Cyclin-dependent kinase-mediated phosphorylation plays a critical role in the oncogenic functions of PELP1

Abstract

Estrogen receptor (ER) signaling plays an important role in breast cancer progression, and ER functions are influenced by coregulatory proteins. PELP1 (proline-, glutamic acid-, and leucine-rich protein 1) is a nuclear receptor coregulator that plays an important role in ER signaling. Its expression is deregulated in hormonal cancers. We identified PELP1 as a novel cyclin-dependent kinase (CDK) substrate. Using site-directed mutagenesis and in vitro kinase assays, we identified Ser(477) and Ser(991) of PELP1 as CDK phosphorylation sites. Using the PELP1 Ser(991) phospho-specific antibody, we show that PELP1 is hyperphosphorylated during cell cycle progression. Model cells stably expressing the PELP1 mutant that lack CDK sites had defects in estradiol (E2)-mediated cell cycle progression and significantly affected PELP1-mediated oncogenic functions in vivo. Mechanistic studies showed that PELP1 modulates transcription factor E2F1 transactivation functions, that PELP1 is recruited to pRb/E2F target genes, and that PELP1 facilitates ER signaling cross talk with cell cycle machinery. We conclude that PELP1 is a novel substrate of interphase CDKs and that its phosphorylation is important for the proper function of PELP1 in modulating hormone-driven cell cycle progression and also for optimal E2F transactivation function. Because the expression of both PELP1 and CDKs is deregulated in breast tumors, CDK-PELP1 interactions will have implications in breast cancer progression.

Figure 1

PELP1 is a novel substrate of interphase CDKs. A, Expression status of PELP1 in the cell cycle was analyzed in IMR-90 cells (left panel) and NIH3T3 cells (right panel). Cells synchronized at G0/G1 phase were released into cell cycle and lysates from different time intervals were used in Western blot analysis with the 220B2 PELP1 antibody. B, Total lysates from MCF7 cells grown in 10% serum were subjected to immunoprecipitation using the PELP1 antibody (left panel), and the CDK4 interaction was verified by Western blotting. T7-tagged PELP1 over-expressing MCF7 cells were treated with 10⁻⁸M estrogen for various periods of time, PELP1 was immunoprecipitated, and the CDK4 interaction was verified by Western blotting (middle panel). In vitro kinase assays for the CDK4/cyclinD complex using baculovirus-expressed GST-tagged full-length PELP1 as a substrate, and phosphorylation was measured by amount of ³²P incorporation (right panel). C, Total lysates from MCF7 cells grown in 10% serum were subjected to immunoprecipitation using the PELP1 antibody and the CDK2 interaction was verified by Western analysis (left panel). MCF7 cells were treated with estradiol (E2) for various periods of time, and the PELP1 interaction with CDK2 was analyzed by using immunoprecipitation (middle panel). In vitro kinase assays using CDK2/CycE complex and CDK2/CycA2 complex (right panel) using full-length PELP1 as a substrate is depicted.

Figure 2

Identification of phosphorylation sites and generation of phosphospecific antibody. A, Bacterially expressed GST-PELP1 deletions were used as substrates for an in vitro kinase assay using CDK4/cycD1 (left panel), CDK2/cycE (middle panel) and CDK2/cycA2 (right panel), and PELP1 domains phosphorylated by each kinase complex was identified by using autoradiography. B, Identification of CDK phosphorylation sites using site-directed mutagenesis (Ser to Ala). Various single mutants in the region of interest were used along with respective wild-type PELP1 deletion fragments. Loss of ³²P incorporation revealed the successful identification of phosphorylation sites. C, Western blot analysis of native, wild-type-tagged PELP1 and tagged phosphomutant PELP1 with the S991 phospho-PELP1 antibody in the presence or absence of the phospho peptide. D, MCF7 cells were either arrested and released by estradiol (E2) stimulation (left panel) or synchronized into the G1-S boundary by double thymidine block were released into cell cycle by addition of thymidine-free media (right panel), and phosphorylation status of PELP1 was analyzed by using the S991 phospho-antibody.

Figure 3

CDKs-mediated phosphorylation is required for optimal PELP1 function in cell cycle. A, Western analysis of ZR75 cells stably expressing GFP-Vec, GFP-PELP1-WT and GFP-PELP1-MT. B, Cell proliferation capacity of parental ZR, PELP1-WT and PELP1-MT stable cells were analyzed after treating the cells with 10% serum (left panel) and with or without E2 (right panel) using Cell Titer-Glo assay. C, Flow cytometric analysis of cell cycle phases in ZR-GFP, wild-type ZR-PELP-WT and mutant PELP1-MT pool clones was done and changes observed in S-phase cells are depicted by percentages. D, ZR75 cells that stably express PELP1 shRNA were transfected with shRNA-resistant PELP1-WT or -MT expression vectors by using nucleofector protocol. After 48 h, transfected cells were subjected to cell proliferation with or without estradiol (E2) and assayed for proliferation using Cell Titer-Glo assay. *, P<0.05, **, P<0.001.

Figure 4

CDK-dependent PELP1 phosphorylation modulates E2F transactivation functions. A, RNA isolated from MCF7-control and stable MCF7-PELP1 shRNA expressing cells were hybridized to the Human Cell Cycle microarray. Changes in the gene expression were analyzed using SABioscience software with actin as a control for normalization. Representative genes downregulated upon PELP1 depletion are shown. Expression of PELP1 in PELP1 shRNA clones is shown as an insert. B, MCF7 and MCF7-PELP1 shRNA cells were transfected with E2F-luciferase reporter along with E2F and DP1 plasmids, and luciferase activity was measured after 36 h (left panel). NIH3T3 cells were transfected with E2F-luc with E2F and DP1 plasmids and with or without PELP1-WT or PELP1-MT. Luciferase activity was measured after 36 h of transfection (right panel). C, Recruitment of PELP1 to E2F target genes promoter was analyzed by using the ChIP assay with MCF7 and ZR75 cells. D, MCF7 cells were transiently transfected with control or PELP1-specific siRNA (left panel). ZR75 cells were stably transfected with control or PELP1-specific shRNA (right panel). Total RNA was isolated and expression of classical E2F target genes was analyzed by real-time qPCR.

Figure 5

CDK phosphorylation is essential for PELP1 oncogenic function. A, Nude mice were implanted with E2 pellet were injected subcutaneously with ZR75-GFP or ZR75-PELP1-WT or ZR75-PELP1-MT cells and tumor growth was measured at weekly intervals. Tumor volume is shown in the graph (left panel). B, The average tumor weight is shown in the graph. Representative images of tumors are shown. Status of PELP1 phosphorylation (C) and PCNA expression as a marker of proliferation (D) was analyzed by immunohistochemistry (IHC). *, P<0.05; **, P<0.001.