Two groups of twenty H. tubulosa adult individuals were collected by scuba-diving at 5–10 meters depth from a rocky site of the Gulf of Naples (40°47’50’’N and 14°12’04’’E) along the southern Tyrrhenian coast in the summer 2023. The animals were collected by the IRM core facility and transferred to the Marine Resources for Research Facility at SZN, where they were acclimated in 150L tanks for one week ( Figure 1A ) and then transferred to 500L tanks for long term maintenance. The tanks are connected to an open system of running seawater directly pumped from the sea (9 meters depth, 250 meters from the coast), therefore water follows seasonal temperature variation (14 ± 1°C to 25 ± 1°C). By keeping adult H. tubulosa in these conditions we could obtain gametes from June to January.

Oocyte maturation was achieved using the Trx peptide (Biomatik) 1mg/mL with a maturation index of 82.88% (on 247 total oocytes counted) assessed counting the number of oocytes that lost follicle cells and presented broken down nuclei. Mature eggs were washed with FSW once, 1 μl of dry sperm was diluted in 10 mL of FSW and some drops were added to the eggs concentrated in 10 mL. After 5 minutes, fertilized eggs were washed several times with FSW. Embryos were reared in 500 mL glass beakers at 24°C under a 12h light/dark cycle. Larval cultures were maintained up to the juvenile stage by exchanging half of the FSW with fresh FSW and by feeding them with a mix of Isochrysis galbana and Rhodomonas salina algae 2–3 times per week.

Specimens were collected and fixed overnight at 4°C in 4% paraformaldehyde (PFA) in FSW, washed x3 in PBS and stored in 70% EtOH for anti-MHC (Andrikou et al., 2013) and anti-acetylated tubulin (Sigma-Aldrich, St Louis, MO, USA) immunohistochemistry. Specimens were fixed in 4% PFA in FSW 3 hours at room temperature (RT) and stored in FSW with 0.01M sodium azide for anti-serotonin (#S5545; Sigma-Aldrich) and anti synaptotagmin (1E11,Nakajima et al., 2004) immunohistochemistry. Samples were then washed in 0.3% triton X-100 in PBS for 15 min at RT, followed by 3–5 washes in 0.1% Tween-20/PBS and finally incubated in blocking solution (3% bovine serum albumin/PBST) for at least 1 hour at RT before adding primary antibodies (anti-MHC 1:100, anti-serotonin 1:500, anti-acetylated tubulin 1:200) and phalloidin 1:300 O/N at 4°C. The appropriate Alexa Fluor secondary antibody was used 1:1000. DAPI was added at a final concentration of 0.01 mg/mL to stain nuclei.

Specimens were imaged using a LEICA DMi8 Inverted Microscope operating in bright field and polarized light mode. For the immunostaining, embryos and larvae were mounted for imaging with a Zeiss (Jena, Germany) LSM 700 confocal microscope and pictures were analyzed using ImageJ (v1.53v) (Schindelin et al., 2012). Statistical analysis was performed using GraphPad Prism Version 10.2.1 (339).

A challenge working with sea cucumbers is the lack of reproducible and sustainable protocols to obtain gametes. Because with natural spawning (Rakaj et al., 2018) it is not possible to control fertilization in vitro, or to obtain gametes from the same animal multiple times, we tested two alternative methods.

First, we tried evisceration, an ejection of organs from the body cavity that represents the sea cucumber’s natural response to stresses such as predation. We induced evisceration by injecting 1mL of potassium chloride (1M) in the body cavity, followed by a few seconds of animal shaking. After less than a minute the animals eviscerated, releasing gonads and other internal organs into the sea water ( Figures 1B, C ). Although highly reproducible, this method has the disadvantage of obtaining a huge amount of gonads that most likely are not needed for developmental studies. Moreover, evisceration leads to empty animals that need months to regenerate their gonads. As a second method, we tested gonad extraction by microsurgery. We made a 1.5 cm cut along the ventral-left side of the animal with a clean scalpel ( Figure 1D ). By gently squeezing the animal we were able to obtain a small amount of gonads ( Figures 1E–H ) that is enough to make thousands of embryos. The advantage of this method we developed is that the wound closes in a few hours and heals in a few days and the same animals can be reused multiple times.

