Evaluation of bacterial uptake, antibacterial efficacy against Escherichia coli, and cytotoxic effects of moxifloxacin-loaded solid lipid nanoparticles

SLNs and NLCs were prepared by ultrasonication (31) using components reported in Table 1. For SLNs we used only solid lipids in the lipid phase. Lipids were first melted and heated to 70 °C with Tween 80 or poloxamer 407 used as emulsifiers. Then we heated 10 mL of water containing 10 mg of MOX to 70 °C and added to the lipid phase. The mixture was sonicated with a probe sonicator (Sonopuls, Bandelin, Germany) at 50 % amplitude for 1 min and let to cool down to room temperature (25 °C) for 2 h to allow formation of particles. A similar method was used to obtain the NLCs, except that solid and liquid lipids were combined as the lipid phase (see Table 1 for details). Moreover, blank nanoparticles (without MOX) were prepared to compare the effects.

To label the nanoparticles, fluorescein (5 mg) was added instead of MOX at the lipid phase step and mixed with magnetic stirrer for 30 s. The water phase and the remaining steps were applied as described above.

Nanoparticle size and polydispersity index (PDI) were measured with photon correlation spectroscopy (aka dynamic light scattering) and zeta potential with laser Doppler velocimetry (both using the Nicomp Nano Z3000 system, PSS, Inc., New York, NY, USA. Each sample was measured in triplicate (32).

To calculate loaded MOX we used an indirect method by measuring MOX content in supernatants. Samples were centrifuged (IEC Centra MP4R, Rockville, USA) at 9000 g for 20 min and MOX detected with a high performance liquid chromatograph (HPLC, Agilent 1260 Infinity, Agilent Technologies Inc., Waldbronn, Germany) using a C18 column (150 mm × 4.6 mm i.d., 5 μm; ACE, Reading, UK) at the wavelength of 302 nm. Column temperature was 30 °C. The mobile phase was a mix of methanol-distilled water-acetonitrile (60:45:5) (v/v/v). It was adjusted to pH 2.7 with o-phosphoric acid. The flow rate was 1 mL/min (33).

MOX release from the nanoparticles was determined with the dialysis bag method (34). One millilitre (1 mL) of each nanoparticle suspension was placed in a dialysis bag (12–14 kDa, Spectrum Labs, USA) and the bag immersed in 50 mL of PBS (pH 7.4) and held in a shaker (Nuve, Istanbul, Turkey) operating at 50 rpm at 37 °C. MOX release was determined in 1 mL of PBS taken out from the pool at 15 and 30 min, and 1, 2, 4, 6, 8, 12, and 24 h. After each reading the PBS buffer was replaced with the same volume of a fresh one. The amount of released MOX was measured with the Agilent HPLC mentioned above.

To characterise the morphology of MOX-loaded SLNs and NLCs, the particles were observed under a FEI Tecnai G2 Spirit BioTwin (Thermo Fisher, Waltham, MA, USA) transmission electron microscope (TEM) operating at 40 kV and 80,000× magnification. Samples were mounted on copper grids with a mesh size of 200 (75 microns), stained with 2 % uranyl acetate for 2 min, and then the excess removed with filter paper. The remainder was dried in a Petri dish for 2 h before microscopy.

Vials with formulations were kept in a dark fridge at +4 °C for one month and their particle size, polydispersity index, and zeta potential rechecked as described above.

E. coli ATCC 25922 were cultivated in tryptic soy agar and tryptic soy broth at 35±1 °C, and the study was carried out with 24-hour fresh bacterial cultures.

To determine the bacterial uptake of SLNs and NLCs we used fluorescent microscopy and flow cytometry. Briefly, 0.1 mL of E. coli suspension at log phase was added to SLNs and NLCs and incubated at 37 ^°C for 2 h. After incubation, the bacterial pellet and culture supernatant were separated by centrifugation at 4137 g. The pellet was washed with sterile distilled water three times to discard free NLCs or SLNs and then resuspended in 1 mL of sterile distilled water. The fluorescent intensity of E. coli cells was observed with a CYTOFlex cytometer (Beckman Coulter, Suzhou, China). Blue fluorescence was collected through a 525/40 BP fluorescent channel with a 488 nm blocking filter. For each sample around ten thousand cells were analysed. Data were measured and histograms created and interpreted using the CytExpert 2.4 Software (Beckman Coulter, Indianapolis, IN, USA). In addition, to detect the position of MOX-loaded NLCs and SLNs in bacteria, the cells were observed under a DM IL inverted fluorescent microscope at 40× magnification and images taken (Leica, Munich, Germany).

