|Home | About | Journals | Submit | Contact Us | Français|
PAX6, a transcription factor, has currently been suggested to function as a tumor suppressor in glioblastoma and to act as an early differentiation marker for neuroendocrine cells. The androgen receptor (AR) plays a pivotal role in prostate cancer development and progression due to its transcriptional activity in regulating genes involved in cell growth, differentiation, and apoptosis. To determine the role of PAX6 in prostate cancer, we investigated whether PAX6 interacts with AR to affect prostate cancer development.
We used immunostaining, RT-PCR, and western blotting assays to show the expression status of PAX6 in prostate tissue and human prostate cancer cell lines. The role of PAX6 in cell growth and colony regeneration potential of LNCaP cells were evaluated by MTT assay and soft agar assay with PAX6-overexpressed LNCaP cells. Mammalian two-hybrid and coimmunoprecipitation (CoIP) assays were used to demonstrate the interaction between PAX6 and AR. Reporter gene and Q-RT-PCR assays were performed to determine the effects of PAX6 on the function of AR.
In prostate cancer tissues, PAX6 expression was stronger in normal epithelial cells than cancer cells, and decreased in LNCaP cells compared to that of DU145 and PC3 cells. Enforced expression of PAX6 suppressed the cell growth of LNCaP cells and also inhibited the colony formation of LNCaP cells. PAX 6 interacted with AR and repressed its transcriptional activity. PAX6 overexpression decreased the expression of androgen target gene PSA in LNCaP cells.
In this study, we found that PAX6 may act as a prostate cancer repressor by interacting with AR and repressing the transcriptional activity and target gene expression of AR to regulate cell growth and regeneration.
PAX6, a member of the paired box gene family, functions as a transcription factor and is involved in various developmental processes. (1). It was first cloned in 1991 with a predicted 422-amino acid polypeptide product possessing a N-terminal paired domain (PD), a homeodomain (HD) in the middle, and a serine/threonine-rich C-terminal domain, all structural motifs characteristic of certain transcription factors (2). The loss of function studies linked the gene to the development of the eye, pituitary gland, neuroendocrine cells in the pineal body, and pancreatic Langerhans islet cells (3-6). As an early differentiation marker, PAX6 is also used to identify the precursors of neuron and neuroendocrine cells (7-10) in current stem cells studies.
In addition to its role on development and cell differentiation, PAX6 was also found to influence the fate of some cancer cells, possessing a tumor suppression function (11). In animal studies, the PAX6 mutant small eye mice were easily susceptible to leukemia (12). In human, Salem et al. identified that the PAX6 is hypermethylated in colon and bladder cancer cells (13). The increased methylation of the PAX6 promoter was observed in bladder cancer as well (14). Furthermore, PAX6 was also found to be hypermethylated in its promoter region and silenced in breast cancer lines and primary tumors, and furthermore, the transfection of PAX6 restored the expression of PAX6 back in MDA-MB-231 cells and consequently suppressed cancer cell growth (15). These results suggest PAX6 could act as a cancer suppressor since the low expression of PAX6 could be linked to cancer development. The similar cancer repressor’s role of PAX6 was also reported in the neuronal cancer, glioblastoma, in which PAX6 was found to suppress the growth and invasiveness of the cancer in vitro and in animal models (16,17). In summary, these studies suggest that the PAX6 acts as a tumor repressor and its silencing promotes the cancer growth. However, the role of PAX6 in prostate cancer is still unknown.
