Anat Embryo1(1992) 186:259-270

Anatomyand Embryology 9 Springer-Verlag 1992

Late postnatal development and differentiation of the ductus epididymidis in a dasyurid marsupial (Antechinus stuartil) D.A. Taggart and P.D. Temple-Smith Department of Anatomy, Monash University, Wellington Rd, Clayton, Victoria, 3168, Australia Accepted April 29, 1992

Summary. The general histology and ultrastructural features of the developing ductus epididymidis were examined in the brown marsupial mouse, Antechinus stuartii, from April, when males were sexually immature, until August, when the adult males were involved in mating activities, just prior to the annual male die-off. Samples were also examined 3 and 6 months after the August die-off period in males kept in isolation from conspecifics during the prebreeding and breeding periods. In April, tubule diameter and epithelial height were largest in the caput and least in caudal segments but the reverse was observed thereafter. Epithelial height increased in caput segments in August and remained high in the post die-off samples. However, caput epithelial height and tubule diameters were low compared with the remainder of the duct from July until February. Luminal shape in caudal segments (10, 11 and 12) changed in June from circular to a narrow slit, and the epithelium became variable in height. The epididymal epithelium was undifferentiated with few cytoplasmic organelles in April. Differentiation occurred mostly from May to June in association with an increased abundance of cytoplasmic organelles, increasing prostatic weight and rising plasma androgen levels. Differentiated principal and basal cells were found in caput and corpus regions in May and in caudal segments in June in association with the de novo development of a brush border of microvilli. Few clear cells were seen in caput and corpus regions of the duct in May but they, and mitochondria-rich cells, were common throughout the duct from June. Development of the unusual structural features of the cauda epididymidis preceded the arrival of spermatozoa in June. The presence of degenerating spermatozoa and cytoplasmic droplets in the cauda at this time suggested that it was not yet capable of supporting sperm viability. There was no evidence to suggest that the presence of spermatozoa has a stimulatory effect on the epididymis. Intact sperm were observed throughout the duct from July. Free cytoplasmic droplets, which showed some evidence of degenCorrespondence to: D.A. Taggart

eration, collected in large masses in the distal corpus/ proximal cauda epididymidis of adult males between aggregates of spermatozoa. Epididymal differentiation appeared complete by mid-July; few ultrastructural changes occurred after this time. Recruitment of spermatozoa into the epididymis ceased by August and was associated with a rapid decline in sperm content in the proximal caput segments. In the November and February samples, spermatozoa were present only in distal corpus and proximal cauda segments. As in some eutherian mammals, differentiation of the epididymis in A. stuartii occurs in a descending fashion from caput to cauda. Development is linked to the onset of fluid and androgen production from the testis, which is essential for developing and maintaining a suitable caudal environment for storage and survival of spermatozoa.

Key words Epididymis - Development - Differentiation Marsupial - Reproduction

Introduction Although the structure und function of the marsupial epididymis has been described in several species (De Mello et al. 1982; Rodger 1982; Jones et al. J984; Temple-Smith 1984a, b; Cummins et al. 1986; Taggart and Temple-Smith 1989, 1990a; Temple-Smith and Taggart 1990) there are no published records of structural changes in the developing ductus epididymidis of any marsupial. This study examines the differentiation of the ductus epididymidis in the brown marsupial mouse, Antechinus stuartii, a small forest-dwelling representitive of the family Dasyuridae. The reproductive biology and life history of Dasyurid marsupials, especially A. stuartii, is very interesting. For example, dasyurid spermatozoa are long by mammalian standards (218 gm-271 lam; Cummins and Woodall 1985), sperm motility is unusual (Taggart and Temple-Smith 1990b) and the testis contains only a few, large-diameter seminiferous tubules

260

(Woolley 1966, 1975). Total spermatogenic failure occurs in the testis of A. stuartii prior to the breeding season (Kerr and Hedger 1983) and the cauda epididymidis has an unusual structure which appears to limit its sperm storage capacity during the breeding season (Taggart and Temple-Smith 1989, 1990a, 1990b). Male A. stuartii are polygynous (Scott and Tan 1985) and have a lek-type mating system (Lazenby-Cohen and Cockburn 1988). A few weeks after mating, all males in wild populations die off, often before the inseminated females have ovulated (Lee et al. 1977, 1982; Selwood and McCallum 1987). A. stuartii seemed an obvious choice for a study of epididymal development in marsupials not only because of the considerable background of information on its reproductive biology, but also because its entire life history is completed within 11 months. Each year males are born in mid to late September and die off in August of the following year (Lee et al. 1977). It is therefore possible to study age-related changes in epididymal structure and function during a highly predictable lifespan. Sampling in this study commenced in April, following the dispersal of all sexually immature male offspring (Lee et al. 1977, 1982) and coinciding with a rise in plasma androgen levels (Kerr and Hedger 1983), which peak in July and remain high until die-off. Particular attention has been given to: 1. The ultrastructure and timing of cellular and regional variation in the epithelium 2. The development of the unusual caudal structure of the epididymis (Taggart and Temple-Smith 1989) 3. Observations on the effect of androgens on development 4. Changes in epididymal ultrastructure in males held after die-off.

ment, stained in toluidine blue and examined with a Leitz Orthoplan microscope. For electron microscopy, silver/grey thin sections, again taken from the distal surface of each epididymal segment, were stained in aqueous uranyl acetate and lead citrate and examined in a Joel 100B electron microscope.

