Endocrine-Related Cancer (2011) 18 245–255

Androgen levels are upregulated by Akt in prostate cancer

Susan Ha1,2*, Rachel Ruoff1,2*, Nicole Kahoud 2,4*, Thomas F Franke1,3 and Susan K Logan1,2

Departments of 1Pharmacology 2Urology and NYU Cancer Institute and 3Psychiatry, New York University School of Medicine, 550 First Avenue, MSB424, New York, New York 10016, USA 4Department of Nephrology, University of Washington, Seattle, Washington 98195, USA (Correspondence should be addressed to S K Logan; Email: [email protected]) *(S Ha, R Ruoff and N Kahoud contributed equally to this work)

Abstract Multiple lines of evidence suggest a functional link between the (AR) and the serine/threonine kinase Akt in the development and progression of prostate cancer. To investigate the impact of Akt activity on AR homeostasis, we treated androgen-dependent LNCaP and LAPC-4 prostate cancer cells with Akt inhibitor. Akt inhibition decreased AR expression, suggesting that Akt activity was required for regulation of AR protein levels. However, while androgen-independent LNCaP-abl cells also showed diminished AR protein levels in response to Akt inhibition, treatment of androgen-independent LNCaP-AI cells failed to alter AR protein levels upon similar treatment, suggesting that AR protein levels in these androgen-independent prostate cells were regulated by mechanisms independent of Akt activation. Regulation of AR, downstream of activated Akt, also was observed in vivo when examining transgenic mice that overexpress constitutively active mutant myristoylated (myr)- Akt1 in the prostate. Transgenic mice expressing activated myr-Akt1 exhibited higher levels of AR mRNA and protein. Expression of activated myr-Akt1 did not alter prostate cell growth and no significant size differences between prostate tissues derived from transgenic animals were observed when comparing transgenic mice with wild-type mice. Still, transgenic mice overexpressing Akt exhibited higher levels of gH2AX and phosphorylated Chk2 in prostate tissue. These changes in markers associated with -induced senescence confirmed significant altered signaling in the transgenic mouse model. Overall, results presented here suggest that AR levels are regulated by the Akt pathway. Endocrine-Related Cancer (2011) 18 245–255

Introduction homolog (PTEN) and activation of Akt, which are The androgen receptor (AR) directs prostate growth strongly correlated with prostate cancer (Di Cristofano and differentiation and, for this reason, anti-androgens et al. 1998, Wang et al. 2003). Synergistic interactions are commonly used to treat prostate cancer. The between AR and Akt in an in vivo prostate regeneration importance of understanding the mechanism of AR model (Xin et al. 2006) provide evidence that the and protein regulation is underscored by the phosphoinositide 3-kinase (PI3K)/PTEN/Akt and AR finding that prostate cancer is reliant on the expression pathways may be linked mechanistically. of AR even after progressing to anti-androgen-resistant It has been previously reported that overexpression disease and increased expression of the AR is the major of myristoylated Akt in prostate results in prostate factor driving prostate cancer recurrence (Chen et al. intraepithelial neoplasia (PIN; Majumder et al. 2003). 2004). Other factors contribute to disease progression, In addition, the PIN lesions occurring in Akt-over- notably, loss of function of phosphatase and tensin expressing transgenic animals invoked an increase in

Endocrine-Related Cancer (2011) 18 245–255 Downloaded from Bioscientifica.comDOI: 10.1530/ERC-10-0204 at 09/27/2021 02:56:54AM 1351–0088/11/018–245 q 2011 Society for Endocrinology Printed in Great Britain Online version via http://www.endocrinology-journals.orgvia free access S Ha, R Ruoff, N Kahoud et al.: Akt regulates AR protein and mRNA levels cellular levels of p27/kip1, resulting in cellular from long-term culture in 10% cFBS. (Culig et al. senescence (Majumder et al. 2008); this is consistent 1999, Gao et al. 1999). R1881 (methyltrienolone) was with reports that cellular senescence is often seen in purchased from Perkin Elmer (Waltham, MA, USA). early or pre-invasive stages of cancer (Chen et al. Akt inhibitor VIII and PI3K inhibitor LY294002 were 2005, Collado et al. 2005, Michaloglou et al. 2005). To purchased from EMD Chemicals (Gibbstown, NJ, explore the link between PI3K/PTEN/Akt and AR USA) and reconstituted in dimethyl sulfoxide (DMSO). pathways, we examined the impact of Akt activity on VCaP cells were transiently transfected using AR protein levels in cultured prostate cells and in a Lipofectamine 2000 (Invitrogen) with vector only or transgenic mouse model. Our findings indicate that AR with pCMV6-myr-Akt1-HA, according to the manu- expression is regulated by Akt in both models, but can facturer’s recommendations. Twenty-four hours be Akt-dependent or Akt-independent in androgen- post-transfection, cells were washed with Dulbecco’s independent cell lines depending on their individual PBS and treated overnight with low serum and characteristics. androgen-free media (0.05% (v/v) cFBS), followed by treatment with R1881 for 2 h. Cells were lyzed as described below 48 h post-transfection. Materials and methods Generation of transgenic lines expressing Hematoxylin and eosin staining active Akt Prostate tissues were dissected and fixed in formalin The ARR2PB rat probasin promoter (Zhang et al. overnight at 4 8C, dehydrated and embedded in 2000), a SV40 t-intron and poly A sequence were paraffin. Five-micron sections were cut, placed on separately subcloned into pBluescript II SK. The charged slides and baked for 1.5 h at 60 8C. Samples minimal ARR2 probasin promoter is a composite of were deparaffinized in xylene and rehydrated two androgen response regions (ARR) of the probasin in ethanol and water. Sections were stained in 0 promoter (ARR1 (K244 to K80) placed 5 to ARR2 non-aqueous hematoxylin and eosin. Samples were (K286 to C24)). A cDNA insert encoding a dehydrated in ethanol and xylene, allowed to air dry, myristoylated mutant of mouse Akt1 (myr-Akt1-HA) mounted (Permount) and coverslipped. was liberated from pCMV6 by double digestion with HindIII and BamHI, blunted and ligated into the EcoRI Immunohistochemistry site of pBluescript vector between the probasin promoter and the SV40 poly-A sequence. Southern Formalin-fixed, paraffin-embedded tissue sections blot analysis of tail DNA derived from transgenic mice were dewaxed in xylene, rehydrated and washed in identified three founder animals from the probasin- PBS, pH 7.4. For antigen retrieval, paraffin sections driven Akt construct (ARR2-myr-Akt). Mice were were heated in a microwave oven (900 W) in 1! citrate maintained in accordance with the guidelines for the buffer, followed by treatment with 3% (v/v) H2O2 and Care and Use of Laboratory Animals. blocked with 20% (v/v) normal goat serum. Sections were incubated with phospho-Chk2 Thr68 (1:50) and Cell culture, transfection, and reagents g-H2AX Ser139 (1:50; Cell Signaling, Danvers, MA, USA), followed by incubation with biotinylated rabbit LNCaP and VCaP cell lines were obtained from the secondary antibody (1:1000 in 2% (w/v) BSA; Vector ATCC (Manassas, VA, USA) and cultured in Labs, Burlingame, CA, USA) and streptavidin per- RPMI-1640 and DMEM, respectively. Media was oxidase. Staining was visualized with 3,30-diamino- supplemented with 10% (v/v) FBS and 1% (v/v) benzidine (DAB) (Vector Labs) and sections were penicillin/streptomycin. LNCaP cells used in this study counterstained with Hematoxylin (Vector Labs) prior to were from passages 23 to 30. LAPC-4 cells (gift from mounting with Crystal/Mount (Electron Microscopy R Reiter, UCLA) and androgen-independent LNCaP-AI Sciences; Hatfield, PA, USA). Microscopy was per- cells (gift from A Ferrari, NYU) were cultured in formed using a Zeiss Axio Imager A2 microscope. IMDM and RPMI-1640, respectively. LNCaP-AI Images were acquired with QCapture Prosoftware. culture medium was supplemented with 10% (v/v) charcoal-stripped FBS (cFBS) and 5 mg/ml insulin (Sigma). Androgen-independent LNCaP-abl cells (gift Protein extraction and western blot analysis from Z Culig) were maintained in RPMI-1640 Whole prostates were dissected from age-matched supplemented with 10% cFBS. Androgen-independent C57BL/6 wild-type (WT; Taconic, Hudson, NY, USA) LNCaP-AI and LNCaP-abl cell lines were both derived and transgenic ARR2-myr-Akt mice. A Tissue Tearor

