Fig 1.
Cellular system for screening of compounds inhibiting STAT3 transcriptional activity.
(A) Overview of the STAT3 luciferase reporter system. (B) A4wt and A4 sublines were transiently transfected with pGL4.27-SIE reporter and pRL-Renilla and either left untreated or were treated with IL6 + IL6R (50ng/mL and 100 ng/mL respectively) or IFNγ (40 IU/mL). After 6h, luciferase activities were measured using Dual-Glo Luciferase assay. The data presented as ratios to the corresponding untreated control. (C) A4 and A4wt sublines were treated with IL6 + sIL6R (50ng/mL and 100 ng/mL, respectively) or IFNγ (40 IU/mL) for 30 min. Western blotting analysis was performed with the indicated antibodies. GAPDH was used as a loading control. (D) A4wt cells were transfected with pGL4.27-SIE reporter and a stable sub-line A4wt-SIE-6 was selected. Cells were either left untreated (control) or were stimulated with IL6 + sIL6R (50 and 100 ng/mL, respectively) either before treatment with STATTIC (10 μM) or Pyr6 (1 μM), or after, as indicated. Luciferase activity was measured 6h after the first treatment in both cases using SteadyLiteTM Plus. The data is presented as fold induction. Error bars represent SE from quadruplicates. (E) A4wt cells stably transfected with pGL4.27-SIE reporter were seeded in 384-well plates and pre-treated with IL6 and sIL6R (50 and 100 ng/mL, respectively) for 1h, whereafter treated with STATTIC (10 μM) or Pyr6 (1 μM) for 5h to estimate the suitability of the screening system for the high-throughput screening and calculate the z’-factor. Luciferase activity was measured as in (C). One column of the 384-well plate was used for each treatment. The data is presented as average of raw luciferase activity ± standard deviation. The relevant z’-values are described in the text. (F) Outline of the luciferase reporter assay used for high-throughput screening. (G) The dot-blot of the results of the primary screening. A4wt-SIE-6 cells stably transfected with pGL4.27-SIE were pre-treated with IL6 and sIL6R and then treated with the drugs. The data are presented as percent inhibition of luciferase activity where the luciferase activity of the cells treated with IL6+sIL6R+mock is regarded as 0% inhibition (red dots), and the activity in cells treated with IL6+sIL6R+STATTIC regarded as 100% inhibition (green dots). The dashed line is the distance of 3 standard deviations from IL6+sIL6R+mock-treated cells. All the library compounds were used at 10 μM. (H) The STAT3 inhibitor screening funnel.
Fig 2.
Characterization of the four hit compounds.
(A-D) Compounds KI1, KI4, KI12 and KI16 and their corresponding activities in the luciferase reporter assays (center) and cell viability assays (right). Luciferase assays were performed in A4wt cells stimulated with IL6 (grey) and in A4 cells stimulated with IFNγ (red). Viability assays were conducted over 48 hours in A4wt cells (grey) and A4 cells (red). (E) Luciferase IC50 values for lead compounds. Curves were fit using GraphPad and error bars indicate the 95% confidence intervals. (F) Cell viability EC50 values in A4 and A4wt cells using CellTiterGlo assay (Promega Biotech AB, Sweden). Curves were fit using GraphPad and error bars indicate the 95% confidence intervals.
Fig 3.
Docking assays and chemical features of the lead compounds.
(A) Chemical structures of KI1 and KI4. The quinazolone core of KI1 and KI4 is also a common feature of clinically approved EGFR- and VEGFR-targeted therapeutics Gefitinib, Erlotinib and Vandetanib. The pharmacophore portion of the clinically used compounds is highlighted in blue, which is highly similar to the structures of KI1 and KI4. (B) STAT3 crystal structure (blue, pdb 1BG1) with residues 702–709 of the opposing monomer shown in cyan. (C) (Upper panel) KI12 docked into the STAT3 SH2 domain using GLIDE docking software. (Lower panel) Ligand interaction diagram showing interactions between KI12 and residues on the STAT3 SH2 domain. KI12 hydrogen bonds with Ser611, Ser613 and Glu612 and forms cation-pi type interactions with the positively charged Lys591. An additional hydrogen bond is formed with Gln635 and a salt bridge with Lys626. (D) (Upper panel) KI16 docked into the STAT3 SH2 domain using GLIDE docking software. (Lower panel) Ligand interaction diagram showing interactions between KI16 and residues on the STAT3 SH2 domain. KI16 makes hydrogen bonds with Ser611, Ser613 and Gln635.
