Abstract
To explore physiological roles of BCL-W, a prosurvival member of the BCL-2 protein family, we generated transgenic (TG) mice overexpressing Bcl-w driven by a chicken β-actin promoter. Male Bcl-w TG mice developed normally but were infertile. The adult TG testes displayed disrupted spermatogenesis with various severities ranging from thin seminiferous epithelium containing less germ cells to Sertoli cell-only appearance. No overpopulation of any type of germ cells was observed during testicular development. In contrast, the developing TG testes displayed decreased number of spermatogonia, degeneration, and detachment of spermatocytes and Sertoli cell vacuolization. The proliferative activity of germ cells was significantly reduced during testicular development and spermatogenesis, as determined by in vivo and in vitro 5′-bromo-2′deoxyuridine incorporation assays. Sertoli cells were structurally and functionally normal. The degenerating germ cells were TUNEL-negative and no typical apoptotic DNA ladder was detected. Our data suggest that regulated spatial and temporal expression of BCL-W is required for normal testicular development and spermatogenesis, and overexpression of BCL-W inhibits germ cell cycle entry and/or cell cycle progression leading to disrupted spermatogenesis.
APOPTOSIS PLAYS AN important role in maintaining testicular homeostasis. It is estimated that up to 25–75% of the expected sperm yield are lost during normal spermatogenesis and the nature of germ cell degeneration during testicular development and spermatogenesis appear to be apoptosis (1, 2). In mice, primordial germ cells (PGCs) colonize the genital ridge at d 11.5 post coitum (pc), and PGCs proliferate actively thereafter and undergo a wave of apoptosis at d 13–17 pc (3). The first wave of spermatogenesis is accompanied by extensive germ cell apoptosis, which peaks at approximately 2–3 wk after birth in the mouse and rat (3–5). The precise homeostasis of each germ cell type in the adult appears to be achieved at the cost of substantial germ cell wastage. The mechanisms by which germ cell apoptosis is regulated are largely unknown. However, studies over the past decade have identified that hormones (FSH, testosterone, human chorionic gonadotropin, estradiol), paracrine factors [stem cell factor (SCF), bone morphogenetic protein 8B], Fas/FasL system, and BCL-2 family proteins are involved in the regulation of germ cell apoptosis (6–8). It appears that endocrine, paracrine, and autocrine factors form a complex network to transduce survival/death signals to germ cells. The signals conveyed by various pathways finally converge on apoptosis machinery of germ cells.
Apoptosis is a strictly regulated process, which can be divided into at least three different phases including sequential activation of specific signaling transduction pathways, disturbance of mitochondrial function causing the release of intermembrane proteins in the cytosol, and finally step-wise degradation of the cell (9–11). The intracellular cascades for apoptosis are likely to be conserved in various cell types and maybe even in different species. BCL-2 family members are key regulators of apoptosis. These proteins, characteristic of the presence of Bcl-2 homology domains, can either support cell survival (BCL-2, BCL-XL, BCL-W, MCL-1, A1/BFL-1) or promote cell death (BAX, BAK, BCL-XS, BAD, BID, BIK, BOK, DIVA, HRK). The Bcl-2 homology domains allow the BCL-2 family members to form heterodimers or homodimers with each other (10). It appears that the BCL-2 family members interact among each other to form a dynamic equilibrium between homo- and heterodimers (9–12). Because members of those opposing fractions can associate and seemingly titrate one another’s function, their relative abundance in a particular cell type may determine its threshold for apoptosis. The competitive action of the pro- and antisurvival Bcl-2 family proteins regulates activation of the proteases (caspases) that dismantle the cell, but how they do so remains unclear.
An increasing body of evidence has shown that BCL-2 family proteins indeed play an important role in controlling germ cell fate. First, most BCL-2 family members are expressed in the testis in a cell type-specific fashion (13–16). Second, they are all expressed in immature testes and their expression sites and levels start to differentiate upon puberty, displaying a cell type-specific expression pattern in the adult testis (13–16). Third, deletion of Bax (17) and Bcl-w (18, 19), overexpression of Bcl-2 (4, 20, 21) or Bcl-xL (4), and reduction of Bcl-xL expression by inhibiting its promoter activity (22), all result in sterility in male mice due to disrupted spermatogenesis. Although Bcl-w is expressed in multiple tissues (23), the Bcl-w-deficient mice developed normally and the only abnormality is the disrupted spermatogenesis leading to sterility (18, 19). In the absence of Bcl-w, testis appears to be normal during embryogenesis and the first 2 wk after birth. The incidence of germ cell apoptosis in Bcl-w knockout mice, however, becomes dramatically elevated between 2 and 4 wk of age. In the adult Bcl-w null males, both Sertoli cells and germ cells of all types are reduced in number due to depletion through apoptosis, and finally most of tubules are devoid of germ cells and Sertoli cells (16, 19). Apparently, Bcl-w is essential for maintaining normal spermatogenesis in adulthood. Recent studies show that loss of Sertoli cells in Bcl-w-deficient mice is independent of germ cells and removal of Bax can suppresses loss of Sertoli cells (24). However, a number of questions concerning the mechanisms behind the testicular phenotype of Bcl-w-deficient mice still remain. For example, is the depletion of germ cells due to intrinsic (lack of prosurvival Bcl-w in germ cells) or extrinsic (secondary to Sertoli cell depletion) mechanism? Are correct temporal and spatial expressions of Bcl-w required for normal spermatogenesis?
