Glycogen synthesis correlates with androgen-dependent growth arrest in prostate cancer
© Schnier et al. 2005
Received: 18 October 2004
Accepted: 24 March 2005
Published: 24 March 2005
Skip to main content
© Schnier et al. 2005
Received: 18 October 2004
Accepted: 24 March 2005
Published: 24 March 2005
Androgen withdrawal in normal prostate or androgen-dependent prostate cancer is associated with the downregulation of several glycolytic enzymes and with reduced glucose uptake. Although glycogen metabolism is known to regulate the intracellular glucose level its involvement in androgen response has not been studied.
We investigated the effects of androgen on glycogen phosphorylase (GP), glycogen synthase (GS) and on glycogen accumulation in the androgen-receptor (AR) reconstituted PC3 cell line containing either an empty vector (PC3-AR-V) or vector with HPV-E7 (PC3-AR-E7) and the LNCaP cell line.
Androgen addition in PC3 cells expressing the AR mimics androgen ablation in androgen-dependent prostate cells. Incubation of PC3-AR-V or PC3-AR-E7 cells with the androgen R1881 induced G1 cell cycle arrest within 24 hours and resulted in a gradual cell number reduction over 5 days thereafter, which was accompanied by a 2 to 5 fold increase in glycogen content. 24 hours after androgen-treatment the level of Glucose-6-P (G-6-P) had increased threefold and after 48 hours the GS and GP activities increased twofold. Under this condition inhibition of glycogenolysis with the selective GP inhibitor CP-91149 enhanced the increase in glycogen content and further reduced the cell number. The androgen-dependent LNCaP cells that endogenously express AR responded to androgen withdrawal with growth arrest and increased glycogen content. CP-91149 further increased glycogen content and caused a reduction of cell number.
Increased glycogenesis is part of the androgen receptor-mediated cellular response and blockage of glycogenolysis by the GP inhibitor CP-91149 further increased glycogenesis. The combined use of a GP inhibitor with hormone therapy may increase the efficacy of hormone treatment by decreasing the survival of prostate cancer cells and thereby reducing the chance of cancer recurrence.
Androgen withdrawal leads to apoptosis of normal prostate cells and is the principal therapy to treat advanced prostate cancer [for a review, ]. Metabolic events known to be associated with androgen withdrawal are reduction in glucose uptake, downregulation of several glycolytic enzymes and of some key enzymes of the pentose-phosphate shunt [2–5]. Androgen withdrawal led to transcriptional downregulation of the pyruvate dehydrogenase E1 alpha (PDH E1α) gene in rat ventral prostate and in PC3 prostate cancer cells transiently transfected with the androgen receptor. Reduced transcription of PDH E1α is associated with a reduction of the glucose oxidative pathway . In contrast, androgen stimulated CO2 production derived from glucose . These results suggest that glucose transporters and several catabolic enzymes are regulated in an androgen-dependent manner.
Glycogen metabolism is regulated by intermediates of glycolysis, by covalent modification and by glycogen and purines. The two major enzymes GS and GP are controlled by phosphorylation and allosterically by effector molecules [7–9]. Glycogen synthase (GS) in its phosphorylated form is inactive but can be activated allosterically by G-6-P. This can facilitate the dephosphorylation by a glycogen-bound PP1-type phosphatase to the active form [10, 11]. Active GS is inactivated by phosphorylation by several important protein kinases: casein kinase II, calmodulin-dependent kinases, protein kinase A (PKA), protein kinase C (PKC) [12, 13]. Glycogen synthase kinase 3 (GSK-3), a major kinase inactivating GS, phosphorylates several sites on GS but only when GS has been phosphorylated at other sites . Partial dephosphorylation of a specific N- or C-terminal residue increases the sensitivity of GS to activation by G-6-P .
Glycogen phosphorylase (GP) also exists in two forms, the active phosphorylated a-form (GP-a) and the inactive b-form (GP-b). cAMP and calcium stimulate the activation of GP through PKA and phosphorylase (PHOS) kinase, which seems to be the only kinase phosphorylating GP . Muscle GP is allosterically activated by the binding of AMP, whereas G-6-P and glucose are allosteric inhibitors .