Fig 4.
The effect of the lead compounds on the IL6-JAK-STAT pathway.
(A) A4wt or A4 cell lines were pre-treated with the indicated compounds (5 and 10μM) for 30 min followed by stimulation with IL6+sIL6R (50 and 100 ng/mL, respectively) for 30 min. Levels of phosphorylated STAT3 (pTyr705) and STAT1 (pTyr701) and total levels of these proteins were determined by Western blotting analysis. GAPDH was used as a loading control. Both left and right panels of western blots were performed simultaneously. (B) A4wt and A4 cell lines were pre-treated with the indicated compounds (5 and 10μM) for 30 min followed by stimulation with IFNγ (40 IU/mL) for 30 min. Levels of phospho-STAT1 and total STAT1 were determined by Western blotting analysis. GAPDH was used as a loading control. (C) A4wt cells were pre-treated with the compounds KI1, KI4, KI12 and KI16 in the concentration of 10 μM and then treated with IL6 + sIL6R (50 ng/mL and 100 ng/mL respectively) for 30 min. The cell pellets were lysed and pJAK1, pJAK2 and total JAK1 and JAK2 levels were assessed by Western blotting. GAPDH was used as a loading control. (D) A4wt cells were stimulated with either IL6+sIL6R (50 and 100 ng/mL, respectively) or IFNγ (40 IU/mL) for 4 h. Expression of indicated STAT3 target genes was assessed by qRT-PCR. The data represents mean of duplicates ± SD. (E) A4wt cells were pretreated with the indicated compounds (5 and 10 μM) for 30 min followed by IL6+sIL6R stimulation for 4 h (as in D). The expression of indicated STAT3 target genes was determined by qRT-PCR. The data is normalized to β-actin expression and presented as fold expression relative to the untreated control. The data represents mean of duplicates ± SD.
Fig 5.
Differential effect of the lead compounds on viability of STAT3-dependent and STAT3-independent cancer cell lines.
(A) Cell pellets from the prostate cancer cells lines PC3 and DU145, and the breast cancer cell lines MCF7 and MDA-MB468, were lysed and subjected to Western blotting analysis. Antibodies against pSerSTAT3, pTyrSTAT3, pTyrSTAT1 and the corresponding antibodies for the total levels of these proteins were used. GAPDH was used as a loading control. (B) PC3, DU145, MCF7 and MDA-MB468 cells were seeded in 384-well white plates using Multidrop. Cells were treated with the range of compounds concentration using Echo liquid dispenser. Cell viability was assessed after 48h using CellTiterGlo viability assay (Promega) according to the instructions of the manufacturer. The viability of the DMSO-treated cells was set as 100%. The data represents IC50 values for the four cell lines calculated from three replicates ± SD. (C, D) DU145 (C) and MDA-MB468 (D) cells were treated with the indicated compounds for 4 h. Levels of phospho-STAT3 and total STAT3 were determined by Western blotting analysis. GAPDH was used as a loading control. (E) DU145 cell line was transfected with 50 nM scrambled siRNA or STAT3 siRNA for 48 h or treated with the indicated compounds for 16 h and 24h, respectively. The concentrations of the compounds were taken from the IC50 values for viability (3B). Expression of the indicated STAT3 target genes was analysed by qRT-PCR. The values are normalized to the expression of GUSB and presented as a relative expression to untreated controls. Error bars represent mean + SE from three independent experiments. (F) A wound healing assay was performed using DU145 cells treated with the indicated compounds at 20 μM. Migration into the denuded area was monitored after 4, 8 and 24 h. The data represents mean of triplicates ± SE from three independent experiments. * p < 0.05.
Table 1.
Primers used for qRT-PCR.