In Vitro and In Vivo Effects of the Polymyxin-Vorinostat Combination Therapy Against Multidrug-Resistant Gram-Negative Pathogens

Abstract

With the stagnancy of antibiotics development, polymyxins have become the last defense for treatment of multidrug-resistant (MDR) Gram-negative bacteria, whereas the effect of polymyxin monotherapy is limited by resistance. The objective of this study was to evaluate the effects of polymyxin B (PMNB)-vorinostat (SAHA) combination therapy against Gram-negative pathogens in vitro and in vivo. The antibacterial activities of PMNB and SAHA were evaluated by susceptibility testing. The synergistic effect was assessed by checkerboard tests and time-killing kinetics experiments. Cellular morphology studies and reactive oxygen species (ROS) assay were conducted to explore potential mechanisms. Also, Galleria mellonella models were made to evaluate the antibacterial effects in vivo. PMNB-SAHA had the synergistic effect against all tested isolates, reducing >2 log₁₀ colony-forming units (CFU)/mL at 40 minutes, and showed more powerful antibacterial effects than PMNB alone in the 24-hour window. Cellular morphology study showed the change of membrane and disruption of integrity. ROS assay showed more oxidative stress in combination than PMNB or SAHA monotherapy. In animal models, PMNB-SAHA showed a higher survival rate than that of monotherapy. This study is the first to report the synergistic antibacterial effect of PMNB-SAHA therapy against MDR Gram-negative bacteria. Further clinical research is needed to confirm the results.

Introduction

With the rapid emergence and spread of drug resistance, Gram-negative pathogens with multidrug-resistant (MDR) have become a serious crisis for nosocomial infection and human health.¹ Treatments for infections by MDR pathogens, including MDR Enterobacteriaceae, Acinetobacter baumannii, and Pseudomonas aeruginosa, find it hard to achieve the desired therapeutic effect, causing higher mortality. For intensive care unit (ICU) patients, 70% of the infections were caused by Gram-negative pathogens.² Previous occupation of ICU rooms by patients with MDR Gram-negative bacterial strains is an independent risk factor for subsequent room occupants to be infected by these bacteria.³ Some MDR bacterial isolates are even resistant to almost all antibiotics. With the lack of effective novel antibiotics, polymyxins have become the last line of antibiotic defense.⁴

Polymyxins (A–E) are a type of cationic cyclic polypeptide antibiotics. As previously reported, polymyxin B (PMNB) and colistin (polymyxin E) have been used in clinical work.⁵ However, PMNB and colistin share the same antibiotic mechanisms, as they compete with Mg²⁺ and Ca²⁺ in lipid A binding, destroy the outer membrane of bacteria, and give rise to cytoplasmic outflow and cell death.⁶ Unfortunately, when using polymyxin monotherapy to treat Gram-negative bacteria, modification of lipid A and loss of lipopolysaccharides may occur, which increases the emergence of polymyxin resistance (PR).⁷

Colistin is a class of inactive antibiotics. Colistimethate sodium injected intravenously is converted to the active version of colistin slowly and incompletely in vivo.⁸ Comparatively, PMNB as the active version itself reaches effective blood concentration rapidly.⁹ Previous studies also reported lower nephrotoxicity of PMNB compared with colistin.¹⁰ Therefore, PMNB was chosen in our study.

Histone deacetylases (HDACs) can remove acetyl groups, causing chromatin condensation and gene silencing, and are associated with inhibition of transcription and regulation of cell signaling pathways generally.^11,12 Vorinostat (suberoylanilide hydroxamic acid or SAHA) is a hydroxamic acid derivative that can inhibit zinc-dependent HDACs (class I, II, and IV).^13,14 Several studies demonstrated that SAHA could induce DNA damage and produce oxidative lesions in many types of tumor.^14–16 It has been approved by the U.S. Food and Drug Administration as an oral drug treating relapsed and refractory cutaneous T cell lymphoma (the recommended treatment protocol is 400 mg/day with food).^17,18

In addition, recent reports indicated that antibacterial activity of polymyxins was enhanced for MDR Gram-negative bacilli via synergistic combination with the anthelmintic drug.^19,20 Meanwhile, there is a strong correlation between rapid target polymyxins therapeutic concentration attainment and antibacterial activity and rapid emergence of PR bacteria.²¹ Thus, to achieve the desired therapeutic effect at a lower drug concentration, a novel PMNB combination therapy is important and imperative. Previous research also showed that SAHA had effects against malaria and human immunodeficiency virus.^22,23 However, the effects of SAHA on bacterial cells are unclear and may be related to enhance oxidative stress.²⁴ We hypothesized that SAHA can reduce the concentration of PMNB without lowering the therapeutic effect. Therefore, here, we investigated the potential synergistic antibacterial activity of PMNB and SAHA against MDR and polymyxin-resistant Gram-negative bacteria.

Materials and Methods

Bacterial isolates, drugs, and Galleria mellonella

The antibacterial activity of the PMNB and SAHA was tested through six Gram-negative bacterial strains. Among these isolates, Escherichia coli K-12 strain BW25113, Acinetobacter baumannii ATCC 17978, Pseudomonas aeruginosa PA14, and Klebsiella pneumoniae ATCC 43816 were wild-type standard strains. Acinetobacter baumannii HR13 and Klebsiella pneumoniae KP024, two clinical MDR wild-type pathogens, were isolated from ICU patients at a tertiary care hospital in Anhui, China. In addition, three Gram-positive bacteria were used to study the antibacterial activity of SAHA including wild-type standard strains Staphylococcus aureus RN450 and Enterococcus faecalis ATCC 29212, as well as one clinical wild-type isolate methicillin-resistant Staphylococcus aureus HR165. All isolates (from the Anhui Center for Surveillance of Bacterial Resistance, Hefei, China) were stored in cation-adjusted Muller-Hinton Broth (CAMHB) with 50% glycerol in cryovials at −80°C and grown in CAMHB or on broth agar at 37°C. PMNB was obtained from Sigma-Aldrich (St. Louis, MO). It was dissolved in ultrapure water and sterilized with a 0.22-μm cellulose acetate membrane filter (Solarbio Science and Technology Co., Ltd., Beijing, China). SAHA was ordered from Yuanye Bio-Technology Co., Ltd. (Shanghai, China) and dissolved in 99% dimethyl sulfoxide (DMSO; Sigma-Aldrich). All prepared solutions were stored at −20°C within 1 month. Galleria mellonella larvae (Kaide Ruixin Co., Ltd., Tianjin, China) in their final-instar stage were stored in the dark and used within 7 days from shipment.

