer Systems, Chelmsford, MA, USA) [see  for details]). Alignment was conducted automatically and where necessary corrected manually prior to manual segmentation of different organ systems. A surface of the segmented structure was created and optimised by iterated steps of simplification which was followed by smoothing. Snapshots of the reconstructions were taken with the Amira software.
Specimens were fixed in 4% paraformaldehyde in 0.01 M phosphate buffer (PBS) containing 0.01% NaN3 for 1 hour at room temperature and thereafter rinsed three times in PBS. Samples were stored at 4°C in this solution until further processing. After three additional washes in PBS, the samples were permeabilized in PBS containing 4% Triton-X (PBT) for approximately 1 hour prior to staining for F-actin. This was followed by overnight incubation at 4°C in a 1:40 dilution of Alexa Fluor 488 phalloidin (Invitrogen, Molecular Probes, Eugene, OR, USA) in PBT. Specimens were then rinsed several times in PBS and mounted in Fluoromount G (Southern Biotech, Birmingham, AL, USA) on standard microscope slides. For alpha-tubulin staining, unspecific binding sites were first blocked in 6% normal goat serum (NGS) in PBT (block-PBT) overnight at 4°C. A mouse anti-acetylated alpha-tubulin antibody (Sigma, Brøndby, Denmark) was applied at a concentration of 1:300 in block-PBT for 24 hours. After several washes in block-PBT for 6 hours, an AlexaFluor 488 conjugated goat-anti mouse secondary antibody (Invitrogen, Molecular Probes, Eugene, OR, USA) was applied at a concentration of 1:200 (diluted in block-PBT) for 24 hours. Then, the samples were rinsed several times in PBS for about 6 hours before embedding in Fluoromount G.
Analysis and image acquisition was performed on a Leica DM IRBE microscope (Leica Microsystems, Wetzlar, Germany) equipped with a Leica TCS SP2 confocal unit. Confocal image stacks were scanned in steps of 0.5 to 1 μm along the Z-axis. Maximum intensity projections were generated with the supplied Leica software.
The outer surface is covered by a cuticle that is thicker and two-layered at the lateral and aboral side of the calyx, as seen in its staining properties on sections (Fig. 2a). Several protuberances or small spines cover the outer cuticle (Fig.2a-c). These are concentrated on the aboral side and are proportionally larger in smaller specimens than in fully grown individuals (Fig. 2b, c). They show staining properties similar to the thickened aboral shield of the calyx. The remaining cuticle of the pedicle shows several rows of lamellae (Fig. 2b-e). A cuticular diaphragm constricts the connection between stalk and calyx. Centrally, several membranous tubules penetrate the opening and spread radially into the calyx. A distinct star-cell complex is not present; instead, a canopy-shaped structure composed of multiple cells is situated at the calyx-stalk junction. It bears several thin radial extensions where the tubular extensions from the stalk transit (Fig. 2f, g).
Apart from nerves and parts of the nephridial system, which will be described below, the body cavity of the calyx is filled with numerous polygonal, sometimes elongated cells. Several thin cytoplasmatic protrusions extend from each cell and interconnect the cells or attach to the bodywall and other organ systems such as the digestive tract or the nephridial system (Figs. 3a, b, d; 4a, b; 5b). Close to the body wall some of these cells exhibit a different appearance by possessing several vesicles (Figs. 3b, d; 5b).
The protonephridial system is probably the most complex organ system in L. sirindhornae. In adult specimens, approximately 90-100 multi-ciliated, bulbous terminal organs with two nuclei are found in the calyx and in the stalk (Figs. 2b; 3a-c, e, f). In the calyx, terminal organs are concentrated at the oral side. One or two terminal organs lie at the base of each tentacle (Fig. 3c), while the remaining terminal organs are located around the esophagus. A few terminal organs are situated at the base of the calyx, close to the connection to the stalk (Fig. 2g) and aborally of the atrial wall. Each terminal organ leads to a thin, slightly convoluted capillary duct (Figs. 3f; 5a, c) that ends in a broad terminal duct. The terminal duct consists of two main lateral branches, each of which starting almost at the distalmost end next to the atrium (Figs. 1b-e, h; 3f). They continue proximally on the lateral sides until the height of the CNS, where they bend medially and then distally to unite into a single short duct that opens into the atrium (Figs. 1d, e, h; 3b; 4c; 5a-c). Two short branches split off where the main branches run medially and continue proximally on the lateral sides of the esophagus. At least parts of the terminal ducts seem to possess long cilia (Fig. 3d, f).
