The koala (Phascolarctos cinereus) and southern hairy-nosed wombat (Lasiorhinus latifrons) are iconic Australian fauna that share a close phylogenetic relationship but there are currently no comparative studies of the seminiferous epithelial cell or testicular microanatomy of either species. Koala and wombat spermatozoa are unusual for marsupials as they possess a curved stream-lined head and lateral neck insertion that superficially is similar to murid spermatozoa; the koala also contains Sertoli cells with crystalloid inclusions that closely resemble the Charcot–Bottcher crystalloids described in human Sertoli cells. Eighteen sexually mature koalas and four sexually mature southern hairy-nosed (SHN) wombats were examined to establish base-line data on quantitative testicular histology. Dynamics of the seminiferous epithelial cycle in the both species consisted of eight stages of cellular association similar to that described in other marsupials. Both species possessed a high proportion of the pre-meiotic (stages VIII, I – III; koala – 62.2 ± 1.7% and SHN wombat – 66.6 ± 2.4%) when compared with post-meiotic stages of the seminiferous cycle. The mean diameters of the seminiferous tubules found in the koalas and the SHN wombats were 227.8 ± 6.1 and 243.5 ± 3.9 μm, respectively. There were differences in testicular histology between the species including the koala possessing (i) a greater proportion of Leydig cells, (ii) larger Sertoli cell nuclei, (iii) crystalloids in the Sertoli cell cytoplasm, (iv) a distinctive acrosomal granule during spermiogenesis and (v) a highly eosinophilic acrosome. An understanding of the seminiferous epithelial cycle and microanatomy of testis is fundamental for documenting normal spermatogenesis and testicular architecture; recent evidence of orchitis and epididymitis associated with natural chlamydial infection in the koala suggest that this species might be useful as an experimental model for understanding Chlamydia induced testicular pathology in humans. Comparative spermatogenic data of closely related species can also potentially reflect evolutionary divergence and differences in reproductive strategies.
Spermatogenesis of mammals results in the timed sequence of cellular associations in the seminiferous epithelium resulting in repeated generations of germ cells surrounded by supporting Sertoli cells (Hess, 1999). This process can be separated into three phases; the mitotic or proliferative phase, the meiotic phase, and the spermiogenic phase, in which three types of germ cells (spermatogonia, spermatocytes and spermatids) are arranged in the seminiferous tubules and whereby their relative association can be characterized as different stages of the seminiferous epithelium cycle (Courot et al. 1970; Russell et al. 1990).
In marsupials, the main events of spermatogenesis are generally similar to those of eutherian mammals (Harding et al. 1979, 1982); however, the spermatids of most marsupials show unusual morphological changes during spermiogenesis and include: (i) a flattening of the nucleus at right angles to the long axis of the flagellum; (ii) nuclear rotation about the point of attachment to the flagellum, possibly caused by invasions by Sertoli cell cytoplasm; and (iii) an incomplete coverage of the acrosome as cap over the spermatid nucleus, whereby it is restricted to a relatively small proportion of the dorsal nuclear surface (Harding et al. 1976, 1979; Tyndale-Biscoe & Renfree, 1987). The mode of nuclear flattening during spermiogenesis in both the koala and wombat follows the basic marsupial pattern except that the flagellar connection is laterally rather than centrally placed and as a result the neck insertion is disto-ventral rather than centro-ventral as in other marsupials (Harding & Aplin, 1990).
Although the koala and the wombat are categorized into the sub-order Vombatiformes, they are also thought to have had a long evolutionary separation (Kirsch et al. 1997; Osborne et al. 2002; Munemasa et al. 2006; Phillips & Pratt, 2008). Consequently, sperm production in the koala and wombat might be expected to exhibit convergent or divergent structural solutions with respect to spermiogenesis and characterisation of the various stages of the seminiferous epithelial cycle. Comparative studies of qualitative and quantitative evaluation of the testicular histology can potentially provide answers to important questions about testicular structure and function and may even reveal insights into differences of reproductive strategy, for example temporal changes and species differences in reproductive hormone secretion (Leydig cell) and sperm production.
