Fig 1.
Baseline features of PANC-1, MIA PaCa2, and HPAF-2 pancreatic cancer cells in EMT related gene expression, CSC population, and resistance to gemcitabine.
A-D. mRNA levels of E-cadherin (E-Cad), Zo-1, Zeb-1 and Snail-1. RT-PCR data was normalized to 18s rRNA and represented as mean ± SD of 2-ΔCt of triplicate determinations of 3 individual experiments. E. Western blots for E-cadherin and N-cadherin expression in 4 pancreatic cancer cell lines. Actin was a loading control. F. Western blots for E-cadherin expression in PANC-1 cells treated with sodium butyrate. G. Flow cytometry identification of CSCs in PANC-1 cells. CD24+/CD44+/EpCAM+ subpopulation were detected as pancreatic cancer CSCs. PANC-1 cells were triple stained with PE-conjugated anti-CD24, PE-Cy7-conjugated anti-CD44 and APC-conjugated anti-EpCAM. DAPI staining was used for identification of living cells. Cells were analyzed with multi-label flow cytometry. The upper-right quadrant showed CD44+/ EpCAM + cells within CD24+ gated population. H. Resistance of PANC-1 cells to gemcitabine treatment. PANC-1 cells viability was determined at 72 hrs of incubation using MTT assays. Data represent Mean ± SD of triplicate measurements of 4 individual experiments. Gemcitabine concentrations up to 5 mM were used but failed to achieve 90% cell death.
Fig 2.
Validation of the immunofluorescent HTS assay.
A. Average signal and standard deviation of the positive and negative controls, compared to the average of all screened compounds. Total 3,168 compounds were from Microsource (2,320 compounds) and Prestwick (848 compounds) compound libraries. Every plate included 16 wells of vehicle (0.35% DMSO) treated cells, and 16 wells of cells treated with 3.5 mM sodium butyrate (SB). Avg +1, +2 and +3 represent signals 1, 2 or 3 standard deviations above average. B. Representative images of E-cadherin immunofluorescence of PANC-1 cells treated by 0.35% DMSO, or 3.5 mM sodium butyrate (SB). E-cadherin was detected by anti-E-cad primary antibody (1:250 dilution) followed by Alexa594 conjugated secondary antibody (1:500 dilution). C. The relative fluorescence (ratio of AlexaFluor to Hoechst, fold over DMSO vehicle) was plotted against individual wells to visualize the data spread. The median for control wells (DMSO vehicle) was 1-fold, with a 0.1-fold standard deviation. The HCS assay cutoff was 3 standard deviations above the median (1.437-fold), marked by the red line. Eighty four compounds had readings greater than or equal to this cutoff.
Fig 3.
A. Scheme of the HTS. B. The percent of red fluorescence, relative to DMSO vehicle, was plotted against individual wells to visualize the data spread. The red line represents 3 standard deviations above average.
Fig 4.
Red curves represent the RFUs using the red fluorophore Alexa Fluor 594 conjugated 2nd antibody in the initial screening, green curves represent RFUs using the green fluorophore Alexa Fluor 488 conjugated 2nd antibody in the secondary screening.
Fig 5.
Structure, cytotoxicity and induction of E-cadherin fluorescence by hit compounds.
E-cadherin immunofluorescence of PANC-1 cells treated by 6 selected compounds was detected at 24 hrs of treatment by anti-E-cad primary antibody (1:250 dilution) and Alexa 488 conjugated secondary antibody (1:500 dilution). Sensitivity of PANC-1, BxPC-3, L3.6 and hTERT-HPNE cells to the compounds were detected at 48 hrs treatment by MTT assay. Data represents Mean ± SD of 1–3 independent experiments each done in triplicate.
Fig 6.
Induction of E-cadherin expression and inhibition of HDACs by hit compounds in PANC-1 cells.
A. Western blot analysis of E-cadherin in PANC-1 cells that were exposed to two different concentrations of cluster#1 and 2 compounds. Concentrations for compound 150 were 0.5 and 1 μM. B. Western blot analysis of E-cadherin in PANC-1 cells that were exposed to 25 μM of cluster#2 compounds for 24 hrs. C. Western blot analysis in PANC-1 and BxPC-3 cells for E-cadherin, N-Cadherin, and Snail. Cells were treated with BSI (25 μM) for 24hrs. D. Western blot analysis in PANC-1 and BxPC-3 cells for H2A-Lys5, H3-Lys9, and H4-Lys8. Cells that were exposed to BSI (25 μM) for 24hrs.
Fig 7.
Inhibition in cell invasion and migration of pancreatic cancer cells by BSI.
A, B. Matrigel invasion assays for PANC-1 cell migration and invasion. Cells were exposed to 25 μM BSI. Cell migration (without Matrigel) and invasion (with Matrigel) were detected at 24hrs. Bar graph (B) shows the average number of migrated/invaded cells per field (Mean ± SD of at least 5 fields per experiment for 3 repeated experiments). C, D. Scratch assay for BxPC-3 pancreatic cancer cell migration. Scratch was made on confluent monolayer and then exposed to 25 μM BSI. Cell migration was measured at 24 and 30 hrs post BSI treatment. Bar graph (D) shows the % distance covered by BxPC-3 cells. Data represents Mean ± SD of 3 repeats.
Fig 8.
Reduction of the pancreato sphere formation by BSI treatment.
BxPC-3 cells (A, B, C) and PANC-1 cells (D, E, F) were seeded into ultra-low attachment 24-well plates at 4,000 cells/well, and were exposed to 25 μM of BSI. Primary Spheres were imaged and counted 14 days post treatment. Primary spheres were dissociated into individual cells and then reseeded into ultra-low attachment plates for secondary spheres formation. Again cells were exposed to 25 μM of BSI. Secondary Spheres were imaged and counted 14 days post treatment. Scale bar 500 μm. Magnification of the images 100X. Bar graph show Mean ± SD of 6–12 repeats.
Fig 9.
Structure, cytotoxicity and effects of BSI analogues on cell migration.
A. Structure of the 6 BSI analogues, and sensitivity of pancreatic cancer cells (PANC-1 and BxPC-3) cells to BSI analogues. Cells were exposed to different concentrations of BSI analogues for 48 hrs. Cell viability detected at 48hrs post treatment by MTT assay. B. Scratch was made on confluent monolayer of BxPC-3 cells using 1.25 ml sterile pipette tip. After washing with media, cells were exposed to 25 μM BSI analogues. Scratch was photographed at 0 and 24 hrs post treatment. C. Bar graph shows the % distance covered by BxPC-3 cells of 3 repeats.
Fig 10.
Inhibition of pancreato-spheres formation by BSI analogues.
BxPC-3 and PANC-1 cells were seeded in ultra-low attachment plates and then cells were exposed to 25 μM of BSI analogues. BxPC-3 (A) and PANC-1 (F) primary spheres were imaged and counted 14 days post treatment. Bar graph representing the average number of primary spheres of BxPC-3 (B) and PANC-1 (G), feret diameter of the primary spheres of BxPC-3 (C) and PANC-1 (H) spheres± SEM. Primary spheres were dissociated into individual cells using trypsin and then reseeded into ultra-low attachment plates for secondary spheres. Again cells were exposed to 25 μM of BSI analogues. BxPC-3 (A) and PANC-1 (F) Secondary Spheres were imaged and counted 14 days post treatment. Bar graphs representing the average number of secondary spheres ± SD of BxPC-3 (D) and PANC-1 (I) cells, feret diameter of BxPC-3 (E) and PANC-1 (J) spheres. Scale bar 500 μm. Magnification of the images 100X.