Physical and Functional Characterization of PLGA Nanoparticles Containing the Antimicrobial Peptide SAAP-148

Synthetic antimicrobial and antibiofilm peptide (SAAP-148) commits significant antimicrobial activities against antimicrobial resistant (AMR) planktonic bacteria and biofilms. However, SAAP-148 is limited by its low selectivity index, i.e., ratio between cytotoxicity and antimicrobial activity, as well as its bioavailability at infection sites. We hypothesized that formulation of SAAP-148 in PLGA nanoparticles (SAAP-148 NPs) improves the selectivity index due to the sustained local release of the peptide. The aim of this study was to investigate the physical and functional characteristics of SAAP-148 NPs and to compare the selectivity index of the formulated peptide with that of the peptide in solution. SAAP-148 NPs displayed favorable physiochemical properties [size = 94.1 ± 23 nm, polydispersity index (PDI) = 0.08 ± 0.1, surface charge = 1.65 ± 0.1 mV, and encapsulation efficiency (EE) = 86.7 ± 0.3%] and sustained release of peptide for up to 21 days in PBS at 37 °C. The antibacterial and cytotoxicity studies showed that the selectivity index for SAAP-148 NPs was drastically increased, by 10-fold, regarding AMR Staphylococcus aureus and 20-fold regarding AMR Acinetobacter baumannii after 4 h. Interestingly, the antibiofilm activity of SAAP-148 NPs against AMR S. aureus and A. baumannii gradually increased overtime, suggesting a dose–effect relationship based on the peptide’s in vitro release profile. Using 3D human skin equivalents (HSEs), dual drug SAAP-148 NPs and the novel antibiotic halicin NPs provided a stronger antibacterial response against planktonic and cell-associated bacteria than SAAP-148 NPs but not halicin NPs after 24 h. Confocal laser scanning microscopy revealed the presence of SAAP-148 NPs on the top layers of the skin models in close proximity to AMR S. aureus at 24 h. Overall, SAAP-148 NPs present a promising yet challenging approach for further development as treatment against bacterial infections.


Introduction
Antimicrobial resistance (AMR) is a leading cause of death worldwide. The World Health Organization (WHO) has declared AMR as one of the top 10 global health threats facing humanity. In 2019, bacterial AMR was responsible for 1.27 million deaths worldwide, particularly in low-income countries [1,2], and the development of effective antimicrobial agents is a top priority in the fight against AMR bacterial infections. Hence, a library of synthetic novel antimicrobial peptide (AMP) analogues was designed, inspired by the sequence of endogenous human cathelicidin LL-37 and screened for antibacterial and anti-biofilm properties against AMR bacteria. The lead peptide, SAAP-148, underwent further testing against AMR bacteria belonging to the ESKAPE panel (Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, and Enterobacter species)-planktonic, in biofilms, or persisters-and tested in biologically

In Vitro Killing Activity
Firstly, we investigated the antimicrobial activity of SAAP-148 NPs and free peptides against planktonic bacteria consisting of AMR bacteria, namely S. aureus and A. baumannii (Table 1). Antimicrobial data revealed that SAAP-148 NPs reached the lethal concentration to kill 99.9% (LC99.9) of planktonic S. aureus and A. baumannii at 12.8 µM and 1.6 µM after 24 h, respectively. Moreover, SAAP-148 solution reached the LC99.9 between 1.6-3.2 µM and ≤0.8 µM against planktonic S. aureus and A. baumannii after 24 h, respectively

