INTRODUCTION
Living organisms are extremely complex systems, the biological components of which interact with each other and with the environment. Their complexity can be a great obstacle to the investigation of animal physiology and biology, particularly in those organisms that are still poorly known, such as certain invertebrates. In particular, studies on basic physiological processes, such as immune responses or hormonal control of reproduction, which often employ experiments involving exogenous administration, can be particularly demanding due to metabolic masking effects. Homeostatic and metabolic mechanisms can prevent the exogenous molecules from exerting their biological functions as they are designed to lower/increase their level through detoxification pathways or excess accumulation of inactive forms. This often results in general uncertainty concerning the data obtained, because it is unknown whether an experimental design has the required features to act effectively in the whole organism and on the target cell/tissues (Sugni et al., Reference Sugni, Motta, Tremolada and Candia Carnevali2012; Mercurio et al., Reference Mercurio, Tremolada, Nobile, Fernandes, Porte and Sugni2015). On the contrary, the in vitro approach allows the simplification of the biological system under investigation, focusing on only a few components in a controlled condition without interference. However, in vitro studies provide a simplified description of the biological problem, which needs to be integrated with in vivo experiments. For these reasons, in the present work we investigated sex-steroid (specifically 17β-oestradiol (E2) and testosterone (T)) involvement in sea urchin reproduction employing both in vivo and in vitro techniques. The role of steroid hormones in echinoid reproductive biology is incompletely understood. This represents a serious deficiency in our knowledge of invertebrate endocrinology in general and has particular ecotoxicological implications. Shedding light on invertebrate (and echinoid in particular) endocrinology could reveal whether these animals represent potential targets of endocrine disruptor (ED) contaminants, a main recent reason of concern in the scientific and public community. Lack of this information makes it impossible to produce a mechanistic interpretation of any ED ecotoxicological study addressed to these ecologically important animals.
The involvement of sex-steroid hormones in echinoid reproduction has been suggested by studies on seasonal changes of steroid levels during the gonadal cycle (Wasson et al., Reference Wasson, Gower, Hines and Watts2000a; Barbaglio et al., Reference Barbaglio, Sugni, Di Benedetto, Bonasoro, Schnell, Lavado, Porte and Candia Carnevali2007). Nevertheless, several hormonal administration experiments have been performed with different and sometimes contrasting results (Unuma et al., Reference Unuma, Yamamoto and Akiyama1999; Wasson et al., Reference Wasson, Gower and Watts2000b; Varaksina & Varaksin, Reference Varaksina and Varaksin2001, Reference Varaksina and Varaksin2002; Sugni et al., Reference Sugni, Motta, Tremolada and Candia Carnevali2012; Mercurio et al., Reference Mercurio, Tremolada, Nobile, Fernandes, Porte and Sugni2015). In juveniles of Pseudocentrotus depressus, oestrone administration induced testis growth and promoted spermatogenesis, whereas E2 treatment did not result in any significant effects (Unuma et al., Reference Unuma, Yamamoto and Akiyama1999). Similarly, in our previous research on the Mediterranean sea urchin Paracentrotus lividus exogenous administration (injection) of E2 influenced significantly neither the reproductive stage of the gonads nor the development of the gametes (Sugni et al., Reference Sugni, Motta, Tremolada and Candia Carnevali2012; Mercurio et al., Reference Mercurio, Tremolada, Nobile, Fernandes, Porte and Sugni2015). On the contrary, in Lytechinus variegatus, E2 and T dietary administration resulted in appreciable effects on some reproductive parameters. In addition, both hormones seemed to increase protein concentration in the gonads, suggesting a steroid influence in protein accumulation (Wasson et al., Reference Wasson, Gower and Watts2000b). Comparable results were reported in Strongylocentrotus nudus (Varaksina & Varaksin, Reference Varaksina and Varaksin2001) and in S. intermedius (Varaksina & Varaksin, Reference Varaksina and Varaksin2002), leading to the hypothesis that steroids control gonad protein expression. This idea has received support from studies of sea urchin Major Yolk Protein (MYP) (Shyu et al., Reference Shyu, Blumenthal and Raff1987; Prowse & Byrne, Reference Prowse and Byrne2012). MYP is a glycoprotein present in almost all echinoid tissues and represents the most abundant protein in sea urchin eggs (Shyu et al., Reference Shyu, Raff and Blumenthal1986; Unuma et al., Reference Unuma, Suzuki, Kurokawa, Yamamoto and Akiyama1998, Reference Unuma, Okamoto, Konishi, Ohta and Mori2001, Reference Unuma, Yamamoto, Akiyama, Shiraishi and Ohta2003). The sequencing of MYP cDNA has revealed the presence of a putative oestrogen responsive element (ERE) upstream of the gene, suggesting an oestrogen involvement in the regulation of its expression (Shyu et al., Reference Shyu, Blumenthal and Raff1987) and, therefore, in sea urchin vitellogenesis processes, as previously reported in asteroids (Schoenmakers et al., Reference Schoenmakers, Van Bohemen and Dieleman1981; Takahashi & Kanatani, Reference Takahashi and Kanatani1981).
