The Buds of Oscarella lobularis (Porifera, Homoscleromorpha): A New Convenient Model for Sponge Cell and Evolutionary Developmental Biology

ABSTRACT

The comparative study of the four non‐bilaterian phyla (Cnidaria, Placozoa, Ctenophora, and Porifera) provides insights into the origin of bilaterian traits. To complete our knowledge of the cell biology and development of these animals, additional non‐bilaterian models are needed. Given the developmental, histological, ecological, and genomic differences between the four sponge classes (Demospongiae, Calcarea, Homoscleromorpha, and Hexactinellida), we have been developing the Oscarella lobularis (Porifera, class Homoscleromorpha) model over the past 15 years. Here, we report a new step forward by inducing, producing, and maintaining in vitro thousands of clonal buds that now make possible various downstream applications. This study provides a full description of bud morphology, physiology, cells and tissues, from their formation to their development into juveniles, using adapted cell staining protocols. In addition, we show that buds have outstanding capabilities of regeneration after being injured and of re‐epithelization after complete cell dissociation. Altogether, Oscarella buds constitute a relevant all‐in‐one sponge model to access a large set of biological processes, including somatic morphogenesis, epithelial morphogenesis, cell fate, body axes formation, nutrition, contraction, ciliary beating, and respiration.

We describe a technique for triggering asexual reproduction by budding in vitro and have monitored bud development. The natural properties of Oscarella lobularis buds and the technical tools we developed make them a new biological system for getting insights into sponge molecular and cellular processes.

1. Introduction

Ctenophora and Porifera are currently the two best candidates as sister groups of all other extant animals (Daley and Antcliffe 2019; Erives and Fritzsch 2019; Laumer et al. 2019; Li et al. 2021; Schultz et al. 2023; Simion et al. 2017; Whelan et al. 2017). Accordingly, they can provide clues to understand the features held by the last common ancestor of Metazoa, and to conceive the early evolution of animal body plans, including the evolution of cell types, of body axes, and of master regulator genes involved (King and Rokas 2017; Renard, Le Bivic, and Borchiellini 2021; Schenkelaars et al. 2019). Nonetheless, several basic aspects of the biology of these animal phyla are still misunderstood (Dunn, Leys, and Haddock 2015), calling for the development of new biological models for evo‐devo research purposes.

Among sponges (Porifera), homoscleromorph sponges (Homoscleromorpha) are of particular interest to study the evolution of epithelia and of epithelial morphogenetic processes because of the histological similarity between their epithelium and that of Bilateria with the clear presence of adherens‐junctions and Type IV collagen basal lamina (Belahbib et al. 2018; Boute et al. 1996; Ereskovsky, Dubois, et al. 2013; Ereskovsky et al. 2009; Fahey and Degnan 2010; Leys, Nichols, and Adams 2009; Leys and Hill 2012; Leys and Riesgo 2012; Miller et al. 2018; Mitchell and Nichols 2019; Nichols et al. 2006, 2012; Renard, Le Bivic, and Borchiellini 2021; Riesgo et al. 2014). In addition, several studies suggested a higher conservation of some bilaterian genes in homoscleromorphs in comparison to other sponge classes (Gazave, Guillou, and Balavoine 2014; Renard et al. 2018; Renard, Rocher et al. 2021; Riesgo et al. 2014). Among Homoscleromorpha, Oscarella lobularis (Schmidt, 1862), the type species of the Oscarella genus, has benefited from in‐depth studies: its morphology, phylogeny, histology, embryology, and ecology have been thoroughly described, leading authors to consider it as “a promising sponge model” (Boury‐Esnault, Sole‐Cava, and Thorpe 1992; Boury‐Esnault et al. 2003; Ereskovsky 2010; Ereskovsky et al. 2009; Ereskovsky, Renard, and Borchiellini 2013b; Ereskovsky and Boury‐Esnault 2002; Fierro‐Constaín et al. 2017; Gaino, Burlando, and Buffa 1986a, 1986b, 1986c; Gazave et al. 2010, 2012, 2013; Gloeckner et al. 2013; Leys and Ereskovsky 2006; Renard et al. 2021b).

O. lobularis adults are commonly found on the rocky substrates of the north‐western Mediterranean Sea. This species is capable of both sexual and asexual reproduction (Ereskovsky and Tokina 2007; Ereskovsky, Renard, and Borchiellini 2013; Fierro‐Constaín et al. 2017) like many other sponges. In sponges, asexual reproduction may proceed by fragmentation, budding, and gemmulogenesis (Ereskovsky, Geronimo, and Pérez 2017; Maldonado and Riesgo 2008; Shaffer et al. 2020; Simpson 1984; Simpson and Fell 1974; Wulff 1991). Budding has been observed in all classes of sponges (Cardone, Gaino, and Corriero 2010; Chen, Chen, and Chang 1997; Connes, 1967, 1971; Ereskovsky 2010; Gaino et al. 2006; Hammel et al. 2009; Schulze, 1887; Singh and Thakur 2015), although the development of buds and the rate of successful development of buds to fully functional individuals differ between species. In all investigated sponges, except Homoscleromorpha, budding involves migration of some (multipotent or differentiated) cells from the inner to the outer regions of the parental sponge, resulting in the formation of cell aggregates (Fell 1993). In contrast, budding in Homoscleromorpha does not involve the migration of cells from the mesohyl; instead, it is based on epithelial morphogenesis (Ereskovsky and Tokina 2007).

In this article, we describe a technique for triggering asexual reproduction by budding in vitro and have monitored bud development from their formation up to the juvenile stage. We also show their important regeneration and reaggregation abilities. Altogether, the natural properties of O. lobularis buds and the technical tools we developed at this stage make them a new, convenient, and suitable biological system for getting insights into molecular and cellular processes occurring during sponge somatic morphogenesis.

2. Materials and Methods

2.1. In Situ Life Cycle Monitoring

The in situ monitoring of a small population in the bay of Marseille (western Mediterranean Sea) was carried out. The six target adults of different sizes and colors were located on the north edge of Maïre Island (43.212518° N, 5.330122° E), in a large, shallow, and rocky cavity (6–10 m). In situ photographs of the sponges were taken monthly from April 2014 until October 2015. In addition, a small fragment of each individual was sampled twice a month, and fixed (in 4% paraformaldehyde [PFA] overnight [O/N]). Histological observations and in situ hybridization were realized on sections of these small fragments (Fierro‐Constaín et al. 2017) to determine the beginning and end of the sexual reproduction period.

2.2. In Vitro Budding and Cultivation of Buds in Lab Conditions

Adult specimens of O. lobularis (Schmidt 1862) were collected by SCUBA diving in the northwestern Mediterranean Sea (Marseille Bay). Individuals were transported in plastic pots to the laboratory in less than 1 h in a cool box. Upon arrival at the laboratory, individuals were carefully cleaned under a stereomicroscope to get rid of other organisms intermingled with sponge tissues (cnidarians, annelid tubes, and algae) and cut into pieces of about 1 cm³ with a sterile scalpel. Each fragment was placed in a well (3.5 cm diameter) of a 6‐well cell culture plate containing 8 mL of natural sea water (NSW). Culture plates were maintained at 17°C in a thermostated room, and NSW was renewed each day until the end of budding.

Once buds were released from the adult fragment, they were transferred to Petri dishes (14 cm diameter) containing NSW and maintained at 17°C. NSW was renewed once a week. In other conditions (e.g., temperature, volume of NSW, frequency of sea water renewal, and the number of buds per well), budding and bud developmental timing might be different.