Downloaded from Bioscientifica.com at 09/27/2021 02:56:54AM via free access 246 www.endocrinology-journals.org Endocrine-Related Cancer (2011) 18 245–255

(BioSpec Products, Bartlesville, OK, USA) was used to A LNCaP homogenize the tissues. Proteins were extracted using AR TNN lysis buffer (50 mmol/l Tris, pH 7.4, 150 mmol/l NaCl, 0.5% (v/v) NP-40, 5 mmol/l EDTA, 5 mmol/l P-Akt S473 NaF, 1 mmol/l dithiothreitol, 5 mg/ml aprotinin, Akt 5 mg/ml leupeptin, 1 mmol/l phenylmethylsulphonyl ERK-1 fluoride (PMSF), 1 mmol/l sodium orthovanadate). ––++10 nM R1881 Lysates were incubated on ice for 30 min, then –+– +20 µM Akti centrifuged and extracted protein was collected. To extract protein from cell lines, cells were lyzed 1 0.45 0.98 0.83 with Triton X-100 lysis buffer (50 mmol/l HEPES, pH B LAPC-4 7.5, 150 mmol/l NaCl, 1 mmol/l EDTA, 1 mmol/l AR EGTA, 10% (v/v) glycerol, 1% (v/v) Triton X-100, 25 mmol/l NaF, 10 mmol/l ZnCl2,10mg/ml aprotinin, P-Akt S473 10 mg/ml leupeptin, 1 mmol/l PMSF, 1 mmol/l sodium orthovanadate). Akt Protein lysates were quantified using Bio-Rad – + + + 10 nM R1881 Protein Assay Dye and normalized for total protein Veh LY Akti concentrations. Total protein was subjected to SDS- 1 1.2 0.62 0.42 PAGE and immunoblotted with the following antibodies: phospho-AR S213 (Taneja et al. 2005), C LNCaP androgen-independent AR (441) and extracellular regulated kinase 1 (ERK-1) AR (K-23; Santa Cruz Biotechnology; Santa Cruz, CA, P-Akt S473 USA); phospho-Akt S473 and Akt (Total; Cell Signaling); Tubulin, HA-epitope tag, and keratin 14 Akt (Covance, Denver, PA, USA); and HA conjugated to ERK-1 HRP (Roche). Protein expression levels were quanti- 10 nM R1881 ––+ + fied using ImageJ 1.42q software (National Institutes 20 µM Akti –+– + of Health, Bethesda, MD, USA). 1 1.06 1.04 1.09

