Targeting Tuberculosis and Malaria through Inhibition of Enoyl Reductase
2003; Elsevier BV; Volume: 278; Issue: 23 Linguagem: Inglês
10.1074/jbc.m211968200
ISSN1083-351X
AutoresMack Kuo, Héctor R. Morbidoni, David Alland, Scott Sneddon, Brian B. Gourlie, Mark M. Staveski, Marina Leonard, Jill S. Gregory, Andrew D. Janjigian, Christopher Yee, James M. Musser, Barry N. Kreiswirth, Hiroyuki IWAMOTO, Remo Perozzo, William R. Jacobs, James C. Sacchettini, David A. Fidock,
Tópico(s)HIV/AIDS drug development and treatment
ResumoTuberculosis and malaria together result in an estimated 5 million deaths annually. The spread of multidrug resistance in the most pathogenic causative agents, Mycobacterium tuberculosis and Plasmodium falciparum, underscores the need to identify active compounds with novel inhibitory properties. Although genetically unrelated, both organisms use a type II fatty-acid synthase system. Enoyl acyl carrier protein reductase (ENR), a key type II enzyme, has been repeatedly validated as an effective antimicrobial target. Using high throughput inhibitor screens with a combinatorial library, we have identified two novel classes of compounds with activity against the M. tuberculosis and P. falciparum enzyme (referred to as InhA and PfENR, respectively). The crystal structure of InhA complexed with NAD+ and one of the inhibitors was determined to elucidate the mode of binding. Structural analysis of InhA with the broad spectrum antimicrobial triclosan revealed a unique stoichiometry where the enzyme contained either a single triclosan molecule, in a configuration typical of other bacterial ENR:triclosan structures, or harbored two triclosan molecules bound to the active site. Significantly, these compounds do not require activation and are effective against wild-type and drug-resistant strains of M. tuberculosis and P. falciparum. Moreover, they provide broader chemical diversity and elucidate key elements of inhibitor binding to InhA for subsequent chemical optimization. Tuberculosis and malaria together result in an estimated 5 million deaths annually. The spread of multidrug resistance in the most pathogenic causative agents, Mycobacterium tuberculosis and Plasmodium falciparum, underscores the need to identify active compounds with novel inhibitory properties. Although genetically unrelated, both organisms use a type II fatty-acid synthase system. Enoyl acyl carrier protein reductase (ENR), a key type II enzyme, has been repeatedly validated as an effective antimicrobial target. Using high throughput inhibitor screens with a combinatorial library, we have identified two novel classes of compounds with activity against the M. tuberculosis and P. falciparum enzyme (referred to as InhA and PfENR, respectively). The crystal structure of InhA complexed with NAD+ and one of the inhibitors was determined to elucidate the mode of binding. Structural analysis of InhA with the broad spectrum antimicrobial triclosan revealed a unique stoichiometry where the enzyme contained either a single triclosan molecule, in a configuration typical of other bacterial ENR:triclosan structures, or harbored two triclosan molecules bound to the active site. Significantly, these compounds do not require activation and are effective against wild-type and drug-resistant strains of M. tuberculosis and P. falciparum. Moreover, they provide broader chemical diversity and elucidate key elements of inhibitor binding to InhA for subsequent chemical optimization. Despite the worldwide ravages of tuberculosis and malaria, chemotherapeutic regimens against these two diseases have remained largely unchanged. For over 40 years, isoniazid (isonicotinic acid hydrazide (INH) 1The abbreviations used are: INH, isonicotinic acid hydrazide; ACP, acyl