Doxorubicin Induces Apoptosis in Germ Line Stem Cells in the Immature Rat Testis and Amifostine Cannot Protect against This Cytotoxicity

Introduction

Improvements in the long-term prognosis for childhood cancer have led to an increasing awareness of the consequences of successful therapy later in life. Male infertility is one of these major effects ( 1– 3), but little is known about the mechanisms by which chemotherapeutic agents used to treat childhood cancer damage spermatogenesis in the prepubertal human testis. Permanent tubular damage may first become apparent during puberty when size of the testis remains prepubertal and sperm production does not initiate ( 1, 4). The lack of appropriate methodology for detecting early cytotoxic damage to the immature testis has made it difficult to identify the exact maturational events on which these anticancer drugs exert their gonadotoxicity.

Induction of programmed cell death (apoptosis) is one of the earliest signs of genotoxic damage to the mature testis ( 5– 8). In the case of the adult rat testis, the cytotoxic drug doxorubicin induces apoptosis in meiotically dividing spermatocytes and type A and intermediate spermatogonia ( 7) by intercalating into DNA to produce strand breaks and by inhibiting topoisomerase II activity ( 9). These genotoxic changes up-regulate expression of tumor suppressor p53, an essential mediator of cell cycle arrest designed to prevent DNA replication in the presence of DNA damage and leading to apoptosis ( 10).

In the immature testis, apoptosis also plays an important physiologic role in connection with maturation of the spermatogenic epithelium ( 11, 12). This physiologic apoptosis culminates at 3 weeks of age in the rat ( 13) and exhibits a pronounced dependency on the stage of spermatogenesis ( 14). Development of a reliable procedure for monitoring and quantitating both this physiologic and drug-induced apoptosis would be of considerable value for research and clinical purposes.

In the present investigation, we have employed light, phase-contrast, and electron microscopy of sequentially dissected microsegments of a single long seminiferous tubule to determine whether physiologic apoptosis in the immature rat testis is influenced by a single i.p. injection of doxorubicin. An intermediate dose of doxorubicin (3 mg/kg), shown earlier to exert long-term testicular toxicity in immature rats, was chosen ( 15, 16). The primary aim was to identify the target cells and maturational stages which are most sensitive to doxorubicin-induced toxicity and to elucidate the underlying biochemical pathways. Finally, possible protection of the immature testis from doxorubicin-induced apoptosis by the recently introduced cytoprotective drug amifostine (Ethyol) was also examined. Contrary to many other normal tissues, testes of young rats do not seem to be protected by this drug against the late toxic effects of doxorubicin ( 15, 17).

Materials and Methods

Animals and treatment. Male Sprague-Dawley rats (BK-Universal, Stockholm, Sweden), at different stages of testicular maturation [6-day-old: onset of spermatogenesis (36 animals); 16-day-old: cessation of Sertoli cell division (28 animals); and 24-day-old: first meiotic cycle (28 animals)], were housed in the animal facilities at the Karolinska University Hospital in Stockholm at 25°C with a 12-hour alternating light/dark cycle. The 6- and 16-day-old rats were housed in the same cages as their mothers and the other animals were placed two to a cage. All of the animals were given free access to tap water and standard laboratory chow. Eight rats at the age of 6, 16, and 24 days were treated with a single i.p. injection of doxorubicin (Pharmacia & Upjohn, Stockholm, Sweden) dissolved in saline 3 mg/kg body weight corresponding to body surface area-related doses of 6.7, 9.8, and 11.6 mg/m², respectively ( 18). Similar numbers of rats received i.p. injection of saline alone. For 6- and 16-day-old rats, treatment/control pairs of the same litter were used (maximum two pairs per litter). After 24 or 48 hours, four rats in each experimental group of each age were asphyxiated with CO₂ and the testes were removed.

