Gonadotropins Regulate Rat Testicular Tight Junctions in Vivo

REPRODUCTION-DEVELOPMENT

Mark J. McCabe, Gerard A. Tarulli, Sarah J. Meachem, David M. Robertson, Peter M. Smooker, and Peter G. Stanton

Prince Henry’s Institute (M.J.M., G.A.T., S.J.M., D.M.R., P.G.S.), Monash Medical Centre, Clayton, Victoria 3168, Australia; School of Applied Sciences (M.J.M., P.M.S.), Royal Melbourne Institute of Technology University, Bundoora, Victoria 3083, Australia; and Department of Biochemistry and Molecular Biology (P.G.S.), Monash University, Clayton, Victoria 3800, Australia

Sertoli cell tight junctions (TJs) are an essential component of the blood-testis barrier required for spermatogenesis; however, the role of gonadotropins in their maintenance is unknown. This study aimed to investigate the effect of gonadotropin suppression and short-term replacement on TJ function and TJ protein (occludin and claudin-11) expression and localization, in an adult rat model in vivo. Rats (n ϭ 10/group) received the GnRH antagonist, acyline, for 7 wk to suppress gonadotropins. Three groups then received for 7 d: 1) human recombinant FSH, 2) human chorionic gonadotropin (hCG) and rat FSH antibody (to study testicular androgen stimulation alone), and 3) hCG alone (to study testicular androgen and pituitary FSH production). TJ proteins were assessed by real-time PCR, Western blot analysis, and immunohistochemistry, whereas TJ function was assessed with a biotin permeation tracer. Acyline treatment significantly reduced testis weights, serum androgens, LH and FSH, and adluminal germ cells (pachytene spermatocyte, round and elongating spermatids). In contrast to controls, acyline induced seminiferous tubule permeability to biotin, loss of tubule lumens, and loss of occludin, but redistribution of claudin-11, immunostaining. Short-term hormone replacement stimulated significant recoveries in adluminal germ cell numbers. In hCG Ϯ FSH antibody-treated rats, occludin and claudin-11 protein relocalized at the TJ, but such relocalization was minimal with FSH alone. Tubule lumens also reappeared, but most tubules remained permeable to biotin tracer, despite the presence of occludin. It is concluded that gonadotropins maintain Sertoli cell TJs in the adult rat via a mechanism that includes the localization of occludin and claudin-11 at functional TJs. (Endocrinology 151: 2911–2922, 2010)

he blood-testis barrier (BTB) is a permeability barrier, Formation of the BTB at puberty in animal models (2, Twhich functions to sequester germ cells undergoing 4) and humans (5) requires the gonadotropins FSH and meiotic and postmeiotic differentiation from the vascular LH (6, 7), a requirement that is also supported by in vitro environment. Endothelial, peritubular myoid, and Sertoli models (8–12). Once formed, Sertoli cell TJs undergo a cells contribute to the barrier (1), but it is recognized that cyclic remodelling during spermatogenesis, as early mei- tight junctions (TJs) found basally between Sertoli cells are otic preleptotene spermatocytes translocate to the adlu- the major component in the adult testis (1). Sertoli cell TJs minal compartment where they complete meiosis and are required for fertility, and it is well established that their spermiogenesis (2). An expanding list of cytokines and disruption leads to germ cell atresia and cessation of sper- growth factors, including TGF-␤, TNF-␣, growth differ- matogenesis (2). Changes in BTB permeability and constit- entiation factor-9, activins, inhibins, and interleukins, uent proteins have been noted in men with infertility (3). produced by both Sertoli and germ cells (10, 13–16; for

ISSN Print 0013-7227 ISSN Online 1945-7170 Abbreviations: BTB, Blood-testis barrier; FSH Ab, FSH antibody; hCG, human chorionic Printed in U.S.A. gonadotropin; hrec, human recombinant; JAM, junctional adhesion molecule; T, testos- Copyright © 2010 by The Endocrine Society terone; TJ, tight junction; ZO, zona occludens. doi: 10.1210/en.2009-1278 Received October 27, 2009. Accepted March 2, 2010. First Published Online March 31, 2010

Endocrinology, June 2010, 151(6):2911–2922 endo.endojournals.org 2911

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 22 January 2015. at 16:54 For personal use only. No other uses without permission. . All rights reserved. 2912 McCabe et al. Regulation of Rat Testicular TJs Endocrinology, June 2010, 151(6):2911–2922 review, see Ref. 17) mediate the local cross talk required Experimental design for this translocation step. Circulating FSH and LH/testosterone (T) were suppressed by Sertoli cell TJs contain three main transmembrane pro- weekly sc injections of the GnRH antagonist, acyline (donated by Richard Blye; National Institutes of Health, Bethesda, MD) for tein groups, the claudins, particularly claudin-11 (18, 19) 7 wk into adult rats at 1.5 mg/kg in 5% mannitol/sterile MilliQ and claudin-3 (6), occludin (20, 21), and the junctional water (34), administered as single injections to the hind flank or adhesion molecule (JAM) family (22). Both androgens and shoulder (400 ␮l/site). Control rats received a sc injection of the FSH can regulate these proteins during TJ formation in vehicle alone (n ϭ 10). ϭ vivo (6, 22) and in vitro (9–11). However, the extent to After 7 wk of gonadotropin suppression, rats (n 10/group) in the eighth week received an additional dose of acyline in con- which gonadotropins contribute to the maintenance of the junction with short-term hormone replacement (daily, for 7 d) by established BTB in the normal adult remains unknown. sc injections with one of the following: 1) human recombinant Animal studies that have previously ablated gonadotropin (hrec) FSH (25 IU/kg) (Puregon; Organon, Oss, The Nether- action on the testis via hypophysectomy (23–25), GnRH lands), 2) human chorionic gonadotropin (hCG) (2.5 IU/kg) ϩ antagonism (26, 27), or GnRH immunization (28, 29) (Pregnyl; Organon) FSH antibody (FSH Ab) (sheep antirat, 2 mg/kg) (29) to study testicular production of androgens alone, have either not examined BTB function or are inconclusive and 3) hCG (2.5 IU/kg) ϩ normal sheep immunoglobulins (2 on this point. Regulation of the BTB has potential impor- mg/kg) to study the production of androgens and endogenous rat tance in understanding the mechanism(s) of action of male FSH (29). Three more groups of rats (n ϭ 10/group) continued hormonal contraception, which suppresses spermatogen- to receive 1) acyline vehicle, 2) acyline ϩ FSH Ab (daily, sheep ϩ esis via suppression of circulating gonadotropins (for re- antirat, 2 mg/kg) to suppress residual FSH, and 3) acyline normal sheep immunoglobulins (daily, 2 mg/kg) for the eighth view, see Ref. 30). week. Doses and treatment intervals were as described elsewhere Precedents exist for a regulatory role for gonadotropins (29, 35). Previous studies demonstrate that endogenous rat FSH on Sertoli cell TJs in adult animals. In the seasonal breed- levels rise under the influence of T stimulation (28, 36, 37). ing, adult Djungarian hamster, circulating gonadotropins, At the end of the experiment, rats were killed by CO2 as- and spermatogenesis are both suppressed during the non- phyxiation, and testes were excised and weighed immediately. Five of the left testes per treatment group were snap-frozen in dry breeding winter months when the BTB is nonfunctional ice for RT-PCR and Western blot analyses. The remaining five (31). We recently demonstrated that claudin-3 and clau- left testes from each group were injected with either the quali- din-11, occludin, and JAM-A were no longer present at tative TJ functional tracer biotin (10 mg/ml, EZ-Link Sulfo- organized Sertoli cell TJs in these animals and that admin- NHS-LC-Biotin; Pierce, Rockford, IL) (n ϭ 4) (6) in a volume of 10% testis weight, or the biotin vehicle, PBS (pH 7.4), and 1 istration of FSH restored BTB functionality and localiza- ϭ mM CaCl2 (n 1). Testes were then incubated at room tem- tion of TJ proteins at Sertoli cell TJs (7, 32). It remains to perature for 30 min before immersion fixation in Bouin’s fixative be seen whether a similar gonadotropin dependence of TJs for 5 h and were then randomly sampled by a systematic uniform occurs in animal models that do not undergo seasonal approach (29, 38) for stereological or immunohistochemical changes. assessment. Shortly after killing rats, blood was taken by cardiac puncture The aim of this study was to profoundly suppress cir- and stored overnight at 4 C before being spun down in a cen- culating gonadotropins and spermatogenesis in adult male trifuge. Serum was then collected by centrifugation and snap- rats with the GnRH antagonist acyline (33, 34) and then frozen before storage at Ϫ80 C for hormone analysis. selectively replace testicular androgens and/or FSH via short-term replacement regimens (29). Changes in germ Immunohistochemistry cell numbers were quantified by stereology, whereas TJ Tissues were embedded in paraffin wax, and 5-␮m sections function was assessed using a qualitative biotin perme- were mounted onto Superfrost-Plus slides (HD Scientific, Mel- ation tracer (6). Changes in TJ proteins were examined bourne, Australia). Antigen retrieval was performed by heating using immunohistochemistry, real-time RT-PCR, and the sections for 10 min in 600 ml 1 mM EDTA-NaOH (pH 8.0) Western blot analysis. (39) in an 800-W microwave. Primary antibodies [rabbit anti- occludin (no. 71-1500, 2.5 ␮g/ml, 30 min; Zymed, South San Francisco, CA), rabbit anti claudin-11 (no. 36-4500, 2 ␮g/ml, 1 h; Zymed), rabbit anti zona occludens (ZO)-1 (no. 61-7300, 1.25 ␮g/ml, 1 h; Zymed), and rabbit anti JAM-A (no. 36-1700, Materials and Methods 0.63 ␮g/ml, overnight; Zymed)] were diluted in 10% normal goat serum (Chemicon International, Temecula, CA) in PBS and Animals then applied to the sections for the times specified at 25 C. Neg- Male outbred Sprague Dawley rats at 75–90 d of age obtained ative controls used an equivalent concentration of nonspecific from Monash University Animal Services were maintained at 20 IgG or normal serum in place of the primary antibody. Detection C in a fixed 12-h light, 12-h dark cycle with free access to food was with either rabbit antigoat Alexa-488 or Alexa-546 (Mo- and water. All animal experimentation was approved by the lecular Probes, Eugene, OR) at dilutions of 1:400 for claudin-11 Monash Medical Centre Animal Ethics Committee. and JAM-A or 1:800 for occludin and ZO-1. In some cases, cells

