Effects of silver nanoparticles in combination with antibiotics on the resistant bacteria Acinetobacter baumannii

Abstract

Acinetobacter baumannii resistance to carbapenem antibiotics is a serious clinical challenge. As a newly developed technology, silver nanoparticles (AgNPs) show some excellent characteristics compared to older treatments, and are a candidate for combating A. baumannii infection. However, its mechanism of action remains unclear. In this study, we combined AgNPs with antibiotics to treat carbapenem-resistant A. baumannii (aba1604). Our results showed that single AgNPs completely inhibited A. baumannii growth at 2.5 μg/mL. AgNP treatment also showed synergistic effects with the antibiotics polymixin B and rifampicin, and an additive effect with tigecyline. In vivo, we found that AgNPs–antibiotic combinations led to better survival ratios in A. baumannii-infected mouse peritonitis models than that by single drug treatment. Finally, we employed different antisense RNA-targeted Escherichia coli strains to elucidate the synergistic mechanism involved in bacterial responses to AgNPs and antibiotics.

Keywords:

• Acinetobacter baumannii
• AgNPs
• synergistic
• antibiotic combination
• antisense RNA

Introduction

Drug-resistant Acinetobacter baumannii is an infectious pathogen that currently presents serious clinical challenges. A. baumannii is particularly associated with hospital-acquired infections such as pneumonia, bloodstream, abdominal, central nervous system, urinary tract, and skin and soft tissue infections.¹ A. baumannii can develop resistance against antibiotics through several mechanisms;² in particular, this bacterium is often resistant to the carbapenems.³ Increasing numbers of carbapenem-resistant A. baumannii isolates have been reported worldwide.⁴ The majority of such bacteria are extensively drug resistant, which may include resistance to carbapenems and all other antibiotics except polymyxins and tigecycline.⁵ Polymyxin B is effective against drug-resistant A. baumannii, but systemic application carries risk of toxicity, primarily kidney toxicity and neurotoxicity.^6,7 A. baumannii infection is common in patients with severe infections, and is often accompanied by other bacterial and/or fungal infections.⁸ Patients infected with resistant A. baumannii have high mortality.⁹ Therefore, there is an urgent need to find suitable therapeutic drugs for the treatment of resistant A. baumannii infections.

The ineffectiveness of synthetic antibiotics against drug-resistant bacteria has led to the reemergence of interest in silver, which has an ancient history as an antibacterial agent.^10–12 The antibacterial activity of silver nanoparticles (AgNPs) had been reported against multiple species of bacteria; for example, Escherichia coli ATCC 8739,¹³ Staphylococcus aureus ATCC1431,¹⁴ Escherichia fergusonii, and Klebsiella aerogenes ATCC 1950,¹⁵ among others. Synthesized AgNPs with capping agents, such as citrate, sodium dodecyl sulfate, and polyvinylpyrrolidone show increased antibacterial activity against S. aureus and E. coli.¹⁶ Huang et al reported antimicrobial activity against A. baumannii with the synergistic combination of chitosan acetate and AgNPs.¹⁷ Jain et al studied the interaction of AgNPs with commonly used antibiotics in Pseudomonas aeruginosa,¹⁵ while Morones-Ramirez et al demonstrated that Ag⁺ treatment sensitized Gram-negative bacteria to the Gram-positive-specific antibiotic vancomycin, both in vitro and in vivo.¹⁸ However, the synergistic antimicrobial activity of antibiotics combined with citrate-capped AgNPs has yet to be studied.

In the present study, we investigated the synergistic combinatorial effects of antibiotics with AgNPs against drug-resistant A. baumannii obtained from clinical patients both in vivo and in vitro. We also investigated the possible mechanisms of this synergistic effect.

Materials and methods

Materials

Trisodium citrate, silver nitrate (AgNO₃), and sodium borohydride (NaBH₄) were used for the synthesis of AgNPs. Rifampicin, tigecyline, polymyxin B (PMB), mucin, dimethyl sulfoxide, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), isopropyl-β-D-thiogalactoside, penicillin–streptomycin, and trypsin–ethylenediaminetetraacetic acid were purchased from Sigma Aldrich (St Louis, MO, USA). A549 cells and HL-7702 cells were purchased from the Type Culture Collection of the Chinese Academy of Sciences (Shanghai, People’s Republic of China). Roswell Park Memorial Institute-1640 medium was purchased from Thermo Fisher Scientific (Waltham, MA, USA) and fetal bovine serum was purchased from HyClone (Logan, UT, USA). No ethical committee approval was required for this set of experiments because the experiments were performed on commercially available cell lines and were considered exempt from full review by the ethics Committee at Shanghai Jiao Tong University.