D LNCaP-abl Quantitative RT-PCR AR Total RNA was extracted from mouse prostates using RNeasy (Qiagen), according to the manufacturer’s P-Akt S473 recommendations. RNA (1 mg) was reverse transcribed Akt using Superscript III Reverse Transcriptase (Invi- trogen) and then 2.5% (v/v) of the resulting cDNA ERK-1 was used for quantitative PCR (qPCR). qPCR was ––++ performed using SYBR green Taq ready mix (Sigma) 10 nM R1881 –+– +20 µM Akti and a LightCyler (Roche). Data were analyzed by 1 0.32 0.87 0.67 the DDCT method (Bookout & Mangelsdorf 2003) using RPL19 or mouse keratin 14 as control Figure 1 Effect of Akt inhibition on prostate cancer cell lines. LNCaP androgen-dependent (A) and androgen-independent , then normalized to naive samples arbitrarily LNCaP-AI (C) and LNCaP-abl (D) cells were serum- and set to 1. The primers used for the qPCR are: mouse androgen-starved overnight in 0.05% cFBS before pretreat- AR F: 50-TACCAGCTCACCAAGCTCCT; mouse ment with 20 mM Akti or vehicle for 30 min. After pretreatment, 0 10 nM R1881 or vehicle was added for 2 h. (B) LAPC-4 cells AR R: 5 -GAACTGATGCAGCTCTCTTGC; RPL19 were serum- and androgen-starved overnight in 0.05% cFBS, F: 50-CACAAGCTGAAGGCAGACAA; RPL19 R: then pretreated with 50 mM LY 294002 (LY), 20 mM Akti or 50-GCGTGCTTCCTTGGTCTTAG; mouse keratin vehicle for 30 min. After pretreatment, 10 nM R1881 or vehicle 0 was added for 2 h. The vehicle for Akti and LY 294002 was 0.1 14 F: 5 -TCCCAATTCTCCTCATCCTC; mouse and 0.05% DMSO, respectively and for R1881, 0.1% ethanol. keratin 14 R: 50-GGTTGGTGGAGGTCACATCT; Total protein was extracted and subjected to western blot with mouse keratin 18 F: 50-CTTGCTGGAGGATGGA- antibodies against AR, phospho-Akt S473, Akt (total) and 0 ERK-1 (loading control). Quantitation of AR protein levels are GAAG; and mouse keratin 18 R: 5 -CTGCACAG shown in the boxed area. AR levels are normalized to the TTTGCATGGAGT. untreated lane arbitrarily set to 1.

Downloaded from Bioscientifica.com at 09/27/2021 02:56:54AM via free access www.endocrinology-journals.org 247 S Ha, R Ruoff, N Kahoud et al.: Akt regulates AR protein and mRNA levels

Results very low. In fact, when the cells were grown in 0.05% cFBS, similar to the experiments shown in Fig. 1,no Akt regulation of AR protein levels in prostate phosphorylated Akt was observed (Fig. 2A). In 10% cancer cells FBS, a small amount of P-Akt S473 was observed; To determine the impact of Akt activity on AR protein however, Akti did not decrease AR levels despite levels, we treated LNCaP (PTEN-deficient), LAPC-4 complete inhibition of phosphorylated Akt. Thus, (PTEN WT) and VCaP (PTEN WT) prostate cancer regulation of phospho-Akt appears tightly regulated cells with an inhibitor of Akt isoforms 1 and 2 (Akti). Figure 1A shows western blot analysis of lysates from A AR LNCaP cells treated with or without the synthetic P-Akt S473 androgen, R1881, in the presence or absence of Akt Akt inhibitor. The results indicate that Akti treatment completely abolished phosphorylation of Akt at S473 ERK-1 ––++–+10 nM R1881 (P-Akt S473), but did not affect total protein levels of –+–+––20 µM Akti Akt. Interestingly, inhibition of Akt activity by Akti 10% FBS 0.05% cFBS resulted in decreased AR protein levels compared with cells treated with vehicle alone. While this decrease B AR may be more apparent in the absence of R1881, both R1881-treated and -untreated cells showed diminished AR darker exp. AR in the presence of the Akt inhibitor (Fig. 1A, AR P-AR S213 panel, 2nd and 4th lane). This result was not specific to one cell type or due to the AR T877A mutation in HA LNCaP cells. LAPC-4 prostate cancer cells, which P-Akt S473 express WT AR, also showed diminished AR protein Akt levels following treatment with the PI3K inhibitor LY 294002 (LY) or Akti. Furthermore, the decrease in AR Akt darker exp. protein levels in the presence of the Akt inhibitor ERK-1 exceeded the effect that was observed after treatment 0 0.5110 0.5 HA-myr-Akt (µg) with LY 294002, which correlates a greater suppres- –––++ + 10 nM R1881 sion of phosphorylation of Akt S473 by Akti (Fig. 1B). In contrast, in the androgen-independent LNCaP C 4 subline (LNCaP-AI), Akti inhibited P-Akt S473 to the 0 µg myr-Akt same extent as in the androgen-dependent LNCaP cells 0.5 µg myr-Akt but did not decrease AR protein expression (Fig. 1C). 3 1 µg myr-Akt This suggests that, in androgen-dependent LNCaP and LAPC-4 cells, AR protein levels are regulated through 2 Akt and that this homeostasis is altered in the LNCaP- AI prostate cancer model. In another model of 1 androgen-independent prostate cancer, LNCaP-abl, Relative AR intensity which was derived in a comparable manner as 0 LNCaP-AI cells (Culig et al. 1999, Gao et al. 1999), 0 nM R1881 10 nM R1881 treatment with Akti decreased expression of AR Figure 2 VCaP response to Akt inhibition and expression of (Fig. 1D), similar to the parental androgen-dependent activated myr-Akt1. (A) VCaP cells were serum- and androgen- LNCaP cells. The different responses to Akt inhibition starved in 0.05% cFBS or treated in 10% FBS overnight. Cells in the androgen-independent models suggest that AR is were then pretreated with 20 mM Akti or vehicle for 30 min. After pretreatment, 10 nM R1881 or vehicle was added for 2 h. regulated by different mechanisms even though both (B) VCaP cells were transiently transfected with increasing LNCaP-AI and LNCaP-abl are capable of growing in amounts of myr-Akt1-HA. Cells were serum- and androgen- the absence of androgen. starved in 0.05% cFBS overnight, then treated with either vehicle or 10 nM R1881 for 2 h. Vehicles were the same as The relationship between Akt activity and AR described in Fig. 1. Total protein was extracted and subjected to expression was also examined in the androgen- western blot with antibodies against AR, phospho-AR S213, dependent VCaP prostate cancer cell line that phospho-Akt S473, Akt (total), HA and ERK-1 (loading control). (C) Quantitation of relative AR protein levels in VCaP cells expresses WT AR. These cells differ from LNCaP overexpressing myr-Akt. Data represents three independent and LAPC-4 cells in that basal levels of P-Akt S473 are experiments and error is represented by S.E.M.

