by ecancer reporter Clare Sansom

Approximately 2-3% of all cancers in adults occur in the kidney; there are several forms of kidney cancer, and the most prevalent of these is clear cell renal cell carcinoma (ccRCC) accounting for 70-80% of all cases.

There will therefore be approximately 189,000-217,000 new cases of this disease diagnosed worldwide each year.

The molecular mechanisms that lead to the development of this tumour are not yet completely understood, and complete cures only occur if the whole of the tumour can be removed with surgery.

Over 90% of clear cell renal cell carcinomas tested have been found to have deletions of part of chromosome 3p or other lesions that cause inactivation of the Von Hippel–Lindau tumor suppressor
gene VHL.

Other genes that have already been implicated in this tumour include some involved in chromatin modification and in ubiquitin-mediated proteolysis.

A large group of scientists led by Seishi Ogawa of the University of Tokyo, Japan, has now performed a complete integrated analysis of the genetic lesions and molecular pathways involved in this tumour.

Paired tumour and normal DNA from ccRCC tumours were analysed using whole-genome, whole-exome and RNA sequencing, gene expression, DNA methylation and copy number analysis.

Whole-genome sequencing of 100 paired ccRCC tumour and normal kidney DNA revealed a total of 71,424 somatic mutations including 68,273 single nucleotide polymorphisms; the average number of non-silent mutations observed in the tumour samples was 47 and, as expected, VHL was the most frequently mutated gene.

The researchers then analysed a total of 106 paired ccRCC tumour-normal samples using whole-exome sequencing, preferentially selecting cases without VHL mutations, and an average of 48.8 non-silent mutations per tumour was observed.

Fourteen samples were analysed using both whole-genome and whole-exome sequencing, and 539 (64%) of the non-silent mutations observed in these cases were detected using both platforms.

Taking both analyses together, a total of 28 genes were found to be significantly mutated in ccRCC cases compared to background mutation rates.

These genes were analysed in more detail in a total of 240 ccRCC cases using deep or Sanger sequencing and DNA methylation and copy number assays.

A vast majority of these cases (224 or 94%) showed loss of heterogeneity of chromosome 3p, which includes four very commonly mutated genes (VHL, PBRM1, BAP1 and SETD2) between segments 3p25 and 3p21. 

Mutations in BAP1, which encodes a protease that cleaves ubiquitin-protein bonds, were found to be mutually exclusive with mutations in the chromatin remodelling gene PBRM1 and to be  associated with poor prognosis.

Genes outside the 3p region that were found to be frequently mutated included TCEB1, which encodes a small protein called elongin C.

As its name implies, elongin C forms part of a protein complex that induces mRNA elongation, and it is also a component of the VHL complex.

Inactivation of the VHL complex through mutation of either VHL itself or TCEB1 disrupts the ubiquitinlyation of hypoxia-inducible factors (HIF proteins), causing them to accumulate.

Expression of two of these proteins, HIF-1 and HIF-2a, was found to be increased in primary ccRCC tumours with mutations in TCEB1 and VHF compared both to normal kidney tissue and to  tumours without such mutations.

Other genes found to be frequently mutated in this tumour type included TET2, which encodes an enzyme that catalyses an important step in DNA demethylation; the well studied tumour suppressor p53; KEAP1 and NRF2.

The latter two genes encode proteins that interact with each other and with CUL3 to form a complex that is involved in the oxidative stress response.

Copy number analysis showed that a large number of the tumours had large chromosomal lesions; loss of heterogeneity in segment 3p was the most frequent of these and other segments showing chromosomal losses included 8p and 9p.

Gene expression profiling of 101 ccRCC cases identified two major clusters, one of which was associated with mutations in PBRM1 and up-regulation of genes associated with hypoxia, and the other with mutations in BAP1.

Finally, the researchers investigated the outcomes for the patients concerned, and identified one subset of tumours with low methylation and loss of heterogeneity at 9p that was associated with excellent prognosis.

Taken together these comprehensive results present a detailed picture of the genetic aberrations associated with clear-cell renal cell carcinoma, further highlighting the VHL pathway and identifying other frequently dysregulated pathways.

These results may enable these tumours to be stratified according to clinical risk and, eventually, to allow more personalised therapies to be developed.

Reference

Sato, Y., Yoshizato, T., Shiraishi, Y, and others (2013). Integrated molecular analysis of clear-cell renal cell carcinoma. Nature Genetics, published online ahead of print 24 June 2013. doi: 10.1038/ng.2699