Endocrinology and Metabolism

Breast cancer resistance protein regulates apical ectoplasmic specialization dynamics stage specifically in the rat testis

Xiaojing Qian, Dolores D. Mruk, Elissa W. P. Wong, C. Yan Cheng

Abstract

Drug transporters determine the bioavailability of drugs in the testis behind the blood-testis barrier (BTB). Thus, they are crucial for male contraceptive development if these drugs (e.g., adjudin) exert their effects behind the BTB. Herein breast cancer resistance protein (Bcrp), an efflux drug transporter, was found to be expressed by both Sertoli and germ cells. Interestingly, Bcrp was not a component of the Sertoli cell BTB. Instead, it was highly expressed by peritubular myoid cells at the tunica propria and also endothelial cells of the microvessels in the interstitium at all stages of the epithelial cycle. Unexpectedly, Bcrp was found to be expressed at the Sertoli-step 18–19 spermatid interface but limited to stage VI-early VIII tubules, and an integrated component of the apical ectoplasmic specialization (apical ES). Apparently, Bcrp is being used by late-stage spermatids to safeguard their completion of spermiogenesis by preventing harmful drugs to enter these cells while they transform to spermatozoa. Also, the association of Bcrp with actin, Eps8 (epidermal growth factor receptor pathway substrate 8, an actin barbed end capping and bundling protein), and Arp3 (actin-related protein 3, a component of the Arp2/3 complex known to induce branched actin polymerization) at the apical ES suggest that Bcrp may be involved in regulating the organization of actin filament bundles at the site. Indeed, a knockdown of Bcrp by RNAi in the testis perturbed the apical ES function, disrupting spermatid polarity and adhesion. In summary, Bcrp is a regulator of the F-actin-rich apical ES in the testis.

  • testis
  • Sertoli cells
  • germ cells
  • spermatids
  • spermatogenesis
  • ectoplasmic specialization
  • F-actin

one of the major obstacles in developing contraceptives for men, in particular drugs that exert their effects in the adluminal compartment, such as adjudin [1-(2,4-dichlorobenzyl)-1H-indazole-3-carbohydrazide] (8, 27) that disrupts germ cell adhesion in the seminiferous epithelium, is the presence of the blood-testis barrier (BTB), which reduces their bioavailability. The BTB in the mammalian testis segregates the seminiferous epithelium into the basal and the adluminal compartment. The BTB restricts paracellular and transcellular transport of substances, including drugs and other biomolecules, besides conferring cell polarity in the seminiferous epithelium (6, 20, 34, 39), limiting the access of drugs to the adluminal (apical) compartment, conferring the “gatekeeper” and the “fence” function of the barrier. Unlike other blood-tissue barriers that are constituted almost exclusively by the endothelial tight junction (TJ) barrier of the microvessel, such as in the brain and eye, which creates the blood-brain and the blood-retina barrier, respectively (4, 15), the BTB is created by TJ between adjacent Sertoli cells near the basement membrane, supported by basal ectoplasmic specialization (basal ES, a testis-specific actin-rich adherens junction), gap junction, and desmosome in the seminiferous tubule (6, 20, 34, 39). The microvessels located in the interstitial space, however, contributed relatively little barrier function to the Sertoli cell BTB (13).

In our quest to develop adjudin into a male contraceptive, it was noted that the bioavailability of this potent drug that induces spermatid depletion from the epithelium is very low, since fewer than 1% of adjudin administered to adult rats by gavage could reach the testis (8). Subsequent studies have shown that this is largely because of the presence of multiple drug efflux transporters in the testis, in particular at the BTB, highly expressed by Sertoli cells to limit the entry of drugs into the seminiferous epithelium (6, 31), such as P-glycoprotein [a member of the multidrug resistance protein (Mdr) family also known as Mdr1 or Abcb1] (42, 45) and Mrp1 (multidrug resistance-related protein 1, a member of the multidrug resistance-related protein family) (3, 19, 42). For instance, P-glycoprotein is an integrated component of the Sertoli cell BTB (42), structurally interacting with integral membrane proteins at the BTB (e.g., occludin, claudin-11, JAM-A). Studies have shown that P-glycoprotein limits the entry of adjudin into the apical compartment, and its knockdown by RNAi was shown to significantly increase the level of [3H]adjudin in the apical compartment (45). Other studies have also shown that Mrp1 is localized to the Sertoli and Leydig cells in mice and humans (3, 19), and it is absent in the endothelial cells of the microvessels in rat testes (49). However, breast cancer resistance protein [Bcrp (also known as ATP-binding cassette, subfamily G, member 2), another ABC (ATP-binding cassette) efflux drug transporter structurally different from Mdr and Mrp families] was earlier shown to be limited to peritubular myoid cells and endothelial cells of microvessels in human testes and not the Sertoli cell (3). Bcrp, as its name implies, was first identified in human breast cancer MCF-7 cells about 15 years ago (10, 36). Subsequent studies have shown that Bcrp is found in many epithelial cells, virtually all cancer cells and endothelial cells, to actively prevent drugs and other substances from entering these cells. Even if drugs can somehow enter cells via influx drug pumps, they are actively being pumped out by Bcrp to maintain cellular homeostasis (32).

Thus, we sought to examine if Bcrp was also restricted to the endothelial cells of microvessels in the rat testis and if it was expressed by Sertoli and germ cells, similar to the other two families of efflux drug transporters. However, we encountered some unexpected findings, which are the subject of this report.

MATERIALS AND METHODS

Animals and antibodies.

Sprague-Dawley rats, both 20-day-old male pups for isolation of Sertoli cells from the testis or adult rats at ∼250–300 g body wt, were purchased from Charles River Laboratories (Kingston, NY) and housed at the Comparative Bioscience Center of Rockefeller University. The use of animals for experiments reported herein was approved by The Rockefeller University Institutional Animal Care and Use Committee (protocol no. 12506). Antibodies used in this study were obtained commercially and listed in Table 1.

View this table:
Table 1.

