Novel Small-Molecule Inhibitors of Anthrax Lethal Factor Identified by High-Throughput Screening Igor A. Schepetkin, ² Andrei I. Khlebnikov, Liliya N. Kirpotina, ² and Mark T. Quinn* Department of Veterinary Molecular Biology, Montana State UniVersity, Bozeman, Montana 59717, and Department of Chemistry, Altai State Technical UniVersity, Barnaul 656038, Russia ReceiVed May 1, 2006 Anthrax lethal factor (LF) is a key virulence factor of anthrax lethal toxin. We screened a chemolibrary of 10 000 drug-like molecules for their ability to inhibit LF and identified 18 novel small molecules with potent LF inhibitory activity. Three additional LF inhibitors were identified through further structure- activity relationship (SAR) analysis. All 21 compounds inhibited LF with an IC 50 range of 0.8 to 11 µM, utilizing mixed-mode competitive inhibition. An evaluation of inhibitory activity against a range of unrelated proteases showed relatively high specificity for LF. Furthermore, pharmacophore modeling of these compounds showed a high degree of similarity to the model published by Panchal et al. (Nat. Struct. Mol. Biol. 2004, 11, 67-72), indicating that the conformational features of these inhibitors are structurally compatible with the steric constraints of the substrate-binding pocket. These novel LF inhibitors and the structural scaffolds identified as important for inhibitory activity represent promising leads to pursue for further LF inhibitor development. Introduction Anthrax is an acute infectious disease caused by toxigenic strains of the spore-forming bacterium Bacillus anthracis. This disease occurs most commonly in agricultural regions, where it can be found in livestock and wild animals. Naturally occurring anthrax is extremely rare in humans and is primarily associated with exposure to infected animals or tissue from infected animals. 1 However, recent events have demonstrated that B. anthracis now poses a significant threat as an agent of biological warfare and terrorism, with a significant capacity to cause mortality. 2 The major virulence factor of B. anthracis is anthrax toxin, which is a binary A-B toxin comprised of protective antigen (PA a , 87.2 kDa) 1 and two enzymatic moieties, lethal factor (LF, 90.2 kDa) and edema factor (EF, 88.8 kDa). 3,4 PA binds to cell-surface receptors and is cleaved by furin-like proteases to form 63 kDa fragments (PA63) that oligomerize into homoheptameric pores and bind LF and EF. 5 Oligomerization also triggers endocytosis of the receptor-bound PA63-LF-EF complex by a clathrin-mediated process. Subsequently, LF and EF are packaged into endosomal carrier vesicles and delivered to the cytoplasm by release from late endosomes. 5 LF appears to be critical for pathogenesis, and bacterial strains lacking LF are not lethal in mice. 6 LF is a Zn 2+ -dependent endopeptidase, which specifically cleaves mitogen-activated protein kinase kinases (MAPKK) near their N-termini, thereby interfering with MAPK-dependent signaling pathways that recruit other immune cells during the response to inflammatory stress. 4 Although antibiotics are effective in clearing B. anthracis from the organism, high levels of the toxin may remain in circulation for several days. Thus, combination therapies of antibiotics and toxin inhibitors have been proposed. 7 There are many potential targets for therapeutic intervention against anthrax lethal toxin (i.e., complex of PA and LF), and new strategies have been exploited on the basis of the recent understanding of the structure and function of the toxin proteins. These approaches include inhibitors of furin-related proteases to block the proteolytic activation of PA, 8 recombinant antibod- ies against PA, 9 and polyvalent inhibitors of PA-LF interac- tions. 10 Because of the key role played by LF in pathogenesis, a number of studies have also focused on the development of LF inhibitors. For example, the lethal action of anthrax toxin can be blocked by synthetic or natural substances that inhibit LF protease activity. 11-13 Peptide and small-molecule LF inhibitors have also been pursued as potential sources of new therapeutics to treat anthrax, 14-24 although relatively few potent competitive and noncompetitive LF inhibitors have been found. Many LF inhibitors have been identified by high-throughput screening (HTS) of libraries composed of a variety of synthetic and natural compounds. 14,17-19,21,22,25 Of note, Panchal et al. 14 used HTS to screen a 1900-compound chemolibrary for LF inhibitors and identified 19 compounds with IC 50 <20 µM. Using structures of six selected compounds that showed a range of LF inhibitory potency, the authors established a six-point pharmacophore model of LF inhibitors. 14 This model suggested several common features essential for optimal LF inhibitor binding and provides a rational approach for optimization of candidate small-molecule inhibitors. In the present study, we utilized HTS to screen a chemical diversity library containing 10 000 drug-like molecules to identify novel inhibitors of LF that have core structures distinct from currently known leads. We identified 21 small molecules that were potent inhibitors of LF protease activity (IC 50 values of 0.5-11 µM), many highly selective for LF. In addition, we used substructure screening, fragment-focusing, and structure- activity relationship (SAR) analyses to further probe the parent chemical library and defined at least three groups of LF inhibitors: carboxylic acid derivatives of 2-phenylfurans, N- phenyldihydropyrazoles, and N-phenylpyrroles. A compound- based pharmacophore modeling of these inhibitors showed a * To whom correspondence should be addressed. Phone: 406-994-5721. Fax: 406-994-4303. E-mail: mquinn@montana.edu. ² Montana State University. Altai State Technical University. a Abbreviations: PA, protective antigen; LF, lethal factor; EF, edema factor; HTS, high-throughput screening; SAR, structure-activity relation- ship; MMP, matrix metalloproteinase. 5232 J. Med. Chem. 2006, 49, 5232-5244 10.1021/jm0605132 CCC: $33.50 © 2006 American Chemical Society Published on Web 07/28/2006