Regulation of the blood–testis barrier by a local axis in the testis: role of laminin α2 in the basement membrane

Abstract

Laminin α2 is one of the constituent components of the basement membrane (BM) in adult rat testes. Earlier studies that used a mouse genetic model have shown that a deletion of laminin α2 impedes male fertility by disrupting ectoplasmic specialization (ES; a testis-specific, actin-rich anchoring junction) function along the length of Sertoli cell in the testis. This includes ES at the Sertoli cell–elongating/elongated spermatid interface, which is known as apical ES and possibly the Sertoli–Sertoli cell interface, known as basal ES, at the blood–testis barrier (BTB). Studies have also illustrated that there is a local regulatory axis that functionally links cellular events of spermiation that occur near the luminal edge of tubule lumen at the apical ES and the basal ES/BTB remodeling near the BM at opposite ends of the seminiferous epithelium during the epithelial cycle, known as the apical ES-BTB-BM axis. However, the precise role of BM in this axis remains unknown. Here, we show that laminin α2 in the BM serves as the crucial regulator in this axis as laminin α2, likely its 80-kDa fragment from the C terminus, was found to be transported across the seminiferous epithelium at stages VIII–IX of the epithelial cycle, from the BM to the luminal edge of the tubule, possibly being used to modulate apical ES restructuring at these stages. Of more importance, a knockdown of laminin α2 in Sertoli cells was shown to induce the Sertoli cell tight junction permeability barrier disruption via changes in localization of adhesion proteins at the tight junction and basal ES at the Sertoli cell BTB. These changes were found to be mediated by a disruption of F-actin organization that was induced by changes in the spatiotemporal expression of actin binding/regulatory proteins. Furthermore, laminin α2 knockdown also perturbed microtubule (MT) organization by considerable down-regulation of MT polymerization via changes in the spatiotemporal expression of EB1 (end-binding protein 1), a +TIP (MT plus-end tracking protein). In short, laminin α2 in the BM seems to play a crucial role in the BTB-BM axis by modulating BTB dynamics during spermatogenesis.—Gao, Y., Mruk, D., Chen, H., Lui, W.-Y., Lee, W. M., Cheng, C. Y. Regulation of the blood–testis barrier by a local axis in the testis: role of laminin α2 in the basement membrane.

Seminiferous tubules are the functional unit of the mammalian testis wherein spermatogenesis takes place through cyclic events of spermatogonial renewal via mitosis, meiosis, postmeiotic spermatid development, and spermiogenesis, which leads to the release of sperm at spermiation (for reviews, see refs. 1, 2). Of interest, cellular events that take place across the seminiferous epithelium are intriguingly coordinated. For instance, spermiation takes place at the luminal edge near the tubule lumen in the adluminal compartment at stage VIII in rodents vs. late stage II in humans, which coincides with remodeling of the blood–testis barrier (BTB) near the basement membrane (BM) to facilitate transport of preleptotene spermatocytes that reside in the basal compartment across the immunological barrier (for reviews, see refs. 3–5). Earlier studies have shown that biologically active fragments, such as F5-peptide (6), that are generated at the apical ectoplasmic specialization (ES; an actin-rich, testis-specific anchoring device at the elongated spermatid-Sertoli cell interface) at spermiation—likely via the action of matrix metalloprotease-2 (MMP-2) (7) on laminin-γ3, a bona fide apical ES adhesion protein in elongated spermatid cell surface (8, 9)—induce BTB remodeling (10). In this context, it is noted that the BTB is composed of an ultrastructure that is similar to that of the apical ES, but is designated basal ES, which coexists with tight junction (TJ), gap junction, and desmosome to constitute the BTB. ES is typified by the presence of an extensive network of actin microfilament bundles, sandwiched between the cisternae of endoplasmic reticulum and the opposing Sertoli-spermatid (apical ES) vs. Sertoli-Sertoli (basal ES) plasma membranes (for reviews, see refs. 5, 11, 12). These findings (6, 10) illustrate the presence of an autocrine-based functional axis that coordinates cellular events, namely spermiation and BTB remodeling, that take place at opposite ends of the seminiferous epithelium at stage VIII of the epithelial cycle in the rat testis (for a review, see ref. 13). The concept of this local regulatory axis has been confirmed by other investigators by using a phthalate-induced Sertoli cell injury model in which phthalate-mediated MMP-2 activation leads to laminin chain degeneration that also modulates BTB restructuring (14, 15).

As BTB is located near the BM, it is likely that there is a local axis in which BM regulates BTB restructuring events to support transport of preleptotene spermatocytes across the barrier during the epithelial cycle. Indeed, a recent report has shown that the NC1 domain released from the collage α3 (IV) chain in BM modulates BTB restructuring (16), which supports the notion that there is a functional axis between the BTB and the BM. In addition to collagen, laminin α1, α2, α4, β1, β2, and γ1 are the major laminin chains found in the BM of seminiferous tubules in rodent testes (9, 17). Furthermore, in a study that used a genetic model, it was shown that transgenic mice that were deficient in laminin α2 chain (also known as merosin) led to a disruption of the apical ES, which caused germ cell exfoliation and male infertility (17), supporting the notion that there is a local axis between apical ES and BM and illustrating the significance of laminin α2 chain in male fertility. Because apical and basal ES share virtually identical ultrastructural features, it is envisioned that laminin α2 chain can modulate basal ES/BTB function as well as apical ES. Here, we demonstrate for the first time, to our knowledge, that laminin α2 plays a role in the regulation of the Sertoli cell TJ-permeability barrier via its effects on F-actin organization at the basal ES, which illustrates the presence of a BTB-BM axis.

MATERIALS AND METHODS

Animals and antibodies

Male Sprague-Dawley rats were purchased from Charles River Laboratories (Kingston, NY, USA). Ten pups were shipped and housed with a foster mother. Use of rats for experiments reported herein was approved by the Rockefeller University Institutional Animal Care and Use Committee (protocols: 12-506, 15-780-H). Antibodies were obtained commercially, unless specified otherwise, and are listed in Table 1.

TABLE 1.