Susceptibility testing

The minimum inhibitory concentrations (MICs) were measured by using the broth microdilution method for PMNB and SAHA in all strains. Overnight cultures were diluted 200-fold into fresh CAMHB, grown to log phase OD₆₀₀ = 0.3 (the optical density of growth cultures was measured at 600 nm), about 5 × 10⁸ CFU/mL, and finally diluted 1,000-fold into broth. A final bacterial inoculum of 5 × 10⁵ CFU/mL was placed in each tube. PMNB and SAHA were added separately to the first tube and diluted serially to the last tube. The ranges of concentrations were 0.25–16 mg/L for PMNB, and 16–1,024 mg/L for SAHA. Tubes were read after overnight incubation at 37°C. The MICs were identified as the lowest drug concentrations that inhibit the visible growth of the bacteria. Susceptibility was determined by using Clinical and Laboratory Standards Institute (CLSI) breakpoints and the European Committee on Antimicrobial Susceptibility Testing (EUCAST).^25,26 After overnight treatment or untreatment (controls), MICs of PMNB were used to evaluate the change of polymyxin susceptibility at 24 hours.

Synergistic effect testing

The synergistic effect was evaluated by the micro-dilution checkerboard assay. A single colony of bacteria was incubated overnight at constant rotation (5 × g) in CAMHB at 37°C. The overnight cultures were adjusted to about 10⁸ CFU/mL and diluted by 1:1,000 with broth (5 × 10⁵ CFU/mL) bacterial suspension. Ninety-six-well microtiter plates were used, and 100 μL of the 5 × 10⁵ CFU/mL bacterial suspensions was prepared and mixed with drugs solution with the same final concentration, respectively, into each horizontal or vertical well. The ranges of selected concentration depended on previously tested MICs, which were 0, 1/16 × , 1/8 × , 1/4 × , 1/2 × , and 1 × MIC from low to high in each well for PMNB, as well as 0, 1/32 × , 1/16 × , 1/8 × , 1/4 × , 1/2 × , and 1 × MIC for SAHA. The microwell plates were incubated at 37°C for 24 hours before reading. The fractional inhibitory concentration index (FICI) was calculated to assess synergistic effect. The FICI was calculated by the equation (MIC of drug_A in combination/MIC of drug_A alone) + (MIC of drug_B in combination/MIC of drug_B alone) and defined as follows: FICI ≤0.5 means synergy; 0.5 <FICI ≤4.0 means indifference; FICI >4 means antagonism.^27,28

Time-kill studies

Time-kill studies were conducted in all strains by using PMNB or SAHA alone and in combination, according to the method previously described.²⁹ Synergy is defined as a reduction of ≥2 log₁₀ CFU/mL by using combined therapy compared with the most active monotherapy.³⁰ Bacterial strains were cultivated overnight in 5 mL CAMHB. The cultures were diluted to 1:200 with 10 mL of fresh broth and grown up to the OD₆₀₀ value 0.25–0.3 (log-phase) with a shaker at 37°C. Then, they were diluted to a concentration of 5 × 10⁵ CFU/mL and transferred to borosilicate glass tubes for each strain. PMNB (2 mg/L), SAHA (50 mg/L), or both compounds were added to achieve corresponding final concentrations. Samples and untreated cultures were removed at 0, 10, 20, 30, and 40 minutes and 2, 4, 8, and 24 hours after 10- and 10⁵-fold dilutions with saline. Samples of bacterial cell suspension (1 mL) were aseptically placed on agar plates and incubated overnight at 37°C to measure the viable colony cells counts (CFU/mL). Controls (untreated) were set for each experiment.

Cellular morphology study

Scanning electron microscopy (SEM) was used for detecting the cell morphologic changes of polymyxin-susceptible Acinetobacter baumannii ATCC 17978 and polymyxin-resistant Acinetobacter baumannii HR13. As the previous time-kill studies, bacteria were treated with neither, one, or both of PMNB (2 mg/L) and SAHA (50 mg/L) for 2 hours in broth. Cultures were centrifuged at 4,600g for 10 minutes; the obtained supernatants were discarded; the bacterial pellets were resuspended in 1 mL of phosphate-buffered saline (PBS), and then they were washed three times in 1 mL PBS as described earlier. After that, the bacterial pellets were resuspended in 0.5 mL PBS with 2.5% glutaraldehyde and fixed for 2 hours at 37°C. Samples were observed and assessed at least 300 cells per assay on the SEM.

Reactive oxygen species assay

To measure intracellular reactive oxygen species (ROS) generation, the fluorescent probe carboxy-H2DCFDA [5(6)-carboxy-2’,7’-dichlorodihydrofluorescein diacetate] from Thermo Fisher (USA) was used. Carboxy-H2DCFDA entered A. baumannii cells, and it was converted into a membrane-impermeable cognate by cellular esterases or oxidized to a fluorescent form by superoxide, hydrogen peroxide (H₂O₂), or hydroxyl radicals.³¹ The signal of fluorescence was analyzed by using a flow cytometer (Beckman Coulter CyAn ADP analyzer). Acinetobacter baumannii ATCC 17978 cells were grown with neither, one, or both PMNB (2 mg/L) and SAHA (50 mg/L) for 30 minutes; washed twice with precooled 1 × PBS to scavenge the agents; and concentrated by centrifugation (4,600g for 2 minutes). Samples were incubated with 5 μM carboxy-H2DCFDA for 10 minutes at 37°C and used to measure ROS accumulation in flow cytometry.

