Ontogeny and cellular origin of an interleukin-1 -like factor in the reproductive tract of the male rat

International Journalof Andrology, 1988,11, pages 437-447

Ontogeny and cellular origin of an interleukin- 1- like factor in the reproductive tract of the male rat

V. S Y E D t , 0. S O D E R , S . ARVER*, M. LINDH, S . KHAN t and E. M. RITZEN and *Department of Endocrinology, Karolinska Hospital, Stockholm, Sweden

Paediatric Endocrinology Unit,

Summary We have recently isolated an interleukin-1 (IL-1)-like factor from the rat testis, which originates from the seminiferous tubules and is a protein with an MW of 17 000 and a PI of 5-6. This paper reports on the appearance of the IL-l-like factor during postnatal development and investigates its cellular origin further. IL-1 activity was measured by a murine thymocyte proliferation assay. Very low IL-1 activity was present in culture medium conditioned by seminiferous tubules from rats aged 10 or 20 days. From 30 days of age, increasing amounts were detected, reaching a maximum level in adult animals (60-90 days). N o IL-1 activity was found in medium conditioned by peritubular cells. Sertoli cell-enriched seminiferous tubules obtained from experimentally cryptorchid or from prenatally irradiated rats produced much higher levels of IL-1 activity than did those obtained from intact testes. IL-1 activity was detected in efferent duct fluid after ligation of the efferent ducts for 24 h, indicating that the IL-l-like factor was secreted into the tubular lumen. Low levels of IL-1 activity were detected in extracts of epididymal tissue and epididymal sperm, whereas ejaculated seminal plasma, seminal vesicle fluid and extracts of seminal vesicles (together with the coagulating glands) and ventral and dorsolateral prostate lacked IL-1 activity. Instead, seminal plasma inhibited testicular IL-1 activity dose-dependently without affecting cell viability in the thymocyte cultures. Although its biological function remains to be defined, our results indicate that the testicular IL-l-like factor is produced by Sertoli cells and that its appearance during development coincides with the initiation of active spermatogenesis in the rat testis.

Keywords: interleukin-1 , testis, Sertoli cells, seminiferous tubules, germ cells, growth factors, developmental changes.

Introduction In a recent paper (Khan et a l . , 1987), we reported the isolation of an interleukin-1 (IL-1)-like factor from the intact testis of the adult rat. The biological function of

t Present address: Max Planck Clinical Research Institute for Reproductive Medicine, Steinfurter str. 107, D-440 Munster, FRG.

Correspondence: Professor E. Martin RitzCn, Paediatric Endocrinology Unit, Karolinska Hospital, S-10401 Stockholm, Sweden.

437

438 V. Syed et a1

this factor is unknown, but since it seems to be produced by the normal testis it might play an important physiological role. The biochemical properties of the testicular IL-l-like factor were very similar to those of macrophage-derived IL-la with an estimated relative molecular mass (MW) close to 17 000, and an isoelectric point (PI) of 5-6. It was also demonstrated that the IL-l-like activity was present in extracts and medium conditioned by cultures of normal and cryptorchid seminifer- ous tubules but not of interstitial cells. Nor was there any activity in culture media conditioned by 10-day-old rat Sertoli cells. Taken together, these findings indicated that the testicular IL-l-like factor was produced by cells in the seminiferous tubules, most probably by Sertoli cells, in adult but not in prepubertal rats.

In the present study we have investigated further the cellular and tissue origin of the IL-l-like factor in the reproductive tract of the male rat. The appearance of the IL-l-like activity during postnatal development of the rat testis was also studied.

Materials and methods

Experimental animals Male Sprague-Dawley rats purchased from ALAB, Stockholm, Sweden, were used as donors of testes and accessory sex glands and secretions. All rats were aged 60-90 days old unless stated otherwise. NMRI mice aged 6-8 weeks were used as thymocyte donors for the IL-1 bioassay.

