Aberrant ERG expression cooperates with loss of PTEN to promote cancer progression in the prostate

Abstract

Chromosomal translocations involving the ERG locus are frequent events in human prostate cancer pathogenesis; however, the biological role of aberrant ERG expression is controversial. Here we show that aberrant expression of ERG is a progression event in prostate tumorigenesis. We find that prostate cancer specimens containing the TMPRSS2-ERG rearrangement are significantly enriched for loss of the tumor suppressor PTEN. In concordance with these findings, transgenic overexpression of ERG in mouse prostate tissue promotes marked acceleration and progression of high-grade prostatic intraepithelial neoplasia (HGPIN) to prostatic adenocarcinoma in a Pten heterozygous background. In vitro overexpression of ERG promotes cell migration, a property necessary for tumorigenesis, without affecting proliferation. ADAMTS1 and CXCR4, two candidate genes strongly associated with cell migration, were upregulated in the presence of ERG overexpression. Thus, ERG has a distinct role in prostate cancer progression and cooperates with PTEN haploinsufficiency to promote progression of HGPIN to invasive adenocarcinoma.

Figure 1. Genetic and molecular alterations ERG and PTEN are frequent and concomitant events in human prostate cancer

FISH break apart probes to 3′ (red) and 5′ (green) ends of both ERG and TMPRSS2 were used. Microscopy (200X), showing hematoxylin and eosin staining of HGPIN (arrowhead) and prostate carcinoma juxtaposed, high grade PIN (arrowhead) with deletion of the 3′ TMPRSS2 region, and prostatic adenocarcinoma (insert) with deletion of the 3′ TMPRSS2 region, as seen by loss of the red signal in at least one of the red-green pairs (A). Bar graph shows the proportion of HGPIN cases by ERG status (A). Tumor tissue microarray demonstrating prostate cancer with normal PTEN staining and prostate cancer with weak/absent PTEN staining (B). FISH staining from consecutive sections of the same specimen demonstrating the ERG genetic rearrangement is shown. The table reports the frequency of PTEN status by ERG rearrangement status (B).

Figure 2. Prostate specific over-expression of ERG cooperates with Pten haploinsufficiency to promote prostate tumorigenesis

Locally advanced mouse (Pten^−/− p53^−/−) prostate cancer demonstrated increased expression of murine Erg at the transcript level (normalized to Hprt) compared to controls (A). A total of 12 Pten^+/−, 12 ERG transgenic mice, and 12 Pten^+/− ERG mice were phenotypically characterized. High (200X) power histologic demonstration of prostatic adenocarcinoma in Pten^+/− ERG mice compared to controls in mice 6 months of age (B). Immunohistochemistry for smooth muscle actin is performed demonstrating invasion (B). Phenotypic characterization of Pten^+/− ERG mice at 2 and 6 months of age (200X). The incidence of prostatic adenocarcinoma is reported for transgenic mice at mice 2, 4, and 6 months of age (B). A bar graph is shown demonstrating the percentage of mice at each time point with the histologic finding of prostatic adenocarcinoma. Immunohistochemistry for Ki67 staining in pre-neoplastic prostate glands (mice 4 months of age) demonstrated no increase in proliferation for transgenic ERG mice while Pten haploinsufficiency provides a proliferative advantage (C). A bar graph is shown demonstrating the percentage of ki67 positive staining cells per gland with mean and standard deviation reported for 3 representative prostate glands with p-value calculated based on Chi-square statistic (C). Immunohistochemistry of Pten revealed no differences in staining for pre-neoplastic prostate glands at 4 months of age and prostate in vivo transcript levels for Pten and AKT were similar for both genotypes (D).

Figure 3. ERG regulates cell migration

Western blot analysis determined expression of Flag-ERG construct and activation of P-AKT in stable BPH-1 cell lines infected with MSCV, MSCV-ERG, MSCV-AKT (AKT-1), and MSCV-AKT MSCV-ERG (A). While over-expression of a constitutively active AKT (AKT-1) promoted cell growth, no proliferative advantage was observed with ERG over-expression (B). A line graph is utilized to demonstrate cell proliferation with the mean and standard deviation represented from 3 experiments (B). A striking increase was observed in cell migration with over-expression of ERG compared to vector control in BPH-1 cells which was not augmented with over-expression of a constitutively active AKT (AKT-1) (C). Bar graphs demonstrate the fold change in the number of migrated cells per high powered field normalized to vector control, and the mean and standard deviation from 3 experiments were calculated with p-value calculated based on total number of cells migrating (C).

Figure 4. ERG directly regulates CXCR4 and ADAMTS1

Putative ETS binding sites were identified in the promoter regions of both human CXCR4 and ADAMTS1 genes (A). ChIP assays were performed demonstrating direct ERG binding to the promoter regions for both CXCR4 and ADAMTS1 (A). Repeat ChIP experiment for CXCR4 in PC3 cells over-expressing ERG (B). CXCR4 siRNA knock-down demonstrates a significant reduction in cell migration in PC3 cells over-expressing ERG (C). Further evaluation revealed that the mRNA expression of murine Cxcr4 was up-regulated in the prostate specimens of ERG transgenic mice (D) and Pten loss models of prostate tumorigenesis with up-regulation of ERG compared to controls (E). The bar graphs demonstrate the fold change in mRNA level following normalization to Hprt and our lowest value, and the mean and standard deviations from 3 experiments are shown.