Cell proliferation controls body size growth, tentacle morphogenesis, and regeneration in hydrozoan jellyfish Cladonema pacificum

Animal cultures

We used Cladonema pacificum (strains 6W and UN2) (Figs. 1–5; Figs. S1–S3), Cytaeis uchidae (strain ♀17) (Fig. 6) and Rathkea octopunctata (strain MF-1) (Fig. 6) medusae for this research. The medusae were cultured in plastic cups (V-type container, V-7 and V-8, AS ONE, Osaka, Japan) at 20 °C (Cladonema and Cytaeis) or 4 °C (Rathkea), and their polyps were maintained in the cups (V-7) at 20 or 4 °C in darkness. Vietnamese brine shrimp (A&A Marine LLC, Elk Rapids, MI, USA) were fed to medusae and polyps every other day ad libitum, with water renewed immediately after feeding. Artificial sea water (ASW) was prepared by SEA LIFE (Marin Tech, Tokyo, Japan). Pictures of medusae were taken through a LEICA S8APO microscope with a Nikon digital camera (D5600).

Immunofluorescence

The medusae were anesthetized with 7% MgCl₂ in ASW for 10 min and fixed 4% paraformaldehyde (PFA) in ASW for 1 h. After fixation, the samples were rinsed in 1× PBS and washed three times (10 min each) in PBS containing 0.1% Triton X-100 (0.1% PBT). The samples were incubated in primary antibodies in 0.1% PBT overnight at 4 °C. The antibodies used were rabbit anti-Phospho-Histone H3 (Ser10) (1:500; 06–570, Upstate, Lake Placid, NY, USA) and mouse anti-α-Tubulin (1:500; T6199, Sigma-Aldrich, St Louis, MO, USA). After the primary antibody incubation, the samples were washed three times (10 min each) in 0.1% PBT and incubated in secondary antibodies (1:500; ALEXA FLUOR Goat anti-mouse IgG, ALEXA FLUOR Goat anti-rabbit IgG; Thermo Fisher Scientific, Waltham, MA, USA) and Hoechst 33342 (1:250; Thermo Fisher Scientific, Waltham, MA, USA) in 0.1% PBT for 1 h in dark. After four washes (10 min each) in 0.1% PBT, the samples were mounted on slides with 70% glycerol. Confocal images were collected through Leica SP8 or SP5 confocal microscopes. Z-stack images were performed using ImageJ/Fiji software.

EdU labeling

The medusae were incubated with 20 μM 5-ethynyl-2′-deoxyuridine (EdU; EdU kit; 1836341; Invitrogen, Carlsbad, CA, USA) in ASW for 24 h (Figs. 1–3 and 6) or 150 μM for 1 h (Fig. S1). After EdU treatment, the medusae were anesthetized with 7% MgCl₂ in ASW for 10 min and fixed 4% PFA in ASW for 1 h. After fixation, the samples were rinsed in 1× PBS and washed three times (10 min each) in 0.1% PBT. The samples were incubated with a EdU reaction cocktail (1× reaction buffer, CuSO₄, Alexa Fluor azide, and 1× reaction buffer additive; all included in EdU kit; 1836341; Invitrogen, Carlsbad, CA, USA) for 30 min in the dark. After the EdU reaction, the samples were washed three times (10 min each) in 0.1% PBT and Hoechst 33342 (1:250; Thermo Fisher Scientific, Waltham, MA, USA) in 0.1% PBT for 1 h in dark. The samples were washed four times (10 min each) in 0.1% PBT and were mounted on slides with 70% glycerol.

Hydroxyurea treatment

The live medusae were incubated with 10 mM hydroxyurea (HU) (085-06653; Wako, Osaka, Japan) in ASW (ASW only for control) (Figs. 3–5; Fig. S3). HU incubation was continued for a maximum of 9 days. Medusae were fed every other day, and HU solution or ASW was renewed after feeding. The medusae treated with HU were able to ingest prey like controls, demonstrating that HU treatment had no effect on feeding behavior (Figs. S3A and S3B).

Measurement of umbrella size and tentacle length

Pictures of medusae were taken with a Nikon D5600, and umbrella size was measured using polygon selections with ImageJ software (Fig. 3C). We measured the length and width of medusae under the microscope using an ocular micrometer and multiplied the length and width to generate a value for umbrella size (Fig. 3M). Tentacle length was measured daily under the microscope with an ocular micrometer (Fig. 5H).

