Bioengineered BERA-Wnt5a siRNA targeting Wnt5a/FZD2 signaling suppresses advanced prostate cancer tumor growth and enhances enzalutamide treatment

Abstract

The next-generation antiandrogen drugs such as enzalutamide and abiraterone extend survival times and improve quality of life in patients with advanced prostate cancer. However, resistance to both drugs occurs frequently through mechanisms which are incompletely understood. Wnt signaling, particularly through Wnt5a, plays vital roles in promoting prostate cancer progression and induction of resistance to enzalutamide and abiraterone. Development of novel strategies targeting Wnt5a to overcome resistance is an urgent need. In this study, we demonstrated that Wnt5a/FZD2-mediated non-canonical Wnt pathway is overexpressed in enzalutamide resistant prostate cancer. In patient databases, both the levels of Wnt5a and FZD2 expression are upregulated upon the development of enzalutamide resistance and correlate with higher Gleason score, biochemical recurrence and metastatic status, and with shortened disease-free survival duration. Blocking Wnt5a/FZD2 signal transduction not only diminished the activation of non-canonical Wnt signaling pathway, but also suppressed the constitutively activated AR and AR variants. Furthermore, we developed a novel bioengineered BERA-Wnt5a siRNA construct and demonstrated that inhibition of Wnt5a expression by the BERA-Wnt5a siRNA significantly suppressed tumor growth and enhanced enzalutamide treatment in vivo. These results indicate that Wnt5a/FZD2 signal pathway plays critical role in promoting enzalutamide resistance and targeting this pathway by BERA-Wnt5a siRNA can be developed as a potential therapy to treat advanced prostate cancer.

Introduction

Targeting androgen biosynthesis and androgen receptor (AR) signaling have been established as mainstay of the therapeutic strategies for castration resistant prostate cancer (CRPC) patients. Several next generation antiandrogens have been developed and approved by the Food and Drug Administration (FDA), including enzalutamide(1) and abiraterone (2). Enzalutamide directly binds to AR and impair AR nuclear translocation, DNA binding and coactivator recruitment (3). Although antiandrogen drugs have achieved significant clinical outcomes initially, unfortunately, the drug resistance is always induced and developed due to the adaptable cancer cells, which circumvent the therapeutics. Several critical mechanisms have been identified to contribute to the development of drug resistance in CRPC, including increased AR expression due to gene amplification or upregulation of mRNA, activation of gain-of-function mutation in receptor or ligand-independent activation of AR variants, and factors that activates the receptor independent of the level of ligand (4,5). Considerable studies have shown that reprogrammed classical AR signaling and emergence of AR splicing variants confers to treatment resistance after enzalutamide and abiraterone treatment(6–10). Upon the development of treatment resistance, patients present complicated genomic status independent of androgen/AR signaling, which renders AR-targeted therapy ineffective (8). It is critical to uncover the mechanism of anti-androgen drug resistance and develop potential strategies to improve the treatment for advanced prostate cancer.

Wnt signaling includes canonical (β-catenin-dependent) and non-canonical (β-catenin-independent) pathways. Non-canonical Wnt signaling is activated by a subset of Wnt ligands (such as Wnt5a and Wnt7b) and controls several downstream pathways, including Ca^2+/calmodulin-dependent protein kinase II, G proteins, Rho GTPases, or c-Jun N-terminal kinase (JNK), which are critical for cell survival, proliferation, and motility (11–13). Numerous studies shown that non-canonical Wnts correlate with aggressiveness and malignancies in melanoma, breast cancer, lung cancer, gastric cancer and prostate cancer (14–17). Wnt5a is a representative ligand that activates the beta-catenin-independent pathway in non-canonical Wnt signaling (17). Wnt5a plays important roles in cell proliferation, differentiation, migration, adhesion, and polarity. Upregulation of Wnt5a promotes epithelial to mesenchymal transition (EMT) and metastasis in pancreatic cancer, melanoma and ovarian cancer (18–20). Emerging evidence suggests that Wnt signaling plays an important role during progression to CRPC (21–23). The oncogenic role of Wnt5a prostate cancer was discovered in single circulating tumor cells from CRPC patients after the treatment of enzalutamide(23). Overexpression of Wnt5a attenuates the anti-proliferative effects of anti-androgen drugs in enzalutamide-resistant LNCaP cells, whereas suppression of Wnt5a could restore the sensitivity of the resistant LNCaP cells(23). Wnt5a was able to co-regulate the transactivation function of mutated AR and promote proliferation of PCa cells in an autocrine fashion (17). A recent study found substantial enrichment for Wnt5a in circulating tumor cells (CTCs) from Enza resistant PCa patients (24). However, the role of Wnt5a in conferring resistance to AR-targeted therapies such as enzalutamide is largely unknown.

Therapeutics targeting on Wnt signaling have been developed and investigated to treat Wnt hyperactive cancers(25). PORCN inhibitors has been developed to block the Wnt ligand secretion and activity, such as LGK974 in patients with solid malignancies (NCT01351103), and ETC-1922159 treating advanced solid tumors by single treatment or in combination with pembrolizumab (NCT02521844). Additionally, Frizzled receptor antagonistic antibody Vantictumab (OMP-18R5) that targeting on FZD1, FZD2, FZD5, FZD7 and FZD8 has been identified to inhibits the growth of human tumor xenografts(26). Several clinical trials are investigating the anti-tumor effects for Vantictumab in treating patients with recurrent or metastatic breast cancer (NCT01973309), non-small cell lung cancer (NCT01957007) and pancreatic cancer (NCT02005315). Ipafricept (IPA, OMP-54F28), a decoy receptor for Wnt ligands, has also been developed to treat advanced solid malignancies by inhibiting Wnt signaling activation(27). Here we developed novel approach to specifically target Wnt5a using bioengineered BERA-Wnt5a siRNA constructs and tested its efficacy in CRPC and improving enzalutamide treatment in resistant CRPC.

In present study, we demonstrated that non-canonical Wnt pathway via Wnt5a and FZD2 are upregulated in enzalutamide-resistant prostate cancer cells and prostate cancer patients, correlating with biochemical recurrence, higher Gleason score and shorten disease-free survival. Inhibition of Wnt5a and FZD2 expression significantly disturbs AR and AR variants gene reprograming and re-sensitizes the resistant cells to enzalutamide treatment both in vitro and in vivo. knocking down the expression of Wnt5a and FZD2 blocks the enrichment of noncanonical Wnt signaling and also the genes associated with cancerous cell survival and proliferation. Furthermore, specifically targeting Wnt5a using bioengineered BERA-Wnt5a siRNA constructs significantly suppresses tumor growth and improved enzalutamide treatment in resistant CRPC. These results suggest that targeting non-canonical Wnt signaling by inhibition of Wnt5a could be a potential therapy either alone or combination with enzalutamide for advanced prostate cancer.

