Acid ceramidase promotes nuclear export of PTEN through sphingosine 1-phosphate mediated Akt signaling

Abstract

The tumor suppressor PTEN is now understood to regulate cellular processes at the cytoplasmic membrane, where it classically regulates PI3K signaling, as well as in the nucleus where multiple roles in controlling cell cycle and genome stability have been elucidated. Mechanisms that dictate nuclear import and, less extensively, nuclear export of PTEN have been described, however the relevance of these processes in disease states, particularly cancer, remain largely unknown. We investigated the impact of acid ceramidase on the nuclear-cytoplasmic trafficking of PTEN. Immunohistochemical analysis of a human prostate tissue microarray revealed that nuclear PTEN was lost in patients whose tumors had elevated acid ceramidase. We found that acid ceramidase promotes a reduction in nuclear PTEN that is dependent upon sphingosine 1-phosphate-mediated activation of Akt. We were further able to show that sphingosine 1-phosphate promotes formation of a complex between Crm1 and PTEN, and that leptomycin B prevents acid ceramidase and sphingosine 1-phosphate mediated loss of nuclear PTEN, suggesting an active exportin-mediated event. To investigate whether the tumor promoting aspects of acid ceramidase in prostate cancer depend upon its ability to export PTEN from the nucleus, we used enforced nuclear expression of PTEN to study docetaxel-induced apoptosis and cell killing, proliferation, and xenoengraftment. Interestingly, while acid ceramidase was able to protect cells expressing wild type PTEN from docetaxel, promote proliferation and xenoengraftment, acid ceramidase had no impact in cells expressing PTEN-NLS. These findings suggest that acid ceramidase, through sphingosine 1-phosphate, promotes nuclear export of PTEN as a means of promoting tumor formation, cell proliferation, and resistance to therapy.

Figure 1. Human prostate tissues lose nuclear PTEN when AC is elevated.

(a-d) A 27 patient tissue microarray with patient-matched prostate adenocarcinoma and adjacent benign tissue was immunostained for AC and PTEN. Staining intensities for AC, nuclear PTEN, and cytoplasmic PTEN were evaluated by a blinded pathologist. The arrows in (b) are examples of nuclei devoid of PTEN. In patients whose cancer tissue had elevated AC than their benign tissue (a), there was also a decrease in the amount of nuclear PTEN, p<.05 (b). Patients whose AC did not increase in their tumor tissue (c) did not have a decrease in nuclear PTEN (d). AC pathology scores and nuclear PTEN percentage for the AC-high and AC-low patient groups are organized in (e).

Figure 2. Acid ceramidase promotes S1P-mediated loss of nuclear PTEN.

PPC1 cells transfected with WT-PTEN were infected with Ad-GFP or Ad-AC for 48 hours in the presence of DMSO (no treatment; NT) or the sphingosine kinase inhibitor SKI-II for 24 hours. A) Cells were immunostained for PTEN (red) and nuclei (blue). Nuclear (N) and cytoplasmic (C) PTEN staining intensity were measured for all cells in a given treatment using ImageJ. N/C indicates the nuclear PTEN to cytoplasmic PTEN ratio. B) The percentage of cells from (A) which had nuclear PTEN in each treatment. C) Nuclear fractions from the indicated treatments were isolated and evaluated for presence of PTEN with Histone H3 as a nuclear loading control and absence of β-tubulin to indicate purity of the nuclear sample. D) PPC1 cells transfected with WT-PTEN were treated with the indicated dose of S1P or PBS for 2 hours prior to fixation and immunostaining for PTEN (red) and nuclei (blue). E) The percentage of cells from (D) which had nuclear PTEN. F) Nuclear fractions from the indicated treatments were isolated and evaluated for presence of PTEN with Histone H3 as a nuclear loading control and absence of β-tubulin to indicate purity of the nuclear sample. One way ANOVA with Bonferroni correction, *p<.05, **p<.01.

Figure 3. AC/S1P promote Akt-dependent export of nuclear PTEN.

