*2.4. Minimum E*ff*ective Concentrations (MEC10) for Tested Bacteria and Fungi*

The minimum effective concentration (MEC10) of CPe, CT80, and CEt on foodborne Gram-positive and Gram-negative bacteria as well as fungi have been determined. The dose–response curve shows a slow killing effect (≤10% of the population) of CPe after 1 h of treatment at a two-fold higher concentration compared with MIC90 data. The MEC10 also highlights the effective killing effect of CPe when compared to CT80 and CEt (Figures 4 and 5) (*P* < 0.01).

**Figure 4.** Minimum effective concentration (MEC10) of CPe, CT80, CEt, and Van (μg/mL) on *E. coli* (**A**), *S. aureus* (**B**), *B. subtilis* (**C**), *P. aeruginosa* (**D**), and *S. pyogenes* (**E**).

#### *2.5. E*ff*ect on Microbial Oxidative Balance*

Reactive oxygen species (ROS) production and accumulation in the cells initiates oxidative stress, leading to cellular structural damage followed by induced apoptosis [6]. We have investigated the relationship between oxidative stress generation after 1 h of treatment and microbial killing activity. The results are demonstrated in Figures 6 and 7 for bacteria and fungi, respectively. Data expressed as % of the control are as follows: the ROS (1085.86 ± 126.36), peroxide (1229.86 ± 164.52) and superoxide (1276.86 ± 165.42) generation were the highest in case of *S. aureus*. The CPe showed an effective increment of ROS, peroxide, and superoxide generation in both Gram-positive and -negative bacteria when compared to CT80 and CEt (*P* < 0.01). CPe showed increased oxidative stress in both bacteria and fungi at least seven-fold higher than the negative control whereas the positive control (menadione) produced an eight- to nine-fold increase in 1 h. The CEt has generated a two to four-fold increment in oxidative stress which is the lowest among all tested compounds.

#### *2.6. Time–Kill Kinetics Study*

The time–kill kinetics curve was performed to quantify living populations after a definite time interval under different sample MEC10 concentrations. A significant reduction (four log-fold) in the cell survivability has been observed in case of CPe when compared to Gc (*P* < 0.01) (Figures 8 and 9). Fifty percent of cell death occurred by CPe at 16 and 36 h in the case of bacteria and fungi, and was most

effective in reducing living colonies in case of *C. albicans* (1.73 ± 0.15 CFU/mL) after 48 h of treatment. At an average of a two-fold higher concentration, CEt was able to show a killing effect compared to CPe (*P* < 0.01).

**Figure 5.** Minimum effective concentration (MEC10) of CPe, CT80, CEt, and Cas (μg/mL) on *S. pombe* (**A**), *C. albicans* (**B**), and *C. tropicalis* (**C**).

**Figure 6.** Percentage oxidative stress generation by CPe, CT80, CEt, and Van on *E. coli* (**A**), *S. aureus* (**B**), *B. subtilis* (**C**), *P. aeruginosa* (**D**), and *S. pyogenes* (**E**). Six independent experiments, each with 3 replicates, compared with menadione (Me) and growth control (Gc) as controls after 1 h of treatment (\*\**P* < 0.01).

**Figure 7.** Percentage oxidative stress generation by CPe, CT80, CEt, and Cas on *S. pombe* (**A**), *C. albicans* (**B**), and *C. tropicalis* (**C**). Six independent experiments, each with 3 replicates, compared with Me and Gc as positive and growth controls after 1 h of treatment (\*\**P* < 0.01).

**Figure 8.** Colony-forming unit (CFU/mL) of CPe, CT80, and CEt on *E. coli* (**A**), *S. aureus* (**B**), *B. subtilis* (**C**), *P. aeruginosa* (**D**), and *S. pyogenes* (**E**). Six independent experiments, each with 3 replicates, compared with Van and Gc as positive and growth controls.

**Figure 9.** Colony-forming unit (CFU/mL) of CPe, CT80, and CEt on *S. pombe* (**A**), *C. albicans* (**B**), and *C. tropicalis* (**C**). Six independent experiments, each with 3 replicates, compared with Cas and Gc as positive and growth controls.

