Biochemical Society Transactions

Genome Instability and Cancer

Modelling the comet assay

Darragh G. McArt, George McKerr, C. Vyvyan Howard, Kurt Saetzler, Gillian R. Wasson


The single-cell gel electrophoresis technique or comet assay is widely regarded as a quick and reliable method of analysing DNA damage in individual cells. It has a proven track record from the fields of biomonitoring to nutritional studies. The assay operates by subjecting cells that are fixed in agarose to high salt and detergent lysis, thus removing all the cellular content except the DNA. By relaxing the DNA in an alkaline buffer, strands containing breaks are released from supercoiling. Upon electrophoresis, these strands are pulled out into the agarose, forming a tail which, when stained with a fluorescent dye, can be analysed by fluorescence microscopy. The intensity of this tail reflects the amount of DNA damage sustained. Despite being such an established and widely used assay, there are still many aspects of the comet assay which are not fully understood. The present review looks at how the comet assay is being used, and highlights some of its limitations. The protocol itself varies among laboratories, so results from similar studies may vary. Given such discrepancies, it would be attractive to break the assay into components to generate a mathematical model to investigate specific parameters.

  • comet assay
  • computer simulation
  • HeLa cell
  • mathematical model
  • microscopy
  • systematic random sampling

The technique

The single-cell gel electrophoresis assay has been used for over 20 years to assess DNA damage. The assay is able to measure both single- and double-strand breaks with versatility and sensitivity [1]. It has been used successfully over the years in key areas from human nutrition to biomonitoring [2,3]. In a typical assay, cells are fixed in agarose on a frosted slide and lysed to remove cellular proteins, leaving a cavity containing DNA with sufficient salt (2.5 M NaCl) and detergent to disrupt and remove membranes, histones and soluble cell constituents [3]. The DNA is then left to unwind in a buffer solution before being electrophoresed to pull the negatively charged, extended and relaxed loops containing strand breaks out of the cavity, forming the tail [4]. Intact DNA remains in the head and supercoiled. Comets are visualized using an intercalating dye such as ethidium bromide or DAPI (4′,6-diamidino-2-phenylindole), which binds in the minor groove, and is normally analysed using comet-specific software [1].

The comet technique has been adapted over the years into different protocols such as the alkaline and neutral versions which were thought to discriminate between double- and single-strand breaks. This theory was revealed by Collins et al. [1] at the most recent International Comet Assay conference to be a misconception as alkaline conditions are not specific to detecting single-strand breaks.

FISH (fluorescent in situ hybridization) techniques were also adapted for application to the comet where the DNA probe can be hybridized to a specific gene sequence or even to a whole chromosome [5,6]. Comet-FISH has been used to study areas as diverse as hypomethylation in human colonic cells and region-specific repair in bladder carcinoma cells [7,8]. Recent developments with whole-chromosome paints allowed Rapp et al. [6] to examine UV-A breakage sensitivity by hybridizing probe to chromosomes 1, 2, 3, 8, 9, 11, 14, 18, 19, 21, X and Y. They were able to show that the numbers of chromatin strand breaks were indifferent to chromosome size [6].

Other such modifications include the BrdUrd (bromodeoxyuridine) comet, a technique of pulse labelling replicating DNA with BrdUrd, a thymidine analogue. McGlynn et al. [9] utilized such a technique to create a novel biomarker for colon carcinogenesis. The comet assay with specific modifications is also able to detect the subtle damages caused by oxidative stress and the protective effects of dietary components [2]. This modification to the assay uses the lesion-specific endonucleases FPG (formamidopyrimidine glycosylase) and EndoIII (endonuclease III) to introduce strand breaks at the site of oxidative damage. Collins et al. [10] have discussed the increased sensitivity and specificity of this assay, which they used to measure the level of oxidative damage in lymphocyte DNA of insulin-dependent diabetes mellitus patients.

The comet assay has moved over the years from being the main focus of an investigation to now being an analytical tool where pertinent results are gleaned from larger investigations. Such a move has merely heightened its reputation as a quick, simple and cheap method of analysis compared with other DNA-damage-investigation tools. Although its rise across differing fields has leant to a broader understanding of the assay, it has also elaborated on the need for standardizations in the technique. It is often used as a quick-fire detection method and Collins [11] states that “it is sometimes used without too much thought regarding how it works and the information it provides”.