Downloaded from Bioscientifica.com at 09/27/2021 02:56:54AM via free access 248 www.endocrinology-journals.org Endocrine-Related Cancer (2011) 18 245–255 in VCaP cells where serum withdrawal is sufficient to A Founders F1-progeny suppress Akt activity. However, while inhibition of the low level of endogenous Akt kinase activity did not myr-Akt WT affect AR protein levels in VCaP cells, overexpression *** of Akt resulted in increased levels of AR protein. Figure 2B shows that transient transfection of VCaP cells with myr-Akt1-HA resulted in a small, repro- ducible increase in AR protein in response to increasing amounts of overexpressed myr-Akt in both 2.5 the presence and absence of R1881 (compare AR panel, 0 mg vs 0.5 and 1 mg myr-Akt1-HA). There was at least a two-fold increase in AR protein expression 0.5 levels in the presence of overexpressed myr-Akt1-HA (Fig. 2C). 123456 1234 Phosphorylation of AR at serine 213, a putative B Kpn Kpn target of Akt, was also examined (Fig. 2B). Ligand- 0.5 2.5 dependent AR phosphorylation at serine 213 was previously shown to occur in prostate epithelial cells ARR1 ARR2 AKT PolyA in vivo (Taneja et al. 2005); however, overexpression PB of Akt resulted in little, if any, AR serine 213 C phosphorylation in VCaP cells. When comparing the 75 impact of Akti on AR levels in LNCaP and LAPC4 versus VCaP (Figs 1 and 2), we cannot rule out that myr-Akt 50 certain cells may be more susceptible to regulation of the AR pathway through Akt than others due to different genetic backgrounds of the cells. However, 37 given the very different levels of P-Akt S473 in LNCaP and LAPC4 cells versus VCaP, there may not be enough Akt activity in VCaP cells to impact AR levels under the experimental conditions. Alternatively, Akti, ERK-1 which is preferential for Akt isoforms 1 and 2 (DeFeo-Jones et al. 2005, Lindsley et al. 2005), may WT not inhibit all of the Akt3 isoform that is present in VCaP cells. VCaP cells express all three isoforms of Control Akt, whereas LNCaP and LAPC4 cells express only Akt1 and 2 (Koseoglu et al. 2007) and Akt3 was not ARR2-myr-Akt detected in either cell line (data not shown). Although Figure 3 Southern blot analysis of founder ARR2-myr-Akt1 Akt activity was assessed by examining the levels transgenic mice. (A) Southern blot of tail DNA from animals of P-Akt S473, it is possible that autophosphorylation microinjected with ARR2-myr-Akt1-HA. Founder animals that have integrated the ARR2-myr-Akt DNA are shown by the at T72 and S246 or other putative phosphorylation sites asterisks. The genomic DNA was cleaved with KpnI and the (Chen et al. 2001, Conus et al. 2002, Li et al. 2006) expected sizes of the DNA fragments are diagramed in (B). contribute to Akt’s effect on AR levels. Overall, Southern blot of F1 progeny (A, right panel) shows expression of ARR2-myr-Akt1 only in the transgenic animals but not in wild- inhibition of Akt in cells expressing constitutively type (WT) animals. (B) Diagram of the probasin promoter high levels of phospho-Akt results in decreased AR regulated, constitutively active Akt1 (myr-Akt1-HA) construct protein levels. used to generate the ARR2-myr-Akt1 transgenic mice. The ARR2 probasin promoter is a composite of two androgen response regions (ARR) of the probasin promoter (ARR1 (K244 to K96) placed 50 to ARR2 (K286 to C24)). DNA Myr-Akt1 expression in the prostate of fragment sizes in kb resulting from KpnI cleavage are indicated. transgenic animals The isolated DNA shown in 3B is used as a probe. (C) Protein lysates were made in TNN lysis buffer using ventral prostates The results presented above indicate that inhibition of of 3-month-old ARR2-myr-Akt1 animals and WT controls. Akt kinase activity resulted in decreased levels of AR Lysates immunoblotted for HA (Top panel) using HA antibody conjugated to HRP. The positive control is overexpressed protein, suggesting crosstalk between these two ARR2-myr-Akt-HA in 293 cells. The same membrane was then pathways that is consistent with published findings re-blotted (bottom panel) for ERK-1 as a loading control.

Downloaded from Bioscientifica.com at 09/27/2021 02:56:54AM via free access www.endocrinology-journals.org 249 S Ha, R Ruoff, N Kahoud et al.: Akt regulates AR protein and mRNA levels

