*Article* **Development of Novel Peptides for the Antimicrobial Combination Therapy against Carbapenem-Resistant** *Acinetobacter baumannii* **Infection**

**Joonhyeok Choi <sup>1</sup> , Ahjin Jang <sup>1</sup> , Young Kyung Yoon <sup>2</sup> and Yangmee Kim 1,\***


**Abstract:** Carbapenem-resistant *Acinetobacter baumannii* (CRAB) infection has a high mortality rate, making the development of novel effective antibiotic therapeutic strategies highly critical. Antimicrobial peptides can outperform conventional antibiotics regarding drug resistance and broad-spectrum activity. PapMA, an 18-residue hybrid peptide, containing N-terminal residues of papiliocin and magainin 2, has previously demonstrated potent antibacterial activity. In this study, PapMA analogs were designed by substituting Ala<sup>15</sup> or Phe<sup>18</sup> with Ala, Phe, and Trp. PapMA-3 with Trp<sup>18</sup> showed the highest bacterial selectivity against CRAB, alongside low cytotoxicity. Biophysical studies revealed that PapMA-3 permeabilizes CRAB membrane via strong binding to LPS. To reduce toxicity via reduced antibiotic doses, while preventing the emergence of multi-drug resistant bacteria, the efficacy of PapMA-3 in combination with six selected antibiotics was evaluated against clinical CRAB isolates (C1–C5). At 25% of the minimum inhibition concentration, PapMA-3 partially depolarized the CRAB membrane and caused sufficient morphological changes, facilitating the entry of antibiotics into the bacterial cell. Combining PapMA-3 with rifampin significantly and synergistically inhibited CRAB C4 (FICI = 0.13). Meanwhile, combining PapMA-3 with vancomycin or erythromycin, both potent against Gram-positive bacteria, demonstrated remarkable synergistic antibiofilm activity against Gram-negative CRAB. This study could aid in the development of combination therapeutic approaches against CRAB.

**Keywords:** antimicrobial peptides; antibiotics; synergistic effect; CRAB

## **1. Introduction**

The emergence of multi-drug resistant (MDR) bacteria, combined with the failure of most antibiotic candidates in clinical trials, poses a serious threat to global public health [1–3]. In particular, diseases caused by Gram-negative bacteria, such as postoperative wound infection, urinary tract infection, hospital-acquired pneumonia, catheter-associated bloodstream infection, meningitis, and sepsis [4,5], have high mortality. Carbapenems such as doripenem, imipenem, and meropenem are generally considered to be the final choice of treatment for MDR Gram-negative bacteria; however, these bacteria have recently begun to show increased resistance to these drugs. MDR bacterial infections featuring carbapenem-resistant *Acinetobacter baumannii* (CRAB) are at the top of the World Health Organization (WHO) priority list for the development of new antibiotics [6–8]. Therefore, there is a need to accelerate the development of new antibiotic therapeutic strategies.

As antibiotic resistance develops rapidly after the introduction of new antimicrobial agents, it is necessary to develop antimicrobial compounds with novel mechanisms that differ from those of conventional antibiotics. Antimicrobial peptides (AMPs) are diverse, and they are produced by various living organisms [9,10], where they are known to

**Citation:** Choi, J.; Jang, A.; Yoon, Y.K.; Kim, Y. Development of Novel Peptides for the Antimicrobial Combination Therapy against Carbapenem-Resistant *Acinetobacter baumannii* Infection. *Pharmaceutics* **2021**, *13*, 1800. https://doi.org/ 10.3390/pharmaceutics13111800

Academic Editor: Clive Prestidge

Received: 31 August 2021 Accepted: 25 October 2021 Published: 27 October 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

participate in the organism's innate immunity [11–13]. Unlike conventional antibiotics, most AMPs have amphiphilic structures, and they exhibit antibacterial activity primarily through interactions with the negatively charged bacterial membrane, making it difficult for the bacteria to develop resistance [14]. In addition, their rapid and broad-spectrum antimicrobial activity make them potential therapeutic alternative to antibiotics [15].

