THE JOURNAL OP Vol. 264, No. 35, Issue of 15, pp. 21386 ... · Testins Are Structurally Related Sertoli Cell Proteins 21387 tained from Charles River Laboratories (Kingston, MA)

THE JOURNAL OP BIOLOGICAL CHEMISTRY @ 1989 by The American Society for Biochemistry and Molecular Biology, Inc

Vol. 264, No. 35, Issue of December 15, pp. 21386-21393, 1989 Printed in U.S.A

Testins Are Structurally Related Sertoli Cell Proteins Whose Secretion Is Tightly Coupled to the Presence of Germ Cells*

(Received for publication, May 19, 1989)

C. Yan Cheng$gn, Josephine GrimaSII, Michael S. StahlerS, Richard A. Lockshinlj, and C. Wayne Bardin$ From $The Population Council, §The Rockefeller University, New York, New York 10021 and the IlDepartment of Biological Sciences, St. John’s Uniuersify, Jamaica, New York 11373

Previous studies from this laboratory have shown that Sertoli cell-enriched culture medium contained two immunologically and structurally related proteins designated CMB-22 and CMB-23 with M, of 37,000 and 40,000, respectively. We have now demonstrated that both CMB-22 and CMB-23 are monomeric proteins with the following NH,-terminal amino acid sequences: CMB-22, NH2-TPDPSLDVEWNEWRTKHG-

EXNEXRTK. These sequences are virtually identical except that CMB-23 has three extra NH, terminus amino acids of X-A-P. Comparison of these sequences with those in the Protein Identification Resource revealed that they are unique proteins. CMB-22 and CMB-23 are highly concentrated in testes and their levels in this tissue increase with age. Studies using [36S]methionine incorporation and immunoprecipitation demonstrated that Sertoli cells synthesize and secrete these proteins in vitro. Because they seem not to have been isolated previously, are concentrated in and synthesized by the testes, and are structurally related, we propose that CMB-22 and CMB-23 be designated testin I and testin 11, respectively. The distribution of these proteins in biological fluids were compared with those of testibumin and rat androgen binding protein (rABP), two other Sertoli cell proteins. The results suggest that testins, unlike testibumin and rABP, are not transported to the epididymis. Although the amount of testins secreted by Sertoli cells in vitro is similar to that of testibumin and rABP, the concentrations in testis and rete testis fluid are several orders of magnitude less than that of testibumin and rABP. These observations suggest that the secretion of these proteins in vivo might be suppressed by germ cells. The fact that 10 times more testins are secreted by tubules from immature rats than by those from adult rats and that there is an increase in the testicular content of testins following a single dose of busulfan, which depleted the germ cells from the seminiferous epithelium, supports this hypothesis. Thus, the secretion of testins by Sertoli cells appears to be tightly coupled to the presence of germ cells; there is an inverse relationship between the amount of testins in the testis and the number of germ cells. These results suggest that testins are unique testicular proteins that can be used to study Sertoli cell-germ cell interactions in the seminiferous epithelium.

KTYNMNEERLKR; CMB-23, NH2-XAPXPDPSLDV-

* This work was supported by National Institutes of Health Grant HD-13541. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisemenl” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

ll To whom all correspondence should be addressed.

Morphological studies have shown that Sertoli cells and germ cells are closely associated in the adluminal compartment of the seminiferous epithelium and that the tight junctional complexes formed by the Sertoli cells create the biood- testis barrier that isolates the processes of spermatogenesis and spermiogenesis from the somatic cells and their products (1-3). Thus, Sertoli cells play an important role in regulating germ cell development by determining the ~icroenvironment of the seminiferous tubules (4). During the past decade, studies have shown that Sertoli cells are the major secretory component in the testis in that they synthesize and secrete most, if not all, of the abundant macromolecules required for spermatogenesis behind the blood-testis barrier including rABP’ (1, 5), transferrin (6), ceruloplasmin (7), clusterin (sulfated glycoprotein-2) (8-ll), testibumin (12), plasminogen activator (13), macroglobulin (141, inhibin ( E ) , growth factor(s) (16), and many others (17, 18). Many of these proteins are also synthesized and secreted by hepatocytes or other cells into the systemic circulation. These macromolecules serve as carriers, protease inhibitors, hormones as well as a variety of unknown functions. We have previously identified two Sertoli cell secretory proteins designated CMB-22 and CMB-23 (18). In the present study we report their unique NH2-terminal amino acid sequences and propose that these proteins be named testin I and testin 11. The secretion of these novel proteins appears to be suppressed by the presence of germ cells.

EXPERIMENTAL PROCEDURES

Biochemicals-’251-Labeled Bolton-Hunter reagent (N-succinimi- dyi 3-(4-hydr0xy-5-(’~~I]iodophenyl)propionate; specific activity 3056-3324 Ci/mmol) was obtained from ICN Radiochemicals. L-[3sS] Methionine (specific activity 1117-1151 Ci/mmol), methyl-14C-meth- ylated myosin (M, 200,000), phosphorylase b (M, 97,000), BSA (M, 69,000), ovalbumin (Mr 45,000), carbonic anhydrase ( M , 30,000), and EnlighteningTM autoradiography enhancer were obtained from DU Pont-New England Nuclear. Tris, EDTA, Ches, bis-Tris, and phenyl- methylsulfonyl fluoride were obtained from Aldrich. Sodium deoxycholate, BSA (RIA grade), and bacitracin were from Sigma. Triton X-100, sodium dodecyt sulfate (SDS), 2-mercaptoethanol, glycine, and N,N’-diallyltartardiamide were from Bio-Rad. DL-Methionine was from United States Biochemical Corp. Acrylamide, N,N’-methylene-hisacrylamide, immunoprecipitin (formalin-fixed Staphylococ- cus aureu cells) and prestained high molecular weight markers were from Bethesda Research Laboratories.

