Due to the limited numbers of PGCs

Due to the limited numbers of PGCs in mouse embryos, it is difficult to dissect the molecular mechanisms of early germ cell development. Recently, using ESC in vitro differentiation models, a number of studies demonstrated that Dazl, Blimp1, Lin28, and possibly Rhox6/Psx1, participated in PGC formation from ESCs (Kee et al., 2009; Liu et al., 2011; West et al., 2009; Yu et al., 2009). However, because of the pluripotency features shared by ESCs and PGCs, many germ cell specific genes, including SSEA1 surface antigen, Oct4, Blimp1, and Stella, are all highly expressed in ESCs. So far, there is a lack of definitive markers to distinguish PGCs and their derivative, EGCs, from ESCs except methylation erasure of imprinted loci. In our study, after day 5 of EB differentiation, EGC colonies were solely derived from SSEA1+ alizarin (data not shown), and our methylation analyses displayed imprinting erasure in EGC colonies (Figs. 2E, F), proving these SSEA1+ EGCs were indeed from germ cell lineage. Additionally, expression data demonstrated that SSEA1+ cells were enriched with pluripotency related genes (Oct4 and Nanog), key regulators in PGC fate specification (Blimp1 and Prdm14), and other PGC markers, (Stella and Fragilis), as well as germ cell genes that expressed in PGCs after migration around 10.5–11.5 dpc (Mvh and Dazl) (Fig. 1C) (Tanaka et al., 2000). Given that EGC culture can only be established from PGCs before sex determination in vivo (McLaren, 2003), SSEA1+ cells in our assays recapitulated the expression profiles and biological properties of post-migrating PGCs before sex determination. Gasz was highly enriched in these SSEA1+ PGCs derived from differentiated ESCs, in contrast to its low expression level in ESCs (Fig. 1C). This result was consistent with its first appearance around 12.5dpc in genital ridges (Figs. 1D–E). PGCs are segregated from somatic cells in mice at 6.5dpc, and settle at the genital ridges before 11dpc (Ginsburg et al., 1990; McLaren, 2003). Therefore, GASZ can complement DAZL and MVH as a unique and useful marker to distinguish post-migrating PGCs from ESCs. We previously reported that a null mutation of Gasz on a C57BL6/J/129 hybrid background led to male infertility through increased expression of retrotransposons and block of meiosis in post-natal spermatocytes (Ma et al., 2009). Our data revealed that GASZ overexpression promotes SSEA1+ PGC derivation from ESCs, and its knockdown compromises EGC formation (Figs. 2A–D, 3B). Altered GASZ expression specifically affects transcript levels of genes in post-migrating PGCs, such as Dazl, Mvh, Mili, Miwi, and premeiotic marker Stra8 in SSEA1+ cells. Results were reproduced in several independent experiments with different ESC lines from E14, C57/BL6, and CCE (Fig. 2, Supplementary Fig. 1H, 2B, 2D), arguing against an artificial phenomenon due to gene manipulation. Consistently with these in vitro findings, we observed that GASZ started to express in post-migrating PGCs from 12.5dpc, and the expression levels of DAZL and MVH were decreased in GASZ−/− gonads at 14.5dpc (Fig. 3E, Supplementary Fig. 2H), suggesting that GASZ is required for maintaining the expression of these key post-migrating PGC regulators during embryonic development. Therefore, our data proposed a previously under-appreciated role of GASZ during embryonic germ cell development. Many genes have been implicated during terminal differentiation and maturation of germ cells during adulthood, while fewer genes were found to participate in the maintenance of post-migrating PGCs before birth. It is possible that mild defects during embryonic germ cell formation may not be manifested due to current technical limitation. In addition, the degree of impaired fertility varies frequently among different genetic backgrounds in gene targeting analyses. Notably, gene knockout analyses of Dazl or Pin1 yield no embryonic abnormality prior to backcrossing with C57/BL6 mice, where both XY and XX Dazl null or Pin1 deficient embryos displayed defects in early germ cell development (Atchison et al., 2003; Lin and Page, 2005). Although no obvious gross abnormalities were detected in GASZ−/− gonads at 14.5dpc, given the interaction and functional relationship between GASZ and DAZL, further studies in Gasz knockout mice with different genetic backgrounds may help us understand the physiological requirement of GASZ during embryonic germ cell development.