Regulation of the functional interaction between cyclin D1 and the estrogen receptor

Abstract

We report that the functional interaction between cyclin D1 and the estrogen receptor (ER) is regulated by a signal transduction pathway involving the second messenger, cyclic AMP (cAMP). The cell-permeable cAMP analogue 8-bromo-cAMP caused a concentration-dependent enhancement of cyclin D1-ER complex formation, as judged both by coimmunoprecipitation and mammalian two-hybrid analysis. This effect was paralleled by increases in ligand-independent ER-mediated transcription from an estrogen response element containing reporter construct. These effects of 8-bromo-cAMP were antagonized by a specific protein kinase A (PKA) inhibitor, indicating that the signaling pathway involved was PKA dependent. Further, we show that culture of MCF-7 cells on a cellular substratum of murine preadipocytes also enhanced the functional interaction between cyclin D1 and ER in a PKA-dependent manner. These findings demonstrate a collaboration between cAMP signaling and cyclin D1 in the ligand-independent activation of ER-mediated transcription in mammary epithelial cells and show that the functional associations of cyclin D1 are regulated as a function of cellular context.

FIG. 1

The physical interaction between cyclin D1 and ER is enhanced by 8b-cAMP in a concentration- and PKA-dependent manner. MCF-7 cells were transfected with pRc/CMV-cyclin D1-HA (6.9 μg) and pcDNA3.1-hER (5.8 μg). (A) At 24 h later, various amounts of 8b-cAMP or vehicle control were added. Incubation was continued for a further 24 h. Whole-cell lysates were prepared, and cyclin D1-ER complex formation was analyzed by immunoprecipitation with ER antibodies (AER314) followed by Western blotting for cyclin D1-HA using 12CA5 antibody against the HA epitope (middle panel). The relative amounts of ER precipitated (top panel) and cyclin D1-HA present in the whole-cell lysates (WCL; bottom panel) were assessed by Western blotting with the same anti-ER and anti-HA antibodies. (B) 8b-cAMP (100 μM) or vehicle control was added at the indicated times before cell lysis. The total posttransfection incubation time was 48 h. Lysates were analyzed as described for panel A. (C) At 24 h after transfection, various amounts of H-89 or vehicle control were added, immediately followed by 8b-cAMP (100 μM) or vehicle control. Incubation was continued for a further 24 h. Lysates were analyzed as described for panel A.

FIG. 2

Treatment with 8b-cAMP reveals a physical interaction between endogenous ER and cyclin D1. (A) D1.13 cells were treated with zinc sulfate (70 μM) or vehicle control, and incubation was continued for a further 18 h. Whole-cell lysates (WCL) were prepared and analyzed by Western blotting with anti-cyclin D1 (Ab-3) and anti-cdk4 (C-22) antibodies. (B) D1.13 cells were treated with 8b-cAMP (100 μM) followed 6 h later by zinc sulfate (70 μM) or vehicle control. Incubation was continued for a further 18 h before whole-cell lysates were prepared, and cyclin D1-ER complex formation was analyzed by immunoprecipitation with ER antibodies (AER314) followed by Western blotting for cyclin D1 (Ab-3). The relative amounts of ER precipitated (top panel) and cyclin D1 present in the lysates (bottom panel) were assessed by Western blotting with the same anti-ER and anti-cyclin D1 antibodies.

FIG. 3

The effect of 8b-cAMP is specific for the cyclin D1-ER interaction. MCF-7 cells were transfected with pRc/CMV-cyclin D1-HA (6.9 μg) and pCMV-cdk4, pCMV5-p27, pcDNA3.1-hER, or pRc/CMV (each at 2.9 μg) in the combinations indicated. At 24 h later, 8b-cAMP (100 μM) or vehicle control was added and incubation was continued for a further 24 h. Whole-cell lysates were prepared, and the formation of complexes between cyclin D1-HA, cdk4, p27, and ER was assessed by immunoprecipitation with antibodies against cdk4 (C-22; lanes 1 to 4), p27 (C-19; lanes 5 and 6), or ER (AER314; lanes 7 to 10) followed by Western blotting with the same antibodies or the 12CA5 antibody against the HA epitope of cyclin D1-HA.

