Loss of PTEN Accelerates NKX3.1 Degradation to Promote Prostate Cancer Progression

Abstract

NKX3.1 is the most commonly deleted gene in prostate cancer and a gatekeeper suppressor. NKX3.1 is a growth suppressor, mediator of apoptosis, inducer of antioxidants, and enhancer of DNA repair. PTEN is a ubiquitous tumor suppressor that is often decreased in prostate cancer during tumor progression. Steady-state turnover of NKX3.1 is mediated by DYRK1B phosphorylation at NKX3.1 serine 185 that leads to polyubiquitination and proteasomal degradation. In this study, we show PTEN is an NKX3.1 phosphatase that protects NKX3.1 from degradation. PTEN specifically opposed phosphorylation at NKX3.1(S185) and prolonged NKX3.1 half-life. PTEN and NKX3.1 interacted primarily in the nucleus as loss of PTEN nuclear localization abrogated its ability to bind to and protect NKX3.1 from degradation. The effect of PTEN on NKX3.1 was mediated via rapid enzyme-substrate interaction. An effect of PTEN on Nkx3.1 gene transcription was seen in vitro, but not in vivo. In gene-targeted mice, Nkx3.1 expression significantly diminished shortly after loss of Pten expression in the prostate. Nkx3.1 loss primarily increased prostate epithelial cell proliferation in vivo. In these mice, Nkx3.1 mRNA was not affected by Pten expression. Thus, the prostate cancer suppressors PTEN and NKX3.1 interact and loss of PTEN is responsible, at least in part, for progressive loss of NKX3.1 that occurs during tumor progression. SIGNIFICANCE: PTEN functions as a phosphatase of NKX3.1, a gatekeeper suppressor of prostate cancer.

Figure 1.

PTEN prolongs NKX3.1 half-life. A. LNCaP cells transfected with expression plasmids as indicated and were exposed to 100 μM cycloheximide. Cells were harvested at the indicated times and processed for immunoblotting. The plots show relative intensity of NKX3.1 bands analyzed by Image J. B. PTEN prevents stress-induced NKX3.1 degradation. LNCaP cells transfected with MYC or MYC-PTEN plasmids were exposed to TNFα, H₂O₂, etoposide, UV or γIR as shown. Cells were processed for immunoblotting.

Figure 2.

PTEN nuclear localization is required for interaction with NKX3.1. A. LNCaP cells transfected with a MYC-PTEN expression plasmid were exposed to 40ng/ml TNFα, 50 μM etoposide, 300 μM H₂O₂, 10 mJ/cm² UV, or 10 Gy γ-irradiation. Cells were processed for immunoblotting to analyze either whole cell lysates (WCL) or nuclear extracts (NE). B. 22RV1 and LAPC4 cells were treated as in Figure 2A without MYC-PTEN transfection to analyze nuclear localization of endogenous PTEN in response to stimuli. C. LAPC4 cells were exposed to 50 μM etoposide or 10 Gy γ-irradiation. Cells were harvested at the indicated times and processed as total lysates or nuclear extracts for immunoblotting (upper panel). In the lower panel cells were pretreated with 100 nM bortezomib prior to etoposide exposure or irradiation and then lysates were subjected to immunoprecipitation with NKX3.1 antiserum and then immunoblotting with the indicated antisera. D. LNCaP cells were transfected with the indicated expression plasmids, exposed to 10 Gy γ-irradiation, and harvested at the indicated times. Cells were processed similar to the experiment shown in panel C. E. LAPC4 cells cultured on polylysine-coated slides were fixed in 10% formalin 1 hour after. Staining was performed with monoclonal anti-PTEN antibody (Santa Cruz) and polyvalent NKX3.1 antiserum. The signals were amplified with HRP-conjugated antimouse IgG plus TSA, or biotin-conjugated antirabbit IgG and Texas Red. Nuclear PTEN was quantitated in three fields by Image J and then normalized to the average intensity in control samples. * Indicates a statistically different result from control by t-test, p<0.05.

