Sertoli cell proliferation in neonatal boars is potentially androgen dependent. Hence, the immediate objective was to evaluate effects of androgen receptor-mediated signaling on the first wave of Sertoli cell proliferation. The experimental design employed littermate pairs of boars with one member assigned to receive a daily oral dose of flutamide, an androgen receptor antagonist, beginning at 1 wk of age and the littermate the canola oil vehicle. Experiment 1 examined the response at 6.5 wk of age after completion of the first wave of Sertoli cell proliferation, and experiment 2 examined the response at 11 wk of age after initiation of the second wave of Sertoli cell proliferation. Experiment 3 was designed to evaluate initial responses at 2, 3, or 4 wk of age. Additional littermates from four of the litters evaluated at 2 wk of age were hemicastrated at 8 days of age. Testis weight increased approximately 50% in the flutamide-treated boars compared with vehicle-treated littermates (P = 0.01) by 6.5 wk of age. Approximately 80% more Sertoli cells/testis were present in flutamide-treated boars at 6.5 wk of age compared with their vehicle-treated littermates (P < 0.01). Animals that were hemicastrated at 8 days of age had more Sertoli cells/testis than their intact littermates at 2 wk of age (P < 0.01), but flutamide inhibited the hemicastration response. Androgen receptor antagonism during postnatal Sertoli cell proliferation increases Sertoli cell numbers, as does hemicastration, but receptor antagonism initially inhibits Sertoli cell proliferation induced by hemicastration.
Sertoli cells are the sole somatic cell located within seminiferous tubules; they provide critical support for the differentiation of spermatozoa through paracrine secretions and the creation of a cellular barrier from the immune system [1-4]. Sertoli cells have a fixed capacity for the number of developing sperm they can support. Hence, Sertoli cell number is a major determinant of sperm production capacity and associated testicular weight [5, 6]. In most species studied, Sertoli cells proliferate only until puberty under normal conditions; therefore, the final number of Sertoli cells and sperm production are determined by the time of puberty [4, 7]. Mammals generally have two postnatal waves of Sertoli cell proliferation prior to puberty , although only one wave can be detected during the short prepubertal period typical of rodents. The two waves of proliferation are distinct in the boar and extend over a few months rather than years as in some primates, facilitating investigation into the regulation of Sertoli cell proliferation during the individual waves . The first wave of Sertoli cell proliferation in the boar occurs from about 1 to 5 wk of age and the second from about 10 to 16 wk of age with sperm present in the epididymis at 16 wk of age in our studies [9-12]. Regulation of Sertoli cell proliferation is an essential component of testicular development and fertility.
Follicle-stimulating hormone (FSH) is generally believed to regulate Sertoli cell proliferation [13-17]. The status of FSH as the predominant and pervasive hormone regulator has been challenged recently in some rodent and porcine studies. For instance, FSH receptor knockout mice have similar numbers of Sertoli cells up to Day 20 of postnatal life as parental strains in one study . Studies using boars indicate that increased Sertoli cell proliferation is not highly correlated with FSH concentrations [19, 20], although Piau pigs did have an increase in FSH that correlated with postnatal Sertoli cell proliferation . Treatment of boars with FSH had no effect on testis weight or Sertoli cell numbers [21, 22]. FSH did partially rescue Sertoli cell numbers in boars treated with a gonadotropin-releasing hormone agonist that inhibited proliferation . Neonatal FSH concentrations in boars with large and small testes are not different . Collectively, available data support the idea that FSH is important for maintaining an established Sertoli cell population, with a more minor role in stimulation of porcine Sertoli cell proliferation.
In contrast to FSH, androgens were thought to be a less important regulator of the size of the Sertoli cell population between birth and puberty, although androgen signaling is essential for the development and maintenance of the male reproductive system . Interpretation of the androgenic role is frequently made more complex by hormone interactions. For example, administration of testosterone to postnatal rats from 1 to 5 days of age caused a decrease in Sertoli cell numbers, accompanied by a decrease in FSH and a decrease in estradiol .
Flutamide, a potent nonsteroidal androgen receptor antagonist, can be used to interfere with androgen receptor activation during specific developmental windows. Several studies have used flutamide to study androgen receptor inactivation prenatally. In pigs, prenatal exposure to flutamide disrupts the tight junctions associated with Sertoli cells and disrupts Leydig cell proliferation [27-29]. To our knowledge, no studies have analyzed the effects of androgen receptor inactivation during this first postnatal window of proliferation.
