Molecular and Cellular Endocrinology, 83 (1992) 219-231 0 1992 Elsevier Scientific Publishers Ireland, Ltd. 0303-7207/92/$05.00

MOLCEL

219

02695

The role of specific germ cell types in modulation of the secretion of androgen-regulated proteins (ARPs) by stage VI-VIII seminiferous tubules from the adult rat Chris McKinnell and Richard M. Sharpe MRC Reproductice Biology Unit, Centre for Reproductice Biology, Edinburgh EH 3 9EW, Scotland, UK (Received

Key words: Seminiferous Testosterone;

2 September

1991; accepted

tubule; Sertoli cell; Peritubular cell; Pachytene Androgen-regulated protein; Testicular cell-cell

25 October

1991)

spermatocyte; Round spermatids; interaction; Germ cell depletion

Elongate

spermatid;

Summary

This study has assessed the role of specific germ cell types in modulating the secretion of seven previously identified androgen-regulated proteins CARPS) as well as the increase in overall level of protein secretion by seminiferous tubules (ST) at stages VI-VIII of the spermatogenic cycle in the rat. Methoxyacetic acid (MAA) was administered at 650 mg/kg to induce 80-100% depletion of pachytene and later spermatocytes; ST at stages VI-VIII were then isolated at 4, 18 and 30 days after MAA treatment when pachytene spermatocytes (PSI, round spermatids (RS) or elongate spermatids (ES), respectively, were depleted selectively. By comparison with ST isolated from controls or from rats 4 days after ethanedimethane sulphonate (EDS)-induced testosterone withdrawal, the overall level of incorporation of 35S-methionine into ST-secreted proteins was assessed after culture for 22 h at 32°C as well as the relative abundance of ARPs identified by two-dimensional sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE). Depletion of PS caused a substantial decrease in abundance of ARP-6 (53-57 kDa; pZ 5.7-5.8) and ARP-7 (56-59 kDa; pZ 5.9-6.2) and minor and more variable decreases in ARP-3 (27-30 kDa; pZ 7.4) and ARP-4 (38 kDa; pZ 6.0-6.2); ARP-6 and ARP-7 are comparable in size to the A and B forms of P-Mod-S, a product of the peritubular cells. Depletion of RS resulted in the complete disappearance of ARP-2 (13-14.5 kDa; pZ 7.3-7.61, a decrease in abundance of ARP-7 and variable decreases in ARP-1 (- 9.5 kDa; pZ 6.9-7.4) and ARP-6; ARP-1 and -2 are thought to be Sertoli cell-secreted proteins. Depletion of ES had no effect on secretion of the ARPs. Depletion of PS, RS or ES resulted in significant (P < 0.001) decreases of 26, 38 and 42% respectively in the overall level of incorporation of methionine into ST-secreted proteins, but this was significantly less than the decrease (55%) induced by EDS-induced testosterone withdrawal. These results show that the androgen-dependent changes in secretion of proteins, which occur specifically at stages VI-VIII of the normal spermatogenic cycle, are largely, if not completely, dependent on a normal germ cell complement, with each germ cell type regulating the secretion of

Correspondence to: Dr. R.M. Sharpe, MRC Reproductive Biology Unit, Centre Edinburgh EH 3 9EW,+Scotland, UK. Tel. 031-229.2575; Fax 031-228.5571.

for Reproductive

Biology,

37 Chalmers

Street,

220

proteins essential for its own development. These findings thus add to the growing evidence for the complex functional inter-dependence of all of the cell types in the testis, and probably explain why previous studies with isolated Sertoli cells have failed to identity the ARPs.

Introduction One of the most important unanswered questions in male reproductive biology is ‘how does testosterone drive spermatogenesis’?‘. Although it has been known for decades that testosterone is all-important for the maintenance of normal spermatogenesis and fertility in the male it remains virtually unknown how testosterone exerts this effect (see Sharpe et al., 1990). We have recently taken the first steps towards answering this question by identifying at least seven androgcn-regulated proteins CARPS) which are secreted in vitro by adult rat seminiferous tubules (ST) at stages VI-VIII of the spermatogenic cycle (Sharpe et al., 1991b). Within 4 days of complete testosterone withdrawal induced by destruction of the Leydig cells with ethanedimethane sulphonate (EDS), secretion of the ARPs by isolated ST at stages VI-VIII is virtually abolished but can be maintained by administration of testosterone. The ARPs are specific for stages VI-VIII of the spermatogenic cycle as, in untreated rats, these proteins are secreted in only minimal amounts at stages II-V or IX-XII. In addition to regulating the secretion of these specific ARPs, testosterone aIso has a general effect on the overall level of protein secretion by isolated ST at stages VI-VIII, which is double that at stages II-V and IX-XII; this difference is abolished within 4 days of EDS-induced testosterone withdrawal but can be maintained by testosterone administration (Sharpe et al., 1991b). As the testosterone-regulated general increase in secretion of proteins and the specific secretion of ARPs occurs at stages VI-VIII but not at earlier or later stages of the normal spermatogenie cycle, it is presumed that the normal germ cell complement at stages VI-VIII is a prerequisite for testosterone to be able to exert its effects. This is in keeping with the growing evidence that the function of the Sertoli cell, either in vitro or in vivo, is governed by the germ cells with which it is associated (Parvinen et al., 1986; Griswold et

