Contribution of retrotransposition to developmental disorders

Abstract

Mobile genetic Elements (MEs) are segments of DNA which can copy themselves and other transcribed sequences through the process of retrotransposition (RT). In humans several disorders have been attributed to RT, but the role of RT in severe developmental disorders (DD) has not yet been explored. Here we identify RT-derived events in 9738 exome sequenced trios with DD-affected probands. We ascertain 9 de novo MEs, 4 of which are likely causative of the patient's symptoms (0.04%), as well as 2 de novo gene retroduplications. Beyond identifying likely diagnostic RT events, we estimate genome-wide germline ME mutation rate and selective constraint and demonstrate that coding RT events have signatures of purifying selection equivalent to those of truncating mutations. Overall, our analysis represents a comprehensive interrogation of the impact of retrotransposition on protein coding genes and a framework for future evolutionary and disease studies.

Fig. 1

The DDD RT call set. a–e Histograms of the total number of variants per individual for the four classes of RT events identified in the DDD cohort (Alu: blue; L1: green; SVA: orange; PPGs: red; combined RT events: grey) in size one bins. f Allele frequency distributions for the RT classes depicted in a–e in log₁₀ allele frequency bins. g Insert size estimates provided by MELT for the MEI classes ascertained in this study in log₁₀ insert size bins. All plots only include variants from unaffected parents

Fig. 2

Coding constraint on MEIs. a Cumulative consequence annotations for Alu, L1, and SVA MEIs in all samples (n = 28,132 individuals) analyzed. The majority of variants identified in this study fell within the noncoding space (either an enhancer or intron). b Comparison of constraint between MEIs and SNVs in unaffected parents. To compare the impact of exonic and intronic Alu (blue) and all MEIs (grey) to varying classes of SNVs (black), we used two metrics: the proportion of variants in genes that have been identified as LoF intolerant as gauged by pLI-score (x-axis) and the proportion of variants identified in only one individual (i.e., singletons; y-axis). Error bars indicate 95% confidence intervals based on population proportion; confidence intervals were calculated for SNVs, but are too small to appear at the resolution displayed in this figure

Fig. 3

RT-derived de novos in the DDD. We identified a total of nine de novo MEIs, four of which disrupted the protein-coding sequence of a known DD gene: a SETD5, b NSD1, c MEF2C, and d ARID2. Shown in each panel is a diagram of the affected gene (blue model) with the relevant insertion indicated with a colored bubble. To the right are PCR validations confirming the de novo status of each mutation; a positive result is indicated by a raised secondary band present only in the proband sample (red arrow). e Circos diagram and PCR results for two identified germ-line de novo PPGs. For each de novo PPG shown is a diagram of the donor gene (gene model), location of duplication as PPG (directional arrow), and new insertion site. Exons from the donor gene included in the PPG are indicated by brackets underneath the donor gene model. To confirm PPG presence, PCR was performed (Methods) on proband, paternal, and maternal gDNA (sample in each lane is shown by pedigree). The band which represents the PPG is marked with a red arrow, and was confirmed via capillary sequencing (Supplementary Fig. 12). Dashed lines indicate intergenic regions, all genes models are shown in sense orientation, and PPG gene diagrams are not to scale

Fig. 4

Estimating enrichment of deleterious MEIs. Depicted are the total number of expected (black) and observed (red) de novo mutations observed in exons (a) and enhancers (b) for all, high pLI (pLI > 0.9), and known monoallelic DD (MA DDG2P) genes. Expectation is based on the Poisson distribution of 100 simulations utilizing the neutral mutation rate (1.2 × 10^–11 μ). P-values are based on the Poisson distribution and used to determine statistical deviation of observed to expected de novo counts for exons and enhancers