Because these methods do not follow the natural spawning cycle, the oocytes obtained from the female ovaries through evisceration and microsurgery are arrested in prophase I of meiosis. A hormonal stimulus is therefore required to induce oocyte maturation for subsequent fertilization (Delroisse et al., 2021). The lack of a universal hormone suitable for every sea cucumber species has been a limiting step for their use in developmental biology. Recently, a 6-aa peptide isolated from the hormone thioredoxin has been successfully used to induce oocyte maturation in Holothuria polii (Léonet et al., 2019). We synthetized this peptide and tested its activity on H. tubulosa oocytes at a range of concentrations from 0.1 mg to 1 mg in sea water. H. tubulosa oocytes are usually wrapped in a layer of follicle cells ( Figure 1I , arrows) that detach during maturation. The oocytes treated with 1 mg/mL peptide all resumed meiotic maturation within two hours, as seen by germinal vesicle breakdown and release of follicle cells ( Figures 1J, K ). No sign of maturation was observed at lower concentrations, even with longer incubation times. Activated oocytes were successfully fertilized (as confirmed by the first cell cleavage observed in most oocytes about two hours after the addition of the sperm ( Figure 1L )) but we did not observe formation of a clear, thick fertilization membrane as previously reported by natural spawning methods (Rakaj et al., 2018). Altogether, we defined a reliable method to obtain and mature oocytes in H. tubulosa that allows researchers to work with the same animal multiple times and without depending on natural spawning events.

We next characterized embryonic and larval development using light and confocal microscopy. As previously reported, the first embryonic divisions give rise to 2 equal blastomeres 2 hours after fertilization (hpf) and 4 blastomeres 3 hpf (Rakaj et al., 2018; Perillo et al., 2024) ( Figures 1L, M ). Blastulae (12 hpf) are round and F-actin staining showed polarization of ectodermal cells, with apical polarity opposite to the blastocoel ( Figure 2A ). At the gastrula stage (29 hpf) the embryo elongates along its antero-posterior axis and the archenteron invagination begins at the vegetal pole ( Figure 2B ). Cells in the archenteron (the future gut) at the early and mid- gastrula stages appear bigger than the ectodermal cells and with a round shape ( Figure 2C , insert). By the end of gastrulation, a straight gut is formed and larvae start feeding (50 hpf, Figures 2D, E ).

By this stage the hydropore is formed, a short tube that connects the blastocoel to the outside environment ( Figures 2D, E , arrow). The early auricularia (3 days) is characterized by new folds in the ectoderm, the formation of the ossicle (described in Figure 3 ) and by the appearance of mesenchymal cells that are rich in actin and are localized around the blastocoel ( Figures 2F arrows, G ). These mesenchymal cells are still rich in actin in the later auricularia stage, with long filamentous projections ( Figure 2H ). At this stage, the tripartite gut is composed of an esophagus covered by parallel circumesophageal muscle fibers that are mononucleated ( Figures 2I ) and are positive to an anti-myosin heavy chain antibody ( Figure 2J ), a round stomach and a tubular intestine ( Figure 2K ). The esophagus and the stomach are separated by the cardiac sphincter constriction, and a pyloric sphincter is present between stomach and intestine, differently from the sea star larval gut (Annunziata et al., 2013) ( Figure 2K ). The lumen of the digestive system is heavily ciliated as shown by acetylated tubulin staining ( Figure 2L ). Summary of these structures are shown in Figure 2M .

As larvae increase in size, the number of ectodermal folds increase ( Figure 2N ). At this stage, the hydropore becomes part of the hydrocoel (the future water vascular canal (Udagawa et al., 2022)), a tissue with an actin-rich apical side towards the central lumen ( Figures 2N’, O and insert). Two additional thin tubes, the left and right somatocoels, form laterally to the stomach ( Figure 2N ’). The function of these additional tissues is unknown, but similar tubular tissues are shared with sea star larvae and are thought to be involved with buoyancy (Potts, 2003; Perillo et al., 2023). At this stage, the hyaline spheres appear ( Figure 2P ), spherical structures that serve as energy storage to be used during metamorphosis (Peters-Didier and Sewell, 2019). Hyaline spheres are located at the tip of the larval arms where numerous nuclei are clustered ( Figure 2P insert).

During metamorphosis there is a dramatic rearrangement of larval structures to transform into a doliolaria and then a pentactula, stages characterized by five ciliary band rings strongly enriched for actin ( Figure 2Q , late doliolaria). Actin staining shows that the tentacles are formed by many parallel muscle fibers ( Figure 2Q’ ). The juvenile is characterized by skeletal elements that are visible under bright field and polarized light, such as adult skeletal structures that wrap around the tentacles and on the body walls, and the larval ossicle that is located posteriorly ( Figures 2R , R’). A ventral oral view of the juvenile shows the central mouth surrounded by the five buccal tentacles ( Figure 2S ).

