Transcribing Chromatin

	[Enlarge Image] SET/TAF-I is specifically required for transcription of chromatin templates Top panel shows a titration of Pol II with a naked DNA template; approximately equal levels of transcription occur in the absence (left) or presence) of recombinant TAF-I. In the bottom panel, the same experiment was performed with a template assembled into chromatin, which can only be efficiently transcribed in the presence of TAF-I.

By biochemical fractionation of human cell extracts, we identified SET/TAF-I, a histone chaperone, as a factor required for Pol II-mediated transcription activated by the vitamin D3 receptor (VDR) on templates assembled into chromatin in vitro. We also showed a requirement for SET/TAF-I in transcription responsive to other activators, suggesting that the requirement for a chaperone is a general one. In contrast, SET/TAF-I had no effect on transcription of naked DNA templates. Although other histone chaperone activities have been proposed to play critical roles in facilitating transcript elongation by Pol II, we showed that the requirement for SET/TAF-I was at an earlier stage, prior to or at the initiation step of the transcription cycle. Thus, SET/TAF-I may function to facilitate access of the Pol II core machinery to the promoter region of active genes in chromatin.

Although SET/TAF-I can facilitate assembly of chromatin, that ability can be dissociated from its stimulatory effect on transcription. The presence of SET/TAF-I in transcription reactions, moreover, does not cause wholesale disassembly of chromatin, which would effectively render the template naked (and the requirement for SET/TAF-I in transcription a possibly trivial one). We are now probing the mechanism by which SET/TAF-I modifies chromatin structure to allow engagment of the transcriptional apparatus. We have also initiated a systematic analysis of the proteins that bind specifcially to SET/TAF-I in a transcriptional extract. Interestingly, the gene encoding SET/TAF-I was first identified as a partner in an oncogene fusion event, possibly implicating SET/TAF-I-dependent programs of gene expression in tumorigenesis. It will therefore be critical to identify the transcriptional signature associated with loss of this histone chaperone in vivo.