The antiestrogen ICI 182,780 induces early effects on the adult male mouse reproductive tract and long-term decreased fertility without testicular atrophy

Background Estrogen receptors (ER) have important physiological roles in both the female and male reproductive systems. Previous studies using the estrogen receptor-α knockout mouse (αERKO) or antiestrogen treatment in adult rodents have shown that ERα is essential for normal function of the male reproductive tract. In the present study, time-response effects of the antiestrogen ICI 182,780 were determined to better understand ERα function in the adult male. Methods Adult male mice, 30 days old, were injected subcutaneously with ICI 182,780 (5 mg) once per week for 17 weeks. Tissues were fixed by vascular perfusion to study the time responses from day 2 to 125 post treatment. Results No difference was seen in body weight due to treatment. Testis weight was decreased 18% on day 59 and 21.4% on day 125. Other significant treatment-related effects included the following: 1) dilation of rete testis and efferent ductule lumen; 2) decreased height of the rete testis and efferent ductule epithelium; 3) decreased height of the supranuclear epithelial cytoplasm in efferent ductules; 4) decreased height of the efferent ductule epithelial microvilli, particularly in the proximal ductules; 5) decrease in the PAS-positive granules and endocytotic vesicles in nonciliated epithelial cells of efferent ductules; 6) capping and vesiculation of narrow cells in the initial segment of the epididymis; 7) accumulation of PAS-positive granules in apical cells of the caput epididymis; 8) increase in lysosomal granules in clear cells of the corpus and cauda epididymis; 9) limited induction of atrophic seminiferous tubules and abnormal spermatogenesis; and 10) decreases in the concentration of cauda sperm, progressive sperm motility and decreased fertility. Conclusions Antiestrogen treatment of the pubertal male mouse resulted in reproductive effects similar to those observed in the αERKO mouse as early as day 4; however, testis weight did not increase substantially and total atrophy was not observed with extended treatment.

The major histopathological effect of ERα knockout in the male mouse reproductive tract was centered in the efferent ductules, a small region between the testis and initial segment of the epididymis [4,5,11]. In the αERKO male, there was a failure of efferent ductule epithelium to reabsorb fluid, which caused fluid to accumulate in the rete testis and seminiferous tubule lumen. The long-term consequence of this pathophysiology was backpressure atrophy of the testis, which lead to azoospermia. The important role of estrogen in the male reproductive tract was further demonstrated by transplanting αERKO testicular germ cells into a wild-type recipient mouse testis depleted of germ cells [12,13]. The transplanted αERKO germ cells gave rise to offspring, thus demonstrating that αERKO sperm are capable of fertilizing eggs if the sperm travel through a wild-type reproductive tract environment.
The αERKO mouse continues to be an important animal for the study of estrogen receptor function in the male, but this transgenic animal is a developmental model, as the males lack ERα expression throughout perinatal and postnatal life. Therefore, to test the hypothesis that ERα is necessary in the adult male, and not just for development, the pure antiestrogen ICI 182,780 (AstraZeneca, Macclesfield, Cheshire, UK) was used to block ERs in the pubertal male. In the first study, mice were treated for 35 days and examined for effects. In general, it was concluded that ICI 182,780 treatment replicated most of the effects seen in the αERKO male, except for testicular atrophy and infertility [9].
The rat was also studied for comparison, and although there were similarities to the ICI 182,780 treated mouse, i.e., dilation of rete testis and efferent ductule lumen and decreased height of the ductal epithelium, the timeresponse approach used in the rat revealed considerable differences between the two species in response to the antiestrogen [14,15]. The rat showed long-term testicular atrophy induced by fluid accumulation and testicular swelling, which led to infertility, similar to the αERKO male. Testicular and efferent ductule effects in the rat were rapid, beginning within 2-4 days after treatment. The rat efferent ductule epithelium also displayed an unusual increase in lysosomal content in the apical cytoplasm of the nonciliated cells, a response not noted after 35 days of treatment in the mouse [9]. Therefore, several questions remained regarding antiestrogen responses in the mouse model: a) Does the mouse require a longer treatment period to show testicular atrophy? b) Is the rat more sensitive to the antiestrogen? c) Would a longer treatment period in the mouse lead to infertility? To answer these and other questions, a time-response study of ICI 182,780 treatment in pubertal male mice was conducted over a 17week period.

