Until recently, gene rearrangements, e.g. the Philadelphia chromosome, were thought to be most prevalent in haematological cancers, such as leukaemia. However, in 2005, most prostate cancers (up to 70%) were found to have fusion of the androgen-responsive genes transmembrane protease, serine 2 (TMPRSS2) and oestrogen-regulated gene (ERG), both on chromosome 21. Soon thereafter, other members of erythroblast transformation-specific (ETS) variant gene (ETV) family were found to have gene fusions, although at much lower frequencies, including ETV1 (chromosome 7), ETV4 (chromosome 17), ETV5 (chromosome 3), and ETS domain-containing protein gene (ELK4, chromosome 1) .
What is a gene rearrangement or fusion? It consists of actual loss of genetic material between two genes on the same chromosome; note the deletion of intervening genes between TMPRSS2 and ERG on chromosome 21 (Fig. 1). A second form of fusion is due to translocation, when a gene moves to another location on the same chromosome or a different chromosome (Fig. 1). Both mechanisms apply for the gene rearrangements in prostate cancer. Given the high prevalence of prostate cancer, this fusion gene is probably the most common fusion gene in human cancer. Further, it appears to be limited exclusively to prostate cancer .
Importantly, TMPRSS2 is an androgen-responsive gene. Consequently, ERG overexpression resulting from fusion may contribute to development of androgen-independence in prostate cancer through disruption of androgen receptor signalling, especially in castrate-resistant prostate cancer.
ERG overexpression can be exploited for the diagnosis of prostate cancer in tissue, urine and blood samples [4-6]. The prognostic value of TMPRESS2-ERG gene fusion is a hotly debated topic in the current literature [4, 5], although most studies indicate that the presence of the fusion gene/gene product denotes an unfavourable outcome.
Various methods can detect TMPRESS2-ERG fusion, including: (i) ERG protein overexpression in nuclei by immunohistochemistry; this is most commonly used, is easy to perform, and is inexpensive; (ii) Fluorescence in situ hybridisation (FISH) , a complex, expensive test that requires experience for proper interpretation; (iii) DNA sequencing, a highly complex test that usually requires fresh tissue (most useful); and (iv) Genechip (present article ), a complex test that requires fresh tissue.
It is expected that the greatest benefit of this new biomarker of prostate cancer gene fusion will be creation of targeted therapies for inhibiting the fusion gene or gene product. By any measure, this represents a significant advance in our understanding of prostate cancer, and promises to improve diagnostic accuracy and treatment outcomes in the foreseeable future.