lable at ScienceDirect

European Journal of Medicinal Chemistry 90 (2015) 436e447

Contents lists avai

European Journal of Medicinal Chemistry

journal homepage: http: / /www.elsevier .com/locate/ejmech

Original article

Piperazine derivatives: Synthesis, inhibition of the Mycobacteriumtuberculosis enoyl-acyl carrier protein reductase and SAR studies

Mariane Rotta a, b, Kenia Pissinate a, b, Anne Drumond Villela a, c, Davi Fernando Back d,Luis Fernando Saraiva Macedo Timmers a, b, e, Jos�e Fernando Ruggiero Bachega a, e,Osmar Norberto de Souza a, b, e, Di�ogenes Santiago Santos a, b, *, Luiz Augusto Basso a, b, c,Pablo Machado a, b, *

a Instituto Nacional de Ciencia e Tecnologia em Tuberculose, Centro de Pesquisas em Biologia Molecular e Funcional, Pontifícia Universidade Cat�olica do RioGrande do Sul, 90619-900, Porto Alegre, Rio Grande do Sul, Brazilb Programa de P�os-graduaç~ao em Biologia Celular e Molecular, Pontifícia Universidade Cat�olica do Rio Grande do Sul, 90619-900, Porto Alegre, Rio Grandedo Sul, Brazilc Programa de P�os-Graduaç~ao em Medicina e Ciencias da Saúde, Pontifícia Universidade Cat�olica do Rio Grande do Sul, 90619-900, Porto Alegre, Rio Grandedo Sul, Brazild Departamento de Química, Laborat�orio de Materiais Inorganicos, Universidade Federal de Santa Maria, 97105-900, Santa Maria, Rio Grande do Sul, Brazile Laboratorio de Bioinform�atica, Modelagem e Simulaç~ao de Biossistemas e LABIO, Pontifícia Universidade Cat�olica do Rio Grande do Sul, 90619-900, PortoAlegre, Rio Grande do Sul, Brazil

a r t i c l e i n f o

Article history:Received 11 July 2014Received in revised form13 November 2014Accepted 18 November 2014Available online 18 November 2014

Keywords:Mycobacterium tuberculosisTuberculosisEnoyl-ACP reductaseInhA inhibitionPiperazines

Abbreviations: TB, tuberculosis; M. tuberculosis,MDR-TB, multidrug-resistant tuberculosis; HIV, humFAS, fatty acid synthase; ACP, acyl carrier protein; Mfrom M. tuberculosis; NADH, nicotinamide adenine dacid hydrazide; KatG, catalase-peroxidase; SAR, structhigh-throughput screening; Kii, inhibitory dissociatioplex; Kis, inhibitory dissociation constant for the EI citory concentration; TEA, triethylamine; HBTU, 2-(1tetramethyluronium hexafluorophosphate; DMF, NCoA, 2-trans-dodecenoyl-CoA; DMSO, dimethyl sulfoDMAP, 4-(N,N-dimethylamino)pyridine; PIPES, 1,4-pip* Corresponding authors. Instituto Nacional de Cie

culose, Centro de Pesquisas em Biologia Molecularversidade Cat�olica do Rio Grande do Sul, 90619-900Sul, Brazil.

E-mail addresses: diogenes@pucrs.br (D.S. Santo(P. Machado).

http://dx.doi.org/10.1016/j.ejmech.2014.11.0340223-5234/© 2014 Elsevier Masson SAS. All rights re

a b s t r a c t

The Mycobacterium tuberculosis NADH-dependent enoyl-acyl carrier protein reductase (MtInhA) cata-lyzes hydride transfer to long-chain enoyl thioester substrates. MtInhA is a member of the mycobacterialtype II dissociated fatty acid biosynthesis system, and is the bona fide target for isoniazid, the mostprescribed drug for tuberculosis treatment. Here, a series of piperazine derivatives was synthesized andscreened as MtInhA inhibitors, which resulted in the identification of compounds with IC50 values in thesubmicromolar range. A structureeactivity relationship (SAR) evaluation indicated the importance of thechemical environment surrounding the carbonyl group for inhibition. In addition, the structure of oneselected compound was supported by crystallographic studies, and experimental geometrical valueswere compared with semi-empirical quantum chemical calculations. Furthermore, the mode of inhibi-tion and inhibitory dissociation constants were determined for the nine most active compounds. Thesefindings suggest that these 9H-fluoren-9-yl-piperazine-containing compounds interact with MtInhA atthe enoyl thioester (2-trans-dodecenoyl-CoA) substrate binding site.

© 2014 Elsevier Masson SAS. All rights reserved.

Mycobacterium tuberculosis;an immunodeficiency virus;tInhA, enoyl-ACP reductase

inucleotide; INH, isonicotinicure activity relationship; HTS,n constant for the ESI com-omplex; MIC, minimal inhib-H-benzotriazole-1-yl)-1,1,3,3-,N-dimethylformamide; DD-xide; Km, Michaelis constant;erazinediethanesulfonic acid.ncia e Tecnologia em Tuber-e Funcional, Pontifícia Uni-Porto Alegre, Rio Grande do

s), pablo.machado@pucrs.br

served.

1. Introduction

Tuberculosis (TB) is an infectious disease primarily caused byMycobacterium tuberculosis (M. tuberculosis), and it remains a majorglobal health concern. According to theWorld Health Organization,an estimated 8.6 million new TB cases with approximately 1.3million deaths were reported in 2012 [1]. The emergence ofM. tuberculosis strains resistant to first- and second-line TB drugshas highlighted the need for novel and effective treatments [2].Additional major concerns include TB-HIV coinfection and latentTB. In 2012, approximately 1.1 million of the estimated TB cases and0.3million deaths were described to occur among peoplewhowereHIV-positive [1]. In addition, one-third of the worldwide popula-tion has been reported to be at risk for reactivation from latent TB,

M. Rotta et al. / European Journal of Medicinal Chemistry 90 (2015) 436e447 437

which presents challenges in the eradication of this disease [3].Within this context, worldwide efforts have been directed at thedevelopment of new anti-TB drugs [4]. Following more than 40years, a new TB drug named bedaquiline was approved for clinicaluse [5]. However, the restrict indications of this drug, possible sideeffects [6], and the rapid appearance of new drug-resistant TBstrains strengthen the need to continuing efforts towards thedevelopment of novel antimycobacterial compounds.

The mycobacterial type II dissociated fatty acid biosynthesissystem (FAS-II) has emerged as an attractive, validated target forthe development of novel anti-TB molecules [7]. The FAS-II systemelongates acyl fatty acid precursors yielding the long carbon chainof the meromycolate branch of mycolic acids, the hallmark ofmycobacteria [8,9]. Mycolic acids are high-molecular-weight a-alkyl, b-hydroxy fatty acids, which appear mostly as bound esters intetramycolylpentaarabinosyl cluster in the mycobacterial cell wall[10,11]. These mycolic acids have been associated withM. tuberculosis virulence [9], the ability ofM. tuberculosis to surviveand replicate inside macrophages and with the inability of manyantimycobacterial compounds to penetrate into the M. tuberculosiscytosol [8,10]. The fatty acid precursors required for M. tuberculosismycolic acid biosynthesis are synthesized by successive rounds ofelongation and reduction reactions via the type I and type II fattyacid synthase (FAS-I and FAS-II) systems to yield, respectively, thea-branch and the meromycolate chain [10,11]. Encoded by the inhAgene, M. tuberculosis enoyl-ACP reductase (MtInhA) catalyzes thefinal essential enzymatic step in fatty acid elongation in the FAS-IIpathway, converting 2-trans-enoyl-ACP to acyl-ACP via a hydridetransfer from the 4S hydrogen of NADH to the C3 position of the 2-trans-enoyl-CoA(ACP) substrate. MtInhA has been identified as theprimary target of isoniazid (INH), which is the frontline drug for TBchemotherapy, thereby validating this target for anti-TB drug dis-covery [12e14]. As a prodrug, INH requires oxidation by myco-bacterial katG-encoded catalase-peroxidase (KatG) [15,16], whichleads to the formation of the final covalent INH-NAD adduct thathas been shown to be a slow, high-affinity competitive inhibitor ofMtInhA [13,17]. Notably, mutations in the katG gene have beenlinked to clinical resistance in 25e50% of newly diagnosed cases ofINH-resistant TB [18,19]. Therefore, compounds able to directlyinhibit MtInhA, without KatG-assisted activation, could in theorytarget INH-resistant M. tuberculosis strains harboring katG genemutations. We have thus been investigating compounds bound totransition metal complexes as self-activating species in the wholemycobacterial cell context with some encouraging results [20,21].Moreover, virtual screening and pharmacophore-based approacheshave resulted in the discovery of new MtInhA inhibitors in themicromolar range [22]. In contrast to the recently reported methylthiazoles that interact withMtInhA in a “Tyr158-out” binding mode

Fig. 1. Chemical structures of piperazine derivatives identified as potent MtInhA in-hibitors through HTS.

[23], direct inhibitors such as pyrrolidine carboxamides [24] andpiperazine-indole derivatives [25] have explored polar interactionsinvolving a ribose hydroxyl, the Tyr158 hydroxyl and a hydrogenbond acceptor in the compounds. Consistent with our strategy [22],this substrateeproteineligand interaction has been observed as apharmacophoric point in virtual screening campaigns for novelMtInhA inhibitors. However, the structural and electronic re-quirements for these hydrogen bond donoreacceptor pairs havenot been extensively examined.

Therefore, in this study, we evaluated the inhibition of MtInhAby piperazine-based compounds combined with structure activityrelationship (SAR) studies. These compounds have been primarilyobtained through high-throughput screening (HTS) approachesand exhibit submicromolar inhibition of MtInhA enzyme activity(Fig. 1) [25e27]. Herein, novel piperazine derivatives were syn-thesized, and the structure of one compound was elucidated usingX-ray diffraction. In addition, the mode of inhibition, inhibitorydissociation constants (Kii and/or Kis) were determined.

2. Results and discussion

2.1. Chemistry

First, we sought to synthesize 1-(9H-fluoren-9-yl)-piperazinederivatives bearing modifications around the carbonyl hydrogenbond acceptor. Using 1-(9H-fluoren-9-yl)-piperazine (3), com-pounds 2 and 4e8 were synthesized in low to good yields usingclassical methods (Scheme 1). The amides 2 and 4aej were ob-tained through acylation reactions of 3 using synthesized orcommercially available benzoyl chlorides in the presence of trie-thylamine (TEA) as the base and dichloromethane (CH2Cl2) as thesolvent. The reaction mixture was maintained at 0 �C during theaddition of the reactants and then allowed to warm to 25 �C withsubsequent stirring for 16 h. The 9H-fluoren-9-yl-piperazines 2 and4aejwere isolated with 7e87% yield. Notably, one-pot synthesis ofbenzoyl chlorides following acylation reactions resulted in lowerproduct yields in comparison with a direct acylation protocol. Inaddition, piperazine 4k was synthesized from an active ester ofpyridine-4-carboxylic acid and 1-(9H-fluoren-9-yl)-piperazine (3).The ester was formed in the presence of 2-(1H-benzotriazole-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate (HBTU) and TEAusing N,N-dimethylformamide (DMF) as solvent. The reactionmixture was stirred at 45 �C for 4 h to afford 4k with 83% yield.

