Introns were originally thought to be ‘junk DNA’ without function but accumulating evidence has shown that they can have important functions in the regulation of gene expression. In humans and other mammals, introns can be extraordinarily large and together they account for the majority of the sequence in human protein-coding loci. However, little is known about their structural variation in human populations and the potential functional impact of this genomic variation. To address this, we have studied how copy number variants (CNVs) differentially affect exonic and intronic sequences of protein-coding genes. Using five different CNV maps, we found that CNV gains and losses are consistently underrepresented in coding regions. However, we found purely intronic losses in protein-coding genes more frequently than expected by chance, even in essential genes. Following a phylogenetic approach, we dissected how CNV losses differentially affect genes depending on their evolutionary age. Evolutionarily young genes frequently overlap with deletions that partially or entirely eliminate their coding sequence, while in evolutionary ancient genes the losses of intronic DNA are the most frequent CNV type. A detailed characterisation of these events showed that the loss of intronic sequence can be associated with significant differences in gene length and expression levels in the population. In summary, we show that genomic variation is shaping gene evolution in different ways depending on the age and function of genes. CNVs affecting introns can exert an important role in maintaining the variability of gene expression in human populations, a variability that could be related with human adaptation.

Source: Widespread population variability of intron size in evolutionary old genes: implications for gene expression variability | bioRxiv