Materials and Methods

Cell culture

C4–2B prostate cancer cells were acquired from the American Type Culture Collection (ATCC) and were maintained in PRMI1640 supplemented with 10% fetal bovine serum (FBS), 100 units per ml penicillin, and 0.1 mg per ml streptomycin. Enzalutamide resistant C4–2B MDVR cells were generated and maintained in 20 μM enzalutamide containing medium as described previously (28). C4–2B cells were passaged alongside with the resistant C4–2B MDVR cells as an appropriate control. C4–2B neo and C4–2B-Wnt5a cells were generated by stable transfection of C4–2B cells with either empty vector pcDNA3.1 or pcDNA3.1 encoding Wnt5a and were maintained in RPMI1640 medium containing 300 μg/ml G418. All cell lines have been routinely tested mycoplasma free by PCR and authenticated by short tandem repeat methods. All cells were cultured in 37 °C humidified incubators with 5% carbon dioxide. Enzalutamide was purchased from Selleck Chemicals. All drugs were dissolved in DMSO and stored at −20°C.

Plasmids and cell transfection

For overexpression of Wnt5a, Wnt5a-expressing plasmids were obtained from Addgene (Catelog#35911) and transfected into C4–2B cells alongside vector control using Lipofectamine as described previously(28). Wnt5a overexpression was confirmed via Western blot analysis as described below.

For small interfering RNA (siRNA) transfection, cells were plated at 60–80% confluence in 24-well plates or 60mm petri dishes and transfected with 20nM of siRNAs targeting the Wnt5a sequence (HSS187692 and HSS111355) and FZD2 sequence (IDT, Catalog# hs.Ri.FZD2.13.1 and hs.Ri.FZD2.13.2) or negative control siRNAs using Lipofectamine-RNAiMAX (Invitrogen). The effect of siRNA-mediated gene silencing was examined using qRT-PCR and western blot 2–3 days after transfection.

Protein extraction and western blot analysis

Cells were harvested and lysed in RIPA buffer and the concentration was determined by Coomassie Plus Protein Assay (Thermo Scientific, Catlog#1856210). Whole cell protein extracts were resolved on SDS-PAGE, and proteins were transferred into nitrocellulose membranes. After blocking in 5% milk in PBS+0.1% Tween-20 for 1 hour at room temperature, targeted proteins were detected by incubating with primary antibodies at 4 °C overnight. AR 441, 1:1000 dilution (Santa Cruz Biotechnology, Santa Cruz, CA. Catelog#22616); Wnt5a/b, 1:1000 dilution (Cell Signaling Technology, Catalog # 2530s); FZD2 (R&D, Catalog# MAB1307–050); Tubulin 1:5000 dilution from Sigma-Aldrich (Catalog#T5168), GAPDH (. Tubulin or GAPDH was used as loading control. Following secondary antibody incubation, immunoreactive proteins were visualized by applying Immobilon Crescendo Western HRP substrate (Millipore, Catalog#WBLUR0500).

Real-time quantitative RT-qPCR

Total RNA was extracted from cells using TriZOL reagent (Invitrogen Catalog# #15596018). cDNA was prepared using reverse transcriptase purchased from Promega. The cDNAs were subjected to quantitative PCR using SsoFast Eva Green Supermix (Bio-Rad) according to the manufacturer’s instruction. Samples were run in triplicates on Bio-Rad CFX-96 real-time cycler. Each reaction was normalized by co-amplification of actin. Primers used for real-time PCR are as follows: Wnt5a, 5’-CGCCTTCTCCGATGTACTGC-3’ (forward) and 5’-ATTCTTGGTGGTCGCTAGGTA-3’(reverse); FZD2, 5’-CACGGACATCGCCTACAACC-3’ (forward) and 5’- GCACCTTCACCAGCGGATAG-3’(reverse); AR, 5’-CCTGGCTTCCGCAACTTACAC-3’(forward) and 5’-GGACTTGTGCATGCGGTACTCA-3’ (reverse); AR-V7, 5’- AACAGAAGTACCTGTGCGCC −3’ (forward) and 5’- TCAGGGTCTGGTCATTTTGA −3’(reverse); PSA, 5’- CACAGACACCCCATCCTATC-3’ (forward) and 5’-GATGACTCCAGCCACGACCT-3’(reverse); NKX3–1, 5’-CCGAGACGCTGGCAGAGACC-3’ (forward) and 5’-GCTTAGGGGTTTGGGGAAG-3’ (reverse); MYC, 5’-TGAGGAGACACCGCCCAC-3’ (forward) and 5’-CAACATCGATTTCTTCATC-3’ (reverse); Actin, 5′-AGAACTGGCCCTTCTTGGAGG-3′ (forward) and 5′-GTTTTTATGTTCCTCTATGGG-3′ (reverse); GAPDH, 5’-GAAATCCCATCACCATCTTCC-3’ (forward) and 5’-ATGAGTCCTTCCACGATACCA-3’ (reverse).

Cell growth and survival assay

Cells were plated on 12-well plates at a density of 0.3 × 10⁵ cells per well in RPMI 1640 media containing 10% FBS and transfected with Wnt5a or FZD2 siRNAs and then treated with 20 μM enzalutamide for 3days. Total cell numbers were determined at 0, 3 and 5 days. C4–2B MDVR cells were plated on 12-well plates at a density of 0.5 × 10⁵ cells per well in RPMI 1640 media containing 10% FBS and transfected with 10 μM Wnt5a or FZD2 with or without 20μM enzalutamide. Cells viability was determined using CCK-8 reagent (Dojindo) according to manufacturer’s instructions.

Migration and invasion assay

C4-2B MDVR cell were treated with siRNAs targeting Wnt5a and FZD2 in 60mm² dishes for 24 hours. Cells 2.5 × 10⁵/ml were suspended in 200 μl serum-free RPMI 1640 medium into the cell culture inserts for 24-well plates (8 μm pore size; Corning Costar). The inserts were coated with Matrigel (Corning, Catelog#356231) and allowed to solidify for 30 min before cell plating. The lower chambers were filled with 500 μl complete growth RPMI 1640 medium. After 48h, cells were fixed with 5% glutaraldehyde in PBS, washed with PBS, and stained with 0.5% toluidine blue in 2% Na₂CO₃ solution. Invasive cells that penetrated the membrane were counted under the microscope; each group was counted at least three visual areas (random 40× magnification fields).

