Eyes in Staurozoa (Cnidaria): a review

Introduction

Staurozoa is a cnidarian class currently represented by 50 species classified in 11 genera (Miranda et al., 2016a; Miranda et al., 2018; Fig. 1). They are all marine, benthic, and generally reported in shallow temperate waters (Miranda et al., 2018). Staurozoa have a life cycle with two main generations (i.e., metagenetic), known as the stauropolyp and the stauromedusa (Wietrzykowski, 1912; Kikinger & Salvini-Plawen, 1995; Miranda, Collins & Marques, 2010). However, metamorphosis in Staurozoa is not so clearly defined as in other medusozoans (i.e., Cubozoa, Hydrozoa, and Scyphozoa). In Staurozoa, metamorphosis is mainly observed in the apical region (i.e., calyx) and the stauromedusa remains attached to the substrate by a basal peduncle (Wietrzykowski, 1912; Kikinger & Salvini-Plawen, 1995; Miranda, Collins & Marques, 2010). Therefore, the medusa stage has polypoid and medusoid characters (Miranda et al., 2016a).

During the metamorphosis of a stauropolyp into an adult stauromedusa, the eight primary tentacles (four interradial and four perradial) can have four fates: (1) they disappear by resorption (e.g., Lucernaria, Craterolophus, and some Calvadosia); (2) they metamorphose into adhesive interradial and perradial rhopalioids (e.g., Manania and Haliclystus); (3) they remain as primary tentacles but with a modified shape (e.g., some Calvadosia); (4) they change their shape (filiform to capitate), migrate and cluster together with the secondary tentacles (e.g., Stylocoronella) (see Miranda et al., 2016a). The secondary tentacles appear between two primary tentacles (one perradial and one interradial), in adradial position, and progressively get united in clusters during arm formation (Wietrzykowski, 1912) (Fig. 2).

We recently hypothesized that rhopalioids are a synapomorphy of the family Haliclystidae, which includes the genera Manania, Haliclystus, Depastromorpha, Depastrum, and Halimocyathus (Miranda et al., 2016a; Fig. 1). The structure likely has a role in temporary substrate attachment, a hypothesis supported by stauromedusan behavior (Larson, 1988) and histology (Miranda et al., 2016b). The genera Lucernaria and Lipkea are closely related, and we hypothesized that Stylocoronella would fit this clade (Fig. 1; see also Kikinger & Salvini-Plawen, 1995) based on the morphology (e.g., presence of interradial longitudinal muscles in peduncle and absence of primary tentacles or rhopalioids in stauromedusa) (Miranda et al., 2016a). Craterolophus and Calvadosia belong to the suborder Amyostaurida as they are the only two genera without interradial longitudinal muscles in peduncle (Miranda et al., 2016a).

In a recent article, Picciani et al. (2018) proposed that eyes (defined as a “region made of photoreceptor cells adjacent to pigment cells”, irrespective of image formation, i.e., including all photoreceptive organs) originated at least eight times in Cnidaria, even in the absence of a central nervous system. Their study was mainly based on a review of the literature on the presence of eyes for adult medusae (their Table S1) in light of an extensive ribosomal and mitochondrial-based molecular phylogeny for Cnidaria. Their analyses covered all of the cnidarian classes (other than the unusual parasitic class Myxozoa), including Staurozoa (i.e., stalked jellyfishes, Fig. 2), which was characterized as lacking eyes (Picciani et al., 2018).

Picciani et al. (2018) considered eyes absent in Staurozoa based on a simple statement by Mayer (1910 p. 520: “eyes [...] are absent in Stauromedusae”). Nevertheless, Picciani et al. (2018) correctly included an exception for the genus Stylocoronella. Polyps of Stylocoronella riedli Salvini-Plawen 1966 and Stylocoronella variabilis Salvini-Plawen 1987 possess dark pigment spots on the oral side of the calyx, at the inner bases of the tentacles (Salvini-Plawen, 1966; Salvini-Plawen, 1987; Blumer et al., 1995; Kikinger & Salvini-Plawen, 1995). Based on histological studies, these pigment spots were interpreted as being ocelli (Salvini-Plawen, 1966). Although the polyps of S. riedli show no distinct reaction to light stimuli, ultrastructural results corroborated the hypothesis that these structures are light-sensitive organs (Blumer et al., 1995). The ocelli are composed of seven to nine monociliary sensory cells, that lie next to the tentacular mesoglea, and one to four pigment cells (Blumer et al., 1995). The monociliary sensory cells of S. riedli show all the features characteristic for photoreceptive cells, including the intraciliary structure, demonstrating that these cilia are immobile (see Blumer et al., 1995). The pigment cells enclose the photoreceptive cilia and have irregularly shaped pigment granules enveloped by a membrane (Blumer et al., 1995). The dark pigment visible in living animals is associated with these membranes (Blumer et al., 1995). In addition, the ocelli in Stylocoronella have characteristics (e.g., arrangement of microtubules of photoreceptive cilia and membranous elements of pigment granules) that were hypothesized as unique for Stylocoronella and unknown within other metazoans (Blumer et al., 1995; Martin, 2002). The dark pigment spots are, in a somewhat modified arrangement, retained in the medusa stage (Blumer et al., 1995; Kikinger & Salvini-Plawen, 1995). The stauromedusa stage of S. riedli has “numerous tiny pigment spots in the basal area of the capitate tentacles and four larger perradial ones at the subumbrellar margin”, which were also presumed to be ocelli (Kikinger & Salvini-Plawen, 1995).

