Bioorganic & Medicinal Chemistry 23 (2015) 2062–2078
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry journal homepage: www.elsevier.com/locate/bmc
Design and synthesis of novel quinoline–aminopiperidine hybrid analogues as Mycobacterium tuberculosis DNA gyraseB inhibitors Brahmam Medapi a, Janupally Renuka a, Shalini Saxena a, Jonnalagadda Padma Sridevi a, Raghavender Medishetti b,c, Pushkar Kulkarni b,c, Perumal Yogeeswari a, Dharmarajan Sriram a,⇑ a b c
Department of Pharmacy, Birla Institute of Technology & Science-Pilani, Hyderabad Campus, Jawahar Nagar, Shameerpet, R.R. District, Hyderabad 500078, Andhra Pradesh, India Dr Reddy’s Institute of Life Sciences, University of Hyderabad Campus, Gachibowli, Hyderabad 500046, India Zephase Therapeutics (an incubated company at the Dr Reddy’s Institute of Life Sciences), University of Hyderabad Campus, Gachibowli, Hyderabad 500046, India
a r t i c l e
i n f o
Article history: Received 9 October 2014 Revised 2 March 2015 Accepted 3 March 2015 Available online 9 March 2015 Keywords: Tuberculosis DNA gyrase Quinoline Aminopiperidine
a b s t r a c t Antibiotics with good therapeutic value and novel mechanism of action are becoming increasingly important in today’s battle against bacterial resistance. One of the popular targets being DNA gyrase, is currently becoming well-established and clinically validated for the development of novel antibacterials. In the present work, a series of forty eight quinoline–aminopiperidine based urea and thiourea derivatives were synthesized as pharmacophoric hybrids and evaluated for their biological activity. Compound, 1-(4chlorophenyl)-3-(1-(6-methoxy-2-methylquinolin-4-yl)piperidin-4-yl)thiourea (45) was found to exhibit promising in vitro Mycobacterium smegmatis GyrB IC50 of 0.95 ± 0.12 lM and a well correlated Mycobacterium tuberculosis (MTB) DNA gyrase supercoiling IC50 of 0.62 ± 0.16 lM. Further, compound 45 also exhibited commendable MTB MIC, safe eukaryotic cytotoxic profile with no signs of cardiotoxicity in zebrafish ether-a-go-go-related gene (zERG). Ó 2015 Elsevier Ltd. All rights reserved.
1. Introduction Infections caused by mycobacteria continue to be the largest cause of death worldwide, while tuberculosis (TB) claimed 1.3 million deaths globally in 2012.1 With an increase in the emergence of multidrug resistant TB (MDR-TB) and extensively drug resistant TB (XDR-TB) there is an urgent need to develop new anti-mycobacterials. While most of the first-line TB drugs currently in use are at least 40 years old, second-line drugs possessed variable efficacy and showed serious side effects.2 Bedaquiline, an ATP synthase targeted drug was recently approved by FDA for the treatment of MDR-TB. Though drugs like delamanid and fluoroquinolones are in the clinical phase of development, there are no effective antitubercular drugs currently in the market.3–5 Hence design of new inhibitors with a novel mechanism of action against mycobacterial enzymes is essential so as to facilitate the development of efficient anti-TB drugs. Presently, bacterial DNA Gyrase, a type II topoisomerase from GHKL (gyrase, HSP 90, histidine kinase, MutL) enzyme family, remain as one of the most investigated and validated targets for the development of novel antibacterial leads. DNA gyrase plays a crucial role in the bacterial DNA replication cycle and
⇑ Corresponding author. Tel.: +91 40663030506; fax: +91 4066303998. E-mail address: [email protected] (D. Sriram). http://dx.doi.org/10.1016/j.bmc.2015.03.004 0968-0896/Ó 2015 Elsevier Ltd. All rights reserved.
subsequently its absence in the eukaryotic cells makes this enzyme a very suitable target for the development of potential drugs.6 Mycobacterium tuberculosis (Mtb) is unusual that it possess only one type II topoisomerase, DNA gyrase (Topoisomerase IV is absent)7 and apart from its main role as supercoiling enzyme, gyrase also performs enhanced relaxation, DNA cleavage, and decatenation activities.8 DNA gyrase is a holo enzyme with two A and B subunits together forming a functional heterodimer structure (A2B2 complex). GyrA is involved in breakage and reunion of the DNA, while GyrB subunit plays a major role in ATPase activity. Fluoroquinolones are the only class of DNA gyrase inhibitors in clinical practice5 till date that target GyrA domain, while the GyrB portion remains under-explored with greater scope for further development as broad spectrum antibacterials. Though novobiocin was licensed for clinical use in the 1960s, it was withdrawn from the market due to severe safety concern and poor pharmacological properties when administered.9 Though GyrB has been genetically demonstrated to be a bactericidal drug target in MTB, till date there has not been any effective therapeutics developed against TB. This prompted us to make an attempt towards synthesis of quinoline–aminopiperidine hybrid analogues targeting GyrB. Ideally, these compounds displayed good activity profile against MTB GyrB ATPase enzyme and had commendable antimycobacterial activity with less cytotoxic effects in eukaryotic RAW 264.7 mouse macrophages cells. Hence we report herein the synthesis
2063
B. Medapi et al. / Bioorg. Med. Chem. 23 (2015) 2062–2078
with this template.15–17 In continuation to our efforts, we further attempted to design novel MTB gyrase inhibitors by hybridizing aminopiperidine linker (3–5) to quinoline (1–2) core to give a desired compound (6). Thus designed ligands possessed three pharmacophoric units, an aminopiperidine, a linker stacked between a left quinoline core and a substituted aryl moiety at the right side as shown in Figure 1. Quinoline nucleus based antimycobacterials were reported earlier (structures 1 and 2 in Fig. 1) and hence utilized in this study as a left hand core.18 Further, the selection of right hand substituent was purely based on the structure–activity relationship (SAR) inputs from a previous GSK report19 wherein the most potent analogues possessed methoxy, fluoro, nitro, methyl and chloro substituted aryl rings on the right hand side, and hence we retained these analogues as such for derivation of SAR and lead optimization in the present study. The titled compounds were synthesized by a series of reactions that has been delineated in Scheme 1. Ethyl 3-(phenyl/4-methoxy/ 4-fluoro/4-trifluoroimino)butanoates (8a–d) were synthesized as intermediates from corresponding anilines (7a–d) by treating with ethylacetoacetate in presence of magnesium sulfate in ethanol at 90 °C. The intermediate 8 was then converted to corresponding unsubstituted/6-subtituted-quinolin-4-ol (9a–d) by heating it at 75 °C with polyphosphoric acid and phosphorous oxy-chloride
of quinoline–aminopiperidine analogues with improved cardiosafety among N-linked aminopiperidine analogues. As the zebrafish ether-a-go-go-related gene (zERG) was orthologous to the human ether-a-go-go-related gene (hERG), we utilized zebrafish model to predict the effect of test compounds on repolarizingpotassium-current-induced arrhythmia.10,11 2. Results and discussion 2.1. Design and synthesis Molecular hybridization is a rational drug design strategy based on the identification of active pharmacophoric sub-units, which facilitate the formation of molecular hybrids of two or more known bioactive inhibitors by fusion of sub-units while retaining desired characteristics of the original templates. The major aim of designing the hybrid architecture was to develop compounds with improved potency towards the target than the parent lead and lower the side effects. Many scientists have explored hybrid design of newer analogues as potential candidates for biological evaluation. N-linked 4-aminopiperidine derivatives were already reported as DNA gyrase inhibitors12–14 and the concept of molecular hybridization was implemented successfully in our previous work
Quinoline based inhibitors reported as antimycobacterial agents O N N
N H
R1
N N
N
O
O S N R1 H O N
O N
N
1.
2.
Amino-piperidine based inhibitors reported for DNA Gyrase Cl
O NH
Cl NH Ar
N
S O
R1
R
N
N
NH
NH
N
N
N
O
R OH
N O
O
3.
4.
5.
