Proboscis apparatus: What do we know (and not know) about nemerteans?

Invertebrate Zoology, 2025, 22(1): 30–43 © INVERTEBRATE ZOOLOGY, 2025 Proboscis apparatus: What do we know (and not know) about nemerteans? A.V. Chernyshev1*, T.Yu. Magarlamov1 1 A.V. Zhirmunsky National Scientific Center of Marine Biology, Far Eastern Branch, Russian Academy of Sciences, ul. Palchevskogo 17, Vladivostok 690041 Russia. Alexey Chernyshev: [email protected] ORCID https://orcid.org/0000-0002-2203-3001 Timur Magarlamov: [email protected] ORCID https://orcid.org/0000-0003-4175-5007 * Corresponding author ABSTRACT: This paper provides an overview of recent data on the morphology and functions of the proboscis in nemerteans. It is still unclear whether rhynchocoel and proboscis are synapomorphies of the phylum Nemertea and whether rhynchocoel is homologous to coelom in other Spiralia. The proboscis initially had a structure similar to the body wall, which is typical of the class Palaeonemertea. However, subsequently, the proboscis evolved in two directions: (1) the development of bilateral symmetry in palaeonemerteans from the family Cephalotrichellidae and Pilidiophora (except some Valenciniidae species) and (2) the development of radial symmetry in Hoplonemertea. Pilidiophorans are characterized by a wide variety of proboscis structures, while hoplonemerteans have the most diverse rhynchocoel morphology. The emergence of the stylet apparatus in hoplonemerteans led to the loss of diagonal musculature and pseudocnidae in the proboscis, as well as to a decrease in the number of the family of peptide toxins. How to cite this article: Chernyshev A.V., Magarlamov T.Yu. 2025. Proboscis apparatus: What do we know (and not know) about nemerteans? // Invert. Zool. Vol.22. No.1. P.30–43. doi: 10.15298/invertzool.22.1.03 KEY WORDS: proboscis, rhynchocoel, coelom, endothelium, pseudocnidae, armature, evolution. Хоботной аппарат: что нам известно (и не известно) о немертинах? А.В. Чернышев1*, Т.Ю. Магарламов1 1 Национальный научный центр морской биологии им. А.В. Жирмунского Дальневосточного отделения Российской академии наук, Пальчевского 17, Владивосток, 690041 Россия. * Ответственный за переписку: [email protected] РЕЗЮМЕ: Данная работа посвящена обзору морфологии и функционированию хоботного аппарата немертин. До сих пор нет однозначного представления является ли наличие ринхоцеля и хобота синапоморфиями типа Nemertea и гомологичен ли ринхоцель целому других Spiralia. Первоначально хобот немертин имел строение сходное со стенкой тела, что типично для представителей класса Palaeonemertea. Однако в дальнейшем хобот эволюционировал в двух направления: (1) развитие билатеральной симметрии у палеонемертин из семейства Cephalotrichellidae и у Pilidiophora (за исключением некоторых Valenciniidae) и (2) формирование радиальной симметрии у представителей Hoplonemertea. Если пилидиофоры характеризуются большим разнообразием в организации хобота, то у гоплонемертин наблюдается Devoted to memory of Claus Nielsen. Proboscis apparatus: What do we know (and not know) about nemerteans? 31 Introduction The proboscis apparatus, a formation consisting of two general components, the rhynchocoel (or proboscis sheath) and the proboscis proper, is specific for Nemertea (Fig. 1A). With contraction of the rhynchocoel walls, the proboscis is everted via rhynchopore and is inverted back by the retractor muscle. The proboscis is an organ of attack and defense; in some nemerteans (primarily terrestrial and semiterrestrial ones) (Fig. 1B), it is also used for locomotion (Gibson, 1972; Moore, Gibson, 1981). The rhynchocoel, in addition to being a proboscis reservoir, performs a supporting function, especially well expressed in benthic polystiliferous nemerteans (Chernyshev, 2011). The nemertean brain surrounds the rhynchocoel like a ring, but not the foregut as in many other Spiralia. In most nemerteans, one or two vessels of the circulatory system enter the rhynchocoel or even the proboscis (Kajihara, 2010; Chernyshev, 2011). In most hoplonemerteans, the mouth merges with the rhynchopore to form the rhynchostomadaeum or the atrium. All the above facts suggest that the proboscis apparatus is an essential component in the ‘Bauplan’ of nemerteans. Thus, another name of the phylum Nemertea is Rhynchocoela, which is preferred by some authors (Hyman, 1951). On the basis of proboscis structure, nemerteans were divided into two groups: Anopla (unarmed nemerteans) and