#### *2.7. Live*/*dead Cell Viability Discrimination*

The effect of CPe, CT80, and CEt on the viability of selected bacteria and fungi were tested (Figures 10 and 11). CPe decreases the viability of the tested bacteria and fungi with an average viability reduction to 42.36% ± 3.74% and 49.62% ± 5.25% of mean percentage viability compared to Gc after 16 and 36 h of treatments in bacteria and fungi respectively (*P* < 0.01), whereas CT80 and CEt were less effective than CPe with mean percentage viabilities of ≥60% and 70%, respectively.

### *2.8. Interaction Study between Cell Model and Di*ff*erent Formulations of Chamomile EO*

The unilamellar liposomes (ULs), consisting of a single phospholipid, can be used as artificial cells or biological membrane model for studying the interactions between cells or cell membranes and drugs or biologically active components [24]. This study was conducted to determine the intracellular delivery ability of active components from chamomile EO for different formulations.

We have studied the interaction of ULs and different forms of chamomile EO for 24 h at 35 ◦C. After 1 h of interaction, 27.2% of EO have penetrated the liposomes from Pickering nanoemulsion, while conventional emulsion and the ethanolic solution did not provide a measurable amount. The next sampling was after 2 h, where Pickering emulsion has delivered 48.3% of EO, conventional emulsion 0.5%, while the amount of EO delivered by the ethanolic solution was not measurable. Final sampling was after 24 h, with 82.2% of chamomile EO found in the ULs when it was introduced in Pickering nanoemulsion form, and this value was 66.8% for conventional emulsion and 32.5% for the ethanolic solution (see Table 2).

**Figure 10.** Mean percentage viability of CPe, CT80, and CEt on *E. coli* (**A**), *S. aureus* (**B**), *B. subtilis* (**C**), *P. aeruginosa* (**D**), and *S. pyogenes* (**E**). Six independent experiments, each with 3 replicates, compared with Van and Gc as positive and growth controls.

**Figure 11.** Mean percentage viability of CPe, CT80, and CEt on *S. pombe* (**A**), *C. albicans* (**B**) and *C. tropicalis* (**C**). Six independent experiments, each with 3 replicates, compared with Cas and Gc as positive and growth controls.



#### **3. Discussion**

The effect of three different formulations on antimicrobial activity of chamomile essential oil has been examined. We have successfully prepared stable Pickering nanoemulsion using silica nanoparticles with appropriate lipophilicity. The emulsion was stable for three months. The effectiveness of Pickering emulsion was compared with conventional emulsion and ethanolic solution.

Based on our antimicrobial activity analyses, CPe shows higher growth inhibitory action and consequently lower MICs compared to CT80 and CEt. Many researchers have studied the antimicrobial activity of chamomile oil [3,7,13,23,25], however, the mechanism of action at subinhibitory concentrations has not previously been studied. Our data suggest an effective killing activity of CPe on selected bacteria and fungi. It is believed that EOs act against cell cytoplasmic membrane and induce stress in microorganisms [26–29]. To visualize the effects of CPe, CT80, and CEt, we introduced different staining methods to understand their mechanism. CPe was able to generate higher oxidative stress compared to the conventional emulsion and ethanolic solution followed by metabolic interference and cell wall disruption and finally caused cell death at subinhibitory concentration [21,30–33].

The results obtained in the model experiment show that CPe is the most effective form for the intracellular delivery of chamomile EO. Based on these results it can be established that the different antibacterial and antifungal effects may be caused by the difference of adsorption properties of EO forms to the microbial cells. The mechanism of delivery has not been revealed in this study, but evidence for the adsorption of Pickering emulsion droplets on the cell membrane has been previously reported [18]. Assuming the adsorption of CPe droplets on the cell membrane of investigated microbes, intracellular delivery of active components from EO is feasible in two ways. Passive diffusion is caused by the higher local concentration gradient of EO on the cell membrane, or fusion of CPe droplets with microbial cells. Overall, our observations demonstrate that CPe facilitates chamomile oil to permeate cells, inducing oxidative stress and disrupting the membrane integrity because of higher adsorption efficacy of chamomile EO. SNP acts as a stabilizer, inhibiting the easy escape of EOs from the emulsion system compared to the conventional emulsion and free oils.

