Introduction

Around 75% of adult cystic fibrosis (CF) patients suffer from the invasion of mucoid, alginate-producing strains of Pseudomonas aeruginosa that wreaks havoc upon the airways leading to accelerated deterioration of pulmonary function and ending in substantial morbidity.1 The mucoid phenotype of P. aeruginosa – being pathognomonic for the disease – poses multiple challenges to effective treatment. Pathogenicity traits allowing P. aeruginosa to endure, proliferate, and profoundly affect all stages of CF pulmonary disease can be predicated by multiple virulence factors and biofilm formation. In addition to the mechanical stability of mature P. aeruginosa biofilms persisting in the respiratory tract, alginate overproduction is involved in the establishment of microcolonies at the beginning of biofilm formation, in addition to mechanical stability of mature P. aeruginosa biofilms persisting in the respiratory tract. P. aeruginosa biofilm is a self-produced matrix of polysaccharides (eg, alginate), proteins, nucleic acids and lipids forming a three-dimensional network responsible for interconnection of a core of inactive “persister” cells and mediating biofilm adherence to surfaces, gel formation and mechanical stability.2 These nonmotile polymicrobial aggregates − elsewhere referred to as “capsulated P. aeruginosa” – are less sensitive to host defenses and possess up to 1000-fold increased tolerance toward antipseudomonal drugs, thus, herald deleterious effects on CF lung.3

Alginate (a high-molecular-mass exopolysaccharide, unbranched heteropolymer) consists of 1,4-linked uronic residues, β-d-mannuronate (M) and α-l-guluronate (G). These components are arranged in homopolymeric blocks of polymannuronate and heteropolymeric sequences with a random distribution of guluronate and partially O-acetylated mannuronate residues.2 As a key component of P. aeruginosa biofilm, alginate plays a key pathogenic role through interrupting the phagocytosis of macrophages and neutrophils, restricting lymphocyte function, and inducing inflammatory reactions, thus, aggravating lung infection.1,4 Specifically, O-acetylation of the M residues at the C2 and/or C3 positions of alginate synthesized by P. aeruginosa ameliorates alginate resistance toward degradation, triggering serious resistance of P. aeruginosa to phagocytic cells and even to aggressive antibiotic regimen.5

Owing to the complexity of the biofilm matrix, the complete degradation of all components compels a range of enzymes, alginate lyase (AL). AL is reported to degrade alginate by cleaving the glycosidic bond via a beta-elimination reaction, yielding oligomers with 4-deoxy-L-erythro-hex-4-enepyranosyluronate at the nonreducing end.6 AL can be classified as endo- and exo-type lyases according to the mode of the cleavage site, where they commonly possess endo-type lyase activity. Currently, most AL ensure poly M-specific lyase activity, preferably by degrading the poly M-alginate block.6 Reduced alginate levels consequently increase the innate immune response and promote phagocytosis.

Several strategies have been developed to eradicate P. aeruginosa biofilms and overcome antibiotic resistance to manage lung infections in CF patients. In the last 25 years, antibiotics have been encapsulated into micelles, liposomes,3,7 chitosan nanoparticles (CS NPs),8,9 PLGA NPs10 and lipid nanocapsules.11 Despite the progress of antibiotic-loaded nanoparticles, broad biofilm eradication could not be achieved presumably due to low antibiotic exposure in the vicinity of P. aeruginosa biofilms.12 Alipour et al13 evidenced that liposomal aminoglycosides (tobramycin, gentamicin and amikacin) did not fare better antipseudomonal activity than conventional forms against P. aeruginosa in CF sputum; the latter overwhelmingly blocked the transport of the nanosystems.7 Some studies even reported (2–8-fold) lower activity of liposomal than free antibiotics. This has led to the development of adjuvant therapeutic agents that can act alongside conventional antibiotics, thereby potentiating their action.

Understanding the host–pathogen interface elucidates the adaptive mechanisms empowering P. aeruginosa evasion of host immunity and spurs the development of novel therapeutics. To date, innovative approaches include the disruption of the DNA−Ca2+−DNA bridges of P. aeruginosa biofilms using lytic enzymes such as recombinant human deoxyribonuclease (rhDNase) to cleave DNA and decrease sputum viscosity, the use of lectin inhibitors to control bacterial adhesion and biofilm formation, in addition to bacteriophages targeting the Pseudomonas genus, quorum-sensing inhibitors (QSI), and antimicrobial peptides.14–18 Reported challenges for these strategies include poor stability, proteolytic sensitivity, immunogenicity, hemolytic activity, bacterial resistance, and high cost.9

P. aeruginosa regulates much of its virulence via quorum-sensing (QS), a mechanism by which individual bacteria communicate with each other through the production and detection of small signal molecules.19,20P. aeruginosa produces Pseudomonas quinolone signal (PQS), which activates PqsR (also referred to as a multiple virulence factor regulator), which drives the coordinated expression of nearly 200 genes related to virulence factors. Based on the fact that PQS is overproduced in CF sputum, blocking the PQS-QS system or “quorum-quenching” should make P. aeruginosa lung infection in CF patients better treatable. Our collaborators demonstrated the potential of various PqsR antagonists as QSI for disrupting cell-to-cell communication and biofilm formation.21–25 Yet, reduction of mucociliary clearance, together with P. aeruginosa biofilm colonization, guard the cargo against reaching its site of action. Impaired penetration could be overcome via mucus-penetrating solid lipid nanoparticles (SLNs), which ensured up to 7-fold superior anti-virulence activity relative to free QSI.26 Moreover, combination of QSI with tobramycin in self-assembled squalenyl NPs demonstrated a significant synergistic effect.24

In line with that, chitosan proved potential antimicrobial and biofilm-disrupting activity.27 The electrostatic interactions of the cationic CS with anionic phospholipids on the bacterial cell membrane, lipopolysaccharides on the cell wall of Gram negative bacteria and/or teichoic acids in Gram positive bacteria, destabilize the cell envelope.27 Coating the surface of SLNs with CS was thus expected to ameliorate the outcome of the system.