Antibodies used for different experiments in this report

Treatment of adult rats with adjudin.

Adult rats, 280∼300 g body wt, were treated with a single dose of adjudin (50 mg/kg body wt) by gavage as described (51). This is an established in vivo model to study anchoring junction dynamics, in particular apical ES, at the Sertoli-germ cell interface in the seminiferous epithelium (8). Animals were killed by CO2 asphyxiation at specified time points (n = 3–4 rats for each time point), and testes were removed immediately, snap-frozen in liquid nitrogen, and stored at −80°C until used. Different samples in a single experiment, including treatment vs. control groups, were processed simultaneously, such as to obtain frozen cross sections for dual-labeled immunofluorescence analysis or to obtain lysates for immunoblotting, to avoid interexperimental variations.

Isolation and culture of seminiferous tubules.

Seminiferous tubules were isolated from testes of adult rats of ∼300 g body wt as previously described (53). For seminiferous tubule isolation, decapsulated testes from an adult rat were incubated in collagenase (0.5 mg/ml in F-12-DMEM) at 35°C for 30 min, and contaminating Leydig cells were removed by sedimentation of tubules under gravity. Seminiferous tubules were trimmed into ∼1- to 2-mm fragments and incubated in F-12-DMEM supplemented with bovine insulin (20 μg/ml), human transferrin (20 μg/ml), gentamicin (100 μg/ml), penicillin (100 IU/ml), and streptomycin (100 μg/ml) in a final volume of 50 ml of F-12-DMEM for each testis. Seminiferous tubules were seeded onto 12-well dishes, each containing ∼300–400 tubules (at ∼1–2 mm in length) suspended in ∼4 ml F-12-DMEM, and incubated at 35°C in a humidified atmosphere of 95% air-5% CO2 (vol/vol) in a CO2 incubator. In short, one testis was sufficient to produce enough tubules to plate a 12-well dish for transfection studies routinely.

Bcrp silencing in primary cultured seminiferous tubules in vitro.

After culture for 2 days, seminiferous tubules were transfected with 150 nM nontargeting control small-interfering RNA (siRNA) duplexes (catalog no. 4390844; Ambion) vs. Bcrp-specific siRNA duplexes (catalog no. 4309771-s161792, sense, 5′-GGUAAUGACUACUUGAUAAtt-3′, antisense, UUAUCAAGUAGUCAUUACCag-3′; Ambion) using Ribojuice siRNA transfection reagent (Novagen) as transfection medium according to the manufacturer's instructions. After 24 h, transfection was terminated by rinsing tubules with fresh F-12-DMEM and then culturing in fresh F-12-DMEM with supplements. About 24 h thereafter, cultures were terminated, and tubules were harvested for lysate preparation to assess the effects of RNAi, such as the efficacy of silencing and any off-target effects.

Bcrp silencing in adult rat testes in vivo.

Experimental conditions obtained from the Bcrp knockdown in tubules in vitro were used for studies in vivo. Adult rats (280∼300 g body wt, n = 3 rats) were treated with control and Bcrp RNAi transfection mix via intratesticular injection using a 28-gauge needle as described (26, 46). Each testis received 100 nM siRNA duplexes (for the nontargeting control and the Bcrp silencing group) suspended in the transfection mix (final volume, ∼200 μl) consisting of 7.5 μl Ribojuice siRNA transfection reagent in 192.5 μl Opti-MEM (Invitrogen). The volume of each testis was assumed to be ∼1.6 ml. Control and Bcrp RNAi transfection mix were administered to each of the two testes of the same rat on days 0, 1, and 2, and rats were killed on day 4 by carbon dioxide asphyxiation. Testes were immediately removed, frozen in liquid nitrogen, and stored at −80°C until used. Changes in phenotypes in the seminiferous epithelium using frozen sections between the Bcrp silencing vs. control groups were assessed by fluorescence microscopy using corresponding antibodies (Table 1) and 4′,6-diamidino-2-phenylindole (DAPI) staining for dual-labeled immunofluorescence analysis.

RNA extraction and RT-PCR.

Semiquantitative RT-PCR was performed as previously described (50). Total RNA from testes, Sertoli cells, and germ cells was extracted by TRIzol reagent (Invitrogen) according to the manufacturer's instructions. Contaminated genomic DNA was eliminated by treatment of RNA samples with RNase-free DNase I (Life Technologies). Reverse transcription was performed using Moloney murine leukemia virus reverse transcriptase (Promega) with a mixture of random hexamers and oligo(dT) at a ratio of 4:1 as described (50). Target cDNA was amplified by PCR using GoTaq DNA polymerase (Promega) with the following Bcrp-specific primer pair and coamplified with S16: Bcrp (Genbank NM_181381.2), sense 5′-GTGCCCTTTACTTTGGTC-3′ (nucleotides 1229–1246) and antisense 5′-ACACTTGGCAAGAACCTC-3′ (nucleotides 1443–1460); ribosomal protein S16 (S16) (Genbank NM_001169146.1), sense 5′-TCCGCTGCAGTCCGTTCAAGTCTT-3′ (nucleotides 87–110) and antisense 5′-GCCAAACTTCTTGGATTCGCAGCG-3′ (nucleotides 448–471). The identities of all PCR products were confirmed by nucleotide sequencing at Genewiz (South Plainfield, NJ).

Immunoblotting and coimmunoprecipitation.

Lysates were obtained from testes of adult rats, Sertoli cells [isolated from 20-day-old rat testes and cultured for 4 days in F-12-DMEM as described (29)], and germ cells [isolated from adult rat testes and detailed elsewhere from this laboratory (2)] using lysis buffer with sonication as described (51). Immunoblotting was performed as described (50, 51) using antibodies listed in Table 1. Equal protein loading was assessed by using β-actin. Coimmunoprecipitation (Co-IP) was performed as described (51). Chemiluminescence was performed using a kit prepared in-house and a Fujifilm LAS-4000 Mini Luminiscent Image Analyzer as described (28). Densitometric analysis was performed using SigmaGel (version 1.0). All immunoblotting and Co-IP experiments reported herein were results of at least three independent experiments using either different batches of cells or testis lysates from different cell preparations or animals.

