Sea cucumbers are widely considered as commercially valuable with a high demand for consumption and use in some traditional medicines (Olivera-Castillo et al., 2013; Fahmy et al., 2015). As a result, many species have been critically overfished, which has led to the rapid exhaustion of wild populations (Lovatelli, 2004; Purcell et al., 2011). With reduced stocks, artificial breeding has become an effective solution to produce high quality seed for aquaculture, stock enhancement, and sea ranching. Spawning induction is a critical procedure in artificial breeding, however, traditional stimulation methods such as thermal shock, drying then rehydration, exposure to conspecific sperm, and a combination of the aforementioned treatments (Al Rashdi et al., 2012) are not efficient. In addition, traditional stimulation methods are thought to reduce the health of broodstock. As such, the employment of reproductive neuropeptide hormones for achieving reproductive maturation and spawning is an innovative approach with great potential to advance hatchery production in sea cucumber aquaculture.

In sea cucumbers, immature oocytes cannot be fertilized and insemination is only successful following germinal vesicle breakdown (GVBD), leading to the formation of the first and second polar bodies (Maruyama, 1980). Oocyte maturation and ovulation can be induced by a hormonal substance present within the radial nerve (Maruyama, 1980). Crude radial nerve extract (RNE) has been shown to stimulate maturation of fully grown oocytes within various sea cucumber species, including Holothuria leucospilota, Holothuria pervicax, Holothuria moebi, Holothuria pardalis, and Apostichopus japonicus (Maruyama, 1980, 1985; Katow et al., 2009). Meanwhile, several peptide hormones have been recognized as inducers of oocyte maturation in sea cucumbers, including the gonad-stimulating substance-like molecule (GSSL), cubifrin (NGIWYamide, NGLWYamide), and the heptapeptide QGLFSGVamide (Kato et al., 2009; Katow et al., 2009; Leonet et al., 2009; Fujiwara et al., 2010; Yamano et al., 2013).

GSSL is a 4.8 kDa single chain peptide which was isolated from the radial nerve of A. japonicus. A synthetic GSSL peptide, at 6 μM, was capable of inducing GVBD at 50% efficiency (Katow et al., 2009). The neuropeptide cubifrin-I (NGIWYamide) and its derivative cubifrin-L (NGLWYamide) have been reported as potent inducers of in vitro oocyte maturation and spawning in A. japonicus (Kato et al., 2009; Fujiwara et al., 2010; Yamano et al., 2013). However, cubifrin may not be effective for all sea cucumber species, where for example, synthetic cubifrin could not induce oocyte maturation in H. leucospilota at concentrations from 1 nM to 10 μM (Chieu et al., 2018). Meanwhile, a second synthesized peptide, QGLFSGVamide could induce GVBD at concentrations ≥1 μM (Kato et al., 2009). However, due to its low potency, this peptide was not suggested as the primary endocrine regulator of oocyte maturation and therefore was not considered for further study (Kato et al., 2009). Lastly, MIF (maturation inducing fractions) collected by gel filtration chromatography from spawned egg extract of the regular sea urchins (e.g., Echinometra mathaei, Stomopneustes variolaris, Paracentrotus lividus) could also induce oocyte maturation and spawning in the sea cucumber (Leonet et al., 2009). The bioactive peptide sequences of MIF are still unknown.

The relaxin-like gonad-stimulating peptide (RGP), originally denoted as GSS (gonad-stimulating substance) in the sea stars due to its classification as a member of the relaxin superfamily, is suggested as the first identified invertebrate gonadotropin to trigger final gamete maturation (Mita et al., 2009; Mita, 2013, 2016). The neuroendocrine mechanism involved in endogenous regulation of oocyte maturation and spawning is considered to be similar among other echinoderm groups, including the sea cucumbers (Smiley, 1990; Mercier et al., 2009). RGP is a heterodimer composed of two peptides (A- and B-chains) with disulphide cross-linkages and the A-chain harboring a cysteine motif “CCXXXCXXXXXXXXC”, where “X” represents any other amino acid (Mita et al., 2009; Mita, 2013). Synthetic RGP (bonded) has been shown to induce oocyte maturation and ovulation within 30 min of incubation, however, no induction occurred when ovarian fragments were incubated with A- or B-chains alone, or as a mixture (unbonded) (Mita et al., 2009; Mita, 2013). As a result, the quaternary structure of RGP is necessary for hormonal action and this structure is complex and difficult to chemically synthesize. Instead of synthetic RGP, an effective approach to obtain RGP with correct quaternary structure is the production of a recombinant RGP in the methylotrophic yeast such as Hansenula polymorpha, Pichia pastoris, and Candida boidinii.