An emerging therapeutic approach encompasses the use of AL to destroy alginate with the aim of effectively dispersing biofilm architecture, thus enabling better antibiotic penetration.13 A promising protocol involves the co-administration of AL with antibiotics to increase bacterial sensitivity towards antimicrobials. Studies proved potential elimination of mucoid bacteria using either gentamicin or ceftazidime with AL28 and applying AL to aminoglycoside antibiotics limited P. aeruginosa biofilm growth and bacterial count in CF sputum.13 Being a biodegradable enzyme, negligible hazards can be speculated. Owing to AL poor thermal stability,9 strategies to stabilize the enzyme included introduction of a disulfide bond29 and the use of glycerol as stabilizer.30 Immobilization of AL onto/into NPs was recently introduced to shield against thermal damage. Nanocomposites containing ceftazidime-, AL-loaded silver NPs and thioether-bridged mesoporous organosilica exhibited strong growth inhibition of PAO1 strain and biofilm degradation.31 In parallel, ciprofloxacin-loaded CS-NPs functionalized with AL reduced biofilm thickness, biomass and density.8

Herein, to bolster the rationale of the QSI-loaded SLNs therapeutic approach, the particles were surface-functionalized with AL and CS coatings. SLNs were then spray-dried to obtain respirable microparticles. Following physicochemical characterization, penetration of the artificial biofilm sputum medium, viability, and uptake in pulmonary cell lines were investigated. The potential of SLNs bearing biofilm-degrading enzyme functionalities to kill P. aeruginosa, prohibit biofilm formation, or disassemble the biofilm matrix was assessed in reference strains and clinical isolates.

Materials and Methods Materials

The quorum-sensing inhibitor (QSI) used in this study was previously reported as compound 3 by Lu et al.23 It is an alkylquinolone-derived inverse agonist of the QS receptor PqsR with the IUPAC name 2-heptyl-6-nitro-4-oxo-1,4-dihydroquinoline-3-carboxamide and has been synthesized and characterized according to previously reported procedures.23,32 Gelucire (43/01) was donated by Gattefossé, Saint Priest, France. Poloxamer 407 (PX) was obtained from BASF (Germany). Low molecular weight ultrapure chitosan chloride (Protasan UP CL 113) and sodium alginate (Pronova ultrapure alginate MVM, <75 kDa, intermediate viscosity >200 mPa*s) were purchased from (Novamatrix, FMC, Sandvika, Norway). Hydroxyethyl cellulose 250M (HEC) was obtained from SE Tylose GmbH & Co.KG, Wiesbaden, Germany). Alginate lyase, coumarin-6 (cou), Calcofluor-White, MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide, mucin from porcine stomach type II, DNA (low molecular weight from salmon sperm), Trizma base, sodium chloride, potassium chloride, 4′,6-diamidino-2-phenylindole (DAPI), and diethylenetriaminepentaacetic acid (DTPA) were purchased from Sigma–Aldrich (Steinheim, Germany). Egg yolk emulsion was obtained from Oxoid Ltd (Hampshire, UK). Bacto Casamino acid was purchased from Difco Laboratories Becton Franklin Lakes, NJ, USA). Alexa Fluor 633 wheat germ agglutinin (WGA) was obtained from Invitrogen (Carlsbad, USA. The live/dead BacLight Bacterial Viability Kit was purchased from Molecular Probes Inc. (Oregon, USA). All other reagents and solvents used were of analytical grade.

Methods Preparation and Characterization of SLNs

SLNs were prepared via a hot homogenization technique using Gelucire 43/1 as a lipid and poloxamer 407 as a stabilizer, as detailed in the Supplementary Materials. Positively charged particles were prepared by adding chitosan (0.05% w/v) to PX solution prior to particle preparation. The selected SLNs were loaded with QSI (50 µg/mL SLNs). For the uptake and biofilm penetration studies, the QSI was substituted with coumarin-6 (30 µg/mL SLNs).

AL-modified SLNs were obtained by adding an aqueous AL solution to a preformulated SLN dispersion (25 U/mL SLNs). Table 1 lists the different SLN formulations used in this study.

Table 1 Composition* and Colloidal Properties of SLNs

SLNs were characterized in terms of size, polydispersity index (PdI) and ζ-potential in addition to morphological examination by TEM, yield, and pH measurements. Thermal properties were investigated by DSC, and surface functionalization with alginate lyase was verified by Fourier-transform infrared (FT-IR) spectroscopy. The experiments are described in the Supplementary Materials.

Entrapment of QSI and in vitro Release

QSI-loaded SLNs were purified by centrifugal ultrafiltration using Centrisart-I® (MWCO 10 kDa, Sartorius AG, Goettingen, Germany) to get rid of free QSI. The concentration of QSI entrapped in SLNs was quantified by liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS), as previously described.26 The entrapment efficiency (EE) and loading efficiency (LE) of QSI were determined after payload extraction from purified loaded particles in methanol/dichloromethane.

The release of QSI from SLNs was investigated in 50 mL simulated lung fluid (SLF; pH 7.4) at physiological temperature for 24 h, as described in the Supplementary Materials. The release profiles were fitted to different kinetic models to determine the relevant release mechanisms.

Cell Viability (MTT Assay)

The effects of different concentrations of plain and QSI-loaded SLNs (±CS, ±AL) on human alveolar lung carcinoma, A549 (CCL-185; ATCC, Manassas, VA, USA), and bronchial Calu-3 (HTB-55; ATCC, Manassas, VA, USA) cell lines were investigated using an MTT viability assay (Supplementary Materials). Cytotoxic concentrations (CC50) of the tested formulations were determined.