Immunofluorescence microscopy and dual-labeled immunofluorescence analysis.

Dual-labeled immunofluorescence analysis was performed essentially as earlier described (24, 51). Frozen sections of testes at 7 μm (in thickness) were obtained with a cryostat at −21°C, fixed with 4% paraformaldehyde (wt/vol) in PBS for 10 min, and permeabilized in 0.1% Triton X-100 (vol/vol) in PBS for 10 min. After being blocked in 1% BSA (wt/vol) in PBS for 1 h, testis sections were incubated with primary antibodies (Table 1) overnight, and then underwent an 1-h incubation of Alexa Fluor-conjugated secondary antibodies (red fluorescence, Alexa Fluor 555; green fluorescence, Alexa Fluor 488; Invitrogen) at a dilution of 1:250. For F-actin staining, sections were incubated with fluorescein isothiocyanate-conjugated phalloidin (Sigma-Aldrich) at a dilution of 1:70 or together with the secondary antibody for dual-labeled immunofluorescence analysis. All incubations were performed at room temperature. Sections were mounted in Prolong Gold antifade reagent with DAPI (Invitrogen). Fluorescent images were examined and visualized in an Olympus BX61 motorized fluorescence microscope. Images were acquired using an Olympus DP70 12.5 MPx digital camera with the Olympus MicroSuite FIVE software (version 1.224; Olympus Soft Imaging Solutions). Brightness/contrast adjustment and images were merged to examine colocalization using Adobe Photoshop in Adobe Creative Suite (version 3.0; Adobe Systems). To avoid interexperimental variations, all samples within a treatment group (e.g., adjudin-treated rats or Bcrp RNAi) vs. controls were processed simultaneously and examined in a single experimental session. All micrographs presented herein were representative findings of an experiment that was repeated at least three times using different animals and yielded similar results.

Semiquantitative analysis of fluorescence images.

To assess the efficacy of knockdown of Bcrp in studies in vivo, the intensity of Bcrp fluorescence signals in cross sections of rat testes at stage VII was quantified using Image J 1.45 (http://rsbweb.nih.gov/ij; U.S. National Institutes of Health, Bethesda, MD) in both Bcrp knockdown and the nontargeting control groups. Stage VII tubules were selected because Bcrp was expressed almost exclusively at this stage of the epithelial cycle at the apical ES, and earlier studies have shown that ∼20% of the tubules were at stage VII when cross sections of testes in Sprague-Dawley rats were randomly examined (22). At least 20 stage VII tubules from each rat were randomly selected, and the Bcrp signal from each treatment was quantified vs. the control group with n = 3 rats.

Assessing phenotypic changes in the seminiferous epithelium following Bcrp knockdown.

To assess changes in the status of spermatogenesis in the seminiferous epithelium, frozen sections (7 μm thickness) of testes obtained in a cryostat at −21°C and stained with DAPI were used. A disruption of spermatogenesis, such as loss of spermatid polarity and defects in spermiation, was assessed by examination of 300–500 randomly selected seminiferous tubules from cross sections of a testis, and a total of three rats were examined. A seminiferous tubule was scored as defective if it met either one of the following criteria: 1) loss of spermatid polarity, defined by the presence of at least five misoriented spermatids per cross section of a tubule in which the heads of these spermatids were not pointing toward the basement membrane but at least 90° deviated from the proper orientation as found in control rat testis; or 2) defects in spermiation, defined by the presence of at least five elongating/elongated spermatids that were embedded in the seminiferous epithelium after spermiation in a stage IX–X tubule. Data (means ± SD) shown in Table 2 were expressed as a percentage of defective tubules in testes transfected with Bcrp-specific siRNA duplexes vs. the corresponding control rats receiving nontargeting siRNA duplexes.

View this table:
Table 2.

Phenotypic changes in the seminiferous epithelium of adult rat testes following a knockdown of Bcrp by RNAi

General methods and statistical analysis.

Protein concentration in samples was determined by using the DC protein assay kit (Bio-Rad Laboratories) and a Bio-Rad spectrophotometer (model 680), using BSA as a standard. Paraffin sections of normal rat testes stained with hematoxylin were performed as described (44). Each experiment reported herein was repeated at least three times, not including pilot experiments. For in vivo studies, each time point has n = 3 rats, including control rats. Statistical analysis was performed using GB-STAT (version 7.0; Dynamic Microsystems, Silver Spring, MD). Student's t-test was used for paired comparison between rats in the treatment group vs. the control group.

RESULTS

Bcrp is highly expressed by germ cells vs. Sertoli cells in the rat testis.

Bcrp was expressed in rat testes, but mostly by germ cells instead of Sertoli cells when examined by RT-PCR using a primer pair specific to Bcrp (Fig. 1A) and also by immunoblotting (Fig. 1B) using a specific anti-Bcrp antibody (Table 1). Bcrp appeared as a duplex of ∼70 kDa, unless the SDS-PAGE was overloaded with Bcrp protein, which is consistent with earlier findings that Bcrp is a half-transporter that requires dimerization of two Bcrp polypeptides (10, 32) to assemble into a fully flux drug transporter, and the minor size differences between the two polypeptides could be the result of differential glycosylation of the two monomers in the testis (Fig. 1B). The specificity of this antibody was assessed by immunoblotting using lysates of adult rat testes (Fig. 1C). This antibody was used to examine the distribution of Bcrp in the seminiferous epithelium using cross sections of adult rat testes. Consistent with the earlier report that examined the cellular distribution of Bcrp in the human testis (3), Bcrp was also restricted to the myoid cell layer in the tunica propria, which is located below the BTB when the colocalization of Bcrp and F-actin was examined by dual-labeled immunofluorescence analysis (Fig. 1D). The relative location of the BTB is shown Fig. 1D. Bcrp was also expressed intensely by endothelial cells of the microvessel in the interstitial space (Fig. 1D). More important, the expression of Bcrp by myoid cells and endothelial cells of the microvessel was not stage-specific, since they were detected at these sites in all stages of the epithelial cycle (Fig. 1D). Figure 1E shows the cross sections of three staged tubules and the relative location of the microvessels in the interstitium. Thus, Bcrp was shown to be restricted to the peritubular myoid cells in the tunica propria and endothelial cells of the microvessels in the interstitium. Bcrp was also detected at the apical ES but was expressed stage specifically. It was first detected in stage VI, became highly expressed in stage VII tubules, but considerably reduced at stage VIII, and was not detectable at the apical ES in all other stages (Fig. 1D).

