E2F is required to prevent inappropriate S-phase entry of mammalian cells

Abstract

E2F is a family of transcription factors that regulates the cell cycle. It is widely accepted that E2F-mediated transactivation of a set of genes is the critical activity that governs cellular progression through G(1) into S phase. In contrast to this hypothesis, we demonstrate that E2F actually suppresses the onset of S phase in two cell types when the cells are arrested by gamma irradiation. Our findings indicate that in these cells, the critical event triggering progression from G(0)/G(1) arrest into S phase is the release of E2F-mediated transrepression of cell cycle genes, not transactivation by E2F. Furthermore, our data suggest that E2F-mediated transactivation is not necessary for the G(1)/S-phase transition in these cells.

FIG. 1

The E2F sites in the DHFR promoter can act as transcriptional repressors. (A) A comparison of the activity of reporter constructs driven by wild-type DHFR and b-myb promoters to reporters driven by DHFR and b-myb promoters from which the E2F sites are deleted demonstrates that E2F site activity varies with cell type. Luciferase values were normalized between cell lines to facilitate presentation. (B) E2F sites are transcriptional repressors in serum-starved NRK52E cells. Twelve hours after transfection, the medium was replaced with serum-free medium; 48 h later, the cells were harvested for luciferase assays. Results are reported in relative light units (RLU).

FIG. 2

Competitor plasmids containing E2F sites bind E2F efficiently and specifically as assessed by EMSA. (A) A plasmid that contains 24 E2F-binding sites efficiently binds the E2F in lysates from 293 and F9 cells, whereas a plasmid that is exactly the same except that it contains 24 mutant E2F sites is ineffective in binding E2F. The plasmid concentrations were adjusted so that the molar concentration of plasmid E2F sites was either 20:1 or 50:1 (as indicated) compared with the molar concentration of the radiolabeled E2F probe. (B) The efficiency with which a plasmid competes for E2F is dependent on the number of E2F sites it contains. One hundred nanograms of each of the competitor plasmids was used in the assays. Two nanograms of a radiolabeled oligonucleotide containing binding sites for E2F was used as a probe. (C) Under the same conditions that E2F competitor plasmids sequester E2F, they fail to compete for other transcription factors. Radiolabeled oligonucleotides containing the indicated transcription factor-binding sites were used as probes. Either 100 ng of E2F competitor plasmid or 50 ng of unlabeled oligonucleotide competitor was used, as indicated.

FIG. 3

E2F competitor plasmids block E2F activity in vivo. A minimal reporter construct containing three E2F sites and a TATA box (p3E2F-CAT) is activated by E1a. A competitor plasmid containing 12 E2F-binding sites blocks E2F-mediated transactivation, whereas a plasmid containing 12 mutant E2F sites does not. The effect of the competitor plasmid is specific for E2F since it has no effect on RSV-CAT, a reporter plasmid that lacks E2F sites. An immunoblot for E1a indicates that the competitor plasmid does not affect the level of E1a expression.

FIG. 4

The E2F in cells treated with gamma irradiation is a potent transcriptional repressor. (A) There is little free E2F in irradiated WS1 cells. A cocktail containing antibodies to the three pocket proteins alters the migration of all of the E2F complexes in an EMSA using lysates from cells treated with 1,000 rads of gamma irradiation, whereas an equal amount of E2F-1 antibody used as a control has only a minimal effect (left panel). Almost all of the E2F in irradiated cells is E2F-4; pRb and p130 are the predominant pocket proteins found in E2F complexes in irradiated cells (middle panel). A comparison of the E2F from irradiated WS1 cells with the E2F from HeLa cells confirms that there is little if any free E2F in the irradiated cells (right panel). (B) The mRNAs from three genes that are thought to be regulated by E2F, the thymidine kinase, DHFR, and b-myb genes, are markedly reduced in irradiated IMR 90 cells as assessed by RT-PCR. RT-PCR of β-actin was used as a control for RNA loading. The PCR for β-actin was performed with two different concentrations of the RT products to confirm that the PCR was in the linear range. The contrast and magnification of the images of the thymidine kinase, DHFR, and b-myb bands was increased to facilitate visualization (the same changes were applied to the entire image). (C) The activity of a DHFR reporter construct is severely repressed in irradiated NRK52E cells compared to the activity of the same construct in cells that are not irradiated, whereas a DHFR promoter construct from which the E2F sites are deleted is unaffected by radiation.

FIG. 5

E2F-mediated transcriptional repressor activity is necessary for gamma-irradiation-induced G₀/G₁ block. (A) An E2F competitor plasmid effectively sequesters the E2F from both nonirradiated and irradiated WS1 cells as assessed by EMSA. (B) An E2F competitor plasmid relieves repression of an E2F-regulated reporter gene. The activity of a b-myb luciferase construct increased approximately fourfold when cotransfected with p-24-E2F-COMPETITOR compared with its activity when cotransfected with p-24-E2Fm-CONTROL. p-24-E2F-COMPETITOR had no effect on the activity of a b-myb luciferase construct from which the E2F sites had been deleted. (C) Transfection of plasmids that bind E2F release gamma-irradiated cells from a G₀/G₁ block. Cells were transfected as indicated, serum starved for 48 h, refed, irradiated, and then assessed for S-phase entry. Values for each data point are shown on the graph. This experiment has now been repeated more than 15 times with WS1 cells using several different preparations of competitor plasmids and controls. The results are completely reproducible. (D) The efficiency with which a competitor plasmid releases cells from the gamma-irradiation-induced G₀/G₁ block depends upon both the number of E2F sites the competitor plasmid contains and the amount of competitor plasmid transfected. This experiment was performed with WS1 cells, as in panel C. (E) Competitor plasmids containing the E2F sites from the DHFR promoter release NRK52E cells from the gamma-irradiation-induced G₀/G₁ block. The experiment was performed as in panel C except that NRK52E cells were transfected with p-COMP-96-D-E2F (contains E2F sites from the DHFR promoter) and pGEM-3 (control), as indicated. The views shown are typical fields. This experiment has now been repeated more than eight times with NRK52E cells with several different preparations of competitor plasmids and controls, including three times with p-24-E2Fm-CONTROL and p-24-E2F-COMPETITOR. The results are completely reproducible. (F) pCOMP-96-E-E2F and pGEM-3 transfect with equal efficiency. An excess of each of these plasmids was cotransfected with a GFP expression vector, and representative fields were photographed after 24 h. p24-E2F-COMPETITOR and p24-E2Fm-CONTROL were assessed in the same manner and were found to transfect with the same efficiency (data not shown).

