Introduction

Life-threatening bacterial infections are a major concern in health care settings worldwide. Because of increase in the emergence of antibiotic resistant bacteria, medical practitioners must look to rarely used antibiotics for alternative treatment options1,2. For instance, bacteria such as Acinetobacter baumannii cause nosocomial infections in immunocompromised patients, which, until recently, could be effectively treated with carbapenems3. The recent emergence of carbapenem-resistant Gram-negative pathogenic bacteria, such as A. baumannii, Escherichia coli, Pseudomonas aeruginosa and Klebsiella pneumonia, has led to the resurgence of polymyxins as last-resort antibiotic treatments3,4,5. Multidrug resistant (MDR) A. baumannii infections are associated with intensive care units and are implicated in ventilator associated pneumonia and urinary tract infections4. In the recent times MDR A. baumannii are reported to exhibit a variety of resistance mechanisms to most classes of antibiotics which demands immediate development of new therapies6.

Polymyxins are cataionic lipopeptide antibiotics produced by non-ribosomal peptide synthetase enzyme complexes and are bactericidal agents with a detergent-like mechanism of action7. Colistin (polymixin E) and polymixin B are the two important forms of polymixnis with clinical significance8. They have largely been used in topical form as otic and ophthalmic solutions and are also used in combination with numerous other antibiotics, such as carbapenems, efflux pump inhibitors, β-lactamase inhibitors and wall teichoic acid synthesis inhibitors7,8,9,10. However, the use of polymyxins is limited by their nephrotoxic and neurotoxic effects and hence administered mostly to prevent rather than treat infections11,12. Furthermore, the pharmacokinetics and pharmacodynamics of polymyxin B remain poorly understood. In addition, there has been a surge in reports of infections caused by naturally-occurring, polymyxin-resistant bacteria, including Providencia spp., Proteus spp., Morganella spp. and Serretia spp.4 There is thus urgency to optimise the clinical use of polymixins by designing effective combination therapies11,12.

In the present study, we conducted a large-scale screening of secondary metabolites isolated from Streptomyces spp. and assessed their synergistic potential when combined with polymyxin B. The most effective secondary metabolite was identified as netropsin. Combination therapy consisting of polymyxin B and netropsin showed synergistic effects when tested on Gram-negative bacteria in vitro, including clinical isolates of colistin (polymyxin E)-resistant A. baumannii. The effectiveness of the combination therapy was verified in vivo using G. mellonella as a model host organism. Netropsin is a pyrrole-amide antitumor drug that binds preferentially to AT-rich regions in the minor groove of DNA and increases positive supercoiling13,14. Thus, the molecular specificity of netropsin makes it a versatile molecule for use not only against bacteria, but also potentially against viruses and tumors15.

Results

Development of novel screening system to identify antibiotics enhancing agents

Certain antibiotics such as polymixins are poor diffusing bactericidal agents on agar medium and hence difficult to obtain consistent results using traditional disc diffusion tests. Here we developed liquid media-based screening technique to screen for novel compounds that show synergetic effects with polymixin B. The strategy of mining antimicrobial agents from secondary metabolites library, their subsequent identification and evaluation both in vitro and in vivo is detailed in a step-by-step procedure (Fig. 1). a. Isolation of Streptomyces spp. from different soil environment. b. Construction of 1800 metabolite library from each isolated Streptomyces sp. c. Mass screening of metabolites with polmyxin B to test synergetic effects on E. coli. d. Identification of synergetic compound and bacteria producing the compound. e. In vitro checkerboard assay. f. Time kill assay using the greater wax moth infection model.

Figure 1
figure 1

Large-scale screening strategy of secondary metabolites for mining agents that enhance the action of polymyxin B.

A to F. Schematic representation of each experimental procedure to identify and characterize compounds that show synergism with polymyxin B. (a) Isolation of various Streptomyces spp. from different soil environmental samples. (b) Isolation of secondary metabolites from individual Streptomyces spp. (c) Mass screening of metabolites with serial concentrations of polmyxin B (250, 25, 2.5 and 0 mg/L), as indicated by violet color shades. Tetrazolium dye (0.5%) is included along with media, to measure the E. coli susceptibility to combined action of secondary metabolite and polymixin B. (d) Identification of synergetic compounds that originated from selected metabolite using 16 s rDNA sequencing and analytical chemistry techniques to identify compound within the metabolite responsible for synergism. (e) Checkerboard assay to test synergistic activity of compounds present in secondary metabolites against various Gram-negative pathogenic bacteria. (f) In vivo time kill assay using the greater wax moth to verify the effectiveness of combination therapy. Bacterial strains were infected to G. mellonella larvae followed by administration of polymixin B or netropsin or both.