Animals and experimental design
This study was conducted in three experiments to fully cover the response period to antiestrogen treatment. The first experiment examined days (d) 8-59 post treatment. Because strong effects were found on d 8, a search for the first response was conducted in an experiment that examined days 1, 2 and 4 post treatment. Infertility was not observed on d 59, therefore, a third experiment was performed to test for effects on days 100 and 125 post treatment.
Adult male C57BL/10 mice at 30 days of age were maintained under constant lighting (12 h light:12 h dark) and temperature (22°C). The University of Illinois Institutional Animal Care and Use Committee approved all experimental protocols. Food and water were available ad libitum for the experimental period. Animals were randomly divided into control (N = 3-6) and treated (N = 3-6) and injected subcutaneously with 0.1 ml of castor oil (controls) or 5 mg of ICI 182,780 (AstraZeneca, Macclesfield, UK) in 0.1 ml, each week for up to 125 days.

Tissue processing
Control and ICI-treated mice were euthanized at specified time points out to d 125 post treatment. The animals were deeply anesthetized with xylazine/ketamine (87 mg/kg and 15 mg/kg, respectively), weighed, and then fixed by vascular perfusion with 4% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.4) for Histopathology. Bouin's fixative or neutral buffered formalin was perfused for immunohistochemistry. Testes, efferent ductules and the epididymides were removed and stored in the same perfusion solution at 4°C until embedded. For microscopic evaluation, tissues were weighed and then washed in cacodylate buffer, dehydrated in an ascending series of ethanol and embedded in glycol methacrylate . Tissue blocks were cut at 2.5 µm thickness, dried at 35°C, stained with periodic acid-Schiff reaction (PAS), counterstained with Mayer's hematoxylin and mounted for observation. Cauda epididymal sections were also stained with Coomassie blue G250, which highlights the sperm acrosome blue and the nucleus pink.

Sperm Concentration and Motility
To determine sperm concentration in the cauda epididymis for the first 59 days of the study, paraffin histological sections of the distal cauda epididymis were sectioned at 20 µm thickness and stained with hematoxylin and eosin. The slides were viewed using a video camera mounted on an Olympus microscope with a planapochromatic 100X objective. Observed fields were selected by a systematic, uniform random sampling method, using the Stereo Investigator software, v3.0 (MicroBrightField Inc., Colchester, VT) and a motorized stage. Sperm heads were counted in a counting frame hav-ing a total of 20,000 µm 3 per tubule. Ten replications per epididymis were made.
On day 125 post treatment, sperm concentrations were determined using minced cauda epididymides and a dualsided semen analyzer system (Hamilton Thorne Research, Beverly MA). The cauda epididymis on one side of the 125 day mouse was dissected prior to perfusion and was placed in a sterile Petri dish containing 1 ml of modified Hank's Balanced Salt Solution (m-HBSS) for sperm motility analysis. The cauda was minced and incubated for 5 minutes at 37°C to allow sperm to swim into the media. At the end of the incubation time, a 100 µl sample of the media containing the sperm was diluted with 900 µl of the m-HBSS solution for sperm analysis. Each sperm analysis chamber was loaded with 14 µl of the diluted sperm by capillary action. Analysis chambers were immediately placed into the specimen holder of TOX-IVOS computer assisted sperm analysis system that had been calibrated for mouse sperm samples. The samples, chambers and specimen holder were maintained at 37°C throughout the entire measurement period. Six fields were analyzed from each of the two chambers and the results averaged.

