Abstract
To clarify the contribution of spontaneous or autolytic post-mortem changes to testis histopathology, the testes of adult rats were examined after animals were left at room temperature for 12, 24, 36, and 48 hours postmortem (n = 2 for all time points except 0 hours postmortem, where n = 3). A progressive decrease in testis weight and seminiferous tubule diameter was observed, as well as detachment of the seminiferous epithelium from the basement membrane. As early as 12 hours postmortem, there was observable clumping and margination of chromatin in Leydig cells, Sertoli cells, spermatogonia, spermatocytes, and step 7–10 spermatids; extensive disintegration of Sertoli cells and residual bodies by 24 hours postmortem; and TUNEL positivity of Leydig cells (by 36 hours postmortem) and step 19 spermatids (at 48 hours postmortem). These findings will aid in ensuring proficient histopathological analysis of testes in toxicity studies.
Introduction
Tissue from animals used in toxicity studies can be affected by fixation and processing protocols, potentially resulting in histopathological alterations that may be incorrectly attributed to toxicity. In toxicity studies where the endpoint is death, the tissue is usually not fixed immediately and may not be fixed for hours to days post-mortem. When such tissue is used from histopathological analysis, it is important to know if the observed changes are caused by delayed fixation or toxicity.
There are many factors that affect fixation and different fixatives are better at preserving tissue for different endpoints of analysis. Fetal rat testes fixed in Davidson’s fluid exhibit more sharply defined histological details than tissue fixed in Bouin’s fluid or 10% neutral buffered formalin while immunostaining was the most intense in formalin fixed tissue (Howroyd et al., 2005).
Penetration of the fixative and duration of fixation are important variables when fixing adult rat testes, which can be approximately one centimeter in diameter. Optimally, thin slices of tissue will be fixed, but the large adult rat testis must be fixed whole to preserve the seminiferous tubule structure within the tunica. It takes at least 24 hours for the fixative to penetrate the center of the testis, during which time metabolic processes may be ongoing. Inadequate fixation can cause shrinkage of the tissue during later processing (Bancroft et al., 2002).
Careful preservation of the testis tissue is important because multiple studies have shown that histopathological analysis is a sensitive method to detect testicular toxicity. There are a variety of histopathological markers that are characteristic of particular toxicants (Linder et al., 1992; Takayama et al., 1995). Examples include the absence or aberrant presence of certain germ cell types in a particular stage, such as the retention of step 19 spermatids in stages IX–XII in adult rats after exposure to deoxynivalenol for 28 days (Sprando et al., 2005). Sertoli cell toxicants such as 2,5-hexanedione and carbendazim cause sertoli cell vacuoles and germ cell sloughing after co-exposure (Markelewicz et al., 2004). Phthalates, which also target Sertoli cells, cause an increase in germ cell apoptosis in adult rats (as shown by TUNEL assay) or multinucleated gonocytes after neonatal exposure (Lee et al., 1997; Li et al., 2000). Toxicants that result in Leydig cell atrophy, such as ethane dimethysulfonate, manifest in the seminiferous tubules as stage specific loss of spermatocytes and spermatids (Creasy, 2003). In long term studies, such as exposing rats to 2,5-hexanedione in drinking water for 5 weeks, seminiferous tubule atrophy may be observed (Allard et al., 1996). Additionally, changes in absolute testis weight can indicate germ cell depletion, altered seminiferous tubule fluid secretion, or obstruction of outflow from the testis (Creasy, 2002, 2003).
We have examined the testes of adult rats 12, 24, 36, and 48 hours postmortem grossly and histopathologically. The testis weights, a sensitive, nonhistological marker of toxicity (Creasy, 2002), progressively decreased over time. Major observed changes in histomorphology were Sertoli cell atrophy and rearrangement of germ cell chromatin. Also, Leydig cells and step 19 spermatids stained positively using the terminal dUTP nick end labeling (TUNEL) assay. Herein is described the post-mortem rat testis phenotype in detail and these changes are compared to commonly observed toxic changes.
