Research Article, J Clin Exp Oncol Vol: 5 Issue: 3
Loss of the Tumor Suppressor NKX3.1 in Prostate Cancer Cells is Induced by Prostatitis Related Mitogens
| Josua Decker‡1, Garima Jain‡1, Tina Kießling1, Philip Sander1, Margit Rid1, Thomas TF Barth1, Peter Möller1, Marcus V Cronauer2 and Ralf B Marienfeld1* | |
| 1Institute of Pathology, University of Ulm, Albert-Einstein-Allee 23, 89070 Ulm, Germany | |
| 2Department of Urology, Prittwitzstraße 43, 89075 Ul m, Germany | |
| ‡ These authors contributed equally to this work. | |
| Corresponding author : Ralf B. Marienfeld Institute of Pathology, University of Ulm, Albert-Einstein-Allee 23, 89070 Ulm, Germany Tel: +49 731 500 56384 E-mail: ralf.marienfeld@uni-ulm.de |
|
| Received: December 15, 2015 Accepted: July 05, 2016 Published: July 08, 2016 | |
| Citation: Decker J, Jain G, Kießling T, Sander P, Rid M, et al. (2016) Loss of the Tumor Suppressor NKX3.1 in Prostate Cancer Cells is Induced by Prostatitis Related Mitogens. J Clin Exp Oncol 5:3. doi:10.4172/2324-9110.1000160 |
Abstract
Objective: Prostate carcinoma (PCa) is the leading causes of cancer-related death in elderly men. Although several risk factors for the development of prostate cancer have been identified, the impact of chronic prostatitis is still a matter of debate. A key event of prostate cancer pathogenesis is the decrease of the homeo box protein NKX3.1 in the luminal epithelial cells of the prostate observed in early pre-cancerous lesions. Furthermore, inactivation of Nkx3.1 in a mouse model led to high incidence of prostatic intraepithelial neoplasia (PIN) formation underscoring the importance of NKX3.1 loss. In this study, we aimed to define the impact of diverse cytokines and growth factors known to be expressed during chronic prostatitis on NKX3.1 expression.
Methods: We determined the NKX3.1 expression in inflamed areas of prostatectomy specimens by immunohistochemistry. NKX3.1 protein and mRNA levels in cytokine and growth factor stimulated PCa cell lines were determined by western blot and RTqPCR. Transcriptional activity of the androgen receptor (AR) was determined by luciferase reporter assays and impact of the AR on NKX3.1 expression by siRNA mediated AR knock down.
Results: Treatment of prostate carcinoma cell lines with epidermal growth factor (EGF) dramatically reduced NKX3.1 protein and mRNA levels, while TNFα or IL-1α had only a moderate effect. Moreover, EGF or a combination of PMA and ionomycin (P+I) also caused diminished levels of the AR. However, while NKX3.1 reduction is observed as early as one hour after stimulation the decrease of AR occurred with a delayed kinetic. We show that P+I-induced NKX3.1 proteolysis is proteasome-dependent and influenced by protein kinase C.
Conclusion: In summary, we provide evidences for a crucial role of inflammatory mitogenic factors leading to reduced NKX3.1 and AR levels which might contribute to the initiation of pre-cancerous PIN lesions.