Following decapsulation, a single long seminiferous tubule selected randomly from one of the testes of each of these animals (four testes in each experimental group) was placed in a Petri dish containing PBS (2 mmol/L NaH₂PO₄, 8 mmol/L Na₂HPO₄, and 150 mmol/L NaCl, pH 7.4) and microdissected into 1-mm sequential segments under a transillumination stereomicroscope. In addition, the contralateral testis of three of these animals (three testes in each experimental group) was fixed for subsequent examination by light and electron microscopy (see below).

Finally, four additional rats at the age of 6 days were treated with the same dose of doxorubicin, and four rats, serving as controls, received saline. One testis of each of these animals (four testes in each experimental group) was removed 48 hours after injection and frozen for subsequent protein analysis (see below).

In a second type of experiment, 6-day-old rats were treated with a single i.p. injection of amifostine (200 mg/kg; Schering-Plough, Stockholm, Sweden) with or without i.p. administration of doxorubicin (3 mg/kg) 15 minutes later (four rats in each group). Forty-eight hours after such treatment, the testes of these animals were removed and decapsulated and segments of a single long seminiferous tubule were prepared as described above.

The animal experiments done in this study were approved by the Northern Stockholm Animal Ethics Committee (project no. 169/97).

Determination of the plasma concentration of doxorubicin. Employing heart puncture, blood samples were collected from 12 animals in each age group 20 minutes after the i.p. injection of doxorubicin (3 mg/kg) for analysis of the plasma levels of this drug using reversed-phase liquid chromatography with fluorometric detection ( 19). Briefly, 100 μL of plasma were mixed with the internal standard (daunorubicin dissolved in 0.1 mol/L phosphoric acid) and this mixture then transferred into a Sep-Pak C18 extraction column (Waters, Inc., Milford, MA). After rinsing the column with 5 mL phosphate buffer (pH 7.0), the anthraquinone glycosides were eluted with 4 mL methanol. Thereafter, this elute was evaporated and the residue redissolved in 0.1 mol/L phosphoric acid and injected into a Nova-Pak Phenyl Radial-Pak Cartridge (Waters). Acetonitrile (normally 40%) in 0.01 mol/L phosphoric acid was employed as the mobile phase. In some runs, minor adjustment of the acetonitrile concentration in this mobile phase was necessary to achieve optimal chromatographic resolution of the anthraquinone glycosides. The fluorometric detector (Shimadzu spectrofluorometric detector, model RF-551, Shimadzu Corporation, Kyoto, Japan) was operated at emission/detection wavelengths of 501/600 nm.

Staging of the seminiferous tubule, identification of apoptotic cells, and in situ 3′-end labeling of DNA (terminal deoxyribonucleotidyl transferase–mediated dUTP nick end labeling analysis). After isolation, each 1-mm tubular segment was carefully squashed under a coverslip and examined under a phase-contrast microscope as described earlier ( 14). The segments from 6-day-old testis all exhibited the same morphologic appearance, whereas samples from 17- to 18-day-old and 25- to 26-day-old testes were divided on the basis of morphologic criteria into spermatogenetic stages II to VI, VII to VIII, and IX to XI ( 14). The characteristic cells present were, in the case of stages II to VI, intermediate and type B spermatogonia and early pachytene spermatocytes; for stages VII to VIII, preleptotene and mid-pachytene spermatocytes; and for stages IX to XI, type A spermatogonia, leptotene, and zygotene spermatocytes. The numbers of apoptotic cells per millimeter of tubule were determined by phase-contrast microscopy. Under a phase-contrast microscope, apoptotic type A spermatogonia displayed accumulation of heterochromatin followed by condensation of DNA at the periphery. Apoptotic spermatocytes are characterized by elevated amounts of heterochromatin inside the nuclear envelope ( 14). The values thus obtained were confirmed by in situ end-labeling of DNA strands [terminal deoxyribonucleotidyl transferase–mediated dUTP nick end labeling (TUNEL) analysis] of these same samples as described earlier ( 14).