For wound healing assay, cells were seeded in 60 mm² dishes and allowed to adhere for 24 h. Cells were treated under the conditions as described above. Scratch wounds were introduced into the confluent monolayers with a 200μl pipette tip and then washed with PBS to clear cell debris and suspension cells. Fresh growth medium was added and wounds closure was monitored over time and photographed using fluorescence microscope (KEYENCE, U.S.A) at 40 × magnification. The wound closure was quantified by measuring the remaining unmigrated area using ImageJ software.

RNA extraction and RNA sequencing

C4-2B MDVR cells were treated with negative control or Wnt5a and FZD2 siRNAs 20 nM for 3 and 5 days before RNA extraction. Total RNA was isolated from C4–2B parental, MDVR cells and MDVR cells transfected with siRNAs targeting Wnt5a and FZD2 by TriZOL reagent (Invitrogen). RNA integrity was assessed using RNA Nano 6000 Assay Kit of the Bioanalyzer 2100 system (Agilent Technologies). RNA sequencing libraries were generated using NEBNext Ultra RNA Library Prep Kit for Illumina (NEB, USA) according to manufacturer’s instruction and index codes were added to attribute sequences to each sample. cDNA fragments of preferentially 150–200 bp in length, were selected using AMPure XP system (Beckman Coulter, USA). The index-coded samples were further clustered using PE Cluster Kit cBot-HS (Illumina) and sequenced on the Illumina platform. Paired-end clean reads were aligned to the reference human genome assembly (GRCh38/hg38) using the Spliced Transcripts Alignment to a Reference (STAR) software. The gene expression level was calculated as FPKM (Fragments Per Kilobase of transcript sequence per millions base pairs sequenced). The common downregulated genes from FZD2 siRNA knocking down was depicted with Venn diagram. The altered genes in MDVR compared with C4–2B parental and knocking down Wnt5a and FZD2 were clustered with hierarchical clustering algorithm by R programming.

Gene Set Enrichment Analysis

Gene expression results were subjected to Gene Set Enrichment Analysis (GSEA) to determine transcriptional program in different group of samples. Gene set enrichment analysis using GSEA desktop software from the Broad Institute was used to determine molecular pathways based on Molecular Signature Database (MSigDB)(29). The pathway analysis was performed using gene sets annotated by Kyoto Encyclopedia of Genes and Genomes (KEGG) and the Pathway Interaction Database (PID). Pathways enriched with a nominal p value lower than 0.05 and FDR q value lower than 0.25 were considered to be significant.

Gene expression omnibus and cBioPortal patient data analysis

Data sets from cBioPortal for Cancer Genomics (www.cbioportal.org) and NCBI’s Gene Expression Omnibus (GEO) were evaluated for the expression levels and clinical relevance of the Wnt5a and FZD2. Tumor sample information and corresponding clinical characteristics were obtained from The Cancer Genome Atlas (TCGA) (cBioPortal for Cancer Genomics http://www.cbioportal.org/study.do?cancer_study_id=prad_tcga_pub); and studies from GEO (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi) under the accession number GSE21032; GSE67980; GSE52169; GSE6099; GSE78201. Transcriptional expression matrix and the corresponding patient Gleason Score were analyzed in the data sets for TCGA, GSE21032, GSE67980, GSE66187. mRNA level in prostate cancer patients with Biochemistry recurrence and treatment refractory were analyzed based on data from GSE21032 and GSE6099. In prostate cancer LuCaP model, the expression level of Wnt5ad and its correlation with FZD2 were determined based on the data from GSE 52169. Prognosis analysis for patients’ Disease-Free Survival was performed on patients from Memorial Sloan-Kettering Cancer (MSKCC) study under the accession number GSE21032. Wnt signaling pathway activity was determined based on the CWR-R1 enzalutamide resistant cell model from GEO study under the accession number GSE78201.

PDX tumor xenografts and organoid culture

All animal experiments were approved by the Institutional Animal Care and Use Committee of UC Davis. LuCaP 35 CR models was obtained from the University of Washington and established at the UC Davis. Briefly, 3–4-week-old C.B17/lcrHsd-Prkdc-SCID mice (ENVIGO) were castrated and implanted with approximately 20 mm³ LuCaP 35 CR tumor pieces after 2 week. For organoids culture, LuCaP PDX-derived tumor tissues were collected and washed twice in cold PBS, and subjected to dissection to 2–4 mm³ with scalpel blade. Tumors were digested using collagenase IV (STEMCELL) in a 60 mm² petri dish and incubated in the 37°C humidified incubators with 5% CO₂ for 15–30 min until tumor cells were dispersed in digestive medium. Advanced DMEM medium supplemented with 1X GlutaMAX (Gibco), 1M HEPES (Gibco) and 100u/ml penicillin and 0.1mg/ml streptomycin were added to the cell suspension, and then filtered through 40μm cell strainers to obtain single cell suspension. The cells were centrifuged at 500g for 3min and the pellet was resuspended in Advanced DMEM complete medium containing GlutaMAX (Gibco), 100units/ml penicillin, 0.1mg/ml streptomycin, B27 (Gibco), N-Acetylcysteine (Thermo Scientific), Human Recombinant EGF (Thermo Scientific), Recombinant FGF-10 (Invitrogen), A-83–01 (Tocris), SB202190 (Bioscience), Nicotinamide (Thermo Scientific), dihydrotestosterone (Sigma), PGE2 (Bioscience), Noggin (Thermo Scientific) and R-spondin (R & D Systems)(30,31). Tumor cell pellet was seeded in 96 well plate with Matrigel diluted with 1:3 ratio of ADMEM complete medium and incubated in the 37°C humidified incubators with 5% CO₂ for 15min to solidify the 3D Matrigel complex. Then ADMEM complete medium mixed with corresponding treatment were added to each well. Viability of the organoids were analyzed using CellTiter-Glo Luminescent assay (Promega) and visualized by immunofluorescence using LIVE/DEAD® Viability/Cytotoxicity Assay Kit (Thermo Scientific) according to the manufacturer’s protocol.

