0013-7227/90/1274-1896$02.00/0 Endocrinology Copyright © 1990 by The Endocrine Society

Vol. 127, No. 4 Printed in U.S.A.

Vectorial Secretion of Inhibin by Immature Rat Sertoli Cells in Vitro: Reexamination of Previous Results* ANDRZEJ JANECKI, ANDRZEJ JAKUBOWIAK, AND ANNA STEINBERGER Department of Obstetrics, Gynecology, and Reproductive Science, University of Texas Medical School, Houston, Texas 77030

ABSTRACT. The vectorial secretion of immunoactive and bioactive inhibin by immature rat Sertoli cells (Sc) cultured in a two-compartment system was investigated using various culture supports. When Sc were cultured on Millipore-HA filters (used in all previous studies on vectorial secretion of inhibin), both immuno- and bioactive inhibin were found almost exclusively in the apical compartment, suggesting predominantly apical secretion of the glycoprotein. However, the cell-free Millipore-HA filters completely blocked the passage of Sc-conditioned medium (SCCM) inhibin, even after pretreatment with BSA and SCCM to saturate the protein-binding sites. On the other hand, polycarbonate Nucleopore filters or Millicell-CM membranes, both exhibiting extremely low protein-binding capacity, did not significantly block the passage of SCCM inhibin. When Sc were cultured on Nucleopore filters, the immunoactive inhibin was detected in both culture compartments; the basal compartment/apical compartment (BC/AC) ratio was about 1.5 (range, 1.2-1.9). The maximal effective dose of FSH or (Bu)2cAMP caused a 6- to 9-fold increase in the total (BC plus AC) secretion of immunoactive inhibin, but only a 60% increase in the secretion of bioactive inhibin, as evaluated by RIA and pituitary cell bioassay, respectively. The latter phenomenon was not accompanied by any significant change in the basal/apical distribution of either bioactive nor immunoactive inhibin. The presence of testosterone alone (10~6 M) did not affect either total

H

ISTORICALLY, the Sertoli cells (Sc) were believed to be predominantly involved in creating a chemical environment in the adluminal compartment of the seminiferous epithelium and having a "feeding" and supporting role for the differentiating germ cells. It was also generally assumed that the secretory activity of these highly polarized cells would be directed apicolaterally. However, data accumulated in recent years showed that most, if not all, Sc products are secreted bidirectionally (1). Little is known about the mechanisms that regulate the polarity of Sc secretions and the role of major secretory products within or outside the seminiferous tubules. A very useful experimental model to study the funcReceived April 20, 1990. Address all correspondence and requests for reprints to: Dr. Andrzej Janecki, Department of Obstetrics, Gynecology, and Reproductive Sciences, University of Texas Medical School, 6431 Fannin Street, Suite 3.204, Houston, Texas 77030. * This work was supported by NIH Grant HD-17802 (to A.S.).

immunoactive inhibin secretion or its BC/AC ratio. The effects of the concomitant presence of FSH and testosterone did not differ significantly from those of FSH alone. Similarly to testosterone, the lack of any significant effect was observed for 17/3estradiol, dihydrotestosterone, androstenediol, and androstenedione regardless of the presence or absence of FSH. The striking dissimilarity of BC/AC ratios of inhibin noted in cultures maintained on Millipore-HA and Nucleopore filters was not due to differences in permeability barrier or Sc functional polarity. When cultured on either support, Sc monolayers developed comparable permeability barriers, as evaluated by measuring the passage of [3H]inulin and development of electrical resistance. The maximal electrical resistance (130-150 ficm2) developed after 6-8 days of culture on either support. Also, total transferrin secretion and transferrin BC/AC ratio were similar on both supports, suggesting comparable cell numbers and functional polarities. These findings demonstrate that immature Sc in vitro secrete inhibin bidirectionally (BC/AC ratio, ~1.5); the polarity of secretion is unaffected by either FSH or various naturally occurring steroids, including testosterone. They also suggest that previously published data concerning the polarity of in vitro inhibin secretion by Sc may have been artifactually affected by inhibin binding to the culture support used. {Endocrinology 127: 1896-1903, 1990)

