Regulation of the pentose phosphate pathway by an androgen receptor-mTOR-mediated mechanism and its role in prostate cancer cell growth

Abstract

Cancer cells display an increased demand for glucose. Therefore, identifying the specific aspects of glucose metabolism that are involved in the pathogenesis of cancer may uncover novel therapeutic nodes. Recently, there has been a renewed interest in the role of the pentose phosphate pathway in cancer. This metabolic pathway is advantageous for rapidly growing cells because it provides nucleotide precursors and helps regenerate the reducing agent NADPH, which can contribute to reactive oxygen species (ROS) scavenging. Correspondingly, clinical data suggest glucose-6-phosphate dehydrogenase (G6PD), the rate-limiting enzyme of the pentose phosphate pathway, is upregulated in prostate cancer. We hypothesized that androgen receptor (AR) signaling, which plays an essential role in the disease, mediated prostate cancer cell growth in part by increasing flux through the pentose phosphate pathway. Here, we determined that G6PD, NADPH and ribose synthesis were all increased by AR signaling. Further, this process was necessary to modulate ROS levels. Pharmacological or molecular inhibition of G6PD abolished these effects and blocked androgen-mediated cell growth. Mechanistically, regulation of G6PD via AR in both hormone-sensitive and castration-resistant models of prostate cancer was abolished following rapamycin treatment, indicating that AR increased flux through the pentose phosphate pathway by the mammalian target of rapamycin (mTOR)-mediated upregulation of G6PD. Accordingly, in two separate mouse models of Pten deletion/elevated mTOR signaling, Pb-Cre;Pten(f/f) and K8-CreER(T2);Pten(f/f), G6PD levels correlated with prostate cancer progression in vivo. Importantly, G6PD levels remained high during progression to castration-resistant prostate cancer. Taken together, our data suggest that AR signaling can promote prostate cancer through the upregulation of G6PD and therefore, the flux of sugars through the pentose phosphate pathway. Hence, these findings support a vital role for other metabolic pathways (that is, not glycolysis) in prostate cancer cell growth and maintenance.

Figure 1

Inhibition of G6PD suppresses prostate cancer cell proliferation. (a) LNCaP cells were treated ± 100 nM 6-aminonicotinamide (6AN) ± the synthetic androgen R1881 for 7 days. Relative cell numbers were quantified after cell lysis by a fluorescent DNA-binding dye. (b) C4-2 cells, a CRPC-derivative of LNCaP cells, were treated ± 100 nM 6AN for 7 days. Relative cell numbers were quantified as in a. (c and d) LNCaP (c) and C4-2 (d) cells were transfected with siRNAs targeting scramble control (siCtrl) or G6PD (nos 1–3). After 72 h transfection, cells were harvested and subjected to immunoblot analysis using GAPDH as a loading control. (e and f) LNCaP (e) and C4-2 (f) cells were transfected with siRNAs as described in c and d, and then treated ± R1881 as indicated for 7 days. Relative cell numbers were quantified as in a. Representative results are expressed as mean relative cell number ±s.e. *, significant (P<0.05) changes from vehicle (no R1881; a and e) or control (b and f).

Figure 2

G6PD is required for maintaining NADPH levels and redox homeostasis. (a and b) LNCaP (a) or CRPC C4-2 (b) cells were treated ± 100 nM 6AN ± 10 nM R1881 as indicated for 3 days. Cells were lysed and NADPH was measured using an enzyme-recycling reaction and subsequent fluorescence measurement and normalized to total protein concentration as described in the Materials and methods. *, significant (P<0.05) changes from vehicle (no 6AN). (c and d) Using siRNA targeting scramble control (siCtrl) or the two most efficacious siRNAs targeting G6PD (nos 1 and 2), as determined in Figure 1, LNCaP (c) and C4-2 (d) cells were transfected with indicated siRNAs alone (d) or then treated for 3 days ± R1881 (c). NADPH levels were then measured as described in a and b. *, significant (P<0.05) changes from vehicle (no R1881; c or siCtrl, d). (e and f) To quantitate intracellular ROS levels, LNCaP (e) and C4-2 (f) cells were treated as described in a and b, respectively. After 3 days treatment, cells (∼300–1000 cells per group) were co-stained with Hoechst (blue) and MitoSOX Red (detects mitochondria-derived ROS) and fixed overnight. Cells were then imaged at × 20 the next day and images were analyzed using ImageJ software. Graphs show relative ROS levels per cell. *, significant (P<0.05) changes from vehicle (no 6AN). Below the graphs are representative images of the ROS staining.