Image analysis
Tissues were photographed using a Spot-II digital camera mounted on an Olympus Microscope and Plan-apochromatic lenses (Olympus America Inc, NY) and saved in Adobe Photoshop (Adobe Systems, Inc., San Jose, CA). Morphometry data from days 8, 16, 59 and 125 post treatment are included. The digital images from both reproductive tracts of 3 control and 3 ICI-treated animals on each post treatment day were analyzed by Scion Image Software (Scion Corporation, Frederick, MD). Luminal diameter of 10 seminiferous tubules (stages VII-VIII) was determined per mouse per testis, by measuring the widest diameter in cross-section. However, for day 125 all sem-  iniferous tubules per testis were measured for diameter because some tubules were atrophic and stage of spermatogenesis was not determined. Luminal diameter of the rete testis was determined in 3 testes per treatment group. Luminal diameter of proximal efferent ductules was determined in 5 sections per mouse. Cell height was measured from the base to the apex in 30 epithelial cells of the rete testis region, or to the microvillus tip in 30 nonciliated cells from non-oblique sections of the proximal efferent ductules per mouse. The supranuclear cytoplasmic height was determined by measurement from the top of the nucleus to the microvillus tip in the nonciliated cells in the proximal efferent ductules.

Fertility
Individual mice of control and ICI-treated groups on day 50, day 93, or day118 after treatment were mated with 3 female mice of the same age and strain for one week, then separated. One week after separation, the females were euthanized by sodium pentobarbital injection and the uterine horns opened. The number of embryos, number of implantation sites and resorptions/stillbirths were counted. If at least one female per mating had live births, the male was considered fertile. Subfertility was indicated by a reduction in the number of live births or an increase in the number of resorptions.

Statistical analysis
Data are presented as means ± SEM. To compare morphological differences between controls and treatment groups, means were analyzed by two-way ANOVA followed by individual paired t-test for each time point (P < 0.05). Proportion data were analyzed by the chi-square test with a probability of less than 5% considered significant.

Body and organ weights
There was no significant difference in body weight between control and the treated groups (Table 1). Testis weight was increased on day 16 but decreased on days 59 and 125. The increase in weight on day 16 is likely due to a discrepancy in animal ages at the time of treatment. When this time point was repeated using mice at 30 day of age, the difference disappeared (data not shown).

Testis and rete testis
Spermatogenesis appeared normal until day 59, as which time focal regions showed seminiferous epithelial disruption and the beginning of isolated tubular atrophy. However, normal spermatogenesis was seen in most seminiferous tubules throughout the 125-day treatment Seminiferous tubular atrophy Figure 2 Seminiferous tubular atrophy. Even control testes show limited atrophic tubules with aging of the males, but beginning on day 59 the number of atrophic tubules was significantly (*) increased in the ICI-treated testes. Means ± SEM; (p < 0.05).
Rete testis luminal diameter Figure 3 Rete testis luminal diameter. The antiestrogen treatment model induced a highly significant increase in the diameter by day 59 (p < 0.05).
Rete testis histology from control and ICI-treated groups on day 59 post treatment period (Fig. 1). Seminiferous tubular diameter showed no difference between treatment groups until day 125, when mean tubular diameter in the treated group was 164.8 ± 0.8 µm compared to 187.1 ± 4.0 µm for the controls, a significant difference of 11.9% (p < 0.01). This change was due to a significant increase in the number of atrophic seminiferous tubules from day 59 to 125 (Fig. 2). The luminal diameter of rete testis was increased 241% on day 59 in the treated group (Figs. 3 and 4A,4B,4C,4D) and remained dilated throughout the 125-day period. Epithelial cells lining the rete testis were variable in height, but showed significant decreases (22% overall decrease across all time periods) beginning on day 8 and continuing to day 125 (Figs. 4B,4D).

