Ectoparasitic diplozoid monogeneans exhibit considerable functional and morphological adaptations to their parasitic lifestyle. This study focuses on body wall organisation and related structures involved in niche searching, host confrontation and self-protection against the environment and excretory/secretory processes.
Self-protection against the environment and excretory/secretory processes
The Neodermata have evolved a unique body covering (tegument) on their surface with numerous functions, including provision of external body support, contributing to the transfer of nutrients and their conversion to energy and protection against host immune responses, enzymes, secretion, excretion and osmoregulation, as well as providing a sensory function . The tegument is the primary interface essential for host-parasite interaction and, as such, exhibits unique modifications related to both nutrition and the worm’s parasitic strategy. Generally speaking, the tegument and related structures of parasitic flatworms exhibit significant variability in organisation and basic function. Unlike cestodes, digeneans and some other monogeneans , where the external surface of the tegument is covered with highly specialised microvilli (microtriches) projecting from the apical plasma membrane essential for nutrient absorption and elimination of waste materials, the surface of the apical plasma membrane in E. nipponicum is smooth. However, the tegument of E. nipponicum, Paradiplozoon homoion and a few other species is equipped with shallow pits [11, 15, 18]. The prominent annular ridges and extensions of the E. nipponicum hindbody are considered to be important for attachment and securing the position of the parasite among the secondary gill lamellae, somewhat analogous to a zip fastener . Less prominent tegumentary folds and three highly mobile lobes have been described on the haptor of P. homoion . The tegument of platyhelminths is reported as apically bound by a plasma membrane (generally described as a trilaminar), underlain by a thick nuclear-free syncytium layer packed with numerous mitochondria, secretory bodies and other inclusions, while the underside of the syncytium is lined with a basal plasma membrane. A fibrous basal lamina supports the entire structure. The basal plasma membrane produces numerous invaginations toward the syncytium, thereby forming a basal labyrinth . Although we confirmed the general organisation of the tegument in E. nipponicum using standard TEM methods, freeze-etching revealed some differences in membrane organisation. The apical plasma membrane previously described as “trilaminar” actually corresponds to a plasma membrane closely associated with a subjacent membrane and separated by a dense layer of proteins. Similar observations were reported in a freeze-fracture study on the tegument of Schistosoma . Obviously, these two membranes are difficult to detect under conventional TEM. In addition, although a membrane-like structure was detected in ultrathin sections, no basal plasma membrane was detected in the proximity of freeze-fractured basal lamina (between the innermost site of the syncytium and the basal lamina); instead, the outer surface of the basal lamina was covered with a number of separating vesicles and projections of endoplasmic reticulum. Prominent body wall musculature separates the basal lamina from the cell bodies (the so-called subtegumentary cells, precursors or cytons) with prominent nuclei that are linked to the syncytium via numerous cytoplasmic projections and reach up to the body parenchyma [17, 19].
Smyth and Halton  considered the tegument of monogeneans to be a secretory epithelium, producing various vesicular and granular inclusions for dispersal on the surface. These authors also suggested that some synthetic activity occurs directly in the syncytium, due to the presence of abundant complexes of granular endoplasmic reticulum and Golgi apparatus in this area. In addition to the endoplasmic reticulum projections on the outer surface of the basal lamina, freeze-etching in this study revealed the presence of vesicles connected to the subjacent membrane, or even interrupting both superficial membranes covering the outer site of the syncytium (Fig. 6C-E). As the presence of accumulated vesicles or bodies within the syncytium has been observed in replicas and histological sections, we consider these to be secretory. Similar vesicles, considered to be exocytotic, were detected at the apical surface of the syncytium in ultrathin sections of Pricea multae . In cestodes, pore-like openings in the protoplasmic face of the basal plasma membrane were considered to be pits from which the membrane-bound channels, extending into the syncytium and opening at its apical surface, emerged, and that these may facilitate transfer between the parenchyma and the parasite surface . Accordingly, other studies identified large cells (considered glands) located within the parenchyma containing abundant electron-dense secretory vesicles and mitochondria, with an opening extending from the subtegumental musculature to the outer syncytial layer . The contents of the vesicles are believed to release externally at the apical surface through these openings. Though the chemical composition and role of such secretory substances (which may differ among species) remains unclear, one theory attributes them to the secretion of a protective glycocalyx, as known for digeneans . In this study, we observed no glycocalyx layer or coat covering the tegument surface; instead, we revealed a dense layer of proteins located between the apical plasma and subjacent membranes. It should be mentioned, however, that the glycocalyx layer could have been washed off during the numerous sample processing steps for electron microscopy and freeze-etching. In addition, this layer is often not visible until stained with ruthenium red, a cationic reagent for electron microscopy [e.g. Fig. 5B vs. 5F in 23]. Interestingly, another study on diplozoids reported the existence of so-called scales, located mainly on the hindbody ridges just beneath the apical plasma membrane . However, we believe that these “scales” correspond to areas of the trilaminar surface with more oblique sectioning, comprising two membranes (apical plasma and subjacent membrane) separated by a dense protein layer (as mentioned above) of various thickness. Our assumption is further supported by the fact that the trilaminar layer in the majority of shown TEM micrographs of higher magnification revealed areas of differing thickness . These formations, corresponding to the local thickenings of the trilaminar surface and more prominent in the hindbody ridges, were also present in our micrographs (compare Figs. 3D-E vs. 7J). Such formations may be due to increased accumulation of proteins in the middle layer and may be related to the (potentially) increased adhesion of this hindbody part.
