Figure 1.
AR activation causes cell cycle arrest and senescence in PC3 cells.
AR activation causes cell cycle arrest and senescence in PC3 cells. (A) In situ immunofluorescence with anti-AR antibody: note diffuse AR staining in control treated PC3-AR cells (left) and nuclear translocation on 3 day of DHT treatment (right). (B) Time-dependent inhibition of cell growth in vitro. PC-3 cells were cultured in Dox ± DHT; viable cells were measured using WST-1 reagent. Each time point represents mean ± standard deviation of three independent experiments. The difference between DHT treatment (red) and control EtOH (green) is statistically significant on day 5 (P<0.01). (C) Cell cycle analysis: the cells were grown in Dox, in the absence (D) and in the presence of DHT (DD) for 3 and 5 days; after 7 days the cells were passed (P1) and incubated for another 5 days. Cell cycle distribution was analyzed by flow cytometry. Note increased cell number in Go/G1 phase accompanied by decreased in S and G2/M populations in DHT-treated cells (DD) compared to Dox alone (D). (D) PC3-AR cells were cultured in Dox and treated with DHT ± Flutamide (Fl, 20 µM). After five days senescence was measured using SA-βGal assay. Note increased βGal positivity upon AR activation and the lack of senescence in the presence of Fl. (E) Senescent cells were counted on the digital images of 5 random fields using Image Tool 3.00 software (UTHSCSA); means of three independent experiments with S.D.M. are shown. (F) RWPE-1 cells were transfected with control lentivirus or lentiviral vector encoding AR. AR expression was measured in whole cell lysates by Western blot. GAPDH antibody was used to assess loading. (G, H) RWPE-AR and vector transfected (RWPE-C) cells were cultured in DHT, DHT and Flutamide (Fl) or with equal volume of EtOH added as control. Senescence was assessed with SA-βGal assay (I) and quantified as in (E).
Figure 2.
AR-dependent tumor suppression was associated with increased cellular senescence.
AR increased cellular senescence in vivo. (A) Experimental design. Male athymic mice (nu/nu, 5 mice/group) were injected s.c. with vector control (PC3-V) or PC3-AR cells and received Dox with drinking water, where indicated, to induce AR expression. Flutamide was given to counteract endogenous testosterone. (B) Mean tumor volumes and standard deviations are shown at 45-day time point. Note lower volume of PC3-AR tumors in male mice treated with Dox, and the lack of the decrease when Fl was added to the treatment. (C) Sections of frozen tumors collected in the experiments above were subjected to SA-βGal assay to detect senescence. Note increased βGal positivity (blue) in PC3-AR tumors treated with Dox and of LNCaP tumors treated with DHT.
Figure 3.
p21 is regulated by AR and required for cellular senescence.
AR-dependent senescence required increased p21. (A) PC3-AR cells were treated with Dox, DHT and Fl, at indicated combinations; cell extracts were analyzed by Western blot for p21, and ppRb. Tubulin was used as a loading control (A). Note the decrease in Rb phosphorylation and p21 increase in DHT-treated cells at every time point. (B) RWPE-1 cells, non-transfected or transfected with pLVX-AR or pLVX vector were treated with vehicle EtOH and DHT with or without Fl. Cell extracts were analyzed for AR, p21 and GAPDH to assess loading. (C) AR binding to p21 promoter: ChIP was performed with AR antibodies and AR-bound DNA amplified with the primers for the promoter region adjacent to the putative ARE element within p21 promoter (top) using real-time PCR. Each sample was run in triplicates, normalized per input DNA, and fold change in occupancy was calculated as FC = 2 (−ΔΔCT [exp-con]). Note increased AR binding to the p21 promoter in the presence of DHT and the reversal by flutamide (Fl, concentration shown in µM). The results of three independent experiments are pulled together (P<0.04). (D) PC3-AR cells were transfected with p21 or non-target control siRNA. Cell extracts were collected after 48 hours and analyzed by Western blot for p21, phospho-Rb (ppRb) and tubulin (loading control). Untransfected cells (-) are shown for comparison. (E, F) The transfectants were allowed 48 hours to recover and treated 5 days with Dox, DHT and Fl, as indicated; senescence was measured by SA-βGal assay. Note a significantly lower senescence levels after p21 knock-down (F, P<0.006). SA-βGal positive cells were quantified as above. Means and S.D. were calculated for three independent experiments. Representative images are shown (F).
Figure 4.
AR-induced senescence requires ROS and Rb.
