Fig 2: Association of PPARd expression with aggressive behavior of breast cancer cells. (a) Overall survival of 2500 breast cancer patients as a function of PPARD gene expression in their biopsies. (b) PPARd expression by immunoblotting in clones of rat mammary adenocarcinomas with ß-actin used as a loading control. Numbers of colonies from plating 5 × 103 cells in soft agar are shown for each clone.12 (c) PPARD expression was measured by RT–PCR in the nine human breast cancer cell lines described in the table. The average and standard error of PPARD expression for the basilar and luminal cell lines is shown in the bottom graph. (d) Two groups of NSG mice (n=5) were injected in the mammary fat pad with MCF-7 cells transfected with either a PPARD expression vector (clone 7 with high PPARD expression) or the vector alone. Mice were killed after 21 days and local tumor volumes measured with calipers. Numbers of tumor colonies in the lungs were determined by visual inspection. **P<0.05.

Fig 3: Migration and growth of PPARDhi, knockout and control MCF-7 cells in conventional glucose conditions. (a) Transwell invasion assays were performed as described in the materials and methods in the presence or absence of the PPARD antagonists DG172 or NXT1511 (3 µM). Cells that migrated to the bottom of the insert were counted after 96 h. Pictures of the stained and fixed inserts are shown in the upper panels ( × 10 magnification) and the average and standard error of the numbers of large, blue migrated cells are shown in the lower graph. (b) Cells were plated at an initial concentration of 104 cells/ml in 24-well plates in DMEM+5% FBS and counted manually on the indicated days. The average and standard error of the results of three different counts per well are shown. Experiments were repeated three times with similar results. **P<0.05.

Fig 4: Survival of breast cancer cells in low glucose and other harsh conditions as a function of PPARD expression. (a) Control, PPARDhi and PPARD knockout MCF-7 cells were cultured in glucose-free RPMI+5% non-dialyzed fetal calf serum. Percentages of viable cells that excluded 7-aminoactinomycin D (7AAD) were then determined at the indicated times by flow cytometry. The numbers in the scatter plots are the percentages of viable 7AAD- cells. Averages and standard errors from three different experiments are shown in the lower graph. (b) Control MCF-7 cells were cultured in low-glucose conditions with or without dexamethasone (DEX) (30 µM) (left panels) or GW0742 (1 µM) or GW50516 (100 nM) (right panels) to increase PPARD expression. PPARD (top panels) and percentages of 7AAD- cells (bottom panels) were measured by RT–PCR or flow cytometry after 72 h. Each line represents the results from a different experiment and the average of all experiments was used for statistical calculations. Results with the two synthetic agonists were pooled, as indicated by GW on the x axis of the right bottom graph. (c) MCF-7 cells infected with lentiviruses (left panel) or SKB-R3 cells infected with retroviruses (right panel) expressing PPARD or the vectors alone were cultured in low-glucose conditions. Viable cells were determined by flow cytometry after 3 days. The average and standard errors of two separate experiments are shown. PPARD expression by RT–PCR under normal and low-glucose conditions after 2 days is shown in the other graphs. (d) Viable cells were determined after 2 days of culture in conventional conditions with or without thapsigargin or hypoxia or in low-glucose conditions. Averages and standard errors of three separate experiments are shown. (e) Cells were cultured in low-glucose conditions and levels of PERK and CHOP were determined at the indicated times by immunoblotting using ß-actin as a loading control. **P<0.05.

Fig 5: Effect of PPARd on glucose deprivation-induced oxidative stress and catalase expression. (a) PPARDhi and vector control MCF-7 cells were cultured in normal or low-glucose conditions for 24 h and then stained with DCFH as a measure of reactive oxygen species (ROS) levels. Examples of the flow cytometric analyses are shown on the left. The numbers in the histograms represent the mean fluorescence intensity (MFI) of DCFH staining. Averages and standard errors of three separate measurements are shown on the right. (b) MCF-7 cells were cultured in low-glucose condition for 0, 4, 8, 24 and 48 h. PPARD and CAT levels were measured at these times by RT–PCR. (c) The cells were cultured in low-glucose with or without catalase (20 µg/ml) for 3 days and the percentages of viable 7-aminoactinomycin D (7AAD)- cells were then determined by flow cytometry. (d) MCF-7 cells infected with PPARD-expressing or control lentiviruses were cultured in normal or low-glucose conditions for 24 h and then stained with DCFH. Averages and standard errors of two to three separate measurements are shown. **P<0.05.