Antibacterial Activities of SAAP-148 NPs 2.2.1. In Vitro Killing Activity
Firstly, we investigated the antimicrobial activity of SAAP-148 NPs and free peptides against planktonic bacteria consisting of AMR bacteria, namely S. aureus and A. baumannii (Table 1). Antimicrobial data revealed that SAAP-148 NPs reached the lethal concentration to kill 99.9% (LC 99.9 ) of planktonic S. aureus and A. baumannii at 12.8 µM and 1.6 µM after 24 h, respectively. Moreover, SAAP-148 solution reached the LC 99.9 between 1.6-3.2 µM and ≤0.8 µM against planktonic S. aureus and A. baumannii after 24 h, respectively (Table 1). Expectedly, blank NPs did not induce any antimicrobial activity against any of the bacterial strains used in this study. In addition, SAAP-148 NPs induced a dose-dependent inhibition of planktonic AMR S. aureus and A. baumannii after 4 h, and they maintained this level of inhibition after 24 h (Supplementary Material Figures S1 and S2). Together, this data  9 concentrations for SAAP-148, in solution, was 25.6 µM regarding both bacterial biofilms for up to 24 h, with no improvement in killing activity at longer exposure times. As before, blank NPs did not exhibit any antibiofilm activity at any of the time-points tested. Together, the in vitro antibiofilm data suggests that increasing SAAP-148 NPs exposure time positively correlates with improved antibiofilm activity.

Cytotoxicity and Hemolytic Activity of SAAP-148 NPs Cytotoxicity against Human Skin Fibroblasts
Subsequently, we evaluated the cytotoxic and hemolytic activities of SAAP-148 NPs compared to free peptide (Figure 2a-d). Results showed that the encapsulation of SAAP-148 in PLGA NPs decreased the peptide's cytotoxicity by at least 41-fold and 24-fold against the primary skin fibroblasts after 4 and 24 h, respectively (Table S1). The effect of the metabolic activity of human skin fibroblasts was also improved by ≥37-fold and ≥128-fold, following exposure to SAAP-148 NPs, after 4 and 24 h, respectively. In agreement, encapsulation of peptide decreased the hemolytic activity by at least 10-fold against isolated human erythrocytes in PBS (Figure 3) Collectively, these results indicate that the encapsulation of SAAP-148 in PLGA NPs drastically decreased the cytotoxic and hemolytic activities of this peptide.

Selectivity Index of SAAP-148 NPs
Subsequently, we calculated the selectivity index of SAAP-148 NPs and free peptide based on the ratio between cytotoxicity and antimicrobial activity (IC 50 /LC 99 ) ( Figure 4). Data showed that the selectivity index of the SAAP-148 NPs was increased by 10-folds after 4 h incubation against S. aureus, and increasing the exposure time to 24 h induced a 3-fold increase compared to the free peptide. Similarly, the selectivity index of SAAP-148 NPs against planktonic A. baumannii improved by 20-fold after 4 h, and increasing the exposure time to 24 h induced a 12-fold increase compared to the free peptide. Furthermore, increasing the exposure time of SAAP-148 NPs significantly increased the selectivity index against S. aureus and A. baumannii biofilms compared to free peptide. More specifically, increasing the exposure time from 4 h to 72 h increased the selectivity index by almost 17-fold against S. aureus biofilm and by up to 12-fold against A. baumannii biofilm compared to the free peptide. Together, these results indicate that encapsulation of SAAP-148 in PLGA NPs substantially increased the selectivity index of the peptide. Figure 3. Hemolytic activity of SAAP-148 NPs (triangles) compared to SAAPand blank PLGA NPs (squares) in PBS after 1 h. Hemolytic activity is based human erythrocytes in PBS, after 1 h incubation at 37 °C, with 95% humidity a retrieved from the mean of three biological replicates performed with three te = 3). Hemolytic activity is based on 5% hemolysis of human erythrocytes.

Selectivity Index of SAAP-148 NPs
Subsequently, we calculated the selectivity index of SAAP-148 NP based on the ratio between cytotoxicity and antimicrobial activity (IC Data showed that the selectivity index of the SAAP-148 NPs was inc after 4 h incubation against S. aureus, and increasing the exposure time 3-fold increase compared to the free peptide. Similarly, the selectivity i NPs against planktonic A. baumannii improved by 20-fold after 4 h, exposure time to 24 h induced a 12-fold increase compared to the free more, increasing the exposure time of SAAP-148 NPs significantly incr and blank PLGA NPs (squares) in PBS after 1 h. Hemolytic activity is based on 5% hemolysis of human erythrocytes in PBS, after 1 h incubation at 37 • C, with 95% humidity and 5% CO 2 . Data are retrieved from the mean of three biological replicates performed with three technical duplicates (n = 3). Hemolytic activity is based on 5% hemolysis of human erythrocytes.

