CH391L/S2013 Tanya Raymond Feb 13 2013

From OpenWetWare
Jump to navigationJump to search

Recombination-restarted replication makes inverted chromosome fusions at inverted repeats


Replication and Genome Instability

Stalled replication forks are a serious threat to eukaryotic cells, since replication forks that become stalled for long periods of time can undergo fork collapse when the replication machinery falls off, or the fork can break and form a DNA double-stranded break. Upon sensing that a replication fork has stalled, a signaling pathway is initiated in order to halt cell cycle progression, allowing time for the fork to restart without additional stress. Since eukaryotes have multiple origins, if the fork is unable to restart, often replication is completed by a second replication fork coming from the other direction or a dormant origin that has been initiated. If all else fails, the cellular machinery responsible for homologous recombination repairs the stalled fork. Sometimes, however, in the presence of homologous regions of DNA, the strand exchanges with an incorrect template which can lead to non-allelic homologous recombination (NAHR) causing gross chromosomal rearrangements (GCRs). This study used Schizosaccharomyces pombe fission yeast as a model organism to study how stalled replication forks can lead to chromosomal rearrangements when improperly repaired. [1,2,3]


The authors created strains of S. pombe with inducible replication fork blockages by inserting the replication terminator sequence RTS1. A protein called Rtf1 is a DNA binding protein that can induce fork arrest by blocking progression of the replisome in a uni-directional manner. The construct RuraR shown in Figure 2 contained two RTS1 sequences and the ura4+ gene between them, a second construct RuiuR was the same as RuraR with an additional inverted ura4+ gene and a spacer. Special probes, labeled pA and pB were used with a Southern blot to determine the amounts and types of chromosomal rearrangements upon induced blockage in these strains. RuiuR was found to have tenfold higher occurrence of non-allelic homologous recombination events than RuraR.

Rpal1 and Rpal2 were two other strains that replaced the centromere proximal portion of ura4+ with his3+ of 0.2 kb and 1.8 kb respectively. Strain Tpal1R and Tpal2R replaced the tel proximal RTS1 with TER2/3 two sequences that cause pausing but not blockage of polymerase. The authors predicted that Rpal1/2 would have reduced levels of GCR and that Tpal1R/2R would have no GCR that occurred. In contrast they found that all of the constructs had some GCR occur, but that strains with double RTS1 sites had higher levels ~30%. In Rpal1/2, however, the sizes of the observed fragments were unexpected. Instead of corresponding to a NAHR event between the two RST1 site was consistent with double the distance between the BglII site and the center of the palindromic sequence. This suggested that, after restart, the polymerase did a “U-turn” resulting in an isodicentric chromosome. Tpal1R and Tpal2R only had one copy of RST1 and were therefore unable to perform NAHR, and were only able to "U-turn" for its chromosomal rearrangements.

In order to investigate this behavior further, a number of constructs were made containing palindromic sequences of varying length with spacers inside the palindrome of varying lengths to observe the effects on likelihood of U-turn occurrence while maintaining the center of symmetry the same distance from the stalled fork. Longer palindromes and smaller spacers had higher occurrence of U-turns. A third experiment using another series of constructs of the same palindrome and spacer length explored the idea that the center of symmetry was important for the likelihood of U-turn occurrence. The most U-turns occurred when the palindrome was directly next to the stall, suggesting that replication restart is initially very error prone, but upon “maturation” is better able to replicate without error although the observed rates of error are still higher than those normally observed. It is important to note that no DSB intermediate was formed for any of these replication forks.

Model for Recombination Restart Repair of Stalled Replication Fork Adapted from [1]
A. Cartoon of RuiuR and RuraR Cen3 indicates the centromere, tel indicates telomere, and red boxes indicate the RNA probe site. B is the BglII restriction sites. B. Southern blot for each strain with and without induced blockage C. Quantification of rearrangements[1]
A. Yeast constructs. B. Southern blot for each strain with and without induced blockage C. Quantification of rearrangements[1]


Stalled replication forks that are repaired erroneously are a major source of global chromosomal rearrangements in eukaryotic cells. Large chromosomal rearrangements like those caused by collapsed replication forks pave the way for tumorigenesis, since an overabundance of this kind of improper repair can cause genome instability. Often, chromosomal rearrangement causes the activation of oncogenes and inactivation of tumor suppressors disturbing the delicate balance between controlled cell growth and the uncontrolled growth of cancerous cells. Understanding the mechanism of replication fork-initiated chromosomal rearrangements is paramount to uncovering how genome instability can occur. This study was able to discover characteristics of a novel type of chromosomal inversion and showed that palindromic sequences of certain length and distance from the fork were more prone to cause GCR. This is not a rare event in these constructs, and the authors data showed that U-turns occurred with a 1 in 40 event frequency between inverted repeats. This U-turn event could explain some of the chromosomal inversions and copy number variations seen in human cancers.


1. Mizuno, Ken’Ichi et al. Recombination-restarted replication makes inverted chromosome fusions at inverted repeats. Nature, 2013. 493:246–249 doi:10.1038/nature11676

2. Sabatinos, S. A.  Recovering a Stalled Replication Fork. Nature Education, 2010. 3(9):31

3. Petermann, Eva et al. Hydroxyurea-Stalled Replication Forks Become Progressively Inactivated and Require Two Different RAD51-Mediated Pathways for Restart and Repair. Mol Cell. 2010 February 26; 37(4): 492–502.