Novel β-carbolines against colorectal cancer cell growth via inhibition of Wnt/β-catenin signaling

Abstract

Wnt signaling pathway is aberrantly activated in a variety of cancers, especially in colorectal cancer (CRC), because of mutations in the genes encoding adenomatous polyposis coli (APC), β-catenin and Axin. Small-molecule antagonists of Wnt/β-catenin signaling are attractive candidates for developing effective therapeutics for CRC. In this study, we have identified a novel Wnt signaling inhibitor, isopropyl 9-ethyl-1- (naphthalen-1-yl)-9H-pyrido[3,4-b]indole-3- carboxylate (Z86). Z86 inhibited Wnt reporter activities and the expression of endogenous Wnt signaling target genes in mammalian cells and antagonized the second axis formation of Xenopus embryos induced by Wnt8. We showed that Z86 treatment inhibits GSK3β (Ser9) phosphorylation, leading to its overactivation and promoting the phosphorylation and degradation of β-catenin. In vitro, Z86 selectively inhibited the growth of CRC cells with constitutive Wnt signaling and caused obvious G1-phase arrest of the cell cycle. Notably, in a nude mouse model, Z86 inhibited dramatically the xenografted tumor growth of CRC. Daily intraperitoneal injection of Z86 at 5 mg/kg resulted in >70% reduction in the tumor weight of HCT116 cell origin that was associated with decreased GSK3β (Ser9) phosphorylation and increased β-catenin phosphorylation. Taken together, our findings provide a novel promising chemotype for CRC therapeutics development targeting the canonical Wnt signaling.

Figure 1

Identification of Z86 as a potent antagonist of Wnt signaling. (a) Structures of the β-carboline compounds. (b) Effects of Z64, Z80 and Z86 on the Topflash reporter activity. The HEK293W cells were treated with different doses of Z64, Z80 and Z86 respectively for 24 h. Relative luciferase activity (Topflash/Renilla) measured represents the level of activated Wnt signaling. (c) The calculated IC₅₀ values of Wnt signaling inhibition of the compounds are displayed in the table. (d) Z86 preferentially inhibits Wnt signaling (ST-Luc) over the NF-κB signaling pathway. NF-κB signaling was stimulated with 25 ng/ml TNFα treatment for 24 h in HEK293T cells that were subsequently treated with Z86 (20 μM) for 24 h and the NF-κB-Luc luciferase activity was measured. Each bar is the mean±S.D. from three independent experiments. NS, not significant, relative to the vehicle control.

Figure 2

Z86 inhibits Wnt signaling in colon cancer cells and the secondary axis formation in Xenopus embryos. (a) Z86 inhibits Topflash reporter activity in a dose-dependent manner in Wnt1-transfected HEK293T cells and CRC cells, SW480 and HCT116. Cells were transfected with luciferase reporters (Wnt1 was co-transfected into HEK293T cells) for 3 h, and subsequently were incubated with Z86 with indicated dosages for additional 24 h. Luciferase activity was then measured. (b) Z86 inhibits Axin2 and cyclin D1 expression in Wnt1-transfected HEK293T cells and CRC cells. The HEK293T cells were transfected with Wnt1 3 h before Z86 treatment. The cells were treated with indicated dosages of Z86 for 12 h. The levels of Axin2 and cyclin D1 mRNAs were determined by real-time PCR and normalized to β-actin. (c) Effects of Z86 on the protein expression of Axin2, Cyclin D1 and Survivin in Wnt1-transfected HEK293T cells and CRC cells. The HEK293T cells were transfected with Wnt1 3 h before Z86 treatment. The cells were treated with Z86 (20 μM) for different time periods and cell extracts were harvested and assayed for the expressions of indicated proteins by western blotting. β-Actin was used as the loading control. (d) Z86 inhibits the secondary axis (white arrowheads) induced in Xenopus embryos by ventrally injected Wnt8 mRNA in a dose-dependent manner. The Wnt8-injected embryos were treated with DMSO or Z86 and were examined for axis duplication (white arrowhead in the upper panel) at late neurula stage (stage 18). Representative embryos are shown in the upper panel (dorsal views with anterior to the bottom). The statistic data are shown in the lower panel and the numbers of embryos in each group are shown on the top of each column. Scale bars=0.5 mm.