carrier protein; Me2SO, dimethyl sulfoxide; ENR, enoyl reductase; FAS-I, fatty-acid synthase type I; FAS-II, fatty-acid synthase type II; InhA, the translation product of the inhA gene; MIC, minimum inhibitory concentration; PfENR, the translation product of the pfenr gene; PIPES, 1,4-piperazinediethanesulfonic acid. ) has been utilized as a frontline agent in drug mixtures to treat Mycobacterium tuberculosis, whereas Plasmodium falciparum malaria has primarily been treated with chloroquine or pyrimethamine-sulfadoxine. The emergence of strains resistant to these and all other widely available and affordable antitubercular and antimalarial drugs, due in part to the limited number of pathways being targeted, makes it essential to identify lead compounds active against novel targets. Of particular interest from a drug discovery perspective, M. tuberculosis and P. falciparum share enzymatic components of the type II fatty acid biosynthetic pathway (FAS-II) (1Bloch K. Adv. Enzymol. Relat. Areas Mol. Biol. 1977; 45: 1-84PubMed Google Scholar, 2Waller R.F. Keeling P.J. Donald R.G. Striepen B. Handman E. Lang-Unnasch N. Cowman A.F. Besra G.S. Roos D.S. McFadden G.I. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 12352-12357Crossref PubMed Scopus (641) Google Scholar). In the case of Mycobacterium, FAS-II co-exists with the type I (FAS-I) pathway, a situation that is unique to this genus. Here FAS-I is responsible for de novo synthesis of the C16–C26 fatty acids, used for production of phospholipids and as primers for complex lipids. The FAS-II system extends these fatty acids up to C56 to make the long chain precursors required for the synthesis of cell wall-associated mycolic acids that are specific to Mycobacteria. For Plasmodium, FAS-II is the only fatty acid pathway whose enzymatic components reside in the apicoplast, a unique plastid organelle that is essential to the development of the parasite (3Fichera M.E. Roos D.S. Nature. 1997; 390: 407-409Crossref PubMed Scopus (504) Google Scholar). The apicoplast is thought to have derived from a cyanobacterial endosymbiont (4Kohler S. Delwiche C.F. Denny P.W. Tilney L.G. Webster P. Wilson R.J. Palmer J.D. Roos D.S. Science. 1997; 275: 1485-1489Crossref PubMed Scopus (615) Google Scholar, 5Palmer J.D. Delwiche C.F. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 7432-7435Crossref PubMed Scopus (88) Google Scholar, 6McFadden G.I. Waller R.F. Bioessays. 1997; 19: 1033-1040Crossref PubMed Scopus (100) Google Scholar), whose function in plasmodial parasites is still undetermined but appears to include fatty acid and isoprenoid biosynthesis (2Waller R.F. Keeling P.J. Donald R.G. Striepen B. Handman E. Lang-Unnasch N. Cowman A.F. Besra G.S. Roos D.S. McFadden G.I. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 12352-12357Crossref PubMed Scopus (641) Google Scholar, 7Jomaa H. Wiesner J. Sanderbrand S. Altincicek B. Weidemeyer C. Hintz M. Turbachova I. Eberl M. Zeidler J. Lichtenthaler H.K. Soldati D. Beck E. Science. 1999; 285: 1573-1576Crossref PubMed Scopus (1031) Google Scholar). The discovery of FAS-II in the malaria parasite was surprising as this pathway had been found previously only in plants, prokaryotes, and Archaea, where it is responsible for the synthesis of fatty acids up to C16 and C18 that are required for membrane biogenesis. No FAS-I homologs can be found in the gene sequence of P. falciparum or, for that matter, in any of the related Apicomplexan parasites. Whereas FAS-II consists of several distinct enzymes, all the enzyme activities required for FAS-I fatty acid elongation reside on a single, large multifunctional enzyme. Several observations demonstrate that the FAS-II enzymes are ideal candidates for drug discovery. First, fatty