Keywords: Prostatitis; NKX3.1; AR; Mitogens; EGF; Prostate cancer
Keywords |
|
| Prostatitis; NKX3.1; AR; Mitogens; EGF; Prostate cancer | |
Abbreviations |
|
| AR: Androgen Receptor; BIM: Bisindolylmaleimide I; EGF: Epidermal Growth Factor; FFPE: Formalin-Fixed Paraffin Embedded; IL-1: Interleukin 1; PKC: Protein Kinase C; PMA: Phorbol Myristate Acetate; Tnfα: Tumor Necrosis Factor Alpha | |
Introduction |
|
| Prostate carcinoma (PCa) is the most common cancer in elderly men in the western hemisphere and is one of the leading causes of cancer-related death in men [1]. While genetic predisposition, age and diet are accepted risk factors for the development of PCa, the role of chronic prostatitis in PCa initiation is still a matter of debate [2-4]. Bacterial and non-bacterial prostatitis is a manifestation characterized by painful inflammation of the prostate appearing with a prevalence of 3 to 16% in elderly men. One potential link between chronic prostatitis and PCa initiation might be the increased expression of cytokines (like TNFα, IL-1α, IL-1β, IL-6, or IL-8) chemokines (like BCA1/CXCL13), and growth factors (like EGF, TGFα and VEGF) in the inflammatory areas [5-8]. Pro-inflammatory cytokines like TNFα and IL-1α, for instance, trigger the decrease of the homeobox protein NKX3.1 [9,10] which is a transcriptional repressor required for the differentiation of luminal cells of prostatic epithelium. A reduced NKX3.1 expression has been observed for the majority of early stages of human primary PCas and even in early pre-cancerous lesions like prostate intraepithelial neoplasias (PINs) or benign prostate hyperplasia (BPH) [11,12] and is thought to be an early event in the pathogenesis of PCa. In contrast to other tumour suppressors, a gradual decrease of cellular NKX3.1 levels is sufficient to promote PIN formation which is underlined by the increased incidence of PIN formation in a conditional prostate specific Nkx3.1 knock out mouse model [13]. Collectively, these data point at a role of NKX3.1 as a “gate-keeper” to block proliferation of differentiated luminal epithelial cells. To get a better insight into the role of chronic prostatitis for the initiation of PCa we set out to test the impact of a panel of cytokines and growth factors produced in the prostate during the different phases of inflammation on NKX3.1 protein and mRNA levels. The pro-inflammatory cytokines TNFα and IL-1α caused a distinct albeit moderate reduction of NKX3.1 protein and mRNA levels in LNCaP cells, while TWEAK and BCA-1 had only a weak impact on NKX3.1 levels. The strongest effect on NKX3.1 protein and mRNA levels was observed after stimulation with the growth factor EGF. As EGF exerts a strong mitogenic effect, we also studied the mitogenic PMA+ionomycin stimulation and observed a similar profound effect on NKX3.1 levels suggesting that mitogens in general lower cellular NKX3.1 levels. | |
Material and Methods |
|
| Tissue samples, tissue culture and proliferation assay | |
| Human PCa cell lines (LNCaP, VCaP, 22Rv1) were from ATCC (Manassas, VA, USA), LNCaP C4.2 cells were kindly provided by Prof. Reske, University of Ulm. Cells were maintained at 37°C in 5% CO2 in RPMI medium supplemented with 10% FBS, 1% penicillin/ streptomycin and 1% glutamine. The cells were checked by STRanalysis for identity and were routinely analysed for mycoplasma infections. For the proliferation assay 8×104 cells were used per well of a 12 well plate and the cells were stimulated for 48 hrs with cytokines or were left untreated. After 48 hrs incubation the cells were counted using a Neubauer chamber. All treatments were done in triplicates. The human PCa specimens were collected at the Institute for Pathology, University of Ulm, and FFPE material was processed using standard procedures. These specimens were derived from therapeutic prostatectomies performed at the Clinic for Urology, University of Ulm. The median age of the patients was 65 years (range, 54–72 years) and the median PSA value was 6.3 ng ml−1 (range, 4.1–25.6 ng ml−1) at diagnosis. On the basis of TNM staging from American Joint Committee on Cancer 6th edition (2002) for prostate cancer, the patients were staged at the time of diagnosis pT2a-3a, pN0, Mx, R0. Initial treatment consisted of radical prostatectomy. The samples were examined following approval of the Ethic commission of the University of Ulm. | |