DAPI poly-γ-glutamate staining

This protocol was adapted from Szczepanek, Cikala & David (2002): The medusae were anesthetized with 7% MgCl₂ in ASW for 10 min and fixed with 4% PFA in ASW for 1 h. After fixation, the samples were rinsed in 1× PBS and washed three times (10 min each) in 0.1% PBT. The samples were incubated in DAPI (1:500; Polysciences, Inc., Warrington, PA, USA) in PBT for 60 min. After the DAPI incubation, samples were washed four times (10 min each) in PBT and mounted on slides with 70% glycerol in DW. Samples were scanned with a combination of 488 nm excitation and 555 nm emission filter using either Leica SP8 or SP5 confocal microscopes. Using ImageJ, we performed Z-stacks and counted nematocysts. Empty nematocysts were counted manually.

Dissection of tentacles for regeneration

Tentacles’ basal sides were dissected with small scissors, leaving the tentacle bulbs intact. Amputated medusae were fed every other day.

Statistical test

An unpaired two-tailed t-test was performed on the data shown in Figs. 3–5.

Average nearest neighbor distance (spatial statistics)

We performed statistical analysis for the proliferating cells’ distribution in umbrellas and tentacles by applying the nearest neighbor distance (NND) test to EdU positive cells (Table S1). Here, we used the images of 1-day-old medusae that had been incubated with EdU 150 μM for 1 h. This analysis was applied to the umbrella area, except for tentacle bulb and manubrium, while the same analysis was applied to the entire main tentacle. The area (S), the signal number (N), and NND in analyzed areas were obtained using ImageJ/Fiji. The average of NND (W), the expectation value of W (E[W]), and the normalized average of NND (w = W/E[W]) were calculated. In this analysis, w > 1 means that EdU signals are distributed uniformly or randomly. In contrast, w < 1 means that EdU signals are distributed clustered or randomly. The spatial distribution of EdU signals were determined by Z score.

Cell proliferation patterns in the medusa Cladonema pacificum

To understand the spatial pattern of cell proliferation in Cladonema medusa, we performed EdU staining, which labels S-phase or the former S-phase cells (Salic & Mitchison, 2008). Given that Cladonema medusa dramatically increases in size and exhibits tentacle branching during development (Figs. 1A–1C), distribution of proliferating cells could change throughout maturation. We thus investigated cell proliferation patterns in both young (day 1) and sexually mature (day 45) medusae.

In young medusae, EdU-positive cells were broadly detected in the whole medusa body including the umbrella, the manubrium (a supporting organ of the oral in medusae), and the tentacles regardless of the incubating time of EdU (Figs. 1D–1K and 1N–1O, EdU: 20 μM for 24 h; Fig. S1, EdU: 150 μM for 1 h). While small numbers of EdU positive cells were detected in the manubrium (Figs. 1F and 1G), EdU positive cells were uniformly distributed in the umbrella (Average NND: uniform, n = 6/7, random, n = 1/7; Table S1), especially in the exumbrella region (Figs. 1H–1K). By contrast, in the tentacles, large numbers of EdU positive cells were identified as clustered (Figs. 1N and 1O; NND: clustered, n = 7/7, Table S1). We further confirmed that these EdU-positive cells were proliferating cells using the mitotic marker, anti-Phospho-Histone 3 (PH3) antibody. PH3-positive cells were detected in both the umbrella and the tentacle bulbs (Figs. 1L and 1P). In tentacles, mitotic cells were primarily detected in the ectoderm (Fig. 1P; Fig. S2), whereas in the umbrella, proliferating cells were located in the exumbrella, which was confirmed by the presence of mitotic spindles, detected with an anti-α Tubulin antibody in PH3-positive cells (Fig. 1M).

As observed in young medusae, EdU-positive cells were broadly detected in the entire body of mature medusae (Fig. 2). In the umbrella, EdU-positive cells were more often located in the exumbrella than in the subumbrella, which is similar to the case of young medusae (Figs. 2A–2D). By contrast, in tentacles, EdU-positive cells were restricted to their base, called the tentacle bulb, where two apparent “clusters” are located on both sides of the bulb (Figs. 2E and 2F). These clusters were also observed in the tentacle bulb of young medusae (Figs. 1N and 1O), suggesting that tentacle bulbs may behave as a proliferation zone throughout the medusa stage. Interestingly, in the manubrium of matured medusae, large numbers of EdU-positive cells were detected (Figs. 2G and 2H). This result likely reflects the presence of germ cells that are produced in the manubrium, a feature of the sexually matured Cladonema medusa (Takeda et al., 2018).

