Mechanisms driving the antibacterial and antibiofilm properties of Hp1404 and its analogue peptides against multidrug-resistant Pseudomonas aeruginosa

Hp1404, identified from the venom of the scorpion Heterometrus petersii, displays antimicrobial activity with cytotoxicity. Several synthetic peptides were designed based on the parent peptide Hp1404 to reduce cytotoxicity and improve activity (deletion of glycine and phenylalanine, substitution with leucine and lysine). The analogue peptides generated comprised 12 amino acids and displayed amphipathic α-helical structures, with higher hydrophobic moments and net positive charge than those of the Hp1404. The analogues showed less hemolytic and toxic effects toward mammalian cells than the Hp1404, especially Hp1404-T1e, which exhibited particularly potent antibacterial and antibiofilm activities against multidrug-resistant Pseudomonas aeruginosa (MRPA) strains. The analogue peptide Hp1404-T1e was more stable against salt and trypsin than the Hp1404. Hp1404’s mechanism of action involves binding to lipopolysaccharide (LPS), thereby killing bacteria through membrane disruption. Hp1404-T1e kills bacteria more rapidly than Hp1404 and not only seems to bind more strongly to LPS but may also be able to enter bacterial cells and interact with their DNA. Additionally, Hp1404-T1e can effectively kill bacteria in vivo. The results of this study indicate that Hp1404-T1e not only displays antimicrobial activity, but is also functional in physiological conditions, confirming its potential use as an effective therapeutic agent against MRPA.

Since their discovery, antibiotics have been widely used as therapeutic agents against infectious diseases. However, their indiscriminate use has led to the rapid emergence of increasingly resistant strains. The best example is penicillin, since its introduction for clinical use in the 1940s, numerous bacterial strains have developed resistance to it 1 . Nowadays, the list of antibiotic-resistant bacteria includes vancomycin-resistant Staphylococcus aureus (VRSA), methicillin-resistant Staphylococcus aureus (MRSA), multidrug-resistant Pseudomonas aeruginosa (MRPA) 2 , multidrug-resistant Acinetobacter baumannii (MRAB), among others. Recently, carbapenem-resistant Enterobacteriaceae (CRE) have also become an issue. As the emergence of multidrug-resistant microbes has grown into a global problem, significant research efforts are currently being channelled into the development of alternative antibiotics. Antimicrobial peptides (AMPs) are promising candidates, as their use does not lead to the development of resistance 3,4 .
AMPs are important components of the innate immune system; they have been isolated from a variety of organisms such as bacteria, insects 5 , reptiles 6 , mammals 5 , and plants 7 , while they can also be chemically

Results
Synthetic peptide design and characterization. Hp1404 was isolated from the venomous gland of the scorpion Heterometrus petersii 20 . The analogue peptides tested in this study were designed by truncating G and F residues, while K and L residues were added 22 . The amino acid sequence, observed and calculated molecular weight, hydrophobic moment, and retention time of each peptide are summarized in Table 1. The hydrophobic moment values of Hp1404-T1, Hp1404-T1a, Hp1404-T1b, Hp1404-T1c, Hp1404-T1d, and Hp1404-T1e were 0.699, 0.761, 0.755, 0.784, 0.768, and 0.831, respectively, thus higher than the ones recorded for the parent peptide (0.677). The net charge values of the analogue peptides ranged from +1 to +6; all but Hp1404-T1 (+1) were more positively charged than Hp1404 (+1). Figure 1 displays helical wheel diagrams and three-dimensional structure projections for the parent peptide and its analogues. Hp1404 and its analogue peptides show amphipathic α-helix conformations with hydrophobic and hydrophilic residues located on opposite sides of the α-helix. Compared to Hp1404, analogue peptides appeared to display stronger amphipathic α-helical structures. Successful peptide synthesis was confirmed by RP-HPLC using C18 columns (Fig. S1), while the molecular weight of each peptide was determined through MALDI-TOF/MS (Fig. S2).
Antibacterial activity. We studied the antibacterial activity of the peptides against P. aeruginosa. As shown in Table 2, the parent peptide exhibited broad-spectrum antimicrobial activity, as it was effective against pseudomonas aeruginosa strains with minimum inhibitory concentration (MIC) values ranging from 3.13 to 12.5 µM. In contrast, Hp1404-T1, Hp1404-T1a, and Hp1404-T1b had no activity at 25 µM against any of them. Three other analogues, Hp1404-T1c, Hp1404-T1d, and Hp1404-T1e, showed antimicrobial activities when tested against P. aeruginosa strains. In the case of ciprofloxacin, which is an antibiotic for the treatment of P. aeruginosa 23 , activity was not observed at a concentration of 400 µM. The peptides that were active against the drug-susceptible strain also proved effective against the resistant strains. Notably, Hp1404-T1e displayed MIC values lower than those of the parent peptide and the other analogues against both drug-susceptible and multidrug-resistant strains of P. aeruginosa. We conclude that Hp1404-T1e has potent antibacterial activity against P. aeruginosa.
