Genomic Stability

Low copy repeats (LCRs) are typically chromosome-locus-specific sequences sharing greater than 95% sequence identity over a range of 10 to 400 kilobases. These repeats make up at least 5% of the human genome (1). Meiotic recombination between these repeats results in translocation, deletion, duplication and / or inversion of the chromosomal regions flanked by the repeats. Several human diseases are the direct result of this process (reviewed in (2)) including Smith-Magenis syndrome (3), Charcot-Marie-Tooth disease (4), DiGeorge / velocardiofacial syndrome (5) and the inversion form of severe hemophilia A (6, 7). In all disease cases the recombination event between repeats necessarily involved crossing-over. Since homologous recombination between repeated sequences is a major form of mitotic double-strand break repair (8), crossing-over between LCRs mitotically would have the potential for significant genome destabilization.

The evidence that crossing-over can happen as the result of mitotic repair is much less direct than for meiotic repair, although a case report of somatic mosaicism in a female hemophilia inversion carrier is highly suggestive (9). In order to directly assay mitotic crossing-over as a double-strand break repair mechanism, I created the XO-GFP reporter gene (figure 1).

Figure 1: XO-GFP crossing over reporter

curved arrow:

transcription promoter

lightning bolt:

I-SceI cleavage site

red arrow:

drug resistance gene

thick bars:

repeated sequences

The reporter consists of two truncated coding sequences for the green fluorescent protein (GFP), one truncated at the 3' end (GFPD3') and the other at the 5' end (SceGFPD5'), inversely oriented with respect to each other (figure 1a). Neither sequence is capable of producing a functional protein. Additionally, the SceGFPD5' sequence contains an 18 basepair recognition site for the homing endonuclease I-SceI (10). An intervening puromycin resistance gene allows selection for stable chromosomal integration. Expression of the endonuclease generates a double-strand break in SceGFPD5', stimulating homologous recombination with the GFPD3' sequence. If the recombination occurs without crossing-over, the I-SceI recognition site is gene converted to GFP sequences, using GFPD3' as a template, changing SceGFPD5' into GFPD5', a 5' truncated, non-functional GFP gene without an I-SceI recognition site (figure 1b). In contrast, if recombination occurs with crossing-over, the I-SceI recognition site in SceGFPD5' is again gene converted to GFP sequences, but the cross-over event also inverts all intervening sequences between the repeats. This inversion brings the front half of GFPD3' into register with the back half of the SceGFPD5' yielding a functional GFP gene (figure 1c). Additionally a doubly-truncated internal GFP fragment (iGFP) is created and the orientation of the drug resistance marker is inverted. Cells in which this crossing-over occurs then constitutively express GFP and are easily detectable by flow cytometry.

I have integrated the XO-GFP reporter stably into the genome of mouse embryonic stem cells and detected GFP+ cells in response to I-SceI transfection (11). Southern blot analysis of these cells directly demonstrated inversion of the drug resistance marker, confirming mitotic crossing-over in response to a double-strand break. Preliminary evidence indicates that the bias between non-cross-over repair and cross-over repair is on the order of 30 : 1 in favor of non-cross-overs. I will explore the genetic factors responsible for this bias by co-transfecting expression vectors of various DNA repair proteins such as the Bloom's syndrome protein and by reducing expression of these proteins by co-transfection of small inhibitory double-stranded RNA sequences (12).

Of all classes of homologous recombination mediated double-strand break repair, those involving crossing-over have the most dire consequences for genome stability. Crossing-over between either direct or inverted repeats on different chromosomes results in reciprocal translocations. Crossing-over between direct repeats on the same molecule results in deletion of all intervening sequences as an episomal circle, while crossing-over between inverted repeats on the same molecule inverts all intervening sequences as described above. Crossing over between direct repeats on a sister chromatid results in deletions and / or expansions. Finally, crossing over between inverted repeats on sister chromatids leads to the obligate formation of both an acentric and a dicentric chromatid. This latter possibility is especially serious as it can result in prolonged breakage-fusion-bridge cycles (13) which have been identified as an important pathway for the generation of tumor genetic heterogeneity (14). The XO-GFP reporter has the potential to rapidly identify cells in which this process is actively occurring, enabling study of possibly very early events in the cellular progression to cancer.

  1. E. E. Eichler, Trends Genet 17, 661-9. (2001).
  2. P. Stankiewicz, J. R. Lupski, Trends Genet 18, 74-82. (2002).
  3. K. S. Chen et al., Nat Genet 17, 154-63. (1997).
  4. L. Pentao, C. A. Wise, A. C. Chinault, P. I. Patel, J. R. Lupski, Nat Genet 2, 292-300. (1992).
  5. L. Edelmann, R. K. Pandita, B. E. Morrow, Am J Hum Genet 64, 1076-86. (1999).
  6. J. Naylor, A. Brinke, S. Hassock, P. M. Green, F. Giannelli, Hum Mol Genet 2, 1773-8. (1993).
  7. D. Lakich, H. H. Kazazian, Jr., S. E. Antonarakis, J. Gitschier, Nat Genet 5, 236-41. (1993).
  8. F. Liang, M. Han, P. J. Romanienko, M. Jasin, Proc Natl Acad Sci U S A 95, 5172-7. (1998).
  9. J. Oldenburg et al., Blood 96, 2905-6 (2000).
  10. M. Jasin, Trends Genet 12, 224-8. (1996).
  11. A. J. Pierce, M. Jasin, unpublished observations.
  12. S. M. Elbashir et al., Nature 411, 494-8. (2001).
  13. B. McClintock, Cold Spring Harbor Symp Quant Biol 16, 13-47 (1951).
  14. D. Gisselsson et al., Proc Natl Acad Sci U S A 97, 5357-62. (2000).