Exposure to Retinoic Acid in the Neonatal but Not Adult Mouse Results in Synchronous Spermatogenesis

Abstract

Retinoic acid (RA) is required for germ cell differentiation, the regulation of which gives rise to a constant production of mature sperm. In testes from 3-day postpartum (dpp) RARE-hsplacZ mice, periodic regions positive for beta-galactosidase activity were observed along the length of the seminiferous tubules. Periodicity was abolished by treatment of neonates with exogenous RA at 2 dpp. To assess the consequences, 2-dpp mice were treated with RA, and the long- and short-term effects were assessed. Long-term effects of neonatal RA exposure included a delay in the appearance of advanced germ cells and the absence of a spermatogenic wave (synchronous spermatogenesis) in the adult. In contrast, RA exposure in vitamin A-sufficient adults did not result in synchronous spermatogenesis but rather induced apoptosis in a subset of spermatogonia. Shortly after (24 h) neonates were exposed, altered expression of known germ cell differentiation and the (Stra8, Kit, Sycp3, and Rec8) meiosis markers and an increase in the number of STRA8 and SYCP3 immunopositive cells were observed relative to those of vehicle controls. However, 48 and 72 h after exposure, a significant reduction in the number of STRA8 and SYCP3 immunopositive cells occurred. Immunohistochemical analysis of a marker for apoptosis demonstrated neonatal exposure resulted in increased germ cell apoptosis, as observed in the adult. Additionally, RA exposure resulted in increased Cyp26a1 expression of the RA-degrading enzyme. Thus, while RA treatment of neonatal and adult mice resulted in apoptosis of spermatogonia, synchronous spermatogenesis occurred only after neonatal RA exposure.

Exposure of neonatal but not adult mice to exogenous retinoic acid leads to synchronous spermatogenesis via altered initiation of the spermatogenic wave.

INTRODUCTION

Spermatogenesis, the process by which spermatogonial stem cells differentiate into spermatozoa, is fundamental for male fertility. In adult men and mice, the production of spermatozoa is constant due to temporal cycling, termed the “spermatogenic cycle,” and spatial phasing, termed the “spermatogenic wave,” of germ cell differentiation [1, 2]. The result of temporal and spatial constraints is the cyclic appearance of specific cellular associations within a given cross-section of a seminiferous tubule. These recurring cellular associations are termed the “stages of the cycle of the seminiferous epithelium,” 12 of which are recognized in the mouse [3, 4]. In normal adult mice, all 12 stages are present at any given time and occur sequentially along the length of the seminiferous tubule. The distance between a given stage and the reappearance of that stage is termed a “spermatogenic wave.” The processes that give rise to the spermatogenic wave result in continuous or asynchronous release of spermatozoa from the seminiferous tubules.

Vitamin A is required for spermatogenesis [5], and vitamin A-deficient (VAD) rats and mice are infertile due to a failure in germ cell differentiation [6]. In both VAD rats and mice, treatment with either vitamin A in the form of retinol (ROL) [7, 8] or multiple treatments with retinoic acid (RA) [9], the biologically active form of vitamin A, in conjunction with dietary retinoid replenishment, results in the reinitiation of the spermatogenic cycle. However, at any single point in time, testes from these animals contained fewer stages of the cycle of the seminiferous epithelium occurring at much higher frequencies than observed in normal individuals. While progression through the stages of the cycle was normal in these animals, the spermatogenic wave had been eliminated, effectively synchronizing spermatogenesis. Synchronous spermatogenesis results in a testis where spermiation occurs episodically as opposed to normal asynchronous spermatogenesis, where spermiation is continuous.

Although questions remain about the exact mechanism driving synchrony in VAD models, it is clear that synchronous spermatogenesis in VAD-ROL-replenished individuals is the result of spatially uniform initiation of spermatogonial differentiation throughout the testis [8]. This appears to be a VAD-specific event, as exogenous RA exposure in vitamin A-sufficient adults results in stage-specific responses in both the spermatogonial and the preleptotene spermatocyte population [10]. Observations of VAD and VAD-ROL-replenished adults led to a large body of evidence linking RA to germ cell differentiation, particularly in embryonic and neonatal systems. In both male and female germ cells, RA appears to be a trigger for meiotic entry [11–13]. Additionally, RA induces the transition from undifferentiated (A_al) spermatogonia to differentiating (A₁) spermatogonia in the neonatal testis [14, 15]. Recent studies examining the distribution of RA signaling within the neonatal testis demonstrate that RA signaling occurs in an uneven pattern across the neonatal testis and may be the foundation for the spermatogenic wave in the adult [15].

