Nitrofuranyl Methyl Piperazines as New Anti-TB Agents: Identification, Validation, Medicinal Chemistry, and PK Studies

Abstract

Whole-cell screening of 20,000 drug-like small molecules led to the identification of nitrofuranyl methylpiperazines as potent anti-TB agents. In the present study, validation followed by medicinal chemistry has been used to explore the structure–activity relationship. Ten compounds demonstrated potent MIC in the range of 0.17–0.0072 μM against H₃₇Rv Mycobacterium tuberculosis (MTB) and were further investigated against nonreplicating and resistant (Rif^R and MDR) strains of MTB. These compounds were also tested for cytotoxicity. Among the 10 tested compounds, five showed submicromolar to nanomolar potency against nonreplicating and resistant (Rif^R and MDR) strains of MTB along with a good safety index. Based on their overall in vitro profiles, the solubility and pharmacokinetic properties of five potent compounds were studied, and two analogues, 14f and 16g, were found to have comparatively better solubility than others tested and acceptable pharmacokinetic properties. This study presents the rediscovery of a nitrofuranyl class of compounds with improved aqueous solubility and acceptable oral PK properties, opening a new direction for further development.

The emergence of resistant tuberculosis has been a serious concern worldwide that has reinvigorated drug discovery efforts in search of novel candidates that are effective against both susceptible and resistant strains as well as safe and potentially faster-acting, with the aim of shortening lengthy TB treatments.¹⁻³ Whole cell screening is an attractive approach to the fast identification of novel compounds active against TB.⁴ The success of the whole cell screening-driven approach to tuberculosis is best illustrated by the discovery of bedaquiline (TMC207, diarylquinoline derivative).^5,6 Similar success has been shown by the discovery of other preclinical candidates such as benzothiazoles TCA-1⁷ and imidazopyridine amides Q-203.⁸ Considering the high attrition rate in clinical trials, further enrichment of the clinical pipeline is greatly needed.

To discover novel and potent anti-TB agents, a whole-cell screening approach was adapted, and 20,000 small drug-like compounds were procured and screened.⁹ The aim of this approach was to find drug-like compounds in this collection and then to chemically modify these compounds to improve their PK/PD behaviors. The library was initially screened against sensitive (H₃₇Rv) and rifampicin resistant (Rif^R) strains of MTB at the concentration of 16 μg/mL, and 707 molecules demonstrated >90% growth inhibition. A minimum inhibitory concentration (MIC) determination for these compounds yielded 233 molecules with ≤8 μg/mL MIC against sensitive and resistant strains of MTB. The chemical clustering of these hits revealed nitrofuranyl methylpiperazine as one of the most potent scaffold (Figure 1). The identified nitrofuranyl methylpiperazine cluster included six compounds 1a–f with a MIC in the range of 0.2–25.3 μM against H₃₇Rv and Rif^R strains of MTB (Figure 1).

Figure 1.

Schematic representation of whole-cell screening of 20,000 molecules.

A literature survey revealed that nitrofuran containing compounds are known to possess anti-TB potential. Lee and co-workers¹⁰⁻¹⁶ have extensively studied nitrofuranylamides 2 and also generated several leads {3 (Lee-562), 4 (Lee-878), and 5 (Lee-1106)} as shown in Figure 2. Despite excellent in vitro profiles, these molecules possessed poor in vivo efficacy because of their low oral bioavailability and aqueous solubility, indicating that further effort is needed to utilize these compounds effectively. Our whole-cell screening program resulted in the identification of nitrofuranyl methyl piperazine 1, wherein a nitrofuranyl ring is directly attached to a piperizine ring through a methylene instead of an amide linkage. The high potency of these compounds, along with their comparatively simpler structure and likeliness of increased aqueous solubility indicated the need for further investigation. A medicinal chemistry program was initiated on nitrofuranyl methyl piperazine 1 using the strategy shown in Figure 3.