Hemolytic and cytotoxic activities of synthetic peptides. To investigate the toxicity of the peptides, their ability to lyse mouse red blood cells (RBCs) was examined (Fig. 2a). Melittin (bee venom) 24 , which was used as a positive control, caused complete hemolysis at 12.5 M. In the case of Hp1404, hemolysis rates were 2, 11, 55, 92, and 100%, at the respective concentrations of 12.5, 25, 50, 100, and 200 µM. In contrast, all analogue peptides did not induce substantial hemolysis, even at the higher tested concentration (200 µM). Cytotoxicity of the peptides was assessed by investigating their toxic effects on HaCaT cells using the MTT method. Results are shown in Fig. 2b. Melittin, which was again used as a positive control, had a strong cytotoxic effect 25 ; the survival rate was 11.29% at 1.56 µM, while no living cells were detected at 3.13 µM. Among the peptides, analogues displayed lower cytotoxicity compared to the parent peptide; the survival rates at 50 µM for Hp1404, Hp1404-T1c, Hp1404-T1d, and Hp1404-T1e were 30, 51, 33, and 100%, respectively. Based on the aforementioned results, all subsequent experiments were performed using Hp1404, Hp1404-T1c, Hp1404-T1d, and Hp1404-T1e.
Bacterial killing kinetics. The effect of peptides on P. aeruginosa was investigated by studying time-kill kinetics (Fig. S3). In general, the peptides kill the bacteria at 4 hours while Hp1404-T1e kills 85% bacteria at 2 hours; Hp1404-T1e effectively killed all bacteria at 3 hours.
Salt sensitivity. Salt environments have a negative impact on antimicrobial peptides. Therefore, the antibacterial activities of peptides were investigated following the addition of physiological concentrations of different salts for the sensitivity assay. Table 3 displays the MIC values obtained for each peptide tested in the presence of salt. Hp1404-T1e showed stable antibacterial activities, with MIC values ranging between 3.13 µM and 12.5 µM for most of the conditions tested. In comparison, Hp1404, Hp1404-T1c, and Hp1404-T1d displayed lower activities. The ability to resist to salt provides the Hp1404-T1e peptide with an advantage when it comes to selecting potential antimicrobial agents to be used in physiological environments.
Trypsin stability. To investigate the protease stability of peptides, these were incubated with trypsin for 1 or 2 h. The products were separated using RP-HPLC. Supplementary Figure 5 displays the profiles obtained for Hp1404, Hp1404-T1c, Hp1404-T1d, and Hp1404-T1e, respectively, at all time points (0, 1, and 2 h). Stability comparisons were based on the percentages of remaining peptide determined by the HPLC peak area corresponding to the native peptide. The parent peptide is much more susceptible to proteolysis compared to its analogues, as the area of the peak corresponding to native Hp1404 is reduced faster. After 2 h of trypsin treatment, only 50% of Hp1404 remains. In contrast, more than 80% of the analogue peptides had escaped proteolytic cleavage at the same time point (Fig. 2c).

Inhibition and visualization of biofilms.
To investigate the antibiofilm activity of the peptides, biofilm formation was examined in the presence of each peptide. P. aeruginosa ATCC 27853 selected to perform the biofilm formation related experiments 26 , Hp1404, Hp1404-T1c, Hp1404-T1d, and Hp1404-T1e inhibited biofilm formation at concentrations of 25, 25, 12.5, and 12.5 µM, respectively. We also examined the antibiofilm activity of the peptides against multidrug-resistant Pseudomonas aeruginosa strains. Minimal biofilm inhibitory concentration (MBIC) values ranged from 6.25 to >50 µM; Hp1404-T1e had the strongest antibiofilm activity among the tested peptides (Table 4).
Finally, we used LIVE/DEAD BacLight staining of bacteria to visualize and examine the viability of cells in biofilms of both standard (ATCC 27853) and multidrug-resistant (1034) strains of P. aeruginosa after treatment with peptides at 0.5× and 1× MBIC. At both concentrations, the peptides significantly impacted biofilm formation, although the lowest concentration was unable to completely remove the biofilm. As expected, ciprofloxacin did not inhibit biofilm formation by the MRPA strain, even at 400 µM (Fig. 3).

LPS binding assay.
