Mapping Six New Susceptibility to Colon Cancer (Scc) Loci Using a Mouse Interspecific Backcross

Chevonne D. Eversley, Yuying Xie, R. Scott Pearsall, David W. ThreadgillG3 2012 · DOI

TL;DR

This study leverages a mouse interspecific backcross to identify six new genetic loci (Scc16-21) that influence susceptibility to colon cancer, bringing the total known loci to 23. By using a more genetically diverse mouse strain (Mus spretus), the researchers were able to uncover loci that might be missed in studies using more closely related strains, paving the way for understanding complex genetic interactions in cancer development.

What Problem Does This Paper Address?

Colorectal cancer (CRC) is a complex disease influenced by a combination of genetic and environmental factors. While many genetic loci contributing to CRC susceptibility have been identified, their individual effects are often small, making them difficult to detect, especially in human studies. This difficulty is compounded by the fact that these loci can interact with each other (epistasis), leading to larger effects than predicted by their individual contributions. Previous mouse studies, often using inbred strains with significant genomic similarity, may have missed susceptibility loci located in regions of lower genetic diversity. This paper addresses the challenge of identifying these low-effect, potentially interacting susceptibility loci by employing a mouse model with greater genetic divergence.

Methods

The researchers utilized an interspecific backcross strategy. This involved crossing two distinct mouse species: Mus musculus (strain A/J, known to be susceptible to chemically induced colon tumors) and Mus spretus (strain SPRET/EiJ, known to be resistant). The initial cross produced infertile males, so females from this F1 generation were then backcrossed to the A/J strain to create an N2 generation population of 235 mice. This interspecific backcross generates a population with a broader range of genetic variation compared to intraspecific crosses.

Cancer phenotyping involved administering the carcinogen azoxymethane (AOM) to the mice, a standard method for inducing colon tumors that mimic human sporadic CRC. Mice were monitored for tumor development, and detailed measurements of tumor number, size, and location along the colon were recorded.

Genotyping was performed using a custom Sequenom MassARRAY SNP Genotyping platform with 183 single nucleotide polymorphism (SNP) markers specifically chosen to be informative between the M. musculus and M. spretus genomes.

Quantitative Trait Loci (QTL) mapping was conducted using the J/qtl software. For the binary phenotype of tumor presence/absence, a maximum likelihood approach was used. For quantitative tumor phenotypes (number, size, load, position), data were transformed (rank z-transformation) to fit a Gaussian distribution, and one-dimensional genome scans were performed. To identify epistatic interactions, two-dimensional modifier scans using Haley-Knott regression were employed, with significance thresholds determined through 1000 permutations.

Datasets

The primary dataset used in this study is derived from the (ASF1)A N2 interspecific backcross population generated by the authors. This population consists of 235 mice, each phenotyped for colon tumor development and genotyped for a panel of informative SNPs. The study also references and aims to validate previously mapped colon cancer susceptibility loci (Scc and Ccs) from existing literature, effectively using these as a benchmark for replication.

Key Insights and Ideas

  • The power of genetic diversity: Using Mus spretus, a species that diverged from Mus musculus over a million years ago, allowed the researchers to explore genomic regions not typically covered by studies using more closely related mouse strains. This highlights the importance of genetic diversity in uncovering a more complete picture of complex disease susceptibility.
  • Colon cancer is truly oligogenic: The identification of six new loci, in addition to replicating six previously mapped ones, reinforces the idea that CRC susceptibility is controlled by a multitude of genes, each with a small effect. This complexity makes it challenging to pinpoint specific genetic drivers in human populations.
  • Location matters: Three of the newly identified loci specifically influence the position of tumors along the proximal-distal axis of the colon. This suggests that genetic modifiers can impact not just the presence or number of tumors, but also their anatomical distribution, which could have implications for disease progression and treatment.
  • Interactions are key: The study provides evidence for epistatic interactions, where combinations of alleles at different loci have a greater effect than the sum of their individual effects. This is crucial for understanding how seemingly low-penetrance alleles can collectively contribute to significant disease risk.

Key Results

  • Six new susceptibility loci identified: The study successfully mapped six novel colon cancer susceptibility loci, designated Scc16 through Scc21, and identified two additional suggestive loci.
  • Replication of known loci: Six previously identified colon cancer modifier loci were replicated in this interspecific backcross population.
  • Tumor location modifiers: Three loci were found to modify the proximal-distal location of tumors within the colon.
  • Major resistance locus identified: Scc16, located on chromosome 11, was identified as a major locus contributing to tumor resistance, with heterozygous mice showing a significantly increased likelihood of being tumor-free.
  • Epistatic interactions detected: The study found significant epistatic interactions influencing tumor load, including a suggestive interaction between Scc17 (tumor number) and Scc21, and a significant interaction between Scc18 and Scc19 (sex-dependent tumor load).

Limitations and Future Work

  • Low-effect alleles and replication challenges: The authors acknowledge that the small effect of many modifiers makes replication difficult and may contribute to Type I errors in some studies. The lack of concordance across different mapping approaches (intrasubspecific linkage, association, and intersubspecific linkage) underscores this challenge.
  • Need for larger studies for interactions: While suggestive epistatic interactions were identified, the authors note that larger studies will be required to validate these findings and to detect other potential interactions.
  • Identifying the underlying genes: The current study identifies the chromosomal locations of these susceptibility loci. Future work will focus on fine-mapping these regions and identifying the specific genes responsible for these effects. This will be crucial for understanding the biological mechanisms underlying CRC susceptibility and for translating these findings to human genetics.
  • Human relevance: While mouse models are powerful, the authors note that none of the identified mouse loci appear to be in conserved syntenic segments with human loci detected in genome-wide association studies. However, they suggest that these mouse loci may function in similar pathways or become detectable in humans when allele interactions are considered.

Summary updated

Abstract

Colorectal cancer (CRC) has a complex etiology resulting from the combination of multiple genetic and environmental factors, each with small effects. Interactions among susceptibility modifier loci make many of the loci difficult to detect in human genome-wide association studies. Previous analyses in mice have used classical inbred strains, which share large portions of their genomes due to common ancestry. Herein, we used an interspecific backcross between the Mus musculus strain A/J and the Mus spretus strain SPRET/EiJ to map 6 additional CRC modifier loci (Scc16-21) and 2 suggestive loci. Three loci modify the location of tumors along the proximal-distal axis of the colon. Six CRC modifiers previously mapped in intraspecific crosses were also replicated. This work confirms genetic models suggesting that CRC is caused by many small effect alleles and brings the catalog of reported CRC modifier loci to 23 spread across 13 chromosomes. Furthermore, this work provides the foundation for large population-level epistatic interaction tests to identify combinations of low effect alleles that may have large effects on CRC susceptibility.
BibTeX
@inproceedings{eversley2012mapping,
  author = {Chevonne D. Eversley and Yuying Xie and R. Scott Pearsall and David W. Threadgill},
  title = {{Mapping Six New Susceptibility to Colon Cancer (<i>Scc</i>) Loci Using a Mouse Interspecific Backcross}},
  booktitle = {G3},
  year = {2012},
  doi = {10.1534/g3.112.002253},
}
Figure