In the clinical setting, different antibiotics are often used in combination therapy to broaden the antimicrobial spectra. The main advantage of combination antibiotic therapy is that it can prevent the emergence of MDR bacteria. Antibiotics can exhibit side effects such as diarrhea, rash, nausea, liver damage, and kidney damage; therefore, decreasing drug toxicity through lowering the doses is beneficial [16–18]. A few novel AMPs exhibit synergistic effects with known antibiotics against MDR bacteria. (p-BthTX-I)<sup>2</sup> and LL-37 in combination with florfenicol and thiamphenicol exert antimicrobial activity against *Citrobacter freundii* [19]. Melittin in combination with clindamycin has shown antimicrobial activity against methicillin-resistant *Staphylococcus aureus* [20]. AMPs have also been combined with antibiotics such as T3, T4 with ampicillin and oxacillin [21], WW304 with ciprofloxacin [22], and G3KL with erythromycin and vancomycin [23]. As carbapenem is the most used front-line antibiotic for treating Gram-negative bacterial infections, the accelerating appearance of CRAB seriously threatens global public health [6–8]. Antimicrobial activity against MDR-Gram-negative bacteria has been improved through the synergistic effects of SET-M33 [24] or melittin [25] with antibiotics; however, such synergistic combinations with antibiotics to combat CRAB infections remains challenging to develop. AMPs have thus demonstrated some potential regarding combination therapy with conventional antibiotics. Additionally, to overcome drug resistance, AMPs can be easily manipulated to design potent novel AMPs by substituting their amino acid residues. Therefore, the development of AMPs that have synergistic effects with antibiotics against MDR Gram-negative bacteria, and especially clinical CRAB isolates, is important but challenging.

AMPs with improved antimicrobial activities include a series of hybrid peptides that were designed by combining the active regions of two AMPs. For example, cecropin A-magainin 2 (CAMA) and cecropin A-melittin (CAME) hybrid peptides have previously been reported to demonstrate high antimicrobial and antitumor activities [26–28]. A hybrid peptide with a broad spectrum of antimicrobial activity (PapMA) was discovered by connecting the N-termini of papiliocin and magainin 2, joined by a proline (Pro) hinge [29]. The structure of PapMA was investigated using nuclear magnetic resonance (NMR) spectroscopy, revealing that it had an N-terminal α-helix from Lys<sup>3</sup> to Lys<sup>7</sup> and a C-terminal α-helix from Lys<sup>10</sup> to Lys17, with a Pro<sup>9</sup> hinge in between. PapMA showed potent antibacterial activity against both Gram-negative and Gram-positive bacteria.

In this study, we aimed to design a novel PapMA analog with increased anti-CRAB activity, while maintaining low cytotoxicity. Its synergistic antibacterial activities against CRAB were then investigated when combined with conventional antibiotics, and the inhibition of biofilm formation was also assessed. In total, six analogs of PapMA were designed by substituting Ala<sup>15</sup> or Phe<sup>18</sup> with Ala, Phe, and Trp at the C-terminus. Among the six analogs, we chose PapMA-3 as a candidate for further investigation, as it showed potent anti-CRAB activity with low cytotoxicity. PapMA-3 was found to depolarize CRAB cell membranes, which disrupted biofilm formation and increased susceptibility to the conventional antibiotics. Therefore, in this study, the key mechanism of action underlying this AMP activity was elucidated, suggesting that they are valuable as an adjuvant pharmaceutical to overcome Gram-negative bacterial resistance and represents a good starting point for the development of new antibiotics against CRAB infection.

#### **2. Materials and Methods**

#### *2.1. Peptides and Materials*

All peptides were synthesized through N-(9-fluorenyl) methoxycarbonyl solid-phase synthesis and were purified using reversed-phase high-performance liquid chromatography (RP-HPLC, YL9100, Younglin, Korea). Peptide purities were over 95%, as evaluated

using an analytical HPLC (C18 column, 4.6 × 250 mm) with two different linear gradients of 0.05% aqueous trifluoroacetic acid (TFA, eluent A) and 0.05% TFA in CH3CN (eluent B) at a flow rate of 1.5 mL per min at 25 ◦C. The molecular masses of the peptides (Table 1) were determined using Axima (Shimadzu Scientific Instruments, Kyoto, Japan) matrixassisted laser-desorption ionization-time of flight mass spectrometry at the Korea Basic Science Institute (KBSI, Ochang, Korea). The conventional antibiotics (purity over 95%) were purchased as follows: imipenem, meropenem, erythromycin, and rifampin from Sigma-Aldrich (St. Louis, MO, USA), vancomycin from BIO BASIC (Markham, Ontario, Canada), and linezolid from Pharmacia & Upjohn Company (Kalamazoo, MI, USA).

#### *2.2. Antimicrobial Activity*

The Gram-negative bacterial strain *Escherichia coli* (KCTC 1682) and Gram-positive bacterial strain *Staphylococcus aureus* (KCTC 1621) were purchased from the Korean Collection for Type Cultures (Jeongeup, Korea). *Acinetobacter baumannii* (KCCM 40203) were purchased from Korea Culture Center of Microorganisms (Seoul, Korea). Additionally, five carbapenem-resistant *Acinetobacter baumanii* C1–C5 (CRAB C1–C5), which have the OXA-23 gene with carbapenem-resistance were collected from the patients with CRAB bacteremia, who presented symptoms and signs of infection at Korea University Anam Hospital (Seoul, Korea) (IRB registration no. 2020AN0157). The minimum inhibitory concentrations (MIC) of the AMP and antibiotics against the various bacterial strains were assessed using the serial dilution method on Muller–Hinton (MH) media, as described previously [30]. In brief, the peptides at 128 <sup>µ</sup>g·mL−<sup>1</sup> and antibiotics at 512 <sup>µ</sup>g·mL−<sup>1</sup> were serially diluted to 1/2 and incubated with a bacterial suspension of 2 <sup>×</sup> <sup>10</sup><sup>5</sup> CFU·mL−<sup>1</sup> in MH media at 37 ◦C for 16 h. Absorbance at 600 nm was measured using a SpectraMAX microplate reader (Molecular Devices, San Jose, CA, USA).

