News

Wilms tumour susceptibility loci discovered through genome wide association scanning

1 May 2012
Wilms tumour susceptibility loci discovered through  genome wide association scanning

by ecancer reporter Clare Sansom

 

Wilms tumour is a relatively rare paediatric cancer of the kidneys which affects approximately one in 10,000 children in Western countries and is most commonly diagnosed between three and four years of age.

 

Its prognosis is very good except in rare cases in which both kidneys are involved, with an overall five-year survival rate of about 90%.

 

There is strong evidence of a genetic component to Wilms tumour development.

 

Approximately 2% of all cases occur in susceptible families, and several genetic syndromes involving an increased risk of the tumour have been identified.

 

Mutations in two genes, WT1 on chromosome 11 and WTX on the X chromosome have been identified in a minority of cases.

 

However, the genetic events associated with most cases of Wilms tumour remain unknown.

 

A large, mainly UK-based group of researchers led by Nazneen Rahman of the Institute of Cancer Research, Sutton, Surrey, UK, has now carried out a genome wide association scan in order to identify genetic regions associated with increased susceptibility to Wilms tumour.

 

They analysed data on about 600,000 SNP positions from a total of 757 Wilms tumour cases recruited from North American oncology clinics, and 1,879 North American controls taken from the Database of Genotypes and Phenotypes (dbGaP). All subjects were of European ancestry and all DNA samples were of high quality. Genotype frequency in cases and controls were compared using the one-degree-of-freedom Cochran-Armitage trend test.

 

The initial scans identified twenty SNPs in nine linkage disequilibrium regions that appeared to be significantly associated with susceptibility to Wilms tumour (p < 5x10-5). Elimination of closely correlated SNPs reduced this set to ten SNPs at nine loci, and these were evaluated in two independent case-control series, one from the UK and the other from the US.

 

The UK series included 769 cases and 2,814 controls, and the US set 719 cases and 1,037 controls, with the US controls again obtained from dbGaP. Two SNPs from chromosomal region 2p24 and one from 11q14 remained significantly associated in both series, and three further SNPs were of clear significance in the US series but not the UK one; the researchers suggested that these, in regions 5q14, 22q12 and Xp22, would be worthy of further evaluation.

 

The genotype probabilities at the remaining loci were estimated using data from HapMap3 and the 1000 Genomes Project, and clear associations with Wilms tumour susceptibility identified for three polymorphisms: rs3755132 and rs807624 at 2p24 and rs790356 at 11q14. rs3755132

 

The strongest evidence of association was found at rs790356 (p = 4.25x10-15). This SNP lies in a region on chromosome 14 that contains the gene DLG2.

 

This gene encodes a guanylate protein kinase that is involved in tissue morphogenesis and is known to interact with a protein, SCRIB, which is one of the targets of the Wilms tumour protein WT1. It is therefore likely that this protein acts to mediate Wilms tumour development within the same pathways as WT1.

 

The correlation between the two SNPs in region 2p24 associated with Wilms tumour was shown to be weak, suggesting that each is independently associated with the disease.

 

It is, however, possible that both arise from a third variant, not observed in this study and in linkage disequilibrium with each. These SNPs are both located in a region that is amplified in many paediatric tumours, most often in neuroblastoma, and rs3755132 is located upstream of the gene DDX1 (DEAD box helicase 1) which encodes a DNA repair protein that can plausibly be linked to carcinogenesis.

 

In summary, Rahman and her colleagues have identified two definite loci associated with a predisposition to Wilms tumour, each of which contains at least one candidate gene, and three further candidate loci worthy of further investigation. They stressed, however, that these represent only part of the genetic variability associated with susceptibility to this tumour.

 

 

Reference 

Turnbull, C., Perdeaux, E.R., Pernet, D. and 29 others (2012). A genome-wide association study identifies susceptibility loci for Wilms tumor. Nature Genetics, published online ahead of print 29 April 2012. doi:10.1038/ng.2251