Alpha-amylase, CaCl2 ∙ 2H2O, 4′,6′-diamidino-2-phenylindole (DAPI), 2′,7′-dichlorofluorescin-diacetate (DCFH-DA), MgCl2 ∙ 6H2O, mucin, ox bile, pancreatin, paraformaldehyde, pepsin, trypsin and ZnO nanopowder (#677450 and #544906) were purchased from Sigma-Aldrich Chemie GmbH, Taufkirchen, Germany. Dulbecco’s Modified Eagle Medium (DMEM), foetal bovine serum (FBS), non-essential amino acids, penicillin/streptomycin, and trypsin/EDTA were obtained from PAN-Biotech GmbH, Aidenbach, Germany. 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT), dimethyl sulfoxide, fluorescein isothiocyanate (FITC)-dextran, hydrogen peroxide (30%), nitric acid (Suprapur), and triton X-100 were acquired from Merck KGaA, Darmstadt, Germany. 2-((3-Chlorophenyl) hydrazinylidene) propanedinitrile (CCCP) and ZnCl2 were procured from Thermo Fisher Scientific Inc., Waltham, MA, USA. Carbamide, ethylene glycol tetraacetic acid (EGTA), KCl, KH2PO4, NaCl, NaHCO3, and Na2HPO4 ∙ 2H2O were purchased from Carl Roth GmbH & Co. KG, Karlsruhe, Germany. Cacodylic acid and glutaraldehyde were bought from Serva Electrophoresis GmbH, Heidelberg, Germany. 5,5′,6,6′-Tetrachloro-1,1′,3,3′-tetraethylbenzimidazolylcarbocyanine iodide (JC-1) was obtained from Enzo Life Sciences GmbH, Lörrach, Germany. Phalloidin-iFlour 488 reagent was acquired from Abcam plc., Cambridge, UK.
Preparation of ZnO NP dispersions and in vitro digestion procedure
ZnO nanopowder was dispersed in 8.5 g/l NaCl (ZnO target concentration 3.34 mg/ml). The dispersions were filled in a glass rosette cell (RZ2; Bandelin electronic GmbH & Co. KG, Berlin, Germany), which was submerged in an ice bath to avoid heat production. An ultrasonic homogenizer (Bandelin Sonopuls HD 2070) was plunged 1 cm deep in the dispersions. The NP were ultrasonicated with a critical sonication energy of 720 J/ml for 1 h to break NP agglomerates and aggregates. The in vitro digestion was conducted after DIN 19738  and Stein et al. . The sonicated dispersions were further diluted with synthetic saliva (8.5 g/l NaCl, 41.55 U α-amylase) to obtain 15 ml of three different ZnO NP concentrations (3.34, 1.67, and 0.67 mg/ml). Skimmed milk powder was additionally added and served as the food matrix. To mimic the mouth phase of digestion, all samples were placed in a shaking water bath (GFL Gesellschaft für Labortechnik GmbH, Burgwedel, Germany) for 5 min at 37 °C and 120 rpm. Afterwards, the gastric phase was simulated by adding 35 ml of gastric juice (0.7 g/l KCl, 0.27 g/l KH2PO4, 2.9 g/l NaCl, 3 g/l mucin, 1 g/l pepsin, pH 2). The samples were placed again in a shaking water bath for 2 h at 37 °C and 120 rpm. Subsequently, the intestinal phase was imitated by the addition of 50 ml intestinal juice (0.5 g/l CaCl2 ∙ 2H2O, 0.3 g/l KCl, 0.2 g/l MgCl2 ∙ 6H2O, 1 g/l NaHCO3, 9 g/l ox bile, 0.3 g/l carbamide, 9 g/l pancreatin, 0.3 g/l trypsin, pH 7.5). All samples were positioned in the shaking water bath for 3 h at 37 °C and 120 rpm. For the cell culture experiments, the digested dispersions were finally mixed 1:10 with the cell culture medium to obtain concentrations of 10, 25, and 50 µg/ml (equivalent to 123, 307, and 614 µM).
