J. Am. Chem. Soc. 134, 20513–20520 (2012)

Celiac disease is caused by an inflammatory response to oligopeptides that remain after the incomplete digestion of the gluten component α-gliadin. As a potential treatment strategy for this condition is to enhance degradation of these ProGln-rich peptides, previous research has focused on engineering proteases specific for ProGln dipeptides to function in the stomach, that is, under highly acidic conditions. Gordon et al. now describe an alternative approach: re-engineering the substrate specificity of an endopeptidase known to be active at low pH. To accomplish this, the authors first searched for enzymes with available crystal structures that functioned in a defined pH range at biologically relevant temperatures. They identified kumamolisin-As from Alicyclobacillus sendaiensis as a strong candidate, as the enzyme is specific for proline at the P2 site and does process substrates—though poorly—with glutamine at the P1 site. To enhance specificity for glutamine, the authors computationally varied residues within 8 Å of glutamine modeled in the active site to find energetically favorable solutions. Expression and activity assays of 261 of the resultant enzymes yielded an improved construct, KumaMax, which was ∼120-fold more active than the wild-type protease. KumaMax was also proteolytically stable, specific for the ProGln motif compared to wild-type substrates and capable of degrading a longer biologically relevant oligopeptide with a half-life of 8.5 min, suggesting its potential utility in the further development of oral enzyme therapies for Celiac disease.