Androgen receptor (AR) is a major survival factor for prostate cancer. Inflammation is implicated in many cancer types, including prostate cancer. Activation of MAP3K7 (also termed TAK1) and downstream IκB kinase β (IKKβ) by proinflammatory cytokines such as TNFα stimulates NF-κB survival pathways. Paradoxically, MAP3K7 is often deleted in human prostate cancer. Here, we demonstrate that AR protein expression is lower in inflammatory tumor areas compared with non-inflammatory tissues in patients with prostate cancer. Map3k7 knockout increased AR protein levels and activity in the mouse prostate, and MAP3K7 and AR protein levels were inversely correlated in prostate cancer patient specimens. TNFα treatment increased AR protein ubiquitination and proteasomal degradation. Mechanistically, activation of IKKβ by TNFα induced phosphorylation and TRCP1/2 E3 ligase–mediated polyubiquitination and degradation of AR protein. TNFα suppressed prostate cancer proliferation, which could be rescued by blockade of AR degradation. These findings reveal a previously unrecognized tumor suppressive function of the inflammation-activated MAP3K7–IKKβ axis in degrading AR protein. Moreover, they suggest that aberrant elevation of AR protein could be a prognostic biomarker and therapeutic target for MAP3K7-deficient prostate cancer.

Significance:

This study identifies that MAP3K7–IKKβ signaling plays a tumor-suppressive role in prostate cancer by degrading AR, revealing potential prognostic and therapeutic strategies for MAP3K7-deficient tumors.

Prostate cancer is the most common male malignancy and the second leading cause of male cancer-related death in western countries. Inflammation is implicated in prostate cancer pathogenesis (1, 2). Increasing evidence suggests that the production of the pro-inflammatory cytokines such as TNFα from inflammatory environment and tumor cells promotes activation of the downstream oncogenic NF-κB and survival of cancerous cells (3). Given the importance of inflammation in prostate cancer development and progression (2), the signaling pathways mediating the activation of NF-κB have been attractive targets for chemo-prevention and -therapy of prostate cancer. Although rationalized chemoprevention trials using anti-inflammation drugs have been conducted, the outcomes have been mixed (2), stressing the importance of investigating the mechanism of action of inflammation in prostate cancer.

MAP3K7, a serine/threonine protein kinase (also termed TGFβ-activated kinase 1, TAK1), is an important component of MAPK signaling pathways (4–6). MAP3K7 is also one of the key mediators of the inflammatory signaling and can be activated by TNFα, TGFβ, and lipopolysaccharides. With the stimulation of these cytokines, MAP3K7 can form a complex with TAK1-binding protein 1 (TAB1), TAB2, and TRAF6 to promote the activation of IκB kinase (IKK) complex and its downstream target NF-κB. The IKK complex contains two catalytic subunits, IKKα and IKKβ and one regulatory subunit IKKγ (also known as NEMO). Activation of IKK by proinflammatory cytokines such as TNFα leads to phosphorylation and subsequent poly-ubiquitination and degradation of IκB mediated by E3 ubiquitin ligases. Recent studies show that IKK can also phosphorylate other cancer-related proteins, including BAD, FOXO3, BCL-10, and A20 and promote ubiquitination and degradation of these proteins (7–11). MAP3K7 gene is also implicated in human cancers such as prostate cancer because it is deleted in more than 10% of prostate cancer in patients (12, 13). However, the precise role of MAP3K7 loss in prostate cancer pathogenesis remains largely unclear.

Androgen receptor (AR), a member of the steroid hormone nuclear receptor family, plays a critical role in initiation and progression of prostate cancer. As a key regulator of prostate homeostasis, AR protein level and its activity are precisely regulated by posttranslational modifications including phosphorylation, acetylation, sumoylation, methylation, and ubiquitination (14, 15). Similar to other transcription factors, AR protein is tightly regulated by the ubiquitin–proteasome pathway. The E3 ubiquitin ligases such as MDM2, C-terminal HSP-interacting protein, and Siah have been implicated in the control of AR stability and activity under basal or stress conditions (16–18).

In the present study, we demonstrated that the proinflammatory cytokine TNFα induces AR protein polyubiquitination and proteasomal degradation. By generating prostate-specific Map3k7 gene knockout (KO) mice, we demonstrated that loss of Map3k7 increased AR protein level and activity in the mouse prostate. We also showed that IKKβ, a downstream effector of MAP3K7, interacts with AR and enhances AR degradation in a manner dependent on β-TRCP E3 ligases. We further demonstrated that blockade of AR degradation enhanced AR activity and prostate cancer cell growth in vitro and in castrated mice under the inflammatory condition.

Cell lines, cell culture, and transfection

LNCaP, LAPC-4, C4–2, 22Rv1, PC-3, DU145, RWPE-1, and 293T cell lines were purchased from the ATCC and authenticated periodically via STR profiling (IDEXX BioResearch). LNCaP, LAPC-4, C4–2, 22Rv1, PC-3, and DU145 cells were cultured in RPMI-1640 medium supplemented with 10% FBS. RWPE-1 cells were cultured in K-SFM medium supplemented with bovine pituitary extract and human recombinant EGF. 293T cells were maintained in DMEM supplemented with 10% FBS. Mycoplasma contamination was regularly examined using Lookout Mycoplasma PCR Detection Kit (Sigma-Aldrich). Plasmocin (InvivoGen) was routinely added to the cell culture medium to prevent or eliminate Mycoplasma contamination. Transfections were performed by electroporation using an Electro Square Porator ECM 830 (BTX; ref. 19) or by using Lipofectamine 2000 (Thermo Fisher Scientific). Approximately 70%–90% transfection efficiencies were routinely achieved.

Generation of prostate-specific map3k7 KO mice

Map3k7 (Tak1) loxp/loxp (Map3k7P/P) mice were kindly provided by Dr. Jiwang Zhang from the Loyola University Medical Center (Maywood, IL). Exon-2 of the Map3k7 gene was flanked by two loxp sites (20). Probasin (Pb)-Cre4 transgenic mice were acquired from the National Cancer Institute (NCI) Mouse Repository, which were originally generated in the laboratory of Dr. Pradip Roy-Burman at University of Southern California (Los Angeles, CA; ref. 21). Pb-Cre+;Map3k7P/P mice were obtained by cross breeding Pb-Cre4 males with Map3k7P/P females. All mice were maintained under standard conditions of food, water, light, and temperature. All experimental protocols were approved by the Institutional Animal Care and Use Committee (IACUC) at Mayo Clinic.

PCR-based genotyping of mice

Genotyping of “wild-type” and Pb-Cre+;Map3k7P/P mice was performed following standard PCR protocol with primers listed in Supplementary Table S1.

Generation of xenografts in mice

The mouse study was conducted according to the NIH guidelines and was approved by the IACUC at Mayo Clinic. Age-matched 6-week-old male SCID mice were castrated before cell injection and randomly separated into different groups. For C4–2 xenograft study, C4–2 cell lines (3 × 106) infected with indicated lentivirus were suspended in 50 μL of PBS and mixed with 50 μL of Matrigel (BD Biosciences) and administered into subcutaneous layer of the left flank of mice. The caliper was used to determine the xenograft size every three day for 27 days. The xenograft volume was determined using the formula (length × width2)/2. After 27-day observation, xenografts were harvested. For LNCaP cell xenograft study, LNCaP cell lines (3 × 106) expressing EV, WT AR or degradation-resistant mutant were suspended in 50 μL of PBS and mixed with 50 μL of Matrigel (BD Biosciences) before they were inoculated into subcutaneous layer of the left flank of mice. Mice were treated with TNFα (10 ng/kg/d) six days after cell injection. A caliper was used to determine the xenograft size every three day for 27 days and xenografts were harvested at the end of measurement.

Tissue section and hematoxylin and eosin staining

4-μm-thick sections were cut from formalin-fixed paraffin-embedded (FFPE) blocks of human prostate cancer specimens, mouse prostate tissues, and xenograft tumors and mounted onto slides. All slides were deparaffinized in xylene for three times and rehydrated through graded ethanol. After stained with hematoxylin and washed with water and ethanol, slides were staining with 1% eosin. Slides were dehydrated through different concentration ethanol washes and xylene washes, and finally sealed over the tissue sections with coverslips.

IHC of human and mouse prostate cancer tissues

A tissue microarray (TMA) was constructed from the FFPE samples of metastatic prostate cancer, identified after a search of pathologic and clinical databases of archival tissues collected originally with written informed consent from patients (22, 23). The Mayo Clinic institutional review board approved the experimental protocols for retrieving pathology blocks/slides and for accessing electronic medical records. 4-μm-thick sections were cut from FFPE blocks of human prostate cancer specimens, mouse prostate tissues, and xenograft tumors and mounted onto slides. IHC was performed as described previously (24). The human TMA contained 157 cores (16 0.6-mm cores and 141 1.0-mm cores) resulting from 53 samples (20 bone metastases and 33 non-bone metastases) of 51 patients. 154 cores were used for IHC data analysis after excluding cores with lost tissue greater than 50%. Staining index (SI) was calculated as follows: Staining intensity (intensity: 0 = no staining, 1 = low staining, 2 = medium staining, and 3 = strong staining) and staining percentage was graded accordingly. A final SI score for each staining was obtained by multiplying values of staining percentage and intensity.

Immunofluorescent cytochemistry

Organoids derived from the prostate tissues of “wild-type” control and Map3k7 KO mice (Pb-Cre+; Map3k7P/P) or 22Rv1 cells were harvested, mounted or seeded on the slides and fixed with ice-cold methanol/acetone (1:1 dilution, stored at −20°C). After washing with IF buffer and blocking with 10% of goat serum, slides were incubated with the primary antibody at 4°C overnight. The next day, slides were incubated with the second antibody and mounted with mounting medium containing DAPI. Immunofluorescence was visualized and photos were taken using LSM700 confocal microscope (Carl Zeiss). The OD values of staining in each cell were measured. 50 cells from each organoid line generated from each mouse were randomly selected and five organoid lines were used for quantitative analysis.