Electron microscopy. In preparation for electron microscopic examination, testes were fixed by immersion in 5% glutaraldehyde in an s-collidine buffer (0.16 mol/L, pH 7.4) at 20°C. Following 30 minutes of such treatment, the tissue samples were cut into ∼1 mm³ cubes and thereafter reimmersed in the same fixative for an additional 2 hours. Postfixation was done with 1% osmium tetroxide in 1.5% aqueous potassium ferrocyanide and the samples then embedded in epoxy resin (Glycidether 100, Merck, Darmstadt, Germany). Ultrathin (70 nm) sections were then prepared (Reichert Ultracut E ultramicrotome, Reichert Jung, Vienna, Austria), stained with uranyl acetate and lead citrate, and finally examined under a JEOL 100SX electron microscope (JEOL, Tokyo, Japan).

Histologic assessment of the numbers of surviving germ cells and Sertoli cells. Semi-thin (1 μm) sections prepared from samples of the most immature testes embedded in epoxy resin as described above were stained with toluidine blue and examined with oil immersion under a light microscope. In each cross section of the seminiferous tubule, cells exhibiting nuclear morphology typical of gonocytes and type A spermatogonia ( 20) were counted and assigned together to the single category of germ cells. The location of these germ cells, either in contact with the basement membrane (peripheral germ cell) or elsewhere in the seminiferous tubule (pericentral germ cell), was noted. The numbers of Sertoli cells present in each cross section of seminiferous tubule were determined on the basis of their typical nuclear morphology.

Western blotting. Total testicular protein was extracted employing a modified radioimmunoprecipitation assay buffer, as described earlier ( 14). Aliquots from each animal, containing a total of 30 μg protein in the case of blotting for p53 and murine double minute-2 (MDM2) and 50 μg protein for caspase 8, were subjected to SDS-PAGE on 12% gels under reducing conditions. Subsequently, the protein bands thus resolved were transferred electrophoretically onto polyvinylidene difluoride (PVDF) membranes and blocked overnight at 4°C in TBS (150 mmol/L NaCl in 10 mmol/L Tris, pH 7.5) containing 5% nonfat dry milk. These membranes were then incubated with rabbit polyclonal antibodies against p53 and caspase 8 and with mouse monoclonal antibody against MDM2 (diluted 1:2,000 in the case of p53 and caspase 8, and 1:1,000 for MDM2; Santa Cruz Biotechnology, Santa Cruz, CA) in TBS at room temperature for 1 hour, washed five times with TBS containing 0.1% Tween 20, and then incubated with horseradish peroxidase–conjugated goat anti-rabbit (diluted 1:10,000; Transduction Laboratories, San Diego, CA) or anti-mouse (Santa Cruz Biotechnology; diluted 1:5,000) secondary antibodies. The antigen-antibody complexes were visualized employing enhanced chemiluminescence (Amersham Biosciences, Uppsala, Sweden), exposed to photographic film, and the signals quantitated by densitometry. The PVDF membranes were also stained with Coomassie blue to confirm equal protein loading.

Statistical analysis. In the case of staging spermatogenic development and microscopic assessment of apoptosis, at least 10 replicate samples, and in the case of TUNEL staining, at least 6 replicate samples, from each of the 4 rats in each experimental group were examined. Three samples from each of three different animals were subjected to electron and light microscopic examination; immunoblotting was done on the total testicular protein fraction from four different animals; and the plasma concentration of doxorubicin in 12 animals was determined. The numbers of germ and Sertoli cells in the testes of 6-day-old treated rats were evaluated in 100 tubular cross sections from each animal and are expressed as mean ± SE per 100 cross sections. Light and electron microscopy was conducted by one observer and TUNEL staining and analysis in a blinded fashion by another observer.

The quantitative data in the figures are presented as mean ± SE; the nonparametric data documented in the table represent median values together with the 25% and 75% percentiles. The Mann-Whitney U test was employed for single statistical comparison of independent groups of samples and the Kruskall-Wallis analysis with Dunn's post hoc test for multiple comparisons of independent groups of samples. P < 0.05 was considered to indicate a statistically significant difference.

