Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 8 (12), 18949-18967

Therapy-induced Developmental Reprogramming of Prostate Cancer Cells and Acquired Therapy Resistance


Therapy-induced Developmental Reprogramming of Prostate Cancer Cells and Acquired Therapy Resistance

Mannan Nouri et al. Oncotarget.


Treatment-induced neuroendocrine transdifferentiation (NEtD) complicates therapies for metastatic prostate cancer (PCa). Based on evidence that PCa cells can transdifferentiate to other neuroectodermally-derived cell lineages in vitro, we proposed that NEtD requires first an intermediary reprogramming to metastable cancer stem-like cells (CSCs) of a neural class and we demonstrate that several different AR+/PSA+ PCa cell lines were efficiently reprogrammed to, maintained and propagated as CSCs by growth in androgen-free neural/neural crest (N/NC) stem medium. Such reprogrammed cells lost features of prostate differentiation; gained features of N/NC stem cells and tumor-initiating potential; were resistant to androgen signaling inhibition; and acquired an invasive phenotype in vitro and in vivo. When placed back into serum-containing mediums, reprogrammed cells could be re-differentiated to N-/NC-derived cell lineages or return back to an AR+ prostate-like state. Once returned, the AR+ cells were resistant to androgen signaling inhibition. Acute androgen deprivation or anti-androgen treatment in serum-containing medium led to the transient appearance of a sub-population of cells with similar characteristics. Finally, a 132 gene signature derived from reprogrammed PCa cell lines distinguished tumors from PCa patients with adverse outcomes. This model may explain neural manifestations of PCa associated with lethal disease. The metastable nature of the reprogrammed stem-like PCa cells suggests that cycles of PCa cell reprogramming followed by re-differentiation may support disease progression and therapeutic resistance. The ability of a gene signature from reprogrammed PCa cells to identify tumors from patients with metastasis or PCa-specific mortality implies that developmental reprogramming is linked to aggressive tumor behaviors.

Keywords: cancer stem cell; hormone resistance; neural crest; neuroendocrine transdifferentiation; prostate cancer.

Conflict of interest statement


The authors declare no conflicts of interest.