The methylotrophic yeast (P. pastoris) has been developed as a high-level production system for recombinant proteins (Hollenberg and Gellissen, 1997), with several advantages such as rapid growth rate, high levels of productivity in an almost protein-free medium, ease of genetic manipulation of well-characterized yeast expression vectors and diverse post-translational modifications of disulphide bonds within the yeasts secretory pathway (Weinacker et al., 2013). Recombinant gonadotropin or gonad-stimulating hormone like peptides have been produced and studied in several aquatic species, such as a recombinant tilapia LH stimulated the release of 11-ketotestosterone from mature testes (Kasuto and Levavi-Sivan, 2005); a recombinant molt-inhibiting hormone-B (MeMIH-B) that could induce yolk protein synthesis in the shrimp (Metapenaeus ensis) (Tiu and Chan, 2007); induction of spermatogenesis in the eel Anguilla japonica by a recombinant goldfish (Carassius auratus) gonadotropins (Hayakawa et al., 2008); regulation of steroidogenesis and early ovarian development in juvenile grouper by a recombinant orange-spotted grouper (Epinephelus coioides) follicle-stimulating hormone (FSH) (Chen et al., 2012); phosphorylation was elevated significantly in the testis of Eastern spiny lobster Sagmariasus verreauxi by a recombinant insulin-like androgenic gland hormone (Aizen et al., 2016); and in vitro and in vivo biological activities in oocyte development and spermatogenesis by a recombinant yellowtail kingfish (Seriola lalandi) FSH (Sanchis-Benlloch et al., 2017). However, production and biological activities of recombinant gonadotropin or gonad-stimulating hormones/peptides have been poorly studied in invertebrates, especially in echinoderm species.

RGP has not been well studied in the sea cucumbers with the exception of the first identification of relaxin-like peptide-1 (gonad-stimulating substance type) genes reported in Holothuria scabra, Holothuria glaberrima and A. japonicus (Suwansa-ard et al., 2018). To date, the bioactive functions of RGP have not been verified in sea cucumbers. In this study, we have investigated the phylogeny of sea cucumber RGPs with other echinoderm species relaxin-like peptides. A recombinant H. scabra RGP was generated in the P. pastoris yeast expression system and subsequently confirmed by mass spectrometry. The recombinant RGP, when injected, induced oocyte maturation and spawning and the spawned oocytes were successfully fertilized leading to larvae and juvenile development.

The enhancement of aquaculture animal breeding using molecular approaches is now a common avenue for replacing or complementing traditional approaches. This includes the use of neuropeptide/hormone manipulation to enhance breeding strategies. However, this approach requires an in-depth knowledge of the relevant molecular machinery. In sea stars, it is well-established that RGP plays an important role in oocyte maturation and spawning (Mita et al., 2009; Mita, 2013, 2016) and we speculated that it could also play a similar role for sea cucumbers. In this study, we have produced a recombinant sea cucumber RGP using a yeast expression system and its bioactivity was confirmed through stimulation of oocyte maturation and spawning.

In the echinoderms, RGP was first characterized in Patiria pectinifera (called PpeRGP) and then identified in other sea star species, including Asterias amurensis RGP (AamRGP) and Aphelasterias japonica RGP (AjaRGP). The protein structure of AamRGP appears to be closer to AjaRGP than it is to PpeRGP (Mita, 2016). Cross-species in vitro oocyte maturation experiments showed that both AamRGP and AjaRGP could not induce oocyte maturation and ovulation in the ovary of P. pectinifera, although the PpeRGP was active in ovaries of A. amurensis and A. japonica (Mita, 2016). To further investigate the conservation of RGP in echinoderms, in this study we explored the holothurian RGP and performed sequence comparisons and phylogenetic analysis. The Holothuria RGP showed highest homology to other echinoderm RGPs, with the highest conservation of amino acid composition found among the various Holothuria species (~87%), suggesting possible cross-species bioactivity for RGP activity within Holothuria species, although further functional assays were required to confirm this. All RGPs contain 6 highly conserved cysteine residues, 4 and 2 of which are present in the A- and B-chains, respectively. This indicates a high selective pressure during the evolution of these critical cysteine residues, which are responsible for disulphide bridge formation. Despite the observed conservation within the A- and B-chains of echinoderm RGPs, the amino acid composition within the RGP-related peptide region was considerably variable. RGP belongs to the insulin/IGF/relaxin family, which include two sub-families such as insulin/IGF and RGP families (Wilkinson and Bathgate, 2007). Our alignment also showed conservation of amino acid composition, especially within the putative A- and B-chain regions, among RGPs and IGFs in echinoderms. The major characteristics of RGP that are different to IGF include: (i) the conservation of a glycine residue prior to the third conserved cysteine residue with the A-chain of RGPs, but not IGFs; and (ii) an amino acid at the C-terminal end of the A-chain, which is the cysteine in RGPs but asparagine in IGFs. The phylogenetic analysis also confirmed a common root for these two peptide families. Within the RGP clades, two distinct sister clades were observed for RGPs from two different echinoderm classes, including the sea cucumber (holothuroids) RGPs and the sea star (asteroids) RGPs (see Figure 2B), which reflects the evolutionary divergence of this peptide during the evolution of echinoderms.

