N1-Benzofused Modification of Fluoroquinolones Reduces Activity Against Gram-Negative Bacteria

Abstract

The fluoroquinolone class of antibiotics has a well-established structure–activity relationship (SAR) and a long history in the clinic, but the effect of electron-rich benzofused substituents at the N1 position remains poorly explored. Because groups at this position are part of the topoisomerase–DNA binding complex and form a hydrophobic interaction with the major groove of DNA, it was hypothesized that an electron-rich benzofused N1 substituent could enhance this interaction. Molecular modeling techniques were employed to evaluate the binding of certain N1-modified fluoroquinolones to DNA gyrase targets from both Staphylococcus aureus and Klebsiella pneumoniae species compared with ciprofloxacin and norfloxacin. Seven N1-modified fluoroquinolones were subsequently synthesized and tested against a panel of Gram-negative pathogens to determine minimum inhibitory concentration (MIC) values. Gram-negative outer membrane penetration was investigated using the membrane permeabilizer polymyxin B nonapeptide and compound efflux via resistance–nodulation–division-family efflux transporters was evaluated using the known efflux pump inhibitor phenylalanine–arginine β-naphthylamide. Additionally, the target inhibitory activity of representative compound 6e was determined in a cell-free environment. A correlation between N1 substituent hydrophobicity and activity was observed across the MIC panel, with compound activity decreasing with increased hydrophobicity. Those compounds with highest hydrophobicity were inactive because of poor solubility profiles whereas compounds with intermediate hydrophobicity were inactive because of impaired outer membrane penetration, and reduced inhibition of topoisomerase targets, the latter in contrast to modeling predictions. This study adds new information to the fluoroquinolone SAR and suggests limited utility of large hydrophobic substituents at the N1 position of fluoroquinolones.

Introduction

Drug design is made far more challenging when structural information about the biological target in question is lacking or the mechanism of action of a hit compound is unknown. While recent advances in the field of protein crystallography have allowed the structures of many challenging targets to be solved,¹ molecular modeling has traditionally been employed in such cases to allow the rational design and improvement of promising drug scaffolds.² Such modeling originally took the form of ligand-based quantitative structure–activity relationship (QSAR) calculations, which were used to more effectively rank analogue compounds and optimize scaffolds without offering any specific target information, and now includes target-based molecular docking for which there are various commercially available software suites. Through this, molecular modeling has become a companion technique to traditional compound screening efforts, allowing cheaper and quicker evaluation of chemical libraries against specified targets.² The caveat, of course, is that such data are only as useful as the model is accurate. While full quantum mechanical simulations would likely provide the highest degree of accuracy, the extreme computational burden involved in realizing such a model on the protein scale continues to make this approach untenable. Instead, modeling suites employ coarse grain molecular dynamics simulations for the majority of the biological target, with the more resource-intensive quantum mechanical simulations reserved for small, specific areas with which the drug is likely to directly interact.³ This compromise inevitably limits the utility of the data generated, the result being that drug discovery efforts are still primarily driven by in vitro and in vivo assays. Despite this, we opine that the refinement of older, successful drug scaffolds developed in the premodeling era represents a task to which current modeling suites are well suited.

One such scaffold is that of the fluoroquinolones. The fluoroquinolones are a family of synthetic, broad-spectrum antibiotic compounds and are important antimicrobial tools within the modern medical arsenal. They remain one of the most frequently prescribed families of antibiotics globally⁴ with a combined 17% share of the $43.9bn global antibiotics market in 2014, second only to the β-lactams.⁵ Since the discovery of nalidixic acid by Lesher and co-workers in 1962,⁶ several generations of fluoroquinolone antibiotics have reached the market including ciprofloxacin, levofloxacin, and moxifloxacin (Figure 1).^7,8 However, the class is not without its controversy. The FDA added black box warnings to fluoroquinolone-class drugs in 2008 in response to evidence of increased risk of tendon damage⁹ and a number of FDA-approved fluoroquinolones including temafloxacin, grepafloxacin, and trovafloxacin have either been withdrawn from the clinic or suffered heavily restricted use because of severe adverse effects.¹⁰ Despite further warnings by the FDA in 2013 and then 2016,¹¹ the increasing threat posed by antimicrobial resistance worldwide has stimulated renewed interest in the fluoroquinolones. In June 2017, the FDA approved Melinta Therapeutics’ delafloxacin;¹² other quinolones currently in clinical trials include lascufloxacin (Kyorin Pharmaceutical Co. Ltd.), finafloxacin (MerLion Pharmaceuticals Pte Ltd.), nemonoxacin (TaiGen Biotechnology Co. Ltd.), OPS-2071 (Otsuka Pharmaceutical Co. Ltd.), and levonadifloxacin (Wockhardt Ltd.).¹³

Figure 1.

Different generations of quinolone family antibiotics. The second generation of quinolones was the first to feature fluorine atoms at the C6 position, hence the adoption of the fluoro-prefix for subsequent compounds.

Fluoroquinolones inhibit the actions of DNA gyrase and topoisomerase IV, enzymes that facilitate local introduction or relaxation of DNA supercoils in order to avoid the occurrence of double stranded DNA breaks in the bacterial genome. If the frequency of breaks is high enough to overwhelm cellular DNA repair pathways, cell death occurs.¹⁴⁻¹⁸ To this end, the SAR of the quinolone core has been studied in detail to determine substituents that allow for a broad-spectrum activity profile. Groups at the N1 position are a part of the topoisomerase–DNA binding complex and form a hydrophobic interaction with the major groove of DNA.¹⁹ A variety of different substituents have found success here, including the cyclopropyl ring present in ciprofloxacin and moxifloxacin (traditionally regarded as the optimal substituent here for potency²⁰) and the 3,5-difluoropyridin-2-amine group in the newer delafloxacin, but the effect of large hydrophobic N1 substituents has yet to be fully explored.

We sought to use a molecular modeling-driven approach to investigate the merit of benzofused substituents at the N1 position of the fluoroquinolone nucleus. It was hypothesized that such substituents could present a novel means of enhancing the fluoroquinolone–DNA interaction to improve upon the levels of activity observed in existing fluoroquinolones. To investigate this hypothesis, representative N1-benzofused fluoroquinolones were first docked in silico against the DNA gyrase enzymes from Staphylococcus aureus and Klebsiella pneumoniae, then synthesized using solution phase chemistry and evaluated in vitro against panels of clinically relevant Gram-negative pathogens.

Results and Discussion

In Silico Modeling

Four novel N1-benzofused fluoroquinolones (6d–g) were modeled against the S. aureus and K. pneumoniae DNA gyrase enzymes. Ciprofloxacin (6a), norfloxacin (6b), and an N1-benzyl analogue (6c) were included as control compounds. The ten most favorable binding poses for each of the seven compounds were selected and compared according to both ChemScore and energy of binding. N1-benzofused compounds were all predicted to possess enhanced binding energy for both gyrase enzymes (best poses ≥ −25.37 kcal/mol vs subunit I) compared to norfloxacin (best poses ≤ −21.54 kcal/mol) and ciprofloxacin (best poses ≤ −23.59 kcal/mol), especially the benzothiophene-containing compounds 6e (best poses ≥ −27.76 kcal/mol) and 6g (best poses ≥ −27.20 kcal/mol) (Table S1). Additionally, visual examination of the binding poses predicted for each compound indicated that the N1-benzofused compounds would bind K. pneumoniae DNA gyrase in a similar manner to ciprofloxacin (Figure 2), though this did not appear to be the case with S. aureus (Figure S1). To test these predictions, all seven compounds were synthesized for microbiological evaluation (Scheme 1).

Figure 2.

(A,B) Molecular modeling results indicate that N1-benzofused compounds (represented by compounds 6e and 6g) bind to K. pneumoniae DNA gyrase in a similar manner to compound 6a (ciprofloxacin). Binding poses shown in 2D (A) and 3D (B) views.

Scheme 1. Synthetic Scheme Used for the Preparation of Fluoroquinolones including Novel N1-Benzofused Analogues.

(i) (EtO)₃CH, Ac₂O, 140 °C, 3 h. (ii) R–NH₂, DCM, rt, 15 h. (iii) DBU, LiCl, DCM, 45 °C—rt, 17.5 h. (iv) HCl, AcOH, reflux, 2.5 h. (v) Boc-piperazine, K₂CO₃, DMF, reflux, 15 h. (vi) TFA, dry DCM, rt, 2 h. (vii) 4 M HCl in dioxane, DCM, rt, 1 h. The seven fluoroquinolone compounds synthesized as a part of this study were three control compounds (ciprofloxacin, norfloxacin, and an N-benzyl analogue; 6a–c) and four novel N1-benzofused analogues (6d–g).

Results

Chemistry

Compounds 6a–g were synthesized using an established solution phase route (Scheme 1) and characterized using liquid chromatography–mass spectrometry (LC–MS), HRMS, infrared (IR), and ¹H and ¹³C NMR techniques. Hydrochloride salt forms were used to improve aqueous solubility for microbiological testing. clog P values were calculated for each compound (Table 1).

Table 1. clog P Values for Compounds 6a–g^a.

^aValues were calculated using ChemDraw Professional 17.1 and ALOGPS 2.1 software,²¹ respectively. Refer to Scheme 1 for full structural details.

Biological Evaluation

The antibacterial activities of the synthesized fluoroquinolones were evaluated against a panel of Gram-negative pathogens including K. pneumoniae (NCTC 13368, M6), Acinetobacter baumannii (AYE, ATCC 17978), Pseudomonas aeruginosa (PAO1, NCTC 13437), and Escherichia coli (NCTC 12923, LEC001). The ability of molecules to permeate the Gram-negative outer membrane was investigated by adding membrane permeabilizer polymyxin B nonapeptide (PMBN) at sub-inhibitory concentrations (Table 2). Phenylalanine–arginine β-naphthylamide (PAβN) was used to investigate efflux of compounds by resistance–nodulation–division (RND)-family efflux pumps (Table 3).

Table 2. Minimum Inhibitory Concentration (MIC) Values (mg/L) of Synthesized Fluoroquinolones^a.

	compound code		6a; ciprofloxacin		6b; norfloxacin		6c		6d		6e		6f		6g
species	strain	S/R	–PMBN	+PMBN	–PMBN	+PMBN	–PMBN	+PMBN	–PMBN	+PMBN	–PMBN	+PMBN	–PMBN	+PMBN	–PMBN	+PMBN
K. pneumoniae	M6	S	0.125	0.125	0.25–0.5	0.5	1	1	1–2	0.5	8	2	>128	>128	>128	>128
	NCTC 13368	R	1	0.5	4	2	16	4	16	2	32	8	>128	>128	>128	>128
A. baumannii	ATCC 17978	S	0.5	0.5	8	4	4–8	4	2	1	8	4	>128	>128	>128	>128
	AYE	R	>32	>32	>32	>32	>32	>32	>32	>32	>32	>32	>128	>128	>128	>128
P. aeruginosa	PAO1	S	0.5–2	0.125	2–4	0.5	8	0.5	8–16	0.5	32	2	>128	16	>128	32
	NCTC 13437	R	>32	16	>32	32	>32	32	>32	32	>32	32	>128	32	>128	16
E. coli	NCTC 12923	S	≤0.03	0.03–0.06	0.125	0.125	1–2	0.5	1	0.25	4	1	>128	32	>128	128
	LEC001	R	16	32	>32	>32	>32	>32	>32	>32	>32	>32	>32	>32	>32	>32

^aS/R refers to strains susceptibility to ciprofloxacin as defined by the EUCAST clinical breakpoints: S indicates a strain is ciprofloxacin-sensitive, while R indicates a strain is ciprofloxacin-resistant.

