Inhibition of CREB binding protein-beta-catenin signaling down regulates CD133 expression and activates PP2A-PTEN signaling in tumor initiating liver cancer cells

Abstract

Background: The WNT-beta-catenin pathway is known to regulate cellular homeostasis during development and tissue regeneration. Activation of WNT signaling increases the stability of cytoplasmic beta-catenin and enhances its nuclear translocation. Nuclear beta-catenin function is regulated by transcriptional co-factors such as CREB binding protein (CBP) and p300. Hyper-activated WNT-beta-catenin signaling is associated with many cancers. However, its role in inducing stemness to liver cancer cells, its autoregulation and how it regulates tumor suppressor pathways are not well understood. Here we have investigated the role of CBP-beta-catenin signaling on the expression of CD133, a known stem cell antigen and PP2A-PTEN pathway in tumor initiating liver cancer cells.

Methods: Human hepatoblastoma cell line HepG2 and clonally expanded CD133 expressing tumor initiating liver cells (TICs) from premalignant murine liver were used in this study. CBP-beta-catenin inhibitor ICG001 was used to target CBP-beta catenin signaling in liver cancer cells in vitro. Western blotting and real time PCR (qPCR) were used to quantify protein expression/phosphorylation and mRNA levels, respectively. CBP and CD133 gene silencing was performed by siRNA transfection. Fluorescence Activated Cell Sorting (FACS) was performed to quantify CD133 positive cells. Protein Phosphatase (PP2A) activity was measured after PP2AC immunoprecipitation.

Results: CBP inhibitor ICG001 and CBP silencing significantly reduced CD133 expression and anchorage independent growth in HepG2 and murine TICs. CD133 silencing in TICs decreased cell proliferation and expression levels of cell cycle regulatory genes, CyclinD1 and CyclinA2. ICG001 treatment and CBP silencing reduced the levels of phospho^{Ser380/Tyr382/383}PTEN, phospho^Ser473-AKT, Phospho-^Ser552beta-catenin in TICs. ICG001 mediated de-phosphorylation of PTEN in TICs was PP2A dependent and partly prevented by co-treatment with PP2A inhibitor okadaic acid.

Conclusions: CBP-beta-catenin signaling promotes stemness via CD133 induction and cell proliferation in TICs. We found a novel functional link between CBP-beta-catenin and PP2A-PTEN-AKT pathway in liver TICs. Therefore, CBP-beta-catenin-PP2A-PTEN-AKT signaling axis could be a novel therapeutic target to prevent liver tumor initiation and cancer recurrence.

Keywords: CBP; CD133; Cancer stem cells; PP2A; PTEN; Tumor initiating cells; WNT-beta catenin signaling.

Fig. 1

CBP-beta-catenin signaling regulates CD133 expression and anchorage independent growth in murine TICs: a Expression levels of CD133 mRNA in CD133 positive clonally expanded TICs from Mat1a−/− murine liver and liver cancer cell line developed from Mat1a−/− mouse liver tumors. Data represents mean ±S.E from three independent experiments (n = 3, #p < 0.0001). b Effect of time course ICG001 treatment on CD133 expression in TICs. Results represent mean ±S.E from four independent experiments (n = 4, *p < 0.005, # p < 0.0005) c Western blot to show the effect of ICG001 on CD133 protein levels. Protein band intensity was calculated by densitometry quantification by Image J Software, NIH. Data is represented as fold difference compared to DMSO treated control cells. Data represents mean ±S.E from three independent experiments (n = 3, #p < 0.005). d FACS analysis to show the effect of ICG001 on CD133 positive cells in TICs. Data represented as % of total live cells in each group. Live cells were gated with DAPI staining. Results represent mean ±S.E from three independent experiments (n = 3, # p < 0.001). e Effect of ICG001 on anchorage independent growth in TICs. After 14 days of culture on soft agar with and without ICG001, colonies were stained with 0.005% crystal violet and images were captured. For each experiment 4–5 images were captured from six different experiments and colonies were counted manually and average number of colonies were shown in the graph. n = 6, # p <  0.005