Following protein expression and purification, an average of 0.7 mg pure RGP was obtained from 1 L of crude supernatant. This yield is comparable with previous studies, including the purification of FSH and LH recombinant proteins by using the same purification technique. For example, 0.4–0.9 mg/L recombinant yellowtail kingfish FSH (Sanchis-Benlloch et al., 2017) and 2.5 mg/L for the grouper FSH (Chen et al., 2012). Lower yields have been reported, such as 0.1 mg/L for the tilapia FSH (Aizen et al., 2007), 0.3 mg/L for the carp LH (Hollander-Cohen et al., 2017) and 0.08 mg/L for the tilapia LH (Kasuto and Levavi-Sivan, 2005). Generally, the yield was suggested to be varied depending on culture conditions (Yu et al., 2010) or the formation of dimeric forms (Aizen et al., 2007). The yield of the heterodimer was usually significantly lower than that of the monomer probably because of the complex structure of the former, for example, 6.5 mg/L were produced for the tilapia FSHβ (monomer), while a yield of only 0.1 mg/L was obtained for the tilapia FSHβα (heterodimer) (Aizen et al., 2007). Production of the recombinant H. scabra RGP in the yeast expression system is an effective approach to obtain the copious amounts of RGP that would be required for commercial applications. However, the process of obtaining the slightly cheaper partially purified RGP and delivering to animals on a mass scale still requires optimisation and development of a method that is less laborious.

To confirm the successful production of the recombinant H. scabra RGP, in-solution digestion followed by LC-MS/MS was performed, which provided evidence of almost complete sequence coverage (see Figure 3A). The spectra additionally suggested the carbamidomethylation on four cysteines, which was potentially the result of reduction on cysteine residues and/or disulphide bridges. No other post-translational modifications were detected for the recombinant RGP. The MD simulation suggested the formation of three disulphide bonds and a disulphide bridge is formed within the A-chain. In terms of secondary structure, there was mainly turn and coil structures formed during the simulation, such as helix and β sheet, which could also be seen from the RMSD calculation that shows significant fluctuation with respect to time. In conclusion, we found that the expression and purification of the recombinant RGP in our study was successful in term of quantity and a properly folded structure as confirmed by MD simulation. Therefore, this established RGP expression and purification procedures that could be applied and/or used as a fundamental procedure for RGP mass production for future uses.

In vitro and in vivo bioassays were used to assess the bioactivity and function of recombinant RGP in the sea cucumbers, H. scabra and H. leucospilota. The in vitro GVBD bioassay demonstrated that our recombinant RGP could induce GVBD, even at low concentrations (see Table 2). The recombinant H. scabra RGP peptide (SR, HR) could induce oocyte maturation/GVBD in both H. scabra and H. leucospilota, which is consistent with the conservation and similarity in the amino acids of RGP precursor (see Figure 2). Hence, a “generic” RGP recombinant may be employed to induce GVBD and spawning in more than one species of sea cucumber, although additional experiments showing compatibility of our RGP recombinant with other commercially important sea cucumber species should be verified in the future. Similarly in previous work, synthetically produced RGP could successfully induce GVBD in the sea stars Asterias aumerensis and Asterina pectinifera using in vitro bioassays (Kanatani et al., 1971; Mita, 1993). However, individual synthetic chains of RGP (either A- or B-chain) did not induce GVBD or spawning, whereas a synthetic heterodimeric form could induce GVBD and ovulation within 1 h of incubation in the Crown-of-Thorns sea star, Acanthaster planci (Mita et al., 2015). We found that partial GVBD and distortion of oocytes occurred when ovarian fragments were incubated with the crude supernatant CS and CR (see Table 2, Figure 4), which was likely due to a high salt concentration within these crude extracts. Meanwhile, a lower efficiency of GVBD was observed in the partially purified RGP (SR). Therefore, we suggest that it is essential to perform a complete or partial purification of the recombinant RGP before being employed as an oocyte maturation stimulant, in order to maximize the mature oocyte quality and quantity for further fertilization.