Bioengineered BERA-tRNA-siWnt5a construct

Bioengineered BERA-siWnt5a constructs (BERA/Wnt5a-siRNA#1 and BERA/Wnt5a-siRNA#2) were developed using corresponding Wnt5a-siRNA expression plasmids through molecular cloning as described previously^(32,33). RNA-loaded liposome-polyethylenimine nanocomplex (Lipopolyplex or LPP) was conducted as described previously (34). 1,2-Di-O-octadecenyl-3-trimethylammonium propane (DOTMA, Avanti, USA), cholesterol (MP Biomedicals, CAS no. 57–88–5) and 1,2-Dimyristoyl-sn-glycerol, methoxypolyethylene glycol (DMG-PEG2000, SUNBRIGHT® GM-020CN) were mixed in chloroform and evaporate to form a lipid film using rotatory evaporator. The lipid film was dissolved in DEPC treated water and subject to sonication to form DOTMA-based cationic liposome. RNA-loaded Polyplex was prepared by incubating PEI (Branched PEI, bPEI10k, Alfa Aesar, USA) solution with RNA solution for 5min. RNA-loaded LPP was freshly prepared by incubating RNA-loaded Polyplex with liposome for 30 min. Zeta potentials and particle sizes of RNA-loaded LPP were determined by dynamic light scattering (Malvern Zetasizer Nano ZS90 instrument; Malvern instruments Ltd.; Malvern, U. K.) before injection to mice.

Animal studies and treatment regimens

All animals used in this study received humane care in compliance with applicable regulations, policies, and guidelines relating to experimental animals. All experimental procedures using animals were approved by the Institutional Animal Care and Use Committee of UC Davis. To determine the effect of bioengineered Wnt5a siRNA construct (BERA/Wnt5a-siRNA #2) on the growth of PDX tumors, the LuCaP 35 CR model was established in the UC Davis Cancer Center. 3–4-week-old C.B17/lcrHsd-Prkdc-SCID mice (ENVIGO) were surgically castrated. After two weeks, approximately 20–30mm³ pieces of LuCaP 35 CR tumor were implanted into the pre-castrated SCID mice. Tumor-bearing mice (tumor volume reaching 50–100 mm³) were randomized into two groups and treated as follows through intravenous (i.v.) injection: 1) tRNA control LSA (30 μg/mouse); 2) BERA/Wnt5a-siRNA#2 (30 μg/mouse). The LSA and BERA/Wnt5a-siRNA#2 were packaged with Lipopolyplex (LPP) immediately before use. Tumors were measured using calipers twice a week and tumor volumes were calculated using length × width × width × 0.52. Tumor tissues were harvested and weighed after 3 weeks of treatment. Tumor samples were paraffin embedded and subjected to IHC and H/E staining. Serum was collected for PSA measurements.

To evaluate the combinational effect of Wnt5a inhibition with enzalutamide on the growth of LuCaP PDX tumors, LuCaP 35 CR model was established on the pre-castrated SCID mice as described above. Once tumors reach 50–100 mm³, mice were randomized into four groups and treated as follows: 1) 0.5% weight/volume (w/v) Methocel A4M solution, p.o and tRNA control LSA (30 μg/mouse, i.v.); 2) 0.5% (w/v ) Methocel A4M solution, p.o BERA/Wnt5a-siRNA#2 (30 μg/mouse, i.v.); 3) enzalutamide (25 mg/kg in 0.5% (w/v) Methocel 4AM solution, p.o. and BERA/Wnt5a-siRNA#2 (30 μg/mouse, i.v.); 4) enzalutamide (25 mg/kg in 0.5% (w/v) Methocel 4AM solution, p.o. and BERA/Wnt5a-siRNA#2 (30 μg/mouse, i.v.). Tumors were monitored and measured using calipers twice a week and tumor volumes were calculated using length × width × width × 0.52. Tumor samples were harvested and weighed after 3 weeks of treatment. Tumor samples were paraffin embedded and subjected to IHC and H/E stained. Serum was collected for PSA determination.

Measurement of serum PSA

Mouse blood from the LuCaP 35 CR tumor model was collected and the serum was isolated. PSA levels were measured using PSA ELISA Kit (United Biotech, Inc., Mountain View, CA) according to the manufacturer’s instructions(35).

Immunohistochemistry

Tumors were fixed by formalin and paraffin-embedded tissue blocks were dewaxed, rehydrated, and blocked for endogenous peroxidase activity. Antigen retrieving was performed in sodium citrate buffer (0.01 mol per Litter, pH 6.0) in a microwave oven at 1000 W for 3 min and then at 100 W for 20 min. Nonspecific antibody binding was blocked by incubating with 10% FBS in PBS for 30 min at room temperature. Slides were then incubated with anti-Wnt5a/b, 1:1000 dilution (Cell Signaling Technology, Catalog # 2530s), anti-Ki67 (1:500, Noemarker) at 4°C overnight. Slides were then washed and incubated with biotin-conjugated secondary antibodies for 30 min, followed by incubation with avidin DH-biotinylated horseradish peroxidase complex for 30 min (Vectastain ABC Elie Kit, Vector Laboratories). The sections were developed with the diaminobenzidine substrate kit (Vector Laboratories) and counterstained with hematoxylin. Nuclear staining of cells was scored and counted in five different fields. Images were collected with all-in-one fluorescence microscope (Keyence).

Statistical analysis

Statistical analyses were performed with GraphPad Prism9.0. Raw data are summarized by means, standard deviations (SD), and graphical summaries and transformed if necessary to achieve normality. Data from the in vitro experiments were presented as means ± SD from three independent experiments. Differences between individual groups were analyzed by two-tailed Student’s t tests for single comparisons or one-way analysis of variance (ANOVA) for multiple group comparisons. In vivo tumor growth experiments, the volume and weight of the tumors at sacrifice serves as the primary response measure. Tumor growth and serum PSA across groups was analyzed by ANOVA. p value less than 0.05 was considered significant (*p<0.05, **p<0.01, ***p<0.005, ****p<0.001) unless otherwise indicated. Kaplan-Meier curves were plot to analyze the survival probability of the two groups. Log-rank test was used to compare the overall survival or disease-free survival between the two groups of different gene expression level. P value lower than 0.05 indicated that two groups differ significantly in disease-free survival.

Data availability

The data generated in this study are available within the article and its supplementary data files.

