Consensus design as a tool for engineering repeat proteins.

Repeat proteins were first identified because of their unusual primary structure, in which short amino acid sequences, typically between 20 and 40 residues, are repeated in tandem, often many times. After identification at the sequence level, the three-dimensional structures of representatives from several classes (e.g., ankyrin, tetratricopeptide, leucine rich repeat) have been solved. The structures indeed reveal unusual, nonglobular structures, a linear "string" of the tandem motifs. Perhaps because of the large surface area that is presented as a consequence of such elongated structures, repeat domains are often involved in mediating protein-protein interactions. Here we udescribe methods of consensus-based design and engineering of repeat proteins. We pay particular attention to the attributes of repeat proteins that make them well-suited to such approaches. In addition, we discuss practical issues related to producing and characterizing such designed proteins. We use the tetratricopeptide repeat, which is well-studied in our group, to illustrate many ideas, but also draw comparisons to other work on repeat proteins, where relevant.