al., 1988; Jirgou et al., 1988; AIlenby et al., 199lb; JCgou, 1991). The aim of the present study was to assess which of the germ cell types present at stages VI-VIII were essential for secretion of the ARPs. As only pachytene primary spermatocytes and round and elongate spermatids at stage VII are testosterone-dependent (see Sharpe et al., 1990, 1991b; Kerr et al., 19911, the effect of selective depletion of each of these germ cell types in vivo on the secretion of ARPs by isolated ST at stages VI-VIII was assessed. Depletion of pachytene spermatocytes was induced selectively by administration of methoxyacetic acid (MAA) (Bartlett et al., 1988; Allenby et al., 1991b), whilst depletion of round or elongate spermatids was studied at later times (18 and 30 days, respectively) after MAA administration when these cell types were depleted selectively as a result of maturation depletion. Materials and methods Animals and treatments

Young adult Wistar rats aged 75-90 days and from our own colony were used for all the studies described below; they were housed under conventional, controlled conditions. To induce selective degeneration/loss of pachytene and later spermatocytes, rats were administered MAA (Aldrich Chemical Co.) as a single oral dose of 650 mg/kg in a volume of 2.5 ml/kg; before administration the pH of the MAA was adjusted to 7.4 using concentrated NaOH and the material then diluted in 0.9% NaCl; age-matched control rats from the same cage received the saline vehicle alone. On the basis of our previous detailed studies (Bartlett et al., 1988; Sharpe, 1989; Allenby et al., 1991b; Sharpe et al., 1991a) a dose of 650 mg/kg MAA was chosen as it results in the selective destruction/loss of > 80% of pachytene and later spermatocytes at all stages, other than early-to mid-stage VII, of the spermatogenic cycle and has no other discernible effect on either the testis or the animals

221

in general. As spermatogenesis proceeds normally and with normal kinetics after MAA administration (Ratnasooriya and Sharpe, 1989; Allenby et al,, 1991b) the initial loss of pachytene and later spermatocytes is manifest as the selective loss of round and then elongate spermatids at successively later time points after treatment (maturation depletion). As the objective of the present studies was to assess the role of individual germ cell types on the secretion of ARPs at stages VI-VIII of the spermatogcnic cycle, days 4 (pachytene spermatocytes missing), 18 (round, step 6-8 spermatids missing) and 30 (elongate, step 18-19 spermatids missing) after MAA administration were chosen. At these time-points, 80-100% of the stated germ cell type should have been absent from stages VI-VIII, whereas all other germ cell types present at these stages should have been present in normal numbers. This was confirmed by both transillumination and morphological assessment of the isolated tubules (set below). In order to compare the effects on ARPs of selective germ cell depletion or androgen withdrawal, in some experiments, other rats were administered EDS at a dose of 75 mg/kg to induce selective but complete destruction of all of the Leydig cells in the testis (Kerr et al., 1985; Sharpe et al., 1990). The EDS was dissolved in dimethyl sulphoxide (DMSO)/H,O (1: 3, v/v) and administered as a single dose by intraperitoneal injection. The EDS treatment reduced intratesticular testosterone to undetectable levels and induces complete androgen withdrawal (Sharpe et al., 1990) as the result of which selected proteins CARPS) are secreted in greatly reduced amounts at stages VI-VIII of the spermatogenic cycle (Sharpe et al., 1991b). Rats were killed by inhalation of CO, followed by cervical dislocation. The testes were then dissected out rapidly, placed into preweighed containers and placed in ice. The testes were then weighed and kept on ice for a maximum of 2 h before dissection of seminiferous tubules (ST) as described below. isolation and culture of ST segments

ST were isolated and cultured as detailed previously (Sharpe et al., 1991b), using methods

which had been optimized by Allenby et al., (1991a, b) such that secretion of immunoactive inhibin by isolated ST in vitro over l-3 days reflected accurately the secretion of inhibin in vivo in a number of experimental situations. These studies had shown that ST should be isolated and cultured in lengths as long as possible (preferably > 1 cm). Testes were kept on ice until dissected and this was always within 2 h of death. In turn, each testis was decapsulated and pIaced into a small plastic Petri dish containing ice-cold Dulbecco’s phosphate-buffered saline (PBS) on a transparent Perspex stage fitted to a dissecting microscope and through which ice-cold water was pumped continuously. Subsequent dissection took place on this stage with illumination from below. The central portion of the testis was teased apart very gently using watchmaker’s forceps and long ST segments (2-5 cm) isolated. These were transferred to PBS in a separate dish and segments of ST at stages VI-VIII were cut using a scaIpe1. Stages were identified according to the criteria of Parvinen (1982) which relies on differences in the transilluminated appearance of the ST at each stage as a consequence of changes in the density and position of the heads of the elongate spermatids. However, the latter were absent or severely depleted from ST at stages VI-VIII at 30 days after MAA treatment making it impossible to isolate these stages based on normal criteria. As this had been anticipated, day 30 after MAA treatment had been chosen especially to circumvent this problem based on our extensive previous work with MAA (Bartlett et al., 1988; Ratnasooriya and Sharpe, 1989; Allenby et al., 1991), and the kinetics of normal spermatogenesis. Thus, at 30 days after MAA treatment stage VI-VIII ST should lack elongate spermatids whilst all other stages should have a normal germ cell complement and thus a normal transillumination pattern. Therefore, in this treatment group, stage VI-VIII tubules were identified based on (a) identification of earlier (I-V) and later (IXXIV) stages either side of VI-VIII, (b) the distinctive transilluminated appearance of stage VI-VIII tubules lacking elongate spermatids, and Cc) the occasional presence of a few isolated foci of elongate spermatids in their normal adluminal