However, the mode of action of steroids in echinoderms is still unknown. There is evidence for receptor-mediated signal transduction (Waal et al., Reference Waal, Portman and Voogt1982; Roepke et al., Reference Roepke, Snyder and Cherr2005, Reference Roepke, Chang and Cherr2006; Köhler et al., Reference Köhler, Kloas, Schirling, Lutz, Reye, Langen, Triebskorn, Nagel and Schönfelder2007) but no definitive evidence was obtained on the existence of classical steroid receptors in echinoids. In particular, no typical vertebrate Oestrogen Receptor (ER) and Androgen Receptor (AR) has been found in the sea urchin genome, where only an Oestrogen Receptor-related Receptor (ERR) is present (Goldstone et al., Reference Goldstone, Hamdoun, Cole, Howard-Ashby, Nebert, Scally, Dean, Epel, Hahn and Stegeman2006) and three orphan members of the steroid nuclear receptor superfamily have been characterized (Kontrogianni-Konstantopoulos et al., Reference Kontrogianni-Konstantopoulos, Vlahou, Vu and Flytzanis1996, Reference Kontrogianni-Konstantopoulos, Leahy and Flytzanis1998; Kontrogianni-Konstantopoulos & Flytzanis, Reference Kontrogianni-Konstantopoulos and Flytzanis2001).
Despite all the above reported investigations, the relationship between sex-steroids and echinoid reproduction is still far from being clearly determined. The highly variable results present in literature may be due in part to the different experimental designs: studies have been conducted on several species and during different reproductive and/or life stages, using different administration types for variable periods (Mercurio et al., Reference Mercurio, Tremolada, Nobile, Fernandes, Porte and Sugni2015). Additionally, such different and even contrasting results could be the consequence of the activation of homeostatic mechanisms. In our previous research on P. lividus, E2 administration induced a marked increase of hormone in the coelomic fluids whereas its level in the gonad appeared to be less affected (Mercurio et al., Reference Mercurio, Tremolada, Nobile, Fernandes, Porte and Sugni2015). These results suggested the presence of homeostatic/detoxification mechanisms involved in E2 elimination, which are particularly active in the gonads, where they could prevent the active hormone from exerting its biological functions and therefore hide their possible effects. In vertebrates, esterification and sulphation are well-known processes that remove excess of bio-available hormones (Strott, Reference Strott1996; Hochberg, Reference Hochberg1998). The ability to produce esterified and sulphated steroids is well documented in echinoids, where these processes have been recognized as belonging to major bio-transformation pathways for sex hormones (Creange & Szego, Reference Creange and Szego1967; Voogt & Van Rheenen, Reference Voogt and Van Rheenen1986; Hines et al., Reference Hines, Watts, McClintock, David, Féral and Roux1994; Janer et al., Reference Janer, LeBlanc and Porte2005a, Reference Janer, Sternberg, LeBlanc and Porteb; Lavado et al., Reference Lavado, Sugni, Candia Carnevali and Porte2006).
In the present study, we investigated T and E2 involvement in the reproductive biology of the common Mediterranean sea urchin Paracentrotus lividus, employing parallel in vivo and in vitro approaches. First, a long-term experiment involving T and E2 dietary administration in adult specimens was performed and different reproductive parameters were analysed. Second, ovarian cells, isolated from ovaries at the same reproductive stages considered in the in vivo study, were cultured and directly exposed to the same steroids. The results, obtained from the two different approaches, were analysed and compared.
MATERIALS AND METHODS
Experimental animals and maintenance
One hundred (84 for the in vivo and 16 for the in vitro experiments) P. lividus adult specimens were collected in the Protected Marine Area of Bergeggi (44°14′N 8°26′E), on the Ligurian coast of Italy (Tyrrhenian Sea), at the end of July and immediately transferred to the laboratory. Animals were randomly distributed in aquaria filled with artificial seawater (Instant Ocean; salinity about 37‰, as in the Mediterranean Sea) and provided with a circulation system as well as mechanical, chemical and biological filters. Animal conditions as well as all physical and chemical parameters were monitored closely throughout the experimental period. Animals were starved for 6 weeks to reset the reproductive cycle to a resting phase and synchronize animals at the same starting reproductive condition (Spirlet et al., Reference Spirlet, Grosjean and Jangoux2000). Throughout the starvation period temperature was set at 16 ± 1°C and photoperiod was fixed at 16:8 h (dark:light), simulating winter conditions. At the beginning of the following feeding period these parameters were gradually increased up to 20°C and 10:14 h (dark:light) respectively, as in summer time, and remained fixed until the end of the experiments.
