Self renewal, expansion, and transfection of rat spermatogonial stem cells in culture

Abstract

The use of a transgenic line of rats that express enhanced GFP (EGFP) exclusively in the germ line has allowed a separation of feeder layers and contaminating testis somatic cells from germ cells and the identification of a set of spermatogonial stem cell marker transcripts. With these molecular markers as a guide, we have now devised culture conditions where rat spermatogonial stem cells renew and proliferate in culture with a doubling time between 3 and 4 days. The marker transcripts increase in relative abundance as a function of time in culture, and the stem cells retain competency to colonize and develop into spermatids after transplantation to the testes of recipient rats. The cells also remain euploid after at least 12 passages. Cell lines could be isolated and cryopreserved and, upon subsequent thawing, continue to self renew. Transfection of the spermatogonial stem cells with a plasmid containing the neomycin phosphotransferase (neo) selectable marker resulted in selection of G418-resistant cell lines that effectively colonize recipient testes, suggesting that gene targeting is now feasible in the rat.

The laboratory rat represents one of the most comprehensively studied mammalian species, with described use in >1 million publications in a wide range of medically relevant areas. Qualities such as size, fecundity, behavior, ease of surgical techniques, tissue sampling, and general laboratory management have contributed to its popularity (1-3). However, a failure to develop technology to produce rat genetic models through gene targeting has resulted in the mouse becoming a widely popular animal model.

Although mouse embryonic stem (ES) cells renew with a sense of immortality, primitive hemapoietic stem cells self renew ineffectively and for only a short period in vitro (4, 5). ES cells from species other than the mouse or human fail to generally self renew effectively and also lose pluripotency in culture (6). We and others have not succeeded in culturing pluripotent ES cells from the rat; however, if spermatogonial stem cells could be cultured under conditions where they self renewed and expanded in numbers, conceivably they could also be genetically modified in vitro in much the same manner as seen with mouse ES cells. This alternative to the use of genetically modified ES cells would result in direct germ line transfer and an escape from the intervening formation of a mosaic animal. After appropriate selection of gene-targeted cells in culture, the chosen spermatogonial stem cells could either be induced to differentiate to the haploid stage in vitro or transplanted to the testes of recipient rats to allow development to the haploid stage. In either case, intracytoplasmic sperm injection into the egg would result in transmission of the genetically modified information (7, 8).

We now show that genetically marked spermatogonial stem cells from transgenic germ-cell-specific EGFP (GCS-EGFP) rats (9) can be placed under defined culture conditions, where they then subsequently renew, possibly indefinitely. Cell lines can be obtained from the cultures, cryopreserved, and upon thawing, continue to self renew with retention of competence to reconstitute spermatogenesis in a recipient testis. After 12 passages, molecular markers for spermatogonial stem cells are enriched, and the cells continue to bind to laminin (a characteristic of spermatogonial stem cells) and to express the deleted azoospermia-like (DAZL) protein and the integrated GCS-EGFP transgene.

Transfection of the spermatogonial stem cells with a DNA construct containing a neomycin phosphotransferase (neo) selection cassette results in the collection of stem cells now resistant to G418; these resistant colonies continue to self renew and retain the ability to colonize a recipient testis. Transfer of this technology to other animals, including the human, could result in corrections of certain forms of male infertility, the preservation of an individual's germ line, and the potential use of these cells for the generation of pluripotent stem cells (10), which would eliminate a need to use ES cells to derive specialized cell types.

Materials and Methods

Materials, chemicals, DNA constructs, and methods to purify, characterize, and transplant germ cells from culture are presented in detail in Supporting Text, which is published as supporting information on the PNAS web site, and in our previous reports on germ cell cultures (11, 12).

