Inhibit thy neighbour

How does a homogeneous population of neurogenic ectodermal cells give rise to a specific pattern of neuronal precursor cells or neuroblasts? We now know that just one cell within a group adopts a different fate and then sends out inhibitory messages, using the Notch–Delta pathway, to stop its neighbours taking on the same fate, but how did we arrive at this understanding?

Back in 1937, Donald F. Poulson was studying embryonic development in Drosophila melanogaster by using deficiencies involving the entire X chromosome or reasonably large portions thereof. The results of one such deficiency, known as Notch-8, were detailed. “The most striking feature of such eggs is that they contain very little or no endoderm or mesoderm ... the process of germ layer formation has been interfered with seriously... The ectoderm proliferates especially along the ventral mid-line and produces what appears to be a semblance of the early nervous system.”

It was almost 50 years before Chris Doe and Corey Goodman, studying the neural ectoderm of grasshoppers, suggested a mechanism that we now know as lateral inhibition. Using laser microbeams to ablate one or more ectodermal cells in a group, they found that neuroblasts are specified by cell interactions. Initially, each undifferentiated cell within a sheet of neural ectodermal cells has an equal chance of becoming a neuroblast, but only one cell within a group takes on this role. Interactions between the cells of a group allow this one cell to enlarge into the neuroblast, which somehow prevents its neighbouring cells from taking on the same identity; these cells instead become support cells or die.

So, cell–cell interactions were obviously important, but what molecules were behind the mechanism? In 1985, Spyros Artavanis-Tsakonas' group used chromosomal walking and the Sanger sequencing method (outlined only a few years earlier) to determine the nucleotide sequence of four overlapping cDNA clones encompassing the Notch locus. From this, the predicted 2,703-amino-acid sequence was determined.

About half of the predicted polypeptide sequence contains 36 tandem cysteine-rich repeats of ∼40 amino acids, with 6 cysteines positioned at specific intervals within each repeat. Searching the Protein Information Resource database identified considerable homology of these repeats with epidermal growth factor (EGF). At the time, the functional significance of this homology was unclear, but Artavanis-Tsakonas and colleagues later showed that these are important in the interaction between Notch and its ligands. Indeed, hydropathy plots by Artavanis-Tsakonas' group implied that Notch spanned the membrane, and the EGF repeats would therefore be extracellular. The authors discussed, on the basis of genetic studies, the possibility that the products of Notch, Delta and Enhancer of split — a known target of Notch signalling — might interact. They proposed that “Notch and all or some of the other neurogenic loci may be involved in a cellular mechanism that mediates and interprets cell–cell interactions, which result in the correct differentiation of the neurogenic region” — a hypothesis that turned out to be correct.

We now know that Notch and Delta are the defining players in the signalling pathway that underlies lateral inhibition. However, there is still much to learn about this process — several other molecules also have key roles, and the mechanism by which Notch–Delta interactions activate gene transcription is still incompletely understood.