To estimate the efficiency of budding in vitro, three adults of O. lobularis sampled in situ in January 2018 were placed in a thermostated room at 17°C O/N, before starting the experiment. The budding experiment was performed as described above. Each sample was cut into 12 fragments. Given the mortality of one of the three individuals, 17 fragments (out of 24) from the two other individuals were monitored. Samples were observed and imaged, and buds counted every day. The bud production per day was analyzed using linear mixed‐effects models that were implemented within the lme4 package in the R software (R Core Team 2013). We considered time as fixed effects and we allowed linear and/or quadratic effects on the dependent variable. We also added individuals (with random slope of time within individuals) and replicates (fragments) nested in individuals as random effects. After observation of the residual's distribution of the models, we retained Gaussian distribution. We tested the significance of fixed and random effects with F tests and likelihood ratio tests, respectively. When two models were equivalent, we retained the most parsimonious model.

2.3. TL Photonic, SEM, SBF‐SEM3, and TEM Microscopy Observations of Budding Adults and Buds

To describe the budding process and bud morphological and histological features, several microscopy techniques have been used. For scanning (SEM), transmission (TEM) electron microscopy and the preparation of semi‐thin sections, a standard fixation method for TEM was used: glutaraldehyde 2.5% in a mixture of 0.4 M cacodylate buffer and post‐fixation in 2% osmium tetroxide (OsO₄) in sea water as first described in Boury‐Esnault et al. (1984). Semi‐thin sections were stained with toluidine blue. For SEM, the specimens were fractured in liquid nitrogen, critical point‐dried, sputter‐coated with gold‐palladium according to the protocol previously used for this species (Johnston and Hildemann 1982). For serial block face‐scanning electron microscope 3 (SBF‐SEM3) observations, samples were prepared using the National Center for Microscopy and Imaging Research (NCMIR) protocol for SBF‐SEM (Deerinck et al. 2010). Imaging was carried out on a FEI Teneo VS running in low vacuum (30 Pa) at 2 kV and using a backscattered electrons detector. Acquisition voxel size was 5 × 5 × 40 nm.

2.4. Observation of Mucus on Buds' Surface

To evaluate the role of body outgrowths in bud settlement, we studied the localization of mucus. Buds were fixed in 3% PFA in phosphate‐buffered saline (PBS) at 4°C O/N, and then washed three times with 1 mL of PBS (15 min each) at room temperature (RT). Buds were then incubated 10 min in 1 mL of Alcian Blue solution (Alcian Blue 1% in acetic acid 3%, pH 2.5) (20:1000 in NSW) at RT, rinsed in two baths with NSW, and immediately mounted with ProLong Antifade Mountant. After polymerization (40 min at 37°C in the dark), the samples were observed with a stereomicroscope.

For TEM observations, we used a special fixation with ruthenium red (Lavrov and Ereskovsky 2022). Buds were fixed in buffered 2.5% glutaraldehyde + 0.1% ruthenium red for 2 h at 4°C, and then rinsed three times with the 0.1 M Na‐cacodylate buffer (30 min each) at RT. Post‐fixation was buffered in 1% OsO₄ + 0.1% ruthenium red in the same buffer for 3 h at RT, followed by rinsing three times with the same buffer (30 min each) at RT, then dehydration in ethanol series and inclusion in a resin.

2.5. Immunofluorescent Assays on Whole Bud

Immunofluorescent assays have been developed to complement bright‐field and electronic microscopy methods to better characterize cell types. Buds are usually fixed (6–24 h) in 3% PFA in PBS at 4°C. Immunostaining of alpha‐tubulin, acetylated alpha‐tubulin, and Type IV collagen were performed as described in Vernale et al. (2021).

Imaging of tissues and cells was performed with an inverted confocal microscope Zeiss Axiovert 200 M or confocal microscope Zeiss Axio Imager Z2 780 or 880 (objective ×63‐NA 1.4). The FIJI software (Schindelin et al. 2012) was used for image editing.

2.6. Time‐Lapse Imaging of Budding, Buds, and Cells

To study the dynamics of budding and bud behavior, we performed time‐lapse imaging. To follow the budding process, adult fragments were placed in a dark chamber at 17°C in 6‐well cell culture plate containing 8 mL NSW. We used a CANON EOS 750D camera with an EF‐S 60 mm 1/2.8 1:2.8 USM Ultrasonic Macro Lens to acquire one picture every 15 min for 6 days.

We used the same protocol for imaging the contractions of the free buds in vitro, in this case the time‐lapse imaging was performed by taking one picture every 5 min for 48 h. For each bud, the maximum and the minimum sizes were measured in pixels to evaluate the amplitude of bud contraction/release. The mean and the standard deviation (SD) were calculated on the R software (https://www.r-project.org/). For analysis of bud motions, Developmental Stage 1 buds were placed in 96‐well cell culture plate containing 200 µL NSW and one picture was taken every 30 s for 24 h. Bud trajectories and volume changes recorded by a camera were made with a Deeplabcut tracking system (Mathis et al. 2018). Prism was used for graphical representations.

To observe cell motility, time‐lapse imaging was performed on Developmental Stage 3 buds placed into 6‐well cell culture plates with NSW at 22°C. We used an inverted microscope Zeiss Axio Observer Z1 in DIC TL, with an objective ×20‐NA 0.7, to acquire one picture every 10 s for 10 min (total 61 pictures).

To observe cilia beating on the exopinacoderm, time‐lapse imaging was performed on a bud with a microscope Zeiss Axio Imager Z1 with DIC II (objective ×40‐NA 1.3) with a rate of one picture every 34.3 ms for 30 s.

2.7. Testing Filtering Activity With Fluorescent Microspheres

To investigate whether buds are capable of water filtration or not, the absorption of fluorescent beads, comparable in size with bacteria, was monitored. Ten Developmental Stage 1 buds and 10 Developmental Stage 3 buds were incubated for 15 min, 30 min, and 1 h with 0.2 µm FluoSpheres Carboxylate‐Modified Microspheres Yellow‐Green (505/515 nm) (Molecular Probes) in 2 mL (1/10,000) of NSW at 17°C (Borchiellini et al. 2021). Buds were then rinsed three times with NSW and fixed (3% PFA) at 4°C O/N. After DAPI/phalloidin counterstaining and mounting as described in the previous section, imaging of samples was performed with a confocal microscope Zeiss LSM 880 Leica. The same experiment was reiterated after a 40 min nocodazole (33 µM) treatment (nocodazole is an inhibitor of tubulin polymerization, a process required for cilia beating). As a control, buds were incubated for 40 min in a dimethyl sulfoxide (DMSO) solution (1/1000, the same concentration as that used to solubilize nocodazole).

2.8. Oxygen Consumption

The physiological activity of buds has also been estimated indirectly by comparing bud and adult oxygen consumption. Oxygen consumption (MO₂) of buds (n = 6) and adult fragments (n = 6) was measured in closed glass vials, connected to a FireSting O₂ sensor (PyroScience; Aachen, Germany). One adult fragment (295–1100 mg fresh weight [FW]) or a pool of buds (23–180 mg FW) was placed in a vial filled with 20 mL of filtered NSW (T = 17°C). After the vial was sealed, connected to the sensor, and placed in a thermoregulated chamber at 17°C, the decrease in dissolved O₂ concentration was monitored in real time for a period of 60 min in the dark. To homogenize O₂ during the experiment, water was gently stirred by a magnetic bar, and separated from the sample with a thin plastic grid. After a 20–30 min period of acclimatization to the experimental conditions, MO₂ was calculated by measuring the decrease in oxygen over a period of 30 min. After the experiment, samples were gently and carefully dried on a paper towel and weighed. MO₂ was expressed as µgO₂/h/mg FW. Differences in oxygen consumption between buds and adults were evaluated using a Mann–Whitney U‐test. Statistical analyses were performed using XLSTAT (Assinsoft, Paris, France).

2.9. Cell Proliferation and Apoptosis Assays

To evaluate the dynamics of cells during bud development, cell division and death were both evidenced via cell staining methods. Cell proliferation was evaluated by phosphohistone H3 (PHH3) immunostaining and 5‐ethynyl‐2′‐deoxyuridine (EdU) incorporation on Developmental Stage 1 and 3 buds, the detailed protocols of which are available in Borchiellini et al. (2021). Phalloidin staining was used to identify cell types and compare their relative division rates.

