Synergistic Efficacy of Aedes aegypti Antimicrobial Peptide Cecropin A2 and Tetracycline against Pseudomonas aeruginosa

ABSTRACT

The increasing prevalence of antibiotic resistance has created an urgent need for alternative drugs with new mechanisms of action. Antimicrobial peptides (AMPs) are promising candidates that could address the spread of multidrug-resistant bacteria, either alone or in combination with conventional antibiotics. We studied the antimicrobial efficacy and bactericidal mechanism of cecropin A2, a 36-residue α-helical cationic peptide derived from Aedes aegypti cecropin A, focusing on the common pathogen Pseudomonas aeruginosa. The peptide showed little hemolytic activity and toxicity toward mammalian cells, and the MICs against most clinical P. aeruginosa isolates were 32 to 64 μg/ml, and its MICs versus other Gram-negative bacteria were 2 to 32 μg/ml. Importantly, cecropin A2 demonstrated synergistic activity against P. aeruginosa when combined with tetracycline, reducing the MICs of both agents by 8-fold. The combination was also effective in vivo in the P. aeruginosa/Galleria mellonella model (P < 0.001). We found that cecropin A2 bound to P. aeruginosa lipopolysaccharides, permeabilized the membrane, and interacted with the bacterial genomic DNA, thus facilitating the translocation of tetracycline into the cytoplasm. In summary, the combination of cecropin A2 and tetracycline demonstrated synergistic antibacterial activity against P. aeruginosa in vitro and in vivo, offering an alternative approach for the treatment of P. aeruginosa infections.

INTRODUCTION

Pseudomonas aeruginosa is a Gram-negative opportunistic bacterial pathogen that causes life-threatening infections with high rates of mortality (1–3). It is associated with nosocomial pneumonia, wound infections, bacteremia, and sepsis among patients with various underlying diseases, including cystic fibrosis, HIV/AIDS, and cancer (4). Antibiotic therapy is challenging in this setting because clinical strains of P. aeruginosa often show extensive intrinsic resistance to a wide range of antimicrobial agents, including tetracyclines, β-lactams, aminoglycosides, and fluoroquinolones (5). This intrinsic resistance has generally been attributed to the low membrane permeability of P. aeruginosa, up to 100 times lower than that of Escherichia coli (6). P. aeruginosa can also withstand antibiotics that attack the outer membrane through the efficient deployment of transmembrane efflux pumps, preventing contact between the antibiotics and their intracellular targets (7–11).

The ubiquitous and relentless clinical challenge of drug resistance has created a pressing need for new antipseudomonal agents. Antimicrobial peptides (AMPs) have been proposed as candidates because of their novel mechanisms of action. AMPs are produced by diverse organisms, including microbes, plants, insects, and mammals, and may replace or complement traditional antibiotics in settings where antibiotics alone have lost efficacy (12–14). Most AMPs interact with bacterial membranes and cause membrane depolarization, electrolyte leakage, and ultimately cell death (15), but others have cytoplasmic targets (16). AMPs that target membranes are less likely to be rendered ineffective by resistance, because redesigning the microbial membrane architecture cannot be achieved by simple mutations (13, 17). Furthermore, combinations of AMPs with traditional antibiotics may achieve enhanced or synergistic activity against bacterial pathogens (18–20).

In this study, we screened a library of 68 insect-derived AMPs (21, 22) for new antibacterial compounds and found that cecropin A2 from the mosquito Aedes aegypti showed potent antimicrobial effects against clinical isolates of P. aeruginosa. Cecropin A2 is 36 amino acids in length (NCBI sequence no. AAL85581.1) and is derived from the mature cecropin A peptide by proteolytic cleavage (Fig. 1). The MIC of cecropin A2 against clinical isolates of P. aeruginosa was high, and so we tested combinations of this AMP with conventional antibiotics to see whether synergic interactions could increase its potency. These tests were carried out in vitro and using larvae of the greater wax moth (Galleria mellonella) as an in vivo infection model. We also investigated the abilities of cecropin A2 to bind to lipopolysaccharides, permeate the bacterial membrane, and interact with genomic DNA to gain insight into its mechanisms of action.

FIG 1.

Amino acid sequence, helical wheel diagram, and predicted structures of cecropin A2. Hydrophobic amino acid residues are shown in green; hydrophilic residues are shown in red and gray.