All animal procedures were approved by the Institutional Animal Care and Use Committee at Shanghai Jiao Tong University. All the animal studies were performed according to the Guiding Principles for the Care and Use of Laboratory Animals according to the Regulations of the People’s Republic of China for Administration of Laboratory Animals. All animal procedures were approved by the Animal Ethics Committee of Shanghai Jiao Tong University. C57BL/6 mice were purchased from Slac Laboratory Animal Co., Ltd. (Shanghai, People’s Republic of China).

Synthesis and characterization of AgNPs

In a three-necked round-bottomed flask, 20 mL trisodium citrate (1%) and 75 mL ultrapure water were mixed for 15 minutes at 70°C. To the solution, 1.5 mL of silver nitrate solution (1%) was added; NaBH₄ (1%) was then added, followed by rapid mixing. This mixed solution was heated for 60 minutes, cooled to room temperature, and water was added to a volume of 100 mL.

To characterize the morphology of the synthesized AgNPs, transmission electron microscopy (TEM) analysis was performed using a Tecnai G2 Spirit 120 kV TEM instrument (0.23 nm resolution) (FEI Company, Hillsboro, OR, USA).

AgNPs were further characterized by scanning the absorbance spectra in 300–500 nm range of wavelength with a multifunction full wavelength microplate analyzer (BioTek Co., Winooski, VT, USA).

A Malvern Zetasizer Nano-ZS instrument (Malvern, Louis, USA) was used to characterize the zeta potential of the nanoparticles in the solution. Data were obtained and analyzed using Zetasizer software (Malvern, Louis, USA).

Determination of minimum inhibitory concentration and fractional inhibitory concentration

The bacterial strain A. baumannii (aba 1604; Fudan University Huashan Hospital, Shanghai, People’s Republic of China) was used as a model test strain to determine the antibacterial activity of AgNPs. Various concentrations of AgNPs were incubated with 4×10⁵ bacteria in Luria Bertani (LB) broth medium in 96-well round-bottomed plates. Bacteria were harvested at the indicated time points and the optical density of the samples was assayed at 600 nm. All samples were plated in triplicate, and values were averaged from three independent trials. The resistant A. baumannii strain (aba1604) was made from clinical patients and following institutional ethical guidelines that were reviewed and approved by the ethics committee at the Huashan hospital clinical ethics committee, Fudan University. Consent from clinical patients was not deemed necessary by the Shanghai Laiya centre for Biopharmaceutical R & D who obtained these strains for antibacterial drug research and collected these samples, as the sample collection was part of normal patient care.

To evaluate the antibacterial activity of AgNPs in combination with antibiotics, a two-dimensional microdilution assay was used.¹⁹ Assays were carried out in LB broth growth medium. Minimum inhibitory concentration (MIC) for each of the antibiotics was first estimated, and the fractional inhibitory concentration (FIC) of a combination of antibiotics and AgNPs was subsequently determined by the checkerboard microtitration method in a 96-well microtiter plate. Antibiotics and AgNPs were diluted to the following concentrations (2MIC, 1MIC, 1/2MIC, 1/4MIC, 1/8MIC, 1/16MIC, and 1/32MIC) in the two-dimensional microdilution assay.

The plates were incubated at 37°C for 18 hours, and results were assayed by measuring the optical density (OD)₆₀₀. The combined antibiotic effect of agents A and B (where A is either AgNO₃ or AgNPs, and B is one of three antibiotic agents) was calculated as follows:

$\begin{array}{l}\text{The FIC index:}\\ =\frac{\text{MIC}(\text{A in combination with B})}{\text{MIC}(\text{A alone})}\\ +\frac{\text{MIC}(\text{B in combination with A})}{\text{MIC}(\text{B alone})}\end{array}$(1)

FIC index values above 4.0 indicate antagonistic effects, values between 0.5 and 4.0 indicate additive effects, and values lower than 0.5 indicate synergistic effects.²⁰