In addition, we also used the fluorescent reporter dye, 3’-(p-hydroxyphenyl) fluorescein (HPF; Thermo Fisher, USA) to further verify, which was oxidized by hydroxyl radicals with high specificity. Cells were treated with PMNB alone, SAHA alone, or PMNB-SAHA, or with 0.15% H₂O₂ at 37°C with gentle shaking for 30 minutes. After being washed twice with precooled 1 × PBS and concentrated by centrifugation, 5 μM HPF was added to treated or untreated cultures. Fluorescence values were measured by a BioTek Synergy HT Multi-Mode Plate Reader (BioTek, USA) with an excitation/emission setting of 490/515 nm.³²

To further determine that ROS is a key factor in lethality for the PMNB-SAHA combination. DMSO and thiourea, effective hydroxyl-radical scavengers, can prevent ROS accumulation, which were used to study ROS-dependent or -independent modes of killing. According to our previously described time-kill experiment, prepared samples (1 mL) were placed on agar plates containing 5% DMSO or thiourea at a final concentration of 600 mM, and they were used to study antibacterial effects in combination. In addition, the effect of growth kinetics was studied in DMSO or thiourea alone (Supplementary Data).^33,34

G. mellonella killing-treatment assay

According to previous studies,³⁵ we built the G. mellonella infection model to evaluate the antibacterial activity of PMNB combined with SAHA in vivo. G. mellonella were stored at 15°C within 5 days from shipment. Sixteen larvae (250–350 mg in weight) were selected randomly and used in each group. Bacterial cells (Acinetobacter baumannii ATCC 17978 and HR13) were washed twice in normal saline (NS; 0.9% saline) and then diluted to OD₆₀₀ 0.35. Bacterial cells were injected into the hemocoels of each larva through the last left proleg by a 50-μL Hamilton syringe (Hamilton Company, Shanghai, China). Each group of larvae was injected with 10 μL of bacterial suspensions containing ∼10⁵ CFU/larva of bacteria in NS. After injection, larvae were incubated at 37°C. Larvae were considered dead if they did not respond to touch, and the number of dead larvae was recorded daily for 6 days. G. mellonella larvae were infected with a lethal dose of A. baumannii, and drugs (PMNB [2.5 mg/kg] or SAHA [8 mg/kg] alone and in combination) were inoculated into a different proleg within 20 minutes. Drug doses were based on those used for humans. Two control groups (NS injection group and no injection group) were set for each experiment.

Statistical analysis

All statistical analyses were performed with SPSS 21.0. One-way analysis of variance with post hoc test was used to calculate the change of antibacterial activity, cell length, and fluorescence values in each drug alone or in combination. Survival curves using Kaplan–Meier methods were drawn and compared with the log-rank test to compare the differences between groups. A p-value of <0.05 indicates that the difference is significant.

Results

Antimicrobial susceptibility of PMNB and SAHA

For all Gram-negative strains, MICs of PMNB and SAHA are shown in Table 1. According to the CLSI and EUCAST, MDR clinical isolates (Acinetobacter baumannii HR13 and Klebsiella pneumoniae KP024) were resistant to PMNB, whereas others were susceptible in this study. SAHA alone had antibacterial activity at a concentration of up to 512 μg/mL for all Gram-negative isolates, which is higher than Gram-positive bacteria (Supplementary Table S1).

Table 1.

Minimum Inhibitory Concentrations of Polymyxin B and Vorinostat

Bacterial isolate	MIC (μg/mL)
Bacterial isolate	PMNB	SAHA
Escherichia coli BW25113	0.5	512
Acinetobacter baumannii ATCC 17978	0.5	512
Acinetobacter baumannii HR13 ^MDR,PR	4	512
Klebsiella pneumoniae ATCC 43816	1	512
Klebsiella pneumoniae KP024 ^MDR,PR	8	512
Pseudomonas aeruginosa PA14	0.5	512

MIC, minimum inhibitory concentration; PMNB, polymyxin B; SAHA, vorinostat.

The changes to the PMNB MICs after overnight treatments are shown in Table 2. For the controls and combination group, MICs of PMNB did not change (within twofold of initial MICs) at 24 hours. With the PMNB alone, MICs of PMNB remained unchanged for polymyxin-resistant isolates whereas they increased remarkably more than eight times for polymyxin-susceptible isolates at 24 hours.

Table 2.

Treatment of Polymyxin B With or Without Vorinostat Impact on Minimum Inhibitory Concentrations at 24 Hours

Bacterial isolate	MICs of PMNB relative to their initial values (μg/mL)
Bacterial isolate	Control^a	PMNB	PMNB + SAHA
Escherichia coli BW25113	1 × MIC (0.5)	8 × MIC (4)	1/2 × MIC (0.25)
Acinetobacter baumannii ATCC 17978	1 × MIC (0.5)	16 × MIC (8)	1 × MIC (0.5)
Acinetobacter baumannii HR13 ^MDR,PR	1 × MIC (4)	1 × MIC (4)	1 × MIC (4)
Klebsiella pneumoniae ATCC 43816	1 × MIC (1)	16 × MIC (16)	1 × MIC (1)
Klebsiella pneumoniae KP024 ^MDR,PR	1 × MIC (8)	2 × MIC (16)	1 × MIC (8)
Pseudomonas aeruginosa PA14	1 × MIC (0.5)	16 × MIC (8)	1 × MIC (0.5)

The values of PMNB MICs are shown in brackets.

Controls, no treatment.

Micro-dilution checkerboard assays

We evaluated synergistic interaction (FICI <0.5) between PMNB and SAHA by calculating the FICIs (Fig. 1). For MDR and polymyxin-resistant clinical isolates, the PMNB-SAHA combination reduced the MIC of PMNB effectively from 4 to 0.5 μg/mL and of SAHA to 64 μg/mL against Acinetobacter baumannii HR13 (Fig. 1D). For Klebsiella pneumoniae KP024, the combination reduced the MIC of PMNB from 8 to 1 μg/mL and of SAHA to 32 μg/mL (Fig. 1F). This indicates that SAHA is a potential companion of polymyxin in treating infections of MDR and polymyxin-resistant Gram-negative pathogens.