Inhibitor designed through molecular hybridisation R LHS
L IN
KE
N
R H N
H N X
R1
RHS
N X = O, S
6. Figure 1. Chemical structure of previously reported synthetic antimycobacterials bearing quinoline nucleus (1–2), DNA gyrase inhibitors bearing aminopiperidine nucleus (3–5) and the inhibitor designed through molecular hybridization (6).
2064
B. Medapi et al. / Bioorg. Med. Chem. 23 (2015) 2062–2078
O O
O R
a NH 2 7a-d
R
OH O
N 8a-d
R
Cl c
b
R
N 9a-d
X NH 2
NH
N
N d
R
R1
e
R
N
N
10a-d
11a-d
12a-d
X= O, S R= H, CH3 O, F, CF3 R 1= Subti. Aryl ring
NH
N f
N
N H
R N
13-36 X=O 37-60 X=S
Scheme 1. Synthetic protocol used to get the target compounds. Reagents and conditions: (a) ethylacetoacetate, anhyd MgSO4, EtOH, reflux, 6 h; (b) PPA, POCl3 (catalytic), 75 °C, 12 h; (c) POCl3, reflux, 5 h; (d) 4-N-Boc amino piperidine, DMF, DIPEA, 130 °C, 3 h; (e) DCM, TFA, 0 °C–rt, 3 h; (f) Et3N, EtOH, substituted aryl isocyanates/ isothiocyanates, 50 °C, 8 h.
(catalytic) for 8 h. The required key intermediate tert-butyl (unsubstituted/6-subtituted-2-methylquinolin-4-yl)piperidin-4-yl)carbamate 11 was obtained by heating 9 with phosphorous oxy-chloride, followed by heating the chloro derivative 10 with N,N-diisopropylethylamine (DIPEA) and 4-N-Boc aminopiperidine in N,N-dimethylformamide (DMF) as solvent at 130 °C for 3 h under N2" atmosphere. The final library of compounds (13–60) were obtained with sequential Boc-deprotection of intermediate 11 with trifluoroacetic acid and reaction with different substituted aryl isocyanates/iso-thiocyanates in presence of triethylamine (Et3N) in ethanol heating at 55 °C for 8 h. Purity of the synthesized compounds was checked using LCMS, structure characterized by 1 H NMR, 13C NMR and elemental analysis. In the spectra 1 ( H NMR and 13C NMR), the signals of the respective protons of the synthesized derivatives were verified on the basis of their chemical shifts, multiplicities, and coupling constants. The elemental analysis results were within ±0.4% of the theoretical values. 2.2. Pharmacological evaluation DNA GyrB from Mycobacterium smegmatis (MS) was utilized for performing ATPase assay. GryB gene was cloned into a prokaryotic expression vector pQE2 and was expressed in BL21 (DE3) pLysS competent cells, and the expressed protein was further induced with 0.1 mM IPTG (Isopropyl-b-D-thiogalactopyranoside), purified by Ni-NTA column and further identified by 10% SDS–PAGE. The kinetics of the MS DNA GyrB enzyme followed greater than first-order kinetics.20 However at a constant enzyme concentration of 15 lM, a hyperbolic dependence of rate on substrate (ATP) concentration was observed as the GyrB ATPase activity of MS did not follow app Michaelis–Menten kinetics. Thus the apparent Kapp M and Vmax determined experimentally were found to be 300 lM and 2.1 lmol/s, respectively, as illustrated in Figure 2. The assay was performed on MS DNA GyrB protein instead of MTB GyrB due to MTB’s slow growing nature when compared to MS [21]. Moreover, the sequence similarity between GyrB protein domains of both the organisms was found to be 87%, with most of the catalytic site residues being conserved which confirmed a higher degree of homology in the ATP binding pocket. Novobiocin was used as a standard inhibitor in these assays, for which a correlation of 3–5 fold variation in IC50 values between the ATPase activity of MS and the supercoiling activity of MTB was observed which had given us enough confidence to utilize MS as a surrogate enzyme for DNA Gyr B ATPase activity.21 All the forty eight synthesized compounds were biologically evaluated for their in vitro MS GyrB ATPase assay, and their IC50 values were found in the range between 0.95 ± 0.12 to 68.22 ± 2.8 lM. Dose response curve was plotted for the most potent lead 45 with an IC50 of 0.95 ± 0.12 lM in MS GyrB using GraphPad Prism software
(GraphPad Software Inc., La Jolla, CA) by taking log (inhibitor concentration) on the X-axis and its response (% inhibition) on the Y-axis as shown in Figure 3, while the standard novobiocin showed an IC50 of 180 ± 3.9 nM. Further, compound 45 was found to exhibit selective inhibition of mycobacterial GyrB when compared to other bacterial GyrB proteins of Staphylococcus aureus and Escherichia coli (Inspiralis) tested at 100 lM concentration confirming their specificity towards mycobacterium species. In order to fully explore the structure–activity relationship associated with the Mycobacterium smegmatis GyrB inhibitors, compounds were docked to the GyrB ATPase domain of MS retrieved from protein data bank (PDB ID: 4B6C) using extra precision mode (XP) of Glide module. The reference ligand 6-(3,4dimethylphenyl)-3-[[4-[3-(4-methylpiperazin-yl)propoxy]phenyl] amino]pyrazine-2-carboxamide was further re-docked with the active site residues of the MS protein to validate the active site cavity. The ligand exhibited a Glide score of 6.93 kcal/mol and was placed in the active site cavity with amino acids like Asn52, Ile84, Val99, Val98, Asp97, Pro85, Arg141, Arg82, Glu56, Ala53, Asp79, Ile171, Val99, Val123, Ser126, Val128, and Glu4. Re-docking results showed that the crystal compound exhibited similar interactions as that of the original crystal structure and exhibited RMSD of 0.86 Å. The crystal ligand 6-(3,4-dimethylphenyl)-3-[[4-[3-(4methylpiperazin-yl)propoxy]phenyl]amino]pyrazine-2-carboxamide exhibited two important hydrogen bonding interactions in the active site pocket, one between amino group of the carboxamide moiety and oxygen atom of Asp79 while the other was seen
Figure 2. ATPase activity of Mycobacterium smegmatis DNA gyraseB protein as a function of substrate (ATP) concentration at a constant enzyme concentration (15 lM).
B. Medapi et al. / Bioorg. Med. Chem. 23 (2015) 2062–2078
Figure 3. Dose response curve of compound 45 with Mycobacterium smegmatis DNA GryB ATPase assay at six various concentrations.
