Androgen receptor signaling fuels prostate cancer and is a major therapeutic target. However, mechanisms of resistance to therapeutic androgen ablation are not well understood. Here, using a prostate cancer mouse model, Ptenpc−/−, carrying a prostate epithelial-specific Pten deletion, we show that the αvβ6 integrin is required for tumor growth in vivo of castrated as well as of noncastrated mice. We describe a novel signaling pathway that couples the αvβ6 integrin cell surface receptor to androgen receptor via activation of JNK1 and causes increased nuclear localization and activity of androgen receptor. This downstream kinase activation by αvβ6 is specific for JNK1, with no involvement of p38 or ERK kinase. In addition, differential phosphorylation of Akt is not observed under these conditions, nor is cell morphology affected by αvβ6 expression. This pathway, which is specific for αvβ6, because it is not regulated by a different αv-containing integrin, αvβ3, promotes upregulation of survivin, which in turn supports anchorage-independent growth of αvβ6-expressing cells. Consistently, both αvβ6 and survivin are significantly increased in prostatic adenocarcinoma, but are not detected in normal prostatic epithelium. Neither XIAP nor Bcl-2 is affected by αvβ6 expression. In conclusion, we show that αvβ6 expression is required for prostate cancer progression, including castrate-resistant prostate cancer; mechanistically, by promoting activation of JNK1, the αvβ6 integrin causes androgen receptor–increased activity in the absence of androgen and consequent upregulation of survivin. These preclinical results pave the way for further clinical development of αvβ6 antagonists for prostate cancer therapy. Cancer Res; 76(17); 5163–74. ©2016 AACR.

Despite considerable progress in the successful management of early-stage disease, prostate cancer remains a life-threatening disease in its advanced stage (1–3). Most, if not all, of aberrant cellular signaling in prostate cancer can be traced back to aberrant androgen receptor (AR) signaling, which dictates disease onset and progression. AR belongs to the nuclear receptor gene superfamily and, in response to binding to its ligand, i.e., androgen, mediates transcription of various genes important for disease maintenance (4). The gain of resistance to therapeutic androgen ablation in its advanced phases, castrate-resistant prostate cancer (CRPC; refs. 5, 6), does not indicate dispensability of AR activity, but rather the acquisition of ancillary mechanism(s) that fuel AR signaling in the absence of androgen (7), such as AR locus amplification, AR variants or mutants that display transcriptional activity without ligand binding, or inappropriate AR activation by ligand-independent and non–androgen-mediated pathways (4, 8).

Changes in the cell matrix interaction during prostate tumorigenesis are involved in disease behaviors and resistance to therapy (9, 10). In this context, prostate cancer cells often overexpress members of the integrin family of adhesion receptors. Integrins not only mediate cell-to-cell communication, but also play a pivotal role in transducing environmental cues to downstream signaling molecules of cell proliferation, survival, and invasiveness, including MAPK, focal adhesion kinase (FAK), src, Akt, and JNK (11, 12). In particular, the αvβ6 integrin is associated with neoplastic and metastatic phenotypes in various cancers (13, 14). It functions as a receptor for extracellular matrix (ECM) proteins, including latency associated peptide transforming growth factor β (LAP-TGFβ), fibronectin (FN), tenascin, and vitronectin (VN), and has been linked to activation of FAK and MAPK (15). However, how the precise arrangement of integrin signaling contributes to resistance to androgen ablation is still largely unknown.

Another downstream protein associated with integrin activation is survivin (16). Survivin is a member of the inhibitor of apoptosis (IAP) family whose expression is associated with conservation of cell division (17). Survivin is not usually expressed in normal prostate epithelium, but gradually increases during tumorigenesis and shows the highest level in lymph node metastasis (18). It has been also reported that androgen induces survivin expression and AR inhibitors prevent survivin expression (19). Intratumoral injection of an antisense reagent against survivin mRNA or an adenovirus expressing a dominant negative form of survivin, T34A, inhibits tumor growth and enhances sensitivity to androgen ablation (19, 20).

In this study, we investigated a potential cross-talk between integrins, AR signaling and survivin in prostate cancer models. We found that αvβ6 integrin expression is required for prostate cancer progression in vivo, including CRPC, and this pathway mediates activation of JNK1, nuclear accumulation and increased activity of AR in the absence of androgen.

Reagents and antibodies

Synthetic androgen R1881 was from PerkinElmer. Enzalutamide was from Selleck Chemicals. The following rabbit Abs (pAbs) were used: AR (N20), phospho-ERK1 (Santa Cruz Biotechnology); Smad3 (Invitrogen); p-src (Y416), Akt, p-Akt, p-JNK and p-p38 (Cell Signaling Technology); prostate-specific antigen (PSA; DAKO Cytomation). A goat Ab against p38 was from Santa Cruz Biotechnology. A rabbit monoclonal Ab against β6 integrin, B1, was from Dr. Dean Sheppard. The following mouse monoclonal Abs (mAbs): AR (441; Santa Cruz Biotechnology); β1 integrin (C-18) and JNK (BD Biosciences); c-src (Cell Signaling Technology); β1 integrin (TS2/16) and β3 integrin (AP3) are from ATCC; β5 integrin (P1F6) from Life Technologies, Inc.; αv integrin (VNR147) and α5 integrin (P1D6) are from Life Technologies; CK8 (Boehringer Mannheim), CK18 (Sigma); β6 integrin Abs 6.2A1, ch2A1 and 6.3G9 were previously described (21); β6 integrin Ab 10D5 from Chemicon. Purified nonimmune mouse IgGs (mIgG) from Pierce, 1E6 (IgG1) from Biogen and 1C10 were used as negative controls.

Cells and culture conditions

LNCaP, PC3, and RWPE cells were obtained from the ATCC, and C4-2B was obtained from UroCor and cultured as previously described (22, 23). LNCaP stable cell lines expressing full-length β3, β6, or empty vector have been described previously (24). PC3 and C4-2B cells stably transfected with human β6-integrin cDNA in pBABE retroviral vector were used (23). Authentication of the cell lines was provided with their purchase. Cellular morphology and anchorage-independent growth assay were visualized using an Olympus IX71 inverted microscope with IPLab V3.55 (Scanalytics, Inc.).

Human tissue specimens

All tissues in this study were discarded, coded, and unidentifiable specimens. Forty-seven human adenocarcinoma tissue specimens from radical prostatectomies provided by: The Cooperative Human Tissue Network, which is funded by the National Cancer Institute (other investigators may have received specimens from the same subjects); the Department of Pathology, University of Massachusetts Medical School and the Department of Pathology, Yale University and were processed according to Institution-approved protocols. Eight normal prostatic tissue sections were obtained from human biopsies. Thirteen additional prostate specimens were snap-frozen in liquid nitrogen for protein extraction. Tissues were processed as previously described (25).

Constructs and cell transfection

Adenoviruses, containing wild-type survivin (WT), dominant negative survivin (T34A) or empty vector (pAd) have been described previously (26). αvβ6-LNCaP cells were transiently transfected with adenoviral constructs for 17 hours. All three constructs contain the GFP gene for determination of transfection efficiency. C4-2B cells were stably transfected human β6 and β3 integrin cDNA generated as described before (24). Clones were selected and maintained in culture medium containing 5% FBS and 100 μg/mL G418. Retroviral constructs, containing wild-type JNK1 or an NH2-terminal JNK-binding domain (JBD), and the transient transfections have been described previously (27). αvβ6-C4-2B, αvβ3-C4-2B or mock-C4-2B cells were transiently transfected with WT-JNK- or JBD-viral construct to determine the effect of JNK on AR activity. WT-JNK (pBABE-puro-JNK1) or JBD containing retrovirus (babe-puro-JBD) were wrapped with VSVG and GAGPol using phoenix cells (ATCC, CRL-3213). The transient transfection was performed with 8 μg/mL of polybrene for 48 hours, then cells were starved in 2% charcoal-stripped (CS) serum-containing medium for 24 hours followed by stimulation with 1 nmol/L R1881 or ethanol for 24 hours.

siRNA transfection

Cells were transfected with β6 integrin siRNA duplexes (IDT Inc.). Two β6 siRNA duplexes, D1 and D2, were described before (22). siRNA duplexes were transfected using oligofectamine at a final concentration of 20 nmol/L. Forty-eight hours after transfection, cells were harvested and analyzed by FACS to confirm downregulation of β6 integrin and to evaluate anchorage-independent cell growth. For downregulation of AR, siRNA smartpool and nonsilencing siRNA (Dharmacon) were used according to the manufacturer's instruction. The protocol used for downregulation of survivin by siRNA in αvβ6-LNCaP cells was based on published methods (28).