Full size image

Identification of compounds that enhance the antibiotic activity of polymixin B

To identify substances capable of enhancing the antibacterial action of polymyxin B, a strategy using a massive library of 1800 secondary metabolites isolated from equal number of Streptomyces spp. culture samples was established (Fig. 1a). A collection of 1800 strains of Streptomyces spp. were isolated from various soil samples that were obtained from forest, land, compost and river water in S. Korea. Humic-vitamin medium was used for selective isolation of Streptomyces spp., which was then maintained as pure culture glycerol stocks. Glucose-Starch-Soybean (GSS) meal medium was used for activating secretory secondary metabolites produced from corresponding Streptomyces spp. culture stocks (Fig. 1b). Combinatorial effect of individual secondary metabolites with different concentrations of polymixin B was tested by measuring physiological change of E. coli oxidative metabolism as indicated by 0.5% tetrazolium dye (indicating oxygen consumption) (Fig. 1c). Of the extracts tested, an extract AN130070 showed the greatest synergistic effect at a minimum inhibitory concentration (MIC) of 0.25 mg/L of polymyxin B (Fig. S1). A checkerboard assay using serially diluted AN130070 metabolite with 0.25 mg/L polymixin B showed maximum synergy at zero dilution of AN130070 (Fig. S1). To identify the bacterial source of AN130070, genomic DNA of the corresponding stock was isolated and subjected to 16s rDNA sequencing. NCBI BLAST search of the sequence revealed that AN130070 metabolite was produced by S. netropsis SN01 (Fig. S2)16,17.

Analysis of AN130070 by HPLC and MS

To identify the compounds responsible for the antibacterial activity of AN130070, a series of chemical fractionations were conducted. After each round, fractions were tested for synergy with polymyxin B to enhance bioactive fractions. Eight of 27 fractions (RB625B-12, 13, 19~24) were initially selected through C18HPLC to show synergism with polymyxin B. Compound 1 (RB625B-12A), obtained from RB625B-12 sub-fractions of AN130070 that is taken from the butanol layer exhibited synergistic effects with polymyxin B (Fig. S3). The Compound 1 was further analysed by HPLC to determine their chemical identity. UV absorption peak of the Compound 1 showed similar retention time compared with that of reference (commercial netropsin) (Fig. 2a). The 1H NMR inspection confirmed chemical structure of the Compound 1 compared with 1H NMR data of netropsin (Fig. S4)18. MS spectrum analysis of the Compound 1 from AN130070 extracted from S. netropsis SN01 showed two major peaks at m/z 431.2264 and 431.2291 and was consistent with the known spectrum for netropsin (Fig. 2b)17,18.

Figure 2
figure 2

Identification of Netropsin.

(a) HPLC spectrum of Compound 1 from AN130070 metabolite (left) and netropsin (right). Inner panel presented UV absorption of each peaks originated from subfraction of AN130070. RB625B12;Varian C18 column; 5~30% MeCN for 20 min, ~60% MeCN for 30 min, ~100% MeCN for 5 min, wash for 10 min, flow rate 1 mL/min, detection by absorption at 297 nm (b) Mass spectrum (positive ESI mode) of Compound 1 (Hao et al. 2013).

Full size image

Checkerboard assay for synergistic activity against Gram-negative bacteria

The MIC values of polymyxin B against several Gram-negative bacteria were lowered by adding netropsin, which indicated synergy. The bacteria tested included E. coli MG1655 and A. baumannii ATCC15150 that are a commensal bacterium and a human pathogen respectively. The fractional inhibitory concentration index (FICI) was then calculated to estimate the strength of the synergism. The FICI was 0.375 for E. coli MG1655 and 0.25 for A. baumannii ATCC15150, indicating strong synergism between netropsin and polymyxin B (Table 1). To test the practical application of a combination therapy using polymixin B and netropsin, a checkerboard assay was performed against clinical isolates of A. baumannii YCSAb5 and YCSAb7 that were obtained from Yonsei Severance Hospital, Seoul, S. Korea. For both isolates of A. baumannii YCSAb5 and YCSAb7, the FICI value was 0.375, showing synergism. In total 6 of 7 strains of A. baumannii showed FICI < 1.