In vivo experimental design
A long-term experiment involving E2 and T dietary administration was performed. After starvation, all the animals were fed with control pellets for 1 week before the hormonal treatment. To evaluate their starting reproductive conditions and to confirm the synchrony within the experimental population, 10 specimens were killed just before the beginning of the dietary administration (T0): animals were weighed and their five gonads were removed for Gonad Index (GI) calculation. One gonad was processed for standard methods of light microscopy. All the remaining animals were distributed in specific aquaria each one internally subdivided into four compartments, each containing only two individuals, in order to reduce competition and help to control daily feeding rates. Four experimental groups were set up in duplicates (8 + 8 specimens for each group): a CTL group, fed with control pellets; an E2 group, fed with pellets containing 17β-oestradiol; an E2–4 weeks group, fed with control pellets for the first 4 weeks and, then, with pellets containing 17β-oestradiol until the end of the experimental period; a T group, fed with pellets containing testosterone. After 4 weeks of dietary administration, 10 control animals were killed (T01) and processed as described above for T0 specimens. T01 animals were necessary to confirm the onset of the reproductive processes had occurred and determine the reproductive conditions in which E2–4 weeks group was at the beginning of the E2 treatment. At the end of the steroid administration period (8 weeks; T1) all the experimental animals were weighed, killed and their gonads were processed for GI calculation and histological analysis.
HORMONAL DIETARY ADMINISTRATION
Control pellets were prepared as described in Wasson et al. (Reference Wasson, Gower and Watts2000b) with some modifications. Briefly, a 30:70% mixture of pulverized sea urchin formulated feed (Wenger Manufacturing, Inc.) and boiled distilled water containing 3% agar was prepared. The mixture was allowed to solidify at room temperature in specific moulds in order to obtain pellets of about 0.2 g. Pellets containing E2 and T were prepared following the same described procedure. E2/T (Sigma) powder was firstly accurately mixed with the pulverized sea urchin feed and, then, distilled water with agar was added only when its temperature decreased to 40°C, in order to avoid steroid thermal degradation. Fresh pellets were prepared every week and stored at −20°C. Each animal was fed daily with a 0.2 g pellet for 8 weeks. Each T pellet contained 10 µg of testosterone whereas each E2 pellet contained 1 µg of 17β-oestradiol. Steroid administered doses were carefully chosen taking into account the physiological T and E2 concentrations in P. lividus gonads, the relevant metabolic activity present in the digestive tube (Lavado et al., Reference Lavado, Sugni, Candia Carnevali and Porte2006; Barbaglio et al., Reference Barbaglio, Sugni, Di Benedetto, Bonasoro, Schnell, Lavado, Porte and Candia Carnevali2007) and previous similar work (Unuma et al., Reference Unuma, Yamamoto and Akiyama1999; Wasson et al., Reference Wasson, Gower and Watts2000b).
In vitro experimental design
The experimental design employed for the in vivo study was adapted for the in vitro experiments. At the end of the starvation period, eight animals were killed and primary cell cultures were developed from gonads of only female individuals (C0). Ovaries were removed from the test and one gonad was processed for standard methods of light microscopy, whereas the remaining four gonads were used to obtain primary cell cultures and test steroid exposure. An additional eight animals were arranged in specific aquaria, as described above for the in vivo experiment, and fed daily with control pellets. After 4 weeks of feeding (C01) specimens were killed and ovaries were used for cell culture development and histological analysis.
CELL CULTURES
Primary cell cultures from P. lividus ovaries were developed, as previously described (Mercurio et al., Reference Mercurio, Di Benedetto, Sugni and Candia Carnevali2014) with some modifications. Briefly, ovaries were washed in sterile Ca2+ Mg2+ Free Sea Water (CMFSW) with antibiotics (40 µg L−1 gentamycin, 100 units mL−1 penicillin and 100 µg mL−1 streptomycin). Ovary pieces were incubated in 0.5 mg mL−1 collagenase dissolved in sterile CMFSW and stirred for 1 h. The resulting cell suspension was centrifuged at 300 × g for 6 min at 15°C and the cell pellet was resuspended in modified Leibovitz-15 medium without supplement addition. Ovarian cells, isolated from each female, were divided into three different experimental groups: CTL group, cells maintained in culture medium without hormones, E2 group, ovarian cells cultured in presence of 20 pg mL−1 17β-oestradiol, and T group, cells cultured in presence of 200 pg mL−1 testosterone.
In order to analyse the E2 and T effects on cell morphology and viability, cells were seeded at a concentration of 3–4 × 105 cells mL−1 in 24-well culture plates, coated with poly-L-lysine (70–150 kDa, 0.01% solution, Sigma). Cultures were developed in triplicates and maintained at 15°C for 2 weeks. Replacement of 50% of the medium was carried out every 2 days and cell behaviour and appearance were observed daily using an inverted phase contrast microscope. At the end of the exposure period, cell viability was determined by staining with fluorescent Calcein AM (viable cell marker, Sigma).