RNAi

Smart pools of siRNAs for human IKKβ, β-TRCP2, and nonspecific control siRNAs were purchased from Thermo Fisher Scientific—Dharmacon. Sh-AR, sh-MAP3K7, sh-MAPK14, sh-JNK, and nonspecific control shRNAs were purchased from Sigma-Aldrich (a list of shRNAs is provided in Supplementary Table S2). siRNA and shRNA transfections of cells were performed following the manufacturer's instruction.

qRT-PCR

Total RNA was isolated from cells or mouse prostate tissues and cDNA was synthesized using the Super-Script kit from Thermo Fisher Scientific. Two-step RT-PCR was performed using the SYBR Green Mix (Bio-Rad) and an iCycler iQTM detection system (Bio-Rad) according to the manufacturer's instructions. Both forward and reverse primers were used at a final concentration of 200 nmol/L. Expression of GAPDH gene in each sample was used as an internal control. Information for primers used is provided in Supplementary Table S1.

Statistical analysis

Experiments were carried out with three or more replicates. Statistical analyses were performed using the Student t test unless otherwise stated. Values with P < 0.05 are considered statistically significant.

Other methods

Details related to other methods are provided in Supplementary Information.

TNFα promotes AR protein degradation in prostate cancer cells

De Marzo and colleagues (25) made a seminal observation that AR protein levels are much lower in luminal cells within the proliferative inflammation atrophy (PIA) lesion compared with tissues outside PIA. We were interested to determine whether there is any association between low AR protein expression and inflammation in prostate cancer. AR protein IHC analysis in prostate cancer patient specimens showed that AR protein level was much lower in inflammatory areas compared those without inflammation (Fig. 1A). Quantitative analysis further confirmed the negative correlation between inflammation and AR protein expression in prostate cancer (Fig. 1B).

Figure 1.

Proinflammation cytokine TNFα promotes AR degradation in prostate cancer. A, Representative images for hematoxylin and eosin (H&E) and AR IHC staining of human prostate cancer samples (n = 30). Red asterisk, inflammation area. B, AR protein expression in prostate epithelial cells from noninflammation areas, inflammation areas, and areas adjacent to inflammation in human prostate cancer samples shown in A was quantified using ImageJ. Fifty cells were calculated per sample. The χ2 test. C, prostate cancer cell lines LNCaP and LAPC-4 were treated with vehicle or TNFα (25 or 50 ng/mL) for 16 hours and harvested for immunoblot analysis to detect the AR protein level. β-Tubulin was used as a loading control. D, LNCaP cell were treated with or without 50 ng/mL of TNFα. After 16-hours incubation, cells were treated with or without MG132 (20 μmol/L) for 8 hours and harvested for immunoblot analysis. E, LNCaP cells were transfected with HA-Ub and treated with various concentrations of TNFα. After 16-hours transfection, cells were treated with MG132 for 8 hours and harvested for ubiquitination assay. F, LNCaP cells were treated with or without 50 ng/mL TNFα in combination with 20 μg/mL of cycloheximide (CHX) and harvested at different time points for immunoblot analysis. G, Quantification of the intensity of immunoblot bands in F. AR band intensity was normalized to β-tubulin first and further normalized to the normalized value at 0-hours time point. Data shown as means ± SD (n = 3 repeats). **, P < 0.01.

Figure 1.

Proinflammation cytokine TNFα promotes AR degradation in prostate cancer. A, Representative images for hematoxylin and eosin (H&E) and AR IHC staining of human prostate cancer samples (n = 30). Red asterisk, inflammation area. B, AR protein expression in prostate epithelial cells from noninflammation areas, inflammation areas, and areas adjacent to inflammation in human prostate cancer samples shown in A was quantified using ImageJ. Fifty cells were calculated per sample. The χ2 test. C, prostate cancer cell lines LNCaP and LAPC-4 were treated with vehicle or TNFα (25 or 50 ng/mL) for 16 hours and harvested for immunoblot analysis to detect the AR protein level. β-Tubulin was used as a loading control. D, LNCaP cell were treated with or without 50 ng/mL of TNFα. After 16-hours incubation, cells were treated with or without MG132 (20 μmol/L) for 8 hours and harvested for immunoblot analysis. E, LNCaP cells were transfected with HA-Ub and treated with various concentrations of TNFα. After 16-hours transfection, cells were treated with MG132 for 8 hours and harvested for ubiquitination assay. F, LNCaP cells were treated with or without 50 ng/mL TNFα in combination with 20 μg/mL of cycloheximide (CHX) and harvested at different time points for immunoblot analysis. G, Quantification of the intensity of immunoblot bands in F. AR band intensity was normalized to β-tubulin first and further normalized to the normalized value at 0-hours time point. Data shown as means ± SD (n = 3 repeats). **, P < 0.01.

Close modal

To determine whether inflammation plays any causal role in downregulating AR protein expression in prostate cancer, we examined the effect of proinflammation cytokines on AR protein expression in prostate cancer cell lines. TNFα is one of the important pro-inflammation cytokines released by the inflammatory environment or cancer cells and plays an essential role in prostate cancer progression (26, 27). We treated AR-positive prostate cancer cell lines LNCaP and LAPC-4 with TNFα and harvested cells for Western blot analysis. We demonstrated that TNFα treatment induced downregulation of AR protein in a dose-dependent manner in both cell lines (Fig. 1C), suggesting that there may be a common mechanism mediating TNFα-induced reduction of AR protein. Next, we examined whether TNFα-induced inhibition of AR protein expression is mediated through the ubiquitination and proteasomal degradation pathway. We treated LNCaP cells with TNFα in the presence or absence of the 26S proteasome inhibitor MG132. TNFα-induced decrease in AR protein level was almost completely blocked by MG132 (Fig. 1D). In agreement with this result, TNFα treatment markedly enhanced AR poly-ubiquitination in LNCaP cells (Fig. 1E). Furthermore, we measured the stability of AR protein in LNCaP cells treated with or without TNFα. The AR protein half-life was approximately 4 hours in vehicle-treated LNCaP cells (Fig. 1F and G), consistent with the previous report (28). Notably, the AR protein half-life was drastically shortened in TNFα-exposed cells (Fig. 1F and G). These results indicate that the proinflammatory cytokine TNFα induces a decrease in AR protein level and this effect is mediated through the ubiquitination and proteasomal degradation pathway in prostate cancer cells.

MAP3K7 KO increases AR protein level and activity in the mouse prostate

MAP3K7 is one of the key factors mediating signaling activated by proinflammatory cytokines such as TNFα (5, 6). Similar to previous reports (12, 13, 29–31), meta-analysis of patient data confirmed that the MAK3K7 gene is often (ranging from 10% to 14% of cases) deleted in prostate cancer specimens (Supplementary Fig. S1A). To determine whether MAP3K7 deletion has any impact on AR protein expression and prostate cancer pathogenesis, we generated prostate-specific Map3k7 gene KO mice (Pb-Cre+;Map3k7P/P) by cross-breeding probasin (Pbsn or termed Pb) promoter-driven Cre males (21) with Map3k7loxp/loxp (Map3k7P/P) conditional females (Fig. 2A; ref. 20). IHC analysis showed that Map3k7 protein was well expressed in the prostate of Cre-negative “wild-type” control mice whereas it was effectively eliminated in the prostate of Map3k7 KO littermates (Fig. 2B). Histological (hematoxylin and eosin) analysis indicated that there were no indications of prostate intraepithelial neoplasia (PIN) lesion, including enlarged nucleus, decreased cytoplasm staining, prominent nucleoli or nuclear atypia in the prostate of Map3k7 KO mice at 12 months of age (Fig. 2B). Hyperplasia or dysplasia was detectable, but only in a very small portion of acini in the ventral prostate of Map3k7 KO mice (Fig. 2B). In agreement with these observations, there was no obvious difference in the size and weight of the prostates between control and Map3k7 KO mice (Fig. 2C and D). Ki-67 expression was slightly increased in the prostate of Map3k7 KO mice compared with control mice, but the difference was not statistically significant (Fig. 2E and F).

Figure 2.

Deletion of MAP3K7 increases AR protein level and activity in the mouse prostate. A, A schematic showing the strategy of generating MAP3K7 knockout mice. B, Hematoxylin and eosin (H&E) and IHC staining of Map3k7 protein in VP of mice with the indicated genotypes at 12 months of age. C, The size of the prostate of mice with the indicated genotypes at 12 months of age. D, Quantification of prostate weight from mice with the indicated genotypes at 12 months of age. The average of prostate weight and the number of mice analyzed for each group are indicated. E, IHC staining of Ki67 in prostate tissues from mice of indicated genotypes at 12 months of age. F, Quantification of Ki67 staining–positive prostate cells in E. Error bars represent means ± SD from three mice in each group. G, IHC staining of AR protein in prostate tissues of mice with the indicated genotypes at 12 months of age. H, Quantification of AR staining–positive prostate cells in G. I, Organoids derived from prostate tissues of mice (n = 5) with the indicated genotypes at 12 months of age were treated with PBS or 25 ng/mL TNFα for 72 hours and subjected to immunofluorescent cytochemistry for Map3k7 and AR proteins. Representative images from organoid lines derived from five mice are shown. J, AR immunofluorescent cytochemistry signal intensity in prostate cells from the organoids shown in I was quantified. Data shown as means ± SD (n = 5 organoid lines/mouse genotype). n.s., no significant difference. *, P < 0.05; **, P < 0.01.