acid biosynthesis has been validated repeatedly as an effective antimicrobial target. For instance, FAS-II enzymes have been identified as the targets of several widely used antibacterials including INH, diazoborines (8Roujeinikova A. Sedelnikova S. de Boer G.J. Stuitje A.R. Slabas A.R. Rafferty J.B. Rice D.W. J. Biol. Chem. 1999; 274: 30811-30817Abstract Full Text Full Text PDF PubMed Scopus (51) Google Scholar), triclosan (9Heath R.J. Yu Y.T. Shapiro M.A. Olson E. Rock C.O. J. Biol. Chem. 1998; 273: 30316-30320Abstract Full Text Full Text PDF PubMed Scopus (323) Google Scholar), and thiolactomycin (10Hayashi T. Yamamoto O. Sasaki H. Kawaguchi A. Okazaki H. Biochem. Biophys. Res. Commun. 1983; 115: 1108-1113Crossref PubMed Scopus (82) Google Scholar, 11Jackowski S. Murphy C.M. Cronan Jr., J.E. Rock C.O. J. Biol. Chem. 1989; 264: 7624-7629Abstract Full Text PDF PubMed Google Scholar, 12Nishida I. Kawaguchi A. Yamada M. J. Biochem. (Tokyo). 1986; 99: 1447-1454Crossref PubMed Scopus (86) Google Scholar, 13Tsay J.T. Rock C.O. Jackowski S. J. Bacteriol. 1992; 174: 508-513Crossref PubMed Google Scholar). Second, FAS-II is absent in humans, superceded by a FAS-I system that is insensitive to many FAS-II inhibitors. Moreover, the cell wall/membrane component fatty acyl chains appear to be essential to microbe division and metabolism, and microbial pathogens are apparently unable to survive simply by scavenging host fatty acids. The enzyme target of most of the known FAS-II antimicrobial compounds is ENR, which catalyzes the final enzymatic step in the elongation cycle of the FAS-II pathway, converting trans-2-enoyl-ACP to acyl-ACP in a NADH-dependent reaction. The M. tuberculosis ENR, commonly known as InhA or FabI, has been shown to be the target of the first line anti-tubercular INH (14Banerjee A. Dubnau E. Quemard A. Balasubramanian V. Um K.S. Wilson T. Collins D. de Lisle G. Jacobs Jr., W.R. Science. 1994; 263: 227-230Crossref PubMed Scopus (1236) Google Scholar). This enzymatic inhibition has been shown in the fast-growing model organism Mycobacterium smegmatis to inhibit mycolic acid synthesis and induce cell lysis (15Vilcheze C. Morbidoni H.R. Weisbrod T.R. Iwamoto H. Kuo M. Sacchettini J.C. Jacobs Jr., W.R. J. Bacteriol. 2000; 182: 4059-4067Crossref PubMed Scopus (244) Google Scholar). The activity of INH is dependent on its activation to an acyl radical by a manganese-dependent reaction catalyzed by the catalase/peroxidase enzyme, KatG (16Zhang Y. Heym B. Allen B. Young D. Cole S. Nature. 1992; 358: 591-593Crossref PubMed Scopus (1097) Google Scholar). The INH acyl-radical forms a covalent adduct with NADH, producing binding interactions with InhA. Mutations in KatG have been linked to clinical resistance in ∼50% of the newly diagnosed cases of INH-resistant tuberculosis (17Escalante P. Ramaswamy S. Sanabria H. Soini H. Pan X. Valiente-Castillo O. Musser J.M. Tubercle Lung Dis. 1998; 79: 111-118Abstract Full Text PDF PubMed Scopus (82) Google Scholar). Compounds that do not require activation and that directly target ENR would circumvent this resistance mechanism. Unlike INH, the diazoborines and triclosan do not require activation, although their utility for human treatment is limited due to their respective toxicity (18Grassberger M.A. Turnowsky F. Hildebrandt J. J. Med. Chem. 1984; 27: 947-953Crossref PubMed Scopus (103) Google Scholar) and poor solubility (19McLeod R. Muench S.P. Rafferty J.B. Kyle D.E. Mui E.J. Kirisits M.J. Mack D.G. Roberts C.W. Samuel B.U. Lyons R.E. Dorris M. Milhous W.K. Rice D.W. Int. J. Parasitol. 2001; 31: 109-113Crossref PubMed Scopus (193) Google Scholar). Triclosan has been reported recently (19McLeod R. Muench S.P. Rafferty J.B. Kyle D.E. Mui E.J. Kirisits M.J. Mack D.G. Roberts C.W. Samuel B.U. Lyons R.E. Dorris M. Milhous W.K. Rice D.W. Int. J. Parasitol. 