The association of TP53 mutations with the resistance of colorectal carcinoma to the insulin-like growth factor-1 receptor inhibitor picropodophyllin

Human colorectal carcinoma cell lines CACAO-2, COLO-205, COLO-320, DLD-1, HCT-8, HT29 and SW948 were purchased from American Type Collection (ATCC; Rockville, MD). Each cell line was grown in RPMI1640 medium (Invitrogen, Carlsbad, CA) supplemented with 10% fetal bovine serum (FBS). Cells were maintained in a humidified 37°C and 5% CO₂ incubator. Human colorectal carcinoma and matched adjacent normal colorectal tissue samples were collected in accordance with the protocols approved by the institutional Review Board of the First Hospital of Jilin University. All patients provided written informed consent for the tissue sample collection. This study was approved by the First Hospital Ethical Committee of Jilin University.

PPP were purchased from Calbiochem (EMD Millipore) and dissolved in dimethyl sulfoxide (DSMO) at the concentration of 10 mM and stored in aliquots at −80°C. Recombinant human IGF-I was also purchased from Calbiochem and stored in aliquots at −80°C. The antibodies used in this study were purchased from Cell Signaling Technology (Beverly, MA) against the human caspase-9, phospho-IRS-1, AKT, phospho-AKT (Ser473), ERK, phopho-ERK (Thr202/Thr204), IGF-1R, phospho-IGF-1R (Y1135/1136), BAD and phospho-BAD (Ser112/Ser136). Other primary antibodies used in the study included those against the human poly (ADP-ribose) polymerase (PARP), caspase-3 (StressGen, Ann Harbor, MI), DNF fragmentation factor-45 (DFF45), β-actin, BCL-2 (Santa Cruz Biotechnology, Santa Cruz, CA), MDM2 (sigma Aldrich) and X-linked inhibitor of apoptosis protein (XIAP; Transduction Laboratories, Lexington, KY). The secondary antibodies used in this study were horseradish peroxidase (HRP)-conjugated goat anti-mouse (Southern Biotech, Birmingham, AL) and goat anti-rabbit antibody (Jackson ImmunoResearch Laboratories, West Grove, PA). Protease inhibitor mixture, Triton x-100 and other chemicals were purchased from Sigma-Aldrich. Chemiluminescence was from Amersham Biosciences (Piscataway, NJ).

Cells were grown in 96-well plates at 8x10³ cells per well in 100 μl of growth medium. Cells were treated or untreated with PPP in the concentrations as indicated in the Results. After incubation for the times indicated in the Results, cells were washed with a phosphate buffer and 100 μl buffer 0.2 M containing sodium acetate (pH 5.5), 0.1% (v/v) Triton X-100 and 20 mM p-nitrophenyl phosphate was added to each of the wells. The plates were incubated at 37°C for 1.5 hours and the reaction was stopped by the addition of 10 μl 1 M NaOH to each well, Absorbance were measured at 405 nm by a microplate reader (BioRad).

Cells were treated with 1 μM PP242 and 2 μM erlotinib, alone or in combination, for 20 hours, harvested, fixed with 70% ethanol, and stained with propidium iodide. The data were acquired using flow cytometry (FACSCanto II Becton Dickinson, Franklin Lakes, NY) and were analyzed using FlowJo software (Tree Star Inc. Ashland, OR). Sub-G1 apoptotic cells were determined as a percentage of the cells.

Western blotting was performed according to our laboratory protocols [32]. In brief, cells were lysed in a cell lysis buffer (20 nM Tris pH7.4, 150 mM NaCL, 1% NP-40, 10% glycerol,1 mM EGTA, 1 mM EDTA, 5 mM sodium pyrophosphate, 50 mM sodium fluoride, 10 mM β-glycerophosphate, 1 mM sodium vanadate, 0.5 mM DTT, 1 mM PMSF, 2 mM imidazole, 1.15 mM sodium molybdate, 4 mM sodium tartrate dihydrate, and 1x protease inhibitor cocktail). Cell lysates were cleared by centrifugation at 18,000 x g for 15 minutes at 4°C. The supernatant was collected and protein concentrations were determined by the Bradford protein assay following the manufacturer^’s protocol (Bio-Rad Laboratories). Equal amounts of protein were separated through SDS-PAGE gels and transferred onto nitrocellulose membranes (Bio-Rad Laboratories). The membranes were incubated overnight at 4°C with primary antibody and then for 1 hour with HP-conjugated secondary antibody. The membranes were developed by chemiluminescence.