Luciferase assay and cell adhesion assay

Luciferase activity was measured as described previously (29). Cell adhesion protocols used here have been described previously (30). For JNK activation assay, LNCaP, C4-2B or PC3 cells stably transfected with pBabe-β6-integrin or pBabe-empty vector were starved overnight in serum-free RPMI culture medium. Cells were seeded on culture plates coated with FN (3 μg/mL), LAP-TGFβ1 (0.5 μg/mL) or BSA. Cells (1 × 106) were plated on each substrate for 1 hour in serum-free RPMI culture medium before cell lysis.

Cell proliferation assay

Cells were seeded in a 96-well plate for 24 hours, starved with medium containing 2% CS serum for 24 hours. Cells were stimulated with either 1 nmol/L R1881 or incubated with the same volume of ethanol and then cultured for 72 hours before performing Sulforhodamine B (SRB) staining.

IHC

IHC staining using formalin-fixed paraffin-embedded specimens was performed as described elsewhere (31). For detection of survivin, IHC was performed according to published protocols (32).

Immunofluorescence

Primary epithelial cells isolated as described (30), were probed using Abs specific to CK8 or CK18, markers of prostate epithelial cells by indirect IF staining. AR nuclear localization was analyzed using IF staining. Cells were seeded on poly-l-Lysine (1 mg/mL) coated coverslips for 2 days. After starvation for 24 hours in 2% CS serum contained medium, cells were grown in the presence of ethanol or R1881 (1 nmol/L) for 2 hours. Cells were probed with an AR-specific Ab N20 (1:500) and 50 cells/condition were counted to determine the localization of AR in nucleus.

FACS

FACS analysis was performed to determine surface expression of integrins using Abs specific to β6 (10D5) or β3 (AP3); nonimmune mIgG, 12CA5, or IC10 were used as negative controls.

Generation of Ptenpc−/− mice and castration

Prostate epithelial-specific Pten conditional knockout mice (Ptenpc−/−) mice were generated as previously reported (33). Ptenpc−/− mice were surgically castrated using published method (33).

Statistical analysis

Unless otherwise indicated, data in the figures are presented as mean ± SEM, and significant differences between experimental groups were determined using the two-tailed Student t test. The Fisher's exact two-tailed test was used to compare the percentage of β6-positive cells. Mantel-extension of Mantel–Haenszel Statistic stratified by subject was used to test the correlation between β6 expression and the types of lesion. The Wilcoxon–Mann–Whitney test was used to analyze the difference in tumor growth between 6.3G9 and control group in vivo. To examine the association between β6 and survivin expression in human specimens, the Spearman correlation coefficient was calculated and tested. A two-sided P value of less than 0.05 was considered statistically significant. SAS statistical software 9.1.2 (SAS Institute, Inc.) was used for statistical analysis.

αvβ6 integrin promotes CRPC growth in vivo

To determine the effect of αvβ6 expression in vivo, we used a CRPC model that uses surgically castrated Ptenpc−/− as delineated in Fig. 1A. αvβ6 expression pattern in castrate-resistant tumors (n = 5) was compared by IHC with noncastrated tumors (n = 5) in Ptenpc−/− mice. We show here that αvβ6 is expressed in mouse CRPC; specifically, we demonstrate that the expression of αvβ6 is highly heterogeneous (18% of the tumor cells) in tumors from noncastrated mice while homogenous (90% of tumor cells) in tumors from castrated mice, indicating a significant enrichment of this integrin upon castration. Accordingly, JNK nuclear localization is increased in αvβ6 positive cells in noncastrated tumors (3% of tumor cells), and becomes dominant in castrated tumors (75% of tumor cells; Supplementary Fig. S1).

Figure 1.

Inhibitory effect of mAb 6.3G9 on tumor growth in castrated Ptenpc−/− mice. A, scheme of surgical castration and Ab treatment of Ptenpc−/− mice. B and C, mAb 6.3G9 against αvβ6 or a negative control mAb 1E6 were injected in castrated Ptenpc−/− mice intraperitoneally both at 10 mg/kg weekly for 5 weeks, then mice (n = 11) were sacrificed and tumor weights were quantified. B, hematoxylin and eosin analysis of tumor specimens from 6.3G9-treated mice (top) or 1E6-treated mice (bottom). Arrow, top right, epithelial cells. Arrowhead, cancer cells sloughed in the lumen. Asterisk, microinvasion. Original magnification, left column, ×200; right column, ×400. C, individual tumor weight is plotted; lines are mean ± SEM. P = 0.004 (Wilcoxon rank-sum two-sided test).

Figure 1.

Inhibitory effect of mAb 6.3G9 on tumor growth in castrated Ptenpc−/− mice. A, scheme of surgical castration and Ab treatment of Ptenpc−/− mice. B and C, mAb 6.3G9 against αvβ6 or a negative control mAb 1E6 were injected in castrated Ptenpc−/− mice intraperitoneally both at 10 mg/kg weekly for 5 weeks, then mice (n = 11) were sacrificed and tumor weights were quantified. B, hematoxylin and eosin analysis of tumor specimens from 6.3G9-treated mice (top) or 1E6-treated mice (bottom). Arrow, top right, epithelial cells. Arrowhead, cancer cells sloughed in the lumen. Asterisk, microinvasion. Original magnification, left column, ×200; right column, ×400. C, individual tumor weight is plotted; lines are mean ± SEM. P = 0.004 (Wilcoxon rank-sum two-sided test).

Close modal

Upon intraperitoneal administration for 5 weeks of 6.3G9, a non ligand-mimetic blocking mAb that does not get internalized upon binding (21), to castrated Ptenpc−/− mice, tumor progression is inhibited and the prostate glands do not show evidence of invasive adenocarcinoma; in contrast, upon intraperitoneal administration of a negative control mAb, 1E6, the prostate glands show invasive adenocarcinoma (Fig. 1B). The results show a significant decrease of tumor weight in castrated Ptenpc−/− mice treated with 6.3G9 (25.3 ± 1.8 mg) as compared with the group treated with 1E6 (39.8 ± 1.5 mg; Fig. 1C). We also carried out xenograft tumor growth experiments in castrated athymic nude mice (Supplementary Fig. S2). For this purpose, we used AR+ LNCaP cells that stably express human β6 (αvβ6-LNCaP), human β3 (αvβ3-LNCaP) or empty vector (mock-LNCaP). These cells exhibit no difference in the expression of other endogenous integrin subunits (Supplementary Fig. S2A). Upon subcutaneous injection, αvβ6-LNCaP xenograft tumors continue to grow significantly (Supplementary Fig. S2B), suggesting that αvβ6 expression is sufficient to confer a resistant phenotype to prostate cancer cells, in vivo. In contrast, αvβ3-LNCaP transfectants do not significantly grow as subcutaneous tumors in castrated mice (Supplementary Fig. S2B).