To further explore the physiological effects of overexpression of BCL-W on mouse development, we used a chicken β-actin promoter to direct exogenous Bcl-w expression in multiple tissues. Interestingly, sterility was the only phenotype in these transgenic (TG) males. In the present study, we analyzed in details how overexpression of Bcl-w in the testis affects testicular development and spermatogenesis.
RESULTS
Male TG Mice Overexpressing Bcl-w Are Infertile
Bcl-w-overexpressing mice were generated using a construct containing the entire coding region of the mouse Bcl-w cDNA driven by a chicken β-actin promoter (Fig. 1A). Nine male and eight female founders were obtained. When breeding with wild-type (WT) females, none of these nine male founders produced any pups, whereas the female founders reproduced normally. F1 and F2 TG males derived from female founders were bred with WT females for 6 months, and none of them sired litters.
Generation of Bcl-w TG Mice A, Schematic presentation of the transgene construct. The construct consists of an EcoRI fragment containing the entire coding region of murine Bcl-w cDNA under the chicken β-actin promoter (containing an intron) and followed by a β-globin poly-A tail. Primers 1 and 2 were used for genotyping, and two riboprobes (1 and 2) were employed for Northern blot analysis. Riboprobe 1, derived from the coding region of Bcl-w cDNA, recognizes both the endo- and exogenous Bcl-w transcripts, whereas riboprobe 2, originated from 3′ noncoding region, can only detect the endogenous Bcl-w. B, Exogenous Bcl-w expression in three adult testes from TG line 2. Note the expression levels of the exogenous Bcl-w are much higher than those of endogenous Bcl-w. Hybridization for 28S rRNA was used as a loading control. C, Expression of exogenous Bcl-w in the developing TG testes (line 2) aging from 1 wk (7d) to 4 months (120d). Hybridization for 28S rRNA was used as a loading control. D, Western blot analysis of BCL-W expression in 1-wk-old TG testes from three different lines (TG1 from line1, TG2 from line 2, and TG3-1 and TG3-2 are two mice from line 3) and their WT littermates. Actin was detected by reprobing the same membrane for monitoring even loading.
By comparing Northern blot analyses using two riboprobes, one of which is endogenous Bcl-w specific (Fig. 1A, riboprobe 2), and the other can detect both endogenous and exogenous Bcl-w (Fig. 1A, riboprobe 1), we found that both endogenous and exogenous Bcl-w displayed similar expression patterns in multiple tissues in the TG males and their WT littermates (data not shown). Despite high levels of exogenous Bcl-w in multiple organs, we did not observe gross or histological abnormalities in other tissues (data not shown) except in the testis. Three lines of TG males displayed different levels of exogenous Bcl-w mRNA in the testes were chosen for detailed characterization (see below). In TG line 2, levels of exogenous Bcl-w mRNA ranged from 5- to 15-folds, as compared with WT (Fig. 1B). During testicular development in TG line 2, levels of exogenous Bcl-w increased, whereas the endogenous Bcl-w levels slightly decreased (Fig. 1C). Using a monoclonal antibody against BCL-W, we also performed Western blot analyses, which showed that BCL-W levels were much higher in the TG testes than in WT at 1 wk of age (Fig. 1D). These data demonstrate that exogenous Bcl-w is overexpressed not only in mRNA levels but also in protein levels in the testes of TG mice.
Testicular Parameters of Bcl-w TG Males
Because the severity of spermatogenic disruption varied among the three lines, and lines 1 and 2 displayed partial disruption (see below and also see Fig. 3), we analyzed the testicular parameters for these two lines (Table 1). Testis weights were significantly reduced in adult TG males, which were less than half of the WT. Sperm counts in adult TG mice were about 1/30 of the WT (Table 1). Both gonadotropins, FSH and LH, were slightly elevated, but the difference was statistically insignificant. Serum testosterone levels in TG mice were slightly lower than those in WT mice. However, the differences were not statistically significant (Table 1).