Patterns of RA signaling within the neonatal testis presumably result from gradients of RA and could potentially be disrupted with exogenous RA, theoretically driving the entire undifferentiated germ cell population to enter the differentiation pathway. This study focuses on the adult murine testis of individuals exposed neonatally or in adulthood to RA, potentially disrupting RA gradients along the seminiferous epithelium and altering germ cell differentiation. This report is the first to demonstrate neonatal but not adult exposure to RA results in synchronous adult spermatogenesis and to examine the mechanism by which synchrony is established.

MATERIALS AND METHODS

Animal Treatment, Tissue Collection, Fixation, and Staining

All animal experiments were approved by the Washington State University Animal Care and Use Committee and carried out in accordance with the standards set by the National Institutes of Health. Testes of Tg(RARE-Hspalb/lacZ)12Jrt (hereafter referred to as RARE-hsplacZ) animals at 2, 5, and 10 days postpartum (dpp) (n ≥ 2) were collected and fixed in a solution of 4% paraformaldehyde (PFA) with 0.25% glutaraldehyde for 1 h. For isolated tubule analysis, RARE-hsplacZ testes were detunicated, mechanically disassociated, and fixed in the PFA plus glutaraldehyde solution for 15 min. Fixed tissue was washed and stained in bromo-chloro-indolyl-galactopyranoside (X-gal) (as previously described [16]) and then washed and imaged. Treated animals consisted of both RARE-hsplacZ and BL6/129 neonates and BL6/129 adults. All treatments and time points of collected tissue were completed with a minimum of biological triplicates. For neonatal exposure, vehicle (dimethyl sulfoxide [DMSO]) or various doses of all-trans-RA (atRA) (Sigma-Aldrich, St. Louis, MO) in vehicle were injected subcutaneously at a single site. Testes were collected at 12, 24, 48, or 72 h and at 8, 28, 58, and 88 days posttreatment for RNA extraction or fixed with either 4% PFA, 4% PFA with 0.25% glutaraldehyde, or Bouin fixative. For adult exposure, vehicle (oil) or 350 μg of atRA was injected subcutaneously every 24 h for 3 days. Testes were collected 28 days after the first exposure and fixed for 24 h in Bouin fixative. Postfixation, samples were washed, dehydrated, and paraffin embedded for histology.

Histology, Immunohistochemistry, and TUNEL Analysis

Four-micrometer sections of BL6/129 testes were used for all histological or immunohistochemical (IHC) procedures. For IHC procedures, the following antibodies were used with the species-appropriate biotinylated secondary (1:500 dilution; Chemicon): activated CASP3 (rabbit antiactivated CASP3 [Cell Signaling] at 1:100 dilution), SYCP3 (rabbit anti-SCP-3 [Santa Cruz Biotechnology] at 1:500 dilution), STRA8 (rabbit anti-STRA8 (Griswold) at 1:1000 dilution with trypsin [1 mg/ml] digestion for 7 min at 37°C in addition to antigen retrieval), and germ cell nuclear antigen (rat anti-germ cell nuclear antigen [courtesy of G. Enders] at 1:100 dilution). In all cases, 0.01 M sodium citrate heat-mediated antigen retrieval was used, unless otherwise noted. A streptavidin-horseradish peroxidase and diaminobenzidine (Invitrogen) solution was used for visualization. In all cases, quantification was performed with a minimum of 100 tubule cross-sections per biological replicate on a minimum of two planes separated by a minimum distance of 50 μm. All IHC procedures and quantifications were performed with a minimum of biological triplicates and technical duplicates. TUNEL analysis was performed using an ApoAlert DNA fragmentation assay kit (BD Biosciences, Franklin Lakes, NJ) according to the manufacturer's instructions. TUNEL-positive cells per tubule were quantified using a Nikon Microphot-FX microscope (Meridian Instrument Co. Inc., Kent, WA) and a minimum of 100 tubules quantified per biological replicate (n ≥ 4).