(Xin et al. 2006). To determine if enhanced Akt A Set 1 Set 2 Set 3 Set 4 activity impacts AR protein levels in vivo,we generated transgenic mice that overexpress constitu- AR tively active myristoylated Akt1, specifically in the prostate. Following pro-nuclear injection of a construct AR (darker exp.) encoding the probasin ARR2 promoter (Zhang et al. Cytokeratin 14 2000), HA epitope-tagged, myristoylated mouse Akt1 and a SV40 poly A sequence, founder animals were Akt WTAkt WT Akt WT Akt WT identified by Southern blot analysis (Fig. 3A). Three 3.00 founders identified by the asterisks in lanes 1, 5 and 6 B (Fig. 3A, left panel) were backcrossed into the C57BL/6 parental strain. Representative samples from transgenic 2.00 F1 males are shown in Fig. 3A, right panel. Mice heterozygous for ARR2-myr-Akt were bred to generate 1.00 homozygous mice. Homozygosity for ARR2-myr-Akt was confirmed by Southern blot analysis, and these Relative AR mRNA mice have been used for studies described below. 0.00 WT Akt To verify expression of myr-Akt-HA protein, western blot analysis was performed using lysates Figure 4 Androgen receptor protein is upregulated in ARR2- from WT and transgenic animals (Fig. 3C). The results myr-Akt1 prostates compared with WT prostates. (A) Pooled indicate that, as expected, the myr-Akt1 transgene was whole prostates from age-matched (3.5 months) myr-Akt1 or WT animals immunoblotted for AR (top panel), a longer expressed in the ventral prostate of transgenic but not exposure (exp.) of the same AR membrane (middle panel) and WT animals. The expression of P-Akt S473 and Akt1 keratin 14 (bottom panel). (B) AR mRNA is upregulated in was also examined in transgenic and WT prostates. ARR2-myr-Akt1 prostates compared with WT prostates. mRNA w was extracted from age-matched (6 months) prostates from P-Akt S473 and Akt1 expression increased 40% in ARR2-myr-Akt1 and WT mice. Quantitative RT-PCR was transgenic mice (data not shown). performed measuring AR mRNA expression (arbitrary units). Samples are normalized to RPL19. Data are representative of three independent experiments and error bars represent S.D. Increased Akt activity results in elevated AR protein and mRNA levels AR transcripts were normalized to RPL19 (Fig. 4B). To determine the effect of increased Akt signaling on Normalization to epithelial cell markers keratin 14 or AR protein levels in vivo, AR levels were examined in 18 showed similar results with upregulation of AR age-matched (3.5 months) WT and transgenic animals mRNA in the ARR2-myr-Akt1 mice (data not shown). expressing myristoylated Akt under the regulation of the probasin promoter. Four separate matched sets of tissue lysates consisting of pools of three prostates Overexpression of activated Akt leads to from either WT or myr-Akt1 transgenic animals were upregulation of senescence markers but not immunoblotted for AR. The samples were also overt changes in cellular morphology immunoblotted for the basal epithelial cell marker As detailed above, transgenic myr-Akt1 mice express keratin 14 and tubulin (data not shown) as internal increased levels of AR, a circumstance associated loading controls. Figure 4A shows that AR protein with development of recurrent prostate cancer. To levels are markedly increased in the Akt transgenic determine if myr-Akt1 mice exhibited signs of compared with WT samples. A dark exposure of the hyperplasia, WT and transgenic mice were killed and AR immunoblot confirmed the presence of AR in WT examined for gross histological changes at 3.5, 6, 9 and mice (Fig. 4A, middle panel). Equivalent levels of 12 months. Prostates were dissected, fixed and paraffin- keratin 14 between the samples indicated comparable embedded for histological analysis. At necropsy, no amounts of epithelial cells in the protein lysates. observable tumors or changes in the mouse prostate Upregulation of AR protein in response to over- were noted. Further, no discernable morphological expressed myr-Akt1 in the transgenic animals correlated differences between ARR2-myr-Akt1 prostates and with upregulation of AR mRNA. RNA from prostates age-matched WT mouse prostates were evident, of age-matched (6 months) ARR2-myr-Akt1 and WT following hematoxylin and eosin staining and animals was examined using qRT-PCR. AR mRNA examination of prostate tissue sections (Fig. 5). Over- increased in transgenic animals compared with WT. expression of ARR2-myr-Akt1 did not affect prostate

Downloaded from Bioscientifica.com at 09/27/2021 02:56:54AM via free access 250 www.endocrinology-journals.org Endocrine-Related Cancer (2011) 18 245–255

Akt WT A B

3.5 months

C D

9 months

E F

12 months

Figure 5 H&E staining of ARR2-myr-Akt1 and WT prostates. Hematoxylin and eosin staining of ARR2-myr-Akt and WT ventral prostates dissected at 3.5 months (A and B), 9 months (C and D), and 12 months (E and F). Panels A–F are shown at 200! magnification. cell size or growth because there was no difference in against phospho-Chk2 and g-H2AX. Prostate tissue the weight or size of the prostate of the transgenic from ARR2-myr-Akt1 animals at all time points animals relative to WT mice. Comparable levels of exhibited more prevalent staining of nuclear phos- keratin 14 (Fig. 4A) suggest that there was no loss of pho-Chk2 and g-H2AX than that from WT animals, basal epithelial cells, consistent with the lack of a suggesting that expression of constitutively active myr- tumorigenic phenotype in the myr-Akt1 animals. Akt1 activated DNA damage response and senescence- The fact that ARR2-myr-Akt1 did not have an inducing pathways even in the absence of any impact on prostate cell growth or cause tumorigenesis histological manifestations of PIN. led us to hypothesize that overexpression of myr-Akt1 induced oncogene-associated stress, leading to cellular senescence in the adult prostate. Recent studies suggest Discussion a biological block to tumorigenesis inhibits the Results presented in this report indicate that an increase progression of preneoplastic lesions to neoplasia in Akt kinase activity correlates with increased levels (Gorgoulis et al. 2005, Michaloglou et al. 2005). of AR protein. These findings are relevant to human Similar observations have been made in mouse prostate cancers, since many have increased Akt models in which oncogene-induced stress is found to activity, due to PTEN mutation, or enhanced growth be associated with signs of replication-induced stress factor receptor signaling. Interestingly, regulation of and leads to cellular senescence as indicated by AR via Akt appears to occur primarily at the level of increased levels of g-H2AX S139 and phospho-Chk2 gene transcription, since transgenic animals expressing (Bartkova et al. 2006). To determine if the ARR2-myr- constitutively active myr-Akt1 have increased levels of Akt1 mice exhibited signaling changes indicative of AR mRNA and protein. While we do not know the cellular senescence, we examined levels of g-H2AX mechanism of Akt-induced AR mRNA upregulation, and phospho-Chk2 Thr 68 in WT versus ARR2-myr- we speculate that this might occur through Akt Akt1 mice. Prostates dissected from 3.5 months activation of NF-kB. Recent findings indicate that (Fig. 6) and 6 and 9 months old mice (Supplementary NF-kB interacts with the 50 regulatory sequence of the Figure 1, see section on supplementary data given at AR gene to upregulate AR mRNA and protein levels the end of this article) were stained with antibodies (Dan et al. 2008). In addition, AR and NF-kB protein

Downloaded from Bioscientifica.com at 09/27/2021 02:56:54AM via free access www.endocrinology-journals.org 251 S Ha, R Ruoff, N Kahoud et al.: Akt regulates AR protein and mRNA levels