Apoptosis activity has been observed via terminal deoxynucleotidyl transferase dUTP nick‐end labeling (TUNEL) staining as described in Borchiellini et al. (2021) on the same developmental stages.

2.10. Regenerative Experiments

To investigate whether buds are able of body regeneration, Developmental Stage 3 buds (n = 58, three independent technical replicates), with a differentiated osculum, were cut into two halves: one half corresponding to the future basal pole of the juvenile, and the other half possessing the osculum and corresponding to the future apical pole of the juvenile. The 116 half‐buds obtained after cutting are placed in 24‐well culture plates containing 2 mL of NSW (renewed every day), and the culture plate is placed at 17°C. Wound healing and regeneration were monitored once a day for 96 h.

Two additional preliminary experiments consisted of staining choanocytes with the Phaseolus vulgaris Erythroagglutinin (PhaE) lectin (as shown in the next section) before cutting to evaluate choanocyte migration during the regenerative process.

2.11. Cell Dissociation–Reaggregation Experiments

To investigate if Oscarella cells are able to reaggregate after complete cell dissociation, a cell dissociation–reaggregation experiment was implemented on a large number of buds. Thousands of buds were collected in a Falcon tube (15 mL) and then slowly centrifuged (10g for 2 min at RT). The supernatant NSW was discarded and the bud solution (1.5 mL) was incubated in 20 mL calcium magnesium–free sea water (CMSFW: 0.54 M NaCl, 10 mM KCl, 7 mM Na₂SO₄, 15 mM Tris‐HCl, 0.2 mM NaHCO₃, pH 8), with 10 mM ethylenediaminetetraacetic acid (EDTA) pH 8 and stirred for 1 h at RT. The cell suspension was filtered on a 17‐μm cell strainer. The concentration of cellular suspension was evaluated using C‐Chip Neubauer. Before counting, the viability of cells can be efficiently evaluated with fluorescein diacetate (FDA) and propidium iodide (PI) staining (Sigma) according to the protocol used by Sipkema et al. (2004). Then, 2 mL of cell suspension (between 0.4 × 10⁶ and 0.56 × 10⁶ cells/mL) were dispatched in a 24‐well cell culture plate and CaCl₂ solution was added to each well (final concentration 20 mM). The culture plate was placed at 17°C and stirred very slowly on a 3D platform shaker. After reaggregation, cell aggregates were carefully transferred in 2 mL artificial sea water (ASW: 10 mM CaCl₂, 50 mM MgCl₂.6H₂O, 0.54 M NaCl, 10 mM KCl, 7 mM Na₂SO₄, 15 mM Tris‐HCl, 0.2 mM NaHCO₃, pH 8) into 6‐well cell culture plates. Culture plates were then placed at 17°C and stirred very slowly on a 3D platform shaker. The reaggregation process was monitored for 7 days. This procedure was repeated three times.

2.12. Cell Staining

With the aim of developing cell tracking during developmental processes in Oscarella, we first carried out cell staining assays. Choanocyte staining, using the lipidic marker Cell Tracker CM‐DiI Dye Red (553/570 nm) (Molecular Probes) and fluorescein‐labeled PhaE, was performed as described in detail in Borchiellini et al. (2021). This specific staining enabled tracking choanocytes during regeneration experiments (described above) via fixation of buds (PFA 4%/PBS) at different time points.

Other lectins and agglutinins provided in Vector Laboratories kits (Kit I, RLK‐2200; Kit II, FLK‐3200; Kit III, FLK‐4100) were tested at different concentrations (1:100, 1:500, 1:1000, and 1:2000 in NSW) for 15 min, 30 min, and 1 h at 17°C. After each treatment, buds were rinsed three times in NSW and immediately observed with an epifluorescence microscope DM2500 Leica and a confocal Zeiss LSM 510 (for CM‐Dil dye treatment). In the second step, the dilution and incubation times were optimized as follows: for Griffonia simplex lectin 1 (Gsl1): dilution 1/700, incubation 1 h, and wheat germ agglutinin (WGA): dilution 1/7000 incubation from 10 to 30 min depending on the expected staining.

Samples were then fixed with 3% PFA/PBS at 4°C O/N, counterstained with DAPI (1:500) (Thermo Fisher Scientific) and 1:1000 iFluor 488, rhodamine, or Alexa‐647 (or 568)‐coupled phalloidin (Abcam, Thermo Fisher Scientific), mounted in Prolong Diamond antifading mountant (Thermo Fisher Scientific), and observed with a confocal microscope (Zeiss AxioImager Z2 780 or 880).

3. Results

3.1. Budding Is a Natural Process

A 2‐year in situ monitoring performed on six target adults from a population of O. lobularis from Marseille (43.212518° N, 5.330122° E) enabled us to monitor natural reproduction events and to complete and refine the life cycle of this species (Figure 1). We observed two types of asexual reproduction: by fragmentation (Figure 1A) and by budding (Figure 1B).

Figure 1.

Reproduction events during the life cycle of Oscarella lobularis. (A) Asexual reproduction by fragmentation (arrow heads) occurs at fall, scale bar = 1 cm. (B) Asexual reproduction by budding yields to the release of free‐floating buds, scale bar = 500 µm. (C–G) Sexual reproduction takes place from April to October. (C) Oocytes (green arrows) near a choanocyte chamber, scale bar = 8 µm. (D) Spermatozoa in spermatocysts, scale bar = 25 µm. (E) Released spermatozoa, scale bar = 2 µm. (F) Embryo dissected from maternal tissues, scale bar = 50 µm. (G) Swimming larva, scale bar = 150 µm (modified from Fierro‐Constain 2016).

Asexual reproduction events occurred naturally several times during the year, from October to April, suggesting an exclusion of asexual and sexual reproduction in the same individual. Unlike sexual reproduction, fragmentation and budding do not seem to be synchronized between individuals. Budding events were observed at several periods in the fall and early spring, whereas fragmentation occurred in the fall after the end of sexual reproduction (Figure 1 and Supporting Information S2: Table S1).

3.2. The Budding Process Follows Three Steps

The budding process triggered in vitro (Section 2) can be divided into three steps (Figure 2) that are identical to those previously observed during natural budding (Ereskovsky and Tokina 2007). First, the smooth body surface of the fragment becomes irregular (Budding Step 1, upper pictures of Figure 2). Then, finger‐like protrusions appear and grow (Budding Step 2). Finally, protruding tissues swell via an increase in volume of the central cavity (Budding Step 3). At the same time, the proximal pole of protrusions becomes tight, resulting in a short stalk. All tissues of the adult are concerned by this process. Accordingly, each just‐release bud contains parts of the exopinacoderm, endopinacoderm, mesohyl, and choanoderm (Figure 3). Time‐lapse imaging of the budding process (Movie S1) shows that budding begins at the most external part of the adult fragment and moves toward the center of the fragment (centripetal progression) until only a shred of tissue remains. The release of free buds starts at 6 days after cutting with a maximum of release at 16 days (Figure 4A and Supporting Information S2: Table S2). The efficiency of budding differs from one individual to another (Figure 4B and Supporting Information S2: Table S3), but a 1 cm³ fragment of adult produce a mean of 450 buds (Figure 4C) that range in diameter from 1 to 2.5 mm, with a mean size of 1.87 mm (Figure 4D). In addition to the large number of buds obtained during this process, their transparency makes them a suitable system for experiments (see the following sections).

Figure 2.

Kinetics of budding. Alive Oscarella lobularis fragments and buds were observed using a stereo microscope. Budding Step 0: Smooth surface of a freshly sampled individual; Budding Step 1: the surface becomes irregular; Budding Step 2: the surface is covered by finger‐like protrusions; Budding Step 3: The swelling up of protrusions results in the formation of buds that are subsequently released. Scale bars = 1 mm (see Movie S1).

Figure 3.