We have recently shown that the cyclin-dependent kinase inhibitor flavopiridol, which is in clinical trials as an anticancer agent, is also a potent GP inhibitor and binds to the purine-nucleotide inhibitor-binding site of GP [17, 18]. Inhibition of glycogen degradation by the specific GP inhibitor CP-91149 also growth inhibited cells that expressed high levels of brain GP but not cells expressing low levels of brain GP . CP-91149 binds at a site located at the subunit interface in the region of the central cavity of the dimeric structure and stabilizes the inactive form of GP [20–23], These observations raised the possibility that glycogen metabolism, and in particular brain GP, may be a potential target for anticancer therapy. Therefore, to understand the regulation and role of glycogen metabolism in prostate cancer in response to androgen we measured intracellular glycogen stores, the activities of GS and GP and G-6-P in prostate cancer cell lines. Our results indicate that glycogen accumulation and reduction in cell growth are associated with the androgen response of prostate cancer cells and can be further enhanced by GP inhibition with the GP inhibitor CP-91149. Thus androgen-dependent growth arrest and cell death can be further enhanced by GP inhibition.
The construction and characterization of PC3 cells reconstituted with the androgen receptor (AR) has been reported . For these experiments, PC3-AR cells were stably transfected with vector pZ16E67 BN containing the human papilloma virus E7 protein cDNA (PC3-AR-E72 and E73) or vector pZipNeoSV(X)1 alone (PC3-AR-V1 and V2). LNCaP cells were obtained from ATCC and experiments were performed with cells around passage 23. All cells were grown in RPMI 1640 lacking phenol red supplemented with charcoal-stripped 5% fetal bovine serum and containing penicillin (100 units/ml), and streptomycin (100 μg/ml). Stably transfected PC3-AR-E7 or V cells were maintained with 100 μg/ml hygromycin and 500 μg/ml G418. Cultures of LNCaP cells were supplemented with 4 nM of the stable testosterone derivative R1881.
Androgen response was induced by either adding 4 nM R1881 (PC3-AR-E7 or V) or by omitting androgen from the culture medium (LNCaP). PC3-AR-E7 or V cells were plated in a density of 5 × 105 cells per 100 mm dish and LNCaP in a density of 3 × 105 cells per 100 mm dish. Cells were kept under 95% air and 5% CO2 at 37°C in a humidified incubator.
GS assays were performed as described with some modifications . For the assay cells from one 100 mm plate were harvested, washed in PBS and lysed in lysis buffer (50 mM Tris/HCl (pH 7.5), 250 mM NaCl, 0.1% NP40, 5 mM EDTA, 50 mM NaF, 1 mM for phenyl-methyl-sulfonyl-fluoride, protease inhibitor cocktail (Boehringer-Mannheim)). Cell lysates were diluted equally with GS dilution buffer. The reaction was started by adding 25 μl assay buffer to 25 μl of the lysate. The assay buffer contained 5 mM UDP-glucose and 1 μCi/ml of [U-14C]UDP-glucose. 10 mM G-6-P was added to determine the activity of allosterically activated GS. The reaction was performed at 30°C for 15 min and stopped by pipetting the mixture on a p81 Whatman filter. The filters were immersed in 66% EtOH, washed several times in EtOH, once in acetone, air-dried and counted in a liquid scintillation counter (LKB Wallac). GP activity and glycogen content were determined as described with modifications [17, 27, 28]. The G-6-P concentration was determined in the following way. Nine 100 mm petri dishes were prepared for the isolation of G-6-P and cells collected, sedimented by centrifugation and 0.75 ml 6% perchloric acid added to the cell pellet. Samples were homogenized, iced for 10 min, vortexed and centrifuged at 6000 rpm for 15 min in the Sorvall at 4°C. 0.6 ml of supernatant was neutralized by adding 134 ul of 4 M KOH/0.8 M imidazole. A pH around 7 was confirmed with pH paper. Samples were then centrifuged for 20 min to precipitate the salts. The supernatant was stored at -80°C until assayed in a Hitachi Fluorometer. The reaction was started by adding G-6-P dehydrogenase from Leuconostroc mesenteroides and incubated at room temperature for 25 min. Samples were measured before the reaction and immediately afterwards. The excitation wavelength was 365 nm and the emission wavelength 435 nm.
Cells were fixed with 70% ethanol, DNA was stained with propidium iodide. The intensity of fluorescence was measured using a Becton Dickenson flow cytometer at 488 nm for excitation and at 650 nm for emission . The cell cycle profile was analyzed using Modifit's Sync Wizard (Verity Software Inc.).