Altogether, our results suggest that cell proliferation may occur uniformly in the medusa umbrella, while a subset of cell proliferation could occur locally in tentacles. Based on these observations, we hypothesized that uniform cell proliferation may control body size growth and tissue homeostasis while clustered cell proliferation in tentacles may contribute to tentacle morphogenesis.

Cell proliferation is necessary for the control of body size

Animal body size increases upon intake of nutrition because nutrition influences cell proliferation and cell growth (Bohnsack & Hirschi, 2004). We first monitored the body size of juvenile medusae by focusing on the size of their umbrella because the umbrella grows in direct proportion with whole body size. Under normal feeding conditions, the medusa umbrella size increased dramatically by 54.8%, from 0.62 ± 0.02 to 0.96 ± 0.02 mm² during the first 24 h, with a subsequent minor increase observed over the following 5 days (0.98 ± 0.03 mm²) (Figs. 3A–3C). By contrast, under starved conditions, the size of medusa umbrella did not increase, compared to controls, and rather gradually decreased over the following 5 days (Fig. 3C). Moreover, fewer EdU positive cells were detected in the starved medusae than in fed controls (Figs. 3D–3H; Control: 1,240.6 ± 214.3, Starved: 433.6 ± 133, t(8) = 3.194, *p < 0.05 (p = 0.0127)), suggesting that, at the cellular level, nutrition affects cell proliferation in medusae. These results indicate that body-size growth in juvenile medusae depends on available nutrition.

To test the hypothesis that uniform cell proliferation in medusae contributes to body-size increase, we performed a pharmacological assay to block cell-cycle progression using HU, a cell-cycle inhibitor that causes G1 arrest (Koç et al., 2004). Under HU treatment, S phase cells detected by EdU staining disappeared from the medusa body (Figs. 3I–3L). By tracking the size of umbrella, we found that HU-treated medusae did not exhibit the size increase that was observed in controls (Fig. 3M). Together, these results suggest that cell-cycle progression affects body size in Cladonema medusae.

Cell proliferation is necessary for tentacle morphogenesis

In Clytia hemisphaerica, another hydrozoan jellyfish, stem-like cells or progenitors are proposed to exist in tentacle bulbs (Denker et al., 2008). The clustered or local cell proliferation observed in tentacles, including those in the bulb, of the Cladonema medusa may reflect such stem or progenitor cell populations (Figs. 1N–1O and 2E–2F; Fig. S1D). Furthermore, many EdU positive cells were frequently detected in small branched tentacles, which were the most recently branched (Figs. 3D–3E and 3I–3J). To test the hypothesis that local cell proliferation in tentacles contributes to tentacle morphogenesis, we first focused on tentacle branching. Although the initial tentacles have one branch in juvenile medusae, the number of branches gradually increases during medusa maturation (Fujiki et al., 2019). In our normal feeding condition, the branching number reached approximately three (2.98 ± 0.05 per tentacle) by day 9 (Figs. 4A and 4C). By contrast, when cell proliferation was blocked with HU, none of the medusae exhibited the typical increase in branched tentacles; rather, all maintained only one branch (Figs. 4B and 4C). Importantly, upon removal of HU, these animals showed an increase in tentacle branching similar to controls, suggesting that the effects of the drug treatment are reversible (Fig. S3C). Combined, these results point to cell proliferation in tentacles as a necessary component for normal tentacle branching.

Cnidarian tentacles have nematocysts, organelles specific to the cnidarian phylum that are utilized for food capture and defense against predators (Kass-Simon & Scappaticci, 2002). In Clytia hemisphaerica, stem-like cells or progenitors in tentacle bulbs seem to supply nematocysts at the tips of tentacles via cell proliferation, migration to the tip, and differentiation (Denker et al., 2008). This evidence raises the possibility that cell proliferation also controls nematocyte development or nematogenesis in hydrozoan jellyfish. To monitor nematocytes in Cladonema tentacles, we utilized DAPI, a nuclear staining dye that can label poly-γ-glutamate synthesized in the nematocyst wall (Szczepanek, Cikala & David, 2002). Using poly-γ-glutamate staining, we discovered nematocyte size variations ranging from 2 to 110 μm² (Figs. 4D–4G). We also found that some of the nematocysts were empty, suggesting that such nematocytes had been depleted (Figs. 4D–4G).