Materials and Methods
Animal Treatment and Experimental Design
The boars used were derived from breeding stock (PIC Lines 65 and C24) provided by PIC USA (Franklin, KY), and semen was provided by Genus PLC (Franklin, KY) with the exception of one litter (Yorkshire × Hampshire crossbred) evaluated at 2 wk of age in experiment 3. Animal experiments were approved by the Institutional Animal Use and Care Advisory Committee at the University of California, Davis, in accordance with the Guide for the Care and Use of Agricultural Animals in Agricultural Research and Teaching and consistent with SSR guidelines. Within litters, boars were assigned to treatment, either a daily oral dose of flutamide (10 mg/kg body weight; Sigma, St. Louis, MO) suspended in canola oil or the canola oil vehicle alone, and the age for tissue collection when treatment effects would be assessed. The dose of flutamide was selected based on previous flutamide studies in dogs and pigs [30-32] and checked in a dose response (0, 1, 5, and 10 mg/kg body weight daily for 1 wk). Testosterone was not elevated at 1 mg flutamide/kg body weight but was increased approximately 40% at 2 wk of age by the 5-mg/kg body weight dose and not further increased by the 10-mg/kg body weight dose. In experiment 1, littermate pairs of boars from five separate litters were treated from 1 to 6.5 wk of age, and testicular tissue was collected at 6.5 wk of age. In experiment 2, littermate pairs of boars from five separate litters were treated as in experiment 1 from 1 to 6.5 wk of age, but testicular tissue was collected at 11 wk of age in order to evaluate whether the treatment effect persisted to the second wave of Sertoli cell proliferation. In experiment 3, littermates from seven separate litters were treated with flutamide or canola oil vehicle daily starting at 1 wk of age, and testicular tissue was collected at castration at 2, 3, or 4 wk of age. Four of these seven litters provided additional littermates that were hemicastrated at 8 days of age to compare the combined effects of androgen receptor inactivation and hemicastration with androgen receptor inactivation only and hemicastration alone on Sertoli cell proliferation at 2 wk of age. A single testis was removed from the intact boars at 2 wk of age, and the remaining testis was removed at 3 wk of age, allowing for a second comparison of androgen receptor inactivation and hemicastration on Sertoli cell proliferation and maturation and their interaction at 3 wk when the remaining testis was removed (four litters). Blood was collected weekly from the jugular vein during treatment. Blood was cooled and centrifuged for a minimum of 10 min at 1300 × g, and the plasma was aliquotted and stored (−20°C). At collection, testes were weighed, and a slice of testis was removed from the equatorial region and fixed overnight in 4% paraformaldehyde in PBS at 4°C. The fixed tissues were then washed in PBS and dehydrated in 30%, 50%, and 70% ethanol and embedded in paraffin. Testicular tissue devoid of the tunica albuginea, and the mediastinum was also frozen on dry ice at collection and stored (−80°C) for subsequent tissue testosterone and RT quantitative PCR (qPCR) analysis.
Sertoli cells were enumerated following immunohistochemical labeling of Sertoli cells by GATA4. This immunolabeling was performed as previously described [19, 33] using the Vectastain Elite ABC kit (Vector Laboratories, Burlingame, CA). Testes were cut into 25-μm-thick paraffin sections and deparaffinized with Citrosolv Hybrid (Fisher Scientific, Pittsburgh, PA). Sections were rehydrated with sequential 2-min immersions in 100%, 95%, and 70% ethanol followed by 5 min in running water. Antigen retrieval was performed by heating sections to 93°C in diluted Antigen Unmasking Solution (Vector Laboratories; catalog no. H3300). After being cooled to room temperature, sections were rinsed in Tris-buffered saline (TBS; 50 mM Tris, 1.5% NaCl, pH 7.6) and incubated for 10 min in 3% hydrogen peroxide in methanol to block endogenous peroxide activity. Sections were blocked in 1% normal serum (Vector Laboratories) in TBS for 30 min at room temperature followed by labeling with a polyclonal goat anti-mouse GATA4 (sc-1237; Santa Cruz Biotechnologies, Santa Cruz, CA) 1:400 in TBS for 2 h at room temperature, washed three times in TBS, and incubated with a biotinylated rabbit anti-goat secondary antibody. Labeled sections were washed three times in TBS and then incubated for 40 min with avidin-biotin-peroxide complex (ABC; Vector Laboratories). Immunoreactivity was visualized with NovaRed (Vector Laboratories). Sertoli cell numbers were determined in 17-μm-thick counting frames randomly selected by the CAST Grid software as previously described . Number of Sertoli cells per testis is a product of the average density per frame and testis weight divided by volume of counting frame. At least 30 counting frames were evaluated for the older testes, and a minimum of 60 frames were evaluated for the younger testes.