Direct sulfonylation was used for the synthesis of benzene-sulfonyl derivatives 5aeb to obtain the desired products in 61e82%yield. Additionally, to synthesize piperazine 6, the cinnamic acidwas previously prepared through condensation between benzal-dehyde and malonic acid as previously described [28]. Thereafter,the a,b-unsaturated acid was converted to the respective chlorideand reacted with 1-(9H-fluoren-9-yl)-piperazine (3) in a one-potsynthesis to afford 6 with 26% yield. The stereochemistry of thedouble bond was assigned based on the observed proton couplingconstants (J) of 15 Hz. Finally, two SN2 nucleophilic substitutionreactions yielded compounds 7 and 8. The 2-bromo-N-phenyl-acetamide precursor was synthesized as previously described [29]and was reacted with amine 3 in DMF in the presence of potassiumcarbonate at room temperature for 24 h. Piperazine 8 was syn-thesized via an identical protocol using (bromomethyl)benzene asthe alkylating agent. Products 7 and 8 were obtained with 51% and55% yield, respectively.

Compounds 10e12were synthesized (Scheme 2) to evaluate theeffect of increasing the degrees of freedom of phenyl-like groupsinserted into the 1-(9H-fluoren-9-yl) ring on MtInhA inhibition.This enhanced conformational flexibility may enable stronger hy-drophobic interactions with the enzyme or result in entropy

Scheme 1. Reactants and conditions: i) ¼ (1) Benzoic acid, SOCl2, toluene, 110 �C, 5 h; (2) TEA, CH2Cl2, 0e25 �C, 16 h. ii) ¼ Benzoyl chloride, TEA, CH2Cl2, 0e25 �C, 16 h. iii) ¼ Benzoicacid, HBTU, TEA, 25 �C, 4 h. iv) ¼ Benzenesulfonyl chloride, TEA, CH2Cl2, 0e25 �C, 16 h. v) ¼ (1) Cinnamic acid, SOCl2, toluene, 110 �C, 5 h; (2) TEA, CH2Cl2, 0e25 �C, 16 h. vi) ¼ 2-Bromo-N-phenylacetamide, K2CO3, DMF, 25 �C, 24 h. vii) ¼ (Bromomethyl)benzene, K2CO3, DMF, 25 �C, 24 h.

M. Rotta et al. / European Journal of Medicinal Chemistry 90 (2015) 436e447438

penalties in molecular recognition [30]. Compound 10 was syn-thesized through acylation of 1-(diphenylmethyl)-piperazine (9)with benzoyl chlorides. The benzoyl chlorides were obtained from acommercial source or synthesized from the desired benzoic acidsusing of thionyl chloride. Independent of the source of the acylatingagent, the reaction with amine 9 was performed in dichloro-methane using TEA as the base at 0e25 �C for 16 h to obtain 10with15e87% yield. It is noteworthy that reactions were accomplished inone-pot procedures when benzoyl chlorides were generated fromthe respective acids. Furthermore, benzenesulfonyl chloride wasreacted with amine 9 as previously described for piperazine 5 toafford compound 11 with 15% yield. Finally, piperazine 12 wasobtained after chloride generation from cinnamic acid withconcomitant acylation of 1-(diphenylmethyl)-piperazine. Com-pound 12 was isolated with 33% yield. Spectroscopic and massspectrometry analyses of all of the synthesized compounds areconsistent with the proposed chemical structures. As we sought tofirst evaluate any MtInhA inhibitory activity by the synthesizedcompounds, the reaction conditions were not yet optimized.

2.1.1. StructureThe crystal structure of 9H-fluoren-9-yl-piperazine 4d was

determined by X-ray diffraction [31]. Carbocyclic 1-(9H-fluoren-9-yl) and 4-(4-fluorobenzoyl) moieties are attached to the piperazinering (Fig. 2). The crystal data and details of the data collection andstructure refinement are summarized in Table S1 (Supportinginformation). Both the 9H-fluoren-9-yl and the 4-fluorobenzoylmoieties adopt a U-like conformation around the heterocyclicscaffold with the carbonyl group pointing out from the mean planeof the 4-fluorobenzyl substituent, as observed in the crystalstructure. The dihedral angle formed between C(24)eC(19)eC(18)eO is 44.5(2)�, whereas the angle observed for OeC(18)eN(2) is121.3�, which results in near planarity of the amide group. As ex-pected, the 9H-fluoren-9-yl ring exhibits a nearly planar structureinferred by the sum of the internal angles of the five-memberedcarbocycle [C(13)eC(12)eC(7)eC(6)eC(1)]. This sum was found tobe 539.98� for compound 4d and deviates only slightly from theideal value of 540� for a regular pentagon.

To investigate the structural requirements of pharmacologicallyactive compounds, utilization of computational methods that are

Scheme 2. Reactants and conditions: i) ¼ (1) Benzoic acid, SOCl2, toluene, 110 �C, 5 h;(2) TEA, CH2Cl2, 0e25 �C, 16 h. ii) ¼ Benzoyl chloride, TEA, CH2Cl2, 0e25 �C, 16 h.iii) ¼ Benzenesulfonyl chloride, TEA, CH2Cl2, 0e25 �C, 16 h. iv) ¼ (1) Cinnamic acid,SOCl2, toluene, 110 �C, 5 h; (2) TEA, CH2Cl2, 0e25 �C, 16 h.

Fig. 2. A view of compound 4d with atomic labeling. Hydrogen atoms are representedby circles of arbitrary radii.

Table 1Determined IC50 values of piperazine derivatives for the MtInhA-catalyzed reaction.

Comp. IC50 (mM)a Comp. IC50 (mM)a

2 0.183 (±0.005) 5b >504a 2.90 (±0.45) 6 >504b 0.222 (±0.03) 7 >504c 1.69 (±0.56) 8 >504d 0.397 (±0.04) 10a 16.03 (±2.86)4e >50 10b >504f 0.250 (±0.02) 10c 7.83 (±1.27)4g 5.99 (±0.94) 10d >504h 0.240 (±0.04) 10e >504i 0.361 (±0.04) 10f >504j 1.57 (±0.17) 11 >504k 9.77 (±1.30) 12 >505a 19.96 (±2.45)

a IC50 is the concentration of inhibitor that reduces the enzyme velocity by halfunder a specific set of reaction conditions.

M. Rotta et al. / European Journal of Medicinal Chemistry 90 (2015) 436e447 439

able to reproduce the experimental data in good agreement iscrucial. Thus, the geometrical parameters observed in piperazine4d were compared with semi-empirical AM1 [32] and PM3 [33,34]quantum chemical calculations. Incidentally, these methods havebeen used to analyze the conformational landscape of heterocycliccompounds, resulting in calculated values that are in good agree-ment with experimentally determined values [35]. The simplelinear regressions of the experimental bond lengths and bond anddihedral angles from X-ray diffraction with semi-empirical calcu-lated values exhibited good correlation coefficients. For bondlengths, both semi-empirical methods exhibited a similar correla-tion with the following experimental values: r ¼ 0.947 for PM3 andr ¼ 0.954 for AM1. These values account for, respectively, 89.8% and91.0% of the variability in the bond lengths in the 4d structure. The

bond angles calculated from semi-empirical methods were also ingood agreement with the experimentally determined bond angles,r¼ 0.935 and r¼ 0.945 for the PM3 and AM1methods, respectively.The PM3 method accounts for 87.4% of the variability in the bondangles, whereas the AM1 approach accounts for 89.3% of thisvariation. However, the correlation between experimental andcalculated values for the dihedral angles exhibited a significantdifference between the two semi-empirical calculations. For thePM3 method, r ¼ 0.891, which accounts for only 79.5% of thevariability in the dihedral angles in 4d, whereas for the AM1method, r¼ 0.988, which accounts for 97.7% of this variability. Thus,according to the magnitude of the correlation coefficients, the AM1method is more suitable than the PM3 method to estimate theexperimental geometrical parameters of the piperazines understudy. In addition, this result represents another example in whichthe crystal structure is at or close to the global minimum energyconformation.

2.2. Inhibitory activity measurements

The reduction reaction of 2-trans-dodecenoyl-CoA (DD-CoA)catalyzed by MtInhA was spectrophotometrically measured moni-toring NADH oxidation to NADþ [36,37]. The enzyme inhibitionassays were carried out in the presence of the synthesized com-pounds in a solution containing a final concentration of 1%dimethyl sulfoxide (DMSO) as the solvent. This concentration didnot interfere with the assay conditions.

The synthesized lead-like 2 was assayed and exhibited an IC50(the concentration of inhibitor that reduces the enzyme velocity byhalf) of 0.18 mM,which is 2-fold higher than the previously reportedvalue of 0.09 mM [26] (Table 1). Modification of the hydrogen bondacceptor character from a carbonyl group to a sulfonyl moiety inpiperazine 5a resulted in a compound with an IC50 value ofapproximately 20 mM. This result may be attributed to a reducedhydrogen-bond basicity of sulfonamide compared to amide groups[38]. Notwithstanding, electron-donating groups tend to improvethis basicity, compound 5b containing a methyl group at the C2position of the phenyl moiety exhibited reduced inhibitioncompared with the non-substituted counterpart (5a); IC50 > 50 mM.In an attempt to increase the hydrogen-bond basicity of thecarbonyl group and improve inhibitory activity, the a,b-unsaturatedamide 6 was synthesized and evaluated. Unfortunately, the conju-gated system was unable to inhibit MtInhA activity and exhibitedan IC50 > 50 mM. These findings suggest that steric factor wasdeterminant on electronic feature of the proposed systems.Displacement of the position of the carbonyl through insertion of amethylene group resulted in drastically reduced inhibition by

M. Rotta et al. / European Journal of Medicinal Chemistry 90 (2015) 436e447440

piperazine 7. Notably, the amide was maintained while the inser-tion of the methylene group altered the conformational possibil-ities of the carbonyl moiety. This result suggests that the position ofthe carbonyl group does not tolerate even small spatial variations.Finally, the synthesis of compound 8 lacking an amide groupafforded a piperazine that did not inhibit MtInhA at 50 mM,emphasizing the importance of this hydrogen bond acceptor forMtInhA inhibition.