222

position. The accuracy of dissection based on these criteria was confirmed by morphological assessment of the isolated ST (see below). Only staged ST segments > 0.5 cm were used for culture and in most instances the segments were l-2 cm. Staged ST segments were transferred to PBS in another Petri dish and, by reference to a transparent grid, a total of 10 cm ST was transferred to the well of a 24-well plastic culture plate containing 0.1 ml ST culture medium (see below). In each experiment ST were dissected from two or three control rats as well as from two or three rats treated either 4, 18 or 30 days previously with MAA. In selected experiments and for comparison, ST were also dissected from rats treated 4 days previously with EDS to induce maximal changes due to testosterone withdrawal (Sharpe et al., 1991b). All experiments were repeated on at least one occasion. ST were cultured for 22 h at 32°C under an atmosphere of 5% CO,/95% air. Each 10 cm ST was incubated in a final volume of 0.4 ml medium, consisting of Eagle’s minimal essential medium without methionine (Flow Labs., Irvine, Scotland, UK) to which was added 4 mM t_-glutamine (Sigma, Dorset, UK), 100 U penicillin/ml and 100 pg streptomycin/ml (both from Flow Labs.), 25 mM Hepes (Gibco, Paisley, Scotland, UK) and 0.1% polyvinyl alcohol (Sigma). Immediately prior to incubation, 60 FCi “S-1abelled methionine (Amersham International, Amersham, UK) was added to each culture. After incubation, the medium was aspirated from the ST, a protease inhibitor (Aprotinin; Sigma) added at 5% final concentration, the medium centrifuged at 1000 x g for 5 min and the aspirate stored at -40°C. Incorporation of “S-methionine into secreted proteins (i.e. in the culture medium) was determined subsequently by precipitation with trichloroacetic acid (TCA) and expressed as cpm/lO cm ST.

Two-dimensional SDS-PAGE All procedures utilized a water-cooled Protean II electrophoresis system and a model 3000 xi power supply from Bio-Rad Laboratories (Hemel Hempstead, UK). A detailed description of the materials and methods used for two-dimensional sodium dodecyl sulphate-polyacrylamide gel elec-

trophoresis (SDS-PAGE) can be found in Sharpe et al. (1991b), and only a brief outline is given here. First-dimension isoelectric focussing (IEF) gels were cast to a height of 12 cm in 160 mm x 2.5 mm internal diameter (i.d.) glass capillary tubes, each 6 ml gel containing 0.6 ml ampholines (0.4 ml pH 5-7,0.15 ml pH 7-9,0.05 ml pH 3-10) (all from Bio-Rad Labs.). IEF gels were prefocussed at 200 V for 15 min followed by 300 V for 30 min and 400 V for 30 min. Each gel was then loaded with equal cpm ‘“S-methionine-labelled proteins (250-300 x 10’ cpm), equivalent to 25-70 ~1 ST-conditioned medium, depending on the treatment group. Gels were then focussed at 400 V for 14 h followed by 800 V at 2 h. They were then extruded, frozen rapidly and stored at -40°C. Second-dimension separations utilized slab gels containing 10% acrylamide and were run at 38 mA/gel for approximately 2 h. Slab gels were then fixed for 30 min in 250 ml methanol/acetic acid/double-distilled water (40 : 10 : 50; v/v) on an orbital shaker set at 40 rpm. After silver-staining, gels were soaked for 20 min in Amplify (Amersham UK) and dried for about 2 h at 62°C on a model 543 gel drier (Bio-Rad Labs.). Dried gels were exposed to X-ray film (Kodak X-OMAT ARS) at -80°C for 21 days. ST-conditioned medium from control and MAA-treated rats from the same experiment were always run in parallel on SDS-PAGE, and subsequent determination of any change in the relative abundance of ARPs on the gels was by comparison of gels from treated animals (MAA or EDS) with their respective control from the same experiment. However, there was remarkably little variation between the control gels from different experiments.