Uptake of SLNs in Pulmonary Cell Lines

To examine the effects of particle charge and surface modification with AL on the cellular uptake of SLNs, qualitative and quantitative in vitro uptake experiments were performed using fluorescently labeled SLNs (Cou-SLNs). Qualitatively, particle internalization after 6 and 24 h was monitored using CLSM 3D-time-lapse imaging of the z-stacks. Data evaluation and 3D reconstruction were performed using ZEN software (Zeiss, Germany). The quantitative uptake efficiency was determined in terms of the fluorescence intensity (I) of the lysed cells, following the protocol described by Shi et al33 (Supplementary Materials).

Penetration of SLNs in Artificial Biofilm-Sputum Model (ABSM)

Artificial Sputum Medium (ASM) was prepared as reported by Kirchner et al34 and supplemented with hydroxyethyl cellulose (1% w/v) to impart high viscosity35 in addition to sodium alginate (1.5% w/v) to mimic the biofilm extracellular matrix and to investigate the effect of alginate lyase on particle penetration.

The diffusion of labeled SLNs was evaluated by the Transwell permeation study (the procedure was modified based on Friedl et al36 and Torge et al11) and 3D-time laps imaging using CLSM (the technique was established in-house as previously described by Nafee et al35), as detailed in the Supplementary Materials.

Preparation and Characterization of Respirable SLN-Embedded Microparticles (MPs)

To prepare inhalable microparticles, cationic and anionic coumarin-labeled SLNs were spray-dried in the presence of carrier system blends of mannitol, maltodextrin, and leucine at a carrier: SLN ratio of 3:1 (w/w). The MP yield and dye content were measured as described in the Supplementary Materials. The size and morphology of MPs were investigated using SEM. To investigate the effect of spray drying on SLN stability, the size of SLNs after mixing with the carrier solution prior to spray drying and that of the recovered SLNs after MP reconstitution was measured. The aerodynamic behavior of spray-dried MPs was investigated using a next-generation impactor (NGI). The relevant in vitro aerosolization parameters were determined, including the mass median aerodynamic diameter (MMAD), emitted dose (ED), emitted dose fraction (EF), fine particle dose (FPD), and fine particle fraction (FPF), in addition to the geometric standard deviation (GSD), as detailed elsewhere.37,38

Antimicrobial Activity of SLNs Antimicrobial Activity of SLNs on P. aeruginosa Strains

Four reference strains of P. aeruginosa were used to study the growth inhibitory effects of plain and QSI-loaded SLNs in the presence and absence of AL. These were wild-type PAO1, highly virulent strain PA14, nonmucinolytic strain RP37, and stable mucoid strain NH57388A. The organisms were grown either in Luria-Bertani (LB) medium or ASM to investigate the effects of nutrients and environmental conditions on microbial growth and biofilm formation using optical density (OD600) measurements. As detailed in the Supplementary Materials, 2-fold serial dilutions of the free or nanoencapsulated QSI (corresponding to QSI concentrations of 4.7–75 μM), in addition to equivalent dilutions of plain nanocarriers, were performed to study the inhibitory effect on planktonic P. aeruginosa, biofilm formation, and P. aeruginosa established biofilms.

Antimicrobial Activity of SLNs on P. aeruginosa Clinical Isolates

P. aeruginosa clinical isolates (17 isolates) were collected from patients presenting to Alexandria Main University Hospital (Alexandria, Egypt) under the discretion of the treating physicians as part of the patients routine medical care (No decisions affecting the patients' immediate care were made based on the current study). Ethical approval was obtained from the Alexandria University Research Ethics Review Committee, Alexandria, Egypt. The isolates were screened for their potential to form biofilm via crystal violet adherence assay described in the Supplementary Materials. Biofilm forming isolates were those having OD values higher than the cut-off OD (ODc) taken as three standard deviations above the OD of the negative control. Five isolates (P10, P11, P13, P20 and P21) proved potential to form biofilm and were selected for further investigations. SLN formulations were incubated with planktonic bacteria and the corresponding biofilms to determine the percent growth inhibition and antibiofilm activity, respectively.

Assessment of P. aeruginosa Viability via Live/Dead (L/D) Assay

Biofilm viability of P. aeruginosa reference strains after incubation with plain and QSI-SLNs (at concentration equivalent to 9.4 μM QSI) in LB medium was investigated by CLSM following Live/Dead staining using SYTO-9 and propidium iodide (PI) to stain live and dead bacteria, respectively (Thermo Fisher Scientific, Waltham, MA, USA), Supplementary Materials.

The Effect of SLNs on Biofilm Extracellular Matrix

The impact of SLNs on the biofilm matrix was verified using calcofluor-white dye, which stains the extracellular polysaccharide matrix in blue in combination with staining of living cells with SYTO-9, as described above, followed by CLSM 3D-time-lapse imaging.

Uptake of SLNs in P. aeruginosa Biofilm

Cou-SLNs were incubated with P. aeruginosa biofilms grown overnight in LB medium at 37 °C for 3 h. The PI-stained biofilms were fixed for CLSM examination.

Statistical Analysis

Data are expressed as means ± standard deviation (SD). Statistical analyses were performed by t-test and one-way analysis of variance (ANOVA) using the SPSS software. Statistical significance was set at p < 0.05.

Results and Discussion Physicochemical Characterization of SLNs Colloidal Properties

Table 1 depicts the impact of payload (QSI/coumarin), chitosan coating, and surface functionalization with alginate lyase on the colloidal properties of SLN formulations. The gelucire-based SLNs were approximately 143 nm in diameter, with a negative surface potential (–19.6 mV). Loading SLNs with either QSI (F2 SLNs) or coumarin-6 (Cou-F1 SLNs) and surface functionalization with AL (F1-AL SLNs) enlarged the particles by 50–60 nm with corresponding reduction in ζ-potential up to –9.5 mV (F2 SLNs), Table 1.