An open system of the 9H-fluoren-9-yl ring was also evaluatedto assess its influence (10e12) on MtInhA enzyme inhibition. Fromthe results summarized in Table 1, substitution of the planar systemin 2 for the freely rotatable diphenylmethyl moiety in 10a produceda compound that is nearly 88-fold less potent (IC50 ¼ 16.0 mM) ascompared to compound 2. Indeed, the hydrophobic 9H-fluoren-9-yl moiety has been described as important for inhibition andmimics the fatty acid substrate within the substrate binding pocket[25]. This moiety is also present in the structure of recentlydescribedMtInhA inhibitors [39,40]. Substitution at the C4 positionof the aryl ring with a methyl group yields piperazine 10c, whichinhibitedMtInhAwith an IC50 of 7.8 mM. By contrast, substitution atthis positionwith amethoxy group (10b) greatly reduced inhibitionby this compound (IC50 > 50 mM). Unexpectedly, even bioisostericreplacement of the 4-methyl (10c) with 4-Cl (10d) yielded amolecule with an IC50 > 50 mM, indicating fine tuning of this sys-tem. Indeed, except to 10a and 10c, none of the modifications onscaffold 10 generated molecules with IC50 values lower than 50 mM.Similarly, sulfonyl (11) and cinnamic acid (12) derivatives did notinhibit MtInhA catalysis at the maximum evaluated concentrationof 50 mM.

In the next round of structural investigations, we evaluatedsubstitutions at the C4 position in the phenyl ring of the lead-likecompound 2. The results summarized in Table 1 for piperazines4aee and 4k indicated that, in general, electron-donating, hydro-phobic and non-bulky groups exhibited enhanced MtInhA inhibi-tion. These findings are in accordance with previous reports, whichhave shown that enzyme inhibition is highly dependent on thechemical characteristics and size of groups around the phenylsubstituent [26,27,41]. In addition, this solvent-exposed cavity isformed by polar and nonpolar groups and has been predicted toaccommodate small and planar chemical groups [27,41].

It is noteworthy that from both ground-state and spectroscopicperspectives, halogens have been described to exhibit electron-donating characteristics from positive mesomeric effects (Mþ)[42]. Accordingly, nitro-substituted compound 4e exhibited anIC50 > 50 mM. This result can be attributed to the ability of electronwithdrawing groups to reduce the ability of the carbonyl to act as ahydrogen bond acceptor. Between the evaluated moieties, the nitromoiety was the most electron withdrawing group evaluated at theC4 position of the phenyl ring presenting negative mesomeric andinductive effects (M� and I�). Such features may account for thereduced inhibition by compound 4e. It is important tomention thatnitro is a planar group in the same way as the indole substituentpresent in piperazine derivative 1, which was able to produce aninhibitor of MtInhA with an IC50 of 0.16 mM [25]. Therefore, for 4eactivity, the electronic effect, instead of steric effects, seems to becrucial.

Piperazine 4kwas proposed with the objective of evaluating theeffect of a compoundwith amolecular volume similar to lead-like 2but showing distinct electronic and hydrophobic characteristics.Pyridine derivative 4k showed moderate inhibition on MtInhA ac-tivity with an IC50 of 9.7 mM, which was nearly 53-fold less potentthan benzyl-substituted compound 2.

Regarding electron-donating groups, the methoxy-substitutedpiperazine 4a exhibited an IC50 of 2.9 mM, whereas the 4-methylgroup in 4b was able to inhibit MtInhA with an IC50 of 0.22 mM

(0.40 mM [26]), which is submicromolar and comparable to lead-like compound 2. Once more, classical bioisosteric replacement ofa methyl group with a chloro substituent yielded compound 4c,which was approximately 10-fold less effective for MtInhA inhibi-tion than the methyl derivative 4b. By contrast, fluoro-substituted4d maintained inhibition in the submicromolar range, with anIC50 of 0.40 mM. This result shows that piperazine 4b (4-methyl) isnearly 2-fold more potent than 4d (4-fluoro), based on themagnitude of the IC50 values. Although 4-fluoro and 4-methyl shareidentical mesomeric effects, their inductive effects are opposite:fluoro substitution reduces the electron density along the sigmabonds, whereas the methyl group increases this electron density.Thus, 4-methyl (4b) can increase the basicity of the carbonyl groupand maintain a more hydrophobic environment around the benzylring, leading to enhanced inhibition of MtInhA compared to the 4-fluoro-substituted compound (4d).

In evaluating the position of the methyl in the benzyl ring, theinhibition data indicated a preference for the C4 and C3 positions incomparison with 2-substituted derivatives. It is noteworthy thathydrogens at the C2 and C3 positions of the phenyl ring have beendescribed as having weak hydrogen bonds with NADH [41]. Piper-azine 4f (3-methyl) exhibited an IC50 of 0.25 mM, whereas com-pound 4g (2-methyl) inhibited MtInhA with an IC50 of 6.0 mM.Considering the fluorinatedmolecules 4d and 4hei, the C3 positionwas slightly preferred for enhanced inhibitory potency in com-parison with the C4 and C2 positions, as 4h (3-fluoro) and 4i (2-fluoro) exhibited IC50 values of 0.24 mM and 0.36 mM, respec-tively. Finally, 2,5-difluoro-substituted 4j exhibited reduced inhib-itory potency on the catalytic activity of MtInhA, with an IC50 of1.6 mM. According to the structures of 4g and 4i studied by dockingsimulations, the presence of the methyl group at the C2 position ofthe aryl moiety causes steric hindrance due to the restricted spacein the cavity formed by Met103, Tyr158, and Met161 residues ofMtInhA (Fig. 3). It is noteworthy that halogen substituents typicallyform multiple interactions, as well as participate in desolvationeffects in ligandeprotein complexes [43]. Based on the observeddistances and the sum of the van der Walls radii, the fluoro groupdoes not appear to form hydrogen bonds with MtInhA or NADH.Therefore, steric hindrance, halogen-mediated interactions, anddesolvation effects may account for the better activity of compound4i compared to 4g.

2.2.1. Determination of Ki values for select MtInhA inhibitorsUsing a threshold of 3 mM for the IC50 values, nine compounds

(2, 4aed, 4f and 4hej) were selected for additional studies on themode of inhibition (uncompetitive, noncompetitive or competitive)of MtInhA. It is noteworthy that none of these molecules exhibitedtime-dependent inhibitory activity up to 30 min of preincubationwith MtInhA (data not shown), suggesting rapid equilibrium pro-cesses. Accordingly, the mode of inhibition could be determinedfrom LineweavereBurk plots and data fitting to appropriate equa-tions to give values for the inhibition constants (Kii and/or Kis) [44](Table 2). When the substrate DD-CoAwas fixed at a non-saturatingconcentration and NADH was varied in the presence of fixed-varying concentrations of the compounds, the double reciprocalplots yielded parallel lines (Supporting information). These resultsindicated that all of the evaluated compounds are uncompetitiveinhibitors with respect to NADH. Fitting the data to Eq. (2) yieldedthe values for the ternary complex inhibition constants (Kii), withvalues ranging from 0.147 to 1.95 mM (Table 2). The inhibitoryprofile suggests that these inhibitors exhibit low affinity to the freeenzyme and a nearly complete binding preference for the enzy-meesubstrate (ES) complex to yield an inactive enzy-meesubstrateeinhibitor (ESI) complex [44,45]. Moreover, thebinding of these compounds to the MtInhA:NADH binary complex

Fig. 3. Best binding modes of inhibitor compounds 4g and 4i (represented by ball and stick models embedded in their meshed molecular surface) to the binary complex MtInhA-NADH (represented as a solid molecular surface colored by CPK) from docking simulations. Closer inspection of the docked complexes suggests a steric hindrance produced by themethyl group which is observed to a minor extent for a fluorine atom. (For interpretation of the references to color in this figure legend, the reader is referred to the web version ofthis article.)

Table 2Inhibition constants of select compounds on MtInhA activity.

Comp. NADH Inhibition mode DD-CoA Inhibition mode

Kii (mM)a Kis (mM)b Kii (mM)

2 0.155 (±0.01) Uncompetitive 0.100 (±0.03) 0.425 (±0.18) Noncompetitive4a 1.95 (±0.16) Uncompetitive 1.60 (±0.53) 7.22 (±6.77) Noncompetitive4b 0.181 (±0.01) Uncompetitive 0.101 (±0.02) e Competitive4c 1.12 (±0.04) Uncompetitive 2.77 (±2.08) 2.00 (±4.62) Noncompetitive4d 0.348 (±0.06) Uncompetitive 0.338 (±0.10) e Competitive4f 0.185 (±0.01) Uncompetitive 0.162 (±0.03) 0.348 (±0.14) Noncompetitive4h 0.147 (±0.004) Uncompetitive 0.118 (±0.02) e Competitive4i 0.332 (±0.04) Uncompetitive 0.149 (±0.06) 0.660 (±0.60) Noncompetitive4j 1.28 (±0.09) Uncompetitive 0.788 (±0.26) 3.25 (±1.56) Noncompetitive

a Kii is the inhibitory dissociation constant for the ESI complex.b Kis is the inhibitory dissociation constant for the EI complex [44].

M. Rotta et al. / European Journal of Medicinal Chemistry 90 (2015) 436e447 441

corroborate the proposed hydrogen bonding network between thepharmacophore-like carbonyl group, hydroxyl group of the nico-tinamide ribose and Y158 in the formation of the ESI complex[25,26,42].

Furthermore, when NADH was fixed at a non-saturating con-centration and the DD-CoA substrate was varied in the presence offixed-varying concentrations of the piperazines 2, 4a, 4c, 4f, and4iej, a noncompetitive inhibition profile was obtained, as indicatedby the lines that intercept on left of y-axis in the double reciprocalplots. Fitting the data to Eq. (3) yielded Kis values ranging from0.100 to 2.77 mM and Kii values ranging from 0.348 to 7.22 mM(Table 2). Except for compound 4c, inhibitors exhibited Kis < Kii,indicating a binding preference for the free enzyme in comparisonto binding to the MtInhA-DD-CoA binary complex. These resultssuggested that the compounds could interact with the DD-CoAbinding site even with a noncompetitive signature [44]. Aimingto shed light on this issue, saturating concentrations of DD-CoA andNADH were used to assay lead-like compound 2. In these assays,when the DD-CoA concentration was maintained at approximately135 mM (Km y 45 mM [36]) and the NADH concentrationwas variedfrom 10 to 120 mM, the uncompetitive inhibition with respect toNADH was maintained (Supporting information). However, whenthe NADH concentration was saturating at 240 mM (Km y 60 mM[36]) and the DD-CoA concentration was varied from 15 to 200 mM,

a competitive inhibition profile with respect to the DD-CoA sub-strate was observed for piperazine 2 (Supporting information).Taken together kinetics results and crystallographic studies of theanalogous compound 1 strengthened the hypothesis that com-pound 2 interacts with the enzyme at the fatty acid binding site[25].