To investigate whether lipopolysaccharides (LPS) bind to peptides, we examined their antimicrobial activities at 1× and 2× MIC against P. aeruginosa, in the presence of increasing LPS concentrations. As expected, LPS alone do not affect the proliferation of P. aeruginosa. In the case of Hp104, the absorbance at 600 nm remained the same, irrespective of LPS concentration, indicating that LPS do not affect the ability of the parent peptide to suppress bacterial growth. In contrast, there is a clear negative effect of LPS on the antimicrobial action of the peptides, beginning at LPS concentrations of 100 µg/mL (Hp1404-T1c), 50 µg/mL (Hp1404-T1d), and 25 µg/mL (Hp1404-T1e). The effect becomes more pronounced as the LPS concentration increases further (Fig. 4a). Antibacterial activities of Hp1404, Hp1404-T1c, Hp1404-T1d, and Hp1404-T1e were significantly lower after the addition of LPS at concentrations of 50, 25, 50, and 6.25 µg/mL, respectively (Fig. 4b).
We used CD spectroscopy to investigate the secondary structures of the parent peptide and its analogues in the presence of LPS from P. aeruginosa. As shown in Fig. 4c, the spectra of all peptides displayed a positive peak at 192 nm, and two negative peaks (at 208-210 nm and 220-225 nm). This pattern indicates the adoption of a helical conformation, especially in the case of Hp1404. Our results suggest that the peptides interact with LPS to produce an α-helix structure. Secondary structures of Hp1404 and its analogues. To determine the secondary structure of peptides in membrane-mimicking environments, circular dichroism (CD) spectra were obtained from peptides dissolved at a final concentration of 40 µM in the following solutions: 10 mM sodium phosphate buffer, pH 7.2 (mimics the aqueous environment); 30 mM sodium dodecyl sulfate (SDS, mimics the negatively charged prokaryotic membrane environment), and 50% trifluoroethanol (TFE, mimics the hydrophobic environment of the microbial membranes). The spectra are shown in Fig. 5a. In the sodium phosphate solution, the parent peptide and its analogues displayed a positive peak at 212 nm and a negative peak at 195 nm, indicating a random coil conformation.
In the other two solutions, all peptides displayed spectra corresponding to α-helices, with two negative peaks at about 208 nm and 222 nm, and a positive peak at 190 nm, suggesting that the peptides adopted an α-helical conformation in the membrane-mimicking environments.
CD spectra analyses and liposome aggregation. We also investigated the peptide structure in the presence of large unilamellar vesicles (LUVs) composed of PE:PG (phosphatidylethanolamine:polyethylene glycol) to mimic bacterial membranes, PC:CH:SM (phosphatidylcholine:cholesterol:sphingomyelin) to mimic mammalian   Table 3. Antibacterial activities of synthetic peptides against P. aeruginosa 27853 under conditions mimicking physiological concentrations of salt.
cell membranes, or PC:CH (phosphatidylcholine:cholesterol) to mimic erythrocyte membranes (Fig. 5b). Hp1404 adopted α-helical structures in the presence of PE:PG and PC:CH liposomes, whereas it displayed a random coil conformation when mixed with PC:CH:SM liposomes. In contrast, all three analogues tested (Hp1404-T1c, Hp1404-T1d, and Hp1404-T1e) exhibited random coil conformations in the presence of all liposome types ( Fig. 5c). Finally, we assessed the ability of the four peptides to induce liposome aggregation. Peptides were mixed with liposome suspensions (prepared from PE:PG, PC:CH, or PC:CH:SM mixtures) at various peptide/liposome (P/L) ratios (Fig. 5b). In all cases, higher P/L ratios led to larger increases in turbidity. PC:CH and PC:CH:SM LUVs were much more resistant to aggregation caused by any of the four peptides compared to PE:PG liposomes. Moreover, Hp1404 was much more effective in promoting aggregation of PE:PG liposomes compared to its analogues. At the higher tested P/L ratio (0.2), the parent peptide increased the turbidity by almost 25%, whereas the respective percentages for the other three peptides were well below 20%.
Calcein leakage assay. The leakage assay is an experiment carried out to assess pore formation and membrane perturbation. We monitored the release of calcein from negatively charged PC:PG (9:1) LUVs, which mimic a bacterial membrane. At a peptide concentration of 4 µM, P/L ratio of 0.4, Hp1404 induced a release of about 70% of the calcein, while the calcein release provoked by the analogue peptides stayed below 30% (Fig. 6a). Therefore, Hp1404 exhibited significantly higher leakage activity than the analogue peptides.
Outer membrane permeabilization activity of peptides against P. aeruginosa. We investigated the ability of the peptides to disrupt bacterial membranes using the NPN uptake assay. The hydrophobic probe NPN is used to assess the permeabilization of the outer membrane of gram-negative bacteria. NPN exhibits weak fluorescence in aqueous environments, but produces a strong signal in hydrophobic environments such as the lipid layer of the outer bacterial membrane, to which it gains access when the membrane is disrupted. Incubation with 4× MIC Hp1404 for 5 min increased NPN uptake by 40%, indicating that the parent peptide was able to effectively permeabilize the membrane. The increases in NPN fluorescence induced by the analogue peptides were significantly lower (Fig. 6b) than the one caused by Hp1404, suggesting that the analogues had a lower outer membrane permeabilizing activity compared to the parent peptide.