Coating particles with CS (F3 SLNs) reverted the ζ-potential to +11.7 mV but obviously did not impact much on the size with an average size of 200 nm, Table 1. Low-molecular-weight chitosan (LMW-CS) was chosen because of its convenient particle size, efficient surface coating, and better water solubility under neutral conditions than those of high-MW CS. It, further, offered high number of free amino groups on the surface to conjugate AL.9 The immobilization of AL is based on the electrostatic effect of overcompensating the surface charges and turning the ζ-potential again negative (–9.4 mV).

Labeled SLNs were comparatively larger (>205 nm), especially those bearing CS and AL; however, considering the size distribution, the effects were not pronounced (Table 1).

pH of SLN formulations ranged from 4.54 to 6.05 with a shift to lower pH in the CS-coated SLNs.

Insight into cystic fibrosis pathophysiology revealed that diminished functionality of the CF transmembrane conductance regulator (CFTR) reduces bicarbonate secretion into the airway lumen, resulting in a decreased pH of the airway surface liquid (ASL) to 6.81–6.89.39 Under these CF environmental conditions, alginate lyase is catalytically active. Meanwhile, the antibacterial and cellular binding activity of chitosan are related to its state of protonation (pKa = 6.3). However, such pH perturbations may be abolished by age,1 thus, reducing the efficiency of both AL and chitosan.

Particle Morphology

SLNs appeared spherical with a smooth surface and an average particle size distribution analogous to the size measurements by dynamic light scattering (Figure 1A). The TEM images showed no signs of particle aggregation. The QSI-loaded SLNs (F2 and F4) did not display any crystals, indicating drug encapsulation in the molecular state. It is believed that the organization of the lipid matrix, stabilizing agents, and active ingredients in SLNs, together with their relationship to one another, dictates their behavior as drug carriers.40

Figure 1 (A) TEM images F1 SLNs, F2 SLNs, F3 SLNs, and F4 SLNs; (B) In vitro release of QSI from F2 and F4 SLNs in SLF (*Statistically significant difference, two-way ANOVA with Tukey’s post-hoc test, p < 0.05); (C) Viability of A549 cells following 6 h incubation with QSI-loaded SLNs; (D) Viability of A549 cells following 24 h incubation with QSI-loaded SLNs; (E) Viability of Calu-3 cells following 6 h incubation with QSI-loaded SLNs; (F) Viability of Calu-3 cells following 24 h incubation with QSI-loaded SLNs. Data represent mean values (n = 3) ± SD.

Thermal Properties of SLNs

Thermograms for F1, F2, F3, and F4 SLNs (Figure S1) showed that all excipients used were completely in crystalline form, with the two sharp peaks around 40 ± 1 and 54.5 ± 1.5°C, corresponding to the lipid and emulsifier, respectively. For a detailed description of the DSC thermograms, refer to the Supplementary Materials. Table S1 shows the melting peak (Tm) and corresponding enthalpies. The crystalline behavior of lipids and possible polymorphic changes, among other factors, strongly influence particle size, stability, and drug entrapment.40 During heating or cooling, lipids tend to undergo structural changes according to their chemical structure. Depending on the composition and preparation process, lipids in the internal structure of the particles can have various conformations, such as liquid crystals, gels, or crystalline lamellar phases.40 Immobilization of AL with LMW-CS was shown to remarkably improve its thermostability at 37 °C and 45 °C for 1 h, while maintaining its enzymatic activity.9

Confirmation of AL Functionalization via FTIR Spectroscopy

A comparison between the FTIR spectra of pure CS, AL, and SLNs before and after immobilization with alginate lyase was carried out to verify surface immobilization, as described in the Supplementary Materials and Figure S2.

Figure 2 Confocal images illustrating the uptake of coumarin-labeled SLNs in: (A) A549 cells following 6 and 24 h incubation, and (B) Calu-3 cells after 24 h incubation (bars at the bottom right represent z-stacks).

The immobilization of alginate lyase on SLNs (F1-AL and F3-AL SLNs) resulted in strong absorption bands between 3000 and 3500 cm−1 indicating the presence of the enzyme, in addition to other important bands between 1541 and 1700 cm−1 corresponding to the N–H bending vibrations of amide II and the C–O stretching vibrations of amide I, respectively (Figure S2). FT-IR data are in line with previous reports.9,31,41 Wan et al31 referred to the characteristic band of amide I at 1670 cm−1 to confirm the presence of AL on the surface of silver nanocomposites. Covalent binding of alginate lyase on cross-linked chitosan nanoparticles was reported for AL-immobilized chitosan nanoparticles by Mohapatra.41

Entrapment and Loading Efficiency of QSI and in vitro Release

Data showed high EE in the range of 90–99% for both negatively and positively charged lipid formulations, with a corresponding loading of 12.48 ± 0.56 and 12.33 ± 0.23%, respectively. This reflects the suitability of the formulation technique, choice of lipid/emulsifier combination, and ratio that offers promising incorporation of payload. Noteworthy, the QSI concentration added is equivalent to 150 µM and represents the maximum loading capacity. The loaded concentration is 75-time the required effective concentration ensuring anti-virulence activity (2 µM) as per the pyocyanin assay data published earlier.22

A release study was carried out in SLF to mimic the microenvironment of the lungs. F2 and F4 SLNs provided a sustained release of the drug for >6 h (Figure 1B). We previously reported a more controlled release rate of QSI in SLF than in PBS; the excessive variety of components in the former might partially restrict particle swelling and limit drug dissolution and diffusion.26

An initial burst between 15 and 30% was recorded within the first hour, Figure 1B), which may be related to the crystallization rate of the lipid and obviously its melting point. Lipids tend to solidify faster during the cooling step, forming the inner core while expelling the drug towards the periphery, causing premature release. An effective concentration of QSI is required soon after administration to act on the bacterial biofilm. The release of QSI from F2 SLNs increased significantly with time to reach a maximum of 60% after 6 h (Figure 1B, two-way ANOVA with Tukey’s post hoc test, p < 0.05), which is a typical behavior of gelucire-based delivery systems (eg, minitablets, granules, and beads).42 QSI displayed comparable release rates from compritol-based SLNs, whereas tristearin-based SLNs allowed the release of no more than 10% QSI over 8 h.26