Characterization of digested ZnO NP dispersions
Inductively coupled plasma optical emission spectrometry (ICP-OES) was used to determine the portion of free zinc ions and bound zinc in the intestinal juice. For this purpose, the digested dispersions were centrifuged for 60 min at 5250×g (Allegra X-15R; Beckman Coulter GmbH, Krefeld, Germany). Supernatants were transferred into new containers. Precipitates were dried for 6 h at 95 °C until mass constancy. Supernatants were further diluted and acidified at a ratio of 1:10 with extra pure nitric acid. Precipitates were suspended in 1 ml 65% extra pure nitric acid and 0.5 ml hydrogen peroxide and placed in a sonication bath (Transsonic 460/H; Elma Schmidbauer GmbH, Singen, Germany) for 30 min at 50 °C. The precipitates were filled up with ultrapure water to 50 ml. A dilute solution for yttrium was added to all samples as an internal standard element. The total zinc content in the supernatants and precipitates was measured by ICP-OES (iCAP™ 7000 equipped with an CETAC ASX-520 autosampler, both from Thermo Fisher Scientific Inc., Waltham, MA, USA) using two different wavelengths, 206.2 and 334.502 nm. Two quality-check samples (QC; certified reference material) within the concentration range of the analytical results were chosen and measured regularly after about 20 samples. The minimum number as well as the acceptance criteria for evaluating the QC are in accordance with the Food and Drug Administration guidelines . The limit of quantification was calculated in compliance with DIN 32645  as 0.3 ± 0.1 mg/l. The measurement of uncertainty was determined with two certified reference materials according to DIN ISO 11352:2013-03  to be 2.3%. Data processing and quality management were conducted with an in-house developed program (from Chemical Analytics, German Environment Agency, Bad Elster, Germany) and Qtegra™ (Thermo Fisher Scientific Inc., Waltham, MA, USA).
Furthermore, the fate of ZnO NP during in vitro digestion was investigated using transmission electron microscopy (TEM). After each digestion step, samples were taken and 4 µl of each sample was placed on a Formvar-carbon filmed 400 mesh grid (Quantifoil Micro Tools GmbH, Großlöbichau, Germany). The samples were dried and analysed using a Zeiss CEM 902 A electron microscope (Carl Zeiss AG, Oberkochen, Germany). Images were acquired with a wide-angle dual-speed 2 K-CCD-camera (TRS, Moorenweis, Germany).
Cell culture and ZnO NP treatment
Mono- and coculture conditions were used to mimic the human intestine, considering that enterocytes and goblet cells represent the two major intestinal cell types . The human epithelial cell line Caco-2 originates from a colon carcinoma and is typically used as a model of the intestinal barrier. Caco-2 cells exhibit the ability to differentiate, and so, they receive morphological and functional similarities to absorptive enterocytes in the small intestine. They develop a characteristic polarization with a brush border containing enterocyte-specific enzymes and cell-cell contacts . Additionally, the human colon adenocarcinoma cell line HT29 was used. HT29 cells show characteristics of mature intestinal cells and are, therefore, used for investigations of food digestion and bioavailability. They can also differentiate to acquire other morphological and functional properties. The addition of methotrexate (MTX) into the cell culture medium induces the differentiation to HT29-MTX cells. They are similar to goblet cells of the small intestine and are able to produce mucus .
Caco-2 and HT29-MTX cells were cultured in DMEM supplemented with 10% FBS, 1% non-essential amino acids, and 1% penicillin/streptomycin in an incubator (37 °C, 95% humidity, 5% CO2; Thermo Fisher Scientific Inc., USA). At regular intervals, the cells were tested for mycoplasma contamination. Both cell lines were verified by short tandem repeat profiling. For all assays, cells were seeded on ThinCert™ cell culture inserts in 12-well plates (3 μm pore size; Greiner Bio-One International GmbH, Germany). Caco-2 cells were used in mono- and also in coculture with HT29-MTX cells (ratio 3:1; altogether 3.3 × 105 cells per well). The cells were cultivated for 23 days to receive a differentiated and stable mono- and coculture. The cell culture medium was changed every 2–3 days. Differentiated cells were treated with 123–614 µM digested ZnO NP dispersed in cell culture medium. The choice of the concentrations was based on the study by Sohal et al. . They calculated physiologically relevant in vitro doses depending on the daily human intake of NP. There is no information about the daily ZnO NP intake. However, TiO2 is applied in comparable areas, especially in the food sector. Therefore, it is assumed that TiO2 and ZnO NP are taken up in similar quantities (5.4 mg/kg bodyweight/day). This results in an estimated in vitro dose of about 246 µM . Zinc chloride served as a source of free zinc ions and was applied in equimolar concentrations as ZnO. Cell culture medium was used for the untreated control. The solvent control consisted of intestinal juice diluted with cell culture medium (1:10) to exclude possible effects caused by the solvent. As positive control, 0.1% triton X-100 and 10 mM EGTA on the apical side and 10 mM EGTA on the basolateral side diluted with cell culture medium were used.
