- Email: [email protected]
Inhibin production by sertoli cell cultures
487
Molecular and Cellular Endocrinology, 28 (1982) 487-498 Elsevier Scientific Publishers Ireland, Ltd.
INHIBIN PRODUCTION
BY SERTOLI CELL CULTURES
F. LE GAC * and D.M. de KRETSER ** Department Received
of Anatomy, Monash University, Clayton, Victoria 3168 (Australia)
30 April
1982; revision
received
20 July 1982; accepted
20 July 1982
The production of inhibin by cultures of Sertoh cells from 21-day-old rats was assessed by the use of an in vitro bioassay using rat pituitary cells in culture. Sertoli cell culture media (SCCM) caused a dose-dependent suppression of the pituitary cell FSH content which was parallel with that of an ovine testis lymph preparation used as an inhibin standard. SCCM also caused a dose-dependent inhibition of FSH secreted by pituitary cells in response to 10 nM GnRH stimulation. The FSH-inhibitory activity in SCCM was destroyed by heat or trypsin digestion and could not be attributable to the steroid content of the medium, since ether extraction caused no change in the inhibitory activity. The inhibin activity in SCCM was not due to cytotoxicity in the bioassay, since the LH cell content was unchanged and the media produced no change in the release of “Cr from labelled pituitary cells, a parameter which has been shown to be a useful test of cytotoxicity. Sertoli cell cultures produced inhibin for the I-day duration of the cultures. The amount of inhibin produced was proportional to the number of Sertoli cells initially plated. If foetal calf serum was included for more than the initial 48 h, the spent medium caused toxic effects in the pituitary cells as evidenced by an increase in 5’ Cr release from ” Cr-labelled pituitary cells. Similar toxic effects were found if the lyophilized spent media contained cellular debris. A dose-dependent increase in inhibin activity was observed in the presence of graded doses of FSH (0.05-5 pg/ml NIH-FSHS13). Keywords:
inhibin;
Sertoli cell; pituitary
cell.
There is increasing evidence to suggest that the testis produces inhibin (for review see de Jong, 1979) and the seminiferous tubules appear to be the source of this activity (Eddie et al., 1978). The inhibition of spontaneous and GnRH-induced FSH secretion by anterior pituitary cells in culture by the media from Sertoli cell cultures strongly suggests that the Sertoli cells are the cell type involved in inhibin secretion (Steinberger
* Present address: Universite de Rennes, Faculte des Sciences, Biologie tion, Avenue de General Leclerc, Baulieu, 3500 Remres (France). ** To whom all correspondence should be addressed.
0303-7207/82/0000-OOOOooo/%O2.75
0 Elsevier Scientific
Publishers
Ireland,
de la Reproduc-
Ltd.
488
F. Le Gac and D.M. de Kretser
and Steinberger, 1976). Additional studies have confirmed that the Sertoli cells produce a substance capable of inhibiting FSH secretion (de Jong et al., 1978; Labrie et al., 1978; Demoulin et al., 1979). There is little evidence available concerning various aspects of inhibin production by Sertoli cells in culture, particularly with reference to hormonal stimulation. Although Steinberger (1980) provided preliminary evidence that FSH could stimulate inhibin production by Sertoli cell cultures, no quantitative data were provided. This study reports the results of investigations designed to characterize and quantify the levels of inhibin secreted by Sertoli cells in culture. It also provides evidence that FSH stimulates inhibin production by Sertoli cell cultures.
MATERIALS AND METHODS Sertoii cell cultures Sertoli cells were isolated from 19-21-day-old Sprague-Dawley rats as described by Dorrington et al. (1975). The cell preparations were plated at a density of l-2 X lo6 cells suspended in 2 ml of Dulbecco Eagle’s minimum essential medium (DMEM), supplemented with non-essential amino acids, glutamine, 10% foetal calf serum, streptomycin (60 pg/ml) and penicillin (100 pg/ml) (all obtained from Commonwealth Serum Laboratories, Australia) and maintained in a humidified atmosphere of 5% CO* in air at 32°C. After 24 h the culture medium was removed and replenished every 2 days by fresh medium without foetal calf serum. The spent media were centrifuged at 10000 X g for 20 min and the supernatant stored at -2O’C. Morphological examination of Sertoli cell monolayers on the 5th day, by light and electron microscopy, showed that more than 90% were Sertoli cells. In vitro inhibin bioassay Inhibin activity in Sertoli cell culture medium was measured using the in vitro bioassay method of Scott et al. (1980) with some modifications. Briefly, trypsin-dispersed rat anterior pituitary cells were prepared in Dulbecco modified Eagle’s medium (DMEM, Commonwealth Serum Laboratories, Australia) containing 10% foetal calf serum (Flow Laboratories Australasia). Aliquots of 500 ~1 (125000 cells) were then added to 16 mm well culture plates (Costar 3524, Cambridge, MA, U.S.A.) and incubated in an enclosed system maintained at 37°C under a watersaturated atmosphere of 5% CO2 in air. The cells were cultured for an initial 48 h to promote cell attachment followed by 72 h in the presence of test substances. All samples were serially diluted in sterile saline and
Inhibin production by Sertoli cell cultures
489
assayed in triplicates of 100 ~1 at each dose level; no differences were noted if DMEM instead of saline was used to dilute the samples. At the end of the assay the medium was discarded and, after a single wash with DMEM (500 ~1, pre-warmed to 37”C), the cells were lysed with Triton X-100 (500 ~1, 0.1% in Dulbecco phosphate buffer). The lysates were either stored at - 20°C or assayed immediately for FSH and LH by RIA. The gonadotropin levels, representing the cell content, were expressed as a percentage of those in control cultures (containing no test substances). The reference standard used in the bioassay is a stable lyophilized preparation of ovine testicular lymph with an arbitrary potency of 1 U/mg (Eddie et al., 1979). A 51Cr-release cytotoxicity test (Robertson et al., 1982) was applied to the bioassay system to assess if the assay response might be influenced by non-specific effects due to toxicity of the Sertoli cell culture medium in the bioassay. This method involved preloading the pituitary cells with radioactive chromium (5’Cr) and then determining the percentage of this radioactivity released into the medium during incubation with test preparations. Cytotoxicity