Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 8 (6), e68003

Androgen Deprivation-Induced Senescence Promotes Outgrowth of Androgen-Refractory Prostate Cancer Cells


Androgen Deprivation-Induced Senescence Promotes Outgrowth of Androgen-Refractory Prostate Cancer Cells

Dominick G A Burton et al. PLoS One.


Androgen deprivation (AD) is an effective method for initially suppressing prostate cancer (PC) progression. However, androgen-refractory PC cells inevitably emerge from the androgen-responsive tumor, leading to incurable disease. Recent studies have shown AD induces cellular senescence, a phenomenon that is cell-autonomously tumor-suppressive but which confers tumor-promoting adaptations that can facilitate the advent of senescence-resistant malignant cell populations. Because androgen-refractory PC cells emerge clonally from the originally androgen-responsive tumor, we sought to investigate whether AD-induced senescence (ADIS) affects acquisition of androgen-refractory behavior in androgen-responsive LNCaP and LAPC4 prostate cancer cells. We find that repeated exposure of these androgen-responsive cells to senescence-inducing stimuli via cyclic AD leads to the rapid emergence of ADIS-resistant, androgen-refractory cells from the bulk senescent cell population. Our results show that the ADIS phenotype is associated with tumor-promoting traits, notably chemoresistance and enhanced pro-survival mechanisms such as inhibition of p53-mediated cell death, which encourage persistence of the senescent cells. We further find that pharmacologic enforcement of p53/Bax activation via Nutlin-3 prior to establishment of ADIS is required to overcome the associated pro-survival response and preferentially trigger pervasive cell death instead of senescence during AD. Thus our study demonstrates that ADIS promotes outgrowth of androgen-refractory PC cells and is consequently a suboptimal tumor-suppressor response to AD.

Conflict of interest statement

Competing Interests: The authors have declared that no competing interests exist.


