Abstract: The impressive efficiency and selectivity of biological catalysts has engendered a long-standing effort to understand the details of enzyme action. It is widely accepted that enzymes accelerate reactions through their steric and electronic complementarity to the reactants in the rate-determining transition states. Thus, tight binding to the transition state of a reactant (rather than to the corresponding substrate) lowers the activation energy of the reaction, providing strong catalytic activity. Debates concerning the fundamentals of enzyme catalysis continue, however, and non-natural enzyme mimics offer important additional insight in this area. Molecular structures that mimic enzymes through the design of a predetermined binding site that stabilizes the transition state of a desired reaction are invaluable in this regard. Catalytic antibodies, which can be quite active when raised against stable transition state analogues of the corresponding reaction, represent particularly successful examples. Recently, synthetic chemistry has begun to match nature's ability to produce antibody-like binding sites with high affinities for the transition state. Thus, synthetic, molecularly imprinted polymers have been engineered to provide enzyme-like specificity and activity, and they now represent a powerful tool for creating highly efficient catalysts. In this Account, we review recent efforts to develop enzyme models through the concept of transition state stabilization. In particular, models for carboxypeptidase A were prepared through the molecular imprinting of synthetic polymers. On the basis of successful experiments with phosphonic esters as templates to arrange amidinium groups in the active site, the method was further improved by combining the concept of transition state stabilization with the introduction of special catalytic moieties, such as metal ions in a defined orientation in the active site. In this way, the imprinted polymers were able to provide both an electrostatic stabilization for the transition state through the amidinium group as well as a synergism of transition state recognition and metal ion catalysis. The result was an excellent catalyst for carbonate hydrolysis. These enzyme mimics represent the most active catalysts ever prepared through the molecular imprinting strategy. Their catalytic activity, catalytic efficiency, and catalytic proficiency clearly surpass those of the corresponding catalytic antibodies. The active structures in natural enzymes evolve within soluble proteins, typically by the refining of the folding of one polypeptide chain. To incorporate these characteristics into synthetic polymers, we used the concept of transition state stabilization to develop soluble, nanosized carboxypeptidase A models using a new polymerization method we term the 'post-dilution polymerization method'. With this methodology, we were able to prepare soluble, highly cross-linked, single-molecule nanoparticles. These particles have controlled molecular weights (39 kDa, for example) and, on average, one catalytically active site per particle. Our strategies have made it possible to obtain efficient new enzyme models and further advance the structural and functional analogy with natural enzymes. Moreover, this bioinspired design based on molecular imprinting in synthetic polymers offers further support for the concept of transition state stabilization in catalysis.
Template and target information: Review - MIPs as catalysts