Figure 3

Z86 promotes the phosphorylation and degradation of β-catenin. (a) Z86 promotes phosphorylation and the subsequent degradation of β-catenin in Wnt3a stably transfected HEK293W cells, as well as in CRC cells (SW480 and HCT116). Cells were treated with Z86 for different times indicated. The cell extracts were harvested and assayed by western blotting for the expressions of phosphorylated β-catenin (Ser33, Ser37 and Thr41) and the total β-catenin. β-Actin was used as the loading control. The blots show representative data derived from three experiments. (b) Cellular distribution and amount of β-catenin in SW480 cells. After incubation with Z86 for 12 h, cells were subjected to immunofluorescent staining for β-catenin. The microscopy images are representative examples from one of three independent experiments. β-Catenin (green) was observed present in the cytosol and nucleus fractions, whereas nucleic acid dyed with 6'-diamidino-2-phenylindole (DAPI; blue) in the nucleus. Scale bars=20 μm. (c) The Wnt signaling (ST-Luc) activity in HCT116 cells, transfected with increasing amount of mutant β-catenin (S37A), with or without Z86 treatment, was measured using dual-luciferase reporter assay.

Figure 4

Z86 inhibits Wnt signaling through suppression of GSK3β phosphorylation. (a) Z86 reduces the phosphorylation of GSK3β (Ser9) in both Wnt3a stably transfected HEK293W cells and CRC cells. Cells were incubated with 20 μM of Z86, and cell extracts were subjected to western blot analysis to detect phosphorylated GSK3β (Ser9) and the total GSK3β. β-Actin was used as the loading control. (b) HEK293T cells were transfected with luciferase reporters (Topflash and Renilla) together with Wnt1 (64 ng per well), DNGSK3β (30 ng per well) in 96-well plates, or incubated 20 mM of LiCl respectively. After transfection or treatment for 3 h, cells were treated with Z86 at indicated dosage for additional 24 h and luciferase activities were then measured. (c, upper panel) Schematic drawing showing the procedure of Xenopus animal cap RT-PCR assays. (c, lower panel) Z86 inhibits Wnt8, but not DNGSK3β-induced expression of Siamois and Xnr3 in Xenopus animal cap cells. The experiment was repeated three times and representative results are shown. Amplification of ornithine decarboxylase (ODC) confirmed RNA integrity. NC, negative control with the reverse transcriptase omitted in the RT reaction; Con, control.

Figure 5

Z86 inhibits the growth of colon cancer cells with G1-phase arrest. (a and b) Growth inhibition of Z86 on various cell lines, including CRC cell lines (SW480, HCT116 and HT29), normal colonic epithelial cell line CCD-841-CoN and human normal bronchus epithelial cell line Beas-2B, determined by MTS assays with cisplatin used as a control. Cells were treated with Z86 up to 72 h at various concentrations. All samples were done in triplicates and the median growth inhibition concentration (GI₅₀) values of Z86 and cisplatin toward the cell lines are presented as mean±S.D. (c and d) Representative histograms depicting cell cycle distribution as analyzed by flow cytometry in HCT116 cells treated with 20 μM of Z86 for 12, 24 and 48 h respectively. Cell counts of different phases were quantified.

Figure 6

Z86 inhibits tumor growth of HCT116 xenografts through suppression of GSK3β phosphorylation in vivo. Mice were injected with 5×10⁶ of HCT116 cells. One week later, tumor-bearing mice were treated intraperitoneally with either vehicle or Z86 (1, 2 and 5 mg/kg respectively) daily for 23 days. (a and b) Tumor growth and body weight were monitored during the administration. (c) Tumors were removed from mice and tumor weight was measured on day 23. Data represent the mean±S.D., *P<0.05, **P<0.01, relative to vehicle control. (d) The expression of β-catenin and phosphorylated β-catenin (Ser33, Ser37 and Thr41) as well as GSK3β and phosphorylated GSK3β (Ser9) in tumor tissues was determined by western blot analysis.