A 400 bp fragment of the spermatogonia‐specific Stra8 locus was sufficient to direct gene expression to the germinal stem cells in transgenic mice. A fractionation procedure was devised, based on immunomagnetic sorting of cells in which the promoter drives the expression of a surface functionally neutral protein tag. The purified cells expressed the known molecular markers of spermatogonia Rbm, cyclin A2 and EP‐Cam, and the β1‐ and α6‐integrins characteristic of the stem cell fraction. A 700‐fold enrichment in stem cells was determined by the ability of the purified fractions to re‐establish spermatogenesis in germ cell‐depleted recipient testes.
A number of adult tissues are in a dynamic state, with new cells continually differentiating from a pool of stem cells, self‐renewed by asymmetric cellular divisions. In the case of the male germ line, the spermatogonial stem cells constitute a small fraction of the testicular germ cells, in the range of 30 000 per adult testis, located in the most peripheral region of the seminiferous tubule (reviewed by de Rooij and Grootegoed, 1998). Mitotic divisions generate with a strict periodicity the first differentiated cell type, the A1 spermatogonia, from which differentiation follows a complex series of steps (reviewed by Bellve et al., 1977). Better knowledge of the properties of the stem cell fraction would be useful in investigating important questions in both basic and applied research, including the molecular controls of germinal differentiation and their pathological alterations (male sterility syndromes), new technologies for transgenesis (Nagano et al., 2001) and the possibility of trans‐differentiation (Terskikh et al., 2001 and references therein). The demonstration that spermatogonial stem cells can proliferate and differentiate in the seminiferous tubules of infertile recipients has provided a functional assay (Brinster and Zimmermann, 1994), and enriched preparations were obtained by selection based on the expression of cell surface molecules (Shinohara et al., 2000). Our initial aim was to identify a promoter that directs transgene expression to the earliest stages of male germ line differentiation. The Stra8 gene, expressed in spermatogonia (Oulad‐Abdelghani et al., 1996), appeared to be a hopeful candidate. We determined that a 400 bp Stra8 promoter fragment was sufficient for testicular expression of a reporter gene. Unlike the endogenous gene, however, this truncated promoter was only expressed in a minor fraction of premeiotic germ cells, which were purified by a procedure based on the expression of a neutral heterologous surface marker. These cells exhibited a high efficiency (∼700‐fold increased) in the establishment of stable spermatogenesis upon transplantation in germ cell‐depleted testes.
Results and Discussion
Promoter activity of an upstream fragment of the Stra8 locus
A mouse genomic library was screened using as a probe the complete Stra8 cDNA. A single 16 kb clone was obtained, which contains five exons corresponding to the 5′‐region of Stra8 cDNA (Figure 1). A fragment spanning nucleotide positions −371 to +29 relative to the transcriptional start site, including a TATA box‐like sequence and two putative binding sequences for retinoic acid receptors, was amplified and cloned into the luciferase reporter vector pXp1. Three lines of transgenic mice carrying the resulting construct (pStra8Luc) showed identical expression patterns, with high luciferase activity in testis extracts, and in no other tissue except brain (Figure 2). Not expected from the initial report (Oulad‐Abdelghani et al., 1996), the latter may be unique to the truncated form of the promoter. It was not further documented, and the following studies addressed the question of expression site(s) in the testis.
The 400 bp Stra8 promoter directs transgene expression to a minor fraction of spermatogonia
Analysis of luciferase expression in the prepubertal testis revealed expression at an early spermatogonial stage in the newborn testis, with identical results again in two independently established transgenic lines. Between birth and puberty, differentiation of germ cells progresses in a synchronous manner (Bellve et al., 1977). Luciferase activity during that period reached high levels prior to the entry into meiosis at day 10 post‐partum (p.p.), a critical period illustrated in Figure 2B by measurements performed on Sycp1‐luc transgenic males, in which the reporter is expressed at the very beginning of meiosis (leptotene to zygotene stages; Sage et al., 1999). Expression appeared to be restricted to the germinal fraction since extracts of pure cultures of Sertoli cells were inactive. In the two transgenic families analyzed, a limited number of positive cells were detected (Figure 3) in the periphery of the tubules (average 0.1 cell per section). A minor fraction contained ≥4 cells (up to 20), often located close to each other, a pattern evocative of sister cells generated by successive cell divisions. The profile of expression of the truncated promoter appeared, therefore, more restricted than that of the endogenous gene, which is expressed in all classes of spermatogonia. The observed increase in luciferase activity between days 1 and 20 p.p. suggests that the number of expressing cells increases after birth, and a detailed analysis is currently in progress.
Expression of a foreign neutral surface marker as a means to purify cells that express a transgene
A method was devised for purification of the cells in which the 400 bp Stra8 promoter is active. It is based on the expression on the cell surface of a protein that is not present in the normal mouse, does not interfere with cellular functions and is efficiently recognized by monoclonal antibodies. A recombinant gene designated CD4HAglo (Figure 4) was engineered for this purpose. This gene encodes a protein whose N‐terminal domain is made of the first two extracellular domains of the human CD4 molecule [insufficient for ligand recognition and recognized by a variety of monoclonal antibodies (Bedinger et al., 1988)], linked to the transmembrane and cytoplasmic regions of the influenza hemagglutinin (HA).