TABLE 1. Primer-specific conditions used for quantitative PCR analysis

Species Size Mg2؉ Anneal Read a (Gene (source) accession no. Primer sequence (5؅-3؅) (bp) (mM) temperature (C) temperature (C Claudin-11 (70) Rat F TTAGACATGGGCACTCTTGG 624 2.5 68 85 NM࿝053457 R ATGGTAGCCACTTGCCTTC Occludin (13) Rat F CTGTCTATGCTCGTCATCG 294 2.5 64 75 NM࿝031329 R CATTCCCGATCTAATGACGC JAM-A (Primer 3) Mouse F TATAGCCGTGGATACTTT 425 2.5 64 72 NM࿝172647 R GTCAGCCCAGCCACTATTACT ZO-1 (Primer 3) Rat F AGAGAGGAAGAGCGAATGTCTA 248 3.0 60 72 NM࿝009386 R TTCATGCTGGGCCTAAGTA INSL-3 (Primer 3) Rat F CTTCCTCACCAGGCTTCTCA 241 2.0 58 72 NM࿝053680 R CTGTGGTCCTTGGTTTCTGG ␤-actin (Primer3) Rat F CCGTAAAGACCTCTATGCCAACA 103 2.0 67 72 NM࿝031144 R GCTAGGAGCCAGGGCAGTAATC GAPDH (71) Mouse F GACCCCTTCATTGACCTCAAC 560 2.5 60 72 NM࿝008084 R GATGACCTTGCCCACAGCCTT F, Forward; R, reverse. a Temperature at which the fluorescence of the PCR product was quantified during LightCycler analysis. were counterstained with the fluorescent nuclear marker 4Ј,6- DNA was removed with a deoxyribonuclease-free kit (Ambion, diamidino-2-phenylindole (Molecular Probes) (100 nM in PBS) Austin, TX), following the manufacturers’ instructions. RT used for 5 min after the secondary antibody incubation. Sections were avian myeloblastosis virus-reverse transcriptase (Roche, Basel, then washed in PBS, mounted with FluorSave (Calbiochem, La Switzerland) and random hexamers as described elsewhere (11). Jolla, CA), and visualized with a conventional fluorescent mi- mRNA expression was quantified using the Roche Light Cy- croscope (Olympus BX-50; Olympus, Mt. Waverley, Australia). cler 380 (Roche) and the FastStart DNA Master Sybr Green 1 systems (Roche). Oligonucleotide primer pairs were obtained Qualitative TJ functional analysis from published sources or were designed using the Primer3 pro- The localization of the biotin tracer in tissue sections was gram (available ar http://frodo.wi.mit.edu/primer3/) and were visualized directly by fluorescence microscopy using streptavidin ordered from Sigma Genosys (Castle Hill, Australia). Primer de- Alexa-488 (Molecular Probes). For colocalization studies with tails and PCR conditions, including anneal temperature and 2ϩ occludin, the probe was diluted 1:100 in PBS along with the goat Mg concentration, are presented in Table 1. Primers produced antirabbit Alexa-546 secondary antiserum. a single band on DNA agarose gels, which corresponded to the target protein as shown by DNA sequencing. Standard curves for Western blot analysis PCR analyses were generated using dilutions of an adult rat testicular cDNA preparation of arbitrary unitage. Unless otherwise Protein was extracted from frozen rat testes using 1% sodium noted, PCR of all samples was performed using triplicate reac- dodecyl sulfate (MP Biomedicals, Solon, OH) (16) and quanti- tions for 38 cycles, after which a melting curve analysis was fied using the bicinchoninic acid protein assay (Pierce). Protein performed to monitor PCR product purity (see Table 1). (40 ␮g for occludin, 120 ␮g for claudin-11) was boiled in Lae- mmli sample buffer, separated by SDS-PAGE on 4–20% precast gels (Bio-Rad, Hercules, CA), and transferred onto poly(vinyli- Serum hormone assays dene difluoride) membranes (Immobilon-P; Millipore, Billerica, Serum androgens, rat FSH, and rat LH concentrations were MA) for Western blot analysis. Membranes were blocked in 5% determined as described previously (29, 35, 40, 41). The cross skim-milk/PBS, 4 C overnight. Primary antibodies [rabbit anti- reactivities of hCG in the FSH and LH immunofluorometric occludin (no. 71-1500, 1:125; Zymed) or rabbit anticlaudin-11 assays were approximately 1 and 4%, respectively (42, 43), (no. 7474, 1:200; Abcam, Cambridge, UK)] were applied to the and assay sensitivities were 0.013 ng/ml (serum androgens), membranes for 2 h with the latter costained with rabbit anti-␤- 0.015ng/ml (rat FSH), and 0.012 ng/ml (rat LH). actin (no. A-2066, 1:1000; Sigma Chemical Co., St. Louis, MO). Negative controls substituted primary antibody with an equiv- Stereological assessment of germ cell type and alent concentration of nonspecific IgG or normal serum. Band number detection was with IRDye 800-conjugated goat antirabbit sec- The total number of cells per testis was calculated using the ondary antibody (1:5000; Rocklands Immunochemicals, Inc., optical disector method as previously described (29, 38). Germ cells Gilbertsville, PA), and membranes were scanned using a near- were identified using the criteria of Russell et al. (44) as published infrared imaging system (Odyssey infrared imaging scanner; Li-Cor previously (45), and approximately 100 germ cells of each Biosciences, Lincoln, NE). type were counted per animal. Germ cells were counted in the following subgroups: type A/intermediate spermatogonia, type Total RNA extraction, RT, and quantitative B/preleptotene spermatocytes, leptotene/zygotene spermato- real-time PCR cytes, pachytene spermatocytes (associated with stages I–VIII Total RNA was extracted from frozen rat testes using the RNA and IX–XIV), round spermatids (steps 1–8), elongating sperma- Isolation kit (QIAGEN, Hildens, Germany), and contaminating tids (steps 9–14), and elongated spermatids (steps 15–19).

