In a genome-wide association study (GWAS) the genomes of a group of individuals are scanned widely to identify a range of points at which they differ. This technique is often used to study differences between the genomes of people suffering from a particular disease (cases) and those without that disease (controls). Many genes with a moderate or strong association with one or more types of cancer have been identified in this way.

Two GWAS studies now reported in Nature and Nature Genetics describe novel genetic loci associated with neuroblastoma and renal cell carcinoma respectively.

Neuroblastoma is a tumour of the sympathetic nervous system that occurs in early childhood, over 50% of cases being under two years old at the time of diagnosis. It is responsible for over 10% of all paediatric deaths from cancer. The etiology of this disease, apart from a rare familial form associated with mutations in the anaplastic lymphoma kinase (ALK) gene, is poorly understood.

A group of researchers led by Kai Wang at the Children's Hospital of Philadelphia, Pennsylvania, USA has conducted a genome-wide association study of neuroblastoma using a total of 2,251 patients and 6,097 genetically matched controls taken from four case series. All cases and controls were of European origin. The two larger cohorts were genotyped at over 500,000 positions using Affymetrix or similar technology.

Significant, independent associations with susceptibility to neurobastoma were obtained at three genomic loci, on chromosomes 2, 6 and 11. The neuroblastoma-associated SNPs on chromosome 11p25.4 were both within the locus of the gene LMA01, which encodes a cysteine-rich transcriptional regulator. Other genes in this family have previously been implicated in cancer.

Patients who carried the susceptibility LMA01 variants were significantly more likely to have metastatic disease, to be diagnosed at over one year of age (which is associated with poor prognosis) and to be classified as "high risk". Together, these findings suggest that variations at this locus are associated not only with a predisposition to develop neuroblastoma but with more aggressive forms of the disease.

A parallel genotype of 701 neuroblastoma samples using Affymetrix arrays detected copy number gain of this gene in about 12% of tumours. This copy number increase in the tumour was associated with the same markers of poor prognosis: metastasis, increased age at diagnosis and high risk status. This confirms the identification of LMA01 as a neurobastoma-associated oncogene.

The second study was conducted by a large, international group of researchers led by the National Cancer Institute in Bethesda, Maryland, USA and the International Agency for Research on Cancer (IARC) in Lyon, France and was a two-stage GWAS study of renal cell carcinoma (RCC). This is by far the most common type of kidney cancer; although a number of rare familial variants are known, the vast majority of cases occur sporadically.

Two parallel scans were conducted, coordinated by the International Agency for Research on Cancer (IARC) and the Centre National de Génotypage and involving a total of 3,772 RCC cases and 8,505 matched controls. All subjects were of European origin and each was genotyped at over 500,000 positions.

A meta-analysis of all data from both centres identified six single nucleotide polymorphisms (SNPs) that were associated with an increased risk of developing RCC with some significance. These could all be replicated in further studies with three more independent case-control series. Three of the associated SNPs were localised to chromosome 2p21, with one each on 3q26, 11q13.31 and 12q24.31; the locus on chromosome 3 was least significant and was not followed up to the same extent as the others.

The identification of 2p21 as a risk locus for RCC was considered particularly interesting because this region contains a gene that has already been associated with this cancer. This gene encodes a transcription factor known as hypoxia-inducible-factor-2 alpha (HIF-2a), part of the VHL/HIF pathway that is involved in sensing concentrations of oxygen.

Increases in the concentration of this transcription factor lead to increased expression of several growth factors. This locus was more highly correlated with RCC in men than in women and in smokers than in non-smokers. The susceptibility locus on chromosome 12 maps to SCARB1, a gene that encodes a cholesterol-binding cell surface receptor; the locus on chromosome 11, however, maps to no known genes.

These studies are both good examples of a type of research that has increasing value in determining genetic loci associated with moderately increased cancer risk and thence, by the identification of relevant genes, in modelling the genetic aetiology of tumour development and, potentially, in identifying novel drug targets.

References

Wang, K., Diskin, S.J., Zhang, H., & 30 others (2010) Integrative genomics identifies LMO1 as a neuroblastoma oncogene Nature published online ahead of print 1 December 2010 doi:10.1038/ng.723

Purdue, M.P., Johansson, M., Zelenika, D. & 101 others (2010) Genome-wide association study of renal cell carcinoma identifies two susceptibility loci on 2p21 and 11q13.3 Nature Genetics published online ahead of print 5 December 2010. doi:10.1038/ng.723