Cell Tissue Res (f992) 270 : 513-519
Cell&Tissue Research 9 Springer-Verlag 1992
Appearance and endocytic activity of epithelial cells from human efferent ducts in primary monolayer culture S. Raczek t, C.H. Yeung 1, L. Hertle 2, H. Schulze 3 and T.G. Cooper 1 i Institut fiir Reproduktionsmedizin, Westf/ilische Wilhelms-Universit/it Mfinster, Steinfurter Strasse 107, W-4400 Miinster, Federal Republic of Germany 2 Institut fiir Urologic Westf/ilische Wilhelms-Universit/it, Mfinster, Federal Republic of Germany 3 Institut ffir Urologic, Ruhr Universit/it Bochum, Federal Republic of Germany Received May 27, 1992 / Accepted July 13, 1992
Summary. The culture of epithelial cells lining h u m a n efferent ducts, obtained from prostatic carcinoma patients, is described. Ciliated cells were observed to beat for at least one m o n t h on plastic. On pervious filters low cuboidal cells characterized the monolayers. Cells comprising monolayers over the filter were 5 to 9 gm in height whereas taller cells were found over the original fragments (14 gin). Some non-ciliated cells contained dark and light vacuoles, others were found to lack them. Both non-ciliated and ciliated cells maintained tight junctional complexes restricting the paracellular movement o f horseradish peroxidase. Both types of cultured cells exhibited fluid-phase and adsorptive endocytosis from both apical and basal surfaces. It is reported for the first time that the monolayers form high resistance barriers (150 O . c m 2) that prevent the apical medium from draining to the basal c o m p a r t m e n t over 24 h.
Key words: Reproductive system, male - Ductuli efferentes - Epithelial cells - Cell culture - Monolayer cultures - M a n
In animals it is clear that the single convoluted tubule of the epididymis is intimately involved in the processing of spermatozoa, so that they become fully functional, potentially fertilising cells. In man, however, clinical observations have fuelled a debate over even the necessity of this organ (see Cooper, 1990). Since no biopsies of it can be taken, the study of h u m a n epididymal physiology is limited. Another complicating factor is the complexity of the a n a t o m y of the h u m a n post-testicular duct system, especially the network of ductuli efferentes comprising the caput epididymidis (Yeung et al. 1991). Although these tubules, and the epididymal epithelium proper, are derived f r o m mesonephric anlage, the former differ f r o m the latter with respect to the presence o f ciliated cells and absence of basal cells. In addition
Correspondence to: T.G. Cooper
to various types of ciliated cells, presumably responsible for transport of testicular fluid and spermatozoa through the tubuli, non-ciliated cells also display variety in the size, number~ location and electron opacity of intracellular vacuoles (Yeung et al. 1991). Whether these structures are absorptive or secretory in nature is not yet known. The demonstration that cells from the hum a n corpus epididymidis can be cultured as monolayers (Cooper et al. 1990) p r o m p t e d this investigation of the culture of cells f r o m the efferent ducts. A preliminary report of this work has been made (Raczek et al. 1992).
Materials and methods Tissues Human testes and epididymides were obtained, with informed consent, over a 15-month period from 20 men (mean age 71 years, range 39-87) undergoing castration for prostatic carcinoma (Table 1). Of the 20 tissues obtained eleven were from men who had not received anti-androgen treatment and eight men received antiandrogen before or during the operation (Table 1). Tissues were transported in Ham's F-12 medium on ice to the laboratory within 90 min of the operation and kept under sterile conditions in a vertical laminar flow hood (TC72, Flow Labs, D-5309 Meckenhelm, FRG). Fourteen of the tissues contained motile spermatozoa (Table 1).
Media and culture procedure Media for enzymic digestion, containing antibiotics and fungizone (ABF), culture medium (D-valine substituted serum-free defined medium, SFDM) and reconstituted basement membrane (Matrigel: Becton-Dickinson, Heidelberg, FRG)-coated culture inserts (Millicell-HA, Millipore GmbH, Eschborn, FRG) were prepared as described in Cooper et al. (1989, 1990). The caput epididymidis comprised of efferent ducts (delineated distally by the furthest extent of the dark tubules: Yeung et al. 1991) was severed from the corpus region, which was processed separately. With fine scissors and forceps the thick white capsule was removed and the brown pigmented tubules under the capsule were exposed. These and the white efferent ducts were incubated
514 Table 1. Age and anti-androgen therapy of patients providing tissues for the study
Tissue
Age
Anti-androgens"
Motile sperm seen ?
