Dorsomorphin Promotes Survival and Germline Competence of Zebrafish Spermatogonial Stem Cells in Culture

Zebrafish spermatogonial cell cultures were established from Tg(piwil1:neo);Tg(piwil1:DsRed) transgenic fish using a zebrafish ovarian feeder cell line (OFC3) that was engineered to express zebrafish Lif, Fgf2 and Gdnf. Primary cultures, initiated from testes, were treated with G418 to eliminate the somatic cells and select for the piwil1:neo expressing spermatogonia. Addition of dorsomorphin, a Bmp type I receptor inhibitor, prolonged spermatogonial stem cell (SSC) survival in culture and enhanced germline transmission of the SSCs following transplantation into recipient larvae. In contrast, dorsomorphin inhibited the growth and survival of zebrafish female germline stem cells (FGSCs) in culture. In the presence of dorsomorphin, the spermatogonia continued to express the germ-cell markers dazl, dnd, nanos3, vasa and piwil1 and the spermatogonial markers plzf and sox17 for at least six weeks in culture. Transplantation experiments revealed that 6 week-old spermatogonial cell cultures maintained in the presence of dorsomorphin were able to successfully colonize the gonad in 18% of recipient larvae and produce functional gametes in the resulting adult chimeric fish. Germline transmission was not successful when the spermatogonia were cultured 6 weeks in the absence of dorsomorphin before transplantation. The results indicate that Bmp signaling is detrimental to SSCs but required for the survival of zebrafish FGSCs in culture. Manipulation of Bmp signaling could provide a strategy to optimize culture conditions of germline stem cells from other species.


Introduction
The zebrafish embryo is an ideal model for in vivo analysis of germ cell determination and migration during early development [1,2], however complementary in vitro studies of zebrafish germ cell growth and differentiation have been lacking due to the absence of suitable cell culture systems. Mouse spermatogonial stem cell (SSC) and female germline stem cell (FGSC) cultures have proven to be valuable in vitro models for studies of germ cell differentiation and been used for the production of transgenic and knockout mice [3,4,5,6]. Recently, our lab used a drug selection strategy to establish zebrafish FGSC cultures that were initiated from Tg(piwil1:neo);Tg(piwil1:DsRed) transgenic fish. The Tg(piwil1: neo);Tg(piwil1:DsRed) fish express Neo and DsRed under the control of the piwil1 promoter enabling the use of G418 selection to isolate the piwil1:neo expressing FGSCs. Using this method a homogeneous population of FGSCs were selected and maintained for more than 6 weeks in culture during which the cells continued to express multiple germ cell markers. Following transplantation into recipient larvae, the cultured FGSCs successfully colonized the gonad of the host and produced functional gametes in the adult chimeric fish [7].
Since piwil1 is also expressed in the germ cell lineage of male zebrafish [8,9,10], in this paper, we apply the same drug selection strategy to establish zebrafish spermatogonial cell cultures from the gonad of male Tg(piwil1:neo);Tg(piwil1:DsRed) fish. Previous attempts to establish long-term spermatogonial cell cultures from zebrafish were ineffective due to spontaneous differentiation of the spermatogonia to non-proliferating spermatids [11,12] and to the cultures becoming overgrown by testicular somatic cells [13]. To avoid these problems, primary cultures were initiated from the testes of Tg(piwil1:neo);Tg(pi-wil1:DsRed) fish and treated with G418 to select Neo-expressing spermatogonia and eliminate somatic cells. A key component of the culture system was the addition of dorsomorphin to block Bmp signaling which prevented the spontaneous differentiation of the zebrafish spermatogonia and prolonged their survival in culture. The action of dorsomorphin on spermatogonia in culture is consistent with results of previous studies showing that mutation of the Bmp type I receptor impairs germ cell differentiation in zebrafish testis [14] and the addition of exogenous BMP promotes mouse spermatogonial differentiation in culture [15,16]. In our study, inhibition of Bmp signaling also enhanced the capacity of cultured spermatogonia to colonize the gonad following transplantation into recipient larvae and produce functional gametes in the adult chimeric fish.

Animals and Ethics
Zebrafish were maintained and staged as previously described [17]. All of the experimental procedures and protocols described in this study were approved by the Purdue University Animal Care and Use Committee and adhered to the National Research Council's Guide for Care and Use of Laboratory Animals.