was inferred when the percentages of “Cr release obtained for cultures exposed to test preparations were significantly elevated above those of control cultures. The inhibitory activity of the test substances on the GnRH-stimulated release of LH and FSH was also assessed using a modification (Scott et al., 1980) of the method of Eddie et al. (1979). In these experiments the pituitary cells were washed once at the end of the incubation with test substance, then reincubated for 6 h in the presence of test substance and 10 nM GnRH. LH and FSH levels were measured in the 6 h incubation media. Hormone measurement FSH and LH were measured in the medium (release) or in the cell lysate (content) by radioimmunoassay using reagents supplied by the National Pituitary Agency, U.S.A. (Lee et al., 1975). Both purified rat FSH (NIAMDD-rat-FSH-13) and rat LH (NIAMDD-rat-LH-16) preparations were iodinated with ‘251by the lactoperoxidase method (Thorell and Johansson, 1971). The labelled hormones were purified by gel filtration (Suginami et al., 1978). FSH and LH levels in lysates were and expressed as ng/ml in terms of NIAMDD-rat-FSH-RP-1 NIAMDD-rat-LH-IS, respectively. The sensitivity of the FSH assay was 75 ng/ml and that of the LH assay 0.10 ng/ml. Within-assay precision for both RIAs ranged from 4.0 to 7.8%. All serum samples and lysates from each bioassay were measured in the same RIA. Statistical analyses For bioassay results, parallel-line bioassay statistics were employed
F. Le Gac and D.M. di Kretser
490
(Finney, 1964). Samples were assayed at 2-5 dilutions and regression analysis was performed on the dose-response lines using a computer program to obtain the slope, index of precision and significance of regression. The relative potencies (and 95% confidence limits) of the Sertoli cell culture media were determined in terms of the ovine testicular lymph reference preparation. Effect of FSH on inhibin production
by Sertoli cell cultures
Ovine FSH (NIH-FSH-S13,0.05-5.0 pg/ml) was added to the culture medium for a 48 h period following the first change of medium 24 h after the initiation of the culture. The medium free of cell debris was bioassayed for inhibin. Addition of ovine FSH in comparable doses directly to the cultures of pituitary cells did not result in any change of the FSH cell content measured after decanting the medium and washing the cells once with DMEM.
RESULTS Demonstration
of inhibin activity in Sertoli cell culture medium
The Sertoli cell culture medium induced a dose-related decrease in the FSH cell content in the inhibin bioassay (Fig. 1) parallel to that induced by the OTL standard. The LH content, although variable, was not inhibited in a dose-dependent manner. Additionally, SCCM caused a dose-dependent inhibition of FSH release stimulated by 10 nM GnRH that was parallel to the activity of the OTL standard (Fig. 2). No effect of SCCM was found on GnRHstimulated LH release, in contrast to the OTL standard which suppressed LH in a dose-dependent manner (Fig. 2). Biochemical
nature of the inhibitory activity
Although the concentrations of testosterone and estradiol in Sertoli cell cultures were low (< 50 pg/ml testosterone and < 10 pg/ml estradiol), the inhibitory activities of untreated SCCM or SCCM previously treated by a double ether extraction were compared. Neither the slope of inhibition curve nor the inhibin activity of SCCM was altered significantly by the ether extraction. Furthermore, total loss of FSH-inhibitory activity was noted when Sertoli cell culture media were submitted to two treatments capable of denaturing proteins: (a) heating at 65°C for 1 h, and (b) digestion with trypsin.
491
Inhibin production by Sertoli cell cultures
-#
“4
70 6Oe
M
6111 sy 60 E
-
100
r
-
-
-
20
SO
E -~
E 0 60 u se
I-
” I
2 1p OTLP
tmg/ml)
$0
170
00261
Cvl/mlI DAY
2
DAY
(
*O
4
0
4
QTp(ms/rnl
B SCCM Crl/ml B
Fig. 1. Effect of 72 h incubation of pituitary cells in culture with increasing concentration of ovine testicular lymph (GTLP) or Sertoli cell culture medium (SCCM) on the FSH and LH intracellular content. Results are expressed as % of control (no test substance added) and represent the mean value for 3 different culture chambers (100% = 700 ng FSH/ml lysate). The percentage of 5’Cr release from pituitw cells exposed to increasing doses of OTLP and SCCM is shown.
Fig 2. The response of FSH and LH secretion by pituitary cells in culture to a 6 h incubation with 10 nM GnRH and increasing concentrations of OTLP or SCCM. The cells had been preincubated for 72 h in identical concentrations of OTLP or SCCM (100% = 1200 ng FSH/ml incubation medium).
Specificity of the inhibitory activity Increasing ~ncentrations of bovine serum albumin did not produce any suppression of FSH, indicating that the effect on pituitary cells was not simply related to protein concentration.’ Furthermore; media from renal glomerular cultures (Glasgow, 1980), either neat or concentrated 10 times, were found to be devoid of FSH-inhibitory activity (data not shown).
F. Le Got and D.M. de Kretser
492
Use was made of the 51Cr cytotoxicity test and, under the conditions described, the inhibition of FSH content due to SCCM or OTLP was not associated with any change in the ” Cr release from pituitary cells
DAY 2 SC
DAY 2 SC Lyophllised
DAY 2 Debris
kKSERTOLI
CELL
CULTURE
( pi/ml )
MEDIA
DILUTIONS
Fig. 3. In Experiment 1 the effect of inclusion of serum in the media of Sertoli cell cultures on the FSH-inhibitory activity contained in SCCM is shown. After the usual preincubation (day O-l), in half of the Sertoli cell culture dishes the media were replaced by normal DMEM (control) while in the other half the DMEM used was supplemented with 10% foetal calf serum. At day 3, media were collected and tested for activity on FSH and LH pituitary cell content and 5’Cr release. In Experiment 2 a preparation enriched in germ cells, dying cells and cellular debris (debris) was maintained under the same culture conditions as that of a normal Sertoli cell preparation (SC). Enrichment with germ cells and cellular debris was achieved by adding the material collected after the collagenase step in the method used to isolate Sertoli cells, this fraction predominantly representing germ cells. After 2 days incubation, media collected from the two types of cultures were tested (after centrifugation) in the pituitary cell culture. The normal Sertoli cell culture medium preparation was also tested after lyophilization (SC lyophilized).
