Glial cell line-derived neurotrophic factor maintains a POZ-itive influence on stem cells.

I n adult males, germ-line stem cells have the remarkable ability to both self-renew and differentiate, ensuring that a continuous population of mature spermatozoa is produced throughout the lifetime of the animal. This balance between self-renewal and differentiation is thought to depend on the proper cellular environment, or stem cell ‘‘niche,’’ that provides the appropriate signals at the right time for these processes. One growth factor shown to be essential for mammalian spermatogonial stem cell (SSC) self-renewal is the glial cell line-derived neurotrophic factor (GDNF) (1). GDNF signals through a receptor complex containing the RET receptor tyrosine kinase and the GDNF family receptor 1 (GFR 1). Gdnf / mouse testes transplanted into WT recipients exhibit a rapid and dramatic loss of SSCs, resulting in testes that contain only the supporting Sertoli cells (2). The development of a culture method and transplantation assay for mouse SSCs has shown that when GDNF is removed from the culture media, the agametic recipient testes into which these cells are transplanted fail to be repopulated, indicating a loss of SSCs (3). Clearly, GDNF is necessary for SSC maintenance, but until now the downstream target genes activated by GDNF and demonstrated as important for SSC self-renewal were unknown. The work of Oatley et al. (4) in a recent issue of PNAS identifies genes regulated by GDNF by using the SSC culture system and microarray analysis. One of these genes, Bcl6b, is a member of the POZ (poxvirus and zinc finger) family of transcriptional repressors that also includes Plzf, previously shown to be required for SSC maintenance (5, 6). These results underscore the importance of the niche in the mammalian testis and extend its molecular characterization by identifying downstream targets of GDNF. The concept of the niche was first proposed by Schofield (7) to describe the relationship between the hematopoietic stem cell and its local microenvironment. For male germ-line stem cells, the best-characterized niche is that of the Drosophila testis. Localizing to the apical somatic hub cells are seven to nine germ-line stem cells that divide asymmetrically. Each stem cell divides to produce one daughter cell that remains attached to the hub and one gonialblast that migrates away and differentiates. The hub generates and releases signals to control this stem cell self-renewal, including two members of the bone morphogenic protein (BMP) growth factor family, glass bottom boat (gbb) and decapentaplegic (dpp) (8). One of the downstream targets of gbb is a dppresponsive gene, Dad, which exhibits transcriptional activation in the stem cells to maintain self-renewal. Through gbb and dpp, BMP signaling represses the expression of bag of marbles (bam), a gene that is transcriptionally active in differentiating spermatocytes but repressed in the mother stem cells and daughter gonialblasts (8). A family of conserved RNA-binding proteins encoded by piwi also represses bam through both BMP-dependent and BMP-independent signaling pathways (9, 10). Thus, through the expression of gbb, dpp, and piwi, and the repression of bam, the BMP growth factor family maintains the stem cell niche in the testis. Another signal generated from the Drosophila hub is Unpaired Outstretched (Os), which activates the Janus kinase (JAK)–signal transducer and activator of transcription (STAT) signaling pathway to maintain stem cell selfrenewal (11, 12). Os is a secreted ligand that is specifically expressed in the hub and upon its release binds to the receptor Domeless on the surface of adjacent stem cells. Receptor binding activates the JAK homologue Hopscotch and the STAT homologue Stat92E, which then translocates to the nucleus where it activates downstream target genes important for stem cell self-renewal (11). When the JAK–STAT signaling pathway is inhibited, germ cells are depleted from the testis after the complete differentiation of all of the stem cells. Recent evidence has also shown that the centrosome protein centrosomin, together with the adenomatous polyposis coli tumor suppressor, regulates the mitotic spindle orientation in stem cells to ensure that asymmetric cell division generates a daughter gonialblast oriented away from the hub (13). Drosophila E-cadherin and -catenin homologue Armadillo colocalize at the hub–stem cell interface, implicating their role in this orientation process. Given these defined pathways that maintain the Drosophila germ-line stem cell niche, one may ask whether there is a similar niche in mammalian testes, and if so, what are the required signaling pathways and downstream target genes? Performing some of the functions of Drosophila hub cells in mammals are the Sertoli cells, which physically interact with SSCs and release growth factors that are proposed to maintain stem cell self-renewal. One growth factor identified for this role is GDNF. As a distant member of the TGFfamily, GDNF was shown to be required for enteric innervation and embryonic kidney development (14, 15). Gdnf / , Gfr 1 / , and Ret / mouse pups all die on postnatal day 1 (14, 16, 17). However, Gdnf / heterozygous mutant males show extensive germ cell depletion, with some seminiferous tubules displaying a progressive Sertoli cell-only phenotype (1). To overcome neonatal lethality and conclusively test whether Gdnf-null mouse testes can support spermatogenesis, Gdnf-deficient testes were transplanted from postnatal day 0 mutants into nude recipients and allowed to develop (2). The resulting testes exhibited severe SSC depletion by postnatal day 7, with an ability of the stem cells to maintain their undifferentiated state. Thus, the GDNF pathway in mammals appears to be as essential as the JAK–STAT pathway in Drosophila in maintaining the niche. Although the activation of JAK– STAT is considered to be a local signaling event in the Drosophila niche, recent experimentation has shown that differentiating cells that are physically detached from the hub can revert to their stem cell identity and migrate back to the hub cells upon receiving JAK–STAT signals (18). It is possible, then, that these signals have a much longer range than previously thought. In the same