Frozen testis tissue was homogenized in 0.1 M potassium phosphate buffer (pH 7.4) containing 20% glycerol, 5 mM β-mercaptoethanol, and 0.5 mM AEBSF as previously described [33-35]. The homogenate was sonicated for 3 sec, centrifuged at 15 000 × g for 10 min, and stored at −20°C until analysis. Protein concentration was determined by the Bradford assay (Bio-Rad Laboratories, Hercules, CA).
Testosterone (plasma and testicular tissue), estradiol, luteinizing hormone (LH), and FSH were measured using radioimmunoassays. A sheep anti-testosterone antibody (Niswender #S250), 3H-testosterone (NET370; Perkin Elmer Life Sciences, Boston, MA), and testosterone standards (A6940; Steraloids, Wilton, NH) were used to assess testosterone. A sheep anti-estradiol 17β-6-BSA antibody (Niswender #224; G. Niswender, Colorado State University, Fort Collins, CO), tritiated estradiol (NET-317; Perkin Elmer Life Sciences), and estradiol standards (E950; Steraloids) were used in the estradiol assays. A mouse monoclonal anti-bovine LH (518B7; courtesy of J.F. Roser) and porcine LH (EX275A; courtesy of H. Papkoff, University of California, Davis, CA) standards and a rabbit anti-porcine FSH (R285; H. Papkoff) and iodinated porcine FSH (EX274B; H. Papkoff) standards were used in the LH and FSH assays, respectively. Sensitivities for the estradiol, testosterone, LH, and FSH radioimmunoassays were 10 pg/ml, 0.1 ng/ml, 0.25 ng/ml, and 0.5 ng/ml, respectively, and extraction efficiencies in estradiol, plasma testosterone, and tissue testosterone assays averaged 82%, 76%, and 86%, respectively. The mean intra-assay coefficients of variation (CVs) for these radioimmunoassays were 12.7% for plasma estradiol, 9.2% for plasma testosterone, and 5.9% for tissue testosterone, and the interassay CVs were 14.5% for estradiol and 9.2% for plasma testosterone. The mean intra-assay CVs were 8.7% and 11.3%, and the interassay CVs were 7.8% and 12% for LH and FSH radioimmunoassays, respectively.
Effectiveness of the flutamide dose was further verified by evaluation of expression of the androgen-responsive gene HSPA5 (heat shock 70-kDA protein 5, previously known as GRP78, glucose-regulated protein, 78 kDa). This gene is androgen responsive in liver, pancreas, prostate, and Sertoli cells [36-38]. RNA was isolated from testicular tissue (four littermate pairs), previously stored in RNAlater (Life Technologies, Grand Island, NY) at −20°C using a mortar and pestle to grind the tissue in liquid nitrogen and using QIAzol lysis reagent (Qiagen Inc., Valencia, CA). The concentration and quality of the isolated nucleic acids were assessed with a NanoDrop 1000 UV-VIS spectrophotometer (NanoDrop Products, Wilmington, DE). One microgram of RNA was treated with deoxyribonuclease (Promega, Madison, WI) prior to cDNA synthesis. Reverse transcription was performed according to the manufacturer's directions using the RevertAid first-strand cDNA synthesis kit (Fisher Scientific). Primers were selected with Primer Blast and products on unintended templates checked against Sus scrofa . Secondary structure of potential primer pairs was checked with NetPrimer software (PREMIER Biosoft International, Palo Alto, CA). The selected forward primer for HSPA5 was TTGAGTGGCTGGAAAGTCACC, and the reverse primer was CATCACCAGTTGGGGGAGG (1943–2074 X92446.1) with a product size of 132 bp. Product size for selected primers was then checked by PCR (Eppendorf Mastercycler Gradient thermal cycler; Eppendorf, Hauppauge, NY) followed by separation on 2% agarose and visualization of the product (UVP ChemiDoc-ItTS2 Imager; UVP Inc., Upland, CA). Absence of a detectable product when water replaced the cDNA template was verified. Sperm-associated antigen 7 (SPAG7) was used as the housekeeping gene  after checking RNAseq analysis of 2-, 3-, and 5-wk-old testes to verify that gene expression was not altered during early development (relative abundance was 1.80, 1.73, and 1.79 for transformed means at 2, 3, and 5 wk of age for four vehicle-treated