p27Kip1 and p21Cip1 collaborate in the regulation of transcription by recruiting cyclin-Cdk complexes on the promoters of target genes

Abstract

Transcriptional repressor complexes containing p130 and E2F4 regulate the expression of genes involved in DNA replication. During the G1 phase of the cell cycle, sequential phosphorylation of p130 by cyclin-dependent kinases (Cdks) disrupts these complexes allowing gene expression. The Cdk inhibitor and tumor suppressor p27(Kip1) associates with p130 and E2F4 by its carboxyl domain on the promoters of target genes but its role in the regulation of transcription remains unclear. We report here that p27(Kip1) recruits cyclin D2/D3-Cdk4 complexes on the promoters by its amino terminal domain in early and mid G1. In cells lacking p27(Kip1), cyclin D2/D3-Cdk4 did not associate to the promoters and phosphorylation of p130 and transcription of target genes was increased. In late G1, these complexes were substituted by p21(Cip1)-cyclin D1-Cdk2. In p21(Cip1) null cells cyclin D1-Cdk2 were not found on the promoters and transcription was elevated. In p21/p27 double null cells transcription was higher than in control cells and single knock out cells. Thus, our results clarify the role of p27(Kip1) and p21(Cip1) in transcriptional regulation of genes repressed by p130/E2F4 complexes in which p27(Kip1) and p21(Cip1) play a sequential role by recruiting and regulating the activity of specific cyclin-Cdk complexes on the promoters.

Figure 1.

Association of p27 with the promoters of target genes. Flow cytometry analysis of NIH3T3 cells (A) or MEFs (B) performed at different times after proliferative activation. The association of p27 with the promoter of the Aurka (C) or Med18 genes (D) in NIH3T3 cells was analyzed by ChIP using anti-p27 or without antibodies as a control. Results are the mean value ± SD of three independent experiments and are relative to the control. The association of p27 with the promoter of the Aurka (E) or Med18 genes (F) in p27^wt and p27^KO MEFs was analyzed by ChIP using anti-p27 or without antibodies as a control. Results are the mean value ± SD of three independent experiments and are relative to the control. Statistical analysis was performed using the Student's t-test. *P < 0.05, ***P < 0.001.

Figure 2.

Differential association of D-type cyclins with the promoters of target genes. (A) The levels of p27 and D-type cyclins, at different times after proliferative activation of NIH3T3 cells, were determined by WB (input). The association of p27 with the D-type cyclins was analyzed by IP using anti-p27. IP using a non-specific IgG was used as a control. (B) Quantification of IP experiments. Results are the mean value of three independent experiments and are relative to p27 levels. The association of cyclin D2 (C), cyclin D3 (D) and cyclin D1 (E) with the Aurka promoter (upper panels) or the Med18 promoter (bottom panels) in NIH3T3 cells was analyzed by ChIP using anti-p27 or without antibodies as a control. Results are the mean value ± SD of three independent experiments and are relative to the control. Graphics showing the association kinetics of p27, cyclin D1, D2 and D3 to the Med18 (F) and Aurka (G) promoters. Statistical analysis was performed using the Student's t-test. **P < 0.01.

Figure 3.

p27 is required for the recruitment of cyclin D2 and D3 on the promoters of target genes in early and mid G₁. (A) The association of p27 with the D-type cyclins at different times after proliferative activation of MEFs was analyzed by IP using anti-p27. IP using a non-specific IgG was used as a control. The association of cyclin D2 (B) and cyclin D3 (C) with the Aurka promoter (upper panels) or the Med18 promoter (bottom panels) in p27^wt and p27^−/− MEFs was analyzed by ChIP using anti-cyclin D2, anti-cyclin D3 or without antibodies as a control. Results are the mean value ± SD of three independent experiments and are relative to the control. Statistical analysis was performed using the Student's t-test. *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 4.

p27 is required for the recruitment of Cdk4 on the promoters of target genes. (A) The levels of p27 and Cdk4 at different times after proliferative activation of NIH3T3 cells were determined by WB (input). The association of p27 with Cdk4 was analyzed by IP using anti-p27. IP using a non-specific IgG was used as a control. (B) The association of Cdk4 to the Aurka (left panel) and the Med18 promoters (right panel) in NIH3T3 cells was analyzed by ChIP using anti-p27 or without antibodies as a control. Results are the mean value ± SD of three independent experiments and are relative to the control. (C) The association of p27 with Cdk4 at different times after proliferative activation of p27^WT MEFs was analyzed by IP using anti-p27. IP using a non-specific IgG was used as a control. (D) The association of p27, cyclin D2 and cyclin D3 with Cdk4 at different times after proliferative activation of p27^WT and p27^KO MEFs was analyzed by IP using anti-p27. IP using a non-specific IgG was used as a control. (E) The association of Cdk4 with the Aurka (upper panel) and the Med18 promoters (bottom panel) in p27^wt and p27^−/− MEFs was analyzed by ChIP using anti-p27 or without antibodies as a control. Results are the mean value ± SD of three independent experiments and are relative to the control. Statistical analysis was performed using the Student's t-test. *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 5.