TABLE 2. Effect of GnRH antagonist acyline treatment (8 wk) and short-term (1 wk) hormone replacement on body and testis weights, and serum hormones (androgens, rat LH, and rat FSH)

Body weight Testis weight Serum LH Serum androgens Serum FSH (g) (mg) (ng/ml) (ng/ml) (ng/ml) Control 439 Ϯ 38 (10) 1755 Ϯ 91a (10) 2.37 Ϯ 1.40a (7) 2.22 Ϯ 1.22a (10) 5.55 Ϯ 1.01a (7) Acyline 403 Ϯ 22 (10) 332 Ϯ 44b (10) 0.38 Ϯ 0.21b (8) 0.28 Ϯ 0.10b (10) 0.68 Ϯ 0.23b (8) Acyline ϩ FSHAb 402 Ϯ 37 (10) 299 Ϯ 47b (10) 0.12 Ϯ 0.01c (5) 0.33 Ϯ 0.15b (10) na Acyline ϩ hrecFSH 416 Ϯ 40 (10) 426 Ϯ 33c (10) 0.32 Ϯ 0.19b (6) 0.31 Ϯ 0.08b (9) 0.40 Ϯ 0.16c (6) Acyline ϩ hCG ϩ FSHAb 413 Ϯ 14 (10) 433 Ϯ 54c (10) 0.020 Ϯ 0.007c (6) 0.78 Ϯ 0.39c (9) na Acyline ϩ hCG 419 Ϯ 20 (10) 482 Ϯ 74c (10) 0.35 Ϯ 0.25b (5) 0.67 Ϯ 0.26c (10) 1.36 Ϯ 0.27d (6)

Data are mean Ϯ SD (n/group). Different letters denote significant differences between groups at P Ͻ 0.05 or greater, as detailed in Results. na, Not assayed.

Cell counting was performed with a ϫ100 objective lens on short-term replacement with hCG ϩ FSH Ab resulted in a an Olympus BX-50 microscope (Olympus, Tokyo, Japan), using further significant decrease of rat LH (P Ͻ 0.05, Table 2). a systematic uniform random sampling method as described elsewhere (29). A set of unbiased frames was used to count the germ cells, with the number of frames for each germ cell type being Androgens determined by their frequency (29). The final screen magnifica- Acyline treatment reduced serum total androgens to 13% tion was ϫ2708, and no correction for tissue shrinkage was (P Ͻ 0.001) of control, and the addition of the FSH Ab necessary as determined previously (29, 45, 46). caused no further decrease (Table 2). Short-term hormone replacement with hrecFSH had no effect on serum androgen Statistical analysis levels, but significant (P Ͻ 0.01) recoveries were achieved Treatments were compared with controls by ANOVA fol- with short-term replacement of hCG (to 30% of control, P Ͻ lowed by the Student’s Newman-Keuls test, or where data were ϩ Ͻ nonparametric, Kruskal-Wallis test, followed by Newman- 0.01) and hCG FSH Ab (to 35% of control, P 0.001). Keuls analog (Equal NЈs) test. P Ͻ 0.05 was used to determine whether results were statistically significant. All statistics were FSH performed using SigmaStat version 3.5 (Systat Software, Inc., Serum FSH was reduced to 12% (P Ͻ 0.01) of controls Ϯ San Jose, CA), and data have been expressed as mean SD with in response to acyline treatment (Table 2). As the FSH Ab n ϭ 5–10 rats per group unless otherwise stated. interfered with the assay (data not shown), no measure- ments were possible in FSH Ab-containing treatment groups. However, the short-term replacement of hrecFSH Results further decreased serum rat FSH levels to 7% of control Body and testis weights and serum hormone (P Ͻ 0.05), whereas the addition of hCG induced a sig- levels nificant recovery to 25% of control (P Ͻ 0.001). Hormone suppression with acyline for 8 wk had no effect on body weights but induced a significant (P Ͻ Changes to germ cell types and numbers 0.001) decrease in testis weights (Table 2) to 19% of con- The effects of chronic GnRH antagonist treatment on trol. The addition of the FSH Ab in the eighth week caused spermatogenesis in the rat are well understood (26, 47). a further 10% decrease in testis weights compared with Because this study used the more recently developed an- acyline alone, but this was no more significant than acyline tagonist, acyline (33, 34), for which quantitative data on alone. germ and Sertoli cell numbers is not yet available, we first Short-term hormone replacement with hrecFSH, hCG ϩ assessed changes in cell populations in response to the FSH Ab, or hCG ϩ control IgG all gave significant (P Ͻ various treatment regimes. 0.001) recoveries in testis weights to 24–27% of controls (Table 2). Spermatogonia and preleptotene spermatocytes Spermatogonia and preleptotene spermatocytes reside LH outside the BTB in the normal animal. Hormone suppres- Acyline treatment significantly decreased rat serum LH sion with acyline for 8 wk significantly (P Ͻ 0.001) re- to 16% (P Ͻ 0.05) (Table 2) of control, and addition of the duced type A and intermediate spermatogonia to 66% of FSH Ab in the eighth week resulted in a further decrease controls, and addition of the FSH Ab caused a further in serum LH to 5% of controls (P Ͻ 0.05 compared with significant reduction compared with acyline alone (P Ͻ acyline alone). Short-term hrecFSH and hCG replacement 0.05) to 46% of controls (Fig. 1). Type B/preleptotene had no effect on suppressed serum LH levels, whereas spermatocytes were similarly suppressed by the acyline

A D 0.001 vs. control, P Ͻ 0.01 vs. acyline alone) (Fig. 1). 20 100 None of the hormone replacement groups resulted in any

a a significant increases compared with acyline alone. 15 75 ) 6 b d Pachytene spermatocytes stages I–VIII er (x 10 10 50 c c f c c Pachytene spermatocytes in stages I–VIII were sup- e ϩ 5 25 pressed by acyline to 15% or by acyline FSH Ab to 8% b

Germ cell numb b (both P Ͻ 0.001) of controls (Fig. 1). Short-term re- 0 0 placement produced significant increases in all treatment B E groups (hrecFSH, hCG ϩ FSH Ab, and hCG, all P Ͻ 0.001 40 100 ac vs. acyline), although animals receiving FSH (either a a ab hrecFSH or via hCG) recovered to a greater extent than the ab ) 6 30 b hCG ϩ FSH Ab group (hrecFSH vs. hCG ϩ FSH Ab, P Ͻ b 10 d e e 0.001; hCG vs. hCG ϩ FSH Ab, P Ͻ 0.001). The increase 20 b

ll number (x 10 in pachytene spermatocyte (stages I–VIII) number in the c 1 Ͻ 10 hrecFSH-treated group was significantly (P 0.001) Germ ce greater than in the hCG-alone group (Fig. 1).