We in parallel investigated whether αvβ6 is required for prostate cancer growth in noncastrated mice. We used Ptenpc−/− mice to independently validate a requirement of αvβ6 for tumor growth in vivo. Intraperitoneal administration of the 6.3G9 Ab for 5 weeks to Ptenpc−/− mice results in acute disruption of epithelial layers of prostate adenocarcinoma, with appearance of necrotic cells filling the glandular lumen (Fig. 2A) and in a significant decrease in tumor weight as compared with mice treated with the 1E6 mAb (P = 0.0042; Fig. 2B). In contrast, prostate specimens collected from Ptenpc−/− mice treated with 1E6 show an intact structure of transformed prostatic glands (Fig. 2A). In addition, histologically normal glands of Ptenpc−/− mice receiving 6.3G9 mAb do not show a disruption of epithelial layers (Fig. 2A). Subcutaneous injection of αvβ6-LNCaP cells in the flanks of immunocompromised SCID mice gives rise to exponentially growing tumors with a statistically significant difference (P < 0.01) compared with αvβ3-LNCaP tumors starting at 54 days after injection (Fig. 2C). In two additional independent experiments, αvβ3-LNCaP cells or mock-LNCaP cells (Supplementary Fig. S3A) produce tumors with statistically slower growth kinetics (P < 0.01) as compared with αvβ6-LNCaP tumors. Accordingly, expression of αvβ3- or αvβ6-integrins on the cell surface was confirmed by FACS (Supplementary Fig. S3B). Tumors generated by the various transfectants retain the expression of the respective β6 or β3 integrins as detected by immunoblotting, whereas tumors generated by mock transfectants remain negative (Supplementary Fig. S3C). Together, these data demonstrate that αvβ6 is required for accelerated prostate cancer growth, in vivo.

Figure 2.

Inhibitory effect of mAb 6.3G9 on tumor growth in noncastrated Ptenpc−/− mice. A and B, Ptenpc−/− mice (5–8 weeks) were intraperitoneally injected with 6.3G9 or 1E6 as described in Fig. 1, then mice were sacrificed and the differences in tumor weight were quantified. A, representative prostate specimens (normal gland or adenocarcinoma) of Ptenpc−/− mice injected with mAb 6.3G9 or control mAb 1E6 were analyzed histologically by hematoxylin and eosin staining. Arrow, cancer epithelial cells. Arrowhead, apoptotic cancer cells in the lumen; none, untreated. Original magnification, left column, ×200; right column, ×400. B, individual tumor weight is plotted (n = 8). Statistical significance was observed in tumor weight, P = 0.0042 (Wilcoxon rank-sum two-sided test). C, αvβ6-LNCaP (filled circle) or αvβ3-LNCaP (open square) cells were injected subcutaneously into male CB17/SCID mice (16 mice/group), and differences in tumor volume were quantified. Statistical analysis shows significant differences in tumor growth between two cell lines from day 54 to 85 (**, P < 0.01); data are shown as mean ± SEM.

Figure 2.

Inhibitory effect of mAb 6.3G9 on tumor growth in noncastrated Ptenpc−/− mice. A and B, Ptenpc−/− mice (5–8 weeks) were intraperitoneally injected with 6.3G9 or 1E6 as described in Fig. 1, then mice were sacrificed and the differences in tumor weight were quantified. A, representative prostate specimens (normal gland or adenocarcinoma) of Ptenpc−/− mice injected with mAb 6.3G9 or control mAb 1E6 were analyzed histologically by hematoxylin and eosin staining. Arrow, cancer epithelial cells. Arrowhead, apoptotic cancer cells in the lumen; none, untreated. Original magnification, left column, ×200; right column, ×400. B, individual tumor weight is plotted (n = 8). Statistical significance was observed in tumor weight, P = 0.0042 (Wilcoxon rank-sum two-sided test). C, αvβ6-LNCaP (filled circle) or αvβ3-LNCaP (open square) cells were injected subcutaneously into male CB17/SCID mice (16 mice/group), and differences in tumor volume were quantified. Statistical analysis shows significant differences in tumor growth between two cell lines from day 54 to 85 (**, P < 0.01); data are shown as mean ± SEM.

Close modal

αvβ6 integrin promotes prostate cancer cell proliferation and activates the AR pathway

To determine the potential basis for αvβ6-mediated aggressive tumor behavior in vivo, we next measured the proliferation kinetics of integrin-expressing cells in the presence or absence of androgen. Mock-LNCaP and αvβ3-LNCaP cells exhibit increased proliferation in response to androgen (R1881; Fig. 3A). Conversely, two independent clones of αvβ6-LNCaP cells show increased rates of proliferation compared to mock-LNCaP or αvβ3-LNCaP cells in the absence of androgen, and this response is unaffected by androgen (Fig. 3A, left). The αvβ6-LNCaP, mock-LNCaP and αvβ3-LNCaP cells do not show morphologic differences (Fig. 3B). In an anchorage-independent assay, we downregulated β6 in AR+ RWPE cells using two independent siRNAs, D1 and D2 (Fig. 3C, right). β6 silencing by two independent siRNAs, under these conditions, suppresses anchorage-independent tumor growth whereas nontargeting siRNA (NS) or oligofectamine (OF) alone has no effect (Fig. 3C, left).

Figure 3.

αvβ6 integrin expression enhances prostate cancer cell proliferation. A, mock-, αvβ6-LNCaP (clones 1 and 2) or αvβ3-LNCaP cells were stimulated with R1881 (1 nmol/L) or vehicle after starvation, and quantified for cell proliferation by SRB staining. B, mock-, αvβ6- and αvβ3-LNCaP cells were cultured in either tissue culture plates or FN (5 μg/mL)-coated plates. After 72 hours, cellular morphology was visualized. C, RWPE cells transfected with β6-directed (D1 or D2) or non-silencing (NS) siRNA were analyzed for anchorage-independent cell growth, and the total number of colonies in 20 fields was scored 14 days after seeding (left). RWPE cells were transfected with β6 integrin-directed siRNA and analyzed by FACS with an mAb to αvβ6 or nonbinding IgG (mIgG; right); OF, oligofectamine. A and C, data are mean ± SEM; **, P < 0.01.

Figure 3.

αvβ6 integrin expression enhances prostate cancer cell proliferation. A, mock-, αvβ6-LNCaP (clones 1 and 2) or αvβ3-LNCaP cells were stimulated with R1881 (1 nmol/L) or vehicle after starvation, and quantified for cell proliferation by SRB staining. B, mock-, αvβ6- and αvβ3-LNCaP cells were cultured in either tissue culture plates or FN (5 μg/mL)-coated plates. After 72 hours, cellular morphology was visualized. C, RWPE cells transfected with β6-directed (D1 or D2) or non-silencing (NS) siRNA were analyzed for anchorage-independent cell growth, and the total number of colonies in 20 fields was scored 14 days after seeding (left). RWPE cells were transfected with β6 integrin-directed siRNA and analyzed by FACS with an mAb to αvβ6 or nonbinding IgG (mIgG; right); OF, oligofectamine. A and C, data are mean ± SEM; **, P < 0.01.

Close modal

To analyze the mechanism by which αvβ6 affects prostate cancer growth, we tested AR activity in the presence or absence of androgen in αvβ6-LNCaP (Fig. 4A, left) or αvβ6-C4-2B cells (Fig. 4B, left); the results show that αvβ6-LNCaP exhibit constitutively high levels of PSA, an established AR target gene, without addition of exogenous androgen, and PSA levels are only minimally affected in the presence of androgen. In control experiments, mock-transfectants, αvβ3-LNCaP or αvβ3-C4-2B cells show negligible basal levels of PSA, whereas androgen (R1881) stimulation elevates PSA expression in these cells (Fig. 4A and B, left; also shown in Supplementary Fig. S4). Silencing AR by siRNA in αvβ6-LNCaP cells (Fig. 4A, right) or αvβ6-C4-2B cells (Fig. 4B, right) abolishes αvβ6-induced PSA, confirming AR specificity of these findings. Similar results are obtained in the analysis of AR promoter activity; αvβ6-LNCaP cells exhibit constitutive high levels of androgen responsive elements (ARE)4-luciferase activity, irrespective of androgen stimulation, as compared with mock-transfected cultures (Supplementary Fig. S5). In addition, neither expression of αvβ6 or αvβ3 integrin affects endogenous AR protein levels, in the presence or absence of androgen (Fig. 4C), indicating that the increase in AR activity observed in αvβ6+ cells does not reflect changes in AR protein expression. Immunofluorescence (IF) analysis shows increased nuclear localization of AR in the absence of androgen in αvβ6-LNCaP cells (Fig. 4D). In contrast, AR reactivity is detected largely in the cytoplasm of unstimulated αvβ3-LNCaP or mock-LNCaP cells, and becomes intense in the nuclei only in response to androgen stimulation.