No. of chambers
No. of days ciliary beating observed
H60 H63 H64 H65 H66 H71 H72 H73 H75 H76 H77 H78 H79 H88 H89 H90 H92 H94 H95 & 96 H97
39 72 85
None None None None None 1 x 150 mg for 7d 1 x 300 mg 1 x 300 mg 3 x 300 rag, > 1 year 3 x 300 mg None Unknown None None None None 1 x 300 mg 1 x 100 mg for 7d 150 mg for 1 & 7d None
No Yes Yes No Yes Yes No Yes Yes No Yes Yes No Yes Yes No No Yes Yes Yes
4 1 1 2 1 2 1 1 3 0 t 2 1 6 4 2 1 1 5 2
Not examined Not examined 29 26 23 22 9 13 21 22 Not examined Not examined Not examined Not examined Not examined 1 21 11 8 None seen
62 74 62 72 87 74 78 82 66 69 51 83 78 72 67 & 56 68
a Androcur, orally
together with elastase (Sigma type I, 13.8 U, Deisenhofen, FRG) and collagenase (Sigma Type IA, Lot. No. 030H-6849, 6.9 U) in 12.5 ml ABF at 37~ at 160 strokes per min for 20 rain. The tissues were then dissected free of connective tissue and digestion was continued until tubules relatively free of connective tissue resulted. The main difference in the culture of human efferent duct cells from that of the corpus epididymidis (Cooper et al. 1990) was the difficulty encountered in removing the knots of extremely tough connective tissue surrounding the small tubules. This necessitated cutting and tearing the tissues and vigorous shaking with the enzyme mixtures. Typically, three changes of enzyme, each for 1 h, were necessary to liberate tubules, which were never so free of connective tissue as tubules obtained from the corpus epididymidis. Further processing of the tubules was by cutting into fine pieces, digestion in collagenase (6.9 U) and hyaluronidase (Sigma Type lS, 1.8 U) in 12.5 ml ABF and vigorous physical disruption by continued aspiration with a 10 ml pipette to liberate epithelial fragments. Periodically one drop of the fragment suspension was examined microscopically to ascertain the size of the fragments. Tissue fragments were harvested from the supernatant after settling out of the larger fragments (Cooper et al. 1990) and washed free of enzyme. Digestion of the larger fragments in the pellet continued until small fragments were obtained. Digestion in this enzyme mixture occurred much faster than in the corpus epididymidis and sometimes hyaluronidase alone sufficed. Fragments of 100-400 gm diameter were centrifuged at 70 g for 5 min, the supernatant discarded and the pellet washed successively in 10 ml ABF, D-valine medium (followed by centrifugation at 130 g) and S F D M (70 g, 5 min). After removal of the final supernatant, the size of the pellet was measured in a 1 ml pipette. A pellet height of 0.1 ml mostly contained sufficient fragments to cover eight, 12 ram-diameter Millicell-HA chambers, and the appropriate volume of S F D M (0.4 ml per chamber) was added to resuspend thepellet. After thorough mixing the fragment suspension was transferred to Matrigel-coated culture chambers in 6-well plates (Falcon Products, Becton Dickinson, New Jersey, USA) and 3 ml S F D M was added to each well. Cultures were kept at 34~ C in 5% CO2. To observe fragments spreading out and the beating of ciliated cells, a drop of the fragment supension was also placed into a Matrigel-coated 6-well plate in S F D M on the day of culture.
Confluent monolayers The basal media were changed for the first time after 48 h and then every 48 h. From day 10 the culture chambers were filled daily with 0.8 ml S F D M (giving a convex meniscus) and the height of the fluid in the chambers was noted after 24 h to determine the status of the monolayer (Onoda et al. 1990). When support of a fluid column indicated complete monolayers, the culture chambers were transferred to 24-well plates ( N U N C GmbH, Wiesbaden, FRG). The medium (0.3 ml in the wells, 0.8 ml in the culture chambers) was changed daily and collected 24 h later into preweighed Eppendorf tubes (for estimation of the fluid volumes) before dentrifugation of the apical medium (2000 g, 5 rain) and freezing both apical and basal media at - 20 ~ C. Some loss of the medium, particularly that adhering beneath the chamber, and apical medium draining through " l e a k y " monolayers, was entailed when the chambers were transferred to Petri dishes for fluid collection. The electrical resistance of the epithelial sheets was measured (Millicell-ERS apparatus, Millipore GmbH, Eschborn, FRG). Resistance measurements from RBM-coated, cell-free filters were subtracted as blanks and values were corrected for the area of the filter. Monolayers in 6-well plates were observed for ciliary beating every 2 days with phase-contrast optics (Zeiss IM inverted microscope, K61n, FRG).
Morphological studies Preparation of tissues for electron microscopy and incubation with markers of fluid phase and adsorptive endocytosis was as previously described (Yeung et al. 1989). In some experiments, horseradish peroxidase was added (0.3 mg/ml in SFDM) to the basal compartment for 20 h in one culture and at 1.0 mg/ml in the apical compartment for 50 min in another; these sections were not stained with uranyl acetate and lead citrate. In another culture, cationic ferritin (1.0 mg/ml in SFDM) was added to the apical compartment for 20 min. Cell heights were measured from 0.5-1 lam-thick Epon sections of the culture filters as previously described (Yeung et al. 1989).
515
Results
Table 3. Classes of monolayers distinguished on the basis of volumes of medium recovered from the apical compartment after 24 h and electrical resistance to current flow
Cell cultures In total 19 cultures were made (Table 1) and success of monolayer formation did not appear to depend on the previous administration of anti-androgens. When examined in one culture, the heights of the cells within epithelial fragments at the end of incubation with enzymes was higher than the cells forming the monolayers after culture for 25 days (Table 2). In this culture taller epithelial cells were found over the tissue clumps, although they were still shorter than those harvested on day zero (Table 2). The ciliated cells were observed to beat on Matrigel-coated plastic dishes for up to 29 days, although the duration of beating was variable and was unrelated to the anti-androgen treatment (Table 1).
