PTEN Deficiency Is Fully Penetrant for Prostate Adenocarcinoma in C57BL/6 Mice via mTOR-Dependent Growth

In Vivo Studies

Mice were housed in suspended polycarbonate cages or individually ventilated cages (Lab Products, Maywood, NJ) on autoclaved hardwood bedding (PJ Murphy Forest Products Corp., Montville, NJ) in an AAALAC-accredited facility (M. D. Anderson Cancer Center, Science Park–Research Division). Room conditions included temperature (20–22°C), humidity (60–70%), and light (14/10 hours; light/dark). Commercial rodent pelleted food (Harlan Teklad, Madison, WI) and autoclaved water were available ad libitum. All procedures were in compliance with the Public Health Service Guide for the Care and Use of Laboratory Animals (National Research Council, 1996). The protocol involving use of these mice was approved by the M. D. Anderson Cancer Center Institutional Animal Care and Use Committee. Male mice were euthanized at various ages from 7 to 12 months by CO₂ asphyxiation, and tissues were harvested and either snap-frozen in liquid N₂ and stored at −80°C or fixed in 10% neutral buffered formalin and paraffin embedded. The intact male reproductive system was transversely sectioned and then paraffin-embedded for histopathological and immunohistological analysis. The Pten^+/− mice were a kind gift from Dr. Ramone Parsons (Columbia University)20 and the Tsc2^+/− mice were obtained from Dr. David Kwiatkowski (Brigham and Women’s Hospital).38

Pten^+/− Mice Treated with Vehicle and Rapamycin

Pten^+/− mice 12 to 14 months old were treated with rapamycin (LC Laboratories, Woburn, MA) (0.15 mg/kg) (n = 4) or vehicle (n = 3) for 14 days daily (i.p) and sacrificed at the end of the study. The vehicle was Tween 80, polyethylene glycol, and ethanol). Tissues were harvested and fixed as described above.

Genetic Background Characterization of C57BL/6-Pten Mice

We performed a genetic background characterization of our C57BL/6.129S1/v-Pten congenic strain containing a Pten _targeted mutation (Pten^tm1Rps).20 For this, we selected 92 microsatellite markers (simple sequence length polymorphism) evenly distributed over all of the 19 autosomal chromosomes (genome scan) and polymorphic between C57BL/6J and 129S1/Sv inbred strains. The average marker spacing was 15 cM. D19Mit88 on chromosome 19 was the only marker showing 129S1/C57BL/6 heterozygosity for all Pten^+/− mice. This is expected since the Pten gene is localized in the same region on chromosome 19 as D19Mit88, and the _targeted allele is 129S1/Sv (W9.5 ES cells) in origin. In agreement with the number of backcross generations performed in our Pten null colony (>N6), the background strain characterization showed that 91 of 92 markers (98.9%) were homozygous C57BL/6.

Histological Analysis

Tissues were stained with hematoxylin and eosin, and prostates were examined microscopically by two study pathologists (J.B. and C.J.C.) blinded as to age and genotype of study animals. Hematoxylin and eosin sections were evaluated for precursor lesions identified as hyperplastic, low-grade PIN, or high-grade PIN and adenocarcinoma whose severity was described by Shapell et al.41 To detect downstream _targets of the PTEN signaling pathway, immunohistochemistry was performed on paraffin-embedded prostate tissue sections using primary antibodies against AKT (1:100; Santa Cruz no. sc-1619, Santa Cruz, CA), phospho-AKT (Ser 473) (1:50; Cell Signaling Technologies, Beverly, MA, no. 3787), S6 (1:50; Cell Signaling Technologies no. 2217), and phospho-S6 (S235/236) (1:50; Cell Signaling Technologies no. 2211). Ki-67 (1:50 Santa Cruz no.15402) was used to determine proliferative index in vehicle and rapamycin treated mice. Primary antibodies were detected with biotinylated secondary antibodies, including anti-goat IgG for AKT, and an Envision plus labeled polymer, anti-rabbit-horseradish peroxidase (Dako Laboratories, Carpinteria, CA), for phospho-AKT, S6 and phospho-S6. This was followed by peroxidase-conjugated avidin/biotin (Vectastain ABC Kit, Vector Laboratories, Burlingame, CA) and DAB substrate (Dako Laboratories). Intensity of immunohistochemistry staining was graded on a scale as follows: “-“ indicating no apparent staining, “+” indicating weak staining, “++” indicating moderate staining and “+++” indicating strong staining. The four point scale for grading immunohistochemistry was used as described previously.42,43

For detection of PTEN, after deparaffinization, sections were pretreated with microwave irradiation in 0.01 mol/L citrate buffer (pH 6.0), followed by blocking 20 minutes with 10% normal donkey serum. Sections were incubated with anti-PTEN antibody (Neomarkers, 1:50 dilution, Freemont, CA) overnight at 4°C. After washing with phosphate-buffered saline, the sections were incubated with the secondary fluorescein isothiocyanate-conjugated donkey anti-mouse IgG (1:200; Jackson ImmunoResearch Laboratories, West Grove, PA). The slides were examined using a Fluoroview laser confocal microscope (Olympus America, Melville, NY).