QSI release was significantly hampered by coating SLNs with chitosan (F4 SLNs), reaching only 38% after 6 h relative to F2 SLNs (Figure 1B, two-way ANOVA with Tukey’s post-hoc test, p < 0.05). As a pH-responsive polymer, chitosan forms a gel network on NP surface under basic conditions (pH 7.4), hindering drug liberation. At pH values below its pKa (6.3), protonated amino groups induce solubility of the surrounding chitosan layer; a faster diffusion can thus be predicted.43 Congruent with these findings; chitosan-coated nanoparticles elicited sustained release of numerous cargos (eg, dexamethasone, curcumin, ibuprofen) compared to naked counterparts.43–45

Effect of QSI-SLNs on Viability of Pulmonary Cell Lines

QSI-SLNs were designed to target bacteria within the biofilm. Thus, the detrimental effect on pulmonary cells is unintended, although it is worth investigating. The effect of drug loading, particle charge, concentration, immobilization with AL, as well as incubation time, were verified on both alveolar (A549) and bronchial (Calu-3) cell lines, Figure 1C–F. Data for unloaded formulations can be checked in the Supplementary Materials, Figure S3A–D.

Following 6 h incubation with loaded SLNs, A549 retained >80% viability in the concentration range (2.4–18.75 µM QSI) in most cases, apart from chitosan-coated SLNs (F4 SLNs) showing viability below 60%, Figure 1C). This reduced viability was not obvious for the corresponding plain particles (F3 SLNs), where above 60% viability was recorded for all SLN concentrations investigated, Figure S3A). Longer incubation period with SLNs (24 h) resulted in around 20% reduction in viability relative to 6 h incubation with either plain or loaded counterparts, Figures 1D and S3B (statistically significant difference, Two-way ANOVA with Tukey’s post hoc test, p < 0.05).

Surface immobilization of SLNs with AL improved A549 cell viability. Although insignificant differences were recorded for negatively charged F1 and F2 SLNs, significant impact was noted in case of CS-coated SLNs (two-way ANOVA with Tukey’s post hoc test, p < 0.05), whereby F4-AL SLNs showed >25% higher viability compared to F4 SLNs after 6 h, Figure 1C and D). Noteworthy, effective anti-virulence activity was confirmed at QSI concentration below 2 µM, where the maximum concentration investigated in the pyocyanin assay was 15 µM.22 This QSI concentration corresponds to SLN concentration in the range of 5.3–40 µg/mL). Reflecting this on the viability data (6 and 24 h), it can be concluded that within the effective concentration range, cell viability above 80 and 50%, respectively, can be ensured for all particle types as well as free QSI.

Determination of the cytotoxic concentration (CC50) after 6h-incubation was not possible because of the high viability of all samples at all concentrations, except for F4-SLNs with a CC50 of 30 µM (equivalent to 80 µg/mL SLNs). One of the main advantages of SLNs is that the lipids used are generally recognized as safe (GRAS) for which metabolic pathways exist. Gelucire is the family of vehicle derived from mixtures of mono, di and triglycerides with PEG esters of fatty acids. Gelucire 43/01 is a hard fat as per European pharmacopoeia (EP) and National formulary (NF). Owing to its extreme hydrophobicity, it was extensively formulated by melt granulation or melt extrusion techniques for capsule filling and tabletting, as an effective carrier in multi-unit floating drug delivery systems, micro- and nano-emulsions. Possible cytotoxic effect was reported for gelucires with higher HLB values and more pronounced surface-active characters. Special consideration might be given to the toxic aspects of surfactants. The concentration-dependent toxicity of poloxamer 188-stabilized SLNs was reported, whereas poloxamer 407 was found to be much safer.46 Poloxamer can interact with multidrug resistant cancer cells resulting in chemosensitisation of cancer cells. The pharmacological target(s) of this cytotoxicity remain(s) unclear. This anticancer action includes complex mechanisms involving fluidisation of the cellular membrane, ATP depletion, inhibition of drug efflux and reduction in GSH/GST detoxification activity.47

Table 2 depicts the CC50 values for the samples tested 24 h post incubation expressed as µg/mL SLNs and equivalent µM QSI. Plain AL-immobilized SLNs can be obviously tolerated at high concentration; CC50 value for F1-AL SLNs was above the tested range (>200 µg/mL), while that for the cationic counterpart F3-AL SLNs was 133.34 µg/mL (Table 2). Loading SLNs with QSI apparently reduced the viability of A549 cells 1–2 fold as revealed by the low CC50 values 17.34–100.8 µg/mL SLNs. It is still to be reminded that the recorded CC50 values for free and nanoencapsulated QSI are >6 µM, which is exceeding the effective concentration for anti-virulence activity.

Table 2 CC50 Values of Viability Testing on A549 Cells Following 24 h Incubation with SLNs

In contrast, incubation of Calu-3 cells with SLNs for 6 h showed equally acceptable viability (>80%) in most cases, even at the maximum particle concentration tested (200 µg/mL), Figure 1E). Unlike A549 cells, F4 SLNs demonstrated the highest viability, whereas AL-CS-coated SLNs showed significantly reduced cell viability when applied at high concentrations (two-way ANOVA with Tukey’s post hoc test, p < 0.05).