The size of the brood pouch largely depends on the size of the embryos. With smaller embryos, the pouch is quite small and has a separate duct that traverses proximo-frontally and terminates at the level of the gonads with an opening close to the nephridial pore (Fig. 1b, d, g). When fully developed, larvae are present and, depending on their position, the brood pouch can extend almost to the lateral sides of the mother animal (Fig. 6c). Additionally, large and expanded brood pouches have been found where no separate duct to the atrium was present. We did not find male gonads in any of the specimens investigated.
In the calyx only few muscle groups are present. Most prominent is the ring musculature of the soft tentacular membrane on the oral side, that acts as a sphincter for enclosing the atrium (Fig. 7a, b, i). Each tentacle possesses two paired muscles, conspicuous main lateral tentacle muscles and very thin inner tentacle muscles (Fig. 7b). The digestive tract only shows ring musculature at the lower part of the esophagus and at the intestine, as well as an anal sphincter (Figs. 4f; 7a, b, f, g). The last group of calyx muscles are a few longitudinal muscle fibres that originate at the proximal bodywall and attach to the lateral sides of the esophagus (esophageal retractors) (Fig. 7f, h).
The general morphology of Loxosomatoides sirindhornae resembles that of other pedicellinids showing a muscular, unsegmented stalk and a calyx. In L. sirindhornae the calyx is obliquely oriented with the tentacle crown tilted orally as also seen in several congeners . Such an orientation is found in most loxosomatids, but the barentsiid Urnatella gracilis likewise bears a very similar shape of the calyx . The digestive tract consists of a wide mouth opening followed by the esophagus, a voluminous stomach, an intestine and a rectum. Thus, the general composition and also the structural details (i.e. ciliation, appearance of the epithelia) are more or less identical to other entoprocts .
The morphology of mesenchyme cells in the body cavity has been described as amoeboid for Pedicellina cernua , which was subsequently adopted for all entoprocts in most compendia on the phylum [16, 17]. However, their existence was subsequently rejected by Emschermann . Also in L. sirindhornae these cells are not amoeboid, but merely possess several spinuous cytoplasmic connections to other cells and organ systems. Since they are most abundant around the esophagus, they appear to act as a stabilizing unit that ensures all organs are kept in place when the animal contracts the esophageal retractors. Adding to the original description , it should be noted that a diaphragm with a primitive 'star-structure' is present in L. sirindhornae. Septa within stolons have previously only been detected at hibernacula in L. sirindhornae, whereas they also regularly occur between individual zooids in the stolon.
With the strict separation of calyx and stalk in most colonial entoprocts and the increased rigidity by a thick cuticle, several muscle systems have become dispensable. Therefore, the muscular system of L. sirindhornae is rather sparsely developed compared to muscular systems of the Loxosomatidae [27, 28]. Most of the occurring muscles, however, are comparable to loxosomatids. Merely the sphincter below the stalk-calyx diaphragm represents a new muscle, which to our knowledge has not been reported from other colonial stolonates. It appears most probable that the sphincter interacts with the star-shaped structure as a circulatory organ, similar to colonial species with a star cell complex.
The tentacle musculature with outer main muscles and thinner inner muscles is identical to the analysed species of the Loxosomatidae [27–29]. There are two atrial ring muscles in both loxosomatids, but only a single thick muscle in L. sirindhornae. The muscles associated with the digestive tract (esophageal ring muscles, intestinal sphincter and anal sphincter) are similar to Loxosomella sp. and Loxosoma sp. [14, 27, 28]. Remaining muscles of the digestive tract as found in loxosomatids, i.e., rectal retractors and rectum musculature, are not present. The esophageal retractors found in L. sirindhornae are most probably derived from longitudinal stalk muscles that extend into the calyx as in loxosomatids [27, 28]. With the formation of a cuticular diaphragm at the calyx-stalk junction, these muscles were separated from the stalk and gained new attachment sites. Functionally, they probably act as atrial depressors enlarging the atrial space during contraction, since the inflexible cuticle of the calyx is unable to give way. This longitudinal calyx musculature has been previously referred to as atrial retractors. Since these muscles are considered remnants of the continuous stalk-calyx musculature in loxosomatids , they are potentially important to clarify the internal relationships of entoproct taxa. Accordingly, Loxosomatoides and related genera are regarded as sister groups to the remaining pedicellinids by the presence of conspicuous and strong longitudinal calyx musculature as well as the structure of the calyx-stalk junction  (Fig. 8). The longitudinal calyx musculature is thought to have been progressively lost in more derived genera such as Pedicellina and Barentsia . However, care should be taken with these assumptions especially since differences in the muscular system of the calyx and stalk seem to occur between the different species. In some of the species the calyx musculature is not as strong or conspicuous as previously assumed. Also, the point of their insertion seems to differ between species. In L. sirindhornae few and in L. colonialis multiple longitudinal muscles run orally from the base of the calyx to the esophagus [, this study]. For Chitaspis athleticus, Annandale  described two muscle strands in the oral part that traverse in the body wall. One of these muscles probably inserts at the stomach wall. Only for L. laevis strong oral calyx muscles have been described that insert at the atrial wall.