Animals-Male Sprague-Dawley rats at 18 days of age were ob-

’ The abbreviations used are: rABP, rat androgen binding protein; HPLC, high performance liquid chromatography; SDS, sodium do- decyl sulfate; PAGE, polyacrylamide gel electrophoresis; Ches, 2-[N- cyclohexylamino]ethanesulfonic acid; PTH, phenylthiohydantoin; RIA, radioimmunoassay; BSA, bovine serum albumin; bis-Tris, 2,2- bis(h~~droxymethyl)-2,2’~~’’-nitrilotriethano~ T, total acrylamide concentration fg/100 mt), acrylamide + methylene-bisacrylamide.

21386

Testins Are Structurally Related Sertoli Cell Proteins 21387

tained from Charles River Laboratories (Kingston, MA). Groups of 20-30 animals a t 20 days of age were used to prepare Sertoli cell- enriched cultures. Animals were killed by COB asphyxiation.

Serum, Tissue Extracts, and Reproductive Tract Fluids-Blood was withdrawn from rats at various ages by cardiac puncture under light ether anesthesia and was allowed to clot at 4 "C overnight. Serum was obtained by centrifugation at 2,500 X g for 10 min at 4 "C. Tissues including brain, liver, kidney, spleen, testis, epididymis, ovary, and uterus were obtained from rats between 10 and 60 days of age and homogenized with a Brinkmann Po'tytron in ice-cold TG buffer (20 mM Tris, 10% glycerol, pH 7.4, at 22 "C) using a tissue to buffer ratio (w/v) of 1:3 except that a ratio of 1:10 was used for epididymis and ovary. In some instances, organs were pooled from several animals. The homogenates were centrifuged at 45,000 X g for 1 h at 4 "C and the supernatants were stored at -20 "C until use. Seminiferous tubular fluid and testicular interstitial fluid were collected by in vitro micropuncture as described previously (19), samples were centrifuged at 12,000 X g for 30 min at 4 "C, and the supernatants were stored at -20 "C until use. Rete testis fluid was collected by micropuncture 4 h after ligation of the efferent ducts (20). All reproductive tract fluids were collected from animals at 60 days of age. To determine the effects of busulfan (butylene dimethanesulfonate) on the contents of testicular testins, a single dose of this drug (10 mg/kg body weight) was administered intraperitoneally on day 0 to a group of 60-day-old rats to deplete germ cells. Control animals received a single intraper- itoneal dose of arachis oil (2 ml/kg body weight). Testes were removed from treated and control animals 7,42,63, and 84 days after treatment and cytosols were prepared as described above.

Preparation of Sertoli Cell-enriched Cultures-Primary Sertoli cell- enriched cultures were prepared as described previously (17, 21). Briefly, Sertoli cell aggregates were seeded at a density of approxi- mately 5 X lo6 cells/9 ml in 100-mm culture dishes in the F-12/ Dulbecco's modified Eagle's serum-free medium obtained from GIBCO supplemented with insulin (10 pg/ml), transferrin (5 pglml), bacitracin ( 5 pg/ml), and epidermal growth factor (2.5 ng/ml), and either testosterone (2 X 10" M) or follicle stimulating hormone (300 ng/ml) or both. Bacitracin was included in the media and used as a nonspecific protease inhibitor (22, 23). The cells were maintained at 35 "C in a humidified culture chamber with 95% air and 5% COS. Spent media were collected on day 4 and fresh media were added and the cells were cultured for an additional 4 days. Thereafter, media were collected, pooled, and stored at -20 "C until used. Concentration and equilibration of media were performed on a Millipore MinitanTM ultrafiltration system (Miilipore) equipped with four MinitanTMplates with M , cutoff at 10,000. All procedures were performed at 4 "C unless otherwise specified.

Purification of Testin I and Testin II-Testin I and testin 11, formally designated CMB-22 and CMB-23, respectively, were purified from Sertoli cell-enriched culture media using sequential anion-exchange, chromatofocusing, and gel permeation HPLC as described previously (18), except that the last step of the purification by gel permeation HPLC was substituted with a Du Pont Zorbax GF-250 column (9.4 X 250 mm inner diameter) which has a particle size of 4-5 pm and a pore size of 150 A, It has a bonded phase of di-alcohol hydrophilic monolayer with a zirconia-stabilized surface. This column had better recovery over the other gel permeation HPLC column for the isolation of testin I and testin 11 and is in the range of 80-90%. The purity of both proteins was confirmed by SDS-PAGE and silver staining.

Immunoprecipitation of p5S/Methwnine-labeled Testins-To determine whether testins were synthesized in Sertoli cell-enriched cultures, primary Sertoli cells were prepared from testes obtained from rats at 20 days of age and grown in F-lZ/DME medium containing %o of the usual concentration of methionine (24). Each plate was incubated with 100 pCi of [35SJmethionine for 24 h in the presence or absence of testosterone (2 X M). In experiments designed to study the kinetics of secretion of testins by Sertoli cell-enriched cultures, cells were incubated with (35S]methionine for 15 min, and 24, 48, or 72 h. Thereafter, a total of X7 ml of spent media were collected for each treatment group and dialyzed extensively against 20 mM Tris, pH 7.4, at 22 "C using a Spectrapor dialysis membrane ( M , cutoff at 6,000-8,000) (Spectrum Medical Industries Inc., Los Angeles, CA) for 24 h at 4 "C to remove unlabeled [35S]methionine. All subsequent procedures were performed at 4 "C unless otherwise specified. Samples were concentrated to 5 ml using an Amicon ultrafiltration unit (model 8010). Five hundred p1 of formalin-fixed S. aureus (Sac) cells which had been washed three times with Sac cell buffer (50 mM Tris. pH 7.4, at, 22 "C containing 150 mM NaCl, 5 mM