Table 3. MIC Values (mg/L) of Synthesized Fluoroquinolone (6e) vs Ciprofloxacin (6a) in the Presence or Absence of the Efflux Pump Inhibitor PAβN^a.

	compound code		6e		6a; ciprofloxacin
species	strain	S/R	–PAβN	+PAβN	–PAβN	+PAβN
K. pneumoniae	M6	S	8	1	≤0.125	≤0.125
	NCTC 13368	R	>32	4	0.25	0.25
A. baumannii	ATCC 17978	S	4	4	≤0.125	≤0.125–0.5
	AYE	R	>32	>32	32	>32
P. aeruginosa	PAO1	S	32	2	≤0.125	≤0.125
	NCTC 13437	R	>32	>32	32	16
E. coli	NCTC 12923	S	0.25–2	0.25	≤0.125	≤0.125
	LEC001	R	>32	>32	16	32

^aS/R refers to strains susceptibility to ciprofloxacin as defined by the EUCAST clinical breakpoints: S indicates a strain is ciprofloxacin-sensitive, while R indicates a strain is ciprofloxacin-resistant.

Across the panel, activity is broadly reduced or abolished in the N1-benzofused analogues compared to ciprofloxacin (6a) and norfloxacin (6b); MICs against ciprofloxacin-sensitive strains are all ≥4× higher than for ciprofloxacin. The uniformly poor activities of compounds 6f and 6g, the most hydrophobic of the compounds synthesized, are likely a result of their poor solubility profiles.

Klebsiella pneumoniae

For the ciprofloxacin-resistant strain NCTC 13368, PMBN rescues the activities of compounds 6d and 6e by 4–8-fold compared to the 2-fold difference seen for both norfloxacin and ciprofloxacin. This pattern extends to the ciprofloxacin-sensitive strain M6 (2–4-fold vs 1-fold potentiation). Using compound 6e as a reference compound, these strains also show significant potentiation with the RND-family efflux inhibitor PAβN, suggesting that the compounds are susceptible to efflux. This was not evident for ciprofloxacin.

Acinetobacter baumannii

The trend observed in both K. pneumoniae strains extends to the ciprofloxacin-sensitive A. baumannii strain ATCC 17978, with 6d and 6e potentiated 2-fold by PMBN versus 1–2-fold for ciprofloxacin and norfloxacin. The ciprofloxacin-resistant strain AYE remains highly resistant to all compounds tested, likely due to mutations in gyrA. Neither ciprofloxacin nor compound 6e showed any difference in MIC when treated in the presence of PAβN.

Pseudomonas aeruginosa

The aforementioned trend is also visible in ciprofloxacin-sensitive P. aeruginosa strain PAO1, with PMBN potentiating 6d and 6e 16 to >32-fold versus 4–16-fold for norfloxacin and ciprofloxacin. Unlike for both K. pneumoniae and A. baumannii, the activities of 6f and 6g were also rescued by over 8-fold against both strains, though all MICs remain well above the EUCAST clinical breakpoint for ciprofloxacin. PAO1, but not NCTC 13437, showed significant potentiation with PAβN.

Escherichia coli

PMBN rescues the activity of all analogues by 2 to >8 fold versus 1 to >4-fold for norfloxacin and ciprofloxacin. No effect was observed with PAβN for either compound 6e or ciprofloxacin.

In vitro activity assays using purified S. aureus, E. coli and P. aeruginosa DNA gyrase, and S. aureus topoisomerase IV enzymes were performed in order to evaluate whether the gains in inhibitory activity predicted for the benzofused compounds were accurate. The benzothiophene-containing 6e was selected as a representative compound for the purposes of this experiment. Results indicate that 6e has a markedly reduced inhibitory activity for the E. coli and P. aeruginosa enzymes (6e IC₅₀ = 2.70 ± 0.42 and 6.68 ± 1.46 μg/mL, respectively) compared to ciprofloxacin (IC₅₀ = 0.069 ± 0.01 and 0.11 ± 0.01 μg/mL, respectively) (Figure 3). An accurate IC₅₀ could not be determined for compound 6e against the S. aureus DNA gyrase or topoisomerase IV enzymes because the compound was poorly soluble in the assay media at the high concentrations required.

Figure 3.

IC₅₀ analysis for 6e against (A) E. coli DNA gyrase and (B) P. aeruginosa DNA gyrase.

Discussion

A general correlation between N1 substituent hydrophobicity and activity is observed; compounds 6c, 6d, and 6e are of intermediate hydrophobicity and display limited activity, whereas the more hydrophobic 6f and 6g show almost complete loss of activity versus ciprofloxacin. Compounds 6d and 6e are less active in K. pneumoniae, P. aeruginosa, and E. coli because of impaired outer membrane penetration, as evidenced by greater potentiation by PMBN versus ciprofloxacin and norfloxacin, lower inhibitory activity for DNA gyrase, as indicated by the E. coli and P. aeruginosa IC₅₀s of 6e, and—in the case of P. aeruginosa PAO1 and both K. pneumoniae strains—efflux by RND-family efflux pumps, as evidenced by conducting MICs for 6e with and without the efflux pump inhibitor PAβN. The outer membrane of A. baumannii strain ATCC 17978 does not seem to impede compound influx in the same way as described above and we can conclude that higher MIC values in the benzofused compounds in this strain are solely due to lower inhibitory activity. The uniformly poor activities of the most hydrophobic compounds 6f and 6g indicate that these analogues are chiefly affected by poor solubility profiles, as was evidenced when undertaking the MIC experiments. None of the N1-benzofused analogues showed superior activity to ciprofloxacin in ciprofloxacin-resistant strains, indicating that the resistance mechanisms that afflict ciprofloxacin in these bacteria are no less effective against the benzofused analogues.

Comparison of the molecular modeling and in vitro data highlights several important points. The weakness of any molecular model lies in the variables excluded from it; since whole-cell systems are far too large and complex to model currently, the lack of a membrane component to simulate compound influx remains a key shortcoming in antibacterial docking efforts. However, this does not mean that the predicted gains in gyrase inhibition are incorrect. For compounds 6d and 6e which were capable of influx into certain strains of bacteria, the problem may be that the ChemScore and binding energy parameters used in the docking are poorly suited to determine whether a fluoroquinolone will inhibit the topoisomerase target. While both parameters offer a guide to the most preferable binding pose of a drug by assessing a plethora of noncovalent binding interactions, neither considers the individual importance of said interactions beyond their relative contributions to likely binding energy. Specifically, fluoroquinolones inhibit topoisomerase enzymes by forming a C3 carboxylate magnesium-water bridge with aspartic/glutamic acid and serine residues in the GyrA/ParC enzyme subunits. Therefore, it is plausible that N1-benzofused analogues 6d and 6e do possess higher binding energies for the enzyme but show reduced inhibition of it because of an altered binding pose. We recommend that this possibility be considered in all future modeling efforts with novel fluoroquinolones. This study illustrates that while molecular modeling is a useful technique in drug design, it cannot yet be a full substitute for traditional in vitro and in vivo compound screening efforts.

Materials and Methods

Synthetic Organic Chemistry

General

Synthetic building blocks and reagents were purchased from a number of suppliers including Fluorochem (UK), Sigma-Aldrich (Merck KGaA, USA), Thermo Fisher Scientific (UK, including Acros Organics, Maybridge and Alfa Aesar), Activate Scientific (UK), Enamine (Ukraine), VWR International (USA), Oxchem (USA), and Apollo Scientific (UK). Solvents were purchased from Sigma-Aldrich and Thermo Fisher Scientific. Thin-layer chromatography (TLC) analysis was performed using silica gel plates (Merck silica gel 60 F254 plates) and visualized using ultraviolet (UV) light (254 nm wavelength) and/or staining with potassium permanganate solution. Manual flash column chromatography was performed using silica gel (Merck 9385, 230–400 mesh ASTM, 40–63 μM) as the stationary phase. TLC was employed to discern solvent systems (mobile phases) with appropriate separation profiles; said profiles were composed of hexanes and ethyl acetate. LC–MS was employed to monitor reaction progression and compound identification. LC–MS analysis was performed on a Waters Alliance 2695 HPLC coupled to a Waters Micromass ZQ instrument with a Waters 2996 PDA. For liquid chromatography, an Onyx monolithic C18 column (50 × 4.6 mm) was used and the mobile phases were composed of water (A) and acetonitrile (B). Formic acid (0.1%) was added to both to ensure acidic conditions throughout the analysis. Gradient conditions used were as follows. Method A (5 min): from 95% A/5% B to 90% B over 3 min. Then, from 90% B to 95% B over 0.5 min and held constant for 1 min. This was then reduced to 5% B over 0.5 min. The flow rate was 1.0 mL/min, 100 μL was split via a zero dead volume T piece which passed into the mass spectrometer. The wavelength range of the UV detector was 220–500 nm. Method B (10 min): from 95% A/5% B to 50% B over 3 min; then, from 50% B to 80% B over 2 min; and then from 80% B to 95% B over 1.5 min and held constant for 1.5 min. This was then reduced to 5% B over 0.2 min and maintained to 5% B for 1.8 min. The flow rate was 0.5 mL/min, 200 μL was split via a zero dead volume T piece which passed into the mass spectrometer. The wavelength range of the UV detector was 220–400 nm. Mass spectrometry data (both ESI+ and ESI– modes) were collected using the following Waters Micromass ZQ parameters: capillary (kV), 3.38; cone (V), 35; extractor (V), 3.0; source temperature (°C), 100; de-solvation temperature (°C), 200; cone flow rate (L/h), 50; and de-solvation flow rate (L/h), 250. High-resolution mass spectra were obtained on a Thermo Navigator mass spectrometer coupled with LC using electrospray ionization (ESI) and time-of-flight mass spectrometry. IR spectra were recorded on a PerkinElmer spectrum 1000 instrument. All NMR spectra were obtained at room temperature using a Bruker Ascend 400 MHz NMR spectrometer and interpreted using ACD/NMR Processor Academic Edition software. Chemical shifts (δH) are expressed in parts per million (ppm) relative to deuterated chloroform (CDCl₃, residual signal ¹Hδ = 7.26, ¹³Cδ = 77.2), deuterated dimethyl sulfoxide ((CD₃)₂SO, residual signal ¹Hδ = 2.54, ¹³Cδ = 40.5), or deuterated trifluoroacetic acid (CF₃CO₂D, residual signal ¹Hδ = 11.50, ¹³Cδ = 164.4, 116.5). Coupling constants are expressed in Hz. Multiplicities in ¹H NMR spectra are quoted as s = singlet, d = doublet, t = triplet, q = quartet, m = multiplet, dd = doublet of doublets, ddd = doublet of doublet of doublets, dt = doublet of triplets, td = triplet of doublets, spt = septet, and br = broad.

