Chemoprevention of colorectal cancer by targeting APC-deficient cells for apoptosis

Abstract

Cancer chemoprevention uses natural, synthetic, or biological substances to reverse, suppress, or prevent either the initial phase of carcinogenesis or the progression of neoplastic cells to cancer. It holds promise for overcoming problems associated with the treatment of late-stage cancers. However, the broad application of chemoprevention is compromised at present by limited effectiveness and potential toxicity. To overcome these challenges, here we developed a new chemoprevention approach that specifically targets premalignant tumour cells for apoptosis. We show that a deficiency in the adenomatous polyposis coli (APC) gene and subsequent activation of beta-catenin lead to the repression of cellular caspase-8 inhibitor c-FLIP (also known as CFLAR) expression through activation of c-Myc, and that all-trans-retinyl acetate (RAc) independently upregulates tumour necrosis factor-related apoptosis-inducing ligand (TRAIL) death receptors and suppresses decoy receptors. Thus, the combination of TRAIL and RAc induces apoptosis in APC-deficient premalignant cells without affecting normal cells in vitro. In addition, we show that short-term and non-continuous TRAIL and RAc treatment induce apoptosis specifically in intestinal polyps, strongly inhibit tumour growth, and prolong survival in multiple intestinal neoplasms C57BL/6J-Apc(Min)/J (Apc(Min)) mice. With our approach, we further demonstrate that TRAIL and RAc induce significant cell death in human colon polyps, providing a potentially selective approach for colorectal cancer chemoprevention by targeting APC-deficient cells for apoptosis.

Figure 1. TRAIL and RAc induce apoptosis in APC-deficient cells

(a) Primary epithelial cells were isolated from normal mouse intestine and polyps from Apc^Min mice and treated with TRAIL (50 ng/ml for 24 h), RAc (6.8 ng/ml for 48 h), or both (RAc for 48 h then TRAIL for 24 h). Cells were stained with annexin V-FITC and propidium iodide. Early apoptotic cells (annexin V positive, PI negative) were counted. The data represent results from three independent experiments. Average and standard deviation are shown. (b) Immortalized normal human colon epithelial cells (NCM356) were transfected with either nonspecific Ns-siRNA or APC-siRNA, and APC protein was detected 48 h after transfection by Western blotting. (c) The transfected NCM356 cells were treated and harvested, and both full-length pro-caspase 8 and cleaved forms (p44/42 and p18) were detected by Western blotting.

Figure 2. Down regulation of cFLIP by APC/β-catenin-mediated activation of c-Myc and modulation of TRAIL receptor expression by RAc contribute to the activation of TRAIL signaling

(a) NCM356 cells were transfected with either β-catenin or c-Myc and treated with TRAIL, RAc, or both. Cleavage of caspase 8 was detected by Western blotting. (b) NCM356 cells were transfected with Ns-siRNA, APC-siRNA (GFP tagged), or APC-siRNA and c-Myc-siRNA. The transfected cells were treated with RAc and TRAIL, and photographs were taken. Expression of APC, c-Myc, and caspase 8 was detected by Western blotting. (c) NCM356 cells were transfected with c-Myc, cFLIP-siRNA, c-Myc and cFLIP, c-Myc and Bcl-2, or c-Myc and Bcl-XL. The transfected cells were treated with RAc and TRAIL. Expression of c-Myc, cFLIP, Bcl-2, Bcl-XL, and caspase 8 was detected by Western blotting. (d) NCM356 cells were treated with RAc for 48 h, and expression of DR4, DR5, DcR1, and DcR2 was detected by Western blotting. (e) NCM356 cells were transfected with either APC-siRNA or APC-siRNA and DR5 or were pretreated with an anti-DcR1 antibody after APC-siRNA transfection. Nonspecific Ns-siRNA was used as control. Transfected cells were treated with TRAIL alone or with TRAIL and RAc, as indicated. Cleavage of caspase 8 was detected by Western blotting.

Figure 3. Treatment with TRAIL and RAc induces cell death in the polyps, inhibits tumor growth, and promote survival in Apc^Min mice

(a) The Apc^Min mice were treated with either control phosphate buffered saline (PBS) or TRAIL and RAc for 15 treatment cycles. Intestinal polyps were examined 1 month after the treatment and photographs were taken. (b) The polyp samples and adjacent sections were stained by H&E and TUNEL stain. Surrounding normal tissue was used as the control. Representative photographs are shown. (c) Intestinal polyps in the entire intestinal track were examined and counted. Each treatment group contained 5-7 mice. Average and standard deviations are shown. An asterisk indicates P<0.005. (d). The Apc^Min mice were treated with either 1X or 10X doses of TRAIL and RAc for two cycles within a week. The intestinal polyps in the entire intestinal track were examined and counted. Each treatment group contained 5-7 mice. Average and standard deviation are shown. An asterisk indicates P<0.005. (e) Kaplan-Meier survival analysis of Apc^Min mice. Apc^Min mice at 4 month were treated with either TRAIL (3 mg/kg) and RAc (68 μg/kg) or control PBS every 3 weeks for 5 times. The endpoint was reached when mice were either moribund or at day 243. Mean survival time was 212.8 ± 5.9 in TRAIL and RAc group (7 mice) and 186 ± 2.6 in the PBS treated group (8 mice) with log-rank test p≤0.001. All statistical analyses were performed using SPSS software (version 16.0).

Figure 4. Effect of TRAIL and RAc on human colon polyps and ISCs, and synthetic lethal interaction of TRAIL, RAc and APC

(a) Tissue slices from both the normal region and colon polyps from FAP patients were treated with RAc (6.8 ng/ml) for 48 h and then with TRAIL (100 ng/ml) for an additional 24 h (TRAIL+RAc). The samples treated with vehicle were used as controls (C). H&E staining, β-catenin staining, and TUNEL assay (green) were performed. Nuclei were revealed by DAPI staining (blue). Scale bar equals 100 μm. Representative photographs are shown. (b) TUNEL-positive cells (green) were counted against the number of nuclei in multiple fields under a microscope. Average and standard deviations are shown. The data are derived from samples of 4 patients. (c) Apc^Min mice were treated with 5 non-continuous cycles of TRAIL and RAc. Both the normal tissue and polyps of the intestine were serially sectioned and stained with either Lgr5 (brown), or activated caspase 3 (red), or both. Multiple serial sections were analyzed and representative photographs are shown.