Androgen-independent prostate cancer frequently acquires the loss of a functional retinoblastoma protein (pRb) . Loss of pRb leads to a reduction of androgen-dependent gene expression, which has been interpreted as a possible mechanism for the androgen-independent growth of these cells . Therefore, we tested whether inactivation of pRb reverses the sensitivity of PC3-AR cells to androgen with respect to glycogen accumulation. PC3-AR cells were constructed that expressed the HPV-E7 protein, which binds and inactivates pRb (PC3-AR-E7) [34, 35]. PC3-AR cells transfected with the empty vector (PC3-AR-V) were used as a control. As shown in Fig. 1A, PC3-AR-E7 cells show a similar 2 to 5 fold increase in the glycogen content as do the control cells lacking E7 expression. The two cell lines differ in their basal glycogen content, which is about 50% lower in PC3-AR-E7. At 120 hours we observed less glycogen accumulation in the PC3-AR-E7 cells compared to control.
In this paper we describe the effect of androgen on glycogen metabolism in different prostate cancers. We used PC3-AR-V and PC3-AR-E7 cells, which ectopically expresses AR and in case of PC3-AR-E7 the HPV-E7 protein. Additionally, LNCaP, which maintains a functional AR, was tested for glycogen content and its correlation to the number of cells. The PC3-AR-V and PC3-AR-E7 cell lines demonstrated a response to androgen leading to G1 arrest with a corresponding increase in the glycogen content (2 to 5 fold). PC3 cells lacking AR did not increase glycogen content in response to androgen. PC3 (AR-) cells demonstrate that intracellular glycogen content corresponds with the growth and/or survival of cells harbouring a functional AR. Intracellular stores of glycogen correlated inversely with the cell number: when cell numbers are low the glycogen content is high. This inverse relationship suggests that glycogenesis participates in growth arrest. However, glycogenesis is most likely not sufficient to induce ATP-dependent apoptosis, as the inhibition of glycogenolysis with the GP inhibitor CP-91149 does not induce cell death in these prostate cell lines. Glycogen content normalized to the number of cells is about 50% higher in PC3-AR cells treated with R1881 than treated with CP-91149. The additional increase in glycogen content of R1881-treated PC3-AR-V/E7 cells upon CP-91149 treatment further results in a reduction of cell number by growth inhibition and reduction in cell viability, which suggests that a certain intracellular glycogen content has to be reached to affect cell viability. Alternatively, certain effects of hormone treatment on cell survival may be enhanced by inhibition of glycogenolysis using CP-91149. Similarly, LNCaP cells responded with glycogen accumulation and growth arrest upon androgen removal, which was further enhanced by the GP inhibitor CP-91149.
One explanation of reduced cell viability in conjunction with glycogen content could be the increase in G-6-P in response to androgen. G-6-P is a key metabolite allosterically activating GS . The G-6-P content was about 3 fold higher 24 hours after androgen addition in PC3-AR cells compared to untreated cells. The increase in the G-6-P level was most probably a consequence of the decrease in activity of some key glycolytic enzymes as well as G-6-P dehydrogenase, which have been reported to be under androgen control . We can, however, not exclude the presence of another allosteric GS activator nor partial activation of GS, which would sensitize GS to allosteric activation. The increase in G-6-P could also explain why an increase in activated GP (GP-a) in PC3-AR cells upon androgen treatment did not prevent the increase in glycogen content, since G-6-P is a GP-a inhibitor [37, 38]. However, the fact that CP-91149 increased the glycogen content twofold more in androgen-treated PC3-AR cells shows that there was still GP-a activity. CP-91149 is known to cause GP-a dephosphorylation and partial GS-activation .
Glycogenesis is part of the androgen response in prostate cancer and further inhibition of glycogen phosphorylase by a specific inhibitor reduces the cell number suggesting that glycogenolysis contributes to cell survival. Thus inhibition of glycogenolysis in combination with hormone therapy may be a more effective treatment for advanced prostate cancer than hormone therapy alone.
Human Papilloma Virus E7
We thank Dr. Judith L. Treadway for CP-91149, and Dr. Hirokazu Inoue for plasmids expressing the E7 cDNA.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.