Comparison of adult (A) and bud tissues (B). Histological comparison of tissues showed the presence of exopinacoderm (Exp), endopinacoderm (Enp), mesohyl (M), inhalant canal (Ic), internal cavity (inc), ostium (Os), and choanoderm (Cc) in adult and bud. Scale bars = 40 µm.

Figure 4.

Kinetics of bud release. (A) Kinetics of budding (Supporting Information S2: Table S2), budding starts 6 days after cutting and reaches a peak of bud release between 13 and 16 days after the beginning of the budding process (red curve). The bud production is correlated with the percentage of fragments that reached the third step of the budding process (blue curve). (B) Daily release (cumulative) of new buds per fragment (17 fragments from two individuals) (see Supporting Information S2: Table S3). (C) Average bud production for each of the two monitored individuals (mean: 451.4; SD: 225.56). (D) Size of just‐released buds (Developmental Stage 1 buds) (Feret's diameter (mm)); the diameter of 84% of the buds, ranges from 1 to 2.5 mm (mean: 1.87 mm; SD: 0.696 mm).

3.3. Bud Morphoanatomical Features and Development

3.3.1. From Bud Release to Settled Juvenile

Once released, buds undergo developmental processes that can be subdivided into four morphologically distinct developmental stages (Figure 5A). The timing of occurrence of each developmental stage is indicated in days since bud release (timing observed in standardized conditions as indicated in Section 2). Just after being released (Day 0, Developmental Stage 1), buds have a spherical shape with a thin tissue layer surrounding a large central cavity (diameter ranging from 160 to 700 µm). The general tissue organization of the bud surrounding the central cavity is similar to that found at the adult stage, despite a simpler organization of the aquiferous system (Figure 5B). The exopinacoderm forms the most external cell layer of the bud and is perforated by ostia (inhalant pores). The endopinacoderm lines short canals of the aquiferous system and the internal cavity. The choanoderm is composed of choanocyte chambers communicating with exhalant canals by apopyles. The chamber dimensions range from 36 to 52 µm (mean 44.7 µm; Supporting Information S2: Table S4 and Supporting Information S1: Figure S1), a size range comparable to the smaller chambers observed in adults (Boury‐Esnault et al. 1984). Between these epithelial layers, the mesohyl, a mesenchymal layer, includes different sponge cell types (next section) and symbiotic bacteria as was previously described for the adult stage (Ereskovsky et al. 2015; Gloeckner et al. 2013). It has been suggested previously that symbiotic bacteria are inherited during budding from the maternal sponge (Ereskovsky and Tokina 2007).

Figure 5.

Development of a bud. (A) Schematic view of development of bud (B) Developmental Stage 1: Spherical just‐released bud. (C) Developmental Stage 2: Bud with outgrowths (blue arrows) devoid of choanocyte chambers (zoom in) and with distal roundish cells (pink arrows). (D) Developmental Stage 3: Bud with osculum (yellow arrows) and outgrowths (blue arrows) clustered on the opposite pole indicating the presence of a polarity. (E) Developmental Stage 4/juvenile: The juvenile is settled owings to basopinacocytes and has a reduced internal cavity (atrium). Scale bars = 200 µm (first column), 100 µm (second column). Ap, apopyle; Cc, choanocyte chamber; Ec, exhalant canal; Enp, endopinacoderm; Exp, exopinacoderm; Ic, inhalant canal; Inc, Internal cavity; M, mesohyl; Osc, osculum; Os, ostium.

Between Days 1 and 7 (Developmental Stage 2), the global shape becomes variable and asymmetric because of the formation of external outgrowths. Outgrowths are only composed of exopinacoderm and mesohyl and, at their distal pole, we notice the presence of more roundish cells which are absent or rare at Stage 1 (Figure 5 C). Histologically, the tissue organization at Developmental Stage 2 is similar to that observed at Developmental Stage 1 (Figure 5B).

The bud's Developmental Stage 3 is observed between Days 8 to 29 and is characterized by the development of an osculum and the establishment of a clear body polarity. The outgrowths are mainly gathered at one pole (the future basal pole), while the opposite pole (the future apical pole) is characterized by the abundance of choanocyte chambers and the presence of an osculum (Figure 5D). Choanocyte chambers are larger compared to that observed at Developmental Stage 1 (up to 70 µm large; Supporting Information S2: Table S4 and Supporting Information S1: Figure S1). The development of a polarized body thereby predates settlement.

Indeed, at Developmental Stage 4 (from 1 month old; Figure 5E), buds become attached to the substrate with the osculum pointing to the surface. The adhesion to the substrate is probably performed by the abundant mucus present at the distal parts of the outgrowths (Figure 6A,E,F). These settled juveniles (with a reduced internal cavity or atrium) are similar to juveniles formed after larval metamorphosis (Ereskovsky and Tokina 2007; Ereskovsky et al. 2007). Some Developmental Stage 3 buds fail to settle; the unsettled buds can nevertheless stay alive for several months in culture plates.

Figure 6.

Cell composition and mucus production in Outgrowths. (A) External view of a Developmental Stage 2 bud showing accumulation of mucus (Alcian blue staining) at the distal part of the outgrowths (blue arrows). (B–D) Observations of a Developmental Stage 2 bud outgrowth on (B) semi‐thin histological section, (C) white light microscopy, and (D) MEB, showing the difference between flat exopinacocytes and roundish basopinacocytes. (E) Basopinacocytes (TEM) are covered by thicker mucus layer (blue arrows) than exopinacocytes (TEM) (F). Scale bars = 100 µm (A); 20 µm (B, C); 43 µm (D); 2 µm (E,F).

In brief, the buds obtained in vitro present the same morphological and histological features as buds resulting from natural budding (Ereskovsky and Tokina 2007) and the juveniles resulting from somatic development via budding are similar to those resulting from embryonic and post‐embryonic development via sexual reproduction (Ereskovsky 2010; Renard, Rocher, et al. 2021).

3.3.2. Bud Cell Types Compared to Adult Cell Types

Bud cells were observed thoroughly by using a combination of semi‐thin histological sections, electron microscopy (TEM, SEM) as already performed in adult tissues, and, in addition, owing to the transparency of buds, fluorescent cell staining on whole‐mount buds was developed and observed by confocal microscopy.

The cell types previously described in adult tissues and in Developmental Stage 1 buds are: exopinacocytes, endopinacocytes including prosopinacocytes and apopinacocytes, choanocytes, apopylar cells, Type I vacuolar cells and Type II vacuolar/archaeocyte‐like cells were observed (Boury‐Esnault, Sole‐Cava, and Thorpe 1992; Ereskovsky 2010; Ereskovsky and Tokina 2007; Fierro‐Constaín et al. 2017; Renard, Rocher, et al. 2021) (Figures 7 and 8A–D and Movie 2).

Figure 7.

Cell types of Oscarella lobularis buds. Position of each cell types in the tissues of a Developmental Stage 2 bud observed on histological sections. For each cell type, the left picture is a max intensity projection of confocal Z‐stack, scale bar = 5 µm; on the right a TEM picture, scale bar = 2.5 µm. Cc, choanocyte chamber; Enp, endopinacoderm; Exp, exopinacoderm; Ic, inhalant canal; Inc, internal cavity; M, mesohyl; Os, ostium.

Figure 8.