Confluent monolayers At least three groups of cultures were distinguishable on the basis of the volume of fluid supported by the monolayers overnight (Table 3). In non-confluent m o n o layers (Grade 0 in our terminology), most of the 0.8 ml medium drained to the basal c o m p a r t m e n t ; in " l e a k y " monolayers (Grades I [meniscus higher than in surrounding compartment] and II [meniscus level with rim of chamber]) some of the medium had passed through the monolayer; and most of the medium remained in the upper chamber in G r a d e - I I I confluent monolayers (Table 3). O f 16 cultures, 5 (31%) produced at least one chamber with confluent monolayers as judged from the overnight support of apical medium, and overall 10 of 39 chambers (26%) were able to support fluid columns. Confluence was first observed between days 16 to 25 (mean 21 days) and was maintained during daily medium changing for 3-15 days (mean 11 days). Cultures were maintained for a mean of 38 days (range 24 to 58) before taken for further examination. The electrical
Table 2. Heights of epithelial cells from human efferent ducts within cultured monolayers, n; Number of measurements Culture
Treatment
Days
Cell height (gm) a
cultured H60 H60 H63 H65 H65 H65 H73 H75 H95/96i H95/96ii
In monolayer Full cell height b In monolayer Before seeding In monolayer Over fragments In monolayer In monolayer In monolayer In monolayer
49 49 28 0 25 25 29 24 35 35
mean
SD
8.7 13.7 6.4 21.0 5.0 14.4 5.6 6.8 8.1 6.8
5.3 5.6 2.3 4.8 1.8 4.4 2.3 2.3 2.4 2.6
a Height of cells in the monolayers at randomly selected sites b Only cells containing cross-sections of nuclei were measured
n 358 223 876 57 1167 44 194 647 334 740
Monolayer grade a
III II I 0
Volume in apical compartment (ml)b
Electrical resistance to current flow (•" c m 2) c
Mean
SEM
n
Mean
SEM
n
0.67 0.53 0.46 0.38
0.07 0.08 0.08 0.08
45 6 12 4
166.7 78.5 44.0 19.6
32.0 18.7 6.9 6.1
12 3 10 22
" Monolayers were categorised after 24 h as Grade III (convex meniscus maintained), Grade II (meniscus level with the rim of the chamber), Grade I (meniscus higher than that in the basal compartment) and 0 (hydrodynamic equilibrium attained) b Original volume 0.79 _+0.02 ml medium, values from 14 chambers from 6 cultures c Values from 54 chambers from 12 cultures resistance of the epithelia exceeded 150 ~?.cm 2 for monolayers that supported fluid columns overnight (Table 3).
Electron microscopy A variety of different cell types were present in cultures of epithelial cells from h u m a n efferent ducts. Cells of the monolayer covering the filter were cuboidal or flatter. These comprised non-ciliated cells with or without dark vacuoles (Fig. 1), with clear vacuoles (Fig. 4), and cells containing no large vacuoles (Fig. 2). Ciliated cells were also observed (Fig.3) but much less frequently than the non-ciliated cells. In some monolayers cells displayed distended intercellular spaces (Figs. 2, 3), but in others lateral cell junctions were in close contact (Figs. 1, 4). Cells in the monolayer over the clumps were taller and richer in cilia and microvilli than cells overlying the filter (Table 2), and thus displayed a more typical columnar appearance, but the same variety of cells were seen: nonciliated cells with dark vacuoles and clear (lipid-rich) vacuoles (Fig. 6), and ciliated cells (not shown). The zonulae occludentes between the cells were morphologically normal (Fig. 8) and effectively prevented m o v e m e n t of horseradish peroxidase from basal to apical c o m p a r t m e n t s (Fig. 5). Cell organelles including mitochondria, Golgi apparatus, endoplasmic reticulum, lysosomes and nuclei appeared normal. Microvilli were stout and sparse. The fluid-phase marker of endocytosis, horseradish peroxidase, was taken up into vesicles, vacuoles, multivesicular bodies and secondary lysosomes Of non-ciliated cells (Fig. 7) and ciliated cells (Fig. 7) f r o m the basal c o m p a r t m e n t during overnight incubation. The same organelles contained H R P after short-term incubation with the tracer in the apical medium (not shown). Although it was rarely found, cationic ferritin was seen lining the apical m e m b r a n e of the cells and within cell vacuoles after incubation in the apical c o m p a r t m e n t (not shown).
