Fungal naphtho-γ-pyrones: Potent antibiotics for drug-resistant microbial pathogens

Four naphtho-γ-pyrones (fonsecinones A and C and aurasperones A and E) were identified as potential antibacterial agents against Escherichia coli, extended-spectrum β-lactamase (ESBL)-producing E. coli, Pseudomonas aeruginosa, Enterococcus faecalis, and methicillin-resistant Staphylococcus aureus (MRSA) in an in vitro antibacterial screen of 218 fungal metabolites. Fonsecinone A (2) exhibited the most potent antibacterial activity, with minimum inhibitory concentrations (MICs) of 4.26, 17.04, and 4.26 μg/mL against ESBL-producing E. coli, P. aeruginosa, and E. faecalis, respectively. The inhibitory effects of fonsecinones A (2) and C (3) against E. coli and ESBL-producing E. coli were comparable to those of amikacin. Molecular docking-based target identification of naphtho-γ-pyrones 1–8 revealed bacterial enoyl-acyl carrier protein reductase (FabI) as an antibacterial target, which was further validated by FabI affinity and inhibition assays. Fonsecinones A (2) and C (3) and aurasperones A (6) and E (7) bound FabI specifically and produced concentration-dependent inhibition effects. This work is the first report of anti-drug-resistant bacterial activities of naphtho-γ-pyrones 1–8 and their possible antibacterial mechanism of action and provides an example of the successful application of in silico methods for drug target identification and validation and the identification of new lead antibiotic compounds against drug-resistant pathogens.

The emergence and spread of antibiotic-resistant bacteria have potentially drastic consequences for human health globally 1 . Methicillin-resistant Staphylococcus aureus (MRSA), Enterococcus faecalis (E. faecalis), extended-spectrum β -lactamase-producing Escherichia coli (ESBL-producing E. coli), Klebsiella pneumoniae, and Pseudomonas aeruginosa express a remarkable array of resistance and virulence factors, which contributes to their prominent roles in hospital-acquired infections 2 . New antibiotics or for further strategies to overcome antibiotic resistance are thus urgently needed. Microorganisms produce diverse antibiotics that function in an antagonistic capacity in nature in competition with other organisms 3 , and most antibacterial agents in clinical or preclinical trials are either microbial products or analogs [3][4][5] . Another effective approach is drug repurposing, in which new useful activities of old drugs are identified by screening against relevant disease targets 6 . Aspergillomarasmine A, a fungal metabolite discovered 50 years ago, was recently repurposed based on its ability to overcome antibiotic resistance associated with metallo-β-lactamase 7 . The development of bioinformatic and computational biology techniques will greatly facilitate, this type of drug repurposing [8][9][10] .
Most of the investigated naphtho-γ-pyrones displayed significant antibacterial activity against gram-negative bacteria. Fonsecinones A (2) and C (3) and aurasperones A (6) and E (7) possessed potential antibacterial activities against E. coli, ESBL-producing E. coli, P. aeruginosa, E. faecalis and MRSA in the micromolar range. Fonsecinone A (2) exhibited the most potent antibacterial activity. The MICs of Fonsecinone A for ESBL-producing E. coli, P. aeruginosa and E. faecalis were 4.26, 17.04, and 4.26 ug/mL, respectively. The inhibitory effects of fonsecinones A (2) and C (3) against E. coli and ESBL-producing E. coli were comparable to those of amikacin 24 , which exhibits MICs of 2.13 and 4.26 μg/mL, respectively. The structure-activity relationships in these compounds indicated that the compounds with C-10′ -C-9 linkages, fonsecinones A and C, had the highest antibacterial activities against the targeted pathogenic bacteria, followed by the compounds with C-10′ -C-7 linkages, fonsecinone B and aurasperones A and E. Asperpyrone C, which features a C-6-C-7′ linkage, had the lowest antibacterial activities. The removal of electron-donating substituents (OH) at C2/C2′ was associated with higher antibacterial activity for these naphtho-γ-pyrones. Flavasperone, a monomer with an angular skeleton, exhibited higher antibacterial activity than its linear isomer (5).