#### *2.3. Peptide-LPS Binding Assay*

The capacity of PapMA series peptides to bind with LPS was analyzed using a fluorescent probe, BODIPY-TR cadaverine (BC) (Thermo Fisher Scientific, MA, USA), as described previously [31]. The probe complex was prepared by incubating LPS (50 <sup>µ</sup>g·mL−<sup>1</sup> ) with BC (5 <sup>µ</sup>g·mL−<sup>1</sup> ) in a 50 mM Tris buffer (pH 7.4) for 6 h at 25 ◦C. Varying concentrations of peptides (1–64 <sup>µ</sup>g·mL−<sup>1</sup> ) were added to a 96-well, dark fluorescence plate and allowed to interact with the LPS–BC complex for 30 min. The fluorescence intensity was recorded at an excitation wavelength of 580 nm and emission wavelength of 620 nm using a Spectra-MAX GeminiTM fluorescence microplate reader (Molecular Devices). The %∆F (A.U.) was calculated using Equation (1):

$$\text{V@}\Delta\text{F (A.U.)} = \left[ \left( \text{F}\_{\text{obs}} - \text{F}\_0 \right) / \left( \text{F}\_{100} - \text{F}\_0 \right) \right] \times 100 \tag{1}$$

Fobs is the observed fluorescence due to the peptide. F<sup>0</sup> is the fluorescence without the addition of the peptide. F<sup>100</sup> is the fluorescence value measured using LL-37, a control peptide with outstanding LPS-neutralizing properties [32].

#### *2.4. Membrane Depolarization*

The membrane depolarization activity of each AMPs at varying concentrations (1–16 <sup>µ</sup>g·mL−<sup>1</sup> ) against CRAB C1 intact cells were measured using 3,30 -dipropylthiadicarbocyanine iodide (diSC3-5). CRAB C1 was washed two times in washing buffer (5 mM HEPES, 20 mM glucose, pH 7.4), the experiment buffer was changed (5 mM HEPES, 20 mM glucose, 0.1 M KCl, pH 7.4), and diSC3-5 dye was added. As a control, 100% depolarization was obtained by treating CRAB C1 with 1% triton X-100 [33]. Spheroplast cells were prepared by the osmotic shock, as previously described [34]. Melittin, which exhibits strong membrane permeabilization, was used for the control treatment at varying concentrations (1–16 <sup>µ</sup>g·mL−<sup>1</sup> ). The corresponding fluorescent were measured using RF-6000PC fluorescent spectrophotometer (Shimadzu Scientific Instruments, Kyoto, Japan).

#### *2.5. Time-Dependent Permeabilization of the Outer Membrane*

Time-dependent outer membrane permeabilization activity of PapMA-3 in CRAB C1 intact cells was evaluated using fluorescence-based 1-N-phenylnaphthylamine (NPN). Melittin was used as the control. CRAB C1 cells were washed twice with buffer (5 mM HEPES, 20 mM glucose, pH 7.4) and diluted to OD<sup>600</sup> = 0.05; 1 µM NPN was added to the cells. Time-dependent NPN uptake was monitored following treatment with PapMA-3 for 30 min. PapMA-3, at varying concentrations (4–32 <sup>µ</sup>g·mL−<sup>1</sup> ), was added to the cells, and the fluorescence was measured using the RF-6000PC fluorescent spectrophotometer (Shimadzu Scientific Instruments).

#### *2.6. Cell Cytotoxicity*

Human embryonic kidney (HEK)-293 cells, purchased from Korean cell line bank (Seoul, Korea) were cultured in Dulbecco's modified Eagle's medium (DMEM) (Welgene, Gyeong-san, Korea) with 10% fetal bovine serum, 1% penicillin-streptomycin at 37 ◦C in a humidified 5% CO<sup>2</sup> incubator as described previously [30]. The cytotoxicity of the six PapMA peptides and melittin was determined using WST-8 Cell Proliferation Assay Kit (Biomax Co, Ltd., Seoul, Korea), according to the manufacturer's instructions. The effects of the most potent peptide, PapMA-3, on mammalian cells were evaluated by measuring the cell activity of HEK-293 cells and human keratinocyte HaCaT cells (Korean cell line bank, Seoul, Korea) after 24 h and 48 h of treatment. The absorbance was measured at 450 nm using a SpectraMAX microplate reader (Molecular Devices).