In Vitro Antimicrobial Activity in 3D HSEs
Subsequently, we studied the efficacy of SAAP-148 NPs in 3D HSE models infected with AMR S. aureus. Antibacterial data from the 3D HSEs showed that SAAP-148 NPs were significantly less effective in eliminating planktonic and cell-associated AMR S. aureus than the peptide solution. Hence, a dual drug delivery strategy was implemented, consisting of combined SAAP-148 NPs and halicin NPs as one treatment ( Figure 5). The data showed that SAAP-148 NPs and halicin NPs induced an improved antimicrobial response compared to SAAP-148 NPs alone against planktonic (p = 0.0005) and cell-associated bacteria (p = 0.0024), whereas their efficacy was similar to that of halicin NPs alone (p > 0.05) after 4 h (Figure 5a). After 24 h exposure, similar data were obtained regarding planktonic and cell-associated bacteria ( Figure 5b). Moreover, analysis of the data showed that the combined SAAP-148 NPs and halicin NPs induced a statistically significant reduction in cell-associated S. aureus compared PBS (p = 0.0018), while observed changes were not statistically significant for either SAAP-148 NPs (p = 0.99) or halicin NPs alone (p = 0.08) after 24 h. Interestingly, the data showed that halicin NPs exhibited a similar antimicrobial activity compared to the halicin solution against planktonic or cell-associated bacteria after 4 or 24 h (p > 0.05 for all). Together, the data indicate that SAAP-148 NPs and halicin NPs, as well as halicin NPs alone, but not SAAP-148 NPs alone, are effective against planktonic and cell-associated bacteria in the 3D HSEs.
OR PEER REVIEW 7 of 19 The data used to calculate the selectivity index was extracted from the median of three independent cytotoxicity and antibacterial activity experiments, each performed in triplicate.

In Vitro antimicrobial activity in 3D HSEs
Subsequently, we studied the efficacy of SAAP-148 NPs in 3D HSE models infected with AMR S. aureus. Antibacterial data from the 3D HSEs showed that SAAP-148 NPs were significantly less effective in eliminating planktonic and cell-associated AMR S. aureus than the peptide solution. Hence, a dual drug delivery strategy was implemented, consisting of combined SAAP-148 NPs and halicin NPs as one treatment ( Figure 5). The data showed that SAAP-148 NPs and halicin NPs induced an improved antimicrobial response compared to SAAP-148 NPs alone against planktonic (p = 0.0005) and cell-associ-

Discussion
The aim of this study was to enhance the selectivity index of SAAP-148 by formulating this peptide in PLGA NPs. The selectivity index is typically used as a pre-clinical