Fig. 1.

Expression of breast cancer resistance protein (Bcrp) by Sertoli and germ cells in the rat testis. A: RT-PCR using primer pair specific to Bcrp illustrated the expression of Bcrp mRNA in the testis vs. Sertoli (SC) and germ (GC) cells. Bcrp was coamplified with S16, which served as a loading control. B: immunoblotting was performed using a specific anti-Bcrp antibody (Table 1) and lysates of adult rat testes (20 μg protein), Sertoli cells (isolated from 20-day-old rat testes and cultured for 4 days in serum-free F-12-DMEM), and germ cells (total germ cells isolated from adult rat testes and cultured for 16 h in vitro) (50 μg protein/lane for lysates of Sertoli and germ cells) to confirm findings shown in A illustrating Bcrp was expressed in the testis, but mostly restricted to germ cells and weakly expressed by Sertoli cells, with β-actin serving as a protein loading control. C: the specificity of the anti-Bcrp antibody was assessed by immunoblotting using lysates of adult rat testes (20 μg protein). Mr, relative molecular weight. D: the anti-Bcrp antibody was used for dual-labeled immunofluorescence analysis to examine the cellular localization of Bcrp (red fluorescence) vs. F-actin (green fluorescence) in the seminiferous epithelium of adult rat testes. Bcrp was localized almost exclusively to the myoid cells in the tunica propria (see “blue” arrowheads in the stage IV tubule that annotate the relative location of the tunica propria in the cross section of a seminiferous tubule), just below the blood-testis barrier (BTB) site where F-actin was highly expressed (see “white” arrowheads in the stage VI tubule that annotate the relative location of the BTB in the panel on the right), and also to the endothelial cells of the microvessels located in the interstitial space as seen in stage VII and VIII tubules (see “yellow” arrowheads in the stage VIII tubule that annotate the endothelial cells of the microvessel in the interstitium in the panel on the right). Bcrp was also expressed stage specifically at the apical ectoplasmic specialization (ES). It first appeared in stage VI tubules with relatively weak staining, and it became highly expressed in stage VII tubules, which also partially colocalized with F-actin at the apical ES. Bcrp expression diminished considerably at stage VIII tubules, and it was no longer detectable in stage IX–XI and also I–V tubules. However, Bcrp did not appear to colocalize with F-actin in the endothelial and/or smooth muscle cells of microvessels. Cell nuclei were visualized by 4′,6-diamidino-2-phenylindole (DAPI). Scale bar = 50 μm, which applies to all micrographs in this panel. E: shown is a normal adult rat testis showing the cross sections of three staged tubules at VII, VIII, and late stage VIII of the epithelial cycle using paraffin sections of testes stained by hematoxylin, which is used to illustrate the relative location of different cellular compartments being referred to in this report. The seminiferous epithelium, composed of Sertoli cells and germ cells at different stages of development (such as spermatocytes, round spermatids, and elongating/elongated spermatids), lies on the tunica propria, which is constituted by the basement membrane (composed by type IV collagen, laminin α2, heparin sulfate proteoglycan, and entactin) and type I collagen layer, to be followed by the myoid cell layer and the lymphatic endothelium (11, 41). Small blood vessels (annotated by “red” arrowheads) in the testis are restricted to the interstitium (In). Scale bar = 150 μm.

Stage-specific expression of Bcrp at the Sertoli-elongated spermatid interface at the apical ES during the seminiferous epithelial cycle of spermatogenesis.

Besides myoid cells and endothelial cells of the microvessels, Bcrp was also detected at the Sertoli-spermatid interface known as the apical ES (which is the similar ultrastructure as the basal ES at the BTB, except that actin filament bundles that are the hallmark structure of the ES are found only on the Sertoli cell side at the apical ES vs. both sides of the two adjacent Sertoli cells in the basal ES), but this expression was highly restrictive, limited only to stage VI–VIII, partially colocalized with F-actin (Fig. 1D). This restrictive spatiotemporal expression of Bcrp at the apical ES was further analyzed by dual-labeled immunofluorescence analysis (Fig. 2), illustrating Bcrp partially colocalized with F-actin. Namely, the actin filament bundles at the apical ES began at late stage VI, most intensely at stage VII, and gradually diminished at stage VIII of the epithelial cycle (Fig. 2). However, it was noted that Bcrp did not express at the basal ES at the BTB (Fig. 1D).

Fig. 2.

Stage-specific expression of Bcrp and its colocalization with F-actin at the apical ES during the epithelial cycle of spermatogenesis in adult rat testes. The expression of Bcrp (red fluorescence) and F-actin (green fluorescence) at the apical ES was examined by dual-labeled immunofluorescence analysis in the seminiferous epithelium of adult rat testes. Bcrp expressed stage specifically. It was first detected in stage VI tubules, and it was most prominently expressed in stage VII tubules and gradually diminished by stage VIII. Bcrp was found to be restricted almost exclusively to the concave side of step 18 (stage VI) and step 19 (stage VII–VIII) spermatid heads, colocalized with F-actin. Scale bar = 10 μm, which applies to all other micrographs in this panel.

Bcrp structurally associates with actin, Eps8, and Arp3 at the apical ES.