| Antibodies, reagents, siRNAs and plasmids | |
| Anti-NKX3.1 (sc-15022) and anti-ERK2 (sc-154) antibodies were purchased from Santa Cruz Biotechnology (Santa Barbara, CA). Anti- AR (AR441), anti-CD3 (M 7254), anti-CD20 (M 0755), and anti- CD68 (M 0876) were from DAKO (Glostrup, Denmark). The siRNAs for human NKX3.1 and AR were purchased from Eurogentech (Cologne, Germany). The sequences of the siRNAs are available upon request. The AR-dependent luciferase reporter plasmids driven by the Probasin promoter (Probasinluc), the PSA promoter (PSAluc), or by the MMTV promoter (MMTVluc) were described elsewhere [14]. Human recombinant TNFα, IL-1α, BCA1, and EGF were obtained by Immunotools (Friesoythe, Germany), recombinant TWEAK was a kind gift of Dr. Harald Wajant (University of Würzburg). Phorbol-12- mysristate-13-acetate (PMA) was purchased from Merck Biosciences (Darmstadt, Germany). Bis-indolylmaleimide I (BIM, G2911) and ionomycin (I0634) were from Sigma-Aldrich (St. Louis, MO, USA). MG132 was obtained from Tocris Biosciences (Ellisville, MO, USA). | |
| Western blot analysis and immunohistochemical stainings | |
| For western blot analysis 25-50 μg of protein extract were loaded on a standard SDS-polyacrylamide (PAA) gel. SDS-PAGE and the transfer to nitrocellulose (Schleicher and Schuell, London, UK) were performed using standard protocols. The membrane was blocked with 5% milk powder in TBS+Tween 20 prior to the incubation with the primary antibody (1:1000 in TBS+Tween 20), subsequently washed three times for 5 minutes each and incubated in a TBS+Tween 20 solution containing either horse-radish peroxidase conjugated secondary antibody (1:5000). The detection was performed using ECL-substrates from Amersham Biosciences (Freiburg, Germany). To ensure an equal loading of the samples, we generally performed anti-ERK2 immunoblot analysis as a control. For the immunohistochemical stainings of the human PCa FFPE tissue samples 2 μm tissue slices were transferred to glass slides. The G-power software was used to determine the appropriate sample size to detect a statistically significant difference. The tissue samples were pre-treated according to the instructions by DAKO. Subsequently, the primary antibodies were applied at a dilution from 1:10 to 1:1000 for 30 minutes. The slides were washed extensively with PBS before applying a PBS solution containing the streptavidin-conjugated secondary antibody. After additional 30 minutes incubation, the slides were washed and stained using the permanent red substrate (K0695, DAKO) and were counter stained with hematoxilin/eosin. | |
| Luciferase reporter assay | |
| To determine the activity of AR, cells were transiently transfected using Lipofectamine LTX (Invitrogen; Carlsbad, CA, USA) with the AR reporter constructs (MMTVluc, Probasinluc, PSAluc; 200 ng per well) along with a plasmid encoding a Renilla-luciferase under the control of the human ubiquitin-promoter (30 ng per well). 24 h after transfection samples were treated with 50 ng EGF for 4h (stimulated) or left untreated. Luciferase activity was measured according to the Dual Luciferase Reporter assay system from Promega (Madison, WI, USA). Firefly luciferase activity was normalized to the approriate Renilla luciferase values. The experiments were done in duplicates and were repeated at least three times. For the siRNA transfection of LNCaP cells, the Viromer transfection kit (blue and green) from Lipocalyx (Halle, Germany) was used according to the manufacturer’s protocol. | |
| RNA extraction and Real-Time qRT-PCR–analysis | |
| Quantitative RT-PCR was performed by using the iCycler PCR instrument (Bio-Rad, Hercules, CA, USA). Total RNA was prepared using RNeasy protect kit from Qiagen (Hilden, Germany) according to the manufacturer’s protocol. 1μg of total RNA was used to generate cDNA using First-Strand Synthesis kit (Invitrogen). qRT-PCR was performed using 0.1 μl of cDNA reaction mix in the IQSYBRGreen supermix (Biorad). PCR was carried out as follows: after an initial three minute pre-incubation step at 95°C, 40 amplification cycles were run (95°C for 30s, 55°C for 30 s, and 72°C for 15 s). Quantification of gene regulation was performed by the ΔΔ Ct method. Results are presented relative to the expression of the house-keeping gene β-ACTIN or GAPDH. The primers were obtained from RealTimePrimers (Elkins Park, PA, USA). The sequences of Primers used for qPCR are available upon request. | |