RESULTS

Initial antibacterial activity screen.

A total of 68 insect-derived AMPs were synthesized by solid-phase synthesis and purified by reversed-phase chromatography by GenScript (Piscataway, NJ), and the purity of the peptides was confirmed using liquid chromatography-mass spectrometry (LC/MS). In preliminary experiments, we screened these AMPs using a broth assay to determine whether the peptides were effective against the Gram-negative bacterium P. aeruginosa strain PA14 at a final concentration of 50 μg/ml (see Table S1 in the supplemental material). Two of the peptides, cecropin A2 and BR021, showed antibacterial activity, with MICs of 32 μg/ml (8.7 μM) and 64 μg/ml (7.6 μM), respectively. Cecropin A2 is a helical cationic AMP derived from the mosquito Aedes aegypti (Fig. 1), while BR021 is one of the coleoptericins from beetles (23).

Therefore, we tested the activity of these peptides against multidrug-resistant clinical strains of P. aeruginosa and observed MICs ranging from 32 μg/ml to more than 64 μg/ml against most of the strains (Table 1). The antimicrobial activities of cecropin A2 and BR021 were also evaluated against a panel of more diverse Gram-negative bacteria (Klebsiella pneumoniae, Enterobacter aerogenes, and Acinetobacter baumannii), as well as against the Gram-positive bacteria Staphylococcus aureus and Enterococcus faecium. Cecropin A2 and BR021 both showed activity against Gram-negative bacteria, particularly cecropin A2, with MICs of 4 μg/ml against K. pneumoniae and 2 μg/ml against A. baumannii, but neither peptide was active against the Gram-positive bacteria, even at the maximum tested concentration of 64 μg/ml (Table 2).

TABLE 1.

MICs of antimicrobial agents against Pseudomonas aeruginosa strains

P. aeruginosa straina	MICs (μg/ml) for:
	Cecropin A2	BR021	Polymyxin B	Tetracyclineb
PA14	32	64	1	64
PA103	64	>64	2	16
PAO1	64	>64	1	ND
PA 2326	64	>64	32	32
PA 1026	32	64	0.5	64
PA 1016	>64	>64	512	8
0054	64	64	1	>32
0064	>64	32	2	32
0090	64	32	1	16
0092	32	32	2	>64
0094	64	32	2	>32
0095	>64	32	2	32
0100	32	64	2	>64
0103	64	64	1	>64
0105	64	64	1	>32
0108	64	32	1	>32
0110	64	32	2	>32
0111	64	>64	1	>32

^aP. aeruginosa clinical isolates obtained from FDA-CDC Antimicrobial Resistance Bank.

^bMICs for tetracycline were obtained from the FDA-CDC Antimicrobial Resistance Bank (58). ND, no detection.

TABLE 2.

MICs of antibacterial peptides against bacterial pathogens

Strain	MICs for (μg/ml):
	Cecropin A2	BR021	Polymyxin B
P. aeruginosa (PA14)	32	64	1
K. pneumoniae (ATCC 77326)	4	64	4
A. baumannii (ATCC 17978)	2	32	2
S. aureus (MW2)	>64	>64	64
E. faecium (E007)	>64	>64	>64
E. aerogenes (EAE 2625)	32	64	8
E. coli (OP50)	4	NDa	ND

^aND, no detection.

Cytotoxicity and hemolytic activity.

Synthetic AMPs have been studied for nearly 3 decades, but few have advanced to the clinic (23, 24), partly reflecting their toxicity toward human cells (14). The hemolytic activities of cecropin A2 and BR021 were therefore tested using a human erythrocyte lysis assay, and cytotoxicity was evaluated against human HepG2 cells (Fig. 2). Neither peptide induced significant hemolysis compared with that in untreated control cells, whereas almost complete hemolysis occurred in the positive-control sample treated with the detergent Triton X-100 (Fig. 2A). Furthermore, cecropin A2 showed no cytotoxicity toward HepG2 cells even at the maximum tested concentration of 64 μg/ml, whereas BR021 showed concentration-dependent in vitro toxicity. HepG2 cells treated with up to 16 μg/ml BR021 were almost 100% viable, but this fell to 39.50% when the cells were exposed to 64 μg/ml BR021. Given the higher MIC and cytotoxicity of BR021 compared with those of cecropin A2, BR021 was selected for further analysis. The hemolytic and cytotoxic 50% lethal dose (LD₅₀) values for cecropin A2 were both >60 μg/ml, suggesting this AMP is ideal for further development as a clinical antimicrobial drug.

