Identification of potential therapeutic antimicrobial peptides against Acinetobacter baumannii in a mouse model of pneumonia

Acinetobacter baumannii-induced nosocomial pneumonia has become a serious clinical problem because of high antibiotic resistance rates. Antimicrobial peptides (AMP) are an ideal alternative strategy due to their broad-spectrum of antimicrobial activity and low incidence of bacterial resistance. However, their application is limited by toxicity and stability in vivo. The present study used a mouse model to directly identify potential AMPs effective for treatment of A. baumannii-induced pneumonia. Fifty-eight AMPs were screened and two identified (SMAP-29 and TP4) to have prophylactic effects which prevented the death of mice with pneumonia. Furthermore, two TP4 derivatives (dN4 and dC4) were found to have therapeutic activity in pneumonia mouse models by peritoneal or intravenous administration. Both dN4 and dC4 also inhibited and/or eliminated A. baumannii biofilms at higher doses. Taken together, these data suggest the AMP derivatives dN4 and dC4 represent a potential treatment strategy for A. baumannii-induced pneumonia.

www.nature.com/scientificreports/ application of AMPs remain, such as potential toxicity to humans 21 , sensitivity to harsh environmental conditions (e.g., extreme pH), and susceptibility to proteases in the circulation 22,23 . Many AMPs work in vitro but fail in vivo. However, because AMPs are composed of amino acids, it is relatively easy to modify their structures in order to lower toxicity and improve stability against proteases 24 . In addition, although previous studies have identified several AMPs exhibiting antimicrobial activity against MDRAB in vitro 25,26 , only a few have tested the effects of AMPs in vivo, especially A. baumannii-induced pneumonia. Therefore, in the present study, we directly used a mouse model of A. baumannii-induced pneumonia to screen for AMPs that can effectively treat MDRAB in vivo. The AMPs having the greatest clinical therapeutic potential were identified in this study.
To evaluate the effect of AMPs, an infection dose of 5 × 10 8 CFU was used and mouse survival monitored. Since studies have indicated AMPs have cellular toxicity or lose antimicrobial activity in the circulation, the prophylactic effect of prescreened AMPs in vivo was examined by intratracheal administration in pneumonia mouse models. Among the 12 AMPs tested, only SMAP-29 and TP4 significantly prevented the mortality of A. baumannii-infected mice (Fig. 1A). Notably, the results of the in vitro time killing assay showed that SMAP-29 and TP4 can rapidly kill A. baumannii within 10 min (Fig. 1B).
In vitro antimicrobial activity of TP4 derivatives against laboratory and clinical MDRAB. SMAP-29 and TP4 have high potential for controlling microbial infections 39,40 . Both exhibit broad-spectrum activity against bacteria and show therapeutic activity in mice. However, their high cellular toxicity and hemolytic activity limit their application. Previous studies have enhanced the application potential of SMAP-29 by modifying its structure 41,42 . Similarly, the hemolytic activity of TP4 was reduced in our previous study by generating truncated and amino acid-substituted analogues 43 . The antimicrobial activities of these derivatives against Pseudomonas aeruginosa, Staphylococcus aureus, and Candida albicans were also investigated in the same previous study 43 . Here, the antimicrobial activity (MIC and MBC) of TP4 derivatives (dN2, dN4, dC2, dC4, F1A/I2A, H3A/  H4A, I5A/I6A, L9A/F10A, I16R, I16E, I16A, L19H/I20H, A12V/A15H, A12I/A15I, R18S/R21H, VHSH, and  TP3) was tested in vitro against laboratory and clinical MDRAB. Most of the TP4 derivatives that exhibited In vivo therapeutic effect of TP4 derivatives in A. baumannii-induced murine pneumonia. Peritoneal administration of TP4 reportedly exhibits a therapeutic effect in S. aureus-induced infection in mice 39 . Therefore, the therapeutic effect of peritoneal administration of TP4 derivatives in A. baumannii-induced pneumonia mouse models was assessed in the present study. Among 17 TP4 derivatives, six (dN4, dC4, H3A/H4A, L9A/F10A, I16R, and R18S/R21H) enhanced the survival rate of pneumonia mouse models (Fig. 2). Notably, dN4 and dC4 enhanced survival the most and had the lowest hemolytic activity 43 . In addition, dN4 and dC4 can also rapidly kill bacteria (Fig. 2B). Therefore, the therapeutic effect of intravenous dN4 and dC4 in mice was examined (Fig. 3). Intravenous administration of dN4 or dC4 4 h after inoculation of mice with A. baumannii reduced bacterial numbers inside the lung and enhanced the survival rate (Fig. 3A,B). Administration of dN4 or dC4 once a day for three days enhanced more mouse survival to chi-square significantly enhanced level (Fig. 3C). These data suggest the potential of dN4 and dC4 in the clinical control of A. baumannii-induced pneumonia.
dN4 and dC4 exhibited antimicrobial activity against A. baumannii biofilm. In addition to colonizing lung tissue, A. baumannii can form a biofilm on the surface of the endotracheal tube which may account for causing pneumonia in patients who require mechanical ventilation 3,7 . Therefore, the antimicrobial activity of dN4 and dC4 against A. baumannii biofilm were further tested. Both of dN4 and dC4 inhibited A. baumannii biofilm formation in a dose-dependent manner (Fig. 4A) and eliminated established biofilm (Fig. 4B,C). These data suggest that both dN4 and dC4 can prevent and/or eliminate A. baumannii biofilm on the surface of the endotracheal tube.

