Biological Characteristics of Fish Germ Cells and their Application to Developmental Biotechnology


Author’s address (for correspondence): G Yoshizaki, Department of Marine Biosciences, Tokyo University of Marine Science and Technology, Tokyo, Japan. E-mail:


We have revealed several unique characteristics of germ cell development using rainbow trout, including the fact that spermatogonia transplanted into the peritoneal cavity of newly hatched embryos migrate toward recipient gonads, that spermatogonia transplanted into female recipients start oogenesis and produce functional eggs and that diploid germ cells transplanted into triploid trout can complete gametogenesis. By combining these unique features of fish germ cells, we established allogeneic and xenogeneic transplantation systems for spermatogonia in several fish species. Spermatogonia isolated from the mature testes of vasa-green fluorescent protein (Gfp) transgenic rainbow trout were transplanted into the peritoneal cavity of triploid masu salmon newly hatched embryos. These spermatogonia migrated toward recipient salmon genital ridges with extending pseudopodia and were subsequently incorporated into them. We further confirmed that the donor-derived spermatogonia resumed gametogenesis and produced sperm and eggs in male and female salmon recipients, respectively. By inseminating the resulting eggs and sperm, we obtained only rainbow trout offspring in the F1 generation, suggesting that the triploid salmon recipients produced functional gametes derived only from donor trout. We further confirmed that this intra-peritoneal transplantation of germ cells is applicable to several marine fishes, which could be of benefit in the production of bluefin tuna that has a large broodstock (>100 kg) and is difficult to maintain in captivity. Gamete production of bluefin tuna could be more easily achieved by generating a surrogate species, such as mackerel, that can produce tuna gametes.


We recently established a method to create masu salmon (Oncorhynchus masou) broodstock that produce both eggs and sperm of rainbow trout (Oncorhynchus mykiss) and succeeded in deriving offspring trout populations through fertilization using these eggs and sperm (Okutsu et al. 2007). This technology has various promising applications in the field of fish biotechnology. For example, the Pacific bluefin tuna (Thunnus orientalis), one of the most commercially valuable fish, takes 3–5 years to reach sexual maturity at which time it reaches a body weight of more than 100 kg (Nagasawa et al. 2009). As a result, gamete production for this species is expensive in terms of time, cost and labour and requires 40–80-m-diameter floating net cages to maintain adult fish. However, if Pacific bluefin tuna spermatogonia could be transplanted into the closely related chub mackerel (Scomber japonicus), which reaches sexual maturity in 1 year at a body weight of <500 g, Pacific bluefin tuna gametes could be produced more easily and rapidly, even in a land-based small fish tank. Another possibility is to produce the eggs and sperm of species facing extinction, such as sturgeon, which requires at least 10 years to mature, in closely related fish species that mature in a shorter period.

The principle of this technology is simple, as shown in Fig. 1. Progenitor cells of eggs and sperm are isolated from the target species and transplanted into the recipient species. Donor-derived eggs and sperm are then produced in the recipient. However, for the technology to be successful, it is necessary to combine several key biological properties of fish. In this review, we introduce these properties while outlining how we generated surrogate masu salmon that produce only rainbow trout offspring.

Figure 1.

 Principle of egg and sperm production from a different species in surrogate parents using germ cell transplantation

Newly Hatched Embryos Lack a Mature Immune System

When transplanting the live cells of another species into recipient fish, the first obstacle to overcome is rejection of transplanted cells by the immune system of the recipient. In many animal species, the thymus and T-cells remain undifferentiated for an extended interval, even as embryonic development progresses. Indeed, as immune immaturity in salmonid fish lasts for approximately 2 weeks after hatching, they lack the ability to reject foreign cells during this time (Manning and Nakanishi 1996). We therefore hypothesized that transplanted rainbow trout primordial germ cells (PGCs) would not be rejected by newly hatched masu salmon. However, the extremely small gonads of salmon embryos (body length is approximately 1.3 cm) presented another challenge as we began PGC transplantation using a micromanipulator. We therefore focused on a truly unique ability of PGCs themselves.

Germ Cells can Seek Out and Migrate to the Gonads

During embryogenesis, all animals first form a gonadal anlagen consisting only of somatic cells. PGCs that are specified outside of the gonads seek out the location of the gonadal anlagen and migrate there by chemotaxis using pseudopodia (Yoshizaki et al. 2002; Raz 2004). On the basis of this knowledge, we hypothesized that if we transplanted PGCs in the vicinity of masu salmon gonads, they would migrate there and become incorporated into them. Indeed, we were able to show that transplantation of PGCs taken from mixed sex donors into the peritoneal cavity of newly hatched recipient embryos (Fig. 2) resulted in their migration and incorporation into recipient gonads in both sexes (Takeuchi et al. 2003, 2004). We also found that the rainbow trout PGCs that were taken up proliferated and began to differentiate within the masu salmon gonads. Some of the male recipient salmon grew to maturity after a full year. As we prepared donor germ cells from transgenic rainbow trout carrying the Gfp gene under the control of the germ cell-specific vasa promoter (Yoshizaki et al. 2000a,b; Takeuchi et al. 2002), we were able to use PCR analysis using Gfp gene-specific primers to determine whether milt from the recipient salmon contained trout sperm. The results indicated that five of 37 salmon produced donor-derived trout sperm (Takeuchi et al. 2004).