Tissue extractions Extracts were prepared from pooled decapsulated testes, epididymides, ventral and dorsolateral prostates, seminal vesicles together with the coagulating glands, and thoroughly washed spermatozoa collected from three rats by retrograde flushing of the caudal portion of the epididymis with saline. The tissues and spermatozoa were frozen and thawed and then homogenized in tissue culture medium (1 : 4 w/v; cx- modified Eagle’s Minimum Essential Medium; a-MEM, Flow Labs, Irvine, U.K.). After centrifugation at 2000 x g for 15 min the supernatants were filtered through a 0.22-pm filter and stored at -20°C until assayed.

Collection of ejaculates and seminal vesicle fZuid Ejaculates delivered upon decapitation of two rats were collected. Seminal vesicle Ruid was obtained by extrusion from the seminal vesicles of the same rats. The collected fluids were diluted 1 : 4 (v/v) in a-MEM, mixed thoroughly and centrifuged at 2000 x g for 10 min. After filtration through a 0.22-p.m filter, samples were frozen at -20°C until assayed.

Preparation of seminiferous tubules from different age groups Rats at 10,20, 30,40,50,60 and 90 days of age were killed by decapitation and the testes collected. An equal weight of decapsulated testicular tissue from each age group was used for isolation of seminiferous tubules. Testes were treated with trypsin (2.5 mglml; Sigma, St Louis, MO, U.S.A.) for 25 min at 32-34°C followed by soybean trypsin inhibitor (5 mglml; Sigma, St Louis, MO, U.S.A.). The isolated seminiferous tubules were then washed in Medium 199 (Flow Labs, Irvine,

Testicular IL-1 -like factor 439

U.K.) and incubated for 2 h at 32-34°C. After washing, the tubules were cut into 2-mm pieces and incubate,d for 24 h at 32-34°C in Medium 199 containing 0.2% BSA, 20 mM L-glutamine, 100 IU/ml penicillin and 50 kg/ml streptomycin. The spent medium was then collected, centrifuged at 1000 x g for 10 min and stored at - 20°C until assayed.

Induction of cryptorchidism Rats aged 50 days were made cryptorchid by suturing the scrotum to the abdominal wall. Four weeks after the operation the cryptorchid testes and the control testes from sham-operated animals were collected. Seminiferous tubules were isolated and incubations performed as described above. The conditioned media were collected and then analysed for IL-1 activity.

Preparation of Sertoli cell-enriched seminiferous tubules Pregnant rats were irradiated with 1.5 Gy as a single dose (6"Cobalt source, focal length 100 cm, 0.011 Gy/sec) at day 19 of gestation in order to kill germ cell precursors in the fetuses (Beaumont, 1960). The male offspring were killed at an age of 90 days and the Sertoli cell-enriched seminferous tubules collected and incubated as described above. The conditioned media were collected and taken for analysis of IL-1 activity.

Preparation of peritubular cells Pertibular cells were prepared from isolated seminiferous tubules from rats aged 90 days by prolonged treatment with collagenase, according to the method of Mather & Phillips (1984). Seminiferous tubules were freed from interstitial cells by incubation with collagenase (0.25 mg/ml; type I; Sigma, St Louis, U.S.A.) for 15 min at 37°C. Peritubular cells were removed from the tubules by prolonged incubation with collagenase for 25 min. The released peritubular cells were collected from the medium, adjusted to a concentration of lo5 cells per ml and plated in petri-dishes (Falcon 35 mm) in a final volume of 2 ml. After 1 h, non- attached cells were removed by washing with medium and the adherent cells further incubated at 32-34°C for 24 h. The conditioned medium was then collected, centrifuged at 1000 X g and stored at -20°C until tested. Seminiferous tubules depleted of peritubular cells by the above procedure were also incubated and the spent media collected. As a control for the effect of prolonged collagenase treatment, isolated peritubular cells were combined with seminiferous tubules depleted of such cells and incubated as described above to produce conditioned medium.