| Statistical analysis | |
| All the results are expressed as the mean ± S.D. of at least 3 independent data sets. Student’s test was used to compare the mean of two groups and to calculate the p-value. P-value<0.05 was considered significant. | |
| Preparation of whole cell extracts, western blot analysis and immunohistochemical staining | |
| Whole cell extracts were prepared by using TNT buffer (20 mM Tris pH8.0, 200 mM NaCl, 1% TritonX100, 1mM DTT, 50 mM NaF, 50 mM α-glycerophosphate, 50 μM leupeptin, 1 mM PMSF). | |
| Quantification of NKX3.1 expression in FFPE specimens | |
| To estimate the nuclear NKX3.1 content in luminal epithelial cells, HE stained FFPE material was validated by an expert pathologist to exclude the presence of tumour areas in the FFPE tissue sample and to identify inflamed regions. After IHC staining of the FFPE slides, the areas in the proximity (in a radius of 5 μm) of CD3 (n=20), CD20 (n=20), or CD68 (n=20) were chosen for evaluation. Randomly selected areas, negative for CD3, CD20 or CD68 staining and without signs of inflammation were selected as negative controls (n=25). The intensity of epithelial NKX3.1 staining was estimated and was assigned to high, medium, and low expression groups. | |
Results |
|
| NKX3.1 expression is decreased in inflamed areas | |
| Reduced expression of NKX3.1 has been observed in early stages of PCa as well as in PIN lesions. To monitor NKX3.1 in chronic prostatitis, we compared inflamed and unaffected areas of human prostate tissue samples marked by immunohistochemical stainings for CD3 (T cells), CD20 (B cells), or CD68 (macrophages) in combination with anti-NKX3.1 stainings (Figure 1A and Supplementary Figure 1). The quantitative analysis revealed increased numbers of luminal epithelial cells with reduced NKX3.1 staining in areas with immune cell infiltration (Figure 1B). However, in areas with normal tissue a fraction of approximately 30% of luminal epithelial cells also displayed low nuclear NKX3.1 levels. Taken together, these results imply that inflammation might cause low nuclear NKX3.1 expression in luminal epithelial cells. | |
 |
Figure 1: Reduced NKX3.1 staining in areas of prostates with immune cell infiltrations. A, Representative images depicting NKX3.1 stainings (middle part) of prostate tissue samples with CD3 (lower part) positive areas. Nuclear NKX3.1 expression in unaffected areas (upper right part) or in CD3 positive areas (upper left part) is marked in higher magnifications of NKX3.1 stainings by arrows. B, Quantification of luminal epithelial cells positive for nuclear NKX3.1 staining. 20 either CD3, CD20, or CD68 positive areas or 25 areas without infiltrations from seven PCa samples were chosen. (ns=not significant, *=p<0.05, **=p<0.01, ***=p<0.001, ****= p<0.0001). |
| EGF stimulation causes NKX3.1 loss in PCa cell lines | |
| To elucidate the importance of cytokines or growth factors known to be expressed during inflammatory processes on NKX3.1 levels, we treated PCa cell lines LNCaP and LNCaP C4.2 with the pro-inflammatory cytokines TNFα and IL-1α, the cytokine BCA1 (CXCL13), TWEAK and the growth factor EGF which are both known to play a role in PCa [15,16]. In contrast to previously published data, we only detected a very moderate reduction in NKX3.1 protein levels upon TNFα and IL-1α stimulation (Figure 2A, lanes 1+3; Figure 2B, lanes 2+3) whereas neither TWEAK nor BCA1 exerted an effect on NKX3.1 expression (Figure 2A and Figure 2B, lanes 5+6). By contrast, EGF stimulation caused a strong NKX3.1 protein reduction (Figure 2A and Figure 2B, lane 4). Similarly, NKX3.1 mRNA levels were strongly reduced upon EGF treatment (approx. 60%), while IL-1α and TNFα were less effective (approx. 40%) and BCA1 or TWEAK had no effect (Figure 2C). EGF, TNFα and IL- 1α have mitogenic potential. However, in our hand only EGF, but not TNFα or IL-1α treatment, caused a distinctively increased proliferation of LNCaP cells (Figure 2D). | |
|
