| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
research-article |
Pfizer Global Research and Development,3 Ann Arbor, Michigan 48105 Division of Animal Sciences4 and Department of Obstetrics & Gynecology,5 University of Missouri at Columbia, Columbia, Missouri 65211-5300
ABSTRACT
Theophylline (THP) and 1,3-dinitrobenzene (DNB) are thought to induce infertility by incapacitating the nurturing Sertoli cells and causing germ cell apoptosis in the testicular seminiferous epithelium, respectively. We hypothesized that THP and DNB exposure would alter the expression of the genes within the ubiquitin-proteasome pathway (UPP), implicated in spermatogenesis and epididymal sperm quality control. Rats were fed 0 or 8000 ppm of THP and necropsied on Days 18, 30, and 42 or administered 0, 2, or 6 mg/kg DNB via oral gavage and necropsied on Day 7. Tissues were collected from the testis and the caput, corpus, and cauda regions of the epididymis for transcriptional profiling by semiquantitative RT-PCR, real-time RT-PCR, and histopathology. Target UPP genes included those encoding for constitutive the 20S proteasomal core subunits Psmb1 (beta1), Psmb2 (beta2), and Psmb5 (beta5); the inducible 20S core subunits Psmb9 (LMP2), Psmb8 (LMP7), and Psmb10 (LMP10); and Ube1 (ubiquitin-activating enzyme E1), Ube2d3 (ubiquitin-conjugating enzyme E2), and Uchl1 (ubiquitin C-terminal hydrolase PGP9.5). Spermatozoa were collected from the cauda epididymis for analysis by light microscopy and flow cytometric evaluation of sperm surface ubiquitin. These data show that reprotoxic exposure alters the tissue-specific expression of UPP genes in the testis and epididymis, which may contribute to the aberrant spermatogenesis and epididymal processing of both normal and defective spermatozoa. Transcriptional profiling and flow cytometric analysis of the UPP thus captures the prodromal effects of reproductive toxicity not captured by conventional histology and functional cytology. Complementing seminal analysis with these measures may be useful in screening drug-induced toxicity or environmental infertility.
epididymis, gamete biology, 1,3-dinitrobenzene, proteasome, sperm maturation, testis, theophylline, toxicology, ubiquitin
The transition from a somatic cell-like phenotype of a spermatogonium to a unique, motile phenotype of a fully differentiated spermatozoon during mammalian spermatogenesis requires regulated proteolysis and organelle degradation [1, 2]. The ubiquitin-proteasome pathway (UPP) fulfills necessary requirements for the substrate specificity and developmental programming of proteolysis within the differentiating male germ cells [3]. Consequently, a set of genes that encode for testis-specific or alternatively spliced, unique ubiquitin-activating and -conjugating enzymes, along with polyubiquitin genes and those encoding for proteasomal subunits, is expressed during spermatogenesis [4–8]. Some of these genes, if disrupted, lead to altered or arrested spermatogenesis.
The UPP is the universal, yet tightly developmentally regulated and highly substrate-specific housekeeping system in charge of the removal of the aberrant, damaged, or outlived protein molecules. Varied roles of the UPP in cellular physiology and pathology include the endoplasmic reticulum-associated quality control during protein sorting into the secretory pathway (ERAD [9]), antigen presentation [10], cell cycle control [11], and programmed cell death or apoptosis. Of particular importance for the present study is the finding that malfunctions of the ubiquitin system, because of molecular misreading during the transcription of polyubiquitin genes [12], are observed in pathological conditions such as Alzheimer disease [13]. A reduced proteasomal proteolytic activity accompanies liver cirrhosis in ethanol-exposed laboratory rodents and patients with a history of alcohol abuse [14]. Within the male reproductive system (reviewed by Baska and Sutovsky [8]), the UPP contributes to a multifaceted gamete-quality control mechanism, providing the means of selective spermatogonial removal at the haploid phase of spermatogenesis [15], protein and organelle degradation during spermatid elongation/spermiogenesis [16–18], and the tagging of defective spermatozoa in the epididymis [15, 19, 20]. For these reasons, the expression of genes that encode for the individual components of the UPP may provide an efficient and highly sensitive indicator of pathology and toxic exposure in the brain, liver, and reproductive tissues. In the present study, we investigated whether a reprotoxic exposure changes the expression of UPP genes in the male reproductive system.
In the canonical ATP-dependent ubiquitin pathway [3], a molecule of ubiquitin is activated by ubiquitin-activating enzyme E1 and then passed onto a ubiquitin carrier, conjugating enzyme E2, and covalently ligated to an internal lysine residue or to an N-terminal residue of a substrate protein flanked with a substrate-specific E3-type ubiquitin ligase [3, 21]. Subsequently, long polyubiquitin chains are formed by the tandem ligation of additional ubiquitin molecules to the one bound to a substrate protein. Polyubiquitin chains mark the substrate for recognition by the 26S proteasome. Proteasome is a multisubunit protease typically composed of two 19S regulatory complexes or particles, capping both sides of a barrel-shaped 20S proteasomal core. The 19S particle is composed of at least 17 proteasomal regulatory proteins, including ATP-dependent and -independent subunits. Some of the 19S subunits are thought to have polyubiquitin-chain binding and deubiquitinating capabilities. The 20S core is composed of 14 subunits, seven 
-type subunits and seven ß-type subunits. Among the latter, the constitutive subunits PMSB1 (ß1), PMSB2 (ß2), and PMSB5 (ß5) are replaced by the inducible subunits PMSB9 (ß1i or LMP2), PMSB8 (ß2i or LMP7), and PMSB10 (ß5i or MECL-1/LMP10) in some cell types, including, but not limited to, professional antigen-presenting leukocytes [22]. Upon binding to the 19S complex, the polyubiquitin chain is removed from the substrate protein and recycled into reusable monoubiquitin molecules by deubiquitinating enzymes, which, in addition to the regeneration of monoubiquitin, fulfill important regulatory functions in ubiquitination and proteasomal degradation (reviewed by Nijman et al. [23] and Wing [7]). The substrate is then unfolded and threaded through the hollow 20S proteasomal core, where it is hydrolyzed into small peptides and released to be disassembled into single amino acids by cytosolic endopeptidases [22].