Efferent duct ligation Rats aged 60 - 70 days were anaesthetized with diethyl ether and the efferent ducts of the testes ligated via abdominal incision, either unilaterally or bilaterally, as described by French & RitzCn (1973). After 24 h, the efferent duct fluid was collected, centrifuged at 1000 x g for 10 min to remove sperm and stored at - 20°C. The testes were then dissected out and seminiferous tubule cultures set up as described above and their culture media collected after 24 h of incubation. The

440 V. Syed et al.

epididymides were also collected, divided into the caudal and caput portions and then homogenized with Medium 199 (1 : 4 w/v) and centrifuged at 1000 x g for 15 min. The supernatant was further centrifuged at 100000 x g for 1 h and then stored at - 20°C until assayed.

Interleukin-1 assay IL-1 activity was assessed by a murine thymocyte proliferation assay, essentially as described previously (Gery, Gershon & Waksman, 1972; Soder, 1987). Material to be tested was added from the start and the cultures were terminated and the cells collected at 48 h, following a 2 h pulse with tritiated thymidine. Incorporated radioactivity was determined by liquid scintillation counting and expressed as mean counts per minute (cpm) of duplicate cultures. Dose-response curves were established for all test samples by assaying serial dilutions of the sample, usually 6 doses of four-fold dilutions starting at a final concentration of 10% (by volume). The IL-1 activity of an individual sample was expressed either as the obtained dose-response curve or as activity units per ml. One unit of IL-1 was defined arbitrarily as the amount needed to double the incorporation of tritiated thymidine by the control cultures. In all cases when IL-1 units of different samples were compared it was ascertained that their dose-response curves were parallel.

Assay of the inhibitory action of seminal plasma Seminal plasma was obtained from ejaculates as described above. The seminal plasma was diluted 1 : 4 in u-MEM and centrifuged and passed through a 0.22-km filter to remove cells and particulates. Low-molecular weight components (MW < 5000) were removed by chromatography on a Sephadex G25 (PD-10) column run in 0.15 M sodium chloride. Material eluting in the void volume was collected and added to thymocyte cultures in serial dilutions in the presence or absence of rat testicular extract containing a suboptimal amount (75% of maximum response) of the testicular IL-1 activity. The cultures were then incubated in the presence of phytohaemagglutinin and collected as described above.

Cytotoxicity nssay Thymocytes are very sensitive to cytotoxic influences (see for example, Shortman & Jackson, 1974; Sandberg & Olsson, 1984). Therefore, murine thymocytes utilized in the IL-1 assay were also employed as target cells in a cytotoxicity test of seminal plasma. Rat seminal plasma collected and chromatographed as described above was added to thymocyte cultures at a final dilution of 1 : 640 (by volume). This dilution of seminal plasma induced 54% inhibition of the uptake of tritiated thymidine by the thymocytes in the presence of a suboptimal concentration of rat testicular IL-1 (see Results). Control cultures were incubated with 100 Fg/ml bovine serum albumin (BSA). After 24 and 48 h of culture the number of living cells was estimated by measuring cell concentration in an automatic cell counter (Cellcounter 134, Analysinstrument, Stockholm, Sweden) and the percentage of unstained cells after incubation with Trypan blue. The fraction of Trypan blue- stained and unstained cells was obtained by examining 200 cells under the light microscope. The results are expressed as mean (+ range) values of duplicate measurements in duplicate cultures.

Testicular IL-1-like factor 441

Results

Interleukin-1 activity in genital tract secretions and organ extracts Aqueous extracts of rat testes contained high levels of IL-1 activity (Fig. 1). Extracts of epididymal sperm and epididymal tissue displayed weak but significant IL-1 activity. No IL-1 activity was found in cell-free seminal plasma, seminal vesicle fluid or extracts of seminal vesicles (together with the coagulating glands) and ventral and dorsolateral prostate (Fig. 1).

