Evaluation of a screening system for obesogenic compounds: screening of endocrine disrupting compounds and evaluation of the PPAR dependency of the effect

Abstract

Recently the environmental obesogen hypothesis has been formulated, proposing a role for endocrine disrupting compounds (EDCs) in the development of obesity. To evaluate this hypothesis, a screening system for obesogenic compounds is urgently needed. In this study, we suggest a standardised protocol for obesogen screening based on the 3T3-L1 cell line, a well-characterised adipogenesis model, and direct fluorescent measurement using Nile red lipid staining technique. In a first phase, we characterised the assay using the acknowledged obesogens rosiglitazone and tributyltin. Based on the obtained dose-response curves for these model compounds, a lipid accumulation threshold value was calculated to ensure the biological relevance and reliability of statistically significant effects. This threshold based method was combined with the well described strictly standardized mean difference (SSMD) method for classification of non-, weak- or strong obesogenic compounds. In the next step, a range of EDCs, used in personal and household care products (parabens, musks, phthalates and alkylphenol compounds), were tested to further evaluate the obesogenicity screening assay for its discriminative power and sensitivity. Additionally, the peroxisome proliferator activated receptor γ (PPARγ) dependency of the positive compounds was evaluated using PPARγ activation and antagonist experiments. Our results showed the adipogenic potential of all tested parabens, several musks and phthalate compounds and bisphenol A (BPA). PPARγ activation was associated with adipogenesis for parabens, phthalates and BPA, however not required for obesogenic effects induced by Tonalide, indicating the role of other obesogenic mechanisms for this compound.

Figure 1. Overview of the experimental setup of one compound and insulin-compound co-treatment experiment of the 3T3-L1 cells.

Medium was changed every 2-3 days. Insulin: 10 µg/mL insulin; MDI: 10 µg/µL insulin, 0.25 µM dexamethasone and 0.5 mM isobutylmethylxanthin.

Figure 2. Pictures of Oil Red O and Nile Red stained 3T3-L1 cells after 10 days of exposure to reference compounds and solvent/positive control.

Light microscopic pictures (A; Oil Red O staining) and fluorescent (b; Nile Red staining) were taken. Pictures were only taken for visualisation of the differentiation and not for quantification. Scale bar represents 50µm.

Figure 3. Effect of reference compounds on adipocyte differentiation in 3T3-L1 cells after 10 days of exposure.

Fluorimetric quantification of lipid accumulation was performed using Nile Red staining. The degree of lipid accumulation induced by TBT in absence (A.) or presence (B.) of insulin and by ROSI in absence (C.) or presence (D.) of insulin are represented as mean (±SD) of three independent experiments Data represent mean (±SD) of the degree of lipid accumulation (with insulin) (DLA(I)) of three independent experiments (each with 4 replicates) (n=3). Significant differences with solvent control (0.1% DMSO or INS) are indicated with asterisks (One Way ANOVA Dunnet’s post hoc test; *p≤0.05; **p≤0.01; ***p≤0.001). Additionally, a sigmoidal dose−response curve was fitted (GraphPad Prism). LAT: Lipid Accumulation Threshold.

Figure 4. Effect of test compounds on adipocyte differentiation in 3T3-L1 cells after 10 days of exposure.

Fluorimetric quantification of lipid accumulation was performed using Nile Red staining. The degree of lipid accumulation (DLA) induced by Alkylphenol (A.), Phthalate (B.) Paraben (C.) and Musk compound (D.) exposure are represented as mean (±SD) of three independent experiments (each with 4 replicates) (n=3). Significant differences with solvent control (0.1% DMSO) are indicated with asterisks (One Way ANOVA Dunnet’s post hoc test; *p≤0.05; **p≤0.01; ***p≤0.001). LAT: Lipid accumulation threshold .

Figure 5. Gene expression of the adipocyte specific gene adipocyte specific protein 2 (aP2) measured by Real-time PCR at day 10.

Data are represented as the mean fold change relative to the solvent control of 3 biological replicates (mean±stdev; n = 3). Significant differences between the conditions were analysed with One Way ANOVA (Dunnet’s post hoc test; *p≤0.05; **p≤0.01; ***p≤0.001).

Figure 6. Dose-response relationships of strong obesogens based on adipocyte differentiation in 3T3-L1 cells.

Data represent the same data as figure 2 and 3 but with 10 concentrations shown. Mean of the degree of lipid accumulation (DLA) of three independent experiments are shown. LAT: Lipid accumulation threshold .

Figure 7. Effect of combined insulin and test compound exposure on adipocyte differentiation in 3T3-L1 cells after 10 days of exposure.

Fluorimetric quantification of lipid accumulation was performed using Nile red staining. The degree of lipid accumulation with insulin (DLAI) induced by Alkylphenol (A.), Phthalate (B.) Paraben (C.) and Musk compound (D.) exposure are represented as mean (±SD) of three independent experiments (each with 4 replicates) (n=3). Significant differences with control (INS) are indicated with asterisks (One Way ANOVA Dunnet’s post hoc test; *p≤0.05; **p≤0.01; ***p≤0.001); DLAI: Degree of Lipid Accumulation with Insulin; LAT: Lipid accumulation threshold .

Figure 8. PPARγ dependency of the obesogenic effects of reference and test compounds.

3T3-L1 cells were exposed to Rosiglitazone 100 nM (A), 500 nM (B) TBT 50 nM (C) and screened obesogenic compounds (D) together with a concentration range of the PPARγ antagonist T0070907(A, B, C) or with and without 10 µM T0070907 (D). After 10 days of exposure fluorescent quantification of lipid accumulation was performed using Nile red staining. Data represent the degree of lipid accumulation (DLA) (mean ±SD) of 4 replicates. Significant differences (p≤0.05) between conditions are indicated with different letters for the optimisation tests (A, B, C) and significant differences between screened obesogens with and without T0070907 (D) are indicated with asterisks (*p≤0.05; **p≤0.01; ***p≤0.001) (One Way ANOVA Tukey’s post hoc test).

Figure 9. Effect of reference and test compounds on PPARγ activation in the PPARγ CALUX cell line.

After 24h of exposure luciferase activity was measured with a luminometer. PPARγ activation of reference compounds ROSI (A.) and TBT (B.) and screened EDCs: Alkylphenol (C.), Phthalate (D.) Paraben (E.) and Musk (F) compounds are represented as mean (±SD) of the fold induction of at least three independent experiments performed in triplicate. Significant differences with solvent control (1% DMSO) are indicated with asterisks (One Way ANOVA Dunnet’s post hoc test; *p≤0.05; **p≤0.01; ***p≤0.001). LOQ limit of quantification.