Figure 2.

Deletion of MAP3K7 increases AR protein level and activity in the mouse prostate. A, A schematic showing the strategy of generating MAP3K7 knockout mice. B, Hematoxylin and eosin (H&E) and IHC staining of Map3k7 protein in VP of mice with the indicated genotypes at 12 months of age. C, The size of the prostate of mice with the indicated genotypes at 12 months of age. D, Quantification of prostate weight from mice with the indicated genotypes at 12 months of age. The average of prostate weight and the number of mice analyzed for each group are indicated. E, IHC staining of Ki67 in prostate tissues from mice of indicated genotypes at 12 months of age. F, Quantification of Ki67 staining–positive prostate cells in E. Error bars represent means ± SD from three mice in each group. G, IHC staining of AR protein in prostate tissues of mice with the indicated genotypes at 12 months of age. H, Quantification of AR staining–positive prostate cells in G. I, Organoids derived from prostate tissues of mice (n = 5) with the indicated genotypes at 12 months of age were treated with PBS or 25 ng/mL TNFα for 72 hours and subjected to immunofluorescent cytochemistry for Map3k7 and AR proteins. Representative images from organoid lines derived from five mice are shown. J, AR immunofluorescent cytochemistry signal intensity in prostate cells from the organoids shown in I was quantified. Data shown as means ± SD (n = 5 organoid lines/mouse genotype). n.s., no significant difference. *, P < 0.05; **, P < 0.01.

Close modal

We also performed IHC analysis of murine AR protein and found that the AR protein level was much higher in the prostate of Map3k7 KO mice compared with WT mice (Fig. 2G and H). In agreement with increased AR protein level, expression of AR target genes, including Pbsn, Nkx3.1, and Tmprss2 was significantly higher in the prostate of Map3k7 KO mice compared with control mice (Supplementary Fig. S1B), which is consistent with the results in MAP3K7 knockdown (KD) prostate cancer cells as reported recently (32). We further generated organoids from the prostate tissues of both "wild-type" and Map3k7 KO mice. Similar to the results in the prostate tissues, AR protein was higher in Map3k7 KO organoids compared with “wild-type” counterparts (Fig. 2I and J). Importantly, TNFα treatment decreased AR protein expression in “wild-type” organoids, a result similarly seen in prostate cancer cell lines (Fig. 1C); however, such effect was abolished by Map3k7 KO (Fig. 2I and J).

That MAP3K7 modulates AR protein level in the mouse prostate prompted us to explore whether this also occurs in human prostate cancer cell lines and patient samples. To this end, we infected 22Rv1 and LNCaP cell lines with lentivirus-expressing nonspecific control small hairpin RNA (sh-NS) or MAP3K7-specific shRNAs and treated cells with TNFα. Similar to the results observed in murine prostate organoids, TNFα-induced downregulation of AR protein was completely abolished by MAP3K7 KD by shRNAs in both 22Rv1 and LNCaP cell lines (Supplementary Fig. S1C and S1D).

To further verify whether MAP3K7 is involved in the regulation of AR protein by proinflammatory signaling, we examined the expression of AR and phosphorylated MAP3K7 (p-MAP3K7) proteins in inflammatory and noninflammatory regions of prostate cancer patient specimens with proficient or deficient expression of MAP3K7. We demonstrated that in MAP3K7-proficient prostate cancer tissues, AR protein level was lower in areas with high infiltration of cells positive for CD3, a T-cell marker and high expression of p-MAP3K7 compared with the areas with little or no infiltration of CD3-positive cells and limited expression of p-MAP3K7 (Supplementary Fig. S1E). In contrast, in MAP3K7-deficient samples AR expression levels in CD3-positive areas were comparable with those in CD3-negative areas (Supplementary Fig. S1E). These data indicate that MAP3K7 activation by proinflammatory signaling plays an important role in mediating AR protein downregulation induced by inflammation in prostate cancer tissues. These data further suggest that although deletion of Map3k7 gene alone is not sufficient to promote prostate tumorigenesis in mice, at least within a one-year timeframe, loss of MAP3K7 induces elevation of AR protein level and activity in both mouse and human prostate.

MAP3K7 loss enhances prostate cancer cell growth in vitro and in mice

Consistent with the results observed in LNCaP prostate cancer cells cultured in charcoal-stripped (TNFα-low/none) medium as reported recently (32), KD MAP3K7 had minimal effect on AR protein expression in RWPE1 cells cultured in serum-free (TNFα-free) media (Supplementary Fig. S2A and S2B); however, MAP3K7 KD largely increased AR protein levels in various cell lines treated with TNFα (Supplementary Fig. S2A–S2C). In addition, KD of MAP3K7 significantly increased AR polyubiquitination in LNCaP cells cultured in regular (TNFα-containing) media (Supplementary Fig. S2D). However, coimmunoprecipitation (co-IP) experiments showed that there was no interaction between MAP3K7 and AR (Supplementary Fig. S2E), suggesting that MAP3K7 regulation of AR protein is mediated through indirect mechanisms.

Castration-resistant prostate cancer (CRPC) often metastasizes to lymph node and bone, two organs heavily enriched with lymphocytes. To explore the clinical relevance of MAP3K7-mediated downregulation of AR protein, we examined the expression of MAP3K7 and AR proteins in CRPC patient samples. We performed MAP3K7 and AR IHC on a TMA, which contains 154 TMA elements derived from 53 samples (20 bone metastases and 33 non-bone metastases) of 51 patients as reported previously (22, 23). We revealed that MAP3K7 protein expression negatively correlated with AR protein level in this cohort and the correlation is statistically significant (Pearson correlation test, r = −0.329, P = 3.0e−5; Fig. 3A–C). Moreover, it has been reported that AR activity is enhanced in TMPRSS2ERG gene fusion-positive prostate cancer due to increased expression of the ETS transcription factor ERG (33). We found that tumors with MAP3K7 deletions, which had almost no overlap with ERG gene fusions in the The Cancer Genome Atlas (TCGA) database (Supplementary Fig. S2F; refs. 13, 34), exhibited much greater AR activity than ERG fusion-positive tumors in primary prostate cancer specimens (Supplementary Fig. S2G). These data indicate that loss of MAP3K7 associates with increased AR protein expression and activity in prostate cancer patient samples.

Figure 3.

MAP3K7 negatively regulates AR activity and human prostate cancer cell growth in vitro and in mice. A, Representative images of hematoxylin and eosin (H&E) and IHC staining of MAP3K7 and AR proteins in prostate cancer patient TMA (n = 154 TMA elements). B, Heat map showing the IHC staining index of MAP3K7 and AR proteins in prostate cancer specimens (n = 154 TMA elements). Scale bar, IHC staining index. Each line in the heat map represents a TMA element. C, Correlation analysis of expression of MAP3K7 and AR proteins in TMA specimens (n = 154). Pearson product–moment correlation coefficiency and the P value are also shown. D, Western blot analysis of indicated proteins in lysate of C4–2 cells transfected with the indicated shRNAs. β-Tubulin was used as a loading control. E, C4–2 cells infected with various shRNAs as in D were subjected to MTS assay at different time points. Data shown as means ± SD (n = 5). P value was calculated at day 5 time point between indicated groups. F and G, C4–2 cells infected with the indicated shRNAs as in D were injected subcutaneously into the left flank of castrated SCID mice. Tumor volume of each xenograft at each time point was measured (F), and tumors at the end of study were harvested and photographed (G). Data shown as means ± SD (n = 6). P value was calculated between the indicated groups at day 32. H and I, Quantification analysis of Ki-67 (H) and cleaved caspase-3 IHC–positive cells (I) in the tissue sections of the indicated groups. Data shown as means ± SD (n = 3 fields/section). P value was calculated between indicated groups. *, P < 0.05; **, P < 0.01, n.s., no significant difference.

Figure 3.

MAP3K7 negatively regulates AR activity and human prostate cancer cell growth in vitro and in mice. A, Representative images of hematoxylin and eosin (H&E) and IHC staining of MAP3K7 and AR proteins in prostate cancer patient TMA (n = 154 TMA elements). B, Heat map showing the IHC staining index of MAP3K7 and AR proteins in prostate cancer specimens (n = 154 TMA elements). Scale bar, IHC staining index. Each line in the heat map represents a TMA element. C, Correlation analysis of expression of MAP3K7 and AR proteins in TMA specimens (n = 154). Pearson product–moment correlation coefficiency and the P value are also shown. D, Western blot analysis of indicated proteins in lysate of C4–2 cells transfected with the indicated shRNAs. β-Tubulin was used as a loading control. E, C4–2 cells infected with various shRNAs as in D were subjected to MTS assay at different time points. Data shown as means ± SD (n = 5). P value was calculated at day 5 time point between indicated groups. F and G, C4–2 cells infected with the indicated shRNAs as in D were injected subcutaneously into the left flank of castrated SCID mice. Tumor volume of each xenograft at each time point was measured (F), and tumors at the end of study were harvested and photographed (G). Data shown as means ± SD (n = 6). P value was calculated between the indicated groups at day 32. H and I, Quantification analysis of Ki-67 (H) and cleaved caspase-3 IHC–positive cells (I) in the tissue sections of the indicated groups. Data shown as means ± SD (n = 3 fields/section). P value was calculated between indicated groups. *, P < 0.05; **, P < 0.01, n.s., no significant difference.