Filopodia and extracellular vesicles in bud tissues. (A) All cell types (whether they are epithelial cell types, endopinacocytes, exopinacocytes, choanocytes, and apopylar cells, or mesenchymal cell types, Type I, II, or III vacuolar cells) possess long and thin actin‐rich cytoplasmic protrusions (white arrows), filopodia, clearly allowing to establish cell‐cell contacts, scale bar = 10 µm. (B, C) In epithelial cell types, such as exopinacocytes and choanocytes visible in this picture, these filopodes (white arrows) pass through the Type IV collagen‐rich (magenta arrows) basement membrane. Actin‐rich adherens‐like junctions are present in epithelial layers, here between exopinacocytes (red arrows), scale bar = 5 µm. (D) Pinacocytes surrounding the ostium with actin protrusions (white arrows), scale bar = 5 µm. (E) Localization of extracellular vesicles (EVs) (yellow arrows) in the mesohyl, scale bar = 5 µm. (F) Extracellular vesicle (yellow arrows) observed by scanning electron microscopy (SEM), scale bar = 5 µm. (G) Very close vicinity of EVs to exopinacocytes observed in semi‐thin histological sections (yellow arrows), scale bar = 10 µm. (H) Variety of shapes of the EVs, scale bars = 5 µm. (I) Contact or very close vicinity of EVs to exopinacocytes observed in confocal imaging, scale bar = 5 µm. Cc, choanocyte chamber; Enp, endopinacoderm; Exp, exopinacoderm; Ic, inhalant canal; Inc, internal cavity; M, mesohyl; Os, ostium. A–I are maximum intensity Z‐stack projections.

As for adults, five of these cell types are engaged in epithelial layers. These epithelial cell types can be easily distinguished and fit the three key criteria defining epithelial cells (Rodriguez‐Boulan and Nelson 1989; Tyler 2003): a clear cell polarity given by an apical cilium or flagellum (Figures 7 and 8B–D) and/or the abundance of basal actin‐rich filopodia (Figure 8A–C), actin‐rich adherens–like junctions (Figure 8C), and a basement membrane containing Type IV collagen (Figure 8B,C). All epithelial cells are covered externally by glycocalyx and mucus layers, the thickness of which can differ (Figure 6E,F).

The choanocytes, lining choanocyte chambers (choanoderm), are composed of a cellular body, an apical collar of microvilli and a flagellum composed of acetylated alpha‐tubulin (Figure 7) and alpha‐tubulin. The basal pole of choanocytes harbors long actin‐rich cytoplasmic protrusions, filopodia, that pass through the basement membrane and contact other cells (Figure 8A,B). We managed to develop several methods for staining choanocytes specifically, thereby providing a convenient way to track these cells (Figures 9 and 10, see next sections): by feeding with fluorescent microspheres (Renard, Rocher, et al. 2021) (Figure 9, see next section), by using the lipophilic marker CMDiI (Borchiellini et al. 2021) as described in Amphimedon queenslandica (Nakanishi, Sogabe, and Degnan 2014) (Figure 10A), or by using the fluorescent lectin PhaE (Phaseolus vulgaris Erythroagglutinin) (Borchiellini et al. 2021; Vernale et al. 2021) (Figure 10B). Here, we showed that GSL1, Griffonia simplifolia Lectin 1, can also be used for a very specific staining of choanocytes (Figure 10C).

Figure 9.

Active filtering activity of buds evidenced by 0.2 µm fluorescent beads. (A) Location of beads in a Developmental Stage 1 buds (devoid of osculum) after 15‐, 30‐ and 60‐min incubation with beads. Arrows of different colors show different locations: canals, choanocyte chambers cavity or microvilli, choanocyte cytoplasm, and mesohylar cell cytoplasm. (B) Location of beads in a Developmental Stage 3 buds (with osculum) after 30‐min incubation with beads, in choanocytes microvilli or choanocyte cytoplasm and in mesohylar cells. (C) Absence of filtering activity after a nocodazole treatment (near absence of beads in Developmental Stage 3 buds after 60‐min incubation) (control: DMSO treatment; Supporting Information S1: Figure S4). Scale bar: 10 µm (A–C). Cc, choanocyte chamber; Exp, exopinacoderm; Ic, inhalant canal; M, mesohyl. Max intensity projection of confocal Z‐stack (A–C).

Figure 10.

Cell staining and tracking. (A) Observation (biphotonic microscope) of choanocyte chambers in an alive bud labelled by the lipidic marker CM‐DiI (red). (B) PhaE lectin staining of choanocytes (green). (C) GSL1 lectin (magenta) staining of choanocytes (white arrows). (D) Rhodamine‐WGA (magenta) staining of exopinacocytes (white arrow). Scale bars = 50 µm (A); 10 µm (B–D); 2 µm (Zoom B); 5 µm (Zoom D). Ap, apopyle; Cc, choanocyte chamber; Ch, choanocyte, Exp, exopinacocytes; Ic, inhalant canal; M, mesohyl; Osc, osculum. Max intensity projection of confocal Z‐stack (B–D).

In the case of pinacocytes, inhalant and exhalant canals are lined with endopinacocytes (prosopinacocytes and apopinacocytes, respectively) (Boury‐Esnault and Rützler 1997; Leys and Hill 2012), whereas exopinacocytes line external parts of the bud. Both have the same ovoid to flattened shape as in adults (Ereskovsky 2010). Each exopinacocyte possesses one very long cilium (up to 20 µm; Figure 7), which is longer than that of endopinacocytes (5–10 µm). The acetylation of the alpha‐tubulin composing the cilia of exopinacocytes is more heterogeneous than that observed in choanocytes, because of a spot‐like distribution along the cilium (Figures 7 and 8C,D). As for choanocytes, both SBF‐SEM (Movie S2) and confocal imaging on whole buds show the presence of long actin‐rich expansions at the basal pole of this cell type passing through the basement membrane and often contacting other cells (Figure 8A–C). Pinacocytes surrounding the ostia often present a fringe of actin protrusions at the edge of the pore (Figure 8D) not reported so far.

Both endo‐ and exopinacocytes contain more vacuoles in buds than in adults as previously described (Ereskovsky et al. 2015). During the development of buds, round‐ to oval‐shaped exopinacocytes, with large vacuoles and very thick glycocalyx and mucus layers, are visible at the distal pole of the external outgrowths (Figure 6A–C). These exopinacocytes possess short apical and long basal actin‐rich cell extensions connected with other cell types (Movie S3). According to their position, their morphology, and their putative function in bud settlement, we consider these cells as basopinacocytes, which are currently defined as pinacocytes affixing the sponge to the substratum by external secretion of a mucous layer (Boury‐Esnault and Rützler 1997; Ereskovsky 2010; Leys and Hill 2012). Mucus‐secreting basopinacocytes have been described in Homoscleromorpha during juvenile‐substrate binding (Ereskovsky 2010). We managed to stain pinacocytes by short incubations with the WGA (Figure 10D). The staining we obtained is definitely less specific than the ones mentioned for choanocytes. Indeed, the staining of other cell types is observed when increasing the incubation time (> 20 min) or concentration (dilution < 1:6000).

The other epithelial cell type is the apopylar cell. This cell type forms the boundary between choanocyte chambers and exhalant canal or central cavity. They are characterized by a flat to triangular shape, and a short flagellum and small microvilli at the apical pole (Figure 7).

In addition to these four epithelial cell types, three mesenchymal amoeboid and motile (Movies S2 and S4) cell types can be distinguished (Figure 7) in contrast to the two types previously mentioned in the adult (Type I vacuolar cells and Type II vacuolar/archaeocyte‐like cells) (Boury‐Esnault, Sole‐Cava, and Thorpe 1992; Ereskovsky et al. 2009; Fierro‐Constaín et al. 2017; Renard, Rocher, et al. 2021). First, Type I vacuolar cells are characterized by the presence of one to four large electron‐translucent vacuoles with few short filopodia, and the nucleus is localized laterally. Second, the Type II vacuolar/archaeocyte‐like cells are characterized by an asymmetric and amoeboid shape, by numerous vacuoles (empty or with fibrous inclusions, few osmiophilic inclusions), and the presence of several long actin‐filopodia, which often contact other cells. Finally, by using fluorescent imaging we were able to distinguish a third vacuolar cell type, which we propose to name “Type III vacuolar cell.” These cells are of a similar size to the two vacuolar cell types mentioned above and contain numerous vacuoles of different sizes like Type II vacuolar/archaeocyte‐like cells. However, in contrast to the two other vacuolar cell types, these cells display a clear polarity with one roundish smooth pole, and, at the opposite pole numerous (> 5) thin actin‐rich filopodia (Figure 7).

