A combinatorial mechanism for determining the specificity of E2F activation and repression

Abstract

Various studies have detailed the role of E2F proteins in both transcription activation and repression. Further study has shown that distinct promoter elements, but comprising the same E2F-recognition motif, confer positive or negative E2F control and that this reflects binding of either activator or repressor E2F proteins, respectively. We now show that the specificity of binding of an activator or repressor E2F protein is determined by adjacent sequences that bind a cooperating transcription factor. We propose that the functional E2F element is a module comprising not only the E2F-binding site but also the adjacent site for the cooperating transcription factor.

Figure 1. Binding of E2F proteins to E2F promoter elements

A. Schematic of the human Cdc6 and Cdc2 promoters. Arrows depict the transcription start sites. Transcription factor binding sites relevant to this study are as indicated (Schlisio et al., 2002; Tommasi and Pfeifer, 1995; Yan et al., 1998; Zhu et al., 2004). B. C. Binding of E2F3 and E2F4 to E2F sites within the Cdc6 promoter and Cdc2 promoter, respectively. EMSAs used to detect binding of in vitro transcribed and translated HA-E2F3 and HA-E2F4 proteins to a Cdc6 promoter fragment containing the −1 E2F site, to a Cdc6 promoter fragment containing the −36 E2F site, to a Cdc2 promoter fragment containing a wild type distal E2F site and a mutated proximal E2F site, or to a Cdc2 promoter fragment containing a wild type proximal E2F site and a mutated distal E2F site, as indicated next to each gel. Lane 1 depicts the control reaction containing only the radiolabeled DNA probe. Lanes 2 and 4 depict binding reactions using 4 μg of in vitro transcribed and translated HA-E2F3 or HA-E2F4 protein, respectively. Lanes 3 and 5 depict antibody supershift assays using 0.2 μg of a polyclonal antibody against HA.

Figure 2. Role of promoter context in determining specificity of interaction of E2F proteins with the Cdc6 promoter

T98G cells were co-transfected with mutant promoter-reporter constructs and an internal control plasmid containing the β-galactosidase gene. Cells were harvested 18 hours after HU block and reporter ChIP experiments were done as described in Materials and Methods. Promoter constructs assayed are indicated above each graph. Positive-acting and negative-acting E2F sites are depicted by green and red boxes, respectively. YY1 sites are depicted by grey boxes. Antibodies used in the reporter ChIP experiments are as indicated, NR=normal rabbit, E2F3N=anti-E2F3, E2F4A=anti-E2F4. A. Interaction of E2F4 with the −36 E2F site within the Cdc6 promoter during G1/S phase of the cell cycle. B. Interaction of E2F3a with the −36 E2F site within the Cdc6 promoter in the presence of an YY1 element adjacent to the −36 E2F site during G1/S phase of the cell cycle. C. Interaction of E2F4 with the −36 E2F site within the Cdc6 promoter in the presence of a mutated YY1 element adjacent to the −36 E2F site during G1/S phase of the cell cycle.

Figure 3. Role of promoter context in determining specificity of interaction of E2F proteins with the Cdc2 promoter

T98G cells were co-transfected with mutant promoter-reporter constructs and an internal control plasmid containing the β-galactosidase gene. Cells were harvested 18 hours after HU block and reporter ChIP experiments were done as described in Materials and Methods. Promoter constructs assayed are indicated above each graph. Positive-acting and negative-acting E2F sites are depicted by green and red boxes, respectively. YY1 sites are depicted by grey boxes. Antibodies used in the reporter ChIP experiments are as indicated, NR=normal rabbit, E2F3N=anti-E2F3, E2F4A=anti-E2F4. A. Interaction of E2F4 with the proximal E2F site within the Cdc2 promoter during G1/S phase of the cell cycle. B. Interaction of E2F3a with the proximal E2F site within the Cdc2 promoter in the presence of an YY1 element adjacent to the proximal E2F site during G1/S phase of the cell cycle. C. Interaction of E2F4 with the proximal E2F site within the Cdc2 promoter in the presence of a mutated YY1 element adjacent to the proximal E2F site during G1/S phase of the cell cycle.

Figure 4. Effect of E2F3 and YY1 knockdown on global gene expression

A. Western blot analysis using cell extracts from asynchronously growing HEK293 cells with untransfected as a control, or transfected with effective YY1 RNAi duplexes, or transfected with effective E2F3 RNAi duplexes. Protein levels were detected by Western blotting using antibodies against YY1 or E2F3, respectively. Levels of tubulin protein were detected by Western blotting using an antibody against α-tubulin as a control. B. Effect of E2F3 and YY1 knockdown on gene expression. Microarray analysis was done to identify genes exhibiting changes in regulation in response to transfection of siRNAs targeting YY1 or E2F3. Fold changes in expression of the twenty genes most effected in E2F3 knockdowns relative to untransfected cells are given (fold changes in expression of the same genes in YY1 knockdown cells relative to untransfected cells are also given). Asterisks indicate genes examined in Figure 4C. C. Asynchronously growing HEK293 cells were harvested and endogenous ChIP experiments were done as described in Materials and Methods. Promoters assayed and antibodies used in the endogenous ChIP experiments are as indicated, NR=normal rabbit.