Inhibin production by Sertoli cell cultures
493
(Fig. 1). Therefore, these results demonstrate that the inhibitory activity of SCCM is unlikely to be due to toxic elements able to interfere with pituitary cell culture activity. Experimental
conditions interfering with inhibin measurement
The addition of serum to the Sertoli cell culture after the preincubation period induced the appearance of an activity strongly inhibiting both the FSH and LH content, and coincidently increasing significantly the release of “Cr (Fig. 3). No effect was found when pituitary cells were incubated with fresh (unused) medium containing 10% FCS (data not shown). Furthermore, the presence of cellular debris, dying cells and/or germ cells in the culture, or the lyophilization of such media before testing, induced the inhibition of both FSH and LH content and increased release of 5’Cr (Fig. 3). The increase of 51Cr release suggests that FSH and LH inhibition was due to the presence of toxic factors in SCCM obtained or processed as described in the legend to Fig. 3.
Fig.4. Inhibin secretion, in units/ml SCCM, as a function of Sertoli cell concentration. Relative potencies were calculated by comparison to the reference OTLP. 95% confidence limits of single estimates are given.
F. Le Gac and D.M. de Kretser
494
Table 1 Variations in the inhibin potency of Sertoli cell culture medium (SCCM) collected at 48 h intervals after the initiation of culture has been established by parallel-hne bioassays. The potency is given in terms of the OTLP standard described in Materials and Methods Days of culture
Inhibin potency (units/ml SCCM)
95% confidence
o-2 2-4 4-6 6-8
19.4 18.7 15.8 25.5
(10.0-32.3) (11.0-31.6) (9.4-25.1) (15.8-40.7)
limits
Factors influencing inhibin production by Sertoli celi cultures the concentration (a) Effect of cell concentration. Increasing cells in the cultures resulted in a linear increase in the amount
of Sertoli of inhibin
25
----_ :1; loo 20
I
CONTROL
FSH (5 ~g/ml)
i 6 15
-. -------_
m
t 3
P E 60 8
E
10
I z z
5
60
40
20---
0 0.25
1
OTLP tmg/mlJ
4
10 40 160 -/-
(PI
10 40
160
InI)
0
0 05
0.5
FSH-pg/ml
Fig. 5. The dose-dependent FSH suppression by the inhibin activity in the OTLP standard is compared to the inhibin activity in media from control cultures of Sertoli cells and those stimulated by 5 pg ovine FSH (NIHSl3). Fig. 6. The dose-dependent FSH stimulation of inhibin activity Data represent the mean and 95% confidence limits. lP ~0.05 inhibin concentration at 0.05 and 0.5 pg/ml FSH.
from Sertoli cell cultures. vs. control, *P CO.05 vs.
495
Inhibin production by Sertoli cell cultures
produced (Fig. 4). Dose-response centrations tested.
curves
were parallel
at all cell con-
(b) Effect of duration of culture. SCCM was collected every 48 h from the initiation of the culture to day 8. Inhibin levels ranged between 15.8 and 25.5 units/ml of culture medium and did not vary significantly throughout the duration of the experiment (Table 1). The inhibin activity showed no departure from parallelism with the standard at any time interval examined, and no changes in the LH levels or 51Cr release were detected (data not shown). (c) Effect of FSH on inhibin production. The characteristics of the inhibin produced during FSH stimulation of Sertoli cells in culture did not differ from that produced in the absence of this hormone (Fig. 5). There was no departure from parallelism in the suppression of FSH cell content, no LH suppression resulted and no increase in “Cr release occurred. Sertoli cells grown in media containing FSH in graded doses showed a significant dose-dependent stimulation of inhibin production (Fig. 6).
DISCUSSION This study shows that the medium from immature Sertoli cells in culture can suppress FSH levels in two bioassays using anterior pituitary cells in culture, and confirms the results of previous investigators (Steinberger and Steinberger, 1976; de Jong et al., 1978; Labrie et al., 1978). The present study, in a properly validated assay system, has shown that the inhibin production is proportional to the number of Sertoli cells in culture, is maintained for 8 days, and is stimulated in a dose-dependent manner by FSH. The present study extends the previous data by demonstrating that the FSH-suppressive activity parallels the activity in a standard preparation derived from ovine testicular lymph. The demonstration of parallelism provides a bioassay criterion that the FSH-suppressive activity may be due to a similar substance which has been accepted as representing inhibin. The present study also provides additional evidence that the inhibin activity is not due to a non-specific toxic effect ,on pituitary cells, since no change occurs in the LH content of the cells and no increase in the “Cr release was found. The latter test has been shown to be a sensitive index of cytotoxicity in the inhibin bioassay and in other cell systems (Robertson et al., 1982; Spangberg, 1973). Furthermore, no
496
F. LQ Got and D.M. de Kretser
alteration of the morphology of the pituitary cells was found in the presence of inhibin-containing extracts, a criterion used by Scott et al. (1980) as indicating the absence of toxicity. The use of the “Cr cytotoxicity test has enabled us to define experimental conditions wherein unknown toxic elements interfere with inhibin measurement. The addition of serum to the culture medium after the preincubation period appears to induce the production of a toxic factor by the Sertoli cells. The purity of the culture is also important since the presence of cell debris, dying cells and/or germ cells may cause release of toxic factors in the medium. Likewise, lyophilization of SCCM may concentrate or denature elements in this medium which, at high concentration, are toxic. The demonstration that toxic materials can be produced under culture conditions makes it essential that investigators using inhibin bioassays should assess activity at multiple dilutions in comparison to a standard and use sensitive tests such as the “Cr cytotoxicity test to detect false positive results. Under conditions which do not cause toxicity in the bioassay system, our results show that increasing concentrations of Sertoli cells in the culture produced increasing amounts of inhibin and that the levels of inhibin production were maintained for at least 8 days of culture. The latter confirms the results of Steinberger (1980), who