Figure 4.

αvβ6 integrin regulates AR activity and nuclear trafficking. A and B, left, LNCaP and C4-2B transfectants were starved in 2% CS serum for 24 hours and stimulated with R1881 (1 nmol/L) or ethanol for 24 hours. PSA level was analyzed by immunoblotting (IB). Right, αvβ6-LNCaP and αvβ6-C4-2B cells were transfected with AR-directed siRNA and the PSA level was evaluated by immunoblotting. Original blots are shown in Supplementary Fig. S4. C, mock-, αvβ3-, αvβ6- and parental (−) LNCaP cells were analyzed for AR expression with or without R1881 by immunoblotting. A–C, ERK1/2 or Akt was used as a loading control. OF, oligofectamine; NS, non-silencing. D, cellular distribution of AR in αvβ6-, αvβ3-, and mock-LNCaP cells. Indicated cells were seeded on poly–L-lysine (1 mg/mL)-coated coverslips for 2 days, starved, and treated with or without R1881 (1 nmol/L). E, LNCaP cell transfectants were treated with the indicated concentrations of enzalutamide in culture medium for 72 hours and cell proliferation rate was calculated by cell counting. Proliferation is valued as the percentage of the control groups treated with DMSO only. Data are mean ± SD.

Figure 4.

αvβ6 integrin regulates AR activity and nuclear trafficking. A and B, left, LNCaP and C4-2B transfectants were starved in 2% CS serum for 24 hours and stimulated with R1881 (1 nmol/L) or ethanol for 24 hours. PSA level was analyzed by immunoblotting (IB). Right, αvβ6-LNCaP and αvβ6-C4-2B cells were transfected with AR-directed siRNA and the PSA level was evaluated by immunoblotting. Original blots are shown in Supplementary Fig. S4. C, mock-, αvβ3-, αvβ6- and parental (−) LNCaP cells were analyzed for AR expression with or without R1881 by immunoblotting. A–C, ERK1/2 or Akt was used as a loading control. OF, oligofectamine; NS, non-silencing. D, cellular distribution of AR in αvβ6-, αvβ3-, and mock-LNCaP cells. Indicated cells were seeded on poly–L-lysine (1 mg/mL)-coated coverslips for 2 days, starved, and treated with or without R1881 (1 nmol/L). E, LNCaP cell transfectants were treated with the indicated concentrations of enzalutamide in culture medium for 72 hours and cell proliferation rate was calculated by cell counting. Proliferation is valued as the percentage of the control groups treated with DMSO only. Data are mean ± SD.

Close modal

The ability of αvβ6 to activate AR, however, does not promote resistance to enzalutamide, an AR antagonist developed for the treatment of metastatic CRPC patients. Mechanistically, enzalutamide inhibits translocation of AR to the nucleus and prevents binding of AR to DNA or to coactivator proteins (34). On the basis of this knowledge, we investigated whether this compound may also inhibit ligand-independent AR activation mediated by αvβ6. We find that αvβ6 integrin expression does not promote resistance to enzalutamide at concentrations up to 100 μmol/L and that enzalutamide is effective in suppressing proliferation of αvβ6-, αvβ3- and mock- LNCaP cell transfectants in a comparable fashion (Fig. 4E).

αvβ6 activation of JNK1 regulates AR activity

To determine the mechanism by which αvβ6 activates AR, we analyzed the kinase cascades typically activated by integrin ligand binding. Adhesion of αvβ6-LNCaP cells to the αvβ6 ligand LAP-TGFβ1 or FN results in JNK1 activation as evaluated by its phosphorylation pattern (Fig. 5A, top). In parallel, we observe that PSA is increased in αvβ6-transfectants attached to LAP-TGFβ1 or FN whereas αvβ3-transfectants show barely detectable PSA expression (Fig. 5A, middle). αvβ3-LNCaP cells adhere to FN or LAP-TGFβ1 but fail to induce JNK1 phosphorylation (Fig. 5A, top). PC3 cells stably transfected with β6 (αvβ6-PC3) also exhibit increased JNK1 phosphorylation upon adhesion to FN or LAP-TGFβ1, whereas mock-transfected cells do not (Fig. 5A, bottom). The function-blocking Ab 6.3G9 inhibits attachment to LAP-TGFβ1 in a concentration-dependent manner, thus identifying αvβ6 as the predominant LAP-TGFβ1 receptor in prostate cancer cells (Fig. 5A, graph; ref. 35). Conversely, ligand binding to αvβ6 does not activate related signaling pathways, including ERK, p38 (Supplementary Fig. S6A and S6B), or JNK2 (data not shown) as determined by IB with phosphorylation specific Abs. Because JNK1 activation has been implicated in controlling AR activity via nuclear translocation (36), we infected αvβ6-C4-2B cells with a retrovirus encoding a JBD, which functions as a dominant negative JNK (27, 37). αvβ6-C4-2B cells exhibit high PSA levels, which is not further modulated by androgen (R1881) stimulation (Fig. 5B; also shown in Supplementary Fig. S7), in agreement with the data presented above. In contrast, JBD inhibits the basal expression of PSA in αvβ6-C4-2B cells in a concentration-dependent manner, and completely abolishes PSA levels after androgen stimulation (Fig. 5B), whereas infection of WT-JNK1 does not affect PSA levels in αvβ6-C4-2B cells, in the presence or absence of androgen. Androgen modulation of PSA in mock- or αvβ3-C4-2B cells is unaffected by WT-JNK1 or JBD (Fig. 5B). To identify a potential link between JNK1 activity and AR-dependent gene expression, we next looked at potential changes in AR localization. IF staining shows that infection of JBD in αvβ6-LNCaP cells significantly impairs AR nuclear localization (Fig. 5C). Overall, this study shows that upon ligand binding, αvβ6 enhances AR transactivation via JNK1 stimulation, which in turn controls AR nuclear trafficking.

Figure 5.

JNK1 mediates transactivation of AR in αvβ6 integrin–expressing cells. A, αvβ6-LNCaP or αvβ3-LNCaP cells (top) were seeded on FN or LAP-TGFβ1 (LAP)–coated plates, and analyzed for JNK1 phosphorylation by immunoblotting (IB). White lines indicate that intervening lanes were spliced out. PSA levels were also analyzed by immunoblotting. c-src was used as a loading control. PC3 cells stably transfected with pBabe-β6 (αvβ6) or pBabe vector (mock) were seeded on FN-, LAP-, or BSA-coated plates and analyzed for JNK1 activation by detection of phospho-JNK1 (p-JNK1). p-JNK1 levels in parental cells (−) exposed to UV irradiation were used as a positive control for JNK1 activation. Total JNK1 was used as a loading control. PC3 cells (bottom) were incubated in the presence of the indicated concentrations of mAb 6.3G9 to αvβ6, and adhesion to LAP-coated plates was quantified. Data are mean ± SEM; **, P < 0.01. B, αvβ6-, αvβ3- or mock C4-2B cells were infected with WT-JNK or decreasing concentrations of JBD and analyzed for PSA expression by immunoblotting. Akt or β-actin was used as a loading control. Uninfected cells (−) were used as controls. Original blots are shown in Supplementary Fig. S7. C, αvβ6-LNCaP cells transiently infected with WT-JNK (WT) or JBD were seeded on a poly–L-lysine–coated coverslip for 24 hours, starved, and analyzed with an Ab to AR by IF (left). DAPI (right).

Figure 5.