Discussion
MDR bacteria-induced infection is one of the toughest issues facing clinics today as several are currently untreatable, including MDRAB. Among alternative strategies under development, AMPs are ideal antimicrobial agents because they do not readily trigger resistance. However, application of AMPs is limited by their high cellular toxicity and loss of function in the circulation 23 ; hence, many AMPs work in vitro but fail in vivo. Therefore, the www.nature.com/scientificreports/  TP4  16  16  16  16  16  16  16  32   dN2  32  32  64  64  32  64  64  64   dN4  32  32  16  64  16  64  32  32   dC2  16  16  16  32  16  32  16  32   dC4  32  32  32  32  32  32  32  www.nature.com/scientificreports/ present study identified AMPs exhibiting potential to treat A. baumannii-induced pneumonia in mice. In particular, intravenous adminstration of TP4 derivatives dC4 or dN4 reduced A. baumannii conlonization in the lung and enhanced survival, suggesting preservation of antimicrobial activity in the circulation. In fact, dN4 and dC4 maintained, though reduced, antimicrobial activity against A. baumannii in the presence of plasma components in vitro (Supplementary Table 4), implying these derivatives may be more resistant to proteases in the plasma relative to other AMPs tested. In addition, our preliminary data showed that the LD 50 of SMAP-29 and TP4 is around 10-20 mg/kg by intratracheal administration and the LD 50 of dN4 and dC4 is approxmiately 80-90 mg/ kg and 90-100 mg/kg, respectively, by intravenous administration. However, additional large scale animal testing is needed to determine the cellular toxicity and stablity/half-life of dN4 and dC4 for pharmaceutical applications. Although intravenous administration of dN4 or dC4 enhanced survival, it should be noted that this required daily administration for three days to achieve significant results. On the other hand, the sample number was small as this was a pilot study conducted to identify AMPs with the potential to treat MDRAB in vivo. More studies with a larger sample size and more dose testing should be performed to confirm the clinical potential of dN4 and dC4. We previously showed that TP4 exhibits broad-spectrum activity against various bacteria, including P. aeruginosa 43 . TP4 binds the P. aeruginosa outer membrane target protein OprI through hydrophobic interactions. Mutation of hydrophobic residues in the main helix or deletion of nonstructural coils at N-and C-termini significantly reduced the hemolytic activity of TP4 and reduced the antimicrobial activity against P. aeruginosa 43 .

