Immunochemical localization of secretory antigens in the human epididymis and their association with spermatozoa

Ejaculated spermatozoa were washed and extracted with 0.6 M NaCl (2 h at 0 degree C) and the extract used to immunize rabbits. The crude antibody reacted with epididymal fluid and cytosol and with prostatic cytosol but did not recognize blood serum and testicular cytosol. After adsorption with prostatic proteins, the serum was specific for epididymis. Using immunoelectrophoresis and affinity chromatography, it was found that the antibody reacted with antigens which co-electrophoresed with androgen-dependent proteins (mobility relative to albumin, Ra) 0.3, 0.43 and 1.0, previously identified in human epididymis. Weak immunofluorescence in the epithelium of proximal caput tubules was detected on tissue sections. In contrast, distal caput and corpus tubules displayed a strong fluorescence in the cytoplasm of basal and principal cells as well as in spermatozoa present in lumen. Intense fluorescence was limited to the luminal content and the apical border and sterociliae of principal cells in caudal tubules. When applied to isolated spermatozoa, the reaction was negative for testicular sperm, while 49%, 82% and 100% of spermatozoa from caput, corpus and cauda, respectively, had a fluorescent acrosomal cap. An apparent gradient of increasing fluorescent intensities was also observed in this sequence. The reaction was strongest over the acrosomal cap, apparently absent in the postacrosomal region and weaker over the midpiece and principal piece. These results are interpreted as suggestive of the progressive coating of human spermatozoa with androgen-dependent epididymal proteins during epididymal transit.     

Immunopathology of murine experimental allergic orchitis

Experimental allergic orchitis (EAO) was induced consistently in BALB/c mice by immunization with homologous testicular tissue homogenate emulsified in complete Freund's adjuvant (CFA) providing that the animals had received simultaneously at least 1 microgram of an extract of Bordetella pertussis rich in pertussigen. All animals thus treated developed orchitis and serum antibody to testicular antigens within 20 days after immunization. The lesions were located in testis (100%), rete testis (37%), cauda epididymis (21%), and vas deferens (37%). Ductus efferentes and caput epididymis were only rarely affected. Early lesions in the seminiferous tubules were characterized by peritubular and/or intratubular accumulation of eosinophils, neutrophils, lymphocytes, and macrophages. This was followed by aspermatogenesis. Late lesions included massive necrosis and extensive fibrosis of the seminiferous tubules. Disruption of blood-testis barrier on day 20 was evidenced by the detection of 1) perfused lanthanum deposits between Sertoli cells and surrounding inflammatory cells inside the seminiferous tubules, 2) deposits of endogenous mouse IgG in germinal epithelium, and 3) probable immune complexes (granular C3) surrounding seminiferous tubules. Murine EAO differed from that of the guinea pig in the lack of involvement of the ductus efferentes, the extensive necrosis, the abundant polymorphonuclear eosinophils in the lesion, and the exquisite requirement of concomitant injection of B. pertussis extract.     

Lectin staining of rat testis and epididymis after ligation of excurrent ducts at different levels

Seven rhodamine-conjugated lectins (PNA, RCA I, SBA, Con A, WGA, UEA I, and DBA; see Table 1) were utilized in studying the staining pattern of glycoproteins in rat testis and epididymis after ligation of ductuli efferentes (DE), corpus epididymidis (CE), and vas deferens (VD) for various time periods. Ductuli efferentes ligation caused a widening of seminferous tubules and detachment of spermatids with formation of multinuclear cells. These cells acquired a strong affinity for all lectins. Corpus epididymidis ligation also caused degeneration of spermatids with increased lectin staining in some tubules, but after 7 days another cell population close to the periphery of seminiferous tubules showed an increased nuclear affinity for some lectins followed by a clear degeneration and strong cytoplasmic staining with all lectins. Vas deferens ligation caused no degenerative changes in testicular spermatids. However, the peripheral cell population showed degenerative changes similar to those found after CE ligation. In both cases this was coincident with the formation of spermatic granulomas at the site of ligation. Ductuli efferentes ligation caused a gradual decrease of intratubular content in caput epididymidis, while the contrary was true after CE ligation. The latter was associated with intratubular accumulation of lectin-positive swollen cells and sperm aggregates as well as an increased lectin staining of narrow cells in initial segment and light cells in distal caput. After VD ligation an increased staining of light cells was initially found in distal cauda and distal caput, but, concomitant with distension of the tubules this reaction decreased.(ABSTRACT TRUNCATED AT 250 WORDS)     

The annual testicular cycle in an equatorial colony of lesser rock hyrax, Heterohyrax brucei.

