F5-Peptide and mTORC1/rpS6 Effectively Enhance BTB Transport Function in the Testis—Lesson From the Adjudin Model

Abstract

During spermatogenesis, the blood–testis barrier (BTB) undergoes cyclic remodeling that is crucial to support the transport of preleptotene spermatocytes across the immunological barrier at stage VIII to IX of the epithelial cycle. Studies have shown that this timely remodeling of the BTB is supported by several endogenously produced barrier modifiers across the seminiferous epithelium, which include the F5-peptide and the ribosomal protein S6 [rpS6; a downstream signaling molecule of the mammalian target of rapamycin complex 1 (mTORC1)] signaling protein. Herein, F5-peptide and a quadruple phosphomimetic (and constitutively active) mutant of rpS6 [i.e., phosphorylated (p-)rpS6-MT] that are capable of inducing reversible immunological barrier remodeling, by making the barrier “leaky” transiently, were used for their overexpression in the testis to induce BTB opening. We sought to examine whether this facilitated the crossing of the nonhormonal male contraceptive adjudin at the BTB when administered by oral gavage, thereby effectively improving its BTB transport to induce germ cell adhesion and aspermatogenesis. Indeed, it was shown that combined overexpression of F5-peptide and p-rpS6-MT and a low dose of adjudin, which by itself had no noticeable effects on spermatogenesis, was capable of perturbing the organization of actin- and microtubule (MT)-based cytoskeletons through changes in the spatial expression of actin- and MT-binding/regulatory proteins to the corresponding cytoskeleton. These findings thus illustrate the possibility of delivering drugs to any target organ behind a blood–tissue barrier by modifying the tight junction permeability barrier using endogenously produced barrier modifiers based on findings from this adjudin animal model.

The blood–testis barrier (BTB), created by the actin-based tight junction (TJ) between adjacent Sertoli cells, which are supported by the basal ectoplasmic specialization (ES), is an important ultrastructure to sustain spermatogenesis [for reviews, see (1–4)]. The BTB also divides the seminiferous epithelium into the basal and the adluminal compartments so that the cellular events pertinent to meiosis I/II and postspermatid development via spermiogenesis all take place in the specialized microenvironment of the adluminal compartment behind the BTB [for reviews, see (5–7)]. In the mammalian body, the BTB also poses a major hurdle for delivery of drugs to the testis because of the “gate” and “fence” function that limits paracellular (between cells) and transcellular (across cells) transport of substances across the barrier. Sertoli cells that create the BTB and germ cells in the testis also express many drug transporters [for reviews, see (1, 8)], such as P-glycoprotein [also known as or multidrug resistance protein 1 (Mdr1)] (9–12) and multidrug resistance-related protein 1 (Mrp1) (10, 13) that actively pump drugs out of the testis. Collectively, this unusual function of the BTB coupled with the network of drug transporters in the testis thus hampers our efforts of delivering drugs to the testis.

Interestingly, during the epithelial cycle, the BTB undergoes cyclic remodeling to facilitate the transport of preleptotene spermatocytes (differentiated from type B spermatogonia), often connected in clones via intercellular bridges (14, 15), across the BTB at stage VIII to IX of the cycle in rodents (16) vs stage VI to VII in humans (17). More importantly, BTB remodeling takes place near the base of the tubules at stage VIII of the cycle and is concomitant with the release of sperm at spermiation near the tubule lumen, but these events occur at opposite ends of the epithelium (5, 7). These observations thus prompted us to investigate whether there are regulatory biomolecules produced in the epithelium to coordinate these cellular events. It was first discovered that the F5-peptide generated from domain IV of the laminin-γ3 chain at the apical ES [for reviews, see (18, 19)] was a potent biologically active peptide, capable of inducing BTB remodeling, making it “leaky” reversibly, and also promoting apical ES degeneration to facilitate spermiation (20–22). Additionally, the laminin-α2 chain in the basement membrane releases an 80-kDa fragment from the N-terminal tail region possibly through the action of matrix metalloproteinase 9 (MMP9), which, in sharp contrast to the F5-peptide, promotes Sertoli cell BTB TJ barrier function through the mammalian target of rapamycin complex 1 (mTORC1)/ribosomal protein S6 (rpS6)/Akt1/2 signaling pathway (23, 24). In short, an inactivation of laminin-α2 by RNA interference (RNAi) was found to perturb Sertoli cell TJ function through an upregulation of phosphorylated (p-)rpS6 concomitant with a downregulation of p-Akt1/2 (23). These findings are consistent with earlier reports that overexpression of a phosphomimetic (i.e., constitutively active) mutant of rpS6, namely p-rpS6-MT by mutating Ser (S) to Glu (E) at S235E, S236E, S240E, and S244E, in Sertoli cell epithelium to activate mTORC1 was capable of perturbing Sertoli cell adhesion and barrier function (25, 26). Collectively, these findings thus illustrate that the testis is producing regulatory peptides/biomolecules that support BTB and apical ES function by modulating Sertoli cell and elongated spermatid adhesion. In this study, we sought to test the hypothesis that these two biomolecules are capable of making the BTB leaky in vivo, thereby modifying its transport function by promoting the entry of a nonhormonal male contraceptive, adjudin [for reviews, see (27, 28)], into the testis to induce germ exfoliation at a considerably reduced dose.

Materials and Methods

Animals

Sprague-Dawley rats [adult males at 275 body weight (b.w.) at ∼70 days of age] were purchased from Charles River Laboratories (Kingston, NY). The use of animals for studies reported herein were approved by the Rockefeller University Institutional Animal Care and Use Committee (protocol nos. 15-780-H and 18-043H). Rats were euthanized at specified time points by CO₂ asphyxiation using slow (20% to 30% per minute) displacement of chamber air from a compressed CO₂ tank via a gas regulator that governed the airflow into a euthanasia chamber that was approved by the Rockefeller University Laboratory Safety and Environmental Health. The function of the gas regulator and euthanasia chamber was inspected by Laboratory Safety and Environmental Health personnel annually. Rats were housed in groups of two animals per cage at the Rockefeller University Comparative Bioscience Center. All rats had free access to standard rat chow and water ad libitum at 20 ± 1°C with a 12-hour light/12-hour dark cycle. The uses of recombinant DNA materials such as cDNAs and transfection medium (e.g., Polyplus in vivo-jetPEI^®), as well as the detailed protocols using Sprague-Dawley rats, were approved by the Rockefeller University Institutional Biosafety Committee (approved protocol no. 2-15-04-007).

Antibodies

Antibodies used for all the experiments reported herein, including their working dilutions, were obtained commercially, unless specified otherwise, and are detailed in an online repository (29). The specificity of these antibodies to their corresponding target proteins used for various experiments in this report had been earlier characterized in our laboratory and noted (with corresponding citations) as follows: Antibodies from Proteintech Group (Rosemont, IL) include: microtubule-affinity regulating kinase 4 [MARK4; immunoblot (IB), 1:500] (30) as earlier characterized (31). Antibodies from Santa Cruz Biotechnology (Dallas, TX) include: dynein cytoplasmic 1 heavy chain 1 [Dync1h1; immunofluorescence (IF), 1:50] (32) as earlier characterized (33), end-binding protein 1 (EB1; IF, 1:100) (34) as earlier characterized (35), myosin VIIa (IF, 1:100) (36) as earlier characterized (37), CAR (IF, 1:50) (38), nectin 3 (IF, 1:100) (39) as earlier characterized (40), β-actin (IB, 1:200) (41), goat IgG–horseradish peroxidase (HRP; IB:1:3000) (42), rabbit IgG-HRP (IB, 1:20,000) (43), and mouse IgG-HRP (IB, 1:3000) (44). Antibodies from Cell Signaling Technology (Danvers, MA) include: rpS6 (IB, 1:1000) (45), p-rpS6-S235/S236 (IB, 1:1000) (46), and p-rpS6-S240/S244 (IB, 1:1000) (47) as earlier characterized (25, 26, 48). Antibodies from Abcam (Cambridge, MA) include: formin 1 (IF, 1:100) (49) as earlier characterized (50, 51), α-tubulin (IB, 1:1000; IF, 1:100) (52) as earlier characterized (35, 53), and GAPDH (IB, 1:1000) (54). Antibodies from Invitrogen (Carlsbad, CA) include: Eps8 (IF, 1:100) (55), ZO-1 (IF, 1:100) (56), N-cadherin (IF, 1:100) (57), and β-catenin (IF, 1:100) (58), which were characterized as earlier reported (59, 60); rabbit IgG–Alexa Fluor 488 (IF, 1:250) (61), rabbit IgG–Alexa Fluor 555 (IF, 1:250) (57), mouse IgG–Alexa Fluor 488 (IF, 1:250) (62), and mouse IgG–Alexa Fluor 555 (IF, 1:250) (63). Antibodies from Sigma-Aldrich (St. Louis, MO) include: Arp3 (IF, 1:100) (64), which was characterized as earlier reported (65). Antibodies against laminin-γ3 (IF, 1:100) (66) was prepared in-house as earlier characterized (67).