In accordance with the study on P. homoion , hydrochloric carmine staining for CLSM and conventional histological staining of E. nipponicum revealed numerous giant cells with prominent nuclei between the GMO, below the buccal suckers and surrounding the pharynx. As these exhibited some differences from parenchymal cells, such as cytoplasm intensively staining with haematoxylin (stains basophilic structures) and netlike chromatin radially attached to the nucleolus (vs. more homogeneous chromatin distribution in parenchymal cells; shown in Fig. 2C), they are generally considered to be unicellular glands . Further, they are at least two times larger than parenchymal cells and their nucleolus stains intensively pink with haematoxylin-eosin staining, in contrast to the purple nucleoli in parenchyma (personal observation). Similar gland cells, described in other monopisthocotyleans, have been shown to secrete a glutinous or sticky material most likely involved in adhesion to the host’s gills, though some of the secretory products may also modulate a host’s immune reaction [22, 24]. Alternatively, the gland cells could secrete pheromones . GMO are also likely to be involved in secretion, though their presence has only been confirmed in developmental stages of E. nipponicum following fusion of diporpae . These apparently hollow organs appear to open ventrally into the corner of the mouth cavity and possibly act as product reservoirs for neighbouring gland cells. Although their glandular function has not been confirmed thus far, the GMO are considered be a part of the digestive tract (e.g. [6, 25]). While we failed to observe any connection with other internal organs in a previous study , CLSM using transmission light mode in this study revealed a single long curved canal starting from the GMO and leading to the area above the buccal sucker (Fig. 2B). Previous work focused on the nerve system of E. nipponicum has suggested that the GMO are active structures as they are adequately innervated . It has previously been speculated that the GMO act either during feeding or serve as the reservoir of secretions produced in gland cells found near the GMO ; however, the absence of GMO in non-fused diporpae exhibiting food intake is at odds with their involvement in parasite feeding. Two apical groups of symmetrically organised, F-actin rich circular structures of unknown function were repeatedly detected in E. nipponicum and P. homoion in the area of apical round projections observed under SEM [6, 15]. Accumulations of numerous drop-like gland cells located in the forebody apical area and between the GMO suggests that these circular structures could in fact be the openings of their drainage canals. However, their involvement in neurosensory activity cannot be ruled out as a pair of sensory nerves appear to interfere with these structures [5, 9]. While the central large circular structure could be the circular rim of a neurosensory receptor , the surrounding smaller circles more likely correspond to the outlet of joined glands, similar to those forming the anterior adhesive apparatus in other monogeneans [26-29]. Furthermore, smaller gland-like cells, located within the parenchyma and free secretory bodies (most likely secreted from a gland cell) in the syncytium, have been observed in the haptor region of E. nipponicum and P. homoion [14, 15, this study]. We assume that secretory bodies in this area serve to transport adhesives from the gland cells to the surface of the haptor. Similar dense secretory bodies, produced by uninucleated gland cells and released to the surface via ducts, have been described in other species [22, 30, 31]. Further, monogeneans equipped with haptoral anchors [30, 31] display similar dense bodies in the syncytial multinucleated haptor gland (hamulus gland). These produce a liquid exudate that is stored in the hook reservoir until leaking into the sleeve cavity to bath the naked hook. The role of the hamulus gland is thought to be histolytic, facilitating penetration of gill tissue . Other studies have proposed that a secretion exocytosed onto its surface gives the haptor adhesive properties .
A recent study on E. nipponicum combining laser capture microdissection with mass spectrometry  identified 2,059 proteins (including 72 peptidases and 33 peptidase inhibitors) in the intestine (1,978), tegument (1,425) and parenchyma (1,302). It is possible that the tegumental proteins may be located within the dense protein layer visualised in freeze-etching replicas and/or the above-mentioned tegument thickenings, while the parenchymatic proteins may accumulate within the giant, possibly secretory, cells and/or the secretory bodies.