0 0.1 CFPachytene spermatocytes stages IX–XIV 48 1000 Acyline treatment reduced pachytene spermatocytes at a a stages IX–XIV to 3% of control (P Ͻ 0.001), with a further 36 )

6 significant suppression to 2% of control numbers with the 100 bd bd FSH Ab treatment (P Ͻ 0.05 vs. acyline) (Fig. 1, note log b bd 24 c d scale). As seen for pachytene spermatocytes in stages c 10 b cell number (x 10 cell I–VIII, short-term hormone replacement produced signif- 12 b

Germ b icant 3- to 5-fold increases in all treatment groups for

0 1 stages IX–XIV pachytene spermatocytes, with the group Control Acyline Control Acyline + + receiving hrecFSH significantly greater than either hCG ϩ - FSH FSHhCG hCG - FSH FSHhCG hCG Ab + Ab + FSH Ab or hCG-alone groups. FSH Ab FSH Ab FIG. 1. Stereological assessment of the number of germ cells (millions/ testis) counted in the following subgroups: A, Type A/intermediate Round spermatids steps 1–8 spermatogonia; B, Type B/preleptotene spermatocytes; C, Leptotene/ Round spermatids at steps 1–8 responded similarly to the zygotene spermatocytes; D, Pachytene spermatocytes stages I–VIII; various treatments as pachytene spermatocytes, although a E, Pachytene spermatocytes stages IX–XIV; and F, Round spermatids steps 1–8. Spermatogenesis was suppressed by GnRH antagonist greater extent of initial suppression with acyline (0.4% of treatment (acyline, 7 wk), then restored over a further 7-d period with control numbers) was observed (Fig. 1). Short-term hrecFSH acyline and either hrecFSH or hCG in combination with a polyclonal replacement resulted in a partial but significant (P Ͻ 0.05) sheep antiserum raised against rat FSH (for details, see Materials and Methods). Data are mean Ϯ SD,nϭ 5 rats/group. Significant recovery of round spermatid numbers to 6.5% of controls differences (P Ͻ 0.05 or greater) between treatment groups are (Fig. 1), whereas androgen alone (hCG ϩ FSH Ab group) denoted by different letters, with full details given in the text. Note the resulted in a nonsignificant increase to 1.4% of control num- logarithmic y-axes for the two lower right graphs. bers. However, the combination of endogenous rat FSH and androgen due to hCG treatment significantly increased and FSH Ab treatment (Fig. 1). Short-term FSH replace- round spermatids to 3.1% of control, but this was signifi- ment was the only treatment that significantly restored cantly (P Ͻ 0.05) less than the number obtained with germ cell numbers in the type B/preleptotene spermatocyte hrecFSH alone (hrecFSH vs. hCG P Ͻ 0.05) (Fig. 1). group to control levels (P Ͻ 0.001 vs. FSH Ab) (Fig. 1). Elongating spermatids steps Leptotene/zygotene spermatocytes Elongating spermatids at steps 9–14 were suppressed Leptotene/zygotene/pachytene spermatocytes and later by acyline to 0.2% of control (P Ͻ 0.001), with no further spermatids all reside inside the BTB in the normal animal. reduction obtained with the FSH Ab. None of the hor- Acyline reduced leptotene/zygotene spermatocyte num- mone replacements gave any recovery in these germ cell bers to 50% of control (P Ͻ 0.001), and the addition of the numbers (Table 3). Elongated spermatids at steps 15–19 FSH Ab caused a further decrease to 34% of control (P Ͻ were not detected in any of the treatment groups (Table 3).

TABLE 3. Stereological data for elongating and elongated spermatids (steps 9–14 and 15–19) and Sertoli cell numbers

Elongating spermatids, Elongated spermatids, Sertoli cells steps 9–14 (cells/testis) steps 15–19 (cells/testis) (cells/testis) Ccontrol 9.45 ϫ 107 Ϯ 0.84 ϫ 107a 3.63 ϫ 108 Ϯ 0.59 ϫ 108a 4.11 ϫ 107 Ϯ 0.31 ϫ 107 Acyline 1.42 ϫ 105 Ϯ 1.93 ϫ 105b Ͻ0.01 ϫ 108b 4.44 ϫ 107 Ϯ 0.88 ϫ 107 Acyline ϩ FSHAb 0.24 ϫ 105 Ϯ 0.35 ϫ 105b Ͻ0.01 ϫ 108b 4.71 ϫ 107 Ϯ 1.51 ϫ 107 Acyline ϩ hrecFSH 3.69 ϫ 105 Ϯ 4.10 ϫ 105b Ͻ0.01 ϫ 108b 5.18 ϫ 107 Ϯ 1.17 ϫ 107 Acyline ϩ hCG ϩ FSHAb 1.22 ϫ 105 Ϯ 2.42 ϫ 105b Ͻ0.01 ϫ 108b 5.54 ϫ 107 Ϯ 1.64 ϫ 107 Acyline ϩ hCG 3.32 ϫ 105 Ϯ 4.04 ϫ 105b Ͻ0.01 ϫ 108b 5.18 ϫ 107 Ϯ 1.15 ϫ 107

Quantification was conducted as per the earlier germ cell types presented in Fig. 2. Data are mean Ϯ SD, n ϭ 5 rats/group. Different letters denote significant differences between groups at P Ͻ 0.05 or greater, as detailed in Results.

Sertoli cell numbers Fig. 3, A and B). In addition, occludin immunostaining Sertoli cell numbers were unchanged with hormone was absent in acyline-treated animals, and tubules were suppression or short-term replacement compared with the permeable to the biotin tracer that surrounded all remain- control (Table 3). ing germ cell types (Fig. 3B). Based on stereological analysis, these cells were spermatogonia, leptotene/zygotene Effect of hormone suppression and short-term spermatocytes, and some (3–15%) pachytene spermato- replacement on TJ structure/localization and cytes (Fig. 1). A similar phenotype was observed for ani- function mals that received acyline plus the FSH Ab (Fig. 3C). In normal adult rats where the BTB is intact, perme- Short-term hormone replacement for 7 d with either ation of the biotin tracer was restricted to the interstitial hrecFSH (Fig. 3D), hCG ϩ FSH Ab (Fig. 3E), or hCG alone space in the testis (Fig. 2A, arrowheads) and only migrated (Fig. 3F) resulted in increased tubule and lumen diameters into the tubules as far as the intact TJs, as visualized by with some occludin staining at the TJ in hCG Ϯ FSH Ab- occludin immunostaining (arrows in Figs. 2, A and B, and treated animals but minimal staining in the hrecFSH group 3A). No biotin staining was detectable in the vehicle con- (Fig. 3, D–F, insets). The biotin tracer generally permeated trol (Fig. 2B), but some nonspecific staining of elongated spermatid heads and interstitial cells was seen with the throughout the tubules (Fig. 3, D–F), but in a low pro- Ͻ occludin antibody (Fig. 2B, inset). In the prepubertal 9-d- portion ( 5% of tubules/cross-sections observed), the old rat testis in which the BTB has yet to form, biotin tracer tracer was predominantly restricted to the interstitium, extensively permeated throughout the testis, including the and occludin immunostaining was more readily apparent seminiferous epithelium, but no occludin immunostaining (Fig. 3, G–I). This latter phenotype (Fig. 3, G–I) was not was discernible (Fig. 2C). observed in the acyline Ϯ FSH Ab hormone-suppressed Acyline treatment markedly altered seminiferous tu- groups. bule morphology, because tubule diameters were sup- The localization of the TJ proteins claudin-11 and pressed and tubule lumens were almost absent (compare JAM-A, and the cytoplasmic plaque protein ZO-1, was

FIG. 2. Assessment of TJ functionality using biotin tracer. A and B, TJs in adult rat testis sections were visualized by staining for the TJ protein occludin (red, indicated by arrows) in conjunction with the nuclear counterstain 4Ј,6-diamidino-2-phenylindole (blue). A, The TJ functional tracer biotin (green) remained localized to the interstitial space and only permeated around cells on the basal side of TJs in the seminiferous epithelium (arrowheads). B, The animal received the vehicle control for the biotin tracer. C, Biotin tracer permeation in a prepubertal (9 d) rat testis that lacks a functional BTB. Nonspecific staining (insets) was assessed by omitting streptavidin Alexa-488 from the probing solutions (A) or by substituting the occludin antibody with a nonspecific rabbit IgG at similar concentration (B). Scale bar,50␮m.