Microbiological Evaluation

Minimal Inhibitory Concentrations

Minimal inhibitory concentrations (MICs) were determined using the microdilution broth method. Briefly, compounds were added to the first two columns of a 96-well plate and diluted two-fold down the plate in tryptic soy broth. Overnight cultures of bacteria were then adjusted to a concentration of 1 × 10⁶ CFU/mL and added to each well for a final concentration of 5 × 10⁵ CFU/mL. The MIC was defined as the lowest concentration of compound which resulted in no visible growth at an optical density of 600 nm after 20 h incubation at 37 °C. Where necessary, MIC determinations were carried out in the presence of the membrane permeabiliser PMBN at a final concentration of 30 μg/mL, or an efflux pump inhibitor, PAβN, at a final concentration of 25 μg/mL, concentrations which did not affect bacterial growth.

Gyrase Inhibition Assay

DNA gyrase from S. aureus and E. coli were treated with compound 6e and the positive control ciprofloxacin using the cell-free S. aureus, E. coli, and P. aeruginosa gyrase supercoiling kits (#SAS4001, K0001, PAS001) and the S. aureus topoisomerase IV relaxation kit (SAR4001), all obtained from Inspiralis (Norwich, UK), according to the manufacturer’s instructions, as described previously.^22,23

Molecular Modeling

Generation of Structures

PDB ID code 2XCT was used for the 3D structure of S. aureus GyrA. The structure of K. pneumoniae GyrA was generated by homology modeling using the FASTA UniProt code A0A0C7K6P2. The sequence identity between the target and template protein with PDB ID code 4TMA was 96.57% for K. pneumoniae GyrA. The final structures of the gyrase and associated DNA structures were amended and assembled using Accelrys Discovery Studio. The 3D structures of ligands were generated using Chem3D 15.0 software and the structures were minimized with SYBYL.

Molecular Docking

Molecular docking was performed to generate several distinct binding orientations and the binding affinity for each binding mode. Subsequently, the binding modes that showed the highest score and lowest binding energy were considered as the most favorable binding modes. AutoDock SMINA was used for blind molecular docking of the ligands to gyrase enzymes to find the best binding pocket by exploring all probable binding cavities in the enzymes. All parameters were kept at their default values for running SMINA. GOLD was then used for molecular docking of the compounds into the SMINA-located binding site for performing flexible molecular docking and determining more precise energies and scores. Based on a fitness function score and ligand binding position, the best-docked pose for each compound was selected; high fitness function scores (generated using GOLD) and low binding energy values are indicative of the best-docked pose for each system. A genetic algorithm (GA) is used in gold ligand docking to thoroughly examine the ligand conformational flexibility along with the partial flexibility of the protein. The maximum number of runs was set to 20 for each compound, and the default parameters were selected (100 population size, 5 for the number of islands, 100,000 number of operations and 2 for the niche size). Default cutoff values of 2.5 Å (d_H–X) for hydrogen bonds and 4.0 Å for van-der-Waals distance were employed. When the top solutions attained the root-mean-square deviation values within 1.5 Å, the GA docking was terminated.

Conclusions

This study suggests large hydrophobic groups, particularly aromatic-benzofused rings, have limited utility at the N1 position of fluoroquinolones as they significantly alter the physicochemical and drug-like properties of the molecules. Electron-rich, hydrophobic, and benzofused substitutions negatively impact penetration of the Gram-negative outer membrane in K. pneumoniae, P. aeruginosa, and E. coli and do not afford protection from existing fluoroquinolone resistance mechanisms present in these species. These modifications also reduce target inhibition (other factors, including substituent size and geometry, are clearly intrinsic determinants of biological activity) and compound solubility in aqueous media. However, the modifications considered herein may not represent the full potential of such modifications; more highly functionalized, polar heteroaromatic groups may yet prove promising N1 substituents for the next generation of fluoroquinolones. The findings also suggest that, while still a powerful technique, molecular modeling alone was unable to successfully predict improvements to the fluoroquinolone scaffold at the N1 position, both because of considerations outside of the scope of the model (membrane penetration and compound solubility) and inaccuracies inherent to current modeling techniques (poor prediction of compound affinity for target with regards to inhibition).

Experimental Section

General Procedure for Compounds 2a–g

Ethyl 2,4,5-trifluorobenzoylacetate (1; 1 g, 4.06 mmol, 1 equiv) was dissolved in triethyl orthoformate (1.15 mL, 6.90 mmol, 1.7 equiv) and heated at 140 °C for 30 min. Acetic anhydride (1.15 mL, 12.19 mmol, 3 equiv) was added and the mixture refluxed at 140 °C for another 3 h. Upon completion, the reaction was cooled to room temperature, dichloromethane (DCM) (3 mL) was added and the mixture was stirred at room temperature for 10 min. Then, cyclopropylamine (703 μL, 10.15 mmol, 2.5 equiv) was added and the reaction was stirred at room temperature until completion. The crude was concentrated in vacuo and the resulting solid was dissolved in DCM (2 mL) and purified by flash chromatography (50% Hex/50% EtOAc) to yield compound 2a, pale yellow solid, 1.158 g (91.0% yield); ¹H NMR (400 MHz, CDCl₃): δ 10.87 (d, J = 12.59 Hz, 0.75H, H4), 9.42 (d, J = 13.60 Hz, 0.25H, H4), 8.20 (d, J = 13.85 Hz, 0.75H, H3), 8.16 (d, J = 14.35 Hz, 0.25H, H3), 7.33 (ddd, J = 10.07, 8.81, 6.29 Hz, 0.25H, H1), 7.19 (ddd, J = 9.82, 8.81, 6.29 Hz, 0.75H, H1), 6.83–6.91 (m, 1H, H2), 4.05 (q, J = 7.13 Hz, 1.5H, H8), 4.00 (q, J = 7.05 Hz, 0.5H, H8), 2.92–3.02 (m, 1H, H5), 1.08 (t, J = 7.18 Hz, 2.25H, H9), 0.77–0.97 (m, 4.75H, H6 + 7 + 9); ¹³C NMR (101 MHz, CDCl₃): δ 188.1, 186.0, 168.4, 166.6, 160.8, 160.4, 154.1 (ddd, 247.22, 9.54, 2.94 Hz), 150.5 (ddd, 253.09, 14.67, 12.47 Hz), 146.6 (ddd, 245.02, 12.47, 3.67 Hz), 127.2–127.6 (m), 117.7 (ddd, J = 20.54, 5.13, 1.47 Hz), 116.8 (ddd, J = 20.54, 5.13, 1.47 Hz), 105.1 (dd, J = 28.61, 21.27 Hz), 104.9 (dd, J = 28.61, 21.27 Hz), 101.6, 101.6, 59.9, 59.6, 30.4, 30.1, 14.0, 13.6, 6.6, 6.5; (ν_max/cm^–1): 1686, 1623, 1569, 1508, 1426, 1406, 1357, 1330, 1294, 1244, 1228, 1174, 1136, 1088, 1058, 1033, 1015, 890, 877, 809, 797, 773, 754, 734, 660, 588. See note A.

Compound 2b, pale yellow solid, 4.916 g (80.3% yield); ¹H NMR (400 MHz, CDCl₃): δ 10.89 (br s, 0.8H, H4), 9.40 (br s, 0.2H, H4), 8.12 (d, J = 14.10 Hz, 0.8H, H3), 8.10 (d, J = 14.60 Hz, 0.2H, H3), 7.32 (ddd, J = 9.82, 8.81, 6.29 Hz, 0.2H, H1), 7.19 (ddd, J = 9.82, 8.81, 6.29 Hz, 0.8H, H1), 6.81–6.92 (m, 1H, H2), 4.05 (q, J = 7.18 Hz, 1.6H, H7), 4.00 (q, J = 7.18 Hz, 0.4H, H7), 3.44–3.55 (m, 2H, H5), 1.36 (t, J = 7.30 Hz, 2.4H, H6), 1.33 (t, J = 7.30 Hz, 0.6H, H6), 1.07 (t, J = 7.18 Hz, 2.4H, H8), 0.94 (t, J = 7.18 Hz, 0.6H, H8); ¹³C NMR (101 MHz, CDCl₃): δ 188.0, 168.6 (d, J = 1.47 Hz), 166.8 (d, J = 1.47 Hz), 160.2, 160.0, 159.8, 154.1 (ddd, J = 247.22, 9.54, 2.20 Hz), 150.4 (ddd, J = 253.09, 14.67, 12.47 Hz), 146.6 (ddd, J = 245.02, 13.20, 3.67 Hz), 127.4–127.7 (m), 116.7 (ddd, J = 20.54, 5.87, 1.47 Hz), 105.0 (dd, J = 28.61, 20.54 Hz), 101.0, 100.8, 59.8, 59.5, 45.1, 44.9, 44.8, 15.8, 15.7, 15.7, 14.0, 13.6; (ν_max/cm^–1): 1678, 1622, 1560, 1510, 1428, 1415, 1380, 1364, 1330, 1310, 1283, 1244, 1219, 1175, 1156, 1136, 1092, 1050, 1036, 884, 1015, 863, 829, 800, 774. See note A.

Compound 2c, pale yellow solid, 1.039 g (70.7% yield); ¹H NMR (400 MHz, CDCl₃): δ 11.10 (br s, 0.75H), 9.63 (br s, 0.25H), 8.14–8.23 (m, 1H), 7.34–7.45 (m, 3H), 7.27–7.33 (m, 2H), 7.21 (ddd, J = 10.01, 8.75, 6.17 Hz, 1H), 6.83–6.92 (m, 1H), 4.58–4.64 (m, 2H), 4.06 (q, J = 7.05 Hz, 1.5H), 4.01 (q, J = 7.05 Hz, 0.5H), 1.08 (t, J = 7.05 Hz, 2.25H), 0.95 (t, J = 7.05 Hz, 0.75H); ¹³C NMR (101 MHz, CDCl₃): δ 188.3, 166.7, 160.4, 160.0, 135.7, 135.5, 129.1, 129.1, 128.5, 128.4, 127.5, 127.5, 116.8 (ddd, J = 1.47, 5.14, 20.54 Hz), 105.1 (dd, J = 21.27, 28.61 Hz), 105.0 (dd, J = 21.27, 28.61 Hz), 101.6, 101.5, 59.9, 59.6, 54.0, 53.8, 14.0, 13.6; (ν_max/cm^–1): 1676, 1628, 1604, 1434, 1379, 1366, 1329, 1311, 1301, 1231, 1175, 1139, 1050, 1019, 855, 844, 802, 777, 732, 721, 692, 652, 580. See notes A, B.