FIG. 1.

The inhibitory concentration is shown in a combination of PMNB and SAHA. (A) Escherichia coli BW25113, FICI = 0.3125. (B) Pseudomonas aeruginosa PA14, FICI = 0.375. (C) Acinetobacter baumannii ATCC 17978, FICI = 0.375. (D) Acinetobacter baumannii HR13, FICI = 0.1875. (E) Klebsiella pneumoniae ATCC 43816, FICI = 0.375. (F) Klebsiella pneumoniae KP024, FICI = 0.1875.

Time-kill studies

The time-killing kinetics of PMNB (2 mg/L) or SAHA (50 mg/L) alone and the combination for all selected isolates are shown in Figs. 2 and 3. Against the four polymyxin-susceptible strains, PMNB monotherapy showed rapid bactericidal effect with ∼2.5 log₁₀ CFU/mL killing for Escherichia coli BW25113 (Fig. 2A), 2.8 log₁₀ CFU/mL killing for Pseudomonas aeruginosa PA14 (Fig. 2B), 2.4 log₁₀ CFU/mL killing for Acinetobacter baumannii ATCC 17978 (Fig. 2C), and 2.5 log₁₀ CFU/mL killing for Klebsiella pneumoniae ATCC 43816 (Fig. 2E) within 40 minutes. However, bacterial counts returned within 24 hours (Fig. 3). With the combination, there were nearly 5 log₁₀ CFU/mL of bacterial killing within 40 minutes (Fig. 2) and they maintained stable antibacterial activity for up to 24 hours (Fig. 3). The bacterial counts of combination were reduced >2 log₁₀ CFU/mL more than that of PMNB monotherapy (Fig. 2). In addition, bacterial killing of SAHA monotherapy was not significant compared with the growth control group (p > 0.05) (Figs. 2 and 3).

FIG. 2.

Synergistic effect of PMNB (2 mg/L) combined with SAHA (50 mg/L) against Gram-negative bacterial strains in 40 minutes. (A) Escherichia coli BW25113. (B) Pseudomonas aeruginosa PA14. (C) Acinetobacter baumannii ATCC 17978. (D) Acinetobacter baumannii HR13. (E) Klebsiella pneumoniae ATCC 43816. (F) Klebsiella pneumoniae KP024 (clinical isolate). (D, F) MDR and PR isolates. Control, no treatment. Error bars represent standard errors of the means. The p-values (compared PMNB and PMNB-SAHA at 40 minutes) are <0.05, <0.01, <0.01, <0.01, <0.01, and <0.01, respectively.

FIG. 3.

Twenty-four hours-kill kinetics of PMNB and SAHA monotherapy and combination therapy against Gram-negative bacterial strains. (A) Escherichia coli BW25113. (B) Pseudomonas aeruginosa PA14. (C) Acinetobacter baumannii ATCC 17978. (D) Acinetobacter baumannii HR13. (E) Klebsiella pneumoniae ATCC 43816. (F) Klebsiella pneumoniae KP024. Controls, no treatment. Error bars represent standard errors of the means. The p-values (compared PMNB and PMNB-SAHA at 24 hours) are all <0.01, respectively.

Against the polymyxin-resistant isolates, Acinetobacter baumannii HR13 (Fig. 2D) and Klebsiella pneumoniae KP024 (Fig. 2F), monontherapy with PMNB reduced far <1 log₁₀ CFU/mL of bacterial counts (Fig. 3D, F). However, synergistic activity was observed in the combination therapy within 40 minutes, with almost 5 log₁₀ CFU/mL of bacteria reduced rapidly (Fig. 2D, F) and no regrowth at 24 hours for Acinetobacter baumannii HR13 (Fig. 3D) and Klebsiella pneumoniae KP024 (Fig. 3F).

Cellular morphology studies

The SEM images of polymyxin-susceptible Acinetobacter baumannii ATCC 17978 and polymyxin-resistant Acinetobacter baumannii HR13 incubated with PMNB, SAHA, or both are shown in Fig. 4. The SEM results revealed that the cell membrane surfaces of bacteria were smooth, and the average cell length remained ∼2 μm in the untreated group (Fig. 4A, E). For Acinetobacter baumannii ATCC 17978, PMNB monotherapy led to visible shortening of the cells (p < 0.01) and the surface of the cell membrane became rough and shrunk (Fig. 4B). Conversely, no significant effects were found in bacterial size (p > 0.05) and membrane surface for Acinetobacter baumannii HR13 (Fig. 4F). SAHA alone had minimal impacts on the surface of bacterial cell membrane and cell length (p > 0.05) for both (Fig. 4C, G). With the PMNB-SAHA combination, bacterial cell length was shortened (p < 0.01), but the bacterial surface was smoother than PMNB monotherapy (Fig. 4D, H).

FIG. 4.

Impact of PMNB and SAHA treatment on the morphology of polymyxin-susceptible and polymyxin-resistant Acinetobacter baumannii. Images from SEM for polymyxin-susceptible Acinetobacter baumannii ATCC 17978 (A–D) and polymyxin-resistant Acinetobacter baumannii HR13 (E–H) treated with 2 mg/L PMNB (B, F), 50 mg/L SAHA (C, G), or both (D, H). (A) and (E) represent the untreated controls.

