Lemur Tyrosine Kinase 2, a novel target in prostate cancer therapy

Abstract

Progression from early forms of prostate cancer to castration-resistant disease is associated with an increase in signal transduction activity. The majority of castration-resistance cancers persist in the expression of the androgen receptor (AR), as well as androgen-dependent genes. The AR is regulated not only by it associated steroid hormone, but also by manifold regulatory and signaling molecules, including several kinases. We undertook evaluation of the role of Lemur Tyrosine Kinase 2 (LMTK2) in modulating AR activity, as several Genome Wide Association Studies (GWAS) have shown a marked association of LMTK2 activity with the development of prostate cancer. We confirm that not only is LMTK2 mRNA reduced in prostate cancer tissue, but also LMTK2 protein levels are markedly diminished. Knockdown of LMTK2 protein in prostate cell lines greatly increased the transcription of androgen-responsive genes. In addition, LMTK2 knockdown led to an increase in prostate cancer stem cell populations in LNCaP cells, indicative of increased tumorogenicity. Using multiple approaches, we also demonstrate that LMTK2 interacts with the AR, thus putting LMTK2 as a component of a signaling complex modulating AR activity. Our finding that LMTK2 is a negative regulator of AR activity defines a novel cellular pathway for activation of AR-responsive genes in castrate resistant-prostate cancer. Moreover, pharmacologic manipulation of LMTK2 activity will provide a novel therapeutic target for more effective treatments for patients with castrate-resistant prostate cancer.

Keywords: LMTK2; androgen receptor; castrate resistant prostate cancer; kinases; prostate cancer.

Figure 1. Predicted structure of Lemur Tyrosine Kinase 2 (LMTK2)

A. Topology of LMTK2 in endosomal membrane is shown with N-terminal and C-terminal either in cytosol or within an endosomal lumen [26]. B. LMTK2 is 1503 amino acid long protein with predicted transmembrane (TM) helices located between 11–29 and 46–63 amino acids while kinase domain is predicted to lie between 94–600 amino acid as shown in yellow. LMTK2 interacts with Protein Phosphatase 1 C (PP1C) via its VTF motif (1356–1358 amino acids) [25].

Figure 2. Expression and localization of LMTK2 in prostate epithelial cells

A. Immunoblot showing expression of LMTK2 in Human Embryonic Kidney Cell (HEK293), Human prostate epithelial cell (PTN1A), Prostate Cancer cell (PC3) & Prostate adenocarcinoma cell (LNCaP). Two concentrations for each cell type were loaded B. Top-left panel shows the mice prostate epithelial cell stained with secondary antibody only; top-right panel shows mice prostate epithelial cell marker cytokeratin 5/8 stained with alexa 488 (Green). LMTK2 is stained with alexa 488(Green) as shown in bottom left panel and bottom right panel shows the AR stained with CY-5 (Red). Nuclei stained in DAPI, appears blue. Bar = 0.25 μm. C. LMTK2 stained with Cy5 (red) and nuclei are stained with DAPI (blue), colocalization between LMTK2 and nuclei appears magenta. Bar = 0.5 μm. D. LNCaP cells grown in androgen deprived media were treated with DMSO or 2.5 nm R1881 (synthetic androgen) for 24 hours; cytoplasmic extract and nuclear extract were immunobloted to measure relative level of LMTK2 and AR. Δ-lamin and Δ-actin were used as nuclear and cytoplasmic protein control, respectively. E. Predicted Multiple Putative NLS and NES by “cNLS Mapper” & NetNES 1.1 (http://www.cbs.dtu.dk/services/NetNES/) and ValidNESs (http://validness.ym.edu.tw/).

Figure 3. LMTK2 is down regulated in human prostate cancer

A. Expression levels of LMTK2 protein in 48 prostate cancer, 8 prostate hyperplasia and 14 normal prostate tissue samples measured by tissue array. LMTK2 protein levels were detected by immunofluorescence staining. Intensity of staining was classified as no (0), low (0–20), medium (20–40), high (40–80) and very high (80–170). B. This box plot gives the cancer status (Normal or prostate adenocarcinoma) on the X-axis, and the protein expression of LMTK2 on the Y-axis. Error bars represent the interquartile range (IQR) of the measurements. The level of significance, P ≤ 0.001, was determined by Mann-Whitney-U analysis for pairwise comparison and circles indicate outliers. C. To generate curves, data were fit using a Best-Fit Gamma distribution (SigmaPlot, Systat, San Jose, CA). D. Representative Immunofluorescence images, LMTK2 stained with secondary antibody linked to Cy-5 (red) and nuclei stained in DAPI. Top panel shows low magnification image, the boxed region is magnified in the bottom panel. Scale bars = 100 μm.

