Impact of Sequence Divergence on Recombinational

DNA Double-Strand Break Repair in Mammalian Cells

Kristina M. Chapman , Megan M. Wilkey , Addison C. Goins , Barbara C. Waldman , and Alan S. Waldman
a b c b b
a
Biomedical Engineering, bBiological Sciences, cCardiovascular Technology
University of South Carolina, Columbia, SC

Abstract Experimental Strategy Recovery of Double-Strand Break Conclusions

We investigated the impact of sequence divergence on DNA double- Repair Events I-SceI

This doesnt work in terms of recombination

strand break repair occurring via homologous recombination (HR) in Colony Frequency: 6 x 10-6
cultured thymidine kinase deficient mouse fibroblasts. We designed Cell Experiment Colony Homeologous recombination: 1.7 x 10-7
and stably transfected cells with a DNA construct harboring a Colonies This works in terms of recombination, but
herpes tk gene (the recipient) that was rendered nonfunctional by
Line No.a Frequencyb not well
insertion of an oligonucleotide containing the recognition site for Colony Frequency: 5 x 10-6
endonuclease I-SceI. The construct also contained a closely linked Analyzed Homeologous recombination: 5 x 10-7
truncated donor tk sequence. The donor could potentially restore
This proved to work very well from our
function to the recipient gene via HR provoked by induction of a results
double-strand break (DSB) at the I-SceI site in the recipient. HR Colony Frequency: 2.7 x 10-5
events were recoverable by selection for tk-positive clones. The Homeologous recombination:1.8 x 10-5
donor sequence contained 33 mismatches relative to the recipient.
The mismatches were clustered, forming a localized segment of
a
Each experiment involved electroporation of 5*106 cells with
DNA sequence displaying 20% divergence relative to the recipient. 20 micrograms of an I-SceI expression plasmid
The mismatched segment was surrounded by regions of high b
Number HATR colonies divided by number of cells plated We concluded that homology search is not based on
homology. Our construct was engineered so that when the donor strand exchange and base-pairing initiated from a DNA
was aligned with the recipient, the DSB site in the recipient was terminus. Our new construct proved to work well
aligned opposite the mismatched segment. This design enabled us comparable to a schematic where the break is made
across from an area of homology that is next to one side of
to potentially capture HR repair events initiating between diverged
a mismatched area (with another area of homology on the
sequences. Previous work demonstrated that mammalian cells other side of the mismatch). The data from the experiment
fastidiously avoid HR between mismatched sequences. In the showed that there is little to no rejection of mismatch
current work, we asked the following question: can flanking regions
of high homology enable genetic exchange between highly diverged Representative PCR Analysis of incorporation due to flanking surrounding homology. This
surrounding homology stabilizes the two sequences, acting
sequences, or is the rejection of exchange between diverged
sequences predominant and not overcome by nearby homologous
Recovered Repair Events by holding the two sequences together. These results
further suggest that homologous sequences are paired
interactions. Our work contributes to a greater understanding of the prior to the formation of a double strand break.
mechanisms that maintain genome stability.
Double-Strand Break Repair
Substrate
(pKM1)
2.5 kb
Design
0.8 kb
Agarose gel
Double-Strand Break Repair Future Directions
Pathways

Our results indicated that neither end of a gene

conversion repair tract must be located within a
region of high homology as long as there are regions
of homology surrounding. To further this research it
The donor on pKM1 has 33 mismatches relative to the recipient, would be beneficial to test the degree of mismatch
and these are represented by the striped box that would still conclude in successful recombination
of the DNA after an induced break. This could be
tested in one of two ways. First the density of the
mismatches could be increased. In the previous
DNA Sequencing Results experiment, there were 33 mismatches within the 230

Figure 1. Non Homologous End Joining
Isolation of Cell Lines Containing a base pairs surrounded by 313 homologous base
pairs and 292 homologous base pairs. The number
Single Integrated Copy of Repair of mismatches within that sequence could be
increased to see if the density or location of the
Substrate pKM1 mismatch has an effect on recombination. The
second test that could be performed would be to
increase the length of the mismatch region. This
would help clarify how much homology is needed on
Figure 2. DSB Repair via Homologous either side of the mismatch region. By comparing
Recombination, Synthesis-Dependent
Strand Annealing Pathway: results of the two different experiments it could
further be determined if the length of the mismatch
region or the density of the mismatches is more
important when repairing a double strand break.

2.5 kb

Acknowledgement
This work was supported by NSF grant MCB:1157416
s Science Foundation to ASW and
from the National
0.8 kb BCW, as well as a Magellan Mini-Grant
to MW. We thank Kendall Potter for her role in
constructing pKM1.

K1B represents the sample of K1 loaded with BAM digestion

38 of the clones analyzed showed a transfer of mismatches to the recipient sequence, and for 13
K1H represents the sample of K1 loaded with HindIII digestion
of these clones, both ends of the repair tract were located within the highly diverged sequences