Morphometry. Mean tubule diameter and epithelial height for each region were determined from 50 duct cross sections per animal, using a Leitz ASM image analyzer (ASM) as described in Taggart and Temple-Smith (1989). Since the epithelium in regions 10, 11 and 12 showed extreme variability, the maximum and minimium heights of the epithelium were measured from 50 cross sections per region for each animal. Results were pooled, and the mean and standard deviation of the measurements were calculated using the ASM.

Results

Monthly weight changes Body weight increased from 23.4 g_+2.7 g in April to a maximum of 31.4 g_+5.9 g in July, and then declined to 28.8 g_+3.6 g in August, and remained at about this level until February. Testicular weight increased from 0.104g_+0.009 g in April to a maximum in June of 0.222 g+0.029 g, and then declined to 0.027 g_+0.005 g in November, and remained at this level until February (Fig. 1). Epididymal weight increased to a peak of 0.128 g_+ 0.011 g in July and then declined to a minimum of 0.05 g_0.004 g in November (Fig. 1). The prostate, however, was difficult to distinguish in April (0.007 g _ 0.0009 g) but showed a 40-fold increase in weight during

0.36 0.34

t--

0.32 0.30 0.28

Materials and methods Animal capture and maintenance. Four male A. stuartii were trapped each month from April (7 months old) to August (11 months old) in wet sclerophyll forest near Powelltown, Victoria. A further eight males (two groups of four individuals) were trapped in May and maintained in isolation until November (14 months old) and February (17 months old). Males isolated prior to the onset of territorial and breeding behavior, can be maintained after the normal die-off period in August and into the next breeding season (Woolley 1966). Trapping, handling and laboratory maintenance of animals have been described previously (Taggart and Temple-Smith 1990a). Tissue preparation and microscopy. Tissue samples were collected approximately in the middle of each month. Animals were anaesthetized and then perfused with a glutaraldehyde-based fixative (Taggart and Temple-Smith 1989). After 20 rain of fixation the epididymides, testes and prostate were removed, dissected free of fat, weighed and placed in 2.5% glutaraldehyde for a further 2 h at room temperature. Epididymides were then divided into twelve segments using the criteria described by Taggart and Temple-Smith (1989). For light (LM) and transmission electron microscopy (TEM) the 12 glutaraldehyde-fixed epididymal segments were routinely processed for TEM (Taggart and Temple-Smith 1989). Sections were cut at 1 gm from the distal surface of each epididymal seg-

PROSTATE i

:

~- 0.22

0.20 111

o.16

'..

i

Z

< o.16

TESTIS

on- 0.14

J

i \l

0.10

~

ii/

O. 12

Post die-off

"~ ".l

I

0.06

0.06

"-,,.

EPiOiO,Mis

o.o4

Ug':

............

0.02 i

0

J

i

i

i....i'"'~

F M A M J

i

J

i

i

i

t

i

i

i

J A S O N D J F M MONTH Fig. 1. Changes in testicular (solid line), epididymal (broken line) and prostatic weight (dotted line) in A. stuartii between April and February. Values are mean_+ SD for four males/month

261 140

APRIL f MAY

30

100 NS e~/ " ~

8O

~ I "~ T T~//

20

60 4O

10

2O i

I I

I r

-g

140, 9JUNE

i

h

i

i

i

L

L

I

I

JULY

120

3O

100

z

20

so

~: &u

=. ==

z

60

lO

40 2(~

0

0

FEBRUARY-Post

AUGUST

-~

_t

x Io. Lu

Epithelia! height. In April epithelial height was relatively constant (16.2 gm+2.3 g.m), except for the caput segments where it reached a maximum height of 22.3 gm + 3.1 gm. Epithelial height increased throughout the duct from May to August, stabilizing after this time. From May to February it was lowest in the distal caput and proximal corpus segments, and greatest in caudal regions (Fig. 2). From June epithelial height in the caudal portion of the duct varied greatly in association with the slit-like shape of the lumen in this region and the method used to estimate height (refer to Materials and methods). In this region of the duct, basement membrane and apical plasma membrane were often separated only by a thin band of cytoplasm- hence the slit-shaped cross-sectional appearance of the duct (Fig. 6b). Variation was largest in July (32.2 ~tm+28.3 ~tm for segment 12) corresponding with the peak in epididymal weight. Ultrastructural changes during development

die-off

140 30

120 100

20

80

60. 10

40 2O 0

i

L

i

i

i

i

i

i

L

~

,

i

0

1 2 3 4 5 S 7 8 9101112 1 2 3 4 5 6 7 8 9101112 EPIDIOYMAL SEGMENT NUMBER

Fig. 2. Changes in tubule diameter (broken line) and epithelial height (solid line) along the epididymisin A. stuartii from April to February post die-off(mean+ SD) the next 3 months to 0.334 g_+0.032 g in August, before declining to 0.169 g-t-0]028 g in November (Fig. 1).