The KID domain of p27 is essential for the proper regulation of Aurka and Med18 transcription. The association of cyclin D2 (A) and cyclin D3 (B) with the Aurka (upper panels) and the Med18 promoters (bottom panels) in p27^WT or p27^CK− MEFs was analyzed by ChIP using anti-cyclin D2, anti-cyclin D3 or without antibodies as a control. Results are the mean value ± SD of three independent experiments and are relative to the control. (C) The levels of Aurka (upper panel) and Med18 (bottom panel) mRNAs were determined by qPCR at different times after proliferative activation in p27^WT, p27^−/− and p27^CK− MEFs. Results are the mean value ± SE of three independent experiments and are relative to the control. (D) The Aurka (upper panel) and Med18 (bottom panel) primary non-spliced transcript (PNS transcript) levels were determined by qPCR in p27^WT, p27^−/− and p27^CK− MEFs. The PNS transcript was determined using a primer of the first exon and another one of the first intron. Results are the mean value ± SD of three independent experiments and are relative to the control. (E) p27^−/− MEFs were infected with an empty vector or with a vector harboring full length p27WT. Levels of p27 in these cells were determined by WB. (F) The levels of Aurka and Med18 mRNAs in p27^−/− MEFs infected with a vector harboring full length p27WT were determined by qPCR. Results are the mean value ± SD of three independent experiments and are relative to the control. Statistical analysis was performed using the Student's t-test. *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 6.

The association of Cyclin D1 and Cdk2 with the promoters of target genes in late G₁ is independent of p27. (A) The association of Cdk2 with the Aurka (left panel) or the Med18 promoter (right panel) in p27^WT and p27^KO MEFs was analyzed by ChIP using anti-Cdk2 or without antibodies as a control. Results are the mean value ± SD of three independent experiments and are relative to the control. (B) The association of Cdk2 with the Aurka (upper panel) and the Med18 promoters (bottom panel) at different times after proliferative activation of NIH3T3 cells was analyzed by ChIP using anti-Cdk2 or without antibodies as a control. Results are the mean value ± SD of three independent experiments and are relative to the control. (C) Graphics showing the association kinetics of p27, cyclin D1, Cdk4 and Cdk2 with the Aurka (upper panel) and the Med18 promoters (bottom panel). (D) The association of Cdk2 with the Aurka (left panel) and the Med18 promoters (right panel) in late G₁ in p27^WT and p27^KO MEFs was analyzed by ChIP using anti-Cdk2 or without antibodies as a control. Results are the mean value ± SD of three independent experiments and are relative to the control. Statistical analysis was performed using the Student's t-test. ns: not significant.

Figure 7.

p21 is required for the association of Cyclin D1 and Cdk2 with the promoters of target genes in late G₁. (A) The levels of p21, cyclin D1 and Cdk2 at different times after proliferative activation of NIH3T3 cells were determined by WB. (B) The interaction of cyclin D1 with p21 and Cdk2 in NIH3T3 cells was analyzed by IP using anti-cyclin D1. IP using a non-specific IgG was used as a control. (C) The association of p21 with the Aurka (upper panel) and the Med18 promoters (bottom panel) in NIH3T3 cells was analyzed by ChIP using anti-p21 or without antibodies as a control. Results are the mean value ± SD of three independent experiments and are relative to the control. (D) The levels of p21, cyclin D1 and Cdk2 along the cell cycle in p27^WT MEFs were determined by WB (input). The interaction of p21 with cyclin D1 and Cdk2 was analyzed by IP using anti-p21. IP using a non-specific IgG was used as a control. (E) The levels of E2F4 and p130 along the cell cycle in p27^WT MEFs were determined by WB (input). The interaction of p21 with E2F4 and p130 at different times after proliferative activation was analyzed by IP using anti-p21. IP using a non-specific IgG was used as a control. (F)The association of Cdk2 with the Aurka (upper panel) and the Med18 promoters (bottom panel) in late G₁ in control and p21^KO MEFs was analyzed by ChIP using anti-Cdk2 or without antibodies as a control. Results are the mean value ± SD of three independent experiments and are relative to the control. (G) The association of cyclin D1 with the Aurka (upper panel) and the Med18 promoters (bottom panel) in late G₁ in control and p21^KO MEFs was analyzed by ChIP using anti-cyclin D1 or without antibodies as a control. Results are the mean value ± SD of three independent experiments and are relative to the control. (H) The levels of Aurka (upper panel) and Med18 mRNAs (bottom panel) were determined by qPCR at different times after proliferative activation in control and p21^KO MEFs. Results are the mean value ± SD of three independent experiments and are relative to the control. (I) p21^−/− MEFs were infected with an empty vector or with a vector harboring full length p21WT. Levels of p21 in these cells were determined by WB. (F) The levels of Aurka and Med18 mRNAs in p21^−/− MEFs infected with a vector harboring full length p21WT were determined by qPCR. Results are the mean value ± SD of three independent experiments and are relative to the control. Statistical analysis was performed using the Student's t-test. *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 8.

p27 and p21 collaborate in the repression of target gene expression. (A) The levels of phosphorylated p130 at different times after proliferative activation in control, p27^−/− and p21^−/− MEFs were determined by WB using anti-phospho-p130. The amount of tubulin was detected with anti-tubulin and was used as a loading control. (B) The levels of Aurka (upper panel) or Med18 mRNAs (bottom panel) were determined by qPCR at different times after proliferative activation in control and the double p27/p21 null MEFs (DKO). Results are the mean value ± SD of three independent experiments and are relative to the control. (C) Model of the sequential role of p27 and p21 in the regulation of transcription. Statistical analysis was performed using the Student's t-test. **P < 0.01, ***P < 0.001.