Theophylline (1,3-dimethylxanthine; THP) is a naturally occurring alkaloid present in tea [24]. As a nonselective PDE4 phosphodiesterase inhibitor, its beneficial pharmacological action is to relax smooth muscle cells in the bronchiolae of asthma patients. At high concentrations, THP is a documented male reprotoxic agent thought to induce infertility by incapacitating the nurturing Sertoli cells, thus resulting in the premature release of late differentiating spermatogenic cells, round spermatids [25]. This leads to the depletion of spermatids and mature spermatozoa from the adluminal compartment of the seminiferous epithelium, ultimately causing testicular atrophy. The effect of THP on the epididymal epithelium is poorly studied, although we have recently shown that an increased accumulation of ubiquitin in the secretory sites along epididymal epithelia of THP-exposed rats coincides with exposure [20]. The 1,3-dinitrobenzene (DNB) is known to induce germ cell apoptosis in the rat testis [26]. Given their respective reprotoxic effects, THP and DNB are model agents for developing novel methods for managing the safety risk associated with developing new pharmacological compounds. In the present study, we tested a hypothesis that the exposure of male rats to THP or DNB would alter the expression of UPP genes in the testis and epididymis of the exposed male animal. We demonstrate that the transcriptional and posttranslational analysis of the UPP component in the rat testis and epididymis can be used to gauge the male reprotoxic effects of THP and DNB with superior sensitivity.
Animals, Treatments, and Sample Isolation
All procedures and experimental protocols were reviewed and approved by the Animal Care and Use Committee at Groton Laboratories (Pfizer, Inc., New York, NY), and animals were housed in facilities accredited by the Association for the Assessment and Accreditation of Laboratory Animal Care. Groups of 10-wk-old male Crl:CD (SD)BR VAF/Plus rats (~175 g) were housed singly and given THP daily at 0 ppm (five animals per group) or 8000 ppm (five animals per group) in feed for 18, 30, or 42 consecutive days (T1633; Sigma, St. Louis, MO). DNB powder was weighted and diluted in acetone and then further diluted in olive oil. Each oral gavage administration had equivalent amounts of acetone (2 µl of acetone for every 1 ml of olive oil; total dose volume limit = 2 ml/kg) and was dosed on the basis of body weight to result in the administration of 0, 2, and 6 mg/kg DNB for 7 days. All rats were observed at least two times daily in their cages for signs of overt toxicity and for any changes in appearance or behavior. Prestudy body weights were collected 1 day prior to the initiation of dosing. Thereafter, body weights were determined weekly during the study period. Food consumption was measured weekly during the study period. The start date of each group was coordinated to accommodate a single necropsy and tissue collection procedure. The left and right testis and epididymis were collected, weighed, and processed for molecular and microscopic end points.
The whole testis and epididymis were released from the tunica albuginea, the testis was nicked with a scalpel, and the whole tissue was submerged for overnight incubation in ultrapure 4% paraformaldehyde and postfixed in Bouin fixative. After fixation and thorough washing, tissues were embedded in paraffin by conventional procedures and positioned within a paraffin block so that the long axis of the testis and all three major epididymal compartments could be sectioned at the same time. Sections were cut on a microtome, mounted on microscopy slides, dewaxed by xylene and ethanol changes, stained by hematoxylin-eosin, and examined and photographed under transmitted visible light illumination with a Nikon Eclipse 800 microscope.
Isolated cauda epididymal spermatozoa were released into a Hepes-buffered Sperm-TL medium by careful nicking of the cauda epididymal tissue and collected by centrifugation. Care was taken to eliminate pieces of tissue and to prevent contamination with epididymal epithelial cells, blood cells, and stromal tissue. Purity of sperm preparation was checked under a light microscope. Spermatozoa were fixed in 2% formaldehyde in PBS, washed in PBS, blocked by a 40-min incubation with 5% normal goat serum (Sigma), and incubated overnight with mouse monoclonal antibody KM691 (diluted 1:100; Kamiya Biomedical Company, Seattle, WA) and then incubated for 1 h with goat anti-mouse immunoglobulin G (IgG)-fluorescein isothiocyanate (FITC). Negative controls were generated by the omission of anti-ubiquitin antibody or by its immunosaturation with purified bovine erythrocyte ubiquitin (Sigma). The efficiency of immunostaining was checked by epifluorescence microscopy (detailed below) by mounting small aliquots (2 µl of sperm pellet from each processed sample) on microscopy slides. Cells were washed and resuspended in 500 µl of filtered PBS and transferred into Falcon flow cytometry tubes equipped with a 50-µm cell strainer to remove clumped cells.