A recombinant RGP has advantages over the use of crude RNE, since it does not require the sacrifice of sea cucumbers, and the potential for other RNE factors that may stress the animals. Our bioassay experiments found that RGP stimulation of spawning did not cause evisceration or any observable stress to broodstock. In contrast, several traditional approaches in spawning stimulation [e.g., thermal shock, drying, and water pressure, sperm-induced spawning, dried alga spirulina bathing treatments (Battaglene et al., 2002; Al Rashdi et al., 2012)], could lead to significant animal stress. Another important factor that needs to be considered is the sea cucumber gonadosomatic growth peaks during the year (Battaglene et al., 2002). The traditional approaches usually require sea cucumbers at fully mature gonad stage (gonadal stage IV) for efficient spawning, but we showed in the current study that RGP could induce spawning as early as gonadal stage III (gonad weight 30–50 g and oocyte diameter ≥150 μm). The receptor for RGP still remains elusive, although in sea stars it has been recognized that the RGP is a ligand to a G protein-coupled receptor that is expressed on follicle cells of growing to fully growing ovaries (stages III to V) (Mita et al., 2011a,b). Thus, we hypothesize that follicle cells and gonadal stages could affect RGP efficiency in sea cucumbers, depending on the presence of an RGP receptor.

In vivo spawning by injection of RGP in H. scabra was successfully achieved using pure RGP (HR), showing that the sea cucumbers exhibited the head waving characteristic at 30–60 min post-injection, and subsequently spawned at either 90, 120, or 170 min post-injection (see Figure 5). Furthermore, spawning was complete within 30–60 min with no subsequent additional spawning observed. This novel method is superior to other traditional stimulating methods since it does not require several days of pre-stimulation, such as thermal stimulation, prior to a spawning event. In addition, the use of RGP is a time and cost-effective procedure since it allows a rapid and highly controllable spawning event.

Injection with the unpurified crude RGP (CS) resulted in no spawning in the first 60 min post-injection, and as such a second injection was performed, leading to spawning within 30, 60, or 120 min following the secondary injection. Spawning after the second injection was completed within 60–80 min. However, we observed that the animals did not show pre-spawning head waving behavior. Hence, we cannot preclude that stress induced spawning might occur when the animals were subjected to high concentration of RGP solution from 2 injections. These results support the observations from the GVBD in vitro assay in which the oocytes treated with high concentration of crude RGP were deformed. Therefore, we suggest that not only purity but also the concentration of RGP is critical and are factors for maximum success for fertilization in artificial breeding of sea cucumbers. Further studies are required to determine the minimum recombinant RGP threshold/concentration for efficient spawning induction in the sea cucumbers.

In this study, we found that the development of the H. scabra larvae produced from RGP-induced spawning proceeded normally and was comparable to previous studies engaged the traditional spawning stimulations for fertilization and larval production (Battaglene et al., 1999; Mercier et al., 2000; Ivy and Giraspy, 2006; Al Rashdi et al., 2012; Ajith Kumara et al., 2013; Robinson et al., 2013; Abidin et al., 2016). For example, the pentacula stage larvae were predominant at 18-21 dpf (see Table 3), which is similar to the previous reported durations of 13–26 dpf of metamorphosis to pentaculae (James, 2004; Dabbagh and Sedaghat, 2012; Mazlan and Hashim, 2015; Abidin et al., 2016). Furthermore, the larvae were reared through to their juvenile stage and appeared normal in their development. This result demonstrates that recombinant RGP for induced spawning is applicable in artificial breeding and that there are potential benefits for its use in sea cucumber hatcheries. However, the survival rates from fertilized oocytes, using the recombinant RGP, until the juvenile stages, need to be studied further.

This is the first study to demonstrate the production and use of a recombinant RGP in oocyte maturation and spawning induction in sea cucumbers. The outcome of this study provides a breakthrough new approach to artificial breeding in the sea cucumbers by using a recombinant RGP. We have developed the experimental procedures for the production of the recombinant RGP that could successfully induce in vitro oocyte maturation as well as spawning activity in sea cucumber broodstock. However, in order to help with hatchery production, further investigations are required in the future in order to obtain the optimal concentration and practicable delivery method. Furthermore, to understand the mechanism of RGP bioactivity in sea cucumbers, molecular signaling pathway analysis should be implemented. This information could help to overcome the seasonality requirements for RGP induced spawning.

The collection and handling of the animals in this study was carried out in accordance with the guidelines for the care and use of laboratory animals at the University of the Sunshine Coast, Australia.

SC and AE conceived the idea and provided oversight of the project from beginning to the final manuscript. HC carried out the laboratory components, in vitro and in vivo experiments and manuscript writing. LT carried out the in vivo spawning assay and was lead to the larvae and juvenile rearing at the Darwin Aquaculture Centre. MS, JN, and PP contributed to the sub-cloning, cloning, and all the steps in the laboratory production of the recombinant RGP. TW performed MS and molecular dynamic simulation. SS constructed sequence alignments and phylogeny analysis images. All authors provided critical feedback and contributed to the writing and review of the final manuscript.