MIC (μg/ml) MBC (μg/ml) MIC (μg/ml) MBC (μg/ml) MIC (μg/ml) MBC (μg/ml) MIC (μg/ml) MBC (μg/ml)
The current results showed that most TP4 derivatives, which showed reduced hemolytic activity and reduced antimicrobial activity against P. aeruginosa 43 , maintained their antimicrobial activity against A. baumannii in the present study (Table 2). This suggests the antimicrobial mechanisms of TP4 against A. baumannii and P. aeruginosa are different. Although TP4 and many of its derivatives have in vitro antimicrobial activity against A. baumannii, only dN4 and dC4 showed clear protection against A. baumannii-induced pneumoniae in vivo (Fig. 2). This may be attributed to the significantly reduced hemolytic activity of dN4 and dC4 compared with TP4 and other TP4 derivatives 43 . In addition, dN4 and dC4 showed comparatively low antimicrobial activity (32 μg/mL) and low in vivo bactericidal effects (Fig. 3A) with a relatively low therapeutic dose, indicating an indirect antimicrobial effect of dN4 and dC4 in vivo. Moreover, our preliminary data showed that intratracheal administration of dN4 or dC4 can only delay death and not enhance survival, which also supports the indirect www.nature.com/scientificreports/ antimicrobial activity of dN4 and dC4 in the circulation. TP4 has been shown to modulate the host immune response thereby enhancing clearance of Helicobacter pylori infection 44 . Therefore, the indirect antimicrobial effect of dN4 and dC4 in vivo may be attributed to the immune modulation ability of these AMPs against A. baumannii infection, which will require further elucidation. In addition to colonizing lung tissue, A. baumannii form a biofilm on the surface of the endotracheal tube, accounting for pneumonia onset in patients requiring mechanical ventilation 3,7 . Biofilm-producing A. baumannii clinical isolates have also been shown to be more resistant to most antibiotics compared with www.nature.com/scientificreports/ non-biofilm-producing isolates 45 . Bacterial biofilm is well-organized and dense so as to support bacteria localization and growth on a specific surface and facilitate escape from immune surveillance and antimicrobial molecule attack 46 . Therefore, bacterial biofilms are even more resistant to antibiotic treatment versus platonic populations. Although the current results showed that the A. baumannii biofilm is more resistant to dN4 and dC4 than platonic populations, both derivatives can successfully inhibit A. baumannii biofilm formation in vitro (Fig. 4A) as well as eliminate established biofilms and the bacteria residing in them (Fig. 4B,C). These results suggest that dN4 and dC4 may be particularly useful for endotracheal tube-associated A. baumannii infections. However, the dose of 256 μg/mL was shown to eliminate almost the bacteria inside the biofilm, which is a little high for clinical use. Therefore, use of dN4 and dC4 to eradicate established biofilms on an endotracheal tube will require further evaluation.
In summary, the present study identified two AMPs (SMAP-29 and TP4) that exhibited in vitro and in vivo prophylactic effects against A. baumannii. Furthermore, TP4 derivatives dC4 and dN4 exhibiting reduced hemolytic activity but high antimicrobial activity against A. baumannii both in vitro and in vivo (intravenously and peritoneally). In addition, our preliminary data showed dN4 and dC4 also have antimicrobial activity against other bacteria, including antibiotic-resistant pathogens. Considering the extremely limited strategies available to treat MDR pathogen-induced infections, the clinical potential of these AMPs in other MDR pathogen-induced infections should be further evaluated.

Materials and methods
Bacterial strains and growth conditions. A. baumannii laboratory strains (ATCC 17,978 and 19,606) and clinical isolates from National Taiwan University Hospital (Taipei City, Taiwan) were grown and maintained in Luria-Bertani (LB) broth and agar plates. Medium containing 50 μg/mL kanamycin was used to propagate A. baumannii carrying plasmids that expressed green fluorescent protein (GFPuv).