Compound 2d, white solid, 1.260 g (79.7% yield); ¹H NMR (400 MHz, CDCl₃): δ 12.83 (d, J = 13.85 Hz, 0.66H), 11.46 (d, J = 13.60 Hz, 0.33H), 8.99 (d, J = 13.60 Hz, 0.66H), 8.85 (d, J = 13.85 Hz, 0.33H), 7.74 (d, J = 2.27 Hz, 0.66H), 7.72 (d, J = 2.01 Hz, 0.33H), 7.44–7.51 (m, 1H), 7.42 (dd, J = 7.81, 1.01 Hz, 0.33H), 7.36 (ddd, J = 9.82, 8.81, 6.29 Hz, 0.66H), 7.24–7.31 (m, 1H), 7.20–7.24 (m, 0.66H), 7.16–7.20 (m, 0.33H), 6.92 (m, 1H), 6.86 (d, J = 2.01 Hz, 0.66H), 6.84 (d, J = 2.27 Hz, 0.33H), 4.12–4.19 (m, 2H), 1.15 (t, J = 7.18 Hz, 2H), 1.04 (t, J = 7.18 Hz, 1H); ¹³C NMR (101 MHz, CDCl₃): δ 189.0, 186.3, 168.2, 166.5, 154.5 (ddd, J = 2.93, 9.54, 247.95 Hz), 153.4, 152.5, 151.0 (ddd, J = 12.47, 14.67, 254.56 Hz), 146.7 (ddd, J = 3.67, 12.47, 245.75 Hz), 145.8, 145.7, 145.0, 144.8, 129.5, 129.4, 126.8–127.0 (m), 124.2, 123.9, 123.8, 118.6, 118.0–118.2 (m), 118.0, 117.2 (ddd, J = 1.47, 5.13, 20.54 Hz), 112.4, 111.5, 107.2, 105.2 (dd, J = 21.27, 28.61 Hz), 105.1 (dd, J = 21.27, 28.61 Hz), 104.8, 104.3, 60.3, 60.2, 14.0, 13.7; (ν_max/cm^–1): 1701, 1624, 1596, 1567, 1506, 1448, 1428, 1307, 1251, 1207, 1179, 1136, 1085, 1060, 1031, 1015, 976, 877, 812, 786, 729, 618, 599, 586, 566, 535, 519. See notes A, B.

Compound 2e, yellow solid, 1.354 g (86.5% yield); ¹H NMR (400 MHz, CDCl₃): δ 12.80 (d, J = 13.35 Hz, 0.66H), 11.31 (d, J = 13.60 Hz, 0.33H), 8.78 (d, J = 13.09 Hz, 0.66H), 8.66 (d, J = 13.60 Hz, 0.33H), 7.75 (dd, J = 7.93, 0.63 Hz, 0.66H), 7.71 (dd, J = 7.93, 0.63 Hz, 0.33H), 7.55 (dd, J = 5.29, 0.50 Hz, 0.66H), 7.53 (dd, J = 5.29, 0.50 Hz, 0.33H), 7.43–7.50 (m, 2H), 7.28–7.43 (m, 2H), 6.88–6.97 (m, 1H), 4.13–4.21 (m, 2H), 1.16 (t, J = 7.18 Hz, 2H), 1.05 (t, J = 7.18 Hz, 1H); ¹³C NMR (101 MHz, CDCl₃): δ 189.1, 186.3, 168.3, 166.3, 153.2, 152.4, 141.7, 141.6, 133.9, 133.8, 131.0, 126.7, 126.4, 125.6, 125.5, 124.9, 124.8, 121.7, 121.1, 117.2 (ddd, J = 1.47, 5.14, 20.54 Hz), 112.6, 111.8, 105.2 (dd, J = 21.27, 28.61 Hz), 105.2 (dd, J = 21.27, 28.61 Hz), 104.4, 60.4, 60.3, 14.0, 13.7; (ν_max/cm^–1): 1693, 1620, 1595, 1507, 1435, 1383, 1324, 1293, 1248, 1186, 1139, 1100, 1020, 978, 893, 787, 761, 734, 701, 666, 593, 564. See notes A, B.

Compound 2f, pale yellow solid, 491 mg (85.7% yield); ¹H NMR (400 MHz, CDCl₃): δ 11.10–11.32 (m, 0.75H), 9.64–9.81 (m, 0.25H), 8.32 (d, J = 13.85 Hz, 0.75H), 8.28 (d, J = 14.35 Hz, 0.25H), 7.68–7.72 (m, 1H), 7.60–7.64 (m, 1H), 7.33 (ddd, J = 10.07, 8.81, 6.29 Hz, 0.25H), 7.16–7.30 (m, 3H), 6.82–6.91 (m, 2H), 4.86–4.92 (m, 2H), 4.07 (q, J = 7.05 Hz, 1.5H), 4.00 (q, J = 7.05 Hz, 0.5H), 1.08 (t, J = 7.05 Hz, 2.25H), 0.94 (t, J = 7.05 Hz, 0.75H); ¹³C NMR (101 MHz, CDCl₃): δ 188.2, 186.1, 168.4, 166.7, 160.7, 160.2, 154.1 (ddd, J = 2.93, 9.54, 247.22 Hz), 152.7, 146.6 (ddd, J = 3.67, 12.47, 245.75 Hz), 145.4, 127.9, 127.9, 127.3–127.6 (m), 123.5, 123.5, 123.2, 121.9, 121.8, 119.6, 119.4, 117.6 (ddd, J = 1.47, 5.14, 20.54 Hz), 116.8 (ddd, J = 1.47, 5.14, 20.54 Hz), 106.9, 106.9, 105.0 (dd, J = 20.54, 28.61 Hz), 105.0 (dd, J = 20.54, 28.61 Hz), 101.6, 101.5, 59.9, 59.6, 49.2, 49.0, 14.0, 13.6; (ν_max/cm^–1): 1690, 1630, 1576, 1519, 1509, 1428, 1385, 1365, 1331, 1305, 1255, 1228, 1170, 1135, 1095, 1071, 1028, 967, 879, 812, 783, 741, 658, 627, 584, 567, 510. See notes A, B.

Compound 2g, white solid, 348 mg (81.3% yield); ¹H NMR (400 MHz, CDCl₃): δ 11.19 (br s, 0.9H), 9.70 (br s, 0.1H), 8.29 (d, J = 13.85 Hz, 0.9H), 8.23 (d, J = 14.35 Hz, 0.1H), 7.82–7.88 (m, 1H), 7.48–7.53 (m, 1H), 7.39–7.45 (m, 2H), 7.31 (d, J = 6.80 Hz, 1H), 7.21 (ddd, J = 10.07, 8.81, 6.17 Hz, 1H), 6.87 (ddd, J = 10.07, 9.06, 6.29 Hz, 1H), 4.83–4.88 (m, 2H), 4.07 (q, J = 7.05 Hz, 1.8H), 4.01 (q, J = 7.05 Hz, 0.2H), 1.08 (t, J = 7.05 Hz, 2.7H), 0.94 (t, J = 7.05 Hz, 0.3H); ¹³C NMR (101 MHz, CDCl₃): δ 188.4, 168.4, 166.6, 160.7, 160.2, 140.5, 138.2, 138.1, 129.6, 126.4, 126.4, 124.9, 124.8, 124.7, 124.6, 124.1, 124.1, 123.4, 123.4, 116.9 (ddd, J = 1.47, 5.14, 20.54 Hz), 105.1 (dd, J = 21.27, 28.61 Hz), 101.9, 60.0, 59.7, 53.1, 52.9, 14.0, 13.6; (ν_max/cm^–1): 1695, 1627, 1573, 1507, 1430, 1383, 1329, 1303, 1256, 1229, 1217, 1171, 1134, 1091, 1070, 1032, 995, 846, 813, 782, 742, 704, 660, 587, 568, 552. See notes A, B.

General Procedure for Compounds 3a–g

Compound 2a (1 g, 3.13 mmol, 1 equiv) was dissolved in DCM (7 mL), then DBU (573 μL, 3.83 mmol, 1.2 equiv) and LiCl (270 mg, 6.38 mmol, 2 equiv) were added and the reaction was stirred at 45 °C for 2.5 h, then at room temperature for 15 h. Upon completion, the mixture was extracted with DCM (2 × 20 mL) and washed with distilled water (15 mL) with the aqueous phase neutralized using a 1 M solution of citric acid (3 mL). The combined organic phases were dried over magnesium sulphate, filtered, and concentrated in vacuo. Crude product 3a was used in the successive reaction without further purification. Compound 3a, pale yellow solid, 1.022 g (>95% crude yield), ¹H NMR (400 MHz, CDCl₃): δ 8.55 (s, 1H, H3), 8.20 (dd, J = 10.45, 8.69 Hz, 1H, H1), 7.72 (dd, J = 11.33, 6.29 Hz, 1H, H2), 4.37 (q, J = 7.05 Hz, 2H, H7), 3.41–3.48 (m, 1H, H4), 1.32–1.43 (m, 5H, H5 + 6 + 8), 1.12–1.19 (m, 2H, H5 + 6); ¹³C NMR (101 MHz, CDCl₃): δ 172.6 (d, J = 1.47 Hz), 165.2, 153.3 (dd, J = 256.03, 15.41 Hz), 148.8, 148.6 (dd, J = 250.89, 13.94 Hz), 137.5 (dd, J = 9.54, 2.20 Hz), 125.6 (dd, J = 5.13, 2.20 Hz), 115.3 (dd, J = 19.07, 2.20 Hz), 110.8, 105.5 (d, J = 22.74 Hz), 61.0, 34.8, 14.3, 8.2; (ν_max/cm^–1): 1723, 1617, 1602, 1490, 1479, 1454, 1446, 1424, 1396, 1386, 1379, 1335, 1314, 1287, 1228, 1209, 1202, 1167, 1121, 1094, 1054, 1033, 1018, 899, 855, 849, 826, 802, 781, 748, 729, 717, 619, 607, 595, 548, 540; LC–MS retention time 3.28 min (method A), purity = 100%, found, 294.0 [M + H]⁺, calcd for C₁₅H₁₃F₂NO₃, 294.27 [M + H]⁺.