FIG 2.

Hemolytic activity and cytotoxicity of the antimicrobial peptides. (A) Human erythrocytes were treated with serial dilutions of Triton X-100 (0.002 to 1%) or AMPs (0.125 to 64 μg/ml). (B) Cytotoxicity of cecropin A2 and BR021 against HepG2 cells. The viability of HepG2 cells was measured after treatment with serially diluted concentrations (0.12 to 64 μg/ml) of the AMPs.

Synergistic efficacy of cecropin A2 combined with antibiotics.

To determine whether cecropin A2 works synergistically with clinical antibiotics, we used the checkerboard assay to measure the efficacy of combinations of cecropin A2 with the antibiotics ampicillin, ceftazidime, ciprofloxacin, erythromycin, gentamicin, vancomycin, nalidixic acid, neomycin, nisin, tetracycline, doxycycline, and polymyxin B against P. aeruginosa strain PA14 (see Table S2). We observed synergistic activity (fractional inhibitory concentration [FIC] < 0.5) for combinations of cecropin A2 with tetracycline or doxycycline, and in the case with tetracycline, the MICs for these compounds were reduced by 8-fold. The absolute values decreased from 32 to 4 μg/ml for cecropin A2, and from 64 to 8 μg/ml for tetracycline (Fig. 3A and B). We also tested the synergic combination of cecropin A2 and tetracycline against a panel of multidrug-resistant clinical strains of P. aeruginosa (Fig. 3C). All the FIC indices were lower than 0.5, confirming that cecropin A2 and tetracycline act synergistically against pathogenic strains of P. aeruginosa despite their overt antibiotic resistance.

FIG 3.

Synergy between cecropin A2 and tetracycline. Heat plot (A) and isobologram (B) showing synergistic growth inhibition of P. aeruginosa strain PA14 by the combination of cecropin A plus tetracycline. (C) Synergetic effects of cecropin A2 plus tetracycline against clinical isolates of P. aeruginosa.

Effects of cecropin A2 and tetracycline on bacterial viability.

The impacts of cecropin A2 and/or tetracycline on P. aeruginosa viability and membrane integrity were tested using the vital stain SYTO 9, which stains all living cells green, and propidium iodide (PI), which specifically penetrates cells with damaged membranes and stains them red. As shown in Fig. 4A, ≥99.85% of the bacterial cells treated with the control compound, dimethyl sulfoxide [DMSO], were stained green, confirming they were viable and featured intact membranes. By contrast, ∼74.3% of the bacterial cells treated with cecropin A2 stained red, suggesting that their membranes were damaged and thus permeable to PI, and even the ∼25.7% green-stained cells were often dim and indistinct, suggesting they were structurally weakened (Fig. 4B). These findings indicate that cecropin A2 can disrupt the integrity of bacterial cell membranes to allow the uptake of PI. Exposure to tetracycline alone did not result in a significant change in viability of membrane permeability compared with that of the bacteria treated with DMSO (Fig. 4C). On the contrary, exposure to both cecropin A2 and tetracycline resulted in the uptake of PI by almost all of the cells (∼98.98%), suggesting a significant increase in cell permeability and hence the synergistic activity of the cecropin A2-tetracycline combination (Fig. 4D).

FIG 4.

Confocal microscopy showing P. aeruginosa PA14 cells exposed to cecropin A2 and/or tetracycline. Bacteria were grown at 37°C in the presence of DMSO (A), 64 μg/ml cecropin A2 (B), 128 μg/ml tetracycline (C), or a combination of 64 μg/ml cecropin A2 and 128 μg/ml tetracycline (D). Bacteria were stained with SYTO 9, which shows live cells as green, and propidium iodide, which stains dead cells red, and cells were visualized by confocal microscopy.

