Allele-specific expression of APC in adenomatous polyposis families

Abstract

Background & aims: Germline mutations in the APC gene cause of most cases of familial adenomatous polyposis (FAP) and a lesser proportion of attenuated FAP (AFAP). Systematic analysis of APC at the RNA level could provide insight into the pathogenicity of identified mutations and the molecular basis of FAP/AFAP in families without identifiable mutations. Here, we analyzed the prevalence of imbalances in the allelic expression of APC in polyposis families with germline mutations in the gene and without detectable mutations in APC and/or MUTYH.

Methods: Allele-specific expression (ASE) was determined by single nucleotide primer extension using an exon 11 polymorphism as an allele-specific marker. In total, 52 APC-mutation-positive (36 families) and 24 APC/MUTYH-mutation-negative (23 families) informative patients were analyzed. Seventy-six controls also were included.

Results: Of the APC-mutation-positive families, most of those in whom the mutation was located before the last exon of the gene (12 of 14) had ASE imbalance, which is consistent with a mechanism of nonsense-mediated decay. Of the APC/MUTYH-mutation-negative families, 2 (9%) had ASE imbalance, which might cause the disease. Normal allele expression was restored shortly after lymphocytes were cultured with puromycin, supporting a 'nonsense-mediated' hypothesis.

Conclusions: ASE analysis might be used to determine the pathogenesis of some cases of FAP and AFAP in which APC mutations are not found. ASE also might be used to prioritize the order in which different areas of APC are tested. RNA-level studies are important for the molecular diagnosis of FAP.

Figure 1. SNaPshot analysis of rs2229992 SNP as a quantitative test

Regression analysis of the amplification products of two homozygous control cDNAs mixed in different proportions and analyzed in triplicate by SNaPshot. Allele frequencies were calculated by the peak heights: freq. C = C/(C+kT), where the correction factor k is determined from the mix simulating an actual allele frequency of 0.5 (5:5). Measured allele frequencies were plotted against the expected values. A near-linear relationship was observed across all data points (R2 = 0.9983).

Figure 2. Quantification of ASE in APC(+) polyposis patients

A. Distribution of APC ASE in controls, non-carriers and carriers of a pathogenic APC mutation from the ICO and UM subsets. ASE values are the proportion between the C and T allele frequencies of the APC coding SNP rs2229992. Each point represents the mean of the triplicate measurements, and each replicate was performed for all samples in each subset. Allelic ratios ranged from 0.646 to 8.706. The normal ASE cut-off range is indicated by dashed horizontal lines (at 1.168 and 0.836) and determined by control ASE values, as indicated in Patients and Methods. Three sample electropherograms obtained from the SNaPshot reaction, and their corresponding ASE values, are also shown. B. Box plots of ASE values for the location of the mutation in the APC gene. Patients were stratified into two groups by mutation site (“Outside e15”, mutation located from exons 1-14; and “Inside e15”, mutation located in exon 15). To give an overview of our results, ASE values are represented as the proportion of allelic frequencies between the underexpressed and the overexpressed allele in each sample, independently if whether it is C or T. The interquartile range includes 50% of the samples and is shown by white boxes. The interdecile range includes 90% of the patients and is shown as whiskers. Outliers are shown by empty dots. Mutated groups were compared using the non-parametric Mann-Whitney test.

Figure 3. Quantification of ASE in APC(−)/MUTYH(−) polyposis patients

Distribution of APC ASE in controls and polyposis patients with no previously identified mutation. ASE values are the proportion between the C and T allele frequencies of the APC coding SNP rs2229992. Each point represents the mean of the triplicate measurements, and each replicate was performed for all samples in each subset. Allelic ratios ranged from 0.756 to 1.284. The normal ASE cut-off range is indicated by dashed horizontal lines (at 1.168 and 0.836).

Figure 4

Proposed diagnostic algorithm for APC mutation screening including ASE measurement in polyposis families.