DNA damage reshapes chromatin even after repair

DNA damage reshapes chromatin well beyond the immediate repair response. In human glioblastoma cells, the effects of widespread DNA breaks were still detectable two weeks later in genome folding, gene expression, DNA methylation and chemical marks on chromatin proteins.

This adds an important layer to the usual picture of DNA repair. When both strands of the double helix are broken, the cell reconnects the DNA ends and attempts to restore the molecule. But “repaired” does not necessarily mean “returned to its original state.” Repair can leave small sequence changes or larger chromosome rearrangements, and the new study shows that it can also be accompanied by longer-lasting changes in how the genome is organized and regulated.

Chromatin is DNA together with the proteins that package it inside the nucleus. That packaging is highly organized. Different regions of chromosomes form loops and spatial neighborhoods that help determine which genes can interact with regulatory elements and which remain less active. Changing this three-dimensional architecture can therefore alter cellular behavior without itself being a new DNA mutation.

Hundreds of breaks at once

The researchers used human glioblastoma stem-like cells and CRISPR–Cas9 to introduce many double-strand breaks at defined genomic locations. Two multi-targeting guide systems were designed around 142 and 483 predicted sites. Direct mapping ultimately detected 339 and 1,386 cut sites, respectively, because Cas9 also acted at similar lower-affinity sequences.

This does not literally reproduce the pattern of damage caused by radiotherapy or chemotherapy. Its advantage is experimental control: because the locations of the breaks are known, changes around them can be followed over time. The investigators compared the cells at three hours and two weeks after damage, with Cas9-expressing cells lacking the targeting guide used as controls.

They examined several layers of genome regulation at the same time. Three-dimensional chromatin contacts were mapped alongside RNA expression, DNA methylation and histone acetylation. The team also used targeted sequencing to identify mutations and large structural rearrangements.

At three hours, the most prominent architectural effect was an increase in relatively local chromatin interactions within genomic neighborhoods. Two weeks later, the pattern had shifted. The early local increase had subsided, while new long-range interactions within chromosomes and contacts between different chromosomes had become more prominent.

The response was therefore not simply a temporary disturbance around individual breaks. Genome organization continued to evolve after the initial damage.

Effects reached far beyond the cut sites

Changes were not confined to the immediate vicinity of a break. At three hours, altered gene expression could be detected as far as 20 megabases from the nearest cut site, and a substantial subset of those transcriptional changes remained two weeks later.

For the two targeting systems, 377 and 1,011 differentially expressed genes remained persistently up- or downregulated at the two-week time point. These represented 30.1% and 70.8% of the genes whose expression changed in the respective conditions.

DNA methylation and histone acetylation also shifted. Regions of the genome that normally occupy more distinct active and inactive spatial compartments showed increased intermingling over time.

The study also documented genetic consequences of repair. Deep sequencing around the breaks found small insertions and deletions at nearly every examined cut site. Larger events—including translocations, inversions, duplications and deletions—also appeared. Each type generally remained below a 5% frequency at any individual site, but large structural variants were found around most sequenced break sites.

That distinction matters. The experiment does not show a genome whose original DNA sequence was perfectly restored everywhere while a purely epigenetic “memory” remained behind. Genetic and non-genetic consequences occurred together. The central finding is that the impact of widespread DNA damage extended beyond local sequence repair to persistent changes in chromatin architecture and transcription.

What this could mean for cancer

The authors propose that such reorganization could contribute to cellular plasticity after severe genotoxic stress. That possibility is particularly relevant to cancer because radiotherapy and many anticancer drugs work, at least in part, by damaging DNA. Surviving tumor cells might therefore emerge not only with new mutations but also with altered regulatory states.

The study did not demonstrate that these chromatin changes cause treatment resistance in glioblastoma. Nor does it establish that a human tumor exposed to radiation will undergo the same pattern of genome reorganization.

There are additional limitations. CRISPR–Cas9 produced controlled breaks at selected genomic sequences, whereas radiation, chemotherapy and spontaneous genomic instability create far more stochastic patterns of damage. The experiments were performed in glioblastoma cell models, so it is not yet clear how broadly the phenomenon extends to other cancers or to nonmalignant cells. And the follow-up lasted two weeks: long enough to establish persistence, but not permanence.

The study therefore changes the question that follows DNA damage. It is not enough to ask whether the broken molecule was rejoined and which mutations were left by the repair process. Widespread damage may also reorganize the three-dimensional genome, leaving the surviving cell with a different regulatory program.