between the nitrogen atom of piperazine and hydrogen atom of guanidine moiety of Arg82. The molecule was stabilized by hydrophobic interaction in the hydrophobic pocket, considered to be important for bringing in specificity observed at the enzyme level as per Figure 4. In the docking studies, all the active compounds from the synthesized compounds, exhibited good docking score with significant polar and non-polar contacts, hydrogen bond interactions with the relevant amino acids and various hydrophobic interactions. Forty eight compounds were then screened for the GyrB inhibitory activity. With respect to structure-activity relationship study, on the left hand core, we attempted phenyl as in 13–18, OCH3 in 19–24, F in 25–30 and CF3 in 31–36 ring system in the sixth position of quinoline nucleus, while the right hand core was substituted with simple/4-F/4-Cl/4-NO2/4-methyl/4-methoxy phenyl ring. In the first subset, various substitutions were attempted on the R and R1 position of quinoline nucleus having oxygen as the X
2065
position (13–36). The substitutions OCH3 and F at R position exhibited good activity, while simple phenyl group showed moderate activity and CF3 was found to be detrimental to activity as compared to other substitutions. These derivatives 13–36 were then evaluated for the Gyr B inhibitory potency using the malachite green assay as described in the materials and methods section. Among the twenty four derivatives evaluated, four compounds showed IC50 less than 10 lM as shown in Table 1. Among the OCH3 substituted analogues at R position, the 4-NO2 of 22 and 4methoxy phenyl of 24 at R1 position showed good activity with IC50s of 6.991 ± 0.85 and 9.255 ± 0.42 lM, respectively. A closer look into the interaction profile of these molecules revealed that the NH of quinoline nucleus was involved in a prominent hydrogen bonding interaction with Asp79, analogous to the one observed in the crystal ligand. Compound 22 also showed additional polar contact with Asn52, due to which was found to display good binding with a docking score of 6.74 kcal/mol, similar to compound 24 that showed hydrogen bonds with Asp79 and Glu48. Both the compounds were further stabilized by various hydrophobic interactions with Val125, Val128, Val49, Ile171, Val77, Ala53, Val99, Met100 and Ile84 as shown in Supplementary information Figure S1. The F substituted compounds at R position as in 25–30, 4-fluoro and 4-nitro phenyl substituted compounds at R1 (26 and 28) turned out to be more promising with IC50s of 1.78 ± 0.23 and 6.826 ± 0.62 lM respectively. Both the compounds showed good binding with a docking score of 6.12 and 5.99 kcal/mol, respectively. The NH group of quinoline nucleus exhibited hydrogen bonding interaction with Asp79, whereas 4-fluoro and 4-nitro groups were exposed to solvent accessible area. Further the compounds were also found to be stabilized by hydrophobic interactions with Val125, Val128, Val49, Ile171, Val77, Ala53, Ile84, Val98, Val99, Met100 and Val123 as shown in Supplementary information Figure S2. However it was surprising to note that replacement of F substituent (26) with chloro (27) on the R1 position led to a drastic reduction in the activity as indicated in their interaction profile where the molecule was found to be in a slightly different orientation/pose compared to that of compound 26 thereby losing
Figure 4. Interaction profile picture of ligand 6-(3,4-dimethylphenyl)-3-[[4-[3-(4-methylpiperazin-yl)propoxy]phenyl]amino]pyrazine-2-carboxamide in the active site of Mycobacterium smegmatis GyrB protein.
2066
B. Medapi et al. / Bioorg. Med. Chem. 23 (2015) 2062–2078
Table 1 In vitro biological evaluation of the synthesized forty eight compounds
X N H
HN
R1
N R N 13-36 X=O 37-60 X=S Compd
R
R1
MS GyrB assay (IC50)
MTB supercoiling assay (IC50)
MTB MIC (lM)
Cytotoxicity at 50 lM (% Inhib.)
13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
H H H H H H OCH3 OCH3 OCH3 OCH3 OCH3 OCH3 F F F F F F CF3 CF3 CF3 CF3 CF3 CF3 H H H H H H OCH3 OCH3 OCH3 OCH3 OCH3 OCH3 F F F F F F CF3 CF3 CF3 CF3 CF3 CF3
Phenyl 4-Fluoro phenyl 4-Chloro phenyl 4-Nitro phenyl p-Tolyl 4-Methoxy phenyl Phenyl 4-Fluoro phenyl 4-Chloro phenyl 4-Nitro phenyl p-Tolyl 4-Methoxy phenyl Phenyl 4-Fluoro phenyl 4-Chloro phenyl 4-Nitro phenyl p-Tolyl 4-Methoxy phenyl Phenyl 4-Fluoro phenyl 4-Chloro phenyl 4-Nitro phenyl p-Tolyl 4-Methoxy phenyl Phenyl 4-Fluoro phenyl 4-Chloro phenyl 4-Nitro phenyl p-Tolyl 4-Methoxy phenyl Phenyl 4-Fluoro phenyl 4-Chloro phenyl 4-Nitro phenyl p-Tolyl 4-Methoxy phenyl Phenyl 4-Fluoro phenyl 4-Chloro phenyl 4-Nitro phenyl p-Tolyl 4-Methoxy phenyl Phenyl 4-Fluoro phenyl 4-Chloro phenyl 4-Nitro phenyl p-Tolyl 4-Methoxy phenyl Novobiocin Isoniazid Rifampicin Moxifloxacin
17.26 ± 0.87 34.88 ± 1.81 12.94 ± 1.13 16.95 ± 0.59 20.48 ± 1.52 22.02 ± 1.7 12.18 ± 0.94 18.48 ± 0.44 16.24 ± 1.13 6.99 ± 0.85 11.48 ± 0.32 9.25 ± 0.42 20.52 ± 1.26 1.78 ± 0.23 48.98 ± 2.66 6.83 ± 0.62 15.05 ± 0.55 10.12 ± 0.85 14.9 ± 0.53 19.77 ± 1.1 33.98 ± 2.3 47.63 ± 2.1 16.6 ± 1.3 41.2 ± 0.96 12.7 ± 2.1 29.81 ± 1.5 12.31 ± 0.92 3.77 ± 0.24 7.26 ± 0.33 19.42 ± 1.91 2.14 ± 0.15 8.7 ± 0.22 0.95 ± 0.12 2.5 ± 0.26 11.14 ± 0.39 10.1 ± 1.26 68.22 ± 2.8 45.92 ± 1.92 41.1 ± 0.98 56.15 ± 1.3 59.3 ± 0.65 39.51 ± 0.87 39.04 ± 1.3 48.5 ± 2.9 16.53 ± 0.91 41.8 ± 0.44 24.68 ± 1.8 32.56 ± 0.52 180 ± 3.9 nM nd nd >50
9.79 ± 0.23 19.74 ± 0.27 8.2 ± 0.24 10.63 ± 0.18 14.82 ± 0.32 18.54 ± 0.41 9.32 ± 0.22 14.8 ± 0.17 14.33 ± 0.21 5.33 ± 0.16 8.63 ± 0.31 5.86 ± 0.22 9.84 ± 0.29 0.78 ± 0.13 29.63 ± 0.65 2.14 ± 0.18 8.83 ± 0.44 7.62 ± 0.31 8.3 ± 0.22 11.8 ± 0.31 17.3 ± 0.33 22.3 ± 0.54 9.2 ± 0.2 19.8 ± 0.36 9.4 ± 0.2 16.72 ± 0.2 8.12 ± 0.13 1.21 ± 0.2 2.66 ± 0.32 13.51 ± 0.35 0.91 ± 0.05 2.31 ± 0.31 0.62 ± 0.16 0.88 ± 0.04 7.94 ± 0.3 6.17 ± 0.23 39.44 ± 0.43 23.54 ± 0.53 21.5 ± 0.55 25.5 ± 0.4 25.41 ± 0.2 17.84 ± 0.31 29.4 ± 0.43 22.84 ± 0.77 4.12 ± 0.14 7.54 ± 0.18 5.1 ± 0.22 34.5 ± 0.87 46 ± 10 nM nd nd 11.2 ± 0.36
25.84 34.09 7.91 15.41 24.14 32.02 32.02 25.66 17.35 14.70 15.45 12.57 33.03 3.82 50.78 9.52 15.97 7.65 29.35 24.89 48.53 52.2 28.05 54.26 16.6 36.14 15.21 7.65 15.68 30.22 3.81 14.72 3.47 1.72 7.28 15.84 63.26 63.26 31.68 63.26 67.94 31.68 35.82 66.53 30.68 51.07 31.68 66.88 >200 0.66 0.23 1.26
33.87 49.39 40.07 30.37 51.44 49.54 66.47 17.31 42.40 35.01 28.17 19.58 48.88 25.39 49.67 47.89 39.73 49.67 20.13 39.75 29.11 48.24 16.23 29.00 11.82 32.93 20.83 31.83 21.63 21.93 41.89 29.28 29.73 24.93 44.93 31.82 37.38 33.93 32.28 40.73 23.93 21.29 23.28 50.72 11.82 19.80 19.27 41.83 nd nd nd nd
ND indicates not determined.
the hydrogen bonding interaction with Asp79 as presented in Supplementary information Figure S3. In the next subset, CF3 substitution at R position as in 31–36 was found to exhibit moderate
activity against MS GyrB ATPase as well as MTB DNA supercoiling. The simple phenyl and 4-methyl phenyl substitution at R1 position as in 31 and 35 respectively showed moderate activity when
B. Medapi et al. / Bioorg. Med. Chem. 23 (2015) 2062–2078
2067
Figure 5. The interaction profile pictures of compounds 43 and 46 at the active pocket of GyrB of MS with respect to the substitutions at R1 position with phenyl and 4nitrophenyl with CH3O substituent at R position.
