Gene activation involves the recruitment of a set of factors to a promoter in response to appropriate signals, ultimately resulting in the formation of an initiation complex by RNA polymerase II and transcription. These events must occur in the presence of nucleosomes, which are compact structures capable of blocking transcription at every step. To circumvent this chromatin block, eukaryotic cells possess ATP-dependent chromatin remodelling machines and nucleosome modifying complexes. The former (e.g. the SWI/SNF complex) use ATP to drive conformational changes in nucleosomes and to move nucleosomes along DNA. The latter contain enzymatic activities which modify the histones post-translationally to alter their DNA-binding properties and to mark them for recognition by other complexes, which have activating or repressive roles (the basis of the "histone code"). Nucleosome modifying enzymes include histone acetylases (HATs), deacetylases (HDACs), methylases, kinases and ubiquitin conjugating enzymes. The current excitement in the chromatin field reflects the recognition that chromatin structure is of central importance in gene regulation and that the cell has dedicated complex systems to manipulate the repressive properties of chromatin structure to maximum effect. Furthermore, multiple connections between chromatin and disease are apparent.
Many low resolution studies of chromatin structure have indicated that major changes in chromatin structure occur at promoters and at other regulatory elements of genes, but whether nucleosomes were conformationally altered, moved around, or simply removed was unclear. Current models are designed to account for these observations: they propose that the primary function of remodelling complexes is to convert the chromatin structure of a promoter to a state conducive to transcript initiation. In contrast, our high resolution chromatin studies indicate that, at least for two yeast genes, the chromatin structure of the entire gene is remodelled, not just the promoter, with important implications for mechanisms of gene regulation.
Initially, we chose the CUP1 gene for our studies 1, 2, 3, because its regulation is well understood, with well-defined basal and activated states. CUP1 encodes a metallothionein required to protect cells from the toxic effects of copper. Much later, it became clear that CUP1 was not the ideal choice, because the remodelling factors acting at CUP1 had not yet been identified, making it difficult to exploit the genetic advantages of yeast. Accordingly, we decided to study a second well-characterised gene, HIS3, for which this information is available. HIS3 encodes an enzyme required for histidine metabolism and is induced by amino acid starvation. HIS3 is activated by the transcriptional activator Gcn4p and is regulated by the SAGA and NuA4 histone acetyltransferase (HAT) complexes) and by the SWI/SNF ATP-dependent remodelling machine.
We discovered that induction of both CUP1 and HIS3 results in the creation of a domain of remodelled chromatin structure that extends far beyond the promoter, to include the entire gene 1, 4, 8. In the case of HIS3, induction results in a dramatic loss of nucleosomal supercoiling, a decompaction of the chromatin, a general increase in the accessibility of the chromatin to restriction enzymes and gene-wide mobilisation of nucleosomes. Formation of this domain of remodelled chromatin requires the SWI/SNF complex and the activator Gcn4p. The NURF-like remodelling complex, Isw1, also mobilises nucleosomes on HIS3, but to different positions. We propose that Gcn4p stimulates the activity of the SWI/SNF complex which then directs remodelling of the surrounding chromatin, generating a highly dynamic structure 8. We propose that this dynamic chromatin structure facilitates access to the DNA for both initiation and elongation factors, as well as promoting transcription through chromatin by RNA polymerase II (Figure 1). Our studies of HIS3 chromatin structure are at an exciting stage and are now focused on understanding the structure of transcriptionally active chromatin (see also ref. 9).
In summary, our work on CUP1 and HIS3 indicates that, at least for these two genes, the target of remodelling complexes is a domain rather than just the promoter. This is an important finding, because it suggests that remodelling complexes act on chromatin domains. In a wider context, the fact that remodelling complexes can participate in the formation of chromatin domains ("gene expression neighbourhoods") might be important in understanding the formation of domains in higher eukaryotes (discussed in 5).
Figure 1. A working model for the transcriptional activation of HIS3 chromatin.
The chromatin structure of the HIS3 gene expressed at basal levels (in the absence of the Gcn4p activator) is characterised by a dominant array of positioned nucleosomes (D1-D5). Alternative (A) arrays composed of quantitatively minor positioned nucleosomes are also present, indicating heterogeneity in HIS3 chromatin structure, even in the basal state. We propose that basal HIS3 chromatin is essentially static in nature. In the presence of the Gcn4p activator, the activity of the SWI/SNF complex is stimulated, resulting in a net mobilisation of nucleosomes from the D-arrays to the A-arrays. The Isw1 complex also affects the distribution of the nucleosomes, particularly at the 3'-end of HIS3. The inference is that HIS3 chromatin structure is highly dynamic. The nucleosomal flux created by the competing activities of the various remodelling complexes should facilitate access to the DNA for both transcript initiation and elongation complexes. Note also that we have shown previously that HIS3 nucleosomes apparently undergo a major conformational change requiring both Gcn4p and the SWI/SNF complex 4 which might increase the transparency of the chromatin still further. Adapted from Ref. 8.
We have shown that nucleosomes on the CUP1 promoter are acetylated in response to induction by copper and that this targeted acetylation is dependent on Spt10p, a putative histone acetyltransferase (HAT) 3. SPT10 was originally identified as one of a set of SPT genes, mutations in which suppress phenotypes associated with insertion of a yeast transposable element into promoters. SPT10 is not an essential gene, but the null allele is associated with very slow growth and global defects in gene regulation. Spt10p activates the histone genes, which it regulates in conjunction with Spt21p, the Hir co-repressor and the SWI/SNF complex. We and others originally proposed that Spt10p might be a co-activator recruited to promoters by activators. However, we have shown recently that Spt10p is in fact a sequence-specific DNA binding protein that recognises the histone UAS elements [(G/A)TTCCN6TTCNC] 6. Spt10p appears to be the activator of the core histone genes, which has been sought after for many years. We found that it binds with high affinity and with extraordinary positive cooperativity to pairs of histone UAS elements. Since pairs of histone UAS elements are found only in the core histone promoters and nowhere else in the yeast genome, there are no other predicted sites for Spt10p binding. We have presented evidence that the effects of Spt10p on other genes are indirect, mediated through global defects in chromatin structure arising from a deficit of histones in spt10 cells 6.
We are making rapid progress in understanding the biological functions of Spt10p 7, 10. Spt10p appears to be only the second example of a sequence-specific DNA binding domain fused to a HAT domain. However, Spt10p has not yet been shown to possess HAT activity in vitro. We are using a variety of approaches to identify the acetyltransferase activity of Spt10p. In addition, we are attempting to identify proteins which interact with Spt10p. Spt10p binds to the histone UAS elements and therefore should be classified as an activator rather than a co-activator, but it does not have a conventional activation domain. Usually activators recruit HAT enzymes as co-activators. In the case of Spt10p, we suggest that Spt10p recruits an activation domain. Our current aim is to place our observations in their biological context of S-phase regulated expression of the histone genes.
Recently, we identified the DNA-binding domain of Spt10p: it comprises about 110 residues and includes an H2-C2 zinc finger 7. A BLAST search identified a homologous H2-C2 zinc finger in the integrase of the human/Simian/primate foamy retrovirus (PFV). This zinc finger is conserved in all retroviruses, including HIV, and is thought to participate in protein-protein interactions. However, the homology of the foamy virus zinc finger with the DNA-binding domain of Spt10p suggests that the PFV zinc finger might in fact be a sequence-specific DNA-binding domain. We are addressing this possibility.
All related news