Predicting tissue-specific enhancers is a tricky business. While histone marks such as H3K4me1 and binding of p300 are good enhancer predictors, pointing out which enhancers are tissue-specific is more of a challenge. That is, of course, unless you have PreSTIGE – an algorithm developed by the Scacheri lab that identifies tissue-specific enhancers by combining H3K4me1 and gene expression data.
Previously we have seen how long non-coding RNAs (lncRNAs) come in two flavors: those that are promoter associated and those that are enhancer associated. It is also known that long non-coding RNAs play a role in enhancer function and show highly tissue-specific expression. In the latest issue of Cell Cycle, the Orom and Scacheri labs put two and two together and find that having an overlapping lncRNA is actually a good indicator of enhancer tissue specificity. Using data from 11 human cell lines with different tissues of origin they show that:
- 28% of all ENCODE annotated human lncRNAs overlap with tissue-specific enhancers.
- Expression of lncRNAs overlapping tissue-specific enhancers is higher in the specific cell lines where the enhancer is active.
- There is a good correlation between lncRNA expression and expression of its putative target protein-coding gene.
- PreSTIGE-predicted tissue-specific enhancers have lower H3K4me1/H3K4me3 ratio, possibly explaining why they were partly missed by previous studies.
Summing it up, Vučićević et al. have used the powerful tissue-specific enhancer prediction abilities of PreSTIGE to suggest a functional link between tissue-specific gene expression and lncRNAs.