MIC and MBC of AMPs.
All methods used in this study were carried out in accordance with relevant guidelines and regulations.
All AMPs and TP4 derivatives used in this study were synthesized by Kelowna International Scientific Inc. (Taipei City, Taiwan) with more than 95% purity, and their molecular weights were verified by mass spectrum analysis. MBC were determined by extending the MIC procedure in sterilized 96-well polypropylene microtiter plates according to the broth microdilution guidelines of the Clinical and Laboratory Standards Institute with some modification 47 . Briefly, 5 × 10 5 CFU/mL of bacteria were inoculated in LB medium containing AMPs or antibiotic at final concentrations ranging from 512 to 1 μg/mL (from serial twofold dilutions). After incubation for 24 h at 37 °C, the lowest concentration showing no visible growth was used as the MIC value. In addition, 10 µL was drawn from each well with no visible turbidity onto LB agar plates and incubated for 24 h at 37 °C. The lowest concentration showing no visible growth on agar subculture was used as the MBC value. All experiments were performed in triplicate.
Experimental A. baumannii-induced pneumonia mouse model. Approval for animal use was obtained from the National Taiwan University Institutional Animal Care and Use Committee prior to initiation of experiments. The study was carried out in compliance with the ARRIVE guidelines (https:// arriv eguid elines. org). The experimental A. baumannii-induced pneumonia mouse model was generated as previously described with some modification 38 . Male BALB/c mice that were 6-8 weeks-old were purchased from the National Laboratory Animal Center (Nangang Dist, Taipei City, Taiwan). Briefly, 5 × 10 8 CFU bacteria were intratracheally inoculated into mice and the survival rate was monitored daily for 10 d. To evaluate prophylactic effects of AMPs, 10 μL of 2 × MIC of each AMP was intratracheally administrated to mice 30 min before inoculation of bacteria. For evaluating the therapeutic effect of TP4 derivatives, 2 mg/kg of each derivative was peritoneally or intravenously administrated 4 h after bacterial infection or intravenously administrated once a day for three days.
To monitor bacterial colocalization in the lung, mice were sacrificed 1 d after bacterial infection. Dissected lung tissue was homogenized in a sterile disposable tissue grinder, serially diluted in phosphate-buffered saline (PBS), and plated on LB agar to determine the CFU/g of lung tissue. For histologic analysis, lung tissue was fixed with 2% paraformaldehyde, embedded in paraffin, cut into 3-μm-thick slices, and stained with hematoxylin and eosin. Immunohistochemistry was performed using primary antibodies against Ly6G (ab21595; Abcam, Cambridge, MA, USA) or F4/80 and developed with diaminobenzidine (Dako, Santa Clara, CA, USA) followed by counterstaining with Mayer's hematoxylin.

Construction of GFP-tagged A. baumannii. For construction of GFPuv-tagged A. baumannii, an A.
baumannii GFPuv-expressing plasmid was generated. Briefly, an A. baumannii promoter PgroEL from groEL of A. baumannii was amplified from A. baumannii chromosomal DNA using primers AB_groELPF_XbaI (5′ CGA TTA TCT AGA GGT GAA GAA CTC TTA ATT ATG-3′) and AB_groELPR_KpnI (5′-CGA TTA GGT ACC CTT GAA ACT TCA CGA ACA AG-3′). The polymerase chain reaction (PCR) product was digested with KpnI and XbaI and inserted upstream of the gene sequence of GFPuv in the pGFPuv plasmid (Clontech Laboratories, Inc., Mountain View, CA). The GFPuv coding sequence containing the promoter PgroEL was further PCR-amplified using primers AB_groELPF_XbaI and GFPR_XbaI (5′-CGA TTA TCT AGA TTA TTT GTA GAG CTC ATC CAT-3′). The PCR product was digested with XbaI and inserted into the E. coli-A. baumannii shuttle plasmid pAbYm3. The resulting plasmid, pABYm3-GFPuv, was transformed into A. baumannii, and positive clones were selected by green fluorescence following ultraviolet excitation (365 nm). www.nature.com/scientificreports/ Biofilm formation assay. Bacterial biofilm assay was performed as previous described 48 with modification. Briefly, approximately 10 7 CFU bacteria in 200 μL of LB medium was added to individual wells of a 96-well polystyrene microtiter plate. To investigate the inhibitory effect of AMPs on biofilm formation, serial concentrations of AMPs were added to the culture medium when inoculating bacteria. After a 16-h incubation at 37 °C, the medium was removed, and biofilms were detected by crystal violet staining assay 48 . Each assay was performed in triplicate, and wells without biofilms were used as blank controls. To investigate the effect of AMPs on established biofilm, the biofilms were grown in LB medium at 37 °C for 16 h. After removing the LB medium containing the planktonic populations of bacteria, the biofilms were washed with sterilized PBS three times, and the medium containing serial concentrations of AMPs was added. After a 16 h incubation at 37 °C, survival bacteria inside the biofilm were detected by plating on LB medium. For observation by confocal microscopy, the bacteria were transformed with pPABYm3-GFPuv. Bacterial biofilms were grown in 24-well plates with round glass coverslips. The biofilms with or without treatment with AMPs were observed using a confocal microscope (Leica TCS SP5).

Statistical analysis.
To analyze the statistical significance of differences between more than two sets of data, 1-way analysis of variance followed by Bonferroni multiple-comparison test was assessed. For analyzing nonparametrically distributed data, a Mann-Whitney U-test was used. Survival curves were compared with logrank tests. A P < 0.05 was considered statistically significant. www.nature.com/scientificreports/ Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.