Randomly induced point mutations provide an effective strategy to recognize and characterize functionally significant features of cloned DNAs. When coupled with sensitive assays for gene function, random mutagenesis permits one to resolve active DNA motifs within a gene’s regulatory and protein-coding gene domains. When gene sequences of interest are short, it is efficient and cost-effective to generate comprehensive libraries of random point mutations by chemical synthesis of oligonucleotides from appropriately “doped” mixtures of DNA substrates (1 –5 ). However, when the target DNAs are long (>50 bp), chemical synthesis of randomly mutated sequences is expensive as well as time-consuming and depends on extensive use of costly instruments and reagents.
