In mouse embryos, germ cells arise during gastrulation and migrate to the early gonad. First, they emerge from the primitive streak into the region of the endoderm that forms the hindgut. Later in development, a second phase of migration takes place in which they migrate out of the gut to the genital ridges. There, they co-assemble with somatic cells to form the gonad. In vitro studies in the mouse, and genetic studies in other organisms, suggest that at least part of this process is in response to secreted signals from other tissues. Recent genetic evidence in zebrafish has shown that the interaction between stromal cell-derived factor 1 (SDF1) and its G-protein-coupled receptor CXCR4, already known to control many types of normal and pathological cell migrations, is also required for the normal migration of primordial germ cells. We show that in the mouse, germ cell migration and survival requires the SDF1/CXCR4 interaction. First, migrating germ cells express CXCR4, whilst the body wall mesenchyme and genital ridges express the ligand SDF1. Second,the addition of exogenous SDF1 to living embryo cultures causes aberrant germ cell migration from the gut. Third, germ cells in embryos carrying targeted mutations in CXCR4 do not colonize the gonad normally. However, at earlier stages in the hindgut, germ cells are unaffected in CXCR4-/-embryos. Germ cell counts at different stages suggest that SDF1/CXCR4 interaction also mediates germ cell survival. These results show that the SDF1/CXCR4 interaction is specifically required for the colonization of the gonads by primordial germ cells, but not for earlier stages in germ cell migration. This demonstrates a high degree of evolutionary conservation of part of the mechanism, but also an area of evolutionary divergence.

In mouse embryos, the primordial germ cells arise during gastrulation prior to, and distant from, the prospective gonads. Observations of PGCs in culture, and in fixed sections, have suggested, but not proved, that they migrate to the gonad by a process of active migration. The opaque nature of the early mouse embryo has precluded direct observation. Using confocal microscopy, we have filmed living PGCs expressing eGFP in tissue slices from mouse embryos at different stages of development. We find four clearly distinct phases of PGC migration. First, until E9.0-E9.5, PGCs are already highly motile, but do not leave the gut. Second, in the E9.0-E9.5 period, before the mesentery forms, PGCs very rapidly exit the gut, but do not migrate towards the genital ridges. Third, during the E10.0-E10.5 period, PGCs migrate directionally from the dorsal body wall into the genital ridges. In contrast to the prevailing model of germ cell migration, very few, if any, PGCs found in the gut mesentery at E10.5 migrate into the genital ridges. Finally, at E11.5, PGCs are slowing and the direction of movement is dependent on the sex of the embryo. This allows, for the first time, a formal description of the events of PGC migration in the mouse.

During germ-cell migration in the mouse, the dynamics of embryo growth cause many germ cells to be left outside the range of chemoattractive signals from the gonad. At E10.5, movie analysis has shown that germ cells remaining in the midline no longer migrate directionally towards the genital ridges, but instead rapidly fragment and disappear. Extragonadal germ cell tumors of infancy, one of the most common neonatal tumors, are thought to arise from midline germ cells that failed to die. This paper addresses the mechanism of midline germ cell death in the mouse. We show that at E10.5, the rate of apoptosis is nearly four-times higher in midline germ cells than those more laterally. Gene expression profiling of purified germ cells suggests this is caused by activation of the intrinsic apoptotic pathway. We then show that germ cell apoptosis in the midline is activated by down-regulation of Steel factor (kit ligand) expression in the midline between E9.5 and E10.5. This is confirmed by the fact that removal of the intrinsic pro-apoptotic protein Bax rescues the germ-cell apoptosis seen in Steel null embryos. Two interesting things are revealed by this: first, germ-cell proliferation does not take place in these embryos after E9.0; second, migration of germ cells is highly abnormal. These data show first that changing expression of Steel factor is required for normal midline germ cell death, and second, that Steel factor is required for normal proliferation and migration of germ cells.

In the mouse embryo, significant numbers of primordial germ cells (PGCs)fail to migrate correctly to the genital ridges early in organogenesis. These usually die in ectopic locations. In humans, 50% of pediatric germ line tumors arise outside the gonads, and these are thought to arise from PGCs that fail to die in ectopic locations. We show that the pro-apoptotic gene Bax,previously shown to be required for germ cell death during later stages of their differentiation in the gonads, is also expressed during germ cell migration, and is required for the normal death of germ cells left in ectopic locations during and after germ cell migration. In addition, we show that Bax is downstream of the known cell survival signaling interaction mediated by the Steel factor/Kit ligand/receptor interaction. Together, these observations identify the major mechanism that removes ectopic germ cells from the embryo at early stages.

