Detection of prostate specific ETS fusion transcripts in cancer samples

L. V. Mevs, G. V. Gerashchenko, E. E. Rosenberg © 2017 L. V. Mevs et al.; Published by the Institute of Molecular Biology and Genetics, NAS of Ukraine on behalf of Biopolymers and Cell. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited UDC 577.218+616.65


Introduction
Gene fusions, which involve the ETS family of transcription factors (ERG, ETV1, ETV4, ETV5) are a frequent event in prostate cancer [1].Usually their 5' fusion partners are androgenregulated genes, particulary TMPRSS2, KLK2, CANT1, SLC45A3 and NDRG1.The most common of them is the TMPRSS2 gene.Because of their expression under the TMPRSS2 promoter the fusion transcripts were demonstrated to be up-regulated by androgenic hormones in the androgen-dependent prostate cancer tissue [2].Transcription factors from the ETS family activate the oncogenic transcriptional programs, such as invasion, induction of DNA breaks and enhance the ETS oncogenic functions [3].The ETS proteins are involved in the signaling pathways associated with cancer development [3,4].It was also found that some members of this family can form the fusion transcripts, which may produce functional fusion proteins and proteins of the 3'-partner genes [3,5].These fusion transcripts can be potentially formed as a result of deletions and translocations [6,7].Although the mechanism of their formation has not been established yet, there is the evidence that noncoding RNAs, hormone levels, genotoxic stress and other factors play an important role in this process [8].It has been shown that different exons and introns of both genes could be involved in the formation of their fusion transcripts in prostate cancer.That is why they usually have several isoforms [2,4,5,9,10].
It was shown that specific fusion transcripts are detected with different frequency within populations of USA [1,2,[8][9][10], Germany [2], Slovenian [11], Sweden [12] and Asia [13], but the frequency of detection of those transcripts for Ukrainian population has not been investigated yet.Therefore, it is important to determine what kind of these fusion transcripts is able to be expressed in tissues of the prostate cancer patients in Ukrainian population.The aim of this study was to detect several types of the fusion transcripts in Ukrainian population and to analyze possible relationships between the ETS fusion transcripts and the clinical characteristics of prostate cancer.Total RNA isolation, cDNA synthesis.Total RNA extraction was done with RNeasy Mini kit (Qiagen, Germany) according to the manufacturer's recommendations after preparation of sample prouder in liquid nitrogen.Total RNA samples quality was determined in 1 % agarose gel by band intensity of 28S vs 18S rRNA (28S/18S ratio) and spectrophotometrically (ratios of 260 nm vs 230 nm and 280 nm respectively).Concentration of total RNA was monitored with absorbance spectra at 260 nm (NanoDrop Technologies Inc. USA).The RNase free DNaseI treatment of total RNA was used to clean samples from genomic DNA.cDNA was synthesized from 1 µg of total RNA using RevertAid H Minus M-MuLV Reverse Transcriptase (Thermo Fisher Scientific, USA) according to the manufacturer's protocol.
The reference genes TBP, HPRT, ALAST1 and K-ALPHA-1 were used for the gene expression normalization [14], the model for expression calculation was Livak 2 -ΔCt and 2 -ΔΔCt method-based relative quantities from the uploaded raw threshold cycle data [15].
Statistical analysis.The Kolmogorov-Smirnov and Lilliefors tests were used for assessing normality of distribution.Kruskal-Wallis test was used to determine differences in multiple comparison between groups.Mann-Whitney test was used for assessment of differences for the expression of TMPRSS2-ERG fusion in groups with various stages, Gleason scores and levels of the prostate specific antigen.Spearman rank test was used to analyze associations between the expression of