FIG. 3. Immunofluorescence microscopy of occludin (red, indicated by arrows) and biotin functional tracer (green) in testes from rats given hormone suppression and replacement. TJ structure and function was initially visualized in a normal control testis (A) and compared with testes which had been hormonally suppressed with acyline alone for 8 wk (B) or with acyline for 8 wk plus the FSH Ab in the final eighth week (C). After hormone suppression, the effect of short-term hormone replacement with hrecFSH alone (D), hCG ϩ FSH Ab to investigate the effect of androgen production alone (E), and hCG to stimulate both androgens and endogenous rat FSH (F) on TJ function and structure was analyzed. G–I, A much smaller proportion of TJ phenotypes observed with each of the short-term hormone replacement treatments listed above, respectively. Such phenotypes were not observed in acyline Ϯ FSH Ab treatments alone. Insets are three times enlarged portions of detail in the same panel. Scale bar,50␮m. also altered after hormone suppression (Fig. 4). In control and ZO-1 was concentrated more basally in hCG-treated animals, staining of claudin-11, JAM-A, and ZO-1 was animals Ϯ FSH Ab (Fig. 4, arrows). predominantly basal, consistent with the localization of intact TJs (Fig. 4, arrows). Unlike occludin, immunostain- Western blot analysis of occludin and claudin-11 ing for all three proteins remained observable in acyline- expression treated and acyline ϩ FSH Ab-treated animals but was Protein extracts from control rat testis displayed bands predominantly observed both at membranes adjacent to corresponding to the expected molecular masses of occlu- germ cells (Fig. 4, asterisks) and as a diffuse pattern in din, which appeared as a doublet at approximately 64 kDa Sertoli cell cytoplasm, indicative of a change in distribu- (Fig. 5A, lane 1), claudin-11 (ϳ22 kDa; Fig. 5B, lane 1), tion of the protein. There did not appear to be any major and ␤-actin (43 kDa; Fig. 5B). Fainter bands were also change in claudin-11 immunolocalization in the hrecFSH- present in each of the gels but were also detectable in the treated group (Fig. 4), but staining for claudin-11, JAM-A, negative controls. Hormone suppression resulted in a

FIG. 5. Effect of hormone suppression and replacement on TJ protein expression. Western blot analysis of occludin (64 kDa, 40 ␮g total protein per lane, reducing conditions) (A) and claudin-11 (22 kDa, 120 ␮g total protein per lane, nonreducing conditions) (B). Protein samples were from single animals from each treatment group. The housekeeping protein ␤-actin (43 kDa) was also probed in each lane on the claudin-11 gel. Lanes are as follows: 1, Control; 2, acyline; 3, FIG. 4. Effect of hormone suppression and replacement on the acyline ϩ FSH Ab; 4, acyline ϩ FSH; 5, acyline ϩ hCG ϩ FSH Ab; 6, localization of claudin-11, JAM-A, and ZO-1 by immunofluorescence acyline ϩ hCG; 7, adult rat testis positive control; and 8, adult rat testis microscopy to the testicular TJ. Inset in control panels shows negative negative control, with nonimmune IgG (occludin) or rabbit serum control for each TJ protein. Scale bar,50␮m. Arrows, Basally located (claudin-11) at matching concentrations. staining of TJ proteins; *, TJ protein staining adjacent to germ cells.

pression of the testicular androgen-regulated protein in- marked decrease in occludin protein expression (Fig. 5A, sulin-like factor 3 (INSL-3) was suppressed 12-fold after lanes 2 and 3), and no changes were observed after short- hormone suppression (Fig. 6E, note log y-axis), and was term hormone replacement (Fig. 5A, lanes 4–6). Simi- restored to control levels by androgen treatment groups larly, claudin-11 protein expression was suppressed to be- (hCG and hCG ϩ FSH Ab) but not hrecFSH (Fig. 6E). yond the limit of detection after hormone suppression ␤-Actin mRNA expression remained unchanged after hor- (Fig. 5B), with no detectable recovery after any of the mone suppression and replacement compared with con- short-term hormone replacement regimens. ␤-Actin pro- trol (Fig. 6F). tein expression did not change with any of the treatments compared with control (Fig. 5B). Discussion Real-time PCR analysis of TJ mRNA expression Hormone suppression with acyline resulted in signifi- This study has established that chronic gonadotropin sup- cant (P Ͻ 0.001) 3- to 4-fold increases in claudin-11, oc- pression in adult rats caused the BTB to become permeable cludin, and JAM-A mRNA expression compared with to a low molecular weight tracer in association with pro- controls (Fig. 6, A–C). Short-term hormone replacement tein redistribution (claudin-11) or complete loss (occlu- with each of the treatments resulted in a significant (P Ͻ din) from the basally located Sertoli cell TJs. Short-term 0.001) decrease in both claudin-11 and occludin mRNA hormone replacement with either FSH and/or androgens expression levels relative to the acyline group, although stimulated a partial reappearance of occludin and clau- there were no differences between groups (FSH alone, din-11 basally and significantly increased the numbers of hCG ϩFSH Ab, or hCG). In contrast, JAM-A levels re- adluminal meiotic and postmeiotic germ cells; however, mained high in all treatment groups (Fig. 6C). Hormone the BTB remained permeable to the tracer. Collectively, suppression and replacement had no effect on ZO-1 these data indicate that the BTB in the adult rat requires mRNA expression compared with controls (Fig. 6D). Ex- ongoing gonadotropic support to remain functional, due

Acyline did not alter body weights, as observed elsewhere (34, 49, 50); however, testis weights were decreased to 19% of control, consistent with suppression studies using a GnRH im- munogen [19–29% (28, 29, 35)]. Acyline also significantly suppressed serum rat LH, FSH, and androgen levels to 12–16% of control val- ues, mirroring other findings with this antagonist in rats (34, 48, 51–54) and also similar to the effects of active GnRH immunization on LH and androgens (29, 35). Interestingly, acyline treatment in this study brought about a 3- to 4-fold greater suppression in serum FSH than GnRH immunization [12 vs. 34–54% of controls (29, 35)], indicating a greater suppres- sive action for the antagonist regime. Further possible decreases in serum FSH in treatment groups that received FSH Ab were not able to be analyzed, but significant decreases in germ cell numbers suggested that an additional suppres- sive effect had occurred. Short-term hormone replacement had differential effects on the serum hormones stud- ied. Administration of hrecFSH to acyline- treated animals caused a further significant reduction in endogenous rat FSH, presumably via negative feedback or via a GnRH-independent pathway (36, 37). Treatment with hCG increased serum androgens (3-fold), but these remained at 30–35% of control, whereas rat