Discussion
The aim of this study was to enhance the selectivity index of SAAP-148 by formulating this peptide in PLGA NPs. The selectivity index is typically used as a pre-clinical screening strategy by determining the ratio of toxic concentration by its effective therapeutic concentration [15,16]. Previous studies have shown that a higher selectivity index is an early indicator of drug success and tends to decrease as it progresses through the development pipeline [17,18]. Although the selectivity index does not allow the extrapolation of doses administered in vivo [19], it is a relevant indicator to use in the screening procedure for further evaluation. Previous studies showed that the encapsulation of AMPs-physiochemically similar to SAAP-148-in PLGA NPs significantly decreased the peptide's cytotoxicity and increased its stability [13,20]. Based on these observations, we hypothesized that encapsulation of SAAP-148 in PLGA NPs would enhance the peptide's selectivity index by improving its stability and inducing controlled release of peptide.
An important functional characteristic of PLGA NPs pertains to its in vitro physiochemical properties, including particle size, surface area, surface chemistry, and release profile. Physiochemical analysis of SAAP-148 NPs revealed a uniform particle distribution with controlled release of peptide over 21 days. Generally, smaller sized nanoparticles (<200 nm) elicit higher antimicrobial activity due to a larger surface area-to-mass ratio, resulting in higher interactions on bacterial surfaces [21]. Furthermore, the release profile of peptide revealed a biphasic pattern consisting of a burst phase after several hours, followed by a sustained release phase lasting several days. Fitting the release profile into the Korsmeyer-Peppas model suggested that the release of peptide was governed by swelling and/or polymer chain relaxation [22,23]. This indicated that peptide release from PLGA NPs is only a function of time regardless of peptide concentration [24]. The higher amount of peptide released during the burst phase is beneficial, as it guarantees that SAAP-148 is rapidly bioavailable to directly interact with bacteria to induce membrane destabilization, whilst the sustained release phase maintains adequate peptide levels lasting for several days. Meanwhile, all the free peptide is directly bioavailable, resulting in greater cytotoxicity to mammalian cells.
Comparing the efficacy of SAAP-148 NPs and this peptide in solution revealed a reduction in antimicrobial and antibiofilm activities of the peptide in NPs against AMR A. baumannii and S. aureus. Interestingly, the SAAP-148 NPs demonstrated a higher antibiofilm activity after 72 h, as compared to 4 h, against AMR S. aureus and A. baumannii, suggesting a dose-response relationship based on the in vitro release data. Impressively, the cytotoxicity of SAAP-148 NPs against human skin fibroblasts and erythrocytes was significantly reduced compared to SAAP-148 solution. This data indicates that the NPs are very effective in shielding this peptide's cytotoxic activities to mammalian cells by reducing the electrostatic interactions to the membrane surface [25]. Combined, these results indicate that SAAP-148 NPs exhibited a significantly higher selectivity index compared to free peptide.
Furthermore, previous data from our laboratory showed that SAAP-148 and halicin induced a synergistic response against AMR planktonic and biofilm bacteria, and these favorable interactions were confirmed for AMR S. aureus infected 3D HSEs [26]. However, synergism was not confirmed in PLGA-formulated SAAP-148 and halicin NPs. To fill this knowledge gap, and because SAAP-148 NPs alone were poorly effective in eliminating bacteria from such models, we tested whether synergism could also be attained with the PLGA NP-formulated counterparts of SAAP-148 and halicin. For this reason, 3D HSEs infected with AMR S. aureus were exposed to combined SAAP-148 NPs and halicin NPs to evaluate the potential synergism in this model and compare them against their respective counterparts. Halicin is a powerful antibiotic that was identified by a machine-learning algorithm using artificial intelligence, and it has low toxicity to human cells [27]. Halicin is a c-Jun N-terminal protein kinase (JNK) inhibitor, which disrupts the electrochemical gradient of bacteria necessary for growth [28]. Combined, SAAP-148 NPs and halicin NPs displayed better antibacterial activity than SAAP-148 NPs alone but showed similar antibacterial activity as halicin NPs alone against planktonic and cell-associated S. aureus in 3D HSE models. It is possible that differences in release of SAAP-148 and halicin from the NPs could explain the absence of synergism between these antibacterial agents at 24 h. Using a fast-release drug delivery system for the delivery of SAAP-148, such as hyaluronic acid nanogels, could improve the therapeutic activity and attain synergism in this setup [29]. Nevertheless, combined, SAAP-148 NPs and halicin NPs were slightly more effective in eradicating cell-associated S. aureus than either SAAP-148 NPs or halicin NPs alone after 24 h. Notably, halicin NPs exhibited a similar antibacterial activity as halicin in solution, suggesting that the physiochemical properties of the encapsulated biomolecule could play a crucial role in determining its antibacterial activity when encapsulated in NPs.
Subsequently, we compared the localization pattern of fluorescently labeled SAAP-148 NPs or peptide solution with GFP-MRSA in the 3D HSE models. Confocal micrographs revealed localization of SAAP-148 NPs and SAAP-148 solution within the proximity of bacteria predominantly situated in the top layers of the epidermal models. Cross-sectional micrographs revealed that SAAP-148 in solution associated better with keratinocytes than SAAP-148 NPs, which was expected since peptide in solution is more readily taken up by the keratinocytes compared to slow releasing peptide NPs [30]. A possible explanation for the formulated peptide's limited efficacy in the skin models is the anionic surface charge of NPs. Previous studies showed that anionic NPs elicit an electrostatic barrier, which limits their adherence to the surface of negatively charged mammalian cells [31][32][33]. Hence, cationic NPs may overcome the initial repulsion forces with the keratinocytes and enhance peptide release into the extracellular milieu [34]: for instance, by incorporating charged groups (amino or carboxyl) into a polymer backbone or by coating the NPs with positively charged polymers/proteins (e.g., polyethylenimine, bovine serum albumin) [35]. However, cationic NPs are limited by their cytotoxicity due to greater plasma membrane destabilization than their anionic counterparts [36]. Hence, further research is needed to fully study the influence of surface charge in relation to the biological effects.
To further improve the efficacy of SAAP-148 NPs, surface functionalization to specifically target bacteria, most specifically biofilms, could be considered. For example, Eissa et al.
(2016) reported on carbohydrate/glycan functionalized polymeric NPs that prevented bacterial adhesion by blocking carbohydrate structures on mammalian cell surface [37]. Tan et al.
(2022) synthesized chitosan nanoparticles that were functionalized with β-1,3-glucanase to degrade the Candida albicans biofilm matrix, which is fundamental for microbial growth and drug resistance [38]. Ivanova et al. (2018) functionalized inert poly(methyl vinyl ether/maleic) acid NPs with the antimicrobial aminocellulose and antifouling hyaluronic acid using layer-by-layer coating technology [39]. Furthermore, Ucak et al. (2020) showed that aptamer-functionalized vancomycin-encapsulated PLGA NPs bind selectively to bacterial cell surface antigens [40,41]. Alternatively, NPs could be functionalized to target (infected) host cells. For instance, PLGA fibers cross-linked to collagen showed enhanced skin cell adhesion by promoting integration [42]. PLGA conjugated to transferrin were shown to selectively target transferrin receptors, which are highly expressed by keratinocytes [43]. To this end, surface functionalization of PLGA NPs presents a promising way to specifically target bacterial biofilms and/or further enhance their association with colonized/infected host cells

