Differentially expressed genes and their promoter region accessiblity

Question

[wyj.jackson]

I have the tissue-specific RNA-seq data and its corresponding ATAC-seq data. I first identify all the differentially expressed genes. Then using ATAC-seq data to check whether the promoter regions of those genes are differentially accessible. It turns out that only a small portion of differentially expressed genes' promoter regions are differentially accessible. I try to find explanations. One reason I can think about is that the data is not perfect so missed some signal. What could be other reasons in the biology perspective?

[ATpoint]

Any mechanism that does not strictly rely on chromatin accessability itself. Can be acetylation, binding of activating or anti-repressive co-factors, post-transcriptional regulation (e.g. miRNAs), etc.
Not really the focus of this community though. I suggest going into reviews of transcriptional regulation.

Answer 1

[i.sudbery]

We've never seen much in the way of correlation between chromatin openness at promoters and gene-expression changes. This is not totally unexpected. Studies of chromatin state find that chromatin in a promoter state is often the same across cell types (while enhancers tend to be more celltype specific).

It becomes obvious that gene expression is not entirely controlled at the chromatin openness level when one thinks about plasmid based reporter assays of promoter function. If you put e.g. AP-1 transcription factoer dependent promoter in a plamsid reporter, and then compare AP-1 activated vs non-activated, you'll see a difference in expression, despite the fact that plasmids don't have chromatin.