Quantification of zinc
Inductively coupled plasma mass spectrometry (ICP-MS) was used to investigate the uptake and permeation of digested ZnO NP through the monolayer. For the examination of the cellular amount of zinc, the cells were harvested after 24 h of exposure to ZnO NP. The cells were washed with phosphate-buffered saline (PBS), and Trypsin/EDTA was added for 5 min at 37 °C. Afterwards, DMEM with 10% FBS was added and the Transwell inserts were rinsed until the cells detached. The detached cells were centrifuged (5 min at 600xg; centrifuge 5810R; Eppendorf AG, Hamburg, Germany). The supernatants were removed, and the cells were washed with PBS. Cell viability and number were determined using the CASY TT (OLS OMNI Life Science GmbH & Co. KG, Bremen, Germany), and the cells were centrifuged again (5 min at 600×g). The cell pellets were diluted with 65% extra pure nitric acid and digested over-night. To support the cellular digestion, the samples were placed in a sonication bath (Transsonic 310/H; Elma Schmidbauer GmbH, Singen, Germany) for 1 h at 50 °C. Subsequently, the samples were diluted with purified water to 2% nitric acid. The ionic zinc content was measured using the ICP-MS iCAP™ RQ (Thermo Fisher Scientific Inc., Waltham, MA, USA) equipped with the 4DX prepFAST autosampler (ESI Elemental Service & Instruments GmbH, Mainz, Germany). The mass signal m/z = 66, which corresponds to the zinc isotope with a mass of 66 amu, was determined. Rhodium was used as an internal standard with a concentration of 2 ppb in 2% nitric acid. It was continuously added to the samples and quantified simultaneously. Three QC were chosen within the measurement range and measured regularly after about 20 samples. The minimum number and acceptance criteria for evaluating the QC were in agreement with the FDA guidelines . The limit of quantification was calculated after DIN 32645  as 0.3 ± 0.3 µg/l, and the measurement uncertainty was calculated according to DIN ISO 11352:2013-03  with three certified reference materials as 6.9%. Data processing was performed with an in-house developed program (from Chemical Analytics, German Environment Agency, Bad Elster, Germany) for quality management as well as Qtegra™ (Thermo Fisher Scientific Inc., Waltham, MA, USA).
In addition, zinc permeation through the monolayer was determined, which allows indirect conclusions about the cellular zinc uptake. The incubation dispersions were measured before treatment of the mono- and cocultured cells and afterwards from the apical and basolateral compartments. All samples were acidified with extra pure nitric acid, and the total ionic zinc content was quantified under the same conditions as described above. The limit of quantification was determined as 1.4 ± 0.7 µg/l.
Transmission electron microscopy evaluation
TEM was used to investigate the possible cellular consequences that result from the uptake of digested ZnO NP. For this, the cells were treated with 614 µM ZnO NP and ZnCl2 for 24 h. Afterwards, the cells were washed twice with PBS to remove NP residues from the cellular surface. For a primary fixation, 2.5% glutaraldehyde was added for 1 h at room temperature. The cells were washed two times with cacodylic acid sodium salt buffer (100 mM, pH 7.2) for 15 min. Finally, the Transwell insert membranes were excised and further processed as already described .
Barrier integrity investigations
Different approaches were used to investigate the influence of ZnO NP on the monolayer barrier integrity. By measuring the transepithelial electrical resistance (TEER), conclusions can be drawn about the monolayer integrity and the integrity of tight junctions as well as the ionic conductance of the paracellular pathway in the epithelial monolayer . The ohmic resistance was measured both before and after 24 h of incubation with ZnO NP to obtain relative TEER values using chopstick electrodes and an epithelial Volt/Ohm meter (EVOM2; World Precision Instruments GmbH, Friedberg, Germany).