compared with other compounds from this series (IC50s of 14.9 ± 0.53 and 16.6 ± 1.3 lM in MS Gyr B, respectively, and 8.3 ± 0.22 and 9.2 ± 0.2 lM, respectively, in supercoiling assays). Other compounds from this set namely 33, 34 and 36, where the R1 position was substituted with 4-chloro/nitro and 4-methoxy phenyl were found to be less active with IC50 P30 lM. Docking studies revealed that these compounds oriented in a different manner, and a closer look at the interaction pattern of the least active analogues in this class showed that the introduction of 4-chlorophenyl as in compound 33 changed the orientation in the active site of protein and also CF3 group was found to be placed outside the pocket which could be the reason behind their lesser activity, which was also supported by a low docking score of 4.12 kcal/mol as presented in Supplementary information Figure S4. The other two compounds substituted with 4-nitro, and 4-methoxy phenyl at R1 were also not active due to the lack of hydrogen bonding interaction with Asp79 though hydrophobic interactions were present as per Supplementary information Figure S4. In the next set of compounds, 37–60, the effects of similar substituents were explored at R and R1 positions on left and right sides of quinoline nucleus having S (sulfur) group instead of O (oxygen) at the X position. Among the 24 synthesized compounds, 6 compounds inhibited Gyr B enzyme activity with IC50s less than 10 lM while compound 45, 1-(4-chlorophenyl)-3-[1-(6-methoxy-2-methylquinolin-4-yl) piperidin-4-yl]thiourea was found to be the most promising with an GyrB IC50 of 0.95 ± 0.12 lM which was well correlated with DNA supercoiling IC50 of 0.62 ± 0.16 lM. In this set of compounds simple phenyl and OCH3 at R position and various simple/4-F/4Cl/4-NO2/4-methyl/4-methoxy phenyl ring at R1 position exhibited good activity, while CF3 substituted compounds showed moderate activity and F substitution was found to be detrimental to activity. In the first subset, simple phenyl was explored at R position and simple 4-F/4-Cl/4-NO2/4-methyl/4-methoxy phenyl rings at R1 position similar to the previous set of compounds (37–42). Here, 4-nitro (40) and 4-methyl phenyl substituted compounds (41) were found to exhibit good activity among the other compounds with IC50s of 3.77 ± 0.24 and 7.26 ± 0.33 lM, respectively. In silico analysis of these compounds indicated that the molecules oriented in a similar manner as that of lead compound retaining hydrogen bonding with Asp79 and other significant polar interactions with
Glu48 and Asn52 amino acid residues. However their good GyrB potency could be accounted due to the hydrophobic interactions which were earlier demonstrated to be crucial for activity and specificity observed at the enzyme level as shown in Supplementary information Figure S5. Both the compounds were well inserted into the active site pocket which made these compounds better among all the compounds that showed good binding with the receptor (6.12 and 6.32 kcal/mol). On the other hand, introduction of 4-fluorophenyl group at R1 position (38) oriented the compound differently, thereby losing key hydrogen bonding interactions which probably resulted in their diminished activity as shown in Supplementary information Figure S6. The OCH3 substituted compounds (43–48) emerged as the most active compounds as they exhibited activity in the range of 0.95 ± 0.12 to 11.14 ± 0.39 lM in MS GyrB and 0.62 ± 0.16 to 7.94 ± 0.3 lM in MTB supercoiling assay. The binding analysis of these compounds within the MS protein in the ATP-binding site revealed that the most promising compounds from the series were 43, 45 and 46 as they showed good binding characteristics with the protein with docking score ranging from 6.87 to 6.33 kcal/kmol. Also, these compounds were well oriented at the active site cavity of the protein as shown in Figures 5 and 6. A closer analysis revealed that all the three active compounds interacted with Asp79 through hydrogen bonding which was important for retaining the bioactivity. Compound 43 was found to show an additional interaction with Glu48, while compound 45 was involved in hydrogen bonding Val99 in addition to Glu48. Furthermore, compound 45 was found to be stabilized by hydrophobic interactions with Ala53, Val128, Val50, Val49, Met100, Ile171 and Val99. Binding pattern of the most active compound 45 is represented in Figure 6. The next set of compounds, where R position was substituted with F or CF3 group (49– 60) were found to be less active when compared to simple phenyl or OCH3 substituted compounds. Binding analysis of the least active, F substituted compounds (49–54) revealed that all these compounds lacked hydrogen bonding interactions with the protein. This was clearly evident with their lesser docking scores of 4.12 to 3.99 kcal/mol. Similarly among the CF3 substituted compounds, only 4-methoxy and 4-chloro substitutions were found to exhibit moderate activity (IC50 of 12.56 ± 0.52 and 16.53 ± 0.91 lM) while other compounds did not possess bioactivity. Analysis of this set
2068
B. Medapi et al. / Bioorg. Med. Chem. 23 (2015) 2062–2078
Figure 6. Interaction profile pictures of the most active compound 45 at the active pocket of GyrB of MS.
Figure 7. Picture depicting the supercoiling assay of compound 45 at four different concentrations of 3.8, 1.9, 0.95 and 0.5 lM; R-Relaxed DNA (substrate + DMSO); CControl (Relaxed DNA substrate + DNA Gyrase + DMSO); N-Novobiocin.
of compounds revealed that the compounds 57 and 60 were well inserted into active site pocket of MS protein and the NH group of quinoline ring was involved in hydrogen bonding interaction with Asp79, while NH of 4-aminopiperidine ring interacted with Glu48 as shown in Supplementary information Figure S7. Binding analysis of least active compounds revealed that the compounds were oriented differently when compared to active compounds. 4-Flouro or nitro group at R1 position was displaced out of the pocket due to which the activity of molecules could have been affected, as shown in Supplementary information Figure S8. Compounds that inhibited ATPase activity of GyrB protein subsequently were expected to also inhibit DNA supercoiling, as Gyr A and GyrB domains constituted as holo enzyme DNA gyrase. Supercoiling assay was performed using the DNA supercoiling kit (Inspiralis Pvt. limited, Norwich) as per the protocol described in Section 4.4. Ideally, to eliminate the possibilities of aggregation and auto-fluorescence artifacts, non-specific inhibiting detergents were added in our biological experiments.22 All the reactions were performed against MTB DNA gyrase enzyme dose dependently at eight concentrations as 100, 50, 25, 12.5, 6.25, 3.125, 1.56 and 0.75 lM, in which the most active compound 45 showed an IC50 of 0.62 ± 0.16 lM as illustrated in Table 1. Novobiocin was employed as a standard compound in this assay for comparison.