FIG. 6. Effect of hormone suppression and replacement on TJ mRNA expression. FSH levels increased 2-fold to approximately Real-time PCR quantitation was conducted for the TJ proteins claudin-11 (A), 25% of control levels. Differences in the effec- occludin (B), and JAM-A (C), as well as the cytoplasmic plaque protein, ZO-1 (D). As tiveness of hCG in elevating androgens and rat a positive control for hormone suppression and replacement, the androgen- FSH between this and previous GnRH immu- regulated testicular gene INSL-3 (note log scale; E) was also assessed. All data are presented as a ratio to the housekeeper protein, glyceraldehyde-3-phosphate nization studies (29, 35) are most probably due dehydrogenase (GAPDH). To check for any changes in GAPDH mRNA across the to the greater level of testicular suppression ␤ various treatment groups, a second housekeeper, -actin (F), was also quantitated. achieved with acyline. It is noted that all hor- Data are mean Ϯ SD,nϭ 5/group. Differences between letters are P Ͻ 0.001. mone replacement treatments induced similar significant increases in testis weights, consistent in part to a regulated localization of TJ proteins to the with previous studies where this, or similar models, have Sertoli cell TJs and, therefore, establishes a new site of been employed (28, 35). gonadotropin action in the adult rat testis. In addition, we show that spermatogenic restoration through meiotic and postmeiotic cell stages can occur even when Sertoli cell TJs Germ cell numbers and BTB function remain permeable to a low molecular weight tracer, sug- The major changes in germ cell numbers observed after gestive of a size-selectivity function of the Sertoli cell TJ. GnRH antagonism occurred in the cell types normally located adluminal to the BTB (leptotene/zygotene sper- Testis weights and serum hormones matocytes, pachytene spermatocytes at stages I–VIII and The gonadotropins LH and FSH were chronically sup- IX–XIV, and round/elongating spermatids) with only pressed for 8 wk in this study using the GnRH antagonist moderate decreases in basally located germ cells (type acyline, which is comparable with other GnRH antago- A/intermediate spermatogonia and type B/preleptotene nists in terms of the extent of suppression of LH and T and spermatocytes). The 10- to 50-fold decreases observed for its effectiveness after multiple weekly doses (33, 34, 48). pachytene spermatocytes and round spermatids are con-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 22 January 2015. at 16:54 For personal use only. No other uses without permission. . All rights reserved. 2920 McCabe et al. Regulation of Rat Testicular TJs Endocrinology, June 2010, 151(6):2911–2922 sistent with the known requirement for FSH and androgen JAM-A) observed after gonadotropin antagonism repre- to maintain spermatocyte survival and spermatid differ- sents similar recycling and/or degradation processes. entiation, respectively (29, 55, 56). All short-term hor- Short-term hormone replacement with FSH Ϯ andro- mone replacements stimulated 3- to 4-fold recoveries in gen for 7 d led to a partial relocalization of occludin and these cell types, with a greater dependency on FSH action claudin-11 to the TJ, although this was not complete and noted, as seen elsewhere (56). Importantly, tubules from remained below the sensitivity of the Western blot anal- both the suppressed (acyline Ϯ FSH Ab) and hormone- ysis, due to the small testis sizes and reduced amount of replaced groups showed a marked disruption in BTB func- total TJ protein available. In occasional tubules, complete tion by biotin tracer permeation, demonstrating that this occludin localization to the TJs with marked biotin ex- hormone-induced restoration of germ cell numbers can clusion was also seen, indicative of either a more rapid occur when Sertoli cell TJs are not fully functional. Lack recovery of tubular function under hormonal stimuli, or of TJ function is also supported by the almost total absence an incomplete suppression in all tubules by the acyline of tubule lumens in the suppressed groups, which is a qual- treatment. As no evidence for this latter point was ob- itative phenotype indicative of disrupted TJ function (1, 7, served, it is concluded that tubule recovery was heteroge- 31, 57). However, tubule lumens were observable in the neous. On the basis of other studies, which show that short-term hormone replacement groups, indicating that a testicular TJs in the suppressed hamster testis remain per- level of TJ function was restored despite the continued meable to biotin tracer after2dofexogenous FSH treat- permeability to the low molecular weight biotin tracer. ment but are impermeable after 10 d (7), it is suggested that These data suggest that Sertoli cell TJs do not need to be longer periods (Ͼ7 d) of hormonal stimulation in this acy- fully functional to allow the early restoration of spermat- line-treatment model would result in a greater restoration ogenesis, particularly at meiotic and postmeiotic stages, to of TJ protein and function. occur. A similar permeability of seminiferous tubules to a In contrast to the protein responses, acyline treatment low molecular weight tracer ([51Cr]-EDTA) has been ob- increased occludin and claudin-11 mRNA levels, whereas served in postpubertal 25-d-old rats (58), which contain short-term gonadotropin replacement decreased expression meiotic and postmeiotic germ cells (46). toward control levels. Similar in vivo responses for TJ mRNA expression have been reported elsewhere (7, 69), suggesting that gonadotropin regulation of TJ protein localization is TJ protein localization and expression more indicative of TJ function than mRNA expression. Immunohistochemical analysis showed that hormone In conclusion, this study demonstrates that mainte- suppression with acyline disrupted the localization of Ser- nance of the adult rat Sertoli cell TJ is regulated by gonad- toli cell TJ proteins. In control testes, both occludin and otropins in vivo. Chronic gonadotropin suppression resulted claudin-11 were present at functional TJs, which excluded in decreased TJ function, as shown by permeability to tracer the biotin tracer from the epithelium. After acyline treat- and loss of tubule lumens, in conjunction with altered or lost ment, occludin protein was difficult to detect by immu- TJ protein localization and a loss of adluminal germ cells. nofluorescence microscopy and Western blot analysis, Short-term hormone replacement with FSH and/or hCG par- whereas claudin-11 protein was suppressed and tended to tially reversed this phenotype, as shown by a reappearance of be redistributed within the Sertoli cell but away from the tubule lumens, some relocalization of TJ proteins to basally basal aspect of the tubules. This phenotype was similar to located TJs, and a renewed proliferation of adluminal germ the gonadotropin-deplete testis of the short-day Djungar- cell types. However, the TJ remained permeable to a low ian hamster (7, 32) and, in both cases, was associated with molecular weight tracer, demonstrating that TJs need not be disrupted TJ function as shown by extensive penetration fully functional for early spermatogenic restoration to occur. of biotin tracer throughout the seminiferous epithelium. Collectively, these data establish a new site of gonadotropin Several recent in vivo and in vitro studies have demon- action at Sertoli cell TJs in the adult rat testis. strated that TJ proteins, including occludin and JAM-A, can undergo endocytosis from the BTB (59) via a clathrin- dependent pathway (60) in response to changes in the cytokines TGF-␤3, TGF-␤2, TNF␣, and androgen (59, 60). Acknowledgments Endocytosed TJ proteins are also seen in other epithelial We thank the excellent technical contributions made by Fiona tissues (6, 7, 60–66), where they can be either recycled McLean (androgen assay), G. A. Balourdos (stereology), and back to the cell surface or targeted for intracellular deg- Enid Pruysers (gonadotropin assays) and the donation of acyline radation (for reviews, see Refs. 67 and 68). Although not from Dr. Richard P. Blye (Contraception and Reproductive investigated in this study, it is speculated that the altered Health Branch, National Institute of Child Health and Human localization of Sertoli cell TJ proteins (claudin-11 and Development/National Institutes of Health).