Cytotoxicity assay

To determine the cytotoxic activity of the AgNPs on mammalian cells, A549 cells and HL-7702 cells (1×10⁴ cells/mL) were grown in Roswell Park Memorial Institute-1640 medium containing 5% fetal bovine serum in a 96-well plate at 37°C in an atmosphere of 5% CO₂ for 24 hours. Cells were treated with AgNPs, AgNO₃, or control solutions at concentrations ranging from 0.625 to 10 μg/mL for another 24 hours. To determine the viability, MTT (at a concentration of 0.1 mg/mL) was added to the wells and incubated for 4 hours at 37°C and 5% CO₂ to allow cell growth.²¹ In metabolically active cells, MTT was reduced to an insoluble, dark purple formazan. The purple formazan was then dissolved in dimethyl sulfoxide. The absorbance was measured at 570 nm using a multifunction full wavelength microplate analyzer and readings were compared from untreated cells. The OD values were used to sort out the percentage of viable cells by using the following formula:

$\begin{array}{c}\text{Percentage}\\ \text{of viability}\end{array}=\frac{\text{OD value of experimental sample}}{\begin{array}{c}\text{OD value of experimental}\\ \text{control (untreated)}\end{array}}\times 100$(2)

Assay for antimicrobial activity in vivo

Minimum lethal dose of AgNO₃ or AgNPs in mice

Six-week-old male C57BL/6 mice (body weight ~20 g) were used for all animal experiments. Mice were housed in a temperature- and humidity-controlled environment, and had free access to food and water. Ten mice per group were given intraperitoneal injections of 100 μL total volume. Mice were treated as follows: no treatment, and 10, 20, 40, and 80 mg/kg AgNPs and AgNO₃. Injected animals were observed for 3 days.

Determination of minimum lethal dose of A. baumannii for peritonitis mouse model

Serial dilutions of A. baumannii ranging from 1×10⁷ to 1×10¹¹ CFU, in 500 μL sterile saline supplemented with 8% mucin, were injected into the peritoneal cavity of mice.¹⁸ Animals were observed for 2 days for determination of the survival rate.

Survival assays

Mice received intraperitoneal injections of the minimum lethal dose (MLD) of A. baumannii, with a total volume of 500 μL with 8% mucin. After 1 hour, ten mice in each group received a 100 μL intraperitoneal injection of either vehicle phosphate buffer saline (PBS) or one of the different antibacterial treatments. Mice were observed for 2 days to evaluate the survival rate.

Bacterial colonization assays

Surviving mice were euthanized and dissected, and their kidneys and lungs were collected. These organs were ground under aseptic conditions, and the homogenates were dissolved in sterilized saline water. These organ homogenates were then cultivated on LB plates at 37°C for 24 hours.

Cytokine profiling

Cytokine concentrations in the mouse plasma were measured at the indicated time after infection by standard enzyme-linked immunosorbent assay kits following the manufacturer’s instructions (Elabscience Biotechnology Co., Ltd, Wuhan, People’s Republic of China).

Antisense RNA models for detecting the synergistic mechanism of AgNPs and antibiotic combinations

To investigate the pathways involved in the bacterial response to AgNPs, we conducted a series of experiments in which the impacts of AgNPs and AgNO₃ on different E. coli antisense RNA-induced gene-silencing strains were examined.^22–24 The various gene-silenced strains were arrayed in microwell plates, and then screened to determine how their sensitivity to each of the different Ag formulations compared to the parent strain. The gene-silenced E. coli strains were treated with isopropyl-β-d-thiogalactoside at appropriate concentrations (Table 1), and portions of the culture were transferred into 96-well plates. Sublethal concentrations of AgNPs, AgNO₃, rifampicin, tigecyline, and PMB were added to the gene-silenced bacteria in 96-well plates. Finally, the plates were incubated at 37°C for 16 hours and shaken at 80 rpm. Absorbance was measured at 600 nm using a multifunction full wavelength microplate analyzer, and OD values were used to calculate the inhibition ratio (I) by using the following formula:

$\begin{array}{l}\mathrm{I}(\%)\\ =\frac{\text{OD value of control}(\text{untreated})\text{OD value of sample}}{\text{OD value of control}(\text{untreated})}\times 100\end{array}$(3)