Figure 1
Figure 1. Developmental reprogramming of PCa Cells to a stem-like intermediate
(A) Chronic AD of LNCaP cells upregulated genes associated with N/NC stem cells and derivative tissues. LNCaP cells (Left) cultured in androgen-depleted medium for 15-days undergo a morphological transformation (Middle), and overexpress genes associated with N/NC stem cells and derivative tissues shown on heat-mapping (Right). (B) Representative phase-contrast images of LNCaP cells over two weeks of culture in STM. (C) Representative phase-contrast images of VCaP, LAPC4 or 22Rv1 spheroids after 14-day STM-mediated reprogramming. Scale bars represent 100 μm. (D) Proliferation assays of parental LNCaP (P) and reprogrammed stem-like LNCaP cells (SL) showed that the doubling time of LNCaP-SL is significantly greater than parentals. Mean ± standard error; **P < 0.01. (E) Clone formation of dispersed LNCaP cells grown in FBS or STM medium indicates that reprogramming occurs in at least 57% of LNCaP cells plated. Mean ± standard deviation; ***P < 0.001. (F) The percentage of spheroids (diameters greater than 100 μm) is significantly greater in STM cultured LNCaP cells compared to FBS cultured LNCaP cells, over 14 days. Mean ± standard deviation shown; ****P < 0.0001. (G) Spheroid colony formation assay of LNCaP cells cultured in FBS, CSS or STM on attachment-free plates shows advantage of STM compared to standard culture medium (RPMI) with (FBS) or without androgen (CSS). Mean ± standard deviation; ***P < 0.001; ****P < 0.0001.
Figure 2
Figure 2. Androgen Receptor axis down-regulation in developmentally reprogrammed PCa cells
(A) AR and PSA mRNA expression, measured by RT-qPCR, is down-regulated in reprogrammed (SL) PCa cells compared to parentals. Bars represent relative expressions compared to parental cell lines. Mean ± standard error; **P < 0.01; ***P < 0.001. (B) Secretion of PSA into cell medium was significantly downregulated in LNCaP-SL cells over 48 hours compared to parental cells. Mean ± standard deviation; ***P < 0.001. (C) Western Blot analyses of parental and reprogrammed PCa cells showed down-regulation of AR and AR-variant protein. (D) Relative growth inhibition (7-days) of LAPC4, LNCaP and VCaP parental and stem-like cells with increasing dosages of enzalutamide. Mean ± standard deviation; *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.
Figure 3
Figure 3. Stem cell-like features of developmentally reprogrammed PCa cells
(A) Expression of mRNAs for OCT3/4, SOX2 and NANOG is increased in reprogrammed PCa cells. Mean ± standard error; *P < 0.05; **P < 0.01; ***P < 0.001. (B) Western Blot analyses showed overexpression of BMI1 and phosphorylated-AKT protein in reprogrammed PCa cells. (C) Sphere-forming assays indicate significantly smaller spheroid formation of LNCaP cells during developmental reprogramming when AKT and BMI1 pathways are targeted with inhibitors LY294002, MK2206, or PTC209. Mean ± standard deviation; ****P < 0.0001. (D) Ingenuity analyses of genes differentially expressed (p < 0.05, fold-change > 2.0) in reprogrammed LNCaP, VCaP, and LAPC4 cells indicate that STM-reprogramming induces a significant enrichment of cellular functions related to cancer stem cell phenotypes. (E) Gene Set Enrichment Analyses (GSEA) of genes differentially expressed (p < 0.05, fold-change > 1.5) in reprogrammed LNCaP, VCaP, and LAPC4 indicate reprogramming is inversely correlated with Androgen Response Genes and significantly enriched in Prostate Cancer-Down Genes (Liu dataset [56]), and Stem Cell Genes (Boquest and Lim datasets [25, 26]).
Figure 4
Figure 4. Developmentally reprogrammed PCa cells have characteristics of NC stem cells
(A) Immunofluorescence staining for NESTIN, a neuroectodermal stem cell marker, in parental and developmentally reprogrammed LNCaP cells. (B) Immunofluorescence staining of BRN3A, a neuronal transcription factor, in parental and developmentally reprogrammed LNCaP cells. Scale bars represent 20 μm. (C) qPCR analyses indicate that reprogrammed stem-like PCa cells overexpress genes associated with NC stem cells compared to parental cells. Mean ± standard error; *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001. (D) GSEA shows significant enrichment of NC stem cell genes in the Kreitzer and Lee neural crest stem cell datasets [27, 28] in reprogrammed LNCaP cells. (E) FACS analyses indicate that reprogramming increases the population of CD29High/CD15 cells. (F) FACS profiling of CD29Moderate and CD29High LNCaP-SL cells showed that reprogrammed CD29High cells overexpress CD271, NRCaM, CD166 and CD57, surface markers of neural crest stem cells. (G) FACS analyses demonstrate that expression of CD133 and CD44 remain unchanged after developmental reprogramming in LNCaP cells.
Figure 5
Figure 5. Reprogrammed PCa cells can differentiate to N/NC-derived cell lineages
(AC) Morphological features and expression of cell lineage biomarkers in LNCaP-SL cells cultured in neuronal (A), oligodendrocyte (B) or osteoblast (C) differentiation mediums (Diff Media) for 14 days, indicating the acquisition of relevant morphology (Left), mRNA expressions (Middle) or protein expressions by IF (Right) indicative of neurons, oligodendrocytes or osteoblasts by LNCaP-SL cells. Mean ± standard error; *P < 0.05; **P < 0.01; ***P < 0.001.
Figure 6
Figure 6. Reprogrammed PCa cells possess increased invasive/metastatic abilities
(A) GSEA indicates a significant enrichment of EMT genes in LNCaP-SL cells. (B) qPCR analyses show that reprogrammed LNCaP-SL cells overexpress mRNA of key EMT markers. Mean ± standard error; *P < 0.05; **P < 0.01; ***P < 0.001. (C) Western Blot demonstrates loss of E-Cadherin in reprogrammed PCa cells indicating increased EMT phenotype. (D) Reprogrammed LNCaP-SL cells demonstrate increased invasion through a Matrigel-coated membrane in in vitro invasion assays compared to parental LNCaP cells. Mean ± standard error; *P < 0.05. (E) Reprogrammed stem-like PCa cells are more invasive/metastatic in zebrafish assays (Left). Representative photomicrographs of VCaP and VCaP-SL xenografted fish illustrates dispersal of fluorescently-labelled VCaP-SL cells within the caudal region (Right).
