Research highlights

Oncogenes: Translocation

Nature Reviews Cancer 7, 9 (01 September 2007) | doi:10.1038/nrc2217

PHOTODISC

Although chromosomal rearrangements creating oncogenic gene fusions are the hallmark of many haematopoietic malignancies, little is known about such rearrangements in solid tumours. Two papers published in Nature have furthered our understanding of gene fusions in prostate cancer and non-small-cell lung cancer (NSCLC).

Fusions of TMPRSS2, which encodes a serine protease, to members of the ETS family of transcription factors (ERG, ETV1 or ETV4) in prostate cancer were first observed by Arul Chinnaiyan and colleagues in 2005. The most frequent fusion is TMPRSS2–ERG; TMPRSS2–ETV1 fusions are rarer, but fewer cases of this fusion have been observed than is expected based on the frequency of ETV1 overexpression in prostate cancer. Now Chinnaiyan and colleagues report that this is because ETV1 has additional 5' fusion partners in prostate tumours. They identified four new partners: a region from 22q11.23 with homology to human endogenous retrovirus family K (referred to as HERV-K_22q11.23), SLC45A3, C15orf21 and HNRPA2B1. Very few or no translated sequences from these 5' partners contribute to the fusion protein, indicating that ETV1 expression is probably driven by regulatory elements in these genes. TMPRSS2 expression is prostate-specific and androgen-induced, and the authors found that SLC45A3 and HERV-K_22q11.23 also fell into this category. Although C15orf21 expression was prostate-specific, it was repressed by androgen, and HNRPA2B1 was not prostate-specific, nor was it androgen-responsive. That these genes respond differently to androgen might have implications for anti-androgen therapy of prostate cancer. Because overexpression of ETV1 was common to all fusions, the authors confirmed an oncogenic effect of ETV1 in normal prostate cells and in mouse prostate. In mice, ETV1 induced mouse prostatic intraepithelial neoplasia (mPIN).

Hiroyuki Mano and colleagues have identified the first recurrent gene fusion in lung cancer, specifically NSCLC. In an attempt to find new oncogenes in NSCLC, they generated a retroviral cDNA library from a patient with NSCLC and carried out focus-formation assays to isolate genes from this library that could transform mouse 3T3 fibroblasts. One gene they identified was a fusion gene containing the N-terminal portion of echinoderm microtubule associated protein like 4 (EML4) and the C-terminal portion of anaplastic lymphoma kinase (ALK), which is commonly translocated in anaplastic large-cell lymphoma. The authors confirmed that EML4–ALK can transform 3T3 cells in vitro, and these cells also form tumours when injected into nude mice. Kinases in fusion proteins are often activated by dimerization driven by the fusion partner, and co-immunoprecipitation experiments showed that the basic domain of EML4 can cause fusion protein dimerization. Furthermore, an ALK kinase inhibitor significantly reduced the growth of BA/F3 cells expressing EML4–ALK. The authors analysed clinical samples from 75 patients with NSCLC and found 5 that contained EML4–ALK. Interestingly, none of these 5 patients had epidermal growth factor receptor (EGFR) mutations, indicating that NSCLC with the EML4–ALK fusion is a new class of NSCLC that might be successfully treated with ALK inhibitors.

The findings by Chinnaiyan and colleagues, and Mano and colleagues suggest that many solid tumours probably contain gene fusions that could be central to their pathogenesis, and thus might represent suitable biomarkers and therapeutic targets.

Author: Sarah Seton-Rogers