(660d) Combinatorial Pairwise Assembly Efficiently Generates High Affinity Binders and Enables a “Mix-and-Read” Detection Scheme

We show that a combinatorial library constructed by random pairwise assembly of low affinity binders can efficiently generate binders with increased affinity. Such a library based on the Sso7d scaffold, from a pool of low affinity binders subjected to random mutagenesis, contained putative high affinity clones for a model target (lysozyme) at higher frequency than a library of monovalent mutants generated by random mutagenesis alone. Increased binding affinity was due to intramolecular avidity generated by linking binders targeting non-overlapping epitopes; individual binders of K_D ~ 1.3 μM and 250 nM produced a bivalent binder with apparent K_D ~ 2 nM. Furthermore, the bivalent protein retained thermal stability (T_M= 84.5 °C) and high recombinant expression yields in E. coli. Finally, it is important to note that efficient identification of binders with non-overlapping epitopes on the target is implicit to our strategy; this in turn can be exploited in biosensing applications. In particular, we show that the two subunit proteins comprising the high affinity bivalent binder can be used to design a simple â??mix-and-readâ? format assay for target quantification. In our scheme, binders comprising the bivalent protein are fused to two out of the three fragments of tripartite split GFP. In the presence of the target, assembly of GFP and fluorescence reconstitution occurs in concentration-dependent manner. Mix-and-read assays, wherein a signal can be obtained by simply mixing reagents with the target analyte, have significant advantages over conventional assay formats such as an enzyme-linked immunosorbent assay (ELISA), wherein time-consuming wash steps and protein immobilization are required.