Efferent ductule
As early as day 4 post treatment, ICI 182,780 treatment caused luminal dilation in the proximal efferent ductules (Fig. 5). However, luminal dilation on day 4 was noted without a decrease in epithelial height. Tubule dilation continued throughout the treatment period, reaching a maximum increased dilation of 140% on day 125 (Fig. 6). The conus and common efferent ductules also showed significant dilation at all time points (data not shown).
Height of epithelial cells in the efferent ductules was also decreased after treatment on days 8-125 (Fig. 7); however, on day 125 epithelial height was highly variable in the treated males and the mean was approaching that of the controls, although areas were nearly squamous in appearance (Fig. 8).
Proximal region of efferent ductules from control (A, C) and ICI-treated (B, D) groups on day 4 post treatment The supranuclear cytoplasm, the region that contains an abundant endocytotic apparatus in nonciliated cells of the efferent ductule, was consistently reduced by nearly 50% in the treated group, compared to controls (Fig. 9). On day 125, the supranuclear region was increased in both treated and controls, but there still remained a significant reduction in height in some proximal ductule areas (Fig.  8).
PAS positive lysosomal granules were numerous in the supranuclear cytoplasm of control nonciliated cell in the proximal region (Fig. 8A,8C,8E). The conus and common regions in control animals contained fewer lysosomal granules (data not shown). ICI 182,780 treatment caused a reduction in the number of lysosomes and vesicles in all regions, but in some regions the effect was more dramatic (Fig. 8B,8D,8F). The reduction in lysosomes was associated with a decrease in supranuclear cytoplasm (Fig. 9). With decreased height of the epithelium, the nuclear shape was changed from nearly round or slightly columnar (controls) to a more flattened morphology (Fig. 8).
ICI 182,780 treatment also caused microvilli along the luminal border of the nonciliated cell to shorten significantly (Fig. 10). Mean height of microvilli in the treated tissue was decreased by 18% in the proximal efferent ductule epithelium. Microvilli in the conus and common regions were decreased in height by 16% and 6%, respectively (data not shown).

Efferent ductule immunohistochemistry
On day 125 post treatment, the expression of ERα was decreased, while ERβ showed no change in expression in the efferent ductule epithelium (Fig. 11A,11B,11C,11D). NHE3 was completely absent along the apical border of the treated nonciliated cells (Fig. 11E,11F). AQP1 were decreased significantly along the apical border in nonciliated cells of the ICI 182,780 treated efferent ductules, but along the basolateral border all staining was eliminated (Fig. 11G,11H). CAII was also decreased significantly along the apical cytoplasm of the efferent ductule epithelium (Fig. 11I,11J).
Proximal efferent ductule luminal diameter Figure 6 Proximal efferent ductule luminal diameter. The ductule lumen was dilated significantly throughout the treatment period. Means ± SEM. Significant differences (*) between control and treated at each time point are noted (p < 0.05).
Proximal efferent ductule epithelial cell height Figure 7 Proximal efferent ductule epithelial cell height. Cell height was reduced significantly at all time points. However, on day 125 post treatment the mean difference was less than previous periods and the height showed greater variation. Means ± SEM (p < 0.05).
Higher magnification of the proximal efferent ductule epithelium in control (A,C,E) and ICI-treated (B,D,F) males