Overall, this detailed analysis of larval morphology with immunostainings, bright field and polarized light allowed us to define the different tissues of H. tubulosa larval stages.

Ossicles, or skeletal rods, are unique structures of sea cucumber larvae that are thought to be involved with orientation of the larval body (Pennington and Strathmann, 1990; Young et al., 2003). In the Holothuria genus, larvae only have one ossicle that is localized near the posterior ectoderm (Young et al., 2003). In H. tubulosa, this single, spherical ossicle can be visualized with polarized light ( Figure 3A 1) and is surrounded by mesenchymal cells ( Figure 3A 2, arrows). In the late auricularia, a hyaline sphere (HS) is visible on the anterior-end of the ossicle ( Figure 3A 3). To explore ossicle morphogenesis, we characterized how this structure forms over developmental time. We first measured the area of the central section of the ossicle from when it first appears (early auricularia, EA) until the late auricularia stage and found that ossicle size consistently increases over time ( Figure 3B ).

In other echinoderm larvae, specialized mesenchymal cells contribute to skeletal element formation (Vidavsky et al., 2014). In H. tubulosa, we observed clusters of mesenchymal cells gathering at the larval posterior end from before the ossicle appears and throughout its growth ( Figure 3C , arrows). We next tested whether ossicle growth correlates with the number of mesenchymal cells clustering around it over time and found that this number is constant, averaging between 2 and 6 from before ossicle formation until later stages ( Figure 3D , see figure legend for statistics). To filter out the effect of the developmental stages and to further define the relationship between ossicle size and number of mesenchymal cells, we correlated ossicle size (from the smallest to the biggest area) with the number of cells around it. This analysis showed that increased ossicle size does not correlate with increased number of mesenchymal cells ( Figure 3E ). Altogether our data show that ossicle growth does not rely on an increased number of cells around it, but it is instead associated with a defined pool of mesenchymal cells that cluster around the posterior end of the larva prior to the formation of the ossicle.

An open question in evo-devo is how functionally related structures such as organs evolved different organizations and shapes. An example is the arrangement of serotonergic neurons, that are among the first type of neurons to differentiate in echinoderm larvae and are involved in the regulation of their swimming behavior (Bisgrove and Burke, 1987; D’Aniello et al., 2020; Obukhova et al., 2024). To explore the development of this neuronal type in H. tubulosa larvae, we performed serotonin immunostaining on larval stages (early auricularia, auricularia and late auricularia) to analyze the distribution of serotonin neurons. The first serotonergic neurons appeared at the end of gastrulation in the early auricularia and are localized in the anterior hood and dorsal anterior ectoderm ( Figures 4A -A’’). At the auricularia stage these neurons extend long projections in both territories, and the neurons of the anterior hood are localized in the ciliary band and ventral ectoderm ( Figures 4B -B’’). By the late auricularia stage serotonergic neurons are restricted to the anterior ventral and dorsal ciliary band only ( Figures 4C -C’’). Using the pan neuronal marker synaptotagmin (1E11 antibody), we confirmed that serotonergic neurons are specifically localized within the anterior ciliary band, including the oral hood ( Figures 4D -D’’).

To provide an example of how valuable the use of echinoderm larvae is to explore the evolution of the nervous system, we performed serotonin immunohistochemistry on larvae of the sea star Patiria miniata and of the sea urchin Paracentrotus lividus. Although sea star larvae have the anterior ectoderm organized in a large oral hood like sea cucumbers, serotonergic neurons are exclusively localized in the anterior dorsal ectoderm and are absent from the oral hood, as reported also for other stages (Cheatle Jarvela et al., 2016; Pagowski, 2024) ( Figures 4E -E’’; 4F -F’’). Conversely, serotonergic neurons in the sea urchin P. lividus are localized in the apical organ of the anterior hood but absent from the dorsal ectoderm ( Figures 4G -G’’; 4H -H’’), similarly to what was previously reported in other species (Beer et al., 2001; Burke et al., 2006; Angerer et al., 2011; Perillo et al., 2018; Wood et al., 2018; Cocurullo et al., 2024). Moreover, in the late pluteus larva we found that these serotonergic neurons extend towards the arms ( Figures 4H -H’).

Together, these results show that while serotonergic neurons are distributed in both the ventral and dorsal anterior ectoderm in H. tubulosa, the same neurons are exclusively expressed in only one domain, being dorsal in the sea star and ventral in sea urchin ( Figure 4I ).