|Home | About | Journals | Submit | Contact Us | Français|
Loss-of-function mutation of the Kit gene causes a severe defect in spermatogenesis that results in infertility due to the inability of its cognate ligand, KIT ligand (KITL), to stimulate spermatogonial proliferation and differentiation. Although self-renewal of mouse spermatogonial stem cells (SSCs) depends on glial cell line-derived neurotrophic factor (GDNF), there is no unequivocal evidence that SSCs with a KIT deficiency can self-renew in vivo or in vitro. In the testis of Wv/Wv mice, in which the KIT tyrosine kinase activity is impaired, spermatogonia with SSC phenotype were identified. When Wv/Wv spermatogonia were cultured in an SSC culture system supplemented with GDNF in a 10% O2 atmosphere, they formed clumps and proliferated continuously. An atmosphere of 10% O2 was better than 21% O2 to support SSC self-renewal. When Wv/Wv clump-forming germ cells were transplanted into testes of infertile wild-type busulfan-treated mice, they colonized the seminiferous tubules but did not differentiate. However, when transplanted into the testes of infertile W/Wv pups, they restored spermatogenesis and produced spermatozoa, and progeny were generated using microinsemination. These results clearly show that SSCs exist in Wv/Wv testes and that they proliferate in vitro similar to wild-type SSCs, indicating that a functional KIT protein is not required for SSC self-renewal. Furthermore, the results indicate that a defect of KIT/KITL signaling of Wv/Wv SSCs does not prevent spermatogonial differentiation and spermatogenesis in some recipient strains.
There are many self-renewing systems supported by stem cells in adult animals, and spermatogenesis is one of the most productive among them . The high productivity depends on spermatogonial stem cells (SSCs) that self-renew and produce daughter spermatogonia that proliferate, undergo meiosis, and morphologically differentiate into sperm. In rodents, SSCs are a small fraction of the most immature type A spermatogonia that consist of Asingle, Apaired, and Aaligned spermatogonia, which are defined primarily by position and cellular association within the seminiferous tubules [2, 3]. Although Asingle spermatogonia are believed to be SSCs, unequivocal identification of SSCs requires a functional transplantation assay. When germ cells are transplanted into seminiferous tubules of infertile mice, only SSCs can colonize the basement membrane and regenerate long-term complete spermatogenesis [4–6]. Strict control of self-renewal and differentiation of SSCs is crucial to maintain continuous spermatogenesis, and elucidation of the regulatory mechanism is a key question in SSC biology. Recently, we developed an in vitro culture system for SSCs using a serum-free defined medium that led to identification of the primary growth factor supporting self-renewal and proliferation of mouse SSCs as glial cell line-derived neurotrophic factor (GDNF) [7, 8]. Our in vitro studies and earlier in vivo studies [9–11] using transgenic mice and germ cell transplantation have established that GDNF is the self-renewal promoting factor for rodent SSCs.
The receptor tyrosine kinase KIT and its ligand, KIT ligand (KITL, also referred to as stem cell factor), have crucial roles in melanogenesis, hematopoiesis, and gametogenesis . In these systems, stem/progenitor cells express KIT, while supporting cells in the surrounding microenvironments express KITL . The KIT/KITL system was believed to be critical for spermatogenesis and appears particularly important in the process of spermatogonial proliferation and differentiation . KIT expression in the adult testis is detectable by immunohistochemical analysis and in situ hybridization in differentiating type A (A1–A4), intermediate, and type B spermatogonia, as well as preleptotene spermatocytes and interstitial Leydig cells, but not in Asingle, Apaired, and Aaligned spermatogonia and Sertoli cells [14–16]. Intravenous injection of anti-KIT monoclonal antibody ACK2, which blocks KIT function, into adult mice caused a depletion of differentiating type A (A1–A4) spermatogonia but did not affect Asingle, Apaired, and Aaligned spermatogonia, suggesting that KIT/KITL signaling mechanisms have little or no role in SSC function . In addition, studies [17, 18] using transplantation experiments in conjunction with fluorescent-activated cell sorting (FACS) for cell fractionation clearly demonstrated that SSCs were enriched exclusively in the KIT-negative cell fraction. Although it has been established that there is little or no KIT expression on mouse SSCs [8, 17, 18], these studies do not directly prove that KIT is not required, perhaps at low levels, for SSC self-renewal. To demonstrate unequivocally that the KIT/KITL system is not required for SSC self-renewal in the mouse, studies using mutant animals defective for KIT or KITL function are necessary.
