A Few Thoughts About Catalyzing Transcription

Abstract

This editorial could be a message to the scientist who has just discovered that his favorite transcription factor interacts with a histone acetylase and who is about to publish that his field is now entering the era of epigenetics. Only 20 years ago, the field of transcription was all about factors interacting with one another, and much of the research effort was aimed at identifying DNA-binding proteins and cis -control elements on the DNA (if nostalgic, see [1]). The logic behind this approach was that specific transcription factors brought to a promoter would in turn interact with general transcription factors and favor recruitment of the preinitiation complex and the RNA polymerase II (RNAPII). The problem with this way of thinking was that the expected interactions between the regulatory domain of the specific transcription factors and the preinitiation complex were convincingly visualized only in very rare cases. Then came the co-activators like GCN5 and p300, and the co-repressors like Sin3a. These proteins were binding to the regulatory domain of the transcription factors and were essential for their activity. At first, the approach to study them remained the same as for the transcription factors: everyone (including me) was looking for interactions between this additional layer of regulators and the preinitiation complex. Again, this gave a lot of disappointing negative results. Then the big breakthrough came from the laboratory of David Allis. Looking for histone acetylases in Tetrahymena, this laboratory cloned a protein with homology to GCN5. This finding made the first direct link between histone acetylation and gene activation [2]. It inferred that co-activators could be enzymes modifying the chromatin structure. The involvement of enzymatic activities in regulation of transcription was not an entirely new idea. In fact, at the time, it was already known that chromatin remodeling machineries such as the SWI–SNF complex could function as transcription regulators. However, the link between these machineries and other types of co-regulators was not defined. From what we know at present, we have to expect that essentially all co-activators or corepressors will either be enzymes, proteins associated with enzymes or proteins able to read the product of an enzymatic reaction. The corollary to this is that transcriptional activation or repression is most likely not about interactions, at least not directly, but rather about enzymatic activities modifying proteins present at the promoter.