The fact that large numbers of individual cells from different tissues can be analysed is a strong advantage of the assay [12]. As data are collected at an individual cell level, the comet assay gives a measure of the heterogeneity in a sample. Compared with other types of DNA-damage-measuring techniques, it is relatively cheap and easy to use, with analysis, using alkaline lysis, being completed within the day. A strong area for the comet assay is that of human investigatory studies as very small numbers of cells are required (<10000), the cells require no prelabelling with radioactivity and cells used can be non-invasive, e.g. lymphocytes [11].

Human lymphocytes have been commonplace in comet investigations over the years in human biomonitoring studies, as it is recognized that they traverse the circulatory system and offer a general overview on the health of the individual. Keretetse et al. [13] have used the comet assay to detect DNA damage and repair in lymphocytes from African petrol attendants from VOCs (volatile organic compounds). Wasson et al. [2] have summarized the impact that the comet assay using lymphocytes has had on the field of human nutrition and cancer.

HeLa cells, an immortal cervical cancer cell line, have long been used as the test bed in comet assay investigations [1416], due to their longevity, ease of growth in culture and because of their wide use in other scientific investigations. The comet assay has also been carried out on plant cells, e.g. Gichner et al. [17,18] testing the alkaline comet assay on tobacco root and leaf seedlings for DNA damage by cadmium and DNA staining with fluorochromes, and Jiang et al. [19] using Spirodela polyrhiza to analyse DNA damage induced by UV radiation.


One of the limitations of the comet assay may be its ability to detect very low or very high levels of DNA damage. Olive and Durand [20] talk about the concerns of using the comet to measure low levels of damage as detecting the ‘signal’ may be awkward. They also discuss the fact that, at such low levels of damage, the fragments and breaks would be of such a size that they would be hard to migrate out of the cavity into the agarose. There can also be problems detecting very high levels of damage as the comet tail length tends to reach a plateau at higher levels of damage and the percentage comet tail parameter also appears to become ‘saturated’. That is not to say that the comet is not a sensitive tool, as its ability to detect a wide range of DNA damage has been demonstrated [21,22], In fact it has been shown to be capable of detecting X-ray damage in the range 0.05–10 Gy [23,24]. However, modification of parameters such as unwinding and electrophoresis time may be necessary to allow for more rounded results.

There are different aspects to the comet assay technique that singularly all contribute to the overall result. Solution molarity, pH and the use of protein digests all make subtle differences to the outcome. Even more intrinsic are the wash times, unwinding times and electrophoresis times, which are all big contributors to the scoring values obtained by visual or computational analysis. Forchhammer et al. [25] discuss the difference caused by different alkaline unwinding and electrophoresis times on visual scoring of comets, with their results displaying variation in the DNA damage detected between investigators. In order for the comet assay to be fully accepted by the wider scientific community, there must be tighter restraints on the protocols used. McKenna et al. [26] discuss the relevance of standardized procedures as the comet assay strives for acceptance in cancer diagnostics, with Taube et al. [27] suggesting criteria for acceptance as needing to be robust and reproducible, so the assay should have the confidence of the community and be of value to the patient. Guidelines are being developed by the OECD (Organisation for Economic Co-operation and Development) for the alkaline version of the comet assay, and a validation study of the in vivo comet is being planned by the Mammalian Mutagenesis Study Group (MMS), which will help standardize the assay and gain greater acceptance [12].

Certain areas of the comet protocol are open to selection bias. The fact that the investigator chooses a fixed number of comets to analyse at an area of their choice increases the likelihood of bias and therefore increases the variability between multiple users. We have investigated a method of analysis that can be used to remove user bias, while at the same time offering better user reproducibility by incorporating a method of SRS (systematic random sampling) into comet analysis [27a]. We demonstrated that the SRS method of analysis led to an increase in the precision of estimates of DNA damage. This area of analysis has been noted by Forchhammer et al. [25] where their results suggested that inter-investigator difference in scoring is a strong determinant of DNA damage levels in comet assay work. McKenna et al. [26] have discussed the effect this variability of result has had on the comet assay's ability to be used as a standard analytical tool in clinical laboratories, stating that there remains doubts about its reliability, reproducibility and validity. They added that the comet assay will not be used in the management of cancer until these problems are addressed.