Following 24 h incubation, maximum tolerability was recorded for both plain and loaded anionic SLNs (F1 and F2 SLNs) as well as plain CS-coated SLNs (F3 SLNs) as demonstrated by their high CC50 values (>160 µg/mL) in Table 3. In contrast to the excellent viability of A549 cells with AL-modified SLNs, the viability of Calu-3 cells was dramatically affected after 24-exposure to either plain or QSI-loaded AL-SLNs (Figures 1F and S3D). The significant negative influence exerted by AL (two-way ANOVA with Tukey’s post hoc test, p < 0.05) is then cell and concentration dependent. Si et al48 recently combined mesoporous silica with alginate and melittin in presence of alginate lyase to control melittin release. The effect on ovarian cancer cells (SKOV3) was evaluated in terms of viability, invasion, migration assays, ROS levels, apoptosis-related proteins, and mitochondrial function tests. Cellular disruption and digestion were observed when 1 μM alginate lyase was employed. In addition, a dose-dependent decrease in the mitochondrial membrane potential was observed. It is to be noted that the concentration of AL used in the current study is far below that reported by Si et al48

Table 3 CC50 Values of Viability Testing on Calu-3 Cells Following 24 h Incubation with SLNs

Uptake of SLNs in Pulmonary Cell Lines

Although internalization of the formulated SLNs in the pulmonary epithelium is not the intention, studying their fate post-administration remains a concern in combination with toxicity evaluation. In parallel to the viability assays, both A549 and Calu-3 cells were designated as model alveolar and bronchial adenocarcinoma cell lines, respectively.

CLSM

As illustrated in Figure 2A, positively charged SLNs (Cou-F3 SLNs) were readily internalized in A549 cells compared to their negatively charged counterparts (Cou-F1 SLNs), which was more obvious in 24 h-treated cells. Surface modification with AL significantly improved the uptake of Cou-F1-AL SLNs compared to that of unmodified candidates (Cou-F1 SLNs), particularly after longer incubation periods. In contrast, extensive uptake of the CS-coated SLNs with and without AL (Cou-F3 SLNs and Cou-F3-AL SLNs) was observed at both time points (6 and 24 h). The potential of chitosan to empower the interactions with anionic cell membranes has been extensively reported.44,49 Analogous uptake behavior was previously reported by Taetz et al50 for chitosan-coated PLGA nanoparticles in A549 cells, confirming their direct dependence on the chitosan content and incubation periods. Labeled SLNs were observed as point-shaped green fluorescent areas, suggesting their entrapment in intracellular vesicles (Figure 2).

In the case of Calu-3 cells, confocal z-stacks (Figure 2B) confirmed the localization of green spots below the cell membrane in the cytoplasmic compartment, whereas no fluorescence was detected with the free dye. In comparison to A549 cells, the uptake of SLNs by Calu-3 cells was low. Their ability to build mucus at the air-liquid interface and firm tight junctions provides robustness and barriers toward particle internalization.

Despite their wide application in preclinical testing, both A549 and Calu-3 are human cancer cell lines that exhibit discrepancies in division rate and barrier properties relative to healthy primary pulmonary epithelium. Previous studies demonstrated extensive uptake of CS-coated PLGA nanoparticles in A549 cells, while they remained colocalized with the cell membrane of primary alveolar epithelial cells (hAEpC) for up to 48 h.50 Furthermore, one can argue that particle uptake predicted upon CF epithelium topped with dense mucus encompassing mucoid bacterial biofilms would be quite diverse.

Quantitative Estimation of Uptake Efficiency

Measurement of fluorescence intensity of lysed A549 cells was carried out to get quantitative estimation of uptake efficiency. In accordance with microscopic observations, overnight exposure offered 2–3 times higher uptake efficiency than 6 h. In addition, immobilization with alginate lyase improved uptake by ≥1.5 times, Figure 3A (statistically significant difference, paired t-test, p < 0.05). Despite the negative surface charge, the increased uptake of AL-immobilized particles suggests uptake mechanisms other than classical ionic interactions with the cell membrane. However, the molecular interactions between AL and human lung epithelium remain unknown. In line with this finding, it was reported that the addition of alginate lyase to melittin-loaded nanoparticles (without surface immobilization) showed concentration-dependent improvement in the cellular uptake studied on ovarian cancer cells.48

Figure 3 (A) Quantitative estimation of uptake efficiency of coumarin-labeled SLNs in A549 cells via Transwell technique following 6 and 24 h incubation (*Statistically significant difference, paired t-test, p <0.05); (B) Permeation efficiency of coumarin-labeled SLNs in artificial biofilm sputum medium; (C) 3D-time lapse confocal stacks showing diffusion of different coumarin-labeled SLNs in artificial biofilm sputum medium over 1–2 h.

Penetration of SLNs Across Artificial Biofilm Sputum Medium

To reach lung-residing bacteria and achieve effective eradication by therapeutic agents, biological barriers such as the extracellular sputum matrix and biofilm need to be overcome.3 The ability of formulated SLNs to penetrate CF sputum dictates their effectiveness against P. aeruginosa in CF lungs. The penetration of coumarin-labeled positively and negatively charged SLNs with and without surface immobilization with alginate lyase was verified on ASM supplemented with alginate as the main component of the extracellular matrix and HEC as a viscosity-imparting agent.51

Efforts to mimic the in vivo nutritional environment of lungs yielded a variety of in vitro culture compositions referred to as “artificial sputum media”. The ASM model has been recently used to investigate bacterial physiology, biofilm morphology, and antibiotic susceptibility in addition to studying the adaptation of pathogens to the CF lung and testing the effectiveness of antibacterial therapeutics.52 Although ASM can be considered more biorelevant than conventional mucin solutions, the absence of bacterial colonies and matrix components would shade probable imperative interactions of nanotherapeutic entities and biofilm. In a previous study, we appraised the application of ASM, real CF sputum and air-interface Calu-3 cells as a platform for penetration experiments.35 In the current study, investigating the effect of enzyme-linked SLNs urged the presence of the substrate (alginate) as supplement in ASM.