Flow cytometry was performed by a procedure developed previously for the measurement of sperm ubiquitin in human semen samples [27]. Samples were analyzed with FACS Scan Analyzers (Becton Dickinson). Relative levels of ubiquitin-induced fluorescence (ubiquitin median) in 10 000 individual cells per sample were recorded. Histograms of relative fluorescence (see Fig. 3) were generated, and the arbitrary thresholds were set on the histograms to gate the small cellular fragments and debris that are ubiquitinated but too small to generate highly ubiquitinated measurement (see marker M1 in Fig. 3), cells of normal size typical of a morphologically normal rat spermatozoa with background levels of surface ubiquitin (marker M2). and defective cells with high surface ubiquitination-generating high readings in flow cytometric analysis (marker M3). Cell subpopulations within those three markers were analyzed by recording their ubiqitin medians and determining what percentage of all measured cells they represented. The information collected on sperm ubiquitin was supplemented with information concerning the size of cells within the screened samples. For this purpose, the diagrams of the visible light scatter were generated by combining the forward and side scatter of the cells passing through the flow cytometer (see Fig. 3). The relative ubiquitin median values of low-fluorescent cells (median M2; x-axis) and high-fluorescent cells (median M3; y-axis) were arranged in scatter plots with regression lines color coded according to treatment (see Fig. 4C). Numeric data (ubiqitin medians and percent cells within each of the three markers) were entered into Microsoft Excel tables and analyzed by statistical analysis tools (e.g., MS Excel and the SAS 8.2 statistical package). The P-values were calculated for ubiquitin medians and percentages of ubiquitin-positive cells in control and toxicant-exposed animals at each time point. Descriptive statistical tools were used to calculate the means and standard errors of the mean.
|
|
Epifluorescence and DIC Microscopy
Fixed epididymal sperm samples and sperm samples processed for flow cytometry with anti-ubiquitin and secondary antibodies were mounted on microscopy slides in VectaShield mounting medium (Vector Laboratories) and examined by differential interference contrast (DIC) optics or epifluorescence illumination in a Nikon Eclipse 800 microscope equipped with infinity-corrected lenses (10x, 20x, 40x, 60x, and 100x primary lens magnification). Slides were photographed with a CoolSnap HQ CCD camera (Roper Scientific, Tucson, AZ) operated by MetaMorph software with appropriate epifluorescence filters. Images were edited by Adobe Photoshop 5.5 (Adobe Systems, Inc., San Jose, CA).
Histology and Immunohistochemistry
Antibodies against the following UPP components were purchased from Biomol (Plymouth Meeting, PA): UBE1 (ubiquitin-activating enzyme E1), UBE2D3 (ubiquitin-conjugating enzyme E2—25 kDa, 59% sequence identity to yeast UBC4 and UBC5 gene products [28]), and 20S proteasomal core subunits PMSB1 (ß1), PMSB2 (ß2), PMSB5 (ß5), PMSB9 (LMP2), PMSB8 (LMP7), PMSB10 (LMP10/MECL-1), PMSA1–7 (1–7), and UCHL1 (ubiquitin C-terminal hydrolase L1/PGP9.5). From MBL (Nagoya, Japan) were purchased anti-bovine erythrocyte ubiquitin MK12–3 (mouse IgG); from Kamiya were purchased anti-human recombinant ubiquitin KM691 (mouse IgM) and anti-ubiquitin-ubiquitin-isopeptide bond domain MC034 (mouse IgM); and from Zymed Inc. (San Francisco, CA) were purchased anti-GFP, anti-His-tag, and all secondary fluorescent antibodies. Testicular and epididymal tissues were immediately fixed in 4% paraformaldehyde and embedded in paraffin by conventional techniques. Tissue sections were mounted on microscopy slides, deembedded [19], and incubated sequentially with the first and second antibodies listed above. Slides were examined under epifluorescence illumination as described above. For conventional histological analysis, deparaffinized tissue sections were stained with hematoxylin-eosin and observed under transmitted illumination.
Total RNA was isolated from 10–30 mg of rat testicular tissue and caput, corpus, and cauda epididymal tissue and rat liver (used as the calibrator) with the Qiagen RNeasy Mini RNA isolation kit (Qiagen, Valencia, CA). The digestion of DNA was performed with the RNase-Free DNase Set from Qiagen according to the manufacturer's specifications. The RNA samples were quantified with a NanoDrop spectrophotometer (NanoDrop, Wilmington, DE) at 260 nm. All RNA samples isolated had an absorbance 260:280 ratio of 1.9–2.2. Random RNA samples were also examined for their integrity by 1% agarose gel electrophoresis stained with 0.5 µg/ml ethidium bromide. The RNAs were immediately stored as aliquots at –80°C after isolation.
Semiquantitative Relative RT-PCR and Multiplex and Singleplex PCR
A two-step RT-PCR was used. An oligo-dT primer (15n labeled; Promega, Madison, WI) was used in the first step of cDNA synthesis. A minimum mixture of total RNA (100 ng) and RNase-free distilled H2O was preheated at 65°C for 5 min and then chilled on ice rapidly for 10 min. The Qiagen Sensiscript RT-PCR kit was used for the RT-PCR reaction, and 20 µl of RT mix was incubated at 37°C for 90 min. The cDNA was stored at –20°C. For singleplex PCR, the samples, along with a calibrator sample (liver), were processed with target and reference gene primers in the same PCR run. All PCR products from both the target and reference genes of the whole sample set were resolved on the same agarose gel for photography and densitometry. For more accurate results with semiquantitative multiplex PCR, amplifying the target gene(s) and reference gene in the same tube was used for most of experiments to reduce the pipetting errors. Depending on the amplicon lengths of the different primer sets, triplex, duplex, or singleplex PCR reactions were chosen for different genes. PCR reactions were performed with a Promega PCR kit, an Eppendorf PCR kit (Eppendorf, Westbury, NY), or a HotMaster Hot Start PCR kit from Eppendorf on an Eppendorf Mastercycler PCR machine. Gene-specific primers were designed to flank intron regions in order to eliminate false-positive PCR fragments generated from genomic DNA. Table 1 shows the primer sequences and the GenBank accession numbers of the genes (mRNA sequences). Actb (ß-actin) was the endogenous control (reference gene) for the normalization of the quantification of the target mRNA to compensate for differences in the amount of total RNA added to each reaction. Validation experiments for the multiplex PCR were performed prior to the actual PCR trial for all primers. The appropriate number of PCR cycles was selected so that the amplification products of both the target and reference genes were clearly visible on agarose gel and could be quantified. Regularly, the amplification was in its exponential phase but was never allowed to reach its plateau. The concentrations of primers used in each reaction were tested and chosen to be the ones that did not compete with each other. A total of 0.7–1.0 µl of the RT-PCR products was used for PCR. Varied PCR conditions and primer concentrations were used for different genes. The PCR products were resolved by electrophoresis on a 2% agarose gel stained poststain with ethidium bromide. Gels were photographed on top of an ultraviolet illuminator. The PCR products were confirmed as target sequences by sequencing with the Applied Biosystems 3730 DNA Analyzer (Applied Biosystems, Foster City, CA) by means of Applied Biosystems Prism BigDye Terminator cycle sequencing chemistry at the DNA core facility of the University of Missouri at Columbia. For semiquantitative RT-PCR, densitometry was performed with the Kodak digital imaging system-Electrophoresis Documentation and Analysis System 290 and Kodak 1-D Image Analysis Software (Kodak, New Haven, CT). The data were organized, and the ratios of the target gene and reference gene were calculated. The final ratios, reflective of the relative gene expression level, were obtained by dividing the ratio of the target and reference genes of the calibrator. These values were analyzed with the SAS 8.2 program by the ANOVA General Linear Models (GLM) procedure and mixed procedures.