P. aeruginosa time-killing kinetics.

The live/dead assay showed that cecropin A2 was lethal to the bacteria with or without tetracycline. Therefore, we conducted a time-course killing experiment and examined the effects of cecropin A2 and/or tetracycline when presented at double and quadruple the nominal MICs. The treatment with 4× MIC of cecropin A2 reduced the number of bacteria by two orders of magnitude during the first 60 min, and no live bacteria remained after 4 h, whereas the 4× MIC tetracycline treatment gradually reduced the cell count by two orders of magnitude over 4 h. The combination of cecropin A2 with tetracycline compounds showed a synergistic activity. More specifically, both the 4× MIC and 2× MIC combinations achieved rapid and sustained bactericidal activity and eliminated bacteria within 3 h (Fig. 5).

FIG 5.

Growth curves of P. aeruginosa strain PA14 in the presence of cecropin A2 and/or tetracycline at 4× MIC (A) and 2× MIC (B) compared with that of a DMSO control. MICs were 32 μg/ml for cecropin A2, 64 μg/ml for tetracycline, and 4 μg/ml and 8 μg/ml for cecropin A2 plus tetracycline, respectively.

In vivo efficacy of cecropin A2 with or without tetracycline.

Next, we evaluated the efficacy of cecropin A2 and/or tetracycline in vivo using the G. mellonella-P. aeruginosa infection model (25–27). As previously reported (25), the inoculation of G. mellonella larvae with P. aeruginosa strain PA14 at 37°C killed the larvae in a dose-dependent manner, with an optimal injection of 10² CFU/ml killing all larvae within 24 h (data not shown). The administration of cecropin A2 protected the larvae and prolonged their survival in a dose-dependent manner (Fig. 6A). More specifically, 90% of the larvae survived the lethal challenge for 24 h when administered 12.5 mg/kg of body weight cecropin A2 (P < 0.001), and 40% survived when administered the lower dose of 6.25 mg/kg cecropin A2 (P = 0.029), but the control larvae injected with phosphate-buffered saline (PBS) were not protected. After 96 h, no larvae survived if they had been injected with either 6.25 mg/kg cecropin A2 or 12.5 mg/kg tetracycline, but 100% of the larvae injected simultaneously with both agents survived for more than 96 h (P < 0.001), clearly demonstrating synergistic protection in vivo (Fig. 6B).

FIG 6.

Effect of cecropin A2 and/or tetracycline on the survival of G. mellonella larvae infected with P. aeruginosa strain PA14. (A) Survival curves for P. aeruginosa strain PA14 following treatment with different concentrations of cecropin A2 or PBS alone. Cecropin A2 at the concentrations of 6.25 mg/kg of body weight and 12.5 mg/kg significantly prolonged the survival of G. mellonella larvae (P < 0.05 for both groups). (B) Survival of P. aeruginosa PA14 following treatment with cecropin A2 with or without tetracycline. The combination of both reagents significantly prolonged larval survival (P < 0.001).

Effect of cecropin A2 with or without tetracycline on the bacterial membrane.

The membrane permeability of P. aeruginosa strain PA14 was tested following exposure to cecropin A2 with or without tetracycline using the nucleic acid stain Sytox green, which only penetrates cells with compromised membranes. As shown in Fig. 7A, cecropin A2 caused a concentration-dependent and time-dependent increase in Sytox green fluorescence, indicating that the AMP can penetrate and kill bacteria by disrupting membrane integrity. By contrast, tetracycline alone (64 μg/ml) had no effect on membrane permeability (Fig. 7B). The presence of both cecropin A2 and tetracycline had a much more potent effect than that of each compound alone, and exceeded the effect of the control antibiotic, polymyxin B.

FIG 7.

Effect of cecropin A2 with or without tetracycline on the P. aeruginosa cell membrane. (A) Time course of membrane permeabilization by cecropin A2 by measuring the increase in Sytox green fluorescence (excitation λ, 485 nm; emission λ, 530 nm). Serial dilutions were tested at concentrations of 4 to 64 μg/ml. (B) Permeabilization of P. aeruginosa PA14 by cecropin A2 (32 μg/ml), tetracycline (64 μg/ml), a combination (32 μg/ml cecropin A2 and 64 μg/ml tetracycline), or polymyxin B (32 μg/ml). Cells treated with DMSO (5% final concentration) were used as the negative control. (C) Cecropin A2 enhanced the uptake of Sytox green into P. aeruginosa cells in a concentration-dependent manner in the presence of a constant concentration of tetracycline (128 μg/ml).