#### *2.7. Stability of PapMA-3 Compared to Melittin in Human Serum*

Serum stability of PapMA-3 was assessed by comparing its activity with that of melittin, based on the effects on *E. coli, A. baumannii*, and CRAB C1. MIC was measured in the presence of 50% human serum (Sigma-Aldrich) in the MH medium, in comparison to that of melittin, as described in Section 2.1. Antibacterial activity of PapMA-3 in combination with imipenem was measured against CRAB C1 in the presence of 50% human serum. The treated cells were incubated for 16 h at 37 ◦C, and the absorbance at 600 nm was measured using a SpectraMAX microplate reader (Molecular Devices).

#### *2.8. Hemolytic Activity*

The hemolytic activity of PapMA series peptides was determined against Sheep red blood cells (sRBC) (KisanBio, Seoul, Korea). Fresh sRBC were washed at least three times with phosphate-buffered saline (PBS), followed by centrifugation for 5 min at 1000× *g* at 4 ◦C. PapMA series peptides (0.25–256 <sup>µ</sup>g·mL−<sup>1</sup> ) dilute in PBS were incubated with 4% (*v*/*v*) sRBC for 1 h at 37 ◦C. The contents were then centrifuged at 4 ◦C for 5 min at 1000× *g*. After transferring the supernatant, absorbance was measured at 405 nm using SpectraMAX microplate reader (Molecular Devices). As a control, 100% hemolysis was obtained by treating sRBC with 1% triton X-100.

#### *2.9. Checkerboard Assays*

The synergistic effects of AMPs and antibiotics were measured using checkerboard assays [35]. Serial dilutions of PapMA-3 and antibiotics were performed from 1 to <sup>1</sup> 2 of the MIC. Samples were then cross-mixed and cultured in MH medium with <sup>2</sup> <sup>×</sup> <sup>10</sup><sup>5</sup> CFU·ml−<sup>1</sup> bacteria. Results were recorded after 16 h of incubation at 37 ◦C. The fractional inhibitory concentration index (FICI) was calculated according to the European Committee on Antimicrobial Susceptibility (EUCAST) [36]. The FICI was calculated using Equation (2):

$$\text{FICI} = \frac{(\text{MIC of PopMA-3 in combination})}{(\text{MIC of PopMA-3 alone})} + \frac{(\text{MIC of antibitic in combination})}{(\text{MIC of antibitic alone})} \quad (2)$$

where FICI ≤ 0.5 indicates synergism, 0.5 < FICI < 1 indicates an additive effect, 1 < FICI ≤ 4 represents indifference, and FICI > 4 shows antagonism [37].

#### *2.10. Time Killing Assay*

CRAB C1 cells at 2 <sup>×</sup> <sup>10</sup><sup>5</sup> CFU·mL−<sup>1</sup> were incubated with selected concentrations of AMP or antibiotic at 37 ◦C. At 5, 10, 15, 30, and 40 min and 1, 2, and 4 h, a ten-fold serially diluted suspensions with MH media were spread on an LB agar plate and incubated at 37 ◦C for 12 h; the number of colonies was counted.

#### *2.11. Scanning Electron Microscope Analysis*

Membrane damage of CRAB C1 was visualized using a field emission scanning electron microscope (FE-SEM), as described previously [38,39], to confirm that the peptides or antibiotics specifically targeted the bacterial membrane. CRAB C1 cells were washed and diluted in PBS to an OD<sup>600</sup> of 0.2 and incubated with either PapMA-3 or erythromycin or with a combination of 4 <sup>µ</sup>g·mL−<sup>1</sup> PapMA-3 and 128 <sup>µ</sup>g·mL−<sup>1</sup> erythromycin for 15 min or 30 min at 37 ◦C. The cells were washed using PBS, fixed in 1% osmium tetroxide for 1 h, and dehydrated using a graded ethanol series. After dehydration, ethanol contents in the sample were replaced with varying ratio (2:1, 1:1, 1:2, 0:1 *v*/*v*) of ethanol–isoamyl acetate mixture. The cells were fixed on a glass slide with hexamethyldisilzane, dried under reduced pressure, and platinum-coated; they were visualized using an FE-SEM (SU8020; Hitachi, Tokyo, Japan).

#### *2.12. Biofilm Inhibition*

Biofilm inhibition activity of PapMA-3 and antibiotics was measured against CRAB C1, as described previously [30]. CRAB C1 cells (2 <sup>×</sup> <sup>10</sup><sup>5</sup> CFU·mL−<sup>1</sup> ) were incubated with PapMA-3 and antibiotics in a tissue-culture well plate in MH medium containing 0.2% glucose for 16 h at 37 ◦C. The cells were stained with 0.1% (*w*/*v*) crystal violet in 0.25% (*v*/*v*) acetic acid for 1 h at room temperature; the dye complex was dissolved with ethanol. Absorbance at 595 nm was measured using SpectraMAX microplate reader (Molecular Devices) to quantify the biofilm formation.