Antibodies used for different experiments

Antibody	Host species	Vendor	Catalog no.	Working dilution
				IB	IF/IHC
Actin	Goat	Santa Cruz Biotechnology (Santa Cruz, CA, USA)	sc-1616	1:300
Arp3	Mouse	Sigma-Aldrich	A5979	1:3000	1:50
α-Tubulin	Mouse	Abcam (Cambridge, MA, USA)	ab7291		1:500 (tissue) 1:300 (cell)
β-Catenin	Rabbit	Thermo Fisher Scientific	71-2700	1:250	1:100
β-Tubulin	Rabbit	Abcam	ab6046	1:1000	1:300
CAR	Rabbit	Santa Cruz Biotechnology	sc-15405	1:200	1:50
Detyrosinated α-tubulin	Rabbit	Abcam	ab48389	1:1000	1:200
Dia1	Goat	Santa Cruz Biotechnology	sc-10885	1:200
EB1	Mouse	BD Biosciences	610534		1:200
EB1	Rabbit	Santa Cruz Biotechnology	sc-15347	1:200
Eps8	Mouse	BD Biosciences	610143	1:5000	1:50
Formin 1	Mouse	Abcam	ab68058	1:500	1:50
JAM-A	Rabbit	Thermo Fisher Scientific	36-1700	1:250
Katanin p80	Rabbit	Santa Cruz Biotechnology	sc-292216	1:200
Laminin α2	Mouse	Millipore	MAB1922	1:300	1:300 (tissue) 1:100 (cell)/1:50
MAP2	Rabbit	Proteintech (Rosemont, IL, USA)	17490-1-AP	1:500
MARK2	Rabbit	Proteintech	15492-1-AP	1:2000
MARK4	Rabbit	Proteintech	20174-1-AP	1:2000
N-cadherin	Mouse	Thermo Fisher Scientific	33-3900	1:200	1:100
Palladin	Rabbit	Proteintech	10853-1-AP	1:1000	1:100
Plastin 3	Rabbit	Abcam	ab137585	1:500
Spastin	Rabbit	Proteintech	22792-1-AP	1:500
ZO-1	Rabbit	Thermo Fisher Scientific	61-7300	1:250	1:100
Goat IgG-HRP	Bovine	Santa Cruz Biotechnology	sc-2350	1:3000
Rabbit IgG-HRP	Bovine	Santa Cruz Biotechnology	sc-2370	1:3000
Mouse IgG-HRP	Bovine	Santa Cruz Biotechnology	sc-2371	1:3000
Rabbit IgG-Alexa Fluor 488	Goat	Thermo Fisher Scientific	A-11034		1:250
Rabbit IgG-Alexa Fluor 555	Goat	Thermo Fisher Scientific	A-21429		1:250
Mouse IgG-Alexa Fluor 555	Goat	Thermo Fisher Scientific	A-11029		1:250
Biotinylated Mouse IgG	Horse	Vector Laboratories (Burlingame, CA, USA)	BA-2000		1:300

Dia1 (also known as mDia1), diaphanous-related formin-1; HRP, horseradish peroxidase; JAM-A, junctional adhesion molecule-A; MAP, microtubule-associated protein; MARK, MAP/microtubule affinity-regulating kinase.

Primary Sertoli cell cultures

Sertoli cells were isolated from the testes of 20-d-old rats as described (18). Cells were seeded on Matrigel (1:5 to ∼1:7, diluted in F12/DMEM; BD Biosciences, San Jose, CA, USA)-coated culture plates (for lysate preparation, RNA isolation, and other biochemical assays), coverslips [for immunofluorescence (IF) microscopy], or bicameral units [for transepithelial electrical resistance (TER) measurement; Millipore, Billerica, MA, USA] at densities 0.5, 0.03 to ∼0.04, or 1.0 × 10⁶ cells/cm², respectively. Sertoli cells were cultured in serum-free F12/DMEM (Thermo Fisher Scientific, Waltham, MA, USA) that was supplemented with growth factors and gentamicin, as previously described (18), in a humidified atmosphere of 95% air/5% CO₂ (v/v) at 35°C. On d 2, Sertoli cells were subjected to a brief hypotonic treatment with 20 mM Tris (pH 7.4) for 2.5 min, as described (19), to lyse residual germ cells. Sertoli cell cultures contained ∼98% Sertoli cells, with negligible contaminations of Leydig cells, peritubular myoid cells, or germ cells when assessed by specific markers of these other cells using either immunoblotting (IB) or RT-PCR as previously described (20). Sertoli cells cultured in vitro were shown to establish a functional TJ-permeability barrier with ultrastructures of TJs, basal ESs, gap junctions, and desmosomes that mimic Sertoli cell BTB in vivo as previously described (21, 22).

Knockdown of Sertoli cell laminin α2 by short hairpin RNA

SureSilencing short hairpin RNA (shRNA) plasmids were purchased from Qiagen (Valencia, CA, USA). shRNAs, including both laminin α2 and negative control, were provided in pGeneClip hMGFP vector. The sequence of shRNA that specifically targeted laminin α2 (clone ID 4) was 5′–ACAGGAAGCTGATCGGCTAAT–3′, and the sequence of negative control was 5′–GGAATCTCATTCGATGCATAC–3′. Clone ID 1–3 of laminin α2 shRNAs were found to be less effective in silencing laminin α2 on the basis of pilot experiments and were not used in subsequent investigations. All plasmids were prepared by ZymoPure Plasmid Midiprep Kit (Zymo Research, Irvine, CA, USA). Sertoli cells were cultured in vitro for 3 d as described above with an established functional TJ-permeability barrier that contained ultrastructures of TJs, basal ES, gap junctions, and desmosomes when examined by electron microscopy (21, 22). Sertoli cells were transfected with plasmid DNA at 0.5 μg (for IF on coverslips placed in 12-well plates, with 2-ml F12/DMEM/well), 1.5 μg (for RT-PCR and IB in 12-well plates, with 3-ml F12/DMEM/well), 0.75 μg (for TJ-barrier function assessment using bicameral units which were placed in 24-well plates, with 0.5-ml F12/DMEM in the apical and basal compartment), or 2.7 μg (for biochemical assays in 6-well plates, with 5-ml F12/DMEM/well) by using Lipojet In Vitro Transfection Reagent (SignaGen Laboratories, Rockville, MD, USA) using a 3-μl transfection medium:1-μg plasmid DNA ratio. After 24 h, cells were rinsed with F12/DMEM twice and then cultured in fresh F12/DMEM that was supplemented with growth factors and gentamicin. RNA and protein lysates were obtained from these cultures 48 h (for IF or RT-PCR) or 72 h (for IB or biochemical assays) post-transfection for analysis.