Prostate cancer remains a great threat to men’s health in the United States. The American Cancer Society estimated that 186,320 men would be diagnosed with prostate cancer in 2008, and approximately 28,660 men would die of the disease in the United States. (18). Therefore, it is of great interests to more clearly understand this cancer. Androgens and androgen receptor (AR) are involved in the normal development and maintenance of the prostate and thus the aberrant androgen/AR signaling plays a critical role in the growth and progression of prostate cancer (19). This recognition has greatly influenced therapeutic concepts for human prostate cancer during the last several decades (20). However, androgen deprivation therapy is usually associated with a gradual transition of the cancer cells from androgen-dependence (sensitive) to androgen depletion-independent (ADI) (21) (refractory) (22). A number of theories and mechanisms leading to the transition have been studied and one possibility is the ligand-independent activation of the AR (23). Neuroendocrine differentiation in prostate cancer is another possible mechanism proposed to lead to androgen insensitivity. The stem cell model was proposed to explain the derivation of neuroendocrine cells that usually are associated with androgen independence and androgen resistancefrom a small stem cell population (24). Thus, androgen deprivation therapy may force these cells to differentiate to androgen-independent neuroendocrine cells (24).
Although PAX6 may act as an early differentiation regulator for neuroendocrine cells as well as a tumor repressor (11), its dual roles in prostate has not been studied. As prostate cancer is an endocrine related cancer and the neuroendocrine differentiation is closely correlated to the prognosis, we hypothesized that PAX6 may play a role in the prostate cancer development.
To test our hypothesis, we used the androgen-dependent prostate cancer cell line, LNCaP, to assess the cellular and molecular effects of PAX6. Our results showed that PAX6 repressed cell growth and colony formation of LNCaP cells and inhibited the transcriptional activity of AR. Using the mammalian two-hybrid system and co-immunoprecipitation (Co-IP) method, we demonstrated interaction between PAX6 and AR. Furthermore, the exogenous expression of PAX6 in LNCaP cells inhibited the expression of prostate specific antigen (PSA) stimulated by androgen at the RNA and protein level. Taken together, these results suggest that PAX6 acts as a negative coregulator for AR by interacting with AR and repressing its function. This phenomenon could establish the tumor suppressor’s role of PAX6 in prostate cancer and help us understand the development and progression of prostate cancer.
The pCMX-PAX6 was a gift from Dr. Richard L. Maas. pSG5AR and VP-16AR were constructed as previously described (25). The insertion fragments of PAX6 were generated by PCR methods. The primers were designed based on the sequence of human PAX6 mRNA available on the Genome Data Base. The BamH1 restriction sequence was inserted at the 5′ end of each primer for cloning purposes and constructed into the pCMX-Gal4, pCMX-VP16, and pcDNA3-flag vector for different experiments. The anti-AR (N-20), anti-PSA, and anti-PAX6 antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). The anti-flag antibody was from Sigma (St. Louis, MO).
The human prostate cancer cell line, LNCaP, was maintained in RPMI 1640 medium containing penicillin (25 units/ml), streptomycin (25 μg/ml) and 10% fetal bovine serum. The COS-1 cells were maintained in Dulbecco’s minimum essential medium containing penicillin (25 units/ml), steptomycin (25 μg/ml) and 10% fetal bovine serum. For the Western blot, 8×105 LNCaP cells were plated on 100-mm dishes and transfected with different plasmids using Superfect (Qiagen, Valencia, CA). After transfection, the cells were maintained in the RPMI1640 supplemented with charcoal-dextran treated fetal bovine serum (CD-FBS) for 16 h, then subjected to different treatments for another 20-24 h. For the transactivation assay and the mammalian two-hybrid assay, the COS-1 cells were transfected using calcium phosphate precipitation method as described previously (26) and the LNCaP cells were transfected using Superfect. For the stable clone selection, the LNCaP cells transfected with either pcDNA-3-flag or pcDNA3-flab-PAX6 were selected using neomycin.
For immunohistochemistry, paraffin-embedded sections were heated at 55°C for at least 2 h, deparaffinized in xylene, rehydrated, and washed in Tris-buffered saline (TBS), pH 8.0. For antigen retrieval, slides were microwaved in 0.01 M sodium citrate/pH 6.0, immersed with 1% hydrogen peroxide in methanol for 30 min, and blocked with 20% normal goat serum in TBS for 60 min. After washing with PBS, sections were incubated for 90 min in PAX6 antibodies diluted 1:100 in TBS containing 1% BSA, followed by goat anti–rabbit biotinylated secondary antibody diluted 1:300 in TBS containing 1% BSA. Sections were incubated with avidin-biotin–peroxidase complex solution for 30 min, followed by development with diaminobenzidine peroxidase substrate kit (Vector Laboratories) for 5 min. Slides were counterstained with hematoxylin for 30 s, dehydrated, cleaned in xylene, and mounted. Primary antibody was replaced with normal rabbit IgGor 1% BSA in TBS for negative controls.