Immunocytochemistry and Histology
Gonads dissected from zebrafish were fixed with either 4% paraformaldehyde in phosphate-buffered saline (PBS) or Bouin's fixative at 4uC overnight, and then processed through successive treatments of ethanol (50%, 70%, 95%, and 100%) followed by two xylene baths and embedded in paraffin. The serial paraffin sections (5 mm) were prepared for immunocytochemical staining to visualize Neo and Vasa within the testicular tissues of Tg(piwil1:neo) [7] according to published protocols [18,19]. Tissue sections were treated with mouse anti-neomycin phosphotransferase II monoclonal IgG (1:500 dilution, Abcam) and rabbit anti-zebrafish Vasa antiserum [20] (1:4000 dilution) and then stained with Alexa Fluor 488 AffiniPure goat anti-mouse IgG and Cy3 AffiniPure goat antirabbit IgG (1:500 dilution, Jackson ImmunoResearch Lab, Inc.). Neo and Vasa expression was evaluated by examining the tissue sections by fluorescence microscopy using a Nikon Eclipse TE200 microscope (Nikon, Tokyo, Japan) equipped with a RT Slider digital camera (Spot Imaging Solution, Sterling Heights, MI, USA). The sections were stained with hematoxylin-eosin for histological examination by light microscopy.

Cell Transplantation and Analysis
Cultured spermatogonia were dissociated (0.25% EDTA in PBS), washed, re-suspended in 50 ml of L-15 medium and immediately transplanted into two-week-old recipient larvae that were treated as embryos with a morpholino (MO) nucleotides against dnd [23]. Cell transplantations were performed under a stereomicroscope using a glass micropipette needle. Spermatogonia were transplanted into the abdominal cavity under the swim bladder close to the gonads according to a published method [19]. Two weeks after transplantation, the recipients were examined by fluorescence microscopy and the potential germline chimeras were identified based on the presence of DsRed-positive cells in the gonadal region. To determine if the transplanted cells were able to generate functional gametes, the recipient fish were raised to sexual maturity and paired with wild-type zebrafish mates.

Statistical Analysis
Data obtained from the growth assays were presented as the mean and standard deviation. For statistical analysis Student t-tests or one-way ANOVA was applied followed by Bonferroni-Dunn tests using SAS program. The significance was accepted at p,0.05.

Derivation of Spermatogonial Cell Cultures from
Tg(piwil1:neo);Tg(piwil1:DsRed) Zebrafish Immunocytochemical analysis of testicular tissue dissected from Tg(piwil1:neo) fish confirmed that Neo expression overlapped with Vasa in the spermatogonial cells (Fig. 1A-C) indicating that a drug selection approach could be used to isolate the germ cells in culture. To initiate spermatogonial cell cultures, an enriched fraction of DsRed-positive testicular germ cells was obtained by Percoll separation of dissociated testes dissected from Tg(piwil1: neo);Tg(piwil1:DsRed) fish. The germ cell fraction was seeded onto confluent growth-arrested OFC3 feeder cells that were engineered to express zebrafish Lif and Fgf2 (OFC3LF) or Lif and Gdnf (OFC3LG) [7] and G418 was added to eliminate any remaining somatic cells. Following 3 weeks of drug selection colonies of 4 or more DsRed-positive spermatogonia were observed in the culture ( Fig. 2A-D). The small (about 10 mm-diameter) DsRed-positive cells possessed morphological characteristics of spermatogonia in the testis [19] with round nucleus and a high nucleus-to-cytoplasm ratio ( Fig. 2A,B). Cell counts performed on 3 week-old cultures revealed that the number of DsRed-positive, G418-resistant spermatogonia was significantly greater on OFC3LF and OFC3LG feeder layers compared to the parent OFC3 feeder line (Fig. 2E). After 3 weeks the cell number began to dramatically decrease in all of the cultures resulting in only 5-10% of the cells remaining in cultures maintained for 6 weeks on OFC3LF and OFC3LG feeder layers (Fig. 2E).

Addition of Dorsomorphin Prolonged Spermatogonial Cell Growth but Inhibited FGSC Survival in Culture
The decrease in cell number observed in the spermatogonial cell cultures after 3 weeks was most likely due to differentiation of the spermatogonia since the DsRed-expressing cells became smaller and formed aggregates that detached from the feeder layer as described in previous studies of zebrafish spermatogonial cell cultures [13]. Analysis of the feeder cells by RT-PCR revealed that bmp2a, bmp2b and bmp4 were all expressed by the OFC3LF and OFC3LG cells (Fig. 3A) raising the possibility that Bmp signaling may be involved in inducing spermatogonial differentiation. To test this hypothesis, spermatogonial cell cultures were initiated and maintained in the presence of dorsomorphin, an inhibitor of the Bmp type I receptor. Results from growth assays revealed that the addition of dorsomorphin significantly enhanced the proliferation of spermatogonia in 3 week-old cultures (Fig. 3B) with the greatest mitogenic effect observed when the inhibitor was added at a concentration of 2 mM. At this concentration of dorsomorphin, the growth rate of DsRed-expressing spermatogonia increased 80% compared to the control with several large colonies containing more than 60 cells present in the cultures (Fig. 3C). In 6 week-old cultures the presence of dorsomorphin increased the number of spermatogonia by 30-fold compared to the control (Fig. 4A) and DsRed-positive colonies containing more than 100 cells were observed (Fig. 4B)  cultured spermatogonia continued to express the germ-cell markers dazl, dnd, nanos3, vasa and piwil1 and the spermatogonial markers plzf and sox17 [24,25] for at least 6 weeks in the presence of dorsomorphin (Fig. 4C). In contrast, when dorsomorphin was added to drug-selected FGSC cultures [7] cell growth was significantly inhibited compared to controls maintained in the absence of dorsomorphin (Fig. 5A-D). Also, the expression of germ cell markers was significantly reduced or completely lost by 3 weeks in the dorsomorphin-treated FGSC cultures (Fig. 5E).

