Human endogenous retrovirus HERV‐K (HML‐2) activity in prostate cancer is dominated by a few loci

W Goering, K Schmitt, M Dostert, H Schaal… - The …, 2015 - Wiley Online Library
W Goering, K Schmitt, M Dostert, H Schaal, R Deenen, J Mayer, WA Schulz
The Prostate, 2015Wiley Online Library
Background Increased expression of human endogenous retroviruses, especially HERV‐K
(HML‐2) proviruses, has recently been associated with prostate carcinoma progression. In
particular, a HML‐2 locus in chromosome 22q11. 23 (H22q) is upregulated in many cases.
We therefore aimed at delineating the extent and repertoire of HML‐2 transcription in
prostate cancer tissues and cell lines and to define the transcription pattern and biological
effects of H22q. Methods Sanger and high throughput amplicon sequencing was used to …
Background
Increased expression of human endogenous retroviruses, especially HERV‐K(HML‐2) proviruses, has recently been associated with prostate carcinoma progression. In particular, a HML‐2 locus in chromosome 22q11.23 (H22q) is upregulated in many cases. We therefore aimed at delineating the extent and repertoire of HML‐2 transcription in prostate cancer tissues and cell lines and to define the transcription pattern and biological effects of H22q.
Methods
Sanger and high throughput amplicon sequencing was used to define the repertoire of expressed HML‐2 in a selected set of samples. qRT‐PCR was used to quantify expression of selected proviruses in an extended set of prostate cancer tissues. Transcription factor binding sites (TFBS) were compared bioinformatically using the Transfac database. Expression of H22q was further characterized by siRNA‐mediated knockdown, 5′ RACE mapping of transcriptional start sites (TSS) and identification of splice sites. Functional effects of H22q knockdown were investigated by viability and apoptosis assays.
Results
In addition to H22q, a limited number of other proviruses were found expressed by sequencing. Of these, provirus ERVK‐5 and to a lesser degree ERVK‐15 were frequently upregulated in prostate cancer. In contrast, expression of ERVK‐24, predominant in germ cell tumors, was not detectable in prostatic tissues. While HML‐2 LTRs contain binding sites for the androgen receptor and cofactors, no consistent differences in transcription factor binding sites were found between expressed and non‐expressed proviruses. The H22q locus contains two 5′‐LTRs of which the upstream LTR is predominantly used in prostatic cells, with an imprecise TSS. Splicing of H22q transcripts is complex, generating, among others, a transcript with an Np9‐like ORF. Knockdown of H22q did not significantly affect proliferation or apoptosis of prostate cancer cells.
Conclusions
Our findings further underline that HML‐2 expression is commonly highly tissue‐specific. In prostate cancer, a limited number of loci become activated, especially H22q and ERVK‐5. As expressed and non‐expressed proviruses do not differ significantly in TFBS, tissue‐ and tumor‐specific expression may be governed primarily by chromatin context. Overexpression of HML‐2 H22q is more likely consequence than cause of prostate cancer progression. Prostate 75:1958–1971, 2015. © 2015 Wiley Periodicals, Inc.
Wiley Online Library