Treatment of adult rats with taxol (paclitaxel)

Taxol is known to stabilize microtubules and inhibit microtubule-dependent cellular processes in the testis (23). Adult male rats were treated with taxol as described (23). Taxol (Mr 853.9; Sigma-Aldrich, St Louis, MO, USA) was prepared by dissolving 5 mg in 200 μl DMSO. Approximately 42.7 μg taxol was diluted in 100 μl PBS compared with PBS alone (negative control) and both were administered intratesticularly to each testis (∼1.6 g in weight with a volume of ∼1.6 ml) by using a 29-gauge needle attached to a 0.5-ml insulin syringe by gently inserting from the apical to near the basal end of the testis vertically. Thus, taxol was used at ∼30 µM in the testis. As the needle was withdrawn apically, solution was released gently and gradually from the syringe so the entire testis was slowly filled with solution to avoid an acute rise in intratesticular hydrostatic pressure. Rats were euthanized by CO₂ asphyxiation 24 h (n = 6) or 36 h (n = 6) thereafter by using slow (20–30%/min) displacement of chamber air with compressed CO₂ in a euthanasia chamber that was approved by the Rockefeller University Comparative Bioscience Center and Rockefeller University Laboratory Safety and Environmental Health. Testes from 6 rats (24 h, n = 3; 36 h, n = 3) were snap-frozen in liquid nitrogen and stored at −80°C until used for IF. Because the phenotypes of these rats following taxol treatment that impeded the transport of laminin α2 chain across the seminiferous epithelium was similar between rats that were terminated at 24 and 36 h, data from both time points were combined for analysis. For α-tubulin staining, testes from 6 rats (24 h, n = 3; 36 h, n = 3) were prefixed with 4% paraformaldehyde (PFA) at 4°C for 1 wk, then snap-frozen in liquid nitrogen.

Lysate preparation and IB

Protein lysates from primary Sertoli cells were obtained by using an immunoprecipitation lysis buffer [50 mM Tris, 0.15 M NaCl, 1% NP-40, 2 mM EGTA, and 10% glycerol (v/v), pH 7.4 at 22°C] that was supplemented with protease and phosphatase inhibitor mixture (Sigma-Aldrich) and phosphatase inhibitor mixture II (Sigma-Aldrich). Protein concentration was determined by using a DC Protein Assay Kit from Bio-Rad (Hercules, CA, USA). Equal amount of protein lysate were resolved by SDS-PAGE and transferred to nitrocellulose membrane (Bio-Rad) for immunoblot analysis using the corresponding primary and secondary antibody (Table 1). Target proteins were visualized by chemiluminescence using a kit prepared in-house as described (24). Images were acquired and quantified by using a Fujifilm LAS-4000 mini-Luminescent Image Analyzer and Multi Gauge software package (version 3.1; Tokyo, Japan).

Immunohistochemistry and dual-labeled IF analysis

Immunohistochemistry (IHC) was performed by using frozen sections of testes (7-μm thick) that were obtained with a cryostat at −22°C. In brief, sections were fixed in 4% PFA in PBS for 10 min, followed by permeabilization with 0.1% Triton X-100 in PBS (v/v) for 10 min, and were blocked with 10% normal goat serum. Sections were then incubated with laminin α2 antibody (Table 1) at 4°C overnight, followed by incubation of biotinylated secondary antibody and streptavidin-horseradish peroxidase (Thermo Fisher Scientific). Color development was performed by using aminoethylcarbazole kits (Thermo Fisher Scientific). To visualize α-tubulin, testes were fixed in 4% PFA for 1 wk at 4°C before being snap-frozen in liquid nitrogen to obtain sections for IF. Dual-labeled IF was performed by using frozen sections of testes (7-μm thick) or Sertoli cells that were cultured on coverslips at a density of 0.03 to ∼0.04 × 10⁶ cells/cm². Sections and/or cells were fixed with 4% PFA or ice-cold methanol, permeabilized with 0.1% Triton X-100 in PBS, and subsequently blocked with 10% normal goat serum or 1% bovine serum albumin in PBS. Sections and/or cells were then incubated with primary antibodies (Table 1) at 4°C overnight, followed by Alexa Fluor 488 (green)– or Alexa Fluor 555 (red)–conjugated secondary antibodies (Thermo Fisher Scientific) for 1 h. To visualize F-actin, sections and/or cells were incubated with rhodamine-conjugated phalloidin (Thermo Fisher Scientific) at 1:50 dilution for 1 h. Slides were mounted in Prolong Gold Antifade reagent with DAPI (Thermo Fisher Scientific) to visualize cell nuclei. Fluorescence and IHC images were examined and acquired by using a Nikon Eclipse 90i Fluorescence Microscope system equipped with Nikon Ds-Qi1Mc and Nikon DS-Fi1 digital cameras, with the Nikon NIS Elements Imaging Software (Nikon, Tokyo, Japan). Image files were then analyzed by using Photoshop in Adobe Creative Suite (version 3.0; San Jose, CA, USA) for image overlay to assess protein colocalization. All sections of testes or Sertoli cells within an experimental group, including controls, were processed simultaneously in a single experimental session to eliminate interexperimental variations, and each experiment was repeated 3 times using different rat testes, yielding similar observations.

Assessment of Sertoli cell TJ-permeability barrier function in vitro

Sertoli cells were seeded on Matrigel (1:5)-coated bicameral units (diameter, 12 mm; pore size, 0.45 μm; effective surface area, 0.6 cm²) at 1 × 10⁶ cells/cm². Bicameral units were then placed in 24-well plates with 0.5 ml F12/DMEM each in the apical and basal compartments. To assess Sertoli cell TJ-permeability barrier, both the silencing and control groups had triplicate bicameral units and TER across the cell epithelium was recorded daily. Of note, TER monitored the ability of the Sertoli cell TJ-barrier to resist conductivity of electrical current that was sent across the electrodes of the Millipore Millicell ERS system as described (18). Each experiment was repeated at least 3 times using different batches of Sertoli cells, yielding similar results.

