Complexity of lung cancer modifiers: mapping of thirty genes and twenty-five interactions in half of the mouse genome.

Abstract

BACKGROUND

Numerous low-penetrance genes control susceptibility to cancer in experimental animals, but the overall genetic information on this group of genes (i.e., number of loci and their mutual interactions) is missing. We performed a systematic search, scanning roughly half of the mouse genome for lung cancer susceptibility (Sluc) genes affecting tumor size or number by using mouse recombinant congenic (RC) strains. In each RC strain (OcB), approximately 12.5% of the genome is derived from the lung cancer-resistant strain B10.O20, whereas the rest is derived from the lung cancer-susceptible strain O20.

RESULTS

We detected a total of 30 Sluc loci (16 new plus 14 previously reported) and 25 two-way interactions. Some of these interactions are counteracting (e.g., Sluc17 and Sluc20), resulting in the partial or total masking of the individual independent effect (main effect) of each involved locus. Seven loci (Sluc1, Sluc5, Sluc12, Sluc16, Sluc18, Sluc20, and Sluc26) and two interactions (Sluc5 x Sluc12 and Sluc5 x Sluc26) were detected in more than one RC strain.

CONCLUSIONS

The extrapolation of our results to the whole genome suggests approximately 60 Sluc loci (90% confidence intervals = 42 to 78). Despite the genetic complexity of lung cancer, use of appropriate mapping strategies can identify a large number of responsible loci and can reveal their interactions. This study provides an insight into the genetic control of lung tumorigenesis and may serve as a paradigm for investigating the genetics of other cancer types.

METHODS

A total of 730 F2 hybrids from five (OcB x O20) crosses were tested. Pregnant mice were treated on day 18 of gestation with a single dose of N-ethyl-N-nitrosourea. When offspring were 16 weeks old, whole lungs were removed and sectioned semiserially, and the size of all lung tumors (n = 2658) was determined. Analysis of variance was used for detection of linkage, and models (including main effect and two-way interactions) were tested with a statistical program.