The animal studies were approved by the Institutional Animal Care and Use Committee of Emory University. The HCT-8 cells or Caco2 cells (7 × 10⁶) were implanted subcutaneously into the flank regions of six-week old (about 20 g of body weight) female athymic (nu/nu) mice (Taconic, Hudson, NY). The mice were allowed to develop subcutaneous xenografts and tumor volumes were measured using caliper measurements. When tumors reached approximately 150–200 mm³, mice were assigned randomly to 2 experimental groups (n = 4 per group) and treated either with saline as control or PPP (50 mg/kg) through oral gavages, twice per week. Tumor volumes were measured once every 3 days and calculated based on the formula: V =4/3 × π × (length/2 × [width/2]²). At the end of treatment, the mice were sacrificed and the tumors were harvested and weighed at necropsy.

Earlier studies have revealed increased levels of the IGF-1R mRNA in human colorectal carcinoma tumors [14, 15]. To examine the expression of IGF-1R protein, we carried out a western blot analysis of human colorectal carcinoma tumors, together with matched normal colorectal tissue. The results showed that IGF-1R proteins were expressed in the carcinoma tumors at much higher levels than in the matched normal tissue (Figure 1A). We then examined a panel of seven colorectal carcinoma cell lines by western blotting and identified the expression of IGF-1R in each of these cell lines. Nearly half of the cell lines expressed much higher levels of IGF-1R as compared with other cell lines (Figure 1B).

Next, we examined how colorectal carcinoma cell lines respond to PPP treatment. To this end, each of the cell lines was treated with a series of PPP concentrations for 72 hours. A cell viability assay showed PPP treatment significantly inhibited the growth of the sensitive cell lines HCT-8 and SW948. Slight inhibition of the growth of the resistant cell lines CACO-2, COLO-205, COLO-320, DLD-1 and HT-29 was found at much higher doses (Figure 1C). The PPP resistant cell lines were reported with TP53 mutations [33] according to the Catalogue of Somatic Mutations in Cancer (http://​cancer.​sanger.​ac.​uk/​cancergenome/​projects/​cosmic). In contrast, HCT-8 [34] and SW948 (http://​cancer.​sanger.​ac.​uk/​cancergenome/​projects/​cosmic) are TP53 wild-type cell lines. These analyses suggest the association of TP53 mutations with the PPP resistance of colorectal carcinoma cells to PPP treatment.

To examine the mechanism of PPP resistance, we evaluated whether PPP treatment blocks IGF-1R auto-phosphorylation [26] and inhibits the downstream AKT and ERK pathways [25]. Since IGF-I and IGF-II activate IGF-1R through paracrine and autocrine loops [6], each of the cell lines was therefore treated with 50 ng IGF-I. Western blotting showed that the IGF-I treatment resulted in the phosphorylation of IGF-1R in both the TP53 wild-type HCT8 and mutated CACO-2 cells (Figure 2A). The cell lines were then treated with 500 nM PPP in the presence of IGF-I and western blotting revealed a decrease in phosphorylation of IGF-1R in a time dependent manner. In contrast, total IGF-1R levels remained unchanged during the treatment. These data indicate that PPP blocks IGF-1R phosphorylation in both TP53 wild-type and mutated cells. PPP treatment reduced the levels of phosphorylated AKT and ERK in the TP53 wild-type HCT-8 but not the TP53 mutated CACO-2 cells; the results suggest that the PPP resistance occurs at IGF-1R downstream intracellular levels in TP53 mutated cells.

Earlier studies have clearly shown that PPP treatment leads to the downregulation of IGF-1R through MDM2-mediated ubiquitination and degradation of the IGF-1R protein [35]. Both IGF-1R and p53 proteins are the substrates of the ubiquitin ligase MDM2 [36]. To explore the role of MDM2 in the resistance of mutated TP53 cell lines to PPP, we examined the protein levels of MDM2 in wild-type and mutated TP53 cell lines by western blotting. The data revealed no difference in the expression of MDM2 protein between TP53 wild-type and mutated cell lines (Figure 2B). Next, we examined the kinetics of IGF-1R degradation under the treatment of IGF-1 and PPP, alone and in combination. To this end, we compared the IGF-1R protein levels between the TP53 wild-type SW948 and mutated CACO-2 since these two cell lines expressed IGF-1R protein at similar levels (Figure 1B). Western blotting revealed that PPP treatment reduced the levels of IGF-1R protein in both SW948 and CACO-2 cells (Figure 2C) due to the similar expression levels of MDM2 protein between these two cell lines (Figure 2B). These results confirm the earlier reports [35, 36] that PPP treatment induces IGF-1R degradation through MDM2-medicated ubiquitination in a p53-independent manner.