516
Fig. 1. Non-ciliated epithelial cells containing dark vacuoles and lacking clear vacuoles from the ductuli efferentes of an 85-year-old man. Monolayer cultured for 41 days. Note sparse stereocilia, wellmaintained Golgi complexes interspersed with vacuoles. F Filter; A apical surface, x 11 600; bar = 1 g m
Fig. 2. Junctions between non-ciliated cells displaying visible intercellular spaces, containing lateral interdigitations, extending from the filter (right side) to the apical surface. A r r o w h e a d Tight junction. Tissue from a 39-year-old male; monolayer in culture for 49 days. x 7 500; bar = 2 gm
517
Fig. 3. Ciliated cell from same tissue as Fig. 2. Note cilia (arrow), prominent nucleolus, mitochondria thoughout the cytoplasm and supranuclear Golgi apparatus (arrowheads), lipofuscin-filled lysosome (L) and distended intercellular spaces, x 8900; bar = 2 lam
Fig. 4. Non-ciliated cells from same tissue as Fig. 2. Note clear vacuoles and short microvilli at apical surface and non-distended intercellular interspaces (arrowheads). x 6200; bar=2 gm
Fig. 5. Lateral intercellular spaces containing per0xidase up to the tight junction (arrowhead) after basal application of the enzyme. Tissue from 87-year-old patient; 24-day culture, x3200; bar= 0.2 gm
518
Fig. 6. Non-ciliated cells overlying epithelial fragments. Note concentration of dark vacuoles near the apex and lighter vacuoles towards the base. Tissue from 72-year-old male; 28-day culture. x 8400; bar = 2 ~tm Fig. 7. Unstained section of ciliated ceti (upper) and non-ciliated cell revealing presence of peroxidase product (dark deposits) in
vesicles (arrowheads) after apical application. C Ciliary rootlet. 24day culture. • 29200; b a r = 0 . 5 I.tm Fig. 8. Well-preserved junctional complex between cells cultured for 41 days. Same tissue as Fig. 1. x 32400; b a r = 0 . 2 gm
Discussion
the knots o f tubules that are present in this region of the epididymal head (Yeung et al. 1991), and a compromise between the a m o u n t of epithelial and connective tissues always had to be made. For example, the fine tubules (lined by epithelium II) branching from the
The culture of h u m a n efferent ducts was more difficult than that of the h u m a n corpus epididymidis, owing to the tough nature of the connective tissue surrounding
519 straight white tubules leaving the testis are likely to be lost during such harsh enzymic and physical treatment. Nonetheless, several epithelial cell types were observed in the cultures: non-ciliated cells with dark vacuoles and clear vacuoles and ciliated cells. Because of the short stature of the cells in the monolayer over the filter, also observed in cultures of human corpus and rat caput and cauda (Cooper et al. 1990; Yeung et al. 1989), the polarised disposition of the vacuoles (dark vacuoles situated apically in epithelium I and basally in epithelium III as described by Yeung et al. 1991) was not apparent. This polarity was maintained, however, in the taller cells over fragments. Low cuboidal epithelia (II, IV) were not distinguishable in cultures and it is not known from which epithelia the ciliated cells arose. That the ciliated cells beat in culture for up to one month attests to the viability of these cells; in cultures of rat efferent ducts they were observed to beat for 2 weeks (Byers et al. 1985). The ultrastructure of cultured cells revealed only sparse cilia on ciliated cells in the monolayer. It is not known if the non-ciliated cells plated out better than the ciliated cells or the latter lost their cilia with age, as was indicated by Byers et al. (1985) in the rat. Uptake of horseradish peroxidase from the apical and basal compartment into ciliated and non-ciliated cells was observed, and HRP-laden vesicles were seen throughout the cytoplasm, as is the case for human corpus epididymal cells (Cooper et al. 1990). However, endocytosis of cationic ferritin was rarely observed in efferent duct cells in culture, in sharp contrast to cells from the corpus region (Cooper et al. 1990). In contrast, rapid endocytosis of androgen-binding protein by rat efferent duct cells has been demonstrated in monolayer culture (Byers et al. 1985), a difference attributable to either species differences, the endocytosed proteins or the age of the cultures. For the first time for human epididymal cultures, p r o o f of confluent monolayers was provided by the observation of the apical medium being supported overnight by the cells comprising the monolayer. This was recently
demonstrated for rat Sertoli cells cultured on similar previous supports (Onoda et al. 1991). Cell-free chambers did not support fluid and the relationship between electrical resistance of the monolayer and the weight of supported fluid suggests that the epithelial cells limit the movement of both ions and fluid. Such polarised epithelia will be useful in future studies of vectorial epithelial secretion and transport.
Acknowledgements. We acknowledge the help of Kfilli Nurmik and Eva M611mann for obtaining the tissues and maintaining cell cultures, and Dr. F. Oberpenning for obtaining patient information, and thank Professor W. Wittkowski for the use of the electron microscope. The study was funded by the Deutsche Forschungsgemeinschaft Grant No. Ni 130/11-3/C. References
Byers SW, Musto NA, Dym M (1985) Culture of ciliated and nonciliated cells from rat ductuli efferentes. J Androl 6:271-278 Cooper TG (1990) In defense of a function for the human epididymis. Fertil Steril 90:965-975 Cooper TG, Yeung CH, Meyer R (1989) Immature rat epididymal epithelial cells grown in static primary monolayer culture on permeable supports. I. Vectorial secretion. Cell Tissue Res 256:567-572 Cooper TG, Yeung CH, Meyer R, Schulze H (1990) Maintenance of human epididymal epithelial cell function in monolayer culture. J Reprod Fert 90:81 91 Onoda M, Suarez-Quian CA, Djakiew D, Dym M (1990) Characterization of Sertoli cells cultured in the bicameral chamber system: relationship between formation of permeability barriers and polarized secretion of transferrin. Biol Reprod 43:672-683 Raczek S, Yeung CH, Hertle L, Schulze H, Cooper TG (1992) Monolayer cell culture of human efferent duct epithelium. Miniposter No. 114, 7th European Workshop on Molecular and Cellular Endocrinology of the Testis, Schlol3 Elmau, FRG, May 1992 Yeung CH, Cooper TG, Meyer R (1989) Immature rat epididymal epithelial cells grown in static primary monolayer culture on permeable supports. II. Histochemistry and ultrastructure. Cell Tissue Res 256:573-580 Yeung CH, Cooper TG, Bergmann M, Schulze H (1991) Organization of tubules in the human caput epididymidis and the ultrastructure of their epithelia. Am J Anat 191:261-279
Cell&Tissue Research 9 Springer-Verlag 1992
Appearance and endocytic activity of epithelial cells from human efferent ducts in primary monolayer culture S. Raczek t, C.H. Yeung 1, L. Hertle 2, H. Schulze 3 and T.G. Cooper 1 i Institut fiir Reproduktionsmedizin, Westf/ilische Wilhelms-Universit/it Mfinster, Steinfurter Strasse 107, W-4400 Miinster, Federal Republic of Germany 2 Institut fiir Urologic Westf/ilische Wilhelms-Universit/it, Mfinster, Federal Republic of Germany 3 Institut ffir Urologic, Ruhr Universit/it Bochum, Federal Republic of Germany Received May 27, 1992 / Accepted July 13, 1992
Summary. The culture of epithelial cells lining h u m a n efferent ducts, obtained from prostatic carcinoma patients, is described. Ciliated cells were observed to beat for at least one m o n t h on plastic. On pervious filters low cuboidal cells characterized the monolayers. Cells comprising monolayers over the filter were 5 to 9 gm in height whereas taller cells were found over the original fragments (14 gin). Some non-ciliated cells contained dark and light vacuoles, others were found to lack them. Both non-ciliated and ciliated cells maintained tight junctional complexes restricting the paracellular movement o f horseradish peroxidase. Both types of cultured cells exhibited fluid-phase and adsorptive endocytosis from both apical and basal surfaces. It is reported for the first time that the monolayers form high resistance barriers (150 O . c m 2) that prevent the apical medium from draining to the basal c o m p a r t m e n t over 24 h.
