Opposing effects of androgen deprivation and targeted therapy on prostate cancer prevention

Abstract

Prostate cancer is an ideal target for chemoprevention. To date, chemoprevention clinical trials with 5α-reductase inhibitors have yielded encouraging yet ultimately confounding results. Using a preclinical mouse model of high-grade prostatic intraepithelial neoplasia (HG-PIN) induced by PTEN loss, we observed unprecedented deteriorating effects of androgen deprivation, in which surgical castration or MDV3100 treatment accelerated disease progression of the otherwise stable HG-PIN to invasive castration-resistant prostate cancer (CRPC). As an alternative, targeting the phosphoinositide 3-kinase (PI3K) signaling pathway via either genetic ablation of genes encoding PI3K components or pharmacologic inhibition of the PI3K pathway reversed the PTEN loss-induced HG-PIN phenotype. Finally, concurrent inhibition of the PI3K and mitogen-activated protein kinase (MAPK) pathways was effective in blocking the growth of PTEN-null CRPC. Together, these data have revealed the potential adverse effects of antiandrogen chemoprevention in certain genetic contexts (such as PTEN loss) while showing the promise of targeted therapy in the clinical management of this complex and prevalent disease.

Significance: Chemoprevention with antiandrogen therapies is attractive for prostate cancer, given its prevalence and established hormonally mediated pathogenesis. However, because PTEN loss has been found in 9% to 45% of HG-PIN in the clinic, the current findings suggest that patients with PTEN-deficient prostate tumors might be better treated with PI3K-targeted therapies.

Fig. 1. Androgen deprivation potentiated the disease progression from HG-PIN to invasive CRPC

(a) Genetic ablation of PTEN in prostatic epithelium caused HG-PIN. IF: pAKT/SMA. (b) Surgical castration induced extensive apoptosis in HG-PIN lesions (left, IF: TUNEL), whereas a subpopulation of tumor cells continued to proliferate (right, IHC: anti-BrdU). (c) PTEN-null prostate tumor mass initially shrank in response to surgical castration but gradually grew back. (d) Androgen deprivation accelerated progression of PTEN-null HG-PIN to invasive CRPC, arrows indicating invasive lesions. Shown are representative lesions observed in 30/32 (93.75%) mice. IHC: anti-SMA. (e) AR staining in CRPC vs. castration naïve HG-PIN. IHC: anti-AR. (f) Western blot of p53 and AR in age-matched wide-type prostate (WT), HG-PIN and CRPC. (g) Chemical castration accelerated progression of PTEN-null HG-PIN to invasive CRPC, arrows indicating invasive lesions. Shown are representative lesions observed in 8/10 (80%) mice. IHC: anti-SMA. Mice harboring HG-PIN at 8 weeks of age were surgically or chemically castrated for another 16–18 weeks, representative data are shown in Fig. 1d, Fig. 1e, Fig. 1f and Fig. 1g. (h) A comparison between the clinical and preclinical trials over the time. High-grade cancer is seen in human trials, whereas invasive CRPC is evident in the preclinical mouse studies.

Fig. 2. Genetic ablation of PI3K components blocked the prostate tumorigenesis caused by PTEN loss

(a–d) In anterior prostates, genetic ablation of p110β, not p110α impaired PTEN loss-induced tumorigenesis. (f–i) In VP, genetic ablation of p110α or p110β failed to affect PTEN loss-induced tumorigenesis. (j) Concurrent ablation of both p110α and p110β restored the normal ductual structure of PTEN-null prostate in VP. Note the HG-PIN lesions labeled in arrow, and regular ductal structure in arrowhead featuring single epithelial layer, small amounts of stroma, abundant secreted proteins. Prostates shown are at age of 12 weeks.

Fig. 3. Pharmacological inhibition of PI3K pathway reversed the PTEN-controlled HG-PIN phenotype

(a) Scheme of BEZ235 drug treatment in PTEN-null HG-PIN (top panel), and histology of ventral prostates (lower panel). Green arrow indicates epithelium and blue arrow for stroma. In BEZ235 treatment group, note the single layer of luminal epithelial cells surrounded by a thin rim of fibromuscular stroma, loose connective tissue extending between individual ducts, and eosinophilic secretions in the gland lumens. (b) Weight of ventral prostates from Fig. 3a. (c) The effect of BEZ235 on cell proliferation in HG-PIN lesions. BrdU was administered at 1-hour post treatment and prostate tissues were harvested 5 hours later. (d) Quantification of apoptosis in HG-PIN lesions after 1 week of daily treatment via oral gavage. (e) Western blotting using HG-PIN prostate tissues harvested at 1-hour post treatment. (f) Histology of prostate tumors after BKM120-treatment, following the same scheme as in Fig. 3a. 12-week-old mice harboring HG-PIN were used above, unless otherwise indicated.

Fig. 4. Combined inhibition of PI3K and MAPK signaling suppressed the growth of CRPC

(a) Tissue western. Mice harboring HG-PIN at 8-weeks were surgically castrated and prostate tissues were harvested 16 weeks later. WT: wild-type prostate, HG-PIN: PTEN-null HG-PIN, and CRPC: PTEN-null CRPC. (b) Proliferation index of CRPC tumors post treatment. Mice harboring PTEN-null HG-PIN at 8 weeks of age underwent surgical castration for 16 weeks. BrdU was administered at 1-hour post drug treatment and prostate tissues were harvested 5 hours later. IHC: anti-BrdU. (c) The effects of drug treatment on the weight of CRPC prostate tissues. (d) Targeted therapy in the clinical management of prostate diseases with PTEN deficiency.