Enopla (armed nemerteans). To date, a high classification has been adopted that divides nemerteans into three classes and seven orders: Palaeonemertea (orders Carinomiformes, Archinemertea, and Tubulaniformes), Pilidiophora (orders Hubrechtiiformes and Heteronemertea), and Hoplonemertea (orders Polystilifera and Monostilifera) (Chernyshev, 2021). The only known nemertean species that lacks the proboscis apparatus is the enigmatic Arhynchonemertes axi described from New Zealand (Riser, 1988). Its systematic position remains unresolved: it is either a representative of an ancient branch of nemerteans that have not yet acquired the proboscis or a specialized nemertean that has lost its proboscis and rhynchocoel (Riser, 1989). If the former assumption is correct, then the proboscis apparatus is not a synapomorphy of Nemertea. However, even if the molecular phylogenetic analysis shows the basal position of A. axi, it still does not answer the question as to whether ancestors of this nemertean had the proboscis apparatus. In the present review, we consider the most debatable issues of the morphology and functions of the proboscis apparatus. These are, first, the origin of the proboscis apparatus and also the evolution of its musculature, nervous system, and some epithelial structures. Do nemerteans have true coelom? The proboscis apparatus was well studied at the light microscopy level in the late 19th century. The magnificent drawings of the proboscis, especially the stylet apparatus, in the books by McIntosh (1873–1894) and Bürger (1895, 1897–1907), still remain unsurpassed. A number of studies with electron microscopy of the proboscis were published in the second half of the 20th century (Ling, 1971; Stricker, Cloney, 1983; Stricker, 1985; Montalvo et al., 1996, 1998; Junoy et al., 2000). The most important were the comparative analyses of microscopic anatomies of nemerteans from different groups, based on which the authors concluded that the nemertean rhynchocoel is a coelomic cavity (Turbeville, Ruppert, 1985; Turbeville, 1991). The statement that the rhynchocoel is a coelom was made by Hyman (1951) and Starobogatov (1983), but Turbeville & Rupert (1985) showed that the endothelium of the rhynchocoel and proboscis has all the morphological features of coelothelium. высокое разнообразие в морфологии ринхоцеля. Появление стилетного аппарата у гоплонемертин привело к потери диагональной мускулатуры и псевдокнид в хоботе, а также к уменьшению числа семейств пептидных токсинов, которые используются при нападении на добычу. Как цитировать эту статью: Chernyshev A.V., Magarlamov T.Yu. 2025. Proboscis apparatus: What do we know (and not know) about nemerteans? // Invert. Zool. Vol.22. No.1. P.30–43. doi: 10.15298/invertzool.22.1.03 КЛЮЧЕВЫЕ СЛОВА: хобот, ринхоцель, целом, эндотелий, псевдокниды, вооружение, эволюция. A.V. Chernyshev, T.Yu. Magarlamov32 Turbeville & Rupert (1985) considered not only the rhynchocoel but also the circulatory system and gonadal sacs as derivatives of coelom. This interpretation has not found wide support (especially as regards gonadal sacs), although it has changed the opinion about nemerteans as exclusively parenchymatous worms (however, in many Russian universities, nemerteans are still considered as acoelomic invertebrates). Ax (1996) insisted on the origin of the coelom (rhynchocoel and blood vessels) in nemerteans independent from that in Spiralia (he did not attribute nemerteans to Spiralia). Nielsen (2001) expressed a similar opinion by placing nemerteans closer to Platyhelminthes. Subsequently, he (Nielsen, 2012) tended to assume that blood vessels can be interpreted as modified coelomic cavities, but rhynchocoel should unambiguously be considered as an independently originated secondary body cavity. Malakhov & Bogomolova (2016), in contrast, stated that rhynchocoel is the same homologue of coelom as blood vessels. Chernyshev (1999, 2011) proposed a hypothetical scenario for the transformation of the paired coelom into the rhynchocoel, blood vessels, and gonadal sacs in the nemertean ancestor. It opposed the hypothesis by Starobogatov (1983) who considered the rhynchocoel homologous to the unpaired acrocoelom (head coelom), and the gonadal funnels homologous to the paired trunk coeloms. The question as to whether the rhynchocoel is homologous to the true coelom of other Spiralia or not remains open. It is also unclear which one was formed earlier, the proboscis or the rhynchocoel. Wijnhoff (1914) and Senz (1997) Fig. 1. Nemerteans with inverted (A) and everted (B) proboscis apparatus. A — diagram of internal morphology of a hoplonemertean (drawn by Oleg Dobrovolsky); B — palaeonemertean Cephalothrix cf. simula with everted proboscis attacking a polychaete. Proboscis apparatus: What do we know (and not know) about nemerteans? 