However, there are no molecular sequences for Stylocoronella species, so the presence of these ocelli was disregarded in the analyses of eye origins within Cnidaria (Picciani et al., 2018). On the other hand, Picciani et al. (2018) included two species of Manania: Manania gwilliami Larson & Fautin 1989 and Manania uchidai (Naumov 1961) (their Table S1) in their phylogeny, and both possess dark pigment spots (=eyes, sensu Picciani et al., 2018) associated with rhopalioids (= anchors) in the stauromedusa stage (Naumov, 1961; Larson & Fautin, 1989). Therefore, Picciani et al. (2018) neglected literature on Staurozoa and our aim is to review it and discuss the evolutionary implications of correctly coding this character in Staurozoa.

Results

Conclusions

Based on our review, all dark pigment spots in staurozoans are associated with the primary tentacles, the region where they used to be, e.g., Stylocoronella and Calvadosia, or with the rhopalioids that are derived from primary tentacles (e.g., Manania and Haliclystus). Because the rhopalia of medusae of Cubozoa and Scyphozoa are also derived via metamorphosis of primary tentacles of cubopolyps and scyphopolyps, respectively, rhopalioids and rhopalia are hypothetically homologous (Thiel, 1966). However, whereas the rhopalia is clearly a sensory structure (reviewed in Katsuki & Greenspan, 2013), an adhesive rather than sensorial function thought to be associated with the benthic habit of staurozoans is often attributed to the rhopalioids of stauromedusae (Larson, 1988; Miranda et al., 2016b; Miranda et al., 2018).

Dark pigment spots associated with the eight rhopalioids have been observed in virtually all species of Manania and in at least some species of Haliclystus. We found no records of dark pigment spots for the other genera with rhopalioids, Depastromorpha, Depastrum, and Halimocyathus (Miranda et al., 2016a), in the literature, but this information could be overlooked, since the dark spots disappear after preservation (Uchida, 1929) and their presence can vary during development (Clark, 1878).

Unlike the pigment spots in polyps of Stylocoronella (Blumer et al., 1995), the ultrastructure of dark pigment spots in stauromedusae of Manania and Haliclystus have never been analyzed, potentially raising doubt about a sensorial function. However, recent evidence (Westlake & Page, 2017) supports the idea of photo reception by these organs in both genera. Westlake & Page (2017) analyzed the neuromuscular morphology of two stauromedusae, Manania handi (Larson & Fautin 1989) and Haliclystus “sanjuanensis” (nomen nudum), using whole mount immunohistochemistry with antibodies against FMRFamide and α-tubulin to label neurons. Comparative observations on cnidarians indicate that photoreceptive organs are consistently associated with the expression of these markers (see Westlake & Page, 2017), although peptides of the RFamide family also occur in species or life stages of cnidarians that do not have ocelli or known structures of photoreception (Plickert & Schneider, 2004). Interestingly, the “transformed primary tentacles” of M. handi had a greatly increased “concentration of FMRFamide-immunoreactive neurons at their base” associated with the dark pigment spots, which the authors concluded to be “consistent with their homology with rhopalia” of cubomedusae and scyphomedusae (Westlake & Page, 2017). Besides, a similar, but less pronounced, knot of FMRFamide- immunoreactive neurons is present at the base of the rhopalioids of H. “sanjuanensis”, although the species lacks a pigment spot in this area (although, as we pointed out, there are changes during development in Haliclystus). This result might be associated with the observed expression of opsin genes (the functional visual pigments in vertebrate and most invertebrate photoreceptors; Martin, 2002) in species of Haliclystus (Picciani et al., 2018), but raises interesting questions about the expression of medusozoan opsin genes reported by Picciani et al. (2018) in staurozoan species whose adult stauromedusae apparently lack both rhopalioids and dark pigment spots, such as Lucernaria quadricornis Müller 1776 and Craterolophus convolvulus (Johnston 1835). For example, What is the diversity of opsin genes across Staurozoa, and how have these genes evolved? Where and when are different opsin genes expressed across staurozoan life cycles and body regions? Are there mechanisms of extraocular photosensitivity in Staurozoa (see Martin, 2002)?