Because of the partial colocalization of Bcrp with F-actin at the apical ES, we next examined if Bcrp indeed structurally interacts with F-actin and several other actin-regulatory and binding proteins, as well as apical ES proteins in the rat testis by Co-IP. Indeed, Bcrp was found to interact with actin (Fig. 3). Interestingly, Bcrp also structurally interacted with Eps8 [epidermal growth factor receptor pathway substrate 8, an actin barbed end capping and bundling protein known to promote and maintain the actin filament bundles at the apical ES (25)], Arp3 [actin-related protein 3, which together with Arp2 and ARPC1–5 (actin-related protein 2/3 complex component 1 to 5) creates a 7-subunit protein complex known to induce branched actin polymerization, altering the actin filament bundles at the apical ES to a “branched” and “unbundled” configuration, thereby destabilizing the apical ES function (24)], and APRC2 but not any of the apical ES proteins examined, including N-cadherin, β-catenin, and JAM-C, in a study using the Co-IP technique (Fig. 3A). The association of Bcrp with Eps8 and Arp3 [because of the lack of a rabbit anti-Arp3 antibody for dual-labeled immunofluorescence analysis, we used an anti-ARPC2 antibody instead of anti-Arp3 antibody for dual-labeled immunofluorescence analysis, since ARPC2 forms a 7-subunit complex with Arp3 to create the functional Arp2/3 complex (6, 35)] was further confirmed by its colocalization with either Eps8 (Fig. 3B) or ARPC2 (Fig. 3C), localized mostly to the concave side of the spermatid head (Fig. 3, B and C).

Fig. 3.

Bcrp, an efflux drug transporter, is an integrated component of the apical ES that expressed stage specifically during the epithelial cycle. A: coimmunoprecipitation (Co-IP) was used to assess any structural interaction between Bcrp and actin, as well as a number of actin-binding and regulatory proteins. It was shown that Bcrp structurally interacted with actin, but also actin regulatory proteins epidermal growth factor receptor pathway substrate 8 (Eps8), actin-related protein 3 (Arp3), actin-related protein 2/3 complex component 2 [ARPC2; but not drebrin E (also an actin-binding protein)], palladin (an actin cross-linking and bundling protein), and several other basal ES (e.g., N-cadherin, β-catenin, afadin, JAM-C) and tight junctions (TJ) (e.g., ZO-1, occludin, JAM-A) proteins at the BTB. Data shown herein are the results of a representative experiment that was repeated three times using different rat testes and yielded similar results. (+), interaction with Bcrp was positively identified; (−), interaction with Bcrp was not identified. B and C: the association of Bcrp (red) with Eps8 (green) (B) and ARPC2 (green) (C) was further confirmed by dual-labeled immunofluorescence analysis to assess their colocalization. Colocalization between Bcrp and F-actin had been shown (see Fig. 2). Colocalization of Bcrp with Arp3 was not done because both anti-Bcrp and anti-Arp3 were mouse antibodies (see Table 1). Bcrp was found to colocalize with Eps8 and APRC2 in stage VII tubules. Similar to Bcrp, the expression of Eps8 and ARPC2 was considerably diminished in stage VIII tubules. Scale bar = 10 μm in B and C, which applies to all other micrographs in both panels.

Bcrp is not a component of the BTB.

To further confirm that Bcrp is not a component of the BTB, dual-labeled immunofluorescence analysis was performed to examine colocalization of Bcrp with either ZO-1 (a TJ adaptor protein), claudin-11 (a TJ integral membrane protein), or actin at the BTB. Indeed, Bcrp was limited to the tunica propria, associated with myoid cells, but not at the Sertoli cell BTB (Fig. 4). Furthermore, Bcrp was also highly expressed by endothelial cells of the microvessels in the interstitium (Fig. 4).

Fig. 4.

Bcrp is not expressed at the BTB, but it is expressed by peritubular myoid cells at the tunica propria and endothelial cells of the microvessels in the interstitial space. A: dual-labeled immunofluorescence analysis of Bcrp (red) with either ZO-1 (green, a TJ adaptor protein), claudin-11 (green, a TJ integral membrane protein), or F-actin (green) in the seminiferous epithelium near the tunica propria was shown here. Bcrp was not associated with either ZO-1 or cluadin-11 in the seminiferous epithelium. In fact, Bcrp and ZO-1/claudin-11 were localized to distinctively different locations near the base of the seminiferous tubule in which Bcrp was localized to the tunica propria and associated with peritubular myoid cells (see yellow arrowheads), whereas ZO-1 and cluadin-11 were at the BTB, at the site above spermatogonia. However, Bcrp displayed partial colocalization with F-actin at the tunica propria (see “orange” arrowheads). B: Bcrp was highly expressed by endothelial cells of the microvessels in the interstitium (see white arrowheads), but it did not colocalize with F-actin in endothelial cells. Scale bar = 10 μm for A and 20 μm for B, which applies to other micrographs in the corresponding panel.

Downregulation of Bcrp expression at the apical ES after treatment of rats with adjudin.

In rats treated with a single dose of adjudin (50 mg/kg body wt, by gavage), which was earlier shown to induce rapid spermatid loss from the epithelium in ∼6–12 h following treatment by disrupting the apical ES (5), the expression of Bcrp was found to be downregulated (Fig. 5, A and B). The findings by immunoblotting were further confirmed by dual-labeled immunofluorescence analysis in which the expression of Bcrp at the apical ES was considerably diminished in similarly staged tubules, namely stage VII, by 12 h after adjudin treatment. By 48 h, virtually all of the elongated spermatids were either depleted or depleting from the epithelium, and the expression, illustrating spermatids that lost adhesive function, of Bcrp was reduced (Fig. 5C). The expression of Bcrp at the tunica propria was also mildly downregulated (Fig. 5D), consistent with findings of immunoblotting (see Fig. 5, A and B).

Fig. 5.