Western Analysis

Normal and tumorigenic anterior prostate tissue lysates were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to polyvinylidene difluoride membranes. Membranes were blocked with 5% nonfat dry milk and separately incubated with phospho-AKT (Thr 308) (1:1000, Cell Signaling Technologies), phospho-AKT (Ser 473) AKT (1:1000, Cell Signaling Technologies), (1:1000, Cell Signaling Technologies), phospho-S6 (1:2000, Cell Signaling Technologies), S6 (1:1000, Cell Signaling Technologies), phospho-TSC2 (Ser 939) (1:1000, Cell Signaling Technologies), and TSC2 (1:1000, Epitomics; Burlingame, CA) for 2 hours, followed by streptavidin horseradish peroxidase-conjugated goat anti-rabbit secondary antibody for one hour at room temperature. Lumiglo (KPL) was used for detection on X-ray film (BioMax, Eastman Kodak, Rochester, NY). As a loading control, blots were stripped and reprobed with an antibody to γ-tubulin (1:5000; Sigma, St. Louis, MO).

C57BL/6-Pten^+/− Mice Develop Prostate Adenocarcinoma

In-house colonies of Pten^+/−20 and Tsc2^+/−38 mice maintained on a C57BL/6 genetic background and confirmed by microsatellite analysis to be homozygous C57BL/6 (see Materials and Methods) were used to generate double heterozygous mice (Pten^+/−; Tsc2^+/−) for analysis. Prostate lesions were evaluated in wild-type, Pten^+/− single heterozygotes, Pten^+/−; Tsc2^+/− double heterozygotes and Tsc2^+/− single heterozygotes with groups of male mice examined at 7–9 and 10–12 months of age. Whole prostates were removed from mice and processed using conventional histology techniques. At necropsy, grossly observable prostate lesions were often noted, in one case in a moribund animal at 6 months of age. As shown in Supplemental Figure 1 (found at http://ajp.amjpathol.org), enlargement and changes in the coloration and texture of the anterior prostrate were the most common observations in these animals. Large, solid masses were also observed, often adjacent to the bladder, which were later confirmed to be tumors arising from the ventral prostate. Although at a lower frequency, grossly observable lesions were also identified in the dorsolateral prostate and seminal vesicle.

Histological analysis confirmed the neoplastic nature of grossly observable lesions and allowed the identification of a variety of microscopic lesions in all three lobes of the prostate. The adenocarcinomas observed in both the 7–9 and 10–12 month groups were moderately to well differentiated, forming in some cases large neoplastic masses with ill-defined glandular structures. Tumors with neuroendocrine differentiation or poorly differentiated histology were not observed.

As shown in Table 1 and Figure 1, adenocarcinomas were observed in the anterior prostates of 100% of the Pten^+/− male mice by 12 months of age, whereas no wild-type or Tsc2^+/− mice developed these lesions. While complete penetrance for adenocarcinoma was observed in the anterior prostate of Pten^+/− mice, adenocarcinomas also arose in the ventral and dorsolateral prostates of 33% and 67% of heterozygous mice, respectively, and in seminal vesicles of 14% of Pten^+/− and 33% of Pten^+/−;Tsc2^+/− mice (Figure 1, A–F). Seminal vesicles with invasive adenocarcinoma and intraepithelial neoplasia exhibited similar features and cytological characteristics as lesions arising in other lobes of the prostate (Figure 1, G–J). Double heterozygous Pten^+/−;Tsc2^+/− animals failed to exhibit any increased incidence of adenocarcinoma relative to Pten^+/− mice. Thus, in the anterior prostate the Pten null allele was fully penetrant for development of prostate adenocarcinoma, and the presence of a Tsc2 null allele did not accelerate lesion development throughout the prostate.

Prostate Cancer Progression in C57BL/6-Pten^+/− Mice

Precursor lesions for prostate adenocarcinoma were also evaluated microscopically. As shown in Table 2 the anterior prostates of wild-type, Pten^+/− and Pten^+/−;Tsc2^+/−, and Tsc2^+/− mice presented with hyperplasia, characterized by increased number of acini and increased epithelial tufting (Figure 1). In addition, a significant number of animals carrying the Pten-null allele showed atypical glandular structures with epithelial stratification and a cribriform pattern, essentially identical to the PIN lesions described previously in Pten^+/− mice and other murine models of prostate cancer. These lesions were then further classified as low grade or high-grade PIN (Figure 1, D–F) according to the severity as described by Shapell et al (Bar Harbor Consensus Report).41

As shown in Table 2, hyperplasias were noted in wild-type and in Tsc2^+/− mice, but progression to PIN lesions was not observed in these animals. Pten^+/− mice also developed hyperplasia by 7–9 months, but in contrast to wild-type and Tsc2^+/− animals, these lesions progressed from hyperplasia to more advanced lesions, which also occurred in Pten^+/−; Tsc2^+/− mice. As shown in Figure 2A–F, in both Pten^+/− and Pten^+/−; Tsc2^+/− mice, other than differing in their anatomical location, the histology of hyperplasias and PIN lesions observed in the dorsolateral prostate, ventral prostate and intraepithelial neoplasia in seminal vesicles were similar to the cognate lesions observed in the anterior prostate.