Key words: Reproductive system, male - Ductuli efferentes - Epithelial cells - Cell culture - Monolayer cultures - M a n
In animals it is clear that the single convoluted tubule of the epididymis is intimately involved in the processing of spermatozoa, so that they become fully functional, potentially fertilising cells. In man, however, clinical observations have fuelled a debate over even the necessity of this organ (see Cooper, 1990). Since no biopsies of it can be taken, the study of h u m a n epididymal physiology is limited. Another complicating factor is the complexity of the a n a t o m y of the h u m a n post-testicular duct system, especially the network of ductuli efferentes comprising the caput epididymidis (Yeung et al. 1991). Although these tubules, and the epididymal epithelium proper, are derived f r o m mesonephric anlage, the former differ f r o m the latter with respect to the presence o f ciliated cells and absence of basal cells. In addition
Correspondence to: T.G. Cooper
to various types of ciliated cells, presumably responsible for transport of testicular fluid and spermatozoa through the tubuli, non-ciliated cells also display variety in the size, number~ location and electron opacity of intracellular vacuoles (Yeung et al. 1991). Whether these structures are absorptive or secretory in nature is not yet known. The demonstration that cells from the hum a n corpus epididymidis can be cultured as monolayers (Cooper et al. 1990) p r o m p t e d this investigation of the culture of cells f r o m the efferent ducts. A preliminary report of this work has been made (Raczek et al. 1992).
Materials and methods Tissues Human testes and epididymides were obtained, with informed consent, over a 15-month period from 20 men (mean age 71 years, range 39-87) undergoing castration for prostatic carcinoma (Table 1). Of the 20 tissues obtained eleven were from men who had not received anti-androgen treatment and eight men received antiandrogen before or during the operation (Table 1). Tissues were transported in Ham's F-12 medium on ice to the laboratory within 90 min of the operation and kept under sterile conditions in a vertical laminar flow hood (TC72, Flow Labs, D-5309 Meckenhelm, FRG). Fourteen of the tissues contained motile spermatozoa (Table 1).
Media and culture procedure Media for enzymic digestion, containing antibiotics and fungizone (ABF), culture medium (D-valine substituted serum-free defined medium, SFDM) and reconstituted basement membrane (Matrigel: Becton-Dickinson, Heidelberg, FRG)-coated culture inserts (Millicell-HA, Millipore GmbH, Eschborn, FRG) were prepared as described in Cooper et al. (1989, 1990). The caput epididymidis comprised of efferent ducts (delineated distally by the furthest extent of the dark tubules: Yeung et al. 1991) was severed from the corpus region, which was processed separately. With fine scissors and forceps the thick white capsule was removed and the brown pigmented tubules under the capsule were exposed. These and the white efferent ducts were incubated
514 Table 1. Age and anti-androgen therapy of patients providing tissues for the study
Tissue
Age
Anti-androgens"
Motile sperm seen ?
