Probes for PTEN

Abstract

Immunohistochemical staining for Phosphatase and Tensin Homolog (PTEN) does not have either an acceptable standard protocol or concordance of scoring between pathologists. Evaluation of PTEN mRNA with a unique and verified sequence probe may offer a realistic alternative providing a robust and reproducible protocol. In this study we have evaluated an in situ hybridization (ISH) protocol for PTEN mRNA using RNAScope technology and compared it with a standard protocol for PTEN immunohistochemistry (IHC). PTEN mRNA expression by ISH was consistently more sensitive than PTEN IHC with 56% of samples on a mixed tumour tissue microarray (TMA) showing high expressionby ISH compared to 42% by IHC. On a prostate TMA 49% of cases showed high expression by ISH compared to 43% by IHC. Variations in PTEN mRNA expression within malignant epithelium were quantifiable using image analysis on the prostate TMAs. Within tumours clear over expression of PTEN mRNA on malignant epithelium compared to benign epithelium was frequently observed and quantified. The use of Spot Studio software in the mixed tumour TMA allowed for clear demonstration of varying levels of PTEN mRNA between tumour samples by the mRNA methodology. This was evident by the quantifiable differences between distinct oropharyngeal tumours (upto 3 fold increase in average number of spots per cell between 2 cases). mRNA detection of PTEN or other biomarkers, for which optimal or standardized immunohistochemical techniques are not available, represents a means by which heterogeneity of expression within focal regions of tumour can be explored with more confidence.

Although a putative role for TGF beta (TGFB) signaling in the pathogenesis of human endometrial cancer has long been proposed, the precise function of TGFB signaling in the development and progression of endometrial cancer remains elusive. Depletion of PTEN in the mouse uterus causes endometrial cancer. To identify the potential role of TGFB signaling in endometrial cancer, we simultaneously deleted TGFB receptor 1 (Tgfbr1) and Pten in the mouse uterus using Cre-recombinase driven by the progesterone receptor (termed Ptend/d ; Tgfbr1d/d ). We found that Ptend/d ; Tgfbr1d/d mice developed severe endometrial lesions that progressed more rapidly compared with those resulting from conditional deletion of Pten alone, suggesting that TGFB signaling synergizes with PTEN to suppress endometrial cancer progression. Remarkably, the Ptend/d ; Tgfbr1d/d mice developed distant pulmonary metastases, leading to significantly reduced life span. The development of metastasis and accelerated tumor progression in Ptend/d ; Tgfbr1d/d mice are associated with increased production of pro-inflammatory chemokines, enhanced cancer cell motility evidenced by myometrial invasion and disruption, and altered tumor microenvironment characterized by recruitment of tumor-associated macrophages. Thus, conditional deletion of Tgfbr1 in PTEN-inactivated endometrium leads to a disease that recapitulates invasive and lethal human endometrial cancer. This mouse model may be valuable for preclinical testing of new cancer therapies, particularly those targeting metastasis, one of the hallmarks of cancer and a major cause of death in endometrial cancer patients.

Genetic alterations that potentiate PI3K signalling are frequent in prostate cancer, yet how different genetic drivers of the PI3K cascade contribute to prostate cancer is unclear. Here, we report PIK3CA mutation/amplification correlates with poor prostate cancer patient survival. To interrogate the requirement of different PI3K genetic drivers in prostate cancer, we employed a genetic approach to mutate Pik3ca in mouse prostate epithelium. We show Pik3caH1047R mutation causes p110α-dependent invasive prostate carcinoma in-vivo. Furthermore, we report PIK3CA mutation and PTEN loss co-exist in prostate cancer patients, and can cooperate in-vivo to accelerate disease progressionvia AKT-mTORC1/2 hyperactivation. Contrasting single mutants that slowly acquire castration-resistant prostate cancer (CRPC), concomitant Pik3ca mutation and Pten loss caused de-novo CRPC. Thus, Pik3ca mutation and Pten deletion are not functionally redundant. Our findings indicate that PIK3CA mutation is an attractive prognostic indicator for prostate cancer that may cooperate with PTEN loss to facilitate CRPC in patients.

The presence and extent of cribriform pattern of prostate cancer portend recurrence and cancer death. Therelative expressions within this morphology of the prognostically adverse loss of PTEN, and the downstream inactivation of cell cycle inhibitor p27/Kip1 had been uncertain. In this study, we examined 52 cases of cribriform cancer by immunohistochemistry (IHC) for PTEN, p27, and CD44 variant (v)7/8, and a subset of 17 casesby chromogenic in situ hybridization (ISH) using probe for PTEN or CDKN1B (gene for p27). The fractions of epithelial pixels positive by IHC and ISH were digitally assessed for benign acini, high grade prostatic intraepithelial neoplasia (PIN), and 8 morphological patterns of cancer. Immunostaining results demonstrated that: 1. PTEN loss was significant for fused small acini, cribriform-central cells, small cribriform acini, and Gleason grade 5 cells in comparison with other acini. 2. p27 loss was significant only for cribriform-peripheral cells; and borderline-significant for fused small acini in comparison with benign acini. 3. CD44v7/8 showed expression loss in cribriform-peripheral cells; other comparisons were not significant. ISH showed thatcribriform cancer had significant PTEN loss normalized to benign acini (P < .02), while Gleason 3 cancer or fused small acini did not. With CDKN1B, the degree of signal loss among various cancer morphologies was insignificant. In conclusion, molecular disparities emerged between the fused small acini and cribriform patterns of Gleason 4 cancer. PTEN or p27 loss as prognostic factors demand distinct assessment in the varieties of Gleason 4 cancer, and in the biphenotypic peripheral versus central populations in cribriform structures.

Abstract

BACKGROUND:

Loss or mutation of PTEN alleles at 10q23 in combination with 8q24 amplification (encompassing MYC) are common findings in aggressive, human prostate cancer. Our group recently developed a transgenic murine model of prostate cancer involving prostate-specific Pten deletion and forced expression of MYC under the control of the Hoxb13 promoter. MYC overexpression cooperated with Pten loss to recapitulate lethal, human prostate cancer.

METHOD:

We now report on the generation of two mouse prostate cancer cell lines, BMPC1 and BMPC2, derived from a lymph node, and liver metastasis, respectively.

RESULTS:

Both cell lines demonstrate a phenotype consistent with adenocarcinoma and grew under standard tissue culture conditions. Androgen receptor (AR) protein expression is minimal (BMPC1) or absent (BMPC2) consistent with AR loss observed in the BMPC mouse model of invasive adenocarcinoma. Growth in media containing charcoal-stripped serum resulted in an increase in AR mRNA in BMPC1 cells with no effect on protein expression, unless androgens were added, in which case AR protein was stabilized, and showed nuclear localization. AR expression in BMPC2 cells was not effected by growth media or treatment with androgens. Treatment with an anti-androgen/castration or androgen supplemented media did not affect in vitro or in vivo growth of either cell line, irrespective of nuclear AR detection.

DISCUSSION:

These cell lines are a novel model of androgen-insensitive prostatic adenocarcinoma driven by MYC over-expression and Pten loss.