Close modal

Next, we sought to determine the importance of MAP3K7 in regulating prostate cancer cell growth in androgen-depleted conditions. Consistent with recent report (32), MAP3K7 KD significantly enhanced growth of 22Rv1 and LAPC-4 cells cultured in charcoal-stripped media (Supplementary Fig. S2H). Also, we knocked down MAP3K7 and AR individually or together in C4–2 and LAPC-4 cells and cultured cells in charcoal-stripped media. Western blot analysis confirmed the effectiveness of MAP3K7 and AR KD in C4–2 and LAPC-4 cells (Fig. 3D and Supplementary Fig. S2I). As expected, AR KD inhibited C4–2 and LAPC-4 cell growth in vitro (Fig. 3E and Supplementary Fig. S2J). Although MAP3K7 KD substantially increased cell growth, this effect was abolished in AR KD cells (Fig. 3E and Supplementary Fig. S2J). Similar to the findings in the Map3k7 KO mouse prostate (Fig. 2I), MAP3K7 KD alone significantly increased expression of AR target genes, including KLK3 (PSA), NKX3.1 and TMPRSS2 in C4–2 cells, a result similar to the recent report (32), but this effect was suspended in AR KD counterparts (Supplementary Fig. S2K).

To assess the effect of MAP3K7 loss on prostate cancer cell growth in vivo, an additional set of C4–2 stable cell lines generated as in Fig. 3E were cultured in androgen-deprived media for 4 weeks before they were injected subcutaneously into castrated SCID mice. We demonstrated that MAP3K7-KD tumors grew much faster than control tumors, but the effect was blocked in AR KD tumors (Fig. 3F and G). IHC analysis showed that depletion of MAP3K7 significantly increased Ki-67 expression; however, this effect was abolished in AR KD tumors (Fig. 3H; Supplementary Fig. S2L). We also examined the effect of MAP3K7 and/or AR KD on expression of cleaved Caspase-3. We found that MAP3K7 KD alone had no obvious effect on the level of cleaved Caspase-3 in these tumors (Fig. 3I; Supplementary Fig. S2L). In contrast, AR KD largely increased Caspase-3 cleavage in both control and MAP3K7 KD tumors (Fig. 3I; Supplementary Fig. S2L). We further examined the expression of AR target genes including KLK3 (PSA), NKX3.1 and TMPRSS2 in these tumors. We found that knocking down MAP3K7 increased expression of these AR target genes without having significant effects on AR mRNA expression (Supplementary Fig. S2M). Expression of these genes was downregulated by AR co-KD (Supplementary Fig. S2M), similar to the results reported recently (32). Collectively, we demonstrate that deletion of MAP3K7 enhances castration-resistant growth of prostate cancer cells under both in vitro and in vivo conditions and this effect is ablated in the absence of AR protein expression.

IKKβ activation by TNFα induces AR protein degradation in prostate cancer cells

Given that there is no interaction between MAP3K7 and AR (Supplementary Fig. S2E), we sought to determine whether its downstream factors, including MAPK14, JNK, and IKKβ (35), play any role in TNFα-induced AR degradation. We found that KD of endogenous IKKβ (IKBKB) increased the level of endogenous AR protein in LNCaP cells cultured in FBS-supplemented (TNFα-contained) medium (Fig. 4A) whereas neither MAPK14 nor JNK KD had any obvious effect on AR protein expression (Supplementary Fig. S3A and S3B). These data suggest that IKKβ is one of the key downstream effectors of MAP3K7-mediating TNFα-induced AR degradation. To further test this hypothesis, we transfected IKKβ expression plasmid into LNCaP cells and demonstrated that ectopic expression of IKKβ decreased AR protein levels, but this effect was completely abolished by MG132 treatment (Fig. 4B), suggesting that IKKβ promotes AR protein degradation through the proteasomal pathway. Moreover, different from the WT IKKβ, the kinase-dead mutant (IKKβ-KD) failed to decrease AR protein expression in LNCaP cells (Fig. 4C). Treatment of LNCaP cells with the IKKβ inhibitor wedelolactone blocked TNFα-induced AR protein destruction (Fig. 4D). These data support the notion that the kinase activity is essential for IKKβ-mediated degradation of AR protein in LNCaP cells. Because MG132 treatment blocked IKKβ-induced AR destruction, we examined whether IKKβ promotes AR protein polyubiquitination. Expression of WT IKKβ enhanced polyubiquitination of ectopically expressed or endogenous AR, but no such effect was observed for the IKKβ-KD mutant (Fig. 4E; Supplementary Fig. S3C). In contrast, genetic silencing of IKKβ by siRNAs or pharmacological inhibition of IKKβ by wedelolactone decreased AR polyubiquitination (Fig. 4F and G). Protein half-life analysis showed that the AR protein stability was increased upon depletion of IKKβ by siRNAs (Fig. 4H–J). IKKβ is widely studied in different cancer types in various conditions (36). Consistent with the previous report that depletion of IKKβ inhibits proliferation of AR-negative prostate cancer cells (37), shRNA-mediated KD of IKKβ also inhibited growth of PC-3 and DU145, two AR-negative prostate cancer cell lines (Supplementary Fig. S3D). In contrast, we demonstrated that IKKβ KD stimulated growth of AR-positive 22Rv1and LNCaP cell lines (Supplementary Fig. S3D). Furthermore, we demonstrated that TNFα-induced downregulation of AR protein was abolished by depletion of MAP3K7 or IKKβ alone or together (Supplementary Fig. S3E). Collectively, all these data indicate that IKKβ promotes polyubiquitination and proteasomal degradation of AR, an effect similar to its upstream activator MAP3K7.

Figure 4.

IKKβ mediates TNFα-induced AR degradation. A, LNCaP cells were transfected with nonspecific siRNA (si-NS) or IKBKB-specific siRNA (si-IKBKB) for 48 hours and harvested for immunoblot analysis. B, LNCaP cells transfected with or without IKKβ for 16 hours were treated with or without MG132 for additional 8 hours and harvested for immunoblot analysis. C, LNCaP cells were transfected with wild-type (WT) IKKβ or its kinase-dead mutant (IKKβ-KD) for 24 hours and harvested for immunoblot analysis. D, LNCaP cells were treated with 50 ng/mL of TNFα in the presence or absence of the IKKβ inhibitor wedelolactone (20 μmol/L) for 24 hours and harvested for immunoblot analysis. E, 293T cells transfected with the indicated plasmids for 16 hours were treated with MG132 for 8 hours and harvested for ubiquitination assay. F, LNCaP cells were transfected as indicated for 48 hours and harvested for ubiquitination assay. G, LNCaP cells transfected with HA-Ub were treated with TNFα (50 ng/mL) in combination with or without Wedelolactone for 24 hours and harvested for ubiquitination assay. H, LNCaP cells were transfected with indicated siRNAs for 48 hours and harvested for immunoblot analysis. I, LNCaP cells were transfected with indicated siRNAs for 48 hours and treated with 20 μg/mL of cycloheximide (CHX). At the indicated time points, cells were harvested for immunoblot analysis. J, Quantification of the intensity of immunoblot bands shown in I. AR band intensity was normalized to β-tubulin first and further normalized to the value at the 0 hours time point. Data shown as means ± SD (n = 3). **, P < 0.01.

Figure 4.

IKKβ mediates TNFα-induced AR degradation. A, LNCaP cells were transfected with nonspecific siRNA (si-NS) or IKBKB-specific siRNA (si-IKBKB) for 48 hours and harvested for immunoblot analysis. B, LNCaP cells transfected with or without IKKβ for 16 hours were treated with or without MG132 for additional 8 hours and harvested for immunoblot analysis. C, LNCaP cells were transfected with wild-type (WT) IKKβ or its kinase-dead mutant (IKKβ-KD) for 24 hours and harvested for immunoblot analysis. D, LNCaP cells were treated with 50 ng/mL of TNFα in the presence or absence of the IKKβ inhibitor wedelolactone (20 μmol/L) for 24 hours and harvested for immunoblot analysis. E, 293T cells transfected with the indicated plasmids for 16 hours were treated with MG132 for 8 hours and harvested for ubiquitination assay. F, LNCaP cells were transfected as indicated for 48 hours and harvested for ubiquitination assay. G, LNCaP cells transfected with HA-Ub were treated with TNFα (50 ng/mL) in combination with or without Wedelolactone for 24 hours and harvested for ubiquitination assay. H, LNCaP cells were transfected with indicated siRNAs for 48 hours and harvested for immunoblot analysis. I, LNCaP cells were transfected with indicated siRNAs for 48 hours and treated with 20 μg/mL of cycloheximide (CHX). At the indicated time points, cells were harvested for immunoblot analysis. J, Quantification of the intensity of immunoblot bands shown in I. AR band intensity was normalized to β-tubulin first and further normalized to the value at the 0 hours time point. Data shown as means ± SD (n = 3). **, P < 0.01.