|
Two-step RT-PCR, multiplex PCR, TaqMan probe-based chemistry, and the Comparative CT Method were used. TaqMan probes and primers (Table 2) were designed with Primer Express (Applied Biosystems). In all cases, the target amplicon spans an intron. A RealMasterMix Probe (Rox) was used as the real-time PCR reagent. On the basis of semiquantitative RT-PCR data from the present study and from a separate pilot study on the reproductive toxicity of DNB (unpublished results), two marker genes were selected: Ube2d3 and Psmb1. Complementary DNAs synthesized from the RT-PCR were diluted and requantified by the NanoDrop spectrophotometer, and optimal quantities for RT-PCR were determined from validation experiments. Standard curves for target and reference genes for both singleplex and multiplex methods were run in a same plate and studied by TaqMan 7500 software to ensure that the efficiencies of target and reference gene expression levels were approximately equal by the template quantity and primer concentrations used in both methods. In the actual experiments, 7 ng of cDNA per reaction for Ube2d3 and 25 ng of cDNA per reaction for Psmb1 were used. Each RT-PCR plate used the same calibrator as the basis for comparable results. Each tube contained the primers and probes of both the target gene and the endogenous control (Actb).
|
Statistical Analysis of Gene Expression Data
For semiquantitative relative RT-PCR, the data from densitometry were densities of bands on agarose gel representing the target gene or the reference gene in each sample. The ratios of the target gene and the reference gene were calculated. The final ratios, reflective of the relative gene expression level, were obtained by dividing the ratio of the target gene and reference gene of the calibrator. All genes used the same calibrator. All values were analyzed with the SAS 8.2 program by the ANOVA GLM procedure and mixed procedures.
For RT relative quantitative PCR with the Comparative CT Method, the amount of target relative expression, referred to as the relative quantity (RQ), was given by 2
CT, which was automatically calculated after each run of an RT PCR experiment, by the Multi-plate Study method analyzed by 7500 TaqMan Relative Quantification Study software from Applied Biosystems. The acquired data were statistically analyzed by the ANOVA GLM procedure and mixed procedures in the SAS 8.2 program.
No changes in the histology of testicular and epididymal tissues were observed in control rats at any time point (Fig. 1, A–C) or in the THP-exposed rats on Day 18 (data not shown; Supplemental Table is available online at www.biolreprod.org). Histological changes of seminiferous epithelia were noted in three of six exposed animals on Day 30 and in five of six rats on Day 42 (Supplemental Table 1; available online at www.biolreprod.org). Sloughing of spermatids in the form of multinucleated bodies (Fig. 1, D–I) was the most frequent testicular anomaly and was consistent with the proposed mechanism of the THP action on Sertoli cells. This translated into the depletion of spermatogonia and spermatids from seminiferous epithelium and a reduction in the number of fully differentiated spermatozoa in the testis and epididymis. Intraluminal sperm granulomas and degenerating sloughed spermatids were observed in the epididymis of some but not all exposed animals with visible seminiferous tubule damage (Fig. 1F).
|
Findings related to DNB treatment (Supplemental Table 2; available online at www.biolreprod.org) included degeneration of round spermatids in the testis of two of the six rats given 6 mg/kg DNB and the increased presence of cellular debris and/or epithelial cell vacuolization in four of the six rats given 6 mg/kg DNB. Two such animals had anomalies in the testis, and two had anomalies in the epididymis. Only one of the eight rats given 2 mg/kg DNB and one of the seven control rats showed similar anomalies. The most frequent pathologies observed in the DNB-exposed rats (Fig. 1, J–O) were the degeneration and loss of spermatocytes and round spermatids, sometimes multifocally distributed, which affected multiple stages. Formation of multinucleated spermatids was also observed (Fig. 1M). Complete sperm depletion was observed in the epididymal tubule lumen of several rats exposed to 6 mg/kg DNB. Other findings, including the presence of some cellular debris in the cauda epididymal lumen, were considered unrelated to treatment, as they were found in both the control and exposed rats.
Sperm head-tail detachment and sperm tail fragmentation were the most commonly observed abnormalities by DIC microcopy in the cauda epididymal spermatozoa of the THP-exposed rats on Days 30 and 42 (Fig. 2, A–F). Sperm fragmentation was not commonly observed after DNB treatment (Fig. 2, G–I). Common defects in the DNB set included sperm head malformations (Fig. 2G) and a pattern of longitudinal bundling of seemingly morphologically normal spermatozoa (Fig. 2H).
|
Samples for flow cytometric sperm evaluation were processed with monoclonal anti-ubiquitin antibody KM691, causing intense fluorescence of the surface (no permeabilization was used during processing) of the defective spermatozoa from both the control and the THP- and DNB-exposed rats (see Fig. 4D for an example). It was expected that a toxic exposure would increase the surface ubiquitination of defective cauda epididymal spermatozoa because of a previously established association between sperm abnormalities and sperm surface ubiquitination [19].