demonstrated prolonged secretion of inhibin by Sertoli cell cultures. The prolonged production of inhibin by Sertoli cell cultures is of some interest since, under the same culture conditions, ABP secretion decreased very rapidly during the first 8 days of culture when Sertoli cells were maintained in the absence of hormonal stimulation (Rommerts et al., 1978; Louis and Fritz, 1979; Le Gac, 1981). This finding suggests that inhibin and ABP secretion are differently regulated, or at least differentially affected by hormones. The term ‘inhibin’ has been given to all the various preparations that suppress FSH secretion even when LH secretion was affected. If the inhibin effect on LH can be explained in vivo by its action on hypothalamus (Demoulin et al., 1979), the effects on LH observed by different investigators using pituitary cells in vitro are still unexplained (for review see Franchimont et al., 1979). When we analyse the LH response to SCCM, in the only cases where LH content was affected, the media appeared to be toxic. Furthermore, and in contradiction with the results of de Jong et al. (1978) and Labrie et al. (1978), we did not observe a dose-dependent inhibition of GnRHstimulated LH release by SCCM, even for doses that reduced FSH maximally. This difference, as yet unexplained, might be due to the low doses of SCCM used in this work or to other experimental differences
Inhibin production by Sertoli ceil cultures
497
such as the purity or the age of the Sertoli cell culture. However, we found that under the same conditions of GnRH stimulation, testicular lymph (OTLP) inhibited LH release, as has already been shown by a number of investigators (Baker et al, 1978; Eddie et al., 1979; Scott et al., 1980; Franchimont et al., 1979). The distinct effects of OTLP and SCCM might be due to species differences (ovine vs. rat), to material differences (lymph vs. SCCM) or to age differences (lymph was collected from adult rams, Sertoli cells were isolated from immature rats). In this regard, OTL may not be the most appropriate standard for evaluating inhibin production in Sertoli cell cultures: a pool of Sertoli cell culture medium may be the more appropriate standard. In conclusion, our results demonstrate that the inhibitory activity contained in SCCM is a specific factor, is produced by Sertoli cells, is peptidic in nature and affects the pituitary cell content of FSH but not LH under our experimental conditions. Therefore, this inhibitory activity strictly meets the definition of inhibin as stated by several investigators such as Setchell et al. (1977), de Jong (1979) and Franchimont et al. (1981). The present study also provides additional evidence that the FSH-suppressing activity is indeed inhibin. The demonstration of the dose-dependent stimulation of inhibin production by FSH is in keeping with the concept of the trophic hormone stimulating a negative feedback signal. The increase of inhibin following FSH stimulation has been shown by parallel-line bioassays and the specificity is supported by the failm * zr?racts to alter 51Cr release by the pituitary cells. It supports the preliminary data presented by Steinberger (1980) indicating a stimulatory effect of FSH on inhibin production. However, the results are contrary to those of Franchimont et al. (1980) who could not demonstrate PMSG stimulation of inhibin production by Sertoli cells, but it should be noted that their conditions of culture were not identical to those of the present study. Further studies are clearly necessary to resolve these differences and to determine which other hormones modify inhibin production by Sertoli cells and whether they act synergistically with FSH.
REFERENCES Baker, H.W.G., Burger, H.G., de Kretser, D.M., Eddie, L.W., Higginson, R.E. and Lee, V.W.K. (1978) Int. J. Androl., Suppl. 2, 115-124. de Jong, F.H. (1979) Mol. Cell. Endccrinol. 13, l-10. de Jong, F.H., Welschen, R., Hennans, W.P., Smith, S.D. and van der Molen, H.J. (1978) Int. J. Androl., Suppl. 2, 125-138.
498
F. Le Gac and D.M. de Krefser
Demo&n, A., Koulischer, L., Hustin, J., Hazee-Hagelstein, M.T., Lambotte, R. and Franchimont, P. (1979) Hormone Res. 10, 177-190. Dorrington, J.H., Roller, N.F. and Fritz, J.B. (1975) Mol. Cell. Endocrinol. 3, 57-70. Eddie, L.W’., Baker, H.W.G., Dulmanis, A., Higginson, R.E. and Hudson, B. (1978) J. Endocrinol. 78,2 17-224. Eddie, L.W., Baker, H.W.G., Higginson, R.E. and Hudson, B. (1979) J. Endocrinol. 81, 49-60. Finney, D.S. (1964) Statistical Method in Biological Assay, 2nd Edn. (Charles Griffin, London). Franchimont, P., Verstraelen-Proyard, J., Hazee-Hagelstein, M.T., Renard, Ch., Demoulin, A., Bourguignon, J.P. and Hustin, J. (1979) Vitam. Hormones 37, 243-302. Franchimont, P., Craze, F. and Verhoeven, G. (1980) C.R. Seances Sot. Biol. 154, 893-897. Franchimont, P., Henderson, K., Hazee-Hagelstein, M.T., Charlet-Renard, Ch., Demoulin, A., Bourguignon, J.P. and LeComte-Yema, J.L. (1981) In: Gonadal Regulation of Reproduction, Eds.: P. Franchimont and C.P. Channing (Academic Press, New York) in press. Glasgow, E.F., Stow, J.L. and Atkins, R.C. (1980) Micron 11, 465-466. Labrie, F., Legace, L., Ferland, L., Kelly, P.A., Drouin, J., Massicotte, J., Bonne, C., Raynaud, J.P. and Dorrington, J.H. (1978) Int. J. Androl., Suppl. 2, 81-89. Lee, V.W.K., de Kretser, D.M., Hudson, B. and Wang, C. (1975) J. Reprod. Fertil. 42, 121-125. Le Gac, F. (1981) Thesis, Universite Pierre et Marie Curie, Paris. Louis, B.G. and Fritz, LB. (1979) Endocrinology 104, 454-461. Robertson, D.M., Au, C.L. and de Kretser, D.M. (1982) Mol. Cell. Endocrinol. 26, 119-127. Rommerts, F.F.G., Kruger-Sewnarain, B.C., van Woerkom-Blik, A., Grootegoed, J.A. and van der Molen, H.J. (1978) Mol. Cell. Endocrinol. 10, 39-55. Scott, RX, Burger, H.G. and Quigg, H. (1980) Endocrinology 107, 1536-1542. Setchell, B.P., Davies, R.V. and Main, S.J. (1977) In: The Testis, Vol. IV, Eds.: A.D. Johnson and W.R. Gomes (Academic Press, New York) pp. 189-238. Spangberg, L. (1973) Oral Surg. 35, 389-401. Steinberger, A. (1980) In: Endocrinology 1980, Eds.: I.A. Cumming, J.W. Funder and F.A.O. Mendelsohn (Australian Academy of Science, Canberra) pp. 259-262. Steinberger, A. and Steinberger, E. (1976) Endocrinology 99, 918-92 1. Suginami, H., Robertson, D.M. and Diczfalusy, E. (1978) Acta Endocrinol. (Kbh.) 89, 506-520. Thorell, J.I. and Johansson, B.G. (1971) B&him. Biophys. Acta 25 1, 363-369.