Inhibition assays for 4b, 4d and 4h indicated that these com-pounds are competitive inhibitors with respect to the DD-CoAsubstrate, as the lines intersected at a single value on the y-axisof the double-reciprocal plots (Supporting information). Fitting thedata to Eq. (4) yielded Kis values ranging from 0.101 to 0.338 mM(Table 2). As proposed for 2, these compounds likely interact withthe fatty acid binding site of MtInhA.

Finally, MIC determination for the most potent MtInhA in-hibitors was accomplished in vitro using the M. tuberculosis H37Rvstrain (Supporting information). Although the poor solubility of theevaluated compounds precluded a more accurate MIC determina-tion, piperazines 4b and 4d exhibited moderate antimycobacterialactivity, with a MIC of 25 mg/mL (z67.8 mM).

3. Conclusion

The inhibition of MtInhA activity by piperazine derivatives wasassessed using kinetic assays. Additionally, insights into the

M. Rotta et al. / European Journal of Medicinal Chemistry 90 (2015) 436e447442

structural requirements for effective enzyme inhibition and thedetermination of the mode of inhibition of the synthesized com-pounds emphasized the pharmacophore role of the amide carbonylgroup in the formation of the ESI complex. The results presentedherein suggest that modulation of the hydrogen bond acceptor maybe a critical step in the optimization of MtInhA inhibitors. Futurechallenges will include the design of novel piperazine derivativecompounds for the fatty acid binding site maintaining the drug-likeproperties that allow for cellular potency. Finally, our results mayprovide important details for the design of novel structures tointeract with a clinically validated TB target, thereby prompting thediscovery of novel antituberculosis agents.

4. Experimental section

4.1. Synthesis and structure: apparatus and analysis

All common reactants and solvents were used as obtained fromcommercial suppliers without further purification. Melting pointswere determined using a Microquímica MQAPF-302 apparatus. 1HNMR spectra were acquired on an Anasazi EFT-60 spectrometer (1Hat 60.13 MHz) at 30 �C. 13C NMR spectra were acquired on a Varian(Federal University of Rio Grande do Sul, UFRGS/Brazil) spectrom-eter (13C at 100.6 MHz) at 25 �C. CDCl3 was used as the solvent, andTMS was used as an internal standard in 5-mm samples tubes.Chemical shifts are expressed in ppm, and J values are given in Hz.High-resolution mass spectra (HRMS) were obtained for all com-pounds on an LTQ Orbitrap Discovery mass spectrometer (ThermoFisher Scientific, Bremen, Germany). This hybrid system combinesthe LTQ XL linear ion trapmass spectrometerwith an Orbitrapmassanalyzer. The experiments were performed using direct infusion ofthe sample in a solution of acetonitrile (50%), methanol (50%), andformic acid (0.1%), with a flow of 5 mL min�1 in positive-ion modeusing electrospray ionization. Elemental composition calculationswere performed using a specific tool included in the Qual Browsermodule of the Xcalibur (Thermo Fisher Scientific, release 2.0.7)software. Fourier transform infrared (FTIR) spectra were recordedusing a universal attenuated total reflectance (UATR) attachmenton a PerkinElmer Spectrum 100 spectrometer in the wavenumberrange of 650e4000 cm�1 with a resolution of 4 cm�1. Crystal datawere collected using a Bruker APEX II CCD area-detector diffrac-tometer and employing graphite-monochromatized MoKa radia-tion. The structures from the ligand and the complex weredetermined by direct methods using SHELXS-97 [46]. SubsequentFourier-difference map analyses yielded the positions of the non-hydrogen atoms. Refinements were performed using the SHELXL-97 package [47]. All refinements were performed using full-matrixleast-squares on F2, with anisotropic displacement parameters forall non-hydrogen atoms. Hydrogen atoms were included in therefinement in calculated positions. Finally, the structure wasrendered using ORTEP-3 for Windows [48].

4.2. General procedure for the synthesis of piperazines 2, 4aek and10aef

4.2.1. Method iBenzoic acid (1 mmol) was reacted with a solution of thionyl

chloride in CH2Cl2 (1 M, 6 mL) in refluxing toluene (3 mL) for 5 h.After cooling to room temperature, the solvent was evaporatedunder reduced pressure. The obtained benzoyl chloride was dis-solved in dry CH2Cl2 (5 mL), and the resultant solution was cooledto 0 �C. To this solution, TEA (4.4 mmol, 0.444 g), 1-(9H-fluoren-9-yl)-piperazine (3) (1.1 mmol, 0.275 g) or 1-(diphenylmethyl)-piperazine (9) (1.1 mmol, 0.277 g) in dry CH2Cl2 (5 mL) was addeddropwise. After the addition, the reaction mixture was stirred for

16 h at room temperature. Subsequently, the solution was washedwith aqueous HCl (10%, v/v, 3 � 15 mL), a saturated NaHCO3 solu-tion (1 �15 mL), and aqueous NaCl (5%, w/v, 1 �15 mL). Finally, theorganic layer was dried over anhydrous magnesium sulfate. Themixture was filtered, and the solvent was evaporated under reducepressure. When necessary, the products were purified using silicagel chromatography with CH2Cl2/methanol (20:5, v/v) as the mo-bile phase. This protocol was used for the synthesis of 9H-fluoren-9-yl-piperazines 2, 4a, 4cee, 4hei and 1-(diphenylmethyl)-piper-azines 10aeb and 10dee. The crystal used for the data collectionwas obtained by recrystallization of compound 4d from hexanefollowed by slow evaporation at room temperature.

4.2.2. Method iiBenzoyl chloride (1.0 mmol) was solubilized in dry CH2Cl2

(5 mL), and the resultant solution was cooled to 0 �C. To this so-lution, TEA (4.4 mmol, 0.444 g), 1-(9H-fluoren-9-yl)-piperazine (3)(1.1 mmol, 0.275 g) or 1-(diphenylmethyl)-piperazine (9)(1.1 mmol, 0.277 g) in dry CH2Cl2 (5 mL) was added dropwise. Afterthe addition, the reaction mixture was stirred for 16 h at 25 �C.Subsequently, the solution was washed with aqueous HCl (10%, v/v,3 � 15 mL), a saturated NaHCO3 solution (1 � 15 mL), and aqueousNaCl (5%, w/v, 1 � 15 mL). Finally, the organic layer was dried overanhydrous magnesium sulfate. The mixture was filtered, and thesolvent was evaporated under reduce pressure. When necessary,the products were purified using silica gel chromatography withCH2Cl2/methanol (20:5, v/v) as the mobile phase. This protocol wasused for the synthesis of 9H-fluoren-9-yl-piperazines 4b, 4feg, 4jand 1-(diphenylmethyl)-piperazines 10c and 10f.

4.2.3. Method iiiPyridine-4-carboxylic acid (0.5 mmol, 0.061 g) was reacted with

a solution of HBTU (0.55 mmol, 0.208 g), TEA (2 mmol, 0.202 g), 1-(9H-fluoren-9-yl)-piperazine (3) (0.5 mmol, 0.125 g) in dry DMF(5 mL). The reaction mixture was stirred for 4 h at 45 �C. Subse-quently, the solution was diluted with aqueous NaCl (5%, w/v,15 mL) and extracted with ethyl acetate (3 � 15 mL). Finally, theorganic layer was dried over anhydrous magnesium sulfate. Themixture was filtered, and the solvent was evaporated under reducepressure. The product was purified using silica gel chromatographywith CH3Cl/methanol (40:1, v/v) as the mobile phase. This protocolwas used for the synthesis of 9H-fluoren-9-yl-piperazine 4k.

4.2.4. (4-(9H-Fluoren-9-yl)piperazin-1-yl) (phenyl)methanone (2)Yellow powder; yield: 0.148 g (42%); m.p. 206e208 �C; 1H NMR

(60 MHz, CDCl3): d 2.63 (m, 4H, CH2), 3.52 (br, 4H, CH2), 4.84 (s, 1H,CH), 7.32e7.74 (m,13H, Ar); 13C NMR (100MHz, CDCl3): d 48.3, 67.9,120.7, 127.2, 128.6, 128.7, 130.6, 131.2, 133.6, 134.4, 142.1, 170.2; FTIR(UATR, cm�1): 3330, 2930, 2817, 1634, 1447, 1425, 740; HRMS (ESI)calcd for C24H22N2O3 þ H: 355.1805, found: 355.1796 (MþH)þ.

4.2.5. (4-(9H-Fluoren-9-yl)piperazin-1-yl) (4-methoxyphenyl)methanone (4a)

Brown powder; yield: 0.309 g (87%); m.p. 143e145 �C; 1H NMR(60 MHz, CDCl3): d 2.59e2.70 (m, 4H, CH2), 3.55e3.71 (m, 4H, CH2),3.77 (s, 3H, OCH3), 4.84 (s, 1H, CH), 6.82 (d, 2H, Ar), 7.25e7.78 (m,10H, Ar); FTIR (UATR, cm�1): 2931, 2798, 1632, 1424, 1250, 738;HRMS (ESI) calcd for C25H24N2O2 þ H: 385.1911, found: 385.1900(MþH)þ.

4.2.6. (4-(9H-Fluoren-9-yl)piperazin-1-yl) (4-tolyl)methanone (4b)White powder; yield: 0.202 g (55%); m.p. 146e147 �C; 1H NMR

(60 MHz, CDCl3): d 2.33 (s, 3H, CH3), 2.25e2.67 (m, 4H, CH2),3.53e3.64 (m, 4H, CH2), 4.86 (s, 1H, CH), 7.20e7.675 (m, 14H, Ar);FTIR (UATR, cm�1): 2920, 2795, 1624, 1429, 1269, 1005, 828, 742;

M. Rotta et al. / European Journal of Medicinal Chemistry 90 (2015) 436e447 443

HRMS (ESI) calcd for C25H24N2O þ H: 369.1961, found: 369.1960(MþH)þ.

4.2.7. (4-(9H-Fluoren-9-yl)piperazin-1-yl) (4-chlorophenyl)methanone (4c)

White powder; yield: 0.027 g (7%); m.p. 133e135 �C; 1H NMR(60 MHz, CDCl3): d 2.60e2.71 (m, 4H, CH2), 3.54e3.62 (m, 4H, CH2),4.86 (s, 1H, CH), 7.25e7.76 (m, 12H, Ar); FTIR (UATR, cm�1): 2808,1624, 1431, 1269, 1004, 836, 740, 730; HRMS (ESI) calcd forC24H21ClN2O þ H: 389.1410, found: 389.1415 (MþH)þ.

