DNA Strand Breaks and the Genome: The Bigger Picture
Accompanying document for the Genetics Board, 05/11/2026. It extends the transcript DNA_Breaks_Transcript.md.
Figures are approximate, drawn from the standard literature and given as ranges. They have not been re-checked against primary sources for this document, so verify any number before it is quoted externally.
1. Summary
- Every human cell is under continuous DNA damage. Single-strand breaks (SSBs) number in the tens of thousands per cell per day; double-strand breaks (DSBs) in the tens.
- The causes are mostly intrinsic: copying DNA, reading it, untangling it, and the chemistry of an oxygen-using cell. They are not the result of a genome that "outgrew" its design. Genome size scales the opportunity for breaks without being their root cause.
- Single-stranded DNA (ssDNA) is a normal but short-lived and hazardous state, and it is the link between minor lesions and major ones: stalled replication produces long ssDNA, which can collapse into a DSB.
- Where a break lands, and how it is repaired, decides the outcome. Breaks in non-coding DNA are less likely to remove a gene but can still disrupt regulation, chromosome structure, or, through repeats, cause large rearrangements.
- Repeats are both a source of function and a source of fragility.
2. Break types and how common they are
| Lesion |
Rough frequency (per cell per day) |
Notes |
| Base damage and abasic sites |
10,000 to 100,000 (estimates vary widely) |
Mostly oxidation and spontaneous chemistry |
| Single-strand breaks (SSBs) |
about 10,000 or more |
Often arise as repair intermediates |
| Double-strand breaks (DSBs) |
about 10 to 50 |
Higher in rapidly dividing cells; about 50 per cell cycle is commonly cited for cultured cells |
DSBs are the most dangerous per event, because both strands are cut and the broken ends can be lost or joined to the wrong partner. SSBs are far more numerous but are repaired quickly using the intact opposite strand as a template.
3. Where breaks come from
Replication. Forks stall at damage, tightly bound proteins, repeats and secondary structures. A collapsed fork leaves a one-ended DSB.
Topology. About two metres of DNA must be packed into the nucleus, copied and read. Topoisomerases relieve twisting by transient cutting. Type II enzymes cut both strands by design, and a cut that fails to reseal persists as a DSB. This is the real kernel of truth in the idea that size is a burden: length raises the topological workload.
Transcription. RNA-DNA hybrids (R-loops) leave the displaced strand single-stranded and exposed. Neuronal activity triggers programmed topoisomerase breaks that are part of normal gene switching.
Oxidative chemistry. Reactive oxygen species from metabolism damage bases and backbones directly. Two nearby single-strand lesions on opposite strands can add up to a DSB.
Programmed breaks. The cell makes some deliberately: V(D)J recombination (building antibody and T-cell receptor diversity), class-switch recombination, and meiotic recombination (made by the enzyme SPO11).
Mobile elements. Active LINE-1 elements cut DNA when they insert, a modest but real contribution.
External. Ionising radiation, topoisomerase-poisoning drugs, and some chemicals.
4. Where single-stranded DNA appears
ssDNA is never a stable genome form in human cells, but it occurs briefly in many places:
- the lagging strand template during replication, between Okazaki fragments
- the displaced strand of an R-loop, and the transcription bubble
- the 3' overhang at each telomere (roughly 50 to 300 bases)
- resected DSB ends, which form long ssDNA tails for homologous recombination
- gaps left during excision repair
- the mitochondrial D-loop
Exposed ssDNA is chemically fragile and prone to hairpins, so it is coated at once by RPA (replication protein A), which protects it and recruits the damage-response kinase ATR. Excess or prolonged ssDNA is itself a danger signal and a precursor to DSBs.
5. How cells repair breaks
| Pathway |
Handles |
Template? |
Accuracy |
| Base excision repair |
Small base damage, SSBs |
Yes (opposite strand) |
High |
| Nucleotide excision repair |
Bulky, helix-distorting lesions |
Yes |
High |
| Mismatch repair |
Copying errors |
Yes |
High |
| Non-homologous end joining (NHEJ) |
DSBs, any cell-cycle stage |
No |
Error-prone; small insertions or deletions |
| Homologous recombination (HR) |
DSBs, after DNA is copied |
Yes (sister chromatid) |
High |
| Microhomology-mediated end joining |
DSBs, backup route |
No |
Error-prone; deletions |
| Single-strand annealing |
DSBs between two repeats |
No |
Deletes the sequence between the repeats |
The choice of pathway depends on the cell-cycle stage and on how far the ends have been resected. This choice is clinically important: HR failure (for example BRCA1 or BRCA2 loss) forces reliance on error-prone routes.
6. Genome architecture: non-coding DNA and repeats
This section records two positions that came up in discussion, so the board can see where they agree and differ.
