Volume 18, Issue 2, Pages 213-224 (April 2005) Selective Ablation of Retinoblastoma Protein Function by the RET Finger Protein  Maja Krützfeldt, Mark Ellis, Daniel B. Weekes, Jonathan J. Bull, Martin Eilers, Maria d M. Vivanco, William R. Sellers, Sibylle Mittnacht  Molecular Cell  Volume 18, Issue 2, Pages 213-224 (April 2005) DOI: 10.1016/j.molcel.2005.03.009 Copyright © 2005 Elsevier Inc. Terms and Conditions

Figure 1 Selective Ablation of Rb Activity by RFP pMMTV-GRE-, TETo-E2F-, or pCYLA-driven luciferase reporters were transfected into C33A cells alongside pRSVgal and expression plasmids as indicated. (A) RFP inhibits GR-dependent MMTV promoter transactivation. Activity in the presence of GR and Dexamethasone (Dex) was set to 1. (B) RFP does not affect activation of GR by Dex. Activation by GR in the absence of Dex was set to 1. (C) RFP specifically inhibits augmentation of transcription by Rb. Activity in the presence of GR and Dex was set to 1. (D and E) RFP does not impair Rb-mediated inhibition of E2F-driven transcription. Activity in the presence of TETR (D) or E2F-1/DP1 (E) was set to 1. Error bars represent the variance within a minimum of three independent assays. Molecular Cell 2005 18, 213-224DOI: (10.1016/j.molcel.2005.03.009) Copyright © 2005 Elsevier Inc. Terms and Conditions

Figure 2 Interaction of Rb and RFP In Vitro and In Vivo (A) Functional domains of RFP and isolates from the two-hybrid screen. (B) Interaction between RFP and Rb in yeast is LXCXE independent and facilitated by Rb-LP. Prey plasmid for full-length RFP was coexpressed with bait plasmids as indicated. Prey plasmid for SV40 T-antigen is used as a control. (C) RFP binds Rb-LP in an LXCXE-independent way in vitro. GST pull-down using GST constructs as indicated and in vitro-translated 35S-methionine-labeled full-length RFP. 5% of the RFP input was loaded in lane 1. (D) Schematic overview of Rb fragments required for RFP binding. (E) Association of recombinant RFP and Rb in mammalian cells. C33A cell lysates were subjected to IMAC. Bound proteins (top and middle panels) or 10% of input lysate (bottom panel) were probed by Western blotting as indicated. The asterisk denotes a cross-reactive contamination with α-HIS antibody. (F and G) Cell-endogenous Rb and RFP interact. Immunprecipitations using α-Rb monoclonal (F) or α-RFP serum (G), lysate from Rb-positive MD-MBA, SW480 and H1299 or Rb-negative C33A cells, and Western blotting using α-RFP serum (F) or α-Rb monoclonal antibody (G) are shown. Buffer + AB and buffer − AB denote immunoprecipitation in the absence of cell lysate and the presence or absence of antibody. Input denotes 1% of input lysate. Molecular Cell 2005 18, 213-224DOI: (10.1016/j.molcel.2005.03.009) Copyright © 2005 Elsevier Inc. Terms and Conditions

Figure 3 Effect of RFP on Rb-Mediated Cellular Responses (A) RFP modulates Rb-induced cell shape response. Cell shape response of U2OS cells with tetracycline-regulated expression of active Rb following infection with RFP encoding adenovirus (RFP) or a vector control (V). The area occupied by individual cells was determined using the SimplePCI software package. Events are ordered by size. All cells present in ten eye fields were scored. Measurements were taken 5 days after Rb inductions. (B) RFP does not influence Rb-associated cell cycle response. Cell cycle response in U2OS. Cells were exposed to nocodazole for 14 hr prior to analysis. (C) RFP perturbs upregulation of moesin expression in Rb-expressing cells. Cells treated as in (A). Cell lysate was processed for immunoblotting by using antibodies as indicated. Molecular Cell 2005 18, 213-224DOI: (10.1016/j.molcel.2005.03.009) Copyright © 2005 Elsevier Inc. Terms and Conditions