For the biochemical analyses, further cell cultures from the same specimens were developed in triplicates and exposed to the hormones for 24 and 48 h. Cells were seeded at a concentration of 3 × 106 cells mL−1 in 6-well culture plates, coated with poly-L-lysine, and maintained at 15°C.
CHEMICALS AND SOLUTION PREPARATION
All chemicals were of reagent grade. 17β-oestradiol and testosterone were purchased from Sigma. Cell culture medium containing E2 was prepared as follows: 2 mg E2 were dissolved in 10 mL acetone (Merck), then 10 µL of this solution was dissolved in 10 mL autoclaved artificial sea water. Ten µL of this latter solution was mixed with 10 mL of modified Leibovitz-15 medium obtaining a stock solution, which was subdivided in aliquots and maintained at −20°C in the dark, in order to prevent steroid degradation. Finally, to reach the final concentration of 20 pg mL−1, 500 µL of the stock solution was dissolved in 4.5 mL of medium. Final acetone concentration was considered negligible (0.00001%). To prepare the culture medium containing 200 pg mL−1 T, the same procedure described for E2 was followed using an initial T amount of 20 mg. Fresh culture medium with hormones was prepared every time. Stock solutions were prepared every week.
Steroid exposure concentrations were defined based on physiological E2 and T levels, previously measured in gonads (Barbaglio et al., Reference Barbaglio, Sugni, Di Benedetto, Bonasoro, Schnell, Lavado, Porte and Candia Carnevali2007) and coelomic fluids of P. lividus (Mercurio et al., Reference Mercurio, Tremolada, Nobile, Fernandes, Porte and Sugni2015).
Determination of reproductive stages
Reproductive stages were determined by histological analysis. Standard methods for light microscopy (paraffin and/or resin) were employed.
Gonads were fixed with Bouin's solution (picric acid, formaldehyde, acetic acid, 75:25:5) for at least 24 h and washed several times in tap water until all the fixative solution was completely removed. Samples were then dehydrated with an ethanol series (70, 90, 95 and 100%), cleared in xylene and left overnight in a solution of xylene and paraffin wax 56–58°C (1:1). Gonads were then immersed in three changes of paraffin wax and finally embedded. Longitudinal sections (5–7 µm) were cut with a Reichert OmE sledge microtome and stained with Milligan's Trichrome. Briefly, before staining, the sections were immersed in xylene then ethanol (100 and 95%) and placed into a solution of potassium dichromate and hydrochloric acid. They were first stained with acid fuchsin, fixed with phosphomolybdic acid (1%), stained with orange G and fast green FCF, and then treated with a solution of acetic acid 1%. Finally, sections were cleared in xylene and dehydrated with ethanol (95 and 100%) before mounting.
Both T0 and C0 samples were embedded in resin and processed for semi-thin sections (which provide better quality and higher resolution) because difficulties in determining animal sex were expected after the starvation period. Briefly, gonads were prefixed with 2% glutaraldehyde in 0.1 M cacodylate buffer and NaCl 1.4% for 2 h and, after overnight washing in the same buffer, postfixed with 1% solution of OsO4 in 0.1 M cacodylate buffer (2 h). After standard dehydration in an ethanol series (25, 70, 90 and 100%), the samples were washed in propylene oxide, left in propylene oxide:resin (1:1) for 2 h. After overnight washing in resin they were embedded in Epon-Araldite 812 resin. The semi-thin (about 1 µm) sections were cut with a Reichert-Jung ULTRACUT E using glass knives and stained with crystal violet and basic fuchsin.
Both resin and paraffin sections were observed and photographed under a Jenaval light microscope to determine the gonad reproductive stage. The following five reproductive stages were determined on the basis of specimen maturation level: Spent (immediately after the spawning event), Recovery (phagocytosis and nutrient accumulation phase), Growing, Premature and Mature (progressive stages of gametogenesis) (Barbaglio et al., Reference Barbaglio, Sugni, Di Benedetto, Bonasoro, Schnell, Lavado, Porte and Candia Carnevali2007; Sugni et al., Reference Sugni, Motta, Tremolada and Candia Carnevali2012).
Morphometric analyses
For each histological sample we analysed gonad acinus area occupied by (1) nutritive phagocytes, (2) developing and mature gametes and (3) germ cells. Briefly, for each sample we selected three histological longitudinal sections from the gonad central layer and we measured the acinal area of the three different cell types using ImageJ program. Percentage of the area occupied by phagocytes/gametes were obtained for each acinus.
