Classifying Enzymes from Selectivity Fingerprints

Fingerprints of lipases and esterases have been recorded by using an array of chiral fluorogenic aliphatic esters of increasing chain length (C(4)-C(16)). Classification of the enzyme series was carried out with selectivity data by clustering and principal component analysis (PCA). Enzymes were classified on the basis of selectivity for chain length (C(4)-C(6) vs. C(10)-C(16)) and of middle-chain-length (C(8)-C(10)) reactivity. A minimum set of nine substrates was defined by cluster analysis of relative reactivities of each substrate for the different enzymes. This selectivity-based analysis is general. It does not require a common reference substrate to react with all enzymes or vice versa, and is independent of knowing the exact concentration of active protein in the enzyme samples.