Figure 2.

 Microinjection of germ cells into the peritoneal cavity of newly hatched embryos

Subsequently, we conducted a progeny test involving the artificial insemination of sperm into normal rainbow trout eggs. We predicted, however, that even though the milt from these five individuals contained trout sperm, it would also contain large amounts of salmon sperm. This would lead to the creation of hybrids following the fertilization of rainbow trout eggs by masu salmon sperm within the recipient salmon milt; we expected that hybrid hatching would be delayed and that the embryos would die soon after hatching. Conversely, we predicted that fertilization of trout eggs by trout sperm in the milt of recipient salmon would produce normal rainbow trout embryos. Although many individuals were indeed hybrids with delayed hatching, 0.4% of F1 individuals hatched normally with exactly the same timing as normal rainbow trout. PCR analyses revealed that these hatchlings carried the vasa-Gfp genes introduced into the donor trout. Further, the PGCs of the vasa-Gfp-positive hatchlings emitted green fluorescence. We therefore concluded that the hatchlings were donor-derived rainbow trout and that rainbow trout sperm had been produced by male masu salmon from the transplanted trout PGCs (Takeuchi et al. 2004).

However, two issues remained. The first was that recipient salmon did not produce trout-derived eggs, and the second was that the trout sperm produced by the salmon accounted for no more than 0.4% of all the sperm. To overcome these problems, we attempted to use spermatogonia.

Spermatogonia of Adult Fish can Colonize Embryonic Gonads

In conducting PGC transplantation experiments, it is often difficult to obtain sufficient PGCs as these cells are only found in hatching-stage embryos. Individual rainbow trout embryos, for example, possess an extremely small number (approximately 50–100) (Nagler et al. 2011). Moreover, only 20–30 PGCs per fish can be isolated through enzymatic treatment. In the above transplant experiments, we required large numbers of hatched embryos to transplant 10–20 PGCs into a single recipient. This posed difficulties for application in commercially valuable species, such as bluefin tuna, compared with farmed fishes such as rainbow trout. We therefore focused on spermatogonia found abundantly in the testes of males regardless of their age and reproductive cycle. In mammalian studies, it has been shown that certain cell populations of spermatogonia have the capacity to act as stem cells, that is, they possess both self-renewal and differentiation abilities (Yoshida et al. 2007). As the prospective identification of spermatogonial stem cells is as yet unavailable in fish, we isolated whole type A spermatogonia (ASG), which were anticipated to contain spermatogonial stem cells, from rainbow trout and transplanted them into masu salmon recipients.

In this study, we again used the vasa-Gfp-transgenic rainbow trout. As the vasa-Gfp gene is predominantly expressed in ASG of adult male fish, flow cytometry was used to enrich vasa-Gfp-expressing ASG (Yano et al. 2008), which were then transplanted into newly hatched rainbow trout embryos (allogeneic transplantation). It was found that even if ASG prepared from adult fish were transplanted into the peritoneal cavity of sexually undifferentiated embryos immediately after hatching, they migrated and were incorporated into recipient gonads (Fig. 3). In this case, it was demonstrated that if the transplant recipient was male, the donor-derived spermatogonia differentiated into functional sperm in the recipient testes. Thus, ASG in the adult fish testes can interact well with the gonadal somatic cells of the embryo and can resume spermatogenesis in an allogeneic somatic microenvironment (Okutsu et al. 2006).

Figure 3.

 Spermatogonia transplanted into the body cavity of a recipient hatchling. The donor-derived spermatogonia attached to the peritoneal wall and extended pseudopodia. Dotted line indicates position of recipient gonad

Spermatogonia of Adult Fish can Differentiate into Functional Eggs

When newly hatched female embryos were used as recipients, the ASG derived from adult fish were incorporated into the ovarian anlagen of the recipient and then differentiated into mature eggs. These eggs were used for artificial insemination, leading to the production of normal offspring, and suggesting that ASG have the capacity to differentiate into both sperm and eggs, at least in rainbow trout. Moreover, this sexual plasticity is not only observed in individuals immediately after sex differentiation, but has been confirmed to exist in ASG of adult fish that have already matured (Okutsu et al. 2006; Yoshizaki et al. 2010a). Recently, similar sexual plasticity was also found in oogonia isolated from the ovaries of female rainbow trout. The oogonia-derived sperm produced by male recipients are fully functional and can produce normal offspring (Yoshizaki et al. 2010b). These results are a major breakthrough in terms of the practical application of germ cell transplantation and enable large quantities of ASG to be obtained from a single male fish.