Seminiferous tubule morphology A portion of the ST isolated from the different treatment groups were fixed by immersion in buffered 3% glutaraldehyde (Kerr and Sharpe, 1985) and embedded in epon-araldite. Semi-thin (1 pm> sections were then cut, stained with toluidine blue and examined using a Zeiss photomicroscope to confirm (1) that ST were from stages VI-VIII, and (2) that ST from rats at 4, 18 and 30 days after MAA treatment exhibited marked (80-100%) but selective depletion of pachytene

223

Results

and 18 days after treatment with MAA were not greatly different. However, in the latter two groups the absen~e/depletion of pachytene spermatocytes (4 days) or round spermatids (18 days) was discernible, as the dark central core of the ST (the elongate spermatids) extended closer to the edge of the ST than in controls (not shown). Examination of cross-sections of isolated, fixed ST confirmed that virtually all ST were at stages VI-VIII and that, in MAA-treated rats, there was severe but selective depletion or absence of either pachytene spermatocytes (+4 days) or round spermatids (+ 18 days) (Fig. 1). At 30 days after MAA treatment, isolated ST appeared very different from normal and their isolation was based predominantly on the identification by transillumination of stages either side of VI-VIII (see Materials and methods). Morphological examination of such tubules after fixation indicated minor contamination (- 12%) with ST at stages IX-X, but most were at stages VI-VIII and lacked most if not all elongate spermatids (Fig. 11.

Testicular weight and se~~n~er~us tubule {ST) ~o~pholo~ At 4 days after EDS treatment, testicular weight was reduced marginaily but non-significantly when compared with controls. In contrast, treatment with MAA induced significant decreases of 16, 22 and 20% in mean testicular weight at 4, 18 and 30 days respectively after treatment, when compared with controls (Table 1). As viewed by transillumination, the size and general appearance of ST at stages VI-VIII from control and EDS-treated rats and from rats at 4

I~cu~orfftiorz of .~~S-~ethi~ni~~einto stage VI-VZII secreted proteins The overall level of incorporation of .3sSmethionine into proteins secreted by ST at stages VI-VIII from control rats is more than double that at stages II-V or IX-XII, and this increase is androgen-dependent (Sharpe et al., 199lb). Therefore, when testosterone is withdrawn by EDS treatment there is a 50-60% decrease in incorporation of methionine and this was clearly evident in the present study (Fig. 2). Selective depletion of pachytene s~ermato~ytes or round or

spermatocytes, round spermatids spermatids, respectively.

or

elongate

Statistical analysis Data for overall incorporation of .75Smethionine into ST-secreted proteins was analysed by analysis of variance and, where a significant difference between treatment groups was indicated, sub-group comparisons were made using Student’s t-test with the within-groups variance as the measure of error. To permit overall comparison of data from several experiments, and to remove differences between experiments in the degree of incorporation of radiolabel, data for ST from each animal were expressed as a percentage of the mean incorporation for the two or three control animals within the same experiment. It is emphasized, however, that within each experiment differences comparable to those depicted in the summary illustration (Fig. 2) were observed.

TABLE I TESTICULAR WEIGHT IN CONTROL RATS AND ANIMALS TREATED WITH EDS OR MAA Means f SD. Treatment group

n

Testis weight (mg)

Control EDS + 4 days

10 4

20855185 1860+221

MAA + 4 days MAA + 18 days MAA + 30 days

5 5 5

1745f182 * 1618flll** 1675f 95 **

* P < 0.01, * * P < 0.001 in comparison with control (Student’s I-test).

Germ cell type depleted

Pachytene spermatocytes Round spermatids Elongate spermatids

u

225

2 e r; 8 B s

-

120.

Germ cell type missing I depleted I PS

RS

ES

1

1

+

+18d

+3od

100 C ***

Control

Ad

EDS Ad -

Days after MAA - treatment

-

Fig. 2. The effect of selective depletion of pachytene spermatocytes (PSI, round spermatids (RS) or elongate spermatids (ES) on the overall incorporation of ssS-methionine into proteins secreted in vitro by isolated seminiferous tubules at stages VI-VIII of the spermatogenic cycle. Data are shown in relation to controls and rats in which testosterone withdrawal had been induced 4 days earlier by EDS treatment. Each column is the mean+SD for the number of animals shown at the base of the columns; the data is from two separate experiments and, to remove differences in the level of incorporation between experiments, data for each animal has been expressed as a percentage of the dear control value from the same experiment. The actual level of incorporation into secreted proteins in controls varied between 2.8 and 5.9~ 10’ cpm/lO cm stage VI-VIII tubuie. *** P < 0.001, in comparison with control group. a P < 0.05,h P < 0.01,' P < 0.001, in comparison with data for EDS +4-day group,

elongate spermatids, following MAA treatment, also reduced significantly the overall level of methionine incorporation into secreted proteins, the magnitude of this reduction increasing with maturity of the depleted germ cell type. However, in all three MAA treatment groups, the magnitude of reduction in incorporation was always significantly less than that induced by EDS-induced testosterone withdrawal (Fig. 2). Effect of selective germ cell depletion on the secretion ofputative androgen-regulated proteins (ARPs)

Before describing results for the ARPs, there are two general points to be noted. First, because

of the reduction in incorporation of radiolabel in MAA-treated (and EDS-treated) groups (Fig. 21, substantially more ST-conditioned medium from these groups had to be loaded to maintain a constant loading of - 300,000 cpm radiolabelled protein per gel; one consequence of this is that if the secretion of a particular protein has not been affected by treatment, its relative abundance will have appeared to increase and there are several obvious proteins which fall into this category (see Fig. 4 and Table 2). Second, in the MAA-treated groups, selective germ cell depletion results in major changes (usually a complete disappearance> in a large number of proteins in addition to