#### *2.13. Isothermal Titration Calorimetry (ITC)*

Binding affinity was measured using MicroCal Auto-iTC200 (Malvern Panalytical, Malvern, UK) at the KBSI (Ochang, Korea). Each peptide (0.2 mM; 370 µL) was added to 0.025 mM of LPS (*E. coli* O111:B4, Sigma-Aldrich) in Dulbecco's phosphate-buffered saline (DPBS, pH 7.0); the injection duration was 2s, the spacing was 150 s, and the temperature was 37 ◦C. ITC data were analyzed using MicroCal Origin 2020b software (MicroCal origin, MA, USA).

#### *2.14. Saturation Transfer Difference (STD)-NMR*

STD-NMR experiments were performed at 25 ◦C on a Bruker 900 MHz spectrometer at KBSI (Ochang, Korea). The STD-NMR spectra were obtained using selective saturation of 15 µM LPS (*E. coli* O111:B4, Sigma-Aldrich, St. Louis, MO, USA) resonances at −4.0 ppm (40 ppm for reference spectra). Peptide was dissolved in 10mM sodium phosphate (pH 6.8, D2O) to a concentration of 0.5 mM. For all STD-NMR experiments, a cascade of 40 selective gaussian-shaped pulses of 50 ms duration were used with a total saturation time of 2 s. Difference spectrum was obtained by subtraction of the two spectra (on resonance-off resonance), which shows signals arising from the saturation transfer.

#### *2.15. Statistical Analysis*

Measurements were taken at least three times, and all statistical analyses were performed using the GraphPad Prism software 8.0 for windows (GraphPad Software, CA, USA). The values are expressed as mean ± standard deviation (SD). Statistical significance (*p* < 0.05) was determined using one-way or two-way ANOVA with Dunnett's test.

#### **3. Results**

#### *3.1. Design of PapMA and Its Analogs*

The cationicity and amphiphilicity of antimicrobial peptides are important regarding their binding to bacterial cell membranes via electrostatic interactions with phospholipid head groups, as well as via hydrophobic interactions with membrane lipids [40]. Papiliocin is a 37-residue AMP that was isolated from the swallowtail butterfly (Papilio xuthus) [41]; magainin 2 is a 23-residue AMP isolated from the skin of the African clawed frog (Xenopus laevis) [42,43]. These two peptides are highly cationic, have amphipathic αhelical structures, and have low cytotoxic effects against mammalian cells. Papiliocin has demonstrated high antibacterial activity against Gram-negative bacteria through bacterial membrane disruption, while magainin 2 has displayed high antimicrobial activity against both Gram-negative and Gram-positive bacteria. An 18-residue hybrid peptide (PapMA) was developed by incorporating N-terminal residues 1–8 of papiliocin and N-terminal residues 4–12 of magainin 2, joined by a proline (Pro) hinge [29]. However, the antibacterial activity of PapMA is not potent enough for it to function as a peptide antibiotic.

**Table 1.** Peptides and their physicochemical properties.


<sup>1</sup> Hydrophobicity <H> was calculated using http://heliquest.ipmc.cnrs.fr/cgi-bin/ComputParams.py (accessed on 17 August 2021) [44]. Bold letters in sequence represent substituted residues. <sup>2</sup> HeliQuest calculates the net charge at pH = 7.4.

To improve and optimize the balance between its antibacterial activity and cytotoxicity, here, analogs were designed by changing the hydrophobicity but maintaining the cationicity. A previous study demonstrated that Trp<sup>2</sup> and Phe<sup>5</sup> in the N-terminus of papiliocin play important roles in its antibacterial activity. Therefore, new analogs of PapMA were designed here by substituting Ala<sup>15</sup> or Phe<sup>18</sup> with Ala, Phe, or Trp at the C-terminus of PapMA to optimize the hydrophobicity and membrane permeabilizing activity, while achieving low cytotoxicity (Table 1). To investigate the role of Phe<sup>18</sup> at the C-terminus, Phe<sup>18</sup> was substituted with Ala or Trp (PapMA-2 and PapMA-3, respectively). To increase the hydrophobicity, Ala<sup>15</sup> was substituted with Phe or Trp (PapMA-4 and PapMA-5, respectively). For PapMA-6, both residues were substituted by Trp. PapMA-2, which had Ala at both positions, exhibited the lowest hydrophobicity (0.312), while PapMa-6, which had Trp at both positions, showed the highest hydrophobicity (0.527; Table 1). The hydrophobic moment of the C-terminal helix was highest in PapMA-6 (0.834), with an order of: PapMA-2 < PapMA < PapMA-3 < PapMA-4 < PapMA-5 < PapMA-6, as shown in Figure 1. The antimicrobial activities and cytotoxicities of peptides were also compared to the parent hybrid peptide, PapMA.