Results

Wnt5a/FZD2 signaling is activated in enzalutamide resistant prostate cancer

We have previously generated enzalutamide-resistant C4–2B MDVR cells via long-term culture C4–2B cells in media containing enzalutamide (28,36). In order to identify potential pathways that are associated with enzalutamide resistance in C4–2B MDVR cells, we interrogated the transcriptomic sequencing data in C4–2B MDVR cells. As shown in Figure 1A, non-canonical Wnt signaling pathways was significantly enriched in enzalutamide resistance C4–2B MDVR cells compared to the parental C4–2B cells. Wnt5a known as the representative non-canonical Wnt ligand, together with its cognate receptor FZD2 was significantly overexpressed in enzalutamide resistant C4–2B MDVR cells. We further validated that Wnt5a and FZD2 expression and showed that both mRNA and protein levels were elevated in C4–2B MDVR cells compared to parental C4–2B cells (Figure 1B and 1C). Similar findings showed that non-canonical Wnt signaling pathways was significantly enriched and Wnt5a/FZD2 expression was elevated in enzalutamide resistant CWR-R1 ENZR cells (Figure 1D). Wnt5a and FZD2 expression were positively correlated in CWR-R1 induced enzalutamide resistance cells (Figure.1E). In addition, short-term treatment of enzalutamide in CWR-R1 cells increased Wnt5a and FZD2 expression, which was further increased in enzalutamide-resistant CWR-R1 ENZR cells (Figure1F). In our another previously established abiraterone-resistant C4–2B AbiR cells, non-canonical Wnt signaling was also significantly enriched (p<0.05) as shown in Figure 1G GSEA enrichment plot and the heatmap. Western blotting analysis confirmed the increased protein of Wnt5a in both enzalutamide (C4–2B MDVR) and abiraterone (C4–2B AbiR) resistant cells (Figure 1H). In summary, Wnt5a and FZD2 significantly upregulated in both mRNA and protein levels and the non-canonical Wnt signaling enriched in enzalutamide resistant prostate cancer cells.

Figure 1. Wnt5a/FZD2 mediated non-canonical Wnt signaling upregulated in enzalutamide resistant prostate cancer.

A. GSEA enrichment plots revealed the significant enrichment of non-canonical Wnt signaling pathway in C4–2B MDVR cells compared to C4–2B parental cells. The heatmap demonstrates the top upregulated genes of non-canonical Wnt signaling pathway in C4–2B MDVR. B. mRNA levels of Wnt5a and FZD2 in C4–2B parental, C4–2B MDVR cells were determined by real-time PCR. C. The protein expression of Wnt5a, FZD2 and AR full length and variants was determined by western blot. D. GSEA enrichment plots and corresponding heatmap demonstrates strong upregulation of non-canonical Wnt signaling pathway in enzalutamide-resistant CWR-R1 cells based on RNA sequencing data from GSE 78201. NES Normalized Enrichment Score. E. In database GSE 78201, the RNA expression level of Wnt5a and FZD2 was determined in CWR-R1 parental, short-term treatment with enzalutamide and enzalutamide resistant ENZR cells. F. The correlation between FZD2 and Wnt5a expression was determined by Spearman rank correlation and the correlation coefficient was determined. G. GSEA enrichment plot shown non-canonical Wnt signaling in enzalutamide and abiraterone resistant cells compared to C4–2B parental cells. The heatmap shown the upregulated genes correspondingly in resistant C4–2B cells and parental cells. H. The protein expression of Wnt5a in C4–2B parental, MDVR and AbiR cells by Western blotting analysis.

Wnt5a/FZD2 significantly overexpressed in advanced prostate cancer patient cohorts

We next analyzed Wnt5a and FZD2 expression in the data bases generated from patients with prostate cancer. We initially analyzed Wnt5a and FZD2 expression in GEO and cBioPortal databases. Wnt5a and FZD2 are significantly upregulated in high Gleason score prostate cancer patients in TCGA(37), MSKCC(38) and single prostate circulating tumor cells(24) (Figure2A). To examine whether FZD2 expression is correlated with hormonal treatment refractory in prostate cancer, we interrogated GSE21032, GSE6099 databases which including 150 prostate tumors (131 primaries, 19 metastases) and 20 metastatic tumors respectively. FZD2 expression was significantly upregulated in prostate cancer patients with recurrence within 1 year and hormone therapy refractory group (Figure2B). To determine whether Wnt5a expression is correlated with enzalutamide treatment induced resistance in castration resistant prostate cancer, we analyzed the RNA-sequencing data in enzalutamide resistant xenograft derived tumor model(39). As shown in Figure2C, Wnt5a was significantly upregulated upon the short-term enzalutamide treatment, and further increased in the long-term resistant xenograft tumors according to the dataset of GSE52169. And in this enzalutamide resistant xenograft tumor model, Wnt5a and FZD2 expression showed a positive correlation (Figure 2D). These results suggest that Wnt5a and its cognate receptor FZD2 are associated with the development of enzalutamide resistance. We also evaluated the prognostic value of Wnt5a and FZD2 expression in patients from GSE21032. As shown in Figure 2E, higher levels of FZD2 expression were significantly associated with worse disease-free survival (DFS), although the p value for Wnt5a was not reaching significance. Our findings are consistent with the previous studies that Wnt5a and FZD2 interacts as non-canonical signaling in colon cancer, breast cancer(40) and prostate cancer(41). Collectively, these results reveal that Wnt5a and FZD2 are significantly overexpressed in advanced prostate cancer, correlating with clinicopathological features such as Gleason score, recurrence status, metastatic status and shortened disease-free survival time.

Figure2. Wnt5a/FZD2 expression is significantly increased in advanced prostate tumors.

A. In three independent GEO data sets (TCGA, GSE21032, GSE67980), Wnt5a and HZD2 gene expression levels were determined in tumor samples of different Gleason Scores. *p<0.05, **p<0.01, ***p<0.001. Statistical analysis was performed using ANOVA. B. In two independent patient databases (GSE21032 and GSE6099), FZD2 gene expression was determined in biochemical recurrence or treatment refractory groups of prostate cancer; in GSE21032 database, Wnt5a gene expression level in primary prostate cancer and metastatic prostate cancer tumor samples were also determined. C. In LNCaP and enzalutamide resistant LREX xenograft tumors from GSE 52169, Wnt5a expression level was determined in LnCaP/AR, LnCaP/AR short-term treated with enzalutamide and enzalutamide-resistant LREX tumors. The Wnt5a/FZD2 correlation coefficient were determined by Spearman rank correlation. the Wnt5a/FZD2 correlation coefficient were determined by Spearman rank correlation. The Wnt5a/FZD2 correlation coefficient were determined by Spearman rank correlation. D. Kaplan-Meier analysis was performed to determine the correlation of high expression of Wnt5a and FZD2 with the Disease-Free Survival (DFS) in MSKCC cohorts (GSE21032). p value less than 0.05 was considered significant (*p<0.05, **p<0.01, ***p<0.005, ****p<0.001). Results are the mean of three independent experiments (±S.D.). Statistical analysis was performed using one-way ANOVA.