Two independent transgenic families were generated in which the CD4HAglo construct is expressed by the 400 bp Stra8 promoter. Expression of the transgene in testicular cells was verified by immunological detection (not shown). All the males of the two families showed fully normal spermatogenesis and complete fertility.
Magnetic cell sorting of CD4‐positive cells
Suspensions of well‐isolated testicular cells (20–30 × 106 cells per testis) were prepared from adult Stra8‐CD4HAglo males. Paramagnetic beads coated with anti‐CD4 monoclonal antibodies were used to sort the positive cells. A small fraction of the input (1 in 103 cells) attached to the beads. Of homogeneous morphology, these cells were determined to be diploid by cytometric analysis (not shown). Preparations from adult testis were initially contaminated by a small amount of elongated and round spermatids (Table 1), a specific problem in the fractionation of adult testicular cells with their largely predominant haploid cell population. Near‐homogeneous preparations (<0.1% contaminants) could be generated by two modifications of the procedure. One was to use testicular cells from 3‐week‐old males, in which the first wave of meiosis has just been completed (Bellve et al., 1977). The second possibility was to perform two successive fractionation steps with anti‐CD4‐coupled beads on adult testis cells. The counterpart of a high degree of purity (close to 100%) was, however, a lower yield, a problem that would clearly not be encountered in large‐scale preparations. The estimated yields (30 000 cells per testis) were those expected for the stem cell fraction (de Rooij and Grootegoed, 1998), with a very minor fraction corresponding to their immediate progeny (Figure 3).
Spermatogonia marker genes
Only a few proteins characteristic of the premeiotic murine germ cells have been identified that can be employed as purification markers. Immunocytochemical analysis (Table 2) detected in 80–90% of the CD4‐positive purified cells the RBM protein (Elliott et al., 2000), which in the mouse is exclusively present in spermatogonia (H. Cooke, personal communication), Ep‐Cam (Anderson et al., 1999), and the β1‐ and α6‐integrins, known markers of the stem cell population (Shinohara et al., 1999). These values are minimal estimates because, in order to count sufficient numbers of cells, the analysis was performed after only one round of immunomagnetic sorting. Labeling with antibodies directed against synaptonemal complex protein 1 (SCP1), a pachytene spermatocyte marker (Sage et al., 1999), did not detect any measurable contamination. RNA of cyclin A2 (spermatogonia) was readily detected by RT–PCR assays (Ravnik and Wolgemuth, 1999), but not that of cyclin A1 (spermatocyte stage) (Figure 5). RNA from the W locus (Kit receptor) could not be detected at significant levels, in agreement with a recent report concluding that the protein is absent in stem cells (Shinohara et al., 2000).
Restoring spermatogenesis in a recipient testis
The stem cell quality of the purified cells was definitely established by assaying their ability to restore complete spermatogenesis cycles in cell‐depleted seminiferous tubules (Brinster and Zimmermann, 1994). CD4‐positive cells were purified by mag‐netic sorting from double‐transgenic Stra8‐CD4HAglo;ROSA26 males (Zambrowicz et al., 1997) and injected into the seminiferous tubules of males previously irradiated to the point at which spermatogenesis was arrested. A total suspension of germ cells was injected in one testis (20 000 cells per injection) and, in the contralateral organ, the same suspension mixed with 2000 CD4‐positive purified cells (one‐round sorting from adult testis). Ten injections were performed in each testis, on two males per experiment, in three independent experiments conducted with fresh preparations of purified cells. The males were killed after 12 weeks and examined for the appearance of ROSA26 β‐galactosidase‐positive germ cells (Figure 6; Table 3). Staining of tubules was clearly stronger and more extensive upon injection of purified cells compared with injection of the total germ cell preparation. Moreover, the number of injection sites at which staining was detected was systematically close to 10 out of 10 in the testes that had received purified cells, and much lower in the controls. Staining of histological sections confirmed the presence of X‐Gal‐positive germ cells at every differentiation stage. When results were quantitated and reported relative to the number of injected cells, the values obtained with purified cells were in the range of 700‐fold greater than those recorded with the initial cell suspension (Table 3). Two series of injections were performed in which cells that had not been retained on the beads in the fractionation procedure were compared with the purified population. In this case (data not shown), the efficiency of colonization was too low to allow an accurate determination. The increase in colonization efficiency (700‐fold as a minimal estimate) was comparable to the yield in purified cells, thus indicating a high degree of purity. The fact that after 12 weeks, corresponding to multiple initiations from the stem population with the 10 day period characteristic of the adult mouse (Bellve et al., 1977), the tubule sections still showed a full complement of ‘blue cells’ demonstrates the establishment of stem colonies of β‐galactosidase‐positive cells. Experiments are in progress to check that fertility will indeed be efficiently restored in the case of sterile W mutants (Ogawa et al., 2000). The availability of homogeneous preparations of active germinal stem cells opens the way to a variety of research avenues, including gene transfer and re‐implantation as a new approach in transgenesis (Nagano et al., 2001), transcriptome studies, and possible transdifferentiation into other cell types depending on the local environment (Terskikh et al., 2001 and references therein).