On the basis of these findings, we attempted to xenotransplant rainbow trout ASG into masu salmon recipients. In this experiment, approximately 18 000 total testicular cells containing approximately 10 000 ASG were transplanted into the peritoneal cavity of a recipient embryo. We previously found that germ cell transplantation is possible using whole testicular suspension as donor cells, instead of FACS-sorted ASGs. By simply increasing the number of transplanted cells, we anticipated a successful transplantation efficiency, compared with an unsuccessful attempt at producing rainbow trout eggs in female masu salmon using PGCs. Thirty-three male recipients had matured by the spawning season in year 2 after transplantation. PCR analyses using DNA extracted from the sperm with Gfp-specific primers revealed that the Gfp gene was present in 16 individuals, indicating that they were producing trout-derived sperm (Okutsu et al. 2008). When normal rainbow trout eggs were fertilized with these sperm, some rainbow trout offspring were produced with the sperm derived from nine fish (in the remaining seven fish, the trout sperm had a low contribution rate and was not detected among F1 individuals tested). The mean occurrence rate of F1 individuals derived from donor trout was 18.9% in these nine groups; thus, the transplantation efficiency was considerably higher than in the earlier experiment (F1 individuals derived from donor, 0.4%).

Two donor-derived rainbow trout offspring were successfully obtained (0.6% of total F1 individuals) from one of 38 mature female recipient salmon. Thus, using spermatogonia for the transplant not only improved the production efficiency of donor-derived rainbow trout in male recipients but also allowed the production of functional rainbow trout eggs in female masu salmon recipients (Okutsu et al. 2008). A remaining issue, however, was that the F1 generation produced by the recipients included both donor-derived individuals and large numbers of individuals originating from recipient gametes. We therefore next attempted to generate masu salmon that produce only rainbow trout gametes.

Triploid Fish are Sterile, but their Somatic Cells are Normal

Masu salmon that produce only rainbow trout offspring would be recipient salmon that do not make their own gametes. We therefore decided to use triploid salmon. In lower vertebrates, including fish, second polar body extrusion occurring after fertilization can be inhibited by immersing the fertilized eggs in warm or cold water, and triploids can be created that have one extra set of chromosomes derived from the second polar body. These triploid individuals develop and grow normally, but are sterile (Thorgaard et al. 1992). We therefore attempted an experiment to transplant spermatogonia derived from diploid rainbow trout into triploid recipient salmon. The resulting triploid recipient male individuals were raised, and an attempt was made to obtain sperm in the spawning season. It was found that of the 29 individuals, 10 of the triploid male salmon recipients showed normal secondary sexual characteristics and sperm could be collected from them. It was impossible to collect any sperm from the normal triploid individuals that had not received trout cells (Okutsu et al. 2007).

In observations of female triploid individuals that received no trout cells and served as controls, no oocytes were observed in which vitellogenesis had advanced. However, triploid salmon that had received trout cells contained colonies of large vitellogenic oocytes with advanced vitellogenesis. Green fluorescence was clearly emitted from all vitellogenic oocytes, indicating that they were derived from rainbow trout. Ovulation was seen in five of 50 individuals, which survived until spawning season. There was, as expected, no ovulation in the triploid control individuals (non-transplanted), and vitellogenesis was not seen. When the sperm and eggs obtained from these mature triploid recipient salmon were artificially inseminated, all F1 individuals hatched with exactly the same timing typical of rainbow trout. Further, these individuals carried the Gfp marker gene (Okutsu et al. 2007), indicating that all of the F1 generation were produced by donor-derived eggs and sperm. Rapid amplification of polymorphic DNA (RAPD) PCR and restriction fragment length polymorphism (RFLP) were used to confirm that the nuclear and mitochondrial genomes of these newly hatched embryos were identical to those of diploid rainbow trout, and development of rainbow trout embryos born from masu salmon parents proceeded normally.

As described above, we succeeded in markedly increasing the production efficiency of donor-derived offspring from germ cell transplantation using spermatogonia as donor cells and triploid sterile recipients.