Na,EDTA, 0.5% Nonidet P-40, 2 mg/ml BSA) were added to each sample. After 1 h, Sac cells were removed by centrifugation at 2,500 x g for 10 min and the supernatant was diluted 1:l (v/v) with buffer A (10 mM sodium phosphate, 0.15 M NaCl, 10 mM DL-methionine, 0.1% sodium deoxycholate, 0.1% Triton X-100, pH 7.4 at 22 "C). Thereafter, the [35S]methionine-labeled testins were immunoprecipitated by incubating with testin I antiserum at a final dilution of 1:300 for 36 h. The immune complexes were precipitat,ed by adding 100 p1 of Sac cells as described above except that the last wash was performed using buffer A instead of the Sac buffer. After centrifugation at 2,500 X g for 10 min, "S-labeled testins were recovered by extraction of the immunoprecipitated 35S-labeled proteins with 100 pl of SDS sample buffer (0.125 M Tris, 1% SDS (w/v), 1.6% 2- mercaptoethanol (v/v), 10% glycerol (v/v), pH 7.4, at 22 "C) at 100 "C for 5 min followed by centrifugation in an Eppendorf Centrifuge (model 5414) a t 16,000 X g for 2 min. An aliquot of 100 p1 was withdrawn from each sample and resolved on a 10% T SDS-polyacrylamide gel under reducing conditions. Control experiments were performed simultaneously using preimmune rabbit serum instead of the rabbit anti-testin antiserum.

Fluorography-Following SDS-PAGE, the gels were fixed in 10% acetic acid, 10% methanol (vfv) for at least 1 h. The gel was then soaked in the EnlighteningTM autoradiographic enhancer (Du Pont- New England Nuclear) for 30 min and washed in several changes of double-distilled water. It was then dried and exposed to Eastman Kodak X-OmatTM AR film at -70 "C.

R a d w i m m u ~ ~ s a y (RIA) of Testins-Since both testin I and testin I1 share common epitopes, antisera prepared against testin I quantitate testin I1 as well. The procedures used for measuring total immunoreactive testins (testin I plus testin 11) in biological extracts were essentially as described previously (18), except that a new antiserum (CYCMB22-B-2) prepared against purified testin I was used in the present study. The monospecificity of this antiserum was determined using immunoelectrophoresis and immunoblots. When this antiserum was used at a final dilution of 1:10,000, the binding of "'1-testin I was about 20% and the sensitivity of the RIA was several orders of magnitude higher than the one previously reported from this laboratory (18). When a pool of Sertoli cell-enriched culture medium (EP-130) was used to calibrate standard curves for the present and previous RIA, the minimal detectable dose and the 50% displacement were 0.05 pi eq/assay tube and 1.2 pl eq, as compared to 2 pl eq/assay tube and 50 ri eq (18). The intra-assay and inter- assay coefficients of variation were 8 and 12%, respectively. When EP-130 was calibrated against a batch of purified testins (CYCP22- l), it was noted that one pl of EP-130 contained 1.1 ng of testins. To determine if the immunoreactive testins contained in testis and serum measured by RIA are authentic proteins and that the displacement of radioactivity was not the result of 'z51-testins degradation due to proteolysis, a constant amount of radioactivity (about 30,000 rpm) was incubated with 0.1, 1, and 2 pl of testicular cytosol/serum or without sample for 48 h at 4 "C. The samples were then resolved by SDS-PAGE on 10% T rod gels, and 1.2-mm gel slices were removed and radioactivity determined by y spectrometry. It was noted that the RF and the amount of radioactivity in samples incubated with testicular cytosol/serum were identical to "'1-testin without treatment.

~ ~ , - t e r r n i ~ l Sequence Analysis o/ Testin I and Testin II-NH2- terminal sequence analyses of testin I and testin I1 were performed by Edman degradation using an Applied Biosystems (model 470A) gas-phase sequencer using 0.1 nmol of purified protein. The protein was equilibrated and concentrated to about 100 pl using an Amicon Centricon-10 microconcentrator and was then resuspended in 0.1% SDS (w/v). During each of the Edman degradation cycles, the NH, terminus amino acid was coupled with phenylisothi~yanate, cleaved by acid hydrolysis, and converted to phenylthiohydantoin (PTH)- derivatives. The PTH-derivatives were then identified and quanti- tated by reverse HPLC using a Hewlett-Packard liquid chromatography terminal (model 1084A) equipped with a Du Pont Zorbax ODS HPLC column (4.6 x 250 mm inner diameter) operated at 50 "C as described previously (10, 11). The repetitive yield was 95%. The NH2- terminal amino acid sequence for each of the proteins was determined using two separat,e batches of purified protein, and identical results were obtained.

General Methods-Protein concentration was estimated using a Coomassie Blue dye binding assay as described by Macart and Ger- baut (25). RIA for rABP and testibumin were performed as described previously (12,26). Analytical polyacrylamide gel electrophoresis was performed according to the procedures of Laemmli (27). Densitomet-

21388 Testins Are Structurally Related Sertoli Cell Proteins ric scanning to determine the relative distribution of [35S]methionine- labeled testin 1:testin I1 following immunoprecipitation was performed on an EC910 densitometer (E-C Apparatus Corp., Petersburg, FL) interfaced with a Hewlett-Packard (model HP3394A) integrator a t 600 nm. The slit size was 0.2 X 3 mm and a scan rate of 1.6 cm/ min was used.