ROS assays

There was a need to confirm whether treatment of the PMNB-SAHA combination induced ROS production. The ROS levels of Acinetobacter baumannii ATCC 17978 are shown in Fig. 5. For the carboxy-H2DCFDA, SAHA monotherapy group had almost no change of fluorescent intensity compared with the control group and PMNB monotherapy showed relatively low production of ROS, whereas the combination group showed notable ROS accumulation (Fig. 5A). For the HPF, we treated cultures with H₂O₂ as a positive control, which was previously demonstrated to induce hydroxyl radicals.³⁶ We observed an about threefold increase in fluorescence, indicating an increase in ROS production (p < 0.001). We next assayed the A. baumannii cultures treated with PMNB alone and observed that monotherapy induced an about twofold increase in fluorescence compared with the untreated controls (p < 0.01). SAHA alone did not alter the level of fluorescence (p > 0.05). However, when treated with PMNB-SAHA, we observed that the combination induced an about twofold increase in fluorescence values compared with PMNB alone (p < 0.001).

FIG. 5.

The generation of ROS in PMNB and SAHA for Acinetobacter baumannii ATCC 17978. (A) The treated or untreated samples were incubated with 5 μM carboxy-H2DCFDA for 10 minutes before observation. Controls, no treatment at 30 minutes. (B) Cultures were untreated or were treated with 0.15% H₂O₂, 50 mg/L SAHA, 2 mg/L PMNB, or both for 30 minutes. After treatment, the hydroxyl radical-specific fluorescent dye was added, and the fluorescence was measured (490/515 nm). Bars represent the means and standard deviations of triplicate samples. **p < 0.001, ***p < 0.0001. NS, no significance.

Then, we sought to determine whether ROS production was required for the rapid killing. Therefore, we utilized the time-killing kinetics experiments (Fig. 6). After the addition of 5% DMSO, the protection was found in PMNB monotherapy (reduced by 1.7 log₁₀ CFU/mL; p < 0.05) and combination group (reduced by 3 log₁₀ CFU/mL; p < 0.01) compared with no DMSO added (Fig. 6A). The bactericidal effect of PMNB-SAHA combination was reduced to about 1/1,000 after 5% DMSO was added. Likewise, when we detected the presence of thiourea, we found a striking decrease in the ability of PMNB-SAHA combination to kill A. baumannii compared with the combination without thiourea (reduced by 1.9 log₁₀ CFU/mL; p < 0.01). PMNB with thiourea was reduced by 0.8 log₁₀ CFU/mL compared with without thiourea (p < 0.05) (Fig. 6B). Obviously, treatment with DMSO or thiourea markedly inhibited the rapid decrease in antibacterial activity. No protection was observed in SAHA alone with DMSO or thiourea (p > 0.05). The addition of DMSO or thiourea alone did not change the growth kinetics (Supplementary Fig. S1).

FIG. 6.

ROS accumulation was inhibited in time-killing curves for Acinetobacter baumannii ATCC 17978. As the hydroxyl-radical scavengers. (A) DMSO and (B) thiourea prevented ROS accumulation. Protection was found in PMNB (p < 0.05) and the combination (p < 0.01) with 5% DMSO added compared with no DMSO at 40 minutes. PMNB (p < 0.05) and the combination (p < 0.01) with 600 mM thiourea added compared with without thiourea at 40 minutes. Error bars represent standard errors of the means.

G. mellonella killing-treatment model

In the infection models of Acinetobacter baumannii ATCC 17978 and HR13, the survival rate of G. mellonella with treatment of PMNB alone was 62.5% (Fig. 7A) and 31.25% (Fig. 7B) at 96 hours after inoculation. However, when SAHA was used as treatment, survival rates were almost identical to the untreated controls, which was consistent with the results in vitro. When the PMNB-SAHA combination was used, it protected the G. mellonella from infection of lethal bacteria more effectively than PMNB monotherapy did (p < 0.001), as 93.75% (Fig. 7A) and 87.5% (Fig. 7B) of G. mellonella were still alive in ATCC 17978 and HR13 infection models.

FIG. 7.

Effect of combined therapy with PMNB and SAHA on the Galleria mellonella infective model. (A) Acinetobacter baumannii ATCC 17978. (B) Acinetobacter baumannii HR13. Statistical significance was calculated with the log-rank test comparisons. PMNB-SAHA has a better therapeutic effect than PMNB (p < 0.001). NS, normal saline.

Discussion

The infections of MDR Gram-negative bacteria are becoming a public health concern. However, the development of new antibiotics has been limited over the recent decades. In addition, polymyxin monotherapy can transform primary infection into polymyxin-resistant bacterial infection, leading to more health care burden and higher mortality rates. Thus, novel treatment strategies are urgently needed. SAHA, as a derivative of hydroxamic acid, is used to treat patients with refractory and relapsed cutaneous T cell lymphoma.¹⁷ The possible mechanisms of SAHA against proliferation include HDAC inhibition and DNA damage.¹⁶ However, the mechanism of potential antibacterial effects still remains uncertain. In this study, we investigated the effects and mechanisms of PMNB-SAHA combination on Gram-negative infections.

In our study, SAHA alone showed no antibacterial activity for all Gram-negative isolates (compared with the controls, p > 0.05). PMNB showed certain antibacterial activity in a short time and gradually reduced in the next few hours. Conversely, PMNB-SAHA combination enhanced this certain antibacterial activity (∼5 log₁₀ CFU/mL) and inhibited the growth of bacteria within 24 hours. Also, PMNB monotherapy reduced the susceptibility for PMNB and thus increased the MICs of originally polymyxin-susceptible isolates, which could explain the regrowth of bacteria observed at 24 hours, and this may account for the clinical failures. However, the combination therapy prevented the reduction of polymyxin susceptibility and showed a long-lasting antibacterial effect. In addition, the MICs of SAHA to Gram-positive isolates were lower than Gram-negative stains, which showed that the potential activity of SAHA for Gram-positive bacteria is worth studying.

For the SEM images, PMNB monotherapy caused a rough and shrunken cell membrane surface on Acinetobacter baumannii ATCC 17978, which indicated that the stability and integrity of the cell membrane was destroyed. However, there was no visible change in Acinetobacter baumannii HR13. It is possible that the PMNB resistance in Acinetobacter baumannii HR13 altered their membrane surface interaction. The combination therapy showed that the outer membrane surface remained smooth on the shortened bacterial cells. It suggested that the combination therapy prevented disruption of the cell membrane integrity and stability, which could be a change in the mechanism of action of PMNB.

