Introduction

Prostate tumours not controlled by radiation therapy1 may present with radiation protective biological characteristics2 whose pre-treatment identification has the potential to predict treatment outcomes and initiate the development of novel, more aggressive, treatment options. Isogenic models of radioresistance are emerging as clinically-relevant models for the study of these tumours characteristics3. This approach has been particularly useful in the characterisation of the radiation-induced DNA damage response4,5. But all identified capabilities or hallmarks of cancer cells can help explain the radiobiological response of tumours3. As a result, the signalling pathways known to regulate several cancer hallmarks, such as p53 and Notch6, may be key to the regulation of radioresistant cancer cells fate.

While the role of p53 in the increased survival of prostate cancer cells to fractionated radiation7, increased cell proliferation8 and treatment outcomes9 in patients with locally recurrent prostate carcinoma after radiation therapy has been documented, implications of the Notch pathway in the radiation response10 is not reported in prostate tumours. The Notch pathway is implicated in angiogenesis11,12 and has been proposed to facilitate prostatic tumourigenesis13, influence the outcome of anti-cancer hormonal14,15 and docetaxel treatments16 and may be particularly involved in the development of prostate cancer in men with high body mass index17,18. Investigation into the regulation of this pathway indicates a possible cross talk with the YB-1 pathway19,20. YB-1 is a multifunctional protein whose expression increases with prostate cancer progression and is predictive of recurrence following surgery21. It is involved in both the transcriptional and translational regulation of gene expression, and controls almost all DNA and mRNA dependent processes in the cell such as cellular differentiation, proliferation and stress response22.

In prostate cancer, exposure to fractionated radiation progressively selected for a 22Rv1 prostate carcinoma cell population enriched in S-phase cells, less susceptible to DNA damage, radiation-induced apoptosis and acquired enhanced migration potential, when compared to wild type and aged-matched control 22Rv1 cells23. These enhanced radioprotective oncogenic properties, also observed in isogenic models of other disease sites3, were associated with an altered miRNA profile common to that of 22Rv1 cells exposed to hypoxia, a known factor associated with radioresistance24,25. This study aimed to further establish the clinical relevance of the model and identify candidate markers of radioresistance for this disease. Ninety proteins associated with the cancer hallmarks, the Notch and the YB-pathways were selected to generate a custom multiplex protein expression profile of radioresistant (RR-22Rv1) and radiosensitive (WT-22Rv1) isogenic prostate cancer cells. Independent validation of differentially expressed PARP-1, p53 and the androgen receptor strengthens the clinical relevance of the model and suggests a role for the Notch-3 intracellular domain (N3ICD) in the radioresponse of these cells. Pilot analysis in pre-treatment biopsies of prostate cancer patients treated with radiation therapy for the first time implicates the YB-1 protein in treatment failure.

Results

Radiation response of 22Rv1 isogenic cells

The change in the radiation response of 22Rv1 cells exposed to 30 × 2Gy- dose fractions (RR-22Rv1), compared to age-matched (AMC-22Rv1) and wild type (WT-22Rv1) cells was confirmed using clonogenic assays. The clonogenic survival of each cell line treated with a 4 Gy single dose and their corresponding unirradiated controls is presented in Fig. 1. With a mean survival of 26.4% ± 0.01, RR-22Rv1 cells were significantly more radioresistant than both AMC-22Rv1 (18.4% ± 0.01) and WT-22Rv1 (10.31% ± 0.01) cells. AMC-22Rv1 showed a non-significant trend towards increased radioresistance, when compared to WT-22Rv1 cells.

Figure 1
figure 1

Radiation response of isogenic 22Rv1 cells. The clonogenic survival of wild type (WT), age-matched controls (AMC) and radioresistant (RR) 22Rv1 prostate cancer cells following a single dose of 4 Gy radiation is presented. N = 4; p < 0.05.

Full size image

Protein profile of the 22Rv1 cells panel

The protein profiles of the isogenic radioresistant cell line (RR-22Rv1) and its parent WT-22Rv1 cell lines were generated for a custom 90 proteins (Fig. 2). Comparison of these profiles identified significant modifications in expression levels for 23 proteins (14 up, 9 down), non-significant trends for 41 (20 up, 21 down) (Table 1), and failed protein detection for 23 (Table 2). The upregulated proteins were associated with the promotion of cell survival, proliferation and invasion (p53, ATR, FKBP12, Bak, Bcl-xL, Beclin-1, Calveolin-1, plasmamix Na+/K+ -ATPase, Claudin-1, Chk1-p-Ser296), the activation of Notch-3 (Notch3 intracellular domain (N3ICD), HES-1) and YB-1 (YB-1, YB-1-p-Ser1-2) signalling. The down regulated proteins were associated with genome instability (plasmamix Histon H3 (diMeth K9), Histone H3), the regulation of tumour growth (Androgen receptor, beta-catenin, Bim, Shh), DNA repair (PARP-1) and Notch signaling (Notch 1 intracellular domain (N1ICD), RBPSUH).