Close modal

Β-TRCP E3 ubiquitin ligases mediate AR protein poly-ubiquitination in prostate cancer cells

SKP1-CULLIN1-F-Box protein (SCF) E3 ubiquitin ligases, including β-TRCP, bind the phosphodegron(s) in substrates and mediate target proteins for polyubiquitination and proteasomal degradation (38). Increasing evidence indicates that IKK phosphorylation sequences often serve as the phosphodegrons recognized by β-TRCP (39, 40). We noticed that there are two putative IKK-β-TRCP–targeting phosphodegrons in AR protein that are similar to those identified in several known β-TRCP degradation substrates, including NF-κB1 protein p105, IκB, FOXO3, and β-Catenin (Fig. 5A). We therefore sought to determine whether β-TRCP1 and β-TRCP2 are involved in regulation of AR protein levels. We found that ectopic expression of wild-type β-TRCP1 decreased AR protein level, and a similar result was observed for β-TRCP2, although co-expression of β-TRCP1 and β-TRCP2 resulted in an additive reduction in AR protein level (Fig. 5B). The F box of β-TRCP is essential for its association with the scaffold protein CULLIN1 (CUL1) via the interaction with the adaptor protein SKP1 (38). Apart from the wild-type β-TRCPs, neither the F-box deletion mutant of β-TRCP1 (TRCP1ΔF-box) nor the mutant of β-TRCP2 (TRCP2ΔF-box) was able to decrease AR protein level (Fig. 5B). Given that β-TRCP2 (BTRC2) expression level was much higher than β-TRCP1 (BTRC) in prostate cancer patient samples in the TCGA dataset (Fig. 5C), we chose to focus on β-TRCP2 in further studies. Co-IP assays revealed that AR and β-TRCP2 proteins interacted with each other at the endogenous level in LNCaP cells (Fig. 5D). Their interaction was substantially enhanced by TNFα treatment (Fig. 5D), but this effect was abolished due to the inhibition of IKKβ by wedelolactone (Fig. 5E), suggesting that IKKβ activation is required for TNFα-mediated interaction between β-TRCP2 and AR. In support of this notion, ectopic expression of IKKβ WT, but not the kinase-dead mutant increased β-TRCP2–AR interaction (Fig. 5F). Consistent with this result, depletion of both endogenous IKKβ and β-TRCP2 by siRNAs resulted in an increase of AR protein (Fig. 5G). Furthermore, depletion of both endogenous IKKβ and β-TRCP2 largely attenuated polyubiquitination of both ectopically expressed and endogenous AR (Fig. 5H; Supplementary Fig. S4A). Overexpression of wild-type β-TRCP2 augmented AR polyubiquitination and no such effect was observed for β-TRCP2ΔF-box, the F box deletion mutant of β-TRCP2 (Fig. 5I). Consistent with a role for β-TRCP2 in control of AR stability, the half-life of the endogenous AR protein was extended from approximately 8 to 24 hours after depletion of the endogenous β-TRCP in LNCaP cells (Fig. 5J–L). In addition, TNFα treatment enhanced β-TRCP2–AR interaction, but this effect was abolished by MAP3K7 KD (Supplementary Fig. S4B). Co-IP experiments confirmed that full-length AR and AR V7 variant can both bind to β-TRCP2 and IKKβ (Supplementary Fig. S4C). Similar to the effect of IKKβ KD (Supplementary Fig. S3E), KD of MAP3K7 resulted in the elevation of AR protein as co-KD of both MAP3K7 and β-TRCP2 did (Supplementary Fig. S4D). Taken together, these data indicate that β-TRCP E3 ligases play a critical role in mediating AR protein polyubiquitination and degradation induced by TNFα-mediated activation of MAP3K7 and IKKβ.

Figure 5.

β-TRCP E3 ubiquitin ligases mediate AR polyubiquitination and degradation. A, Sequence alignment of putative phosphodegron motifs targeted by IKKβ and β-TRCP in human AR protein and other known ubiquitination targets of β-TRCP. S, Serine; T, threonine; ψ, hydrophobic amino acid; X, any amino acid. B, 293T cells were transfected with the indicated plasmids for 24 hours and harvested for immunoblot analysis. C, Comparison of BTRC (β-TRCP1) and BTRC2 (β-TRCP2) mRNA expression in prostate cancer from the TCGA cohort (Firehose Legacy test, 499 cases). D, LNCaP cells were treated with vehicle (PBS) or TNFα (50 ng/mL) for 24 hours and harvested for co-IP assay and immunoblot analysis. E, LNCaP cell were treated with TNFα (50 ng/mL) and/or wedelolactone for 24 hours and harvested for co-IP and immunoblot analysis. F, 293T cells were transfected with indicated plasmids for 24 hours and harvested for co-IP assay and immunoblot analysis. G, LNCaP cells transfected with indicated siRNAs for 48 hours were treated with TNFα (50 ng/mL) for 24 hours and harvested for immunoblot analysis. H, LNCaP cells transfected with indicated siRNAs and HA-Ub for 40 hours were treated with MG132 for additional 8 hours and harvested for IP and immunoblot analysis. I, LNCaP cells transfected with indicated plasmids for 16 hours were treated with MG132 for 8 hours and harvested for ubiquitination assay. J, LNCaP cells were transfected with indicated siRNAs for 48 hours and harvested for immunoblot analysis. K, LNCaP cells transfected with indicated siRNAs for 48 hours as in J and treated with 20 μg/mL of cycloheximide (CHX). At indicated time points, cells were harvested for immunoblot analysis. L, Quantification of the intensity of the AR immunoblot bands shown in K. AR immunoblot band intensity was normalized to β-tubulin first and further normalized to the value at the 0-hours time point. Data shown as means ± SD (n = 3). **, P < 0.01; ***, P < 0.001.

Figure 5.

β-TRCP E3 ubiquitin ligases mediate AR polyubiquitination and degradation. A, Sequence alignment of putative phosphodegron motifs targeted by IKKβ and β-TRCP in human AR protein and other known ubiquitination targets of β-TRCP. S, Serine; T, threonine; ψ, hydrophobic amino acid; X, any amino acid. B, 293T cells were transfected with the indicated plasmids for 24 hours and harvested for immunoblot analysis. C, Comparison of BTRC (β-TRCP1) and BTRC2 (β-TRCP2) mRNA expression in prostate cancer from the TCGA cohort (Firehose Legacy test, 499 cases). D, LNCaP cells were treated with vehicle (PBS) or TNFα (50 ng/mL) for 24 hours and harvested for co-IP assay and immunoblot analysis. E, LNCaP cell were treated with TNFα (50 ng/mL) and/or wedelolactone for 24 hours and harvested for co-IP and immunoblot analysis. F, 293T cells were transfected with indicated plasmids for 24 hours and harvested for co-IP assay and immunoblot analysis. G, LNCaP cells transfected with indicated siRNAs for 48 hours were treated with TNFα (50 ng/mL) for 24 hours and harvested for immunoblot analysis. H, LNCaP cells transfected with indicated siRNAs and HA-Ub for 40 hours were treated with MG132 for additional 8 hours and harvested for IP and immunoblot analysis. I, LNCaP cells transfected with indicated plasmids for 16 hours were treated with MG132 for 8 hours and harvested for ubiquitination assay. J, LNCaP cells were transfected with indicated siRNAs for 48 hours and harvested for immunoblot analysis. K, LNCaP cells transfected with indicated siRNAs for 48 hours as in J and treated with 20 μg/mL of cycloheximide (CHX). At indicated time points, cells were harvested for immunoblot analysis. L, Quantification of the intensity of the AR immunoblot bands shown in K. AR immunoblot band intensity was normalized to β-tubulin first and further normalized to the value at the 0-hours time point. Data shown as means ± SD (n = 3). **, P < 0.01; ***, P < 0.001.

Close modal

The 297DSAGKS302 motif in AR acts as a phosphodegron for AR destruction

Phosphorylation of the substrate plays a key role in recognition, ubiquitination, and subsequent proteolysis by the E3 ubiquitin enzyme β-TRCP. The IKKβ phosphorylation consensus sequence DSψXXS/T, where ψ represents the hydrophobic amino acid and X represents any amino acid, often functions as a phosphodegron in β-TRCP degradation substrates, such as β-catenin and IκB (41–44). We identified two putative β-TRCP–targeting phosphodegrons 157DSAAPS162 and 297DSAGKS302 in AR protein (Figs. 5A and 6A). To determine whether these sites are required for β-TRCP–mediated AR degradation, we constructed two AR mutants, S158A/S162A and S298A/S302A, by mutating two serine (S) residues in each motif (Ser158/Ser162 and Ser298/Ser302) into alanine (A) (Fig. 6A). Western blotting analysis revealed that coexpression of IKKβ and β-TRCP2 decreased protein level of both WT AR and S158A/S162A, but not the S298A/S302A mutant (Fig. 6B). Analogous to this finding, co-IP assays showed that β-TRCP2 interacted with WT AR and the S158A/S162A mutant, but not the S298A/S302A mutant (Fig. 6C). Ubiquitination assays showed that coexpression of IKKβ and β-TRCP2 induced polyubiquitination of WT AR and the S158A/S162A mutant, but not the S298A/S302A mutant (Fig. 6D). Moreover, we demonstrated that TNFα treatment increased phosphorylation of both WT AR and the S158A/S162A mutant, but not the S298A/S302A mutant became resistant in 293T cells (Fig. 6E). Furthermore, we found that the half-life of the S298A/S302A mutant was much longer than those of WT AR and the S158A/S162A mutant (Fig. 6F and G). Together, these data indicate that the 297DSAGKS302 motif functions as a phosphodegron that is essential for AR destruction induced by IKKβ and β-TRCP2 in prostate cancer cells.

Figure 6.

The 297DSAGKS302 motif is important for IKKβ-mediated degradation of AR. A, Schematic diagram showing two putative IKKβ phosphorylation motifs (157DSAAPS162 and 297DSAGKS302) in AR protein and the alanine (A) mutants of these motifs. B, 293T cells were transfected with indicated plasmids for 24 hours and harvested for immunoblot analysis. C, 293T cells transfected with indicated plasmids for 24 hours were treated with MG132 for 8 hours and harvested for IP assay and immunoblot analysis. D, 293T cells were transfected with various plasmids as indicated. Sixteen hours post-transfection, cells were treated with MG132 for 8 hours. The cell lysates were IP with anti-AR and immunoblot analysis was performed. E, 293T cells transfected with indicated plasmids for 16 hours were treated with MG132 for 8 hours and harvested for IP and immunoblot analysis. F, PC-3 cells were transfected with AR-WT or two indicated AR mutants for 24 hours and treated with 20 μg/mL of cycloheximide (CHX). At different time points, cells were harvested for immunoblot analysis. G, Quantification of the intensity of immunoblot bands as shown in F. AR immunoblot band intensity was normalized to β-tubulin first and further normalized to the value at the 0-hours time point.