The Kit and Kitl genes are encoded by the dominant white spotting (W) and steel (Sl) loci in the mouse, respectively [19, 20]. Many mutations, including point mutations and deletions that result in loss of function of W or Sl, have been reported [12, 13]. The original W mutant has a deletion in the KIT transmembrane domain that results in no surface expression of the KIT protein, while Wv mutants have a point mutation in the tyrosine kinase domain of KIT that impairs signaling activity. W/W mice die during the perinatal or late fetal stage of development, whereas Wv/Wv mice are viable but are characterized by mild macrocytic anemia, loss of coat pigmentation, and sterility. The original Sl mutation is missing the entire Sl gene, and the Sl/Sl mutant animals are embryonic lethal. In contrast, the Sld mutation deletes the transmembrane and cytoplasmic domains of KITL protein, resulting in the ability to produce only soluble KITL. The Sld/Sld mice are viable and show a phenotype similar to the Wv/Wv mice, including sterility in males.
KIT and KITL are also necessary for efficient proliferation and migration of primordial germ cells (PGCs) during embryogenesis [12, 19, 20]. The number of PGCs in mice with loss-of-function mutations in KIT or KITL is dramatically reduced; however, a few germ cells can survive and colonize the embryonic gonad. In normal male gametogenesis, PGCs stop dividing in the seminiferous tubules of fetal testes and become gonocytes. Shortly after birth, the gonocytes resume proliferation and migrate to the basement membrane of the seminiferous tubules. Some of these gonocytes differentiate into SSCs, while others become non-stem type A spermatogonia that do not have the ability to self-renew, which then initiate the first wave of spermatogenesis . In the Wv or Sld mutant mice, surviving gonocytes develop into type A spermatogonia, which can be identified in the postnatal testis . Transplantation experiments of Sl/Sld testis cells expressing LacZ into infertile W or busulfan-treated recipients demonstrate that SSCs were present in Sl/Sld mice, indicating that membrane KITL is not required for SSC maintenance [23, 24]. A study using transplantation of green fluorescent protein-expressing W/Wv testis cells into W/Wv recipient mice testis demonstrated that the donor W/Wv spermatogonia settled and proliferated in the recipient seminiferous tubules . Although the findings of that study suggested the existence of SSCs in W/Wv testes, it was unclear whether the donor germ cells colonized were functional SSCs, as colonized germ cells did not differentiate to generate spermatozoa.
Previous studies demonstrated a critical role for the KIT/KITL system in differentiating spermatogonia, spermatocytes, and Leydig cells but none or a small role in Asingle, Apaired, and Aaligned spermatogonia . However, those investigations used indirect experimental techniques and cannot exclude a small but critical function of the KIT/KITL signaling in self-renewal of SSCs. The possibility of such a role is suggested by the importance of the KIT/KITL system for earlier developmental stages of germ cell differentiation. Therefore, the objectives of this study were to determine if Kit mutants contain SSCs that are able to self-renew and proliferate in a manner similar to wild-type SSCs and to identify a suitable transplant recipient that would allow for production of spermatozoa from putative SSCs carrying a mutant Kit gene. To complete these objectives, we used Wv/Wv testis cell preparations to determine whether the SSCs would proliferate in vitro without testis somatic cell support and, if so, whether they would undergo complete spermatogenesis when transplanted into recipient testes. The results of this study confirm that SSCs exist in Wv/Wv testes and can proliferate in culture and that they can contribute to all stages of spermatogenesis when transplanted into suitable recipients.