Antibacterial activity against E. coli was tested with the broth microdilution method for minimal inhibition concentration (MIC) and with the disc diffusion test for the inhibition zone according to the European Committee on Antimicrobial Susceptibility Testing standards (35). For the disc diffusion test we used 4.0±0.5 mm deep Mueller-Hinton agar plates (25 mL in a 90 mm circular Petri dish). Bacterial suspensions were prepared from fresh E. coli cultures in sterile saline (0.9 % NaCl) to the density of a 0.5 McFarland measured with a DEN-1B densitometer (Biosan SIA, Riga, Latvia). A sterile cotton swab was dipped into bacterial suspensions and spread on the agar surface in three directions. Discs with MOX-loaded SLNs and NLCs were then placed on the surface of inoculated agar plates and incubated at 35±1 °C for 18±2 h. After incubation, inhibition diameters around the discs were measured and compared to standard MOX as control.

For the broth microdilution (MIC) test 100 μL of cation-adjusted Mueller-Hinton broth added to all U-bottom microplate wells. We added 100 μL of MOX-loaded nanoparticle solution to the first well and diluted it further across 12 wells. Then the bacterial suspension (in saline, 0.9 % NaCl) with the density of the 0.5 McFarland (1x10⁸ CFU/mL) was 1:100 diluted and added to all wells to obtain the final bacterial concentration of 5x10⁵ CFU/mL. The microplates were then incubated at 35±1 °C for 18±2 h, and MIC determined as the lowest concentration of MOX-loaded SLN and NLC solutions that visibly inhibited the growth of E. coli.

RAW 264.7 cells were seeded in 25 cm² flasks filled with 7 mL of DMEM supplemented with 10 % FBS, and 1 % penicillin-streptomycin and then grown in an incubator (Sanyo, Osaka, Japan) at 37 °C in an atmosphere supplemented with 5 % CO₂ for 24 h.

Cytotoxicity was determined with the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay as described earlier by Bacanli et al. (36). Cells were detached with a cell scraper and a total of 10⁵ cells/well seeded in 96-well tissue-culture plates. The cells were incubated in full medium with various concentrations (0.010, 0.025, 0.05, and 0.1 μg/mL) of MOX alone, MOX-free SLNs, MOX-loaded SLNs, MOX-free NLCs, and MOX-loaded NLCs at 37 °C in an atmosphere supplemented with 5 % CO₂ for 24, 48, and 72 h. For negative control we used a medium containing 10 % FBS and 1 % penicillin-streptomycin. After exposure, the medium was aspirated, cells washed with PBS, 10 μL of MTT (5 mg/ mL of stock solution with PBS) added to 100 μL of cell suspension per well, and cells incubated for another 3 h. The MTT dye was then carefully removed and 100 μL of DMSO added to each well. The absorbance of each well was measured with a microplate reader (Epoch, BioTek Instruments, Winooski, VT, USA) at 570 nm.

For statistical analysis we used the SPSS for Windows v. 20.0 software (IBM, New York, NY USA). All data are expressed as means ± standard deviations of measurements in three biological replicates. As the distribution was normal, differences between the groups were compared using Student’s t test. Statistical significance was set to p<0.05.

The composition of lipid nanoparticles is given in Table 1. Visual inspection did not reveal any separation between the oil and water phase in the MOX-loaded SLNs and NLCs, that is, emulsification was successful.

We aimed for small particles for better uptake by bacteria. SLN1, which contains Tween 80 as emulsifier, had significantly lower particle size and PDI compared to SLN2, which contains poloxamer 407 (p<0.05, Table 2). These differences in particle size and PDI are likely owed to the lower molecular weight of Tween 80 than that of poloxamer 407 (37), as suggested by similar studies (38, 39). PDI values of SLNs were below 0.3 characteristic of a monodisperse system (40). SLN1 had lower encapsulation efficiency than SLN2, which may be related to its lower particle size. Zeta potentials were significantly different (p<0.05), possibly due to different emulsifiers used in the SLNs.