FIG. 5

E2F-mediated transcriptional repressor activity is necessary for gamma-irradiation-induced G₀/G₁ block. (A) An E2F competitor plasmid effectively sequesters the E2F from both nonirradiated and irradiated WS1 cells as assessed by EMSA. (B) An E2F competitor plasmid relieves repression of an E2F-regulated reporter gene. The activity of a b-myb luciferase construct increased approximately fourfold when cotransfected with p-24-E2F-COMPETITOR compared with its activity when cotransfected with p-24-E2Fm-CONTROL. p-24-E2F-COMPETITOR had no effect on the activity of a b-myb luciferase construct from which the E2F sites had been deleted. (C) Transfection of plasmids that bind E2F release gamma-irradiated cells from a G₀/G₁ block. Cells were transfected as indicated, serum starved for 48 h, refed, irradiated, and then assessed for S-phase entry. Values for each data point are shown on the graph. This experiment has now been repeated more than 15 times with WS1 cells using several different preparations of competitor plasmids and controls. The results are completely reproducible. (D) The efficiency with which a competitor plasmid releases cells from the gamma-irradiation-induced G₀/G₁ block depends upon both the number of E2F sites the competitor plasmid contains and the amount of competitor plasmid transfected. This experiment was performed with WS1 cells, as in panel C. (E) Competitor plasmids containing the E2F sites from the DHFR promoter release NRK52E cells from the gamma-irradiation-induced G₀/G₁ block. The experiment was performed as in panel C except that NRK52E cells were transfected with p-COMP-96-D-E2F (contains E2F sites from the DHFR promoter) and pGEM-3 (control), as indicated. The views shown are typical fields. This experiment has now been repeated more than eight times with NRK52E cells with several different preparations of competitor plasmids and controls, including three times with p-24-E2Fm-CONTROL and p-24-E2F-COMPETITOR. The results are completely reproducible. (F) pCOMP-96-E-E2F and pGEM-3 transfect with equal efficiency. An excess of each of these plasmids was cotransfected with a GFP expression vector, and representative fields were photographed after 24 h. p24-E2F-COMPETITOR and p24-E2Fm-CONTROL were assessed in the same manner and were found to transfect with the same efficiency (data not shown).

FIG. 5

E2F-mediated transcriptional repressor activity is necessary for gamma-irradiation-induced G₀/G₁ block. (A) An E2F competitor plasmid effectively sequesters the E2F from both nonirradiated and irradiated WS1 cells as assessed by EMSA. (B) An E2F competitor plasmid relieves repression of an E2F-regulated reporter gene. The activity of a b-myb luciferase construct increased approximately fourfold when cotransfected with p-24-E2F-COMPETITOR compared with its activity when cotransfected with p-24-E2Fm-CONTROL. p-24-E2F-COMPETITOR had no effect on the activity of a b-myb luciferase construct from which the E2F sites had been deleted. (C) Transfection of plasmids that bind E2F release gamma-irradiated cells from a G₀/G₁ block. Cells were transfected as indicated, serum starved for 48 h, refed, irradiated, and then assessed for S-phase entry. Values for each data point are shown on the graph. This experiment has now been repeated more than 15 times with WS1 cells using several different preparations of competitor plasmids and controls. The results are completely reproducible. (D) The efficiency with which a competitor plasmid releases cells from the gamma-irradiation-induced G₀/G₁ block depends upon both the number of E2F sites the competitor plasmid contains and the amount of competitor plasmid transfected. This experiment was performed with WS1 cells, as in panel C. (E) Competitor plasmids containing the E2F sites from the DHFR promoter release NRK52E cells from the gamma-irradiation-induced G₀/G₁ block. The experiment was performed as in panel C except that NRK52E cells were transfected with p-COMP-96-D-E2F (contains E2F sites from the DHFR promoter) and pGEM-3 (control), as indicated. The views shown are typical fields. This experiment has now been repeated more than eight times with NRK52E cells with several different preparations of competitor plasmids and controls, including three times with p-24-E2Fm-CONTROL and p-24-E2F-COMPETITOR. The results are completely reproducible. (F) pCOMP-96-E-E2F and pGEM-3 transfect with equal efficiency. An excess of each of these plasmids was cotransfected with a GFP expression vector, and representative fields were photographed after 24 h. p24-E2F-COMPETITOR and p24-E2Fm-CONTROL were assessed in the same manner and were found to transfect with the same efficiency (data not shown).

FIG. 6

Model for the mechanism by which a transfected competitor plasmid releases cells from the gamma-irradiation-induced G₀/G₁ block. The upper panel is supported by our finding that the E2F in gamma-irradiated cells is a potent transcriptional repressor, and the lower panel is supported by our finding that competitor plasmids containing E2F-binding sites bind and sequester E2F complexes in vitro and in vivo.