Dorsomorphin Enhanced the Capacity of Spermatogonial Cell Cultures to Colonize the Gonads and Produce Functional Gametes Following Transplantation to Recipient Fish
Cell transplantation experiments were performed to evaluate the germline competency of the spermatogonial cell cultures. Spermatogonia obtained from 3-or 6-week-old cultures were transplanted into recipient larvae that were treated as embryos with a MO nucleotides against dnd to block endogenous germ cell formation [23]. For each experiment approximately 20 to 60 spermatogonia were transplanted into the recipient larva. Due to the small number of cells available, only approximately 10 or less DsRed-expressing cells were transplanted in experiments involving spermatogonia cultured for 6 weeks in the absence of dorsomorphin. Two weeks after transplantation, the recipients were screened by fluorescence microscopy to identify those that possessed DsRed-positive cells in the gonadal region (Fig. 6A). Survivors were raised to sexual maturity and successful germline transmission of the transplanted cells was confirmed by pairing the chimeric fish with wild-type zebrafish mates. All of the surviving adult recipient fish that were obtained from the transplantation experiments were male. A total of 36 recipient male fish were obtained from 2 independent cell transplantations using 3 weekold spermatogonia cultured in the absence of dorsomorphin. Five of the fish (13.8%) were fertile and generated normal offspring (Table 1). When spermatogonia that were cultured for 3 weeks in the presence of dorsomorphin were transplanted, 10 of the 52 (19.2%) adult recipients obtained were fertile ( Table 2). The presence of dorsomorphin had a dramatic effect on promoting the germline competency of 6 week spermatogonial cell cultures. None of the 28 adult recipients obtained by transplanting 6-week-old spermatogonia cultured without dorsomorphin were fertile while 9 of 50 (18%) recipients (Table 3) produced from the 6 week dorsomorphin-treated spermatogonia were able to reproduce. Genomic PCR analysis revealed 100% inheritance of piwil1-neo confirming that all of the offspring were derived from the transplanted spermatogonia (Fig. 6B). All of the fertile recipient fish obtained were able to produce multiple batches of normal offspring for at least 6 months. Dissection of three fertile recipients that were produced from transplanted 6-week-old spermatogonial cell cultures revealed that two of the fish developed only a single testis on one side of the body and the third fish developed two testes of unequal size (Fig. 6C). All of the testes examined contained DsRed-positive cells (Fig. 6C1). Histological examination of the fully-developed testes revealed the presence of a significant number of mature sperm along with spermatogonial cells at different developmental stages present in individual tissue sections (Fig. 6D). Results from cell transplantation experiments demonstrated the presence of SSCs in the spermatogonial cell cultures derived from the testes of Tg(piwil1:neo);Tg(piwil1:DsRed) zebrafish.

