With Halloween coming up, perhaps it is fitting that the authors of two recent papers describe a “tangled web” of epigenetic regulations. In one paper, researchers suggest that DNA methylation and perhaps other epigenetic modifications disrupt long non-coding RNA (lncRNA) activity. In the other paper, a different team shows that DNAm messes with the miRNA regulation program.
A team from the U.S., The Netherlands, Poland, Australia, Belgium, and the U.K. looked at lncRNAs and their role in “alveolar capillary dysplasia with misalignment of pulmonary veins,” a rare lethal disorder of the lungs. Mutations in the FOXF1 gene showed up in patient samples, but some samples also had deletions of upstream regions. The team looked more closely at this region to see what was going on.
They examined samples from patients with normal FOXF1 genes and found deletions in an overlapping upstream region that they call the “shared deletion region” (SDR). There was lots of action in the SDR—lncRNA genes that are highly expressed in normal lungs, GLI2 binding sites, and differentially methylated CpG islands.
Also, the chromosome appears to loop in that region. Messing up any of these things could make lncRNA-mediated chromatin interactions go haywire. The researchers also say that this could be the basis for other human developmental disorders.
The other group, which is from the U.S. and Taiwan, studied miRNA genes in hepatocellular carcinoma (HCC) samples. They knew that some miRNAs are downregulated in HCC tumors and that DNA methylation can silence a few miRNAs in HCC tissues. So the group decided to look genome-wide at miRNA genes to see whether DNAm really changes their regulation.
They found a ton of differential DNAm in the patient samples. Most of the miRNA genes didn’t show a significant methylation difference. But eight genes that host miRNAs were hypermethylated. They tested three of these for miRNA expression levels. miR-10a was significantly downregulated in tumor tissues, as expected for a hypermethylated gene. Surprisingly, miR-10b was expressed more (though the difference was not significant), as was miR-196b. They say that DNAm can affect miRNA levels in HCC tumors.
Read all the data at Genome Research, October 2012 (lncRNA study) and Epigenetics, November 2012 (miRNA paper).