Testicular Histology of the Adult Bcl-w TG Mice Bouin-fixed, paraffin-embedded testis samples were sectioned and stained with hematoxylin-eosine. A, Normal testicular histology in adult WT mice, characteristic of robust spermatogenesis; B–E, four types of testicular histology observed in 3 lines of TG mice: B (line 1), homogeneous appearance of seminiferous tubules with thin epithelium containing all types of germ cells, but much less in number; C (line 2), heterogeneous appearance of two types of tubules: one devoid of elongating and elongated spermatids, but containing few round spermatids and numerous degenerating spermatocytes; the other with underpopulated epithelium resembling the tubules in B; D (line 3), homogeneous appearance of severely depleted epithelium containing multinuclear giant cells (representing degenerated round spermatids), massively degenerating spermatocytes, numerous vacuoles, and no elongating or elongated spermatids. E (line 3), Represents the most severe testicular phenotype, characteristic of nearly complete depletion of all types of germ cells and Sertoli cell-only appearance. All microphotographs are in the same magnification (×200).
Testicular Parameters of Bcl-w TG Mice from Lines 1 and 2a
| WT | Bcl-w TG | |
|---|---|---|
| Testis weight (mg) | 118.6 ± 7.7 | 57.4 ± 18.2b |
| Sperm count (×106) | 25.7 ± 11.1 | 0.8 ± 0.2b |
| FSH (ng/ml) | 39.8 ± 2.5 | 42.3 ± 1.9 |
| LH (ng/ml) | 0.17 ± 0.1 | 0.19 ± 0.2 |
| Testosterone (ng/ml) | 4.7 ± 1.3 | 3.9 ± 2.9 |
| WT | Bcl-w TG | |
|---|---|---|
| Testis weight (mg) | 118.6 ± 7.7 | 57.4 ± 18.2b |
| Sperm count (×106) | 25.7 ± 11.1 | 0.8 ± 0.2b |
| FSH (ng/ml) | 39.8 ± 2.5 | 42.3 ± 1.9 |
| LH (ng/ml) | 0.17 ± 0.1 | 0.19 ± 0.2 |
| Testosterone (ng/ml) | 4.7 ± 1.3 | 3.9 ± 2.9 |
Values represent mean ± sem, n = 6 for each group;
P < 0.05, by t test.
Testicular Parameters of Bcl-w TG Mice from Lines 1 and 2a
| WT | Bcl-w TG | |
|---|---|---|
| Testis weight (mg) | 118.6 ± 7.7 | 57.4 ± 18.2b |
| Sperm count (×106) | 25.7 ± 11.1 | 0.8 ± 0.2b |
| FSH (ng/ml) | 39.8 ± 2.5 | 42.3 ± 1.9 |
| LH (ng/ml) | 0.17 ± 0.1 | 0.19 ± 0.2 |
| Testosterone (ng/ml) | 4.7 ± 1.3 | 3.9 ± 2.9 |
| WT | Bcl-w TG | |
|---|---|---|
| Testis weight (mg) | 118.6 ± 7.7 | 57.4 ± 18.2b |
| Sperm count (×106) | 25.7 ± 11.1 | 0.8 ± 0.2b |
| FSH (ng/ml) | 39.8 ± 2.5 | 42.3 ± 1.9 |
| LH (ng/ml) | 0.17 ± 0.1 | 0.19 ± 0.2 |
| Testosterone (ng/ml) | 4.7 ± 1.3 | 3.9 ± 2.9 |
Values represent mean ± sem, n = 6 for each group;
P < 0.05, by t test.
Altered Expression Sites and Levels of Bcl-w mRNA in TG Testes during Testicular Development and Spermatogenesis
We performed in situ hybridization (ISH) analyses using two riboprobes described above to examine the expression levels and sites of endogenous and exogenous Bcl-w in the TG and WT testes. The differences between the ISH results derived from two probes represent the expression of exogenous Bcl-w. The ISH analyses revealed that the magnitude of exogenous Bcl-w expression (Fig. 2, C and D) in different cell types within the testis appeared as follows: Leydig cells > Sertoli cells > spermatogonia = spermatocytes > spermatids, in comparison to that of endogenous Bcl-w (Fig. 2, A and B): Sertoli cells > spermatogonia > spermatocytes > Leydig cells (13, 14). In the immature testes at ages of 2–3 wk, the expression levels followed the same order as in the adult TG mice, but different from the expression pattern in the WT immature testes (Sertoli cells = spermatogonia > spermatocytes > Leydig cells) (data not shown). Sense controls gave no or very weak background signals (data not shown).
ISH Analysis of Bcl-w mRNA in WT (A and B) and TG (C and D) Testes Dark-field (A and C, at ×100) and bright-field (B and D, at ×400) images are shown. The intensity of hybridization signals in the WT testis (A and B) follows the following order: Sertoli cells > spermatogonia > spermatocytes > Leydig cells, in comparison to the pattern in the TG testis (C and D): Leydig cells > Sertoli cells > spermatogonia = spermatocytes > spermatids. Sc, Sertoli cell; Sg, spermatogonium; Sp, spermatocyte; Sd, spermatid; Lc, Leydig cell.