COS-1 cells were transfected with 5 μg of pSG5 AR plus 5 μg of pcDNA3-flag-PAX6 and harvested after 48 hr. Nuclear extracts were prepared for Co-IP assays. Four hundred microliters of each extract were first incubated with anti-Flag antibodies or anti-AR antibodies (N20, Sant Cruz, SantaCruz, CA, USA) overnight at 4 °C, and then with Protein G Sepharose for an additional 16 h at 4 °C. Protein G-Sepharose containing the immune complex was then washed three times with the washing buffer (50 mM Tris-HCl/pH 7.4, 100 mM NaCl, 5 mM CaCl2, 5 mM MgCl2, 0.1% Nonidet P-40) and resuspended in SDS-containing sample buffer. The flag and AR associated complexes were immunoprecipitated by anti-flag and anti-AR antibody, respectively. The isolated protein complexes were fractionated, transferred, and immunodetected with anti-AR or anti-flag antibodies. The proteins were resolved through a 6% SDS-PAGE and immunoblotted with anti-flag antibody or anti-AR antibody.
We suspended 5× 103 pcDNA3-Flag-PAX6 or pcDNA3-Flag transfected LNCaP cells in 0.3% Bacto-agar (BD, Lot. 5089021) and after transfection, layered the cells on top of 1 ml of 0.6% agarose in 6-well cultureplates, incubated with 1 ml RPMI 1640 supplemented with 10% FBS for 4 wk. After 4 wk, the colonies were visualized by staining with 0.5% crystal violet. The experiment was analyzed in triplicate, and colonies larger than 100 μm in diameter were counted.
LNCaP cells stably transfected with and without pcDNA3-PAX6 were incubated in 96- well dishes (1.5*103 cells/ well) for 5 days, MTT reagent (Promega, Madison, WI) was added at day 1, 3, and 5 per the manufacturer’s instruction. After 16 h of reaction, absorbance of the converted dye was measured at a wavelength of 595 nm with background subtraction at 650 nm.
Western blotting was performed as described previously (27). In short, cell extracts from LNCaP cells transfected with various vectors were prepared for electrophoresis run on SDS/PAGE gel and then transferred onto nitrocellulose (Minipore). The antibodies, anti-AR (Santa Cruz SC-816), anti-flag (Sigma F-3165), anti-PSA (Santa Cruz SC-7638), anti-Nkx 3.1 (Santa Cruz T 19, sc-15022), or anti-tubulin (Santa Cruz SC-5274) were used for the detection. The images were shown using alkaline phosphatase substrate color development method (Bio-Rad, Hercules, CA).
For COS-1 cells, 2 × 105 cells were plated on 6-well plates 12 h before being transfected with 0.5 μg of luciferase reporter and other expression vectors as indicated in the figures and legends. After 24 h transfection, the cells were treated with 10−8 M 5α-dihydrotestosterone (DHT) or ethanol for another 24 h. For each transfection, SV40 promoter driven Renilla luciferase (SV40RL) was used as an internal control. For LNCaP cells, the same volume of cells were seeded in 6-well plates 36 h before transfection, and the cells were transfected by Superfect (Qiagen) as described by the manufacturer. The treatment after transfection was the same as with the experiments performed in COS-1 cells.