Q C ._ 0 E 5 I

0 I

i 0 160 6iO 2560 102LO Reciprocal dilution

Fig. 1. IL-1 activity in secretions and organ extracts of the genital tract frommale rats. Testicular extract (A) used as a positive control showed high IL-1 activity, and activity was also detected in ductal fluid from ligated testes (W) and in extracts of epididymides (* 7 ) and spermatozoa (0). NO IL-1 activity was detected in seminal plasma, seminal vesicle fluid, or in extracts of seminal vesicle tissue together with the coagulating glands or extracts of the dorsolateral and the ventral prostate (W). IL-1 activity has been expressed as dose-response curves obtained from thymocyte cultures given serial dilutions of the above material. Values are mean cpm f range of duplicate cultures.

Cellular origin of testicular IL-1 activity The cellular origin of the testicular IL-1 activity was explored by measuring IL-1 activity in conditioned media from various testicular cell incubations. Sertoli cell- enriched testes produced the largest amount of IL-1 activity. Thus, conditioned media from seminiferous tubules isolated from testes of prenatally irradiated rats or experimentally cryptorchid rats contained higher IL-1 activity when compared with the respective controls (Fig. 2a,b). By contrast, very low IL-1 activity was found in media conditioned by peritubular cell incubations. Removal of peritubular cells did not alter the production of IL-1 activity by the cultured seminiferous tubules (Fig. 2a). Nor was there any change in the production of IL-1 activity when peritubular cell-depleted tubules and separated peritubular cells were recombined and incubated for 24 h (data not shown).

442 V. Syed et al.

I O O O r

Tested materiol

Fig. 2. Cellular origin of the interleukin-1 (IL-1)-like factor in the rat testis. Conditioned media from incubations of various testicular cell preparations were analysed for IL-1 activity in thymocyte cultures. The results are expressed as IL-1 units k SEM ( n = 4) per ml of conditioned medium. A: IL-1 activity in conditioned media from cultures of intact (a ) and peritubular cell-depleted (b) seminiferous tubules, peritubular cells (c) and seminiferous tubules from prenatally irradiated rats (d). B: IL-1 activity in conditioned media from cultures of seminiferous tubules from sham-operated (a) or experimentally cryptorchid (b) rats.

IL-1 activity in efferent duct-ligated testes Unilateral ligation of the efferent ducts of the testis for 24 h caused an increase in testicular weight when compared with the non-ligated side (2.26 vs 1.76 g) indicating accumulation of seminiferous and rete testis fluid. IL-1 activity was detected in efferent duct fluid from ligated testes demonstrating secretion of the IL-1-like factor into the tubular lumen. The amounts present are difficult to judge as the dose-response curve indicates the presence of inhibitory material interfering in the assay (Fig. 1). Extracts of ligated and non-ligated testes showed no difference in IL- 1 activity (data not shown) suggesting that testicular IL-1 is not accumulated in the seminiferous tubule lumen under the present experimental conditions.

Ontogeny of testicular IL-I activity Seminiferous tubules were isolated from the testes of rats of different ages, incubated for 24 h and their conditioned media analysed for IL-1 activity. Little if any IL-1 activity was detected in conditioned media of seminiferous tubules from rats aged 10 or 20 days. Increasing amounts were present from the age of 30 days, reaching a maximum level at adult age (60-90 days) (Fig. 3).

