Novel RNA hybridization method for the in situ detection of ETV1, ETV4, and ETV5 gene fusions in prostate cancer

Abstract

The genetic basis of 50% to 60% of prostate cancer (PCa) is attributable to rearrangements in E26 transformation-specific (ETS) (ERG, ETV1, ETV4, and ETV5), BRAF, and RAF1 genes and overexpression of SPINK1. The development and validation of reliable detection methods are warranted to classify various molecular subtypes of PCa for diagnostic and prognostic purposes. ETS gene rearrangements are typically detected by fluorescence in situ hybridization and reverse-transcription polymerase chain reaction methods. Recently, monoclonal antibodies against ERG have been developed that detect the truncated ERG protein in immunohistochemical assays where staining levels are strongly correlated with ERG rearrangement status by fluorescence in situ hybridization. However, specific antibodies for ETV1, ETV4, and ETV5 are unavailable, challenging their clinical use. We developed a novel RNA in situ hybridization-based assay for the in situ detection of ETV1, ETV4, and ETV5 in formalin-fixed paraffin-embedded tissues from prostate needle biopsies, prostatectomy, and metastatic PCa specimens using RNA probes. Further, with combined RNA in situ hybridization and immunohistochemistry we identified a rare subset of PCa with dual ETS gene rearrangements in collisions of independent tumor foci. The high specificity and sensitivity of RNA in situ hybridization provides an alternate method enabling bright-field in situ detection of ETS gene aberrations in routine clinically available PCa specimens.

Figure 1

Schematic diagram showing the genomic organization of A) ERG, B) ETV1, C) ETV4 and D) ETV5 genes. Horizontal black bar indicate the location of the probe in the gene and the numbers in the bar indicate the probe coverage region.

Figure 2

Comparison of ERG RNA ISH and immunohistochemistry. Left panel showing the images of ERG RNA ISH and right panel showing images of ERG immunohistochemistry in benign prostate tissue (A&B); strong positive ERG by RNA ISH with score 4 (C) and positive ERG by immunohistochemistry (D) in a localized prostate carcinoma; weak staining of ERG by RNA ISH with score 1and weak ERG by immunohistochemistry in a metastatic prostate carcinoma (E and F); ERG negative both by RNA ISH (G) and immunohistochemistry (H) in a metastatic prostate carcinoma.

Figure 3

Unique expression of ETV1, ETV4 and ETV5 genes detected by RNA ISH in corresponding positive cases (ETVI RNA-ISH score 2, ETV4 RNA-ISH score 4 and ETV5 RNA ISH score 2 with no cross reactivity or nonspecific hybridization of the probes in the negative cases.

Figure 4

Whole section view of a prostate carcinoma tissue showing tumor molecular heterogeneity with a small foci showing ERG expression by immunohistochemistry (A) and a large tumor foci showing ETV1 expression by RNA ISH (Score 4).Confirmation of ERG and ETV1 rearrangements by FISH (Insets) in the corresponding tumor foci. Green and red color signals indicate 5’ and 3’ probes, respectively.

Figure 5

Whole section view of a second prostate carcinoma tissue showing tumor molecular heterogeneity with a large tumor foci showing ETV1 expression by RNA ISH (Score 4) and two independent small foci showing ERG expression by immunohistochemistry (B). Confirmation of ERG and ETV1 rearrangements by FISH (Insets) in the corresponding tumor foci. Green and red color signals indicate 5’ and 3’ probes, respectively.

Figure 6

A prostate adenocarcinoma sample showing ERG and ETV4 rearrangement in two independent tumor foci. A) ERG IHC; B) ETV1 RNA ISH; C) ETV4 RNA ISH with score 4; D) ETV5 RNA ISH.