Depolarization of cytoplasmic membrane in P. aeruginosa. The membrane potential-sensitive dye DisC 3 -5 was used to evaluate the ability of the peptides to depolarize the bacterial cytoplasmic membrane (Fig. 6c). DisC 3 -5 accumulates in polarized cytoplasmic membranes, whereas disruption of the membrane potential releases the probe, resulting in an increase in fluorescence intensity. After a 10-min stabilization period, peptides were added and changes in membrane potential were monitored for another 30 min. All peptides dose-dependently increased fluorescence intensity, suggesting that they had a significant membrane depolarization effect on P. aeruginosa. Hp1404 or Hp1404-T1e at 4× MIC increased fluorescence intensity to almost 200, whereas 4× MIC Hp1404-T1c and Hp1404-T1d increased fluorescence intensity up to 150. PI uptake assay. Propidium iodide (PI) uptake was performed to observe the permeabilization activity on the Pseudomonas aeruginosa membrane. Increasing PI fluorescence indicates the degree of permeability of the bacterial membrane (Fig. S6). Melittin, which is known to quickly disrupt the membrane, generated an increased in PI fluorescence within 2 min. In contrast, the fluorescence generated by ciprofloxacin was low, indicative of the absence of effect on membrane permeabilization. The permeabilization effect of peptides was concentration-dependent. Similarly to melittin, Hp1404 exhibited an activity on the membrane, while in contrast, analogue peptides at 1× MIC showed relatively low activities.
FACS analysis. FACS analysis of propidium iodide-stained bacteria was applied to compare the effect of Hp1404 and one of its analogues, Hp1404-T1e, on the integrity of bacterial membranes, as described by Jang et al. 27 . Propidium iodide (PI) is a membrane impermeant dye that binds to double-stranded DNA by intercalating between base pairs. In the absence of peptides, the percentage of PI-positive P. aeruginosa cells was  Table 4. MBIC values of the parent Hp1404 peptide and three of its analogues, as well as ciprofloxacin, against P. aeruginosa strains.
DNA binding assay. After revealing the significant differences among the peptides with respect to their effects on bacterial membranes, we investigated their intracellular effects; specifically, we evaluated their DNA binding ability using the gel retardation assay, as previously described 28 . Buforin 2, used as a positive control, is known to bind to DNA and kill bacteria without damaging the cell wall 29 . As seen in Figs 7 and S7, Hp1404 and Hp1404-T1c did not bind to DNA. In contrast, Hp1404-T1c and Hp1404-T1e exhibited a clear inhibitory effect on DNA migration at peptide:DNA ratios of 1.0 and higher. These results suggest that Hp1404-T1c and Hp1404-T1e can interact with DNA, similarly to buforin 2 (Figs 7a and S7a) 30 . Efficacy of Hp1404-T1e in vivo. In order to observe the antimicrobial effect of Hp1404-T1e in vivo, BALB/c mice were used (Fig. 8). In the absence of bacterial infection, PBS treated mice, Hp1404-T1e treated mice and ciprofloxacin-treated mice served as controls. Compared with the control group, P. aeruginosa-infected mice showed colonies until 4 days. The number of colonies in Hp1404-T1e treated mice decreased to 60% after 3 days and to 40% after 4 days, but ciprofloxacin showed a decrease of about 60% until 4 days. On the 5 th and 6 th days, the number of bacterial colonies was slightly reduced in P. aeruginosa-infected mice and ciprofloxacin-treated mice, but Hp1404-T1e treated mice bacterial colony formation was strongly inhibited. On day 7, mice treated with Hp1404-T1e, the bacteria were completely killed, whereas P. a-infected mice and ciprofloxacin treated mice did not kill the bacteria.

Discussion
The emergence and rapid worldwide spread of antibiotic-resistant bacteria poses major public health issues 31 . New antibacterial agents are being developed to counter the growing resistance of bacteria against conventional antibiotics. AMPs, which are essential elements of the human innate immune system, are excellent candidates for the development of novel therapeutic agents.
AMPs are found in a variety of organisms (e.g., insects, fungi, plants, and mammals), and characterized by cationic amphipathic properties. Up to now, 50 peptides have been isolated from the scorpion venom and most of them exhibit broad spectrum antimicrobial activities against bacteria, including multidrug-resistant strains 32 . However, scorpion peptides are toxic toward eukaryotic cells and therefore, it is necessary to develop AMPs that maintain high activity while displaying lower toxicity 33,34 .