Figure 7
Figure 7. Reprogramming in vitro by androgen-deprivation and acquired therapy resistance
(A) LNCaP-SL cells re-differentiate to neuroendocrine-like cells within 5-days in RPMI with CSS. (B) LNCaP-SL cells regain the typical morphology of LNCaP cells within 5-days in RPMI with FBS. Scale bars represents 100 μm. (C) Western Blot shows that re-differentiated LNCaP cells regain AR and E-Cadherin expression while retaining higher phosphorylated-AKT and BMI1 protein expression. (D) Loss of growth inhibition when LNCaP-R6 cells were grown in RPMI with CSS medium compared to parental LNCaP demonstrates acquired resistance to androgen-deprivation. (E) Loss of growth inhibition of LNCaP-R6 cells grown in RPMI with FBS supplemented with 1 μM enzalutamide compared to parental cells demonstrates acquired resistance to enzalutamide. Mean ± standard error; *P < 0.05; **P < 0.01; ***P < 0.001. (F) FACS analyses shows the increase of CD29SuperHigh populations in reprogrammed LNCaP cells and in parental LNCaP cultured in RPMI with CSS (P+CSS) or RPMI with FBS and 10 μM enzalutamide (P+Enz) for 3 days, LNCaP androgen-independent cells (AI) cultured in CSS (AI) or in the presence of 10 μM enzalutamide (AI+Enz). Mean ± standard error; *P < 0.05; **P < 0.01; ***P < 0.001.
Figure 8
Figure 8. Reprogramming in vivo by androgen-deprivation and acquired therapy resistance
(AB) Genes overexpressed in reprogrammed LNCaP cells are positively enriched in the transcriptomes of LNCaP xenografts after castration of hosts. GSEA-determined normalized enrichment scores shows significant enrichment of stem-like overexpressed genes in RNAs from regressing LNCaP xenografts after castration of the host (n = 6) and at PSA nadir (n = 9) whereas these genes were not enriched in xenografts prior to castration (n = 11) or in recurrent castrate-resistant tumors (n = 12). Mean ± standard error. (C) The Normalized Enrichment Score (NES) of LNCaP-SL cells compared to LNCaP xenografts before and after castration is inversely and significantly correlated with PSA levels in the blood of murine xenograft hosts.
Figure 9
Figure 9. Reprogrammed PCa cells share a gene signature that correlates with adverse outcomes in patients
(A) Venn Diagram of genes differentially expressed (p < 0.05, fold-change > 2) in three reprogrammed PCa cell lines share a common 132 gene signature. (B) Heatmap of the top 90 differentially expressed genes (p < 0.05, fold-change > 2) in reprogrammed LNCaP, VCaP and LAPC4 PCa cell lines. The complete gene list is found in Supplementary Table 2. (CF) Enrichment of the 132 gene reprogrammed PCa cell signature was assessed in the MCII dataset (see supplementary data for results in other patient cohorts) and correlated with PCa patient clinical outcomes, including biochemical recurrence (BCR), development of metastasis (METS) and Prostate Cancer Specific Mortality (PCSM). (C) Patient annotation matrix and waterfall plots showing the 132 gene signature score in waterfall plots above a patient annotation matrix. Each column of the waterfall plot relates to a column in the annotation matrix and represents one patient. The rows of the annotation matrix indicate the events which the patients experienced. (D) The 132 gene signature score is significantly higher (p = 2.55 × 10−5, AUC = 0.66) in patients with BCR and a high gene signature score (green line) in primary tumors predicts BCR (p = 0.046). (E) The 132 gene signature score is significantly higher (p = 3.97 × 10−8, AUC=0.72) in patients with METS and a high gene signature score (green line) in primary tumors predicts the development of metastases (p = 1.74 × 10−6). (F) The 132 gene signature score is significantly correlated (p = 4.98 × 10−4, AUC = 0.69) to prostate cancer specific mortality and a high gene signature score (green line) in primary tumors is significantly correlated with poor prognosis (p = 7.93 × 10−4).
Figure 10
Figure 10. Schematic pathway for the neural/neural crest stem cell-mediated transdifferentiation of prostate cancer cells
Based on the outcomes of this study, we propose a model whereby androgen receptor pathway inhibition triggers a developmental reprogramming process that results in a metastable cancer stem-like state that resembles neural/neural crest stem cells. These N/NC stem-like cells may acquire features of differentiated N/NC derived cell lineages when placed into cell-specific differentiation mediums. Depending on the cellular microenvironment, these developmentally reprogrammed cells may transdifferentiate to neuroendocrine-like or other therapy resistant PCa cell types.

Similar articles

See all similar articles

Cited by 7 PubMed Central articles

See all "Cited by" articles


    1. Alva A, Hussain M. The changing natural history of metastatic prostate cancer. Cancer J. 2013;19:19–24. - PubMed
    1. Ferraldeschi R, Welti J, Luo J, Attard G, de Bono JS. Targeting the androgen receptor pathway in castration-resistant prostate cancer: progresses and prospects. Oncogene. 2015;34:1745–1757. - PMC - PubMed
    1. Ware KE, Garcia-Blanco MA, Armstrong AJ, Dehm SM. Biologic and clinical significance of androgen receptor variants in castration resistant prostate cancer. Endocr Relat Cancer. 2014;21:T87–T103. - PMC - PubMed
    1. Beltran H, Tomlins S, Aparicio A, Arora V, Rickman D, Ayala G, Huang J, True L, Gleave ME, Soule H, Logothetis C, Rubin MA. Aggressive variants of castration-resistant prostate cancer. Clin Cancer Res. 2014;20:2846–2850. - PMC - PubMed
    1. Hu CD, Choo R, Huang J. Neuroendocrine differentiation in prostate cancer: a mechanism of radioresistance and treatment failure. Front Oncol. 2015;5:90. - PMC - PubMed