Subsequently, IC50 were calculated based on relative quantification using Image lab software, Bio-Rad (Fig. 7). Further, the synthesized derivatives were screened for their in vitro anti-tubercular activity against Mycobacterium tuberculosis H37Rv strain (ATCC 27294) by agar dilution method as described in Section 4.23 The minimum inhibitory concentration (MIC) was determined for each compound and was found to be within a range of 0.78–50 lg/mL, as tested in seven different concentrations in triplicate. Overall, a good correlation was observed between the MTB DNA gyrase supercoiling IC50 values and the in vitro MTB MICs. However a slight deviation in the MS Gyr B IC50 was observed, probably owing to the slight difference in the proteins of the two organisms, as MIC was evaluated on MTB (M. tuberculosis H37Rv) and the Gyr B assay was performed on MS protein respectively.21 Ethambutol, ofloxacin and novobiocin were used as standard compounds. All the synthesized compounds showed good activity against MTB with MIC values ranging from 1.72 to 66.8 lM as shown in Table 1. Closer observation of MIC values confirmed that these compounds were better in activities than the first-line anti-tubercular drug ethambutol (MIC = 9.84 lM) but were less active when compared to isoniazid and moxifloxacin. The standard compound novobiocin was found to be inactive up to 200 lM concentration. Twenty compounds showed commendable MIC
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry journal homepage: www.elsevier.com/locate/bmc
Design and synthesis of novel quinoline–aminopiperidine hybrid analogues as Mycobacterium tuberculosis DNA gyraseB inhibitors Brahmam Medapi a, Janupally Renuka a, Shalini Saxena a, Jonnalagadda Padma Sridevi a, Raghavender Medishetti b,c, Pushkar Kulkarni b,c, Perumal Yogeeswari a, Dharmarajan Sriram a,⇑ a b c
Department of Pharmacy, Birla Institute of Technology & Science-Pilani, Hyderabad Campus, Jawahar Nagar, Shameerpet, R.R. District, Hyderabad 500078, Andhra Pradesh, India Dr Reddy’s Institute of Life Sciences, University of Hyderabad Campus, Gachibowli, Hyderabad 500046, India Zephase Therapeutics (an incubated company at the Dr Reddy’s Institute of Life Sciences), University of Hyderabad Campus, Gachibowli, Hyderabad 500046, India
a r t i c l e
i n f o
Article history: Received 9 October 2014 Revised 2 March 2015 Accepted 3 March 2015 Available online 9 March 2015 Keywords: Tuberculosis DNA gyrase Quinoline Aminopiperidine
a b s t r a c t Antibiotics with good therapeutic value and novel mechanism of action are becoming increasingly important in today’s battle against bacterial resistance. One of the popular targets being DNA gyrase, is currently becoming well-established and clinically validated for the development of novel antibacterials. In the present work, a series of forty eight quinoline–aminopiperidine based urea and thiourea derivatives were synthesized as pharmacophoric hybrids and evaluated for their biological activity. Compound, 1-(4chlorophenyl)-3-(1-(6-methoxy-2-methylquinolin-4-yl)piperidin-4-yl)thiourea (45) was found to exhibit promising in vitro Mycobacterium smegmatis GyrB IC50 of 0.95 ± 0.12 lM and a well correlated Mycobacterium tuberculosis (MTB) DNA gyrase supercoiling IC50 of 0.62 ± 0.16 lM. Further, compound 45 also exhibited commendable MTB MIC, safe eukaryotic cytotoxic profile with no signs of cardiotoxicity in zebrafish ether-a-go-go-related gene (zERG). Ó 2015 Elsevier Ltd. All rights reserved.
1. Introduction Infections caused by mycobacteria continue to be the largest cause of death worldwide, while tuberculosis (TB) claimed 1.3 million deaths globally in 2012.1 With an increase in the emergence of multidrug resistant TB (MDR-TB) and extensively drug resistant TB (XDR-TB) there is an urgent need to develop new anti-mycobacterials. While most of the first-line TB drugs currently in use are at least 40 years old, second-line drugs possessed variable efficacy and showed serious side effects.2 Bedaquiline, an ATP synthase targeted drug was recently approved by FDA for the treatment of MDR-TB. Though drugs like delamanid and fluoroquinolones are in the clinical phase of development, there are no effective antitubercular drugs currently in the market.3–5 Hence design of new inhibitors with a novel mechanism of action against mycobacterial enzymes is essential so as to facilitate the development of efficient anti-TB drugs. Presently, bacterial DNA Gyrase, a type II topoisomerase from GHKL (gyrase, HSP 90, histidine kinase, MutL) enzyme family, remain as one of the most investigated and validated targets for the development of novel antibacterial leads. DNA gyrase plays a crucial role in the bacterial DNA replication cycle and
⇑ Corresponding author. Tel.: +91 40663030506; fax: +91 4066303998. E-mail address: [email protected] (D. Sriram). http://dx.doi.org/10.1016/j.bmc.2015.03.004 0968-0896/Ó 2015 Elsevier Ltd. All rights reserved.
subsequently its absence in the eukaryotic cells makes this enzyme a very suitable target for the development of potential drugs.6 Mycobacterium tuberculosis (Mtb) is unusual that it possess only one type II topoisomerase, DNA gyrase (Topoisomerase IV is absent)7 and apart from its main role as supercoiling enzyme, gyrase also performs enhanced relaxation, DNA cleavage, and decatenation activities.8 DNA gyrase is a holo enzyme with two A and B subunits together forming a functional heterodimer structure (A2B2 complex). GyrA is involved in breakage and reunion of the DNA, while GyrB subunit plays a major role in ATPase activity. Fluoroquinolones are the only class of DNA gyrase inhibitors in clinical practice5 till date that target GyrA domain, while the GyrB portion remains under-explored with greater scope for further development as broad spectrum antibacterials. Though novobiocin was licensed for clinical use in the 1960s, it was withdrawn from the market due to severe safety concern and poor pharmacological properties when administered.9 Though GyrB has been genetically demonstrated to be a bactericidal drug target in MTB, till date there has not been any effective therapeutics developed against TB. This prompted us to make an attempt towards synthesis of quinoline–aminopiperidine hybrid analogues targeting GyrB. Ideally, these compounds displayed good activity profile against MTB GyrB ATPase enzyme and had commendable antimycobacterial activity with less cytotoxic effects in eukaryotic RAW 264.7 mouse macrophages cells. Hence we report herein the synthesis
2063
B. Medapi et al. / Bioorg. Med. Chem. 23 (2015) 2062–2078
with this template.15–17 In continuation to our efforts, we further attempted to design novel MTB gyrase inhibitors by hybridizing aminopiperidine linker (3–5) to quinoline (1–2) core to give a desired compound (6). Thus designed ligands possessed three pharmacophoric units, an aminopiperidine, a linker stacked between a left quinoline core and a substituted aryl moiety at the right side as shown in Figure 1. Quinoline nucleus based antimycobacterials were reported earlier (structures 1 and 2 in Fig. 1) and hence utilized in this study as a left hand core.18 Further, the selection of right hand substituent was purely based on the structure–activity relationship (SAR) inputs from a previous GSK report19 wherein the most potent analogues possessed methoxy, fluoro, nitro, methyl and chloro substituted aryl rings on the right hand side, and hence we retained these analogues as such for derivation of SAR and lead optimization in the present study. The titled compounds were synthesized by a series of reactions that has been delineated in Scheme 1. Ethyl 3-(phenyl/4-methoxy/ 4-fluoro/4-trifluoroimino)butanoates (8a–d) were synthesized as intermediates from corresponding anilines (7a–d) by treating with ethylacetoacetate in presence of magnesium sulfate in ethanol at 90 °C. The intermediate 8 was then converted to corresponding unsubstituted/6-subtituted-quinolin-4-ol (9a–d) by heating it at 75 °C with polyphosphoric acid and phosphorous oxy-chloride
of quinoline–aminopiperidine analogues with improved cardiosafety among N-linked aminopiperidine analogues. As the zebrafish ether-a-go-go-related gene (zERG) was orthologous to the human ether-a-go-go-related gene (hERG), we utilized zebrafish model to predict the effect of test compounds on repolarizingpotassium-current-induced arrhythmia.10,11 2. Results and discussion 2.1. Design and synthesis Molecular hybridization is a rational drug design strategy based on the identification of active pharmacophoric sub-units, which facilitate the formation of molecular hybrids of two or more known bioactive inhibitors by fusion of sub-units while retaining desired characteristics of the original templates. The major aim of designing the hybrid architecture was to develop compounds with improved potency towards the target than the parent lead and lower the side effects. Many scientists have explored hybrid design of newer analogues as potential candidates for biological evaluation. N-linked 4-aminopiperidine derivatives were already reported as DNA gyrase inhibitors12–14 and the concept of molecular hybridization was implemented successfully in our previous work
Quinoline based inhibitors reported as antimycobacterial agents O N N
N H
R1
N N
N
O
O S N R1 H O N
O N
N
1.
2.
Amino-piperidine based inhibitors reported for DNA Gyrase Cl
O NH
Cl NH Ar
N
S O
R1
R
N
N
NH
NH
N
N
N
O
R OH
N O
O
3.
4.
5.
