Biocatalysis becomes more and more popular but few enzymes can directly catalyze the reaction efficiently. Thus, it is essential to modify the catalyst to satisfy the practical needs. Tyrosinases demonstrate great potential to generate l-3,4-dihydroxyphenylalanine (l-DOPA), the most promising therapeutic drug for Parkinson’s disease, using their monophenolase activity with l-tyrosine as substrate. However, the naturally discovered tyrosinases exhibit poor activities towards l-tyrosine. Here, we proposed a structure-guided mutation strategy (iterative saturation mutagenesis (ISM) based on intra- and inter-functional clusters) to redesign the tyrosinase (baTRY) from Bacillus aryabhattai into a biocatalyst for producing l-DOPA. Firstly, key residues (three clusters including (1) the sites nearby the active center, (2) substrate binding pocket, and (3) water molecule interaction domain) for baTRY’s function were identified by computational structural analysis. Secondly, ISM with the order of cluster 1 (ISM) → cluster 2 (ISM) → cluster 3 (ISM) was developed to reconstruct baTRY’s spatial structure. A highly active variant MT6 (G43R/M61H/A232C/Q214D/V217A/F197W) was created, exhibiting a 11.27-fold higher monophenolase activity than WT. Quantum mechanics and molecular dynamics simulations suggested that the improved catalytic efficiency was attributed to the reduced catalysis energy barrier, the enlarged substrate-binding pocket, and the increased interactions with the substrate and water molecules. Finally, ortho-hydroxylation efficiency of MT6 was tested with 2 mM l-tyrosine as substrate, demonstrating the l-DOPA conversion rate of 90.2% and productivity of 177.6 mg/L/h. This study provided an effective strategy for guiding the engineering of biocatalysis and expanded the applications of biocatalyst on value-added chemicals.