Actin bundling assay

Actin bundling assay using lysates of Sertoli cells was performed as previously described (25) by using Actin Binding Protein Spin-Down Assay Biochem Kits from Cytoskeleton (Denver, CO, USA). In brief, 10 µl of cell lysates from 6-well dishes vs. 10 μl Tris lysis buffer (serving as a negative control) was added to F-actin microfilaments that were generated by using the kit at ∼21 µM, as detailed previously (25), to assess the ability of cell lysates to induce actin bundling. This mixture was centrifuged at 14,000 g at 24°C for 5 min to sediment bundled F-actin in the pellet, whereas linear and unbundled actin microfilaments remained in the supernatant. The entire pellet from each sample tube was resuspended in 30 μl sterile water, and 5 μl supernatant from each sample was analyzed by immunoblotting for actin. This experiment was repeated with 4 independent experiments using different cell preparations.

Microtubule polymerization assay

Microtubule (MT) polymerization assay was performed as described (26). Primary Sertoli cells were seeded on 6-well plates at 0.4 × 10⁶ cells/cm². Both the control and treatment groups had duplicate wells. On day 3, cells were transfected with 2.7 μg laminin α2 shRNA vs. control shRNA. After 24 h, cells were rinsed with F12/DMEM twice and then cultured in fresh F12/DMEM that was supplemented with growth factors and gentamicin for an additional 48 h. On day 6, cells were harvested in prewarmed lysis and MT stabilization buffer (100 mM PIPES, 5 mM MgCl₂, 1 mM EGTA, 0.1% NP-40, 0.1% Triton X-100, 0.1% Tween 20, 0.1% 2-mercaptoethanol, 30% glycerol, pH 6.9). Thereafter, cells were homogenized with a 22-gauge syringe needle and centrifuged at 100,000 g for 30 min at 35°C to separate free tubulin monomers (supernatant) from MTs/polymerized tubulin (pellet). Supernatant was collected and 40 μl supernatant of each sample were analyzed by immunoblotting for β-tubulin. The pellet was resuspended in 250 μl 2 mM CaCl₂, and 30 μl of each sample was used for IB. This experiment was repeated with a total of 4 independent experiments using different cell preparations.

RNA isolation and RT-PCR

Total RNA was isolated from Sertoli cells by using Trizol reagent (Thermo Fisher Scientific) and was reverse transcribed with M-MLV reverse transcriptase (Promega, Madison, WI, USA). Reverse transcriptase products were used as templates for PCR with specific primer pairs for laminin α2 and S16 (Table 2). Coamplifications of laminin α2 and S16 were both in linear phase, and PCR was performed as described (8). Authenticity of the PCR product was confirmed by direct nucleotide sequencing at Genewiz (South Plainfield, NJ, USA).

TABLE 2.

Primers used for RT-PCR

Gene	Primer, 5′–3′				Length (bp)	Tm (°C)	Cycle no.	GenBank accession no.
	Sense	Position	Antisense	Position
Laminin α2	CACCAGAGAGGCTCCTTCAGTTGGC	5009–5033	CCCGCGAGTCGGTTGGCTTCATCAA	5642–5666	658	62	30	XM_008758643.1
S16	TCCGCTGCAGTCCGTTCAAGTCTT	15–38	CCAAACTTCTTGGTTTCGCAGCG	376–399	385			XM_341815

Image analysis

For in vitro experiments that assessed the effects of laminin α2 silencing on target protein distribution in Sertoli cells at the cell-cell interface, at least 200 cells were randomly selected and examined in experimental vs. control groups with 3–5 independent experiments using different batches of cell preparations. Fluorescence intensity of a target protein in Sertoli cells was quantified by using ImageJ 1.45 software (National Institutes of Health, Bethesda, MD, USA).

Statistical analysis

For studies that used Sertoli cell cultures, triplicate coverslips, dishes, or bicameral units were used. Each data point (or bar graph) is given as the mean ± sd of 3–5 experiments. Statistical analysis was performed by using the GB-STAT software package (version7.0; Dynamic Microsystems, Silver Spring, MD, USA). Statistical analysis was performed by 2-way ANOVA followed by Dunnett’s test. In selected experiments, Student’s t test was used for paired comparisons.

RESULTS

Localization and expression of laminin α2 in the rat testis

Laminin α2, also known as merosin, is a ∼390-kDa protein that can be cleaved at amino acid residue 2580 by MMPs to generate a 300-kDa N-terminal and a 80-kDa C-terminal fragment in both humans (27) and rodents (17). Laminin α2 is mostly found in the BM of skeletal muscle, testis, brain, and thymus; its deletion and/or mutation leads to muscular dystrophy and male infertility in humans and rodents (17) (for reviews, see refs. 28, 29). Using a primer pair that is specific to laminin α2, we found laminin α2 to be expressed by Sertoli and germ cells in the rat testis (Fig. 1A). Using an antibody that is specific to the 80-kDa C-terminal fragment of laminin α2 (Table 1), laminin α2 protein of 80 kDa was also detected in lysates of Sertoli and germ cells in the rat testis (Fig. 1A), which suggested the presence of an 80-kDa fragment of laminin α2 chain in lysates of Sertoli and germ cells, as well as in testis lysates. In Sertoli cells that were cultured in vitro for 4 d, laminin α2 indeed was found at the cell-extracellular matrix (ECM) interface when examined by immunofluorescence microscopy (Fig. 1B), consistent with earlier reports that Sertoli cells produced and deposited ECM when cultured in vitro (30, 31). Partial colocalization of laminin α2 with F-actin was observed, but considerably less colocalization of laminin α2 with β-tubulin (Fig. 1B). Using the antibody that is specific to the 80-kDa C-terminal fragment of laminin α2 for IHC (Fig. 1C), which thus recognized the entire laminin α2 chain (Fig. 1D), laminin α2 was indeed localized to the base of the seminiferous epithelium, which is consistent with its localization at the BM (Fig. 1C). However, it was noted that immunoreactive laminin α2 chains appeared as aggregates and were found in the adluminal compartment in stage VIII to early IX (Fig. 1C). Presence of laminin α2 chain in the adluminal compartment at stages VIII–IX was further confirmed by fluorescence microscopy (Fig. 1E). As noted in the IF micrographs, these aggregates did not appear to be localized at the apical ES—the ultrastructure at the Sertoli-elongating/elongated spermatids interface (Fig. 1E, boxed area). Instead, these aggregates were found mostly near the elongating spermatids (Fig. 1E), which suggested that they localized in the Sertoli cell cytosol.

Figure 1.

