6-Hydrogen-8-Methylquinolones Active Against Replicating and Non-replicating Mycobacterium tuberculosis


  • Oriana Tabarrini,

    Corresponding author
    1. Dipartimento di Chimica e Tecnologia del Farmaco, Università di Perugia, Via del Liceo 1, 06123 Perugia, Italy
    Search for more papers by this author
  • Stefano Sabatini,

    1. Dipartimento di Chimica e Tecnologia del Farmaco, Università di Perugia, Via del Liceo 1, 06123 Perugia, Italy
    Search for more papers by this author
  • Serena Massari,

    1. Dipartimento di Chimica e Tecnologia del Farmaco, Università di Perugia, Via del Liceo 1, 06123 Perugia, Italy
    Search for more papers by this author
  • Marco Pieroni,

    1. Dipartimento di Chimica e Tecnologia del Farmaco, Università di Perugia, Via del Liceo 1, 06123 Perugia, Italy
    Search for more papers by this author
    • Present address: Dipartimento Farmaceutico, Via GP Usberti 27/A-Campus Universitario Università di Parma, Parma, Italy.

  • Scott G. Franzblau,

    1. Institute for Tuberculosis Research, College of Pharmacy, University of Illinois at Chicago, 833 S. Wood St., Chicago, IL 60612, USA
    Search for more papers by this author
  • Violetta Cecchetti

    1. Dipartimento di Chimica e Tecnologia del Farmaco, Università di Perugia, Via del Liceo 1, 06123 Perugia, Italy
    Search for more papers by this author

Corresponding author: Oriana Tabarrini,oriana.tabarrini@unipg.it


The screening of an in-house quinolones library against Mycobacterium tuberculosis (Mtb) H37Rv, followed by a first cycle of optimization, yielded 6-hydrogen-8-methyl derivatives endowed with good potency. The antitubercular activity also encompassed the bacteria in a non-replicating state (NRP-TB) with minimum inhibitory concentration values lower than those of the reference agent, moxifloxacin. Among the best compounds, 11w and 11ai, characterized by a properly substituted piperidine at the C-7 position, were active against single-drug-resistant (SDR-TB) Mtb strains, maintaining overall good potency also against ciprofloxacin-resistant Mtb. This study expands the body of SAR around antitubercular quinolones leading to reconsider the role played by the usual fluorine atom at the C-6 position. Further elaboration of the 6-hydrogen-8-methylquinolone scaffold, with a particular focus on the C-7 position, is expected to give even more potent congeners holding promise for shortening the current anti-TB regimen.






Directly Observed Therapy Short-course








low oxygen recovery assay


microplate Alamar Blue assay


minimum bactericidal concentration


multidrug-resistant Mtb


minimum inhibitory concentration


Mycobacterium tuberculosis


new chemical entities


non-replicating persistent Mtb






structure–activity relationships


single-drug-resistant Mtb




extensively drug-resistant Mtb

Tuberculosis (TB) is an old and re-emerging disease that spreads at alarming rate. Mycobacterium tuberculosis (Mtb) is responsible for 8.8 million new active cases and 1.5 million deaths in 2010, and one-third of the world’s population is estimated to be latently infected with TB. About 95% of the tuberculosis cases and deaths occur in the developing countries, but there has also been a recent, highly publicized, resurgence of TB also in developed ones.a This pandemic is complicated by the co-infection with HIV (15% of new cases worldwide are HIV positive) and the emergence of multidrug-resistant (MDR-TB) and extensively drug-resistant (XDR-TB) Mtb strains.b The current recommended therapeutic strategy, namely Directly Observed Therapy, Short-course (DOTS) is based on the co-administration of four drugs, namely rifampin (RMP), isoniazid (INH), pyrazinamide (PZA), and ethambutol (EMB) for 2 months, followed by a prolonged treatment with INH and RMP for an additional 4–7 months.c The cause of the long-period treatment is the peculiar ability of Mtb to survive in a non-replicating persistent (NRP-TB) state while withstanding chemotherapy. These features highlight the need to develop well-tolerated drugs able to shorten the anti-TB treatment by targeting bacteria in a NRP state and effective also in the treatment of MDR-TB and XDR-TB strains.

