Cells that originate in one place in the embryo often have to migrate some distance to reach their ultimate place of residence. A good example of this is the Drosophila germ cells. These cells are specified by maternal factors and lie outside the posterior of the embryo, segregated from the soma until gastrulation. At this stage, they undergo a complex pattern of migration that brings them inside the embryo and into contact with two lateral clusters of somatic gonadal precursor (SGP) cells — an association that creates a functional gonad. But how do germ cells know where to go? In their study of this process, Girish Deshpande and colleagues have found that, by secreting Hedgehog (Hh), SGP cells could provide the attractive cue that guides germ cells to their destination.

The study, published in Cell, was based on the premise that germ cells would follow attractive signals derived from the SGP cells, and that the identity of such molecules could best be found by looking at genes involved in specifying SGP cell fate. Of several such genes, Hh seemed a particularly promising candidate — it is expressed in the SGPs (as shown by the distribution of Hh–lacZ) and is known to be a short- and long-range signalling molecule. A crucial experiment reinforced this prediction: when Hh was ectopically expressed in the embryo, using four different Gal4 drivers, it redirected the migration of a subset of germ cells towards the new source of Hh. An alternative explanation for these results — that germ cells were drawn to ectopic SGP cells newly induced by Hh — is unlikely, as the Hh-expressing cells were negative for a SGP-cell-specific marker.

If germ cells respond directly to Hh, rather than to a secondary signal, then it follows that the germ line should require cell-autonomous components of the Hh pathway for proper migration. Indeed, when mutant for positively acting members of the pathway (such as smoothened or fused ), germ cells scattered randomly and failed to associate with SGP cells, as if they had lost all sense of direction — consistent with loss of Hh. Conversely, patched or protein kinase A mutant germ cells, in which Hh signalling is constitutively active, clumped together in the middle of the embryo and failed to migrate at all.

Interest in germ-cell migration is not new and so several genes that affect this process, such as wunen and Columbus , have already been identified. How does Hh fit in with previous models of germ-cell migration? As there are countless sources of Hh in the embryo, how is specificity of migration achieved? This is probably only one leg of a longer journey to find out how germ cells reach their targets.