Table 1 In vitro MIC and FICI values for treatments against Gram-negative bacterial strains.
Full size table

Colistin (also referred to as polymyxin E) and polymyxin B are the same class peptide antibiotics and share similar structures and physicochemical properties. To test the combination of polymyxin B and netropsin against colistin-resistant, clinical isolates of A. baumannii, a checkerboard assay was carried out on five isolates. On colistin-resistant A. baumannii YCRAb257, the synergism was strong, with a FICI value of 0.266. Results with other isolates YCRAb 49, 269, 301 and 615 also indicated partial synergism of the polymyxin B and netropsin combination (Table 1).

The checkerboard assay was conducted for other Gram-negative bacteria- S. flexneri SF2a, S. typhimurium LT2 and P. aeruginosa PA14 and its clinical isolate YPa2. For S. flexneri SF2a and S. typhimurium LT2, there was partial synergism, with FICI values of 0.75 and 0.625, respectively (Table 1). When compared with the MICs of SF2a using polymyxin B alone and netropsin alone, the combination of netropsin with polymyxin B yielded 8- and 2-fold decrease in MICs, respectively. The FICI values for PA14 and YPa2 were 1 and 1.003, which indicated lack of synergism.

Combination therapy using the greater wax moth, an alternative animal model

We verified the results obtained above by performing a survival assay in G. mellonella infected with A. baumannii ATCC1515019. Polymyxin B and netropsin treatment concentrations were calculated from the MICs determined by in vitro assays. Concentrations of 0.5 mg/L of polymyxin B and 100 mg/L of netropsin were identified as the lowest active concentrations based on MIC values under in vitro conditions. The survival of G. mellonella was higher after 24 hours under treatment with the combination therapy (90%) than it was when polymyxin B (66.7%) or netropsin (76.67%) were administered individually. In addition, the survival of G. mellonella treated with the combination therapy was higher than that of untreated insects (46.7%; Fig. 3a).

Figure 3
figure 3

Synergistic effects of polymyxin B combination with netropsin against A. baumannni and its clinical isolates.

The survival of G. mellonella infected with A. baumannii, its clinical isolates (YCSAb5 and YCSAb7) and colistin resistant clinical isolates (YCRAb49, 257, 269, 301, 615). (a) A. baumanii ATCC 15150, (b) YCSAb2, (c) YCSAb7, (d) YCRAb 49, (e) YCRAb 257, (f) YCRAb 269, (g) YCRAb 301 and (h) YCRAb 615. Treatments were: polymyxin B 4 mg/L (▲), netropsin 12.5 mg/L(△), polymyxin B 4 mg/L + netropsin 12.5 mg/L (■), control (○) and mock (●). Results represent means of three independent determinations ± standard deviations for 20 insects per treatment. An asterisk (*) indicates a significant difference between combination and monotherapy (p < 0.05).

Full size image

Two clinical A. baumannii isolates YCSAb5 and YCSAb7 were tested in vivo for efficacy of combination therapy (Fig. 3b,c). The combination treatment exhibited potent synergism (YCSAb5, 40% and YCSAb7, 36.7%) as compared to polymyxin B (YCSAb5, 10% and YCSAb7, 26.7%) or netropsin (YCSAb5, 10% and YCSAb7, 10%) monotherapy 24 hours after treatment (Fig. 3b,c).

The combination therapy was also tested against five clinical isolates of colistin-resistant A. baumannii. We used the concentrations of single or double antibiotics to 4 mg/L of polymyxin B and 12.5 mg/L of netropsin based on Table 1. A. baumannii isolates YCRAb49, YCRAb257 and YCRAb615 showed relatively higher survival after 12 hours when treated with the combination therapy (YCRAb49, 60%; YCRAb257, 53.3%; and YCRAb615, 73.3%) than when treated with individual antibiotics (Fig. 4d,e,h). With YCRAb 269, the effects of the combination therapy lasted for 24 hours after treatment and survival was higher (43.3%) than it was under monotherapy (polymyxin B, 23.3% and netropsin, 26.7%; Fig. 4f). However, the effect of the combination therapy on YCRAb 301 was not significantly different than that of monotherapy (Fig. 4g). A combination therapy using colistin and netropsin was also evaluated. In contrast to the combination of polymixin B and netropsin, the combination of colistin and netropsin did not exhibit a synergistic effect, except in the case of A. baumannii ATCC 15150 and YCRAb615 (Fig. S5).