Molecular and Cellular Endocrinology, 28 (1982) 487-498 Elsevier Scientific Publishers Ireland, Ltd.
INHIBIN PRODUCTION
BY SERTOLI CELL CULTURES
F. LE GAC * and D.M. de KRETSER ** Department Received
of Anatomy, Monash University, Clayton, Victoria 3168 (Australia)
30 April
1982; revision
received
20 July 1982; accepted
20 July 1982
The production of inhibin by cultures of Sertoh cells from 21-day-old rats was assessed by the use of an in vitro bioassay using rat pituitary cells in culture. Sertoli cell culture media (SCCM) caused a dose-dependent suppression of the pituitary cell FSH content which was parallel with that of an ovine testis lymph preparation used as an inhibin standard. SCCM also caused a dose-dependent inhibition of FSH secreted by pituitary cells in response to 10 nM GnRH stimulation. The FSH-inhibitory activity in SCCM was destroyed by heat or trypsin digestion and could not be attributable to the steroid content of the medium, since ether extraction caused no change in the inhibitory activity. The inhibin activity in SCCM was not due to cytotoxicity in the bioassay, since the LH cell content was unchanged and the media produced no change in the release of “Cr from labelled pituitary cells, a parameter which has been shown to be a useful test of cytotoxicity. Sertoli cell cultures produced inhibin for the I-day duration of the cultures. The amount of inhibin produced was proportional to the number of Sertoli cells initially plated. If foetal calf serum was included for more than the initial 48 h, the spent medium caused toxic effects in the pituitary cells as evidenced by an increase in 5’ Cr release from ” Cr-labelled pituitary cells. Similar toxic effects were found if the lyophilized spent media contained cellular debris. A dose-dependent increase in inhibin activity was observed in the presence of graded doses of FSH (0.05-5 pg/ml NIH-FSHS13). Keywords:
inhibin;
Sertoli cell; pituitary
cell.
There is increasing evidence to suggest that the testis produces inhibin (for review see de Jong, 1979) and the seminiferous tubules appear to be the source of this activity (Eddie et al., 1978). The inhibition of spontaneous and GnRH-induced FSH secretion by anterior pituitary cells in culture by the media from Sertoli cell cultures strongly suggests that the Sertoli cells are the cell type involved in inhibin secretion (Steinberger
* Present address: Universite de Rennes, Faculte des Sciences, Biologie tion, Avenue de General Leclerc, Baulieu, 3500 Remres (France). ** To whom all correspondence should be addressed.
0303-7207/82/0000-OOOOooo/%O2.75
0 Elsevier Scientific
Publishers
Ireland,
de la Reproduc-
Ltd.
488
F. Le Gac and D.M. de Kretser
and Steinberger, 1976). Additional studies have confirmed that the Sertoli cells produce a substance capable of inhibiting FSH secretion (de Jong et al., 1978; Labrie et al., 1978; Demoulin et al., 1979). There is little evidence available concerning various aspects of inhibin production by Sertoli cells in culture, particularly with reference to hormonal stimulation. Although Steinberger (1980) provided preliminary evidence that FSH could stimulate inhibin production by Sertoli cell cultures, no quantitative data were provided. This study reports the results of investigations designed to characterize and quantify the levels of inhibin secreted by Sertoli cells in culture. It also provides evidence that FSH stimulates inhibin production by Sertoli cell cultures.
MATERIALS AND METHODS Sertoii cell cultures Sertoli cells were isolated from 19-21-day-old Sprague-Dawley rats as described by Dorrington et al. (1975). The cell preparations were plated at a density of l-2 X lo6 cells suspended in 2 ml of Dulbecco Eagle’s minimum essential medium (DMEM), supplemented with non-essential amino acids, glutamine, 10% foetal calf serum, streptomycin (60 pg/ml) and penicillin (100 pg/ml) (all obtained from Commonwealth Serum Laboratories, Australia) and maintained in a humidified atmosphere of 5% CO* in air at 32°C. After 24 h the culture medium was removed and replenished every 2 days by fresh medium without foetal calf serum. The spent media were centrifuged at 10000 X g for 20 min and the supernatant stored at -2O’C. Morphological examination of Sertoli cell monolayers on the 5th day, by light and electron microscopy, showed that more than 90% were Sertoli cells. In vitro inhibin bioassay Inhibin activity in Sertoli cell culture medium was measured using the in vitro bioassay method of Scott et al. (1980) with some modifications. Briefly, trypsin-dispersed rat anterior pituitary cells were prepared in Dulbecco modified Eagle’s medium (DMEM, Commonwealth Serum Laboratories, Australia) containing 10% foetal calf serum (Flow Laboratories Australasia). Aliquots of 500 ~1 (125000 cells) were then added to 16 mm well culture plates (Costar 3524, Cambridge, MA, U.S.A.) and incubated in an enclosed system maintained at 37°C under a watersaturated atmosphere of 5% CO2 in air. The cells were cultured for an initial 48 h to promote cell attachment followed by 72 h in the presence of test substances. All samples were serially diluted in sterile saline and
Inhibin production by Sertoli cell cultures
489
assayed in triplicates of 100 ~1 at each dose level; no differences were noted if DMEM instead of saline was used to dilute the samples. At the end of the assay the medium was discarded and, after a single wash with DMEM (500 ~1, pre-warmed to 37”C), the cells were lysed with Triton X-100 (500 ~1, 0.1% in Dulbecco phosphate buffer). The lysates were either stored at - 20°C or assayed immediately for FSH and LH by RIA. The gonadotropin levels, representing the cell content, were expressed as a percentage of those in control cultures (containing no test substances). The reference standard used in the bioassay is a stable lyophilized preparation of ovine testicular lymph with an arbitrary potency of 1 U/mg (Eddie et al., 1979). A 51Cr-release cytotoxicity test (Robertson et al., 1982) was applied to the bioassay system to assess if the assay response might be influenced by non-specific effects due to toxicity of the Sertoli cell culture medium in the bioassay. This method involved preloading the pituitary cells with radioactive chromium (5’Cr) and then determining the percentage of this radioactivity released into the medium during incubation with test preparations. Cytotoxicity was inferred when the percentages of “Cr release obtained for cultures exposed to test preparations were significantly elevated above those of control cultures. The inhibitory activity of the test substances on the GnRH-stimulated release of LH and FSH was also assessed using a modification (Scott et al., 1980) of the method of Eddie et al. (1979). In these experiments the pituitary cells were washed once at the end of the incubation with test substance, then reincubated for 6 h in the presence of test substance and 10 nM GnRH. LH and FSH levels were measured in the 6 h incubation media. Hormone measurement FSH and LH were measured in the medium (release) or in the cell lysate (content) by radioimmunoassay using reagents supplied by the National Pituitary Agency, U.S.A. (Lee et al., 1975). Both purified rat FSH (NIAMDD-rat-FSH-13) and rat LH (NIAMDD-rat-LH-16) preparations were iodinated with ‘251by the lactoperoxidase method (Thorell and Johansson, 1971). The labelled hormones were purified by gel filtration (Suginami et al., 1978). FSH and LH levels in lysates were and expressed as ng/ml in terms of NIAMDD-rat-FSH-RP-1 NIAMDD-rat-LH-IS, respectively. The sensitivity of the FSH assay was 75 ng/ml and that of the LH assay 0.10 ng/ml. Within-assay precision for both RIAs ranged from 4.0 to 7.8%. All serum samples and lysates from each bioassay were measured in the same RIA. Statistical analyses For bioassay results, parallel-line bioassay statistics were employed
F. Le Gac and D.M. di Kretser
490
(Finney, 1964). Samples were assayed at 2-5 dilutions and regression analysis was performed on the dose-response lines using a computer program to obtain the slope, index of precision and significance of regression. The relative potencies (and 95% confidence limits) of the Sertoli cell culture media were determined in terms of the ovine testicular lymph reference preparation. Effect of FSH on inhibin production
by Sertoli cell cultures
Ovine FSH (NIH-FSH-S13,0.05-5.0 pg/ml) was added to the culture medium for a 48 h period following the first change of medium 24 h after the initiation of the culture. The medium free of cell debris was bioassayed for inhibin. Addition of ovine FSH in comparable doses directly to the cultures of pituitary cells did not result in any change of the FSH cell content measured after decanting the medium and washing the cells once with DMEM.
RESULTS Demonstration
of inhibin activity in Sertoli cell culture medium
The Sertoli cell culture medium induced a dose-related decrease in the FSH cell content in the inhibin bioassay (Fig. 1) parallel to that induced by the OTL standard. The LH content, although variable, was not inhibited in a dose-dependent manner. Additionally, SCCM caused a dose-dependent inhibition of FSH release stimulated by 10 nM GnRH that was parallel to the activity of the OTL standard (Fig. 2). No effect of SCCM was found on GnRHstimulated LH release, in contrast to the OTL standard which suppressed LH in a dose-dependent manner (Fig. 2). Biochemical
nature of the inhibitory activity
Although the concentrations of testosterone and estradiol in Sertoli cell cultures were low (< 50 pg/ml testosterone and < 10 pg/ml estradiol), the inhibitory activities of untreated SCCM or SCCM previously treated by a double ether extraction were compared. Neither the slope of inhibition curve nor the inhibin activity of SCCM was altered significantly by the ether extraction. Furthermore, total loss of FSH-inhibitory activity was noted when Sertoli cell culture media were submitted to two treatments capable of denaturing proteins: (a) heating at 65°C for 1 h, and (b) digestion with trypsin.
491
Inhibin production by Sertoli cell cultures
-#
“4
70 6Oe
M
6111 sy 60 E
-
100
r
-
-
-
20
SO
E -~
E 0 60 u se
I-
” I
2 1p OTLP
tmg/ml)
$0
170
00261
Cvl/mlI DAY
2
DAY
(
*O
4
0
4
QTp(ms/rnl
B SCCM Crl/ml B
Fig. 1. Effect of 72 h incubation of pituitary cells in culture with increasing concentration of ovine testicular lymph (GTLP) or Sertoli cell culture medium (SCCM) on the FSH and LH intracellular content. Results are expressed as % of control (no test substance added) and represent the mean value for 3 different culture chambers (100% = 700 ng FSH/ml lysate). The percentage of 5’Cr release from pituitw cells exposed to increasing doses of OTLP and SCCM is shown.
Fig 2. The response of FSH and LH secretion by pituitary cells in culture to a 6 h incubation with 10 nM GnRH and increasing concentrations of OTLP or SCCM. The cells had been preincubated for 72 h in identical concentrations of OTLP or SCCM (100% = 1200 ng FSH/ml incubation medium).
Specificity of the inhibitory activity Increasing ~ncentrations of bovine serum albumin did not produce any suppression of FSH, indicating that the effect on pituitary cells was not simply related to protein concentration.’ Furthermore; media from renal glomerular cultures (Glasgow, 1980), either neat or concentrated 10 times, were found to be devoid of FSH-inhibitory activity (data not shown).
F. Le Got and D.M. de Kretser
492
Use was made of the 51Cr cytotoxicity test and, under the conditions described, the inhibition of FSH content due to SCCM or OTLP was not associated with any change in the ” Cr release from pituitary cells
DAY 2 SC
DAY 2 SC Lyophllised
DAY 2 Debris
kKSERTOLI
CELL
CULTURE
( pi/ml )
MEDIA
DILUTIONS
Fig. 3. In Experiment 1 the effect of inclusion of serum in the media of Sertoli cell cultures on the FSH-inhibitory activity contained in SCCM is shown. After the usual preincubation (day O-l), in half of the Sertoli cell culture dishes the media were replaced by normal DMEM (control) while in the other half the DMEM used was supplemented with 10% foetal calf serum. At day 3, media were collected and tested for activity on FSH and LH pituitary cell content and 5’Cr release. In Experiment 2 a preparation enriched in germ cells, dying cells and cellular debris (debris) was maintained under the same culture conditions as that of a normal Sertoli cell preparation (SC). Enrichment with germ cells and cellular debris was achieved by adding the material collected after the collagenase step in the method used to isolate Sertoli cells, this fraction predominantly representing germ cells. After 2 days incubation, media collected from the two types of cultures were tested (after centrifugation) in the pituitary cell culture. The normal Sertoli cell culture medium preparation was also tested after lyophilization (SC lyophilized).