In a previous study, Hp1404, composed of 14 amino acids and displaying an amphipathic and αhelical structure, was isolated from scorpion. However, although this peptide displays a potent antimicrobial activity against gram-positive bacteria, including Staphylococcus aureus, it also exhibits toxicity toward eukaryotic cells 20 .
In this study, we designed new analogue peptides (Hp1404-T1, Hp1404-T1a, Hp1404-T1b, Hp1404-T1c, Hp1404-T1d, and Hp1404-T1e) based on Hp1404. These peptides are composed of 12 amino acids after removal of Glycine 1 and Phenylalanine 14 from the parent peptide. Subsequently, the analogue peptides form a pronounced amphipathic structure, displaying higher cationic properties, as a result from the substitution of the removed residues with lysine and leucine residues 35 . Peptides comprising the hydrophilic amino acid leucine and positively charged lysine residues have strong antimicrobial activity and form a helix structure 36,37 . The hydrophobic amino acid Leucine increases the hydrophobicity in an amphipathic structure, thereby conferring an important role in antimicrobial activity through the interaction with bacterial membranes 38 . The lysine residue, a positively charge amino acid, enhances the binding between AMP and the negatively charged bacterial membrane by increasing the electrostatic attraction 39 . Therefore, the leucine and lysine residues were located on the hydrophobic and hydrophilic faces, respectively (Fig. 1). In the 3D representation, the N-terminal random coil structure disappears and a strong α-helical structure is formed. As a result, the analogue peptides display a well-defined amphipathic α-helical structure, with higher hydrophobic moment, and increased net positive charge 37,40 .
We examined the antibacterial activities of Hp1404 and its analogues against microorganisms. The parent peptide and three of the analogues (Hp1404-T1c, Hp1404-T1d, and Hp1404-T1e) exhibited antimicrobial activities. The parent peptide displayed antibacterial activity against gram-positive bacteria, including methicillin-resistant staphylococcus aureus (MRSA); however, it exhibited hemolytic and toxic activities toward mammalian cells. The antimicrobial activity assays demonstrated that Hp1404-T1e was the analogue with the most potent activity against Pseudomonas aeruginosa strains. In contrast, the conventional antibiotic, ciprofloxacin, showed no antibacterial activity, even at high concentrations (Table 2). At active concentrations, Hp1404-T1e displayed no hemolytic activity and no toxicity toward mouse RBCs or HaCaT cells, respectively. We also showed that Hp1404-T1e was not toxic through calcein leakage assay, indicating that Hp1404-T1e could be developed into a safe therapeutic agent (Fig. S4).
The development of antimicrobial peptides involves overcoming several limiting factors. Firstly, antimicrobial activity decreases in high-salt conditions, such as those found in physiological environments. Secondly, their sensitivity toward digestion by proteases can also represent a major drawback 41 . After testing their stability upon trypsin treatment for 2 h, the analogue peptides proved to be more resistant toward proteolytic degradation than Hp1404 (Fig. S5), suggesting that Hp1404-T1e might be able to overcome this common limitation 42 . Additionally, we confirmed that the activity of this analogue peptide was not overly impacted by high-salt environments, which is an important feature regarding its use under physiological conditions (Table 3) 43 . Taken together, these results show that Hp1404-T1e is stable in vivo and can be used as a therapeutic agent.
Bacterial biofilms are involved in a wide range of microbial infection events. A typical biofilm infection is usually chronic because of the low sensitivity of constituting cells toward antibiotics 44 . P. aeruginosa is known as a major biofilm-forming pathogen. We investigated the ability of the peptides to inhibit P. aeruginosa biofilm formation (Table 4). Hp1404-T1e could inhibit biofilm formation at a concentration lower than the one required when using the conventional antibiotic. The degree of biofilm inhibition was visualized using SYTO9/PI staining. As a result, we conclude that the peptide analogues are effective agents against P. aeruginosa biofilm-related infections (e.g., cystic fibrosis, keratitis) 45 .
LPS, also known as endotoxins, are located on the outer membrane of gram-negative bacteria and can cause inflammations 46 . Peptide-LPS interaction is important toward allowing bacterial death and reducing inflammation 47 . Owing to their cationic properties (positive net charge), AMPs can bind to, and accumulate on, the surface of bacteria that contain anionic polymers such as LPS. The CD spectra obtained suggest that Hp1404 and its analogues display an α-helical structure when bound to LPS. Unlike Hp1404, which antimicrobial activity was not affected by LPS, the analogue peptides showed lower activities in the presence of 100 µg/mL LPS, suggesting that LPS binds to the peptides and interfere with their antibacterial activity. The absorbance recorded after adding 2× MIC of peptides increased compared to the value obtained at 1× MIC. We conclude that increasing the net charge significantly enhances peptide affinity toward the bacterial surface. Taken together, these data suggest that peptides could prove effective against LPS-induced inflammations.