Inhibitor designed through molecular hybridisation R LHS
L IN
KE
N
R H N
H N X
R1
RHS
N X = O, S
6. Figure 1. Chemical structure of previously reported synthetic antimycobacterials bearing quinoline nucleus (1–2), DNA gyrase inhibitors bearing aminopiperidine nucleus (3–5) and the inhibitor designed through molecular hybridization (6).
2064
B. Medapi et al. / Bioorg. Med. Chem. 23 (2015) 2062–2078
O O
O R
a NH 2 7a-d
R
OH O
N 8a-d
R
Cl c
b
R
N 9a-d
X NH 2
NH
N
N d
R
R1
e
R
N
N
10a-d
11a-d
12a-d
X= O, S R= H, CH3 O, F, CF3 R 1= Subti. Aryl ring
NH
N f
N
N H
R N
13-36 X=O 37-60 X=S
Scheme 1. Synthetic protocol used to get the target compounds. Reagents and conditions: (a) ethylacetoacetate, anhyd MgSO4, EtOH, reflux, 6 h; (b) PPA, POCl3 (catalytic), 75 °C, 12 h; (c) POCl3, reflux, 5 h; (d) 4-N-Boc amino piperidine, DMF, DIPEA, 130 °C, 3 h; (e) DCM, TFA, 0 °C–rt, 3 h; (f) Et3N, EtOH, substituted aryl isocyanates/ isothiocyanates, 50 °C, 8 h.
(catalytic) for 8 h. The required key intermediate tert-butyl (unsubstituted/6-subtituted-2-methylquinolin-4-yl)piperidin-4-yl)carbamate 11 was obtained by heating 9 with phosphorous oxy-chloride, followed by heating the chloro derivative 10 with N,N-diisopropylethylamine (DIPEA) and 4-N-Boc aminopiperidine in N,N-dimethylformamide (DMF) as solvent at 130 °C for 3 h under N2" atmosphere. The final library of compounds (13–60) were obtained with sequential Boc-deprotection of intermediate 11 with trifluoroacetic acid and reaction with different substituted aryl isocyanates/iso-thiocyanates in presence of triethylamine (Et3N) in ethanol heating at 55 °C for 8 h. Purity of the synthesized compounds was checked using LCMS, structure characterized by 1 H NMR, 13C NMR and elemental analysis. In the spectra 1 ( H NMR and 13C NMR), the signals of the respective protons of the synthesized derivatives were verified on the basis of their chemical shifts, multiplicities, and coupling constants. The elemental analysis results were within ±0.4% of the theoretical values. 2.2. Pharmacological evaluation DNA GyrB from Mycobacterium smegmatis (MS) was utilized for performing ATPase assay. GryB gene was cloned into a prokaryotic expression vector pQE2 and was expressed in BL21 (DE3) pLysS competent cells, and the expressed protein was further induced with 0.1 mM IPTG (Isopropyl-b-D-thiogalactopyranoside), purified by Ni-NTA column and further identified by 10% SDS–PAGE. The kinetics of the MS DNA GyrB enzyme followed greater than first-order kinetics.20 However at a constant enzyme concentration of 15 lM, a hyperbolic dependence of rate on substrate (ATP) concentration was observed as the GyrB ATPase activity of MS did not follow app Michaelis–Menten kinetics. Thus the apparent Kapp M and Vmax determined experimentally were found to be 300 lM and 2.1 lmol/s, respectively, as illustrated in Figure 2. The assay was performed on MS DNA GyrB protein instead of MTB GyrB due to MTB’s slow growing nature when compared to MS [21]. Moreover, the sequence similarity between GyrB protein domains of both the organisms was found to be 87%, with most of the catalytic site residues being conserved which confirmed a higher degree of homology in the ATP binding pocket. Novobiocin was used as a standard inhibitor in these assays, for which a correlation of 3–5 fold variation in IC50 values between the ATPase activity of MS and the supercoiling activity of MTB was observed which had given us enough confidence to utilize MS as a surrogate enzyme for DNA Gyr B ATPase activity.21 All the forty eight synthesized compounds were biologically evaluated for their in vitro MS GyrB ATPase assay, and their IC50 values were found in the range between 0.95 ± 0.12 to 68.22 ± 2.8 lM. Dose response curve was plotted for the most potent lead 45 with an IC50 of 0.95 ± 0.12 lM in MS GyrB using GraphPad Prism software
(GraphPad Software Inc., La Jolla, CA) by taking log (inhibitor concentration) on the X-axis and its response (% inhibition) on the Y-axis as shown in Figure 3, while the standard novobiocin showed an IC50 of 180 ± 3.9 nM. Further, compound 45 was found to exhibit selective inhibition of mycobacterial GyrB when compared to other bacterial GyrB proteins of Staphylococcus aureus and Escherichia coli (Inspiralis) tested at 100 lM concentration confirming their specificity towards mycobacterium species. In order to fully explore the structure–activity relationship associated with the Mycobacterium smegmatis GyrB inhibitors, compounds were docked to the GyrB ATPase domain of MS retrieved from protein data bank (PDB ID: 4B6C) using extra precision mode (XP) of Glide module. The reference ligand 6-(3,4dimethylphenyl)-3-[[4-[3-(4-methylpiperazin-yl)propoxy]phenyl] amino]pyrazine-2-carboxamide was further re-docked with the active site residues of the MS protein to validate the active site cavity. The ligand exhibited a Glide score of 6.93 kcal/mol and was placed in the active site cavity with amino acids like Asn52, Ile84, Val99, Val98, Asp97, Pro85, Arg141, Arg82, Glu56, Ala53, Asp79, Ile171, Val99, Val123, Ser126, Val128, and Glu4. Re-docking results showed that the crystal compound exhibited similar interactions as that of the original crystal structure and exhibited RMSD of 0.86 Å. The crystal ligand 6-(3,4-dimethylphenyl)-3-[[4-[3-(4methylpiperazin-yl)propoxy]phenyl]amino]pyrazine-2-carboxamide exhibited two important hydrogen bonding interactions in the active site pocket, one between amino group of the carboxamide moiety and oxygen atom of Asp79 while the other was seen
Figure 2. ATPase activity of Mycobacterium smegmatis DNA gyraseB protein as a function of substrate (ATP) concentration at a constant enzyme concentration (15 lM).
B. Medapi et al. / Bioorg. Med. Chem. 23 (2015) 2062–2078
Figure 3. Dose response curve of compound 45 with Mycobacterium smegmatis DNA GryB ATPase assay at six various concentrations.