Fig. 1. Representative morphology of seminiferous tubules (ST) at stages VI-VIII of the snermatoaenic cvcle after their isolation from control rats (panel A) or rats at 4 (panel Bf, 18 (panel Cl or 30 days (panel Dl after-h&A treatment. Note that the control ST exhibits a full complement of germ cells (P = pachytene spermato~tes; RS = round spermatids; ES = elongate spermatids) whereas the ST from MAA-treated rats exhibit selective depletion (SO-lo%) of either PS (day 41, RS (day 18) or ES (day 30) with apparently normal numbers of the other germ cell types (arrows = preleptotene spermatocytesf. Note that the ‘vacuolation’ of the seminiferous epithelium, especially around the base, and the detachment of the peritubular cell layer (arrowheads) are artifacts resulting from the immersion fixation. All x 250.

226

the ARPs, but as these are not the focus of this study they will not be referred to or described. Following two-dimensional SDS-PAGE, the seven ARPs identified from a previous study (Sharpe et al., 1991b) are clearly evident on the two fluorograms of control ST-secreted proteins (Figs. 3 and 4). Following EDS-induced testosterone withdrawal these ARPs virtually disappear, with the exception of ARP-5 which is regulated negatively by testosterone and which therefore appears only after testosterone withdrawal (Fig. 3). At 4 days after MAA treatment, when pachytene spermatocytes were absent/depleted, ARP-6 and ARP-7 were both reduced substantially in relative abundance whereas ARP-1 and ARP-2 were unaffected (Fig. 4); the relative abundance of ARP-3 and ARP-4 was also decreased at this time but this was rather variable between experiments. At 18 days after MAA treatment, when round spermatids were absent/ depleted, the most notable change was the com-

PH

4.7

5.1

5.8

6.2

6.5

7,O

7,2

plete disappearance of ARP-2 whilst ARP-3 and ARP-4 showed reproducible increases in relative abundance (Fig. 4); ARP-7 also showed a decrease in relative abundance at this time whilst ARP-1 and ARP-6 showed smaller decreases which were variable between experiments. In contrast to these changes, at 30 days after MAA treatment, coincident with specific depletion of elongate spermatids, the relative abundance of all of the ARPs was completely unaffected (Fig. 4). ARP-5 was absent from all fluorograms in the MAA experiment. A summary of the effect of specific germ cell depletion on the relative abundance of the ARPs is presented in Table 2. Discussion This study follows on from our recent identification of several putative androgen-regulated proteins CARPS) which are secreted in a stagespecific manner by isolated seminiferous tubules (ST) from the adult rat (Sharpe et al., 1991b).

7.4

4.7 5.1 5;8 I -. Ir.._

6;2

6;5

7;O

7;2

7.4 I

67-

3 8 X

43-

;

30-

r’ 20.1-

Fig. 3. Representative fluorograms of “S-1abelled proteins secreted by isolated seminiferous tubules at stages VI-VIII isolated from control rats (left) or from rats in which testosterone withdrawal had been induced by treatment with EDS 4 days earlier (right), to demonstrate how the latter leads to major alterations in the secretion of previously identified androgen-regulated proteins CARPS; circled and numbered on the gels). Note that identification of the ARPs is based both on their modulation by testosterone and on the stage-specificity of their secretion (see Sharpe et al., 1991b). Note also that in contrast to the other ARPs, ARP-5 is regulated negatively rather than positively by testosterone. The Sertoli cell secretory proteins sulphated glycoprotein-1 (SGP l), sulphated glycoprotein-2 (SGP 2) and cyclic protein-2 (CP 2) are indicated on the left-hand panel for reference.

227

essential if the normal level of secretion of severa1 of the ARPs is to occur at stages VI-VIII under androgen control: (2) the androgen-regu-

The findings of the present study demonstrate that: (1) the presence of one or more specific germ cell types in the seminiferous epithelium is

pH

4.7

I

5.1

I

5.8

I

6.2

I

rji5 7.0

I

7;2

7.4

I

9467-

43-

30-

43--

20. l-

14.4Fig, 4. Representative fluorograms of “%-tabelled proteins secreted by seminiferous tubules (ST) at stages VI-VIII isolated from control rats (top left) or from rats at 4 (top right), 18 (bottom left) or 30 days (bottom right) after MAA treatment, to illustrate how specific germ cell depletion affects the secretion of previously identified androgen-regulated proteins (ART’s; circled and numbered on the gels). For the purpose of illustration, only a single ‘representative’ control gel is shown, but it is emphasized that determination of any change in abundance of ARPs at each of the three time-points after MAA treatment was done by comparison with the appropriate control gel from the same experiment. A summary of the change(s) in secretion of the ARPs in relation to the particular germ cell type depleted is given in Table 2. Note that many proteins, in addition to the ARPs, are affected selectively by germ cell depletion but are not described here. The Sertoli cell-secreted proteins SW I, SGP 2 and CP 2 are indicated in the top left panel for reference.

