The evolution of multi cellular organism is guided by changes on the molecular level in individual cells influenced by the underlying protein interaction network. During the course of evolution this intra-cellular network increases in complexity and flexibility and is responsible for complex environmental responses like the regulation of growth, signal processing, cell polarity and replication. As a key element of such networks, PDZ-domains (PSD-95/Discs-large/ZO-1) bind specifically to the C-terminus of proteins and are believed to originate from common progenitors resulting in a functionally diverse protein-family. In our hypothesis we assume that a common progenitor in the PDZ-family existed and finally gave rise to the complex pattern of PDZ specificities we observe today. In the present work we report that the Erbin-PDZ can indeed be altered to bind specifically to a multitude of different peptides providing proof that such a hypothetical progenitor for the PDZ-domain family exists. To probe our hypothesis we constructed a protein-library directed to 10 positions in the binding site of the Erbin-PDZ and isolated 237 unique domains not biased for functionality i.e. peptide binding. We asked then the question how many of these domains, carrying in average 7.7 mutations compared to the wt-sequence, are still able to recognize a C-terminal peptide. Surprisingly we found that 61 domains, one fourth, are still binding to peptides however with completely different specificities compared to the wt-Erbin-PDZ. Most of the specificities are not found in nature although we could also identify several peptide motifs similar to naturally occurring ones. Therefore, this study shows as well that the specificity repertoire found in nature is under-exploited demonstrating that the complexity of the intra-cellular networks established in evolution so far is not yet at maximum.

Extended Abstract Status: Not Uploaded