JNK1 mediates transactivation of AR in αvβ6 integrin–expressing cells. A, αvβ6-LNCaP or αvβ3-LNCaP cells (top) were seeded on FN or LAP-TGFβ1 (LAP)–coated plates, and analyzed for JNK1 phosphorylation by immunoblotting (IB). White lines indicate that intervening lanes were spliced out. PSA levels were also analyzed by immunoblotting. c-src was used as a loading control. PC3 cells stably transfected with pBabe-β6 (αvβ6) or pBabe vector (mock) were seeded on FN-, LAP-, or BSA-coated plates and analyzed for JNK1 activation by detection of phospho-JNK1 (p-JNK1). p-JNK1 levels in parental cells (−) exposed to UV irradiation were used as a positive control for JNK1 activation. Total JNK1 was used as a loading control. PC3 cells (bottom) were incubated in the presence of the indicated concentrations of mAb 6.3G9 to αvβ6, and adhesion to LAP-coated plates was quantified. Data are mean ± SEM; **, P < 0.01. B, αvβ6-, αvβ3- or mock C4-2B cells were infected with WT-JNK or decreasing concentrations of JBD and analyzed for PSA expression by immunoblotting. Akt or β-actin was used as a loading control. Uninfected cells (−) were used as controls. Original blots are shown in Supplementary Fig. S7. C, αvβ6-LNCaP cells transiently infected with WT-JNK (WT) or JBD were seeded on a poly–L-lysine–coated coverslip for 24 hours, starved, and analyzed with an Ab to AR by IF (left). DAPI (right).

Close modal

Survivin is a downstream effector of the αvβ6 integrin

To identify potential downstream effectors of integrin-mediated prostate cancer progression, we focused on survivin, that has been implicated in prostate cancer maintenance and resistance to therapy (19, 38). αvβ6-C4-2B cells exhibit higher levels of survivin, compared with αvβ3-C4-2B (Fig. 6A), whereas XIAP or Bcl2, two other molecules known to inhibit apoptosis, are not affected (Fig. 6A). Similar results were obtained with LNCaP cells stably transfected with a β3/6 integrin chimera (Fig. 6B and Supplementary Fig. S8), containing the β6 intracellular domain and the β3 extracellular domain expressed at the cell surface, suggesting that the β6 cytoplasmic domain is required for survivin induction.

Figure 6.

AR regulates survivin expression and anchorage-independent-growth in αvβ6 integrin–expressing cells. A, αvβ3- and αvβ6-C4-2B cells were analyzed for survivin and XIAP expression by immunoblotting (IB; top). αvβ3- and αvβ6-C4-2B (clone 1 and 2) cells were stimulated with or without R1881 (1 nmol/L) and analyzed for Bcl-2 levels by immunoblotting (bottom). B, mock-LNCaP (lane 1, 3), αvβ6-LNCaP (lane 2), or chimera αvβ3/6-LNCaP (lane 4) cells were analyzed for survivin expression by immunoblotting. The white line indicates that intervening lanes were spliced out. C, αvβ6-LNCaP cells were transfected with nonsilencing (NS) or AR-directed siRNA for 24 hours, starved, and analyzed for expression of AR or survivin by immunoblotting. A–C, Akt or ERK1/2 was used as a loading control. OF, oligofectamine. D, αvβ6-LNCaP cells were transiently infected with WT-JNK or JBD and analyzed for survivin expression by immunoblotting. Akt was used as a loading control. Uninfected cells (−) were used as controls. E, αvβ6-LNCaP cells were infected with adenovirus containing survivin wild-type (WT) or a survivin mutant (T34A) and analyzed for anchorage-independent growth. The average of colony diameters (left) of the three groups ± SEM are plotted. αvβ6-LNCaP cells were transiently transfected with nonsilencing or survivin-directed (SV) siRNA for 48 hours and analyzed for anchorage-independent growth. The average number of the colonies (right) in the three groups ± SEM are plotted. **, P < 0.01.

Figure 6.

AR regulates survivin expression and anchorage-independent-growth in αvβ6 integrin–expressing cells. A, αvβ3- and αvβ6-C4-2B cells were analyzed for survivin and XIAP expression by immunoblotting (IB; top). αvβ3- and αvβ6-C4-2B (clone 1 and 2) cells were stimulated with or without R1881 (1 nmol/L) and analyzed for Bcl-2 levels by immunoblotting (bottom). B, mock-LNCaP (lane 1, 3), αvβ6-LNCaP (lane 2), or chimera αvβ3/6-LNCaP (lane 4) cells were analyzed for survivin expression by immunoblotting. The white line indicates that intervening lanes were spliced out. C, αvβ6-LNCaP cells were transfected with nonsilencing (NS) or AR-directed siRNA for 24 hours, starved, and analyzed for expression of AR or survivin by immunoblotting. A–C, Akt or ERK1/2 was used as a loading control. OF, oligofectamine. D, αvβ6-LNCaP cells were transiently infected with WT-JNK or JBD and analyzed for survivin expression by immunoblotting. Akt was used as a loading control. Uninfected cells (−) were used as controls. E, αvβ6-LNCaP cells were infected with adenovirus containing survivin wild-type (WT) or a survivin mutant (T34A) and analyzed for anchorage-independent growth. The average of colony diameters (left) of the three groups ± SEM are plotted. αvβ6-LNCaP cells were transiently transfected with nonsilencing or survivin-directed (SV) siRNA for 48 hours and analyzed for anchorage-independent growth. The average number of the colonies (right) in the three groups ± SEM are plotted. **, P < 0.01.

Close modal

In this context, silencing of AR significantly reduces survivin levels in αvβ6-LNCaP cells, whereas a nontargeting siRNA is ineffective (Fig. 6C). Inhibition of JNK activity in these cells via infection with JBD also decreases survivin levels independently of androgen, whereas expression of WT JNK1 moderately up-regulates survivin in αvβ6-LNCaP cells (Fig. 6D). Silencing of survivin by siRNA, or interference with survivin function via adenoviral transduction of a survivin Thr34Ala dominant–negative mutant (Fig. 6E, left), abolishes prostate cancer colony formation in soft agar, that is, anchorage-independent growth. In contrast, a non-targeting siRNA or adenoviral transduction of WT survivin does not affect anchorage-independent growth under the same conditions (Fig. 6E, right). We also analyzed the expression of αvβ6 integrin in human primary specimens of prostatic adenocarcinoma or normal prostate. Using an Ab that specifically recognizes αvβ6, 6.2A1, we found by IHC that this integrin is expressed in prostate cancer (35/47 specimens, 74%; Fig. 7A), consistent with its epithelial-specific expression, although in a highly heterogeneous pattern with 5%∼31% of the glands being αvβ6 positive. Identical results were obtained with an independent mAb (B1) to αvβ6 (Supplementary Fig. S9). In contrast, αvβ6 is undetectable in 8 out of 8 cases of normal prostate (Fig. 7A), and does not correlate with Gleason Score (5–10), or patient age (range, 41–83 years). In addition, patient-derived primary prostate cancer cells, isolated as described (30) and reactive for cytokeratin (CK) 8 and 18 (Fig. 7B, middle), express high levels of αvβ6, as detected by FACS analysis (Fig. 7B, top). This result was confirmed using 15 fresh prostatic adenocarcinoma tissue lysates by IB using an Ab to αvβ6 (6.2A1), which detects a 110 kDa protein consistent with β6 reactivity (Fig. 7B, bottom, two representative specimens are shown) albeit with variable levels of expression. Survivin expression was also studied in 34 specimens and is found to be expressed in 97% of adenocarcinoma specimens (Fig. 7C, 1–4); the association between αvβ6 and survivin is statistically significant in 34 cases of adenocarcinoma (P < 0.028) and is independent of Gleason scores (5–10) or disease stage (T2-4; Supplementary Table S1). Overall, our in vitro and IHC analysis show that survivin is a downstream effector of the αvβ6 integrin.

Figure 7.