Silencing Wnt5a/FZD2 expression inhibited migration and invasion and restored sensitivity to enzalutamide treatment

The non-canonical Wnt signaling regulates cellular migration and invasion in cancer (40) and neural development process(42). To determine the potential activity of Wnt5a in prostate cancer cells, we knocked down Wnt5a using siRNAs and examined the effects on cell growth, migration and invasion in C4–2B MDVR cells. Knockdown of Wnt5a robustly decreased the migration and invasion in C4–2B MDVR cells (Figure 3A–3B, Suppl Figure 1A–1B), and significantly inhibited cell proliferation and improved enzalutamide treatment in C4–2B MDVR cells (Figure 3C and 3D). Performing transcriptomic analysis, knocking down Wnt5a in C4–2B MDVR cells significantly suppressed the enrichment of Wnt noncanonical signaling (Supplementary Figure 1C). Gene program regulating AR targeted transcriptional factors, AR variant 7 associated genes and cancerous cell survival and growth were significantly decreased in Wnt5a knocking down group (Supplementary Figure 1D and 1E). We next examined if blocking FZD2 expression could also affect cell migration and invasion in C4–2B MDVR cells. Knocking down of FZD2 expression suppressed cellular migration and invasion in treatment resistant C4–2B MDVR cells (Figure 3E–3F, and Suppl Figure 1F–1G). Furthermore, knocking down FZD2 restored the sensitivity of C4–2B MDVR cells to enzalutamide treatment (Figure 3G and 3H). Conversely, Wnt5a overexpression dramatically increased cellular invasion in C4–2B cells (Figure 3I). In addition, knocking down of FZD2 expression by FZD2 specific siRNA significantly decreased cell invasion of the stable clones (Wnt5a#10 and Wnt5a#11) of C4–2B cells overexpression Wnt5a (Figure 3J). Taking together, these results suggest that Wnt5a/FZD2 pathways play a critical role in promoting cell proliferation and invasion, and sensitivity to enzalutamide treatment in enzalutamide-resistant C4–2B MDVR cells.

Figure 3. Wnt5a/FZD2 suppressing inhibited non-canonical Wnt signaling activation in enzalutamide resistance.

A. C4–2B MDVR cells were treated with 20nM siRNAs targeting Wnt5a in media containing FBS after wound were introduced. The wound closure was quantified at the time indicated. B. C4–2B MDVR cells were allowed to migrate through Matrigel coated membranes with 8 μm pores for 48 hours after the transfection with siRNAs targeting Wnt5a. Invasive cells were counted and quantified under microscope. C. C4–2B MDVR cells were transiently transfected with Wnt5a siRNAs and then treated with 20μM enzalutamide for 3 days and the proliferation rate was determined by CCK8. D. qRT-PCR analysis of Wnt5a expression in C4–2B MDVR cells after transfection with siRNAs targeting Wnt5a for 48h. E. C4–2B MDVR cells were transiently knocking down FZD2 and the wounds were introduced into the 60 mm² petri dishes. The wound closure was quantified at the time indicated. F. C4–2B MDVR cells were allowed to migrate through Matrigel coated membranes with 8 μm pores for 48 hours after transiently knocking down FZD2. The invasive cells in groups of control and knocking down siRNA were quantified. G. Proliferation rate was determined in C4–2B MDVR cells after treated with siRNAs targeting FZD2 and then treated with 20μM enzalutamide for 3 days. H. mRNA expression and protein levels of full-length AR and its variants, FZD2 and Wnt5a were examined by RT-qPCR. I. Representative images of the invasive C4–2B cells with Wnt5a transfection. Then cells were allowed to migrate through Matrigel coated membranes with 8 μm pores for 48 hours after transiently transfect with Wnt5a-expressing plasmids. Invasive cells were counted under the microscope. J. Representative images of the invasive C4–2B-Wnt5a stable clones after knocking down FZD2. The cells were seeded in the 24-well plate insert to migrate through Matrigel coated membranes with 8 μm pores for 48 hours after knocking down FZD2. Invasive cells were quantified and imaged under the microscope. p value less than 0.05 was considered significant (*p<0.05, **p<0.01, ***p<0.005, ****p<0.001). Results are the mean of three independent experiments (±S.D.). Statistical analysis was performed using one-way ANOVA.

Silencing cognate receptor FZD2 blocks Wnt5a-mediated non-canonical Wnt and AR/AR-V7 signature gene program

Our previous studies showed that enzalutamide resistant C4–2B MDVR cells overexpress AR-V7 and that knocking down AR-V7 expression restored cell sensitivity to enzalutamide treatment. To determine if Wnt5a/FZD2 pathways regulate AR/AR-V7 signaling, we performed RNA sequencing analyses using C4–2B MDVR cells treated with either FZD2 or Wnt5a specific siRNAs. There are 11021 genes and 11494 genes were significantly altered after knocking down FZD2 by two different siRNAs respectively, and 8953 genes were commonly regulated by both FZD2 specific siRNAs (FKPM>1) (Figure4A). Altered genes were clustered together in the gene expression pattern compared with control siRNA group as shown in the hierarchical clustering heatmap of altered genes after knocking down FZD2, indicating a consistence of the gene alteration were affected by both two FZD2 targeted siRNAs (Figure4B). Further gene set enrichment analysis (GSEA) revealed that suppressing FZD2 expression significantly blocked the hallmark genes of androgen response (NES=1.3676, FDR-q value<0.05), AR-V7 related signature genes (NES=2.4348, FDR-q value<0.001), and genes regulating cancerous cell survival and proliferation (NES=1.6972, FDR-q value<0.005) (Figure4C). Silencing FZD2 by specific siRNAs impeded the activity of non-canonical Wnt signaling indicated by decreased gene expression of downstream transcription factors NFAT and MYC (Figure4D). Looking at the individual gene expression, our data indicated that classical androgen response and AR-V7 signature genes, such as ASCL3, AKT1, ELK4, UBE2C were suppressed after inhibition of FZD2. Gene program regulating tumor cell proliferation such as E2F1, MYC, ALDH1B1 and FAS was also decreased after silencing FZD2. Furthermore, the levels of AR-FL/AR-Vs, Wnt5a proteins were inhibited by Western blot analysis, and AR and AR-V7 associated genes including PSA, NKX3.1 and MYC were also decreased after knocking down FZD2 in C4–2B MDVR (Figure4E). Similar results by GSEA analysis showed that silencing Wnt5a by specific siRNAs significantly downregulated the AR regulated transcription factors (NES=1.7256, FDR-q value=0.01961), AR-V7 related signature genes (NES=1.3193, FDR-q value=0.06276), and genes regulating cancerous cell survival and proliferation (NES=2.1158, FDR-q value<0.001) (Suppl Fig 1). Taking together, that data indicate that suppressing Wnt5a/FZD2 expression inhibited the gene programs that regulate non-canonical Wnt signaling, classical androgen response, AR-V7 target genes and cancerous cell proliferation and survival.