Materials and Methods
Animals
Fischer rats from Charles River Lab were housed at 70 ± 2°F on a 12 hour light/dark schedule with access to water and Purina Rodent Chow 5001 ad libitum. Animals were killed by isofluorine (Baxter, Deerfield, IL) overdose and the right testis was immediately removed after tying off the spermatic cord. The abdomen was closed and the animal was left at room temperature until the left testis was removed at 12, 24, 36, or 48 hours postmortem.
Sample Preparation
All reagents were from Sigma-Aldrich (St. Louis, MO), unless otherwise noted. Testes were fixed in Davidson’s modified media for 48 hours at room temperature and then transferred to 10% neutral buffered formalin and kept at 4°C, which is an acceptable alternative to fixation in Bouin’s fluid (as described by Latendresse et al., 2002). Horizontal cross-sections taken from the middle of each testis were embedded in 2-hydroxyethyl methacrylate (Technovit 7100, Kulzer, Germany) and 3 μm sections were stained with PAS/hematoxylin. For the TUNEL assay, tissue was embedded in paraffin and 5 μm sections were stained using an ApopTag Peroxidase In Situ Apoptosis Detection Kit (Cat. #: S7100, Chemicon International, Temecula, CA) per the manufacture’s protocol.
Sample Analysis
Samples were viewed on a Zeiss standard microscope and searched for signs of germ cell sloughing, multi-nucleated gonocytes, and any other notable morphologic changes. The minor axis of 50 randomly selected seminiferous tubules per cross-section (basement membrane to basement membrane) was measured using a stage micrometer (n = 3 at 0 hour, n = 2 at all other time points). The TUNEL assay was quantified by scoring all tubules in a cross-section as having zero, one-to-three, or greater than three TUNEL positive nuclei. Pictures were taken on a Zeiss Axiovert 35 microscope using a Spot RT camera (Diagnostic Instruments Inc., Sterling Heights, MI) and resized in Adobe Photoshop 6.0 (Adobe Systems Incorporated, San Jose, CA). Arrows, lettering, and bars on all figures were drawn using Canvas 8 (Deneba Systems, Inc., Miami, FL).
Results
Rat testis excised at 0, 12, 24, 36, and 48 hours after death showed progressive gross and histopathological changes. Testis weight decreased throughout the time course. Historical data from our lab showed that the left testis weighed 2.5% ± 2% more than the right, based on the testis weights of 30 animals. At 12 hours postmortem, the ratio between the left testis and the right testis (which was removed immediately after death) had decreased to 0.96. The left testis weighed 10% less than the right testis by 48 hours post-mortem (Figure 1A). Concomitantly, there was a continual decrease in the average seminiferous tubule diameter throughout the time course from 331 μm at 0 hours to 304 μm at 48 hours postmortem (Figure 1B). Ranges of measured seminiferous tubule diameters are available online (supplemental Table 1).
The histopathology will be discussed at each time point based on methacrylate embedded sections, as they show better resolution than paraffin sections. For each time point of analysis (0, 12, 24, 36, and 48 hours postmortem), the overall changes in the seminiferous tubule cross-sections observed at lower power by light microscopy will be described first (Figure 2), followed by changes in specific cell types observed at high power (Figure 3).
Histopathology of testes excised immediately after death showed seminiferous tubules with full cell populations adherent to the basement membrane with few gaps between the seminiferous tubules and interstitium (Figure 2A). There were 10–15 Sertoli cells per seminiferous tubule cross section with angular nuclei adjacent to the basement membrane and a moderate density of heterochromatin (Figure 3B). Type A spermatogonial nuclei were uniform and adjacent to the basement membrane (Figure 3C). Chromatin of spermatocytes looked uniform with punctuate condensations (Figure 3E). Residual bodies were condensed and small at stages VII/VIII (Figure 3I).
At 12 hours postmortem, the seminiferous tubule epithelium began to detach from the basement membrane; however cell-to-cell interactions appeared intact and cell membranes were discernable (Figure 2B). Sertoli cell nuclei were less angular and were no longer flattened against the basement membrane. The chromatin was partially condensed into nodules along the nuclear envelope (Figure 3B). At this time point, chromatin began to be rearranged in spermatogonia, spermatids, and Leydig cells. Type A Spermatogonial chromatin was more condensed (Figure 3C) and type B spermatogonial chromatin showed nuclear margination (Figure 3D). The nuclei of step 7–8 spermatids had marginated chromatin and cleared nucleoplasms (Figure 3F). Residual bodies were larger and looked more diffuse (Figure 3I). Leydig cell nuclei were less uniform, with chromatin condensation within the nucleus (Figure 3A).