Histograms of ubiquitin-induced fluorescence were generated for all samples on the basis of 10 000 measurements per sample and were divided into three marker areas (M1–M3). Marker M1 included low-ubiquitin cells and fragments, marker M2 was predominantly normal spermatozoa of typical size and shape, and marker M3 represented defective spermatozoa with high ubiquitin levels. Scatter diagrams of the visible light (to the right from the histogram) showed a tight focus of cells corresponding to the spermatozoa of normal size in the control rats in the THP trial at all time points (Fig. 3A). Many spermatozoa of a THP-exposed rat on Day 18 were still within marker M2; however, the peak was wider and shifted to the right toward the higher ubiquitin level (Fig. 3B), where a distinct second peak of highly fluorescent, high-ubiquitin cells was noted within marker M3.
On Day 30 (Fig. 3C) and Day 42 (Fig. 3D), the THP-exposed rat sperm samples displayed an increased amount of sperm fragments with lesser relative fluorescence and a reduced contribution of normal spermatozoa within marker M2. Note the diminishing cell size in the light scatter analysis, appearing as an increased number of dots or cells in the lower left corner, due to sperm fragmentation and decapitation. A shift toward M1 reflects abundant sperm sample decapitation. A flow cytometric histogram of a blank, negative control sample processed with a second antibody only is included for reference (Fig. 3E). The average median values of ubiquitin-induced fluorescence, as well as the percentage of presumed normal spermatozoa within marker M2, were significantly reduced on both Days 30 and 42 (P = 0.02; Fig. 3F).
To evaluate the effect of DNB on mature sperm, flow cytometric analysis was performed in sperm samples from the control and DNB-exposed rats (Fig. 4, A–D). Each line represented peaks of a 5000-measured-cell histogram. The exposure to 6 mg/kg DNB significantly reduced the levels of ubiquitin in the sperm sample, causing the shift of the histogram peak to the left, toward low-fluorescent cells (Fig. 4, A and B). This was not caused by sperm fragmentation, as seen after THP exposure. Compared to the control rats, the average ubiquitin median levels (Fig. 4C) were increased slightly in the 2-mg/kg DNB group (P = 0.46) and significantly (P = 0.04) in the 6-mg/kg group.
Compounded transcriptional profiles of all evaluated genes in all four segments of the male reproductive system (the testis and the caput, corpus, and cauda epididymis) showed a significant reduction in the expression of all three examined constitutive 20S core subunits after THP treatment (Supplemental Fig. 1; available online at www.biolreprod.org). Expression of Psmb1 and Psmb2 genes encoding for the constitutive 20S core subunits PSMB1 (ß1) and PSMB2 (ß2) was reduced in both the DNB-exposed animals (Supplemental Fig. 1, A and B, and Supplemental Table 3A; available online at www.biolreprod.org) and the THP-exposed animals (Supplemental Fig. 1C and Supplemental Table 3B; available online at www.biolreprod.org). Expression of Psmb5 encoding the PSMB5 (ß5) subunit was increased in the DNB-exposed animals but decreased in the THP-exposed animals. There was a significant reduction in the expression of Ube2d3 after DNB treatment (Supplemental Table 3A; available online at www.biolreprod.org).
In segment analysis (Fig. 5 and Supplemental Table 3, A and B; available online at www.biolreprod.org), there was a significantly (P < 0.01) reduced expression of all three constitutive proteasomal subunits in the corpus epididymis after THP treatment. Also, Psmb8 showed a significant reduction in the testis and caput epididymis and a significant increase in the cauda epididymis. Exposure to DNB had an even more profound effect on UPP gene expression. A highly significant reduction in the expression of Psmb1 in all four segments at both 2- and 6-mg/kg DNB exposure levels was observed (Fig. 6 and Supplemental Table 3A; available online at www.biolreprod.org). Furthermore, the expression of Psmb2 was significantly reduced in all four segments after the 6-mg/kg DNB exposure, and the expression of Ube1 and Ube2d3 was altered in the caput, corpus, and cauda of some of the DNB exposure categories.
|
|
On the basis of the compounded relative gene expression in all animals, wherein each gene was compared against Actb, the highest induction of expression of the UPP genes was observed in the caput epididymis and testis, followed by the corpus epididymis in both trials (Fig. 6, A and B, and Supplemental Fig. 2, A and B; available online at www.biolreprod.org). Overall, the lowest levels of UPP gene expression relative to actin gene expression were noted in the cauda epididymis (Fig. 6, A and B). The mRNAs transcribed from Ube1, Ube2d3, and Uchl1 genes were most prominent in the testis and caput epididymis. The mRNAs encoding for constitutive 20S proteasomal core subunits were most abundant in the corpus epididymis (Fig. 6, A and B).