Fig. 2

CBP-beta-catenin signaling regulates CD133 expression, cell survival proliferative genes and anchorage independent growth in HepG2 cells. a CD133 mRNA expression levels in human hepatoblastoma cell line HepG2 and cholangiocarcinoma cell line MzCha-1. Expression level was normalized to MzCha-1 cell line and fold expression was calculated, (n = 3 #p < 0.005). b Effect of time course ICG001 treatment on CD133 mRNA expression (n = 3 *p < 0.001). c Western blot to show the effect of ICG001 on CD133 protein levels. Protein band intensity was calculated by densitometry quantification by Image J Software, NIH. Results represent mean ±S.E from three independent experiments (n = 3, # p < 0.01). d FACS analysis to show the effect of ICG001 on CD133 positive cells in HepG2 cells. Data represented as % of total live cells in each group. Live cells were gated with DAPI staining. Results represent mean ±S.E from three independent experiments (n = 3, # p < 0.001). e Effect of ICG001 on CYCLIND1 and SURVIVIN expression levels in HepG2 cells. Results represent mean ±S.E from three independent experiments #p < 0.001, *p < 0.05. f Effect of ICG001 on anchorage independent growth in HepG2 cells. After 14 days of culture on soft agar with and without ICG001, colonies were stained with 0.005% crystal violet and images were captured. For each experiment 4–5 images were captured from six different experiments and colonies were counted manually and average number of colonies were shown in the graph, n = 6, # p < 0.005

Fig. 3

CD133 expression is associated with cell proliferation in TICs a siRNA mediated gene silencing of CD133 for 72 h in TICs. Results represent mean ±S.E from three independent experiments (n = 3 # p < 0.005). b Effect of CD133 silencing on cell proliferation as determined by BrdU incorporation assay (n = 6 *p < 0.05). c Effect of CD133 silencing on the expression level of cell cycle regulatory genes, CyclinD1 and CyclinA2. Expression level was normalized to scrambled siRNA control (Sc) and represented as relative fold, n = 3 *p < 0.05. d. Immuno-staining for proliferation marker, Ki67 (red) in Scrambled and CD133 silenced TICs. Isotype IgG was used as antibody staining control. Nuclei stained with DAPI (Blue). Scale bar = 25 µm. Images represent four independent experiments

Fig. 4

ICG001 treatment activates PTEN and inactivates AKT and beta-catenin in murine TICs: a Western blot showing PhosphoSer380/Thr382/383-PTEN and total PTEN in ICG001 treated and DMSO control samples. b Western blot of PhosphoSer473-AKT and total AKT in ICG001 treated and DMSO controls. Protein band intensity was calculated by densitometry quantification by Image J Software, NIH. Results represent mean ±S.E from four independent experiments. (n = 4, #p < 0.005). Data was normalized to DMSO controls and represented as relative expression fold. c Western blot analysis for pSer552-beta-catenin in ICG001 treated samples and DMSO controls. Data was normalized to DMSO controls and represented as relative expression fold (n = 4 *p < 0.05). d Immunofluorescence staining for phospho-Ser552-beta-catenin in DMSO and ICG001 treated TICs. Isotype IgG staining was performed as staining control. Number of phospho-Ser552-beta-catenin positive cells per high power field (HPF) was manually counted and compared to DMSO control. Results represent mean ±S.E from four independent experiments (n = 5, *p < 0.01). Scale bar = 25 µm

Fig. 5

CBP silencing downregulates CD133, cell proliferative genes and modulates PTEN-AKT pathways in TICs. a mRNA expression of CBP, CD133 and CyclinD1 in CBP silenced TICs. Results represent mean ±S.E from three independent experiments. Data was normalized to scrambled controls (SC) and represented as relative expression fold. (n = 3 #p < 0.01, *p < 0.05). b Western blot analysis to show changes in protein phosphorylation in CBP silenced cells. c Densitometry quantification of protein band intensity by Image J Software, NIH. Results represent mean ±S.E from three independent experiments. n = 3, *p < 0.05

Fig. 6

ICG001 mediated activation of PTEN is PP2A dependent a Western blot analysis to show the effect of ICG001 and PP2A inhibitor Okadaic acid (OA) on PTEN phosphorylation. Results represent mean ±S.E from four independent experiments. (*p < 0.01, DMSO vs ICG001; #p < 0.05, ICG001 vs ICG001 + OA). b Western blot analysis showing Survivin expression in ICG001−/+ OA treated samples. Results represent mean ±S.E from four independent experiments. *p < 0.05 (DMSO vs ICG001 and ICG001 vs ICG001 + OA). c ICG001 activates PP2A in murine TICs. PP2Ac was immuno-pelleted from cell extracts and PP2A activity was measured using PP2A activity assay kit as described in material and methods. Results represent mean ±S.E from four independent experiments