Many studies in the past have reported that Steel factor plays an essential role in PGC development 
. Most recently, work from our lab revealed that the Steel factor-expressing cells establish a “traveling niche” surrounding migratory PGCs which provides a short-range signal to regulate PGC survival, proliferation and motility throughout their migration 
. However, how these different behaviors are regulated by the same ligand, and how Steel factor functions as a short-range signal remain uncertain. In other cell populations, the two forms of Steel factor, membrane-bound and soluble form, have been shown to have different functions 
, although the mechanisms whereby they do so have remained unclear. This suggested that different aspects of PGC behavior might also be regulated through different forms of Steel factor. The Steel-dickie
mutation, which only produces soluble Steel factor due to a deletion in cytoplasmic and transmembrane domain of Steel
allele, offers a great tool to address this question. Previous work has shown that transcription from the Steel-dickie
allele is at the same level as wild type. Here we showed, by western blotting using supernatant fractions of E12.5 whole embryos, that Steel factor protein is produced in the mutant embryos, and that there is an increased amount of the soluble form, compared to the wild type embryo. This is the expected result if the mutation generates only the soluble form, the translation levels are similar, and if protein stability is not altered. In this case, all the translated protein will be in the soluble form in the Steel-dickie
mutant embryo. To test the functions of the two forms, we analyzed PGC number and motility in E7.5 Steeld/d
embryos. The results showed that the soluble form of Steel factor present in Steeld/d
embryos maintained the survival but not the motility of PGCs. This shows that Steel factor does not need to be membrane-bound for PGC survival, but that the membrane-bound form is essential for their motility.
The results presented here also offer a mechanistic explanation of the somewhat puzzling effects of the Steel-dickie
mutation. It has long been known that this mutant is viable but sterile, implying that membrane-bound Steel factor plays different roles in the PGCs and hematopoietic cell lineages. However, Steel factor is known to be a survival factor in both cell types 
, and the soluble Steel factor protein present in Steeld/d
embryos is clearly sufficient to maintain their survival (shown in this work). We propose that the difference is due to the fact that membrane-bound Steel factor provides a niche for maintaining active motility, and in its absence, the PGCs lose motility, and do not move to the next location in the pathway. We have shown previously that Steel factor maintains PGC survival only at short range 
, and that expression of Steel factor moves along the PGC migratory pathway 
. With reduced motility, PGCs in Steeld/d
embryos will not keep up with the wave of Steel factor expression, and will die because this essential survival factor is turned off before they have moved to the next site of expression. All the observations to date support this hypothesis. First, in E7.5 Steeld/d
embryos, no obvious reduction of PGC number were observed, implying that the soluble Steel factor produced in the Steel-dickie
mutants is adequate to maintain PGC survival at this stage. Although there is no significant decrease in cell number, PGCs in Steeld/d
embryos show remarkably reduced motility and instead form a large clump of non-motile cells. Second, Steeld/d
PGCs remain clustered at the boundary between allantois and hind gut at E8.0, while most PGCs in wild type embryos have already migrated anteriorly along the hind gut endoderm. Third, many PGCs are found in ectopic locations with apoptotic morphology in Steeld/d
The finding that addition of soluble Steel factor to E7.5 Steeld/d
embryos is capable of restoring PGC motility is interesting, because it suggests the membrane-bound Steel factor is essential to maintain an optimal local concentration in a defined region, rather than play a distinct role to regulate PGC motility. The addition of large amounts of soluble recombinant Steel factor can restore PGC motility in Steeld/d
embryos, showing that the local ligand concentration is crucial for Steel factor regulation. In support of this view, a previous study showed that in vivo
administration of soluble recombinant Steel factor into Steeld/-
mice caused a notable reduction of the severity of their macrocytic anemia and a great increase of mast cell number at the injection sites 
. The high levels of Steel factor required for PGC motility is normally maintained not by increased synthesis, but by maintenance of a high local concentration by presenting it on the membranes of adjacent cells in a restricted area, thus defining the region in which PGCs can maintain their motility. Movement of this “motility niche” would thus create the pathway along which PGCs are able to move.