Antimicrobial peptides are reported to act by forming α-helical, β-sheet, and random coil structures 48,49 . Even though peptides were predicted to form α-helices (Fig. 1b), the secondary structure is dependent on the environment. Thus, we used CD spectroscopy to investigate the peptide conformation in different solutions. Peptides formed random coils in aqueous solutions and α-helixes in bacterial membrane-mimicking environments. Furthermore, LUVs were used to investigate the relationship between membrane lipid composition and AMP-induced aggregation. CD spectroscopy was used to investigate the peptide secondary structure in the presence of LUVs. These data suggest that a high concentration of Hp1404 is required to allow the α-helical structure to appear in the presence of PE:PG (bacterial membrane mimicking) and PC:CH (mimicking RBCs) liposomes. There is no relationship with the 3D structure (Fig. 1b). Hp1404 showed strong affinity to PE:PG liposomes compared to analogue peptides. In contrast, Hp1404-T1c, Hp1404-T1d, and Hp1404-T1e only displayed weak affinities for these LUVs. In the cases of PC:CH and PC:CH:SM liposomes, neither of the four tested peptides caused aggregation. These data indicate that low concentrations of the peptides do not cause toxicity. Combined with the results obtained from the CD spectroscopy assay, we conclude that Hp1404 acts strongly on the bacterial membrane while the effects induced by the analogue peptides are very limited.
In a more detailed mechanism study, three Hp1404 analogues were compared to the parent peptide with respect to their biological activities and action mechanisms. To explore the action mechanism of peptides, their effects on bacterial membranes were investigated using the calcein leakage assay, the NPN outer membrane permeabilization assay, the DisC 3 -5 cytoplasmic membrane depolarization assay, and PI FACS analysis. PC:PG liposomes were used to mimic negatively charged bacterial membranes. The analogue peptides did not induce calcein leakage from these LUVs, whereas Hp1404 did. Hp1404 significantly increased NPN fluorescence, implying its damaging effect on the outer membrane. The analogue peptides caused a much smaller increase in NPN uptake percentage compared to the parent peptide, indicating the absence of effect on outer membrane permeabilization.  In P. aeruginosa, incubation with parent or analogue peptides resulted in similar levels of fluorescence intensity being generated. Taken together, these results suggest that there is a difference between the parent peptide and its analogues with respect to their effect on membranes. FACS analysis showed that Hp1404 damaged membranes while Hp1404-T1e did not. Based on these results, we investigated the potential presence of intracellular peptide targets using the DNA binding assay 50 . Surprisingly, Hp1404-T1d and Hp1404-T1e interacted with DNA, suggesting that their differences with the parent peptide may imply the existence of a distinct action mechanism 51,52 .
We observed the antibacterial effect of Hp1404-T1e in vivo and found that Hp1404-T1e has an inhibitory activity against P. aeruginosa in vivo compared to ciprofloxacin antibiotics; thus, Hp1404-T1e showed potential as a good therapeutic agent.
In conclusion, among the analogue peptides, Hp1404-T1e exhibited potent activity against P. aeruginosa strains. The peptides displayed lower hemolytic and toxic activities towards mammalian cells compared to the parent peptide. Hp1404-T1e showed better antibiofilm formation properties than the parent peptide. Moreover, Hp1404 kills bacteria through membrane disruption, while Hp1404-T1e may be able to enter the cells and interact with their DNA. Furthermore, Hp1404-T1e has high stability against salts and enzymes and has antibacterial activity in vivo. Overall, Hp1404-T1e is a promising candidate for the development of a novel therapeutic agent against P. aeruginosa infections. Peptide design and sequence analysis. The peptides were designed by truncating and substituting residues, based on the helical wheel diagram and three-dimensional structure of Hp1404. Projections of the predicted three-dimensional structures were constructed online using the Mobyle@RPBS bioinformatics portal (http:// mobyle.rpbs.univ-paris-diderot.fr/cgi-bin/portal.py#welcome), whereas the HeliQuest site (http://heliquest. ipmc.cnrs.fr) was used to create helical wheel diagrams and determine the relative hydrophobic moments of the peptides.

Materials and Methods
Antimicrobial activity assay. We examined the activity of each peptide against both pseudomonas aeruginosa and multidrug-resistant strains. The minimal inhibitory concentrations (MICs) of peptides were determined using the broth micro-dilution assay as described previously 53 . Briefly, bacterial strains were cultivated with shaking overnight at 37 °C. The next day, they were diluted to a density of 2 × 10 5 CFU/mL with growth medium. Peptides were serially diluted with 10 mM sodium phosphate buffer. 50 µL bacterial aliquots were mixed with an equal volume of peptide solution, and incubated at 37 °C for 18-24 h. Inhibition of bacterial growth was determined by measuring the absorbance at 600 nm using a Versa Max microplate reader (Molecular Devices, Sunnyvale, CA, USA). The test was performed in triplicate.
