Proteomic changes in rat spermatogenesis in response to in vivo androgen manipulation; impact on meiotic cells

Abstract

The production of mature sperm is reliant on androgen action within the testis, and it is well established that androgens act on receptors within the somatic Sertoli cells to stimulate male germ cell development. Mice lacking Sertoli cell androgen receptors (AR) show late meiotic germ cell arrest, suggesting Sertoli cells transduce the androgenic stimulus co-ordinating this essential step in spermatogenesis. This study aimed to identify germ cell proteins responsive to changes in testicular androgen levels and thereby elucidate mechanisms by which androgens regulate meiosis. Testicular androgen levels were suppressed for 9 weeks using testosterone and estradiol-filled silastic implants, followed by a short period of either further androgen suppression (via an AR antagonist) or the restoration of intratesticular testosterone levels. Comparative proteomics were performed on protein extracts from enriched meiotic cell preparations from adult rats undergoing androgen deprivation and replacement in vivo. Loss of androgenic stimulus caused changes in proteins with known roles in meiosis (including Nasp and Hsp70-2), apoptosis (including Diablo), cell signalling (including 14-3-3 isoforms), oxidative stress, DNA repair, and RNA processing. Immunostaining for oxidised DNA adducts confirmed spermatocytes undergo oxidative stress-induced DNA damage during androgen suppression. An increase in PCNA and an associated ubiquitin-conjugating enzyme (Ubc13) suggested a role for PCNA-mediated regulation of DNA repair pathways in spermatocytes. Changes in cytoplasmic SUMO1 localisation in spermatocytes were paralleled by changes in the levels of free SUMO1 and of a subunit of its activating complex, suggesting sumoylation in spermatocytes is modified by androgen action on Sertoli cells. We conclude that Sertoli cells, in response to androgens, modulate protein translation and post-translational events in spermatocytes that impact on their metabolism, survival, and completion of meiosis.

Figure 1. Schematic diagram of study design and rationale.

Three treatment groups (TE, TE+Flut, TE+T24) were utilised in this study, as well as an untreated control group. Each group consisted of 4–5 adult rats. These treatments have been used previously, and their effects on intratesticular testosterone levels (iTT) and spermatogenic cell populations have been described, see Results. At completion of treatment, enriched meiotic cell preparations were prepared from each animal and total protein was isolated. Equal amounts of protein from each rat was subjected to 2-dimensional Difference In-Gel Electrophoresis (2D-DIGE) analysis. Protein spots were considered to be differentially expressed in response to in vivo androgen manipulation if they showed a statistical (p<0.05) difference between the TE+Flut and TE+T24 groups. For full details, see Materials and Methods.

Figure 2. Proteomic analysis of enriched meiotic cell preparations.

A. Representative 2D-DIGE image of rat spermatocyte proteins (first dimension pH 4–7, second dimension 8–16% polyacrylamide gradient). The image shown is the Cy2-labelled internal standard which was added to all gels, and was prepared by mixing equal amounts of all samples (see Materials and Methods for details). Proteins shown to be significantly different (p<0.05) between groups by image analysis and subsequently identified by mass spectrometry are indicated (see Table 1 for details of spot identity). B. Principle component analysis of the expression patterns of all 738 spots identified four separate clusters corresponding to the four treatment groups.

Figure 3. Localisation of oxidised DNA adducts and RKIP during androgen manipulation in vivo.