αvβ6 integrin and survivin are coexpressed in human prostatic lesions. A, human adenocarcinoma (1–4) and normal prostate (5) specimens were analyzed for αvβ6 expression by IHC using mAb 6.2A1 (1–3) or mIgG was used as a negative control (4). Representative cases are shown. Arrows, αvβ6-expressing gland; arrowheads, αvβ6-negative glands (1 and 2). Original magnification, ×200. The percentage of αvβ6-expressing adenocarcinoma tissue specimens at different Gleason scores is shown. B, primary prostatic epithelial cells were isolated from three cancer patients and analyzed for expression of αvβ6 by FACS analysis with mAb 10D5 (thick line; left). IF staining of primary prostatic epithelial cells was carried out using Abs to CK 8 or 18. Ab 1C10 against an endothelial marker was used as a negative control (top right). Primary tumor extracts were analyzed for expression of β6 integrin using mAb 6.2A1 (bottom). Akt was used as loading control. C, representative images of adenocarcinoma (1–4) showing co-expression of αvβ6 integrin and survivin in the same gland. D, schematic diagram for integrin regulation of AR-mediated prostate tumorigenesis independently of androgen.

Figure 7.

αvβ6 integrin and survivin are coexpressed in human prostatic lesions. A, human adenocarcinoma (1–4) and normal prostate (5) specimens were analyzed for αvβ6 expression by IHC using mAb 6.2A1 (1–3) or mIgG was used as a negative control (4). Representative cases are shown. Arrows, αvβ6-expressing gland; arrowheads, αvβ6-negative glands (1 and 2). Original magnification, ×200. The percentage of αvβ6-expressing adenocarcinoma tissue specimens at different Gleason scores is shown. B, primary prostatic epithelial cells were isolated from three cancer patients and analyzed for expression of αvβ6 by FACS analysis with mAb 10D5 (thick line; left). IF staining of primary prostatic epithelial cells was carried out using Abs to CK 8 or 18. Ab 1C10 against an endothelial marker was used as a negative control (top right). Primary tumor extracts were analyzed for expression of β6 integrin using mAb 6.2A1 (bottom). Akt was used as loading control. C, representative images of adenocarcinoma (1–4) showing co-expression of αvβ6 integrin and survivin in the same gland. D, schematic diagram for integrin regulation of AR-mediated prostate tumorigenesis independently of androgen.

Close modal

In this study, we show that expression of an epithelial-specific integrin, αvβ6, in prostate cancer as well as in castrated tumors, is sufficient to promote aggressive tumor growth and castrate-resistance disease. The molecular requirements of this pathway are centered on αvβ6 activation of JNK1, which in turn promotes AR nuclear shuttling, concomitant ligand-independent AR transcription and subsequent survivin-regulated tumor growth (Fig. 7D).

At the molecular level, this transition to castrate resistance is an as yet poorly understood process (4). Our data unravel a novel pathway of castrate-resistant aggressive prostate cancer that originates in the tumor via αvβ6 integrin ligand binding, and ultimately promotes AR-modulated gene expression, independent of androgen.

There is ample precedent for the role of integrins in controlling steroid hormone receptor expression and function. In a prostate cancer in vitro model, our group recently found that αvβ6 is efficiently transferred via exosomes to αvβ6-negative recipient cells and localizes to the prostate cancer cell surface, promoting cell adhesion and migration on LAP-TGFβ, suggesting a systemic role of αvβ6 in tumor progression (39) in addition to its local effect. These effects, however, do not require transfer of JNK via exosomes, because recent data show that JNK is not detected in exosomes (L.R. Languino; unpublished data). In this study, our data indicate that expression of αvβ6 in prostate cancer cells is sufficient to confer a “sub-optimal” androgen-independent phenotype, characterized by proliferation and PSA expression in the absence of exogenous androgen. This is further supported by our in vivo studies, showing a significant increase in the growth of αvβ6-expressing tumors in castrated as well as noncastrated mice. Finally, because αvβ6 effect does not promote resistance to enzalutamide, targeting this integrin may thus be a potential adjuvant in enzalutamide treatment.

The molecular requirements of signal integration from αvβ6 to AR-dependent gene expression in prostate cancer remain to be fully elucidated. However, one of the key requirements in this pathway is αvβ6 integrin-mediated activation of the MAP kinase family member, JNK1. Despite its conflicting, and likely cell-type specific, roles in both cell survival and apoptosis (37), the JNK pathway has recently emerged as an important regulator of prostate cancer growth via AR (36). The JNK1 downstream target, c-Jun, strongly potentiates AR-mediated gene transcription (40, 41), and silencing of JNK1 inhibits prostate cancer tumor growth (42). The data presented here suggest that downstream kinase activation by αvβ6 may be specific for JNK1, with no involvement of p38 or ERK kinase. In addition, differential phosphorylation of Akt is not observed under these conditions consistent with a context-dependent role of Akt in AR activation as described by the following studies. Although PI3K/Akt signaling promotes AR activity (43), examples of PTEN regulation of JNK signaling independently of Akt have been reported (44); furthermore, JNK deficiency promotes androgen-independent progenitor cell growth and metastatic prostate cancer more rapidly than PTEN deficiency alone (33). Our data from several prostate cancer cell lines (LNCaP, C4-2B, PC3) show that JNK1 promotes AR nuclear translocation and activation. With respect to how active JNK1 modulates AR function in αvβ6 integrin-expressing cells, the data presented here are in line with several reports linking AR phosphorylation to its nuclear-cytoplasmic shuttling. AR can be phosphorylated at tyrosine, serine, and threonine residues by a wide range of kinases, such as cyclin-dependent kinases (CDK), Akt, PIM1, Fer, Aurora-A, Src, Etk, PKC, PAK6, JNK and p38 (45). Akt, for example, can phosphorylate AR at S210/S790 and inhibit androgen-induced apoptosis (46); CDK1 phosphorylates AR S81 and regulates AR stability (47); in addition, soluble factors derived from stroma activate ERK, which phosphorylates AR at S81, promote AR-dependent transcription and anchorage-independent growth (48). The data presented in our study show that αvβ6 predominantly promotes AR nuclear translocation and activation through JNK1.

It should be stressed that AR is already activated in castrate-resistant cells, such as C4-2B, but the activity and/or nuclear distribution are not saturated or irreversible. Upon proper stimulation (49), the fraction of AR in the nucleus can be further increased, as shown in αvβ6 transfectants in this study. Our data suggest that integrins promote a novel mechanism for castrate resistance, other than the mechanisms dependent on AR mutations. We believe the use of the castrate-resistant cells in this article supports the major conclusion that αvβ6 is one of the key factors promoting prostate cancer progression in castrate settings, in addition to its protumor effect in androgen-sensitive cells such as RWPE cells.

An additional downstream effector of aggressive prostate cancer phenotype of αvβ6 integrin-expressing cells is identified here as survivin, a critical prostate cancer–promoting molecule (18) regulated by PI3K/Akt signaling, as a mediator of drug resistance (19), and as an effector of tumor progression. Although Interleukin-4–induced survivin is independent of JNK (32), in our model JNK is required for upregulation of survivin by αvβ6 integrin, suggesting a different mechanism in this context. Against this backdrop, survivin expression correlates here with αvβ6 integrin levels in primary prostate cancer specimens, is directly regulated by AR, and is required for anchorage-independent growth, in agreement with an emerging role of survivin in this disease (50).

In summary, we have identified a novel signaling circuit linking integrins with modulation of nuclear receptor gene expression, critically important to confer a castrate-resistant phenotype and aggressive behavior of prostate cancer in humans. The identification of mechanisms of castrate resistance amenable to therapeutic intervention is an urgent and as yet unmet need in advanced prostate cancer, a disease stage where therapeutic options are few and only minimally effective. The translational relevance of our findings is highlighted by the recognized role of integrins as bona fide therapeutic targets for drug development. The preclinical ability of an mAb to αvβ6 integrin to inhibit tumor growth and disrupt the epithelial architecture of prostate cancer, without toxicity to normal prostate, as presented here, bodes well for further clinical development of αvβ6 antagonists.