For experiments to investigate effects of O2 concentration on in vitro SSC self-renewal, donor cells were isolated from B6.129S7-Gt(ROSA)26Sor (designated C57BL/6 ROSA; The Jackson Laboratory), which express the Escherichia coli lacZ gene and thus express β-galactosidase (β-gal) in virtually all cell types, including germ cells . Donor C57BL/6 ROSA cells can be visualized by staining with the β-gal substrate 5-bromo-4-choloro-3-indolyl β-D-galactoside (X-gal). Recipient male mice were immunologically compatible C57BL/6 × 129/SvCP adult mice, which were treated with 60 mg/kg busulfan 4–6 wk before use [4, 5]. For experiments to evaluate SSC activity of Wv/Wv germ cells, β-gal-expressing Wv/Wv mice were used. To obtain Wv/Wv mice carrying the β-gal transgene of C57BL/6 ROSA (designated Wv/Wv ROSA), Wv/+ mice (background C57BL/6) were mated to C57BL/6 ROSA mice. β-gal-expressing Wv/+ male and female mice were selected by X-gal staining of tail snips, and the Wv/+ ROSA males and females were mated to obtain Wv/Wv ROSA male mice. Several types of recipient mice were used for transplantation of Wv/Wv germ cells, including busulfan-treated immunologically compatible C57BL/6 and FAS ligand (FASL)-deficient adult mice (B6Smn.C3-Faslgld/J, designated B6-gld; The Jackson Laboratory, Bar Harbor, ME), which were treated with 44 mg/kg busulfan, and congenitally infertile W54/Wv  and W/Wv pups (6–10 days postpartum).
Culture of wild-type mouse SSCs was performed as described previously , and a detailed protocol is available elsewhere . Briefly, THY1+ testis cells enriched for SSCs were isolated from C57BL/6 ROSA mouse pups (5 days postpartum) using magnetic-activated cell sorting (MACS; Miltenyi Biotec, Auburn, CA) and cultured in 12-well plates with mitotically inactivated STO (SIM mouse embryo-derived thioguanine- and ouabain-resistant) cell feeders (SNL76/7; gift from Dr. A. Bradley, The Wellcome Trust Sangar Institute, London, UK) using a serum-free medium (SFM) supplemented with 20 ng/ml GDNF (R&D Systems, Minneapolis, MN), 150 ng/ml GDNF family receptor α1 (GFRA1; R&D Systems), and 1 ng/ml basic fibroblast growth factor (FGF2; BD Biosciences, San Jose, CA). The SFM consisted of minimum essential medium α (Invitrogen, Carlsbad, CA) with 0.2% bovine serum albumin (194774, lot No. R14550; MP Biomedicals, Solon, OH ), 5 μg/ml insulin (Sigma-Aldrich, St. Louis, MO), 10 μg/ml iron-saturated transferrin (Sigma-Aldrich), 3 × 10−8 M Na2SeO3 (Sigma-Aldrich), 60 μM putrescine (Sigma-Aldrich), 7.6 μEq/L free fatty acid mixture (Sigma-Aldrich), 50 μM 2-mercaptoethanol (Sigma-Aldrich), 2 mM l-glutamine (Invitrogen), 10 mM Hepes (Sigma-Aldrich), 50 units/ml penicillin (Invitrogen), and 50 μg/ml streptomycin (Invitrogen). For SSC culture from Wv/Wv mice, testis cells were prepared from 2-wk-old Wv/Wv male mice. The isolated testis cells were cultured in SFM on STO feeder cells without enrichment for THY1+ cells. Normal humidified incubators with 5% CO2 were used for the 21% O2 culture condition. For low oxygen cultures, plates were placed into gas-tight modular incubator chambers (Billups-Rothenberg, Del Mar, CA) that were gassed with mixture of one part 21% O2/5% CO2/balance N2 and one part 5% CO2/balance N2 to produce a 10% O2 culture condition.
Cultured germ cells were suspended in SFM and transplanted into recipient testes via the efferent tubules. Wild-type germ cells (10 μl of 1.25 × 106 cells/ml) were transplanted into the testes of busulfan-treated immunologically compatible C57BL/6 × 129/SvCP male mice [4, 5]. Cultured Wv/Wv ROSA cells were transplanted into C57BL/6 and B6-gld adult male mice, as well as W54/Wv and W/Wv pup testes. Adult testes received approximately 10 μl of donor cell suspension (30 × 106 cells/ml), resulting in 70%–80% filling of the tubules , whereas W54/Wv and W/Wv pup testes received approximately 2 μl of donor cell suspension (75–91 × 106 cells/ml) . The Institutional Animal Care and Use Committee of the University of Pennsylvania approved all experimental procedures.