animals at each age, P > 0.25; T. Berger, unpublished data). The selected forward primer for SPAG7 was GAGCTGGATTCCTACCGTCG, and the reverse primer was CCTTCAGCCTCCGTTTCTCC (NM_001243921.1, nucleotides 357–426, which are primarily in exon 5 with one nucleotide in exon 6) with a product size of 70 bp. The RT-qPCR reaction was performed using an Applied Biosystems 7500 fast Real Time PCR system with a Fast SYBR Green PCR master mix (catalog no. 4385612; Applied Biosystems, Foster City, CA), 200-nM primers, and the cDNA corresponding to product synthesized from 0.005 μg RNA. The final reaction volume was 20 μl and was incubated at 95°C for 10 min followed by 40 cycles of 95°C for 15 sec and 60°C for 60 sec. Dilution curves prepared from aggregated cDNA indicated calculated efficiencies of 98.8% for the HSPA5 and 95.3% for SPAG7 with appropriate melt curves. The Pfaffl method was used to normalize data to the housekeeping gene .
Effects of treatments on Sertoli cell numbers were analyzed with SAS using the Proc Mixed procedure (SAS Institute, Cary, NC) with treatment as a fixed factor and litter as a random factor. Separate analyses for Sertoli cell number, testis weight, and body weight were conducted at each age to meet normality and homogeneity of variance constraints for ANOVA. A P-value less than or equal to 0.05 was regarded as significant. Hormone data were analyzed with time as a repeated measure, and testosterone, estradiol, FSH, and LH values were log transformed to meet normality constraints. The means and SEM for the hormone analyses are reported from the nontransformed data, and P-values are reported from the log-transformed data when treatment differences were detected by linear contrasts of transformed and nontransformed data at individual time points. Log-transformed Pfaffl scores and ΔCt scores were similarly analyzed with treatment as the fixed factor.
Experiment 1: 6.5-Wk-Old Boars
Testis weight was increased by approximately 50% in flutamide-treated boars compared with the vehicle-treated littermates at 6.5 wk of age (P = 0.01; Fig. 1). The same flutamide-treated boars had approximately 80% more Sertoli cells at 6.5 wk of age compared with their vehicle-treated littermates (P < 0.01; Fig. 2). Body weight was unaffected by continuous flutamide treatment through 6.5 wk of age (P > 0.05), and inclusion of body weight as a covariate in the statistical model had minimal effect on the statistical significance of flutamide treatment on Sertoli cell numbers or testis weight.
Experiment 2: 11-Wk-Old Boars
At 11 wk of age, 4.5 wk after treatment with flutamide ceased, the number of Sertoli cells and testis weights did not differ between the flutamide-treated boars and their vehicle-treated littermates (Figs. 1 and 2). Again, body weight was not affected by treatment (P > 0.50).
Experiment 3: 2-, 3-, and 4-Wk-Old Boars and Hemicastration
Treatment with flutamide increased Sertoli cell numbers at 2 wk of age (P < 0.01; Fig. 2), although testis weight was not affected (P > 0.05; Fig. 1). However, continuing flutamide treatment through 3 and 4 wk of age did not demonstrate effects on Sertoli cell numbers at 3 or 4 wk of age compared with vehicle-treated littermates. Body weight was again not affected by treatment (P > 0.05). Animals that were hemicastrated at 8 days of age had more Sertoli cells than their intact littermates at 2 wk of age (P < 0.01; Fig. 3). However, the combination of hemicastration at 8 days of age and flutamide treatment reduced the number of Sertoli cells at 2 wk of age compared with the vehicle-treated littermates that were hemicastrated (P < 0.05; Fig. 3). Overall, animals that were hemicastrated at 2 wk of age had significantly more Sertoli cells per testis than their intact littermates at 3 wk of age (P = 0.01). More Sertoli cells were present in animals that were hemicastrated and treated with flutamide than in their littermates that were treated with flutamide, in contrast to the response observed at 2 wk of age (Fig. 3).