General Histology, Spermatid Analysis, Stage Frequency Analysis, and Determination of Synchronization Factor

Four-micrometer sections of paraffin-embedded tissue were counterstained with hematoxylin for histological evaluation. Cross-sections of seminiferous tubules from 30-, 60-, and 90-dpp vehicle control and RA-treated testes were assessed for their spermatid content or stage of the cycle of the seminiferous epithelium based on the criteria described by Russell et al. [3]. A minimum of 200 cross-sections per animal were assessed on a minimum of two planes separated by a minimum of 50 μm. Synchronization factor was determined using the method described by Siiteri et al. [17] and stage frequencies established by Bianchi and Tiglao [18].

RNA Isolation and Real-Time Reverse Transcriptase PCR

Total RNA was extracted using TRIzol (Invitrogen, Carlsbad, CA) (for whole-tissue samples) or Pico Pure (Applied BioSystems, South San Francisco, CA) and quantified by using an ND-1000 spectrometer (Thermo Scientific, Wilmington, DE). For real-time reverse transcriptase (RT)-PCR, forward and reverse primers (shown in Supplemental Table S1, available online at www.biolreprod.org) were designed using Primer Express version 2.0 (Applied BioSystems, Foster City, CA). An iScript cDNA synthesis kit (Bio-Rad, Hercules, CA) and 200 ng of sample RNA were used for cDNA production. Real-time RT-PCR was carried out with a 7500 Fast real-time PCR system and Fast SYBR GREEN Master Mix (Applied BioSystems). Individual real-time RT-PCR sample reactions were analyzed in triplicate, and all treatments or ages were analyzed with a minimum of biological duplicates. Relative fold-change values were calculated using the ΔΔC_⊤ method as described previously [19] with the ribosomal protein S2 (Rps2) gene as the endogenous control and vehicle-treated samples as biological controls.

Neonatal Testis Cultures

Two-days postpartum BL6/129 testes were collected into sterile PBS, detunicated, and cut into four equal pieces under sterile conditions. Four pieces from a total of two testes were placed into the groove of an embryonic gonad culture agar mold (as described previously [20]) presoaked in Dulbecco minimal Eagle medium containing 10% fetal calf serum (Gibco) and 100 mg/ml ampicillin with either vehicle (DMSO) or 0.7 μM atRA. At 24 h after specimens were placed in molds and every 24 h thereafter medium was removed, and fresh medium containing no vehicle or treatment was added. Testes were flash-frozen on dry ice for RNA isolation. All treatment regimens contained a minimum of seven biological replicates obtained across a minimum of three technical replicates.

Statistical Analyses

For all analyses, data were analyzed using Student t-tests to determine significance and to calculate P values. All statistical analysis was completed using JMP version 7.0.1 software (SAS Institute Inc.).

RESULTS

Determination of RA Dose

The optimal dose of RA in the 2-dpp neonate was determined by examining Stra8 expression and the number of germ cells immunopositive for STRA8 at 24 h after treatment with various doses of RA (Supplemental Fig. S1, available at www.biolreprod.org). Stra8 expression was most highly induced with 100 μg of RA; however, induction was also observed at 50- and 200-μg doses of RA. The lowest dose resulting in a significant increase in the number of STRA8-positive cells, 50 μg of RA per individual, was selected for all further analyses.

Physiological Effects of Neonatal Exposure to RA

Testis and body weights as well as testis-to-body weight ratios were assessed for individuals exposed to 50 μg of RA at 2 dpp (Supplemental Fig. S2). With the exception of 8 days after exposure (10 dpp), no significant difference was observed in testis or body weight with exposure to vehicle or RA. At 10 dpp, RA exposure significantly reduced both testis and body weight but had no effect on testis-to-body weight ratio.