No. of chambers
No. of days ciliary beating observed
H60 H63 H64 H65 H66 H71 H72 H73 H75 H76 H77 H78 H79 H88 H89 H90 H92 H94 H95 & 96 H97
39 72 85
None None None None None 1 x 150 mg for 7d 1 x 300 mg 1 x 300 mg 3 x 300 rag, > 1 year 3 x 300 mg None Unknown None None None None 1 x 300 mg 1 x 100 mg for 7d 150 mg for 1 & 7d None
No Yes Yes No Yes Yes No Yes Yes No Yes Yes No Yes Yes No No Yes Yes Yes
4 1 1 2 1 2 1 1 3 0 t 2 1 6 4 2 1 1 5 2
Not examined Not examined 29 26 23 22 9 13 21 22 Not examined Not examined Not examined Not examined Not examined 1 21 11 8 None seen
62 74 62 72 87 74 78 82 66 69 51 83 78 72 67 & 56 68
a Androcur, orally
together with elastase (Sigma type I, 13.8 U, Deisenhofen, FRG) and collagenase (Sigma Type IA, Lot. No. 030H-6849, 6.9 U) in 12.5 ml ABF at 37~ at 160 strokes per min for 20 rain. The tissues were then dissected free of connective tissue and digestion was continued until tubules relatively free of connective tissue resulted. The main difference in the culture of human efferent duct cells from that of the corpus epididymidis (Cooper et al. 1990) was the difficulty encountered in removing the knots of extremely tough connective tissue surrounding the small tubules. This necessitated cutting and tearing the tissues and vigorous shaking with the enzyme mixtures. Typically, three changes of enzyme, each for 1 h, were necessary to liberate tubules, which were never so free of connective tissue as tubules obtained from the corpus epididymidis. Further processing of the tubules was by cutting into fine pieces, digestion in collagenase (6.9 U) and hyaluronidase (Sigma Type lS, 1.8 U) in 12.5 ml ABF and vigorous physical disruption by continued aspiration with a 10 ml pipette to liberate epithelial fragments. Periodically one drop of the fragment suspension was examined microscopically to ascertain the size of the fragments. Tissue fragments were harvested from the supernatant after settling out of the larger fragments (Cooper et al. 1990) and washed free of enzyme. Digestion of the larger fragments in the pellet continued until small fragments were obtained. Digestion in this enzyme mixture occurred much faster than in the corpus epididymidis and sometimes hyaluronidase alone sufficed. Fragments of 100-400 gm diameter were centrifuged at 70 g for 5 min, the supernatant discarded and the pellet washed successively in 10 ml ABF, D-valine medium (followed by centrifugation at 130 g) and S F D M (70 g, 5 min). After removal of the final supernatant, the size of the pellet was measured in a 1 ml pipette. A pellet height of 0.1 ml mostly contained sufficient fragments to cover eight, 12 ram-diameter Millicell-HA chambers, and the appropriate volume of S F D M (0.4 ml per chamber) was added to resuspend thepellet. After thorough mixing the fragment suspension was transferred to Matrigel-coated culture chambers in 6-well plates (Falcon Products, Becton Dickinson, New Jersey, USA) and 3 ml S F D M was added to each well. Cultures were kept at 34~ C in 5% CO2. To observe fragments spreading out and the beating of ciliated cells, a drop of the fragment supension was also placed into a Matrigel-coated 6-well plate in S F D M on the day of culture.
Confluent monolayers The basal media were changed for the first time after 48 h and then every 48 h. From day 10 the culture chambers were filled daily with 0.8 ml S F D M (giving a convex meniscus) and the height of the fluid in the chambers was noted after 24 h to determine the status of the monolayer (Onoda et al. 1990). When support of a fluid column indicated complete monolayers, the culture chambers were transferred to 24-well plates ( N U N C GmbH, Wiesbaden, FRG). The medium (0.3 ml in the wells, 0.8 ml in the culture chambers) was changed daily and collected 24 h later into preweighed Eppendorf tubes (for estimation of the fluid volumes) before dentrifugation of the apical medium (2000 g, 5 rain) and freezing both apical and basal media at - 20 ~ C. Some loss of the medium, particularly that adhering beneath the chamber, and apical medium draining through " l e a k y " monolayers, was entailed when the chambers were transferred to Petri dishes for fluid collection. The electrical resistance of the epithelial sheets was measured (Millicell-ERS apparatus, Millipore GmbH, Eschborn, FRG). Resistance measurements from RBM-coated, cell-free filters were subtracted as blanks and values were corrected for the area of the filter. Monolayers in 6-well plates were observed for ciliary beating every 2 days with phase-contrast optics (Zeiss IM inverted microscope, K61n, FRG).
Morphological studies Preparation of tissues for electron microscopy and incubation with markers of fluid phase and adsorptive endocytosis was as previously described (Yeung et al. 1989). In some experiments, horseradish peroxidase was added (0.3 mg/ml in SFDM) to the basal compartment for 20 h in one culture and at 1.0 mg/ml in the apical compartment for 50 min in another; these sections were not stained with uranyl acetate and lead citrate. In another culture, cationic ferritin (1.0 mg/ml in SFDM) was added to the apical compartment for 20 min. Cell heights were measured from 0.5-1 lam-thick Epon sections of the culture filters as previously described (Yeung et al. 1989).
515
Results
Table 3. Classes of monolayers distinguished on the basis of volumes of medium recovered from the apical compartment after 24 h and electrical resistance to current flow
Cell cultures In total 19 cultures were made (Table 1) and success of monolayer formation did not appear to depend on the previous administration of anti-androgens. When examined in one culture, the heights of the cells within epithelial fragments at the end of incubation with enzymes was higher than the cells forming the monolayers after culture for 25 days (Table 2). In this culture taller epithelial cells were found over the tissue clumps, although they were still shorter than those harvested on day zero (Table 2). The ciliated cells were observed to beat on Matrigel-coated plastic dishes for up to 29 days, although the duration of beating was variable and was unrelated to the anti-androgen treatment (Table 1).
