Androgens and spermatogenesis: lessons from transgenic mouse models

Abstract

Transgenic mouse models have contributed considerably to our understanding of the cellular and molecular mechanisms by which androgens control spermatogenesis. Cell-selective ablation of the androgen receptor (AR) in Sertoli cells (SC) results in a complete block in meiosis and unambiguously identifies the SC as the main cellular mediator of the effects of androgens on spermatogenesis. This conclusion is corroborated by similar knockouts in other potential testicular target cells. Mutations resulting in diminished expression of the AR or in alleles with increased length of the CAG repeat mimick specific human forms of disturbed fertility that are not accompanied by defects in male sexual development. Transcriptional profiling studies in mice with cell-selective and general knockouts of the AR, searching for androgen-regulated genes relevant to the control of spermatogenesis, have identified many candidate target genes. However, with the exception of Rhox5, the identified subsets of genes show little overlap. Genes related to tubular restructuring, cell junction dynamics, the cytoskeleton, solute transportation and vitamin A metabolism are prominently present. Further research will be needed to decide which of these genes are physiologically relevant and to identify genes that can be used as diagnostic tools or targets to modulate the effects of androgens in spermatogenesis.

Figure 1.

Generation and phenotype of mice with a general (ARKO) or Sertoli cell-selective (SCARKO) knockout of the androgen receptor. (a) Female mice, heterozygous for an AR allele with a floxed exon 2 (AR^flox/+; loxP sites are indicated as red arrowheads) were crossed either with PGK-Cre mice, expressing the Cre recombinase ubiquitously, or with AMH-Cre mice, expressing Cre selectively in SC, to generate ARKO or SCARKO mice, respectively. (b) SCARKO mice have normally descended testes (t) that are markedly reduced in size (28% of control). Ductus deferens (dd), seminal vesicles (sv) and prostate (p) are normally developed. ARKO mice have very small testes that are located intraabdominally and male accessory sex tissues are absent.

Figure 2.

Expression pattern (from days 8 to 20) for a subset of genes originally identified as differentially expressed in SCARKO and control mice on day 10. The genes studied are Rhox5 (Pem), Eppin, Galgt1 (β-1,4-acetylgalactosaminyltransferase), Drd4 (dopamine receptor D4), Tsx (testis-specific X-linked), Gpd1 (glycerol-3-phosphate dehydrogenase 1), Tubb3 (tubulin β3), PCI (protein C inhibitor) and Tpd52l1 (tumour protein D52-like 1). Expression was assessed by microarray analysis and quantitative RT-PCR (qPCR). Left axis (bars): expression levels measured by qPCR in testes of control and SCARKO mice of the indicated ages (n = 3). Data were normalized to an external luciferase standard. All values are expressed as a percentage of the highest value measured for the corresponding gene arbitrarily set at 100. Values represent the mean ± s.e.m. of three measurements. Right axis (lines): gene expression measured by microarray analysis on a pool of mRNA from three testes of three control or SCARKO mice of the indicated ages. Data were expressed as a percentage of the highest signal observed for the studied gene, arbitrarily set at 100. Notice that qPCR confirms differential expression between SCARKO and control on day 10 for all the genes identified by microarray analysis. While the microarray data (reflecting the number of transcripts in a given amount of RNA) suggest a decrease in the transcript levels for most of the studied genes, and for some of them a loss of differential expression, the qPCR measurements (corrected for exogenously added luciferase and accordingly reflecting transcript levels per testis) show that this is an artefact caused by the increased contribution of developing germ cells to the total amount of RNA selectively in the control. The experiment illustrates that genes differentially expressed in SC may be missed by microarray analysis on samples with different degrees of germ cell maturation. Unfilled bar, control (qPCR); filled bar, SCARKO (qPCR); unfilled circle, control (microarray); filled circle, SCARKO (microarray).

Figure 3.

Molecular mediators of androgen action on spermatogenesis. (b) Indicates the main steps in spermatogenesis affected by androgens. In (a), a number of processes and molecular targets are summarized that are identified as potential targets of SC-mediated androgen action by the microarray experiments discussed in the text. The numbers in the centrally represented SC (yellow) refer to the processes summarized on (a). Developing germ cells are indicated in blue. Microtubules and intermediate filaments are represented in green and dark blue, respectively. Tight junctions are indicated in purple and the ectoplasmic specializations are represented by actin bundles (red) and a cistern of endoplasmic reticulum.