tional and morphological polarity of Sc has been the twocompartment culture system, in which the Sc become morphologically and functionally polarized, closely resembling their counterparts in situ (2, 3). They secrete several proteins bidirectionally, including androgenbinding protein (ABP), transferrin (Trf), and plasminogen activator (4-6). Hormones (i.e. FSH or testosterone), coculture with germ cells or peritubular myoid cells, or the addition of germ cell-conditioned medium were shown to significantly alter the polarity of secretion of certain Sc proteins (7-9). Previously reported data (10-12) indicated that immature rat Sc cultured in two-compartment chambers secrete both bioactive and immunoactive inhibin almost exclusively apically. If this were to reflect the situation in vivo, little, if any, inhibin would be secreted into testicular interstitial space. Other results, however, suggest that inhibin may be involved in regulating the

1896

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 October 2015. at 05:00 For personal use only. No other uses without permission. . All rights reserved.

INHIBIN SECRETION BY SERTOLI CELLS IN VITRO number of spermatogonial cells as well as steroidogenesis in immature rat Leydig cells (13, 14). Moreover, inhibin appears to enter the general circulation predominantly via the interstitial fluid, with reabsorption in the rete testis playing a negligible role during puberty (15). These data suggest that immature Sc in situ secrete inhibin basolaterally. In analyzing the discrepancy between the in vivo and in vitro observations, we considered the possibility that binding of inhibin to Millipore-HA filters used for cell support in the two-compartment culture chambers may have prevented its appearance in the basal compartment, resulting in an artifactually high apical/ basal ratio of inhibin secretion in vitro. Therefore, we reexamined the vectorial secretion of inhibin in the twocompartment culture system using membranes with very low protein-binding capacity (polycarbonate Nucleopore filters). In this paper we present new and more accurate data concerning the polarity of inhibin secretion and its hormonal regulation in vitro. We also provide evidence that previously reported data have been artifactually affected by the properties of filter support used for Sc culture.

Materials and Methods Sc culture Sc were isolated from testes of 18-day-old Sprague-Dawley rats, as described previously (4). The final cell preparation contained approximately 95% Sc; the residual 5% were predominantly germ cells. The cells were plated at density of 1.2-1.5 X 106 cells/cm2 and cultured at 34 C for up to 7 days in a humidified atmosphere of 2% CO2 and 98% air. The medium was replaced after the initial 24 h (day 1) and then every 48 h. The culture medium was Dulbecco's Modified Eagle's MediumHam's F-12 (1:1; Flow Laboratories, McLean, VA) supplemented with HEPES (15 mM), NaHCO3 (1.2 g/liter), ITS [insulin (5 mg/liter), transferrin (5 mg/liter), and sodium selenate (5 Mg/liter); Collaborative Research, Inc., Bedford, MA] , epidermal growth factor (EGF; 10 ^g/liter; Collaborative Research), vitamins A and E (200 ^g/liter), gentamycin (25 mg/ liter), and fungizone (2.5 mg/liter; all Sigma Chemical Co., St. Louis, MO). In some experiments this medium was supplemented from day 1 with either FSH (NIH oFSH S17; 4 or 100 ng/ml), FSH (100 ng/ml) plus testosterone (10~6 M), testosterone, 17/3-estradiol, dihydrotestosterone, androstenediol, or androstenedione (all 10"6 M; Sigma Chemical Co.). For evaluating the Sc response to various doses of FSH (NIH oFSH S17) and (Bu)2cAMP (Sigma Chemical Co.), the cultures were exposed on day 5 to FSH (0.05-400 ng/ml) or (Bu)2cAMP (0.05-400 Mg/ml), and immunoactive inhibin was measured in both compartments on day 7. In all other experiments, the media were collected at designated times separately from the apical (AC) and basal (BC) compartments, spun at 3000 X g for 15 min at 4 C, and frozen at -20 C for further use. Three types of cell support were used: 1) Millipore-HA filters (0.45-^m pore size; Millipore Co., Bedford, MA) that were