2001; 31: 109-113Crossref PubMed Scopus (193) Google Scholar, 20Surolia N. Surolia A. Nat. Med. 2001; 7: 167-173Crossref PubMed Scopus (404) Google Scholar, 21Perozzo R. Kuo M. Sidhu A.S. Valiyaveettil J.T. Bittman R. Jacobs Jr., W.R. Fidock D.A. Sacchettini J.C. J. Biol. Chem. 2002; 277: 13106-13114Abstract Full Text Full Text PDF PubMed Scopus (212) Google Scholar) to inhibit P. falciparum ENR (PfENR), confirming the FAS-II pathway as a promising antimalarial chemotherapeutic target. Inhibition with thiolactomycin, a fairly nonspecific inhibitor of the FAS-II condensing enzyme, was also reported, albeit at high (∼50 μm) concentrations (2Waller R.F. Keeling P.J. Donald R.G. Striepen B. Handman E. Lang-Unnasch N. Cowman A.F. Besra G.S. Roos D.S. McFadden G.I. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 12352-12357Crossref PubMed Scopus (641) Google Scholar). In an extension of this research, the crystal structure of PfENR complexed with triclosan was recently solved, indicating a high degree of structural conservation between this malarial enzyme and microbial and plant ENRs (21Perozzo R. Kuo M. Sidhu A.S. Valiyaveettil J.T. Bittman R. Jacobs Jr., W.R. Fidock D.A. Sacchettini J.C. J. Biol. Chem. 2002; 277: 13106-13114Abstract Full Text Full Text PDF PubMed Scopus (212) Google Scholar). In vitro whole-cell studies indicated killing of multidrug-resistant and drug-sensitive P. falciparum strains with low micromolar concentrations of triclosan and recently with triclosan analogs (19McLeod R. Muench S.P. Rafferty J.B. Kyle D.E. Mui E.J. Kirisits M.J. Mack D.G. Roberts C.W. Samuel B.U. Lyons R.E. Dorris M. Milhous W.K. Rice D.W. Int. J. Parasitol. 2001; 31: 109-113Crossref PubMed Scopus (193) Google Scholar, 20Surolia N. Surolia A. Nat. Med. 2001; 7: 167-173Crossref PubMed Scopus (404) Google Scholar, 21Perozzo R. Kuo M. Sidhu A.S. Valiyaveettil J.T. Bittman R. Jacobs Jr., W.R. Fidock D.A. Sacchettini J.C. J. Biol. Chem. 2002; 277: 13106-13114Abstract Full Text Full Text PDF PubMed Scopus (212) Google Scholar). To extend the range of pharmacophores from which to develop more potent and pharmacologically suitable ENR inhibitors, we have undertaken a high-throughput screen against the M. tuberculosis enzyme, InhA. Here we report the identification and preliminary evaluation of two novel inhibitors derived from this screen, Genz-8575 and Genz-10850, which demonstrate activity against drug-sensitive and drug-resistant strains of both M. tuberculosis and P. falciparum. We also present structural data on the binding of one of these compounds, and triclosan, to purified InhA and PfENR. These structural data provide insights into the structure-activity relationships of these compounds and serve for further chemical refinement of inhibitors of this established antimicrobial target. Analysis of [14C]NAD+Binding to InhA—Gel filtration chromatography was utilized to separate InhA-bound [14C]NAD+ from free [14C]NAD+ (22Heath R.J. Rubin J.R. Holland D.R. Zhang E. Snow M.E. Rock C.O. J. Biol. Chem. 1999; 274: 11110-11114Abstract Full Text Full Text PDF PubMed Scopus (431) Google Scholar). For this, a 1 × 12-cm Sephadex G-25 column was equilibrated in 0.1 m PIPES, pH 6.8, at 25 °C. Samples (100 μl) containing combinations of InhA, [14C]NAD+, and triclosan were then applied. The column was eluted with the same buffer, and fractions were collected. Fractions were counted in ScintiVerse scintillation solution. Crystallization and Data Collection—Recombinant InhA was produced, purified, and crystallized using methods described previously (23Dessen A. Quemard A. Blanchard J.S. Jacobs Jr., W.R. Sacchettini J.C. Science. 