Figure 4
figure 4

Synergistic effects of polymyxin B combination with netropsin.

The survival of G. mellonella infected with (a) S. flexneri SF2a, (b) S. typhimurium LT2, (c) P. aeruginosa PA14 and (d) clinical isolates of P. aeruginosa YPa2 at 12 and 24 hour post-infection. Treatments were: polymyxin B 0.5 mg/L (▲), netropsin 100 mg/L (△), polymyxin B 0.5 mg/L + netropsin 100 mg/L (■), control (○) and mock (●). Results represent means of three independent determinations ± standard deviations for 20 insects per treatment. An asterisk (*) indicates a significant difference between combination and monotherapy (p < 0.05).

Full size image

Next, we tested the effectiveness of combination of polymyxin B and netropsin against other Gram-negative pathogenic bacteria (Fig. 4). The combination therapy resulted in 30% and 23.3% higher survival than monotherapy when used against S. flxneri SF2a (polymyxin B 10% and netropsin 0%) and S. typhimurium LT2 (polymyxin B 0% and netropsin 0%) infection, respectively (Fig. 4a,b). In the case of P. aeruginosa PA14 infection, there was no difference in survival under combination therapy (0%) compared to monotherapy or the control (0%; Fig. 4c). However the survival of G. mellonella inoculated with these clinical isolates was higher under treatment with combination therapy (80%) than when treated with monotherapy (polymyxin B 20% and netropsin 3.3%) or the untreated control (0%) in the case of P. aerugnosa YPa2 (Fig. 4d).

Discussion

In this study, a novel combination therapy consisting of a last resort antibiotic (polymixin B) and an antitumor and antiviral drug (netropsin) is developed as an antimicrobial recipe against drug-resistance bacteria through in vitro screening of Streptomyces spp. secondary metabolites followed by in vivo evaluation (Fig. 1). Standard disc diffusion technique for polymixin susceptibility test is unreliable due to slow diffusion of polymixins on agar medium20. Here we utilized highly sensitive tetrazolium dye-based detection of synergism between natural compounds and polymixin B that allowed us to minimize antibiotic usage in the combination therapy. Bacterial secondary metabolites are the gold mine of antibiotics and are produced mainly by Streptomyces spp. in the idiophase21. We found that one of the metabolite among the screens; AN130070 is majorly composed of netropsin that markedly enhanced combinatorial effect with polymixin B against E. coli. Mono-treatment of AN130070 metabolite did not show significant inhibition, suggesting a potential adjuvant activity of netropsin (Fig. S1). The combination therapy was most effective against clinical isolates of A. baumannii wherein 6 out of 7 isolates tested showed FICI < 1 including colistin resistant isolates (Table 1). Presently colistin (polymyxin E) and polymixin B are the only two most clinically relevant polymixins with nearly identical spectrum of activity20. Despite similarities in the mechanism of action and toxicity between colistin and polymixin B, chemical structure, formulation, dosage and pharmacokinetic properties vary to a considerable extent between the two biochemical agents22. In our study, polymixin-netropsin combination therapy was effective against colistin resistant clinical isolates of A. baumannii with FICI < 0.5 (Table 1) and the colistin-netropsin therapy did not show same improvement, indicating that efficacy of combination therapy depends on formulation of type of polymixin used with netropsin (Figs 3 and S5). In addition, since the efficacy of colistin mono-therapy is low against colistin-resistant isolates, a combination therapy of colistin and netropsin can require further in vivo optimization that may vary from that of polymixin B-netropsin combination therapy22.

As such, role of netropsin has been primarily attributed to its antitumor and antiviral properties23. The crescent shaped molecule binds to the AT rich region of the DNA minor grove, thus inhibiting cellular processes such as replication, transcription and thus killing the aggressively dividing cancer cells rapidly24. The mechanism of action of netropsin has been extensively studied in the past13,14,25. Netropsin affects positive supercoiling of DNA and increases average twist per base but do not influence the contour length of the DNA in vitro14. However, some report suggests that netropsin alone do not significantly inhibit bacterial growth26. Of all the species tested, A. baumannii ATCC15150 showed highest synergy of FICI 0.25, while strains of P. aeruginosa showed FICI 1. The variations in the FICI values among different species tested are attributed to species-specific breakpoints of polymixin MIC values20. In particular, high efficacy against A. baumannii clinical isolates (FICI < 0.5) is of utmost clinical significance since polymyxin resistant A. baumannii are on a global rise at an alarming rate3,4. Overall, the decrease in the MIC of polymyxin B-netropsin combination therapy is statistically significant when compared to the reversed counterpart treatment.