At 24 hours, the post-mortem changes resembled those from 12 hours postmortem, but with a more pronounced rearrangement of the chromatin patterns (Figure 2C). Type A spermatogonia had more globular chromatin and were lifted away from the basement membrane (Figure 3C). Round spermatid nuclei had condensed chromatin along the nuclear envelope, with clearing out of the nucleoplasm (Figure 3F). Individual residual bodies were no longer distinct and often not discernable, appearing as a fuzzy haze (Figure 3I). Leydig cell nuclei were pyknotic and showed condensed, globular, fragmented chromatin (Figure 3A).
At 36 hours, the seminiferous tubule epithelium was intact, save for chromatin rearrangement, while individual cells in the interstitium were hard to distinguish (Figure 2D). The occasional Sertoli cell nuclei that could be identified had lifted off of the basement membrane and had lost all angularity (Figure 3B). Type A spermatogonial looked smaller and were detached from the basement membrane (Figure 3C). Type B spermatogonia had dense globules of chromatin and no apparent nuclear membrane (Figure 3D). Intermediate spermatogonia and preleptotene spermatocytes had very condensed chromatin and some nuclei contained globular chromatin (Figure 2D). The chromatin in pachytene spermatocytes looked more fragmented, a change that continued to progress throughout the time course (Figure 3E). Leydig cells were almost completely disintegrated (Figure 3A).
At the last time point, 48 hours postmortem, cells in the interstitium were indistinct, but the seminiferous tubule epithelium was remarkably intact. There was evidence of some breakdown of cell-cell junctions, but no evidence of luminal sloughing of the apical seminiferous tubule epithelium (Figure 2E). Sertoli cells had largely disappeared and only a few nuclei were discernable (Figure 3B). Type A spermatogonia collapsed into pyknotic foci by 48 hours (Figure 3C), and the nuclei of spermatocytes looked fragmented (Figure 3E). The chromatin in spermatogonia and spermatocytes showed the dense compact staining characteristic of apoptosis; however there was no increase in the incidence of TUNEL positive nuclei near the basement membrane. Early spermatids (step 1–6) remained intact, however step 7–10 spermatid nuclei had collapsed and appeared crumpled (Figure 3F). The elongated structure of step 11/12 spermatids became jagged and angular (Figure 3G). Leydig cell nuclei were condensed into pyknotic foci (Figure 3A).
TUNEL staining revealed that at 0 hours postmortem, 12% of the seminiferous tubules in a cross-section had one to three TUNEL positive spermatogonia or spermatocytes, indicating normal ongoing germ cell apoptosis. There was no evidence of Leydig cell apoptosis (Figure 4A). At 24 hours, an occasional Leydig cell was TUNEL positive, and at 36 hours many stained positively (Figure 4B). At 48 hours postmortem, some step 19 spermatids became diffuse and stained positively in the TUNEL assay (Figure 4C). And, 85% of the stage VII/VIII seminiferous tubule cross-sections contained TUNEL positive step 19 spermatids.
We did not observe any multi-nucleated gonocytes, germ cell sloughing, Sertoli cell vacuoles, or an increase in early germ cell apoptosis by TUNEL assay at any of the timepoints.
Discussion
Testicular toxicants can cause a variety of histopathologic changes after exposure. Testis tissue that was not fixed for 12–48 hours postmortem showed a distinct histopathological phenotype, as described in this paper. Postmortem autolytic changes can make it difficult to discern different cell types in a particular tissue, making histopathological analysis difficult, as is the case with rat carotid body Type I cells (Brown et al., 1996); however, some cell types are very resistant to autolysis, such at atrial specific granules in pigs, which appeared intact 24 hours postmortem (Castagnaro et al., 1992). This paper characterizes the postmortem testicular changes in rats to help distinguish between real toxicity and artifact of delayed tissue processing.