Testicular Histology of the Bcl-w TG Mice
At 3 months of age, TG testes displayed disrupted spermatogenesis with four types of histological appearances in three lines studied. The severity of disruption appears to correlate with the expression levels of exogenous Bcl-w. In TG line 1 (12 mice analyzed) where the exogenous Bcl-w mRNA levels were 2- to 5-fold higher than the endogenous Bcl-w, the seminiferous epithelium appeared to be much thinner than that in the WT (Fig. 3B). It contained all types of germ cells, but in reduced numbers, which made the lumen appear larger (Fig. 3B), as compared with the WT (Fig. 3A). The exogenous Bcl-w levels were 5- to 15-fold higher than the endogenous Bcl-w in TG line 2 (16 mice analyzed). In this line, approximately two thirds of the tubules were devoid of spermatids, but contained spermatogonia and few spermatocytes (Fig. 3C); the remaining one third of the tubules contained all types of germ cells, but in reduced numbers (Fig. 3C), similar to the epithelial pattern in the line 1 (Fig. 3B). The TG line 3 (10 mice analyzed) expressed considerably high levels of exogenous Bcl-w (15- to 30-fold over endogenous Bcl-w). In these mice, all seminiferous tubules were severely depleted of germ cells (Fig. 3, D and E). Numerous degenerating and detached spermatocytes, multinucleated giant cells (representing degenerating spermatids), and vacuoles were present within the seminiferous epithelium (Fig. 3D). In some of these mice, the seminiferous epithelium contained only Sertoli cells and very few spermatogonia (Fig. 3E), representing the most severe germ cell depletion in Bcl-w TG males.
Onset of Germ Cell Depletion in Developing Bcl-w TG Testes
As suggested in TG mice overexpressing Bcl-2 (4, 20, 21) or Bcl-XL (4), the depletion of germ cells may result from overpopulation of germ cells during testicular development due to the presence of excessive amount of prosurvival BCL-2 family proteins. We therefore analyzed the histology of developing testes in young TG mice. To our surprise, no discernable morphological abnormalities were observed in TG testes from birth to 1 wk of age (Fig. 4). However, by wk 1 there seems to be less germ cells in the TG seminiferous epithelium than in WT (Fig. 4). From wk 2 onward, germ cell depletion became obvious and vacuoles started to appear in the epithelium in TG males (Fig. 4). The depleted germ cells were mainly spermatocytes and some spermatocytes appeared to detach from Sertoli cells and slough into tubule lumen. Germ cell depletion became more severe at wk 3 and 4 and more vacuoles and less germ cells were present within the epithelium (data not shown). Thus, we did not observe overpopulation of any type of germ cells during any stages of testicular development in the TG mice.
Histology of Developing Testes in Bcl-w TG and WT Mice Paraformaldehyde-fixed, paraffin-embedded testis samples were sectioned and stained with hematoxylin and eosin. No discernable abnormality was observed in the TG and WT testes before 1 wk of age. By wk 1, the TG seminiferous tubules seem to contain less germ cells. Numerous vacuoles (V) and degenerating/degenerated germ cells (D) are present in the seminiferous epithelium at wk 2. All microphotographs are in the same magnification (×400).
Normal Ultrastructure of Sertoli Cells in Bcl-w TG Mice
Electron microscopy revealed that Sertoli cells displayed a columnar shape, intact tight junctions, and normal organelles in the TG mice (Fig. 5). Most of the vacuoles contained membrane-like debris resembling undigested cytoskeleton after phagocytosis (Fig. 5).
Electron Microphotographs of Basal Compartment of Seminiferous Epithelium of Bcl-w TG Mice Arrows point to Sertoli-Sertoli tight junctions. A, Sertoli cell (S) contains a degenerating (D) and phagocytozed germ cell. B, Sertoli cells (S) contain large vacuoles (V) with some membrane debris of phagocytozed and dissolved germ cells. Uranyl acetate and lead citrate staining was used. Original magnification: A, ×4000; B, ×2500.
Lack of TUNEL Staining and Absence of Apoptotic DNA Fragmentation in the Degenerating Germ Cells in TG Mice
No enhanced germ cell apoptosis was detected during the first 3 wk of postnatal testicular development (Fig. 6A) using TUNEL assay. The numbers and types of TUNEL-positive germ cells in both immature and mature TG testes were not different from those in WT mice. Those morphologically apparently degenerating spermatocytes and spermatids were TUNEL-negative, suggesting that these cells were depleted in a nonapoptotic fashion. This was further corroborated by DNA laddering analyses, which showed that random DNA degradation, rather than typical apoptotic DNA ladder, was present in the TG testis (Fig. 6B).