The observed synergy was effective and consistent in the G. mellonella host. Few ethical constraints, a short life-cycle and an innate immune response similar to that of higher vertebrates led us to use G. mellonella as an attractive model host19. Polymyxin B and netropsin combination therapy increased the survival of G. mellonella infected with A. baumannii upto 80% compared to survival without antibiotic treatment (Fig. 3). Monotherapies of polymixin B and netropsin showed lesser survival in most of the clinical isolates of A. baumannii and P. aeruginosa YPa2 compared to the combination therapy suggesting in vivo efficacy of the combination therapy (Figs 3b,c and 4d). Variation in survival number of individual positive controls of clinical isolates is reflected in the efficacy of the combination therapy. We did not see significant efficacy in case of P. aeruginosa Pa14 both in vitro and in vivo (Table 1, Fig. 4). In a pharmacokinetic study involving P. aeruginosa infection in G. mellonella larvae, little therapeutic benefit was seen treated with colistin27. Notably, there are no documented studies that show suitability of G. mellonella in studying combination therapies against Gram-negative bacteria. Treatment, optimization and evaluation in clinically relevant model hosts are currently being explored to further develop the polymixin-netropsin combination therapy. Polymyxin B has been used in combination with numerous other antibiotics, such as carbapenems, efflux pump inhibitors, β-lactamase inhibitors and wall teichoic acid synthesis inhibitors10. However, less clinical experience with polymixin B usage and resuscitation of polymixin resistant pathogenic strains demands more robust combination therapies, which minimize the use of multiple antibiotics20.

Off late polymixin B usage is increasingly promoted in the human health care. For example, a 3.5-year study extensively evaluated the clinical efficacy and safety profile of of polymixin B mono- and combination therapy for the respiratory tract infections caused by MDR P. aeruginosa and A. baumannii28. Of the 25 patients administered with 29 courses of polymixin B, 76% displayed favorable response while 52% survived the hospitalization with minimal nephro- and neurotoxcity. A total of 41% achieved microbiological clearance and only 21% displayed end-of-treatment mortality. Similarly more experiences that evaluate polymixin B with new adjuvants and therapeutic agents are required to minimize the dosage and maximize the effectiveness of the polymixin B combination therapy.

Presently, there are no DNA-binding, antibacterial agents qualified for clinical trial. Polymyxins have high affinity for lipopolysaccharides, as they competitively displace divalent cations from the bacteria’s outer membrane20,29. This allows transient uptake of a variety of molecules, leading to changes in membrane permeability and, ultimately, cell death7,30. Therefore, we speculate that polymyxin B enhances netropsin uptake through the outer membrane, thereby increasing the binding of netropsin to bacterial DNA and inhibiting cell growth7,20. However, this speculated model seems to have species-specific limitations. For example, synergy was not significant in case of P. aeruginosa both in vitro and in vivo (Table 1, Fig. 4) which could be attributed to highly evolved efflux pump systems and increased outer membrane impermeability that may render polymixin B and netropsin combination therapy ineffective31. Nevertheless, to the best of our knowledge, this is the first study to investigate in vivo efficacy of a DNA-binding antitumor agent with polymyxin B against Gram-negative bacteria that is demonstrated using the clinical MDR isolates of A. baumannii.

In conclusion, the combination therapy of polymxin B and netropsin provides a potential alternative to the existing limitations of polymixin usage and polymixin resistance. Mass screening of secondary metabolites isolated from Streptomyces spp. is a simple and effective method and hosts a plethora of useful compounds that are effective against a number of drug resistant, Gram-negative bacterial species including polymyxin B and colistin resistant A. baumannii. By using 96-well-containing tetrazolium dye-based in vitro high throughput screening and G. mellonella-based time kill studies we have demonstrated for the first time the efficacy of this novel combination therapy against polymixin B and colistin resistant clinical isolates of Gram-negative pathogenic bacteria.

Methods

Ethical considerations

Under the Korean Law of ‘Bioethics and Safety’, we had verbal approval from the hospital IRB committee to use the bacterial isolates from clinical samples without informed consents if we eliminated all the patients’ identifiers. The Committee approved all the experimental protocols using clinical isolates. All experimental procedures and methods were carried out in accordance with the approved protocols and the Committee’s guidelines.