1995; 267: 1638-1641Crossref PubMed Scopus (402) Google Scholar). Hanging drop methods were utilized to co-crystallize InhA·inhibitor complexes. For the InhA·Genz-10850 complex, the product analog was solubilized in 100% Me2SO and added dropwise at room temperature to a dilute InhA solution containing NADH. The final molar ratios of the mixture were 1 InhA active site to 100 NADH to 1.5 Genz-10850, in 1% (v/v) Me2SO. The mixture was concentrated to 10 mg/ml, and the crystals were produced in wells containing 12% polyethylene glycol 3350, 150 mm ammonium acetate, and 100 mm [(carbamoylmethyl)imino]diacetic acid, pH 6.8, using the vapor diffusion method. The crystals were of the space group C2 with unit cell dimensions a = 101.01 Å, b = 83.31 Å, c = 193.07 Å (α = 90°, β = 95.21°, γ = 90°) and contained 6 molecules per asymmetric unit. X-ray diffraction data of the InhA:Genz-10850 crystal was collected using a MacScience DIP2030 image plate detector coupled to a Rigaku x-ray generator utilizing a copper rotating anode (CuKα, λ = 1.54 Å). The detector was placed 200 mm from the crystal with no offset in the 2θ angle. Each data frame was exposed for 10 min and consisted of a 1.5° rotation of the crystal. For the InhA·triclosan complex, crystals bound to triclosan were formed in the I212121 space group, unit cell a = 94.8 Å, b = 103.9 Å, c = 189.6 Å, α = β = γ = 90°, via hanging drop vapor diffusion experiments, with 100 mm Tris, pH 8.0, in the well solution. InhA was concentrated to 5 mg/ml and crystallized upon the introduction of triclosan to a solution with a 1:2:1 stoichiometric ratio of InhA, NAD+, and triclosan. Increasing the concentration of triclosan to yield a 1:2:2 (or higher) stoichiometric ratio resulted in microcrystals that did not diffract. Data were collected to 2.6 Å resolution at 120 K. X-ray diffraction data of the InhA:triclosan crystal were collected at the Advanced Photon Source (Beamline 14-BM-C) at a wavelength of 1.0 Å. The detector was placed 200 mm from the crystal. Each data frame was exposed for 30 s and consisted of a 1.0° rotation of the crystal. All images were autoindexed, integrated, and scaled together using the DENZO and SCALEPACK software packages (24Otwinowski Z. Minor W. Methods Enzymol. 1997; 276: 307-326Crossref PubMed Scopus (38617) Google Scholar). Structure Determination and Refinement—Initial phases were obtained using molecular replacement, with a single subunit of InhA derived from the hexagonal crystal form of InhA (Protein Data Bank code 1ENY) as a search model. Molecular replacement solutions for the translation and rotation function were obtained from Crystallography and NMR software (CNS) (25Brunger A.T. Adams P.D. Clore G.M. DeLano W.L. Gros P. Grosse-Kunstleve R.W. Jiang J.S. Kuszewski J. Nilges M. Pannu N.S. Read R.J. Rice L.M. Simonson T. Warren G.L. Acta Crystallogr. Sect. D Biol. Crystallogr. 1998; 54: 905-921Crossref PubMed Scopus (16979) Google Scholar). The CNS software package was utilized for molecular replacement, rigid-body refinement, simulated annealing, minimization, and B-factor refinement. The nucleotide cofactor was easily identified in a symmetry-averaged difference Fourier (Fo - Fc) electron density map, verifying the correctness of the molecular replacement solution. Non-crystallographic symmetry restraints were applied after rigid body refinement and initial simulated annealing. NAD+, triclosan, and water molecules were added during several cycles of minimization followed by B-factor refinement, resulting in an Rfactor of 19% and an Rfree of 28% for the InhA·Genz-10850 complex, and an Rfactor of 22.5% and an Rfree of 28% for the InhA·triclosan complex. Non-crystallographic restraints were removed when significant differences were observed between