A rapid and considerable loss of Bcrp expression at the apical ES during adjudin-induced germ cell loss from the seminiferous epithelium. Rats were treated with a single dose of adjudin (50 mg/kg body wt, by gavage) to induce germ cell loss from the epithelium by disrupting apical ES function in the testis. A: immunoblot analysis to illustrate a time-dependent downregulation of Bcrp in the testis following adjudin treatment. Actin served as a protein loading control. B: this histogram summarizes results of immunoblotting shown in A, with each bar a mean ± SD of n = 3–5 rats. A time-dependent downregulation on the expression of Bcrp was noted. *P < 0.05. C: localization of Bcrp at the apical ES and its colocalization with F-actin was noted (see orange merged staining annotated by yellow arrowheads); however, a considerable downregulation of Bcrp expression at the apical ES was noted as early as 12 h after adjudin treatment. By 48 and 96 h, the expression of Bcrp at the apical ES was virtually undetectable. Scale bar = 20 μm, which applies to other micrographs in this panel. D: in contrast to the Bcrp at the apical ES, the expression of Bcrp at the tunica propria in the testis from rats following treatment of adjudin at 12, 48, and 96 h only displayed gradual downregulation, consistent with data shown in A and B. Scale bar = 20 μm, which applies to other micrographs in this panel.

Knockdown of Bcrp in the testis in vivo perturbs the apical ES function, leading to a loss of spermatid polarity in stage VII tubules.

We first used seminiferous tubules cultured in vitro to assess the efficacy of Bcrp knockdown (Fig. 6, A and B). It was noted that a knockdown of Bcrp by ∼60% using seminiferous tubules did not lead to any off-target effects when the steady-state levels of Eps8, Arp3, and β-actin were assessed (Fig. 6, A and B). These experimental conditions were used for the knockdown of Bcrp in the testis in vivo in which the Bcrp fluorescence signal in the seminiferous epithelium was shown to be diminished by ∼60% in seminiferous tubules, including apical ES in stage VII tubules and tunica propria in all tubules examined (Fig. 6, C and D), consistent with the findings by immunoblotting in which Bcrp was knocked down in tubules in vitro (Fig. 6, A and B). We focused on stage VII tubules since this is the stage in which Bcrp expression was upregulated at the apical ES (see Figs. 1 and 3). It was noted that a knockdown of Bcrp in stage VII tubules led to a loss of elongated spermatid polarity, as well as their premature release from the epithelium into the tubule lumen (Fig. 6C). Furthermore, a loss of spermatid polarity and adhesion thus induced defects in spermatid transport across the epithelium, causing defects in spermiation in which elongated spermatids were “trapped” inside the seminiferous epithelium in stage IX tubules (Fig. 6C). To further investigate the underlying mechanism by which spermatid polarity and adhesion in the seminiferous epithelium were affected following the knockdown of Bcrp in the testis in vivo, we examined changes in the localization of F-actin, Eps8, and Arp3 in testes transfected with Bcrp-specific siRNA duplexes vs. nontargeting control siRNA duplexes (Fig. 7). Following the knockdown of Bcrp in the seminiferous epithelium in which the expression of Bcrp was considerably downregulated, spermatids were misoriented, and this loss of spermatid polarity was associated with changes in the organization of the actin filament bundles at the apical ES, in which F-actin at the apical ES was considerably diminished (Fig. 7) [Note: in the control testis where nontargeting siRNA duplexes were administered and spermatids were aligned with correct orientation (or polarity) in the seminiferous epithelium, the heads of elongated spermatids were pointing toward the basement membrane in the tunica propria (Fig. 7); however, in the treatment groups, the heads of spermatids were pointing randomly at different directions.]. This change was shown to be associated with a mislocalization of Eps8 (1, 25, 30) and Arp3 (7, 24, 47) in which these two proteins were no longer localized at the concave side near the tip of the spermatid heads to confer the needed spermatid adhesion and polarity. Instead, they were mislocalized and “diffused” away from the apical ES (Fig. 7). Thus, actin filament bundles could not be properly maintained at the apical ES, leading to a loss of spermatid polarity and spermatid adhesion. Table 2 summarizes these findings in selected staged tubules, illustrating the defects in polarity as the result of mislocalization of Eps8 and Arp3, which in turn perturbed a proper organization of actin filament bundles at the apical ES. This thus impeded spermatid adhesion such that elongated spermatids failed to undergo proper release at spermiation, and they were trapped in the seminiferous epithelium in stage IX tubules, after spermiation is complete. These findings, however, show that Bcrp is involved in the apical ES function by conferring spermatid polarity and adhesion via its structural and functional interaction with Eps8 and Arp3.

Fig. 6.

Effects of Bcrp knockdown in the adult rat testis on the apical ES function. A and B: Bcrp was knocked down by RNAi with specific Bcrp siRNA duplexes (Bcrp RNAi) vs. nontargeting control duplexes (Ctrl RNAi) using seminiferous tubules isolated from adult rat testes and cultured in vitro as described in materials and methods. As shown in A, the knockdown of Bcrp in seminiferous tubules in vitro by ∼60% (B) did not lead to any off-target effects when the steady-state levels of Eps8, Arp3, and actin were assessed using 20 μg protein/lane in which actin served as a protein loading control. As shown in the histogram in B, each bar is a mean ± SD of n = 3 independent experiments using different batches of tubules isolated from adult rat testes. **P < 0.01. The experimental conditions shown in A and B were then used for studies in vivo shown in C and D in which each testis was transfected with either Bcrp specific siRNA duplexes vs. control duplexes as described in materials and methods. In short, each testis was transfected daily beginning on days 0, 1, and 2 with a total of three transfections, and rats were euthanized by CO2 asphyxiation on day 4 to assess any changes in phenotypes following the knockdown of Bcrp (C). It was noted that Bcrp was silenced by at least ∼60% (C and D) when the intensity of Bcrp signals in the seminiferous epithelium in Bcrp knockdown vs. control groups were assessed (D). C: a considerable decline in Bcrp signal at the apical ES in a stage VII tubule was noted in the Bcrp knockdown group (Bcrp RNAi) vs. the control group (Ctrl RNAi), which was also associated with a loss of spermatid adhesion (see white arrowheads, illustrating spermatids found in the tubule lumen in stage VII and early VIII tubules) and loss of spermatid polarity (see yellow arrowheads, illustrating misoriented spermatids). In stage IX tubules, elongated spermatids were found to be “trapped” in the seminiferous epithelium due to defects of spermiation because of a disruption of the apical ES function following Bcrp knockdown (see “green” arrowheads). The white boxed area in the panel on top was magnified and shown on bottom. Scale bar in the micrograph in panels on the top and bottom is 100 and 25 μm, respectively, which also apply to all other micrographs in the corresponding panel. Each bar in D is a mean ± SD of n = 60 tubules from 3 rats (20 randomly selected stage VII–VII tubules from each rat). **P < 0.01.

