The lipid A biosynthesis pathway is essential in Pseudomonas aeruginosa. LpxA and LpxD are the first and third enzymes in this pathway respectively, and are regarded as promising antibiotic targets. The unique structural similarities between these two enzymes make them suitable targets for dual-binding inhibitors, a characteristic that would decrease the likelihood of mutational resistance and increase cell-based activity. We report the discovery of multiple small molecule ligands that bind to P. aeruginosa LpxA and LpxD, including dual-binding ligands. Binding poses were determined for select compounds by X-ray crystallography. The new structures reveal a previously uncharacterized magnesium ion residing at the core of the LpxD trimer. In addition, ligand binding in the LpxD active site resulted in conformational changes in the distal C-terminal helix-bundle, which forms extensive contacts with acyl carrier protein (ACP) during catalysis. These ligand-dependent conformational changes suggest a potential allosteric influence of reaction intermediates on ACP binding, and vice versa. Taken together, the novel small molecule ligands and their crystal structures provide new chemical scaffolds for ligand discovery targeting lipid A biosynthesis, while revealing structural features of interest for future investigation of LpxD function.

Carbapenem-hydrolyzing class D β-lactamases (CHDLs) are a diverse family of enzymes that are rapidly becoming the predominant cause of bacterial resistance against βlactam antibiotics in many regions of the world. OXA-48, an atypical member of CHDLs, is one of the most frequently observed in the clinic and exhibits a unique substrate profile. We applied Xray crystallography to OXA-48 complexes with multiple β-lactam antibiotics to elucidate this enzyme's carbapenemase activity and its preference of imipenem over meropenem and other substrates such as cefotaxime. In particular, we obtained acyl-enzyme complexes of OXA-48 with imipenem, meropenem, faropenem, cefotaxime, and cefoxitin, and a product complex with imipenem. Importantly, the product complex captures a key reaction milestone with the newly generated carboxylate group still in the oxyanion hole, and represents the first such complex with a wild-type serine β-lactamase. A potential hydrogen bond is observed between the two carboxylate groups from the product and the carbamylated Lys73, representing the stage immediately after the breakage of the acyl-enzyme bond where the product carboxylate would be neutral. The placement of the product carboxylate also illustrates the approximate transient location of the deacylation water that has long eluded structural characterization in class D β-lactamases. Additionally, comparing the product complex with the acyl-enzyme intermediates provides new insights into the various mechanisms by which specific side chain groups hinder the access of the deacylation water to the acyl-enzyme linkage, especially in meropenem. Taken together, these data offer valuable information on the substrate specificity of OXA-48 and the catalytic mechanism of CHDLs.

Gram-negative pathogens expressing serine β-lactamases (SBLs) and metallo-β-lactamases (MBLs), especially those with carbapenemase activity, threaten the clinical utility of almost all β-lactam antibiotics. Here we describe the discovery of a heteroaryl phosphonate scaffold that exhibits noncovalent cross-class inhibition of representative carbapenemases, specifically the SBL KPC-2 and the MBLs NDM-1 and VIM-2. The most potent lead, compound 16, exhibited low nM to low μM inhibition of KPC-2, NDM-1, and VIM-2. Compound 16 potentiated imipenem efficacy against resistant clinical and laboratory bacterial strains expressing carbapenemases while showing some cytotoxicity toward human HEK293T cells only at concentrations above 100 μg/mL. Complex structures with KPC-2, NDM-1, and VIM-2 demonstrate how these inhibitors achieve high binding affinity to both enzyme classes. These findings provide a structurally and mechanistically new scaffold for drug discovery targeting multidrug resistant Gram-negative pathogens and more generally highlight the active site features of carbapenemases that can be leveraged for lead discovery.

Serine and metallo-carbapenemases are a serious health concern due to their capability to hydrolyze nearly all β-lactam antibiotics. However, the molecular basis for their unique broad-spectrum substrate profile is poorly understood, particularly for serine carbapenemases, such as KPC-2. Using substrates and newly identified small molecules, we compared the ligand binding properties of KPC-2 with the noncarbapenemase CTX-M-14, both of which are Class A βlactamases with highly similar active sites. Notably, compared to CTX-M-14, KPC-2 was more potently inhibited by hydrolyzed β-lactam products (product inhibition), as well as by a series of novel tetrazole-based inhibitors selected from molecular docking against CTX-M-14. Together with complex crystal structures, these data suggest that the KPC-2 active site has an enhanced ability to form favorable interactions with substrates and small molecule ligands due to its increased hydrophobicity and flexibility. Such properties are even more pronounced in metallo-carbapenemases, such as NDM-1, which was also inhibited by some of the novel tetrazole compounds, including one displaying comparable low μM affinities against both KPC-2 and NDM-1. Our results suggest that carbapenemase activity confers an evolutionary advantage on producers via a broad β-lactam substrate scope but also a mechanistic Achilles' heel that can be exploited for new inhibitor discovery. The complex structures demonstrate, for the first time, how noncovalent inhibitors can be engineered to simultaneously target both serine and metallo-carbapenemases. Despite the relatively modest activity of the current compounds, these studies also demonstrate that hydrolyzed products and tetrazole-based chemotypes can provide valuable starting points for broad-spectrum inhibitor discovery against carbapenemases.