the individual subunits present in the asymmetric unit. Inspection of Fo - Fc electron density maps calculated from an InhA:NADH model revealed significant additional density in the active sites of some of the subunits in the asymmetric unit. For the InhA·Genz-10850 complex, four of the six subunits contained product analog, whereas the remaining subunits lacked bound inhibitor. For the InhA·triclosan complex, two molecules of triclosan were clearly identified and placed in one subunit of the enzyme, whereas the other subunit contained only one molecule of triclosan. The differences may be related to negative cooperativity (23Dessen A. Quemard A. Blanchard J.S. Jacobs Jr., W.R. Sacchettini J.C. Science. 1995; 267: 1638-1641Crossref PubMed Scopus (402) Google Scholar), which may translate into differences in inhibitor occupancy. Data collection statistics are presented in Table I.Table IData collection and refinement statisticsData collectionInhA:Genz-10850InhA:TriclosanMaximum resolution (Å)2.72.6Space groupC2I212121a (Å)101.094.8b (Å)83.3103.9c (Å)193.1189.6α (°)9090β (°)95.290γ (°)9090Unique reflections38,31327,704Rsym (%)aRsym = ΣhΣi|Ihi— |/ΣhΣiIhi where Ihi is the intensity of observation I of reflection h.8.58.0Completeness (%)85.295.1Redundancy2.94.5I/σ13.210.3Refinement statisticsResolution range (Å)30-2.730-2.6No. reflections38,31327,704No. atoms/subunitProtein11,96411,964NAD+4444Ligand(s)3017/34Rcryst (%)bRcryst = Σh||Fobs| —|Fcalc||/Σ|Fobs| where Fobs and Fcalc are observed and calculated structure factors, respectively.19.022.5Rfree (%)cRfree was calculated on 10% of the data omitted at random.28.828.2Average B-factors (Å2)Protein31.626.5NAD+/NADH44.145.8Inhibitors47.738/52a Rsym = ΣhΣi|Ihi— |/ΣhΣiIhi where Ihi is the intensity of observation I of reflection h.b Rcryst = Σh||Fobs| —|Fcalc||/Σ|Fobs| where Fobs and Fcalc are observed and calculated structure factors, respectively.c Rfree was calculated on 10% of the data omitted at random. Open table in a new tab High Throughput Screening, Synthesis of Inhibitors, and KiMeasurements—The high throughput screen measured the NADH-dependent catalysis of an octenoyl (C8:1 Δ2)-CoA substrate as a decrease in 340 nm absorbance resulting from conversion of NADH to NAD+. This screen was run using samples synthesized as mixtures of up to 100 compounds. For the mixture containing Genz-10850, indole-5 carboxylic acid was reacted with 86 different amines in equimolar and limiting concentration. This mixture showed 40% inhibition when tested at a total concentration of 40 μm. The 86 compounds in this mixture were then individually synthesized and tested. The most potent analogs from this mixture were amides of indole-5-carboxylic acid and 4-aryl-substituted piperazines. Subsequent synthesis of an array of piperazine analogs resulted in Genz-10850, which showed potent inhibition of InhA with an IC50 value of 0.15 μm (Table II). In the case of Genz-8575 (Table III), a mixture of 13 malondialdehydes was reacted with an excess of (4-trifluoromethyl-pyrimidin-2-yl)-hydrazine. The resulting mixture had an inhibition of 38% against InhA at 40 μm. Deconvolution of this mixture yielded several active pyrazoles, the best of which (Genz-5542, 2-[3-(4-chloro-2-nitrophenyl)-pyrazol-1-yl]-4-trifluoromethyl-pyrimidine) showed 82% inhibition at 40 μm. Several analogs of this compound were prepared, among them Genz-8575, the result of replacing the chlorine atom with a nitro group. Genz-8575 displayed 91% inhibition at 40 μm against InhA (Table III). Substrate and co-enzyme concentrations in these screens were 250 and 100 μm, respectively. IC50 values were determined by measuring the initial velocity over a broad range of inhibitor concentrations, plotting the fractional activity