Glycogen synthase kinase-3β inhibition depletes the population of prostate cancer stem/progenitor-like cells and attenuates metastatic growth

Abstract

Cancer cells with stem or progenitor properties play a pivotal role in the initiation, recurrence and metastatic potential of solid tumors, including those of the human prostate. Cancer stem cells are generally more resistant to conventional therapies thus requiring the characterization of key pathways involved in the formation and/or maintenance of this malignant cellular subpopulation. To this end, we identified Glycogen Synthase Kinase-3β (GSK-3β) as a crucial kinase for the maintenance of prostate cancer stem/progenitor-like cells and pharmacologic inhibition of GSK-3β dramatically decreased the size of this cellular subpopulation. This was paralleled by impaired clonogenicity, decreased migratory potential and dramatic morphological changes. In line with our in vitro observations, treatment with a GSK-3β inhibitor leads to a complete loss of tumorigenicity and a decrease in metastatic potential in preclinical in vivo models. These observed anti-tumor effects appear to be largely Wnt-independent as simultaneous Wnt inhibition does not reverse the observed antitumor effects of GSK-3β blockage. We found that GSK-3β activity is linked to cytoskeletal protein F-actin and inhibition of GSK-3β leads to disturbance of F-actin polymerization. This may underlie the dramatic effects of GSK-3β inhibition on prostate cancer migration. Furthermore, GSK-3β inhibition led to strongly decreased expression of several integrin types including the cancer stem cell-associated α2β1 integrin. Taken together, our mechanistic observations highlight the importance of GSK-3β activity in prostate cancer stemness and may facilitate the development of novel therapy for advanced prostate cancer.

Figure 1. The effect of GSK-3β inhibition on Wnt signaling in prostate cancer cell lines

Wnt reporter activity in transiently transfected PC-3 and PC-3M-Pro4 cells after pretreatment with 30-100 μM PNU-74654 and/or 50-100 nM GIN for 24 hours. The figure comprises two independent experiments. * p < 0.05 versus control; *** p < 0.001 versus control; $ p < 0.05 versus GIN-treated.

Figure 2. The effect of GSK-3β inhibition on the stem/progenitor subpopulation, clonogenic and migratory potential

(a) The effect of treatment with 30-100 μM PNU-74654 and/or 50-100 nM GIN on (a) ALDH^HIGH subpopulation after 48 hours (% ALDH^HIGH subpopulation of untreated: PC-3, 4.5±1.0%; PC-3M-Pro4, 25.8±7.1%; C4-2B4, 19.6±2.9%; DU-145, 2.3±0.5%), (b) clonogenic potential after 7 days and (c) migratory capacity after 16 hours. All figures comprise two or three independent experiments. * p < 0.05 versus control; ** p < 0.01 versus control; *** p < 0.001 versus control; $S p < 0.01 versus GIN-treated.

Figure 3. The effect of GIN pretreatment in vitro on tumorigenic potential in vivo

PC-3M-Pro4luc2 cells were pretreated with vehicle or 100nM GIN for 48 hours prior to subcutaneous inoculation in nude mice (n=10 per group). (a) Experimental schedule. (b) Subcutaneous tumor burden. (c) Representative examples of bioluminescent images. * p < 0.05 versus control.

Figure 4. The effect of GIN pretreatment in vitro on metastatic potential in vivo

PC-3M-Pro4luc2 cells were pretreated with vehicle or 100nM Gin for 48 hours prior to inoculation in the left cardiac ventricle of nude mice (n=9 per group). (a) Experimental schedule. (b) Quantification of the total number of metastases and (c) total tumor burden. ** p < 0.01 versus control.

Figure 5. The effect of GSK-3β inhibition on F-actin polymerization and integrin expression

(a) PC-3 or PC-3M-Pro4 cells were treated with 100 μM PNU-74654 and/or 100 nM GIN for 48 hours. F-actin polymerization was visualized using Phalloidin (green) and the nucleus was visualized using DAPI (blue). (b) PC-3M-Pro4 cells were treated with 100nM GIN for 48 hours and integrin expression was monitored using flow cytometry. *** p < 0.01 versus control.