Results and Discussion
It is known that fusion transcripts can be formed by the TMPRSS2 gene and the ETS family members in prostate cancer [2] and expression of some fusion transcripts may correlate with the clinical data [11].The gene fusion products are involved in the regulation of cell cycle, prostate cancer initiation, progression and invasive potential of cells [4,17].Thus, we decided to detect their expression in Ukrainian population and to test whether they can be used as markers of the development and progression of prostate cancer.We studied expression of 6 fusion transcripts: TMPRSS2-ERG, TMPRSS2-ETV1a, TMPRSS2-ETV1b, TMPRSS2-ETV4a, TMPRSS2-ETV5b, TMPRSS2-ETV5c (Fig. 1) in prostate cancer samples (Table 2).These gene fusions typically result in the fusion of the 5′-part of androgen-regulated genes to the oncogenic ETS family genes, causing their over-expression [17].Taken TMPRSS2-ERG is namely TMPRSS2-ERG, isoform 2 (EF194202.1),which was detected with a high frequency in the study of Lapointe J. et al. [12].
Bioinformatic analysis using ORF finder (https://www.ncbi.nlm.nih.gov/orffinder/) and BLASTP https://blast.ncbi.nlm.nih.gov/Blast.cgi) has shown that some fusion transcripts can produce proteins: TMPRSS2-ERG potentially can produce the ERG protein, TMPRSS2-ETV4a -ETV4 protein; some of them may even be able to produce fusion proteins (e,g, TMPRSS2-ETV5 fusion transcripts).Our data did not show the existence of any proteins from the TMPRSS2-ETV1a and TMPRSS2-ETV1b fusion transcripts, but an experimental study indicated that TMPRSS2-ETV1a can produce the ETV1 protein [3].
Our study of 6 fusion transcripts in 37 prostate carcinoma samples allowed us to detect only one fusion transcript -TMPRSS2-ERG (Table 2).This transcript was extracted from agarose gel, cleaned by GeneJET Extraction Kit (Thermo Scientific) and sequenced.The presence of the expected fusion transcript in prostate cancer samples was confirmed (Fig. 2).
In our study this fusion transcript was detected in 21 out of 37 prostate cancer samples, i.e. with a frequency of 56.8 %, confirming the frequency of its detection in other studies (40-80 %) [1,2,4].
The androgen signaling has been shown to induce the proximity of TMPRSS2 and ERG locus in androgen responsive cells, and in combination with agents causing DNA double strand breaks induces the TMPRSS2-ERG gene fusions [8,19,20].So the level of androgens in the body plays an important role in the expression of fusion transcripts formed with the TMPRSS2 gene.Also the TMPRSS2-ERG fusion transcript product binds to the ERG locus and drives the over-expression of wild-type ERG in prostate cancers [18].Moreover, androgen signaling-mediated up-regulation of TMPRSS2-ERG resulted in the concomitant up-regulation of the wild-type ERG transcription in VCaP cells.The loss of wild-type ERG expression was associated with a decrease in their invasive potential [18].
The expression of other fusion transcripts was not detected.Notably, their expression depends on many factors, including features of prostate carcinogenesis and studied populations.For instance, the TMPRSS2-ETV1, TMPRSS2-ETV4 and TMPRSS2-ETV5 fusion transcripts were detected in the highest frequency in heterogeneous American populations by Tomlins S. et al. [2,9] and Helgeson B. et al. [10], but these three fusion tran-   Notes: "+" -detected fusion transcript; "-" -no expression of fusion transcript.place of the junction scripts were not detected in the study of Dong J. et al. in Eastern Chinese population [13].They usually have a low frequency (1-10 %), so their detection requires a large number of samples.We studied whether an expression frequency of the] TMPRSS2-ERG fusion transcript is associated with the clinical characteristics (stages, Gleason scores and levels of the PSA) of prostate cancer patients though no such correlation was found.We observed approximately the same expression frequency of TMPRSS2-ERG in all prostate adenocarcinoma samples of different stages and three groups of Glisone score.So, there is the hypothesis that these changes can occur at an earlier stage and proceed at all stages of the tumor development and growth .There is evidence that the expression of the fusion transcripts can be observed at the earliest stages of cancer [2].It is most likely that the emergence of the fusion transcript is typical for a specific mechanism of oncogenic transformation.That is why we investigated the expression of the detected fusion transcript in paired conventionally normal tissue samples and adenomas.2 -ddCt model was used for calculation of the TMPRSS2-ERG relative gene expression, normalized to the mean of the adenomas group.The levels of relative TMPRSS2-ERG expression in three types of prostate tissues are shown in Fig. 3.
As previously reported, this fusion transcript was detected in 21 samples of prostate cancer.The expression of TMPRSS2-ERG was detected in 16 normal prostate tissue samples and 4 samples of prostate adenoma.The fusion transcript, detected in these samples, was sequenced, and the TMPRSS2-ERG fusion transcript was confirmed (Fig. 2).Kruskal-Waillis test determined differences in the TMPRSS2-ERG relative expression levels among investigated groups with p = 0.001.All fusion positive samples were distributed into 7 groups: 3 adenocarcinoma groups with Gleason score < 7, = 7, > 7 and 3 respective groups of normal tissues divided by Gleason score < 7, = 7, > 7, and the adenoma samples group.Mann-Witney test has shown significant differences between 9 pairs of investigated groups.All 3 groups of prostate adenocarcinoma with 3 types of Gleason score (< 7, = 7, > 7) and one group of normal tissue with Gleason score > 7 demonstrated a significant increase in the TMPRSS2-ERG relative expression levels compared with the adenoma group (p = 0.04; p = 0.04; p = 0.01; p = 0.012 respectively).Among paired groups of prostate adenocarcinoma and normal tissue with the same Gleason score the only significant differences were detected between adenocarcinoma and normal tissue with Gleason score < 7 (p = 0.045).Two other prostate adenocarcinoma groups with Gleason score = 7 and Gleason score > 7 had significant differences (increased relative expression levels) with normal tissue Gleason score < 7 (p = 0.025; p = 0.003 respectively).Furthermore, we found the evidence of significant increase in the TMPRSS2-ERG relative expression level in the prostate adenocarcinoma groups with Gleason score > 7 compared to normal tissue with Gleason score = 7 (p = 0.03).Among groups of normal tissue with different Gleason score significant differences between Gleason score < 7 and Gleason score > 7 (p = 0.08) were detected.No differences were detected between the prostate adenocarcinoma groups with different Gleason score.The Benjamini-Hochberg procedure with false dis-covery rate 0.10 was used to correct p value under multiple comparison detection [21].All our data with p value (p < 0.05) were detected as significant in this procedure with 0.1 false discovery rate.
In 21 samples of fusion positive prostate adenocarcinoma samples we detected 16 paired fusion positive normal samples.Those were not false positive results, because all TMPRSS2-ERG transcripts from normal samples were sequenced.This is consistent with the notion that some fusion transcripts can be detected in normal prostate tissue [3].This may mean that there are tumor cells in normal tissue, or the normal tissue has already undergone oncogenic changes and is no more the normal one.Besides, the fusion transcripts are already found in the adenomas (4 fusion positive specimens in the prostate adenomas of 20 samples).
Our results have shown that the TMPRSS2-ERG fusion transcript is present at early stages of the cancer development and its expression level significantly increases in a high Gleason score group.
It was detected that the TMPRSS2-ERG fusion transcript has an oncogenic effect on the development of prostate cancer, since it can potentially produce the oncogenic ERG protein [3].It is well known that ERG is involved in the Wnt signaling pathway, which regulates gene transcription, cell proliferation and cell migration [4].There is evidence that one of its isoforms is capable of production of a chimeric protein that is able to regulate the activity of the insulin-like growth factor-1 receptor (IGF1R).It plays a key role in cell growth and tumorigenesis, and is overexpressed in most malignancies, including prostate cancer, and the IGF1 signaling pathway, an important cell survival network.Trans activa tion of the IGF1R gene by oncogene ERG constitutes a key event in the prostate cancer development.Enhanced activation of the overexpressed IGF1R by locally produced or circulating IGF1 or IGF2 may provide a selective advantage to tumor cells [20].
Additionally, our data suggest the formation of this transcript at the early stages of carcinogenesis of the prostate.In future it can be used in the development of a method for early diagnosis and prognosis of the course of the disease, taking into account the data obtained in other studies [11,22].
The recent studies demonstrate the connection between androgen receptor expression and expression of the TMPRSS2-ERG fusion in prostate cancer, giving arguments in favor of the fact that the expression of this fusion transcript depends on the influence of androgens in hormone-dependent tumors [4,23,24].It is also known that the expression of TMPRSS2-ERG fusion cooperates with such signaling pathways as RAS/RAF and PI3K/PTEN/AKT [24].Krstanoski Z. et al. tested association between the TMPRSS2-ERG fusion transcript by FISH method and several clinicopathological variables, i.e., pT stage (describes the size of the primary tumour), extended lymph node dissection status, and Gleason score, correcting for multiple comparisons.Only the association with pT stage was significant (p = 0.05) [16].
Because the credible mechanisms of formation of fusion transcripts are still unknown, our objective for further research is to reveal the mechanism of their formation in the patients with prostate cancer, as well as the mechanisms by which fusion transcripts affect the development of prostate cancer disease.