The excretory (protonephridial) system of E. nipponicum is of typical organisation and similar to that in other platyhelminths. The role of the terminal cells in protonephridia (termed flame cells) is still not completely understood; however, they are likely to function in excretory/secretory processes and/or participate in maintenance of the osmotic environment . In addition to conventional TEM (described elsewhere [12, 35]) and CLSM-based tubulin labelling , this study is the first to use the freeze-etching technique to visualise the flame cells, their tufts and the protonephridial duct in monogeneans. Aside from the remarkably prolonged lamellae lining the protonephridial and the excretory ducts, the ultrastructure of the protonephridial system generally corresponds to that already published for diplozoids . These lamellae have previously been reported as an extensive reticulum formed by numerous interconnected cavities within the epithelium of the protonephridial duct . According to other works on platyhelminths [36, 37], lateral flames were observed within the protonephridial ducts. The organisation of the flame bulb corresponds to that observed in other monogeneans (except Udonella), aspidogastreans and digeneans [36, 38].
Niche searching and sensing the host environment
In addition to the posteriorly located haptor, most monogeneans are thought to use their anterior ends (for diplozoids, specifically the buccal suckers) for transient attachment to the host (e.g. during feeding or translocation on the host’s gills) . It is likely that all monogenean ectoparasites can change their location on the host via leech-like locomotion based on temporary attachment of the apical forebody during detachment and relocation of the haptor . In monogeneans studied thus far (e.g. Leptocotyle), this leech-like locomotion is usually achieved through cooperation between adhesive secretion (discussed above) and the haptor .
In E. nipponicum, the order of individual muscle layers appears to differ slightly from the generalised muscle organisation (i.e. outer circular, intermediate diagonal and inner longitudinal muscle fibres) of parasitic worms . In E. nipponicum, we documented outer circular and inner longitudinal muscles interwoven by diagonal muscle bundles in a basket-like manner. In P. homoion, similarly perpendicular muscles were anchored to the tegument . These appear to correspond to the so-called dorsoventral muscles , whose function may be related to the dorsoventral flattening of the worm. Also, of interest were the abundant muscle fibres located within the tegumentary ridges, which could make it easier for the worm to attach itself between the secondary gill lamellae.
While a basic diagram showing the nervous system of E. nipponicum was published as early as 1891 , later immunomicroscopic studies focusing on the central nerve elements in paired adults demonstrated peptidergic and serotoninergic innervation (via indirect immunocytochemistry), cholinergic components (via enzyme cytochemistry) and neuropeptide immunoreactivity at the subcellular level (via TEM immunogold) . In this study, in addition to identifying components of the central nervous system, we were able to visualise the network of peripheral nerves and innervation of sensory structures using α-tubulin labelling for CLSM. These results are generally consistent with our observations on P. homoion . Furthermore, this study was also able to provide visualisation of the peripheral nerve fibres, along with the innervation of sensory structures and clamps, using TEM and freeze-etching TEM. The fine structure of the nerve fibres corresponds with that observed in other studies on monogeneans [14, 20, 42, 43] and other platyhelminths [44, 45].
Diplozoid sensory structures, which appear to be numerous and of several types, may serve for the evaluation of external and internal conditions and/or assist in the search for a competent host. In accordance with our study on P. homoion, no microvilli (reported on the surface of some monogeneans) are present on the body surface of E. nipponicum (except for the buccal cavity) and were able to confirm that the well-innervated sensory structures are distributed over the entire body, despite being concentrated in the forebody and hindbody [11, 15]. The single uniciliated receptors with a tegumentary rim accumulated around the mouth opening, while also occurring on the E. nipponicum and P. homoion forebody, appear to be involved in contacting the host tissue and subsequent feeding [11, 15]. These structures are generally believed to have a tango- and/or rheoreceptory function ; however, in skin monogeneans, it has been suggested that they may function as mechanoreceptors detecting turbulence or vibration in water currents generated by fish in close proximity . Generally speaking, the uniciliated receptor architecture corresponds with that in other monogeneans [13, 15, 20, 46-48]. Of particular interest is the organisation of the base in receptors with a prominent tegumentary rim (Fig. 9J), which differs in having much more massive innervation and the presence of muscles. Similar raised receptors with muscles surrounding the central innervation were detected by immunofluorescence (inset in Fig. 10B). It seems likely that such musculature would allow receptor positioning. In SEM micrographs of P. homoion, the cilium of single uniciliated receptors was seen to be anchored by radially organised septa embedded in the tegument . In this study, similar radial septa were occasionally detected in ultrathin sections of E. nipponicum. These appear to correspond to the transitional striated fibres arising from the basal body of the cilium (in a similar manner to wheel spokes) and extending to the dense collar of the nerve bulb . In addition to uniciliated sensory structures, both this study and previous studies have shown that the lateral side of the hindbody in E. nipponicum bears a row of non-ciliated papillae that we assume are involved in reception of environmental stimuli [11, 15]. Similar non-ciliated receptors, including the clusters of papillae above the clamps in P. homoion  and papillae in Entobdella soleae [2, 48], are generally assumed to function as mechanoreceptors while contacting the host . In our previous study, we suggested that the proprioreceptors may have an alternative role in sensing the relative position of the haptor during movement .