Turning an asparaginyl endopeptidase into a peptide ligase

• Main Authors: Hemu, Xinya, El Sahili, Abbas, Hu, Side, Zhang, Xiaohong, Serra, Aida, Goh, Boon Chong, Darwis, Dina Amallia, Chen, Ming Wei, Sze, Siu Kwan, Liu, Chuan-Fa, Lescar, Julien, Tam, James P.
• Other Authors: School of Biological Sciences
• Format: Article
• Language: English
• Published: 2021
• Subjects: Science; Peptide Asparaginyl Ligase (PAL); Asparaginyl Endopeptidase (AEP)
• Online Access: https://hdl.handle.net/10356/148032
• Institution: Nanyang Technological University

Butelase-1 is a peptide asparaginyl ligase (PAL) that efficiently catalyzes peptide bond formation after Asn/Asp (Asx), whereas its homologue butelase-2 is an asparaginyl endopeptidase (AEP) that catalyzes the reverse reaction, hydrolyzing Asx-peptide bonds. Since PALs and AEPs share essentially similar overall structures, we surmised that the S2 and S1′ substrate-binding pockets immediately flanking the catalytic S1 site constitute the major ligase activity determinants (LADs) that control the catalytic directionality of these enzymes. Here, we report the successful conversion of butelase-2 into butelase-1-like ligases based on the LAD hypothesis. We prepared 23 LAD mutants by mutating residues of the S2 pocket (LAD1) and/or the S1′ pocket (LAD2) of butelase-2 into homologous residues in PALs. These LAD mutants markedly diminished protease activity and increased ligase activity. In contrast, substituting 12 non-LAD residues to the corresponding residues in butelase-1 did not change their protease profiles. At physiological pH, mutations targeting both LADs resulted in shifting the catalytic directionality from 95% hydrolysis to >95% peptide ligation. This results in the engineering of efficient recombinant peptide ligases that were demonstrated to be useful for macrocyclization and site-specific labeling of bioactive peptides and proteins. Five high-resolution crystal structures of butelase-2 and its engineered mutants reveal subtle changes proximal to the catalytic S1 site that account for the reversal in enzymatic activity. Computational simulations of enzyme-substrate complexes suggest how the S-acyl intermediate is positioned at the S2 site and the substrate orientated, controlling accessibility of either water or incoming nucleophiles from the prime-side (S1′ site) of the catalytic center. Together, these features determine the catalytic directionality between hydrolysis and ligation in AEPs and PALs. Overall, this work validates the LAD hypothesis as a central guide for making peptide ligases from their corresponding widely available protease counterparts, to produce precision tools for exquisite site-specific conjugation and the biomanufacturing of biologics.