In addition to semiquantitative RT-PCR analysis, the relative expression of the select UPP genes most prominently changed by reprotoxic exposure was examined by real-time RT-PCR (RT-RT). The expression of Psmb1 relative to Actb after THP exposure followed the same pattern (testis > corpus > caput > cauda) by both methods (Fig. 7 and Supplemental Table 4A; available online at www.biolreprod.org), although the RT-RT method captured finer differences between segments than the semiquantitative method. Similarly, the relative expression of Psmb1 after DNB exposure was the same in the semiquantitative and RT-RT methods, except for the increased expression shown by the semiquantitative method in the cauda epididymis. Similar to our semiquantitative studies, the RT-RT analysis of Psmb1 showed a significantly reduced expression in the THP-treated animals when compared with the control animals (Supplemental Table 4B, line 1). While semiquantitative PCR did not show a significant difference among treatment days (Supplemental Table 4B, line 2), RT-RT showed that the mean RQ value of 42 days was significantly lower than in the 18- and 30-day animals, while the 18- and 30-day groups were not significantly different from each other. When only the treated animals were included in the analysis (Supplemental Table 4B, line 3), RT-RT showed a significant reduction of Psmb1 expression by the length of treatment; the semiquantitative PCR, a less sensitive method, did not capture such an outcome. Similarly, the RT-RT method showed significant differences for feed and treatment between the control and DNB-exposed animals (Supplemental Table 4B, lines 4 and 5). Numerically, the expression of Ube2d3 was reduced by THP in the testis and cauda epididymis and increased in the caput and corpus epididymis by the RT-RT method (Supplemental Table 4C).
|
The RT-RT PCR method confirmed that the expression of Ube2d3 was reduced by THP exposure in some of the examined compartments (Supplemental Table 4D). Similarly, the expression of Ube2d3 was reduced in the testis and cauda epididymis of the 2- and 6-mg/kg DNB-exposed animals (Fig. 7).
Epifluorescence Localization of the UPP Gene Products in the Testis and Epididymis of Control and Exposed Rats
Detailed immunohistochemical analyses were conducted in the DNB trial to demonstrate the distribution of the target gene products throughout individual segments of the rat reproductive system. The DNB sample set was chosen because of its direct action on male germ cells, as opposed to the indirect effect of THP, mediated by Sertoli cell toxicity. The UBE1 protein was localized mainly in the cytoplasmic lobes of elongating spermatids and residual bodies rejected by fully differentiated spermatozoa in the control testis (Fig. 8A). In the DNB testis, the multinucleated spermatids showed a distinct accumulation of UBE1 in the perinuclear region (Fig. 8A'), although some localization to residual bodies was still observed (Fig. 8A', insert). Similar to UBE1, UBE2 was localized in the residual body and cytoplasmic lobe (Fig. 8B) but also within the developing acrosome of round and elongating spermatids (Fig. 8B, top insert) and in the chromatin of the pachytene and dividing spermatocytes (Fig. 8B, bottom insert). In the DNB testis, UBE2 and ubiquitin showed a distinct accumulation in the multinucleated spermatids (Fig. 8B'). While the acrosomal localization was still observed in the spermatids, there was no labeling in the pachytene and dividing spermatocytes (Fig. 8B', insert). UCHL1 was concentrated in the residual bodies and within the developing acrosomal caps of the round spermatids in the control testis (Fig. 8C). While the acrosomal cap localization was retained in the DNB-treated animals (Fig. 8C', insert), these animals also showed a distinctly diffuse labeling in the cytoplasm and nuclei of most cell types within the seminiferous tubules (Fig. 8C'). Except for the absence of nuclear/chromosome labeling in the pachytene and dividing spermatocytes, the localization of inducible 20S proteasomal core subunits within the seminiferous tubules, as exemplified by PSMB9 (Fig. 8, D and D'), followed the localization patterns of the ubiquitin and UPP enzymes described above. In the control testis, the distribution of PSMB9 showed a diffuse pattern in the nuclei and cytoplasm of all types of spermatogenic cells. Distinct accumulation was observed in the cytoplasmic lobe of elongated spermatids, in the residual bodies (Fig. 8D), and in the acrosomal cap of the round and elongated spermatids (Fig. 8D, insert). Acrosomal and residual body localization of PSMB9 (LMP2) was also seen in the DNB testis (Fig. 8D', top insert), while there seemed to be an increased diffuse labeling in the cytoplasm and nuclei of the degenerated, single, and multinucleated spermatids. In addition, the accumulation of PSMB9 was observed within the heterochromatin body inside the nuclei of the secondary spermatocytes (Fig. 8D', bottom insert).
|
In the control epididymis, the UBE1 protein showed an even, diffuse distribution in the cytoplasm of all epithelial cells, while the proteasomes labeled by a monoclonal antibody against 20S core subunits 1–7 appeared to be sequestered mainly in the apical compartment (Fig. 9A). In the DNB epididymis, a distinct accumulation of UBE1 and proteasomes was observed within the cytoplasm of the clear cells of the epididymal epithelia (Fig. 9A'), while the accumulation of UBE1 was also observed in the perinuclear region of the principal cells (Fig. 9A', insert). UBE2 showed a distinct perinuclear focus in the principal cells of the control epididymis, reminiscent of endoplasmic reticulum localization pattern (Fig. 9B). Some of the principal cells showed a nuclear localization of the ubiquitinated protein species recognized by antibody MK12–3 (Fig. 9B). The endoplasmic reticulum-like localization pattern was absent from the DNB epididymis (Fig. 9B'), while the nuclear accumulation of anti-ubiquitin immunoreactive proteins was more pronounced (Fig. 9B', insert). UCHL1 showed similar patterns in both the control (Fig. 9C) and DNP-exposed epididymis (Fig. 9C'). In both types of tissue, a diffuse labeling was present in the cytosol of the principal cells, while a concentrated labeling that coincided with that of the ubiquitinated proteins was seen in the clear cells. A similar distribution pattern was followed by the proteasomes in the principal and clear cells of the control (PSMA/PMSB; Fig. 9D) and DNB-exposed (Fig. 9D') epididymis. In this example, proteasomes were identified by a polyclonal antibody that recognized multiple 20S proteasomal and ß subunits. Noteworthy was the accumulation of ubiquitin and/or ubiquitinated substrates on the apical surface of the epididymal epithelium, recognized by anti-ubiquitin antibody KM691 [20]. Negative control slides did not show any of the patterns described above for the testis or epididymis in the control or DNB-exposed animals.
|
The present study demonstrates the reproductive toxic and spermatotoxic effects of THP and DNB by conventional means (histology and DIC microcopy) and by novel approaches, including flow cytometric analysis of isolated epididymal spermatozoa and transcriptional profiling of select gene products within the proteolytic UPP. Our goal was to determine if transcriptional analysis and flow cytometry could capture minor changes in the male reproductive system of rats exposed to reprotoxic chemicals. We were particularly interested in examining threshold toxic exposures at which a conventional histological examination did not reveal any pathology. No damage to testicular or epididymal tissues was identified by histology prior to Day 30 of THP exposure (Supplemental Table 1). Similarly, only some of the 2-mg/kg DNB-exposed animals showed testicular or epididymal histopathology (Supplemental Table 2).