This figure depicts the role of ROS and RB in AR-dependent senescence. (A). PC3-AR cells were treated with Dox, DHT and Fl, as indicated and ROS assessed by DHE fluorescence. Note increased ROS upon DHT treatment and lower DHE fluorescence in the presence of Fl (20 µM). ROS quencher, NAC (10 mM) was used as a negative control. (B, C) The effect of NAC on AR-induced senescence: PC3-AR cells were cultured in Dox, AR activated with DHT ± NAC, and senescent cells visualized with SA-βGal assay. Representative images (B) and quantitative assessment (C) are shown. (D) PC3-AR cells were cultured in Dox, treated 24 hours with DHT and/or NAC (20 mM) where indicated. Cell extracts were analyzed by western blot for phospho-Rb. PCNA was used to assess loading. Note increased Rb phosphorylation in the presence of NAC. (E) PC3-AR cells were transfected with Rb siRNA and Rb levels measured after 24 hours by western blot. Note ∼2-fold reduction after transfection with Rb siRNA. (F, G) Senescence was measured by SA-βGal assay on day 6 after Rb knock-down (values calculated as in Fig. 1, S.D. values are shown; P<0.0002). Note the reduced senescence of the DHT-treated cells after Rb knock-down. Representative images are shown below (G).
Figure 5.
p63 expression is blocked by P21 and contributes to AR-dependent senescence.
This figure illustrates the link between p21 and p63 n AR-driven senescence. (A) PC3-AR cells were transfected with non-target (NT), p21 or p63 siRNA. Dox ± DHT were added to the media 24 hours post-transfection. Whole cell lysates were collected after 3 days and p63 expression measured by Western blot. Arrows point to TAp63-α and ΔNp63; * indicates non-specific band. Note the lack of p63 downregulation in p21 knock-down cells. (B) RWPE-1 cells were transfected with pLVX-AR or control pLVX and treated with vehicle EtOH or DHT. TAp63-α and ΔNp63 were measured as in (A). (C) PC3-AR cells were transfected with p21 or non-target (NT) siRNA. The cells were placed in Dox, ± DHT. RNA was isolated 3 days post-transfection. P63 message was measured by real-time RT-PCR. Note an approximately seven-fold reduction upon DHT treatment and increased expression after p21 knock-down. (D) In situ immunofluorescence of PC3-AR cells with pan-p63 antibody. Note pronounced p63 nuclear staining in Dox-treated cells, the lack of nuclear staining after 3 days in DHT and restored nuclear staining in the presence of Fl (top). Below: P63 nuclear localization (green pseudocolor), nuclei are counterstained with DAPI (red pseudocolor); colocalization appears in yellow. (E) PC3-AR cells were transfected with p21, p63 or control siRNA, cultured 48 hours in Dox ± DHT and stained with p63 antibodies. Note similar, diffuse p63 localization with the nuclear presence (filled arrows) in Dox-treated cells, the lack of nuclear localization in the presence of DHT (empty arrows), which is lost after p21 knock-down (filled arrows). P63 knock-down results in a weak, residual cytoplasmic staining. (F, G) Parental PC3 cells lacking AR expression were transfected with NT and p63 siRNA, cultured 5 days and senescence measured by SA-βGal assay. Representative images (left) and quantitative analysis (right) are shown. * indicates P<0.05.
Figure 6.
AR alters PML aggregation via p63 pathway.
PML tumor suppressor was linked to AR dependent senescence. (A, C) PC3-AR cells were cultured in Dox (D) and treated with DHT (DD) or DHT and Fl (DDF) for 48 hours and stained for AR, to assess activation state, and for PML (A). The number of PML bodies per 20× field was measured using MetaMorph software and recalculated per single cell (C, left). Note a significant increase in the number of PML bodies/cell in the presence of DHT (P<0.02), but not in the presence of Fl. (B, C) PC3-AR cells were transfected with NT or p63 siRNA, cultured for 3 and 5 days and stained for PML and Rb. The number of PML bodies per cell was calculated as above (C, right). Note a robust decrease in PML staining upon p63 knock-down (day 3, P<0.05; day 5, P<0.0001). (D) PC3-AR cells were transfected with non-target (NT) or p21 siRNA, treated for 3 days with DHT and stained for PML. Note decreased staining upon p21 knock-down. (E) PC3-AR treated 3 days with Dox ± DHT and stained for AR, Rb and PML. Note AR activation and nuclear translocation, concomitantly increased intensity of Rb staining with appearance of partial aggregation, and co-localization of Rb aggregates with PML nuclear bodies.
Figure 7.
Molecular mechanism of the AR-induced senescence.
This figure summarizes our findings. (A) mTOR activity in PC3-AR cells. Cells were treated with Dox alone (D) or in combination with DHT. Cell extracts were collected at days 3 and 5 of treatment and analyzed by Western blot for phosphorylated mTOR and p70S6K. Tubulin and total p70S6K served as loading controls. (B) Schematic representation of pathways leading to AR-indiced senescence: AR triggers two parallel pathways, necessary to ensure senescence and possibly engaged in a cross-talk: (1) Activated AR directly binds p21 promoter and thus causes protein expression and accumulation. P21 attenuates the levels of TAp63-α, increasing the number of PML nuclear bodies and causing senescence (2) AR enhances ROS production thus reducing phospho-Rb levels, while active Rb binds and sequesters E2F, causing senescence. Elevated p21 levels may additionally decrease phospho-Rb by blocking Cdks. Conversely, transcriptionally inactive Rb/E2F/HDAC complexes may be sequestered in PML bodies to maintain irreversible growth arrest/senescence. These pathways cumulatively contribute to quiescence, which progresses to senescence due to constitutive high mTOR activity.