Haplotype structure and selection of the MDM2 oncogene in humans

Abstract

The MDM2 protein is an ubiquitin ligase that plays a critical role in regulating the levels and activity of the p53 protein, which is a central tumor suppressor. A SNP in the human MDM2 gene (SNP309 T/G) occurs at frequencies dependent on demographic history and has been shown to have important differential effects on the activity of the MDM2 and p53 proteins and to associate with altered risk for the development of several cancers. In this report, the haplotype structure of the MDM2 gene is determined by using 14 different SNPs across the gene from three different population samples: Caucasians, African Americans, and the Ashkenazi Jewish ethnic group. The results presented in this report indicate that there is a substantially reduced variability of the deleterious SNP309 G allele haplotype in all three populations studied, whereas multiple common T allele haplotypes were found in all three populations. This observation, coupled with the relatively high frequency of the G allele haplotype in both and Caucasian and Ashkenazi Jewish population data sets, suggests that this haplotype could have undergone a recent positive selection sweep. An entropy-based selection test is presented that explicitly takes into account the correlations between different SNPs, and the analysis of MDM2 reveals a significant departure from the standard assumptions of selective neutrality.

Fig. 1.

Schematic diagram of the MDM2 gene and the SNPs genotyped in the present study. The three SNPs common to both the Celera and Sheba data sets are indicated by the dotted lines.

Fig. 2.

Inferred haplotype frequencies in the Caucasian population.

Fig. 3.

Inferred haplotype frequencies in the African American population.

Fig. 4.

Inferred haplotype frequencies in the Ashkenazi Jewish population.

Fig. 5.

Pairwise linkage disequilibrium for Ashkenazi Jewish (a), Caucasian (b), and African American (c) populations. The upper right triangle reports the |D′| measure, and the lower left triangle reports the r² measure. SNP309 is highlighted in all populations.

Fig. 6.

Monte Carlo simulation of haplotypes. (a) The conditional multientropy of the adjacent SNPs is plotted for a particular locus where the minor allele (lower curve) had a frequency of 0.19 and the major allele (upper curve) had a frequency of 0.81. The unit of distance is measured in rescaled base pairs. (b) Plot of the DLE values versus frequency of each allele. One thousand Monte Carlo genome samples of size 250,000 base pairs were generated. The mutation rate was set to produce 8,000 SNPs, and the recombination rate was set to 1 cM Mb⁻¹. (c) Selected allele (indicated by arrow) underwent a selective advantage per copy per generation of 0.1