Contrary to ambiguous histopathological findings, the THP-induced sperm damage and breakdown were manifested by altered flow cytometric median values of ubiquitin on Days 30 and 42 (Fig. 3). While the difference on Day 18 was not significant (P = 0.79), several treated samples already exhibited a flattened, shifted flow cytometry histogram, suggestive of sperm fragmentation, and a subpopulation of highly fluorescent cells, suggestive of increased sperm ubiquitination (see Fig. 3B). Similarly, several components of the UPP showed a statistically significant expression change on Day 18 in expression analysis. On the basis of our studies of sperm ubiquitin in infertile men [29], we could expect an overall reduction in the percentage of presumably normal spermatozoa with background levels of surface ubiquitin (see Fig. 3, A and B) and predict an overall increase of the sperm ubiquitin levels in THP- and DNB-exposed rats. While we indeed observed a significant reduction in the normal sperm percentage after THP exposure, the overall sperm ubiquitin values were actually lower in the exposed animals than in the control animals on Days 30 and 42. Sperm fragmentation and widespread tubular disease with THP exposure could explain this seeming paradox. Although the detached heads and tail fragments did show intense ubiquitin immunostaining in epifluorescence microscopy, their small size reduced the individual cell reading in the flow cytometer to a level that was only a small fragment of the reading obtained for intact, whole spermatozoa, regardless of their ubiquitin expression. While this shortcoming could be compensated for by dual DNA ubiquitin flow cytometry, we found the current analysis, which was based on the distribution within markers M1–M3, sufficiently informative and reflective of THP exposure.
Because of the previously established association between sperm abnormalities and sperm surface ubiquitination [19], we expected that a toxic exposure to DNB in the absence of sperm fragmentation (as observed in THP rats) would increase the surface ubiquitination of defective cauda epididymal spermatozoa. However, there was actually a decrease (see Fig. 4). This could be because of the reduced expression of ubiquitin-conjugating enzymes within the epididymal epithelia and epididymal fluid, among which the reduction in the expression of Ube2d3 was impaired in a highly significant manner (P = 0.003). Accordingly, we have observed the perinuclear accumulation of UBE2, perhaps reflecting the association of UBE2 with the ubiquitin-dependent, endoplasmic reticulum-associated protein-quality control mechanism (ERAD), in the principal cells of the control epithelia (Fig. 9B) but not in the DNB-exposed epithelia. It could also be argued that the apocrine-secretory capabilities of the epididymal epithelium were impaired, resulting in a reduced content of UBE2 and other ubiquitin system enzymes in the epididymal fluid. We did not observe any major morphological changes in the epithelia, and the accumulation of presumed secretory ubiquitin on the apical secretory sites of the DNB epididymis appeared comparable to the control rats.
On the basis of compounded transcriptional profiling of all tissues by semiquantitative RT-PCR, the most pronounced THP- and DNB-induced change overall was a significant reduction in the expression of constitutive 20S proteasomal core subunit genes Psmb1, Psmb2, and Psmb5 across the male reproductive system. This is in agreement with the reduced expression of proteasomal subunits in the epididymis of aging rats [30]. Also significant are the observations linking a reduced proteasomal subunit expression or an impaired proteasomal proteolytic activity in various human pathologies including, but not limited to, liver cirrhosis [14] and neuronal Alzheimer disease [13]. Alzheimer disease has been linked to the aberrant transcription of the ubiquitin-B (UB-B) gene, caused by a +1 frame-shift during transcription. This misreading results in the translation of the dysfunctional UB-B+1 protein with the elongated C-terminus not capable of ligation to a substrate protein [12]. Consequently, the amyloid protein within neurons is not properly ubiquitinated and degraded by proteasomes, causing the formation of amyloid plaques in Alzheimer disease-affected brain tissue. Intriguingly, the UB-B+1 frame shift product has been found in the aging human epididymis [12].
In segment analysis, the reduction in 20S core subunit expression was most significant in the corpus epididymis, which is in good agreement with our immunohistochemical data showing that proteasome-rich, endocytotic clear cells are particularly abundant within this epididymal segment [20]. Clear cells have been implicated in the removal of intraluminal debris, derived from the breakdown of rejected sperm cytoplasmic droplets [31]. The mechanism of epididymal sperm maturation and disposal of defective spermatozoa is not completely understood and may involve defective sperm marking by the UPP [18, 19, 32] as well as the coating of defective spermatozoa by a procoagulatory glycoprotein, HEP64/fgl2 [33]. We have shown that defective spermatozoa become ubiquitinated in the caput epididymis [19], presumably by the enzymatic ubiquitination machinery residing within epididymal fluid [8]. The percentage of defective spermatozoa is reduced after passage through the corpus epididymis [19, 34, 35]. Since proteasomes have also been detected in the epididymal fluid [36], it is possible that defective spermatozoa are partially degraded intraluminally in the caput and corpus epididymis and that the resident clear cells of the corpus epididymis take up and degrade the ubiquitinated proteins released from moribund spermatozoa. Such a mechanism of defective sperm ubiquitination in the caput and removal in the corpus epididymis is consistent with the transcriptional profiles of the individual epididymal segment from the present study (see Fig. 7). The mRNAs of Ube1, Ube2d3, and Uchl1 are most prominent in the caput epididymis, whereas the constitutive proteasomal core subunits prevail in the corpus epididymis. THP- and DNB-reduced expression of Psmb1, Psmb2, and Psmb5—the ones with actual protease activities—within the caput and corpus epididymis could alter the epididymal fluid composition and disposal of defective spermatozoa in several ways: by reducing the population of the assembled proteasomes in clear cells, by diminishing the release of assembled proteasomes from secretory epithelial cells into epididymal fluid, by saturating the proteolytic capacity of the remaining proteasomes, or by directly reducing the proteasomal proteolytic activities of assembled proteasomes in the epididymal fluid and/or the clear cell cytoplasm.