RESULTS

Characterization of Testin I and Testin 11 Testin I and testin I1 were isolated from a 6-liter pool of

Sertoli cell-enriched culture media using the Mono QTManion- exchange and Mono PTM chromatofocusing HPLC columns. These proteins were further purified on a Du Pont Zorbax GF-250 gel permeation HPLC column. Fractionation of the purified proteins on SDS-polyacrylamide gels under reducing (Fig. lA) and nonreducing conditions (Fig. 1B) suggested that they were greater than 98% pure and that testin I and testin I1 consisted of a single polypeptide chain of M , 37,000 and 40,000, respectively. The NH2-terminal amino acid sequence analysis showed that testin I1 is an NH2-terminal extended version of testin I by three amino acids (Fig. 2, A and B). Comparison of these sequences with those in the data base at the Protein Identification Resource (PIR), which consists of 7,396 sequences containing 1,855,982 residues (release 16.0), showed no similarity. Thus, testin I and testin I1 are unique proteins.

Tissue Distribution of Immunoassayable Testins in the Reproductive Tract

A highly sensitive radioimmunoassay was used to measure the concentration of testin I plus testin I1 in biological fluids (total testins) and tissue extracts. Rete testis fluid, testicular cytosol, and serum contain immunoreactive testins that com-

A B I 2 3 4 5 6 1 lh

- 68 61

"45

4! i im - 31

31

C i r r r - D

FIG. 1. Isolation of testin I and testin I1 from Sertoli cell- enriched cultures. A, Sertoli cell-enriched cultures (about 6 liters/ batch) prepared from 20-day-old rats were fractionated by preparative anion-exchange HPLC using a preparative Mono QTM column (HR 10/16) as described previously (12, 18). Fractions containing testin I and testin I1 were pooled, concentrated, and equilibrated against 0.025 M bis-Tris, pH 7.1, with iminodiacetic acid at 22 "C and fractionated by chromatofocusing HPLC column (Mono PTM, 5 X 200 mm inner diameter) using a pH gradient of pH 7.1-4. Testin I and testin I1 were separated and eluted at different pH values of 5.8 and 5.2-5.4, respectively. Testin I and testin I1 were then purified to apparent homogeneity by gel permeation HPLC column using a Du Pont Zorbax GF-250 (9.4 X 250 mm inner diameter) column, and the purified proteins were resolved by SDS-PAGE on a 10% T polyacrylamide gel under reducing conditions. Purified testin I is shown in lanes 1-3 and purified testin I1 is shown in lanes 5-7. B, purified testin I ( l a n e 1 ) and testin I1 (lane 2) obtained in A were fractionated by SDS-PAGE on a 10% T polyacrylamide gel under nonreducing conditions without 2-mercaptoethanol. Both proteins displayed similar apparent molecular weights of the reduced proteins shown in A.

A

) TErnNI

$ 1 a \ 1 * \

0 5 10 15 20 25 30 RESIDUE NUMBER

FIG. 2. NHz-terminal sequence analysis of purified testin I (A) and testin I1 (B). Each analysis was performed twice on two separate batches of samples; identical results were obtained in each experiment. X represents an amino acid that cannot be unequivocally identified. The abscissa represents the residue number from each Edman degradation cycle; the ordinate represents the PTH-derivative obtained in each cycle. 0.1 nmol of purified protein was used for each sequence analysis.

peted with the binding of the '261-testin I to testin antibodies. The proteins in these fluids appear to share common epitopes since they generate parallel competition curves with purified testin I standard (Fig. 3A). No measurable testins were detected in the epididymal cytosol (Fig. 3A) suggesting that these proteins are not synthesized or taken up by epididymal cells as is the case with other proteins that are present in rete testis fluid and media from Sertoli cell-enriched cultures (1).

The small amounts of testins detected in various tissue extracts also generated parallel displacement curves (Fig. 3B). The concentration of testins was the highest in testes compared with serum and extracts of other organs. The concentrations of these proteins in testes appeared to increase with the age of the animals (Fig. 3C). Testins contained in rete testis fluid were examined by immunoblots using anti-testin I antibodies; it was noted that they had the same M, of the testins contained in Sertoli cell-enriched cultures (Fig. 3 0 ) .

Testicular Contribution to Serum Testin Levels A study was performed to determine whether immunoreac-

tive testins in serum originated from testes. The results showed that serum levels did not decline in adult rats up to 336 h following orchiectomy (Fig. 4). When the study was repeated using immature rats of 12 and 21 days of age, serum testins also did not decline after orchiectomy (not shown). These observations suggest that organs other than testes produce most of the testins in blood. To determine whether a small portion of the testins in blood was of testicular origin, the concentrations of testins in serum and interstitial fluid were measured and compared with those of testibumin and rABP (Table I). It was noted that the concentrations of testins


A

h

v N

0 m 2

60 ..

'O-. 0-OSCCM A-ARTF

20.. ~ - - . T c

'

V-7 SERUM *-e EC

0.01 0.1 1 10 100

IMMUNOREACTIVE TESTINS 0 1 1 oq/aaaay tub.)

A-Asplaan A-A kldnay

W--.liver

0.001 0.01 0.1 1.0 10.0 100.0

IMMUNOREACTIVE TESTINS ( p l aq/assay tuba)

C

ACE OF RAT (DAYS)

D 1 2 3

- 200

- 97 - 68

F -45 -.

