Much is known about early patterning in Drosophila melanogaster development, but the subsequent events are still a mystery. For example, which genes control the migration and differentiation processes required to generate specialized tissues? Weiss and colleagues recently addressed this question, and, reporting in Cell, they describe the discovery of a novel Drosophila gene called jelly belly (jeb). They propose that the Jeb protein acts as a positive extracellular signalling molecule and that it is essential for the development of midgut muscles (visceral mesoderm).

Visceral mesoderm precursors are formed when the products of the genes decapentaplegic (dpp) and hedgehog (hh) act in combination to activate bagpipe (bap), the product of which is an NK class homeodomain protein.

Weiss and collegues did a screen to identify Drosophila genes that are transcriptionally controlled by Tinman (Tin), another member of the NK class of homeodomain proteins and an essential regulator of visceral mesoderm development. One of the DNA fragments isolated in this screen lay adjacent to a novel gene, jeb, which is expressed in somatic muscle precursors. The authors showed that, although sufficient, Tin is not necessary for jeb expression in these precursor cells. The jeb gene is expressed in somatic muscle precursors from stages 8–12 of Drosophila embryo development, but Jeb is not required for somatic muscle development.

To investigate Jeb function, Weiss et al. looked at jeb mutants and found that no differentiated visceral mesoderm could be detected, although other muscular components of the mesoderm developed normally. Visceral mesoderm precursors were specified in the mutants, as indicated by normal staining for Bap, but they failed to migrate and did not differentiate to form the visceral mesoderm. The authors propose that, in jeb mutants, these precursors default to a somatic mesoderm cell fate, as they found no evidence of apoptosis and jeb-mutant embryos showed an increase in the number of nuclei in the positions of somatic muscle precursors.

The somatic muscle precursors in which Jeb is produced lie next to the visceral mesoderm precursors that depend on Jeb function. Jeb contains a secretory signal sequence and a type A LDL receptor repeat. Based on these features, the authors proposed that Jeb is secreted from somatic precursor cells and that it acts in the extracellular compartment. They confirmed this by showing that Jeb is secreted from Drosophila tissue culture cells and that bap-expressing visceral muscle precursors, which do not express jeb but depend on jeb function, contain Jeb protein. The authors showed that visceral mesoderm precursors accumulate Jeb by receptor-mediated endocytosis. This endocytosis requires the type A LDL receptor repeat in Jeb; a dynamin-related GTPase, Shibire, that is required for microtubule-mediated endocytosis; and, probably, a Jeb-specific receptor.

To rule out the possibility that Jeb acts in somatic muscle precursors to produce a signal that is not Jeb, the authors expressed Jeb in the visceral muscle precursors of jeb-mutant embryos. They showed that this expression rescued differentiation of these precursor cells, but that migration of the precursors remained defective. Weiss and co-workers therefore propose that Jeb acts as a signal, and that, by being secreted from somatic muscle cell precursors, it conveys positional information to visceral muscle precursors.

In conclusion, Jeb is required for Drosophila midgut muscle development. It acts as a positive migratory or differentiation signal for visceral mesoderm precursors, and it is possible that this new signalling system has been conserved in evolution. It seems that Jeb might act as a developmental signal in many contexts, as Weiss et al. have shown that jeb messenger RNA is expressed in the embryonic central nervous system and that Jeb can be transported along axons. The authors are now investigating Sco-spondin in mice, a secreted protein that is highly similar to Jeb in the functionally important LDL receptor repeat region. This study, combined with the work presented in Cell, should further unravel the complexities and evolutionary origins of specialized tissue development.