Figure 2.

Nitrofuranyl containing anti-TB molecules.

Figure 3.

Medicinal chemistry approach.

Results and Discussion

Chemistry

Initially, the potent hit 1a was prepared by the reductive amination of 5-nitrofuraldehyde 6 with 4-phenyl piperazine 7 (Scheme 1). Next, to ascertain the role of the nitro group, the des-nitro derivative 13 was prepared using the synthetic strategy shown in Scheme 2.¹⁷ The role of the furan ring was also investigated by synthesizing analogues 14a–c, in which the furan ring was replaced with a thiophene ring using the same synthetic strategy. The effect of ring C was also studied, and analogues 14d–l were synthesized with varying substituents. Through further modification, analogues 16a–h were prepared by replacing the phenyl ring (ring C) with alkyl/aryl sulfonyl groups as per the synthetic strategy shown in Scheme 3. In another modification, the piperazine ring (ring B) was replaced with piperidine and morpholine rings using the synthetic scheme shown in Scheme 4.

Scheme 1. Synthesis of Identified Hit 1a.

Reagents and conditions: (a) Na(OAc)₃BH, AcOH, DCM, rt, 12 h, 85%.

Scheme 2. Synthesis of Rings A and C Modified Nitrofuranyl Methyl Piperazines.

Reagents and conditions: (a) Pd(OAc)₂,Cs₂(CO)₃, rac-BINAP, toluene, reflux, 4 h, 60–65%;¹⁷ (b) TFA, DCM, rt, 2 h; (c) Na(OAc)₃BH, AcOH, DCM, rt, 12 h, 80–90%.

Scheme 3. Synthesis of Alkyl/Aryl Sulfonyl Groups Containing Nitrofuranyl Methyl Piperazines.

Reagents and conditions: (a) Na(OAc)₃BH, AcOH, DCM, rt, 12 h, 80%; (b) TFA, DCM, rt, 15 min; (c) R-SO₂CI, TEA, DMAP, DCM, rt, 12 h, 90–95%.

Scheme 4. Synthesis of Nitrofuranyl Analogues.

Reagents and conditions: (a) Na(OAc)₃BH, AcOH, DCM, rt, 12 h, 75–80%.

Biological Evaluation

The synthesized analogues 1a, 13, 14a–l, 15, 16a–g, and 18a–b were screened for in vitro activity against MTB H₃₇Rv (ATCC27294 strain) using the microbroth dilution method. The MIC was determined as the minimum concentration of the compound required to inhibit 90% of bacterial growth. The MIC values of all of the synthesized compounds are summarized in Table 1.

Table 1. In Vitro Activity of All Synthesized Compounds^a.

compd	MIC (H37Rv) (μM)	compd	MIC (H37Rv) (μM)
13	>16.5	15	0.048 ± 0
14a	>13.1	16a	23.07 ± 7.9
14b	>11.8	16b	0.17 ± 0
14c	>11.5	16c	0.16 ± 0
14d	13.2 ± 6.8	16d	0.31 ± 0
14e	0.12 ± 0.4	16e	0.15 ± 0
14f	0.0072 ± 0.03	16f	0.5 ± 0.1
14g	0.37 ± 0.18	16g	0.08 ± 0
14h	0.02 ± 0	16h	1.3 ± 0
14i	0.3 ± 0.1	18a	0.59 ± 0.2
14j	0.047 ± 0	18b	31.4 ± 10.8
14k	0.019 ± 0.006	Rifampicin	0.07 ± 0.03
14l	25.2 ± 7.6

^aValues reported are the average of three individual measurements ± SD.

Synthesized compound 1a exhibited an MIC value of 0.2 μM against the H₃₇Rv strain of MTB, similar to that observed during the library screen. The removal of nitro group destroyed the activity of the compound 13, which revealed that the presence of the nitro group is essential. The replacement of furan with a thiophene ring was also unfavorable, and none of the thiophene ring-containing analogues 14a–c demonstrated any inhibition at concentrations up to 10 μM. The results suggested that the nitro-furan moiety is essential for activity.