Preparation of PLGA NPs
PLGA NPs were formulated using the solvent evaporation double emulsion waterin-oil-in-water (W/O/W) technique using a previously described method [45,46]. Briefly, 50 mg of PLGA was dissolved in 1.5 mL of ethyl acetate. SAAP-148 was dissolved in Milli-Q water at a concentration of 2 mg/mL. Additionally, 0.5 mL of SAAP-148 2 mg/mL solution was added dropwise while vortexing at 1000 rpm to the dissolved PLGA, and then, it was sonicated at 25% amplitude for 1 min (20 s on/off cycle) on ice water to form the primary emulsion. Subsequently, 2.5 mL of PVA solution (25 mg/mL) was added dropwise to the primary emulsion while vortexing at 1000 rpm, and then, it was sonicated again at the same setting. Organic solvent was removed by evaporation using a rotatory evaporator for 1 h at 100 mbar at room temperature (RT). The formulated NPs were then washed in Milli-Q and lyophilized overnight before storage at −20 • C.

Peptide Analysis with Ultra-High Performance Liquid Chromatography
SAAP-148 was detected by ultra-high performance liquid chromatography (UPLC, Waters, Milford, MA, USA) equipped with a BEH C 18 , 1.7 µM, 130 Å, 2.1 × 100 mm column over 8 min with a linear gradient of two buffers: (A) Milli-Q and (B) acetonitrile, both containing 0.05% trifluoroacetic acid. Starting at a ratio of 95% A and 5% B to 25% A and 75% B and ending after 8 min at a flow rate of 0.5 mL/min and UV detection of 214 nm, data were analyzed using Mass-Lynx Software V4.2.