Inhibitory action of seminal plasma When cell-free seminal plasma was tested for the presence of IL-1 activity an inhibitory effect on thymidine incorporation by the cultured thymocytes was observed. This inhibitory activity remained after removal of low-molecular weight (MW < 5000) components by passage through a Sephadex G2.5 column. The chromatographed seminal plasma exhibited a dose-dependent inhibitory action when added to murine thymocyte cultures that had been stimulated with a suboptimal amount (75% of maximum response) of testicular IL-1 activity (Fig. 4). This inhibitory effect was not due to a cytotoxic action of seminal plasma since the number of living cells were identical in treated and control cultures at both 24 and

Testicular IL-1 -like factor 443

250 r 1

10 20 30 LO 50 60 90

Age (days1

Fig. 3. Ontogeny of the interleukin-I, (IL-1)-like factor in the rat testis. Seminiferous tubules were isolated from rats of different age grolrps, incubated at 32-34°C for 24 h and their conditioned media collected and analysed for IL-1 activity. Results have been expressed as mean units of IL-1 ir SEM (n = 4) per ml of conditioned medium. Each age group represents testes from 8-35 rats. The inset figure shows the dose-response curve of IL-1 activity (mean cpm f SEM of duplicate cultures) obtained with medium conditioned by seminiferous tubules from rats aged 90 days. All dose-response curves of the different age groups were parallel.

0 0 01 O O L 0 16 0 64

Concentration of seminal plasma ( % v l v l

Fig. 4. Inhibitory effect of seminal plasma on testicular IL-1-like activity. Ejaculated rat seminal plasma was separated on a Sephadex G25 column and serial dilutions of material recovered in the void volume (MW > 5000) was added to thymocyte cultures together with a suboptimal amount (75% of maximum activity) of the rat testicular IL-1-like factor. The cultured cells were collected at 48 h after a 2 h pulse with tritiated thymidine. The horizontal axis gives the final reciprocal dilution of seminal plasma in the thymocyte culture medium. Results have been expressed as IL-1 activity (cpm k SEM of duplicate cultures) as a percentage of controls containing testicular IL-1-like factor alone (= 100%: 68 080 ir 6440 cpm). Negative controls (no testicular IL-1-like factor) produced 4023 f 140 cpm.

48 h of incubation. Thus, in the presence of seminal plasma (1 : 640 final dilution causing a 54% inhibition of thymidine uptake by the murine thymocytes) the number of living cells was 5.2 (4.9-5.5) x lo5 (mean and range of duplicate determinations) per culture at 24 h and 4.1 (3.7-4.5) X lo5 at 48 h. The corresponding figures in control cultures (100 ~g/mI BSA) were 5.0 (4.9-5.1) x105 and 4.2 (4.0-4.4) X lo5, respectively.

444 V. Syed et al.

Discussion We have shown recently that the adult rat testis produces large amounts of an IL-1- like factor with molecular properties similar to those of macrophage-derived IL-la (Khan et al., 1987). It was also shown that the IL-1 activity was present in cryptorchid testes depleted of maturing germ cells but not in culture media of Sertoli cells from rats aged 10 days. These previous findings indicated that the testicular IL-l-like activity originated from Sertoli cells of adult but not prepubertal rats. The results presented here further support this assumption. Thus, little IL-1 activity was produced by seminiferous tubules from rats aged 10-20 days, whereas from day 30 IL-1 activity increased gradually to reach a maximum level in adult animals. The highest IL-1 activity was detected in Sertoli cell-enriched testes. Both cryptorchid and prenatally irradiated testes contain very few if any germ cells, and as peritubular cells failed to produce IL-1 activity in culture, no cells other than Sertoli cells are left as major candidates for the production of testicular IL-1. However, since we have not studied isolated germ cells, it cannot be excluded that these cells may also produce IL-1 activity in amounts too small to be detected by comparison of intact and germ cell-depleted testes. Although monocytes/macro- phages are major producers of IL-1 in inflamed tissues and in cell cultures (Dinarello, 1984; Krakauer, 1986), several observations make it unlikely that these cells constitute a major source of testicular IL-1 activity. Firstly, macrophages are scarcely found in the seminiferous tubules under normal conditions (El-Demiry et al . , 1986; Pollanen & Niemi, 1987), and no IL-1 is produced unless macrophages are activated (Dinarello, 1984; March et al., 1985). Secondly, although mononuclear phagocytes produce both IL-la (PI 5 ) and IL-1p (PI 7), the p-form is predominant and constitutes more than 90% of the IL-1 protein secreted by macrophages (March et al. , 1985). This is in contrast to the testicular IL-1 activity which seems to be only of a-type without detectable amounts of IL-1P-like material (Khan et al., 1987).