Figure 4 Structure Activity Relation for Inhibition of Rb-Mediated GR Augmentation by RFP (A) RFP domains required for Rb binding. GST-Rb pull-down using 35S-methionine-labeled RFP fragments. Five percent of the RFP input was run alongside. (B) Inhibition of Rb-mediated augmentation of MMTV promoter activity. Expression of RFP constructs was monitored in parallel. Error bars represent variance between three independent samples Molecular Cell 2005 18, 213-224DOI: (10.1016/j.molcel.2005.03.009) Copyright © 2005 Elsevier Inc. Terms and Conditions

Figure 5 Functional Interaction between RFP and EID-1 (A) Association of RFP and EID-1 in mammalian cells. C33A cells transfected with expression plasmids, as indicated, were subjected to IMAC. Bound proteins (top panels) and input lysate (one tenth, bottom panels) were probed with antibodies as indicated. (B) Association of transfected EID-1 with cell endogenous RFP. C33A cells were transfected as indicated. Immunoprecipitations were performed with α-RFP serum and analyzed by Western blotting using α-T7 antibody. (C) Association of cell endogenous EID-1 and RFP immunoprecipitations were performed using α-RFP serum. Immunoprecipitations were analyzed by α-EID-1 Western blotting alongside 1% of input material. Buffer + AB and buffer − AB denote controls as in Figure 2G. (D) RFP associates with transfected MMTV promoter in a GR-dependent way. ChIP from C33A cells transiently transfected with MMTV promoter and plasmids as indicated. ChIPs were performed in the presence of α-xpr antibody (top and bottom: lanes 2 and 4), absence of antibody (top: lanes 1 and 3), or nonimmune mouse IgG (bottom: lanes 1 and 3). Bound DNA was probed by PCR for MMTV promoter sequences. 1%, 0.3%, and 0.1% of DNA input was analyzed alongside (lanes 5–10). Molecular Cell 2005 18, 213-224DOI: (10.1016/j.molcel.2005.03.009) Copyright © 2005 Elsevier Inc. Terms and Conditions

Figure 6 Stabilization of EID-1 by RFP (A) Coexpression of RFP increases the half-life of recombinant EID-1. CX-chase of T7-EID-1 and HIS/xpr-RFP-expressing U20S cells. Lysates were subjected to α-T7 Western blotting. Loading of samples and exposure were adjusted to achieve similar signal intensity at 0 min. (B) Densitometry-based signal quantification derived from three independent experiments. Experimental details as in (A). Error bars represent variance within four independent experiments. (C and C′) Effect of RFP subdomains on the half-life of EID-1. EID-1 stability in U2OS cells expressing RFP fragments that inhibit Rb-mediated augmentation of GR-dependent transcription. Experimental details described in (A). Fragment behavior was consistent in three independent experiments. Samples from 0 min were probed for expression of RFP by using α-6HIS antibody (C′). (D and D’) Stabilization of endogenous EID-1 following adenovirus-mediated transduction of RFP. U20S cells were infected with RFP-encoding adenovirus (Ad-RFP) or empty vector (Ad-vector). Lysates were probed with α-EID-1 monoclonal antibody. Other experimental details described in (C). (E) siRNA-mediated ablation of RFP results in loss of endogenous EID-1. U2OS cells were transfected with RFP siRNA oligonucleotide (si-RFP1) or a scrambled control (si-scr). Lysates were analyzed 48 hr later by Western blot using antibodies as indicated, followed by Amidoblack staining of the membrane. (F) Mapping of RFP domains required for EID-1 binding. C33A cells were transfected with expression plasmids for T7-EID-1 and HIS/xpr-RFP subdomain fragments. Cell lysate amounts equalized for T7-EID-1 expression prior to IMAC. Bound proteins and 1% of input probed as indicated. (G) Summary of results. Molecular Cell 2005 18, 213-224DOI: (10.1016/j.molcel.2005.03.009) Copyright © 2005 Elsevier Inc. Terms and Conditions

Figure 7 Mechanism of RFP Action RFP uncouples Rb-mediated gene activation, gene repression, and associated cell responses, by preventing EID-1 degradation at the site of transcription. Molecular Cell 2005 18, 213-224DOI: (10.1016/j.molcel.2005.03.009) Copyright © 2005 Elsevier Inc. Terms and Conditions