Adult males from a colony of lesser rock hyrax found near the equator in Kenya exhibited an annual cycle of testicular activity characterized by intense spermatogenesis and elevated androgen status from May to July. Average masses of testes and seminal vesicles taken in these months were almost fourfold greater than those from September to January. During the months of peak testicular activity average diameters of Leydig cells and seminiferous tubules were increased by approximately one half and total tubule length was doubled, compared with values for the quiescent months. Variable testicular development occurred during transitional intervals preceding and following peak testicular activity. From February to Aril thickening of the seminiferous epithelium and appearance of spermatozoa in the caput epididymidis signalled re-establishment of sperm production. In August shedding of germinal cells from the epithelium heralded impending failure of spermatogenesis. Evidence of an annual testicular cycle contradicted the prevalent belief that equatorial hyrax breed all year and suggested that the testicular cycle is a conservative element of hyracoid reproductive strategy.     

Effects of a highly purified antiserum to FSH on testicular function in immature rats

Effects of highly purified antiserum (AS) to follicle stimulating hormone (FSH) on testicular function was studied in immature rats. Treatment with FSHAS for 10 days, from 25-34, decreased weights of the testis (p .001) and increased weights of the epididymis (p .05). Numbers of the cell types in the seminiferous epithelium, particularly Type A spermatogonia pachytene spermatocytes and spermatids, were markedly reduced, possibly due to: 1) decreased division of the initial stem cells, 2) impairment of division of Type B spermatogonia and their transformation to pachytene spermatocytes, and 3) desquamation and degeneration of pachytene spermatocytes and spermatids. FSHAS also affected the sertoli cell function which was reflected in the decreased binding of androgens to supernatant fraction of the testis and epididymides. Treatment with luteinizing hormone-AS for 5 days did not affect testicular function but the binding of androgens to the supernatants of the caput and cauda epididymides and ventral prostate was significantly reduced (p .001). These data indicate that FSH is necessary for the maintenance of the cellular integrity of the seminiferous epithelium during the completion of the 1st wave of spermatogenesis.     

Organization of tubules in the human caput epididymidis and the ultrastructure of their epithelia

The structure of the human caput epididymidis was examined by gross morphological and light and electron microscopic techniques. There were at least seven types of tubules, each characterized by a different epithelium. These tubules were connected with one another by at least eight types of junctions to form a network. Most of the caput epididymidis was composed of efferent ducts. Within these, five types of tubules, each with a different ciliated epithelium, were found in different regions; and four types of junctions between the efferent ducts and the epididymal tubule were observed. The efferent ducts left the testis, initially as parallel straight tubules containing both ciliated and non-ciliated cells in an epithelium of irregular height. Each efferent duct then coiled tortuously into lobules that folded over one another. These efferent ducts then branched out as thin tubules to join a network of dark tubules which were lined by a regular epithelium containing prominently vacuolated, non-ciliated cells. These tubules anastomosed via common cavities characterized by a ciliated cuboidal epithelium and sometimes joined tubules exhibiting a non-vacuolated ciliated epithelium. The latter, as well as typical efferent ducts, made connection with the epididymis proper in both end-to-end and end-to-side junctions. In the more distal junctions with the epididymis, the efferent ducts joined to a transitional epididymal ductule before joining to the side of the epididymis proper. Post-junctional epithelia in the beginning of the epididymis occasionally contained patches of cells characteristic of efferent ducts. Tall cells with long stereocilia constituted a discontinuous "initial segment"-like region of the epididymis. This is the most detailed study so far of the epithelia and the tubule organization in the caput epididymidis of any species, and most of the results are reported for the first time for the human. Although the pattern of the tubule network resembles that of some domestic species, the rich variety of epithelia has not been appreciated before.     

Effects of caput ligation on rat sperm and epididymis: protein thiols and fertilizing ability.