As the results of ROS assay, when using carboxy-H2DCFDA, PMNB monotherapy showed an obvious reduction of ROS production compared with the control group. As previous research demonstrated, polymyxins can permeabilize the outer membrane of Gram-negative bacteria and destabilize it.³⁷ This may cause the intracellular substance to flow out, so the fluorescence intensity was not completely detected in the carboxy-H2DCFDA, which explains the reduction of ROS detected. In the combination, bacterial fluorescence intensity obviously increased for both, and the HPF fluorescence values were remarkably stronger than the sum of the fluorescence levels of PMNB and SAHA alone. These indicate that SAHA can increase production of ROS. However, fluorescence intensity of the SAHA monotherapy group was no change for the control group, and there was no difference between SAHA monotherapy with and without DMSO or thiourea. This may be because SAHA alone cannot pass through the outer membrane of Gram-negative bacteria, and thus cannot induce the increase ROS production. Also, compared with the combination without DMSO group, the antibacterial activity was significantly inhibited after adding ROS inhibitors. This means that DMSO and thiourea cleared the production of oxidative stress caused by SAHA alone, which is the underlying main mechanism for PMNB-SAHA combination.

Hence, combining our results with previous morphological studies, we supposed that PMNB destroys the integrity of the outer membrane, letting SAHA into the bacterial cell, which causes oxidative stress and enhances the generation of ROS. Interestingly, the HPF fluorescence values of PMNB alone significantly increased (p < 0.01), and antibacterial activity was also reduced after ROS inhibitors were added. These indicate that an increase of ROS production may also be a potential bactericidal mechanism of PMNB, which was consistent with previous reports.^34,36 SAHA enhances the production of ROS, which is accompanied by the hyperpolarization of mitochondrial membrane,³⁸ and causes the outflow of cations to the cytoplasm, such as Mg²⁺ and Ca²⁺. It also inhibits the electrostatic interaction between the positively charged (PMNB) and the negatively charged lipid A, and therefore, reduces the extent of disruption of the cell membrane integrity and stability. This may also explain why the cell membrane of the combination therapy group was smoother than that of the PMNB monotherapy group. Thus, in the combination group, the main bactericidal mechanism of PMNB may be production of ROS as well as disrupting the cell membranes to some extent and thus letting SAHA in and exhibiting synergistic antibacterial activity.

Recently, G. mellonella has been used as a well-accepted invertebrate insect model to study the effects of antimicrobial combination in vivo. The main advantages of this model include the low cost of maintenance, the fast life cycle, the possibility of using a large number of caterpillars, and the innate immune system, which are evolutionarily conserved relative to mammals. We investigated PMNB or SAHA alone and in combination in the infection model of Acinetobacter baumannii ATCC 17978 and HR13. SAHA monotherapy showed no effects in both infection models (p > 0.05). PMNB monotherapy showed some effects, whereas PMNB-SAHA combination was significantly more effective (p < 0.001), especially in the polymyxin-resistant infection model (HR13). This was consistent with our in vitro results that provided preliminary evidence in vivo. This combination could be a useful promotion of existing therapeutic methods.

Overall, this study is the first to report the strong and constant bactericidal effects of PMNB-SAHA combination against MDR Gram-negative bacteria. However, we should acknowledge some limitations. First, we chose MDR, a polymyxin-resistant A. baumannii strain, and a K. pneumoniae strain, which were reported as a top bacterial “superbug” by the World Health Organization for research.^39,40 There still remains other polymyxin-resistant Gram-negative isolates, such as P. aeruginosa and E. coli, which could also be tested to further confirm our results. Second, the concentration of SAHA (50 mg/mL) we used in vitro was higher than the blood concentration of clinical treatment and changes in pharmacokinetics of the combination are unclear. Then, we used G. mellonella models to test the effects of drug combination at clinically treated concentrations in vivo; however, more effective mammal models are further needed to validate the results at clinically relevant blood concentrations of SAHA and explore other possible mechanisms of combination as well as the effect of pharmacokinetics in vivo when combined therapy is used.

With the stagnant development of new antibacterial drugs, this study gives hope for using novel drug combinations to treat MDR Gram-negative bacteria. Further clinical studies are needed to confirm the therapeutic effects of PMNB-SAHA combination.

Conclusion

PMNB-SAHA combination, as a novel therapy, can increase antibacterial activity of PMNB against MDR Gram-negative pathogens mainly via the generation of ROS and reduce the emergence of PR. Further clinical research is needed to confirm its efficacy.

Footnotes

Acknowledgments

The authors wish to thank all the students in the Institute of Bacterium Resistance of Anhui Medical University (Hefei, China) for their helpful assistance.

Ethical Statement

This study was approved by the Ethical Committee of the Second Hospital of Anhui Medical University (Hefei, China) and received Joint Commission International certification (approval no.: PJYX2018-001).

Disclosure Statement

The author reports no conflicts of interest in this work.

Funding Information

Funding for this work is supported by grant 81673242 from the National Natural Science Foundation of China.

Supplementary Material

References

Boucher

H.W.

, Talbot

G.H

, Benjamin

D.K

Jr. , Bradley

, Guidos

R.J

, Jones

R.N

, Murray

B.E

, Bonomo

R.A

, and Gilbert

. 2013. 10 x ‘20 Progressdevelopment of new drug active against Gram-negative bacilli: an update from the Infectious Diseases Society of America. Clin. Infect. Dis. 56:1685–1694.

Gaynes

, and Edwards

J.R.

. 2005. Overview of nosocomial infections caused by Gram-negative bacilli. Clin. Infect. Dis. 41:848–854.

Nseir

, Blazejewski

, Lubret

, Wallet

, Courcol

, and Durocher

. 2011. Risk of acquiring multidrug-resistant Gram-negative bacilli from prior room occupants in the intensive care unit. Clin. Microbiol. Infect. 17:1201–1208.

Nation

R.L.