Akt WT A B

P-chk2 3.5 months

C D

γ H2AX 3.5 months

Figure 6 Senescence markers are upregulated in ARR2-myr-Akt1 prostates. Immunohistochemistry was performed on 3.5-month- old prostates from ARR2-myr-Akt1 and WT animals using phospho-Chk2 Thr68 (A and B) and gH2AX Ser139 (C and D) antibodies. Panels A–D are shown at 200! magnification. Insets show the same sample (boxed area) at 400! magnification and demonstrate enhanced nuclear staining in transgenic mice compared with WT controls. levels are strongly correlated in prostate cancer (Dan Previous studies have shown that the impact of Akt et al. 2008), supporting the idea that NF-kB may on AR differed in low-passage versus high-passage regulate AR expression during prostate cancer LNCaP cells and depended on the activation of progression. forkhead , FOXO3a. In low-passage Although expression of the different isoforms of Akt LNCaP cells (i.e. passage numbers !35), AR and is shown to correlate with cancerous lesions and prostate-specific antigen were shown to be upregulated clinical outcomes in prostate cancer (Le Page et al. due to FOXO3a activation after treatment with the 2006), the ARR2-myr-Akt1 transgenic mice described PI3K inhibitor LY294002 (Yang et al. 2005). In in this report did not display an obvious phenotype in addition, overexpression of constitutively active Akt contrast to previous reports showing that expression of (cAkt) in LNCaP cells at low passage numbers (!38) activated Akt in the murine prostate induces highly suppressed AR activity as assessed by MMTV- penetrant PIN in the ventral prostate (Majumder et al. luciferase and AR protein expression when compared 2003). It is unlikely that the difference is due to the with high passage numbers (O60), at which point genetic backgrounds, since other studies were also MMTV-luciferase and AR expression was enhanced in based on experiments in a C57BL/6 background response to overexpression of cAkt (Lin et al. 2003). (Xu et al. 2007) similar to that used in our study. Our While studies presented in this report did not study differs from others in the promoter used examine the impact of Akt on AR target gene (probasin K426 to C28 (Majumder et al. 2003) transcription, we use lower passage number LNCaP versus the ARR2 promoter containing two copies of cells (23–34) to show that an Akt inhibitor results in the enhancer used here) and the inclusion of a diminished AR expression, a result further supported polyadenylation sequence in our transgenic construct. by the observation that transgenic expression of Furthermore, it is likely that the significant increase in myristoylated Akt results in increased AR protein nuclear expression of g-H2AX and phospho-Chk2 in levels. We speculate that differences between studies our ARR2-myr-Akt1 animals is contributing to cellular (Yang et al. 2005) may be due to the use of an Akt- senescence, thus blocking tumorigenesis. Still, the specific inhibitor to suppress endogenous Akt activity most likely explanation for the observed phenotypic as opposed to the effect of overexpression of cAkt or variations between studies using similar transgenic inhibition of PI3K, upstream of Akt. Reduced mouse lines may be found in variations of myr-Akt1 expression of AR in response to Akt inhibition is expression levels due to the site of integration or the likely due to the diminished pro-survival signaling promoter used. (Manning & Cantley 2007), altered cell cycle

Downloaded from Bioscientifica.com at 09/27/2021 02:56:54AM via free access 252 www.endocrinology-journals.org Endocrine-Related Cancer (2011) 18 245–255 regulation (Liang & Slingerland 2003) or increased showing that phase II clinical trials of androgen- degradation of AR. Indeed, inhibition with independent or biochemically recurrent prostate cancer MG132 could partially rescue AR levels in the patients using the Akt inhibitor perifosine did not presence of Akti (data not shown). significantly improve clinical outcomes (Posadas et al. Phosphorylation-dependent degradation of AR has 2005, Chee et al. 2007). Thus, one might speculate that been reported in response to overexpression of cAkt the window of opportunity for the clinical use of Akt and resulted in phosphorylation-dependent AR inhibitors to treat prostate cancer may be restricted and degradation (Lin et al. 2002). While we observed that these agents may be useful to prevent progression ligand-dependent phosphorylation of AR S213 (com- of androgen-dependent disease to the anti-androgen- parable to S210) in human prostate tissue and LAPC4 resistant disease stage. cells, we did not observe this in LNCaP cells. In fact, when we previously overexpressed the LNCaP AR Supplementary data T877A mutant in 293 cells, we observed robust phosphorylation of S213 in WT AR, but greatly This is linked to the online version of the paper at http://dx. diminished phosphorylation of the mutant (Taneja doi.org/10.1530/ERC-10-0204. et al. 2005). However, we have not ruled out the possibility that S213 is constitutively phosphorylated Declaration of interest at low levels in LNCaP cells. The authors declare that there is no conflict of interest that Regulation of AR in the LNCaP-AI subline appears could be perceived as prejudicing the impartiality of the to be independent of Akt. Interestingly, the androgen- research reported. independent sublines of LNCaP (AI and abl) responded differently to Akt inhibition. These cell lines have differing characteristics that may impact androgen- Funding independent growth. Silencing of the cyclin-dependent This work was supported by the American Cancer Society kinase inhibitor p21WAF1 contributes to the androgen- and NIH R01CA112226 (S K Logan). independent phenotype of LNCaP-AI cells (Wang et al. 2001), whereas M-phase cell cycle genes such as Acknowledgements UBE2C are upregulated in LNCaP-abl cells (Wang et al. 2009). In addition, other authors have presented We thank members of the Logan laboratory (Paolo Mita, evidence of gross differences in AR protein and mRNA Julia Staverosky and Eric Schafler) for critical reading of the regulation in androgen-dependent versus -independent manuscript. We thank R Reiter (University of California, Los cells, the latter expressing more stable AR protein and Angeles) for providing LAPC4 cells, A Ferrari (New York University School of Medicine) for the LNCaP-AI cells and mRNA. For example, pulse chase experiments show Z Culig (Innsbruck Medical University, Austria) for the that AR protein is 2–4 times more stable in cells LNCaP-abl cells. derived from recurring prostate tumors than in LNCaP cells (Gregory et al. 2001). There are also differences in regulation of AR mRNA in androgen-dependent References versus -independent cells: AR transcription is Bartkova J, Rezaei N, Liontos M, Karakaidos P, Kletsas D, decreased in response to cytokines such as tumor Issaeva N, Vassiliou LV, Kolettas E, Niforou K, necrosis factor-a in LNCaP cells but not in androgen- Zoumpourlis VC et al. 2006 Oncogene-induced senes- independent cells (Ko et al. 2008). cence is part of the tumorigenesis barrier imposed by Conventional anti-androgen treatments inhibit the DNA damage checkpoints. Nature 444 633–637. (doi:10. activity of AR but activation of AR through other 1038/nature05268) signaling molecules such as Akt may still lead to Bookout AL & Mangelsdorf DJ 2003 Quantitative real-time disease progression. Multiple studies have shown a PCR protocol for analysis of signaling correlation between phosphorylated Akt and prostate pathways. Nuclear Receptor Signaling 1 e012. (doi:10. cancer progression and recurrence, making Akt an 1621/nrs.01012) Chee KG, Longmate J, Quinn DI, Chatta G, Pinski J, attractive therapeutic target. Unfortunately, our finding Twardowski P, Pan CX, Cambio A, Evans CP, Gandara that AR protein levels are not decreased in all DR et al. 2007 The AKT inhibitor perifosine in androgen-independent prostate cancer cells examined biochemically recurrent prostate cancer: a phase II suggests that the AR pathway would be fully intact California/Pittsburgh cancer consortium trial. Clinical even in the presence of Akt inhibitors in some late- Genitourinary Cancer 5 433–437. (doi:10.3816/CGC. stage prostate cancers. This is supported by studies 2007.n.031)