as a function of the log of the inhibitor concentration, and curve-fitting with a sigmoidal function.Table IIGenz-10850 structure-activity relationship data Open table in a new tab Table IIIGenz-8575 structure-activity relationship dataView Large Image Figure ViewerDownload Hi-res image Download (PPT) Open table in a new tab Inhibition constants were determined under conditions of saturating substrate (200 μm substrate and 600 μm NADH) and variable inhibitor concentrations. Ki values were determined from the x intercept of a Dixon plot, assuming uncompetitive inhibition. For triclosan binding of InhA, we found a Ki of 8.5 μm, which differs from an earlier measurement of 0.22 μm (26Parikh S.L. Xiao G. Tonge P.J. Biochemistry. 2000; 39: 7645-7650Crossref PubMed Scopus (215) Google Scholar). These discrepant results may in part be due to differences in the assay conditions, including the use of different chain length acyl-CoA substrates (C12:1 CoA versus C8:1 CoA). Minimal Inhibitory Concentration Determination—Serial 10-fold dilutions of each M. tuberculosis culture were prepared in 7H9 medium (Difco) containing 0.05% glycerol, 0.02% Tween 80, and 10% oleate-albumin dextrose complex/saline. Each dilution was then plated on Middlebrook 7H10 medium containing 0.05% glycerol, and 10% oleate-albumin dextrose. Test compounds were added across a range of concentrations. The MIC99 value was defined as the lowest concentration of the test compound that resulted in <1% growth as compared with drug-free controls. Mycobacterial Growth Inhibition Experiments—Klebsiella pneumoniae (ATCC-12658), Pseudomonas aeruginosa (ATCC-27853), Streptococcus pneumoniae (ATCC-33400), Staphylococcus aureus (ATCC-29213), Candida albicans (ATCC-10231), and Aspergillus niger (ATCC-16404) were obtained from the ATCC. To determine bacterial growth inhibition, fresh cultures were grown at 37 °C in TB broth to an optical density of 0.2 at 550 nm. By using 96-well plates, 100 μl of this culture was added to 100 μl of TB broth containing compounds dissolved in 2 μl of Me2SO. When the optical density of the control wells reached 0.9, the optical density of experimental wells was determined and used to derive percent inhibition of growth. Plasmodial Growth Inhibition Assays—P. falciparum asexual blood-stage parasites were assayed in vitro in RPMI 1640-based culture medium containing 2.5 mg/ml hypoxanthine and 0.5% albumax (In-vitrogen). Parasite cultures were then exposed to serial drug dilutions over a 72 h period, with addition of tritiated hypoxanthine (7.5 μCi/ml) after 48 h (27Fidock D.A. Nomura T. Wellems T.E. Mol. Pharmacol. 1998; 54: 1140-1147Crossref PubMed Scopus (142) Google Scholar). After accounting for background counts/min with uninfected red blood cell controls, the percentage reduction in tritiated hypoxanthine equaled 100 ×(mean cpm of samples without drug - mean cpm of drug dilution samples)/(mean cpm of samples without drug). This measurement represents a standard surrogate marker of growth inhibition. The IC50 and IC90 values were calculated by curve fitting and regression analyses. Characteristics of Triclosan Binding to InhA—To investigate the mode of triclosan binding to InhA, we purified and expressed this enzyme, performed initial velocity experiments, and measured the inhibition constants for triclosan. These experiments showed that triclosan binding was reversible and promoted by NAD+ binding, suggesting that triclosan was uncompetitive with respect to NAD+ and competed with the fatty acyl substrate. Previous studies with Escherichia coli FabI have used gel filtration experiments to investigate whether triclosan is a slow binding inhibitor (22Heath R.J. Rubin J.R. Holland D.R. Zhang E. Snow M.E. Rock C.O. J. Biol. Chem. 