Figure 4. LMTK2 and AR interact in prostate cancer epithelial cells and co-localizes in human prostate tissue

A. Protein within interacting distance i.e. <40 nm is detected using proximity ligation assay (PLA). Cartoon shows two interacting protein being detected by fluorescent signal. PLA between non-interacting proteins, CFTR and BAX was used as negative control. PLA between LMTK2 and AR in LNCaP cells grown in DMSO or 2.5 nM R1881 (androgen) shows PLA signal (Yellow dots), representing interaction. B. LMTK2-AR PLA signals per nucleus in LNCaP cells are counted using Image J software & compared with CFTR-BAX (non interacting proteins). Error bars indicate SD, which is given for mean of n ≥ 30 cells per condition. *P < 0.05 for difference from CFTR-BAX by student-unpaired t-test. C. Lysates from LNCaP cells was precipitated using mouse anti-AR Ab or control mouse Ab and blotted for LMTK2 or AR. 10% of lysate was loaded in the first lane as lysate control. D. Representative figure for co-localization between LMTK2 (red) and AR (green) in normal human prostate tissue is shown. Nuclei were stained with DAPI (blue). Arrows in the merge image shows some, but not all areas of co-localization (yellow) between LMTK2 and AR. Top panel shows low magnification image, the boxed region is magnified in the bottom panel.

Figure 5. LMTK2 inhibits AR transcriptional activity

A. Immunoblot comparing LMTK2 expression between HEK293 cells stably transfected with control shRNA (Parental 293), LMTK2 shRNA (LMTK2-KD) or PCI-LMTK2 (Overexpressing LMTK2). β-actin was used as endogenous loading control. B. Immunoblot comparing LMTK2 expression between LNCaP cells and LNCaP cell stably transfected with either control shRNA (LNCaP-control) or LMTK2 shRNA (LNCaP-KD). β-actin was used as endogenous loading control. C. Modulation of AR responsive gene activity by LMTK2 (mean ± SD, n = 3) *p < 0.05 for difference from endogenous LMTK2 levels in HEK293 cells. D. ELISA was performed in duplicates to analyze the concentrations of PSA protein in culture supernatant and cell lysate from LNCaP sublines cultured in androgen deprived medium for 3 days and treated with either DMSO or 2.5 nM synthetic androgen (R1881) for 16 hours (R1881) for 24 hours. Data is presented as % change in relation to the control (DMSO treated), (mean ± SD, n = 3) **p < 0.005 for difference from DMSO treatment group. E. Real-time PCR comparing transcript level of AR and genes regulated by AR between LNCaP sublines grown under androgen deprivation for 3 days was performed in duplicates (KLK2, S100P, TMPRS22, PSA), (mean ± SD, n = 3) **p < 0.01, ***p < 0.001 value for difference from endogenous LMTK2 levels in LNCaP cells are denoted on the graph. F, G. Immunobloting of FKBP51 in control and knock down cells grown under androgen deprivation for 3 days and treated with 1 nm R1881 for time points shown in figure. Quantitative analysis of normalized protein levels of FKBP51 between control and knock down LNCaP cells is shown in panel G.

Figure 6. LMTK2 down-regulation promotes tumor forming capacity and proliferation in LNCaP cells

A. Panel showing affects of endogenous LMTK2 levels on LNCaP tumor forming capacity. Representative colonies are shown in the figure. Quantitative analysis of colony numbers is shown in the panel B. (mean ± SD, n = 3) *p < 0.05 for difference from endogenous LMTK2 in LNCaP cells. C. Androgen-independent and androgen-stimulated cell viability of LNCaP-control and LNCaP-KD cell lines measured by the ATP-Glo™ Bioluminometric cell viability assay, (mean ± SD, n = 3) ****p < 0001, **p < 0.01, ****p < 0.0001 value for difference from endogenous LMTK2 in LNCaP cells.