Results

Effects of a single i.p. dose of doxorubicin on the testis of 6-day-old rats. Forty-eight hours after a single i.p. injection of doxorubicin, the number of apoptotic germ cells in sequential segments of single long seminiferous tubules, as determined by phase-contrast microscopy ( Fig. 1E-G ), was significantly greater than in control samples ( Fig. 1A-C; Table 1 ). Only physiologic apoptosis of gonocytes was observed 24 hours following doxorubicin treatment or without such treatment. There was no statistically significant difference in the number of apoptotic cells in the same samples as determined morphologically or employing the TUNEL procedure ( Table 1). This enhanced apoptosis in 6-day-old testis 48 hours after doxorubicin treatment was accompanied by an elevation in the level of cleaved caspase 8 ( Fig. 4C).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Phase-contrast micrographs of living cells in microsegments of the testis of control and doxorubicin-treated rats. A to C, testis from 8-day-old saline-treated rats, containing gonocytes (G₀) surrounded by type A spermatogonia (A) and Sertoli cells (Sc). E to G, testis from 8-day-old rats 48 hours after treatment with doxorubicin (3 mg/kg). Visualization was done before (A and E) and after squashing (B-C and F-G) these segments. In the doxorubicin-treated testes (E), apoptotic cells (arrow) appeared in patches and, after squashing (G), could be identified as spermatogonia demonstrating accumulation of heterochromatin inside the nuclear envelope (a), followed by condensation of DNA at the periphery, resulting in the formation of bright, phase-negative apoptotic bodies (arrows). Gonocytes (G₀) exhibiting both intact and apoptotic morphology can be seen. Bar, 20 μm. TUNEL staining of squashed seminiferous tubules from control (D) and doxorubicin-treated rats (H).

Doxorubicin-induced apoptosis could be observed clearly through the basement membrane under a phase-contrast microscope even without squashing ( Fig. 1E). Following squashing ( Fig. 1F), the morphology of the apoptotic cells ( Fig. 1G) was similar to that of gonocytes and spermatogonia undergoing physiologic apoptosis, as described earlier ( 14). However, the distribution of these apoptotic cells was patchy ( Fig. 1E), giving rise to large variations in their numbers in different tubular segments ( Table 1).

Examination of the ultrastructural morphology of the spermatogonia confirmed the apoptotic nature of the death they were undergoing. Nuclear fragmentation and condensation were visible 24 hours after doxorubicin treatment ( Fig. 2C ) and maximal 48 hours following such injection, at which time many of the apoptotic germ cells were being phagocytized by surrounding Sertoli cells ( Fig. 2D). Spermatogonia were still migrating to the basement membrane 24 hours after doxorubicin treatment, but after 48 hours only individual surviving spermatogonia in contact with the basal lamina were observed ( Fig. 2D). At this same time point, round gonocytes located pericentrally in the seminiferous tubules still showed normal morphology ( Fig. 2D). Although no Sertoli cells with apoptotic morphology were observed, the nucleus of many Sertoli cells did exhibit ultrastructural changes. The large reticular nucleus with central condensation typical of control Sertoli cells ( Fig. 2A and B) was small and fragmented both at 24 ( Fig. 2C) and 48 hours ( Fig. 2D) following doxorubicin treatment.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Electron micrographs of the testis of control and doxorubicin-treated rats. In the testis of a 7-day-old control rat, a nonapoptotic gonocyte (A) and a smaller type A spermatogonia (B), both in close contact with the basement membrane and containing large mitochondria with vertical lamellae (arrow), can be seen. The morphology of the nucleolus of Sertoli cells in these same samples (*) is normal. C, in the testis of the 7-day-old rat 24 hours after treatment with doxorubicin (3 mg/kg), early alterations in the morphology of the nucleolus (*) of Sertoli cells, together with an apoptotic germ cell still containing large identifiable mitochondria with vertical lamellae (arrow) and being phagocytized by the Sertoli cells, can be seen. D, in the testis of the 8-day-old rat 48 hours after treatment with doxorubicin, the phagocytized cells have been transformed into dense apoptotic bodies and germ cells have disappeared from the vicinity of the basement membrane. Only a large round gonocyte located pericentrally shows normal morphology (G₀). Bar, 4 μm.