Recently, Deif (2011) has reported the first description of orchitis and epididymitis associated with ascending chlamydiosis in the koala, with this organism being identified using histochemistry and PCRq. Deif (2011) has also reported the histopathology of these lesions and identified the organism in the tissues at the ultra structural level. An understanding of the histopathology of this infection requires a detailed comprehension of normal spermatogenesis and testicular cellular architecture. In addition, and given that chlamydiosis in the male koala is a natural infection and more widespread than previously thought (Deif, 2011), the koala may in fact be a useful model for studying the pathology of this organism in the human testis.
Although spermiogenesis in the koala and the wombat has been described at the electron microscopy level (Harding et al. 1987; Harding & Aplin, 1990), the kinetics of spermatogenesis has not yet been investigated. This study aimed to establish base-line data on the quantitative testicular histology of koala and southern hairy-nosed (SHN) wombat, with a focus on light microscopic characterisation of the seminiferous epithelial cycle and relative frequency of cellular associations. Our findings will be discussed in light of the close phylogenetic relationship between the koala and wombat, functional differences in testicular microanatomy between these species, and the potential use of the koala as a model for understanding Chlamydia mediated testicular histopathology in humans.
Materials and methods
Testes were collected from 18 sexually mature necropsied koalas (Queensland in August and September in from 1992 to 1994) and four sexually mature southern hairy-nosed wombats (South Australia in September in 2010). Koala testes were recovered from animals euthanized at the Moggill Koala Hospital (Environment Protection and Heritage, Queensland Government) and wombat testes were collected as part of population management program conducted under the authority of the South Australian Department of Environment and Heritage. Experimental procedures were approved by Animal Ethics Committee in the University of Queensland (SAS/261/10) and Nippon Veterinary and Life Science University (AEC1).
Following dissection, each testis was immediately placed in Bouin's fixative (Humason, 1977) after removing the top and bottom of the testis to facilitate penetration of the fixative. After an initial fixation of 2–3 h, testes were cut into about 5 mm thick slices and then re-fixed in Bouin's fixative overnight. Fixed tissues were transferred to 70% ethanol. For studies of testicular quantitative histology, six to eight tissue blocks were obtained from different sections of the right testis that were separated by at least 200 μm. Sections for histology were prepared using conventional histology and stained with haematoxylin-eosin staining (Humason, 1977). Histological preparations were viewed under standard light microscopy (Olympus BX51) for both quantitative measurements of microanatomy and photomicroscopy (Olympus DP2-BSW).
Random samples of 100 tubules from each testis were examined and measured using a light microscopy at 100× magnification with a calibrated micrometer eyepiece and high resolution photography using imagej 1.45 software (http://rsbweb.nih.gov/ij/). Transverse seminiferous tubule diameters were determined by taking two measurements at right angles to each other. The longitudinal and transverse lengths of the Sertoli cell nucleus were also measured in the same way using light microscopy at 200× magnification. Relative proportions of the various stages of the seminiferous epithelial cycle observed in this study were also determined by light microscopy and high resolution images (200× magnification) using photographic software (gimp 2.6.12, http://www.gimp.org/) with a Weibel randomized grid system (Weibel, 1979); a total of 2000 points were made for each individual animal.
The epithelium of seminiferous tubules in marsupials has essentially the same cell types as those found in eutherian mammals, namely, the various germinal cells (spermatogonia, spermatocytes and spermatids) and Sertoli cells. Descriptions of the various stages of the seminiferous epithelial cycle in the present study were therefore partly based on observations of eutherian mammals (Courot et al. 1970; Russell et al. 1990) and other marsupials (Mason & Blackshaw, 1973; Setchell & Carrick, 1973; Orsi & Ferreira, 1978; Queiroz & Nogueira, 1992; Lin et al. 2004; Lloyd et al. 2008; Phillips et al. 2008). The relative frequency of each stage was calculated.
Mean seminiferous tubule diameters, mean longitudinal and transverse length of Sertoli cell nucleus, mean ratios of the seminiferous tubule and intersititial tissue, and mean stage ratios of the various stages of the seminiferous epithelium in the koala and wombat were compared with a nonparametric U-test (P < 0.05) using pasw statistics 18.0 (SPSS Japan Inc., Tokyo, Japan).