Figure 4. FZD2 knocking down diminishes noncanonical Wnt signaling and AR/AR-V7 signature.

A. Venn diagrams of RNA-sequencing analysis of the two FZD2 knocking down groups in C4–2B MDVR depicting the common altered genes, upregulated and downregulated genes. B. GSEA enrichment plots of Hallmark Androgen Response, AR-V7 associated genes and gene sets for cancerous cell growth in C4–2B MDVR with knocking down FZD2. C. Heatmap and hierarchical clustering of the differentially expressed genes between FZD2 targeted knocking down in C4–2B MDVR cells compared with control siRNA group. The top downregulated genes in C4–2B MDVR transfected with FZD2 siRNAs were displayed corresponding to the gene sets of Hallmark Androgen Response, AR-V7 associated signature gene and genes regulating cancerous cell survival/proliferation/growth. D. GSEA enrichment plots of the PID-noncanonical Wnt pathway and KEGG Wnt pathway suggested a decrease in the enrichments in C4–2B MDVR cells after knocking down of FZD2 gene expression. Heatmap presents the downregulated genes of non-canonical Wnt signaling in C4–2B MDVR cells with downregulating FZD2 expression. E. Western blotting analysis present the knocking down effects in C4–2B MDVR cells after transfected with siRNAs targeting FZD2. F. qRT-PCR analysis of the indicated genes in C4–2B MDVR cells transfected with siRNAs targeting FZD2 for 48 hours.

Targeting WNT5a suppressed tumorigenesis in vivo

In an attempt to target WNT5a signaling, we have developed a bioengineered BERA-Wnt5a siRNA construct using corresponding Wnt5a-siRNA expression plasmids through molecular cloning as described previously (32). We have purified recombinant Wnt5a-siRNA agents (BERA/Wnt5a-siRNA#1 and BERA/Wnt5a-siRNA#2) to high homogeneity (>95%) using an established anion exchange FPLC method (Suppl Figure 2A–2C) (32,33). In addition to spectrometry and gel electrophoresis analyses of target Wnt5a-siRNAs during the purification, we have conducted HPLC analysis to validate the purity of isolated siRNAs and LC-MS studies to verify RNA sequence and identify possible posttranscriptional modifications, as described in our recent publications (33,43). We have purified bioengineered siWnt5a agents (BERA/Wnt5a-siRNA#1 and BERA/Wnt5a-siRNA #2), and functional studies on BERA/Wnt5a-siRNA#2 showed that BERA/Wnt5a-siRNA#2 downregulated Wnt5a expression in C4–2B MDVR cell growth (Figure 5A and 5B). To further characterize the effects of BERA/Wnt5a-siRNA#2 on tumor growth in vivo, the enzalutamide resistant LuCaP35 CR model was used which have increased protein expression of Wnt5a and FZD2 (Figure 5C, Supplementary Figure 2D and 2E). As shown in Figure 5D–F BERA/Wnt5a-siRNA#2 significantly suppressed LuCaP 35CR growth and tumor weight but did not alter mouse body weights. The BERA/Wnt5a-siRNA#2 treatment also significantly suppressed serum PSA level (Figure 5G). Immunohistochemical staining of Wnt5a and Ki67 showed that BERA/Wnt5a-siRNA#2 significantly decreased Wnt5a and Ki67 expression in tumors (Figure 5H and 5I). These results suggest that targeting Wnt5a by the novel BERA-Wnt5a siRNA construct can significantly inhibit enzalutamide resistant tumor growth in vivo.

Figure 5. Wnt5a inhibition suppressed enzalutamide resistant LuCaP tumor growth.

A. C4–2B MDVR cells were transfected with bioengineered Wnt5a siRNA construct (BERA/Wnt5a-siRNA #2) together with LSA vector control for 48 hours. The RNA was extracted and the mRNA level of Wnt5a was analyzed by qRT-PCR. B. The corresponding whole cell lysates were harvested and subjected western blotting analysis. C. Western blot shown Wnt5a protein expression in LuCaP 35CR patient derived tumor xenograft tumor (s.e., short exposure; l.e. long exposure). D. Mice bearing LuCaP 35 CR xenografts were treated with control LSA and BERA/Wnt5a-siRNA #2 (30μg/mouse, i.v.) for 3 weeks. Tumor volumes were measured twice a week. E. Body weight was measured. F. Tumor weight were measured at time of sacrifice. G. PSA level in the mice serum was determined in the two groups. IHC staining of Ki67 and Wnt5a in each group was performed and quantified H-I. IHC staining of Ki67 and Wnt5a in each group was performed and quantified. *p<0.05. Results are the mean of three independent experiments (±S.D.). Statistical analysis was performed using one-way ANOVA.

Targeting WNT5a/FZD2 enhance enzalutamide effects in LuCaP 35CR organoids and PDX model

To further examine if targeting Wnt5a enhance the enzalutamide treatment in resistant prostate cancer, we initially determined the combinational effects of Wnt5a inhibition and enzalutamide in ex vivo model. LuCaP35CR organoids was developed in 3D Matrigel conditions and treated with BERA/Wnt5a-siRNA#2 and the corresponding negative control LSA with or without enzalutamide. LuCaP 35CR organoids remained resistant to enzalutamide treatment, however, combinational treatment with BERA/Wnt5a-siRNA#2 significant decreased the viability of resistant prostate cancer organoids (Figure 6A). To further validate the combinational effects of Wnt5a inhibition and enzalutamide in vivo, we employed LuCaP 35CR PDX model and treatment with bioengineered BERA/Wnt5a-siRNA#2 in combination with enzalutamide. As shown in Figure 6B, LuCaP 35CR tumors were resistant to enzalutamide treatment, and single treatment of BERA/Wnt5a-siRNA#2 significantly inhibited the tumor growth (p<0.05). Combinational of BERA/Wnt5a-siRNA#2 with enzalutamide further suppressed the tumor growth in LuCaP 35CR PDX tumors (Figure 6B and 6D), while the body weights were not affected by either LSA or BERA/Wnt5a-siRNA#2 treatments (Figure6C). Enzalutamide treatment affected the serum PSA expression without reaching significance (p>0.05), however the combinational treatment using BERA/Wnt5a-siRNA#2 with enzalutamide significantly reduced the PSA level (p<0.05) (Figure 6E). Moreover, immunohistochemical staining of Ki67 verified that prostate cancer cell proliferation was significantly inhibited by Wnt5a inhibition alone, which was further inhibited by the combination treatment with enzalutamide (Figure 6F–6H). Collectively, our results suggest that inhibition Wnt5a expression by bioengineered siRNA constructs reduces enzalutamide-resistant prostate tumor growth, the effects of which could be further enhanced by the combinational treatment with enzalutamide.