β-Galactosidase Activity and the Effect of Exogenous RA in the Transgenic RARE-hsplacZ Mouse Model

β-Galactosidase activity in the RARE-hsplacZ neonatal testis was determined by whole-tissue staining at 2, 5, and 10 dpp (Fig. 1). At all ages examined, β-galactosidase activity was periodic with the frequency and length of patches appearing to increase with age. The effect of exogenous RA on β-galactosidase activity was determined by treating 2-dpp RARE-hsplacZ animals with 50 μg of RA, followed by whole-tissue or whole-tubule staining at 3 dpp (Fig. 2). In vehicle-treated animals, β-galactosidase activity was patchy on the whole-tissue level and roughly periodic in isolated tubules. RA treatment completely altered the normal appearance of β-galactosidase activity on the whole-tissue and the tubule levels.

FIG. 1.

Whole neonatal RARE-hsplacZ testes stained for β-galactosidase activity are shown at (A) 2 dpp, (B) 5 dpp, and (C) 10 dpp (n ≥ 2). Bars = 400 μm.

FIG. 2.

Effects of exogenous RA treatment at 2 dpp on 3-dpp RARE-hsplacZ whole testes and on isolated tubules stained for β-galactosidase activity are shown in (A) vehicle control of whole testis; (B) RA-treated whole testis; (C) vehicle control of isolated tubules; and (D) RA-treated isolated tubules (n = 3). Bars = 400 μm.

Long-Term Effects of Neonatal Exposure to RA

Initial histological analysis of 60- and 90-dpp testes exposed at 2 dpp to either vehicle or RA revealed a significant perturbation of spermatogenesis in RA-treated individuals that was not observed in vehicle controls. The frequency of abnormal tubules in vehicle controls was similar to that in RA-treated individuals; however, the distribution of spermatogenic stages appeared altered with RA treatment. In order to determine if this was the case, stage frequency analysis and synchrony factor calculations (Fig. 3) were used to demonstrate that spermatogenesis in RA-treated individuals was significantly more synchronous than in vehicle-exposed individuals and that this synchrony was maintained until at least 90 dpp.

FIG. 3.

Stage frequencies of 60- and 90-dpp animals exposed to vehicle (white bars) or 50 μg of RA (gray to black bars) at 2 dpp are shown. Vehicle stage frequency is the average of a minimum of biological triplicates. Stage frequency graphs are shown in 60-dpp (A) or 90-dpp (C) individuals. Calculated synchronization factors for vehicle- or RA-exposed individuals are shown at 60 dpp (B) or 90 dpp (D). Error bars represent standard deviations (n ≥ 3). Asterisks represent significance (P < 0.05), and reported values are P values calculated from comparisons between vehicle and treated individuals.

RA Effects in the Adult, Vitamin A-Sufficient Testis

The effects of RA exposure on spermatogenesis in adult, vitamin A-sufficient testis was assessed by exposing mice to three doses of 350 μg of RA over the course of 3 days. After the final dose, TUNEL analysis demonstrated a significant increase in the frequency of TUNEL-positive cells associated with the periphery but not the center of the seminiferous tubule (Fig. 4A). Histological analysis of animals exposed to RA at 28 days after the initial exposure revealed a reduction in round spermatids associated with 20% to 30% of tubules in stages VII and VIII (Fig. 4B). No definitive alteration of stage frequencies was observed in vitamin A-sufficient adults exposed to RA in adulthood.

FIG. 4.

Effects of RA exposure in vitamin A-sufficient adults are shown by (A) quantification of TUNEL-positive cells 3 days after initial RA exposure. Periphery and Center indicate positions within the tubule cross-section of a TUNEL-positive cell. Error bars represent standard deviations (n = 4). Asterisks represent significance (P < 0.05), and reported values represent calculated P values compared to vehicle control. B) Histology of stage VIII tubule cross-sections in control (left panel) are compared with those in RA-exposed (right panel) individuals. Bars = 20 μm.