4.2.8. (4-(9H-Fluoren-9-yl)piperazin-1-yl) (4-fluorophenyl)methanone (4d)

Yellow powder; yield: 0.029 g (8%); m.p. 140e142 �C; 1H NMR(60 MHz, CDCl3): d 2.57e2.71 (m, 4H, CH2), 3.47e3.61 (m, 4H, CH2),4.86 (s, 1H, CH), 6.84e7.75 (m, 12H, Ar); FTIR (UATR, cm�1): 2806,1625, 1603, 1218, 1149, 1003, 840, 739; HRMS (ESI) calcd forC24H21N2OF þ H: 373.1711, found: 373.1732 (MþH)þ.

4.2.9. (4-(9H-Fluoren-9-yl)piperazin-1-yl) (4-nitrophenyl)methanone (4e)

Yellow powder; yield: 0.095 g (24%); m.p. 119e120 �C; 1H NMR(60 MHz, CDCl3): d 2.52e2.73 (m, 4H, CH2), 3.29e3.71 (m, 4H, CH2),4.86 (s, 1H, CH), 7.25e7.76 (m,10H, Ar), 8.21 (d, 2H, Ar); FTIR (UATR,cm�1): 3330, 2931,1630,1599,1516,1439, 740; HRMS (ESI) calcd forC24H21N3O3 þ H: 400.1656, found: 400.1644 (MþH)þ.

4.2.10. (4-(9H-Fluoren-9-yl)piperazin-1-yl) (3-tolyl)methanone(4f)

Yellow powder; yield: 0.0856 (77%); m.p. 124e126 �C; 1H NMR(60 MHz, CDCl3): d 2.32 (s, 3H, CH3), 2.71 (br, 4H, CH2), 3.62 (br, 4H,CH2), 4.94 (s, 1H, CH), 7.16e7.841 (m, 12H, Ar); FTIR (UATR, cm�1):3373, 293, 2808, 1632, 1450, 1421, 1275, 1201, 1005, 810, 737; HRMS(ESI) calcd for C25H24N2O þ H: 369.1961, found: 369.1960 (MþH)þ.

4.2.11. (4-(9H-Fluoren-9-yl)piperazin-1-yl) (2-tolyl)methanone(4g)

Yellow powder; yield: 0.0917 (83%); m.p. 63e64 �C; 1H NMR(60 MHz, CDCl3): d 2.24 (s, 3H, CH3), 2.54 (br, 2H, CH2), 2.91 (br, 2H,CH2), 3.28 (br, 2H, CH2), 3.89 (br, 2H, CH2), 5.96 (s, 1H, CH),7.16e7.77 (m, 12H, Ar); FTIR (UATR, cm�1): 2923, 2817, 1627, 1426,1255, 1001, 729; HRMS (ESI) calcd for C25H24N2O þ H: 369.1691,found: 369.1966 (MþH)þ.

4.2.12. (4-(9H-Fluoren-9-yl)piperazin-1-yl) (3-fluorophenyl)methanone (4h)

Brown powder; yield: 0.029 g (8%); m.p. 165e167 �C; 1H NMR(60 MHz, CDCl3): d 2.53e2.69 (m, 4H, CH2), 3.41e3.66 (m, 4H, CH2),4.85 (s, 1H, CH), 6.99e7.78 (m, 13H, Ar); FTIR (UATR, cm�1): 3063,2824, 1639, 1582, 1439, 1277, 1004, 740; HRMS (ESI) calcd forC24H21FN2O þ H: 373.1773, found: 373.1747 (MþH)þ.

4.2.13. (4-(9H-Fluoren-9-yl)piperazin-1-yl) (2-fluorophenyl)methanone (4i)

White powder; yield: 0.119 g (32%); m.p. 169e171 �C; 1H NMR(60 MHz, CDCl3): d 2.37e2.53 (m, 2H, CH2), 2.71e2.85 (m, 2H, CH2),3.18e3.34 (m, 2H, CH2), 3.70e3.90 (m, 2H, CH2), 4.85 (s, 1H, CH),6.99e7.76 (m, 12H, Ar); FTIR (UATR, cm�1): 3342, 2835, 1631, 1610,1451, 1431, 1276, 1007, 741; HRMS (ESI) calcd for C24H21FN2O þ H:373.1773, found: 373.1747 (MþH)þ.

4.2.14. (4-(9H-Fluoren-9-yl)piperazin-1-yl) (2,5-difluorophenyl)methanone (4j)

White powder; yield: 0.339 g (87%); m.p. 167e168 �C; 1H NMR(60 MHz, CDCl3): d 2.64e2.99 (m, 4H, CH2), 3.41e3.59 (m, 2H, CH2),

3.89e4.04 (m, 2H, CH2), 5.10 (s, 1H, CH), 6.93e7.87 (m, 11H, Ar); 13CNMR (100 MHz, CDCl3): d 48.6, 69.2, 115.8 (dd, J ¼ 25.2, J ¼ 8.4),117.0 (dd, J ¼ 24.4, J ¼ 8.4), 118.1 (dd, J ¼ 24.0, J ¼ 8.8), 120.1, 126.5,127.6, 129.2, 141.3, 153.9 (d, J ¼ 244.9), 158.6 (d, J ¼ 244.9), 163.6;FTIR (UATR, cm�1): 3061, 2929, 1634, 1611, 1449, 1008, 737; HRMS(ESI) calcd for C24H20F2N2O3 þ H: 391.1616, found: 391.1616(MþH)þ.

4.2.15. (4-(9H-Fluoren-9-yl)piperazin-1-yl) (pyridin-4-yl)methanone (4k)

White powder; yield: 0.148 g (83%); m.p. 156e157 �C; 1H NMR(400 MHz, CDCl3): d 2.42e2.44 (m, 2H, CH2), 2.83e2.85 (m, 2H,CH2), 3.27e3.29 (m, 2H, CH2), 3.78e3.80 (m, 2H, CH2), 4.87 (s, 1H,CH), 7.23 (d, J¼ 6.3, 2H, Ar), 7.30 (t, J¼ 7.8, 2H, Ar) 7.39 (t, J¼ 7.4, 2H,Ar), 7.60 (d, J ¼ 7.4, 2H, Ar), 7.69 (d, J ¼ 7.4, 2H, Ar), 8.63 (d, J ¼ 5.9,2H, Ar). 13C NMR (100 MHz, CDCl3): d 42.8, 48.2, 48.5, 49.4, 69.8,119.9, 121.2, 125.8, 127.2, 128.4, 141.0, 143.1, 143.4, 150.1, 167.6; FTIR(UATR, cm�1): 2815, 1636, 1440, 1279, 1000, 832, 751; HRMS (ESI)calcd for C23H21N3O þ H: 356.1757, found: 356.1756 (MþH)þ.

4.2.16. (4-Benzhydrylpiperazin-1-yl) (phenyl)methanone (10a)White powder; yield: 0.309 g (87%); m.p. 145e146 �C; 1H NMR

(60 MHz, CDCl3): d 2.33e2.47 (m, 4H, CH2), 3.47e3.68 (m, 4H, CH2),4.26 (s, 1H, CH), 7.61e7.35 (m, 15H, Ar); FTIR (UATR, cm�1): 3024,2808, 1631, 1446, 1265, 996, 706; HRMS (ESI) calcd forC24H24N2O þ H: 357.1961, found: 357.1937 (MþH)þ.

4.2.17. (4-Benzhydrylpiperazin-1-yl) (4-methoxyphenyl)methanone (10b)

White powder; yield: 0.142 g (37%); m.p. 64e66 �C; 1H NMR(60 MHz, CDCl3): d 2.34e2.49 (m, 4H, CH2), 3.63e3.63 (m, 4H, CH2),3.78 (s, 3H, OCH3), 5.27 (s, 1H, CH), 6.78e6.92 (d, 2H, Ar), 7.21e7.43(m, 12H, C5H6); FTIR (UATR, cm�1): 3421, 2808, 1605, 1427, 1230,996, 704; HRMS (ESI) calcd for C25H26N2O þ H: 387.2120, found:387.2146 (MþH)þ.

4.2.18. (4-Benzhydrylpiperazin-1-yl) (4-tolyl)methanone (10c)White powder; yield: 0.203 g (55%); m.p. 62e64 �C; 1H NMR

(60 MHz, CDCl3): d 2.33 (s, 3H, CH3), 2.35e2.49 (m, 4H, CH2),3.73e3.79 (m, 4H, CH2), 4.36 (s, H, CH), 7.20e7.47 (m, 14H, Ar); FTIR(UATR, cm�1): 2807, 1627, 1426, 1285, 1253, 996, 748, 705; HRMS(ESI) calcd for C27H28N2O2 þ H: 371.2118, found: 371.2131 (MþH)þ.

4.2.19. (4-Benzhydrylpiperazin-1-yl) (4-chlorophenyl)methanone(10d)

Brown powder; yield: 0.234 g (60%); m.p. 130e132 �C; 1H NMR(60 MHz, CDCl3): d 2.36e2.51 (m, 4H, CH2), 3.53e3.78 (m, 4H, CH2),4.28 (s, 1H, CH), 7.26e7.45 (m, 15H, Ar); FTIR (UATR, cm�1): 3027,2808, 1629, 1431, 1254, 1087, 996, 832, 745, 696; HRMS (ESI) calcdfor C24H23ClN2O þ H: 391.1572, found: 391.1609 (MþH)þ.

4.2.20. (4-Benzhydrylpiperazin-1-yl) (4-fluorophenyl)methanone(10e)

Brown powder; yield: 0.202 g (54%); m.p. 54e57 �C; 1H NMR(60 MHz, CDCl3): d 2.42e2.52 (m, 4H, CH2), 2.92e3.06 (m, 2H, CH2),3.36e3.60 (m, 2H, CH2), 4.24 (s, 1H, CH), 6.88e7.51 (m, 14H, Ar);FTIR (UATR, cm�1): 2811, 1628, 1603, 1451, 1432, 1254, 1223, 996,843, 705; HRMS (ESI) calcd for C24H23N2OF þ H: 375.1867, found:372.1874 (MþH)þ.

4.2.21. (4-Benzhydrylpiperazin-1-yl) (2,5-difluorophenyl)methanone (10f)

White powder; yield: 0.321 g (82%); m.p. 78e79 �C; 1H NMR(60 MHz, CDCl3): d 2.28e2.52 (br, 4H, CH2), 3.25e3.41 (m, 2H, CH2),3.70e3.88 (m, 2H, CH2), 4.26 (s, 1H, CH), 6.93e7.41 (m, 13H, Ar);

M. Rotta et al. / European Journal of Medicinal Chemistry 90 (2015) 436e447444

FTIR (UATR, cm�1): 2919, 1627, 1427, 1286, 1253, 996, 705; HRMS(ESI) calcd for C24H22F2N2O þ H: 393.1773, found: 393.1747(MþH)þ.