Delayed fixation of testes for histopathological analysis showed few of the typical markers for testicular toxicity, with the exception of decreased testis weight. The progressive decrease in the testis weight with time postmortem correlated with a decrease in seminiferous tubule diameter, and is explained by a failure of seminiferous tubule fluid formation (Creasy, 2002). The observed relatively rapid loss of Sertoli cell structural integrity would account for the decrease in seminiferous tubule fluid formation and secretion.
There was remarkable preservation of the cell-to-cell adhesions in the seminiferous tubules, although cell-to-basement membrane adhesions began to deteriorate within 12 hours of death. There was no evidence of sloughing of the apical seminiferous tubule epithelium into the lumen. Sloughing usually occurs when germ cell-to-Sertoli cell adhesions are lost (Creasy, 2002). Although the Sertoli cells rapidly deteriorated, some amount of the adhesion remained or other germ cell-to-germ cell adhesions kept the epithelium of the seminiferous tubule intact for at least two days after death. Internal cytoskeletal structures remained intact enough to prevent the formation of multinucleated gonocytes, which were not observed in any of the cross-sections. In addition, no Sertoli cell vacuoles were observed.
The TUNEL staining pattern was very interesting. In a healthy testis, the occasional spermatogonia and spermatocytes will undergo apoptosis as part of the germ cell culling process (Russell et al., 1990). Characteristically, in the normal testis, 12% of seminiferous tubules in a cross-section contain one or more apoptotic (TUNEL positive) cells. There was no difference in the occurrence of germ cell apoptosis throughout the postmortem time course, as indicated by TUNEL staining, but the observed Leydig cell apoptosis is unusual for most toxicologic studies. Perhaps there was a loss of a “survival” signal, leading to default apoptosis in Leydig cells. Apoptosis is an energy-dependent process. It is unknown how Leydig cells would have enough energy to undergo this process 36–48 hours after death.
There was also TUNEL staining of late spermatids at 48 hours post-mortem, particularly the decondensed step 19 spermatids. Our results are consistent with previous studies that have observed TUNEL staining in elongated spermatids and spermatozoa (Smith et al., 1998). Spermatid TUNEL positivity could be a result of endogenous DNA breaks or lesions caused by ROS, but is probably not a result of apoptosis. It is unlikely that such a differentiated germ cell has the machinery to undergo apoptosis. Endogenous DNA breaks are formed in elongating spermatids by topoisomerase II as part of the chromatin condensation and packaging process. Also, low levels of ROS have been shown to be present in all germ cells, particularly during the end stages of spermiogenesis. ROS levels increase in spermatozoa during development in the epididymis (Gomez et al., 1996). ROS can cause DNA damage, and 48 hours may be enough time for the damage to accumulate to levels detectable by TUNEL assay.
Step 19 spermatid chromatin is normally condensed into compact nuclei. The decondensed phenotype of some step 19 spermatids at 48 hours post mordem is an unusual phenomenon. As part of the condensation process of spermatid heads in spermiogenesis, histones are replaced with protamines, a cysteine rich histone variant (Govin et al., 2004). Cysteines are oxidized to disulfide bonds in the epididymis, but remain reduced before spermiation (Perreault et al., 1987). The absence of these bonds in elongated spermatids may allow for a reversal of the chromatin packing process which occurs during spermiogenesis, resulting in a diffuse, or decondensed, phenotype two days after death.
In conclusion, most cell types in the postmortem testis underwent major chromatin rearrangement while many structural (cytoskeletal and adhesion) components remained intact. The histopathological alterations described herein will be helpful in distinguishing toxicant injury from postmortem changes in the testis when test animals are not fixed immediately after death. Endpoints that we observed and are often observed in toxicologic studies include: decreased testis weight, decreased seminiferous tubule diameter, and Leydig cell TUNEL positivity (which can be induced after exposure to high levels of corticosterone) (Gao et al., 2003). Interestingly, histopathological changes
In toxicity studies, rats are routinely checked, and often refrigerated if an autopsy is not possible immediately after death. In any case, there may be some delay in fixation. These results highlight the importance of immediate fixation after the death of animals used for testicular toxicity studies, particularly when the most frequently observed histopathological markers are decreased testis weight, Leydig cell apoptosis, and chromatin rearrangement.
Footnotes
Acknowledgments
The project described was supported by grant number 5 P42 ES013660-02 from the National Institute of Environmental Health Sciences (NIEHS), NIH.