Germ Cell Apoptosis in Developing and Adult Bcl-w TG and WT Testes A, TUNEL assay of the Bcl-w TG and WT testes during testicular development and spermatogenesis. Arrows point to TUNEL-positive germ cells. All microphotographs are in the same magnification (×400). B, DNA ladder analysis of the testes from four adult TG mice (line 2). Mouse testes at 12 h (control 1), 18 h (control 2), and 24 h (control 3) after methoxyacetic acid treatment (ip 650 mg/kg body weight), which induced spermatocyte apoptosis, were employed as positive controls. M, Molecular marker.
Inhibited Proliferative Activity of Spermatogonia in Bcl-w TG Mice in Vivo
As described above, no obvious germ cell degeneration was observed in TG testes from birth to postnatal d 7 (P7). However, there seems to be less germ cells in the TG seminiferous epithelium than in WT by P7. To determine the proliferative activity of spermatogonia during testicular development, we performed quantitative analyses using immunohistochemical detection of incorporated 5′-bromo-2′deoxyuridine (BrdU) and a Southern-Western-based method to quantify total BrdU incorporation (Fig. 7). By counting BrdU-positive spermatogonia per 100 Sertoli cells, we found less BrdU-positive spermatogonia in TG testes than in WT at P3–7 (Fig. 7, A and B). The newborn WT and TG testes contain comparable number of spermatogonia per 100 Sertoli cells (data not shown), indicating that there is no defect before birth in germ cell proliferation. Before P3, spermatogonia are in a nonproliferative status, as demonstrated by absence of BrdU incorporation in both WT and TG testes (Fig. 7B). This also supported by similar levels of c-kit mRNA expression in both WT and TG testes between birth and P2 (Fig. 8A). By P3, some spermatogonia start to resume proliferation. After P3, the proliferative activity of spermatogonia increased dramatically. Comparing WT and TG testes, we found that BrdU-positive spermatogonia in the TG testes were only approximately 50% of the WT (Fig. 7B). By the ages of 2 and 3 wk, TG testes contained much less BrdU-positive cells than WT testes (Fig. 7A).
Reduced BrdU Incorporation during Testicular Development and Spermatogenesis in Bcl-w TG Mice A, Immunohistochemical detection of incorporated BrdU in the immature WT and TG testes at the ages of 3–7 d (3d-7d), 2 wk (2w) and 3 wk (3w). Arrows point to BrdU-positive spermatogonia. All microphotographs are in the same magnification (×200). B, Quantitative analyses of BrdU-positive spermatogonia in TG and WT testes during the first postnatal week of development. Bars represent the numbers of BrdU-positive spermatogonia per 100 Sertoli cells (means ± sem, n = 3). C, Southern-Western-based BrdU content assay for developing testes. An aliquot of DNA was blotted onto nylon membrane (1/1000 of the DNA isolated from the whole testes) and BrdU was detected by an alkaline phosphatase (AP)-conjugated anti-BrdU monoclonal antibody in conjunction with chemiluminescence (upper panel). The results were scanned and quantified (lower panel). Bars represent percentage of BrdU contents of the WT (means ± sem, n = 3).
SCF/c-kit System in the TG Testes A, Northern blot analysis of SCF mRNA expression in the testes of Bcl-w TG and WT mice during the first postnatal week of development. The same membrane was reprobed for 18S rRNA as a loading control. B, Semiquantitative RT-PCR analyses of c-kit mRNA expression in TG and WT testes during first postnatal week of development. Mouse β-actin was amplified as a loading control.
Severe degeneration of spermatogenic cells after 2 wk of age changes the proportional composition of different germ cell types within the testis. Therefore, we cannot quantify gene expression and BrdU incorporation based on signals from similar amount of starting materials from mice older than 2 wk. Because no obvious germ cell degeneration before wk 1, all gene expression and BrdU incorporation analyses reported from herein were performed using testes before 1 wk of age. Using a Southern-Western blotting-based assay (25), we quantified BrdU contents of the testis during the first 3 wk of development (Fig. 7C). The quantitative analyses showed that BrdU contents in TG testes (three TG mice from line 3) were dramatically reduced, as compared with WT testes (Fig. 7C). For example, BrdU contents of the TG testes were approximately 50% of the WT during d 4–7 after birth (Fig. 7C). Data from this assay are consistent with our morphological quantification of numbers of BrdU-positive spermatogonia described above. Both reveal a reduction in proliferation rates of spermatogonia in TG mice, suggesting that overexpression of BCL-W could affect cell cycle entry and/or cell cycle progression of spermatogonia. We also performed immunohistochemical staining of Ki-67 and quantitative analyses revealed similar reduction in proliferative activity of spermatogonia in the TG testes during P3–7 (data not shown).