The presence of oblique stalk musculature has been used as taxonomic character for the genus Myosoma, but it also occurs in Chitaspis athleticus. The genus Chitaspis was characterised by the absence of stalk muscles on the aboral side . In several species of Loxosomatoides (L. colonialis, L. laevis, L. sirindhornae) the oral stalk musculature is somewhat thicker than the aboral musculature [, this study]. Whether the aboral stalk musculature in Chitaspis athleticus is absent or only thinner as in loxosomatoids and whether or not this genus should be synonymised with Loxosomatoides as recently suggested  remains unresolved until thorough reinvestigations will be performed.
Adult entoprocts usually possess a single pair of protonephridia between the stomach and esophagus, that opens into the atrium by separate nephropores in loxosomatids or a single pore in pedicellinids and barentsiids [14, 30]. By contrast, freshwater species such as Urnatella gracilis and Loxosomatoides sirindhornae show a complex branched nephridial system [, this study]. The systems of both species show striking similarities in the calyx and the stalk. In the calyx, both possess long terminal ducts that branch off small capillary ducts to numerous terminal organs. The terminal ducts are much longer and branched in L. sirindhornae, extending even into the lateral border of the tentacle crown. In U. gracilis they are restricted to the area between stomach and esophagus and are arranged in the shape of an inverted Y . The protonephridia in the stalk show a similar structure as in the calyx, with bulbous terminal organs with contorted tubules. In both species they exit through the epidermis and in U. gracilis function in ion regulation . Direct evidence for the function in L. sirindhornae is lacking, but most likely similar to U. gracilis. The two freshwater species are the only known species possessing protonephridia in the stalk. Many other colonial entoprocts have several pore organs with rudimentary cilia underneath the cuticle of the stalk which act as ion regulatory organs . It has been assumed that they represent rudimentary protonephridia and that the ancestor of colonial entoprocts possibly had several protonephridia along the body surface . Although Loxosomatoides appears to be a sister group to the remaining stolonate entoprocts, it seems rather unlikely to assume such an ancestor from the current state of knowledge. Unless similar conditions will be found in marine or brackish species of this genus, it seems more likely that protonephridia in the stalk are a special adaptation to freshwater.
Terminal organs in entoprocts are always flame-bulb nephridia composed of two cells and are usually small and thin . In the freshwater species, however, they are large and bulbous, not to mention numerous . As previously mentioned, the freshwater species currently belong to two different families and convergent adaptation to freshwater is most likely. It seems that a highly sophisticated nephridial system is a prerequisite for conquering freshwater in these species. Even developing trochophore-like larvae of L. sirinhornae possess four large terminal organs (Schwaha: unpublished observations), while all other known larvae, even that of Urnatella gracilis, have only one pair [7, 14, 32]. In U. gracilis the nephridial system appears to have solely osmoregulatory function. Furthermore, exposure to different salinities has shown that the terminal organs of U. gracilis decrease their beating frequencies in higher salinities and even stop beating at a certain concentration . Albeit direct observation is missing, it is very likely that the protonephridial system functions similarly in L. sirindhornae and that, as in other entoprocts, all excretory processes take place in the stomach roof and rectum as indicated by their similar structure [8, 33]. Similar to U. gracilis, L. sirindhornae is highly intolerant towards higher salinities . Obviously, both freshwater entoprocts are hypertonic towards their surrounding medium and remove inflowing water with the protonephridial system. Despite that, both species are so well adapted to freshwater, that they are seemingly unable to properly 'turn off' this water-removal system. Unfortunately, concise information on the nephridial system of brackish species, e.g., of other loxosomatoids, is still lacking, but following our line of thought it is tempting to speculate that these species represent an intermediate type concerning the differentiation of the nephridial system.