Short-Term Effects of Neonatal RA Exposure

The effect of RA exposure on the expression and number of cells immunopositive for several differentiation and meiosis markers was assessed in 3-, 4-, and 5-dpp animals exposed to RA at 2 dpp (Fig. 5). At 3 dpp (24 h after exposure), the expression of the differentiation marker, Stra8, and the meiotic marker, Sycp3, and the number of cells that were immunopositive for STRA8 or SYCP3 were increased in RA-treated subjects versus vehicle controls. By 4 and 5 dpp (48 and 72 h after exposure, respectively), RA-treated individuals had a general reduction in the expression of differentiation (Stra8 and Kit) and meiosis markers (Sycp3 and Rec8) relative to vehicle controls. This reduction coincided with a dramatic decrease in the number of cells immunopositive for STRA8 at both 4 and 5 dpp and a reduction in the number of cells immunopositive for SYCP3 at 5 dpp. The number of apoptotic germ cells in vehicle- or RA-exposed individuals was assessed at 3 dpp by IHC detection of the apoptotic marker activated CASP3 (Fig. 6). A significant increase in the number of activated CASP3-positive germ cells was detected after RA exposure compared to that of vehicle controls.

FIG. 5.

Quantitative analysis of expression (A–D) and immunopositive cells (E and F) are shown for markers of differentiation and meiosis after exposure to RA. Age is indicated as days postpartum (dpp). Fold-change values are shown relative to age for (A) Stra8, (B) Sycp3, (C) Kit, and (D) Rec8 genes. All fold-change values are relative to 3-dpp vehicle control. Errors bars represent standard errors of the means (n ≥ 2). Asterisks represent significance (P < 0.05), and reported values represent calculated P values. Numbers of immunopositive cells per tubule cross-section relative to age are shown for (E) STRA8 and (F) SYCP3. Error bars represent standard deviations (n ≥ 3). Asterisks represent significance (P < 0.05), and reported values represent calculated P values compared to vehicle control within a time point.

FIG. 6.

IHC detection and quantitative analysis of germ cells immunopositive for activated CASP3 at 24 h after RA exposure are shown in (A) vehicle controls and (B) cells exposed to 50 μg RA. Bars = 40 μm. C) Quantification of immunopositive germ cells per tubule cross-section are shown. Error bars represent standard deviations (n = 4). Asterisk represents significance (P < 0.05), and reported value represents calculated P value.

Delayed but Synchronized Spermatogenesis with Neonatal RA Exposure

Examination of germ cell maturation at 10 dpp by analysis of Stra8 and Sycp3 expression using real-time RT-PCR determined that the expression of both markers had recovered to control levels in RA-exposed individuals 8 days after exposure. However, IHC analysis of STRA8 and SYCP3 at the same time point demonstrated that tubule-to-tubule distribution of both was more even in RA-exposed than in vehicle control testis (Fig. 7). Additionally, the number of STRA8 or SYCP3 immunopositive cells not associated with the basement membrane was significantly reduced with RA treatment relative to that of control. As a second measure of germ cell maturation, the spermatid population in 30-dpp individuals exposed to RA at 2 dpp was assessed by histological evaluation. No dramatic loss or increase in spermatid number was observed with RA exposure (data not shown); however, while step-16 spermatids were observed in the majority of vehicle-exposed animals, testes of RA-exposed individuals rarely contained spermatids more advanced than step 12. Second, the range of observed spermatid types was significantly reduced in RA-exposed individuals relative to that of controls (P < 0.0005 by t-test), with testes of RA-exposed individuals containing an average of 7 different spermatid steps and vehicle-exposed individuals containing an average of 11.17 spermatid steps.

FIG. 7.

Effects of RA exposure at 2 dpp on spermatogenesis at 10 dpp are shown for Stra8 and Sycp3 expression and STRA8 (A) and SYCP3 (B) immunopositive cells 8 days after neonatal exposure to RA (10 dpp). Bars = 40 μm. Insets: secondary antibody-only controls, scaled to 20%. C) Results are shown for Stra8 and Sycp3 real-time RT-PCR analyses. All fold-change values are relative to vehicle control. Errors bars represent standard errors of the means (n = 3). D) Quantification of immunopositive cells not associated with the basement membrane is shown. Error bars represent standard deviations (n ≥ 4). Asterisks represent significance (P < 0.05), and reported values represent calculated P values compared to vehicle control within a time point.

Alteration of RA Metabolism Enzyme Expression with Neonatal RA Exposure

The expression of two known RA-degrading enzyme-encoding genes, Cyp26a1 and Cyp26b1, was assessed at 12 h (2.5 dpp) and at 1 (3 dpp), 2 (4 dpp), 3 (5 dpp), and 8 (10 dpp) days after RA exposure (Fig. 8). The Cyp26a1 gene expression rapidly increased shortly (12 h) after RA exposure; however, by 4 and 5 dpp, its expression was reduced in RA-exposed individuals compared to that in vehicle controls. In contrast Cyp26b1 expression remained relatively stable throughout the examined time points.