The effect of ring C and its substituents were also investigated. The nature and position of the substituents greatly influenced activity. The presence of substituents at para- and meta-positions on ring C was found to be favorable. Among the different para-groups, compounds with bulkier groups, 14f (4-tert-butyl) and 14h (4-morpholinyl), demonstrated comparatively better MIC values of 0.0072 and 0.02 μM against H₃₇Rv strain of MTB, respectively. Among the meta-groups, compounds with methoxy group 14j and cyano group 14k demonstrated the most potent MIC values of 0.047 and 0.019 μM, respectively (Table 1). The activity results suggested that both the nature and position of substituents influenced activity, and the presence of bulkier groups at the para-position and smaller groups at the meta-position led to enhanced potency.

In further modifications, the phenyl ring (ring C) was replaced by alkyl/aryl sulfonyl groups. Compound 16a, with a methane sulfonyl group on the piperazine ring, demonstrated decreased activity. However, nitrofuranyl methyl piperazines with un/substituted phenyl sulfonyl groups 16b–h showed good to excellent activity (Table 1). Moreover, due to the presence of a sulfonyl group between ring B and ring C, analogues with an even smaller group at the para-position of ring C exhibited good activity. Overall, these results suggest that replacement of the phenyl ring with an arylsulfonyl moiety is acceptable. The replacement of the piperazine ring with piperidine in compound 18a led to a MIC of 0.59 μM, but replacement with morpholine in compound 18b led to a complete loss of activity. These results indicate that the piperazine ring is preferred over piperidine or morpholine rings. A brief summary of SAR is shown in Figure 4.

Figure 4.

Brief SAR of nitrofuranyl methylpiperzine series.

Among all of the tested compounds, the 10 most active nitrofuranyl methyl piperazine analogues with activities of ≤0.2 μM (Table 1) were further screened against nonreplicating (streptomycin starved M. tuberculosis 18b) and resistant (Rif^R and MDR) strains of MTB. The cytotoxic potential of these compounds was also investigated in a HepG2 cell line. The results are shown in Table 2. Five analogues viz., 14f, 14h, 14k, 15 and 16g demonstrated potent MIC values against nonreplicating and resistant strains of MTB. None of the tested compounds were toxic in HepG2 cell lines, and all have acceptable safety indices.

Table 2. Activity against Nonreplicating and Resistant Strains of MTB and Cytotoxicity Studies^a.

compd	NRPb μM	MIC (RifR) μM	MIC (MDR) μM	CC50c μM	selectivity index (Sl)d
14e	12.6 ± 0	1.32 ± 0.45	>25.2	>63	>700
14f	1.4 ± 0	0.072 ± 0.02	0.029 ± 0.01	>58	>11600
14h	0.56 ± 0.2	0.08 ± 0	0.08 ± 0	>53	>2650
14j	2.11 ± 0.9	0.37 ± 0	0.09 ± 0	>63	>1575
14k	1.34 ± 0.4	0.19 ± 0	0.06 ± 0.02	>64	>3200
15	1.6 ± 0	0.32 ± 0.1	0.12 ± 0.05	>64	>1600
16b	>22.7	2.13 ± 0.8	>22.7	>56	>329.4
16c	>21.9	0.68 ± 0	14.59 ± 6.32	>54	>337.5
16e	20.1 ± 0	0.63 ± 0	4.19 ± 1.46	>50	>333.3
16g	>21.6	1.35 ± 0	1 ± 0.47	>54	>675
Rifampicin	2.4 ± 1.4	311.07 ± 0	155.5 ± 0
Gatifloxacin	2.66 ± 0	2.66 ± 1.5	1.33 ± 1.5

^aValues reported are the average of three individual measurements ± SD.