Position A: everything in the genome matters. Much non-coding DNA is regulatory: enhancers, silencers, insulators, promoters, and the sequences that organise 3D chromosome domains. Non-coding RNAs have real functions. Most variants linked to common disease by genome-wide association studies lie outside coding regions. Centromeres and telomeres are non-coding repeats that chromosomes cannot do without.
Position B: substantial portions are tolerant or neutral. Comparisons across species suggest roughly 5 to 15% of the human genome (depending on method) is under measurable selective constraint, against about 1.5% that codes for protein. Biochemical activity is not proof of function, since cells produce a lot of noisy transcripts. Some large non-coding deletions in mice produced no obvious phenotype.
Where they meet. Neither position says a break is trivial. A break in non-coding DNA is less likely to delete a gene, but it can disrupt regulation or domain boundaries, and repeats can make misrepair rearrange large regions. Both positions agree that location and repair outcome decide the consequence.
Repeats
About half of the human genome, and by some estimates more, derives from transposable elements: sequences that copy themselves and insert elsewhere. Alu elements make up roughly 10% and LINE-1 roughly 17%.
- Functional repeats: telomeric repeats (TTAGGG), alpha satellite at centromeres, ribosomal DNA repeats, and a retroviral gene (syncytin) needed for the placenta. Some Alu and other element relics act as enhancers, p53 binding sites, and insulator (CTCF) sites.
- Origin versus use: most repeats exist because they replicated, and the host recruited some later (exaptation). Function followed origin, which is why blanket purpose is hard to defend. Genome size varies from about 400 million bases in a pufferfish to tens or over a hundred billion in lungfish, largely through repeats, between animals of similar complexity.
- Cost: repeats are fragile spots for replication, and misaligned repair between them (for example Alu-Alu) causes deletions and duplications behind many genetic diseases.
7. Consequences
- Cancer. Unrepaired or misrepaired DSBs drive translocations, deletions and amplifications. Inherited defects in repair genes (BRCA1, BRCA2, ATM, TP53 and others) raise cancer risk.
- Neurodegeneration. Neurons are long-lived, non-dividing and metabolically active. Repair defects cause severe neurological disease (ataxia telangiectasia is an example).
- Ageing. Accumulating damage and declining repair capacity are leading features of most ageing models.
- Immunity. Controlled DSBs generate antibody and receptor diversity; errors here seed lymphoma.
- Chromosome shattering. A micronucleus can break apart and be reassembled in a single event (chromothripsis), producing complex rearrangements at once.
- Innate immune signalling. DNA fragments leaking into the cytoplasm are sensed by cGAS-STING, linking damage to inflammation.
- Evolution. Breaks and their repair are a source of structural variation, gene duplication and, over long times, genome expansion and contraction.
8. Clinical and technological relevance
- PARP inhibitors. PARP enzymes help repair SSBs. In cells already lacking HR (such as BRCA-mutant tumours), blocking PARP is selectively lethal. This is a clinical example of synthetic lethality.
- Radiotherapy and many chemotherapies work by causing DSBs or by poisoning topoisomerases, so tumour sensitivity depends on repair capacity.
- CRISPR-Cas9 relies on creating a DSB and on the cell's own choice of repair route. The pathway chosen determines whether the edit is a disruption (NHEJ) or a precise correction (HR). Large deletions and rearrangements at the cut site are a recognised safety concern.
- Base editors and prime editors avoid a full DSB, which is a main reason for their appeal.
9. Open questions
- How much of the genome is truly functional, and how should functional, neutral and costly sequence be defined and measured?
- Why do some regions (fragile sites, repeat-rich zones) break repeatedly, and does the cell treat them differently?
- How does the cell choose between repair pathways in practice, and how does that vary by tissue and age?
- Can repair capacity be raised safely, given that excess or misdirected repair also has costs?
- How much of the age-related rise in damage reflects more breaks, and how much reflects weaker repair?
10. Glossary
- DSB: double-strand break, both strands cut.
- SSB: single-strand break, one strand cut.
- ssDNA: single-stranded DNA.
- NHEJ: non-homologous end joining.
- HR: homologous recombination.
- RPA: replication protein A, the protein that coats ssDNA.
- ATR: a damage-response kinase activated by ssDNA.
- R-loop: a stretch where RNA pairs with one DNA strand, displacing the other.
- Topoisomerase: an enzyme that cuts and reseals DNA to relieve twisting.
- Transposable element: a self-copying DNA sequence that can move within the genome.
- Exaptation: a feature co-opted for a function other than its original one.
- PARP: poly(ADP-ribose) polymerase, an enzyme that detects and helps repair SSBs.
- Synthetic lethality: a pair of defects that kill the cell together but not alone.
- Chromothripsis: one-off shattering and reassembly of chromosomes.