Because tetracycline preferentially inhibits the biosynthesis of envelope proteins, we speculated that cecropin A2 may permeabilize the bacterial membrane and facilitate the uptake of tetracycline, enhancing its bactericidal efficiency. In confirmatory experiments, we measured Sytox green fluorescence with a constant concentration of tetracycline and various concentrations of cecropin A2. The fluorescence intensity of the cells exposed to tetracycline alone was almost negligible, whereas the fluorescence intensity increased with increasing concentrations of cecropin A2 (Fig. 7C). The highest concentrations of cecropin A2 tested (32 and 64 μg/ml) caused rapid increases in fluorescence, reaching the maximum at 60 min, which is similar to that in the time-killing curve described above (Fig. 5).

Interaction between cecropin A2 and bacterial LPS or DNA.

Lipopolysaccharide (LPS) is the major outer membrane constituent of Gram-negative bacteria, and so we used a fluorescence-based displacement assay with BODIPY TR cadaverine to investigate whether cecropin A2 binds to LPS (Fig. 8A). Cecropin A2 behaved in a manner similar to that of polymyxin B, which is known to bind LPS and was used as a positive control (28, 29). The LPS-binding activity of cecropin A2 was dose dependent, and the peptide was able to displace BODIPY TR cadaverine, confirming that cecropin A2 binds to P. aeruginosa LPS with a high affinity.

FIG 8.

The binding activities of cecropin A2 to bacterial LPS and DNA. (A) The affinities of cecropin A2 and control reagents to P. aeruginosa LPS were determined using the BODIPY TR cadaverine displacement method. (B) Interaction of cecropin A2 and control reagents with P. aeruginosa genomic DNA. To highlight the effects of high concentrations of cecropin A2 on its DNA-binding activity, we ran multiple gels that demonstrated a decreasing intensity of DNA bands from cells that were treated with a high concentration of the peptide (representative bands from two different gels with complete pictures of the gels included in Fig. S1 in the supplemental material). (C) The DNA-binding ability of cecropin A2. DNA bands were quantified with ImageJ software and DNA binding (%) was defined as the ratio of nonmigrating DNA to total genomic DNA.

AMPs often target multiple cell components, such as DNA. We used DP-Bind (http://lcg.rit.albany.edu/dp-bind) to predict the specific DNA-binding position of cecropin A2. We found that cecropin A2 had DNA-binding ability, and its DNA-binding residues were G1, G2, K4, K5, G7, G12, G14, K15, R16, V17, F18, N19, G30, K32, A33, R35, and K36. To confirm this, we evaluated the ability of cecropin A2 to bind the bacterial DNA. The results of the DNA-binding assay of cecropin A2 shown in Fig. 8B and C confirmed the concentration-dependent displacement of genomic DNA bands, suggesting that the genomic DNA of P. aeruginosa strain PA14 was tightly bound by high doses of cecropin A2 (64 and 128 μM). However, LL-37, a known DNA-binding AMP, showed good DNA-interacting ability at a low concentration (12.5 μM). Accordingly, we could cautiously conclude that cecropin A2 at high doses had a strong affinity for DNA.

DISCUSSION

New antibacterial compounds with diverse mechanisms of action are required to address the challenge of multidrug-resistant bacterial infections. This is particularly important in the case of Gram-negative bacteria such as P. aeruginosa, which possess seemingly impermeable protective membranes (9). One proposed strategy is the exploitation of natural AMPs (23). Cecropin A2, a derivative of cecropin A from the mosquito Aedes aegypti, contains 10 cationic and 16 hydrophobic amino acids and protects the model nematode host Caenorhabditis elegans from infection with A. baumannii (22). Here we characterized the in vitro activity of cecropin A2 against several Gram-negative bacterial pathogens and found that the peptide acts synergistically with tetracycline against the common pathogen P. aeruginosa both in vitro and in vivo. In addition to binding to LPS and disrupting the bacterial membrane, cecropin A2 at a high dose also interacts with bacterial genomic DNA. This is advantageous because multiple targets reduce the likelihood that resistance will emerge as a result of the selection of individual mutations.