X

0. w

;-31

FIG. 3. Tissue distribution of testin I and testin 11. A, rete testis fluid (RTF), testicular cytosol (TC), serum, epididymal cytosol (EC), and Sertoli cell-enriched culture medium (SCCM) (EP-130) were compared based on their ability to compete with the binding of "'1-testin to testin antibody. On the abscissa is the log dose of sample

g [ : ; : ; ; ; J 5 0 I 0 50 100 150 200 250 300 350 5 HOURS AFTER ORCHIECTOMY

FIG. 4. The concentration of testins in the serum following orchiectomy in adult rats at 60 days of age. Blood was collected from a group of a t least three animals per time point at 0 , 6, 12, 24, 48, 72, 168, and 336 h following orchiectomy.

in the serum and interstitial fluid were several orders of magnitude less than that of testibumin and rABP (Table I); however, like testibumin and rABP, the concentration of testins was higher in interstitial fluid than serum. These observations suggest that testins are secreted by Sertoli cells and contribute to the pool in the interstitial fluid (and thus into blood), but the testicular contribution to the pool of this protein in blood is small.

Secretion of Testins by Sertoli Cells in Vitro

Effects of Testosterone-Since both testin I and I1 were isolated from Sertoli cell-enriched cultures, we sought to distinguish whether these cells actively synthesize these proteins or release them in vitro after in vivo accumulation. Using [35S]methionine incorporation and immunoprecipitation, it was shown that Sertoli cells indeed synthesize and secrete testins in vitro and that testosterone increased the amount of these 35S-labeled proteins accumulating in medium (Fig. 5A). The effect of testosterone on the accumulation of immunoassayable testins in vitro was also examined using cultures in which the cells were exposed to an increasing dose of testosterone ranging between 0 and 1 X M. A dose-response

used per assay tube; on the ordinate is B/Bo, where B and Bo are counts bound in the presence (B) and absence (Bo) of unlabeled competitor. B, immunoreactive testins contained in spleen, kidney, brain, liver, and Sertoli cell-enriched cultures were compared for their ability to compete with the binding of 12sI-testin to its antibody. C, the concentrations of total testins in testis, serum, liver, and brain were compared in rats between 10 and 90 days of age. The number above the data point represents the number of animals used per group and the bar represents standard error bars. D, visualization of immunoreactive testins in Sertoli cell-enriched culture medium and rete testis fluid by immunoblot using antiserum prepared against purified testin I. Lane 1, Sertoli cell-enriched culture medium (10 pl, about 0.5 pg of protein); lane 2, rat rete testis fluid (20 pl, about 18 pg of total protein of which about 8 ng is testins as determined by radioimmunoassay). Following electrophoresis, proteins were transferred onto nitrocellulose paper and incubated with 1% anti-testin I antibody diluted in 0.1% BSA-phosphate-buffered saline-Tris buffer (10 mM sodium phosphate, 10 mM Tris, pH 7.4, at 22 "C containing 0.15 M NaCI, 0.1% BSA (w/v)). Testins were visualized using goat anti- rabbit IgG-alkaline phosphatase (diluted 1:200 with 0.1% BSA-phosphate-buffered saline-Tris buffer). The substrate used for visualization was 25 ml of 0.15 M Ches buffer (pH 9.6 at 22 "c), 50 pl of 2 M MgC12, 12.5 pl of nitro blue tetrazolium (75 mg of nitro blue tetrazolium/0.7 ml dimethylformamide plus 0.3 ml of water), 50 pl of 5- bromo-4-chloro-3-indolyl phosphate (p-toluidine salt) (10 mg/ml dimethylformamide). The sensitivity of this staining procedure is about 5 ng of protein which is about 5-fold more sensitive than the protein A-peroxidase procedure as detailed elsewhere (10).

2 1390 Testins Are Structurally Related Sertoli Cell Proteins

TABLE I Tissue distribution of testins in adult ruts

“

Sample” Testins* rABP‘ Testibumid

In vivo (pg protein/ml) Seminiferous 0.002 f 0.0005 0.014 rt 0.003 0.35 f 0.11 -

tubule culture medium‘

riched culture medium’

Sertoli cell-en- 0.92 t 0.23 0.16 2 0.042 4.3 r+ 0.92

In L,itro Organ (pg protein/g tissue) Testis 0.28 -t 0.04 3.2 rt 0.73

Epididymis <0.0001 84 f 24

Kidney 0.041 t 0.003 NDp 52 2 13

ND Spleen 0.029 zk 0.01 ND ND Liver 0.062 rt 0.02 ND ND Brain 0.012 -t 0.01 ND ND

Interstitial fluid 11.5 rt 2.12

0.0032 t 0.0005 Serum (male) 0.0022 t 0.0003

0.71 f 0.13 7.3 rt 1.4 0.23 t 0.04 4.2 f 0.28

38 t 12

Fluid (pg protein/ml) Rete testis fluid 0.16 t 0.09 5.1 rt 0.7

Two to six animals were used for each sample. Data were the means f S.D. Testins concentration was determined by RIA, as described under “Experimental Procedures.” One microliter

of the pooled Sertoli cell-enriched culture medium (EP-130) used for calibration of standard curve contained 1.1 ng of purified testins.

e Testibumin concentration was measured by RIA as described previously (12).

‘Five-cm segments of seminiferous tubules were cultured in 0.5 ml of F-12/Dulbecco’s modified Eagle’s serum-

’Sertoli cell-enriched cultures were prepared from rats a t 20 days of age, as described under “Experimental

rABP concentration was measured by RIA as detailed elsewhere (26).

free medium for 24 h at 35 “C with 95% 02, 5% Cog. Thereafter, an aliquot was withdrawn for RIA.

