4 Mimicking Enzymes with Antibodies Scheme 4.3 Antibody 39-A11 catalyzes a DielsAlder reaction between an electron-rich acyclic diene (8) and an N-aryl maleimide (9). It was elicited with the bicyclo[2.2.2]octene hapten 7. The ethano bridge locks the cyclohexene ring into the requisite boat conformation but has no counterpart in the substrates or transition state. hexachloronorbornene hapten was achieved by introducing two rare somatic mutations into the germline scaffold, one at position L89 in the light chain (SerL89Phe) and another at H47 in the heavy chain (TrpH47Leu) . These seemingly subtle substitutions, coupled with minor adjustments in CDR H3 residues, were sufficient to convert a relatively non-specific, catalytically inactive germline antibody into a potent catalyst.

This same family of germline antibodies has also given rise to a second DielsAlderase, antibody 39-A11. In this case, the substituted bicyclo[2.2.2]octene derivative 7 served as hapten . This antibody accelerates the reaction between an electron-rich acyclic diene (8) and an N-aryl maleimide (9) to give a cyclohexene derivative (10) (Scheme 4.3), but catalysis is relatively inefficient as judged by an EM of 0.35 M. Examination of the 39-A11 structure points to an explanation for the low efficiency: the large bicyclooctene unit used to mimic the reacting [4+2] system is relatively poorly packed by the protein . Most binding interactions in the complex are directed instead to the aryl group shared by the substrate and transition state. The binding pocket in which the cycloaddition takes place is simply too large to restrict all the degrees of freedom available to the diene substrate, so low rates result. By improving packing interactions with the cycloaddition transition

These examples illustrate the important interplay between binding energy and catalysis, underscoring the utility of complementary packing interactions at the transition state. For 1E9, the mechanistic information contained in the transition state analog is effectively mirrored in the induced binding pocket, allowing efficient catalysis. Even so, 1E9 is not an evolutionary optimum with respect to activity. Despite the seemingly

4.5 General AcidBase Catalysis perfect shape complementarity between protein and transition state analog, replacement of MetH100b at the floor of the pocket by phenylalanine increased the affinity for 3 by a factor of two and augmented kcat 7-fold . Even higher activity is likely to be attainable if the affinity of the antibody for the transition state (analog) can be further increased.