Interaction of Alzheimer's A beta peptide with an engineered binding protein--thermodynamics and kinetics of coupled folding-binding.

The oligomerization and aggregation of the amyloid-beta (A beta) peptide, a cleavage product of the amyloid precursor protein predominantly 40 or 42 amino acids in length, has been implicated in the pathogenesis of Alzheimer's disease. The identification of A beta-binding agents, e.g., antibodies or peptides, constitutes a promising therapeutic approach. However, the amount of structural and biophysical data on the underlying A beta interactions is currently very limited. We have earlier determined the structure of A beta (1-40) in complex with the affibody protein Z(A beta 3), a selected binding protein based on a three-helix bundle scaffold (Z domain). Z(A beta 3) is a dimer of affibody subunits linked via a disulfide bridge involving a selected cysteine mutation at position 28. Z(A beta 3) binds to the central and C-terminal part of A beta (residues 17-36), which adopts a beta-hairpin conformation in the complex. Here we present a detailed biophysical analysis of the Z(A beta 3):A beta (1-40) interaction, employing NMR, circular dichroism spectroscopy, 8-anilino-1-naphthalenesulfonic acid and tyrosine fluorescence, size-exclusion chromatography, thermal denaturation profiles and isothermal titration calorimetry. We conclude that (i) free Z(A beta 3) is characterized by conformational exchange and the loss of helix 1 of the three-helix bundle scaffold; (ii) a high-energy barrier is associated with the conversion of an initial Z(A beta 3):A beta (1-40) recognition complex into the native complex structure, entailing slow binding kinetics; (iii) both A beta and Z(A beta 3) fold upon binding, which, e.g., becomes manifest in the binding thermodynamics that feature a large negative change in heat capacity; (iv) the C28-disulfide does not merely afford dimerization, but its impact on the binding interfaces of the affibody subunits and A beta is a prerequisite for tight binding. The extensive folding coupled to binding observed here likely constitutes an obligate feature of biomolecular interactions involving the central and C-terminal part of A beta. Options for improvement of Z(A beta) binding proteins are discussed.