Figure 3.

In vitro dephosphorylation of NKX31 by PTEN. His-tagged recombinant NKX3.1 was incubated with DYRK1B for 30 minutes and then incubated with PTEN for 60 minutes. A. Phosphorylated NKX3.1 were resolved on 8% of Phos-tag SDS-PAGE and revealed by NKX3.1 antiserum. B. Phosphorylated NKX3.1 was resolved on SDS-PAGE and stained with Coomassie blue. Bands were extracted for mass spectrometer analysis to identify dephosphorylation sites. Dephosphorylation was determined by subtracting the phosphorylated fraction of peptides in the reaction with both DYRK1B and PTEN from phosphorylation fraction of peptides in the sample with DYRK1B alone. The resulting differences were normalized to the phosphorylation fraction of NKX3.1 peptides extracted from the exposure to DYRK1B alone. C. LNCaP cells were transfected with expression plasmids for MYC-PTEN, pcDNA3 or MYC-NKX3.1(S185A). After 24 hours cells were treated with 100μM cycloheximide for the indicated times. Immunoblotting was performed with polyvalent NKX3.1 antiserum and PTEN antibody.

Figure 4.

Intracellular association of NKX3.1 and PTEN. A. Either 293T cells transfected with both Flag-PTEN and MYC-NKX3.1 plasmids or LNCaP cells transfected with a MYC-PTEN expression plasmid were processed for immunoprecipitation and immunoblotting for interaction of exogenously expressed proteins. LAPC4 cells were analyzed similarly for interaction of endogenous PTEN and NKX3.1. B. 293T cells were transfected with MYC-NKX3.1 or MYC-NKX3.1(S185A) plasmids with or without MYC-PTEN and were used for interaction analysis. Immunoprecipitation (right panel) was done with PTEN antibody. C. Cell lysates were treated with 100 units calf intestinal alkaline phosphatase for 60 minutes. The interaction of NKX3.1 and PTEN was abrogated.

Figure 5.

Effect of PTEN on NKX3.1 transcription. A. LNCaP cells were cultured and transfected with the indicated expression plasmids. After three days cells were processed for immunoblotting (inset in histogram) and quantitative PCR. B. 22Rv1 cells were cultured and transfected with either siRNA or guide RNA vectors to decreased expression of PTEN. After three days cells were processed as in A. C. LNCaP cells were cultured and transfected as in A except cells were processed at both 24 and 72 hours. D. Prostates from 5-month old mice were processed for RNA and analyzed by quantitative PCR. Expression of Nkx3.1 relative to Gapdh is shown. E. Effect of Pten phosphatase-deficient mutants on NKX3.1 mRNA 24 hours after transfection of the indicated constructs into LNCaP cells.

Figure 6.

Conditional knockout of Pten decreases both Pten and Nkx3.1 expression. Nkx3.1^CreERT2/+;Pten^flox/flox mice and Nkx3.1^CreERT2/+;Pten^+/+ mice were administered tamoxifen and analyzed by immunohistochemical staining as shown. Microscopic images of the comparator mice are juxtaposed for ease of comparison. The Pten signal was amplified by incubation with HRP-conjugated IgG and revealed with tyramide. Nkx3.1 was amplified with biotin conjugated anti-rabbit IgG and Texas red. Ki67, CK5 and CK8 were detected by Alexa fluor 488-, Alexa fluor 633- and Alexa fluor 568-conjugated IgG, respectively. The signals are shown as pseudocolors.

Figure 7.

Attenuation of Pten phosphatase reduces Nkx3.1 expression. Mice with the indicated Pten genotype were analyzed at 9 months of age as shown. The far right set of micrograms show prostates from Nkx3.^1−/− mice at 9 months of age. Immunofluorescent staining was performed as in Figure 6. B. The upper panel shows image J quantitation of Ki-67 staining of the different prostates and the lower panel staining for Nkx3.1. Results of t-test for significant differences are shown in the graphs. The data show general inverse relationship between Nkx3.1 and Ki-67 expression.