Furthermore, FITC-dextran was used to investigate changes in the monolayer permeability due to the treatment with ZnO NP. Dextran will be transported transcellularly through the monolayer depending on its permeability. Dextran can then be quantified photometrically because of the linked FITC marker. After 24 h exposure to ZnO NP, the incubation medium was replaced by phenol red-free DMEM on the apical and basolateral compartments for 10 min to wash the cells. DMEM was removed again, and 500 µl 1 mg/ml FITC-dextran was added on the apical side and 1000 µl phenol red-free DMEM was added on the basolateral compartment. The cells were incubated at 37 °C on an orbital shaker for 6 h. After 30 min, 60 min, and 6 h, the basolateral DMEM was pipetted in triplicate onto a 96-well plate. The fluorescence was measured with a microplate reader (excitation: 485 nm, emission: 528 nm; Synergy 2; BioTek Instruments, Inc., Bad Friedrichshall, Germany), and the FITC-dextran concentration was calculated using a standard series (0.001–1000 µg/ml).
To examine morphological changes after ZnO NP treatment, the cells were stained with phalloidin-iFluor 488 and DAPI to visualize actin filaments in the cytoskeleton and cell nuclei. For this purpose, mono- and cocultured cells were treated with digested ZnO NP and ZnCl2 for 24 h. After two wash steps with PBS, they were fixed with 4% paraformaldehyde for 20 min. They were washed again and permeabilized with 0.1% triton-X 100 for 10 min. The cells were washed once more and stained with 0.1 µg/ml DAPI and 0.1% phalloidin-iFluor 488 for 1 h. To analyze the samples, the Transwell insert membranes were excised and transferred onto a microscopic slide. The evaluation was carried out using a laser scanning microscope (excitation: 488 nm, emission: 445 nm for DAPI and 525 nm for phalloidin-iFluor 488; LSM 780; Carl Zeiss AG).
The potential cytotoxic impact of digested ZnO NP was assessed by conducting different investigations. To determine the cellular metabolic activity, the MTT assay as an accepted and common cytotoxicity test was used . After 24 h of treatment with ZnO NP, the cell culture medium was removed and 0.5 mg/ml MTT was added to the apical and basolateral compartments (stock solution: 50 mg MTT powder in 10 ml PBS; working solution diluted with DMEM). The cells were incubated for 2 h at 37 °C. MTT was removed, dimethyl sulfoxide (≥ 99.8%) was added to the apical and basolateral compartments to solubilize the resulting formazan, and the plate was placed on an orbital shaker for 10 min. Finally, apical and basolateral dimethyl sulfoxide was mixed, and the samples were placed in triplicate onto a transparent 96-well plate. The absorbance was measured at 570 nm using a microplate reader (reference wavelength: 630 nm; Synergy 2; BioTek Instruments, Inc., Germany).
To detect ROS, the DCFH-DA assay was used. After 24 h incubation with digested ZnO NP, the cells were harvested as described above. The cells were centrifuged (5 min at 600×g), and the cell pellet was mixed with 1 µM DCFH-DA. The cells were incubated for 30 min at 37 °C. Hydrogen peroxide (5 mM, 30 min) served as a positive control after the DCFH-DA treatment. A CytoFLEX flow cytometer (Beckman Coulter GmbH, Krefeld, Germany) was used to measure 1 × 104 viable cells per sample (excitation: 488 nm, emission: 525 nm). Data processing was carried out using the CytExpert software (Beckman Coulter GmbH, Krefeld, Germany).
The membrane-permeant JC-1 dye was used to investigate the MMP. After 24 h of exposure to digested ZnO NP, the cells were harvested as described above. The cells were centrifuged (5 min at 600×g) and the supernatants were removed. The positive control was incubated with 100 µM CCCP for 5 min. All samples were treated with 8 µM JC-1 for 15 min at 37 °C. After washing the cells with PBS, 1 × 104 viable cells per sample were measured using the CytoFLEX flow cytometer (Beckman Coulter GmbH, Krefeld, Germany). The excitation wavelength was 488 nm. The emission was measured using two different wavelengths to differentiate between JC-1 monomers (525 nm) and aggregates (585 nm). The MMP results from the proportion of monomers and aggregates, which was evaluated using the CytExpert software (Beckman Coulter GmbH, Krefeld, Germany).
At least three independent experiments were performed for each assay. IBM SPSS Statistics (Version 26; IBM Corporation, Armonk, NY, USA) was used for data processing. To detect significant differences, one-way analysis of variance with the Ryan–Einot–Gabriel–Welsh post hoc test or the unpaired t test were conducted. The figures were generated using GraphPad Prism 5 (GraphPad Software, San Diego, CA, USA). The results shown in the figures represent means and standard deviations.