When treating the E7.5 Steeld/d and wild type embryos with added soluble Steel factor, some PGCs in the most distal region of allantois tended to migrate randomly, instead of moving proximally towards the posterior epiblast. It may be that these PGCs are already too far away from the normal chemotactic signals, which attract PGCs to the posterior epiblast. Without the added soluble Steel factor, these cells would normally die by apoptosis because they are in an ectopic location where Steel factor expression is being turned off. However, the globally added soluble Steel factor allowed them to survive and migrate randomly. These findings accentuate the importance of maintaining a high local concentration only in a defined region to make sure that PGCs which receive the chemotactic signals maintain motile, whereas the ones out of the range of direction cues cease motility and die.
One observation that remains unexplained is the formation of large clumps of PGCs in the allantois in both Steel-/-
embryos. We have previously shown that PGCs activate expression of E-cadherin when they leave the hind gut. There is a failure of the normal aggregation of germ cells as they enter the genital ridges when E-cadherin function is blocked 
. In a previous study of Steel-null
mutant embryos, we stained for E-cadherin protein in PGCs of different genotypes to check whether PGC clumping at E7.5 was due to an upregulation of E-cadherin in the absence of Steel factor. However, no significant change in E-cadherin expression level was observed 
. To further confirm that E-cadherin is not responsible for the increased cluster formation and decreased motility, we treated wild type E7.5 embryos with Ack2, the inhibitor of c-kit receptor, and ECCD1, an antibody which blocks the function of E-cadherin, to test whether loss of E-cadherin function could rescue the defects caused by loss of Steel factor signaling. Compared to embryos which were only treated with Ack2, the dual treatment group did not show any significant change in motility, displacement and cluster formation (data not shown), indicating the defects shown in Steel-/-
embryos are not due to a premature up-regulation of E-cadherin function. The mechanism of PGC cluster formation in Steel-null
mutants needs to be investigated in future studies.
It is important to maintain PGC motility during their migration. It is also crucial to ensure PGCs stop migration when they coalesce in the genital ridges. Interestingly, addition of soluble Steel factor into embryo slices at E11.0 resulted in reacquisition of motility in PGCs, indicating that PGCs can still respond to high levels of Steel factor, and that modulation of Steel-ckit signaling may control the ending of migration. However, there is no decrease in the expression of Steel factor mRNA in E11.5 genital ridges as determined by real-time RT-PCR, suggesting that the ending of PGC migration is not due to down-regulation of Steel factor expression at transcriptional level. There are a number of possible explanations for this paradoxical finding. First, PGCs could still be motile in the E11.5 genital ridge, but are constrained by other factors. Second, there may be less membrane-bound Steel factor at the protein level. Third, an inhibitor of Steel factor signaling could be present in the genital ridges, which acts independently of Steel factor expression level. Fourth, when PGCs arrive in the genital ridges, the amount of membrane-bound Steel factor available per PGC may decrease. As PGCs adhere to each other, they will be less likely to be adjacent to a membrane-bound Steel factor-expressing somatic cell. Whatever the mechanism of clumping, it is overcome by addition of increased soluble Steel factor, so is likely to include the inhibition of Steel factor signaling. E-cadherin function has been shown to be essential for PGC coalescence in the genital ridges 
. Although there is no obvious change in E-cadherin expression in PGCs with Steel factor addition, we cannot exclude the possibility that Steel factor promotes PGCs to migrate out of the cluster by interrupting E-cadherin function without affecting its membrane expression. Further studies need to be performed to address these questions.
These studies reveal two important aspects of the “traveling niche”. First, the Steel factor activates two different functions in PGCs at different concentrations, and second that the moving expression of membrane-bound Steel factor maintains PGC motility in a defined region of the embryo. PGCs escaping from this region, or that get left behind on the migratory route, cease motility, and are removed by apoptosis as the expression wave moves on towards the genital ridges. Thus the colonization of the genital ridges is as much due to the control of survival and motility as it is to chemotactic signals.