33 suggested the proboscis as an invagination of the anterior end of the head that was formed first, and then the rhynchocoel appeared around it. This hypothesis does not explain why the coelomic epithelium lines both the rhynchocoel and the inverted proboscis. Another hypothesis seems more plausible: the elongated anterior end of the head in nemertean ancestors served to capture food (Fig. 2A) and subsequently began to invert into the already existing coelomic cavity (the future rhynchocoel) (Chernyshev, 2011). As a result, the proboscis became lined inside by the coelomic epithelium of this cavity (Fig. 2B–E). It is still impossible to confirm the correctness of such a pattern, and it remains purely hypothetical. The endothelium of the rhynchocoel and proboscis has cilia characteristic of coelothelium. These are mostly rare and short, but in some nemerteans they can be long and numerous (Magarlamov, Chernyshev, 2015) (Fig. 3D, E). The endothelium structures in the proboscis and the rhynchocoel differ. The proboscis endothelium is pseudostratified coelothelium, where peritoneal cells overlap myocytes that form the outer (endothelial) circular musculature (Magarlamov, Chernyshev, 2015). Initially, myocytes are few in number and scattered, and can only be found in ultrathin sections or by confocal laser scanning microscopy (CLSM) with phalloidin labeling (Fig. 3A). However, in many nemerteans, myocytes are numerous and arranged into several layers (Fig. 3B); in this case, outer circular musculature becomes clearly visible in histological sections. This has created an erroneous impression that some of nemerteans have the endothelial musculature in the proboscis while others do not. The endothelium of the rhynchocoel is true peritoneum, and myocytes are separated from the endothelium by basal lamina and extracellular matrix (Fig. 3C). If we assume that myoepithelium is the initial state for coelothelium (Kuzmina et al., 2018), then the proboscis endothelium has retained more archaic features than the rhynchocoel endothelium. The proboscis morphology as a reflection of the anterior body end While the origin of the rhynchocoel remains unclear, there are almost no disagreements as regards the origin of the proboscis: it is a modified anterior outgrowth of the head inverted inside. As Hyman (1951) noted, this explains why the structure of the nemertean proboscis repeats the structure of the nemertean’s body wall. Is it so? Seventy years ago, the locations of the muscular Fig. 2. Origin and early evolution of Nemertea. A — hypothetical ancestor of nemerteans (drawn by Oleg Dobrovolsky); B–E — diagrams of presumptive origin of proboscis apparatus (yellow — rhynchocoel, red — endothelium; blue — digestive tract, green — muscle-retractor). A.V. Chernyshev, T.Yu. Magarlamov34 layers of the proboscis and the body wall were examined only in histological sections. The CLSM and phalloidin labeling methods have revealed new details in the structure of both the proboscis and the body walls in nemerteans (Chernyshev, 2010, 2011, 2015). The proboscis structure fully repeats the structure of the body wall only in palaeonemerteans that have four layers of musculature (outer circular, diagonal, longitudinal, and inner circular) and a pair of intraepithelial or subepithelial nerve cords both in the proboscis and in the body wall (Fig. 4A, B). The exception is representatives of the order Archinemertea whose diagonal musculature is located under the outer circular musculature in the proboscis and under the epidermis in the body wall (Chernyshev, Kajihara, 2019). Furthermore, the proboscis of archinemerteans Cephalotrichella (Fig. 5A) and Balionemertes has the outer longitudinal musculature that is absent from the body wall (Chernyshev, 2015; Chernyshev, Kajihara, 2019). In a great number of species in the class Pilidiophora, both the proboscis and the body wall have the outer longitudinal musculature (Fig. 4C). However, there is no complete similarity here because the diagonal musculature in the proboscis is located between the outer circular and inner longitudinal musculatures (Fig. 4C), while the diagonal musculature in heteronemerteans is Fig. 3. Endothelium of proboscis apparatus in nemerteans. A — diagram of proboscis endothelium of Cephalothrix cf. simula, longitudinal section; B — diagram of proboscis endothelium in Tortus tokmakovae, longitudinal section; C — diagram of rhynchocoel endothelium in Hubrechtella juliae, transverse section; D, E — CLSM micrographs of proboscis endothelium labeled with α-tubulin antibodies (D — Baseodiscus cf. princeps; E — Cerebratulus sp.). Scale — 10 µm. Proboscis apparatus: What do we know (and not know) about nemerteans? 