When semi-thin sections were examined under a light microscope, a statistically significant (P < 0.05) decrease in the total number of surviving germ cells was observed 48 hours after doxorubicin treatment ( Fig. 3 ). The number of germ cells located in the vicinity of the basement membrane was also reduced by doxorubicin treatment, whereas no change in the mean number of gonocytes located pericentrally was observed. The total numbers of germ cells in control samples and 24 hours after doxorubicin treatment were comparable. A physiologic increase (P < 0.05) in peripherally located spermatogonia was observed during postnatal days 6 to 8 in the saline-treated controls ( Fig. 3).

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

Numbers of spermatogonia located pericentrally and peripherally in the testis 24 and 48 hours after injection of saline (ctrl) or doxorubicin (doxo) into 6-day-old rats. Points, mean per 100 cross sections of seminiferous tubules; bars, SE. *, P < 0.05, compared with the corresponding value for doxorubicin-treated animals at the same time point.

The statistically significant (P < 0.05) physiologic increase in Sertoli cell number that occurred in saline-treated controls was apparently delayed by doxorubicin. A statistically significant (P < 0.05) decrease in the total number of Sertoli cells was observed 48 hours after doxorubicin treatment. Sertoli cells per 100 cross sections were 306 ± 12 on the day of injection; 335 ± 7 and 404 ± 11 at 24 and 48 hours after saline treatment, respectively; and 318 ± 1 and 346 ± 10 at 24 and 48 hours following injection of doxorubicin, respectively.

The testicular level of p53 ( Fig. 4A ) was elevated significantly (P < 0.05), whereas that of the MDM2 protein ( Fig. 4B) that down-regulates p53 was reduced (P = 0.057) 48 hours after injection of doxorubicin into 6-day-old rats.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.

The patterns of expression of p53 (A), MDM2 (B), and cleaved caspase 8 (C) in the testis of 6-day-old rats 48 hours following treatment with doxorubicin (3 mg/kg). Top, densitometric quantitation of the immunoblotted bands; columns, mean densities (n = 4); bars, SE. Bottom, autoradiograms.

Lack of effect of a single dose of doxorubicin on the testes of 16- and 24-day-old rats. In the case of 16-day-old rats, the number of cells in the testis exhibiting apoptotic morphology or TUNEL staining 24 and 48 hours after doxorubicin treatment was similar to that observed in saline-treated control animals ( Table 1). When the seminiferous tubules of doxorubicin-treated and control 16-day-old rats were dissected into three different stages of spermatogenesis as described earlier ( 14), a normal pattern of physiologic apoptosis and normal progression of spermatogenesis were observed in both cases ( Table 1). The level of apoptosis in stages IX to XI was significantly higher than in stages II to VI ( Table 1) and the apoptotic cells observed in stages VII and VIII could be identified as mid-pachytene spermatocytes, whereas those in stages IX to XI were type A spermatogonia. The Sertoli cells were elongated, with the nucleoli characteristic for mature cells of this type, and exhibited no apoptosis or mitosis. In addition, electron microscopic examination revealed that the ultrastructural morphology of the testes of 16-day-old rats was also unaffected by a single doxorubicin treatment (not shown).

In the case of 24-day-old rats as well, morphologic characterization revealed no statistically significant alterations in total or stage-specific apoptosis in the testis 24 or 48 hours after injection of doxorubicin ( Table 1). Furthermore, neither the progression of spermatogenesis nor the ultrastructural morphology of the testis of 24-day-old rats was influenced by doxorubicin (not shown).