Gross morphological changes during development Tubule diameter. Tubule diameter, which was relatively constant in April, increased in caudal segments (especially segments 11 and 12) in May. This trend continued from June through to February, when tubule diameter remained relatively constant in caput and proximal corpus regions, increased in distal corpus segments and reached a maximum in the caudal portion of the duct (Fig. 2). Luminal shape. In April the lumen of the duct was circular and varied from open in the caput to almost completely closed in the caudal regions (Figs. 2, 3). In May it enlarged, but remained circular in transverse section through the duct. In June luminal shape in the caudal regions changed from circular in some animals to oval and slit-like in others. From July until February luminal shape was characteristically narrow and slit-like in caudal regions (Fig. 6).

The undifferentiated epithelium. The epididymal epithelium in April consisted predominantly of a single layer of low columnar cells (Fig. 3 a, b). Ultrastructurally the epithelial cells in April appeared undifferentiated. Nuclei, often with prominent nucleoli, occupied most of the cell and were roughly circular in shape. Mitotic figures were rarely seen. The apical plasma membrane was relatively smooth throughout the duct, with the exception of the caudal region where it became irregular with obvious cytoplasmic protrusions containing small-fluidfilled vacuoles. Small vacuoles, mitochondria and the occasional vesicular inclusion were also observed in the apical cytoplasm. Rough (RER) and smooth endoplasmic reticulum (SER), Golgi and numerous ribosomes were also present (Fig. 3 c, d). A general increase in the abundance of cytoplasmic organelles in all cells was evident between April and June. Basal and mitochondria-rich cells appear to have started to differentiate in April when the remainder of the epididymal epithelium is characteristically undeveloped. Development of the mature epididymis appeared to have been completed by July as very few ultrastructural changes occurred after this time. Principal cells. By mid-May most of the low columnar cells observed in the caput and proximal corpus regions of the duct were developing into principal cells. Small numbers of long, unbranched or simple branching stereocilia projected into the lumen with the occasional cytoplasmic protrusion. Various organelles were common in the apical and supranuclear cytoplasm, including multivesicular bodies, SER, RER, mitochondria, Golgi bodies and lipid (Fig. 4). Nuclei, often containing prominent nucleoli, RER and a few small electron-dense membrane-bound vesicles were commonly observed in the basal region of these cells (Fig. 4). In the caput regions there was a marked increase in the abundance of both RER and SER, and a variety of vesicular inclusions appeared at various levels in the

262

Fig. 3 a-d. Changes in duct shape between (a) caput and (b) cauda epididymidis in April. Note the presence of spermatocytes (S) in the lumen (L) of the duct in caput regions and the small luminal opening in the duct in caudal segments at this time. x 120. e, d The undifferentiated epididymal epithelium in April. e Distal caput

epididymidis ( x 10,500). d Proximal cauda epididymidis ( x 8700). Note the lack of cell organelles and luminal surface specializations (large arrow) in caput regions and the cytoplasmic protrusions (small arrow) at the lumen in caudal regions. M, Mitochondria; iV, nucleus; B, basement membrane

263

Fig. 4. Oblique section through the proximal caput epididymidis of A. stuartii during May, showinga varietyof vesicular inclusions (/), large amounts of RER (R), basal cells (B), stereocilia(S), mitochondria (small arrow), lumen (L) and basement membrane (large arrow). Note also profiles of the apical portions of mitochondriarich cells (M), x 3000

differentiating principal cells. These inclusions were located in the apical and supranuclear regions (Fig. 4) and were characterised by either tiny membrane-bound vesicles (apical region) or membranous material in short parallel arrangements (supranuclear region). Both structures contained numerous electron-dense granules. The inclusions were of varying densities, up to 2 gm in diameter and were often associated with large Golgi complexes, membranous whorls, mitochondria and RER (Fig. 4). In the distal corpus and caudal regions of the epididymis, the epithelium appeared relatively undifferentiated, similar to that seen in April (Fig. 3d). Short stereocilia in low numbers however were present on the apical surface of some epithelial cells. Principal cells were identified throughout the epididymis in June, and with the exception of the caudal segments appeared similar to those described in May (Fig. 4). A lipid-rich zone, however, was present in the supranuclear region of the distal caput segments and a variety of supranuclear electrondense inclusions were present at various levels along the duct. An obvious increase in the abundance of cytoplasmic organelles, particularly endoplasmic reticulum, mul-

