UC Berkeley

Transcription by RNA Polymerase II (Pol II) represents a major control point for regulation of eukaryotic gene expression. Yet, the mechanistic details and dynamics of a large number of transcriptional regulatory processes are currently unknown. Many of these processes, such as chromatin remodeling and epigenetic silencing, are mediated through nucleosomes, which comprise the repeating units of chromatin. Thus, it is of great interest to investigate the real-time dynamics of Pol II when it encounters a nucleosome, and to determine what happens to nucleosomes upon the passage of a transcribing Pol II.

It has been shown that a single nucleosome is sufficient to halt or greatly slow transcription by Pol II in vitro, and factors that restrict transcriptional backtracking by Pol II also relieve nucleosome-induced pauses and arrests. These observations suggest that the influence of the nucleosome is mediated through polymerase backtracking. Unfortunately, the temporal resolution provided by these biochemical studies was not sufficient to provide mechanistic information about the nucleosomal barrier. Using optical tweezers, we studied nucleosomal transcription of single Pol II complexes in real time, and obtained direct evidence for the first time that a nucleosome acts as mechanical fluctuating barrier that both increases the tendency of the polymerase to enter a backtracked pause and slows its recovery from these pauses. These changes in pause durations quantitatively agree with a model where temporary rewrapping of the nucleosomal DNA immediately downstream of the backtracked Pol II prevents enzyme recovery from its paused state. Furthermore, our studies revealed that Pol II does not actively separate the nucleosomal DNA from the surface of the histones, but, instead, acts as a ratchet that rectifies local nucleosomal unwrapping events to gain access to downstream DNA and overcome the nucleosomal barrier.

In vivo, the histone tails are essential for the regulation of gene expression, so we investigated their direct effect on the dynamics of transcription elongation. We found that removal of the tails favors progression of Pol II into the entry region of the nucleosome, by increasing the DNA fluctuations in this region. However, since our data shows that the magnitude of the barrier to transcription is highest in the central region of the nucleosome, and the tails only affect the entry region of the nucleosome, we investigated what interactions control the strength of the barrier near the nucleosome dyad. To this end, we used nucleosomes with point mutations in the histone-fold domains of H3 and H4, mutations that affect histone-DNA contacts at the dyad and that have been shown to partially relieve the requirement of the chromatin remodeling factor SWI/SNF in vivo. We found that these mutations abolish the barrier to transcription in the central region by increasing the local unwrapping rate of the DNA from the surface of the histones near the dyad. We speculate that factors that could bind to the nucleosome and specifically disrupt even a single DNA-histone contact in this region would have a profound effect on transcription.

Nucleosomes are disrupted to varying degrees by transcription elongation, with outcomes ranging from partial loss to complete removal and exchange of histones. In vitro studies with the phage SP6 RNA polymerase and RNA Polymerase III have shown that upon transcription the histone octamer moves upstream, while Pol II leads to the formation of a hexamer whose position on DNA is unchanged. The histone transfer process is believed to involve looping of the DNA template, but claims of template looping have so far relied on indirect evidence. Using atomic force microscopy, we obtained direct evidence of looping in the form of polymerase-nucleosome complexes in which the histones bridge the DNA upstream and downstream of the polymerase simultaneously. We also showed that a small fraction of the transcribed nucleosomes moved upstream of their original position. Significantly, we found that the fraction of the transcribed nucleosomes that are remodeled to hexasomes versus the ones that are transferred as intact octamers depends on the speed of elongation. A simple model involving the kinetic competition between the rates of transcription elongation, histone transfer, and histone-histone dissociation quantitatively rationalizes our observations and unifies results obtained with other polymerases.

In conclusion, we show how the finely tuned interplay between polymerase dynamics and nucleosome fluctuations determines the outcome of transcription, and we propose that factors affecting the relative magnitude of these processes provide the physical basis for the regulation of gene expression.