Testicular Cell Cultures. Enriched populations of WT or GCS-EGFP laminin-binding (Lam_B) or -nonbinding (Lam_NB) germ cells were isolated by our previously established methods (11, 12). Freshly isolated Lam_B and Lam_NB germ cells are referred to as day 0 (D0) cultures. D0 Lam_B germ cells are highly enriched in spermatogonial stem cell activity and contain >90% type A spermatogonia and ≈5% somatic testis cells (11, 12). In contrast, D0 Lam_NB germ cells are depleted of stem cell activity and consist of ≈96% differentiated spermatogonia plus spermatocytes and only ≈4% type A spermatogonia (11, 12). We have previously defined culture conditions that maintain spermatogonial stem activity of the Lam_B population but that do not lead to expansion of germ cell numbers and include these descriptions in Supporting Text.

Transfection of Spermatogonia. D0 WT or GCS-EGFP Lam_B spermatogonia or cultures of expanding populations of GCS-EGFP⁺ spermatogonia were suspended to ≈1.25 × 10⁶ cells/ml in their respective culture medium. The transfection mixture containing Lipofectamine 2000 (Invitrogen) and plasmid DNA were prepared in Opti-MEM-I (Invitrogen) according to the manufacturer's protocol. The mixture (Lipofectamine/DNA = 1 μl/0.5 μg in 100 μl of Opti-MEM-I) was added to the cell suspension (20% vol mix/80% vol cell suspension) and incubated (37°C, 5% CO₂) for 40-120 min (routinely 80 min). During transfections lasting longer than 1 h, the cells were resuspended by gently pipetting them up and down two times midway through the incubation period. After the incubation period, the cells were washed by suspending them to 10 times the initial volume with fresh culture medium and then were pelleted for 5 min at 500 × g. The supernatant fluid was discarded, and the pellets were washed two additional times. After the third wash, the cell pellets were suspended in fresh medium from their respective culture conditions and then plated on fresh fibroblast feeder layers. Before transfecting the EGFP⁺ spermatogonia by this method, they were passaged onto gelatin-coated plates for a 30-min period to reduce the number of feeder cells present in the transfection.

Spermatogonial Stem Cell Index (SSCI). To determine SSCI values, the relative abundance of 10 rat SSCI transcripts (12) in GCS-EGFP⁺ germ cell samples was determined by real-time PCR. Real-time PCR analysis was performed by using the Applied Biosystems Prism 7900HT Sequence Detection System (13). Germ cell (GCS-EGFP⁺) samples were purified from different experimental cultures by FACS, and then total RNA was isolated from each sample by using the RNAqueous Micro kit (Ambion, Austin, TX). Purified RNA samples were quantified by using the RiboGreen RNA quantitation method (Invitrogen), and then 5-35 ng of each sample was reverse-transcribed to cDNA in a 20-μl reaction by using the SuperScript III First Strand Kit (Invitrogen). The cDNA samples were diluted 1:10 or 1:100, and then 5 μl was used for PCR by using the rat SSCI and 18S ribosomal RNA subunit primer sets (Table 1, which is published as supporting information on the PNAS web site). PCR was performed by using the SYBR Green PCR Master Mix (BioRad). The threshold cycle (C_T) generated by each primer set for SSCI marker transcripts was normalized to the C_T generated by the primer set specific for the 18S ribosomal RNA subunit by the following formula: Relative transcript abundance = 1/2^{(C_T of marker transcript - C_T of 18S RNA subunit)} (14, 15). Relative abundance values were normalized to 0.02 as a maximal value from each primer pair within a set of germ-cell samples. Each set of germ-cell samples analyzed included standard cDNA prepared from Lam_B (D0) and Lam_NB (D0) germ cells. SSCI values are equal to the mean of the relative abundance values determined for all 10 of the SSCI marker transcripts in a given germ cell sample.

Results

The generation of the GCS-EGFP rat line that expresses EGFP exclusively in the germ line (9) has greatly facilitated our research on male germ cells in culture (11, 12). However, in addition to using the EGFP signal to identify germ cells, we also used an antibody generated to rat DAZL in immunocytochemistry experiments to confirm that only EGFP⁺ cells were DAZL⁺ (11, 12).