These Biological Properties are Well Conserved Throughout Evolution

The key biological properties, which newly hatched embryos lacking mature immune systems and germ cells that can seek out and migrate to gonads, are well conserved throughout vertebrate evolution (Raz 2004). Thus, intra-peritoneal ASG transplantation using newly hatched embryos as recipients is expected to be applicable to any other fish species. We therefore applied the above-mentioned ASG transplantation technique to the nibe croaker (Nibea mitsukurii) (Takeuchi et al. 2009; Higuchi et al. 2011), belonging to the Scianidae family, the chub mackerel (Scomber japonicus) (Yazawa et al. 2010), belonging to the Scombridae family, and the yellowtail (Seriola quiqueradiata) (Morita et al. 2012), belonging to the Carangidae family. In these experiments, we used total testicular cell suspension prepared from allogeneic, non-transgenic individuals as donor cells because no transgenic strain carrying green germ cells is available in these three species. The donor testicular cells were labelled with a red fluorescent dye, PKH26, to visualize them in recipient embryos.

In all cases, intra-peritoneally transplanted ASG showed no sign of immune rejection and migrated to the recipient genital ridges where they were incorporated and commenced proliferation and differentiation. To further confirm whether donor ASG-derived gametes could be produced in the recipients, we focused on the yellowtail. This species is the most commonly cultivated fish in Japan with an annual production of 150 000 metric tons in 2003, constituting approximately 57% of the total farmed marine finfish production. In this study, diploid fertile recipients were used as there is no current method to induce triploidy in yellowtail. We raised the yellowtail recipients that received allogeneic ASGs for two and a half years and observed that nine males and four females matured out of 43 recipients (a mixture of males and females). The progeny tests and the following parentage tests with microsatellite markers revealed that all mature males and females produced donor-derived offspring. The average percentages of donor-derived offspring in the F1 generation were 66.6% (19.6–98.8%) for male recipients and 63.2% (17.0–97.5%) for female recipients. The resulting offspring developed normally with normal external morphology (Morita et al. 2012). We are currently working on transplanting yellowtail ASG into closely related species of small body size and with a short generation time.

In addition to the xenogeneic germ cell transplantation, the above-mentioned allogeneic transplantation of ASG could be a powerful tool for the aquaculture industry in the preservation of genetic resources from particular strains carrying desirable genetic traits. Although the cryopreservation of fish eggs is impossible because of their large size, high yolk content and low permeability of the egg membrane, testicular cells containing ASG can be semipermanently cryopreserved in a similar manner to somatic cells (Okutsu et al. 2007; Yoshizaki et al. 2011). As frozen ASG can be converted into functional eggs and sperm via transplantation experiments, these techniques are a powerful alternative to the cryopreservation of eggs or embryos (Fig. 4). It would be particularly valuable to establish a germ cell repository using the ASG of endangered fish species and strains carrying commercially valuable genetic traits. Further, the intra-peritoneal transplantation of oogonia has been successful in zebrafish, with the resulting male recipients producing sperm derived from transplanted oogonia (Wong et al. 2011). Thus, it was confirmed that the intra-peritoneal transplantation of germ cells originally established using salmonid fish can be applied to a wide range of marine and freshwater fish species.

Figure 4.

 Application of spermatogonial transplantation to the conservation of endangered fish species. Restoration of an extinct species could be achieved by transplanting frozen spermatogonia into recipient hatchlings of a closely related species

Conclusions and Perspectives

In this article, we have introduced a novel germ cell transplantation technique for use in fish. These experiments have combined the following biological characteristics of living organisms to build an effective method: (i) newly hatched fish embryos are immunologically immature; (ii) germ cells can seek out and migrate to gonads; (iii) fish spermatogonia differentiate into eggs as well as sperm; and (iv) although triploid fish are sterile, their somatic cells are normal. This transplantation procedure is both simple and rapid, enabling us to transplant into 30–50 newly hatched recipients in an hour.

Recently, successful transplantation of spermatogonia into the testes of adult recipients has been performed in tilapia (Lacerda et al. 2006, 2010, 2012), pejerrey (Majhi et al. 2009) and zebrafish (Nóbrega et al. 2010). If these techniques can become applicable to the production of donor-derived eggs, they would also be useful in the creation of surrogate broodstock and preservation of genetic resources in fish. Furthermore, the development of methods for the in vitro culture of spermatogonia in several fish species could improve the efficiency of germ cell transplantation because donor cells could be amplified in vitro prior to use (Hong et al. 2004; Shikina et al. 2008; Shikina and Yoshizaki 2010). Other improvements include our identification of a cell surface protein that is predominantly expressed in rainbow trout ASG (Nagasawa et al. 2010), which could enable the enrichment of ASG from fish testes using specific antibody-mediated flow cytometry or magnetic cell sorting. Finally, we recently developed a simple flow cytometry protocol to enrich ASG using light scattering properties (Kise et al. 2012). We predict that these research efforts will lead to rapid advances in novel developmental biotechnology for aquacultural and conservation applications.

Conflicts of interest

None of the authors have any conflicts of interest to declare.