The nervous system of entoprocts has only been the subject of very few detailed analyses. The dumbbell-like shape of the CNS, as for example found in representatives of the solitary Loxosomatidae, probably represents the plesiomorphic condition [27, 34]. In some of the investigated species of the colonial Barentsiidae and Pedicellinidae an oval CNS has been described [15, 35]. It has therefore been assumed that the oval ganglion results from a fusion of the paired ganglia . Loxosomatoides sirindhornae still shows the plesiomorphic condition. Since the genus is considered as an early offshoot of stolonate species, this finding is congruent with the sparse solid entoproct phylogenies currently available . However, the other freshwater species, Urnatella gracilis, likewise exhibits a dumbbell-shaped CNS , although it is currently recognized as a derived form of Barentsia [2, 12]. Accordingly, the dumbbell-like shape would have re-evolved from the oval CNS of barentsiids. This scenario seems unlikely and it is more parsimonious to assign U. gracilis a more basal position within stolonate entoprocts, although phylogenetic analyses that corroborate such a scenario are currently lacking.
The general organisation of the female genital tract resembles those of other entoprocts [14, 15]. As in L. sirindhornae, maturation of few germinal cells in the ovary is common , but in Urnatella gracilis only a single oocyte matures at a time . Glandular portions of the oviduct or 'shell glands' are common and produce egg envelopes and attachment cords [14, 33]. While the shell gland is a compact large organ in Loxosomella elegans , it appears grape-shaped in Loxosomatoides sirindhornae, with several grouped cells that each lead into the oviduct by a thin cellular extension. As in most other species, direct evidence of fertilization is lacking, but can be assumed to take place in the ovary before the formation of the egg envelope .
By the absence of male gonads in all analysed specimens, it can be concluded that L. sirindhornae is a protandric hermaphrodite, similar to other members of the genus (L. colonialis and L. evelinae; see [22, 36]). All adult specimens contained several embryos or larvae brooded inside the brooding pouch which occupies extensive space that might make further fertilization events unnecessary.
Unlike other freshwater organisms (e.g. ectoprocts, bivalves and rotifers), the entoprocts show a drastically modified nephridial system. This probably explains why only two species have hitherto been found in freshwater. Previous studies of the Loxosomatoides-Myosoma species complex were mainly based on external features such as spines and ornamentation of the aboral shield . Our study shows that the internal anatomy of these species such as the myoanatomy, structure of the stalk-calyx junction as well as the nephridial system holds more promising information for taxonomic and perhaps even evolutionary inferences. Thus, besides external features, more attention should be paid to sectioning methods and muscle staining analysed with confocal microscopy. In particular, features such as spines are troublesome as they might underlie ecological factors and therefore may vary greatly in size between individual zooids, as, e.g., seen in L. colonialis . Also, with the exception of L. evelinae, all species of Loxosomatoides investigated so far possess polygonal ornamentation on the aboral shield, which is lacking in L. sirindhornae. Therefore, it should be questioned if the investigated species belongs to Loxosomatoides or not, especially since it also shows differences with respect to the, probably primitive, differentiation of the stalk-calyx junction. This junction, as well as differences in the musculature, seem to yield most information relevant for taxonomic analyses, but these characters are also important for understanding the evolution of colonialism in Entoprocta. Comparative data is unfortunately mostly missing or in definite need for revision for this neglected phylum, but hopefully will become available in the near future.
TS and TW gratefully acknowledge the generous logistical support for this research by the Department of Environmental Science, Faculty of Science, Kasetsart University, Bangkok. We especially thank the department chair, Dr. Jukkrit Mahujcharyawong, as well as Dr. Patana Anurakpongsatorn, and Dr. Ratcha Chaichana. Without their gracious hospitality this work would not have been possible. TS' trip to Thailand was supported by the KWA-scholarship of the University of Vienna. TS is indebted to Manfred Walzl for his continuous support and encouragement. Research in the lab of AW is funded by the EU Early Stage Research Training Network MOLMORPH (contract grant number MEST-CT-2005 - 020542). The authors would like to thank Claus Nielsen for providing literature and for valuable comments on the manuscript.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.