by ecancer reporter Clare Sansom

Prostate cancer is the second most common cancer in men.

It is a heterogeneous disease; while many patients can live with indolent prostate cancer for many years, a minority of cases are very aggressive.

Accurate prediction of the prognosis of individual tumours is important as over-treatment of indolent disease can cause significant morbidity.

The genetic basis of prostate cancer development, progression and response to therapy is not yet fully understood.

Two large studies published in Nature journals have now shed further light on the genetic landscape of this disease.

Both used the technique of exome sequencing, or the selective sequencing of the coding regions of the tumour genome.

The first study, led by Levi Garraway at the Dana-Farber Cancer Institute, Cambridge, Massachusetts, USA, sequenced the exomes from matched tumour and normal tissue from 112 prostate cancer patients.

All patients were diagnosed with adenocarcenoma of the prostate, with tumours at a wide range of stages, grades and risks of progression. A total of 5.764 somatic mutations were detected in these tumours, 997 in one highly mutated sample; other tumours harboured from 10 to 105 total mutations.

The researchers identified a total of twelve genes that were significantly enriched in non-synonymous mutations in this tumour set, including the well-known “cancer genes” PIK3CA, TP53 and PTEN.

However, the most frequently mutated gene, mutated in 13% of the tumours, was SPOP, which encodes a subunit of a ubiquitin ligase. Garraway and his colleagues then sequenced this gene in multiple cohorts comprising a further 300 primary prostate tumours and metastases. Between six and 13% of tumours in each cohort were found to harbour non-synonymous mutations in SPOP, and all these mutations found to affect conserved residues in the substrate-binding cleft of its protein product.

Prostate cancers bearing SPOP mutations were found to have a specific pattern of further genetic alterations, indicating a potential new subtype of prostate cancer. Recurrent mutations were also observed in some other genes, including a transcription regulator, MED12 and a DNA-binding protein that interacts with chromatin, FOXA1.

The second study, by Arul Chinnaiyan of the University of Michigan, Ann Arbor, MI, USA and his colleagues, sequenced the exomes from prostate tumour samples obtained from 61 patients who had died of their disease. Fifty of these were heavily treated castration-resistant prostate cancers (CRPC) and the other 11 were treatment-naïve, high-grade localised tumours. CRPC is a form of prostate cancer that arises when tumours become resistant to androgen deprivation therapy.

All tumours, even the heavily treated ones, were found to have relatively low mutation rates, confirming that CRPC has a monoclonal origin. Nine genes were found to be significantly mutated in this tumour set. Six of these, TP53, AR, ZFHX3, RB1, PTEN and APC, have previously been identified as being involved in the development of prostate cancer.

The other three, without previously described roles in prostate cancer, were MLL2, CDK12 and OR5L1. The first two of these have been implicated in some other tumour types; MLL2 encodes a histone methyltransferase, involved in chromatin regulation, and CDK12 encodes a cyclin-dependent kinase and is therefore involved in cell cycle regulation.

Copy number analysis identified CHD1 as a gene that is deleted or mutated in a significant proportion of these cancers, and showed that deletion of CHD1 was associated with disruption of the ETS family of transcription factors. Interestingly, however, tumours with CHD1 deletions lacked the fusions between ETS genes and androgen-driven genes that are found in many prostate tumours. This was seen to define a CHD1 negative, ETS fusion negative genetic subtype of prostate cancer.

The analysis also identified a number of genes involved in chromatin and histone regulation other than MLL2 as being mutated in a number of the prostate cancers. Some of these encode components of the MLL complex.

The protein products of these genes were shown using immunoprecipitation to interact with the androgen receptor (AR). Other genes identified as recurrently mutated in the prostate tumours are also known to encode proteins that physically interact with this receptor; these include the AR collaborating factor FOXA1.

Both these papers provide important information about the genetic landscape of prostate cancer that may lead to improved prognostic prediction or even to novel therapies for defined subtypes of this disease.

References

[1]: Barbieri, C.E., Baca, S.C., Lawrence, M.S. and 43 others (2012). Exome sequencing identifies recurrent SPOP, FOXA1 and MED12 mutations in prostate cancer. Nature Genetics, published online ahead of print 20 May 2012. doi:10.1038/ng.2279

[2]: Grasso, C.S., Wu, Y-M., Robinson, D.R. and 24 others (2012). The mutational landscape of lethal castration-resistant prostate cancer. Nature, published online ahead of print 20 May 2012. doi:10.1038/nature11125