Combinations of antibiotics are widely used for the management of P. aeruginosa infections, staphylococcal infections, leprosy, tuberculosis, and malaria (30, 31). Some AMPs are known to exert synergistic antimicrobial effects with conventional antibiotics, resulting in greater efficacies in vitro and in vivo, including combinations with lactoferricin-derived peptides (32) and polymyxin B (33). The combination of polymyxin B and doxycycline reduced the MIC of the peptide against carbapenemase-producing K. pneumoniae by at least 4-fold (33). We found that cecropin A2 displayed in vitro synergy with tetracycline, which interferes with bacterial protein synthesis (34–36), leading to an 8-fold reduction in the MIC of each agent. This synergistic effect was observed when testing clinical P. aeruginosa strains despite their overt antibiotic resistance, suggesting that cecropin A2 combined with tetracycline offers an attractive option for the treatment of recalcitrant P. aeruginosa infections.

Our findings support the hypothesis that combinatorial antimicrobial drugs are beneficial because AMPs penetrate cell membranes as a direct antibacterial mechanism of action, resulting in cytoplasmic leakage and cell death (14, 16, 22). Therefore, we reasoned that cecropin A2 should disrupt the P. aeruginosa cell envelope and promote the uptake of tetracycline, which would inhibit protein synthesis and induce cell death. Indeed, the combination of cecropin A2 and tetracycline rapidly permeabilized the membrane and was more efficacious than polymyxin B or cecropin A2 alone, which could be ascribed to the properties of tetracycline, such as inhibiting protein synthesis and chelating divalent cations (37). We also found that cecropin A2 binds to LPS, the main outer membrane component of Gram-negative bacteria. Taken together, these results agree with those from previous studies indicating that AMPs can disrupt membrane integrity and encourage the uptake of antibiotics, allowing the antibiotics to interact with their intracellular targets and exert a bactericidal effect (38–40).

The bactericidal activity of AMPs often involves the disruption of membrane integrity, but some AMPs target additional intracellular components, such as bacterial DNA (41–43). Therefore, we tested the affinity of cecropin A2 for DNA extracted from P. aeruginosa strain PA14. Accordingly, we found that cecropin A2 induced a concentration-dependent DNA-binding ability during electrophoresis, indicating that the peptide may kill bacterial cells by first permeabilizing the membrane and then targeting intracellular components, including the genomic DNA (44). Like other members of the cecropin family, the linear cationic α-helical peptide cecropin A2 is therefore likely to interact with the negatively charged components of the bacterial membrane to disrupt its integrity (16, 45) before entering the cell and interacting directly with the DNA. When combined with tetracycline, the ability of cecropin A2 to permeabilize the cell facilitates the internalization of the antibiotic, and thus potentiating its ability to inhibit protein synthesis. There is little opportunity for bacterial pathogens to develop resistance to attacks against the envelope, because this would require the complete modification of the cell membrane, which would in turn require changes to multiple biochemical pathways (42, 46, 47). The ability of cecropin A2 and/or other insect-derived AMPs to disrupt the bacterial envelope could therefore restore the susceptibility of even multidrug-resistant pathogens to conventional antibiotics that are presently becoming less effective (23).

The development of new antimicrobial agents requires the confirmation of their efficacy in vivo. We found that cecropin A2 prolonged the survival of the model host, G. mellonella, in a dose-dependent manner, and that the combination of cecropin A2 and tetracycline was significantly more effective than monotherapy with either reagent. Antibiotic combinations can protect larvae from a broad spectrum of pathogens, including S. aureus, P. aeruginosa, and A. baumannii (26, 48, 49). However, few studies have tested the efficacy of AMP combinations in vivo (32), and the G. mellonella model could thus provide a facile alternative for the study of AMP-antibiotic combinations.

In conclusion, our results demonstrated that cecropin A2 has potent activities against clinical strains of P. aeruginosa in vitro and in vivo, which are achieved by binding to LPS, penetrating the bacterial membrane, and interacting with the bacterial DNA. When combined with tetracycline, cecropin A2 facilitated the intracellular uptake of the antibiotic and enhanced the bactericidal activity of both agents against P. aeruginosa but showed no toxicity toward human cells at the concentrations we tested. Although further experiments are necessary to explore the clinical potential of cecropin A2 in more detail, the cecropin A2-tetracycline combination offers a promising alternative approach for the treatment of infections caused by recalcitrant Gram-negative pathogens.

