A fine-scale recombination map of the human genome enables sequence-level analysis of recombination-based mutagenesis

Meiotic recombination and de novo variants underlie genetic diversity and evolution. To investigate the factors involved in recombination and the genetic changes that result, researchers need high-resolution maps of the locations at which crossovers occur. In a recent article in Science (https://doi.org/10.1126/science.aau1043), Halldorsson et al. used data from single-nucleotide polymorphism (SNP) microarrays and genome sequencing to create a fine-scale map of crossover events in the human genome and to evaluate whether de novo variants occur more frequently near sites of recombination. Using the Icelandic genealogical database, the authors defined the boundaries of 761,981 meiotic crossovers to an average resolution of 682 base pairs. They found that crossovers occurred within recombination hotspots for 71.1% of maternal and 74.9% of paternal crossovers and that crossovers occurred more frequently in regulatory regions of the genome compared with coding regions. Complex crossovers—crossovers that include a gene conversion event within 100 kilobases (kb)—comprised 1.24% of maternal and 0.53% of paternal crossovers. In maternal meioses, the rate of recombination and the fraction of complex crossovers increased with age. Crossovers in older mothers also shifted toward locations with lower guanine–cytosine (GC) content, later-replicating regions, and regions enriched for C-to-G transversions. Genome sequence data from 2976 parent–child trios revealed 200,435 de novo variants. An age-related increase in de novo variants was observed for both sexes, with an increase of 0.38 variants per maternal year at birth and 1.39 variants per paternal year at birth. The rate of de novo variants within 1 kb of crossovers was 58.4 times the genome-wide average for maternal crossovers and 41.5 times the genome-wide average for paternal crossovers, with maternal variants consisting largely of C-to-thymine (T) transitions outside of CG dinucleotides and paternal variants consisting largely of C-to-T transitions within CG dinucleotides. At 1–3 kb from crossovers, the rate of de novo variants was 11.9 times the genome-wide average for maternal crossovers and 6.9 times the genome-wide average for paternal crossovers. At 3–40 kb away from crossover events, only maternal crossovers demonstrated an increased rate of de novo variants, which was 2.2 times the genome-wide average and occurred in the context of complex crossovers. The team linked 35 loci to variability in the rate of recombination and the localization of crossovers. The genes at many of these loci encode proteins involved in the formation or composition of the synaptonemal complex. The authors conclude that crossovers are mutagenic, that some features of recombination and recombination-related mutagenesis differ between males and females and according to age, and that the synaptonemal complex is an important regulator of meiotic recombination. —Raye Alford, News Editor

Credit: PhotoDisc/Getty

Variation in gene expression and splicing as a result of intronic copy-number variations

Introns comprise a much larger proportion of the human genome than exons. They vary substantially in length and harbor noncoding functional RNA genes, regulatory elements, and pathogenic variants. Yet introns tend to be one of the last places that scientists look for disease-related genetic alterations. In a recent article in PLOS Genetics (https://doi.org/10.1371/journal.pgen.1007902), Rigau et al. characterized the structural variation of introns and investigated the impact of intronic copy-number variants (CNVs) on gene expression and RNA splicing. The research team used five published data sets to evaluate CNVs in protein coding genes and found that, compared with CNVs overlapping coding sequences, intronic CNVs—CNVs wholly contained within introns—were the most frequent type, comprising 63% of the CNVs. Of 12,986 intronic CNVs identified, 12,334 (95%) were losses. Despite their lack of coding sequence, however, intronic losses occurred significantly less often than deletions in similarly sized intergenic regions, suggesting selective pressure to maintain the size and sequence of introns. The team further assessed the relative frequency of coding and intronic deletions based on the evolutionary age of genes and found that losses involving the coding regions of genes were more prevalent in primate-specific genes, i.e., recent genes, compared with ancient genes. In contrast, intronic losses showed a more even distribution throughout evolution. Among essential genes, the authors found intronic losses in 1154 genes, contrary to expectations that essential genes, which tend to have shorter introns and be evolutionarily older, would be depleted of deletions. Intronic losses were also found in 1638 disease-related genes. To assess the impact of intronic losses on gene expression and splicing, the authors evaluated RNA sequencing data from lymphoblastoid cell lines; the cell lines were derived from 445 individuals for whom CNV data were also available. For 53 of 1505 genes with intronic losses, the team found significant variances in expression, with 38% of the genes upregulated and 62% downregulated. For 185 genes with intronic deletions, the team identified 217 differentially expressed transcripts. They also determined that intronic losses significantly reduced intronic guanine–cytosine (GC) content, suggesting that intronic losses could impact splicing through alteration of intron length, reduction of GC content, or both. Additionally, the authors found 12 genes whose expression was significantly altered—and 65 genes whose splicing was disturbed—due to intronic losses in distant genes, suggesting a long-range effect. The authors conclude that intronic CNVs contribute to variation among humans through modulation of the size and expression of genes and the splicing of RNAs. They suggest that intronic CNVs contribute to human evolution and that future studies of disease genes would benefit from evaluation of intronic variation. —Raye Alford, News Editor

Credit: iStock/Getty