Transwell Permeation Study

SLN permeation through the artificial biofilm-sputum model was generally a slow process, remaining below 10% during the first 8 h. As illustrated in Figure 3B, labeled SLNs decorated with alginate lyase (Cou-F1-AL and Cou-F3-AL SLNs) showed faster migration than undecorated SLNs, whereby the fluorescence intensity almost doubled in the presence of AL at all time points (statistically significant, two-way ANOVA with Tukey’s post hoc test, p < 0.05).

Interestingly, significantly higher fluorescence intensity was observed for negatively charged SLNs (Cou-F1 and Cou-F1-AL SLNs) than for the CS-coated candidates (Cou-F3 and Cou-F3-AL SLNs), Figure 3B, (two-way ANOVA with Tukey’s post hoc test, p < 0.05). Meanwhile, free coumarin-6 displayed negligible fluorescence, presumably because of the quenched fluorescence of the hydrophobic probe in the aqueous milieu.53

Nanocarriers penetrate the biofilm matrix. The potential for biofilm diffusion is a function of their size, charge, and surface chemistry. Nevertheless, both positively and negatively charged NPs display reduced diffusion velocities and binding to bacterial biofilms via diverse mechanisms. While anionic polystyrene NPs can bind to the proximity of bacterial cells via hydrophobic interactions, their cationic counterparts might attach to wire-like components, presumably biofilm polymers, and/or negatively charged DNA.15

Relevant to this discussion, the emulsifier layer on nanoparticle surface plays a crucial role in biofilm penetration process. We previously demonstrated that SLNs stabilized with poloxamer 407 exerted higher diffusion in both ASM and CF sputum compared to Tween 80- and PVA-stabilized SLNs.35 In accordance with that, coating polystyrene NPs with PEG was described to exhibit a 90-fold higher diffusion rate in 10 µm-thick mucus versus noncoated NPs.54 Poloxamer 407 – the triblock copolymer (PEO-PPO-PEO) comprising polyethylene oxide (PEO) and polypropylene oxide (PPO) – covers F1-NP surface with a dense hydrophilic brush of PEO that incites excellent penetrability.35 Coating SLNs with the mucoadhesive polymer chitosan was certainly allied to a different performance. Owing to the advantages of mucoadhesion and prolonged residence time, weak penetrability remains a major concern.

3D Time-Lapse Confocal Imaging

3D time-lapse confocal imaging represents a promising approach to visually investigate nanoparticulate diffusion in biofilms, as reviewed by Schlafer and Meyer.55 Alexa Fluor-WGA is a well-reported fluorescently labeled lectin with selective-binding characteristics to the carbohydrate-containing extracellular matrix of P. aeruginosa biofilms.56

The z-stacks (Figure 3C) showed gradual time-dependent particle diffusion at varying rates and extent over 2 h. The maximum green fluorescence was observed for Cou-F1-AL, followed by Cou-F1 > Cou-F3-AL > Cou-F3 SLNs. Confocal stacks were in line with the Transwell diffusion experiment, confirming the potential of alginate lyase to break down the alginate mesh (the primary component in bacterial biofilms), resulting in reduced viscosity and weakened biofilms with faster and deeper particle penetration. According to Messiaen et al,57 the presence of DNase I also improved the diffusion of charged polymeric NPs 10 times in the biofilm.

In contrast, the chitosan-coated SLNs exhibited minimal penetrability. Indeed, there are multiple lines of evidence that support a predilection toward this behavior; electrostatic interactions between cationic chitosan and anionic alginate/DNA in biofilms and/or sialic acid residues in mucin would presumably capture the particles on the top and hinder their diffusion. Nonetheless, several studies have reported the potential of chitosan-coated PLGA NPs to penetrate artificial mucus.54,58 Penetration of labeled chitosan nanoparticles in Staphylococcus aureus and P. aeruginosa biofilms was improved when functionalized with DNase.59,60 Analogous behavior was observed with AL-functionalized SLNs. The reported approaches for AL-immobilized chitosan NPs lack data on sputum and/or biofilm penetration.8,9

Characterization of Respirable SLN-Embedded Microparticles (MPs)

To ensure sufficient deposition and avoid premature exhalation of the inhaled powder nanocarriers owing to their small size, SLNs were assembled into respirable microparticles via spray drying. Cou-F1 and Cou-F3 SLNs were sprayed with a carbohydrate carrier mixture (mannitol, leucine, and maltodextrin). The carrier ratio was previously optimized elsewhere.61 MPs were characterized in terms of yield, dye content, particle morphology, and aerodynamic behavior.

Cou-F1 and Cou-F3 MPs ensured a powder yield of 71.6 and 76.3%, respectively, with a percentage dye recovery of 98.14 and 97.84%, respectively.

SEM micrographs (Figure 4A) illustrate the morphology of the spray-dried microparticles as spherical corrugated microspheres with an average geometric diameter of 3 µm. Surface wrinkles are usually defined by collapse owing to the low glass transition temperature of mannitol counterbalanced by maltodextrin.61

Figure 4 (A) SEM micrographs of SLN-embedded microparticles (scale bar 2 µm); (B) Colloidal stability of SLNs in the spray drying solution and after spray drying; (C) Deposition pattern of cou-F1 MPs and cou-F3 MPs on stages of next generation impactor.

The stability of SLNs after spray drying was verified. As depicted from Figure 4B, SLNs mixed with the carrier solution were around 20 nm larger than the original SLNs, presumably attributed to leucine, mannitol and maltodextrin coating particle surface or dispersed in solution. Further slight increase in size and PdI was reported for SLNs recovered following reconstitution of microparticles. Of interest, this increase in size and PdI cannot reflect particle agglomeration and/or aggregation.

Investigating the in vitro deposition pattern of MPs using NGI revealed comparable behavior for both formulations (statistically insignificant difference, paired t-test, p < 0.05); maximum deposition (44–60%) in stages 3–5 (equivalent to 0.94–2.82 µm cutoff diameter), Figure 4C). Relevant aerodynamic diameters are summarized in Table 4, which depicts an MMAD of 3.54 and 2.48 µm for Cou-F1 and Cou-F3 MPs, respectively.