4.3. General procedure for the synthesis of piperazines 5aeb and 11

Sulfonyl halide (1.2 mmol) was dissolved in dry CH2Cl2 (5 mL),and the resultant solution was cooled to 0 �C. To this solution, TEA(2.2 mmol, 0.222 g), 1-(9H-fluoren-9-yl)-piperazine (3) (1.1 mmol,0.275 g) or 1-(diphenylmethyl)-piperazine (9) (1.1 mmol, 0.277 g),in dry CH2Cl2 (5 mL) was added dropwise. After the addition, thereaction mixture was stirred for 16 h at room temperature.Extraction and purificationwere performed as previously describedfor piperazines 2, 4 and 10.

4.3.1. 1-(9H-Fluoren-9-yl)-4-(phenylsulfonyl)piperazine (5a)White powder; yield: 0.238 g (61%); m.p. 152e154 �C; 1H NMR

(60 MHz, CDCl3): d 2.58e2.70 (m, 4H, CH2), 2.85e3.11 (m, 4H, CH2),4.91 (s, 1H, CH), 7.13e7.89 (m, 13H, Ar); FTIR (UATR, cm�1): 3418,2887, 2835, 1634, 1348, 1171, 735; HRMS (ESI) calcd forC23H22N2O2S þ H: 391.1475, found: 391.1467 (MþH)þ.

4.3.2. 1-(9H-Fluoren-9-yl)-4-(o-tolylsulfonyl)piperazine (5b)White powder; yield: 0.331 g (82%); m.p. 68e70 �C; 1H NMR

(60 MHz, CDCl3): d 2.61e2.74 (m, 7H, CH2), 3.09e3.24 (m, 4H, CH2),4.81 (s, 1H, CH), 7.49e7.75 (m, 13H, Ar); FTIR (UATR, cm�1): 1447,1314, 1158, 1133, 940, 722; HRMS (ESI) calcd for C24H24N2O2S þ H:405.1631, found: 405.1631 (MþH)þ.

4.3.3. 1-Benzhydryl-4-(phenylsulfonyl)piperazine (11)White powder; yield: 0.058 (15%); m.p. 153e155 �C (m.p.

157e159 �C [49]); 1H NMR (60 MHz, CDCl3): d 2.43e2.50 (m, 4H,CH2), 3.03e3.63 (m, 4H, CH2), 4.24 (s, 1H, CH), 7.25e7.83 (m, 15H,Ar); FTIR (UATR, cm�1): 2855, 2811, 1447, 1347, 1331, 1169, 944, 741,689; HRMS (ESI) calcd for C25H26N2O2S þ H: 393.1631, found:393.1607 (MþH)þ.

4.4. General procedure for the synthesis of piperazines 6 and 12

The cinnamic acid precursor was synthesized as previouslydescribed [28]. Thereafter, this acid (1 mmol, 0.148 g) was reactedwith a solution of thionyl chloride in CH2Cl2 (1 M, 6 mL) for 5 h inrefluxing toluene (5 mL). The obtained acyl chloride was dissolvedin dry CH2Cl2 (5 mL), and the resultant solution was cooled to 0 �C.To this solution, TEA (4.4 mmol, 0.444 g), 1-(9H-fluoren-9-yl)-piperazine (3) (1.1 mmol, 0.275 g) or 1-(diphenylmethyl)-pipera-zine (9) (1.1 mmol, 0.277 g) in dry CH2Cl2 (5 mL) was addeddropwise. After the addition, the reaction mixture was stirred for16 h at room temperature. Finally, ethyl ether (15 mL) was added tothe reaction mixture, and the precipitate was isolated on a filter.The products were purified using silica gel chromatography withCH2Cl2/methanol (20:5, v/v) as the mobile phase.

4.4.1. (E)-1-(4-(9H-Fluoren-9-yl)piperazin-1-yl)-3-phenylprop-2-en-1-one (6)

White powder; yield: 0.098 g (26%); m.p. 169e170 �C; 1H NMR(60 MHz, CDCl3): d 2.34e2.50 (m, 4H, CH2), 3.61e3.75 (m, 4H, CH2),4.86 (s, 1H, CH), 6.77 (d, J ¼ 15 Hz, 1H, CHtrans), 7.25e7.53 (m, 17H,Ar), 7.77 (d, J ¼ 15, CHtrans); FTIR (UATR, cm�1): 3060, 2813, 1649,1605, 1410, 1202, 740; HRMS (ESI) calcd for C26H24N2O þ H:381.1961, found: 381.1967 (MþH)þ.

4.4.2. (E)-1-(4-Benzhydrylpiperazin-1-yl)-3-phenylprop-2-en-1-one (12)

White powder; yield: 0.126 g (33%); m.p. 141e143 �C; 1H NMR(60 MHz, CDCl3): d 2.34e2.50 (m, 4H, CH2), 3.61e3.75 (m, 4H, CH2),4.26 (s, 1H, CH), 6.80 (d, J ¼ 15 Hz, 1H, CHtrans), 7.17e7.36 (m, 18H,Ar), 7.66 (d, J ¼ 15 Hz, 1H, CHtrans); FTIR (UATR, cm�1): 3023, 2809,1644, 1595, 1452, 1439, 1229, 705; HRMS (ESI) calcd forC26H26N2O þ H: 383.2123, found: 383.2118 (MþH)þ.

4.5. General procedure for the synthesis of 9H-fluoren-9-yl-piperazine 7

Bromo-N-phenylacetamide was synthesized from 2-bromoace-tyl chloride and aniline as previously described [29]. A solutioncontaining bromo-N-phenylacetamide (1.04 mmol, 0.221 g), po-tassium carbonate (3.12 mmol, 1.08 g), and 1-(9H-fluoren-9-yl)-piperazine (3) (1.1 mmol, 0.275 g) in DMF (8 mL) was stirred underargon atmosphere for 24 h at room temperature. Thereafter, thesolution was diluted with distilled water (50 mL). The obtainedsolid was isolated on a filter and dried under reduce pressure. Theproduct was purified using silica gel chromatography with chlo-roform/methanol (50:50, v/v) as the mobile phase.

4.5.1. 2-(4-(9H-Fluoren-9-yl)piperazin-1-yl)-N-phenylacetamide(7)

White powder; yield: 0.172 g (51%); m.p. 174e175 �C; 1H NMR(60MHz, CDCl3): d 2.60e2.75 (m, 8H, CH2), 3.10 (s, 2H, CH2), 4.86 (s,1H, CH), 7.03e7.78 (m, 15H, Ar); FTIR (UATR, cm�1): 2546, 1682,1447, 1313, 1133, 952, 736; HRMS (ESI) calcd for C25H25N3O þ H:384.2048, found: 384.2070 (MþH)þ.

4.6. General procedure for the synthesis of 9H-fluoren-9-yl-piperazine 8

A solution containing (bromomethyl)benzene (1.04 mmol,0.124 mL), potassium carbonate (3.12 mmol, 1.08 g), and 1-(9H-fluoren-9-yl)-piperazine (3) (1.1 mmol, 0.275 g) in DMF (8 mL) wasstirred under argon atmosphere and at room temperature for 24 h.Thereafter, the mixture was diluted with distilled water (20 mL).The obtained solid was isolated on a filter and dried under reducepressure. The product was purified using silica gel chromatographywith chloroform/methanol (50:50, v/v) as the mobile phase.

4.6.1. 9-(4-Benzylcyclohexyl)-9H-fluorene (8)Yellow powder; yield: 0.055 g (55%); m.p. 129e131 �C; 1H NMR

(60 MHz, CDCl3): d 2.38 (br, 8H, CH2), 3.48 (s, 2H, CH2), 4.82 (s, 1H,CH), 7.62e7.73 (m, 15H, Ar); FTIR (UATR, cm�1): 2807, 1449, 1130,1007, 736, 697; HRMS (ESI) calcd for C24H24N2 þ H: 341.2012,found: 341.2016 (MþH)þ.

4.7. Theoretical calculations and docking simulations

The geometry of the compound was optimized using semi-empirical AM1 and PM3 methods implemented in the Hyper-Chem 7.52 package (2002) [50]. The structureswere fully optimizedwithout fixing any parameter, thus, bringing all geometric variablesto their equilibrium values. The energy minimization protocolemployed the PolakeRibiere conjugated gradient algorithm.Convergence to a local minimum was achieved when the energygradient was �0.01 kcal mol�1. The RHF method was used in thespin pairing for the two semi-empirical tools.

Flexible ligand docking, carried out using AutoDock4.2.5.1 andAutoDockTools4 software [51,52], employed the crystal structure ofthe piperazine 1 (Genz-10850) bound to MtInhA (PDB ID: 1P44).The NADH coenzyme was treated as part of the protein receptor in

M. Rotta et al. / European Journal of Medicinal Chemistry 90 (2015) 436e447 445

all docking simulations. Water molecules were removed, polarhydrogens were added, and Gasteiger partial atomic charges forligands were assigned by AutoDockTools4 [51]. For the dockingsimulation, the active site was set on a cubic grid, spaced at 0.375 Å,with dimensions of 60 points � 60 points � 60 points centered onthe crystallographic bound ligand 1. This grid is large enough toinclude the NADH coenzyme as part of the protein receptor, as wellas theMtInhA substrate binding pocket in 1P44. The results of eachdocking experiment were derived from 25 independent runs, forwhich a maximum number of 27,000 LGA (Lamarckian GeneticAlgorithm) generations was generated on the initial population of150 individuals, a maximum number of 2.5 � 106 energy evalua-tions, with an elitism value of 1, a mutation rate of 0.02, and a cross-over rate of 0.8. For the local search, the pseudo-Solis and Wetsmethod was applied using default parameters [52]. Each run pro-vides one predicted binding mode. At the end of the dockingexperiment, binding modes with root-mean-square deviation(RMSD) of 2.0 Å within each other and to the crystallographicbound ligand 1 were placed in the same cluster.

4.8. Enzymatic activity assays

4.8.1. Inhibition assessed by steady-state kinetics measurementsThe expression and purification of recombinant MtInhA were

performed as previously described [36,37]. The substrate DD-CoAwas synthesized from 2-trans-dodecenoic acid and coenzyme Avia anhydride formation following acylation, as previouslydescribed [36]. DD-CoAwas purified by reverse-phase HPLC using a19 � 300 mm C18 mBondapak column (Waters Associates, Milford,MA) as previously described [53]. Protein concentration wasdetermined by the Bradford method [54] using serum albumin asthe standard.

The steady-state velocities in the presence of the synthesizedcompounds were determined using a UV-2550 UV/visible spec-trophotometer (Shimadzu) bymonitoring NADH oxidation to NADþ

at 340 nm (εb-NADH ¼ 6.22 M�1 cm�1) due the reduction of the DD-CoA substrate. Experiments were performed at 25 �C in 100 mMPIPES, pH 7.0, and measurement of the enzyme-catalyzed reactionstarted with the addition of MtInhA at 2.2 mM to the assay mixture(500 mL final volume) and data collected for 1 min.