Fig. 7.

A knockdown of Bcrp in the testis impedes spermatid polarity and adhesion via changes in F-actin organization at the apical ES mediated by mislocalization of Eps8 and Arp3. A knockdown of Bcrp in the testis with Bcrp-specific siRNA duplexes (Bcrp RNAi) vs. nontargeting control duplexes (Ctrl RNAi) (see materials and methods) led to reduced Bcrp signals in the tunica propria and the apical ES in stage VII tubules when the expression of Bcrp was the highest. It was noted that a knockdown of Bcrp in these stage VII tubules by ∼60% led to loss of spermatid polarity (see yellow arrowheads). This loss of spermatid orientation was likely the result of changes in the organization of actin filament bundles at the apical ES, since F-actin was prominent at the apical ES in stage VII tubules in the Ctrl RNAi group but not in the Bcrp RNAi group. These changes in F-actin organization at the apical ES appeared to be mediated by mislocalization of: 1) Eps8 at the apical ES, which failed to maintain the actin filament bundles at the apical ES {but not at the BTB in the seminiferous epithelium, since Bcrp was not found at the BTB [see Figs. 1 and 4; see white arrowheads annotating the expression of Eps8 at the BTB in both the Bcrp RNAi and Ctrl RNAi groups, consistent with earlier reports that Eps8 is highly expressed (25) but Arp3 is downregulated (24) at the BTB in stage VII tubules]} in which Eps8 was no longer tightly localized to the concave side of spermatid heads, and also 2) Arp3 in which Arp3 was not restricted to the concave side of spermatid heads but diffused away from this site (see enlarged images in selected panels for Eps8 and Arp3 vs. Bcrp and F-actin). Boxed areas in micrographs were enlarged to illustrate changes in localization of F-actin, Eps8, and Arp3 in Bcrp RNAi vs. Ctrl RNAi groups. Cell nuclei were visualized by DAPI staining. Scale bar = 20 μm in micrograph and scale bar = 10 μm in enlarged micrograph, which apply to remaining micrographs in corresponding panels.

DISCUSSION

The BTB, unlike other blood-tissue barriers that are constituted almost exclusively by the endothelial TJ barrier of the capillary, is contributed almost exclusively by specialized junctions between adjacent Sertoli cells near the basement membrane in the mammalian testis (6, 20, 34, 39), in short, the basal ES that coexists with TJ and gap junction, which together with desmosome constitute the BTB in the mammalian testis. The basal ES is an F-actin-rich ultrastructure in which actin filament bundles that lie perpendicular to the opposing Sertoli cell plasma membrane are sandwiched in between cisternae of endoplasmic reticulum and the plasma membrane. This ultrastructural feature unique to the BTB was first reported in the literature in the early 1970s (17, 18), and the name basal ES was not used until late in the decade when a similar structure was also found at the Sertoli-spermatid interface designated apical ES (37), except that the actin filament bundles at the apical ES were restricted to the Sertoli cell and any ultrastructure residing in the spermatid was not apparent. Since then, the detailed morphology of the ES, both the apical and the basal ES, has been described in detail (6, 34, 38, 40, 48). Although the BTB confers one of the tightest blood-tissue barriers, it undergoes restructuring at stage VIII of the epithelial cycle so that preleptotene spermatocytes differentiated from type B spermatogonia connected in “clones” via intercellular bridges (14) residing in the basal compartment can traverse the BTB to enter the adluminal compartment to differentiate into zygotene and pachytene spermatocytes to prepare for meiosis I and II (21, 33). Thus, it is not unusual that many drug transporters are expressed by endothelial cells of microvessels that create the endothelial TJ barrier in other blood-tissue barriers (e.g., blood-brain barrier and the blood-retinal barrier), such as efflux drug transporters P-glycoprotein and Mrp1 (6, 31, 39), since they are being used to protect the organs (e.g., the brain and the eye) behind these barriers by actively preventing toxic drugs and substances from gaining entry into the corresponding organs and keeping them out of the endothelial TJ barrier. Because the BTB was constituted by junctions between Sertoli cells, thus, P-glycoprotein, Mrp1, and other influx drug transporters (e.g., Oatp3) were reported to be highly expressed by Sertoli cells. They were detected at the BTB in the seminiferous epithelium, colocalized with known TJ (e.g., occludin, claudin-11, JAM-A, and ZO-1) and basal ES (e.g., N-cadherin and β-catenin) at the BTB (42, 43) and used to “guard” the BTB, protecting meiosis and postmeiotic spermatid development from potential harmful substances in the systemic circulation. Interestingly, Bcrp is also an ABC efflux transporter, but it was not detected at the Sertoli cell BTB in the seminiferous epithelium as reported herein, consistent with earlier reports (3, 9, 16). Instead, Bcrp was expressed by peritubular myoid cells in the tunica propria in the rat testis in all stages of the epithelial cycle, consistent with an earlier study reporting the localization of Bcrp in myoid cells in the human testis (3). Interestingly, in rodent testes, the peritubular myoid cell layer was found to effectively block the passage of tracers/markers (e.g., colloidal carbon, thorium, lanthanum salt), and these electron-opaque markers could penetrate only ∼15% of the tubules beyond the myoid cell layer when examined by electron microscopy (13, 18), even though the myoid cell layer was not effective in blocking the penetration of these markers in primates (12), perhaps humans. Taken collectively, these findings suggest that the testis, at least in rodents, has two “layers” of barrier in place, one at the Sertoli cell and the other at the peritubular myoid cells by equipping different drug transporters in both cell types to protect developing germ cells from being exposed to toxicants during spermatogenesis.

