Chromosomes with DRT/DMC were detected in five of seven tumor
cell lines and five of 13 primary tumor samples, suggesting
that chromosomes with DRT/DMC are common in tumor cells in vitro
and in vivo
). In addition, following exposure to IR, chromosomes
with DRT/DMC were detected in as many as 25% of surviving cells,
indicating that DRT/DMC is also quite common in cells exposed
to IR (9
). Furthermore, IR-induced chromosomes with DRT/DMC
in primary cells, SV40 immortalized cells and tumor-derived
cells at similar frequencies, indicating that the DRT/DMC phenotype
is not restricted to cells with unregulated cell cycles or defective
cell-cycle checkpoints (9
). Here, we show that
10% of random
chromosome translocations generated using a chromosome engineering
strategy display the DRT/DMC phenotype. These results are consistent
with our previous observations that
5% of chromosome translocations
induced by IR display DRT/DMC (9
). In addition, we found that
two distinct methods for generating the same translocations,
site-specific homologous recombination or NHEJ, give rise to
DRT/DMC, but only on certain derivative chromosomes. Moreover,
the majority of chromosome translocations do not display DRT/DMC,
indicating that DRT/DMC is not the result of a stochastic process
that could occur at any translocation breakpoint. Taken together,
our data suggest that
5–10% of all chromosome translocations
induced by different mechanisms, site-specific recombination,
NHEJ or exposure to IR display DRT/DMC.
Genetic instability can occur at distinct levels. In most cancers,the instability occurs at the chromosome level, resulting ingains or losses of whole chromosomes (4). This type of instabilityis a dominant trait and is independent of p53 mutations (18).Another common form of genetic instability found in cancer cellsis characterized by the generation of frequent chromosome rearrangements,including the formation of marker chromosomes and gene amplifications.We believe that these two forms of genetic instability are distinct,and refer to the process that results in gains or losses ofintact chromosomes as chromosome instability (CIN) (4) and tothe process that results in the generation of frequent translocationsand rearrangements as translocation instability (TIN). Unfortunately,the molecular mechanisms responsible for either CIN or TIN intumor cells are still poorly understood (4,19).
Another example of genetic instability found in mammalian cellsoccurs at a delayed time following exposure to IR (20,21). Thisdelayed or persistent chromosomal instability occurs in vitroand in vivo and is characterized by the appearance of new chromosometranslocations and rearrangements in subsequent generationsafter the initial exposure to IR (22). Interestingly, many aspectsof this delayed instability cannot be explained by simple mutationalinactivation of trans-acting factors. For example, sub-cloningexperiments of irradiated cells indicate that this type of genomicinstability is unevenly transmitted to sibling sub-clones andthat the chromosomal rearrangements that occur within the unstableclones are non-randomly distributed throughout the karyotype(23,24). So, alternative epigenetic and/or cis-acting mechanismshave been proposed to explain this poorly understood process(reviewed in 25,26).
Given our limited understanding of genomic instability in mammaliancell systems, it is currently not known whether the delayedchromosomal instability observed in irradiated cells and theTIN associated with cancer cells are caused by similar or distinctmechanisms. However, we previously showed that chromosomes withDRT/DMC are common in tumor cells in vitro and in vivo (8).In addition, we found that chromosomes with DRT/DMC were presentin as many as 25% of cells exposed to IR (9). In this study,we found that chromosomes with DRT/DMC have a 30–80-foldincrease in the rate at which new GCRs occur on the affectedchromosomes. Therefore, because chromosomes with DRT/DMC arecommon in tumor cells and in cells exposed to IR, we proposethat chromosomes with DRT/DMC represent a common source of thegenomic instability observed in cancer cells (i.e. TIN) andin cells exposed to IR (i.e. delayed chromosomal instability).In addition, because the DRT/DMC phenotype occurs only on certainderivative chromosomes, our data provide support for previousmodels that genomic instability is driven by a cis-acting mechanism(26).
DRT/DMC is regulated in cis
How do ICT events at specific chromosome locations result inthe DRT/DMC phenotype? This is an intriguing question giventhat the translocation breakpoints that induced DRT/DMC in thisstudy occurred within the plasmid cassette sequences and notwithin the chromosomal DNA. In addition, it appears that inter-chromosomalexchanges are required for the phenotype, as plasmid cassetteinsertions and small Cre-mediated intra-chromosomal deletionsat the same chromosome locations did not result in DRT/DMC.Furthermore, on certain balanced translocations, only one ofthe derivative chromosomes displayed the phenotype. Taken together,these observations indicate that the DRT/DMC phenotype is regulatedby a cis-acting mechanism that occurs following specific chromosomalexchange. However, it is important to point out that intra-chromosomalrearrangements, involving chromosomal DNA instead of plasmidDNA, may be capable of generating DRT/DMC. Regardless, we areconsidering two possibilities to explain how specific chromosomalexchanges can give rise to the DRT/DMC phenotype. First, itis possible that these chromosomal exchanges result in deletionor mutation of a cis element that normally establishes earlyreplication timing for the entire chromosome. Loss of this elementwould then result in delayed replication of the entire chromosome.Second, it is possible that these specific chromosomal exchangesgenerate dominant interfering elements that act in cis to delaynormal chromosome replication timing by some unknown mechanism.Although we cannot distinguish between these possibilities atthe present time, the chromosome engineering strategy describedhere, combined with ‘mixing and matching’ of loxP-taggedchromosomes and directing loxP sites to specific chromosomelocations, should provide for a system in which the molecularmechanisms responsible for the DRT/DMC phenotype can be elucidated.