Cellular and stage-specific expression of laminin α2 in the rat testis. A) Steady-state mRNA and protein levels of laminin α2 in adult rat testes (T), Sertoli cells (SC), and germ cells (GC) were analyzed by RT-PCR and IB. For RT-PCR, laminin α2 mRNA was analyzed by using specific primer pairs (Table 2). S16 served as a loading control. For IB, β-actin served as loading control. Each bar in the histogram is mean ± sd of 3 experiments in which the level of laminin α2 in the testis was arbitrarily set at 1. B) Sertoli cells were stained with anti–laminin α2 antibody (red), FITC-phalloidin (green), or β-tubulin antibody (green). Sertoli cell nuclei were stained with DAPI (blue). Scale bar, 30 μm. C) Localization of laminin α2 in the seminiferous epithelium of tubules at stages I–XIV of the epithelial cycle, as shown by IHC. Boxed area was magnified, illustrating that laminin α2 was detected near the round and elongating spermatids in the adluminal compartment in stage VIII-IX tubules, besides the basement membrane. Negative control was performed by using mouse IgG to substitute the primary antibody, illustrating the specificity of IHC staining. The specificity of anti–laminin α2 antibody was supported by IB using lysates of Sertoli cells that yielded a prominent band with an apparent Mr corresponding to the 80-kDa laminin fragment. Scale bars, 200 μm (upper panel), 80 μm (lower panel), 20 μm (inset). D) A schematic drawing illustrating the different domains of laminin α2. The short arm of laminin α2 from the N terminus is comprised of rod domains of epidermal growth factor [EGF; laminin EGF like domain a (LEa), LEb, LEc] and globular domains [laminin N-terminal domain (LN), laminin 4a domain (L4a), laminin 4b domain (L4b)]. Long arm of laminin α2 is comprised of laminin coiled-coil (LCC) domain and 5 C-terminal laminin globular (LG) domains at the C-terminus. Proteolytic cleavage is close to the N terminus of LG3 (black arrow) (72). The anti–laminin α2 antibody recognizes the C-terminal LG domains and detects the 80-kDa fragment of laminin α2 (Table 1). E) Stage-specific expression of laminin α2 (red) in the seminiferous epithelium of adult rat testes was also shown by IF analysis. Cell nuclei were visualized by DAPI. Scale bars, 80 μm, 15 μm (inset).

Transport of laminin α2 across the seminiferous epithelium is an MT-dependent event

Studies have shown that intracellular protein trafficking and/or organelle transport that includes endosomes are MT-dependent events (for reviews, see refs. 32, 33). We next examined whether the laminin α2 chain that was detected in the adluminal compartment was derived and being transported from the BM. In this study, adult rats were treated with taxol (also known as paclitaxel, an MT-stabilizing agent; for a review, see ref. 34) for 24–36 h (n = 12 rats each for control vs. taxol-treated group), and control rats received the same volume of PBS only via intratesticular administration as described (23). Concentration and treatment duration of taxol used in our study were selected on the basis of the earlier report (23). Consistent with the earlier finding (23), most seminiferous tubules in these taxol-treated testes seemed to be relatively normal (Fig. 2, top), except that taxol was shown to disrupt the track-like structures conferred by MTs when MTs were visualized by α-tubulin staining (Fig. 2, middle) in virtually all tubules examined. It was also found that some spermatids failed to be transported to the luminal edge in some stage VIII tubules, likely as a result of the disruption of MTs (Fig. 2, top and middle, yellow arrowheads). This notion is also supported by findings in a recent report that a disruption of the MT-based cytoskeleton by adjudin would impede spermatid transport as a result of the disruption of the MT-based tracks (35). In these taxol-treated testes, the level (or expression) of laminin α2 that was found at the adluminal compartment in stage VIII–IX tubules was considerably reduced in most of the tubules, and even some of the laminin α2 that managed to be transported out of the BM failed to reach the end of the adluminal compartment but remained trapped near the basal region of the epithelium (Fig. 2, top). In addition, taxol had no noticeable effect on F-actin network in the testis (Fig. 2, bottom). This also confirmed that the cell viability was not affected by taxol using this treatment regimen, which is consistent with the earlier report (23). These findings thus support the notion that the immunoreactive laminin α2 found in the adluminal compartment is derived from the BM and might be transported via the track-like structures provided by the MTs, and after treatment with taxol, which specifically disrupted the MT-based track-like structures, the transport of immunoreactive laminin α2 across the epithelium was grossly perturbed.

Figure 2.

Laminin α2 in the adluminal compartment is likely derived from the BM via an MT-dependent transport mechanism. To investigate the origin of laminin α2 (Lam α2) in the adluminal compartment found in stage VIII tubules, testes were treated with taxol (paclitaxel, an MT-stabilizing reagent). In stage VIII tubules from control (ctrl) testes, laminin α2 was localized predominantly in the BM but some were found in the adluminal compartment; however, in some stage VIII tubules at 24–36 h after taxol treatment, laminin α2 failed to reach the luminal edge of the adluminal compartment but dispersed across the epithelium, and in some stage VIII tubules, considerably fewer laminin α2 reached the luminal edge of the adluminal compartment (top). This defect in transport of laminin α2 across the epithelium might be an MT-dependent event as the track-like ultrastructures conferred by MTs (green fluorescence for α-tubulin, the building block of MT) that served as tracks for laminin α2 transport were grossly disrupted by taxol (middle). However, the track-like ultrastructures conferred by F-actin across the epithelium remained relatively intact in ctrl vs. taxol-treated testes (bottom). Cell nuclei were visualized by DAPI. Yellow arrowheads illustrate some step 19 spermatids in the late stage VIII tubules that remained embedded inside the seminiferous epithelium, failing to be transported near the tubule lumen to prepare for spermiation. Scale bar, 30 μm.