Figure 3

G6PD regulates ribose synthesis in prostate cancer cell models. (a and b) LNCaP (a) and C4-2 (b) cells were treated for 3 days as indicated ± 100 nM 6AN ± 10 nM R1881 prior to starvation and treatment with ¹⁴C-labeled glucose for an additional 24 h. Total RNA extraction was performed and the fraction of newly radiolabeled RNA was quantitated via a scintillation counter and normalized to the total RNA pool. *, significant (P<0.05) changes from vehicle (no R1881; a) or vehicle (no 6AN; b). (c and d) LNCaP (c) and C4-2 (d) cells were initially transfected and treated as in Figures 2c and d. Cells were then treated for an additional 24 h with ¹⁴C-labeled glucose and subjected to the ¹⁴C RNA incorporation assay as described above. *, significant (P<0.05) changes from vehicle (no R1881; c). ^#, significant (P<0.05) changes from control (siCtrl; c). *, significant (P<0.05) changes from control (siCtrl; d).

Figure 4

AR signaling regulates G6PD levels. (a and b) Hormone-sensitive LNCaP (a) or LAPC4 (b) cells were treated with increasing concentrations of R1881 (0, 0.1, 1 and 10 nM) for 72 h. Cells were harvested and subjected to immunoblot analysis using GAPDH as a loading control. (c–f) CRPC C4-2 (c and e) and 22Rv1 (d and f) cells were transfected with siRNAs targeting control (siCtrl) or AR (nos 1 and 2). Note, siAR #1 targets both full-length AR and the constitutive active, C-terminal-truncated AR splice variants found in 22Rv1 cells while siAR #2 targets only full-length AR. (c and d) After 72 h transfection, cells were lysed and subjected to immunoblot analysis using GAPDH as a loading control. Representative images (top) and graphs representing relative band density normalized to loading control (bottom) are shown for each subfigure. (e and f) Cell growth was then quantitated after 7 days as described in Figure 1. *, significant (P<0.05) changes from control (siCtrl).

Figure 5

AR regulates G6PD levels via mTOR signaling. (a–d) LNCaP cells were treated ± 10 nM rapamycin (inhibitor of mTOR) ± 10 nM R1881 for 72 h. Cells were then subjected to immunoblot analysis (a), assessed for intracellular levels of NADPH (b), flux of ¹⁴C-labeled glucose into RNA (c) or relative cell numbers (d) as described above. *, significant (P<0.05) changes from vehicle (no R1881). (e and f) C4-2 cells were treated ± 10 nM rapamycin for 72 h and then subjected to immunoblot analysis (e) or assayed for cell growth (f). *, significant (P<0.05) changes from vehicle (no rapamycin).

Figure 6

G6PD expression increases during prostate cancer progression in vivo. (a) Paraffin-embedded prostate tissues from tumors derived from the dorsolateral lobes of 7-week-old Pb-Cre;Pten^f/f mice (Pb-Pten) or the dorsolateral lobe of 7-week-old C57BL/6 control mice (wt) were stained for G6PD (brown) and counterstained with hematoxylin (blue). (b) Immunohistochemical analysis of G6PD from the dorsolateral lobes of K8-CreER^T2;Pten^f/f inducible transgenic mice (K8-Pten) 1, 4 or 6 months after tamoxifen treatment in intact mice or mice that were castrated 6 months after initial tamoxifen treatment and then killed 2 months after castration. Hematoxylin was again used as a counterstain. White bars=50 μm.