FIG. 4

Cyclin D1 and 8b-cAMP act synergistically to stimulate ligand-independent ER-mediated transcription. MCF-7 cells, cultured under estrogen-free conditions, were transfected with p(ERE)₂-tk-luc (0.1 μg), pcDNA3.1-hER (0.1 μg), and pCMV-β-gal (0.25 μg), together with either pRc/CMV-cyclin D1-HA or pRc/CMV (each at 0.1 μg). At 24 h later, various amounts of 8b-cAMP, vehicle control, or 17β-estradiol (10 nM) were added, and incubation was continued for a further 24 h. Luciferase and β-galactosidase activities were determined, and normalized fold activations were calculated relative to the corrected luciferase activity in the absence of cyclin D1, 8b-cAMP, or estradiol. Results are means and standard deviations from three independent experiments each performed in triplicate.

FIG. 5

Coculture with murine preadipocytes, but not fibroblasts, enhances the physical interaction between cyclin D1 and ER. (A) MCF-7 cells were transfected as described for Fig. 1 before being trypsinized and replated on confluent cultures of NIH 3T3 fibroblasts or 3T3-L1 preadipocytes (lanes 3 and 7, respectively) or culture plastic (lanes 2 and 6). At 24 h later, whole-cell lysates (WCL) were prepared from each. Whole-cell lysates were also prepared from NIH 3T3 or 3T3-L1 cells cultured in isolation (lanes 1 and 5, respectively). For mixing experiments, lysates of transfected MCF-7 cells grown on plastic were combined with those from NIH 3T3 or 3T3-L1 cells cultured in isolation and rocked for 60 min at 4°C (lanes 4 and 8, respectively). Cyclin D1-ER complex formation was assessed in all lysates by immunoprecipitation with ER antibodies (AER314) followed by Western blotting for cyclin D1-HA using 12CA5 antibody against the HA epitope (middle panel). The relative amounts of ER precipitated (top panel), and cyclin D1-HA present in the whole-cell lysates (bottom panel) were assessed by Western blotting with the same anti-ER and anti-HA antibodies. (B) Mammalian two-hybrid analysis. MCF-7 cells were transfected with pGAL4-TATA-luc, pCMV-β-gal, pCMX-GAL4 or pCMX-GAL4-cyclin D1, and pCMX-VP16 or pCMX-VP16-ER (each at 0.25 μg) before being trypsinized and replated on confluent cultures of NIH 3T3 fibroblasts, 3T3-L1 preadipocytes, or culture plastic. At 24 h later, 8b-cAMP was added to a final concentration of 100 μM as indicated, and incubation was continued for a further 24 h. Luciferase and β-galactosidase activities were determined, and normalized fold activations were calculated relative to the corrected luciferase activity resulting from expression of GAL4-cyclin D1 and VP16 under each culture condition. Results are means and standard deviations from three independent experiments each performed in triplicate. (C) Cocultures were prepared as described for panel A, except that H-89 (50 nM) was added as soon as the cells had become attached (lanes 3 and 5), immediately followed by 8b-cAMP (100 μM) where indicated (lane 2). At 24 h later, whole-cell lysates were prepared and analyzed as described for panel A. (D) MCF-7 cells were transfected with pRc/CMV-cyclin D1-HA (6.9 μg) and pCMV-cdk4 or pcDNA3.1-hER (each at 2.9 μg) before being trypsinized and replated on confluent cultures of NIH 3T3 fibroblasts or 3T3-L1 preadipocytes. At 24 h later, whole-cell lysates were prepared, and complex formation between cyclin D1-HA and cdk4 or ER was assessed by immunoprecipitation with antibodies against cdk4 (C-22; lanes 1 and 3) or ER (AER314; lanes 2 and 4) followed by Western blotting with the same antibodies or the 12CA5 antibody against the HA epitope of cyclin D1-HA. The relative amounts of cyclin D1-HA present in the whole-cell lysates (bottom panel) were assessed by Western blotting with the same anti-HA antibodies.

FIG. 6

Coculture with preadipocytes enhances ligand-independent ER-mediated transcription induced by cyclin D1. MCF-7 cells, cultured under estrogen-free conditions, were transfected with p(ERE)₂-tk-luc (0.1 μg), pcDNA3.1-hER (0.1 μg), and pCMV-β-gal (0.25 μg), together with either pRc/CMV-cyclin D1-HA or pRc/CMV (each at 0.1 μg) before being trypsinized and replated on confluent cultures of NIH 3T3 fibroblasts, 3T3-L1 preadipocytes, or culture plastic. At 24 h later, 8b-cAMP was added to a final concentration of 100 μM as indicated and incubation was continued for a further 24 h. Estrogen-free conditions were maintained throughout. Luciferase and β-galactosidase activities were determined, and normalized fold activations were calculated relative to the corrected luciferase activity in the absence of cyclin D1 under each culture condition. Results are means and standard deviations from three independent experiments each performed in triplicate.