Overexpression of the rpS6 quadruple phosphomimetic mutant and/or F5-peptide in adult rat testes in vivo

The quadruple phosphomimetic (i.e., constitutively active) mutant of rpS6 (p-rpS6-MT) was obtained by site-directed mutagenesis by converting Ser (S) 235, S236, S240, and S244 from the N terminus to Glu (E) 235, E236, E240, and E244 by PCR, and this mutant cDNA was cloned into the mammalian expression vector pCI-neo (Promega, Madison, WI) as detailed elsewhere (25, 26). Plasmid DNA was obtained by using Plasmid Plus Midi kits (Qiagen, Boston, MA), which also removed any endotoxin according to the protocol provided by the manufacturer. The biologically active F5-peptide used in this study was derived from domain IV of the laminin-γ3 chain at the apical ES of the rat testis, which was also cloned into the pCI-neo as detailed elsewhere (21), and overexpressed in the testis to assess its biological effects on testis function as described (22). Overexpression of either p-rpS6-MT, F5-peptide, or p-rpS6-MT plus F5-peptide (i.e., dual overexpression) was performed by intratesticular injection with 50 µL of transfection solution per testis, containing either 10 µg (for singular overexpression of either p-rpS6-MT or F5-peptide) or 20 µg (for dual overexpression of both p-rpS6-MT and F5-peptide using 10 µg each) of plasmid DNA, respectively (for both treatment groups and control group) together with 1.8 µL of Polyplus in vivo-jetPEI^® transfection reagent (PolyPlus-transfection, Illkirch-Graffenstaden, France) according to the manufacturer’s protocol as described (22). Control testes were transfected with the same amount of plasmid DNA (10 or 20 µg, but empty pCI-neo vector without a cDNA insert had no noticeable phenotypic effects in the seminiferous epithelium based on pilot experiments because this treatment only affected <8% of the tubules examined; see Fig. 1). This transfection solution (50 µL) was administered into each testis from the apical toward the basal end, using a 28-gauge 0.5-mL syringe with a 12.7-mm-long needle. When the syringe was withdrawn from the testis, plasmid DNA in the transfection medium loaded inside the syringe was gently released in the testis to avoid rapid hydrostatic pressure build-up. The first transfection was set as day 0, to be followed by two additional transfections on day 4 and day 8, with a total of three transfections in our regimen. On days 2, 6 and 10, rats were treated with or without adjudin at 10 mg/kg b.w. via oral gavage, and at this dose, earlier studies had shown adjudin to have no effects on the status of spermatogenesis, lacking any phenotypic changes in the seminiferous epithelium (27, 28). Rats were euthanized on day 21 (n = 10 rats for each treatment group of either p-rpS6-MT, F5-peptide, p-rpS6-MT plus F5-peptide, or p-rpS6-MT plus F5-peptide plus low-dose adjudin vs pCI-neo/Ctrl (empty vector) or low-dose adjudin only; from three independent experiments of n = 3 to 4 rats for each treatment vs control group to a total of n = 10 rats). Transfection efficiency was estimated to be ∼70% in parallel experiments wherein rats were transfected with a red fluorescence protein (DsRed2) cloned into the pCI-neo vector using the regimen shown in Fig. 1A as described (22) from n = 3 rats. Overexpression of either F5-peptide or rpS6 (and its corresponding phosphorylated/activated forms) was confirmed by RT-PCR or immunoblotting, respectively, using corresponding specific primer pairs (Table 1) or antibodies (29). This thus quantified the steady-state mRNA levels of F5-peptide (derived from domain IV of laminin-γ3 chain) or the steady-state protein level of rpS6 and its phosphorylated forms in lysates of testes in treatment and control groups.

Figure 1.

Combined overexpression of p-rpS6-MT and F5-peptide in adult rat testes considerably promotes the effects of adjudin to induce germ cell exfoliation from the epithelium. (A) Regimen used for the study reported herein. (B) Defects in spermatogenesis in each treatment vs control groups were scored using n = 4 rats for histological analysis as noted in (C) using the criteria outlined in Materials and Methods (top panel). In the bottom panel, the thinning of the epithelium [see (C)] was used as the parameter to define tubule damage, which also reflected the severity of the tubule damage. At least 100 seminiferous tubules were randomly scored using cross-sections from each rat for a total of 400 tubule sections. Each bar is mean ± SD of n = 4 rats. *P < 0.05, **P < 0.01, by ANOVA. (C) Histological analysis using cross-sections of testes. Boxed areas in the first, second, and third columns were magnified and are shown in the second, third, and fourth columns, respectively. Testes of control (pCI-neo/Ctrl, empty vector) and low-dose adjudin (10 mg/kg b.w.) had no notable defects in the status of spermatogenesis. Overexpression of p-rpS6-MT, F5-peptide, or p-rpS6-MT plus F5-peptide in the testis without adjudin induced considerable defects in the testis. These included: (i) the appearance of multinucleated spermatids (green arrowheads); (ii) failure in phagosome transport, which should have been distributed close to the base of the tubule for lysosomal degradation (68), but instead, phagosomes were found near the tubule lumen (white arrowhead) or in the lumen (blue arrowhead); (iii) considerable thinning of the epithelium, which was more remarkable in the group overexpressed with p-rpS6 plus F5-peptide and also treated with low-dose adjudin, and in this last treatment group, virtually no advanced germ cells were detected except early spermatogonia (red arrowhead), differentiated spermatogonia (pink arrowhead), and Sertoli cells (yellow arrowhead). Data shown herein are representative of an experiment of n = 4 rats that yielded similar results. Scale bars, 250 µm, 100 µm, 80 µm, and 40 µm in the first, second, third, and fourth columns (or inset), respectively, which apply to corresponding micrographs.

Table 1.

Primer Pairs Used for RT-PCR

Gene	Primer Sequence	Position	Length, bp	Temp., °C	No. of Cycles	GenBank Accession No.
F5-peptide (from laminin-γ3 chain)	Sense: 5′-CCTCCAATCTACCCAGCTC-3′	1790–1808	162	54	29	NM_001107830
	Antisense: 5′-GGCCACTGGTCCAGATGCT-3′	1933–1951
S16	Sense: 5′-TCCGCTGCAGTCCGTTCAAGTCTT-3′	15–38	385	54	29	XM_341815
	Antisense: 5′-GCCAAACTTCTTGGTTTCGCAGCG-3′	376–399

Lysate preparation and IB analysis

IB analysis was performed as described (69, 70) with testis lysates. In brief, testis lysates were obtained via sonication using IP lysis buffer [50 mM Tris, 0.15 M NaCl, 1% Nonidet P-40 (v/v), 2 mM EGTA, and 10% glycerol (v/v), pH 7.4 at 22°C] freshly supplemented with protease and phosphatase inhibitors, which included 1 mM 4-(2-aminoethyl)benzenesulfonyl fluoride hydrochloride, 1 mM sodium orthovanadate, 0.05 mM bestatin, 0.05 mM sodium EDTA, 15 μM E64, 1 mM pepstatin, 4 mM sodium tartrate dehydrate, 5 mM NaF, and 3 mM β-glycerophosphate disodium salt (22, 24). Protein concentration was determined with DC protein assay kits (Bio-Rad Laboratories, Hercules, CA) using BSA as a standard for calibration. Equal amounts of lysate (40 µg of protein) from each sample, including treatment and control groups, were used for SDS-PAGE and IB analysis with the corresponding primary and secondary antibodies (29). Targeted protein bands were detected by enhanced chemiluminescence using in-house–made kits (71) with the GE ImageQuant LAS 4000 mini luminescent image analyzer and the ImageQuant (version 1.3) software package. Relative intensities of protein bands were quantified by ImageJ 1.45s software (National Institutes of Health, Bethesda, MD). For IB analysis, all samples within an experimental vs control group were processed simultaneously, thus avoiding interexperimental variations. GAPDH served as the protein loading control. Each sample had triplicates of n = 4 rats for analysis.