In previous work, we determined that rat Lam_B spermatogonia cultured on the STO (SIM mouse embryo-derived thioguanine and ouabain resistant)-neomycin-LIF 76/7 (SNL) fibroblast line lost stem cell activity, whereas cells plated on the mouse Sertoli cell line-1 (MSC-1) retained stem cell activity as a function of time in culture (12). However, the germ cells in these cultures failed to proliferate after the first passage. Testicular transplantations as a means to determine whether cells in culture retain stem cell characteristics are tedious and require a considerable lag period before evaluation (16, 17). Therefore, we developed a group of molecular markers that appeared to serve as indicators of “stemness.” The marker transcripts were identified by microarray analysis of RNA from FACS-purified (GCS-EGFP⁺) Lam_B spermatogonia isolated after culture on SNL or MSC-1 feeders (12). We speculated that the Lam_B germ cell transcripts that declined in relative abundance on SNL feeders but that were maintained on MSC-1 feeders could serve as markers for spermatogonial “stemness” (12).

The relative abundance of each target transcript was measured by real-time PCR. Signals for target transcripts were normalized to a signal generated by amplification of transcripts encoding the rat 18S ribosomal RNA subunit (14, 15). The 18S ribosomal transcript is expressed at essentially the same relative abundance per ng of total RNA in either premeiotic (Lam_B) or meiotic (Lam_NB) rat germ cell populations (C_T = 24.1±.46, ± SD, n = 13 different primary cultures). Fig. 1 represents the relative abundance of 10 transcripts that mark spermatogonial stem cell activity in Lam_B germ cells after culture on SNL or MSC-1 cells. Consistent with microarray results (12), real-time PCR analysis established that these 10 transcripts are relatively abundant in D0 Lam_B germ cells and do not decline in abundance after maintenance for 13 days on MSC-1 feeders (Fig. 1 A). In contrast, the abundance of these same transcripts decreased markedly in Lam_B germ cells by 11 days after culture on SNL cells, acquiring a profile more comparable to Lam_NB germ cells (Fig. 1 A). The mean of the relative abundance values determined for these 10 marker transcripts yielded the SSCI (Fig. 1B).

Fig. 1.

The SSCI for rat germ cell cultures. (A) Relative abundance of 10 transcripts associated with spermatogonial stem cell activity was measured in cultures of GCS-EGFP⁺ Lam_B germ cells by real-time PCR. D0 Lam_B germ cells (D0 LB) were isolated fresh or purified from somatic feeder cells by FACS after 13 days on MSC1 cells (D13 MSC) or 11 days on SNL cells (D11 SNL). D0 Lam_NB germ cells (D0 LNB) were also analyzed. Rat gene symbols are listed for each transcript on the x axis. (B) Rat SSCI values for germ cell cultures analyzed in A. The SSCI values are equal to the mean of the relative abundance of the transcripts in A (±SEM of relative abundance values, n = 10 marker transcripts).

Our attempts to propagate rat spermatogonial stem cells in culture were initially tested by culturing D0 Lam_B spermatogonia on a variety of feeder cell types and extracellular matrices (Supporting Text) by using DMEM:HAMS-F12 (1:1) medium supplemented with 10% FBS and 2-mercaptoethanol. Although some culture conditions yielded a high stem-cell index, such as on MSC1 feeder layers, none supported the proliferation of germ cells after the first passage. Media used to culture germ cells in the mouse include the addition of serum, basic fibroblast growth factor, glial cell-line-derived neurotrophic factor (GDNF), and/or GDNF family receptor α1 (10, 18-29). However, use of these media in the cultures of D0 rat Lam_B spermatogonia resulted in decreased numbers of germ cells after each passage.