Downloaded from Bioscientifica.com at 09/27/2021 02:56:54AM via free access www.endocrinology-journals.org 253 S Ha, R Ruoff, N Kahoud et al.: Akt regulates AR protein and mRNA levels

Chen R, Kim O, Yang J, Sato K, Eisenmann KM, McCarthy J, Gregory CW, He B, Johnson RT, Ford OH, Mohler JL, Chen H & Qiu Y 2001 Regulation of Akt/PKB activation French FS & Wilson EM 2001 A mechanism for androgen by tyrosine phosphorylation. Journal of Biological receptor-mediated prostate cancer recurrence after Chemistry 276 31858–31862. (doi:10.1074/jbc. androgen deprivation therapy. Cancer Research 61 C100271200) 4315–4319. Chen CD, Welsbie DS, Tran C, Baek SH, Chen R, Vessella R, Ko S, Shi L, Kim S, Song CS & Chatterjee B 2008 Interplay Rosenfeld MG & Sawyers CL 2004 Molecular of nuclear factor-kappaB and B-myb in the negative determinants of resistance to antiandrogen therapy. Nature regulation of androgen receptor expression by tumor Medicine 10 33–39. (doi:10.1038/nm972) necrosis factor alpha. Molecular Endocrinology 22 Chen Z, Trotman LC, Shaffer D, Lin HK, Dotan ZA, Niki M, 273–286. (doi:10.1210/me.2007-0332) Koutcher JA, Scher HI, Ludwig T, Gerald W et al. 2005 Koseoglu S, Lu Z, Kumar C, Kirschmeier P & Zou J 2007 Crucial role of -dependent cellular senescence in AKT1, AKT2 and AKT3-dependent cell survival is cell suppression of Pten-deficient tumorigenesis. Nature 436 line-specific and knockdown of all three isoforms 725–730. (doi:10.1038/nature03918) selectively induces apoptosis in 20 human tumor cell Collado M, Gil J, Efeyan A, Guerra C, Schuhmacher AJ, lines. Cancer Biology & Therapy 6 755–762. (doi:10. Barradas M, Benguria A, Zaballos A, Flores JM, Barbacid 4161/cbt.6.5.3995) M et al. 2005 Tumour biology: senescence in premalig- Le Page C, Koumakpayi IH, Alam-Fahmy M, Mes-Masson nant tumours. Nature 436 642. (doi:10.1038/436642a) AM & Saad F 2006 Expression and localisation of Akt-1, Conus NM, Hannan KM, Cristiano BE, Hemmings BA & Akt-2 and Akt-3 correlate with clinical outcome of Pearson RB 2002 Direct identification of tyrosine 474 as a prostate cancer patients. British Journal of Cancer 94 regulatory phosphorylation site for the Akt protein kinase. 1906–1912. (doi:10.1038/sj.bjc.6603184) Journal of Biological Chemistry 277 38021–38028. Li X, Lu Y, Jin W, Liang K, Mills GB & Fan Z 2006 (doi:10.1074/jbc.M203387200) Autophosphorylation of Akt at threonine 72 and serine Culig Z, Hoffmann J, Erdel M, Eder IE, Hobisch A, Hittmair 246. A potential mechanism of regulation of Akt kinase A, Bartsch G, Utermann G, Schneider MR, Parczyk K activity. Journal of Biological Chemistry 281 et al. 1999 Switch from antagonist to agonist of the 13837–13843. (doi:10.1074/jbc.M602060200) androgen receptor bicalutamide is associated with Liang J & Slingerland JM 2003 Multiple roles of the prostate tumour progression in a new model system. PI3K/PKB (Akt) pathway in cell cycle progression. Cell British Journal of Cancer 81 242–251. (doi:10.1038/sj. Cycle 2 339–345. (doi:10.4161/cc.2.4.433) bjc.6690684) Lin HK, Wang L, Hu YC, Altuwaijri S & Chang C 2002 Dan HC, Cooper MJ, Cogswell PC, Duncan JA, Ting JP & Phosphorylation-dependent ubiquitylation and Baldwin AS 2008 Akt-dependent regulation of NF-kBis degradation of androgen receptor by Akt require Mdm2 controlled by mTOR and Raptor in association with IKK. E3 ligase. EMBO Journal 21 4037–4048. (doi:10.1093/ Genes and Development 22 1490–1500. (doi:10.1101/ emboj/cdf406) gad.1662308) Lin HK, Hu YC, Yang L, Altuwaijri S, Chen YT, Kang HY & DeFeo-Jones D, Barnett SF, Fu S, Hancock PJ, Haskell KM, Leander KR, McAvoy E, Robinson RG, Duggan ME, Chang C 2003 Suppression versus induction of androgen Lindsley CW et al. 2005 Tumor cell sensitization to apoptotic receptor functions by the phosphatidylinositol stimuli by selective inhibition of specific Akt/PKB family 3-kinase/Akt pathway in prostate cancer LNCaP cells members. Molecular Cancer Therapeutics 4 271–279. with different passage numbers. Journal of Biological Di Cristofano A, Pesce B, Cordon-Cardo C & Pandolfi PP Chemistry 278 50902–50907. (doi:10.1074/jbc. 1998 Pten is essential for embryonic development and M300676200) tumour suppression. Nature Genetics 19 348–355. Lindsley CW, Zhao Z, Leister WH, Robinson RG, Barnett SF, (doi:10.1038/1235) Defeo-Jones D, Jones RE, Hartman GD, Huff JR, Huber Gao M, Ossowski L & Ferrari AC 1999 Activation of Rb and HE et al. 2005 Allosteric Akt (PKB) inhibitors: discovery decline in androgen receptor protein precede retinoic and SAR of isozyme selective inhibitors. Bioorganic and acid-induced apoptosis in androgen-dependent LNCaP Medicinal Chemistry Letters 15 761–764. (doi:10.1016/j. cells and their androgen-independent derivative. bmcl.2004.11.011) Journal of Cellular Physiology 179 336–346. (doi:10. Majumder PK, Yeh JJ, George DJ, Febbo PG, Kum J, Xue Q, 1002/(SICI)1097-4652(199906)179:3!336::AID- Bikoff R, Ma H, Kantoff PW, Golub TR et al. 2003 JCP11O3.0.CO;2-Q) Prostate intraepithelial neoplasia induced by Gorgoulis VG, Vassiliou LV, Karakaidos P, Zacharatos P, prostate restricted Akt activation: the MPAKT Kotsinas A, Liloglou T, Venere M, Ditullio RA Jr, model. PNAS 100 7841–7846. (doi:10.1073/pnas. Kastrinakis NG, Levy B et al. 2005 Activation of the 1232229100) DNA damage checkpoint and genomic instability in Majumder PK, Grisanzio C, O’Connell F, Barry M, Brito JM, human precancerous lesions. Nature 434 907–913. Xu Q, Guney I, Berger R, Herman P, Bikoff R et al. 2008 (doi:10.1038/nature03485) A prostatic intraepithelial neoplasia-dependent p27 Kip1

