Distinctive attributes of beta-lactam target proteins in Acinetobacter baumannii relevant to development of new antibiotics
Han, S., Caspers, N., Zaniewski, R.P., Lacey, B.M., Tomaras, A.P., Feng, X., Geoghegan, K.F., Shanmugasundaram, V.(2011) J Am Chem Soc 133: 20536-20545
- PubMed: 22050378 
- DOI: https://doi.org/10.1021/ja208835z
- Primary Citation of Related Structures:  
3UDF, 3UDI, 3UDX, 3UE0, 3UE1, 3UE3 - PubMed Abstract: 
Multi-drug-resistant forms of the Gram-negative pathogen Acinetobacter baumannii are an emerging threat to human health and further complicate the general problem of treating serious bacterial infections. Meeting this challenge requires an improved understanding of the relationships between the structures of major therapeutic targets in this organism and the activity levels exhibited against it by different antibiotics. Here we report the first crystal structures of A. baumannii penicillin-binding proteins (PBPs) covalently inactivated by four β-lactam antibiotics. We also relate the results to kinetic, biophysical, and computational data. The structure of the class A protein PBP1a was solved in apo form and for its covalent conjugates with benzyl penicillin, imipenem, aztreonam, and the siderophore-conjugated monocarbam MC-1. It included a novel domain genetically spliced into a surface loop of the transpeptidase domain that contains three conserved loops. Also reported here is the first high-resolution structure of the A. baumannii class B enzyme PBP3 in apo form. Comparison of this structure with that of MC-1-derivatized PBP3 of Pseudomonas aeruginosa identified differences between these orthologous proteins in A. baumannii and P. aeruginosa. Thermodynamic analyses indicated that desolvation effects in the PBP3 ligand-binding sites contributed significantly to the thermal stability of the enzyme-antibiotic covalent complexes. Across a significant range of values, they correlated well with results from studies of inactivation kinetics and the protein structures. The structural, biophysical, and computational data help rationalize differences in the functional performance of antibiotics against different protein targets and can be used to guide the design of future agents.
Organizational Affiliation: 
Pfizer Worldwide Research, Eastern Point Road, Groton, Connecticut 06340, USA. seungil.han@pfizer.com