35 Fig. 4. Diagrams of transverse sections of everted proboscis (A, C, E) and body (B, D, F). A, B — Palaeonemertea; C, D — Heteronemertea; E, F — Hoplonemertea. siomorphic position between the outer circular and inner longitudinal muscle layers (Schwartz, Norenburg, 2005; Chernyshev, 2010). Many pilidiophorans lack the outer longitudinal musculature in the proboscis: in hubrechtiids, this lack is primary, while in heteronemerteans, it is, located in the cutis, and also between the outer longitudinal and outer circular musculatures of the body wall in some species (Chernyshev, 2011; Hookabe, Kajihara, 2020). Only hubrechtiids and the heteronemertean Archimicrura ignae have the body diagonal musculature with a ple- A.V. Chernyshev, T.Yu. Magarlamov36 Fig. 5. Light (A) and CLSM (B–G) micrographs of proboscis labeled with phalloidin and 5-HT antibodies. A — Cephalotrichella echinicola (transverse section); B — Lineus viridis (transverse section); C — Carinoma sp. (transverse section, dark arrows show muscle crosses); D — Cephalothrix cf. simula (longitudinal section); E — Cerebratulus sp. (longitudinal section); F — Baseodiscus sp. (longitudinal section); G — Kurilonemertes dilutebasisae (longitudinal section). Abbreviations: er — epithelial ridge, ilm — inner longitudinal musculature, mc — muscle crosses, ocm — outer circular musculature. Scale: A, C — 50 µm, B, D–G — 100 µm. apparently, secondary. In species of the genera Hubrechtella and Baseodiscus, the proboscis diagonal musculature is located between the inner circular and longitudinal musculatures, while the outer circular musculature is absent, i.e., the proboscis musculature differs significantly from the body wall musculature (Chernyshev et al., 2013). A total of at least eight different variants of arrangement of the muscular layers in the proboscis have been recorded among Pilidiophora, which is twice as many as in the other two classes combined (Chernyshev, 2015). Another feature of the proboscis musculature in pilidiophorans is the presence of one or two muscle crosses, connected to the diagonal musculature (Fig. 5B) (Chernyshev, 2010, 2015). Muscle crosses are present in most heteronemerteans and recently have been described in Hubrechtella ijimai (Kajihara, 2006; Chernyshev et al., 2017). In the body wall of some of heteronemerteans, a dorsal “cross” is present between the outer circular muscle layer and the rhynchocoel musculature. The dorsal and ventral muscle crosses are characteristic of the body wall in many palaeonemerteans which, however, lack any similar structures in their proboscises, except those in Carinoma (Fig. 5C) and Parahubrechtia (see Chernyshev, 2010; Chernyshev et al., 2017). In nemerteans of the class Hoplonemertea, the proboscis repeats the body wall to an even lesser extent: the proboscis lacks diagonal musculature, and several nerve cords are oriented radially in the longitudinal musculature (Fig. 3E, F). Thus, the musculature structure and the location of the nervous system in the anterior part of the proboscis in hoplonemerteans is very conserved, with, however, some deviations in Malacobdella and Ototyphlonemertes valentinae (Magarlamov, Chernyshev, 2010; Chernyshev, 2015). The hoplonemerteans are characterized by a diversity of rhynchocoel morphologies. In a number of hoplonemerteans, the rhynchocoel has lateral, ventral, or dorsal pouches; in species of the genus Uniporus, the lateral pouches are branched. At least nine different variants of the rhynchocoel muscular wall structure have Proboscis apparatus: What do we know (and not know) about nemerteans? 