4.8.2. IC50 determinationsThe IC50 values were determined by adding different concen-

trations of the compounds (dissolved in DMSO) to the reactionmixture, and the enzyme velocity was used to determine the %inhibition. As a control, the maximal rate of the enzymatic reaction(100% MtInhA activity) was determined in the absence of inhibitorand in the presence of fixed non-saturating concentrations ofNADH (Km y 60 mM) and DD-CoA (Km y 45 mM), in the presence of1% DMSO [36,37]. The IC50 values were estimated using Eq. (1),where [I] is the inhibitor concentration, vi is the initial velocity inthe presence of the inhibitor, vo is the initial velocity in the absenceof inhibitor, and the IC50 value is defined as the concentration of theinhibitor that reduces the enzyme velocity by half.

viv0

¼ 1

1þ�

½I�IC50

� (1)

4.8.3. Time-dependent inhibitionTo evaluate whether or not enzyme inhibition may be time

dependent, 2.2 mM recombinant MtInhA was preincubated with0.2 mM inhibitor (final concentration), which was then added at

different times (up to 30min) to the reaction mixture, as previouslydescribed. The change in the initial velocity as a function of timewas monitored, and the percent inhibition was determined. Thisanalysis was performed to determine if inhibition follows a rapidequilibrium mode (classical competitive, uncompetitive andnoncompetitive inhibition) or if there is a slow step in the equi-librium process. Control experiments were also carried out toevaluate whether or not DMSO may have any inhibitory effect onMtInhA at the maximum concentration (1%). For these experi-ments, identical concentrations of recombinant MtInhA and in-hibitor were preincubated with 0.1 mMNADH (final concentration),0.1 mM DD-CoA (final concentration) or 1% DMSO (final concen-tration) [20].

4.8.4. Mode of inhibition and determination of the overall inhibitionconstant

The compounds that exhibited IC50 values below 3 mM wereanalyzed for themode of inhibition and overall inhibition constants(Ki). LineweavereBurk (double-reciprocal) plots were employed todetermine the mode of inhibition (competitive, noncompetitive oruncompetitive) and data fitting to appropriate equations gavevalues for the inhibition constants (Kis and/or Kii). In short, the in-hibition mode was proposed based on the effects on Vmax and Kmvalues for each inhibition type, resulting in plots with distinctstraight line patterns toward either NADH or DD-CoA as follows:lines intercept on y axis for competitive inhibition (does not affectthe apparent Vmax and increases apparent Km), lines intercept onleft of y axis for noncompetitive inhibition (decreases apparentVmax, and does not affect apparent Km values if Kis ¼ Kii, increasesthe apparent Km values if Kis < Kii, or decreases the Km values ifKis > Kii), and parallel lines for uncompetitive inhibition (decreasesboth apparent Vmax and Kmvalues) [44]. Initial rates weremeasuredat varying NADH concentrations (10, 20, 40, 60, 100, and 160 mM) atfixed non-saturating DD-CoA concentrations (Km y 45 mM), andfixed-varied inhibitor concentrations (0.05e4 mM). The Ki valueswith respect to NADH were calculated by fitting the data to anequation that describes uncompetitive inhibition (Eq. (2)), inwhich[I] is the inhibitor concentration, [S] is the substrate concentration,Km is the Michaelis constant, Vmax is the maximal velocity, and Kii isthe overall inhibition constant for the ESI complex [44].

1v0

¼ Km

Vmax

�1½S�

�þ 1Vmax

�1þ ½I�

Kii

�(2)

Initial rates were measured as a function of DD-CoA concen-tration (15, 30, 45, 75, 105, and 135 mM) at fixed non-saturatingNADH concentrations (Km y 60 mM) and fixed-varied inhibitorconcentrations (0.05e4 mM). The Ki values with respect to the DD-CoA substratewere calculated by fitting the data to an equation thatdescribes noncompetitive (Eq. (3)) or competitive (Eq. (4)) inhibi-tion, in which [I] is the inhibitor concentration, [S] is the substrateconcentration, Km is the Michaelis constant, Vmax is the maximalvelocity, Kii is the overall inhibition constant for the ESI complex,and Kis is the overall inhibition constant for the EI complex [44].

1v0

¼ Km

Vmax

�1þ I

Kis

��1½S�

�þ 1Vmax

�1þ ½I�

Kii

�(3)

1v0

¼ Km

Vmax

�1þ I

Kis

��1½S�

�þ 1Vmax

(4)

An independent experiment was performed using compound 2.Initial rates were measured as a function of the NADH concentra-tion (10e120 mM) at a fixed saturating DD-CoA concentration(135 mM), and fixed-varied inhibitor concentrations (0, 0.100, 0.250,

M. Rotta et al. / European Journal of Medicinal Chemistry 90 (2015) 436e447446

and 0.450 mM). The Ki values with respect to NADH were calculatedby fitting the data to an equation that describes uncompetitiveinhibition (Eq. (2)). Initial rates were measured as a function of theDD-CoA concentration (15e200 mM), fixed saturating NADH con-centration (240 mM), and fixed-varied inhibitor concentrations (0,0.05, 0.150, and 0.450 mM). The Ki values with respect to the DD-CoAsubstrate were calculated by fitting the data to an equationdescribing competitive inhibition (Eq. (4)). Higher concentrationsof DD-CoA could not be used because increasing the concentrationabove 135 mM resulted in a decrease in the rate, possibly due tosubstrate inhibition [37]. For all Ki determinations, each initial ve-locity was determined in duplicate, and at least five different sub-strate concentrations were examined.

Acknowledgments

This work was supported by funds from the National Institute ofScience and Technology on Tuberculosis (INCT-TB), Decit/SCTIE/MS-MCT-CNPq-FNDCT-CAPES (Brazil) to D. S. Santos and L. A. Basso.O. N. Souza (CNPq, 305984/2012-8), D. S. Santos (CNPq, 304051/1975-06), and L. A. Basso (CNPq, 520182/99-5) are Research CareerAwardees of the National Research Council of Brazil (CNPq). Thefellowships from CNPq (M. Rotta, K. Pissinate, and L.F.S.M. Timers),CAPES (A. D. Villela), and FAPERGS/CAPES (J. F. R. Bachega) are alsoacknowledged.

Appendix A. Supplementary data

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.ejmech.2014.11.034.

References

[1] World Health Organization, Global Tuberculosis Report 2013, 2013.[2] N.R. Gandhi, P. Nunn, K. Dheda, H.S. Schaaf, M. Zignol, D. van Soolingen,

P. Jensen, J. Bayona, Multidrug-resistant and extensively drug-resistanttuberculosis: a threat to global control of tuberculosis, Lancet 375 (9728)(2010) 1830e1843.

[3] Y. Zhang, W.W. Yew, M.R. Barer, Targeting persisters for tuberculosis control,Antimicrob. Agents Chemother. 56 (5) (2012) 2223e2230.

[4] A. Zumla, P. Nahid, S.T. Cole, Advances in the development of new tuberculosisdrugs and treatment regimens, Nat. Rev. Drug Discov. 12 (5) (2013) 388e404.

[5] J.C. Palomino, A. Martin, TMC207 becomes bedaquiline, a new anti-TB drug,Future Microbiol. 8 (9) (2013) 1071e1080.

[6] J. Avorn, Approval of a tuberculosis drug based on a paradoxical surrogatemeasure, J. Am. Med. Assoc. 309 (13) (2013) 1349e1350.

[7] H. Lu, P.J. Tonge, Inhibitors of FabI, an enzyme drug target in the bacterial fattyacid biosynthesis pathway, Acc. Chem. Res. 41 (1) (2008) 11e20.

[8] P.J. Brennan, H. Nikaido, The envelope of mycobacteria, Annu. Rev. Biochem.64 (1995) 29e63.

[9] K. Takayama, C. Wang, G.S. Besra, Pathway to synthesis and processing ofmycolic acids in Mycobacterium tuberculosis, Clin. Microbiol. Rev. 18 (1) (2005)81e101.

[10] C.E. Barry III, R.E. Lee, K. Mdluli, A.E. Sampson, B.G. Schroeder, R.A. Slayden,Y. Yuan, Mycolic acids: structure, biosynthesis and physiological functions,Prog. Lipid Res. 37 (2e3) (1998) 143e179.

[11] E.K. Schroeder, O.N. de Souza, D.S. Santos, J.S. Blanchard, L.A. Basso, Drugs thatinhibit mycolic acid biosynthesis in Mycobacterium tuberculosis, Curr. Pharm.Biotechnol. 3 (3) (2002) 197e225.

[12] A. Banerjee, E. Dubnau, A. Qu�emard, V. Balasubramanian, K.S. Um, T. Wilson,D. Collins, G. de Lisle, W.R. Jacobs Jr., inhA, a gene encoding a target forisoniazid and ethionamide in Mycobacterium tuberculosis, Science 263 (5144)(1994) 227e230.

[13] D.A. Rozwarski, G.A. Grant, D.H. Barton, W.R. Jacobs Jr., J.C. Sacchettini,Modification of the NADH of the isoniazid target (InhA) from Mycobacteriumtuberculosis, Science 279 (5347) (1998) 98e102.

[14] C. Vilch�eze, F. Wang, M. Arai, M.H. Hazb�on, R. Colangeli, L. Kremer,T.R. Weisbrod, D. Alland, J.C. Sacchettini, W.R. Jacobs Jr., Transfer of a pointmutation in Mycobacterium tuberculosis inhA resolved the target of isoniazid,Nat. Med. 12 (9) (2006) 1027e1029.

[15] K. Johnsson, P.G. Schultz, Mechanistic studies of the oxidation of isoniazid bythe catalase peroxidase from Mycobacterium tuberculosis, J. Am. Chem. Soc.116 (16) (1994) 7425e7426.

[16] B. Lei, C.J. Wei, S.C. Tu, Action mechanism of antitubercular isoniazid: acti-vation by Mycobacterium tuberculosis KatG, isolation, and characterization ofInhA inhibitor, J. Biol. Chem. 275 (4) (2000) 2520e2526.

[17] R. Rawat, A. Whitty, P.J. Tonge, The isoniazid-NAD adduct is a slow, tight-binding inhibitor of InhA, the Mycobacterium tuberculosis enoyl reductase:adduct affinity and drug resistance, Proc. Natl. Acad. Sci. U. S. A. 100 (24)(2003) 13881e13886.

[18] Y. Zhang, B. Heym, B. Allen, D. Young, S. Cole, The catalase-peroxidase geneand isoniazid resistance of Mycobacterium tuberculosis, Nature 358 (6387)(1992) 591e593.

