A signalling scientist marvels at perfect patterns.

The formation of patterns during animal development depends to a great extent on cells, or groups of cells, sending a specific signal that activates a cascade of reactions in the cells that receive and respond to it. Studies of this process in the fruitfly Drosophila have provided many insights into the nature of the molecules involved and the mechanisms underlying cell–cell signalling.

The cell surfaces of almost all animals are decorated extensively with large molecules known as heparan sulphate proteoglycans (HSPGs). These modulate most developmental signalling pathways and comprise protein cores modified by the addition of long carbohydrate chains called glycosaminoglycans (GAGs). GAGs are key to mediating interactions between HSPGs and the molecules that they bind.

Recently, Rahul Warrior at the University of California, Irvine, and his colleagues (Development 135, 1039–1047; 2008) explained the puzzling observation that although HSPGs are required for signalling by the protein BMP in certain tissues, they are not required for BMP signalling during very early fly development.

The authors demonstrate that GAG synthesis does not occur in early embryos because the messenger RNAs that encode two enzymes involved in its construction are not translated.Preventing GAG synthesis at this stage allows an 'activity gradient' of BMP to be generated across the embryo that patterns the dorso–ventral axis of the fly. A few hours later, the GAG enzymes are produced, allowing the modified HSPGs to participate in other signalling pathways.

This study illustrates how temporal control of the synthesis of a ubiquitous set of enzymes is used to modulate the activity of signalling pathways in different tissues.

Discuss this paper at http://blogs.nature.com/nature/journalclub