A. Immunohistochemical localisation of oxidised DNA adducts as detected by 8OHdG (green) labelling in testis from Control and - Androgen (androgen suppressed, TE+Flutamide) rats. A negative control for the primary antibody is also shown (1°Ab Control) in a TE+Flutamide-treated testis. Positively labelled pachytene spermatocytes were only apparent during androgen suppression (arrowheads). B. Immunohistochemical localisation of RKIP (green) in testis from Control and - Androgen (androgen suppressed, TE+Flutamide). A negative control for the primary antibody is also shown (1°Ab Control). In controls, staining was most apparent in Sertoli cells (SC) and elongating spermatid cytoplasm (eST), but was faintly present in pachytene spermatocyte cytoplasm (arrowheads). During androgen suppression (- Androgen) a marked increase in immunostaining for RKIP was noted throughout the epithelium, with cytoplasmic staining more obvious in pachytene spermatocytes (arrowhead). In A and B, nuclei are labelled blue (TOPRO). C. Confirmation of changes in expression of androgen-responsive RKIP isoforms; the left panel shows representative images of the 2D-Western during androgen blockade (-Androgen, TE+Flutamide) compared to androgen replacement (+ Androgen, TE+T24). Blots were performed on pooled samples from the same individual animals used for the 2D-DIGE proteomics. Five distinct isoforms (#1– # 5) were resolved. Results of the densitometric analysis (right panel) from 2D western blots revealed that 3 isoforms showed significant (* p<0.05, ** p<0.01) differences between the –Androgen and + Androgen groups. Data is shown as mean ± SD (n = 4 separate experiments).

Figure 4. DDX4 during androgen manipulation.

A. Immunohistochemical localisation of DDX4 (green) in control testis. Staining is observed in late pachytene spermatocyte (PSC) cytoplasm and chromatoid body precursor structures in the perinuclear region (arrowheads). Inset shows control for the primary antibody. B. During androgen blockade (TE+Flutamide), DDX4 immunostaining intensity increased in late pachytene spermatocyte (PSC) cytoplasm. In panels A and B, cell nuclei were visualized with TOPRO (red). C. Evaluation of androgen-responsive pI isoforms of DDX4; the upper panel shows a representative image for the 2D-Western during androgen blockade (-Androgen, TE+Flutamide) compared to androgen replacement with T24 (TE+T24, +Androgen). Fourteen distinct pI isoforms were resolved. Results of the densitometric analysis of pooled samples (lower panel) from –Androgen and +Androgen groups (for details see Figure 3 legend) revealed that one isoform showed a significant (p<0.05, t-test) difference between these groups (asterix), however the other isoforms showed trends to increase or decrease with androgen replacement. Data is shown as mean ± SD (n = 3 separate experiments).

Figure 5. SUMO1 during androgen manipulation.

A. Immunohistochemical localisation of SUMO1 (green) in control testis. Staining is observed in the cytoplasm of pachytene spermatocytes (PSC) in this stage VII tubule, whereas no staining was observed in the primary antibody control (inset). Cell nuclei were visualized with TOPRO (red). B. SUMO1 immunostaining in pachytene spermatocyte (PSC) cytoplasm was reduced during androgen suppression, however staining associated with the nuclei and the XY body (arrowheads) was preserved. C. Densitometric analysis of 15 kDa SUMO1 (i.e. ‘free’ SUMO1) in 1D-western blots with n = 4 separate animals/group from the four different treatments. Different letters denote statistical differences (p<0.01) between groups. During androgen blockade (TE+Flut), there was a significant decrease in free SUMO1 compared to control. Data is shown as mean ± SD.

Figure 6. PCNA during androgen manipulation.

A. PCNA immunostaining (green) in control testis. Representative images from stages I-III and VIII are shown, with visualization of cell nuclei using TOPRO (blue). Pachytene spermatocytes (arrowheads) were immuno-positive in the early stages, but became immuno-negative around stages VII-VIII. PCNA was also observed in proliferating spermatogonia (asterix), whereas no staining was observed in the primary antibody control (inset). B. PCNA immunostaining was more intense in pachytene spermatocytes (arrowhead) in stages I-VI when androgen action was suppressed. C. Densitometric analysis of PCNA in 1D Western blots revealed a significant (p = 0.001, asterix) increase in PCNA protein during androgen blockade (TE+Flut) compared to control. Data is shown as mean ± SD (n = 5).