Identification of FabI as a possible antibacterial target by inverse docking. Compounds 1-8,
which exhibited potent antimicrobial activity, were subjected to further investigation to determine their possible antimicrobial mechanisms. A thorough literature survey indicated that several flavones 25 and flavonoids 26,27 with similar polycyclic skeleton structures and cephalochromin 28 , a bis-naphtho-γ-pyrone with a chaetochromin-skeleton, are inhibitors of cellular fatty acid biosynthesis in bacteria. Bacterial fatty acid biosynthesis enzymes are excellent targets for novel antibacterial agent discovery because of their essential roles in the synthesis of the cell membrane 29 . Several antibiotics, such as diazaborines, isoniazid, and triclosan, inhibit fatty acid synthetic enzymes 29 . Selective targeting of key bacterial fatty acid metabolism enzymes was performed via in silico target identification by small-scale inverse docking of 14 enzymes involved in fatty acid metabolism 29,30 ( Fig. S25) with compounds 1-8. Targets with lower calculated binding energies are considered to have higher binding affinities for a specific compound 31,32 . The docking score results are presented in Table 2. FabI 29 , a single enoyl-ACP reductase enzyme that is the key control point within the bacterial fatty acid elongation cycle, was predicted to exhibit significantly higher binding affinities for the investigated naphtho-γ-pyrones. As the negative control, four unrelated targets [33][34][35][36] were also performed docking with compounds 1-8, the results were shown in Table S1, no significant binding was predicted for these targets.
FabI is highly conserved among most bacteria, including E. coli, S. aureus, E. faecalis and P. aeruginosa, and has no homologue in mammals 30,37 . FabI is therefore an attractive target for antibacterial drug development 29,37 . Notably, fonsecinones A and C, which possessed the most potent antibacterial activities, had the lowest binding   Bioactive naphtho-γ-pyrones maintain specific binding with FabI. The ligand-protein networks suggested that the bioactive naphtho-γ-pyrones could bind FabI. Microscale thermophoresis (MST) was then performed to quantitatively measure binding 40 . The overall structures of FabI in E. coli and other pathogens have been reported to be very similar 29 ; therefore, we expressed and purified E. coli FabI to quantify the dissociation constant (K d ) for compound binding to the protein. Fonsecinones A (2) and C (3) and aurasperones A (6) and E (7) had K d values of 215 ± 28.8, 270 ± 9.1, 289 ± 31.1, and 329 ± 29.7 μM, respectively (Table 3), confirming their specific binding to E. coli FabI (Fig. 3).

Bioactive naphtho-γ-pyrones inhibit FabI in a concentration-dependent manner.
To further characterize their activity against FabI, compounds 1-8 were assayed. Six naphtho-γ-pyrones (1-4, 6, and 7) inhibited E. coli FabI; their IC 50 values are presented in Table 3. Fonsecinone A (2) exhibited the most potent inhibition, with an IC 50 of only 2.78 μM. The order of inhibitory potencies was 2 > 3 > 6 > 7 > 1 > (≈ ) 4. The antibacterial activities of these compounds generally agreed with the experimentally determined FabI inhibitory activities. Among these compounds, fonsecinones A and C, had moderate antibacterial activities against E. coli and S. aureus, which express FabI as the sole enoyl reductase in the FASII system; however, no obvious inhibition of K. pneumonia, which produces only FabK, was observed 29,38 . The inhibitory activity of fonsecinone A against E. coli FabI was comparable to that of cephalochromin 28 , possibly indicating that the abilities of fonsecinone A and cephalochromin to access the active site of E. coli FabI and exert selective inhibitory effects are comparable, as observed in the molecular docking experiment (Figs 2 and S26). Fonsecinone A and cephalochromin have similar modes of binding in the active site of FabI and a potent π -π interaction with the nicotine ring of NAD + 38, 39 . The results of the FabI inhibition assay are in agreement with the predicted binding free energies (Table 2), further validating the success of the inverse-docking target prediction and confirming FabI as an antibacterial target of the investigated naphtho-γ-pyrones.
In conclusion, a combination of in vitro antibacterial screening and molecular docking techniques were employed to identify several bioactive naphtho-γ-pyrones (2, 3, 6, and 7) as potent antibiotics against a panel of drug-resistant microbial pathogens. These compounds exhibit FabI inhibition as one antibacterial mode of   15,17 . The active naphtho-γ-pyrones found in this study may provide novel chemical scaffolds for the discovery of antibacterial agents.