Figure 6.

The 297DSAGKS302 motif is important for IKKβ-mediated degradation of AR. A, Schematic diagram showing two putative IKKβ phosphorylation motifs (157DSAAPS162 and 297DSAGKS302) in AR protein and the alanine (A) mutants of these motifs. B, 293T cells were transfected with indicated plasmids for 24 hours and harvested for immunoblot analysis. C, 293T cells transfected with indicated plasmids for 24 hours were treated with MG132 for 8 hours and harvested for IP assay and immunoblot analysis. D, 293T cells were transfected with various plasmids as indicated. Sixteen hours post-transfection, cells were treated with MG132 for 8 hours. The cell lysates were IP with anti-AR and immunoblot analysis was performed. E, 293T cells transfected with indicated plasmids for 16 hours were treated with MG132 for 8 hours and harvested for IP and immunoblot analysis. F, PC-3 cells were transfected with AR-WT or two indicated AR mutants for 24 hours and treated with 20 μg/mL of cycloheximide (CHX). At different time points, cells were harvested for immunoblot analysis. G, Quantification of the intensity of immunoblot bands as shown in F. AR immunoblot band intensity was normalized to β-tubulin first and further normalized to the value at the 0-hours time point.

Close modal

Degradation-resistant AR mutant enhances AR activity and prostate cancer growth

We further examined whether AR degradation mediated by MAP3K7, IKKβ, and β-TRCP also affects AR transcriptional activity and AR-dependent prostate cancer growth. Luciferase reporter assays showed that the activities of both PSA promoter/enhancer (PSA-Luc) and composite three-copy androgen response element (3X ARE-Luc)–based luciferase reporters were enhanced by KD of IKKβ, β-TRCP1/2 or together in LNCaP cells cultured in androgen-depleted TNFα-containing medium, and similar results were obtained in cells treated with the synthetic androgen mibolerone (Fig. 7A). KD of IKKβ, β-TRCP1/2 or together also substantially increased expression of AR target genes including KLK3 (PSA), TMPRSS2, NKX3.1 and KLK2 (HK2) in LNCaP cells treated with or without mibolerone (Fig. 7B). Although MAP3K7 KD significantly increased AR target gene expression, this effect was abolished by overexpression of S177/181E, a phospho-mimicking mutant of IKKβ (Supplementary Figs. S2K, S5A, and S5B). We also infected LNCaP cells with lentivirus-expressing empty vector (EV), wild-type AR (AR-WT), S158A/S162A or S298A/S302A mutant. Ectopic expression of both AR-WT and the two mutants significantly enhanced proliferation in LNCaP cells without TNFα treatment (Fig. 7C, left); however, TNFα treatment largely inhibited the growth of LNCaP cells expressing EV, WT AR, and the S158A/S162A mutant, but not S298A/S302A mutant (Fig. 7C, right). To validate our finding in vivo, LNCaP cells expressing EV, WT AR or the S298A/S302A mutant were subcutaneously injected into SCID mice followed by TNFα treatment. We demonstrated that tumors expressing the S298A/S302A mutant grew much faster than tumors expressing EV or WT AR (Fig. 7D–F). These results indicate that AR degradation induced by IKKβ and β-TRCP suppresses AR activity and prostate cancer cell growth in vitro and in mice.

Figure 7.

Degradation-resistant AR mutant increases AR transcriptional activity and prostate cancer growth in vitro and in mice. A, LNCaP cells transfected with indicated siRNAs in combination with PSA-Luc or 3X ARE-Luc and Renilla luciferase reporters were cultured in charcoal-stripped (androgen-depleted) TNFα-containing media for 32 hours. Cells were further treated with vehicle (ethanol) or mibolerone (Mib) for 8 hours and harvested for dual luciferase assay. Data are normalized to the value of nonspecific (N.S.) siRNA group and shown as means ± SD (n = 6). B, LNCaP cells were transfected with siRNAs and treated with Mib as in A. Cells were harvested for qRT-PCR analysis of expression of KLK3, TMPRSS2, NKX3.1, and HK2 mRNA. Data shown as means ± SD (n = 6). C, LNCaP cells infected with lentivirus-expressing control vector, WT AR, S158A/S162A, or S298A/S302A mutants were treated with or without TNFα (50 ng/mL). Cell proliferation was measured using MTS assay at the indicated time points. The absorbance values were normalized with day-0 value in each group and are presented as fold change. Data are means ± SD (n = 6). D–F, LNCaP cells infected with lentivirus-expressing control plasmid, WT AR, or S298A/S302A mutant were injected subcutaneously into the left flank of SCID mice and mice were treated with TNFα (10 ng/kg/d) 6 days after cell injection. The tumor volume of each xenograft at each time point (D), the weight of tumors at the end of treatment (E), and tumors at the end of treatment (F) are shown. Data are means ± SD (n = 7). The unpaired Student t test. G, A hypothetical model for proinflammation cytokine TNFα-induced AR degradation. Left, upon the binding of the receptors by the proinflammation cytokines such as TNFα, activation of the MAP3K7–IKKβ signaling axis leads to the degradation of IκB and activation of the NF-κB pathway as well as the degradation of AR, thereby inhibiting AR signaling and prostate cancer progression. Right, the MAP3K7 gene is frequently deleted in human prostate cancer. Although loss of MAP3K7 leads to inhibition of the NF-κB pathway, it results in the stabilization and elevation of AR protein, thereby promoting AR activation and castration-resistant progression of prostate cancer. *, P < 0.05; **, P < 0.01.

Figure 7.

Degradation-resistant AR mutant increases AR transcriptional activity and prostate cancer growth in vitro and in mice. A, LNCaP cells transfected with indicated siRNAs in combination with PSA-Luc or 3X ARE-Luc and Renilla luciferase reporters were cultured in charcoal-stripped (androgen-depleted) TNFα-containing media for 32 hours. Cells were further treated with vehicle (ethanol) or mibolerone (Mib) for 8 hours and harvested for dual luciferase assay. Data are normalized to the value of nonspecific (N.S.) siRNA group and shown as means ± SD (n = 6). B, LNCaP cells were transfected with siRNAs and treated with Mib as in A. Cells were harvested for qRT-PCR analysis of expression of KLK3, TMPRSS2, NKX3.1, and HK2 mRNA. Data shown as means ± SD (n = 6). C, LNCaP cells infected with lentivirus-expressing control vector, WT AR, S158A/S162A, or S298A/S302A mutants were treated with or without TNFα (50 ng/mL). Cell proliferation was measured using MTS assay at the indicated time points. The absorbance values were normalized with day-0 value in each group and are presented as fold change. Data are means ± SD (n = 6). D–F, LNCaP cells infected with lentivirus-expressing control plasmid, WT AR, or S298A/S302A mutant were injected subcutaneously into the left flank of SCID mice and mice were treated with TNFα (10 ng/kg/d) 6 days after cell injection. The tumor volume of each xenograft at each time point (D), the weight of tumors at the end of treatment (E), and tumors at the end of treatment (F) are shown. Data are means ± SD (n = 7). The unpaired Student t test. G, A hypothetical model for proinflammation cytokine TNFα-induced AR degradation. Left, upon the binding of the receptors by the proinflammation cytokines such as TNFα, activation of the MAP3K7–IKKβ signaling axis leads to the degradation of IκB and activation of the NF-κB pathway as well as the degradation of AR, thereby inhibiting AR signaling and prostate cancer progression. Right, the MAP3K7 gene is frequently deleted in human prostate cancer. Although loss of MAP3K7 leads to inhibition of the NF-κB pathway, it results in the stabilization and elevation of AR protein, thereby promoting AR activation and castration-resistant progression of prostate cancer. *, P < 0.05; **, P < 0.01.

Close modal

It is well established that AR plays essential roles in prostate cancer survival (45, 46). Acute inflammation is critical for host defense against pathogen exposure and disease conditions (3). In the present study, we identify AR protein as a degradation target of inflammatory signaling. We first noticed that decreased expression of AR associates with inflammation in prostate cancer patient specimens. This result is corroborated by our findings that the treatment of prostate cancer cells with the proinflammatory cytokine TNFα induces proteasomal degradation of AR protein and such effects are mediated by the activation of the MAP3K7–IKKβ inflammatory signaling axis (Fig. 7G, left). This observation provides a possible explanation for the seminal discovery that the AR protein level is lower in the majority of prostatic cells within the PIA lesion compared with areas without inflammation (25). We further show that genetic deletion of MAP3K7 or pharmacological inhibition of IKKβ blocks inflammation signaling-induced AR degradation, thereby providing a plausible explanation for the mixed outcome of the rationalized anti-inflammation chemoprevention trials in prostate cancer (2) because inhibition of inflammation may lead to elevation of AR protein level that thereby antagonizes the potential inhibition of prostate cancer growth and progression imposed by anti-inflammation drugs.