Figure 2: Reduced NKX3.1 expression levels in PCa cell lines treated with EGF. Anti-NKX3.1 (upper panels) or anti-ERK2 (lower panels) western blot analyses using whole cell extracts from LNCaP (A) and LNCaP C4.2 cells (B) treated with the indicated stimuli for 18 hrs or left untreated (control, C). Anti- ERK2 western blot analyses were performed to ensure an equal loading of the samples. C, Quantitative RT-PCR analysis of NKX3.1 mRNA levels of LNCaP cells treated with the indicated stimuli for 18 hrs. All samples were measured in triplicates and relative NKX3.1 expression levels were determined by theΔΔCt method. The relative NKX3.1 mRNA level in the control samples were arbitrarily set to 1 (ns=not significant, *=p<0.05, **=p<0.01, ***=p<0.001, ****= p<0.0001) D, EGF increases LNCaP cell proliferation. LNCaP cells were either left untreated (Co.) or were stimulated with either TNFα (50ng/ml), IL-1α (20ng/ ml) or EGF (100ng/ml) for 24 hours. Subsequently, the cells were harvested, the cell numbers were determined and the increase in comparison to the cell numbers at begin is depicted. All stimulations were done in triplicates. |
| Loss of NKX3.1 is caused by PMA+ionomycin stimulation | |
| As these results point to a potential role of the mitogenic potential of EGF for the reduction of NKX3.1, we next used a combination of the strong mitogens PMA and ionomycin (P+I) which imitates the effects of a phospholipase C mediated hydrolysis of phosphatidylinositol- diphosphate (Figure 3A and 3B). Similar to EGF treatment, stimulation with P+I led to similarly reduced NKX3.1 protein levels in all analysed PCa cell lines (Figure 3B). Furthermore, PMA stimulation alone is sufficient to drive NKX3.1 reduction whereas calcium influx augmented the PMA induced NKX3.1 reduction but had no effect on its own (Figure 3C). A slight NKX3.1 reduction was already seen after 1 hour P+I stimulation (Figure 3D, upper panel), and after 5 hours NKX3.1 was almost completely absent. By contrast, EGF mediated NKX3.1 reduction was slightly delayed and detectable after 4 hours stimulation (Figure 3E, lane 3), but was comparable after 6 hours EGF and P+I stimulation (Figure 3E, lanes 4+6). Importantly, AR levels also dropped in P+I stimulated LNCaP cells albeit with a somewhat delayed kinetic showing reduced AR levels after 7 hours stimulation (Figure 3D, middle panel). Similarly, P+I or EGF stimulation had a more dramatic effect on NKX3.1 mRNA (Supplementary Figure 1B) than on AR mRNA levels (Supplementary Figure 1D). By contrast, the expression of RAMP-1 (Supplementary Figure 1A) or TWIST-1 (Supplementary Figure 1C), known to be suppressed by NKX3.1, was found to be augmented upon EGF (RAMP-1 and TWIST-1) or P+I (RAMP-1) stimulation. To exclude that NKX3.1 reduction is due to attenuated AR activity, we performed luciferase reporter assays using three different AR reporter constructs. None of these AR reporter assays revealed a diminished AR activity upon EGF stimulation implying that NKX3.1 decrease occurs independent of AR activity. To determine whether the delayed AR reduction is a consequence of reduced NKX3.1 levels, we analyzed AR or NKX3.1 expression after transfecting siRNAs for both factors. While suppression of AR reduced basal NKX3.1 mRNA levels (Figure 4A), AR mRNA levels remained unaffected by knock-down of NKX3.1 (Figure 4B). Interestingly, however, EGF stimulation had only a very moderate effect on AR and NKX3.1 mRNA and/or protein levels in NKX3.1siRNA transfected LNCaP cells, whereas the EGF effect on NKX3.1 mRNA levels was still observed in ARsiRNA transfected cells. | |
|
Figure 3: P+I stimulation causes NKX3.1 loss in PCa cell lines. A+B, Western blot analyses of whole cell extracts from 22Rv1 cells (left), LNCaP cells (center), and LNCaP C4.2 cells (right) either left untreated (C) or stimulated with either EGF (A) or P+I (B). C, PMA is sufficient to induce NKX3.1 loss. LNCaP cells were either left untreated, or were treated with PMA, ionomycin, PMA+ionomycin, or TNFα. The resulting whole cell extracts were subjected to western blot analyses with the indicated antibodies. D, Western blot analyses of LNCaP cells treated with P+I for indicated time periods. Whole cell extracts were monitored with anti-NKX3.1 (upper panel), anti-AR (middle panel), or anti-ERK2 (lower panel). E, Comparison of EGF and P+I induced NKX3.1 loss. LNCaP cells were either left untreated or were stimulated with P+I or EGF as indicated. Anti-NKX3.1 and anti–ERK2 western blot analyses were performed with the resulting western blots. |