Effect of Flutamide Treatment on Hormone Levels and Gene Expression
Plasma testosterone concentration was increased in flutamide-treated boars during the overall treatment period of 1–6.5 wk (P = 0.02) compared with vehicle-treated littermates. These differences in plasma testosterone were significant at Weeks 3 and 4 (Fig. 4). Tissue levels of testosterone at 2 and 3 wk of age were not affected by treatment (Fig. 3). Similarly, testicular tissue levels of testosterone did not differ between vehicle and flutamide-treated littermates at 6.5 wk of age (4 vs. 7 ng/mg protein; SEM = 5). Overall, plasma estradiol concentration was also increased in flutamide-treated boars compared with vehicle-treated littermates (P < 0.01; Fig. 4). Plasma FSH and LH concentrations were increased at Weeks 4 and 5 (Fig. 4). Flutamide treatment reduced expression of the androgen-responsive gene HSPA5 in the testis to 25 ± 13% of the level in the vehicle-treated littermates using the Pfaffl correction (P < 0.05). Results were similar without the Pfaffl correction (ΔCt of −2.45 vs. 0.36 for vehicle and flutamide-treated littermates, respectively; SEM = 0.86; P < 0.05; expression in flutamide-treated boars was approximately 23% of that in vehicle-treated littermates). This is an additional indication that the dose of flutamide was effective at inhibiting androgen receptor activation.
Androgen receptor antagonism (flutamide treatment) in postnatal boars led to a substantial increase in Sertoli cell numbers per testis detectable at the end of treatment at 6.5 wk of age. The presence of androgen receptors on Sertoli cells as well as peritubular myoid cells and interstitial cells is consistent with local mediation of an androgen response during this interval . These data suggest that androgens may limit Sertoli cell population size in pigs, at least during the first postnatal wave of Sertoli cell proliferation. We have previously shown that estrogens also limit Sertoli cell proliferation in the neonatal pig by using both an inhibitor of estrogen synthesis and an estrogen receptor antagonist [11, 33]. This period of Sertoli cell proliferation up to 6.5 wk of age corresponds with declining estrogen and androgen secretion by the testis [9, 11, 43] and as observed here. Our past and present data therefore indicate that blockade of both androgen and estrogen receptor activation further promotes Sertoli cell proliferation during this crucial developmental interval in the boar. This further suggests that the converse may also be true, namely, that androgens and estrogens may promote differentiation of Sertoli cells by closing the proliferative window . Estrogens acting through ESR2 were recently hypothesized to promote Sertoli cell differentiation, although estrogen signaling through ESR1 was believed to stimulate proliferation in cultured rat Sertoli cells, an effect of estrogen that directly contrasts with that observed in vivo in the pig .
The mechanism by which androgen receptor blockade promotes Sertoli cell proliferation in neonatal boars is unknown, and, indeed, the result was somewhat unexpected. Since the initial increase in Sertoli cell numbers observed at 2 wk of age following daily flutamide treatment in this study occurred before any detectable change in gonadotropins, the response appears to be mediated at the testicular level. Few studies investigating the influence of androgen receptor activation on testicular development have been conducted in domestic animals. Flutamide has been given to late pregnant sows and early postnatal boars but for a much shorter interval, and Sertoli cell numbers were not evaluated. However, Sertoli cell numbers were increased in young ram lambs following prenatal exposure to testosterone [46-48]. While this might seem to contrast with the results obtained here, administration of exogenous testosterone increases negative hypothalamic feedback and may well reduce testicular testosterone concentrations and hence reduce testosterone signaling within the testis [49-51]. The observed increases in circulating LH and FSH associated with flutamide treatment observed in the current experiments are consistent with reduced negative feedback at the hypothalamus with concomitant increased trophic drive to the anterior pituitary. These data and the changes in testicular expression of HSPA5, a known androgen receptor-responsive gene, argue strongly that androgen receptor antagonism was effective both systemically and locally within the gonad.