Figure 2
figure 2

DigiWest protein profiling of wildtype (WT) and radioresistant (RR) 22Rv1 cell lines. (A) Typical representation of DigiWest raw data showing signals for each molecular weight group. Peak signals at the correct molecular weight are marked and quantified in Relative Fluorescence Units (RFU). (B) Heat map of Log2 transformed DigiWest data in WT and RR cell lines. Hierarchical Clustering was performed using complete linkage and Eucledian distance, applying the Multi Experiment Viewer (MeV) software. (C) Bar graphs for DigiWest data from selected proteins in WT vs RR cells. (D) Representation of DigiWest data as Western Blot mimics.

Full size image
Table 1 Differential protein expression patterns in RR-22Rv1 compared to WT-22Rv1 cell samples.
Full size table
Table 2 List of analysed Proteins that were below detection levels.
Full size table

Determination of the expression levels of PARP-1 and the androgen receptor in independent sets of WT-22Rv1, AMC-22Rv1 and RR-22Rv1 validated the decreased expression of these two proteins in cells with a radioresistant phenotype (Fig. 3).

Figure 3
figure 3

Representative cropped immunoblots of wild type (WT), age-matched controls (AMC) and radioresistant (RR) 22Rv1 prostate cancer cells. The detection levels of (A) PARP-1 in whole cell lysates, and (B) the androgen receptor (AR) in the nuclear (N) and cytoplasmic (C) protein fractions are presented. GAPDH or Lamin B1 were used as cytoplasmic and nuclear loading controls respectively. The full length blots are presented in Supplementary Figs. 1A,B and 2A,B.

Full size image

Validation of the p53 expression profile

The expression levels of the p53 protein were confirmed to increase in radioresistant compared to wild type cells in independent cell lysates (Fig. 4A). The expression levels of the p53 pathway regulatory miRNAs, miR-141 and miR-1285 were measured by qPCR in the cell lines. While expression levels of miR-141 were significantly downregulated in RR-22Rv1 and AMC-22Rv1 compared to WT-22Rv1 cells, no significant change in miR-1285 was observed (Fig. 4). The commercially available p53 inhibitor Pifithrin-α (PFT-α) was next evaluated for its ability to affect the clonogenic survival of the cell lines. WT and RR-22Rv1 cells were treated with two separate concentrations of PFT-α (20 µM, 40 µM, 2 h) alone or in combination with radiation (4 Gy single dose) (Fig. 4). The surviving fractions of WT-22Rv1 cells treated with either 20 µM (38.7 ± 0.04) or 40 µM (37.5% ± 0.03) PFT-α alone were significantly higher than those of cells treated with radiation alone (10.31% ± 0.01) (P < 0.05). The combination of PFT-α and 4 Gy resulted in a significant reduction in surviving fraction when compared to PFT-α alone at both 20 µM (12.2% ± 0.02) and 40 µM (5.87% ± 0.01) (p < 0.05). The surviving fraction of cells exposed to the combined treatment was significantly reduced at 40 µM (p < 0.05), indicating a very small radiosensitising effect of PFT-α at 40 µM. RR-22Rv1 cells were significantly more sensitive to PFT-α than WT-22Rv1 at both concentrations alone or in combination with radiation (p < 0.05). The surviving fraction of RR-22Rv1 cells (22% ± 0.1) treated with either PFT-α concentration or 4 Gy alone were not significantly different. In response to the combination of PFT-α and 4 Gy, a very small but significant radiosensitising effect was observed as a reduction in clonogenic survival at both concentrations (p < 0.05).