22s TABLE

2

SUMMARY OF THE EFFECT OF MAA-INDUCED SELECTIVE DEPLETION OF PACHYTENE SPERMATOCYTES (PSI, ROUND SPERMATIDS (RS) OR ELONGATE SPERMATIDS (ES) ON THE SECRETION OF PUTATIVE ANDROGENREGULATED PROTEINS CARPS) BY SEMINIFEROUS TUBULES AT STAGES VI-VIII OF THE SPERMATOGENIC CYCLE ARP No.

Celllilar origin *

Mr * (kDa)

PI *

I

?SC

- 9.5

2 3 4 5 6 7

?SC Unknown Unknown SC il ?PC h ?PC h

6.9-7.4 7.3-7.6 7.4 6.0-6.2 4.5-5.2 5.7-5.8 5.9-6.2

13-14.5 27-30 38-38.5 39-46 53-57 56-59

Effect of specific germ cell depletion

**

(day 4) PS

(day 18) RS

(day 30) ES

_ _

_ _ _

T-1 n.d.

?i ilk Tf TT n.d.

J. 1

?I 1

?1

n.d. _

* Based on Sharpe et al. (1991b) and unpublished data (SC = Sertoli cell; PC = peritubular cell). * * Based on the comparison of gels for control and MAA-treated groups from two experiments, utilizing ST-conditioned medium from separate animals. The magnitude of change is indicated by the number of arrows; ? indicates a small but inconsistent decrease between gels. Note also that T t, indicating an increase in relative abundance, could be due in part to the fact that more ST-conditioned medium had to be run on these gels than in the controls (Fig. I) to normalize the total added cpm radiolabelled protein/gel. ~’ ARP-5 is an isomer of sulphated glycoprotein-2 (Griswold et al., 1988). ” ARP-6 and ARP-7 are the same molecular weight as the A and B forms, respectively, of P-Mod-S (Skinner et al., 19881. n.d. = not detectable on any of the fluorograms from control and MAAtreated rats.

lated increase in total protein secretion by ST at stages VI-VIII is dependent on a full complement of the different germ cell types. These findings also have some important general implications as to which of the androgen-regulated protein changes are essential for the development of individual germ cell types during their passage through stages VI-VIII of the spermatogenic cycle. To investigate the potential modulatory role of selective germ cell types on secretion of the ARPs, pachytene and later spermatocytes were depleted using the testicular toxicant MAA (Greasy et al., 1985; Foster et al., 1987). MAA has been used extensively in our laboratory in recent years to investigate in vivo the effects of selective germ cell depletion (e.g. Bartlett et al., 1988; Ratnasooriya and Sharpe, 1989; Sharpe 1989; Allenby et al., 1991b; Sharpe et al., 1991a). These studies have demonstrated the remarkable selectivity, effectiveness and reproducibility of MM in depleting ST of pachytene and later spermatocytes. Because spermatogenesis proceeds with normal kinetics after MAA treatment, particular times

can be identified post-treatment when ST at certain stages will be depleted of later germ cell types due to maturation depletion. Based on previous temporal studies (Bartlett et al., 19881, days 4 (pachytene spermatocytes depleted), 18 (round spermatid depleted) and 30 (elongate spermatids depleted) were chosen for the present study. The primary reason for selection of these time-points was that ST at stages VI-VIII should show SO100% specific depletion of only one of the particular germ cell types at each of the times in question, and this was confirmed by morphological examination of isolated ST. Two additional factors influenced the choice of time-points. First, potential direct effects of MAA on ST-secreted proteins should be minimal by day 4 due to elimination of MAA (see Allenby et al., 1991b). Second, day 30 was chosen to provide ‘normal’ stages either side of VI-VIII to enable isolation of stages VI-VIII by transillumination in the near absence of elongate spermatids. The latter strategy worked remarkably well as indicated both by morphological examination of the isolated ST and the secretion of the ARPs in vitro; if the

229

isolated ST had not been at stages VI-VIII, secretion of the ARPs would have been minimal (Sharpe et al., 1991b). The present findings show that depletion of round spermatids results in cessation of the secretion of ARP-2 whereas the secretion of ARP-6 and ARP-7 is decreased when either pachytene spermatocytes or round spermatids are depleted. The secretion of ARP-3 and ARP-4 was decreased variably in the absence of pachytene spermatocytes but was increased substantially in the absence of round spermatids. The latter may not be a real increase but may be a consequence of having to load twice as much ST-conditioned medium on the gels because of the reduction in into secreted incorporation of “‘S-methionine proteins at day 18 (see Results). Perhaps the most unexpected finding was that depletion of elongate spermatids had no effect whatsoever on secretion of the ARPs. This was considered surprising in view of the strong evidence that the presence or absence of elongate spermatids can exert major effects on a number of Sertoli cell functions such as the secretion of inhibin and androgen-binding protein as well as the production of seminiferous tubule fluid (JCgou et al., 1984; Pinon-Lataillade et al., 1988; Pineau et al., 1989; Sharpe, 1989; reviewed by JCgou, 1991). However, although depletion of elongate spermatids clearly had no effect on secretion of the ARPs, it did have the largest effect on the overall level of protein secretion (see below). One of the most prominent androgen-regulated proteins, ARP-2, virtually disappeared from stage VI-VIII secreted proteins when round spermatids were depleted, but was unaffected by depletion of other germ cell types. This could mean either that ARP-2 is a secretory product of round spermatids or that it is a product of the Sertoli cells (or peritubular cells), the secretion of which is absolutely dependent upon the presence of round spermatids (and testosterone). It remains to be shown conclusively which of these possibilities is correct, but our present (indirect) evidence favours the second possibility because: (1) ARP-2 is present in both seminiferous tubule fluid and testicular interstitial fluid (unpublished data), a feature characteristic of bidirectionallysecreted Sertoli cell proteins (see Sharpe, 1988);

