How Common Is Parallel Intron Gain? Rapid Evolution Versus Independent Creation in Recently Created Introns in Daphnia

The evolutionary history of the spliceosomal introns that interrupt nuclear genes in eukaryotes has been debated for four decades. Positions of introns show a high degree of coincidence between various eukaryotes, implying either than many modern introns are very old and/or that intron creation is highly biased toward certain sites, leading to rampant parallel intron gain. A series of articles in this and other journals reported evidence for a strikingly high degree of parallel insertion of introns in different alleles of the water flea Daphnia pulex. Here, I report several new analyses of these data. Among the 23 loci reported to be undergoing parallel intron gain, I find that in five cases introns reported to be unrelated show extended sequence similarity strongly suggesting that the introns are in fact homologous. Five additional cases show extended conserved motifs between allegedly unrelated introns. For nearly all loci including the 13 remaining loci, at least one intron shows hallmarks of rapid sequence evolution, thwarting confident inference about homology. In addition, I reanalyze gene trees reconstructed from flanking exonic sequences, claimed by the original authors as additional evidence for parallel gain. I show that these phylogenetic trees frequently fail to recover expected relationships, and in any case show relationships not consistent with parallel intron gains. In total, I conclude that the data do not support widespread parallel intron gain in D. pulex. These findings strengthen the notion that shared intron positions generally reflect ancestral introns, and thus the notion of complex genes in early eukaryotes.