between the nitrogen atom of piperazine and hydrogen atom of guanidine moiety of Arg82. The molecule was stabilized by hydrophobic interaction in the hydrophobic pocket, considered to be important for bringing in specificity observed at the enzyme level as per Figure 4. In the docking studies, all the active compounds from the synthesized compounds, exhibited good docking score with significant polar and non-polar contacts, hydrogen bond interactions with the relevant amino acids and various hydrophobic interactions. Forty eight compounds were then screened for the GyrB inhibitory activity. With respect to structure-activity relationship study, on the left hand core, we attempted phenyl as in 13–18, OCH3 in 19–24, F in 25–30 and CF3 in 31–36 ring system in the sixth position of quinoline nucleus, while the right hand core was substituted with simple/4-F/4-Cl/4-NO2/4-methyl/4-methoxy phenyl ring. In the first subset, various substitutions were attempted on the R and R1 position of quinoline nucleus having oxygen as the X
2065
position (13–36). The substitutions OCH3 and F at R position exhibited good activity, while simple phenyl group showed moderate activity and CF3 was found to be detrimental to activity as compared to other substitutions. These derivatives 13–36 were then evaluated for the Gyr B inhibitory potency using the malachite green assay as described in the materials and methods section. Among the twenty four derivatives evaluated, four compounds showed IC50 less than 10 lM as shown in Table 1. Among the OCH3 substituted analogues at R position, the 4-NO2 of 22 and 4methoxy phenyl of 24 at R1 position showed good activity with IC50s of 6.991 ± 0.85 and 9.255 ± 0.42 lM, respectively. A closer look into the interaction profile of these molecules revealed that the NH of quinoline nucleus was involved in a prominent hydrogen bonding interaction with Asp79, analogous to the one observed in the crystal ligand. Compound 22 also showed additional polar contact with Asn52, due to which was found to display good binding with a docking score of 6.74 kcal/mol, similar to compound 24 that showed hydrogen bonds with Asp79 and Glu48. Both the compounds were further stabilized by various hydrophobic interactions with Val125, Val128, Val49, Ile171, Val77, Ala53, Val99, Met100 and Ile84 as shown in Supplementary information Figure S1. The F substituted compounds at R position as in 25–30, 4-fluoro and 4-nitro phenyl substituted compounds at R1 (26 and 28) turned out to be more promising with IC50s of 1.78 ± 0.23 and 6.826 ± 0.62 lM respectively. Both the compounds showed good binding with a docking score of 6.12 and 5.99 kcal/mol, respectively. The NH group of quinoline nucleus exhibited hydrogen bonding interaction with Asp79, whereas 4-fluoro and 4-nitro groups were exposed to solvent accessible area. Further the compounds were also found to be stabilized by hydrophobic interactions with Val125, Val128, Val49, Ile171, Val77, Ala53, Ile84, Val98, Val99, Met100 and Val123 as shown in Supplementary information Figure S2. However it was surprising to note that replacement of F substituent (26) with chloro (27) on the R1 position led to a drastic reduction in the activity as indicated in their interaction profile where the molecule was found to be in a slightly different orientation/pose compared to that of compound 26 thereby losing
Figure 4. Interaction profile picture of ligand 6-(3,4-dimethylphenyl)-3-[[4-[3-(4-methylpiperazin-yl)propoxy]phenyl]amino]pyrazine-2-carboxamide in the active site of Mycobacterium smegmatis GyrB protein.
2066
B. Medapi et al. / Bioorg. Med. Chem. 23 (2015) 2062–2078
Table 1 In vitro biological evaluation of the synthesized forty eight compounds
X N H
HN
R1
N R N 13-36 X=O 37-60 X=S Compd
R
R1
MS GyrB assay (IC50)
MTB supercoiling assay (IC50)
MTB MIC (lM)
Cytotoxicity at 50 lM (% Inhib.)
13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
H H H H H H OCH3 OCH3 OCH3 OCH3 OCH3 OCH3 F F F F F F CF3 CF3 CF3 CF3 CF3 CF3 H H H H H H OCH3 OCH3 OCH3 OCH3 OCH3 OCH3 F F F F F F CF3 CF3 CF3 CF3 CF3 CF3
Phenyl 4-Fluoro phenyl 4-Chloro phenyl 4-Nitro phenyl p-Tolyl 4-Methoxy phenyl Phenyl 4-Fluoro phenyl 4-Chloro phenyl 4-Nitro phenyl p-Tolyl 4-Methoxy phenyl Phenyl 4-Fluoro phenyl 4-Chloro phenyl 4-Nitro phenyl p-Tolyl 4-Methoxy phenyl Phenyl 4-Fluoro phenyl 4-Chloro phenyl 4-Nitro phenyl p-Tolyl 4-Methoxy phenyl Phenyl 4-Fluoro phenyl 4-Chloro phenyl 4-Nitro phenyl p-Tolyl 4-Methoxy phenyl Phenyl 4-Fluoro phenyl 4-Chloro phenyl 4-Nitro phenyl p-Tolyl 4-Methoxy phenyl Phenyl 4-Fluoro phenyl 4-Chloro phenyl 4-Nitro phenyl p-Tolyl 4-Methoxy phenyl Phenyl 4-Fluoro phenyl 4-Chloro phenyl 4-Nitro phenyl p-Tolyl 4-Methoxy phenyl Novobiocin Isoniazid Rifampicin Moxifloxacin
17.26 ± 0.87 34.88 ± 1.81 12.94 ± 1.13 16.95 ± 0.59 20.48 ± 1.52 22.02 ± 1.7 12.18 ± 0.94 18.48 ± 0.44 16.24 ± 1.13 6.99 ± 0.85 11.48 ± 0.32 9.25 ± 0.42 20.52 ± 1.26 1.78 ± 0.23 48.98 ± 2.66 6.83 ± 0.62 15.05 ± 0.55 10.12 ± 0.85 14.9 ± 0.53 19.77 ± 1.1 33.98 ± 2.3 47.63 ± 2.1 16.6 ± 1.3 41.2 ± 0.96 12.7 ± 2.1 29.81 ± 1.5 12.31 ± 0.92 3.77 ± 0.24 7.26 ± 0.33 19.42 ± 1.91 2.14 ± 0.15 8.7 ± 0.22 0.95 ± 0.12 2.5 ± 0.26 11.14 ± 0.39 10.1 ± 1.26 68.22 ± 2.8 45.92 ± 1.92 41.1 ± 0.98 56.15 ± 1.3 59.3 ± 0.65 39.51 ± 0.87 39.04 ± 1.3 48.5 ± 2.9 16.53 ± 0.91 41.8 ± 0.44 24.68 ± 1.8 32.56 ± 0.52 180 ± 3.9 nM nd nd >50
9.79 ± 0.23 19.74 ± 0.27 8.2 ± 0.24 10.63 ± 0.18 14.82 ± 0.32 18.54 ± 0.41 9.32 ± 0.22 14.8 ± 0.17 14.33 ± 0.21 5.33 ± 0.16 8.63 ± 0.31 5.86 ± 0.22 9.84 ± 0.29 0.78 ± 0.13 29.63 ± 0.65 2.14 ± 0.18 8.83 ± 0.44 7.62 ± 0.31 8.3 ± 0.22 11.8 ± 0.31 17.3 ± 0.33 22.3 ± 0.54 9.2 ± 0.2 19.8 ± 0.36 9.4 ± 0.2 16.72 ± 0.2 8.12 ± 0.13 1.21 ± 0.2 2.66 ± 0.32 13.51 ± 0.35 0.91 ± 0.05 2.31 ± 0.31 0.62 ± 0.16 0.88 ± 0.04 7.94 ± 0.3 6.17 ± 0.23 39.44 ± 0.43 23.54 ± 0.53 21.5 ± 0.55 25.5 ± 0.4 25.41 ± 0.2 17.84 ± 0.31 29.4 ± 0.43 22.84 ± 0.77 4.12 ± 0.14 7.54 ± 0.18 5.1 ± 0.22 34.5 ± 0.87 46 ± 10 nM nd nd 11.2 ± 0.36
25.84 34.09 7.91 15.41 24.14 32.02 32.02 25.66 17.35 14.70 15.45 12.57 33.03 3.82 50.78 9.52 15.97 7.65 29.35 24.89 48.53 52.2 28.05 54.26 16.6 36.14 15.21 7.65 15.68 30.22 3.81 14.72 3.47 1.72 7.28 15.84 63.26 63.26 31.68 63.26 67.94 31.68 35.82 66.53 30.68 51.07 31.68 66.88 >200 0.66 0.23 1.26
33.87 49.39 40.07 30.37 51.44 49.54 66.47 17.31 42.40 35.01 28.17 19.58 48.88 25.39 49.67 47.89 39.73 49.67 20.13 39.75 29.11 48.24 16.23 29.00 11.82 32.93 20.83 31.83 21.63 21.93 41.89 29.28 29.73 24.93 44.93 31.82 37.38 33.93 32.28 40.73 23.93 21.29 23.28 50.72 11.82 19.80 19.27 41.83 nd nd nd nd
ND indicates not determined.
the hydrogen bonding interaction with Asp79 as presented in Supplementary information Figure S3. In the next subset, CF3 substitution at R position as in 31–36 was found to exhibit moderate
activity against MS GyrB ATPase as well as MTB DNA supercoiling. The simple phenyl and 4-methyl phenyl substitution at R1 position as in 31 and 35 respectively showed moderate activity when
B. Medapi et al. / Bioorg. Med. Chem. 23 (2015) 2062–2078
2067
Figure 5. The interaction profile pictures of compounds 43 and 46 at the active pocket of GyrB of MS with respect to the substitutions at R1 position with phenyl and 4nitrophenyl with CH3O substituent at R position.