Figure 1
Figure 1. Androgen deprivation induces senescence markers in LNCaP cells.
(A) Senescence associated beta-galactosidase (SA-beta-gal) staining under indicated conditions and timepoints. FBS, fetal bovine serum indicating androgen-replete culture; CSS, charcoal-stripped serum indicating androgen- deprived culture. 9D FBS post-SEN denote LNCaP cells that were subjected to 14 days CSS culture and then switched to FBS media for 9 days. Quantitation of positively stained cells is presented to the right of the representative images. (B) Ki67 pan-proliferation marker staining is shown at indicated conditions and time points. DAPI staining is used to mark cell nuclei for purposes of counting Ki67-positive cells. Percentage of stained cells is graphed on the right. (C) Propidium iodide cell cycle analysis, at indicated timepoints and culture conditions. Note the S-phase fraction drops from 15.6% in FBS to 0.4% at 10d CSS and to 0.8% after restoration to FBS media for 9 days (9D FBS post-SEN). (D) Proliferation curve for LNCaP cells cultured in CSS for 3, 8 or 14 days before being switched back to FBS media. (E) Measurement of total ROS levels via flow cytometry in LNCaP cells at day 1, day 4 and day 7 of indicated culture conditions. Note the right-shifted peak in red corresponding to increasing ROS levels in the CSS-cultured LNCaP cells relative to FBS-cultured cells. (F) DNA double-strand break (DSB) foci detected via H2AX/53BP1 co-staining (green = H2AX, red = 53BP1) for LNCaP cells cultured under the indicated conditions. Representative images for cells with the different numbers of counted foci are shown. Note that the percentage of cells with 5+ foci increases with the duration of CSS culture.
Figure 2
Figure 2. ADIS is mediated by the p16INK4a pathway.
(A) Immunoblotting was carried out on approximately 35 µg of the indicated protein samples with antibodies against the denoted proteins. Coomassie blue staining was used to normalize for total protein loading. Note that AR expression decreases as expected upon CSS culture. The trend of p53, p21 and p16 protein expression is maintained following restoration of androgen-replete culture post-senescence (post-SEN) indicating the permanency of the changes. (B) Immunoblotting to show effects of shRNA-mediated knockdown of p16 on ADIS. Note elevated AR and cyclin A levels in shp16 cells under CSS culture relative to control or shp53 knockdown cells. (C) Effect of p16 suppression on SA-beta-gal staining. The indicated cells were cultured in CSS for 14 days and then switched to FBS- containing media for a further 9 days. Note the observed decrease in SA-beta-gal staining in shp16 cells. * p<0.05.
Figure 3
Figure 3. Repeated exposure of LNCaP cells to senescence-inducing stimuli via cyclic AD induces outgrowth of an androgen-refractory subpopulation.
(A) Colony formation in GFP-labeled LNCaP cells and by crystal violet staining. Colonies indicates clones that exist in a transiently arrested state in the bulk population and resume division when androgen-replete conditions are restored after initial CSS culture. The indicated cells underwent one round of switch back to FBS media (SB1) following senescence in CSS culture. Following 21 days in FBS culture, clonal outgrowths were observed via crystal violet staining (∼0.1%). The insets show light microscopy and FITC images of a representative colony that has begun to proliferate in FBS media. (B) Schematic of the switch back (SB) method for generating androgen-refractory population of cells (designated LNCaP-SB5). (C) Proliferation curves for SB0 and SB5 cells in CSS. Cells were harvested on Day 7 and immunoblotted for cyclin A expression (inset) as a proliferation marker. (D) Morphology differences between LNCaP SB0 and SB5 cell populations. (E) Percentage SA-beta-gal positive cells: 87% LNCaP- SB0, 43% LNCaP-SB5. (F) PI cell cycle analysis (S phase: 0.7% LNCaP-SB0, 6.5% LNCaP-SB5). (G) Immunoblotting of 35 µg protein lysates from indicates samples against the denoted proteins. Note that AR expression remains elevated in SB5 versus SB0 cells. Androgen-refractory LNAI cells are shown as a positive control. GAPDH expression is used to normalize for protein loading. (H) Relative proliferation ability of SB5 in different co-cultures under AD. FL1-A indicates intensity of FITC signal and the right hand gate (V2-R) contains GFP-positive (SB5) cells whereas V2-L contains the co-mixed unlabeled cells. Representative flow profiles with the percentages (in red) of each co-mixed population are shown at the indicated time points. (I) Quantitation of GFP-positive (SB5) cell numbers under different co-culture conditions after 7 days of CSS culture are graphed. Quantitation was accomplished via the cell counting parameters from the CFlow software for counterpart cytometric profiles to those shown in (I).
Figure 4
Figure 4. Senescent LNCaP cells display a pro-survival phenotype and upregulate IL-8 mRNA.
(A) Western blot of survival markers in LNCaP cells growing in FBS and CSS day 5, 10 and 17. Cells were cultured under indicated conditions and 50 µg proteins immunoblotted against the indicated antibodies. Note that pro-apoptotic factor Bax is decreased and pro-survival factors MCL-1 and phospho-AKT are increased. Coomassie blue staining is used to normalize for loading. (B) RT-PCR showing IL-8 message levels in the indicated samples. ‘SEN’ samples were cultured in CSS media for 10 days to senescence. ‘Prolif’ indicates samples cultured in parallel in FBS media. (C) Trypan blue staining to indicate cell death in response to indicated treatments for 48 hours (n = 3): Docetaxel (1 nM, Doc); Flavopiridol (0.5 µM, FP); Bortezemib (1 µM, Btz). LNCaP cells were kept in CSS culture for 10 days prior to treatment. (D) Western blotting for pro-survival proteins from the indicated lysates. Note elevated Mcl-1, TAp63 and survivin levels in SB5 cells relative to SB0 cells. The TAp63 band corresponds to the 77 kDa alpha isoform; the other lower mobility isoforms are negligible or absent. (E) Trypan blue staining to indicate cell death. LNCaP SB0 and ADIS–resistant LNCaP SB5 cells in FBS media were treated with 1 nM Doc for 48 hours (n = 6), ** p<0.01.
Figure 5
Figure 5. AR knockdown induces senescence in the LNCaP SB0 but not the LNCaP SB5 cells.
LNCaP cells were transduced with the indicated shRNA constructs and kept continuously in 2 µg/ml puromycin selection, starting at 48 hours after transduction. (A) Western blotting for senescence-associated and pro-survival molecular markers. Six days following transduction with shGFP or shAR, SB0 and SB5 LNCaP cells were harvested and lysed. Prior to lysis, cells were kept continuously in 2 µg/ml puromycin selection, starting 48 hours transduction. Approximately 25 µg of protein was immunoblotted against the indicated antibodies. GAPDH expression was used to normalize for loading. The TAp63 signal is from the 77 kDa alpha isoform and the other two isoforms are absent. (B) SA-beta-gal assay. Cells were stained 7 days following transduction with shRNA constructs. Representative images are shown above the quantitation. (C) Proliferation rates. Cells were seeded in duplicate at a density of 7.5×104 and counted at the indicated intervals. Note that the SB5 shGFP and shAR cells grow at near-identical rates, as expected.
Figure 6
Figure 6. Activating p53 with Nutlin-3 prior to AD triggers cell death rather than senescence in LNCaP cells.
(A) ADIS inhibits p53/Bax activation. LNCaP cells cultured in FBS or in CSS culture for 10 days till senescent (CSS-Sen) were subjected to 10 µM Nutlin-3 treatment for 72 hours. ‘Pre-treat’ indicates cells in FBS culture were pre-treated with a non-cytotoxic Nutlin-3 dose (2 µM, 48 hours) to activate p53 before being switched to fresh FBS or CSS as indicated and treated with 10 µM Nutlin for 72 hours. % Trypan blue staining indicates extent of cell death. (B), (C) Immunoblotting of counterpart samples from (A) against indicated proteins. Approximately 35 µg of samples were immunoblotted. Note increased p53, Bax and cl-PARP in Nutlin-3 treated compared to vehicle (DMSO) cells in LNCaP/FBS cells. Despite increased p53 expression in LNCaP/CSS-SEN cells, note the lower expression of cl-PARP and Bax under Nutlin-3 treatment when compared to the LNCaP/FBS cells. GAPDH expression is used to normalize for protein loading. (D) Schematic indicating stressors and tumor suppressor pathways evoked upon AD and the potential means of switching response from cell senescence to cell death.