Address all correspondence and requests for reprints to: Peter 17. Xia W, Mruk DD, Lee WM, Cheng CY 2005 Cytokines and junction G. Stanton, Ph.D., Prince Henry’s Institute of Medical Research, restructuring during spermatogenesis—a lesson to learn from the P.O. Box 5152, Clayton 3168, Victoria, Australia. E-mail: testis. Cytokine Growth Factor Rev 16:469–493 18. Morita K, Sasaki H, Fujimoto K, Furuse M, Tsukita S 1999 Claudin- [email protected]. 11/OSP-based tight junctions of myelin sheaths in brain and Sertoli This work was supported by National Health and Medical cells in testis. J Cell Biol 145:579–588 Research Council (Australia) Program Grants 241000 and 19. Mazaud-Guittot S, Meugnier E, Pesenti S, Wu X, Vidal H, Gow A, 494802 (to D.M.R. and P.G.S.) and by the Victoria Govern- Le Magueresse-Battistoni B 2010 Claudin 11 deficiency in mice ment’s Operational Infrastructure Support Program. results in loss of the Sertoli cell epithelial phenotype in the testis. Biol Reprod 82:202–213 Disclosure Summary: The authors have nothing to disclose. 20. Saitou M, Ando-Akatsuka Y, Itoh M, Furuse M, Inazawa J, Fujimoto K, Tsukita S 1997 Mammalian occludin in epithelial cells: its expression and subcellular distribution. Eur J Cell Biol 73:222–231 21. Moroi S, Saitou M, Fujimoto K, Sakakibara A, Furuse M, Yoshida References O, Tsukita S 1998 Occludin is concentrated at tight junctions of mouse/rat but not human/guinea pig Sertoli cells in testes. Am J 1. Setchell BP 2008 Blood-testis barrier, junctional and transport pro- Physiol 274:C1708–C1717 teins and spermatogenesis. In: Cheng CY, ed. Molecular mecha- 22. Xia W, Wong CH, Lee NP, Lee WM, Cheng CY 2005 Disruption of nisms in spermatogenesis. Austin, TX: Landes Bioscience; 202–233 Sertoli-germ cell adhesion function in the seminiferous epithelium of 2. Russell LD, Peterson RN 1985 Sertoli cell junctions: morphological the rat testis can be limited to adherens junctions without affecting and functional correlates. Int Rev Cytol 94:177–211 the blood-testis barrier integrity: an in vivo study using an androgen 3. Cavicchia JC, Sacerdote FL, Ortiz L 1996 The human blood-testis suppression model. J Cell Physiol 205:141–157 barrier in impaired spermatogenesis. Ultrastruct Pathol 20:211–218 23. El Shennawy A, Gates RJ, Russell LD 1998 Hormonal regulation of 4. Russell LD, Bartke A, Goh JC 1989 Postnatal development of the spermatogenesis in the hypophysectomized rat: cell viability after Sertoli cell barrier, tubular lumen, and cytoskeleton of Sertoli and hormonal replacement in adults after intermediate periods of hy- myoid cells in the rat, and their relationship to tubular fluid secretion pophysectomy. J Androl 19:320–334; discussion 341–322 and flow. Am J Anat 184:179–189 24. Franca LR, Parreira GG, Gates RJ, Russell LD 1998 Hormonal regu- 5. de Kretser DM, Burger HG 1972 Ultrastructural studies of the hu- lation of spermatogenesis in the hypophysectomized rat: quantitation man Sertoli cell in normal men and males with hypogonadotropic of germ-cell population and effect of elimination of residual testoster- hypogonadism before and after gonadotropic treatment. In: Saxena one after long-term hypophysectomy. J Androl 19:335–340; discussion BB, Beling CG, Gandy HM, eds. Gonadotropins. New York: Wiley- 341–332 Interscience; 640–656 25. Ghosh S, Bartke A, Grasso P, Reichert Jr LE, Russell LD 1992 Struc- 6. Meng J, Holdcraft RW, Shima JE, Griswold MD, Braun RE 2005 tural manifestations of the rat Sertoli cell to hypophysectomy: a cor- Androgens regulate the permeability of the blood-testis barrier. Proc relative morphometric and endocrine study. Endocrinology 131:485– Natl Acad Sci USA 102:16696–16700 497 7. Tarulli GA, Meachem SJ, Schlatt S, Stanton PG 2008 Regulation of 26. Hikim AP, Swerdloff RS 1995 Temporal and stage-specific effects of testicular tight junctions by gonadotrophins in the adult Djungarian recombinant human follicle-stimulating hormone on the mainte- hamster in vivo. Reproduction 135:867–877 nance of spermatogenesis in gonadotropin-releasing hormone an- 8. Janecki A, Jakubowiak A, Steinberger A 1991 Regulation of trans- tagonist-treated rat. Endocrinology 136:253–261 epithelial electrical resistance in two-compartment Sertoli cell cul- 27. Sinha-Hikim AP, Swerdloff RS 1993 Temporal and stage-specific tures: in vitro model of the blood-testis barrier. Endocrinology 129: changes in spermatogenesis of rat after gonadotropin deprivation by 1489–1496 a potent gonadotropin-releasing hormone antagonist treatment. 9. Hellani A, Ji J, Mauduit C, Deschildre C, Tabone E, Benahmed M Endocrinology 133:2161–2170 2000 Developmental and hormonal regulation of the expression of 28. McLachlan RI, Wreford NG, Tsonis C, De Kretser DM, Robertson oligodendrocyte-specific protein/claudin 11 in mouse testis. Endo- DM 1994 Testosterone effects on spermatogenesis in the gonado- crinology 141:3012–3019 tropin-releasing hormone-immunized rat. Biol Reprod 50:271–280 10. Florin A, Maire M, Bozec A, Hellani A, Chater S, Bars R, Chuzel F, 29. Meachem SJ, Wreford NG, Stanton PG, Robertson DM, McLachlan Benahmed M 2005 Androgens and postmeiotic germ cells regulate RI 1998 Follicle-stimulating hormone is required for the initial claudin-11 expression in rat Sertoli cells. Endocrinology 146:1532– phase of spermatogenic restoration in adult rats following gonad- 1540 otropin suppression. J Androl 19:725–735 11. Kaitu’u-Lino TJ, Sluka P, Foo CF, Stanton PG 2007 Claudin-11 30. Matthiesson KL, McLachlan RI 2006 Male hormonal contracep- expression and localisation is regulated by androgens in rat Sertoli tion: concept proven, product in sight? Hum Reprod Update 12: cells in vitro. Reproduction 133:1169–1179 463–482 12. Sluka P, O’Donnell L, Bartles JR, Stanton PG 2006 FSH regulates the 31. Bergmann M 1987 Photoperiod and testicular function in Phodopus formation of adherens junctions and ectoplasmic specialisations be- sungorus. Adv Anat Embryol Cell Biol 105:1–76 tween rat Sertoli cells in vitro and in vivo. J Endocrinol 189:381–395 32. Tarulli GA, Stanton PG, Lerchl A, Meachem SJ 2006 Adult sertoli 13. Lui WY, Lee WM, Cheng CY 2001 Transforming growth factor-␤3 cells are not terminally differentiated in the Djungarian hamster: perturbs the inter-Sertoli tight junction permeability barrier in vitro effect of FSH on proliferation and junction protein organization. possibly mediated via its effects on occludin, zonula occludens-1, Biol Reprod 74:798–806 and claudin-11. Endocrinology 142:1865–1877 33. Hild SA, Attardi BJ, Reel JR 2004 The ability of a gonadotropin- 14. Nicholls PK, Harrison CA, Gilchrist RB, Farnworth PG, Stanton PG releasing hormone antagonist, acyline, to prevent irreversible infer- 2009 Growth differentiation factor 9 is a germ cell regulator of tility induced by the indenopyridine, CDB-4022, in adult male rats: Sertoli cell function. Endocrinology 150:2481–2490 the role of testosterone. Biol Reprod 71:348–358 15. Lui WY, Lee WM, Cheng CY 2003 Transforming growth factor ␤3 34. Porter KL, Shetty G, Meistrich ML 2006 Testicular edema is asso- regulates the dynamics of Sertoli cell tight junctions via the p38 mito- ciated with spermatogonial arrest in irradiated rats. Endocrinology gen-activated protein kinase pathway. Biol Reprod 68:1597–1612 147:1297–1305 16. Wong CH, Mruk DD, Lui WY, Cheng CY 2004 Regulation of blood- 35. Pratis K, O’Donnell L, Ooi GT, Stanton PG, McLachlan RI, testis barrier dynamics: an in vivo study. J Cell Sci 117:783–798 Robertson DM 2003 Differential regulation of rat testicular 5␣-