#### *3.2. Antimicrobial Activities of PapMA Analogs*

Measurement of the minimum inhibitory concentrations (MIC) was conducted to determine the effect of the hydrophobicity of each antimicrobial peptide on its antimicrobial activity. MIC was defined as the minimum concentration that killed more than 99% of bacteria; it was measured against two standard Gram-negative bacteria (*E. coli* and *A. baumanii*), five clinically isolated CRAB (C1–C5), and one Gram-positive bacteria (*S. aureus*). The antimicrobial activities of PapMA and its analogs are listed in Table 2.

**Figure 1.** Helical wheel diagram of C-terminal helix from 10th to 18th residue of PapMA and its analogs after Pro hinge generated using HeliQuest at pH 7.4 [44]. Residues at the N-terminus and C-terminus of C-terminal helix are marked as N and C in the figure. Hydrophilic residues are shown in blue. Hydrophobic residues are shown in yellow. Uncharged His is shown in cyan, and Ser is shown in purple. The arrows represent the helical hydrophobic moment.


**Table 2.** Antibacterial activities of antimicrobial peptides and antibiotics against microorganisms.

<sup>1</sup> The geometric means (GMs) are the mean minimum inhibitory concentration (MIC) values of Gram-negative bacterial strains. The GMs were assumed to be 256 <sup>µ</sup>g·mL−<sup>1</sup> for MIC > 128 <sup>µ</sup>g·mL−<sup>1</sup> and 1024 <sup>µ</sup>g·mL−<sup>1</sup> for MIC > 512 <sup>µ</sup>g·mL−<sup>1</sup> .

> In this study, six conventional antibiotics were selected for analysis. Imipenem and meropenem are carbapenem antibiotics that have demonstrated potency against Gramnegative bacteria; they inhibit cell wall synthesis [45]. They have very strong antibacterial activity against *E. coli* and *A. baumanii;* however, their antibacterial activity against carbapenem-resistant CRAB strains is very low. Rifampin has been shown to be potent against *Mycobacterium* and *S. aureus*; it inhibits bacterial deoxyribonucleic acid (DNA) dependent ribonucleic acid (RNA) polymerase [46]. The antibiotic vancomycin is only potent against Gram-positive bacteria; it inhibits cell wall peptidoglycan synthesis [47]. Ery

thromycin and linezolid, meanwhile, can bind to 50s ribosome RNA, causing Gram-positive bacterial death through the inhibition of protein synthesis [48]. Compared to PapMA, PapMA-2, which had a lower hydrophobicity due to substitution with Ala, showed a reduced antimicrobial activity. However, PapMA-3, -4, -5, and -6, which exhibited increased hydrophobicities, demonstrated enhanced antimicrobial activities. PapMA and its analogs showed more potent antibacterial activity against Gram-negative bacteria than against Gram-positive bacteria.

Geometric means (GM) were calculated to compare the relative antimicrobial activities of the analogs against Gram-negative bacteria. The GM values were in the order of PapMA-4 < PapMA-6 < PapMA-5 < PapMA-3 < PapMA < PapMA-2, confirming the improved activities of the analogs compared to PapMA (except for PapMA-2). These results suggest that the increased hydrophobicity had a positive effect on the antimicrobial activity. CRAB C1–C5 are carbapenem-resistant to imipenem and meropenem. In contrast, erythromycin [49], vancomycin [50], and linezolid have shown potent antibacterial activity against Gram-positive bacteria, but much lower antimicrobial activity against Gram-negative bacteria. Next, the antibacterial mechanisms of peptides were investigated.

## *3.3. Antibacterial Mechanisms of PapMA Analogs against CRAB*

## 3.3.1. Binding Assay of LPS-PapMA Analogs

LPS is a major component of the outer membrane of Gram-negative bacteria. It is the permeability barrier of conventional antibiotics, and results in the complication of antibiotic development. Therefore, it is useful to design AMPs that can perturb the bacterial membrane by interacting with LPS. To confirm the antibacterial mechanisms of the developed PapMA analogs against Gram-negative bacteria, the LPS binding mechanisms of the PapMA analogs were investigated using the BC displacement assay (Figure 2). LL-37, which is well-known as the most efficient LPS-neutralizing peptide, was used as a control; its fluorescence intensity at 64 <sup>µ</sup>g·mL−<sup>1</sup> of LL-37 was selected as 100% for comparison. The activity was compared to that of polymyxin B, which is also a well-known LPS-neutralizing peptide [51]. As a result of incubating the LPS-BC complex and the peptides together, all the peptides produced stronger dose-dependent enhancements in fluorescence intensity. All the PapMA analogs showed higher LPS binding interactions than that of polymyxin B. The results showed that LPS interaction increased following the substitution of Ala with Phe or Trp at the C-terminus. Comparing the interactions of PapMA and PapMA-3, LPS interactions increased slightly when Phe<sup>18</sup> was replaced with Trp. LPS interactions with PapMA-4, -5, and -6 with two aromatic rings at the C-terminus were slightly higher compared to those of PapMA, PapMA-2, and -3.