$\mathrm{\Delta }\mathrm{I}(\%)={\mathrm{I}}_{\text{antisense RNA-induced gene-silencing strains}}-{\mathrm{I}}_{E.coli\text{DH5α/pHN}678}$(4)

Table 1 The optimal concentration of IPTG for antisense RNA strains

Genes	Target genes function	IPTG (μmol/L)
ligA	DNA biosynthesis	800
dnaB	Replicative DNA helicase, DNA biosynthesis	20
rpsR	Small subunit ribosomal protein S18, ribosome	10
rpsA	Small subunit ribosomal protein S1, ribosome	40
rpsL	Large subunit ribosomal protein L7/L12, ribosome	10
rplC	Large subunit ribosomal protein L3, ribosome	20
rplT	Large subunit ribosomal protein L20, ribosome	40
rplS	Large subunit ribosomal protein L19, ribosome	40
rpmA	Large subunit ribosomal protein L27, ribosome	60
murA	UDP-N-acetylglucosamine 1-carboxyvinyltransferase, peptidoglycan biosynthesis	20
murB	UDP-N-acetylenolpyruvoylglucosamine reductase, peptidoglycan biosynthesis	40
murG	Peptidoglycan biosynthesis	40
murE	UDP-N-acetylmuramoylalanyl-D-glutamate–2,6-diaminopimelate ligase, peptidoglycan biosynthesis	40
leuS	Leucyl-tRNA synthetase, aminoacyl-tRNA biosynthesis	40
tufA	Elongation factor Tu, protein biosynthesis	40
rpoD	Principal σ factor, RNA biosynthesis	40
fabI	Enoyl-[acyl-carrier protein] reductase I, fatty acid biosynthesis	40
kdsA	Lipopolysaccharides biosynthesis	1,600
kdsB	Lipopolysaccharides biosynthesis	1,600
lpxC	Lipopolysaccharides biosynthesis	1,600
lepB	Signal peptidase, protein export	40
mutL	DNA mismatch repair protein MutL, mismatch repair	40
menD	Ubiquinone and other terpenoid-quinone biosynthesis	40

Abbreviations: IPTG, isopropyl-β-d-thiogalactoside; tRNA, transfer RNA; UDP, uridine diphosphate.

Statistical analysis

Each assay was repeated three times. Data are presented as mean ± standard deviation, unless otherwise noted. Comparisons between multiple groups were made using one-way analysis of variance (ANOVA) and all analyses were performed using SPSS 21.0 statistical software (21.0; IBM Corporation, Armonk, NY, USA). The threshold for statistical significance was set at P<0.05.

Results

Characterization of AgNPs

Based on the image in Figure 1A, AgNPs stabilized by citrate had good dispersion. The particle size distribution was shown by counting AgNPs particle numbers based on TEM images and the histogram of particle size distribution was in Figure 1B. AgNPs were 5–12 nm in diameter, with an average size of 8.4 nm. The nanoparticles were found to be stable for over 6 months, even at 37°C. The zeta potential of the synthesized AgNPs is summarized in Figure 1C. The ultraviolet–visible (UV–vis) spectra of the solution samples are reported in Figure 1D. A single strong peak was observed at 392 nm, which indicates the synthesis of spherical nanoparticles. In practice, dispersion was stable if the zeta potential was higher than 30 mV or less than −30 mV. Supporting the earlier statement, we observed that AgNPs dispersed in water were highly stable with a zeta potential value of −44.5 mV.

Figure 1 Appearance and physicochemical characteristics of AgNPs.

Notes: (A) TEM of AgNPs. (B) Size distribution of AgNPs based on TEM images. (C) Zeta potential analysis of AgNPs. (D) UV–visible absorption spectroscopy showed the maximum absorbance at 392 nm for AgNPs.

Abbreviations: AgNPs, silver nanoparticles; OD, optical density; TEM, transmission electron microscopy; UV, ultraviolet.

Antibacterial activities of AgNP combination treatments

Combination antibiotic therapy is a strategy often employed in the treatment of multiple drug resistance (MDR) A. baumannii. Because PMB, rifampicin, and tigecycline are all commonly used against MDR A. baumannii in combination with other antibiotics,^25–27 we selected these three antibiotics to evaluate potential combinatorial effects with AgNPs. It was seen that AgNPs displayed potent antimicrobial activity against A. baumannii, with an MIC of approximately 2.5 μg/mL, similar to that of AgNO₃. Typical drug MICs, summarized in Table 2, were 0.25 μg/mL for PMB, 3.12 μg/mL for rifampicin, and 3.12 μg/mL for tigecycline.

