A study of enzyme catalysis of cycloaddition reactions

Abstract

This thesis presents an investigation into the development of enzyme catalysis of cycloaddition reactions. Tyrosyl-tRNA synthetase was initially investigated as a catalyst for the cycloaddition between azides and alkynes and nitrile oxides and alkenes. From modeling studies, substrates were proposed based on the natural substrates of tyrosyltRNA synthetase. Each substrate was appended with either an azide, alkene, alkyne or nitrile oxide moiety. Tyrosyl-tRNA synthetase was expressed in E. coli and purified by FPLC. A variety of commercial and literature based assays were attempted to determine the activity of tyrosyl-tRNA synthetase but each assay failed to give a satisfactory result. An HPLC based assay was developed that could determine the amount of AMP produced by tyrosyl-tRNA synthetase and therefore the activity of the enzyme. The assay was found to be broadly applicable with the activities of pyruvate kinase and potato apyrase also being determined. The binding affinities of the azide and alkyne substrates and their corresponding cycloadducts were determined using the HPLC assay and these data were used to plan experiments to investigate catalysis of the cycloaddtion between the azide and alkyne. Under the experimental conditions, it was found that the rate of non-enzyme catalysed cycloaddition between the azide and alkyne was too slow to be measured. The rate of reaction in the presence of tyrosyl-tRNA synthetase was also too slow to be measured, therefore, it could not be determined if catalysis was occurring in this system. Copper catalysis of the cycloaddtion reaction was attempted to increase the non-enzyme catalysed rate, however, in the presence of the enzyme the reaction was found to be inhibited. Due to the instability of the nitrile oxide and the alkene, catalysis was not investigated in this system. Also, as tyrosyl-tRNA synthetase was only accessible in small quantities, it was decided to continue investigating catalysis of cycloaddition reactions with Chymotrypsin as this enzyme could be bought in bulk and this meant the concentration of the enzyme could be increased and would allow a greater proportion of the substrates to be enzyme bound. Chymotrypsin was investigated as a catalyst for the cycloaddition between benzonitrile oxide and alkenes. Initial modeling studies provided aliphatic alkenes as the complementary substrates to benzonitrile oxide. The binding affinities of the alkenes and the benzonitrile oxide precursor benzaldoxime as well as the corresponding cycloadducts were directly measured using isothermal titration calorimetry. The binding data were used to plan the experiments to investigate catalysis of the cycloaddition between the aliphatic alkene and benzonitrile oxide by Chymotrypsin. The results showed that the reaction between the alkene and benzonitrile oxide was too slow. Although the formation of the cycloadduct was detectable, impurities present in the mixture containing Chymotrypsin made it difficult to isolate the product by HPLC. The amount of cycloadduct produced in the presence of Chymotrypsin could not be accurately determined but at most was similar to that produced in the non-enzyme catalysed reaction. The alkene substrate was then redesigned to be more reactive towards cycloaddition with benzonitrile oxide to increase the rate of reaction. The reaction rate was found to be greater than that of the previous system by a factor of approximately 105. The reaction between the activated alkene and benzonitrile oxide produced a mixture of regioisomers with the 5-substituted regioisomer being the major product. Under the chosen reaction conditions, the reaction between the alkene and benzonitrile oxide was found to be too fast with the reaction being complete after an hour. On lowering the concentrations of both reactants it was found that the reaction still did not proceed beyond half an hour and also that the yields of the cycloadducts were lower. It was determined from these results that the benzonitrile oxide was taking part in a competing reaction which was likely to be first order rather than the second order dimerisation. As the benzonitrile oxide was reacting too quickly, the reaction rate between the alkene and benzonitrile oxide could not be determined. However, an increase in the ratio of the 5- to 4-substituted cycloadducts in the presence of Chymotrypsin indicated that approximately 10% of the reaction may be occurring on the enzyme if the reaction on the enzyme is completely regioselective.