tivesicular bodies and smooth and coated vesicles was also observed in this phase of development. The most intriguing structural change in June occurred along the luminal surface of principal cells in the cauda epididymidis in association with the change in luminal shape. These principal cells were lined by a brush border of closely packed microvilli, each with internal filaments running longitudinally into the apical cytoplasm (Fig. 5 b). Canaliculi and coated vesicles were a common feature of the apical cytoplasm of principal cells in the caudal region at this time. Principal cells in more proximal regions of the duct were lined by stereocilia (Fig. 5 a), which showed no regional differentiation. During July there was an obvious reduction in the number of multivesicular bodies, vesicular inclusions and apical vesicles in principal cells of the caput region of the duct. The relative abundance of other cytoplasmic organelles appeared unaltered in July, although Golgi bodies were more commonly enountered in corpus segments. Principal cell ultrastructure in August was similar to that described in July, except for the presence of low numbers of small electron-dense inclusions which appeared in the perinuclear and supranuclear regions of the cell in the distal corpus and cauda epididymidis. The most prominent features of the duct in August were the large aggregations of lipid situated basally and in the supranuclear regions of principal cells in caput segments (Fig. 6a). In the November and February post-die-off samples, principal cell ultrastructure resembled that seen during the breeding season with four notable exceptions: 1. Numerous small vacuoles occurred throughout the duct in supranuclear and apical regions (Fig. 5 a). 2. Large numbers of translucent inclusions were seen in the apical cytoplasm of the distal caput. 3. Dense inclusions, particularly membranous whorls, appeared in the perinuclear, supranuclear and basal regions of the distal caput, distal corpus and proximal cauda segments (Fig. 5 b). 4. A marked increase in small, circular, electron-dense vesicles was observed in the supranuclear region of principal cells in distal cauda epididymidis. Mitochondria-rich cells. In contrast to the generally undifferentiated appearance of the cells lining the duct in April were a few dark-staining low columnar cells restricted to the caput and corpus regions (Fig. 7a). These cells had a centrally placed irregular-shaped nucleus with large amounts of clumped heterochromatin. Nucleoli were not visible in any profiles of cell nuclei, and cell surface specializations appeared absent. Extensive apical vacuolations were also common in these cells. A small Golgi apparatus in the supra-nuclear region, scattered mitochondria and moderate amounts of RER and SER were present throughout the cytoplasm. This cell type is an immature form of the mitochondria-rich cell which is observed in later months. Mature mitochondria-rich cells were present in May. The immature form, however, was still found in very low numbers at this time, but was not observed in the

264

Fig. 5. a Ultrastructural features of principal cells in the proximal caput epididymidis in February (6 months post male die-off). Note the large numbers of small vacuoles (large arrow) in supranuclear and apical regions of principal cells at this time. Stereocilia (small arrow); N, nucleus; B, basal cell; L Lumen. x 4400. b Ultrastructural features of principal cells in the proximal cauda epididymidis

in February (6 months post die-off). Note the dense inclusions and membranous whorls (large arrow) in supranuclear regions of the cell. Also note the presence of particulate matter (small arrow) in the lumen (L) and apical and supranuclear regions of the cell. MV, microvillus brush border; N, nucleus; B, basal cells, x 4000

following months. From June until February narrow, columnar mitochondria-rich cells occurred in low numbers throughout the duct. These cells contained dense aggregations of mitochondria above and below a basal nucleus, but had few other cytoplasmic organelles. The apical surface was relatively smooth and undulating throughout (Fig. 7 b).

caput region and lacked the characteristic deep nuclear infoldings which are typical of mature basal cells in this species. The nucleus is quite dense surrounded by little cytoplasm containing few organelles. Mature basal cells were a common feature of the caput and corpus segments of the epididymis in May. In general they appear similar in structure of those from April samples, although by May the basal cell nuclei had developed deep infoldings. Basal cells could not be identified in the cauda epididymidis in May but were found in moderate numbers throughout the epididymis from June until Feb-

Basal cells. Small triangular-shaped cells located along the basal lamina in April appeared to be an early form of the basal cell. This cell type was found only in the

265

Fig. 6. a Section through the proximal caput epididymidis of A.

stuartii in August, showing numerous small vacuoles in the lumen (L) of the duct and accumulations of dense lipid inclusions (arrow) in the basal and supranuclear region of principal cells. Note the absence of spermatozoa. N, principal cell nucleus, x 160. b Section through the proximal cauda epididymidis of A. stuartii in August, showing changing duct shape, dense luminal aggregations of spermatozoa (S) and microvillus brush border (arrow). E, epithelium. x 180

Fig. 7a, b. Ultrastructural features of a an immature (April, • 12,450) and b a mature (June, x 6450) mitochondria-rich cell. Note the absence of cell surface specializations (small arrow), extensive apical vacuolation (V and large arrow) and the abundance of mitochondria (M). BC, basal cell; PC, principal cell; L, lumen; N, nucleus

266

Fig. 8. Degenerating spermatozoa in the lumen (L) of the cauda epididymidisof A. stuartii in June. Note the degenerating sperm tails (Dst), disorganized axoneme (?), abundance of particulate matter (P), (possibly chromatin strands from decondensed nuclei) in the lumen, and the breakdown of connecting fibres (small arrow). F, dense fibres, x 20,540

ruary (Figs. 5, 7). No ultrastructure differences were observed between basal cells from any region of the mature epididymis. Clear cells. Clear cells which spanned the height of the epithelium were first seen in low numbers in May. These cells featured a large, central, heterochromatic nucleus, large cytoplasmic vacuoles and a few mitochondria, lipid droplets and membrane-bound vesicles. Other organelles were scarce or absent, and few luminal surface specializations were observed. Clear cells were in low numbers and of similar structure the duct between June and August but were most numerous in caudal segments. Clear cells were more commonly observed in November and February samples. Luminal contents and the arrival of spermatozoa

The lumen in April was fluid-filled and, when open, often contained particulate or cellular matter and occasional primary spermatocytes (Fig. 3 a). In contrast, in May the lumen was relatively clear and free of debris. A number of significant events occurred during June which affected the luminal contents; in particular the arrival of sperm in the lumen. Sperm and their associated cytoplasmic droplets were found in all regions of the duct in June, though in caudal segments they were in various stages of degeneration, which included disruption of the sperm plasma membrane and mitochondria of the midpiece, and the disarray of dense fibres and other structures of the axoneme. Sperm nuclei were occasionally seen decondensing, and a considerable amount of particulate and membranous debris was present in the lumen at this time, including material released from the breakdown of cytoplasmic droplets (Fig. 8).

