Template switching between the leading and lagging strands at replication forks generates inverted copy number variants through hairpin-capped extrachromosomal DNA

Abstract

Inherited and de novo copy number variants (CNVs) are increasingly found to be correlated with human developmental and cancerous phenotypes. Several models for template switching during replication have been proposed to explain the generation of these gross chromosomal rearrangements. We proposed a model of template switching (ODIRA--origin dependent inverted repeat amplification) in which simultaneous ligation of the leading and lagging strands at diverging replication forks could generate segmental inverted triplications through an extrachromosomal inverted circular intermediate. Here, we created a genetic assay using split-ura3 cassettes to trap the proposed inverted intermediate. However, instead of recovering circular inverted intermediates, we found inverted linear chromosomal fragments ending in native telomeres--suggesting that a template switch had occurred at the centromere-proximal fork of a replication bubble. As telomeric inverted hairpin fragments can also be created through double strand breaks we tested whether replication errors or repair of double stranded DNA breaks were the most likely initiating event. The results from CRISPR/Cas9 cleavage experiments and growth in the replication inhibitor hydroxyurea indicate that it is a replication error, not a double stranded break that creates the inverted junctions. Since inverted amplicons of the SUL1 gene occur during long-term growth in sulfate-limited chemostats, we sequenced evolved populations to look for evidence of linear intermediates formed by an error in replication. All of the data are compatible with a two-step version of the ODIRA model in which sequential template switching at short inverted repeats between the leading and lagging strands at replication forks generates inverted interstitial triplications. AUTHOR SUMMARYChromosomal rearrangements are a potent source of genetic variation in humans and other organisms. One specific type of rearrangement involves the increase in copies of segments of the genome. The variation in gene dosage that these rearrangements can cause has been associated with a wide range of neurological and other human disorders. A specific puzzling form of copy number increase consists of three tandem copies with the central copy in inverted orientation. How this rearrangement occurs is of great interest, yet the mechanisms responsible are only inferred by examining the sequence of final inverted products. Yeast provides a unique model system to explore the underlying molecular defects that give rise to inverted triplications. While the favored hypothesis suggests that double stranded DNA repair is the causative agent, we find that a particular form of template switching between strands at the replication fork, not a double stranded DNA break, is the initiating event. Using the awesome power of yeast genetics, we provide evidence in two different assays for this unique replication error that we call ODIRA (for Origin Dependent Inverted Repeat Amplification) and propose that it can also explain this form of copy number variant seen in human evolution and disease.