People concentrated on this region, according to Shaw, because it has several negatively charged groups that are generally thought to interact with the positively charged groups in hydroxyapatiteor HAP, the crystals that make up bone and enamel.

A series of experiments confirmed that this region played an important role in shaping HAP crystals. Armed with this information, Shaw and colleagues set out to prove that this carboxyl group was indeed the business end of the protein.

To do that, they selected a form of amelogenin called LRAP and isotopically labeled one of the charged amino acids thought to be near LRAPs surface. They put the protein into contact with hydroxyapatite, a proxy for developing enamel crystals, then took its picture. In this case, the camera was a powerful nuclear magnetic resonance instrument capable of recording the positions of tagged protein atoms in relation to the forming HAP crystals.

There are only a handful of labs capable of doing this, Shaw said, and there are more proteins than there are people to look at them all.

The NMR data complement previous results, suggesting that proteins function is to interact with HAP specifically. The carboxyl terminus of the protein is later cleaved by an enzyme, disrupting the protein-HAP interaction and allowing the long, thin crystals to grow outward as well, in three dimensions. The protein is cleaved further still, Shaw said, and by the time the process is complete, enamel is 99.9 percent crystal and no protein.