Table 4 Aerosolization Parameters of SLN-Embedded Microparticles as Determined by NGI

Both particles ensured a very promising emitted dose fraction above 92% (statistically insignificant difference, one-way ANOVA, p < 0.05). Cou-F3-SLNs-MPs showed higher FPF of 64.52 compared to Cou-F1-SLNs-MPs (47.88) indicating major deposition in the deep lung.

Anti-Pseudomonal Activity of SLNs

The impact of the formulated SLNs on the growth of mucoid and non-mucoid strains of P. aeruginosa (PA01, PA14, RP37, and NH57388A) and biofilm-forming clinical isolates was explored. Predetermined concentrations of QSI-loaded negatively charged particles (F2 SLNs) were compared to those of their positively charged counterparts (F4 SLNs). In addition, the influence of AL coating (F2-AL and F4-AL SLNs) was investigated. For comparison, free QSI and plain SLNs (F1, F1-AL, F3, and F3-AL SLNs) were studied at concentrations equivalent to those of the corresponding loaded formulations. In addition to the different formulations, the effects of the culture medium (LB and ASM) on P. aeruginosa strains in presence of the formulations were investigated. The studied endpoints included growth inhibition of planktonic bacteria, effects on biofilm formation, and eradication of established biofilms.

Antimicrobial Activity of SLNs on P. aeruginosa Reference Strains

The growth inhibition graphs (Figures 5 and S4) revealed strong dependence on the P. aeruginosa strain, planktonic versus biofilm cells, growth medium, type of SLNs, and AL coating, as detailed below. A concentration range of 4.7–75 µM free and nanoencapsulated QSI was tested. Figure 5 shows the antimicrobial activity of SLN formulations equivalent to 18.75 µM QSI. The rest of the data are shown in the Supplementary Materials (Figure S4).

Figure 5 Antimicrobial activity of SLNs (equivalent to 18.75 µM QSI) on P. aeruginosa reference strains (A) P. aeruginosa planktonic reference strains in LB medium; (B) P. aeruginosa planktonic reference strains in ASM; (C) Antibiofilm activity in LB medium; (D) Antibiofilm activity in ASM; (E) Increase in biofilm formation in LB medium; (F) Increase in biofilm formation in ASM.

Effect of SLNs on Growth of P. aeruginosa Planktonic Cells

In LB medium (Figure 5A), both PA14 and NH57388A were more sensitive than PAO1 to SLN treatment ensuring 70–95% growth inhibition when treated with AL-immobilized SLNs (F2-AL SLNs). In comparison, incubation of RP73 with F4-AL SLNs (equivalent to 18.75 µM QSI) repressed bacterial growth by 75%. Of interest, no growth inhibition of PA01 planktonic cells at QSI concentrations of up to 18.75 µM was recorded in LB medium regardless of the particle type (Figure 5A). These data are coherent with the goal of the study. Indeed, bacterial killing that ends up with robust antibiotic resistance is not the target. One advantage to quorum quenching therapeutic approach is that it inhibits QS without killing the bacteria, which does not pressure the bacteria to develop resistance against the inhibitors, thus permitting the host immune system to clear the infection.62

At the QSI concentration previously reported for promising anti-virulence potential (up to 15 µM), no growth inhibition was achieved. The dependence of anti-infective effect on bacterial strain, growth medium, and incubation period, among others, is extensively reported.62,63

PAO1, isolated from a wound >50 years ago, is a moderately virulent strain. The broad use of PAO1 created opportunities for mutational events to yield numerous PAO1 sub-lines with various phenotypic and genetic differences.64 In comparison, the clinical strain PA14 is highly virulent and represents the most common clonal group worldwide.65 Although PA14 is the more virulent strain, it displays a high level of genomic conservation with that of PAO1. On the genomic level, PA14 possesses two pathogenicity islands, which carry virulence-associated genes that are not present in the genome of PAO1.65

Another aspect is the key difference between PA14 and PAO1 quorum sensing (QS) mechanisms. The less virulent PAO1 strain produces high levels of QslA protein, a QS deactivator, which was minimally expressed in the hypervirulent PA14 strain.66 In addition, some strain genetic and phenotypic inconsistencies have been observed including MexT, a significant QS factor and fitness regulator, and LasR found in PAO1 strains from different laboratories. Their reported mutations affected QS signal, virulence factor, biofilm formation, and motility expression.67

Indeed, there is evidence that the growth behavior of P. aeruginosa is influenced by the growth medium. While the nutrient-rich, standard laboratory medium (LB medium) comprises tryptone, yeast extract, and sodium chloride in water,68 ASM includes different components that mimic the lung environment, namely mucin, DNA, amino acids, and lipid source.69 Growing P. aeruginosa strains in ASM instead of LB medium significantly increased their susceptibility to SLNs, where almost full growth inhibition was observed in all strains, especially NH57388A mucoid planktonic cells. Interestingly, 75–95% growth inhibition of PA14 was achieved regardless of the formulation type (Figure 5B).

It is to be carefully considered that numerous ASM recipes were reported in the literature with varying types, sources and concentrations of nutrients and were found to yield different phenotypes and chemotypes of P. aeruginosa, in other words, affecting microbial growth, metabolism and virulence.70 In this study, for instance, porcine gastric mucin, the main component of most ASM, represents an unexpected source of iron, which alters the production of specialized metabolites referred to as siderophores, mainly acting as metal scavenging virulence factors in P. aeruginosa.70 Siderophore production is primarily triggered by diminished levels of environmental iron. ASM with limited or no amino acid content restricts the ability of P. aeruginosa to produce the redox active virulence factors (phenazines).70 Conversely, ASM supplemented with N-acetyl glucosamine induces significant production of the virulence factor pyocyanin as N-acetyl glucosamine promotes a mechanism t