After 40 years of neglect, because of a combination of medical and marketing evaluations, novel therapeutics are fuelling the anti-TB pipeline, with ten compounds in clinical trials and about the same number in preclinical development (1,2). Some of these molecules are New Chemical Entities (NCEs), while others are drugs already marketed for purposes other than the antimycobacterial chemotherapy. Belonging to the latter class, the 8-methoxyfluoroquinolones, moxifloxacin (3) and gatifloxacin (4), have earned a place in the antitubercular chemotherapy as second-line drugs, in the case of intolerance or resistance to the first-line drugs (5,6).d In addition, because of the high oral bioavailability, generally good tolerability, and low-to-moderate cost, they are currently in phase III of clinical trials for an evaluation of their potential to shorten TB treatment when in combination with other anti-TB agents (7). If successful, these drugs could shift from second-line antituberculars to first-line (8–10). However, mainly because of the recurrent use of quinolones for infections in patients where TB initially was not suspected, resistance to these two agents is a matter of concern (11,12). Cross-resistance is likely to occur because all the fluoroquinolones (FQs) target gyrA (13,14). However, an analysis of the gyrA mutations has shown that about a half of the isolated strains resistant to ofloxacin are susceptible to moxifloxacin (15,16) and to high doses of levofloxacin (16). So it is not unlikely that novel quinolones might be active also against FQ-resistant TB strains.

Challenging the C-6 fluorine dogma, our group has synthesized a relevant number of 6-amino- and 6-hydrogenquinolone derivatives, collectively referred as 6-desfluoroquinolones (6-DFQs) (17–22), with potent broad antibacterial spectra, including multidrug-resistant Gram-positive pathogens, such as methicillin-resistant and ciprofloxacin-resistant Staphylococcus aureus and penicillin-resistant Streptococcus pneumoniae.

The encouraging evidences obtained with regard to the antibacterial activity and the availability of a large series of variously substituted 6-DFQs prompted us to evaluate their capability to inhibit the growth of Mtb, with the aim of contributing to the enrichment of the SAR regarding their antitubercular activity.

In particular, in this study, we report the inhibitory activity against Mtb strain H37Rv of about 60 selected 6-DFQs (114, Figure 1). The results obtained from the primary screening guided the design of new analogues (11ah, 11ai, and 11aj) for which the synthesis and antimycobacterial activity are also herein reported. To provide a comprehensive profile of their antimycobacterial properties, the activity against single-drug-resistant strains (SDR-TB), the minimum bactericidal concentration (MBC), and the apparent cytotoxicity to Vero cells have been also evaluated for selected molecules. In addition, a low oxygen recovery assay (LORA) was used to test their activity against NRP-TB.

Figure 1.

 Chemical structure of 6-DFQs tested for their activity against Mtb H37Rv.

Material and Methods


All reactions were routinely checked by thin-layer chromatography (TLC) on silica gel 60F254 (Merck) and visualized by using UV or iodine. Flash column chromatography separations were carried out on Merck silica gel 60 (mesh 230–400). Melting points were determined in capillary tubes using an Electrothermal apparatus 9100 (Electrothermal, Rochford, UK) and are uncorrected. Elemental analyses were performed in the elemental analyzer EA 1108 CHN (Fisons Instruments, Beverly, MA, USA), and the data for C, H, and N are within ±0.4% of the theoretical values. 1H NMR and 13C NMR spectra were recorded at 200 MHz (Bruker Avance DPX-200) and 400 MHz (Bruker Avance DRX-400) using residual solvents such as chloroform (δ = 7.26) or dimethylsulfoxide (δ = 2.48) as an internal standard. Chemical shifts are given in ppm (δ), and the spectral data are consistent with the assigned structures. Reagents and solvents were purchased from common commercial suppliers and were used as such. After extraction, organic solutions were dried over anhydrous Na2SO4, filtered, and concentrated with a Büchi rotary evaporator at reduced pressure. Yields are of purified product and were not optimized. Microwave-irradiated reactions were carried out in a Biotage Initiator 2.0 apparatus (Biotage AB, Uppsala, Sweden) using 5-, 10-, and 20-mL pyrex vials. All starting materials were commercially available unless otherwise indicated.

General procedure for the preparation of 6-hydrogen-8-methylquinolones (11ah, 11ai, and 11aj)

The mixture of 1-cyclopropyl-7-fluoro-8-methyl-4-oxo-1,4-dihydroquinoline-3-carboxylic acid BF2 chelate (20) (1.0 equiv), piperidine-4-one, piperidine-4-one oxime (23) or pyrrolidine-3-ol (1.5 equiv), and Et3N (1.5 equiv) in dry DMSO was heated at 120 °C for 24 h. The reaction mixture was then poured into ice water and extracted with CHCl3. The combined organic layers were purified by column chromatography eluting with chloroform followed by trituration with ethyl acetate.

