Advanced Respiratory Models for Hazard Assessment of Nanomaterials-Performance of Mono-, Co- and Tricultures

Abstract

Advanced in vitro models are needed to support next-generation risk assessment (NGRA), moving from hazard assessment based mainly on animal studies to the application of new alternative methods (NAMs). Advanced models must be tested for hazard assessment of nanomaterials (NMs). The aim of this study was to perform an interlaboratory trial across two laboratories to test the robustness of and optimize a 3D lung model of human epithelial A549 cells cultivated at the air-liquid interface (ALI). Potential change in sensitivity in hazard identification when adding complexity, going from monocultures to co- and tricultures, was tested by including human endothelial cells EA.hy926 and differentiated monocytes dTHP-1. All models were exposed to NM-300K in an aerosol exposure system (VITROCELL^® cloud-chamber). Cyto- and genotoxicity were measured by AlamarBlue and comet assay. Cellular uptake was investigated with transmission electron microscopy. The models were characterized by confocal microscopy and barrier function tested. We demonstrated that this advanced lung model is applicable for hazard assessment of NMs. The results point to a change in sensitivity of the model by adding complexity and to the importance of detailed protocols for robustness and reproducibility of advanced in vitro models.

Keywords: 3D lung model; NM-300K; air–liquid interface; nanotoxicology; tricultures.

Figure 1

Scheme of mono-, co-, and triculture lung models of A549, EA.hy926, and differentiated THP-1 (dTHP-1) cells cultivated at the air–liquid interface (ALI) on membrane inserts.

Figure 2

Triculture model investigated by confocal microscopy. (a,b) Z-stack image series (2D x-y view and respective side views) showing the distribution of A549, EA.hy926, and dTHP-1 (*, green) cells on the opposite sides of a transwell insert (arrow). (a) The 2D x-y view from the A549 and dTHP-1 side (z-stack thickness 44.5 µm). (b) The 2D x-y view from the EA.hy926 side (z-stack thickness 27.5 µm). (c–e) dTHP-1 cells in different morphologies on top of the A549 cells. Red: cellular membranes stained with Cell Mask red dye, blue: nuclei counterstained with DAPI, green: dTHP-1 cells stained with Cell tracker green dye. Magnification: 40×. Scale bar 10 µm (a,b) and 50 µm (c–e). dTHP-1: differentiated THP-1.

Figure 3

Transmission electron microscopy micrographs (a,b) and confocal picture © of unexposed triculture with A549 and dTHP-1 cells at the apical side and EA.hy926 cells at the basolateral side of a cell insert with 1 µm pore size membrane. LM: lamellar bodies; v: vesicles; N: nuclei; black arrows: microvilli. (a) Scale bar: 5 µm. (b) Scale bar: 1 µm. (c) Confocal image of the apical side. A549 were stained with ZO-1 antibody (green) and differentiated THP-1 with CD11b (red). White arrows: tight junctions. **: cytoplasmic ZO-1. Nuclei are stained in blue (DAPI). Scale bar: 50 µm; Magnification: 40×.

Figure 4

Permeation of fluorescein through the cellular layers. Shown is mean recovery of fluorescein ± SD in the basolateral compartment relative to the total deposition of fluorescein at the apical side of the membrane after exposure of cocultures in the cloud system to phosphate buffered saline (PBS), dispersant NM-300K DIS, or NM-300K at 1 or 10 µg/cm². The results are based on 4 technical replicates in n = 3 independent experiments. No statistically significant difference was seen between PBS treatment and the other samples, analyzed by ordinary one-way ANOVA with multiple comparisons, post-test Dunnett’s, p > 0.05. SD: standard deviation.

Figure 5

Relative viability measured by the AlamarBlue assay after exposure of monocultures (a,d), cocultures (b,e), and tricultures (c,f) of A549, EA.hy926, and dTHP-1 cells to NM-300K and control solutions at the air–liquid interface (ALI). Experiments were performed at Lab 1 (a–c) and Lab 2 (d–f) for comparison. A reduction in viability was seen after exposure to control solutions and NM-300K at ALI. Results are normalized against negative incubator control (NC, set to 100%) and presented as boxplots with mean (+), median/50th percentile (line), 25th and 75th percentiles (box), and minimum and maximum values (whiskers). A total of (a) n = 6–9, (b) n = 4–5, (c) n = 4, (d,e) n = 3, and (f) n = 4 independent experiments with each 1–2 replica cell culture inserts were performed. Statistically significant differences compared to PBS control were analyzed by ordinary one-way ANOVA with multiple comparisons post-test Dunnett´s, and are indicated by * p < 0.5, ** p < 0.1. NC: negative control (incubator control), PBS: phosphate buffered saline, dispersant: NM-300K dispersion medium (NM-300K DIS); dTHP-1, differentiated THP-1 cells.

Figure 6

DNA strand breaks and oxidized DNA lesions, measured as DNA tail intensity, by the comet assay with Fpg in cells exposed at the air–liquid interface (ALI) to NM-300K. Cells were cultured as (a) monocultures, (b,d) cocultures, and (c,e) tricultures. Results are presented as mean of median ± SD of (a) n = 3–5, (b–e) n = 3 independent experiments. From each experiment, the median DNA tail intensity (%) was calculated from 50 cells per 2–6 gels from 1–2 cell culture inserts. Significantly different effects on DNA damage compared to PBS control were analyzed by ordinary one-way ANOVA followed by Dunnett’s post-hoc test (*** p < 0.001). SBs: strand breaks, NC: negative control, PBS: phosphate buffered saline, SD: standard deviation.

Figure 7

Confocal microscopy investigation of tricultures with and without NM-300K exposure. Confocal pictures show the apical side of triculture model in (a) negative incubator control, and (b,c) after exposure to NM-300K at 10 µg/cm². A549 cells were stained with pro-surfactant protein C (green) and dTHP-1 are marked with CD11b (red). DNA is stained with DAPI (blue). (c) Phase contrast picture combined with immunofluorescence staining. (a–c) Magnification: 40×. Scale bar: 50 µm (a,b); 20 µm (c).

Figure 8

Representative transmission electron micrographs of A549 cells on the apical side of the triculture model. (a) Unexposed cells from incubator control (NC) showed no electron dense vacuoles. (b) Electron dense vacuoles-like structure and lamellar bodies were seen in cells exposed to NM-300K 10 µg/cm². (c,d) Higher magnification of the cells exposed to NM-300K 10 µg/cm² showed that NM-300K were found in the cell cytoplasm (c) or inside of a vacuole (d). (d) NM-300K were found as single particles (5–20 nm) (1,2) or in small aggregate of 100 nm (3). Scalebar: 2 µm (a,b); 1 µm (c); 500 nm (d). N: nucleus, LM: lamellar bodies, v: vacuoles-like structure, p: NM-300K nanoparticles, m: mitochondria.