Mammalian spermatozoa mature while passing through the epididymis. Maturation is accompanied by thiol oxidation to disulfides. In rats, sperm become motile and fertile in the cauda. We have previously demonstrated that rat caput sperm contain mostly thiols and that upon passage from the corpus to the cauda epididymidis, sperm protein thiols are oxidized. The present work was undertaken to study the role of the regions of the epididymis in sperm maturation as reflected in the thiol status, fertility, and motility of the spermatozoa. The distal caput epididymidis of mature albino rats was ligated on one side. After 5 days, sperm were isolated from the ligated caput and from caput and cauda of the control side. Thiol groups in sperm, epididymal luminal fluid (EF), and epididymal tissue were labeled using the fluorescent thiol-labeling agent monobromobimane. After ligation, changes were observed in a) sperm proteins, sperm nuclear proteins, and epididymal fluid by electrophoresis; b) epididymal tissues by histochemistry; c) progressive motility by phase microscopy; and d) fertilizing ability after insemination into uteri of immature females. We found that after ligation, caput sperm thiols, especially protamine thiols, are oxidized, rendering them similar to mature sperm isolated from the cauda epididymidis. Spermatozoa from ligated caput epididymidis gain progressive motility and partial fertilizing ability. Morphology of epithelial cells of ligated caput is similar to that of cauda cells. However, other changes in caput EF and epithelium induced by ligation render the ligated caput epididymidis different from either control caput or cauda. Hence, sperm thiol oxidation, along with the development of fertilizing ability, can occur in sperm without necessity for sperm transit through the corpus and cauda epididymidis.     

Ghrelin attenuates heat-induced degenerative effects in the rat testis

This study was conducted to examine the efficacy of ghrelin in prevention of deleterious effects of heat stress in rat testicular tissue. Forty five adult male rats were scheduled for this study and were divided equally into three groups: heat-saline, heat-ghrelin and control-saline. The scrota of heated-designed rats were immersed once in water bath at 43 °C for 15 min. Immediately upon heating, 2 nmol of ghrelin were given subcutaneously to heat-ghrelin animals every other day up to day 60 and physiological saline to the other two groups using the same method. The animals were sacrificed at 10, 30 and 60 days after heat treatment and their testes were taken for later photomicrograph and immunohistochemical analysis. Testicular histopathology revealed a significant reduction in the means of seminiferous tubules and Sertoli cell nucleus diameters as well as germinal epithelium height on day 10 in both heated groups. Furthermore, other testicular components including miotic index, spermatogenesis rate, presence of spermatocytes and volume densities were dramatically decreased following heat exposure. Notably, ghrelin caused a partial recovery in all of the above-mentioned parameters and accelerated testicular regeneration process by day 30 compared to the heat-saline group (P<0.05). Because of testicular progressive recovery, these indices were similar among groups on day 60 (P>0.05). However, immunohistochemistry evaluation for in situ detection of Bcl-2 protein did not exhibit any germ cells-positive of this factor among groups at different experimental days. In conclusion, the results of the present study indicate for the first time the novel evidences of ghrelin ability in attenuation of heat-induced testicular damage and also that ghrelin therapy may be useful as a suppressor of degenerative effects following testicular hyperthermia.     

Testicular torsion and its effects on the spermatogenic cycle in the contralateral testis of the rat

The aim of the present study was to evaluate the effects of unilateral testicular torsion on the contralateral testis with respect to the stages of the cycle of the seminiferous epithelium (CSE). Fifty-five male Wistar rats, 60 days old, were used. The animals were divided into 11 groups. Groups 1-5 were subjected to unilateral testicular torsion from 3 to 48 h, followed by detorsion. Groups 6-10 had unilateral orchiectomies after unilateral testicular torsion for 3 to 48 h. Animals constituting group 11 served as the control sham-operated group. All animals were killed after 2 months. The percentage of affected tubules (tubules showing pathological changes) in the contralateral testis was estimated based on the CSE stages. In the torsion/detorsion group, the percentage of affected tubules was significantly greater (58.6%) than in torsion/orchiectomy group (48.0%). Stages VI-XI of the spermatogenic cycle were the most affected when compared with the rest of the stages in each experimental group (P <0.05). These results show that stages VI-XI of the spermatogenic cycle, the stages associated with low antioxidant capacities, are the most sensitive to the effects of testicular torsion on the contralateral testis.     

The histochemical localization of prostaglandin synthetase activity in reproductive tract of the male rat.