S.M. Violette and P.H. Weinreb have ownership interest (including patents) in Biogen, Inc. No potential conflicts of interest were disclosed by the other authors.

The Pennsylvania Department of Health specifically disclaims responsibility for any analyses, interpretations, or conclusions.

Conception and design: H. Lu, T. Wang, J. Li, R.V. Iozzo, S.M. Violette, R.J. Davis, T.J. FitzGerald, L.R. Languino

Development of methodology: H. Lu, T. Wang, J. Li, S.M. Violette, R.J. Davis, T.J. FitzGerald

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): H. Lu, T. Wang, J. Li, C. Fedele, J. Zhang, Z. Jiang, T.J. FitzGerald

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): H. Lu, T. Wang, J. Li, C. Fedele, Q. Liu, Z. Jiang, D. Jain, R.V. Iozzo, S.M. Violette, P.H. Weinreb, D. Gioeli, T.J. FitzGerald, L.R. Languino

Writing, review, and/or revision of the manuscript: H. Lu, T. Wang, J. Li, C. Fedele, Q. Liu, Z. Jiang, D. Jain, S.M. Violette, P.H. Weinreb, R.J. Davis, D. Gioeli, T.J. FitzGerald, D.C. Altieri, L.R. Languino

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): T. Wang, S.M. Violette, D. Gioeli, T.J. FitzGerald

Study supervision: L.R. Languino

We thank Drs. Dean Sheppard for B1, rabbit mAb to β6; Michael Lu for ARE constructs; Lan Hart for β6 integrin constructs; Shutsing Liao for retrovirus containing pLNCX2-human AR; Hong Wu for Ptenloxp/loxp mice; Takehiko Dohi for preparation of adenovirus constructs; David Garlick and Irwin Leav for evaluation of Ptenpc−/− prostate specimens and Rebecca Galanti for technical help with the 6.3G9 Ab in vivo analysis. We are grateful to Dr. Chung C. Hsieh for biostatistical analysis. For this study, Sidney Kimmel Cancer Center Bioimaging and Histology Core Facilities, which is supported in part by NCI Cancer Center-Support Grant P30 CA56036, were used. We are grateful to Mrs. Cecilia Deemer for administrative support with this manuscript.

NIH CA89720 and CA109874 (L.R. Languino), CA78810 and CA90917 (D.C. Altieri), CA140043 (L.R. Languino and D.C. Altieri), CA65861 (R.J. Davis), CA39481 (R.V. Iozzo), Prostate Cancer Foundation Challenge Award (L.R. Languino and D.C. Altieri), Danny Cancer Funds P000100033 (T. Wang), Postdoctoral Research Fellowship from the American Italian Cancer Foundation (C. Fedele). This project is also funded, in part, under a Commonwealth University Research Enhancement Program grant with the Pennsylvania Department of Health (H.R.).