FIG. 8.

Effect of RA exposure at 2 dpp on the expression of RA-degrading enzymes is shown by real-time RT-PCR analysis of (A) Cyp26a1 expression. Value given at 2.5 dpp represents the average fold-change ± the standard errors of the means. B) Cyp26b1 expression is shown. Fold-change values are shown relative to within-age vehicle controls. Error bars represent standard errors of the means (n ≥ 3). Asterisks indicate significance (P < 0.05), and reported values represent calculated P values relative to vehicle control.

RA Effects on Long-Term Cultures of Neonatal Testis

The effect of a single dose of RA followed by a recovery period on Stra8 and Cyp26a1 expression in cultured 2-dpp neonatal testis was assessed using a 72-h culture regimen that consisted of 24-h exposure to vehicle or 0.7 μM RA, followed by 48 h of culture in the absence of treatment (Fig. 9). In the case of Stra8 expression, a single exposure of RA followed by an absence of treatment resulted in a significant reduction in expression, compared to Cyp26a1 expression, which was induced with RA treatment.

FIG. 9.

Effect of 24-h RA exposure followed by a 48-h recovery in cultured neonatal testis is shown. Real-time RT-PCR analyses show Stra8 and Cyp26a1 expression levels. Error bars represent standard errors of the means (n ≥ 7). Asterisks indicate significance (P < 0.05), and reported values represent calculated P values relative to vehicle control.

DISCUSSION

The current study demonstrated that RA availability is roughly periodic along the length of the neonatal seminiferous tubule and that periodicity is eliminated with neonatal RA exposure. Additionally, observations in this work show that exposure of the neonate but not the adult to RA resulted in synchronized adult spermatogenesis. As in other synchronized systems, synchrony was ultimately the result of elimination of the spermatogenic wave but maintenance of the spermatogenic cycle.

β-Galactosidase activity in the neonatal testis of RARE-hsplacZ mice has previously been shown to be an accurate measure of available RA and was associated predominantly with the germ cell population [15]. The intermittent appearance of β-galactosidase activity along the seminiferous tubule further suggests that localized RA signaling along the seminiferous tubule plays a fundamental role in establishing the spermatogenic wave. Treatment with exogenous RA eliminated the periodic appearance of β-galactosidase activity, demonstrating that exogenous RA can effectively alter the distribution of RA in the neonatal testis and potentially alter the establishment of the spermatogenic wave.

The effect of neonatal RA exposure on adult spermatogenesis was both striking and consistent. Unlike in adult vitamin A-sufficient animals exposed to RA during adulthood, all adult individuals exposed to RA as neonates displayed highly synchronous spermatogenesis. Several RA-induced events were observed during the neonatal period including germ cell differentiation, germ cell apoptosis, and alteration of RA metabolism enzyme expression. RA-induced events did not appear to coincide with altered germ cell proliferation. RA-induced germ cell differentiation has been demonstrated by multiple groups in both embryonic gonads and neonatal testis [10–14, 21]. Induction of genes associated with germ cell differentiation (Stra8 and Sycp3) and an increase in the number of germ cells immunopositive for differentiation markers (STRA8 and SYCP3) at 24 h after RA exposure suggest that exogenous RA treatment induced germ cell differentiation in vivo. However, as observed in this work, 48 and 72 h after RA exposure, the expression of genes associated with differentiation or meiosis (Stra8, Kit, Rec8, and Sycp3) and the number of germ cells immunopositive for differentiation or meiosis markers (STRA8 and SYCP3) decreased relative to those of vehicle controls. This outcome appears to be the result of a direct action by RA on the testis and not systemic RA exposure effects, as cultured neonatal testes exposed to a single dose of RA followed by a recovery period displayed a reduction in Stra8 expression similar to that observed in vivo.