Inhibin production by Sertoli cell cultures
493
(Fig. 1). Therefore, these results demonstrate that the inhibitory activity of SCCM is unlikely to be due to toxic elements able to interfere with pituitary cell culture activity. Experimental
conditions interfering with inhibin measurement
The addition of serum to the Sertoli cell culture after the preincubation period induced the appearance of an activity strongly inhibiting both the FSH and LH content, and coincidently increasing significantly the release of “Cr (Fig. 3). No effect was found when pituitary cells were incubated with fresh (unused) medium containing 10% FCS (data not shown). Furthermore, the presence of cellular debris, dying cells and/or germ cells in the culture, or the lyophilization of such media before testing, induced the inhibition of both FSH and LH content and increased release of 5’Cr (Fig. 3). The increase of 51Cr release suggests that FSH and LH inhibition was due to the presence of toxic factors in SCCM obtained or processed as described in the legend to Fig. 3.
Fig.4. Inhibin secretion, in units/ml SCCM, as a function of Sertoli cell concentration. Relative potencies were calculated by comparison to the reference OTLP. 95% confidence limits of single estimates are given.
F. Le Gac and D.M. de Kretser
494
Table 1 Variations in the inhibin potency of Sertoli cell culture medium (SCCM) collected at 48 h intervals after the initiation of culture has been established by parallel-hne bioassays. The potency is given in terms of the OTLP standard described in Materials and Methods Days of culture
Inhibin potency (units/ml SCCM)
95% confidence
o-2 2-4 4-6 6-8
19.4 18.7 15.8 25.5
(10.0-32.3) (11.0-31.6) (9.4-25.1) (15.8-40.7)
limits
Factors influencing inhibin production by Sertoli celi cultures the concentration (a) Effect of cell concentration. Increasing cells in the cultures resulted in a linear increase in the amount
of Sertoli of inhibin
25
----_ :1; loo 20
I
CONTROL
FSH (5 ~g/ml)
i 6 15
-. -------_
m
t 3
P E 60 8
E
10
I z z
5
60
40
20---
0 0.25
1
OTLP tmg/mlJ
4
10 40 160 -/-
(PI
10 40
160
InI)
0
0 05
0.5
FSH-pg/ml
Fig. 5. The dose-dependent FSH suppression by the inhibin activity in the OTLP standard is compared to the inhibin activity in media from control cultures of Sertoli cells and those stimulated by 5 pg ovine FSH (NIHSl3). Fig. 6. The dose-dependent FSH stimulation of inhibin activity Data represent the mean and 95% confidence limits. lP ~0.05 inhibin concentration at 0.05 and 0.5 pg/ml FSH.
from Sertoli cell cultures. vs. control, *P CO.05 vs.
495
Inhibin production by Sertoli cell cultures
produced (Fig. 4). Dose-response centrations tested.
curves
were parallel
at all cell con-
(b) Effect of duration of culture. SCCM was collected every 48 h from the initiation of the culture to day 8. Inhibin levels ranged between 15.8 and 25.5 units/ml of culture medium and did not vary significantly throughout the duration of the experiment (Table 1). The inhibin activity showed no departure from parallelism with the standard at any time interval examined, and no changes in the LH levels or 51Cr release were detected (data not shown). (c) Effect of FSH on inhibin production. The characteristics of the inhibin produced during FSH stimulation of Sertoli cells in culture did not differ from that produced in the absence of this hormone (Fig. 5). There was no departure from parallelism in the suppression of FSH cell content, no LH suppression resulted and no increase in “Cr release occurred. Sertoli cells grown in media containing FSH in graded doses showed a significant dose-dependent stimulation of inhibin production (Fig. 6).
DISCUSSION This study shows that the medium from immature Sertoli cells in culture can suppress FSH levels in two bioassays using anterior pituitary cells in culture, and confirms the results of previous investigators (Steinberger and Steinberger, 1976; de Jong et al., 1978; Labrie et al., 1978). The present study, in a properly validated assay system, has shown that the inhibin production is proportional to the number of Sertoli cells in culture, is maintained for 8 days, and is stimulated in a dose-dependent manner by FSH. The present study extends the previous data by demonstrating that the FSH-suppressive activity parallels the activity in a standard preparation derived from ovine testicular lymph. The demonstration of parallelism provides a bioassay criterion that the FSH-suppressive activity may be due to a similar substance which has been accepted as representing inhibin. The present study also provides additional evidence that the inhibin activity is not due to a non-specific toxic effect ,on pituitary cells, since no change occurs in the LH content of the cells and no increase in the “Cr release was found. The latter test has been shown to be a sensitive index of cytotoxicity in the inhibin bioassay and in other cell systems (Robertson et al., 1982; Spangberg, 1973). Furthermore, no
496
F. LQ Got and D.M. de Kretser
alteration of the morphology of the pituitary cells was found in the presence of inhibin-containing extracts, a criterion used by Scott et al. (1980) as indicating the absence of toxicity. The use of the “Cr cytotoxicity test has enabled us to define experimental conditions wherein unknown toxic elements interfere with inhibin measurement. The addition of serum to the culture medium after the preincubation period appears to induce the production of a toxic factor by the Sertoli cells. The purity of the culture is also important since the presence of cell debris, dying cells and/or germ cells may cause release of toxic factors in the medium. Likewise, lyophilization of SCCM may concentrate or denature elements in this medium which, at high concentration, are toxic. The demonstration that toxic materials can be produced under culture conditions makes it essential that investigators using inhibin bioassays should assess activity at multiple dilutions in comparison to a standard and use sensitive tests such as the “Cr cytotoxicity test to detect false positive results. Under conditions which do not cause toxicity in the bioassay system, our results show that increasing concentrations of Sertoli cells in the culture produced increasing amounts of inhibin and that the levels of inhibin production were maintained for at least 8 days of culture. The latter confirms the results of Steinberger (1980), who