Lack of protection of the testis of 6-day-old rats by amifostine from doxorubicin-induced apoptosis. When 22 sequential segments of a single long seminiferous tubule were examined morphologically 48 hours after injection of doxorubicin into 6-day-old rats that also received a single dose of amifostine (200 mg/kg) 15 minutes before doxorubicin treatment, the median number of apoptotic germ cells per millimeter of tubule (confidence limit: 25%,75%) was found to be 20 (5,32). This value did not differ from that obtained after treatment with doxorubicin alone ( Table 1) and was significantly higher (P < 0.05) than the corresponding value for animals treated with saline ( Table 1) or amifostine alone [3 (1,10); n = 20]. Thus, the testicular morphology observed 48 hours after injection of doxorubicin or saline into 6-day-old rats was unaffected by pretreatment with amifostine.

Plasma levels of doxorubicin. Twenty minutes after i.p. injection of doxorubicin (3 mg/kg), the plasma concentration of this drug in 6-day-old rats (190 ± 23 ng/mL) was found to be significantly lower (P < 0.05) than in 16-day-old (487 ± 40 ng/mL) or 24-day-old rats (390 ± 50 ng/mL).

Discussion

The present observation that a single i.p. dose of doxorubicin (3 mg/kg) increases the level of apoptosis significantly only in testes of the youngest (6-day-old) rats tested suggests that the initiation phase of spermatogenesis is especially sensitive to the toxic effects of this compound. Doxorubicin-induced loss of germ cells was localized to the migrating spermatogonia, which constitute the major stem cell pool for the germ line in the immature testis ( 20– 22). The patchy arrangement of apoptotic cells along the seminiferous tubule ( Fig. 1E) most likely represents cohorts of synchronously dying damaged spermatogonia. The physiologic increase in spermatogonial number that normally occurs on postnatal day 8 was also blocked by doxorubicin. Enhanced expression of the p53 protein, in association with reduced expression of the MDM protein which down-regulates p53, suggests that the tumor suppressor p53 mediates the arrest in the germ cell cycle and subsequent apoptosis.

Observed morphologically unaffected gonocytes were still in the central compartment of the seminiferous tubules 48 hours after doxorubicin treatment. These round, nonmigrating gonocytes in the 8-day-old rat testis are believed to undergo physiologic degeneration ( 20, 22) and seem to be incapable of replacing losses in the stem cell pool because no restoration of spermatogenesis after sexual maturation was observed in rats treated in a similar fashion in a previous study ( 15). Our present findings indicate that the ablation of the spermatogonial pool in the immature testis caused by doxorubicin is immediate and may be selective with regards to the type and localization of affected germ cells within the seminiferous tubule.

In the older rats, the same relative dose of doxorubicin (3 mg/kg) had no influence on the physiologic, stage-specific apoptosis occurring in type A spermatogonia and mid-pachytene spermatocytes in the spermatogenic epithelium ( 14). Nor did doxorubicin treatment increase the frequency of apoptosis in other types of germ cells. These observations suggest that the stem cell spermatogonia of 16- and 24-day-old rats are more resistant to the toxicity of doxorubicin than are the migrating spermatogonia in 6-day-old rats.

This difference cannot be explained by pharmacokinetic factors because 20 minutes after injection, the plasma levels of doxorubicin in the older rats were higher than in the 6-day-old animals. Higher plasma levels in the older rats could be expected because the increase in body weight associated with increasing age is known to lead to higher drug dose when dosing is based on body weight instead of body surface area ( 18). However, the local concentration of doxorubicin in the intratubular compartment may nonetheless be highest in the youngest rats because the blood-testis barrier is known to develop during puberty. Thus, tight junctions between Sertoli cells, which create a specialized environment for differentiating spermatocytes with respect to electrolytes, substrates for intermediary metabolism, and regulatory factors, are known to appear in rats first on postnatal day 15 ( 23). The endothelial cells are an additional component of the blood-testis barrier that matures in connection with pubertal development ( 24) and can influence the local conditions to which spermatogonia are exposed. Moreover, the physiologically cryptorchid testis of the youngest rats may also experience higher local concentrations of doxorubicin due to the fact that i.p. administration was employed here ( 15).