In this context, it is noted that, while we have not examined in this study if the knockdown of Bcrp would modulate the bioavailability of adjudin in the testis, in particular at the adluminal compartment, a recent study has shown that a knockdown of P-glycoprotein by RNAi, another efflux drug transporter in the testis but confined mostly to the Sertoli cell BTB, was shown to promote the apparent permeability of [3H]adjudin across the BTB, significantly increasing the bioavailability of [3H]adjudin in the apical compartment (45). Thus, it is tempting to speculate that the knockdown of Bcrp would facilitate the transport of [3H]adjudin across the peritubular myoid cell layer in the tunica propria. However, this possibility must be carefully evaluated in future studies using seminiferous tubules isolated from Bcrp−/− mice (23), such as from Taconic Farms (Hudson, NY), versus the wild-type to examine if the knockout of Bcrp would facilitate the bioavailability of [3H]adjudin in the adluminal compartment behind the BTB, since the presence of P-glycoprotein at the Sertoli cell BTB may still pose a significant barrier to adjudin.

The findings that Bcrp is an integrated component of the actin filament-rich apical ES, structurally associated with actin and the actin barbed end capping and bundling protein Eps8, and the branched actin polymerization protein Arp3, are intriguing. First, the expression of Bcrp at the apical ES is highly stage specific. Bcrp first appeared at the apical ES at stage VI but at a very low level of expression; it was highly expressed and most predominant at stage VII, but it rapidly diminished thereafter at stage VIII; by stage IX–XIV and also stage I–V, virtually no Bcrp immunoreactive signal was detected at the apical ES. This finding suggests that Bcrp may be involved in protecting spermatid development, in particular the final stage of spermiogenesis, such as the transformation of step 19 spermatids at stage VII–VIII tubules to spermatozoa, so that any drug(s) that can possibly affect these spermatids adversely can be actively pumped out (or be prevented from gaining entry in these spermatids at the final stage of their development before spermiation). Second, the stage-specific association of Bcrp with Eps8 and Arp3, as well as with F-actin, at the apical ES suggests that Bcrp may be involved in apical ES restructuring at stage VII to prepare for its degeneration at spermiation that occurs at stage VIII of the epithelial cycle. This possibility is supported by studies in which the expression of Bcrp was silenced by RNAi, since a knockdown of Bcrp by ∼60% perturbed the F-actin organization via changes in the localization of Eps8 and Arp3 at the apical ES in stage VII tubules when the expression of Bcrp is high. It is conceivable that the mislocalization of Eps8 and Arp3 at the apical ES leads to changes in the configuration of the actin filament bundles at the apical ES. Thereby, F-actin is no longer properly maintained at the apical ES as reported herein. This concept is also supported by the observations that some elongated spermatids were found to be misoriented, losing their polarity, and were either depleted from the seminiferous epithelium prematurely or trapped in the epithelium, leading to defects in spermiation in rats following the knockdown of Bcrp by RNAi, which are the result of a loss of structural support via F-actin at the apical ES. It is noted that these changes in phenotypes, namely loss of spermatid adhesion and polarity, following the knockdown of Bcrp, are also two of the crucial functions of the apical ES (27, 38, 48), supporting the notion that Bcrp is involved in F-actin organization at the apical ES. Furthermore, recent findings from our laboratory have shown that the interplays of Eps8 and the Arp2/3 complex (24, 25), possibly involving Par- and Scribble-based polarity complexes (46, 51, 52), are crucial to regulate the organization of actin filament bundles at the ES. Taking these findings collectively, Bcrp may play an important role in the organization of the unique actin filament bundles at the apical ES via its structural and possibly functional interactions with Eps8 and Arp3 during the epithelial cycle. Third, since the actin filament bundles are ultrastructurally identical between the basal and the apical ES, but Bcrp was not found at the basal ES that constitutes the BTB, it is suggested that there may be some functional differences between the basal and apical ES in the seminiferous epithelium. Indeed, when Bcrp was knocked down in the seminiferous epithelium via its transfection in the testis in vivo, the stage-specific expression of Eps8 and Arp3 at the basal ES, in contrast to the apical ES, was not affected, suggesting that the organization of the actin filament bundles at the basal ES was not altered.

In summary, we have demonstrated herein Bcrp is a testis-associated efflux drug transporter; however, it is localized to the endothelial TJ barrier in microvessels at the interstitium and also the peritubular myoid cells in the testis, but not at the Sertoli cell BTB. Interestingly, Bcrp is also an integrated component of the apical ES and structurally interacts with Eps8, Arp3, and F-actin. Bcrp appears to play a role in regulating the configuration of the actin filament bundles at the apical ES via its structural and possibly functional interactions with Eps8 and Arp3, involving in the organization of the actin filaments from their “bundled” to their “unbundled/branched” configuration.

GRANTS

Studies from the authors' laboratory were supported by grants from the National Institutes of Health (R01-HD-56034 to C. Yan Cheng; U54-HD-029990 Project 5 to C. Yan Cheng)

DISCLOSURES

The authors have nothing to disclose.

AUTHOR CONTRIBUTIONS

Author contributions: X.Q., E.W.W., and C.Y.C. performed experiments; X.Q., D.D.M., and C.Y.C. analyzed data; X.Q., D.D.M., and C.Y.C. interpreted results of experiments; X.Q. and C.Y.C. prepared figures; C.Y.C. conception and design of research; C.Y.C. and X.Q. drafted manuscript; C.Y.C. and X.Q. edited and revised manuscript; C.Y.C. approved final version of manuscript.

R01-HD-56034U54-HD-029990 Project 5National Institutes of Health

REFERENCES

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
  45. 45.
  46. 46.
  47. 47.
  48. 48.
  49. 49.
  50. 50.
  51. 51.
  52. 52.
  53. 53.
View Abstract