Fig 1.
The obesogen tributyltin (TBT) increases lipid content in HepaRG human liver cells.
A) HepaRG cells were grown in the presence of TBT (5–50 nM) or vehicle (DMSO) for 14 days, lipid droplets were stained with LipidTox and imaged using deconvolution epifluorescence microscopy; max projection images are shown. B-E) High content image analysis of the effects of TBT at the single cell and single lipid droplet level. n> 400 cells/condition analyzed from two independent experiments. *p<0.05 by ANOVA (Kruskal-Wallis test). Scale bar: 25 μm (10 μm inset).
Fig 2.
TBT increases lipid content in a time- and dose-dependent manner in HepaRG cells.
A) HepaRG cells were grown in the presence of TBT (50 nM) or vehicle for 1 to 14 days and lipid droplets were stained with LipidTox. Y axis is fold change over DMSO treated cells. B) heatmap showing total lipid content/cell after treatment with TBT (50 pM-50 nM) for 1 to 3 days. C) dose-response curve and EC50 calculation at the 72 hr time point. D-F) box plots representing single cell data from panel B. *p<0.05 by ANOVA (Kruskal-Wallis test).
Fig 3.
TBT increase in lipid droplets is mediated through glycolysis and activation of some lipogenic genes.
A) HepaRG cells were grown in the presence of TBT (50 nM), 2-deoxyglucose (2DG, 2 mM) or vehicle for 3 days, then lipid content was analyzed as in Fig 1. Representative 40x images are shown. Scale bar: 25μm. B) quantification of images from the experiment performed in panel A, data is z-normalized based on vehicle (DMSO) treatment and shown as integrated intensity of lipid droplets/cell. *p<0.01 compared to Vehicle; **p<0.01 compared to TBT. C-D) Single molecule RNA FISH for SREBF1 and FASN was performed in HepaRG cells after 8 and 24 h (shown in the panels) of TBT or vehicle treatment. Images (60x/1.4NA) shown are max projections after deconvolution. In panel D, the number of single RNA molecules/cell was counted and represented as a box plot. Scale bar: 10 μmM. E) RT-qPCR validation of TBT induction of SREBF1 and FASN mRNAs after 24h of 50nM TBT treatment. *p<0.01 compared to Vehicle.
Fig 4.
TBT down-regulates RXRA nuclear levels in a time- and dose-dependent manner in HepaRG cells.
A) HepaRG cells were grown in the presence of TBT (50 nM) or vehicle for 1 to 3 days, then immunofluorescence was performed using specific RXRA antibody. Representative 20x images (max projection) are shown. Scale bar: 25 μm. B) Fold change of RXRA down-regulation at 72h is reproducible across five independent experiments. C-D) box plots representing single cell data from panel A after high throughput microscopy and image analysis. *p<0.05 by ANOVA (Kruskal-Wallis test with Dunn post test); in panel B a non-parametric t-test was used.
Fig 5.
TBT down-regulates RXRA in a fast, proteasome-dependent and PPARG-independent manner.
A) HepaRG cells were treated TBT (50 nM) or vehicle for 2–8 hours, then immunofluorescence using specific RXRA antibody was performed and single cell nuclear levels of RXRA measured. B-C) Effect of MG132 and TO070907 on TBT (50 nM, 8 h) down-regulation of RXRA represented as box-plots. *p<0.05 by ANOVA (Kruskal-Wallis test with Dunn post test).
Fig 6.
RXRA levels do not correlate with lipid content before or after TBT treatment.
A) HepaRG cells were treated TBT (50 nM) or vehicle for 72 h hours, then immunofluorescence using specific RXRA antibody (red), together with LipidTox staining (green), was performed and single cell nuclear levels of RXRA, lipid droplet counts and total lipid content were measured. Scale bar: 25μm. B) Scatter plots showing no significant Spearman r correlation coefficients under both measurement and treatments. C) Table showing the % of cells having either high or low RXRA nuclear level and number of lipid droplets as measured considering the mean values in the vehicle control samples.