Materials and Methods
General experimental procedures. UV spectra were measured on a Shimadzu UV-2401A spectrophotometer. IR spectra were determined on a Bruker Vertex 70 FT-IR spectrophotometer. ECD spectra were obtained with a JASCO J-810 spectrometer. HRESIMS was performed on an APIQSTAR Pulsar spectrometer mass spectrometer. NMR spectra were recorded on a Bruker AM-400 NMR spectrometer, and chemical shifts were referenced to the solvent peaks for CDCl 3   Computational methods for ECD spectra. The conformational analyses were carried out for compounds 3, 4 and 7 using BALLOON11 and confab12 programs. The BALLOON program searches conformational space with a generic algorithm, whereas the confab program systematically generates diverse low energy conformations that are supposed to be close to crystal structures. Conformations generated by both programs were grouped together by removing duplicated conformations in which the root mean square (RMS) distance was less than 0.5 Å. Semi-empirical PM3 quantum mechanical geometry optimisations were performed on the conformations using the Gaussian 0913 program. The duplicated conformations after geometry optimisation were then identified and removed. Remaining conformations were further optimised at the B3LYP/6-31G* level of theory in methanol solvent with the IEFPCM314 solvation model using the Gaussian 09 program, and duplicated conformations emerging after these calculations were removed according to the same RMS criteria indicated above. The harmonic vibrational frequencies were performed to confirm the stability of obtained conformers (Figures S106 and S107). Oscillator strengths and rotational strengths of the 20 weakest electronic excitations of each conformer were calculated using the TDDFT methodology at the B3LYP/6-311+ + G** level of theory with methanol as the solvent by the IEFPCM solvation model implemented in the Gaussian 09 program. The ECD spectra for each conformers were then simulated using Gaussian function with a bandwidth σ = 0.45 eV. The calculated spectra for each conformations were combined after Boltzmann weighting according to their population contributions.
Antimicrobial activity. Whole-cell antimicrobial activity was determined by the broth microdilution 41 . Test strains were grown to the mid-log phase in Mueller-Hinton broth and diluted 1000-fold in the same medium. Cells (10 5 /mL) were inoculated into Mueller-Hinton broth and dispensed at 0.1 mL/well into a 96-well microtitre Scientific RepoRts | 6:24291 | DOI: 10.1038/srep24291 plate. MICs were determined in triplicate by serial 2-fold dilutions of test compounds. The MIC was defined as the concentration of a test compound that completely inhibited cell growth during a 24h incubation at 30 °C. Bacterial growth was determined by measuring the absorption at 600 nm using a microtitre ELISA reader.
FabI cloning, expression, and purification. The full-length FabI gene (E. coli str. K-12 substr. MG1655, complete genome) was amplified from E. coli genomic DNA using the following forward and reverse primers: 5′-GGAATTCATATGGGTTTTCTTTCCGGTAAGCGC-3′ and 5′ -GGGTGCCTCGAGTTATTTCAGTTCGAGTTCGTT-3′ . The gene was cloned into a modified pET21b vector containing a 6 His-tag coding region at the N-terminus of the insert. After verifying the recombinant plasmids by sequencing, the plasmids were used to transform E. coli BL21 (DE3) cells. The transformed cells were grown in LB medium at 37 °C to an OD of 0.8-1.0 and induced with 0.5 mM isopropyl-D-thiogalactopyranoside (IPTG) at 18 °C for 18 h. The cells were harvested by centrifugation at 4,100 rpm for 10 min and re-suspended in lysis buffer containing 20 mM Tris, pH 8.5, 200 mM NaCl, and 10 mM imidazole, followed by disruption on a French press. Cell debris was removed by centrifugation at 21,000 rpm for 30 min. The protein was bound to Ni-agarose affinity resin, washed with buffer containing 20 mM Tris, pH 8.5, 200 mM NaCl, and 10 mM imidazole, and eluted with buffer containing 20 mM Tris, pH 8.8, 250 mM NaCl, and 150 mM imidazole. The protein was further purified by anion-exchange chromatography using a linear gradient of 10 mM to 1 M NaCl and size exclusion chromatography at pH 8.5 in 200 mM NaCl.

Microscale thermophoresis. Recombinant E. coli FabI was labeled with the Monolith NT ™ Protein
Labeling Kit RED (Cat#L001) according to the supplied labeling protocol. Labelled FabI was kept constant at 50 nM, while all samples tested were diluted in a 20 mM HEPES (pH 7.4) and 0.05 (v/v)% Tween-20. After 10 min incubation at room temperature, samples were loaded into Monolith TM standard-treated capillaries and the thermophoresis was measured at 25 °C after 30 min incubation on a Monolith NT.115 instrument (NanoTemper Technologies, München, Germany). Laser power was set to 20% or 40% using 30 seconds on-time. The LED power was set to 100%. The dissociation constrant Kd values was fitted by using the NTAnalysis software (NanoTemper Technologies, München, Germany). IC 50 assay of compounds with E. coli FabI. NADH, NADPH, crotonoyl-CoA, and HEPES reagents were obtained from Sigma. All other chemicals were of analytical grade. The assay was performed in half 96-well microtiter plates. All compounds in this study were dissolved in DMSO to produce 100 mM stocks. The IC 50 values of the investigated compounds were determined using 1 μg of enzyme, 50 μM crotonoyl-CoA as the substrate, and 100 μM NADH as a cofactor 30,40 . Nine concentrations of the compounds were analyzed using a log-probit analysis program in GraphPad Prism 4.0. Equal volumes of DMSO and triclosan were used in the untreated and positive controls.

Molecular Docking.
For the small scale inverse docking 31,32 , crystal structures of docking targets (Table 2) were obtained from the Protein Data Bank (http://www.rcsb.org) 38,39,[42][43][44][45][46][47][48][49][50][51] . The docking was performed by using ICM 3.8.2 modeling software on an Intel i7 4960 processor (MolSoft LLC, San Diego, CA) 52 . Ligand binding pocket residues were selected by using graphical tools in the ICM software, to create the boundaries of the docking search. In the docking calculation, potential energy maps of the receptor were calculated using default parameters. Compounds were imported into ICM and an index file was created. Conformational sampling was based on the Monte Carlo procedure 53 , and finally the lowest-energy and the most favorable orientation of the ligand were selected. Statistical analysis. Statistical analysis of the data was performed using Graph Pad Prism 4.0 software. The data were expressed as the means ± SD. Values were analyzed using SPSS version 12.0 software by one-way analysis of variance (ANOVA), and p < 0.05 was considered statistically significant.