Structural Basis for Substrate Selectivity of the E3 Ligase COP1 Sacha Uljon, Xiang Xu, Izabela Durzynska, Sarah Stein, Guillaume Adelmant, Jarrod A. Marto, Warren S. Pear, Stephen C. Blacklow  Structure  Volume 24, Issue 5, Pages 687-696 (May 2016) DOI: 10.1016/j.str.2016.03.002 Copyright © 2016 Elsevier Ltd Terms and Conditions

Structure 2016 24, 687-696DOI: (10.1016/j.str.2016.03.002) Copyright © 2016 Elsevier Ltd Terms and Conditions

Figure 1 The COP1 WD40 Domain Is Necessary and Sufficient for Trib1 Binding (A) Schematic illustrating forms of human COP1 tested in immunoprecipitation (IP) experiments. (B) Trib1 immunoprecipitations. Flag-tagged full-length or C-terminally truncated Trib1 proteins were recovered using anti-Flag antibodies. The whole-cell extracts (WCE) and the immunoprecipitates were probed with anti-Flag (top) and anti-COP1 antibodies (bottom). NS indicates a non-specific band seen in all immunoprecipitates probed with anti-COP1. Structure 2016 24, 687-696DOI: (10.1016/j.str.2016.03.002) Copyright © 2016 Elsevier Ltd Terms and Conditions

Figure 2 Structures of Human and Arabidopsis COP1 WD40 Domains (A) Cartoon representation. Top and side views are shown for unbound human COP1 (left), the human COP1-peptide complex (middle) and the plant COP1-peptide complex (right). The Trib peptide is shown in magenta as sticks. Numbers denote blades from the N to C terminus (colored on a rainbow scale from blue to red), and letters A–D indicate strands of a typical blade. A dashed arrow in the top view of human COP1 points to the helical insert between strands 1C and 1D. A second arrow in the top view of the plant COP1 indicates the 6D to 7A loop. (B) Surface representation, colored according to sequence conservation on a sliding scale from maroon (highest) to cyan (lowest) using the program Consurf (Ashkenazy et al., 2010). (C and D) Close-up views of the Arabidopsis COP1 (C) and human COP1 (D) complexes. The protein backbones are shown as cartoons and the surface is in white, with key interface side chains labeled and rendered as sticks. The Trib peptide is shown in yellow and residue side chains are represented as sticks. Structure 2016 24, 687-696DOI: (10.1016/j.str.2016.03.002) Copyright © 2016 Elsevier Ltd Terms and Conditions

Figure 3 Comparison of WD40 β Propellers from E3 Ligases Bound to Peptide Motifs (A) Cartoon representation, showing a top view (clockwise from N to C terminus) of seven-bladed WD40 β propellers from COP1, β-TRCP1 (1P22), and CDC20 (4GGA), and eight-bladed propellers from CDC4 (1NEX) and FBW7 (2OVQ). (B) Surface representation of these propeller domains bound to cognate peptides represented as space-filling spheres. The propeller domains are colored according to electrostatic surface potential, and the bound substrates are shown in CPK colors (carbon, green). Structure 2016 24, 687-696DOI: (10.1016/j.str.2016.03.002) Copyright © 2016 Elsevier Ltd Terms and Conditions

Figure 4 Scanning Mutagenesis and Importance of Residue Position (A) Fluorescence polarization assay measuring the binding affinity of the human COP1 β propeller (376–731) for the Trib1 consensus motif. The change in fluorescence polarization is plotted as a function of COP1 protein concentration. (B) Competition assay comparing binding of a Trib1 protein that includes the kinase-homology domain and the COP1-binding motif with the isolated C-terminal peptide. (C) Alanine scanning mutagenesis. Displacement of the consensus Trib1 wild-type peptide FITC-SEIGTSDQIVPEYQEDSDI was monitored using a fluorescence polarization assay, and the change in polarization plotted for the indicated peptides as a function of unlabeled peptide concentration. (D) Competition assay using arginine and isoleucine mutagenesis, using the same fluorescence polarization as in (A). (E) Plot of ΔΔG (in kcal/mol) for the alanine scan using the formula ΔΔG = −RTln(i) where i is the ratio of the competing peptide IC50 value to that of the wild-type peptide under the same conditions. Graphs represent ΔΔG derived from triplicate measurements of IC50 values with error bars showing ±SEM calculated using standard propagation of error calculations. Structure 2016 24, 687-696DOI: (10.1016/j.str.2016.03.002) Copyright © 2016 Elsevier Ltd Terms and Conditions

Figure 5 Trib1 Peptide Binds Human COP1 and Competes with COP1 Substrates (A) Peptide sequences used in a competition experiment to compare binding affinities of reported COP1 substrates with binding of the Trib1 peptide. Red text indicates the COP1 consensus binding residues. IC50 values were calculated from the data in (B). Phosphorylated serine residues are shown in blue in the ETS1P peptide. (B) Competition experiment in which Trib2, ETS, and Jun family peptides displace FITC-Trib1 at a fixed human COP1 concentration. Structure 2016 24, 687-696DOI: (10.1016/j.str.2016.03.002) Copyright © 2016 Elsevier Ltd Terms and Conditions

Figure 6 Model for Trib-COP1 Assembly (Top left) The Trib1 COP1-binding motif is restrained by interaction with the N lobe in autoinhibitory interactions. Release of the Trib1 tail from the N lobe allows it to bind to the COP1 propeller. C/EBPα is then recruited by the assembled complex, and C/EBPα is subsequently ubiquitinated and undergoes proteasome-dependent degradation. The Trib1 models were generated using coordinates with PDB: 5CEK and 5CEM. The green dashed line represents an unstructured region of Trib1 that connects the kinase-homology domain to the COP1-binding motif. N-term, N terminus. Structure 2016 24, 687-696DOI: (10.1016/j.str.2016.03.002) Copyright © 2016 Elsevier Ltd Terms and Conditions