Bacterial strains and media

E. coli MG1655, A. baumannii ATCC15150, P. aeruginosa strain PA14, Shigella flexneri SF2a and Salmonella typhimurium LT2 were grown in Luria-Bertani (LB) broth at 37 °C. Clinical specimens were isolated from blood, sputum and faeces of patients suffering from P. aeruginosa YPa2, A. baumanii YCSAb5 and YCSAb7 infections. Clinical isolates of colistin-resistant A. baumannii YCRAb49, 257, 269, 301 and 615 were analysed in the hospital (Yonsei Severance Hospital, Seoul, S. Korea) and were grown at 37 °C in LB medium or agar (Asan Company) and Muller-Hinton agar (Asan Company)32,33.

Construction of Streptomyces spp. secondary metabolite library

1800 of Streptomyces spp. were isolated from various soil samples that were obtained from forest, land, compost and river water in S. Korea (http://www.mrscc.net). One gram wet weight of each soil sample was suspended in 9 ml of saline solution, serially diluted down to 10−6 and aliquots of each dilution (100 μl) was plated on humic-vitamin agar34, or fivefold-diluted R2A agar35, all supplemented with cycloheximide (50 μg/ml) to reduce fungal contamination. Sample dilution plates were incubated at 28 °C for 14~21 days until sporulated or non-sporulated Streptomyces sp. colonies were observed and morphologically distinct colonies were isolated (Fig. 1a). A total of 1800 pure bacterial cultures were maintained on Bennett’s agar (10 g glucose, 1 g yeast extract, 2 g peptone, 1 g beef extract and 15 g of agar per liter; pH 7.0) and stored as 20% (v/v) glycerol suspensions at −80 °C. Individual bacterial samples were then inoculated into GSS liquid medium (10 g soluble starch, 20 g glucose, 25 g soybean meal, 1 g beef extract, 4 g yeast extract, 2 g NaCl, 0.25 g K2HPO4 and 2 g CaCO3 per liter; pH 7.2) to activate secondary metabolites secreted from correspondent Streptomyces sp. isolates and incubated at 28 °C with continuous shaking for 5–7 days at 140 rpm (Fig. 1b). Cultures were centrifuged at 10000 rpm and the supernatant was collected as the source of secretory secondary metabolite and stored at −4 °C for future use.

Mass screening of secondary metabolites

A total of 50 μL of the individual secondary metabolites isolated from respective culture broths of Streptomyces spp. were added to 100 μL Muller-Hinton (MH) broth of minimal media containing 250, 25, 2.5, or 0 mg/L of polymyxin B and 0.5% tetrazolium dye as an indicator for oxidatively physiological change in individual wells of 96-well plates. After inoculation with 0.5% of a fresh culture of E. coli strain MG1655 (O.D.600 = 0.3), the plates were incubated at 37 °C. The survival of MG1655 was determined by checking the colour change into violet (alive) or dead (transparent) and by measuring the OD600 nm in a spectrophotometer (Fig. 1c). Selected metabolites were tested in triplicate. The checkerboard assay with polymyxin B and secreted metabolite from selected AN130070 was conducted to optimize the serial concentration of polymyxin B (1, 0.5, 0.25, 0.125, 0.625, 0.031, 0.016 and 0 mg/L) with serially diluted AN130070 metabolite (1, 0.5, 0.25, 0.13, 0.06 and 0; 5 dilution factors from initial concentration of AN130070 metabolite produced from S. netropsis SN01) (Fig. S1)36.

Chemical fractionation of AN130070

A culture (10 L) of strain S. netropsis AN130070 was extracted three times with butanol (5.0 L). The extracts were combined and concentrated under reduced pressure to yield 19.0 g of dark brown gel. The gel was fractionated on a reversed-phase middle pressure liquid chromatography (MPLC) (Teledyne Isco, HP C18 Aq. Redisef Rf gold, 40 id × 300 mm; flow rate 7 mL/min, 100% H2O for 60 min, ~50% aq. MeOH for 300 min, MPLC 2. 50~100% MeOH for 300 min, wash for 60 min), yielding 27 sub-fractions. Sub-fractions 12 (120 mg, a brown gel) was purified by repeated, preparative C18 HPLC (Microsorb 100-5, C18, 21.4 id × 250 mm; gradient elution of 10 to 25% aqueous MeCN for 30 min, flow rate: 7 mL/min; detection by absorption at 297 nm) to afford Compound 1 (1.2 mg, tR 33.0 min) (Figs 1d and S3).