Similar articles

See all similar articles

Cited by 12 PubMed Central articles

See all "Cited by" articles


    1. Jenster G (1999) The role of the androgen receptor in the development and progression of prostate cancer. Semin Oncol 26: 407–421. - PubMed
    1. Feldman BJ, Feldman D (2001) The development of androgen-independent prostate cancer. Nature Reviews Cancer 1: 34–45. - PubMed
    1. Pienta KJ, Bradley D (2006) Mechanisms underlying the development of androgen-independent prostate cancer. Clin Cancer Res 12: 1665–1671. - PubMed
    1. Linja MJ, Savinainen KJ, Saramaki OR, Tammela TL, Vessella RL, et al. (2001) Amplification and overexpression of androgen receptor gene in hormone-refractory prostate cancer. Cancer Research 61: 3550–3555. - PubMed
    1. Chmelar R, Buchanan G, Need EF, Tilley W, Greenberg NM (2007) Androgen receptor coregulators and their involvement in the development and progression of prostate cancer. Int J Cancer 120: 719–733. - PubMed

Publication types

MeSH terms

Grant support

This work was funded by a Florida Biomed Bankhead Coley New Investigator Award, a University of Miami/Sylvester Comprehensive Cancer Center Pap Corps Developmental Cancer Research Grant and a University of Miami/Stanley Glaser Foundation award (to PR). MGG was partially supported by a Sylvester Comprehensive Cancer Center Summer Undergraduate Research Award. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.