NLC1 had bigger particles and distribution than NLC2. As known, particle size and distribution can be modified with sonication, which breaks coarse emulsion droplets to form a nano-emulsion, so it is one of important parameters to bear in mind while preparing lipid nanoparticles (41). If sonication is less effective, nanoparticles containing more liquid lipids, like NLC1, tend to form bigger particles, as liquid lipids increase the viscosity of coarse formulations.

Zeta potential of formulations may differ as particles contain different liquid phases. For instance, oleic acid, which was used in NLC2 as liquid phase, could shift the zeta potential from -6.3±0.38 to -11.6±0.61. However, there was no significant difference in encapsulation efficiency between NLCs (Table 2).

Table 3 shows that there were no significant changes in particle size, polydispersity index, and zeta potential measured in the formulations after one month, save for the drop in zeta potential in the SLN1 formulation. However, it is clear that particles tend to aggregate during storage, which leads to a gain in size and polydispersity index and loss in zeta potential.

TEM images and photon correlation spectroscopy of LNPs show spherical particles of similar size (Figure 1).

Figure 1

Figure 2 shows MOX release from nanoparticles. The release of standard (not nanoparticle-loaded) MOX was high in the first two hours, after which it kept dropping by the end of hour 4, at which point 95 % of MOX was released, and no more drug was until hour 24. LNPs slowed down its release thanks to its even dispersion in the lipid matrix and diffusion from it, as previously proposed by other authors (42, 43).

Figure 2

Figure 3 shows flow cytometry findings of E. coli cells. There is no fluorescence signal in control cells (Figure 3a), which were incubated with unlabelled nanoparticles, whereas the rest shows a strong signal (Figure 3b-f). Bacterial uptake of SLN1 and SLN2 (57.29 % and 50.74 %, respectively) was significantly more efficient than that of fluorescein solution alone (21.47 %) or either NLC (39.26 % for NLC1 and 32.79 % for NLC2) (p<0.05).

Figure 3

Judging by higher uptake of lipid nanoparticles than the fluorescein solution alone, drug encapsulation in nanoparticles should enhances drug absorption owing to the small size, prolonged release, and hydrophobic nature of lipid nanoparticles, which is similar to the Gram-negative bacterial cell wall (44). Apparently, smaller size of SLN1 is also the reason for higher bacterial uptake than that of SLN2, which is in line with previous reports (45).

These findings are confirmed by fluorescence microscopy (Figure 4), as only a small number of bacterial cells absorbed fluorescein solution alone (Figure 4a), whereas lipid nanoparticles show better absorption (Figure 4b-e) and therefore confirm that they can better enter bacteria than the free drug, as suggested in a related study (14).

Figure 4

Table 4 shows the antibacterial activity of nanoparticles in terms of inhibition zone (diameter) and MIC. Only the SLN1 formulation was significantly more effective against E. coli (MIC 0.020 μg/mL, 34 mm) than MOX delivered as a standard solution (p<0.05). The NLC1 formulation showed better inhibition diameter (36 mm) than standard MOX but not MIC. In fact, SLN2 and both NLC formulations showed the same MIC as the corresponding standard MOX.

We know that E. coli with its thinner peptidoglycan layer and an outer lipopolysaccharide membrane has a limited permeability to drugs (46). Nanostructure with controlled drug release weakens its membrane resistance and increases drug uptake over time. In this sense, our findings are consistent with the ceftriaxone study by Kumar et al. (47).

We believe that the higher antibacterial activity of MOX loaded into SLN1 is related to the smaller size of SLN1 and distinct lipid and surfactant features, as earlier studies suggest that lipid nanoparticles with Tween 80 can generate higher antibacterial activity against E. coli (42). Furthermore, SLN1 can carry MOX directly to the target within the bacterium and also act as efflux pump inhibitor, that is, inhibit drug clearance from the cell (14, 48).

None of the nanoparticles, whether loaded with MOX or not, lowered RAW 264.7 cell viability below 50 % after 24, 48, and 72 h of exposure to all studied concentrations. The highest concentration of standard MOX (not loaded into nanoparticles), however did lower cell viability below 50 %. Our findings therefore suggest that lipid nanoparticles in addition to having low cytotoxicity improve lower the cytotoxicity of MOX and lower its biocompatibility.

Similar observations of nanocarriers lowering MOX cytotoxicity were reported by several studies (49–51).