1-Cyclopropyl-8-methyl-4-oxo-7-(4-oxopiperidine-1-yl)-1,4-dihydro-3-quinoline carboxylic acid (11ah): 40% yield, mp 218–220 °C (dec), 1H NMR (dimethylsulfoxide-d6)δ 0.80–1.00 and 1.20–1.40 (m, each 2H, cyclopropyl CH2), 2.65–2.80 and 3.45–3.60 (t, = 6.5 Hz, each 4H, piperidine CH2), 2.85 (s, 3H, CH3), 4.10–4.25 (m, 1H, cyclopropyl), 7.30 (d, = 9 Hz, 1H, H-6), 8.30 (d, = 9 Hz, 1H, H-5), 9.00 (s, 1H, H-2). 13C NMR (DMSOd6) δ 10.9, 18.4, 41.3, 41.7, 48.9, 107.7, 118.4, 120.3, 121.4, 124.7, 144.1, 152.5, 156.2, 166.3, 177.8, 208.7. Analytical calculated for C19H20N2O4: C, 67.05; H, 5.92; N, 8.23. Found C, 67.36; H, 6.00; N, 8.50.

1-Cyclopropyl-7-[4-(hydroxyimino)-1-piperidinyl]-8-methyl-4-oxo-1,4-dihydro-3-quinoline carboxylic acid (11ai): 38% yield, mp 277–279 °C (dec), 1H NMR (DMSOd6/acetone-d6) δ 0.80–0.95 and 1.20–1.30 (m, each 2H, cyclopropyl CH2), 2.40–2.50 (m, 2H, piperidine CH2), 2.65–2.80 (m, 7H, CH3 and piperidine CH2), 3.10–3.30 (m, 4H, piperidine CH2), 4.25–4.40 (m, 1H, cyclopropyl), 7.25 (d, = 9 Hz, 1H, H-6), 8.10 (d, = 9 Hz, 1H, H-5), 8.80 (s, 1H, H-2), 10.35 (bs, 1H, OH), 15.00 (bs, 1H, COOH). 13C NMR (DMSOd6) δ 10.9, 18.6, 25.1, 31.8, 41.4, 50.9, 52.4, 107.5, 118.0, 120.2, 121.5, 124.6, 144.2, 152.5, 154.1, 158.3, 166.3, 177.9. Analytical calculated for C19H21N3O4: C, 64.21; H, 5.96; N, 11.82. Found C, 64.03; H, 6.10; N, 12.15.

1-Cyclopropyl-7-(3-hydroxy-1-pyrrolidinyl)-8-methyl-4-oxo-1,4-dihydro-3-quinoline carboxylic acid (11aj): 45% yield, mp 252–253 °C (dec), 1H NMR (DMSOd6) δ 0.60–0.80 and 1.05–1.25 (m, each 2H, cyclopropyl CH2), 1.90–2.10 (m, 2H, pyrrolidine CH2), 2.45 (s, 3H, CH3), 3.00–3.25 and 3.50–3.70 (m, each 2H, pyrrolidine CH2), 4.20–4.35 (m, 1H, pyrrolidine CH), 4.40–4.50 (m, 1H, cyclopropyl CH), 5.00 (bs, 1H, OH), 7.10 (d, = 9 Hz, 1H, H-6), 7.95 (d, = 9 Hz, 1H, H-5), 8.75 (s, 1H, H-2), 15.00 (bs, 1H, COOH). 13C NMR (DMSOd6) δ 10.8, 18.8, 41.3, 37.9, 50.5, 57.7, 70.6, 107.6, 118.4, 120.2, 121.5, 124.5, 144.1, 152.5, 158,4, 166.3, 177.9. Analytical calculated for C18H20N2O4: C, 65.84; H, 6.14; N, 8.53. Found C, 65.76; H, 5.98; N, 8.41.

Biologic evaluation

The MICs were determined using the BACTEC 460 (24) or MABA (24) and LORA (25) assays according to published procedures. Similarly, cytotoxicities were determined on Vero cells according to a published procedure (26). The MBC was determined for Mtb H37Rv by subculturing onto a drug-free solid medium and counting of CFU following exposure in Middlebrook 7H9 medium supplemented with drug concentrations equivalent to and higher than the previously determined MICs against the same strains. The MBC was defined as the lowest concentration resulting in a reduction in colony-forming units of 99% relative to the pretreatment level.