Normal Production of SCF and Reduced Levels of c-Kit in Bcl-w TG Testes
The SCF/c-kit system plays important roles in stimulating germ cell proliferation and survival during testicular development and spermatogenesis (26, 27). Northern blot analyses showed no differences in SCF mRNA production in both TG and WT testes during the first postnatal week of development (Fig. 8A). A semiquantitative RT-PCR analysis uncovered that c-kit mRNA levels were reduced in TG testes (Fig. 8B) after P3. Normal c-kit production before P2 suggests that there are similar numbers of c-kit-expressing spermatogonia in both TG and WT testes at these time points. Decreased levels of c-kit thereafter may result from reduced number of c-kit-positive spermatogonia, consistent with our BrdU incorporation assays described above. Based on normal SCF production and normal structure of Sertoli cells (Figs. 3, 4, 5), we suggest that Sertoli cells of the TG testes function normally. Therefore, the defects in germ cells must be intrinsic (due to overexpression of BCL-W).
Inhibited Proliferative Activity of Bcl-w-Overexpressing Spermatogonia in Vitro
SCF can stimulate DNA synthesis in the immature testes (28, 29) and in cultured adult seminiferous tubules (30). To further confirm our in vivo findings, we isolated seminiferous tubule segments from TG males at P4 and P7, and cultured in the presence or absence of a recombinant mouse SCF (100 ng/ml) for 24 h. BrdU (10 ng/ml) was added 10 h before harvest. At both P4 and P7 the basal proliferative activity (absence of SCF, SCF−) was lower in TG testes than in WT. The stimulated proliferative activity (presence of SCF, SCF+) was much higher in WT testes than in TG testes (Fig. 9). The data suggest that germ cells in the TG tubules are less responsive to DNA synthesis-stimulating signals, and cell cycle progression is indeed compromised in TG testes. Interestingly, the decreased proliferative activity correlated with reduced number of c-kit-positive spermatogonia (Fig. 8), suggesting that the number of differentiating spermatogonia was reduced in the TG testes.
Reduced Proliferative Activity of Bcl-w-Overexpressing Spermatogonia in Vitro Seminiferous tubules dissected from TG or WT testes at P4 and P7 were cultured for 24 h in the presence or absence of recombinant mouse SCF (100 ng/ml). BrdU was added at a final concentration of 10 ng/ml 10 h before harvest and incorporated BrdU was measured using a proliferation assay kit (Roche Molecular Biochemicals). Bars represent OD values (means ± sem, n = 4).
DISCUSSION
Taking advantage of the size difference between endogenous and exogenous Bcl-w transcripts, we could simultaneously analyze the expression levels of both transcripts using Northern blot analyses. Despite high levels of exogenous Bcl-w expression in multiple tissues (data not shown), the male TG mice displayed normal development, indicating that overexpression of Bcl-w does not affect embryogenesis and early development. Sterility is the only phenotype observed in the Bcl-w overexpressing males. Given the pro-survival roles of BCL-W, we initially assumed that the testicular phenotypes of these TG males would be similar to those of Bcl-2-overexpressing testes, e.g. germ cell overpopulation followed by depletion through apoptosis (4, 20, 21). To our surprise, we did not find any time points, when any type of germ cells was overpopulated during testicular development. On the contrary, all types of germ cells appear to be underpopulated in the seminiferous epithelium and constantly degenerate throughout testicular development and spermatogenesis. This observation prompted us to investigate other mechanisms underlying the spermatogenic defects in the Bcl-w TG mice.
Although BCL-2 and its homologs inhibit apoptosis, their overexpression can delay entry into cell cycle and promote exit of cell cycle (31–36). The findings are based upon the analysis of cell cycle distribution, cell cycle kinetics, and relative phosphorylation of the retinoblastoma tumor suppressor protein, using primary tissues in vivo, ex vivo, and in vitro, as well as continuous cell lines (31–36). The effects of BCL-2 overexpression on cell cycle progression appear to be focused at the G1 to S phase transition, which is a critical control point in the decision between continued cell cycle progression or the induction of programmed cell death. It has been shown that Bcl-2 overexpression resulted in a 30–60% increase in the length of G1 phase; such an increase is very substantial in the context of other regulators of cell cycle progression (31, 32).
