Role of eIF4E on epithelial-mesenchymal transition, invasion, and chemoresistance of prostate cancer cells.

(1‐aminFoethyl)iminobis[N‐oleicylcysteinyl‐1‐aminoethyl) propionamide]

eukaryotic translation initiation factor 4E

epithelial‐mesenchymal transition

mesenchymal‐epithelial transition

matrix metalloproteinase 3

prostate cancer

paclitaxel

small interfering RNA

Dear Editor,

Prostate cancer (PCa) is the most common cancer and second leading cause of cancer death for men in the United States [1]. PCa with similar Gleason score has been reported to show substantial interpatient heterogeneity and differential prostate cancer‐specific mortality rate [2]. Such heterogeneity in PCa often results in different therapeutic responses among patients, including therapy resistance, therapeutic failure, relapse, and metastasis [3]. Numerous oncogenes, such as eukaryotic translation initiation factor 4E (eIF4E), have been reported to be involved in epithelial‐mesenchymal transition (EMT) and/or drug resistance in PCa [4]. We previously demonstrated that eIF4E overexpression was involved in chemoresistance of triple‐negative breast cancer and silencing eIF4E significantly inhibited cancer cell proliferation and sensitized cancer cells to chemotherapy in a patient‐derived xenograft mouse model [5]. In addition, eIF4E phosphorylation is known to stimulate the translation of matrix metalloproteinase 3 (MMP3) and Snail mRNAs to promote EMT in PCa [6]. Furthermore, the complexity and dynamic nature of EMT contributes to the heterogeneity of aggressive cancer cells [7]. The precise role of eIF4E in EMT, invasion, and chemoresistance in PCa is still to be established with consideration of different subpopulations in order to develop precision medicine for PCa. In this work, we aimed to explore the role of eIF4E in EMT, invasion, and chemoresistance in PCa for establishing a promising new therapeutic strategy by regulating eIF4E expression using (1‐aminoethyl)iminobis[N‐oleicylcysteinyl‐1‐aminoethyl) propionamide] (ECO)/small interfering RNA (siRNA) nanoparticles previously developed in our lab [8, 9, 10] for PCa therapy in the context of tumor heterogeneity.

Two PCa cell lines (PC3 and DU145) and their corresponding paclitaxel (PTX)‐resistant cell lines (PC3‐DR and DU145‐DR) were investigated to assess the role of eIF4E in EMT, invasion, and chemoresistance of PCa cells (more details in Supplementary information). The average IC50 of PTX for PC3‐DR cells (577.88 nmol/L) was about 69 times higher than that for PC3 cells (8.37 nmol/L), and the average IC50 of PTX for DU145‐DR cells (379.15 nmol/L) was about 85 times higher than that for DU145 cells (4.44 nmol/L), indicating that both the drug‐resistant cell lines acquired significant resistance to PTX treatment. Interestingly, PC3‐DR and DU145‐DR cells displayed different morphology and invasiveness from their corresponding parental cell lines (Figure 1A‐D). PC3 cells were inherently invasive and displayed spindle‐like mesenchymal morphology, while PC3‐DR cells exhibited a squamous or epithelial morphology (Figure 1A and B). In addition, wound healing and transwell invasion assays revealed that PC3‐DR cells were less migratory and invasive than PC3 cells (Figure 1C and D, Supplementary Fig. S1). On the contrary, DU145 cells displayed the epithelial hallmark of densely packed squamous morphology, whereas DU145‐DR cells exhibited an elongated and spindle‐like morphology, which is characteristic of mesenchymal cells (Figure 1A and B). DU145‐DR cells were more invasive than DU145 cells, as evidenced by an increase in migrated cells in transwell migration assay and narrower gap in wound healing assay at 24 h (Figure 1C and D, Supplementary Fig. S1).Moreover,significant down‐regulation of mesenchymal markers, such as N‐cadherin and ZEB‐1, was observed in PC3‐DR cells as compared to PC3 cells. On the other hand, significant up‐regulation of N‐cadherin, vimentin, and ZEB‐1 was observed in DU145‐DR cells as compared to DU145 cells (Figure 1E and F). Taken together, these results demonstrated that although both PC3‐DR and DU145‐DR cells developed significant chemoresistance to PTX, only DU145‐DR cells acquired features of EMT, characteriing by the transition to mesenchymal phenotype and increased invasion in response to PTX resistance. For the inherently mesenchymal PC3 cell line, its corresponding PTX‐resistant cell line PC3‐DR underwent mesenchymal‐epithelial transition (MET) and acquired epithelial morphology with decreased migration and invasion.