The relative importance of physiological inputs that ultimately determine the size of Sertoli cell populations in the testis likely differs among species. Our previous data demonstrate the particular significance of estrogen signaling as a limiting factor in pigs, which contrasts with the lack of evidence for regulation by thyroid hormone, a well-known determinant of Sertoli cell proliferation in rodents [52-55]. Species differences in Sertoli cell response to altered hormone levels might partially reflect the separation of two distinct waves of proliferation or the presence of a single wave or overlapping waves. Hence, differences in the relative importance of androgen signaling in the establishment of Sertoli cell populations among species should not be surprising. Sertoli cell numbers are reduced in global androgen receptor knockout mice perinatally, suggesting that reduced androgen signaling during fetal development decreases Sertoli cell numbers in rodents [18, 50, 56]. In addition, postnatal treatment of rats with flutamide also decreased Sertoli cell numbers, even though this treatment increased FSH , which typically stimulates Sertoli cell proliferation in rodents . The androgenic effect on rodent Sertoli cells is unlikely to be mediated directly by Sertoli cell androgen receptor function since neonatal rodent Sertoli cells apparently do not express the androgen receptor [57-60] and specific loss of androgen receptor expression in Sertoli cells does not influence the size of the Sertoli cell population . Involvement of peritubular myoid cells has been suggested . In contrast, the presence of androgen receptors in Sertoli cells of neonatal pigs  suggests that restrictive androgen-mediated signaling directly at the Sertoli cell might exist in these young pigs, consistent with apparent differences in how testicular growth and differentiation are regulated in pigs and rodents, as supported by our current and previously published observations.
Although the vehicle-treated boars evaluated at 11 wk of age had more than twice as many Sertoli cells as the vehicle-treated littermates at 6.5 wk of age, the number of Sertoli cells appeared similar in the 6.5-wk and the 11-wk flutamide-treated boars. These animals were of the same genetic background; hence, it appears that the initiation of the second wave of Sertoli cell proliferation was either delayed or halted in the flutamide-treated animals. If the initiation of this second wave was delayed rather than stopped, the delay appears greater than that following letrozole treatment . Both the rapid response to androgen receptor antagonism and the lack of a sustained response in terms of increased Sertoli cell numbers at 11 wk of age argue that the mechanisms involved differ, as do hemicastration and letrozole for similar reasons based on our recent report.
Hemicastration has been a much-utilized strategy for investigating compensatory hypertrophy of the remaining testis and testicular development in general. Postnatal hemicastration causes a compensatory increase in Sertoli cells in the remaining testis to a remarkable degree, depending on the age at which the testis is removed [24, 61-63]. This response is also generally considered to be fairly rapid and was detected within 1 wk in boars hemicastrated at 8 days of age. Recent studies in our laboratory confirm the rapid response to hemicastration but show that the response induced by estrogen deprivation takes much longer, suggesting that different mechanisms may be involved. Moreover, the increase in testicular size and Sertoli cell number in response to hemicastration is even further enhanced by concomitant estrogen deprivation. Therefore, we concluded that Sertoli cell proliferation can be enhanced through different activational pathways, in part characterized by differences in the time course of the response. The response to flutamide in this study was as rapid as hemicastration, suggesting that, unlike the response to estrogen deprivation, the responses to androgen receptor antagonism and hemicastration likely share a similar activational pathway at some level. Since antagonism of the androgen receptor initially prevented the hemicastration response, we conclude that the hemicastration response is also dependent on androgen receptor activation. Hemicastration immediately halves testicular mass, but any reduction in circulating androgen was so short term that it was not detectable at the next blood sampling .
Observations in this study indicate that androgens regulate porcine Sertoli cell proliferation. Androgen receptor antagonism during postnatal Sertoli cell proliferation results in increased Sertoli cell numbers during treatment. Although we cannot exclude a role for FSH in the increased Sertoli cell numbers observed at 6.5 wk of age, the observed increase at 2 wk of age occurs without detectable changes in FSH concentrations. Further exploration of androgen regulation of Sertoli cell proliferation in multiple species will contribute to a better understanding of testicular development.