Based on these previous studies, we propose that overexpression of BCL-W may have inhibitory effects on cell cycle entry and/or cell cycle progression in male germ cells in vivo. Spermatogonia are in a nonproliferative state at birth, and they resume proliferative activity by P3-P4 in the mouse. In terms of cell cycle phases, spermatogonia are in G0 status at birth. When the resumption of proliferation starts, these cells reenter cell cycle (G0→G1) and then proceed to S phase. Reduced number of BrdU-positive spermatogonia in TG testes in the first week after birth implicates that overexpression of BCL-W may hinder the reentry of cell cycle (G0→G1) and/or G1 to S phase transition. More convincingly, the in vivo observation was further confirmed by our in vitro studies using seminiferous tubule culture system and BrdU incorporation assays, which showed that spermatogonia were less responsive to proliferation-stimulating signals, such as SCF. Therefore, we suggest that overexpression of BCL-W can affect cell cycle entry and/or cell cycle progression of spermatogonia. Interestingly, the severity of the spermatogenic disruption seems to correlate with the expression levels of exogenous Bcl-w. This observation is consistent with the previous studies showing that the cell cycle effect of overexpression of the prosurvival members of BCL-2 protein family is dose dependent (31, 32). Given the fact that the most severely degenerating germ cells are spermatocytes in adult TG testes, we also postulate that meiotic prophase may be another sensitive window to the overexpression of BCL-W. Thus, it appears that overexpression of BCL-W has toxic effects on spermatogonia and spermatocytes, which result in decreased proliferative activity in spermatogonia and degeneration of spermatocytes. Isolation of different germ cell types from the TG testes and analysis of cell cycle-related gene expressions may further help us understand the mechanisms underlying the global inhibitory effects of overexpression of Bcl-w on germ cell cycle progression in the Bcl-w TG testes.
Although our Bcl-w TG males and the previously reported male TG mice overexpressing Bcl-2 (4, 20, 21) or Bcl-xL (4), all displayed sterility due to degeneration of germ cells, the mechanisms are different. The previous studies suggest that the depletion of germ cells are due to overpopulation of germ cells, whereas our findings show underpopulation of germ cells due to inhibitory effects of BCL-W overexpression on cell cycle entry and/or cell cycle progression. A question arises: why do those TG mice overexpressing Bcl-2 (4, 20, 21) or Bcl-xL (4) fail to display similar cell cycle effects? The following factors may contribute to the discrepancy: first, considerably high levels of the prosurvival proteins may be needed for the cell cycle effects because of dose-dependent characteristics. Our TG mice have very high (5- to 30-folds) levels of exogenous Bcl-w mRNA and protein. We could not compare our data with previous studies (4, 20, 21) because no data on transgene expression levels were reported. Second, the cDNAs for Bcl-2 and Bcl-xL used in the previous studies are all from humans, which may compromise the ability of the transgene products to induce cell cycle effects. Third, different promoters including inhibin-α (20), elongation factor-1 (21), and phosphoglycerate kinase-1 (pgk-1) (4), were used in the previous studies. These promoters may have different temporal and spatial activities in germ cell lineages during testicular development and spermatogenesis, which could affect the magnitude and timing of the overexpression of these prosurvival BCL-2 family proteins. Fourth, most of published data so far suggest that no or very little expression of BCL-2 in WT testes in both immature and adult mice (4, 5, 14, 19, 22, 24). Therefore, overexpression BCL-2 may not be able to activate the same pathway leading to cell cycle effects as overexpression of BCL-W dose.
In summary, the present study reveals that regulated spatial and temporal expression of Bcl-w in the testis is required for normal testicular development and spermatogenesis. Our TG models manifest the cell cycle effects of overexpression of BCL-W in vivo. We propose that overexpression of BCL-W in the testis disrupts spermatogenesis by hindering cell cycle entry and/or cell cycle progression of germ cells. Therefore, BCL-W not only functions as a prosurvival protein but also has a role in the regulation of germ cell cycle during testicular development and spermatogenesis.
MATERIALS AND METHODS
Generation of TG Mice
The transgene construct is schematically illustrated in Fig. 1. Briefly, a 1.1-kb EcoRI fragment, containing the entire coding region of mouse Bcl-w cDNA (kindly provided by Dr. Jerry M. Adams, the Walter and Eliza Hall Institute of Medical Research, Victoria, Australia) was subcloned into a vector and flanked by a chicken β-actinpromoter, an intone, and a β-globin polyadenylate tail (kindly provided by Dr. Junichi Miyazaki, Department of Nutrition and physiological Chemistry, Osaka University Medical School, Japan). The 4.4-kb transgene fragment isolated by digestion with SalI/HindIII was microinjected into the pronuclear of fertilized eggs from the FVB/N strain. Microinjected oocytes were implanted into oviducts of psuedopregnant female mice (NMRI strain) and carried to term. Integration of the transgene was detected by PCR using β-actin-promoter-specific upstream primer (5′-actggtcgctagatacagat-3′), a Bcl-w-specific downstream primer (5′-acccagttatagatagatagcgta-3′), and DNA isolated from tail biopsies. All animal experiments were approved by the Committee of the Ethics of Animal Experimentation in the University of Turku.
Gross and Microscopic Histology
Mice received a single ip injection of BrdU (50 mg/kg body weight) 1 h before they were killed. Testes and seminal vesicles were collected and weights were measured. One testis was snap frozen in liquid nitrogen for RNA and protein isolation and the other was fixed with Bouin’s fixative or 4% paraformaldehyde at 4 C overnight followed by dehydration and embedding into paraffin. Sections were cut at 5 μm and stained with hematoxylin and eosin for microscopic analysis.