In addition, in the mesohyl we observed numerous anucleated cell‐like structures, not yet described in adults or Developmental Stage 1 buds, whose size ranges from 2 to 5 µm. We interpret them as being extracellular vesicles (EVs) (Figures 7 and 8E–I) (Raposo and Stahl 2019; Sheta et al. 2023). EVs have a variety of shapes, from spherical and ovoid to bean‐shaped with an obvious actin‐rich pole. These EVs seem to originate from cell budding at the basal pole of exopinacocytes (Figure 8F,G) (Clancy, Schmidtmann, and D'Souza‐Schorey 2021). Some of them seem to contain rare small DAPI–stained spots that may suggest the presence of bacteria inside them (Figure 12D), while other EVs are brightly stained by TUNEL assays (Figure 12E), suggesting that they may contain fragmented DNA.

Figure 12.

Staining of apoptotic cells in buds. (A) TUNEL staining (red) at Developmental Stage 1: Apoptotic DNA fragmentation is mainly observed in basopinacocytes (cyan arrow) and in cells localized in the mesohyl (white arrows). (B) TUNEL staining (red) at Developmental Stage 3 in mesohylar cells (white arrows), in bacteria, and extracellular vesicles (EVs). (C) The proportion of cells undergoing apoptosis is not significantly different between Developmental Stages 1 and 3 (values in Supporting Information S2: Table S2), each dot represents the proportion observed for one sample, and numbers indicate the total number of nuclei observed. (D) Positive TUNEL detection control using recombinant DNase treatment. (E) A part of the TUNEL staining observed in the mesohyl corresponds to extracellular vesicles (EVs, see Figure 4A and Supporting Information S1: S3E–I) shown by yellow arrows. Scale bars = 10 µm (A, B); 30 µm (D); 20 µm (E). Cc, choanocyte chamber; Exp, exopinacoderm; M, mesohyl. Max intensity projection of confocal Z‐stack (D, E).

In brief, by using complementary staining and microscopy methods and according to their positions and morphological features, we identified in buds (according to Boury‐Esnault and Rützler 1997) six epithelial cell types: choanocytes; endopinacocytes (including prosopinacocytes and apopinacocytes); exopinacocytes; basopinacocytes; and apopylar cells. Additionally, we identified three mesohylar cells (Type I, II, and III vacuolar cells) and EVs in the mesohyl.

3.3.3. Cell Dynamics During Bud Development and Growth

According to PHH3 immunolocalization and EdU incorporation (Figure 11A,B and Supporting Information S1: Figure S2), only 1% of cells divide at the same time (Figure 11C and Supporting Information S2: Tables S5 and S6). Division occurs more frequently in choanocytes, and to a lesser extent in mesohyl cells than in pinacocytes (Figure 11A–C, Supporting Information S1: Figure S2A,B, and Supporting Information S2: Tables S5 and S6). This may explain the slight increase in the number of choanocytes per chamber and the increase in size of choanocyte chambers from Developmental Stages 1 to 3 (Supporting Information S2: Table S4 and Supporting Information S1: Figure S1).

Figure 11.

Cell proliferation during bud development. (A) Cell proliferation (green: EdU) concerns mainly choanocytes (yellow arrows) and, in less extent, mesohylar cells (white arrows), at Developmental Stage 1 (Table S5). (B) Cell proliferation (magenta: PHH3) also mainly concerns choanocytes at stage 3. (C) Proportion of cells observed simultaneously in the division by PHH3 immunostaining at Developmental Stage 3 (Supporting Information S2: Table S6). Each dot represents the proportion observed for one sample, and numbers indicate the total number of nuclei observed. Cell division is rarely observed in exopinacocytes (Supporting Information S1: Figure S3) and has never been observed in endopinacocytes. The difference is significant (*) between choanocytes and endopinacocytes (Kruskall–Wallis: p = 0.0973; Dunn test: p _adj = 0.0197). Scale bars = 10 µm (A, B). Cc, choanocyte chamber; M, mesohyl.

The adaptation of the TUNEL staining protocol in O. lobularis (Borchiellini et al. 2021) also allowed us to observe that, during bud development, apoptosis seems to occur more frequently in mesohylar cells and basopinacocytes (Figure 12 and Supporting Information S2: Table S7). The observation of apoptosis activity in mesohylar cells has already been reported in other sponge species (Melnikov et al. 2022). The rates of apoptosis for tissue homeostasis and during morphogenetic processes seem different between sponge species (Baghdiguian et al. 2023; Melnikov et al. 2022). It would seem that the significance of apoptosis during bud development still remains to be evaluated.

3.4. Buds Are Physiologically Active Individuals

3.4.1. Buds Present Rhythmic Contractions

It has been previously shown that O. lobularis, like other sponges, contract their body rhythmically (Nickel 2010). Here, using time‐lapse experiments, we observed that, once free, buds also present rhythmic contractions that at first seem to be synchronized between buds and the adult they come from (Movie S1), and they continue to contract later, at Developmental Stages 2 and 3 in a non‐synchronized way (Movie S5). The amplitude of contraction/release is variable among clonal buds (Table S8). Figure 13A shows the evolution of one bud volume, based on the measurement of bud area over time. It illustrates a typical spontaneous 25% volume contraction event. Longer observations of Developmental Stage 1 buds suggest a rhythm of around 3.4 spontaneous contractions per day with a 14 min mean time of contraction.

Figure 13.

Analysis of bud motions. On the left, one image of a Developmental Stage 1 bud in a well of a 96‐well plate extracted from a 12‐h movie. Schematics illustrate each observed parameter of bud motion. On the right, a graph of the evolution of each parameter for one bud. (A) Observation of a contraction event. The analysis of the volume is based on the observed area of the bud during the recording. We plotted the evolution through time (on x‐axis) of relative volume (on y‐axis) that corresponds to the area of the bud at time t divided by the maximal observed area of the bud at the beginning of the video. (B) Bud translational motions. We plotted the motion of the bud center of mass in cartesian coordinates. The color code indicates the evolution in time starting from dark blue to yellow (see also the color scale on the right). (C) Bud rotational motions were plotted following the rotation of the internal cartesian coordinate of the bud (in dotted line) compared to the fixed coordinate used in (B). The coordinate in pixel represents the distance and orientation of the vector between the center of mass of the bud (the red dot) and a surface point of the bud identified in Deeplabcut, relative to the Cartesian coordinates of the 96‐well plate. The same color code is used as in (B) for the time scale.

3.4.2. Buds Are Mobile

Time‐lapse imaging focused on exopinacocyte cilia (in white light) clearly showed their beating (Movie S6). Time‐lapse experiments enabled us to show for the first time evidence that these cilia are involved in the mobility of buds. Indeed, we showed in vitro that buds are moving in the wells (from Developmental Stages 1–3). A protocol to track trajectories of Developmental Stage 1 buds was developed. It allows us to follow and quantify bud translational and rotational motions (Figure 13B,C), thereby offering future perspectives for light or chemotaxis experiments.

3.4.3. Buds Filter Water

To establish if buds are able to filter water from the first developmental stage despite the absence of a complete aquiferous system (i.e., devoid of osculum), we compared the entrance of fluorescent microspheres (size 0.2 µm to mimic bacteria size, as described in Borchiellini et al. 2021) in the aquiferous system between Developmental Stage 1 buds (devoid of osculum) and Developmental Stage 3 buds (possessing an osculum) after 15, 30, and 60 min of incubation (Figure 9). The localization of fluorescent microspheres in inhalant canals and choanocyte chamber cavities shows that Developmental Stage 1 buds are already capable of water filtering (Figure 9A). This process is fast. Indeed, after only 15 min of incubation, beads were present in inhalant canals and in the cavity of choanocyte chambers, indicating that filtration occurred. After 30 min, fluorescent microspheres were observed both at the level of the microvilli of choanocytes and in the cytoplasm of choanocytes (Figure 9A). This means they were actively internalized, likely by phagocytosis. After 30 min, in Developmental Stage 3 buds, some mesohylar cells contain beads (Figure 9B), suggesting that choanocytes that integrated microspheres have migrated into the mesohyl.

