Browsing by Person "Fricke, Florian"

Now showing 1 - 1 of 1

Analyse altersabhängiger Änderungen der DNA-Methylierung
(2024) Holländer, Olivia; Fricke, Florian
This study investigated age-dependent DNA methylation patterns to estimate chronological age. DNA methylation is a covalent bond of a methyl group at the 5'-carbon atom of a cytosine, which in mammals occurs almost exclusively in sequences of cytosine-phosphate-guanine, so-called CpG sites. For criminal proceedings, it is highly relevant whether a person has reached a legally relevant age limit of 14, 18 or 21 years. Estimating the age is also relevant for unaccompanied refugee minors who are entitled to special protection by law. There is often a lack of valid documents that can confirm that they are minors. Furthermore, estimating the age of unknown perpetrators of biological crime scene evidence can support police investigations in order to narrow down the circle of possible trace donors. In 2016, the German Ethics Commission expressed that they do not consider the medical methods currently used to estimate age to be suitable, as they violate the fundamental rights to physical integrity, human dignity and the general right to privacy. As buccal mucosal swabs or saliva samples represent a practicable, cost-effective and, above all, non-invasive alternative, an estimation model based on buccal mucosa was sought. The first research hypothesis covered age estimation of adolescents and young individuals. First, 88 age-dependent CpG sites of the eight markers PDE4C, ELOVL2, ITGA2B, ASPA, EDARADD, SST, KLF14 and SLC12A5 were analysed by bisulphite conversion and subsequent pyrosequencing. For this purpose, 141 buccal mucosal swabs from individuals between 21 and 69 years of age were analysed. In 54 of the 88 CpG sites analysed (in six markers PDE4C, ELOVL2, EDARADD, SST, KLF14 and SLC12A5), age-dependent methylation patterns were identified. This study showed that a reliable estimation of age is possible using as little as three CpG sites (PDE4C, EDARADD and KLF14) and that this model can also be successfully transferred to another sequencing method, namely minisequencing. In the second step, research focused on estimating the age of adolescent individuals. For this purpose, buccal swabs from 230 individuals between 1 and 88 years of age were used. Eight CpG sites of already established markers (PDE4C, EDARADD, SST, KLF14, ELOVL2, FHL2, C1orf132 and TRIM59) were analysed using mini sequencing. After analysing the correlation of the methylation status of the individual CpG sites with the chronological age of the test subjects, a logarithmic rather than linear relationship from birth to adulthood was observed, particularly for the markers ELOVL2 and TRIM59. To deal with this phenomenon, 20 years was defined as the ‘cut-off’ value, at which the chronological age is transformed in order to adjust the chronological age to the epigenetic age. A MAD of 4.680 years was thus achieved in the training set and a MAD of 4.695 years in the independent validation set. The estimates were most accurate in the lower age segment from 0 to 19 years (MAD of only 2.64 years), thus achieving the goal of a model that is suitable for age estimation for adolescents and young adults. To answer the second research hypothesis, the contribution of hydroxymethylation to the improvement of age estimation models was analysed. The term DNA methylation usually refers to 5-methylcytosine (5mC), but bisulphite conversion cannot distinguish between 5- hydroxymethylcytosine (5hmC) and 5mC. By subjecting a portion of the sample to an oxidation step prior to bisulphite conversion, in which 5hmC is oxidised and later treated like an unmethylated cytosine, 5mC and 5hmC can be analysed separately. Methylation values of 40 CpG sites in eight markers (EDARADD, PDE4C, SST, KLF14, SLC12A5, TOX2, LRRN2 and STK12A) were analysed by pyrosequencing in 108 individuals. Estimation models including 5hmC markers or take ydroxymethylation into account reveal similar accuracy as those that only take 5mC markers (so-called ‘true 5mC methylation’) into account and models based on the gold standard (undifferentiated 5mC+5hmC values). In any case, this makes 5hmC another promising marker for forensic age estimation. However, no significant improvement in estimation accuracy was achieved, which is why the second research hypothesis could not be confirmed. The 5hmC methylation values of individual CpG sites showed weak to moderate correlations with chronological age. The literature indicates that 5hmC shows better genome- wide correlation with age than 5mC and that it also regulates different genes. This suggests that the best 5hmC markers for age estimation may not yet have been found. Methods that can analyse longer sections of the genome are better suited for a future search for the best markers.