Seminiferous tubule diameter and relative proportions of testicular tissue
The seminiferous tubule cross-sectional diameter and relative proportions of interstitial tissue in the koala and the SHN wombat are shown in Table 1. There were no significant differences in the seminiferous tubule diameter between the koala and the SHN wombat. The proportion of mean interstitial tissue in the koala was greater than that in the SHN wombat (Table 1, Fig. 1).
Table 1. Seminiferous tubule diameter (STD), longitudinal (SCLL) and transverse length (SCTL) of Sertoli cell nucleus (µm ± SEM), and mean proportions (% ± SEM) of the seminiferous tubule (ST) and intersititial tissue (IT) in the koala and the SHN wombat
Seminiferous tubule refers to all tissue within the seminiferous tubules including the lumen. Interstitial tissue includes Leydig cells, blood vessels, and connective tissue.
Statistically significant larger value (nonparametric U-test; P < 0.05).
Eight stages of the seminiferous epithelial cycle in the koala and the SHN wombat as defined in terms of spermatid nuclear morphology (Fig. 2) and cellular associations were recognized and are summarized in Table 2 and Figs 3 and 4. Type-A spermatogonia possessed an ovoid nucleus, their chromatin was homogeneous and dust-like and contained a prominent nucleolus. Intermediate spermatogonia had a small ovoid nucleus equipped with a small amount of heterochromatin lying along the inner aspect of the nuclear membrane. Type-B spermatogonia possessed a rounded nucleus and contained a moderate amount of heterochromatin lying along the inner aspect of the nuclear membrane. Preleptotene spermatocytes appeared similar to Type-B spermatogonia except that they were slightly smaller. Preleptotene–leptotene transition nuclei had a speckled appearance. Leptotene spermatocytes contained a round nucleus with fine chromatin threads, whereas the zygotene spermatocytes had thicker chromosomes than those of leptotene spermatocytes. Pachytene spermatocytes had a larger nucleus and nucleolus, and their chromosomal material was more disperse. Secondary spermatocytes had a round nucleus and were larger in nuclear diameter than round spermatids. The nuclear morphology of rod-shaped spermatids was different from that of round spermatids, possessing an elongated and cylindrical nucleus. Condensed spermatids had condensed chromatin and a small spherical nucleus. Elongated spermatids were defined by a dorsal-ventral flatting of the nucleus and the long axis of the nucleus orientated perpendicular to the flagellum but characteristically possessed a neck insertion in the caudal region of the nucleus. As the spermatids developed during the later stage of spermiogenesis, the nucleus became progressively curved proximally, ultimately resulting in a hook or sickle-shaped mature sperm nucleus.
Table 2. Cell types and cellular associations of the eight stages of the seminiferous epithelial cycle in the koala and the SHN wombat
The lumen of the early stage I tubule was lined with round spermatids in both species that elongated into rod-shaped nuclei prior to nuclear condensation. Round spermatids initially possessed a large round acrosomal vacuole that progressively collapsed during nuclear protrusion. The acrosomal granule in the koala spermatid during stage I was substantially more prominent and eosinophilic than that found in the SHN wombats. Outside the spermatids, there was a layer of pachytene spermatocytes and then a layer of Sertoli cells and ovoid type-A spermatogonia, which were present in all stages. Preleptotene spermatocytes lay adjacent to the basement membrane, as were preleptotene–leptotene transition cells.
The spermatids lining the lumen of the tubule had condensed round nuclei that had not yet elongated. The nucleus of the koala spermatid possessed a highly eosinophilic acrosomal matrix that was not discernible in the wombat nucleus. Pachytene spermatocytes were present outside the spermatid nuclei and the preleptotene spermatocytes had now developed to leptotene spermatocytes adjacent to the basement membrane.
The nuclei of the spermatids in stage III were dorso-ventrally flattened and elongated parallel to the basement membrane. Pachytene spermatocytes remained essentially unchanged but the nuclei of the outer leptotene spermatocytes had now formed homologous chromosomal pairs and entered the characteristic ‘bouquet configuration’ (Clermont, 1972) of the zygotene spermatocytes.
While the spermatid nuclei of stage IV remained essentially unchanged from stage III, stage IV was clearly differentiated by the presence of meiotic figures from the previous pachytene spermatocytes or, in some sections, the presence of the newly produced secondary spermatocytes and round spermatids; zygotene spermatocytes were also present along the adjacent basement membrane. It was not possible to distinguish the inter chromosomal clear areas of the diplotene spermatocytes.