MATERIALS AND METHODS

Bacterial strains and culture conditions.

The bacterial strains we used (Tables 1 and 2) were cultured in lysogeny broth (LB) or on LB agar at 37°C. In addition to clinical isolates of P. aeruginosa, we tested the following bacterial strains: Enterococcus faecium (E007), Klebsiella pneumoniae (ATCC 77326), Acinetobacter baumannii (ATCC 17978), Staphylococcus aureus (MW2), and Enterobacter aerogenes (EAE 2625). The P. aeruginosa clinical isolates were obtained from the FDA-CDC Antimicrobial Resistance Bank (Atlanta, GA) (50). The other bacterial strains were selected from an existing strain collection maintained in our laboratory and were originally obtained from the American Type Culture Collection (Manassas, VA) or established laboratory strains (21, 22, 51, 52).

Inhibition of bacterial growth.

MICs were determined using the broth microdilution method in 96-well microtiter plates with an assay volume of 100 μl (53). Briefly, 2-fold serial dilutions of the test compounds spanning the concentration range of 0.125 to 64 μg/ml were prepared, and 50-μl aliquots were transferred to the microtiter plate wells. Overnight bacterial cultures were diluted to a final concentration of ∼5 × 10⁵ CFU/ml in non-cation-adjusted Mueller-Hinton (MH) broth (Difco; BD, Franklin Lakes, NJ), and 50-μl aliquots were mixed with the contents of each well. After further overnight incubation at 37°C, we measured the absorbance at 600 nm to determine cell growth as an indicator of antimicrobial activity. The MIC was defined as the minimum concentration at which no bacterial growth was observed in each well.

Human blood hemolysis.

The hemolytic activities of the compounds were determined as previously described with assay volumes of 100 μl (54). Microtiter plates were prepared with each well containing 50 μl of the serially diluted test compounds. Human erythrocytes (Rockland Immunochemicals, Limerick, PA) were suspended in PBS to a final concentration of 2%, and 50-μl aliquots were transferred to the plate wells, mixed gently, and incubated at 37°C for 1 h. The plates were then centrifuged at 500 × g for 5 min, and 50 μl of the supernatant was transferred to a new plate to determine the absorbance at 540 nm. Each assay was carried out in triplicates.

Cytotoxicity assay.

Cytotoxicity was measured as previously described (55) with some modifications. The human hepatocellular liver carcinoma cell line HepG2 (ATCC HB 8065; ATCC, Manassas, VA) was cultivated in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin at 37°C in a 5% CO₂ humidified atmosphere. The cell culture reagents were supplied by Gibco (Thermo Fisher Scientific, Waltham, MA). The cells were harvested and resuspended in DMEM at a concentration of 2.5 ×10⁵ cells/ml. We transferred 100-μl aliquots (2.5 ×10⁴ cells) to each well of a microtiter plate and incubated them as above. The test compounds were serially diluted in antibiotic-free DMEM containing 10% FBS and were transferred to the cell monolayers. The plates were incubated at 37°C in a 5% CO₂ humidified atmosphere for 24 h, and for the final 4 h, we added 10 μl 2-(4-iodophenyl)-3-(4-nitrophenyl)-5-(2,4-disulfophenyl)-2H-tetrazolium (WST-1) solution (Roche, Mannheim, Germany) to each well. The reduction of WST-1 was confirmed by measuring the absorbance at 450 nm. The assay was carried out in triplicates, and the proportions of surviving cells were calculated by comparison with a control treated with DMSO.

Checkerboard assay.

The synergy between pairs of compounds was tested using a checkerboard assay (56). Bacterial suspensions were prepared with a final concentration of ∼5 × 10⁵ CFU/ml. The two compounds for combinatorial testing were arrayed as serial dilutions, vertically for one compound and horizontally for the other, in the same 96-well microplate. The assay was carried out in triplicates as described above for the measurement of individual MICs. The FIC indices were calculated according to the following formulae: FIC = X/MIC_X, and FIC index = FIC_{cecropin A2} + FIC_antibiotic, where X is the lowest inhibitory concentration of the drug in the presence of the codrug. An FIC index of less than 0.5 indicates synergy between the test compounds (30, 57).

Time-killing curve.