Review of published protocols suggested that serum effectively prevented the expansion of germ cells in a medium supplemented with GDNF, GDNF family receptor α1, and basic fibroblast growth factor (19), although these reports were in contrast to the results of others where serum appeared to be needed for initial derivation and expansion of mouse germ stem cells grown on primary mouse embryonic fibroblasts (MEFs) (18, 20, 21). Without the addition of serum, the somatic feeders could not be removed from the germ cell population by an initial selection step on gelatin (18, 20, 21). We speculated that a combination of testis somatic cells and serum might attenuate the proliferation of rat germ stem cells in vitro. Based on this, we replaced the serum in S medium (20) with B27 minus vitamin A supplement to generate SA medium and removed essentially all testis somatic cells from the cultures. To do this, we first isolated Lam_B spermatogonia from GCS-EGFP rats and then further purified EGFP⁺ cells by FACS (Fig. 2A). A population of >99.7% pure EGFP⁺ cells was obtained and then cultured for 2 days on gelatin in the serum-containing S medium to further remove contaminating somatic cells (Fig. 2B). The purified germ cells did not adhere to the gelatin plates and were passaged onto monolayers of MEFs in the serum-free SA medium. As an example of the success of these cultures, we have passaged cells 12 times while in SA medium, spanning 151 days of culture, and have expanded the numbers of germ cells by ≈20,000-fold relative to the initial cell number (Fig. 2C). Thus, from 200,000 freshly isolated Lam_B germ cells, we have already generated ≈4 billion EGFP⁺ germ cells. After 110-140 days in culture, the EGFP⁺ cells remained positive for DAZL (Fig. 3A), maintained a normal number of rat chromosomes (42 per cell), and are type A spermatogonia based on both morphology (Fig. 3B) and DNA content (Fig. 3C). The marker transcripts associated with spermatogonial stem cell activity described above are ≈2-fold more abundant in the 60-day-old EGFP⁺ spermatogonial cultures than in D0 Lam_B spermatogonia (Fig. 4A). Transplantation of the D74 EGFP⁺ spermatogonia to the testes of busulfan-treated rats showed no loss in stem cell activity (Fig. 4 B-D). Additionally, the EGFP⁺ spermatogonia from these cultures developed into spermatids (EGFP⁺, Crem-τ⁺) within 62 days after transplantation to recipient rat testes (Fig. 4E). Testicular tumors were not observed after transplantation (n = 15 testes transplanted with a total of ≈3 × 10⁵ EGFP⁺ cells). Also, as with D0 Lam_B spermatogonia, the EGFP⁺ cells grown in these cultures bound avidly to laminin but not collagen (Fig. 6, which is published as supporting information on the PNAS web site).

Fig. 2.

Purification and proliferation of EGFP⁺ spermatogonia in vitro. (A) Flow cytometric properties of D0 GCS-EGFP⁺ Lam_B germ cells before and after purification by FACS. (Left) Side (SSC) versus forward (FSC) scatter plot. Region 1 (R1) contained ≈92% of the total population. (Center) ≈91% of Lam_B cells from R1 were EGFP⁺ germ cells (R2). (Right) Lam_B cells collected from R1 and R2 were >99% EGFP⁺ germ cells. (B) Selection of vimentin⁺ (red) gelatin-binding somatic cells from the D0 GCS-EGFP⁺ Lam_B germ cell (green) population before (Left) and after (Right) collection of the EGFP⁺ population by FACS. Cells were selected for 3 days on gelatin-coated plates in S medium before harvesting unbound germ cells and labeling bound cells with antivimentin IgG and Hoechst 33342 (blue). (C) Expansion of FACS purified EGFP⁺ spermatogonia in culture on MEFs. Cultures were initiated with 2 × 10⁵ FACS-purified EGFP⁺ cells (dashed line).

Fig. 3.

Properties of EGFP⁺ spermatogonia. (A) (Left) Colonies of EGFP⁺ spermatogonia (GCS-EGFP) expressing DAZL (DAZL IgG) after 11 passages and 140 days in culture. Nuclei of MEF feeder cells (EGFP^-, DAZL^-) and germ cells are labeled blue with Hoechst 33342. (Right) Preimmune IgG negative control antiserum. (B) Transmission electron micrographs of FACS-purified D0 (Upper), and D112 (Lower) EGFP⁺ spermatogonia displaying properties of undifferentiated type A spermatogonia. Images are magnified ≈×4,000. (C) DNA content of FACS-purified, EGFP⁺ spermatogonia after 120 days of culture in SA medium (Top) and of freshly isolated testes cells from 17- (Middle) and 45- (Bottom) day-old rats. DNA content values are expressed as 1C, 2C, and 4C for haploid, diploid, and tetraploid populations, respectively.

