Flipping the switch

In mice, the pulmonary adenoma susceptibility 1 ( Pas1 ) locus confers susceptibility to lung tumours and comprises six genes, including Kras2 , which is frequently mutated in chemically induced lung tumours. During the course of their attempt to show that Kras2 is the Pas1 gene, Allan Balmain and colleagues unexpectedly found that the expression levels of Kras2 determine whether Pas1 confers either susceptibility or resistance to tumorigenesis.

M. spretus (SPRET/Ei) mice carry a dominant Pas1 resistance haplotype, making them resistant to chemically induced lung tumours. Mice that carry the mutant Kras2^LA2 allele develop lung tumours with complete penetrance and without the need for chemical carcinogenesis. To fix the origin of the mutant allele to a particular mouse strain, Kras2^LA2 mice were crossed to FVB/N mice. The FVB/N mice have a Pas1 susceptibility haplotype and, as expected, the Kras2^LA2FVB/N mice all developed lung tumours. To test whether or not Kras2 is the Pas1 gene, the authors backcrossed female FVBSPRETF1 hybrids with Kras2^LA2FVB/N males. They genotyped the mice to test for linkage to Pas1, and expected to see one of two possibilities. If the Pas1 gene was Kras2, they expected to see no linkage to the region because every mouse had the same parental mutant Kras2 allele. If the Pas1 gene was not Kras2, but another gene in the locus, they expected to see linkage to the region and resistance to tumours in those mice that inherited the SPRET/Ei Pas1 resistance allele. To their surprise, neither of these possibilities occurred. They detected linkage to the region, but found that mice that carried the SPRET/Ei Pas1 resistance allele were actually more susceptible to lung tumours than those that carried the FVB/N Pas1 susceptibility allele.

Why does the SPRET/Ei Pas1 allele sometimes confer resistance and sometimes susceptibility to lung tumours? The authors propose that Kras2 is indeed Pas1, and that suppression of the oncogenic effect of a mutant Ras allele can be mediated by expression of the wild-type allele. Owing to a polymorphism, the SPRET/Ei Kras2 allele is poorly transcribed compared with the FVB/N Kras2 allele. Therefore, when mice carry both wild-type and mutant Kras2 alleles from FVB/N mice, they are less susceptible to tumours because the wild-type allele is more highly expressed and therefore more able to suppress the mutant allele. This does not occur in mice with the less actively transcribed SPRET/Ei wild-type Kras2 when they inherit the FVB/N Kras2^LA2 mutation. Importantly, in chemical carcinogenesis, SPRET/Ei mice are resistant to lung tumour formation as mutation at Kras2 on the SPRET/Ei allele results in low expression of the mutant protein.

What are the potential implications of this finding? Allele-specific transcriptional activity is common in the human genome. It is not clear how widespread context-dependent susceptibility will be, but the data presented by Balmain and colleagues indicate that further biological data might need to be taken into account in genetic-association studies to determine cancer susceptibility genes.