Confluent monolayers At least three groups of cultures were distinguishable on the basis of the volume of fluid supported by the monolayers overnight (Table 3). In non-confluent m o n o layers (Grade 0 in our terminology), most of the 0.8 ml medium drained to the basal c o m p a r t m e n t ; in " l e a k y " monolayers (Grades I [meniscus higher than in surrounding compartment] and II [meniscus level with rim of chamber]) some of the medium had passed through the monolayer; and most of the medium remained in the upper chamber in G r a d e - I I I confluent monolayers (Table 3). O f 16 cultures, 5 (31%) produced at least one chamber with confluent monolayers as judged from the overnight support of apical medium, and overall 10 of 39 chambers (26%) were able to support fluid columns. Confluence was first observed between days 16 to 25 (mean 21 days) and was maintained during daily medium changing for 3-15 days (mean 11 days). Cultures were maintained for a mean of 38 days (range 24 to 58) before taken for further examination. The electrical
Table 2. Heights of epithelial cells from human efferent ducts within cultured monolayers, n; Number of measurements Culture
Treatment
Days
Cell height (gm) a
cultured H60 H60 H63 H65 H65 H65 H73 H75 H95/96i H95/96ii
In monolayer Full cell height b In monolayer Before seeding In monolayer Over fragments In monolayer In monolayer In monolayer In monolayer
49 49 28 0 25 25 29 24 35 35
mean
SD
8.7 13.7 6.4 21.0 5.0 14.4 5.6 6.8 8.1 6.8
5.3 5.6 2.3 4.8 1.8 4.4 2.3 2.3 2.4 2.6
a Height of cells in the monolayers at randomly selected sites b Only cells containing cross-sections of nuclei were measured
n 358 223 876 57 1167 44 194 647 334 740
Monolayer grade a
III II I 0
Volume in apical compartment (ml)b
Electrical resistance to current flow (•" c m 2) c
Mean
SEM
n
Mean
SEM
n
0.67 0.53 0.46 0.38
0.07 0.08 0.08 0.08
45 6 12 4
166.7 78.5 44.0 19.6
32.0 18.7 6.9 6.1
12 3 10 22
" Monolayers were categorised after 24 h as Grade III (convex meniscus maintained), Grade II (meniscus level with the rim of the chamber), Grade I (meniscus higher than that in the basal compartment) and 0 (hydrodynamic equilibrium attained) b Original volume 0.79 _+0.02 ml medium, values from 14 chambers from 6 cultures c Values from 54 chambers from 12 cultures resistance of the epithelia exceeded 150 ~?.cm 2 for monolayers that supported fluid columns overnight (Table 3).
Electron microscopy A variety of different cell types were present in cultures of epithelial cells from h u m a n efferent ducts. Cells of the monolayer covering the filter were cuboidal or flatter. These comprised non-ciliated cells with or without dark vacuoles (Fig. 1), with clear vacuoles (Fig. 4), and cells containing no large vacuoles (Fig. 2). Ciliated cells were also observed (Fig.3) but much less frequently than the non-ciliated cells. In some monolayers cells displayed distended intercellular spaces (Figs. 2, 3), but in others lateral cell junctions were in close contact (Figs. 1, 4). Cells in the monolayer over the clumps were taller and richer in cilia and microvilli than cells overlying the filter (Table 2), and thus displayed a more typical columnar appearance, but the same variety of cells were seen: nonciliated cells with dark vacuoles and clear (lipid-rich) vacuoles (Fig. 6), and ciliated cells (not shown). The zonulae occludentes between the cells were morphologically normal (Fig. 8) and effectively prevented m o v e m e n t of horseradish peroxidase from basal to apical c o m p a r t m e n t s (Fig. 5). Cell organelles including mitochondria, Golgi apparatus, endoplasmic reticulum, lysosomes and nuclei appeared normal. Microvilli were stout and sparse. The fluid-phase marker of endocytosis, horseradish peroxidase, was taken up into vesicles, vacuoles, multivesicular bodies and secondary lysosomes Of non-ciliated cells (Fig. 7) and ciliated cells (Fig. 7) f r o m the basal c o m p a r t m e n t during overnight incubation. The same organelles contained H R P after short-term incubation with the tracer in the apical medium (not shown). Although it was rarely found, cationic ferritin was seen lining the apical m e m b r a n e of the cells and within cell vacuoles after incubation in the apical c o m p a r t m e n t (not shown).
