DNA methylation topology differentiates between normal and malignant in cell models, resected human tissues, and exfoliated sputum cells of lung epithelium

Abstract

Global DNA hypomethylation is a prominent feature of cancer cells including lung cancer, that has not been widely explored towards cancer diagnosis. In this study we assess the comparative distribution of global DNA methylation in normal cells versus cancer cells in various specimen models. We used in situ immunofluorescence labeling of overall 5-methylcytosine (5mC) and covisualization of global DNA (gDNA) by 4',6-diamidino-2-phenylindole (DAPI), confocal microscopy and 3D image analysis to derive 5mC/DAPI colocalization patterns in human cell lines (BEAS-2B, A549, H157) and upper respiratory epithelial cells derived from various sources (i.e., sputum from healthy and cancer patients, and resected tissues from normal parenchyma and lung tumors). By introducing 5mC/DAPI colocalization index as a metric we could distinguish between normal epithelial cells and aberrantly hypomethylated cancer cells. Cultured lung cancer cells (H157 and A549) had significantly lower indices compared to normal cells (BEAS-2B). Furthermore, we were able to identify such extensively hypomethylated low-index cells in tumor tissues and the matching sputum from cancer patients. In contrast, the indices of cells derived from sputum of healthy individuals had more similarity to epithelial cells of normal parenchyma and the phenotypically normal BEAS-2B cells. The results suggest that 5mC topology using high-resolution image cytometry shows potential for identifying hypomethylated cancerous cells in human tissues and amongst normal cells in matching sputum, which may render a valuable surrogate for biopsied tissues. This promising feature deserves further validation in more comprehensive studies.