The elongated spermatid nucleus on the luminal surface of the tubule had taken on a slightly curved morphology approaching the appearance to the nuclei of mature spermatozoa, with their long axes orientated perpendicular to the basement membrane of the seminiferous tubule. These curved spermatids surrounded a layer of early stage round spermatids that showed no light microscope evidence of early acrosomal formation. There were no pachytene spermatocytes in this stage, as zygotene spermatocytes were still present. Lining the basement membrane of stage 5 tubules were type-A and intermediate spermatogonia, although the later cells were difficult to differentiate.
The spermatids lining the lumen of stage VI were similar in morphology to those of stage V, although the curvature of the nucleus was slightly more pronounced. Outside the layer of curved spermatids were round spermatids with distinct acrosomal vacuoles. In the koala, the round spermatids possessed a distinctive acrosomal granule within its vacuole, whereas this granule was rarely found in the wombat spermatids. Stage VI was differentiated from stage V by the appearance of pachytene spermatocytes and type-B spermatogonia.
The characteristically hooked, re-curved spermatid nuclei of the koalas and the SHN wombats were clearly visible by stage VII on the luminal extremity of the tubule epithelium. The flagella of the spermatids were arranged radially and extended into the lumen. Outside this layer, round spermatids were found that possessed larger acrosomal vacuoles than those in stage VI. The round acrosomal granules found in the koala acrosomal vacuole in stage VI had flattened against the nuclear surface; smaller sized, less distinctive granules were only rarely observed in the wombat round spermatids in stages VII to I. The acrosomal vacuole of both species appeared to have flattened and there was a corresponding a nuclear concavity. Pachytene spermatocytes and type-A and type-B spermatogonia were also present in stage VII.
This was considered the stage of spermiation and was characterized by the presence of spermatozoa free within the lumen and the presence of residual bodies of cytoplasm on the luminal epithelium. The remainder of the stage VIII tubule was essentially similar to stage VII. Round spermatids with large acrosomal vacuoles were present, as were pachytene spermatocytes and type-A and type-B spermatogonia. Prominent flattened granules were found in the acrosomal vacuoles of round spermatids of the koala but were usually not detectable in the SHN wombat.
Relative frequencies of stages of the seminiferous epithelial cycle
The mean proportions (% ± SE) of each stage of the seminiferous epithelium cycle are shown in Table 3. The relative proportion of stages IV, VII and VIII was higher in the koala than in the SHN wombat, whereas stages III, VI were respectively lower. However, both the koala and the SHN wombat had a higher frequency of stages in the pre-meiotic phase (VIII, I–III) relative to the post-meiotic phase (V–VII) (Table 4).
Table 3. Mean (% ± SEM) ratios of the stage of the seminiferous epithelial cycle in the koalas and the SHN wombats
* Statistically significant larger value (nonparametric U-test; P < 0.05).
16.1 ± 1.46
22.1 ± 1.43
14.1 ± 1.14
5.0 ± 0.60*
11.5 ± 1.35
8.0 ± 0.71
13.3 ± 0.85*
10.0 ± 0.93*
15.9 ± 1.28
28.4 ± 2.29
19.7 ± 1.97*
2.6 ± 0.44
10.5 ± 0.46
12.6 ± 1.29*
7.7 ± 1.15
2.6 ± 0.74
Table 4. Mean (% ± SEM) proportions of the pre-meiotic and post-meiotic phases of the seminiferous epithelial cycle in the koalas and the SHN wombats
*Statistically significant larger value (nonparametric U-test; P < 0.05).
62.2 ± 1.65
5.0 ± 0.60*
32.8 ± 1.67
66.6 ± 2.40
2.6 ± 0.44
30.8 ± 1.99
Microanatomy of the Sertoli cell
The Sertoli cell nuclei of the koalas were significantly larger than those of the SHN wombats and possessed strongly eosinophilic crystalloid inclusions (Table 1, Fig. 5); no crystalloid inclusions were found associated with the wombat Sertoli cells.