Inducible proteasomal core subunits PSMB9, PMSB8, and PSMB10 replace their constitutive counterparts PMSB5, PMSB1, and PMSB2 in the professional antigen-presenting cells [22] but are also present in other cell types, such as eye lens cells [37] and the sperm acrosome [29]. The expression of inducible subunit Pmsb8 as measured by semiquantitative RT was significantly reduced within the testis and corpus epididymis of the THP-exposed rats, and yet overall, the changes in the expression of the other two inducible subunits were not significant. It is possible that the inducible subunit genes are less sensitive to THP. The loss of inducible 20S subunit transcripts could also be masked by an increased infiltration of white blood cells expressing inducible subunits in the testicular or epididymal tissues of exposed animals; such a white blood cell infiltration, however, was not detected by our histological analysis. Only Psmb1, Psmb2, and Psmb5 and the genes encoding for the constitutive 20S proteasomal core subunits and the Ube2d3 gene showed significant expression level changes between the control and treated samples when all segments were analyzed simultaneously. However, several other genes showed significant expression level changes in certain segments alone after either THP or DNB exposure. This indicates that THP and DNB cause significant changes in gene expression in specific compartments of the male reproductive system. For example, Psmb8 showed a P-value of 0.156 between the control and THP-treated animals when all four segments (i.e., the testis and the caput, corpus, and cauda epididymis) were analyzed together. However, according to the feed-segment interaction statistical analysis, the expression change of Psmb8 in the feed-segment interaction was statistically significant (P = 0.0047).
The subunit composition of the proteasomes may vary, and the 20S core can be found uncapped, capped with one 19S complex, capped with two 19S complexes, capped with an 11S activator complex, or capped with one 19S and one 11S complex (20S + 19S + 11S chimera). Given this variability and the paucity of knowledge about the significance of such varied proteasomal composition with regard to the testicular and epididymal function, we decided to focus our attention solely on the 20S proteasomal core. We chose to follow all three constitutive ß-type subunits that can be substituted for by the inducible ß subunits. While we found that both constitutive ß subunits and their inducible counterparts were expressed in the testis and epididymis, only the constitutional subunits showed significant expression changes in the tissues of the exposed animals. Products of two of the three examined UPP enzyme genes not directly associated with the proteasome, the Ube1 and Ube2d3 genes, also showed significant changes overall or in some segments. This indicates that the screening of the ubiquitin system provides valuable information about the effect of toxic exposure on the male reproductive system. In the future, we will explore the simultaneous screening of all 19S, 20S, and 11S subunits as well as an expanded repertoire of ubiquitin-conjugating and deubiquitinating enzymes. Such an analysis could be greatly facilitated by automated transcriptional profiling—for example, by a microarray designed specifically to capture all gene products within the UPP.
In summary, our data indicate that DNB and THP exposures alter the expression of select genes within the UPP in a time- and dose-dependent manner. In particular, the Psmb1, Psmb2, Psmb5, and Ube2d3 genes that encode the constitutive proteasomal core and ubiquitin-activating enzyme UBE2 were affected. This change is most significant in the testis and corpus epididymis, but the magnitude of effect differs slightly by the administered toxicant. Such an altered gene expression may correlate with aberrant spermatogenesis in the testis and impair the processing of both normal and defective spermatozoa in the epididymis. Transcriptional profiling and flow cytometric analysis of the UPP thus capture the subtle effects of reproductive toxicity not observed by conventional histology and functional cytology and offer a prospective new tool to detect and manage reproductive toxicology. In addition to mapping the effect of reprotoxic exposure on the expression of a specific subset of functionally related genes in the testis and epididymis, the present study showed a gradient in the expression of the UPP pathway genes that, for most genes, seemed to follow the pattern of testis > caput > corpus > cauda or caput > testis > corpus > cauda.
ACKNOWLEDGMENTS
We thank Kathryn Craighead (University of Missouri at Columbia [UMC]) for clerical assistance, Nicole Leitman (UMC) for technical assistance, Dr. Mark R. Ellersieck for help with statistical analysis, Jill Long (PFE) and Dr. David Reim (PFE) for study execution, Dr. Dan Morton (PFE) for histopathological review, Dr. Bob Chapin (PFE) for critical discussions, and the staff of DNA Core and Histology Core of the UMC for sample processing.
FOOTNOTES
1Supported by a research grant from Pfizer Inc. (PFE; New York, NY) to P.S. and by matching funds from the Food for the 21st Century Program of the University of Missouri at Columbia. 
Correspondence: 2 Mark W. Tengowski, 2800 Plymouth Rd., MS16-1A/6, Ann Arbor, MI 48105. FAX: 734 622 5196; e-mail: mark.w.tengowski{at}pfizer.com
Received: 13 April 2006.
First decision: 7 May 2006.
Accepted: 13 September 2006.
REFERENCES
This article has been cited by other articles:
|
|
 |