Table 2 FIC index of combinations among silver and antibiotics against Acinetobacter baumannii

Compounds	FIC of AgNPs	FIC of AgNO3	MIC of antibiotics (μg/mL)
Polymyxin B	0.19	0.19	0.25
Rifampicin	0.38	0.38	3.12
Tigecycline	0.75	0.75	3.12
AgNPs	–	–	2.5
AgNO3	–	–	2.5

Note: Data are presented as mean ± SD, unless otherwise specified.

Abbreviations: FIC, fractional inhibitory concentration; MIC, minimum inhibitory concentration; AgNPs, silver nanoparticles; SD, standard deviation.

The FIC of AgNPs and the various antibiotic combinations were investigated and are summarized in Table 2. These experiments showed that PMB and rifampicin acted synergistically (P<0.5) with AgNPs and AgNO₃, while tigecycline did not show synergy (P>0.5) with either AgNPs or AgNO₃.

Cytotoxicity of AgNPs in vitro

Many investigations have reported on the inhibitory effects of AgNPs on cells. For example, Beer et al found that AgNPs inhibited the proliferation of A549 cells in a dose-dependent manner,²⁸ whereas Foldbjerg et al reported that AgNPs induced increase in reactive oxygen species (ROS) level in A549 cells.²⁹ Here, we used the method described by Foldbjerg et al to evaluate the cytotoxicity of AgNPs.²⁹

As shown in Figure 2, high concentrations of AgNO₃ significantly affected cell growth. By comparison, exposure to AgNPs at a higher concentration of 10 μg/mL did not exhibit significant cytotoxicity in A549 and HL-7702 cells. These results demonstrate that the cytotoxicity of AgNPs is lower than that of AgNO₃.

Figure 2 Relative survival of A549 and HL-7720 cells exposed to AgNPs.

Notes: Relative survival of cells as affected by different doses AgNPs or AgNO₃. (A) MTT assay results confirmed the in vitro cytotoxicity of AgNPs and AgNO₃ against A549 cells. (B) Effects of AgNPs or AgNO₃ on HL-7720 cell growth. Results are shown as the mean ± SD of three independent experiments. *P<0.05, **P<0.01 and ***P<0.001 vs control.

Abbreviations: AgNPs, silver nanoparticles; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; SD, standard deviation.

Effects of AgNPs on antimicrobial activity in vivo

The acute toxicity of AgNPs and AgNO₃ were measured in vivo to establish the median lethal dose (LD₅₀) (Figure 3). The LD₅₀ of AgNPs was determined to be between 20 and 40 mg/kg, and the LD₅₀ of AgNO₃ was also found to be between 20 and 40 mg/kg. These LD₅₀ values for AgNO₃ are similar to those reported in an earlier toxicity study.¹⁸

Figure 3 Toxicity of AgNPs and AgNO₃ in mice.

Notes: (A) Survival of mice given the following treatments: control, 10, 20, 40, and 80 mg/kg AgNO₃. (B) Survival of mice given the following treatments: control, 10, 20, 40, or 80 mg/kg AgNPs. Survival assays were performed with ten mice per group.

Abbreviations: AgNPs, silver nanoparticles; h, hours.

Infection was established in mice through intraperitoneal delivery of 5×10⁹ A. baumannii cells suspended in an aqueous solution containing 8% mucin. Within 24 hours after the time of injection, all infected mice had died. Thus, A. baumannii was used at this same concentration of 5×10⁹ cells for all subsequent experiments.

The PMB and AgNPs combination and the PMB and AgNO₃ combination showed the same synergistic antibiotic effects that we had earlier observed in vivo (Figure 4). We observed the mice living condition for 1 week, and calculated 2 days survival rate. A mixture of AgNO₃ and PMB (3 mg/kg and 10 μg/kg, respectively) resulted in a survival rate of 40%. When either of these two compounds was given alone as a single-dose treatment, survival rates were 0%. By comparison, even a low-dosage mixture of AgNPs and PMB (2 mg/kg and 10 μg/kg, respectively) resulted in a high survival rate of 60%.