PPC1 cells were transfected with WT-PTEN were infected with Ad-GFP or Ad-AC for 48 hours in the presence of DMSO (NT) or the indicated compounds for 24 hours. A) Nuclear fractions from the indicated treatments were isolated and evaluated for presence of PTEN with Histone H3 as a nuclear loading control and absence of β-tubulin to indicate purity of the nuclear sample. B) Cells were immunostained for PTEN (red) and nuclei (blue). C) The percentage of cells from (B) which had nuclear PTEN in each treatment. D) PPC1 cells transfected with WT-PTEN were treated with 1µM JTE013 or 5µM AktX for 24 hours prior to treatment with the indicated dose of S1P or PBS for 2 hours followed by fixation and immunostaining for PTEN (red) and nuclei (blue). E) The percentage of cells from (D) which had nuclear PTEN. F) Nuclear fractions from the indicated treatments were isolated and evaluated for presence of PTEN with Histone H3 as a nuclear loading control and absence of β-tubulin to indicate purity of the nuclear sample. One way ANOVA with Bonferroni correction, *p<.05**p<.01.

Figure 4. AC/S1P promotes Leptomycin B sensitive export of nuclear PTEN.

PPC1 cells transfected with WT-PTEN were infected with Ad-GFP or Ad-AC for 48 hours in the presence of EtOH (NT) or 100 nM Leptomycin B (LMB) for 24 hours. A) Cells were immunostained for PTEN (red) and nuclei (blue). B) The percentage of cells from (A) which had nuclear PTEN in each treatment. C) Nuclear fractions from the indicated treatments were isolated and evaluated for presence of PTEN with Histone H3 as a nuclear loading control and absence of β-tubulin to indicate purity of the nuclear sample. D) PPC1 cells transfected with WT-PTEN were treated with 100nM LMB for 24 hours and 500nM S1P or PBS for 2 hours prior to fixation and immunostaining for PTEN (red) and nuclei (blue). E) The percentage of cells from (D) which had nuclear PTEN. F) Nuclear fractions from the indicated treatments were isolated and evaluated for presence of PTEN with Histone H3 as a nuclear loading control and absence of β-tubulin to indicate purity of the nuclear sample. One way ANOVA with Bonferroni correction, *p<.05**p<.01.

Figure 5. PTEN and Crm1 form an S6K mediated complex upon S1P stimulation.

A) PPC1 cells were transfected with WT-PTEN and FLAG-Crm1. Cells were collected after 2 hours stimulation with 500nM S1P or PBS and submitted to immunoprecipitation of the FLAG-Crm1 protein. The negative control (Neg) indicates lysate from cells not transfected with FLAG-Crm1. B) PPC1 cells were transfected with FLAG-PTEN and collected after 2 hour stimulation with 500nM S1P or PBS and submitted to immunoprecipitation of the FLAG-PTEN protein. The negative control (Neg) indicates lysate from cells not transfected with FLAG-PTEN. C) PPC1 cells were infected with Ad-AC or Ad-GFP and analyzed for S6K phosphorylation. D) PPC1 cells were treated with 500nM S1P for 2 hours and analyzed for S6K phosphorylation. E) PPC1 cells transfected with WT-PTEN were treated with water (NT) or 2.5 µM S6K1 inhibitor DG2 (S6Ki) for 24 hours then stimulated with 500 nM S1P for 2 hours prior to immunostaining for PTEN (red) and nuclei (blue). F) The percentage of cells from (E) which had nuclear PTEN in each treatment. G) PPC1 cells were transfected with WT-PTEN and FLAG-Crm1 and treated for 24 hours with 2.5 µM S6Ki. Cells were collected after 2 hour stimulation with 500nM S1P or PBS and submitted to immunoprecipitation of the FLAG-Crm1 protein. One way ANOVA with Bonferroni correction, *p<.05**p<.01.

Figure 6. AC promotes tumor formation and Docetaxel resistance in cells with wild type, but not nuclear localized, PTEN.

A-D) PPC1 cells were infected with 25 MOI Ad-GFP or Ad-AC and either 25 MOI Ad-GFP, Ad-WT-PTEN, or Ad-PTEN-NLS. A) After 24 hours plating, cells were treated with 1.5nM Docetaxel and after a further 48 hours, stained with propidium iodide and analyzed for apoptotic cells using FACS. B) After 24 hour attachment, cells were treated with a dose course (.01 to 100nM) Docetaxel and analyzed for relative cell viability using MTS assay after a further 48 hours. The EC50 was estimated using Prism 4 software. C) Cells were counted on the indicated day (day 0 being the day of plating). Student’s t-test, *p<.05, **p<.01. D) 4x106 cells were injected into the flanks of nu/nu mice and observed for 6 weeks. We monitored the mice each week for the formation of palpable tumors and graph the number of mice in the indicated treatment that had established palpable tumor at the indicated day.