RNA extraction and RT-PCR

Total RNA was extracted from testes in control and treatment groups with Trizol reagent (Life Technologies) for reverse transcription to obtain total cDNAs which were then used for PCR using primer pairs specific to F5-peptide and S16 (served as a PCR control) as described (72). PCR products were verified by direct DNA sequencing at Genewiz (South Plainfield, NJ). RT-PCR was performed with n=3 experiments which yielded similar results, and representative findings were shown herein.

Histological analysis

Histological analysis was performed as described (53). In brief, testes were fixed in modified Davidson’s fixative (73) and paraffin-embedded, and cross-sections of testes were obtained with a microtome (∼5 µm in thickness), then followed by staining in hematoxylin and eosin. Images of testis sections were acquired using an Olympus BX61 microscope with a built-in Olympus DP-71 digital camera and processed with the Olympus MicroSuite Five software package (version 1224; Olympus, Tokyo, Japan) without any further manipulations. All histological data reported herein are representative findings from experiments of n = 4 rats, including treatment and control groups, which yielded similar results. To avoid interexperimental variations, all treatment and control groups in an experimental set were processed simultaneously.

Fluorescence microscopy and dual-labeled IF analysis

Fluorescence microscopy and dual-labeled IF were performed as described (53, 69, 74). In brief, frozen cross-sections (7 µm thick) of testes were obtained in a cryostat at −22°C. Sections were then fixed in 4% paraformaldehyde, permeabilized with 0.1% Triton X-100, and nonspecific binding sites were blocked in 1% BSA (w/v) in PBS [10 mM sodium phosphate, 0.15 M NaCl (pH 7.4) at 22°C] and then incubated overnight with the corresponding primary antibodies (29) at room temperature. To confirm the staining noted for different target proteins based on IF analysis, negative controls were included and performed as earlier described (75) wherein the primary antibody was substituted with the corresponding IgG of the same animal species, such as rabbit, mouse, or goat, and at the same concentration as for the primary antibody, all of which are shown in an online repository (29). Note that negative controls were included in the same experimental sessions in parallel with the use of the corresponding primary antibody and on the same cross-sections of testes as earlier described (75, 76) to confirm staining specificity, and according to the guidelines of immunohistochemical analysis (77). Thereafter, sections were incubated with Alexa Fluor 555 (red fluorescence)– or Alexa Fluor 488 (green fluorescence)–conjugated secondary antibodies for 1 hour at room temperature (53, 69, 74). F-actin was visualized using a rhodamine/phalloidin conjugate (red fluorescence) at a 1:200 dilution (53). Tissue sections stained with rhodamine B (i.e., without phalloidin) (Sigma-Aldrich) served as a negative control. Cell nuclei were stained with 4′,6-diamidino-2-phenylindole. Fluorescence images were captured with a Nikon Eclipse 90i fluorescence microscope equipped with a Nikon DS-Qi1Mc digital camera. Images were subsequently analyzed using Adobe Photoshop for image overlays to assess protein–protein colocalization. Data shown herein are representative findings from n = 4 rats from different experiments that yielded similar results. All samples within an experimental set, including both treatment and control groups, were simultaneously process to eliminate interexperimental variations. In selected experiments, we also assessed changes in the distribution of F-actin (or a target protein, such as N-cadherin, ZO-1) near the basal ES/BTB by measuring its distribution at the site. At least 80 randomly selected tubules from cross-sections of testes were scored from n = 4 rats for comparison between treatment and control groups.

BTB integrity assay

A functional in vivo assay as detailed elsewhere (22) was performed to assess the BTB integrity in treatment vs the control groups. In brief, this assay assessed the ability of a functional BTB found in control testes to block the entry of a small membrane-impermeable biotin reagent, EZ-Link Sulfo-NHS-LC-Biotin (relative molecular mass, 556.58; Thermo Fisher Scientific), into the adluminal compartment of the epithelium to biotinylate any primary amines in proteins behind the BTB. Positive control was rats received cadmium chloride (CdCl₂; 3 mg/kg b.w., via intraperitoneal IP administration) for 3 days because CdCl₂ was earlier shown to disrupt BTB integrity irreversibly (78, 79). For this study, the regimen shown in Fig. 1A was used except that the BTB integrity assay was performed on day 12 with n = 4 rats for each group including the treatment, as well as the negative and positive control groups. Rats were under anesthesia following administration of ketamine hydrochloride (60 mg/kg b.w., IM) with xylazine (10 mg/kg b.w., IM). Thereafter, 100 µL of 10 mg/mL EZ-Link Sulfo-NHS-LC-Biotin dissolved in PBS was loaded into the testis under the tunica albuginea using a 28-gauge needle to allow diffusion of the biotin reagent across all of the seminiferous tubules in the testis in 30 minutes. Rats were then euthanized by CO₂ asphyxiation. Testes were immediately removed, snap-frozen in liquid nitrogen, and frozen cross-sections (7 µm thickness) were obtained. Sections were then fixed in 4% paraformaldehyde for 10 minutes and incubated with Alexa Fluor 555–streptavidin (red fluorescence), which had high affinity for biotin for 1 hour at room temperature. Cell nuclei were stained using 4′,6-diamidino-2-phenylindole. Fluorescence images were captured using the Nikon 90i fluorescence microscope. Semiquantitative analysis was performed that compared changes in BTB integrity between treatment and various control groups. Distance of the red fluorescence (biotin) that diffused from the base (i.e., basement membrane) of a seminiferous tubule at the site of the BTB vs radius of the tubule was obtained from at least 100 randomly selected tubules from a rat with a total of n = 4 different rat testes (i.e., 400 randomly selected tubules) from each treatment and control group. For an oval-shaped tubule section, because it was obliquely cut, the tubule radius was obtained using the average of the longest and the shortest radii. This thus semiquantitatively assessed the extent of BTB leakage.

Assessment of spermatogenesis defects

Defects in spermatogenesis were assessed based on histology data using cross-sections of paraffin-embedded (and modified Davidson’s fixative–treated) testes. We used the following criteria to determine whether a tubule was scored as a tubule with defects in spermatogenesis: (i) germ cell exfoliation in which >15 elongating/elongated/round spermatids and/or spermatocytes were detected in the lumen of a tubule; (ii) thinning of the epithelium of a tubule wherein <50% fewer germ cell layers were noted vs similarly staged tubules in control groups, and this also reflected a considerable reduction in tubule diameter because of germ cell loss; (iii) appearance of multinucleated round spermatids due to failure of microtubule (MT) and/or F-actin dynamics—to be phagocytosed by Sertoli cells to form phagosomes, undergoing lysosomal degradation; (iv) defects in spermatid polarity in the seminiferous epithelium wherein >10 elongating/elongated spermatids were found to have their heads pointed at least 90° to 180° away from the intended orientation of aligning perpendicularly to the basement membrane, as noted in control testes; and (v) defects in spermatid transport wherein >10 step 19 spermatids were found to be trapped deep inside the epithelium, some even close to the basement membrane, in late stage VIII, IX, X, and XI tubules when spermiation had already taken place. At least 100 tubules were randomly scored from n = 5 rats from both treatment and different control groups. The phenotypes reported herein were found in >95% of the tubules examined because transfections for overexpression and adjudin treatment were performed thrice.

Statistical analysis

Each data point was a mean ± SD of n = 4 to 8 rats from two to three independent experiments. Statistical significance was determined by two-way ANOVA using the repeated measures model followed by a Dunnett test with the GB-STAT statistical analysis software package (version 7.0; Dynamic Microsystems, Silver Spring, MD). This thus compared changes between treatment groups and their corresponding controls by assessing within-experiment effects, which were the focus of the present analysis. A P value <0.05 was considered to be statistically significant.