Encapsulation Efficiency and Drug Loading
The amount of encapsulated peptide was calculated by measuring the amount of free peptide using an indirect method adapted from the literature [47]. After lyophilization, the SAAP-148 NPs were suspended in 1 mL of PBS in 2 mL microtube, and then, they were mixed by vortexing for 10 min to extract the free peptide in the aqueous phase. Subsequently, the suspension was centrifuged under 10,000 rpm for 30 min at RT to separate the polymeric debris and the amount of free peptide was measured in the supernatant with UPLC, as previously described (Section 4.3).
The encapsulation efficiency (EE) of SAAP-148 was calculated by the following equation:

%EE :
Total The in vitro release profile of SAAP-148 from PLGA NPs was analyzed with SpectraPor ® Float-A-Lyzer ® membrane G2 dialysis device with MWCO of 100 kD. The membranes were soaked in Milli-Q before adding 10% ethanol to remove present salts in the dialysis instrument. The membranes were then washed with Milli-Q before coating with 1 mg/mL lysozyme, for 1 h, while stirring at 37 • C. The lysozyme was then removed, and the membrane was washed again with Milli-Q. Then, 1 mL of 100 µM SAAP-148 NPs were added to the inner well, and 5 mL of PBS was added to the outer well. At each time-point, 200 µL of fluid was removed from the outer compartment and replaced with fresh PBS of the same volume. All samples were stored at −20 • C until further analysis with UHPLC.

Cryo Electron Microscopy (cryo-EM)
The 300 mesh EM grids (Quantifoil R2/2, Jena, Germany) were glow-discharged by 0.2 mbar air for two minutes using the glow discharger unit of an EMITECH K950X. The 3 µL of the sample was applied per glow-discharged grid, and the grid was vitrified using an EMGP (Leica, Wetzlar, Germany) at room temperature and 100% humidity. For vitrification, excess sample was removed by blotting for one second to Whatman #1 filter paper directly followed by plunging the grid into liquid ethane (−183 • C), after which the grid was stored under liquid nitrogen until further use. Cryo EM imaging was performed at 120 kV on a Tecnai 12 electron microscope (FEI Company, Eindhoven, The Netherlands) after mounting the grid in a Gatan 626 cryo-holder. A 4 k × 4 k Eagle camera (FEI Company) was used to record images with focus between 5-10 µm and an 18,000× magnification (pixel size 1.2 nm).

Antibacterial Killing Studies
Colonies of Gram negative (A. baumannii RUH875) and Gram positive (methicillin resistant S. aureus LUH14616; NCCB100829) were incubated for 2.5 h in TSB to midlogarithmic phase. The mid-logarithmic phase bacteria were washed and suspended in PBS before measuring the optical density at 600 nm with a spectrometer. Bacteria were then diluted in PBS to a concentration of 5 × 10 6 colony-forming units (CFU)/mL and kept on ice. To each 96-well, 30 µL of SAAP-148, SAAP-148 NP, or blank PLGA NPs was added. To make a final plasma concentration of 50%, 50 µL/well of human plasma was added, followed by 20 µL of bacterial suspension previously prepared. The plate was then covered with a plastic sealer and lid and mixed for 10 s using a plate shaker, and it was centrifuged for 1 min at 12,000 rom at RT. The plate was then incubated for 4 or 24 h at 37 • C. Each well was then diluted 10× in PBS and plated on a Mueller-Hinton agar plate. Then, they were incubated overnight at 37 • C. The bacterial colonies were manually counted and converted to CFU/mL. All bacterial studies performed under sterile conditions.