Figure 6. Wnt5a inhibition enhanced enzalutamide therapy in vivo.

A. Organoids from LuCaP 35 CR xenograft tumors were collected and plated in 96 well plate in format of 3D Matrigel and then treated with BERA-Wnt5a siRNA constructs (BERA/Wnt5a-siRNA#2) at the indicated dose with or without 20μM enzalutamide for 14 days. The viability of the organoids was determined by CellTiter Glo and visualized by LIVE/DEAD staining. LIVE= Calcein staining of live cells in green, DEAD= Ethidium homodimer-1 staining of dead cells in red. B. Mice bearing LuCaP 35 CR xenografts were treated with BERA/Wnt5a-siRNA#2 (30μg/mouse, i.v.), enzalutamide (25mg/kg, p.o) or their combination for 3 weeks. The tumors were collected at the end of the experiment. Tumor volumes were measured twice a week and quantified. Data present means ±S.D. from five mice per group. C. Body weight of mice was determined for each group. D. Tumor weight was assessed and quantified at the day of sacrifice. E. PSA level in the mice serum was determined by ELISA assay. F-H. IHC staining of Ki67 and Wnt5a for each group were performed and quantified. *p<0.05, **p<0.01, ***p<0.005, ****p<0.001. Results are the mean of three independent experiments (±S.D.). Statistical analysis was performed using one-way ANOVA.

Discussion

Our present study reveals the contributing role of Wnt5a/FZD2 signaling in the development of resistance to enzalutamide treatment, and provides a novel therapeutic strategy to target Wnt5a overcoming treatment resistance. We found that Wnt5a and its receptor FZD2 mediated non-canonical Wnt signaling pathway are enriched in enzalutamide resistant prostate cancer C4–2B MDVR cells and in advanced prostate cancer, and function as critical molecules bypassing the AR activation in antiandrogen treatment resistance. Non-canonical Wnt signaling activation could be suppressed by targeting Wnt5a, effectively decreasing the hallmark androgen response and AR V7 signature genes and restored the sensitivity to enzalutamide treatment. Gene programs regulating tumor cell survival and proliferation could also be downregulated after blocking Wnt5a and FZD2 expression. Furthermore, we demonstrated that targeting Wnt5a by a novel bioengineered therapeutic siRNA construct suppressed tumor growth and improved enzalutamide treatment in resistant LuCaP patient derived tumor model.

Although AR abnormalities such as AR amplification/mutation, AR variants, and intratumoral biogenesis of androgens have been the major mechanisms for hyperactive AR during the development of antiandrogen resistance(6,10,44–46), AR independent mechanisms have emerged during the progression of CRPC including Wnt signaling activation(39,47). Numerous studies have shown that non-canonical Wnt ligands including Wnt5a and Wnt7b upregulated in the advanced castration resistant prostate cancer(24,48,49). High expression of Wnt5a in prostate cancer of higher Gleason scores is associated with EMT and biochemical recurrence(23,41). In enzalutamide-resistant LNCaP cells, overexpression of Wnt5a attenuates the anti-proliferative effects of anti-androgen drugs, whereas Wnt5a suppression restores the sensitivity to the antiandrogens (23,24,50). Our data not only support these findings but also demonstrate that FZD2, the binding receptor of Wnt5a, is highly upregulated in enzalutamide-resistant C4–2B MDVR cells and CRPC patients, and the higher levels of expression of both Wnt5a and FZD2 are associated with biochemical recurrence, higher Gleason score and shorten disease-free survival in prostate cancer patients. Furthermore, our data shown that FZD2 overexpression in advanced prostate cancer is consistent with the reports that FZD2 is frequently overexpressed in late-stage mesenchymal-type cancer cells, such as hepatocellular carcinoma, breast cancer, lung cancer and colon cancer(40,51,52), and higher levels of FZD2 expression is correlated with poor survival rate of cancer patients(53).

Several strategies have been explored for targeting Wnt signaling in prostate cancer (54). However, none of them directly target noncanonical Wnt/Wnt5a signaling. The development of new classes of therapeutics such as small interfering RNAs (siRNA) is a specific way with less off-target effects during treatment. In recent years, therapeutic RNAs has advanced, such as small interfering RNA (siRNA) givosiran(55), lumasiran(56), and most recent mRNA-based vaccines for SARS-CoV-2(57). Transfer RNA (tRNA) scaffold can act as a prodrug for antitumor treatment(33) and polyplexes carriers can achieve desirable delivery for RNA therapeutics(58). However, unfavorable pharmacokinetic properties and side effects are major drawbacks for siRNAs to advance into in vivo studies and clinical trials (59). Motivated by the idea to produce biological RNAs to perform RNA actions, we have developed a novel RNA bioengineering technology to achieve high-yield (e.g., 10–20% of total RNAs) and large-scale (mg of ncRNAs per liter of bacterial culture) production of biological miRNA/siRNA agents (32), which is based upon an optimal tRNA/pre-miRNA noncoding RNA scaffold (OnRS) (33,60,61). The bioengineered tRNAs have shown their activity in the control of human carcinoma cell proliferation, target gene expression, xenograft tumor progression, and safety profiles (33,60,61). In this study, we demonstrated that bioengineered BERA-Wnt5a siRNA constructs effectively inhibited Wnt5a expression and suppressed tumor growth and enhanced enzalutamide treatment in treatment resistant LuCaP35CR xenograft model.

In summary, our results demonstrated the contributing role of Wnt5a/FZD2 signaling in the development of resistance to enzalutamide treatment, and provides a potential therapeutic strategy to target Wnt5a via a novel bioengineered therapeutic siRNA construct either alone or in combination with anti-androgens such as enzalutamide to treat advanced prostate cancer.