#### **4. Materials and Methods**

#### *4.1. Synthesis, Surface Modification, and Characterization of Stöber Silica Nanoparticles*

We have performed the synthesis of 20 nm hydrophilic silica nanoparticles with the previously reported modified Stöber method [9]. Briefly, a solution of tetraethoxysilane (TEOS) and ultrapure water in ethanol was prepared by using tetraethoxysilane, (Thermo Fisher GmbH, Kandel, Germany, pur. 98%); absolute ethanol AnalaR Normapur ≥99.8% purity (VWR Chemicals, Debrecen, Hungary) and water (membraPure Astacus Analytical with UV, VWR Chemicals, Debrecen, Hungary). The solution was stirred for 20 min and sonicated for another 20 min (Bandelin Sonorex RK 52H, BANDELIN electronic GmbH & Co. KG, Berlin, Germany). An appropriate amount of NH3 solution (28% (*w*/*w*) ammonium solution, VWR Chemicals, Debrecen, Hungary) was added to the reaction mixture and was stirred at 1000 rpm for 24 h at room temperature (25 ◦C). The molar ratio of components was water/ethanol/TEOS/NH3 = 100:300:5.2:1. The surface of hydrophilic silica nanoparticles was modified with propyltriethoxysilane (PTES Alfa Aesar, Haverhill, MA, USA, pur. 99%) in a post-synthesis modification reaction [32]. The ethanolic solution of the modifying agent was added to the freshly prepared hydrophilic silica nanoparticle suspensions; the mixtures were stirred for 6 h with 1000 rpm at room temperature. Before further use of the SNPs, the ammonium hydroxide and ethanol were always removed from the reaction mixture by distillation (Heidolph Laborota 4000, Heidolph Instruments GmbH & CO. KG, Germany). The water content was supplemented three times. The concentration of silica nanoparticle water-based suspension was finally adjusted to 1 mg/cm3.

The size distribution and zeta potential of silica nanoparticles were determined by dynamic light scattering (DLS) using Malvern Zetasizer NanoS and NanoZ instruments (Malvern Instruments-Malvern Panalytical, Worcester, UK). The morphology and size distribution were also examined with transmission electron microscopy (TEM), (JEOL JEM-1200 EX II and JEM-1400, JEOL Ltd., Tokyo, Japan). The samples were dropped onto 200 mesh copper grids coated with carbon film (EMR Carbon support grids, Micro to Nano Ltd, Haarlem, The Netherlands) from diluted suspensions.

#### *4.2. Preparation and Characterization of Pickering Nanoemulsion*

As stabilizing agents, surface-modified silica nanoparticles or Tween 80 surfactant (Polysorbate80, Acros Organics, New Jersey, NJ, USA) were used. The concentration of stabilizing agents and chamomile essential oil (bluish *Matricaria chamomilla* oil, Aromax Ltd., Budapest, Hungary) was kept constant for all experiments, the values were 1 mg/mL and 100 μg/mL, respectively. The first step of the emulsification process was sonication for 2 minutes (Bandelin Sonorex RK 52H, BANDELINelectronic GmbH & Co. KG, Germany), then emulsification using UltraTurrax (IKA Werke T-25 basic, IKA®-Werke GmbH & Co. KG, Germany) for 5 min at 21,000 rpm. To compare the different formulations, an ethanolic solution was also prepared; chamomile essential oil was added to absolute ethanol at 100 μg/mL concentration, and the solution was sonicated for 5 min.

The stability of Pickering emulsion was studied from periodical droplet size determination using DLS measurements (Malvern Zetasizer Nano S, Malvern Panalytical Ltd, Worcester, UK).