Northern Blotting Analysis
RNA preparation, gel fractionation, transfer, and hybridization were performed as described previously (37).
TUNEL and DNA Ladder Analyses
TUNEL and DNA ladder analyses were performed as described previously (38). Methoxyacetic acid-treated testes were used as positive controls in DNA laddering analysis. MAA treatment and sample collection were performed as described previously (13).
In Situ Hybridization
Five-micrometer sections were cut from paraffin-embedded testis samples and mounted onto SuperFrost Plus glass slides (Menzel-Gläser, Steinheim, Germany). The slides were incubated at 37 C overnight and then stored at 4 C before use. In situ hybridization was performed as described previously (13).
Semiquantitative RT-PCR
A semiquantitative RT-PCR method (39) was performed to measure levels of c-kit mRNA using a pair of primers: 5′-GTGCCCGAAACAAGTCATCTCC-3′ and 5′-GTGCCTCCTTCTGTGCCTTTCAAT-3′. As an internal control, mouse β-actin was amplified using a primer pair purchased from CLONTECH Laboratories, Inc. (Palo Alto, CA). The cycle numbers that keep the PCR at exponential phase are as follows: 20 cycles for c-kit; 18 cycles for β-actin.
Immunohistochemistry
Two 5-μm consecutive sections were cut from each sample and mounted onto poly-lysine-coated slides. Microwave antigen retrieval was employed as described previously (25). After blocking, an aliquot of a monoclonal anti-BrdU antibody (DAKO, Glostrup, Denmark) diluted at 1:200 was applied to each section and incubated at 4 C overnight. Incubation with secondary antibody and visualization of positive cells were performed using VectaStain Elite Kit (Vector Laboratories, Burlingame, CA) according to the manufacturer’s instructions.
Western Blot Analysis
Protein was isolated from 1-wk-old testes of TG (three different lines) mice and their WT littermates. Western blot analyses were performed as described previously (5). A rat monoclonal anti-BCL-W antibody (Alexis Biochemicals, Inc., San Diego, CA), which was shown to be highly specific (23), was used at a dilution of 1:100.
Southern- and Western-Based BrdU assay
BrdU injection, DNA isolation, slot blotting, and Western detection were performed as described previously (25).
Tissue Culture and Proliferation Assay
Seminiferous tubules isolated from testes at P4 and P7 of TG and WT mice were cultured in the DMEM/F12 medium (1:1) (GIBCO BRL, Paisley, Scotland, UK) supplemented with 15 mm HEPES, 1.25 g/liter sodium bicarbonate, 10 mg/liter gentamycin sulfate, 60 mg/liter G-penicillin, 1 g/liter BSA, and 0.1 mm 3-isobutyl-1-methylxanthin (Aldrich Chemie, Steinheim, Germany) for 24 h in the presence or absence recombinant mouse SCF (Genzyme Transgenics Corp., Cambridge, MA) at 10 ng/ml. BrdU was added into the medium 10 h before harvest. Upon harvest, the cultured tubules were homogenized by passing through a needle (G23) attached to a syringe for 5–10 times. The homogenates were centrifuged at 7000 rpm for 5 min at 4 C. Supernatant was discarded. BrdU contents were then measured using a digoxigenin-based proliferation measurement kit according to the manufacturer’s instructions (Roche Molecular Biochemicals, Indianapolis, IN).
Electron Microscopy
Sampling, fixation, and electron microscopic analysis were conducted as described previously (40).
Acknowledgments
We would like to thank Dr. Jerry M Adams of the Walter and Eliza Hall Institute of Medical Research (Victoria, Australia) for providing us with the mouse full-length Bcl-w cDNA and Dr. Junichi Miyazaki, Department of Nutrition and Physiological Chemistry, Osaka University Medical School (Osaka, Japan) for a gift of the vector containing chicken β-actin promoter and β-globin polyadenylate tail. We are grateful to Johanna Vesa, Nina Messner, and Tarja Laiho for their excellent technical assistance. Henna-Riikka Leinonen and Suvi Haka are acknowledged for taking care of our TG mice. We are also grateful to Dr. John Eriksson, Ph.D., for inspiring discussions.
This work was supported by grants from the Academy of Finland and European Union contract QLK4-1999-01422 and QLK4-CT-2000-00684.
Abbreviations:
- BrdU,
5′-Bromo-2′deoxyuridine;
- ISH,
in situ hybridization;
- P,
postnatal day;
- pc,
post coitum;
- PGC,
primordial germ cell;
- TG,
transgenic;
- TUNEL,
terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate nick end labeling;
- SCF,
stem cell factor;
- WT,
wild-type.
