Identification of genes that promote PI3K pathway activation and prostate tumour formation.

We have performed a functional in vivo mutagenesis screen to identify genes that, when altered, cooperate with a heterozygous Pten mutation to promote prostate tumour formation. Two genes, Bzw2 and Eif5a2, which have been implicated in the process of protein translation, were selected for further validation. Using prostate organoid models, we show that either Bzw2 downregulation or EIF5A2 overexpression leads to increased organoid size and in vivo prostate growth. We show that both genes impact the PI3K pathway and drive a sustained increase in phospho-AKT expression, with PTEN protein levels reduced in both models. Mechanistic studies reveal that EIF5A2 is directly implicated in PTEN protein translation. Analysis of patient datasets identified EIF5A2 amplifications in many types of human cancer, including the prostate. Human prostate cancer samples in two independent cohorts showed a correlation between increased levels of EIF5A2 and upregulation of a PI3K pathway gene signature. Consistent with this, organoids with high levels of EIF5A2 were sensitive to AKT inhibitors. Our study identified novel genes that promote prostate cancer formation through upregulation of the PI3K pathway, predicting a strategy to treat patients with genetic aberrations in these genes particularly relevant for EIF5A2 amplified tumours.

Comparison of different methodologies and cryostat versus paraffin sections for chromogenic immunohistochemistry.

Immunohistochemistry (IHC) specifically localizes proteins in cells and tissues, but methodologies vary widely. Therefore, we performed a methodological IHC optimization and validation study. First, we compared advantages and disadvantages of cryostat sections versus paraffin sections. Second, we compared and optimized antigen retrieval in paraffin sections using citrate buffer and Tris/EDTA buffer. Third, aminoethyl carbazole (AEC) and 3,3'-diaminobenzidine (DAB) were tested as horseradish peroxidase (HRP) substrates to obtain a water-insoluble coloured end product to visualize antigens. Fourth, secondary antibodies conjugated with either mono-HRP or poly-HRP were compared. The study was performed using serial sections of human tonsil. IHC was performed with primary antibodies against endothelial cell marker CD31, smooth muscle actin (SMA), chemokine stromal-derived factor-1α (SDF-1α) and its receptor C-X-C receptor type 4 (CXCR4), macrophage marker CD68 and proliferation marker Ki67. DAB rather than AEC, and cryostat sections rather than paraffin sections gave optimum staining at highest primary antibody dilutions, whereas tissue morphology in paraffin sections was superior. Loss of antigenicity in paraffin sections by formaldehyde fixation, heat and/or masking of epitopes was counteracted by antigen retrieval but not for all antigens. Two out of six antigens (CD31 and CD68) could not be retrieved irrespective time and type of retrieval. Tris-EDTA was superior to citrate buffer for antigen retrieval. The use of mono-HRP or poly-HRP depended on the affinity of the primary antibody for its antigen. We conclude that IHC methodology optimization and validation are crucial steps for each antibody and each research question.