reductase type 1 and 2 isoforms by testosterone and FSH. J En- germ cell apoptosis due to depletion of gonadotropins caused by docrinol 176:393–403 GnRH antagonists. Apoptosis 12:1085–1100 36. Bhasin S, Fielder TJ, Swerdloff RS 1987 Testosterone selectively 54. Vera Y, Erkkila¨ K, Wang C, Nunez C, Kytta¨nen S, Lue Y, Dunkel L, increases serum follicle-stimulating hormonal (FSH) but not lutein- Swerdloff RS, Sinha Hikim AP 2006 Involvement of p38 mitogen- izing hormone (LH) in gonadotropin-releasing hormone antagonist- activated protein kinase and inducible nitric oxide synthase in ap- treated male rats: evidence for differential regulation of LH and FSH optotic signaling of murine and human male germ cells after hor- secretion. Biol Reprod 37:55–59 mone deprivation. Mol Endocrinol 20:1597–1609 37. Rea MA, Weinbauer GF, Marshall GR, Nieschlag E 1986 Testos- 55. Sinha Hikim AP, Rajavashisth TB, Sinha Hikim I, Lue Y, Bonavera terone stimulates pituitary and serum FSH in GnRH antagonist- JJ, Leung A, Wang C, Swerdloff RS 1997 Significance of apoptosis suppressed rats. Acta Endocrinol 113:487–492 in the temporal and stage-specific loss of germ cells in the adult rat 38. Wreford NG 1995 Theory and practice of stereological techniques after gonadotropin deprivation. Biol Reprod 57:1193–1201 applied to the estimation of cell number and nuclear volume in the 56. McLachlan RI, O’Donnell L, Meachem SJ, Stanton PG, de Kretser testis. Microsc Res Tech 32:423–436 DM, Pratis K, Robertson DM 2002 Identification of specific sites of 39. Pileri SA, Roncador G, Ceccarelli C, Piccioli M, Briskomatis A, hormonal regulation in spermatogenesis in rats, monkeys, and man. Sabattini E, Ascani S, Santini D, Piccaluga PP, Leone O, Damiani S, Recent Prog Horm Res 57:149–179 Ercolessi C, Sandri F, Pieri F, Leoncini L, Falini B 1997 Antigen 57. Pelletier RM 1990 A novel perspective: the occluding zonule encir- retrieval techniques in immunohistochemistry: comparison of dif- cles the apex of the Sertoli cell as observed in birds. Am J Anat ferent methods. J Pathol 183:116–123 188:87–108 40. O’Donnell L, Stanton PG, Wreford NG, Robertson DM, McLachlan 58. Setchell BP, Po¨ lla¨nen P, Zupp JL 1988 Development of the blood- RI 1996 Inhibition of 5 ␣-reductase activity impairs the testosterone- testis barrier and changes in vascular permeability at puberty in rats. dependent restoration of spermiogenesis in adult rats. Endocrinology Int J Androl 11:225–233 137:2703–2710 59. Yan HH, Mruk DD, Lee WM, Cheng CY 2008 Blood-testis barrier 41. Robertson DM, Pruysers E, Stephenson T, Pettersson K, Morton S, dynamics are regulated by testosterone and cytokines via their dif- McLachlan RI 2001 Sensitive LH and FSH assays for monitoring ferential effects on the kinetics of protein endocytosis and recycling low serum levels in men undergoing steroidal contraception. Clin in Sertoli cells. FASEB J 22:1945–1959 Endocrinol 55:331–339 60. Xia W, Wong EW, Mruk DD, Cheng CY 2009 TGF-␤3 and TNF␣ 42. Haavisto AM, Pettersson K, Bergendahl M, Perheentupa A, Roser perturb blood-testis barrier (BTB) dynamics by accelerating the clath- JF, Huhtaniemi I 1993 A supersensitive immunofluorometric assay rin-mediated endocytosis of integral membrane proteins: a new concept for rat luteinizing hormone. Endocrinology 132:1687–1691 of BTB regulation during spermatogenesis. Dev Biol 327:48–61 43. Jimenez M, Spaliviero JA, Grootenhuis AJ, Verhagen J, Allan CM, 61. Ando-Akatsuka Y, Yonemura S, Itoh M, Furuse M, Tsukita S 1999 Handelsman DJ 2005 Validation of an ultrasensitive and specific Differential behavior of E-cadherin and occludin in their colocal- immunofluorometric assay for mouse follicle-stimulating hormone. ization with ZO-1 during the establishment of epithelial cell polar- Biol Reprod 72:78–85 ity. J Cell Physiol 179:115–125 44. Russell LD, Ettlin RA, Sinha Hikim AP, Clegg ED 1990 Histological 62. Bruewer M, Luegering A, Kucharzik T, Parkos CA, Madara JL, and histopathological evaluation of the testis. Clearwater, FL: Hopkins AM, Nusrat A 2003 Proinflammatory cytokines disrupt Cache River Press epithelial barrier function by apoptosis-independent mechanisms. 45. McLachlan RI, Wreford NG, de Kretser DM, Robertson DM 1995 J Immunol 171:6164–6172 The effects of recombinant follicle-stimulating hormone on the res- 63. Bruewer M, Utech M, Ivanov AI, Hopkins AM, Parkos CA, Nusrat A toration of spermatogenesis in the gonadotropin-releasing hor- 2005 Interferon-␥ induces internalization of epithelial tight junction mone-immunized adult rat. Endocrinology 136:4035–4043 proteins via a macropinocytosis-like process. FASEB J 19:923–933 46. Meachem SJ, McLachlan RI, de Kretser DM, Robertson DM, 64. Ivanov AI, Nusrat A, Parkos CA 2004 The epithelium in inflam- Wreford NG 1996 Neonatal exposure of rats to recombinant matory bowel disease: potential role of endocytosis of junctional follicle stimulating hormone increases adult Sertoli and spermato- proteins in barrier disruption. Novartis Found Symp 263:115–124; genic cell numbers. Biol Reprod 54:36–44 discussion 124–132, 211–218 47. Hikim AP, Swerdloff RS 1994 Time course of recovery of spermat- 65. Lu L, Reinach PS, Kao WW 2001 Corneal epithelial wound healing. ogenesis and Leydig cell function after cessation of gonadotropin- Exp Biol Med 226:653–664 releasing hormone antagonist treatment in the adult rat. Endocri- 66. Morimoto S, Nishimura N, Terai T, Manabe S, Yamamoto Y, Shi- nology 134:1627–1634 nahara W, Miyake H, Tashiro S, Shimada M, Sasaki T 2005 Rab13 48. Broqua P, Riviere PJ, Conn PM, Rivier JE, Aubert ML, Junien JL mediates the continuous endocytic recycling of occludin to the cell 2002 Pharmacological profile of a new, potent, and long-acting surface. J Biol Chem 280:2220–2228 gonadotropin-releasing hormone antagonist: degarelix. J Pharma- 67. Ivanov AI, Nusrat A, Parkos CA 2005 Endocytosis of the apical col Exp Ther 301:95–102 junctional complex: mechanisms and possible roles in regulation of 49. Herbst KL 2003 Gonadotropin-releasing hormone antagonists. epithelial barriers. Bioessays 27:356–365 Curr Opin Pharmacol 3:660–666 68. Yu D, Turner JR 2008 Stimulus-induced reorganization of tight 50. Herbst KL, Coviello AD, Page S, Amory JK, Anawalt BD, Bremner junction structure: the role of membrane traffic. Biochim Biophys WJ 2004 A single dose of the potent gonadotropin-releasing hormone Acta 1778:709–716 antagonist acyline suppresses gonadotropins and testosterone for 2 69. Abel MH, Baban D, Lee S, Charlton HM, O’Shaughnessy PJ 2009 weeks in healthy young men. J Clin Endocrinol Metab 89:5959–5965 Effects of follicle stimulating hormone on testicular mRNA transcript 51. Herbst KL, Anawalt BD, Amory JK, Bremner WJ 2002 Acyline: the levels in the hypogonadal mouse. J Mol Endocrinol 42:291–303 first study in humans of a potent, new gonadotropin-releasing hor- 70. Chung NP, Mruk D, Mo MY, Lee WM, Cheng CY 2001 A 22-amino mone antagonist. J Clin Endocrinol Metab 87:3215–3220 acid synthetic peptide corresponding to the second extracellular 52. Jiang G, Stalewski J, Galyean R, Dykert J, Schteingart C, Broqua P, loop of rat occludin perturbs the blood-testis barrier and disrupts Aebi A, Aubert ML, Semple G, Robson P, Akinsanya K, Haigh R, spermatogenesis reversibly in vivo. Biol Reprod 65:1340–1351 Riviere P, Trojnar J, Junien JL, Rivier JE 2001 GnRH antagonists: 71. Tso JY, Sun XH, Kao TH, Reece KS, Wu R 1985 Isolation and a new generation of long acting analogues incorporating p-ureido- characterization of rat and human glyceraldehyde-3-phosphate de- phenylalanines at positions 5 and 6. J Med Chem 44:453–467 hydrogenase cDNAs: genomic complexity and molecular evolution 53. Pareek TK, Joshi AR, Sanyal A, Dighe RR 2007 Insights into male of the gene. Nucleic Acids Res 13:2485–2502