Conclusions
Among 6 investigated fusion transcripts (TMPRSS2-ERG, TMPRSS2-ETV1a, TMPRSS2-ETV1b, TMPRSS2-ETV4a, TMPRSS2-ETV5b, TMPRSS2-ETV5c) only the TMPRSS2-ERG fusion transcript was detected.Subsequent sequencing confirmed the presence of TMPRSS2-ERG fusion transcript.Its frequency in prostate tumors was determined to be 56.8 % in our group.Additionally, we observed the expression of TMPRSS2-ERG in 16 normal prostate tissue samples (43.2 %) and 4 samples of prostate adenoma samples (20 %).It suggests that the frequency of TMPRSS2-ERG fusion transcript would be rather the same as in other European populations and higher than in Asian populations.The obtained results show that the fusion transcript TMPRSS2-ERG is present at the early stages of cancer.
Whereas the ETS fusions were discovered several years ago, and are important in several aspects of the prostate cancer initiation and progression, the formation of TMPRSS2-ERG fusion is still under investigation [4].Thus, in our further studies we intend to identify the mutation which could form the TMPRSS2-ERG fusion transcript and test whether it can be used as a diagnostic marker of the formation and development of prostate cancer, in investigation of the fusion positive tumors carcinogenic features.

Fig. 1 .
Fig. 1.Schematic representation of the fusion transcripts used in the study.Note: ATG -start codon for protein.