Intact spermatozoa and their associated cytoplasmic droplets were a prominent feature of the lumen in July, appearing in greatest concentrations in the distal corpus and proximal caudal segments. Free cytoplasmic droplets occurred in large aggregates, particularly in distal corpus segments, which were interspersed between aggregates of spermatozoa in distal regions of the duct. In August few, if any, spermatozoa were present in the proximal caput epididymidis (Fig. 6). However, along with large quantities of particulate and membranous material they were a common feature of the luminal content in distal caput, corpus and caudal regions (Fig. 6). Degenerating spermatozoa were not seen in the duct in August, but degeneration of some free cytoplasmic droplets was observed in distal corpus and caudal regions during these months. The presence of vesicles within the principal cells which contained material similar to that released by degenerating cytoplasmic droplets suggested that uptake of this material from the lumen was occurring. Similar observations of the uptake of luminal material by principal cells were made from samples taken after the period of male die-off (Fig. 5b). Some degenerating spermatozoa were found in the distal corpus and proximal cauda epididymidis (segments 8-10) in both the November and February postdie-off samples (Fig. 9). The nuclei of these sperm had decondensed and the plasma membrane, mitochondria and axoneme were frequently seen in disarray. Spermatozoa were absent from all other regions of the duct. The lumen throughout the epididymis, including the regions with spermatozoa present, contained large quantifies of particulate and membranous material.

267 the breeding season is unusual in mammals. Testicular and epididymal weight usually remain high throughout the breeding season, and either decline soon after mating finishes and rise again in a cyclical manner during the next breeding season (e.g. ferret), or alternatively (e.g. rat) remain high from the onset of sexual naaturity until death (Setchell 1978).

The influence of androgens

Fig. 9. Section through the luminal region of the proximal cauda epididymidis in February (6 months post male die-off). Note the abundance of particulate matter and membranous material (P) present in the lumen (L), along with degenerating sperm tails in transverse section. Breakdown of mid-piece fibre network (F), degeneration of axonemal elements (latge arrow) and mitochondrial sheath (small arrow). D, dense fibres; MV, microvillus brushborder; T, terminal web; V, vacuoles, x 7850

Discussion

The testis and epididymal development Changes in testis weight of A. stuartii described in this study correspond closely with those reported by Woolley and Kerr and Hedger. In each study testis weight increased from April to June and then declined in July in association with the release of spermatozoa into the epididymis (Woolley 1966; Kerr and Hedger 1983). In contrast epididymal weight began to increase one month later in May, as plasma androgen levels begin to rise (Kerr and Hedger 1983), and peaked in July. The decline in epididymal weight in August is probably associated with the cessation of flow of spermatozoa and other testicular products in males at this time (Woolley 1966; Kerr and Hedger 1983; Taggart and Temple-Smith 1990a). Declining testicular weight prior to or during

Androgens are essential for the early post-natal differentiation of the epididymis and for the maintenance of its differentiated state during adulthood (Orgebin-Crist et al. 1975; Podesta et al. 1975). Androgens in both testicular fluid and blood plasma are known to influence the developing epididymis (Alexander 1972; Setty and Jehan 1977). The epididymis itself has also been shown to be capable of synthesizing and metabolizing steroids (Hamilton et al. 1969; Hamilton 1972) which are added to the luminal pool. The presence of spermatocytes in the lumen of the caput epididymidis in April indicated that testicular fluids, and possibly androgens, flow into the epididymis well before sperm release in A. stuartii. The presence of spermatids in the epididymal ducts of 21-day-old rats suggests a similar conclusion (Setchell 1970; Cooper and Waites 1974). As testicular activity commenced in A. stuartii there was a progressive increase in organ weight down the tract. Testicular weight reached a maximum about a month in advance of the epididymis, which in turn peaked before that of the prostate, suggesting that development and normal functioning of the epididymis is directly influenced by materials produced by the testis (de Kretser et al. 1982), or alternatively that the epididymis is more sensitive to androgens than the prostate. Hedger (1979) found that Leydig cell volume and the amount of smooth endoplasmic reticulum and lipid inclusions within the Leydig cell cytoplasm of Antechinus, from the Powelltown population, progressively increased from April to a maximum in June. During this period both the steroidogenic and spermatogenic activity of the testis increased to a peak in June together with sperm production by the seminiferous epithelium. Since development of the caput epididymidis in brown marsupial mice begins in April/May in advance of more distal regions of the epididymis, when circulating plasma androgens are low (Kerr and Hedger 1983), epididymal development in this region at that time appears to be influenced primarily by androgens in testicular fluids rather than by circulating androgens. Similar observations have been made in the rat (Podesta and Rivarola 1974). In 21-day-old rats, Leydig cells have been shown to be functional, although the epididymis is still histologically undifferentiated and circulating plasma androgen levels are low (Knorr et al. 1970; Podesta and Rivarola 1974). The caput epididymidis, however, has both dihydrotestosterone and cytoplasmic receptors (Calandra et al. 1974) along with increased levels of androgen-binding protein. This suggests that androgens, which are probably of testicular origin, are in-

