Sintered indium-tin oxide particles induce pro-inflammatory responses in vitro, in part through inflammasome activation

Abstract

Indium-tin oxide (ITO) is used to make transparent conductive coatings for touch-screen and liquid crystal display electronics. As the demand for consumer electronics continues to increase, so does the concern for occupational exposures to particles containing these potentially toxic metal oxides. Indium-containing particles have been shown to be cytotoxic in cultured cells and pro-inflammatory in pulmonary animal models. In humans, pulmonary alveolar proteinosis and fibrotic interstitial lung disease have been observed in ITO facility workers. However, which ITO production materials may be the most toxic to workers and how they initiate pulmonary inflammation remain poorly understood. Here we examined four different particle samples collected from an ITO production facility for their ability to induce pro-inflammatory responses in vitro. Tin oxide, sintered ITO (SITO), and ventilation dust particles activated nuclear factor kappa B (NFκB) within 3 h of treatment. However, only SITO induced robust cytokine production (IL-1β, IL-6, TNFα, and IL-8) within 24 h in both RAW 264.7 mouse macrophages and BEAS-2B human bronchial epithelial cells. Our lab and others have previously demonstrated SITO-induced cytotoxicity as well. These findings suggest that SITO particles activate the NLRP3 inflammasome, which has been implicated in several immune-mediated diseases via its ability to induce IL-1β release and cause subsequent cell death. Inflammasome activation by SITO was confirmed, but it required the presence of endotoxin. Further, a phagocytosis assay revealed that pre-uptake of SITO or ventilation dust impaired proper macrophage phagocytosis of E. coli. Our results suggest that adverse inflammatory responses to SITO particles by both macrophage and epithelial cells may initiate and propagate indium lung disease. These findings will provide a better understanding of the molecular mechanisms behind an emerging occupational health issue.

Fig 1. Transmission electron micrographs of cells treated with ITO production particles.

Representative images of A. RAW and B. BEAS-2B cells after 3 h exposures to particle suspensions at 50 μg/ml. Images were acquired at 10,000 times magnification using a 5 kV accelerating voltage. Insets show magnified engulfed particles. Scale bars, 2 μm or 500 nm.

Fig 2. Pre-uptake of SITO or VD impairs phagocytosis.

A. RAW cells were treated with particle suspensions for 3, 6, or 12 h or Cytochalasin D (Cyto D) 3 h. Cells were washed, pHrodo Red E. coli BioParticles were added for 2 h, then plates were read to measure changes in fluorescence. Error bars represent the mean ± SD (n = 4 for 3 h and 6 h, n = 3 for 12 h). *, p < 0.05 compared to PBS. B. The same as in A, except cells were exposed to particle suspensions for 6 h on coverslips, fixed with formaldehyde following E. coli particle incubations, and mounted onto slides. Images were acquired at 60x magnification using dual-mode fluorescence and enhanced darkfield microscopy to show phagocytosed E. coli (red) and cell-associated particles (bright spots). Scale bar, 10 μm.

Fig 3. ITO production particles cause IκBα degradation.

Representative western blot images of IκBα and actin from treated RAW (A) and BEAS-2B (C) cell lysates. Cells were treated with particle suspensions for 1 or 3 h, washed twice with PBS, and lysed. B, D. Densitometry analysis was performed and is represented as the percentage of IκBα compared to levels present in PBS control-treated lysates. Error bars represent the mean ± SD (n = 4). *, p < 0.05 compared to PBS.

Fig 4. Nuclear accumulation of p65 following particle exposures.

RAW (A) and BEAS-2B (B) cells were grown on coverslips and treated with particle suspensions for 3 h as in Fig 3. Prior to adding lysis buffer to wells, coverslips were removed, fixed, and stained with antibodies against NFκB subunit p65. Coverslips were mounted onto slides with Fluoromount G (DAPI) and sealed with nail polish. Experiments were performed in duplicate, and cells were imaged at 40x. Scale bar, 20 μm.

Fig 5. Pro-inflammatory cytokine production with particle exposures.

Cells were treated with particle suspensions or 1 μg/ml of LPS as a positive control for 24 h. Media were collected, and ELISAs were run to measure levels of IL-6, TNFα, and IL-1β from RAW cells (A) or IL-6 and IL-8 from BEAS-2B cells (B). Error bars represent the mean ± SD (n = 4). *, p < 0.05 compared to PBS.

Fig 6. Inflammasome activation in RAW cells by SITO particles.

A. RAW cells were pre-treated with either 40 μM Z-YVAD-FMK (caspase-1 inhibitor) or 200 μM Glybenclamide for 1 h. Cells were then treated for 24 h with SITO (50 μg/ml), LPS alone (1 μg/ml), or Min-U-sil following LPS priming (LPS + Si; 150 μg/ml). Media were collected and ELISAs were run to measure levels of IL-1β (n = 4). *, p < 0.05 compared to no inhibitor conditions. B. Cells were treated as in A (without inhibitors) for the time points listed, lysed, and 100 μg lysates were assayed for caspase-1 activity. Values were normalized to PBS controls. *, p < 0.05 compared to PBS at that time point. C. Media collected from (B) were measured for IL-1β at the 12 h time point (n = 3). Error bars represent the mean ± SD (n = 3). *, p < 0.05 compared to PBS.

Fig 7. Particle-induced inflammasome activation in LPS-primed cells.

RAW cells were primed with LPS for 3 h, followed by particle treatments for 24 h. Media were collected, and ELISAs were run to measure levels of IL-1β. Error bars represent the mean ± SD (n = 4). *, p < 0.05 compared to LPS priming alone.

Fig 8. Cytokine production from baked SITO exposures.

Cells were treated with SITO, baked SITO, or LPS for 24 h as in Fig 5. Media were collected, and ELISAs were run to measure levels of IL-6, TNFα, and IL-1β from RAW cells (A) or IL-6 and IL-8 from BEAS-2B cells (B). Error bars represent the mean ± SD (n = 4). *, p < 0.05 compared to PBS.

Fig 9. Baked SITO pre-treatment impairs phagocytosis.

The phagocytosis assay was carried out as in Fig 2, but with baked SITO for comparison. Error bars represent the mean ± SD (n = 3). *, p < 0.05 compared to PBS at that time point.

Fig 10. Baked SITO cytotoxicity in RAW and BEAS-2B cells.

Treatments with SITO, baked SITO, LPS alone, or Min-U-sil (Si) were carried out for 24 and 48 h in RAW (A) and BEAS-2B (B) cells. MultiTox-Fluor Reagent (Promega) was added to all wells for 1 h at 37° C, and plates were read at 400ex/505em to determine the fluorescence signal from live cells. Values were normalized to PBS controls. Error bars represent the mean ± SD (n = 4). *, p < 0.05 compared to PBS.