In contrast, when buds are exposed to nocodazole before incubation with beads, almost no fluorescent microspheres are found in the aquiferous system (whatever the bud developmental stages and incubation times) (Figures 9C and Supporting Information S1: Figure S3 for control). This observation shows that the filtering activity of buds (at all developmental stages) is an active process depending on the beating of flagella, as for adults.

3.4.4. Buds Respiration

Oxygen consumption is a fundamental indicator of the metabolic rate and represents the overall energy expenditure in organisms over a period of time (Darveau et al. 2002). We therefore decided to measure the oxygen consumption (MO₂) of buds and to compare it to that of adults. In NSW and in the dark, the mass‐specific oxygen consumption rate in buds was significantly (p < 0.05) higher (MO₂ = 0.139 ± 0.069 µgO₂/h/mg FW) than that measured in adults (MO₂ = 0.044 ± 0.009 µgO₂/h/mg FW; p = 0.002) (Figure 14). These results show that buds are not only metabolically active but also have an even higher energy expenditure than adults. This finding is in agreement with what is observed in most animals: Mass‐specific metabolic rate usually decreases throughout ontogeny, mainly as a consequence of the additional energetic cost of growth and development (Rosenfeld et al. 2015).

Figure 14.

Oxygen consumption. Mass‐specific oxygen consumption rate (MO₂) in Oscarella lobularis juvenile (buds) and adult specimens. A significant difference was found between bud and adult MO₂ (Mann–Whitney U‐test; p < 0.05).

3.5. Buds Are Able of Regenerative Processes In Vitro

3.5.1. Wound Healing and Regeneration

After bisection, all basal and apical halves of buds survived and completed wound healing within 24 h, with wound healing defined as the presence of a continuous epithelium (exopinacoderm) at the bud surface. Then, in less than 96 h after bisection (Figure 15), bud halves had regenerated the missing part, namely, an apical pole with a whole differentiated osculum or a basal pole with outgrowths. During this regenerative process, PhaE–stained cells were observed in the mesohyl (Figure 16), suggesting that choanocytes could be involved. Further experiments are required to evaluate the rate of migration of this and other cell types.

Figure 15.

Regenerative experiments. (A) Developmental Stage 3 bud with osculum at apical pole and outgrowths at basal pole. The dotted arrow represents the section level. (B) One half with primary osculum at apical pole and devoid of basal outgrowths and (C) one half without osculum but with basal outgrowths. After 96 h, both halves regenerate the missing part: new secondary osculum or new basal outgrowth. Scale bar = 500 µm (A–C).

Figure 16.

Choanocyte migration during the regeneration process. The presence of PhaE–stained cells in the mesohyl (green arrows) indicates a migration of choanocytes from mesohyl, after 72 h of regeneration. Scale bar = 10 µm. Max intensity projection of confocal Z‐stack. Cc, choanocyte chamber; M, mesohyl.

3.5.2. Reaggregation Capabilities of Dissociated Cells

The disorganization of bud choanocyte chambers has been shown to be a reversible calcium‐dependent process that can be triggered by a 1 h incubation in calcium magnesium free sea artificial water (CMFSW) (Vernale et al. 2021). Here, we obtained a complete cell dissociation of buds by a 1 h incubation in CMFSW supplemented with EDTA combined with stirring. This treatment resulted in a dense cell suspension (about 0.5 × 10⁶ cells/mL) (Figure 17A). The addition of calcium chloride triggered cell reaggregation. Twenty‐four hours after calcium addition, compact aggregates were obtained (Figure 17B). Cell aggregation was then followed by re‐epithelialization within 48 h after calcium addition (Figure 17C): The surface of the aggregates becomes smooth, and they can be considered as primmorphs (Custodio et al. 1998).

Figure 17.

Reaggregation of dissociated cells. (A) Buds are dissociated in CMFSW and EDTA to obtain a cell suspension. (B) Twenty‐four hours after the addition of calcium chloride, compact aggregates are observed. (C) After 48 h in ASW, primmorphs are obtained; scale bar = 50 µm.

4. Discussion and Conclusion

4.1. The Limits of Cell Type Definition

In the literature, according to light and electron microscopy observations, the number of cell types described in sponges ranges from six to at least 16 (Ereskovsky and Lavrov 2021; Simpson 1984). In O. lobularis, in both adult and Developmental Stage 1 buds, previous histological observations described seven cell types (Boury‐Esnault, Sole‐Cava, and Thorpe 1992; Ereskovsky and Tokina 2007; Fierro‐Constaín et al. 2017; Renard, Rocher, et al. 2021). In the present study, our additional observations using fluorescent confocal imaging on whole mount buds of O. lobularis, enabling Z‐stacking and 3D projections, allowed us to identify additional cell types (basopinacocytes and Type III vacuolar cells). This finding shows that the distinction between cell types is highly technique dependent, calling for a reevaluation of cell types in different sponge species using complementary tools.

In addition, because of terminology issues, the homology of cell types, between different sponge species, is unclear. Sponge cell biology therefore expects a lot of the single‐cell transcriptomic approaches, to enable the definition of cell types based on shared regulatory networks (Arendt et al. 2016). Such data were recently acquired for two demosponge species: A. queenslandica (Sebé‐Pedrós et al. 2018; Sogabe et al. 2019; Wong et al. 2019) and Spongilla lacustris (Musser et al. 2021). These preliminary results tell us more about cell functions in sponges and suggest that there are more cell types or cell subpopulations (cluster) than previously thought. Similar single‐cell transcriptomic approaches are currently in progress in O. lobularis.

4.2. Hypotheses Concerning Cell–Cell Communication in Sponges

Though the capability of sponges to react to external chemical or mechanical stimuli has been thoroughly documented, the way they perform signal transduction largely remains a mystery (Elliott and Leys 2007; Leys 2015; Leys and Meech 2006; Leys et al. 2019; Mah and Leys 2017; Maldonado 2006; Nickel 2010; Renard et al. 2009). A few sensory cells were identified in some demosponges, such as the light‐sensing pigmented cells of larvae (Collin et al. 2010; Leys et al. 2002; Leys and Degnan 2001; Maldonado et al. 2003) or mechanical‐sensory ciliated cells on adult osculum (Ludeman et al. 2014). Concerning signal transduction, action potentials were reported in Hexactinellida only (Leys, Mackie, and Meech 1999) and neurotransmitters were evidenced in demosponges (Burkhardt and Sprecher 2017; Elliott and Leys 2010; Ellwanger, Eich, and Nickel 2007; Ellwanger and Nickel 2006).

We report the existence in O. lobularis of very long filopodia that establish cell contacts between pinacoderm, mesohylar cells, and choanoderm. Similar actin‐rich filipodia were recently described in exopinacocytes and porocytes in calcareous sponge Leucosolenia variabilis (Skorentseva et al. 2023), in Oscarella pearsei (Miller et al. 2018), in the demosponges Hymeniacidon heliophile (Coutinho et al. 2017) and S. lacustris (Musser et al. 2021). These filopodia are likely a key element of cell–cell communication in sponges, and the significance of such a developed cell–cell interacting network may have been previously underestimated because of the common use of sections to study sponge anatomy.