268 fluencing the development of the proximal region of the epididymis at this age. Also in the rat (Setty and Jehan 1977), early differentiation of the initial segment of the epididymis may be associated with the fact that the initial segment is the first region to be exposed to testicular fluids and therefore increasing concentrations of androgens (Cooper and Waites 1974). The progressive development of stereocilia along the epididymis in prepubertal rats is thought to be androgen-mediated because treatment with exogeneous testosterone hastens their appearance (Setty and Jehan 1977). Gupta et al. (1974) suggested that the delay in development between caput and caudal regions in the rat may be related to a low receptor concentration in the cauda epididymidis of maturing animals and/or a higher requirement for androgens by this region of the duct. The wave of epididymal development that occurs down the duct from April to June in Antechinus, as shown by the changing distributions of principal and basal cells during this period, and the appearance of luminal surface specializations, is also therefore likely to be directly related to one or a combination of factors. These include the passage of testicular fluids and androgens along the epididymis, regional sensitivity to androgens by the epididymal epithelial cells, and/or a higher requirement for androgens by the caudal region of the duct (Woolley 1966; Gupta et al. 1974; Suzuki and Racey 1976). Regional sensitivity to androgens may also account for the open and closed nature of the duct in April when epididymal development and subsequent luminal enlargement commences earlier at some sites along the duct than at others. The structural differences may, however, have developed in response to androgens produced by the epididymal epithelium at various sites (Hamilton et al. 1969; Hamilton 1972). The period of maximum differentiation and development in the epididymis of A. stuartii occurs between May and June. All adult cell types can be observed throughout the duct at this time, in addition to a change in duct shape, the appearance of microvilli on the apical surface of principal cells in caudal regions, a large increase in the abundance of cytoplasmic organelles and the arrival of sperm. These changes appear superficially similar to those detected in the rat during the period of maximum differentiation (Setty and Jehan 1977; Sun and Flickinger 1979) - between the onset of androgen production at 21 days, and the stabilization of the Leydig cell population at 60 days. In A. stuartii, they correspond to a rise in circulating plasma androgen levels (Kerr and Hedger 1983) which also causes substantial increases in prostatic weight in this phase of maturation. Bradley et al. (1980), have suggested that the weight of the prostate in brown marsupial mice can be used as a possible index of androgen activity on target organs. As the rise in androgen levels comes in a period of active differentiation of the epididymis, it suggests that androgens have an influence on the development of the epididymis in this animal which is similar to that reported in the rat (Knorr et al. 1970; Setty and Jehan 1977; Sun and Flickinger 1979).

Sperm arrival and its relationship to epididymal development The arrival of sperm in the epididymis appears to have very little effect on the morphology of the duct (Sun and Flickinger 1979, this study). Degeneration of sperm and cytoplasmic droplets in the cauda epididymis in June supports the idea discussed previously of descending development and indicates that, in contrast to more proximal regions of the epididymis, the cauda epididymidis in June is not yet functionally mature and therefore not yet able to provide a suitable environment to maintain sperm viability. This is an interesting observation since it suggests that principal cells lining the caudal regions of the epididymis may be unable to regulate the luminal environment sufficiently at this time to provide conditions suitable for sperm storage and maintenance. Degenerating spermatozoa are also found in the terminal segment of the mature honey possum, Tarsipes rostratus, in association with large masses of disintegrating cytoplasmic droplets (Cummins et al. 1986). Lytic enzymes contained within mammalian cytoplasmic droplets (Dott and Dingle 1968; Garbers etal. 1970) may be responsible for sperm degeneration. Wilton et al. (1985) described a type of human male infertility termed epididymal necrozoospermia which is characterized by the degeneration of sperm during passage through or storage in the epididymis. Their observations suggested that this may result from either a hostile environment within the epididymis, or an inherent structural instability in the spermatozoa which causes them to break down with time resulting in dead and degenerating spermatozoa in the ejaculate. In A. stuartii spermatozoa are found intact throughout the duct in July, although principal cells in all regions appear structurally similar to those described in June. A slight increase in epithelial height and the modification of duct shape in caudal segments between June and July may indicate that differentiation in the cauda epididymidis is, as suggested earlier, not complete until July. Because total spermatogenic failure and the associated collapse of the seminiferous epithelium had occurred by early August in brown marsupial mice (Kerr and Hedger 1983) it was not surprising to see a depletion of spermatozoa in proximal caput regions of the epididymis, as the last of the spermatozoa passed out from the testis. Taggart and Temple-Smith (1990 a) have demonstrated that even in unmated males very few spermatozoa remained in the proximal caput region after mid August, and that less than 0.2 x 106 spermatozoa/epididymis were found in the distal corpus and proximal caudal regions of the duct in late August. Spermatozoa were present only in these segments in the November and February after die-off. This supports previous suggestions (Taggart and Temple-Smith 1989, 1990a) that the storage region in this species is proximal to the distal cauda epididymidis. Ultrastructural observations showed that these spermatozoa were in advanced stages of degeneration. This degradation of spermatozoa is perhaps due to the deleterious effects on the epididymis of declining androgen levels, as shown by a marked fall

269 in prostatic weight after the die-off period (Fig. 3 ; Woolley 1966).

