Statistical Analysis of Alkaline Comet Assay Data for DNA Damage Assessment

1. Statistical analysis of the alkaline comet assay data for DNA damage assessment, also with the use of enzymes for evaluating specific DNA damage (e.g. ENDO III, hOGG1, FPG) and the assessement of DNA repair IRCCS San Raffaele Pisana, Rome, Italy,28 February - 2 March 2018

2. The comet or single-cell gel lectrophoresis assay is now widely used in regulatory, mechanistic and biomonitoring studies using a range of in vitro and in vivo systems. Each of these has issues associated with the experimental design which determine to a large extent the statistical analyses than can be used. Biomonitoring studies, being observational rather than experimental, are vulnerable to confounding and biases. Critical factors in any statistical analysis include the identification of suitable end points, the choice of measure to represent the distribution of the comet end point in a sample of cells, estimates of variability between experimental units and the identification of the size of effects that could be considered biologically important.

4. Which endpoint ?? The comet has a complex form which after visualization can be simplified to a set of multivariate data representing the shape. Automated scanners can measure many different components relating to the shape of the comet.

5. Three measures of DNA migration are commonly used: 1) Tail length, 2) Tail moment, 3) % of the DNA in tail (% tail DNA), Metrics on tail or head length and moment are measured in arbitrary units and may vary from study to study or from laboratory to laboratory. The % tail DNA is a measure of the relative fluorescent intensity in the head and tail (3). The % tail DNA values are constrained to a maximum of 100 and a minimum of 0 with no variability at the extremes and maximum variability at intermediate values such as 50%. An alternative scoring system is to classify DNA migration data using a five-category classification scheme (0, no damage to 4, almost all the DNA in the tail). 4) Damage index The system is manual and relies on a subjective assessment based upon comparisons with standard images.

6. Statistical unit is the person ! Not the cell !! It is common practice to measure DNA migration in at least 50 Comets per Gel. There are often two replicate Gels per experiment (i.e., one day of analysis). Consequently, there are usually 100 measurements of DNA migration per Sample. All comments and examples from now on refers to the use of the Comet Assay in human biomonitoring studies

7. Many distributions of comet measures (e.g. % tail DNA) within a sample (intra-sample) from a culture or animal are not normally distributed but rather may be asymmetric, skewed, bi- or multi-modal. Histograms of the distribution of log-transformed tail lengths in empirical in vitro data. Data can be transformed to try to make the distribution of the data conform to normality. The logarithmic transformation has the convenient property that back transformation (taking antilogs) to biologically meaningful values is easier than with other transformations. The problem of the logarithm of zero values can be overcome by the addition of small positive values (such as 0.001) to the data. It is important to appreciate, though, that while a transformation may correct one violation, say normality, it can result in another, such as heterogeneity of variances.

8. Uni- and Bi-variate analysis of Comet assay data. CAN PARAMETRIC TESTS BE USED FOR COMET ASSAY DATA? The distribution of Comets is typically nonormal.This sometimes leads to the misconception that comet assay data cannot be analyzed by parametric tests. Comet data can be analyzed by either parametric or non-parametric tests, depending on the homogeneity of variance and distribution of residuals(i.e., the unexplained variation).There are a range o different post-hoc tests parametric tests, including Dunnett’s, Fisher’s least statistical difference, Scheffe’s and Tukey’s tests. Comparison of 2+ means can be done with one-way analysis of variance (ANOVA), implying no a priori hypothesis of a linear relationship. First we test for homogeneity of variance between the groups (e.g., by Levene’s test). Post hoc testing and log transformation has to be applied

9. All endpoints of the comet assay should be tested in a study ?

10. Multivariate analysis of Comet assay data.

11. Hierarchical design, random effect models and generalized linear modelling

12. Multivariate analysis of Comet assay data. The robustness of the major parametric techniques and the propensity of comet assay data to be analysed after logarithmic transformation makes lognormal analyses extremely suitable for standard analyses of comet assay data. The use of measures of effect expresseed as ratio (including their 95% confidence interval is particularly meaningful for population studies, since it resembles the values expressed by Relative Risk of Odds Ratios.

13. Effect of Air Pollution on the Basal DNA-Damage of Mother-Newborns Couples of México City. Valverde M et al. In this study, we used the alkaline comet assay for evaluating the basal DNA damage in cord blood cells of newborns from different zones in the Metropolitan Area of Mexico City and in their mothers, possibly due to exposure in uterus to high levels of air pollution, principally ozone and air particles (PM10,PM2.5). Establishing correlation with the DNA damage of their mothers, and different parameters of the environmental exposure and complications during pregnancy. The study population was composed of 104 newborns: 64 located in the Southwest zone of the city, and 40 located in the Northeast zone. Thirty-nine mothers of the first group of children and all 40 of the second group were also included.

14. The metropolitan area of Mexico City was divided into 4 zones, namely northeast (NE), northwest (NW), southeast (SE) and southwest (SW). The accumulated concentrations over a nine-month period of the different pollutants (ozone, PM10, sulfur dioxide, carbon monoxide and dioxide, and nitrogen oxide and dioxide) were calculated for each zone. Over the past few years, ozone concentrations were indistinctly higher in the southern part of the city. Therefore, the metropolitan area was also divided into 2 global zones (north (N) vs. south (S)), and calculations for pollutant concentrations in these zones were performed. Separated analyses were conducted when considering the zone of residence of the the mothers, based on two or four subgroups.. 50 cells per slide were evaluated on two slides per individual. The comet head and tail were defined as described elsewhere (Rojas et al. 2000). The percentage of scored cells was calculated, and DNA migrations in the tail as well as comet head diameter were measured with an ocular micrometer. Categories of damage were established as follows: 0) low damage (tail length less than once comet head diameter), 1) intermediate damage (tail length from once to twice comet head diameter), 2) high damage (tail length more than two comet head diameters, and 3) extremely damaged cells forming clouds).

15. Univariate analysis was carried out to evaluate the effect of each variable on the level of DNA damage, measured as Tail Length (TL). A parallel analysis was performed using Damage index as dependent variable, but since results of the two endpoints overlapped only those referring to the Tail Length were reported Logarithmic transformation of data was applied to normalize the distributions of variables, which showed a significant departure from normality according to the Kolmogorov-Smirnov test.

17. Nine-months cumulative concentration of Nox, PM10 and Ozone in the group of mothers living in Mexico City. For Ozone it was calculated the cumulative concentration of the first trimester of pregnancy and the value of ozone at the day of delivery. Maximum values ​​allowed by Mexican legislation: PM10 : 50 ug/m3, Ozone 0.110 ppm, Nox 0.210 ppm

20. Scatter plot of the level of DNA damage (TL) in mothers vs. DNA damage (TL) in their newborns. Next Slide; Multivariate Log Normal modeling of the association between selected predictive variables and DNA damage (TL) (Mothers).

22. Multivariate Log Normal modeling of the association between selected predictive variables and DNA damage (TL) (Newborns).

23. In conclusion, our data show that basal DNA damage in both newborns and their mothers is low, probably due to a combination of the elevated antioxidant status of pregnant women and the protection offered by the placenta. We did observe that health complications in the newborns tend to increase the amount of damage in their DNA. However, the only pregnancy risk factor that increased the amount of damage in the newborns was maternal smoking. This report provides additional support for the use of Comet assay as a technique for assessing newborn health problems. Further research with a follow up design could validate the Comet assay as a powerful tool for evaluating DNA damage in exposome studies.