Figure 4
figure 4

Evaluation of p53 as a marker of radioresistance. (A) Representative cropped immunoblots for the detection of p53 in wild type (WT), age-matched controls (AMC) and radioresistant (RR) 22Rv1 prostate cancer cells. GAPDH was used as the loading control. Full length blots are presented in Supplementary Fig. 3A,B. (B) miR-141 and miR-1285 expression profiles in WT-, AMC- and RR-22Rv1 prostate cancer cell lines was determined by qPCR. MiR-141 and miR-1285 expression levels for each cell line were normalised to the endogenous control miR-16 and calibrated to that of WT-22Rv1 cells to establish the relative change in microRNA expression (% miR-expression). Results are expressed as mean +/− SD, N = 3. Clonogenic survival of (C) WT-22Rv1 and (D) RR-22Rv1 treated with Pifithrin-α (20 µM, 40 µM, 2hrs) alone or in combination with a single dose of 4 Gy radiation. The clonogenic survival of untreated cells was used as a control. UT: untreated controls, *p < 0.05.

Full size image

Validation of the Notch-3 expression profile

The independent validation of the expression of the Notch-3 receptor indicated strong expression of the cleaved intracellular domain of the Notch-3 receptor (N3ICD) in WT-22Rv1, AMC-22Rv1 and RR-22Rv1 cells. The expression of the full-length receptor appeared to increase in AMC-22Rv1 and RR-22Rv1 cells (Fig. 5A). Evaluation of the cellular localization of N3ICD following radiation exposure (5 Gy) identified a loss of expression in irradiated WT-22Rv1 prostate cancer cells, but maintained HES-1 expression (Fig. 5B). Treatment of WT-22Rv1 and RR-22Rv1 cells with two commercially available Notch inhibitors 3,5-Difluorophenacetyl)-L-alanyl]-S-phenylglycine t-butyl ester (DAPT) and Batimastat (60 µM) failed to significantly decrease clonogenic survival when compared to untreated controls. When the inhibitors were combined with a single 4 Gy radiation dose, a mild radiosensitising effect was observed in RR-22Rv1 treated with DAPT, with a reduction in clonogenic survival from 28.3% ± 4.4 (radiation alone) to 11.1% ± 1.1 (p < 0.05) (Fig. 5C).

Figure 5
figure 5

Evaluation of Notch-3 as a marker of radioresistance. (A) Representative cropped immunoblots of Notch-3 and its intracellular domain in wild type (WT), age-matched controls (AMC) and radioresistant (RR) 22Rv1 prostate cancer cells. The full length blots are presented in Supplementary Fig. 4. (B) The detection levels of the Notch-3 intracellular domain (N3ICD) and HES-1 in the protein fractions of untreated and 5Gy-irradiated WT-22Rv1 cells are presented. Actin or Tubulin were used as loading controls. The full blots are presented in Supplementary Fig. 5A–C. (C) Clonogenic survival of RR-22Rv1-cells treated with the Notch-inhibitors Batimastat or DAPT (60 µM, 24hrs), alone or in a combination with a 4 Gy single radiation dose. *p < 0.05.

Full size image

Validation of the YB-1 expression profile

The independent validation of the expression of the YB-1 protein indicated strong cytoplasmic and nuclear expression in all three cell lines, with a slight increase in nuclear expression following 4Gy- irradiation (Fig. 6A). Immunostaining for YB-1 was next performed in a series of 22 formalin-fixed parafilm-embedded biopsy specimens from prostate cancer patients. Nuclear expression was detected in 3 out of 4 specimen obtained from patients diagnosed with biochemical failure following radiation therapy. Nuclear expression was not detected in any biopsy specimens from the patients in remission five years post treatment (n = 18) (Fig. 7). Treatment of RR-22Rv1 cells with the YB-1 inhibitor Fisetin (60 µM, 24 h) resulted in a significant reduction in surviving fraction in unirradiated and 4-Gy irradiated cells (Fig. 6D). The combination of Fisetin with radiation was associated with a reduction in clonogenic survival from 28.4% ± 4.4 (radiation alone) to 8.5% ± 5.1 (p < 0.05).

Figure 6
figure 6

Evaluation of YB-1 as a marker of radioresistance. (A) Representative cropped immunoblots of YB-1 in the cytoplasmic (C) and nuclear (N) protein fractions of wild type (WT), age-matched controls (AMC) and radioresistant (RR) 22Rv1 prostate cancer cells which received a single dose of 4 Gy radiation or were left untreated. Vinculin and Lamin B1 were used as cytoplasmic and nuclear loading controls respectively. The full length blots are presented in Supplementary Fig. 6A–C. (B) Clonogenic survival of unirradiated and 4-Gy irradiated RR-22Rv1 cells treated with the YB-1 inhibitor Fisetin (60 μM, 24hrs). *p < 0.05.