Developmental changes in the epithelium of the cauda epididymidis As cellular and regional differentiation o f the epididymis progressed in Antechinus stuartii the differences in tubule diameter a n d epithelial height between the various segments became m o r e apparent, especially in the caudal region. The significance o f the slit-shaped lumen o f the duct in this region, the extreme variability observed in principal cell height, and the presence o f a microvillus brush b o r d e r have all been discussed in detail previously (Taggart and Temple-Smith 1989). D e v e l o p m e n t o f these caudal structures and shape changes occurred whilst the duct was fluid-filled, prior to the arrival o f s p e r m a t o z o a in the epididymis. The modification o f duct shape in caudal regions is p e r h a p s m a d e easier in the absence o f spermatozoa. The luminal surface o f principal cells in c a u d a epididymidis remains relatively s m o o t h with the exception o f the occasional cytoplasmic protrusion, until the appearance o f a squat yet highly ordered brush b o r d e r o f microvilli in June ( H a r d i n g et al. 1982; Taggart and Temple-Smith 1989). This brush b o r d e r does n o t develop f r o m stereocilia or any modification t h e r e o f but originates de n o v o f r o m a relatively featureless luminal surface and is thought, t h r o u g h its terminal web, to give structural s u p p o r t to m a i n t a i n luminal shape in this region (Taggart a n d Temple-Smith 1989). The differentiation, development, and functional m a t u r a t i o n o f the male reproductive tract in A. stuartii appears therefore, to be completed in July in close sync h r o n y with reproductive events in the female and in time for the c o m m e n c e m e n t o f the breeding season in late July/early August.

Acknowledgements. We thank Jenny McKervey and Brian Lloyd for photographic assistance; Sue Simpson for help with preparation of diagrams, and the Victorian Department of Conservation, Forests and Lands (permit no. 84/53) for approval to conduct this investigation.

References Alexander NJ (1972) Prenatal development of ductus epididymidis in the Rhesus monkey. The effects of fetal castration. Am J Anat 135:119-134 Bradley AJ, McDonald IR, Lee AK (1980) Stress and mortality in a small marsupial (Antechinus stuartii, Macleay). Gen Comp Endocrinol 40 : 188-200 Calandra RS, Podesta EJ, Rivarola MA (1974) Tissue androgens and androphilic proteins in rat epididymis during sexual development. Steroid 24:507-518 Cooper TG, Waites GMH (1974) Testosterone in rete testis fluid and blood of rams and rats. J Endocrinol 62:619-629 Cummins JM, Woodatl PF (1985) On mammalian sperm dimensions. J Reprod Fertil 75:153-175 Cummins JM, Temple-Smith PD, Renfree MB (1986) Reproduction in the male honey possum (Tarsipes rostratus: Marsupialia). The epididymis. Am J Anat 177:385-401

