Test specificity between repressors and activators on a genome-wide scale.
To test for specificity between transcriptional repressors and active regulatory elements I have modified the STAP- and STARR-seq protocols that were previously developed in the host lab. Practically I have cloned 4xUAS sites next to the enhancer library, allowing for the recruitment of any selected Co-Repressor or GFP (control) by expressing Gal4-DNA-binding-domain CoR fusions. We then generated rep-STARR-seq libraries containing either a selected part (~5%) of the Drosophila Melanogaster genome (using BACs) or the entire genome for the final screens. Using the smaller BAC libraries, which contained around 200 active enhancers, I did a pre-screening with 17 Gal4-repressor proteins and Gal4-GFP as control. We found that the rep-STARR-seq method was highly reproducible and we identified specificity between transcriptional repressors and active enhancers. I then selected 5 prominent corepressors, CoRest, CtBP, Rbf, Rbf2 and Sin3A and performed the genome-wide screens, now testing the effect of these CoRs on all (3094) active enhancers in Drosophila S2 cells. I processed and analysed the sequencing data, again confirming that there is clear specificity between repressors and enhancers.
Identification of activators that confer resistance and sensitivity towards each of the CoRepressors.
Based on this differential response of the enhancers towards these CoRs, I then clustered the enhancers into 5 groups, and we found that these enhancer groups differ in chromatin marks and transcription factor (TF) motifs, using public ChIP-seq data combined with transcription factor motif enrichment analyses. By training and evaluating Generalized Linear Models, we showed that TF motif content could largely explain the sensitivity towards the different repressors. These analyses led us to identify several TF motifs that were specifically enriched in enhancers that were resistant to a given CoR, while at the same time being sensitive to other CoRs.
Validation of TF motifs that are required and sufficient for resistance/sensitivity towards repression.
Suggesting that these TF motifs could protect/be resistant against repression by a specific CoR. To further test this, performed additional rep-STARR-seq screens, this time using a designed oligonucleotide library of enhancer mutants. In particular, through motif mutant and motif paste experiments we could convincingly show that certain motifs are indeed protecting enhancers from repression by specific CoRs. These findings uncovered an additional layer in transcriptional regulation, where the specific interplay between transcriptional repressors and the proteins (TFs, CoFs) at active enhancers determines the transcriptional outcome.
This work will soon (upcoming month) be sent out for publication.