It has long been recognized that activation of the NF-κB pathway due to the aberrant activation of the MAP3K7-IKKβ inflammatory signaling is important for the survival of cells exposed with inflammatory cytokines (47). Surprisingly, in both TRAMP and myc-CaP prostate cancer mouse models conditional deletion of Ikbkb (encoding murine Ikkβ) in prostate epithelial cells failed to trigger tumor regression in both intact and castrated mice, highlighting the relevance of non-cell autonomous pathways in driving castration-resistant progression of prostate cancer (48). However, an alternative explanation could be that although loss of IKKβ results in the inhibition of the NF-κB signaling, it may also augment other survival pathways that can compensate NF-κB inhibition in a cell autonomous manner. We demonstrate that IKKβ binds to AR and promotes AR protein polyubiquitination and proteasomal degradation and this process is mediated by β-TRCP E3 ubiquitin ligases. We provide evidence that depletion of IKKβ by shRNAs or its inhibition by pharmacological inhibitor induces AR stabilization. Importantly, we further show that blocking AR degradation mediated by MAP3K7 and IKKβ promotes prostate cancer cell growth in vitro and in mice. Thus, our findings identify a previously unrecognized tumor-suppressive activity of MAP3K7 and IKKβ in mediating AR protein destruction and prostate cancer cell growth (Fig. 7G, left).

It is generally accepted that activation of the NF-κB pathway by the MAP3K7–IKKβ signaling axis promotes cancer survival and progression. It seems counterintuitive that the MAP3K7 gene is frequently deleted in prostate cancer patient specimens and that MAP3K7 deletion associates with high-grade prostate cancer (13, 29). Given the predominant role of AR in prostate cancer cell growth and survival, our findings support a model that loss of MAP3K7 favors prostate cancer progression due to enhanced AR activity, thereby compensating the inhibition of the NF-κB survival pathway (Fig. 7G, right), a discovery consistent with the findings in a recent report (32). Consistent with this notion, it has been shown previously that loss of MAP3K7 increases proliferation, migration, and invasion of prostate cancer cells (30). Moreover, deletion of the MAP3K7 gene often co-occurs with deletion of the chromatin modifier gene CHD1 in prostate cancer (13, 31, 49), and coordinate loss of MAP3K7 and CHD1 promotes aggressive prostate cancer (31). Intriguingly, it has been shown recently that CHD1 loss induces AR cistrome reprogramming (50). Thus, it is possible that increased AR expression due to loss of MAP3K7 and aberrant AR cistrome due to deletion of CHD1 may cooperate to promote aggressive phenotypes in prostate cancer and further investigation is warranted.

We have shown previously that treatment of prostate cancer cells with TNFα decreased AR expression and that the TNFα receptor-associated death domain (TRADD) protein plays an important role in mediating TNFα inhibition of AR expression (51). However, the downstream signaling pathway that mediates TNFα-induced downregulation of AR expression was unclear. Our present study not only uncovers a pivotal role of the MAP3K7–IKKβ signaling cascade in regulation of AR protein degradation and stability, but also suggests that the TNFα receptor–adaptor proteins such as TRADD may also play a role in TNFα-induced AR protein degradation.

In summary, we demonstrate that the proinflammatory cytokines such as TNFα promote AR protein degradation in prostate cancer cells and this effect is mediated by the MAP3K7–IKKβ signaling. Mechanistically, we show that IKKβ binds to AR and triggers phosphorylation-dependent interaction of AR with the E3 ubiquitin ligases β-TRCP1/2 and subsequent polyubiquitination and proteasomal degradation of AR. Our findings reveal a previously uncharacterized tumor-suppressive role of the MAP3K7–IKKβ signaling in inducing the destruction of AR, a pro-cancer protein in prostate cancer. The relevance of this discovery is supported by the observations in the clinical specimens that the MAP3K7 gene is often deleted in patients with prostate cancer. Thus, identification of IKKβ-mediated β-TRCP–dependent AR degradation enhances our understanding of how AR activity is influenced by tumor microenvironment especially inflammation. These results also suggest that aberrantly elevated AR protein can not only serve as a valuable prognostic factor for the progression of MAP3K7-deficient prostate cancer, but also is a viable therapeutic target of this subtype of prostate cancer.

R.J. Karnes reports other support from Decipher Biosciences outside the submitted work. No disclosures were reported by the other authors.

Z. Huang: Data curation, formal analysis, validation, investigation, methodology, writing–original draft. B. Tang: Resources, data curation, investigation, writing–original draft, writing–review and editing. Y. Yang: Resources, data curation, investigation, methodology. Z. Yang: Resources, data curation, software, formal analysis, investigation, methodology. L. Shi: Investigation, methodology. Y. Bai: Investigation, methodology. B. Yan: Investigation, methodology. R.J. Karnes: Investigation, methodology. J. Zhang: Investigation, methodology. R. Jimenez: Investigation, methodology. L. Wang: Software, methodology. Q. Wei: Data curation, methodology. J. Yang: Resources, data curation, methodology. W. Xu: Conceptualization, resources. Z. Jia: Conceptualization, resources. H. Huang: Conceptualization, supervision, funding acquisition, writing–review and editing.

