CeO2 nanoparticles induce pulmonary fibrosis via activating S1P pathway as revealed by metabolomics

Abstract

CeO₂ nanoparticles (NPs) have been shown to cause lung fibrosis, however, the exact underlying molecular mechanisms are poorly understood. In this study, we have conducted a mass spectrometry-based global metabolomic analysis of human bronchial epithelial BEAS-2B cells treated by CeO₂ NPs with different aspect ratios and assessed their toxicity on the bronchial epithelial cells by various cell-based functional assays. Although CeO₂ NPs at doses ranging from 12.5 μg/mL to 25 μg/mL displayed low cytotoxicity on the bronchial epithelial cells, the metabolomic analysis revealed a number of metabolites in the cellular metabolic pathways of sphingosine-1-phosphate, fatty acid oxidation, inflammation, etc. were significantly altered by CeO₂ NPs, especially those with high aspect ratios. More importantly, the robustness of metabolomics findings was further successfully validated in mouse models upon acute and chronic exposures to CeO₂ NPs. Mechanistically, CeO₂ NPs upregulated transforming growth factor beta-1 (TGF-β1) levels in BEAS-2B cells in an aspect ratio-dependent manner through enhancing the expression of early growth response protein 1 (EGR-1). In addition, both in vitro and in vivo studies demonstrated that CeO₂ NPs significantly induced the expression of sphingosine kinase 1 (SHPK1), phosphorylated Smad2/3 and lung fibrosis markers. Moreover, targeting SPHK1, TGFβ receptor or Smad3 phosphorylation significantly attenuated the fibrosis-promoting effects of CeO₂ NPs, and SPHK1-S1P pathway exerted a greater effect on the TGF-β1-mediated lung fibrosis compared to the conventional Smad2/3 pathway. Collectively, our studies have identified the metabolomic changes in BEAS-2B cells exposed to CeO₂ NPs with different aspect ratios and revealed the subtle changes in metabolic activities that traditional approaches might have missed. More importantly, we have discovered a previously unknown molecular mechanism underlying CeO₂ NP-induced lung fibrosis with different aspect ratios, shedding new insights on the environmental hazard potential of CeO₂ NPs.

Keywords: CeO2 nanorods/nanowires; aspect ratio; lung fibrosis; mass spectrometry; metabolomics; nanotoxicity.

Figure 1.

Physicochemical characterization of CeO₂ NPs. (A) The morphology of CeO₂ #1, CeO₂ #2, and CeO₂ #3 was analyzed by transmission electron microscopy (TEM). (B) The particle size and ζ-potential of the CeO₂ NPs in deionized water. (C) Ultrastructural images to depict the interaction of CeO₂ nanorods/nanowires with BEAS-2B cells using TEM. BEAS-2B cells were cultured with CeO₂ nanorods/nanowires in BEGM for 24 h at 25 μg/mL. CeO₂ # 1 (a, d); CeO₂ # 2 (b, e); CeO₂ # 3 (c, f). Panels d, e, and f show a higher magnification view of the labeled square area in panels a, b, and c, respectively. The circles in panels a/d highlight the CeO₂ # 1 nanorods taken up into membrane-lined subcelluar compartments. Scale bar for panels a, b and c = 2 μm; scale bar for panels d, e and f = 1 μm.

Figure 2.

Cell-based functional assays of untreated BEAS-2B cells (CTRL) and the BEAS-2B cells treated with three types of CeO₂ NPs. (A) The percentage of EdU positivity of untreated and CeO₂ NP-treated BEAS-2B cells (scale bar=50 μm). (B) The cell cycle distribution of untreated and CeO₂ NP-treated BEAS-2B cells. (C) The immunofluorescence intensity of PCNA in untreated and CeO₂ NP-treated BEAS-2B cells (scale bar=20 μm). (D) MTS assays of untreated and CeO₂ NP-treated BEAS-2B cells. (E) The GSH levels in untreated and CeO₂ NP-treated BEAS-2B cells. (F) The levels of IL-6, IL-8 and IL-1β in untreated and CeO₂ NP-treated BEAS-2B cells. (G) The relative mitochondria potential in untreated and CeO₂ NP-treated BEAS-2B cells.

Figure 3.

Metabolomic analysis of the BEAS-2B cells exposed to different types of CeO₂ NPs (high dose, 25 μg/mL). (A) Volcano plots of the significantly altered metabolic features between CeO₂ NP treated and untreated (CTRL) cells (n=6 per group). (B) Heat maps of differentially expressed metabolites between CeO₂ NP treated and untreated cells (n=6 per group). (C) The number of identified metabolites with significant changes between CeO₂ NP treated and untreated cells. (D) The number of commonly and differentially altered metabolites in the BEAS-2B cells treated with different types of CeO₂ NPs.

Figure 4.