Figure 4 Survival of mice given AgNPs or AgNO₃ with PMB in a peritonitis infection model.

Notes: (A) PMB treatment concentration in peritonitis infection model. (B) AgNO₃ and PMB treatment concentration in peritonitis infection model. (C) AgNPs and PMB treatment concentration peritonitis infection model.

Abbreviations: AgNPs, silver nanoparticles; PMB, polymyxin B; h, hours.

Two days after administration in mice, we dissected a portion of mice for bacterial colonization assay. Representative images of bacterial growth are presented in Figure 5, which show that AgNO₃ and AgNPs enhanced the action of PMB against A. baumannii in peritonitis infection model. Figure 5A and B shows that there are plenty of bacteria present in both the kidney and lungs when animals are treated with PMB alone at a dose of 250 μg/kg. With the addition of AgNO₃ (3 mg/kg, 18 μM) or AgNPs (2 mg/kg, 18 μM), no bacteria were detected in the kidneys or lungs. When we reduced the PMB dose to 50 μg/kg, we found that the kidney and lung tissues contained only a small amount of bacteria. Further reduction of the PMB dose to 10 μg/kg was less effective, leaving a substantial bacterial burden in the kidney and lungs. Moreover, we also tested the blood and ascites of infected mice and found that they did not contain bacteria under the combined administration of PMB (50 μg/kg) with either AgNO₃ (3 mg/kg, 18 μM) or AgNPs (2 mg/kg, 18 μM) (Figure 5C and D). These results indicate not only that AgNO₃ can enhance the antibacterial activity of antibiotics but also that AgNPs possess antibacterial capability in vivo.

Figure 5 Bacterial burdens in Acinetobacter baumannii-infected mice after treatment with PMB combined with AgNO₃ or AgNPs.

Notes: Bacterial burdens in Acinetobacter baumannii-infected mice after treatment with PMB (250 μg/kg) (a); the infected mice after treatment with PMB combined with AgNO₃ (3 mg/kg, 18 μM) (b1: 250 μg/kg; b2: 50 μg/kg; b3: 10 μg/kg), and with AgNPs (2 mg/kg, 18 μM) (c1–c3). (A) kidney; (B) lung; (C) blood; (D) ascitic fluid.

Abbreviations: AgNPs, silver nanoparticles; PMB, polymyxin B.

To analyze whether AgNO₃ or AgNPs combined with antibiotics regulate inflammation during A. baumannii infection, we assessed proinflammatory cytokines in the mouse plasma using enzyme-linked immunosorbent assay. The levels of tumor necrosis factor alpha (TNF-α) and interleukin (IL)-6 decreased significantly in the mouse plasma of treated mice compared to those of model mice at 18 hours post infection (Figure 6). AgNO₃ and AgNPs could enhance the action of PMB against A. baumannii in vivo.

Figure 6 AgNO₃ or AgNPs combined with antibiotics-modulated Acinetobacter baumannii-induced inflammatory reaction.

Notes: ELISA was used to measure IL-6 (A) and TNF-α (B) in the mouse plasma 18 hours post infection. Results are shown as the mean ± SD of three independent experiments. ***P<0.001 vs control; ^#P<0.05 and ^##P<0.01 vs infection model; ^$P<0.05 vs PMB only group. We obtained P-values by a Students’ t-test.

Abbreviations: ELISA, enzyme-linked immunosorbent assay; IL, interleukin; PMB, polymyxin B; TNF-α, tumor necrosis factor alpha; SD, standard deviation.

Mechanisms for enhanced antimicrobial activity of AgNPs in combination with antibiotics

To better understand the mechanisms supporting the synergistic effects of AgNPs combined with antibiotics, we used antisense RNA-induced gene silencing to silence the expression of a number of genes in E. coli (Figure 7). The difference in inhibition rate between a set of gene-silenced E. coli strains and the control E. coli DH5α/pHN678 strain is represented by color in our heat map analysis; dark red represents the most sensitive strains, whereas dark blue represents the least sensitive strains, relative to the parent strain. We found that silencing of rpoD, kdsA, kdsB, lpxC, and mutL resulted in sensitivity to both AgNO₃ and AgNPs. Silencing of rpsR, rpsL, rpsA, murB, murA, leuS, and dnaB led only to sensitivity to AgNO₃, which indicates that AgNPs act at a higher selectivity than AgNO₃. In addition, the rpoD-silenced strain was sensitive to rifampicin, while the silencing of kdsA, kdsB, and lpxC increased sensitivity to PMB.