The dynamics of the seminiferous epithelial cycle in the koala and the SHN wombat were fundamentally similar to the cellular associations previously described in eutherian mammals (Courot et al. 1970; Russell et al. 1990) and other marsupials (Mason & Blackshaw, 1973; Setchell & Carrick, 1973; Orsi & Ferreira, 1978; Queiroz & Nogueira, 1992; Lin et al. 2004; Lloyd et al. 2008; Phillips et al. 2008). Marsupial seminiferous epithelial cycles have been broadly characterized based upon the cumulative sum of frequency of stages between pre- and post-meiotic stages (Setchell & Carrick, 1973; Phillips et al. 2008). The koala and the SHN wombat have a greater proportion of pre-meiotic (62 and 67%, respectively) than post-meiotic stages and are therefore similar to the brush-tailed possum (Trichosurus vulpecula; 53%), the bandicoot (Permeles nausta; 56%), the red-necked wallaby (Macropus rufogriseus; 53%) and the greater bilby (Macrotis lagotis; 61%), but differ with respect to the tammar wallaby (Macropus eugenii; 39%) and the common ring-tail possum (Pseudocheirus peregrinus; 29%) (Setchell & Carrick, 1973; Phillips et al. 2008; Johnston et al. 2010).
The mean diameters of the seminiferous tubules found in the koalas and the SHN wombats (227.8 and 243.5 μm, respectively) were generally similar to those found in marsupials, e.g. the Virginia opossum (Didelphis virginiana; 230–250 μm), the common brush-tailed possum (250 μm), and the tammar wallaby (254 μm) (Setchell, 1977). These diameters are greater than the maximum diameter in the greater glider (Petauroides volans; 205 μm; Setchell, 1977), but generally smaller than the mean diameters of the Dasyuridae, which ranged from 360 μm in the Tasmanian devil (Sarcophilus harrisii) to 510 μm in the fat-tailed dunnart (Sminthopsis crassicaudata) (Woolley, 1975) and the ring-tail possum (Pseudocheirus peregrinus; 274 μm) (Phillips et al. 2008). Although the biological significance of such differences in tubular diameter requires further investigation, Setchell & Breed (2006) have suggested that the large cross-sectional diameter of some dasyurid seminiferous tubules may be related to the large size of their spermatozoa.
The proportion of interstitial tissue was different between the koalas (33%) and the SHN wombats (15%). This large proportion of interstitial tissue has previously been noted by Temple-Smith & Taggart (1990) but these authors gave no quantitative data. Similar high relative proportions have also been noted in the swamp antechinus (Antechinus minimus; 23–37%; Wilson & Bourne, 1984) and the greater bilby (37%; Johnston et al. 2010). The lower proportion of interstitial tissue reported for the SHN wombat in this study is similar to that found in the common ringtail possum (15%; Phillips et al. 2008). Although the function of such a high proportion of interstitial tissue of these marsupials is equivocal, the high proportion of Leydig cells in the testis may be associated with higher androgen secretion (Akingbemi et al. 1999) or perhaps even the production of pheromones. The Leydig cells of the domestic boar, for example, are also thought to be involved with the synthesis of odoriferous steroids (Fawcett et al. 1973). It is therefore interesting that koala also possesses a large sternal gland, the size and secretion of which appears to be under androgen control (Allen et al. 2010). In the domestic boar and the stallion, large amounts of oestrogen are also apparently secreted by the Leydig cells (Fawcett et al. 1973; Raeside & Reneud, 1983); perhaps a similar function occurs in the koala, as Lee & Carrick (1989) have noted that more oestradiol is usually found in males than is typical in females.
The round spermatid nuclei in the koalas and the SHN wombats both passed through an elongation phase in late stage I to produce what we have described as ‘rod’-shaped uncondensed spermatid nuclei; the formation of a ‘rod’-shaped spermatid nuclei has not been previously reported for any other marsupial. Another unique feature of spermiogenesis in the koalas and the SHN wombats was the progressive curvature of the proximal sperm head, which appeared to commence in stage V and continued through to stage VII to spermiation. The curved nature of the mature sperm head was first noted by Hughes (1965) and then Harding & Aplin (1990). Temple-Smith & Taggart (1990) also noted heterogeneity in the shape of koala sperm heads during spermiogenesis and a progressive increase in the proportion of sperm with recurved nuclear morphology after epididymal transit; these authors have suggested that the variable nature of nuclear shaping during spermiogenesis and epididymal transit in the koala was likely to result in a greater level of heterogeneity in the shape of the sperm head, a phenomenon which has also subsequently been noted by Wildt et al. (1991) and Johnston et al. (1994) in ejaculated semen. It is interesting that the same degree of sperm nuclear pleiomorphy has not been described in the ejaculated spermatozoa of the common wombat (MacCallum & Johnston, 2005).