Results

p-rpS6-MT and F5-peptide, which enhance BTB permeability, promote adjudin-mediated defects in spermatogenesis through modification of BTB transport function

Using the regimen summarized in Fig. 1A, adult rats in groups of n = 3 to 4 animals per treatment vs control groups in three separate experiments (i.e., a total of at least 10 rats per group) were used for studies. Transfection of testes with the corresponding plasmid DNA (10 µg of DNA per testis) was performed on days 0, 4, and 8 (thrice) and rats were also treated with adjudin at a low dose on days 2, 6, and 10 (also thrice), which by itself was ineffective in inducing any notable phenotypes, including germ cell loss, disruption of the apical ES and/or basal ES/BTB, or infertility as earlier reported (27, 28). Rats were then euthanized on day 21, that is, 3 weeks after the initial transfection, and testes were removed and processed for histological analysis. The percentage of abnormal tubules in different treatment groups vs the control and low-dose adjudin groups are shown in Fig. 1B (top panel), using the criteria depicted in Fig. 1C. In the bottom panel of Fig. 1B, the parameter that was used to score abnormal tubules was thinning of the seminiferous epithelium in treatment group vs control testes by quantifying the width of the epithelium, which is also noted in Fig. 1C. About 100 tubules were randomly selected for examination from each rat testis (and n = 4 rats were examined, totaling ∼400 tubules) and are summarized in both panels in Fig. 1B. The low percentage of defective tubules noted in the control group (pCI-neo/Ctrl, i.e., empty vector), ∼18%, following triple transfection was consistent with the results of earlier reports (22, 73) using Polyplus in vivo-jetPEI^® transfection medium for overexpression. However, combined transfections of testes with pCI-neo/rpS6-MT and pCI-neo/F5-peptide with adjudin (by oral gavage) were highly effective to induce defects in tubules (Fig. 1B) that led to aspermatogenesis as noted in Fig. 1C. As shown in Fig. 1C, most tubules (∼85%) had normal spermatogenesis in the epithelium from testes of control (pCI-neo/Ctrl)- and low-dose adjudin (10 mg/kg b.w.)–treated groups. However, overexpression of either rpS6-MT, F5-peptide, or combined overexpression of rpS6-MT and F5-peptide induced considerable seminiferous epithelial damage, mostly notably in tubules devoid of germ cells (Fig. 1C). Interestingly, combined overexpression of p-rpS6-MT and F5-peptide coupled with adjudin at a low dose caused more remarkable seminiferous epithelial damage wherein tubules were devoid of virtually all germ cells except Sertoli cells (yellow arrowhead) and possibly undifferentiated spermatogonia (red arrowhead) and some early spermatogonia (blue arrowhead). In short, defects of spermatogenesis that were noted included: (i) loss of elongated spermatids in prestage VIII tubules; (ii) loss of round spermatids and spermatocytes that were found in tubule lumen; (iii) appearance of multinucleated round spermatids (green arrowhead); (iv) defects in phagosome transport when phagosomes should have been transported to the base of the epithelium for lysosomal degradation, as noted in the third row (i.e., pCI-neo/p-rpS6-MT) in the last two panels (see the green boxed image, wherein the phagosomes were enlarged in the red boxed image and are indicated with white arrowheads); (v) defects in spermatid polarity when spermatids no longer pointed toward the basement membrane but deviated by 90° to 180° from the intended orientation; and (vi) thinning of the seminiferous epithelium due to germ cell loss. Because some of these defects in spermatogenesis were overlapping such that differences between the p-rpS6-MT plus F5-peptide group vs p-rpS6-MT plus F5-peptide plus adjudin group might be an underestimate. Thus, when the thinning of the epithelium was used as the only parameter to assess spermatogenesis defects, the combined use of p-rpS6-MT, F5-peptide, and adjudin at a low dose was found to be most effective to induce defects in the tubules when compared with any other treatment groups (Fig. 1B, bottom panel).

Overexpression of p-rpS6-MT and F5-peptide in the testis modifies BTB function in vivo, facilitating adjudin transport across the BTB

We next confirmed overexpression of F5-peptide by RT-PCR using primer pairs specific to F5-peptide derived from domain VI of laminin-γ3 chain (Table 1; Fig. 2A, left panel). Overexpression of rpS6 and its two phosphorylated forms was also confirmed by IB using the corresponding specific antibodies (29) (Fig. 2A, right panel). The regimen shown in Fig. 1A that was used to assess the effects of overexpression of either p-rpS6-MT, F5-peptide, or both biomolecules (with and without low-dose adjudin) on spermatogenesis was also used to assess their effects on BTB integrity based on a functional assay, except that rats were euthanized on day 12 instead of day 21 to make sure that the BTB had not been “resealed.” As noted in testes of control rats transfected with pCI-neo empty vector (pCI-neo/Ctrl) or treated with adjudin at a low dose, which had no effects on the status of spermatogenesis (Fig. 1C), the functional BTBs in these rats were able to exclude biotin from entering the adluminal compartment so that biotin (red fluorescence) was kept at the base of the epithelium (see white brackets in the enlarged images in the third column of each panel) near the basement membrane (annotated by a dashed white line) (Fig. 2B). However, biotin penetrated all the way into the epithelium in rats treated with CdCl₂ at 3 mg/kg b.w. (for 5 days) as noted in the positive control group (Fig. 2B), which rendered the barrier leaky (78, 79) (see yellow bracket). However, overexpression of either p-rpS6-MT or F5-peptide also made the barrier leaky, and the combined use of both biomolecules was more effective to make the barrier leaky based on the extent of diffusion of biotin in the adluminal compartment (Fig. 2B) (see yellow vs white brackets). Although adjudin at a low dose per se had no effects on BTB integrity, its use with both p-rpS6-MT and F5-peptide was more potent to make the BTB leaky (Fig. 2B). The findings noted in Fig. 2B are also consistent with the semiquantitative data shown in Fig. 2C, because the combined use of overexpression of both biomolecules and adjudin at a low dose indeed was more effective to make the BTB leaky vs either biomolecule or adjudin alone.

Figure 2.

Overexpression of p-rpS6-MT and F5-peptide together with adjudin is capable of perturbing the BTB function, considerably more severely than p-rpS6-MT, F5-peptide, or p-rpS6-MT plus F5-peptide. (A) Samples were obtained using the regimen shown in Fig. 1A on day 21 for RT-PCR (left panel) and immunoblot analysis (right panel) to confirm overexpression of either F5-peptide or rpS6 (total) vs the two phosphorylated/activated forms of rpS6. S16 and GAPDH served as the corresponding PCR and protein loading control. Uncropped DNA and IB gel blots are shown in an online repository (29). (B) The regimen shown in Fig. 1A was used for this experiment except that rats were used on day 12 for the BTB integrity assay instead of on day 21 to avoid the perturbed BTB being “resealed.” In testes of control rats (pCI-neo/Ctrl, empty vector) or adjudin alone at a low dose (10 mg/kg b.w., by oral gavage), the functional BTB was capable of limiting the biotin (red fluorescence) from entering the adluminal compartment, restricting the biotin at the base of the tubule at the basement membrane (annotated by dashed white line), as noted by white brackets (see the magnified image boxed in green). However, overexpression of either p-rpS6-MT or F5-peptide in the testis was able to perturb the BTB function, but combined overexpression of p-rpS6-MT and F5-peptide appeared to be more effective in perturbing the barrier function. However, combined overexpression of p-rpS6-MT and F5-peptide coupled with adjudin at a low dose (which by itself had no detectable effect) was able to perturb the barrier function considerably better, as noted herein. Scale bars, 400 µm, 100 µm, and 60 µm in the low-, median-, and high-magnification images. (C) Semiquantitative data by comparing the extent of various treatment groups that perturbed the Sertoli cell BTB function vs the control group using the ratio on the distance traveled by biotin behind the BTB near the basement membrane (annotated by dashed white line) (i.e., D_Biotin)/radius of the corresponding seminiferous tubule (i.e., ST_Radius). At least 50 tubules were randomly selected in all groups from each experiment with n = 4 different rat testes. For obliquely sectioned tubules, the average of the longest and shortest radii was used as the radius. Each bar is a mean ± SD of n = 4 testes. *P < 0.05, **P < 0.01, by ANOVA. DAPI, 4’,6-diamidino-2-phenylindole.