RA-induced germ cell apoptosis is observed with treatment of both adults and neonates. This result fits well with previous reports showing increased germ cell apoptosis in neonatal testicular suspensions cultured with RA [22] and transgenic animals lacking RA-degrading enzymes [23, 24]. One observable outcome of neonatal germ cell apoptosis is an apparent delay in spermatogenesis, observable 8 and 28 days after RA exposure as a decrease in preleptotene and leptotene spermatocytes and a delay in the appearance of advanced spermatids, respectively. The outcome of RA-induced apoptosis in the adult was a loss, in at least some tubule cross-sections, of steps 7 and 8 round spermatids. Based on the time frame of recovery (28 days), the missing germ cells were stage VII and VIII A_al (undifferentiated) spermatogonia at the time of treatment, suggesting this cell population is specifically sensitive to RA-induced apoptosis. Premature differentiation resulting from RA treatment may be the driving factor in RA-induced A_al apoptosis, as previous reports have linked premature germ cell differentiation to increased germ cell apoptosis [23–25].

CYP26 enzymes degrade RA, thus regulating RA availability in both the embryonic and neonatal gonads [12, 13, 15]. The abnormal expression of CYP26 enzymes resulting from neonatal RA exposure may potentially be involved in permanent alteration of RA gradients along the seminiferous tubule. Previous reports have demonstrated both the Cyp26a1 and Cyp26b1 genes are induced in response to RA exposure in isolated gonocytes [14]. Thus, transient induction and subsequent suppression of Cyp26a1 expression after neonatal RA exposure suggest that exogenous RA may alter testicular RA distribution via CYP26 enzymes.

Several observations reported in this work (Fig. 10) potentially shed light on the mechanisms driving the onset and maintenance of the spermatogenic wave. RA-induced synchrony by neonatal exposure is similar to that observed in VAD-ROL-replenished individuals. One notable commonality between the VAD and neonatal testes is the lack of advanced germ cells or an established spermatogenic wave. This is not the case in a vitamin A-sufficient testis, where the presence of advanced germ cells and an established wave may be preventing RA-induced synchrony. While it is tempting to postulate that advanced germ cells in the adult maintain the spermatogenic wave established by localized RA availability in the neonate, further research is needed to definitely demonstrate a role for advanced germ cells in the maintenance of asynchronous spermatogenesis. Additionally, the role of the somatic environment must not be overlooked in developing a complete understanding of the mechanisms giving rise to asynchronous spermatogenesis.

FIG. 10.

Summary of exogenous RA effects and potential mechanism of RA-induced synchrony. A) Untreated 2-dpp testes have roughly periodic regions of RA response that represent the onset of RA-driven spermatogonial differentiation and are the precursors to the (B) adult spermatogenic wave. C) Exposure to RA in the neonate results in a more even distribution of RA response and onset of differentiation in many spermatogonia (presumably A_al). D) Spermatogonia prematurely entering differentiation undergo apoptosis (red, stage X). E) Adult testes from individuals exposed to RA as neonates have a normal spermatogenic cycle but lack a spermatogenic wave (synchronous spermatogenesis), similar to (F) adults who have undergone vitamin A deficiency (VAD) followed by (G) retinoid replenishment, resulting in synchronous initiation of germ cell differentiation. H) Similar RA treatment in vitamin A-sufficient adults results in apoptosis of A_al spermatogonia (I) leading to a loss of round spermatids (red box) 28 days after exposure but with no affect on the spermatogenic wave. Blue indicates RA response or availability. Roman numerals represent stages of the cycle of the seminiferous epithelium. Shapes represent maturing germ cells: horizontal ovals are spermatogonia; small circles are preleptotene and leptotene spermatocytes; large circles are pachytene spermatocytes; small, partially filled circles are round spermatids; crescents are elongated spermatids; and vertical ovals are elongating spermatids.

The observations reported herein demonstrate a novel method for the induction of synchronous spermatogenesis. Previous work has demonstrated the fundamental role of RA in driving germ cell differentiation and has implicated RA as a potential initiator of asynchronous spermatogenesis. Induction of synchronous spermatogenesis by RA exposure in the neonate further supports the notion that RA availability in the neonatal testis is fundamental to the initiation of asynchronous adult spermatogenesis and suggests a potential role for advanced germ cells in the maintenance of the spermatogenic wave in the adult.