Whole-protein random mutation and substrate tunnel evolution have recently been applied to the pharmaceutically relevant esterase RhEst1 for the synthesis of a cilastatin precursor. The mutant RhEst1(M1) (=RhEst1(AI47I/V148F/G254A)) was identified from a large library consisting of 1.5 x 10(4) variants. Though the activity of this mutant was improved 5-fold, the enantioselectivity for biohydrolysis decreased at the same time. Herein a smart library (3.0 X 10(3)) focused on the cap domain of RhEst1 was constructed to improve its catalytic performance comprehensively. As a result, a variant designated as RhEst1(M2) (=RhEst1(M1-A143T)), showed a 6-fold increase in specific activity compared with the wild type. Meanwhile, the decreased enantioselectivity for enzymatic resolution was recovered to the native enzyme level. The melting temperature of RhEst1(M2) was nearly 11 degrees C higher than that of the wild type. This work provides detailed insight into the vital role of alpha/beta hydrolase cap domains in influencing all aspects of enzyme characteristics. Furthermore, the commercial resin ESR-1 with free amino groups was used for enzyme immobilization to enhance the operational performance of RhEst1(M2). No obvious activity loss was observed when the immobilized enzyme was incubated at 30 degrees C for 200 h. The immobilized enzyme could be repeatedly used for up to 20 batches, and the total turnover number (TTN) reached up to 8.0 X 10(5).