Compound 3b, pale yellow solid, 4.55 g (>95% crude yield), ¹H NMR (400 MHz, CDCl₃): δ 8.47 (s, 1H, H3), 8.27 (dd, J = 10.32, 8.81 Hz, 1H, H1), 7.27 (dd, J = 11.21, 6.17 Hz, 1H, H2), 4.38 (q, J = 7.05 Hz, 2H, H6), 4.21 (q, J = 7.22 Hz, 2H, H4), 1.55 (t, J = 7.30 Hz, 3H, H5), 1.40 (t, J = 7.05 Hz, 3H, H7); ¹³C NMR (101 MHz, CDCl₃): δ 172.6 (d, J = 2.20 Hz), 165.3, 153.5 (dd, J = 256.02, 15.41 Hz), 148.7, 148.3 (dd, J = 251.62, 13.94 Hz), 135.5 (dd, J = 8.80, 2.20 Hz), 126.4 (dd, J = 5.13, 2.20 Hz), 115.7 (dd, J = 18.34, 2.20 Hz), 111.0, 104.6 (d, J = 22.01 Hz), 61.0, 49.3, 14.3, 14.3; (ν_max/cm^–1): 1720, 1617, 1566, 1466, 1449, 1375, 1369, 1311, 1288, 1228, 1217, 1209, 1173, 1157, 1137, 1094, 1071, 1049, 1016, 902, 863, 829, 814, 802; LC–MS retention time 3.18 min (method A), purity = 100%, found, 281.9 [M + H]⁺, calcd for C₁₄H₁₃F₂NO₃, 282.26 [M + H]⁺.

Compound 3c, pale yellow solid, 540 mg (57.2% crude yield); ¹H NMR (400 MHz, CDCl₃): δ 8.59 (s, 1H), 8.28 (dd, J = 10.45, 8.69 Hz, 1H), 7.38–7.41 (m, 2H), 7.33–7.43 (m, 3H), 7.09–7.20 (m, 3H), 5.35 (s, 2H), 4.40 (q, J = 7.22 Hz, 2H), 1.41 (t, J = 7.18 Hz, 3H); ¹³C NMR (101 MHz, CDCl₃): δ 172.7 (d, J = 1.47 Hz), 165.2, 153.3 (dd, J = 15.41, 256.03 Hz), 149.9, 148.5 (dd, J = 13.94, 251.62 Hz), 136.1 (d, J = 11.00 Hz), 133.3, 129.6, 128.9, 126.4 (dd, J = 2.20, 5.14 Hz), 126.0, 115.6 (dd, J = 2.20, 18.34 Hz), 111.1, 105.7 (d, J = 22.74 Hz), 61.1, 57.9, 14.3; (ν_max/cm^–1): 1718, 1617, 1598, 1492, 1454, 1387, 1287, 1228, 1206, 1168, 1077, 922, 894, 825, 800, 731, 703, 620, 536; LC–MS retention time 3.62 min (method A), purity = 100%, found, 344.1 [M + H]⁺, calcd for C₁₉H₁₅F₂NO₃, 344.33 [M + H]⁺.

Compound 3d, pale yellow solid, 846 mg (89.1% crude yield); ¹H NMR (400 MHz, CDCl₃): δ 8.56 (s, 1H), 8.33 (dd, J = 10.32, 8.56 Hz, 1H), 7.88 (dd, J = 7.81, 1.01 Hz, 1H), 7.66 (d, J = 2.01 Hz, 1H), 7.49 (t, J = 7.68 Hz, 1H), 7.41 (dd, J = 7.68, 1.13 Hz, 1H), 6.98 (d, J = 2.27 Hz, 1H), 6.68 (dd, J = 11.08, 6.29 Hz, 1H), 4.39 (q, J = 7.05 Hz, 2H), 1.39 (t, J = 7.05 Hz, 3H); ¹³C NMR (101 MHz, CDCl₃): δ 172.9 (d, J = 2.20 Hz), 165.0, 153.4 (dd, J = 14.67, 256.02 Hz), 149.4, 149.0, 148.7 (dd, J = 13.94, 251.62 Hz), 146.6, 137.2 (dd, J = 2.20, 9.54 Hz), 130.4, 125.4–125.5 (m), 124.6, 124.2, 123.9, 122.9, 115.2 (dd, J = 2.20, 18.34 Hz), 111.8, 107.5, 106.3 (d, J = 22.74 Hz), 61.2, 14.4; (ν_max/cm^–1): 1685, 1646, 1612, 1566, 1491, 1436, 1279, 1211, 1168, 1116, 1085, 1048, 1009, 904, 855, 800, 738, 631, 604, 561; LC–MS retention time 3.78 min (method A), purity = 100%, found, 370.0 [M + H]⁺, calcd for C₂₀H₁₃F₂NO₄, 370.32 [M + H]⁺.

Compound 3e, pale yellow solid, 817 mg (85.8% crude yield); ¹H NMR (400 MHz, CDCl₃): δ 8.56 (s, 1H), 8.32 (dd, J = 10.32, 8.56 Hz, 1H), 8.09 (dd, J = 7.93, 0.88 Hz, 1H), 7.61–7.66 (m, 1H), 7.54 (m, 2H), 7.47 (dd, J = 7.55, 0.76 Hz, 1H), 6.64 (dd, J = 10.83, 6.42 Hz, 1H), 4.38 (q, J = 7.05 Hz, 2H), 1.38 (t, J = 7.18 Hz, 3H); ¹³C NMR (101 MHz, CDCl₃): δ 172.9 (d, J = 1.47 Hz), 164.8, 153.5 (dd, J = 14.67, 256.03 Hz), 148.8 (dd, J = 13.94, 251.62 Hz), 148.7, 142.4, 137.4, 136.5–136.6 (m), 134.3, 128.1, 126.5, 125.9, 125.5 (dd, J = 2.20, 5.14 Hz), 124.9, 123.4, 115.3 (dd, J = 2.20, 19.07 Hz), 112.0, 106.2 (d, J = 22.74 Hz), 61.2, 14.3; (ν_max/cm^–1): 1729, 1619, 1556, 1492, 1397, 1383, 1318, 1285, 1243, 1217, 1202, 1170, 1072, 1029, 901, 851, 803, 727, 699, 618, 605; LC–MS retention time 3.88 min (method A), purity = 100%, found, 385.9 [M + H]⁺, calcd for C₂₀H₁₃F₂NO₃S, 386.38 [M + H]⁺.

Compound 3f, pale yellow solid, 405 mg (94.3% crude yield); ¹H NMR (400 MHz, CDCl₃): δ 8.77 (s, 1H), 8.22–8.30 (m, 1H), 7.68–7.71 (m, 1H), 7.62 (d, J = 7.81 Hz, 1H), 7.38 (dd, J = 11.33, 6.04 Hz, 1H), 7.20–7.26 (m, 1H), 7.04 (d, J = 7.55 Hz, 1H), 6.82–6.86 (m, 1H), 5.61 (s, 2H), 4.41 (q, J = 7.05 Hz, 2H), 1.42 (t, J = 7.05 Hz, 3H); ¹³C NMR (101 MHz, CDCl₃): δ 172.8, 165.3, 153.3 (dd, J = 13.94, 256.02 Hz), 152.0, 150.2, 148.4 (dd, J = 13.94, 251.62 Hz), 145.6, 136.1 (dd, J = 1.47, 9.54 Hz), 128.3, 126.3 (dd, J = 2.20, 5.14 Hz), 123.5, 122.5, 122.3, 117.0, 115.6 (dd, J = 2.20, 18.34 Hz), 111.1, 107.1, 105.4 (d, J = 23.47 Hz), 61.1, 53.1, 14.4; (ν_max/cm^–1): 1674, 1650, 1612, 1493, 1427, 1317, 1290, 1247, 1220, 1126, 1093, 1031, 905, 874, 827, 802, 784, 750, 724, 711, 631, 606, 555, 525; LC–MS retention time 3.73 min (method A), purity = 100%, found, 383.9 [M + H]⁺, calcd for C₂₁H₁₅F₂NO₄, 384.35 [M + H]⁺.

Compound 3g, pale yellow solid, 290 mg (>95% crude yield); ¹H NMR (400 MHz, CDCl₃): δ 8.69 (s, 1H), 8.26–8.32 (m, 1H), 7.84–7.90 (m, 1H), 7.53 (d, J = 5.54 Hz, 1H), 7.47 (m, 1H), 7.34–7.40 (m, 1H), 7.08 (dd, J = 11.08, 6.29 Hz, 1H), 6.96–7.02 (m, 1H), 5.54 (s, 2H), 4.40 (q, J = 7.05 Hz, 2H), 1.41 (t, J = 7.18 Hz, 3H); ¹³C NMR (101 MHz, CDCl₃): δ 172.8 (d, J = 1.47 Hz), 165.1, 153.4 (dd, J = 15.41, 256.76 Hz), 150.2, 148.6 (dd, J = 13.94, 252.36 Hz), 140.9, 136.7, 136.1–136.3 (m), 127.3, 126.6, 126.2–126.4 (m), 125.1, 124.9, 124.5, 121.8, 115.7 (dd, J = 2.20, 18.34 Hz), 111.4, 105.3 (d, J = 22.74 Hz), 61.3, 57.0, 14.4; (ν_max/cm^–1): 1674, 1647, 1613, 1477, 1395, 1316, 1290, 1223, 1172, 1085, 1051, 830, 802, 786, 680, 605, 564; LC–MS retention time 3.85 min (method A), purity = 100%, found, 400.1 [M + H]⁺, calcd for C₂₁H₁₅F₂NO₃S, 400.41 [M + H]⁺.

General Procedure for Compounds 4a, 4b

Compound 3a (700 mg, 2.39 mmol, 1 equiv) was refluxed with concentrated HCl (3.5 mL) and concentrated AcOH (13 mL) for 2.5 h. The mixture was allowed to cool to room temperature and the resulting precipitate was filtered, washed with distilled water (3 mL), and dried. Crude product 4a was used in the successive reaction without further purification. Compound 4a, white solid, 573 mg (90.4% yield); ¹H NMR (400 MHz, CF₃CO₂D): δ 9.43 (s, 1H, H3), 8.41–8.52 (m, 2H, H1 + 2), 4.07–4.18 (m, 1H, H4), 1.64–1.73 (m, 2H, H5 + 6), 1.41–1.49 (m, 2H, H5 + 6); ¹³C NMR (101 MHz, CF₃CO₂D): δ 174.8 (d, J = 4.01 Hz), 171.7, 160.1 (dd, J = 270.27, 15.26 Hz), 154.6 (dd, J = 263.79, 14.50 Hz), 152.7, 142.7 (d, J = 10.87 Hz), 120.8 (dd, J = 8.20, 1.72 Hz), 115.8 (dd, J = 20.79, 3.24 Hz), 110.5 (d, J = 23.84 Hz), 106.9, 41.5, 10.0; (ν_max/cm^–1): 1719, 1614, 1556, 1421, 1332, 1303, 1289, 1231, 1204, 1056, 1033, 1020, 891, 806, 778, 748, 719, 606; LC–MS retention time 3.37 min (method A), purity = 100%, found, 265.9 [M + H]⁺, calcd for C₁₃H₉F₂NO₃, 266.22 [M + H]⁺.