Overexpression of p-rpS6-MT and F5-peptide in the testis, which modifies BTB function, promotes adjudin transport, intensifying defects in actin cytoskeletal organization

Changes in F-actin organization

In testes of rats transfected with pCI-neo empty vector (pCI-neo/Ctrl, i.e., control testes) or treated with adjudin at a low dose, F-actin was orderly aligned across the seminiferous epithelium, as noted by its conspicuous presence at the apical and basal ES. F-actin appeared as a bulb-like structure at the concave (ventral) side of spermatid heads in stage VII to early VIII tubules at the apical ES in both control and adjudin low-dose groups (Fig. 3A). At the basal ES (annotated by white brackets in the third column) near the basement membrane (annotated by a dashed white line), it coexisted with TJ, gap junction, and desmosome to crease the BTB (which collectively are referred to as basal ES/BTB), F-actin was tightly localized at the site to support the BTB function. However, following overexpression of rpS6-MT, F5-peptide, or rpS6-MT plus F5-peptide, F-actin at the apical ES was no longer expressed orderly, as noted in control testes (or adjudin low-dose group). Instead, F-actin was grossly mislocalized, possibly because of the apical ES degeneration that led to spermatid loss from the epithelium (see magnified images in the last column in Fig. 3A). Also, F-actin was no longer tightly packed at the basal ES/BTB, as noted in control testes (white bracket); instead, it was loosely organized (see yellow brackets) in treatment groups. F-actin disorganization at the basal ES/BTB was considerably more severe in the rpS6-MT plus F5-peptide plus adjudin treatment group vs other groups as noted in the bar graphs when disruptive changes in the organization of F-actin were used as the parameter to score abnormal (i.e., damaged) tubules (Fig. 3B, see right panel). The visualization of F-actin was specific because no staining was noted in the negative control (Fig. 3A).

Figure 3.

Changes in the organization of F-actin across the epithelium following overexpression of p-rpS6-MT, F5-peptide, or their combined use with or without low-dose adjudin. (A) In testes of control (pCI-neo/Ctrl) and low-dose adjudin groups, F-actin was robustly expressed and orderly aligned in the seminiferous epithelium, such as at the actin-rich anchoring junctions apical ES and basal ES, wherein actin appeared as bulb-like structures on the concave (ventral) side of spermatid heads and also at the basal ES (annotated by white bracket) near the basement membrane (annotated by dashed white line). In testes overexpressed with p-rpS6-MT, F5-peptide, or p-rpS6-MT plus F5-peptide, F-actin was grossly disorganized at the apical ES because most of the elongating/elongated spermatids had undergone spermiation in nonstage VIII tubules, and F-actin at the basal ES/BTB was also grossly misdistributed, no longer tightly localized at the basal ES/BTB (white bracket), as noted in the control and low-dose adjudin groups, but diffusely localized (yellow bracket) in the treatment groups. However, in the testes overexpressed with p-rpS6-MT plus F5-peptide and treated with low-dose adjudin (which by itself had no effects), there was a considerable loss of F-actin in the epithelium except for a thin layer of F-actin near the basement membrane. Tissue sections stained with rhodamine B (i.e., without phalloidin) (Sigma-Aldrich) served as a negative control (see last panel). Micrographs are representative of n = 3 independent experiments (different rat testes) that yielded similar results. Scale bars, 100 µm, 80 µm, 40 µm, and 40 µm in the first (also negative control) and third columns and the green and yellow insets, respectively, enlarged from the corresponding sites of a lower magnified micrograph. These scale bars apply to corresponding micrographs in the same panels. (B) Bar graph illustrating the percentage of abnormal tubules that were scored based on disruptive changes on the organization of F-actin as noted in (A), which was used as the only parameter to identify damaged tubules. Each bar is a mean ± SD of n = 4 rats, and at least 100 tubules were randomly scored from each rat. *P < 0.05, **P < 0.01, by ANOVA. DAPI, 4’,6-diamidino-2-phenylindole.

Changes in spatial expression of actin-regulatory proteins

These changes in F-actin organization across the seminiferous epithelium in the treatment vs control groups were associated with changes in the spatial expression of the actin-regulatory proteins. These included changes in Eps8 [an actin barbed end capping and bundling protein that confers actin filaments the bundled configuration at the ES (59)] and Arp3 [a branched actin polymerization protein that induces barbed end nucleation of an existing linear actin microfilament, thereby causing actin filament branching, which in turn destabilizes the ES function (65)] (Fig. 4). For instance, Eps8 and Arp3 that appeared as bulb-like structures at the concave side of spermatid heads to support apical ES as noted in testes of control rats or rats treated with adjudin alone (at a low dose) were considerably mislocalized in each treatment group and grossly disorganized (Fig. 4). Also, these two proteins also diffusely localized at the basal ES/BTB, thereby perturbing proper distribution of F-actin at the site, no longer tightly organized to support the BTB (see white brackets in the second columns) but diffusely organized and localized instead (see yellow brackets in the second columns) (Fig. 4). Additionally, distribution of a linear actin filament, nucleation protein formin 1 [to generate linear actin filaments to be assembled into actin bundles at the ES (51)], and an actin-based cargo protein, myosin VIIa, known to direct cargoes to the plus/barbed end (i.e., rapidly growing end) of F-actin (80), were also affected in treatment vs control groups (Fig. 5). These changes not only affected F-actin organization, but they also impeded numerous cellular events, such as intracellular endosome-mediated protein trafficking events, to support spermatogenesis.

Figure 4.

Changes in F-actin organization across the epithelium following overexpression of p-rpS6-MT, F5-peptide, or their combined use with or without low-dose adjudin are associated with changes in spatial expression of actin regulatory proteins Eps8 and Arp3. In testes of the control (pCI-neo/Ctrl) or low-dose adjudin groups, both Eps8 and Arp3 robustly expressed at the concave side of step 19 spermatid heads at the apical ES in stage VII to early stage VIII tubules as bulb-like structures, colocalizing with F-actin, as noted herein. At the basal ES, Eps8 and Arp3 also localized at the basal ES near the basement membrane (annotated by dashed white line), supporting F-actin that aligned tightly at the BTB (see white brackets) in these two groups. However, in testes overexpressed with p-rpS6-MT, F5-peptide, or p-rpS6-MT and F5-peptide, there were considerable changes in the expression of Eps8 and Arp3, such that either actin regulatory protein was no longer robustly expressed at the apical or basal ES. Owing to spermatid exfoliation, Eps8 and Arp3 were no longer orderly expressed near the tubule lumen but were diffusely localized, perhaps due to the absence of the apical ES. Although basal ES remained, Eps8 and Arp3 also were no longer robustly and orderly expressed at the basal ES/BTB but were considerably diminished (see yellow brackets). These changes thus led to disruptive expression/localization of F-actin across the epithelium, as shown herein. In testes overexpression with p-rpS6-MT and F5-peptide coupled with low-dose adjudin, Eps8 and Arp3 were virtually undetectable at the apical and basal ES, and F-actin at the basal ES was considerably diminished, consistent with data shown in Fig. 3. Negative controls for Eps8/F-actin and Arp3/F-actin are shown in an online repository (29) using corresponding normal IgG of the same animal species and rhodamine B (without phalloidin) for staining. Micrographs are representative of a single experiment derived from n = 3 independent experiments that yielded similar results. Scale bars, 80 µm in the first column and 40 µm in the inset in the fourth column, which apply to corresponding micrographs in the same panel. DAPI, 4’,6-diamidino-2-phenylindole.

Figure 5.

Changes in the spatial expression of formin 1 and myosin VIIa impede F-actin organization following overexpression of p-rpS6-MT, F5-peptide, or their combined use with or without low-dose adjudin. In testes of both control (pCI-neo/Ctrl) and low-dose adjudin (10 mg/kg b.w., via oral gavage) groups, formin 1 and myosin VIIa were expressed robustly at the concave side of step 19 spermatid heads at the apical ES, and also basal ES/BTB near the basement membrane (annotated by dashed white line), consistent with earlier reports (50, 51, 81), colocalizing with F-actin at the site. However, in testes of rats following overexpression of p-rpS6-MT, F5-peptide, or p-rpS6-MT plus F5-peptide, both formin 1 and myosin VIIa became grossly misdistributed across the seminiferous epithelium, as apical ES was no longer recognizable because there were no elongating/elongated spermatids in ∼80% to 90% of the tubules. In testes of rats overexpressed with p-rpS6-MT plus F5-peptide with low-dose adjudin, the expression of both formin 1 and myosin VIIa was considerably diminished and they were difficult to find at the apical ES, and the expression at the basal ES was also remarkably diminished. Negative controls for formin 1/F-actin and myosin VIIa/F-actin are shown in an online repository (29) using corresponding normal IgG of the same animal species and rhodamine B (without phalloidin) for staining. Micrographs are representative of a single experiment, which were derived from n = 3 independent experiments using n = 3 rat testes and yielded similar results. Scale bars, 80 µm in the first column and 40 µm in inset magnified from the corresponding boxed area in the third column, which apply to corresponding micrographs in the same panel. DAPI, 4’,6-diamidino-2-phenylindole.