Results and Discussion

Synthesis of 6-hydrogen-8-methylquinolones 11ah, 11ai, and 11aj

Compounds 11ah, 11ai, and 11aj were synthesized as shown in Scheme 1, accordingly with a highly efficient procedure already reported by us for deactivated 6-hydrogen-8-methyl analogues (20), reacting 1-cyclopropyl-7-fluoro-8-methyl-4-oxo-1,4-dihydroquinoline-3-carboxylic acid BF2 chelate (20) with piperidine-4-one, piperidine-4-one oxime (23), and pyrrolidine-3-ol in DMSO at 120 °C overnight using dry triethylamine as the base.

                Scheme 1:

 Reagents and conditions: (i) Et3N, DMSO, 120 °C, 24 h.

Mycobacterium tuberculosis inhibitory activity

The 6-DFQs tested in this study (Figure 1) can be structurally grouped into two main classes: 6-aminoquinolones 110 (17–19,21), variously functionalized at N-1, C-7, and C-8 position, and 6-hydrogenquinolones 11 (20), all bearing a cyclopropyl group at the N-1 position coupled with a C-8 methyl group and different heterocycles at C-7 position. A few examples of 6-nitro (12c, 12t, 12u) (19), 6-methylamino (13h) (19), and 6-hydroxy (14q) (22) derivatives were also tested. Moreover, the library also included two anti-HIV 6-aminoquinolones, 1o (27) and 4s (28), selected within our in-house library of antiviral quinolones (29).

All the compounds were preliminary tested against Mtb H37Rv, at the single concentration of 12.5 μg/mL using the BACTEC 460 radiometric system. In this primary screening, an inhibition value higher than 90% was shown by 24 derivatives for which the successive antimycobacterial evaluation was performed in a microplate Alamar Blue assay (MABA) (24), so as to allow a rank in potency to be delineated (Table 1). Some of the compounds resulted inactive including the anti-HIV quinolones, 1o and 4s, characterized by a 4-arylpiperazine as C-7 substituent. More than half of the compounds showed only a modest antimycobacterial activity with minimum inhibitory concentration (MIC) values ranging from 12.5 to 6.25 μg/mL; two 6-amino derivatives, 1e and 5u, had good activity (MIC = 3.13 μg/mL), while the 6-hydrogenquinolones, 11v, 11y, 11b, 11h, 11ag, 11af, and 11w, showed the best activity with MIC values ranging from 1.6 to 0.1 μg/mL. The 7-(4-hydroxy)-1-piperidinyl derivative 11w was found to show the best activity with a MIC value of 0.1 μg/mL, the same as that of moxifloxacin. The better activity of 6-hydrogenquinolones compared with the 6-amino counterparts (11h, 11q, 11v, 11w, and 11yversus5h, 5q, 5v, 5w, and 5y) may be due to their higher lipophilicity (20), a characteristic likely to enhance the penetration of the molecule through the thick and waxy Mtb cell wall. However, the nature of the C-7 substituent also significantly influences the antitubercular activity with the following rank: 4-hydroxypiperidine (w) > 3-aminopyrrolidine (af, ag) > 4-methylpiperidine (h) = piperazine (b) = 3,5-dimethylpiperidine (y) > piperidine (v) > tetrahydroisoquinoline (q). These data lead to speculate that the presence of moieties such as the hydroxyl and the amino group, which have a hydrogen bond donor/acceptor ability, may play a key role in the interaction with the active site of the target.

Table 1.   Antimycobacterial activity of 6-DFQs against Mtb H37Rv
CompoundsMIC (μg/mL)
  1. aDetermined as described in Reference 24.

  2. bDetermined as described in Reference 25.

11w 0.11.95
11af 0.39 
11ag 0.3919.40
11b, 11h0.78 
11y 0.7825.63
11v 1.6 
1e, 5u3.13 
5t, 5q, 11q, 12t, 12u6.25 
1i, 2c, 2i, 5d, 5v, 5w, 5y, 11ae, 8c, 5h12.5 
11ah 0.2 
11ai 0.12.95
11aj 0.3913.69

Based on these structural considerations, many 6-hydrogen-8-methyl derivatives, variously functionalized at the C-7 positions, can be planned. Using 11w as a template, we initially synthesized three close analogues: 11ah, 11ai, and 11aj. In particular, the 4-hydroxypiperidine was changed to the piperidone analogue 11ah, which in turn was derivatized to its corresponding oxime 11ai, while it was contracted to a 3-hydroxy-1-pyrrolidine ring in compound 11aj.