Knockdown of laminin α2 perturbs the Sertoli cell TJ-permeability barrier via changes in localization of TJ and basal ES proteins at the Sertoli cell-cell interface

Laminin α2 was silenced by RNAi. Laminin α2 expression was reduced by ∼60 and ∼50% (P < 0.01) on the basis of analysis by RT-PCR and IB, respectively (Fig. 3A). This knockdown of laminin α2 in the Sertoli cell epithelium with an established functional TJ-barrier did not affect the steady-state levels of proteins at the TJ, basal ES, actin regulatory proteins, or MT regulatory proteins, except TJ-integral membrane protein JAM-A (junctional adhesion molecule-A), which was down-regulated by ∼35% (P < 0.01; Fig. 3A). However, a knockdown of laminin α2 was found to perturb the distribution of TJ proteins CAR (coxsackievirus and adenovirus receptor; an integral membrane protein) and ZO-1 (zonula occludens 1; an adaptor protein) as well as basal ES proteins N-cadherin (an integral membrane protein) and β-catenin (an adaptor protein) such that these proteins no longer tightly localized to the Sertoli cell-cell interface to confer cell adhesion function (Fig. 3B). Instead, these TJ- and basal ES-proteins were diffusively localized at the cell-cell interface and moved toward the cell cytosol (as illustrated in the semiquantitatively analyzed data shown in the right panel of Fig. 3B). These changes thus destabilized the Sertoli cell TJ-permeability barrier, which contributed to a partial disruption of the barrier function (as noted in Fig. 3C).

Figure 3.

Knockdown of laminin α2 perturbs the Sertoli cell TJ-permeability barrier function. Sertoli cells that were cultured for 3 d were transfected with control shRNA (ctrl) vs. laminin α2 (Lam α2) shRNA for 24 h. Cells were then rinsed and cultured in fresh F12/DMEM and were terminated 48 and 72 h thereafter for RNA extraction (or for IF) and preparation of cell lysates (for IB), respectively. A) RT-PCR confirmed considerable knockdown of laminin α2 by shRNA. S16 served as a loading control. Studies by IB also illustrated that laminin α2 protein was silenced. All BTB-associated proteins and actin vs. MT regulatory proteins examined were not affected by laminin α2 knockdown, with the exception of JAM-A (junctional adhesion molecule-A), which was reduced. Each bar in the histogram is the mean ± sd of 6 independent experiments. **P < 0.01 by Student’s t test. B) Study by IF confirmed that laminin α2 knockdown considerably reduced its expression in Sertoli cells. Laminin α2 knockdown also affected the localization of TJ proteins CAR and ZO-1, as well as basal ES proteins N-cadherin and β-catenin. In control cells, these proteins localized at the cell-cell interface (white brackets). Laminin α2 knockdown caused rapid internalization of these proteins so that these BTB-associated proteins no longer tightly localized at the Sertoli cell-cell interface (yellow brackets). Sertoli cell nuclei were visualized by DAPI (blue). Expression of GFP (green) illustrated successful transfection as GFP is an integrated component of the vector DNA (see Materials and Methods). Scale bar, 30 μm. Histograms on the right panel summarize IF results for each target protein shown on the two left panels, with each bar a mean ± sd of 3 experiments. **P < 0.01 by Student’s t test. C) Laminin α2 knockdown perturbed the Sertoli cell TJ-permeability barrier, with each data point a mean ± sd of TER measurement of a representative experiment with triplicate bicameral units. A total of 3 experiments yielded similar results. Dia1 (also known as mDia1), diaphanous-related formin-1; JAM-A, junctional adhesion molecule-A; MAP, microtubule-associated protein; MARK, MAP/microtubule affinity-regulating kinase. **P < 0.01.

Laminin α2 knockdown perturbs actin microfilament organization via changes in spatiotemporal expression of actin binding and regulatory proteins at the Sertoli cell BTB

Cell adhesion protein complexes of the TJ (i.e., CAR/ZO-1) and the basal ES (i.e., N-cadherin/β-catenin) all used F-actin for attachment. We next examined whether their disruptive localization at the cell cortical zone (seen in Fig. 3B) was related to changes in actin microfilament organization in Sertoli cells. As shown in Fig. 4A, organization of actin microfilaments across the Sertoli cell cytosol was disrupted after laminin α2 knockdown in Sertoli cells as microfilaments no longer stretched across the entire Sertoli cell cytosol, but became grossly truncated. This observation is consistent with an earlier report that laminin α2^−/− mice were found to have a disrupted F-actin network at the apical ES in their testes (17). As noted here, these disruptive changes in F-actin organization were likely the result of changes in the spatiotemporal expression of branched actin polymerization protein Arp3 (actin-related protein 3) and actin barbed end capping and bundling protein Eps8 (epidermal growth factor receptor pathway substrate 8), but not the actin bundling protein palladin, after laminin α2 knockdown (Fig. 4A). For instance, Arp3 was found mostly inside the cell cytosol instead of at the cell-cell interface, which facilitated branched actin polymerization that caused the bundled actin microfilaments to form into a branched but truncated network in laminin α2 knockdown cells as noted in F-actin stained cells (Fig. 4A). The actin bundling protein Eps8 also shifted its localization from the cell cortical zone to the cell cytosol, thereby impeding the organization of actin microfilaments into bundles in these cells (Fig. 4A). In short, these changes led to an overall reduction of the relative amount of bundled actin microfilaments as confirmed in a biochemical assay that assessed the level of bundled actin microfilaments, which was reduced considerably after laminin α2 knockdown (Fig. 4B).

Figure 4.

Laminin α2 (Lam α2) knockdown in Sertoli cells induces F-actin disorganization mediated by changes in the spatial expression of actin binding and regulatory proteins. A) After laminin α2 knockdown, F-actin (red or gray) organization was grossly affected. Localization of branched actin polymerization protein Arp3 and actin barbed end capping and bundling protein Eps8 were also considerably perturbed after laminin α2 knockdown. Localization of actin cross-linking/bundling protein palladin was not visibly affected. Sertoli cell nuclei were visualized by DAPI (blue). Expression of GFP (green), which is an integrated component of the transfection vector, illustrated successful transfection. B) Actin bundling assay illustrated that laminin α2 knockdown caused considerable reduction of actin bundling capability in Sertoli cells. In this biochemical assay, pellet contained bundled F-actin, whereas supernatant contained unbundled and linear actin microfilaments. ctrl, control. Each bar in the histogram is mean ± sd of 3 independent experiments. Scale bar, 30 μm. **P < 0.01 by Student’s t test.