**Figure 2.** LPS interaction of PapMA analogs. Binding affinity of PapMA derivatives and polymyxin B to LPS based on displacement assays with BODIPY-TR-cadaverine fluorescent dye. Statistical analysis was performed using two-way ANOVA with Dunnett's comparison test. The values are expressed as the mean ± SEM of three independent experiments and are statistically significant at \*\*\**p* < 0.001; ns, not significant.

#### 3.3.2. Depolarization of PapMA and Its Analogs against CRAB C1

To elucidate the antibacterial mechanisms of the PapMA analogs on the CRAB C1 membrane, depolarization experiments were performed using intact CRAB C1, as well as CRAB C1 spheroplasts that were created by removing LPS and peptidoglycan; melittin was used as a control. Figure 3A shows that, at a concentration of 8 <sup>µ</sup>g·mL−<sup>1</sup> , the depolarization of PapMA analogs and melittin occurred close to the maximum. At 4 <sup>µ</sup>g·mL−<sup>1</sup> (i.e., half of the concentration of maximum depolarization), depolarization values of 70.7, 57.6, 75.7, 76.5%, 73.2, 70.5, and 86.4% were achieved, respectively. Melittin showed the highest depolarization, while PapMA-2, which had the lowest hydrophobicity, showed the lowest depolarization among all peptides. Interestingly, the PapMA analogs induced bacterial membrane damage even at concentrations much lower than their MICs. When LPS was removed from the CRAB C1 membrane, all peptides displayed approximately 30–40% lower depolarization for CRAB C1 spheroplasts than for the intact membrane (Figure 3B). These results, along with those from the BC displacement assays, indicate that PapMA and its analogs interacted with LPS, major outer membrane component of CRAB, implying that the PapMA peptides targeted and disrupted the outer bacterial membrane more efficiently than the inner membrane.

**Figure 3.** CRAB C1 membrane destruction caused by PapMA analogs. The concentration dependent depolarization of (**A**) intact CRAB C1 and its (**B**) spheroplast induced by PapMA and its analogs, determined using the membrane potentialsensitive fluorescent dye diSC<sup>3</sup> -5. Dye release was monitored by measuring fluorescence, at an excitation wavelength of 654 nm and an emission wavelength of 670 nm.

#### *3.4. Cytotoxicities of PapMA Analogs*

To utilize AMPs as therapeutic agents, they should exhibit low toxicity against mammalian cells [15]. Antibiotics could cause kidney damage; polymyxins, the last-resort antibiotics to treat Gram-negative bacterial infections, have limited use due to its nephrotoxicity [52]. Therefore, to assess the cytotoxicity and to select safe candidates, the cytotoxicities of PapMA and its analogs were investigated against the HEK-293 cell line (Figure 4A). PapMA, PapMA-2, and PapMA-3 showed 100% survival rates at concentrations below <sup>64</sup> <sup>µ</sup>g·mL−<sup>1</sup> , whereas PapMA-4, PapMA-5, and PapMA-6 showed survival rates of 32.1, 21.4, and 34.8%, respectively, at 64 <sup>µ</sup>g·mL−<sup>1</sup> . These results show that cytotoxicity increased proportional to increasing hydrophobicity.

PapMA, PapMA-2, and PapMA-3 showed 100% survival rates at concentrations below <sup>64</sup> <sup>µ</sup>g·mL−<sup>1</sup> , whereas PapMA-4, PapMA-5, and PapMA-6 showed survival rates of 32.1, 21.4, and 34.8%, respectively, at 64 <sup>µ</sup>g·mL−<sup>1</sup> . These results show that cytotoxicity increased proportional to increasing hydrophobicity.