Recipient testes were collected at least 2 mo after donor cell transplantation and analyzed by X-gal staining [4, 5]. The number of donor-derived spermatogenic colonies from testes receiving wild-type SSCs was counted using a dissection microscope, and colonies per 105 cells cultured were determined.
Selected transplanted testes stained with X-gal and testes from 2- and 8-wk-old Wv/+ and Wv/Wv mice were evaluated histologically to determine the extent of spermatogenesis. Briefly, testis tissues were fixed in formalin or Bouin, washed in ethanol, embedded in paraffin, and sectioned. Sections were observed without staining or after staining with hematoxylin-eosin or nuclear fast red.
Cell staining procedures for flow cytometry of freshly isolated or cultured testis cells were described previously [7, 8, 18]. Antibodies used for surface antigens were biotin-conjugated anti-THY1 (53.2.1, 1:100 dilution; BD Biosciences, San Jose, CA), allophycocyanin-conjugated anti-ITGA6 (GoH3, 1:200 dilution; BD Biosciences), and R-phycoerythrin-conjugated anti-ITGAV (RMV-7, 1:200 dilution; BD Biosciences). Alexa Fluor 488-conjugated streptavidin (Invitrogen, Carlsbad, CA) was used as the secondary reagent for biotin-conjugated anti-THY1 antibody. Stained cells were analyzed using a FACSCalibur (BD Biosciences) cytometer.
Recipient W/Wv mice transplanted with cultured Wv/Wv ROSA cells were euthanized 3–6 mo after transplantation, and testes were removed, detunicated, and quartered. The tissue was chopped, washed, and filtered over a 40-μm nylon mesh. Spermatozoa or nuclei from round spermatids were microinjected into oocytes derived from C57BL/6 × DBA/2 F1 mice using a piezoelectric actuator (PrimeTech, Ibaraki, Japan) . Injected oocytes were cultured overnight and transferred to the oviducts of Day 1 pseudopregnant ICR female mice. Fetuses were retrieved on Day 19 or were allowed to be born and nursed by foster mothers.
All data are presented as the mean ± SEM. Data were analyzed by one-way ANOVA. Statistical difference among means was determined using Student t-test.
Differentiating germ cells are not present in the seminiferous tubules of Wv/Wv mouse testes because of impairment of the KIT/KITL signaling system (Fig. 1A). Previous work using a histological approach has demonstrated that undifferentiated spermatogonia were identified in W/Wv testes . However, it was unclear if these cells express antigenic markers characteristic of SSCs. To determine whether germ cells in Wv/Wv testes express surface antigens characteristic of SSCs (THY1+ ITGA6+ ITGAV−), flow cytometric analysis was performed on cell populations from 2-wk-old Wv/+ and Wv/Wv testes. Approximately 8% and 1.5% of cells from Wv/+ and Wv/Wv testes were THY1+ ITGA6+ ITGAV−, respectively (Fig. 1B). These results suggested the existence of an SSC population in Wv/Wv testes. Wild-type mouse SSCs self-renew and proliferate as tightly packed cell clumps consisting of THY1+ ITGA6+ ITGAV− germ cells when cultured in a serum-free defined medium supplemented with GDNF, soluble GFRA1, and FGF2 . Thus, we attempted to culture the putative SSCs of Wv/Wv mice. Because O2 levels have been shown to affect growth of other cell types, we cultured the Wv/Wv germ cells in 10% and 21% O2 conditions. Testis cells prepared from 2-wk-old Wv/Wv mice were placed on STO feeder cells in an SFM-containing GDNF, soluble GFRA1, and FGF2. Because the cell number obtained from the Wv/Wv testes was low, it was impossible to enrich a cell fraction for SSCs using MACS. Our previous studies [7, 11] demonstrated that testicular fibroblasts had a detrimental effect on SSC self-renewal and tend to take over germ cell cultures; therefore, removal of the fibroblasts and avoidance of their outgrowth were critical for establishing continuous proliferation of SSCs in culture . To eliminate testicular fibroblasts from the cultures, weakly attached germ cells were removed 2 days after seeding and subcultured onto fresh STO feeders. Eight days after subculture, a few germ cell clumps appeared (Fig. 2A). The size of germ cell clumps appeared to be larger