Changes in apical and basal ES and TJ protein distribution

Changes in F-actin distribution across the epithelium considerably impeded distribution of apical ES proteins laminin-γ3 chain (67, 82) and nectin 3 (83) (Fig. 6A), the basal ES adhesion protein complex N-cadherin/β-catenin (85, 86), and also the TJ adhesion protein complex CAR/ZO-1 (87, 88) (Fig. 6B) since these proteins all used F-actin for their attachment. For instance, laminin-γ3 and nectin 3 (both adhesion proteins that are specifically expressed by elongated spermatids at the apical ES) were no longer restrictively expressed at the convex (dorsal) side and the tip of spermatid heads, as noted in control (pCI-neo/Ctrl, group 1) and low-dose adjudin (group 2) groups (Fig. 6A) following overexpression of either p-rpS6-MT (group 3), F5-peptide (group 4), or p-rpS6-MT with F5-peptide without adjudin (group 5). Either one of these treatments perturbed their distribution, but the effects of overexpression of p-rpS6-MT plus F5-peptide was more remarkable, wherein the expression of either laminin-γ3 or nectin 3 was either diminished or their distribution was affected (Fig. 6A). For example, these proteins were no longer robustly expressed on the convex side of spermatids but diffused away, and virtually no expression was detected at the tip of spermatid heads in groups 3, 4, or 5 (Fig. 6A). Virtually no expression of either laminin-γ3 or nectin 3 was found in group 6 when low-dose adjudin was used in conjunction with overexpression of p-rpS6-MT and F5-peptide (Fig. 6A). Additionally, many elongated spermatids with defects in polarity that orientated their heads by turning 90° to 180° away from the basement membrane instead of aligning perpendicularly to the basement membrane were noted (see yellow arrowheads in Fig. 6A). For the basal ES proteins N-cadherin and β-catenin and TJ proteins ZO-1 and CAR, they were no longer tightly associated with the BTB near the basement membrane (annotated by a dashed white line), as noted in the testes of control and low-dose adjudin groups (Fig. 6B). Owing to defects in F-actin organization (Fig. 3), these proteins diffused away from the BTB, as noted in yellow brackets in treatment groups vs control and adjudin low-dose groups (white brackets) (Fig. 6B). The bar graphs in Fig. 6C provide the semiquantitative data of Fig. 6B, illustrating that the combined use of p-rpS6-MT and F5-peptide in the presence of low-dose adjudin was effective to modify the BTB function by inducing redistribution of both basal ES and TJ proteins at the site.

Figure 6.

Changes in the spatial expression and distribution of apical and basal ES/BTB proteins in the seminiferous epithelium following overexpression of p-rpS6-MT, F5-peptide, or their combined used with or without low-dose adjudin. (A) Using the regimen shown in Fig. 1A, in testes of control (pCI-neo/Ctrl) (group 1) and low-dose adjudin (group 2) groups, laminin-γ3 and nectin 3 were localized predominantly on the convex (dorsal) side and the tip of spermatid heads, similar to earlier reports (67, 84). However, considerable changes in their distribution were noted following overexpression of p-rpS6-MT (group 3), F5-peptide (group 4), or p-rpS6-MT plus F5-peptide (group 5) wherein their expressions were considerably diminished, no longer robustly expressed on the convex side of spermatid heads but moved away from the site, and virtually no expression was detected on the spermatid tip (see groups 3 to 5 vs groups 1 and 2). In group 6, where low-dose adjudin was administered together with overexpression of p-rpS6-MT and F5-peptide, virtually no laminin-γ3 and nectin 3 were associated with any remaining elongating spermatids. Furthermore, many elongated spermatids had gross defects in their polarity because their heads were pointed 90° to 180° away from the intended orientation (yellow arrowheads) of pointing toward the basement membrane, as noted in control testes. Scale bars, 80 µm and 40 µm in first and third columns, respectively, which apply to other micrographs. Negative controls for laminin-γ3 and nectin 3 using the corresponding normal IgG of the same animal species are shown in an online repository (29). (B) In testes of control and low-dose adjudin groups, basal ES (N-cadherin/β-catenin) and TJ (CAR/ZO-1) adhesion protein complexes were tightly associated with the BTB (see white brackets) near the basement membrane (annotated by dashed white line), but these proteins became diffusely localized at the site (see yellow brackets) owing to the defective F-actin organization at the BTB (see Fig. 3) following overexpression of p-rpS6-MT, F5-peptide, or both without adjudin. However, in the presence of low-dose adjudin plus p-rpS6-MT and F5-peptide, the distribution of either adhesion protein complex was considerably farther away from the basement membrane (annotated by dashed white line). Scale bars, 80 µm, which apply to corresponding micrographs in the same panel. Negative controls for N-cadherin/β-catenin and CAR/ZO-1 using the corresponding normal IgG of the same animal species are shown in an online repository (29). (C) Changes in the distribution of the basal ES adhesion protein complex N-cadherin/β-catenin and the TJ adhesion protein complex CAR/ZO-1 between the treatment (yellow brackets) and control (white brackets) groups. Each bar is a mean ± SD of n = 4 rats in which ∼100 cross-sections of tubules in each rat were randomly quantified for a total of 400 tubules. *P < 0.05, **P < 0.01, by ANOVA. DAPI, 4’,6-diamidino-2-phenylindole.

Overexpression of p-rpS6-MT and F5-peptide in the testis in vivo, which modifies BTB function, promotes adjudin-induced defects in MT cytoskeletal organization

Changes in MT organization

In this experiment, MTs across the seminiferous epithelium were visualized by staining α-tubulin (red fluorescence) (Fig. 7A). Note that α-tubulin together with β-tubulin, which create the α-tubulin/β-tubulin heterodimers, are the building blocks of MTs (89–91). In testes of control (pCI-neo/Ctrl, empty vector) rats or low-dose adjudin–treated rats, MTs appeared as tracks that stretched across the epithelium that lay perpendicular to the basement membrane (see dashed white line, encircling the base of the seminiferous tubule). However, overexpression of either p-rpS6-MT, F5-peptide, or both p-rpS6-MT and F5-peptide perturbed MT organization considerably wherein these tracks were almost collapsed, and in some cases, MTs encircled the multinucleated cells composed of round spermatids that were then phagocytosed by Sertoli cells to form phagosomes to be degraded (Fig. 7A; see also Fig. 1). However, in rats treated with low-dose adjudin (which by itself had no notable effect) and coupled with overexpression of p-rpS6-MT and F5-peptide, MTs were collapsed and compacted to the base of the seminiferous tubule (Fig. 7A). When the distribution of MTs across the epithelium was semiquantified, as summarized in the bar graph in Fig. 7B, the p-rS6-MT/F5-peptide/low-dose adjudin group was found to be most effective to perturb MT organization. The bar graph in Fig. 7B summarizes results noted in Fig. 7A when disruptive changes in the organization of MTs were used as the parameter to identify abnormal (i.e., damaged) tubules. These findings thus support the notion that p-rpS6-MT and F5-peptide that were used to modify BTB also promoted transport function at the site to enhance the entry of adjudin into the testis to perturb spermatogenesis.

Figure 7.