Anti-Biofilm Studies
Bacteria were grown to the mid-logarithmic phase in brain heart infusion broth (BHI). Then, 100 µL of 1 × 10 7 CFU/mL in BHI was added in a number of wells of a flat-bottom polypropylene plate. The plates were covered with breathable seals and incubated for 24 h at 37 • C in a humidified controlled environment. The bacterial suspension was then discarded, and the wells were washed 2 × 100 µL PBS before treatment with peptide. The plate was then covered with a sealer and incubated for either 4, 24, or 72 h at 37 • C under the same humidified environment. The treatment wells were then washed twice with 100 µL PBS, covered with an aluminum seal, and placed in a waterproof plastic bag before sonicating for 10 min to dislodge the biofilm. To further disrupt the biofilm, vigorous pipetting was introduced. The biofilm colonies were diluted and plated on a Mueller-Hinton plate and converted to CFU/mL.

Cytotoxic Activities of SAAP-148 NPs
Human skin fibroblasts (HSFs) were cultured in culture flasks and sustained with DMEM supplemented with 1% (v/v) GlutaMAX™, 1% (v/v) P/S, and 5% (v/v) FCS. HSFs were seeded at a density of 20,000 cells in DMEM medium, supplemented with 1% P/S and 0.5% human serum, and incubated overnight at 37 • C at 5% CO 2 . Peptide dilutions were made in the DMEM medium supplemented with 1% P/S and 0.5% human serum. Additionally, 75 µL of the medium from the wells were removed before adding 25 µL of peptide dilutions to the remaining 25 µL medium in the wells. As a negative control, 25 µL medium without peptide was used. As a positive control, 25 µL of 1% Triton in the medium was used. The treated cells were returned to the incubator for either 4 or 24 h. At the end of the incubation period, 25 µL of the supernatant was added to a new round bottom 96-well plate to measure LDH release.

Cytotoxicity Based on LDH Release
The collected supernatant (or subnatants from 3D human skin equivalents; Supplementary Material, Figure S3) was centrifuged at 2000 rpm for 3 min and diluted 1:10, in PBS, in a 96-well flat bottom plate. The LDH reaction mix was added at 50 µL per well according to the manufacturer's instructions, the diluted plate was incubated at RT for 30 min, and then, the absorbance was read at 490 nm and 600 nm as a reference.

Cell Metabolism Based on WST-1
To the remaining cells, 20 µL of fresh medium, supplemented with 0.5% human serum, was added per well. Then, 5 µL of WST reagent was added according to the manufacturer's instructions. The cells were incubated at 37 • C, for 30-60 min, until a color change could be observed. The absorption was measured at 440 nm, and 590 nm was used a reference wavelength. The data provided is based on % of the positive control (LDH: Triton X-100100, WST-1: PBS) calculated using the following formula: % Cytotoxicity or metabolic activity = test sample − negative control positive control − negative control × 100 4.6.3. Hemolytic Activity Whole blood was obtained from healthy volunteers in citrate tubes and spun for 10 min at 3000 rom to pellet the erythrocytes (Sanquin, NVTO128.02, with informed consent). The erythrocyte pellet was then transferred to a plastic 15 mL tube, and the cells were washed 3× with PBS by vortexing, then spinning at 1000 rpm, and then discarding the supernatant. The erythrocyte suspension was then diluted to 20% in PBS and, then, in PBS to a 2% erythrocyte suspension. Stock peptide and SAAP-148 NPs solutions were diluted separately to their respective concentrations in Milli-Q, using a 96-well polypropylene V-bottom plate, to a final volume of 25 µL. As positive control, 25 µL of Triton-X was used instead. Then, 50 µL of PBS was added, followed by 25 µL of the 2% erythrocyte suspension, and the plate was covered with a lid and mixed by shaking for 10 s at 300 rpm before incubation at 1 h at 37 • C. Subsequently, the plan was spun for 3 min at 1200 rpm, and 85 µL of the supernatant was transferred to a 96-well flat-bottom plate to measure the optical density at 415 nm. The following formula was used to determine the % hemolysis based on the manufacturer's instructions: IC 50 LC 99.9 % hemolysis = test sample − negative control positive control − negative control × 100

