Spectroscopic Investigation of Novel New Delhi Metallo‐β‐Lactamase‐1 Inhibitors

The prevalence of antimicrobial‐resistant pathogens is an ever‐growing threat to world health, and the pipeline supplying new classes of antibiotics has begun to run dry. The β‐lactams are a class of antibiotics that revolutionized healthcare and have been a mainstay in medicine for decades. However, their efficacy towards pathogenic bacteria is threatened by the transmission and evolution of β‐lactamases, enzymes capable of inactivating this important class of antibiotics. The metallo‐β‐lactamases (MBLs) are a class of β‐lactamases of which there exist no clinically known inhibitors. Yet, they are capable of hydrolyzing almost all β‐lactams. These enzymes rely on either one or two Zn(II) ions within their active site which are necessary for catalysis. New Delhi MBL‐1 (NDM‐1) is of particular concern due to its global spread and rapid transmission between pathogens. Recently, 2,6‐dipicolinic acid (DPA), as well as several derivatives, were identified as potent inhibitors of the MBLs. Here we report the spectroscopic techniques used to elucidate the mode of inhibition, which aided in the discovery process. This investigation aimed to fit the mode of inhibition of NDM‐1 into one of two models: through metal scavenging, or through competitive inhibition. EDTA, a well‐known metal sequestering agent, was selected as a standard for interpreting spectroscopic data of an inhibitor fitting the first model. Alternatively, L‐captopril, a known competitive inhibitor of the MBLs, was selected for the second model. Using 1 H NMR, CW‐EPR, and equilibrium dialysis coupled ICP‐AES on either the di‐Zn(II) or di‐Co(II) metalloforms of NDM‐1, it was discovered that DPA best fit a model of non‐selective metal scavenging. However, the data suggest that several of its derivatives form a stable ternary complex between NDM‐1:active site ions:inhibitor in a concentration dependent manner. At lower concentrations, closer to that which would be found in a cell, there is little evidence of metal scavenging effects. However, at high concentrations they better fit a metal‐sequestering model, though with a preference for targeting the Zn2 site before the Zn1 site. These spectroscopic techniques allowed for the minimization of non‐selective metal scavenging effects during the drug discovery process and may be useful for guiding future investigations into other Zn(II) metalloproteins. Support or Funding Information This work was supported in part by the National Institutes of Health (GM111926) and by the Robert A. Welch Foundation (F‐1572 to WF).