Epididymis
In the initial segment, narrow cells show slight protrusion of their apical vesiculated cytoplasm into the lumen, while in the ICI 182,780 treated tissues the narrow cell cytoplasm contained more vesicles and protruded further into the lumen (Fig. 12). By day 125 post treatment, apical cells in the caput epididymis were engorged with PASpositive lysosomal-like granules (Fig. 13), with some granules nearly as large as the nucleus. In the corpus and cauda regions post treatment, the clear cells showed an increase in the number and size of vesicles in both principal and the clear cells (Fig. 14). This observation was present as early as day 8 post treatment. The vesicles were mainly supranuclear in location. Clear cells also contained an increase in the number of PAS positive granules in their basal cytoplasm after treatment (Fig. 14). After antiestrogen treatment, principal cells of the distal cauda also showed an increase in the number and size of cyto-plasmic vesicles, as early as day 8, as well as an increase in the number of PAS-positive granules (Fig. 14).
Histological sections of cauda epididymis on day 59 post treatment were also stained with Coomassie blue G250 to highlight the sperm acrosome and nucleus. Several abnormalities were noted in the ICI-treated sections (Fig. 15). In control cauda epididymis, the sperm were aligned nearly parallel and the cytoplasmic droplet was located on the mid-tail region. At higher magnification, the sperm heads were attached to the tail and displayed a prominent staining of the acrosomic region along the dorsal edge of the nucleus (Fig. 15A,15B). In the ICI-treated samples, sperm were less concentrated and appeared disorganized, with few nuclei attached to tails (Fig. 15C). At higher magnification, detached sperm nuclei were misshaped and often appeared to have reduced staining of the acrosome (Fig 15D,15E) and the tails were looped around the cytoplasmic droplets (Fig. 15F).
Proximal efferent ductule nonciliated cell supranuclear cyto-plasmic height Figure 9 Proximal efferent ductule nonciliated cell supranuclear cytoplasmic height. The data represent means ± SEM. The increase seen on day 125 may be due to this day being a separate experiment, although there may be some increase due to aging. Significant differences (*) between control and ICItreated means are noted (p < 0.05).
Proximal efferent ductule nonciliated cell microvillus height Figure 10 Proximal efferent ductule nonciliated cell microvillus height. The data represent means ± SEM. At all time points there was a significant decrease in height (*) between control and ICI-treated cells (p < 0.05).
Immunohistochemical staining of proximal efferent ductules on day 125 post treatment

Reproductive performance
There was no difference in fertility, the number of embryos or the number of resorptions and stillbirths per female between the control and treated groups on day 59 and 100 (data not shown). However, cauda sperm concentrations in the treated groups did not show the normal increase associated with increasing age and beginning on day 16 through 125 post treatment there was a significant difference from controls (Fig. 16). On day 125 post treatment, cauda sperm concentration in the ICI-treated males was decreased 40%, sperm motility was decreased 80% and 4 of 6 treated males were infertile ( Table 2). The treated males that sired offspring had a lower mean number of births (litter size), but due to large variation in both groups, the means were not significantly different. However, there a significant increase in the incidence of resorptions and stillbirths (Table 2).