Discussion
In adult mammals, the SSC population in the testis is represented by the type A undifferentiated (A und ) spermatogonia that are present as single (A s ), paired (A pr ) or multiple aligned (A al ) cells [26,27]. The type A pr and A al spermatogonia are able to dissociate to generate type A s cells [28]. In contrast, the zebrafish testis possesses two subtypes of undifferentiated type A spermatogonia, A und and A und* , that possess stem cell characteristics based on cell transplantation experiments [29,30]. In this study we showed that Neo is expressed in the spermatogonia of the Tg(piwil1:neo) fish enabling us to use a drug selection approach to derive spermatogonial cell cultures initiated from the testes of Tg(piwil1:neo);Tg(piwil1:DsRed) fish. Using G418 selection to eliminate the somatic cells that were present in the testicular primary cultures enabled us to avoid the problem of the spermatogonia becoming overgrown in culture which was encountered in previous studies [13,31]. Another advantage of using the Tg(piwil1:neo);Tg(piwil1:DsRed) fish was that the expression of piwil1:DsRed made it possible to visually evaluate the growth of the spermatogonia in culture by fluorescence microscopy. The zebrafish spermatogonial cell cultures were derived in conditions that were similar to those previously used to propagate zebrafish FGSCs in culture [7]. A unique requirement of the spermatogonial cell culture system, however, was to block Bmp signaling by the addition of dorsomorphin to the medium. Previous studies have shown that Bmp functions to promote spermatogonial differentiation both in vivo and in spermatogonial cultures. Mutation of the zebrafish Bmp type I receptor resulted in an impairment of spermatogonial differentiation in vivo while addition of BMP4 to mouse SSC cultures promoted the production of differentiated spermatids [14,15,16]. Since our analysis of the OFC3 feeder cells, used to derive our spermatogonial cell cultures, showed that the cells expressed bmp2a, bmp2b and bmp4, we decided to block Bmp signaling by adding dorsomorphin to the culture medium. Dorsomorphin, also known as compound C, is a small molecular weight inhibitor of the Bmp type I receptor [32]. The compound has been shown to disrupt dorsoventral patterning when it is administered to zebrafish embryos [33]. We found that in cultures maintained for 3 weeks in the absence of dorsomorphin the zebrafish spermatogonia became smaller and formed aggregates that detached from the feeder cells indicating that the cells had differentiated to non-proliferating spermatids [11]. Addition of dorsomorphin to the cultures inhibited differentiation and enabled the spermatogonia to proliferate and express appropriate markers for more than 6 weeks in vitro. Further evidence that dorsomorphin maintained the SSC characteristics in culture was obtained from the cell transplantation experiments showing that the presence of dorsomorphin extended the time that the cultured SSCs could be transplanted and colonize the gonads of recipient larvae and produce functional gametes in the adult chimeric fish. In contrast, when dorsomorphin was added to zebrafish FGSC cultures, the cells stopped proliferating and lost expression of the germ cell markers within 3 weeks, demonstrating a requirement for Bmp signaling by the female germ cells in culture. These results showing that Bmp signaling has the opposite effect on zebrafish male and female germline stem cells in culture may be useful in optimizing conditions that support the growth of germline stem cells from other species. The OFC3 feeder cell line used to initiate the spermatogonial cell cultures was derived from ovaries obtained from Tg(gsdf:neo) zebrafish [7]. Gsdf is produced in somatic cells of the fish ovary and testis and functions to promote the growth of germ cells [34]. To derive OFC3, G418 selection was applied to Tg(gsdf:neo) ovarian primary cultures to isolate the Gsdf-expressing cells. Since OFC3 cells express Gsdf, it was expected that they may also produce other unknown factors that promote the growth and survival of germ cells in culture and therefore serve as an ideal feeder layer for spermatogonia. To optimize the spermatogonial cell culture conditions, OFC3 was transfected to express zebrafish Lif and Fgf2 (OFC3LF) or Gdnf (OFC3LG). Recombinant LIF, FGF2 and GDNF have all been shown to be important factors in establishing mouse SSC cultures [22,35]. Previously, we demonstrated that optimal growth of zebrafish FGSCs was achieved using the OFC3LF and OFC3LG feeder layers indicating that the cells produce factors that support zebrafish germ cells in culture [7]. A previously reported zebrafish spermatogonial cell culture system employed a feeder cell line that was derived from a spontaneously arising zebrafish testicular tumor [11]. Although we made several attempts to establish a continuously-growing G418 resistant feeder cell line initiated from healthy Tg(gsdf:neo) testicular tissue, we were not successful.
Using the drug selection strategy and culture conditions that were developed in this study, spermatogonia could be propagated for several weeks in vitro and then transplanted into infertile recipient larvae where the cells colonized the gonads and generated fertile adult fish at a high frequency. Since recipient larvae were treated as embryos with a MO that blocked endogenous germ cell formation and gonad development resulting exclusively in the production of infertile male [23], the transplanted spermatogonia were able to direct the formation of a functional gonad and rescue fertility of the host fish. Although only approximately 60 spermatogonia were transplanted into each recipient, the fertile adult fish that resulted from the transplantations were able to produce multiple batches of offspring over a 6 month period. These results demonstrate that the transplanted cells re-established the entire spermatogonial cell lineage in the host testis and therefore possessed germline stem cell qualities that are characteristics of SSCs. Only male fertile recipients were produced from successful germline transmission of transplanted spermatogonial cells. Using germ cell ablation, a recent zebrafish study has suggested that signals from female germ cells, particularly the oocytes, may be needed to promote primary female sex determination and maintain female development throughout adulthood [36]. It is most likely that the cultured spermatogonial cells did not produce these signals that are required for female development. As such, all the fertile recipients completely developed into males. The spermatogonial culture system described in this study will provide an in vitro model for studies of extracellular factors controlling zebrafish spermatogonial differentiation and may form the basis of gene transfer approach.