In the early stages of spermiogenesis, the round spermatids of the koala possessed a acrosomal vacuole with a distinctive eosinophilic acrosomal granule, but this structure was not as prominent and certainly not as eosinophilic in the early wombat round spermatids. Our inability to detect a prominent acrosomal granule in the wombat spermiogenesis is perplexing, as Harding & Aplin (1990) clearly noted this structure when using the electron microscope but were not able to detect a proacrosomal granule characteristic of early acrosome formation in the koala. The strongly eosinophilic nature of the acrosome granule and its persistence with the development of the mature acrosome in the koala spermatid, compared with that in the wombat, is also perplexing and may be related to the presence of the highly eosinophilic Sertoli cell crystalloid inclusions only found in the koala.
The crystalloid inclusions previously reported in the Sertoli cell of the koala (Harding et al. 1982, 1987; Kerr et al. 1987; Harding & Aplin, 1990) were not present in the SHN wombats. Our results support the incidental observations of Harding & Aplin (1990), who noted the lack of crystalloids in the testis of a single SHN wombat. Similar crystalloids have also been reported to occur in humans (Lubarsch, 1896; Sohval et al. 1971), the pig (Toyama, 1975, 1976; Toyama et al. 1979) and sloth (Toyama et al. 1990), although there were differences in the ultrastructure and arrangement of filaments constituting these crystalloids. Although the function of these crystalloids remains unknown, the cytoplasmic processes of the Sertoli cell containing the crystalloids are cross-linked to the plasma membrane of germ cells, suggesting that they may stabilize the relationship between the Sertoli cell cytoplasm and the surrounding germ cells (Kerr et al. 1987; Temple-Smith & Taggart, 1990).
Although there is a growing literature on the effect of Chlamydia trachomatis on human males (Villegas et al. 1991; Satta et al. 2006; Fernandez et al. 2007; Cunningham & Beagley, 2008; Gallegos et al. 2008), studies of the effect of Chlamydiae on male non-human mammalian fertility are, by comparison, extremely limited (Deif, 2011). To date, there has been no systematic review of existing information on the pathology associated with chlamydial infection in the genital tract or semen in male non-human mammals. Inflammation of the testis usually occurs within the seminiferous tubule and can disrupt spermatogenesis or even result in complete degeneration of the seminiferous tubule (e.g. humans – Allam et al. 2008). In severe chronic orchitis, seminiferous tubules may be obliterated by connective tissue (fibrosis) or granulomas (granulomatous orchitis) so that it is likely in these situations that this pathology could result in irreversible damage to the testis and sterility (Deif, 2011). Recent observations of orchitis in the koala related to natural chlamydial infection (Deif, 2011) suggest that the koala might potentially be an important animal model for studying this disease, and therefore descriptions of normal testicular histology will help in assessing the severity of pathology.
This study has revealed that the microanatomy and seminiferous epithelial cycle in the koala and wombat are essentially the same and confirms the close phylogenetic relationship between these species. There were, however, some interesting differences between the species, including the koala possessing (i) a greater proportion of Leydig cells, (ii) larger Sertoli cell nuclei, (iii) crystalloids in the Sertoli cell cytoplasm, (iv) a distinctive acrosomal granule during spermiogenesis and (v) a highly eosinophilic acrosome. Our observations will provide important insights for understanding the evolution of testis structure and function in future comparative studies with other mammals and become the basis for the assessment of the impact of testis histo-pathology associated with reproductive diseases such as chlamydiosis.
The authors would like to thank Dr Allan McKinnon, Mr Peter Theilemann, Mr Peter Brooks, Mrs Valerie Salmon, M. Darryl Sleep and M. Grant Fox for kindly helping us to collect the testes from the koalas and the SHN wombats. This study was partly supported by Australian Animals Care and Education Inc.