In order to investigate whether cell proliferation in tentacle also contributes to nematocyte maturation, we examined the emptiness of nematocytes after cell-cycle blocking with HU. We detected that the proportion of the empty nematocysts was higher in the medusae with HU treatment than in controls (HU−: 11.4% ± 2.0%; HU+: 19.7% ± 2.0%, Figs. 4D–4G and 4H). This result indicates that even after discharge, nematocytes are still actively supplied by progenitor cell proliferation and that this refill is prevented when cell proliferation is blocked. Taken together, our data suggest that cell proliferation in tentacle plays an important role in both tentacle branching and nematogenesis.

Cell proliferation is necessary for tentacle regeneration

Cnidarians are known to have a high regenerative capacity (Galliot & Schmid, 2002; Holstein, Hobmayer & Technau, 2003), and the hydrozoan jellyfish Cladonema species exemplifies this typical regenerative ability (Weber, 1981). Given the localization of proliferative cells in the tentacle bulb of matured Cladonema medusae (Figs. 2E and 2F), we decided to investigate the nature of tentacle regeneration. After dissecting tentacles at their base, we monitored the process of tentacle regeneration (Figs. 5A–5D). During the first 24 h, wound healing occurred at the dissected area (Fig. 5B). Subsequently, the tip of tentacle became elongated and started branching on day 2 (Fig. 5C). At day 4, fully branched tentacles were observed (Fig. 5D), suggesting that tentacle regeneration may follow normal tentacle morphogenesis after elongation.

To examine the initial stage of tentacle regeneration, we examined the distribution of proliferating cells using PH3 staining to visualize mitotic cells. While dividing cells were frequently observed near the amputated area, mitotic cells were dispersed in uncut control tentacle bulbs (Figs. 5E and 5F). We quantified the number of PH3-positive cells present in the tentacle bulbs and found a significant increase in PH3-positive cells in the tentacle bulbs of amputee medusae, compared to controls (Fig. 5G). These observations indicate that initial regenerative responses accompany the active increase of cell proliferation in tentacle bulbs.

In order to test the role of cell proliferation in tentacle regeneration, we blocked cell-cycle progression using HU after dissection and monitored the length of regenerating tentacles. While the tentacles continued to elongate from the bulb structure after dissection in controls, tentacles in animals treated with HU were not able to elongate despite displaying normal wound healing (Fig. 5H). These results demonstrate that cell proliferation in tentacle bulbs is required for proper tentacle regeneration.

Cell proliferation patterns across different hydrozoan jellyfish

Hydrozoan jellyfish constitute the most broadly varied class of cnidarian jellyfish with approximately 1,150 species worldwide featuring highly diverse morphological and physiological characteristics (Cartwright & Nawrocki, 2010; Schuchert, 2019). For instance, Cytaeis uchidae has four tentacles, and their polyps live exclusively on one type of shell: Niotha livescens (Takeda, Deguchi & Itabashi, 2018; Takeda et al., 2013). Another species, R. octopunctata, has eight grouped-tentacles, and their juvenile medusae asexually produce medusae that grow out of the manubrium (Berrill, 1952; Schuchert, 2007).

To gain insight into the conserved and diversified nature of cell proliferation in hydrozoan jellyfish, we investigated the spatial pattern of cell proliferation in Cytaeis and Rathkea medusae. In Cytaeis medusa, EdU-positive cells were observed in manubrium, tentacle bulbs, and at the top of the umbrella (Figs. 6A and 6B). PH3-positive cells were also detected in the same regions, suggesting that proliferating cells in Cytaeis are distributed in a pattern similar to that observed in Cladonema, although there are some discrepancies (Figs. 6C and 6D). By contrast, in R. octopunctata, EdU-positive cells and PH3-positive cells were mostly restricted to the manubrium and tentacle bulbs (Figs. 6E–6H). Of note, proliferating cells were frequently detected in the medusa buds that grew out of the manubrium (Figs. 6E–6G), which may reflect asexual reproduction in Rathkea medusae. These results suggest that cell proliferation may occur in tentacle bulbs across hydrozoan medusae commonly, while cell proliferation patterns may vary in a species-specific manner with physiology.