(2) the secretion of ARP-2 at stages VI-VIII is minimal at 4 days following EDS-induced testosterone withdrawal (see Fig. 3), yet at this timepoint round spermatids are still present in virtually normal numbers (Sharpe et al., 1991b). It will be interesting in future studies to ascertain whether ARP-2 acts on round spermatids and whether this explains their modulation of its secretion (see below). It is also concluded tentatively that the decrease in secretion of ARP-6 and ARP-7 when either pachytene spermatocytes or round spermatids were depleted is not because it is a secretory product of these germ cells. Based on its molecular weight, pl, the presence of several charge isomers and its complete androgen-dependence (Sharpe et al., 1991b1, ARP-6 and ARP-7 are comparable to the A and B forms, respectively, of P-Mod-S, a peritubular cell product (Skinner et al., 1988). This is only a possibility, but if it is correct then the regulation of secretion of ARP-6 and ARP-7 at stages VI-VIII must involve a complex interaction between testosterone, pachytene spermatocytes, round spermatids, the Sertoli cells and peritubular cells, as it is presumed that the pachytene spermatocytes and/or round spermatids could not modulate the peritubular cells directly because of the intervening barrier of the inter-Sertoli cell tight junctions. Some of the ARPs (1, 3 and 4) did not show major repeatable changes in secretion after depletion of any one particular germ cell type. This might indicate either that their secretion is regulated co-ordinately by more than one germ cell type or that the presence of residual germ cells (germ cell depletion by MAA varies from 80 to 100%; see Fig. 1 and Bartlett et al., 1988) is able to maintain secretion of the ARPs at only marginally subnormal levels. The fact that the magnitude of change in secretion of these ARPs (1, 3 and 4) did not match the degree of depletion of the different germ cell types also provides reasonable (if indirect) evidence that these ARPs do not originate from any single one of the three most mature germ cell types. In contrast to the selective effects of germ cell depletion on secretion of the ARPs, the effect on overall protein secretion (as judged by the level of incorporation of ““S-methionine) was far less germ

230

cell-specific. Our previous study (Sharpe et al., 1991b) had shown that, during the normal spermatogenic cycle, protein secretion by isolated ST at stages VI-VIII was more than double that at earlier or later stages, a difference that was completely androgen-dependent. The present findings confirm this and, in addition, show that a full normal germ cell complement is essential if the normal increase in protein secretion at stages VI-VIII is to occur (Fig. 2). This conclusion is based on the finding that depletion of pachytene spermatocytes or round or elongate spermatids all led to a significant decrease in overall protein secretion by ST at stages VI-VIII when compared with controls; interestingly, this decrease was most pronounced when elongate spermatids were depleted, yet secretion of the ARPs was unaffected by the loss of this germ cell type. It is presumed that the androgen-regulated increase in secretion of proteins by ST at stages VI-VIII is due to parallel changes in the abundance and organization of Sertoli cell organelles such as the rough endoplasmic reticulum and mitochondria at stages VII-VIII (Kerr, 1988; Ueno and Mori, 1990). The present data could therefore be interpreted as suggesting that the three most mature germ cell types at stages VI-VIII may modulate the stage-specific changes in intracellular organelles of the Sertoli cell. However, it must also be recognized that such an effect of germ cells might occur independently of the stage of the spermatogenic cycle, based on findings in vitro using isolated Sertoli cells from immature rats (Djakiew and Dym, 1988; Onada and Djakiew, 1990, 199 1). These have shown that addition to the Sertoli cells of enriched preparations of either pachytene spermatocytes or round spermatids, or medium conditioned by these germ cells, leads to increases of up to 3-fold in the incorporation of radiolabelled methionine into proteins secreted by the Sertoli cells. An important deduction can be made about androgen action on spermatogenesis based on the present findings. Although MAA-induced depletion of any single germ cell type caused important changes in total protein secretion and the secretion of certain ARPs, these ‘abnormal’ changes

