Our overall aim is to understand the mechanisms of gene regulation in the context of chromatin structure. The packaging of DNA into chromatin is generally repressive for transcription. Much work has been done on chromatin structure but comparatively little is known about the detailed structure of native chromatin and how it responds to activation. To address this problem, we have invested a considerable amount of time in developing a new model system based on the yeast CUP1 gene. CUP1 encodes a copper-inducible metallothionein and its regulation is very well understood. The long term aim is to reconstitute the activation of CUP1 in vitro using native chromatin as substrate and then to dissect the mechanism of activation. The chromatin structure of inactive CUP1 is relatively ordered with clusters of alternative nucleosome positions separated by linkers of various lengths. This probably represents the undisturbed chromatin structure laid down during DNA replication. In the presence of the copper-activated transcriptional activator, Ace1p, nucleosomes are moved from their original positions to new positions including the linkers. This re-positioning of nucleosomes required the activator but is independent of the TATA boxes and transcription. We inferred that a dynamic chromatin structure had been created, in which nucleosomes are moved around on CUP1 by a chromatin remodelling activity recruited by Ace1p (3). The movement of nucleosomes is suggested to facilitate access to DNA for transcription factors, particularly for the formation of the transcription initiation complex. We have also shown that the nucleosome positions observed in native chromatin are encoded in the DNA sequence (2). The relationship between nucleosome re-positioning, histone acetylation and transcription complex formation at CUP1 was investigated (submitted). Targeted acetylation of both histones H3 and H4 at the CUP1 promoter was observed. Positioned nucleosomes containing regulatory elements (i.e., the Ace1p binding sites and the TATA boxes) were the targets for acetylation. Targeted acetylation required the presence of Ace1p and, surprisingly, the TATA boxes. This implicates a TATA-box dependent histone acetylase, which we are attempting to identify.Thus, targeted acetylation is not required for nucleosome movement (because the latter occurs in the absence of the TATA boxes). These observations have important implications for the mechanism of remodelling at CUP1: targeted acetylation occurs at the moment of TATA box recognition by TBP, or afterwards. Thus, acetylation occurs after most of the important events at the CUP1 promoter. Since Ace1p binds to acetylated nucleosomes with much higher affinity than to unacetylated nucleosomes, we propose that the function of targeted histone acetylation at CUP1 is to boost subsequent rounds of transcription by facilitating the re-binding of Ace1p and possibly TBP. We have also begun work on reconstituting CUP1 transcription in vitro. A variety of yeast transcription factors have been purified. Purified RNA polymerase II was capable of specific initiation from the CUP1 promoter, provided the template DNA is negatively supercoiled (1). This was surprising in view of the plethora of transcription factors apparently required for initiation. We suggested that the ultimate function of these other factors is to create a locally destabilized DNA helix to facilitate initiation by RNA polymerase II at the CUP1 initiation element.