A time-kill assay was carried out in non-cation-adjusted MH broth according to the protocol from the Clinical and Laboratory Standards Institute (Wayne, PA). The assay was carried out in triplicates. An overnight bacterial culture was diluted in MH broth to an absorbance of 0.05 at 600 nm. Different concentrations of test compounds were then added and the culture was incubated while shaking at 37°C. Aliquots periodically sampled from the cultures were serially diluted, plated on MH agar, and incubated overnight at 37°C. The colonies were then counted to determine viability.

Live/dead assay.

P. aeruginosa strain PA14 was cultured at 37°C until the cells reached the logarithmic growth phase. The cells were harvested and washed, and then a suspension comprising ∼10¹⁰ CFU/ml was treated with an equal volume of cecropin A2 (64 μg/ml), tetracycline (128 μg/ml), a combination (64 μg/ml cecropin A2 plus 128 μg/ml tetracycline), or the negative control (5% DMSO). After incubating at 37°C for 4 h, the cultures were harvested and stained with 0.3% SYTO 9 and PI at ratio of 1:1 from the Live/Dead BacLight bacterial viability and counting kit (Thermo Fisher Scientific). After mixing, the samples were mounted on microscope slides and observed using a Nikon C1si confocal microscope (Nikon, Melville, NY) with diode lasers of 488 and 561 nm.

P. aeruginosa-G. mellonella infection model.

Ten G. mellonella larvae were inoculated with 10 μl of the bacterial suspension at various concentrations, as previously reported (25, 58). The larvae were considered dead if they failed to respond to touch. Infected larvae were treated with cecropin A2, tetracycline, a combination of both, or a placebo. All reagents were injected through the proleg using different injection sites for each delivery (starting with the last left proleg), and there was a 30-min interval between injections. Untreated larvae and larvae injected with 10 μl sterile PBS were used as controls. Deaths were scored during an incubation period of 144 h at 37°C using two biological replicates. Survival curves were plotted using the Kaplan-Meier method, and differences in survival were calculated using the log-rank test in OriginPro v9 (OriginLab, Northampton, MA). A P value of < 0.05 was considered to be statistically significant.

Bacterial cell membrane permeabilization assay.

The permeability of bacterial membranes was determined by measuring the uptake of Sytox green (Thermo Fisher Scientific) into bacterial cells in 96-well plates as previously described (21, 22). The assay was carried out in triplicates. Briefly, overnight cultures of P. aeruginosa strain PA14 were harvested, washed twice, and resuspended in PBS before adding Sytox green to a final concentration of 5 μM and incubating the cells in the dark for 30 min. Next, 50 μl of serially diluted compounds in PBS was added to 50 μl of the bacterial suspension. Sytox green fluorescence (excitation wavelength, 485 nm; emission wavelength, 530 nm) was measured every 5 min for 2 h at room temperature using a SpectraMax M2 Multi-mode microplate reader (Molecular Devices, Sunnyvale, CA).

LPS binding assay.

The affinities of the test compounds for LPS were determined using the BODIPY TR cadaverine displacement assay (28). Polymyxin B, a peptide antibiotic known to bind and neutralize LPS, was used as the positive control (29), and meropenem was used as the negative control. Stock solutions of P. aeruginosa LPS (5 mg/ml) and BODIPY TR cadaverine (500 μM) were prepared in 50 mM Tris buffer (pH 7.4). The two stocks were mixed and diluted in Tris buffer to yield final concentrations of 50 μg/ml LPS and 5 μM BODIPY TR cadaverine. Then, we mixed 50 μl of the BODIPY TR cadaverine:LPS mixture and 50 μl of the peptide serial dilutions in the wells of black-bottomed 96-well plates. Fluorescence was measured at 25°C at an excitation wavelength of 580 nm and an emission wavelength of 620 nm.

DNA-binding assay.

P. aeruginosa strain PA14 genomic DNA was extracted using the bacterial genomic DNA extraction kit (Sigma-Aldrich Biotechnology, St. Louis, MO), and 20 μl of the purified genomic DNA was incubated with cecropin A2, LL-37, or tetracycline for 1 h at room temperature. We then added 2 μl of native loading buffer, and a 10-μl aliquot was separated by 1.5% agarose gel electrophoresis in 0.5× Tris-borate-EDTA buffer (45 mM Tris-borate, 1 mM EDTA, pH 8.0) to detect DNA binding.