compared with other compounds from this series (IC50s of 14.9 ± 0.53 and 16.6 ± 1.3 lM in MS Gyr B, respectively, and 8.3 ± 0.22 and 9.2 ± 0.2 lM, respectively, in supercoiling assays). Other compounds from this set namely 33, 34 and 36, where the R1 position was substituted with 4-chloro/nitro and 4-methoxy phenyl were found to be less active with IC50 P30 lM. Docking studies revealed that these compounds oriented in a different manner, and a closer look at the interaction pattern of the least active analogues in this class showed that the introduction of 4-chlorophenyl as in compound 33 changed the orientation in the active site of protein and also CF3 group was found to be placed outside the pocket which could be the reason behind their lesser activity, which was also supported by a low docking score of 4.12 kcal/mol as presented in Supplementary information Figure S4. The other two compounds substituted with 4-nitro, and 4-methoxy phenyl at R1 were also not active due to the lack of hydrogen bonding interaction with Asp79 though hydrophobic interactions were present as per Supplementary information Figure S4. In the next set of compounds, 37–60, the effects of similar substituents were explored at R and R1 positions on left and right sides of quinoline nucleus having S (sulfur) group instead of O (oxygen) at the X position. Among the 24 synthesized compounds, 6 compounds inhibited Gyr B enzyme activity with IC50s less than 10 lM while compound 45, 1-(4-chlorophenyl)-3-[1-(6-methoxy-2-methylquinolin-4-yl) piperidin-4-yl]thiourea was found to be the most promising with an GyrB IC50 of 0.95 ± 0.12 lM which was well correlated with DNA supercoiling IC50 of 0.62 ± 0.16 lM. In this set of compounds simple phenyl and OCH3 at R position and various simple/4-F/4Cl/4-NO2/4-methyl/4-methoxy phenyl ring at R1 position exhibited good activity, while CF3 substituted compounds showed moderate activity and F substitution was found to be detrimental to activity. In the first subset, simple phenyl was explored at R position and simple 4-F/4-Cl/4-NO2/4-methyl/4-methoxy phenyl rings at R1 position similar to the previous set of compounds (37–42). Here, 4-nitro (40) and 4-methyl phenyl substituted compounds (41) were found to exhibit good activity among the other compounds with IC50s of 3.77 ± 0.24 and 7.26 ± 0.33 lM, respectively. In silico analysis of these compounds indicated that the molecules oriented in a similar manner as that of lead compound retaining hydrogen bonding with Asp79 and other significant polar interactions with
Glu48 and Asn52 amino acid residues. However their good GyrB potency could be accounted due to the hydrophobic interactions which were earlier demonstrated to be crucial for activity and specificity observed at the enzyme level as shown in Supplementary information Figure S5. Both the compounds were well inserted into the active site pocket which made these compounds better among all the compounds that showed good binding with the receptor (6.12 and 6.32 kcal/mol). On the other hand, introduction of 4-fluorophenyl group at R1 position (38) oriented the compound differently, thereby losing key hydrogen bonding interactions which probably resulted in their diminished activity as shown in Supplementary information Figure S6. The OCH3 substituted compounds (43–48) emerged as the most active compounds as they exhibited activity in the range of 0.95 ± 0.12 to 11.14 ± 0.39 lM in MS GyrB and 0.62 ± 0.16 to 7.94 ± 0.3 lM in MTB supercoiling assay. The binding analysis of these compounds within the MS protein in the ATP-binding site revealed that the most promising compounds from the series were 43, 45 and 46 as they showed good binding characteristics with the protein with docking score ranging from 6.87 to 6.33 kcal/kmol. Also, these compounds were well oriented at the active site cavity of the protein as shown in Figures 5 and 6. A closer analysis revealed that all the three active compounds interacted with Asp79 through hydrogen bonding which was important for retaining the bioactivity. Compound 43 was found to show an additional interaction with Glu48, while compound 45 was involved in hydrogen bonding Val99 in addition to Glu48. Furthermore, compound 45 was found to be stabilized by hydrophobic interactions with Ala53, Val128, Val50, Val49, Met100, Ile171 and Val99. Binding pattern of the most active compound 45 is represented in Figure 6. The next set of compounds, where R position was substituted with F or CF3 group (49– 60) were found to be less active when compared to simple phenyl or OCH3 substituted compounds. Binding analysis of the least active, F substituted compounds (49–54) revealed that all these compounds lacked hydrogen bonding interactions with the protein. This was clearly evident with their lesser docking scores of 4.12 to 3.99 kcal/mol. Similarly among the CF3 substituted compounds, only 4-methoxy and 4-chloro substitutions were found to exhibit moderate activity (IC50 of 12.56 ± 0.52 and 16.53 ± 0.91 lM) while other compounds did not possess bioactivity. Analysis of this set
2068
B. Medapi et al. / Bioorg. Med. Chem. 23 (2015) 2062–2078
Figure 6. Interaction profile pictures of the most active compound 45 at the active pocket of GyrB of MS.
Figure 7. Picture depicting the supercoiling assay of compound 45 at four different concentrations of 3.8, 1.9, 0.95 and 0.5 lM; R-Relaxed DNA (substrate + DMSO); CControl (Relaxed DNA substrate + DNA Gyrase + DMSO); N-Novobiocin.
of compounds revealed that the compounds 57 and 60 were well inserted into active site pocket of MS protein and the NH group of quinoline ring was involved in hydrogen bonding interaction with Asp79, while NH of 4-aminopiperidine ring interacted with Glu48 as shown in Supplementary information Figure S7. Binding analysis of least active compounds revealed that the compounds were oriented differently when compared to active compounds. 4-Flouro or nitro group at R1 position was displaced out of the pocket due to which the activity of molecules could have been affected, as shown in Supplementary information Figure S8. Compounds that inhibited ATPase activity of GyrB protein subsequently were expected to also inhibit DNA supercoiling, as Gyr A and GyrB domains constituted as holo enzyme DNA gyrase. Supercoiling assay was performed using the DNA supercoiling kit (Inspiralis Pvt. limited, Norwich) as per the protocol described in Section 4.4. Ideally, to eliminate the possibilities of aggregation and auto-fluorescence artifacts, non-specific inhibiting detergents were added in our biological experiments.22 All the reactions were performed against MTB DNA gyrase enzyme dose dependently at eight concentrations as 100, 50, 25, 12.5, 6.25, 3.125, 1.56 and 0.75 lM, in which the most active compound 45 showed an IC50 of 0.62 ± 0.16 lM as illustrated in Table 1. Novobiocin was employed as a standard compound in this assay for comparison.
Subsequently, IC50 were calculated based on relative quantification using Image lab software, Bio-Rad (Fig. 7). Further, the synthesized derivatives were screened for their in vitro anti-tubercular activity against Mycobacterium tuberculosis H37Rv strain (ATCC 27294) by agar dilution method as described in Section 4.23 The minimum inhibitory concentration (MIC) was determined for each compound and was found to be within a range of 0.78–50 lg/mL, as tested in seven different concentrations in triplicate. Overall, a good correlation was observed between the MTB DNA gyrase supercoiling IC50 values and the in vitro MTB MICs. However a slight deviation in the MS Gyr B IC50 was observed, probably owing to the slight difference in the proteins of the two organisms, as MIC was evaluated on MTB (M. tuberculosis H37Rv) and the Gyr B assay was performed on MS protein respectively.21 Ethambutol, ofloxacin and novobiocin were used as standard compounds. All the synthesized compounds showed good activity against MTB with MIC values ranging from 1.72 to 66.8 lM as shown in Table 1. Closer observation of MIC values confirmed that these compounds were better in activities than the first-line anti-tubercular drug ethambutol (MIC = 9.84 lM) but were less active when compared to isoniazid and moxifloxacin. The standard compound novobiocin was found to be inactive up to 200 lM concentration. Twenty compounds showed commendable MIC