1999; 274: 11110-11114Abstract Full Text Full Text PDF PubMed Scopus (431) Google Scholar). To test whether the same holds true for triclosan inhibition of InhA, we performed similar experiments, testing [14C]NAD+, InhA, and triclosan in co-elution assays (Fig. 1). These showed that [14C]NAD+ co-eluted with InhA only in the presence of triclosan. In the absence of triclosan, [14C]NAD+ eluted in the included volume. These results demonstrate that the ternary inhibitory complex is in fact triclosan, NAD+, and InhA, consistent with our previously published Km data that compared NAD+ with NADH (28Quemard A. Sacchettini J.C. Dessen A. Vilcheze C. Bittman R. Jacobs Jr., W.R. Blanchard J.S. Biochemistry. 1995; 34: 8235-8241Crossref PubMed Scopus (334) Google Scholar). These data indicate that triclosan is a slow binding inhibitor of InhA, by nature of its requirement to form a complex with NAD+ in order to bind to the active site. These findings are consistent with earlier reports (9Heath R.J. Yu Y.T. Shapiro M.A. Olson E. Rock C.O. J. Biol. Chem. 1998; 273: 30316-30320Abstract Full Text Full Text PDF PubMed Scopus (323) Google Scholar) of similar Ki values of triclosan inhibition of InhA and E. coli FabI, which were in the low micromolar range. In vitro whole-cell assays demonstrated good activity of triclosan against INH-resistant and -sensitive M. tuberculosis strains (Table IV). Consistent with triclosan acting without the requirement for KatG-mediated activation, this compound was as effective against a strain carrying the KatG mutation S315T (conferring high-level INH resistance) as it was against wild-type strains. When tested against a low level, INH-resistant line containing an inhA promoter mutation leading to enzyme overexpression, triclosan activity was partially compromised at 35 μm; however, this compound remained fully active at 70 μm. These data indicated that triclosan acts against InhA, as has now been confirmed by structural analysis (see below).Table IVTriclosan is active against isoniazid-susceptible and -resistant M. tuberculosisStrainMutationIsoniazid MICaMIC, minimal inhibitory concentration, expressed in μM.Triclosan MICM305Wild type1.520M306Wild type1.525M290inhA promoter7.560M307bThis strain is resistant to isoniazid, rifampicin, pyrazinamide, ethambutol, streptomycin, and kanamycin.katG S315T20020a MIC, minimal inhibitory concentration, expressed in μM.b This strain is resistant to isoniazid, rifampicin, pyrazinamide, ethambutol, streptomycin, and kanamycin. Open table in a new tab Development of Chemically Tractable Inhibitors—To identify additional lead compounds that act directly on M. tuberculosis InhA without the requirement for activation, we performed a high-throughput screen of ∼500,000 compounds from a combinatorial library, using purified InhA as a target. This screen resulted in the identification of Genz-8575 and Genz-10850, which represent two new classes of InhA inhibitors (Table II and Table III; Fig. 2). In testing a series of about 300 Genz-10850 analogs, a structure-activity relation emerged whereby substitution was not allowed at carbon positions 2 or 3 of the piperazine ring (suggesting a steric clash with this part of the molecule) and where polar substitutions were allowed at the 2-position of the fluorenyl group (suggesting that this site is exposed to solvent). Alkylation or acylation of the indole nitrogen was not tolerated, in agreement with the role of this moiety as a hydrogen-bonding group. For the Genz-8575 analogs, a stringent requirement was detected at the trifluoromethylpyrimidine substituent, with more latitude available at the dinitrophenyl site. When tested against purified InhA, Genz-8575 and Genz-10850 displayed IC50 values of 2
Referência(s)