|
Figure 4: AR mRNA levels remain unchanged by knock-down of NKX3.1 in LNCaP cells. LNCaP cells transfected with the indicated siRNAs and were either left untreated or were stimulated with EGF for 16 hours. mRNA levels of NKX3.1 (A) or AR (B) were determined by quantitative RT-PCR analysis (ns=not significant, *=p<0.05, *=p<0.01, ***=p<0.001, ****=p<0.0001) C, A fraction of the LNCaP cells analysed was subjected to western blot analyses using the indicated antibodies. |
| P+I and EGF induced NKX3.1 loss is mediated by the proteasome and regulated by PKC | |
| Having established P+I and EGF as potent inducers of NKX3.1 loss, we next aimed at defining the molecular mechanisms underlying this process. Since PMA is a strong inducer of protein kinase C (PKC) activity, we first dissected the role of PKC in P+I induced NKX3.1 loss. Indeed, pre-treatment of LNCaP cells with the pan-PCK inhibitor bisindolylmalmeide I (BIM) stabilized NKX3.1 protein levels (Figure 5A) as well as NKX3.1 mRNA levels (Figure 5B) after PMA or P+I stimulation. By contrast, BIM had no effect upon EGF stimulation suggesting that PKCs are involved in the P+I, but not in the EGF induced NKX3.1 reduction. Since TNFα and IL-1α have been reported to cause a phosphorylation-dependent proteasomal degradation of NKX3.1 [9], we next determined the role of the proteasome by using the proteasome inhibitor MG132. Indeed, MG132 pre-treatment stabilized the NKX3.1 protein during P+I or EGF stimulation to a similar degree (Figure 6A). However, although MG132 pre-treatment further augmented EGF induced reduction of NKX3.1 mRNA levels, the loss of NKX3.1 after P+I stimulation was partially reverted, while AR mRNA remained unaffected upon P+I and MG132 treatment (Figure 6B and Figure 6C). Taken together, these results clearly show that the EGF and P+I induced loss of NKX3.1 is based on two different processes: Reduced mRNA levels (reduced transcription or mRNA stability) and a proteasomal degradation of pre-existing NKX3.1 protein. | |
|
Figure 5: P+I induced NKX3.1 loss is mediated by protein kinase C.A, Western blot analyses of whole cell extracts from LNCaP cells pre-treated with the PKC inhibitor bis- indolylmaleimide I (BIM) for one hour prior to a stimulation with either PMA or P+I for six additional hours. B, Comparison of the effect of PKC inhibition on EGF and P+I induced NKX3.1 loss. mRNA from LNCaP cells pre-treated with (BIM) and subsequently stimulated with either EGF or P+I were subjected to qRT-PCR analysis. (ns=not significant, *=p<0.05, **=p<0.01, ***=p<0.001, ****=p<0.0001). |
|
Figure 6: NKX3.1 proteolysis is mediated by the proteasome. A, NKX3.1 protein levels were determined by western blot analyses of whole cell extracts from LNCaP cells pre-treated with either MG132 prior to stimulation with EGF or P+I. A fraction of the LNCaP cells pre-treated with MG132 were subjected to qRT-PCR analyses for AR (B) or NKX3.1 mRNA (C) levels. (ns=not significant, *=p<0.05, **=p<0.01, ***=p<0.001, ****=0.0001). |
Discussion |
|
| The impact of chronic prostatitis on PCa initiation still remains a matter of discussion [3,17,18]. One potential link between chronic prostatitis and initiation of PCa could be the loss of the tumour suppressor NKX3.1 caused by cytokines or growth factors which has already been shown in murine models of bacterial prostatitis [19,20]. During chronic prostatitis a multiplicity of growth promoting cytokines and growth factors (like TNFα, IL-1α, EGF, etc.) are expressed which in turn tip the balance toward a prolonged loss of NKX3.1 causing reduced p53 levels and as a consequence an increased mutation rate [21]. Thus, the increased and prolonged expression of cytokines and growth factors might be the reason why prostatitis represents a risk factor for PCa initiation, quite similar to the role of inflammatory processes in the development of other types of cancer. Indeed, a TNFα or IL1α induced NKX3.1 proteolysis has been reported previously [9]. In addition to reduced NKX3.1 protein levels (Figure 2A and Figure 2B), we observed also a distinct reduction of the NKX3.1 mRNA levels in TNFα or IL-1α