PG synthetase activity was assessed histochemically in the reproductive tract of male rats. Moderate activity was observed in tails of spermatozoa within the corpus and cauda epididymidis but there was no activity in the caput epididymidis or the seminiferous tubules. The sperm tail activity was maximal for cells within the vas deferens. PG synthetase activity was also observed in individual adipose cells adhering to the testicular capsule, epididymis and vas deferens, and in isolated interstitial cells of the testis and the caput, corpus and cauda epididymidis. Specific cells in the capsules of the testes, epididymis and vas deferens also produced PGs. The activity observed in the interstitial cells of the testis and the caput epididymidis was less than that for the other tissues in terms of the proportion of possible cells. The demonstration of PG synthetase activity paralleled to known loss of arachidonic acid from the phospholipids of the spermatozoa as they pass through the male tract. Endogenous substrate was not limiting in the assay system, even in the testis and caput epididymidis where PG synthesis was not normally observed, indicating that a PG synthesis inhibitor may be present in these two tissues. PG synthetase activity within teased seminiferous tubules was markedly increased by physical trauma. Indomethacin diminished but did not eliminate synthesis.     

Spermatogenesis in white-lipped peccaries (Tayassu pecari)

We describe here morphological and functional analyses of the spermatogenic process in sexually mature white-lipped peccaries. Ten sexually mature male animals, weighing approximately 39 kg were studied. Characteristics investigated included the gonadosomatic index (GSI), relative frequency of stages of the cycle of seminiferous epithelium (CSE), cell populations present in the seminiferous epithelium in stage 1 of CSE, intrinsic rate of spermatogenesis, Sertoli cell index, height of seminiferous epithelium and diameter of seminiferous tubules, volumetric proportion of components of the testicular parenchyma and length of seminiferous tubules per testis and per gram of testis. The GSI was 0.19%, relative frequencies of pre-meiotic, meiotic and post-meiotic phases were, respectively 43.6%, 13.8% and 42.6%, general rate of spermatogenesis was 25.8, each Sertoli cell supported an average 18.4 germinative cells, height of seminiferous epithelium and diameter of seminiferous tubules were, respectively, 78.4 microm and 225.6 microm, testicular parenchyma was composed by 75.8% seminiferous tubules and 24.2% intertubular tissue, and length of seminiferous tubules per gram of testis was 15.8m. These results show that, except for overall rate of spermatogenesis, the spermatogenic process in white-lipped peccaries is very similar to that of collared peccaries, and that Sertoli cells have a greater capacity to support germinative cells than most domestic mammals.     

Dye permeability and alkaline phosphatase activity of testicular capillaries in the postnatal rat

Summary Staining of testicular and epididymal tissues after intravenous, intraperitoneal or subcutaneous administration of a number of dyes was investigated in rats at different stages of postnatal development. After light green injections heavy staining of both testis and epididymis was visible to the naked eye in neonatal animals up to the age of 10 days, while in rats over 15 days old no appreciable staining of the testis could be seen, although the caput epididymis was strongly coloured. From 3–8 hours after subcutaneous acriflavine administration, the nuclei in the blood vessel walls of the testis, as well as the nuclei in the rete testis, tubuli efferentes and caput epididymis, fluoresced in all age groups. The nuclei of the interstitial and tubular cells were stained intensely until the age of 5 days. Thereafter the intensity gradually diminished until the age of 20 days, when no nuclear fluorescence was visible in the seminiferous tubules and even the interstitial nuclei fluoresced weakly or not at all.The histochemical alkaline phosphatase activity of the testicular capillaries was studied by Gomori's method, using fresh and postfixed cryostat sections from postnatal rat testes. The testicular capillaries exhibited appreciable activity at the age of 10 days.On the basis of the present and previous observations on the permeability of the testicular capillaries, the existence of a blood-testis barrier in the puberal and adult rat testis is suggested.Development of the blood-testis barrier and the alkaline phosphatase activity of the testicular capillaries are suggested to reflect general vascular maturation at the beginning of puberty in the rat.Supported by grants from Yrjö Jahnsson's Foundation and P. O. Klingendahl Foundation.     

Effects of intratesticular zinc gluconate treatment on testicular dimensions, echodensity, histology, sperm production, and testosterone secretion in American black bears (Ursus americanus)