Hemolysis and cytotoxicity assay. The hemolytic activities of the peptides were determined as previously described 54 . Fresh mouse red blood cells (RBCs) were washed and diluted to 8% with PBS. RBC suspension was added to 96-well plates (100 µL/well), mixed with an equal volume of serially diluted peptides, and incubated at 37 °C for 1 h with gentle shaking. The plate was centrifuged and the supernatant was transferred to a 96-well plate. The absorbance at 414 nm was measured with a Versa-Max ELISA reader (Molecular Devices). PBS and 0.1% Triton X-100 were used as negative and positive controls, respectively. The experiment was performed in triplicate. The hemolysis rate was calculated using the following formula 55  We used the MTT assay to evaluate the toxicity of the peptides toward human skin epithelial cells (HaCaT) as previously described 56 . Briefly, HaCaT cells were seeded at a density of 2 × 10 4 cells per well in 96-well microtiter plates in Dulbecco's modified Eagle's medium (DMEM) containing 10% foetal bovine serum (FBS) at 37 °C under 5% CO2 atmosphere. The next day, peptides were added to each well at various final concentrations (higher concentration of 200 µM). After 24 h incubation, culture supernatants were removed. 0.5 mg/mL MTT was added to each well, and the plates were incubated for 4 h at 37 °C. The medium was replaced by dimethyl sulfoxide (DMSO) to dissolve the formazan crystals formed in the wells. The absorbance at 570 nm was measured using a Versa-Max microplate ELISA reader (Molecular Devices). The viability rate was calculated using the following formula: Cell viability (%) = (absorbance of treated sample)/(absorbance of control) × 100.
Salt sensitivity. The effects of various cations on the antibacterial activities of the parental and analogue peptides against P. aeruginosa 27853 were evaluated as previously described 57 . Bacteria was treated with peptides at a final concentration 25 µM in the presence of physiological salts such as NaCl (50 mM, 100 mM, and 150 mM), CaCl 2 (1.25 mM, 2.5 mM, and 5 mM), or MgCl 2 (0.5 mM, 1 mM, and 2 mM). The MIC values were determined as described above.
Trypsin stability assay. Each sample with trypsin was analysed by RT-HPLC on a Jupiter C18 column in turn was determined by comparing the HPLC peak areas, corresponding to the peak detected for the native peptide, before and after trypsin treatment.
Antibiofilm activity assay and SYTO9/PI staining. To evaluate the ability of bacterial strains to form biofilms either in the absence or presence of peptides, antibiotic-susceptible P. aeruginosa (ATCC 27853), and various strains of MDR P. aeruginosa were cultured in nutrient broth (NB). Bacterial suspensions (5 × 10 5 CFU/ mL) with 0.2% glucose were added to 96-well tissue culture plates, mixed with peptide solution (at a final concentration of 25 µM), and incubated for 24 h at 37 °C. After the incubation, culture supernatants were discarded and wells were gently washed with PBS to remove planktonic cells. The biofilms were fixed with 100% methanol for 15 min, stained with 0.1% crystal violet (CV) for 1 h, and rinsed three times with distilled water (dH 2 O). After the addition of 95% ethanol, biofilm mass was quantified by measuring OD at 595 nm using a Versa-Max microplate ELISA reader (Molecular Devices). Results were expressed as the percentage of biofilm formed with respect to the control. We also calculated the minimum biofilm inhibitory concentration (MBIC), which was defined as the lowest peptide concentration at which biofilm formation was completely inhibited. All experiments were performed at least in triplicate.
Visualization of the remaining bacteria in the biofilm was achieved using the LIVE/DEAD BacLight Bacterial Viability Kit (Molecular Probes, Eugene, OR, USA), and an IX71 inverted fluorescence microscope (Olympus, Tokyo, Japan) was used for observation 58,59 . Briefly, bacterial suspensions (5 × 10 5 CFU/mL) were incubated with 0.5× MBIC or 1× MBIC of the peptides, for 24 h at 37 °C. Following incubation, biofilms were stained with pre-mixed dye (SYTO-9 and propidium iodide) for 30 min in the dark; images were obtained using the aforementioned microscope.

LPS binding assay.
To study the effect generated by interactions between lipopolysaccharides (LPS) and peptides, we monitored the antimicrobial activity against P. aeruginosa ATCC 27853. The LPS (final concentration 100 µg/mL) solution was serially diluted in 10 mM sodium phosphate buffer (pH 7.2), and then mixed with 1× and 2× MIC peptides. A P. aeruginosa suspension at 2 × 10 5 CFU/mL was prepared and added to the mixtures. The plate was incubated overnight at 37 °C, and bacteria growth was measured at 600 nm.