Xenopus Vg1, a transforming growth factor ␤ (Tgf␤) family member, was one of the first maternally localized mRNAs identified in vertebrates. Its restriction to the vegetal pole of the egg made it the ideal candidate to be the mesoderm-inducing signal released by vegetal cells, but its function in vivo has never been resolved. We show that Vg1 is essential for Xenopus embryonic development, and is required for mesoderm induction and for the expression of several key Bmp antagonists. Although the original Vg1 transcript does not rescue Vg1-depleted embryos, we report that a second allele is effective. This work resolves the mystery of Vg1 function, and shows it to be an essential maternal regulator of embryonic patterning.

No abstract

Members of the bone morphogenetic protein (BMP) family play diverse roles in multiple developmental processes. However, in the mouse, mutations in many BMPs, BMP receptors and signaling components result in early embryonic lethality making it difficult to analyze the role of these factors during organogenesis or tissue homeostasis in the adult. To bypass this early lethality, we used an organ culture system to study the role of BMPs during primordial germ cell (PGC) migration. PGCs are the embryonic precursors of the sperm and eggs. BMPs induce formation of primordial germ cells within the proximal epiblast of embryonic day 7.5 (E7.5) mouse embryos. PGCs then migrate via the gut to arrive at the developing gonads by E10.5. Addition of BMP4 or the BMP-antagonist Noggin to transverse slices dissected from E9.5 embryos elevated PGC numbers or reduced PGC numbers, respectively. Noggin treatment also slowed and randomized PGC movements, resulting in a failure of PGCs to colonize the urogenital ridges (UGRs). Based on p-Smad1/5/8 staining, migratory PGCs do not respond to endogenous BMPs. Instead, the somatic cells of the urogenital ridges exhibit elevated p-Smad1/5/8 staining revealing active BMP signaling within the UGRs. Noggin treatment abrogated p-Smad staining within the UGRs and blocked localized expression of Kitl, a cytokine known to regulate the survival and motility of PGCs and Id1, a transcription factor expressed within the UGRs. We propose that BMP signaling regulates PGC migration by controlling gene expression within the somatic cells along the migration route and within the genital ridges.

GP130 is the shared receptor for members of the IL6 family of cytokines. Members of this family have been shown to enhance the survival of migratory (E10.5) or postmigratory (E12.5) murine primordial germ cells (PGCs) in culture; however, it is uncertain what role these cytokines play during PGC development in vivo. We have examined PGC numbers in E13.5 GP130-deficient mouse embryos and found that males exhibited a slight decrease in PGC numbers; females were normal. Also, we used the Cre-loxP system to inactive GP130 specifically in germ cells and found that this resulted in a fertility defect in females. These animals were found to have a slight reduction in the number of primary follicles and a major defect in ovulation. This data suggests that GP130 is required in female germ cells for their normal function, but is dispensable in male germ cells. These cytokines share common receptors and common signal transduction machinery, and this might explain the relatively mild phenotypes resulting from inactivation of a single family member or a single high affinity receptor. Ablation of the common receptor (GP130), or a common cytoplasmic component (STAT3), should affect all IL6 family members and result in stronger phenotypes. This appears to be the case. STAT3-knockout animals die by E7.5 (Takeda et al., 1997), and GP130-null animals die of cardiac and hematological disorders by E15.5 (Yoshida et al., 1996), or at birth (Kawasaki et al., 1997), depending on the genetic background. Intriguingly, GP130-null animals have been reported to have fewer numbers of PGCs (T. Taga, unpublished).In order to clarify the role of the IL6 family in PGC development, we examined PGC numbers in GP130-deficient males and females. In addition, we have used the Cre-loxP system to generate germ-cell-specific ablations of GP130. Surprisingly, our data demonstrate that GP130-mediated signaling is not required for the early stages of PGC development, but reveal a novel role for GP130-mediated signaling late in oogenesis. MATERIALS AND METHODS Mouse strains and genotypingThe TNAP-Cre, GP130Flox and Oct4∆PE:GFP lines have been described previously (Anderson et al., 1999; Betz et al., 1998;Lomeli et al., 2000). For breed tests, CD1 males and females were purchased from Charles River. Genotyping of animals was performed by PCR. DNA was isolated from tail snips (adults) or from heads (embryos) using the Wizard Genomic DNA purification kit (Promega). PCR was performed on 1-100 ng of genomic DNA using RedMix Plus (PGC Scientific) or Platinum PCR Supermix (Invitrogen) as a source of Taq, buffer and dNTPs. Final primer concentrations were 0.4 µM. PCR consisted of: an initial denaturing step of 5 minutes at 95°C; followed by 5 cycles of 30 seconds at 95°C, 1 minute at 65°C and 30 seconds at 72°C; followed by 35 cycles of 30 seconds at 95°C, 1 minute at 60°C and 30 seconds at 72°C; followed by a 10 minute extension step at 72°C. Primers for genotyping the Oct4∆PE:GFP strain are described by Anderson et al. (Anderson et al., 1999). Primers for the TN...