The presence of IL-1 activity in sperm could reflect uptake from Sertoli cell secretions, although no direct evidence of such a transfer exists at present. The rat sperm used for extraction were isolated by flushing of the epididymides and thus contained no particulate or cellular contributions from the genital tract distal to the epididymis. The present findings are in accordance with our previous observations of IL-1 activity in ejaculated human sperm (Arver & Soder, 1986). A spermatozoa1 origin might be suggested for the observed weak epididymal activity since the content of the epididymal duct was included in the extract.

Our observation that the IL-1 content of the testis was unchanged after ligation of the efferent ducts, suggests that it is not accumulated in the tubular lumen, and that it is produced in the testis to exert a local function. This is in contrast to, for example, androgen-binding protein, which accumulates in the testis after efferent duct ligation (French & RitzCn, 1973). An interesting finding in the present study is the coincidence of the first appearance of IL-1 activity in the testis and the onset of active spermatogenesis. This indicates that testicular IL-1 might be involved in the spermatogenic process. Since IL-1 is mitogenic for a wide variety of target cells (see for example, Oppenheim & Gery, 1982; Soder & Madsen, 1988), it could be

Testicular IL-1 -like factor 445

speculated that it is involved in regulation of active proliferation in the testis. Recently, several other peptide growth factors have been detected in the testis. Of these, insulin-like growth factor-1 (Chatelain, Naville & Saez, 1987; Smith et al., 1987) and epidermal growth factor (Holmes, Spotts & Smith, 1986; Tsutsumi, Kurachi & Oka, 1986) are particularly interesting as they have been suggested as germ cell growth and maturation factors. Although these factors are molecularly different from the testicular IL-l-like factor, i t remains to be determined whether they exert similar functions in the testis. Even if testicular IL-1 originates from the seminiferous tubules, part of its action may be directed towards the interstitium. One possibility is that it is involved in the response of the testis to an acute gonadotrophic stimulation. In studies on testicular microcirculation, Widmark, Damber & Bergh (1986) noticed recently that an injection of hCG or LH causes an acute inflammatory reaction in the testis within a few hours. This aspect of IL-1 function in the testis is currently under investigation.

We were unable to detect any IL-1 activity in seminal plasma. Instead, a potent inhibitory action was found, which was not due to prostaglandins, polyamines and other low molecular weight inhibitors as the inhibitory activity was recovered in the void volume (MW > 5000) after Sephadex G25 chromatography. N o cytotoxic effect of seminal plasma was found. Whether this inhibitory effect is specific for IL- 1 is presently unknown. Inhibition was observed also in the absence of the testicular IL-l-like factor, but this effect might well be due to inhibition of the small amounts of endogenously produced IL-1 (Gery, Gershon & Waksman, 1972) in the thymo- cyte cultures. Recently, several authors have reported on the isolation of IL-1 binding proteins from various sources (Berman et al., 1986; Muchmore & Decker, 1986). These binding proteins inhibit IL-1 bioactivity in several assays and have been suggested to take part in the regulation of IL-1 actions in different tissues. Such a regulatory effect might also be suggested for the inhibitory activity in seminal plasma, but its relationship, if any, to the reported IL-1 binding proteins remains to be shown. Our present findings in the rat are supported by ongoing studies with human seminal plasma demonstrating similar high MW IL-1 inhibitory activity lacking thymocytotoxic effects (Soder & Arver, 1987).