37 been distinguished within the class (vs. only four variants found in palaeonemerteans and pilidiophorans), and five, where the circular musculature is transformed into the spiral one, are unique for pelagic nemerteans (Norenburg, Roe, 1998; Chernyshev, 2011; Chernyshev, Polyakova, 2018, 2019). There is still no consensus as to which arrangement of the rhynchocoel musculature is primary for Hoplonemertea: from the outer circular and inner longitudinal musculatures as in a sister group, Pilidiophora, or from the intertwined longitudinal and circular muscles as in benthic Polystilifera (see Kajihara, 2021). Some authors suggested that initially the rhynchocoel wall should have also reflected the body wall, i.e., consisted of separate layers (Wijnhoff, 1914; Senz, 1997). The proboscis nervous system in palaeonemerteans consists of two nerve cords located intraepithelially or subepithelially, repeating the nervous system of the body wall (Figs 4A, B; 5D). Heteronemerteans usually have two or more pairs of main nerve cords that are randomly connected via secondary nerves, forming a nerve plexus with a pronounced bilateral symmetry (Figs 4C; 5E). In Baseodiscus, Sonnenemertes, and Hubrechtella, the nerve plexus of the proboscis lost its bilateral symmetry (Fig. 5F) and became similar to the proboscis nervous system of hoplonemertean Malacobdella (Magarlamov et al., 2011; Chernyshev et al., 2013; Chernyshev, Polyakova, 2018). In hoplonemerteans, seven to 36 nerve cords embedded in the longitudinal musculature and numerous connectives between them form a more or less ordered radially symmetrical nerve plexus (Fig. 5G) (Chernyshev, 2011; Magarlamov et al., 2011) (except Ototyphlonemertes valentinae with its secondarily bilaterally symmetrical proboscis — see Chernyshev, 2015). Proboscis armature The main distinguishing feature of the hoplonemertean proboscis is armature, which is also found, however, in some of Palaeonemertea and Pilidiophora. The trowel-like armature is present in the proboscis of palaeonemerteans of the genus Callinera (Kajihara, 2006; Chernyshev, 2011, 2015) (Fig. 6A). It was first reconstructed and described by Hiroshi Kajihara and, therefore, we suggest referring to it as Kajihara’s stylet. This structure is apparently composed of chitin and is flexible. Functions of the Kajihara’s stylet remain unknown, but, when the proboscis is everted, it is located subapically and can potentially cause wounds to prey. An armature consisting of a multitude of hooked parastylets has been found in the proboscis of the heteronemertean Heteroenopleus enigmaticus (Wern, 1998). Thus, there are armed “unarmed” nemerteans. Moreover, there are also unarmed hoplonemerteans such as species of the genus Malacobdella that completely lost the stylet apparatus. The true stylet apparatus is present only in hoplonemerteans and is located in the middle proboscis chamber. Two types of this apparatus are distinguished: polystiliferous (with numerous short stylets on a falciform basis) (Fig. 6B) and monostiliferous (with a single long stylet on the longitudinally stretched basis) (Fig. 6C). Fig. 6. Proboscis armature. A — Callinera sp. (arrow shows Kajihara’s stylet); B — polystiliferous armature of Drepanophoridae sp. (inset shows stylets at high magnification); C — monostiliferous armature of Amphiporus sp. (arrows show accessory stylets); D — monostiliferous armature of Nipponnemertes cf. rubella (arrow shows accessory stylet inside basis); E — “bistiliferous” armature of Cratenemertidae sp. IZ-45644. Scale: A, B, D — 50 µm, C — 100 µm. A.V. Chernyshev, T.Yu. Magarlamov38 Fig. 7. Monociliated sensory cells (A–E) and pseudocnidae (F–I). A — diagram of monociliated sensory cell of Lineus viridis; B — diagram of transverse section of the collar part; C — fragment of everted proboscis in Cephalothrix cf. simula with monociliated sensory cells (arrows show sensory processes); D — CLSM micrograph of proboscis epithelium in Nipponomicrura uchidai labeled with phalloidin (with collar microvilli visible); E — SEM micrograph of proboscis epithelium of monociliated sensory cells in Hubrechtella juliae (white arrow shows bulb-like structure, black arrow shows microvilli); F — diagram of pseudocnida structure in Lineus viridis; G — CLSM micrograph of proboscis epithelium in Cephalothrix cf. simula with large (arrows) and small (arrowheads) pseudocnidae (autofluorescence); H — pseudocnida of Hubrechtella juliae with extruded core (arrow); I — SEM micrograph of everted proboscis in Micrura kulikovae showing epithelial ridge (arrows) with pseudocnida clusters (inset shows pseudocnidae cluster at high magnification). Abbreviation: ar — axial rootlet. Scale: C, D, G, H — 10 µm, E — 2 µm, I — 50 µm.