Eight adult American black bears were used to evaluate the effects of chemical castration by intratesticular zinc gluconate treatment on testicular dimensions, echodensity, histology, sperm production, and testosterone secretion. Treatment did not affect testicular dimensions and did not result in decreased resting or GnRH-stimulated testosterone secretion. Multifocal hyperchoic areas in the testicular parenchyma were observed on ultrasound examination, and white foci were observed on gross pathology examination after zinc gluconate treatment. Histologically, there were normal seminiferous tubules containing either round or elongated spermatids, along with abnormal tubules in all bears after treatment. Vacuolation of the seminiferous epithelium, sloughing of germ cells into the tubules' lumen, presence of multinuclear giant cells, and reduced height of the seminiferous epithelium with missing generations of germ cells were commonly observed. The most severe testicular changes were multifocal and included fibrosis, complete degeneration of the seminiferous epithelium with shrinkage of the tubule, and sperm stasis. Epididymal sperm reserve was 982.74 ± 654.16 × 10⁶ sperm (mean ± SEM) and motile sperm were observed in the epididymis of all but one of the bears. In conclusion, although intratesticular zinc gluconate treatment in black bears resulted in testicular degenerative changes detected by ultrasound and histology examinations, sperm production was not completely ablated. We inferred that normal fertility might have been compromised, but treatment unlikely resulted in sterility.     

In vivo secretion and association of clusterin (SGP-2) in luminal fluid with spermatozoa in the rat testis and epididymis.

Clusterin (sulfated glycoprotein-2) is a heterodimeric glycoprotein synthesized and secreted by rat Sertoli cells. An antigenically similar form is synthesized and secreted by the epididymis. The goal of this study was to define the epididymal regions in which clusterin is present and the regions in which clusterin is secreted and interacts with developing spermatozoa. Seminiferous tubule (STF), caput, corpus, and cauda fluids were collected by micropuncture and/or microperfusion and two-dimensional Western blot analysis was performed with a polyclonal antibody directed against Sertoli cell clusterin. Clusterin was found in both STF and epididymal fluid. STF contained predominantly the clusterin heavy chain (45 kd); however, a 70 Kd heterodimer was present under nonreducing conditions. Two subunits of clusterin with lower molecular weights (41 kd, heavy chain; 32 kd, light chain) and higher isoelectric points were present in the luminal fluid of all epididymal regions. The intraluminal levels of the heavy and light chains decreased from caput to cauda. Analysis by two-dimensional gel electrophoresis of proteins secreted directly into the epididymal luminal fluid revealed that clusterin was secreted by caput epithelium and not by the corpus and cauda epithelium. Western blots of membrane extracts from testicular, caput, and cauda spermatozoa revealed that testicular clusterin was associated with testicular sperm and epididymal clusterin with predominantly caput sperm. Our findings suggest that clusterin is secreted into the caput epididymal lumen, where it binds to sperm and then dissociates from sperm to be endocytosed by cells of the distal epididymal epithelium.     

Ultrastructure of the reproductive system of the black swamp snake (Seminatrix pygaea). VI. Anterior testicular ducts and their nomenclature

In this study, the anterior testicular ducts of the North American natricine snake Seminatrix pygaea are described using light and electron microscopy. From the seminiferous tubules, the rete testis passes into the epididymal sheath, a structure along the medial border of the testis heavily invested with collagen fibers. The rete testis consists of simple, nonciliated cuboidal epithelium (principal cells). The intratesticular ducts of the rete testis are narrow (50–70 μm) at their junction with the seminiferous tubules, widen (80–100 μm) as they extend extratesticularly, and divide into smaller branches as they anastomose with the next tubules, the ductuli efferentes. The ductuli efferentes are lined by simple cuboidal epithelium but possess nonciliated principal cells as well as ciliated cells. These are the only ducts in the male reproductive system with ciliated cells. The ductuli efferentes are narrow (25–45 μm), divide into numerous branches, and are highly convoluted. The ductus epididymis is the largest duct in diameter (240–330 μm), and the diameter widens and the epithelium thins posteriorly. The ductus epididymis is lined by nonciliated, columnar principal cells and basal cells. No regional differences in the ductus epididymis are apparent. Ultrastructural evidence suggests that all of the nonciliated principal cells in each of the anterior testicular ducts function in both absorption and secretion. Absorption occurs via small endocytic vesicles, some of which appear coated. Secretion is by a constitutive pathway in which small vesicles and a flocculent material are released via a merocrine process or through the formation of apocrine blebs. The secretory product is a glycoprotein. Overall, the characteristics of the anterior testicular ducts of this snake are concordant with those of other amniotes, and the traditional names used for snakes are changed to conform with those used for other sauropsids and mammals. J. Morphol., 2010. © 2009 Wiley‐Liss, Inc.     

Testicular lesions of streptozotocin diabetic rats.