Laminin α2 knockdown perturbs MT organization

Studies have shown that proper MT organization in Sertoli cells is crucial to support the transport of spermatids and cell organelles (e.g., residual bodies, phagosomes, and endosomes) across the seminiferous epithelium during spermatogenesis (35–37). Thus, MT also plays a crucial role in Sertoli cell BTB function. In fact, a knockdown of end-binding protein 1 (EB1)—a +TIP (MT plus-end tracking protein) that promotes MT polymerization—was recently shown to perturb the Sertoli cell TJ-barrier function (26). We thus examined whether a knockdown of laminin α2 would impede MT organization in Sertoli cells. Of interest, a knockdown of laminin α2 was found to induce obvious phenotypic changes in MT organization in Sertoli cells wherein MT filaments no longer stretched across the cell cytosol but retained close to the cell nucleus (Fig. 5A), which seemed to be the result of reduced polymerization. This notion was supported by a biochemical assay that assessed the extent of MT polymerization and showed that laminin α2 knockdown caused considerable reduction in MT polymerization in Sertoli cells, almost to the level in which cells were treated with CaCl₂ that inhibited MT polymerization vs. taxol, an MT-stabilizing agent that prevented MT to undergo depolymerization (Fig. 5B). Localization of detyrosinated α-tubulin—a form of stabilized MT that is less dynamic, with reduced capability of undergoing catastrophe—was also perturbed after laminin α2 knockdown. For instance, detyrosinated α-tubulin no longer stretched across the Sertoli cells, but surrounded the cell nuclei instead. Moreover, laminin α2 knockdown perturbed the localization of EB1 such that this +TIP no longer associated with MTs in Sertoli cells to confer MT growth at their plus (+) ends (Fig. 5A). In short, laminin α2 knockdown in Sertoli cells perturbed MT organization via a disruptive effect on MT polymerization, thereby impeding intracellular transport and protein/cell organelle trafficking, which led to a disturbance of TJ-permeability barrier function.

Figure 5.

Laminin α2 (Lam α2) knockdown in Sertoli cells induces MT disorganization mediated by changes in detyrosinated α-tubulin and EB1 distribution. A) After laminin α2 knockdown, α-tubulin—the building block of MTs—localization (red or gray) was considerably disorganized. Furthermore, organization of detyrosinated α-tubulin, which caused MTs to become stabilized, but less dynamic, across the Sertoli cell cytosol was also grossly affected. +TIP (plus-end tracking protein) EB1 localization was also grossly affected such that EB1 no longer associated with the long stretches of MTs but considerably diminished along the MTs. Sertoli cell nuclei were visualized by DAPI (blue). Expression of GFP (green), an integrated component of the transfection vector, illustrated successful transfection. Scale bar, 30 μm. B) MT polymerization assay was performed to quantify the ability of cell lysates to induce MT polymerization in Sertoli cells. Laminin α2 knockdown considerably reduced polymerized MTs found in the pellet [note: β-tubulins are constituents of MTs via their polymerization; for reviews, see refs. 73, 74; see also the bar graph (right)] but not in the supernatant, which quantified the free β-tubulins as monomers. Taxol and CaCl₂ served as the corresponding positive and negative controls, respectively. ctrl, control. Each bar graph is a mean ± sd of 6 independent MT polymerization experiments. **P < 0.01 by Student’s t test.

DISCUSSION

BTB requires extensive remodeling during the epithelial cycle to support the transport of preleptotene spermatocytes across the immunological barrier at stage VIII of the epithelial cycle in the rat vs. stage II in the human testis (for a review, see ref. 4). Studies have shown that the event of BTB restructuring that is essential to spermatogenesis is supported by timely disassembly of the old BTB behind the spermatocytes under transport at the immunological barrier and assembly of the new BTB above these germ cells (38, 39). This process involves intricate differential endocytosis, transcytosis, and recycling of integral membrane proteins and their associated adaptor proteins at the site (39, 40), as well as participation of signaling proteins, such as c-Src and c-Yes (41, 42), and testosterone, cytokines (e.g., TGF-β2, TGF-β3), and small GTPases (e.g., Cdc42) (39, 40, 43). Here, we report for the first time that there is a local axis that functionally links the BM and Sertoli cell BTB in which changes in the expression of laminin α2 chain, such as by using specific laminin α2 shRNA to knock down its steady-state level, can induce BTB restructuring. This facilitates transport of preleptotene spermatocytes across the immunological barrier. These findings are consistent with a recent report in which laminin α2^−/− mice displayed defective blood-brain barrier wherein tracer that was administered systemically was capable of penetrating the barrier, leaking into the brain parenchyma (44). In short, these findings illustrate that laminin α2 might play an important role in maintaining the barrier function in different blood-tissue barriers.

Laminin is a heterotrimeric glycoprotein composed of one α, β, and γ chain to form a functional protein, which usually serves as the putative ligand for integrin-based receptor and is mostly restricted to the cell-ECM interface to induce signaling function (for reviews, see refs. 45–47). Studies have shown that a peptide derived from domain IV of laminin γ3 chain (10), designated F5-peptide (6), induces aspermatogenesis, which is mediated via disruption of the underlying actin- and MT-based cytoskeletal network (48). These earlier findings thus demonstrate the presence of a local autocrine-based regulatory axis to coordinate cellular events that take place across the seminiferous epithelium in the testis. For instance, this axis coordinates the release of sperm at the apical ES during spermiation and the remodeling of the basal ES/BTB to support transport of preleptotene spermatocytes across the immunological barrier at stage VIII of the epithelial cycle (for a review, see ref. 13). These earlier findings led us to investigate whether laminin α2 that was found in the BM, which is analogous to the laminin chains at the apical ES, would be involved in the apical ES-BTB-BM functional axis. Our results support this notion that laminin α2 chain in the BM is an integral component of this functional axis. In this context, it is of interest to note that the BM that surrounds the seminiferous tubule is a modified form of ECM (49, 50). In addition to being found in the BM of seminiferous epithelium, laminin α2 chain is also a major component of the BM of muscle and brain, and its deficiency or mutation in humans is known to cause congenital muscular dystrophy, which is characterized by severe muscle weakness, inability to walk, an elevated creatine kinase in skeletal muscle, and characteristic white matter hypodensity on cerebral MRI (28, 51–53). In mice, deletion or spontaneous mutations of laminin α2 chain also lead to muscular dystrophy (54–57). Furthermore, laminin α2–deficient male mice are infertile. These knockout mice are found to have a down-regulation in expression of laminin γ3 chain [a component of the apical ES (8, 9)], defects in lumen formation, and defects at the apical ES as a result of a disrupted actin microfilament organization, which leads to a considerable reduction in the number of spermatids that are capable of adhering to the seminiferous epithelium in adult mutant mice, leading to infertility (17). Collectively, these other observations also suggest the presence of a possible physiological link between the BM and the ES in the testis.