in 10% O2 compared with 21% O2. After 5 wk in culture, the number of germ cells appeared higher in 10% O2, and flow cytometric analysis confirmed expansion of THY1+ ITGA6+ ITGAV− germ cells (Fig. 2, A and B). The number of THY1+ ITGA6+ ITGAV− cells cultured for 5 wk in 10% and 21% O2 conditions was determined using flow cytometry. In 10% O2, the number of THY1+ ITGA6+ ITGAV− cells increased from 2.0 ± 0.22 × 104 (n = 3) to 200 ± 3.2 × 104 (n = 3). In contrast, the number of THY1+ ITGA6+ ITGAV− cells cultured in 21% O2 did not increase (2.0 ± 0.22 × 104 before culture and 2.1 ± 0.13 × 104 after 5 wk of culture [n = 3]). While the flow cytometric analysis showed no difference in the cell surface phenotype of cultured cells in the two conditions, the number of THY1+ ITGA6+ ITGAV− cells in 10% O2 was almost 100-fold higher than that in 21% O2 after 5 wk of culture, indicating that the clump-forming germ cells doubled every 5.3 days (35/log2100) in a 10% O2 environment. This rate is similar to the in vitro doubling time of wild-type SSCs (5.6 days), which was calculated based on transplantation assay . Although transplantation assays were not performed, it appeared that THY1+ ITGA6+ ITGAV− Wv/Wv germ cells proliferated at a doubling time similar to that of wild-type SSCs in vitro. Moreover, the cells continued to proliferate over 6 mo, strongly suggesting the presence of SSCs within the culture. These results demonstrated that THY1+ ITGA6+ ITGAV− Wv/Wv clump-forming germ cells were able to proliferate continuously in vitro using the same growth factors for self-renewal as wild-type SSCs.
To confirm that the effect of 10% O2 concentration on Wv/Wv clump-forming germ cells had a similar effect on wild-type SSCs within the culture, SSC cultures established from C57BL/6 ROSA mouse pup testes maintained in a 21% O2 atmosphere were subcultured in 10% and 21% O2 concentrations. The SSCs seemed to proliferate in both conditions; however, after 15 days, it appeared that SSCs cultured in 10% O2 formed larger clumps than those in 21% O2 (Fig. 3A). To determine the effect of O2 concentration on SSC number, cultured germ cells were transplanted into recipient mouse testes 15 days after subculture to the different O2 conditions. Only SSCs are able to colonize the seminiferous tubules and generate individual colonies of complete spermatogenesis [31–33], and each colony of donor-derived spermatogenesis represents the clonal expansion of a single SSC [31–33]. Therefore, germ cell transplantation analysis provided a direct examination of both total SSC number and SSC concentration within the two treatments. Quantification of donor-derived colonies generated from the transplanted cells identified by X-gal staining (Fig. 3B) indicated that the number of stem cells was greater in 10% O2 (Fig. 3C). Using the same cell population that was transplanted, the concentration of SSCs within the clump-forming germ cell population was determined. The concentration of SSCs was calculated by applying the percentage of clump-forming germ cells that was identified as THY1+ ITGA6+ ITGAV− cells by flow cytometry to the total number of SSCs determined by transplantation. After 2 wk of culture in 10% O2 and 21% O2, the numbers of spermatogenic colonies generated per 105 THY1+ ITGA6+ ITGAV− cells transplanted into individual recipient testes were 225 ± 28 and 236 ± 34 (n = 16), respectively. After 6.5 wk of culture in 10% O2 and 21% O2, the numbers of spermatogenic colonies generated per 105 THY1+ ITGA6+ ITGAV− cells transplanted were 245 ± 35 and 233 ± 38 (n = 16), respectively. No significant difference in stem cell activity was found between the two groups at any time point. Thus, the concentration of SSCs within the clump-forming germ cell population was not different between the two culture conditions at least for 6.5 wk. Moreover, low O2 concentration did not alter the surface antigenic phenotype of cultured SSCs as seen in Wv/Wv clump-forming germ cells (Fig. 3D). These results indicate that wild-type clump-forming germ cells, including SSCs, grow better in a 10% O2 culture condition.