Figure 7 Heat map analysis of sensitivity to AgNPs and AgNO₃ in gene-silenced Escherichia coli strains relative to the control E. coli DH5α/pHN678 strain.

Notes: Each unit represents the difference in inhibition rate between a single antisense RNA-induced gene silencing and the control E. coli DH5α/pHN678 strain. Red units represent the most sensitive strains for compounds, and blue represents strains that grow similarly to the control E. coli DH5α/pHN678 strain.

Abbreviations: AgNPs, silver nanoparticles; TIGE, tigecycline; RIF, rifampicin; PMB, polymyxin B.

Discussion

A. baumannii has been proved to be resistant to many kinds of antibiotics, which is suitable for genetic exchange. Fournier et al reported an 86 kb genomic region naming AbaR1 resistance island in AYE which had 45 resistance genes in the MDR isolate.³⁰ The key resistance genes were those coding for AmpC, VEB-1, and OXA-10 beta-lactamases, tetracycline efflux pumps, and various aminoglycoside-modifying enzymes. Moreover, Fournier et al found that 17 out of the 22 clinical A. baumannii isolates showed an original ATPase ORF. These 17 isolates contained eleven isolates that are resistant to several antibiotic families, including β-lactams, and six other isolates susceptible to β-lactams. AgNPs could damage the membrane potential, prevent ATP production, increase the level of ROS, and damage the membrane lipids as well as DNA, which demonstrated that AgNPs have broad-range antibacterial properties, including A. baumannii.³¹

It has been reported that citrate-capped AgNPs are less toxic to mammalian cells and show increased antimicrobial activity against S. aureus and P. aeruginosa.³² Our results support these findings, and further build from them to explore the toxicity of AgNPs to A549 and HL-7702 cells, and to describe the effects of AgNPs on drug-resistant A. baumannii. Our study demonstrates that the cytotoxicity of AgNPs is lower than that of AgNO₃.

The combination of AgNPs and either PMB or rifampicin showed strong synergistic antimicrobial effects. Moreover, the pairing of AgNPs and PMB showed an enhanced effect against A. baumannii in vivo, which suggests the possibility of a clinical application for this combination therapy.³³

IL-6 and TNF-α are cytokines that have been shown to play an important role in the host immune response against intracellular pathogens in murine models. Smani et al demonstrated that A. baumannii induced the release of TNF-α and IL-6 and increased the Ca²⁺ influx.³⁴ The levels of TNF-α and IL-6 decreased significantly in the mouse plasma of treated mice compared to those of model mice, which further proved that inhibiting proinflammatory signals could be protective during A. baumannii infections. Sarkar et al reported that AgNPs could modulate human macrophage responses to Mycobacterium tuberculosis.³⁵ A. baumannii is also highly correlated with the host immune status.³⁵ We have verified the role of AgNPs on A. baumannii. Hence, we hypothesized that AgNPs could modulate human macrophage responses to A. baumannii, as the mode of M. tuberculosis.

E. coli is one of the most representative model organisms in experimental biology and medical study; there are a large number of experimental studies on the drug mechanisms and targets using E. coli.^22–24 On the one hand, A. baumannii is a common clinical pathogen, which has received increased attention.³⁶ A. baumannii and E. coli are Gram-negative bacteria, and they have similar cell structure.³⁷ So we probe the mechanism of synergistic effect of AgNPs and antibiotics combination against A. baumannii using E. coli as a model system.

The probable role of PMB in such drug synergy is its rapid permeabilization of the outer cell membrane, allowing enhanced penetration by AgNPs. Polymyxin B can displace Mg²⁺ or Ca²⁺, and also binds to the Lipid A component of lipopolysaccharide (LPS), resulting in changes to the outer membranes of bacteria. Our gene-silencing experiments also suggested that LpxC plays a role in sensitivity toward AgNPs and AgNO₃. Lin et al found that an LpxC inhibitor blocked the ability of bacteria to activate the sepsis cascade, enhanced opsonophagocytic killing of A. baumannii, and protected mice from lethal infection.³⁸ Moreover, the potential contributions of PMB and AgNPs both involve Lipid A of LPS, a convergence which supports their synergistic effects.