Discussion
The present study demonstrates that the mouse ICI 182,780 treatment model, when extended from 4 days to 125 days post treatment, resembles the αERKO model in nearly every feature, except for the transient increase in testis weight and testicular atrophy. The decrease in sperm concentration in cauda epididymis and luminal dilation of the rete testis and efferent ductules following treatment clearly support the overall hypothesis that a functional ERα is required for normal fluid reabsorption in the efferent ductules, as concluded from prior studies of the αERKO male [4,7].
In the present model, treatment was started at 30 days of age. This age represents early puberty and the males were still growing, as evidenced by the increase in body and testis weights in controls and treated animals over the next 60 days. Except for an unexplained difference in testis weight on day 16 post treatment, there was no effect on testis weight out to day 25, similar to a previous shortterm study [9]. However, on day 59 of treatment testis weight differences began to appear, with a decrease in the treated animals. Thus, with a longer treatment period the mouse showed limited atrophy of seminiferous tubules, but not total atrophy as seen in the αERKO males [4].
Initial segment of the epididymis on days 10 and 59 post treatment At the beginning of puberty, spermatogenesis appears normal in the αERKO testis, but as the male ages seminiferous tubules begin to degenerate [1]. The αERKO testis weight nearly doubles in approximately 40 days, starting at puberty; then testis weight declines and eventually total testicular atrophy occurs at 185 days of age [4]. ICI 182,780 treatment in the rat induces a similar pattern of effect, with both testicular swelling and long-term atrophy of the testis, nearly identical to the αERKO male [14,15]. Why the mouse testis did not respond with a similar increase in testis weight and then atrophy remains a major question. However, the rete testis was dilated in the mouse, as in αERKO and ICI-treated rat males, suggesting that ability of the mouse testis to handle fluid accumulation is dependent upon developmental factors that are lacking in the αERKO male. One possibility is that the mouse epididymis is capable of compensatory reabsorption of the excess fluid. Future studies are required to understand these differences in response.
The effects observed on spermatogenesis did not appear to be due to fluid accumulation because there was no dilation of the seminiferous tubular lumen, in contrast to what was discovered in the αERKO testis [4]. Furthermore, abnormal spermatogenesis did not occur until day 59, suggesting that the degenerative process required long-term inhibition of ER. Because ICI 182,780 blocks both ERα and ERβ [20], the delayed testicular response may be due to ICI's inhibition of ERβ and not ERα. ERα is not expressed in the seminiferous epithelium, but ERβ is abundant in Sertoli and germ cells [16,[21][22][23][24][25]. Therefore, ICI's effects on the epithelium may be through long-term inhibition of ERβ. It is interesting that the aromatase knockout mouse also shows a delayed disruption of spermatogenesis [26,27], similar to the effects of ICI 182,780 in the mouse testis. Although the ERβ knockout mouse shows no apparent phenotype [3,10,28], in light of the current findings, it would be worthwhile to examine the βERKO male for specific aging effects on the testis.
Dilation of the rete testis lumen was less extensive following ICI 182,780 treatment, compared to the αERKO mouse [9]. The present study demonstrated that this effect begins sooner than previously thought, with dilation being observed on day 10. On the other hand, the rete epithelium was reduced in height as early as day 8. It is assumed that the decrease in epithelial height is related to the accumulation of luminal fluid; however, it may also be due to overgrowth of the rete epithelium because neonatal exposure to environmental estrogens has produced a similar dilation and overgrowth of the rete testis [29,30].
A prior study did not find these early effects of ICI 182,780 treatment [9]. However, here we report effects as early as day 4, at which time proximal regions of the efferent ductules were dilated. This observation is important because the dilatation did not induce a reduction in epithelial cell height, suggesting that the two effects are not directly linked. Others have suggested that the decrease in epithelial height is due to alterations in ion transporter expression [8], but the present study does not support this hypothesis, as fluid reabsorption detected by luminal dilation occurred without a decrease in epithelial height. Other early effects were subsequent reduction in efferent ductule epithelial height (day 8) and protrusion of the narrow cells in the epididymis. Effects on the testis were not observed until much later. These early morphological effects were maintained throughout the 125-day treatment period and are in agreement with our previous studies of αERKO mouse [4,5] and ICI 182,780 treatment in mice [9]. Thus, it may be possible to use the mouse anti-estrogen treatment model during shorter periods of treatment for the study of fluid dynamics within the efferent ductules, but also to study testicular effects that require more than 50 days of exposure.
Overall effects of ICI 182,780 on the mouse efferent ductules were identical to those observed in the αERKO male, including the loss of NHE3 staining and decrease in staining for AQP-1 and CAII [7]. These data further support the hypothesis that estrogen's regulation of sodium transport by this epithelium is essential for maintenance Cauda epididymal region on day 59 stained with Coomassie blue G250 to highlight the sperm acrosome and nucleus Figure 15 Cauda epididymal region on day 59 stained with Coomassie blue G250 to highlight the sperm acrosome and nucleus. A. Control section showing normal spermatozoa (arrow) with intact sperm tails and normal heads. Cytoplasmic droplets (*) are found along the mid-tail region. Sperm are aligned nearly parallel within the lumen. B. Higher magnification of a normal sperm from the control section showing an intact acrosome (arrows) along the dorsal edge of the sperm nucleus (pink stain). C. ICItreated section showing more dilute sperm, isolated sperm heads (circles), disorganized appearance and sperm tails looped around the cytoplasmic droplets (*). D, E. higher magnification of ICI-treated sperm heads showing abnormal shapes (arrow) and decreased staining of the acrosome region. F. Higher magnification of the tail looped around cytoplasmic droplets (arrows). Bars = 10 µm for A, C. Bars = 2.5 µm for B, D, E, F.
of fluid reabsorption. Antiestrogen treatment also caused a decrease in ERα staining, which has been reported in female tissues [31], but the compound had no effect on ERβ.
The effects noted in epididymis could be due to direct action of ICI on the epididymal epithelium, as recent studies have shown that ERα is expressed more extensively in the epididymis of the mouse [21], than in the rat [11]. Cell types that responded to treatment, the narrow and apical cells of initial segment, apical cells of the caput, and clear cells of the corpus and cauda, are all strongly ERα-positive [21]. In the ICI treated group, apical vesicles in the narrow cells and vesicles in clear and principal cells in corpus and distal caudal epididymis were increased in size and number. An accumulation of large PAS-positive granules was also observed in the apical cells of the caput. These findings are nearly identical to epididymal changes seen in αERKO [5] and may indicate an attempt to increase reabsorption of fluid by the epididymis, to compensate for reduced reabsorption found in the proximal region of the efferent ductules [4,7].
In the control animals, epididymal sperm concentrations reached their peak on day 59, showing a gradual increase starting at 38 days of age. However, the ICI treated group did not experience this normal increase in sperm concentration and beginning on day 16 showed a divergence from controls so that on days 59 and 125 post treatment there were decreases of 30 and 41%, respectively. These differences are considerably more than the 22% decline seen in αERKO males [9]. On day 25 (data not shown), there was a 12.4% decline in sperm concentration, which was similar to our previous experiment, which showed an 11% decrease after 35 days of antiestrogen treatment [9]. Thus, the overall trend in the mouse antiestrogen treatment model is for increasing dilution of cauda sperm and accumulation of fluids in the efferent ductules and testis. It is noteworthy that the decrease in fertility did not occur until day 125 post treatment, the time period that coincided with the maximum decrease in sperm concentration. In the rat, decreased fertility coincided with the onset of testicular atrophy [14,15], therefore it was questioned whether infertility and subfertility was due to abnormal sperm maturation due to dilution in the epididymis or if atrophy of the seminiferous epithelium was required before sperm concentrations could reach levels that would produce infertility. The current study suggests that subfertility in the mouse antiestrogen treatment model is related to the sperm dilution effects that occur in the epididymis, which appears to induce abnormal maturation of the cauda sperm. Even in the αERKO male, testicular sperm are fertile when transplanted into germ cell depleted normal testes [12,13], which suggests that the epididymal environment is one of the more important factors leading to infertility in the absence of ERα expression in the male.