The characteristics of continuously proliferating Wv/Wv germ cells in vitro were similar to those of wild-type clump-forming germ cells, which are enriched for SSCs, suggesting that they might have SSC potential, which can be confirmed only by transplantation into recipient testes and generation of complete spermatogenesis. Therefore, we sought a host to support the differentiation of Wv/Wv clump-forming germ cells into spermatozoa. FAS is a surface molecule that belongs to the tumor necrosis factor receptor family and can induce apoptosis when bound by FASL . Ten percent of seminiferous tubules of FAS-deficient Wv/Wv mice contained spermatocytes, spermatids, and spermatozoa , suggesting that the lack of differentiating germ cells in KIT/KITL-deficient mice is partly due to FAS-dependent germ cell apoptosis. In the B6-gld mice, Fas-expressing cells cannot receive the death signal via ligation with FASL; therefore, it is possible that when transplanted into FASL-deficient B6-gld recipients Wv/Wv germ cells might colonize and undergo spermatogenesis. We also used W54/Wv or W/Wv pup testes as recipients for Wv/Wv germ cells, as these mice provide one of the most suitable environments for colonization by wild-type SSCs . Although complete spermatogenesis cannot be seen in the testes of W54/Wv or W/Wv adult mice, absence of spermatogenesis in these mice may result from decreased colonization of the genital ridge by PGCs and thus fewer SSCs. Therefore, we also hypothesized that transplanting many Wv/Wv germ cells might result in some level of spermatogenesis.
To determine the colonization ability of cultured Wv/Wv germ cells, we transplanted Wv/Wv ROSA clump-forming germ cells into busulfan-treated C57BL/6 and B6-gld adult or W54/Wv and W/Wv pup testes. Two months after transplantation, C57BL/6 and B6-gld recipient testes were stained with X-gal; only spermatogonia were observed, and no colonies appeared to have differentiating germ cells (Fig. 4A). W54/Wv recipient testes showed predominantly the same staining pattern as wild-type C57BL/6 recipients, but occasionally colonies with a more intense staining pattern indicative of more advanced spermatogenesis were observed (Fig. 4B, three of 21 testes examined). Histological analyses did not detect complete spermatogenesis, but it appeared that more germ cells were present (Fig. 4, D and E). Surprisingly, when W/Wv recipient pup testes were used, many dark blue colonies were identified in the seminiferous tubules following X-gal staining 2 mo after transplantation (Fig. 4C, eight of eight testes examined). This unexpected result suggested that Wv/Wv ROSA clump-forming germ cells differentiated and generated spermatozoa in the W/Wv seminiferous tubules, which was confirmed histologically (Fig. 4F). Moreover, when examined 3–5 mo after transplantation, six of seven testes contained dark blue colonies, suggesting that the donor SSCs were undergoing self-renewal in the recipient testes. One of seven testes analyzed stained only weakly, and many faint colonies were identified in the recipient testis. These transplantation experiments were repeated using three independent Wv/Wv ROSA germ cell cultures.
To examine whether the Wv/Wv spermatozoa generated in the W/Wv seminiferous tubules were functionally normal, in vitro microinsemination was performed to generate offspring . A testis cell suspension was prepared from the W/Wv recipient testes in which spermatozoa were identified (Fig. 4G). After microinsemination and transfer of embryos, three offspring were born; two were cannibalized at birth, but one male survived and was fertile. This male had the characteristic Wv/+ phenotype, including a white belly (Fig. 4H), and expressed LacZ (data not shown), which clearly indicate the male was from spermatozoa arising from the cultured Wv/Wv ROSA SSCs that were transplanted to the W/Wv recipient. When the male was mated to a Wv/+ female, Wv/Wv pups (white coat [Fig. 4I]), Wv/+ pups (black coat with a white spot [Fig. 4I]), and +/+ pups (black coat [data not shown]) of both sexes were generated.