Downloaded from Bioscientifica.com at 09/27/2021 02:56:54AM via free access 254 www.endocrinology-journals.org Endocrine-Related Cancer (2011) 18 245–255

checkpoint induces senescence and inhibits cell prolifer- receptor regulates a distinct transcription program in ation and cancer progression. Cancer Cell 14 146–155. androgen-independent prostate cancer. Cell 138 245–256. (doi:10.1016/j.ccr.2008.06.002) (doi:10.1016/j.cell.2009.04.056) Manning BD & Cantley LC 2007 AKT/PKB signaling: Xin L, Teitell MA, Lawson DA, Kwon A, Mellinghoff IK & navigating downstream. Cell 129 1261–1274. (doi:10. Witte ON 2006 Progression of prostate cancer by synergy 1016/j.cell.2007.06.009) of AKT with genotropic and nongenotropic actions of the Michaloglou C, Vredeveld LC, Soengas MS, Denoyelle C, androgen receptor. PNAS 103 7789–7794. (doi:10.1073/ Kuilman T, van der Horst CM, Majoor DM, Shay JW, pnas.0602567103) Mooi WJ & Peeper DS 2005 BRAFE600-associated Xu Q, Majumder PK, Ross K, Shim Y, Golub TR, Loda M & senescence-like cell cycle arrest of human naevi. Nature Sellers WR 2007 Identification of prostate cancer 436 720–724. (doi:10.1038/nature03890) modifier pathways using parental strain expression Posadas EM, Gulley J, Arlen PM, Trout A, Parnes HL, mapping. PNAS 104 17771–17776. (doi:10.1073/pnas. Wright J, Lee MJ, Chung EJ, Trepel JB, Sparreboom A 0708476104) et al. 2005 A phase II study of perifosine in androgen Yang L, Xie S, Jamaluddin MS, Altuwaijri S, Ni J, Kim E, independent prostate cancer. Cancer Biology & Therapy 4 Chen YT, Hu YC, Wang L, Chuang KH et al. 2005 1133–1137. (doi:10.4161/cbt.4.10.2064) Induction of androgen receptor expression by Taneja SS, Ha S, Swenson NK, Huang HY, Lee P, Melamed J, phosphatidylinositol 3-kinase/Akt downstream Shapiro E, Garabedian MJ & Logan SK 2005 Cell-specific substrate, FOXO3a, and their roles in apoptosis of regulation of androgen receptor phosphorylation in vivo. LNCaP prostate cancer cells. Journal of Biological Journal of Biological Chemistry 280 40916–40924. Chemistry 280 33558–33565. (doi:10.1074/jbc. (doi:10.1074/jbc.M508442200) M504461200) Wang LG, Ossowski L & Ferrari AC 2001 Overexpressed WAF1 Zhang J, Thomas TZ, Kasper S & Matusik RJ 2000 A small androgen receptor linked to p21 silencing may be responsible for androgen independence and resistance to composite probasin promoter confers high levels apoptosis of a prostate cancer cell line. Cancer Research of prostate- specific through 61 7544–7551. regulation by androgens and glucocorticoids in vitro Wang S, Gao J, Lei Q, Rozengurt N, Pritchard C, Jiao J, and in vivo. Endocrinology 141 4698–4710. (doi:10.1210/ Thomas GV, Li G, Roy-Burman P, Nelson PS et al. 2003 en.141.12.4698) Prostate-specific deletion of the murine Pten tumor suppressor gene leads to metastatic prostate cancer. Cancer Cell 4 209–221. (doi:10.1016/S1535- Received in final form 17 January 2011 6108(03)00215-0) Accepted 8 February 2011 Wang Q, Li W, Zhang Y, Yuan X, Xu K, Yu J, Chen Z, Made available online as an Accepted Preprint Beroukhim R, Wang H, Lupien M et al. 2009 Androgen 11 February 2011

Downloaded from Bioscientifica.com at 09/27/2021 02:56:54AM via free access www.endocrinology-journals.org 255