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.
Siegel
RL
,
Miller
KD
,
Jemal
A
. 
Cancer statistics, 2016
.
CA Cancer J Clin
2016
;
66
:
7
30
.
2.
Keller
ET
,
Dai
J
,
Escara-Wilke
J
,
Hall
CL
,
Ignatoski
K
,
Taichman
RS
, et al
New trends in the treatment of bone metastasis
.
J Cell Biochem
2007
;
102
:
1095
102
.
3.
Mehra
R
,
Kumar-Sinha
C
,
Shankar
S
,
Lonigro
RJ
,
Jing
X
,
Philips
NE
, et al
Characterization of bone metastases from rapid autopsies of prostate cancer patients
.
Clin Cancer Res
2011
;
17
:
3924
32
.
4.
Feldman
BJ
,
Feldman
D
. 
The development of androgen-independent prostate cancer
.
Nat Rev Cancer
2001
;
1
:
34
45
.
5.
Culig
Z
. 
Targeting the androgen receptor in prostate cancer
.
Expert Opin Pharmacother
2014
;
15
:
1427
37
.
6.
Isaacs
JT
. 
The biology of hormone refractory prostate cancer. Why does it develop?
Urol Clin North Am
1999
;
26
:
263
73
.
7.
Dehm
SM
,
Tindall
DJ
. 
Molecular regulation of androgen action in prostate cancer
.
J Cell Biochem
2006
;
99
:
333
44
.
8.
Taplin
ME
,
Balk
SP
. 
Androgen receptor: a key molecule in the progression of prostate cancer to hormone independence
.
J Cell Biochem
2004
;
91
:
483
90
.
9.
Felding-Habermann
B
. 
Integrin adhesion receptors in tumor metastasis
.
Clin Exp Metastasis
2003
;
20
:
203
13
.
10.
Nagle
RB
,
Knox
JD
,
Wolf
C
,
Bowden
GT
,
Cress
AE
. 
Adhesion molecules, extracellular matrix, and proteases in prostate carcinoma
.
J Cell Biochem
1994
;
19
:
232
7
.
11.
Hynes
RO
. 
Integrins: bidirectional, allosteric signaling machines
.
Cell
2002
;
110
:
673
87
.
12.
Cox
BD
,
Natarajan
M
,
Stettner
MR
,
Gladson
CL
. 
New concepts regarding focal adhesion kinase promotion of cell migration and proliferation
.
J Cell Biochem
2006
;
99
:
35
52
.
13.
Sheppard
D
. 
Epithelial integrins
.
Bioessays
1996
;
18
:
655
60
.
14.
Bates
RC
,
Bellovin
DI
,
Brown
C
,
Maynard
E
,
Wu
B
,
Kawakatsu
H
, et al
Transcriptional activation of integrin β6 during the epithelial–mesenchymal transition defines a novel prognostic indicator of aggressive colon carcinoma
.
J Clin Invest
2005
;
115
:
339
47
.
15.
Yokosaki
Y
,
Monis
H
,
Chen
J
,
Sheppard
D
. 
Differential effects of the integrins α9β1, αvβ3, and αvβ6 on cell proliferative responses to tenascin. Roles of the β1 subunit extracellular and cytoplasmic domains
.
J Biol Chem
1996
;
271
:
24144
50
.
16.
Fornaro
M
,
Plescia
J
,
Chheang
S
,
Tallini
G
,
Zhu
YM
,
King
M
, et al
Fibronectin protects prostate cancer cells from tumor necrosis factor α-induced apoptosis via the AKT/survivin pathway
.
J Biol Chem
2003
;
278
:
50402
11
.
17.
Altieri
DC
. 
Targeted therapy by disabling crossroad signaling networks: the survivin paradigm
.
Mol Cancer Ther
2006
;
5
:
478
82
.
18.
Shariat
SF
,
Lotan
Y
,
Saboorian
H
,
Khoddami
SM
,
Roehrborn
CG
,
Slawin
KM
, et al
Survivin expression is associated with features of biologically aggressive prostate carcinoma
.
Cancer
2004
;
100
:
751
7
.
19.
Zhang
M
,
Latham
DE
,
Delaney
MA
,
Chakravarti
A
. 
Survivin mediates resistance to antiandrogen therapy in prostate cancer
.
Oncogene
2005
;
24
:
2474
82
.
20.
O'Connor
DS
,
Grossman
D
,
Plescia
J
,
Li
F
,
Zhang
H
,
Villa
A
, et al
Regulation of apoptosis at cell division by p34cdc2 phosphorylation of survivin
.
Proc Natl Acad Sci U S A
2000
;
97
:
13103
7
.
21.
Weinreb
PH
,
Simon
KJ
,
Rayhorn
P
,
Yang
WJ
,
Leone
DR
,
Dolinski
BM
, et al
Function-blocking integrin αvβ6 monoclonal antibodies: distinct ligand-mimetic and nonligand-mimetic classes
.
J Biol Chem
2004
;
279
:
17875
87
.
22.
Dutta
A
,
Li
J
,
Fedele
C
,
Sayeed
A
,
Singh
A
,
Violette
SM
, et al
αvβ6 integrin is required for TGFβ1-mediated matrix metalloproteinase2 expression
.
Biochem J
2015
;
466
:
525
36
.
23.
Dutta
A
,
Li
J
,
Lu
H
,
Akech
J
,
Pratap
J
,
Wang
T
, et al
The αvβ6 integrin promotes an osteolytic program through upregulation of MMP2
.
Cancer Res
2014
;
74
:
1598
608
.
24.
Manes
T
,
Zheng
DQ
,
Tognin
S
,
Woodard
AS
,
Marchisio
PC
,
Languino
LR
. 
αvβ3 integrin expression up-regulates cdc2, which modulates cell migration
.
J Cell Biol
2003
;
161
:
817
26
.
25.
Fornaro
M
,
Manzotti
M
,
Tallini
G
,
Slear
AE
,
Bosari
S
,
Ruoslahti
E
, et al
β1C integrin in epithelial cells correlates with a nonproliferative phenotype: forced expression of β1C inhibits prostate epithelial cell proliferation
.
Am J Pathol
1998
;
153
:
1079
87
.
26.
Mesri
M
,
Wall
NR
,
Li
J
,
Kim
RW
,
Altieri
DC
. 
Cancer gene therapy using a survivin mutant adenovirus
.
J Clin Invest
2001
;
108
:
981
90
.
27.
Dickens
M
,
Rogers
JS
,
Cavanagh
J
,
Raitano
A
,
Xia
Z
,
Halpern
JR
, et al
A cytoplasmic inhibitor of the JNK signal transduction pathway
.
Science
1997
;
277
:
693
6
.
28.
Ghosh
JC
,
Dohi
T
,
Raskett
CM
,
Kowalik
TF
,
Altieri
DC
. 
Activated checkpoint kinase 2 provides a survival signal for tumor cells
.
Cancer Res
2006
;
66
:
11576
9
.
29.
Lu
ML
,
Schneider
MC
,
Zheng
Y
,
Zhang
X
,
Richie
JP
. 
Caveolin-1 interacts with androgen receptor. A positive modulator of androgen receptor mediated transactivation
.
J Biol Chem
2001
;
276
:
13442
51
.
30.
Zheng
DQ
,
Woodard
AS
,
Fornaro
M
,
Tallini
G
,
Languino
LR
. 
Prostatic carcinoma cell migration via αvβ3 integrin is modulated by a focal adhesion kinase pathway
.
Cancer Res
1999
;
59
:
1655
64
.
31.
Garlick
DS
,
Li
J
,
Sansoucy
B
,
Wang
T
,
Griffith
L
,
FitzGerald
TJ
, et al
αvβ6 integrin expression is induced in the POET and PTENpc−/− mouse models of prostatic inflammation and prostatic adenocarcinoma
.
Am J Transl Res
2012
;
4
:
165
74
.
32.
Roca
H
,
Craig
MJ
,
Ying
C
,
Varsos
ZS
,
Czarnieski
P
,
Alva
AS
, et al
IL-4 induces proliferation in prostate cancer PC3 cells under nutrient-depletion stress through the activation of the JNK-pathway and survivin up-regulation
.
J Cell Biochem
2012
;
113
:
1569
80
.
33.
Hubner
A
,
Mulholland
DJ
,
Standen
CL
,
Karasarides
M
,
Cavanagh-Kyros
J
,
Barrett
T
, et al
JNK and PTEN cooperatively control the development of invasive adenocarcinoma of the prostate
.
Proc Natl Acad Sci U S A
2012
;
109
:
12046
51
.
34.
Tran
C
,
Ouk
S
,
Clegg
NJ
,
Chen
Y
,
Watson
PA
,
Arora
V
, et al
Development of a second-generation antiandrogen for treatment of advanced prostate cancer
.
Science
2009
;
324
:
787
90
.
35.
Sheppard
D
. 
Integrin-mediated activation of latent transforming growth factor β
.
Cancer Metastasis Rev
2005
;
24
:
395
402
.
36.
Gioeli
D
,
Black
BE
,
Gordon
V
,
Spencer
A
,
Kesler
CT
,
Eblen
ST
, et al
Stress kinase signaling regulates androgen receptor phosphorylation, transcription, and localization
.
Mol Endocrinol
2006
;
20
:
503
15
.
37.
Weston
CR
,
Davis
RJ
. 
The JNK signal transduction pathway
.
Curr Opin Cell Biol
2007
;
19
:
142
9
.
38.
Cao
C
,
Mu
Y
,
Hallahan
DE
,
Lu
B
. 
XIAP and survivin as therapeutic targets for radiation sensitization in preclinical models of lung cancer
.
Oncogene
2004
;
23
:
7047
52
.
39.
Fedele
C
,
Singh
A
,
Zerlanko
BJ
,
Iozzo
RV
,
Languino
LR
. 
The αvβ6 integrin is transferred intercellularly via exosomes
.
J Biol Chem
2015
;
290
:
4545
51
.
40.
Shemshedini
L
,
Knauthe
R
,
Sassone-Corsi
P
,
Pornon
A
,
Gronemeyer
H
. 
Cell-specific inhibitory and stimulatory effects of Fos and Jun on transcription activation by nuclear receptors
.
EMBO J
1991
;
10
:
3839
49
.
41.
Bubulya
A
,
Chen
SY
,
Fisher
CJ
,
Zheng
Z
,
Shen
XQ
,
Shemshedini
L
. 
c-Jun potentiates the functional interaction between the amino and carboxyl termini of the androgen receptor
.
J Biol Chem
2001
;
276
:
44704
11
.
42.
Yang
YM
,
Bost
F
,
Charbono
W
,
Dean
N
,
McKay
R
,
Rhim
JS
, et al
C-Jun NH(2)-terminal kinase mediates proliferation and tumor growth of human prostate carcinoma
.
Clin Cancer Res
2003
;
9
:
391
401
.
43.
Wen
Y
,
Hu
MC
,
Makino
K
,
Spohn
B
,
Bartholomeusz
G
,
Yan
DH
, et al
HER-2/neu promotes androgen-independent survival and growth of prostate cancer cells through the Akt pathway
.
Cancer Res
2000
;
60
:
6841
5
.
44.
Vivanco
I
,
Palaskas
N
,
Tran
C
,
Finn
SP
,
Getz
G
,
Kennedy
NJ
, et al
Identification of the JNK signaling pathway as a functional target of the tumor suppressor PTEN
.
Cancer Cell
2007
;
11
:
555
69
.
45.
Koryakina
Y
,
Ta
HQ
,
Gioeli
D
. 
Androgen receptor phosphorylation: biological context and functional consequences
.
Endocr Relat Cancer
2014
;
21
:
T131
45
.
46.
Lin
HK
,
Yeh
S
,
Kang
HY
,
Chang
C
. 
Akt suppresses androgen-induced apoptosis by phosphorylating and inhibiting androgen receptor
.
Proc Natl Acad Sci U S A
2001
;
98
:
7200
5
.
47.
Chen
S
,
Xu
Y
,
Yuan
X
,
Bubley
GJ
,
Balk
SP
. 
Androgen receptor phosphorylation and stabilization in prostate cancer by cyclin-dependent kinase 1
.
Proc Natl Acad Sci U S A
2006
;
103
:
15969
74
.
48.
Shigemura
K
,
Isotani
S
,
Wang
R
,
Fujisawa
M
,
Gotoh
A
,
Marshall
FF
, et al
Soluble factors derived from stroma activated androgen receptor phosphorylation in human prostate LNCaP cells: roles of ERK/MAP kinase
.
Prostate
2009
;
69
:
949
55
.
49.
Saxena
P
,
Trerotola
M
,
Wang
T
,
Li
J
,
Sayeed
A
,
Vanoudenhove
J
, et al
PSA regulates androgen receptor expression in prostate cancer cells
.
Prostate
2012
;
72
:
769
76
.
50.
Mehrotra
S
,
Languino
LR
,
Raskett
CM
,
Mercurio
AM
,
Altieri
DC
. 
IAP regulation of metastasis
.
Cancer Cell
2010
;
17
:
53
64
.