A recent paper by Lovell and Omori [12] has focused on the statistical design of the comet assays protocol in order to address some of the problems arising from collection and interpretation of data. In 1999, Lovell et al. [28] published a paper on the issues related to experimental design and statistical analysis of comet data in order to provide a template of analysis, stating that the experimental unit is the sample rather than the cell.

Choice of cell line may also be a very important determinant in comet assay sensitivity. Some cells are more radiosensitive than others. Moneef et al. [29] looked at the radiosensitivity of six bladder cancer cell lines and showed variability in the cells that were damaged. Radioresistant cell lines also demonstrated greater repair capacity and they concluded that the alkaline comet assay is a good predictor of bladder cancer cell radiosensitivity [29].

Delincée et al. [30] showed that certain plant cells perform better as indicators of damage to irradiation treatment than others, with some needing modifications of technique in lysis times in order to better separate the DNA.

Lymphocytes have been used in comet assay experiments for human nutrition and biomonitoring studies as they are easily obtained and highly specialized. They are limited, however, by their tendency to phenotypic variation and may not be suitable to detect organ-specific levels of damage [1].

In our opinion, one of the main limitations of the comet methodology is the lack of understanding of the underlying mechanisms of comet formation and how strongly the comet shape reflects the DNA damage that has occurred. The current thinking on how a comet is formed is that lysis (particularly when combined with a protease digestion step) removes the majority of the cell components, including the nuclear matrix, but the DNA remains supercoiled. Where a single-strand break occurs, this supercoiling is allowed to relax, and when an electrophoretic current is applied, this loop of supercoiled DNA is pulled out into the comet tail [1]. In this theory, the DNA in the tail remains attached to that in the head which explains the phenomena which is generally observed when tail length fails to increase further after a certain amount of damage. However, the appearance of the comet tail is more like a collection of fragments, which led to another theory that the tail was made up of unattached DNA fragments. This is unlikely as it would require breaks to occur quite close together to generate fragments.

Comet assay simulation

Using statistical and computational methods, we believe that a simulation of the comet assay would ascertain important information about DNA damage at a cellular level as well as about the fundamental basis of the comet assay while offering impartial judgement and reproducibility. Information is available on some aspects of comet structure in published literature as well as information on cellular structure and damaging agents [3134]. With new techniques both computationally and experimentally, it has become possible to examine intrinsic information on comet structure. Information is also available via novel or existing laboratory techniques and also through microscopic techniques to examine architecture.

Code has been generated to create a simulated comet that will imitate the shape of the nucleus, whether it has damaged or undamaged DNA. The initial creation of a virtual comet required data on chromatin loop size and chromosome lengths in base pairs [33]. The authors have developed a rudimentary three-dimensional model of the comet, with virtual attachment points based on loop size and location (Figures 1 and 2) [35].

Figure 1 Laboratory-generated undamaged comet
Figure 2 Computer-generated undamaged comet

This model simulation in the future could then be used to ask important questions on aspects of comet formation. For instance, it could be used to test the loop versus fragment theories on comet formation or to answer the question as to what size of loops are likely to be involved. It should allow for simulations based on hypothetical situations that may not be feasible in a laboratory environment and allow for a better understanding of cellular architecture.


The comet assay is a quick and simple method of analysing DNA damage on single cells. Its versatility has seen the technique gain wider acceptance across multiple fields as a useful tool. In the present review, we have noted some of the limitations of the assay, which are being addressed by the comet community in a bid to achieve a wider acceptance. Standardized, unbiased and statistically reliable methods are being developed to make the comet assay more acceptable to the wider scientific community and to offer reproducible results. It is our opinion that advancements in computational modelling and the abundance of literary information available could be used to address some of the issues surrounding the comet assay and help drive our understanding of the technique.


This work was supported in part by a Ph.D. bursary from the Department of Education and Learning Northern Ireland.


  • Genome Instability and Cancer: Biochemical Society Irish Area Section Focused Meeting held at National University of Ireland, Galway, Ireland, 4 December 2008. Organized and Edited by Michael Carty (National University of Ireland, Galway, Ireland).

Abbreviations: BrdUrd, bromodeoxyuridine; FISH, fluorescent in situ hybridization; SRS, systematic random sampling


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 27a.
  29. 28.
  30. 29.
  31. 30.
  32. 31.
  33. 32.
  34. 33.
  35. 34.
  36. 35.
View Abstract