^bNonreplicating phase of M.tb.

^cCytotoxicity (concentration causing death of 50% of cells; CC₅₀) to HepG2 cells.

^dSelectivity index (CC₅₀(nM)/MIC(nM)).

Based on the measured activity levels against all of the tested strains, the five most active compounds, 14f, 14h, 14k, 15, and 16g, were further tested for solubility and pharmacokinetic activity. Previous studies have reported that nitrofuran based compounds suffer from poor absorption, which might be caused by their poor aqueous solubility. We studied the solubility of potent compounds in water and in other media (basic and acidic), and all of the tested compounds except 14h possess good to excellent aqueous solubility. The oral in vivo pharmacokinetic properties of these compounds were also studied in mice at a dose of 10 mg/kg, and the results are summarized in Table 3 (detail given in SI).

Table 3. Solubility and PK Profiles of Potent Compounds.

	solubility (μM)a				PKb
compd	water	PBS	SGF	SIF	Cmax (nM)	Tmax (h)	AUC0-t (nM·h)	t1/2(h)
14f	116.5	58.2	1165.5	2.9	2352.11	0.25	2994.63	4.04
14h	2.68				5342.18	0.25	4512.72	0.88
14k	256.3	256.3	2563.1	32	ndc	ndc	ndc	ndc
15	1285.5	2571.1	2571.1	2571.1	3806.90	0.25	2709.06	1.69
16g	108.3	54.1	2167.6	13.5	3263.51	0.25	4221.42	1.46

^aSolubility data were the average of three determinations (SD values < 1%).

^bOral dose at 10 mg/kg, and the data were average of five determinations.

^cnd: not detected.

Among the five tested compounds, three compounds 14f, 15, and 16g possess promising PK properties, with C_max values of 2352.11, 3806.90, and 3263.51 nM, respectively, and AUC_0-t values of 2994.63, 2709.06, and 4221.42 nM·h, respectively. The half-lives of 14f, 15, and 16g were calculated to be 4.04, 1.69, and 1.46 h, respectively. Compound 14h had a high C_max of 5342.18 and an AUC_0-t of 4512.72, but a comparatively short elimination half-life. Although compound 14k demonstrated good aqueous solubility, it was not detected in an in vivo oral PK experiment. This result might be due to the position and nature of the substituents present on ring C. The presence of substituents at the para-position provides stability and prevents the compound from being metabolized quickly.¹⁸ In 14k, a cyano group is present at the meta-position, while in the other four compounds (14f, 14h, 15, and 16g), substituents are present at the para-position. Thus, compound 14k may undergo fast metabolism, preventing its detection. The PK results obtained suggest that compounds 14f and 16g demonstrate good in vivo exposure and half-lives.

In conclusion, we have rediscovered nitrofuranyl methyl piperazine as a potent scaffold for compounds effective against sensitive and resistant strains of MTB. The present study demonstrates the promise of the nitrofuranyl-based class of compounds against sensitive, resistant, and nonreplicating strains of MTB. The reported compounds have optimal PK properties and comparatively better aqueous solubility than other reported analogues in this class. Studies to determine the in vivo efficacy of these compounds are currently underway.

Supporting Information Available

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsmedchemlett.5b00141.

• Full experimental details for the compounds synthesized, along with NMR and MS spectra and descriptions of biological assays (PDF)

Author Contributions

^# K.R.Y., G.M., and S.S. have equally contributed to this work. K.R.Y. and G.M. performed the chemical syntheses. S.S., S.K., and V.S.R. performed biological screening. A.M., M.T., and G.D.S. performed in vivo PK. S.S.B. performed the solubility study. P.P.S., I.A.K., and R.A.V. participated in the design and execution of this study.

This work was supported by the Council of Scientific and Industrial Research (CSIR)-New Delhi with research grant # HCP 0001. This work is assigned the IIIM Publication No: IIIM/1761/2015.

The authors declare no competing financial interest.