Compound 4b, pale yellow solid, 2.964 g (73.2% yield); ¹H NMR (400 MHz, CF₃CO₂D): δ 9.36 (s, 1H, H3), 8.39 (t, J = 8.43 Hz, 1H, H1), 7.99 (dd, J = 9.95, 6.17 Hz, 1H, H2), 4.79 (q, J = 7.13 Hz, 2H, H4), 1.66 (t, J = 7.05 Hz, 3H, H5); ¹³C NMR (101 MHz, CF₃CO₂D): δ 171.8 (d, J = 4.01 Hz), 169.1, 157.7 (dd, J = 270.27, 15.26 Hz), 151.9 (dd, J = 263.79, 14.50 Hz), 149.4 (d, J = 1.15 Hz), 137.9 (d, J = 9.15 Hz), 118.8 (dd, J = 7.63, 1.33 Hz), 113.4 (dd, J = 20.41, 3.43 Hz), 107.1 (d, J = 23.27 Hz), 104.7, 53.4, 12.9; (ν_max/cm^–1): 1719, 1617, 1484, 1396, 1385, 1361, 1306, 1289, 1231, 1213, 1094, 1042, 948, 900, 874, 808; LC–MS retention time 3.28 min (method A), purity = 100%, found, 253.8 [M + H]⁺, calcd for C₁₂H₉F₂NO₃, 254.2 [M + H]⁺.

General Procedure for Compounds 4c–g

Crude compound 3c (450 mg, 1.31 mmol, 1 equiv) was refluxed with LiOH (157 mg, 6.55 mmol, 5 equiv), water (3.28 mL), and dioxane (6 mL) at 100 °C for 2 h. The mixture was allowed to cool to room temperature. The resulting precipitate was acidified using 1 M HCl, filtered, washed with water (3 mL), and dried in vacuo. Crude product 4c was used in the successive reaction without further purification. Compound 4c, white solid, 170 mg (41.2% yield); ¹H NMR (400 MHz, CF₃CO₂D): δ 9.39 (s, 1H), 8.45 (t, J = 8.39 Hz, 1H), 8.05 (dd, J = 10.27, 6.24 Hz, 1H), 7.38–7.48 (m, 3H), 7.18–7.27 (m, 2H), 5.92 (s, 2H); (ν_max/cm^–1): 1718, 1615, 1502, 1465, 1387, 1363, 1287, 1223, 1168, 974, 919, 837, 807, 773, 753, 731, 716, 692, 607, 526, 522; LC–MS retention time 3.70 min (method A), purity = 100%, found, 316.0 [M + H]⁺, calcd for C₁₇H₁₁F₂NO₃, 316.28 [M + H]⁺. See note C.

Compound 4d, white solid, 625 mg (84.4% yield); ¹H NMR (400 MHz, CF₃CO₂D): δ 9.48 (s, 1H), 8.56 (t, J = 8.25 Hz, 1H), 8.07 (d, J = 7.79 Hz, 1H), 7.53–7.68 (m, 3H), 7.33 (dd, J = 9.90, 6.24 Hz, 1H), 7.04 (d, J = 2.02 Hz, 1H); ¹³C NMR (101 MHz, CF₃CO₂D): δ 175.7 (d, J = 3.82 Hz), 171.6, 160.4 (dd, J = 15.64, 271.03 Hz), 154.7 (dd, J = 14.31, 263.4 Hz), 153.8, 150.2, 149.8, 142.5 (d, J = 10.11 Hz), 133.8, 128.7, 126.8, 124.6, 123.8, 120.8 (d, J = 8.39 Hz), 115.5–115.7 (m), 111.2 (d, J = 23.84 Hz), 109.8, 107.7; (ν_max/cm^–1): 1718, 1610, 1492, 1456, 1334, 1290, 1263, 1220, 1208, 1017, 897, 851, 806, 793, 736, 720, 598, 562; LC–MS retention time 3.78 min (method A), purity = 100%, found, 342.0 [M + H]⁺, calcd for C₁₈H₉F₂NO₄, 342.27 [M + H]⁺.

Compound 4e, pale yellow solid, 513 mg (73.6% yield); ¹H NMR (400 MHz, CF₃CO₂D): δ 9.51 (s, 1H), 8.58 (t, J = 8.21 Hz, 1H), 8.28 (d, J = 8.07 Hz, 1H), 7.77 (t, J = 7.84 Hz, 1H), 7.58–7.67 (m, 3H), 7.31 (dd, J = 9.86, 6.28 Hz, 1H); ¹³C NMR (101 MHz, CF₃CO₂D): δ 175.6 (d, J = 4.20 Hz), 171.3, 161.7 (dd, J = 15.64, 271.23 Hz), 154.6 (dd, J = 14.11, 264.36 Hz), 153.0, 145.6, 141.4–141.6 (m), 138.0, 134.8, 130.0, 128.1, 127.2, 124.5, 120.7–120.8 (m), 115.5 (dd, J = 3.05, 20.60 Hz), 110.9, (d, J = 24.03 Hz), 107.7; (ν_max/cm^–1): 1612, 1491, 1477, 1460, 1388, 1279, 1218, 1023, 895, 850, 807, 746, 730, 697, 663, 648, 620; LC–MS retention time 3.93 min (method A), purity = 100%, found, 357.9 [M + H]⁺, calcd for C₁₈H₉F₂NO₃S, 358.33.33 [M + H]⁺.

Compound 4f, white solid, 176 mg (50.0% yield); ¹H NMR (400 MHz, CF₃CO₂D): δ 9.69 (s, 1H), 8.42 (t, J = 8.34 Hz, 1H), 8.27 (dd, J = 10.41, 6.10 Hz, 1H), 7.68 (d, J = 7.34 Hz, 1H), 7.55 (d, J = 2.11 Hz, 1H), 7.25–7.39 (m, 2H), 6.78 (d, J = 2.11 Hz, 1H), 6.21 (s, 2H); (ν_max/cm^–1): 1615, 1492, 1431, 1387, 1360, 1286, 1217, 1179, 1124, 1050, 1033, 1020, 957, 919, 903, 873, 840, 807, 740, 700, 607, 526; retention time 3.83 min (method A), purity = 100%, found, 356.0 [M + H]⁺, calcd for C₁₉H₁₁F₂NO₄, 356.30 [M + H]⁺. See note C.

Compound 4g, white solid, 170 mg (68.0% yield); ¹H NMR (400 MHz, CF₃CO₂D): δ 9.43 (s, 1H), 8.45 (t, J = 8.39 Hz, 1H), 8.10 (dd, J = 10.27, 6.24 Hz, 1H), 7.93 (d, J = 8.07 Hz, 1H), 7.38–7.49 (m, 3H), 7.24 (d, J = 7.43 Hz, 1H), 6.16 (s, 2H); ¹³C NMR (101 MHz, CF₃CO₂D): δ 174.9 (d, J = 4.20 Hz), 171.6, 160.0 (dd, J = 14.88, 270.84 Hz), 154.4 (dd, J = 14.88, 264.36 Hz), 152.6, 144.2, 141.1–141.3 (m), 140.1, 128.4, 128.2, 127.6, 127.3, 126.8, 126.0, 121.2–121.3 (m), 115.9 (dd, J = 3.81, 20.22 Hz), 110.1 (d, J = 23.27 Hz), 107.0, 62.9; (ν_max/cm^–1): 1617, 1340, 1289, 1213, 1049, 821, 788, 743, 687, 605; LC–MS retention time 3.93 min (method A), purity = 100%, found, 371.9 [M + H]⁺, calcd for C₁₉H₁₁F₂NO₃S, 372.36 [M + H]⁺.

General Procedure for Compounds 5a, 5b

A mixture of 4a (500 mg, 1.89 mmol, 1 equiv), tert-butyl piperazine-1-carboxylate (1.06 g, 5.67 mmol, 3 equiv), and potassium carbonate (552 mg, 3.78 mmol, 2 equiv) was stirred in dimethylformamide (DMF) (18 mL) at 140 °C for 15 h. Upon completion, the mixture was extracted with DCM (2 × 15 mL) and washed with distilled water (10 mL) with the aqueous layer neutralized using a 1 M citric acid solution (3 mL). Combined organic layers were dried over magnesium sulphate, filtered, and concentrated in vacuo. The crude solid was recrystallized from DMF (5 mL) to yield Boc-protected ciprofloxacin. This compound (200 mg, 0.46 mmol, 1 equiv) was then dissolved in dry DCM (7 mL), the mixture cooled to 0 °C and TFA (710 μL, 9.27 mmol, 20 equiv) was added, and then allowed to warm to room temperature over 2 h. Upon completion, the solution was concentrated in vacuo and washed with toluene (4 × 3 mL). The crude was washed with EtOAc (5 mL) and MeOH (5 mL) to give pure compound 5a, pale orange solid, 150 mg (34.5% yield over two steps); ¹H NMR (400 MHz, CF₃CO₂D): δ 9.29 (s, 1H, H3), 8.24 (d, J = 12.34 Hz, 1H, H1), 7.90 (d, J = 6.80 Hz, 1H, H2), 4.03–4.12 (m, 1H, H4), 3.90–3.99 (m, 4H, H8 + 9), 3.70–3.78 (m, 4H, H10 + 11), 1.60–1.68 (m, 2H, H5 + 6), 1.34–1.43 (m, 2H, H5 + 6); ¹³C NMR (101 MHz, CF₃CO₂D): δ 173.2, 172.4, 157.6 (d, J = 258.64 Hz), 151.5, 150.5 (d, J = 10.87 Hz), 143.5, 117.9 (d, J = 10.11 Hz), 114.2 (d, J = 25.18 Hz), 108.0, 105.7, 48.5 (d, J = 5.34 Hz), 46.8, 40.8, 9.9; (ν_max/cm^–1): 1685, 1627, 1612, 1490, 1454, 1341, 1272, 1259, 1184, 1138, 1107, 1056, 1034, 941, 894, 886, 829, 807, 793, 785, 749, 723, 708, 665, 637, 609; LC–MS retention time 2.48 min (method A), purity = 100%, found, 332.0 [M + H]⁺, calcd for C₁₇H₁₈FN₃O₃, 332.35 [M + H]⁺.

Compound 5b, pale brown solid, 44 mg (33.3% yield over two steps); ¹H NMR (400 MHz, CF₃CO₂D): δ 9.30 (s, 1H, H3), 8.29 (d, J = 12.09 Hz, 1H, H1), 7.48 (d, J = 6.55 Hz, 1H, H2), 4.85 (q, J = 7.22 Hz, 2H, H4), 3.91–3.99 (m, 4H, H7+8), 3.71–3.79 (m, 4H, H9 + 10), 1.74 (t, J = 7.30 Hz, 3H, H5); ¹³C NMR (101 MHz, CF₃CO₂D): δ 170.3, 169.7, 154.9 (d, J = 258.64 Hz), 148.2, 147.9 (d, J = 10.68 Hz), 138.8, 115.9 (d, J = 9.54 Hz), 111.7 (d, J = 27.47 Hz), 104.6, 103.3, 52.6, 45.9 (d, J = 6.10 Hz), 44.2, 12.7; (ν_max/cm^–1): 1696, 1624, 1610, 1508, 1474, 1453, 1421, 1399, 1382, 1366, 1310, 1265, 1202, 1125, 1104, 1088, 1053, 1033, 990, 934, 916, 900, 828, 808, 797, 748, 720; LC–MS retention time 2.38 min (method A), purity = 85.5%, found, 320.0 [M + H]⁺, calcd for C₁₆H₁₈FN₃O₃, 320.33 [M + H]⁺.