Identification of netropsin

NMR spectra were recorded on a Varian Mercury 400 spectrometer with standard pulse sequences operating at 400 MHz in 1H NMR. Chemical shifts measured in ppm, were referenced to solvent peak (δH 3.31 for CD3OD). Compound 1; 1H NMR(CD3OD, 400 MHz): δ 7.19 (1 H, s), 7.17 (1 H, s), 6.90 (1 H, s), 6.89 (1 H, s), 4.06 (2 H, s), 3.89 (3 H, s), 3.87 (3 H, s), 3.65 (2 H, m), 2.72 (2 H, s). High Resolution-ESI-MS spectra were recorded in the positive ESI (electrospray ionization) mode using a Waters SYNAPT G2 at the Korea Basic Science Institute. HR-TOF-MS (positive ESI mode) m/z 432.2264 [M + H]+ (calculated for C18H27N10O3+, 431.2262) (Fig. 1d).

16S rRNA gene sequencing

Genomic DNA from the strain was prepared using the method by Tamaoka and Komagata37. The 16S rRNA gene was amplified by PCR with the forward primer 27F and the reverse primer 1492R16. Direct sequence determination of the PCR amplified DNA was carried out using an automated DNA sequencer (model ABI 3730XL, Applied Biosystems). The 16S rRNA gene sequence of the strain was compared with available sequences from GenBank using the BLAST program (http://www.ncbi.nlm.nih.gov/BLAST/) to determine an approximate phylogenetic affiliation and sequence similarity values were computed using the EzTaxon server (http://www.eztaxon.org/) (Fig. 1d)38.

In vitro MIC tests

The checkerboard assay as in vitro assay was performed using commercially available netropsin in combination with polymyxin B (Fig. 1e). The initial concentrations of each antibiotic were determined from previously determined MIC90 values39,40,41. Wells containing the lowest fractional inhibitory concentrations were used to determine the fractional inhibitory concentration indices ( where A1 is the MIC value of antibiotic A monotherapy, A2 is the MIC value of antibiotic A combination therapy, B1 is the MIC value of antibiotic B monotherapy and B2 is the MIC value of antibiotic B combination therapy). Synergistic effects were defined in terms of FICI values as ≤0.5 (perfect synergy), >0.5 to <1 (partial synergy), a value of 1 (additive effect) and >1 (loss of synergism). MIC values were interpreted as previously described36,42. MICpol/net range value were determined the series of concentration of first-indicated antibiotics (polymyxin) concentration range in the presence of the second-indicated antibiotics (netropsin) showing synergism in the checkerboard assay in Table 1 (opposite to MICnet/pol)9.

G. mellonella time kill assay

To confirm the synergism of polymyxin B with netropsin in vivo, time kill assay of greater wax moth, Galleria mellonella, as a model system were carried out (Fig. 1f). G. mellonella caterpillars were grown at 37 °C until they reached the fourth larval instar (Ecowin). Caterpillars were then stored in the dark at 37 °C for 3 days for stabilization. Bacterial cells were washed with phosphate-buffered saline and diluted to 106 cfu/larva. Bacterial inocula were injected into the left and right abdominal cavities of 20 randomly selected larvae using 5 μL micro-syringes (MicroliterTM #701, Hamilton). Within 30 min of infection with the microbial suspension, monotherapy or combination therapy was performed on the caterpillars by injecting antibiotic into different prolegs. The amount of antibiotic (polymyxin B and netropsin, Sigma) administered to individual caterpillars was determined by larval weight and MIC results from the in vitro assay. Infected caterpillars were kept in a 37 °C incubator to assess survival 12 and 24 hours after inoculation43. All experiments were repeated thrice.

Data analysis

All experiments were repeated thrice. The results were reported as the mean ± SD from at least three representative experiments. Statistical analyses were carried out with JMPIN 4.0.4 (SAS Institute) using one-way ANOVA followed by a Student’s t-test for treatment comparisons. Differences were considered to be significant at p < 0.05.

Additional Information

How to cite this article: Chung, J.-h. et al. Combination therapy with polymyxin B and netropsin against clinical isolates of multidrug-resistant Acinetobacter baumannii. Sci. Rep. 6, 28168; doi: 10.1038/srep28168 (2016).