New insights into the generation and role of de novo mutations in health and disease

Abstract

Aside from inheriting half of the genome of each of our parents, we are born with a small number of novel mutations that occurred during gametogenesis and postzygotically. Recent genome and exome sequencing studies of parent-offspring trios have provided the first insights into the number and distribution of these de novo mutations in health and disease, pointing to risk factors that increase their number in the offspring. De novo mutations have been shown to be a major cause of severe early-onset genetic disorders such as intellectual disability, autism spectrum disorder, and other developmental diseases. In fact, the occurrence of novel mutations in each generation explains why these reproductively lethal disorders continue to occur in our population. Recent studies have also shown that de novo mutations are predominantly of paternal origin and that their number increases with advanced paternal age. Here, we review the recent literature on de novo mutations, covering their detection, biological characterization, and medical impact.

Fig. 1

Mechanisms of de novo mutations. De novo mutations can arise because of static properties of the genome, such as the underlying sequence (deamination of methylated CpGs, transitions versus transversions) or due to erroneous pairing of nucleotides during DNA replication. However, de novo mutations can also occur in relation to cell-specific properties such as the chromatin state, transcriptional status, and gene expression levels. Mutational hotspots for genomic rearrangements are largely determined by the underlying genomic architecture. One such example is given for non-allelic homologous recombination (NAHR). Arrows represent the influence of each feature on the de novo mutation rate. Green arrows pointing upwards indicate elevated mutability; red arrows pointing downwards indicate lower mutability. M methyl group modifying cytosine

Fig. 2

Timing of de novo mutations (DNMs). Sperm cells have undergone approximately 100 to 150 mitoses in a 20-year-old man, whereas oocytes have gone through 22 mitoses in a woman of the same age (left). As a result of errors in both replication of the genome and repair of DNA damage occurring during parental embryogenesis, gametogenesis, or as postzygotic events in the offspring, DNMs arise in each new generation. Advanced parental age is associated with an increase in the number of de novo mutations (right). The male germline adds 23 mitoses per year, entailing that a spermatogonial stem cell in a 40-year-old man has undergone more than 600 cell mitoses. Each additional year in paternal age at conception adds one to three de novo mutations to the genome of the offspring. Oogenesis has a fixed number of mitoses, but mutations accumulate over time possibly owing to failure to repair DNA damage. The increase in number of de novo mutations with maternal age is lower: 0.24 extra de novo mutations for each additional year of maternal age at conception. Cell lineages modified from [238]. Somatic cells are showed in orange, the male germline is shown in blue, and the female germline is shown in purple. Blue stars represent postzygotic mutations present in the germline and in somatic cells; yellow stars represent mutations arising exclusively in the germline; red stars represent somatic mutations arising during embryonic development or post-natal life which are absent from germline cells. Figure footnotes: ¹The ratio of paternal to maternal mutations originating from parental gonosomal mosaicism is 1:1; ²the ratio of paternal to maternal germline de novo mutations is 4:1; ³the ratio of paternal to maternal postzygotic de novo mutations is 1:1; ⁴this range is based on the average number of de novo mutations published elsewhere [9, 10, 12, 13, 15] irrespective of parental age

Box Fig. 1

Technical improvements to the detection of de novo mutations (DNMs). a Trio-based sequencing allows the identification of de novo mutations in an individual. b Increased sequencing coverage benefits the detection of de novo mutations (in blue). Low coverage (upper) reduces the probability that a de novo mutation will be sequenced and called, compared with high sequencing coverage (lower). c Using random tags or unique molecular identifiers (UMIs) decreases the number of false positives (in red) by making consensus calls from all reads with the same UMI. Furthermore, UMIs can be used to remove PCR-derived duplicate reads to determine accurately the allelic ratio. d Long sequencing reads improve mappability, even across difficult genomic regions such as those containing repeats (gray boxes). Additionally, long reads can be used to phase mutations (shown in blue and in green) and generate haplotypes, to help identify the parent of origin of a mutation. IV inherited variant.