A previous study about testicular susceptibility to the toxicity of doxorubicin at critical stages of maturation ( 16) suggests that the elevation in germ cell apoptosis observed in the present investigation may reflect a more long-lasting damage to the spermatogenic epithelium. Thus, a similar single dose of doxorubicin (3 mg/kg) was reported to impair the development of reproductive functions in the youngest 6-day-old rats as well as to reduce the fertility of 24-day-old rats, if only temporarily ( 16). The reduced rate of increase in the number of Sertoli cells and the ultrastructural changes exhibited by the nucleoli of these cells in response to doxorubicin observed here, together with the reduction in the level of androgen-binding protein caused by this drug, reported previously ( 16), suggest that doxorubicin damages immature Sertoli cells. Disruption of Sertoli cell function by toxicants can reduce their capacity to nurse germ cells, leading to apoptosis of the latter ( 25). An age-dependent decrease in busulfan-induced damage to stem cell niches has earlier been reported in immature rats ( 26). Altogether, 80% of such rats treated prenatally and 40% of those treated at postnatal days 8 to 12 were not able to support endogenous or grafted stem cell spermatogonia and became infertile. This same period of testicular maturation is also known to be sensitive to other cancer treatments, including irradiation and chemotherapy with procarbazin ( 27, 28). Thus, the presently observed more pronounced sensitivity of germ cells in the 6-day-old rat testis to doxorubicin-induced apoptosis may also be due to the presence of particularly sensitive stem cell niches.

In humans, the phase of testicular development, during which gonocytes resume their proliferation, migrate to the basement membrane, and give rise to spermatogonia, lasts much longer (i.e., from birth up to the age of 4-5 years) and does not overlap with the initiation of puberty, unlike the case in rodents ( 29– 31). An additional period of proliferation and migration of spermatogonia located pericentrally has been described in boys 8 to 9 years of age, suggesting that the spermatogonial population in humans may be constantly renewed in preparation for the onset of puberty ( 30).

Neither of these maturational periods is known to be associated with increased sensitivity of the human testis to cancer therapy. However, interpretation of the clinical data is difficult because pediatric studies often do not adequately document pubertal status at the time of therapy ( 3). Reported variations in the risk for infertility in individuals exposed to similar doses of cytotoxic drugs may possibly be explained, at least in part, if the sensitivity of maturing human spermatogonia to such drugs were better characterized ( 31, 32).

The cytoprotective organic thiophosphate amifostine exerted no influence on the apoptosis induced by doxorubicin in gonocytes and spermatogonia in the testes of 6-day-old rats. This observation is in good agreement with previous findings that amifostine does not protect the rat testis against the severe impairment of reproductive functions caused by doxorubicin ( 15, 16). In adult mice, treatment with amifostine alone has been associated with cytotoxicity and a decrease in stem cell survival ( 33). In contrast to this finding, we found that physiologic apoptosis in the immature 6-day-old rat testis was not affected by amifostine (200 mg/kg) either, leading support to our previous observation that amifostine does not disrupt normal testicular maturation in the rat ( 15).

In summary, the present investigation reveals that the initiation phase of rat spermatogenesis is more sensitive to doxorubicin-induced apoptosis than are the later stages in this process. The stem cell pool of the germ line, together with migrating gonocytes and spermatogonia, are the cells that are vulnerable to this induced apoptotic death, and the size of the stem cell pool is reduced by this cytotoxic insult. Our findings suggest that a decrease in the number of spermatogonia, enhanced germ cell apoptosis, and elevated levels of p53 are early indicators of a permanent cytotoxic damage to the immature spermatogenic epithelium in the rat.