#### *4.3. Materials for Biological Experiments*

In these experiments, the sterile 96-well microtiter plates were from Greiner Bio-One (Kremsmunster, Austria), potassium phosphate monobasic, glucose, adenine, 96% ethanol (Et), peptone, yeast extract, agar-agar, and Mueller Hinton agar were from Reanal Labor (Budapest, Hungary), modified RPMI 1640 (contains 3.4% MOPS, 1.8% glucose, and 0.002% adenine), SYBR green I 10,000×, propidium iodide, dihydrorhodamine 123 (DHR 123), 2 ,7 -dichlorofluorescin diacetate (DCFDA), dihydroethidine (DHE) and menadione (Me) were from Sigma-Aldrich Chemie GmbH (Steinheim, Germany), disodium phosphate and dimethyl sulfoxide (DMSO) were from Chemolab Ltd. (Budapest, Hungary), sodium chloride from VWR Chemicals (Debrecen, Hungary), potassium chloride was from Scharlau Chemie S.A (Barcelona, Spain), 3-(*N*-morpholino) propanesulfonic acid (MOPS) was from Serva Electrophoresis GmbH (Heidelberg, Germany), caspofungin (Cas) from Merck Sharp & Dohme Ltd (Hertfordshire, UK), vancomycin (Van) from Fresenius Kabi Ltd. (Budapest, Hungary), 0.22 μm vacuum filters from Millipore (Molsheim, France) and the cell spreader was from Sarstedt AG & Co. KG (Numbrecht, Germany). All other chemicals used in the study were of analytical or spectroscopic grade. For fungi, we used an in-house nutrient agar medium [34] while phosphate-buffered saline (PBS, pH 7.4) was from Life Technologies Ltd. (Budapest, Hungary). Highly purified water (<1.0 μS) was applied throughout the studies.

#### *4.4. Determination of Minimum Inhibitory Concentration (MIC90)*

#### 4.4.1. Microorganisms

*Escherichia coli* (*E. coli*) PMC 201, *Pseudomonas aeruginosa* (*P. aeruginosa*) PMC 103, *Bacillus subtilis* (*B. subtilis*) SZMC 0209, *Staphylococcus aureus* (*S. aureus*) ATCC 29213, *Streptococcus pyogenes* (*S. pyogenes*) SZMC 0119, *Schizosaccharomyces pombe* (*S. pombe*) ATCC 38366, *Candida albicans* (*C. albicans*) ATCC 1001, and *Candida tropicalis* (*C. tropicalis*) SZMC 1368 were obtained from Szeged Microbial Collection, Department of Microbiology, University of Szeged, Hungary (SZMC) and Department of General and Environmental Microbiology, Institute of Biology, University of Pecs, Hungary (PMC).

#### 4.4.2. Antimicrobial Activity Tests

The antibacterial activity of the tested drugs was separately evaluated on *E. coli, P. aeruginosa*, *B. subtilis*, *S. aureus*, and *S. pyogenes* according to our previously published protocol [35]. In brief, bacterial populations of ~10<sup>5</sup> CFU/mL were inoculated into RPMI media and incubated for 16 h at 35 <sup>±</sup> 2 ◦C with test compounds (CPe, CT80, CEt, and Van) over a wide concentration range (0.3–0.01 μg/mL). The absorbance was measured by a Thermo Scientific Multiskan EX 355 plate reader (InterLabsystems, Budapest, Hungary) at 600 nm.

The antifungal activity against *S. pombe*, *C. albicans*, and *C. tropicalis* species were also carried out according to our previously published method [35]. Briefly, ~10<sup>3</sup> cells/mL were incubated for 48 h at 30 ± 2 ◦C with test compounds (CPe, CT80, CEt and Cas) at wide concentration range (20–0.01 μg/mL) in modified RPMI media. The absorbance was measured by a Thermo Scientific Multiskan EX 355 plate reader (InterLabsystems, Budapest, Hungary) at 595 nm. Absorbance values were converted to percentages compared to growth control (~100%) and data were fitted by nonlinear dose–response curve method to calculate the dose producing ≥90% growth inhibition (MIC90). All the measurements were performed by applying three technical replicates in six independent experiments. Van and Cas were used as a standard antibacterial and antifungal drug, respectively, throughout the experiments.