As has been discussed previously (Khan et al., 1987), final identification of the testicular IL-l-like factor must await disclosure of its primary structure. Our own studies in progress have confirmed that the molecular weight is close to 17 000 (SDS-PAGE; HPLC gel exclusion) and have demonstrated a more precise PI of 5.7 (chromatofocussing). These figures are identical to those reported for an IL-la-like factor isolated from rat skin (Schmitt et al., 1986). The seminiferous growth factor (SGF), a protein with an MW of 15-17 000 isolated from seminiferous tubule cytosols and Sertoli cell cultures, and reported to stimulate the proliferation of cultured fibroblasts (Feig et al . , 1980), obviously shares some molecular properties with the testicular IL-l-like factor. As IL-1 also stimulates fibroblast proliferation (Postlethwaite, Lachman & Kang, 1984) it is important to note that, in contrast to the testicular IL-1 activity, SGF is produced predominantly by the prepubertal testis (BellvC et al., 1986), indicating that SGF and the testicular IL-l-like factor are separate entities.

446 V. Syed et al.

Acknowledgments The skilful technical assistance of Ms C. Fahlgren and Ms M. Widing is acknowledged gratefully. This work was supported by grants from the Swedish Medical Research Council (project nos 3168 and 08282), the Swedish Society of Medicine, Ake Wiberg Foundation and the King Gustav V Jubilee Foundation.

References Arver, S. & Soder, 0. (1986) Interleukin-1 activity in human testis and spermatozoa. Abstract, 4rh

International Forum of Andrology, Paris. Bellve, A. R., Stewart, T. A., Braunhut, S. J., Ernisse, B. J. & Rufo, Jr, G. A. (1986) Regulation of

cell proliferation during development of the mammalian testis. In: Developmenf and Funcfion of the Reproductive Organs, Serono Symposia Review No. 11 (eds A. Eshkol, B. Eckstein, N . Dekel, H . Peters, and A. Tsafriri), pp. 171-183. Ares-Serono Symposia, Rome 1986.

Beaumont, H. M. (1960) Changes in the radiosensitivity of the testis during the neonatal period. Journal of Reproduction and Fertility, 44,25-42.

Berman, M. A., Sandborg, C. I., Calabia, B. S., Andrews, B. S. & Friou, G. J . (1986) Studies of an interleukin 1 inhibitor: characterization and clinical significance. Clinical and Experimental Immunology, 64,136-145.

Chatelain, P. G. , Naville, D. & Saez, J . M. (1987) Somatomedin-Ciinsulin-like growth factor 1-like material secreted by porcine Sertoli cells in vitro: characterization and regulation. Biochemical and Biophysical Research Communications, 146, 1009-1017.

Dinarello, C. A. (1984) Interleukin-1. Review oflnfecfious Diseases, 6, 51-95. El-Demiry, M., Hargreave, T. B., Busuttil, A. , Crisholm, G . D. & James, K. (1986) Abstract, 6rh

International Congress of Immunology, Toronto, Canada 1986. Feig, L. A., Bellve, A. R., Horbach-Erickson, N. & Klagsbrun, M. (1980) Sertoli cells contain a

mitogenic polypeptide. Proceedings of the National Academy of Sciences, U.S.A., 77.47744778. French. F. S. & RitzCn, E. M. (1973) Androgen-binding protein in efferent duct fluid of the rat testis.

Journal of Reproduction and Fertility, 32,479483. Gery, I., Gershon. R. K. & Waksman. B. H. (1972) Potentiation of the T-lymphocyte response to

mitogens. I . The responding cell. Journal ofExperimenta1 Medicine, 136, 128-142. Holmes, S. D., Spotts, G . & Smith, R. G. (1986) Rat Sertoli cells secrete a growth factor that blocks

epidermal growth factor (EGF) binding to its receptor. Journal of Biological Chemistry, 261. 40764080.

Khan, S., Soder, O., Syed, V.. Gustafsson, K., Lindh, M. & RitzCn, E. M. (1987) The rat testis produces large amounts of an interleukin-1-like factor. InternationalJournal ofAndrology, 10,495-503.