RNA was extracted by using the Trizol reagent (Invitrogen), and first-strand cDNA was synthesized from 2 μg of total RNA in 20 μl reactions containing RT buffer with dNTPs, Oligo-dT, RNase inhibitor, and Moloney murine leukemia virus reverse transcriptase according to the manufacturer’s protocol (Invitrogen). Reverse transcription reactions were incubated at 72 °C for 5 min and then 25 °C for 10 min, followed by 42 °C for 60 min, and aliquots of the reaction products were used in later quantitative real-time PCR analysis. The quantitative PCR analysis was done using an ABI PRISM 7700 sequence detector system and the SYBR green PCR master mix kit (Applied Biosystems), according to the manufacturer’s suggestions. The housekeeping gene, β-actin was used for normalization. Relative mRNA expression was calculated by 2−ΔΔCt method as described (28). The following primer pairs were used: PSA forward 5′-CCGCATCTAATCGCTGGAGAG-3′, reverse 5′-CGATGTAGTTGGCGAAGCG-3′; β-actin forward 5′-CATATTCACCACCTCGGACAA-3′, reverse 5′- TGACGCCACAGACCACAC-3′, KLK2 forward 5′-CCTGGCAGGTGGCTGTGTAC-3′, reverse 5′- TGTGCCGACCCAGCCA-3′(29); and Nkx3.1 forward 5′-CGCAGCGGCAAGGC-3′, reverse 5′-GGTGCTCAGCTGGTCGTTCT-3′(30).
Since higher PAX6 expression level indicates improved prognosis in malignant astrocytic gliomas, whereas lower levels correlate with unfavorable outcomes (11), PAX6 was proposed to play an essential role in tumor progression. To determine whether PAX 6 affects prostate cancer development, we examined the expression of PAX6 in normal prostate tissues by immunostaining. Expression of PAX6 in the normal prostate epithelial cells (Fig. 1A, upper panel) was higher compared to PAX expression in adjacent cancer cells in the same tissue sections (Fig. 1A, middle panel) . We also examined the PAX6 expression in 3 different types of prostate cancer cells (PC3, DU145, and LNCaP cells). The RNA level of PAX6 was determined in prostate cells using reverse transcription-PCR analysis. The protein level of PAX6 was determined in prostate cells using western blotting analysis. As shown in Fig. 1B&C, PAX6 was expressed in all prostate cancer cells, but PAX6 was expressed higher in PC3 and DU145, which are AR negative cells compared to that of LNCaP cells, which are AR positive cells. Therefore, these results showed that PAX6 is expressed in prostate tissues and the expression level is lower in AR positive prostate cancer cells, as compared to AR negative cells.
Overexpression of PAX6 in glioblastoma cells suppresses cell growth (16), suggesting that PAX6 may act as a tumor repressor for certain cancers. Therefore, we evaluated the effects of PAX6 on prostate cancer growth. Because of the lower PAX6 expression in LNCaP cells, we stably transfected pcDNA3-flag-PAX6 into the LNCaP cells to establish PAX6-overexpressed LNCaP cells. The cell growth of PAX6-overexpressed LNCaP cells was assayed by the MTT assay and compared to that of parental LNCaP cells. As shown in Figure 2A, through days 1-4, PAX6-overexpressed LNCaP cells grew slower than parental LNCaP cells and on day 5, the growth of PAX6-overexpressed LNCaP cells showed about 57% lower than that of parental LNCaP cells. To determine whether the modulation of PAX6 expression affected the tumorigenic properties of the prostate cancer cells, we then tested whether PAX6 overexpression could affect the ability of LNCaP cells to form colonies in soft agar assay. LNCaP cells were transfected with either pcDNA3-Flag-PAX6 or pcDNA3-Flag. At 48 h after transfection, the cells were placed into medium with soft agar, and colonies were counted after 4 weeks. The colony numbers in the PAX6 transfected cells significantly decreased (about 48% reduction) compared with those of LNCaP cells transfected with control vectors (Fig. 2B and C). This result showed that the exogenous expression of PAX6 decreased the ability of LNCaP cells to regenerate in soft agar. Taken together, these results suggest that PAX6 may suppress the LNCaP cell growth and reproduction.