FIGURE 1

Cell morphology, migration and invasion ability, and EMT features of parental prostate cancer cell lines (PC3 and DU145) and their corresponding paclitaxel‐resistant cell lines (PC3‐DR and DU145‐DR). A. Representative images of cell morphology under microscope. Scale bar: 100 µm. B. Representative images of cytoskeleton immunofluorescence staining using anti‐F‐actin (red) antibody. Nuclei were labeled with DAPI (blue). Scale bar: 100 µm. C. Cell invasion ability assessed by transwell invasion assays. Scale bar: 200 µm. D. Cell migration potential assessed by wound healing assays. Scale bar: 500 µm. E‐F. Expression of EMT markers at mRNA (E) and protein (F) levels determined by semiquantitative real‐time PCR and Western blotting, respectively. β‐actin was used as loading control. The data are represented as mean ± SD from 4 independent experiments. *P < 0.05, ***P < 0.001, vs parental cells. Abbreviations: EMT, epithelial‐mesenchymal transition

Similar to mesenchymal markers, significant down‐regulation of eIF4E expression was observed in PC3‐DR cells as compared to PC3 cells. Conversely, DU145‐DR cells showed up‐regulated eIF4E expression and phosphorylation as compared to DU145 cells (Figure 2A and B). The expression of eIF4E followed the same trend with the expression of mesenchymal markers in all four cell lines, regardless of drug resistance. Considering the results of morphology, EMT, and invasiveness of all four cell lines, it was found that eIF4E expression was associated with EMT of PCa cells, irrespective of whether they were PTX‐resistant or not, indicating a possible role of eIF4E in EMT and invasion and a limited role in the development of PTX resistance of PCa cells. Regarding the tumor heterogeneity of PCa, different subtypes of PCa may develop PTX resistance through different routes.

FIGURE 2

eIF4E expression in prostate cancer cells. A‐B. Expression of eIF4E in parental prostate cancer cell lines (PC3 and DU145) and their corresponding paclitaxel‐resistant cell lines (PC3‐DR and DU145‐DR) at mRNA (A) and protein (B) levels determined by semi‐quantitative real‐time PCR and western blotting, respectively. C‐D. The eIF4E expression at mRNA (C) and protein (D) levels in parental prostate cancer cell lines (PC3 and DU145) and their corresponding paclitaxel‐resistant cell lines (PC3‐DR and DU145‐DR) after treatment with ECO/siNS and ECO/sieIF4E nanoparticles determined by semi‐quantitative real‐time PCR and western blotting, respectively. β‐actin was used as loading control. The data are represented as mean ± SD from 4 independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001, vs parental cells. Abbreviations: ECO, (1‐aminoethyl) iminobis [N‐(oleicylcysteinyl‐1‐amino‐ethyl) propionamide]

ECO/siRNA nanoparticles have been shown to mediate effective RNAi of various oncogenic targets in vitro and in vivo [8, 9, 10]. To further confirm the role of eIF4E in PCa cells, the effects of down‐regulation of eIF4E were assessed using RNAi based on ECO/siRNA nanoparticles. A multifunctional pH‐sensitive amino lipid ECO (Supplementary Fig. S2A) was employed as a carrier to formulate stable nanoparticles with sieIF4E via self‐assembly for silencing eIF4E in PCa cells. Both ECO/sieIF4E and ECO/siNS (control) nanoparticles displayed uniform size distribution and positive zeta potential (Supplementary Fig. S2B and C). The gel retardation assay demonstrated efficient encapsulation of siRNA in the nanoparticles, with negligible free siRNA bands (Supplementary Fig. S2D). Compared to the negative control ECO/siNS nanoparticles, ECO/sieIF4E nanoparticles resulted in significant eIF4E down‐regulation in all four PCa cell lines, both at mRNA and protein levels, with more significant silencing in the eIF4E‐rich PC3 and DU145‐DR cells (Figure 2C and D).