General Procedure for Compounds 5c–5g

A mixture of compound 4c (150 mg, 0.48 mmol, 1 equiv), piperazine (248 mg, 2.88 mmol, 6 equiv), and potassium carbonate (133 mg, 0.96 mmol, 2 equiv) was stirred in DMF (2.5 mL) at 140 °C for 2 h. Upon completion (as monitored by LC–MS), the mixture was extracted with DCM (2 × 15 mL) and washed with water (10 mL) with the aqueous layer neutralized using a 1 M solution of citric acid (2 mL). The combined organic layers were dried over magnesium sulfate and concentrated in vacuo. The crude compound was recrystallized from DMF (500 μL) and the crude solid washed with EtOAc (5 mL) and MeOH (5 mL) to give pure compound 5c, white solid, 140 mg (76.5% yield); ¹H NMR (400 MHz, CF₃CO₂D): δ 9.29 (s, 1H), 8.23 (d, J = 12.47 Hz, 1H), 7.41–7.47 (m, 3H), 7.39 (d, J = 6.88 Hz, 1H), 7.18–7.28 (m, 2H), 5.91 (s, 2H), 3.69–3.86 (m, 4H), 3.52–3.69 (m, 4H); ¹³C NMR (101 MHz, CF₃CO₂D): δ 173.1, 172.2, 157.2 (d, J = 258.06 Hz), 151.4, 149.9 (d, J = 10.87 Hz), 141.7, 133.0, 132.3, 132.1, 128.9, 118.3 (d, J = 9.92 Hz), 114.0 (d, J = 26.13 Hz), 108.4, 105.5, 63.4, 48.1 (d, J = 5.53 Hz), 46.4; (ν_max/cm^–1): 1622, 1578, 1498, 1447, 1390, 1337, 1278, 1262, 1215, 1199, 1179, 1137, 1032, 943, 825, 800, 778, 741, 696, 625, 536, 516; LC–MS retention time 2.68 min (method A), purity = 100%, found, 382.0 [M + H]⁺, calcd for C₂₁H₂₀FN₃O₃, 382.41 [M + H]⁺.

Compound 5d, white solid, 231 mg (32.2% yield); ¹H NMR (400 MHz, CF₃CO₂D): δ 9.35 (s, 1H), 8.41 (d, J = 12.29 Hz, 1H), 8.11 (dd, J = 7.79, 0.83 Hz, 1H), 7.71 (d, J = 2.20 Hz, 1H), 7.66 (t, J = 7.84 Hz, 1H), 7.57–7.62 (m, 1H), 7.10 (d, J = 2.29 Hz, 1H), 6.83 (d, J = 6.88 Hz, 1H), 3.55–3.77 (m, 8H); ¹³C NMR (101 MHz, CF₃CO₂D): δ 173.7, 172.0, 157.4 (d, J = 258.45 Hz), 152.4, 150.4 (d, J = 10.49 Hz), 150.1, 149.4, 143.1, 133.5, 128.2, 126.6, 124.5, 123.6, 117.7 (d, J = 10.87 Hz), 113.8 (d, J = 25.37 Hz), 109.7, 108.3, 106.2, 47.9 (d, J = 4.77 Hz), 46.3; (ν_max/cm^–1): 1595, 1493, 1433, 1379, 1330, 1289, 1259, 1202, 1032, 909, 876, 800, 736, 688, 624, 561, 552; LC–MS retention time 2.72 min (method A), purity = 100%, found, 408.0 [M + H]⁺, calcd for C₂₂H₁₈FN₃O₄, 408.40 [M + H]⁺.

Compound 5e, pale yellow solid, 194 mg (36.4% yield); ¹H NMR (400 MHz, CF₃CO₂D): δ 9.27 (s, 1H), 8.32 (d, J = 12.38 Hz, 1H), 8.22 (d, J = 8.07 Hz, 1H), 7.72 (t, J = 7.84 Hz, 1H), 7.52–7.60 (m, 3H), 6.71 (d, J = 6.88 Hz, 1H), 3.44–3.65 (m, 8H); ¹³C NMR (101 MHz, CF₃CO₂D): δ 174.0, 172.2, 157.7 (d, J = 258.83 Hz), 152.1, 150.7 (d, J = 10.30 Hz), 145.8, 142.4, 138.1, 135.3, 130.1, 130.0, 128.5, 127.5, 124.8, 117.8–118.0 (m), 114.2 (d, J = 27.08 Hz), 108.6, 106.6, 48.1 (d, J = 4.77 Hz), 46.6; (ν_max/cm^–1): 1722, 1664, 1492, 1454, 1395, 1352, 1326, 1285, 1251, 1100, 1042, 885, 802, 700, 661, 616, 563, 540; LC–MS retention time 2.75 min (method A), purity = 100%, found, 424.2 [M + H]⁺, calcd for C₂₂H₁₈FN₃O₃S, 424.46 [M + H]⁺.

Compound 5f, pale pink solid, 167 mg (94.4% yield); ¹H NMR (400 MHz, CF₃CO₂D): δ 9.56 (s, 1H), 8.19 (d, J = 12.38 Hz, 1H), 7.63–7.74 (m, 1H), 7.49–7.63 (m, 2H), 7.29 (d, J = 4.22 Hz, 2H), 6.82 (br s, 1H), 6.18 (s, 2H), 3.65–3.77 (m, 4H), 3.54–3.65 (m, 4H); ¹³C NMR (101 MHz, (CD₃)₂SO): δ 176.5, 166.3, 152.8 (d, J = 250.44 Hz), 152.0, 150.0, 146.6, 145.4 (d, J = 10.30 Hz), 137.6, 127.9, 123.7, 123.5, 121.9, 119.0 (d, J = 7.06 Hz), 118.7, 111.3 (d, J = 23.08 Hz), 107.3, 107.2, 106.0 (d, J = 1.91 Hz), 52.7, 50.4 (d, J = 4.77 Hz), 45.1; (ν_max/cm^–1): 1624, 1583, 1488, 1448, 1341, 1321, 1262, 1215, 1173, 1032, 1013, 940, 807, 793, 747, 628; LC–MS retention time 2.78 min (method A), purity = 100%, found, 422.0 [M + H]⁺, calcd for C₂₃H₂₀FN₃O₄, 422.43 [M + H]⁺.

Compound 5g, pale yellow solid, 41 mg (23.4% yield); ¹H NMR (400 MHz, CF₃CO₂D): δ 9.45 (s, 1H), 8.20 (d, J = 12.29 Hz, 1H), 7.91 (d, J = 7.70 Hz, 1H), 7.44–7.52 (m, 2H), 7.41 (t, J = 7.79 Hz, 1H), 7.32 (d, J = 6.60 Hz, 1H), 7.19 (d, J = 7.34 Hz, 1H), 6.13 (s, 2H), 3.53–3.64 (m, 4H), 3.44–3.53 (m, 4H); (ν_max/cm^–1): 1621, 1583, 1488, 1448, 1338, 1279, 1266, 1204, 1182, 1049, 1032, 940, 825, 793, 744, 689, 622, 554, 517; LC–MS retention time 5.53 min (method B), purity = 100%, found, 437.9 [M + H]⁺, calcd for C₂₃H₂₀FN₃O₃S, 438.49 [M + H]⁺. See note C.

General Procedure for Compounds 6a–6g

Compound 5a (150 mg, 0.45 mmol, 1 equiv) was stirred in DCM (7 mL) for 5 min (method A), then 4 M HCl in dioxane (2.26 mL, 9.05 mmol, 20 equiv) was added dropwise and the mixture stirred for 1 h. Upon completion, the mixture was washed with hexane (3 × 1 mL) and lyophilized overnight to give compound 6a, pale brown solid; ¹H NMR (400 MHz, (CD₃)₂SO): δ 15.13 (br s, 1H, H7), 9.46 (br s, 2H, H13), 8.67 (s, 1H, H3), 7.94 (d, J = 13.09 Hz, 1H, H1), 7.61 (d, J = 7.30 Hz, 1H, H2), 3.86 (br s, 1H, H4), 3.53–3.60 (m, 4H, H8 + 9), 3.27–3.36 (m, 4H, H10 + 11), 1.28–1.37 (m, 2H, H5 + 6), 1.15–1.24 (m, 3H, H5+6 + 12); ¹³C NMR (101 MHz, (CD₃)₂SO): δ 176.4 (d, J = 2.93 Hz), 165.9, 152.9 (d, J = 249.42 Hz), 148.2, 144.1 (d, J = 10.27 Hz), 139.1, 119.3 (d, J = 8.07 Hz), 111.2 (d, J = 23.47 Hz), 106.9 (d, J = 2.93 Hz), 106.8, 46.3 (d, J = 5.14 Hz), 42.5, 36.0, 7.6; (ν_max/cm^–1): 1701, 1624, 1491, 1458, 1383, 1341, 1272, 1142, 1106, 1034, 941, 909, 889, 853, 829, 804, 774, 749, 703, 665, 636, 619; LC–MS retention time 1.89 min (method A) and 4.70 min (method B), purity = 97.6% (found, 332.1 [M + H]⁺) and 100% (found, 332.0 [M + H]⁺), respectively, calcd for C₁₇H₁₈FN₃O₃, 332.35 [M + H]⁺; HRMS observed 332.1404 [M + H]⁺, theoretical value 332.1405 [M + H]⁺.

Compound 6b, pale orange solid; ¹H NMR (400 MHz, (CD₃)₂SO): δ 15.31 (br s, 1H, H6), 9.37 (br s, 2H, H12), 8.97 (s, 1H, H3), 7.96 (d, J = 13.09 Hz, 1H, H1), 7.26 (d, J = 7.30 Hz, 1H, H2), 4.62 (q, J = 7.13 Hz, 2H, H4), 3.52–3.59 (m, 4H, H7 + 8), 3.25–3.33 (m, 4H, H9 + 10), 1.41 (t, J = 7.18 Hz, 3H, H5), 1.22 (d, J = 6.55 Hz, 1H, H11); ¹³C NMR (101 MHz, (CD₃)₂SO): δ 176.3 (d, J = 3.05 Hz), 166.3, 152.9 (d, J = 249.48 Hz), 148.9, 144.6 (d, J = 10.68 Hz), 137.3, 120.1 (d, J = 7.63 Hz), 111.6 (d, J = 23.65 Hz), 107.3, 106.7 (d, J = 3.43 Hz), 49.3, 46.6 (d, J = 4.58 Hz), 42.8, 14.6; (ν_max/cm^–1): 1701, 1696, 1626, 1507, 1454, 1345, 1340, 1332, 1273, 1130, 1053, 1033, 933, 899, 859, 829, 804, 746, 665; LC–MS retention time 1.89 min (method A) and 3.29 min (method B), purity = 100% (found, 320.1 [M + H]⁺) and 98.5% (found, 320.1 [M + H]⁺), respectively, calcd for C₁₆H₁₈FN₃O₃, 320.34 [M + H]⁺; HRMS observed 320.1404 [M + H]⁺, theoretical value 320.1405 [M + H]⁺.