Two questions were addressed in the present study. The first question was whether spermatogonia of Wv/Wv mice with the SSC phenotype can expand continuously in a culture system that supports self-renewal and proliferation of wild-type SSCs. The second question was whether these cultured Wv/Wv germ cells can differentiate to fully functional spermatozoa to prove definitively that the proliferating cells contain SSCs. We clearly demonstrated that THY1+ ITGA6+ ITGAV− Wv/Wv spermatogonia contain SSCs that have the ability to expand in vitro continuously in the presence of GDNF and generate functional spermatozoa following transplantation into an appropriate recipient. To our knowledge, this is the first demonstration that Wv/Wv SSCs are phenotypically similar to wild-type SSCs based on antigenic profile and characteristic culture attributes. These results demonstrate that mouse SSCs do not require activation of the KIT signaling pathway for self-renewal. Furthermore, the data indicate that in an appropriate recipient testis Wv/Wv SSCs undergo complete spermatogenesis.
Wv/Wv mice possess a point mutation in the Kit gene, in which a threonine at position 660 is replaced by a methionine [36, 37]. The mutated position of the Wv allele is in a highly conserved subdomain that involves ATP binding of protein tyrosine kinases. Although the mutant protein is expressed on the cell surface, the kinase activity is impaired, and homozygous mutant cells do not respond to KITL [36, 37]. Although KIT signals are required for proliferation and migration of PGCs during embryonic development, a few male germ cells in W/Wv mouse testes survive until birth and some into adulthood . A previous study  showed that surviving spermatogonia in mature W/Wv mice slowly proliferate in the seminiferous tubules. However, there was no evidence that the proliferating cells are functional SSCs that can contribute to all stages of spermatogenesis. In addition, it is unclear whether the proliferation of W/Wv spermatogonia is specific to the mouse strain or if spermatogonia from other KIT-deficient mice (such as Wv/Wv or W54/Wv) also slowly proliferate in the testis. Moreover, because the KIT receptor is involved in proliferation of several stem or progenitor cells in vivo and in vitro , the lack of the signaling pathway might prevent self-renewal of SSCs in culture. Faint expression of KIT protein and transcript has been observed on a subpopulation of cultured wild-type SSCs [8, 38], and STO feeder cells used in culture express KITL ; therefore, we could not exclude a small but critical role of the KIT/KITL signaling pathway in SSC self-renewal. We found that cultures of clump-forming germ cells established from Wv/Wv mouse testes continuously proliferated over 6 mo. Moreover, although KIT activation is required to proceed through complete spermatogenesis, when Wv/Wv clump-forming germ cells were transplanted into W/Wv testes, colonies of spermatogenesis containing functional spermatozoa were produced.
The mechanism that supports Wv/Wv SSC-derived spermatogenesis in W/Wv recipient testes, but not C57BL/6, B6-gld, and poorly in W54/Wv testes, is unclear. Seminiferous tubules of KIT-deficient pup testes provide a superior microenvironment for SSC colonization compared with those of busulfan-treated or KIT-deficient adult testes . Thus, W/Wv and W54/Wv pups used in the present study should be equally hospitable for transplanted SSCs. Finding the remarkable difference between W/Wv and W54/Wv mice as recipients for Wv/Wv SSCs was unexpected. One possibility is that W/Wv Sertoli cells provide a stimulatory signal to Wv/Wv germ cells to bypass a critical step involving KIT/KITL signaling. The W54/Wv testis may also provide a stimulatory signal, but it might be compromised by the difference in the genetic background. The genetic backgrounds of Wv, W, and W54 are C57BL/6, WB/ReJ, and 129/SvCP, respectively. 129/Sv mice are prone to generate testicular germ cells tumors , and they appear to show more severe phenotypes in spermatogenic defects in some gene-targeted mice [41, 42]. Therefore, W54/Wv might not be as ideal a recipient for Wv/Wv SSCs as W/Wv because of contributions from the 129/SvCP genetic background.