Maturity Index and Gonad Index
Maturity Index (MI) was determined based on the results obtained from microscopical analysis. It was calculated as the mean value of the numerical reproductive stage of each experimental group based on histological analysis (Spent = 0, Recovery = 1, Growing = 2, Premature = 3, Mature = 4). The numerical reproductive stage of the animal was corrected with +0.25 or −0.25 when in advanced or precocious conditions, respectively. This allowed us to more accurately indicate the actual maturative state.
Gonad Index (GI; Spirlet et al., Reference Spirlet, Grosjean and Jangoux2000) was determined as a percentage of the ratio between gonads wet weight (GW) and total wet weight (TW): GI = (GW/TW) × 100.
Biochemical analyses
The ovarian cells, cultured in 6-well culture plates, were processed for electrophoresis analysis after 24 and 48 h of in vitro steroid exposure. To prepare samples for running on gel, cells were gently scraped off the dish, using an ice-cold plastic cell scraper. The cell suspension was transferred into a pre-cooled tube and centrifuged at 300 × g for 6 min at 4°C; the resulting cell pellet was resuspended in 100 µL of lysis buffer. Cells were lysed in ice-cold 20 mM Tris-HCl (pH 7.5) with protease inhibitors (5 mM EDTA, 0.2 mM PMSF and 4 mM NEM), maintaining them in constant agitation for 30 min. Then, samples were centrifuged at 12,000 × g for 20 min at 4°C and the supernatant was transferred into a fresh tube kept on ice. Finally, samples were dialysed, protein assayed (BCA protein assay kit, Sigma) and stored at −20°C until electrophoresis was performed.
Sodium dodecylsulphate-polyacrylamide gel electrophoresis (SDS-PAGE) was performed using 6% slab gel (Laemmli, Reference Laemmli1970) in order to verify and compare the sample protein content in C0 and C1 experimental groups (CTL, E2 and T). Samples containing 10 µg of total proteins were diluted with sample buffer (SDS reducing buffer), boiled for 5 min and then applied to each lane. Gels were run at 100 V at room temperature. Protein bands were visualized with Coomassie brilliant blue R-250. SDS-PAGE standards (StoS Protein Marker, Genespin s.r.l.) were also run for molecular weight calibration. Each sample was analysed twice. The described protocol was initially optimized by performing SDS-PAGE analyses on P. lividus egg extracts (which have a rich MYP content), obtained with the same extraction procedure.
Statistical analysis
Results are presented as mean values ± SEM. Statistical significance was assessed using one-way ANOVA (Tukey's post hoc test). Statistical analysis was performed by the computer program GraphPad Prism 4. To compare the acinal area occupied by the gonad cell types in the different experimental groups we employed generalized linear mixed models (GLMs). Phagocytes/gametes area was used as dependent variable and the treatment (CTL, E2, E2–4w, T) was used as dependent variable. Gonad from which each set of data was determined was added as random factor. To assess the significance of the treatment we used a likelihood ratio test. Analyses were performed in R environment using the nlme package.
A P-value of less than 0.05 was considered statistically significant.
RESULTS
In vivo experiments
ANIMAL HEALTH CONDITIONS
During both the in vivo and in vitro experimental periods, physical and chemical parameters of the aquaria were frequently monitored and, when necessary, promptly adjusted. All the animals appeared healthy and none died throughout the experiments. As far as the experiment involving steroid dietary administration is concerned, the animal daily feeding rate, calculated as the mean percentage of the eaten pellets, was close to 100% for all the experimental groups (CTL: 99.4%; E2: 99.6%; T: 99.7%; E2–4 weeks: 100%), ensuring that all the groups consumed the same quantity of food and that daily E2 and T administration occurred properly. All the control and hormonally treated groups displayed a similar sex ratio, close to 1:1.5 (males:females).
GONAD INDEX (GI)
Mean GI values (Figure 1) measured in T0 and T01 animals were significantly lower than that measured in the T1 CTL group (ANOVA, F(3,42) = 25.07, P < 0.0001, Tukey's test: P < 0.001). A slight and progressive increase of GI was observed from T0 (3.8 ± 0.5) to T01 (5.2 ± 0.4) (P > 0.05). The highest mean GI was registered in T1 CTL (8.5 ± 0.5).
Gonad Index of untreated groups (T0, T01, T1 CTL) during the experimental period. Data are expressed as mean ± SEM. ** = P < 0.001 (N = 10–16).
Considering T1 animals, no significant difference in GI was recorded between CTL and hormonally treated groups (ANOVA, F(3,60) = 0.46, P > 0.05): mean GI values were almost the same in the different experimental groups, varying between 8.5 ± 0.5 and 9.6 ± 0.7. No statistically significant differences were found between males alone (ANOVA, F(3,24) = 1.07, P > 0.05) or between females alone (ANOVA, F(3,31) = 0.28, P > 0.05) (Figure 2). In females, mean GI values oscillated between 8.9 ± 0.9, registered in CTL specimens, and 9.9 ± 1.2, measured in E2 group. In males, GI values were less uniform, varying between 7.2 ± 0.7, recorded in E2 group, and 9.6 ± 0.8 found in T samples.