In addition, we report the presence of EVs in the mesohyl of O. lobularis buds. Different types of vesicles have been observed in another homoscleromorph species (Ereskovsky et al. 2007), during the oogenesis of Chondrilla (Maldonado et al. 2005), for bacteria uptake or transmission (Carrier et al. 2022; Marulanda‐Gomez et al. 2023). The understanding of the diversity and significance of EVs in cell–cell communication in Eukaryotes including animals is recent (Mulcahy, Pink, and Carter 2014). Considering their numerous functions from cell–cell signaling to cell–symbiont interactions (Fu et al. 2023; Latifkar et al. 2019; Li, Liao, and Tian 2020; Liu and Wang 2023; Margolis and Sadovsky 2019; van Niel, D'angelo, and Raposo 2018; Raposo and Stahl 2019; Sheta et al. 2023), molecular studies are now needed to evaluate the types (apoptotic EVs or otherwise) and roles of these vesicles in sponges.

4.3. Significant Prospects for Morphogenetic Studies in Sponges

The capability of sponge cells to reaggregate after cell dissociation was reported decades ago (Custodio et al. 1998; Grice et al. 2017; Lavrov and Kosevich 2014, 2016; Le Pennec et al. 2003; Müller et al. 1999, 2004; Rady, Salem, and Ez El‐Arab 2019; Simpson 1984).

As far as we know, complete dissociation–reaggregation has not been evidenced so far in homoscleromorph species (Grice et al. 2017). This study is thereby the first report of successful total dissociation–reaggregation experiments in this class, in addition to the partial dissociation previously published (Vernale et al. 2021). This process offers a convenient experimental context to understand the molecular and cellular mechanisms involved in self/non‐self‐recognition (Grice et al. 2017) and in cell adhesion. Because Homoscleromorpha is the only sponge class with obvious basement membrane and adherens‐like junctions, this simple experimental setup provides the opportunity to study the dynamics of these key epithelial structures. Moreover, the resulting primmorph can be considered as a 3D cell‐culture system (Rady, Salem, and Ez El‐Arab 2019), which can be useful considering the difficulties in establishing sponge cell cultures (Conkling et al. 2019).

The second somatic morphogenetic process reported here is the regeneration of half‐cut buds. Most experiments performed on sponges (including the adults of O. lobularis; Ereskovsky et al. 2015; Fierro‐Constaín et al. 2017) are wound‐healing experiments instead of regenerative experiments (Ereskovsky et al. 2021). Indeed, the authors focus on the restoration of exopinacoderm integrity. Here, we went further by performing regenerative studies on buds by cutting off the osculum, a key functional structure of sponges. The reproducible and quick osculum regeneration observed in O. lobularis buds provides a simple and convenient model to study cellular and molecular mechanisms involved during sponge regeneration. This new biological system offers promising perspectives for comparative studies aimed at studying the origin and evolution of regenerative properties in animals (Lai and Aboobaker 2018).

During regenerative experiments, some choanocytes initially present in the choanoderm (epithelial layer) migrated to the mesohyl (mesenchymal layer). This simple experimental setup should thereby enable the study of epithelial–mesenchymal transition (EMT), transdifferentiation, and the putative multipotent stem properties of choanocytes (Ereskovsky 2010; Ereskovsky et al. 2015; Fierro‐Constaín et al. 2017).

4.4. O. lobularis Reaches the Top 3 of Sponge “Models”

The embryology, life cycle, development, regeneration, and histology of the homoscleromorph sponge O. lobularis are well described. Moreover, a chromosome‐level assembly of its genome is now publicly available (bioproject PRJEB57964) as well as an annotated transcriptome (Vernale et al. 2021). In this study, we managed to artificially induce in the lab a natural asexual reproduction event: budding, whih occurs several times a year in natural conditions. The buds obtained from O. lobularis are numerous and easy to maintain over several months in a laboratory in NSW or up to 1 week in ASW and can travel for up to 10 h in a cold box. These buds therefore offer the possibility to perform a wide range of assays on individuals whether genetically identical or not. Buds are very small, making incubation in small volumes possible (100 µL), which is convenient when expensive or toxic reagents are used (De Pao Mendonca et al. 2024). Moreover, we show here that Oscarella buds are physiologically active and can therefore be used for physiological assays.

The transparency of buds enabled us to develop various fluorescent staining methods on both fixed and live specimens (Borchiellini et al. 2021; Vernale et al. 2021; this study). This is a significant step forward in evaluating cell dynamics (division, migration, death) during different developmental or regenerative morphogenetic processes described here.

Altogether, this technical progress in O. lobularis buds provides new tools to better understand sponge cell biology and the cellular and molecular mechanisms involved in the Homoscleromorpha sponge class. The variety of resources and experimental tools available will establish O. lobularis buds as a biological model comparable to that of the juveniles of the demosponges A. queenslandica and Ephydatia muelleri (Table 1). Because of the diversity of sponges, the establishment of a new sponge “model” (as defined by Lanna 2015; Russell et al. 2017) pertaining to a distinct class is useful for evo‐devo comparative approaches, in particular, to study the evolution of epithelia and epithelial morphogenetic processes.

Table 1.

Data and tools available for the main sponge species used for evo‐devo studies (for a review Schenkelaars et al. 2019).

	Species	Environment	Development in the lab	Other morphogenetic experiments	Genome	Transcriptome	Single‐cell transcriptomic	Karyotype	In situ hybridization	Immunofluorescence	Cell type–specific staining	Live/death staining	TUNEL assays	PH3/EDU/BrdU proliferation assays	Pharmacological assays	Acid nucleic transfection	References
Homoscleromorpha	Oscarella lobularis	M	●	●	●	●	❍	❍	●	●	●	●	●	●	●	❍	Belahbib et al. (2018); Fierro‐Constaín et al. (2017); Gazave et al. (2008); Lapébie et al. (2009); Ereskovsky et al. (2015); This paper
Demospongiae	Amphimedon queenslandica	M	●	●	●	●	●		●	●	●		●	●	●❍		Adamska et al. (2007); Dunn, Leys, and Haddock (2015); Grice et al. (2017); Nakanishi, Sogabe, Nakanishi, and Degnan (2016); Sogabe et al. (2019); Sogabe et al. (2014, 2016, 2019); Ueda et al. (2016); Srivastava et al. (2010)
	Ephydatia muellieri	F	●		●	●		●	●	●	●				●	●	Adams et al. (2010); Elliott and Leys (2007); Imsiecke et al. (1995); Ishijima et al. (2008); Ludeman et al. (2014); Peña et al. (2016); kelaars et al. (2016); Schippers et al. (2018); Rivera et al. (2011, 2013); Windsor and Leys (2010); Windsor Reid et al. (2018); Hall et al., (2019)
	Ephydatia fluviatilis	F	●			●	●	●	●		●	●		●	●		Alié et al. (2015); Funayama et al. (2005); Ishijima et al., (2008); Nakayama et al., (2015)
	Spongilla lacustris	F	●		●	●	●	●	●			●				●	Adams et al., (2010); Kahn and Leys, (2016); Ishijima et al. (2008); Mah, Christensen‐Dalsgaard, and Leys (2014); Pfannkuchen and Brmmer (2009); Schill et al. (2006); Windsor Reid et al. (2018)
Calcarea	Leucosolenia complicata	M		●	●	●			●	●				●			Fortunato et al. (2015); Lavrov et al. (2018); Voigt et al. (2014)
	Sycon ciliatum	M	●	●	●	●			●								Fortunato et al. (2012); Dunn, Leys, and Haddock (2015); Leininger et al. (2014); Voigt et al. (2014)
Hexactinellida	Aphrocallistes vastus	M				●			●					●			Grant et al. (2018); kahn and Leys (2016); Windsor Reid et al. (2018)

Note: Filled circles mean acquired, empty circles mean in the process of being acquired.

Conflicts of Interest

The authors declare no conflicts of interest.

Ethics Statement

The authors have nothing to report.

Consent

The authors have nothing to report.

Supporting information

Supporting information.

Supporting information.

Supporting information.

Supporting information.

Supporting information.

Supporting information.

Supporting information.

Supporting information.

Supporting information.

Supporting information.

Data Availability Statement

All data supporting the conclusions of this article are included within the article and its additional file(s).