demonstrated prolonged secretion of inhibin by Sertoli cell cultures. The prolonged production of inhibin by Sertoli cell cultures is of some interest since, under the same culture conditions, ABP secretion decreased very rapidly during the first 8 days of culture when Sertoli cells were maintained in the absence of hormonal stimulation (Rommerts et al., 1978; Louis and Fritz, 1979; Le Gac, 1981). This finding suggests that inhibin and ABP secretion are differently regulated, or at least differentially affected by hormones. The term ‘inhibin’ has been given to all the various preparations that suppress FSH secretion even when LH secretion was affected. If the inhibin effect on LH can be explained in vivo by its action on hypothalamus (Demoulin et al., 1979), the effects on LH observed by different investigators using pituitary cells in vitro are still unexplained (for review see Franchimont et al., 1979). When we analyse the LH response to SCCM, in the only cases where LH content was affected, the media appeared to be toxic. Furthermore, and in contradiction with the results of de Jong et al. (1978) and Labrie et al. (1978), we did not observe a dose-dependent inhibition of GnRHstimulated LH release by SCCM, even for doses that reduced FSH maximally. This difference, as yet unexplained, might be due to the low doses of SCCM used in this work or to other experimental differences
Inhibin production by Sertoli ceil cultures
497
such as the purity or the age of the Sertoli cell culture. However, we found that under the same conditions of GnRH stimulation, testicular lymph (OTLP) inhibited LH release, as has already been shown by a number of investigators (Baker et al, 1978; Eddie et al., 1979; Scott et al., 1980; Franchimont et al., 1979). The distinct effects of OTLP and SCCM might be due to species differences (ovine vs. rat), to material differences (lymph vs. SCCM) or to age differences (lymph was collected from adult rams, Sertoli cells were isolated from immature rats). In this regard, OTL may not be the most appropriate standard for evaluating inhibin production in Sertoli cell cultures: a pool of Sertoli cell culture medium may be the more appropriate standard. In conclusion, our results demonstrate that the inhibitory activity contained in SCCM is a specific factor, is produced by Sertoli cells, is peptidic in nature and affects the pituitary cell content of FSH but not LH under our experimental conditions. Therefore, this inhibitory activity strictly meets the definition of inhibin as stated by several investigators such as Setchell et al. (1977), de Jong (1979) and Franchimont et al. (1981). The present study also provides additional evidence that the FSH-suppressing activity is indeed inhibin. The demonstration of the dose-dependent stimulation of inhibin production by FSH is in keeping with the concept of the trophic hormone stimulating a negative feedback signal. The increase of inhibin following FSH stimulation has been shown by parallel-line bioassays and the specificity is supported by the failm * zr?racts to alter 51Cr release by the pituitary cells. It supports the preliminary data presented by Steinberger (1980) indicating a stimulatory effect of FSH on inhibin production. However, the results are contrary to those of Franchimont et al. (1980) who could not demonstrate PMSG stimulation of inhibin production by Sertoli cells, but it should be noted that their conditions of culture were not identical to those of the present study. Further studies are clearly necessary to resolve these differences and to determine which other hormones modify inhibin production by Sertoli cells and whether they act synergistically with FSH.
REFERENCES Baker, H.W.G., Burger, H.G., de Kretser, D.M., Eddie, L.W., Higginson, R.E. and Lee, V.W.K. (1978) Int. J. Androl., Suppl. 2, 115-124. de Jong, F.H. (1979) Mol. Cell. Endccrinol. 13, l-10. de Jong, F.H., Welschen, R., Hennans, W.P., Smith, S.D. and van der Molen, H.J. (1978) Int. J. Androl., Suppl. 2, 125-138.
498
F. Le Gac and D.M. de Krefser
Demo&n, A., Koulischer, L., Hustin, J., Hazee-Hagelstein, M.T., Lambotte, R. and Franchimont, P. (1979) Hormone Res. 10, 177-190. Dorrington, J.H., Roller, N.F. and Fritz, J.B. (1975) Mol. Cell. Endocrinol. 3, 57-70. Eddie, L.W’., Baker, H.W.G., Dulmanis, A., Higginson, R.E. and Hudson, B. (1978) J. Endocrinol. 78,2 17-224. Eddie, L.W., Baker, H.W.G., Higginson, R.E. and Hudson, B. (1979) J. Endocrinol. 81, 49-60. Finney, D.S. (1964) Statistical Method in Biological Assay, 2nd Edn. (Charles Griffin, London). Franchimont, P., Verstraelen-Proyard, J., Hazee-Hagelstein, M.T., Renard, Ch., Demoulin, A., Bourguignon, J.P. and Hustin, J. (1979) Vitam. Hormones 37, 243-302. Franchimont, P., Craze, F. and Verhoeven, G. (1980) C.R. Seances Sot. Biol. 154, 893-897. Franchimont, P., Henderson, K., Hazee-Hagelstein, M.T., Charlet-Renard, Ch., Demoulin, A., Bourguignon, J.P. and LeComte-Yema, J.L. (1981) In: Gonadal Regulation of Reproduction, Eds.: P. Franchimont and C.P. Channing (Academic Press, New York) in press. Glasgow, E.F., Stow, J.L. and Atkins, R.C. (1980) Micron 11, 465-466. Labrie, F., Legace, L., Ferland, L., Kelly, P.A., Drouin, J., Massicotte, J., Bonne, C., Raynaud, J.P. and Dorrington, J.H. (1978) Int. J. Androl., Suppl. 2, 81-89. Lee, V.W.K., de Kretser, D.M., Hudson, B. and Wang, C. (1975) J. Reprod. Fertil. 42, 121-125. Le Gac, F. (1981) Thesis, Universite Pierre et Marie Curie, Paris. Louis, B.G. and Fritz, LB. (1979) Endocrinology 104, 454-461. Robertson, D.M., Au, C.L. and de Kretser, D.M. (1982) Mol. Cell. Endocrinol. 26, 119-127. Rommerts, F.F.G., Kruger-Sewnarain, B.C., van Woerkom-Blik, A., Grootegoed, J.A. and van der Molen, H.J. (1978) Mol. Cell. Endocrinol. 10, 39-55. Scott, RX, Burger, H.G. and Quigg, H. (1980) Endocrinology 107, 1536-1542. Setchell, B.P., Davies, R.V. and Main, S.J. (1977) In: The Testis, Vol. IV, Eds.: A.D. Johnson and W.R. Gomes (Academic Press, New York) pp. 189-238. Spangberg, L. (1973) Oral Surg. 35, 389-401. Steinberger, A. (1980) In: Endocrinology 1980, Eds.: I.A. Cumming, J.W. Funder and F.A.O. Mendelsohn (Australian Academy of Science, Canberra) pp. 259-262. Steinberger, A. and Steinberger, E. (1976) Endocrinology 99, 918-92 1. Suginami, H., Robertson, D.M. and Diczfalusy, E. (1978) Acta Endocrinol. (Kbh.) 89, 506-520. Thorell, J.I. and Johansson, B.G. (1971) B&him. Biophys. Acta 25 1, 363-369.