The hemolytic activity was analyzed against sheep red blood cells (sRBC; Figure 4B). The incubation of sRBC with 256 <sup>µ</sup>g·mL-1 for PapMA-4, -5, and -6 induced 1.4, 2.2, and 3.8% hemolysis, respectively. However, PapMA, -2, and -3 caused almost no hemolysis (much

lower than 1%). In contrast, melittin exhibited more than 90% hemolysis at <sup>32</sup> <sup>µ</sup>g·mL−<sup>1</sup> . These results also confirmed that increases in hydrophobicity led to increases in toxicity, through strong hydrophobic interactions occurred between the aromatic residues of peptides and phospholipids in the mammalian cell membranes (which have higher compositions of neutral phospholipids). Among all six peptides studied, PapMA-3 exhibited the highest bacterial cell selectivity, with a GM of 18.3 and a 100% survival rate at 64 <sup>µ</sup>g·mL−<sup>1</sup> in HEK-293 cells. Even though PapMA-4, -5, and -6 showed potent antibacterial activities, with GMs of 10.3–18.3, they showed severe cytotoxicity (<35% survival rates at 64 <sup>µ</sup>g·mL−<sup>1</sup> in HEK-293 cells). Therefore, PapMA-3 was selected as a candidate therapeutic peptide for further investigation.

**Figure 4.** Cytotoxicity of PapMA analogs. (**A**) Cytotoxicity of PapMA and its analogs against HEK-293 cell. The peptide was serially diluted and incubated with cells for 24 h. (**B**) Hemolysis activity of PapMA analogs against sRBC. The peptide was serially diluted and incubated with sRBC for 1 h with melittin as a control. Statistical analysis was performed using two-way ANOVA with Dunnett's comparison test. The values are expressed as the mean ± SEM of three independent experiments and are statistically significant at \*\* *p* < 0.01; and \*\*\* *p* < 0.001. ns, not significant.

#### *3.5. Synergistic Effects of PapMA-3 with Antibiotics against Five CRAB*

The appearance of CRAB has accelerated the usage of combination therapy as a new therapeutic approach for its treatment [53–55]. As PapMA-3 was selected as a candidate peptide antibiotic, the synergistic effects of PapMA-3 with conventional antibiotics were investigated using checkerboard assays against five clinically isolated CRAB (C1–C5) [35,56]. Regarding the combinations with front-line conventional antibiotics, imipenem and meropenem were used for Gram-negative infections. The synergistic effects of PapMA-3 were also investigated with rifampin, erythromycin, vancomycin, and linezolid, which are well-known antibiotics that are potent against Gram-positive bacteria.

The ability of PapMA-3 to facilitate these antibiotics in penetrating the bacterial membranes of Gram-negative bacteria was investigated. Each peptide was serially diluted to 1/16 from 1 MIC; the experiment was carried out by cross-mixing them. As shown in Figure 5 and Figure S1, the checkerboard assays revealed that PapMA-3 displayed an outstanding synergistic effect with all antibiotics against CRAB C1. PapMA-3 at 4 <sup>µ</sup>g·mL−<sup>1</sup> (1/4 MIC) displayed synergistic effects towards CRAB C1, exhibiting FICI values lower than 0.50 when combined with all six antibiotics (Table 3). PapMA-3 also showed synergistic

effects against CRAB C2 with imipenem (0.38), rifampin (0.25), vancomycin (0.25), and linezolid (0.50; Figure S2). For CRAB C3, PapMA-3 only showed a synergistic effect when combined with rifampin (FICI = 0.38; Figure S3). Among all cases, the combination of PapMA-3 (1 <sup>µ</sup>g·mL−<sup>1</sup> ) and rifampin (16 <sup>µ</sup>g·mL−<sup>1</sup> ) showed the most effective synergistic effect against CRAB C4, with a FICI value of 0.16 (Figure S4). Figure S5 shows that PapMA-3 demonstrated synergistic effects against CRAB C5 with rifampin, vancomycin, erythromycin, and linezolid; the FICI value was 0.5, respectively. Interestingly, combining PapMA-3 at 2 <sup>µ</sup>g·mL−<sup>1</sup> (1/8 MIC) and vancomycin at 32 <sup>µ</sup>g·mL−<sup>1</sup> (1/8 MIC) demonstrated an effective synergistic effect (FICI = 0.25) against both CRAB C1 and C2 (Figure S2). Antibiotics potent for Gram-positive bacteria, such as erythromycin, vancomycin, and linezolid, cannot pass the outer membrane barriers of Gram-negative bacteria; they only demonstrate antibacterial activity against Gram-positive bacteria. Combining PapMA-3 with these antibiotics demonstrated significant antibacterial effects on CRAB, confirming that the interaction of PapMA-3 with the Gram-negative CRAB membrane allowed these antibiotics to penetrate it.

**Figure 5.** Checkerboard assays of PapMA-3 in combination with six conventional antibiotics against CRAB C1. PapMA-3 and antibiotics were subjected to 1/2 dilution vertically and horizontally from the MIC concentration at the upper left corner. White (0.5 < FICI < 2) indicates a partial synergistic effect, yellow (FICI = 0.5) and orange (FICI < 0.5) indicate a synergistic effect, and gray indicates growth of bacteria. We defined MIC that inhibits completely over 99% of CRAB C1 bacterial growth.


**Table 3.** FICI for the synergistic effect of PapMA-3 in combination with antibiotics against CRABs.