Mean Gonad Index values in females and males of T1 experimental groups. Data are expressed as mean ± SEM (N = 5–11).
MATURITY INDEX (MI) AND REPRODUCTIVE STAGES
Mean MI values of T0 and T01 animals were significantly lower than those measured at the end of the experimental period in T1 CTL (ANOVA, F(2,33) = 8.78, P = 0.007, Tukey's test: T0 vs CTLT1: P < 0.001; T01 vs T1: P < 0.05). MI gradually increased from T0 (1.4 ± 0.2) to T01 groups (2.3 ± 0.1) and, finally, reached the highest mean value in T1 samples (2.8 ± 0.3) (Figure 3).
Maturity Index of T0, T01, T1 CTL groups. Data are expressed as mean ± SEM. ** = P < 0.001 (N = 10–16).
Considering MI values of T1 experimental sets, no significant difference was recorded between CTL and hormonally treated groups (ANOVA, F(3,60) = 0.49, P > 0.05). A similar result was found when considering males alone (ANOVA, F(3,25) = 0.01, P > 0.05) and females alone (ANOVA, F(3,30) = 2.26, P > 0.05) (Figure 4).
Mean Maturity Index values in females and males of T1 experimental groups. Data are expressed as mean ± SEM (N = 5–11).
Further analysis of the distribution of the reproductive stage frequencies (Figure 5) showed that at the end of the starvation period, almost all the samples were in Recovery stage, regardless of sex. After 4 weeks of feeding, in T01 specimens, the reproductive processes had started: all females were in Growing stage whereas males presented a high percentage of samples in Premature stage (67%).
Distribution of reproductive stages of T1 experimental groups in females and males. Number on the top of the column indicates the number of female/male samples for each experimental group. Number of samples for each reproductive stage ranges from 1 to 6.
Focusing on T1 females, control animals were all found in Premature (60%) and Mature (40%) stages whereas in treated groups a low percentage of samples still in Growing stage was observed (E2: 30%; T: 45% and E2–4 weeks: 27%). Resting non-gametogenic stages (Spent and Recovery stages) were occasionally found in E2–4 weeks and T groups. In males, a higher variability could be noticed. Both resting and active gametogenic stages were present in each experimental group and no evident difference was observed between control and hormonally treated groups. Overall, no significant difference was observed between control and hormonally treated groups in the relative frequencies of the reproductive stages.
GONAD STRUCTURES
Analysing the relative areas occupied by nutritive phagocytes, developing and mature gametes and germ cells we did not find any effects of the treatment (phagocytes: GLMs; F = 0.79, P = 0.5; gametes area: GLMs; F = 0.02, P = 0.99). The medians as well as the variation range observed in each experimental group were comparable (Figure 6).
Boxplot of the acinal areas occupied by phagocytes (A) and mature and developing gametes (B) in T1 experimental groups.
In vitro experiments
C0 and C01 cell cultures were obtained from ovaries at the same reproductive conditions displayed by T0 and T01 animals (in vivo experiment). In particular, the histological analysis of ovaries confirmed that all C0 individuals were in Recovery stage whereas C01 animals were in Growing stage, similarly to T0 and T01 animals respectively (data not shown).
In vitro analyses were mainly focused on morphological aspects. Considering C0 cell cultures (Figure 7), no marked difference was found between control and steroid exposed cells. E2 and T did not affect cell morphology and behaviour during the considered culture period. In particular, no difference in size of either nutritive phagocytes or oocytes was observed between the experimental groups. Similar results were obtained in C01 cell cultures. Ovary cells, cultured in presence of sex-steroids, appeared comparable with the controls in terms of both size and behaviour.
Phase contrast microscopy. C0 ovarian heterogeneous cells cultured with: no steroid hormone (A, D, G), 20 pg mL−1 E2 (B, E, H) and 200 pg mL−1 (C, F, I), during the experimental period. 1d = 1 day after cell isolation (A, B, C); 1w = 1 week cell cultures (D, E, F); 2w = 2 week cell cultures (G, H, I). O = growing oocytes at different level of maturation; P = nutritive phagocytes. Scale bars: 20 µm.
Cell viability gradually decreased in the experimental groups as no serum was added to the culture medium. At the end of the exposure period (2 weeks) no striking difference between treated and control cell cultures were observed in cell viability (Figure 8) for both C0 and C01 experimental groups.
Fluorescence microscopy. Calcein AM method: viable cells appear fluorescent. C0 primary cell cultures from P. lividus ovaries after 2 weeks with no steroid (A) and 20 pg mL−1 E2 (B).
