Improving upon nature: active site remodeling produces highly efficient aldolase activity toward hydrophobic electrophilic substrates.

Journal Article

The substrate specificity of enzymes is frequently narrow and constrained by multiple interactions, limiting the use of natural enzymes in biocatalytic applications. Aldolases have important synthetic applications, but the usefulness of these enzymes is hampered by their narrow reactivity profile with unnatural substrates. To explore the determinants of substrate selectivity and alter the specificity of Escherichia coli 2-keto-3-deoxy-6-phosphogluconate (KDPG) aldolase, we employed structure-based mutagenesis coupled with library screening of mutant enzymes localized to the bacterial periplasm. We identified two active site mutations (T161S and S184L) that work additively to enhance the substrate specificity of this aldolase to include catalysis of retro-aldol cleavage of (4S)-2-keto-4-hydroxy-4-(2'-pyridyl)butyrate (S-KHPB). These mutations improve the value of k(cat)/K(M)(S-KHPB) by >450-fold, resulting in a catalytic efficiency that is comparable to that of the wild-type enzyme with the natural substrate while retaining high stereoselectivity. Moreover, the value of k(cat)(S-KHPB) for this mutant enzyme, a parameter critical for biocatalytic applications, is 3-fold higher than the maximal value achieved by the natural aldolase with any substrate. This mutant also possesses high catalytic efficiency for the retro-aldol cleavage of the natural substrate, KDPG, and a >50-fold improved activity for cleavage of 2-keto-4-hydroxy-octonoate, a nonfunctionalized hydrophobic analogue. These data suggest a substrate binding mode that illuminates the origin of facial selectivity in aldol addition reactions catalyzed by KDPG and 2-keto-3-deoxy-6-phosphogalactonate aldolases. Furthermore, targeting mutations to the active site provides a marked improvement in substrate selectivity, demonstrating that structure-guided active site mutagenesis combined with selection techniques can efficiently identify proteins with characteristics that compare favorably to those of naturally occurring enzymes.

Full Text

Duke Authors

Cited Authors

  • Cheriyan, M; Toone, EJ; Fierke, CA

Published Date

  • February 28, 2012

Published In

Volume / Issue

  • 51 / 8

Start / End Page

  • 1658 - 1668

PubMed ID

  • 22316217

Electronic International Standard Serial Number (EISSN)

  • 1520-4995

Digital Object Identifier (DOI)

  • 10.1021/bi201899b

Language

  • eng

Conference Location

  • United States