Procedures,” containing both follicle stimulating hormone (300 ng/ml) and testosterone (2 X 10” M). ND, not determined.

-

curve was obtained with a half maximal effect at 1 X lo-’ M (Fig. 5B).

Kinetics of Secretion-The secretion of testins by Sertoli cells in vitro was studied by measuring the amount of immunoassayable testins secreted by Sertoli cell-enriched cultures after 0, 15, 30 min; 2 h; and 1, 3, 4, and 7 days of culture. It was shown that the amount of testins accumulated in culture increased with time and was maximal after 4 days; thereafter, the amount of secreted testins remained relatively constant (Fig. 6A). When the amounts of testins secreted were examined by immunoprecipitation, it was also shown that their synthesis and secretion increased with time (Fig. 6B). When this fluorogram was examined by densitometric scanning, it was noted that the ratios of testin 1:testin I1 at 15 min, 24, 48, and 72 h following incubation with [“Slmethionine were 1:1.4, 1:1.5, 1:1.5, 1:1.3, respectively (Fig. 623). This observa- tion suggests that if testin I is a degradation product due to cleavage of testin 11, then the rate of testin I1 cleavage is equivalent to its rate of accumulation. However, testin I might be cleaved from testin I1 in Sertoli cells before their secretion.

The Relationship between Testins and Germ Cells Distribution of Testins in Different Reproductive Tract

Fluids Relative to Other Sertoli Cell Proteins-Studies on the relative epididymal and testicular concentrations of testins and two other Sertoli cell secretory proteins, testibumin and rABP, revealed several interesting correlations. First, rABP and testibumin concentrations in the testis were several orders of magnitude greater than the concentration of testins (Table I). Second, the amount of testins in the epididymis was virtually undetectable. By contrast, the concentration of rABP and testibumin in testis was only 0.1 and 1.6 times those in the epididymis. The large amount of these latter proteins in the epididymis is related to their secretion into the tubular fluid and transport to the epididymis via the rete testis (Table I). These observations are consistent with the hypothesis that the testins accumulate much less in the testis

than the other two proteins and that a relatively smaller fraction of the testicular testins is secreted into the tubular lumen for transport to the epididymis. As a consequence, the amount of testins in the epididymis is almost undetectable (Table I). The concentrations of testins, testibumin and rABP in media from seminiferous tubular and Sertoli cell-enriched cultures were next examined (Table I). The relative amounts of the three proteins in tubular culture media were similar t.0 those in rete testis fluid and testis (Table I). That is, the amount of testins secreted by seminiferous tubules in comparison with testibumin and rABP was several orders of magnitude lower (Table I). By contrast, the amount of testins secreted by Sertoli cell-enriched cultures was at a level com- parable with those of testibumin and rABP. These latter observations suggest that the secretion of testins in vitro increased when germ cells were depleted.

The Relation of Testin Secretion in the Presence of Germ Cells-The amount of testins secreted by intact seminiferous tubules prepared from rats at 10, 20, and 60 days of age was examined. The results summarized in Fig. 7A indicate that these proteins secreted by tubules of mature rats were %O of those secreted by tubules from immature rats when the number of germ cells was smaller (Fig. 7A). To further define the relationship between the secretion of testins and germ cells, a group of adult rats were treated with busulfan which selectively depleted the germ cells in the seminiferous epithelium without a marked effect on Leydig cell testosterone produc- tion. The testicular contents of testins increased following this treatment (Fig. 7B). These observations are consistent with the hypothesis that the secretion of testins by the intact seminiferous tubule is suppressed by germ cells as they accumulate after puberty.

DISCUSSION

The Sertoli cell is the major secretory component of the seminiferous epithelium. These cells determine and maintain a unique micro-environment in the adluminal compartment


A e o o - l - - - 1 200 -

68 - CI)

I 0 45- r X

s 31 -

D-

FIG. 5. The secretion of immunoreactive testins by Sertoli cell-enriched cultures in the presence of testosterone. A, Sertoli cell-enriched cultures incubated with ["S]methionine in the absence (lanes 1 and 2) and presence (lanes 3 and 4 ) of 1 X M testosterone. Lanes 1 and 3 are immunoprecipitates prepared with testin I antiserum. Lanes 2 and 4 are corresponding controls immunoprecipitated using hyperimmune rabbit serum. R, Sertoli cells were cultured in the absence or presence of various doses of testosterone for 24 h. The amount of testins in the media was measured by radioimmunoassay.

of the seminiferous epithelium behind the blood-testis barrier. This barrier is formed by the junctional complexes between the basal portions of adjacent Sertoli cells (2, 3), and it is behind this barrier that the complex processes of spermatogenesis and spermiogenesis take place (1-4). It is, therefore, conceivable that germ cell development could be regulated via the secretory proteins of Sertoli cells. It is now clear that the study of multiple Sertoli cell proteins is necessary to provide a better understanding of testicular physiology. Recent studies reveal that the secretion of rABP, transferrin, and testibumin are differentially regulated (28) and that the secretion of these proteins is under the control of different hormones (1). In addition, secretion of rABP and transferrin by Sertoli cells in vitro is bidirectional with differing amounts of each protein being released apically and basally (29). We now have reported another group of proteins, designated testins, that appear to

68 -

X

D-

FIG. 6. Kinetics of testin secretion by Sertoli cell-enriched cultures. A, Sertoli cell-enriched cultures were prepared from 20- day-old rats and plated onto 100-mm plastic Petri dishes (8 ml each) a t a cell density of 5 X lo6 cells/dish. A total of three dishes were terminated at 0, 15, 30 min, and 2 h, and 1, 3, 4, and 7 days after plating, and the amounts of testins in the spent media were measured by RIA. B, Sertoli cell-enriched cultures were incubated with ["SI methionine for 15 min, 24,48, and 72 h as shown in lanes 1 and 5 , 2 and 6, 3 and 7, and 4 and 8, respectively. [35S]Methionine-labeled proteins were immunoprecipitated using anti-testin I antiserum and visualized by fluorography (lanes 1-4). Lanes 5-8 represent the corresponding control experiments immunoprecipitated using hyperimmune rabbit serum.

be unique markers for studying Sertoli cell-germ cell interactions.