Changes in MT organization across the epithelium following overexpression of p-rpS6-MT, F5-peptide, or their combined use with or without low-dose adjudin. (A) In testes of the control (pCI-neo/Ctrl) or low-dose adjudin (10 mg/kg b.w.) groups, MTs, visualized by α-tubulin (red fluorescence) staining, appeared as tracks that lay across the epithelium orderly and aligned perpendicular to the basement membrane (annotated by dashed white line) to support spermatogenesis. MTs were also detected at the basal ES, tightly organized to support the BTB (annotated by white brackets). Following overexpression of p-rpS6-MT, F5-peptide, or p-rpS6-MT+F5 peptide, there was gross disorganization of MT-based tracks across the epithelium. For instance, linear tracks that lay across the epithelium and aligned perpendicular to the basement membrane were no longer detected; these tracks were truncated, sometimes wrapped around multinucleated round spermatids, which would become phagosomes to undergo lysosomal degradation by Sertoli cells. Interestingly, MT-based tracks were remarkably collapsed in testes overexpressed with p-rpS6-MT plus F5-peptide and treated with low-dose adjudin compared with all other groups. MTs also were no longer tightly expressed at the basal ES/BTB but were diffusely localized (see yellow brackets) in the treatment groups. The negative control (-ve Ctrl) is shown in the bottom panel of cross-sections of normal rat testes wherein the primary α-tubulin antibody (29) was substituted with normal mouse IgG, which yielded no notable staining. Micrographs are representative of a single experiment from n = 3 independent experiments that yielded similar results. Scale bars, 100 µm in the first column, 60 µm in the third column, and 40 µm in insets in either green or yellow, which apply to corresponding micrographs in the same panel. Scale bar, 150 µm in the -ve Ctrl panel. (B) Bar graph illustrates the percentage of abnormal tubules in the treatment vs control groups by randomly scoring 100 cross-sections of tubules per rat of n = 4 rats for a total of 400 tubules, wherein defects in MT organization as noted in (A) served as the parameter for determining tubule damage. *P < 0.05, **P < 0.01, by ANOVA. DAPI, 4’,6-diamidino-2-phenylindole.

Changes in spatial expression of MT regulatory proteins

Changes in EB1 and MARK4 distribution

We next investigated whether changes in MT organization noted in Fig. 7 were associated with changes in spatial expression of MT-regulatory proteins. We elected to examine changes in the distribution of two MT-regulatory proteins: EB1 [an MT plus-end tracking protein, +TIP, known to promote MT stability (35, 92)] and MARK4 [a cytosolic Ser/Thr protein kinase known to phosphorylate MT affinity proteins, such as MAP1a, causing their detachment from MTs to promote MT catastrophe (31, 93)], owing to their differential effects on MT dynamics to confer MT plasticity (Fig. 8). EB1 and MARK4 appeared as tracks that lay across the epithelium and aligned perpendicular to the basement membrane (annotated by a dashed white line), and they also colocalized with MTs (visualized by α-tubulin staining), as noted in testes of control (pCI-neo/Ctrl) and low-dose adjudin rats (Fig. 8). However, overexpression of either p-rpS6-MT, F5-peptide, or p-rpS6-MT plus F5-peptide grossly perturbed their distribution and downregulated their expression across the epithelium, thereby rendering MTs as a disorganized network of MT protofilaments (Fig. 8). However, overexpression of p-rpS6-MT plus F5-peptide and low-dose adjudin induced notable changes in MT organization wherein the MT network appeared as crumbled structures that were densely packed at the base of the seminiferous tubules, consistent with data shown in Fig. 7. These findings thus support the notion that the overexpression of p-rpS6-MT and F5-peptide, which modified the BTB permeability, promoted the entry of adjudin into the epithelium, causing more subtle changes in MT organization that impeded spermatogenesis as noted in Fig. 1B.

Figure 8.

Changes in the spatiotemporal expression of EB1 and MARK4 impede MT organization following overexpression of p-rpS6-MT, F5-peptide, or their combined use with or without low-dose adjudin. In testes of rats from the control (pCI-neo/Ctrl) or low-dose adjudin (10 mg/kg b.w., oral gavage) groups, EB1 (green fluorescence) and MARK4 (green fluorescence) lay across the seminiferous epithelium as tracks, perpendicular to the basement membrane (annotated by dashed white line). They also colocalized with MTs (red fluorescence, visualized by α-tubulin staining; see merge images). However, following overexpression of p-rpS6-MT, F5-peptide, or p-rpS6-MT plus F5-peptide, there were considerable disruptive changes in the organization of either EB1 or MARK4 (and also MTs) across the epithelium vs the control testis. For instance, EB1, MARK4, and also MTs no longer lay across the epithelium as undisrupted tracks, perpendicular to the basement membrane. Instead, these tracks were either truncated or they lay as short stretches in parallel (not perpendicular) to the basement membrane. More importantly, overexpression of p-rpS6-MT and F5-peptide coupled with a low-dose adjudin (note that adjudin by itself had no notable effects on the phenotypes in the testis) was capable of inducing more cytoskeletal changes regarding the organization of MTs across the epithelium wherein MTs were found to be almost collapsed, making the phenotypes considerably worse than testes found in p-rpS6-MT, F5-peptide, or p-rpS6-MT plus F5-peptide. Negative controls for EB1/α-tubulin and MARK4/α-tubulin using the corresponding normal IgG of the same animal species are shown in an online repository (29). Micrographs are representative of a single experiment, which were derived from n = 3 independent experiments using n = 3 rat testes and yielded similar results. Scale bars, 40 µm, and 30 µm in inset in green, which apply to corresponding micrographs and insets. DAPI, 4’,6-diamidino-2-phenylindole.

Changes in MT-specific motor protein dynein 1

We also examined changes in the distribution of an MT-specific cargo protein dynein 1 known to promote intracellular cargo transport (including spermatids) across the seminiferous toward the minus end of the MT-based tracks (33, 94). Unlike testes from control- or low-dose adjudin–treated rats wherein dynein 1 appeared as tracks, colocalized with α-tubulin, and lay perpendicular to the basement membrane that stretched across the epithelium, overexpression of either p-rpS6-MT, F5-peptide, or p-rpS6-MT plus F5-peptide rendered gross disruption of dynein 1 distribution along these MT-based tracks (Fig. 9). Consistent with findings noted in Figs. 7 and 8, combined overexpression of p-rpS6-MT and F5-peptide and low-dose adjudin induced more severe damage to MT-based cytoskeletal organization in which the MTs were crumbled into a network of densely packed protofilaments that lay at the base of the tubule, along with a considerable downregulation of dynein 1 expression (Fig. 9).

Figure 9.

Changes in the spatial expression of MT-dependent motor protein dynein 1 impede MT organization following overexpression of p-rpS6-MT, F5-peptide, or their combined use with or without low-dose adjudin. In testes of control (pCI-neo/Ctrl) and low-dose adjudin groups, dynein 1 (green fluorescence), a cytosolic MT minus end–directed motor protein, lay across the seminiferous epithelium perpendicular to the basement membrane (annotated by dashed white line) and colocalized with MTs (visualized by α-tubulin staining, red fluorescence). However, following overexpression of p-rpS6-MT, F5-peptide, or p-rpS6-MT plus F5-peptide, there were considerable disruptive changes in the cytoskeletal organization of MTs across the epithelium vs the control testis. For instance, dynein 1 (and also MTs) no longer lay across the epithelium as undisrupted tracks, perpendicular to the basement membrane. Instead, these tracks were either truncated or they lay as short stretches in parallel (not perpendicular) to the basement membrane, in sharp contrast to the control testis. More important, overexpression of p-rpS6-MT and F5-peptide coupled with a low-dose adjudin (note that low-dose adjudin by itself had no notable effects on the phenotypes in the testis) was capable of inducing more cytoskeletal changes regarding the organization of MTs across the epithelium wherein MTs were found to be almost collapsed, making the phenotypes considerably worse than testes found in p-rpS6-MT, F5-peptide, or p-rpS6-MT+F5-peptide. Negative controls for dynein 1/α-tubulin using the corresponding normal IgG of the same animal species are shown in an online repository (29). Micrographs are representative of a single experiment derived from n = 3 independent experiments using n = 3 rat testes and yielded similar results. Scale bars, 40 µm, and 30 µm in inset in green, which apply to corresponding micrographs and insets in the same panel. DAPI, 4’,6-diamidino-2-phenylindole.