MDM2-mediated ubiquitination of IGF-1R with PPP treatment leads to the activation of ERK pathway [37], resulting in the resistance of Ewing’s sarcoma to the treatment of the anti-IGF-1R antibody figitumuab [38]. To explore this mechanism in colorectal carcinoma, we treated SW948 and CACO-2 cell lines with PPP in a dose-dependent manner and found that PPP treatment increased the levels of p-ERK in the TP53 mutated CACO-2 but not in the TP53 wild-type SW948 cells (Figure 2D). Taken together, the results suggest that PPP treatment bocks the phosphorylation of IGF-1R and inhibits the downstream ERK pathway in TP53 wild type colorectal carcinoma cells. In contrast, TP53 mutated carcinoma cells are resistant to the PPP treatment in part due to its failure of inhibition of the intracellular ERK pathway.

Earlier studies have shown that PPP treatment inhibits cell growth and induces apoptosis in different types of cancer cells [25, 27]. To examine this in colorectal carcinoma cells, we analyzed PPP-treated cells by flow cytometry. The results showed that PPP treatment led to a significant increase of sub-G1 apoptotic cells in the TP53 wild-type but not mutated cell lines (Figure 3A,B). The results further suggest that TP53 mutated carcinoma cells are resistant to PPP treatment in part due to its failure of induction of apoptosis in these cells.

IGF-1R activation leads to the inhibition of apoptosis through the AKT/ERK-mediated phosphorylation of BAD [3]. The failure in AKT/ERK activation and apoptosis induction in TP53 mutated cells under PPP treatment suggests the possibility that BAD phosphorylation may play a role in the PPP resistance. To test this notion, we treated the TP53 wild type HCT-8 and mutated CACO-2 cells with 500 nM PPP. Lysates from the treated cells were examined by western blot analysis using antibodies against the phosphorylated and unphosphorylated form of BAD. The results showed that the PPP treatment inhibited BAD phosphorylation in TP53 wild-type but not mutated cells (Figure 4A).

Unphosphorylated BAD interacts with the BCL2 family of proteins and releases their inhibition of the mitochondrial membrane potential [4], leading to the mitochondrial release of apoptosis factors and resulting in caspase-9 activation and initiation of apoptosis through cleavage of the downstream effectors caspase-3, DFF45, and PARP [39]. In addition, the second mitochondria-derived activator of caspase/direct inhibitor of apoptosis binding protein with low pI (SMAC/DIABLO) interacts with THE X-linked inhibitor of apoptosis protein (XIAP), which releases XIAP from binding to caspase-3 and allows caspase-9 cleavage of caspase-3 [40, 41]. To examine this mitochondrial pathway in PPP-induced apoptosis, we showed that the treatment of PPP led to the cleavage of XIAP (Figure 4A) and caspase-9, caspase-3, PARP, and DFF45 in the TP53 wild-type HCT-8 but not the mutated CACO-2 cells (Figure 4B). Collectively, the PPP resistance is in part due to the inhibition of BAD-mediated mitochondrial apoptosis in TP53 mutated colorectal carcinoma cells.

To examine the potential of PPP in treatment of colorectal carcinoma, we first injected the TP53 wild-type HCT-8 cells subcutaneously in athymic (nu/nu) mice for the generation of subcutaneous flank xenografts. The mice were closely monitored and once xenografts reached approximate size of 150–200 mm³, the mice were divided into two groups. In the treatment group, mice were treated with PPP (50 mg/kg) and in the control group, mice were treated with saline. The mice were treated through oral gavage, twice per week for three weeks. Tumor volumes were measured and the results showed that PPP treatment significantly inhibited the growth of the TP53 wild-type HCT-8 colorectal carcinoma xenografts (Figure 5A). At necropsy, a significant difference in the tumor sizes was observed between the control and treatment mice (Figure 5B). The xenografts were removed and tumor lysates were subjected to western blot analysis. The results showed that PPP treatment inhibited the phosphorylation of IGF-1R, AKT, and ERK in the TP53 wild-type HCT-8 colorectal carcinoma xenografts (Figure 5C).

To examine whether the TP53 mutated colorectal carcinoma xenografts resist the treatment of PPP, we injected CACO-2 cells subcutaneously in athymic (nu/nu) mice. Once subcutaneous xenografts were formed approximately 150–200 mm³ in size, the mice were treated either with PPP (50 mg/kg) or saline through oral gavage, twice per week for three weeks. The results showed no significant difference in the tumor sizes between the treatment and control group of mice, as indicated by the measured tumor volumes (Figure 6A) and the tumor sizes as observed at necropsy (Figure 6B). Western blot analysis of the representative xenograft tissues showed that PPP treatment failed to inhibit the phosphorylation of IGF-1R, AKT and ERK in the TP53 mutated CACO-2 colorectal carcinoma xenografts (Figure 6C). Taken together, these studies suggest that TP53 wild type colorectal carcinoma may respond to the treatment of PPP whereas TP53 mutated carcinomas are most likely resistant to the treatment.