Recent studies have provided strong evidence that macromolecular complexes are used in the cell to remodel chromatin structure during activation and to create an inaccessible structure during repression, Although there is not yet any rigorous demonstration that modification of chromatin structure plays a direct, causal role in either activation or repression, there is sufficient smoke to indicate the presence of a blazing inferno nearby. It is clear that complexes that remodel chromatin are tractable in vitro; hopefully this will allow the establishment of systems that provide a direct analysis of the role that remodeling might play in activation. These studies indicate that establishment of functional systems to corroborate the elegant genetic studies on repression might also be tractable. As the mechanistic effects of these complexes are sorted out, it will become important to understand how the complexes are regulated. In many of the instances discussed above, the genes whose products make up these complexes were identified in genetic screens for effects on developmental processes. This implies a regulation of the activity of these complexes in response to developmental cues and further implies that the work to fully understand these complexes will occupy a generation of scientists.

The Polycomb group (Pc-G) genes are essential for maintaining the proper spatially restricted expression pattern of the homeotic loci during Drosophila development. The Pc-G proteins appear to function at target loci to maintain a state of transcriptional repression. The murine oncogene bmi-1 has significant homology to the Pc-G gene Posterior sex combs (Psc) and a highly related gene, Suppressor two of zeste [Su(z)2]. We show here that the proteins encoded by bmi-1 and the Pc-G genes Polycomb (Pc) and Psc as well as Su(z)2 mediate repression in mammalian cells when targeted to a promoter by LexA in a cotransfection system. These fusion proteins repress activator function by as much as 30-fold, and the effect on different activation domains is distinct for each Pc-G protein. Repression is observed when the LexA fusion proteins are bound directly adjacent to activator binding sites and also when bound 1,700 bases from the promoter. These data demonstrate that the products of the Pc-G genes can significantly repress activator function on transiently introduced DNA. We suggest that this function contributes to the stable repression of targeted loci during development.

We describe a system for screening cDNA expression libraries in Escherichia coli based on protein-protein interactions. The system utilizes fusion proteins containing the DNA binding domain of the lambda phage cl repressor and a heterologous dimerization domain, which is the target of the screen. Such chimeric proteins were functional as transcriptional repressors in E.coli; function was dependent on the presence of the heterologous dimerization domain, and function of the chimeras was disrupted by expression of excess dimerization domain. A screen was designed to identify factors that could interact with the heterologous dimerization domain and thereby inactivate the chimeric repressor. We used this screen to identify factors that could interact with the basic helix-loop-helix/leucine zipper domains of c-Myc, and isolated the cDNA for a previously characterized HMG domain protein that interacts specifically with c-Myc in this system. This screening method could be used with proteins that have the ability to homo- or heterodimerize.

DNA binding by transcriptional activators is typically an obligatory step in the activation of gene expression. Activator binding and subsequent steps in transcription are repressed by genomic chromatin. Studies in vitro have suggested that overcoming this repression is an important function of some activation domains. Here we provide quantitative in vivo evidence that the activation domain of GAL4-VP16 can increase the affinity of GAL4 for its binding site on genomic DNA in mammalian cells. Moreover, the VP16 activation domain has a much greater stimulatory effect on expression from a genomic reporter gene than on a transiently transfected reporter gene, where factor binding is more permissive. We found that not all activation domains showed a greater activation potential in a genomic context, suggesting that only some activation domains can function in vivo to alleviate the repressive effects of chromatin. These data demonstrate the importance of activation domains in relieving chromatin-mediated repression in vivo and suggest that one way they function is to increase binding of the activator itself.Binding of transcriptional activators to DNA is frequently inhibited by the presence of nucleosomes. GAL4 derivatives, glucocorticoid receptor, progesterone receptor, USF, Spl, and Myc/Max heterodimers are all able to bind to nucleosomal DNA, but the affinity of these activators for nucleosomal DNA is decreased by up to three orders of magnitude over that observed on naked DNA (1-12). Other activators, such as NF1 and human heat shock factor 1, cannot bind to nucleosomal DNA under conditions tested so far, raising the possibility that nucleosomal inhibition of binding might prevent the function of these activators under certain circumstances (3,10,13). The mechanisms that facilitate activator binding to chromatin are therefore a key aspect of transcriptional control in eukaryotes.Several considerations suggest that activation domains might increase the ability of activators to bind to nucleosomal DNA. Activation domains can contact general transcription factors (GTFs; for review, see refs. 14 and 15), leading to the possibility that cooperative interactions between activators and GTFs might facilitate activator binding in an activationdomain-dependent manner. Activation domains are also capable of alleviating the repressive effects of nucleosomes on transcription in vitro (16-18), and can alter chromatin structure in intact yeast cells (19,20 We demonstrate here that the VP16 activation domain can enhance the ability of the corresponding GAL4 derivative to bind to sites located in genomic chromatin in mammalian cells. In addition, we have measured the relative activation potentials of these GAL4 derivatives and others containing different activation domains on chromosomal and transiently transfected reporters. We found that the relative potencies of these activators were dramatically different when compared under these two conditions, such that only certain activation domains were effective in stimulating t...

The Polycomb group (Pc-G) genes are essential for maintaining the proper spatially restricted expression pattern of the homeotic loci during Drosophila development. The Pc-G proteins appear to function at target loci to maintain a state of transcriptional repression. The murine oncogene bmi-1 has significant homology to the Pc-G gene Posterior sex combs (Psc) and a highly related gene, Suppressor two of zeste [Su(z)2]. We show here that the proteins encoded by bmi-1 and the Pc-G genes Polycomb (Pc) and Psc as well as Su(z)2 mediate repression in mammalian cells when targeted to a promoter by LexA in a cotransfection system. These fusion proteins repress activator function by as much as 30-fold, and the effect on different activation domains is distinct for each Pc-G protein. Repression is observed when the LexA fusion proteins are bound directly adjacent to activator binding sites and also when bound 1,700 bases from the promoter. These data demonstrate that the products of the Pc-G genes can significantly repress activator function on transiently introduced DNA. We suggest that this function contributes to the stable repression of targeted loci during development.