There are several features of testins that may make them interesting markers for studying the physiology of the testis and epididymis. First, they are among the few Sertoli cell proteins whose synthesis and secretion are regulated by testosterone. Second, they are highly concentrated in the testis as compared with other organs examined; but unlike other Sertoli cell secretory proteins, they do not appear to be concentrated in the epididymal compartment. Sertoli cells secrete several proteins that accumulate in the epididymis including rABP (1)) testibumin (12), cr2-macroglobulin (14)) and clusterin.* In this regard, there is a protein concentration

C. Y. Cheng, unpublished observations.

21392 Testins Are S t r u ~ ~ u r ~ l l y Related Sertoli Cell Proteins

A 0'030 T

10 20 60

Age of Rats (Days)

'oool" - -

50L--"""" ,

20 40 60 80 1 0 0 ""(

Orryr after tnatm+nt wfth 01 w8hout busulfan

FIG. 7. The relation of testins to the presence of germ cells. A, seminiferous tubules (5-cm segment/culture dish (1 ml)) were prepared from rats a t 10, 20, and 60 days of age and maintained in F-lB/Dulbecco's modified Eagle's medium for 24 h; thereafter, the concentrations of testins in the media were determined by RIA. B, the amount of testins in the testes were determined in adult rats (60 days of age) a t 7, 42, 63, and 84 days after treatment with a single dose of busulfan (10 mg/kg body weight) in comparison with control rats receiving an injection vehicle (arachis oil, 2 ml/kg body weight) intraperitoneally. This dose was sufficient to reduce mature germ cells to 20% of normal by 7 days of treatment (33).

gradient between the lumen of the seminiferous tubule and the rete testis; seminiferous tubular fluid has a protein content about 10-50-fold higher than that of rete testis fluid (5.4-29 uersus 0.6-1.1 mg/ml) (30, 31). Despite this, both rABP and testibumin are concentrated in the epididymis and the concentration of rABP in seminiferous tubular fluid and rete testis fluid versus testibumin are 3.1 and 4.2 uersus 40 and 12 pg/ml, respectively (18). Testins appear to behave in a similar manner in that they are not diluted as they pass from the seminiferous tubular lumen into the rete testis compartment (0.1 uersus 0.15 gg/ml). Even though testins appear to be processed in the same way as rABP and testibumin in the seminiferous tubular lumen and the rete testis, it is not detectable in the epididymal lumen which is in striking contrast with rABP and testibumin. It is possible that studies of the factors that regulate the concentrations of these proteins in the various compartments of the male reproductive tract may provide some insights into their function and/or metab- olism.

Testins are one of a growing number of proteins that are synthesized and secreted by Sertoli cells behind the blood- testis barrier that are also found in serum. As is the case with testibumin, clusterin, macroglobulin, and transferrin; immunoreactive testins detected in serum did not appear to originate from the testis. Furthermore, the levels of testins in the blood do not change before and after the closure of the blood-testis barrier. This suggests that the levels of these proteins in blood or testis are regulated by mechanisms other than the blood-testis barrier. This is strikingly different from the concentrations of rABP that rise in the testis and epidid-

ymis but decline in the blood by as much as 10-fold following closure of the blood-testis barrier (1, 26).

A single dose of busulfan selectively destroys differentiated spermatogonia (32, 33). By 42 days after treatment, there is selective depletion of the seminiferous tubules so that they contain basically Sertoli cells and mature germ cells. This is associated with a 40% decrease in testicular weight without a change in testicular sperm counts (33). Sixty-three days after treatment, testicular sperm count declined to 2% of controls without a further change in testicular weight (33). By 84 days, recovery of the seminiferous epithelium is in progress. These effects of busulfan do not alter Leydig cell function (34, 35). In the present study, it was noted that busulfan treatment elevated the testicular testin levels. The increase in these proteins was modest and appeared unrelated to which types of germ cells were missing from the epithelium at any given time. By contrast, rABP content was depleted from the testis and the loss of this protein begins at 42 and is maximal at 63 days when mature germ cells are low. These observations argue that drugs that reduce germ cells in the mature testis will increase the content of testins. These and other studies suggest either the secretion of these proteins is inhibited or they are actively metabolized by the developing germ cells.

In conclusion, we have identified two testosterone-respon- sive Sertoli cell secretory proteins designated testin I and testin I1 which are immunolo~cally and structurally related. A sensitive and specific RIA has been developed to quantitate these proteins in biological fluids and tissue extracts. These latter studies have shown that there was an inverse relationship between the secretion of these proteins and the presence of germ cells. Additional studies are now in progress to delin- eate the functional relationship between these cells. This will greatly facilitate the study of cell-cell interactions within the seminiferous epithelium.

Acknow4edgmen~-We thank Dr. Terry Turner at the University of Virginia for providing us with rete testis fluid, seminiferous tubular fluid, and interstitial fluid to make this study possible. We thank Drs. Martti Parvinen at the University of Turku and Carla Boitani of the University of Rome for providing us with seminiferous tubular culture medium. We also thank Dr. Ian Morris at the University of Man- chester for providing us with samples for the busulfan study. We thank Jean Schweis for her assistance in the preparation of this manuscript.