[19] B. Heym, N. Honor�e, C. Truffot-Pernot, A. Banerjee, C. Schurra, W.R. Jacobs Jr.,J.D. van Embden, J.H. Grosset, S.T. Cole, Implications of multidrug resistancefor the future of short-course chemotherapy of tuberculosis: a molecularstudy, Lancet 344 (8918) (1994) 293e298.

[20] J.S. Oliveira, E.H. Sousa, L.A. Basso, M. Palaci, R. Dietze, D.S. Santos, I.S. Moreira,An inorganic iron complex that inhibits wild-type and an isoniazid-resistantmutant 2-trans-enoyl-ACP (CoA) reductase from Mycobacterium tuberculosis,Chem. Commun. 3 (2004) 312e313.

[21] V.S. Rodrigues-Junior, A. dos Santos Jr., A.J. dos Santos, C.Z. Schneider,J.B. Calixto, E.H.S. Sousa, L.G. de França Lopes, A.A. Souto, L.A. Basso,D.S. Santos, M.M. Campos, Activity of IQG-607, a new orally active compound,in a murine model of Mycobacterium tuberculosis infection, Int. J. Antimicrob.Agents 40 (2) (2012) 182e185.

[22] I. Pauli, R.N. dos Santos, D.C. Rostirolla, L.K. Martinelli, R.G. Ducati,L.F.S.M. Timmers, L.A. Basso, D.S. Santos, R.V.C. Guido, A.D. Andricopulo,O.N. de Souza, Discovery of new inhibitors of Mycobacterium tuberculosis InhAenzyme using virtual screening and a 3D-pharmacophore-based approach,J. Chem. Inf. Model. 53 (9) (2013) 2390e2401.

[23] P.S. Shirude, P. Madhavapeddi, M. Naik, K. Murugan, V. Shinde,R. Nandishaiah, J. Bhat, A. Kumar, S. Hameed, G. Holdgate, G. Davies,H. McMiken, N. Hegde, A. Ambady, J. Venkatraman, M. Panda, B. Bandodkar,V.K. Sambandamurthy, J.A. Read, Methyl-thiazoles: a novel mode of inhibitionwith the potential to develop novel inhibitors targeting InhA in Mycobacte-rium tuberculosis, J. Med. Chem. 56 (21) (2013) 8533e8542.

[24] X. He, A. Alian, R. Stroud, P.R. Ortiz de Montellano, Pyrrolidine carboxamidesas a novel class of inhibitors of enoyl acyl carrier protein reductase fromMycobacterium tuberculosis, J. Med. Chem. 49 (21) (2006) 6308e6323.

[25] M.R. Kuo, H.R. Morbidoni, D. Alland, S.F. Sneddon, B.B. Gourlie, M.M. Staveski,M. Leonard, J.S. Gregory, A.D. Janjigian, C. Yee, J.M. Musser, B. Kreiswirth,H. Iwamoto, R. Perozzo, W.R. Jacobs Jr., J.C. Sacchettini, D.A. Fidock, Targetingtuberculosis and malaria through inhibition of enoyl reductase: compoundactivity and structural data, J. Biol. Chem. 278 (23) (2003) 20851e20859.

[26] X. He, A. Alian, P.R. Ortiz de Montellano, Inhibition of the Mycobacteriumtuberculosis enoyl acyl carrier protein reductase InhA by arylamides, Bioorg.Med. Chem. 15 (21) (2007) 6649e6658.

[27] P. Pan, P.J. Tonge, Targeting InhA, the FASII enoyl-ACP reductase: SAR studieson novel inhibitor scaffolds, Curr. Top. Med. Chem. 12 (7) (2012) 672e693.

[28] K. Misra, H.S. Maity, S. Chanda, A. Nag, New greener alternatives for bio-reduction of aromatic aldehydes and decarboxylation of aromatic acids usingjuice of fruits, J. Mol. Catal. B Enzym. 82 (2012) 92e95.

[29] I.S. Macpherson, S. Kirubakaran, S.K. Gorla, T.V. Riera, J.A. D'Aquino, M. Zhang,G.D. Cuny, L. Hedstrom, The structural basis of Cryptosporidium-specific IMPdehydrogenase inhibitor selectivity, J. Am. Chem. Soc. 132 (18) (2010)1230e1231.

[30] E. Freire, Isothermal titration calorimetry: controlling binding forces in leadoptimization, Drug Discov. Today Technol. 1 (3) (2004) 295e299.

[31] CCDC number 1000538 contains the supplementary crystallographic data forcompound reported in this paper. These data can be obtained free of chargevia www.ccdc.cam.ac.uk/conts/retrieving.html (or from the Cambridge Crys-tallographic Data Centre, 12, Union Road, Cambridge CB2 1EZ, UK; fax: þ441223 336033).

[32] M.J.S. Dewar, E.G. Zoebisch, E.F. Healy, J.J.P. Stewart, Development and use ofquantum mechanical molecular models. 76. AM1: a new general purposequantum mechanical molecular model, J. Am. Chem. Soc. 107 (13) (1985)3902e3909.

[33] J.J.P. Stewart, Optimization of parameters for semiempirical methods I.Method, J. Comput. Chem. 10 (2) (1989) 209e220.

[34] J.J.P. Stewart, Optimization of parameters for semiempirical methods II. Ap-plications, J. Comput. Chem. 10 (2) (1989) 221e264.

[35] P. Machado, P.T. Campos, G.R. Lima, F.A. Rosa, A.F.C. Flores, H.G. Bonacorso,N. Zanatta, M.A.P. Martins, Experimental and calculated structural parametersof 5-trihalomethyl-4,5-dihydro-1H-pyrazole derivatives, novel analgesicagents, J. Mol. Struct. 917 (2e3) (2009) 176e182.

[36] A. Qu�emard, J.C. Sacchettini, A. Dessen, C. Vilch�eze, R. Bittman, W.R. Jacobs Jr.,J.S. Blanchard, Enzymatic characterization of the target for isoniazid inMycobacterium tuberculosis, Biochemistry 34 (26) (1995) 8235e8241.

[37] L.A. Basso, R. Zheng, J.M. Musser, W.R. Jacobs Jr., J.S. Blanchard, Mechanisms ofisoniazid resistance in Mycobacterium tuberculosis: enzymatic characteriza-tion of enoyl reductase mutants identified in isoniazid-resistant clinical iso-lates, J. Infect. Dis. 178 (3) (1998) 769e775.

[38] C. Laurence, K.A. Brameld, J. Graton, J.Y. Le Questel, E. Renault, The pKBHxdatabase: toward a better understanding of hydrogen-bond basicity for me-dicinal chemists, J. Med. Chem. 52 (14) (2009) 4073e4086.

[39] T. Matviiuk, F. Rodriguez, N. Saffon, S. Mallet-Ladeira, M. Gorichko, A.L.L.R. deJesus, M.R. Pasca, C. Lherbet, Z. Voitenko, M. Baltas, Design, chemical synthesis

M. Rotta et al. / European Journal of Medicinal Chemistry 90 (2015) 436e447 447

of 3-(9H-fluoren-9-yl)pyrrolidine-2,5-dione derivatives and biological activityagainst enoyl-ACP reductase (InhA) and Mycobacterium tuberculosis, Eur. J.Med. Chem. 70 (2013) 37e48.

[40] T. Matviiuk, G. Mori, C. Lherbet, F. Rodriguez, M.R. Pasca, M. Gorichko,B. Guidetti, Z. Voitenko, M. Baltas, Synthesis of 3-heteryl substitutedpyrrolidine-2,5-diones via catalytic Michael reaction and evaluation of theirinhibitory activity against InhA and Mycobacterium tuberculosis, Eur. J. Med.Chem. 71 (2014) 46e52.

[41] A. Punkvang, P. Saparpakorn, S. Hannongbua, P. Wolschann, H. Berner,P. Pungpo, Insight into crucial inhibitor-enzyme interaction of arylamides asnovel direct inhibitors of the enoyl ACP reductase (InhA) from Mycobacteriumtuberculosis: computer-aided molecular design, Monatsh. Chem. 141 (9)(2010) 1029e1041.

[42] D.T. Clark, J.N. Murrell, J.M. Tedder, 234. The magnitudes and signs of theinductive and mesomeric effects of the halogens, J. Chem. Soc. (1963)1250e1253.

[43] C. Bissantz, B. Kuhn, M. Stahl, A medicinal chemist's guide to molecular in-teractions, J. Med. Chem. 53 (2010) 5061e5084.

[44] W.W. Cleland, P.F. Cook, Enzyme Kinetics and Mechanism, Garland SciencePublishing, New York, 2007, pp. 121e204 (chapter 6).

[45] R.A. Copeland, Evaluation of Enzyme Inhibitors in Drug Discovery: a Guide forMedicinal Chemists and Pharmacologists, first ed., John Wiley & Sons,Hoboken, NJ, 2005 (Chapters 3 and 5).

[46] G.M. Sheldrick, SHELXS-97: Program for Crystal Structure Solution, Universityof G€ottingen, Germany, 1997.

[47] G.M. Sheldrick, SHELXL-97: Program for Crystal Structure Refinement, Uni-versity of G€ottingen, Germany, 1997.

[48] L.J. Farrugia, ORTEP-3 for windows e a version of ORTEP-III with a graphicaluser interface (GUI), J. Appl. Cryst. 30 (5) (1997) 565.

[49] C.S.A. Kumar, S.B.B. Prasad, N.R. Thimmegowda, K.S. Rangappa, S. Chandrappa,S. Naveen, M.A. Sridhar, J.S. Prasad, Synthesis and crystal structure of 1-ben-zenesulfonyl-4-benzhydrylpiperazine, Mol. Cryst. Liq. Cryst. 469 (1) (2007)p.111ep.119.

[50] HYPERCHEM(™) Professional 7.52, Hypercube, Inc., 1115 NW 4th Street,Gainesville, FL 32601, USA, 2002.

[51] G.M. Morris, R. Huey, W. Lindstrom, M.F. Sanner, R.K. Belew, D.S. Goodsell,A.J. Olson, Autodock4 and AutoDockTools4: automated docking with selectivereceptor flexibility, J. Comput. Chem. 16 (2009) 2785e2791.

[52] G.M. Morris, D.S. Goodsell, R.S. Halliday, R. Huey, W.E. Hart, R.K. Belew,A.J. Olson, Automated docking using a Lamarckian genetic algorithm andempirical binding free energy function, J. Comput. Chem. 19 (1998)1639e1662.

[53] S. Parikh, D.P. Moynihan, G. Xiao, P.T. Tonge, Roles of tyrosine 158 and lysine165 in the catalytic mechanism of InhA, the enoyl-ACP reductase fromMycobacterium tuberculosis, Biochemistry 38 (41) (1999) 13623e13634.

[54] M.M. Bradford, A rapid and sensitive method for the quantification ofmicrogram quantities of protein utilizing the principle os protein-dye binding,Anal. Biochem. 72 (1976) 248e254.