The authors thank Dr. Jiwang Zhang from the Loyola University Medical Center for kindly providing Map3k7 (Tak1) loxp/loxp conditional knockout mice. This work was supported by funding from the Mayo Clinic Foundation (to H. Huang).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Sfanos
KS
,
Yegnasubramanian
S
,
Nelson
WG
,
De Marzo
AM
. 
The inflammatory microenvironment and microbiome in prostate cancer development
.
Nat Rev Urol
2018
;
15
:
11
24
.
2.
Bardia
A
,
Platz
EA
,
Yegnasubramanian
S
,
De Marzo
AM
,
Nelson
WG
. 
Anti-inflammatory drugs, antioxidants, and prostate cancer prevention
.
Curr Opin Pharmacol
2009
;
9
:
419
26
.
3.
De Marzo
AM
,
Platz
EA
,
Sutcliffe
S
,
Xu
J
,
Grönberg
H
,
Drake
CG
, et al
Inflammation in prostate carcinogenesis
.
Nat Rev Cancer
2007
;
7
:
256
69
.
4.
Sato
S
,
Sanjo
H
,
Takeda
K
,
Ninomiya-Tsuji
J
,
Yamamoto
M
,
Kawai
T
, et al
Essential function for the kinase TAK1 in innate and adaptive immune responses
.
Nat Immunol
2005
;
6
:
1087
95
.
5.
Ninomiya-Tsuji
J
,
Kishimoto
K
,
Hiyama
A
,
Inoue
J
,
Cao
Z
,
Matsumoto
K
. 
The kinase TAK1 can activate the NIK-I kappaB as well as the MAP kinase cascade in the IL-1 signalling pathway
.
Nature
1999
;
398
:
252
6
.
6.
Yamaguchi
K
,
Shirakabe
K
,
Shibuya
H
,
Irie
K
,
Oishi
I
,
Ueno
N
, et al
Identification of a member of the MAPKKK family as a potential mediator of TGF-beta signal transduction
.
Science
1995
;
270
:
2008
11
.
7.
Yan
J
,
Xiang
J
,
Lin
Y
,
Ma
J
,
Zhang
J
,
Zhang
H
, et al
Inactivation of BAD by IKK inhibits TNFα-induced apoptosis independently of NF-κB activation
.
Cell
2013
;
152
:
304
15
.
8.
Hutti
JE
,
Turk
BE
,
Asara
JM
,
Ma
A
,
Cantley
LC
,
Abbott
DW
. 
IkappaB kinase beta phosphorylates the K63 deubiquitinase A20 to cause feedback inhibition of the NF-kappaB pathway
.
Mol Cell Biol
2007
;
27
:
7451
61
.
9.
Xie
M
,
Zhang
D
,
Dyck
JR
,
Li
Y
,
Zhang
H
,
Morishima
M
, et al
A pivotal role for endogenous TGF-beta–activated kinase-1 in the LKB1/AMP-activated protein kinase energy-sensor pathway
.
PNAS
2006
;
103
:
17378
83
.
10.
Ohkawara
B
,
Shirakabe
K
,
Hyodo-Miura
J
,
Matsuo
R
,
Ueno
N
,
Matsumoto
K
, et al
Role of the TAK1–NLK–STAT3 pathway in TGF-beta–mediated mesoderm induction
.
Genes Dev
2004
;
18
:
381
6
.
11.
Hoffmann
A
,
Preobrazhenska
O
,
Wodarczyk
C
,
Medler
Y
,
Winkel
A
,
Shahab
S
, et al
Transforming growth factor-beta–activated kinase-1 (TAK1), a MAP3K, interacts with Smad proteins and interferes with osteogenesis in murine mesenchymal progenitors
.
J Biol Chem
2005
;
280
:
27271
83
.
12.
Armenia
J
,
Wankowicz
SAM
,
Liu
D
,
Gao
J
,
Kundra
R
,
Reznik
E
, et al
The long tail of oncogenic drivers in prostate cancer
.
Nat Genet
2018
;
50
:
645
51
.
13.
Cancer Genome Atlas Research Network
. 
The Molecular taxonomy of primary prostate cancer
.
Cell
2015
;
163
:
1011
25
.
14.
Chen
S
,
Gulla
S
,
Cai
C
,
Balk
SP
. 
Androgen receptor serine 81 phosphorylation mediates chromatin binding and transcriptional activation
.
J Biol Chem
2012
;
287
:
8571
83
.
15.
Ponguta
LA
,
Gregory
CW
,
French
FS
,
Wilson
EM
. 
Site-specific androgen receptor serine phosphorylation linked to epidermal growth factor-dependent growth of castration-recurrent prostate cancer
.
J Biol Chem
2008
;
283
:
20989
1001
.
16.
Lin
HK
,
Wang
L
,
Hu
YC
,
Altuwaijri
S
,
Chang
C
. 
Phosphorylation-dependent ubiquitylation and degradation of androgen receptor by Akt require Mdm2 E3 ligase
.
EMBO J
2002
;
21
:
4037
48
.
17.
Sarkar
S
,
Brautigan
DL
,
Parsons
SJ
,
Larner
JM
. 
Androgen receptor degradation by the E3 ligase CHIP modulates mitotic arrest in prostate cancer cells
.
Oncogene
2014
;
33
:
26
33
.
18.
Qi
J
,
Tripathi
M
,
Mishra
R
,
Sahgal
N
,
Fazil
L
,
Ettinger
S
, et al
The E3 ubiquitin ligase Siah2 contributes to castration-resistant prostate cancer by regulation of androgen receptor transcriptional activity
.
Cancer Cell
2013
;
23
:
332
46
.
19.
Chen
S
,
Bohrer
LR
,
Rai
AN
,
Pan
Y
,
Gan
L
,
Zhou
X
, et al
Cyclin-dependent kinases regulate epigenetic gene silencing through phosphorylation of EZH2
.
Nat Cell Biol
2010
;
12
:
1108
14
.
20.
Tang
M
,
Wei
X
,
Guo
Y
,
Breslin
P
,
Zhang
S
,
Zhang
S
, et al
TAK1 is required for the survival of hematopoietic cells and hepatocytes in mice
.
J Exp Med
2008
;
205
:
1611
9
.
21.
Wu
X
,
Wu
J
,
Huang
J
,
Powell
WC
,
Zhang
J
,
Matusik
RJ
, et al
Generation of a prostate epithelial cell-specific Cre transgenic mouse model for tissue-specific gene ablation
.
Mech Dev
2001
;
101
:
61
9
.
22.
Blee
AM
,
He
Y
,
Yang
Y
,
Ye
Z
,
Yan
Y
,
Pan
Y
, et al
TMPRSS2-ERG controls luminal epithelial lineage and antiandrogen sensitivity in PTEN and TP53-mutated prostate cancer
.
Clin Cancer Res
2018
;
24
:
4551
65
.
23.
Jin
X
,
Ding
D
,
Yan
Y
,
Li
H
,
Wang
B
,
Ma
L
, et al
Phosphorylated RB promotes cancer immunity by inhibiting NF-kappaB activation and PD-L1 expression
.
Mol Cell
2019
;
73
:
22
35
.
24.
Zhang
H
,
Pan
Y
,
Zheng
L
,
Choe
C
,
Lindgren
B
,
Jensen
ED
, et al
FOXO1 inhibits Runx2 transcriptional activity and prostate cancer cell migration and invasion
.
Cancer Res
2011
;
71
:
3257
67
.
25.
De Marzo
AM
,
Marchi
VL
,
Epstein
JI
,
Nelson
WG
. 
Proliferative inflammatory atrophy of the prostate: implications for prostatic carcinogenesis
.
Am J Pathol
1999
;
155
:
1985
92
.
26.
Alexander
RB
,
Ponniah
S
,
Hasday
J
,
Hebel
JR
. 
Elevated levels of proinflammatory cytokines in the semen of patients with chronic prostatitis/chronic pelvic pain syndrome
.
Urology
1998
;
52
:
744
9
.
27.
Nakashima
J
,
Tachibana
M
,
Ueno
M
,
Miyajima
A
,
Baba
S
,
Murai
M
. 
Association between tumor necrosis factor in serum and cachexia in patients with prostate cancer
.
Clin Cancer Res
1998
;
4
:
1743
8
.
28.
Lee
DK
,
Chang
C
. 
Endocrine mechanisms of disease: Expression and degradation of androgen receptor: mechanism and clinical implication
.
J Clin Endocrinol Metab
2003
;
88
:
4043
54
.
29.
Liu
W
,
Chang
BL
,
Cramer
S
,
Koty
PP
,
Li
T
,
Sun
J
, et al
Deletion of a small consensus region at 6q15, including the MAP3K7 gene, is significantly associated with high-grade prostate cancers
.
Clin Cancer Res
2007
;
13
:
5028
33
.
30.
Wu
M
,
Shi
L
,
Cimic
A
,
Romero
L
,
Sui
G
,
Lees
CJ
, et al
Suppression of Tak1 promotes prostate tumorigenesis
.
Cancer Res
2012
;
72
:
2833
43
.
31.
Rodrigues
LU
,
Rider
L
,
Nieto
C
,
Romero
L
,
Karimpour-Fard
A
,
Loda
M
, et al
Coordinate loss of MAP3K7 and CHD1 promotes aggressive prostate cancer
.
Cancer Res
2015
;
75
:
1021
34
.
32.
Jillson
LK
,
Rider
LC
,
Rodrigues
LU
,
Romero
L
,
Karimpour-Fard
A
,
Nieto
C
, et al
MAP3K7 loss drives enhanced androgen signaling and independently confers risk of recurrence in prostate cancer with joint loss of CHD1
.
Mol Cancer Res
2021
Apr 12. [Epub ahead of print]
.
33.
Chen
Y
,
Chi
P
,
Rockowitz
S
,
Iaquinta
PJ
,
Shamu
T
,
Shukla
S
, et al
ETS factors reprogram the androgen receptor cistrome and prime prostate tumorigenesis in response to PTEN loss
.
Nat Med
2013
;
19
:
1023
9
.
34.
Abida
W
,
Cyrta
J
,
Heller
G
,
Prandi
D
,
Armenia
J
,
Coleman
I
, et al
Genomic correlates of clinical outcome in advanced prostate cancer
.
Proc Natl Acad Sci U S A
2019
;
116
:
11428
36
.
35.
Takaesu
G
,
Surabhi
RM
,
Park
KJ
,
Ninomiya-Tsuji
J
,
Matsumoto
K
,
Gaynor
RB
. 
TAK1 is critical for IkappaB kinase-mediated activation of the NF-kappaB pathway
.
J Mol Biol
2003
;
326
:
105
15
.
36.
Prescott
JA
,
Cook
SJ
. 
Targeting IKKβ in cancer: challenges and opportunities for the therapeutic utilisation of IKKβ inhibitors
.
Cells
2018
;
7
:
115
.
37.
Zhang
Y
,
Lapidus
RG
,
Liu
P
,
Choi
EY
,
Adediran
S
,
Hussain
A
, et al
Targeting IkappaB kinase beta/NF-kappaB signaling in human prostate cancer by a novel ikappab kinase beta inhibitor CmpdA
.
Mol Cancer Ther
2016
;
15
:
1504
14
.
38.
Huang
H
,
Tindall
DJ
. 
Regulation of FOXO protein stability via ubiquitination and proteasome degradation
.
Biochim Biophys Acta
2011
;
1813
:
1961
4
.
39.
Karin
M
,
Ben-Neriah
Y
. 
Phosphorylation meets ubiquitination: the control of NF-[kappa]B activity
.
Annu Rev Immunol
2000
;
18
:
621
63
.
40.
Hu
MC
,
Lee
DF
,
Xia
W
,
Golfman
LS
,
Ou-Yang
F
,
Yang
JY
, et al
IkappaB kinase promotes tumorigenesis through inhibition of forkhead FOXO3a
.
Cell
2004
;
117
:
225
37
.
41.
Liu
C
,
Kato
Y
,
Zhang
Z
,
Do
VM
,
Yankner
BA
., 
He X. beta-Trcp couples beta-catenin phosphorylation-degradation and regulates Xenopus axis formation
.
PNAS
1999
;
96
:
6273
8
.
42.
Hart
M
,
Concordet
JP
,
Lassot
I
,
Albert
I
,
del los Santos
R
,
Durand
H
, et al
The F-box protein beta-TrCP associates with phosphorylated beta-catenin and regulates its activity in the cell
.
Curr Biol
1999
;
9
:
207
10
.
43.
Rothwarf
DM
,
Zandi
E
,
Natoli
G
,
Karin
M
. 
IKK-gamma is an essential regulatory subunit of the IkappaB kinase complex
.
Nature
1998
;
395
:
297
300
.
44.
Malinin
NL
,
Boldin
MP
,
Kovalenko
AV
,
Wallach
D
. 
MAP3K-related kinase involved in NF-kappaB induction by TNF, CD95, and IL-1
.
Nature
1997
;
385
:
540
4
.
45.
Watson
PA
,
Arora
VK
,
Sawyers
CL
. 
Emerging mechanisms of resistance to androgen receptor inhibitors in prostate cancer
.
Nat Rev Cancer
2015
;
15
:
701
11
.
46.
Wen
S
,
Niu
Y
,
Huang
H
. 
Posttranslational regulation of androgen dependent and independent androgen receptor activities in prostate cancer
.
Asian J Urol
2020
;
7
:
203
18
.
47.
Karin
M
,
Cao
Y
,
Greten
FR
,
Li
ZW
. 
NF-kappaB in cancer: from innocent bystander to major culprit
.
Nat Rev Cancer
2002
;
2
:
301
10
.
48.
Ammirante
M
,
Luo
JL
,
Grivennikov
S
,
Nedospasov
S
,
Karin
M
. 
B-cell–derived lymphotoxin promotes castration-resistant prostate cancer
.
Nature
2010
;
464
:
302
5
.
49.
Boysen
G
,
Barbieri
CE
,
Prandi
D
,
Blattner
M
,
Chae
SS
,
Dahija
A
, et al
SPOP mutation leads to genomic instability in prostate cancer
.
Elife
2015
;
4
:
e09207
.
50.
Augello
MA
,
Liu
D
,
Deonarine
LD
,
Robinson
BD
,
Huang
D
,
Stelloo
S
, et al
CHD1 loss alters AR binding at lineage-specific enhancers and modulates distinct transcriptional programs to drive prostate tumorigenesis
.
Cancer Cell
2019
;
35
:
603
17
.
51.
Wang
D
,
Montgomery
RB
,
Schmidt
LJ
,
Mostaghel
EA
,
Huang
H
,
Nelson
PS
, et al
Reduced tumor necrosis factor receptor-associated death domain expression is associated with prostate cancer progression
.
Cancer Res
2009
;
69
:
9448
56
.