The conserved transcriptome in human and rodent male gametogenesis

Abstract

We report a cross-species expression profiling analysis of the human, mouse, and rat male meiotic transcriptional program, using enriched germ cell populations, whole gonads, and high-density oligonucleotide microarrays (GeneChips). Among 35% of the protein-coding genes present in rodent and human genomes that were found to be differentially expressed between germ cells and somatic controls, a key group of 357 conserved core loci was identified that displays highly similar meiotic and postmeiotic patterns of transcriptional induction across all three species. Genes known to be important for sexual reproduction are significantly enriched among differentially expressed core loci and a smaller group of conserved genes not detected in 17 nontesticular somatic tissues, correlating transcriptional activation and essential function in the male germ line. Some genes implicated in the etiology of cancer are found to be strongly transcribed in testis, suggesting that these genes may play unexpected roles in sexual reproduction. Expression profiling data further identified numerous conserved genes of biological and clinical interest previously unassociated with the mammalian male germ line.

Fig. 1.

Schematic representation of the mammalian testis. (a) Drawing of the testis containing seminiferous tubules and the epididymis. (b) Sertoli nurse cell associated with a mitotic spermatogonium, a meiotic pachytene spermatocyte, and round and elongated postmeiotic spermatids.

Fig. 2.

The core testicular transcriptome in mammals (CDET). (a) Venn diagram of conserved probesets (top number) and corresponding genes (bottom number) represented on each GeneChip. (b) Summarizes the filtration and clustering procedure. (c) Heatmaps of expression patterns in four clusters for three species as indicated. Numbers of probe sets and corresponding genes in each expression cluster are shown at the left. Sample names are Sertoli cells (SE), spermatogonia (SG), spermatocytes (SC), spermatids (ST), tubules (TU), and total testis (TT). Human samples include chondrocytes (CC) and vascular smooth muscle (SM) cells. To produce a multispecies heatmap where each line contains the mouse, human, and rat homologs, only the mouse genes were clustered, whereas the data for corresponding human and rat orthologs are displayed on the same line. Log2-transformed signals are shown according to the scale bar.

Fig. 3.

Tissue profiling identifies potentially testis-specific genes. (a) Heatmap showing expression in mouse germ cells and somatic controls of genes expressed specifically in Sertoli cells or mitotic, meiotic, and postmeiotic germ cells. The loci are organized into four expression clusters across species as shown. The corresponding probe set and gene numbers per cluster are indicated. Sample names are abbreviated as in Fig. 2 except for A-type spermatogonia (SG A), B-type spermatogonia (SG B), and cumulus-oocyte complex (COC). (b) Shown is the specific number of probe sets (y axis) falling into four expression clusters for which expression was detected in none of the somatic controls (0) or in 1–17 tissues as indicated (x axis). (c) Shows the total sum of probe-sets identified in testis (0) or in testis and 1–17 somatic controls. Expression clusters are shown in blue (SO, somatic), red (MI, mitotic), green (ME, meiotic), and yellow (PM, postmeiotic) as indicated in the legend.

Fig. 4.

Biological Process GO term enrichment in four expression clusters containing genes differentially expressed in testis (DET), conserved and differentially expressed in testis (CDET), or conserved, differentially expressed, and specific to testis (CDEST). These abbreviations are also used in the BioMart query form of GermOnline. Probe set and corresponding gene numbers are given above each map. Each cluster is matched with enriched GO terms from the ontology “biological process” that are ordered according to peak expression in somatic (SO), mitotic (MI), meiotic (ME), and postmeiotic (PM) clusters. Numbers of genes associated with a specific GO term and enriched in each cluster are given within rectangles in bold as observed and as expected. A color code indicates overrepresentation (red) and underrepresentation (blue) as indicated in the scale bar.