apparently have no effect on the remaining germ cell types at stages VI-VIII as these complete their development normally (see Bartlett et al., 1988; Ratnasooriya and Sharpe, 1989). This is in marked contrast to what happens when testosterone is withdrawn, as most of the meiotic and post-meiotic germ cells which pass through stage VII in androgen-depleted rats degenerate within the next 9 days (see Kerr et al., 1991). This implies that each of the ‘androgen-dependent’ germ cell types at stages VI-VIII (i.e. pachytene spermatocytes, round spermatids and elongate spermatids) controls androgen-dependent aspects of Sertoli (and peritubular) cell function which are specifically essential for its own continued development, but not those functions which are specific for other germ cell types; this would not of course preclude more than one germ cell type controlling the same function. If this interpretation is correct then it suggests, for example, that if ARP-2 is a secretory product of the Sertoli cell (see above), then its regulation specifically by round spermatids is indicative of its requirement by these germ cells alone. In conclusion, the present study has shown that androgen-dependent changes in ST-secreted proteins, which occur selectively at stages VI-VIII of the normal spermatogenic cycle in the rat, are largely dependent on a normal germ cell complement. Individual germ cell types may regulate the secretion of selected ARPs but there is also a modulatory effect of the overall germ cell complement on total protein secretion. As well as demonstrating the role of germ cells in modulating androgen-dependent, stage-specific changes in protein secretion, these findings add to the growing evidence for the complex functional inter-dependence of all of the cell types in the testis (Sharpe et al., 1990). This complexity probably explains the failure of previous studies, using isolated Sertoli cell cultures, to identify any major androgen-regulated proteins. Acknowledgements We are grateful to Denis Doogan and Mike Millar for skilled help and to Dr. Philippa Saunders for constructive comments.

231

References Allenby, G., Foster, P.M.D. and Sharpe, R.M. (1991a) Fundam. Appl. Toxicol. 16, 710-724. Allenby, G., Foster, P.M.D. and Sharpe, R.M. (1991b) Endocrinology 128, 467-476. Bartlett, J.M.S., Kerr, J.B. and Sharpe, R.M. (1988) J. Androl. 9, 31-40. Creasy, D.M., Flynn, J.C., Gray, T.J.B. and Butler, W.H. (I9851 Exp. Mol. Pathol. 43, 321-336. Djakiew, D. and Dym, M. (19881 Biol. Reprod. 39, 1193-1205. Foster, P.M.D., Lloyd, S.C. and Blackburn D.M. (1987) Toxicology 43, 17-30. Griswold, M.D., Morales, C. and Sylvester, S.R. (1988) in Oxford Reviews of Reproductive Biology, Vol. 10 (Clarke, J.R., ed.), pp. 124-161, Oxford University Press, Oxford. JCgou, B. (1991) Ann. NY Acad. Sci. (in press). JCgou, B., Laws, A.O. and de Kretser, D.M. (19841 Int. J. Androl. 7, 244-257. JCgou, B., Le Magueresse, B., Sourdaine, P., Pineau, C., Velez de la Calle, J.F., Garnier, D.-H., Guillou, F. and Boisseau, C. (1988) in The Molecular and Cellular Endocrinology of the Testis, Serono Symposium, Vol. 50 (Cooke, B.A. and Sharpe, R.M., eds.), pp. 255-270, Raven Press, New York. Kerr, J.B. (1988) Anat. Embryol. 179, 191-203. Kerr, J.B. and Sharpe, R.M. (1985) Endocrinology 116, 25922604. Kerr, J.B., Donachie, K. and Rommerts, F.F.G. (1985) Cell. Tissue Res. 242, 145-156.

Kerr, J.B., Millar, M., Maddocks, S. and Sharpe, R.M. (1992) Anat. Rec. (in press). Onoda, M. and Djakiew, D. (1990) Mol. Cell. Endocrinol. 73, 35-44. Onoda, M. and Djakiew, D. (1991) In Vitro Cell. Dev. Biol. 27A, 215-222. Parvinen, M. (1982) Endocr. Rev. 3, 404-417. Parvinen, M., Vihko, K.K. and Toppari, J. (1986) Int. Rev. Cytol. 104, 115-129. Pineau, C., Velez de la Calle, J.F., Pinon-Lataillade, G. and JCgou, B. (1989) Endocrinology 124, 2720-2728. Pinon-Lataillade, G., Velez de Calle, J.F., Viguier-Martinez, MC., Garnier, D.H., Folliot, R., Maas, J. and JCgou, B. (1988) Mol. Cell. Endocrinol. 58, 51-63. Ratnasooriya, W.D. and Sharpe. R.M. (19891 Int. .I. Androl. 12, 44-57. Sharpe, R.M. (1988) Int. J. Androl. 11, 87-91. Sharpe, R.M. (1989) J. Androl. IO, 304-310. Sharpe, R.M., Maddocks, S. and Kerr, J.B. (1990) Am. J. Anat. 188, 3-20. Sharpe, R.M., Bartlett, J.M.S. and Allenby, G. (1991a) J. Endocrinol. 128, 359-367. Sharpe, R.M., Maddocks, S., Millar, M., Saunders, P.T.K., Kerr, J.B. and McKinnell, C. (1991b) J. Androl. (in press). Skinner, M.K., Fetterrolf. P.M. and Anthony, C.T. (1988) J. Biol. Chem. 263, 2884-2890. Ueno. H. and Mori, H. (1990) Biol. Reprod. 43, 769-776.