Loss of Pten Function in Prostate Adenocarcinomas

The PTEN tumor suppressor protein functions to inhibit PI3K/AKT activation. On loss of PTEN function, AKT becomes activated, phosphorylating and inactivating TSC2 to activate other signaling pathways such the mTOR pathway and its effectors, including ribosomal S6. Thus, activation of AKT is a reliable indicator of loss of PTEN function in the prostate.26 Anterior prostates obtained from animals histologically confirmed to have adenocarcinoma were used to determine AKT activation and mTOR signaling by Western analysis. As shown in Figure 3 and Supplemental Table 1 (found at http://ajp.amjpathol.org), all prostates from Pten^+/− and Pten^+/−;Tsc2^+/− mice exhibited elevated Akt activity, as assessed by phosphorylation of Akt at S473 and T308, and/or phosphorylation of S6 and Tsc2. In approximately 50% of the anterior prostate lesions examined, Pten expression was absent or barely detectable (summarized in Supplemental Table 1 at http://ajp.amjpathol.org). Importantly, Tsc2 expression was retained in both Pten^+/− and Pten^+/−; Tsc2^+/− double heterozygotes, and consistent with loss of PTEN and activation of AKT, was phosphorylated at S939 (Figure 3), indicating functional inactivation of Tsc2 occurred in these lesions rather than loss of heterozygosity.

Because prostates contain both tumor and normal tissues, to determine whether the wild-type Pten allele was lost in prostate adenocarcinomas, we isolated macroscopic tumors from Pten^+/− mice and Western analysis was performed to assess Pten expression in these lesions. In addition, we performed immunohistochemistry on anterior prostates from Pten^+/− mice confirmed to have microscopic adenocarcinomas. Pten expression was lost or barely detectable in adenocarcinomas from Pten^+/− mice (Figure 4a and Figure 5), consistent with elevated S6 phosphorylation observed in these tumors (Figure 4b) and activation of AKT signaling observed in the prostates of Pten^+/− animals (Figure 6). Similarly, while normal epithelial cells in the anterior prostate exhibited strong immunoreactivity with an antibody directed against Pten, this was not observed in adenocarcinomas of Pten^+/− mice (Figure 5).

Progression of Prostate Lesions in Pten-Haploinsufficient Mice Correlates with Elevated AKT and mTOR Signaling

To explore when during the progression of prostate lesions loss of PTEN function and activation of AKT and mTOR signaling occurred, normal prostate and hyperplasias, PIN lesions and adenocarcinomas from Pten^+/− and Pten^+/−; Tsc2^+/− mice were stained for phospho-AKT and phospho-S6, markers for loss of PTEN function and mTOR activation, respectively. As shown in Figures 6 and 7, all adenocarcinomas stained positively for phospho-AKT (Ser 473) and phospho-S6 (Ser 235/236), whereas normal anterior prostate was negative for both phospho-AKT and phospho-S6 immunoreactivity. Hyperplasias were also negative for phospho-AKT and phospho-S6, indicating that development of these lesions was not associated with loss of PTEN function, which was also consistent with the presence of hyperplasias as background lesions in this strain. However, PIN lesions that developed in both Pten^+/− and Pten^+/−; Tsc2^+/− mice contained detectable levels of phospho-AKT and phospho-S6, indicating that loss of PTEN function had occurred during this stage of progression. Interestingly in concordance with a previous study,40 in Pten^+/−; Tsc2^+/− mice, both high-grade PIN and adenocarcinoma had levels of phospho-AKT that were lower than comparable lesions arising in Pten^+/− mice. Thus heterozygosity at the Tsc2 locus appeared to have a dampening effect on AKT activation following loss of PTEN function.

Growth of Prostate Lesions Is mTOR-Dependent

To further understand the implications of elevated mTOR signaling in prostate lesions that develop as a result of loss of PTEN function, we treated Pten^+/− mice for 14 days with 0.15 mg/kg of rapamycin and stained the prostate for phospho-S6 to monitor mTOR signaling (Figure 8A). We observed that rapamycin given for even a short duration (14 days) was sufficient to down-regulate mTOR signaling as evidenced by reduction of S6 phosphorylation in PIN lesions and adenocarcinomas. Also, the total number of lesions was reduced in rapamycin-treated mice (Figure 8), although the number of animals examined were too small to evaluate statistically. However, we further investigated the functional significance of inhibition of mTOR signaling by examining cell proliferation (Ki-67 levels) in these lesions (Figure 8, B and C). Dramatic reductions in cell proliferation in response to rapamycin confirmed that these lesions were dependent on mTOR for growth.