Krakauer, T. (1986) Human interleukin 1. CRC Critical Reviews in Imnzunology, 6,213-244. March, C. J . . Mosley, B., Larsen, A , , Pat Ceretti, D., Braedt, G., Price, V.. Gillis, S. , Henney, C. S . ,

Kronheim, S . R., Grabstein, K., Conlon, P. J . , Hopp, T. P. & Cossman, D. (1985) Cloning, sequence and expression of two distinct human interleukin-1 complementary DNAs. Nature, 315, 641-647.

Mather, J. P. & Phillips, D. M. (1984) Primary culture of testicular somatic cells. In: Methods for Serum-free Culture of Cells of the Endocrine System Vol I1 (eds D . W. Barnes, D. A. Sirbasku and G. H. Sato). pp. 29-45. Alan Liss Inc., New York.

Muchmore, A. V . & Decker, J . M. (1986) Uromodulin; an immunosuppressive 85-kilodalton glycoprotein isolated from human pregnancy urine is a high affinity ligand for recombinant interleukin-la. Journal of Biological Chemistry, 261, 13404-13407.

Oppenheim, J . J. & Gery, I. (1982) Interleukin-1 is more than an interleukin. Immunology Today, 3,

Pdlanen, P. & Niemi, M. (1987) Immunohistochemicai identification of macrophages, lymphoid cells and HLA antigens in the human testis. It~temafionalJournaI ofAndrology, 10,3742.

Postlethwaite, A. E., Lachman, L. B. & Kang, A. H. (1984) Induction of fibroblast proliferation by interleukin-1 derived from human monocyte leukemia cells. Arthritis and Rheumatism, 27, 995- 1001.

113-119.

Testicular IL-1 -Like factor 447

Sandberg. G. & Olsson, M. (1984) Studies on thymocyte subpopulations in guinea pigs. IV. A subpopulation sensitive to a natural, cold reactive cytotoxic factor in autologous serum. International Archives of Allergy and Applied Immunology, 13,32-37.

Schmitt, A., Hauser. C., Jaunin, F.. Dayer, J . M. & Saurat, J. H. (1986) Normal epidermis contains high amounts of natural tissue IL-1. Biochemical analysis by HPLC identifies a MW 17Kd form with a PI 5.7 and a MW 30 Kd form. Lymphokine Research, 5,105-118.

Shortman, K. & Jackson, H. (1974) The differentiation of T lymphocytes. I . Proliferation kinetics and interrelationships of subpopulations of mouse thymus cells. Cellular Immiinology, 12,230-246.

Smith, E . M. , Svoboda, M. E., Van Wyk. J. J . , Kierszenbaum, A. L. & Tres, L. L. (1987) Partial characterization of a somatomedin-like peptide from the medium of cultured rat Sertoli cells. Endocrinology, 120,186-193.

Soder. 0. (1987) Isolation of interleukin-I from human milk. International Archives of Allergy and Applied Immunology, 83, 19-23.

Soder, 0. & Arver, S. Inhibition of testicular interleukin-1 activity by seminal plasma (1987) Abstract, VII Workshop on Development and Function of the Reproductive Organs, Turku, Finland, 1987.

Soder, 0. & Madsen, K. (1988) Stimulation of chondrocyte DNA synthesis by interleukin-1. British Journal of Rheumatology, 27-21-26.

Tsutsumi, O. , Kurachi, H. & Oka, T. (1986) A physiological role of epidermal growth factor in male reproductive function. Science, 233,975-977.

Widmark, A.. Damber, J.-E. & Bergh, A. (1986) Relationship between human chorionic gonado- trophin-induced changes in testicular microcirculation and the formation of testicular interstitial fluid. Journal of Endocrinology, 109,419425.

Received 27 July 1987; accepted (after revision) 15 March 1988