A step-by-step comet assay protocol for DNA damage runs in one day for the alkaline version: embed 10,000–20,000 single cells in 0.7% low-melting agarose on a frosted slide, lyse in cold high-salt buffer with 1% Triton X-100, unwind DNA in alkaline buffer (pH >13) for 20 minutes, electrophorese at 25 V and 300 mA for 30 minutes in a cold tank, neutralize, stain with SYBR Gold or ethidium bromide, and score % tail DNA or olive tail moment by epifluorescence microscopy. Damage appears as a comet tail whose length and intensity scale with single-strand breaks, alkali-labile sites, and double-strand breaks.
The Single-Cell Electrophoresis Mechanism: Why DNA Forms a Tail
The comet assay works because of a size-dependent physical race. Intact genomic DNA is roughly 10⁹ base pairs long and cannot migrate through an agarose lattice. When you lyse the cell and strip histones in high-salt buffer, the DNA collapses into a compact supercoiled nucleoid that stays in the "head" of the comet.
Break the DNA and the supercoils relax. Fragments below roughly 10⁷ base pairs can now be dragged toward the anode, and they form the tail. The more breaks, the more small fragments, and the longer and brighter the tail becomes. In plain terms: the gel is a sieve, and broken DNA is small enough to squeeze through it.
Electrophoresis is the engine of the separation. At pH >13, single-strand breaks and alkali-labile sites (abasic sites, oxidized bases after repair glycosylase action) all register as strand breaks. That is why the alkaline assay detects far more lesions than the neutral version, which resolves only frank double-strand breaks above about 10 Gy equivalent.
The quantified threshold matters: the alkaline comet assay detects damage from roughly 0.1 Gy up to 10 Gy of ionizing radiation in mammalian cells, while neutral comets need 5–10 Gy before a visible tail appears. Flow cytometry and pulsed-field gel electrophoresis sit above and below this window, which is why comet fills the low-dose gap in genotoxicity testing.
Wet-Lab Timeline: Stages, Timing, and Working Temperatures
Timing is where most failed experiments die, so treat this timeline as the backbone of the protocol. Morning lysis is non-negotiable for same-day scoring.
| Stage | Duration | Working temperature | Critical parameter |
|---|---|---|---|
| Cell preparation and viability check | 20–40 min | 4 °C | >90% viability by trypan blue |
| Slide coating, first agarose layer | 15 min + 30 min dry | 37–42 °C agarose | 1% standard agarose |
| Cell embedding in second layer | 10 min | 37 °C agarose | 0.7% low-melting agarose |
| Lysis | 60 min (minimum 1 h) | 4 °C, dark | Fresh 1% Triton X-100 |
| Alkaline unwinding | 20–40 min | 4 °C, dark | pH >13 confirmed by meter |
| Electrophoresis | 20–30 min | 4 °C | 25 V, 300 mA, ~0.7 V/cm |
| Neutralization and fixation | 15 min + dry | 4 °C | 0.4 M Tris, pH 7.5 |
| Staining and scoring | 60 min | Room temp, dark | SYBR Gold 1:10,000 |
The whole alkaline assay, including scoring, fits in 8–10 hours. The neutral assay and the enzyme-modified variants (FPG for oxidized purines, ENDO III for oxidized pyrimidines) add 2–3 hours because enzyme incubation runs 30–45 minutes at 37 °C after lysis.
Preparing Cells and Slides Without Introducing Artifacts
Cell handling is the quiet source of false positives. Any mechanical shearing during pipetting, vortexing, or centrifugation manufactures strand breaks that you will then score as "damage." Use wide-bore tips, cut pipette tips, and low-speed spins at 200 × g or less.
Aim for 10,000–20,000 cells per gel, which is 1–10 µL of a 10⁵–10⁶ cells/mL suspension. Too few cells and you cannot find enough comets; too many and they overlap, making scoring impossible. For adherent lines, trypsinize gently and neutralize thoroughly; for blood cells, layer directly onto the agarose without washing more than once.
Embedding temperature is the second artifact source. Low-melting agarose must be held at 37 °C. If it is too hot, you cause thermal damage; if it cools below about 30 °C, it sets before you can spread the gel. A 37 °C water bath or heat block is essential equipment, not a luxury.
The unterminated-slide trap catches first-time users. Even with the recommended protocol, some cells are lost during lysis and electrophoresis. Always run at least two gels per condition per slide, and use frosted slides or pre-coated slides to anchor the agarose.
Scoring and the Quantified Comparison: Choosing Your Metric
Scoring choice determines whether your data are comparable across labs. The field-standard metric is % tail DNA, which is the fraction of fluorescence intensity in the tail relative to the whole comet. It is nearly linear with dose from 0.1 to 10 Gy, making it the metric of choice for genotoxicity testing per the OECD Test Guideline 489 (2014) for in vivo mammalian alkaline comet assays.
| Metric | What it measures | Linearity with dose | Best use case |
|---|---|---|---|
| % tail DNA | Fraction of DNA in tail | 0.1–10 Gy, near linear | Regulatory and dose-response work |
| Olive tail moment | % tail DNA × tail center of mass | Curvilinear | Historical comparison |
| Tail length | Migration distance | Saturates above ~2 Gy | Quick visual screening |
| Tail moment (% Tail DNA × length) | Composite | Curvilinear | Legacy datasets only |
The non-obvious caveat: tail length saturates far earlier than % tail DNA, so old datasets reporting tail length underestimate high-dose damage. If you are comparing against literature from before about 2005, convert carefully or reanalyze raw images.