, Li

, Cars

, Couet

, Dudley

M.N

, Kaye

K.S

, Mouton

J.W

, Paterson

D.L

, Tam

V.H

, Theuretzbacher

, Tsuji

B.T

, and Turnidge

J.D.

. 2015. Framework for optimisation of the clinical use of colistin and polymyxin B: the Prato polymyxin consensus. Lancet Infect. Dis. 15:225–234.

Stansly

P.G.

, Shepherd

R.G

, and White

H.J.

. 1947. Polymyxin: a new chemotherapeutic agent. Bull. Johns Hopkins Hosp. 81:43–54.

Schindler

, and Osborn

M.J.

. 1979. Interaction of divalent cations and polymyxin B with lipopolysaccharide. Biochemistry. 18:4425–4430.

Arroyo

L.A.

, Herrera

C.M

, Fernandez

, Hankins

J.V

, Trent

M.S

, and Hancock

R.E.

. 2011. The pmrCAB Operon mediates polymyxin resistance in Acinetobacter baumannii ATCC 17978 and clinical isolates through phosphoethanolamine modification of Lipid A. Antimicrob. Agents Chemother. 55:3743–3751.

, Li

J.C

, Nation

R.L

, Jacob

, Chen

, Lee

H.J

, Tsuji

B.T

, Thompson

P.E

, Roberts

, Velkov

, and Li

. 2013. Pharmacokinetics of four different brands of colistimethate and formed colistin in rats. J. Antimicrob. Chemother. 68:2311–2317.

Sandri

A.M.

, Landersdorfer

C.B

, Jacob

, Boniatti

M.M

, Dalarosa

M.G

, Falci

D.R

, Behle

T.F

, Saitovitch

, Wang

, Forrest

, Nation

R.L

, Zavascki

A.P

, and Li

. 2013. Pharmacokinetics of polymyxin B in patients on continuous venovenous haemodialysis. J Antimicrob. Chemother. 68:674–677.

10.

Zavascki

A.P.

, and Nation

R.L.

. 2017. Nephrotoxicity of polymyxins: is there any difference between colistimethate and polymyxin B? Antimicrob. Agents Chemother. 61:e02319-16.

11.

Marks

, Rifkind

R.A

, Richon

V.M

, Breslow

, Miller

, and Kelly

W.K.

. 2001. Histone deacetylases and cancer: causes and therapies. Nat. Rev. Cancer. 1:194–202.

12.

Ceccacci

, and Minucci

. 2016. Inhibition of histone deacetylases in cancer therapy: lessons from leukaemia. Br. J. Cancer. 114:605–611.

13.

W.S.

, Parmigiani

R.B

, and Marks

P.A.

. 2007. Histone deacetylase inhibitors: molecular mechanisms of action. Oncogene. 26:5541–5552.

14.

Petruccelli

L.A.

, Dupéré-Richer

, Pettersson

, Retrouvey

, Skoulikas

, and Miller Jr

W.H.

. 2011. Vorinostat induces reactive oxygen species and DNA damage in acute myeloid leukemia cells. PLoS One. 6:e20987.

15.

Namdar

, Perez

, Ngo

, and Marks

P.A.

. 2010. Selective inhibition of histone deacetylase 6 (HDAC6) induces DNA damage and sensitizes transformed cells to anticancer agents. Proc. Natl. Acad. Sci. U. S. A. 107:20003–20008.

16.

Lee

J.H.

, Choy

M.L

, Ngo

, Foster

S.S

, and Marks

P.A.

. 2010. Histone deacetylase inhibitor induces DNA damage, which normal but not transformed cells can repair. Proc. Natl. Acad. Sci. U. S. A. 107:14639–14644.

17.

Wozniak

M.B.

, Villuendas

, Bischoff

J.R

, Aparicio

C.B

, Martínez Leal

J.F

, de La Cueva

, Rodriguez

M.E

, Herreros

, Martin-Perez

, Longo

M.I

, Herrera

, Piris

M.A

, and Ortiz-Romero

P.L.

. 2010. Vorinostat interferes with the signaling transduction pathway of T-cell receptor and synergizes with phosphoinositide-3 kinase inhibitors in cutaneous T-cell lymphoma. Haematologica. 95:613–621.

18.

Prebet

, and Vey

. 2011. Vorinostat in acute myeloid leukemia and myelodysplastic syndromes. Expert. Opin. Investig. Drugs. 20:287–295.

19.

Domalaon

, De Silva

P.M

, Kumar

, Zhanel

G.G

, and Schweizer

. 2019. The anthelmintic drug niclosamide synergizes with colistin and reverses colistin resistance in Gram-Negative bacilli. Antimicrob. Agents. Chemother. 63:e02574-18.

20.

Domalaon

, Okunnu

, Zhanel

G.G

, and Schweizer

. 2019. Synergistic combinations of anthelmintic salicylanilides oxyclozanide, rafoxanide, and closantel with colistin eradicates multidrug-resistant colistin-resistant Gram-negative bacilli. J. Antibiot. 72:605–616.

21.

Cheah

S.E.

, Li

, Tsuji

B.T

, Forrest

, Bulitta

J.B

, and Nation

R.L.

. 2016. Colistin and polymyxin B dosage regimens against Acinetobacter baumannii: differences in activity and the emergence of resistance. Antimicrob. Agents. Chemother. 60:AAC.02927-15.

22.

Mota

T.M.

, Rasmussen

T.A

, Rhodes

, Tennakoon

, Dantanarayana

, Wightman

, Hagenauer

, Roney

, Spelman

, Purcell

D.F.J

, McMahon

, Hoy

J.F

, Prince

H.M

, Elliott

J.H

, and Lewin

S.R.

. 2017. No adverse safety or virological changes 2 years following vorinostat in HIV-infected individuals on antiretroviral therapy. AIDS. 31:1137–1141.

23.

Marfurt

, Chalfein

, Prayoga

, Wabiser

, Kenangalem

, Piera

K.A

, Fairlie

D.P

, Tjitra

, Anstey

N.M

, Andrews

K.T

, and Price

R.N.

