| Yeast surface display (YSD) is a powerful method for the directed evolution of proteins. A key factor in the success of directed evolution approaches is the selection conditions. To gain a better understanding of the thermodynamic factors that influence protein-protein affinity, a thermodynamic characterization of the association of a model single chain variable fragment (scFv) antibody (D1.3) and a higher affinity mutant (M3) with their antigen (henegg lysozyme, HEL) was performed. Fluorescence quenching titrations revealed that the mutant had five-fold higher affinity for HEL, but it had nearly the same enthalpy and heat capacity change upon binding, as measured by isothermal titration calorimetry. Thus, much of the binding free energy difference could be attributed to entropic effects. Fluorescence quenching with acrylamide indicated this more favorable entropy change may have resulted from a more flexible mutant-HEL complex, and thus be a configurational entropy effect. Osmotic stress experiments indicated that nearly twice as many water molecules were taken up into the M3-HEL complex upon binding as were taken up into the D1.3-lysozyme complex. A review of the literature revealed that water molecules can play a critical role in high-affinity protein interactions.; Different methodologies were explored for improving YSD affinity maturation of anti-protein antibodies, using scFv D1.3. Using equilibrium sorted YSD with the D1.3/M3 antibodies, a single-pass enrichment factor of 125-fold (+/-65-fold) was achieved, indicating excellent differentiation between clones of only slightly different affinity. Optimal equilibrium affinity screening of a randomly mutated D1.3 library was performed with and without an osmotic stressor present in the binding reactions. Osmotic stress yielded a more diverse set of mutant clones from the screen, and recombination of selected mutations using site-directed mutagenesis produced a four-fold higher affinity mutant (MEC1). MEC1 was randomly mutagenized and screened by kinetic selection, identifying a large number of improved mutants. In both screens, library size was approximately 5 x 106, and mutagenesis was by error-prone PCR of the entire scFv gene. This affinity compares favorably to the highest affinities attained previously with phage display for anti-protein antibodies, without the necessity for large libraries or site-directed saturation mutagenesis. |