Full size image
Figure 7
figure 7

Representative images of positive and negative YB1 staining. Top left - Colon control tissue stained without primary antibody. Top Right - Colon control tissue displaying positive staining. Centre Left - WT-22RV1 cells exhibiting predominantly cytoplasmic staining for YB1. Centre right - RR-22RV1 cells which display cytoplasmic and enhanced nuclear YB1 staining. Bottom left - A pre-treatment prostate biopsy from a patient who subsequently underwent radiation therapy and exhibited a favourable response which display predominantly cytoplasmic YB1 staining. Bottom right - A biopsy from a non-responder/radioresistant patient displaying enhanced nuclear YB1 staining. Images were taken at 40X magnification using a Leica DM 3000 Microscope and Leica DFC 425 camera.

Full size image

Discussion

The effects of radiation exposure on cancer cells and subsequent regulation of cell fate has been linked to DNA damage induction4,5, all known cancer hallmarks3, and the tumour microenvironment26. This study proposed to compare the protein expression profiles for a selection of signaling proteins involved in these biological processes in a pre-existing isogenic model of 22Rv1 prostate cancer cells23. High content multiplex protein profiling identified twenty-three proteins within the selected panel whose expression was significantly modified in the radioresistant line, compared to the wild type cell line.

The DNA repair protein PARP-1 was independently identified as one of the protein whose expression was down-regulated in radioresistant cells, alongside the androgen receptor (AR). PARP-1 has been implicated in epithelial-mesenchymal-transition (EMT) during prostate tumorigenesis27, and its elevated expression was associated with poor prognosis following radiotherapy28. PARP-1 inhibition appears most effective in tumours with elevated PARP-1, but reduced AR and p53 expression levels29,30. Our data support these findings and suggest that tumour recurrence following exposure to fractionated radiation may result from the selection of cells with reduced PARP-1 and AR as well as elevated p53 expression. Further evaluation of this molecular combination in radiotherapy patients is warranted.

The expression of p53 was increased in RR-22Rv1 cells and correlated with the downregulation of miR-141. Reports of the upregulation of miR-141 in prostate cancer suggest its involvement in the metastatic process31, the regulation of AR activity32 and modifications in cell proliferation rates and apoptosis susceptibility33, but no previous reports have shown evidence of its involvement in the response to radiation therapy. Treatment of RR-22Rv1 cells with PFT-α, a known suppressor of p53 as well as glucocorticoid and heat shock receptor factor34 was associated with a very small but significant radiosensitisation effect. This is the first report of a radiosensitisation effect of this compound in prostate cancer and further evaluation of the magnitude of radiosensitisation that may be achieved is warranted.

Our data identifies the up regulation of the intracellular domain of the Notch-3 receptor in RR-22Rv1, which correlates with a mild radiosensitisation effect following exposure to the Notch inhibitor DAPT. Notch inhibitors are increasingly explored as novel anti-cancer strategies35,36,37 and their potential as radiosensitisers has been proposed10. We38,39, like others40, have previously reported the upregulation of the Notch-3 receptor in prostate cancer, and its role in hypoxic tumours41. Although Notch activity has been implicated in the radiation-induced stress response of prostate cancer cells42, no such strategy exists in prostate cancer. Exposure to a single 5Gy radiation dose led to undetectable expression of the intracellular domain of the Notch-3 receptor in wild type 22Rv1 cells. This loss did not correlate with a reduction in HES-1 expression, a known downstream target of the Notch pathway. This data indicates that the regulation of HES-1 is not dependent on Notch-3 activation. In prostate tumours, HES-1 elevated expression was also related to Notch-1 receptor activation43. In prostate cancer cells, while the selective genetic manipulation of members of the pathways (Notch-144) impaired differentiation and tumour growth45,46, the pharmacological inhibition of the pathway with gamma-secretase inhibitors only caused mild modifications47,48. Examples of Notch targeted radiosensitising strategies are beginning to emerge in lung cancer49 and glioblastoma42,50.