Compound 6c, white solid; ¹H NMR (400 MHz, (CD₃)₂SO): δ 15.14 (br s, 1H), 9.57 (br s, 1H), 9.19 (s, 1H), 7.92 (d, J = 12.93 Hz, 1H), 7.06–7.48 (m, 6H), 5.90 (s, 2H), 3.41 (br s, 4H), 3.22 (br s, 4H); ¹³C NMR (101 MHz, (CD₃)₂SO): δ 176.4 (d, J = 1.53 Hz), 166.0, 152.6 (d, J = 250.05 Hz), 149.6, 143.9 (d, J = 10.11 Hz), 137.5, 135.3, 129.0, 128.2, 127.0, 120.1 (d, J = 7.44 Hz), 111.4 (d, J = 22.89 Hz), 107.3, 107.2 (d, J = 2.86 Hz), 56.5, 46.2 (d, J = 4.20 Hz), 42.3; (ν_max/cm^–1): 1718, 1627, 1507, 1453, 1383, 1367, 1268, 1254, 1207, 1059, 1033, 1007, 960, 927, 903, 833, 802, 763, 739, 698, 535, 524; LC–MS retention time 2.17 min (method A) and 4.24 min (method B), purity = 98.9% (found, 382.1 [M + H]⁺) and 98.8% (found, 382.1 [M + H]⁺), respectively, calcd for C₂₁H₂₀FN₃O₃, 382.41 [M + H]⁺; HRMS observed 382.1559 [M + H]⁺, theoretical value 382.1561 [M + H]⁺.

Compound 6d, pale yellow solid; ¹H NMR (400 MHz, (CD₃)₂SO): δ 14.91 (br s, 1H), 9.31 (br s, 2H), 8.87 (s, 1H), 8.05–8.10 (m, 2H), 8.01 (dd, J = 7.81, 1.01 Hz, 1H), 7.74 (dd, J = 7.81, 1.01 Hz, 1H), 7.56 (t, J = 7.81 Hz, 1H), 7.22 (d, J = 2.27 Hz, 1H), 6.37 (d, J = 7.30 Hz, 1H), 3.09–3.24 (m, 8H); ¹³C NMR (101 MHz, (CD₃)₂SO): δ 176.8 (d, J = 2.20 Hz), 165.4, 153.0 (d, J = 250.89 Hz), 149.4, 148.3, 147.5, 144.6 (d, J = 11.00 Hz), 138.4, 129.9, 124.2, 123.5, 123.3, 119.4 (d, J = 8.07 Hz), 111.5 (d, J = 23.48 Hz), 108.2, 107.7, 106.5 (d, J = 2.93 Hz), 45.9 (d, J = 4.40 Hz), 42.3; (ν_max/cm^–1): 1724, 1627, 1505, 1451, 1380, 1331, 1271, 1201, 1172, 1116, 1033, 1017, 908, 871, 804, 734, 625, 559, 549; LC–MS retention time 2.23 min (method A) and 5.17 min (method B), purity = 100% (both), found, 408.1 [M + H]⁺ (both), calcd for C₂₂H₁₈FN₃O₄, 408.40 [M + H]⁺; HRMS observed 408.1352 [M + H]⁺, theoretical value 408.1354 [M + H]⁺.

Compound 6e, pale orange solid; ¹H NMR (400 MHz, (CD₃)₂SO): δ 9.47 (br s, 2H), 8.83 (s, 1H), 8.23 (d, J = 7.89 Hz, 1H), 8.07 (d, J = 12.93 Hz, 1H), 7.89 (d, J = 5.41 Hz, 1H), 7.79–7.84 (m, 1H), 7.68–7.76 (m, 2H), 6.32 (d, J = 7.15 Hz, 1H), 3.04–3.26 (m, 8H); ¹³C NMR (101 MHz, (CD₃)₂SO): δ 176.8 (d, J = 2.48 Hz), 165.3, 153.0 (d, J = 250.05 Hz), 148.5, 144.7 (d, J = 10.68 Hz), 142.1, 137.8, 136.3, 133.7, 128.8, 126.1, 126.1, 125.2, 123.9, 119.5 (d, J = 7.63 Hz), 111.7 (d, J = 23.08 Hz), 108.4, 106.2 (d, J = 2.67 Hz), 45.9 (d, J = 4.77 Hz), 42.2; (ν_max/cm^–1): 1722, 1666, 1627, 1499, 1439, 1396, 1354, 1268, 1181, 1032, 951, 901, 866, 832, 799, 713, 684, 615, 555, 537; LC–MS retention time 2.33 min (method A) and 5.27 min (method B), purity = 100% (both), found, 424.1 [M + H]⁺ (both), calcd for C₂₂H₁₈FN₃O₃S, 424.46 [M + H]⁺; HRMS observed 424.1124 [M + H]⁺, theoretical value 424.1126 [M + H]⁺.

Compound 6f, pale orange solid; ¹H NMR (400 MHz, (CD₃)₂SO): δ 15.11 (br s, 1H), 9.62 (br s, 2H), 9.33 (s, 1H), 8.10 (d, J = 2.11 Hz, 1H), 7.90 (d, J = 13.11 Hz, 1H), 7.65 (d, J = 7.52 Hz, 1H), 7.38 (d, J = 7.24 Hz, 1H), 7.34 (d, J = 7.24 Hz, 1H), 7.28 (t, J = 7.57 Hz, 1H), 7.00 (d, J = 2.20 Hz, 1H), 6.17 (s, 2H), 3.32–3.43 (m, 4H), 3.22 (br s, 4H); ¹³C NMR (101 MHz, (CD₃)₂SO): δ 176.4 (d, J = 2.48 Hz), 165.9, 152.6 (d, J = 249.67 Hz), 152.0, 150.0, 146.5, 143.9 (d, J = 10.30 Hz), 137.4, 127.8, 123.9, 123.4, 121.9, 119.9 (d, J = 7.63 Hz), 118.5, 111.4 (d, J = 22.89 Hz), 107.2, 107.1, 106.8 (d, J = 3.05 Hz), 52.5, 46.2 (d, J = 4.77 Hz), 42.2; (ν_max/cm^–1): 1712, 1624, 1546, 1496, 1448, 1387, 1268, 1209, 1179, 1032, 1015, 932, 801, 756, 741, 557, 525, 510; LC–MS retention time 2.33 min (method A) and 4.70 min (method B), purity = 100% (both), found, 422.1 [M + H]⁺ (both), calcd for C₂₃H₂₀FN₃O₄, 422.43 [M + H]⁺; HRMS observed 422.1507 [M + H]⁺, theoretical value 422.1511 [M + H]⁺.

Compound 6g, red solid; ¹H NMR (400 MHz, (CD₃)₂SO): δ 15.11 (br s, 1H), 9.31 (s, 1H), 9.24 (br s, 2H), 7.95 (d, J = 13.11 Hz, 1H), 7.91 (d, J = 7.15 Hz, 1H), 7.78 (d, J = 5.41 Hz, 1H), 7.53 (d, J = 5.50 Hz, 1H), 7.36–7.47 (m, 2H), 7.15 (d, J = 7.34 Hz, 1H), 6.16 (s, 2H), 3.24–3.30 (m, 4H), 3.16 (br s, 4H); ¹³C NMR (101 MHz, (CD₃)₂SO): δ 176.6 (d, J = 2.48 Hz), 166.0, 152.7 (d, J = 250.05 Hz), 150.4, 143.8 (d, J = 10.30 Hz), 140.8, 137.8, 136.8, 129.2, 127.7, 125.1, 124.6, 124.2, 123.7, 119.9 (d, J = 7.82 Hz), 111.6 (d, J = 23.08 Hz), 107.4, 106.9 (d, J = 1.53 Hz), 56.7, 46.2 (d, J = 49.6 Hz), 42.4; (ν_max/cm^–1): 1707, 1617, 1507, 1395, 1300, 1268, 1206, 1107, 1056, 1033, 940, 832, 802, 771, 725, 699, 620, 554, 516; LC–MS retention time 2.40 min (method A) and 4.81 min (method B), purity = 100% (both), found, 438.1 [M + H]⁺ (both), calcd for C₂₃H₂₀FN₃O₃S, 438.49 [M + H]⁺; HRMS observed 438.1278 [M + H]⁺, theoretical value 438.1282 [M + H]⁺.

Notes

A

Compounds 2a–g display complex and unusual splitting patterns and integrals in their associated NMR spectra. Because fluorine-decoupled carbon-13 NMR spectra could not be generated, the highly fluorinated nature of the aromatic rings in these compounds causes splitting of the signals of nearby carbons. Additionally, compounds 2a–g can all form two distinct structural isomers because of their common 3-aminoacrylate moiety. These pairs of isomers exist in equilibrium with an interconversion rate significantly less than the difference in frequency between the isomers (“slow on the NMR timescale”) meaning that both isomers are fully resolved in ¹H and ¹³C NMR spectra.

B

Continuing from note A; the reduction in signal-to-noise ratio caused by the high degree of splitting of aromatic carbon signals, in combination with the twinning of each peak, precluded resolution of many of these peaks for some compounds. In some cases, several small baseline multiplets are visible in the correct chemical shift range and are most likely the aforementioned missing signals.

C

Poor compound solubility in all deuterated solvents tested precluded analysis via carbon-13 NMR.

Glossary

Abbreviations

ATCC

American Type Culture Collection

DNA

deoxyribonucleic acid

HRMS

high-resolution mass spectrometry

IR

infrared spectroscopy

LC–MS

liquid chromatography–mass spectrometry

MIC

minimum inhibitory concentration

NCTC

National Collection of Type Cultures

NMR

nuclear magnetic resonance

PAβN

phenylalanine–arginine β-naphthylamide

PMBN

polymyxin B nonapeptide

TLC

thin layer chromatography

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.9b03910.

• Characterization data for new N1-benzofused fluoroquinolones, tabular data for modeled N1-benzofused fluoroquinolone-DNA gyrase interactions, and top binding poses for compounds 6a, 6e and 6g docked against S. aureus DNA gyrase (PDF)

Author Present Address

^§ Department of Pharmaceutical Sciences, University of Padua, Via Francesco Marzolo 5, 35131 Padua, Italy.

Author Present Address

^∥ Cancer Research UK Manchester Institute, The University of Manchester, Alderley Park, SK10 4TG, UK.

Author Contributions

M.L. and C.H. contributed equally to the work. M.L. wrote the manuscript with input from all the authors; M.L. and A.F. undertook the synthesis, purification, and characterization of compounds; C.H., B.E., and M.C. carried out the microbiological evaluation of the compounds; S.J. completed the computational modeling work; K.M.R. and J.M.S. conceived of, supervised and oversaw completion of the project.

The authors declare no competing financial interest.