Circular dichroism spectroscopy toward examining secondary structures.
To investigate the secondary structure within different environments, 40 µM peptide solutions were prepared in 30 mM sodium dodecyl sulfate (SDS; Sigma-Aldrich), 50% trifluoroethanol (TFE; Sigma-Aldrich), or 10 mM sodium phosphate buffer (pH 7.2), with or without the addition of 0.1% LPS. Circular dichroism (CD) assays were then performed using a JASCO 810 spectropolarimeter (Jasco, Tokyo, Japan). CD spectra were recorded from 190 nm to 250 nm using a quartz cell with a 1.0 mm path length 60 .
Large unilamellar vesicles preparation. Large unilamellar vesicles (LUVs) were prepared using the freeze-thaw method 61 . Briefly, we prepared the following lipid mixtures: phosphatidylethanolamine (PE) and polyethylene glycol (PG) at a 7:3 w/w ratio; phosphatidylcholine (PC), cholesterol (CH), and sphingomyelin (SM) at a 1:1:1 w/w ratio, and; PC:CH at a 2:1 w/w ratio. The mixtures were dissolved in chloroform, and then dried in argon gas. The remaining chloroform was removed by overnight lyophilization. The next day, dry lipid films were resuspended in 2.6 mL PBS by gently vortexing for 30 min. After nine freeze-thaw cycles under liquid nitrogen and a water bath at 50 °C, lipid suspensions were extruded 15-20 times through 0.2 µm polycarbonate membranes. Lipid concentration was determined with the phosphate assay, using a standard curve.
Liposome aggregation and CD assay. Aggregation of LUVs were confirmed by measuring turbidity change. Peptides were prepared at concentrations ranging from 10 µM to 80 µM, and added to 400 µM LUV composed of PE:PG (7:3, w/w), PC:CH:SM (1:1:1, w/w), and PC:CH (2:1, w/w). Aggregation causes an increase in turbidity, which is quantified by measuring absorbance at 405 nm. The secondary structure was assessed using circular dichroism (CD) spectroscopy, as previously described 60 . CD spectra were measured in aqueous solution with 100 µM peptide in the presence of LUVs (400 µM).
Calcein leakage assay. Peptide-induced membrane permeabilization was quantified by monitoring the calcein release from LUVs 62 . PC:PG liposome (9:1, w/w, mimicking a bacterial membrane surface) was prepared with 70 mM calcein as described above. The calcein-entrapped vesicles were removed by gel filtration chromatography on a Sephadex G-50 column. The calcein-loaded LUVs (10 µM) were mixed with different concentrations of peptides (0.00625-4 µM) in a 96-well black plate. The fluorescence generated from the released calcein was measured at an excitation wavelength of 480 nm and an emission wavelength of 520 nm.
Outer membrane permeabilization assay. The outer membrane permeabilization activities of the peptides were determined using the NPN assay 63 . P. aeruginosa (ATCC 27853) was grown to mid-log phase at 37 °C, and harvested by centrifugation in 5 mM HEPES buffer (pH 7.2). Subsequently, the cell suspension (OD 600 = 0.2) was mixed with 10 μM NPN in 96-well black plates. After addition of peptides at various final concentrations, or Triton X-100 at a final concentration of 0.1% (positive control), NPN fluorescence was measured (excitation wavelength of 350 nm, emission wavelength of 420 nm).
Cytoplasmic membrane depolarization assay. The membrane depolarization activities of the peptides were determined using P. aeruginosa (ATCC 27853), and the membrane potential-sensitive fluorescent dye 3,3′-Dipropylthiadicarbocyanine iodide (DiSC 3 -5), as previously described 64  containing 5 mM HEPES (pH 7.3), 20 mM glucose, and 0.1 M KCl. DiSC 3 -5 was then added at a final concentration of 1 µM. After 1 h incubation, mixture was transferred into 96-well plates and left until maximal dye uptake had been reached. Subsequently, peptides were added to the wells, and fluorescence intensity was monitored (excitation wavelength of 622 nm, emission wavelength of 670 nm).
Flow cytometry. P. aeruginosa (ATCC 27853) cells were washed three times with PBS buffer (pH 7.4), and resuspended in the same buffer to OD = 0.2. Bacteria were incubated for 1 h at 37 °C in the absence (negative controls) or presence of each tested peptide (at 2× MIC). Cells were centrifuged and then treated with propidium iodide (PI; Sigma-Aldrich) at a final concentration of 10 µg/mL for 30 min at 4 °C. After the incubation, cells were centrifuged at 10 000× g and resuspended in 1 mL PBS. Analysis was performed using a CytoFLEX flow cytometer (Beckman Coulter, Brea, CA, USA.).