De Kretser DM, Temple-Smith PD, Kerr JB (1982) Anatomical and functional aspects of the male reproductive organs. In: Frick J, Bandhauer M (eds) Disturbances in male fertility. Handbook of urology, vol xvi. Springer, Berlin Heidelberg New York, pp 1-131 De Mello VR, Orsi AM, MelloDias S, Oliveira MC (1982) Fine structure of principal cells and basal cells of the cauda epididymidis of the South American Opossum (Didelphis azarae). Anat Anz Jena 151:497-502 Dott HM, Dingle JT (1968) Distribution of lysosomal enzymes in the spermatozoa and cytoplasmic droplets of bull and ram. Exp Cell Res 52:523 540 Garbers DL, Wakabayashi T, Reed PW (1970) Enzyme profile of the cytoplasmic droplets from bovine epididymaI spermatozoa. Biol Reprod 3:327-337 Gupta G, Rajalakshmi M, Prasad MRN (1974) Regional differences in androgen thresholds of the epididymis of the castrated rat. Steroids 24:575 586 Hamilton DW (1972) The mammalian epididymis. In: Balwin W, Glasser S (eds) Reproductive biology, Excerpta Medica, Amsterdam, pp 268-337 Hamilton DW, Jones AL, Fawcett DW (1969) Cholesterol biosynthesis in the mouse epididymis and ductus deferens: a biochemical and morphological study. Biol Reprod 1:167-184 Harding HR, Woolley PA, Shorey CD, Carrick FN (1982) Sperm ultrastructure, spermiogenesis and epididymal sperm maturation in dasyurid marsupials: phylogenetic implications. In: Archer M (ed) Carnivorous marsupials. R Zool Soc NSW, Sydney, pp 659-673 Hedger MP (1979) Structural, hormonal and biochemical studies on the testis of the marsupial mouse Antechinus stuartii: a model of spontaneous spermatogenic failure. Thesis, Department Physiology, Monash University, Melbourne, Australia Jones RC, Hinds LA, Tyndale-Biscoe CH (1984) Ultrastructure of the epididymis of the tammar wallaby, Macropus eugenii, and its relationship to sperm maturation. Cell Tissue Res 237:525 535 Kerr JB, Hedger MP (1983) Spontaneous spermatogenic failure in the marsupial mouse Antechinus stuartii Macleay (Dasyuridae: Marsupialia). Aust J Zool 31:445-466 Knorr DW, Vanha-Pettula T, Lipsett MB (1970) Structure and function of the rat testis through pubescence. Endocrinology 86:1298-1304 Lazenby-Cohen KA, Cockburn A (1988) Lek promiscuity in a semelparous mammal, Antechinus stuartii (Marsupialia: Dasyuridae)? Behav Ecol Sociobiol 22:195-202 Lee AK, Bradley AJ, Braithwaite RW (1977) Corticosteroid levels and male mortality in Antechinus stuartii. In: Stonehouse B, Gilmore D (eds) The biology of marsupials. Macmiilan, London, pp 209 220 Lee AK, Woolley P, Braithwaite RW (1982) Life history strategies of Dasyurid marsupials. In : Archer M (ed) Carnivorous marsupials, Roy Zool Soc NSW, Sydney, pp 1-11 Orgebin-Crist MC, Danzo BJ, Davies J (1975) Endocrine control of the development and maintenance of sperm fertilizing ability in the epididymis. In: Greep RO, Astwood EB (eds) Handbook of physiology, vol v, Endocrinology, section 7. Am Physiol Soc, Washington, pp 319-338 Podesta EJ, Calandra RS, Rivarola MA, Blaquier JA (1975) The effect of castration and testosterone replacement on specific proteins and androgen levels of the rat epididymis. Endocrinology 97:2, 399-405 Podesta EJ, Rivarola MA (1974) Concentration of androgens in whole testis, seminiferous tubules and interstitial tissue of rats at different stages of development. Endocrinology 95:455-461 Rodger JC (1982) The testis and its excurrent ducts in American caenolestid and didelphid marsupials. Am J Anat 163:269-282 Scott MP, Tan TN (1985) A radiotracer technique for the determination of male mating success in natural populations. Behav Ecol Sociobiol 17 : 29-33

270 Selwood L, McCallum F (1987) Relationship between longevity of spermatozoa after insemination and the percentage of normal embryos in the brown marsupial mouse (Antechinus stuartii). J Reprod Fertil 79:495-503 Setchell BP (1970) The secretion of fluid by the testis of rats, rams and goats with some observations on the effect of age, cryptorchidism and hypophysectomy. J Reprod Fertil 23:7%85 Setchell BP (1978) The mammalian testis. Elek, London, pp 1-29 Setty BS, Jehan O (1977) Functional maturation of the epididymis in the rat. J Reprod Fertil 49:317-322 Sun EL, Flickinger CJ (1979) Development of cell types and of regional differences in the postnatal rat epididymis. Am J Anat 154: 27-56 Suzuki F, Racey PA (1976) Fine structural changes in the epididyreal epithelium of moles (Talpa europaea) throughout the year. J Reprod Fertil 47 :47-54 Taggart DA, Temple-Smith PD (1989) Structural features of the epididymis in a dasyurid marsupial (Antechinus stuarti O. Cell Tissue Res 258:203-210 Taggart DA, Temple-Smith PD (1990a) The effects of breeding season and mating on total number and relative distribution of spermatozoa in the epididymis of the brown marsupial mouse, Antechinus stuartii. J Reprod Fertil 88:81 91

Taggart DA, Temple-Smith PD (1990b) An unusual mode of progressive motility in spermatozoa from the dasyurid marsupial, Antechinus stuartii. Reprod Fertil Dev 2:107-114 Temple-Smith PD (1984a) Phagocytosis of sperm cytoplasmic droplets by a specialized region of the epididymis of the brushtailed possum, Trichosurus vulpecula. Biol Reprod 30:707-720 Temple-Smith PD (1984b) Reproductive structures and strategies in male possums and gliders. In: Smith AP, Hume ID (eds) Possums and gliders, Surrey Beatty, with Australian mammal society, Sydney pp 89-106 Temple-Smith PD, Taggart DA (1990) On the male generative organs of the koala (Phascolarctos cinereus) : an up date. In: Lee A, Handasyde KA, Sanson G (eds) The biology of the koala. Surrey Beatty, Sydney, pp 33-54 Wilton LJ, Teichtahl H, Temple-Smith PD, De Kretser DM (1985) Structural heterogeneity of the axonemes of respiratory cilia and sperm flagella in normal men. J Clin Invest 75:825-831 Woolley PA (1966) Reproduction in Antechinus spp and other dasyurid marsupials. Syrup Zool Soc London, 15:281-294 Woolley PA (1975) The seminiferous tubules in dasyurid marsupials. J Reprod Fertil 45 : 255-261