MYP CONTENT
MYP content in C0 and C01 cell cultures was investigated after 24 and 48 h of T/E2 exposure, performing SDS-PAGE analysis (Figure 9).
6% SDS-PAGE analysis of C0 and C01 cell extracts.
Firstly, a 170 kDa protein was present in all the samples. Considering the molecular weights and current knowledge (Unuma et al., Reference Unuma, Sawaguchi, Yamano and Ohta2011; Prowse and Byrne, Reference Prowse and Byrne2012) we concluded that the band corresponded to EGMYP. In addition, two other protein bands with slightly different molecular weights were found. The 180 kDa protein band (Figure 8) had the expected molecular weight of CFMYP (Unuma et al., Reference Unuma, Sawaguchi, Yamano and Ohta2011).
Comparing C0 experimental groups, no difference in EGMYP content was observed in cells exposed to the hormones for either 24 or 48 h. Steroids seemed also not to affect EGMYP content in C01 ovary cells: in E2 and T samples the intensity of EGMYP bands was comparable to the controls.
DISCUSSION
In this study, E2 and T involvement in echinoid reproduction was explored by applying both in vivo and in vitro approaches.
First, a long-term experiment involving steroid dietary administration was performed in adult specimens of P. lividus and different reproductive parameters were investigated. The experiment was designed in order to synchronize the initial reproductive conditions of the animals and reduce as much as possible the variability present in field populations. The nutritional conditions of the experimental animals were optimal, as a progressive increase of GI mean values was observed from starved animals (T0) to sea urchins well-fed for 4 (T01) and 8 weeks (T1 CTL). Indeed, these changes represent the classic response to feeding, as GI values appear to be strongly influenced by feeding rates (Mercurio et al., Reference Mercurio, Tremolada, Nobile, Fernandes, Porte and Sugni2015). MI values (determined from histological analysis) of control groups (T0, T01 and T1 CTL) displayed a similar trend: after starvation gametogenesis processes restarted and sea urchins progressively reached higher reproductive stages. These data underlined the success of the experimental design. Experimental animals were divided into four groups each consisting of 16 animals: CTL, E2, T and E2–4 weeks. In particular, the latter was chosen in order to verify the existence of a specific reproductive stage sensitive to E2, as proposed for different asteroid species. In both Asterias rubens and Asterina pectnifera E2 seemed to promote oocyte growth only during vitellogenesis, suggesting the presence of a threshold oocyte size for E2 effectiveness (Schoenmakers et al., Reference Schoenmakers, Van Bohemen and Dieleman1981; Takahashi & Kanatani, Reference Takahashi and Kanatani1981). To investigate the existence of similar mechanisms in echinoids, E2–4 weeks specimens were treated with E2, only after they reached the Growing stage, during which gametogenesis and vitellogenesis processes start.
Considering the analysed reproductive parameters, E2 and T administration did not affect either gonad growth, expressed as GI, and maturation or acinal cell composition. Microscopical observations did not reveal marked differences in gonad structure and/or gamete amount and all the experimental groups displayed a similar distribution of the reproductive stages at the end of the experiment (T1). These results are comparable to those obtained in our previous works (Sugni et al., Reference Sugni, Motta, Tremolada and Candia Carnevali2012; Mercurio et al., Reference Mercurio, Tremolada, Nobile, Fernandes, Porte and Sugni2015) on E2 but partially disagree with findings in other echinoid species (Wasson et al., Reference Wasson, Gower and Watts2000b; Varaksina & Varaksin, Reference Varaksina and Varaksin2001, Reference Varaksina and Varaksin2002). In L. variegatus E2 dietary administration for 36 days resulted in an inhibited growth of individual oocytes whereas T was found to promote oocyte growth (Wasson et al., Reference Wasson, Gower and Watts2000b). In S. intermedius and S. nudus the administration of oestradiol dipropionate stimulated gonad development and gamete maturation (Varaksina & Varaksin, Reference Varaksina and Varaksin2001, Reference Varaksina and Varaksin2002). No response to E2 and T was instead observed in P. depressus yearling females (Unuma et al., Reference Unuma, Yamamoto and Akiyama1999), in agreement with our results. This heterogeneity of results could be due to species-specific hormonal mechanisms as well as to different experimental conditions. Treatment length, steroid concentrations, administration type and specimen conditions could affect the results. In particular, the asynchronous starting conditions of the animals can strongly affect the results (Sugni et al., Reference Sugni, Motta, Tremolada and Candia Carnevali2012) and has to be considered. In addition, results obtained from long-term experiments could be influenced by metabolic masking effects that can cover the real effects by preventing the exogenous hormones from reaching their target organs/tissues (Mercurio et al., Reference Mercurio, Tremolada, Nobile, Fernandes, Porte and Sugni2015). Thus, to confirm these findings and explore more deeply the possible involvement of E2 and T in echinoid reproduction, we also performed in vitro exposure experiments. Although the in vitro conditions