ECO/sieIF4E treatment significantly enhanced the response of PC3 and DU145‐DR cells to PTX at low concentrations when compared with the ECO/siNS treatment. As for DU145 and PC3‐DR cells, which displayed epithelial features and low expression of eIF4E, ECO/sieIF4E treatment did not show a significant change in response to PTX treatment, compared with ECO/siNS treatment (Supplementary Fig. S3A). The average IC50 of PTX for ECO/sieIF4E‐treated DU145‐DR cells (236 nmol/L) was about 35% less than that for ECO/siNS‐treated DU145‐DR cells (365 nmol/L), suggesting that silencing eIF4E can resensitize PTX‐resistant DU145‐DR cells. However, treatment with ECO/sieIF4E did not have the same resensitizing effect on PC3‐DR cells, possibly due to their low eIF4E expression and epithelial nature. Silencing eIF4E with ECO/sieIF4E nanoparticles also resulted in changes in molecular signatures of PC3 and DU145‐DR cells. Reduced expression of the mesenchymal markers, ZEB‐1, N‐cadherin, and vimentin, was observed in the western blots of ECO/sieIF4E‐treated cells. In contrast, there was no change in these markers in DU145 and PC3‐DR cells after treatment with ECO/sieIF4E (Supplementary Fig. S3B).

ECO/sieIF4E‐treated PC3 and DU145‐DR cells also showed a significant reduction in their migration and invasion abilities, evidenced by the significantly reduced number of migrated cells stained in purple compared with those treated with ECO/siNS (Fig. 3A, Supplementary Fig. S4). The ECO/sieIF4E‐treated PC3 and DU145‐DR cells also migrated slower than those treated with ECO/siNS in wound healing assay (Fig. 3B, Supplementary Fig. S5). For DU145 and PC3‐DR cells, ECO/sieIF4E treatment did not result in significant changes in migration or invasion, compared with ECO/siNS treatment (Fig. 3A, Supplementary Fig. S4). Moreover, no significant differences were observed in the wound closure between ECO/siNS and ECO/sieIF4E treatment for these cell lines (Fig. 3B, Supplementary Fig. S5). Taken together, these results indicate that silencing eIF4E by ECO/siRNA nanoparticles can overcome chemoresistance and inhibit EMT or invasion of PCa cells with a mesenchymal phenotype (PC3 and DU145‐DR), rather than an epithelial phenotype (DU145 and PC3‐DR).

FIGURE 3

Cell migration and invasion ability of normal prostate cancer cell lines (PC3 and DU145) and their corresponding paclitaxel (PTX)‐resistant cell lines (PC3‐DR and DU145‐DR) after treatment with ECO/siNS and ECO/sieIF4E nanoparticles. A) Cell migration and invasion ability assessed by transwell study. Scale bar: 200 µm. B) Cell migration potential assessed by standard scratch‐wound assays. Scale bar: 200 µm.

In the context of tumor plasticity, these results have important implications in the treatment planning for PCa, which is highly heterogeneous. As a result of this heterogeneity, the bulk tumor might include a diverse collection of cells harboring distinct molecular signatures with different characteristics and thus result in differential levels of sensitivity to treatment [11]. In the present study, eIF4E expression is high in cells undergoing EMT, irrespective of drug resistance. Silencing eIF4E by ECO/siRNA nanoparticles can overcome chemoresistance and inhibit EMT or migration in PCa cells with a mesenchymal phenotype (PC3 and DU145‐DR), rather than an epithelial phenotype (DU145 and PC3‐DR). These results have important implications on targeting eIF4E for the treatment of PCa with EMT features.

Supporting Information

Click here for additional data file.