Score at least 50–100 comets per gel and at least two gels per condition. Report the median rather than the mean, because comet distributions are right-skewed by cells in S-phase or with apoptosis-related DNA fragmentation. Automated image analysis software (Comet Assay IV, CASP, or OpenComet) reduces scorer bias by 15–30% compared with manual scoring in published inter-laboratory trials.
Edge Cases, Modifications, and the Enzyme-FPG Variant
The standard alkaline assay measures strand breaks, but most oxidative DNA damage (8-oxoguanine) does not create a strand break directly. To detect it, incubate gels after lysis with formamidopyrimidine DNA glycosylase (FPG) or human 8-oxoguanine DNA glycosylase (hOGG1); the enzyme cuts at oxidized purines, converting them to measurable breaks. Subtract the alkaline-only value from the enzyme value to isolate oxidized base damage.
There are three important edge cases. First, apoptotic cells produce highly fragmented DNA that looks like extreme damage and inflates averages; exclude them by their "hedgehog" morphology (>90% tail DNA) or by TUNEL counterstaining. Second, cells in S-phase (DNA replication) can show intermediate tails even without damage; this is why synchronized populations are preferable. Third, non-mammalian systems (fish, plants, invertebrates) require species-specific lysis times (often 2–3 hours) and different hemolymph volumes.
The neutral comet assay is the wrong tool for most DNA damage studies, but the right tool for double-strand break studies after 5–10 Gy. If you care about double-strand breaks below 1 Gy, comet cannot see them; switch to γ-H2AX immunofluorescence, which reaches detection down to about 0.01 Gy.
Common Misconceptions and Professional Caveats
The most persistent misconception is that longer electrophoresis equals more sensitivity. Beyond about 30 minutes at 25 V, tails saturate and small fragments migrate out of the frame entirely, reducing apparent damage. Sensitivity increases come from longer unwinding (up to 40 minutes) and higher pH, not from longer running.
A second misconception is that low % tail DNA means no damage. The assay has a background of 5–10% tail DNA even in untreated controls, partly from endogenous single-strand breaks and partly from processing. Always run concurrent negative and positive controls (e.g., 10 µM H₂O₂ for 5 minutes on ice as a positive control) on the same slide batch.
A professional caveat: inter-laboratory reproducibility of the comet assay is poor. Published ring trials show coefficients of variation of 20–40% between labs for identical samples, driven by differences in agarose concentration, lysis time, and electrophoresis voltage. Fix your parameters and never compare absolute values across labs; compare relative induction in matched controls.
The final caveat is regulatory. Per OECD Test Guideline 489 (2014), the in vivo alkaline comet assay is validated for regulatory genotoxicity testing in animal studies, but the in vitro version is not fully validated for regulatory submissions without supplementary data. In plain terms: it is a superb research tool and a hard sell for regulatory dossiers on its own.
What to Actually Do: A Practical Benchmarking Checklist
Before you call your first dataset publishable, do these five things on a fresh batch of cells.
- Run a positive control (10 µM H₂O₂, 5 minutes, 4 °C) and a negative control in parallel on every slide batch; if the positive control % tail DNA is below 30%, re-check lysis buffer and electrophoresis buffer pH.
- Verify viable cells are above 90% before embedding; dead cells leak DNA and generate false tails.
- Pre-cool the electrophoresis buffer to 0–4 °C and confirm the current reaches 300 mA at 25 V; a cold, low-conductance buffer prevents unwanted heat damage.
- Score at least 50 comets per gel with a single operator, or use automated software to eliminate scorer drift.
- Keep raw images for every scored gel; regulators and reviewers increasingly request them, and reanalysis is impossible without them.
When results look implausible, the diagnostic order is: viability check, positive control, buffer pH, agarose temperature, electrophoresis current. In our experience, 80% of failed comet runs trace to one of these five variables, and lysis buffer that is more than 24 hours old is the single most common cause of weak positive controls.
Limitations and What the Assay Cannot Tell You
The comet assay is a broad-spectrum detector, not a specific one. It cannot distinguish single-strand breaks from double-strand breaks, from abasic sites, or from repair intermediates unless you use enzyme variants. It cannot tell you whether the damage is in a specific gene, chromosome, or cell-cycle phase. And it cannot detect mutations, only lesions.
The sensitivity ceiling is another limit. Below about 0.1 Gy equivalent of strand breaks, the assay signal approaches its background noise floor. Above 10 Gy, tails detach and the measurement saturates. That is a working window of roughly two orders of magnitude, which is narrower than many researchers assume.
Finally, no single comet assay study should stand alone. For genotoxicity claims in a publication, pair the comet result with a clastogenicity assay (micronucleus) and, where relevant, an Ames test. The comet assay tells you DNA was damaged; only the downstream assays tell you whether it matters.