Transgenic lines were generated by microinjection in C57BL/6 × DBA/2 (B6D2) F1 fertilized eggs according to standard procedures (Hogan et al., 1994).
(i) pStra8Luc. The −371/+29 fragment amplified from Stra8 genomic DNA (primers Stra8‐5′ and ‐3′) was inserted in the SmaI site of pXp1 (Li et al., 1994). (ii) pStra8 HAglo. The CD4HAglo marker has been assembled from PCR fragments amplified using primers CD4‐1 and ‐2 (first two domains of the human CD4), HA‐1 and ‐2 for the HA coding region with a Gly‐Gly‐Ser‐Ser spacer (Bodmer et al., 1994), and glo‐1 and ‐2 for the 3′‐UTR of the β‐globin gene.
Reverse transcription and PCR amplification.
RNA analysis was performed as described previously (Vidal et al., 2001), using primers gapdh‐1 and ‐2 for Gapdh expression; cycA1‐s and ‐r for Cyclin A1, confirmed by hybridization with cycA1‐h; cycA2‐s and ‐r for Cyclin A2, confirmed by hybridization with cycA2‐h; and ckit‐s and ‐r for W (Kit receptor) expression (Table 4).
Luciferase activity in cell extracts.
Activity was determined using the Dual‐Luciferase Reporter (DLRTM) Assay System according to the manufacturer's instructions (Promega). Protein concentrations were determined according to Bradford (1976).
Immunocytochemical localization of luciferase in testis sections.
In situ detection of the protein was performed on 12‐μm‐thick cryosections, using rabbit anti‐luciferase IgGs (Sigma) and FITC‐conjugated goat anti‐rabbit IgGs (Sigma).
Immunocytochemical analysis of isolated cells.
Cells were fixed in 4% paraformaldehyde for 30 min and permeabilized with 0.1% Triton X‐100 for 10 min. Primary antibodies were a polyclonal rabbit anti‐mouse RBM protein (kindly provided by Dr H. Cooke, MRC, Edinburgh, UK), rat anti‐mouse CD29 (β1 integrin) (BD Biosciences, Heidelberg, Germany) and rat G8.8 monoclonal antibody against the mouse Ep‐CAM protein (Anderson et al., 1999), obtained from the Developmental Studies Hybridoma Bank, Iowa City, IA. Depending on the primary antibody, the secondary antibody was either FITC‐conjugated goat anti‐rabbit IgG (Sigma) or anti‐rat R‐phycoerythrin‐conjugated antibodies (Sigma). FITC‐conjugated rat anti‐human CD49f (α6‐integrin) monoclonal antibody was from BD Biosciences.
Total germ cells were prepared as described previously (Vidal et al., 2001). Immunomagnetic isolation of CD4‐positive cells from total testicular cells of Stra8‐CD4HAglo mice was performed using the CD4 Positive Isolation Kit (Dynal, Oslo, Norway) according to the manufacturer's instructions.
Germ cell transplantation.
Recipient 3‐week‐old B6D2 males were γ‐irradiated [60Co, 3Gy, 1 Gy/min, 1.17–1.33 MeV on a Theratron 780 (Theratronics)], a dose that had been determined to be sufficient to arrest germinal differentiation. Three to four weeks later, after examination of control animals for the absence of spermatogenesis, purified Stra8‐CD4HAglo;ROSA26 cells (1× sorting) were injected in the tubules. CD4‐positive purified cells (104) from one male were resuspended in 100 μl of buffer together with 105 total germ cells from the same animal. Total germ cells only (106/ml, same preparation used for immunomagnetic sorting) were injected in the contralateral testis. Ten microlitres of cell suspension were injected at 10 distinct sites per testis. Ten to twelve weeks later, the transplanted males were killed and seminiferous tubules were stained with X‐Gal. Three measurements were found to be important to evaluate the extent of colonization of the recipient testis: the number of injection sites from which β‐galactosidase‐positive cells had developed; the length of stained tubule extending from these sites; and the intensity of X‐Gal staining determined using MacBAS v2.2 (Fujiphotofilm Co. Ltd) and Adobe Photoshop 5.5 (Adobe Systems Inc.) applications. The product of the three values was taken as an estimate of the colonization potential of the injected cells.
We thank D. Mathis and C. Benoist for helpful discussions, D.G. de Rooij for critical comments, and H. Cooke for the communication of unpublished observations and the generous gift of anti‐Rbm antibody. We are indebted to A. Costa and P.‐Y. Marcie (Centre Antoine Lacassagne, Nice, France) for access to equipment and technical help for irradiation of mice. This work was made possible by grants from Ministère de la Recherche (France) and Association pour la Recherche sur le Cancer.
- Copyright © 2002 European Molecular Biology Organization