treated LNCaP cells (Figure 2C). However, a more prominent effect on NKX3.1 mRNA levels was obtained by EGF or P+I stimulation. EGF is a growth factor known to be expressed during late stages of inflammation and during wound healing but is also thought to be important for the homeostasis of the prostate epithelium [16,22] Thus, the approximately 30 percent of luminal epithelial cells with low or absent NKX3.1 expression observed in normal prostate tissue areas might be due to EGF driven homeostasis. A mutual exclusive expression of NKX3.1 and the AR in PCa samples was shown previously [21] suggesting that the proliferative effect of the AR also requires low NKX3.1 levels. As NKX3.1 is a well-known AR target gene, it is tempting to speculate that an EGF or P+I induced attenuation of AR activity might by responsible for the observed down regulation of NKX3.1 mRNA levels. EGF treatment has been shown previously to reduce AR protein and mRNA levels and impairs the expression of AR target genes in differentiated vas efference epithelial cells [23]. Indeed, we also observed a distinct reduction in AR protein levels after P+I stimulation (Figure 3D) and occasionally also reduced AR mRNA levels after P+I or EGF stimulation (Figure 4B, Figure 6B, Supplementary Figure 1D). However, a decrease of the AR protein was visible not until six hours of P+I stimulation, whereas NKX3.1 protein and mRNA levels dropped as soon as one hour of P+I stimulation (Figure 3D and data not shown). Furthermore, the transcriptional activity of AR remained unchanged after treatment of LNCaP cells with EGF as shown by luciferase reporter assays with three different AR-dependent luciferase reporter constructs (Supplementary Figure 1E). Given these results, it seems unlikely that the loss of NKX3.1 mRNA is only caused by reduced AR protein levels or an attenuation of AR activity. Importantly, the EGF and P+I induced NKX3.1 loss is based on two mechanisms: A down-modulated mRNA production and a proteasomal degradation of the preexisting NKX3.1 (Figure 6). Proteasomal degradation has already been reported for the NKX3.1 loss mediated by pro-inflammatory cytokines [9]. This TNF α induced NKX3.1 proteolysis depends on serine-phosphorylation, while an ATM mediated phosphorylation at Thr134 and Thr166 regulates the NKX3.1 loss after UV induced DNA damage [24]. By contrast, CK2 or PIM-1 mediated phosphorylation stabilizes NKX3.1 in PCa cell lines [25,26]. Whether EGF or PMA+ionomycin induced NKX3.1 loss requires specific NKX3.1 phosphorylations or to a removal of stabilizing phosphorylations remains currently unknown. Although EGF or P+I stimulation induce NKX3.1 losses to a comparable degree, there are some mechanistic differences. Inhibition of the proteasome by MG132 stabilized NKX3.1 upon EGF and P+I stimulation. However, while NKX3.1 mRNA levels were even further reduced after EGF + MG132 stimulation, a partial rescue was observed in case of P+I stimulation. Moreover, while an inhibition of the PKC family by the pan-PKC inhibitor bis-indolylmaleimide I (BIM) blocked P+I induced reduction of NKX3.1 protein or mRNA levels, no effect on EGF induced reduction of NKX3.1 mRNA was observed. Collectively, these data show that EGF and P+I induced NKX3.1 loss involves divergent signal transduction pathways and suggest that NKX3.1 stabilization by PKC and proteasome inhibition in case of P+I stimulation might occur on the transcriptional level. Taken together, we end up to propose that chronic prostatitis might serve as a risk factor for the development of PINs and PCa by causing increased expression of growth factors involved in repair processes (like EGF). Growth factors (or i. g. mitogenic stimuli) subsequently cause reduced levels of the tumor suppressor NKX3.1 which in turn promote cell proliferation, de-differentiation of luminal epithelial cells, and attenuate DNA repair capabilities. | |
Acknowledgments |
|
| We thank Mrs. Julia Melzner and Mrs. Iwona Nerbas for the help with the immunohistochemical stainings. This work was supported by a grant from the Dr.-Robert-Pfleger-Stiftung (to R.B.M. and M.V.C.) and by a fellowship from the Deutsche Forschungsgemeinschaft (GRK1041, to G.J.). | |
Conflict of Interest |
|
| The authors declare no conflicts of interests. | |
References |
|
|
|