KDO 2-keto-3-deoxyoctanoic acid (KDO) plays an essential role in LPS biosynthesis, and may serve a universal role in group 2 capsule biosynthesis by linking the polysaccharide region to the lipid domain.³⁹ The kdsA gene encodes the protein KDO 8-phosphate synthetase, which catalyzes the first step of the KDO synthetic reaction.⁴⁰ The kdsB gene encodes CMP-KDO synthetase, which is essential to the LPS biosynthesis pathway.⁴¹ Our gene-silencing results suggested that kdsA and kdsB are involved in bacterial sensitivity toward AgNPs and AgNO₃, as well as PMB.

Similarly, we observed that silencing of RpoD, which is involved in promoter localization and plays a crucial role in transcription initiation,⁴² increased the sensitivity of E. coli toward AgNPs and AgNO₃. Most of the factors investigated belong to the σ70 family, and all bacteria express one or more σ70 factors. The σ70 factor sequence is highly conserved and plays an important role in bacterial growth. Rifampicin has a molecular mechanism of action that involves inhibition of DNA-dependent RNA polymerase.⁴³ In E. coli, this enzyme is a complex oligomer comprising four subunits: α, β, β′, and σ, encoded, respectively, by rpoA, rpoB, rpoC, and rpoD, and their disruption interferes in the transcription process.⁴³ The potential mechanisms of both rifampicin and AgNPs involve effects on DNA-dependent RNA polymerase, which is the evidence that supports their synergistic effects (Figure 8).

Figure 8 Proposed mechanisms of the combination of AgNPs/Ag⁺ with antibiotics against G⁺ negative bacterium.

Notes: AgNPs/Ag⁺ can enhance PMB-induced damage of the membrane lipids. AgNPs/Ag⁺ and RIF also may bind intracellular proteins and RNA polymerase, upon entering the cytosol.

Abbreviations: AgNPs, silver nanoparticles; PMB, polymyxin B; RIF, rifampicin; LPS, lipopolysaccharide; CMP-KDO, cytosine monophosphate-2-keto-3-deoxyoctanoic acid.

Silver and silver-containing compounds have recently drawn increasing interest as antimicrobial agents for treating bacterial infections. AgNPs showed synergy of inhibiting P. aeruginosa biofilms when combined with sub-MIC levels of aztreonam.⁴⁴ Combination of AgNPs with ceftazidime also showed a synergy to inhibit P. aeruginosa.⁴⁵ AgNPs prepared as described by Tiwari et al exhibited tremendous antibacterial activity against a carbapenem-resistant strain of A. baumannii.³ This required an efficient treatment regimen, and the combination of rifampin with imipenem had been evaluated in clinical infections caused by a highly imipenem-resistant A. baumannii strain.⁴⁶ Yoon et al showed that the combination of PMB plus imipenem was as effective as PMB plus rifampin against a carbapenem-resistant strain of A. bau-mannii.²⁶ Therefore, drug treatment with newer antimicrobials or antimicrobial combinations has become increasingly important to eradicate these infections. According to our study, the combination of AgNPs with antibiotics could be an effective solution to the problem of carbapenem-resistant strains of A. baumannii, potentially at lower and less-toxic doses than what is now typically used clinically.

Investigators have previously suggested that combination drug therapy could be an effective tool to prevent the emergence of bacterial resistance, especially for patients infected with Gram-negative bacteria that have developed resistance to a single therapy.⁴⁷ Our study showed the synergistic effects of combining AgNPs and PMB or AgNPs and rifampicin against drug-resistant A. baumannii isolated from clinical patients. Considering the lower toxicity of AgNPs compared to other treatment options, these drug combinations have potential as useful tools for the clinic.

Disclosure

The authors report no conflicts of interest in this work.

Acknowledgments

The authors are very grateful to Professor Zhu Demei from Fudan University Huashan Hospital, who provided the resistant A. baumannii strain (aba1604) from clinical patients.

This work was funded by The National Major Scientific and Technological Special Project for “Significant New Drugs Development” (2012ZX09301002-003-007).