Laminin α2 (merosin) is a ∼390-kDa protein that is composed of an N-terminal ∼300-kDa fragment and a C-terminal ∼80-kDa fragment (27, 58–60) and can be readily detected by antibodies prepared against either the N- or C-terminal epitope(s). In humans, laminin α2 is a 3119 aa polypeptide (vs. rat laminin α2, which is a 3165 aa polypeptide) that can be cleaved at the N terminus of the LG3 domain to generate a 300-kDa N-terminal vs. an 80-kDa C-terminal fragment (27). Because the expression of these two fragments is differentially regulated in patients with different severities of muscular dystrophy, it is highly likely that these fragments, once generated, may have different biological function(s) (58–60). Studies in other laminin chains have shown that biologically active laminin fragments are mostly generated via the action of MMPs and that these fragments are known to modulate cell adhesion, cell migration, blood-tissue barrier function, inflammatory response, junction assembly, cell differentiation, and others [for reviews, see refs. 28, 45, 47, 61). Sertoli cells that are cultured in vitro are known to secrete glycoproteins, including laminins, to form the BM (31, 62, 63). The 80-kDa laminin α2 fragment located near the C-terminal region of the ∼390 kDa laminin α2 chain which is possibly released via the action of MMP-9 at the BM and is perhaps modulated by TNF-α (64), which was earlier shown to activate MMP-9 in the rat testis—is possibly used to regulate BTB dynamics. This axis is likely an additional local regulatory axis between the BM and the Sertoli cell BTB to modulate BTB dynamics in addition to the apical ES-BTB axis that was identified earlier (6, 10, 65). This 80-kDa laminin α2 fragment was detected in lysates of testes vs. Sertoli and germ cells as reported herein; however, staining of laminin α2 by IHC and fluorescence microscopy likely represents the entire laminin α2 chain or a combination of the 390-kDa and the 80-kDa forms at the BM in virtually all stages of the epithelial cycle. However, at stage VIII to early IX, laminin α2 (or the 80-kDa fragment) was detected in the adluminal compartment when F-actin at the apical ES underwent extensive reorganization, which leads to apical ES degeneration. Laminin α2 (or its 80-kDa fragment) seemed to be transported to the apical ES via an MT-dependent mechanism as treatment of testis with taxol (paclitaxel), an MT-stabilizing chemical known to disrupt MT-dependent cellular transport function [for reviews, see refs. 66, 67), via intratesticular injection was found to block the transport of laminin α2 across the seminiferous epithelium. The stage-stage expression of laminin α2 in the adluminal compartment suggests that laminin α2 (or its 80-kDa fragment) may exert its effects by disrupting microfilament organization at the apical ES to facilitate spermiation. This notion is supported by findings that laminin α2 knock down in the Sertoli cell epithelium in vitro led to disruption of the Sertoli cell TJ-permeability barrier, which was mediated by disorganizing actin microfilaments across the Sertoli cell cytosol such as defragmentation of actin filaments. The bundling capability of actin microfilament was also considerably reduced through changes in the spatiotemporal expression of actin regulatory proteins Arp3—which together with Arp2 forms the Arp2/3 complex that is known to induce branched actin nucleation by effectively converting actin microfilaments into a branched configuration—and Eps8, an actin barbed end capping and bundling protein that confers actin microfilaments into a bundled configuration. Such mislocalizations of Arp3 and Eps8 thus impede the proper organization of actin microfilaments after a knockdown of laminin α2 chain, failing to support cell adhesion protein complexes, such as CAR/ZO-1 and N-cadherin/β-catenin which utilize F-actin for their attachment, thereby perturbing the TJ-permeability barrier. In this context, it is of interest to note that earlier studies have shown that dystroglycan serves as a link between ECM and F-actin to maintain epithelial homeostasis (68, 69). In the mouse testis, α/β-dytroglycan is localized in the BM of seminiferous tubules and is colocalized with laminin α2 (70). Thus, it is possible that laminin α2 knockdown perturbs F-actin organization via changes in dystroglycan, which should be carefully examined in future studies. Other than its effects on F-actin organization, a knockdown of laminin α2 was also shown to impede MT organization such that α-tubulins no longer stretch across the Sertoli cell cytosol but are retracted from the cell periphery. These changes likely are the result of changes in the spatiotemporal expression of EB1. Laminin α2 knockdown caused mislocalization of MT-stabilizing protein EB1, which is a known +TIP protein (plus-end tracking protein) (71, 72). This change thus perturbed MT organization. In short, our findings demonstrate that BTB integrity is maintained by laminin α2 chain and its knockdown impedes F-actin and MT organization, possibility involving signaling protein kinases, which will require additional investigations.

In summary, we have shown that laminin α2 chain in the BM is a crucial regulator of BTB dynamics, possibly via its effects—likely as an 80-kDa fragment—on the organization of actin microfilaments and MTs at the ES. These findings also support earlier findings and the idea that there is a local regulatory axis in the seminiferous epithelium to coordinate cellular events that take place simultaneously across the epithelium during the epithelial cycle of spermatogenesis (for reviews, see refs. 13, 73, 74).

Glossary

Arp3

actin-related protein 3

BM

basement membrane

BTB

blood–testis barrier

CAR

coxsackievirus and adenovirus receptor

EB1

end-binding protein 1

ECM

extracellular matrix

Eps8

epidermal growth factor receptor pathway substrate 8

ES

ectoplasmic specialization

IB

immunoblotting

IF

immunofluorescence

IHC

immunohistochemistry

MMP-2

matrix metalloprotease-2

MT

microtubule

PFA

paraformaldehyde

shRNA

short hairpin RNA

TER

transepithelial electrical resistance

+TIP

microtubule plus-end tracking protein

TJ

tight junction

ZO-1

zonula occludens 1

AUTHOR CONTRIBUTIONS

Y. Gao and C. Y. Cheng designed research; Y. Gao, H. Chen, and C. Y. Cheng performed research; D. Mruk, W.-Y. Lui, W. M. Lee, and C. Y. Cheng contributed new reagents and analytic tools; Y. Gao and C. Y. Cheng analyzed data; Y. Gao and C. Y. Cheng wrote the paper; Y. Gao and C. Y. Cheng performed in vivo animal experiments; Y. Gao and C. Y. Cheng prepared all figures; and all authors reviewed the manuscript.