516
Fig. 1. Non-ciliated epithelial cells containing dark vacuoles and lacking clear vacuoles from the ductuli efferentes of an 85-year-old man. Monolayer cultured for 41 days. Note sparse stereocilia, wellmaintained Golgi complexes interspersed with vacuoles. F Filter; A apical surface, x 11 600; bar = 1 g m
Fig. 2. Junctions between non-ciliated cells displaying visible intercellular spaces, containing lateral interdigitations, extending from the filter (right side) to the apical surface. A r r o w h e a d Tight junction. Tissue from a 39-year-old male; monolayer in culture for 49 days. x 7 500; bar = 2 gm
517
Fig. 3. Ciliated cell from same tissue as Fig. 2. Note cilia (arrow), prominent nucleolus, mitochondria thoughout the cytoplasm and supranuclear Golgi apparatus (arrowheads), lipofuscin-filled lysosome (L) and distended intercellular spaces, x 8900; bar = 2 lam
Fig. 4. Non-ciliated cells from same tissue as Fig. 2. Note clear vacuoles and short microvilli at apical surface and non-distended intercellular interspaces (arrowheads). x 6200; bar=2 gm
Fig. 5. Lateral intercellular spaces containing per0xidase up to the tight junction (arrowhead) after basal application of the enzyme. Tissue from 87-year-old patient; 24-day culture, x3200; bar= 0.2 gm
518
Fig. 6. Non-ciliated cells overlying epithelial fragments. Note concentration of dark vacuoles near the apex and lighter vacuoles towards the base. Tissue from 72-year-old male; 28-day culture. x 8400; bar = 2 ~tm Fig. 7. Unstained section of ciliated ceti (upper) and non-ciliated cell revealing presence of peroxidase product (dark deposits) in
vesicles (arrowheads) after apical application. C Ciliary rootlet. 24day culture. • 29200; b a r = 0 . 5 I.tm Fig. 8. Well-preserved junctional complex between cells cultured for 41 days. Same tissue as Fig. 1. x 32400; b a r = 0 . 2 gm
Discussion
the knots o f tubules that are present in this region of the epididymal head (Yeung et al. 1991), and a compromise between the a m o u n t of epithelial and connective tissues always had to be made. For example, the fine tubules (lined by epithelium II) branching from the
The culture of h u m a n efferent ducts was more difficult than that of the h u m a n corpus epididymidis, owing to the tough nature of the connective tissue surrounding
519 straight white tubules leaving the testis are likely to be lost during such harsh enzymic and physical treatment. Nonetheless, several epithelial cell types were observed in the cultures: non-ciliated cells with dark vacuoles and clear vacuoles and ciliated cells. Because of the short stature of the cells in the monolayer over the filter, also observed in cultures of human corpus and rat caput and cauda (Cooper et al. 1990; Yeung et al. 1989), the polarised disposition of the vacuoles (dark vacuoles situated apically in epithelium I and basally in epithelium III as described by Yeung et al. 1991) was not apparent. This polarity was maintained, however, in the taller cells over fragments. Low cuboidal epithelia (II, IV) were not distinguishable in cultures and it is not known from which epithelia the ciliated cells arose. That the ciliated cells beat in culture for up to one month attests to the viability of these cells; in cultures of rat efferent ducts they were observed to beat for 2 weeks (Byers et al. 1985). The ultrastructure of cultured cells revealed only sparse cilia on ciliated cells in the monolayer. It is not known if the non-ciliated cells plated out better than the ciliated cells or the latter lost their cilia with age, as was indicated by Byers et al. (1985) in the rat. Uptake of horseradish peroxidase from the apical and basal compartment into ciliated and non-ciliated cells was observed, and HRP-laden vesicles were seen throughout the cytoplasm, as is the case for human corpus epididymal cells (Cooper et al. 1990). However, endocytosis of cationic ferritin was rarely observed in efferent duct cells in culture, in sharp contrast to cells from the corpus region (Cooper et al. 1990). In contrast, rapid endocytosis of androgen-binding protein by rat efferent duct cells has been demonstrated in monolayer culture (Byers et al. 1985), a difference attributable to either species differences, the endocytosed proteins or the age of the cultures. For the first time for human epididymal cultures, p r o o f of confluent monolayers was provided by the observation of the apical medium being supported overnight by the cells comprising the monolayer. This was recently
demonstrated for rat Sertoli cells cultured on similar previous supports (Onoda et al. 1991). Cell-free chambers did not support fluid and the relationship between electrical resistance of the monolayer and the weight of supported fluid suggests that the epithelial cells limit the movement of both ions and fluid. Such polarised epithelia will be useful in future studies of vectorial epithelial secretion and transport.
Acknowledgements. We acknowledge the help of Kfilli Nurmik and Eva M611mann for obtaining the tissues and maintaining cell cultures, and Dr. F. Oberpenning for obtaining patient information, and thank Professor W. Wittkowski for the use of the electron microscope. The study was funded by the Deutsche Forschungsgemeinschaft Grant No. Ni 130/11-3/C. References
Byers SW, Musto NA, Dym M (1985) Culture of ciliated and nonciliated cells from rat ductuli efferentes. J Androl 6:271-278 Cooper TG (1990) In defense of a function for the human epididymis. Fertil Steril 90:965-975 Cooper TG, Yeung CH, Meyer R (1989) Immature rat epididymal epithelial cells grown in static primary monolayer culture on permeable supports. I. Vectorial secretion. Cell Tissue Res 256:567-572 Cooper TG, Yeung CH, Meyer R, Schulze H (1990) Maintenance of human epididymal epithelial cell function in monolayer culture. J Reprod Fert 90:81 91 Onoda M, Suarez-Quian CA, Djakiew D, Dym M (1990) Characterization of Sertoli cells cultured in the bicameral chamber system: relationship between formation of permeability barriers and polarized secretion of transferrin. Biol Reprod 43:672-683 Raczek S, Yeung CH, Hertle L, Schulze H, Cooper TG (1992) Monolayer cell culture of human efferent duct epithelium. Miniposter No. 114, 7th European Workshop on Molecular and Cellular Endocrinology of the Testis, Schlol3 Elmau, FRG, May 1992 Yeung CH, Cooper TG, Meyer R (1989) Immature rat epididymal epithelial cells grown in static primary monolayer culture on permeable supports. II. Histochemistry and ultrastructure. Cell Tissue Res 256:573-580 Yeung CH, Cooper TG, Bergmann M, Schulze H (1991) Organization of tubules in the human caput epididymidis and the ultrastructure of their epithelia. Am J Anat 191:261-279
