P272 Geographical variation in DNA metylation: a comparative analysis of regional and intra-urban differences in two Italian birth cohorts

Résumé fourni par la source

Aim Previous studies have shown that the epigenome-wide average methylation correlates between twins, brothers and even spouses sharing the same household conditions, and that this correlation increased with time spent in the same household. With this study we aimed to provide a broader vision to the topic, by investigating whether these epigenetic similarities persisted at larger geographical scales. We thus examined the spatial footprint of the methylome across two levels: first between-cities (comparing five Italian centers) and then locally (assessing variation within a single city). This allowed us to determine if geographical proximity is reflected in similarity in methylation profiles, due to similar external conditions (e.g. socioeconomic, environmental, urban). Methods The between-city analysis was conducted on the PiccoliPiù cohort recruiting pregnant women from five different Italian birth centres (Firenze, Roma, Torino, Trieste, Viareggio), for which 364 cord blood samples were analysed. The within-city analysis was conducted on 171 cord blood samples from the PiccoliPiù cohort, collected from children residing in Torino at birth, and 135 saliva samples from the NINFEA birth cohort, collected from children aged 6-18 months residing in the Torino province at the time of sample collection. At both levels, we performed differentially methylated probes analyses and additional geostatistical analyses at the within-city level using variograms. Results For the first analysis we selected the top 500 differentially methylated probes between different birth centres, stratified by buffer of DNA extraction (maximum unadjusted p-value: 0.0005), and looked for intersections in at least three comparisons. This resulted in a set of 43 candidate probes that have been previously associated with external exposures, according to the EWAS Atlas. In the within-city analysis we did not find any differentially methylated probes among the candidates across the different local health authorities of the children’s residence. Furthermore, the variograms did not show any pattern of methylation similarity related to the residential distance between subjects. Conclusion To our knowledge, this is the first study to investigate the direct dependence between geographical distance and methylation differences. Our results suggest that this relationship is scale-dependent: while broad between-city differences leave a measurable biological footprint in the methylome, the lack of variation within the city of Torino suggests that administrative borders may not capture sufficient environmental heterogeneity to induce detectable changes. We conclude that geographical impacts on the methylome are evident at large scales, but limited at the sub-city level.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.