. 2011. Ex vivo activity of histone deacetylase inhibitors against multidrug-resistant clinical isolates of Plasmodium falciparum and P. vivax. Antimicrob. Agents. Chemother. 55:961–966.

24.

Masadeh

M.M.

, K.H. Alzoubi

S.I.

, Al-azzam, and Al-buhairan

A.H.

. 2017. Possible involvement of ROS generation in vorinostat pretreatment induced enhancement of the antibacterial activity of ciprofloxacin. Clin. Pharmacol. 9:119–124.

25.

Wayne, P.A. 2018. Performance Standards for Antimicrobial Susceptibility Testing. 28th ed. Clinical and Laboratory Standards Institute (CLSI) Document M100. Wayne, PA: CLSI.

26.

The European Committee on Antimicrobial Susceptibility Testing (EUCAST). 2019. Breakpoint Tables for Interpretation of MICs and Zone Diameters. Version 9.0. Available at www.eucast.org (accessed February 1, 2020).

27.

Odds, F.C. 2003. Synergy, antagonism, and what the chequerboard puts between them. J. Antimicrob. Chemother. 52:1.

28.

Bremmer

D.N.

, Bauer

K.A

, Pouch

S.M

, Thomas

, Smith

, Goff

D.A

, Pancholi

, and Balada-Llasat

J.M.

. 2016. Correlation of checkerboard synergy testing with time-kill analysis and clinical outcomes of extensively drug-resistant Acinetobacter baumannii respiratory infections. Antimicrob. Agents Chemother. 60:6892–6895.

29.

Tran

T.B.

, Cheah

S.E

, Yu

H.H

, Bergen

P.J

, Nation

R.L

, Creek

D.J

, Purcell

, Forrest

, Doi

, Song

, Velkov

, and Li

. 2016. Anthelmintic closantel enhances bacterial killing of polymyxin B against multidrug-resistant Acinetobacter baumannii. J. Antibiot. 69:415–421.

30.

Huang

, Yu

, Diep

J.K

, Sharma

, Dudley

, Monteiro

, Kaye

K.S

, Pogue

J.M

, Abboud

C.S

, and Rao

G.G.

. 2017. In vitro assessment of combined polymyxin B and minocycline therapy against Klebsiella pneumoniae carbapenemase (KPC)-producing K. pneumoniae. Antimicrob. Agents Chemother. 61:e00073-00017.

31.

, Dai

, Xue

, Zhang

, Niu

, Hong

, Drlica

, and Zhao

. 2016. Dimethyl sulfoxide protects Escherichia coli from rapid antimicrobial-mediated killing. Antimicrob. Agents Chemother. 60:5054–5058.

32.

Dwyer

D.J.

, Kohanski

M.A

, Hayete

, and Collins

J.J.

. 2007. Gyrase inhibitors induce an oxidative damage cellular death pathway in Escherichia coli. Mol. Syst. Biol. 3:91.

33.

Repine

J.E.

, Pfenninger

O.W

, Talmage

D.W

, Berger

E.M

, and Pettijohn

D.E.

. 1981. Dimethyl sulfoxide prevents DNA nicking mediated by ionizing radiation or iron/hydrogen peroxide-generated hydroxyl radical. Proc. Natl. Acad. Sci. U. S. A. 78:1001–1003.

34.

Sampson

T.R.

, Liu

, Schroeder

X.R

, Kraft

C.S

, Burd

E.M

, and Weiss

D.S.

. 2012. Rapid killing of Acinetobacter baumannii by polymyxins is mediated by a hydroxyl radical death pathway. Antimicrob. Agents Chemother. 56:5642–5649.

35.

Peleg

A.Y.

, Jara

, Monga

, Eliopoulos Jr

G.M

., M.R., and Mylonakis

. 2009. Galleria mellonella as a model system to study Acinetobacter baumannii pathogenesis and therapeutics. Antimicrob. Agents Chemother. 53:2605–2609.

36.

Kohanski

M.A.

, Dwyer

D.J

, Boris

, Lawrence

C.A

, and Collins

J.J.

. 2007. A common mechanism of cellular death induced by bactericidal antibiotics. Cell. 130:797–810.

37.

Bakthavatchalam

Y.D.

, Pragasam

A.K

, Biswas

, and Veeraraghavan

. 2018. Polymyxin susceptibility testing, interpretative breakpoints and resistance mechanisms: an update. J. Glob. Antimicrob. Resist. 12:124–136.

38.

Fedoseeva

I.V.

, Pyatrikas

D.V

, Stepanov

A.V

, Fedyaeva

A.V

, Varakina

N.N

, Rusaleva

T.M

, Borovskii

G.B

, and Rikhvanov

E.G.

. 2017. The role of flavin-containing enzymes in mitochondrial membrane hyperpolarization and ROS production in respiring Saccharomyces cerevisiae cells under heat-shock conditions. Sci. Rep. 7:2586.

39.

Tacconelli

, Carrara

, Savoldi

, Harbarth

, Mendelson

, Monnet

D.L

, Pulcini

, Kahlmeter

, Kluytmans

, Carmeli

, Ouellette

, Outterson

, Patel

, Cavaleri

, Cox

E.M

, Houchens

C.R

, Grayson

M.L

, Hansen

, Singh

, Theuretzbacher

, Magrini

, and Pathogens Priority List Working Group

WHO

. 2017. Discovery, research, and development of new antibiotics: the WHO priority list of antibiotic-resistant bacteria and tuberculosis. Lancet Infect. Dis. 18:318–327.

40.

García-Fernández

, Morosini

M.I

, Marco

, Gijón

, Vergara

, Vila

, Ruiz-Garbajosa

, and Cantón

. 2015. Evaluation of the eazyplex^® SuperBug CRE system for rapid detection of carbapenemases and ESBLs in clinical Enterobacteriaceae isolates recovered at two Spanish hospitals. J. Antimicrob. Chemother. 70:1047–1050.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.04 MB