Conclusions
The antiestrogen, ICI 182,780 induces very early effects on rete testis, efferent ductules and epididymis, which resemble morphological changes observed in the αERKO male. These early effects are maintained throughout the treatment period of 125 days and thus, suggest that a much shorter model of 4-8 days of treatment may be used to study the role of estrogen in fluid reabsorption in the male tract. Although the inhibition of estrogen receptors for 125 days induced structural and physiological changes in the reproductive tract, it did not induce total testicular atrophy, but did cause a decrease in fertility and showed evidence of early induction of abnormal spermatogenesis. Thus, it is possible to induce infertility by blocking ER function in the male, without causing azoospermia. If such treatment effects are observed at lower dosages and if the responses are reversible, it should be possible in the future to target ERα for male contraception.

Authors' contributions
HWC carried out experiment one, analyzed the data and wrote the first draft of the manuscript. RN assisted with the immunohistochemistry. KC provided general labora-Sperm concentration from cauda epididymis in control and ICI-treated mice Figure 16 Sperm concentration from cauda epididymis in control and ICI-treated mice. The left Y-axis applies to days 8-59, which were analyzed by the optical dissector method. The right Yaxis applies to day 125, which were counted using the Hamilton Thorne semen analyzer system (Beverly MA). The data are presented as means ± SEM. Significant differences (*) between control and ICI-treated means are noted (p < 0.05).
tory assistance, carried out experiments 2-3 and assisted with photography. NAQS provided valuable laboratory assistance. RAH served as PI on this project, designed the study and wrote the final draft. All authors read and approved the final manuscript.