Androgens can activate the transcriptional activity of AR to affect the expression of genes that influence the prostate cancer cell growth. Therefore, the development and progression of prostate cancer are affected by the androgen/AR signaling pathway. Because PAX6 was demonstrated to repress LNCaP cells growth in previous results, the interactions between PAX6 and AR were assessed for the functional association between PAX6 and AR. To determine whether PAX6 affects AR transcriptional activity, three sets of different androgen response elements constructed in luciferase (Luc) reporter plasmids were used for reporter gene assays. AR expression vectors with various amounts of PAX6 expression vectors were transfected into COS-1 cells together with either mouse mammary tumor virus (MMTV), 4 copies of androgen-response element (ARE4), or PSA Luc reporter plasmids. Renilla SV40 Luc reporter was used as an internal control. The results shown in Fig. 3 A-C revealed that the transfection of PAX6 suppressed the AR transactivation in all three sets of reporter gene assays in the presence of DHT. The dose dependent effect was shown in MMTV-Luc. However, there was no significant dose-dependent effect in PSA-Luc and ARE4-Luc since the suppression is too profound to display the difference of the dose-dependent effect. These findings indicate that PAX6 represses AR transactivation.
The physical interaction between two transcriptional factors is one proposed mechanism to inhibit each other’s transcriptional activity. To clarify whether PAX6 directly interacts with AR, Co-IP and mammalian two-hybrid assays were performed. The interaction was first confirmed in the mammalian two-hybrid system. In the mammalian two-hybrid assay, pGal4-PAX6 and pVP16-AR plasmids were transfected into the COS-1 cells to generate Gal4-PAX6 fusion protein with the Gal4 DNA binding domain as well as VP16 activation domain. As shown in Fig. 4A, the luciferase activity was increased 4 fold over the DHT treated cells with co-transfection of pGal4-PAX6 and pVP16-AR plasmids comparing with the cells transfected with pGal4 and pVP16, suggesting that AR and PAX6 interact in the system. The directly physical interaction was demonstrated in the co-IP, pcDNA3-flag-PAX6 and pSG5-AR were transfected to COS-1 cells. After two days of transfection, the protein in the cells was harvested. The anti-AR antibody or anti-flag antibody was used to pull down the precipitate. Then the Western blot was performed with the AR or flag antibody (Fig. 4B). The results showed that using anti-flag antibody was able to immunoprecipitate PAX6 and pulled down AR and using anti-AR antibody pulled down AR and PAX6. These experiments indicate that PAX6 can interact with AR and form a protein complex in intact cells.
To determine the significance of PAX6 inhibitory effect on AR signaling, we examined the expression of AR target genes: prostate specific antigen (PSA), kallikrein 2 (KLK2) and the homeobox gene Nkx 3.1 in LNCaP cells (31,32). We compared the AR regulated gene level induced by DHT in LNCaP cells with or without the exogenous expression of PAX6 by transfecting PAX6 expression vectors or parent vectors. Using q-RT-PCR to determine the expression of PSA, we showed that PAX6 decreased the PSA (Fig. 5A), KLK2 (Fig. 5B), Nkx 3.1 (Fig. 5C) mRNA expression stimulated by DHT. In order to check whether the protein expression was also affected by PAX6, we performed Western blot analysis in the LNCaP cells transfected with vector or pcDNA3-flag-PAX6 and treated with either ethanol or 10 nM DHT. As shown in Figure 5D, cells transfected with pcDNA3-flag-PAX6 expressed a flag-tagged PAX6 fusion protein of 47 kDa detected by anti-flag Abs, but cells transfected with PCDNA3-flag only expressed flag protein of 1012 Da ( which was not detected in the position of flag-tagged PAX 6 fusion protein on western blot. Similar to the down-regulation of mRNA levels, we found that the LNCaP cells transfected with PAX6 and treated with DHT expressed less PSA and Nkx 3.1 proteins than the cells transfected with parental vectors (Fig. 5D). These results suggest that PAX6 is capable of repressing AR action on its positively regulated genes.