were obviously very different from those of the in vivo model, cell cultures allow the use of controlled experimental conditions without most of the metabolic interference that can occur in the whole animal, ensuring that the hormone reach the putative target cells. Primary cell cultures were developed from ovaries at two different starting reproductive conditions (C0: Recovery stage; C01: Growing stages) in order to make the results comparable with those of the in vivo experiment. Analysing cell morphology and behaviour, no variation was observed between controls and hormonally exposed groups. In both C0 and C01 cultures oocyte size and viability were similar, suggesting that T and E2 are probably not involved in P. lividus gametogenesis and no temporal window/critical oocyte size of E2 sensitivity is present, contrary to what was observed in starfish (Schoenmakers et al., Reference Schoenmakers, Van Bohemen and Dieleman1981; Takahashi & Kanatani, Reference Takahashi and Kanatani1981; Barker & Xu, Reference Barker and Xu1993). In addition, in P. lividus, MYP content was apparently not influenced by E2 exposure as it is in different asteroid species where oestrogen treatment affected vitellogenesis and protein incorporation into oocytes (Schoenmakers et al., Reference Schoenmakers, Van Bohemen and Dieleman1981; Takahashi & Kanatani, Reference Takahashi and Kanatani1981; Van der Plas et al., Reference Van der Plas, Koenderman, Deibel-van Schijndel and Voogt1982). These different results could be due to class-specific mechanisms: echinoids and asteroids could have different hormonal regulation systems for oocyte development and growth. This is supported by recent findings regarding echinoderm yolk protein differences. A yolk protein has been characterized and found to be a vitellogenin-like molecule (vtg), whereas in echinoids no vitellogenin-like molecule has been found to date and, although the sea urchin genome contains a predicted vtg, this is probably a pseudogene as suggested by its several atypical features (Prowse & Byrne, Reference Prowse and Byrne2012). These findings strongly underline the diversity present between the two classes, supporting the hypothesis that not only yolk proteins but probably several other aspects, including hormonal mechanisms, in sea urchin reproductive biology could be completely different from those present in asteroids. However, the protein content alone might fail to reveal steroid effects on MYP expression. Indeed, in the sea cucumber Apostichopus japonicus, although the intensity of MYP band in SDS-PAGE gel remained stable throughout ovarian development, a clear increase of MYP transcription was observed. The inconsistency of the levels of MYP mRNA and MYP protein was proposed to be due to MYP delivery from the ovary to other animal regions such as coelomic fluids or its rapid consumption (Fujiwara et al., Reference Fujiwara, Unuma, Ohno and Yamano2010). Thus, further research is needed to confirm our results: steroid involvement in protein expression/synthesis has been proposed by several authors (Harrington & Ozaki, Reference Harrington and Ozaki1986; Barker & Xu, Reference Barker and Xu1993; Varaksina & Varaksin, Reference Varaksina and Varaksin2001, Reference Varaksina and Varaksin2002) and requires further examination, for example by investigating other target organs (MYP is synthesized mainly in the gonads and the digestive tube) and employing different technical and methodological approaches.
In conclusion, under the conditions of this work, our results indicate that E2 do not strongly influence echinoid gonad maturation and, particularly, do not promote gamete maturation, in contrast to their effects on vertebrates (Lange et al., Reference Lange, Hartel and Meyer2002). No specific reproductive stage sensitive to E2 was found, suggesting that asteroids and echinoids have different hormonal mechanisms. These findings seem to confirm our previous results: although E2 is present in sea urchin tissues it does not play a key role in the control of the sea urchin reproductive cycle (Mercurio et al., Reference Mercurio, Tremolada, Nobile, Fernandes, Porte and Sugni2015). The role of testosterone in echinoderm gametogenesis has been less explored and further studies appear necessary to confirm our results. Our experiments were designed in order to reduce as much as possible individual variability and synchronize animal reproductive conditions. In this way, the heterogeneity present in field populations did not affect our results, supporting the reliability of our findings. In addition, their consistency was further demonstrated by performing in vitro studies that suggested ovarian cell insensitivity to the tested hormones. Nevertheless, since sex-steroids are present in echinoderms and are endogenously synthesized, it is highly likely that they are involved in some aspect of echinoderm biology and physiology, such as the regulation of lipid metabolism and protein synthesis (Wasson et al., Reference Wasson, Gower and Watts2000b; Varaksina & Varaksin, Reference Varaksina and Varaksin2001, Reference Varaksina and Varaksin2002).
ACKNOWLEDGEMENTS
We are grateful to the Protected Marine Area ‘Isola di Bergeggi’ (SV) for giving us permission to collect experimental animals. This research did not receive any grant from any funding agency in the public, commercial or not-for-profit sector.