1897

preincubated with 5% BSA in Hanks' Balanced Salt Solution (HBSS) for 2 h at 37 C and washed twice with HBSS; this treatment was shown to almost completely prevent nonspecific binding of transferrin (Trf) and ABP (21); 2) Nucleopore filters (0.4 nm pore size; Nucleopore Co., Pleasanton, CA), which minimally absorb proteins; and 3) Millicell-CM chambers (12mm diameter; Millipore Co.) equipped with recently developed transparent membrane claimed to have extremely low proteinbinding capacity. Before use, all supports were covered with a film of Matrigel (Collaborative Research, Inc., Waltham, MA) diluted 1:6 with water and dried in sterile air. After the completion of experiments, the monolayers were fixed and stained for light microscopy as described previously (2). The cell density on different types of culture supports was evaluated at the end of each experiment by counting Sertoli cell nuclei (recognized by their characteristic morphology) in 10 randomly selected high magnification fields. Since in a given experiment the cells were plated at the same density per unit surface area, this method provided a reasonable estimate of the cell number per culture chamber. Electrical resistance and permeability barrier Electrical resistance of the Sc monolayers was measured in triplicate cultures at designated times using the Millicell-ERS device (Millipore Co.), which allows multiple measurements of electrical resistance without interfering with the culture protocol. The resistance of Sc monolayers was calculated by subtracting the resistance of control cell-free filters from the total resistance measured, and was expressed as ohms per cm2 (Ocm2). Immediately after the measurement of electrical resistance, the cultures were also tested for permeability to inulin, as described previously (4). Approximately 0.6 fid (1.3 nmol) of [3H]inulin (SA, 90 mCi/g; ARC, Inc., St. Louis, MO) was added to the BC, and after 5 h of incubation at 34 C medium samples were collected from the AC for scintillation counting. The results were expressed as the percentage of [3H]inulin found in AC of corresponding control cell-free chambers. Binding of inhibin and Trf to the culture supports To compare the degree of protein binding by various cell supports, Sc-conditioned medium (SCCM) containing known amounts of immunoactive inhibin and Trf was added to AC of cell-free culture chambers and incubated at 34 C for 24 h. The media were then collected separately from AC and BC (filtrate), and inhibin and Trf concentrations were determined by specific RIAs. The cell supports tested were 1) Millipore-HA filters that had been pretreated with BSA and covered with Matrigel, 2) similar filters preincubated for 48 h with SCCM (replaced four times during the incubation period) and then extensively washed with fresh medium, 3) Nucleopore filters covered with Matrigel, and 4) Matrigel-covered Millicell-CM membranes. Inhibin bioassay Inhibin bioactivity was measured in cultures of pituitary cells from 18-day-old male rats. The cells were plated (2 X 105 cells/cm2) in plastic 96-well plates (Falcon, Becton Dickinson Labware, Lincoln Park, NJ) that had been covered with a dried

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 October 2015. at 05:00 For personal use only. No other uses without permission. . All rights reserved.

INHIBIN SECRETION BY SERTOLI CELLS IN VITRO

1898

Endo • 1990 Vol 127 • No 4

film of Matrigel (as described for Sc cultures). The use of Matrigel resulted in excellent cell attachment and formation of confluent monolayers in serum-free medium. Unknown samples or serial dilutions of inhibin standard were added to triplicate culture wells just before cell plating for a total volume of 220 fd. After 3 days of incubation at 37 C in an atmosphere of 2% CO2 and 98% air, the media were collected, spun, and stored at -20 C until RIA of FSH and LH. The inhibin bioassay standard was a partially purified ovine follicular fluid inhibin preparation, WLG-4-117, kindly provided by Dr. D. Ward (MDA, Houston, TX). The bioactivity of 1 fig of this praparation was equivalent to 50 U WHO 86/690 inhibin standard. The ED50 was 0.06 ± 0.01 ng/m\, and the maximal effective dose (0.8 ± 0.2 jug/ml) resulted in 75% suppression of FSH secretion, without affecting LH secretion. The inter- and intraassay variations were 6% and 16%, respectively (n = 7).

FIG. 1. Displacement curves obtained for N-terminal peptide of porcine inhibin a-chain [(l-30)aINH], partially purified ovine follicular

RIAs

FSH, and TGF/3 did not displace (l-30)

Vectorial secretion of inhibin by immature rat Sertoli cells in vitro: reexamination of previous results.

The vectorial secretion of immunoactive and bioactive inhibin by immature rat Sertoli cells (Sc) cultured in a two-compartment system was investigated...
1MB Sizes 0 Downloads 0 Views