Fig. 4.

Stem cell activity of EGFP⁺ spermatogonia. (A) SSCI values for GCS-EGFP⁺ Lam_B germ cells before (D0 LB) and after (D0, D26, and D60 LB) purification from somatic cells by FACS. After FACS, EGFP⁺ cells were analyzed directly (D0 LB) or plated onto MEFs. After 26 and 60 days on MEFs (D26 and D60 LB), EGFP⁺ cells were purified again by FACS, and their SSCI values were determined. SSCI values are also shown for D0 GCS-EGFP⁺ Lam_NB germ cells (D0 LNB). (B) The number of EGFP⁺ colonies formed per testis by donor GCS-EGFP⁺ germ cells. One thousand EGFP⁺ germ cells were transplanted per testis before (D0 LNB, D0 LB) or after (D0 and D74 LB) purification by FACS. Purified germ cells were maintained on MEFs for 0 (D0) or 74 days (D74) before transplantation. Parentheses contain the number of testes and recipient rats analyzed. (C) (Upper Left, D74 LB) Right testis from a WT rat transplanted 63 days earlier with ≈10⁵ D74 EGFP⁺ spermatogonia. (Upper Right) Bright-field image of the same testis. (Lower Left and Right) Respective images of the nontransplanted (NT) left testis from the same rat. Note green fluorescence only in the transplanted testis. (Bar = 5 mm.) (D) (Left) Seminiferous tubule from a WT recipient rat containing an EGFP⁺ colony derived from D74 EGFP⁺ spermatogonia. (Right) Higher magnification of the same colony. (E) (Left) Cross section through the transplanted testis in top of C showing seminiferous tubules colonized by D74 EGFP⁺ spermatogonia (green). (Center) Nuclei of cells in the same section stained with Hoechst 33342 (blue). (Right) The same section showing donor cells that have developed into round spermatids (Crem-T, red). Asterisks denote tubules of the WT recipient that were not colonized by donor cells.

Addition of GDNF family receptor α1 (GFRα1) and/or basic fibroblast growth factor to a serum-free medium that contained GDNF was essential for expansion of mouse spermatogonial stem cells in a C57BL/6 background (19). When we increased the concentrations of GDNF and GFRα1 in SA medium (SB medium), even more robust expansion of EGFP⁺ rat germ cells occurred (Fig. 7, which is published as supporting information on the PNAS web site). Under these conditions, the EGFP⁺ population doubled at least once every 3.5 days (Fig. 7A), remained positive for DAZL (Fig. 7B), and maintained their SSCI values (Fig. 7C). Initial SSCI values for EGFP⁺ spermatogonia were also maintained after 1 month of storage in liquid nitrogen, subsequent thawing, and culture in SB medium (Fig. 7 C and D).

We optimized conditions for transfecting WT D0 Lam_B spermatogonia with Lipofectamine 2000 by using a DNA reporter plasmid encoding EGFP (pUBC-EGFP) by lipofection (Fig. 8, which is published as supporting information on the PNAS web site) and then applied these methods to expanding cultures of EGFP⁺ spermatogonia (Fig. 5A). As with WT spermatogonia, >20% of the EGFP⁺ spermatogonia cultured in SB medium were transfected with pUBC-DsRed2.1, which encodes a red fluorescent marker protein (Fig. 5A).

Fig. 5.

Transfection of rat spermatogonia. (A) GCS-EGFP⁺-spermatogonia (Upper, green) that were transfected with pUBC-DsRed2.1 (Lower, red) for 80 min. More than 20% of the EGFP⁺ spermatogonia were transfected based on counting red-fluorescent cells on a hemocytometer. (B) Diagram of OMP gene replacement vector (pOMP-TV). (C) (Left) Number of G418-resistant, EGFP⁺ colonies obtained after transfecting rat EGFP⁺ spermatogonia without DNA (-Neo) or with pOMP-TV (+Neo) and selection in 200 μg/ml (SA medium) and 150 μg/ml (SB medium) G418. (Right) Images of EGFP⁺ spermatogonia either mock-transfected without DNA (-Neo) or transfected with pOMP-TV (+Neo) after selection without (-G418) or with (+G418) 150 μg/ml G418 in SB medium. Colonies were scored and images taken at day 14 posttransfection and after 10 days of selection in G418-containing medium.