The changes of representative metabolites in the BEAS-2B cells exposed to CeO₂ NPs. (A) The alterations of amino acids in the BEAS-2B cells treated with different types of CeO₂ NPs. (B) The altered levels of representative nucleosides/nucleotides in the BEAS-2B cells exposed to CeO₂ NPs. (C) The altered levels of representative organic acid metabolites in the BEAS-2B cells exposed to CeO₂ NPs.

Figure 5.

The changes of acylcarnitines, PGE2 and metabolites in S1P pathway in the BEAS-2B cells exposed to CeO₂ NPs. (A) The altered levels of acylcarnitines in cells treated with different types of CeO₂ NPs. (B) The alteration of PGE2 and metabolites in S1P pathway in the BEAS-2B cells exposed to CeO₂ NPs. (C-E) The heatmaps revealed that most of the representative metabolites were altered by CeO₂ NPs with different aspect ratios in a dose-dependent manner (high dose: 25 μg/mL; low dose: 12.5 μg/mL).

Figure 6.

Validation of the changes in representative metabolites in the mouse models upon acute exposures (40 h) to CeO₂ NPs. (A) H&E staining of the lung tissues from the mice exposed to the three types of CeO₂ NPs with high (2.0 mg/kg) or low dose (0.5 mg/kg) (scale bar=200 μm). (B) Enzymatic assays of S1P, PGE2 and arachidonic acid in the lung tissues of the mice exposed to the CeO₂ NPs with high or low doses.

Figure 7.

Lung fibrosis induced by chronic exposures (44 days, high dose, 2.0 mg/kg) to CeO₂ NPs and validation of the changes in representative metabolites. (A) Masson’s trichrome staining of the lung tissues from the mice with indicated treatments (CeO₂ NPs #1, CeO₂ NPs #1, CeO₂ NPs #1 and positive control SiO₂ NPs) (scale bar = 200 μm). (B) The levels of TGF-β1 in the BALF from the mice exposed to CeO₂ NPs with different aspect ratios. (C) The amount of total collagen in the lung tissues from the mice with indicated treatments. (D) The amount of CeO₂ NPs in the lung tissues from the mice exposed to CeO₂ NPs with different aspect ratios. (E) The ratio of collagen/CeO₂ NPs in the lung tissues with indicated treatments. (F-G) Enzymatic assays of alanine, asparagine, glutamine, histidine, serine, S1P, PGE2, lactate, arachidonic acid and fumaric acid in the lung tissues from the mice exposed to CeO₂ NPs.

Figure 8.

EGR1 is a upstream regulator of TGF-β1 in CeO₂ NP treated BEAS-2B cells. (A) EGR1 was enriched in the promoter region of TGF-β1, and knockdown of EGR1 significantly reduced its enrichment. (B) Luciferase reporter assays indicated that EGR1 directly induced the promoter activity of TGF-β1 in the BEAS-2B cells. (C) Both TGF-β1 and EGR1 were increased by CeO₂ NPs in an aspect ratio dependent manner, and EGR1 depletion suppressed the expression of TGF-β1. (D) EGR1 downregulation significantly decreased the levels of TGF-β1 in the culture media derived from the BEAS-2B cells with indicated treatments.

Figure 9.

CeO₂ NPs induced the expression of SPHK1, pSmad2, pSmad3 and fibrosis markers COL1A1 and α-SMA in BEAS-2B cells. (A) Western blot analysis of the expression of SPHK1, pSmad2, Smad2, pSmad3 and Smad3 in the BEAS-2B cells exposed to CeO₂ NPs. (B) Immunofluorescence of COL1A1 and α-SMA in the BEAS-2B cells exposed to different types of CeO₂ NPs (scale bar=50 μm).

Figure 10.

As indicated by Western blot analysis, the expression of EGR1, SPHK1, pSmad2 and pSmad3 were significantly increased in the lung tissues of the mice upon chronic exposures (44 days, high dose, 2.0mg/kg) to CeO₂ NPs in an aspect ratio dependent manner.

Figure 11.

SPHK1-S1P pathway exhibited a greater effect on the TGF-β1-mediated lung fibrosis compared to the conventional Smad2/3 pathway. (A) Blocking the interaction between TGF-β1 and TGFBR suppressed the expression of SPHK1 and phosphorylation of Smad2 and Smad3. As expected, SPHK1 inhibtor SKI II significantly decreased SPHK level and the Smad3 inhibitor SIS3 dramatically reduced the levels of pSmad2 and pSmad3 in BEAS-2B cells. The protein expression levels were measured by Western blotting. (B) Targeting TGFBR almost completely obliterated the fibrosis-enhancing effects of CeO₂ NPs, and targeting SPHK1 exhibited a more attenuation effect on the expression of COL1A and α-SMA, when compared to targeting Smad3 phosphorylation (scale bar=50 μm).

Figure 12.

A schematic diagram illustrating the proposed mechanism for CeO₂ NP-induced lung fibrosis.