Diabetes was induced in adult male albino rats by a single intravenous injection of streptozotocin (75 mg/kg body weight). The diabetes was allowed to stabilize for at least 15 days, whereafter the testicular and seminal vesicle histology was studied at various time intervals. Reduction in testis weights and tubule diameters was significant after 2 weeks of diabetes. The changes in seminiferous tubules ranged from premature sloughing of epithelium to total cessation of spermatogenesis. The testicular histology of diabetic animals frequently greatly simulated the situation described following hypophysectomy. By subjective visual assessment the number of Leydig cells was found to be normal or reduced in all of the diabetic animals. Diabetes was also demonstrated to induce seminal vesicle atrophy, which did not show any correlation with the degree of testicular lesions. The possible etiology of testicular damage in diabetic animals is discussed.     

Location of enzymes of androgen and estrogen biosynthesis in the testis of the ground squirrel (Citellus lateralis)

The intratesticular localization of enzymes of androgen and estrogen biosynthesis was studied in the ground squirrel (Citellus lateralis). In mature animals, interstitium and tubules were isolated by manual dissection. Microsomes were prepared and enzymes assayed by analysis of product formation after incubation with appropriate 3H-labeled substrates. In the immature testis, tubules and interstitium are not readily separable; thus, distribution was inferred after analysis of whole testicular microsomes from control, follicle-stimulating hormone (FSH)-treated, and luteinizing hormone (LH)-treated animals. To verify the cellular composition of tissues and the status of steroidogenic organelles in Leydig and Sertoli cells, samples were also analyzed by light and electron microscopy. In mature squirrels, enzymes of androgen biosynthesis were concentrated in the interstitium; however, levels present in the tubules were sufficient to account for a substantial fraction of whole testicular activity (1/3 to 1/5). By contrast, virtually all of the testicular aromatase was accounted for by that in the seminiferous tubules. The purity of these fractions was checked by light microscopy; they showed little cross-contamination. In whole testicular microsomes of immature squirrels, androgen biosynthetic enzymes had a much lower specific activity than in mature animals; however, the opposite was true for aromatase, its activity being approximately 5-fold higher in prepubertal animals. Luteinizing hormone treatment markedly stimulated hydroxylase and lyase but not aromatase. Luteinizing hormone also induced an increase in Leydig cell size and a dramatic proliferation of smooth endoplasmic reticulum. These changes were correlated with increased serum testosterone. As shown previously in rats, 3 beta-hydroxysteroid dehydrogenase was independent of LH control. Follicle-stimulating hormone had no effect on any of the enzymes studied, but induced some increase of agranular reticulum in Sertoli cells. Results from immature squirrels thus corroborate data from mature animals, showing a predominant interstitial location of androgen biosynthetic enzymes. While we cannot explain the absence of FSH stimulation of aromatase activity, the data do not refute the findings in mature animals showing a predominant tubular location of this enzyme. We conclude that the distribution of steroidogenic enzymes in the testis of squirrels differs in several important respects from rats, although both are members of the order Rodentia.     

Testosterone immunoreactivity in the seminiferous epithelium of rat testis: effect of treatment with ethane dimethanesulfonate

Androgens drive spermatogenesis by processes that are largely unknown. Direct effects on germ cells and indirect effects mediated via testicular somatic elements are currently under consideration, and specific localization of androgens in seminiferous tubules may provide information as regards this. Adult male rats were injected with ethane dimethanesulfonate (EDS; 75 mg/kg body weight) or vehicle. Testes were fixed and paraffin-embedded for localization of testosterone immunoreactivity 1 and 2 weeks after treatment, using the unlabeled antibody (PAP) technique. Plasma testosterone dropped from a pre-treatment level of 2.3 ng/ml to below 0.2 ng/ml 3 days after EDS injection and remained at low levels until the end of observation, accompanied by a progressive decrease in testicular weight. In the seminiferous tubules of vehicle-injected males, testosterone immunoreactivity was found in nuclei of spermatocytes and spermatids and in nuclei and the cytoplasm of Sertoli cells, and showed typical variations according to the stage of spermatogenesis. One week after EDS treatment, immunoreactivity had disappeared from the seminiferous epithelium. Two weeks after treatment, staining of germ cells was detected in two out of four males. The disappearance and reappearance of immunoreactivity coincided with the time course of EDS effects on rat Leydig cells, and we conclude that it corresponds to androgen specifically localized in fixed, paraffin-embedded tissue. Because staining of germ cell nuclei varied with the stage of spermatogenesis, the technique may detect a physiologically relevant androgen fraction; its location suggests that androgens may also directly affect certain germ cell stages.