New enzymes from combinatorial library modules.

Publisher Summary This chapter describes the strategies to generate large patterned libraries from long synthetic oligonucleotides. It is generally believed that native-like proteins occur extremely infrequently in random sequence space. To explore the extent to which sequence can be varied without loss of catalytic activity, all the secondary structural elements in the homodimeric, helical bundle AroQ class chorismate mutase are replaced from Methanococcus jannaschii with modules of random sequence and selected functional variants by complementation of our chorismate mutase-deficient E. coli strain. The modules themselves can be fully randomized or biased in various ways, for example, to follow the inherent binary pattern of hydrophobic and hydrophilic residues in the parent enzyme or by using restricted sets of building blocks. The helices can be replaced individually or in combination, and separately constructed library modules can be crossed to obtain more comprehensively randomized enzymes. Strategies for optimizing the efficiency of library construction and for avoiding common artifacts, such as false positives, plasmid mixtures, and unplanned mutations are also addressed.