In this study, we found that PAX6 acts as a prostate cancer repressor by inhibiting AR action. Beside the tissue sites such as neuron, eye, pancreas, and embryo, the expression of PAX6 is found in small intestine, testis, and thymus (33). The diverse expression pattern of PAX6 suggests that it has a versatile role in tissue development and maintenance. Here, the results of immunohistochemistry performed on tissues sections (Fig. 1A) demonstrated that PAX6 expression is higher in normal prostate tissue adjacent to the cancer cells than cancer cells. After examining the expression of PAX6 in prostate cancer cells, we also found that the expression of PAX6 varied among 3 different prostate cancer cell lines with lowest expression in androgen-sensitive LNCaP cells (Fig. 2B&C). The differential expression profile of PAX6 in prostate cancer tissues and prostate cancer cell lines may be linked to prostate cancer development and progression related to androgen sensitivity.
Besides being an important protein for regulating differentiation, it is also possible that PAX6 plays a critical role in control of cell growth and differentiation in prostate. Since PAX6 is a key player in lens developmental processes and is involved in almost every step as the development proceeds, PAX6 frequently shifts its role between the repressor and the activator in different stages (34,35). PAX6 also interacts with numerous transcriptional factors during the processes of eye, neuron, and pancreas development, forming a possible network of transcriptional factors, which cooperate with PAX6 to fine-tune the expression of key genes related to development (5,34,36,37). However, in adult tissues, PAX6 may have another role. In pancreatic islet cells, the direct interaction between PAX6 and PPARγ resulted in decreasing PAX6 transactivation activity and reducing the glucagons secretion (19). This study and our results suggest the regulatory role of PAX6 by interacting with transcriptional factors such as PARRγ and AR to modulate gene expression, which could be the function of PAX6 after finishing tissue development in adult tissues.
In addition, the results of our growth assay and soft agar assay showed that PAX6 suppressed LNCaP cell growth and reproduction (Fig. 2), possibly by repressing the transcriptional activity of AR (Fig. 3). This is consistent with its tumor suppression role in glioblastoma (16). However, little is known about the mechanism of PAX6 suppression in cancer growth. One proposed mechanism regarding the influence of PAX6 on cancer physiology is that it may suppress the matrix metalloproteinase-2 (MMP-2) to reduce the invasiveness of glioblastoma cells (38). In contrast, androgen was reported to increase MMP-2 expression in prostate cancer (39). As PAX6 suppresses androgen action, it may also reduce MMP-2 in prostate cancer and suppress the progression of prostate cancer.
The AR mediates the action of androgens in the normal development and maintenance of prostate (40). Proteins that interact with AR and affect its transcriptional activity could contribute to prostate cancer progression due to the aberrant regulation of AR activity. Here, we demonstrated that PAX6 suppresses the AR transcriptional activity through the direct interaction with AR. As shown in Fig. 5, PAX6 was also able to decrease the expression of PSA, one of AR’s target genes in prostate and its expression is linked to the progression of prostate cancer. These findings establish a role of PAX6 in regulating prostate cancer development and progression. AR can interact with the other transcriptional factors as both belong to the family of DNA binding regulatory proteins. For examples, AR has been reported to interact with AP-1 and consequently disrupt the formation of the AR-ARE complex, resulting in the repression of PSA gene induced by androgen (41). Nuclear factor κB was shown to form a complex with AR, causing the repression of AR transactivation. (42,43) And sex-determining region Y was demonstrated to interact with and negatively regulate AR transcriptional activity (44). The transcriptional factor, AP-1/c-Jun, was also shown to interact with AR and suppress AR transcriptional activity (41). In summary, these findings provide the mutually interactive links among transcriptional factors, which may be essential for normal cell functions and once deregulated, may contribute to the development and progression of cancer.
In conclusion, we have found that PAX6 was expressed in the prostate tissue. It may have tumor cell suppression roles in the androgen-sensitive LNCaP cells. The direct interaction between PAX6 and AR, which leads to the decreasing AR transactivation activity and its target gene expression, may contribute to the tumor cell suppression effect of PAX6.
Grant sponsor: National Institute of Health Grant DK073414 and CA122295.