Interestingly, the novel compounds, 11ah, 11ai, and 11aj, showed a 100% inhibitory activity at 12.5 μg/mL and at MICs of 0.2, 0.1, and 0.39 μg/mL, respectively, on the same order of those of the hit 11w and the clinical candidate moxifloxacin (Table 1). These encouraging results prompted us to carry out further antimicrobial evaluations for some of the best 6-hydrogenquinolones, namely 11y, 11w, 11ag, 11ai, and 11aj. In particular, we tested these derivatives in a LORA (25), which is an in vitro luminescence-based high-throughput assay suggested for assessment of activity against NRP-TB in oxygen-deprived conditions (Table 1). To some extent, this assay can predict whether or not the compound tested has potential for killing Mtb in a non-replicating state, that is, a condition required to shorten the TB regimen. We were pleased to notice that, while in the MABA, the activities of 11w and 11ai are similar to that of moxifloxacin, in the LORA, the compounds tested resulted, respectively, approximately fourfold (MIC 1.95 μg/mL) and approximately threefold (2.95 μg/mL) more active than moxifloxacin against NRP-TB.

Compounds 11y, 11w, 11ag, 11ai, and 11aj were also tested against a panel of SDR-TB, and as expected, all compounds showed the same activity against strains resistant to INH, RMP, KAN, and EMB as on drug-sensitive H37Rv strain (data not shown). More importantly, it must be noticed that compounds 11w and 11ai maintained a good activity (MIC 3.25 and 1.6 μg/mL, respectively) also against the ciprofloxacin-resistant strain, whereas MICs ranging from 4 to >8 μg/mL have been reported for moxifloxacin (30), as well as for many other antitubercular FQs reported to date (13–15,31) (Table 2). Compound 11ai was also tested against a moxifloxacin-resistant strain, showing an encouraging MIC of 2.94 μg/mL. As expected, the evaluation of MBCs against Mtb H37Rv (Table 2) showed that the 6-hydrogenquinolones, like clinical FQs, are bactericidal and kill Mtb at concentrations close to those that cause inhibition. Finally, all the compounds resulted to be not toxic against Vero cells (26) at the concentration tested of 102 μm.

Table 2.   Antimycobacterial activity against ciprofloxacin-resistant (CipR) Mtb strain and MBC against Mtb H37Rv of selected 6-DFQs
CompoundsMIC (μg/mL)aMBC (μg/mL)b
  1. aDetermined as described in reference 24.

  2. bSee materials and methods.

11y 0.78>251.56
11w 0.23.25>3.25
11ag 0.78>12.512.5
11ai ≤0.051.600.1
11aj 1.56>12.51.56

Conclusions and Future Directions

In summary, FQs play an important role in the treatment of TB infection with early agents being used as a second-line therapeutic option for the treatment of MDR-TB, while the newer compounds are in clinical trials for their potential to shorten the treatment regimens. Two quinolones, TBK-613 and DC-159a, are currently in preclinical development (1,2), and other promising analogues continue to enrich the literature (32,33).

In this study, the screening of an in-house quinolones library against Mtb H37Rv, followed by a first cycle of optimization, led to disclose 6-hydrogen-8-methylquinolones as very promising compounds together with additional structural information that definitely improved the SAR of the antimycobacterial quinolones: (i) potent and selective agents can be obtained that even lacks the usual fluorine atom at the C-6 position; however, it is important to emphasize that only the right substitution pattern around the 6-hydrogenquinolone nucleus can give a good antimycobacterial agent; indeed, only a moderate activity has been recently reported for a non-fluorinated quinolone, nemonoxacin (34); (ii) as a C-6 substituent, the hydrogen atom is better than the amino group, probably for the higher degree of lipophilicity, which warrants a better cell penetration; (iii) for the C-7 position, the basic nature of the substituent seems not required; the 4-arylpiperazine is not tolerated, while various piperidines or pyrrolidines resulted in suitable substituents; (iv) for the C-8 position, the methyl group is a good replacer of the standard methoxyl.

In conclusion, the preliminary findings stemmed from this work encourage further investigation around the 6-hydrogen-8-methylquinolone class of antitubercular compounds with the aim to obtain even more potent congeners holding promise of shortening the current anti-TB regimen.



The authors thank R. C. Reynolds of the TAACF, coordinated by the Southern Research Institute (Birmingham, Alabama) under the direction of the U. S. National Institute of Allergy and Infectious Diseases for arranging the following tests: activity of the whole library of compounds against Mtb H37Rv, activity of selected compounds against resistant strains, and their MBC. The authors are also very grateful to A. Fravolini for his pioneering efforts in the quinolone field and to Mr R. Bianconi (Department of Chemistry and Technology of Drug, Perugia, Italy) for his excellent technical assistance.