The growth of polymers from the surface of proteins via controlled radical polymerization depends on the attachment of small molecule initiators to amino acid residues. Our ability to control and harness the power of polymer-based protein engineering is reliant on the accuracy of prediction where and how fast atom transfer radical polymerization (ATRP) initiators will react with a protein surface. We performed a systematic characterization of the reaction between a bromine-functionalized N-hydroxysuccinimide amine-reactive ATRP initiator and the amino groups in lysozyme and chymotrypsin. The tertiary structures of the proteins were used to predict computationally α-amino group and lysine side-chain accessibility and analyze the chemical and structural environment of the amino groups. To predict reactivity from accessibility calculations, a probe radius that resembled the size of the initiator molecule was used. Experimental data showed that the rate of initiator-protein modification at each amine site was related to surface accessibility but not the pKa of amino groups. Further refinements of the prediction of where the initiator modified the protein and in what sequence were achieved by considering the local environment of each amino group.
Keywords: ATRP; amines; initiator; modification; prediction; protein; reactivity; structure-based.