Infection and Treatment of 3D HSE Models
GFP-MRSA (USA300 JE2) was transferred from the glycerol stock to TSB containing 50 µg/mL tetracycline for overnight culture at 37 • C. The next day, the bacterial culture was split and grown for an additional 2.5 h, to the mid-log phase at 37 • C, before washing and optical density measurement. On the day of the infection, any residual liquid above the model was removed. The models were then infected by adding 300 µL of 3.3 × 10 5 CFU/mL AMR S. aureus (10 5 CFU/model) or GFP-MRSA for confocal microscopy experiments and incubated for 1 h at 37 • C. As a control, the inoculum was diluted and counted. After 1 h, the supernatant was removed, and the models were washed 1× with 500 µL PBS. Then, two models were retained to measure the starting planktonic and cell-associated bacterial concentration after 1 h (T0). Thereafter, 300 µL of the desired peptide solution (plain or fluorescently labeled peptide) was added to the model and incubated for 24 h at 37 • C and 5% CO 2 . After 24 h of incubation, the supernatant was removed and stored for the counting of the planktonic fraction. The skin models were extracted by carefully removing the skin models from the inserts with a scalpel and transferring them to a potter system with 0.5 mL of PBS. To extract the cell-associated bacteria fraction, the models were homogenized with a bead-beater (Precellys 24 tissue homogenizer, Bertin Technologies, Montigny-le-Bretonneux, France) for 3 × 5000 rpm for 10 s with 10 s pause. The planktonic and cell-associated bacterial fractions that were separately diluted in PBS and CFU on MH agar plates were stored overnight in 37 • C overnight, and colonies were manually counted.

Confocal Microscopy
The infected and treated 3D HSE models were washed with PBS before fixing with 1% paraformaldehyde (PFA), 500 µL apically and 1 mL basally, and incubated for 1 h at 4 • C. The models were washed with PBS and stored overnight in PBS at 4 • C. The skin models were, then, blocked with 1% BSA + 0.3% Triton X-100 in PBS (PBT) for 10 min at RT. The plastic inserts around the models were, then, carefully removed with a scalpel. The skin models were, then, stained with 50 µL of 1:50 diluted Alexa Fluor™ 405 dye Phalloidin (ThermoFischer Scientific) in PBT solution for 2 h at 4 • C to stain filamentous actin (F-actin) of the keratinocytes. The models were then washed 3× in PBS and 3× in Milli-Q before they were mounted on a glass microscope slide with the epidermis facing upwards. Diamond Antifade mountant was added to each sample, and a cover slip was placed on top. Mounted samples were kept at RT in the dark until imaging. All samples were imaged on a Leica SP8 up right confocal microscope, 3D images were produced using Leica SP8 WLL-2 inverted confocal microscope at 40× magnification, and images were process using LASX software version 3.7.6 (Leica Microsystems, Wetzlar, Germany).

Statistics
Statistical significance between groups was evaluated by the nonparametric Kruskal-Wallis test with Dunn's post-hoc test with GraphPad Prism software version 9.3.1 (Graph-Pad Software, San Diego, CA, USA). Differences between groups were considered statistically significant at p ≤ 0.05.

Conclusions
To conclude, SAAP-148 NPs were successfully formulated with favorable physiochemical properties and the sustained release of peptide for up to 21 days. Combined studies of the antibacterial and cytotoxic activities of SAAP-148 NPs demonstrated higher selectivity index SAAP-148 in PLGA NPs compared to the free peptide. Antibiofilm studies indicated that SAAP-148 NPs exert optimal antibacterial activity after 72 h, suggesting a dose-response relationship, which corroborates the in vitro release data. Unfortunately, SAAP-148 NPs were not as effective as free peptide in eliminating bacteria from colonized 3D HSE, thus requiring further optimization of these NPs. Together, PLGA NPs present an interesting approach to further development as a treatment to combat AMR bacterial infections that require long-term sustained local release of peptide.

Patents
The SAAP-148 peptide is patented (WO2015/088344) by Leiden University Medical Center with co-inventors J.W.D. and P.H.N.

Data Availability Statement:
The data presented in this study are available from the corresponding author upon request.