Niclosamide's potential direct targets in ovarian cancer†

Abstract

Recent evidence indicates that niclosamide is an anti-cancer compound that is able to inhibit several signaling pathways. Although niclosamide has previously been identified by high-throughput screening platforms as a potential effective compound against several cancer types, no direct binding interactions with distinct biological molecule(s) has been established. The present study identifies key signal transduction mechanisms altered by niclosamide in ovarian cancer. Using affinity purification with a biotin-modified niclosamide derivative and mass spectrometry analysis, several RNA-binding proteins (RBPs) were identified. We chose the two RBPs, FXR1 and IGF2BP2, for further analysis. A significant correlation exists in which high-expression of FXR1 or IGF2BP2 is associated with reduced survival of ovarian cancer patients. Knockdown of FXR1 or IGF2BP2 in ovarian cancer cells resulted in significantly reduced cell viability, adhesion, and migration. Furthermore, FXR1 or IGF2BP2 deficient ovarian cancer cells exhibited reduced response to most doses of niclosamide showing greater cell viability than those with intact RBPs. These results suggest that FXR1 and IGF2BP2 are direct targets of niclosamide and could have critical activities that drive multiple oncogenic pathways in ovarian cancer.

Keywords: FXR1 and IGF2BP2; RNA-binding protein; niclosamide; ovarian cancer.

Figure 1

Biotinylated niclosamide (A) The O-ethylamino niclosamide, 3, was prepared by Mitsunobu coupling of niclosamide with N-Boc-aminoethanol followed by Boc deprotection using trifluoroacetic acid (TFA) followed by Chen et al. [21]. Then, 3 was biotinylated through amide bond formation with either EZ-Link-NHS-Biotin® (SR-248) or EZ-Link-Sulfo-NHS-LC-Biotin® (SR-247) in presence of Et3N in DMF at room temperature. (B) IC50 of niclosamide, SR-247 and SR-248 in OvCa cell lines. (C) Effects of niclosamide, niclosamide-NH2, (3 in Figure 1A), SR-247, and SR-248 on proliferation of OVCAR4.

Figure 2

RNA-binding proteins in OvCa. (A) Genetic alterations of FXR1, IGF2BP2, PIK3CA, PRKCI, ECT2, SOX2, and TP63 in OvCa. OncoPrint was created by cBioPortal for Cancer Genomics using total 591 OvCa patients from TCGA, which are aligned by genetic alteration side-by-side on the X axis. (B) FXR1 and IGF2BP2 are associated with poor outcome in OvCa. Overall survival rate was calculated in 276 patients from GSE9891 by Kaplan–Meier method using Prism 7. The graphs were shown at the lowest P-value/cutoff which was determined by PrognoScan. (C) OVCAR4 cell lysate was incubated with SR-248 or SR-248 + niclosamide (10-fold competitor) for 16 h, and then M-280 Streptavidin Dynabeads were added and incubated for 1 h. Unbound proteins were removed by washing the beads, and the bound proteins were separated by SDS-PAGE gel followed by western blot using anti-FXR1 or IGF2BP2.

Figure 3

Oncogenic activity of FXR1 and IGF2BP2 in OvCa. (A) Generation of FXR1 or IGF2BP2 deficient HGSOC cell lines. The CRISPR/Cas9 system was used to ablate FXR1 or IGF2BP2 in OVCAR3, OVCAR4, and OVCAR8 cells. Two gRNAs were cloned to a lentiviral vector for each gene. After drug selection (puromycin, Cas9) and (blasticidin, gRNA), loss of FXR1 or IGF2BP2 was verified by western blot. (B) Cell proliferation: Cells were seeded in 24-well plates and grown for the indicated culture time. Viable cells were counted using a Countess II FL Automated Cell Counter with trypan blue exclusion. (n = 4 per time). (C) Sensitivity to niclosamide in FXR1 or IGF2BP2 deficient OvCa cells. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001 vs control at the same time points or doses.

Figure 4

Oncogenic activity of FXR1 and IGF2BP2 in OvCa. (A) Colony formation: Cells (10 000/well) were seeded on 0.6% agar media in 6-well plates and overlaid with 0.3% agar media. After 3 weeks, wells were stained with methylene blue, and colonies were counted (n = 5). (B) Cell adhesion: cells were seeded in 24-well plates and allowed to attach for 1 or 4 hr. Cell adhesion was assessed by Amido Black absorbance (n = 4). (C) Wound healing capability: Wound healing was assessed 24 h after scratch (n = 4) and expressed as the percentage of gap closure. (D) Cell migration: Cell migration assay was performed using a modified Boyden Chamber method with an 8-μm pore size polycarbonate membrane 24-well transwell (n = 4) ^*P < 0.05, ^**P < 0.01, ^***P < 0.001 vs control.