Discussion

Targeted delivery of drugs or therapeutic agents across a blood–tissue barrier to a specific organ, such as the brain and the testis, remains a challenge to investigators in the field. However, advances in nanotechnology have offered some promises, such as magnetic- and silica-based nanoparticles, as well as nanoparticles coated with an antibody (e.g., anti-FSH receptor to target drugs to the testis wherein Sertoli cells express FSH receptors) (95–100). Note that delivery of nonhormonal male contraceptives (e.g., adjudin) that exert their effects behind the BTB in the seminiferous epithelium (such as by perturbing germ cell adhesion to disrupt spermatogenesis) remains a challenge because the immunological barrier poses an obstacle due to the presence of dozens of drug transports, capable of pumping drugs actively out of the testis (1, 8, 101). To provide a more rational approach of drug delivery to the testis, we opted to better understand BTB biology. First, emerging evidence has shown that the testis is producing a BTB modifier designated F5-peptide, which is a biologically active peptide of 50–amino acid residues derived from domain IV of laminin-γ3 chains expressed by elongated spermatids at the apical ES, possibly via the action of matrix metalloproteinase 2 (MMP2) (102), capable of modifying BTB integrity to facilitate the transport of preleptotene spermatocytes across the immunological barrier (20–22, 67). Furthermore, the F5-peptide is capable of potentiating apical ES degeneration, thereby facilitating the release of fully developed spermatids (i.e., spermatozoa) at spermiation (21, 22). Second, the rpS6 signaling protein, downstream of the mTORC1 signaling protein (103, 104), has been shown to work with mTOR to modulate BTB permeability function, because its overexpression in Sertoli cell epithelium in vitro promotes BTB remodeling, causing the TJ barrier to become leaky (25, 26, 48). These findings are also consistent with studies using genetic models in mice wherein a specific deletion of mTOR in Sertoli cells was found to cause an upregulation of p-rpS6-S240/S244 (i.e., an activation of rpS6), but not total rpS6, and was associated with BTB disruption, epithelial damage, loss of cell polarity, and infertility (105). Furthermore, specific deletion of regulatory-associated protein of mTOR (Raptor, an adaptor protein whose binding to mTOR creates the mTORC1 signaling protein) in Sertoli cells in mice also led to extensive epithelial damage due to cytoskeletal disorganization of F-actin, MTs, and vimentin, as well as loss of cell polarity, germ cell exfoliation, and infertility (106). Collectively, studies based on the use of genetic models support the notion that the homeostasis of the mTORC1/rpS6 signaling complex is crucial to support BTB and spermatogenic function.

Recent studies have also shown that besides F5-peptide, the testis is producing another biologically active peptide fragment endogenously, designated LG3/4/5 (an 80-kDa fragment) derived from laminin-α2, a major constituent component of the basement membrane (107). It was shown that an inactivation of laminin-α2 by RNAi using a laminin-α2–specific short hairpin RNA was found to induce Sertoli cell TJ barrier disruption through changes in the cytoskeletal organization of both F-actin and MTs (23, 24). These include a considerable decline in the ability of the Sertoli cell epithelium to polymerize MTs and to assemble actin filaments into bundles to support Sertoli cell and spermatid adhesion (23, 24). More importantly, this inactivation of LG3/4/5 following its knockdown by RNAi was accompanied by a considerable activation of rpS6, such as an upregulation on the expression of p-rpS6-S235/S236 and p-rpS6-S240/S244 (23). In this context, note that laminin chains in mammalian epithelia were earlier shown to be capable of modifying cell permeability barrier function. For instance, a fragment derived from laminin-511, designated iMatrix-511, was found to bind to integrin receptor and was capable of stimulating proliferation and differentiation of MDPC-23 cells into odontogenic differentiation through integrin activation (108). Alternatively, the NH₂ fragment derived from laminin-γ2 chain was shown to increase vascular permeability in endothelial barrier through the p38 MAPK signaling pathway (109). Furthermore, fragments of laminin-332 or laminin-γ2 chain (110) and laminin-α4 chain (110) were capable of promoting neutrophil cell migration in vitro and inhibiting de novo adipogenesis in mice in vivo, respectively. Collectively, these findings thus support the notion that F5-peptide (20–22) and rpS6 (and its constitutively active and quadruple phosphomimetic mutant, p-rpS6-MT, of the mTORC1/rpS6 signaling complex) (25, 26, 111) earlier discovered in our laboratory, which are capable of modulating BTB function and spermatogenesis, are in agreement with published findings regarding these two biomolecules.

We have reported our findings in this study based on the use of these two biologically active biomolecules, namely F5-peptide and p-rpS6-MT, that were cloned into the mammalian expression vector pCI-neo, whose overexpression in the testis was capable of modifying BTB permeability barrier function. However, their combined use was proven to be more potent to modify the BTB function, making it more leaky, and induced more subtle epithelial damage across the epithelium. However, the most important observation in this study is that these BTB modifiers are capable of modifying the transport function at the BTB by using the nonhormonal male contraceptive adjudin as a candidate drug, which was earlier shown to have poor bioavailability based on a study to use [³H]adjudin to monitor its tissue distribution (28). The combined use of these two BTB modifiers along with adjudin at a dose (which by itself had no effects in perturbing spermatogenic function in the testis) was found to induce considerably better phenotypic changes regarding defects in spermatogenic function and BTB integrity based on semiquantitative data analysis. Interestingly, in this context note that adjudin is an effective nonhormonal and reversible male contraceptive, capable of inducing germ cell exfoliation via its effects on actin- and MT-based cytoskeletons through its disruptive effects by perturbing the organization of actin filament bundles (27, 101), but also MTs at the ES as noted herein. However, its entry into the testis was considerably negated by the drug transporters at the BTB site, which were able to actively pump [³H]adjudin out of the Sertoli cell BTB (12, 112). As such, a relative high dose of adjudin is needed to exert its antifertility effects in rodents following its administration by oral gavage (27), thereby causing systemic toxicity as noted in a 29-day subchronic toxicity test (50 mg/kg b.w. for 29 days in both male and female rats of n = 10 rats each), even though its acute toxicity (1 g/kg b.w. in mice vs 2 g/kg b.w. via oral gavage in rats) was acceptable in rodents (113). The rationale behind this study was that if the BTB permeability function was transiently modified, such as via the combined use of F5-peptide and p-rpS6-MT that caused the BTB to become leaky transiently, this would improve the bioavailability of adjudin (and possibly other therapeutic drugs) to the testis. It is also noted that adjudin, even at doses of 125 mg/kg b.w. and 250 mg/kg b.w. (via oral gavage), had no noticeable effects on the population of spermatogonial stem cells (SSCs) when assessed by corresponding markers of SSC and undifferentiated spermatogonia as earlier reported (114). Thus, it is unlikely that the SSCs and undifferentiated spermatogonia were affected by adjudin when it was used at 10 mg/kg b.w. in this report.

In summary, we have demonstrated unequivocally that BTB integrity can be modified following overexpression of F5-peptide plus p-rpS6-MT, which is more potent than the use of either F5-peptide or p-rpS6-MT alone. More importantly, the combined use of these two biomolecules also modifies the transport function at the BTB because the use of a low-dose adjudin, which by itself had no noticeable effects on spermatogenic function, was more capable of perturbing spermatogenic function in the testis than adjudin alone (at a low dose), and was also more potent than either one of the biomolecules alone or their combined use, even more potent that adjudin was used at 50 mg/kg b.w. or higher doses (e.g., 150 mg/kg b.w.) as earlier reported (28, 114, 115). For instance, there is extensive epithelial damage in the testis that leads to having >90% of tubules that are devoid of germ cells as noted herein. Interestingly, these findings thus support the notion that this is a promising approach to deliver drugs to a tissue barrier by modifying the barrier function. Work is now in progress to examine whether these two BTB modifiers were able to induce similar barrier modifications in other blood–tissue barriers such as the blood–brain barrier.

Glossary

Abbreviations:

BTB

blood–testis barrier

b.w.

body weight

CdCl₂

cadmium chloride

EB1

end-binding protein 1

ES

ectoplasmic specialization

HRP

horseradish peroxidase

IB

immunoblot

IF

immunofluorescence

MARK4

microtubule affinity-regulating kinase 4

Mdr1

multidrug resistance protein 1

Mrp1

multidrug resistance-related protein 1

MT

microtubule

mTORC1

mammalian target of rapamycin complex 1

p-

phosphorylated

p-rpS6-MT

quadruple phosphomimetic mutant of rpS6

RNAi

RNA interference

rpS6

ribosomal protein S6

SSC

spermatogonial stem cell

TJ

tight junction