Although the dark pigment spots in Manania species were often associated with hypotheses of light perception, Westlake & Page (2017) mentioned that M. handi showed no immediate response to changes in light intensity, and hypothesized that “light sensitive neurons in staurozoans may detect light to trigger synchronous spawning, rather than to modify immediate behavior” (Westlake & Page, 2017). However, they highlighted that many stauromedusae spawn in response to light (see Otto, 1976; Otto, 1978; Miranda et al., 2018) despite having no obvious pigment spots (e.g., species of Haliclystus), which could be interpreted as a counterargument for this hypothesis (see examples of extraocular photosensitivity associated with cnidarian spawning in Martin, 2002). Besides, polyps of Stylocoronella also have ocelli, show no distinct reaction to light stimuli (Blumer et al., 1995) and do not spawn (as gonads are only fully developed in the stauromedusa stage; Kikinger & Salvini-Plawen, 1995). Therefore, although the ultrastructure of ocelli are compatible with light perception, at least in polyps of S. riedli (Blumer et al., 1995), the complete understanding of their function is a challenge and further studies that specifically address photosensitive behavior are necessary.

The presence of dark pigment spots (or ocelli) in Staurozoa suggests that rhopalioids in stauromedusae can have both adhesive (see Miranda et al., 2016b) and sensorial functions. The reasons related to the wide occurrence of these structures in the stauromedusa stage of species of Manania needs further investigation. The homology of dark pigment spots in Staurozoa is questionable, since the genera Manania, Haliclystus, Stylocoronella, and (possibly?) Calvadosia do not form a monophyletic group (Miranda et al., 2016a; Fig. 1). In this context, ultrastructural studies on the dark pigment spots of different genera, especially Manania, and at different stages (i.e., polyp and medusa) should also be encouraged as it could provide relevant data on the evolution of this character in Staurozoa. However, in all these genera, the dark pigment spots of stauromedusae were associated with primary tentacles, their region (perradial/interradial) or structures derived from them. Therefore, based on our review and on the phylogenetic topology obtained by Picciani et al. (2018), in which Staurozoa formed a sister group relationship to the remaining medusozoans (Fig. 3A), the presence of these structures in Staurozoa could indicate at least a ninth independent origin of eyes in Cnidaria.

Alternatively, the observations reviewed here highlighting eyes associated with structures derived from primary tentacles (rhopalioids/rhopalia) might indicate shared ancestry of eyes of medusae across Cubozoa, Scyphozoa, and Staurozoa, which were recently shown to form the clade Acraspeda based on phylogenomic data (Kayal et al., 2018; Fig. 3B). The Acraspeda hypothesis linking Staurozoa with Cubozoa and Scyphozoa goes back to the late 1880’s (Haeckel, 1880; Claus, 1883) and was also suggested by a cladistic analysis based on morphology and life history characteristics (Marques & Collins, 2004). This explicit phylogenetic analysis supported the assertion of Thiel (1966) that rhopalia and rhopalioids, both derived from primary polyp tentacles, are shared across Acraspeda due to common ancestry. That these apparently homologous structures are also always the position of eyes across Acraspeda raises the possibility that light sensitivity has specific components that are shared across the group due to common ancestry. On the other hand, the strength of the Picciani et al. (2018) analysis was that it looked at species level observations of eyes, and while eyes are ubiquitous across Cubozoa, documented eyes are less than universal in both staurozoans and scyphozoans, supporting the idea that eyes may have evolved independently several times within Acraspeda. We support that eyes in hydromedusae have an independent origin from those in the medusae of Acraspeda, as sense organs in the former are not associated with metamorphosis of primary polyp tentacles (Thiel, 1966; Salvini-Plawen, 1987; Marques & Collins, 2004; Picciani et al., 2018). In short, correctly coding this character in Staurozoa, a lineage of benthic medusae, has profound consequences for understanding the evolution of eyes and nervous systems in Cnidaria and studies disregarding the presence of staurozoan ocelli should be reassessed.