Finally, the comparative analysis of these protein profiles identifies upregulation of the YB-1 protein. Nuclear localization of YB-1 has been proposed as an indicator of poor prognosis51, with high protein levels of YB-1 and MTA1 associated with a 3-fold increased risk for requiring future hormone therapy or radiation therapy in prostate cancer21. Our data confirmed detection of the protein in the cytoplasm and nucleus of all three cell lines, and a slight elevation in the RR-22Rv1 cells. Exposure of RR-22Rv1 cells to a 4 Gy single radiation dose resulted in a modest increase in nuclear YB-1 expression levels. Pilot analysis identified elevated nuclear expression in 3 out of 4 formalin-fixed-parafilm-embedded biopsy specimen from high risk radiotherapy prostate cancer patients whose tumours progressed post-treatment and none in those with no evidence of residual disease at five years. The inhibition of YB-1 by Fisetin was associated with microtubule stabilising properties, an ability to reverse mesenchymal features and inhibit cell proliferation, migration, and invasion52. Treatment of RR-22Rv1 with this dietary flanovol resulted in a significant reduction in the clonogenic survival of irradiated cells. This is the first report of the radiosensitisation potential of this YB-1 inhibitor.

In conclusion, the exposure of 22Rv1 prostate cancer cells to fractionated radiation and emergence of a radioresistant cell line is associated with the deregulation of signalling proteins associated with more aggressive disease and poor prognosis following radiotherapy. This work supports the clinical relevance of this model and identifies actionable markers of the radioresistant phenotype of prostate cancer.

Materials and Methods

Cell culture

Newly acquired, authenticated human 22Rv1 (WT-22Rv1) prostate cancer cells (American Type Culture Collection) were cultured at 37 °C in 95% humidified air containing 5% CO2 in RPMI cell culture medium containing l-glutamine (Lonza, Dublin, Ireland) with 10% foetal bovine serum (Gibco, Dublin, Ireland) and 1% pen/strep (Lonza). Fractionated 2 Gy X-rays doses (250 keV, 15 mA) using an RS225 cabinet (XStrahl, Surrey, UK) were delivered weekly to a cumulative dose of 60 Gy to generate the RR-22RV1 cell model, as described in23. Mock irradiated cells were cultured alongside to generate age-matched controls cells (AMC-22Rv1)23.

Clonogenic assays

Cell survival was evaluated using a standard colony forming assay53. Cells were treated with Pifithrin-α (Sigma Aldrich, Dublin, Ireland) at concentrations of 20 µM and 40 µM for 2 hours; batimastat, DAPT or Fisetin (Sigma Aldrich) at a concentration of 60 µM for 24 hours. The cells were then irradiated at a single dose of 4 Gy and incubated for 7–14 days until formation of colonies. The colonies washed with PBS, were fixed and stained using a 0.05% Crystal Violet (Sigma Aldrich) in 70% Methanol solution. Individual colonies (defined as a cluster of greater than 50 cells) were counted using the ColCount software (Oxford Optronix Ltd, Abingdon,UK).

DigiWest multiplex protein profiling

DigiWest analyses were performed as previously described54. Briefly, gel electrophoresis and Western blotting onto PVDF membranes was conducted using the NuPAGE system (Life Technologies, Carlsbad, USA). Blots were washed in PBST and proteins were biotinylated on the membrane using 50 µM NHS-PEG12-Biotin in PBST for 1 h, then washed in PBST and dried. Each sample lane was cut into 96 molecular weight fractions of 0.5 mm each and proteins were eluted in 96 well plates using per well 10 µl elution buffer (8 M urea, 1% Triton-X100 in 100 mM Tris-HCl pH 9.5). Eluted proteins from each molecular weight fraction were loaded onto one distinct color of neutravidin coated MagPlex beads (Luminex, Austin, TX, USA) and then pooled. A total of 10 µg of protein per cell sample was used for generation of 90 different antibody incubations.

Aliquots of the DigiWest bead mixes were added to 96 well plates containing 50 µl assay buffer (Blocking Reagent for ELISA (Roche, Rotkreuz, Switzerland) supplemented with 0.2% milk powder, 0.05% Tween-20 and 0.02% sodium azide). After discarding the assay buffer, 30 µl of diluted primary antibody was added per well. Following overnight incubation at 15 °C on a shaker, bead mixes were washed twice with PBST, and Phycoerythrin-labelled secondary antibodies (diluted in assay buffer) were added and incubated for 1 h at 23 °C. Beads were washed twice prior to readout on a Luminex FlexMAP 3D.