After our initial success using lentiviral constructs to transduce D0 rat Lam_B spermatogonial stem cells in culture (11), we proceeded to generate a gene replacement construct to modify rat spermatogonial stem cells at a desired genomic locus by homologous recombination. The targeting construct generated, pOMP-TV (Fig. 5B), contains a total of 7.35 kbp of rat genomic sequence from chromosome 1q32 that includes a 5.2-kb 5′ arm and a 2.15-kb 3′ arm, which flank the ORF of the rat gene, olfactory marker protein (Omp). The construct also contains a positive selection cassette (neomycin phosphotransferase) inserted between the 5′ and 3′ arms and a negative selection cassette (thymidine kinase) positioned directly outside the 3′ arm (Fig. 5B). These markers are also commonly used to select for mouse ES cells with targeted genomic modifications (30). Because isogenic DNA increases the rate of homologous recombination in mouse lines by 10-100 times (31), we generated a colony of GCS-EGFP rats that are isogenic for the Sprague-Dawley Omp allele used in the Omp targeting construct.

The rat Omp targeting construct was transfected into EGFP⁺ spermatogonia, and selection was initiated with G418. After 10 days, >20 G418-resistant colonies were selected per 10⁵ transfected EGFP⁺ cells when cultured in SA medium, whereas under separate conditions, >450 G418-resistant colonies were selected per 10⁵ transfected EGFP⁺ cells after 10 days in SB medium (Fig. 5C). Under conditions of mock transfection for either condition, no EGFP⁺ colonies survived the selection protocol (Fig. 5C). In recipient rats, the G418-resistant germ cells generated 1,130 ± 550 (± SEM, n = 4) GCS-EGFP⁺ colonies per testis per 10⁵ donor cells.

Discussion

We were unable to induce self renewal and expansion of rat spermatogonial stem cell numbers by a number of the published protocols used for the mouse (18-21, 23, 25, 29, 32). Successful propagation of rat spermatogonial stem cells requires a number of important features. First, in our studies, serum addition, even with minor contamination by testicular somatic cells, results in a loss of stem cell numbers after growth on a variety feeder layers. Previous studies suggested that factors detrimental to spermatogonial stem cell maintenance were present in serum (19, 25, 29), whereas in other studies, it seemed that, although spermatogonial stem cells could expand in the absence of serum when grown on MEFs, serum was essential for proliferation on a laminin matrix (21). This suggested that possibly serum stimulates the contaminating somatic cells to produce factors that block stem cell proliferation. Examination of the components of the B27 supplement used to replace serum in cultures of mouse germ-line stem cells (21) revealed that it contained vitamin A. Removal of vitamin A from the diet of rodents has been known for decades to block the differentiation of spermatogonial stem cells (33). Therefore, it seemed conceivable that the inclusion of vitamin A in the B27 supplement might contribute to the loss of rat spermatogonial stem cells during culture. The establishment of culture conditions free of serum, contaminating testes somatic cells, and external sources of vitamin A resulted in a population of rat germ cells that now proliferated while retaining stem cell characteristics, in much the same manner as seen with mouse or human ES cells.

The ability to apparently expand spermatogonial stem cell numbers indefinitely in culture may now result in a comprehensive understanding of the physiology and molecular attributes of the spermatogonial stem cell, gene targeting at the level of the spermatogonial stem cell (ensuring direct germline transmission), an ability to select against detrimental genetic defects, a capability to correct various forms of male infertility, the discovery of specific gene targets for development of male-directed contraceptives, and an understanding of the signaling pathways required for progression through the various stages of spermatogenesis.

Note. While this manuscript was under review, a paper appeared (34) that also obtained rat spermatogonial stem cell self renewal, although the methods to achieve this were considerably different from those presented here.