To date, there are two studies demonstrating complete spermatogenesis derived from Wv/Wv SSCs. In each case, the SSCs have another genetic defect in apoptotic signaling pathways in addition to the KIT deficiency. In one study , FAS-deficient Wv/Wv SSCs (C57BL/6 genetic background) were used; in the other study , TRP53-deficient Wv/Wv SSCs (CBA and 129/Ola mixed genetic background) were used. Wv/Wv SSCs with FAS or TRP53 deficiency could complete spermatogenesis, suggesting that one of the roles of the KIT/KITL system in spermatogenesis is to block death signals during spermatogonial differentiation. In our study, the Wv/Wv SSCs do not have such a second genetic defect; therefore, W/Wv Sertoli cells must provide factors or an environment that prevents apoptosis of the Wv/Wv germ cells. Understanding this mechanism would help to elucidate the signals regulating apoptosis during spermatogenesis. Just as remarkable as the ability of the Wv/Wv stem cells to undergo in vitro self-renewal is the ability of type A and later-stage spermatogonia to undergo marked proliferation and differentiation, a process that has been regarded as critically dependent on KIT/KITL signaling .
We found that 10% O2 atmosphere was more suitable for SSC self-renewal. Low oxygen conditions had significant benefits on self-renewal of various types of stem cells, including embryonic stem cells [44, 45], and findings from studies [46, 47] suggest that oxidative stress has detrimental effects on cultured cells. Therefore, we examined the effect of low O2 concentrations on in vitro Wv/Wv and wild-type SSC proliferation. Our results indicated that 10% O2 was better than 21% O2 for in vitro SSC proliferation. Thus, like some other stem cells, lower oxygen concentrations appeared to have a beneficial effect for in vitro culture of SSCs. When normal embryonic fibroblasts were cultured in low oxygen conditions, less DNA damage was accumulated in the cultured cells compared with those in 21% O2 . Findings from a recent study  suggested that excessive accumulation of DNA damage led to activation of a cell cycle inhibitor, p21Cip1/Waf1, resulting in a self-renewal defect of SSCs. Proliferating SSCs in 10% O2 may have less DNA damage than in 21% O2.
Previous work has demonstrated that SSC transplantation or gene therapy can rescue defective spermatogenesis using W and/or Sl mutant mice. Transplantation of Sl/Sld SSCs into W recipient testes resulted in restoration of spermatogenesis and generation of progeny of the Sl mouse  because of the ability of the W testis to produce KITL. In addition, when a normal KITL gene was introduced into Sertoli cells of Sl/Sld mutant males using viral vectors or electroporation [49–51], spermatogenesis was restored. In a third study , Sertoli cells from the W mouse, in which KITL production is normal, were transplanted into Sl/Sld mouse testes, again resulting in restoration of mutant spermatogenesis. However, all of these studies corrected a mutation causing a loss of KITL production, which is intrinsic to the somatic Sertoli cells. The SSCs and their progeny do not express KITL; therefore, in these three studies, it was unnecessary to manipulate SSCs themselves to obtain functional spermatozoa. In contrast, there are no reports in which a mutation that causes a direct defect in the SSC itself has been overcome. The results of the present experiments indicate that it is possible to expand defective SSCs in culture and to generate functional gametes following transplantation to a suitable recipient.
Elucidation of the mechanisms governing fate determination of SSCs is important for understanding the basic biology and potential clinical uses of SSCs [53, 54]. The present study demonstrates that the KIT/KITL signaling pathway is not required for self-renewal of SSCs and indicates, contrary to previous belief, that functional KIT is not an absolute requirement for differentiation of germ cells. To our knowledge, this study is the first report of the use of germ cell transplantation to a specific recipient to overcome a mutation directly related to spermatogonial maturation. Although the mechanism regulating this functional recovery remains unknown, this work provides new insight for further understanding the mechanism controlling SSC fate decision and could serve as a model for the development of techniques to correct a genetic defect in the germline.
We thank C. Freeman and R. Naroznowski for assistance with animal maintenance and experimentation, and J. Hayden for photography.
1Supported by the National Institute of Child Health and Human Development ( HD 044445 and HD 052728) and the Robert J. Kleberg, Jr. and Helen C. Kleberg Foundation.