Fig. 1.

DNA methylation differences associated with newborns exposed to the COVID-19 pandemic in utero. A Dendrogram representation of unsupervised hierarchical clustering of normalized and batch-corrected beta values of pre-pandemic controls (CTL, yellow), RES pandemic COVID-19-negative pregnancies (solid pink), and RES COVID-19-positive pregnancies (hatched pink). Timing of approximate start of gestation is indicated at the bottom: light blue, entirety of pregnancy occurred prior to start of the pandemic (December 2019); dark blue; gestation that started September 2019-February 2020, prior to the onset of the COVID-19 pandemic; medium blue, gestation started between March and May 2020, after the US declared a national disaster. B Principal component analysis (PCA) of global DNA methylation differences of normalized and batch-corrected beta values between newborns exposed to the COVID-19 pandemic during pregnancy (RES, n = 32, triangles) and pre-pandemic healthy controls (CTL, n = 12, circles). Subcategorization by maternal COVID-19 infection status during pregnancy is denoted by color: coral indicates COVID-19 negative pregnancies (n = 17), green indicates COVID-19 positive pregnancies (n = 15), and blue indicates pre-pandemic pregnancies, unexposed to COVID-19 (n = 12). C Heatmap of unsupervised, hierarchical clustering of the differentially methylated probes identified between pre-pandemic CTL and RES pandemic newborns. Z-score scale represents transformed intensity values between differentially methylated probes with red being a negative z-score, white being a z-score of 0, and blue a positive z-score. The grid below the heatmap represents various clinical metrics, including timing of pregnancy (same color code as in 1A), COVID-19 infection status and severity of disease, and prenatal maternal mental health questionnaire data (STAI-S, STAI-T, PSS, PTSD), keyed in the legend to the right of the heatmap. Each column represents a sample, while rows represent specific probes that are differentially methylated between RES and CTL after performing linear regression analysis (FDR p-value ≤ 0.05, log₂ fold-change threshold ≥ 1 or ≤ -1). D IPA analysis of annotated differentially methylated probes between CTL and RES cohort. Z-score scale was calculated from differential intensity values and red and blue shades indicate pathways with a negative or positive z-score, respectively. White indicates pathways with a z-score of 0 (genes in the pathway are differentially methylated, some positively, some negatively, resulting in a Z-score = 0). Absolute log(p-value) represents statistical significance for the identified pathway to be associated with the imported list of annotated probes