For quantification of antibody specific signals, the DigiWest analysis tool described in54 was used to identify peaks and to calculate peak areas. For comparative analyses, protein expression values were normalized to the total protein per lane.

miRNA expression analysis

Total RNA was isolated as per the manufacturer’s instructions using the mirVana™ miRNA Isolation Kit (Applied Biosystems, USA). RNA concentration was determined using the nanodrop 1000 spectrophotometer. Reverse transcription was carried out using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems) on the Gene Amp PCR System 9600 (Perkin Elmer). TaqMan RT-PCR was then performed in 384 well plates using the 7900HT Real-Time PCR System (Applied Biosystems) using primers and probes for miR-141, miR-1285 and miR-16 and TaqMan™ Universal PCR Master Mix, no AmpErase™ UNG (Thermo Fisher Scientific) mixed together with amplified cDNA samples as per the manufacturer’s protocols. MiRNA expression levels were calculated using the ΔΔCT method55 relative to the endogenous control, miR-16. A change in miRNA expression was considered significant if at least a 2-fold change (above 200% expression or below 50% expression) was observed, with a p value of ≤0.05 compared to WT-22Rv1 cells.

Western blotting

Proteins were extracted from cells using either RIPA Lysis buffer (Santa Cruz, USA) or the NE-PER Nuclear and Cytoplasmic Extraction kit (Thermo Fisher Scientific, USA) modified with phenylmethanesulfonyl fluoride (PMSF) (200 mM), a protease inhibitor cocktail, and sodium orthovanadate (100 mM), according to the manufacturer’s instructions Protein concentration was determined using the Pierce™ BCA Protein Assay Kit ((Thermo Fisher Scientific). 20 µg of each protein sample was then resolved by SDS-PAGE on a 5% stacking gel and 10% resolving gel using a Mini-Protean Tetra Cell electrophoresis rig (BioRad, USA). Resolved proteins were then transferred to 0.2 µM Hybond PVDF membranes (Amersham, UK) using the XCell II™ Blot module (Invitrogen, Dublin, Ireland). Following transfer membranes were blocked using 5% w/v milk protein and probed using primary antibodies directed against PARP1 (#sc-8007; 1:100, Santa Cruz, UK56), AR (#sc-377546, 1:250, Santa Cruz), p53 (#sc-126, 1:500, Santa Cruz, UK57), Notch-3 (#ab23426, 1:250, Abcam, UK58), HES-1 (#ab71559, 1:1000, Abcam59) or YB-1 (#ab12148, 1:2000, Abcam60). After washing, the membrane was incubated with horseradish peroxidase conjugated goat secondary antibody (sc-2020, 1:1000, Santa Cruz61) or Anti-Rabbit IgG HRP Linked (#7074, 1:1000, Cell Signalling62) antibodies. Following incubation with the primary and secondary antibodies, a detection reagent luminol (SC-2048, Santa Cruz) was applied to blots and chemiluminescence images were then developed using a Fujifilm LAS-4000 luminescent image analyser. Molecular weight was confirmed using the MagicMark™ XP Western Protein Standard (LC5602, Thermo Fisher Scientific, Dublin, Ireland). Membranes were stripped prior to reprobing with either anti-actin (1:10000, Sigma-Aldrich63), anti-GAPDH (#ab9485, 1:10,000, Abcam63), Lamin (#GTX103292, 1:1,000, Genetex, UK64) or Vinculin (#GTX113294, 1:1,000, Genetex64).

Patient specimens

Diagnostic biopsies (n = 22) were obtained retrospectively with informed consent from high-risk prostate cancer patients with a primary Gleason score of 8–10 who subsequently underwent androgen deprivation therapy and external beam radiation therapy. Ethical approval was obtained from St James’s/Tallaght Hospital and the St Luke’s Radiation Oncology Network, Dublin, Ireland. All research was performed in accordance with relevant guidelines/regulations. Four patients were diagnosed with biochemical failure and disease recurrence, according to the Phoenix criteria. The remaining patients (n = 18) were in remission at 5 years post- treatment.

Immunohistochemistry

4 µM sections were cut from the 22 selected FFPE patient tissue blocks and were baked onto charged slides overnight in an oven at 65 °C prior to storage until needed. Slides were stained using the Rabbit specific HRP/DAB (ABC) Detection IHC Kit (Abcam, ab64261) and a rabbit polyclonal YB-1 antibody (Abcam, ab12148) diluted 1:100 in PBS. Immunostained sections were blindly examined by two trained histopathologists (SF & DC). Images were taken using a Nikon Eclipse E200 microscope and OC view 7 software.

Statistical analysis

A Student’s t-test or one way ANOVA was used to compare means between treatment and control groups as appropriate. Statistical analysis was performed using Prism, Version 5.01 (GraphPad Software Inc. CA). A p-value of <0.05 was considered statistically significant.