REFERENCES

1. Bardin, C. W., Cheng, C. Y., Musto, N. A., and Gunsalus, G. L. (1988) in The Physiology of Reproduction (Knobil, E., Neill, J. D., Ewing, L. L., Greenwald, G. S., Markert, C. L., and Pfaff, D. W., eds) Vol. 1, pp. 933-974, Raven Press, New York

2. Fawcett, D. W. (1975) Handb. Physiol. 5 , 21-55 3. Dym, M., and Fawcett, D. W. (1970) Biol. Reprod. 3,308-326 4. Dym, M. (1977) in Male Reproductive System: Fine Structure

Analysis by Scanning and Transmission Electron Microscopy (Yates, R. D., and Gordon, M., eds) pp. 155-169, Masson Publishing Inc., New York

5. Musto, N. A., Larrea, F., Cheng, S.-L., Kotite, N., Gunsalus, G. L., and Bardin, C. W. (1982) Ann. N. Y. Acad. Sci. 383, 343- 359

6. Perez-Infante, V., Bardin, C. W., Gunsalus, G. L., Musto, N. A., Rich, K. A., and Mather, J. P. (1986) Endocrinology 118,383- 392

7. Wright, W. W., Musto, N. A., Mather, J . P., and Bardin, C. W. (1981) Proc. Natt. Acad. Sci. U. S. A. 78, 7565-7569

8. Blaschuk. 0.. Burdzv, K., and Fritz, I. B. (1983) J. Bioi. C h m . - . . ". ..

258,7?14:7720 9. Griswold, M. D., Roberts, K., and Bishop, P. (1986) Biochemistry

10. Cheng, C. Y., Mathur, P. P., and Grima, J. (1988) Biochemistry 25,7265-7270

27,4079-4088 11. Cheng, C. Y., Chen, C.-L. C., Feng, Z. M., Marshall, A., and

Testins Are Structurally Related Sertoli Cell Proteins 2 1393

Hardin, C. W. (1988) Biochem. Biophys. Res. Commun. 155, 23. Simmons, W. H., and Ritzmann, R. F. (1980) Pharrnacol.

12. Cheng, C. Y., and Bardin, C. W. (1986) Biochemistry 25, 5276- 24. Cheng, C. Y., Grima, J., Lee, W. M., and Bardin, C . W. (1987) 5288 Biol. Reprod. 37,875-885

13. Lacroix, M., Smith, E. E., and Fritz, I. B. (1977) Mot. Cell. 25. Macart, M., and Gerbaut, L. (1982) Clin. Chim. Acta 122, 93- Endocrinol. 9, 227-236 101

14. Cheng, C . Y., Grima, J., Stahler, M. S., Guglielmotti, A., Silves- 26. Cheng, C. Y., Gunsalus, G. L., Musto, N. A., and Bardin, C. W. trini, B., and Bardin, C. W. (1990) Biochemistry, in press (1984) EndocrinoZogy 114, 1386-1394

15. Toebosch, A. M. W., Robertson, D. M., Trapman, J., Klaassen, 27. Laemmli, U. K. (1970) Nature 227,680-685 P., dePaus, R. A., declong, F. H., and Grootegoed, J. A. (1988) 28. Rossi, V., Cheng, C. Y., Gunsalus, G. I,., Bardin, C. W., and Spitz, Mol. Cell. Endocrinol. 55, 101-105 I. M. (1989) J. Androl., in press

16. Holmes, S. D., Spotts, G., and Smith, R. G. (1986) J. Biol. Chem. 29. Janecki, A., and Steinberger, A. (1987) Endocrinology 120, 291-

17. Cheng, C. Y., Mather, J. P., Byer, A. L., andBardin, C. W. (1986) 30. Cheng, C. Y., Gunsalus, G. L., Morris, I. D., Turner, T. T., and

18. Cheng, C. Y., and Bardin, C. W. (1987) J. Biol. Chern. 262, 31. Hinton, B. T., and Keefer, D. A. (1983) Exp. Cell Res. 230,367-

19. Howards, S. S., Johnson, A., and Jessee, S. (1975) Fertil. Steril. 32. Jackson, J. A., Partington, M., and Fox, B. W. (1962) Nature

20. Turner, T. T., Jones, C . E., Howards, S. S., Ewing, L. L., Zegeye, 33. Morris, I. D., Bardin, C. W., Musto, N. A., Thau, R. B., and

21. Mather, J. P., and Sato, G. H. (1979) Exp. Cell Res. 120, 191- 34. Debeljuk, L., Arimura, A., and Schally, A. V. (1973) Endocrinology

22. Patty, A., Graf, L., Kenessey, A., Szekely, J. I., and Bajusz, S. 35. Gomes, W. R., Hall, R. W., Jain, S. K., and Boots, L. R. (1973)

398-404 Biochem. Behau. 13,715-718

26 1,4076-4080 298

Endocrinology 118, 480-488 Bardin, C. W. (1986) J. Androl. 7, 175-179

12768-12779 375

26,13-19 194, 1184-1185

B., and Gunsaius, G. L. (1984) E n ~ o c r ~ ~ o ~ g y 115,1925-1932 Gunsalus, G. L. (1987) Int. J . Androl. 10, 691-700

200 92,48-54

(1977) Biochem. Bwphys. Res. Commun. 79,254-259 Endocrinology 93,800-809

Documents

THE JOURNAL OP Vol. 264, No. 35, Issue of 15, pp. 21386 ... · Testins Are Structurally Related Sertoli Cell Proteins 21387 tained from Charles River Laboratories (Kingston, MA)