Whole-genome sequencing in autism identifies hot spots for de novo germline mutation

Abstract

De novo mutation plays an important role in autism spectrum disorders (ASDs). Notably, pathogenic copy number variants (CNVs) are characterized by high mutation rates. We hypothesize that hypermutability is a property of ASD genes and may also include nucleotide-substitution hot spots. We investigated global patterns of germline mutation by whole-genome sequencing of monozygotic twins concordant for ASD and their parents. Mutation rates varied widely throughout the genome (by 100-fold) and could be explained by intrinsic characteristics of DNA sequence and chromatin structure. Dense clusters of mutations within individual genomes were attributable to compound mutation or gene conversion. Hypermutability was a characteristic of genes involved in ASD and other diseases. In addition, genes impacted by mutations in this study were associated with ASD in independent exome-sequencing data sets. Our findings suggest that regional hypermutation is a significant factor shaping patterns of genetic variation and disease risk in humans.

Figure 1. Paternal age effect explains 44% of variation in genome-wide mutation rates

Data points represent the total number of autosomal DNMs detected in offspring. See also Supplemental Figure 1 and Supplemental Table 1.

Figure 2. Non-random distribution of DNMs in the genome

Quantile-quantile plots of the observed distribution of inter-DNM distances within and between individuals and the expected distribution based on a random mutation model. Differences are statistically significant at α=0.05 by the KS test. See also Supplemental Figure 4 and Supplemental Table 2.

Figure 3. Individual associations of genome features with de novo mutation

Predisposition to de novo mutation is influenced by sequence and chromatin characteristics. A variety of quantitative genome data were tested for associations with de novo mutation sites, including conservation, DNase hypersensitivity, GC content, histone marks, lamin B1 association, nucleosome occupancy, recombination rate, replication timing, transcription in human embryonic stem cells, simple repeats at the site of DNM, and the particular trinucleotide sequence centered at the site of DNM. The data were tested for association at different scales (i.e. window sizes at which the genome data were averaged), indicated on the x-axis. The strength and direction of association between the features and DNM are indicated by logistic regression coefficients (y-axis), which are shown with their standard errors. Significant associations (FDR < 0.10) are indicated in bold type. A summary of the relationship between these features and the principal components used in the predictive model is provided in Supplemental Figure 6. A detailed legend of the feature names and their descriptions is provided in Supplemental Table 6, and further details relating to the origin and construction of the features can be found in methods.

Figure 4. Intrinsic characteristics of the genome explain variation in observed mutation rates

Mutability index at the site level (1 bp) is highly predictive of the mutation rate in (A) ASD genomes in this study, (B) Control genomes in Conrad et al., 2011, and in ASD cases (C) and controls (D) of previous exome studies (combined data from O’Roak et al. 2011 and , Iossifov et al., 2012, Sanders et al., 2012, and Neale et al., 2012). Mutability index explains a majority of the variability in site specific mutation rates, and the degree of mutation rate variation was similar in cases and controls. CpG sites and non-CpG sites varied widely in their mutability and the range of CpG mutability overlapped considerably with the range for non-CpG sites (Supplemental Fig. 2). Mutability index was also predictive of regional mutation rates (Supplemental Fig. 3).

Figure 5. Landscape of mutability in the genome

(A) The 1 kb average mutability index (MI) across a 20 Mb genomic region of chromosome 8p21–23 indicates the existence of extended regions of hypermutability. “Hotspots” (red), “warm spots” (orange) as well as “cold spots” were defined by segmenting the MI scores using a 5 state HMM (see Supplemental Table 7). Predicted mutation rates (y-axis) were computed by multiplying the arithmetic mean MI by the baseline mutation rate of 10⁻⁸, then transforming to the log₁₀ scale. The genome- and exome-wide distributions of MI are depicted in Supplemental Figure 7. The locations of DNMs are also shown and include a dense cluster of DNMs from individual 74–0355 (red, DNMs < 100 kb apart marked by asterisk). (B) The lower panel displays segmentation results for a second genomic region at 15q11–13. This region is notable for having a high rate of recurrent structural mutation. In the same region, the predicted rate of nucleotide substitutions is highly elevated.

Figure 6. U-shaped relationship of mutability and evolutionary conservation

(A) Throughout the genome, we observe a correlation of hypermutability, hyperdivergence and hyperdiversity, consistent with previous studies. By contrast, in highly conserved regions the opposite trend is evident. MI and conservation were averaged in 1kb windows genome-wide. Windows were then binned according to percentiles of conservation. (B) Specifically within exons, there is a strong positive correlation of mutability and evolutionary conservation (also binned by percentiles of conservation). (C) The positive correlation between mutation rate and average exon conservation was confirmed by data from exome studies. Note that the positive relationship exists for both cases and controls. Under the null hypothesis, in which exons are hit with probability proportional to their length, this relationship is not observed.

Figure 7. Disease genes are characterized by high mutability

Disease genes are more mutable than non-disease genes (A) within genes and (B) within exons. In both cases, mutability is highest for genes involved in dominant disorders and mutability is increased to a lesser extent for genes involved recessive and polygenic traits. Mutability is significantly elevated for genes preferentially expressed in the brain (C- D) as well as genes involved in ASD (see methods for details). An asterisk indicates a significant difference compared to the respective background set (at α=0.01 by a two-sided t-test). See also Supplemental Table 3 and Supplemental Table 4 for the mean mutability index of exons and genes, respectively.